A3GALT2 encodes a predicted GT6-family single-pass type II membrane protein homologous to mammalian isoglobotriaosylceramide synthase. Human catalytic activity remains unresolved: rat-human catalytic-domain chimeras lacked detectable activity, whereas a reconstructed human coding sequence produced only very weak iGb3 and iGb4 after overexpression in CHO cells. Endogenous full-length transcript and protein expression, subcellular localization, and physiological glycosphingolipid flux remain unestablished.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0016757 glycosyltransferase activity | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: A reconstructed human A3GALT2 coding sequence produces weak iGb3/iGb4 in CHO cells, supporting secondary glycosyltransferase and biosynthetic capacity. Reason: Full Methods and Figure 6/Table 1 of PMID:23378701 explicitly test a reconstructed human ORF in sorted CHO cells against mock-transfected controls. The focused report incorrectly dismissed this result from the pig-focused title/abstract. The 2008 human-domain chimera and rat Y252N experiments remain genuine contrary evidence, but do not erase a later weak positive in a different construct/readout. Independent sequence mapping confirms human U3KPV4 N253 corresponds to rat A0A4Z3 Y252; the report mislabeled mouse Q3V1N9 as rat. Retain the broad activity/process as non-core without asserting endogenous human flux. The live PTHR10462 tree places human leaf PTN002470899 beneath PTN000049004 and PTN000049131, with no negative PAINT row on that path. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN000049004 Β· PTN000049004 SUPPORTS TRANSFER Actual target descent verified and saved in A3GALT2-paint-lineage.json. Weak human-construct activity supports retained capacity; no inference from donor count or absent endogenous assays. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0005794 Golgi apparatus | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Retain the inferred secretory-pathway localization while distinguishing it from direct localization of endogenous human protein. Reason: Golgi and vesicle localization descend from PTN000049004 in PTHR10462, and the predicted type-II membrane architecture is compatible with these compartments. The negative catalytic chimera experiments do not demonstrate loss of trafficking or membrane insertion. No target-specific localization contradiction was established; missing endogenous-human imaging alone does not invalidate the curated phylogenetic inference. The more precise Golgi-cisterna IEA remains an inference. The live reference tree independently confirms target leaf PTN002470899 descends from PTN000049004; the focused report supplies no localization-loss evidence. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN000049004 Β· PTN000049004 SUPPORTS TRANSFER Confirmed target path in PTHR10462; negative catalytic assays do not establish loss of Golgi/vesicle localization. |
| GO:0006688 glycosphingolipid biosynthetic process | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: A reconstructed human A3GALT2 coding sequence produces weak iGb3/iGb4 in CHO cells, supporting secondary glycosyltransferase and biosynthetic capacity. Reason: Full Methods and Figure 6/Table 1 of PMID:23378701 explicitly test a reconstructed human ORF in sorted CHO cells against mock-transfected controls. The focused report incorrectly dismissed this result from the pig-focused title/abstract. The 2008 human-domain chimera and rat Y252N experiments remain genuine contrary evidence, but do not erase a later weak positive in a different construct/readout. Independent sequence mapping confirms human U3KPV4 N253 corresponds to rat A0A4Z3 Y252; the report mislabeled mouse Q3V1N9 as rat. Retain the broad activity/process as non-core without asserting endogenous human flux. The live PTHR10462 tree places human leaf PTN002470899 beneath PTN000049004 and PTN000049131, with no negative PAINT row on that path. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN000049131 Β· PTN000049131 SUPPORTS TRANSFER Actual target descent verified and saved in A3GALT2-paint-lineage.json. Weak human-construct activity supports retained capacity; no inference from donor count or absent endogenous assays. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0031982 vesicle | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Retain the inferred secretory-pathway localization while distinguishing it from direct localization of endogenous human protein. Reason: Golgi and vesicle localization descend from PTN000049004 in PTHR10462, and the predicted type-II membrane architecture is compatible with these compartments. The negative catalytic chimera experiments do not demonstrate loss of trafficking or membrane insertion. No target-specific localization contradiction was established; missing endogenous-human imaging alone does not invalidate the curated phylogenetic inference. The more precise Golgi-cisterna IEA remains an inference. The live reference tree independently confirms target leaf PTN002470899 descends from PTN000049004; the focused report supplies no localization-loss evidence. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN000049004 Β· PTN000049004 SUPPORTS TRANSFER Confirmed target path in PTHR10462; negative catalytic assays do not establish loss of Golgi/vesicle localization. |
| GO:0005975 carbohydrate metabolic process | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: A reconstructed human A3GALT2 coding sequence produces weak iGb3/iGb4 in CHO cells, supporting secondary glycosyltransferase and biosynthetic capacity. Reason: Full Methods and Figure 6/Table 1 of PMID:23378701 explicitly test a reconstructed human ORF in sorted CHO cells against mock-transfected controls. The focused report incorrectly dismissed this result from the pig-focused title/abstract. The 2008 human-domain chimera and rat Y252N experiments remain genuine contrary evidence, but do not erase a later weak positive in a different construct/readout. Independent sequence mapping confirms human U3KPV4 N253 corresponds to rat A0A4Z3 Y252; the report mislabeled mouse Q3V1N9 as rat. Retain the broad activity/process as non-core without asserting endogenous human flux. The live PTHR10462 tree places human leaf PTN002470899 beneath PTN000049004 and PTN000049131, with no negative PAINT row on that path. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0016020 membrane | IEA GO_REF:0000002 | ACCEPT | Summary: A3GALT2 retains the predicted single-pass type-II membrane architecture of the GT6 family. Reason: The UniProt sequence annotation supports an N-terminal membrane anchor. Uncertainty about endogenous catalytic flux does not refute this structural localization property; membrane is an accurate broad compartment. |
| GO:0016758 hexosyltransferase activity | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: A reconstructed human A3GALT2 coding sequence produces weak iGb3/iGb4 in CHO cells, supporting secondary glycosyltransferase and biosynthetic capacity. Reason: Full Methods and Figure 6/Table 1 of PMID:23378701 explicitly test a reconstructed human ORF in sorted CHO cells against mock-transfected controls. The focused report incorrectly dismissed this result from the pig-focused title/abstract. The 2008 human-domain chimera and rat Y252N experiments remain genuine contrary evidence, but do not erase a later weak positive in a different construct/readout. Independent sequence mapping confirms human U3KPV4 N253 corresponds to rat A0A4Z3 Y252; the report mislabeled mouse Q3V1N9 as rat. Retain the broad activity/process as non-core without asserting endogenous human flux. The live PTHR10462 tree places human leaf PTN002470899 beneath PTN000049004 and PTN000049131, with no negative PAINT row on that path. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0032580 Golgi cisterna membrane | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Retain the inferred secretory-pathway localization while distinguishing it from direct localization of endogenous human protein. Reason: Golgi and vesicle localization descend from PTN000049004 in PTHR10462, and the predicted type-II membrane architecture is compatible with these compartments. The negative catalytic chimera experiments do not demonstrate loss of trafficking or membrane insertion. No target-specific localization contradiction was established; missing endogenous-human imaging alone does not invalidate the curated phylogenetic inference. The more precise Golgi-cisterna IEA remains an inference. |
| GO:0047276 N-acetyllactosaminide 3-alpha-galactosyltransferase activity | IEA GO_REF:0000120 | UNDECIDED | Summary: The precise LacNAc reaction remains unresolved independently of weak human iGb3-synthase capacity. Reason: GO:0047276 requires Gal-beta1,4-GlcNAc, whereas the verified human CHO experiment in PMID:23378701 measures products derived from lactosylceramide, Gal-beta1,4-Glc-ceramide. Positive activity on that acceptor neither proves nor excludes LacNAc use. The ISS source and corresponding UniProt RHEA:13013 statement trace specifically to rat A0A4Z3 and PMID:10854427; the accessible source abstract establishes preferential glycolipid use but does not resolve the exact LacNAc assay. The report neither inspected that assay nor reconciled the real human positive result. Retain UNDECIDED, without claiming GGTA1 mispropagation or treating substrate preference as an exclusion test. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0047276 N-acetyllactosaminide 3-alpha-galactosyltransferase activity | ISS GO_REF:0000024 | UNDECIDED | Summary: The precise LacNAc reaction remains unresolved independently of weak human iGb3-synthase capacity. Reason: GO:0047276 requires Gal-beta1,4-GlcNAc, whereas the verified human CHO experiment in PMID:23378701 measures products derived from lactosylceramide, Gal-beta1,4-Glc-ceramide. Positive activity on that acceptor neither proves nor excludes LacNAc use. The ISS source and corresponding UniProt RHEA:13013 statement trace specifically to rat A0A4Z3 and PMID:10854427; the accessible source abstract establishes preferential glycolipid use but does not resolve the exact LacNAc assay. The report neither inspected that assay nor reconciled the real human positive result. Retain UNDECIDED, without claiming GGTA1 mispropagation or treating substrate preference as an exclusion test. Supporting Evidence: PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
| GO:0001962 alpha-1,3-galactosyltransferase activity | IDA NOT PMID:18630988 Humans lack iGb3 due to the absence of functional iGb3-synth... | UNDECIDED | Summary: The original NOT remains represented, with conflicting construct-specific negative and weak positive assays unresolved. Reason: PMID:18630988 Figures 3-4 show negative human-domain chimeras and loss after rat Y252N substitution. Figure 5 already tested reverse single/combined substitutions without rescue, contrary to the report's proposed missing resurrection experiment. Human sequence-version-1 N253 aligns to rat Y252; mouse Q3V1N9 is a different accession/species. Full PMID:23378701 Methods and Figure 6 demonstrate weak human-ORF iGb3/iGb4 production in CHO cells, despite its pig-organ title. Preserve the source negation while declining to generalize either construct-specific result to all human products or robust endogenous flux. Supporting Evidence: PMID:18630988 The single isolated substitution of rat Y252N resulted in the complete elimination of GalΞ±(1,3)Gal staining PMID:23378701 However, the human iGb3 synthase showed very weak activity even in the CHO cell transfection experiment and could not further elongate iGb3 to B4, B5, and B6 GSLs |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Is an endogenous full-length A3GALT2 transcript and protein expressed in any human cell type, tissue, or developmental state?
Q: Does endogenous human A3GALT2 causally generate iGb3 or iGb4, rather than merely retaining very weak catalytic capacity when an assembled coding sequence is overexpressed?
Q: If endogenous A3GALT2 protein exists, does it localize to Golgi cisternae or vesicles with the predicted type-II membrane topology, and what native biochemical activity and substrate specificity does it have?
Experiment: Survey fetal and adult thymic populations, monocyte-derived dendritic cells, and other candidate tissues using long-read RNA sequencing or RACE across every exon junction, followed by targeted proteomics and an independently validated A3GALT2-specific antibody to establish endogenous full-length transcript and protein.
Hypothesis: A full-length A3GALT2 transcript and protein are expressed only in a restricted human cell type or developmental window.
Type: expression analysis
Experiment: In a human cell type with verified endogenous A3GALT2 protein, combine CRISPR knockout with wild-type and catalytic-site-mutant rescue, stable-isotope UDP-galactose tracing, and isomer-resolved quantitative LC-MS/MS of iGb3, iGb4, and precursor lipids to test causal glycosphingolipid flux.
Hypothesis: Endogenous A3GALT2 is responsible for low-level human iGb3 and iGb4 biosynthesis.
Type: genome editing and lipidomics
Experiment: Introduce a minimally disruptive tag at the endogenous locus for super-resolution microscopy, organelle fractionation, and membrane-topology mapping; in parallel, compare verified full-length human, rat, and catalytic-site-mutant proteins in the same reconstituted membrane enzymology assay with lactosylceramide and N-acetyllactosaminide acceptors.
Hypothesis: Native human A3GALT2 is a Golgi type-II membrane protein with weak lactosylceramide alpha-1,3-galactosyltransferase activity.
Type: endogenous localization and comparative enzymology
What is not known β curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: It is unknown whether an intact endogenous full-length A3GALT2 transcript and protein are expressed in any human cell type or developmental state.
OPEN BIOLOGY WHOLLY_DARK
What is known: The 2008 study failed to detect correctly spliced RNA in the adult tissues and dendritic cells tested, whereas the 2013 study recovered separate pediatric-thymus cDNA fragments and supplied exon 1 and part of exon 2 by primer; UniProt remains at transcript-level protein evidence.
Significance: Without endogenous full-length expression, neither catalytic nor localization claims can be assigned physiological significance.
What would resolve it: Demonstrate every exon junction in a native transcript and detect an A3GALT2-specific endogenous protein in the same human cell or tissue.
Provenance (the field's own admissions):
Gap: Whether endogenous human A3GALT2 catalyzes any measurable reaction or contributes causally to physiological iGb3 or iGb4 flux is unresolved.
OPEN BIOLOGY MF_DARK
What is known: A reconstructed human coding sequence produced MS-detectable iGb3 and iGb4 after CHO-cell overexpression, but the authors characterized the activity as very weak; an earlier human-catalytic-domain chimera study was negative.
Significance: This determines whether A3GALT2 is an active human glycosyltransferase or a retained, weakly competent sequence without a physiological catalytic role.
What would resolve it: In a human cell with verified endogenous protein, combine A3GALT2 knockout and wild-type rescue with isomer-resolved quantitative lipidomics and isotope tracing; independently reproduce activity with verified full-length protein and matched acceptor panels.
Provenance (the field's own admissions):
Gap: The native subcellular location and membrane topology of endogenous human A3GALT2 protein are unknown.
OPEN BIOLOGYCURATION CC_DARK
What is known: The sequence predicts an N-terminal type-II membrane signal anchor, but Golgi-like localization was observed only for chimeras whose targeting segments were rat-derived; UniProt transfers the Golgi statement by similarity.
Significance: Compartment and topology determine which lipid or glycoconjugate acceptors a putative lumenal glycosyltransferase can encounter.
What would resolve it: Endogenously tag verified A3GALT2 protein and combine organelle-resolved imaging with protease protection and glycosylation-accessibility assays.
Provenance (the field's own admissions):
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)