ganglioside GM3 + UDP-GalNAc -> ganglioside GM2 + UDP + H(+) (EC 2.4.1.92, RHEA:12588) [file:human/B4GALNT1/B4GALNT1-uniprot.txt "a ganglioside GM3 (d18:1(4E)) + UDP-N-acetyl-alpha-D-"].This session supersedes the earlier annotation-action summary above. All 21 distinct GOA assertions were checked; the repeated raw Golgi-membrane ISS tuple represents one distinct annotation. The proposed decisions are 16 ACCEPT and 5 MODIFY, with no NEW assertions. Broad electronic catalytic terms are refined to the already supported GO:0003947; broad metabolic terms to GO:0001574; and the generic membrane term to GO:0000139. A general term is not an over-annotation simply because a more specific term is available. The valid broad IBA catalytic assertion remains core. Its PAINT node placement has not been independently reconstructed, and the target's occurrence among IBA sources is valid provenance.
The normal GOA WITH/FROM fields were restored to supporting_entities for ten assertions. In particular, Q10468 is the rat ortholog supporting the ISS locations; Q09200 is mouse and supports the automatic orthology transfer. The three UniProt alternative products remain unchanged. The correct schema field is sequence_note, singular, and both VSP identifiers for Q00973-3 were already complete.
The six pre-existing PMID caches contain abstracts rather than full experimental Methods. Their identifiers and titles were checked against PubMed. PMID:7487055 distinguishes mouse B16 and human MeWo recipient cells; those recipient names do not establish the species of the cloned cDNA. PMID:7890749 supports efficient GM2/GD2 synthesis but also context-dependent neutral asialo-GM2 production, so low activity on an alternate acceptor in vitro does not establish biological irrelevance. GA2 itself is a neutral glycosphingolipid. The catalytic description now distinguishes GM3/GD3-derived gangliosides from the LacCer-to-GA2 branch. PMID:7487055(https://pubmed.ncbi.nlm.nih.gov/7487055/), PMID:7890749(https://pubmed.ncbi.nlm.nih.gov/7890749/)
PubMed links the original cloning paper PMID:1601877 to the published erratum PMID:8120069. The normal UniProt record explicitly associates the latter with sequence revision of residues 412–533. This explains why the original abstract's open-reading-frame length should not be equated with the current 533-residue sequence. The correction text has not been read, and no replacement PMID, retraction, or wrong-gene conclusion is inferred. Original cloning paper, erratum metadata
The ordinary Falcon research command, with its configured perplexity-lite fallback, exited unsuccessfully and produced no provider report. This reassessment therefore uses explicitly manual literature reading and an independent annotation audit. The raw UniProt, GOA, and publication caches are preserved.
The recovered normal PMID:40593514 cache agrees with the previously read primary article. The human ΔN27/ΔN46 lumenal constructs were expressed in HEK293-F cells. Crystallography supports a homodimer and three intrachain disulfide bonds, challenging the 2000 interchain assignment. The newer authors explicitly leave the discrepancy with nonreducing SDS-PAGE unresolved. The older reference therefore receives a DISPUTED finding for bond topology, while homodimerization remains supported. Selected Results, Discussion and figure legends, and the relevant expression, purification, crystallography, activity and molecular-dynamics Methods were read; figure pixels, coordinates, supplements and raw data were not independently analyzed. 2025 structure, 2000 disulfide study
The new study's His-tag-tethered liposome assays support more activity on GM3/GD3 than LacCer under the tested conditions. These are endpoint comparisons, not a complete Michaelis–Menten treatment. Polar changes in the F93/W491 loops impaired processing of lipid substrates, and hydrophobic substitutions restored or increased activity. Membrane insertion comes from simulations supplemented by mutagenesis, not direct imaging of the native Golgi protein. R300C folding and D433A dimer defects are structure-based predictions; the direct active-site panel tested R288H, R505H and D358A with R66A as a control. The questions now distinguish these predictions from measured cellular phenotypes. The proposed extracellular remodeling mechanism is not promoted to a new function. Primary article
The recovered PMID:8120069 record confirms only the published-erratum citation and its link to the 1992 paper. It contains neither a scientific abstract nor the correction body. The sequence-revision statement continues to be attributed to UniProt. No original GOA reference or raw cache was rewritten. The final review retains all 21 distinct source assertions and all three alternative products, with 16 ACCEPT, 5 MODIFY, one catalytic core and no NEW annotation.
The review now states why the broad PAINT catalytic assertion is retained while the gene-level ARBA assertion is refined: an experimentally supported human substrate preference does not, by itself, revise an ancestral function across a clade. The WITH/FROM list contains Q8NHY0, identified as B4GALNT2 in the cached PANTHER entry table; this is compatible with a shared GalNAc-transferase class, but the exact PAINT node has not been reconstructed. No claim of a reconstructed phylogeny is made.
The Golgi-apparatus ISS row is now MODIFY to the already present Golgi-membrane term, supported by the same rat donor and corroborating human topology. This treats the two general location terms consistently without declaring a core location non-core. The resulting tally is 15 ACCEPT and 6 MODIFY across the same 21 distinct source assertions, with one catalytic core, three unchanged alternative products and no NEW annotation.
The biochemical excerpts now name the GM2/GD2 products and the GM3/GD3 acceptors, and the cloning MF row quotes the GalNAc transfer and beta-1,4 linkage. The core cites that cloning paper without duplicating its excerpt and uses the 2025 paper for the human homodimer. A complete short clause supports the specifically disputed older disulfide assignment. Quotations remain exact cached text; duplication is limited rather than cutting a clause before its products. The erratum remains identifier-verified but has LOW functional relevance because its correction body is unavailable. EC 2.4.1.92 and CAZy GT12 are restored from the raw UniProt record, and the description explicitly identifies GM2/GD2 as gangliosides. Raw sources and prior journal entries are unchanged.