FUT8

Summary: Retain the Golgi membrane location of the human catalyst.

Reason: The actual ordinary Reactome reaction identifies human FUT8 residues 1–575 in Golgi membrane through its catalyst/physical-entity chain. Selected original PMID:14514715 Results and Figure 6 independently show a C-terminally tagged human enzyme retaining activity and colocalizing with giantin in COS7 cells. The tag-dependent ER retention of an N-terminal fusion is not generalized into normal ER function. These observations support the core Golgi membrane location without assigning every alternative product the same topology.

Summary: Retain the correctly identified Golgi catalyst location in the viral glycan model.

Reason: Reactome R-HSA-9696980 uses the same verified FUT8 physical entity and Golgi membrane compartment as the ordinary glycan reaction. The location is consistent with human construct localization and the type-II membrane enzyme architecture. Its retention does not validate the separate viral peptide-cleavage process propagation or claim a virus-specific location.

Summary: Retain prenatal developmental consequences as non-core context.

Reason: The selected original PMID:11698403 reports high fetal expression with an earlier citation, while PMID:14514715 adds developmental transcript distribution and recombinant-enzyme activity/localization; these observations alone are not a human embryogenesis perturbation assay. The later full main text of PMID:29304374 describes three unrelated individuals with biallelic FUT8 variants and prenatal growth restriction, with shortened fetal limbs and polyhydramnios in two pregnancies. Patient fibroblast and serum assays separately establish deficient FUT8 protein/core fucosylation. Together these findings support developmental consequences of FUT8 deficiency, so retain the original developmental annotation as non-core rather than marking it over-annotated. Keep the original NAS evidence code and PMID:11698403 unchanged: later clinical and biochemical evidence supports the gene-level judgment, without converting the original source into an experimental developmental assay or making embryogenesis a defining catalytic function.

Summary: Retain the positive ZBTB10 isoform 3 association as non-core.

Reason: Three source-specific IntAct records for PMID:32296183 identify human FUT8 and human ZBTB10 isoform 3 (Q96DT7-3) in positive yeast two-hybrid assays. The partner suffix belongs to ZBTB10 and is preserved; it does not specify a FUT8 natural product. The records retain the assay host, fusion tags and differing experimental-role fields. They support a physical interaction in the assay, without establishing an endogenous Golgi–nuclear mechanism, a FUT8 adaptor activity or SH3-domain binding. Exact primary Table 9 and FUT8 construct boundaries were not independently recovered. Retain this positive generic binding context under the campaign instruction.

Summary: Retain Golgi apparatus as an informative root-organelle location.

Reason: The Golgi is the established site of the membrane-bound FUT8 core-fucosylation reaction. Although the older NAS source is not a direct localization experiment, selected human FUT8 activity/localization experiments in PMID:14514715 and the verified Reactome entity provide independent support. Root-organelle coverage remains informative alongside Golgi membrane and cisterna terms under the project organelle policy. No natural ER role is inferred from an artificially N-terminally tagged construct.

Summary: Retain the inherited biology through the existing N-glycan-processing term.

Reason: This IBA is supported by the actual AN0/PTN000992822 ancestral placement above the exact human FUT8 leaf. The older four-donor GOA set and later five-donor IBD set are preserved separately, with all known donors below the node and no negative target-lineage record. FUT8 modifies an already N-linked glycan during maturation. The accepted GO:0006491 N-glycan processing is an official part_of child of GO:0006487 and captures this step more specifically. This specificity decision does not challenge the curator’s ancestral function placement.

Summary: The established processing step provides more specific N-linked-glycosylation coverage.

Reason: Human FUT8 structural/acceptor evidence in the PMID:17172260 abstract and independent catalytic work support participation in N-linked glycan maturation. Full assay detail from that structural paper was inaccessible, and no wrong-gene claim is made. The accepted GO:0006491 is an official part_of child of this term and identifies the conversion of an already attached N-glycan toward a mature structure, the step FUT8 performs. The original IDA/source assertion remains intact.

Summary: Use the retained processing step for this traceable N-glycosylation context.

Reason: The PMID:9133635 abstract specifically supports transfer of fucose to N-linked glycopeptides by human FUT8. The existing accepted GO:0006491 N-glycan processing describes that maturation reaction and is an official part_of child of GO:0006487. The broader process is sound but adds no separate coverage here; this is a granularity decision, not a rejection of the source or an assertion that FUT8 attaches the initial glycan directly to asparagine.

Summary: Retain N-glycan processing as the core biological process.

Reason: FUT8 catalyzes fucose transfer onto the innermost GlcNAc of an already N-linked glycan, converting its structure during maturation. This is exactly the processing of N-linked glycan structures by glycosyltransferases described by GO:0006491. Independent human purification, glycan-product assays and the actual Reactome catalyst support the ARBA-derived assertion. No protein-folding or trafficking activity is inferred merely because those processes appear in the term’s broader explanatory text.

Summary: Retain traceable N-glycan maturation by human FUT8.

Reason: The selected original PMID:11698403 Results distinguish human FUT8 addition to the proximal GlcNAc of N-linked glycans, forming core-fucosylated hybrid or complex structures. That supports the existing N-glycan-processing assertion and agrees with the direct human catalytic literature. The paper’s fly-genome emphasis does not erase its explicit human-enzyme comparison.

Summary: Retain directly measured human FUT8 glycan-transfer activity.

Reason: The original PMID:36280670 Results/Figure 4 and Methods test an engineered soluble human FUT8 construct produced in bacterial hosts and identify the fucosylated glycan product by MALDI-MS. This directly supports the alpha-1,6 core-fucose transfer chemistry. The fusion and truncation are experimental engineering, not a natural FUT8 isoform; the local substrate-number discrepancy between Results and Methods is retained in the source guide without changing the measured reaction.

Summary: Retain the specific catalytic function supported by human structural work.

Reason: The authentic PMID:17172260 abstract describes a human FUT8 crystal structure and N-linked GlcNAc acceptor recognition. Full Methods and individual mutant experiments were inaccessible, so this review does not claim to have independently repeated that paper’s IDA assignment. The specific catalytic function is nevertheless clearly established by independent human purification and recombinant glycan-product evidence; retaining the curator’s row is appropriate. Mouse phenotype discussion is not treated as a human assay.

Summary: Retain the specific N-glycan core-fucosyltransferase reaction.

Reason: The EC:2.4.1.68, RHEA:12985 and InterPro contexts agree with the defined transfer of fucose from GDP-L-fucose to the innermost N-glycan GlcNAc in an alpha-1,6 linkage. Human purification/cloning and engineered-human-enzyme glycan-product measurements independently support this chemistry. The reaction itself is appropriate even though the separate current ontology link from this MF to direct protein O-fucosylation is problematic. This gene-level activity is not assigned to all four natural products.

Summary: Retain the traceable human FUT8 catalytic statement.

Reason: The selected original Results section of PMID:11698403 explicitly distinguishes human FUT8 transfer to the proximal GlcNAc of N-linked glycans from the other fucosylation pathways. The paper is principally a Drosophila genome/pathway analysis, but it genuinely discusses the human enzyme; its title is not evidence of misattribution. The TAS row is consistent with direct human catalytic work reported separately in PMID:9133635.

Summary: Retain human enzyme purification and cloning support for core fucosylation.

Reason: The verified original abstract of PMID:9133635 describes purification from human MKN45 culture supernatant and cloning of a 575-residue enzyme that transfers fucose from GDP-fucose to N-linked glycopeptides. This supports the specific catalytic term and its TAS evidence scope. Full article access was unavailable; purification from supernatant is not elevated into a normal extracellular localization claim or equivalent activity of alternative products.

Summary: Retain the FUT8 catalytic activity in ordinary N-glycan maturation.

Reason: The actual Reactome R-HSA-1028788 model traces CatalystActivity 1028794 through physical entity R-HSA-1028782 to human Q9BYC5, residues 1–575, in Golgi membrane. The modeled reaction transfers GDP-fucose to N-glycan GlcNAc, agreeing with human biochemical evidence. FUT8 is the catalyst and the glycoprotein is the substrate; this is direct reaction participation, not an inference from pathway membership alone.

Summary: Retain the same catalytic reaction in the modeled viral glycan context.

Reason: Reactome R-HSA-9696980 explicitly includes the same human FUT8 catalyst as the ordinary core-fucosylation reaction. The catalyst/physical-entity join supports the enzyme identity and N-glycan transfer activity. The pathway’s limits in direct evidence for individual host enzymes on spike do not overturn FUT8’s established chemistry. This row does not claim that FUT8 cleaves the viral protein or establish a new virus-specific molecular function.

Summary: The retained N-glycan-processing assertion is a more specific biosynthetic process.

Reason: FUT8 genuinely contributes to glycoprotein biosynthesis by catalyzing a glycan modification; it is not simply cargo or a necessary substrate. Nevertheless, the already accepted GO:0006491 N-glycan processing is a direct official is_a child of GO:0009101 and specifies its work. The InterPro mapping is retained as original provenance without inventing a separate biosynthetic role.

Summary: Retain the broad oligosaccharide-biosynthesis context as non-core.

Reason: Human FUT8 directly adds a fucosyl residue to an N-linked glycan acceptor, so the broad biosynthetic context is biologically compatible and is not rejected as a mere loss-of-function consequence. The PMID:9133635 abstract specifies N-linked glycopeptide acceptors, and PMID:36280670 independently measures fucosylated glycan products using engineered human enzyme. Retain this broad oligosaccharide-biosynthesis statement as non-core, while the defining process is represented by the existing N-glycan-processing annotations and core function. No direct typed ancestry from GO:0006491 to GO:0009312 was established in the retained official query, so this is a scope judgment rather than an unsupported TOO_GENERAL path claim; no redundant new annotation is proposed.

Summary: Membrane localization is valid but already covered more specifically.

Reason: The original PMID:19946888 abstract describes membrane-fraction proteomics in human NK-like YTS cells, but the exact FUT8 identification table/peptides were inaccessible. No wrong-gene, absence or specific membrane-topology claim is made from that gap. Independent human localization and the actual Reactome catalyst support Golgi membrane at gene level; its verified is_a path through GO:0098588 and GO:0031090 entails the original generic membrane term. Retain the source/evidence context while marking only its redundant location granularity.

Summary: The retained Golgi membrane assertion provides a specific membrane location.

Reason: The InterPro-derived membrane assignment is consistent with FUT8’s type-II membrane organization and Golgi catalytic role. The existing accepted GO:0000139 Golgi membrane entails GO:0016020 by the verified is_a path through GO:0098588 and GO:0031090. The generic membrane row adds no separate localization coverage. This is not a challenge to the InterPro source.

Summary: Reject peptide-cleavage processing as the function of the modeled glycosyltransferase.

Reason: The official GO:0019082 definition specifically requires cleavage of peptide bonds within a viral protein. The actual Reactome R-HSA-9694548 pathway carries that term, but its child R-HSA-9696980 assigns FUT8 a GDP-fucose-to-N-glycan transfer step through the verified human catalyst. Neither that reaction nor the retained human biochemical evidence gives FUT8 peptide-cleaving activity. Remove this propagated process assertion, while preserving the original TAS source and the legitimate modeled viral glycan context. The upstream query concerns the pathway-to-participant GO propagation, not mistaken FUT8 identity.

Summary: Retain the annotated Golgi cisterna membrane location with its inference scope.

Reason: The official UniProt subcellular record places membrane-bound FUT8 in Golgi stack membrane and identifies trans cisternae by similarity; the incoming combined IEA sources preserve that inference. Human construct colocalization supports the Golgi but does not independently resolve cisternal position or native isoform topology. There is no target-specific contrary evidence requiring rejection. Retain the cisterna location at its original inferred scope, without upgrading it to a direct human cisterna-imaging assay.

Summary: Reject direct protein O-fucosylation for the N-glycan core enzyme.

Reason: GO:0036066 starts with fucose linked directly to a protein serine or threonine oxygen. FUT8 instead transfers fucose to the C6 hydroxyl of the innermost GlcNAc in an N-linked glycan, as its accepted GO:0008424 definition and independent human assays specify. The current official GO response actually contains GO:0008424 part_of GO:0036066, explaining this GO_REF:0000108 inference; the edge is preserved as an upstream ontology inconsistency rather than denied or silently repaired. The accepted N-glycan-processing rows already cover the supported biology, so no redundant replacement or new annotation is added. The full main text of PMID:29304374 additionally reports unchanged O-glycans in affected individuals (data not shown), alongside measured loss of core N-glycan fucosylation; that source-scoped observation corroborates the reaction distinction without claiming that O-glycan data were independently inspected.

Summary: Reject L-fucose breakdown as a FUT8-catalyzed process.

Reason: GO:0042355 describes chemical breakdown of L-fucose. Human FUT8 transfers an intact fucosyl group from GDP-fucose to an N-glycan, preserving the sugar in the product; it does not catalyze degradation of free L-fucose. The selected original PMID:11698403 explicitly separates fucosyltransferase-mediated synthesis from fucosidase-mediated degradation and names FUCA1 in the latter context. This reaction-level contradiction supports removal of the NAS process row, without guessing a replacement PMID or changing the original reference. Donor GDP-fucose turnover is assessed separately under GO:0046368. The PMID:29304374 introduction independently assigns fucosylated-glycan degradation to FUCA1; this distinguishes enzyme roles but is not itself an assay of free L-fucose breakdown.

Summary: Retain direct chemical involvement in GDP-L-fucose metabolism.

Reason: The official definition includes reactions involving GDP-L-fucose, not only its biosynthesis. FUT8 consumes this nucleotide-sugar donor in its own catalytic reaction, transferring the fucosyl group and releasing GDP. It therefore performs a chemical step involving GDP-L-fucose rather than merely responding to its abundance. The human structural abstract and independent purification/product assays support the chemistry, although detailed PMID:17172260 Methods were inaccessible. No GDP-fucose synthesis or feedback-regulatory function is inferred.

Summary: The inherited alpha-1,6 transfer activity is covered by the specific glycoprotein reaction.

Reason: The actual PTHR13132 human leaf AN19/PTN002490629 descends from AN0/PTN000992822, the literal origin of this IBA. The 575-residue target sequence is exact. All four incoming donors map below that node; human self-donor is legitimate descendant evidence. The later retained IBD adds a zebrafish donor, so its donor set is explicitly not equated to the older GOA snapshot. No loss or phylogenetic failure is inferred. The accepted GO:0008424 is a direct official is_a child of GO:0046921 and preserves the supported reaction with its N-glycoprotein acceptor specificity.

Summary: Prefer the already retained acceptor-specific alpha-1,6 transfer term.

Reason: The combined ARBA/mouse-source electronic context is biologically consistent with human FUT8 catalysis. GO:0008424, already retained with direct human biochemical support, is a direct official is_a child of this broader linkage-specific activity and specifies the N-glycoprotein acceptor. The broader row adds no separate molecular function. Its literal mouse accession, Ensembl identifier and rule provenance remain unchanged; no pairwise donor-count criticism is made.

Summary: Retain the positive parotid-exosome-fraction detection.

Reason: Authentic PMID:19199708 Table S2 row 397 identifies historical FUT8 IPI00302523.2, with two sequences/two spectra; Table S1 rows 3993–3994 supply the two peptides. Both match the current canonical sequence, while individual peptides also match different shorter current products. The historical 446-residue database form therefore does not identify a unique present-day natural isoform. Preserve the original part_of UBERON:0001831 parotid scope. Detection in this fraction is supported and non-core, without implying exosome biogenesis, vesicle membrane orientation or physiological secretion as the normal enzyme role.