FKTN source evidence, context coverage and experimental boundaries

The supported core is first ribitol-phosphate transfer on the GalNAc residue of the O-mannose-derived core M3 glycan, at the Golgi. Source-specific evidence and independent corroboration are distinguished. All 43 genuine seed source objects, both natural products and the 44-row normal TSV are preserved. No new publication quotations or NEW annotations are authored.

Primary source access and limits

PMID:15213246(https://pubmed.ncbi.nlm.nih.gov/15213246/)

Normal cache is abstract-only. The original scanned article was separately read on pages 709–712 in an official government-hosted report. Transfected myc-fukutin is the cis-Golgi comparator; exact insert species was not stated in the recovered Methods. Endogenous human/mouse rough-ER evidence in that paper concerns FKRP and is not reassigned to FKTN.

PMID:17034757(https://pubmed.ncbi.nlm.nih.gov/17034757/)

Verified original abstract plus official IntAct human-pair assay records; full publisher Methods remained inaccessible. Co-IP/two-hybrid association, colocalization and the Y371C redistribution phenotype support bounded interpretation. IntAct records tagged human proteins mainly in COS-7 and one C2C12 colocalization experiment. No purified allosteric activation is claimed.

PMID:18808525(https://pubmed.ncbi.nlm.nih.gov/18808525/)

Official full HTML Methods, Results and Discussion read. Human HeLa/ZR-75-1 nuclear/cytoplasmic fractionation and immunogold include absorption/RNAi controls and residual specificity uncertainty. F697 at 20 nM/five days increased Ki-67/phospho-c-Jun; no proliferation rescue or direct JNK assay. The later 1321N1 findings are contextual differences, not blanket refutation.

PMID:25279699(https://pubmed.ncbi.nlm.nih.gov/25279699/)

Full Results/Table 1/Fig1 caption and cloning Methods read. Human FKTN pTW322 and mouse Fkrp pTW323 are distinct Myc constructs in HEK293T. Human patient fibroblast glycan-maturation defects are separate from the construct-localization experiment.

PMID:26923585(https://pubmed.ncbi.nlm.nih.gov/26923585/)

Normal cache remains abstract-only. The original Kobe-deposited PDF was separately recovered and ROOT read selected Results/Methods and visually inspected pages 8,12,13. NMR establishes GalNAc-C3 attachment; pNp NMR acceptors lack Man6-phosphate. Main text does not establish every enzyme insert species. Mouse-liver activity fractionation supports Golgi context; the supplement was not inspected.

PMID:27194101(https://pubmed.ncbi.nlm.nih.gov/27194101/)

Full primary Results and Methods support purified mouse Fktn transfer and human knockout/mouse-Fktn rescue. Rabbit dystroglycan was produced in human host cells. The paper alone does not resolve exact attachment site; human host is not human enzyme provenance.

PMID:27601598(https://pubmed.ncbi.nlm.nih.gov/27601598/)

Normal abstract-only cache plus separate official full HTML/selected Methods/Results and Figures 6–7 read in independent consultation. Tagged human-protein association in HEK293T is distinct from HCT116 glycoform experiments. Two residues can occupy separate core-M3 sites; tandem extension, tentative GroP identity and alternative donor hypotheses are distinguished.

PMID:28514442(https://pubmed.ncbi.nlm.nih.gov/28514442/)

Normal primary Methods support human tagged-bait AP-MS. Exact FKTN–RXYLT1 supplemental edge, bait direction and spectra were not replayed. Trusted curated association is independently corroborated by targeted complex studies.

PMID:29477842(https://pubmed.ncbi.nlm.nih.gov/29477842/)

Verified abstract and author-deposited doctoral summary read; journal Methods remain unavailable. Tagged triple co-IP, GM130 colocalization, recovered activities and FKTN product MS support complex/function. Endogenous FKTN, but not FKRP, was visualized in the summary immunoblot. Exact insert provenance, obligatory stoichiometry and channeling remain unresolved.

PMID:33961781(https://pubmed.ncbi.nlm.nih.gov/33961781/)

The normal full-text flag accompanies abstract/introduction/discussion rather than recovered Methods. Exact edge, cell-line allocation and spectra were not replayed. Retain curated association with targeted corroboration; no assertion that both cell lines tested this pair.

PMID:9690476(https://pubmed.ncbi.nlm.nih.gov/9690476/)

Verified original abstract and metadata. The discovery links FKTN to congenital disease and reports transfection results suggesting secretion. Full construct design, processing, endogenous secretion and detailed developmental assays were not inspected.

PMID:29884773(https://pubmed.ncbi.nlm.nih.gov/29884773/)

Normal abstract-only cache plus separate official full HTML Results/Experimental procedures read. Soluble Myc-tagged enzyme assays establish ribitol/glycerol donor competition and altered acceptor extension in vitro. FKTN kinetic fits are distinct from FKRP relative-product assays; physiological regulation remains unestablished.

PMID:34830034(https://pubmed.ncbi.nlm.nih.gov/34830034/)

Official full HTML selected Methods/Results/Discussion read. Human astrocytoma 1321N1 uses 40 nM F697/four days and tagged overexpression; proliferation direction differs from HeLa/ZR-75-1, while p-JNK/p-c-Jun remain unchanged. Cell-lysate association/ELISA does not establish purified direct DNA binding or a universal transcription-coactivator function.

Context projection and source boundaries

Forty-four original QuickGO contexts map to 43 stored decisions. The table preserves each original identifier and assigning database; indices are zero-based. The two POMGNT1-binding contexts map to index 10. IntAct context 447030760 contains two alternative original extension groups: occurs_in GO:0005783 (endoplasmic reticulum), or occurs_in GO:0005794 (Golgi apparatus). UniProt context 447030759 has no extension. The 16-column fetched TSV retains both assigning-database records but has no extension column; the genuine seed represents their shared fields once. No source field or duplicate NEW row was manually created. These extension alternatives describe the source assertion and are not new exclusive native-location claims.

Stored actions: 26 ACCEPT, 13 KEEP_AS_NON_CORE, 2 TOO_GENERAL, 1 MARK_AS_OVER_ANNOTATED, 1 UNDECIDED. Original-context actions: 26 ACCEPT, 14 KEEP_AS_NON_CORE, 2 TOO_GENERAL, 1 MARK_AS_OVER_ANNOTATED, 1 UNDECIDED.

Supported proliferation and developmental observations are retained as context, not as additional catalytic cores. The JNK-cascade term is marked over-annotated because the earlier phospho-c-Jun readout does not isolate regulation of that cascade. Structured CURRENT findings preserve both proliferation outcomes and record opposing directions and context limits without marking either source result as superseded. The unresolved ER annotation is not a claim that ER localization is false. Both nuclear and extracellular observations are separated from the established Golgi glycan-transfer function. The source-specific reasons retain experimental species, constructs and access limits.

The two natural UniProt products are O75072-1 and O75072-2; VSP_045961 changes the isoform 2 C-terminal segment. The seed product objects are retained without assigning engineered soluble/tagged proteins, secretion or nuclear localization to either natural product. No primary study here establishes that those pools are isoform-specific.

For original links, ontology definitions, PAINT provenance and immutable normal-source hashes, see the source guide. These notes summarize primary findings without inserting new supporting-text quotations.

The follow-up retains both human-cell outcomes as CURRENT findings, with reference-only support and explicit opposing-direction notes. The earlier paper accurately proposed suppression in its tested cells; no hypothetical universal claim is attributed to it as a DISPUTED finding and no superseded_by link is added. The earlier 20 nM/five-day and later 40 nM/four-day protocols were re-read; context dependence is plausible, but no single differing variable was isolated as the explanation.