The displayed human product is Q7L5A8-1 (372 residues); Q7L5A8-2 is the natural 159-residue product lacking residues 1–213. The experimental Met93-start construct and the NeuroNet region 93–372 are different from that natural product. Historical accession labels, expression hosts and domain similarity do not establish that both natural products share activity, localization or partners.
The core reaction is free-fatty-acid 2-hydroxylation in the ER membrane. Its hydroxylated acyl products feed the GalCer and GlcCer biosynthetic branches, both descendants of ceramide biosynthesis. These are precursor contributions, not downstream sugar-transfer reactions. Fatty-acid metabolism and CNS myelin maintenance describe distinct supported aspects of the same enzyme's work; none is an ancestor of another selected process. No separate core is introduced solely for heme binding. The cytochrome-b5 domain/signature and human deletion experiment support this enzyme facet without direct human heme-occupancy measurements or a complete physiological electron-transfer mechanism.
Alderson et al., 2004, PMID:15337768, Experimental Procedures and Results/Figures 2–6, uses human FA2H expressed in COS7 cells. The original full-length construct was assembled with its N-terminal cytochrome b5 region and sequence-verified; a Met93-start form and a FLAG-tagged form were tested separately. Membrane fractionation places the expressed protein in the membrane pellet. Microsomes convert deuterated free tetracosanoic acid into the corresponding 2-hydroxy acid in an NADPH/reductase-supported assay. Empty-vector, deletion and electron-supply comparisons bound the attribution. Precise ER localization and topology were predictions in this paper.
Alderson et al., 2005, PMID:15863841, Methods and Results/Figures 2–7, specifies the human construct in COS7 microsomes and identifies the free hydroxy-acid product by retention time and mass fragments. The source distinguishes this direct assay from older assays that measured label in ceramide after multiple reactions. Neither paper is a purified isolated-protein assay. The assay does not by itself establish all physiological electron donors or identical activity of the shorter natural product.
For both papers the immutable normal cache is abstract-only. Selected operative author-upload main text and captions were actually read; direct PDF routes remained inaccessible. No figure pixels were inspected. Public author copies are the 2004 article and the 2005 assay article.
Eckhardt et al., PMID:15658937, Experimental sections and Results/Figure 4, provides independent mouse localization evidence. A sequenced FVB/N 372-residue ORF, expressed as an N-terminal EGFP fusion in CHO-CGT cells, colocalizes with calnexin and retains lipid-product formation. Human membrane fractionation plus this conserved localization supports the ER-membrane refinement; it does not become a human marker-imaging experiment. Reactome's human FA2H entity explicitly places residues 1–372 in that compartment. The mouse paper's Discussion calls ceramide usage probable while acknowledging that the immediate substrate was not established.
PMID:22517924(https://pmc.ncbi.nlm.nih.gov/articles/PMC3371244/), Methods and Figure 1, demonstrates R-2-hydroxypalmitate production in FLAG-human-FA2H CHO cells. The complete main was read. Its separate mouse 3T3-L1 experiments address enantiomer rescue, raft mobility, GLUT4/glucose uptake and lipid abundance (Figures 2–6). Those outcomes do not isolate turnover of a ceramide-bound acyl substrate. GO:0120520 therefore captures the directly established free-substrate reaction. The more specific existing GO:0120521 assertion remains undecided; this is not a claim that such activity has been disproved. Hydroxylated acyl precursors and increased 2-hydroxyceramides support participation in ceramide biosynthesis without assigning ceramide-synthase activity.
PMID:17355976(https://pubmed.ncbi.nlm.nih.gov/17355976/) remains available only as metadata and a positive abstract in this review. It reports human-keratinocyte enzyme/lipid changes and lamellar-body/extracellular-membrane phenotypes after FA2H knockdown. Operative Methods, controls and microscopy were not recovered. At the initial checkpoint the skin-barrier assertion was undecided; the feedback assessment below retains it as KEEP_AS_NON_CORE with the same abstract-only access limit. Independent human biochemical originals support the specific enzyme refinement without being attributed to the unread keratinocyte assay.
The original main of PMID:17901466(https://pmc.ncbi.nlm.nih.gov/articles/PMC2662131/) was recovered and selected Methods/Results/captions/Discussion read. Primary rat Schwann cells and D6P2T cells, knockdown, free deuterated C24 assays and lipid measurements link FA2H activity to hydroxylated GalCer. This supports a conserved precursor-supply step, not galactosyltransferase chemistry or a direct human Schwann-cell experiment.
Selected original Methods and the Results section “Analysis of Skin and Epidermal Sphingolipids,” including the Figure 5 caption, were read for PMID:21628453(https://pmc.ncbi.nlm.nih.gov/articles/PMC3138264/). Complete export was not established and its normal cache remains abstract-only. In mouse whole skin, selected 2-hydroxy GlcCer species decreased; isolated epidermis retained the measured hydroxylated lipids. The MS analysis selected a d18:1 backbone, and the effects differed by acyl-chain length: C20 was absent, C22–24 decreased, and C16/C26 were not significantly changed. The transferred GlcCer process therefore retains tissue and species limits. It does not assert that every glucosylceramide or every tissue depends on FA2H. No Figure 5 pixels were inspected. An unrelated initial search hit was excluded from the target evidence.
Complete originals were recovered for PMID:18815260(https://pmc.ncbi.nlm.nih.gov/articles/PMC6671223/) and PMID:21491498(https://pmc.ncbi.nlm.nih.gov/articles/PMC3094470/). Selected knockout, lipid-analysis and morphology Methods, Results and captions distinguish loss of hydroxylated myelin lipids from initially formed myelin and later degeneration. The second paper includes glial conditional deletion. These support conserved CNS myelin maintenance through FA2H's executed lipid-precursor chemistry; they are not human perturbation experiments. No microscopy pixels or supplements were read.
Actual PANTHER v19 PTHR12863 tree and MSA recover target PTN002489952 with an ungapped sequence exactly matching the displayed human 372 residues. The path contains asserted PTN000304948 and PTN002656510. Four current IBD records match the imported ER, hydroxylase, ceramide-biosynthesis and CNS-myelin-maintenance assertions; no NOT lies on the target path. Tree GO-slim display fields are not substituted for the actual IBD records. Target self-WITH is valid descendant experimental grounding, and donor count is not used as an evidential score. Target-specific chemistry supports the free-substrate refinement while retaining valid ancestry.
Current mouse Q5MPP0 and rat Q2LAM0 GOA were traced to their experimental sources. Each transfer keeps the actual mouse/rat cell, tissue, construct and assay boundary. The official GO_REF methods were read for InterPro, manual ortholog transfer, PAINT, UniPathway, subcellular-location mapping, Ensembl Compara, Rhea and combined IEA. UniPathway is no longer maintained; the current human UPA00199/UPA00787 links and source assertions are recorded, but the attempted standalone mapping-file recovery failed.
The actual human InterPro matches include IPR014430 across the displayed protein and IPR018506 at its cytochrome-b5 heme-binding signature. IPR014430's Scs7 family name does not assign every yeast substrate preference to human FA2H. ARBA00084402 and ARBA00027419 each have one matching Eukaryota/FunFam condition; unrelated condition sets are excluded. SL-0097 maps directly to ER membrane. The preserved curated PANTHER-family review is never used as scientific evidence.
Reactome 5693761 links through catalystActivity 5693796 to human FA2H entity 5693771, residues 1–372, with its own ER-membrane compartment. The source reaction converts a free fatty acid to a 2-hydroxy fatty acid, with distinct CYB5B redox complexes. The parent sphingolipid pathway 1660661 includes this reaction; its aggregate compartments and other events are not assigned wholesale to FA2H. The target reaction closure contains 12 fully fetched objects plus inline terminal O2, whose standalone request returned 429. This is not a claim to have expanded every parent-pathway event.
Actual Rhea reaction pages 38551, 38559 and 38855 show palmitate, C24 and generalized free-substrate equations and their GO:0120520 mapping. The requested RDF URLs returned usable HTML, which is recorded as HTML rather than relabeled as RDF. Separate electron carriers in that model are not measurements of FA2H-bound heme.
Current official definitions and ancestry support membrane/ER→ER membrane, lipid metabolism→fatty-acid metabolism, sphingolipid biosynthesis→ceramide biosynthesis, and broad hydroxylase→free-fatty-acid 2-hydroxylase. The ceramide branch is supported by an executed precursor-supply step, not merely by FA2H being necessary or being a substrate. Existing GalCer and GlcCer contexts remain source-specific descendants; no new process annotation is proposed. GO:0002933 lipid hydroxylation was inspected but not selected because its wording about fatty acids in a lipid could imply an immediate substrate not established by these human free-fatty-acid assays.
HuRI, PMID:32296183, Figure 1/Extended Data Figure 1 and the public interaction data, reports a systematic human binary yeast two-hybrid screen. All 26 imported FA2H partner contexts were matched to exact source records. FA2H is labelled Q7L5A8-1, human ORFeome clone 513, with historical Ensembl identifiers and partial DNA-sequence identification. Array, pooling and validated-two-hybrid entries are distinct source records, not three independent measurements of a native complex. The Gal4 fusion configurations and yeast host are explicit. Study-wide MAPPIT/GPCA benchmarks are not individual validation of each FA2H pair. Main text was read; the attempted supplementary Methods DOCX returned HTML and was not treated as a document.
NeuroNet, PMID:32814053, STAR Methods and Table S2, is also a yeast two-hybrid study. NN1.0 contains DYNC1H1/FA2H and KLK6/FA2H; the extended network retains both. Official IntAct IM-28217 participant records identify the FA2H binding region as residues 93–372. The final main PDF and workbooks were available, and pairwise reporter controls were read. No FA2H-specific DULIP/LuTHy result or complete historical plasmid sequence was verified.
Supported contacts are retained as non-core under the campaign's explicit generic-binding exception. Specific receptor, enzyme-regulator, transcription-factor, dynein-complex and adaptor interpretations were considered against current definitions. The screens do not establish those mechanisms, native partner processing states or physiological assemblies. In particular, contact with KLK6 does not assign protease inhibition to FA2H, and contact with DYNC1H1 does not establish engagement of an assembled dynein motor. The natural 159-residue product is not assigned the screen results by equivalence.
The full_text_unavailable flags follow the selected normal-cache metadata: true means that the immutable cached record does not contain the full article. External original availability and actual reading are recorded separately. The older human enzyme papers and the mouse skin paper have actual selected original access despite abstract-only normal caches; the complete NeuroNet final main is available outside its abstract-only cache. Its flag is therefore true under this cache convention, while its complete external-main availability and actual STAR Methods/Table S2 reading remain unchanged. Other recovered full originals were read at the explicitly listed scopes; cache metadata does not assert that every section, figure pixel or supplement was inspected. The four shared reference assessments preserve both direct and database reading scopes with labels. The source notes distinguish genuine source acquisition from its scientific interpretation, and no publication cache was edited.
The 372-residue displayed and 159-residue natural products, separate transcript models, engineered fragments and processing notation are not interchangeable. UniProt heme-ligand assignments are inferred, and Zn-site assignments by similarity do not establish direct human metal occupancy. Biological questions address product activity, immediate substrate choice and tissue-specific precursor use; they do not create new annotations. Supported generic associations are retained under the campaign instructions, with no new publication quotations.
The three refined IBA contexts retain valid ancestral assertions. Their structured metadata uses NO_FAILURE_CORE with the inspected PTN000304948 node marked SUPPORTS_TRANSFER; no failure mode is invented for target-supported specificity. The remaining generic-binding advisories concern the 28 experimentally supported HuRI/NeuroNet contexts, retained as non-core under the standing campaign convention.
The human keratinocyte study PMID:17355976(https://pubmed.ncbi.nlm.nih.gov/17355976/) has an accessible positive abstract, including FA2H-siRNA lipid changes, abnormal lamellar bodies, failure of normal extracellular lamellar membranes and fatty-acid hydroxylation before ceramide/GlcCer production. That performed precursor role supports contextual skin-barrier retention. Full operative Methods, controls and microscopy remain unread; the abstract is not presented as a complete main. The earlier UNDECIDED checkpoint is superseded by KEEP_AS_NON_CORE. Mouse study PMID:21628453 is a separate whole-skin/sebaceous study with retained epidermal lipids, not the experimental basis of the human barrier row.
Current GO:0046513 directly includes GlcCer GO:0006679 and GalCer GO:0006682 as is_a children. Both are descendants of sphingolipid biosynthesis, and neither is an ancestor of the other. The four broad assertions are now MODIFY with both replacements, as required for overly general terms by the user instructions. The actual source of each broad assertion remains intact. Additional branch evidence is explicitly attributed: mouse FA2H/CHO-CGT and rat Schwann-cell GalCer experiments, versus human keratinocyte and mouse whole-skin GlcCer results. The core includes those specific branches once, and FA2H supplies hydroxylated precursors rather than executing downstream glycosylation.
The Reactome R-HSA-5693761 reaction and human PMID:17355976 abstract independently corroborate free-substrate ordering. The direct human microsomal assays remain primary substrate evidence despite their abstract-only normal caches and separate original captures. Those observations do not exclude a second substrate route. GO:0120521 remains unresolved; neither missing terminology nor FlyBase being the assigning database establishes an erroneous human assertion. No source defect is claimed.
Human disease context now includes Edvardson et al., 2008, Dick et al., 2010 and Kruer et al., 2010. Selected original clinical/genotyping/assay/Results/Discussion text was read for the first and third; only the accessible abstract was read for the second. Biallelic FA2H variants link leukodystrophy, SPG35 and FAHN, with variable brain-iron findings. These clinical observations complement the mouse maintenance experiments without substituting for them. D35Y lost activity in the reported COS7 assay, whereas R154C had no significant assay difference (data not shown) and reduced abundance in a separate cell system; variant effects are not interchangeable. No new figure pixels, patient table-cell audit or supplementary panels were inspected.
The unchanged UniProt record provides checkable curated locators at CC:CATALYTIC ACTIVITY (tetracosanoate), CC:BIOPHYSICOCHEMICAL PROPERTIES (KM), CC:DOMAIN (cytochrome-b5 domain) and CC:COFACTOR. Its two-Zn proposal is ECO:0000250 similarity evidence from Q03529, not direct human metal stoichiometry. A focused question/experiment retains that uncertainty. Both natural-product objects remain source-exact, with no inferred functional equivalence.
Cache availability, independent external-original availability and actual reading remain separate. PMID:17355976 and PMID:20104589 flags stay true because their selected caches are abstract-only; PMID:19068277 and PMID:20853438 are false because complete mains are cached. All 21 prior reference identities/findings remain, complete prior notes are retained and PMID:17355976's superseded action wording is explicitly historical. Three existing genuine normals are newly selected, with no cache rewrite or new provider claim.
The saved official QuickGO evidence makes the ontology claim independently checkable offline. Its unmodified HTTP-200 responses from 2026-10-25 09:21 UTC explicitly give GO:0006679 is_a GO:0046513, GO:0006682 is_a GO:0046513, and GO:0046513 is_a GO:0030148. The projection preserves complete returned ancestor lists and exact response selectors; both branch terms entail the broad ceramide process under this retrieved ontology, and neither is an ancestor of the other. The receipts identify this snapshot by URLs, times and hashes; no unspecified release number is inferred. The repository's generated GO label cache is unchanged.
The four MODIFY decisions are gene-level refinements synthesized across explicitly identified sources. They do not replace the imported source object, recode an IDA result, revise an IBD, or claim that the source of a broad assertion performed each branch-specific experiment. In particular, PMID:15337768 expressed human FA2H in monkey COS7 host cells and measured free-fatty-acid hydroxylation and increased 2-hydroxyceramides. Its GalCer proposal is not treated as a branch-specific assay. The original GO:0046513 IDA assertion and PMID remain present in the imported row; the review's narrower proposals cite separate evidence.
| Imported context (zero-based review index) | What that source supports | Separate branch support used in the synthesis |
|---|---|---|
| 42: GO:0030148, Reactome TAS | The human FA2H free-fatty-acid hydroxylation step within sphingolipid biosynthesis | The GalCer and GlcCer evidence listed below |
| 46: GO:0046513, PAINT IBA | Valid inherited ceramide-precursor contribution at PTN000304948 | The same separately cited branch evidence; no narrower IBD is invented |
| 47: GO:0046513, PMID:15337768 IDA | Human FA2H-dependent hydroxylation and 2-hydroxyceramide production in COS7 cells | The branch-specific sources below, not an expanded claim about the 2004 experiment |
| 48: GO:0046513, combined IEA | The matched ARBA rule and mouse donor support broad precursor contribution | The same separately cited branch evidence, not an assertion that the rule independently tests both branches |
For GalCer, PMID:15658937(https://pmc.ncbi.nlm.nih.gov/articles/PMC1186713/) uses mouse FA2H in CHO-CGT cells, and PMID:17901466(https://pmc.ncbi.nlm.nih.gov/articles/PMC2662131/) links FA2H knockdown, free-fatty-acid hydroxylation and hydroxylated GalCer in rat Schwann-cell systems. For GlcCer, the accessible human-keratinocyte abstract of PMID:17355976(https://pubmed.ncbi.nlm.nih.gov/17355976/) and selected mouse whole-skin Results of PMID:21628453(https://pmc.ncbi.nlm.nih.gov/articles/PMC3138264/) support the affected hydroxylated lipid branch; the latter's retained epidermal pool and chain-length dependence remain explicit. The earlier sections give each original-access and actual-reading limit. This is conserved precursor-hydroxylation work by FA2H, not sugar transfer, unrestricted dependence of every tissue/lipid species, or equivalence of both natural products.
NO_FAILURE_CORE and SUPPORTS_TRANSFER retain the positive PAINT judgment: the broad ancestral function and inspected target placement are sound. They do not claim that the ancestral term is the most specific summary supported by all later target and ortholog evidence. The narrower recommendations therefore do not allege a propagation failure. Under the campaign instructions and user-supplied review guidance, supported overly general terms use MODIFY with explicit replacements. The alternative TOO_GENERAL suggestion is not adopted. The direct is_a evidence preserves the broad ceramide claim by entailment while the imported broad assertions remain unchanged and inspectable.