Historical draft below contains superseded claims. See the 2026-09-26 full source audit appended below for the current evidence and decisions.
Human very-long-chain specific acyl-CoA dehydrogenase (VLCAD), mitochondrial. EC 1.3.8.9 (very-long-chain) / EC 1.3.8.8 (long-chain). HGNC:92. NCBITaxon:9606.
VLCAD catalyzes the first (rate-limiting committed) step of each cycle of mitochondrial fatty acid beta-oxidation (FAO): the FAD-dependent alpha,beta-dehydrogenation of acyl-CoA thioesters to the corresponding trans-2-enoyl-CoA, with the electron-transfer flavoprotein (ETF) as the physiological electron acceptor.
This audit supersedes the earlier annotation and ortholog interpretations above. In particular,
P50544 is mouse Acadvl/VLCAD, not Acadl/LCAD. The earlier removal recommendations,
no-cross-linking interpretation, and ACAD9-paper miscitation claim are withdrawn. The opening
“committed”/“rate-limiting” wording and claim that long-chain activity is a parent of very-long-chain
activity should also be read as superseded by the definition-based synthesis below.
488555581d3642ba24843fc05bcb6d6517dabcd9; local review matched remote blobf840362956e5df1b69588af45347a479c287c1a9. HGNC:92/NCBI Gene37 confirms ACADVL; historical--fallback perplexity-lite --timeout 1200,/tmp uv cache/tool directories. Both attempts failed before invoking a provider becauseuvx could not resolve pypi.org to install deep-research-client. The wrapper exited 1;full_text_available flags.NCBI mouse Acadvl, Gene11370 links P50544 to Acadvl
and MGI:895149. The source GOA identifies P50544, with ENSMUSP00000099634 on Ensembl transfers.
The MGI Acadvl graph displays the
four disputed process terms as mouse IMP assertions with J:95532. This is a graph generated in
March 2023; its date is recorded rather than represented as a new 2026 annotation release.
PMID:15639194(https://pubmed.ncbi.nlm.nih.gov/15639194/) directly reports loss of cold tolerance
in both VLCAD- and LCAD-deficient mice. Temperature-homeostasis transfers are therefore retained
as non-core organismal consequences of fatty acid oxidation. The same abstract reports increased
hepatic oxidation-gene expression in both mutants, while its non-fasted lipogenesis discussion
foregrounds LCAD. Full study results and the exact J:95532-to-PMID mapping could not be recovered.
The six transfers for negative regulation of fatty acid synthesis, negative regulation of fatty
acid oxidation and regulation of cholesterol metabolism are UNDECIDED. Neither a wrong-paralog
claim nor the argument that a catalytic enzyme cannot exert indirect negative feedback is valid.
The original source identifiers remain intact.
Live AmiGO definitions distinguish
medium-chain GO:0070991,
long-chain GO:0004466, and
very-long-chain GO:0017099: their aliphatic
tail ranges are respectively 6–12, 13–22, and greater than 22 carbons. Long-chain and very-long-chain
activities are sibling specializations of acyl-CoA dehydrogenase activity, not parent/child terms.
The definitions concern substrate chemistry rather than the historical names MCAD, LCAD or VLCAD.
The human palmitoyl-CoA mutant assays PMID:9461620, fibroblast flux rescue PMID:7668252 and patient
enzyme assays PMID:8466512 substantiate long-chain catalysis. They do not by themselves establish
the modern very-long-chain threshold. The original IDA very-long-chain row is retained with curator
deference rather than rejected from an incomplete abstract. The machine-fetched UniProt record
separately assigns C24-CoA reaction RHEA:47232 with PMID:21237683, and the human structure explains
how the cavity accommodates extended chains PMID:18227065. Structural cavity capacity is not
misrepresented as a turnover measurement.
The primary ACAD10/11 article includes purified
VLCAD comparators (Figure 4F) and human muscle membrane activity in section 3.7. Its individual
figure rates were not re-extracted. UniProt's RHEA:47296 dodecanoyl-CoA reaction and the current GO
medium-chain definition support retention of boundary-range activity as NON_CORE, instead of
calling the Rhea mapping erroneous solely because ACADM is the main medium-chain enzyme.
Four broad catalytic MF rows are refined to the experimentally grounded long-chain reaction. The
broad energy-derivation process is refined to existing GO:0033539 coverage. That term describes the
beta-oxidation pathway employing an acyl-CoA dehydrogenase initial step; VLCAD performs that step,
not every reaction in the pathway. No NEW process annotation is introduced.
PAINT IBDs in interpro/panther/PTHR43884/PTHR43884-paint.tsv at PTN000856877 confirm the inherited
VLCAD-activity and fatty-acyl-CoA-binding assertions. The human descendant evidence is legitimate
support for node placement. No donor-count or circularity objection is made. InterPro catalytic
and FAD mappings, ARBA provenance, and Rhea substrate mappings were traced from GOA/UniProt.
All organelle and mitochondrial-membrane rows retain their source resolution and are ACCEPT.
The live HPA page reports supported
mitochondrial localization (HPA019006/HPA020595); this does not resolve matrix or inner membrane.
The MitoCoP full main article was inspected, but its individual ACADVL supplementary entry was not
independently re-extracted; the curated mitochondrial assertion is retained with targeted support.
PMID16020546, author-uploaded full primary article
Methods and Figure 5 Results directly include VLCAD in human muscle fractionation. VLCAD sediments
with the membrane fraction. The old MISCITED designation is withdrawn; the paper's ACAD9-focused
title did not exclude experimental VLCAD controls.
The two Reactome records are distinct: R-HSA-1791069 is an expression event, whereas R-HSA-77299
is palmitoyl-CoA dehydrogenation. Matrix TAS rows remain contextual/non-core, distinguishing imported
protein and reaction-compartment placement from mature membrane-associated topology. Matrix is
not an ancestor of inner membrane. Wild-type membrane-dependent assembly and the soluble-matrix
S583W mutant are distinguished PMID:9599005; the structural analysis supports a matrix-facing
peripheral enzyme PMID:18227065.
PMID18063578, author-uploaded full primary article
Table 1 places ACADVL/NP_000009 in Class I, present in both native and cross-linked preparations.
Native anti-TFAM/anti-mtSSB peptide counts are 3/0 and cross-linked preparations 1/2 are 3/0. Thus
the previous review's failure-to-cross-link claim was false. The Methods used HeLa mitochondria,
formaldehyde cross-linking, gradient purification and peptide identification. Retain nucleoid
association as non-core without proposing DNA-binding, replication or genome-maintenance functions.
The GO nucleoid definition
is regional; direct DNA contact is not required of every localized protein.
PMID21492153, full Wiley primary article
Methods compare proliferating and day-17 differentiated Caco-2 cultures. Table 1 identifies ACADVL
with a 3.25-fold increase; the Discussion interprets lipid-enzyme changes as altered metabolic
turnover. Keep MARK_AS_OVER_ANNOTATED for differentiation participation. This is now based on
full study inspection; the IEP evidence code correctly records expression-pattern evidence.
The HuRI source PMID:32296183 includes repeated screens, pair retests and broader validation.
GOA identifies TAF1B/Q53T94 as the partner; UniProt records three experiments. Remove generic
protein binding as functionally uninformative, without labeling the interaction an artifact from
compartment annotations. No unsupported replacement binding function is proposed. Pair-specific
raw assays were not independently reanalyzed. Homodimerization and substrate binding remain real
non-core properties, while FAD binding is integrated with the core catalytic mechanism.
Two core entries describe the same enzyme's long-chain and very-long-chain substrate ranges,
with cofactor, homodimer and membrane information integrated rather than adding a separate FAD-only
physiological function. All 42 source annotation objects remain unchanged outside review.
Final actions: 20 ACCEPT, 9 KEEP_AS_NON_CORE, 6 UNDECIDED, 5 MODIFY, 1 REMOVE,
1 MARK_AS_OVER_ANNOTATED. No NEW annotations or proposed ontology terms were added.
Final just validate human ACADVL passed without curation warnings; just render human ACADVL
passed. All 42 source-field objects and the GOA/UniProt bytes were verified unchanged. YAML trailing
spaces were removed with parsed-data equality checked. History validation and final file hashes are
recorded in /tmp/ACADVL-audit-manifest.json. No Git state or shared project file was changed.
Coordinator inspection of all 42 rows and both core entries found no blocking biological issue. Added the exact cached UniProt C24 catalytic reaction and its experimental attribution to the very-long-chain rows/core, distinguishing curated turnover evidence from structural cavity capacity. Abstract-only local PMID caches now carry explicit reference flags; externally accessed full text remains separately described above. Source assertions and action counts are unchanged.
This append-only entry supersedes the earlier final action count and energy-process refinement above. It also makes the nucleoid evidence explicit; the prior phrase “both native and cross-linked preparations” meant preparation categories, not every experiment. The current YAML reasons and biological questions now stand independently of the earlier draft. The history of the P50544/paralog and paper-attribution corrections remains in this journal.
Independently re-read the author-uploaded full primary article, PMID:18063578, Table 1 and its legend, Results, and Discussion. The table is legible in the indexed full article. The row is Acyl-CoA dehydrogenase,VLC ACADVL NP_000009 3 0 3 0. Its columns are:
| Preparation | Native, anti-TFAM | Native, anti-mtSSB | Cross-linked, preparation 1 | Cross-linked, preparation 2 |
|---|---|---|---|---|
| Independent ACADVL peptides | 3 | 0 | 3 | 0 |
The legend defines these as independent LC-MS/MS peptide identifications with confidence above 90%. They are not abundance measurements. Detection occurred in one native preparation and one cross-linked preparation; the two zero columns preclude claiming identification in every preparation. Class I is the paper's operational core-nucleoid group and includes metabolic enzymes. Class II contains the native-only set, including ATAD3, discussed in the abstract. Thus the abstract does not assign all metabolic enzymes to the native-only set. Cross-linking can capture indirect contacts, and neither the table nor the classification establishes an ACADVL DNA-binding or genome-maintenance activity. The nucleoid row remains KEEP_AS_NON_CORE with this bounded biochemical support.
The exact table row is now quoted in supporting_text_fulltext; the generic abstract set-level quote was removed from this annotation. The immutable local cache is abstract-only, so full_text_unavailable: true remains correct. No cache was altered.
Re-read PMID:16020546, full original Figure 5 Results and caption. Human muscle mitochondrial matrix and membrane fractions were immunoblotted using anti-VLCAD, with purified enzyme controls. The result states that VLCAD pelleted with the membrane fraction; a short direct excerpt is now recorded in supporting_text_fulltext. The local cache remains abstract-only. This assay supports membrane-level localization; separate assembly and structural sources provide the finer inner-membrane context.
The live AmiGO GO:0033539 page and GO:0006635 page, viewed 2026-09-26, display an inferred is_a ancestry through fatty acid oxidation/catabolism without GO:0015980. The energy derivation page places that process under GO:0006091. This checks the displayed is_a relationship, not an exhaustive assertion that no other relation can connect these processes. QuickGO/OLS API requests were unavailable. PMID:7479827 explicitly links VLCAD deficiency to diminished myocardial oxidation and energy production. Accordingly, retain GO:0015980 as KEEP_AS_NON_CORE; replacing it with already-present GO:0033539 would lose a supported distinct assertion.
The MitoCoP finding now quotes an actual cached dataset statement, while explicitly recording that the ACADVL supplementary identification was not independently re-extracted. Core descriptions positively distinguish C16-based long-chain activity from the C24 reaction and extended binding cavity. Both describe the same FAD-dependent membrane-associated homodimer. EC 1.3.8.9 and EC 1.3.8.8 match the fetched UniProt record; the membrane-associated architecture is distinguished from the soluble matrix tetramers SCAD, MCAD and LCAD.
Final actions are 20 ACCEPT, 10 KEEP_AS_NON_CORE, 6 UNDECIDED, 4 MODIFY, 1 REMOVE, 1 MARK_AS_OVER_ANNOTATED. Only the energy row action changed in this follow-up. All 42 source assertions and all fetched GOA/UniProt/cache artifacts remain unchanged. Targeted validation, rendering, new history validation, and byte-based publication hashes are recorded in the follow-up manifest supplied to the coordinator. No Git or remote mutation was performed.