Current judgments are in the 2026-09-26 full annotation audit below. Earlier entries are a historical journal and include superseded mechanistic interpretations and provider-access outcomes.
UniProtKB: Q99424; HGNC:120; gene ACOX2 (chromosome 3p14.3). 681 aa; C-terminal SKL
peroxisomal targeting signal (PTS1). EC 1.3.3.6. FAD flavoprotein. Homodimer (by similarity).
Deep research note: falcon deep-research provider is out of credits (HTTP 402), so no
-deep-research-falcon.md was generated. This review is grounded in the UniProt record,
the seeded GOA, and cached publications only.
ACOX2 is the human peroxisomal branched-chain acyl-CoA oxidase (originally "hBRCACox";
homolog of rat trihydroxycoprostanoyl-CoA / THCC oxidase). It is a FAD-dependent oxidase
that catalyses the first, oxygen-dependent, rate-limiting dehydrogenation step of
peroxisomal beta-oxidation, removing two hydrogens and introducing a trans-2,3 double bond
into the acyl-CoA, producing H2O2 as a byproduct (UniProt FUNCTION; EC 1.3.3.6).
Two physiological substrate classes:
1. C27 bile-acid intermediates — (25S)-THCA-CoA and DHCA-CoA. ACOX2 initiates the
side-chain shortening that converts C27 cholestanoic acids to the mature C24 bile acids
(cholic / chenodeoxycholic). PMID:27884763
2. 2-methyl-branched-chain fatty acyl-CoAs — e.g. (2S)-pristanoyl-CoA; stereospecific
for (2S)-methyl isomers. Redundant with ACOX3 for pristanoyl-CoA (PMID:29287774, per
UniProt). Also mono-methyl branched-chain FAs (mmBCFA) in brown adipose tissue (by
similarity to mouse Q9QXD1). Low-efficiency straight-chain activity (C10, C16).
The 1990 human-liver enzymology paper PMID:2079609 established that human liver contains a
separate trihydroxycoprostanoyl-CoA oxidase distinct from palmitoyl-CoA oxidase (ACOX1),
explaining why bile-acid metabolism is normal in ACOX1 deficiency — this separate THCC
oxidase is ACOX2.
Peroxisome / peroxisomal matrix. Direct protein evidence in human liver
PMID:8943006 (immunocytochemistry; SKL PTS1; absent in Zellweger livers) and
PMID:2079609. IMP peroxisomal localization confirmed for both WT and R225W mutant in
HepG2 PMID:27884763. The Reactome TAS cytosol annotations refer to the peroxisomal-import
transit state (PEX5 cargo in cytosol before matrix translocation), not the site of
catalytic activity.
ACOX2 deficiency / Congenital bile acid synthesis defect 6 (CBAS6; MIM:617308) —
autosomal recessive inborn error of bile-acid synthesis. Accumulation of toxic C27
intermediates (THCA, DHCA), negligible C24 bile acids, persistent hypertransaminasemia /
liver fibrosis; variable neurological features (ataxia, cognitive impairment). Known
variants: R225W PMID:27884763 and premature termination p.Y69* PMID:27647924.
The latter corresponds to loss of residues 69–681 in UniProt's protein-consequence
annotation, not an N-terminal in-frame deletion. Phytanic/pristanic acids remain normal (redundancy with ACOX3 for BCFA).
Read every source annotation and relevant evidence, including primary full texts for disputed reactions and the exact PAINT target path. See ACOX2-primary-source-checks.md for assay context, access limits, short exact excerpts, and pending focused adjudication. All original source fields are preserved; no NEW annotation is added. Broad true chemistry and localization are retained separately from exact substrate/reaction conflicts.
This section supersedes conflicting assertions in the earlier journal and the September 20 primary-source notes. Those historical files are retained; no generated source was rewritten. In particular, GO:0033791 is not equivalent to the UniProt oxidase reaction, homodimerization is unresolved, weak straight-chain oxidation is not disproved, and generic protein-binding annotations are removed as uninformative. The current synthesis does not assign mouse brown-fat physiology as an established human function.
Approved human symbol ACOX2, HGNC:120, NCBI Gene 8309 and UniProt Q99424 were checked against primary nomenclature records; aliases include BRCACOX, BRCOX and THCCox. The parent verified main 9e34ed516aaa199a283a5042f6adf3b60ae45d54 and no overlapping open PR. The review baseline was INITIALIZED with 40 source rows and 20 references. Local YAML blob 1811ae9458b2b1ab4d27b670ae414dd3a05d4063 and notes blob 7b94ca1fd30c55d359a3a2c68c83445bad63a7bd matched main. Local HTML was an older derived rendering; the publication base is remote blob d6264f4213cfd4c447528aa5f58435322d996f9a.
The required genuine Falcon request with a 1200-second timeout and automatic perplexity-lite fallback was attempted using writable temporary UV tool/cache directories. Both failed before reaching a provider because PyPI could not resolve deep-research-client (three retries, 6.4 and 9.8 seconds). The earlier HTTP 402 note describes a different session. No provider output was produced or authored manually. Concurrent normal publication fetching found all seven existing PMID caches. Normal additional fetches of 8654595, 8026493, 9218493 and 8387517 also failed DNS and produced no files. The last two are required new review references, so the publication must remain draft until they are cached normally. The first two are unresolved donor leads only. These notes are manual research, not provider output.
Live GO:0033791 specifies 25R-THCA-CoA hydroxylation using water and an acceptor, while the ACOX2 UniProt RHEA:46728 reaction consumes oxygen and produces a 24E-enoyl product from 25S substrate. Historical oxidase synonyms therefore do not settle the formal reaction mismatch. The human PMID:27884763(https://pubmed.ncbi.nlm.nih.gov/27884763/) publisher methods/results describe 10 micromolar THCA supplied to HepG2 cells over 48 hours, with cholic-acid production measured by HPLC-MS. This supports pathway participation, but does not isolate the elementary hydroxylation specified by GO. Its IDA and the exact IBA chemistry remain UNDECIDED; no self-evidence circularity is alleged.
The ISS donor O02767 is rabbit ACOX2, not rat. Full PMID:9218493(https://pdfs.semanticscholar.org/25a0/7b6451608e1ad668bdbeb8c04fe2dc8ae4e7.pdf?skipShowableCheck=true) verifies the enoyl product by chromatography and mass spectrometry. Its small hydroxylated product is attributed probabilistically to endogenous COS-cell hydratase. Accordingly the donor-supported ISS is MODIFY to GO:0003997, with the nomenclature problem retained as a question. This does not assert universal impossibility of coupled hydroxylation. The annotation-reviewer peer independently checked the product analysis and endorsed this bounded refinement.
Full PMID:29287774(https://www.researchgate.net/publication/322072163_A_novel_case_of_ACOX2_deficiency_leads_to_recognition_of_a_third_human_peroxisomal_acyl-CoA_oxidase) was read through methods, results and discussion. The human recombinant assays use a yeast fox1 deletion background, 100 micromolar substrate, and HPLC measurement of enoyl plus hydroxylated products. The panel includes C10, C16, pristanoyl and THC-CoA, not C24/C26. It supports weak straight-chain capacity as NON_CORE and shared ACOX2/ACOX3 pristanoyl activity; only ACOX2 handled THC-CoA in that tested comparison. Assay substrate breadth is kept separate from predominant physiological flux.
The primary human purification study PMID:8387517(https://www.researchgate.net/publication/14810824_The_CoA_esters_of_2-methyl-branched_chain_fatty_acids_and_of_the_bile_acid_intermediates_Di-_and_trihydroxycoprostanic_acids_are_oxidized_by_one_single_peroxisomal_branched_chain_Acyl-CoA_oxidase_in_h) supplies two additional qualifications. Table III uses partially purified preparations chromatographically separated from palmitoyl-CoA oxidase; its lignoceroyl-CoA rate is 0.14 versus 59.4 nmol H2O2/min/ml for 2-methylpalmitoyl-CoA (0.2%). This supports retaining limited VLCFA capacity as NON_CORE, not a claim of in-vivo flux. The paper also contrasts roughly 67–70 kDa preparations with an earlier 138 kDa estimate and explicitly leaves native assembly unresolved, including possible purification-induced dissociation. Homodimer ISS from rat ACOX1 therefore remains UNDECIDED. The short external-text excerpts are identified as such in the YAML; no uncached body text is impersonated as a repository cache.
Live GO:0000038 describes an aliphatic tail longer than 22 carbons; the total carbon count of a C27 steroid does not itself meet that criterion. Normal patient VLCFA levels do not prove absent side activity. The earlier generalization that every ACOX2-deficient patient lacks C24 bile acids is also narrowed to the measured family and assay.
All seven PAINT rows were assessed using the cached ancestral path rather than donor counting; the source_entities use actual PTN nodes. The fatty-acid-binding IBD draws on rat Acox2/Acox3. NCBI source tracing leads to PMID:8654595 and PMID:8026493, but the accessible abstracts do not settle free-acid recognition, so that exact binding annotation remains UNDECIDED. Acyl-CoA preference alone cannot disprove free-acid binding. The rat ligand-bound structure called ACO-II in PMID:16672280 belongs to ACOX1/P07872, not ACOX2, and is not repurposed as direct ACOX2 evidence.
Electronic mappings retain actual Rhea, InterPro and ARBA source identities. Unread rule predicates remain UNRESOLVED while the annotation judgment follows independent biology. The oxidoreductase and broad metabolic terms are refined to their demonstrated oxidase chemistry or pathway. Live GO:0071949 specifically denotes oxidized FAD and is a child of the broader GO:0050660; the two are not synonyms. The cofactor transfer is consistent with the flavoprotein reaction, while FAD supplementation is not presented as an independent affinity measurement.
The P07872 ISS donor is rat ACOX1, a paralog: cofactor and location can transfer with human corroboration, while assembly needs separate assessment. PMID:8943006 is abstract-only for this session and its original localization is retained with later human PMP70 colocalization. PMID:2079609 is pre-cloning human-liver enzymology, not sequence identification. Five cached Reactome records were read: matrix locations describe the catalytic compartment, and the two cytosolic rows describe PEX5 cargo before import. No import-machinery process is invented for that cargo.
The BioPlex 2.0 and HuRI full cached texts support study context; the BioPlex 3.0 cache is truncated despite metadata claiming full text, so its review flag records unavailable full text. GOA/UniProt identify STRN3 and DYNLT1 partners. All three generic binding rows are REMOVE because no informative ACOX2 molecular function is established, without denying interaction or pretending the pair-level supplements were revalidated.
All 40 source assertions, alternative products, UniProt/GOA files, PAINT evidence and historical primary-source notes are preserved. One core function integrates matrix oxidase chemistry, bile-acid synthesis and oxidase-dependent beta-oxidation. No NEW annotation is added. Validation and independent-review outcomes follow below.
Independent focused consultation by the annotation-reviewer peer confirmed the rabbit product-analysis refinement and separately read PMID:8387517 Table III and Discussion. The peer agreed that the small C24 rate supports only preparation-level in-vitro capacity and that the native-mass discrepancy warrants UNDECIDED. Its 0.2% figure is a substrate-rate comparison within that preparation, not an organ-level physiological contribution.
Final local checks: just validate human ACOX2, just validate-history and just render human ACOX2 pass. Gene validation has two warnings: the required uncached PMID:9218493/PMID:8387517 references and the deliberate source-specific GO:0033791 split (rabbit ISS MODIFY; unresolved human IDA/IBA UNDECIDED). Final actions are 23 ACCEPT, 6 KEEP_AS_NON_CORE, 4 MODIFY, 4 UNDECIDED and 3 REMOVE. All 40 source assertions and alternative products were compared against the baseline; immutable source hashes match. Trailing spaces were removed with parsed-YAML equality. Exact four-file handoff: /tmp/ACOX2-local-manifest.json. No Git or remote writes were made; the parent independently reviews and publishes the draft.
Coordinator review found no further biological blocker after inspecting all 40 decisions, the core and reference assessments. The reference-level full-text flags for PMID:27884763 and PMID:29287774 now record their abstract-only local caches; PMID:9218493 and PMID:8387517 likewise remain unavailable locally despite successful external primary reading. This flag describes the cache, not the truth or verification of a citation.
The citation audit also includes notes and historical journal entries. The full missing-cache list is PMID:16672280, PMID:27647924, PMID:8026493, PMID:8387517, PMID:8654595 and PMID:9218493. Two support new decisions; the remainder document donor leads, an excluded paralog source or historical disease context. Keep the PR draft until these required records are retrieved normally. No missing source or provider report was fabricated.
Read the full review and formal changes-requested summary against published head dda08a47bbe1ebb4a788a579d5a410f2456afd55. The coordinator independently confirmed the three starting blobs: YAML ff99313a1f911d286e931dc861d8b1bb89d4afc3, notes 080e8b821db260aa1b782bc9f40c8d39b8d13825, and HTML 616fbc609bc946e729d69e2907c4c0f8a8a34d69. This section supersedes the prior C24 retention, mixed hydroxylase decisions, and one-core synthesis. Source annotations, alternative products, and machine caches remain unchanged.
The C24 rate in PMID:8387517 is a real reported preparation-level result, independently read by two reviewers, but it is not securely attributable to ACOX2 itself. Table III's 0.14 versus 59.4 nmol H2O2/min/ml (0.2%) compares substrates within a partially purified preparation. The experiment does not by itself discriminate a small intrinsic side activity from residual activity of another oxidase. Residual contamination is a possibility, not an observed explanation. The VLCFA-process IBA is therefore UNDECIDED. The generic acyl-CoA reaction quote was removed from that row because it does not establish C24 activity. The purified/recombinant human panel in PMID:29287774 did not test C24/C26, so it cannot close this question. Neither normal patient VLCFA concentrations nor the total carbon count of steroid substrates is negative evidence for intrinsic VLCFA activity.
The current GO:0033791 definition was rechecked: its formal reaction and RHEA:15733/EC:1.17.99.3 cross-references specify hydroxylation, while its synonyms include oxidase names. The chemistry attributed to ACOX2 by the human recombinant study and UniProt RHEA:46728 is enoyl formation with oxygen and peroxide. The three rows now consistently use MODIFY to the verified oxidase term GO:0003997. This is a target-level reaction-term adjudication supported independently of the unresolved rat IBD assay; it is not a conclusion that all evidence codes must always have matching actions. Their primary-source limits remain distinct: human PMID:27884763 is a whole-cell bile-acid endpoint; the rat IBD chemistry is uninspected; rabbit PMID:9218493 directly identifies the enoyl product and offers endogenous hydratase as a probable explanation of the minor hydroxylated product. The human target's contribution to IBD is legitimate, not circular.
The loss of substrate precision in the broad replacement is explicitly recorded as an ontology question. GO:0003997 already occurs in the review; the replacement repairs the chemical assertion rather than adding functional coverage. No fabricated THCA-CoA-specific term is introduced, and no exact hydroxylation assertion is retained solely to preserve an informative label. The THCA-CoA substrate scope remains explicit in evidence and core prose.
The core now separates two substrate scopes: C27 bile-acid side-chain oxidation under the verified generic oxidase MF, and pristanoyl-CoA oxidation GO:0016402. The first core is specifically scoped to the bile-acid reaction, not an extra umbrella duplicating the pristanoyl core. The branched-chain core retains the precise oxidase-dependent beta-oxidation process GO:0033540. The broader GO:0006635 annotation remains ACCEPT in the original set and is covered by this more specific core process; a redundant core-process parent was not added solely for repetition.
All YAML aliases were expanded to literal evidence entries without changing their content. The rabbit reference title now correctly joins cholestanoyl without an inserted space. The IMP row again includes the measured family C27/C24 phenotype alongside the mutant/WT cellular comparison, and the bile-acid row includes the source's pathway statement. Mutation-versus-WT cellular experiments were not reclassified as inherently incapable of supporting IMP. Core prose now describes biology rather than curation scope; EC 1.3.3.6 and the UniProt-supported heart/liver/kidney expression summary were restored to the standalone description.
The reference reviews were updated to match these judgments, especially the preparation-level C24 evidence. Local full-text flags remain true for unavailable/abstract-only caches even when an external full paper was read. All six notes-inclusive missing PMIDs remain a hard draft gate: 16672280, 27647924, 8026493, 8387517, 8654595, and 9218493. The PR must remain draft and must not merge until normal source recovery. The reference requirement is not waived by manual primary reading, and no cache/provider artifact was fabricated. The review reply is prepared for coordinator publication, not sent by this author.
CI for head f47396930afe7737f2eb0149e491b44ca529fb8a reported a PMID:9218493 title mismatch after the review-requested word-join correction. An independent lookup of the official PubMed-indexed record confirms that its title itself contains chole stanoyl-CoA; its PMID, authors, year, journal and DOI 10.1074/jbc.272.29.18481 match the intended rabbit paper. Direct page opening encountered a browser challenge, but the indexed primary record exposed the title and abstract. This supersedes the earlier description of the space as a YAML-folding defect. The reference title now preserves that source spelling verbatim, with an explicit reference-review explanation. Scientific prose continues to use cholestanoyl-CoA. No source annotation, action, core function or supporting quotation changes. The six required missing caches and draft status remain unchanged.
All 12 canonical artifacts and four published history records match main
d2d8c9043b082a62378eff620ec0122d4118173b; none of the previous histories or
provider artifacts was edited. The six prior review/notes cache gates are now
recovered: PMID:16672280, PMID:27647924, PMID:8026493, PMID:8387517,
PMID:8654595 and PMID:9218493. Each publication file is byte-identical to the
normal-fetch Actions artifact receipt in
tmp/verified-reference-records/local-import-receipt.json: source run
36286975328, head 5946477c8ac79ade0709264c775ea1262b108438, artifact
10920674630, ZIP SHA-256
c0ffe4a66b80278af34b44aab6a3ae354ffd5699236b3a486ca95527be5e9713.
No source text was reconstructed or normalized.
All six recovered records are abstract-only. The actual recovered abstracts
were read; recovery does not imply new full-text verification. The rabbit record's
exact title contains the literal word break chole stanoyl, which remains
unchanged. The human purification abstract agrees with the original full-paper
assessment but does not itself resolve the weak C24 preparation result or native
assembly discrepancy. Rat donor abstracts describe CoA-ester specificity and
purification, without resolving the exact free-fatty-acid-binding assertion. The
excluded ACO-II structure remains rat ACOX1 evidence, not human ACOX2 evidence.
The disease abstract supports the historical p.Y69* context. No annotation
action, source object, core function or original reference identity changed.
PMID:33961781 full_text_unavailable is corrected to false because the existing
cache metadata records full text; its incomplete extraction and unresolved
individual STRN3 pair remain explicit in the reference review. The recovered
PMID:9218493 and PMID:8387517 availability notes now distinguish their actual
abstract caches from the prior externally read full sources.
A broader recursive scan found two additional missing citations only in the
immutable pre-existing openscientist hypothesis report and its companion output:
PMID:7929456 and PMID:15769750. Neither is a reference in the current biological
YAML, and neither is used to revise this review. Their artifact-only scope is
recorded separately from the six now-closed review/notes gates. Normal retrieval
was attempted once; the outcome and final validation are recorded below.
The normal fetch of those two artifact citations failed DNS, producing zero of
two files (/tmp/ACOX2-artifact-only-fetch.log). Targeted gene validation passes
with zero warnings because all current YAML references are cached and match
exact generated titles. Status nevertheless remains DRAFT while the two broader
citation gaps remain; passing YAML checks is not a claim of directory-wide cache
completion. The final manifest records all 40 unchanged source assertions and
annotation reviews, both unchanged cores, 22 unchanged reference identities,
all protected artifacts and all four append-only published histories.
This bounded follow-up starts from PR #3248 head
f0fa38d08c79573176858f128210e75e2fe7adba. The published YAML, notes,
render, six previously recovered publication records, protected source artifacts,
and all five existing history records matched their recorded hashes before edits.
The remaining two hypothesis-artifact citations now have real normal-fetch caches.
Both canonical files are byte-identical to the source3 import receipt
tmp/source3-canonical-import-receipt.json: Actions run 36292249952, source head
3f234ebd4b540057fb287b27efc11341cfb102b1, artifact 10924402765, ZIP SHA-256
d8908403ad407495b26d23af7007ee4da1fc643ebd61728d99ce9755bcfacdc2.
Neither record existed at the exact PR base. Their source bytes and metadata remain
unchanged; both records are abstract-only.
The recovered PMID:7929456(https://pubmed.ncbi.nlm.nih.gov/7929456/) abstract
separates the first oxidase step from the third dehydrogenase step of peroxisomal
beta-oxidation. Its experiments purify rat-liver 3-hydroxyacyl-CoA dehydrogenases;
this is pathway context, not a direct human ACOX2 substrate or product assay.
The recovered PMID:15769750(https://pubmed.ncbi.nlm.nih.gov/15769750/) abstract
examines L- and D-bifunctional-protein knockout mice and identifies DBP with the
second and third steps. It explicitly reports an oxidase import defect in the
double knockout, which prevents a firm conclusion about the proposed alternative
LBP pathway. That confound is retained when interpreting the study. The primary
PubMed-indexed identifiers, titles and abstracts were also checked independently.
No full-paper access is claimed for either recovered record.
These sources support pathway-step distinctions without turning the immutable
provider report's stronger verdicts into independent human reaction measurements.
The review's existing human and rabbit chemistry assessments, source-specific
uncertainties, all 40 annotation decisions, both cores, 22 reference identities and
all reference assessments remain unchanged. The two artifact sources are not added
to the top-level reference list because they do not supply a new review assertion.
The final census covers the review YAML, append-only notes, primary-source notes,
provider Markdown and citation list, hypothesis CSV/HTML/PDF, and rendered review.
All 15 distinct cited PMIDs and five cited Reactome records are cached. The PDF's
explicit citations were checked through text extraction and introduce no extra
source gap. Unused bibliography and pathway cross-references inside the immutable
UniProt record were inventoried separately; they are not authored or provider
citations. Historical cache-failure entries above remain as provenance, superseded
by this successful recovery. Status is now COMPLETE, subject to the final targeted
validator and history checks recorded with the frozen manifest. No biological
assertion, quotation or original source field changed.
Final targeted gene validation passed with zero curation warnings; the emitted
pkg_resources deprecation notice is a runtime dependency warning. History
validation and rendering also passed. The parsed YAML delta is exactly the status
change; there are no annotation, reference, core, description, isoform or quotation
changes. The incremental six-file manifest records the two exact new source
records, YAML, notes, rendered HTML and the newly scaffolded history.
The human index report PMID:27647924
now explicitly supports the bile-acid annotation and core alongside the independent
R225W study. Its cached abstract describes p.Y69, absent liver ACOX2 staining,
and accumulated upstream bile-acid intermediates. This corroborates pathway
participation and disease involvement without being an isolated-enzyme chemistry
assay. The earlier disease paragraph's deletion wording is corrected: p.Y69 is a
premature stop, and UniProt's missing residues 69–681 describe its consequence.
The rat structural study PMID:16672280
and original structure 2DDH provide positive
family-level context: rat ACOX1/ACO-II binds a free C12 fatty acid after hydrolysis
of the cocrystallized CoA ester. That evidence is now stated directly in the
fatty-acid-binding IBA rationale. It does not resolve the uninspected Acox2/Acox3
experimental descendants behind the target's PTN000097533 assertion, so its
UNDECIDED judgment remains. Both newly listed references remain marked as
abstract-only locally; no recovered full experiment is implied.
Historical local receipt paths above are working audit aids, not tracked public
source links. The durable recovery records are source1 run36286975328
and source3 run36292249952,
with their source commits, artifact identifiers and archive hashes already recorded
above. Existing journal entries remain as provenance. This follow-up adds two
reference assessments and corroborating quotes; all 40 source assertions/actions,
both core functions' biological scopes and all 22 original reference identities
remain unchanged. The complete citation census remains 15 cached PMIDs and five
cached Reactome records: these two papers were already counted in the notes.
PMID:8654595(https://pubmed.ncbi.nlm.nih.gov/8654595/) is now an explicit reference on the three hydroxylase-to-oxidase refinements and the pristanoyl core. Its independently verified PubMed abstract reports rat-liver oxidases acting only on the S-isomer of 2-methylpentadecanoyl-CoA. It also reports desaturation of one THCA-CoA epimer, but the accessible abstract does not name that epimer or the enoyl geometry. The exact human 25S-to-24E statement is therefore supported separately by the cached UniProt RHEA:46728 reaction, now quoted in full; the existing pristanoyl reaction quote remains. This distinguishes the primary comparative stereoselectivity experiment from the curated human reaction, without changing any annotation action or core definition. Processing of both THCA-CoA epimers by whole peroxisomes supports an upstream racemase requirement, not racemase activity of ACOX2. The existing experiment proposal already calls for purified human epimer-specific kinetics and product identification, so no redundant process or research question is added.
The live RCSB 2DDH entry, read at its Literature and Macromolecules panels, links PMID:16672280, rat P07872/ACOX1, an E. coli expression host and ligand HXD, (3R)-3-hydroxydodecanoic acid. The abstract describes accommodation of a C12-derived acid after cleavage of cocrystallized dodecanoyl-CoA. This is not an assay supplying free acid as the binding input. The human fatty-acid-binding IBA remains UNDECIDED; the rat ACOX1 comparator is explained in its reason and reference assessment rather than attached as direct target support. Neither the abstract-only cache flag nor the unresolved Acox2/Acox3 ancestral-assay linkage changes.
Current status supersedes the earlier validation snapshot: all three GO:0033791 rows use MODIFY, and PMID:9218493/PMID:8387517 are now normally cached. The earlier mixed-action and missing-cache statements remain dated journal history. The live GO:0033791 definition, rechecked here, still specifies 25R hydroxylation with water and an acceptor; it does not specify the oxygen-dependent enoyl-forming reaction. The 40 source assertions and decisions, two core definitions, 24 prior reference identities and raw source files remain unchanged; the reference list now has 26 entries, including the explicit notes receipt described below.
These receipts were rechecked against publicly accessible original articles. They preserve access to the original evidence; the notes are not an independent study. The normal publication records remain abstract-only.
The three public passages above were re-read at their original Table III, Discussion and Results locations. Their annotation evidence objects now name the primary PMIDs directly. The normal caches remain abstract-only, so those objects record the access limitation without a cached quote. The excerpts above remain external-reading receipts, not independent evidence or a substitute for validation against an unmodified primary cache. This supersedes the earlier routing of these quotes through the notes file. The schema reserves supporting_text_fulltext for text that cannot be shared publicly; these public originals do not establish that condition.
The 2DDH deposition again confirms the rat ACOX1/P07872 construct, E. coli expression host and primary citation PMID:16672280. Its HXD chemical component is (3R)-3-hydroxydodecanoic acid, C12H24O3. The paper's abstract separately attributes the retained C12 fatty-acid moiety to thioester hydrolysis of dodecanoyl-CoA. Hydrolysis alone would give dodecanoic acid, not introduce a 3-hydroxyl group. The inspected abstract and deposition therefore do not resolve the hydroxyl's origin; no additional in-crystal chemistry is inferred. This clarification supersedes the earlier wording that joined the two accounts without identifying that gap. The observed rat ligand remains family-level accommodation evidence, not a direct human ACOX2 free-acid binding assay.