Current review (2026-11-11): DRAFT; 22 ACCEPT, 6 KEEP_AS_NON_CORE, 7 MODIFY and 3 UNDECIDED across 38 contexts; zero NEW or PENDING. Earlier sections below are historical snapshots. The YAML and latest follow-up record the current decisions.
HAAO encodes 3-hydroxyanthranilate 3,4-dioxygenase (3-HAO; also "3-hydroxyanthranilic acid dioxygenase / oxygenase"; EC 1.13.11.6), a 286-aa cytosolic enzyme of the kynurenine pathway of tryptophan degradation.
The product ACMS (2-amino-3-carboxymuconate-6-semialdehyde) cyclises non-enzymatically to quinolinate, the universal precursor of de novo NAD+ biosynthesis (via QPRT), unless it is diverted by ACMSD toward picolinate/glutaryl-CoA. Thus HAAO sits directly upstream of quinolinate and de novo NAD+ from tryptophan.
PMID:7514594 (Malherbe et al. 1994, JBC) — molecular cloning and functional expression of human 3-HAO; recombinant enzyme active; "The Km value of 3-HANA for recombinant h3-HAO (approximately 2 microM)"; catalyzes "the synthesis of QUIN from 3-hydroxyanthranilic acid" PMID:7514594. Supports catalytic MF (GO:0000334), quinolinate biosynthesis (GO:0019805), and cytosolic location (basis for UniProt's ECO:0000305 cytosol call). Abstract-only in cache (full_text_available: false); curator (UniProtKB) read the full paper.
PMID:12007609 (Calderone et al. 2002, BBA) — E. coli-expressed purified human 3-HAO; "3-hydroxyanthranilic acid oxygenase (3-HAO) catalyses the conversion of 3-hydroxyanthranilic acid to quinolinic acid"; "enzymatic activity which can occur only in the presence of Fe(II); several other metals have been tested but in no case the formation of the product has been observed"; "two of the ions tested inhibit the catalytic reaction and one of them, Zn2+, could be of physiological relevance" PMID:12007609. Supports catalytic MF (GO:0000334), Fe(II)/ferrous iron binding (GO:0008198), and Zn2+ inhibition (basis of the GOA "response to zinc ion" and "response to cadmium ion" IDA calls — see caveats below).
PMID:28375145 (Pidugu et al. 2017, Acta Cryst D) — first crystal structures of human 3HAO with native iron and with zinc (an inhibitor); "3HAO is a non-heme iron-containing, ring-cleaving extradiol dioxygenase that catalyzes the addition of both atoms of O2 to ... 3-hydroxyanthranilic acid (3-HANA) to form quinolinic acid (QUIN)" PMID:28375145. Supports ferrous iron binding (GO:0008198, IDA) and the extradiol dioxygenase mechanism. Active-site Fe residues His47/His91/Glu53 (BINDING features) [file:human/HAAO/HAAO-uniprot.txt].
PMID:28792876 (Shi/Dunwoodie et al. 2017, NEJM) — biallelic loss-of-function HAAO (and KYNU) variants cause a congenital NAD-deficiency malformation syndrome (VCRL1). "Variants were identified in two genes that encode enzymes of the kynurenine pathway, 3-hydroxyanthranilic acid 3,4-dioxygenase (HAAO) and kynureninase (KYNU)"; "The mutant enzymes had greatly reduced activity in vitro"; "NAD is synthesized de novo from tryptophan through the kynurenine pathway"; "The patients had reduced levels of circulating NAD"; Haao-null and Kynu-null mice reproduce the defects; "Niacin supplementation during gestation prevented the malformations in mice" PMID:28792876. Supports catalytic MF IDA (in-vitro enzyme assay of variants), quinolinate biosynthesis IDA, and NAD+ biosynthetic process IMP. UniProt DISEASE: "Vertebral, cardiac, renal, and limb defects syndrome 1 (VCRL1) [MIM:617660]" [file:human/HAAO/HAAO-uniprot.txt].
PMID:2967497 (Schwarcz et al. 1988, PNAS) — 3-hydroxyanthranilate oxygenase activity detected in human brain and increased in Huntington disease striatum: "The activity of 3-hydroxyanthranilate oxygenase is increased in Huntington disease brains as compared to control brains" PMID:2967497. This is the basis of the GOA IMP "neuron cellular homeostasis" (GO:0070050) call — a correlational disease-association observation, not a demonstration that HAAO participates in maintaining neuronal homeostasis. Weak support for that specific BP term; kept as non-core.
All five bare "protein binding" (GO:0005515) IPI annotations come from high-throughput interactome/screen datasets, per GOA WITH/FROM and UniProt INTERACTION block (partners GAD1 = Q99259/Q8IVA8, POT1 = Q9NUX5):
- PMID:16189514 — Rual et al. 2005, first CCSB Y2H human interactome map (CCSB-HI1) PMID:16189514.
- PMID:21044950 — Lee et al. 2011, genome-wide split-Venus BiFC telomere-interactome screen; POT1 (Q9NUX5) among six telomere baits PMID:21044950.
- PMID:25416956 — Rolland et al. 2014, HI-II-14 systematic binary interactome map PMID:25416956.
- PMID:31515488 — Fragoza et al. 2019, ExAC-variant Y2H interaction screen PMID:31515488.
- PMID:32296183 — Luck et al. 2020, HuRI reference binary interactome PMID:32296183.
None assigns a specific molecular function; the GAD1/POT1 pairings have no established biological rationale for a cytosolic kynurenine-pathway enzyme. Per policy, bare "protein binding" IPI is not removed but marked as over-annotated (uninformative for MF).
GO:0010043 (response to zinc ion) and GO:0046686 (response to cadmium ion) IDA both trace to PMID:12007609, which showed that Zn2+ (and another tested ion) inhibit the enzyme in vitro. This is an in-vitro inhibitor-sensitivity result, not evidence that HAAO participates in a cellular "response to zinc/cadmium ion" process. These are over-interpretations of an inhibition assay; kept but flagged as over-annotated (Zn) / mark-as-over-annotated (Cd).
GO:0009055 (electron transfer activity) NAS is mechanistically wrong: 3-HAO is a non-heme Fe(II) dioxygenase that incorporates both O2 atoms into the substrate (extradiol ring cleavage) — it is not an electron carrier/electron transfer protein. UniProt itself retains the NAS DR line but the assignment does not reflect the current mechanistic understanding (PMID:28375145 "ring-cleaving extradiol dioxygenase") PMID:28375145. NAS (author statement) — mark as over-annotated rather than accept.
Disease: biallelic LoF causes congenital NAD-deficiency malformation syndrome (VCRL1), rescuable in mice by gestational niacin (PMID:28792876).
This historical snapshot superseded the earlier annotation-disposition commentary above. At that stage the DRAFT review recorded: 26 ACCEPT, 6 KEEP_AS_NON_CORE, 2 MODIFY and 4 UNDECIDED judgments across 38 stored contexts, with no PENDING or NEW annotations. The authored source guide gives source-access limits and precise evidence locations; it is not a provider report. Falcon failed with HTTP 403 and Perplexity-lite with HTTP 401/insufficient quota. Existing publication caches and all 12 inherited supporting quotations are preserved.
The native fetcher archived 35 current QuickGO records and produced the corresponding 35-row TSV. The historical and current 35-context inventories share 32, yielding 38 retained contexts. Native seeding restored the second donor context from PMID:25416956 and added three updated electronic mappings; three older mappings remain explicit historical contexts with retired: true lifecycle flags. See the context audit and literal archive. No extension, NOT or isoform-qualified assertion is introduced.
The core is consolidated into one Fe(II)-dependent dioxygenase activity. HAAO produces ACMS; subsequent cyclization to quinolinate is spontaneous. Curated cytosolic localization is accepted with explicit source limits. The Fe(III) state of the air-prepared human crystal is distinguished from physiological Fe(II) [PMID:28375145, original Methods §2.2 and Results §3.3; PMID:12007609, abstract]. Human disease evidence and mouse niacin rescue are kept distinct PMID:28792876.
Five HAAO–GAD1 contexts are supported by exact IntAct/IMEx records and retained as non-core under the ClinGen project instruction. Both Q8IVA8 and Q99259 contexts from PMID:25416956 are retained. HAAO–POT1 remains UNDECIDED because the paper-specific target/supplementary evidence was not recovered, despite the positive UniProt interaction entry. Missing returned records are not treated as evidence against that interaction.
The older electron-carrier-only objection to GO:0009055 is withdrawn after checking its current definition, which includes enzymatic redox reactions. The broad annotation is non-core. Iron/metal binding is refined to ferrous iron binding. Zinc response, cadmium response and neuron cellular homeostasis remain UNDECIDED rather than being rejected from abstract-only evidence [PMID:12007609; PMID:2967497].
GO:0034354 has a QuickGO/OLS version discrepancy. Its incoming IDs and stored labels are retained; acceptance concerns the supported biological pathway claim, while the authored core uses active terms. The source-evidence record contains the service checks. An inherited PTHR15497/BNA1 prose error misidentifies P46952 as bacterial NbaC; this reassessment uses the literal PAINT donors and verified human identity and does not borrow that label.
This follow-up supersedes the earlier POT1 uncertainty. Updated IntAct evidence resolves target verification; the source guide records the access limits.
The existing binding context is now KEEP_AS_NON_CORE under the project instruction. That follow-up recorded 26 ACCEPT, 7 KEEP_AS_NON_CORE, 2 MODIFY and 3 UNDECIDED across the same 38 stored contexts, with zero NEW or PENDING. The catalytic core, all machine source fields, all 12 inherited quotations and all publication caches remain unchanged; status remains DRAFT.
Two cytoplasm rows now propose cytosol, preserving their distinct source qualifiers. Two GO:0034354 rows propose active NAD+ biosynthetic process GO:0009435 while retaining the incoming IDs and the documented ontology-version discrepancy. Electron transfer activity now proposes the demonstrated dioxygenase activity as a mechanistic replacement; no ontology descendant relationship is asserted. Two short, exact ontology-definition excerpts were added, while all 12 inherited quotations remain unchanged.
The three experimental process judgments remain UNDECIDED. The inspected abstracts establish inhibition or correlation, but do not verify the full process assertions or exhaust the inaccessible original methods. This follows the standing incomplete-evidence instruction. Correctness of neuronal homeostasis is not asserted by keeping its evidence unresolved.
The source guide is retained as a locator guide in these notes; per-annotation citations now point to the underlying literature, curated records and literal interaction archive. Interaction-paper relevance is LOW relative to the established catalytic function. Earlier history records and their then-current counts remain unchanged; a new history entry captures this follow-up.