ADTRP (androgen-dependent TFPI-regulating protein) is a six-pass integral membrane enzyme of the ADTRP/AIG1 family that hydrolyses fatty acid esters of hydroxy fatty acids (FAHFAs), a class of endogenous bioactive signalling lipids with anti-inflammatory and insulin-sensitising properties. It is an atypical hydrolase: it has no sequence or structural relationship to the classical serine-hydrolase folds, and instead uses a threonine nucleophile (Thr47) activated by a histidine general base (His131), both of which sit within transmembrane helices rather than in a soluble catalytic domain. Substitution of either residue abolishes FAHFA hydrolysis, and the enzyme is inactive against phospholipids, lysophospholipids and mono-, di- and triglycerides, preferring FAHFA isomers branched distal to the carboxylate head group. Together with its paralogue AIG1 it is one of the two principal FAHFA hydrolases in mammals; knockout mice accumulate FAHFAs, particularly the 9-position ester isomers, with other lipid classes unchanged. ADTRP is expressed in vascular endothelium and placenta, is transcriptionally induced by androgen, and localises to the plasma membrane, concentrating with tissue factor pathway inhibitor (TFPI) and caveolin-1 in lipid rafts and caveolae. In endothelial cells it raises TFPI expression and the cell-associated anticoagulant activity of TFPI against factor Xa, and its depletion increases monocyte adhesion and transendothelial migration through a PIK3R3-AKT-MIA3 axis. The gene lies at a coronary-artery-disease GWAS locus. How the lipid-hydrolase activity relates to the endothelial TFPI and vascular phenotypes is not established.
Definition: The chemical reactions and pathways resulting in the breakdown of a fatty acid ester of a hydroxy fatty acid (FAHFA), a class of endogenous bioactive signalling lipids in which a fatty acid is esterified to a hydroxyl group on a second fatty acid backbone.
Justification: GO gained the molecular function GO:0120573 FAHFA hydrolase activity on 2026-03-14 but has no biological-process counterpart; a QuickGO text search for 'FAHFA' returns GO:0120573 and nothing else. The nearest available process term, GO:0042758 long-chain fatty acid catabolic process, is chemically true for ADTRP because ChEBI classifies the substrate 9-PAHSA(1-) (CHEBI:83670) as a long-chain fatty acid anion, but it is far broader than the demonstrated specificity and it is the term PAINT is currently propagating from a pan-eukaryotic node to 86 gene products including 14 fungi and 5 plants. A FAHFA-level process term would let PAINT and manual curators state the specific claim where it is supported and withhold it where it is not. GO may reasonably decline this on the grounds that a single-step hydrolysis is adequately covered by the molecular function; that is recorded here as the counter-argument rather than suppressed.
Parent term: long-chain fatty acid catabolic process
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0016787 hydrolase activity | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Correct but maximally general: the root of the hydrolase branch, propagated from a pan-eukaryotic node whose only characterised members are ADTRP and AIG1. True of this gene, but subsumed by the experimentally grounded GO:0120573. Reason: The IBD node PANTHER:PTN001659973 is scoped by PAINT to taxon:2759 (Eukaryota) and seeded by exactly two proteins, human ADTRP (Q96IZ2, the target itself) and human AIG1 (Q9NVV5). It reaches 86 gene products: 40 Vertebrata, 25 invertebrate Metazoa, 14 Fungi, 5 Viridiplantae and 2 other Eukaryota. Only 7 of the 86 are Swiss-Prot reviewed, and two of those are uncharacterised UPF0641 fungal proteins (P38842 YHR140W, Q96WV4 SPBPJ4664.05). The character that is conserved across the node is the catalytic dyad: aligning all 85 other recipients to ADTRP and requiring the aligned column to land on ADTRP's own annotated SITE positions, 73 of 85 retain Thr47/His131, including 39/39 vertebrates and both Dictyostelium members (positive control: AIG1 scores dyad-intact at 36.5% identity). All 14 fungal recipients fall below 25% identity, so their dyad status is undetermined rather than negative. GO:0016787 is a verified is_a ancestor of GO:0120573 FAHFA hydrolase activity, which this gene already holds by IMP (mutagenesis of both catalytic residues) and by IEA from 11 RHEA reactions, so the IBA adds no information about ADTRP. It is kept rather than removed because at the node the general term is correctly scoped to the evidence: what is conserved family-wide is the Thr/His catalytic dyad, and a conserved dyad licenses a claim about catalytic MECHANISM but not about SUBSTRATE. GO:0016787 states mechanism only, so it is exactly what the residue conservation supports; the substrate is known only for the four characterised animal members. PAINT placed this term on 2026-05-28, ten weeks after GO:0120573 was created (2026-03-14), so the general term is a deliberate judgement and not a stale-term artefact. The action is KEEP_AS_NON_CORE rather than ACCEPT only because it should not be counted as this gene's core molecular function. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN001659973 · PANTHER IBD node, taxon:2759 Eukaryota, 86 recipients SUPPORTS TRANSFER The conserved character across the node is the Thr/His catalytic dyad (73/85 recipients), which supports a mechanism-level term such as this one but not a substrate-level one. UniProtKB:Q96IZ2 · human ADTRP (the target itself) SUPPORTS TRANSFER Self-referential IBD seed: a PAINT curator judged the hydrolase function to be a family-level character. Not circular. UniProtKB:Q9NVV5 · human AIG1, reviewed (Swiss-Prot), catalytic Thr43/His134 SUPPORTS TRANSFER Genuine co-seed characterised in the same paper; carries its own IMP to GO:0120573 and GO:0042758. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md | `PTN001659973` | taxon:2759 Eukaryota | GO:0016787, GO:0042758 | 86 | PMID:27018888 We have discovered herein using ABPP that the poorly characterized multipass transmembrane proteins AIG1 and ADTRP represent a new family of hydrolytic enzymes that degrade the FAHFA class of signaling lipids. |
| GO:0005901 caveola | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Propagated from a node whose reach is exactly the vertebrate ADTRP orthologue set and whose sole seed is this gene's own IDA. Taxon-consistent and not over-reaching, but the relation was upgraded from located_in to is_active_in without any assay placing the FAHFA hydrolase activity in caveolae. Reason: PANTHER:PTN002591065 is scoped taxon:117571 (Euteleostomi), seeded by UniProtKB:Q96IZ2 alone, and reaches 25 gene products which resolve to ADTRP orthologues in 25 vertebrate species and nothing else; the paralog AIG1 is correctly excluded. Caveolae are a vertebrate structure requiring caveolins, so the node's taxon scope and the term agree - this is the benign answer to the reciprocal 'which node's reach is exactly my gene set' question. The reservation is the relation, not the compartment: the seeding evidence is a located_in IDA (colocalisation with TFPI and caveolin-1 in endothelial lipid rafts), whereas the IBA asserts is_active_in, i.e. that the molecular function executes there. ADTRP's catalytic activity was measured in bulk membrane lysates of transfected HEK293T cells, which carry no caveola-specific information, and the caveola observation comes from a TFPI-regulation study that did not assay hydrolase activity. Kept as non-core: the location is real and experimentally supported, but caveolar residence is a property of the endothelial TFPI role rather than a demonstrated requirement of FAHFA hydrolysis. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN002591065 · PANTHER IBD node, taxon:117571 Euteleostomi, 25 recipients, all ADTRP orthologues SUPPORTS TRANSFER Node reach matches the term's taxon requirement exactly; no invertebrate or fungal recipient receives a caveola annotation. UniProtKB:Q96IZ2 · human ADTRP (the target itself) SUPPORTS TRANSFER Self-referential IBD seed backed by this gene's own IDA from PMID:21868574. The seed's evidence is located_in; the propagation states is_active_in. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md | `PTN002591065` | taxon:117571 Euteleostomi | GO:0005901 | 25 | Vertebrata 25 | 3 | PMID:21868574 Imaging and Triton X-114-extraction confirm TFPI and ADTRP association with lipid rafts/caveolae. |
| GO:0042758 long-chain fatty acid catabolic process | IBA GO_REF:0000033 | ACCEPT | Summary: True for human ADTRP, which holds the identical term by IMP from the same characterisation paper. Flagged upstream: the node placing it spans Eukaryota, so 84 of its 86 recipients receive a specific lipid-catabolism claim with no supporting data. Reason: For this gene the term is right, and the substrate chemistry was checked rather than assumed: ChEBI classifies the substrate 9-PAHSA(1-) (CHEBI:83670) as a long-chain fatty acid anion, so hydrolysing a FAHFA is literally the breakdown of a long-chain fatty acid, and UniProt gives PhysiologicalDirection=left-to-right for all 12 catalytic-activity lines, i.e. the annotated direction is removal, not addition. The upstream concern does not change this gene's verdict but is recorded in suggested_questions: the same node that carries this four-step-deep biological process declines to give anything more specific than GO:0016787 for the molecular function. The IBD node PANTHER:PTN001659973 is scoped by PAINT to taxon:2759 (Eukaryota) and seeded by exactly two proteins, human ADTRP (Q96IZ2, the target itself) and human AIG1 (Q9NVV5). It reaches 86 gene products: 40 Vertebrata, 25 invertebrate Metazoa, 14 Fungi, 5 Viridiplantae and 2 other Eukaryota. Only 7 of the 86 are Swiss-Prot reviewed, and two of those are uncharacterised UPF0641 fungal proteins (P38842 YHR140W, Q96WV4 SPBPJ4664.05). The character that is conserved across the node is the catalytic dyad: aligning all 85 other recipients to ADTRP and requiring the aligned column to land on ADTRP's own annotated SITE positions, 73 of 85 retain Thr47/His131, including 39/39 vertebrates and both Dictyostelium members (positive control: AIG1 scores dyad-intact at 36.5% identity). All 14 fungal recipients fall below 25% identity, so their dyad status is undetermined rather than negative. The asymmetry is a category distinction, not a curation lapse: the conserved dyad licenses the mechanism-level MF the node carries, whereas this BP asserts a SUBSTRATE, and the substrate is established only for the four characterised animal members. So the node propagates a substrate claim on evidence that can only support a mechanism claim. Note this corrects a first-pass reading of mine, which had argued from clade heterogeneity instead; measurement showed the dyad is broadly conserved, so heterogeneity is the wrong basis for the objection. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN001659973 · PANTHER IBD node, taxon:2759 Eukaryota, 86 recipients SUPPORTS TRANSFER Supports the transfer to ADTRP, which has its own IMP, but the node's taxon scope carries a substrate-level claim to 46 non-vertebrate recipients whose substrate has never been determined; their catalytic dyad is largely conserved, which supports mechanism but not substrate. UniProtKB:Q96IZ2 · human ADTRP (the target itself) SUPPORTS TRANSFER Self-referential IBD seed; the target independently holds this term by IMP from PMID:27018888. UniProtKB:Q9NVV5 · human AIG1, reviewed (Swiss-Prot) SUPPORTS TRANSFER Carries its own IMP to GO:0042758 from the same paper. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md | `CHEBI:83670` | 9-PAHSA(1-) | yes | PMID:27018888 In contrast, both AIG1 and ADTRP-transfected cell membrane lysates robustly hydrolyzed several fatty-acid esters of hydroxy-fatty acids (FAHFAs) |
| GO:0005886 plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Swiss-Prot subcellular-location mapping of SL-0039 (Cell membrane), which UniProt asserts with experimental evidence. Consistent with the caveola and cell-surface IDAs from PMID:21868574. Reason: UniProt's SUBCELLULAR LOCATION line reads 'Cell membrane ... Multi-pass membrane protein', and the feature table gives six TRANSMEM helices with alternating cytoplasmic and extracellular topological domains, so a plasma-membrane call is well founded for this protein. The mapping route names its source explicitly (UniProtKB-SubCell:SL-0039) and is not an unattributed bulk import. Supporting Evidence: PMID:21868574 Imaging and Triton X-114-extraction confirm TFPI and ADTRP association with lipid rafts/caveolae. |
| GO:0016020 membrane | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro2GO from IPR006838. The signature is family-specific (ADTRP/AIG1), not a bare fold, and it maps only to a cellular component - the predicted fold-to-activity error is absent here. Reason: IPR006838 'ADTRP/AIG1' covers 5788 proteins of which 8 are reviewed (Swiss-Prot): four ADTRP orthologues, two AIG1 orthologues and two uncharacterised UPF0641 fungal proteins. interpro2go maps the entry to GO:0016020 membrane and to nothing else - no molecular function, despite the family now being a characterised hydrolase family. That is the correct restraint: a quarter of its own reviewed members are curated as uncharacterised, and the entry spans thousands of unstudied proteins. The location claim it does make is independently confirmed by an IDA from PMID:27018888. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md interpro2go maps it to GO:0016020 |
| GO:0120573 FAHFA hydrolase activity | IEA GO_REF:0000116 | ACCEPT | Summary: Rhea mapping from the 11 FAHFA hydrolysis reactions UniProt curates for this protein with experimental evidence. This is the gene's core molecular function, anchored to explicit reaction identifiers. Reason: The term is RHEA-anchored on both sides: GO:0120573 carries 12 RHEA cross-references and this annotation's WITH/FROM lists 11 of them (RHEA:52048/52052/52056/52060/52064/52068/52072/52076/52080/52084/52096), each curated in UniProt as ECO:0000269|PubMed:27018888 with PhysiologicalDirection=left-to-right. The direction and the substrate identity are therefore both explicit and both correct: the enzyme removes an acyl group by hydrolysis rather than forming one. GO:0120573 is current (not obsolete, no secondary ids) and was created 2026-03-14. Supporting Evidence: PMID:27018888 Both AIG1 and ADTRP hydrolyze bioactive fatty acid esters of hydroxy fatty acids (FAHFAs) but not other major classes of lipids. |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | MARK AS OVER ANNOTATED | Summary: Uninformative bare protein-binding row for TMED8 (Q6PL24), resting on one yeast two-hybrid screen logged as three sub-methods. Reason: Both GO:0005515 rows come from the single HuRI binary interactome screen. IntAct returns 10 interactions for Q96IZ2; the two HuRI partners are each logged three times, as 'two hybrid array', 'two hybrid prey pooling approach' and 'validated two hybrid' - three sub-methods of one experiment, which is what UniProt's NbExp=3 is counting. All are yeast two-hybrid in Saccharomyces cerevisiae with an MI-score of 0.56, and there is no orthogonal assay and no follow-up anywhere in the ADTRP literature. ADTRP is a six-pass integral membrane protein whose extracellular and cytoplasmic loops are 13-22 residues long, which makes it a poor Y2H substrate. The partner accession itself is sound - it resolves to the reviewed canonical Swiss-Prot entry with matching length, so this is not an ORFeome-clone substitution - but a single unreplicated screen hit is not an informative molecular function. Marked over-annotated rather than removed: nothing refutes the interaction, it is simply unsupported beyond one screen. Q6PL24 is reviewed TMED8_HUMAN, 325 aa, matching the canonical entry. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md | TMED8 | two hybrid array, two hybrid prey pooling approach, validated two hybrid | 3 | 1 | |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | MARK AS OVER ANNOTATED | Summary: Uninformative bare protein-binding row for CMTM7 (Q96FZ5), resting on one yeast two-hybrid screen logged as three sub-methods. Reason: Both GO:0005515 rows come from the single HuRI binary interactome screen. IntAct returns 10 interactions for Q96IZ2; the two HuRI partners are each logged three times, as 'two hybrid array', 'two hybrid prey pooling approach' and 'validated two hybrid' - three sub-methods of one experiment, which is what UniProt's NbExp=3 is counting. All are yeast two-hybrid in Saccharomyces cerevisiae with an MI-score of 0.56, and there is no orthogonal assay and no follow-up anywhere in the ADTRP literature. ADTRP is a six-pass integral membrane protein whose extracellular and cytoplasmic loops are 13-22 residues long, which makes it a poor Y2H substrate. The partner accession itself is sound - it resolves to the reviewed canonical Swiss-Prot entry with matching length, so this is not an ORFeome-clone substitution - but a single unreplicated screen hit is not an informative molecular function. Marked over-annotated rather than removed: nothing refutes the interaction, it is simply unsupported beyond one screen. Q96FZ5 is reviewed CKLF7_HUMAN, 175 aa, matching the canonical entry; it is itself a MARVEL-domain membrane protein. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md | CMTM7 | two hybrid array, two hybrid prey pooling approach, validated two hybrid | 3 | 1 | |
| GO:0120573 FAHFA hydrolase activity | IMP PMID:27018888 AIG1 and ADTRP are atypical integral membrane hydrolases tha... | ACCEPT | Summary: The gene's core molecular function, established by loss-of-activity mutagenesis of both catalytic residues against a panel of non-FAHFA lipid controls. Reason: ADTRP is a six-pass integral membrane threonine hydrolase that cleaves the ester bond of fatty acid esters of hydroxy fatty acids (FAHFAs). Catalysis requires Thr47 and His131, both of which UniProt records as SITE features with experimental evidence from PubMed:27018888 and both of which lie inside predicted transmembrane helices. The evidence is specific in both directions: transfected membrane lysates hydrolysed FAHFAs but showed negligible activity against (lyso)-phospholipids and mono-, di- and tri-glycerides, and both T47A and H131A abolished FAHFA hydrolysis. The term matches the measured reaction exactly and is anchored to RHEA, so no granularity question arises. IMP is the right code: the assignment rests on the mutant phenotype. Supporting Evidence: PMID:27018888 The FAHFA hydrolase activities of AIG1 and ADTRP were abolished by mutating their putative catalytic nucleophilic residues Thr-43 and Thr-47, respectively PMID:27018888 We also tested the H134A mutant of AIG1 and H131A mutant of ADTRP and found that these proteins showed no detectable FAHFA hydrolase activity above a mock-transfected control PMID:27018888 The membrane lysates of hAIG1- and hADTRP-transfected HEK293T cells showed negligible hydrolytic activity above a mock-transfected control proteome with the majority of tested lipid substrates, including common classes of (lyso)-phospholipids and neutral lipids PMID:27018888 AIG1 and ADTRP displayed a preference for FAHFAs with branching distal from the carboxylate head group of the lipids |
| GO:0051897 positive regulation of phosphatidylinositol 3-kinase/protein kinase B signal transduction | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: Downstream transcriptional consequence of ADTRP knockdown in endothelial cells, not a direct signalling activity of the protein. Reason: The mechanism reported is that ADTRP raises PIK3R3 transcript levels, which then activates AKT. ADTRP is a membrane lipid hydrolase with no kinase, adaptor or scaffold module, so its effect on PI3K/AKT is indirect and mediated by gene expression. involved_in is defensible for an siRNA phenotype and the IMP code is correct, but this is a pathway consequence rather than a core function. Note the cached record for this reference is abstract-only, so nothing here is asserted beyond what the abstract states. Supporting Evidence: PMID:28341552 ADTRP positively regulates expression of PIK3R3 encoding the regulatory subunit 3 of PI3K, which leads to activation of AKT, resulting in up-regulation of MIA3/TANGO1. |
| GO:0005886 plasma membrane | EXP PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | ACCEPT | Summary: Plasma-membrane localisation supported by this paper's imaging and detergent-partitioning of ADTRP in endothelial lipid rafts. Reason: The abstract reports imaging plus Triton X-114 extraction placing ADTRP with TFPI in lipid rafts/caveolae of endothelial cells, and caveola is a verified descendant of plasma membrane, so the compartment is properly supported by this reference. Supporting Evidence: PMID:21868574 We confirm ADTRP expression and colocalization with TFPI and caveolin-1 in ECs. |
| GO:0005886 plasma membrane | EXP PMID:27018888 AIG1 and ADTRP are atypical integral membrane hydrolases tha... | ACCEPT | Summary: Plasma-membrane localisation is correct for ADTRP and stands on PMID:21868574 and the SL-0039 mapping, so the term is accepted. Recorded caveat: this particular reference contains no localisation experiment and supports only GO:0016020 membrane, for which the same paper is separately cited - a UniProt evidence-attribution question, not a GO error. Reason: Accepted on the term, with a reference-attribution caveat rather than a GO action, because the plasma-membrane call stands independently on PMID:21868574 and on the SL-0039 mapping. The caveat: the cached full text of PMID:27018888 (full_text_available: true) contains no localisation experiment at all - zero occurrences of 'plasma membrane', 'cell surface', 'immunofluorescence', 'confocal', 'localization' - against positive controls in the same file of 'membrane lysates' 9, 'membrane fraction' 3, 'transmembrane' 29, 'HEK293T' 36 and 'FAHFA' 67, so the scan is working. What that paper shows is recovery of recombinant ADTRP in the membrane fraction of transfected HEK293T cells plus six topology predictors placing the catalytic residues in transmembrane helices; it supports 'membrane', not 'plasma membrane'. This row appears to descend from UniProt's SUBCELLULAR LOCATION line, which cites both PubMed:21868574 and PubMed:27018888 for 'Cell membrane'; GOA then split that into two EXP rows. Reported as a UniProt evidence-attribution question in suggested_questions. Supplementary figures are not in the cache, so the scan is scoped to the cached full text. Supporting Evidence: file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md no plasma-membrane or cell-surface localisation experiment in the cached full text PMID:27018888 we next analyzed the protein sequences for both AIG1 and ADTRP using six different transmembrane topology prediction programs (CCTOP, Phobius, PSORT II, TMHMM, TMpred, and Uniprot) and found that these programs consistently predicted that both the conserved Thr and His residues of AIG1 and ADTRP were located within transmembrane domains of these proteins |
| GO:0002042 cell migration involved in sprouting angiogenesis | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: Endothelial migration phenotype of ADTRP knockdown. Kept as non-core; note the abstract reports EC migration without using the words sprouting or angiogenesis. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. The abstract states that ADTRP knockdown 'inhibited EC proliferation and migration'. It does not mention sprouting or angiogenesis, and the full text is not cached, so the specific sprouting-angiogenesis context cannot be verified here; the BHF-UCL curator read the full text and the term is left in place. Either way this is a downstream cellular phenotype, not a core function of a lipid hydrolase. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:0002686 negative regulation of leukocyte migration | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: Monocyte transendothelial migration increases on ADTRP knockdown. Correct at the leukocyte level and the appropriate granularity for the data. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. Monocytes are leukocytes, so this term is satisfied by the reported phenotype and sits at a level the data supports. It is also the verified common ancestor covering the cell-type problem flagged on the GO:2000402 row below. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:0003332 negative regulation of extracellular matrix constituent secretion | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: ADTRP lowers collagen VII levels in endothelial and HepG2 cells. Indirect, via the MIA3/TANGO1 collagen-export pathway. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. Collagen VII is an extracellular matrix constituent and MIA3/TANGO1 is a collagen export factor, so the term matches the reported effect. The effect runs through ADTRP's regulation of MIA3/TANGO1 expression rather than any direct role in secretion. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:0050709 negative regulation of protein secretion | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: ADTRP lowers ApoB levels in endothelial and HepG2 cells; indirect, through the same MIA3/TANGO1 axis. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. Same evidence and same indirection as the extracellular-matrix row: ApoB is a secreted protein and MIA3/TANGO1 is the mediator. A regulatory consequence, not a core function. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:0140052 cellular response to oxidised low-density lipoprotein particle stimulus | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: The monocyte-adhesion phenotype was measured under oxidised-LDL stimulation. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. The abstract describes 'oxidized-LDL-mediated monocyte adhesion to ECs' being promoted by ADTRP knockdown, which places ADTRP within the cellular response to oxLDL. This is a context of the assay rather than a demonstrated ADTRP-dependent oxLDL-sensing step, so it is retained as non-core. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:1903038 negative regulation of leukocyte cell-cell adhesion | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: Monocyte adhesion to endothelial cells increases on ADTRP knockdown. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. Directly supported by the reported adhesion phenotype, and monocytes are leukocytes so the cell-type term is right here. Downstream of the PIK3R3/AKT/MIA3 axis rather than of hydrolase activity. Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis |
| GO:2000402 negative regulation of lymphocyte migration | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | MODIFY | Summary: Wrong leukocyte lineage, not merely imprecise. Every migration experiment described in the abstract used monocytes; lymphocyte and monocyte terms are disjoint siblings under negative regulation of leukocyte migration. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. Cell-type words in the cached abstract: 'monocyte' 4, 'lymphocyte' 0, 'leukocyte' 0. The phenotype is 'transendothelial migration of monocytes'. Both ancestor closures were fetched before calling this a granularity problem, and neither term contains the other: GO:2000402 (lymphocyte) is not an ancestor of GO:2000438 (monocyte extravasation) and GO:2000438 is not an ancestor of GO:2000402. Monocytes are myeloid mononuclear phagocytes and lymphocytes are lymphoid, so the annotated term names a different cell lineage from the one assayed - it is wrong rather than general. The proposed replacement GO:2000438 negative regulation of monocyte extravasation matches 'transendothelial migration of monocytes' precisely. If the uncached full text does contain a lymphocyte migration assay, the correct resolution is GO:0071676 negative regulation of mononuclear cell migration, a verified ancestor of both terms that would assert strictly less than either. Nothing is lost by the change in any case: the leukocyte-level claim is separately annotated from this same paper as GO:0002686. Proposed replacements: negative regulation of monocyte extravasation Supporting Evidence: PMID:28341552 Knockdown of ADTRP expression by siRNA promoted oxidized-LDL-mediated monocyte adhesion to ECs and transendothelial migration of monocytes, inhibited EC proliferation and migration, and increased apoptosis file:human/ADTRP/ADTRP-bioinformatics/RESULTS.md **Neither closure contains the other** |
| GO:0010628 positive regulation of gene expression | IMP PMID:28341552 Identification of a molecular signaling gene-gene regulatory... | KEEP AS NON CORE | Summary: ADTRP knockdown lowers PIK3R3 and MIA3/TANGO1 transcript levels. A real but very general and wholly indirect regulatory effect. Reason: One of nine biological-process rows curated by BHF-UCL from a single siRNA-knockdown study of ADTRP in endothelial cells. The cached record is abstract-only, so this review does not assert anything about the paper's figures. The term is the most general positive-regulation-of-expression node in GO and records nothing about which genes or by what mechanism. It is retained because the phenotype is real, but a membrane lipid hydrolase has no direct transcriptional activity and GO offers no target-specific child, so this cannot be a core function. Supporting Evidence: PMID:28341552 We showed that knockdown of ADTRP expression markedly down-regulated expression of MIA3/TANGO1. |
| GO:0016020 membrane | IDA PMID:27018888 AIG1 and ADTRP are atypical integral membrane hydrolases tha... | ACCEPT | Summary: Recombinant ADTRP partitions into the membrane fraction, consistent with six predicted transmembrane helices. This is the compartment claim this paper actually supports. Reason: All ADTRP activity assays in this paper were run on membrane lysates of transfected HEK293T cells, and the protein's catalytic Thr47 and His131 are predicted by six independent topology programs to lie within transmembrane helices. GO:0016020 is the right granularity for that evidence; the more specific plasma-membrane claim rests on PMID:21868574 instead. Supporting Evidence: PMID:27018888 we use activity-based profiling to discover that the poorly characterized multipass transmembrane proteins AIG1 and ADTRP are atypical hydrolytic enzymes that depend on conserved threonine and histidine residues for catalysis |
| GO:0042758 long-chain fatty acid catabolic process | IMP PMID:27018888 AIG1 and ADTRP are atypical integral membrane hydrolases tha... | ACCEPT | Summary: The biological-process expression of the gene's core activity. The substrate is itself a long-chain fatty acid by ChEBI, so the term names the right chemistry. Reason: 9-PAHSA(1-) (CHEBI:83670) is classified in ChEBI as a long-chain fatty acid anion, and ADTRP cleaves it into palmitate and 9-hydroxystearate, so FAHFA hydrolysis is the breakdown of a long-chain fatty acid and the direction is degradative. The IMP rests on the catalytic-residue mutants. The term is nonetheless broader than the measured specificity - GO has no FAHFA-level biological process, only the molecular function GO:0120573 - which is filed under proposed_new_terms. For ADTRP specifically the in vivo demonstration is PMID:32152231, where Adtrp-knockout mouse tissues accumulate FAHFAs with unchanged levels of other lipid classes. Supporting Evidence: PMID:32152231 Tissues from mice lacking ADTRP (Adtrp-KO), or both AIG1 and ADTRP (DKO) had higher concentrations of FAHFAs particularly isomers with the ester bond at the 9th carbon due to decreased FAHFA hydrolysis activity. PMID:32152231 The levels of other lipid classes were unaltered indicating that AIG1 and ADTRP specifically hydrolyze FAHFAs. |
| GO:0005901 caveola | IDA PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | KEEP AS NON CORE | Summary: Imaging plus Triton X-114 partitioning place ADTRP with TFPI and caveolin-1 in endothelial lipid rafts/caveolae. Solid evidence for a real location, kept as non-core because caveolar residence belongs to the endothelial TFPI role rather than to the catalytic function. Reason: Two independent methods in the same study support the compartment, and caveola is a verified descendant of plasma membrane so it is consistent with the other location rows. located_in is the right relation for a colocalisation result; see the GO:0005901 IBA row for why the is_active_in upgrade is treated more cautiously. Non-core rather than accepted-as-core so that this term matches the IBA row on the same term, and because the core_functions entry lists the plasma membrane and membrane as the locations of the FAHFA hydrolase activity: no assay has placed that activity in caveolae specifically, and the caveolar observation comes from a TFPI-regulation study in endothelial cells. Supporting Evidence: PMID:21868574 Imaging and Triton X-114-extraction confirm TFPI and ADTRP association with lipid rafts/caveolae. PMID:21868574 We confirm ADTRP expression and colocalization with TFPI and caveolin-1 in ECs. |
| GO:0009986 cell surface | IDA PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | KEEP AS NON CORE | Summary: Retained on curator authority. Plausible for a polytopic plasma-membrane protein with three extracellular loops, though the cached abstract does not state it. Non-core: unlike GO:0016020, cell surface is not an ancestor of plasma membrane but an independent claim, and the catalytic residues sit inside the bilayer rather than on the external face. Reason: GO:0009986 sits in its own branch: closures were fetched and it is not a descendant of GO:0005886 plasma membrane or of GO:0016020 membrane, so it is a distinct claim rather than a coarser version of the other location rows. The cached record for this reference is abstract-only and the abstract does not use the phrase 'cell surface'; the study does measure cell-associated TFPI activity and FXa inhibition at the endothelial surface, and the UniProt curator worked from the full text. Retained with that limitation stated rather than marked undecided, because the topology makes an externally exposed surface pool unproblematic. Supporting Evidence: PMID:21868574 Dihydrotestosterone up-regulates TFPI and ADTRP expression, and increases FXa inhibition by TFPI in an ADTRP- and caveolin-1-dependent manner. |
| GO:0010628 positive regulation of gene expression | IMP PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | KEEP AS NON CORE | Summary: ADTRP knockdown lowers and overexpression raises TFPI mRNA. Real, reproducible, and the origin of the gene's name - but an indirect effect and a very general term. Reason: This is the observation the gene was named for, and it is supported in both directions by shRNA and overexpression. It remains non-core for two reasons: the term says nothing about the target or the mechanism, and no direct transcriptional or promoter-binding activity of ADTRP has ever been shown. The mechanism was later resolved: the DNA-binding protein is the transcription factor POU1F1, which binds an ADTRP-response element 806 to 756 bp upstream of the TFPI start site, and deleting that site or knocking down POU1F1 abolishes ADTRP-mediated TFPI transcription. That places ADTRP upstream of a transcription factor rather than at the DNA, so an effect-on-expression process term is the correct aspect and no DNA-binding or transcription-factor molecular function should be inferred. 'Regulates the expression of TFPI' is a biological process, not a molecular function, and it is correctly annotated as one here. Supporting Evidence: PMID:21868574 ADTRP-shRNA reduces, while over-expression of ADTRP enhances, TFPI mRNA and activity and the colocalization of TF-FVIIa-FXa-TFPI with caveolin-1. PMID:21868574 We demonstrate that this protein regulates both the native and androgen-enhanced TFPI expression and activity in cultured ECs, and we named it androgen-dependent TFPI-regulating protein (ADTRP). PMID:32445923 Deletion of POU1F1-binding site or knockdown of POU1F1 expression abolished ADTRP-mediated transcription of TFPI. PMID:32445923 ChIP and EMSA demonstrated that POU1F1 binds to the ADTRP response element. file:human/ADTRP/ADTRP-deep-research-affinage.md ADTRP regulates TFPI transcription through transcription factor POU1F1 |
| GO:0030195 negative regulation of blood coagulation | IMP PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | KEEP AS NON CORE | Summary: ADTRP raises TFPI-dependent FXa inhibition at the endothelial surface. An indirect, tissue-level consequence of regulating TFPI, not an activity of ADTRP. Reason: The anticoagulant readout is inhibition of factor Xa by TFPI, and the ADTRP dependence is on TFPI expression and its caveolar localisation. ADTRP is not itself a protease inhibitor or a coagulation factor, so this is a downstream physiological consequence. It is a genuine, experimentally supported process role and is retained, but it must not be read as evidence for any molecular function: the vascular and thrombosis associations of this locus, including its coronary-artery-disease GWAS signal, establish phenotype rather than activity. Supporting Evidence: PMID:21868574 the ADTRP-dependent up-regulation of TFPI expression and activity by androgen represents a novel mechanism of increasing the anticoagulant protection of the endothelium |
| GO:0071383 cellular response to steroid hormone stimulus | IEP PMID:21868574 Novel protein ADTRP regulates TFPI expression and function i... | KEEP AS NON CORE | Summary: Dihydrotestosterone raises ADTRP expression. An expression-response annotation, correctly coded IEP. Reason: IEP is the right code for an expression-pattern inference and the term matches: dihydrotestosterone is a steroid hormone and ADTRP transcript rises in response. UniProt records the same fact as 'INDUCTION: By androgens'. This describes how the gene is regulated rather than what the protein does, so it is non-core by construction. Supporting Evidence: PMID:21868574 Dihydrotestosterone up-regulates TFPI and ADTRP expression, and increases FXa inhibition by TFPI in an ADTRP- and caveolin-1-dependent manner. |
Loading supporting content…
Download this section (compressed HTML)Q: PAINT recommendation for PTHR10989. At node PTN001659973, scoped taxon:2759 (Eukaryota) and seeded only by human ADTRP (Q96IZ2) and human AIG1 (Q9NVV5), the molecular function is held at GO:0016787 hydrolase activity - the root of the branch - while the biological process is the far more specific GO:0042758 long-chain fatty acid catabolic process. Both reach all 86 recipients, including 14 fungal, 5 plant and 2 other-eukaryote proteins, of which the only reviewed representatives are two uncharacterised UPF0641 proteins (P38842, Q96WV4). The character conserved across the node is the Thr/His catalytic dyad: 73 of 85 recipients retain it when aligned onto ADTRP's own annotated SITE positions. A conserved dyad supports a claim about catalytic mechanism, which is what GO:0016787 states, but not a claim about substrate - and GO:0042758 is a substrate-level term whose substrate has been established only for the four characterised animal members. Should the process term therefore sit on the vertebrate or mammalian subclade where the IBD seeds actually are, as PTN002591065 already does for GO:0005901, leaving the mechanism-level MF at the Eukaryota node? Note the MF placement is recent and deliberate (2026-05-28, after GO:0120573 was created on 2026-03-14), so this is a question about the process term rather than a suggestion that the MF is stale.
Suggested experts: GO Central / PAINT curators
Q: UniProt evidence attribution for Q96IZ2. The SUBCELLULAR LOCATION line cites both PubMed:21868574 and PubMed:27018888 as ECO:0000269 evidence for 'Cell membrane', and GOA has split this into two EXP rows for GO:0005886. The cached full text of PubMed:27018888 contains no localisation experiment - no immunofluorescence, confocal imaging or fractionation beyond bulk membrane lysates of transfected HEK293T cells, and no occurrence of the phrases 'plasma membrane' or 'cell surface'. Should that reference be attached to 'Multi-pass membrane protein' only, leaving PubMed:21868574 as the evidence for the plasma-membrane compartment?
Suggested experts: UniProt curators
Q: Is GO:2000402 negative regulation of lymphocyte migration correct for ADTRP from PMID:28341552? The abstract describes monocyte adhesion and monocyte transendothelial migration and never mentions lymphocytes. If the full text has no lymphocyte migration assay, GO:2000438 negative regulation of monocyte extravasation is the matching term; GO:0071676 negative regulation of mononuclear cell migration is a verified ancestor of both and would be the conservative choice.
Suggested experts: BHF-UCL curators
Q: Does ADTRP's FAHFA hydrolase activity have anything to do with its effect on TFPI expression? The two literatures have never been joined: no study has asked whether the catalytically dead T47A or H131A mutant still supports TFPI up-regulation, nor whether FAHFA levels change in endothelial cells on ADTRP knockdown. Until that is done, the gene has two unconnected functional descriptions and GO records both without a causal link.
Suggested experts: Lipid signalling, Vascular biology
Q: Should human ADTRP carry a caveola annotation with the is_active_in relation? The seeding evidence is a located_in colocalisation IDA in endothelial cells, and no assay has placed FAHFA hydrolase activity in caveolae specifically.
Suggested experts: GO Central / PAINT curators
Experiment: Reconstitute ADTRP knockdown endothelial cells with wild-type ADTRP, the T47A nucleophile mutant and the H131A general-base mutant, and measure TFPI mRNA, cell-associated TFPI anticoagulant activity against factor Xa, and cellular FAHFA levels by LC-MS in parallel. A catalytically dead mutant that still restores TFPI would separate the two functions; one that does not would unify them.
Hypothesis: TFPI regulation by ADTRP requires its FAHFA hydrolase activity.
Type: structure-function rescue
Experiment: Fractionate endothelial membranes into caveolar and non-caveolar pools by detergent-free density gradient, and assay FAHFA hydrolase activity in each with and without the ADTRP-active inhibitors KC01 and JJH260 alongside their inactive structural controls THL and ABC34. This would test the is_active_in caveola claim that PAINT currently propagates to 25 vertebrate orthologues.
Hypothesis: ADTRP hydrolyses FAHFAs within caveolae rather than across the plasma membrane generally.
Type: subcellular fractionation and activity assay
Experiment: Express the two reviewed uncharacterised family members, S. cerevisiae YHR140W (P38842) and S. pombe SPBPJ4664.05 (Q96WV4), and assay them against the same FAHFA and non-FAHFA lipid panel used for ADTRP and AIG1, checking fluorophosphonate probe reactivity and the conservation of the Thr/His pair. A negative result would justify restricting the node's process annotation to the clade where activity is demonstrated.
Hypothesis: The fungal ADTRP/AIG1 members that PAINT annotates as long-chain fatty acid catabolic enzymes are not FAHFA hydrolases.
Type: heterologous expression and substrate panel
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: The physiological substrate-to-phenotype link for human ADTRP is unknown: no study connects its measured FAHFA hydrolase activity to any of the endothelial, coagulation or atherosclerosis phenotypes attributed to the gene.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: Known: the enzyme hydrolyses FAHFAs in vitro and controls FAHFA levels in mouse tissues; and separately, its knockdown alters TFPI expression, monocyte adhesion and PIK3R3/AKT signalling in human endothelial cells. Unknown: whether either phenotype is downstream of the catalytic activity.
Significance: The gene has two independent functional literatures that GO records side by side without a causal relation. Any core-function statement that merges them would be asserting a link nobody has tested.
Provenance (the field's own admissions):
Gap: GO can express FAHFA hydrolysis as a molecular function but not as a biological process, so FAHFA catabolism has to be annotated with the much broader GO:0042758 long-chain fatty acid catabolic process.
OPEN ONTOLOGY BP_DARK
What is known: Known: GO:0120573 FAHFA hydrolase activity exists and is RHEA-anchored. Unknown/absent: any FAHFA-level process term; a QuickGO search for 'FAHFA' returns exactly one term.
Significance: The missing term is why a pan-eukaryotic PAINT node propagates a specific lipid-catabolism claim to 86 proteins: there is no way to state the narrower fact.
Provenance (the field's own admissions):
Gap: The ADTRP/AIG1 family is pan-eukaryotic with 5788 members, but only four proteins have ever been characterised biochemically: human and mouse ADTRP and human and mouse AIG1. No substrate is known for any fungal, plant or invertebrate member.
OPEN BIOLOGYCURATION MF_DARK
What is known: Known: two human enzymes, their mouse orthologues, a mouse knockout phenotype, and that the Thr/His catalytic dyad is retained in 73 of 85 other members of the PANTHER node - so most of the family is very likely to be catalytically competent. Unknown: what any non-animal member hydrolyses, and whether FAHFAs exist as metabolites outside animals. Undetermined rather than negative for the 14 fungal members, whose alignments to ADTRP all fall below 25% identity.
Significance: Two of the eight reviewed members of the family signature IPR006838 are curated as uncharacterised UPF0641 proteins, yet PAINT gives all 86 recipients of PTN001659973 a long-chain fatty acid catabolic process annotation - a substrate claim, where the conserved evidence supports only a mechanism claim.
Provenance (the field's own admissions):
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)