ABHD12

Summary: ABHD12 hydrolyzes the oxidized sn-2 acyl chain of phosphatidylserine.

Reason: Human ABHD12 membrane preparations hydrolyze oxidized PS in PMID:30643283, with inactive S246A and mock controls. The substrate assays quantify both fatty-acid and lysoPS products; the source-linked UniProt reactions RHEA:59364/59368 retain the sn-1 acyl group and release the oxidized sn-2 fatty acid. The A2-type term resolves this chemistry more precisely than the glycerophospholipase parent. This does not imply activity on all phospholipids or unoxidized PS.

Proposed replacements: A2-type glycerophospholipase activity

Summary: The cited human experiment measures lysophosphatidylserine hydrolysis.

Reason: PMID:25290914 explicitly uses human ABHD12 as a positive control in Figure 4, despite focusing on BAT5/ABHD16A. It measures oleate release from 1-oleoyl-lysoPS, with LPA and neutral-lipid comparisons. Current GO:0004622 specifies a phosphocholine substrate. Replace it with the phosphoserine-specific activity demonstrated by this source; this is a substrate-resolution correction, not a claim that the paper studied the wrong gene. The replacement is already represented among the seeded rows; this corrects the source-specific assertion and adds no new functional coverage.

Proposed replacements: phosphatidylserine lysophospholipase A1 activity

Summary: The broad cytoplasmic location is compatible with the established intracellular membrane pool.

Reason: The mouse donor annotation traces to the expressed ABHD12 localization experiment in PMID:27307232. Cytoplasm includes intracellular organelles and is not synonymous with soluble cytosol. Human MCF7 fractionation and microscopy in PMID:30237167 establish an ER-associated intracellular pool, supporting the broad location without replacing the source assertion by a finer compartment. This is an inclusive location of the core membrane enzyme, not a secondary pool merely because the integrated core names the ER membrane more precisely.

Summary: ABHD12 is enriched at the endoplasmic reticulum membrane.

Reason: The PAINT assertion at PTN000957701 represents inherited ER-membrane localization, not a pairwise transfer or a vote among descendants. Direct target evidence from PMID:30237167 corroborates this location: endogenous human MCF7 fractionation and ER-marker microscopy agree with mouse brain, Neuro-2a and primary-macrophage experiments. The node is retained without claiming an independently reconstructed ancestral tree or exclusive localization in every cell type.

Summary: ABHD12 is enriched at the endoplasmic reticulum membrane.

Reason: PMID:30237167 combines organelle fractionation with ER-marker immunofluorescence. The study includes endogenous human MCF7 ABHD12 as well as mouse brain, Neuro-2a cells and primary mouse macrophages. These observations support the ER-membrane assertion at its original resolution; they do not establish exclusive localization in every cell type.

Summary: ABHD12 is enriched at the endoplasmic reticulum membrane.

Reason: The combined source mapping includes mouse Q99LR1/ENSMUSP00000053558 and UniProt location SL-0097. Its ER-membrane assertion agrees with independently measured endogenous human MCF7 fractionation and ER-marker microscopy in PMID:30237167. Retain this compartment at the source resolution; agreement among mapping inputs is not independent experimental replication.

Summary: Reactome places the monoacylglycerol-hydrolyzing ABHD12 pool at the plasma membrane.

Reason: R-HSA-5694462 and its ABHD12 entity R-HSA-5694474 explicitly use the plasma-membrane compartment. Retain this curated pathway location as contextual. The dominant ER localization in PMID:30237167 and the lack of plasma-membrane staining of overexpressed mouse ABHD12 in one neuronal experiment do not exclude a plasma-membrane pool in other settings. Direct quantitative localization of endogenous human ABHD12 at this surface remains less well resolved than the ER pool. The positive location source is the official physical-entity page https://reactome.org/content/detail/R-HSA-5694474, whose Compartment is plasma membrane; the ER study is a contextual limitation, not positive evidence for this surface location.

Summary: ABHD12 directly catabolizes oxidized phosphatidylserine.

Reason: PMID:30643283 demonstrates hydrolysis of chemically characterized oxidized PS by human ABHD12, with catalytically inactive and mock controls, and altered oxidized-PS abundance after expression or mouse gene deletion. The process assertion is supported by a catalytic step. Its substrate scope here is oxidized PS; the same assays find negligible hydrolysis of unoxidized PS.

Summary: ABHD12 directly catabolizes oxidized phosphatidylserine.

Reason: PMID:30643283 demonstrates hydrolysis of chemically characterized oxidized PS by human ABHD12, with catalytically inactive and mock controls, and altered oxidized-PS abundance after expression or mouse gene deletion. The process assertion is supported by a catalytic step. Its substrate scope here is oxidized PS; the same assays find negligible hydrolysis of unoxidized PS.

Summary: ABHD12 directly catabolizes oxidized phosphatidylserine.

Reason: PMID:30643283 demonstrates hydrolysis of chemically characterized oxidized PS by human ABHD12, with catalytically inactive and mock controls, and altered oxidized-PS abundance after expression or mouse gene deletion. The process assertion is supported by a catalytic step. Its substrate scope here is oxidized PS; the same assays find negligible hydrolysis of unoxidized PS.

Summary: Mouse ABHD12 expression changes PSD-95 palmitate labeling, but direct human protein depalmitoylase chemistry is not established.

Reason: The MGI-linked mouse Abhd12 NM_024465 construct in PMID:27307232 reduces PSD-95 palmitate labeling in HEK293T cells. This positive screen is preserved. However, the ABHD12 experiment does not isolate protein-thioester hydrolysis with ABHD12 catalytic-mutant or purified-substrate controls; the detailed catalytic and neuronal tests concern ABHD17. PMID:25580854 supports a lumenal/extracellular ABHD12 catalytic domain, consistent with the human UniProt topology placing catalytic residues 246/333/372 within residues 96-398. That orientation separates the active site from cytosolic PSD-95 palmitoyl cysteines, so the cellular labeling change does not by itself establish direct access and hydrolysis. GO:0008474 explicitly asserts palmitoyl-protein hydrolysis. Its transfer as a demonstrated human molecular function overstates this evidence; an indirect lipid effect or altered topology under overexpression is possible but was not demonstrated. This does not declare the mouse screen false or prove that no protein substrate can ever be hydrolyzed.

Summary: Lysophospholipid hydrolysis is direct phospholipid catabolism.

Reason: The human lysoPS substrate assays in PMID:30237167 directly demonstrate phospholipid breakdown, with a catalytic-serine control and quantitative product detection. The broader process term is true for this core reaction and need not be removed merely because glycerophospholipid and substrate-specific terms are also present. The integrated core uses the narrower glycerophospholipid process to summarize the same directly performed lipid degradation. That summary choice does not make the broader experimental or ortholog assertion a different, non-core process.

Summary: Lysophospholipid hydrolysis is direct phospholipid catabolism.

Reason: The human lysoPS substrate assays in PMID:30237167 directly demonstrate phospholipid breakdown, with a catalytic-serine control and quantitative product detection. The broader process term is true for this core reaction and need not be removed merely because glycerophospholipid and substrate-specific terms are also present. The integrated core uses the narrower glycerophospholipid process to summarize the same directly performed lipid degradation. That summary choice does not make the broader experimental or ortholog assertion a different, non-core process.

Summary: Membrane association is supported for human and mouse ABHD12.

Reason: The mouse-to-human membrane assertion agrees with human ABHD12 activity in membrane preparations and endogenous MCF7 microsomal enrichment in PMID:30237167. Preserve the broad source resolution; the independent ER-membrane annotations describe its more specific location. This is an inclusive location of the core membrane enzyme, not a secondary pool merely because the integrated core names the ER membrane more precisely.

Summary: Membrane association is supported for human and mouse ABHD12.

Reason: The mouse-to-human membrane assertion agrees with human ABHD12 activity in membrane preparations and endogenous MCF7 microsomal enrichment in PMID:30237167. Preserve the broad source resolution; the independent ER-membrane annotations describe its more specific location. This is an inclusive location of the core membrane enzyme, not a secondary pool merely because the integrated core names the ER membrane more precisely.

Summary: Monoacylglycerol hydrolysis can release arachidonate.

Reason: R-HSA-426048 includes conversion of arachidonoylglycerol to arachidonate and glycerol, and its ABHD6/12 reaction is supported by human-enzyme assays in PMID:22969151. Retain this biochemical contribution as contextual, rather than assigning ABHD12 the upstream DAG-lipase step. Whole-brain bulk 2-AG regulation is limited in the mouse knockout study, while newer microglial work reports a cell-state-dependent role.

Summary: Rodent ABHD12 is a verified constituent of native AMPA-receptor preparations.

Reason: The original PMID:22632720 institutional PDF lists the mouse protein identifier Q99LR1 as ABHD12 in Table 1. The experiments used total-brain membrane fractions from adult rats, wild-type mice and antibody-target-knockout mice, with multiple anti-GluA antibodies and IgG controls. The accession identifies the protein; it does not by itself identify the species of each preparation. This verifies the donor complex-membership evidence and supports a contextual ortholog transfer. It does not establish ABHD12 as an obligatory pore subunit or assign it receptor-gating chemistry.

Summary: The donor localization experiment examines ABHD12 in neuronal dendrites.

Reason: MGI curates dendrite cytoplasm from PMID:27307232, whose Methods specify mouse Abhd12 and whose Figure 3A compares FLAG-tagged constructs in rat hippocampal neurons. Retain the source-specific neuronal localization with these expression and host-species bounds. Neither the assay nor this ortholog transfer establishes an endogenous human synaptic-spine pool; the paper explicitly distinguishes ABHD12 from the spine-localized ABHD17 proteins. The primary Results state: "ABHD12 and ABHD13 were neither present at the plasma membrane nor in the dendritic spines." This restricts the observed pattern rather than excluding the broader dendritic cytoplasm (https://pmc.ncbi.nlm.nih.gov/articles/PMC5015780/).

Summary: The demonstrated acylglycerol catabolism is monoacylglycerol hydrolysis.

Reason: Human ABHD12 hydrolyzes a panel of monoacylglycerols in PMID:22969151, whereas the tested di- and triacylglycerols do not yield detectable hydrolysis. Refine the true parent process to monoacylglycerol catabolism. This is a secondary substrate class relative to the central lysoPS pathway, with a context-dependent physiological contribution to 2-AG metabolism. The replacement is already represented among the seeded rows; this corrects the source-specific assertion and adds no new functional coverage.

Proposed replacements: monoacylglycerol catabolic process

Summary: The demonstrated acylglycerol catabolism is monoacylglycerol hydrolysis.

Reason: The mouse Q99LR1 ortholog supports the transferred acylglycerol-catabolism parent. Human ABHD12 hydrolyzes a panel of monoacylglycerols in PMID:22969151, whereas the tested di- and triacylglycerols do not yield detectable hydrolysis. Refine the true parent process to monoacylglycerol catabolism. This is a secondary substrate class relative to the central lysoPS pathway, with a context-dependent physiological contribution to 2-AG metabolism. The replacement is already represented among the seeded rows; this corrects the source-specific assertion and adds no new functional coverage.

Proposed replacements: monoacylglycerol catabolic process

Summary: Human ABHD12 directly degrades a glycerophospholipid substrate.

Reason: The cited PMID:25290914 human ABHD12 control releases oleate from 1-oleoyl-lysoPS. Lysophospholipids satisfy the GO definition of glycerophospholipids, which requires at least one O-acyl, O-alkyl or O-alkenyl group. Thus this process is directly performed by ABHD12 and is not inferred merely from a mutant phenotype.

Summary: Human ABHD12 has experimentally demonstrated monoacylglycerol lipase activity.

Reason: PMID:22969151 assays recombinant human ABHD12 in HEK293 lysates using glycerol production and identifies its catalytic triad by mutagenesis. The MAG activity is real, including arachidonoylglycerols, but the integrated central lipid-homeostasis role is lysoPS hydrolysis. This classification does not dismiss its cell-specific endocannabinoid role.

Summary: The BAT5 study independently measures human ABHD12 MAG hydrolysis.

Reason: PMID:25290914 directly uses human ABHD12 as a positive control and measures time-dependent fatty-acid release from 1-arachidonoylglycerol and 1-linoleoylglycerol. Retain this secondary catalytic activity with the actual ABHD12 experiment, without borrowing BAT5 results.

Summary: Human ABHD12 hydrolyzes multiple MAG substrates.

Reason: PMID:30237167 measures human ABHD12 activity against 1-MAGs spanning C10 to C24 and selected 2-MAGs, using mock and S246A controls. The membrane-lysate data show a preference for very-long-chain 1-MAGs, but substrate accessibility may contribute to the apparent kinetics. This is a supported secondary activity alongside stronger lysoPS preference.

Summary: The EC, Rhea and ortholog mappings agree with direct human MAG assays.

Reason: The combined mapping represents the same MAG ester hydrolysis measured in PMID:22969151, PMID:25290914 and PMID:30237167. Retain the enzymatic activity as secondary rather than treating EC or rule mappings as independent experiments or assuming all mapped acyl species dominate physiological flux.

Summary: The Reactome reaction captures ABHD12 hydrolysis of arachidonoylglycerol.

Reason: R-HSA-5694462 explicitly assigns ABHD6/12 the hydrolytic conversion of arachidonoylglycerol to fatty acid and glycerol. Human-enzyme assays support this activity and isomer preference. Its contextual role is retained without equating ABHD12 with the quantitatively dominant bulk-brain MAG lipase.

Summary: Degradation of immunomodulatory lipids contributes to inflammatory regulation.

Reason: In PMID:25580854, thioglycollate-elicited mouse peritoneal macrophages lacking Abhd12 have elevated secreted lysoPS and cytokine release, and reducing ABHD16A-dependent lipid production suppresses these changes. PMID:30420694 adds active DO264 versus inactive (S)-DO271 controls, primary-human-macrophage lipid measurements, and heightened inflammatory lung responses in DO264-treated or Abhd12-null mice during LCMV clone 13 infection. These connect ABHD12 lipid degradation to inflammatory regulation without assigning cytokine-receptor activity. Retain the broad ortholog process as contextual: species, cell state, lipid substrate and stimulus constrain the immune direction.

Summary: MAG catabolism is a supported secondary lipid reaction.

Reason: ABHD12 itself cleaves monoacylglycerol ester bonds, providing direct process participation. The human substrate assays support this assertion, while lysoPS turnover is its major established homeostatic role. Retain the source assertion and its contextual endocannabinoid contribution; limited bulk-brain 2-AG effects do not disprove cell-specific MAG turnover.

Summary: MAG catabolism is a supported secondary lipid reaction.

Reason: ABHD12 itself cleaves monoacylglycerol ester bonds, providing direct process participation. The human substrate assays support this assertion, while lysoPS turnover is its major established homeostatic role. Retain the source assertion and its contextual endocannabinoid contribution; limited bulk-brain 2-AG effects do not disprove cell-specific MAG turnover.

Summary: MAG catabolism is a supported secondary lipid reaction.

Reason: ABHD12 itself cleaves monoacylglycerol ester bonds, providing direct process participation. The human substrate assays support this assertion, while lysoPS turnover is its major established homeostatic role. Retain the source assertion and its contextual endocannabinoid contribution; limited bulk-brain 2-AG effects do not disprove cell-specific MAG turnover.

Summary: MAG catabolism is a supported secondary lipid reaction.

Reason: ABHD12 itself cleaves monoacylglycerol ester bonds, providing direct process participation. The human substrate assays support this assertion, while lysoPS turnover is its major established homeostatic role. Retain the source assertion and its contextual endocannabinoid contribution; limited bulk-brain 2-AG effects do not disprove cell-specific MAG turnover.

Summary: The inferred depalmitoylation process is an obsolete duplicate of molecular-function coverage.

Reason: The seeded term is explicitly obsolete. The GO editors confirm in go-ontology issue 32290 and go-announcements issue 1174 that this activity should be represented as molecular function rather than this biological process. Remove the inferred BP while preserving the separately assessed GO:0008474 source row, now marked over-annotated at the human target. This does not negate the mouse PSD-95 palmitoylation experiment.

Summary: The Rhea mapping specifies lysoPE hydrolysis, but its exact donor assay remains incompletely checked.

Reason: RHEA:40895 is present on the human UniProt record by similarity to mouse Q99LR1. PMID:23297193 assays several lysophospholipids with mouse ABHD12 and reports no change in bulk-brain LPE hydrolysis after knockout; that negative tissue result does not exclude recombinant catalytic capacity. The exact LPE substrate-level Figure 5A result and transfer have not been independently resolved, so neither a core assignment nor a rejection is warranted.

Summary: Lysophosphatidylserine hydrolysis is the central ABHD12 activity.

Reason: Human ABHD12 directly releases fatty acid from lysoPS in PMID:25290914 and PMID:30237167. Catalytic-mutant controls, patient-cell lipid studies and the mouse knockout model support the physiological lysoPS pathway. This exactly matches the phosphatidylserine lysophospholipase A1 reaction and does not mean ABHD12 cleaves unoxidized diacyl PS.

Summary: Lysophosphatidylserine hydrolysis is the central ABHD12 activity.

Reason: Human ABHD12 directly releases fatty acid from lysoPS in PMID:25290914 and PMID:30237167. Catalytic-mutant controls, patient-cell lipid studies and the mouse knockout model support the physiological lysoPS pathway. This exactly matches the phosphatidylserine lysophospholipase A1 reaction and does not mean ABHD12 cleaves unoxidized diacyl PS.

Summary: Lysophosphatidylserine hydrolysis is the central ABHD12 activity.

Reason: Human ABHD12 directly releases fatty acid from lysoPS in PMID:25290914 and PMID:30237167. Catalytic-mutant controls, patient-cell lipid studies and the mouse knockout model support the physiological lysoPS pathway. This exactly matches the phosphatidylserine lysophospholipase A1 reaction and does not mean ABHD12 cleaves unoxidized diacyl PS.

Summary: Mouse ABHD12 supports a secondary lysophosphatidylinositol hydrolase capacity.

Reason: The RHEA:44588 mapping traces through the human UniProt by-similarity assertion to mouse Q99LR1. PMID:23297193 Figure 5 and Results explicitly report LPI hydrolysis and its decrease in knockout brain membranes. This supports a bounded ortholog-based secondary activity; modest lipid changes and the dominant lysoPS phenotype do not establish equivalent physiological importance for every lysophospholipid headgroup.