COX15 is a multipass mitochondrial inner membrane enzyme that converts heme O to heme A by oxidation of its C8 methyl group to a formyl group. Heme A supplies the prosthetic groups of cytochrome c oxidase, linking this biosynthetic reaction to respiratory complex IV assembly and activity. Human COX15.1 partially restores heme A content and oxidase activity in deficient patient fibroblasts. Alternative splicing produces a 388-residue product with a different C terminus from the 410-residue displayed product; equal catalytic competence of the two products has not been established. Biallelic pathogenic variants cause mitochondrial complex IV deficiency, including severe infantile cardiomyopathy and neurological disease.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0120547 heme A synthase activity | IBA GO_REF:0000033 | ACCEPT | Summary: Conserved heme A synthase activity is supported by phylogeny and human rescue. Reason: Current PTHR23289 topology places the exact human 410-residue sequence beneath the recorded PTN000590786 ancestor. Current fission-yeast and budding-yeast same-term records include experimental support, while human COX15.1 rescue independently increases heme A. The current v19 GO-slim chemistry fields differ from current GOA; topology and MSA do not replay the historical IBD curation session. No target-specific loss of this activity is established. The tree and MSA response URLs/hashes, exact target node and root-to-target path are recorded under paint in COX15-source-evidence.yaml. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000590786 SUPPORTS TRANSFER Current PTHR23289 v19 places the exact human sequence at descendant PTN002510931. Experimental yeast same-term records and human heme A rescue support conserved synthase activity. Current GO-slim fields do not reconstruct the historical IBD session. |
| GO:0005743 mitochondrial inner membrane | IBA GO_REF:0000033 | ACCEPT | Summary: The mitochondrial inner membrane is the conserved site of COX15 function. Reason: The current target lies below PTN000590786. Current experimental location records are present for both yeast sources and human COX15; the target appearing among descendant evidence is legitimate grounding. Human import evidence and curated inner-membrane location agree with the multipass enzyme architecture. This is not an inference based on the number of donors. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000590786 SUPPORTS TRANSFER The current target is a descendant of this recorded ancestral node. Yeast experimental inner-membrane records and direct human import/location evidence support the placement; human descendant evidence is legitimate grounding. |
| GO:0006784 heme A biosynthetic process | IBA GO_REF:0000033 | ACCEPT | Summary: COX15 performs the heme A-forming step of the pathway. Reason: The current family tree supports the recorded ancestral placement, with experimental heme A process records in both yeast sources and on the human target. Human COX15.1 complementation increases heme A, providing performed catalytic work rather than merely a downstream respiratory phenotype. The human self-source is not circular. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000590786 SUPPORTS TRANSFER The current tree places human COX15 within this ancestral clade. Yeast experimental heme A process records and human COX15.1 rescue support performed pathway chemistry, without claiming a replay of the historical IBD curation. |
| GO:0005743 mitochondrial inner membrane | IEA GO_REF:0000044 | ACCEPT | Summary: The curated UniProt location maps specifically to the mitochondrial inner membrane. Reason: The complete current source identifies UniProtKB-SubCell:SL-0168, a location vocabulary entry rather than a donor gene. Its mapping agrees with the curated target location, conserved multipass architecture and the reaction-specific Reactome entity. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0168 SUPPORTS TRANSFER This is the curated UniProt inner-membrane vocabulary mapping, not a donor protein. It agrees with target architecture and the reaction-specific Reactome entity. Supporting Evidence: GO_REF:0000044 Reactome:R-HSA-2995334 |
| GO:0006784 heme A biosynthetic process | IEA GO_REF:0000120 | ACCEPT | Summary: The heme A process mapping agrees with the enzyme family and human evidence. Reason: Current IPR003780 and IPR023754 both explicitly map to heme A biosynthesis; the complete source also retains UniPathway UPA00269. The human protein performs the terminal conversion of heme O to heme A, as supported by complementation and heme measurements. This does not assign all heme pathway reactions to COX15. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR003780 SUPPORTS TRANSFER The inspected current entry explicitly maps to heme A biosynthesis; the target carries this enzyme-family signature. InterPro:IPR023754 SUPPORTS TRANSFER The inspected current entry maps the heme A synthase family to heme A biosynthesis. UniPathway:UPA00269 UNRESOLVED The complete current source lists this pathway identifier. Its independent historical pathway record was not recovered; the process is supported by the two inspected InterPro mappings and human rescue. |
| GO:0016020 membrane | IEA GO_REF:0000002 | MODIFY | Summary: Refine the retired broad membrane assertion to the inner membrane. Reason: Independent target localization and the reaction-specific Reactome entity support mitochondrial inner membrane, as for the active combined-source record. The historical InterPro-only GO_REF:0000002 source is absent from the current source set and remains retired. Retirement describes source turnover; it does not alter the biological location or justify a different specificity judgment. The original omitted WITH/FROM field remains omitted rather than being backfilled from a distinct current source. Propagation Review Root cause: TERM SCOPING PROBLEM Sources checked: GO_REF:0000002 SOURCE STALE OR MISSING This historical source lacks a retained WITH/FROM field and is absent as an exact source identity from the current set. The distinct current combined mapping is reviewed separately; source turnover does not refute membrane localization. Proposed replacements: mitochondrial inner membrane |
| GO:0120547 heme A synthase activity | IEA GO_REF:0000120 | ACCEPT | Summary: The specific heme A synthase reaction is supported by current electronic mappings. Reason: The complete current record identifies IPR023754 and EC:1.17.99.9. The actual InterPro entry maps to GO:0120547 and describes oxidation of the C8 methyl group of heme O to a formyl group. Human rescue supports the same function, without establishing equal activity of the alternative C-terminal product. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR023754 SUPPORTS TRANSFER The actual current entry maps to GO:0120547 and describes the heme O C8 methyl-to-formyl conversion. EC:1.17.99.9 SUPPORTS TRANSFER The complete current source names this heme A synthase enzyme-class mapping, consistent with the verified current GO reaction definition and human complementation. |
| GO:0005515 protein binding | IPI PMID:29128334 A Map of Human Mitochondrial Protein Interactions Linked to ... | MODIFY | Summary: The positive PRDX6 association supports phospholipase binding. Reason: Three exact human COX15–PRDX6 records from PMID:29128334 include an anti-bait co-immunoprecipitation record and two spoke-expanded cross-link associations in HEK293 cells. Their sequence ROGIDs match the 410-residue COX15 and 224-residue PRDX6 products. Independent reviewed human PRDX6 annotation establishes phospholipase A2 activity; the primary PMID:10893423 abstract corroborates the bifunctional enzyme class, without resolving every construct detail. GO:0043274 is binding to a phospholipase and has no returned current child. This counterpart-class refinement preserves positive association without claiming phospholipase regulation, contact with its catalytic site, purified binary affinity, or a second COX15 core. PRDX6 peroxidase activity does not negate its independently established phospholipase class: GO:0043274 requires binding to a phospholipase, not binding that depends on phospholipase catalysis. This follows the CLINGEN_MENDELIAN standing instruction to refine positive binding evidence by an independently verified counterpart class; it does not create annotations for every PRDX6 activity. Assay and product details are recorded under binding and products in COX15-source-evidence.yaml. Proposed replacements: phospholipase binding Supporting Evidence: PMID:10893423 protein that hydrolyzed the sn-2 acyl chain of phospholipids |
| GO:0005739 mitochondrion | IDA GO_REF:0000052 | ACCEPT | Summary: Human immunofluorescence curation supports mitochondrial localization. Reason: The current HPA ICC/IF record reports mitochondria as the main location with supported reliability and nucleoplasm as an additional location. The returned experimental summary was inspected; image pixels were not independently scored. The mitochondrial annotation is retained without assigning localization to both splice products individually. Supporting Evidence: GO_REF:0000052 |
| GO:0006783 heme biosynthetic process | TAS Reactome:R-HSA-189451 | MODIFY | Summary: Specify heme A biosynthesis within the broad heme pathway. Reason: Current R-HSA-189451 contains R-HSA-2995334. Its catalyst and physical-entity path resolves to human Q7KZN9 and the heme O-to-heme A reaction. The more specific heme A process captures COX15 enzymatic participation without attributing upstream porphyrin synthesis to it. Proposed replacements: heme A biosynthetic process Supporting Evidence: Reactome:R-HSA-189451 Reactome:R-HSA-2995334 |
| GO:0120547 heme A synthase activity | TAS Reactome:R-HSA-2995334 | ACCEPT | Summary: Reactome assigns COX15 the heme O-to-heme A reaction. Reason: The actual event reaches catalyst 2995322, physical entity 2995350 and reference entity 145870/Q7KZN9. The reaction-specific annotation agrees with current GO:0120547 chemistry and human rescue. It is curated reaction evidence, not a newly measured purified human enzyme mechanism. |
| GO:0005743 mitochondrial inner membrane | EXP PMID:9878253 Identification and characterization of human cDNAs specific ... | ACCEPT | Summary: Retain the established inner-membrane location with the original access limit explicit. Reason: The complete PMID:9878253 abstract reports mitochondrial import and protease protection of the human gene products. Its detailed original subcompartment Methods were not recovered. Curated target location, multipass architecture and the specific reaction entity independently support the inner membrane; lack of the full paper does not justify reversing the experimental annotation. |
| GO:0120547 heme A synthase activity | ISS GO_REF:0000024 | ACCEPT | Summary: The manually transferred synthase activity agrees with the authenticated bacterial donor. Reason: The complete current WITH/FROM identifies Bacillus subtilis CtaA/P12946, not yeast. Its current same-term records include EXP to PMID:15236569 and PMID:16321940 and IDA to PMID:30397130. Those raw source records were inspected, without claiming their full primary papers were read. Conserved synthase identity and independent human rescue support the transfer. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:P12946 SUPPORTS TRANSFER Official identity is Bacillus subtilis CtaA. Current same-term records include EXP references PMID:15236569 and PMID:16321940 and IDA PMID:30397130. Those curator records were inspected; the full donor papers were not independently read. |
| GO:0005739 mitochondrion | HTP PMID:34800366 Quantitative high-confidence human mitochondrial proteome an... | ACCEPT | Summary: Retain the high-throughput mitochondrial annotation. Reason: The original MitoCoP Table S1 workbook was recovered from the publisher and authenticated against its size and MD5 in the original PMC XML. Sheet (A), row 239 identifies protein group 7459, Q7KZN9/COX15 (including Q7KZN9-2 in the group), with subtractive Class 1, spatial Cluster 1 and importomics Class 1 assignments; sheet (B), row 214 confirms MitoCoP inclusion. These are positive target-specific proteomics observations, supporting the mitochondrial annotation. The Integral membrane category is also reported, but the grouped measurement does not distinguish isoform activity or itself resolve the inner versus outer membrane. Workbook URL/hash and exact cell locators are recorded under mitocop_table_s1 in COX15-source-evidence.yaml. |
| GO:0016627 oxidoreductase activity, acting on the CH-CH group of donors | IMP PMID:12474143 Mutations in COX15 produce a defect in the mitochondrial hem... | MODIFY | Summary: Correct the chemical subclass to heme A synthase activity. Reason: The original human study measures deficient heme A and partial rescue by sequence-verified COX15.1. Current GO:0120547 specifies oxidation of the C8 methyl group of heme O to the formyl group of heme A and descends from GO:0016725, acting on CH or CH2 groups. The existing CH-CH subclass is chemically misplaced. MODIFY preserves the demonstrated redox function while correcting the subclass; this is not a claim that the replacement is a literal child of GO:0016627. PR #836 removed this annotation because the CH-CH chemistry was wrong; that chemical objection remains valid. The present decision instead uses MODIFY to replace the misplaced subclass with the demonstrated heme A synthase function. GO:0120547 already occurs under five other source records; this is consolidation of an existing assertion, not a NEW annotation or additional functional coverage. Proposed replacements: heme A synthase activity Supporting Evidence: PMID:12474143 COX15 plays an essential role in mitochondrial heme modification |
| GO:0017004 cytochrome complex assembly | IMP PMID:12474143 Mutations in COX15 produce a defect in the mitochondrial hem... | KEEP AS NON CORE | Summary: Human complementation supports the existing complex-assembly context. Reason: Human COX15.1 increased heme A and restored assembled oxidase toward controls. The original native-gel antibodies detected respiratory subunits, and the rescue does not isolate an autonomous COX15 heme-insertion or scaffold mechanism. The existing IMP assembly association is retained outside the direct core process, which is heme A biosynthesis. Supporting Evidence: |
| GO:0070069 cytochrome complex | IDA PMID:12474143 Mutations in COX15 produce a defect in the mitochondrial hem... | UNDECIDED | Summary: The exact human cytochrome-complex assertion remains unresolved. Reason: GO:0070069 covers any complex containing a redox-active heme protein, not only mature respiratory complex IV. The original PMID:12474143 Electrophoretic Methods and Results/Figure 3 assay respiratory subunits and heme profiles, not the exact COX15 complex assertion. Independent yeast Cox15 oligomers and porcine complexes in PMID:26940873 make a broader heme-containing assembly plausible without resolving this human IDA assertion. PR #836 removed the record by excluding COX15 from respiratory-complex membership; that argument does not settle this broader term. UNDECIDED records the unresolved evidence-to-assertion relationship, not an inability to read the 2003 original. That original was recovered externally; the repository cache itself contains Introduction, one Electrophoretic Methods paragraph and Discussion but omits the Results body. Neither acceptance of the exact assertion nor permanent exclusion from all cytochrome complexes is established. |
| GO:0006784 heme A biosynthetic process | IGI PMID:12474143 Mutations in COX15 produce a defect in the mitochondrial hem... | ACCEPT | Summary: Human COX15.1 complementation restores heme A synthesis. Reason: The sequence-verified human COX15.1 construct increased patient fibroblast heme A and respiratory oxidase activity. Rescue was partial relative to controls expressing COX15, and heme O accumulated in patient heart but was undetectable in patient fibroblasts. The source does not test equivalent function of the shorter product. The isoform WITH/FROM is preserved in the separately recorded complete current source. Supporting Evidence: |
| GO:0020037 heme binding | TAS PMID:26940873 Analysis of Oligomerization Properties of Heme a Synthase Pr... | KEEP AS NON CORE | Summary: Retain heme binding at the source’s mechanistic scope. Reason: The full original PMID:26940873 studies conserved histidines and heme synthesis in yeast, porcine mitochondrial oligomers, and protein abundance in human patient fibroblasts. It supports the heme-binding model and substrate relationship, but does not directly measure binding by purified human COX15 or distinguish all catalytic versus heme-ligating roles of those residues. This TAS annotation remains non-core beside the specific synthase activity. Supporting Evidence: PMID:26940873 investigated Cox15 in the model eukaryote Saccharomyces cerevisiae |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-2995334 | ACCEPT | Summary: The heme A reaction entity is in the mitochondrial inner membrane. Reason: The actual catalyst physical entity 2995350 resolves to Q7KZN9 and explicitly has GO:0005743 as its compartment. Matrix also occurs in the event-level compartment list because the reaction has other participants; that does not independently locate COX15 in the matrix lumen. Supporting Evidence: Reactome:R-HSA-2995334 |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-9865449 | ACCEPT | Summary: Reactome complex membership supports the inner-membrane location. Reason: The current event for MT-CO1 copper insertion reaches catalyst 9865627, complex 9865441, component 2995350 and reference Q7KZN9. The COX15 entity is explicitly located in the mitochondrial inner membrane. This supports the location annotation; a modeled metallochaperone-complex role is not evidence that isolated COX15 catalyzes copper insertion or establishes every mechanistic statement in the event summary. Supporting Evidence: Reactome:R-HSA-9865449 |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-9865579 | ACCEPT | Summary: Reactome complex membership supports the inner-membrane location. Reason: The current event for association of MT-CO1 and MT-CO2 modules reaches catalyst 9865627, complex 9865441, component 2995350 and reference Q7KZN9. The COX15 entity is explicitly located in the mitochondrial inner membrane. This supports the location annotation; a modeled metallochaperone-complex role is not evidence that isolated COX15 catalyzes copper insertion or establishes every mechanistic statement in the event summary. Supporting Evidence: Reactome:R-HSA-9865579 |
| GO:0005743 mitochondrial inner membrane | TAS Reactome:R-HSA-9865630 | ACCEPT | Summary: Reactome complex membership supports the inner-membrane location. Reason: The current event for MT-CO2 copper insertion reaches catalyst 9865627, complex 9865441, component 2995350 and reference Q7KZN9. The COX15 entity is explicitly located in the mitochondrial inner membrane. This supports the location annotation; a modeled metallochaperone-complex role is not evidence that isolated COX15 catalyzes copper insertion or establishes every mechanistic statement in the event summary. Supporting Evidence: Reactome:R-HSA-9865630 |
| GO:0005739 mitochondrion | IDA PMID:9878253 Identification and characterization of human cDNAs specific ... | ACCEPT | Summary: The original human cDNA study supports mitochondrial import. Reason: The complete original abstract explicitly reports an in vitro mitochondrial import and protease-protection assay for the human gene products. The unavailable full Methods limit reconstruction of product and subcompartment details, but the positive broad mitochondrial localization is retained. Supporting Evidence: PMID:9878253 import and protease-protection assay. |
| GO:0098803 respiratory chain complex | TAS PMID:9878253 Identification and characterization of human cDNAs specific ... | UNDECIDED | Summary: The original respiratory-chain-complex membership basis is unresolved. Reason: GO:0098803 specifically describes a complex in a respiratory chain. PMID:9878253 establishes human mitochondrial targeting in its complete abstract, but legitimate original full-text recovery did not expose the evidence underlying this TAS complex annotation. Cox15 homo-oligomers and biochemical dependence of complex IV do not by themselves resolve this more specific membership question. PR #836 removed this record on the premise that an assembly enzyme is never a respiratory-complex component. Without the original evidence, that categorical exclusion is not justified; mitochondrial targeting alone also cannot establish the narrower complex location. UNDECIDED preserves both limits pending the missing original evidence. Supporting Evidence: |
| GO:0016020 membrane | IEA GO_REF:0000120 | MODIFY | Summary: Refine the current broad membrane assertion to the inner membrane. Reason: The exact current source combines IPR003780, IPR023754 and ARBA00028763. Both InterPro entries explicitly map to membrane. The current ARBA rule has 1,269 branches; no match is established from the inspected InterPro/PANTHER/taxonomic features, while branches requiring uncollected FunFam assignments remain unresolved. The refinement rests on independent target location and the explicit Reactome entity, not on a fabricated ARBA branch match. This is an existing current GOA record appended after retirement of a different historical source, not a NEW annotation. The preserved raw-response URL/hash and exhaustive branch partitions are in COX15-source-evidence.yaml under arba; these are current-rule observations, not reconstruction of the historical rule. Propagation Review Root cause: TERM SCOPING PROBLEM Sources checked: InterPro:IPR003780 SUPPORTS TRANSFER The current entry supports broad membrane localization. Independent target evidence supplies the more precise inner-membrane replacement. InterPro:IPR023754 SUPPORTS TRANSFER The current heme A synthase family entry explicitly maps to membrane; it does not alone specify the inner membrane. ARBA:ARBA00028763 UNRESOLVED Across 1,269 current branches, none is confirmed from the inspected features; 999 retain unresolved FunFam conditions and 270 are nonmatching. No target FunFam assignment was recovered. The refinement relies on independent target location, not an asserted rule-branch match. Proposed replacements: mitochondrial inner membrane |
Loading supporting content…
Download this section (compressed HTML)Q: Does the 388-residue C-terminal alternative product retain heme A synthase activity, and does it cooperate with COX15.1 under endogenous expression conditions?
Suggested experts: Antonicka H, Shoubridge EA
Q: What direct human evidence distinguishes a heme-containing COX15 oligomer from transient membership in respiratory-chain assembly intermediates?
Suggested experts: Khalimonchuk O, Fox JL
Q: What role, if any, does the observed COX15–PRDX6 association play in mitochondrial lipid or redox physiology?
Q: How does newly synthesized heme A move from COX15 to cytochrome c oxidase without conflating catalytic synthesis with an untested insertase or scaffold activity?
Suggested experts: Khalimonchuk O, Shoubridge EA
Q: How is COX15 abundance and activity coordinated with COX10 (heme O synthase) and with CIV assembly demand to prevent heme O accumulation and mis-incorporation into non-COX targets?
Suggested experts: Antonicka H, Shoubridge EA
Experiment: Compare sequence-verified COX15.1 and COX15.2 at matched expression in deficient human cells, measuring import, membrane topology, heme O/heme A by calibrated chromatography and respiratory rescue with catalytic-loss controls.
Hypothesis: The alternative C terminus changes COX15 product activity or import under matched expression conditions.
Type: matched human isoform complementation with heme quantification
Experiment: Resolve endogenous COX15 assemblies by reciprocal affinity isolation and native mass spectrometry with respiratory-subunit and heme measurements, distinguishing stable homo-oligomers from transient assembly intermediates.
Hypothesis: Human COX15 forms distinguishable stable oligomers and transient respiratory assembly associations.
Type: endogenous complex isolation and native mass spectrometry
Experiment: Test the PRDX6 association by reciprocal endogenous co-immunoprecipitation and purified membrane-compatible reconstitution, then separately assay phospholipase and peroxidase effects with appropriate catalytic controls.
Hypothesis: The COX15–PRDX6 association affects one or both PRDX6 catalytic activities under defined membrane conditions.
Type: reciprocal interaction assays with separate phospholipase and peroxidase measurements
Experiment: Use proximity labeling (BioID/TurboID) of catalytically active vs catalytically dead COX15 in human cells, combined with native-PAGE of CIV assembly intermediates, to map COX15 partners along the MT-CO1 hemylation pathway and detect a putative heme A channeling complex.
Hypothesis: COX15 hands off heme A directly to nascent MT-CO1 within an assembly super-complex, rather than releasing heme A into the mitochondrial inner membrane bulk lipid bilayer.
Type: proximity labeling combined with native CIV assembly profiling
Experiment: Express and purify recombinant human COX15 (and the conserved-histidine mutants) in membrane-mimetic systems; characterise heme content and electronic structure by UV-Vis, EPR, and resonance Raman; compare wild-type and mutant turnover with heme O substrate.
Hypothesis: The four conserved histidines of COX15 are direct heme-A-pathway ligands whose mutation traps a covalent or non-covalent heme-bound intermediate.
Type: spectroscopy of purified recombinant human heme A synthase
Experiment: Compare isogenic human iPSC-derived cardiomyocytes carrying R217W and splice variants with wild-type controls. Measure heme A, oxidase activity, respiratory assembly intermediates and COX15 oligomeric state, using matched expression and independent catalytic and stability controls to distinguish the competing mechanisms.
Hypothesis: Pathogenic COX15 substitutions can affect catalytic competence and oligomer stability to different extents.
Type: iPSC-cardiomyocyte disease modelling with native complex profiling
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)