COX10

Summary: Phylogenetically supported mitochondrial localization.

Reason: The current PTHR43448 tree places human Q12887 below PTN000114245; its ungapped MSA sequence exactly matches the 443-residue reference protein. The current mouse donor has experimental mitochondrial evidence, and the target has independent HPA and proteome support. Target self-with/from is legitimate descendant evidence. Current topology supports placement but does not reconstruct every historical narrow-term IBD decision.

Summary: COX10 performs a step of heme A biosynthesis.

Reason: Human COX10 lies below the recorded PTN000114245 node in the current family tree. Current donor records for mouse, fission yeast and budding yeast and the human target retain heme A pathway support. Human patient-cell HPLC and complementation independently establish the pathway contribution. COX10 makes heme O, which is converted to heme A by COX15; it need not perform every pathway step. The target self-source is not circular.

Summary: The conserved specific reaction is protoheme IX farnesylation.

Reason: The current human sequence descends from PTN000114243, with mouse and E. coli source members in the authenticated family. Current E. coli CyoE IDA evidence explicitly measures heme O formation from protoheme and farnesyl diphosphate. Human yeast complementation and patient-cell heme rescue corroborate the orthologous function. Bacterial enzymology and phylogenetic inference are not presented as a purified human-enzyme assay.

Summary: Replace broad prenyltransferase activity with the specific heme reaction.

Reason: Current ARBA00027066 branches 1 and 20 match the official InterPro/FunFam features and human taxonomic lineage. The broad reaction is sound, but GO:0008495 specifies the experimentally grounded heme substrate and heme O product. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: protoheme IX farnesyltransferase activity

Summary: Specify heme A biosynthesis within the broader heme pathway.

Reason: The actual InterPro IPR006369 and IPR016315 descriptions identify the COX10/heme O synthase family. COX10 performs the farnesylation step leading to heme A, so the existing more-specific process GO:0006784 captures its contribution without assigning upstream porphyrin-ring synthesis. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: heme A biosynthetic process

Summary: The current ARBA rule supports the specific heme A pathway.

Reason: ARBA00087087 branch 0 matches the exact target FunFam and eukaryotic lineage. Its heme A process inference agrees with human heme measurements and complementation; COX10 contributes enzymatic work through heme O production.

Summary: Independent electronic sources converge on protoheme IX farnesyltransferase activity.

Reason: The current combined source object retains the mouse UniProt/Ensembl transfer, COX10 InterPro signatures, RHEA:28070 and EC:2.5.1.141. The mouse accession and Ensembl product cross-reference were authenticated, and complete current same-term records are preserved. These support the same reaction as the human complementation and bacterial direct assay, without converting electronic evidence into human IDA.

Summary: Refine membrane localization to the mitochondrial inner membrane.

Reason: UbiA-family and COX10-specific InterPro signatures establish an integral membrane protein. The human reaction-specific Reactome catalyst entity explicitly locates COX10 at the mitochondrial inner membrane; current sequence topology is consistent with that site. This refinement does not assert mitochondrial import or equal activity for the untested shorter product.

Proposed replacements: mitochondrial inner membrane

Summary: Specify protoheme IX farnesyltransferase activity.

Reason: IPR000537 is a broad UbiA prenyltransferase family that also contains enzymes with other substrates. COX10-specific family and reaction evidence resolves the transferred alkyl-group activity to protoheme IX farnesylation; GO:0008495 is the supported specific descendant. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: protoheme IX farnesyltransferase activity

Summary: Specify the complex IV assembly context.

Reason: Current ARBA00086146 branch 3 matches the exact target FunFam and chordate lineage. The relevant cytochrome complex is respiratory complex IV. Human COX10-deficient fibroblasts have altered complex IV subassembly patterns, supporting the existing broad assembly context. This does not establish a heme-insertase activity or make assembly a separate direct catalytic core function. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: respiratory chain complex IV assembly

Summary: Refine mitochondrial membrane to the inner membrane.

Reason: The combined current record identifies COX10-specific IPR016315 and the UniProt mitochondrial inner-membrane location source. The reaction-specific Reactome physical entity is also explicitly in the inner membrane. The broader mitochondrial-membrane term can therefore be specialized.

Proposed replacements: mitochondrial inner membrane

Summary: Mitochondrial localization is supported across orthologs and on the human target.

Reason: The current Ensembl/UniProt source resolves to mouse Q8CFY5, whose official record carries the stated Ensembl protein identifier. Full current donor records and their provenance are preserved. The broad organelle location agrees with the mouse mitochondrial experiments and independent human HPA evidence, without importing a matrix-specific interpretation.

Summary: The transferred matrix location exceeds the localization evidence in its experimental donor paper.

Reason: The current mouse donor match is IDA to PMID:16103131. Its actual Methods and Results measure respiratory proteins in mitochondrial fractions and muscle sections. The Discussion explicitly reports that Cox10-specific antibody attempts failed to produce a specific immunoblot signal; Cox1p was used as a functional proxy. These observations support mitochondrial physiology but do not locate Cox10 in the gel-like matrix lumen defined by GO:0005759. This is a source-specific specificity judgment, not a claim that a membrane enzyme cannot expose domains to the matrix.

Summary: Human immunofluorescence curation supports mitochondria.

Reason: The current Human Protein Atlas record authenticates COX10/ENSG00000006695 and reports mitochondria as the main ICC/IF location with supported reliability. Additional nucleolar/cytosolic observations are retained in the evidence record. The present assessment reads the returned experimental summary rather than independently interpreting the images, and does not infer isoform-specific localization.

Summary: Refine the broad Reactome pathway annotation to heme A biosynthesis.

Reason: The current R-HSA-189451 hasEvent edge directly reaches R-HSA-2995330; that reaction has a catalyst-to-physical-entity-to-UniProt chain ending at Q12887. Thus COX10 is a catalyst in the pathway, not merely a named neighbor. Its specific contribution is heme O production within heme A biosynthesis. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: heme A biosynthetic process

Summary: Reactome explicitly assigns COX10 the heme O synthesis reaction.

Reason: The actual R-HSA-2995330 catalyst activity is GO:0008495 and its physical entity resolves to Q12887, with end coordinate 443. The reaction converts protoheme and farnesyl diphosphate to heme O. This is curated reaction evidence, not a new experimental assay or evidence of activity for all alternative products.

Summary: Retain the experimentally curated broad mitochondrial location.

Reason: The genuine full MitoCoP article describes spatial proteomics, subtractive profiling and importomics, and explicitly discusses COX10 among mitochondrial disease proteins. Its final inventory also includes a literature-supported pool. The complete target row of supplementary Table S1 was not recovered, so a particular COX10 classification pool or quantitative signal is not claimed. The broad mitochondrial annotation is consistent with the independent human HPA record and established COX10 biology; incomplete supplement access is not grounds to overrule the HTP curator.

Summary: Correct the reaction annotation to protoheme IX farnesyltransferase activity.

Reason: The historical GO:0004311 label is archived unchanged in the evidence index; the displayed label now matches the current definition, geranylgeranyl diphosphate synthase activity, converting farnesyl diphosphate plus isopentenyl diphosphate to free geranylgeranyl diphosphate. Neither that reaction nor the historical squalene-synthase-like label describes heme O synthesis. The actual original paper cloned a 443-residue human cDNA and tested respiratory rescue of yeast cox10 mutants, with segregation/retransformation controls and sequence comparison. This supports correction to GO:0008495; it is not purified human enzymology or a wrong-gene inference from the title. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: protoheme IX farnesyltransferase activity

Summary: The source establishes cofactor biosynthesis and assembly consequences, not cytochrome-complex membership.

Reason: The original human patient study measures HPLC heme content, COX activity and respiratory-subunit immunoblots, with rescue by COX10 expression. Those assays establish a heme A biosynthetic defect and impaired holoenzyme accumulation. They do not demonstrate that COX10 is a constituent of the cytochrome-containing protein complex defined by GO:0070069. Transient association with assembly machinery is not excluded. The term itself is not restricted to respiratory complexes.

Summary: The reaction catalyst is explicitly located in the mitochondrial inner membrane.

Reason: The current Reactome catalyst entity 2995338 resolves to Q12887 and has the inner-membrane compartment. The event also lists matrix because it includes other reaction participants; that event-level compartment list is not a separate matrix localization assay for COX10.

Summary: The inferred mitochondrial location agrees with the source and independent target evidence.

Reason: The original patient-fibroblast study examines mitochondrial respiratory complex IV assembly. Its inference of COX10 mitochondrial location is consistent with heme A biosynthesis and independent human HPA evidence. The native-gel blots detect respiratory subunits; they are not direct imaging of COX10 or a matrix-specific location assay.

Summary: Human loss-of-function and rescue establish heme A pathway participation.

Reason: Patient muscle and fibroblast mitochondria had reduced heme A. Human COX10 transduction raised patient fibroblast heme A toward control levels as well as rescuing COX activity and holoenzyme abundance. COX10 performs the heme O-producing step of the pathway; the source does not require attributing the later COX15 oxidation to COX10.

Summary: Retain the experimentally grounded complex IV assembly context.

Reason: The actual human fibroblast study found diminished holoenzyme and altered early subassemblies in the COX10-deficient patient. The authors infer an assembly order from these patterns. This supports the existing IMP physiological assembly assertion, while COX10 directly catalyzes heme O synthesis. The observations do not identify COX10 as an insertase or prove direct physical assembly of respiratory subunits, so assembly remains outside the direct core process list.

Summary: Correct the historical reaction term to the specific heme farnesyltransferase.

Reason: The historical GO:0004311 label is archived unchanged in the evidence index; the displayed label now matches the current definition, geranylgeranyl diphosphate synthase activity, converting farnesyl diphosphate plus isopentenyl diphosphate to free geranylgeranyl diphosphate. Neither that reaction nor the historical squalene-synthase-like label describes heme O synthesis. The actual original paper cloned a 443-residue human cDNA and tested respiratory rescue of yeast cox10 mutants, with segregation/retransformation controls and sequence comparison. This supports correction to GO:0008495; it is not purified human enzymology or a wrong-gene inference from the title. The replacement term is already annotated on COX10; this corrects the source-specific row and adds no NEW assertion.

Proposed replacements: protoheme IX farnesyltransferase activity