COA4 (cytochrome c oxidase assembly factor 4 homolog, formerly CHCHD8/E2IG2) is a small (87 aa) twin CX9C protein of the mitochondrial intermembrane space. Its CHCH fold is stabilized by two disulfides and it is imported by the MIA40/CHCHD4 disulfide-relay pathway characteristic of this family. COA4 acts as an accessory factor in the mitochondrial copper delivery pathway to cytochrome c oxidase (complex IV), where it binds the intermembrane-space-facing domain of the copper metallochaperone COX11 and is required for COX11-dependent metallation of the CuB site of the mtDNA-encoded subunit COX1. It is not itself a copper metallochaperone - it lacks the copper-coordinating cysteine motif of COX17, and its loss produces a milder defect than loss of the bona fide chaperones. Loss of COA4 destabilizes COX11, lowers mitochondrial copper without affecting iron, zinc or manganese, sensitizes COX1 to copper limitation, depletes complex IV-containing supercomplexes, and reduces respiration. The function is conserved - human COA4 complements the yeast coa4-null mutant, and yeast coa4 deletion causes cytochrome c oxidase deficiency and reduced mitochondrial copper. No catalytic activity has been established for COA4, and its precise molecular role - stabilizing COX11, loading it with copper, or gating handoff to COX1 - remains undetermined.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0033617 mitochondrial respiratory chain complex IV assembly | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic propagation of a complex IV assembly role across the COA4 family, seeded from the experimentally characterized yeast COA4 (SGD:S000004208). This is the core biological process for COA4 and is now independently supported by human loss-of-function data. Reason: The IBD node placement is sound - the COA4 family is narrow, and the yeast ortholog has two decades of experimental support for a cytochrome c oxidase assembly role. Human-specific confirmation now exists - CRISPR COA4 knockout in MCH58 fibroblasts causes a drastic, specific loss of complex IV-containing supercomplexes and reduced oxygen consumption, and human COA4 complements the yeast coa4-null mutant, establishing that the inherited function is retained. This is the core function of the gene. Supporting Evidence: PMID:20624914 Cells lacking Coa4 are depressed in CcO activity but show no impairment in Cox1 maturation or formation of the Shy1-stabilized Cox1 assembly intermediate PMID:35666203 we demonstrate that human COA4 can replace the function of yeast Coa4 indicating its evolutionarily conserved role DOI:10.1038/s41467-026-77112-z reduction in the levels of complex IV-containing supercomplexes in |
| GO:0005758 mitochondrial intermembrane space | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic propagation of intermembrane space localization. Correct and specific; COA4 is a twin CX9C protein imported by the MIA disulfide relay, and the yeast ortholog is experimentally localized to the IMS. Reason: The twin CX9C motif and the two intramolecular disulfides are the sequence signature of MIA40-dependent IMS import, and the yeast ortholog is directly localized there. The is_active_in qualifier is appropriate - COA4 acts in the IMS, contacting the IMS-facing domain of COX11. Supporting Evidence: PMID:20624914 Coa4 is a twin CX 9 C motif mitochondrial protein localized in the intermembrane space and associated with the inner membrane PMID:19703468 which we demonstrated to be dependent for import...on the Mia40-Erv1 disulfide relay |
| GO:0005739 mitochondrion | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: Automated organelle-level localization from UniProt subcellular location and ARBA. Correct but subsumed by the more specific intermembrane space annotation. Reason: Not wrong, but uninformative next to GO:0005758. Retained as a correct parent term rather than removed, since the organelle assignment is independently well established. |
| GO:0005758 mitochondrial intermembrane space | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro-based localization from the Coa4-like family signature IPR039870. Consistent with the IBA and with the twin CX9C/MIA import route. Reason: The InterPro2GO mapping for the Coa4-like family is appropriate; family membership genuinely predicts IMS localization for this narrow, well-behaved family. Corroborated independently by the yeast localization data. Supporting Evidence: PMID:20624914 Coa4 is a twin CX 9 C motif mitochondrial protein localized in the intermembrane space and associated with the inner membrane |
| GO:0033617 mitochondrial respiratory chain complex IV assembly | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro-based process assignment from the Coa4-like family signature IPR039870, duplicating the IBA call. Reason: Correct, and the family signature is a reliable predictor here because the Coa4-like family is narrow and functionally homogeneous. Redundant with the IBA row but not erroneous. |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MARK AS OVER ANNOTATED | Summary: Affinity-purification interaction with COX11 (UniProtKB:Q9Y6N1) captured in the BioPlex proteome-scale network. The interaction itself is real and biologically central - it has since been confirmed by reciprocal targeted co-immunoprecipitation in both yeast and human cells - but the GO term records none of that. Reason: Per curation guidelines, GO:0005515 protein binding conveys no molecular function and should not stand as an MF annotation. The informative content of this row - that COA4 binds the copper metallochaperone COX11 - is captured by the complex IV assembly process annotation and in core_functions, and the partner is preserved in supporting_entities. Note this is NOT a case of a weak or spurious high-throughput hit - the same interaction was independently predicted by AlphaFold-Multimer and validated by reciprocal co-IP (DOI:10.1038/s41467-026-77112-z). The objection is to the term, not the evidence. No supporting_text is attached: the cached record is a proteome-scale study that does not name COA4 in its text - the assignment lives in its supplementary tables - so no verbatim quotation can carry this claim. The partner recorded in supporting_entities is the informative content here. |
| GO:0005739 mitochondrion | HTP PMID:34800366 Quantitative high-confidence human mitochondrial proteome an... | KEEP AS NON CORE | Summary: High-throughput mitochondrial proteome membership. Correct but generic, and less specific than the IMS annotations. Reason: A quantitative high-confidence mitochondrial proteome is good evidence for organelle membership, but the sub-organelle assignment from the twin CX9C family and the yeast data is more informative. |
| GO:0005739 mitochondrion | IDA PMID:23676665 Protein import and oxidative folding in the mitochondrial in... | KEEP AS NON CORE | Summary: Direct assay localization from a study of MIA40-dependent import and oxidative folding in intact mammalian cells; also the source of the UniProt disulfide-bond assignments for COA4. Reason: An experimental localization from curators who read the full text, and consistent with everything else known about this protein. Retained rather than removed; marked non-core only because GO:0005758 is the specific compartment. Note the cached text of this reference does not name COA4 in the body - the supporting data are in the study's large-scale substrate analysis - so the curator assignment is not second-guessed here. |
Loading supporting content…
Download this section (compressed HTML)Q: Should COA4 carry an intracellular copper ion homeostasis annotation (GO:0006878)? Reduced mitochondrial or cellular copper is reproducible in yeast coa4-null cells and in human COA4 knockout cells (with iron, zinc and manganese unaffected), but it may be a downstream consequence of failed cytochrome c oxidase metallation rather than a separate role. Comparable copper deficiency is seen in SCO1, SCO2, COX10 and COX15 patient fibroblasts, where it is attributed to CTR1 turnover or mislocalization.
Suggested experts: GO copper/metal curation group, Gohil lab
Q: What is COA4's actual molecular function? No catalytic activity is established and metallochaperone activity is positively excluded. Is it stabilizing COX11, loading COX11 with copper, or gating the handoff to COX1? None of the existing data distinguish these, and no current GO molecular function term expresses the accessory role well.
Suggested experts: GO molecular function curation group
Q: Does the reciprocal abundance regulation between COA4 and COX11 warrant a protein stabilization annotation, or is the COX11 loss in COA4-knockout cells a secondary consequence of a stalled assembly intermediate being degraded?
Suggested experts: Gohil lab
Experiment: Measure copper occupancy of immunopurified COX11 from wild-type and COA4-knockout cells by ICP-MS, normalizing to recovered COX11 protein, so that metallation is separated from abundance.
Hypothesis: COA4 is required for copper loading of COX11 rather than for COX11 stability per se.
Type: biochemical
Experiment: Quantify CTR1/SLC31A1 abundance and plasma-membrane localization in COA4-knockout cells, and test whether the copper deficit is phenocopied by knockout of an unrelated late complex IV assembly factor.
Hypothesis: The reduced cellular copper in COA4-deficient cells is secondary to cytochrome c oxidase failure, not a direct COA4 role in copper homeostasis.
Type: cell biology
Experiment: Map Coa4 phosphosites in yeast, test phospho-null and phospho-mimetic mutants for respiratory growth and cytochrome c oxidase assembly, and assay whether the Coa4-Cox11 interaction is phosphorylation-dependent.
Hypothesis: Ptc5-mediated dephosphorylation regulates Coa4 activity in copper delivery.
Type: genetic
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)