COG4 is a noncatalytic subunit of the eight-subunit conserved oligomeric Golgi tethering complex. It helps assemble the complex and coordinates the Sec1/Munc18 protein SCFD1/Sly1 with the Golgi SNARE syntaxin 5 through distinct binding surfaces. These interactions support retrograde trafficking and maintenance of Golgi architecture, enabling the correct distribution of glycosylation machinery. COG4 occupies both cytosolic and peripheral Golgi-membrane pools. Biallelic loss-of-function variants cause a congenital disorder of glycosylation, whereas a recurrent dominant variant causes Saul-Wilson syndrome with a different trafficking disturbance.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000139 Golgi membrane | IEA GO_REF:0000044 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The UniProt Golgi-membrane mapping agrees with endogenous human COG4 at the Golgi and recruitment of replacement COG4 to the Golgi in the HEK293T knockout-rescue study. This is peripheral association with the cytoplasmic membrane surface, not a transmembrane segment. |
| GO:0000139 Golgi membrane | NAS PMID:27066481 COG Complex Complexities: Detailed Characterization of a Com... | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The cited human HEK293T study observes replacement COG4 at the Golgi and documents COG-dependent Golgi architecture. The separate human fibroblast fractionation and microscopy study independently supports membrane-associated COG4 while retaining its larger cytosolic pool. Supporting Evidence: PMID:27066481 each replacement COG subunit was observed on the Golgi |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-6809006 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-6809010 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-6809011 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-6811433 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-8847544 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-8847635 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000139 Golgi membrane | TAS Reactome:R-HSA-8849748 | ACCEPT | Summary: COG4 associates with Golgi membranes as a COG-complex subunit. Reason: The exact Reactome event includes Q9H9E3 through the COG complex at the Golgi membrane. Its physical-entity chain resolves through COG4 entity 6808722 and reference 52518. This supports the location without assigning every biochemical operation in the event to the COG4 subunit. |
| GO:0000301 retrograde transport, vesicle recycling within Golgi | IBA GO_REF:0000033 | ACCEPT | Summary: Conserved COG4-dependent intra-Golgi retrograde recycling. Reason: Human COG4 participates in tether-complex assembly and coordinates Sly1/STX5, providing a performed molecular role in retrograde recycling. Current human and yeast same-term records support this interpretation. The historical PAINT tree and IBD placement were not reconstructed; the target in WITH/FROM is legitimate experimental grounding. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000613595 UNRESOLVED The historical IBD node and family alignment were not inspected. Positive current target/donor same-term records and target primary evidence support the retained biological assertion independently; this status does not dispute the PAINT placement. |
| GO:0000301 retrograde transport, vesicle recycling within Golgi | IMP PMID:27066481 COG Complex Complexities: Detailed Characterization of a Com... | ACCEPT | Summary: Human COG4 supports retrograde trafficking within the Golgi. Reason: COG4 knockout is included in the human HEK293T panel, with defective Golgi architecture and glycosylation rescued by replacement subunit. The paper measures retrograde toxin trafficking at pathway level; altered toxin binding limits some probes, and SubAB passage is an integrated endosome/Golgi/ER assay. The specific intra-Golgi role is supported by COG4-mediated organization of the Sly1/STX5 machinery, rather than by attributing glycosyltransferase chemistry to COG4. |
| GO:0005515 protein binding | IPI PMID:15047703 The binary interacting network of the conserved oligomeric G... | MODIFY | Summary: COG4 supplies a structural scaffold within the COG tethering complex. Reason: The original human COG-subunit study goes beyond the positive COG7 pair: its assembly analysis places COG4 at a network of COG1/2/5/7 contacts connecting the subcomplex. This supports the integral-complex scaffold activity. Pairwise reticulocyte-lysate co-immunoprecipitation is not a purified-protein affinity measurement; the functional refinement uses the whole assembly analysis. Proposed replacements: protein complex scaffold activity Supporting Evidence: PMID:15047703 crucial role of COG4 in the assembly/function of the complex |
| GO:0005515 protein binding | IPI PMID:15047703 The binary interacting network of the conserved oligomeric G... | KEEP AS NON CORE | Summary: Positive human COG4–COG2 assembly contact. Reason: The original study tests human subunits produced in rabbit reticulocyte lysate and recovers this pair by reciprocal co-immunoprecipitation. Retain the interaction as a distinct source assertion; the integrated scaffold mechanism is represented in the COG7-source refinement and core. The authors explicitly reserve conclusive direct binding for purified-protein assays. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:15047703 The binary interacting network of the conserved oligomeric G... | KEEP AS NON CORE | Summary: Positive human COG4–COG1 assembly contact. Reason: The original study tests human subunits produced in rabbit reticulocyte lysate and recovers this pair by reciprocal co-immunoprecipitation. Retain the interaction as a distinct source assertion; the integrated scaffold mechanism is represented in the COG7-source refinement and core. The authors explicitly reserve conclusive direct binding for purified-protein assays. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:15047703 The binary interacting network of the conserved oligomeric G... | KEEP AS NON CORE | Summary: Positive human COG4–COG5 assembly contact. Reason: The original study tests human subunits produced in rabbit reticulocyte lysate and recovers this pair by reciprocal co-immunoprecipitation. Retain the interaction as a distinct source assertion; the integrated scaffold mechanism is represented in the COG7-source refinement and core. The authors explicitly reserve conclusive direct binding for purified-protein assays. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:19536132 Direct interaction between the COG complex and the SM protei... | KEEP AS NON CORE | Summary: COG4 and COG7 are associated within the endogenous COG complex. Reason: The cited experiments recover COG4/COG7 association while identifying COG4 as the principal Sly1-contacting subunit. This is positive complex-association evidence, distinct from a claim of purified COG4–COG7 affinity. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:19536132 Direct interaction between the COG complex and the SM protei... | MODIFY | Summary: COG4 binds the Golgi syntaxin STX5. Reason: Human Golgi syntaxin 5 is tested with COG4 fragments in the original study. COG4 residues 1–153 bind STX5, whereas 1–84 retain Sly1 binding but not STX5 binding; E53/71A also separates Sly1 from syntaxin binding. Syntaxin binding is a precise partner-class refinement. It does not assign COG4 the SNARE helical-bundle zippering activity. Proposed replacements: syntaxin binding Supporting Evidence: PMID:19536132 interacts with Syntaxin 5 through a different binding site |
| GO:0005515 protein binding | IPI PMID:19536132 Direct interaction between the COG complex and the SM protei... | KEEP AS NON CORE | Summary: Human COG4 binds the rat Sly1 construct. Reason: The original Methods explicitly use recombinant rSly1 supplied by JC Hay. Purified bacterial His-COG4 (1–231) binds GST-rSly1, establishing a positive cross-species pair even though the current interaction export lacks an exact Q62991 record. Preserve the rat partner identity; do not relabel that construct as human. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:19536132 Direct interaction between the COG complex and the SM protei... | KEEP AS NON CORE | Summary: Human COG4 associates with endogenous human SCFD1/Sly1. Reason: Endogenous HeLa co-immunoprecipitation supports the human SCFD1 association, alongside transfected and recombinant rat-Sly1 assays that resolve the mechanism. COG4 has separable Sly1- and syntaxin-binding surfaces. The positive source-specific human pair is retained without conflating it with the purified rat construct experiment. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with TXNDC9. Reason: The exact HuRI records include validated two-hybrid detection. The official partner is a regulator of the TCP1 chaperonin; a thioredoxin-domain name alone does not establish catalytic redox activity or a more specific binding class. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | MODIFY | Summary: Positive HuRI interaction with SORBS3/vinexin. Reason: The exact positive HuRI pair maps to the official 671-residue displayed Alpha product of SORBS3/vinexin. Its documented modulation of actin stress-fiber organization supports the more specific cytoskeletal regulatory protein binding term, an is_a child of cytoskeletal protein binding. This identifies the bound partner class without assigning COG4 an actin-regulatory role or a mapped regulatory-domain interface. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Proposed replacements: cytoskeletal regulatory protein binding Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with PSMA1. Reason: The positive HuRI partner is the noncatalytic alpha subunit of the proteasome. The official record assigns proteolysis to beta subunits; an isolated PSMA1 pair does not by itself establish binding to the assembled proteasome or an enzyme-binding MF. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with FAM110A. Reason: The verified partner is a centrosomal/spindle-pole protein. The available class evidence establishes localization, without securely establishing the cytoskeletal-component or catalytic class needed for a narrower binding assignment. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with BEX2. Reason: The exact HuRI pair is positive. The official BEX2 record describes context-dependent regulatory associations, but this assay does not identify a catalytic class, a mapped interaction domain, or a COG4-mediated apoptotic mechanism. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with CTNNA3 isoform 2. Reason: The exact Q9UI47-2 partner reconstructs to 516 residues, with a changed 511–516 tail and loss of 517–895. Its positive interaction is retained, but whole-length alpha-catenin actin-linkage functions cannot simply be assigned to this truncated product. A specific cytoskeletal-component refinement remains product-limited. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | MODIFY | Summary: Positive HuRI interaction with FARSA. Reason: The exact HuRI partner is the catalytic alpha subunit of phenylalanyl-tRNA synthetase. Its verified enzyme identity supports enzyme binding at the assay scope. No tRNA aminoacylation activity or physiological translation function is attributed to COG4. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Proposed replacements: enzyme binding Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:32296183 A reference map of the human binary protein interactome. | KEEP AS NON CORE | Summary: Positive HuRI interaction with ANKRD11 fragment X5D778. Reason: The source maps specifically to the 421-residue X5D778 fragment. Its sequence-derived ROGID matches the interaction record. The pair is retained without transferring full-length ANKRD11 transcriptional-corepressor activity to that fragment. All corresponding HuRI COG4 records specify the 253–785 sufficient-binding fragment. Dataset-level MAPPIT/GPCA validation is not asserted for this individual pair. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | KEEP AS NON CORE | Summary: BioPlex recovers COG4 association with COG7. Reason: Exact positive anti-tag co-immunoprecipitation records support this COG-subunit association in the assayed human cell context. AP-MS establishes co-purification, not necessarily an isolated direct interface. The scaffold core captures the integrated complex role while this independent pair remains represented. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | KEEP AS NON CORE | Summary: BioPlex recovers COG4 association with COG2. Reason: Exact positive anti-tag co-immunoprecipitation records support this COG-subunit association in the assayed human cell context. AP-MS establishes co-purification, not necessarily an isolated direct interface. The scaffold core captures the integrated complex role while this independent pair remains represented. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | KEEP AS NON CORE | Summary: BioPlex recovers COG4 association with COG1. Reason: Exact positive anti-tag co-immunoprecipitation records support this COG-subunit association in the assayed human cell context. AP-MS establishes co-purification, not necessarily an isolated direct interface. The scaffold core captures the integrated complex role while this independent pair remains represented. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MODIFY | Summary: COG4 co-purifies with the E3 ligase MUL1. Reason: The exact BioPlex pair is positive, and the partner sequence checksum matches official MUL1, a RING E3 ubiquitin ligase. Ubiquitin-protein-ligase binding expresses the verified partner class at AP-MS scope; neither MUL1-dependent ubiquitination of COG4 nor a mitochondrial mechanism is established here. Proposed replacements: ubiquitin protein ligase binding Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | KEEP AS NON CORE | Summary: BioPlex recovers COG4 association with COG5. Reason: Exact positive anti-tag co-immunoprecipitation records support this COG-subunit association in the assayed human cell context. AP-MS establishes co-purification, not necessarily an isolated direct interface. The scaffold core captures the integrated complex role while this independent pair remains represented. Supporting Evidence: |
| GO:0005515 protein binding | IPI PMID:40205054 Multimodal cell maps as a foundation for structural and func... | KEEP AS NON CORE | Summary: The U2OS AP-MS network supports a COG2–COG4 association. Reason: The original linked network has the exact Ensembl-matched COG2-bait to COG4-prey edge. This authenticates the source-specific relationship despite its absence from the queried IntAct/MINT exports. Retain co-complex association without assigning all cellular-map predictions to COG4. Supporting Evidence: |
| GO:0005829 cytosol | IEA GO_REF:0000044 | ACCEPT | Summary: COG4 has a substantial cytosolic pool. Reason: The UniProt mapping agrees with direct human fibroblast fractionation: much of COG4 is in cytosolic COG subcomplexes, with a smaller Golgi-membrane-associated pool. Cytosol and Golgi membrane are complementary distributions. Supporting Evidence: PMID:19494034 most is recovered in the cytosol |
| GO:0006890 retrograde vesicle-mediated transport, Golgi to endoplasmic reticulum | IMP PMID:19536132 Direct interaction between the COG complex and the SM protei... | ACCEPT | Summary: COG4 coordinates machinery needed for Golgi-to-ER retrograde traffic. Reason: Human COG4 depletion delays BFA-induced mannosidase II redistribution and KDELR recycling. Wild-type COG4 restores KDELR transport, whereas the Sly1-binding-defective E53/71A mutant does not despite preserved Golgi localization and complex incorporation. This links the performed COG4 recruitment role to the transport process. Supporting Evidence: |
| GO:0007030 Golgi organization | IBA GO_REF:0000033 | ACCEPT | Summary: Conserved COG4 activity maintains Golgi organization. Reason: The current human descendant evidence includes source-specific perturbation and restoration of Golgi-associated machinery. COG4 supplies structural assembly and recruitment within the COG complex, so this is performed organization rather than a phenotype-only association. The historical PAINT node remains uninspected. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN001166029 UNRESOLVED The historical IBD node and family alignment were not inspected. Positive current target/donor same-term records and target primary evidence support the retained biological assertion independently; this status does not dispute the PAINT placement. |
| GO:0007030 Golgi organization | IMP PMID:19536132 Direct interaction between the COG complex and the SM protei... | ACCEPT | Summary: COG4/Sly1 coordination supports Golgi organization. Reason: COG4 knockdown disrupts compact Golgi organization and redistributes Golgi SNAREs; overexpressing the Sly1-binding N-terminal fragment also fragments the Golgi. The binding-separation mutant and rescue assays connect these effects to COG4-mediated molecular assembly. Supporting Evidence: |
| GO:0007030 Golgi organization | IMP PMID:27066481 COG Complex Complexities: Detailed Characterization of a Com... | ACCEPT | Summary: Loss of COG4 disrupts Golgi cisternal organization. Reason: The COG4 knockout in human HEK293T cells shows dilated and fragmented Golgi cisternae by electron microscopy. Light-microscopy localization of several markers can remain perinuclear, so normal-looking marker distribution does not negate the ultrastructural defect. COG4 complex assembly supplies the structural role. Supporting Evidence: |
| GO:0017119 COG complex | IBA GO_REF:0000033 | ACCEPT | Summary: COG4 is a conserved constituent of the COG tethering complex. Reason: Direct human assembly evidence and current same-term yeast, fly and human records support COG-complex membership. The extant donor list is not a reconstruction of the PAINT ancestral node, and the target as its own descendant evidence is expected. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000613595 UNRESOLVED The historical IBD node and family alignment were not inspected. Positive current target/donor same-term records and target primary evidence support the retained biological assertion independently; this status does not dispute the PAINT placement. Supporting Evidence: |
| GO:0017119 COG complex | IDA PMID:15047703 The binary interacting network of the conserved oligomeric G... | ACCEPT | Summary: Human COG4 is an integral COG-complex subunit. Reason: The original human-subunit co-translation and assembly analysis supports COG4 contacts with multiple subunits and its inclusion in the larger complex. Rabbit reticulocyte lysate is the expression/assay system, not the species of the cloned COG4. Supporting Evidence: |
| GO:0017119 COG complex | NAS PMID:27066481 COG Complex Complexities: Detailed Characterization of a Com... | ACCEPT | Summary: COG4 is the lobe-A component tested in the human COG-knockout panel. Reason: The source explicitly disrupts each COG subunit and demonstrates the interdependence of COG3/COG4 and other lobe-A proteins. The original complex-assembly study independently establishes membership, rather than deriving it solely from similar knockout phenotypes. |
| GO:0032588 trans-Golgi network membrane | TAS Reactome:R-HSA-6811431 | ACCEPT | Summary: The modeled COG complex associates with trans-Golgi-network membranes. Reason: The exact Reactome event resolves through TGN COG complex 6814058 to COG4 entity 6813868 and UniProt Q9H9E3. Retain this location at the curated pathway scope. Reactome explicitly cautions that joint GARP/COG capture of a single vesicle is not established; this annotation does not assert that mechanism. Supporting Evidence: Reactome:R-HSA-6811431 |
| GO:0032588 trans-Golgi network membrane | TAS Reactome:R-HSA-6814682 | ACCEPT | Summary: The modeled COG complex associates with trans-Golgi-network membranes. Reason: The exact Reactome event resolves through TGN COG complex 6814058 to COG4 entity 6813868 and UniProt Q9H9E3. Retain this location at the curated pathway scope. Reactome explicitly cautions that joint GARP/COG capture of a single vesicle is not established; this annotation does not assert that mechanism. Supporting Evidence: Reactome:R-HSA-6814682 |
Loading supporting content…
Download this section (compressed HTML)Q: How do the short Cog4S and N-terminally truncated COG4 products differ in complex incorporation and Sly1/STX5 coordination?
Q: How is COG4-mediated Sly1 recruitment coupled to SNARE assembly during an individual tethering/fusion event?
Experiment: Compare sequence-verified COG4 products at matched expression levels in COG4-deficient human cells, measuring complex incorporation, Sly1/STX5 association and retrograde cargo transport rather than glycan phenotypes alone.
Hypothesis: Product-specific loss of interaction regions uncouples complex assembly from Sly1-dependent trafficking.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)