COL4A4 encodes the alpha4 chain of type IV collagen, a secreted structural component of basement membranes. It assembles with alpha3 and alpha5 chains into collagen IV protomers whose terminal interactions organize the glomerular basement-membrane network. The collagenous region and C-terminal NC1 domain contribute to this architecture; COL4A4 variants are associated with thin-basement-membrane nephropathy and Alport-spectrum kidney disease. Isolated recombinant human NC1, termed tetrastatin, binds integrin alphaVbeta3 and inhibits melanoma-cell growth and invasion in experimental systems. A synthetic C-terminal peptide, QS-13, binds alpha5beta1 and reduces endothelial migration and experimental vascularization, a fragment activity distinct from the intact structural chain.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005201 extracellular matrix structural constituent | IEA GO_REF:0000120 | MODIFY | Reason: The combined ortholog/domain assignment correctly identifies a matrix structural protein. Refine it to the collagen-specific structural term already supported by the target IBA and human NC1 assembly evidence. GO:0030020 explicitly permits extracellular-matrix collagen proteins; this is a type IV network contribution, not an assertion that this study measured the force response of isolated alpha4 chains or that COL4A4 is fibrillar collagen. Proposed replacements: extracellular matrix structural constituent conferring tensile strength |
| GO:0005201 extracellular matrix structural constituent | IMP PMID:17942953 COL4A3/COL4A4 mutations producing focal segmental glomerulos... | MODIFY | Reason: The complete genuine PMID:17942953 abstract explicitly includes COL4A4 as well as COL4A3 in families with thin basement membranes, hematuria and renal disease. Full original methods were not recovered, and the abstract leaves progression versus modifier explanations open. The established physical role of the alpha4-containing collagen IV network nevertheless supports a more specific collagen structural term. Retain the positive clinical evidence without presenting it as a direct tensile-strength assay. Proposed replacements: extracellular matrix structural constituent conferring tensile strength |
| GO:0005201 extracellular matrix structural constituent | IMP PMID:19675380 Identification of novel variants in the COL4A4 gene in Korea... | MODIFY | Reason: The original publisher PDF studies 45 Korean patients with biopsy-confirmed thin-basement-membrane nephropathy and 143 controls, using COL4A4 sequencing and family/clinical comparisons. G199R segregation and the proposed G1606E interpretation support a matrix-disease link, but the authors call for functional confirmation and do not directly measure network mechanics. Refine the established structural role to the collagen-specific term using independent NC1/complex evidence; do not declare every detected variant pathogenic. Proposed replacements: extracellular matrix structural constituent conferring tensile strength |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-1454757 | ACCEPT | Reason: The gelatin-degradation event contains extracellular collagen-derived material, including a current entity path to COL4A4. Metalloproteinases perform the proteolysis; the chain is the extracellular substrate. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-1454757 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-1564142 | ACCEPT | Reason: The collagen-IV degradation event contains extracellular alpha3/alpha4/alpha5 network material. Preserve the extracellular location without assigning metalloproteinase activity to its structural substrate. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-1564142 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2213207 | ACCEPT | Reason: The network-formation event places alpha4-containing type IV protomers in extracellular networks. The chain supplies matrix structure through association, rather than functioning as an assembly enzyme. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2213207 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2327695 | ACCEPT | Reason: The integrin association event contains collagen IV as one of several extracellular collagen ligands. Its exact current entity path reaches COL4A4, but this set membership does not by itself establish an isolated alpha4 interaction with every listed integrin. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2327695 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2327733 | ACCEPT | Reason: The FN1 event includes extracellular collagen IV-containing material. The verified target entity path supports location; the family-level event is not an alpha4-specific affinity measurement. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2327733 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2327738 | ACCEPT | Reason: The DDR1 event includes type IV collagen among extracellular ligands. Retain the location without confusing a type IV network with a fibrillar collagen or assigning receptor kinase activity to COL4A4. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2327738 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2396234 | ACCEPT | Reason: The collagen VII event connects extracellular anchoring material with collagen IV-containing networks. The target is a chain within the collagen IV participant, not a demonstrated autonomous receptor or enzyme. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2396234 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2426450 | ACCEPT | Reason: The laminin/nidogen event places the alpha4-containing type IV network in extracellular matrix. Current complex-membership paths substantiate the target location without implying that each isolated alpha4 chain directly contacts every component. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2426450 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-2559639 | ACCEPT | Reason: The sulfilimine-crosslink event is extracellular and includes alpha4-containing collagen IV. The matrix supplies residues and architecture; peroxidasin generates the oxidizing chemistry. This location is not a COL4A4 catalytic annotation. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-2559639 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-375151 | ACCEPT | Reason: The NCAM1 event represents extracellular collagen IV-containing material. Its target path supports extracellular location, not an independently established neuronal signaling function of COL4A4. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-375151 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-382054 | ACCEPT | Reason: The PDGF association event places collagen-containing material outside cells. Retain the location while distinguishing broad matrix-participant membership from a direct assay of purified human alpha4 binding to PDGF. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-382054 |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-4088218 | ACCEPT | Reason: The tetrastatin event describes the extracellular NC1 fragment binding alphaVbeta3. Direct recombinant human NC1 binding in PMID:22539938 independently supports this fragment context. Preserve the source location without equating the soluble experimental NC1 preparation with all forms of intact collagen IV. The complete cached event summary and current participant traversal were inspected. |
| GO:0005576 extracellular region | TAS Reactome:R-HSA-9638628 | ACCEPT | Reason: The DraE event places host collagen IV extracellularly as a bacterial adhesin ligand. The verified COL4A4 entity path does not make the human matrix chain a bacterial adhesin. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-HSA-9638628 |
| GO:0005576 extracellular region | TAS Reactome:R-NUL-4084497 | ACCEPT | Reason: This distinct R-NUL source identifier is preserved. Its current record reports Homo sapiens and its collagen VII/collagen IV participant paths reach P53420 in extracellular material. The identifier prefix is not used to invent a nonhuman assay or to collapse this assertion with the separate R-HSA event. The complete cached event summary and current participant traversal were inspected. Supporting Evidence: Reactome:R-NUL-4084497 |
| GO:0005581 collagen trimer | IEA GO_REF:0000002 | MODIFY | Reason: The generic collagen-trimer domain assignment can be refined to the type IV heterotrimer. Independent alpha3/alpha4/alpha5 NC1 composition and association evidence supports alpha4 membership; it does not imply an alpha4 homotrimer. The noncanonical alpha4 crystal assembly in PMID:30443360 is kept separate from native protomer composition. Proposed replacements: collagen type IV trimer |
| GO:0005587 collagen type IV trimer | IBA GO_REF:0000033 | ACCEPT | Reason: The PAINT type IV trimer placement agrees with target alpha3/alpha4/alpha5 composition and current same-term mouse and human records. Human COL4A4 in WITH/FROM is legitimate descendant grounding, not circular evidence. The historical IBD node was not reconstructed; neither donor count nor a current annotation inventory is used to claim a node-placement error. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN008601296 UNRESOLVED The original PAINT ancestral placement is preserved. Current resolved descendant identities and same-term records corroborate collagen type IV trimer; the historical IBD/tree/MSA was not reconstructed, so no source-node error or circularity is asserted. |
| GO:0005587 collagen type IV trimer | IDA PMID:7523402 Complete primary structure of the human type IV collagen alp... | ACCEPT | Reason: The complete original abstract establishes the human 1,690-residue alpha4(IV) chain and its collagenous/NC1 organization, but is primarily sequence/expression work and is not described as a new trimer assay. Type IV heterotrimer membership is independently supported by alpha3/alpha4/alpha5 NC1 characterization in PMID:12193605. Retain the curator assertion with these distinct access and experimental scopes. |
| GO:0005587 collagen type IV trimer | IEA GO_REF:0000107 | ACCEPT | Reason: The mouse Q9QZR9 ortholog identity and current same-term records agree with established human type IV heterotrimer membership. Preserve the Ensembl and UniProt source identifiers; current supporting records corroborate the class without claiming to reconstruct every historical electronic transfer. Supporting Evidence: |
| GO:0005587 collagen type IV trimer | NAS PMID:29853175 Collagen Fibril Assembly and Function. | ACCEPT | Reason: Type IV trimer membership is well established independently, including the positive complete PMID:12193605 abstract. The original PMID:29853175 review is available only as an abstract despite publisher/repository recovery attempts, so its precise target passage remains unverified. Retain this correct component assignment on integrated evidence rather than treating missing access as a miscitation. |
| GO:0005604 basement membrane | IDA PMID:2211832 Molecular heterogeneity of basal laminae: isoforms of lamini... | ACCEPT | Reason: The complete original abstract explicitly reports subunit-specific antibody localization of alpha4(IV) in basal laminae, including synaptic and renal compartments. Its abstract does not state the experimental species, and the accessible PMC body is only a PDF notice; no species-specific full-methods claim is made. This positive localization agrees with independent alpha4-containing glomerular basement-membrane architecture. |
| GO:0005604 basement membrane | IEA GO_REF:0000120 | ACCEPT | Reason: The combined basement-membrane mapping is consistent with the alpha4-containing collagen IV network and positive subunit-localization evidence. Current Q9QZR9 same-term records corroborate the ortholog context. Basement membrane is extracellular matrix, not a lipid bilayer; the source ARBA, Ensembl and subcellular-location identifiers remain unchanged. |
| GO:0005788 endoplasmic reticulum lumen | IEA GO_REF:0000117 | KEEP AS NON CORE | Reason: The secreted precursor occupies the ER lumen during synthesis and assembly before entering basement membrane. Retain this biosynthetic location as non-core relative to the mature matrix and isolated extracellular fragment activities; it does not assign ER chaperone or export-machine activity. Supporting Evidence: Reactome:R-HSA-2022073 Reactome:R-HSA-2089971 |
| GO:0005788 endoplasmic reticulum lumen | TAS Reactome:R-HSA-2022073 | KEEP AS NON CORE | Reason: The complete Reactome triple-helix formation event and literal current component/member paths reach P53420 in ER-luminal collagen material. This supports precursor localization during biosynthesis. The chain provides or carries the collagen structure, while other proteins perform processing or transport; retain the original location without assigning their catalytic/export functions. Supporting Evidence: Reactome:R-HSA-2022073 |
| GO:0005788 endoplasmic reticulum lumen | TAS Reactome:R-HSA-2089971 | KEEP AS NON CORE | Reason: The complete Reactome collagen secretion event and literal current component/member paths reach P53420 in ER-luminal collagen material. This supports precursor localization during biosynthesis. The chain provides or carries the collagen structure, while other proteins perform processing or transport; retain the original location without assigning their catalytic/export functions. Supporting Evidence: Reactome:R-HSA-2089971 |
| GO:0005788 endoplasmic reticulum lumen | TAS Reactome:R-HSA-8944265 | KEEP AS NON CORE | Reason: The complete Reactome type IV procollagen association event and literal current component/member paths reach P53420 in ER-luminal collagen material. This supports precursor localization during biosynthesis. The chain provides or carries the collagen structure, while other proteins perform processing or transport; retain the original location without assigning their catalytic/export functions. Supporting Evidence: Reactome:R-HSA-8944265 |
| GO:0030020 extracellular matrix structural constituent conferring tensile strength | IBA GO_REF:0000033 | ACCEPT | Reason: The PAINT structural classification agrees with the type IV network role and GO:0030020 explicitly allows extracellular-matrix collagen proteins. Human alpha1(IV) and worm let-2 source identities and current same-term records were resolved; their presence does not convert IBA into pairwise transfer, and the original ancestral node was not independently reconstructed. Actual alpha4 NC1 contacts support structural assembly without implying fibrillar-collagen mechanics or intrinsic crosslink catalysis. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN001166351 UNRESOLVED The original PAINT ancestral placement is preserved. Current resolved descendant identities and same-term records corroborate extracellular matrix structural constituent conferring tensile strength; the historical IBD/tree/MSA was not reconstructed, so no source-node error or circularity is asserted. |
| GO:0030020 extracellular matrix structural constituent conferring tensile strength | RCA PMID:25037231 Extracellular matrix signatures of human primary metastatic ... | ACCEPT | Reason: The original human ECM proteomics workbook contains COL4A4/P53420 in both the complete MS dataset and the Matrisome sheet, classified as a core-matrisome collagen. This RCA source supplies classification and protein detection, not a direct force measurement. The collagen-specific structural activity is supported by the integrated target network/NC1 evidence and the explicit GO collagen convention. |
| GO:0030199 collagen fibril organization | NAS PMID:29853175 Collagen Fibril Assembly and Function. | UNDECIDED | Summary: The available review abstract discusses fibril assembly; the COL4A4-specific basis of this NAS assertion remains unverified. Reason: GO:0030199 concerns processes determining collagen fibril size and arrangement, and is a child of extracellular matrix organization. COL4A4 forms a type IV network, but not being a fibrillar collagen does not establish that it cannot influence fibrils. The complete PMID:29853175 abstract describes fibril assembly generally and does not identify a COL4A4-performed step; the original full text remains unavailable after the documented retrieval failures. Retain source-specific uncertainty rather than infer either participation or contradiction from collagen class alone. Independently established type IV trimer membership is a separate claim. This differs from the older merged COL4A1 REMOVE judgment because that collagen-class-based exclusion exceeds the available source evidence; the reopened COL4A1 assessment likewise retains UNDECIDED for this source. Supporting Evidence: |
| GO:0031012 extracellular matrix | HDA PMID:25037231 Extracellular matrix signatures of human primary metastatic ... | ACCEPT | Reason: Exact COL4A4/P53420 rows were read in the original human colon/liver ECM mass-spectrometry workbook, after the paperβs sequential extraction and insoluble-ECM enrichment. This directly supports protein detection in the sampled ECM fraction. Shared collagen biology and the basement-membrane network are consistent, but the proteomic measurement does not demonstrate cancer causality, secretion mechanism or an isolated mechanical activity. Supporting Evidence: |
| GO:0031012 extracellular matrix | IBA GO_REF:0000033 | ACCEPT | Reason: The ancestral extracellular-matrix placement is consistent with direct human target proteomics and collagen IV architecture. All listed extant donor identities were resolved to literal current same-term records, whose evidence types vary; they are corroboration, not a replacement reconstruction of the historical IBD. The target among the descendants is legitimate positive grounding. No divergence or loss that would overturn matrix placement was established. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN008390299 UNRESOLVED The original PAINT ancestral placement is preserved. Current resolved descendant identities and same-term records corroborate extracellular matrix; the historical IBD/tree/MSA was not reconstructed, so no source-node error or circularity is asserted. |
| GO:0032836 glomerular basement membrane development | IBA GO_REF:0000033 | ACCEPT | Reason: The alpha4 chain contributes the physical alpha3/alpha4/alpha5 network from which the glomerular basement membrane is formed and maintained. Current mouse and human same-term records support the context, while the historical PAINT node remains uninspected. Retain the inherited structural participation rather than treating the chain as merely a passive processing substrate or the targetβs own descendant evidence as circular. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN008601295 UNRESOLVED The original PAINT ancestral placement is preserved. Current resolved descendant identities and same-term records corroborate glomerular basement membrane development; the historical IBD/tree/MSA was not reconstructed, so no source-node error or circularity is asserted. |
| GO:0032836 glomerular basement membrane development | IMP PMID:17942953 COL4A3/COL4A4 mutations producing focal segmental glomerulos... | ACCEPT | Reason: The source explicitly includes COL4A4-associated thin basement membranes and renal phenotypes in human families. Its complete abstract supports the positive matrix-disease association, but not every variant-specific developmental mechanism, and full access remains unavailable. Holistically, alpha4 supplies the glomerular basement-membrane network structure; retain the established process contribution with curator deference rather than infer an unsupported source error or an independent catalytic developmental activity. |
| GO:0032836 glomerular basement membrane development | IMP PMID:19675380 Identification of novel variants in the COL4A4 gene in Korea... | ACCEPT | Reason: The original full human study uses biopsy-confirmed thin-basement-membrane nephropathy, COL4A4 sequencing and segregation/clinical comparisons. It supports the structural disease context; it does not directly time embryonic basement-membrane formation or establish functional causality for every variant. Retain the existing glomerular basement-membrane contribution on the independently established material role of the alpha4-containing network, with that limitation on the particular IMP study. |
| GO:0060090 molecular adaptor activity | EXP PMID:30443360 Structures of collagen IV globular domains: insight into ass... | MODIFY | Reason: Actual human NC1 constructs and crystallographic interfaces establish structural assembly information. Isolated alpha4 eluted as a monomer and crystallized in a noncanonical dodecamer whose physiological relevance the authors question; mixed alpha3/alpha4/alpha5 yielded alpha5 crystals, and SPR concerned alpha3/alpha5. This supports the collagen structural role but not a separately established alpha4 molecular-adaptor mechanism or native alpha345 reconstitution in this experiment. Refine to the existing collagen-specific structural activity without discarding the positive target structural data. Proposed replacements: extracellular matrix structural constituent conferring tensile strength |
| GO:0005178 integrin binding | IDA PMID:22539938 Tetrastatin, the NC1 domain of the Ξ±4(IV) collagen chain: a ... | NEW | Summary: Sequence-checked recombinant human NC1 binds immobilized alphaVbeta3 by SPR; an integrin-blocking antibody partly reduces melanoma-cell adhesion to coated NC1. Reason: Recombinant human NC1, termed tetrastatin and reported as NC1(1β229), directly bound purified immobilized alphaVbeta3 in surface plasmon resonance assays; EDTA dissociated the complex and receptor-blocking antibody partially reduced melanoma-cell adhesion. The construct was amplified from human kidney mRNA and expressed with a bacterial affinity tag. This is an isolated-domain ligand activity, not a claim that every intact precursor state binds the receptor, that the insert exactly equals the current annotated NC1 interval, or that natural release was demonstrated. No original annotation or final replacement covers this binding activity. Synthetic QS-13 binds a different integrin, alpha5beta1, in separate experiments and is kept at its precise sequence/pharmacological scope. Supporting Evidence: PMID:22539938 The NC1 Ξ±4(IV) domain bound to the immobilized Ξ±vΞ²3 integrin |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Which COL4A4 NC1 fragments are released naturally in human basement membranes, at what concentrations, and do they retain the integrin interactions measured with recombinant tetrastatin?
Q: Is the QS-13 sequence liberated in a stable receptor-active form in vivo, and does it regulate vascularization at endogenous concentrations rather than only after synthetic-peptide administration?
Q: How do native alpha3/alpha4/alpha5 NC1 contacts differ from the noncanonical isolated alpha4 crystal assembly, and which disease substitutions perturb protomer assembly versus later network incorporation?
Q: What specific COL4A4-performed mechanism, if any, controls collagen fibril size or arrangement in the source underlying the existing NAS annotation, as distinct from type IV network formation?
Experiment: Use sequence-resolved mass spectrometry of human basement-membrane turnover samples to identify endogenous COL4A4 fragments, then test purified native and matched recombinant preparations in receptor-controlled binding assays.
Hypothesis: Natural cleavage products differ in abundance or receptor specificity from the full recombinant NC1 and synthetic QS-13 preparations.
Experiment: Reconstitute sequence-defined human alpha3/alpha4/alpha5 protomers with wild-type or disease-associated alpha4, measuring incorporation, NC1 contacts and network ultrastructure separately from crosslink chemistry supplied by peroxidasin.
Hypothesis: Some alpha4 variants disrupt chain assembly or network architecture without conferring a catalytic processing defect on COL4A4.
Experiment: Compare full NC1 and exact QS-13 in receptor-loss/rescue endothelial assays, separating adhesion, migration, proliferation and vascularization endpoints at matched exposure.
Hypothesis: The synthetic C-terminal ligand and full NC1 have distinct receptor and endothelial-response profiles.
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)