AP5S1 encodes sigma-5 (also called C20orf29), the 200-residue small subunit of AP-5, the fifth and most divergent of the heterotetrameric adaptor protein complexes. AP-5 pairs two large subunits, zeta-5 (AP5Z1) and beta-5 (AP5B1), with a medium subunit mu-5 (AP5M1) and sigma-5, and unlike AP-1 and AP-2 it neither binds clathrin nor requires ARF1 for membrane recruitment. In cells it does not work alone: the tetramer is stably associated with the two large hereditary spastic paraplegia proteins spatacsin (SPG11) and spastizin (ZFYVE26/SPG15), and the six proteins co-purify at approximately equal stoichiometry from both cytosol and membranes. The assembly cycles between a cytosolic pool and the membrane of late endosomes and lysosomes, onto which it is recruited by coincidence detection that requires phosphatidylinositol 3-phosphate, read by the spastizin FYVE domain, together with Rag GTPases, and which is enhanced when cells are starved. Its cellular job is retrieval out of the late endocytic system: loss of AP-5 impairs the return of the cation-independent mannose 6-phosphate receptor and of the Golgi membrane proteins GOLIM4 and GOLM1 from endosomes to the Golgi, acting as a parallel route to retromer, and cells lacking the complex accumulate enlarged, multilamellar endolysosomes filled with storage material. Within the complex, sigma-5 is the partner of zeta-5, forming one of the two hemicomplexes from which AP-5 is assembled, and in the cryo-EM structure of the core it also contributes part of the surface that binds the spatacsin WD40-hairpin, an interface whose spatacsin side is required for the complex to assemble. It has no known catalytic activity, and no sorting signal has yet been identified for AP-5: the cargo-binding regions that are conserved in the other adaptor complexes are altered in AP-5, so whether and how it recognises cargo directly remains open. Mutations in the other AP-5 subunits and in SPG11 and ZFYVE26 cause hereditary spastic paraplegia and a lysosomal macular dystrophy; no disease-causing variant has yet been reported in AP5S1 itself.
Definition: A protein-containing complex consisting of the four subunits of the AP-5 adaptor complex (zeta-5, beta-5, mu-5 and sigma-5) stably associated with spatacsin and spastizin, found on the cytoplasmic face of late endosomal and lysosomal membranes, where it acts in retrieval of membrane proteins out of the late endocytic system.
Justification: AP-5 does not work as an isolated heterotetramer. The four subunits co-purify with spatacsin and spastizin at approximately 1:1:1:1:1:1 from both cytosol and detergent-extracted membranes, all six knockdowns give the same cargo phenotype, and the assembly has now been reconstituted and solved by cryo-EM, with the N-terminal region of spatacsin required for AP-5 complex interaction and assembly. ComplexPortal curates it as CPX-20045 (AP5Z1:AP5B1:ZFYVE26:SPG11:AP5M1:AP5S1) but that entry carries no GO cross-reference, whereas the tetramer CPX-5181 maps to GO:0044599; QuickGO text searches for 'spatacsin', 'spastizin' and 'AP-5' return no term for this six-protein assembly, the only AP-5 complex term being GO:0044599 for the tetramer. The absence is visible in this gene's own annotation set: the three rows ComplexPortal derives from the hexamer's structural paper reach AP5S1 as location and process statements because there is no complex for it to be part_of. A term would also give the SPG11 and ZFYVE26 gene products, which are not AP-5 subunits, a correct home.
Parent term: protein-containing complex
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000724 double-strand break repair via homologous recombination | IEA GO_REF:0000002 | MARK AS OVER ANNOTATED | Summary: InterPro2GO row. IPR029392 is the AP5S1-specific family entry (PANTHER PTHR16120 plus Pfam PF15001), and one of the three GO terms it maps to is homologous-recombination repair. The InterPro entry's own description states that the family 'is required for efficient homologous recombination DNA double-strand break repair' and cites PMID:20613862 for it, which is the same single HeLa esiRNA experiment that already supplies this gene's IMP row. Reason: The electronic row is not independent of the experimental one beside it: the mapping traces to one DR-GFP recombination readout, performed on this very protein, so it adds no evidence about AP5S1 while doubling the apparent support for the weakest process claim on the gene. It also carries that human cell-line phenotype to every member of the family - the InterPro entry matches 790 proteins across 2,294 taxa - for a process with no mechanistic account, no replication in the fifteen years since, and no coherence even inside the original experiment, where two of the other four core proteins of the same complex gave no effect. Marked as over-annotated rather than removed because the underlying human observation is real; the family-wide mapping is the part that should be revisited, and that is raised with InterPro in suggested_questions. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: SOURCE EVIDENCE WEAK Sources checked: InterPro:IPR029392 Β· InterPro IPR029392, AP-5 complex subunit sigma-1 (PANTHER PTHR16120 + Pfam PF15001) CIRCULAR OR REDUNDANT The signature is correctly assigned - AP5S1 is the only human gene in PTHR16120 and in of PF15001 - so the problem is not the match but the term attached to it. The InterPro description cites PMID:20613862 for the repair claim, i.e. the source of the family-level mapping is the target's own single RNAi experiment, which the target already carries as an IMP. Supporting Evidence: PMID:20613862 significant reduction of GFP positive cells were observed upon silencing of C20orf29 and SPG15 with two independent esiRNAs PMID:20613862 Knockdown of SPG11 and DKFZp761E198, however, did no have an effect on the percentage of GFP positive cells. |
| GO:0000724 double-strand break repair via homologous recombination | IMP PMID:20613862 A genome-scale DNA repair RNAi screen identifies SPG48 as a ... | KEEP AS NON CORE | Summary: The only direct experimental claim for a nuclear process. C20orf29 (AP5S1) was not a hit in the genome-scale screen itself; it was tested because it co-purified with KIAA0415/AP5Z1, and two independent esiRNAs then reduced homologous recombination in the DR-GFP reporter assay. Reason: The observation is on-target for this gene and is not second-guessed here: the paper states it for C20orf29 explicitly, with two independent silencing triggers, and the curator read the full text. It is nonetheless peripheral rather than core, on grounds internal to the same paper and to the subsequent literature. Within that experiment the effect does not behave like a property of the complex - knocking down SPG11 or beta5 changed nothing - so it cannot be read as 'AP-5 acts in recombination'. The group that went on to characterise AP-5 argued the repair phenotype is likely indirect and showed the complex acting on late endosomes with an endosomal knockdown phenotype. Fifteen years later no work has connected AP-5 to recombination, while five papers have built out its endolysosomal role. Retained as a real but non-core observation. Supporting Evidence: PMID:20613862 significant reduction of GFP positive cells were observed upon silencing of C20orf29 and SPG15 with two independent esiRNAs PMID:20613862 Knockdown of SPG11 and DKFZp761E198, however, did no have an effect on the percentage of GFP positive cells. PMID:22022230 these observations suggest that the site of action of AP-5 is the late endosome and/or lysosome, as defined by the presence of LAMP1 |
| GO:0005515 protein binding | IPI PMID:20613862 A genome-scale DNA repair RNAi screen identifies SPG48 as a ... | MARK AS OVER ANNOTATED | Summary: Partner resolved: O43299 is AP5Z1 (zeta-5/SPG48), the large adaptin of AP-5 and the bait in the experiment. This is the one pairing in the set that is a direct, structurally resolved contact - sigma-5 sits on the zeta trunk - but the row records it only as protein binding. Reason: Bare GO:0005515 tells us nothing about what sigma-5 does, and here it cannot be upgraded. The reference-projection test shows PMID:20613862 annotating exactly five distinct entities in QuickGO - AP5Z1, AP5B1, AP5S1, SPG11 and ZFYVE26 - so these IPI rows are the internal partner list of one complex rather than independent binding activities. The informative statement is complex membership, GO:0044599, which the gene already carries. The architectural molecular function sigma-5 does have is asserted separately, as a NEW GO:0005198 row grounded in the structure; it is not reached by modifying an IPI row, because an IPI records a pairwise partnership and the binding-specificity function that would fit such a row - recognition of a sorting signal - is not established for AP-5 at all. The one candidate binding term, GO:0140312 cargo adaptor activity, requires binding a clathrin- or COPII-like scaffold and bridging a cargo receptor, neither of which holds here. For this partner specifically the contact is real and direct: in the AP-5 cryo-EM structure sigma-5 buries ~2,620 square angstroms against the zeta trunk and touches no other AP-5 subunit. Supporting Evidence: PMID:20613862 Immunoprecipitation experiments followed by spectrometric identification of co-isolated proteins revealed interactions of KIAA0415-LAP with SPG11, SPG15, C20orf29, and DKFZp761E198 file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. |
| GO:0005515 protein binding | IPI PMID:20613862 A genome-scale DNA repair RNAi screen identifies SPG48 as a ... | MARK AS OVER ANNOTATED | Summary: Partner resolved: Q2VPB7 is AP5B1 (beta-5), the other large subunit of AP-5. The evidence is co-immunoprecipitation of the five-protein assembly, not a binary assay. Reason: Bare GO:0005515 tells us nothing about what sigma-5 does, and here it cannot be upgraded. The reference-projection test shows PMID:20613862 annotating exactly five distinct entities in QuickGO - AP5Z1, AP5B1, AP5S1, SPG11 and ZFYVE26 - so these IPI rows are the internal partner list of one complex rather than independent binding activities. The informative statement is complex membership, GO:0044599, which the gene already carries. The architectural molecular function sigma-5 does have is asserted separately, as a NEW GO:0005198 row grounded in the structure; it is not reached by modifying an IPI row, because an IPI records a pairwise partnership and the binding-specificity function that would fit such a row - recognition of a sorting signal - is not established for AP-5 at all. The one candidate binding term, GO:0140312 cargo adaptor activity, requires binding a clathrin- or COPII-like scaffold and bridging a cargo receptor, neither of which holds here. This pairing is in addition indirect at the structural level: in the AP-5 core sigma-5 makes no contact at all with beta5 (or with mu5), so the co-purification reflects shared membership of the heterotetramer, which GO:0044599 already states. Supporting Evidence: PMID:20613862 Immunoprecipitation experiments followed by spectrometric identification of co-isolated proteins revealed interactions of KIAA0415-LAP with SPG11, SPG15, C20orf29, and DKFZp761E198 file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. |
| GO:0005515 protein binding | IPI PMID:20613862 A genome-scale DNA repair RNAi screen identifies SPG48 as a ... | MARK AS OVER ANNOTATED | Summary: Partner resolved: Q68DK2 is ZFYVE26/spastizin (SPG15), an accessory protein that is stably associated with AP-5 rather than a subunit of it, and that supplies the PI3P-binding FYVE domain by which the assembly is recruited to endolysosomal membrane. Reason: Bare GO:0005515 tells us nothing about what sigma-5 does, and here it cannot be upgraded. The reference-projection test shows PMID:20613862 annotating exactly five distinct entities in QuickGO - AP5Z1, AP5B1, AP5S1, SPG11 and ZFYVE26 - so these IPI rows are the internal partner list of one complex rather than independent binding activities. The informative statement is complex membership, GO:0044599, which the gene already carries. The architectural molecular function sigma-5 does have is asserted separately, as a NEW GO:0005198 row grounded in the structure; it is not reached by modifying an IPI row, because an IPI records a pairwise partnership and the binding-specificity function that would fit such a row - recognition of a sorting signal - is not established for AP-5 at all. The one candidate binding term, GO:0140312 cargo adaptor activity, requires binding a clathrin- or COPII-like scaffold and bridging a cargo receptor, neither of which holds here. Spastizin is not present in the AP-5 cryo-EM structure, so there is no structural evidence of a direct sigma-5 contact; what is established is that the six proteins form a stable assembly at roughly 1:1:1:1:1:1 stoichiometry. That assembly has no GO cellular-component term, which is recorded as an ontology gap rather than papered over here. Supporting Evidence: PMID:20613862 Immunoprecipitation experiments followed by spectrometric identification of co-isolated proteins revealed interactions of KIAA0415-LAP with SPG11, SPG15, C20orf29, and DKFZp761E198 PMID:23825025 we show that the four AP-5 subunits can be coimmunoprecipitated with SPG11 and SPG15, both from cytosol and from detergent-extracted membranes, with a stoichiometry of βΌ1:1:1:1:1:1 |
| GO:0005515 protein binding | IPI PMID:20613862 A genome-scale DNA repair RNAi screen identifies SPG48 as a ... | MARK AS OVER ANNOTATED | Summary: Partner resolved: Q96JI7 is SPG11/spatacsin, the second accessory protein of the assembly. Of the four partners recorded from this paper, this is the one for which a direct sigma-5 contact is now visible: the SPG11 WD40-hairpin packs against sigma-5 in the AP-5 structure. Reason: Bare GO:0005515 tells us nothing about what sigma-5 does, and here it cannot be upgraded. The reference-projection test shows PMID:20613862 annotating exactly five distinct entities in QuickGO - AP5Z1, AP5B1, AP5S1, SPG11 and ZFYVE26 - so these IPI rows are the internal partner list of one complex rather than independent binding activities. The informative statement is complex membership, GO:0044599, which the gene already carries. The architectural molecular function sigma-5 does have is asserted separately, as a NEW GO:0005198 row grounded in the structure; it is not reached by modifying an IPI row, because an IPI records a pairwise partnership and the binding-specificity function that would fit such a row - recognition of a sorting signal - is not established for AP-5 at all. The one candidate binding term, GO:0140312 cargo adaptor activity, requires binding a clathrin- or COPII-like scaffold and bridging a cargo receptor, neither of which holds here. The contact itself is better evidenced than the term suggests - sigma-5 buries ~755 square angstroms against the SPG11 WD40-hairpin over 20 residues, and SPG11's N-terminal region is what allows the two AP-5 hemicomplexes to assemble into a tetramer - but 'protein binding' expresses none of that, and the assembly-licensing role it belongs to has no molecular-function term. Supporting Evidence: PMID:20613862 Immunoprecipitation experiments followed by spectrometric identification of co-isolated proteins revealed interactions of KIAA0415-LAP with SPG11, SPG15, C20orf29, and DKFZp761E198 file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is also part of the SPG11 binding surface, burying 755.4 A^2 against the SPG11 WD40-hairpin over 20 residues. PMID:40175557 the N-terminal region of SPG11 is required for AP5 complex interaction and assembly |
| GO:0005515 protein binding | IPI PMID:40175557 Structural basis for membrane remodeling by the AP5-SPG11-SP... | MARK AS OVER ANNOTATED | Summary: Partner resolved: Q3U829 is mouse Ap5z1. The pairing is cross-species because the reconstitution solved by cryo-EM is: in the deposited structure 8YAB, chain C is human sigma-5 (Q9NUS5) and chain A is mouse zeta (Q3U829), which was verified here by matching each chain's modelled sequence against its UniProt record. So this is not a stray heterologous hit but the zeta/sigma-5 heterodimer resolved at 3.26 A. Reason: This is the best-evidenced interaction on the gene and it is recorded with the least informative term available. It still cannot be upgraded: the molecular statement it supports is that sigma-5 is the small subunit of the zeta/sigma-5 hemicomplex, and GO has no molecular function for that role. GO:0140312 cargo adaptor activity would be the tempting target and is wrong on its own definition, which requires binding clathrin- or COPII-like scaffolding and bridging a cargo receptor: AP-5 binds no clathrin, and whether it recognises cargo at all is explicitly open. The complex membership statement that does apply is already on the gene as GO:0044599. Supporting Evidence: file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md sigma5 (AP5S1, human) -- the target file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md zeta5 (AP5Z1, mouse) PMID:40175557 We employed cryogenic-electron microscopy and in silico predictions to investigate the structural assemblies of the SPG11-SPG15 and AP5-SPG11-SPG15 complexes. PMID:33464297 Even highly conserved regions of the subunits, such as the domains that bind to sorting signals, are different in AP-5, indicating that cargo recognition (assuming it occurs) must be by a different molecular mechanism. |
| GO:0005764 lysosome | IBA GO_REF:0000033 | ACCEPT | Summary: PAINT inference of lysosomal localisation across the sigma-5 orthology group, from the family's single IBD node. It restates, as an ancestral property, what was measured directly on the human protein: GFP-tagged sigma-5 overlaps the late endosomal/lysosomal marker LAMP1. Reason: Correct and core. AP-5 was described on discovery as the only coat then known to associate with late endosomes, and every subsequent study places the AP-5/SPG11/SPG15 assembly on late endosomes and lysosomes. The node is seeded by a single donor, the target itself, which is the expected pattern for a family with one well-studied member and is not circular: that donor's evidence is an IDA, not another propagation. Human AP5S1 is inside the clade by construction, being the only human gene in PTHR16120, and there is no IRD or IKR anywhere in the slice. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000413531 Β· PTHR16120 AP-5 COMPLEX SUBUNIT SIGMA-1 IBD node SUPPORTS TRANSFER The family's only PTN node, and the local PAINT slice shows it carrying exactly five IBDs (GO:0005764, GO:0005770, GO:0005829, GO:0030119, GO:0016197), each with a single seed: UniProtKB:Q9NUS5, the target itself. PANTHER assigns AP5S1 to PTHR16120:SF0, so the target is inside the inheriting clade by construction; the node asserts that what was measured on the human protein is ancestral to the sigma-5 orthology group. No IRD or IKR appears anywhere in the slice. UniProtKB:Q9NUS5 Β· AP5S1 (Homo sapiens, the target) SUPPORTS TRANSFER The target's own accession appearing among the sources is expected, not circular: the target's own IDA for lysosome (PMID:22022230: GFP-tagged sigma-5 overlaps LAMP1, Pearson 0.789) is one of the descendant evidences the PAINT curator used to place the IBD. It marks that experimental grounding exists on the target, and the IBA adds the separate claim that the property is inherited rather than human-specific. Supporting Evidence: PMID:22022230 0.789Β±0.003 for C20orf29-GFP and LAMP1 PMID:22022230 these observations suggest that the site of action of AP-5 is the late endosome and/or lysosome, as defined by the presence of LAMP1 PMID:23825025 AP-5, SPG11, and SPG15 colocalize on a late endosomal/lysosomal compartment file:human/AP5S1/AP5S1-uniprot.txt DR PANTHER; PTHR16120; AP-5 COMPLEX SUBUNIT SIGMA-1; 1. |
| GO:0005764 lysosome | IDA PMID:22022230 The fifth adaptor protein complex. | ACCEPT | Summary: Direct imaging evidence on sigma-5 itself: GFP-tagged C20orf29 shows substantial overlap with LAMP1 (Pearson 0.789 +/- 0.003), against markedly less overlap with markers of other compartments. Reason: This is the primary observation the whole lysosomal assignment rests on, and it was made on this subunit rather than inherited from the complex. LAMP1 marks late endosomes and lysosomes together, which is why the curator made both rows; the authors' own reading is that AP-5 acts at the late endosome and/or lysosome. Core. Supporting Evidence: PMID:22022230 0.789Β±0.003 for C20orf29-GFP and LAMP1 PMID:22022230 In both cases, there was relatively little colocalisation with most markers, but substantial colocalisation with LAMP1 PMID:22022230 these observations suggest that the site of action of AP-5 is the late endosome and/or lysosome, as defined by the presence of LAMP1 |
| GO:0005764 lysosome | NAS PMID:40175557 Structural basis for membrane remodeling by the AP5-SPG11-SP... | ACCEPT | Summary: ComplexPortal statement carried from the AP5-SPG11-SPG15 cryo-EM paper, whose subject is the hexameric assembly that this subunit belongs to and which is described as acting in lysosome formation and autolysosome tubulation. Reason: Author-statement evidence that agrees with the direct imaging on this subunit and with every localisation study of the complex since 2011, so nothing rests on the NAS alone. Reference-projection check: QuickGO shows PMID:22022230 annotating 10 distinct entities (the four human AP-5 subunits, their four mouse orthologues and the two ComplexPortal AP-5 complexes) and PMID:40175557 annotating 10 (the AP-5 subunits plus SPG11, ZFYVE26 and the hexamer complex CPX-20045). These are complex-scoped papers, not catalogue depositions. Supporting Evidence: PMID:40175557 The human spastizin (spastic paraplegia 15, SPG15) and spatacsin (spastic paraplegia 11, SPG11) complex is involved in the formation of lysosomes PMID:40175557 SPG11-SPG15 can cooperate with the fifth adaptor protein complex (AP5) involved in membrane sorting of late endosomes. PMID:33464297 Adaptor protein complex 5 (AP-5) and its partners, SPG11 and SPG15, are recruited onto late endosomes and lysosomes. |
| GO:0005765 lysosomal membrane | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location mapping: the record's 'Lysosome membrane; Peripheral membrane protein; Cytoplasmic side' line maps to GO:0005765. The underlying UniProt statement is ECO:0000305 - curator inference from PMID:22022230 rather than a separate observation. Reason: The inference is sound and the mapping is exact. AP-5 is a peripheral coat that cycles on and off membrane - fractionation puts roughly half of the complex on membranes and half in cytosol - and the membrane it is recruited to is the LAMP1-positive one, by coincidence detection on PI3P and Rag GTPases. A cytoplasmic-face peripheral assignment is exactly right for this kind of protein. Recorded here rather than silently accepted: the ECO:0000305 status means this row is a reading of the same experiment as the lysosome IDA, not additional evidence. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0157 Β· UniProt subcellular location SL-0157, Lysosome membrane SUPPORTS TRANSFER Verified at the UniProt locations endpoint: SL-0157 is 'Lysosome membrane' and its declared GO mapping is GO:0005765 lysosomal membrane, so the vocabulary step is exact. The source statement is on AP5S1's own record and is flagged ECO:0000305 from PubMed:22022230. Supporting Evidence: PMID:22022230 Like clathrin and AP-1, C14orf108 partitions approximately equally between membranes and cytosol, indicating that it cycles on and off the membrane PMID:33464297 recruitment of AP-5/SPG11/SPG15 is enhanced in starved cells and occurs by coincidence detection, requiring both phosphatidylinositol 3-phosphate (PI3P) and Rag GTPases PMID:22022230 0.789Β±0.003 for C20orf29-GFP and LAMP1 |
| GO:0005770 late endosome | IBA GO_REF:0000033 | ACCEPT | Summary: The late-endosome half of the same PAINT node's compartment assignment, seeded by the target's own IDA. Reason: Correct and core. The late endosome is where AP-5's function has been located by every subsequent approach - colocalisation of the six-protein assembly, the retrieval defect on knockout, and recruitment by PI3P and Rag GTPases. Same node and same single seed as the lysosome row; the target is inside the clade and there is no evidence of lineage-specific loss. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000413531 Β· PTHR16120 AP-5 COMPLEX SUBUNIT SIGMA-1 IBD node SUPPORTS TRANSFER The family's only PTN node, and the local PAINT slice shows it carrying exactly five IBDs (GO:0005764, GO:0005770, GO:0005829, GO:0030119, GO:0016197), each with a single seed: UniProtKB:Q9NUS5, the target itself. PANTHER assigns AP5S1 to PTHR16120:SF0, so the target is inside the inheriting clade by construction; the node asserts that what was measured on the human protein is ancestral to the sigma-5 orthology group. No IRD or IKR appears anywhere in the slice. UniProtKB:Q9NUS5 Β· AP5S1 (Homo sapiens, the target) SUPPORTS TRANSFER The target's own accession appearing among the sources is expected, not circular: the target's own IDA for late endosome (PMID:22022230, the same LAMP1-positive compartment) is one of the descendant evidences the PAINT curator used to place the IBD. It marks that experimental grounding exists on the target, and the IBA adds the separate claim that the property is inherited rather than human-specific. Supporting Evidence: PMID:22022230 Tagged AP-5 localises to a late endosomal compartment in HeLa cells. PMID:23825025 AP-5, SPG11, and SPG15 colocalize on a late endosomal/lysosomal compartment PMID:33464297 Adaptor protein complex 5 (AP-5) and its partners, SPG11 and SPG15, are recruited onto late endosomes and lysosomes. |
| GO:0005770 late endosome | IDA PMID:22022230 The fifth adaptor protein complex. | ACCEPT | Summary: Direct imaging on sigma-5: the GFP-tagged subunit concentrates on a perinuclear, LAMP1-positive compartment, which the paper identifies as late endosomal. Reason: Primary evidence on this subunit for the compartment where AP-5 acts. Core. Supporting Evidence: PMID:22022230 0.789Β±0.003 for C20orf29-GFP and LAMP1 PMID:22022230 Tagged AP-5 localises to a late endosomal compartment in HeLa cells. PMID:22022230 these observations suggest that the site of action of AP-5 is the late endosome and/or lysosome, as defined by the presence of LAMP1 |
| GO:0005770 late endosome | NAS PMID:22022230 The fifth adaptor protein complex. | ACCEPT | Summary: ComplexPortal statement for the AP-5 tetramer (CPX-5181), whose curated location is the late endosome, carried down to each subunit. Reason: Agrees with the subunit's own IDA from the same paper, so the complex-level statement adds redundancy rather than risk. Reference-projection check: QuickGO shows PMID:22022230 annotating 10 distinct entities (the four human AP-5 subunits, their four mouse orthologues and the two ComplexPortal AP-5 complexes) and PMID:40175557 annotating 10 (the AP-5 subunits plus SPG11, ZFYVE26 and the hexamer complex CPX-20045). These are complex-scoped papers, not catalogue depositions. Supporting Evidence: PMID:22022230 Tagged AP-5 localises to a late endosomal compartment in HeLa cells. PMID:22022230 Thus, we propose that like the other AP complexes, AP-5 is a heterotetramer, consisting of two large subunits, a medium subunit, and a small subunit |
| GO:0005770 late endosome | NAS PMID:40175557 Structural basis for membrane remodeling by the AP5-SPG11-SP... | ACCEPT | Summary: The same location statement made for the hexameric AP5-SPG11-SPG15 complex (CPX-20045) from the structural paper. Reason: Consistent with the direct evidence on this subunit and with the recruitment mechanism. Worth noting for curation: because GO has no term for the AP5-SPG11-SPG15 assembly, ComplexPortal's curation of that complex reaches AP5S1 only as location and process rows, never as a part_of. That gap is recorded in knowledge_gaps and proposed_new_terms. Supporting Evidence: PMID:40175557 SPG11-SPG15 can cooperate with the fifth adaptor protein complex (AP5) involved in membrane sorting of late endosomes. PMID:33464297 Adaptor protein complex 5 (AP-5) and its partners, SPG11 and SPG15, are recruited onto late endosomes and lysosomes. |
| GO:0005829 cytosol | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: PAINT inference that the sigma-5 group is cytosolic, from the same single-seed node. For a coat subunit this describes the unrecruited pool rather than a site of action. Reason: True but peripheral. AP-5 partitions roughly equally between membranes and cytosol and cycles on and off; the cytosolic fraction is the reservoir, while the function happens once the complex is recruited onto late endosomal and lysosomal membrane. The inference itself is sound - the node is correctly placed, the target is inside the clade, and the seed is the target's own direct observation - so this is a scope judgement, not a propagation failure; hence NO_FAILURE_NON_CORE rather than a failure mode. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: PANTHER:PTN000413531 Β· PTHR16120 AP-5 COMPLEX SUBUNIT SIGMA-1 IBD node SUPPORTS TRANSFER The family's only PTN node, and the local PAINT slice shows it carrying exactly five IBDs (GO:0005764, GO:0005770, GO:0005829, GO:0030119, GO:0016197), each with a single seed: UniProtKB:Q9NUS5, the target itself. PANTHER assigns AP5S1 to PTHR16120:SF0, so the target is inside the inheriting clade by construction; the node asserts that what was measured on the human protein is ancestral to the sigma-5 orthology group. No IRD or IKR appears anywhere in the slice. UniProtKB:Q9NUS5 Β· AP5S1 (Homo sapiens, the target) SUPPORTS TRANSFER The target's own accession appearing among the sources is expected, not circular: the target's own cytosolic pool, seen both by fractionation of AP-5 and by HPA immunofluorescence is one of the descendant evidences the PAINT curator used to place the IBD. It marks that experimental grounding exists on the target, and the IBA adds the separate claim that the property is inherited rather than human-specific. Supporting Evidence: PMID:22022230 Like clathrin and AP-1, C14orf108 partitions approximately equally between membranes and cytosol, indicating that it cycles on and off the membrane PMID:33464297 recruitment of AP-5/SPG11/SPG15 is enhanced in starved cells and occurs by coincidence detection, requiring both phosphatidylinositol 3-phosphate (PI3P) and Rag GTPases |
| GO:0005829 cytosol | IDA GO_REF:0000052 | KEEP AS NON CORE | Summary: Human Protein Atlas immunofluorescence assignment of cytosolic staining. GO_REF:0000052 is the large-scale HPA curation reference; a projection count in QuickGO gives 3,640 distinct entities for it, so this is a systematic localisation call rather than a statement about AP5S1 in particular. Reason: Consistent with the biochemistry - about half of AP-5 is cytosolic at steady state - and therefore worth keeping, but it is the unrecruited pool and not where the complex works. The low abundance of AP-5 and the fact that tagged constructs appear mostly cytosolic unless endogenous protein is depleted also make a diffuse cytosolic call the expected outcome of this kind of screen. Supporting Evidence: PMID:22022230 Like clathrin and AP-1, C14orf108 partitions approximately equally between membranes and cytosol, indicating that it cycles on and off the membrane |
| GO:0005829 cytosol | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: UniProt subcellular-location mapping of the record's 'Cytoplasm, cytosol' line (ECO:0000269 from PMID:22022230) to GO:0005829. Reason: The vocabulary step is exact and the underlying statement is experimentally grounded, but as with the other two cytosol rows this describes the unrecruited pool of a cycling coat rather than the compartment where sigma-5 acts. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: UniProtKB-SubCell:SL-0091 Β· UniProt subcellular location SL-0091, Cytosol SUPPORTS TRANSFER Verified at the UniProt locations endpoint: SL-0091 is 'Cytosol' and maps to GO:0005829. The source line on AP5S1's record is ECO:0000269|PubMed:22022230, i.e. experimentally backed, so the transfer is faithful; the demotion is about biological scope, not about the mapping. Supporting Evidence: PMID:22022230 Like clathrin and AP-1, C14orf108 partitions approximately equally between membranes and cytosol, indicating that it cycles on and off the membrane |
| GO:0007040 lysosome organization | NAS PMID:40175557 Structural basis for membrane remodeling by the AP5-SPG11-SP... | KEEP AS NON CORE | Summary: ComplexPortal process statement from the structural paper, which frames the SPG11-SPG15 and AP5-SPG11-SPG15 complexes as acting in lysosome formation and as essential for initiating autolysosome tubulation. Reason: Defensible but secondary for this subunit. The supporting work is complex-level and largely in vitro or on other subunits: reconstituted AP5-SPG11-SPG15 binds PI3P, senses curvature and remodels membrane, and loss of AP-5 leaves cells with enlarged, storage-filled endolysosomes. Sigma-5 has never been perturbed in an autophagic lysosome reformation assay. Lysosome organization is also better read as a consequence of the sorting role than as a separate activity, so it is kept as a real but non-core process. Supporting Evidence: PMID:40175557 Our findings reveal that the AP5-SPG11-SPG15 complex can bind PI3P molecules, sense membrane curvature and drive membrane remodeling in vitro. PMID:40175557 which is essential for the initiation of autolysosome tubulation PMID:26085577 the loss of AP-5 results in an accumulation of enlarged multilamellar structures that are filled with aberrant storage material. |
| GO:0016192 vesicle-mediated transport | NAS PMID:22022230 The fifth adaptor protein complex. | KEEP AS NON CORE | Summary: ComplexPortal process statement for the AP-5 tetramer: the parent vesicle-transport term, assigned from the paper that established AP-5 as an adaptor complex. Reason: True, and the right level for a complex-scoped statement made in 2011, when AP-5's specific step was not yet defined. It is kept rather than removed, but not as core: the gene separately carries the more specific child GO:0016197 endosomal transport by IBA, IEA and IMP, and a late-endosome-to-Golgi term is proposed as a NEW row below, so this grandparent is true but superseded by annotations already on the gene. ACCEPT would assert it as representing the core function, which the child does better. Supporting Evidence: PMID:22022230 Knocking down AP-5 subunits interferes with the trafficking of the cation-independent mannose 6-phosphate receptor and causes the cell to form swollen endosomal structures with emanating tubules. PMID:22022230 Thus, we propose that like the other AP complexes, AP-5 is a heterotetramer, consisting of two large subunits, a medium subunit, and a small subunit PMID:22022230 Together with the C14orf108 localisation data, these observations point to a role for the C14orf108-DKFZp761E198 complex in endosomal trafficking. |
| GO:0016197 endosomal transport | IBA GO_REF:0000033 | ACCEPT | Summary: The node's single process inference: the sigma-5 orthology group acts in endosomal transport, seeded by the target's own knockdown phenotype. Reason: Correct and core, and at a defensible level of specificity. Sigma-5 knockdown reproduces the phenotype of knocking down any other AP-5 subunit - the CIMPR and the retromer subunit Vps26 collapse into enlarged perinuclear puncta - and knockdown of all six members of the assembly traps the CIMPR in early endosome clusters. GRANULARITY_MISMATCH does not apply: the node's more specific destination, late endosome to Golgi transport, was only established seven years after this IBD was placed and by a knockout of a different subunit, so the parent term was the right call on the evidence available to the curator. The finer term is added separately as a NEW row rather than being demanded of the IBA. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000413531 Β· PTHR16120 AP-5 COMPLEX SUBUNIT SIGMA-1 IBD node SUPPORTS TRANSFER The family's only PTN node, and the local PAINT slice shows it carrying exactly five IBDs (GO:0005764, GO:0005770, GO:0005829, GO:0030119, GO:0016197), each with a single seed: UniProtKB:Q9NUS5, the target itself. PANTHER assigns AP5S1 to PTHR16120:SF0, so the target is inside the inheriting clade by construction; the node asserts that what was measured on the human protein is ancestral to the sigma-5 orthology group. No IRD or IKR appears anywhere in the slice. UniProtKB:Q9NUS5 Β· AP5S1 (Homo sapiens, the target) SUPPORTS TRANSFER The target's own accession appearing among the sources is expected, not circular: the target's own IMP for endosomal transport (PMID:22022230: sigma-5 knockdown redistributes the CIMPR and Vps26) is one of the descendant evidences the PAINT curator used to place the IBD. It marks that experimental grounding exists on the target, and the IBA adds the separate claim that the property is inherited rather than human-specific. Supporting Evidence: PMID:22022230 knocking down either KIAA0415/SPG48 or C20orf29 produces the same phenotype as knocking down ΞΌ5 or Ξ²5 PMID:23825025 all six knockdowns cause the cation-independent mannose 6-phosphate receptor to become trapped in clusters of early endosomes PMID:22022230 Together with the C14orf108 localisation data, these observations point to a role for the C14orf108-DKFZp761E198 complex in endosomal trafficking. |
| GO:0016197 endosomal transport | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro2GO row from the same family signature that supplies the repair mapping. This one is well founded: IPR029392's description cites PMID:22022230 for the endosomal-sorting role of the family. Reason: The signature is AP5S1-specific rather than a generic adaptor domain, the term is the same one the target carries by IMP and IBA, and the family really is defined by membership of a complex that acts in endosomal sorting. An electronic route to a conclusion that experimental and phylogenetic evidence reach independently is worth keeping. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR029392 Β· InterPro IPR029392, AP-5 complex subunit sigma-1 (PANTHER PTHR16120 + Pfam PF15001) SUPPORTS TRANSFER Fetched from the InterPro API: IPR029392 is the AP-5 sigma-1 family, built from PANTHER PTHR16120 and Pfam PF15001, and its GO mapping to GO:0016197 is cited to PMID:22022230. AP5S1 is the only human gene in it, so the signature match is exact and the mapped term is the one its own experimental evidence supports. Supporting Evidence: PMID:22022230 knocking down either KIAA0415/SPG48 or C20orf29 produces the same phenotype as knocking down ΞΌ5 or Ξ²5 PMID:22022230 Together with the C14orf108 localisation data, these observations point to a role for the C14orf108-DKFZp761E198 complex in endosomal trafficking. |
| GO:0016197 endosomal transport | IMP PMID:22022230 The fifth adaptor protein complex. | ACCEPT | Summary: The subunit-level perturbation evidence. siRNA depletion of C20orf29 reproduces the trafficking phenotype seen for mu5 and beta5: the cation-independent mannose 6-phosphate receptor and Vps26 redistribute into larger, brighter, fewer perinuclear puncta, and cells accumulate swollen multivesicular bodies with emanating tubules. Reason: This is the strongest process evidence that exists for sigma-5 specifically rather than for the complex, and it is what makes endosomal transport a core annotation for this gene rather than one inherited by complex membership. Corroborated by the 2013 study, where all six knockdowns - including sigma-5 - trap the CIMPR in early endosome clusters. Supporting Evidence: PMID:22022230 knocking down either KIAA0415/SPG48 or C20orf29 produces the same phenotype as knocking down ΞΌ5 or Ξ²5 PMID:22022230 Knocking down AP-5 subunits interferes with the trafficking of the cation-independent mannose 6-phosphate receptor and causes the cell to form swollen endosomal structures with emanating tubules. PMID:23825025 all six knockdowns cause the cation-independent mannose 6-phosphate receptor to become trapped in clusters of early endosomes |
| GO:0030119 AP-type membrane coat adaptor complex | IBA GO_REF:0000033 | ACCEPT | Summary: PAINT inference that the sigma-5 group are subunits of an AP-type membrane coat adaptor complex, seeded by the target's own co-immunoprecipitation evidence. The interesting question for this row is whether sigma-5 is a real sigma-adaptin, and the structure says it is. Reason: Correct at family level. Sigma-5 is a genuine AP small subunit: profile searches recover sigma2, sigma1 and zeta-COP as its closest structural homologues, and in the AP-5 cryo-EM structure it adopts the sigma fold and occupies the canonical position on the large adaptin's trunk, burying ~2,620 square angstroms against zeta and contacting neither beta5 nor mu5. Two things that could have made this transfer wrong do not: no clathrin-related term leaks in from the AP-1/AP-2 sigmas, because sigma-5 sits in its own PANTHER family (PTHR16120) rather than in PTHR11753, and the clathrin-linking clause of this term's definition is permissive ('or another coat-forming molecule') with AP-5 explicitly represented by the child term GO:0044599, which QuickGO confirms is in this term's descendant closure. What does not transfer with the fold is cargo recognition: the residues that build sigma2's acidic dileucine pockets are not conserved in sigma-5 and the equivalent surface is partly covered by the zeta N-terminus, recorded as residue_claims below. Accepted as the correct general parent; the direct-evidence row for the same term is modified to the AP-5-specific child instead. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN000413531 Β· PTHR16120 AP-5 COMPLEX SUBUNIT SIGMA-1 IBD node SUPPORTS TRANSFER The family's only PTN node, and the local PAINT slice shows it carrying exactly five IBDs (GO:0005764, GO:0005770, GO:0005829, GO:0030119, GO:0016197), each with a single seed: UniProtKB:Q9NUS5, the target itself. PANTHER assigns AP5S1 to PTHR16120:SF0, so the target is inside the inheriting clade by construction; the node asserts that what was measured on the human protein is ancestral to the sigma-5 orthology group. No IRD or IKR appears anywhere in the slice. UniProtKB:Q9NUS5 Β· AP5S1 (Homo sapiens, the target) SUPPORTS TRANSFER The target's own accession appearing among the sources is expected, not circular: the target's own IDA for the AP-type complex (PMID:22022230: GFP-sigma-5 co-immunoprecipitates mu5 and beta5) is one of the descendant evidences the PAINT curator used to place the IBD. It marks that experimental grounding exists on the target, and the IBA adds the separate claim that the property is inherited rather than human-specific. Supporting Evidence: PMID:22022230 C20orf29 has a similar predicted secondary structure to the Ο subunits, but with longer loops connecting the folded domains PMID:22022230 iterative PSI-BLAST and HHpred searches using KIAA0415/SPG48 and C20orf29 identify AP large subunits and small subunits, respectively, as principal homologues PMID:22022230 the anti-GFP antibody not only brings down the construct itself, but also C14orf108 and DKFZp761E198 file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. file:human/AP5S1/AP5S1-uniprot.txt DR PANTHER; PTHR16120; AP-5 COMPLEX SUBUNIT SIGMA-1; 1. PMID:19140243 The major recognition events are the two leucine residues binding in hydrophobic pockets on Ο2. PMID:19140243 Mutation of a number of these to hydrophilic residues (Ο2L65S, Ο2V88D, Ο2V98S or Ο2L103S), or filling in the pocket by replacement of Ο2A63 or Ο2N92 with tryptophan, strongly inhibited binding of recombinant AP2 core complexes to different dileucine motifs PMID:23222959 missense mutations of AP2 Ο subunit (AP2S1) affecting Arg15, which forms key contacts with dileucine-based motifs of CCV cargo proteins, result in familial hypocalciuric hypercalcemia type 3 (FHH3) |
| GO:0030119 AP-type membrane coat adaptor complex | IDA PMID:22022230 The fifth adaptor protein complex. | MODIFY | Summary: Direct evidence of complex membership: immunoprecipitating GFP-tagged C20orf29 brings down mu5 (C14orf108) and beta5 (DKFZp761E198), which together with the shared knockdown phenotype is what led the authors to name this protein sigma-5. Reason: The evidence identifies a specific complex, not the general AP-type class: the proteins that co-precipitate are the other AP-5 subunits, and GO has a term for exactly that complex. GO:0044599 'AP-5 adaptor complex' is defined as 'An AP-type membrane coat adaptor complex that in humans consists of beta5, zeta, mu5 and sigma5 subunits', it is a descendant of GO:0030119 (checked in QuickGO), and it is the term the direct evidence supports. Modifying loses nothing - the parent remains on the gene through the IBA and IEA rows - and gains the AP-5-specific statement, which matters here because the parent's definition is written around linking clathrin to a membrane and AP-5 does not associate with clathrin at all. Proposed replacements: AP-5 adaptor complex Supporting Evidence: PMID:22022230 the anti-GFP antibody not only brings down the construct itself, but also C14orf108 and DKFZp761E198 PMID:22022230 we are calling the small subunit (C20orf29) Ο5 PMID:22022230 Thus, we propose that like the other AP complexes, AP-5 is a heterotetramer, consisting of two large subunits, a medium subunit, and a small subunit PMID:22022230 AP-5 does not associate with clathrin and is insensitive to brefeldin A. |
| GO:0030119 AP-type membrane coat adaptor complex | IEA GO_REF:0000002 | ACCEPT | Summary: InterPro2GO row mapping the AP-5 sigma family signature to the AP-type adaptor complex parent. Reason: True of every member of this family by definition, and the general parent is an appropriate level for an electronic family mapping even though the specific child exists. The IEA is not the place to demand AP-5-level specificity; the direct-evidence row carries that. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR029392 Β· InterPro IPR029392, AP-5 complex subunit sigma-1 (PANTHER PTHR16120 + Pfam PF15001) SUPPORTS TRANSFER IPR029392's GO mapping includes GO:0030119, cited in the InterPro entry to PMID:22022230, the paper that identified the family as the small subunit of a fifth AP complex. AP5S1 is the family's only human gene, so there is no risk of the mapping reaching a non-subunit. Supporting Evidence: PMID:22022230 C20orf29 has a similar predicted secondary structure to the Ο subunits, but with longer loops connecting the folded domains PMID:22022230 the anti-GFP antibody not only brings down the construct itself, but also C14orf108 and DKFZp761E198 |
| GO:0031902 late endosome membrane | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location mapping of the record's 'Late endosome membrane; Peripheral membrane protein; Cytoplasmic side' line (ECO:0000305 from PMID:22022230) to GO:0031902. Reason: The membrane-side refinement of the compartment where AP-5 works, and it is the correct reading of the underlying data: the complex partitions between membrane and cytosol, is recruited to PI3P-containing endolysosomal membrane through the SPG15 FYVE domain, and faces the cytosol. As with the lysosomal-membrane row this is a curator inference from the same imaging rather than an independent observation, which is worth recording but is not a reason to reject it. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0151 Β· UniProt subcellular location SL-0151, Late endosome membrane SUPPORTS TRANSFER Verified at the UniProt locations endpoint: SL-0151 is 'Late endosome membrane', mapping to GO:0031902. The source statement sits on AP5S1's own record with ECO:0000305|PubMed:22022230, so it is a curator inference from the localisation study rather than a second experiment. Supporting Evidence: PMID:22022230 Like clathrin and AP-1, C14orf108 partitions approximately equally between membranes and cytosol, indicating that it cycles on and off the membrane PMID:33464297 recruitment of AP-5/SPG11/SPG15 is enhanced in starved cells and occurs by coincidence detection, requiring both phosphatidylinositol 3-phosphate (PI3P) and Rag GTPases PMID:22022230 Tagged AP-5 localises to a late endosomal compartment in HeLa cells. |
| GO:0044599 AP-5 adaptor complex | NAS PMID:22022230 The fifth adaptor protein complex. | ACCEPT | Summary: ComplexPortal statement of membership in the AP-5 adaptor complex (CPX-5181: AP5B1:AP5M1:AP5S1:AP5Z1), from the paper that defined the complex. Reason: The single most informative annotation on the gene, and core. Although recorded as an author statement, it is backed by direct evidence on this subunit: GFP-tagged sigma-5 co-precipitates mu5 and beta5, the four subunits plus SPG11 and SPG15 co-purify at roughly 1:1:1:1:1:1 from both cytosol and membranes, and the tetramer has since been resolved by cryo-EM with sigma-5 as one of its four chains. Supporting Evidence: PMID:22022230 the anti-GFP antibody not only brings down the construct itself, but also C14orf108 and DKFZp761E198 PMID:23825025 we show that the four AP-5 subunits can be coimmunoprecipitated with SPG11 and SPG15, both from cytosol and from detergent-extracted membranes, with a stoichiometry of βΌ1:1:1:1:1:1 PMID:33464297 The SPG15-GFP construct assembles into a complex containing SPG11 and the four subunits of AP-5 (AP5Z1, AP5B1, AP5M1, and AP5S1 file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. |
| GO:0005198 structural molecule activity | IDA PMID:40175557 Structural basis for membrane remodeling by the AP5-SPG11-SP... | NEW | Summary: GOA gives AP5S1 no molecular function at all beyond bare protein binding, and the same is true of every other AP-5 subunit (checked in QuickGO: AP5Z1, AP5B1 and AP5S1 have only GO:0005515, AP5M1 has no molecular-function annotation). The one molecular role that is established for sigma-5 is architectural, and it is established structurally: in the cryo-EM structure of the AP-5 core bound to the spatacsin WD40-hairpin, sigma-5 is one of the four chains of the complex and packs against the zeta trunk over 60 residues, burying about 2,620 square angstroms. Reason: GO:0005198 is defined as 'the action of a molecule that contributes to the structural integrity of a complex', which is what a large, fully buried subunit interface demonstrates. Three lines converge: the four subunits plus spatacsin and spastizin co-purify at roughly 1:1:1:1:1:1, so sigma-5 is a stoichiometric member rather than a substoichiometric associate; immunoprecipitating tagged sigma-5 brings down mu5 and beta5; and the structure resolves it in the canonical small-subunit position on the large adaptin, the same position sigma1, sigma2 and sigma3 occupy. The claim is deliberately bounded and no stronger term is used: unlike sigma4, sigma-5 has never been removed from cells, so whether AP-5 can assemble without it is unknown, and that limit is recorded in knowledge_gaps. No cargo-recognition or adaptor activity is asserted. Supporting Evidence: file:human/AP5S1/AP5S1-bioinformatics/RESULTS.md Sigma-5 is the zeta-adaptin partner: it buries 2621.8 A^2 against the zeta trunk over 60 residues, and touches neither beta5 nor mu5. PMID:40175557 We employed cryogenic-electron microscopy and in silico predictions to investigate the structural assemblies of the SPG11-SPG15 and AP5-SPG11-SPG15 complexes. PMID:22022230 Thus, we propose that like the other AP complexes, AP-5 is a heterotetramer, consisting of two large subunits, a medium subunit, and a small subunit PMID:23825025 we show that the four AP-5 subunits can be coimmunoprecipitated with SPG11 and SPG15, both from cytosol and from detergent-extracted membranes, with a stoichiometry of βΌ1:1:1:1:1:1 PMID:22022230 the anti-GFP antibody not only brings down the construct itself, but also C14orf108 and DKFZp761E198 |
| GO:0034499 late endosome to Golgi transport | IC PMID:29381698 Role of the AP-5 adaptor protein complex in late endosome-to... | NEW | Summary: The pathway AP-5 is now assigned to, which no AP-5 subunit currently carries in GOA (checked in QuickGO for AP5Z1, AP5B1, AP5M1, AP5S1, SPG11 and ZFYVE26: zero annotations to GO:0034499 or its parent GO:0042147). Inferred by curator from two annotations the gene already has - GO:0016197 endosomal transport, which it carries by its own IMP, and GO:0044599 AP-5 adaptor complex - combined with the complex-level demonstration that losing AP-5 impairs retrieval of the CIMPR, GOLIM4 and GOLM1 from endosomes back to the Golgi. Reason: GO:0034499 'late endosome to Golgi transport' is a descendant of GO:0016197, the term the gene already carries, so this refines an existing annotation rather than making a new kind of claim. The evidence code is deliberately IC rather than IMP. Sigma-5's own perturbation (PMID:22022230) shows the CIMPR and Vps26 redistributing, which the authors read as 'a role ... in endosomal trafficking'; it does not establish the direction, and that paper's own introduction notes that the direction of transport is uncertain even for the far better-studied AP-1. The late-endosome-to-Golgi direction comes from the 2018 knockout study, which perturbed AP5Z1, not AP5S1. Coding that as IMP on this gene would assert that this gene was the one perturbed, so the honest encoding is IC with the two annotations the inference rests on named in supporting_entities. What licenses carrying the complex-level result to this subunit is that all six members of the assembly give the same CIMPR phenotype when depleted. Left at GO:0034499 rather than the broader GO:0042147 because the sorting step was localised specifically to late endosomes. Propagation Review Root cause: NO FAILURE CORE Sources checked: GO:0016197 Β· endosomal transport (this gene's own IMP, PMID:22022230) SUPPORTS TRANSFER The parent term, held by AP5S1 on subunit-level experimental evidence: siRNA depletion of sigma-5 redistributes the CIMPR and Vps26. It establishes that this gene acts in endosomal transport; the child term adds only the direction, which comes from the second source. GO:0044599 Β· AP-5 adaptor complex (this gene's own complex membership) SUPPORTS TRANSFER Membership is what carries the complex-level result to this subunit, and it is not a bare assumption of equivalence: knocking down any of the six members of the AP-5/SPG11/SPG15 assembly produces the same CIMPR phenotype, so the subunits are not separable by this readout. Supporting Evidence: PMID:22022230 knocking down either KIAA0415/SPG48 or C20orf29 produces the same phenotype as knocking down ΞΌ5 or Ξ²5 PMID:23825025 all six knockdowns cause the cation-independent mannose 6-phosphate receptor to become trapped in clusters of early endosomes PMID:29381698 Immunolocalisation showed that loss of AP-5 led to impaired retrieval of the cation-independent mannose 6-phosphate receptor (CIMPR), GOLIM4, and GOLM1 from endosomes back to the Golgi region. PMID:29381698 our findings suggest that AP-5 functions in a novel sorting step out of late endosomes, acting as a backup pathway for retromer PMID:29381698 We have used CRISPR-Cas9 to knock out the AP-5 ΞΆ subunit gene, AP5Z1, in HeLa cells, and then analysed the phenotype by subcellular fractionation profiling and quantitative mass spectrometry. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Should InterPro's mapping of IPR029392 to GO:0000724 be retained? The family entry matches 790 proteins across 2,294 taxa and the mapping is cited in the entry itself to PMID:20613862, a single DR-GFP experiment in HeLa cells in which two other members of the same complex showed no effect. The result is that every AP-5 sigma subunit in every eukaryote is electronically annotated to homologous recombination repair.
Suggested experts: InterPro curators, GOA
Q: Should GO create a cellular-component term for the AP-5/spatacsin/spastizin complex? ComplexPortal curates it as CPX-20045 but the entry has no GO cross-reference, so the hexamer's curation reaches the subunits only as location and process statements, and SPG11 and ZFYVE26 have no complex term at all.
Suggested experts: GO Ontology editors, ComplexPortal curators
Q: How should the molecular activity of a non-clathrin coat adaptor be expressed? GO:0140312 cargo adaptor activity is defined as binding the structural scaffolding of a vesicle coat 'such as clathrin or COPII' and bridging the cargo receptor. ComplexPortal asserts it for the AP-5 tetramer, but AP-5 binds no clathrin, its scaffold is thought to be spatacsin, and whether it engages cargo at all is unresolved - so no AP-5 subunit carries any molecular function term beyond bare GO:0005515 protein binding.
Suggested experts: GO Ontology editors, Gene Ontology membrane-trafficking curators
Q: Is the zeta N-terminus a regulatory element over a latent sigma-5 site? In closed AP-2 the sigma2 dileucine pocket is occupied by the beta2 N-terminus and released on activation; beta5 lacks that segment, and in the AP-5 structure the equivalent sigma-5 surface is instead in contact with the zeta N-terminus, in what the authors describe as the active, super-open conformation. Does deleting it expose a binding site?
Suggested experts: AP complex structural biologists, Membrane traffic biochemists
Experiment: Use the reconstituted AP-5 core (zeta/sigma-5 plus beta5/mu5, as assembled for cryo-EM with the spatacsin WD40-hairpin) in liposome-binding or ITC assays against a panel of sorting motifs - CD4 and LIMP-II dileucine peptides, a TGN38 YxxPhi peptide - with the AP-2 core as the positive control and PI3P-containing liposomes to provide the membrane context. Repeat with a zeta N-terminal deletion (residues 1-10) and with sigma-5 T143V/R151V substitutions that restore the sigma2 pocket chemistry. A negative result with wild type and a positive one with the restored pocket would show the site is vestigial rather than absent.
Hypothesis: AP-5 does not read acidic dileucine or YxxPhi sorting signals, and the surface that would read them in sigma-5 is both chemically altered and occluded by the zeta N-terminus.
Type: biophysical binding assay
Experiment: Make CRISPR AP5S1 knockouts in HeLa and run the same two readouts used for the AP5Z1 knockout: dynamic organellar mapping and vesicle-fraction proteomics, plus monensin-washout imaging of CIMPR, GOLIM4 and GOLM1 retrieval, with and without retromer knockdown. Blot for the other three subunits to establish whether the tetramer survives loss of sigma-5, which is the question the siRNA experiments cannot answer.
Hypothesis: Sigma-5 is required for AP-5 function, and an AP5S1 null phenocopies the AP5Z1 null.
Type: genetic knockout with quantitative proteomics
Experiment: Mutate the 20 sigma-5 residues identified at the spatacsin interface in the structure, in blocks of three to four, and test each variant for (i) co-expression assembly of the heterotetramer in the reconstitution system, (ii) co-immunoprecipitation with spatacsin from cells, and (iii) rescue of CIMPR retrieval in an AP5S1 knockout. Separating an assembly-defective allele from a spatacsin-binding-defective one would say whether sigma-5's contribution to the SPG11 contact is structural or regulatory.
Hypothesis: The sigma-5 surface that contacts the spatacsin WD40-hairpin is required for assembly of the AP-5 tetramer, not merely adjacent to it.
Type: structure-guided mutagenesis
Experiment: Repeat the DR-GFP assay in AP5S1-null cells alongside AP5Z1-null cells, with cell-cycle profiling and proliferation controls, and score gamma-H2AX resolution after irradiation. Then ask whether a membrane-recruitment-defective but assembly-competent condition - for example wortmannin treatment or loss of the spastizin FYVE domain - reproduces the repair phenotype. If repair tracks membrane recruitment, the phenotype is downstream of traffic.
Hypothesis: The homologous-recombination phenotype of sigma-5 knockdown is an indirect consequence of endolysosomal dysfunction.
Type: genetic separation of function
What is not known β curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: No sorting signal, and no direct cargo, has ever been identified for AP-5, so sigma-5 has no molecular function that can be stated - not merely none that GO can express.
OPEN BIOLOGY MF_DARK
What is known: Firmly established on the negative side: mu-5 lacks the residues that read YxxPhi signals; the sorting-signal-binding regions are different throughout AP-5; the structurally equivalent positions of the sigma2 dileucine site in sigma-5 mostly carry different residues (3 of 19 identical) and three of them are covered by the zeta N-terminus in the assembled core; and cargo binding in this system has so far been attributed to the accessory protein spastizin rather than to the heterotetramer. Firmly established on the positive side: the complex is needed for retrieval of the CIMPR, GOLIM4 and GOLM1 from endosomes to the Golgi, and sortilin and the CIMPR come down in spastizin pulldowns, so cargo is being selected somehow.
Significance: Until it is closed, every AP-5 subunit is annotatable only by location, complex membership and pathway participation; the complex's defining property as an adaptor - which cargo it selects and through which surface - is unrepresentable, and the five protein-binding rows on this gene cannot be upgraded to anything informative.
What would resolve it: Test the reconstituted AP-5 core against a panel of sorting motifs with AP-2 as a positive control, including a zeta N-terminal deletion to ask whether a latent site is being occluded; or find the cargo determinant by comparing the cytosolic tails of the proteins whose retrieval fails when AP-5 is lost.
Provenance (the field's own admissions):
Gap: GO has no term for the AP-5/spatacsin/spastizin assembly, although that six-protein complex, not the AP-5 tetramer, is the entity that has been purified, phenotyped and solved.
OPEN ONTOLOGY CC_DARK
What is known: Established: the six proteins associate stably and at equal stoichiometry, all six knockdowns give the same cargo phenotype, and the assembly has been reconstituted for cryo-EM. Established on the ontology side: GO:0044599 exists for the tetramer and is cross-referenced by ComplexPortal's CPX-5181, whereas CPX-20045 for the hexamer carries no GO cross-reference, and no GO term returned by searches for spatacsin, spastizin or AP-5 covers it.
Significance: ComplexPortal's curation of the hexamer currently reaches AP5S1 only as location and process rows, never as complex membership, and spatacsin and spastizin have no complex term at all despite being the two proteins whose mutations define the disease.
What would resolve it: Create the complex term set out in proposed_new_terms and map CPX-20045 to it.
Provenance (the field's own admissions):
Proposed term (ontology gap):
Gap: AP5S1 is the only AP-5 subunit with no reported disease-causing variant, and no AP5S1 knockout has been characterised, so it is unknown whether losing sigma-5 has the same consequences as losing the other three subunits.
OPEN BIOLOGY
What is known: Established: bi-allelic variants in AP5Z1, AP5M1 and AP5B1 cause a recessive macular dystrophy, AP5Z1 variants also cause SPG48, and the authors of the genetics study state explicitly that whether AP5S1 variants cause disease is still unclear and would be very rare. Established for cell biology: sigma-5 siRNA phenocopies the other subunits, but every deeper phenotype - the knockout proteomics, the organellar maps, the patient fibroblasts, the mouse models - was obtained on AP5Z1 or on SPG11/SPG15.
Significance: It leaves open whether sigma-5 is dispensable for AP-5 function - a real possibility for a small subunit whose only measured contacts are with zeta and spatacsin - and it means the strongest phenotypic evidence on this gene is a 2011 siRNA experiment.
What would resolve it: Knock out AP5S1 in HeLa and repeat the dynamic organellar mapping and vesicle-fraction proteomics used for the AP5Z1 knockout, asking whether the AP-5 tetramer still forms and whether the retrieval defect is reproduced.
Provenance (the field's own admissions):
Gap: Whether the reduction in homologous recombination on sigma-5 knockdown reflects a direct role or a downstream consequence of disrupted endolysosomal traffic has never been tested.
OPEN BIOLOGYCURATION
What is known: Established: two independent esiRNAs against C20orf29 reduced recombination in the DR-GFP reporter. Also established: in the same experiment, knockdown of two other members of the same complex had no effect; the protein localises to late endosomes and lysosomes, not to the nucleus; and no follow-up in fifteen years has connected AP-5 to DNA repair.
Significance: The claim is currently annotated twice on this gene, once as IMP and once electronically from the family signature, and the electronic route carries it to all members of the family across eukaryotes.
What would resolve it: Re-test recombination in AP5S1-null cells with cell-cycle and proliferation controls, and ask whether a separation-of-function allele that leaves the complex intact but blocks membrane recruitment also affects repair.
Provenance (the field's own admissions):
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)