AP3S2 encodes sigma-3B (also called sigma-2), one of the two interchangeable small subunits of adaptor protein complex 3 (AP-3), the heterotetrameric coat adaptor that selects transmembrane cargo for delivery from endosomes to lysosomes and to lysosome-related organelles such as melanosomes, platelet dense granules and lung lamellar bodies. AP-3 comprises two large adaptins (delta and beta-3), a medium adaptin (mu-3A or mu-3B) and a single small adaptin, which in human cells is either sigma-3A or sigma-3B, so the sigma-3B form is a variant of the same machine rather than a separate one. The complex is cytosolic at rest and is recruited onto endosomal membranes by ARF1-GTP acting principally through the delta subunit, together with phosphoinositides; it then polymerises with ARF1 into spiralling arrays that deform the membrane into tubular carriers. Unlike AP-1 and AP-2, AP-3 builds those carriers without a clathrin lattice. Within the complex the sigma subunit supplies half of the composite site that reads acidic dileucine sorting signals in the cytosolic tails of cargo proteins. Such signals bind neither the delta nor the sigma chain alone but the delta-sigma3 hemicomplex, and the hydrophobic pockets that receive the motif's two bulky residues lie on sigma: in sigma-3A, substituting Val94 or Leu109 abolishes signal binding, and sigma-3B carries both residues unchanged at the same positions. A delta-sigma3B hemicomplex binds the dileucine signal of the lysosomal membrane protein LIMP-II with the same sequence requirements as its sigma-3A counterpart. In the resting complex that pocket is capped by the N-terminal extension of beta-3, and cargo binding is part of what opens the complex up. Sigma-3B is a 193-residue longin/roadblock-fold protein with no catalytic activity; it is expressed in every adult tissue examined, and it is identical in sequence to its mouse ortholog. The biology AP-3 carries out through this subunit is cargo-specific rather than organelle-generic. In neurons it packages synaptic-vesicle membrane proteins such as ZnT3 and ClC-3 at cell bodies for delivery to nerve terminals; in pigment cells it routes tyrosinase and TYRP1 from early and recycling endosomal tubules to maturing melanosomes; and mutations in AP-3 subunits cause Hermansky-Pudlak syndrome, with abnormal melanosomes, platelet dense granules and lung lamellar bodies. What is not known is anything that distinguishes the sigma-3B form of AP-3 from the sigma-3A form: no cargo, tissue, phenotype or structure separates them.
Definition: Binding to an acidic dileucine sorting signal, a [DE]XXXL[LI] peptide motif in the cytosolic tail of a transmembrane cargo protein, which directs that protein into carriers formed by heterotetrameric adaptor protein complexes.
Justification: GO can express the tyrosine-motif counterpart read by the medium adaptin (GO:0089710 endocytic targeting sequence binding, whose definition requires an essential tyrosine), but it has no term for the dileucine motif read by the small adaptin together with the large one. The result is that the one activity a sigma subunit demonstrably participates in cannot be stated: AP3S2's only molecular-function row in GOA is AP-3 adaptor complex binding, which describes the protein as binding the complex it is part of. The gap is not speculative - the site is mapped by substitution in all three complexes, and sigma-3B itself was assayed. The absence was checked against the GO release served by QuickGO by retrieving the definitions of GO:0035615, GO:0089710, GO:0060090, GO:0030674, GO:0038024 and GO:0035459; none of them names a dileucine or sorting-signal binding activity.
Parent term: protein binding
Supporting Evidence:
Definition: Bringing together a cargo protein and a membrane coat, thereby selecting that cargo for incorporation into a coated transport carrier.
Justification: GO:0035615 clathrin-cargo adaptor activity is the term GO uses for AP-complex cargo selection, but its definition - bringing a cargo protein together with clathrin, responsible for the formation of endocytic vesicles - has two commitments that AP-3 does not meet. AP-3 does not copurify with clathrin-coated vesicles and reconstituted AP3:ARF1 forms carriers with no clathrin lattice, which is now offered as the explanation for the clathrin independence of AP-3-mediated trafficking; and AP-3 acts on endosomes, not in endocytosis. UniProt already describes AP-3 as not clathrin-associated. The consequence is that the AP-3 subunits have no complex-level cargo-adaptor term at all and fall back on GO:0030674 protein-macromolecule adaptor activity, which does not say that the partner is a coat. A coat-agnostic parent, with GO:0035615 as its clathrin child, would serve AP-3 and AP-4 as well as any other non-clathrin coat adaptor. The same revision would fix GO:0035654, whose definition asserts that AP-3 loads cargo into clathrin-coated vesicles.
Parent term: molecular adaptor activity
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0016192 vesicle-mediated transport | IBA GO_REF:0000033 | ACCEPT | Summary: AP-3 is a vesicle coat adaptor, and vesicle-mediated transport is the process it acts in; sigma-3B is a constituent subunit of it [PMID:9118953]. The phylogenetic inference is sound at the node and general for the right reason. Reason: The IBD behind this row sits on PTHR11753 node PTN000204281 and is seeded by ten sigma subunits, two of which (S. cerevisiae APS3, C. albicans APS3) are AP-3 sigma orthologs; the rest are AP-1 and AP-2 sigmas. AP3S2 sits squarely inside the clade, and there is no target-specific evidence of loss or divergence: the longin/roadblock fold is intact, the dileucine-binding pocket is intact at the same positions as in AP3S1, there are no NOT rows, and the protein is expressed in every adult tissue examined. The term's generality is not an under-call - the seeds disagree about destination (trans-Golgi/endosome for AP-1, plasma membrane for AP-2, endosome/lysosome for AP-3), so vesicle-mediated transport is the correct least common ancestor rather than a missed opportunity for a child term. One donor on the row, FB:FBgn0043012 (D. melanogaster AP-2sigma), is not in the current GO:0016192 seed list but is a seed of the same node's GO:0043231 IBD, whose slice line is three months older; that is a release-timing difference between GOA and the PAINT slice, not a donor the tree lacks. Core. Propagation Review Root cause: NO FAILURE CORE Sources checked: CGD:CAL0000182525 SUPPORTS TRANSFER Q59QC5 APS3_CANAL, the Candida albicans AP-3 sigma subunit - a seed from the same complex as the target. FB:FBgn0039132 SUPPORTS TRANSFER Drosophila AP-1sigma (CG5864). FB:FBgn0043012 SUPPORTS TRANSFER Drosophila AP-2sigma (CG6056, Q9VDC3). Present on the GOA row but absent from the current GO:0016192 seed list in the PAINT slice; it is a seed of the same node's older GO:0043231 IBD line, so the tree does contain it. MGI:MGI:1098244 SUPPORTS TRANSFER P61967 AP1S1_MOUSE, mouse AP-1 sigma-1A. MGI:MGI:1889383 SUPPORTS TRANSFER Q9DB50 AP1S2_MOUSE, mouse AP-1 sigma-1B. PANTHER:PTN000204281 SUPPORTS TRANSFER PTHR11753 IBD node PTN000204281, seeded by ten sigma subunits spanning AP-1, AP-2 and AP-3 across fungi, nematode, fly, rodent and human. Two seeds are AP-3 sigma orthologs (S. cerevisiae APS3, C. albicans APS3), so AP-3 is inside the clade the function is asserted for, and AP3S2 is an AP-3 sigma subunit. The remaining eight seeds act on different itineraries, which is why the node carries the parent process term rather than a destination-specific child. PomBase:SPAP27G11.06c SUPPORTS TRANSFER Q9P7N2 AP1S1_SCHPO (vas2/aps1), fission-yeast AP-1 sigma; same role. RGD:620188 SUPPORTS TRANSFER P62744 AP2S1_RAT, rat AP-2 sigma. SGD:S000003561 SUPPORTS TRANSFER P47064 AP3S_YEAST (APS3), the budding-yeast AP-3 sigma subunit. The most directly relevant seed on the row: same subunit, same complex. SGD:S000004160 SUPPORTS TRANSFER P35181 AP1S1_YEAST (APS1), an AP-1 sigma. Supports the family-level process term, not an AP-3-specific itinerary. UniProtKB:P53680 SUPPORTS TRANSFER AP2S1_HUMAN, the human AP-2 sigma subunit - a within-species paralog of the target, contributing the human experimental evidence to the node. WB:WBGene00000157 SUPPORTS TRANSFER C. elegans aps-2 (F02E8.3, Q19123), AP-2 sigma. Not resolvable through UniProt's xref index; identified via the GO API bioentity record. Supporting Evidence: PMID:9118953 sigma3A and sigma3B are components of a large complex, named AP-3, that also contains proteins of apparent molecular masses of 47, 140 and 160 kDa. PMID:39705307 Adaptor protein complex-3 (AP-3) mediates cargo sorting from endosomes to lysosomes and lysosome-related organelles. file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md Two of the ten seeds are genuine AP-3 sigma orthologs file:human/AP3S2/AP3S2-uniprot.txt DR PAN-GO; P59780; 1 GO annotation based on evolutionary models. |
| GO:0005794 Golgi apparatus | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt places AP-3 at the Golgi apparatus, and the antibody localisation that underlies it saw the sigma-3-containing complex in the TGN area [PMID:9118953]. Reason: The SubCell mapping reproduces the UniProt SUBCELLULAR LOCATION line faithfully, and that line is grounded in the original localisation of the sigma-3-containing complex to the TGN region and peripheral structures. Mammalian AP-3's main site of action is endosomal rather than Golgi, so this is the broader of the two compartments it occupies, but it is not wrong and the term is a parent of the TGN. Accepted as a genuine location. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0132 SUPPORTS TRANSFER SubCell 'Golgi apparatus', mapped from the UniProt SUBCELLULAR LOCATION line, which is itself an ECO:0000250 by-similarity statement backed by the 1997 immunofluorescence of the sigma-3 complex. Supporting Evidence: file:human/AP3S2/AP3S2-uniprot.txt CC -!- SUBCELLULAR LOCATION: Golgi apparatus. Cytoplasmic vesicle membrane PMID:9118953 Immunofluorescence microscopy analyses show that the sigma3-containing complex is present both in the area of the TGN and in peripheral structures, some of which contain the transferrin receptor. |
| GO:0006886 intracellular protein transport | IEA GO_REF:0000002 | ACCEPT | Summary: Moving transmembrane proteins between intracellular compartments is what AP-3 is for, so this term is core even though it is reached from a family signature [PMID:9545220]. Reason: IPR000804 is the clathrin small-chain signature shared by all AP-complex sigma subunits, and intracellular protein transport is true of every one of them - but being family-wide is not the same as being peripheral. Signal-mediated sorting of membrane proteins between intracellular compartments is the function of AP-3 [PMID:9545220], and the term is carried in this review's core_functions as one of the processes the subunit is directly involved in. Accepted as core; the more specific destination is supplied by the replacement proposed on the GO:0006896 row, and there is no conflict between holding a parent and a child. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR000804 SUPPORTS TRANSFER Clathrin adaptor small-chain signature, present on the target (PROSITE PS00989 / Pfam PF01217 on the same record). The mapped term is a family-wide generality and applies. Supporting Evidence: PMID:9545220 A heterotetrameric complex termed AP-3 is involved in signal-mediated protein sorting to endosomal-lysosomal organelles. PMID:39705307 Adaptor protein complex-3 (AP-3) mediates cargo sorting from endosomes to lysosomes and lysosome-related organelles. |
| GO:0006896 Golgi to vacuole transport | IEA GO_REF:0000002 | MODIFY | Summary: The right pathway for fungal AP-3, the wrong compartment of origin for the human complex. Mammalian AP-3 sorts from endosomes to lysosomes and lysosome-related organelles [PMID:39705307], not from the Golgi to the vacuole. Reason: IPR027155 is the APS3 signature, so the mapping is correctly targeted at AP-3 sigma subunits; the problem is that the GO term it maps to describes the budding-yeast ALP pathway, in which AP-3 genuinely does run Golgi-to-vacuole. In mammals the characterised itinerary begins at early/recycling endosomal tubules [PMID:23247405] and ends at lysosomes and lysosome-related organelles [PMID:39705307]. GO:0008333 endosome to lysosome transport states the human route without importing the fungal compartment. This is a term-scoping problem in the InterPro2GO mapping rather than a wrong family assignment. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: COMPARTMENT OR COMPLEX MISMATCH Sources checked: InterPro:IPR027155 SUPPORTS SOURCE BUT NOT TARGET APS3, the AP-3 sigma signature; the family call is right and the target does carry it. What does not transfer is the mapped term's compartment of origin, which reflects the fungal Golgi-to-vacuole pathway rather than the mammalian endosome-to-lysosome one. Proposed replacements: endosome to lysosome transport Supporting Evidence: PMID:39705307 Adaptor protein complex-3 (AP-3) mediates cargo sorting from endosomes to lysosomes and lysosome-related organelles. PMID:23247405 Rab32 and Rab38 interact with AP-1, AP-3 and BLOC-2 on early/recycling endosome tubules, where cargo such as tyrosinase and Tyrp1 are loaded into vesicles or transport intermediates. PMID:23247405 Packaging of the tyrosinases into transport vesicles at early/recycling endosome-associated tubules is dependent on ubiquitous adaptor protein complex (AP)-1 and AP-3 |
| GO:0015031 protein transport | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: The most general true statement about an AP-complex subunit. Reason: IPR016635 is the AP-complex small-subunit signature and protein transport is the process every member serves. Correct, and the UniProt keyword block carries the same claim, so there is nothing to remove or repair. But GO:0015031 is the parent of the intracellular transport term this review does take as core, and its definition extends to movement into, out of and between cells - a scope AP-3 does not cover. Kept, marked non-core, so that the grading matches the reasoning rather than promoting the broadest available parent to a core function. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: InterPro:IPR016635 SUPPORTS TRANSFER AP_complex_ssu, the family signature listed on the target's own record. The mapped term is correct at family level. Supporting Evidence: PMID:9545220 A heterotetrameric complex termed AP-3 is involved in signal-mediated protein sorting to endosomal-lysosomal organelles. file:human/AP3S2/AP3S2-uniprot.txt DR PANTHER; PTHR11753; ADAPTOR COMPLEXES SMALL SUBUNIT FAMILY; 1. |
| GO:0016192 vesicle-mediated transport | IEA GO_REF:0000002 | ACCEPT | Summary: Same conclusion as the IBA row, reached from the domain signature rather than from the tree. Reason: IPR000804 maps to vesicle-mediated transport for all AP-complex small chains, and AP-3 is a vesicle coat adaptor [PMID:42139345]. Duplication with the IBA and NAS rows for the same term is normal and not a defect. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR000804 SUPPORTS TRANSFER Clathrin adaptor small-chain signature on the target. The mapped process is the same one the PANTHER node asserts, arrived at independently. Supporting Evidence: PMID:42139345 AP3:ARF1 spontaneously remodels membranes containing cargo and the phosphoinositide PI(3,5)P2 into tubular structures coated in spiraling rows of AP3 arches and ARF1 dimers PMID:39705307 Adaptor protein complex-3 (AP-3) mediates cargo sorting from endosomes to lysosomes and lysosome-related organelles. |
| GO:0030117 membrane coat | IEA GO_REF:0000002 | ACCEPT | Summary: AP-3 forms the inner layer of a membrane coat, and the GO definition explicitly covers non-clathrin coats. Reason: GO:0030117 is defined to include coats other than clathrin's, which matters here because AP-3 builds carriers without a clathrin lattice [PMID:42139345]. Cryo-electron tomography shows AP-3 and ARF1 polymerising into spiralling arches on tubular membranes, which is a membrane coat in the literal sense. UniProt makes the same statement. Accepted. Propagation Review Root cause: NO FAILURE CORE Sources checked: InterPro:IPR000804 SUPPORTS TRANSFER Clathrin adaptor small-chain signature. The mapped CC term is deliberately coat-agnostic, so it survives the finding that AP-3 coats are clathrin-free. Supporting Evidence: PMID:42139345 AP3:ARF1 spontaneously remodels membranes containing cargo and the phosphoinositide PI(3,5)P2 into tubular structures coated in spiraling rows of AP3 arches and ARF1 dimers file:human/AP3S2/AP3S2-uniprot.txt CC cytoplasmic face of coated vesicles located at the Golgi complex. |
| GO:0030123 AP-3 adaptor complex | IEA GO_REF:0000120 | ACCEPT | Summary: The automatic route to the same conclusion as the IDA row, and both halves of its evidence were checked and hold. Reason: InterPro:IPR027155 is the APS3 signature, which is AP-3-sigma-specific. The other half, ARBA00033921, fires on FunFam 3.30.450.60:FF:000001 plus taxon Primates - the shape of rule that can over-generalise, because FunFam signatures carry family-level names. Tested against the complete set of reviewed human AP-complex sigma subunits, it does not: AP1S1 and AP1S3 carry FF:000005, AP1S2 FF:000009, AP2S1 FF:000004, AP4S1 FF:000010 and AP5S1 none at all, while only AP3S1 and AP3S2 carry FF:000001. Zero non-AP-3 subunits would be mis-called. The row is sound. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00033921 SUPPORTS TRANSFER Fetched from the UniProt ARBA service: conditions are FunFam 3.30.450.60:FF:000001 AND taxon Primates, asserting GO:0030123. The FunFam discriminates - among the eight reviewed human AP-complex sigma subunits only AP3S1 and AP3S2 carry it. InterPro:IPR027155 SUPPORTS TRANSFER APS3, the AP-3 sigma subunit signature, present on the target. AP-3-specific, so it grounds the complex assignment directly. Supporting Evidence: file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md exactly the two AP-3 sigma subunits match file:human/AP3S2/AP3S2-uniprot.txt DR FunFam; 3.30.450.60:FF:000001; AP complex subunit sigma; 1. PMID:9118953 sigma3A and sigma3B are components of a large complex, named AP-3, that also contains proteins of apparent molecular masses of 47, 140 and 160 kDa. |
| GO:0030659 cytoplasmic vesicle membrane | IEA GO_REF:0000044 | ACCEPT | Summary: AP-3 is a peripheral membrane protein on the cytoplasmic face of coated vesicles; the UniProt record says so and the SubCell mapping reproduces it. Reason: The location is intrinsic to what a coat adaptor is: it sits on the cytosolic leaflet of the carrier it coats, which the 2026 tomography visualises directly as rows of AP-3 arches on tubular membranes [PMID:42139345]. Accepted as a core location. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0089 SUPPORTS TRANSFER SubCell 'Cytoplasmic vesicle membrane', mapped from the UniProt SUBCELLULAR LOCATION line, whose Note describes the protein as a component of the coat on the cytoplasmic face of coated vesicles. Supporting Evidence: file:human/AP3S2/AP3S2-uniprot.txt CC cytoplasmic face of coated vesicles located at the Golgi complex. PMID:42139345 AP3:ARF1 spontaneously remodels membranes containing cargo and the phosphoinositide PI(3,5)P2 into tubular structures coated in spiraling rows of AP3 arches and ARF1 dimers |
| GO:1904115 axon cytoplasm | IEA GO_REF:0000108 | KEEP AS NON CORE | Summary: A mechanical inter-ontology inference off the anterograde axonal transport row; it inherits that row's grade rather than adding evidence of its own. Reason: GO_REF:0000108 derives a location from a process annotation: if the protein is involved in anterograde axonal transport, it is in the axon cytoplasm. The inference is valid, and AP-3 is genuinely present in neuronal processes, so the row is not wrong. But it is only as strong as its input, and that input is a complex-level, neuron-specific claim derived from a delta-subunit mutant [PMID:21998198]. Kept, non-core: a ubiquitously expressed Golgi/endosomal coat subunit is not characterised by its presence in axoplasm. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: GO:0008089 SUPPORTS TRANSFER The source is a GO term, not a gene product: this row is a logical inference from AP3S2's own anterograde axonal transport annotation, which is reviewed on rows 12 and 24 and kept as non-core. The inference step itself is sound; its strength is inherited. Supporting Evidence: PMID:21998198 Deficiencies in AP-3 and BLOC-1 affect synaptic vesicle composition. PMID:21998198 Synaptosome fractions from control brains (lanes 1β8) and AP-3βdeficient mocha (Ap3d1mh/mh) brains |
| GO:0008021 synaptic vesicle | IEA GO_REF:0000107 | MARK AS OVER ANNOTATED | Summary: Transferred from a rat synaptic-vesicle proteomic detection whose own authors list AP3 among endosomal 'visitor' proteins rather than vesicle residents [PMID:33376223]. Reason: The ortholog transfer is mechanically fine - the donor is rat Ap3s2 - but the donor annotation is a mass-spectrometric detection in a purified synaptic-vesicle fraction, and the paper reporting it explicitly offers endosomal contamination and transient visitors as the explanation for AP3's presence. A qualifier of is_active_in asks for more than detection: it asks that the protein performs its function there. AP-3's function is to bud carriers from endosomal membranes, which is upstream of the mature vesicle. The detection is real and reproducible, so this is over-annotation rather than error; not removed. Propagation Review Root cause: SOURCE WEAK OR INFERRED Failure modes: COMPARTMENT OR COMPLEX MISMATCH Sources checked: UniProtKB:A0A0G2K302 SOURCE WEAK OR INFERRED Rat Ap3s2 (unreviewed). Its GO:0008021 rests on a SynGO EXP/IDA from a synaptic-vesicle proteome in which the authors themselves list AP3 among endosomal 'visitor' proteins rather than vesicle residents [PMID:33376223]. ensembl:ENSRNOP00000072451 SOURCE WEAK OR INFERRED The Ensembl protein model for the same rat gene as UniProtKB:A0A0G2K302 - one donor, two identifiers. Supporting Evidence: PMID:33376223 This may explain the presence of endosomal-related proteins (e.g., Stx7, AP3) or proteins of the AZ (e.g., Piccolo, Bassoon) in the SV proteome. PMID:39705307 Adaptor protein complex-3 (AP-3) mediates cargo sorting from endosomes to lysosomes and lysosome-related organelles. |
| GO:0008089 anterograde axonal transport | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: AP-3 packages synaptic cargo at neuronal cell bodies for delivery to nerve terminals, so its subunits participate in the process - but as the packaging step, not as the motor, and the underlying genetics is on the delta subunit. Reason: The donor is mouse Ap3s2, which is identical in sequence to the human protein, so there is no orthology objection at all. The mouse annotation traces to PMID:21998198, where the AP-3 loss is Ap3d1 mocha and the sigma subunit is detected with a pan-sigma-3 antibody - the same antibody class that resolves sigma-3A and sigma-3B only as a doublet [PMID:9151686]. The claim that reaches AP3S2 is therefore complex-level and neuronal. GO:0008089 is moreover defined as movement along microtubules from cell body to periphery, a motor-driven process in which a coat adaptor participates by loading the cargo rather than by moving it. Correct to keep, wrong to call core for a ubiquitously expressed subunit. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: CONTEXT OR TISSUE MISMATCH Sources checked: UniProtKB:Q8BSZ2 SUPPORTS TRANSFER Mouse Ap3s2, the one-to-one ortholog and the only gene-product donor on this row. The transfer is as safe as an ortholog transfer gets: human and mouse AP3S2 are identical in sequence over all 193 residues. What the donor annotation itself rests on is a delta-subunit mutant read out with a pan-sigma-3 antibody, so the inherited claim is complex-level. ensembl:ENSMUSP00000075082 SUPPORTS TRANSFER The Ensembl protein model for the same mouse gene as UniProtKB:Q8BSZ2 - one donor, two identifiers, not two independent lines of evidence. Supporting Evidence: PMID:21998198 Synaptosome fractions from control brains (lanes 1β8) and AP-3βdeficient mocha (Ap3d1mh/mh) brains PMID:21998198 antibodies against synaptic vesicle markers (SV2, synaptophysin), AP-3βdependent synaptic vesicle cargoes (PI4KIIΞ±, VAMP7, ZnT3), and AP-3 Ο3 subunit file:human/AP3S2/AP3S2-uniprot.txt CC cell bodies for delivery into neurites and nerve terminals. file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md Human and mouse AP3S2 are byte-for-byte identical proteins |
| GO:0035651 AP-3 adaptor complex binding | IEA GO_REF:0000107 | MARK AS OVER ANNOTATED | Summary: AP3S2 is not a binder of AP-3; it is a constituent of AP-3, which the same record already states experimentally as part_of GO:0030123 [PMID:9118953]. Reason: GO:0035651 is defined as binding to an AP-3 adaptor complex. Applying it to a core subunit of that complex conflates being a part with binding a partner, and it adds nothing the part_of IDA row does not already say more precisely. The mouse source is an MGI IDA from a study that purified cross-linked AP-3 complexes and identified associated proteins by mass spectrometry [PMID:19010779] - an experiment in which a constituent subunit necessarily appears, so the curator's reading of the assay is not in question; what does not follow is that the subunit enables complex binding. Marked over-annotated rather than removed: the underlying observation is genuine and the ortholog transfer is faithful. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: ROLE CONFLATION Sources checked: UniProtKB:Q8BSZ2 SUPPORTS SOURCE BUT NOT TARGET Mouse Ap3s2, identical in sequence to the human protein, so the ortholog step is faultless. The problem is upstream of it: the donor term casts a constituent subunit as a binder of its own complex. ensembl:ENSMUSP00000075082 SUPPORTS SOURCE BUT NOT TARGET The Ensembl protein model for the same mouse gene as UniProtKB:Q8BSZ2 - one donor, two identifiers, not two independent lines of evidence. Supporting Evidence: PMID:19010779 purification of cross-linked AP-3 complexes and mass spectrometric identification of associated proteins PMID:9118953 sigma3A and sigma3B are components of a large complex, named AP-3, that also contains proteins of apparent molecular masses of 47, 140 and 160 kDa. PMID:9151686 Antibodies raised against recombinant delta and sigma3 show that they are the other two subunits of the adaptor-like complex. |
| GO:0048490 anterograde synaptic vesicle transport | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: The synaptic-vesicle-specific form of the axonal transport row, with the same source and the same limits. Reason: Real AP-3 biology, but neuronal and complex-level rather than a function of this ubiquitously expressed subunit: the perturbation behind the mouse source annotation is the delta subunit (Ap3d1 mocha) and the sigma detection uses a pan-sigma-3 antibody that does not separate sigma-3A from sigma-3B [PMID:21998198]. Kept, marked non-core. The specific process - movement of synaptic vesicles along axonal microtubules to the presynapse - again describes motor-driven transport, whereas the AP-3 contribution demonstrated in the source is the loading of vesicle membrane proteins at the cell body [PMID:21998198]. Real, inherited, peripheral. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: CONTEXT OR TISSUE MISMATCH Sources checked: UniProtKB:Q8BSZ2 SUPPORTS TRANSFER Mouse Ap3s2, the one-to-one ortholog and the only gene-product donor on this row. The transfer is as safe as an ortholog transfer gets: human and mouse AP3S2 are identical in sequence over all 193 residues. What the donor annotation itself rests on is a delta-subunit mutant read out with a pan-sigma-3 antibody, so the inherited claim is complex-level. ensembl:ENSMUSP00000075082 SUPPORTS TRANSFER The Ensembl protein model for the same mouse gene as UniProtKB:Q8BSZ2 - one donor, two identifiers, not two independent lines of evidence. Supporting Evidence: PMID:21998198 Deficiencies in AP-3 and BLOC-1 affect synaptic vesicle composition. PMID:15537701 beta3B deficiency compromised synaptic zinc stores assessed by Timm's staining and the synaptic vesicle targeting of membrane proteins involved in zinc uptake (ZnT3 and ClC-3) file:human/AP3S2/AP3S2-uniprot.txt CC cell bodies for delivery into neurites and nerve terminals. |
| GO:0005769 early endosome | NAS PMID:23247405 Cell type-specific Rab32 and Rab38 cooperate with the ubiqui... | ACCEPT | Summary: Early/recycling endosomal tubules are where AP-3 loads cargo, which the cited review states directly [PMID:23247405]. Reason: This is the compartment modern work puts AP-3 in: cargo packaging occurs at early/recycling endosome-associated tubules, and the complex sorts from there to lysosomes and lysosome-related organelles [PMID:39705307]. The original localisation of the sigma-3-containing complex already showed peripheral structures alongside the TGN, some of them transferrin-receptor positive [PMID:9118953] - i.e. early/recycling endosomes. A ComplexPortal NAS is an author-statement-grade annotation made on the complex, but the statement is well founded and the subunit is in the complex. Core location. Supporting Evidence: PMID:23247405 Rab32 and Rab38 interact with AP-1, AP-3 and BLOC-2 on early/recycling endosome tubules, where cargo such as tyrosinase and Tyrp1 are loaded into vesicles or transport intermediates. PMID:23247405 Packaging of the tyrosinases into transport vesicles at early/recycling endosome-associated tubules is dependent on ubiquitous adaptor protein complex (AP)-1 and AP-3 PMID:9118953 Immunofluorescence microscopy analyses show that the sigma3-containing complex is present both in the area of the TGN and in peripheral structures, some of which contain the transferrin receptor. |
| GO:0016183 synaptic vesicle coating | NAS PMID:15537701 Genetic analysis of the neuronal and ubiquitous AP-3 adaptor... | MODIFY | Summary: The term means clathrin-coated pit formation at the presynaptic plasma membrane, which is not what AP-3 does and not what the cited paper reports. Reason: Read by its definition rather than its label, GO:0016183 is about the formation of clathrin-coated pits in the presynaptic membrane endocytic zone. AP-3 acts on endosomal membranes, and the cited paper measured the synaptic-vesicle content of ZnT3 and ClC-3 in beta-3 isoform knockout mice [PMID:15537701] - vesicle cargo composition, not presynaptic pit formation. Two further facts point away from the term: AP-3 does not copurify with clathrin-coated vesicles and builds its carriers without a clathrin lattice [PMID:42139345]. GO:0016182 synaptic vesicle budding from endosome names the step AP-3 is actually credited with and keeps the annotation in the synaptic-vesicle domain the source paper is about. Two caveats on the replacement. It is as neuronal and as complex-level as the GO:0008089, GO:0048490 and GO:0036465 rows this review grades KEEP_AS_NON_CORE, and it should be read with the same reservation - MODIFY cannot carry the non-core marker, so it is stated here instead. And GO:0035459 vesicle cargo loading, the replacement chosen on the GO:0035654 row, arguably matches the cited paper's actual readout more literally, since it measured vesicle cargo content rather than budding; GO:0016182 is preferred because it keeps the annotation in the synaptic-vesicle branch, which is the whole point of the original term, and because endosomal budding is the step AP-3 is independently credited with in neurons. Proposed replacements: synaptic vesicle budding from endosome Supporting Evidence: PMID:15537701 beta3B deficiency compromised synaptic zinc stores assessed by Timm's staining and the synaptic vesicle targeting of membrane proteins involved in zinc uptake (ZnT3 and ClC-3) PMID:15537701 Our results suggest that concerted nonredundant functions of neuronal and ubiquitous AP-3 provide a mechanism to control the levels of selected membrane proteins in synaptic vesicles. PMID:42139345 AP3 does not copurify with clathrin-coated vesicles PMID:42139345 By demonstrating that AP3:ARF1 can generate carriers without using a clathrin lattice, we explain the clathrin independence of AP3-mediated trafficking. |
| GO:0016192 vesicle-mediated transport | NAS PMID:23247405 Cell type-specific Rab32 and Rab38 cooperate with the ubiqui... | ACCEPT | Summary: The same correct process term as the IBA and InterPro rows, asserted at complex level by ComplexPortal. Reason: AP-3 loads cargo into vesicles and transport intermediates at endosomal tubules [PMID:23247405] and polymerises with ARF1 into membrane-deforming coats [PMID:42139345]. Vesicle-mediated transport is the process, and sigma-3B is in the complex. Duplicate evidence for a term is not a defect. Supporting Evidence: PMID:23247405 Rab32 and Rab38 interact with AP-1, AP-3 and BLOC-2 on early/recycling endosome tubules, where cargo such as tyrosinase and Tyrp1 are loaded into vesicles or transport intermediates. PMID:42139345 AP3:ARF1 spontaneously remodels membranes containing cargo and the phosphoinositide PI(3,5)P2 into tubular structures coated in spiraling rows of AP3 arches and ARF1 dimers |
| GO:0030123 AP-3 adaptor complex | NAS PMID:9151686 Characterization of the adaptor-related protein complex, AP-... | ACCEPT | Summary: The second 1997 paper to identify sigma-3 as an AP-3 subunit, with antibodies raised separately against sigma-3A and sigma-3B [PMID:9151686]. Reason: Simpson et al. cloned sigma-3B independently, raised an antibody against it, and showed by reciprocal immunoprecipitation from brain cytosol that the sigma-3 chains are subunits of the adaptor-like complex; the sigma-3 signal in the delta and beta-3 immunoprecipitates runs as a doublet because both isoforms are present. The GO definition of the term names sigma3A and sigma3B as alternative occupants of the sigma slot, so the annotation matches the term exactly. Core, and independent of the IDA row's evidence. Supporting Evidence: PMID:9151686 Antibodies raised against recombinant delta and sigma3 show that they are the other two subunits of the adaptor-like complex. PMID:9151686 Ο3 (which appears as a doublet, presumably because there are two isoforms of the protein) PMID:9151686 The Ο3 antibodies bring down these four subunits as well |
| GO:0035654 clathrin-coated vesicle cargo loading, AP-3-mediated | NAS PMID:9545220 Association of the AP-3 adaptor complex with clathrin. | MODIFY | Summary: Cargo loading by AP-3 is exactly what sigma-3B contributes to; the clathrin-coated-vesicle half of the term is the part that has since been overturned [PMID:42139345]. Reason: The term is defined as the formation of a complex between AP-3 subunits and the proteins that will be transported by a clathrin-coated vesicle. The first half is the best available description of what a sigma subunit does: it supplies the pockets that receive a dileucine sorting signal [PMID:21097499, PMID:39705307]. The second half rests on the 1998 finding that AP-3 binds clathrin through the beta-3 appendage [PMID:9545220], which became consensus by 2013 [PMID:23247405] and has now been reversed: AP-3 does not copurify with clathrin-coated vesicles, and reconstituted AP3:ARF1 builds carriers with no clathrin lattice at all [PMID:42139345]. UniProt already describes the complex as not clathrin-associated. GO:0035459 vesicle cargo loading is the same assertion with the superseded commitment removed. The alternative was considered and rejected: keeping GO:0035654 and pursuing the definition fix would preserve AP-3 specificity, and that fix is proposed in proposed_new_terms and knowledge_gaps. It is not the right call for this row, because an annotation should not assert a mechanism its own evidence contradicts while the ontology catches up; generalising loses granularity, whereas keeping it would keep a falsehood. Proposed replacements: vesicle cargo loading Supporting Evidence: PMID:9545220 In vitro binding assays showed that mammalian AP-3 did associate with clathrin by interaction of the appendage domain of its beta3 subunit with the amino-terminal domain of the clathrin heavy chain. PMID:42139345 AP3 does not copurify with clathrin-coated vesicles PMID:42139345 By demonstrating that AP3:ARF1 can generate carriers without using a clathrin lattice, we explain the clathrin independence of AP3-mediated trafficking. file:human/AP3S2/AP3S2-uniprot.txt CC not clathrin-associated. The complex is associated with the Golgi PMID:41985787 A conserved N-terminal acidic dileucine motif (LEDRGL69L70) in TMEM163 is essential for interactions with BLOC-1 and AP-3 but dispensable for binding with AP-1, AP-2 and BLOC-2. PMID:41985787 our findings established TMEM163 as a cargo protein sequentially sorted by AP-3 and BLOC-1 via a shared dileucine-based sorting signal, which is essential for its proper trafficking to platelet dense granules |
| GO:0036465 synaptic vesicle recycling | NAS PMID:15537701 Genetic analysis of the neuronal and ubiquitous AP-3 adaptor... | KEEP AS NON CORE | Summary: AP-3 regenerates synaptic vesicles from endosomes and sets their membrane protein content, which is part of the recycling cycle but is neuronal and complex-level. Reason: The cited paper shows that removing either the ubiquitous or the neuronal AP-3 changes the synaptic-vesicle content of ZnT3 and ClC-3, and concludes that the two complexes control the levels of selected membrane proteins in synaptic vesicles [PMID:15537701]. That places AP-3 in the endosomal arm of synaptic vesicle recycling, so the term is defensible for the complex. It is not a core function of a ubiquitously expressed sigma subunit: the isoform variable in that paper is beta-3, and nothing there distinguishes sigma-3B from sigma-3A. Supporting Evidence: PMID:15537701 Neurons express adaptor (AP)-3 complexes assembled with either ubiquitous (beta3A) or neuronal-specific (beta3B) beta3 isoforms. PMID:15537701 beta3B deficiency compromised synaptic zinc stores assessed by Timm's staining and the synaptic vesicle targeting of membrane proteins involved in zinc uptake (ZnT3 and ClC-3) PMID:15537701 Our results suggest that concerted nonredundant functions of neuronal and ubiquitous AP-3 provide a mechanism to control the levels of selected membrane proteins in synaptic vesicles. |
| GO:0060155 platelet dense granule organization | NAS PMID:23247405 Cell type-specific Rab32 and Rab38 cooperate with the ubiqui... | KEEP AS NON CORE | Summary: A genuine AP-3 function in a specialised cell type: dense granules are lysosome-related organelles, and AP-3 subunit mutations disrupt them in Hermansky-Pudlak syndrome [PMID:23247405]. Reason: AP-3 belongs to the machinery that builds lysosome-related organelles, and HPS patients and models carrying AP-3 subunit mutations have abnormal platelet dense granules alongside abnormal melanosomes and lamellar bodies. The annotation is therefore right about the complex. It is cell-type restricted and is one downstream output of the generic endosome-to-lysosome sorting activity rather than a distinct function, and no sigma-3B-specific evidence exists for it; non-core. Supporting Evidence: PMID:23247405 Hermansky-Pudlak Syndrome (HPS) patients and the corresponding animal models have abnormal melanosomes, platelet dense granules and lamellar bodies of lung type II epithelial cells. PMID:23247405 Mutations in subunits of AP-3, BLOC-1, BLOC-2 and BLOC-3 underlie many forms of HPS. PMID:41985787 Transmembrane protein 163 (TMEM163), a zinc transporter, is drastically reduced in platelets of AP-3-, BLOC-1-, and BLOC-2-deficient Hermansky-Pudlak syndrome mice and patients PMID:41985787 our findings established TMEM163 as a cargo protein sequentially sorted by AP-3 and BLOC-1 via a shared dileucine-based sorting signal, which is essential for its proper trafficking to platelet dense granules |
| GO:1903232 melanosome assembly | NAS PMID:23247405 Cell type-specific Rab32 and Rab38 cooperate with the ubiqui... | KEEP AS NON CORE | Summary: The best-worked example of AP-3's lysosome-related-organelle role: tyrosinase and TYRP1 are packaged at endosomal tubules for delivery to maturing melanosomes [PMID:23247405]. Reason: Melanosome biogenesis depends on AP-3 acting with AP-1 and the BLOC complexes at early/recycling endosomal tubules, and loss of AP-3 produces a partial pigmentation defect. The complex-level claim is solid. As with the platelet dense granule row it is a specialised, cell-type-restricted expression of the same cargo-sorting activity, and is graded non-core for that reason rather than because it is doubted. Supporting Evidence: PMID:23247405 Packaging of the tyrosinases into transport vesicles at early/recycling endosome-associated tubules is dependent on ubiquitous adaptor protein complex (AP)-1 and AP-3 PMID:23247405 Rab32 and Rab38 interact with AP-1, AP-3 and BLOC-2 on early/recycling endosome tubules, where cargo such as tyrosinase and Tyrp1 are loaded into vesicles or transport intermediates. PMID:23247405 Mutations in subunits of AP-3, BLOC-1, BLOC-2 and BLOC-3 underlie many forms of HPS. |
| GO:0030123 AP-3 adaptor complex | IDA PMID:9118953 AP-3: an adaptor-like protein complex with ubiquitous expres... | ACCEPT | Summary: The one experimental annotation this gene has, and it is correct: sigma-3B was identified biochemically as a component of AP-3 [PMID:9118953]. Reason: Dell'Angelica et al. cloned both human sigma-3 paralogs, raised antibodies, and showed by Western blot and immunoprecipitation that sigma-3A and sigma-3B are components of a large complex containing 47, 140 and 160 kDa chains, which they named AP-3. The GO term's definition names sigma3A and sigma3B explicitly as the alternative occupants of the sigma slot, so term and evidence match without interpretation. This is the anchor of the whole review and the core cellular component assignment. The identical row is assigned to AP3S1 from the same paper, which is correct - the paper identified both. Supporting Evidence: PMID:9118953 sigma3A and sigma3B are components of a large complex, named AP-3, that also contains proteins of apparent molecular masses of 47, 140 and 160 kDa. PMID:9118953 Northern and Western blot analyses demonstrate that the products of both the sigma3A and sigma3B genes are expressed in a wide variety of tissues and cell lines. PMID:9118953 These results suggest that the sigma3 chains are components of a novel, ubiquitous adaptor-like complex involved in the recognition of tyrosine-based sorting signals. |
| GO:0008089 anterograde axonal transport | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: A manual ortholog transfer from mouse Ap3s2 - a paralogy-free transfer between sequence-identical proteins - carrying a complex-level neuronal claim. Reason: It is worth being explicit that the donor here is the mouse ortholog and not the within-species paralog AP3S1: no row in this gene's record transfers from AP3S1. Human and mouse AP3S2 are identical across all 193 residues, so the similarity premise of an ISS is satisfied about as completely as it can be, and a paralog-confusion objection does not arise. What limits the row is the donor evidence, not the transfer: the mouse IMP comes from Ap3d1 mocha brains probed with a pan-sigma-3 antibody [PMID:21998198], so it is a statement about neuronal AP-3. Kept as non-core, for the same reason as the Ensembl row for this term. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: CONTEXT OR TISSUE MISMATCH Sources checked: UniProtKB:Q8BSZ2 SUPPORTS TRANSFER Mouse Ap3s2, the one-to-one ortholog and the only gene-product donor on this row. The transfer is as safe as an ortholog transfer gets: human and mouse AP3S2 are identical in sequence over all 193 residues. What the donor annotation itself rests on is a delta-subunit mutant read out with a pan-sigma-3 antibody, so the inherited claim is complex-level. Supporting Evidence: file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md Human and mouse AP3S2 are byte-for-byte identical proteins PMID:21998198 Synaptosome fractions from control brains (lanes 1β8) and AP-3βdeficient mocha (Ap3d1mh/mh) brains PMID:21998198 antibodies against synaptic vesicle markers (SV2, synaptophysin), AP-3βdependent synaptic vesicle cargoes (PI4KIIΞ±, VAMP7, ZnT3), and AP-3 Ο3 subunit |
| GO:0048490 anterograde synaptic vesicle transport | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: The synaptic-vesicle form of the same manual mouse-to-human transfer, with the same strengths and the same ceiling. Reason: Donor is mouse Ap3s2, sequence-identical to the human protein; the transfer step is unimpeachable. The donor annotation is a complex-level neuronal result [PMID:21998198], and the term names a microtubule-motor process rather than the cargo-loading step AP-3 performs. Real but peripheral for this gene. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: CONTEXT OR TISSUE MISMATCH Sources checked: UniProtKB:Q8BSZ2 SUPPORTS TRANSFER Mouse Ap3s2, the one-to-one ortholog and the only gene-product donor on this row. The transfer is as safe as an ortholog transfer gets: human and mouse AP3S2 are identical in sequence over all 193 residues. What the donor annotation itself rests on is a delta-subunit mutant read out with a pan-sigma-3 antibody, so the inherited claim is complex-level. Supporting Evidence: file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md Human and mouse AP3S2 are byte-for-byte identical proteins PMID:21998198 Deficiencies in AP-3 and BLOC-1 affect synaptic vesicle composition. PMID:15537701 beta3B deficiency compromised synaptic zinc stores assessed by Timm's staining and the synaptic vesicle targeting of membrane proteins involved in zinc uptake (ZnT3 and ClC-3) |
| GO:0030674 protein-macromolecule adaptor activity | IPI PMID:14691137 Recognition of dileucine-based sorting signals from HIV-1 Ne... | NEW | Summary: The activity the gene actually performs and that its GO record does not record at all: as part of a delta-sigma3B hemicomplex, sigma-3B brings a cargo protein's sorting signal into the AP-3 coat. Reason: AP3S2's molecular function is unrepresented in GOA - the only MF row it carries is AP-3 adaptor complex binding, which mis-types a constituent as a binder. The activity that is documented is cargo selection: acidic dileucine signals bind no single AP subunit but a composite site formed by the large and small chains together [PMID:21097499], and Janvier et al. cloned sigma-3A and sigma-3B separately and showed in a yeast three-hybrid assay that the LIMP-II cytosolic tail interacts with delta-sigma3B as well as with delta-sigma3A and gamma1-sigma1A, with the same three residues of the signal required in each case [PMID:14691137]. Sequence analysis confirms the structural basis transfers from the paralog: sigma-3A Val94 and Leu109, whose substitution abolishes signal binding, are present unchanged at the same positions in sigma-3B [file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md]. The qualifier is contributes_to because the binding site is shared with delta and the coat function belongs to the tetramer. GO:0035615 clathrin-cargo adaptor activity, which the AP-1 sigma subunits use, is not available here: its definition commits to clathrin and to endocytic vesicles, and AP-3 builds carriers without a clathrin lattice [PMID:42139345]. GO:0030674 is the nearest term that states the adaptor activity without importing a contradicted mechanism; the precise activity - binding a dileucine sorting signal - has no GO term and is proposed as a new one. The IPI interactors are recorded: UniProtKB:O14617 (AP3D1, the delta chain that forms the hemicomplex) and UniProtKB:Q14108 (SCARB2/LIMP-II, whose cytosolic tail is the bait). Supporting Evidence: PMID:14691137 yeast three-hybrid analyses showed that the cytosolic tail of LIMP-II interacted with Ξ³1βΟ1A, Ξ΄βΟ3A, and Ξ΄βΟ3B PMID:14691137 Of 12 residues that were mutated, only three were essential for interactions with Ξ³1βΟ1A, Ξ΄βΟ3A, and Ξ΄βΟ3B, the leucine-isoleucine pair and the glutamate residue at position β4 PMID:14691137 ΞΌ3A and Ο1A (EcoRIβSalI), and Ο2, Ο3A, Ο3B, and Ο4 (BamHIβXhoI) were cloned into pGAD424. PMID:21097499 signals, on the other hand, do not bind to any single AP subunit but to combinations of Ξ³-Ο1, Ξ±-Ο2, and Ξ΄-Ο3 subunits, as demonstrated by the use of yeast three-hybrid (Y3H) and in vitro binding assays PMID:21097499 This is evidenced by the loss of signal binding by the Ο2 V88D or L103S substitutions and the homologous Ο1A V88D and I103S and Ο3A V94D and L109S substitutions. file:human/AP3S2/AP3S2-bioinformatics/RESULTS.md AP3S2 retains both pocket residues at the same native positions as AP3S1 PMID:42139345 By demonstrating that AP3:ARF1 can generate carriers without using a clathrin lattice, we explain the clathrin independence of AP3-mediated trafficking. |
| GO:0005198 structural molecule activity | IC PMID:9118953 AP-3: an adaptor-like protein complex with ubiquitous expres... | NEW | Summary: Sigma-3B occupies the small-chain position of the AP-3 heterotetramer and holds that position in the assembled complex. Reason: Inferred by curator from this gene's own experimentally supported part_of GO:0030123 row: sigma-3B is one of the four chains of AP-3 [PMID:9118953], and a core subunit of a heterotetramer contributes to the structural integrity of that complex, which is what GO:0005198 states. Independent support that the subunit participates in forming the complex rather than merely copurifying with it comes from the yeast three-hybrid reconstitution, in which the LIMP-II interaction is read out only when delta and sigma-3B are co-expressed [PMID:14691137], and from the cryo-EM core, where sigma-3's pocket is capped by the beta-3 N-terminal extension - an intersubunit contact [PMID:39705307]. Supporting Evidence: PMID:9118953 sigma3A and sigma3B are components of a large complex, named AP-3, that also contains proteins of apparent molecular masses of 47, 140 and 160 kDa. PMID:9151686 Antibodies raised against recombinant delta and sigma3 show that they are the other two subunits of the adaptor-like complex. PMID:14691137 yeast three-hybrid analyses showed that the cytosolic tail of LIMP-II interacted with Ξ³1βΟ1A, Ξ΄βΟ3A, and Ξ΄βΟ3B PMID:39705307 Surprisingly, the dileucine cargo-binding site on Ο3 is occupied by the N-terminal extension of Ξ²3 |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Do delta-sigma3A and delta-sigma3B hemicomplexes differ quantitatively in affinity or specificity for dileucine cargo tails, as the AP-1 gamma-sigma1 variants do, or are they interchangeable?
Suggested experts: Bonifacino J.S., Mattera R., Janvier K.
Q: Is the sigma slot of AP-3 filled stochastically in a given cell, or do sigma-3A and sigma-3B partition between distinct AP-3 populations - for example between the beta-3A and beta-3B complexes, or between different endosomal subdomains?
Suggested experts: Robinson M.S., Faundez V., Dell'Angelica E.C.
Q: Given that reconstituted AP3:ARF1 forms carriers without clathrin, does the beta-3 clathrin box do anything in cells, and should GO:0035654 continue to assert that AP-3 loads cargo into clathrin-coated vesicles?
Suggested experts: Owen D.J., Kaufman J.G.G., Baker R.W.
Q: Is AP3S2 the effector gene at the 15q26 type 2 diabetes locus, or is the signal carried by C15orf38, by the C15orf38-AP3S2 readthrough, or by PLIN1?
Suggested experts: Kooner J.S., Chambers J.C.
Experiment: Express and purify the human AP-3 core twice, once with AP3S1 and once with AP3S2, using the published baculovirus construct set in which the sigma subunit is a single interchangeable component. Measure binding to a panel of dileucine cargo tails (LIMP-II, TMEM163, RNF13, LAMP1) by surface plasmon resonance or isothermal titration calorimetry. A difference in any tail would be the first functional distinction between the paralogs; identity across the panel would justify treating the AP3S1 literature as transferable and would close the largest gap in this gene's record.
Hypothesis: Sigma-3B and sigma-3A confer indistinguishable dileucine-cargo specificity on AP-3.
Experiment: Generate single and double AP3S1/AP3S2 knockouts in a non-specialised human cell line and assay AP-3-dependent trafficking (LAMP1 and PI4KIIalpha distribution, delivery of a dileucine-tagged reporter to lysosomes) alongside AP-3 complex integrity by blue-native gel. Single knockouts that are silent and a double knockout that phenocopies AP3D1 loss would establish redundancy directly, which is currently assumed rather than shown.
Hypothesis: The two sigma-3 paralogs are individually dispensable but jointly essential for AP-3-dependent cargo delivery.
Experiment: Endogenously tag AP3S1 and AP3S2 with distinguishable epitopes in the same cell line and in primary neurons, then determine the beta-3 and mu-3 composition of each purified complex by quantitative mass spectrometry. This tests the combinatorial-assembly hypothesis directly, on endogenous protein, in the one cell type where the beta-3 isoforms are known to differ.
Hypothesis: Sigma-3B occupies AP-3 complexes containing either mu-3 paralog and either beta-3 paralog, rather than partitioning into a defined subset.
Experiment: Take the membrane proteins recovered with AP3S2 in affinity-purification interactomes, score their cytosolic tails for acidic dileucine motifs, and test the top candidates for AP-3-dependent localisation in AP3D1-null cells and for direct binding to a reconstituted delta-sigma3B hemicomplex. This would convert a list of high-throughput associations into the ubiquitous cargo set that the gene's expression pattern implies but that no study has defined.
Hypothesis: The lysosomal and endolysosomal membrane transporters that co-purify with AP3S2 in interactome datasets are genuine AP-3 cargoes selected through dileucine signals.
What is not known β curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: Nothing is known that distinguishes the sigma-3B form of AP-3 from the sigma-3A form. No cargo has been found that one selects and the other does not, no tissue or cell type is known to use one preferentially, and no phenotype has been assigned to loss of either paralog individually.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: What is established is that both paralogs exist, that both are incorporated into AP-3, and that both are expressed ubiquitously - measured by Northern and Western blot in 1997 and confirmed by a second group in the same year, with sigma-3 appearing as a two-isoform doublet in AP-3 immunoprecipitates. Combinatorial assembly would in principle give eight AP-3 variants from the beta-3, mu-3 and sigma-3 pairs, and the field has said in print that it does not know whether these variants occur in cells or differ functionally. Human AP3S2 also co-purifies with both mu-3 paralogs in high-throughput affinity-purification interactomes, so sigma-3B is not restricted to one mu-3 variant.
Significance: This is the question that decides whether AP3S2 deserves any annotation of its own beyond the complex's. Every process and location term it now carries is inherited from AP-3; if the paralogs are fully redundant, that inheritance is the whole truth about the gene, and if they are not, the annotations are silently attributing sigma-3A results to sigma-3B.
What would resolve it: Comparative proteomics of AP-3 cargo in cells expressing only one sigma-3 paralog at a time, paired with quantification of the two proteins (not transcripts) across tissues, would either name the discriminating cargo or establish that the difference is quantitative.
Provenance (the field's own admissions):
Gap: No structure, knockout, knockdown, patient allele or cargo screen has ever used sigma-3B. Every mechanistic statement about the AP-3 sigma subunit in the structural literature is an observation on sigma-3A that is assumed to transfer.
NARROWING BIOLOGYCURATION MF_DARK
What is known: The assumption is well supported but has been tested only in silico here: the two paralogs are 83.9% identical, and the two residues whose substitution abolishes dileucine-signal binding in sigma-3A are present unchanged at the same positions in sigma-3B, with only four differences anywhere in the 46-residue window around them. The single functional assay ever performed on sigma-3B - a yeast three-hybrid with the LIMP-II tail - gave the same result as sigma-3A. What is missing is any measurement on the sigma-3B protein itself: no affinity, no structure, no cellular phenotype.
Significance: The whole GO record for this gene rests on the transfer holding. A quantitative difference in cargo affinity between the paralogs, of the kind seen among the AP-1 sigma variants, would not be visible in any current annotation.
What would resolve it: Reconstitute delta-sigma3B alongside delta-sigma3A and measure binding to a panel of dileucine cargo tails; or solve the AP-3 core with sigma-3B substituted for sigma-3A, which is a single-subunit swap in an already-published expression system.
Provenance (the field's own admissions):
Gap: The reason AP3S2 appears in the literature at all - a type 2 diabetes association at 15q26 - has never been connected to AP-3 function, and the evidence available points away from AP3S2 being the effector gene.
OPEN BIOLOGY BP_DARK
What is known: The association itself is solid and has replicated in several populations. What is absent is any causal link to the protein: the lead variant rs2028299 is an eQTL for the neighbouring C15orf38 rather than for AP3S2, a second gene at the locus (PLIN1) is an equally plausible candidate on biological grounds, and the locus additionally produces a C15orf38-AP3S2 readthrough transcript, which UniProt lists among this entry's alternative products. The gene is pharmacologically dark and is not commonly required in CRISPR screens.
Significance: A locus name is repeatedly read as a gene function. If AP-3 really does act in insulin-granule or adipocyte trafficking it would be a genuine biological process annotation for this subunit; if the effector is C15orf38 or PLIN1, then none of the diabetes literature should ever inform this gene's curation.
What would resolve it: Fine-mapping with allele-specific expression in pancreatic islet and adipocyte models, plus a test of whether the readthrough transcript produces a protein, would assign the effector. A direct test of AP-3 requirement for insulin granule biogenesis would settle the functional half independently of the locus.
Provenance (the field's own admissions):
Gap: Whether AP-3's documented clathrin binding has any residual physiological role is unresolved, and GO currently encodes the superseded answer.
NARROWING BIOLOGYONTOLOGY BP_DARK
What is known: The 1998 in vitro result is not disputed: the beta-3 appendage binds the clathrin heavy chain N-terminal domain, and AP-3 colocalises with clathrin by microscopy. By 2013 this was the consensus reading. The 2026 reconstitution shows AP3:ARF1 deforming membranes into coated tubules with no clathrin present, and notes that AP-3 does not copurify with clathrin-coated vesicles. Both observations stand; what is unsettled is whether the clathrin interaction does anything in cells.
Significance: GO:0035654 asserts in its definition that AP-3 loads cargo into clathrin-coated vesicles, so every gene annotated to it - including this one - carries a mechanism the field has moved away from. The annotation cannot be fixed at the gene level alone.
What would resolve it: A clathrin-box mutation in beta-3, assayed for AP-3-dependent cargo delivery in cells, would decide the biology; on the ontology side, revising GO:0035654 or migrating its annotations to GO:0035459 would decouple the two questions.
Provenance (the field's own admissions):
Gap: It is not known which cargoes the ubiquitous, non-neuronal, non-pigment-cell AP-3 delivers, so the function of this ubiquitously expressed subunit in the tissues where it is most abundant is uncharacterised.
OPEN BIOLOGY BP_DARK
What is known: The characterised AP-3 cargoes come from specialised systems: ZnT3 and ClC-3 in neurons, tyrosinase and TYRP1 in melanocytes, LAMP1 and PI4KIIalpha in cell lines. Sigma-3B, by contrast, is present in all adult tissues examined, has low tissue specificity in Human Protein Atlas, and its protein is detected by proteomics. Human affinity-purification data place it in complexes with a list of lysosomal and endolysosomal membrane transporters, which are candidate cargoes rather than established ones.
Significance: All of this gene's process annotations are neuronal or pigment-cell specific and are graded non-core for that reason, which leaves the ubiquitous function - the one that matches the expression pattern - with no annotation at all.
What would resolve it: Quantitative proteomics of AP-3-coated carriers isolated from a non-specialised cell line, compared with AP3S1/AP3S2 double-depleted cells, would define the ubiquitous cargo set.
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)