Current synthesis: gamma-1 adaptin contributes cargo-recognition, coat and accessory contacts within the AP-1 complex. The initial pre-seed research below is retained as history; the later imported-seed and consulted-review sections supersede its no-seed/PENDING statements.
Written by codex on 2026-09-28 UTC, before Seed9 import. This is a bounded research notebook, not a provider-generated report or an annotation review. No seed, GOA row, canonical review YAML, or publication cache has been created or changed. No ordinary seed/reference fetch was attempted. The eventual source assertions and isoforms must come from the verified seed.
The previously checked Seed9 identity receipt is tmp/seed9-initial/alpk3-independent-primary-identity-review.json: AP1G1, HGNC:555, UniProtKB:O43747. The current official UniProt entry identifies human gamma-1 adaptin, an AP-1 heterotetramer subunit. AP1G1 associates with AP1B1, a mu subunit (AP1M1 or AP1M2), and a sigma subunit (AP1S1, AP1S2, or AP1S3). Gamma-1 is a peripheral coat component, not a cargo transmembrane receptor. Its trunk, hinge and appendage contribute different interactions. Preserve the distinction between AP-1 as a complex and AP1G1 as its gamma subunit. The current entry lists two isoforms; reconcile these with the actual seed before making any isoform-specific assertion. Official UniProt O43747
A plausible eventual central activity is contribution to clathrin-cargo adaptor activity in AP-1. Official GO:0035615 is a child of cargo adaptor activity, GO:0140312; its definition concerns bringing cargo and clathrin together. Do not manufacture a NEW ancestor alongside it, and do not assume that this term is absent before seeing the seed. Gamma/sigma cargo recognition and gamma-mediated coat/accessory recruitment provide mechanistic support, with the construct limits below. Official AmiGO GO:0035615
The gamma/sigma1 and alpha/sigma2 hemicomplexes of clathrin adaptors AP-1 and AP-2 harbor the dileucine recognition site. DOI:10.1091/mbc.e07-01-0012. Read the complete existing normal cached abstract and official indexed original Methods/Figure 1 caption. The Methods explicitly specify full-length human gamma, with mouse beta1, mu1 and sigma1; baculoviral expression used Sf9 cells. Gamma/sigma1 bound tested dileucine motifs strongly, unlike beta/mu alternatives. This establishes a functional cargo-binding hemicomplex containing human AP1G1; it does not establish autonomous gamma-only cargo recognition. The reported cellular internalization correlation concerns alpha/sigma2 and must not be reassigned to gamma/sigma1. Whole original Results, figure pixels and supplements remain unread. Existing cache is abstract-only and remains untouched. PubMed · Original PMC article, Methods and Figure 1
Gamma subunit of the AP-1 adaptor complex binds clathrin: implications for cooperative binding in coated vesicle assembly. DOI:10.1091/mbc.12.7.1925. Read the official abstract, indexed original opening Results and Figure 1/2 captions. Human and mouse sequences contain the compared hinge motifs, but the actual tested GST-gamma fragments in the Results are mouse, with rat liver cytosol and purified cytosolic clathrin. Purified-protein binding supports a direct interaction. Hinge and appendage cooperate in clathrin recruitment/polymerization; appendage alone binds but does not polymerize clathrin. This is useful ortholog mechanistic evidence, not a demonstrated human construct assay. Complete Methods/Results and figure pixels remain unread. PubMed · Original PMC article
Structural basis for recruitment and activation of the AP-1 clathrin adaptor complex by Arf1. DOI:10.1016/j.cell.2012.12.042. Read the complete existing cached abstract and indexed original opening structural Results. The reconstituted core combines murine gamma1 residues 1–595, human beta1 trunk, human sigma1C and murine mu1A; human Arf1 is engineered with Q71L and an N-terminal deletion. Recruitment and allosteric activation involve distinct surfaces. Gamma contacts Arf1, but the gamma N-terminal recruitment site is not itself the activating site. AP1G1 is not an Arf GTPase, GEF or GAP on this evidence. The source supports a complex-level recruitment/activation mechanism with explicit ortholog transfer; full biochemical Methods and all controls remain to be read. Existing abstract cache is preserved. PubMed · Original PMC article · Primary article PDF with construct paragraph
Structural basis for the accessory protein recruitment by the gamma-adaptin ear domain. DOI:10.1038/nsb808. Read the article abstract through its primary PDB deposition and the actual RCSB construct metadata. PDB 1IU1 is explicitly a human gamma1 ear-domain fragment, linked to O43747, expressed in E. coli, with a 1.8-Angstrom X-ray structure. The abstract reports structure-guided mutagenesis of accessory-protein recruitment involving gamma-synergin/Rabaptin-5. This corroborates an accessory-binding appendage function. The monomeric isolated fragment is not evidence that full AP-1 or intact AP1G1 is monomeric. Original binding Methods and all mutant controls have not yet been read. Original publisher · Primary structure deposition 1IU1
Gamma-adaptin appendage domain: structure and binding site for Eps15 and gamma-synergin. DOI:10.1016/S0969-2126(02)00801-8. Complete official abstract read. It reports structure-guided mapping of gamma-synergin and Eps15 binding to the same appendage site, with Golgi colocalization context. The construct species and detailed assays were not established by this abstract; do not call this direct human evidence yet. PubMed
De novo and bi-allelic variants in AP1G1 cause neurodevelopmental disorder with developmental delay, intellectual disability, and epilepsy. DOI:10.1016/j.ajhg.2021.05.007. Read the official indexed summary and targeted original passages about transient tagged expression, coexpression, localization and transferrin chase. Recessive variants alter recycling without an apparent defect in association with other AP-1 subunits. WT and mutant trafficking behavior must remain separate; disease-associated aggregation is not normal localization. Human variant nomenclature uses an alternative transcript numbering, so reconcile residues with O43747 isoforms. The paper contains a four-transmembrane-domain prediction inconsistent with the established soluble adaptin architecture; do not propagate it uncritically. Complete supplementary construct provenance and Methods remain unread, so the HEK293T/COS7 host alone does not establish every reagent species. Zebrafish rescue is organismal necessity evidence, not a new direct developmental molecular function. Original PMC article · Author-institution primary article PDF
The aftiphilin/p200/gamma-synergin complex. DOI:10.1091/mbc.e04-12-1077. Complete official abstract and Figure 9/10 captions read. AP-1 depletion slows transferrin recycling, whereas aftiphilin depletion accelerates it; both perturbations affect selected other cargo. Figure 10 identifies the AP-1 knockdown as mu1, not gamma1-specific depletion. The accessory complex interacts with AP-1 but is not a replacement name for the core AP-1 heterotetramer. This source provides context for AP1G1-associated trafficking, not a license to assign every accessory phenotype directly to gamma1. Original Methods and target-specific binding experiments remain to be read. PubMed
These notes establish research priorities, not annotation actions. No whole-paper completion claim is made: PMC direct opens often returned a browser challenge, and the Molecular Biology of the Cell publisher returned HTTP 403. Targeted original indexed passages are distinguished from abstract-only and deposit metadata above. No figure pixels, supplemental tables, raw diffraction data or PAINT phylogeny were inspected. No negative annotation judgment follows from these access limits.
Before curation: inspect the actual seed and all its references; preserve source tuples, NOT assertions and products; check existing core/annotation coverage; consult an independent annotation reviewer before decisions. Locate actual PAINT node evidence for any IBA challenge. Test precise compartment, cargo, and motor-binding assertions against their own sources. Cargo recognition, clathrin assembly, Arf recruitment and accessory recruitment can support one coherent adaptor core without proliferating overlapping NEW terms. No NEW biological process is proposed here.
Existing normal caches found and read: PMID17360967 and PMID23415225, both full_text_available: false. The other five priority records above were absent locally at this read. This is an observation only, not a missing-source request or a completed retrieval attempt. Wait for Seed9 auxiliary inventory before deciding whether a finite normal recovery batch is necessary.
The preceding section is preserved as pre-seed research. Seed9 has now imported the exact normal AP1G1 seed (82 source annotations, two products O43747-1 and O43747-2); source objects remain PENDING pending independent annotation consultation. Fresh main/alias/history/branch and complete open-PR preflight passed (tmp/AP1G1-initial/preflight-verification.json). Root imported the four previously absent normal caches as part of the 70-record Seed9 auxiliary import. All 35 required normal PMID/Reactome caches are now present. Fifteen existing canonical records differ from staged candidate bytes and are preserved; staged fuller bodies are a separate read source, never replacements.
The proper installed-client Falcon attempt with perplexity-lite fallback terminated in 4.789 seconds with exit 1 and a connection error, producing no provider report. The actual receipt is tmp/AP1G1-initial/deep-research-receipt.json. Raw debug output was not retained because such output may include credential-bearing local variables; only its digest and controlled error indicators were saved. These are manual notes.
All 19 seeded PMID abstracts and 16 Reactome summaries have been read. This does not establish that the original experiment underlying each row has been read. Initial molecular-function scope: cargo-binding assays for tyrosinase, OCA2 and PSEN2 use an AP-1 complex or gamma/sigma hemicomplex; do not describe free gamma as recognizing all cargo autonomously. PMID16162817 actual canonical Methods and Results establish mouse tyrosinase cytoplasmic tails and gamma1/sigma1A three-hybrid binding, with conserved dileucine controls; adaptor constructs are inherited from earlier methods, so their species are not reassigned from the assay host. PMID19841138 actual Results show reciprocal gamma-adaptin/KIF13A co-immunoprecipitation in human MNT-1 cells and epitope-blocking control; this establishes association, not the binary interface. That paper's AP-1 depletion targets mu1A.
Separately read the preserved Seed9 candidate full body for PMID19116314: human OCA2 N-terminal GST constructs, MNT-1/HeLa lysate pull-downs, gamma detection and dileucine controls. This supports cargo recognition by the AP-1 complex, not a purified gamma-only interaction. The canonical cache remains abstract-only. Separately read targeted Methods/Results/Discussion in the fuller Seed9 PMID22511774 candidate: human MNT-1 gamma-subunit knockdown; endogenous Rab32/Rab38 co-IP; GDP/GTPgammaS-loaded GST-Rab pull-downs from cytosol. AP-1 prefers the GTP analog less strongly than AP-3/BLOC-2. The authors explicitly do not exclude a bridging protein. Its gamma knockdown alters Tyrp1 recycling/melanin and is actual target evidence, not a phenotype inferred solely from the two Rabs. The canonical cache remains abstract-only.
The rat collagen-binding electronic donor is identified by official UniProt/NCBI as Ap1g1 (A0A8I5Y697, RGD619785), but the exact primary annotation chain has not been recovered. RGD and QuickGO endpoint attempts were inaccessible. The compartment difference alone is not proof that the donor experiment is wrong; this source remains unresolved.
Three additional normal publications were requested in one finite 45-second batch: PMID11451993 (mouse gamma clathrin-binding mechanics), PMID12042876 (human gamma ear structure/accessory recruitment), and PMID34102099 (AP1G1 variants and recycling). The batch timed out with no output cache; no claim is made that all three IDs completed individual attempts. No canonical cache was written and no repeated attempt is planned.
The indexed original PMID:23632890(https://pmc.ncbi.nlm.nih.gov/articles/PMC3674668/) Methods, adaptin-gamma Results, Figure 7 caption and related Discussion were inspected. The direct PMC open returns a browser challenge, but the indexed original body exposes these passages. Human YTS cells receive adaptin-gamma RNA interference; Figure 7 measures perforin retention in CI-MPR-positive transport vesicles and reduced target-cell lysis. This establishes gamma-subunit perturbation, rather than merely assigning every LAMP1 result to AP1G1. The direct LAMP1/AP-1 interaction is attributed to earlier work. Supplementary methods, exact RNAi sequence and all image pixels remain uninspected. The distinct degranulation assertion therefore retains its curator provenance without claiming that Figure 7 is a direct gamma-dependent release assay. The normal abstract cache remains unchanged.
The three added AP1G1 normal records are part of the fixed nine-source recovery batch; publication of that retrieval workflow does not establish successful source recovery. The PMID:11451993 construct scope must distinguish rat liver cytosol from the separately purified bovine-brain clathrin preparation. Independent identity review verified that distinction; a common mammalian source is not assumed.
The independent all-82 consultation accepted the refined v2 map: 55 ACCEPT, 18 KEEP_AS_NON_CORE and nine UNDECIDED, with no NEW annotation. Two process assertions were made explicitly uncertain after challenge: the gamma-specific NK degranulation experiment is not exposed by the inspected Figure 7, and the platelet review paragraph does not establish gamma1 participation. Cytotoxicity and cargo transport remain distinct, positively supported outcomes. Three lysosomal-membrane assertions retain their uncertainty because lysosome-destined coated carriers do not by themselves establish residence on mature lysosomal membrane. The known cargo-adaptor activity is retained as a contribution of gamma in the assembled AP-1 complex.
All three exact Source36 normal records were inspected before parent import. PMID:11451993, PMID:12042876 and PMID:34102099 are abstract-only; complete abstracts, source identities and retrieval logs were read, and eligible bytes equal the raw normal-output copies. Previous indexed original-body/PDB observations remain separately bounded above. In particular, mouse gamma fragments, rat liver cytosol and bovine-brain purified clathrin in the 2001 study are not merged into a human purified assay. Human gamma in the 2007 cargo hemicomplex and human ear-domain structure provide separate target evidence. No cache was overwritten or additional normal fetch attempted.
The current official HPA AP1G1 summary and assay table states: "Mainly localized to the Golgi apparatus and vesicles. In addition localized to the cytosol." Both main locations and the cytosolic pool carry enhanced reliability, with siRNA validation reported. HPA041224 and HPA070312 cover human cell lines including A-431, MCF-7, U2OS, HAP1 and SiHa; not every line has every pool. The table was read; no image pixels were rescored.
Two richer immutable Seed9 candidate bodies supply exact supporting passages absent from the preserved canonical abstracts. In the PMID:19116314 original study, human OCA2-tail material was incubated with cell extracts: "GST-OCA2 NT bound the adaptor proteins AP-1, AP-2, and AP-3". This does not establish purified gamma-only cargo binding. In PMID:22511774(https://pmc.ncbi.nlm.nih.gov/articles/PMC3365991/), "AP-1 also showed a preference for the GTPγS form of Rab32 and Rab38 but less dramatically than BLOC-2 or AP-3". The authors do not exclude an intervening bridging protein; gamma knockdown and the individual Rab perturbations remain different experiments. These exact short passages were checked in the separately preserved candidate records, not inserted into a normal cache.
The current official clathrin-coated-pit definition includes TGN/endosomal budding structures, so the broad coat-pit term is not restricted to AP-2 plasma-membrane uptake. No exact PAINT node placement or donor experiment was invented. All original source objects, ordinary qualifiers and both products remain unchanged; no isoform-specific function is inferred from the mere presence of two products. The unresolved collagen and neuronal donor chains remain UNDECIDED rather than being rejected from compartment or donor-count arguments.
All 82 source assertions have a reviewed decision. Independent prospective and final authored consultations passed; all three recovered normal references are closed at their actual abstract-only scope. Initial full gene validation passed with eight deliberate generic-binding NONCORE advisories: the verified cargo/subunit associations are retained under the explicit experimental-removal criteria, without inventing isolated interfaces or mapping every association to autonomous adaptor activity. Final validation, rendered output and the scaffolded codex history are checked and recorded with the frozen handoff. Original source objects, both products, raw seed files and all 40 normal source caches remain unchanged.
The first review retained cargo and assembly associations conservatively. This follow-up uses each cargo paper's own inspected experiments to distinguish the four cargo-recognition contributions from three still-uninspected sigma1B screen records. It supersedes the earlier generic cargo/subunit dispositions only as described here. All 82 original source assertions, ordinary qualifiers, two products and published history remain unchanged.
The original CLN3 study, PMID:15598649 was read at its recombinant-construct and pull-down Methods, Figure 6 Results, trafficking Results and Discussion. The DOI in the article header is 10.1074/jbc.M411862200. Human CLN3 cytoplasmic-loop GST material captures native gamma-containing AP-1 from HeLa cytosol; L253A/I254A abolishes capture. Mouse liver and mu1A-deficient mouse fibroblasts supply separate binding/sorting contexts. The paper does not use a reconstituted gamma/sigma hemicomplex. Methods and caption use different loop endpoints, so no unnecessary residue-range claim is added. Figure pixels and unexposed supplements were not reanalyzed. The abstract-only normal cache remains untouched.
The other three own-source cargo experiments were rechecked: PMID:16162817(https://pubmed.ncbi.nlm.nih.gov/16162817/) uses tyrosinase sorting-motif mutants and gamma1/sigma1A Y3H, alongside native-complex and coated-bud localization; PMID:19116314(https://pmc.ncbi.nlm.nih.gov/articles/PMC2649270/) captures native AP-1 with human OCA2 tails in HeLa/MNT-1 extracts; PMID:27293189(https://pubmed.ncbi.nlm.nih.gov/27293189/) combines PSEN2 motif-dependent gamma1/sigma1 binding, rat-brain AP-1 capture and cargo targeting. None is described as autonomous gamma-only cargo-to-clathrin bridging. The four refinements express contribution to the assembled clathrin-cargo adaptor activity; original source qualifiers are preserved and the core retains its explicit contributes-to relationship. Rabaptin-5 remains a supported secondary partner interaction, not GTPase activity or a newly inferred cargo bridge.
The three AP1G1/AP1S2 screen records remain UNDECIDED because their exact pair, variant and construct data were not inspected. The earlier PMID:9733768 support overread a gamma2/sigma1B statement and is removed from those rows. Its reference assessment is corrected without claiming that the paper or gamma1 interaction is biologically false. The independently read PMID:34102099 main coimmunoprecipitation Results name sigma1 generically; this is not a substitute for the specific original screen records. No generic-only removal is made.
Cytoplasm, coat-vesicle and recycling-endosome rationales now distinguish their original source types. The PMID:15758025 normal cache is abstract-only, so the previously checked positive human HPA cytosol passage remains the independent supporting attachment; an available record is not equivalent to an available target-specific localization experiment. Human MNT-1 recycling-endosome IDA is no longer described with an irrelevant unresolved mouse-transfer caveat. The core process pointer is narrowed to the already accepted intracellular protein transport term. No new annotation, process, source cache or isoform claim is created.
Full follow-up validation passes. One generic-binding NONCORE advisory remains for the directly supported Rabaptin-5 association; the explicit experimental-removal criteria take precedence over removal based only on genericity. Final source preservation, rendered output and independent authored consultation are recorded in the new follow-up history and handoff. Earlier completion counts above describe the initial published review rather than this revision.