Human DEPDC5 is the large GATOR1 subunit. The biochemical distinction is between binding and organizing Rag GTPases, which DEPDC5 performs, and the catalytic arginine finger supplied by NPRL2. Human GATOR1 structures and reconstitution experiments support one core: small GTPase binding, contribution to complex GAP activity, and negative regulation of TORC1 signaling at the lysosomal surface. DEPDC5 alone was inactive in the tested GAP assay. This does not imply that DEPDC5 is dispensable for complex function (PMID:23723238(https://pubmed.ncbi.nlm.nih.gov/23723238/), PMID:29590090(https://pubmed.ncbi.nlm.nih.gov/29590090/), PMID:35338845(https://pubmed.ncbi.nlm.nih.gov/35338845/)).
KICSTOR recruits GATOR1 to lysosomes. Several recruitment experiments image NPRL2 as a complex marker, while the earlier study images tagged DEPDC5 itself. Native human placental proteomics separately identifies DEPDC5 in enriched lysosomal membranes. These observations support membrane association; they do not establish membrane-spanning topology. Mouse neuronal cytosolic/perinuclear staining and corroborating human neural-culture experiments identify another observed pool, with the localization study’s stated preliminary limitation (PMID:28199306(https://pubmed.ncbi.nlm.nih.gov/28199306/), PMID:17897319(https://pubmed.ncbi.nlm.nih.gov/17897319/), PMID:23542697(https://pubmed.ncbi.nlm.nih.gov/23542697/)).
The eight HuRI annotations all refer to the 559-residue O75140-2 product. Each exact human pair has validated two-hybrid, array and pooling records, with sequence checksums matching the current target and partner. The main study describes quadruplicate retesting and sequence confirmation; no pair-specific orthogonal-validation result is claimed. Binding to USP2, CCT3 and the identified MEOX2 clone supports finer partner-class terms. The other five positive associations remain non-core. None establishes a native cellular complex, full-length DEPDC5 activity, or a transcription, repair or phosphatase mechanism for DEPDC5 (PMID:32296183(https://pubmed.ncbi.nlm.nih.gov/32296183/)).
The exact MEOX2 partner is Q6FHY5, an unreviewed 303-residue sequence differing from reviewed MEOX2 by one glutamine deletion outside the homeodomain. Human MEOX2 promoter-binding and reporter assays establish its DNA-binding transcription-factor class. The original distinguishes DNA-binding-dependent p16/CDKN2A activation from DNA-binding-independent p21/CDKN1A activation; the class refinement does not assert identical activity of the HuRI clone (PMID:22206000(https://pubmed.ncbi.nlm.nih.gov/22206000/)).
PAINT IBD nodes PTN000323637 and PTN000995343 both occur on the human lineage. Actual donor records and selected yeast originals support complex membership, TORC1 inhibition and downstream autophagy regulation. Human target evidence legitimately appears among the donors; it is not circular. Fission-yeast nutrient-response details and historical Iml1 catalytic interpretations are not automatically transferred to the human subunit (PMID:21900499(https://pubmed.ncbi.nlm.nih.gov/21900499/), PMID:25046117(https://pubmed.ncbi.nlm.nih.gov/25046117/), PMID:29199950(https://pubmed.ncbi.nlm.nih.gov/29199950/), PMID:33534698(https://pubmed.ncbi.nlm.nih.gov/33534698/)).
KLHL22-dependent DEPDC5 degradation remains disputed. The 2018 public extended-data legends report tagged DEPDC5/CUL3/RBX1 association, ubiquitylation and DEPDC5-dependent mTORC1 epistasis. The 2025 challenge and reply disagree about antibody bands, endogenous tags and nutrient protocols. Their complete public Figure 1 legends were inspected, while their main bodies were unavailable. The qualified colocalizes_with association is retained as non-core; the independently supported negative TORC1 process is retained without treating degradation as settled. No retraction is asserted (PMID:29769719(https://pubmed.ncbi.nlm.nih.gov/29769719/), PMID:39780017(https://pubmed.ncbi.nlm.nih.gov/39780017/), PMID:39780003(https://pubmed.ncbi.nlm.nih.gov/39780003/)).
All eight official products retain their source identifiers, names and sequence notes. Sequence differences were reconstructed from UniProt features and matched against all-isoform FASTA. Their lengths are 1603, 559, 1581, 1552, 1229, 1503, 1594 and 1572 residues. Functional equivalence is not inferred from preservation of a domain, and engineered truncations are not relabeled as natural products.
The 44 source rows are preserved: 26 ACCEPT, 7 MODIFY and 11 KEEP_AS_NON_CORE; no NEW assertions. Broad supported terms are refined with MODIFY even when the specific term already occurs elsewhere. Positive generic interactions without a justified finer function follow the standing project curation instructions. That instruction is a project/user policy, not a claim of independent human sign-off.
Normal source retrieval and publication caching completed. Falcon returned an authorization failure and the configured fallback returned an insufficient-quota error; neither produced a research file. The manual investigation is recorded here and in the source guide. Existing machine caches and separately recovered originals retain their distinct access states.
The single Rag-binding/GATOR1 core now explicitly includes cellular response to amino acid starvation (GO:0034198). Human depletion, reconstitution and interface-mutant experiments connect DEPDC5 assembly and Rag engagement to mTORC1 inhibition after amino acid withdrawal (PMID:23723238(https://pmc.ncbi.nlm.nih.gov/articles/PMC3728654/), PMID:29590090(https://pubmed.ncbi.nlm.nih.gov/29590090/), PMID:35338845(https://pubmed.ncbi.nlm.nih.gov/35338845/)). The broader protein-containing complex binding annotation remains supported as non-core association context. It does not require a second DEPDC5-enabled activity alongside the directly resolved Rag contacts. In particular, GATOR1–GATOR2 co-IP does not establish direct DEPDC5 contact with every component; the 2018 subunit experiment identifies NPRL3 as necessary and sufficient for GATOR2 association.
For the imported enabled GAP annotation, the preferred GOA correction is GO:0005096 with contributes_to. The imported enables qualifier remains unchanged. Small GTPase binding is the supported enabled-function replacement, while the core separately records the contribution to complex GAP activity. The UniProt FUNCTION record states “it does not execute the GAP activity”; the original human studies identify NPRL2 as the catalytic subunit. The short exact publication anchors now attached to the YAML distinguish the catalytic residue, DEPDC5–RagC engagement, starvation response and lysosomal recruitment. Their original sources and assay limits are listed in the source guide.
The broad cytoplasm refinement includes both cytosol and lysosomal membrane. Neuronal immunostaining and human neural cultures support the cytosolic pool; human DEPDC5 imaging and GATOR1 recruitment studies support the lysosomal pool. The neural localization evidence remains preliminary, and some KICSTOR experiments image NPRL2 as a complex marker. These specific locations were already represented in the imported annotations; no new location assertion is added.
The cached GO-CAM 646ff70100001312, activity 646ff70100001542, and 646ff70100001561, activity 646ff70100001735, both place DEPDC5 small GTPase binding at the lysosomal membrane within negative TORC1 regulation and an enclosing starvation-response process. In the latter model, DEPDC5 is also an input to the KLHL22-enabled activity. These curated roles corroborate the functional distinction; they do not assign ubiquitination activity to DEPDC5 or settle the later KLHL22/degradation dispute.
SAMTOR is a regulatory input with a direction that must be distinguished from the UniProt SUBUNIT wording. The 2018 GATOR1 original demonstrates association. The dedicated SAMTOR original, Figures 1, 3 and 4 shows that methionine starvation increases SAMTOR–GATOR1/KICSTOR association and that methionine or SAM restoration reduces it. SAM binds SAMTOR; direct methionine sensing is not assigned to DEPDC5 (PMID:29123071(https://pubmed.ncbi.nlm.nih.gov/29123071/)). Separately, the PMID:31548394 abstract and UniProt PTM record describe AKT1/PIM1 phosphorylation that relieves DEPDC5-mediated mTORC1 inhibition in the tested cancer-cell models. That context is based on the abstract and curated record here, not an independently inspected original or evidence for every natural product.
The publisher-indexed lysosomal proteomics evidence is the original main Table 1 in PMID:17897319(https://doi.org/10.1111/j.1600-0854.2007.00643.x), read separately from its abstract-only normal cache. It identifies native human placental DEPDC5 with 20 peptides, 18.4% coverage and enrichment ratio 4.13. The historical product designation does not establish unique current-isoform peptides. The source guide now provides public article, figure, table and model locators.
All 44 source objects and eight official products remain unchanged. The current decisions are 25 ACCEPT, 7 MODIFY and 12 KEEP_AS_NON_CORE; the single action change is the broad complex-association annotation. There are no NEW annotations. Earlier entries above remain as the record of the preceding assessment.