C9orf72 is a native protein regulator of small-GTPase-dependent membrane traffic. Its long product forms the C9orf72–SMCR8–WDR41 complex. C9orf72 also has experimentally reported Rab39A exchange activity outside that complex and acts as an active-Rab1A effector during autophagy-initiation machinery recruitment. These are distinct biochemical operations, not interchangeable uses of the legacy GEF name. Pathogenic repeat RNA and translated dipeptide-repeat species are separate from the two source-record C9orf72 protein products.
The normal seed contains 110 logical GO assertions from 114 raw GOA records. The difference comprises four presynaptic-cytosol records with different ECO codes that map to one logical source object, and a duplicate complex-GEF assertion differing in date/assigning organization. All raw records remain unchanged. Every logical machine object, original PMID, qualifier and partner accession is preserved. No NOT or product-specific annotation flag is introduced. Product names are preserved literally: Q96LT7-1 is 1 (C9-L, C9(LF)); Q96LT7-2 is 2 (C9-S, C9(SF)). A partner suffix such as Q86UV6-2 belongs to TRIM74, not C9orf72. Family exports remain quarantined.
The original PMID:28195531 seed reference carried TODO: Fetch title because the latest normal attempt failed. A complete older normal cache already existed. Its correct title is used without replacing or rewriting that cache. The other source titles remain unchanged.
PMID:27103069(https://pubmed.ncbi.nlm.nih.gov/27103069/) reports complex-level GDP release on Rab8A and Rab39B from human constructs purified using insect cells. C9orf72 alone had little effect on those substrates in that experiment. Rab39B-QL rescue supports autophagy regulation, but does not uniquely identify catalytic direction. PMID:27617292(https://pubmed.ncbi.nlm.nih.gov/27617292/) also reports exchange with complex preparations. Those are genuine positive experiments. They cannot be dismissed from a later title or from the observation that SMCR8 supplies a catalytic residue in a different assay.
PMID:37821429(https://pubmed.ncbi.nlm.nih.gov/37821429/) directly tests bacterially expressed C9orf72 and Rab39A, including purified binding, Mant-GDP release, titration, cellular loss/rescue and HOPS recruitment. The exchange assays used near-equimolar concentrations and long observation windows, approximately 7,200–14,000 seconds. The supplied C9orf72/SMCR8 preparation behaved differently. This supports a bounded C9orf72-alone Rab39A mechanism, not automatic replication of holo-complex Rab8A/Rab39B exchange or ARF-GAP experiments. HOPS performs tethering and SNAREs execute membrane fusion; those activities are not transferred to C9orf72.
PMID:27334615(https://pubmed.ncbi.nlm.nih.gov/27334615/) supports a separate recruitment mechanism: C9orf72 preferentially binds active Rab1A, interacts with ULK1-complex components, and is required for recruitment of that machinery in human cell lines and rat neurons with human rescue. This is Rab1A-effector behavior, not evidence for Rab1A exchange. Loss of C9orf72 did not prevent the measured ULK1 activation response. The synthesis uses the existing specific small-GTPase-binding and assembly-regulation terms; it does not manufacture another adaptor or kinase annotation merely to label the mechanism.
PMID:27559131(https://pubmed.ncbi.nlm.nih.gov/27559131/) shows nutrient-dependent lysosomal recruitment of endogenously tagged C9orf72. Its mTORC1 phenotypes differ between C9orf72 and SMCR8 single knockouts and the double knockout. The broad signaling-regulation assertion is retained with those context limits, without inventing a uniform activating or inhibitory role, Rag-GAP activity or intrinsic mTOR kinase activity.
PMID:32848248(https://pubmed.ncbi.nlm.nih.gov/32848248/) and PMID:34145292(https://pubmed.ncbi.nlm.nih.gov/34145292/) are now normal canonical full-text caches. Their unique Results/Discussion and selected mechanistic Methods were read. Human HEK293-purified complex promotes ARF1/5/6 hydrolysis; SMCR8 supplies Arg147 and C9orf72 forms part of the substrate pocket. C9orf72 interface mutations impair GAP activity. The 2021 structure uses an ARF1–SMCR8 fusion and GDP–BeF3 analog, while separate biochemical assays support the function. ARF1 lacks its N-terminal membrane anchor. Rab8A required substantially more complex than ARF1, and physiological local substrate pools remain unresolved. Lysosomal recruitment does not prove ARF hydrolysis occurs in cis at the lysosomal membrane.
The 2020 negative exchange experiment tested Rab1A/5A/7A/8A/39B with a positive Rabex5 control over 1,400 seconds, not Rab39A. The older positive complex assays and the later C9orf72-alone Rab39A assay therefore require preparation- and substrate-specific reconciliation. The three contributes_to exchange rows and GEF-complex classification remain UNDECIDED; two enables exchange rows retain the bounded Rab39A finding. Finding-level DISPUTED records point to the actual negative experiment without declaring the original papers invalid.
PMID:30075745(https://pubmed.ncbi.nlm.nih.gov/30075745/) provides mouse knockout validation of the new 1C1/12E7 antibodies, fractionation and histology, with distinct human iPSC-derived motor-neuron observations. A soluble presynaptic cytosolic pool is supported; the mouse LS2 signal is not a vesicle-lumen or integral-membrane assignment. Human iPSC experiments are not human knockout controls. Low-abundance short products remain possible.
The same study tests commercial sc-138763, used in PMID:24549040(https://pubmed.ncbi.nlm.nih.gov/24549040/). Its specificity concern materially affects old endogenous nuclear, secreted and fine vesicular assignments. The three endosome assertions remain unresolved: independently established lysosomal/autolysosomal pools and trafficking functions do not by themselves identify a separate endosomal pool. It also affects the hnRNPA1/A2B1 co-IP evidence: the source's reverse-IP prose and figure-caption direction differ, but both depend on the C9orf72 reagent for capture or detection. Tagged C9orf72 induction of stress granules is not an independent tagged-protein hnRNP-binding validation. These two binding assertions remain undecided because of that concrete concern, not because an extra native or orthogonal assay is demanded for every curated interaction.
The 2017 correction, DOI:10.1093/hmg/ddx309, replaces Fig. 5E internalized-TrkB imagery and restores the Fig. 7C legend; the corrected hnRNP co-IP is Fig. 7F. Authors state the correct blot was used for quantification and conclusions were unchanged. Correction prose was inspected through the official index and the institutional publisher-version PDF. Corrected images were not independently reanalyzed. The notice neither retracts the paper nor validates its antibody.
PMID:26174152(https://pubmed.ncbi.nlm.nih.gov/26174152/) used custom isoform-specific antibodies. Its complete abstract supports the authors' long/short-product distinction, but original specificity controls and images were unavailable. The later commercial-antibody tests do not establish that these custom reagents were tested. The fine short-product nuclear-membrane pool is therefore unresolved rather than disproved by mouse long-product results.
PAINT assertions are phylogenetic curator judgments, not pairwise transfers. A target accession in its own descendant evidence is legitimate; no circularity or donor-count objection is made. The actual PAINT trees were not independently re-curated.
Mouse Q6DFW0 evidence is kept separate from direct human experiments. The official mouse UniProt record, PMID:35876881 primary abstract, and indexed PMID:31651360 primary passage support contextual synaptic claims. These donor sources are web-reading evidence, not newly installed normal caches or full-paper reading. They are not added as unsupported YAML citations. Fine late-endosome-to-lysosome transport remains unresolved because broad autophagic fusion does not establish that exact route.
The accessible PMID:27723745 abstract supports an ARF6/RAC1–LIMK/cofilin pathway and axonal actin effects, not C9orf72 kinase or phosphatase activity. Exact growth-cone/main-axon images were not inspected. PMID:27037575(https://pubmed.ncbi.nlm.nih.gov/27037575/) supports native long-product stress-granule assembly and stress-dependent localization. Repeat-associated RNA effects are described separately rather than assigned to the normal protein.
Supported generic binding is retained as non-core under the standing published project instruction. Source-specific screen frameworks and exact UniProt edges support curator deference without claiming target supplement cells or clone images were inspected. CRX/c-Rel, HSC70 and Rab39 partners permit five more informative binding refinements. Complex coassociation with WDR41 does not mean a direct C9orf72–WDR41 interface.
Canonical PMID:27103069 is abstract-only; selected detailed complex, HSC70 and exchange passages came from the authenticated fuller normal-fetch variant kept in TMP. Fuller variants of the screen sources supplied selected Methods to the binding consultation. No existing cache was overwritten, and every YAML quotation comes from the normal canonical cache. Publication availability flags are distinct from actual reading coverage.
One ordinary Falcon attempt with the prescribed perplexity-lite fallback failed DNS and created no provider output. Research was completed manually from the primary records and explicit consultations; no provider-branded manual report was created. All 110 source assertions have been assessed: 35 ACCEPT, 47 KEEP_AS_NON_CORE, five MODIFY and 23 UNDECIDED, with no NEW assertion. The three core units describe exchange, complex GAP participation and active-Rab1A recruitment separately. DRAFT preserves the supported generic-binding advisories and unresolved evidence. Normal candidate validation and independent whole-science approval are recorded separately before any canonical application.
Full normal schema, references, GOA, best-practice and ontology-term validation passed on the completed candidate. The 21 advisories comprise 18 supported generic-binding policy warnings, two unresolved IBA assertions without a fabricated structured phylogeny assessment, and one same-term action-consistency warning. The latter reflects preserved contributes_to complex-GEF assertions with conflicting assays versus enables Rab39A evidence, not an unexamined evidence-code preference. DRAFT remains appropriate. An initial validation subprocess failed at the unavailable default uv cache; the unchanged candidate passed using the established writable offline/no-sync environment. Independent whole-science peer remains required before application.
The exact candidate was applied after independent ROOT whole-science approval and preimage checks. Full normal candidate and canonical validation, status reporting, targeted rendering and standard codex/gpt-6 CREATE history validation passed. DRAFT preserves 21 disclosed advisories and 23 unresolved assertions. All 110 machine source objects, both named products, normal caches and earlier histories remain unchanged. The source-specific mechanism and antibody limits above remain in force. This local application record does not claim remote publication.