This preliminary whole review preserves the 23 normal GOA assertions, including their qualifiers and source tokens, and the two UniProt products O95389-1 and O95389-2. The reviewed human entry is O95389, HGNC:12771, NCBI Gene 8838, 354 residues. The second product retains the fetched VSP_037803 note; no construct is assigned to it by inference. The frozen ClinGen association is autosomal recessive progressive pseudorheumatoid arthropathy of childhood, Definitive, dated 2024-09-04. That clinical classification supplies disease context, not an additional molecular function or developmental-process assertion.
ROOT imported the normal sources without changing their bytes. The ordinary deep-research attempt failed offline because the locked deep-research-client[cyberian] dependency could not be resolved; the actual failure is recorded at tmp/CCN6-artifact-transport/root-deep-research-tool-result.json. No provider report was generated. These are manual research notes. The preliminary candidate is not a final validated or independently approved review: four normal supplemental publication caches are still pending. Their titles, access metadata and short anchors will be reconciled to actual imported bytes in a separate final snapshot.
The standing ClinGen generic protein-binding policy is recorded in projects/CLINGEN_MENDELIAN.md under Curation instructions. None of the 23 original rows is a generic GO:0005515 assertion. This policy does not turn unverified integrin/heparin specificity into a verified target assay.
WISP3–IGF1 interaction, PMID:23424195 supplies the strongest own-activity evidence. Selected original Results, captions and Methods were read on the publisher page, including human WISP3 from a synovial-fibroblast cDNA, human IGF1, C-28/I2 chondrocytes, transfected HEK293 cells, recombinant coimmunoprecipitation and surface plasmon resonance. The cell line is derived from human juvenile costal cartilage and immortalized with SV40; it is not primary adult articular cartilage. Mouse ATDC5 positive-control material is separate. SPR used micromolar IGF1 concentrations and reported approximately 30 μM affinity for immobilized WISP3; this must not be described as the high affinity of canonical IGFBPs. The full multidomain protein is important, and the IGFBP-like domain alone is insufficient. Poor expression of a deletion construct cannot be interpreted as a clean binding-negative mutant. The cellular experiments link WISP3 to IGF1 secretion and IGF1-induced hypertrophic markers, without showing proteolytic cleavage of IGF1 or an intrinsic enzyme activity.
The original Wisp3 BMP/Wnt study, PMID:17823661 provides an independent ligand context. Selected original Methods, Results and figure captions distinguish human WISP3-containing conditioned medium inhibiting human BMP2-driven alkaline-phosphatase responses in mouse 10T1/2 cells from human WISP3/mouse BMP4 coassociation. Reciprocal immunoprecipitation, conditioned-medium depletion and disease-variant comparisons are informative. Mixing conditioned media is not a purified two-protein binding assay. The Wnt receptor binding and TopFlash experiments principally use zebrafish Wisp3, even when carried out in mammalian cells. They are not relabeled as direct human CCN6–LRP6 or human CCN6–Frizzled measurements. The zebrafish cartilage phenotype and the absence of an obvious mouse phenotype remain distinct from human recessive disease.
The single provisional NEW is GO:0019838 growth factor binding. Its official definition and protein-binding parent were checked in AmiGO. It describes CCN6's own interaction with growth-factor ligands, not merely a response to added growth factor. The existing integrin- and heparin-binding terms do not cover IGF1/BMP ligand binding. Only this broad activity is proposed; its IGF-binding child is not added redundantly. Dated local IGFBP3 annotations provide positive IGF-binding/growth-factor-binding comparator coverage, and the official MGI Nog record provides positive predicted growth-factor-binding coverage for a separate extracellular antagonist. These are explicit, limited positive precedents, not a complete current QuickGO survey or proof that curators missed a gene. The inspected local GO-CAM index has no O95389/CCN6/WISP3 match; that literal result is not proof of global model absence. No NEW biological process is proposed.
The core uses extracellular growth-factor binding to summarize modulation of signaling and communication. It does not assign receptor kinase activity, BMP-receptor binding, a universal Wnt response, or a structural extracellular-matrix activity. Broad inherited integrin/heparin binding and cell-adhesion assertions remain unresolved where the original PAINT node evidence and target specificity were not established. Integrin-blocking cellular responses found in selected external target papers are useful context, but are not represented as purified interface measurements. No donor-count, target-self-inclusion or wrong-paralog argument is used to reject PAINT.
The canonical 2016 CCN6 study, PMID:27252383 is abstract-only. Selected original Results and Methods were read separately from an author-uploaded primary reproduction. The available evidence includes endogenous and expressed-protein microscopy, fractionation, mitoplast protease protection and siRNA depletion in human C-28/I2 cells. Carbonate extraction of membrane-associated protein used CCN6-Myc-transfected HEK293 cells, not the chondrocyte line. Isotype/empty-vector and compartment-marker controls were considered; figure pixels and all supplements were not independently analyzed. These experiments support a mitochondrial pool without uniquely fixing its membrane topology.
In the chondrocyte experiment, partial CCN6 depletion increases respiration, mitochondrial ROS, ATP-associated measures and potential-sensitive dye signals. Mitochondrial mass also changes, so a dye response should not be described as an independently calibrated voltage measurement unaffected by organelle abundance. The primary experiments do not measure CCN6 catalyzing an electron-transfer reaction. The nondirectional membrane-potential and ROS regulatory assertions are retained as non-core cellular processes.
For PMID:32686858(https://pubmed.ncbi.nlm.nih.gov/32686858/), the complete official abstract has been read; the complete original Methods and Results remain pending. Size-exclusion/native-gel coassociation and perturbation-linked assembly effects do not by themselves make CCN6 a stoichiometric respiratory-complex subunit. The different effects of partial depletion and disease-variant contexts should not be compressed into one universal direction.
For PMID:34811632(https://pmc.ncbi.nlm.nih.gov/articles/PMC9411309/), the official abstract and selected original PMC Results were read. Human mammary-cell depletion and ectopic-expression experiments show respiration/ROS responses that differ from the chondrocyte depletion direction. Clinical expression correlations are distinct from the cellular interventions. The description therefore says context-dependent mitochondrial regulation, not a universal mitochondrial brake. A second mitochondrial molecular-function core is not invented merely to accommodate a well-supported compartment and perturbation phenotype.
The current Human Protein Atlas CCN6 page reports supported mitochondrial staining with HPA064196 in RT4, SK-MEL-30 and U2OS. This independently corroborates the broad compartment. The displayed image pixels and historical GO_REF:0000052 submission have not been reanalyzed. Extracellular prediction fields on that page are not treated as microscopy of secretion.
The complete canonical abstract of PMID:12082632 describes human WISP3 transfection in SUM149 inflammatory-breast-cancer cells, reduced soft-agar growth and angiogenic potential, and xenograft effects in nude mice. The original full assay methods were not recovered. The two experimental suppressive process assertions are retained as non-core with curator deference and explicit context, not rewritten as universal antiproliferative or antiangiogenic molecular activities.
PMID:9843955 is also abstract-only. Its complete abstract distinguishes WISP1/2 induction by Wnt1 in mouse mammary systems from WISP3 identification and expression in human colon tumors. Later direct human ligand experiments support retention of broad signaling, intercellular communication and extracellular localization. The original title is not taken as evidence that WISP3 itself was induced by Wnt1 in the same experiment.
Positive differentiation remains unresolved: inhibition of IGF1-induced hypertrophic markers or BMP-induced alkaline phosphatase does not disprove a positive role at a different step or in another tissue. Equally, cartilage disease alone does not establish the precise positive differentiation process. Broad developmental regulation is retained as contextual and distinct from a specific embryonic development assertion. No PAINT node placement or ARBA rule execution is claimed to have been inspected.
T's bounded consultation at tmp/CCN6-matrix-consultation/consultation.json independently read the complete cached narrative of PMID:23825719, official GO:0140149/GO:0140151 definitions and available donor routes. The normal cache contains XML full text, but its abstract section is bibliography only. The narrative provides CCN-family context and foregrounds CCN1 retinal biology; no inspected target-specific CCN6 interstitial solid-phase localization assay was identified. This is a concrete source-access/scope limit, not a title-only wrong-paralog assertion.
The precise non-collagenous interstitial-matrix term concerns material in the matrix solid phase. A regulatory protein can reside there, so nonstructural signaling is not an exclusion. The broad IBA extracellular-matrix association is retained non-core with curator deference, while the two precise interstitial-matrix rows remain undecided. The combined electronic annotation retains ARBA00096895, D3Z5L9 and ENSMUSP00000076003 exactly. The uninspected rule/donor chain is not labeled false merely because its detailed provenance was inaccessible.
The preliminary candidate contains 8 ACCEPT, 9 KEEP_AS_NON_CORE, 6 UNDECIDED and one proposed NEW molecular function, with one extracellular core. Four normal supplemental caches and their final quote anchors are pending. All original source fields/products and the nine current canonical/frozen preimages are preserved. No old or provider-named research file is authored; no canonical, history, remote or source-cache mutation was performed. Final validation, quote aggregation, rendered YAML equality, a whole independent peer and ROOT acceptance remain separate gates.
ROOT has now authenticated and exclusively imported PMID:17823661, PMID:23424195, PMID:32686858 and PMID:34811632. All four normal caches contain complete abstracts, not full text, including the two records with PMC identifiers. Their exact cached titles agree with the preliminary bibliography. This supersedes the pending-cache statements in the historical preliminary checkpoint above. The original seven gene/publication files and two frozen inventories remained byte-identical through import; these four additions expand protection to thirteen files.
The final snapshot preserves every original assertion, both products, all actions, the one proposed growth-factor-binding activity and the core term set. It removes the evidence-commentary sentence from the biological description, updates the four reference access assessments and adds four short verbatim anchors from the actual imported abstracts: extracellular conditioned-medium context on the existing IDA location, IGF1 secretion on the existing broad signaling TAS, and the direct ligand-binding statement on the NEW activity and its core. The two uses of the binding statement count separately in the quote audit. No notes quotation was added.
The underlying human construct and assay claims still depend on the separately saved original publisher Results/Methods for PMID:17823661 and PMID:23424195. The broad mixed-species abstract wording in PMID:17823661 does not override the body-level separation of human BMP experiments from zebrafish Wnt-coreceptor experiments. The selected original mammary-cell Results for PMID:34811632 remain an external reading; complete Methods, all figure pixels and supplements were not inspected. The final four-cache factual source peer verifies only identity and header/abstract access, not these detailed biological interpretations.
For PMID:32686858 the correct source is the FASEB Journal paper DOI 10.1096/fj.202000405RR. The official abstract and canonical abstract are available; the full Methods/Results are not. One exploratory wrong-DOI request was retained as an access failure and excluded from scientific support. The candidate makes no assay-level statement obtained from that failed request. Coassociation and assembly/respiration phenotypes remain insufficient to assign an intrinsic respiratory enzyme or fixed complex subunit activity.
The final counts remain 8 ACCEPT, 9 KEEP_AS_NON_CORE, 6 UNDECIDED and one NEW molecular function, with one extracellular core. The normal status tool assigns COMPLETE because validation reports zero curation warnings. This workflow status does not resolve the six UNDECIDED biological/source-specific assessments, which remain explicit. Normal validation/status/render, the exact source and quotation audit, and the distinct whole scientific peer are recorded alongside this snapshot. The author has made no canonical/source/history or remote mutation.
This entry supersedes the earlier general growth-factor-binding proposal and the four inherited-role uncertainty judgments. The original 23 machine assertions and two alternative products remain unchanged. The single existing authored NEW is narrowed to insulin-like growth factor I binding, with human IGF1 as partner; no second NEW or new biological-process row is added. The three broad signaling assertions are refined to negative regulation of insulin-like growth factor receptor signaling pathway, and the extracellular core uses that process. Three additional primary references were recovered through the unchanged normal fetcher and imported by ROOT after independent archive and source checks. Their exact cached titles and access flags are included in the reference assessments.
The original IGF1 study reports approximately 30 micromolar SPR affinity for immobilized WISP3. The IGFBP-like domain alone does not confer full binding, and deletion of GCGCC and deletion of GLYC have different interpretive limits: the latter protein expressed too poorly to test binding cleanly. These caveats now accompany the molecular-function judgment. The separate BMP4 conditioned-medium association remains biological context, not support for an IGF1-specific IPI row. UniProt annotates IGFBP_N, TSP1 and CTCK features but no VWC feature; that annotation absence is not proof that the experimentally discussed intervening region or its contribution is absent.
The human receptor-pathway study uses full-length WISP3-transfected SUM149 and HEK293 cells, empty-vector conditioned-medium controls, 24-hour recipient exposure and 15-minute human IGF-I stimulation. Selected complete Methods, Results, captions and Discussion were read. IGF1R phosphorylation is normalized to total receptor, and downstream IRS1/ERK responses decrease. The authors explicitly leave another induced soluble mediator possible. The later direct IGF-I-binding experiment establishes own molecular work, but neither paper by itself proves that extracellular sequestration is the only cause of the conditioned-medium phenotype. This combined evidence supports a bounded process refinement without adding a generic process from necessity alone.
The four IBA annotations are retained as non-core. Uninspected PAINT metadata does not nullify a curated ancestral assertion, and no target loss or erroneous placement was demonstrated. The CTCK domain is present, but distinct CCN-family integrin and glycosaminoglycan sites cannot be assigned to CCN6 from domain presence alone. The target gastric-cell study supplies a real adhesion endpoint: one-hour adhesion to collagen I/IV-coated plates decreases after WISP3 knockdown, whereas fibronectin adhesion is spared. The inspected Methods and Results distinguish this from migration/proliferation assays; the experiment does not show purified integrin binding or isolate the physical adhesive step. Negative hypertrophy/BMP readouts in other CCN6 contexts do not prove loss of every inherited positive differentiation role. The exact inherited interfaces and positive cell state remain research questions.
The complete cached CCN1 narrative was reread. It discusses family matricellular properties and CCN1 retinal biology but supplies no CCN6-specific interstitial solid-phase localization experiment. Its reference assessment is MISCITED for that narrow IDA evidential use while both precise location rows remain UNDECIDED. Nonstructural signaling proteins can reside in matrix; the opinion's structural-versus-regulatory language does not justify removal. The electronic row has a separate unresolved rule/donor boundary.
The original human PPRD genetics paper identifies disease-associated WISP3 variants in unrelated affected individuals. Its official abstract was read; full genetic pedigrees and functional assays were not inspected. This is clinical provenance for the description, not a molecular-function experiment.
Public article links above provide durable provenance for the biological reassessment. Earlier TMP references document local execution or consultations and are not substitutes for publications. The historical journal text is preserved; this entry supersedes its earlier action summaries and workflow-dependent evidential wording. No new verbatim quotation was added in this follow-up, and no source cache was edited.
The newly imported genetics and receptor-pathway caches are abstract-only; the gastric-cell source is an HTML full-text extraction. The latter’s selected cached siRNA, adhesion Methods and Results were checked against the external source. The receptor-pathway Methods and captions remain external access, not contents of its abstract cache. Final actions are 5 ACCEPT, 13 KEEP_AS_NON_CORE, 3 MODIFY, 2 UNDECIDED and the one refined existing NEW; all 23 source assertions, two products and the historical notes prefix are preserved. No additional quotation words were introduced.