Dystroglycan is a processed receptor system. The official Q14118 sequence is an 895-residue precursor, with signal peptide 1–29, alpha chain 30–653 and beta chain 654–895. These chains have complementary functions and are recorded as cleavage products, not splice isoforms. (UniProt Q14118) Glycosylated alpha-DG presents the matrix-ligand-binding surface; beta-DG anchors it in the membrane and engages intracellular partners. One integrated laminin-receptor core captures this physical linkage and its contribution to matrix organization. (PMID:20044576(https://pubmed.ncbi.nlm.nih.gov/20044576/)) (PMID:23940118(https://pubmed.ncbi.nlm.nih.gov/23940118/)) Developmental, synaptic and signaling effects retain their tissue and experimental contexts rather than becoming universal core activities.
The distinction between intrinsic maturation and cleavage by another enzyme is decisive. The 2008 SEA study supports intramolecular precursor cleavage but does not establish the catalytic triad in the serine-endopeptidase definition. Cleavage mutants retain or enhance laminin binding and assembly, so cleavage is not described as universally necessary for ligand competence. (PMID:17905726(https://pubmed.ncbi.nlm.nih.gov/17905726/)) In the separate MMP-9 study, murine beta-DG is the cleaved substrate and human MMP-9 performs proteolysis. (PMID:19946898(https://pubmed.ncbi.nlm.nih.gov/19946898/)) Neither experiment assigns glycan-transfer activity to DAG1. Public original locators and precise read limits are in the evidence guide.
Beta-chain partner evidence has different levels of resolution. Utrophin and dystrophin are established cytoskeletal partners. (PMID:10767429(https://pubmed.ncbi.nlm.nih.gov/10767429/)) (PMID:11495720(https://pubmed.ncbi.nlm.nih.gov/11495720/)) Human DAG1–CLASP co-immunoprecipitation is distinguished from chick epiblast perturbation, and complex association is not described as a purified binary interface. (PMID:23940118(https://pubmed.ncbi.nlm.nih.gov/23940118/)) The NCK1 screen tests a synthetic peptide corresponding to alpha-chain residues 300–313; its interaction with intracellular SH3 domains does not establish accessibility in the native mature receptor. (PMID:17474147(https://pubmed.ncbi.nlm.nih.gov/17474147/)) The ankyrin records test a human beta tail against rat ankyrin-G. (PMID:19109891(https://pubmed.ncbi.nlm.nih.gov/19109891/)) Positive AGR2/AGR3 yeast interactions and the caveolin-3 tail interaction are retained with their assay limits. (PMID:12592373(https://pubmed.ncbi.nlm.nih.gov/12592373/)) (PMID:10988290(https://pubmed.ncbi.nlm.nih.gov/10988290/)) No partner enzyme activity is transferred to DAG1.
The platelet original remains inaccessible. (PMID:18341635(https://pubmed.ncbi.nlm.nih.gov/18341635/)) General cytoskeletal interplay in its abstract is insufficient to reconstruct individual actin, tubulin, MAP2 or vinculin assays and the fine localization/anchoring endpoints. These assertions remain unresolved. A separately inspected human ACTN1–DAG1 curated BioGRID record supports that specific interaction, and plasma-membrane localization is independently established. (BioGRID 616031) (PMID:8017170(https://pubmed.ncbi.nlm.nih.gov/8017170/)) Neither a missing current interaction record nor an unavailable original is evidence of nonbinding.
The human 8UF4 structure supplies direct evidence for calcium coordination. One modeled ion is contacted by five native DAG1 residues in a recombinant 491–748 fragment. Refolding and purification used calcium-rich buffers. This supports contextual calcium binding while leaving native occupancy, affinity and physiological regulation unresolved. It is independent of calcium dependence on the ligand side of an adhesion assay; the paper’s wording about both molecules is kept distinct from the single calcium modeled in the current deposit. (PMID:39305901(https://pubmed.ncbi.nlm.nih.gov/39305901/)) (8UF4 coordinates)
Compartment annotations are interpreted by product and source. ER/Golgi locations describe normal receptor maturation and trafficking, and extracellular alpha-derived fragments can be released. (PMID:17905726(https://pubmed.ncbi.nlm.nih.gov/17905726/)) (PMID:18201566(https://pubmed.ncbi.nlm.nih.gov/18201566/)) Nuclear and cytoplasmic beta-chain pools are supported in cultured human cells without establishing a universal nuclear function. (PMID:18764929(https://pubmed.ncbi.nlm.nih.gov/18764929/)) Two defective-POMT models instead label alpha-DG 30–653 cytosolic; their summations do not supply an operative localization measurement, so those particular assignments remain unresolved. (R-HSA-5615556, R-HSA-5615604) A separate beta-tail observation does not answer that question.
The human parotid exosome workbooks contain IPI00028911.1/DAG1 with two sequences and two spectra. Both measured peptides match the alpha chain. (PMID:19199708(https://pubmed.ncbi.nlm.nih.gov/19199708/)) This supports a preparation-level observation with contamination and product limits, not intact beta-chain cargo or a required extracellular-vesicle function. The prostatic-secretion survey has relevant main methods, but attempted target spreadsheet files were HTML/error bodies; no target values were inferred from them. (PMID:23533145(https://pubmed.ncbi.nlm.nih.gov/23533145/)) The guide gives the original parotid download links, sheet names, rows and literal peptide cells.
The actual PTHR21559 tree and alignment reproduce the 895-residue target. All nine IBD assertions lie on its ancestral path. (PANTHER PTHR21559, GO_REF:0000033) The authentic full GAF was captured on October 12 and reused; this is not presented as a new October 16 download. Human descendant evidence within an IBA is legitimate, and a short donor list is not a defect. Complete current donor queries distinguish experimental provenance from propagated records; donor paper identifiers in the guide identify the curators’ sources, not papers independently read here. Rat expression-pattern evidence is kept separate from mouse perturbation and direct human assays.
The normal default Falcon request failed with an authentication response, and the configured perplexity-lite fallback failed with a quota response. No generated provider report was fabricated or replaced. Manual synthesis uses inspected primary material, explicit independent consultation scopes and the raw database reconstructions described in the guide. Immutable fetched caches and all original annotation source objects are preserved.
The donor-paper citations explicitly named in the review have genuine publication caches. They document the provenance of inspected GO annotation records; obtaining those caches is not a claim to have independently read all donor originals. Three already-existing main-branch caches were preserved exactly.
The alpha and beta chains now have separate, complementary core representations. Alpha-dystroglycan presents the glycan-dependent laminin receptor surface; beta-dystroglycan spans the membrane and connects alpha to dystrophin or utrophin. This makes the performed cytoskeletal adaptor activity explicit without changing the underlying receptor architecture or asserting direct beta-chain binding to laminin or actin. The complete utrophin and dystrophin study abstracts support the cytoplasmic pair interactions, while human muscle immunogold supports the membrane linkage. The dystrophin phosphorylation experiment uses mouse myotubes and peptide assays, not a human full-length ternary-complex measurement. These two cores supersede the earlier single-core representation described above. (PMID:10767429(https://pubmed.ncbi.nlm.nih.gov/10767429/), PMID:11495720(https://pubmed.ncbi.nlm.nih.gov/11495720/), PMID:8017170(https://pubmed.ncbi.nlm.nih.gov/8017170/))
Human DAG1 from an MGC cDNA and human CLASP constructs associate in HEK293 co-immunoprecipitation. The corresponding functional basement-membrane experiments were performed in chick epiblast; they support the conserved structural linkage without becoming direct human embryonic assays. Co-immunoprecipitation establishes association, not an isolated molecular interface. The receptor core uses the already supported basement-membrane organization and extracellular-matrix assembly processes, whose current ontology paths are separate descendants of matrix organization. DAG1 contributes the receptor/scaffold rather than synthesizing matrix ligands. (PMID:23940118(https://pubmed.ncbi.nlm.nih.gov/23940118/), PMID:10366602(https://pubmed.ncbi.nlm.nih.gov/10366602/), PMID:17905726(https://pubmed.ncbi.nlm.nih.gov/17905726/))
The complete PMID9117347 abstract explicitly reports focal-adhesion recruitment in human aortic smooth-muscle cells and rat REF52 fibroblasts. Recruitment depends on cell type and extracellular ligand; some other cell lines do not recruit dystroglycan even on laminin. This independent positive human observation supports contextual retention of the focal-adhesion annotations. It does not recover the separate PMID21423176 DAG1 proteomics assignment or the PMID18341635 platelet localization panel, both of which remain unresolved at the source-assay level. The original rat donor is preserved rather than relabeled as human. (PMID:9117347(https://pubmed.ncbi.nlm.nih.gov/9117347/), PMID:21423176(https://pubmed.ncbi.nlm.nih.gov/21423176/), PMID:18341635(https://pubmed.ncbi.nlm.nih.gov/18341635/))
Supported caveolin/AGR2/AGR3 binding remains contextual under the project policy. The NCK1 peptide observation retains its extracellular alpha-fragment versus intracellular-partner accessibility limit. SEA self-processing remains distinct from MMP9-performed cleavage of DAG1, and no additional process annotation is manufactured by this core representation. (PMID:10988290(https://pubmed.ncbi.nlm.nih.gov/10988290/), PMID:12592373(https://pubmed.ncbi.nlm.nih.gov/12592373/), PMID:17474147(https://pubmed.ncbi.nlm.nih.gov/17474147/), PMID:17905726(https://pubmed.ncbi.nlm.nih.gov/17905726/), PMID:19946898(https://pubmed.ncbi.nlm.nih.gov/19946898/))
Current QuickGO and OLS, both checked on 2026-10-16, name GO:0008093 cytoskeletal adaptor activity. QuickGO lists the older anchor wording as a narrow synonym. The older repository cache explains the label advisory; the label was verified against current primary ontology services.
The explicit user instruction in the ClinGen project policy is to retain supported generic binding as KEEP_AS_NON_CORE when a specific refinement is not established. The same paragraph explicitly calls this an intentional task-specific departure from the skill's informational-exclusion recommendation. This is why the caveolin-3, AGR2 and AGR3 rows remain non-core; no change to the global validation policy is requested or implied.
The two utrophin and dystrophin IPI rows retain cytoskeletal protein binding. Their recovered evidence measures pair binding, while the integrated beta core synthesizes the bridge mechanism with human muscle localization and human co-IP evidence. A core-coverage advisory is not sufficient reason to make each pair assay assert a complete adaptor mechanism. NCK1 likewise remains a source-scoped isolated-peptide binding observation with an explicit physiological-access limitation.
The beta-chain core description now states the muscle structural role as sarcolemmal integrity; no third duplicate core is needed. The MMP9 ectodomain row remains MARK_AS_OVER_ANNOTATED, with the performing enzyme and substrate distinguished in its reason. Intrinsic SEA processing remains separate. All 166 source annotations, 94 reference identities, two processed-product records and eight quote occurrences are unchanged. The two core functions and ten UNDECIDED judgments are retained.