AHNAK, also called desmoyokin, is a 5,890-residue (~700 kDa) intrinsically disordered scaffolding phosphoprotein with three parts: an N-terminal PDZ domain, a very large central region built from roughly 26 to 30 tandem 128-residue repeats, and a ~1,000-residue C-terminal regulatory domain carrying several nuclear localisation signals. It has no catalytic activity and no folded interaction module other than the PDZ domain; its role is to hold other proteins together at the cytoplasmic face of membranes. The C-terminal domain binds the annexin A2/S100A10 heterotetramer through a short motif resolved crystallographically at about 30 nM affinity, binds monomeric actin, cosediments with and bundles actin filaments, and binds the CACNB2 (Cav-beta2) regulatory subunit of L-type voltage-gated calcium channels with 55-60 nM affinity through a proline-rich motif. Through these contacts AHNAK couples the L-type channel to the cortical actin cytoskeleton and acts as a brake on L-type calcium current that beta-adrenergic PKA signalling relieves; it is also required for surface delivery of L-type channels in T lymphocytes and neurons. The central repeats bind and activate protein kinase C-alpha by displacing it from an inhibitory complex with protein phosphatase 2A, and bind and activate phospholipase C-gamma1 in the presence of arachidonic acid. In striated muscle AHNAK is a sarcolemmal and costameric protein and a component of the dysferlin complex, binding the dysferlin C2A domain directly; it is lost from the sarcolemma in dysferlinopathy and appears in calcium-stimulated shed vesicles. In the nucleus it binds the 53BP1 oligomerization domain and restrains 53BP1 multimerisation and phase separation, damping the p53 DNA-damage response. Its distribution between nucleus, cytoplasm and plasma membrane is set by phosphorylation: PKB/Akt phosphorylation drives nuclear export through an adjacent nuclear export signal, while protein kinase C activation and raised calcium drive it toward the membrane. A short alternatively spliced isoform of about 17 kDa, comprising essentially the PDZ domain, shuttles to the nucleus and feeds back on splicing at the AHNAK locus; the same arrangement occurs at the Periaxin locus, with which AHNAK and AHNAK2 share a common ancestor.
Definition: A molecular function in which a protein binds the self-association interface of another protein and limits the size of the oligomer or condensate that partner forms, without preventing its binding to other partners, displacing it from its site of action, or targeting it for degradation.
Justification: AHNAK's nuclear activity has no MF term. It binds the 53BP1 oligomerization domain and caps 53BP1 multimerisation and phase separation; in its absence 53BP1 nuclear bodies enlarge and the p53 response is amplified (PMID:33961796). The two candidate existing terms both fail on their differentia. GO:0140311 protein sequestering activity is defined as binding a protein 'to prevent it from interacting with other partners or to inhibit its localization to the area of the cell or complex where it is active' - but AHNAK does neither: 53BP1 stays on chromatin and stays active, it simply assembles less far. GO:0060090 molecular adaptor activity is the opposite direction, bringing molecules together. This is not an AHNAK-specific gap: the same shape recurs wherever a chaperone-like binder tunes the extent of a biomolecular condensate rather than blocking its formation, and that class of regulator is growing faster than the ontology's vocabulary for it. A biological-process parallel already exists (GO:0032460 negative regulation of protein oligomerization), which this review uses; what is missing is the activity that carries it out. Parenting is the hard part, and the difficulty is itself the gap. The obvious homes do not hold: GO:0140678 molecular function inhibitor activity and GO:0098772 molecular function regulator activity both presume the thing being regulated is a molecular function, and 53BP1 oligomerization is not one - GO models it as a biological process (GO:0051260 protein homooligomerization). GO:0140313 molecular sequestering activity fails on the same differentia as its child GO:0140311. So the term is proposed under GO:0005515 protein binding, which is where it structurally belongs and which is deliberately uninformative, to make the point: GO's molecular-function regulator vocabulary is built to regulate molecular functions, and has no way to say that an activity modulates a partner's self-assembly. If GO were willing to read limiting a partner's self-association as regulating its function, GO:0098772 would become the natural parent; that is a decision for an ontology editor rather than something this review should assume.
Parent term: protein binding
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0003723 RNA binding | HDA PMID:22658674 Insights into RNA biology from an atlas of mammalian mRNA-bi... | KEEP AS NON CORE | Summary: RNA binding, from mRNA interactome capture. Kept as non-core with the caveat the source paper itself supplies. Reason: Interactome capture is a stringent method - covalent UV crosslinking to poly(A) RNA followed by oligo-dT capture - so the datum is not dismissed. But the paper's own headline caveat describes AHNAK exactly: it reports that many proteins in the HeLa mRNA interactome are highly intrinsically disordered and enriched in short repetitive motifs, and AHNAK is the extreme case of that architecture, with 26-30 tandem 128-residue repeats and MobiDB-lite disorder called across the protein. No orthogonal evidence for AHNAK RNA binding exists; the one nucleic-acid measurement in the literature reports weak DNA binding (PMID:15177040). Kept, not removed - the full text was not available and the curator's HDA stands - but it should not be read as a molecular function of AHNAK. Supporting Evidence: PMID:22658674 we find that many proteins of the HeLa mRNA interactome are highly intrinsically disordered and enriched in short repetitive amino acid motifs |
| GO:0005515 protein binding | IPI PMID:15607035 Systematic identification of hepatocellular proteins interac... | KEEP AS NON CORE | Summary: Hepatitis C virus NS5A, from a yeast two-hybrid screen of a liver library. Reason: A single yeast two-hybrid hit, with the partner recorded as the mature NS5A chain (P27958-PRO_0000037576). Host-factor Y2H screens against a viral bait recover large abundant proteins readily, and nothing followed this up. Retained as a recorded interaction, not as evidence of a viral function for AHNAK. |
| GO:0005515 protein binding | IPI PMID:16417406 Hem-1 complexes are essential for Rac activation, actin poly... | KEEP AS NON CORE | Summary: NCKAP1L/Hem-1, from proteomic analysis of the neutrophil Hem-1 (WAVE) complex. Reason: AHNAK was recovered in Hem-1 immunoprecipitates from neutrophils. This is the most biologically suggestive of the screen-derived partners, because the Hem-1/WAVE complex drives Rac-dependent cortical actin polymerisation and AHNAK independently scaffolds Rac activation through PAK and beta-PIX in aortic smooth muscle (PMID:23042471). But the record is one co-IP proteome with no reciprocal test, so the connection is a lead rather than a finding. |
| GO:0005515 protein binding | IPI PMID:17185750 AHNAK, a novel component of the dysferlin protein complex, r... | ACCEPT | Summary: Dysferlin. A direct, domain-mapped, Ca-independent interaction - the strongest IPI on this gene. Accepted despite the uninformative term. Reason: This row is everything the screen-derived GO:0005515 rows are not: the interaction was found by co-immunoprecipitation from skeletal muscle, the binding sites were mapped to the dysferlin C2A domain and the AHNAK C-terminal domain by GST pulldown, and UniProt records it with an experimental evidence tag. It is normally right to replace bare 'protein binding' with an informative child, and it is not done here because GO has no term for binding a ferlin or a membrane-repair C2 protein; GO:0019904 protein domain specific binding would be technically true and no more informative. The scaffolding conclusion this interaction supports is recorded in core_functions instead, and the missing term is raised in suggested_questions. Supporting Evidence: PMID:17185750 We defined the binding sites in dysferlin and AHNAK as the C2A domain in dysferlin and the carboxyterminal domain of AHNAK by glutathione S-transferase (GST)-pull down assays. file:human/AHNAK/AHNAK-uniprot.txt Interacts with DYSF; the interaction is direct |
| GO:0005515 protein binding | IPI PMID:17185750 AHNAK, a novel component of the dysferlin protein complex, r... | ACCEPT | Summary: Mouse dysferlin, same study. Accepted. Reason: The same interaction recorded against the mouse ortholog (Q9ESD7, Swiss-Prot, 2090 aa against human 2080 aa). Cross-species conservation of a mapped interaction strengthens rather than weakens it. |
| GO:0005515 protein binding | IPI PMID:17185750 AHNAK, a novel component of the dysferlin protein complex, r... | ACCEPT | Summary: Myoferlin, same study. Accepted. Reason: Myoferlin is dysferlin's paralogue and its N-terminal domain binds the same AHNAK C-terminal domain. This makes the AHNAK-binding surface a shared property of the ferlin family rather than a dysferlin idiosyncrasy. Supporting Evidence: PMID:17185750 As expected, the N-terminal domain of myoferlin also interacts with the carboxyterminal domain of AHNAK. |
| GO:0005515 protein binding | IPI PMID:20029029 Regulation of epidermal growth factor receptor trafficking b... | KEEP AS NON CORE | Summary: EGFR, from a ubiquitin-reconstruction screen of EGFR trafficking. Real but unmapped; kept as non-core. Reason: Three separate EGFR-bait interactome studies place AHNAK in the EGFR neighbourhood (this row, PMID:24189400 by TAP and PMID:31980649 by ubiquitin reconstruction), which is more replication than most partners on this gene have. None of them maps a binding site, measures affinity, or tests directness, and AHNAK is an abundant cortical protein that is recovered in a large fraction of plasma-membrane bait pulldowns - IntAct lists 608 distinct AHNAK partners. Retained as a proximity observation, not promoted to a molecular function. |
| GO:0005515 protein binding | IPI PMID:23415230 SMARCA3, a chromatin-remodeling factor, is required for p11-... | MODIFY | Summary: S100A10. This is the one GO:0005515 row on AHNAK with a crystal structure behind it; replaced with the informative term. Reason: Bare 'protein binding' understates the evidence. The AHNAK C-terminal motif bound to the S100A10/annexin A2 heterotetramer has been solved twice - at 3.5 A (PMID:22940583, ~30 nM) and at 2.5 A (PMID:23275167) - and the minimal 20-residue motif was mapped biochemically before that (PMID:16984913). GO:0044548 S100 protein binding is already on this gene from PMID:14699089 and is the correct term here too. IntAct records this particular row's method as x-ray diffraction, consistent with a structural rather than a screen-level datum. Proposed replacements: S100 protein binding Supporting Evidence: PMID:22940583 The AHNAK peptide adopts a coil conformation that arches across the heterotetramer contacting both annexin A2 and S100A10 protomers with tight affinity (∼30 nM) and establishing a structural rationale whereby both S100A10 and annexin proteins are needed in AHNAK recruitment. PMID:23275167 Here, the 2.5 Å resolution crystal structure of this 20-amino-acid peptide of AHNAK bound to the AnxA2-S100A10 heterotetramer (1:2:2 symmetry) is presented |
| GO:0005515 protein binding | IPI PMID:23764002 Epidermal growth factor receptor potentiates MCM7-mediated D... | KEEP AS NON CORE | Summary: MCM7, from an EGFR/MCM7 co-immunoprecipitation study. Kept as non-core. Reason: An anti-bait co-immunoprecipitation in a paper about EGFR-driven MCM7 phosphorylation; AHNAK is a bystander in that experiment rather than its subject. MCM7 also appears as a holdup hit for AHNAK's PDZ domain (PMID:36115835), so there are two records, but they are a lysate co-IP and an isolated-domain peptide assay - neither establishes a cellular function. |
| GO:0005515 protein binding | IPI PMID:24189400 Perturbation of the mutated EGFR interactome identifies vuln... | KEEP AS NON CORE | Summary: EGFR, from tandem affinity purification of the mutant EGFR interactome. Reason: Second of three independent EGFR-bait screens recovering AHNAK. Same reasoning as the PMID:20029029 row: replicated proximity, no mapped interface, no functional test on AHNAK. |
| GO:0005515 protein binding | IPI PMID:26496610 A human interactome in three quantitative dimensions organiz... | KEEP AS NON CORE | Summary: HSPA5/BiP, from a quantitative stoichiometry interactome. Kept as non-core. Reason: An anti-tag co-IP in a proteome-scale interactome. BiP is an ER-lumenal chaperone and AHNAK is cytosol-facing, so a direct interaction would require one of them to be somewhere it is not normally found; the likelier reading is co-purification. The second record for this pair is the holdup PDZ screen, which tests a peptide rather than the protein. Retained rather than removed because the measurement itself is real and the full text of the screen was not read. |
| GO:0005515 protein binding | IPI PMID:26618866 ∆F508 CFTR interactome remodelling promotes rescue of cystic... | KEEP AS NON CORE | Summary: CFTR, from the delta-F508 CFTR interactome. Kept as non-core. Reason: CFTR has the deepest record of any AHNAK partner in IntAct - three publications by three methods (anti-bait co-IP here, proximity biotinylation in PMID:36012204, holdup in PMID:36115835) - and UniProt's INTERACTION block gives it NbExp=12, the highest on the entry. That is worth noting, but all three are interactome-scale surveys of CFTR, not studies of AHNAK, and none tests whether AHNAK does anything to CFTR. |
| GO:0005515 protein binding | IPI PMID:30021884 Histone Interaction Landscapes Visualized by Crosslinking Ma... | MODIFY | Summary: S100A10, from in-nucleus crosslinking mass spectrometry. Replaced with the informative term. Reason: The same partner as the structurally resolved row, detected independently by crosslink-MS. GO:0044548 S100 protein binding is the informative term and is already used on this gene for the same interaction from PMID:14699089. Keeping this row as bare protein binding hides that AHNAK's best-characterised molecular interaction has now been seen by pulldown, crosslinking and crystallography. Proposed replacements: S100 protein binding Supporting Evidence: PMID:14699089 The S100A10 subunit serves to mediate the interaction between annexin 2 and the COOH-terminal regulatory domain of AHNAK. |
| GO:0005515 protein binding | IPI PMID:30021884 Histone Interaction Landscapes Visualized by Crosslinking Ma... | KEEP AS NON CORE | Summary: ITK, from in-nucleus crosslinking mass spectrometry. Kept as non-core. Reason: A crosslink-MS hit from a study of histone interaction landscapes. ITK is a T-cell tyrosine kinase and AHNAK is required for calcium signalling in T cells (PMID:18191595, PMID:19497879), so the pairing is not absurd - but those studies are mouse knockouts that work through the L-type channel, not through ITK, and nothing connects the crosslink to a function. ITK is also a holdup hit for the PDZ domain, i.e. two records from two screens. |
| GO:0005515 protein binding | IPI PMID:30833792 A protein-interaction network of interferon-stimulated genes... | KEEP AS NON CORE | Summary: COMMD3, from a pulldown in an interferon-stimulated-gene interaction network. Reason: A pulldown in a network-scale screen, plus a holdup hit for the same pair. No follow-up, no mapped interface, and no proposed biology connecting AHNAK to the CCC/commander complex. |
| GO:0005515 protein binding | IPI PMID:31041561 A pathogenic CtBP1 missense mutation causes altered cofactor... | KEEP AS NON CORE | Summary: CTBP1, from a co-IP in a study of a pathogenic CtBP1 missense variant. Reason: An anti-tag co-IP where CtBP1 is the subject and AHNAK one of many recovered proteins, plus a holdup hit. AHNAK is nuclear as well as cortical, so a nuclear co-IP is plausible; nothing here tests it. |
| GO:0005515 protein binding | IPI PMID:31980649 Extensive rewiring of the EGFR network in colorectal cancer ... | KEEP AS NON CORE | Summary: EGFR, from ubiquitin reconstruction in KRAS(G13D) colorectal cells. Reason: Third of the three independent EGFR-bait screens. Recorded for completeness; the argument is the same as for the other two. |
| GO:0005515 protein binding | IPI PMID:36012204 Differential CFTR-Interactome Proximity Labeling Procedures ... | KEEP AS NON CORE | Summary: CFTR, from proximity-dependent biotin labelling. Kept as non-core. Reason: BioID reports proximity within roughly 10 nm, not binding. Useful as corroboration that AHNAK is in CFTR's cortical neighbourhood; not evidence of a complex. |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of HSPA5/BiP in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. HSPA5 also appears in an anti-tag co-IP interactome (PMID:26496610). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of CFTR in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. CFTR is also seen by co-IP and by BioID, so it is the best-replicated partner on this gene. The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of U2AF2 in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. U2AF2 also appears by proximity biotinylation (PMID:39251607). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of MCM7 in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. MCM7 also appears in a co-IP (PMID:23764002). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of ITK in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. ITK also appears by crosslink-MS (PMID:30021884). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of CTBP1 in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. CTBP1 also appears in a co-IP (PMID:31041561). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | MARK AS OVER ANNOTATED | Summary: RPS6KA1. The only AHNAK partner whose entire evidence is one holdup measurement, and the weakest-scoring of the eight. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. RPS6KA1 is the one partner of the eight with no orthogonal record anywhere in IntAct - no co-IP, no proximity labelling, no crosslink - and it carries the lowest MI-score of the block (0.44 against 0.57-0.67 for the others). Stated as a reusable rule rather than left as a judgement call: within a set of rows that share one experiment and one method, single out the row that has no orthogonal record in any other publication AND sits below the rest of the set on the database's own confidence score. Both conditions are needed - the first alone would catch honest singletons, the second alone would catch ordinary scatter. Here exactly one of eight meets both. A single isolated-domain affinity against a synthetic decapeptide, at the bottom of the screen's own confidence range, should not stand as a GO protein-binding annotation on its own. Marked over-annotated rather than removed because the measurement is real and reproducible in principle; what is missing is any evidence that it happens between the full-length proteins in a cell. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | KEEP AS NON CORE | Summary: AHNAK's PDZ domain retained the C-terminal peptide of COMMD3 in the holdup screen. Kept as non-core: one in-vitro fragment measurement. Reason: One of eight GO:0005515 rows imported from a single holdup screen (PMID:36115835). That screen expressed a recombinant library of all 266 human PDZ domains and tested them against a synthetic library of 448 C-terminal 10-mer PDZ-binding motifs; IntAct holds 415 AHNAK records from it, every one with detection method 'holdup assay' and AHNAK as the prey. So this row reports an affinity between AHNAK's isolated ~80-residue PDZ domain and a 10-residue peptide, not an interaction of the 5,890-residue protein. COMMD3 also appears in a pulldown (PMID:30833792). The partner accession resolves to the reviewed canonical Swiss-Prot entry at full length, so this is not an ORFeome-fragment substitution; the fragment here is on AHNAK's side. Supporting Evidence: PMID:36115835 we expressed a recombinant PDZome library covering all the 266 known human PDZs PMID:36115835 we used the holdup method, a high-throughput comparative chromatographic retention assay that we developed previously |
| GO:0005515 protein binding | IPI PMID:39251607 Systematic identification of post-transcriptional regulatory... | KEEP AS NON CORE | Summary: U2AF2, from proximity labelling in a post-transcriptional regulatory module screen. Reason: Proximity biotinylation, plus a holdup hit for the same pair. AHNAK does have a documented link to splicing regulation, but through its short nuclear isoform acting on its own locus (PMID:21940993), and nothing connects that to U2AF2. Note this publication carries an unflagged 2024 Author Correction (PMID:39468017); nothing in this review rests on it beyond the existence of the proximity hit. |
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: Nuclear localisation, propagated from the PTHR23348 node. Accepted: AHNAK's nuclear residence is its oldest and best-replicated human observation. Reason: The donor set is not a family-level guess. Querying each donor's own evidence shows mouse Prx holds GO:0005634 by EXP and IDA, human PRX by IDA, rat Prx by EXP and mouse Ahnak by its own experimental annotation - that is all four protein donors, the fifth WITH/FROM token being the PANTHER node rather than a protein. Human AHNAK also has its own record: the gene's cloning paper reported predominantly nuclear residence (PMID:1608957), and PKB/Akt phosphorylation driving nuclear export through an adjacent NES is the mechanism that controls it (PMID:11535620). Three of the four protein donors are Periaxin rather than AHNAK orthologs, which is legitimate for IBA but means the ortholog-strength inference on this row rests on mouse Ahnak alone; it does not matter here because the target's own evidence is stronger than the transfer. Propagation Review Root cause: NO FAILURE CORE Sources checked: MGI:MGI:108176 · Mus musculus Prx / Periaxin (O55103, Swiss-Prot, 1391 aa) SUPPORTS TRANSFER A paralogue, not the ortholog - legitimate for IBA but it means this token carries no ortholog-strength inference. MGI:MGI:1316648 · Mus musculus Ahnak (E9Q616, TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The only true ortholog in this WITH/FROM set. Note the entry is unreviewed: there is no Swiss-Prot mouse Ahnak, so its protein NAME is an automatic label even though its GO annotations are curated experimental ones. PANTHER:PTN001156025 · PANTHER tree node in PTHR23348; reaches 74 gene products across metazoa, human reach exactly AHNAK, AHNAK2 and PRX NOT RELEVANT A tree node, not a protein. QuickGO shows it carries exactly three terms (GO:0005634, GO:0005737, GO:0043484) and assigns each to the same 74 entities, so it is the inference structure rather than an evidence source. RGD:619960 · Rattus norvegicus Prx / Periaxin (Q63425, Swiss-Prot, 1383 aa) SUPPORTS TRANSFER A paralogue, not the ortholog. UniProtKB:Q9BXM0 · Homo sapiens PRX / Periaxin (Swiss-Prot, 1461 aa) SUPPORTS TRANSFER The human paralogue in the same PANTHER node. Supporting Evidence: PMID:11535620 Protein kinase B phosphorylates AHNAK and regulates its subcellular localization. |
| GO:0005634 nucleus | IEA GO_REF:0000120 | ACCEPT | Summary: Nucleus from UniProt's combined automatic pipeline. Correct, and redundant with the IBA and the human NAS. Reason: GO_REF:0000120 is the combinatorial multi-method pipeline, not ARBA. Its WITH/FROM combines an Ensembl ortholog projection from mouse Ahnak with UniProt's own SubCell keyword SL-0191 (Nucleus) - and that keyword is on the entry because UniProt's SUBCELLULAR LOCATION line for AHNAK reads simply 'Nucleus'. So the electronic route and the manual route are drawing on the same underlying claim. Correct, and worth keeping for machine consumers, but it adds no independent evidence. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION Sources checked: UniProtKB-SubCell:SL-0191 · UniProt controlled-vocabulary subcellular-location term SL-0191 (Nucleus) NOT RELEVANT A vocabulary term, not an evidence source; it records that the UniProt entry carries the Nucleus location, which is itself the claim being transferred. UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. ensembl:ENSMUSP00000090633 · Ensembl mouse Ahnak protein ENSMUSP00000090633 - the Compara projection source SUPPORTS TRANSFER The Ensembl identifier for the same mouse Ahnak protein as UniProtKB:E9Q616; it is the same donor counted twice in the WITH/FROM field, not a second one. Supporting Evidence: file:human/AHNAK/AHNAK-uniprot.txt CC -!- SUBCELLULAR LOCATION: Nucleus. |
| GO:0005634 nucleus | ISS PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Nucleus by ISS from mouse Ahnak. Correct, and redundant with the IBA and the human NAS. Reason: One of five ISS rows that UniProt/CAFA attached to this gene in one block, all citing PMID:21940993 and all using UniProtKB:E9Q616 (mouse Ahnak) as the donor. E9Q616 is an unreviewed TrEMBL entry - there is no Swiss-Prot mouse Ahnak - so its protein name is an automatic label, although its GO annotations are curated experimental ones. For nucleus the transfer is sound - the donor holds the term experimentally and so does human AHNAK - so this row is redundant rather than wrong. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. |
| GO:0005634 nucleus | NAS PMID:1608957 A human gene (AHNAK) encoding an unusually large protein wit... | ACCEPT | Summary: Nucleus, NAS from the gene's cloning paper. Accepted. Reason: PMID:1608957 is the paper that named AHNAK and reported its predominantly nuclear distribution; a companion study from the same group confirmed the protein is located primarily in the nucleus (PMID:8381120). This is a traceable, gene-specific human observation and the historical anchor for every later nuclear row. Note PubMed records an erratum for this paper; the erratum has no PMID in the record and its scope could not be established, so nothing load-bearing rests on details beyond the nuclear localisation, which later work reproduced. Supporting Evidence: PMID:8381120 The human gene AHNAK encodes a large phosphoprotein located primarily in the nucleus. |
| GO:0005737 cytoplasm | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Cytoplasmic localisation, propagated from the PTHR23348 node. Correct, but the informative terms are cytosol, plasma membrane and sarcolemma, so non-core. Reason: The campaign check for a term whose only support is the sibling paralogue comes back negative here. UniProtKB:Q8IVF2 (AHNAK2) and MGI:MGI:2144831 (mouse Ahnak2) are both in the WITH/FROM list, but AHNAK2 carries GO:0005829 cytosol by its own IDA (GO_REF:0000052) and mouse Ahnak2 carries GO:0030018 and GO:0030659 by IDA from PMID:15007166, so neither is a bare family-level inference. All six non-self protein donors hold their own experimental cytoplasm or cytosol annotation; the seventh protein token is AHNAK itself and the eighth is the PANTHER node. Human AHNAK has an independent IDA of its own (PMID:22057634, primary human myotubes). Propagation Review Root cause: NO FAILURE NON CORE Sources checked: MGI:MGI:108176 · Mus musculus Prx / Periaxin (O55103, Swiss-Prot, 1391 aa) SUPPORTS TRANSFER A paralogue, not the ortholog - legitimate for IBA but it means this token carries no ortholog-strength inference. MGI:MGI:1316648 · Mus musculus Ahnak (E9Q616, TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The only true ortholog in this WITH/FROM set. Note the entry is unreviewed: there is no Swiss-Prot mouse Ahnak, so its protein NAME is an automatic label even though its GO annotations are curated experimental ones. MGI:MGI:2144831 · Mus musculus Ahnak2 (A0A7N9VR94, TrEMBL, 3501 aa; two shorter TrEMBL entries also map) SUPPORTS TRANSFER A paralogue. The MGI id maps to three unreviewed accessions of different lengths (3501, 770 and 409 aa); reported rather than silently resolved to one. PANTHER:PTN001156025 · PANTHER tree node in PTHR23348; reaches 74 gene products across metazoa, human reach exactly AHNAK, AHNAK2 and PRX NOT RELEVANT A tree node, not a protein. QuickGO shows it carries exactly three terms (GO:0005634, GO:0005737, GO:0043484) and assigns each to the same 74 entities, so it is the inference structure rather than an evidence source. RGD:619960 · Rattus norvegicus Prx / Periaxin (Q63425, Swiss-Prot, 1383 aa) SUPPORTS TRANSFER A paralogue, not the ortholog. UniProtKB:Q09666 · AHNAK itself - self-referential WITH/FROM SUPPORTS TRANSFER A self-referential IBA records the PAINT curator's own judgement that the term is ancestral for this gene; it is not circular evidence. UniProtKB:Q8IVF2 · Homo sapiens AHNAK2 (Swiss-Prot, 5795 aa) SUPPORTS TRANSFER The human paralogue. It carries GO:0005829 cytosol by its own IDA (GO_REF:0000052), so this row is not supported only by the sibling. UniProtKB:Q9BXM0 · Homo sapiens PRX / Periaxin (Swiss-Prot, 1461 aa) SUPPORTS TRANSFER The human paralogue in the same PANTHER node. |
| GO:0005737 cytoplasm | IDA PMID:22057634 Ahnak1 abnormally localizes in muscular dystrophies and cont... | KEEP AS NON CORE | Summary: Cytoplasm, IDA in primary human myotubes and patient muscle. The best human evidence for the term, but the term itself is a non-core parent. Reason: A human study: ahnak1 localisation was followed in muscle from LGMD2A and LGMD2B patients and in primary human myotubes, and it redistributes from the sarcolemma to the cytoplasm and into shed vesicles. This is the strongest human cytoplasmic evidence on the gene. Supporting Evidence: PMID:22057634 We found that ahnak1 has lost its sarcolemmal localization in LGMD2B but not in LGMD2A. |
| GO:0005737 cytoplasm | IEA GO_REF:0000117 | KEEP AS NON CORE | Summary: Cytoplasm from an ARBA association rule. Correct, uninformative, non-core. Reason: GO_REF:0000117 is ARBA. The WITH/FROM token is the rule identifier, not a gene product, so there is no donor whose evidence could be interrogated. The term is true and is independently supported by an IDA on this gene (PMID:22057634); this row simply restates it at the least informative level available. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: GRANULARITY MISMATCH Sources checked: ARBA:ARBA00026971 · UniProt ARBA association rule ARBA00026971 NOT RELEVANT A machine-learned association rule, not a gene product. |
| GO:0005765 lysosomal membrane | HDA PMID:17897319 Integral and associated lysosomal membrane proteins. | MARK AS OVER ANNOTATED | Summary: Lysosomal membrane, from a placental lysosome-membrane proteome. No supporting biology anywhere; marked over-annotated. Reason: The paper separates 58 proteins already known to reside in the lysosomal membrane from 'additional 86 proteins that were significantly enriched in the lysosomal membrane fraction'; AHNAK is in the second, enrichment-only set. AHNAK is an abundant peripheral cortical protein with no transmembrane segment, no lipid anchor and no lumenal domain, which is the profile that co-purifies with organelle membrane fractions. Nothing else in the AHNAK literature places it on lysosomes, no follow-up exists, and the term has no functional role to play in any AHNAK model. Marked over-annotated rather than removed because a single HDA whose full text was not read does not justify contradicting the curator outright. Supporting Evidence: PMID:17897319 We report on additional 86 proteins that were significantly enriched in the lysosomal membrane fraction. |
| GO:0005829 cytosol | IDA GO_REF:0000052 | ACCEPT | Summary: Cytosol by HPA immunofluorescence. Accepted. Reason: GO_REF:0000052 is the Human Protein Atlas immunofluorescence curation set - a direct assay on human cells with the antibody validation that comes with it. Consistent with every other cytoplasmic record on this gene. |
| GO:0005829 cytosol | IEA GO_REF:0000107 | ACCEPT | Summary: Cytosol, projected from mouse Ahnak by Ensembl Compara. Correct, redundant. Reason: The donor holds GO:0005829 by IDA from PMID:21940993. The two WITH/FROM tokens are the UniProt and Ensembl identifiers for the same mouse protein, so this row has one donor, not two. Human AHNAK independently has an HPA immunofluorescence IDA to cytosol (GO_REF:0000052), which is why this is redundant rather than load-bearing. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. ensembl:ENSMUSP00000090633 · Ensembl mouse Ahnak protein ENSMUSP00000090633 - the Compara projection source SUPPORTS TRANSFER The Ensembl identifier for the same mouse Ahnak protein as UniProtKB:E9Q616; it is the same donor counted twice in the WITH/FROM field, not a second one. |
| GO:0005829 cytosol | ISS PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Cytosol by ISS from mouse Ahnak. Redundant with the HPA IDA on this gene. Reason: One of five ISS rows that UniProt/CAFA attached to this gene in one block, all citing PMID:21940993 and all using UniProtKB:E9Q616 (mouse Ahnak) as the donor. E9Q616 is an unreviewed TrEMBL entry - there is no Swiss-Prot mouse Ahnak - so its protein name is an automatic label, although its GO annotations are curated experimental ones. The donor's cytosol evidence is an IDA from the cited paper, so this particular row's reference and its donor evidence do agree - which is not true of the costamere and elasticity rows in the same block. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. |
| GO:0005886 plasma membrane | IDA GO_REF:0000052 | ACCEPT | Summary: Plasma membrane by HPA immunofluorescence. Accepted; this is AHNAK's core location. Reason: A direct human immunofluorescence assay for the localisation that all of AHNAK's characterised biochemistry depends on - the annexin A2/S100A10 platform, the cortical actin link and the Cav-beta2 interaction all occur at the inner leaflet. |
| GO:0005886 plasma membrane | IDA PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Plasma membrane, IDA. Accepted. Reason: A direct assay on this gene for its principal location; consistent with the HPA and human-muscle IDAs. |
| GO:0005886 plasma membrane | IDA PMID:22057634 Ahnak1 abnormally localizes in muscular dystrophies and cont... | ACCEPT | Summary: Plasma membrane, IDA in human muscle. Accepted. Reason: Same human study as the cytoplasm and sarcolemma rows; the sarcolemma row is the informative one. |
| GO:0005886 plasma membrane | IDA PMID:3126079 The submembranous location of p11 and its interaction with t... | ACCEPT | Summary: Plasma membrane, IDA. The term is right and core; this row's reference cannot support it, because the paper predates AHNAK's discovery. Reason: PMID:3126079 is Osborn et al. 1988 on the submembranous location of p11 and p36 - that is, S100A10 and annexin A2. It contains no AHNAK experiment and could not: the AHNAK gene was not cloned until 1992 (PMID:1608957). Querying QuickGO by reference shows the row is a ComplexPortal projection: PMID:3126079 carries GO:0005886 across eight entities by IDA - CPX-850, CPX-853 and CPX-898, plus ANXA2 and S100A10 in both human and mouse, plus AHNAK - with four further NAS rows on CPX-856, ANXA2, S100A10 and HLTF, every one assignedBy ComplexPortal; ten entities in all. The complex CPX-850 (ahnak-anxa2-s100a10) is genuinely at the plasma membrane and the term therefore stands - it is independently held on this gene by two IDAs and an HPA assay - so this is a reference defect, not a term defect, and the row is accepted with the provenance recorded rather than downgraded. Supporting Evidence: PMID:3126079 p36, a major cytoplasmic substrate of pp60 src kinase, is present beneath the plasma membrane. PMID:1608957 A human gene (AHNAK) encoding an unusually large protein |
| GO:0005886 plasma membrane | IEA GO_REF:0000117 | ACCEPT | Summary: Plasma membrane from an ARBA rule. Accepted - this is where AHNAK does most of what it does. Reason: Unlike the cytoplasm ARBA row, this one names AHNAK's principal site of action. It is corroborated four ways on this gene alone: an HPA immunofluorescence IDA (GO_REF:0000052), an IDA in primary human myotubes (PMID:22057634), an IDA from PMID:21940993, and the crystallographically defined AHNAK/annexin A2/S100A10 complex that sits on the inner leaflet (PMID:22940583, PMID:23275167). The ARBA rule identifier is not a gene product, so there is no donor evidence to query. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00027801 · UniProt ARBA association rule ARBA00027801 NOT RELEVANT A machine-learned association rule, not a gene product. |
| GO:0005925 focal adhesion | HDA PMID:21423176 Analysis of the myosin-II-responsive focal adhesion proteome... | KEEP AS NON CORE | Summary: Focal adhesion, from the myosin-II-responsive focal adhesion proteome. Reason: A high-throughput direct assay: isolated focal-adhesion fractions analysed by mass spectrometry. AHNAK is cortical and actin-associated, so its recovery is expected, but no AHNAK-specific focal-adhesion function has been reported and the paper's cached full text does not discuss it. |
| GO:0015629 actin cytoskeleton | IDA PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Actin cytoskeleton, IDA. Accepted. Reason: Independently corroborated by biochemistry that GOA does not otherwise capture: the AHNAK C-terminal region binds G-actin and cosediments with F-actin (PMID:12153988) and bundles actin filaments into paracrystalline arrays (PMID:15001564), and AHNAK redistributes with the actin cytoskeleton when cells are treated with cytochalasin D (PMID:14699089). This review proposes the corresponding molecular-function and process terms, which are currently absent. Supporting Evidence: PMID:12153988 The same carboxyl-terminal ahnak region (aa 5262-5643) bound to G-actin and cosedimented with F-actin. |
| GO:0016020 membrane | HDA PMID:19946888 Defining the membrane proteome of NK cells. | KEEP AS NON CORE | Summary: Membrane, from an NK-cell membrane proteome. True but subsumed by plasma membrane. Reason: The paper isolated membranes and 'treated with reagents that have been reported to remove peripheral membrane proteins', so recovering AHNAK - a peripheral, cytosol-facing protein with no transmembrane segment and no lipid anchor in UniProt's feature table - says it resisted that treatment, not that it is an integral membrane protein. The general term is kept rather than MODIFYed to GO:0005886, and the rule this review applies is: refine a term when the same experiment that produced the row identifies the more specific entity, and keep it when the term is general because the assay itself cannot discriminate. The two GO:0005515 rows moved to GO:0044548 qualify because those experiments name S100A10; a bulk membrane preparation names no membrane, so replacing this row with plasma membrane would assert something PMID:19946888 did not show. AHNAK's plasma-membrane localisation is certain, but it is carried by the five rows that measured it. Supporting Evidence: PMID:19946888 Isolated membranes were treated with reagents that have been reported to remove peripheral membrane proteins. |
| GO:0030315 T-tubule | NAS PMID:17185750 AHNAK, a novel component of the dysferlin protein complex, r... | ACCEPT | Summary: T-tubule, NAS. Supported in human cardiac muscle and contested in skeletal muscle; accepted with both observations recorded. Reason: The cited paper states that dysferlin and AHNAK co-localise at the sarcolemma and T-tubules of normal skeletal muscle. A later study using AHNAK1- and AHNAK2-specific antibodies reports the opposite for skeletal muscle - 'no AHNAK expression was detected in the T-tubule system' (PMID:20833135) - and since that study distinguishes the two paralogues while the earlier one did not, isoform cross-reactivity is a plausible reconciliation. Crucially the disagreement is confined to skeletal muscle: in human left ventricle, confocal microscopy localises the AHNAK C-terminal portion to the sarcolemma including the T-tubular system (PMID:12153988). The term is therefore supported for cardiac muscle and contested for skeletal muscle, which is a tissue difference rather than a contradiction, and the NAS code understates evidence that exists in two papers. Supporting Evidence: PMID:17185750 In normal skeletal muscle, dysferlin and AHNAK colocalize at the sarcolemmal membrane and T-tubules. PMID:12153988 Confocal microscopy of human left ventricular tissue localized the carboxyl-terminal ahnak portion to the sarcolemma including the T-tubular system and the intercalated disks of cardiomyocytes. PMID:20833135 In contrast, no AHNAK expression was detected in the T-tubule system. |
| GO:0031982 vesicle | IDA PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | KEEP AS NON CORE | Summary: Vesicle, IDA. True but the least informative of AHNAK's vesicle records. Reason: The same rule as the GO:0016020 row: refine when the experiment identifies the specific entity, keep when it cannot. The substantive vesicle biology on this gene - Ca-stimulated shedding of ~150 nm vesicles from human myotubes (PMID:22057634) and a requirement for carcinoma extracellular-vesicle production (PMID:27374178) - is captured by the extracellular exosome rows, but it comes from different experiments. PMID:21940993's full text is not available, so which vesicle class this IDA refers to cannot be established, and no replacement term is proposed rather than guessing at GO:0031410 or an extracellular one. |
| GO:0042383 sarcolemma | IDA PMID:22057634 Ahnak1 abnormally localizes in muscular dystrophies and cont... | ACCEPT | Summary: Sarcolemma, IDA in human muscle. Accepted and core for the muscle role. Reason: Sarcolemmal AHNAK is where the dysferlin-complex and costamere biology sits, and this row is a direct assay on human tissue rather than a projection. Its loss in dysferlinopathy is the observation that first placed AHNAK in the dysferlin complex (PMID:17185750). Supporting Evidence: PMID:22057634 We found that ahnak1 has lost its sarcolemmal localization in LGMD2B but not in LGMD2A. |
| GO:0042383 sarcolemma | NAS PMID:17185750 AHNAK, a novel component of the dysferlin protein complex, r... | ACCEPT | Summary: Sarcolemma, NAS. Redundant with the human IDA for the same term. Reason: The same term is held on this gene by IDA from human muscle (PMID:22057634), which is the row that should carry it. This NAS duplicate adds nothing and its evidence code understates the cited paper, which reports the localisation directly. |
| GO:0043034 costamere | IEA GO_REF:0000107 | ACCEPT | Summary: Costamere, projected from mouse Ahnak by Ensembl Compara. Accepted - the donor's evidence is a direct, specific IDA. Reason: The donor mouse Ahnak holds GO:0043034 by IDA from PMID:20833135, where AHNAK1 and AHNAK2 were shown to co-localise with vinculin in mouse skeletal muscle. This is the rare propagated row that lands on exactly the term its donor holds, at exactly the donor's specificity, from a true ortholog. Human corroboration is indirect but consistent: AHNAK is sarcolemmal in human muscle and is lost from the sarcolemma in dysferlinopathy (PMID:22057634). Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. ensembl:ENSMUSP00000090633 · Ensembl mouse Ahnak protein ENSMUSP00000090633 - the Compara projection source SUPPORTS TRANSFER The Ensembl identifier for the same mouse Ahnak protein as UniProtKB:E9Q616; it is the same donor counted twice in the WITH/FROM field, not a second one. Supporting Evidence: PMID:20833135 Co-localization of AHNAK1 and AHNAK2 with vinculin clearly demonstrates that both proteins are components of the costameric network. |
| GO:0043034 costamere | ISS PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Costamere by ISS. The term is right, but this row cites a paper that does not contain the costamere experiment. Reason: One of five ISS rows that UniProt/CAFA attached to this gene in one block, all citing PMID:21940993 and all using UniProtKB:E9Q616 (mouse Ahnak) as the donor. E9Q616 is an unreviewed TrEMBL entry - there is no Swiss-Prot mouse Ahnak - so its protein name is an automatic label, although its GO annotations are curated experimental ones. Here the block's uniform referencing breaks down. The donor E9Q616 holds GO:0043034 by IDA from PMID:20833135 - the vinculin co-localisation study - whereas this row cites PMID:21940993, which is about alternative splicing at the AHNAK locus and contains no costamere experiment. The same mismatch affects the GO:0097493 row in the block. The term itself is well founded and is separately and correctly propagated by Ensembl Compara, so this row is redundant as well as mis-referenced. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: SOURCE MISCITATION CIRCULAR PROPAGATION Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. Supporting Evidence: PMID:20833135 Co-localization of AHNAK1 and AHNAK2 with vinculin clearly demonstrates that both proteins are components of the costameric network. |
| GO:0043484 regulation of RNA splicing | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Regulation of RNA splicing, propagated from the PTHR23348 node to 74 gene products. Kept, but non-core and isoform-restricted. Reason: The observation is real but singular and isoform-restricted. QuickGO shows that across the whole family the only experimental annotation to GO:0043484 anywhere is mouse Ahnak (E9Q616) by IDA from PMID:21940993; human PRX, rat Prx, mouse Prx and human AHNAK2 all hold the term by IBA alone, propagated from that one study. Human AHNAK then receives it three times over - IBA (GO_REF:0000033), Ensembl Compara IEA (GO_REF:0000107) and CAFA ISS - so three GOA rows are three routes out of one experiment, not three findings. The study also assigns the activity to the small ~17 kDa AHNAK isoform that shuttles to the nucleus, which corresponds to UniProt isoform Q09666-2 (residues 1-149 with an alternative C-terminus, i.e. the PDZ domain alone), not to the 700 kDa protein the canonical accession represents. The PAINT call itself is more defensible than a single-gene reading suggests, because the same paper reports the behaviour for Periaxin as well as AHNAK, so the node has two family members behind it rather than one. What is not defensible is treating this as a function of the 700 kDa protein: it is a property of the short nuclear isoform, and it is peripheral to everything AHNAK does at the cell cortex. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: GRANULARITY MISMATCH Sources checked: MGI:MGI:108176 · Mus musculus Prx / Periaxin (O55103, Swiss-Prot, 1391 aa) SUPPORTS TRANSFER A paralogue, not the ortholog - legitimate for IBA but it means this token carries no ortholog-strength inference. MGI:MGI:1316648 · Mus musculus Ahnak (E9Q616, TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The only true ortholog in this WITH/FROM set. Note the entry is unreviewed: there is no Swiss-Prot mouse Ahnak, so its protein NAME is an automatic label even though its GO annotations are curated experimental ones. PANTHER:PTN001156025 · PANTHER tree node in PTHR23348; reaches 74 gene products across metazoa, human reach exactly AHNAK, AHNAK2 and PRX NOT RELEVANT A tree node, not a protein. QuickGO shows it carries exactly three terms (GO:0005634, GO:0005737, GO:0043484) and assigns each to the same 74 entities, so it is the inference structure rather than an evidence source. Supporting Evidence: PMID:21940993 During muscle differentiation the small AHNAK is strongly increased, thereby establishing a positive feedback loop to regulate mRNA splicing of its own locus. PMID:21940993 These proteins interact in the cytoplasm, but the small AHNAK is also present in the nucleus. PMID:21940993 A small 17-kDa isoform of Periaxin similarly traffics between the cytoplasm and the nucleus to regulate mRNA splicing. |
| GO:0043484 regulation of RNA splicing | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Regulation of RNA splicing, projected from mouse Ahnak by Ensembl Compara. Reason: The observation is real but singular and isoform-restricted. QuickGO shows that across the whole family the only experimental annotation to GO:0043484 anywhere is mouse Ahnak (E9Q616) by IDA from PMID:21940993; human PRX, rat Prx, mouse Prx and human AHNAK2 all hold the term by IBA alone, propagated from that one study. Human AHNAK then receives it three times over - IBA (GO_REF:0000033), Ensembl Compara IEA (GO_REF:0000107) and CAFA ISS - so three GOA rows are three routes out of one experiment, not three findings. The study also assigns the activity to the small ~17 kDa AHNAK isoform that shuttles to the nucleus, which corresponds to UniProt isoform Q09666-2 (residues 1-149 with an alternative C-terminus, i.e. the PDZ domain alone), not to the 700 kDa protein the canonical accession represents. This row is the Compara route of the three. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION GRANULARITY MISMATCH Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. ensembl:ENSMUSP00000090633 · Ensembl mouse Ahnak protein ENSMUSP00000090633 - the Compara projection source SUPPORTS TRANSFER The Ensembl identifier for the same mouse Ahnak protein as UniProtKB:E9Q616; it is the same donor counted twice in the WITH/FROM field, not a second one. Supporting Evidence: PMID:21940993 During muscle differentiation the small AHNAK is strongly increased, thereby establishing a positive feedback loop to regulate mRNA splicing of its own locus. PMID:21940993 These proteins interact in the cytoplasm, but the small AHNAK is also present in the nucleus. |
| GO:0043484 regulation of RNA splicing | ISS PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | KEEP AS NON CORE | Summary: Regulation of RNA splicing by ISS from mouse Ahnak. The third route out of one experiment. Reason: The observation is real but singular and isoform-restricted. QuickGO shows that across the whole family the only experimental annotation to GO:0043484 anywhere is mouse Ahnak (E9Q616) by IDA from PMID:21940993; human PRX, rat Prx, mouse Prx and human AHNAK2 all hold the term by IBA alone, propagated from that one study. Human AHNAK then receives it three times over - IBA (GO_REF:0000033), Ensembl Compara IEA (GO_REF:0000107) and CAFA ISS - so three GOA rows are three routes out of one experiment, not three findings. The study also assigns the activity to the small ~17 kDa AHNAK isoform that shuttles to the nucleus, which corresponds to UniProt isoform Q09666-2 (residues 1-149 with an alternative C-terminus, i.e. the PDZ domain alone), not to the 700 kDa protein the canonical accession represents. One of five ISS rows that UniProt/CAFA attached to this gene in one block, all citing PMID:21940993 and all using UniProtKB:E9Q616 (mouse Ahnak) as the donor. E9Q616 is an unreviewed TrEMBL entry - there is no Swiss-Prot mouse Ahnak - so its protein name is an automatic label, although its GO annotations are curated experimental ones. Propagation Review Root cause: EVIDENCE CIRCULAR OR REDUNDANT Failure modes: CIRCULAR PROPAGATION GRANULARITY MISMATCH Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. Supporting Evidence: PMID:21940993 During muscle differentiation the small AHNAK is strongly increased, thereby establishing a positive feedback loop to regulate mRNA splicing of its own locus. PMID:21940993 These proteins interact in the cytoplasm, but the small AHNAK is also present in the nucleus. |
| GO:0044291 cell-cell contact zone | IDA PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | ACCEPT | Summary: Cell-cell contact zone, IDA. Accepted. Reason: Independently supported: AHNAK and the annexin 2/S100A10 complex are recruited to calcium-dependent cell-cell contacts and dissociate from the membrane when those contacts are broken with EGTA (PMID:14699089). Consistent with the E-cadherin interactome row. Supporting Evidence: PMID:14699089 calcium-dependent cell–cell adhesion provides a signal for initiating the localized recruitment of AHNAK and annexin 2/S100A10 to sites of cell–cell contacts |
| GO:0044548 S100 protein binding | IPI PMID:14699089 AHNAK interaction with the annexin 2/S100A10 complex regulat... | ACCEPT | Summary: S100 protein binding, IPI with S100A10. Accepted - this is AHNAK's best-characterised molecular interaction. Reason: The evidence chain behind this single row is unusually complete for this gene: S100A10 pulled down by recombinant GST-AHNAK C-terminal domain (PMID:14699089), the minimal 20-residue binding motif mapped and shown to require both annexin A2 and S100A10 (PMID:16984913), and two independent crystal structures of that motif bound to the heterotetramer at ~30 nM affinity (PMID:22940583, PMID:23275167). AHNAK is also a specific Ca- and Zn-dependent target of S100B (PMID:11312263), so the term's scope is right as well as its substance. Supporting Evidence: PMID:14699089 When expressed in rabbit reticulocyte, S100A10 protein binds to and is pulled down by the recombinant GST-COOH-terminal domain of AHNAK PMID:16984913 Identification of an AHNAK binding motif specific for the Annexin2/S100A10 tetramer. |
| GO:0045121 membrane raft | IDA PMID:14699089 AHNAK interaction with the annexin 2/S100A10 complex regulat... | KEEP AS NON CORE | Summary: Membrane raft, IDA by detergent-resistant flotation and cholera-toxin colocalisation. Reason: Directly assayed - AHNAK cofractionated with annexin 2, S100A10 and caveolin-1 in OptiPrep flotation and colocalised with the GM1-binding cholera toxin B subunit - but in MDCK (canine) cells, and only a portion of the annexin 2 pool cofractionated. A raft subdomain is a refinement of the plasma-membrane localisation rather than a separate site of action. Supporting Evidence: PMID:14699089 In MDCK cells, a significant amount of AHNAK partitioned with annexin 2 and S100A10 into the lipid raft fraction together with caveolin-1, a known cholesterol-binding protein |
| GO:0045296 cadherin binding | HDA PMID:25468996 E-cadherin interactome complexity and robustness resolved by... | KEEP AS NON CORE | Summary: Cadherin binding, from the E-cadherin interactome. Kept as non-core. Reason: A high-throughput co-IP proteome of E-cadherin complexes. It is coherent with AHNAK's independently demonstrated recruitment to calcium-dependent cell-cell contacts (PMID:14699089) and with its GO:0044291 IDA, so this is not an isolated oddity. But an HDA from a junctional co-IP cannot distinguish binding a cadherin from being present in the junctional cortex, and no direct AHNAK-cadherin measurement exists. Supporting Evidence: PMID:14699089 calcium-dependent cell–cell adhesion provides a signal for initiating the localized recruitment of AHNAK and annexin 2/S100A10 to sites of cell–cell contacts |
| GO:0070062 extracellular exosome | HDA PMID:19056867 Large-scale proteomics and phosphoproteomics of urinary exos... | KEEP AS NON CORE | Summary: Extracellular exosome, from urinary exosome proteomics. Reason: A second independent human exosome proteome. Same reasoning as the parotid row. |
| GO:0070062 extracellular exosome | HDA PMID:19199708 Proteomic analysis of human parotid gland exosomes by multid... | KEEP AS NON CORE | Summary: Extracellular exosome, from parotid gland exosome proteomics. Reason: A mass-spectrometry catalogue of human exosomes. Consistent with the human-myotube IDA and with AHNAK being the dominant protein of carcinoma extracellular vesicles, so this is corroboration rather than noise - but it remains a cargo observation. |
| GO:0070062 extracellular exosome | IDA PMID:22057634 Ahnak1 abnormally localizes in muscular dystrophies and cont... | KEEP AS NON CORE | Summary: Extracellular exosome, IDA from Ca-stimulated vesicle shedding by human myotubes. Reason: A real human observation, and AHNAK is reported elsewhere as the most abundant protein of carcinoma-derived extracellular vesicles and necessary for their formation (PMID:27374178) - so extracellular vesicles are a genuine part of AHNAK biology. Two caveats keep it non-core. The vesicles measured here are ~150 nm Ca-stimulated shed microvesicles, which are not exosomes in the strict endosomal sense; and being cargo of a vesicle is a fate, not a site of action. Supporting Evidence: PMID:22057634 Electron microscopy revealed a homogenous population of vesicles with a diameter of about 150 nm. |
| GO:0097493 structural molecule activity conferring elasticity | IEA GO_REF:0000107 | MARK AS OVER ANNOTATED | Summary: Structural molecule activity conferring elasticity, projected from mouse Ahnak. An elastomeric molecular function has never been measured for this protein. Reason: GO:0097493 is defined as 'The action of a molecule that contributes to the structural integrity of a complex or assembly within or outside a cell, providing elasticity and recoiling'. The donor's entire evidence for it is a single IMP: atomic force microscopy found that whole AHNAK1-deficient mouse skeletal muscle fibres are stiffer than wild-type (PMID:20833135). That is a tissue-mechanics phenotype in a knockout. It is consistent with the term's direction - losing AHNAK does reduce compliance - but it does not measure any mechanical property of the AHNAK molecule, and no force spectroscopy, reconstitution or extensibility measurement on AHNAK or its central repeats has ever been published. An elastomeric molecular function is a large inferential step from a knockout fibre, and it reaches human AHNAK only through a mouse ortholog that is itself unreviewed. Propagation Review Root cause: SOURCE WEAK OR INFERRED Failure modes: GRANULARITY MISMATCH SOURCE EVIDENCE WEAK Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. ensembl:ENSMUSP00000090633 · Ensembl mouse Ahnak protein ENSMUSP00000090633 - the Compara projection source SUPPORTS TRANSFER The Ensembl identifier for the same mouse Ahnak protein as UniProtKB:E9Q616; it is the same donor counted twice in the WITH/FROM field, not a second one. Supporting Evidence: PMID:20833135 Using atomic force microscopy (AFM), we observed a significantly higher transverse stiffness of AHNAK1⁻/⁻ fibers. |
| GO:0097493 structural molecule activity conferring elasticity | ISS PMID:21940993 Self-regulated alternative splicing at the AHNAK locus. | MARK AS OVER ANNOTATED | Summary: Structural molecule activity conferring elasticity by ISS, and citing a paper that contains neither the experiment nor the term. Reason: GO:0097493 is defined as 'The action of a molecule that contributes to the structural integrity of a complex or assembly within or outside a cell, providing elasticity and recoiling'. The donor's entire evidence for it is a single IMP: atomic force microscopy found that whole AHNAK1-deficient mouse skeletal muscle fibres are stiffer than wild-type (PMID:20833135). That is a tissue-mechanics phenotype in a knockout. It is consistent with the term's direction - losing AHNAK does reduce compliance - but it does not measure any mechanical property of the AHNAK molecule, and no force spectroscopy, reconstitution or extensibility measurement on AHNAK or its central repeats has ever been published. An elastomeric molecular function is a large inferential step from a knockout fibre, and it reaches human AHNAK only through a mouse ortholog that is itself unreviewed. One of five ISS rows that UniProt/CAFA attached to this gene in one block, all citing PMID:21940993 and all using UniProtKB:E9Q616 (mouse Ahnak) as the donor. E9Q616 is an unreviewed TrEMBL entry - there is no Swiss-Prot mouse Ahnak - so its protein name is an automatic label, although its GO annotations are curated experimental ones. This row compounds the problem by citing PMID:21940993, the alternative-splicing paper, while the donor's evidence for the term is an IMP from PMID:20833135. Propagation Review Root cause: SOURCE WEAK OR INFERRED Failure modes: GRANULARITY MISMATCH SOURCE EVIDENCE WEAK SOURCE MISCITATION Sources checked: UniProtKB:E9Q616 · Mus musculus Ahnak (TrEMBL/unreviewed, 5656 aa) - the mouse ortholog SUPPORTS TRANSFER The single donor for every Ensembl-Compara and CAFA-ISS row on this gene. Supporting Evidence: PMID:20833135 Using atomic force microscopy (AFM), we observed a significantly higher transverse stiffness of AHNAK1⁻/⁻ fibers. |
| GO:0098797 plasma membrane protein complex | IPI PMID:23275167 Structure of a C-terminal AHNAK peptide in a 1:2:2 complex w... | ACCEPT | Summary: Plasma membrane protein complex, from the AHNAK/S100A10/annexin A2 crystal structure. Accepted and core. Reason: ComplexPortal CPX-850. Unlike the other two ComplexPortal rows on this gene, this one cites a paper that actually contains the AHNAK experiment: a 2.5 A structure of the AHNAK C-terminal 20-mer bound to the annexin A2/S100A10 heterotetramer. The reference annotates exactly three entities - ANXA2, S100A10 and AHNAK - which is the complex's real composition rather than a projection across a larger set. Supporting Evidence: PMID:23275167 Structure of a C-terminal AHNAK peptide in a 1:2:2 complex with S100A10 and an acetylated N-terminal peptide of annexin A2. PMID:23275167 a minimal binding motif has been mapped to a 20-amino-acid peptide corresponding to residues 5654-5673 of the AHNAK C-terminal domain |
| GO:1905686 positive regulation of plasma membrane repair | NAS PMID:22940583 Structure of an asymmetric ternary protein complex provides ... | MARK AS OVER ANNOTATED | Summary: Positive regulation of plasma membrane repair. The cited paper contains no membrane-repair experiment, and the one experiment that tested AHNAK in membrane repair found no requirement. Reason: PMID:22940583 is a crystallography paper. Its AHNAK content is a 3.5 A structure of a C-terminal peptide bound to the S100A10/annexin A2 heterotetramer; membrane repair appears only in its opening framing sentence, which is why the curator correctly coded it NAS. Querying QuickGO by reference rather than by gene shows this is a ComplexPortal complex-to-subunit projection: PMID:22940583 carries GO:1905686 by NAS on ten entities - four complexes (CPX-850, CPX-853, CPX-898, CPX-905) plus ANXA2, S100A10 and AHNAK in both human and mouse - every one assignedBy ComplexPortal with the same evidence code. The phenotype spreads across the whole set, which is the signature of a projection rather than of per-protein curation. Against that, the only published functional test is negative: a laser-wounding assay on AHNAK1-deficient muscle fibres found no repair defect (PMID:20833135). The negative is a single assay in mouse skeletal muscle and is hedged by its authors, so it does not license removal - AHNAK could still contribute in other cell types, and its partners annexin A2, S100A10 and dysferlin are genuine repair proteins. But a positive-regulation assertion with no supporting experiment anywhere and a loss-of-function negative in the obvious assay is over-annotated, and the guilt-by-complex-membership route by which it arrived is worth recording. Supporting Evidence: PMID:20833135 A laser wounding assay with AHNAK1-deficient fibers suggests that AHNAK1 is not involved in membrane repair. PMID:22940583 Here, we present the three-dimensional structure of a multiprotein complex that includes S100A10, annexin A2, and AHNAK, which along with dysferlin, functions in muscle and cardiac tissue repair. |
| GO:0005246 calcium channel regulator activity | IDA PMID:16319140 Ahnak is critical for cardiac Ca(V)1.2 calcium channel funct... | NEW | Summary: Proposed. AHNAK binds the CACNB2 (Cav-beta2) subunit of the L-type voltage-gated calcium channel and acts as a brake on L-type current that PKA phosphorylation relieves. GOA records nothing about calcium channels on this gene. Reason: This is AHNAK's most thoroughly dissected activity and it is entirely absent from the GO record. The interaction is mapped and quantified with recombinant AHNAK fragments: the C-terminal region binds beta2a with Kd 55 nM (PMID:12153988), the most C-terminal 188 residues carry a PxxP motif that binds Cav-beta2 with Kd ~60 nM and lose ~20-fold affinity when the prolines are mutated (PMID:20607281), and binding is to the SH3-HOOK-GK core of Cav-beta2 and is tuned by PKA phosphorylation of Cav-beta2 Ser-296 (PMID:22497893). The functional consequence was established with a naturally occurring human AHNAK variant, Ile5236Thr in the 5,643-residue sequence used in that study: the mutant fragment halves beta2 binding, mimics the PKA effect on I(CaL) and blocks further up-regulation by isoprenaline, from which the authors conclude the ahnak C1 domain is a physiological brake on the current (PMID:16319140). The specificity control is strong - neither wild-type nor mutant fragment affects I(CaL) in ahnak1-deficient cardiomyocytes, so the effect requires endogenous AHNAK rather than being a peptide artefact. Evidence code IDA: the binding and the electrophysiology are direct assays on defined AHNAK fragments. The electrophysiology itself was performed in rat and mouse cardiomyocytes. Supporting Evidence: PMID:16319140 Hence, we suggest the ahnak-C1 domain serves as physiological brake on I(CaL). PMID:12153988 Equilibrium sedimentation analysis revealed Kd values of 55 +/- 11 nM and 328 +/- 24 nM for carboxyl-terminal (aa 195-606) and amino-terminal (aa 1-200) truncates of the beta2a subunit, respectively. PMID:20607281 In vitro binding studies revealed that 188-PSTP conferred high-affinity binding to Cavbeta(2) (K (d) approximately 60 nM). |
| GO:1901385 regulation of voltage-gated calcium channel activity | ISS PMID:20607281 Ahnak1 modulates L-type Ca(2+) channel inactivation of roden... | NEW | Summary: Proposed. The process counterpart of the Cav-beta2 interaction: AHNAK sets the amplitude and inactivation kinetics of L-type calcium current, and is required for surface delivery of L-type channels in several cell types. Reason: Loss-of-function evidence comes from four independent systems. In cardiomyocytes, AHNAK fragments alter I(CaL) density and inactivation in wild-type cells but have no effect at all in ahnak1-deficient cells (PMID:20607281). In CD4 T cells, AHNAK1 is required for plasma-membrane expression of the Cav1.1 alpha1S subunit and for calcium influx after TCR crosslinking (PMID:18191595), with the same requirement in cytotoxic CD8 T cells (PMID:19497879). In neurons, Ahnak knockout reduces alpha1 surface expression and L-type current (PMID:30760886). In osteoblastic cells, AHNAK knockdown impairs calcium influx without disassembling the Cav1.2/beta2 complex (PMID:19261907), which locates AHNAK's contribution downstream of complex assembly. All of these are knockout or knockdown experiments in rodent cells. Evidence code ISS, not IMP: the perturbations are all on the mouse or rat ortholog, and an experimental code on the human gene would assert an experiment nobody has done. UniProtKB:E9Q616 (mouse Ahnak) is given as the donor. This is the same discipline applied to the GOA rows elsewhere in this review, and the first version of this row failed it. Supporting Evidence: PMID:20607281 By contrast, neither 188-PSTP nor 188-ASTA affected any I (CaL) characteristics in ahnak1-deficient mouse cardiomyocytes. PMID:18191595 We found that AHNAK1 is required for plasma membrane expression of L-type calcium channels |
| GO:0003779 actin binding | IDA PMID:12153988 The carboxyl-terminal region of ahnak provides a link betwee... | NEW | Summary: Proposed. The AHNAK C-terminal region binds G-actin and cosediments with F-actin. GOA gives this gene the actin cytoskeleton location but no actin-binding function. Reason: Measured directly with purified components: the same C-terminal AHNAK fragment that binds Cav-beta2 also bound G-actin and cosedimented with F-actin (PMID:12153988), and a C-terminal fragment added to actin filaments bundled them into paracrystalline arrays visible by electron microscopy (PMID:15001564). In cells, disrupting actin with cytochalasin D redistributes AHNAK (PMID:14699089). The gene already carries GO:0015629 actin cytoskeleton by IDA, so GO records where AHNAK sits relative to actin but not that it binds actin - the molecular function that makes the localisation mean something. GO:0051015 actin filament binding would also be defensible on the cosedimentation and bundling data; GO:0003779 is proposed because G-actin binding was shown as well and GO:0003779 covers both. Supporting Evidence: PMID:12153988 The same carboxyl-terminal ahnak region (aa 5262-5643) bound to G-actin and cosedimented with F-actin. PMID:15001564 Addition of ahnak-C2 to actin filaments induced filament bundling into paracrystalline-like structures as revealed by electron microscopy. |
| GO:0051017 actin filament bundle assembly | IDA PMID:15001564 The carboxyl-terminal ahnak domain induces actin bundling an... | NEW | Summary: Proposed. A recombinant AHNAK C-terminal fragment bundles purified actin filaments into paracrystalline arrays. Reason: A reconstitution with purified proteins, read out by electron microscopy, and with a physiological correlate: the same fragment attenuated the decline in isometric force over repeated contraction-relaxation cycles in demembranated muscle fibres, and an endogenous 72 kDa C-terminal AHNAK fragment copurifies with human cardiac myofibrillar proteins and localises to intercalated discs and the Z-line. Bundling is a specific, assayable process that GO:0015629 actin cytoskeleton does not imply. Supporting Evidence: PMID:15001564 Addition of ahnak-C2 to actin filaments induced filament bundling into paracrystalline-like structures as revealed by electron microscopy. PMID:15001564 In the human heart, we observed an endogenous carboxyl-terminal 72 kDa ahnak fragment that copurified with myofibrillar proteins. |
| GO:0030674 protein-macromolecule adaptor activity | IDA PMID:12153988 The carboxyl-terminal region of ahnak provides a link betwee... | NEW | Summary: Proposed. The molecular function the whole AHNAK literature converges on: a scaffold that holds separate partners together so they act as a unit. GOA currently gives this gene no molecular function beyond protein binding, cadherin binding, S100 protein binding, RNA binding and a disputed elasticity term. Reason: GO:0030674 is defined as 'An adaptor activity that brings together two or more macromolecules in contact, permitting those molecules to function in a coordinated way', with the definition explicitly covering both two proteins and a protein plus a lipid. The generic parent GO:0060090 molecular adaptor activity would also be true, but the child is supported and more informative, and there is no heterogeneous-donor argument for staying broad here because this is a direct measurement on the protein itself. AHNAK satisfies that definition with measured, mapped, simultaneous binding sites rather than by analogy. One and the same C-terminal region binds the Cav-beta2 channel subunit with nanomolar affinity and binds G- and F-actin, which is precisely a link between two otherwise unconnected partners (PMID:12153988); a defined 20-residue motif in that region binds the annexin A2/S100A10 heterotetramer, with the crystal structure showing both protomers contacted at once and neither sufficient alone (PMID:22940583, PMID:16984913); and knocking down annexin 2/S100A10 removes AHNAK from the plasma membrane while knocking down AHNAK leaves that complex in place but blocks cortical actin reorganisation (PMID:14699089). The protein has no catalytic activity, no enzyme motif, and a feature table that is one PDZ domain plus 26-30 tandem 128-residue repeats and long disordered stretches - adaptor activity is not a fallback here, it is the architecture. Supporting Evidence: PMID:12153988 These results suggest that ahnak provides a structural basis for the subsarcolemmal cytoarchitecture and confers the regulatory role of the actin-based cytoskeleton to the L-type Ca2+ channel. PMID:14699089 The S100A10 subunit serves to mediate the interaction between annexin 2 and the COOH-terminal regulatory domain of AHNAK. |
| GO:0030866 cortical actin cytoskeleton organization | ISS PMID:14699089 AHNAK interaction with the annexin 2/S100A10 complex regulat... | NEW | Summary: Proposed. AHNAK knockdown prevents the cortical actin reorganisation that establishes epithelial cell height. Reason: The central loss-of-function result of the paper that GOA already cites twice on this gene, and the process the S100A10/annexin A2 and actin interactions exist to serve. Reciprocal knockdowns separate the two arms: annexin 2 siRNA removes AHNAK from the plasma membrane, while AHNAK siRNA leaves the annexin complex in place and blocks the cortical actin change. Evidence code ISS, not IMP. The AHNAK knockdown and the cortical-actin phenotype are both in MDCK cells, which are canine - only the annexin 2 arm used human material - so an experimental code on human AHNAK would assert a perturbation of the human gene that was never performed. UniProtKB:A0A8I3NCQ4 (dog AHNAK, TrEMBL, 6134 aa) is given as the donor. An earlier version of this row read 'hence IMP' about the canine knockdowns, which inverted the convention exactly backwards: a non-human perturbation is the reason the code is not IMP. Supporting Evidence: PMID:14699089 AHNAK interaction with the annexin 2/S100A10 complex regulates cell membrane cytoarchitecture. PMID:14699089 The interaction of the membrane-bound form of AHNAK with the cortical actin cytoskeleton was confirmed by indirect immunofluorescence analysis of MDCK cells treated with cytochalasin D |
| GO:0032460 negative regulation of protein oligomerization | IMP PMID:33961796 AHNAK controls 53BP1-mediated p53 response by restraining 53... | NEW | Summary: Proposed. AHNAK binds the 53BP1 oligomerization domain and restrains 53BP1 multimerisation and phase separation, tuning the p53 DNA-damage response. Reason: A human loss-of-function study in AHNAK-knockout U2OS cells: losing AHNAK makes 53BP1 hyper-accumulate on chromatin, enlarges 53BP1 nuclear bodies, enhances phase separation in an optoDroplet assay, and elevates the p53 response to the point of senescence in non-transformed cells. The interaction is domain-mapped, since the 53BP1 oligomerization domain is required for stable association with AHNAK. This gives AHNAK's nuclear pool a concrete function, which GOA does not record at all: TP53BP1 is absent from AHNAK's GO:0005515 set and no DNA-damage, p53 or senescence term appears anywhere on the gene. Supporting Evidence: PMID:33961796 We demonstrate that AHNAK binds to the 53BP1 oligomerization domain and controls its multimerization potential. PMID:33961796 Loss of AHNAK results in hyper-accumulation of 53BP1 on chromatin and enhanced phase separation, culminating in an elevated p53 response |
| GO:0008047 enzyme activator activity | IDA PMID:15177040 AHNAK interacts with the DNA ligase IV-XRCC4 complex and sti... | NEW | Summary: Proposed. AHNAK binds the DNA ligase IV-XRCC4 complex and increases its double-stranded ligation activity. The only nuclear function GOA records for this gene is regulation of RNA splicing, which this review shows is one experiment on a short isoform. Reason: Measured with purified proteins and corroborated in cells: AHNAK interacts specifically with the LIG4-XRCC4 complex - not with other DNA ligases or other non-homologous end-joining components - the two co-immunoprecipitate from cells, and AHNAK stimulates the double-stranded ligation activity of the complex. GO:0008047 is defined as 'A molecular function regulator that increases a catalytic activity', which is exactly what was assayed. Two narrower terms were considered and rejected: GO:0051106 positive regulation of DNA ligation is obsolete (checked at QuickGO), and involved_in GO:0006303 double-strand break repair via nonhomologous end joining would over-reach, because no cellular repair assay was performed - the evidence is an in-vitro ligation stimulation plus a co-immunoprecipitation, so the activity is annotated and the pathway role is left as a knowledge gap. The alternative mechanism has to be met rather than ignored, and the paper raises it itself in the sentence immediately after the DNA-binding one: AHNAK links two DNA molecules to about the same extent as Ku. If the stimulation worked purely by juxtaposing substrate ends, that would be substrate presentation rather than activation of the ligase, and the adaptor framing used elsewhere in this review would fit better than an enzyme-activator MF. Two things keep GO:0008047 as the better call without settling the mechanism. The binding is selective - specifically for LIG4-XRCC4 and not for other DNA ligases or other non-homologous end-joining components - which a generic DNA-bridging activity does not predict; and AHNAK forms a stable complex with LIG4-XRCC4 on DNA, so it is a component of the ligating assembly rather than a tether acting at a distance. A tempting precedent was checked and discarded: Ku80 (P13010) does carry GO:0008047, but by TAS from a telomerase paper (PMID:19188702), so it is not evidence about how GO treats end-bridging stimulation of ligation. The mechanism is recorded as an open question rather than asserted. This row also completes the review's own argument: PMID:15177040 was already cited for AHNAK's weak DNA binding, and a purified-protein enzymatic result on this gene supporting no annotation is precisely the coverage defect this review describes. Supporting Evidence: PMID:15177040 Using immunoaffinity purification or purified proteins, we show that AHNAK interacts specifically with the DNA ligase IV-XRCC4 complex, a complex that functions in DNA non-homologous end-joining. PMID:15177040 We characterised AHNAK as a protein that stimulates the double-stranded (DS) ligation activity of DNA ligase IV-XRCC4. PMID:15177040 We show that AHNAK has weak DNA-binding activity and forms a stable complex with the DNA ligase IV-XRCC4 complex on DNA. PMID:15177040 AHNAK is also able to link two DNA molecules to a similar extent to that previously reported for Ku. |
| GO:0005080 protein kinase C binding | ISS PMID:18174170 Ahnak protein activates protein kinase C (PKC) through disso... | NEW | Summary: Proposed. AHNAK's central repeated units bind PKC-alpha and displace it from its inhibitory complex with protein phosphatase 2A. Reason: Two papers from the same group establish it with recombinant AHNAK repeat-unit fragments: a four-CRU derivative binds and activates PKC-alpha independently of phosphatidylserine and diacylglycerol (PMID:18174170), building on the earlier demonstration that the CRUs scaffold PLC-gamma1 and PKC together (PMID:15033986). The loss-of-function control is informative: Ahnak-null MEFs show enhanced PKC-PP2A complex formation and reduced PKC-alpha membrane translocation, and adding Ahnak back abolishes the complex. This assigns a function to the central repeat region, which is roughly three quarters of the protein and currently carries no GO annotation of any kind. Evidence code ISS, not IPI: the study is mouse on both sides - NIH3T3 cells and Ahnak-null MEFs - so the partner is mouse Prkca (P20444) rather than human PRKCA (P17252), and the donor given is mouse Ahnak. The first version of this row asserted IPI against the human PRKCA accession, which claimed a human interaction measurement that does not exist. The term proposed is the binding activity rather than GO:0030295 protein kinase activator activity, because the assays used isolated repeat fragments. Supporting Evidence: PMID:18174170 Here, we demonstrate that an Ahnak derivative consisting of four central repeated units binds and activates PKC-alpha in a phosphatidylserine/1,2-dioleoyl-sn-glycerol-independent manner. PMID:18174170 Consistently, Ahnak null MEF cells stimulated by PMA showed enhanced PKC-PP2A complex formation, and add-back expression of Ahnak into Ahnak null MEF cells abolished the PKC-PP2A complex formation in response to PMA. |
Loading supporting content…
Download this section (compressed HTML)Q: For ComplexPortal: GO:1905686 positive regulation of plasma membrane repair is asserted by NAS on ten entities - CPX-850, CPX-853, CPX-898, CPX-905, and ANXA2, S100A10 and AHNAK in human and mouse - all from PMID:22940583, a 3.5 A crystal structure that performs no membrane-repair experiment. Meanwhile the only published functional test, a laser-wounding assay on AHNAK1-deficient fibres, found no requirement (PMID:20833135). Should the projection be restricted to the complexes, or to the subunits for which repair has actually been measured?
Suggested experts: ComplexPortal curator, muscle membrane repair
Q: For ComplexPortal: the GO:0005886 IDA on AHNAK cites PMID:3126079, a 1988 paper on p11 and p36 that predates the 1992 report of the AHNAK gene (PMID:1608957) and contains no AHNAK experiment. The term is correct for AHNAK on other evidence, so this is a provenance problem rather than a wrong annotation - but a complex-to-subunit projection should probably not place a subunit in a paper that could not have studied it. Can projections be restricted to references whose date postdates the subunit's identification?
Suggested experts: ComplexPortal curator
Q: For UniProt/CAFA: five ISS rows on AHNAK (GO:0005634, GO:0005829, GO:0043034, GO:0043484, GO:0097493) were assigned as one block, all citing PMID:21940993 and all using mouse Ahnak E9Q616 as donor. For two of them the donor's own evidence lies in a different paper - GO:0043034 and GO:0097493 are both IDA/IMP from PMID:20833135, and PMID:21940993 contains neither experiment. Should block-assigned ISS rows carry the reference that supports the specific term rather than a single reference for the block?
Suggested experts: UniProt curator, CAFA
Q: For PAINT/GO Central: PANTHER node PTN001156025 assigns GO:0043484 regulation of RNA splicing to 74 gene products across metazoa, with human reach exactly AHNAK, AHNAK2 and PRX. The entire experimental basis is one paper (PMID:21940993), and within it the activity belongs to short nuclear isoforms of AHNAK and Periaxin - for AHNAK that is isoform Q09666-2, essentially the PDZ domain alone. Should the node term be restricted, or at least recorded as isoform-specific? A separate consequence: the paper's Periaxin result is experimental, yet human, mouse and rat Periaxin all hold GO:0043484 only by IBA or ISO, so an experimental annotation is missing where the data exist.
Suggested experts: PAINT curator, RNA splicing
Q: Should GO:0097493 structural molecule activity conferring elasticity be retained for AHNAK? Its whole basis is that AHNAK1-knockout mouse muscle fibres are stiffer by AFM (PMID:20833135) - a tissue-level phenotype from which a molecular elastomeric activity was inferred. No force spectroscopy, extensibility measurement or reconstitution has ever been done on AHNAK or its central repeats. If the term is kept, is there a case for an 'inferred from organ mechanics' caveat that distinguishes it from titin and elastin, where the molecular property was measured?
Suggested experts: muscle mechanics, GO ontology editor
Q: Does GO need a term for binding a ferlin family protein, or more generally a membrane-repair C2-domain protein? AHNAK's direct, domain-mapped, Ca-independent interactions with dysferlin and myoferlin (PMID:17185750) are among the best-supported on the gene, yet the only available term is bare GO:0005515 protein binding, so three of the most informative IPI rows on AHNAK are indistinguishable from screen hits.
Suggested experts: GO ontology editor, muscle cell biology
Q: For UniProt: the FUNCTION line of Q09666 reads in full 'May be required for neuronal cell differentiation', which reflects the 1992 cloning paper rather than the thirty years since. The same entry's own reference list and its INTERACTION block already point at the Cav-beta2, annexin A2/S100A10, dysferlin and actin interactions. Could the FUNCTION comment be updated, and the SUBCELLULAR LOCATION line - currently just 'Nucleus' - extended to the plasma membrane and sarcolemma that four IDAs on the same protein support?
Suggested experts: UniProt curator
Q: Is the T-tubule localisation of AHNAK tissue-specific? PMID:17185750 reports sarcolemma-plus-T-tubule co-localisation with dysferlin in skeletal muscle; PMID:20833135, using AHNAK1- and AHNAK2-specific antibodies, reports no AHNAK in the skeletal T-tubule system; and PMID:12153988 localises the AHNAK C-terminal portion to the T-tubular system of human left ventricle. Cardiac-versus-skeletal difference and AHNAK1-versus-AHNAK2 antibody cross-reactivity are both plausible explanations and have not been distinguished.
Suggested experts: muscle cell biology
Experiment: Settle the membrane-repair question properly. Repeat the laser-wounding assay on AHNAK-knockout human myotubes and on a non-muscle cell type, with FM1-43 dye entry quantified against dysferlin- and annexin A2-knockout controls in the same experiment. The existing negative is a single assay in mouse skeletal fibres; the existing positive annotation rests on no assay at all.
Experiment: Measure AHNAK's mechanical properties directly rather than inferring them from fibre stiffness. Single-molecule force spectroscopy on a recombinant construct of several tandem 128-residue central repeats would establish whether the repeat region is elastomeric, which is the claim GO:0097493 makes and which has never been tested.
Experiment: Establish which interactions of the full-length protein are real. All eight of the holdup GO:0005515 rows measure AHNAK's isolated PDZ domain against a synthetic decapeptide. Endogenous AHNAK immunoprecipitation followed by quantitative mass spectrometry, with an AHNAK-knockout line as the specificity control, would separate real partners from the 608 accumulated IntAct associations of a very large, very abundant cortical protein.
Experiment: Test whether the splicing-regulation activity is isoform-restricted, as the annotation implicitly assumes. Express the short isoform Q09666-2 and the full-length protein separately in AHNAK-null cells and assay splicing at the endogenous AHNAK locus; the current GO record attaches the activity to the canonical 700 kDa accession.
Experiment: Perturb AHNAK in a human excitable cell and measure L-type current. Knockout or acute degron depletion in human iPSC-derived cardiomyocytes, with whole-cell I(CaL) recorded before and after beta-adrenergic stimulation, would convert the strongest claim about AHNAK from a cross-species inference into a human measurement.
Experiment: Ask whether AHNAK's 53BP1 activity is separable from its cortical one. A C-terminal truncation that removes the annexin A2/S100A10 and actin-binding region, or a nuclear export signal mutant that traps AHNAK in the nucleus, expressed in AHNAK-knockout U2OS cells, would show whether restraining 53BP1 requires membrane-derived regulation or is an independent activity of the nuclear pool.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)