wu:fi36a10 (ZFIN ZDB-GENE-030131-6139, chromosome 23; NCBI and RefSeq symbol magi3a) is one of two zebrafish co-orthologs of vertebrate MAGI3; PANTHER places the duplication at the teleost genome duplication. Its paralog is magi3b. It encodes a cytoplasmic membrane-associated guanylate kinase with inverted domain order (MAGI): an N-terminal PDZ domain, a guanylate kinase-like (GK) domain that lacks the nucleotide-binding P-loop and serves as a protein-interaction module rather than an enzyme, two WW domains, five further PDZ domains and a long disordered C-terminal tail. MAGI proteins scaffold receptors and signalling proteins at the cytoplasmic face of cell-cell junctions. In a genome-wide yeast two-hybrid screen the zebrafish protein (reported as Magi3a) bound the C-terminal ETTV PDZ-binding motif of the Wnt7 co-receptor Gpr124. The gene is expressed in endothelial cells, and F0 crispants have reduced brain vascularization that phenocopies loss of that motif, as do magi3b crispants. Public RNA-Seq calls cover a broad range of embryonic and adult tissues, as for the single gar magi3 gene, and are more extensive than those of magi3b. Of the two copies it is the closer in sequence to gar and human MAGI3.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000166 nucleotide binding | IEA GO_REF:0000104 | REMOVE | Summary: Nucleotide binding from UniRule UR000400325 (MAGUK guanylate kinase-like domain). Reason: UniRule UR000400325 transfers the guanylate-kinase chemistry of the GK fold to all MAGUKs. MAGI-family GK domains are degenerate: they lack the LID and CORE subdomains, and in wu:fi36a10 (as in magi3b, gar and human MAGI3) the segment that UniProt labels as the ATP-binding site has no Walker A/P-loop motif. The GK domain of MAGIs is a phosphopeptide-binding module, not a kinase. No kinase or nucleotide-binding activity has been shown for any MAGI protein. This is a family-level IEA overextension to a pseudoenzyme domain. Supporting Evidence: PMID:37163606 Unlike other MAGUK GKs, which are composed of three subdomains including guanosine monophosphate (GMP)–binding, LID, and CORE subdomains, MAGI GKs lack the last two subdomains file:DANRE/wu_fi36a10/wu_fi36a10-bioinformatics/RESULTS.md Neither zebrafish copy, nor gar or human MAGI3, has a Walker A (P-loop) motif at the position UniProt annotates as the ATP-binding site |
| GO:0005524 ATP binding | IEA GO_REF:0000104 | REMOVE | Summary: ATP binding from UniRule UR000400325, based on the predicted P-loop of the GK-like domain. Reason: UniRule UR000400325 transfers the guanylate-kinase chemistry of the GK fold to all MAGUKs. MAGI-family GK domains are degenerate: they lack the LID and CORE subdomains, and in wu:fi36a10 (as in magi3b, gar and human MAGI3) the segment that UniProt labels as the ATP-binding site has no Walker A/P-loop motif. The GK domain of MAGIs is a phosphopeptide-binding module, not a kinase. No kinase or nucleotide-binding activity has been shown for any MAGI protein. This is a family-level IEA overextension to a pseudoenzyme domain. Supporting Evidence: PMID:37163606 Unlike other MAGUK GKs, which are composed of three subdomains including guanosine monophosphate (GMP)–binding, LID, and CORE subdomains, MAGI GKs lack the last two subdomains file:DANRE/wu_fi36a10/wu_fi36a10-bioinformatics/RESULTS.md Neither zebrafish copy, nor gar or human MAGI3, has a Walker A (P-loop) motif at the position UniProt annotates as the ATP-binding site |
| GO:0005634 nucleus | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Nucleus (UniProt subcellular location; for human MAGI3 this is 'by similarity'). Reason: Human MAGI3 is reported to concentrate in the nucleus as well as at junctions, but only by similarity. There is no zebrafish evidence. Plausible, but not where the scaffold function characterized in zebrafish acts. magi3b lacks this IEA row because of different rule coverage, not because of any known biological difference. |
| GO:0005737 cytoplasm | IBA GO_REF:0000033 | ACCEPT | Summary: Cytoplasm (IBA, PANTHER PTN000817101 MAGI node). Reason: MAGI proteins are cytoplasmic scaffolds at the inner face of the plasma membrane. wu:fi36a10 is inside the MAGI clade, keeps all its domains, and nothing argues for loss of the function. Its paralog magi3b has a direct IDA for this location. Supporting Evidence: PMID:10748157 localizes to epithelial cell tight junctions PMID:35649360 the PDZ proteins membrane-associated guanylate kinase, WW, and PDZ domain containing 3 (Magi3a) and discs large MAGUK scaffold protein 4 (Dlg4a) bound to Gpr124 via its C-terminal ETTV motif. |
| GO:0005886 plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Plasma membrane (UniProt subcellular location, peripheral membrane protein). Reason: MAGI3 is a peripheral protein on the cytoplasmic face of the plasma membrane at cell-cell junctions. In zebrafish it binds the C terminus of the transmembrane receptor Gpr124. Supporting Evidence: PMID:10748157 localizes to epithelial cell tight junctions PMID:35649360 the PDZ proteins membrane-associated guanylate kinase, WW, and PDZ domain containing 3 (Magi3a) and discs large MAGUK scaffold protein 4 (Dlg4a) bound to Gpr124 via its C-terminal ETTV motif. |
| GO:0005911 cell-cell junction | IBA GO_REF:0000033 | ACCEPT | Summary: Cell-cell junction (IBA, PANTHER PTN002569180). Reason: Junctional localization is the conserved property of MAGI1-3. wu:fi36a10 keeps the full MAGI domain set, and nothing argues for loss of the function. This is the correct level of specificity for a family-level inference. Supporting Evidence: PMID:10748157 localizes to epithelial cell tight junctions |
| GO:0005923 bicellular tight junction | IEA GO_REF:0000044 | ACCEPT | Summary: Bicellular tight junction (UniProt subcellular location). Reason: Human MAGI3 localizes to epithelial tight junctions. There is no zebrafish test of this, but the domain architecture is conserved. It is a reasonable transfer, more specific than the IBA cell-cell junction. Supporting Evidence: PMID:10748157 localizes to epithelial cell tight junctions |
| GO:0007165 signal transduction | IBA GO_REF:0000033 | ACCEPT | Summary: Signal transduction (IBA, PANTHER PTN000817101). Reason: Very general, but correct. MAGI3 scaffolds signalling complexes: PTEN and AKT in mammals. In zebrafish this copy (Magi3a) binds the Gpr124 ETTV motif in yeast two-hybrid, and its crispants phenocopy loss of that motif in Wnt7-dependent brain angiogenesis. A more specific process term would need evidence beyond F0 crispants. Supporting Evidence: PMID:10748157 cooperate to modulate the kinase activity of AKT/PKB PMID:35649360 the PDZ proteins membrane-associated guanylate kinase, WW, and PDZ domain containing 3 (Magi3a) and discs large MAGUK scaffold protein 4 (Dlg4a) bound to Gpr124 via its C-terminal ETTV motif. PMID:35649360 Targeting magi3a, magi3b, or dlg4b, but not dlg4a, partially impaired brain vascularization (Figures 4H and 4I), in agreement with the expression data. |
| GO:0016301 kinase activity | IEA GO_REF:0000104 | REMOVE | Summary: Kinase activity from UniRule UR000400325. Reason: UniRule UR000400325 transfers the guanylate-kinase chemistry of the GK fold to all MAGUKs. MAGI-family GK domains are degenerate: they lack the LID and CORE subdomains, and in wu:fi36a10 (as in magi3b, gar and human MAGI3) the segment that UniProt labels as the ATP-binding site has no Walker A/P-loop motif. The GK domain of MAGIs is a phosphopeptide-binding module, not a kinase. No kinase or nucleotide-binding activity has been shown for any MAGI protein. This is a family-level IEA overextension to a pseudoenzyme domain. Supporting Evidence: PMID:37163606 Unlike other MAGUK GKs, which are composed of three subdomains including guanosine monophosphate (GMP)–binding, LID, and CORE subdomains, MAGI GKs lack the last two subdomains file:DANRE/wu_fi36a10/wu_fi36a10-bioinformatics/RESULTS.md Neither zebrafish copy, nor gar or human MAGI3, has a Walker A (P-loop) motif at the position UniProt annotates as the ATP-binding site |
| GO:0016740 transferase activity | IEA GO_REF:0000104 | REMOVE | Summary: Transferase activity from UniRule UR000400325 (parent of kinase activity). Reason: UniRule UR000400325 transfers the guanylate-kinase chemistry of the GK fold to all MAGUKs. MAGI-family GK domains are degenerate: they lack the LID and CORE subdomains, and in wu:fi36a10 (as in magi3b, gar and human MAGI3) the segment that UniProt labels as the ATP-binding site has no Walker A/P-loop motif. The GK domain of MAGIs is a phosphopeptide-binding module, not a kinase. No kinase or nucleotide-binding activity has been shown for any MAGI protein. This is a family-level IEA overextension to a pseudoenzyme domain. Supporting Evidence: PMID:37163606 Unlike other MAGUK GKs, which are composed of three subdomains including guanosine monophosphate (GMP)–binding, LID, and CORE subdomains, MAGI GKs lack the last two subdomains file:DANRE/wu_fi36a10/wu_fi36a10-bioinformatics/RESULTS.md Neither zebrafish copy, nor gar or human MAGI3, has a Walker A (P-loop) motif at the position UniProt annotates as the ATP-binding site |
Loading supporting content…
Download this section (compressed HTML)Q: Do stable wu:fi36a10 mutants (ideally RNA-less, to avoid transcriptional adaptation of magi3b) reproduce the brain vascularization defect seen in F0 crispants?
Q: Is the broader expression of wu:fi36a10 the ancestral MAGI3 pattern, with magi3b having lost domains? Gar magi3 RNA-Seq calls resemble those of wu:fi36a10.
Q: Should the ZFIN symbol wu:fi36a10 be updated to magi3a, the symbol NCBI and RefSeq already use?
Experiment: Generate stable single, double and compound-heterozygous mutants and count hindbrain central arteries; test cross-rescue by endothelial (kdrl promoter) expression of each coding sequence.
Hypothesis: magi3b and wu:fi36a10 are both required, non-redundantly and dose-dependently, for Gpr124-dependent brain angiogenesis.
Type: genetic interaction and cross-rescue
Experiment: Single-cell or FACS-sorted RNA-seq and in situ hybridization for both paralogs in endothelium, retina and kidney, compared with gar magi3.
Hypothesis: The restricted expression of magi3b relative to wu:fi36a10 reflects expression partition, not reduced function.
Type: expression profiling
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)