just fetch-gene human MAGI3 retrieved 59 GOA rows, grouped by the review seeder into 23 annotation records.just fetch-gene-pmids human MAGI3 cached all nine PMID references present in the seeded GOA review.just deep-research-falcon human MAGI3 --fallback perplexity-lite) could not run: Falcon/Edison returned HTTP 402 (payment required), and Perplexity Lite returned HTTP 401 (quota exceeded). No provider-named research file was created. The evidence synthesis below was therefore performed manually from the local UniProt record, cached primary publications, GOA, and current GO definitions. A standalone synthesis is in MAGI3-deep-research-manual.md.MAGI3 is a large, soluble MAGUK-family scaffold with an N-terminal guanylate-kinase-like region, two WW domains, and six PDZ domains. The guanylate-kinase-like region is not a nucleotide kinase: a full-text MAGI3 study explicitly calls it a "catalytically inactive region of homology to the yeast guanylate kinase" PMID:27205883. This is decisive against GO:0004385 GMP kinase activity and the automatically inferred GMP/GDP metabolic-process annotations.
The founding human paper localized MAGI3 to epithelial tight junctions and showed PDZ-mediated association with PTEN. MAGI3 and PTEN cooperate to modulate AKT/PKB, consistent with MAGI3 positioning a phosphatase in a membrane-proximal signaling context [PMID:10748157 "MAGI3 and PTEN/MMAC cooperate to modulate the kinase activity of AKT/PKB"; PMID:10748157 "MAGI3 allows for the juxtaposition of PTEN/MMAC to phospholipid signaling pathways involved with cell survival"]. A second study directly localized MAGI3 at plasma-membrane, nuclear, tight-junction, adherens-junction, and focal-adhesion sites and showed that it links receptor tyrosine phosphatase beta to a substrate PMID:12615970. Together these papers support enzyme-substrate adaptor activity, cell-junction localization, and regulation of signaling, but not intrinsic phosphatase or kinase activity.
MAGI3 also organizes receptor-specific signaling through its PDZ domains. It binds beta1- and beta2-adrenergic receptors and attenuates their ERK1/2 output in expression systems [PMID:16316992 "MAGI-3 co-expression with beta1AR profoundly impaired beta1AR-mediated ERK1/2 activation"; PMID:20353789 "beta2AR-stimulated extracellular signal-regulated kinase-1/2 (ERK1/2) activation was substantially retarded by MAGI-3 expression"]. In contrast, endogenous MAGI3 supports LPA2-driven ERK and RhoA activation in SW480 cells, illustrating receptor- and cell-context dependence PMID:16904289. A mouse study further showed a FZD4-MAGI3-Ltap/Vangl2 complex that activates JNK through Rac PMID:15195140. Neutral regulation terms are preferable where MAGI3 has opposite effects in different receptor contexts.
An additional direct activity was reported in colorectal-cancer cells: MAGI3 recognizes c-Myc through PDZ5, binds SKP1 through PDZ2, and assembles a c-Myc-MAGI3-SKP1-CUL1 complex that promotes c-Myc K48 polyubiquitylation and proteasomal degradation [PMID:35864508 "these results demonstrate the existence of a macromolecular complex c-Myc-MAGI3-SKP1-CUL1"; PMID:35864508 "MAGI3 modulates c-Myc protein stability by promoting c-Myc ubiquitin–proteasome degradation"]. The molecular activity is best represented as GO:1990756 ubiquitin-like ligase-substrate adaptor activity, not ubiquitin-protein transferase activity: MAGI3 supplies substrate recognition and bridging rather than the catalytic E3 center.
Full-length MAGI3 also binds and inhibits YAP in mammary epithelial cells. A cancer-associated premature-polyadenylation product lacks the C-terminal PDZ6 domain, associates with full-length MAGI3, and acts dominantly to release YAP inhibition PMID:27205883. This is important isoform/pathology context but is not mapped onto the four normal UniProt splice isoforms in the current record.
Current Gene Ontology records were checked before use: