MAGI1 is a large, nonenzymatic membrane-associated guanylate kinase-family scaffold containing six PDZ domains, two WW domains, and a guanylate-kinase-like domain. It concentrates at epithelial tight and adherens junctions and other cell-cell contacts, where it recruits adhesion receptors, signaling regulators, and actin-associated proteins into multiprotein assemblies. Through these complexes MAGI1 couples junctions to the actin cytoskeleton, promotes mature cell-cell adhesion and tight-junction integrity, and modulates junctional RhoA, Rap, and Wnt/beta-catenin signaling. Tissue- and isoform-dependent roles also occur at podocyte slit-diaphragm and neuronal postsynaptic compartments. High-risk papillomavirus E6 proteins bind MAGI1 PDZ domains and can drive MAGI1 degradation, thereby disrupting tight junctions.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005737 cytoplasm | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Phylogenetic inference places MAGI1 activity in the cytoplasm. Human studies detect cytoplasmic pools, but the best-defined core activity is concentrated at junctional membrane sites. Reason: The location is plausible and experimentally compatible with MAGI1 trafficking and cytoplasmic interaction partners, but it is too broad to define the core junctional scaffold function. |
| GO:0005911 cell-cell junction | IBA GO_REF:0000033 | ACCEPT | Summary: Cell-cell junction is the principal site of MAGI1 scaffold activity. Human experiments place MAGI1 at epithelial and endothelial contacts and show that junctional recruitment promotes actin polymerization and adhesion. Reason: The phylogenetic annotation is strongly corroborated by direct human localization and functional evidence (PMID:9647739; PMID:20298433). Supporting Evidence: PMID:9647739 both products were co-localized at the cytoplasmic membrane PMID:20298433 Overexpression of ESAM recruited MAGI-1 to the cell-cell contact area. |
| GO:0007165 signal transduction | IBA GO_REF:0000033 | ACCEPT | Summary: MAGI1 scaffolds RAPGEF2/Rap GEP at tight junctions and ESAM-dependent MAGI1 recruitment activates RhoA during junction maturation. Reason: Although broad, signal transduction accurately captures a core consequence of MAGI1 molecular-adaptor activity and is supported by two independent human junctional signaling studies. Supporting Evidence: PMID:11168587 serves as a scaffolding molecule for Rap GEP at tight junctions PMID:20298433 Interaction of ESAM with MAGI-1 activated RhoA |
| GO:0005737 cytoplasm | IEA GO_REF:0000117 | KEEP AS NON CORE | Summary: The ARBA prediction is compatible with experimental cytoplasmic MAGI1 pools, including cytoplasmic extensions in podocytes. Reason: Retain as a broad secondary location; the active junctional pools are more informative for the core function. |
| GO:0005886 plasma membrane | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location mapping places MAGI1 at the plasma membrane, consistent with direct localization at membrane-associated cell contacts. Reason: MAGI1 is a peripheral membrane scaffold recruited to the cytoplasmic face of junctional plasma membrane domains (PMID:9647739; PMID:20298433). |
| GO:0005911 cell-cell junction | IEA GO_REF:0000117 | ACCEPT | Summary: ARBA predicts the experimentally established cell-cell-junction localization. Reason: Direct human experiments independently validate this location (PMID:9647739; PMID:20298433). |
| GO:0005923 bicellular tight junction | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt mapping assigns MAGI1 to bicellular tight junctions, the best-defined epithelial site of its scaffold activity. Reason: MAGI1 is directly localized to tight junctions, and MAGI1 loss disrupts tight- junction integrity (PMID:11168587; PMID:21123374). Supporting Evidence: PMID:21123374 MAGI-1 degradation is a loss of tight-junction integrity |
| GO:0016020 membrane | IEA GO_REF:0000117 | MODIFY | Summary: The generic membrane prediction captures a real feature but loses the strong evidence for junctional plasma-membrane localization. Reason: Replace the overly broad membrane term with plasma membrane and bicellular tight junction, both already supported by direct and curated evidence. Proposed replacements: plasma membrane bicellular tight junction |
| GO:0005515 protein binding | IPI PMID:11168587 Membrane-associated guanylate kinase with inverted orientati... | MODIFY | Summary: MAGI1 interacts with RAPGEF2/Rap GEP and functions as its tight-junction scaffold. The experiment supports an adaptor activity more informative than generic protein binding. Reason: MAGI1's defining molecular role is to assemble interacting proteins into junctional complexes. Molecular adaptor activity captures this causal role without treating MAGI1 as an enzyme. Proposed replacements: molecular adaptor activity Supporting Evidence: PMID:11168587 MAGI-1/BAP1 serves as a scaffolding molecule for Rap GEP file:human/MAGI1/MAGI1-deep-research-manual.md The best-supported core role is organization of cell-cell junction complexes. |
| GO:0005515 protein binding | IPI PMID:12097473 Association of the kinesin superfamily motor protein KIF1Bal... | UNDECIDED | Summary: The abstract reports KIF1Balpha binding to the neuronal scaffold S-SCAM, but the cached text does not resolve the accession or older MAGI/S-SCAM nomenclature sufficiently to verify that the assayed product was human MAGI1 Q96QZ7. Reason: The full evidence needed to resolve the exact MAGI family member is unavailable. Per curation policy, the experimental annotation is not rejected from an abstract-level nomenclature ambiguity. |
| GO:0005515 protein binding | IPI PMID:12444549 Chimaeric HPV E6 proteins allow dissection of the proteolyti... | MARK AS OVER ANNOTATED | Summary: Papillomavirus E6 proteins carrying PDZ-binding sequences bind MAGI1 and may target it for degradation. Reason: The interaction is credible, but generic protein binding is uninformative for MAGI1's normal molecular function and the viral interaction is a perturbation rather than a core activity. Supporting Evidence: PMID:12444549 enables them to bind Dlg and a second MAGUK family member, MAGI-1 |
| GO:0005515 protein binding | IPI PMID:14627983 Carom: a novel membrane-associated guanylate kinase-interact... | MARK AS OVER ANNOTATED | Summary: FCHSD2/Carom binds the fifth PDZ domain of MAGI1 in vitro and transiently overlaps MAGI1 before epithelial junction maturation. Reason: The domain-specific interaction supports the scaffold model, but the generic protein-binding term adds no informative molecular function. Supporting Evidence: PMID:14627983 Carom binds to the fifth PDZ domain of MAGI-1 |
| GO:0005515 protein binding | IPI PMID:16316992 Proteomic analysis of beta1-adrenergic receptor interactions... | UNDECIDED | Summary: The cached abstract describes beta1-adrenergic receptor interactions with MAGI2 and MAGI3 but does not identify MAGI1; the MAGI1 evidence may be confined to the proteomic array or supplementary material. Reason: Without the accession-level result, the specific MAGI1 interaction cannot be verified. The experimental annotation is retained without alleging curator or paralog error. |
| GO:0005515 protein binding | IPI PMID:17267502 Structures of a human papillomavirus (HPV) E6 polypeptide bo... | MARK AS OVER ANNOTATED | Summary: Crystal structures establish direct recognition of an HPV18 E6 peptide by MAGI1 PDZ domains. Reason: The structural interaction is well supported but describes viral hijacking; generic protein binding should not be treated as the core MAGI1 function. Supporting Evidence: PMID:17267502 peptide from HPV18 E6 bound to three PDZ domains from MAGI-1 |
| GO:0005515 protein binding | IPI PMID:22662192 IQGAP1 interacts with components of the slit diaphragm compl... | MARK AS OVER ANNOTATED | Summary: Co-immunoprecipitation and proximity assays detect MAGI1 association with IQGAP1 in cultured human podocytes. Reason: This supports a podocyte junctional complex context, but generic protein binding is not an informative molecular-function annotation for MAGI1. Supporting Evidence: PMID:22662192 interactions between IQGAP1 and SD proteins (nephrin, MAGI-1 |
| GO:0005515 protein binding | IPI PMID:22898364 Comparative analysis of virus-host interactomes with a mamma... | MARK AS OVER ANNOTATED | Summary: This large-scale virus-host interaction study includes papillomavirus E6-MAGI1 pairs in its interaction dataset. Reason: A high-throughput viral interaction does not define MAGI1's physiological molecular activity, and protein binding is too generic to retain as core. |
| GO:0005515 protein binding | IPI PMID:23873930 Cellular localization and characterization of cytosolic bind... | MARK AS OVER ANNOTATED | Summary: WW-domain screening and follow-up assays identify PRRG4 as a MAGI1-interacting membrane protein. Reason: The interaction is compatible with multivalent adaptor activity, but generic protein binding does not describe the biological function of the complex. |
| GO:0005515 protein binding | IPI PMID:30126976 Proteome-wide analysis of phospho-regulated PDZ domain inter... | MARK AS OVER ANNOTATED | Summary: Proteome-wide peptide-phage screening maps PDZ-domain interactions and their phosphoregulation, including MAGI1 domain interactions in the dataset. Reason: The assay is valuable interaction evidence but generic protein binding is neither specific nor a core functional description. |
| GO:0005515 protein binding | IPI PMID:33208464 The ATPase ATP6V1A facilitates rabies virus replication by p... | UNDECIDED | Summary: MAGI1 appears in a rabies-virus matrix-protein interactome, but the cached article's validated mechanistic work focuses on ATP6V1A and does not expose the accession-level MAGI1 result in the main text. Reason: The exact MAGI1 evidence cannot be verified from the available article text; the experimental interaction is therefore retained without a confident functional interpretation. |
| GO:0005515 protein binding | IPI PMID:33864728 Host PDZ-containing proteins targeted by SARS-CoV-2. | UNDECIDED | Summary: A SARS-CoV-2 PDZome affinity screen is cited for MAGI1, but the abstract's named high-confidence binders do not include MAGI1 and the cache lacks the supplementary accession-level table. Reason: The specific MAGI1 interaction cannot be confirmed from the accessible abstract-only record. It is not removed because the curator may have used supplementary data unavailable here. |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MARK AS OVER ANNOTATED | Summary: BioPlex affinity-purification networks report MAGI1 interaction partners in cell-line-specific, proteome-scale datasets. Reason: The interactions may be valid, but a broad protein-binding label derived from a proteome-scale network does not identify MAGI1's molecular activity. |
| GO:0005515 protein binding | IPI PMID:35271311 OpenCell: Endogenous tagging for the cartography of human ce... | MARK AS OVER ANNOTATED | Summary: OpenCell combines endogenous tagging, microscopy, and interaction proteomics to map cellular organization; the MAGI1 pair is a high-throughput dataset result. Reason: This is useful interaction context but does not justify retaining generic protein binding as a defining MAGI1 molecular function. |
| GO:0005515 protein binding | IPI PMID:36012204 Differential CFTR-Interactome Proximity Labeling Procedures ... | MARK AS OVER ANNOTATED | Summary: MAGI1 was recovered in CFTR proximity-labeling datasets. Reason: Proximity labeling does not by itself establish direct binding or a MAGI1 function, and the generic protein-binding term overstates the functional information supplied by the assay. |
| GO:0005515 protein binding | IPI PMID:36115835 Quantitative fragmentomics allow affinity mapping of interac... | MARK AS OVER ANNOTATED | Summary: Quantitative fragmentomics measured affinities for hundreds of MAGI1 PDZ-domain interactions with PDZ-binding motifs, providing extensive biochemical binding evidence. Reason: The dataset strongly supports MAGI1's multivalent interaction architecture, but collapsing hundreds of motif affinities into generic protein binding is not an informative functional annotation. Supporting Evidence: PMID:36115835 65,000 interactions involving PDZ domains and their target PDZ-binding motifs |
| GO:0005515 protein binding | IPI PMID:37100772 Large-scale phage-based screening reveals extensive pan-vira... | MARK AS OVER ANNOTATED | Summary: A pan-viral phage screen identifies viral short linear motifs that engage host interaction domains, including a MAGI1-virus pair. Reason: Viral motif mimicry is a perturbational interaction context, and generic protein binding is not a suitable core function for MAGI1. |
| GO:0030054 cell junction | IDA GO_REF:0000052 | ACCEPT | Summary: Human Protein Atlas immunofluorescence detects MAGI1 at cell junctions. Reason: This direct localization is consistent with the independently established tight-junction and cell-cell-contact localization. |
| GO:0001886 endothelial cell morphogenesis | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Orthology-based transfer assigns MAGI1 a role in endothelial cell morphogenesis, consistent with the conserved AMOTL2-MAGI1-CDH5 junction-actin complex. Reason: The process is biologically plausible but is a tissue-specific developmental consequence of the general junctional scaffold function and lacks direct human morphogenesis evidence in the cited transfer. |
| GO:0005737 cytoplasm | IDA PMID:20298433 Interaction of endothelial cell-selective adhesion molecule ... | KEEP AS NON CORE | Summary: ESAM-transfection experiments detect a non-junctional cytoplasmic MAGI1 pool in addition to recruitment at cell contacts. Reason: Retain the experimentally observed broad pool, but the contact-associated plasma-membrane localization is the functionally informative site. |
| GO:0005886 plasma membrane | IDA PMID:20298433 Interaction of endothelial cell-selective adhesion molecule ... | ACCEPT | Summary: ESAM recruits MAGI1 to the cell membrane and cell-cell contact area during mature adhesion. Reason: The localization is direct and mechanistically linked to RhoA-dependent adhesion maturation. Supporting Evidence: PMID:20298433 ESAM-mediated MAGI-1 recruitment to the cell membrane |
| GO:0005911 cell-cell junction | IDA PMID:20298433 Interaction of endothelial cell-selective adhesion molecule ... | ACCEPT | Summary: ESAM expression recruits MAGI1 to cell-cell contacts in a PDZ-dependent junctional adhesion system. Reason: This is direct localization at the core site of MAGI1 activity. Supporting Evidence: PMID:20298433 Overexpression of ESAM recruited MAGI-1 to the cell-cell contact area. |
| GO:0022409 positive regulation of cell-cell adhesion | IDA PMID:20298433 Interaction of endothelial cell-selective adhesion molecule ... | ACCEPT | Summary: ESAM-MAGI1 colocalization promotes actin polymerization and firm cell-cell adhesion through RhoA signaling. Reason: Functional adhesion and inhibitor experiments directly establish this as a core downstream process of junctional MAGI1 scaffolding. Supporting Evidence: PMID:20298433 resulted in firm cell-cell adhesion PMID:20298433 Interaction of ESAM with MAGI-1 activated RhoA |
| GO:0071944 cell periphery | IDA PMID:20298433 Interaction of endothelial cell-selective adhesion molecule ... | KEEP AS NON CORE | Summary: MAGI1 is detected at the cell periphery after recruitment by ESAM. Reason: The localization is valid but less specific than the accepted plasma-membrane and cell-cell-junction terms from the same study. |
| GO:0005912 adherens junction | ISS GO_REF:0000024 | ACCEPT | Summary: Orthology transfer places MAGI1 at adherens junctions. Human cell studies show MAGI1 stabilizes E-cadherin and beta-catenin at cell-cell junctions. Reason: The transferred location is consistent with direct human functional evidence for adherens-junction stabilization (PMID:21666716). Supporting Evidence: PMID:21666716 stabilized E-cadherin and beta-catenin localization at cell-cell junctions |
| GO:0005515 protein binding | IPI PMID:16464232 Postsynaptic recruitment of Dendrin depends on both dendriti... | UNDECIDED | Summary: The abstract reports Dendrin interaction with MAGI/S-SCAM, but older nomenclature and the unavailable full text prevent accession-level confirmation for Q96QZ7. Reason: Retain the experimental annotation conservatively without claiming a wrong family member; the accessible evidence does not resolve the exact MAGI protein. |
| GO:0005737 cytoplasm | IDA PMID:16464232 Postsynaptic recruitment of Dendrin depends on both dendriti... | KEEP AS NON CORE | Summary: MAGI/S-SCAM retains Dendrin in the cytoplasm, but the abstract does not resolve whether the assayed scaffold corresponds specifically to MAGI1 Q96QZ7. Reason: Cytoplasmic MAGI1 is independently supported by other experimental and phylogenetic annotations, so the location can be retained as non-core even though this paper's historical MAGI/S-SCAM nomenclature is unresolved. |
| GO:0042995 cell projection | IDA PMID:16464232 Postsynaptic recruitment of Dendrin depends on both dendriti... | UNDECIDED | Summary: The study concerns postsynaptic Dendrin recruitment in neuronal projections, but the cached abstract does not provide accession-resolved MAGI1 localization. Reason: Neuronal/postsynaptic localization may be isoform-specific and the available text cannot distinguish the relevant MAGI family member. |
| GO:0051393 alpha-actinin binding | IPI PMID:16464232 Postsynaptic recruitment of Dendrin depends on both dendriti... | ACCEPT | Summary: The original abstract does not clearly establish a direct MAGI1-alpha-actinin interaction, but an independent human MAGI1 study directly maps alpha-actinin-4 binding to MAGI1's fifth PDZ domain. Reason: PMID:12042308 provides direct in-vitro binding and cellular colocalization evidence for the specific molecular-function term, rescuing the term despite limitations of the original cached abstract. Supporting Evidence: PMID:12042308 alpha-actinin-4 is also capable of binding to MAGI-1 |
| GO:0005886 plasma membrane | TAS PMID:9647739 Cloning and characterization of BAI-associated protein 1: a ... | ACCEPT | Summary: Coexpressed BAI1 and MAGI1/BAP1 colocalize at the cytoplasmic membrane. Reason: The localization is directly reported and corroborated by later human studies. Supporting Evidence: PMID:9647739 both products were co-localized at the cytoplasmic membrane |
| GO:0005911 cell-cell junction | TAS PMID:9647739 Cloning and characterization of BAI-associated protein 1: a ... | ACCEPT | Summary: BAI1 and MAGI1/BAP1 colocalize preferentially at cell-cell junctions. Reason: This foundational observation is replicated and functionally extended by later adhesion studies. Supporting Evidence: PMID:9647739 especially at cell-cell junctions |
| GO:0007155 cell adhesion | TAS PMID:9647739 Cloning and characterization of BAI-associated protein 1: a ... | ACCEPT | Summary: The original paper proposed a cell-adhesion role from BAI1-MAGI1 interaction and junctional localization; later ESAM experiments directly demonstrate that MAGI1 recruitment promotes mature cell-cell adhesion. Reason: Later direct functional evidence resolves the initially tentative TAS claim and establishes adhesion as a core process (PMID:20298433). Supporting Evidence: PMID:20298433 subsequently promote actin polymerization and mature cell-cell adhesion |
| GO:0007166 cell surface receptor signaling pathway | TAS PMID:9647739 Cloning and characterization of BAI-associated protein 1: a ... | KEEP AS NON CORE | Summary: MAGI1 binds the cytoplasmic tail of the adhesion GPCR BAI1, and the original paper suggested a signaling role, but did not directly assay BAI1 signaling. Reason: Receptor-associated signaling is plausible for a junctional adaptor, but this broad pathway term is indirect and context-specific relative to the core adhesion/junction function. |
| GO:0065003 protein-containing complex assembly | NAS PMID:9647739 Cloning and characterization of BAI-associated protein 1: a ... | ACCEPT | Summary: PDZ-dependent BAI1-MAGI1 association and the many independently mapped MAGI1 complexes support a direct role in assembling protein-containing complexes. Reason: Multivalent complex assembly is the process-level consequence of MAGI1 molecular-adaptor activity, although molecular adaptor activity is the more informative molecular-function term. Supporting Evidence: PMID:9647739 Interaction between BAI1 and BAP1 was mediated by a QTEV motif |
| GO:0005515 protein binding | IPI PMID:9647693 Atrophin-1, the DRPLA gene product, interacts with two famil... | MARK AS OVER ANNOTATED | Summary: Yeast two-hybrid and in-vitro assays identify MAGI1/AIP3 as an atrophin-1 interactor through its WW-domain-containing architecture. Reason: The interaction is credible but generic protein binding is uninformative and the disease-protein interaction is not the defining physiological MAGI1 activity. Supporting Evidence: PMID:9647693 AIP3/WWP3 are MAGUK-like multidomain proteins |
| GO:0070830 bicellular tight junction assembly | IDA PMID:21123374 A systematic analysis of human papillomavirus (HPV) E6 PDZ s... | NEW | Summary: Proposed new annotation. HPV E6-mediated MAGI1 degradation disrupts ZO-1 organization, while E6 ablation restores tight junctions only when MAGI1 is present. Reason: This loss-and-restoration experiment establishes a direct requirement for MAGI1 in forming or restoring intact bicellular tight junctions and adds a process term missing from the seeded GOA set. Supporting Evidence: PMID:21123374 Ablation of E6 expression restores tight junctions PMID:21123374 restoration is dependent on the presence of MAGI-1 |
| GO:0005634 nucleus | IDA PMID:21123374 A systematic analysis of human papillomavirus (HPV) E6 PDZ s... | NEW | Summary: Proposed new non-core localization annotation. The study directly reports preferential E6 targeting of MAGI1 in nuclear as well as membrane pools. Reason: The experimentally observed nuclear pool is absent from the seeded GOA review. It should remain non-core because its normal function is not established. Supporting Evidence: PMID:21123374 E6 preferentially targets MAGI-1 within the nucleus |
Loading supporting content…
Download this section (compressed HTML)Q: Which MAGI1 isoforms execute epithelial tight-junction, podocyte slit-diaphragm, and neuronal postsynaptic scaffold functions, and which interactions are genuinely isoform-specific rather than simply assayed with one isoform?
Suggested experts: epithelial junction biologists, podocyte biologists, neuronal MAGI scaffold specialists
Q: Does the MAGI1 guanylate-kinase-like domain bind nucleotides or partners under physiological conditions, and should ATP binding remain annotated without a demonstrated catalytic role?
Suggested experts: MAGUK structural biologists, GO molecular-function curators
Q: Which of the hundreds of PDZ-motif interactions detected by fragmentomics occur at endogenous abundance in polarized tissues and contribute to junction function?
Suggested experts: quantitative interactomics experts, epithelial polarity experts
Experiment: Create isoform-specific MAGI1 knockout/rescue panels in human epithelial cells, podocytes, and neurons; quantify junction or synapse localization, barrier resistance, actin coupling, and endogenous partner recruitment.
Hypothesis: Alternative MAGI1 termini and internal segments partition a shared molecular- adaptor activity into tissue-specific junctional and postsynaptic complexes.
Type: isoform-specific endogenous rescue and functional imaging
Experiment: Reconstitute selected endogenous MAGI1 PDZ/WW interactions at native expression levels using domain-selective mutations, proximity labeling, quantitative co-immunoprecipitation, and junction barrier assays.
Hypothesis: A small subset of high-affinity interactions, rather than the full fragmentomics interactome, is required for MAGI1-dependent tight-junction assembly and adhesion.
Type: domain-resolved interaction-to-function perturbation
Experiment: Compare purified MAGI1 guanylate-kinase-like domain with canonical guanylate kinases using nucleotide-binding, catalytic, and structural assays, followed by rescue with nucleotide-pocket mutants in polarized cells.
Hypothesis: The MAGI1 guanylate-kinase-like domain functions as a noncatalytic interaction module rather than an active guanylate kinase.
Type: biochemical activity and structure-function rescue
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)