FAT4 is a large transmembrane atypical cadherin that forms calcium-dependent heterophilic cell contacts with DCHS1. Its extracellular cadherin domains bind DCHS1 on neighboring cells, while its intracellular region contributes to the accumulation and internalization of these boundary complexes. FAT4-dependent cell contacts and context-dependent signaling help organize developing epithelia and neural progenitors. Its effects on YAP differ among tissues and are not restricted to a single canonical Hippo kinase mechanism.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005509 calcium ion binding | IEA GO_REF:0000002 | ACCEPT | Summary: FAT4 has a calcium-binding extracellular cadherin architecture. Reason: The actual FAT4 sequence matches the four cited cadherin/EGF InterPro records. Independently, the human FAT4 EC1–4–DCHS1 crystal structures resolve calcium ions at interdomain junctions, and purified human fragments bind heterophilically. This supports calcium binding by the extracellular protein; it does not imply that every EC linker binds calcium or that all splice products preserve the tested N-terminal region. |
| GO:0005515 protein binding | IPI PMID:22153077 Structural basis and sequence rules for substrate recognitio... | KEEP AS NON CORE | Summary: Two human FAT4 cytoplasmic peptides bind tankyrase-2 directly. Reason: PMID:22153077 validates two human FAT4 peptides by fluorescence polarization against the tankyrase-2 (TNKS2; Q9H2K2) ARC4 substrate-recognition domain (Figure S7A, Table S6 and Extended Experimental Procedures). Each FAT4 series used three experiments performed in technical duplicate. The separate Figure 6 co-immunoprecipitation and PARylation experiments used a mouse Fat4 fragment lacking most of its extracellular region, not endogenous full-length human FAT4; endogenous FAT4 was not detected in HEK293T cells. This supports the imported interaction as a non-core binding observation under the project binding exception. It does not establish an adaptor or catalytic molecular function for FAT4, and TNKS2-mediated substrate recognition does not transfer TNKS2 enzyme activity to FAT4. Supporting Evidence: |
| GO:0005886 plasma membrane | IBA GO_REF:0000033 | ACCEPT | Summary: Retain the phylogenetically inherited plasma-membrane location. Reason: The actual PANTHER v19 FAT4 leaf PTN002513196 descends from the term-specific IBD node PTN009075295 in PTHR24026. The target alignment matches canonical Q6V0I7. Human full-length FAT4 fusion proteins accumulate at cell boundaries, and rodent endogenous tissue localization and membrane topology corroborate this location. Current InterPro family PTHR24027 is a separate source snapshot, not a reason to reject the recovered PAINT ancestry. The claim is not experimentally established for every splice product. |
| GO:0005886 plasma membrane | IEA GO_REF:0000002 | ACCEPT | Summary: Retain the electronic plasma-membrane annotation. Reason: IPR020894 genuinely matches FAT4, but the generic conserved-site mapping is corroborated by full-length human FAT4 boundary localization and endogenous rodent tissue observations. These support plasma membrane as the working location of the characterized adhesion receptor; no universal location is assigned to the N-terminally shortened isoform 2. |
| GO:0005912 adherens junction | IBA GO_REF:0000033 | MODIFY | Summary: Refine the classical adherens-junction inference to the apicolateral plasma membrane. Reason: FAT4 is inside the actual PTN001931937 IBD clade, so this is a biological scope refinement, not a missing-tree or donor-count objection. In mouse embryonic cortex, endogenous Fat4 lies more apically than the beta-catenin-marked adherens junction. Knockdown and electron microscopy distinguish a Fat4-dependent subapical membrane apposition from the neighboring adherens junction. PMID:25355906 Figure 5 and the corresponding Results independently identify these appositions at the apical end of lateral epithelial membranes: transfected MDCK contacts lie above the tight junction, and mouse neuroepithelial contacts lie above the adherens junction. GO:0016327 apicolateral plasma membrane describes that membrane compartment without assigning a classical adherens-junction identity. The broader apical part of cell term is already retained at source index 15; the replacement preserves membrane-level resolution that the broader term lacks. This scoped localization inference does not deny context-dependent junctional interactions elsewhere. Propagation Review Root cause: TERM SCOPING PROBLEM Failure modes: COMPARTMENT OR COMPLEX MISMATCH Sources checked: PANTHER:PTN001931937 · PTN001931937 UNRESOLVED The actual IBD node lies above FAT4 in the recovered PTHR24026 tree. This verifies the ancestry; it does not resolve whether the classical adherens-junction compartment matches the experimentally distinct Fat4 subapical apposition. UniProtKB:Q2PZL6 · mouse Fat4 UNRESOLVED Independent endogenous cortex localization and EM distinguish Fat4 subapical apposition from beta-catenin-defined adherens junctions. This supports the proposed apicolateral-membrane refinement; it does not establish universal exclusion from all junctional contexts. Proposed replacements: apicolateral plasma membrane |
| GO:0007155 cell adhesion | IEA GO_REF:0000002 | TOO GENERAL | Summary: The broad adhesion term is covered by retained specific adhesion annotations. Reason: The IPR020894 mapping is biologically compatible with FAT4, but cell adhesion is an ancestor of the retained heterophilic cell-cell adhesion (GO:0007157, row 7) and cadherin-mediated cell-cell adhesion (GO:0044331, row 14). Human FAT4–DCHS1 assays provide the specific mechanism; retaining the broad parent adds no functional resolution. |
| GO:0007156 homophilic cell-cell adhesion | IEA GO_REF:0000002 | MODIFY | Summary: Replace the generic homophilic cadherin mapping with heterophilic adhesion. Reason: The InterPro IPR002126 match is real, but its homophilic-adhesion consequence does not describe the characterized FAT4 interaction. Human full-length FAT4 and DCHS1 accumulate at heterotypic boundaries, with no corresponding homotypic accumulation in the tested co-cultures; purified human N-terminal fragments bind directly. The earlier full-length aggregation assay used rat constructs and showed efficient mixed-cell, calcium-dependent aggregation with little aggregation of either single transfectant. These target/ortholog experiments support heterophilic adhesion, not a universal claim that FAT4 can never self-associate under any condition. The replacement GO:0007157 is already retained at source index 7; MODIFY identifies the supported heterophilic interpretation of this imported homophilic mapping rather than adding another annotation. Proposed replacements: heterophilic cell-cell adhesion |
| GO:0007157 heterophilic cell-cell adhesion | ISS GO_REF:0000024 | ACCEPT | Summary: Retain heterophilic adhesion mediated by FAT4–DCHS1. Reason: The cited mouse Q2PZL6 has a genuine experimental GOA link to PMID:19506035. Its full text separates mouse endogenous cortex work from rat full-length cDNA aggregation assays, so the exact construct species must not be relabeled as mouse. Independent human full-length co-culture experiments and purified human FAT4–DCHS1 structural/ITC measurements directly corroborate the conserved heterophilic mechanism. This is not inferred merely from disease association or classical cadherin family membership. |
| GO:0007409 axonogenesis | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Retain axonogenesis as a qualified ancestral developmental inference. Reason: The recovered FAT4 target is genuinely below the axonogenesis IBD at PTN008390614. The named donor is fly shotgun/DE-cadherin (Q24298), whose IMP annotation is grounded in PMID:12716940, not fly Fat. Accessible original Results describe perturbations of secondary axon trajectories after neuron/glia adhesion disruption; they allow secondary distortion from altered cortical cell placement. This is a considered ancestral inference, not direct human FAT4 axon-outgrowth evidence. FAT4 cortical-neuron generation experiments do not by themselves establish axonogenesis, but neither they nor the different adhesion partner establish a target-specific loss of the inherited developmental role. Retain the IBA as non-core with that uncertainty rather than reject it from donor count or paralog naming alone. |
| GO:0016020 membrane | IEA GO_REF:0000120 | TOO GENERAL | Summary: Membrane is covered by the retained plasma-membrane annotations. Reason: The two actual InterPro matches and the UniProt subcellular-location mapping are compatible with FAT4. GO:0016020 is an ancestor of GO:0005886, retained in rows 2 and 3 and supported by characterized full-length protein localization. The broad membrane term adds no separate location; this does not remove or overwrite its original combined-source provenance. |
| GO:0021987 cerebral cortex development | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Retain the conserved cerebral-cortex developmental role. Reason: The mouse Q2PZL6 donor has a genuine IMP annotation to PMID:19506035. Its mouse in-utero knockdown, endogenous immunolocalization and ultrastructure show that Fat4 maintains neuroepithelial subapical membrane organization. The later cortical knockdown/rescue experiments corroborate a role in progenitor differentiation and neuron placement. This is a tissue-specific developmental consequence of the receptor/adhesion system, not a separate molecular activity or proof of an identical mechanism in all human cell types. |
| GO:0022008 neurogenesis | ISS GO_REF:0000024 | TOO GENERAL | Summary: Neurogenesis is covered by the more specific retained generation-of-neurons annotation. Reason: Mouse Fat4 perturbation and Yap co-knockdown support the neuronal-generation/differentiation context in PMID:24056717, including the accessible original supplement. GO:0022008 is an ancestor of GO:0048699, retained at row 16. The latter preserves the supported developmental scope; neither term is treated as direct evidence of axon extension. |
| GO:0030855 epithelial cell differentiation | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Retain epithelial differentiation as a conserved developmental role. Reason: FAT4 descends from the actual epithelial-differentiation IBD at PTN008601578. The later IBD snapshot adds a Dachsous donor relative to the frozen GOA list; both source versions are preserved. Mammalian corroboration comes from Fat4-null nephron progenitors and co-culture with Fat4-expressing cells: extracellular Fat4 signaling promotes an epithelial differentiation program, including E-cadherin, rather than merely being consumed by differentiation. This kidney stromal–progenitor mechanism is distinct from neuronal differentiation out of neuroepithelium and is retained as non-core. The identity of the neighboring receptor and all downstream kinase steps were not resolved by that study. |
| GO:0035329 hippo signaling | ISS GO_REF:0000024 | KEEP AS NON CORE | Summary: Retain context-dependent Hippo/YAP signaling with explicit mechanistic limits. Reason: Mouse Q2PZL6 carries an IMP source annotation to PMID:24056717. The accessible original supplement shows reduced phospho-Yap after Fat4 knockdown in P19 cells (Figure S7), and cortical Yap co-knockdown rescues proliferation/differentiation defects (Figures S4 and S8). Kidney experiments independently link Fat4 to Yap phosphorylation and localization. These are positive signaling results sufficient to retain this broad, tissue-dependent annotation; they do not identify every MST/LATS step. In contrast, the heart study finds an AMOTL1-dependent route that bypasses altered Hippo-kinase/Yap phosphorylation, and liver experiments do not show the fly-like growth phenotype. No universal fly Fat–Dachs mechanism or universal absence of Hippo participation is inferred. The main body of PMID:24056717 remains inaccessible; its cited supplement experiments were read directly. |
| GO:0044331 cell-cell adhesion mediated by cadherin | IBA GO_REF:0000033 | ACCEPT | Summary: Retain cadherin-mediated cell-cell adhesion. Reason: The actual target descends from the cadherin-adhesion IBD at PTN001931937. The newer IBD donor list and frozen GOA list differ by one added Dachsous donor, which is a snapshot distinction rather than a provenance rewrite. Human full-length FAT4–DCHS1 boundary accumulation, direct purified-fragment binding and the cadherin-domain interface independently support this adhesion function. Heterophilic specificity is retained separately; cadherin-mediated adhesion does not require the classical homophilic mechanism. |
| GO:0045177 apical part of cell | IEA GO_REF:0000117 | ACCEPT | Summary: Retain the apical compartment supported by endogenous ortholog localization. Reason: The exact ARBA00028419 rule assigns apical part of cell; its FunFam-dependent target eligibility was not independently reproduced. The biological conclusion is independently supported by endogenous mouse Fat4 immunolocalization at the apical/subapical region of embryonic neural progenitors, with knockdown-based antibody controls and ultrastructural context. This positive compartment is not equated with the adjacent classical adherens junction or with a universal location of all splice products. |
| GO:0048699 generation of neurons | IEA GO_REF:0000117 | KEEP AS NON CORE | Summary: Retain generation of neurons as a developmental output of FAT4 signaling. Reason: The captured ARBA00027116 consequence is genuine, although all FunFam eligibility conditions were not reproduced. Independent mouse cortical Fat4 knockdown, differentiation-marker and Yap rescue experiments in the original PMID:24056717 supplement support the neuronal-generation/differentiation role. FAT4 functions as a cell-surface signaling/adhesion component in that process; this does not infer a catalytic neurogenic activity, a direct axon-growth mechanism, or equivalent activity of each natural splice product. |
| GO:0060429 epithelium development | IEA GO_REF:0000117 | TOO GENERAL | Summary: Epithelium development is covered by retained epithelial differentiation. Reason: The captured ARBA00027680 rule has this consequence, but its complete target FunFam eligibility was not reproduced. Positive mammalian Fat4 signaling evidence supports epithelial differentiation (GO:0030855, row 12), a descendant of GO:0060429. That retained specific process captures the relevant kidney developmental biology; the broad parent contributes no additional functional resolution. |
| GO:0070062 extracellular exosome | HDA PMID:23533145 In-depth proteomic analyses of exosomes isolated from expres... | UNDECIDED | Summary: The FAT4-specific exosome proteomics assignment remains unresolved. Reason: The main text of PMID:23533145 describes shotgun proteomics of differential-ultracentrifugation preparations from pooled expressed prostatic secretions in urine, with peptide/protein identification at FDR below 1%. However, the target-bearing Supplemental Table 2 and supplementary Methods could not be recovered as actual spreadsheet/document content in this review: available links returned download-gate HTML, access errors or unavailable-source responses. The main text does not independently identify the FAT4 protein/peptide entry. Retain the imported HDA context as undecided until that entry and its identification criteria can be checked. This access limitation does not demonstrate that the curator was wrong. Detection in a mixed vesicle preparation would not by itself establish a FAT4 exosome mechanism, cancer-specific role or localization of every natural product. Supporting Evidence: |
| GO:0098609 cell-cell adhesion | IEA GO_REF:0000120 | TOO GENERAL | Summary: Cell-cell adhesion is covered by retained cadherin-mediated and heterophilic adhesion. Reason: IPR039808 genuinely matches FAT4 and ARBA00026305 has the stated consequence; the complete FunFam rule path was not reproduced. GO:0098609 is an ancestor of the retained cadherin-mediated cell-cell adhesion GO:0044331 at row 14, and direct FAT4–DCHS1 experiments supply the specific heterophilic mechanism retained at row 7. Preserve this combined-source object while recognizing its redundant breadth. |
| GO:0045296 cadherin binding | IDA PMID:36797229 Structure of the planar cell polarity cadherins Fat4 and Dac... | NEW | Summary: FAT4 directly binds DCHS1 to mediate heterophilic contacts between cells. Reason: Purified human FAT4 EC1–4 binds human DCHS1 in isothermal titration calorimetry, and the interface-disrupting FAT4 L379R mutation abolishes detectable binding and reduces boundary accumulation in full-length-protein co-cultures. The earlier full-length human constructs were cloned from MCF7 and form heterotypic FAT4–DCHS1 boundary complexes. FAT4 itself performs the cadherin-binding step rather than being merely necessary for a downstream tissue phenotype. Cadherin binding (GO:0045296) is the molecular activity for this interaction, with the cell-cell adhesion context carried by the core functions rather than by the precomposed GO:0098641 (projects/INVOLVED_IN_PRECOMPOSITION.md); GO:0086080 would additionally require distinct cell types, which is not established by nonidentical cadherin partners. This added MF is distinct from the retained TNKS2 interaction and from the already imported adhesion processes. It does not assert that DCHS1 is the only partner or that all natural FAT4 products have the measured activity. |
Loading supporting content…
Download this section (compressed HTML)Q: In native mammalian cell contacts, which intracellular partners link FAT4 to the boundary-complex internalization behavior measured in the engineered human-cell system of PMID:39955614? The cell-culture measurements do not by themselves establish the corresponding tissue-polarity mechanism.
Q: Where is the N-terminally shortened natural product Q6V0I7-2 expressed and localized, and can it support DCHS1-dependent adhesion despite lacking the N-terminal EC1–4 region studied in the full-length and fragment assays? Sequence architecture alone does not establish its function.
Experiment: Compare sequence-verified full-length FAT4 with the published intracellular deletion construct at matched cell-surface abundance in a defined endogenous-expression mammalian system. Measure DCHS1 binding, boundary accumulation and internalization separately, then test candidate intracellular partners without treating actin inhibition as evidence of direct FAT4–actin binding.
Hypothesis: The contribution of the FAT4 intracellular region to boundary-complex turnover depends on cellular context.
Type: Controlled interaction and trafficking comparison
Experiment: First establish isoform-specific transcript and protein expression in a defined human context. Compare sequence-verified Q6V0I7-1 and Q6V0I7-2 for abundance, surface localization, DCHS1 binding and cell-boundary accumulation, without assigning a canonical signal-peptide cleavage site or functional equivalence in advance.
Hypothesis: The N-terminally shortened Q6V0I7-2 product may differ from full-length FAT4 in surface delivery or adhesion.
Type: Isoform-resolved expression and adhesion analysis
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)