FERMT2

FERMT2 (Kindlin-2) is a FERM domain-containing focal adhesion protein that co-activates β1 and β3 integrins, couples integrin-mediated adhesion to the actin cytoskeleton, and transduces mechanical and signaling cues to downstream effector pathways including RhoA/actomyosin, FAK–YAP/TAZ, Wnt/β-catenin, and TGFβ signaling. FERMT2 binds integrin β-subunit cytoplasmic tails through its PTB/FERM domain, synergizes with talin to drive integrin activation, and connects to the actin cytoskeleton via interactions with migfilin and filamin; loss of FERMT2 causes aberrant RhoA activation, membrane blebbing, and impaired adhesion and migration (PMID:12679033, PMID:18458155, PMID:17513299, PMID:29337051). FERMT2 directly interacts with APP to regulate its surface trafficking and Aβ production, and caspase/calpain-mediated cleavage of its F0–F1 domains modulates this activity, linking FERMT2 to Alzheimer's disease-associated amyloid and tau proteostasis (PMID:33144711, PMID:40273529, PMID:30371777). In cancer contexts, FERMT2 activates YAP/TAZ nuclear accumulation through integrin–FAK signaling independently of canonical Hippo kinases, stabilizes SOX2 to upregulate fibronectin and promote anoikis resistance, and drives Wnt/β-catenin-dependent epithelial–mesenchymal transition (PMID:41792242, PMID:40024947, PMID:36480537).

1. 2003 High

   Identifying FERMT2's first interacting partners established that focal adhesion scaffolding — rather than enzymatic activity — is its primary molecular role, answering how cell-matrix adhesions connect to the actin cytoskeleton through the migfilin–filamin axis.

   Evidence Co-immunoprecipitation, pulldown assays, and dominant-negative experiments in cultured cells showing migfilin bridges FERMT2 to filamin and actin

   PMID:12679033

   Open questions at the time

   • Integrin-binding mechanism of FERMT2 not yet characterized
   • Whether FERMT2 directly activates integrins or only scaffolds was unknown
2. 2007 High

   Mapping FERMT2's FERM domain as the direct binding interface for β1 and β3 integrin cytoplasmic tails resolved how FERMT2 is recruited to focal adhesions and demonstrated that this interaction promotes integrin activation and cell-ECM adhesion.

   Evidence Pulldown assays with integrin tail peptides, domain mutagenesis, αIIbβ3 activation in CHO cells, focal adhesion imaging

   PMID:17513299

   Open questions at the time

   • Whether FERMT2 synergizes with talin for integrin activation was not addressed
   • Structural basis of the FERM–integrin tail interaction at atomic resolution was lacking
3. 2008 High

   Demonstrating that FERMT2 synergizes with talin head domain to co-activate αIIbβ3 and that its PTB domain recognizes specific motifs (NITY759) in the β3 tail established the talin–kindlin cooperativity model of inside-out integrin signaling.

   Evidence In vitro binding assays, co-transfection of talin head + FERMT2 in CHO cells, siRNA knockdown with flow cytometry integrin activation readout

   PMID:18458155

   Open questions at the time

   • How FERMT2 and talin simultaneously engage the same integrin tail sterically was unresolved
   • Relevance to β1-integrin co-activation in non-hematopoietic cells not directly tested
4. 2014 Medium

   Identifying the F3 subdomain as necessary and sufficient for AKT/JNK-dependent anti-apoptotic signaling revealed that FERMT2 has signaling functions beyond mechanical adhesion scaffolding.

   Evidence Overexpression/siRNA in glioma cells, domain-deletion mutagenesis, kinase inhibitor pharmacology, apoptosis assays

   PMID:25152024

   Open questions at the time

   • Direct binding partner for F3 in the AKT pathway not identified
   • Whether this anti-apoptotic function is integrin-dependent was not determined
5. 2016 Medium

   A genome-wide siRNA screen uncovered an unexpected role for FERMT2 in APP metabolism, showing that FERMT2 underexpression increases cell-surface mature APP and Aβ production — the first link between FERMT2 and Alzheimer's disease biology.

   Evidence High-content siRNA screen, cell-surface APP quantification, APP recycling assays, CSF Aβ correlation

   PMID:27933404

   Open questions at the time

   • Whether FERMT2 directly binds APP or acts indirectly through integrins was unknown
   • Mechanism of APP recycling regulation by FERMT2 not defined
6. 2018 High

   FERMT2 loss in podocytes was shown to elevate RhoA activation and actomyosin contractility, causing membrane blebbing reversible by actomyosin inhibition — establishing FERMT2 as a negative regulator of RhoA-dependent cortical tension downstream of integrin adhesion.

   Evidence CRISPR KO in human podocytes and conditional mouse KO, RhoA activity assay, pharmacological rescue with actomyosin inhibitors

   PMID:29337051

   Open questions at the time

   • Identity of the RhoGEF or RhoGAP regulated by FERMT2 not determined
   • Whether this mechanism operates in cell types beyond podocytes was untested
7. 2019 Medium

   Validation in iPSC-derived neurons showed that FERMT2 modulates both Aβ and phospho-tau levels, but with direction depending on genetic background (knockdown reduces Aβ in wild-type neurons; CRISPR KO in familial AD neurons elevates Aβ42:40 ratio), highlighting context-dependent effects.

   Evidence shRNA and CRISPR-Cas9 in iPSC-derived neurons, ELISA for Aβ species and phospho-tau

   PMID:30371777

   Open questions at the time

   • Mechanism underlying opposing effects in different genetic backgrounds not resolved
   • Whether tau phosphorylation is a direct or indirect consequence of FERMT2 loss unclear
8. 2020 High

   Demonstrating direct FERMT2–APP protein interaction by Co-IP, and showing that FERMT2 underexpression impairs axonal growth, synaptic connectivity, and LTP in an APP-dependent manner, established FERMT2 as a direct regulator of APP function in neurons; the AD-risk SNP rs7143400-T was shown to reduce FERMT2 expression via miR-4504.

   Evidence Co-immunoprecipitation, genome-wide screens, shRNA in neurons with axonal/LTP readouts, luciferase 3'UTR reporter for miR-4504

   PMID:33144711

   Open questions at the time

   • Binding interface between FERMT2 and APP not mapped at domain level
   • Whether integrin co-activation is required for the FERMT2–APP interaction not tested
9. 2022 Medium

   Epistasis experiments in colorectal carcinoma cells showed that FERMT2 drives migration, invasion, and EMT through the Wnt/β-catenin pathway, with β-catenin overexpression rescuing FERMT2-knockdown phenotypes — establishing a second oncogenic signaling axis for FERMT2.

   Evidence FERMT2 overexpression/knockdown, Wnt/β-catenin western blotting, β-catenin rescue of migration and invasion

   PMID:36480537

   Open questions at the time

   • Whether FERMT2 regulates β-catenin via integrin-dependent or integrin-independent mechanism unknown
   • Direct molecular link between FERMT2 and β-catenin stabilization not identified
10. 2025 Medium

    Caspase and calpain I were identified as proteases that cleave FERMT2 between its F0 and F1 domains, reducing its control of APP processing — revealing a post-translational regulatory mechanism relevant to neurodegeneration.

    Evidence In vitro caspase/calpain cleavage assays, domain-dissociation analysis, APP processing readouts

    PMID:40273529

    Open questions at the time

    • In vivo relevance of FERMT2 cleavage at the synapse not demonstrated
    • Cleavage site residues not precisely mapped
11. 2025 Medium

    In gastric cancer, FERMT2 was shown to stabilize SOX2 by suppressing its ubiquitination, leading to FN1 transcriptional upregulation and anoikis resistance, with TGFβ-1/TGFβ-RI forming a positive feedback loop — defining a complete FERMT2–SOX2–FN1–TGFβ signaling circuit for peritoneal metastasis.

    Evidence Ubiquitination assays, ChIP/reporter for FN1, in vivo peritoneal metastasis models, TGFβ inhibitor experiments

    PMID:40024947

    Open questions at the time

    • Whether FERMT2 directly binds SOX2 or acts through an E3 ligase intermediate not resolved
    • Generalizability to cancers beyond gastric not tested
12. 2025 High

    FERMT2 was shown to activate YAP/TAZ nuclear accumulation through integrin–FAK signaling independently of canonical Hippo kinases LATS1/2, with glucocorticoid-driven FAK activation rescuing YAP/TAZ in FERMT2-depleted cells — establishing a Hippo-independent mechanotransduction pathway in breast cancer.

    Evidence CRISPR screen (in vitro and in vivo), FERMT2 KO, phospho-FAK blotting, FAK activator rescue, epistasis analysis

    PMID:41792242

    Open questions at the time

    • How FAK phosphorylation leads to YAP tyrosine phosphorylation at specific residues not mapped
    • Whether this Hippo-independent YAP regulation applies to non-malignant cells not tested

• Key open questions include the structural basis of simultaneous FERMT2–talin–integrin engagement, the direct binding interface between FERMT2 and APP, the identity of RhoGEFs/GAPs linking FERMT2 to RhoA, and whether FERMT2's diverse signaling outputs (Wnt, TGFβ, YAP/TAZ) are all integrin-dependent or reflect integrin-independent scaffolding functions.
• No high-resolution structure of FERMT2 in complex with integrin tail and talin
• Binding interface between FERMT2 and APP not mapped
• Whether Wnt/β-catenin and SOX2 stabilization functions require integrin binding unknown