ARHGAP26

ARHGAP26 (GRAF/GRAF1) is a multidomain RhoGAP that restrains Rho-family GTPase signaling to control actomyosin contractility, membrane trafficking, and cytoskeletal remodeling (PMID:9858476, PMID:18954304, PMID:33835025). In vivo it acts as a GAP selective for RhoA, with substrate specificity dictated structurally by recognition of Glu-95/97 of the GTPase, and its catalytic GAP domain is held inactive in cis by an autoinhibitory intramolecular interaction with its BAR domain — a domain that independently binds and tubulates membranes (PMID:9858476, PMID:10982819, PMID:18954304). Through this RhoGAP activity ARHGAP26 inactivates RhoA upon integrin α5β1 engagement and PI3K signaling to stabilize cortical actin, and balances RhoGEF2–Rho-driven contractility at the actomyosin ring (PMID:19308677, PMID:33835025). ARHGAP26 is integrated into growth-factor and stress signaling: it is phosphorylated at Ser510 by MAP kinase downstream of EGF/NGF, engages FAK/PYK2 and the Rho effector kinase PKNβ via its SH3 domain, and limits EGFR signaling by promoting ubiquitylation-dependent receptor internalization and degradation (PMID:9525907, PMID:9494093, PMID:11432776, PMID:28993397). It also functions in mitochondrial quality control, where it is recruited to damaged mitochondria as a PRKN/Parkin-binding protein, is phosphorylated by PINK1, and coordinates actin remodeling for mitophagy, and it interacts with Cofilin1 to maintain mitochondrial integrity via DRP1 dynamics (PMID:38855880, PMID:39313581). ARHGAP26 protein levels are controlled by SMURF1-mediated ubiquitination and by promoter CpG methylation, the latter silencing GRAF in AML/MDS (PMID:31004081, PMID:16404424). Recurrent CLDN18-ARHGAP26 fusions in diffuse gastric cancer act paradoxically as gain-of-function oncogenes that activate RHOA and downstream FAK and YAP-TEAD signaling, disrupt epithelial adhesion, and drive invasion and signet-ring transformation (PMID:26146084, PMID:38621923).

1. 1998 High

   Establishing how GRAF couples to upstream signaling addressed whether this RhoGAP is regulated by, and physically linked to, growth-factor and adhesion kinases.

   Evidence In vitro MAP kinase assay with S510A mutagenesis and EGF/NGF-stimulated mobility shifts in PC12 cells; GST-SH3 pulldown and co-IP of PYK2/CAKβ from rat brain

   PMID:9494093 PMID:9525907

   Open questions at the time

   • Functional consequence of Ser510 phosphorylation on GAP activity not defined
   • PYK2 interaction not linked to a downstream cytoskeletal output
2. 1999 High

   Defining GRAF's in vivo substrate clarified which Rho-family GTPase it regulates, anchoring its cellular role in actin cytoskeletal control.

   Evidence Microinjection of wild-type vs GAP-dead GRAF into Swiss 3T3/PC12 cells with selective agonist challenge and C3 inhibition

   PMID:9858476

   Open questions at the time

   • In-cell assay does not exclude Cdc42 GAP activity under other conditions
   • Physiological trigger of GRAF activation not addressed
3. 2000 High

   Solving the GAP domain structure explained the molecular basis of substrate selectivity at atomic resolution.

   Evidence X-ray crystallography at 2.4 Å plus reciprocal GTPase mutagenesis (Cdc42 E95A, Rac1 A95E)

   PMID:10982819

   Open questions at the time

   • Structure of full-length protein and BAR-GAP arrangement not resolved
   • RhoA-specific contacts inferred rather than co-crystallized
4. 2001 Medium

   Discovery of PKNβ binding and phosphorylation revealed a feedback loop linking a Rho effector kinase back to its GAP.

   Evidence Yeast two-hybrid, GST-SH3 pulldown, co-IP from COS-7, and in vitro kinase assay with active PKNβ

   PMID:11432776

   Open questions at the time

   • Effect of PKNβ phosphorylation on GAP activity unresolved
   • Single lab, no in vivo validation
5. 2006 Medium

   Demonstrating epigenetic silencing established how GRAF is inactivated in myeloid malignancy without mutation.

   Evidence Promoter reporter assay, bisulfite methylation analysis, and pharmacological reactivation with demethylating agent plus HDAC inhibitor in leukemia lines

   PMID:16404424

   Open questions at the time

   • Causal contribution of silencing to leukemogenesis not tested functionally
   • Downstream RhoA consequences in leukemia not measured
6. 2009 High

   Characterizing BAR-mediated autoinhibition and dual signal-driven recruitment explained how GRAF GAP activity is spatially and temporally controlled.

   Evidence In vitro GAP and liposome tubulation assays with domain dissection; integrin blocking and PI3K inhibition with RhoA activity readouts in a breast cancer dormancy model

   PMID:18954304 PMID:19308677

   Open questions at the time

   • Mechanism relieving BAR autoinhibition at membranes not defined
   • Dormancy-model findings are single-study, cell-based
7. 2017 Medium

   Ortholog work positioned GRAF in receptor endocytosis, showing it limits EGFR signaling by promoting receptor internalization and degradation.

   Evidence Drosophila Graf loss-of-function, GEEC localization, EGFR internalization/degradation assays, and Graf-EGFR co-IP

   PMID:28993397

   Open questions at the time

   • Conservation of EGFR regulation in mammalian cells not shown
   • Whether GAP activity is required for endocytic function unclear
8. 2018 Medium

   Loss-of-function studies in human smooth muscle cells linked ARHGAP26 to proliferation/migration via the RhoA-ROCK1-PTEN axis under hypoxia.

   Evidence siRNA knockdown in primary DASMCs, hypoxia culture, RhoA/ROCK1/phospho-PTEN Westerns, and ROCK inhibition

   PMID:30592323

   Open questions at the time

   • Direct GAP-substrate relationship in this axis not biochemically dissected
   • Single cell type, single lab
9. 2019 Medium

   Identifying SMURF1-mediated ubiquitination established post-translational control of ARHGAP26 abundance and its consequence for cancer cell motility.

   Evidence Co-IP, ubiquitination assay, overexpression/knockdown rescue in ovarian cancer lines, and in vivo metastasis model

   PMID:31004081

   Open questions at the time

   • Ubiquitination sites and degradation route not mapped
   • Link to RhoA GAP activity vs β-catenin pathway not fully separated
10. 2021 High

    Genetic dissection in Drosophila placed GRAF in a RhoGEF2-Rho-GRAF balance governing actomyosin ring constriction and in EGFR-MAPK-dependent neural patterning.

    Evidence Graf RNAi/mutants with domain-specific rescue, RhoGEF2/ROCK epistasis, live imaging, Rho-GTP pulldown; mosaic analysis with human OPHN1 rescue and memory assays

    PMID:33835025 PMID:33892766

    Open questions at the time

    • Mammalian conservation of the cellularization role untested
    • Whether neural phenotype reflects direct EGFR-MAPK regulation by GRAF unproven
11. 2024 High

    Recent work expanded ARHGAP26 into mitochondrial quality control and oocyte integrity, and defined CLDN18-ARHGAP26 as a gain-of-function oncogene rather than a GAP loss.

    Evidence PRKN co-IP, PINK1 phosphosite mapping, KO mouse hearts (mitophagy); Cofilin1 co-IP and Arhgap26 KO oocytes (DRP1/ROS); LSL-CLDN18-ARHGAP26 transgenic mice/organoids with RHOA-FAK-YAP-TEAD analysis and inhibitor rescue

    PMID:38621923 PMID:38855880 PMID:39313581

    Open questions at the time

    • How a RhoGAP fusion activates RHOA mechanistically not resolved
    • Whether mitochondrial and cytoskeletal roles share a common biochemical activity unclear

• It remains unresolved how the BAR-GAP autoinhibition, SH3-partner engagement, and phosphorylation/ubiquitination inputs are integrated to switch ARHGAP26 between its GAP, membrane-tubulation, and mitophagy functions in mammalian cells.
• No structure of full-length regulated protein
• Mechanism converting fusion into RHOA activation unknown
• Crosstalk between cytoskeletal and mitochondrial roles undefined