CNF1 toxin is a virulence factor produced by uropathogenic Escherichia coli. Upon cell binding and introduction into the cytosol, CNF1 deamidates glutamine 63 of RhoA (or 61 of Rac and Cdc42), rendering constitutively active these GTPases. Unexpectedly, we measured in bladder cells a transient CNF1-induced activation of Rho GTPases, maximal for Rac. Deactivation of Rac correlated with the increased susceptibility of its deamidated form to ubiquitin/proteasome-mediated degradation. Sensitivity to ubiquitylation could be generalized to other permanent-activated forms of Rac and to its sustained activation by Dbl. Degradation of the toxin-activated Rac allowed both host cell motility and efficient cell invasion by uropathogenic bacteria. CNF1 toxicity thus results from a restricted activation of Rho GTPases through hijacking the host cell proteasomal machinery.

Plant water homeostasis is maintained by the phytohormone abscisic acid (ABA), which triggers stomatal pore closure in response to drought stress. We identified the Arabidopsis small guanosine triphosphatase (GTPase) protein AtRac1 as a central component in the ABA-mediated stomatal closure process. ABA treatment induced inactivation of AtRac GTPases and disruption of the guard cell actin cytoskeleton. In contrast, in the ABA-insensitive mutant abi1-1, which is impaired in stomatal closure, neither AtRac inactivation nor actin cytoskeleton disruption was observed on ABA treatment. These observations indicate that AtRac1 inactivation is a limiting step in the ABA-signaling cascade leading to stomatal closure. Consistent with these findings, expression of a dominant-positive mutant of AtRac1 blocked the ABA-mediated effects on actin cytoskeleton and stomatal closure in wild-type plants, whereas expression of a dominant-negative AtRac1 mutant recapitulated the ABA effects in the absence of the hormone. Moreover, the dominant-negative form of AtRac1 could also restore stomatal closure in abi1-1. These results define AtRac1 as a central element for plant adaptation to drought.

Adhesion to fibronectin through the alpha5beta1 integrin enables endothelial cells to proliferate in response to growth factors, whereas adhesion to laminin through alpha2beta1 results in growth arrest under the same conditions. On laminin, endothelial cells fail to translate Cyclin D1 mRNA and activate CDK4 and CDK6. Activated Rac, but not MEK1, PI-3K, or Akt, rescues biosynthesis of cyclin D1 and progression through the G(1) phase. Conversely, dominant negative Rac prevents these events on fibronectin. Mitogens promote activation of Rac on fibronectin but not laminin. This process is mediated by SOS and PI-3K and requires coordinate upstream signals through Shc and FAK. These results indicate that Rac is a crucial mediator of the integrin-specific control of cell cycle in endothelial cells.

Rac1 small GTPase controls essential aspects of cell biology and is a direct target of numerous bacterial virulence factors. The CNF1 toxin of pathogenic Escherichia coli addresses Rac1 to ubiquitin-proteasome system (UPS). We report the essential role of the tumor suppressor HACE1, a HECT-domain containing E3 ubiquitin-ligase, in the targeting of Rac1 to UPS. HACE1 binds preferentially GTP-bound Rac1 and catalyzes its polyubiquitylation. HACE1 expression increases the ubiquitylation of Rac1, when the GTPase is activated by point mutations or by the GEF-domain of Dbl. RNAi-mediated depletion of HACE1 blocks the ubiquitylation of active Rac1 and increases GTP-bound Rac1 cellular levels. HACE1 antagonizes cell isotropic spreading, a hallmark of Rac1 activation, and is required for endothelial cell monolayer invasion by bacteria. Together, these data establish the role of the HACE1 E3 ubiquitin-ligase in controlling Rac1 ubiquitylation and activity.