The high frequency and high cost of urinary tract infections (UTIs) make these pathologies a major public health problem in developed countries. Uropathogenic Escherichia coli (UPEC), the major causative agent of UTIs, and its frequently associated cytotoxic necrotizing factor-1 (CNF1) toxin, have therefore attracted much attention. Investigations carried out in Aktories and Boquet's laboratories, together with Horiguchi's laboratory working on the CNF1-related dermonecrotic toxin of Bordetella, have clarified the molecular mechanism of action of this toxin family towards Rho GTPase proteins [1xBacterial toxins that modify the actin cytoskeleton. Barbieri, J.T. et al. Annu. Rev. Cell Dev. Biol. 2002; 18: 315–344Crossref | PubMed | Scopus (107)See all References, 2xSee all References].In vitro, these toxins cause permanent activation of Rho GTPases by modifying the glutamine residue responsible for GTP hydrolysis. Recent observations have revealed that in fact, CNF1 has a dual mechanism of action, comprising activation followed by the proteasome-dependent degradation of RhoA, Rac and Cdc42, thereby resulting in long-term, low-grade stimulation of the Rho GTPases [3xCNF1 exploits the ubiquitin-proteasome machinery to restrict Rho GTPase activation for bacterial host cell invasion. Doye, A. et al. Cell. 2002; 111: 553–564Abstract | Full Text | Full Text PDF | PubMed | Scopus (163)See all References, 4xProteasomal degradation of cytotoxic necrotizing factor 1-activated Rac. Lerm, M. et al. Infect. Immun. 2002; 70: 4053–4058Crossref | PubMed | Scopus (69)See all References]. This dual mechanism has been hidden until now owing to the lower capacity of some carcinoma cell lines used for CNF1 studies to efficiently carry out ubiquitination and proteasomal degradation of permanently activated Rho GTPases [3xCNF1 exploits the ubiquitin-proteasome machinery to restrict Rho GTPase activation for bacterial host cell invasion. Doye, A. et al. Cell. 2002; 111: 553–564Abstract | Full Text | Full Text PDF | PubMed | Scopus (163)See all References[3].Previous work has highlighted a role for CNF1 as an invasive factor [5xInduction of phagocytic behaviour in human epithelial cells by Escherichia coli cytotoxic necrotizing factor type 1. Falzano, L. et al. Mol. Microbiol. 1993; 9: 1247–1254Crossref | PubMedSee all References[5]. In addition, we could demonstrate that reaching a low level of Rac activation is required for efficient bacterial entry into non-professional phagocytic cells, together with associated cellular motility and cellular junction dynamic [3xCNF1 exploits the ubiquitin-proteasome machinery to restrict Rho GTPase activation for bacterial host cell invasion. Doye, A. et al. Cell. 2002; 111: 553–564Abstract | Full Text | Full Text PDF | PubMed | Scopus (163)See all References[3]. Both cellular motility and bacterial invasion thus appeared tightly coupled, requiring cell polarization [6xQuo vadis: polarized membrane recycling in motility and phagocytosis. Mellman, I. J. Cell Biol. 2000; 149: 529–530Crossref | PubMed | Scopus (24)See all References[6]. Our observations also raised an alternative hypothesis that pathogenic bacteria have developed antagonist virulence factors toward Rho GTPases to protect host cells from toxicity arising from massive activation of Rho GTPases. We suggest that instead, these virulence factors might cooperate for efficient bacterial invasion. Consistent with this model, different pathogenic bacteria appear to have developed different strategies to achieve a local polymerization of the actin cytoskeleton for cell penetration. For example, Salmonella injects both SopE/E2 and SptP virulence factors for the respective activation and inhibition of Rac and Cdc42 [7xStriking a balance: modulation of the actin cytoskeleton by Salmonella. Galan, J.E. and Zhou, D. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8754–8761Crossref | PubMed | Scopus (179)See all References[7]. Likewise, Shigella injects IpaC for Rac and Cdc42 activation, and IpaA, which produces indirect disassembly of local actin filaments, both factors acting together to produce efficient bacterial entry into cells [8xBacterial signals and cell responses during Shigella entry into epithelial cells. Tran Van Nhieu, G. et al. Cell Microbiol. 2000; 2: 187–193Crossref | PubMed | Scopus (99)See all References[8].The crosstalk between bacterial virulence factors and the ubiquitin/proteasome machineries is a major point of interest for future research. Consistent with this, the YopJ/P virulence factor of Yersinia has been reported to cleave SUMO-1 ubiquitin-like polypeptide, and SigD/SopB inositol phosphatase of Salmonella appeared regulated by ubiquitylation [9xDisruption of signaling by Yersinia effector YopJ, a ubiquitin-like protein protease. Orth, K. et al. Science. 2000; 290: 1594–1597Crossref | PubMedSee all References, 10xSalmonella enterica serovar Typhimurium effector SigD/SopB is membrane-associated and ubiquitinated inside host cells. Marcus, S.L. et al. Cell Microbiol. 2002; 4: 435–446Crossref | PubMed | Scopus (70)See all References].Finally, in addition to the demonstration that CNF1 is a bona fide invasive factor, we have evidence that an oncogenic protein, exemplified by Rac, is rapidly withdrawn upon permanent activation by the ubiquitin-proteasome machinery, pointing out the proteasome to be a guardian of the cellular homeostasis.