Bacterial Invasion

Developmental Cell 314

Sagot, I., Rodal, A.A., Moseley, J., Goode, B.L., and Pellman, D. (2002). Nat. Cell Biol. 4, 626–631.

Xu, Y., Sagot, I., Moseley, J.B., Poy, F., Goode, B.L., Pellman, D., and Eck, M.J. (2004). Cell, in press.

Shimada, A., Nyitrai, M., Vetter, I.R., Kuhlmann, D., Bugyi, B., Narumiya, S., Geeves, M.A., and Wittinghofer, A. (2004). Mol. Cell, in press.

Yang, H.C., and Pon, L.A. (2002). Proc. Natl. Acad. Sci. USA 99, 751–756.

Wallar, B.J., and Alberts, A.S. (2003). Trends Cell Biol. 13, 435–446.

Bacterial Invasion: A New Strategy to Dominate Cytoskeleton Plasticity Some bacterial pathogens enter mammalian cells by injecting, directly into the host cytosol, proteins that trigger cytoskeletal rearrangements necessary for internalization. In the February 27 issue of Molecular Cell, McGhie et al. identify mechanisms by which the Salmonella protein SipA interferes with ADF/cofilin and gelsolin function. Thus Salmonella not only triggers actin polymerization but also counteracts the major F-actin destabilizing proteins. Some bacterial pathogens that replicate extracellularly can also enter into nonphagocytic cells, where they freely live and grow, away from host cell defenses. Bacterial entry into cells, like all phagocytic events, relies on the host actin cytoskeleton and on its capacity to be remodeled during formation and closure of the phagocytic cup. Bacteria have evolved various ways to trigger these actin rearrangements. In some cases, bacteria express on their surfaces proteins that interact with transmembrane receptors connected to the cytoskeleton. Engagement of these receptors leads to the progressive wrapping of the bacterium by the cell membrane and the formation of a phagocytic vacuole. Such a “zipper mechanism” takes place during entry of Yersinia and Listeria, which use an invasin-integrin interaction and an internalin-E-cadherin interaction, respectively, to trigger their entry into cells (Isberg and Barnes, 2001; Mengaud et al., 1996). In both cases, focal adhesion plaque and adherens junction machineries are maximally exploited by bacteria to stimulate actin rearrangements. A variation on the theme is the interaction of another invasion protein of Listeria, InlB, with the tyrosine kinase receptor protein Met (Shen et al., 2000). In other cases, e.g., Salmonella and Shigella, bacteria seem to bypass the need for a transmembrane receptor. As soon as they contact the host cell, they deliver directly into the cell cytosol, via type III secretion systems, effector proteins that induce transient, actin-rich membrane ruffles. These ruffles then engulf the infecting bacteria in a process that resembles macropinocytosis. Salmonella directly activates Rho GTPases by injecting SopE2 and SopE, proteins that act as guanine nucleotide exchange factors (GEFs) for the small GTPases Cdc42 and Rac, two well-known activators of the actin polymerization machinery, via WASP family proteins and the Arp2/3 complex (Galan, 2001). Salmonella also injects another effector protein SptP, a GTPase activating

Zigmond, S.H., Evangelista, M., Boone, C., Yang, C., Dar, A.C., Sicheri, F., Forkey, J., and Pring, M. (2003). Curr. Biol. 13, 1820–1823.

protein (GAP) for Cdc42 and Rac, which inactivate them at the end of the internalization process. A third injected protein SopB is a lipid phosphatase that also promotes actin rearrangements. Two other proteins, SipC and SipA, have also been implicated in the actin rearrangements induced by Salmonella (Galan, 2001; Zhou et al., 1999; Hayward and Koronakis, 1999). SipC, which is essential for entry, inserts into the plasma membrane during infection. It directly nucleates actin polymerization and bundles actin filaments in vitro. These SipC activities are activated by SipA, suggesting that the two proteins cooperate in vivo. SipA is not essential for entry but enhances its efficiency. Purified SipA binds to F-actin and prevents actin depolymerization in vitro. It promotes actin assembly by lowering the critical G-actin concentration required for nucleating actin filaments and enhances F-actin bundling by fimbrin/plastin and SipC. These activities lie in the C-terminal 226 amino acid fragment of SipA, a region of the protein which has been proposed to connect consecutive actin monomers as does eukaryotic nebulin. In their study, McGhie et al. (2004) have addressed the role of SipA in cell extracts, by taking advantage of the well-known actin-based motilities of Listeria and Shigella. They demonstrate that it binds F-actin, by showing that the bacterial actin tails could be decorated with, and stabilized by, SipA. That is, the tails appeared longer, much like those obtained when ADF/cofilin is depleted from extracts or when it is inactivated by LIMkinase overexpression (Bierne et al., 2001). This suggests that SipA protects F-actin from depolymerization by ADF/cofilin. It is indeed well established that ADF/cofilin induces actin depolymerization, thereby increasing actin turnover and actin-based motility (Loisel et al., 1999). McGhie et al. show that SipA not only protected F-actin from ADF/cofilin depolymerization but also displaced ADF/cofilin pre-bound to F-actin in vitro. Gelsolin is another actin binding protein that severs F-actin and caps barbed ends, thereby participating in the actin dynamics. In contrast to the situation with ADF/cofilin, SipA and gelsolin binding to F-actin were not mutually exclusive. SipA prevented severing by gelsolin but did not inhibit gelsolin binding. Furthermore, SipA re-annealed gelsolin-severed and capped F-actin filaments. Taken together, these findings highlight that SipA has multiple activities and show for the first time that a bacterial protein can counteract the activity of ADF/cofilin, a key regulator of actin cytoskeleton plasticity. It remains now to determine if during Salmonella internalization SipA actually displaces cofilin from actin filaments and if it protects the actin filaments from cofilin binding. Another important issue concerns the actin nucleation in vivo. Are two nucleation processes occurring in cells, one mediated by SipC and SipA, and one by Rac and

Previews 315

Cdc42? Does SipA only stabilize the filaments nucleated by SipC? Does it also stabilize the filaments induced via SopE2, SopE, Rac, and Cdc42? Finally, it will also be important for the establishment of the definitive role of SipA during entry to decipher when this protein functions. Is SipA, like SopE and SptP, a substrate for the proteasome (Kubori and Galan, 2003), and is it degraded to facilitate disruption of actin filaments after internalization? There are still important gaps to fill to understand the coordinated spatio-temporal action of the bacterial effectors during Salmonella entry. Pascale Cossart Unite´ des Interactions Bacte´ries Cellules Institut Pasteur 28 Rue du Docteur Roux INSERM U604 Paris 75015 France

Tipping the Balance toward Longevity Genetic experiments in C. elegans suggested that SIR2, an NAD-dependent protein deacetylase, acts through FOXO/DAF-16 transcription factor to prolong life. Recent studies show that mammalian SIR2 deacetylates FOXO, and may maximize survival by tempering cell death and increasing stress resistance. The provocative finding that single gene mutations can dramatically extend life span in several model organisms reveals that aging is plastic, regulated, and under genetic control. At least two pivotal regulators conserved across taxa robustly extend organismal survival, forkhead transcription factor FOXO, and the SIR2 (silencing information regulator 2) protein deacetylase. FOXO mediates the transcriptional output of insulin/ IGF-1 signal transduction, which regulates the longevity of species as diverse as worms, flies, and mice (reviewed in Tatar et al., 2003). Insulin/IGF-1 signaling is pro-aging, while mild inhibition of the pathway prolongs life. When insulin/IGF-1 signaling is active, a PI3 kinase/AKT kinase cascade phosphorylates FOXO, leading to its nuclear exclusion. When insulin signaling is inhibited, unphosphorylated FOXO enters the nucleus where it promotes programs of somatic endurance, stress resistance, and longevity. But how should blunted insulin/IGF-1 signaling promote longevity when normally it is associated with frank diabetes and premature senescence? One key is moderation: a modest downregulation of the pathway leads to healthy old age, while strong downregulation leads to frailty and morbidity. The other key is specificity. While insulin/IGF-I signaling is traditionally known for its role in metabolic and growth control, it may be a general sensor for all kinds of environmental stress. In turn,

Selected Reading Bierne, H., Gouin, E., Roux, P., Caroni, P., Yin, H.L., and Cossart, P. (2001). J. Cell Biol. 155, 101–112. Galan, J.E. (2001). Annu. Rev. Cell Dev. Biol. 17, 53–86. Hayward, R.D., and Koronakis, V. (1999). EMBO J. 18, 4926–4934. Isberg, R.R., and Barnes, P. (2001). J. Cell Sci. 114, 21–28. Kubori, T., and Galan, J.E. (2003). Cell 115, 333–342. Loisel, T.P., Boujemaa, R., Pantaloni, D., and Carlier, M.F. (1999). Nature 401, 613–616. McGhie, E.J., Hayward, R.D., and Koronakis, V. (2004). Mol. Cell 13, 497–510. Mengaud, J., Ohayon, H., Gounon, P., Mege, R.M., and Cossart, P. (1996). Cell 84, 923–932. Shen, Y., Naujokas, M., Park, M., and Ireton, K. (2000). Cell 103, 501–510. Zhou, D., Mooseker, M.S., and Galan, J.E. (1999). Science 283, 2092– 2095.

FOXO induces a whole suite of genes, including those involved in gluconeogenesis (PEPCK), detoxification of reactive oxygen species (MnSOD), heat shock, DNA damage repair (GADD45), growth control (4EBP), cell cycle arrest (p27KIP), cell death (BIM, FAS ligand), and innate immunity (see references within Motta et al., 2004). Of all the traits controlled by FOXO, stress resistance is most tightly correlated with longevity. How then might moderation and specificity be achieved? Here SIR2 may provide some answers. SIR2 was originally discovered in yeast as an NAD-dependent histone deacetylase with a role in silencing at the rDNA, telomeres, and mating type loci (Hekimi and Guarente, 2003). Notably, increased SIR2 dose prolongs yeast replicative life span by about 40%, while loss of function abrogates extended survival brought about by glucose limitation, a model of caloric restriction. Caloric restriction prolongs life span in numerous species, and it is proposed that SIR2 couples metabolic signals to downstream regulatory events. Resounding evidence that SIR2 regulates animal life span came from studies in the worm, where extra doses of sir-2.1 (one of four sirtuin homologs) extend adult life by 50% (Tissenbaum and Guarente, 2001). Moreover, extension was dependent on FOXO/daf-16, bringing these pivotal regulators together in a single pathway. But the specific nature of this interaction was unknown. Recently, several reports (Brunet et al., 2004; Motta et al., 2004) have explored the physical and biochemical interaction of mammalian FOXO homologs with SIRT1, the closest SIR2 homolog among the seven mammalian sirtuins. Indeed, consistent with unified action, it was shown that SIRT1 and FOXO physically interacted in coimmunoprecipitation experiments. Moreover, in response to stress FOXO was a substrate for acetylation by protein acetylases p300 and PCAF and subsequent deacetylation by SIRT1 and other deacetylases (Brunet et al., 2004; Motta et al., 2004).