- Email: [email protected]
Subversion of Myosin Function by E. coli
Developmental Cell
Previews principles of such a recognition code will be more fully mapped out. As these principles emerge, we can expect MIT domains to take their place alongside ubiquitin- and phosphoinositide-binding domains as a major organizing theme in the ESCRT system and related pathways.
Hurley, J.H., and Emr, S.D. (2006). Annu. Rev. Biophys. Biomol. Struct. 35, 277–298.
son, C.M., and Urbe, S. (2007). J. Biol. Chem. 282, 30929–30937.
Lottridge, J.M., Flannery, A.R., Vincelli, J.L., and Stevens, T.H. (2006). Proc. Natl. Acad. Sci. USA 103, 6202–6207.
Scott, A., Gaspar, J., Stuchell-Brereton, M.D., Alam, S.L., Skalicky, J.J., and Sundquist, W.I. (2005). Proc. Natl. Acad. Sci. USA 102, 13813–13818.
REFERENCES
Obita, T., Saksena, S., Ghazi-Tabatabai, S., Gill, D.J., Perisic, O., Emr, S.D., and Williams, R.L. (2007). Nature 449, 735–739.
Azmi, I.F., Davies, B.A., Xiao, J., Babst, M., Xu, Z., and Katzmann, D.J. (2008). Dev. Cell 14, this issue, 50–61. Ciccarelli, F.D., Proukakis, C., Patel, H., Cross, H., Azam, S., Patton, M.A., Bork, P., and Crosby, A.H. (2003). Genomics 81, 437–441.
Nickerson, D.P., West, M., and Odorizzi, G. (2006). J. Cell Biol. 175, 715–720.
Phillips, S.A., Barr, V.A., Haft, D.H., Taylor, S.I., and Haft, C.R. (2001). J. Biol. Chem. 276, 5074–5084. Row, P.E., Lui, H., Hayes, S., Welchman, R., Charalabous, P., Hofmann, K., Claugue, M.J., Sander-
Stuchell-Brereton, M.D., Skalicky, J., Kieffer, C., Karren, M.A., Ghaffarian, S., and Sundquist, W.I. (2007). Nature 449, 740–744. Tsang, H.T.H., Connell, J.W., Brown, S.E., Thompson, A., Reid, E., and Sanderson, C.M. (2006). Genomics 88, 333–346. Xiao, J., Xia, H., Zhou, J., Azmi, I.F., Davies, B.A., Katzmann, D.J., and Xu, Z. (2008). Dev. Cell 14, this issue, 37–49.
Subversion of Myosin Function by E. coli Seema Mattoo,1 Neal M. Alto,3 and Jack E. Dixon1,2,* 1The
Howard Hughes Medical Institute, Leichtag Biomedical Research Building of Pharmacology, School of Medicine University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0721, USA 3University of Texas Southwestern Medical Center, 6000 Harry Hines Boulevard, NA6.308, Dallas, TX 75390, USA *Correspondence: [email protected] DOI 10.1016/j.devcel.2007.12.016 2Department
Enteropathogenic E. coli establish close contact with host cells by nucleating localized actin rearrangements and directly evading phagocytosis. Iizumi et al. now show in a recent issue of Cell Host and Microbe that the type III secretion effector EspB, initially thought to be involved in the translocation of other bacterial effectors, mediates antiphagocytosis and microvilli lesions by inhibiting myosin function. Enteropathogenic Escherichia coli (EPEC) colonize the gut mucosa and elicit diarrheal disease (Caron et al., 2006). The intestinal and colonic epithelia colonized by EPEC form a highly specialized cell type consisting of an apical brush border surface that faces the gut lumen and a basolateral membrane that makes contact with the interstitial tissue environment. EPEC infection is characterized by attaching and effacing (A/E) lesions, caused by bacterially mediated effacement of the intestinal brush border microvilli and subsequent intimate attachment of the bacterium to the host epithelium. EPEC belong to a large class of Gramnegative bacterial pathogens that utilize a specialized type III secretion system (T3SS) to directly inject ‘‘effector’’ proteins into the host cell; these effectors can severely alter the host’s cell-signaling
mechanisms (Caron et al., 2006). For EPEC, this arsenal of T3SS effectors is critical for virulence. EPEC effectors initiate signaling events that promote actin rearrangement and formation of a characteristic ‘‘pedestal’’-shaped membrane protrusion from the host cell. The bacteria make close contact with these pedestals, which move along the surface of the host cell in an actin-dependent manner. This movement is accompanied by the concurrent clearing of the microvillus brush border as EPEC establish their colonization niches. Several proteins involved in actin dynamics (Figure 1) are present in EPEC-induced pedestals (Freeman et al., 2000). Efficient establishment of infection by EPEC also requires crosstalk between host cytoskeletal networks, and a role for disruption of microtubules by EPEC effectors has been shown (Caron
8 Developmental Cell 14, January 2008 ª2008 Elsevier Inc.
et al., 2006). Interestingly, EPEC establish pedestal formation and microvilli effacement while simultaneously avoiding phagocytosis by macrophages. This antiphagocytic activity directly correlates with T3SS-dependent tyrosine dephosphorylation of several host proteins, for which a responsible effector has not yet been identified (Goosney et al., 1999). While the dynamics of EPEC-induced actin rearrangement have been extensively studied, the mechanism by which these pathogens evade phagocytosis remain unknown. EspB is another EPEC effector implicated in actin-mediated cytoskeletal rearrangement. EspB is a bifunctional protein that is targeted to the plasma membrane and cytosol of host cells, where it modulates the cell cytoskeleton by decreasing actin stress fibers (Taylor et al., 1999). It functions as a translocator protein
Developmental Cell
Previews
Figure 1. Schematic Depiction of the Events Carried Out by EPEC’s Type III Secreted Effectors for Manipulating the Host Cytoskeletal and Signaling Networks EPEC infection activates the host’s NFkB-dependent immune responses, leading to recruitment of phagocytic cells to the site of EPEC infection. EspB targets myosin to evade phagocytosis and induce A/E lesions. Host proteins involved in actin dynamics that are employed during EPEC infection are listed on the left. Straight green lines represent actin bundles; shapeless green lines represent EPEC-induced actin rearrangements; gray knobs represent myosin.
essential for the translocation of T3SS effectors into host cells, and is thus required for A/E lesions. EspB contains a putative transmembrane domain and three coiledcoil domains that allow it to form pores in the host cell membrane through which other bacterial effectors can be translocated (Warawa et al., 1999). It interacts with host a-catenin (Kodama et al., 2002) and a1-antitrypsin (Knappstein et al., 2004), but the physiological relevance of these interactions is not understood. In a recent manuscript appearing in Cell Host and Microbe, Iizumi et al. establish an unexpected role for EspB (Iizumi et al., 2007). By binding HeLa cell extracts to purified EspB immobilized on latex beads, and subjecting putative binding proteins to mass spectrometry analysis, Iizumi et al. identified myosin-1c as a novel binding partner for EspB. In fact, EspB bound to all myosin family members tested. EspB bound directly to the domain of myosin-1c that is typically involved in binding to actin filaments with an affinity nearly 20-fold greater than F-actin. In addition, EspB biochemically competed with actin in vitro to bind myosin-1c’s ac-
tin-binding domain in a dose-dependent manner. The authors identified an internal region of EspB (amino acids 159–218) that bound myosin and inhibited the interaction between myosin and actin. This region is highly conserved among EspB molecules from EPEC, the related Enterohemorrhagic E. coli (EHEC), and a closely related mouse pathogen Citrobacter rodentium. Since a natural infection model (other than humans) does not exist for EPEC/EHEC, infection of mice or mammalian tissue culture cell lines by C. rodentium is an accepted model system to reflect physiological aspects of EPEC/ EHEC pathogenesis. EspB was shown to coimmunoprecipitate with various myosin family members when delivered into Caco-2 intestinal epithelial cells using the C. rodentium infection model, while an EspB mutant protein lacking amino acids 159–218 (EspB-Dmid) did not. The authors further demonstrated that EspB partially colocalized with myosin in vivo, without altering the actin polymerization events triggered by the other EPEC effectors.
Perhaps the most interesting observation was the involvement of EspB’s myosin binding domain in the evasion of phagocytosis and the induction of the A/E phenotype. Myosin family members colocalize to the constricting end of phagosomes and are implicated in their contractile activity (Swanson et al., 1999). The authors, therefore, analyzed the uptake of EPEC expressing EspB or EspB-Dmid by bone marrow-derived macrophages, and found that the myosin binding region of EspB was essential for resisting phagocytosis. Similarly, EPEC expressing EspBDmid failed to efface microvilli in mammalian tissue culture cells following bacterial infection in vitro, as well as in a mouse model of infection with C. rodentium. Interestingly, the myosin binding region of EspB was not required for bacterial adherence to the gut epithelium or for delivery of other T3SS effectors. While actin filaments provide the structural core and mechanical support for brush border microvilli, class 1 myosin family members (mainly myosin-1a and -1c) line the microvillus shaft and anchor actin bundles to the plasma membrane. This allows the microvillus brush border to stand upright. Accordingly, Myo1a knockout mice display abnormal microvillar packing and nonuniform length (Tyska et al., 2005). Additionally, myosin and tropomyosin are found at the base of the microvillus shaft during pedestal formation following EPEC infection (Freeman et al., 2000). Targeting myosin and interfering with its ability to interact with the actin architecture seems like an obvious and simple tactic for a bacterium to destroy the microvillus brush border so that it can establish tight attachment to the underlying epithelium. Surprisingly, EspB can subvert myosin function without affecting the actin polymerization events occurring at the apical surfaces of the microvillus shaft. The most likely explanation is that EspB is delivered in highly concentrated doses to its site of action. This may also explain how EspB is able to outcompete and thereby prevent a highly abundant protein like actin from binding to myosin. Previous studies have suggested a role for dephosphorylation of host proteins by EPEC T3SS effectors as a mechanism for ablating phagocytosis (Goosney et al., 1999). This aspect was not tested by Iizumi et al., and it would be interesting
Developmental Cell 14, January 2008 ª2008 Elsevier Inc. 9
Developmental Cell
Previews to determine if EspB affected the phosphorylation state of any host protein. EspB’s myosin binding region alone appears to be essential for evading phagocytosis, and it appears to do so by physically competing with actin and interrupting myosin-induced phagosome constriction. Although the biochemistry behind this phenomenon needs further analysis, this observation suggests that the tyrosine dephosphorylation phenotype associated with EPEC’s antiphagocytic capabilities may simply be coincidental. It would be worthwhile to analyze EspB’s role in EPEC infection using Myo1a knockout mice. It also remains to be determined how EspB’s function fits with the overall actin dynamics coordi-
nated by the other T3SS effectors. Still, the simplicity of the interaction of this bacterial effector with myosin leading to such profound phenotypes that are essential for bacterial survival suggests that EPEC may not be alone in targeting myosin in this manner. REFERENCES Caron, E., Crepin, V.F., Simpson, N., Knutton, S., Garmendia, J., and Frankel, G. (2006). Curr. Opin. Microbiol. 9, 40–45.
Iizumi, Y., Sagara, H., Kabe, Y., Azuma, M., Kume, K., Ogawa, M., Nagai, T., Gillespie, P.G., Sasakawa, C., and Handa, H. (2007). Cell Host Microbe 2, 383–392. Knappstein, S., Ide, T., Schmidt, M.A., and Heusipp, G. (2004). Infect. Immun. 72, 4344–4350. Kodama, T., Akeda, Y., Kono, G., Takahashi, A., Imura, K., Iida, T., and Honda, T. (2002). Cell. Microbiol. 4, 213–222. Swanson, J.A., Johnson, M.T., Beningo, K., Post, P., Mooseker, M., and Araki, N. (1999). J. Cell Sci. 112, 307–316. Taylor, K.A., Luther, P.W., and Donnenberg, M.S. (1999). Infect. Immun. 67, 120–125.
Freeman, N.L., Zurawski, D.V., Chowrashi, P., Ayoob, J.C., Huang, L., Mittal, B., Sanger, J.M., and Sanger, J.W. (2000). Cell Motil. Cytoskeleton 47, 307–318.
Tyska, M.J., Mackey, A.T., Huang, J.D., Copeland, N.G., Jenkins, N.A., and Mooseker, M.S. (2005). Biol. Cell 16, 2443–2457.
Goosney, D.L., Celli, J., Kenny, B., and Finlay, B.B. (1999). Infect. Immun. 67, 490–495.
Warawa, J., Finlay, B.B., and Kenny, B. (1999). Infect. Immun. 67, 5538–5540.
10 Developmental Cell 14, January 2008 ª2008 Elsevier Inc.
Previews principles of such a recognition code will be more fully mapped out. As these principles emerge, we can expect MIT domains to take their place alongside ubiquitin- and phosphoinositide-binding domains as a major organizing theme in the ESCRT system and related pathways.
Hurley, J.H., and Emr, S.D. (2006). Annu. Rev. Biophys. Biomol. Struct. 35, 277–298.
son, C.M., and Urbe, S. (2007). J. Biol. Chem. 282, 30929–30937.
Lottridge, J.M., Flannery, A.R., Vincelli, J.L., and Stevens, T.H. (2006). Proc. Natl. Acad. Sci. USA 103, 6202–6207.
Scott, A., Gaspar, J., Stuchell-Brereton, M.D., Alam, S.L., Skalicky, J.J., and Sundquist, W.I. (2005). Proc. Natl. Acad. Sci. USA 102, 13813–13818.
REFERENCES
Obita, T., Saksena, S., Ghazi-Tabatabai, S., Gill, D.J., Perisic, O., Emr, S.D., and Williams, R.L. (2007). Nature 449, 735–739.
Azmi, I.F., Davies, B.A., Xiao, J., Babst, M., Xu, Z., and Katzmann, D.J. (2008). Dev. Cell 14, this issue, 50–61. Ciccarelli, F.D., Proukakis, C., Patel, H., Cross, H., Azam, S., Patton, M.A., Bork, P., and Crosby, A.H. (2003). Genomics 81, 437–441.
Nickerson, D.P., West, M., and Odorizzi, G. (2006). J. Cell Biol. 175, 715–720.
Phillips, S.A., Barr, V.A., Haft, D.H., Taylor, S.I., and Haft, C.R. (2001). J. Biol. Chem. 276, 5074–5084. Row, P.E., Lui, H., Hayes, S., Welchman, R., Charalabous, P., Hofmann, K., Claugue, M.J., Sander-
Stuchell-Brereton, M.D., Skalicky, J., Kieffer, C., Karren, M.A., Ghaffarian, S., and Sundquist, W.I. (2007). Nature 449, 740–744. Tsang, H.T.H., Connell, J.W., Brown, S.E., Thompson, A., Reid, E., and Sanderson, C.M. (2006). Genomics 88, 333–346. Xiao, J., Xia, H., Zhou, J., Azmi, I.F., Davies, B.A., Katzmann, D.J., and Xu, Z. (2008). Dev. Cell 14, this issue, 37–49.
Subversion of Myosin Function by E. coli Seema Mattoo,1 Neal M. Alto,3 and Jack E. Dixon1,2,* 1The
Howard Hughes Medical Institute, Leichtag Biomedical Research Building of Pharmacology, School of Medicine University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0721, USA 3University of Texas Southwestern Medical Center, 6000 Harry Hines Boulevard, NA6.308, Dallas, TX 75390, USA *Correspondence: [email protected] DOI 10.1016/j.devcel.2007.12.016 2Department
Enteropathogenic E. coli establish close contact with host cells by nucleating localized actin rearrangements and directly evading phagocytosis. Iizumi et al. now show in a recent issue of Cell Host and Microbe that the type III secretion effector EspB, initially thought to be involved in the translocation of other bacterial effectors, mediates antiphagocytosis and microvilli lesions by inhibiting myosin function. Enteropathogenic Escherichia coli (EPEC) colonize the gut mucosa and elicit diarrheal disease (Caron et al., 2006). The intestinal and colonic epithelia colonized by EPEC form a highly specialized cell type consisting of an apical brush border surface that faces the gut lumen and a basolateral membrane that makes contact with the interstitial tissue environment. EPEC infection is characterized by attaching and effacing (A/E) lesions, caused by bacterially mediated effacement of the intestinal brush border microvilli and subsequent intimate attachment of the bacterium to the host epithelium. EPEC belong to a large class of Gramnegative bacterial pathogens that utilize a specialized type III secretion system (T3SS) to directly inject ‘‘effector’’ proteins into the host cell; these effectors can severely alter the host’s cell-signaling
mechanisms (Caron et al., 2006). For EPEC, this arsenal of T3SS effectors is critical for virulence. EPEC effectors initiate signaling events that promote actin rearrangement and formation of a characteristic ‘‘pedestal’’-shaped membrane protrusion from the host cell. The bacteria make close contact with these pedestals, which move along the surface of the host cell in an actin-dependent manner. This movement is accompanied by the concurrent clearing of the microvillus brush border as EPEC establish their colonization niches. Several proteins involved in actin dynamics (Figure 1) are present in EPEC-induced pedestals (Freeman et al., 2000). Efficient establishment of infection by EPEC also requires crosstalk between host cytoskeletal networks, and a role for disruption of microtubules by EPEC effectors has been shown (Caron
8 Developmental Cell 14, January 2008 ª2008 Elsevier Inc.
et al., 2006). Interestingly, EPEC establish pedestal formation and microvilli effacement while simultaneously avoiding phagocytosis by macrophages. This antiphagocytic activity directly correlates with T3SS-dependent tyrosine dephosphorylation of several host proteins, for which a responsible effector has not yet been identified (Goosney et al., 1999). While the dynamics of EPEC-induced actin rearrangement have been extensively studied, the mechanism by which these pathogens evade phagocytosis remain unknown. EspB is another EPEC effector implicated in actin-mediated cytoskeletal rearrangement. EspB is a bifunctional protein that is targeted to the plasma membrane and cytosol of host cells, where it modulates the cell cytoskeleton by decreasing actin stress fibers (Taylor et al., 1999). It functions as a translocator protein
Developmental Cell
Previews
Figure 1. Schematic Depiction of the Events Carried Out by EPEC’s Type III Secreted Effectors for Manipulating the Host Cytoskeletal and Signaling Networks EPEC infection activates the host’s NFkB-dependent immune responses, leading to recruitment of phagocytic cells to the site of EPEC infection. EspB targets myosin to evade phagocytosis and induce A/E lesions. Host proteins involved in actin dynamics that are employed during EPEC infection are listed on the left. Straight green lines represent actin bundles; shapeless green lines represent EPEC-induced actin rearrangements; gray knobs represent myosin.
essential for the translocation of T3SS effectors into host cells, and is thus required for A/E lesions. EspB contains a putative transmembrane domain and three coiledcoil domains that allow it to form pores in the host cell membrane through which other bacterial effectors can be translocated (Warawa et al., 1999). It interacts with host a-catenin (Kodama et al., 2002) and a1-antitrypsin (Knappstein et al., 2004), but the physiological relevance of these interactions is not understood. In a recent manuscript appearing in Cell Host and Microbe, Iizumi et al. establish an unexpected role for EspB (Iizumi et al., 2007). By binding HeLa cell extracts to purified EspB immobilized on latex beads, and subjecting putative binding proteins to mass spectrometry analysis, Iizumi et al. identified myosin-1c as a novel binding partner for EspB. In fact, EspB bound to all myosin family members tested. EspB bound directly to the domain of myosin-1c that is typically involved in binding to actin filaments with an affinity nearly 20-fold greater than F-actin. In addition, EspB biochemically competed with actin in vitro to bind myosin-1c’s ac-
tin-binding domain in a dose-dependent manner. The authors identified an internal region of EspB (amino acids 159–218) that bound myosin and inhibited the interaction between myosin and actin. This region is highly conserved among EspB molecules from EPEC, the related Enterohemorrhagic E. coli (EHEC), and a closely related mouse pathogen Citrobacter rodentium. Since a natural infection model (other than humans) does not exist for EPEC/EHEC, infection of mice or mammalian tissue culture cell lines by C. rodentium is an accepted model system to reflect physiological aspects of EPEC/ EHEC pathogenesis. EspB was shown to coimmunoprecipitate with various myosin family members when delivered into Caco-2 intestinal epithelial cells using the C. rodentium infection model, while an EspB mutant protein lacking amino acids 159–218 (EspB-Dmid) did not. The authors further demonstrated that EspB partially colocalized with myosin in vivo, without altering the actin polymerization events triggered by the other EPEC effectors.
Perhaps the most interesting observation was the involvement of EspB’s myosin binding domain in the evasion of phagocytosis and the induction of the A/E phenotype. Myosin family members colocalize to the constricting end of phagosomes and are implicated in their contractile activity (Swanson et al., 1999). The authors, therefore, analyzed the uptake of EPEC expressing EspB or EspB-Dmid by bone marrow-derived macrophages, and found that the myosin binding region of EspB was essential for resisting phagocytosis. Similarly, EPEC expressing EspBDmid failed to efface microvilli in mammalian tissue culture cells following bacterial infection in vitro, as well as in a mouse model of infection with C. rodentium. Interestingly, the myosin binding region of EspB was not required for bacterial adherence to the gut epithelium or for delivery of other T3SS effectors. While actin filaments provide the structural core and mechanical support for brush border microvilli, class 1 myosin family members (mainly myosin-1a and -1c) line the microvillus shaft and anchor actin bundles to the plasma membrane. This allows the microvillus brush border to stand upright. Accordingly, Myo1a knockout mice display abnormal microvillar packing and nonuniform length (Tyska et al., 2005). Additionally, myosin and tropomyosin are found at the base of the microvillus shaft during pedestal formation following EPEC infection (Freeman et al., 2000). Targeting myosin and interfering with its ability to interact with the actin architecture seems like an obvious and simple tactic for a bacterium to destroy the microvillus brush border so that it can establish tight attachment to the underlying epithelium. Surprisingly, EspB can subvert myosin function without affecting the actin polymerization events occurring at the apical surfaces of the microvillus shaft. The most likely explanation is that EspB is delivered in highly concentrated doses to its site of action. This may also explain how EspB is able to outcompete and thereby prevent a highly abundant protein like actin from binding to myosin. Previous studies have suggested a role for dephosphorylation of host proteins by EPEC T3SS effectors as a mechanism for ablating phagocytosis (Goosney et al., 1999). This aspect was not tested by Iizumi et al., and it would be interesting
Developmental Cell 14, January 2008 ª2008 Elsevier Inc. 9
Developmental Cell
Previews to determine if EspB affected the phosphorylation state of any host protein. EspB’s myosin binding region alone appears to be essential for evading phagocytosis, and it appears to do so by physically competing with actin and interrupting myosin-induced phagosome constriction. Although the biochemistry behind this phenomenon needs further analysis, this observation suggests that the tyrosine dephosphorylation phenotype associated with EPEC’s antiphagocytic capabilities may simply be coincidental. It would be worthwhile to analyze EspB’s role in EPEC infection using Myo1a knockout mice. It also remains to be determined how EspB’s function fits with the overall actin dynamics coordi-
nated by the other T3SS effectors. Still, the simplicity of the interaction of this bacterial effector with myosin leading to such profound phenotypes that are essential for bacterial survival suggests that EPEC may not be alone in targeting myosin in this manner. REFERENCES Caron, E., Crepin, V.F., Simpson, N., Knutton, S., Garmendia, J., and Frankel, G. (2006). Curr. Opin. Microbiol. 9, 40–45.
Iizumi, Y., Sagara, H., Kabe, Y., Azuma, M., Kume, K., Ogawa, M., Nagai, T., Gillespie, P.G., Sasakawa, C., and Handa, H. (2007). Cell Host Microbe 2, 383–392. Knappstein, S., Ide, T., Schmidt, M.A., and Heusipp, G. (2004). Infect. Immun. 72, 4344–4350. Kodama, T., Akeda, Y., Kono, G., Takahashi, A., Imura, K., Iida, T., and Honda, T. (2002). Cell. Microbiol. 4, 213–222. Swanson, J.A., Johnson, M.T., Beningo, K., Post, P., Mooseker, M., and Araki, N. (1999). J. Cell Sci. 112, 307–316. Taylor, K.A., Luther, P.W., and Donnenberg, M.S. (1999). Infect. Immun. 67, 120–125.
Freeman, N.L., Zurawski, D.V., Chowrashi, P., Ayoob, J.C., Huang, L., Mittal, B., Sanger, J.M., and Sanger, J.W. (2000). Cell Motil. Cytoskeleton 47, 307–318.
Tyska, M.J., Mackey, A.T., Huang, J.D., Copeland, N.G., Jenkins, N.A., and Mooseker, M.S. (2005). Biol. Cell 16, 2443–2457.
Goosney, D.L., Celli, J., Kenny, B., and Finlay, B.B. (1999). Infect. Immun. 67, 490–495.
Warawa, J., Finlay, B.B., and Kenny, B. (1999). Infect. Immun. 67, 5538–5540.
10 Developmental Cell 14, January 2008 ª2008 Elsevier Inc.