Salmonella invasion and delivery of protein effectors to mammalian cell cytoplasm

Abstract

I nvasion of epithelial cells by Salmonella and Shigella spp. occurs by macropinocytosis, a process whereby fluid and particles are internalized by membrane ruffles, probably induced through a pathway that includes CDC42 (Ref. 1). This is in contrast to bacteria such as Listeria monocytogenes that invade cells by stimulating receptormediated endocytosis through a pathway involving phosphoinositide 3-kinase 2. Analogous to several other Gram-negative bacterial pathogens, Salmonella spp. invasion is dependent upon proteins secreted by highly conserved type lII secretion systems (TTSS), termed Ssp(or Sip)B, C and D (reviewed in Ref. 3). The best-studied TFSS is the Yop secretion system of Yersinia. Using both morphological and biochemical techniques, it has been demonstrated that several Yops are translocated from bacteria into the cytosol of mammalian host cells 4,5. Yop translocation requires an intact secretion apparatus and Yops B, D and N. Rosqvist et al. demonstrated that Salmonella typhimurium translocates YopE into HeLa cells, indicating that the translocation machinery is conserved 6. Thus, it seems likely that translocated proteins play a role in Salmonella pathogenesis. Recently, Wood et al. presented data that a Salmonella dublinsecreted protein (SopE) promotes invasion and is translocated into HeLa cell cytoplasm by a mechanism that requires the Salmonella TTSS-secreted proteins (Ssp) essential to invasion, SspB-D (Ref. 7). This work raises interesting questions concerning the role of secreted and translocated proteins in the process of Salmonella spp. invasion. In the context of this paper, we will discuss two hypotheses for the mechanism of induction of macropinocytosis by Salmonella, as well as the mechanisms by which bacteria respond to host cell signals to initiate synthesis of TTSS. The data presented by Wood et al. suggest a model in which SspB and SspC, and perhaps other Ssps, are part of a translocation apparatus that delivers effectors of bacterial invasion, including SopE, to the mammalian cell cytosol (Fig. 1 a). These authors detected SopE within the cytoplasm of HeLa cells 4-6 h after infection with wild-type but not sspBCDA mutant bacteria, suggesting a role for Ssps in translocation. These data, although intriguing, are limited in that the multiplicity of infection was very high. In addition, bacterial overexpression of sopE was required to detect SopE within the mammalian cell cytoplasm. Although cytosolic SopE was not detected in sspBCDA mutants, suggesting an association with TTSS, it is possible that soluble extracellular SopE became internalized as a result of the induction of macropinocytosis. Wood et al. also showed that a sopE mutant is twofold less invasive, which the authors argue is consistent with multiple translocated proteins inducing macropinocytosis. An alternative hypothesis, based on data obtained with the SspB and SspC homologues from Shigella flexneri (IpaB and IpaC), is that SspB and SspC are the effectors of inva-