Sandra L. Marcus

Elimination of host cell PtdIns(4,5)P 2 by bacterial SigD promotes membrane fission during invasion by Salmonella

Salmonella invades mammalian cells by inducing membrane ruffling and macropinocytosis through actin remodelling. Because phosphoinositides are central to actin assembly, we have studied the dynamics of phosphatidylinositol-4,5-bisphosphate (PtdIns(4,5)P2) in HeLa cells during invasion by Salmonella typhimurium. Here we show that the outermost parts of the ruffles induced by invasion show a modest enrichment in PtdIns(4,5)P2, but that PtdIns(4,5)P2 is virtually absent from the invaginating regions. Rapid disappearance of PtdIns(4,5)P2 requires the expression of the Salmonella phosphatase SigD (also known as SopB). Deletion of SigD markedly delays fission of the invaginating membranes, indicating that elimination of PtdIns(4,5)P2 may be required for rapid formation of Salmonella-containing vacuoles. Heterologous expression of SigD is sufficient to promote the disappearance of PtdIns(4,5)P2, to reduce the rigidity of the membrane skeleton, and to induce plasmalemmal invagination and fission. Hydrolysis of PtdIns(4,5)P2 may be a common and essential feature of membrane fission during several internalization processes including invasion, phagocytosis and possibly endocytosis.

SifA permits survival and replication of Salmonella typhimurium in murine macrophages.

SifA was originally identified as a virulence factor required for formation of Salmonella‐induced filaments (Sifs), elongated tubules rich in lysosomal glycoproteins that extend from the Salmonella‐containing vacuole in infected epithelial cells. Here, we demonstrate that deletion mutants of ssaR, a component of the SPI‐2 type III secretion system, do not form Sifs in HeLa epithelial cells. This suggests that SifA is a translocated effector of this system, acting within host cells to form Sifs. In support of this hypothesis, transfection of HeLa cells with a vector encoding SifA fused to the green fluorescent protein caused extensive vacuolation of LAMP‐1‐positive compartments. Filamentous tubules that closely resembled Sifs were also observed in transfected cells, demonstrating that SifA is sufficient to initiate alteration of host cell endosomal structures. ΔsifA mutants were impaired in their ability to survive/replicate in RAW 264.7 murine macrophages, a phenotype similar to ssaR mutants. Our findings suggest that SifA is an effector of the SPI‐2 type III secretion system and allows colonization of murine macrophages, the host niche exploited during systemic phases of disease in these animals. A family of SifA‐related proteins and their importance to Salmonella pathogenesis is also discussed.

Structural and biochemical characterization of the type III secretion chaperones CesT and SigE.

Several Gram-negative bacterial pathogens have evolved a type III secretion system to deliver virulence effector proteins directly into eukaryotic cells, a process essential for disease. This specialized secretion process requires customized chaperones specific for particular effector proteins. The crystal structures of the enterohemorrhagic Escherichia coli O157:H7 Tir-specific chaperone CesT and the Salmonella enterica SigD-specific chaperone SigE reveal a common overall fold and formation of homodimers. Site-directed mutagenesis suggests that variable, delocalized hydrophobic surfaces observed on the chaperone homodimers are responsible for specific binding to a particular effector protein. Isothermal titration calorimetry studies of Tir–CesT and enzymatic activity profiles of SigD–SigE indicate that the effector proteins are not globally unfolded in the presence of their cognate chaperones.

A synaptojanin-homologous region of Salmonella typhimurium SigD is essential for inositol phosphatase activity and Akt activation

The Ser–Thr kinase Akt is activated in epithelial cells by Salmonella enterica serovar typhimurium. The bacterial effector SigD, which is translocated into host cells via the specialized type III secretion system, is essential for Akt activation. Here, we investigated the inositol phospholipid substrate preferences of SigD. Recombinant SigD preferentially dephosphorylated phosphatidylinositol 3,5‐biphosphate and phosphatidylinositol 3,4,5‐triphosphate over other phosphatidylinositol lipids. Phosphatidylinositol 3‐phosphate was not a substrate, suggesting the 5′ phosphate moiety is one of the preferred substrates. Database searches revealed that SigD bears a small region of homology to the mammalian type II inositol 5‐phosphatase synaptojanin. Mutation of two conserved residues in this region, Lys527 and Lys530, decreased or abrogated phosphatase activity, respectively. The Shigella flexneri SigD homologue, IpgD, displayed a similar activity in vitro and also activated Akt when used to complement a ΔsigD Salmonella strain. A mutation in IpgD at Lys507, analogous to Lys530 of SigD, also failed to activate Akt. Thus, we have characterized a region near the carboxyl‐terminus of SigD which is important for phosphatase activity. We discuss how dephosphorylation of inositol phospholipids by SigD in vivo might contribute to the activation of Akt.

Salmonella enterica serovar Typhimurium effector SigD/SopB is membrane-associated and ubiquitinated inside host cells

SigD/SopB is an effector protein translocated into host cells by one of the type III secretion systems of Salmonella enterica serovar Typhimurium (serovar Typhimurium). It is an inositol phosphatase that has activity towards several inositol phospholipids in vitro, including phosphatidylinositol 3,4,5‐ triphosphate. SigD activates Akt in epithelial cells and indirectly activates Cdc42 through one of its products, inositol 1,4,5,6‐tetrakisphosphate. As phospholipid targets of SigD activity are localized to host cell membranes, we sought to investigate the intracellular localization of translocated SigD. We show here that SigD is a membrane‐associated protein that is ubiquitinated inside host cells. SigD was extracted from host cell membranes with a high pH buffer but not by high salt. Fractionation and deletion analysis using transfected SigD‐green fluorescent protein fusions revealed that amino acid residues 117–167 of SigD are essential for membrane association, and that a fragment containing residues 29–116 was ubiquitinated. This is the first direct evidence of a bacterial effector protein being ubiquitinated. Treatment of cells with the proteasome inhibitor MG‐132 revealed that, unlike the host cell protein inhibitor of nuclear factor kappa B (IκBα), SigD does not appear to be rapidly degraded by the proteasome. We speculate that ubiquitination serves to downregulate SigD activity by an alternative mechanism, such as by targeting it for lysosomal degradation.