SopD2 is a Novel Type III Secreted Effector of Salmonella typhimurium That Targets Late Endocytic Compartments Upon Delivery Into Host Cells

Authors

  • John H. Brumell,

    1. Biotechnology Laboratory and Departments of Biochemistry and Molecular Biology, Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T-1Z3, Canada
    Search for more papers by this author
  • Sonya Kujat-Choy,

    1. Biotechnology Laboratory and Departments of Biochemistry and Molecular Biology, Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T-1Z3, Canada
    Search for more papers by this author
  • Nat F. Brown,

    1. Biotechnology Laboratory and Departments of Biochemistry and Molecular Biology, Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T-1Z3, Canada
    Search for more papers by this author
  • Bruce A. Vallance,

    1. Biotechnology Laboratory and Departments of Biochemistry and Molecular Biology, Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T-1Z3, Canada
    Search for more papers by this author
  • Leigh A. Knodler,

    1. Biotechnology Laboratory and Departments of Biochemistry and Molecular Biology, Microbiology and Immunology, University of British Columbia, Vancouver, BC, V6T-1Z3, Canada
    Search for more papers by this author
  • B. Brett Finlay

    Corresponding authorSearch for more papers by this author

Corresponding author: B. Brett Finlaybfinlay@interchange.ubc.ca

Abstract

Salmonella typhimuriumis a facultative intracellular pathogen that utilizes two type III secretion systems to deliver virulence proteins into host cells. These proteins, termed effectors, alter host cell function to allow invasion into and intracellular survival/replication within a vacuolar compartment. Here we describe SopD2, a novel member of the Salmonella translocated effector (STE) family, which share a conserved N-terminal type III secretion signal. Disruption of the sopD2 gene prolonged the survival of mice infected with a lethal dose of Salmonella typhimurium, demonstrating a significant role for this effector in pathogenesis. Expression of sopD2 was induced inside host cells and was dependent on functional ssrA/B and phoP/Q, two component regulatory systems. HA-tagged SopD2 was delivered into HeLa cells in a SPI-2-dependent manner and associated with both the Salmonella-containing vacuole and with swollen endosomes elsewhere in the cell. Subcellular fractionation confirmed that SopD2 was membrane associated in host cells, while the closely related effector SopD was localized to the cytosol. A SopD2 fusion to GFP associated with small tubular structures and large vesicles containing late endocytic markers, including Rab7. Surprisingly, expression of N-terminal amino acids 1–150 of SopD2 fused to GFP was sufficient to mediate both binding to late endosomes/lysosomes and swelling of these compartments. These findings demonstrate that the N-terminus of SopD2 is a bifunctional domain required for both type III secretion out of Salmonella as well as late endosome/lysosome targeting following translocation into host cells.

Salmonella enterica serovar typhimurium (S. typhimurium) is a facultative intracellular pathogen capable of causing a variety of diseases in different hosts (1,2). Throughout the course of disease, S. typhimurium inhabits a vacuolar niche inside host cells, referred to as the Salmonella-containing vacuole (SCV) (3,4). A hallmark of this pathogen is its ability to modify intracellular trafficking of the SCV, a strategy shared with other intracellular pathogens (5). Recent studies have demonstrated that modified trafficking of the SCV is unique within different cell types of the host (6). In macrophages, S. typhimurium blocks fusion of the SCV with late endocytic compartments including late endosomes and lysosomes [see (7) for a review of this subject]. Assembly of the NADPH oxidase, a free radical generating system that plays a major role in host defense, is also blocked on the SCV membrane (8,9). In epithelial cells, early (< 3 h post infection) trafficking of the SCV is characterized by an avoidance of late endocytic compartments (10–12) and the recruitment of lysosomal glycoproteins from the cell surface (13). At later times post infection, when the bacteria have begun intracellular replication, extensive fusion of late endocytic compartments with the SCV is observed (14). This mediates delivery of large amounts of membrane that cause tubulation of the SCV into large (up to 40 μm in length) structures known as Salmonella-induced filaments (Sifs) (15,16).

Modulation of SCV trafficking is mediated, in part, by the actions of two different type III secretion systems (TTSS) expressed by S. typhimurium at different stages of disease. These secretion systems mediate the delivery of bacterial proteins, called effectors, into host cells where they act to subvert host cell machinery, including signal transduction cascades (17). The Salmonella Pathogenicity Island (SPI)-1 encoded TTSS and its effectors play a major role in animal models of gastroenteritis (18–22) and mediate the invasion of S. typhimurium into nonphagocytic cells in vitro (23). The latter function is accomplished by a battery of translocated SPI-1 effectors that initiate actin rearrangements in the host cell, leading to ruffling of the plasma membrane and uptake of the bacteria into nascent SCVs (24). Our recent studies have also demonstrated the role of SPI-1 effectors in intracellular survival/replication in epithelial cells, possibly through modulation of SCV trafficking at early times post invasion (25).

Upon uptake by host cells, S. typhimurium initiates expression of a second TTSS, encoded in SPI-2. This system mediates delivery of a unique set of effectors into the host cell. However, this occurs across the SCV inside the host cell and not across the plasma membrane. The SPI-2 TTSS plays a major role in systemic phases of disease in infected mice and is required for replication within macrophages (26–33). Effectors of SPI-2 include the Sse (Secretion system effector) family of proteins, which are encoded within the SPI-2 pathogenicity island (34–36). Recent studies have also demonstrated the existence of a subset of SPI-2 effectors encoded throughout the chromosome. This includes some members of the Salmonella Translocated Effector (STE) family, which share a conserved N-terminal sequence that serves as a type III secretion signal for both the SPI-1- and SPI-2-encoded TTSS (37,38). Despite the importance of the SPI-2 TTSS in causing disease, identification of its secreted effectors and their function in disease remains incomplete.

Principal to understanding the function of SPI-2 effectors is determining where these proteins act within the host cell. We recently discovered that SPI-2 effectors can associate with the SCV after translocation into host cells. For example, at late times postinfection SifA associates with the SCV membrane, including tubular extensions of the SCV that form Sifs (39). Transfection of HeLa cells with a SifA fusion to GFP causes swelling and tubulation of endocytic compartments (40). Thus, it is tempting to speculate that SifA acts at the SCV to promote membrane fusion and/or inhibit membrane fission, leading to membrane accumulation and Sif formation. This would also allow for maintenance of the SCV, another recently described function of this virulence factor (41). Another SPI-2 effector, PipB, was also found to associate with the SCV in infected cells, though its role in disease and mechanism of action remain unknown (42). How each SPI-2 effector is targeted to specific compartments within host cells remains to be determined.

Here we describe the identification and characterization of SopD2, a novel member of the STE family. SopD2 was found to associate with late endocytic compartments upon delivery into host cells. This targeting was mediated by a discrete portion of its N-terminus, that has previously been shown to function as a type III secretion signal for members of the STE family (37,38). Thus, the N-terminus of SopD2 can serve as a subcellular targeting domain within prokaryotic and eukaryotic cells. These findings demonstrate that SPI-2 effectors can specifically target subcellular compartments of the host other than the SCV and highlight the complexity by which S. typhimurium interacts with its host to alter endosomal processes.

Results

SopD2 is a novel member of the Salmonella Translocated Effector family

We (40,43) and others (38,41,44) have previously identified a conserved N-terminal sequence present in the STE family of type III secreted effectors of S. typhimurium. This sequence of 140–150 amino acids mediates effector translocation into host cells by both the SPI-1- and SPI-2-encoded type III secretion systems (37,38). We searched the genomic database of S. typhimurium (45), using this conserved N-terminal sequence, and identified a novel gene, which we named sopD2. The protein encoded by sopD2 is 42% identical to SopD, a known effector of the SPI-1 TTSS that plays a major role in gastroenteritis in animal models of Salmonella infection (19). As with other members of the STE family, sopD2 is present in a low (G + C) islet, suggesting recent acquisition by a horizontal transfer event (Figure 1A). This islet is inserted within the pflAB operon, which encodes pyruvate formate lyase and its activating enzyme (conserved in E. coli) and maps to 19.5 cs of the S. typhimurium chromosome. Homologous sequences are also present in the genomic databases of S. paratyphi, S. dublin and S. enteriditis. Interestingly, sopD2 in S. typhi has a frameshift mutation at codon 48 and has been classified as a pseudogene by Parkhill et al. (NCBI accession #003198), suggesting that sopD2 may have a host-adapted function (46). The sequence similarity between SopD2 and SopD spans the entire length of both proteins and includes a common putative coiled coil domain in the C-terminal region (Figure 1B).

Figure 1.

Figure 1.

A novel islet-encoded type III effector of S. typhimurium.A. sopD2 is encoded within a low (G + C) pathogenicity islet of the S. typhimurium chromosome. Per cent (G + C) content for each region of the chromosome is indicated. B. Alignment of SopD and SopD2. Alignments were performed with CLUSTAL W and colored with BOXSHADE through the Biology Workbench 3.2 server (http://workbench.sdsc.edu/cgi/bw.cgi). The predicted coiled coil region shared by SopD and SopD2 was identified using the Coils program at (http://www.ch.embnet.org/software/COILS_form.html).

SopD2 contributes to virulence in mice

To test a role for SopD2 in disease, mice were infected by intraperitoneal injection with a lethal dose of S. typhimurium bearing a nonpolar disruption in the sopD2 gene. Mutation of sopD2 led to a prolonged survival of infected mice compared to those infected with the isogenic wild-type strain. Results from a typical experiment are shown in Figure 2. The average survival times for three such experiments was 4.9 ± 0.2 days for mice infected with wild-type bacteria and 8.2 ± 0.5 days for mice infected with the sopD2 mutant. Furthermore, in a competitive index assay (41) the sopD2 mutant was recovered from spleens at significantly lower levels than the wild-type strain, after the two were coinfected (at a ratio of 1 : 1) into the same mouse (data not shown). These findings demonstrate that sopD2 plays an important role in pathogenesis of S. typhimurium in the mouse model of Salmonella infection.

Figure 2.

Figure 2.

Role of sopD2 in virulence of S. typhimurium in mice. Female BALB/c mice (6–8 weeks old) were infected by intraperitoneal injection with 5 × 104 cfu of either wild-type S. typhimurium SL1344 (closed boxes) or isogenic mutants of this strain with a disruption in the sopD2 gene (open diamonds). Survival times are shown for a representative experiment in which groups of five mice were infected with either strain.

SopD2 expression is up-regulated inside host cells

The expression of sopD and sopD2 was examined in vitro by growing S. typhimurium under conditions that mediate induction of either the SPI-1- or the SPI-2-encoded TTSS and their effectors (see Experimental Procedures). RT-PCR analysis demonstrated that sopD transcription was optimal under conditions that induce the SPI-1 TTSS but was down-regulated in medium that induces expression of SPI-2 (Figure 3A). Conversely, sopD2 transcription was up-regulated under SPI-2-inducing conditions, indicating that these genes are differentially regulated.

Figure 3.

Figure 3.

Analysis of sopD2 expression.A. RT-PCR analysis of sopD and sopD2 in S. typhimurium cultures in vitro under conditions that induce the SPI-1 or SPI-2 encoded TTSS and their effectors. B. Expression of HA-tagged SopD and SopD2 in vitro in either wild-type S. typhimurium or mutants deficient in the ssrA/B or phoP two component regulatory systems, as indicated. Cultures were grown under conditions that induce expression of either the SPI-1 or SPI-2 TTSS and their respective effectors, as indicated. Expression was analyzed by immunoblotting samples with anti-HA antibodies. Induction of SPI-1 or SPI-2 was confirmed by immunoblotting samples with antibodies to SigD (SPI-1 effector) and SseB (a SPI-2 encoded protein), respectively. Equivalent protein loading was confirmed by blotting samples with antibodies to Dna kinase. C. HeLa cells were infected with S. typhimurium expressing HA-tagged SopD2 for the indicated time and then subjected to SDS PAGE and immunoblotting with anti-HA antibodies. Protein loading was normalized to an equivalent number of intracellular bacteria.

Protein expression in vitro was examined with the use of plasmids encoding SopD and SopD2 with two tandem C-terminal hemagglutinin (HA) tags (see Experimental Procedures). As shown (Figure 3B), expression of HA-tagged SopD in wild-type bacteria was maximal under conditions that induce the SPI-1 TTSS and its effectors, including SigD. Growth conditions that induce the SPI-2 TTSS (and the SPI-2 encoded protein SseB) led to a relative decrease in the expression of SopD-2HA. It should be noted, however, that SopD-2HA expression occurred to some extent under SPI-2-inducing conditions, while expression of the SPI-1 effector SigD was not detectable under these conditions. This suggests that SopD, originally identified as an effector of the SPI-1 TTSS, may also be translocated into host cells by the SPI-2 TTSS.

Consistent with sopD2 transcript analysis, SopD2–2HA expression increased in wild-type bacteria under SPI-2-inducing conditions. Expression was dependent on ssrA, the transcriptional activator for the ssrA/B two-component regulatory system encoded within SPI-2. SopD-2HA expression, however, was not affected by deletion of ssrA under these growth conditions. These findings demonstrate that sopD2 is a novel member of the ssrA/B regulon, which regulates genes encoded throughout the S. typhimurium chromosome including SPI-2 effectors (47). Expression of SopD2–2HA was also dependent on the phoP/Q two-component global regulatory system, which plays a major role in the virulence of S. typhimurium (48). The expression profile of both SopD-2HA and SopD2–2HA was largely unaffected by deletion of hilA, a regulator of some SPI-1 effectors (49,50) including SigD (51) (data not shown).

Induction of SopD2 in SPI-2-inducing medium, which is thought to mimic the chemical composition of the SCV (52), suggests that S. typhimurium expresses this protein upon uptake into its vacuolar niche in host cells. To confirm that SopD2 expression is induced inside host cells, HeLa epithelial cells were infected with S. typhimurium expressing SopD2–2HA, and lysates of infected cells were analyzed by SDS PAGE and immunoblotting with antibodies to the HA epitope. Protein loading was normalized such that equivalent numbers of bacteria were present in each lane (determined by measuring colony forming units from cell lysates prior to preparation of samples for SDS PAGE). As shown in Figure 3(C), expression of SopD2–2HA was detected by 5 h post infection and was maintained at all time points tested thereafter.

Membrane association of SopD2 in host cells

The localization of SopD2 in infected HeLa cells was analyzed by indirect immunofluorescence. SopD2–2HA was consistently observed to associate with the Salmonella-containing vacuole, both immediately surrounding bacteria (Figure 4A, upper panels), as well as tubular extensions of this compartment known as Salmonella-induced filaments (data not shown). Association of SopD2–2HA with the SCV was highest at later times post infection (18–21 h) and when intracellular bacteria had undergone swelling and elongation, a variable phenotype in HeLa cells that is more commonly seen in macrophages and melanoma cells (53,54). SopD2–2HA also localized to large endosomes that costained with the lysosomal glycoprotein LAMP-2, a marker of late endocytic compartments (see arrowheads). In parallel experiments, HeLa cells were infected with an S. typhimurium mutant lacking ssaR, an essential component of the SPI-2 type III secretion apparatus (31). No specific signal was detected with the HA antibody in cells infected with ssaR mutants expressing SopD2–2HA (Figure 4A, lower panels).

Figure 4.

Figure 4.

Delivery of SopD2 into host cells and association with membranes.A. HeLa cells were infected for 21 h with S. typhimurium expressing HA-tagged SopD2. Infected cells were then fixed and co-immunostained for LAMP-2 and the HA epitope and analyzed by indirect immunofluorescence. The SopD2–2HA protein localized to Salmonella-containing vacuoles (arrows) and swollen endosomes throughout infected cells (see arrowheads). No specific signal for the HA epitope was observed in cells infected with ssaR (SPI-2 TTSS defective) bacteria expressing SopD2–2HA, demonstrating specificity of the HA antibody. Size bar indicates 10 μm. B. Cells infected as in A were subjected to mechanical lysis and fractionation of membranes by centrifugation (see Experimental Procedures). Fractions were subjected to SDS-PAGE and immunoblotting for the HA epitope tag. As shown, SopD2–2HA associated predominantly with membranes (M) in cells infected with the wild-type strain but not with the ssaR (SPI-2 TTSS mutant) strain of S. typhimurium. Expression of SopD2–2HA in the ssaR mutant under these conditions was observed in the low-speed pellet (P) containing intact bacteria. SopD-2HA was localized to the cytosolic (C) fraction.

The preceding results suggested that SopD2 is targeted to specific membrane compartments in host cells. To further analyze subcellular distribution of SopD2, fractionation of infected cells was performed by mechanical disruption and differential centrifugation, as previously described by our laboratory (55). Immunoblotting with antibodies to the HA epitope revealed that SopD2–2HA was present predominantly in the membrane (M) fraction of infected HeLa cells (Figure 4B). In contrast, SopD was observed exclusively in the cytosol (C) fraction of infected cells. The ssaR (SPI-2 TTSS) mutant was capable of expressing HA-tagged SopD2, which was localized to the pellet (P) fraction containing intact bacteria and unbroken cells. However, delivery of SopD2–2HA into either the cytosol or membrane fractions of HeLa cells was not observed following infection with the ssaR mutant. These findings demonstrate that HA-tagged SopD2 is translocated into host cells in an SPI-2-dependent manner where it associates with cellular membranes.

SopD2 targets late endocytic compartments

To further examine intracellular targeting of SopD and SopD2, we transfected HeLa cells with GFP fusions of both proteins. We observed that SopD-GFP localized exclusively to the cytosol (Figure 5A), consistent with the fractionation of bacterially delivered HA-tagged SopD in infected cells. In contrast, SopD2-GFP localized to vesicular structures that tended to concentrate in the perinuclear area but were also present throughout the cell (Figure 5B). This included association with small tubular structures and large vesicles (see arrowheads). The majority of structures containing SopD2-GFP also colocalized with the lysosomal glycoprotein LAMP-1 (Figure 5C). Interestingly, the tubular structures containing SopD2-GFP appeared as tubules of SopD2-GFP extending from LAMP-1 positive vesicles (see inset). These structures were also observed in association with lysosomes loaded with fluorescent dextran (Figure 5D, see inset). In a similar manner, SopD2-GFP colocalized with the late endosome-specific lipid lysobisphosphatidic acid (Figure 5E) and with cathepsin D (Figure 5F).

Figure 5.

Figure 5.

Association of SopD2-GFP with late endocytic compartments. HeLa cells were transfected with GFP fusions (green) to either SopD (A) or SopD2 (B–F). SopD2-GFP associated with membrane structures in the perinuclear region and throughout the cell, including large vesicles (B, see arrowheads). C. Association of tubulovesicular structures of SopD2-GFP with vesicles containing LAMP-1 (red, see inset). D. SopD2–GFP association with lysosomes loaded with Alexa 568-conjugated dextran (red, see inset). E. Late endosomes were visualized by staining transfected cells with monoclonal antibodies to lysobisphosphatidic acid (red, see inset). F. Colocalization of SopD2-GFP with cathepsin D (red, see inset). Size bar indicates 10 μm.

The Rab5 GTPase regulates early endosome trafficking, including budding, cytoskeletal transport and docking/fusion activities (56). To determine if SopD2 interacts with early endosomes, HeLa cells were cotransfected with a Rab5 fusion to GFP and a mammalian expression vector encoding SopD2 with two tandem C-terminal HA tags. As shown in Figure 6 (upper panels), Rab5-GFP was localized to punctate endocytic vesicles throughout the cell, including the cell periphery. With the exception of areas of perinuclear accumulation, little colocalization of Rab5-GFP and SopD2–2HA was observed. By contrast, SopD2–2HA colocalized extensively with Rab7-GFP (middle panels, see arrows), a regulator of trafficking to late endocytic compartments (56). Expression of the N125I dominant-negative Rab7 mutant (57–59) did not block membrane association of SopD2–2HA (lower panels). This demonstrates that SopD2 associates with late endocytic compartments, including late endosomes and/or lysosomes, in a Rab7-independent manner. Furthermore, these findings demonstrate that SopD2 encodes a targeting signal specific for late endosomes/lysosomes and does not require other bacterial factors for association with these compartments.

Figure 6.

Figure 6.

Colocalization of SopD2-GFP with Rab7. HeLa cells were cotransfected with a vector encoding SopD2 with 2 C-terminal HA epitope tags and GFP fusions to either Rab5, Rab7 or the N125I (dominant-negative) mutant of Rab7, as indicated. Arrows indicate areas of colocalization of SopD2–2HA with Rab7-GFP. Size bar indicates 10 μm.

The N-terminus of SopD2 mediates targeting of late endocytic compartments in host cells and type III secretion by the SPI-2 TTSS

To determine which portion of SopD2 mediates targeting to late endocytic compartments, we generated a series of truncation mutants of this bacterial effector fused to GFP (Figure 7A). Amino acids 150–319 from SopD2 were localized to the cytosol in transfected HeLa cells and did not colocalize with lysosomal glycoproteins (Figure 7B). Other constructs with larger deletions of the N-terminus (e.g. expression of amino acids 250–319) were similarly localized to the cytosol (data not shown). In contrast, amino acids 1–150 were associated with vesicles that costained with LAMP-1 (Figure 7C), lysobisphosphatidic acid (Figure 7D) and cathepsin D (not shown). Interestingly, binding of this SopD2 construct to late endocytic compartments led to their swelling (see arrowheads in Figure 7C,D). It should also be noted that this construct did not display the tubular morphology witnessed with the full-length SopD2-GFP fusion (compare Figure 5B–F with Figure 7C–D). Amino acids 1–125 from SopD2 were similarly localized to swollen late endocytic compartments that were positive for LAMP-1 (Figure 7E). C-terminally truncated peptides that retained only the first 75 or 100 amino acids of SopD2 were also associated with LAMP-1 positive vesicles, but had a lesser effect on late endosome/lysosome swelling (Figure 7F,G). Constructs limited to amino acids 1–50 localized primarily to the cytosol of transfected cells and did not colocalize with LAMP-1 (Figure 7H). These observations demonstrate that the N-terminus of SopD2 is sufficient to mediate late endosome/lysosome targeting and swelling of these compartments.

Figure 7.

Figure 7.

The N-terminus of SopD2 mediates targeting to late endocytic compartments.A. HeLa cells were transfected with the indicated SopD2 truncations fused to GFP for 16–24 h, fixed with paraformaldehyde and co-immunostained with antibodies to LAMP-1 (B, C, E–H) or lysobisphosphatidic acid (D). Arrowheads in C and D indicate swollen late endosomes/lysosomes that result from expression of amino acids 1–150 of SopD2. Size bar indicates 10 μm.

To identify the region of SopD2 needed for type III secretion, HeLa cells were infected with S. typhimurium expressing N-terminal fragments of SopD2 with two C-terminal HA tags. Consistent with other members of the STE family of TTSS effectors (37,38), the N-terminal 200 amino acids of SopD2 was sufficient to mediate its translocation into host cells. As shown in Figure 8(A), SopD2 (1–200)-2HA associated with the SCV at late times post infection (see arrow). Shorter N-terminal fragments of SopD2 (amino acids 1–100 and 1–150) were expressed in S. typhimurium but were not translocated into host cells (data not shown). Subcellular fractionation confirmed that SopD2 (1–200)-2HA was associated with host cell membranes in infected cells (Figure 8B). Thus, the N-terminus of SopD2 is a bifunctional domain that mediates both type III secretion out of Salmonella and late endosome/lysosome targeting upon translocation into host cells.

Figure 8.

Figure 8.

The N-terminus of SopD2 mediates type III secretion from S. typhimurium and membrane association in vivo. A. HeLa cells were infected for 21 h with S. typhimurium expressing the N-terminal 200 amino acids of SopD2 with 2 C-terminal HA epitope tags. Infected cells were then fixed and processed for immunofluorescence imaging of the HA tagged SopD2, which can be seen in association with the S. typhimurium-containing vacuole (arrow). Size bar indicates 10 μm. B. Infected cells were also subjected to subcellular fractionation, as performed in Figure 4(B), (see Experimental Procedures). As shown, the N-terminus of SopD2 was sufficient to direct SPI-2-mediated translocation and membrane binding in infected cells.

Discussion

In this study we have characterized SopD2, a novel SPI-2 effector of S. typhimurium. SopD2 was specifically targeted to late endocytic compartments after delivery into host cells by bacteria. The significance of this targeting is underlined by the fact that SopD2 shares no homology to any identified eukaryotic protein. SopD2 likely represents an example of convergent evolution that has allowed Salmonella to exquisitely manipulate host cell machinery and enact its pathogenic strategy (24). Targeting of SopD2 was mediated by a discrete portion of its N-terminal type III secretion signal, which is conserved in other effectors of the STE family (38). Thus, our findings demonstrate that this signal sequence can mediate intermolecular interactions in the host in addition to its role in type III secretion from the bacteria. Other members of the STE family may also utilize this conserved sequence for intracellular targeting or, possibly in the case of SopD (which was localized to the cytosol), may mediate other intermolecular interactions in the host required for pathogenesis. These findings broaden our understanding of how SPI-2 effectors are targeted within host cells during intracellular occupation by Salmonella. The discovery of a specific late endosome/lysosome targeting sequence will also provide a useful tool for the study of late endocytic trafficking events in eukaryotic cells.

Based on our studies of SopD2-GFP transfected cells, it seems likely that SopD2 contributes to the aggregation and swelling of late endosomes/lysosomes witnessed in infected cells. Indeed, SopD2–2HA associated with swollen endosomes that costained with LAMP-2 at late times post infection (Figure 4A). Because of its ability to target late endocytic compartments, it would be tempting to speculate that SopD2 plays a role in blocking SCV fusion with these compartments in macrophages. However, recent studies have demonstrated that the SPI-2 TTSS is not required for this phenotype (60). Furthermore, fusion of late endosomes/lysosomes with the SCV in epithelial cells occurs at late times post invasion (14), despite the delivery of SopD2 in this cell type. While the molecular function of SopD2 remains undetermined, it is exciting to note that at least three different SPI-2 effectors, SifA, SseJ and SopD2, have been implicated in altering late endocytic trafficking inside host cells (39–41,61). As with cytoskeletal rearrangements mediated by effectors of the SPI-1 TTSS (24), modulation of the host endosomal system is likely to be mediated by a battery of translocated SPI-2 effectors that act in a coordinated fashion.

Our studies revealed that SopD is expressed under conditions that induce both the SPI-1 and SPI-2 TTSS and their effectors. In fact, expression of SopD was comparable to that of SopD2 in infected cells at late times post infection (e.g. Figure 4B). This suggests strongly that SopD is a dual effector for both the SPI-1 and SPI-2 TTSS. It is interesting that despite significant similarity at the level of their amino acid sequence, the two proteins have unique intracellular sites of action in host cells. SopD has previously been shown to play a major role in an animal model of gastroenteritis (19). Our finding that SopD expression is maintained inside host cells is consistent with recent studies demonstrating a significant role for the SPI-2 TTSS in this disease (62,63). Identification of the host cell target(s) of SopD will promote a better understanding of how this effector contributes to gastroenteritis.

In summary, we have identified a novel effector of the SPI-2 TTSS that specifically targets late endocytic compartments after delivery into host cells. Future work will be directed at understanding the mechanisms by which SopD2 targets late endosomes/lysosomes and promotes pathogenesis in coordination with other SPI-2 effectors.

Materials and Methods

Cell culture and bacterial strains

HeLa (human epithelial cell line) cells were obtained from ATCC. Cells were maintained in DMEM (Gibco BRL, Burlington, ON, Canada) supplemented with 10% FCS (Gibco BRL) at 37 °C in 5% CO2 without antibiotics. Cultures were used between passage numbers 10–25. Cells were seeded (5 × 104 cells/well) into 24-well plates with 1-mm sterile coverslips 18 h prior to each experiment. Where indicated, cells were incubated with either 2 μg/ml nocodazole (Sigma, St. Louis, MO, USA) or 5 μg/ml cytochalasin D (Sigma). Wild-type S. typhimurium SL 1344 was used for these studies. The ssaR mutant has been previously described (40). SL1344 ssrA::mTn5 and SL1344 phoP::Tn10d-Tc were constructed by P22 transduction from 14028 s ssrA::mTn5 (37) and 14028 s phoP::Tn10d Tc (64), respectively. SL1344 hilA::kan-339 has been described previously (50). To construct the sopD2 mutant with an in-frame deletion of codons 9–312 (319 total), we first cloned the intact gene and 2 kb of flanking DNA from S. typhimurium SL 1344. This 2.95 kb DNA fragment was obtained by PCR amplification using the Elongase system (Gibco BRL) with the oligonucleotide primers SKC05 (5′-CAG TTC ATC AAT CAC CGG GT-3′) and SKC06 (5′-ACT GAA AAT GCA GGT TGG TCC-3′), cloned into pCRTOPO2.1 vector (Invitrogen, Burlington, ON, Canada), and confirmed by sequencing. The resultant plasmid, pTOPOsopD2, served as the template DNA for inverse PCR using the primers SKC07 (5′-ACG CGT CGA CAC CAA ACT TAA CGT AAC TGG CAT-3′) and SKC08 (5′-ACG CGT CGA CAG TTG TCG CAA TAT GCT TAT ATA-3′). These primers overlap the first and last eight codons of the sopD2 ORF, respectively, and introduce a SalI site (underlined in SKC07 and SKCO8) to replace the intervening coding sequence. The inverse PCR reaction yielded a 6-kb product with SalI sites at the termini, which were subsequently digested and ligated to yield a nonpolar deletion of 912 bp of the sopD2 coding sequence. The remaining 2 kb of S. typhimurium sequence was released by digestion with SacI and XbaI (sites within pCRTOPO2.1), ligated into the corresponding sites of the positive selection suicide vector pRE112 (CmR) (65) and transformed by electroporation into E. coli SY327 λpir (66). The S. typhimurium SL1344 sopD2 mutant was then constructed by allelic exchange as described (65). Briefly, recombinant pRE112 plasmids were purified from SY327 λpir and transformed into SM10 λpir (66). A transformant was conjugated with wild-type SL1344, and SmR/CmR SL1344 colonies were selected, grown for 4 h in LB without antibiotic selection, plated on LB agar containing 5% sucrose and incubated overnight at 30 °C. Sucrose-resistant colonies were chosen and the presence of the in-frame deleted sopD2 gene was confirmed by PCR analysis.

Bacterial infection of mice

Female BALB/c mice (6–8 weeks old) were purchased from Harlan Laboratories (Indianapolis, IN, USA). Mice were kept in sterilized, filter-topped cages, handled in tissue culture hoods and fed autoclaved food and water under specific pathogen-free (SPF) conditions at our animal facilities. Sentinel animals were routinely tested for common pathogens. The protocols used were in direct accordance with guidelines drafted by the University of British Columbia's Animal Care Committee and the Canadian Council on the Use of Laboratory Animals. Bacterial cultures were grown overnight and then diluted in phosphate-buffered saline (PBS), before injection. Groups of five mice were infected by intraperitoneal injection with approximately 5 × 104 cfu in 0.3 ml of PBS. To assess the virulence of the tested strains, mice were monitored throughout the course of the infection, and any that showed extreme distress or became moribund were euthanized. Three such experiments were performed; the results from a typical experiment are presented in Figure 2.

Reverse transcriptase-polymerase chain reaction (RT-PCR)

Salmonella typhimurium was grown under SPI-1- or SPI-2-inducing conditions as described previously (38). Total bacterial RNA was isolated as described (67). For RT-PCR analysis, total bacterial RNA was DNase (Ambion, Austin, TX, USA) treated and 1 μg was reverse transcribed with a gene-specific primer and Superscript II reverse transcriptase (Gibco BRL) according to the manufacturer's instructions. The gene-specific primers were: for sopD, sopD-R1 (5′-AAT AGG AGA TGC ACG GAT GC-3′); for sopD2, sopD2-R1 (5′-GTA GCA GAT AGC TAA GTG TTG G-3′). An aliquot (1/20th) of the cDNA reaction was subject to 35 cycles of PCR amplification using AmpliTaq Gold DNA polymerase (Perkin Elmer, Boston, MA, USA) with the following oligonucleotides: for sopD, sopD-F (5′-AGG ACC ATG CGC TGG AAG TG-3′) and sopD-R2 (5′-ATC GAC TCC TGA AGT AAT T-3′) (359 bp product); for sopD2, sopD2-F (5′-AAC AAG AGG TGC TGG AGG TA-3′) and sopD2-R2 (5′-TTC GAT CTT CTC CAG ATC GC-3′) (443 bp product).

Plasmids and transfection

The low copy plasmid pACYC184 (NCBI accession # X06403) was used to express epitope-tagged SopD1 and SopD2 in S. typhimurium with two C-terminal HA tags.

Primers JBO95 (5′-CGC GGA TCC TTA CGC ATA ATC CGG CAC ATC ATA CGG ATA CGC ATA ATC CGG CAC ATC ATA CGG ATACTC GAG TAT AAG CAT ATT GCG ACA ACT CGA CTT-3′) and JBO91(5′-ACG CGT CGA CGC GGT ACT GCG AGC GTA AAT TTT GGA C-3′) were used to amplify sopD2 and its upstream promoter from a plasmid containing sopD2. The PCR product was digested with BamHI and SalI (sites underlined in JBO95 and JBO91) and cloned into the corresponding sites in pACYC184. This yielded plasmid psopD2–2HA, which includes a XhoI restriction site (underlined in JBO95) preceding two encoded HA epitope tags (bold within JBO95) and a stop codon for the tagged protein. To make HA-tagged SopD, psopD2–2HA was digested with SalI and XhoI to remove sopD2. Next, sopD and its promoter were amplified using primers JBO101 (5′-ACG CGT CGA CTT ATA GTC ACC ACA AAG GAT TAC CAA C-3′) and JBO103 (5′-CCG CTC GAG TGT CAG TAA TAT ATT ACG ACT GCA-3′), digested with SalI and XhoI (sites underlined) and ligated into the HA-tagging vector, generating psopD-2HA. Plasmids were transformed into S. typhimurium by electroporation, and expression of epitope-tagged SopD and SopD2 was confirmed by growth in vitro under SPI-1 [growth in LB to late log phase (12)] or SPI-2-inducing conditions [growth in N-minimal medium to late log/stationary phase (38)], as described previously. Samples of these cultures were sedimented, resuspended in protein sample buffer and subjected to SDS PAGE and immunoblotting on PVDF membranes.

For expression of SopD and SopD2 in HeLa cells, N-terminal fusions to the enhanced GFP mutant were constructed by PCR amplification of sopD and sopD2 from a genomic library of S. typhimurium, using the following primers: for sopD1, JBO109 5′-CCG CTC GAG AGC ATG CCA GTC ACT TTA AGC TTC GGT AAT-3′; JBO110 5′- CGT GGA TCC GTC AGT AAT ATA TTA CGA CTG CAC CC-3′; for sopD2, JBO111 5′- CCG CTC GAG AGC ATG CCA GTT ACG TTA AGT TTT GGT AAT-3′; JBO112 5′- CGT GGA TCC AGT ATA AGC ATA TTG CGA CAA CTC GAC TT-3′. After digestion with XhoI and BamHI (sites underlined above), the PCR products were cloned into the multiple cloning region of the pEGFP-N1 N-terminal protein fusion vector from Clontech (Palo Alto, CA, USA). Truncation mutants of SopD2 fused to GFP were generated by PCR amplification and cloning into the pEGFP-N1 vector, as above for the full-length protein, except that PCR products were digested with Sal1 and BamH1 (sites underlined) prior to ligation. The primers used for amplification of sopD2 fragments were as follows: for amino acids 1–150 of SopD2 (JBO111 and SKCO4 5′-CGT GGA TCC CGA GAA ATT TTT TCT CCA TAC TT-3′); for amino acids 150–319 (SKCO1 5′- CCG GTC GAC AGC ATG AAG TAT GGA GAA AAA ATT TCT CGC − 3′ and JBO112); for amino acids 1–125 (JBO111 and SKCO11 5′-CGT GGA TCC AGA CGA TGC AAA TTA CAT TTA TCA AT-3′), for amino acids 1–100 (JBO111 and SKCO12 5′-CGT GGA TCC ATA ACA AAG CGG TCT TGT TGA GA-3′); for amino acids 1–75 (JBO111 and SKCO13 5′-CGT GGA TCC AGG GTT ATA TCA ACA TTA AGA ACA GC-3′); for amino acids 1–50 (JBO111 and SKCO14 5′-CGT GGA TCC TCT TGT TTT TTG TGT GTT CTG AA-3′).

The GFP-Rab5 construct was provided by Dr C. Roy, Yale University, USA, and the wild-type and N125I Rab7 GFP fusions were provided by Dr A. Wandinger-Ness, University of New Mexico Health Sciences Center, USA. Plasmid DNA was purified using the Midiprep kit from Qiagen (Valencia, CA, USA) and used for transfection of cells with the FuGene 6 transfection reagent (Boerhinger Mannheim, Laval, Quebec, Canada) according to the manufacturer's instructions.

Bacterial infection of cell cultures

Late-log bacterial cultures were used for infections and prepared using a method optimized for bacterial invasion (12). Bacteria were grown for 16 h at 37 °C with shaking and then subcultured (1 : 33) in Luria-Bertani broth for 3 h. Bacterial inoculum were prepared by pelleting at 10 000 × g for 2 min and then directly resuspending in PBS, pH 7.4. The inoculum was diluted and added to cells at a multiplicity of infection of ≈?100 : 1 for 10 min at 37 °C. After infection, extracellular bacteria were removed by extensive washing with PBS and addition of growth medium containing gentamicin (50 μg/ml). For experiments in which cells were incubated for more than 2 h following bacterial infection, the gentamicin concentration was subsequently decreased to 5 μg/ml.

Mechanical fractionation of infected HeLa cells was performed essentially as described (55). Briefly, infected HeLa cells were gently scraped and mechanically disrupted by vigorous passage through a 22-gauge needle using a 1-ml syringe in a buffer containing 3 mm imidazole (pH 7.4), 250 mm sucrose, 0.5 mm EDTA, 2 mm sodium orthovanadate, 5 mm sodium fluoride and protease inhibitors (Boerhringer Mannheim, Mannheim, Germany). A low-speed centrifugation step (3000 × g, 15 min) was used to pellet bacteria and unbroken cells (P fraction in Figure 4), followed by ultracentrifugation (41 000 × g, 20 min) to separate cellular membranes (M) from the cytosol fraction (C). Subcellular fractions were analyzed by SDS PAGE and immunoblotting on PVDF membranes.

Processing cells for immunofluorescence microscopy

Cells were fixed in 2.5% paraformaldehyde in PBS pH 7.2 for 10 min at 37 °C. Fixed cells were washed twice with PBS and permeabilized/blocked by treatment with 0.2% saponin (Calbiochem, San Diego, CA, USA) in PBS containing 10% normal goat serum (SS-PBS) for 1–16 h. Primary and secondary antibodies were overlaid on coverslips in SS-PBS for 1–2 h, followed by three washes with PBS. Coverslips were mounted onto 1-mm glass sides using Mowiol (Aldrich, Wilwaukee, WI, USA). Samples were analyzed using a Zeiss Axiovert S100 TV microscope (63 × objective) attached to a Bio-Rad Radiance Plus confocal microscope. Confocal sections were assembled into flat projections using NIH Image, imported into Adobe Photoshop in RGB format and assembled in Canvas 5 for labeling.

Antibodies

Rabbit polyclonal antibodies to human LAMP-2 were generously provided by Dr M. Fukuda (La Jolla Cancer Research Foundation, USA) (68). Murine monoclonal antibodies to lysobisphosphatidic acid were provided by Dr Jean Gruenberg, Department of Biochemistry, University of Geneva, Switzerland (69). Murine monoclonal anti-human LAMP-1 (clone H4A3), anti-human LAMP-2 (H4B4) and monoclonal rat anti-murine LAMP-1(1D4B) were obtained from the Developmental Studies Hybridoma Bank, Iowa City, USA. Rabbit polyclonal antibodies to human cathepsin D (RC242; Scripps Laboratories, San Diego, CA, USA) and S. typhimurium LPS (Difco, Detroit, MI, USA) were obtained from commercial sources. Mouse monoclonal antibodies to E. coli Dna kinase (which cross-reacts with the S. typhimurium protein) and rabbit polyclonal antibodies to calnexin were purchased from Stressgen (Victoria, BC, Canada). Mouse monoclonal antibodies to the hemagglutinin epitope tag (MMS-101R; Covance, Richmond, CA, USA) were used at a final dilution of 1 : 400 for immunofluorescence, following removal of nonspecific cross-reactivity by incubation with an acetone powder from HeLa cells. Secondary antibodies used were: Alexa 488-conjugated donkey anti-rabbit IgG, Alexa 488-conjugated goat anti-murine IgG, Alexa 568-conjugated donkey anti-rabbit IgG, Alexa 568-conjugated goat anti-murine IgG and Cy5-conjugated donkey anti-rabbit IgG, all from Molecular Probes.

Acknowledgments

The authors wish to thank members of the Finlay lab for careful reading of this manuscript and Dr Sandra Marcus for providing antibodies to SigD. We also thank Drs Ed Miao, Sam Miller and David Holden for providing strains, Drs Craig Roy and Angela Wandinger-Ness for providing Rab reagents and Drs Fukuda and Gruenberg for providing antibodies. Special thanks to Dr Elaine Humphrey of the Electron Microscopy Laboratory, University of British Columbia, for her assistance with confocal microscopy. This work was supported by grants (to BBF) and a postdoctoral fellowship (to JHB) from the Canadian Institutes of Health Research. BBF is an International Research Scholar of the Howard Hughes Medical Institute and a Distinguished Investigator of the Canadian Institutes for Health Research. JHB is an honorary fellow of the Izaac Walton Killam Memorial Foundation.

Ancillary