SEARCH

SEARCH BY CITATION

Keywords:

  • intracellular pathogen;
  • microtubule motor protein;
  • type III secretion system

Abstract

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Intracellular replication of Salmonella enterica requires the formation of a unique organelle termed Salmonella-containing vacuole (SCV). The type III secretion system (T3SS) encoded by Salmonella Pathogenicity Island 2 (SPI2–T3SS) has a crucial role in the formation and maintenance of the SCV. The SPI2–T3SS translocates a large number of effector proteins that interfere with host cell functions such as microtubule-dependent transport. We investigated the function of the effector SseF and observed that this protein is required to maintain the SCV in a juxtanuclear position in infected epithelial cells. The formation of juxtanuclear clusters of replicating Salmonella required the recruitment of dynein to the SCV but SseF-deficient strains were highly reduced in dynein recruitment to the SCV. We performed a functional dissection of SseF and defined domains that were important for translocation and the specific effector functions of this protein. Of particular importance was a hydrophobic domain in the C-terminal half that contains three putative transmembrane (TM) helices. Deletion of one of these TM helices ablated the effector functions of SseF. We observed that this domain was essential for the proper intracellular positioning of the SCV to a juxtanuclear, Golgi-associated localization. These data show that SseF, in concert with the effector proteins SifA and SseG mediate the precise positioning of the SCV by differentially modulating the recruitment of microtubule motor proteins to the SCV.

Salmonella enterica is a facultative intracellular pathogen that resides in a membrane-bound compartment after internalization by host cells. This compartment, referred to as Salmonella-containing vacuole (SCV), deviates from the default endocytic pathway and allows for the massive intracellular replication of S. enterica[reviewed in (1,2)]. The intracellular pathogenesis of Salmonella is a complex, multifactorial trait that involves metabolic flexibility to adapt to nutrient limitation as well as repair mechanisms to compensate for damage induced by host defense mechanisms. In addition, intracellular Salmonella modify the biogenesis of the SCV by interference with normal host cell processes. This modification is dependent on the function of a type III secretion system (T3SS) that is encoded by Salmonella Pathogenicity Island 2 (SPI2) (3). T3SS are complex molecular machines that translocate effector proteins from the bacterial cytoplasm into the host cell [for a review, see (4)]. These effector proteins act as ‘injected toxins’ and modify various host cell processes for the benefit of the pathogen. Salmonella enterica possesses two T3SS, with the SPI1–T3SS involved in invasion of host cells and gastrointestinal pathogenesis and the SPI2–T3SS being active during intracellular life of Salmonella. The SPI2–T3SS is specifically activated by intracellular Salmonella and translocates a group of effector proteins across the membrane of the SCV (3).

Several recent studies have indicated that the successful intracellular life of S. enterica depends on the modification of the host cell microtubule cytoskeleton and the interference with cellular transport processes of the host cell (5–10). In mammalian cells, microtubules are organized in a polarized fashion with the minus-ends located in the cell center and the plus-ends at the cell periphery. The correct functioning of microtubules is essential for both the transport and spatial organization of membrane-bound organelles within the cell. For example, microtubules guide the endoplasmic reticulum, control the distribution of melanophores and depolymerization of microtubules results in a collapse of the Golgi apparatus. The microtubule-associated motor proteins dynein and kinesin are thought to play a key role in this process by facilitating the directional transport of organelles to either the minus- or plus-ends of microtubules, respectively, and in so doing ensure that the correct cellular distribution of the organelles is maintained [reviewed by Hirokawa (11)].

While the function of the SPI2–T3SS is essential for systemic pathogenesis and intracellular lifestyle of S. enterica, the contribution of the several effector proteins to these virulence traits is only partially understood. SifA is an effector that is encoded by genes outside of the SPI2 locus. The function of this protein is important for the induction of tubular aggregates of late endosomal/lysosomal compartments referred to as ‘Salmonella-induced filaments’ (SIF) (12). SIF form along microtubules (13) and the function of SifA is required for the maintenance of the SCV (14) as SifA-deficient Salmonella are released into the host cell cytoplasm. Recent work suggested that SifA is required to prevent the interaction of the microtubule motor protein kinesin with the SCV (10). This interference is thought to be crucial to maintain the integrity of the SCV [for recent reviews, see (15,16)].

The formation and dynamics of SIF is controlled by PipB2 (17) and SopD2 (18), further SPI2–T3SS effector proteins encoded by genes outside the SPI2 locus and by SPI2-encoded SseF and SseG (19). Our group has previously demonstrated that SseF and SseG are associated with late endosomal/lysosomal membrane compartments as well as with microtubules. It was found that both effectors were required for the induction of SIF (20). In the absence of SseF or SseG, the aggregation of endosomal compartments to SIF was incomplete and so-called ‘pseudo-SIF’ were observed (19). Furthermore, the formation of massive bundles of microtubules was observed in infected HeLa cells. This phenotype was dependent on the function of SPI2 and in particular required the function of SseF and SseG (6). Work by Salcedo and Holden (7) showed that SseG is required to direct the SCV to a juxtanuclear, trans-Golgi network (TGN)-associated subcellular position. It was also observed that translocated SseG is targeted to the TGN, a finding that was not corroborated by another study (6). The molecular functions of SseF have not been analyzed in detail.

Based on the interference of intracellular Salmonella with microtubules, we speculated that the SPI2–T3SS and more specifically the effector proteins may control the trafficking or subcellular localization of the SCV as a prerequisite for intracellular proliferation. We set out to determine the contribution of these SPI2 effectors to the positioning of the SCV and the interference with the organization and function of the microtubule cytoskeleton. We observed that SseF is required for the positioning of the SCV to a juxtanuclear position and the formation of tight clusters of replicating bacteria, previously referred to as microcolonies (21). Our observations further suggest that these phenotypes are mediated by the SseF-dependent recruitment of the dynein motor complex to the SCV.

Results

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Formation of juxtanuclear microcolonies requires the function of the SPI2–T3SS

In order to investigate the dynamics of the intracellular replication and formation of microcolonies of Salmonella in mammalian cells, HeLa cells were infected with Salmonella wild-type (WT) or a SPI2 null mutant strain (ssaV) and the subcellular localization of the SCV was examined. We observed that the majority of internalized bacteria appeared in a juxtanuclear localization within 2 h after infection. This initial juxtanuclear localization was independent of the function of the SPI2–T3SS (Figure 1A). A large number of SCV harboring wild-type Salmonella (WT SCV) remained in a juxtanuclear position throughout the time course of the experiment and, at 16 h post-infection (p.i.), 81 ± 5% of the WT SCV formed microcolonies positioned in the juxtanuclear region. By contrast, whereas most of the SCV of the ssaV mutants strains were similarly located in the juxtanuclear region up to 4 h p.i., notable scattering of the SCV was observed at 6 h p.i., as well as all later time points examined. At 16 h p.i. the majority of ssaV SCV displayed a scattered distribution throughout the cytoplasm, with only 19 ± 1% forming discernible microcolonies in the juxtanuclear region. In order to investigate which SPI2 effector proteins were responsible for this phenotype, HeLa cells were infected with Salmonella harboring mutations in genes for various effector proteins. Similar to the WT strain, sseJ, sseI, pipB2, sopD2, sifB, sspH1, sspH2 and slrP mutant strains accumulated in the juxtanuclear region of infected cells up to 16 h p.i. (data not shown). By contrast, Salmonella strains deficient in sseF or sseG displayed a predominantly scattered distribution in infected HeLa cells (Figure 1B,C). To analyze if the observed effect was restricted to the epithelial cell line HeLa, we also analyzed the localization of SCV in the fibroblast cell line COS-7 and the murine macrophage-like cell line RAW264.7 after infection with WT Salmonella and various mutant strains. A reduction of juxtanuclear SCV formation for mutant strains deficient in ssaV, sseF or sseG was observed in COS-7 cells, similar to the phenotype in HeLa cells (Figure 1D). As macrophages are important host cells for Salmonella, we also investigated the formation of microcolonies in the macrophage-line cell line RAW264.7. However, this analysis was hampered by the more compact cell morphology of RAW264.7 cells and phagocytic uptake of multiple bacteria (data not shown). Thus, we focused on HeLa cells as an infection model.

image

Figure 1. Role of SPI2 in positioning of intracellular Salmonella.(A) HeLa cells were infected with S. typhimurium wild-type (WT) and an ssaV strain deficient in SPI2–T3SS function (SPI2) (LPS, green). At various time points post-infection (p.i.), cells were fixed and immunostained for Salmonella LPS (green) and the trans-Golgi network using antibodies against Golgin97 (red). Micrographs show merged immunofluorescence and phase-contrast images for representative infected host cells. Scale bar: 10 µm. (B) HeLa cells were infected with WT S. typhimurium, the ssaV strain or various strains deficient in genes encoding SPI2 effector proteins. Cells were fixed 16 h p.i. and immunostained as for panel (A). Cells showing representative phenotypes for S. typhimurium WT and mutant strains deficient in ssaV, sseF, sseG or sseJ are presented. Scale bars correspond to 10 µm. (C) HeLa cells were infected with various S. typhimurium strains and 16 h p.i., the appearance and subcellular localization of the bacteria were analyzed. Infected cells with clusters of replicating bacteria (microcolonies) in a juxtanuclear position were scored positive, while infected cells with a scattered distribution of individual bacteria or microcolonies in the cell periphery were scored negative. At least 50 infected cells per infecting strain were analyzed and mean values and standard deviation of three independent experiments are shown. (D) COS-7 cells were infected with WT S. typhimurium and various mutant strains at an MOI of 50. Quantification of juxtanuclear microcolonies was analyzed 16 h after infection basically as described for HeLa cells. The means and standard deviations of three independent experiments are shown. Statistical analyses of mutant strains versus WT: *, p < 0.001; ns, not significant.

To test if intracellular replication of Salmonella is affected by the positioning of the SCV, we compared the proliferation of Salmonella WT with strains deficient in ssaV, sseF and a plasmid-complemented sseF strain (Figure 2). These analyses indicated that the intracellular proliferation of an sseF strain is reduced to an extent similar to that of the ssaV strain deficient in the translocation of the entire set of SPI2 effector proteins. The intracellular proliferation of the sseF strain could be restored by plasmid-borne SseF. A comparable effect on intracellular proliferation has been previously reported for effector protein SseG (7).

image

Figure 2. Role of SseF for intracellular proliferation of Salmonella in HeLa cells. HeLa cells were infected at an MOI of 1 with Salmonella wild-type (WT), ssaV and sseF strains, and the sseF strain harboring a plasmid for the complementation of the sseF mutation. At 2 and 14 h after infection, infected host cells were lysed by addition of Triton-X-100 and the colony-forming units (CFU) of intracellular Salmonella were quantified by plating serial dilutions onto agar plates. The data show representative mean CFU/mL of lysates and the standard deviation of mean and are representative of three independent experiments.

Effector protein SseF affects the distribution of the motor protein dynein

The role of SPI2 in the positioning of the SCV described above and previous observations on the interference of SPI2 effector proteins with microtubule organization prompted us to analyze the distribution of microtubule motor proteins. In accordance with previous observations (8), dynein was found to accumulate in the immediate vicinity of the bacterial microcolonies (clusters consisting of five or more bacteria) in HeLa cells infected with WT Salmonella for 16 h (Figure 3A). Dynein recruitment to the SCV was not detectable prior to 8 h after infection and the proportion of cells displaying dynein recruitment to the SCV increased over time from about 8% to 62% at 8 and 12 h p.i., respectively. The observation of dynein recruitment to the SCV [(8), this study] is in contrast to the work of Boucrot et al. (10) that did not reveal dynein recruitment to the SCV at 10 h after infection. A further study (9) did not indicate dynein recruitment to the SCV, but the analyses were performed at 3 h p.i. After infection with the ssaV or the sseF strain, the number of microcolonies was highly reduced and intracellular Salmonella frequently appeared scattered throughout the host cells as shown for an sseF-infected HeLa cell (Figure 3AsseF scattered’). For the analyses of dynein recruitment, only those host cells were scored that showed formation of microcolonies as shown in Figure 3sseF microcolony’. Microcolonies formed by the ssaV strain, which is deficient in translocation of all known SPI2–T3SS effector proteins, displayed a marked decrease in its ability to recruit dynein. To identify the SPI2 effector proteins involved in the recruitment of dynein to the SCV, strains harboring mutations in genes for various SPI2 effector proteins were examined (Figure 3B). While mutant strains deficient in SseJ, SseI, PipB2, SopD2, SifB, SspH1, SspH2 or SlrP were found to accumulate dynein at a level similar to that found for the WT strain, SseF- or SseG-deficient strains accumulated dynein at a markedly reduced level. The inability of the SseF-deficient strain to recruit dynein was shown to be specific, because a plasmid encoding SseF could restore the ability of the SseF mutant strain to recruit dynein to the SCV (Figure 3B, ‘sseF[sseF ]’).

image

Figure 3. SPI2 effectors mediate the recruitment of microtubule motor dynein to Salmonella microcolonies. HeLa cells were infected with various S. typhimurium strains and the distribution of cytoplasmic dynein was analyzed. Cells were fixed 16 h post-infection (p.i.) and processed for staining of Salmonella (green) or dynein (red). (A) Confocal micrographs showing dynein distribution in uninfected cells (mock) or cells infected with wild-type (WT) Salmonella or the indicated mutant strains. Note the accumulation of dynein at a microcolony of the WT strain. The frequency of microcolony formation was 81 ± 5% for the WT strain and 19 ± 1% for the ssaV strain. Microcolony formation was also reduced after infection with sseF or sseG strains. Infection with these mutant strains frequently resulted in the appearance of a scattered distribution of the intracellular bacteria (shown for sseF‘scattered’). For the subsequent analysis of dynein accumulation, only those infected cells were considered that showed typical microcolonies of at least five bacteria (as shown for sseF‘microcolony’). Scale bars represent 10 µm. (B) The recruitment of dynein to microcolonies was quantified. After infection with various Salmonella strains, cells harboring microcolonies were identified and scored for the appearance of condensations of dynein. At least 50 microcolonies per infecting strain were scored and the means and standard deviation for three independent experiments are shown. Statistical analyses of mutant strains versus WT: *, p < 0.001; ns, not significant.

Formation of juxtanuclear microcolonies requires the function of microtubule motor dynein

To investigate the role of dynein in more detail, we performed experiments to interfere with dynein function. Dynein activity can be inhibited by sodium o-vanadate (22,23), and we analyzed the formation of juxtanuclear microcolonies in the presence of this compound. In HeLa cells infected with WT Salmonella, the preincubation of cells with o-vanadate resulted in a dose-dependent reduction of microcolony formation (Figure 4A). o-Vanadate at a non-toxic concentration of 100 µm is frequently used to inhibit dynein function (24,25) and as we also did not observe negative effects on host cells, this concentration was used for further studies. To further investigate whether dynein function is required for formation of juxtanuclear microcolonies, o-vanadate was added at various time points after infection and the number of microcolonies quantified 16 h p.i. (Figure 4B). o-Vanadate addition at early time points after infection caused a strong reduction in microcolony formation, while the effect of the inhibitor decreased at later time points of addition. We observed a linear dependency between the time point of inhibition of dynein function and the formation of microcolonies, indicating that dynein function is continuously required for formation of the SCV and the growth of microcolonies.

image

Figure 4. Formation of Salmonella microcolonies requires dynein motor activity.(A) HeLa cells were infected with wild-type (WT) Salmonella. Various amounts of sodium o-vanadate as indicated were added 30 min prior to infection to HeLa cells and maintained throughout the experiment. The cells were fixed 12 h post-infection (p.i.) and the formation of juxtanuclear microcolonies was quantified by analyses of at least 100 infected cells. Statistical analyses: *, p < 0.001. (B) HeLa cells were infected with WT Salmonella and 100 µm sodium o-vanadate was added at various time points after infection as indicated. Cells were fixed 12 h after infection and the formation of juxtanuclear microcolonies was quantified in at least 100 infected cells per time point. Statistical analyses (inhibitor versus mock: *, p < 0.001; ns, not significant. (C) To further investigate the role of the dynein motor complex in SCV localization, the formation of juxtanuclear microcolonies of WT Salmonella was quantified after transfection of HeLa cells with constructs for expression of p50/dynamitin, RILP or RILP-C33. As a control, HeLa cells were transfected with the vector pEGFP-N3 and infected with WT or SPI2-deficient Salmonella. The formation of juxtanuclear microcolonies was quantified 16 h after infection for at least 50 transfected and infected cells. Statistical analyses of constructs versus vector transfection: *, p < 0.001; ns, not significant.

Dynein function can also be altered by overexpression of the dynactin complex subunit p50/dynamitin (26) or the presence of a dominant-negative allele of the Rab7 effector protein RILP (27). We generated a transfection construct for expression of p50/dynamitin–EGFP and observed that in transfected HeLa cells the organization and subcellular localization of the Golgi and late endosomal/lysosomal compartments was dramatically altered, a phenotype consistent with the disruption of the dynactin complex and dynein motor function (28) (see also SupplementaryFigure S1). Overexpression of p50/dynamitin was found to reduce the formation of juxtanuclear SCV by approximately 40% relative to cells transfected with a control EGFP-expressing plasmid (Figure 4C). Similar to previous observations (8), disruption of dynein function by overexpression of p50/dynamitin was also found to inhibit the intracellular replication and the ability of Salmonella to form SIF in infected cells (data not shown). In addition, expression of dominant-negative RILP-C33 strongly reduced the formation of juxtanuclear microcolonies, while expression of WT RILP had no effect (Figure 4C). A similar level of dynein recruitment to microcolonies formed by WT Salmonella was observed in cells transfected with control and RILP vectors. For most of the microcolonies formed in cells transfected with the RILP-C33 or p50/dynamitin constructs, dynein recruitment was not detectable (data not shown). Taken together, these observations indicate that the function of the dynactin–dynein motor complex is required for the formation of juxtanuclear microcolonies of replicating Salmonella.

Functional dissection of SseF

The biological function of the SPI2 effector protein SseF is only partially understood and the observed function of SseF in modifying microtubule motor distribution suggested a novel role for this protein. In order to investigate the molecular functions of SseF, a deletional analysis was performed. We previously observed a strict association of translocated SseF with endosomal membrane systems containing lysosomal glycoproteins (lgp) such as LAMP-1 or LAMP-2 (19). Furthermore, subcellular fractionation revealed the presence of translocated SseF in the membrane fraction of the host cell. Sequence analyses of SseF indicated the presence of several hydrophobic regions that may act as transmembrane (TM) domains (Figure 5A). To gain further insight into the potential role of these domains in the subcellular localization and biological function of SseF, a series of in-frame, HA epitope-tagged deletion mutants of SseF consisting of eight internal deletions (SseFΔ1 to SseFΔ8) and four C-terminal deletions (SseFΔC1 to SseFΔC4) were generated (see Figure 5B for schematic representation of the deletions). The internal deletions were designed to delete, entirely or in part, the putative TM domains. To examine whether the various mutant proteins were synthesized by Salmonella in vitro, the various constructs were introduced into a mutant strain deficient in synthesis of SseF (strain HH107) and the bacterial strains grown under conditions previously shown to induce the expression of SPI2 genes. Western blot analysis with an antibody to the HA epitope revealed single bands of the predicted molecular weights for each of the deletion constructs, barring SseFΔ1 and SseFΔ2, which did not appear to give rise to detectable amounts of protein under the experimental conditions used (data not shown, results summarized in Figure 5C).

image

Figure 5. Functional dissection of SPI2 effector protein SseF.(A) Hydropathy plot of SseF and localization of putative membrane-spanning domains. The hydrophobicity was calculated using the Kyte–Doolittle algorithm and putative transmembrane (TM) segments were predicted using TMpred. Two TM segments were predicted: TM1 (aa 69–96) with one TM helix and TM2 (aa 131–205) with three TM helices. (B) Schematic representation of mutant alleles of sseF analyzed in this study. A set of plasmids was constructed harboring wild-type (WT) sseF or various mutant alleles. In all constructs, a sequence encoding the HA tag was fused to the 3′ end of sseF. The positions of internal deletions in sseF (sseFΔ1 to sseFΔ8) and C-terminal truncations of sseF (sseFΔC1 to sseFΔC4) are indicated by arrows and dashed lines. The extent of the deletion is further specified by the first and last deleted codon. (C) Analysis of synthesis and translocation of mutant constructs of SseF. For analyses of expression, S. typhimurium WT harboring plasmids for the expression of WT sseF or various mutant alleles was grown in PCN (non-inducing media) and PCN-P minimal media (inducing media). Equal amounts of bacterial cells were harvested and processed for Western blot analyses with an antibody against the HA tag. The presence or absence of SseF-HA or mutant proteins is indicated by + and –, respectively. For analyses of translocation, HeLa cells were infected with S. typhimurium WT harboring the various plasmids. At 16 h p.i., the cells were fixed and processed for immunostaining of Salmonella LPS and the HA tag. The presence or absence of translocated SseF is indicated by + or –, respectively (for further details, refer to SupplementaryFigure S2). To determine the localization of the C terminus of SseF and derivatives, HeLa cells were infected with the sseF-deficient strain harboring plasmids for the expression of WT sseF (WT) or various mutant alleles of sseF. Sixteen hours after infection, cells were subjected to digitonin permeabilization and immunostaining of the HA tag was performed. CP indicates that the HA tag at the C-terminal of SseF and derivatives was accessible to immunostaining; nd, not determined.

Translocation of the epitope-tagged derivatives of SseF by intracellular Salmonella was next analyzed by confocal microscopy following infection of HeLa cells. In accordance with previous results, full-length epitope-tagged SseF was translocated by intracellular Salmonella, where it was found to co-localize extensively with LAMP-1 or LAMP-2 present in the membranes of the SCV, as well as in SIF. A similar intracellular distribution was found for four of the internal deletion mutants (SseFΔ5 to SseFΔ8, see SupplementaryFigure S2 for representative infected cells). In contrast, no signal for HA-tagged proteins was observed for mutant proteins encoding either the first N-terminal 55 amino acids (aa) of SseF (SseFΔC1), or those lacking TM1 either in part (SseFΔ4) or in its entirety (SseFΔ3), suggesting a lack of translocation by Salmonella of these particular mutant proteins or their rapid degradation within eukaryotic cells. The mutant alleles of sseF were also subcloned into pTre-Tight for inducible expression after eukaryotic cell transfection. After transfection, the constructs SseFΔ1 to SseFΔ8 were all detected by immunofluorescence, indicating that the epitope tag of SseFΔ3 and SseFΔ4 is accessible in eukaryotic cells, but that these constructs cannot be translocated by intracellular Salmonella. We observed that SseFΔ1 to SseFΔ8 preferentially co-localized with TGN markers after expression by transfection vectors (SupplementaryFigure S3). Thus, the subcellular distribution of SseFΔ5 to SseFΔ8 after bacterial translocation was different from those of transfection constructs. A similar observation of an artificial Golgi localization after transfection has been made for SseG (6), indicating that eukaryotic transfection may be of limited use for the study of these particular SPI2–T3SS effectors. To test if the various deletions result in a major alteration of the topology of the SseF derivatives, localization of the C-terminus was probed by immunostaining for the HA tag after selective permeabilization of the plasma membrane. The C-terminus of WT SseF and of various mutant proteins was accessible from the cytoplasm (Figure 5C, SupplementaryFigure S4), suggesting that the deletions did not cause massive alterations of the topology of SseF. For further studies, constructs SseFΔ5 to SseFΔ8 and SseFΔC1 to SseFΔC4 were used for functional analyses in the background of the sseF strain.

A hydrophobic region in SseF is essential for effector functions

The formation of SIF in Salmonella-infected HeLa cells was shown to be dependent on a functional SPI2–T3SS. Although the sseF strain induced the formation of filamentous structures in infected cells, these displayed only a punctate distribution of lgp markers, as opposed to the continuous distribution found in cells infected with the WT strain. This phenotype was referred to as pseudo-SIF (19). Consistent with previous results, cells infected with strain HH107 expressing WT sseF induced SIF that displayed a continuous distribution of LAMP-2 (SupplementaryFigure 2). In contrast, an sseF mutant strain was severely compromised in its ability to induce the continuous distribution of LAMP-2 along SIF. The formation of SIF was fully restored in the sseF mutant strain by complementation with a plasmid harboring the epitope-tagged, full-length version of sseF. In order to determine whether the translocated mutant proteins were impaired in their ability to form SIF, HeLa cells were infected with WT and various mutant strains and the formation of SIF and pseudo-SIF was quantified (Figure 6A,B). Similar to results observed for plasmid-borne WT SseF, SIF were predominantly observed when cells were infected with strains harboring the SseFΔ8 or SseFΔC4 constructs. In contrast, the number of SIF observed in Salmonella strains translocating proteins that lacked part of, or the entire, TM2 (SseFΔ5 to SseFΔ7 and SseFΔC2 and SseFΔC3) was markedly reduced, a phenomenon that was accompanied by an increase in the number of pseudo-SIF formed.

image

Figure 6. A short hydrophobic segment in SseF is essential for SseF-mediated phenotypes. To investigate the role of SseF and its deletion derivatives in Salmonella-induced filament (SIF) formation and microtubule bundling, HeLa cells were infected with S. typhimurium wild-type (WT) or mutant strains deficient in ssaV, sseF, sseG, sseI or sseJ (A,C), or the sseF strain harboring plasmids for the expression of WT sseF or various mutant alleles (B,D). Sixteen hours post-infection (p.i.), cells were fixed and processed for immunostaining of S. typhimurium and LAMP-2, and the formation of SIF (filled bars) or pseudo-SIF (open bars) was enumerated by confocal microscopy (A,B). SIF and pseudo-SIF formation was enumerated for at least three experiments with at least 50 infected cells per infecting strain, and the values shown are the average ± the standard deviation for four experiments. For analyses of Salmonella-induced microtubule bundling, cells were fixed 16 h p.i. and processed for immunostaining of Salmonella and β-tubulin (C,D). Cells infected with Salmonella were analyzed for the appearance of the microtubule (MT) cytoskeleton and cells showing massive bundles of MT were scored as positive for MT bundling. For each strain, at least 50 infected cells were scored and the percentage of infected cells showing bundling of microtubules is given. Statistical analyses in A and C of mutant strains versus WT, and in B and D of the sseF strain complemented with various mutant alleles versus WT sseF: *, p < 0.001; ns, not significant.

The infection of HeLa cells with WT Salmonella also induced the formation of massive bundles of microtubules in a fraction of the host cells, and previous work demonstrated that this phenotype is dependent on the function of the SPI2–T3SS and a subset of SPI2 effectors including SseF (6). In accordance with the previous study, SPI2–T3SS effectors other than SseF and SseG did not contribute to this phenotype (Figure 6C,D and data not shown). We analyzed the effect of deletions of various domains of SseF for the induction of microtubule bundling (Figure 6D). The deletion of sseF could be complemented by plasmid-borne WT SseF as well as by SseFΔ8 or SseFΔC4. Translocation of SseF derivatives with deletions affecting the TM2 domain, however, resulted in highly reduced microtubule bundling. These observations indicate that the TM2 domain is essential for the effector function of SseF.

A hydrophobic domain in SseF is essential for positioning of intracellular Salmonella and recruitment of dynein to Salmonella-containing vacuoles

In order to investigate the molecular requirements of SseF in the positioning of the SCV, we investigated the effect of mutations in SseF on the formation of juxtanuclear microcolonies in HeLa cells (Figure 7A,C). The reduced appearance of microcolonies was restored to WT levels by complementation with plasmid-borne WT SseF. Microcolony formation similar to that in WT Salmonella-infected cells was also observed for the sseFΔ8 and sseFΔC4 alleles, while the strains expressing other sseF alleles showed a highly reduced appearance of microcolonies (Figure 7C).

image

Figure 7. A hydrophobic domain in SseF is required for dynein recruitment to the SCV and intracellular positioning. HeLa cells were infected with the sseF mutant strain harboring plasmids for the expression of wild-type (WT) sseF or various mutant alleles. (A) Immunostaining was performed for Salmonella LPS (green) and TGN (stained for Golgin97, red) as described in the legend to Figure 1. Micrographs of merged images with the phase contrast image show the typical appearance of Salmonella in infected cells. Scale bar: 10 µm (B) Immunostaining for Salmonella LPS (green) and dynein (red) was performed with cells fixed 16 h p.i. The distribution of dynein is shown in representative cells infected with the sseF strain complemented with sseF, sseFΔ5 or sseFΔ8 with microcolony formation. (C) The formation of juxtanuclear microcolonies as shown in panel (A) was quantified. The histograms show the mean values with standard deviation obtained from at least three independent experiments. (D) Dynein accumulation at microcolonies was quantified after infection with the sseF harboring plasmids for the expression of WT sseF or various mutant alleles. Representative infected cells observed after infection with sseF[sseFΔ5] or sseF[sseFΔ8] are shown in (B). For each infecting strain, 50 host cells harboring microcolonies were scored for dynein distribution. The average ± the standard deviation was enumerated for at least three independent experiments. Statistical analyses of the sseF strain complemented with various mutant sseF alleles versus WT sseF: *, p < 0.001; ns, not significant.

Our initial observations (Figures 1 and 3) suggested a relationship between the SPI2-dependent recruitment of dynein to the SCV and the formation of juxtanuclear microcolonies. To further investigate this correlation, we established whether strains translocating any of the mutant SseF derivatives were defective in their ability to recruit dynein to the Salmonella microcolonies (Figure 7B). The percentage of microcolonies associated with this motor protein in cells infected with the sseF strain harboring plasmids for expression of WT sseF or mutant alleles was enumerated (Figure 7D). Whereas the sseF strain translocating SseFΔ8 recruited dynein at levels similar to the complemented sseF mutant strain, deletion mutants lacking regions located within TM2 showed reduced levels of dynein association at the SCV as found in the sseF mutant strain. Reduced dynein recruitment was observed for all mutant SseF derivatives including sseFΔ7 with the smallest deletion within the TM domain (Figure 7D).

image

Figure 8. Models for the organization of functional domains in SseF and the role of SseF in the intracellular fate of Salmonella. (A) Functional domains in SseF. Based on the mutational analyses and functional assays, a domain structure of SseF is proposed. An N-terminal region (aa 12–49) is essential for the stability of SseF. SscB, the specific chaperone of SseF may bind in this region. Domain TM1 (aa 63–110) is required for the secretion and translocation of SseF. Essential effector functions of SseF are located in the TM2 domain (aa 128–179). However, this domain is not important for the subcellular localization of SseF. No important function was attributed to the C-terminal part (aa 210–260), and this region appeared dispensable for stability, translocation and all cellular phenotypes analyzed here. (B) Model for inference of intracellular Salmonella with microtubule motor proteins. Dynein and kinesin motors mediate minus-end- and plus-end-directed transport of cellular cargo along microtubules, respectively (a). After internalization, endosomes containing Salmonella enter a subcellular localization in juxtaposition to the nucleus and Golgi. By activities of effectors of the SPI2–T3SS, the SCV is modified and replication initiates (b). The function of SifA is required to interfere with kinesin activity and the integrity of the SCV containing a sifA mutant strain is lost (c). Effector proteins SseF and SseG are required to maintain the SCV in a juxtanuclear position. SseF is required for the recruitment of dynein to the SCV and SCV containing the sseF strains are displaced to the cell periphery (d). Formation of an SCV allowing Salmonella replication requires a fine-tuned balance between recruitment of dynein and kinesin by the concerted activity of SifA, SseF and SseG.

Taken together, these data demonstrate that the integrity of the hydrophobic TM2 domain of SseF is crucial for the recruitment of dynein to the SCV, the subcellular localization of the SCV and the ability to form replicative clusters of bacteria in a juxtanuclear position within host cells.

Discussion

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

The intracellular life of Salmonella depends on a highly dynamic interaction with transport processes of the host cell. Internalized bacteria have to avoid the default endocytic pathway and the effect of antimicrobial mechanisms of the host cells. In addition, the successful intracellular phase of Salmonella requires the access to nutrients and a constant supply of membrane compartments to allow for extension of the SCV. In this work, we demonstrate that intracellular WT Salmonella can manipulate the positioning of the SCV in such a manner that the formation of juxtanuclear microcolonies is favored, while mutant strains with defects in intracellular pathogenesis predominantly showed a dispersed distribution of SCV containing single bacteria.

A number of recent studies have demonstrated that intracellular Salmonella interferes with both microtubule- and motor protein-dependent functions, and that this interference plays an essential role in ensuring the proper maturation of the SCV (5,8,10). In line with these observations, our group observed that a subset of effector proteins of the SPI2–T3SS, including SseF and SseG, is targeted to microtubules and are involved in the alteration of the structure of the microtubule cytoskeleton, which can lead to the formation of massive bundles of microtubules (6). The data presented in this study suggest that the SCV behaves in a manner analogous to that of many other cellular organelles, in that its intracellular positioning is controlled in a host cell motor protein-dependent manner. By following the fate of WT Salmonella and various mutant strains in infected cells, we could show that the bacteria assume a juxtanuclear location 2 h after infection. It was reported that the dynactin–dynein complex is required for the initial juxtanuclear localization of the SCV (5) and we observed that this process was independent from the SPI2 phenotype of the infecting strain. While the majority of WT Salmonella remained in the juxtanuclear region where replicative clusters developed, strains deficient in the SPI2–T3SS frequently assumed an aberrant intracellular position. The SCV containing the SPI2 strain frequently displayed a dispersed distribution within the cytoplasm, in stark contrast to the focal microcolonies formed by WT Salmonella during intracellular growth. This phenotype was found to specifically depend on the effector proteins SseF and SseG, as mutant strains deficient in the translocation of other effector proteins, other than SifA, behaved in a manner similar to the WT strain with respect to their intracellular localization.

Our results demonstrate that the function of SseF is required for the recruitment of dynein to the SCV. Dynein recruitment also required the function of SseG, an effector protein previously described as being required for the targeting of Salmonella to the Golgi apparatus in infected cells (7). While a subset of sseF- and sseG-deficient Salmonella was still able to form microcolonies in a juxtanuclear position, the SCV containing these strains were mostly devoid of detectable levels of dynein recruitment. In contrast, SCV harboring clusters of WT Salmonella usually displayed a distinct accumulation of dynein in the immediate vicinity of the SCV. As dynein is the major minus-end-directed motor protein responsible for the transport to, and subsequent retention of organelles in the juxtanuclear region of the cell, our results suggest that the dispersal of these strains arises due to their inability to actively recruit dynein to the membrane of the SCV. Support for this hypothesis was provided by the observation that that inhibition of dynein motor activity results in the reduced formation of juxtanuclear microcolonies, resembling the phenotypes observed for an sseF- or sseG-deficient strain.

Based on these findings, we propose that the recruitment of dynein to the SCV is associated with the proper positioning of the SCV in infected cells, a process that facilitates and promotes the intracellular replication and formation of microcolonies of Salmonella. The retrograde movement of bacteria towards the microtubule-organizing center (MTOC) was demonstrated for several facultative intracellular pathogens such as Campylobacter jejuni (24), Orientia tsutsugamushi (29) and Chlamydia trachomatis (30,31) and an involvement of dynein was observed. In the latter example, transport of C. trachomatis inclusions to the juxtanuclear region was, however, shown to occur independently of the dynactin complex, seeing that overexpression of p50/dynamitin had no effect on the juxtanuclear positioning of intracellular bacteria. This led to the proposal that chlamydial proteins are involved in the recruitment of dynein to the membrane of the inclusion in a manner that is independent of the cargo-binding subunits of dynactin. Previous observations by Harrison et al. (5) indicated that, following invasion of host cells, the dynactin–dynein complex is required for initial transport of the SCV to a juxtanuclear location. A similar requirement for recruitment of the dynactin–dynein complex in the maintenance of the SCV within the juxtanuclear region was observed in this study, because increased bacterial dispersal was observed in response to overexpression of either RILP-C33 or p50/dynamitin. This would suggest that the recruitment of dynein to the SCV occurs in a manner mechanistically distinct from that observed for C. trachomatis. At present, the molecular mechanisms underlying the SseF- and SseG-mediated recruitment of dynein to the SCV are unknown. Possible explanations could be the modification of the vacuolar membrane that increases its affinity for the dynein motor protein or, alternatively, the interaction with upstream effectors involved in the recruitment of dynein to membrane-bound cargo. Further work aimed at distinguishing between these possibilities is currently underway.

We performed a mutational analysis of SseF to identify domains of this effector important for the intracellular phenotypes. SseF derivatives with deletions in the N-terminal region, i.e., aa 12–49 or 39–49 (SseFΔ1 and SseFΔ2, respectively) were neither detectable in Salmonella nor translocated into host cells, but expressed after transfection. Recent work by Dai and Zhou (32) demonstrated that SscB functions as chaperone for SseF and is required for the stability of SseF in the bacterial cytoplasm as well as for the efficient secretion. The deletions of N-terminal regions of SseF may interfere with SscB binding and render the protein sensitive to proteolysis.

Our previous studies have demonstrated that translocated SseF is targeted to both the SCV and SIF. The first 127 aa of SseF are sufficient to mediate translocation and subsequent targeting of SseF to these intracellular compartments. Mutations affecting the integrity of TM1 (aa 63–110) also affected the translocation of the mutant protein or their stability after translocation (e.g., compare SseFΔC1 with SseFΔC2 or SseFΔ3 and SseFΔ4 with SseFΔ5, Figure 6) and prevented further analyses of this region with respect to host cell phenotypes. It is likely that aa 60–127 play a critical role in the translocation of SseF. A possible role of the region aa 1–127 in functioning either as targeting determinants or in additional modulation of SseF activity following translocation will have to be addressed by future mutagenesis experiments that alter the primary sequence of SseF without compromising its stability and/or translocation.

Our functional dissection of SseF further revealed that the effector functions critically depend on the integrity of a complex hydrophobic domain encompassed by TM2 (for a model, see Figure 8A). Even a short deletion of 33 aa located within TM2 resulted in the loss of SseF-dependent phenotypes. While another experimental setup was used for analysis of SseG, an effector sharing 30% identity with SseF, a hydrophobic region was found to be essential for the function of this effector (7). Interestingly, deletions in the TM2 region of SseF not only affected the recruitment of dynein and consequently the intracellular positioning of the SCV, but also other SseF-dependent phenotypes such as SIF formation and microtubule bundling (6,19). This may indicate that the loss of SseF-mediated phenotypes is due to the same basic defect in SseF function. While the actual function of TM2 is unknown, it is possible that disruption of TM2 interferes with the ability of SseF to interact with host cell proteins required for its proper function, or with other SPI2–T3SS effectors. Alternatively, this region may be important for the proper integration of SseF into endosomal membranes. These possibilities will need to be addressed by further studies.

Many organelles possess the ability to move in a bidirectional manner along microtubules and regulate their cellular positioning by controlling the recruitment or relative activities of motor proteins [reviewed by Welte (33)]. A recent study by Boucrot et al. (10) described the role of SifA, another SPI2–T3SS effector protein, in preventing the excessive accumulation of kinesin at the SCV. This effect was mediated by the recruitment of SKIP by SifA, and the interference of the SKIP–SifA complex with kinesin accumulation. Furthermore, the recruitment of kinesin to the SCV was postulated to be an active process mediated by SPI2-encoded translocated effector proteins other than SifA, given that kinesin accumulation to the SCV was only rarely observed in a strain deficient for the translocation of all known SPI-2 effectors. In contrast to the study of Boucrot et al. (10), we observed a prominent recruitment of dynein to the SCV containing WT Salmonella. Dynein accumulation was reduced on SCV containing a SPI2–T3SS null mutant strain or mutants deficient in sseF or sseG, but dynein was present on SCV containing strains deficient in other effectors of the SPI2–T3SS. Taken together, these observations suggest that the positioning and the integrity of the SCV depend on the SPI2-dependent modulation of both dynein and kinesin recruitment. The proper intracellular localization of SCV and intracellular replication requires the recruitment of the minus-end-directed motor protein dynein through the action of SseF and SseG and the prevention of plus-end-directed motor protein through the action of SifA. Only the interference with both motor activities allows for the proper positioning of the SCV and efficient replication within host cells.

Based on our results on the role of SseF, and previous studies on SifA and SseG we propose a model for the concerted action of the three effector proteins (Figure 8B). Translocated SifA interferes with the excessive recruitment of kinesin to the SCV and this activity prevents the displacement of the SCV towards the cell periphery, as well as the disruption of the SCV by the pulling force of the plus-end-directed kinesin motor complex. SseF and SseG counteract this activity by recruiting dynein as a motor protein with an opposite, minus-end-directed activity. The simultaneous activity of these effector proteins results in the correct balance of plus- and minus-end-directed activities being maintained, thereby ensuring the steady-state positioning of the SCV in a predominantly juxtanuclear location. A defect in the translocation of either one of these translocated effector proteins results in plus-end-directed activity predominating, which manifests itself in the increased centrifugal movement of the SCV. We therefore postulate that the successful intracellular proliferation of Salmonella depends on the prevention of microtubule-dependent movement of the SCV that will lead to dispersal of the bacteria.

It is well established that the maturation of the SCV occurs via selective interactions with the endocytic pathway. Recent observations in our laboratory have indicated that Salmonella is also capable of interfering with exocytic events in an SPI2-dependent manner (34), an activity that may contribute to the biogenesis of the SCV. The ability of Salmonella to manipulate the microtubule cytoskeleton and its associated motor proteins to traffic towards, and subsequently establish a replication niche near the minus-ends of microtubules, is likely to place the SCV in close apposition to a variety of organelles such as late endosomes, lysosomes and the Golgi apparatus, which maintain a predominantly juxtanuclear position at steady state. Such an arrangement would allow Salmonella to interfere with both endocytic and exocytic trafficking events, thereby allowing it to gain privileged access to membrane compartments required for vacuolar membrane biogenesis and SIF formation, as well as nutrients required for intracellular growth that may otherwise be limiting within the intracellular environment.

Further evaluation of the mechanisms by which SPI2-encoded effector proteins exploit both host cell motor proteins and intracellular trafficking pathways to facilitate its own growth and replication is likely to provide a better understanding of intracellular bacterial pathogenesis and of basic eukaryotic cell functions.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Bacterial strains

Bacterial strains and plasmids used in this study are listed in Table 1. Salmonella enterica serotype typhimurium (S. typhimurium) strain 12023 was used as the WT strain and all mutant strains used are isogenic derivatives thereof. Routine cloning and plasmid propagation were performed using Escherichia coli strains DH5-α or XL-1 blue. Bacteria were routinely cultured in LB broth or on LB agar plates supplemented with carbenicillin (50 µg/mL) and/or kanamycin (50 µg/mL) when required. To test the expression of sseF alleles, cultures were grown in non-inducing (PCN) and inducing (PCN-P) minimal media as previously described (35).

Table 1. Bacterial strains and plasmids used in this study
Strain or plasmidRelevant characteristicsReference
Salmonella enterica serovar typhimurium strains
 NCTC 12023Wild-typeNCTC, Colindale
 P2D6ssaV::mTn5(38)
 HH107ΔsseF::aph(39)
 HH108ΔsseG::aph(39)
 MvP373ΔsscB sseFG::aph(6)
 MvP377ΔsseJ::aph(40)
 MvP378ΔsseI::aph(40)
Plasmid
 p2643PsseAsscB sseF::HA in pWSK29(6)
 p3009PsseAsscB sseFΔ1::HA in pWSK29This study
 p3010PsseAsscB sseFΔ2::HA in pWSK29This study
 p3011PsseAsscB sseFΔ3::HA in pWSK29This study
 p3012PsseAsscB sseFΔ4::HA in pWSK29This study
 p3013PsseAsscB sseFΔ5::HA in pWSK29This study
 p3014PsseAsscB sseFΔ6::HA in pWSK29This study
 p3015PsseAsscB sseFΔ7::HA in pWSK29This study
 p3016PsseAsscB sseFΔ8::HA in pWSK29This study
 p3093PsseAsscB sseFΔC1::HA in pWSK29This study
 p3094PsseAsscB sseFΔC2::HA in pWSK29This study
 p3095PsseAsscB sseFΔC3::HA in pWSK29This study
 p3096PsseAsscB sseFΔC4::HA in pWSK29This study
 p3153p50/dynamitin in pEGFP-N3This study
 pEGFP-RILPEGFP-RILP(8)
 pEGFP-RILP-C33EGFP-RILP-C33(8)

Recombinant DNA methods

DNA manipulations were performed according to standard procedures (36). DNA restriction and modification enzymes were purchased from MBI Fermentas (St Leon-Rot, Germany) and used in accordance with the manufacturer's instructions. PCR reactions were performed using proof-reading polymerases (‘High Fidelity PCR Enzyme Mix’, MBI Fermentas) so as to minimize the introduction of errors during the amplification process. Genomic DNA, plasmids, PCR products, and DNA fragments were purified using Qiagen (Hilden, Germany) kits according to the instructions of the manufacturer. Plasmid constructs were introduced into E. coli and S. typhimurium competent cells by electroporation.

For the creation of a p50/dynamitin-GFP-expressing plasmid, HeLa cDNA was generated using total RNA extracted from HeLa cells with the Total RNA Isolation Kit (MBI Fermentas) and the first-strand synthesis kit (MBI Fermentas). The complete p50/dynamitin cDNA was subsequently PCR amplified using primers p50-Dynamitin-EcoRI-For and p50-Dynamitin-BamHI-Rev, into which the appropriate restriction sites for cloning were introduced. The resulting PCR product was cloned between the EcoRI and the BamHI sites of the pEGFP-N3 vector (BD Clontech, Heidelberg, Germany) and confirmed by DNA sequencing.

Generation of epitope-tagged derivatives of SseF

The construction of p2643, which encodes a full-length version of SseF harboring a C-terminal HA epitope tag from the low-copy vector pWSK29, has been previously described (6). The construction of in-frame deletion derivatives of sseF was performed by the ‘splice-by-overlap-extension’ PCR method, using p2643 as a template. Oligonucleotides are listed in Table 2. Briefly, first-round PCR reactions were performed using primer SscB109-SmaI-For in conjunction with the relevant reverse primers [SseF12-Rev, SseF39-Rev (for SseFΔ1 and SseFΔ2, respectively) SseF63-Rev, SseF87-Rev (for SseFΔ3 and SseFΔ4, respectively), SseF127-Rev, SseF148-Rev, SseF179-Rev (for SseFΔ5, SseFΔ6 and SseFΔ7, respectively) or SseF227-Rev (for SseFΔ8)], or with the T7-terminator primer in conjunction with the relevant forward primer [SseF49-For (for SseFΔ1 and SseFΔ2), SseF110-For (for SseFΔ3 and SseFΔ4), SseF212-For (for SseFΔ5, SseFΔ6 and SseFΔ7) or SseF253-For (for SseFΔ8)]. PCR fragments generated in the first-round PCR were gel purified, and those containing the corresponding overlapping ends were combined and used as a template in a second-round PCR using the SscB109-SmaI-For and T7-terminator primer as the forward and reverse primers, respectively. For generation of C-terminal deletions of SseF, primer SscB109-SmaI-For was used in conjunction with primer SseF60-HA-XbaI-Rev, SseF120-HA-XbaI-Rev, SseF180-HA-XbaI-Rev or SseF210-HA-XbaI-Rev. Each of the resulting PCR products were gel purified, digested with SmaI and XbaI and cloned into SmaI/XbaI-digested plasmid p2643 from which the WT sseF gene had been excised. All of the constructions were confirmed by DNA sequencing.

Table 2. Primers used in this study
DesignationSequence (5′−3′)
SseF12-RevATCCCTCTGCTGCCTTATTTGTTCTATATTACTTGCCGCTGACGGAAT
SseF39-RevATCCCTCTGCTGCCTTATTTGTTCGGTGCCAGGCGCTGGAATTTCAGG
SseF49-ForGAACAAATAAGGCAGCAGAGGGAT
SseF63-RevTATCGATTGATAATTATGATACGCTTGCATAAAATGTATCGCATAATC
SseF87-RevTATCGATTGATAATTATGATACGCCCCGCCAGAAATTACCGCTGCAGC
SseF110-ForGCGTATCATAATTATCAATCGATA
SseF127-RevATTTTCCTGATCGTCGCCAGAGGGGGCGGTTTGTAATGGCTCCTTTTG
SseF148-RevATTTTCCTGATCGTCGCCAGAGGGGCAGTTAAGACTTGCCCCACATTT
SseF179-RevATTTTCCTGATCGTCGCCAGAGGGCGCGGGCAGTGGAAACTGTAGGGG
SseF212-ForCCCTCTGGCGACGATCAGGAAAAT
SseF227-RevTCCCCGAGATGTATGATCAGAACTATCGGCATGAAGTTCATCAACAGA
SseF253-ForAGTTCTGATCATACATCTCGGGGA
SseF60-HA-XbaI-RevGAGTCTAGATTAAGCGTAGTCTGGGACGTCGTATGGGTACCTCTGC
TGCCTTATTTG
SseF120-HA-XbaI-RevGAGTCTAGATTAAGCGTAGTCTGGGACGTCGTATGGGTAACTGG
CGGTTTGTAATG
SseF180-HA-XbaI-RevGAGTCTAGATTAAGCGTAGTCTGGGACGTCGTATGGGTACAAAG
AGGCCGCAATATTT
SseF210-HA-XbaI-RevGAGTCTAGATTAAGCGTAGTCTGGGACGTCGTATGGGTAGGCATGAA
GTTCATCAAC
T7 terminator primerTATGCTAGTTATTGCTCAG
SscB109-SmaI-ForGGAACCCGGGTTGGCGAGAG
p50-Dynamitin-EcoRI-ForATAGAATTCATGGCGGACCCTAAATACGCC
p50-Dynamitin-BamHI-RevATAGGATCCCTTTCCCAGCTTCTTCATCCG

Cell culture

The human epithelial cell line HeLa were obtained from ATCC and used between passage numbers 5 and 25. Cells were grown in Dulbecco's modified Eagle's medium (DMEM, PAA Laboratories, Cölbe, Germany) supplemented with 10% heat-inactivated fetal calf serum (FCS, Sigma, Taufkirchen, Germany) and 2 mm glutamine at 37 °C in an atmosphere of 5% CO2.

Bacterial infection of HeLa cells

For infection experiments, HeLa cells were seeded on glass coverslips in 24-well tissue culture plates at a density of about 8 × 104 cells/well 24 h before infection. Salmonella typhimurium strains were grown in LB broth containing the necessary antibiotics at 37 °C with agitation to stationary phase. The cultures were then diluted 1:30 with fresh LB broth and incubated for another 3.5 h at 37 °C with agitation to reach late logarithmic phase. The OD600 of the cultures was adjusted with LB to 0.2 and the bacteria were washed once with phosphate-buffered saline (PBS). Cells were then diluted in DMEM containing FCS and glutamine and added to the HeLa cells at a multiplicity of infection of 10. The bacteria were centrifuged onto the cells at 500 × g for 5 min and incubated for 25 min at 37 °C in 5% CO2. After infection, the epithelial cells were washed three times with PBS and incubated for 1 h in medium containing FCS, glutamine and 100 µg/mL gentamicin. The medium was replaced with medium containing FCS, glutamine and 10 µg/mL gentamicin for the remainder of the experiment. At least 50 infected host cells were quantified for each condition in each experiment, and all experiments were repeated three times. For the quantification for intracellular proliferation, the same infection procedure was used but at various time points after infection, cells were washed with PBS and lysed by addition of 1 mL PBS containing 0.1% Triton-X-100. Serial dilutions of the lysates were plated onto Mueller–Hinton agar plates to determine the number of viable intracellular bacteria.

Transfection of HeLa cells

HeLa cells were seeded on glass coverslips in 24-well tissue culture plates 24 h before transfection at a density of 4 × 104 cells/well. Cells were transiently transfected with Polyfect Transfection Reagent (Qiagen) using 0.2–0.4 µg DNA according to the manufacturer's instructions. Cells were subsequently incubated at 37 °C in 5% CO2 for the indicated time periods prior to fixation and antibody staining.

Immunofluorescence

For immunofluorescence, cells were fixed in 3% p-formaldehyde in PBS for 15 min at room temperature. For immunostaining of dynein, cells were fixed in methanol for 5 min at −20 °C, and then washed three times with PBS. Antibodies were diluted in a blocking solution consisting of 2% bovine serum albumin (BSA)/2% goat serum and 0.1% saponin (Sigma) in PBS. Cells were incubated with primary antibodies at the recommended dilution for 1–3 h at RT, washed three times with PBS, and incubated for 1 h with the appropriate Cy2-, Cy3- or Cy5-conjugated secondary antibodies. Coverslips were mounted on Fluoroprep (bioMèrieux, Nürtingen, Germany) and sealed with Entellan (Merck, Darmstadt, Germany). Samples were analyzed using a confocal laser scanning microscope (TCS-NT Leica, Bensheim, Germany) or with an Axiovert 200 microscope equipped with an Axiocam MR and an Apotome (Zeiss, Göttingen, Germany). Image analyses were performed using axiovision 4.3 software (Zeiss) and Adobe Photoshop.

Selective permeabilization of the plasma membrane with digitonin

To analyze the localization of the HA epitope-tagged C terminus of WT SseF and mutant proteins, the previously described digitonin treatment (37) was modified. All incubation steps were carried out on ice in ice-cold solutions. The cells were washed twice in KHM buffer (110 mm KAc, 20 mm Hepes pH 7.2, 2 mm MgAc) and then incubated for 5 min in KHM, containing 10 µg/mL digitonin (Fluka, Taufkirchen, Germany). The detergent was removed and the cells were incubated for 20 min in KHM without digitonin to allow permeabilization. After a further washing step with KHM, fixation was performed with 3% PFA. The subsequent immunostaining was carried out in blocking solution without saponin. To control the procedure, immunostaining of Salmonella within the SCV was performed after permeabilization with digitonin or saponin.

Antibodies

The following primary antibodies were used at the specified dilutions: rabbit anti-Salmonella-O4 Bacto testsera (Difco by BD, Heidelberg, Germany), 1:1000; rat anti-HA (Roche, Mannheim, Germany), 1:500; 1:300; mouse anti-human LAMP-1, 1:100 (clone A4H3, DSHB, Iowa City, IA, USA); mouse anti-human LAMP-2 (clone H4B4, DSHB), 1:500; mouse anti-human LAMP-3 (clone H5C6, DSHB), 1:500; mouse anti-dynein (clone 1618, Chemicon, Chandlers Ford, UK) 1:100; mouse anti-β-tubulin Cy3 (Sigma), 1:200; mouse anti-human Golgin97 (Molecular Probes by Invitrogen, Karlsruhe, Germany), 1:300.

Fluorochrome-conjugated secondary antibody were obtained from Dianova (Hamburg, Germany) and used at the following dilutions: donkey anti-rat Cy2, 1:500; goat anti-rabbit Cy2 at 1:1000; goat anti-mouse Cy3 1:500; goat anti-rat Cy5 at 1:200; goat anti-mouse Cy2 at 1:300; rabbit and goat anti-mouse Cy5, 1:200.

Statistical analyses

Statistical analyses were performed by one-way anova using SigmaStat 3.1. Statistical significance was defined as p < 0.001.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

This work was supported by grants HE1964/9-1 and 9-2 as part of the priority program ‘Signal pathways to the cytoskeleton and bacterial pathogenicity’ of the Deutsche Forschungsgemeinschaft. GLA was a recipient of a post-doctoral fellowship from the National Research Foundation of South Africa. MH likes to thank the ‘Fonds der Chemischen Industrie’ for support. The initial observations on dynein recruitment were made by Volker Kuhle and his contribution is kindly acknowledged. We like to thank Cecillia Bucci and Trina Schroer for providing plasmids for transfection.

Supplementary Material

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Figure S1. Overexpression of human p50 Dynamitin-EGFP affects Golgi organization and subcellular localization of endosomes. HeLa cells were transfected with a plasmid for the expression of a p50/Dynamitin-EGFP fusion (green). Cells were fixed 16 h after transfection and processed or immunostaining for LAMP1 (red, upper panel) or Golgin97 (red, lower panel). Note the displacement of LAMP-1-positive compartments to the cell periphery and the disruption of peri-nuclear trans-Golgi stack in transfected cells.

Figure S2. Translocation of SseF and mutant derivatives of SseF by intracellular Salmonella. HeLa cells were infected with the sseF-deficient strain harboring plasmids for the expression of wild-type sseF (WT) or various mutant alleles of sseF. Cells were fixed 16 h after infection and processed for immunostaining for Salmonella LPS (blue), HA-tagged SseF (red) and LAMP-1 (green). Scale bars represent 10 µm.

Figure S3. Localization of SseF and mutant derivatives of SseF after transfection. HeLa cells were transfected with plasmids for the expression of WT sseF::HA or various mutant alleles. The expression was induced by addition of doxycyclin directly after transfection. Cells were fixed 24 h after transfection and immunostained for the HA epitope tag (red) and the TGN marker Golgin97 (green). Representative transfected cells are shown. Note the preferential accumulation of SseF and derivatives in the Golgi. Similar observations were made with other SseF derivatives (SseFΔ2, SseFΔ4, SseFΔ6, SseFΔ7, data not shown).

Figure S4. Analysis of the subcellular localization of SseF and mutant derivatives of SseF after translocation. (A) To control the selectivity of the permeabilization procedure, HeLa cells were infected with Salmonella typhimurium wild-type (WT) and fixed 16 h after infection. Cells were permeabilized with 1 mg/mL saponin or 10 µg/mL digitonin as indicated. Subsequently, immunostaining of Salmonella LPS (blue) and host cell β-tubulin (red) was performed. For the detection of Salmonella-infected cells, DAPI staining was performed (purple) and intracellular bacteria are indicated by arrows. (B) S. typhimurium sseF strains harboring plasmids for the expression of WT sseF or various mutant alleles of sseF were used to infect HeLa cells. Infected cells were fixed 16 h after infection and subjected to permeabilization by saponin or digitonin as indicated. Immunostaining was performed for Salmonella LPS (blue), the HA tag to detect translocated SseF (green) and host cell β-tubulin (red). Representative infected cells are shown. Scale bars correspond to 5 µm.

These materials are available as part of the online article from http://www.blackwell-synergy.com

References

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgments
  7. Supplementary Material
  8. References
  9. Supporting Information

Cloning and analysis of transfection constructs for sseF and mutant alleles For expression analysis following transfection, DNA fragments harboring WT and deletion variants of SseF were digested with SmaI and XbaI and cloned into doxycycline-inducible eukaryotic expression vector, pTre-Tight vector (BD Clontech, Heidelberg, Germany), which had been digested SmaI and XbaI. Expression of genes under control of the doxycycline-regulated promoter was achieved via the addition of 100 – 1000 ng/ml doxycycline (BD Clontech) in order to adjust the expression levels of the different constructs. The resulting constructs are listed in Table S1.

Fig. S 1. Over-expression of human p50 Dynamitin-EGFP affects Golgi organization and subcellular localization of endosomes. HeLa cells were transfected with a plasmid for the expression of a p50/Dynamitin-EGFP fusion (green). Cells were fixed 16 h after transfection and processed or immuno-staining for LAMP1 (red, upper panel) or Golgin97 (red, lower panel). Note the displacement of LAMP-1-positive compartments to the cell periphery and the disruption of peri-nuclear trans Golgi stack in transfected cells.

Fig. S 2. Translocation of SseF and mutant derivatives of SseF by intracellular Salmonella. HeLa cells were infected with the sseF -deficient strain harboring plasmids for the expression of wild-type sseF (WT) or various mutant alleles of sseF. Cells were fixed 16 h after infection and processed for immuno-staining for Salmonella LPS (blue), HA-tagged SseF (red) and LAMP-1 (green). Scale bars represent 10 μm.

Fig. S 3. Localization of SseF and mutant derivatives of SseF after transfection. HeLa cells were transfected with plasmids for the expression of WT sseF ::HA or various mutant alleles. The expression was induced by addition of doxycyclin directly after transfection. Cells were fixed 24 h after transfection and immuno-stained for the HA epitope tag (red) and the TGN marker Golgin97 (green). Representative transfected cells are shown. Note the preferential accumulation of SseF and derivatives in the Golgi. Similar observations were made with other SseF derivatives (SseFΔ2, SseFΔ4, SseFΔ6, SseFΔ7, data not shown).

Fig. S 4. Analysis of the subcellular localization of SseF and mutant derivatives of SseF after translocation. A) To control the selectivity of the permeabilization procedure, HeLa cells were infected with S. Typhimurium wild type and fixed 16 h after infection. Cells were permeabilized with 1 mg/ml saponin or 10 μg/ml digitonin as indicated. Subsequently, immuno-staining of Salmonella LPS (blue) and host cell β-tubulin (red) was performed. For the detection of Salmonella-infected cells, DAPI staining was performed (purple) and intracellular bacteria are indicated by arrows. B) S. Typhimurium sseF strains harboring plasmids for the expression of wild type sseF or various mutant alleles of sseF were used to infect HeLa cells. Infected cells were fixed 16 h after infection and subjected to permeabilization by saponin or digitonin as indicated. Immuno-staining was performed for Salmonella LPS (blue), the HA tag to detect translocated SseF (green) and host cell β-tubulin (red). Representative infected cells are shown. Scale bars correspond to 5 μm.

FilenameFormatSizeDescription
tra_506_sm_7_8f_Fig1.jpg276KSupporting info item
tra_506_sm_7_8f_Fig2.jpg667KSupporting info item
tra_506_sm_7_8f_Fig3.jpg411KSupporting info item
tra_506_sm_7_8f_Fig4.jpg573KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.