OUP user menu

Expressed Salmonella antigens within macrophages enhance the proliferation of CD4+ and CD8+ T lymphocytes by means of bystander dendritic cells

Masahiro Eguchi, Yukie Sekiya, Yuji Kikuchi, Akiko Takaya, Tomoko Yamamoto, Hidenori Matsui
DOI: http://dx.doi.org/10.1111/j.1574-695X.2007.00275.x 411-420 First published online: 1 August 2007

Abstract

ATP-dependent Lon protease-deficient Salmonella enterica serovar Typhimurium (strain CS2022) appeared to invade successfully the mesenteric lymph nodes (MLN) and Peyer's patches (PP) of BALB/c mice and appeared to be easily eradicated by the host after oral immunization. As detected by flow cytometry, the population of major histocompatibility complex class I (MHC-I)-expressing macrophages and dendritic cells (DCs) was increased in the PP of mice immunized with CS2022 on day 6 after immunization. Thereafter, the population of splenic surface CD69+ T lymphocytes prepared from mice immunized with CS2022 6 weeks prior to measurement increased as a result of the administration of the extracellular vesicles of RAW264.7 macrophage-like cells derived by Salmonella challenge. In addition, the proliferation of CD8+ and even of CD4+T cells isolated from mouse spleens immunized with CS2022 was enhanced after cocultivation with naive DCs in the presence of the extracellular vesicles. These findings indicate that the extracellular vesicles prepared from the Salmonella-challenged macrophages carried salmonellae antigens to bystander DCs, thereby stimulating T-cell responses. Therefore, as antigen presentation after phagocytosis should be a central process in the T-cell activation that occurs in response to Salmonella infection, an oral immunization with CS2022 sufficiently induces T cell-mediated immunity in mice.

Keywords
  • Salmonella pathogenicity island 1 proteins
  • dendritic cells
  • CD4 and CD8T lymphocytes
  • macrophage
  • Salmonella live vaccine

Introduction

ATP-dependent protease Lon-deficient Salmonella enterica serovar Typhimurium (strain CS2022) has been found to exhibit increased production and secretion of three Salmonella pathogenicity island 1 (SPI1) proteins, namely SipA, SipC, and SipD. This strain is thereby capable of dramatically enhancing its expression of hilA, which is thought to function as a central regulator of SPI1 gene expression (Takaya et al., 2002, 2005a). Indeed, Lon has been shown to target both HilC and HilD, which are positive regulators of hilA expression (Takaya et al., 2005a). In vitro experiments have revealed that the expression of invasion genes associated with CS2022 induces dramatic increases in the ability of the bacteria to invade Intestine-407 cells (Takaya et al., 2002). However, because Lon-deficient S. enterica serovar Typhimurium exhibits a decreased ability to survive within HEp-2 epithelial cells, Lon protease has also been thought to play an important role in the intracellular survival of S. enterica serovar Typhimurium after its invasion of epithelial cells (Boddicker & Jones, 2004). Moreover, CS2022 has been found to be unable to either survive or proliferate within murine macrophages owing to its enhanced susceptibility to an oxygen-dependent killing mechanism-associated respiratory burst and to the low phagosomal pH (Takaya et al., 2003).

Salmonella infection induces antigen-specific CD4+ and CD8+ T-cell and B-cell responses, all of which can contribute to protective immunity (Ravindran & McSorley, 2005). A better understanding of T-cell activation during Salmonella infection should therefore lead to the development of more effective vaccines. In the present study, we show that an oral immunization with CS2022 induced the expression of major histocompatibility complex (MHC) molecules by macrophages and dendritic cells (DCs) in Peyer's patches (PP) on day 6 after immunization, and thereafter the 6-week-primed CD4+ and CD8+ T cells were activated by the addition of salmonellae antigens prepared from Salmonella-challenged macrophages and the naive DCs.

Materials and methods

Bacterial strains and growth conditions

Salmonella enterica serovar Typhimurium SR-11 χ3306 (virulent strain, resistant to nalidixic acid) and CS2022 (vaccine strain, lon::chloramphenicol resistance) were routinely grown in L-broth (1% Bacto Tryptone [Difco, Detroit, MI], 0.5% Bacto Yeast extract [Difco], 0.5% sodium chloride, 0.1% glucose, pH7.4) or on L-agar, respectively, at 37°C (Matsui et al., 2003; Kodama & Matsui, 2004; Kodama et al., 2005). To differentiate the rate of replication of the strain used for vaccination from that of the virulent strain during infection, the temperature-sensitive marker plasmid pHSG422 (Hashimoto-Gotoh et al., 1981) was used as previously described (Gulig & Doyle, 1993; Gulig et al., 1997) in strains CS2022 and χ3306. In order to culture the strains containing pHSG422, the bacteria were incubated at 30°C as previously described (Yamamoto et al., 2001). When appropriate, the following antibiotics were added to the cultures: ampicillin (100 µg mL−1), nalidixic acid (25 µg mL−1), and chloramphenicol (30 µg mL−1).

Oral inoculation of mice

Six-week-old female BALB/c (H-2d) mice (Charles River Japan, Yokohama, Japan) were orally inoculated with 1 × 108 CFU of exponential-growth-phase salmonellae of the CS2022 or χ3306 strain, which were concentrated in 20-µL doses in phosphate-buffered saline (PBS), pH 7.4, containing 0.01% (w/v) gelatin (BSG), as described previously (Gulig & Doyle, 1993; Matsui et al., 2000). One or 6 days later, the mesenteric lymph nodes (MLN), PP, and spleens were removed, homogenized in BSG, and plated on L-agar containing the relevant antibiotics. Infections were reported as follows: log10 total CFU, log10 CFU carrying pHSG422, and log10 in proportion with CFU carrying pHSG422 per total CFU of MLN, PP, or spleen sample (Gulig & Doyle, 1993; Gulig et al., 1997).

MHC analysis

The expression levels of MHC class I (MHC-I) and MHC class II (MHC-II) were evaluated by flow cytometry analysis by referring to previously published procedures (Yrlid & Wick, 2002) as follows. Mixed-cell suspensions prepared from PP or spleens of five mice from each group (naive or 6-day either χ3306- or CS2022-infected mice) by passage over sterile glass wool were placed onto 96-well plates (2 × 106 cells well−1, in triplicate), and the cells were triply labelled for 1 h at 4°C with R-phycoerythrin (PE)-conjugated antimouse H-2 Kd(MHC-I) (clone, SF1-1.1; BD Pharmingen, San Jose, CA) used at a dilution of 1 : 200, with fluorescein isothiocyanate (FITC)-conjugated anti-MHC-II(I-A/I-E) (clone, M5/114.15.2; eBioscience, San Diego, CA) used at a dilution of 1 : 200, and with biotin-conjugated antimouse CD11b (clone, M1/70; BD Pharmingen) used at a dilution of 1 : 100 or with biotin-conjugated antimouse CD11c (clone, HL3; BD Pharmingen) used at a dilution of 1 : 500. The biotinated cells were subsequently labelled with streptavidin-PE-Cy5 (eBioscience) for 1 h at 4°C used at a dilution of 1 : 1 000. Finally, 1 × 105 cells obtained from the labelled 2 × 106 cells well−1 were analysed by multi-flow cytometry (EPICS ELITE; Beckman-Coulter, Fullerton, CA) using the expo 32 elite software package (Beckman-Coulter).

CD4+ and CD8+ T-cell analyses of splenic cells

Mouse splenocytes were placed onto 96-well plates (2 × 106 cells well−1, in triplicate) and were labelled with FITC-conjugated anti-CD4 (clone, RM4-5; BD Pharmingen) used at a dilution of 1 : 200, or with R-PE-Cy5-conjugated antimouse CD8a (clone, 53–6.7; eBioscience) used at a dilution of 1 : 200. The population of CD4+ or CD8+ T cells among samples containing 1 × 105 cells was measured by flow cytometry (Lee & Woodward, 1996). Three mice were used, and each spleen was analysed independently in the same experiment.

Preparation of extracellular vesicles released by Salmonella-challenged RAW264.7 cells

Macrophage infection experiments were performed using murine macrophage-like RAW264.7 cells (ATCC TIB-71), as described previously (Takaya et al., 2005b). Briefly, RAW264.7 cells maintained in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich, St Louis, MO) containing 8% heat-inactivated fetal calf serum (FCS) and 50 mM β-mercaptoethanol were seeded onto 24-well plates (5 × 105 cells well−1) and were challenged with χ3306 or CS2022 at a multiplicity of infection (MOI) of 10 salmonellae per cell. The challenged cell monolayers were incubated for 30 min at 37°C to permit phagocytosis. The cell monolayers were rinsed three times with Hank's balanced salt solution (HBSS; Sigma-Aldrich) to remove the free bacteria. DMEM containing 8% FCS and 100 µg of gentamicin (Gibco BRL, Grand Island, NY,) per milliliter was added to the cell monolayers, which were then incubated for 1.5 h at 37°C to kill the extracellulr bacteria. The cell monolayers were rinsed with HBSS three times, followed by incubation with DMEM containing 8% FCS and 10 µg gentamicin mL−1. To prepare the extracellular vesicles from the χ3306- or CS2022-challenged RAW264.7 cells, the culture supernatants obtained 24 h after challenge were collected by consecutive centrifugations at 800 g for 10 min, 1800 g for 10 min, and 25 000 g for 10 min at 4°C. The final supernatants were precipitated by ultracentrifugation (100 000 g for 1 h at 4°C). High-speed centrifugal pellets were suspended in DMEM (390 mg protein mL−1) and were stored at −85°C until use (Schaible et al., 2003). The prepared extracellular vesicles were free of live salmonellae, as verified by culture.

T-cell proliferation assay in splenocytes

T-cell proliferation assay in splenocytes was performed by referring to previously published procedures (Yrlid et al., 2001a) as follows. The mixed splenocytes prepared separately from three BALB/c mice from each group (naive or 6-week CS2022 immunization) were seeded onto 96-well plates (2 × 106 cells well−1, in triplicate), and these cultures were simultaneously supplemented with extracellular vesicles prepared from CS2022-challenged RAW264.7 cells (final concentration: 20 ng protein mL−1). The plates were incubated for 6 h at 37°C in 5% CO2. Each T-cell population was analysed by multi-flow cytometry using the expo 32 elite software package after triple labelling had been carried out for 1 h at 4°C with FITC-conjugated antimouse CD4, R-PE-Cy5-conjugated antimouse CD8a, and biotin-conjugated antimouse CD69 (clone, H1.2F3, BD Pharmingen, San Jose, CA), each of which was used at a dilution of 1 : 200. The biotinated cells were subsequently labelled for 1 h at 4°C with streptavidin-PE (Molecular Probes, Eugene, OR) used at a dilution of 1 : 200. The CD69 activation antigen is a phosphorylated 28- to 32-kDa disulfide-linked homodimer that is rapidly induced after T-cell activation (Testi et al., 1989).

Preparation of DCs from bone marrow

DCs were generated from the bone marrow cells of uninfected BALB/c mice according to the method of Schaible et al., (2003) as follows. The bone marrow cells were seeded onto 12-well plates (10 wells per mouse) and were cultured for 1 week in RPM-I (Sigma) containing 10% FCS supplemented with interleukin-4 (IL-4; 2 ng mL−1; R & D Systems, Minneapolis, MN) and granulocyte macrophage colony stimulating factor (GM-CSF; 10 ng mL−1; R & D Systems) at 37°C in 5% CO2. After cultivation, the CD11c-positive cells were purified as DCs by a positive selection method using magnetic cell sorting (MACS) columns (Miltenui Biotec, Bergisch Gladbach, Germany) following the manufacturer's protocol.

T-cell proliferation assay in the DC culture

T-cell proliferation assay in the DC culture was performed by referring to previously published procedures (Schaible et al., 2003) as follows. The CD4+ and CD8+ T cells were isolated from splenocytes of three to five mice at week 6 after immunization with CS2022 by a positive selection method using MidiMACS columns (Miltenui Biotec) or AutoMACS columns (Miltenui Biotec) following the manufacturer's protocol. The purified DCs (5 × 104 cells well−1) were seeded onto 96-well plates in triplicate, and the cells were cultured overnight in DMEM containing 10% FCS supplemented with the extracellular vesicles prepared from either χ3306- or CS2022-challenged RAW264.7 cells (final concentration: 20 or 200 ng protein mL−1) at 37°C in 5% CO2. The isolated CD4+ or CD8+ T cells (5 × 105 cells well−1) were then added to the DC culture, and the mixed-cell cultures were incubated for 3 days at 37°C in 5% CO2. Before being harvested, the T cells were pulsed with [3H]thymidine (1.85 × 104 Bq well−1) for 6 h for liquid scintillation counting to measure their proliferation.

Results

Elimination of CS2022 in MLN and PP

As shown in Fig. 1, after the mice were orally inoculated with either wild-type χ3306 (pHSG422) or CS2022 (pHSG422) at a dose of 1 × 108 CFU, the total number of salmonellae, the total number of pHSG422-carrying salmonellae, and the proportion of salmonellae still carrying pHSG422 in the tissue samples were determined. Small numbers of salmonellae of χ3306 or CS2022 were detected in the PP but not in the spleen or MLN on day 1 after inoculation (Fig. 1a). The log10 total number of bacterial cells was 1.84±0.41 or 1.94±0.56 (log10 CFU±SD) per PP sample of χ3306- or CS2022-inoculated mice, respectively. These differences were not statistically significant (P=0.8). Therefore, χ3306 and CS2022 appear to be equally invasive during the early stages of infection of these mice. On day 6 after inoculation, CS2022 was, as expected, greatly attenuated, and the number of bacteria in each of the tissue types examined (i.e. the spleen, MLN, and PP) was significantly lower than that for χ3306 (Fig. 1b). CS2022 was gradually cleared from each tissue, until it was no longer detected by week 8, 6, or 4 after inoculation in the spleen, MLN, or PP, respectively (data not shown). An even greater difference between the total log10 CFU of χ3306 and CS2022 was observed in the spleen on day 6 (7.17±0.49 and 1.76±0.71, respectively; P=0.0000003), and the log10 total number of pHSG422-carrying salmonellae was 3.11±0.66 and 1.10±0.17 in the case of χ3306 and CS2022, respectively. Therefore, the log10 proportion of pHSG422-carrying salmonellae was −4.28±0.72 and −1.09±0.42 in the case of χ3306 and CS2022, respectively. The rates of carriage of pHSG422 in the case of the χ3306 and CS2022 strains in the spleen samples were c. 1/19 000 and 1/12, respectively. The rate of carriage of pHSG422 in the case of CS2022 was more than 1500-fold higher than that of χ3306 in the spleen; this finding indicates that it is possible that the growth rate of C2022 in the spleen was extremely slow compared with that of χ3306 in the spleen. We have previously reported that the doubling times for the growth of χ3306 and CS2022 are each 24 min in L-broth at 37°C (Matsui et al., 2003). Greater numbers of χ3306 than of CS2022 were detected in both the MLN and PP on day 6. However, no pHSG422-carrying salmonellae were detected in either strain in the MLN or PP on day 6. These results suggest the possibility that, in the MLN and PP, CS2022 exhibited a normal growth rate, but the host readily eliminated these salmonellae.

Figure 1

Infection of BALB/c mice by χ3306 or CS2022. On days (a) 1 and (b) 6 after oral inoculation with χ3306 (pHSG422) or CS2022 (pHSG422) at a dose of 1 × 108 CFU, (1) the total CFU of salmonellae, (2) the total CFU of pHSG422-carrying salmonellae, and (3) the proportion of each strain containing pHSG422 were measured. There were six (χ3306-inoculated group) or seven (CS2022-inoculated group) mice, except in the case of the χ3306-inoculated group on day 6 (when there were four mice owing to two deaths). The data are shown as means±SD (SD). The differences between the means of the various groups were determined with the two-tailed Student's t-test. ND, not detected.

Increase in the population of MHC-I-expressing macrophages and DCs derived from the PP of immunized mice

Because it was observed that CS2022 was less able to infect the MLN and PP than was χ3306, it was predicted that this lower ability of CS2022 to grow in mice could influence the efficiency of antigen presentation processed by phagosomes. It is well known that two types of phagocytic cells that act as antigen-presenting cells (APCs), namely macrophages and immature DCs in the PP, are particularly important in the interface between the innate and adaptive immune responses to Salmonella infection (Yrlid et al., 2001b). On day 1 after inoculation with either χ3306 or CS2022, no increases in DCs or in macrophages derived from the PP were detected by flow cytometry, compared with the population of either macrophages or DCs derived from the PP of naive mice (data not shown). As shown in Fig. 2a, on day 6 after inoculation, the population of macrophages derived from the PP of χ3306-inoculated mice increased (4.5%) compared with the population of macrophages derived from the PP of naive mice (0.8%; P=0.0009, the statistical significance between naive mice and χ3306-inoculated mice from three experiments) and CS2022-inoculated mice (0.9%; P=0.001, the statistical significance between CS2022-inoculated mice and χ3306-inoculated mice from three experiments). In contrast, there was no difference in the population of macrophages derived from the PP between naive and CS2022-inoculated mice (P=0.4). We can only assume that the difference of the population of macrophages derived from the PP between χ3306- and CS2022-inoculated mice was the result of the greater difference in the number of salmonellae between χ3306 and CS2022 in the PP on day 6 after inoculation (Fig. 1b). However, as shown in Fig. 2b, the population of DCs derived from the PP was much the same among naive, χ3306-, and CS2022-inoculated mice (1.1, 1.1, and 1.0% of total cells, respectively; P>0.2, the statistical significance of each combination from three experiments).

Figure 2

Flow cytometric analysis of the populations of MHC-I- and/or MHC-II-expressing cells on day 6 after immunization. The populations of MHC-I- and MHC-II-expressing (a) CD11b+ or (b) CD11c+ cells in the PP prepared from naive, χ3306-inoculated, or CS2022-inoculated mice are shown. The results of three experiments, each with 5 mice, were similar, and the results of a representative experiment are shown here. The differences among the means of the three experiments were determined with the two-tailed Student's t-test.

A comparison of CS2022-inoculated mice with naive mice demonstrated that the population of MHC-I-expressing macrophages (single positive: 49%; P=0.0003, the statistical significance from three experiments; Fig. 2a) or DCs (single positive: 51%; P=0.0003, the statistical significance from three experiments; Fig. 2b) derived from the PP increased (in naive mice, single positive: 38% or 43% of macrophages or DCs, respectively). In contrast, no increases in the population of MHC-I-expressing macrophages (single positive: 37%; P=0.5, the statistical significance from three experiments; Fig. 2a) nor any increases in DCs (single positive: 43%; P=0.2, the statistical significance from three experiments; Fig. 2b) derived from the PP of χ3306-inoculated mice were observed, as compared with the population of the MHC-I-expressing macrophages (single positive: 38%; Fig. 2a) or with the population of MHC-I-expressing DCs (single positive: 43%; Fig. 2b) from naive mice. In addition, no differences were observed in terms of the population of MHC-II-expressing macrophages (single positive: 11%) or DCs (single positive: 12%) derived from the PP of CS2022-inoculated mice, as compared with naive mice (single positive: 13% or 14% of macrophages or DCs, respectively; P>0.3, the statistical significance of each combination from three experiments). Based on these results, it is reasonable to assume that the digested CS2022 might be primarily processed for antigen presentation on MHC-I molecules in both macrophages and DCs in the PP. Unfortunately, the increase in the population of MHC-II-expressing macrophages and DCs in the PP was not clearly detected on day 6 after immunization.

Increases in the CD4+ and CD8+ T-cell populations in spleens at week 6 after immunization

It has been shown that recovery from Salmonella infection occurs in two stages: an early phase that is T-cell-independent and a late phase that requires T cells (Eisenstein, 1999). Therefore, we next demonstrated changes in the CD4+ and CD8+ T-cell populations in spleens after inoculation of mice with CS2022. Here, greater responses were consistently observed in immunized mice than in naive mice; in immunized mice, increases in both CD4+ (single positive: 31%) and CD8+ (single positive: 17%) T cells were observed in the spleen at week 6 after inoculation (Fig. 4b), whereas no such increases were observed in CD4+ (single positive: 23%; P=0.008, the statistical significance between naive mice and CS2022-inoculated mice from three experiments) and CD8+ (single positive: 12%; P=0.006, the statistical significance between naive mice and CS2022-inoculated mice from three experiments) T cells in the spleens of naive mice (Fig. 4a). Indeed, the populations of both CD4+ and CD8+ T cells had already increased to a certain extent at weeks 2 and 4 after inoculation (data not shown). These findings suggest that both CD4+ and CD8+ T cells were elicited in the spleen by oral immunization with CS2022.

Figure 4

Activation of T-cell responses in splenocytes associated with supplementation with extracellular vesicles. The extracellular vesicles prepared from the culture supernatant of CS2022-challenged RAW264.7 cells were mixed (final concentration: 20 ng protein mL−1) with splenocytes prepared from mice. (a) Naive mice. (b) Mice at 6 weeks postimmunization. The results of three experiments, each with three mice, were similar, and the results of a representative experiment are shown here. The differences among the means of three experiments were determined with the two-tailed Student's t-test.

Increases in splenic CD4+ and CD8+ T cells in response to a supply of salmonellae antigens

We have already reported that an oral immunization with CS2022 induces a robust protective immunity against the challenge with virulent S. enterica serovar Typhimurium at week 4 after immunization (Matsui et al., 2003; Kodama et al., 2005). We suppose that immunization with CS2022 leads to the generation of Salmonella-specific memory T cells. We therefore aimed to confirm the activation of CD4+ and CD8+ T cells in response to a supply of salmonellae antigens. Accordingly, we prepared extracellular vesicles from the culture supernatants of CS2022-challenged RAW264.7 cells as the antigen source. For the target T cells, we prepared splenocytes from either 6-week postinfection CS2022-immunized mice or naive mice. As expected, CD69-expressing CD4+ (single positive: 18%; P=0.006, the statistical significance between naive mice and the 6-week postinfection CS2022-immunized mice from three experiments) or CD8+ (single positive: 24%; P=0.005, the statistical significance between naive mice and the 6-week postinfection CS2022-immunized mice from three experiments) T cells in splenocyte populations prepared from the 6-week postinfection CS2022-immunized mice were activated by the supply of extracellular vesicles (Fig. 4b), whereas this was not the case with CD69-expressing CD4+ (single positive: 4.5%) or CD8+ (single positive: 9.5%) T cells prepared from the spleens of naive mice (Fig. 4a). In addition, if there was a lack of supplementation with extracellular vesicles, there were no activations of CD69-expressing CD4+ or CD8+ T-cell populations among splenocytes prepared from the 6-week postinfection CS2022-immunized mice (single positive CD4+: 2.4% and single positive CD8+: 9.3%; P=0.9) compared with naive mice (single positive CD4+: 3.8% and single positive CD8+: 10.1%). These differences were not statistically significant (P>0.1). These findings suggest the possibility that the CD4+ and CD8+ T cells in splenocytes prepared from CS2022-immunized mice responded to salmonellae antigens supplemented with the extracellular vesicles.

Increase in the proliferation of CD4+ and CD8+ T cells through DCs in response to a supply of salmonellae antigens

We attempted to obtain in vitro evidence that salmonellae antigens supplemented with the extracellular vesicles would induce the activation of splenic CD4+ and CD8+ T cells via DCs. The DCs were generated from the bone marrow cells of naive mice. Then, extracellular vesicles were prepared from the culture supernatants of either CS2022- or χ3306-challenged RAW264.7 cells, and these vesicles were mixed with DCs. Subsequently, the CD4+ or CD8+ T cells purified from the spleens of CS2022-immunized mice were added to the DC culture. As shown in Fig. 5, the proliferation of both CD4+ and CD8+ T cells appeared to depend on the extracellular vesicles prepared from χ3306- or CS2022-challenged RAW264.7 cells. Statistically significant differences were observed between the levels of CD4+ or CD8+ T-cell proliferation associated with extracellular vesicles prepared from CS2022-challenged RAW264.7 cells and those associated with a lack of vesicle supplementation (P=0.0007–0.05). Furthermore, significant differences were also observed between the levels of CD4+ or CD8+ T-cell proliferation associated with extracellular vesicles (200 ng mL−1) prepared from χ3306-challenged RAW264.7 cells and those with a lack of vesicle supplementation (P=0.003 or 0.03, respectively). In contrast, when the extracellular vesicles (20 ng mL−1) were prepared from χ3306-challenged RAW264.7 cells, the proliferation of CD4+ and CD8+ T cells did not differ from the values for cells lacking such supplementation (P=0.8 or 0.2, respectively). Under these experimental conditions, the proliferation of CD4+ or CD8+ T cells with vesicle supplementation did not differ from that of nonsupplemented cells when the cultures contained only purified T cells and the extracellular vesicles were prepared from CS2022-challenged RAW264.7 cells (this result excludes the DCs) (data not shown).

Figure 5

Proliferation of primed T cells induced by cocultivation with naive DCs and extracellular vesicle supplementation. The (a) CD4+ or (b) CD8+ T cells responded to the extracellular vesicles prepared from the culture supernatants of (2) 20 ng protein mL−1 of χ3306-, (3) 200 ng protein mL−1 of χ3306-, (4) 20 ng protein mL−1 of CS2022-challenged, or (5) 200 ng protein mL−1 of CS2022-challenged RAW264.7 cells. (1) No extracellular vesicle supplementation. The results of the three experiments were similar, and the results of a representative experiment are shown here. The data represent the means±SD in triplicate samples. The differences among the means of the various groups were determined with the two-tailed Student's t-test.

Discussion

In female BALB/c mice, the oral 50% lethal doses (LD50) of CS2022 and the virulent strain (χ3306) of S. enterica serovar Typhimurium SR-11 have been reported to be greater than 109 CFU (maximum infection dose) and 3 × 105 CFU, respectively (Takaya et al., 2003). When CS2022 was used for oral inoculation at the maximum dose (109 CFU), the mice did not die. A previously reported theoretical strategy (Gulig & Doyle, 1993; Gulig et al., 1997) was consulted for measuring the relative replication rates of bacteria in mice using the plasmid pHSG422. As a result, Fig. 1 shows that χ3306 was capable of growing in the spleen, MLN, and PP of infected mice, with the present result that two out of six χ3306-inoculated mice died by day 6 after oral inoculation. Figure 1 also provides compelling evidence that, although CS2022 was unable to grow in the lymphoid tissues during the infection of mice, CS2022 was able to invade murine gut-associated lymphoid tissue (GALT) after a single oral immunization. As a result, a sufficient level of both humoral and cellular immunity was found to be induced to dramatically protect the mice against subsequent oral challenge with virulent S. enterica serovar Typhimurium at a lethal infection dose (Matsui et al., 2003; Kodama et al., 2005).

In the present study, we initially expected that CS2022 would be digested by the resident APCs derived from the PP, resulting in the presentation of bacteria-encoded antigens on MHC-I and MHC-II molecules after the internalization of the antigens by bystander APCs. Indeed, the populations of both MHC-I-expressing DCs and MHC-I expressing macrophages derived from the PP increased on day 6 after immunization with CS2022 (Fig. 2). These results indicate that the DCs and macrophages derived from the PP were capable of processing CS2022, but not χ3306, for peptide presentation on MHC-I molecules. In contrast, the population of MHC-II-expressing DCs nor that of MHC-II-expressing macrophages increased on day 6 after immunization with CS2022. However, the populations of both CD4+ and CD8+ T cells increased in the spleen at week 6 after immunization with CS2022 (Fig. 3). Therefore, it could be assumed that the populations of both MHC-1- and MHC-II-expressing APCs in the PP were actually increased by immunization with CS2022. Otherwise, the intestinal epithelial cells secreted exosomes from their apical and basolateral surfaces, and these exosomes may have carried molecules involved in antigen presentation with MHC-II expression (van Niel & Heyman, 2002). In this context, intracellular virulent Salmonella have been variously described as actively inhibiting antigen processing and presentation (Cheminay et al., 2004; Mitchell et al., 2004; Qimron et al., 2004; Tobar et al., 2004).

Figure 3

Flow cytometric analysis of the population of CD4+ or CD8+ T cells after immunization of mice with CS2022. The population of CD4+ or CD8+ T cells among splenocytes prepared from mice is shown. (a) Naive mice. (b) Mice at 6 weeks postimmunization. The results of a representative experiment are shown here. The differences among the means of three experiments were determined with the two-tailed Student's t-test.

It was clearly observed that the addition of extracellular vesicles prepared from the culture supernatants of CS2022-challenged RAW264.7 cells to splenocyte cultures prepared from CS2022-immunized mice triggered the proliferation of CD4+ and CD8+ T cells (Fig. 4). Moreover, the addition of limited amounts of proteins (20 ng mL−1) involved in the extracellular vesicles prepared from the culture supernatants of CS2022-challenged RAW264.7 cells to the DC culture activated the proliferation of CD4+ and CD8+ T cells purified from the spleens of CS2022-immunized mice. However, the addition of these large amounts of proteins (200 ng mL−1) involved in the extracellular vesicles prepared from the culture supernatants of χ3306-challenged RAW264.7 cells to the DC culture also activated the proliferation of CD4+ and CD8+ T cells purified from the spleens of CS2022-immunized mice (Fig. 5). Taken together, the findings suggest that the Salmonella-expressed proteins within macrophages induced the activation of T-cell responses after uptake by bystander DCs.

When salmonellae were phagocytosized in RAW264.7 cells, the SPI1 genes of CS2022 were continuously expressed; however; those of χ3306 were depressed (Takaya et al., 2005b). Thus, the accumulated SPI1 proteins within RAW264.7 cells may have acted as the T-cell antigens. In principle, memory CD8+ T cells are an important component of protective immunity against viral and bacterial infections. It has been demonstrated that DCs in PP confer gut-homing specificity to CD8+ T cells, and thus these DCs license effector/memory cells to access the anatomical sites most likely to contain their cognate antigen (Mora et al., 2003). The infection of mice with lymphocytic choriomaningitis virus or Listeria monocytogenes has been shown to result in long-term protective immunity and the generation of a memory CD8+ T-cell population that is maintained in the absence of antigen (Busch et al., 1998; Murali-Krishna et al., 1998). In the course of the clearance of these intracellular pathogens, there was a linear differentiation path from effector into central T cells, indicating that memory CD8+ T-cell subsets participated in a continuum of T-cell differentiation (Wherry et al., 2003).

Recently, Schaible et al., (2003) reported that, upon infection with mycobacteria, APCs release extracellular vesicles during their progression towards apoptotic cell death. These vesicles carry the enclosed mycobacterial antigens to uninfected bystander APCs, thus promoting presentation to MHC-I- and CD1b-restricted T cells. In the case of Salmonella infection, Miller & Mekalanos (1990) were the first to demonstrate that PhoPc (a constitutive phoQ mutation resulting in unregulated expression of phoP-activated genes and in the unregulated repression of phoP-repressed genes) S. enterica serovar Typhimurium was attenuated with respect to mouse virulence and survival within macrophages. However, the PhoPc strain did not induce apoptosis in infected macrophages, and thus this mutant strain could not afford to be processed for antigen presentation to T cells by bystander DCs, unless the infected macrophages were induced to undergo apoptosis by treatment with LPS and ATP (Yrlid & Wick, 2000). We have previously demonstrated (Takaya et al., 2005b) that infection with CS2022 induces rapid and massive apoptosis in cultured RAW264.7 macrophage cells via both caspase-1- and caspase-3-dependent pathways. The increase in macrophage apoptosis induced by infection with CS2022 might be the result of the sustained expression of SPI1 genes in a context in which they would normally be repressed. Under the experimental conditions of this study, there was a siginificant increase in the number of apoptotic cells as early as 6 h after challenge with CS2022. However, at this time point, the apoptotic cells still adhered to the cell culture plates. Most of these apoptotic cells were then released into the culture medium by 24 h after the challenge. It is unclear whether CS2022 can induce the apoptosis of infected macrophages in mice.

We believe that our data will provide guidance for the rational design of new vaccine strategies against salmonellosis.

Acknowledgements

We are grateful to Paul A. Gulig (University of Florida) for his critical reading of the manuscript and for his helpful suggestions. This work was supported by a Grant-in-Aid for Scientific Research C (15590398) to H.M. from the Ministry of Education, Culture, Sports, Science, and Technology of the Japanese Government. Support was also received in the form of a Kitasato University Research Grant for Young Researchers in 2004 to M.E.

Footnotes

  • Editor: Willem van Eden

References

View Abstract