Efficient T cell activation requires both TCR signals and costimulatory signals. CD28 is one of the molecules that provide costimulatory signals for T cells. We used mice deficient in CD28 expression (CD28−/− mice) to analyze the role of CD28 in the immune response against the intracellular bacterium Salmonella typhimurium, the causative agent of murine typhoid fever. CD28−/− mice were highly susceptible to infection with wild-type S. typhimurium and even failed to control infection with attenuated aroA S. typhimurium. More detailed analysis revealed that CD28−/− animals did not mount a T-dependent Ab response and were highly impaired in the production of IFN-γ. Thus, CD28 cosignaling is crucial for immunity against S. typhimurium. To our knowledge, this is the first report describing an essential role for CD28 in protective immunity against an intracellular microbial pathogen.

Salmonella typhimurium (Salmonella enterica serotype Typhimurium) causes murine typhoid fever, which is considered a valuable model for human typhoid (1). After oral uptake, S. typhimurium crosses the intestinal epithelium and enters the Peyer’s patches. From there, bacteria spread via the mesenteric lymph nodes to spleen and liver. S. typhimurium infects and survives in different cell types, notably macrophages and hepatocytes (1). The initial phase of infection is characterized by the production of inflammatory cytokines and phagocyte activation. A critical component of the innate immune response is the expression of a functional Nramp1 molecule in phagocytes (1, 2). Mice eventually acquire specific immunity against S. typhimurium, and both B and T lymphocytes are important for this process (1). The specific response results in clearance of S. typhimurium from the host and protection against reinfection. S. typhimurium infection induces CD4+ and CD8+ T cells, and CD4+ T cells appear to be particularly important for protection against S. typhimurium (3, 4, 5, 6, 7). Although the mechanisms by which T cells mediate protection are not fully understood, cytokines that activate bactericidal mechanisms in macrophages are critical. IFN-γ has been shown to be especially crucial for the defense against S. typhimurium (4, 5, 8, 9). Besides NK cells, T cells are the most important source of this cytokine, and the development of a Th1 response is considered essential for successful immunity against S. typhimurium (1). The relevance of additional T cell mechanisms is less clear. These mechanisms may include help for B cells, organization of granuloma formation by either cytokine secretion or direct cell-cell contact, and cytotoxicity against infected cells.

Ag recognition by the TCR induces activation of T lymphocytes. However, TCR-mediated signals alone are insufficient for efficient T cell activation, and additional costimulatory signals are required. One of the most important surface molecules that delivers costimulatory signals for T cells is CD28. CD28 is expressed on T cells and NK cells, and ligands for CD28 and the structurally related CTLA4 (CD152) are the molecules B7.1 (CD80) and B7.2 (CD86) (10). B7.1 and B7.2 molecules are expressed on professional APC, and their expression is up-regulated during the immune response (10). Stimulation of T cells in the absence of CD28-mediated cosignaling results in impaired proliferation, reduced cytokine production, and altered generation of CD4+ Th cell subsets (11, 12, 13). Moreover, CD28 plays an important role in T-B cell cooperation. Mice deficient in CD28 fail to develop germinal centers and have changes in the basal serum levels of different Ab isotypes and impaired specific Ab production (11, 14). Despite overwhelming evidence of the importance of CD28 for T cell activation and differentiation in vitro and in vivo, there is only limited evidence for an essential role of CD28 in protection against infection. The role of CD28 has been analyzed in different infection models for intracellular pathogens, including Leishmania major and Listeria monocytogenes. Surprisingly, the absence of CD28 did not change the response against L. major in CD28-deficient mice of C57BL/6 and BALB/c backgrounds (15, 16). CD28-deficient C57BL/6 mice produced IFN-γ in response to L. major and were still able to control the infection, whereas CD28-deficient BALB/c mice remained highly susceptible and produced large amounts of IL-4 (15). Treatment of mice with Abs against the B7 molecules reduced IFN-γ and IL-2 production in response to L. monocytogenes, but left the control of L. monocytogenes unimpaired during both primary and secondary infections (17). These results would be consistent with CD28 cosignaling not being essential for protective immunity against intracellular pathogens and would indicate that CD28 signaling is either not a critical component for mobilization of antimicrobial effector functions or that CD28 signaling can be compensated for by other signaling pathways.

Here we report that mouse mutants deficient in CD28 are highly susceptible to infection with a wild-type (wt)2 (2) strain of S. typhimurium and even fail to control infection with an attenuated strain of S. typhimurium. CD28−/− mice suffered from impaired salmonella-specific Ab responses with a complete lack of specific IgG1 and IgG2a. Furthermore, the reduced IFN-γ production, which was at least in part due to diminished numbers of IFN-γ-producing cells, indicates that CD28 is crucial for the development of a Th1 response to S. typhimurium. In summary, our findings that CD28 is essential for immunity against S. typhimurium demonstrates for the first time a critical role of this molecule in antimicrobial immunity.

CD28−/− mice back-crossed 10 times onto the C57BL/6 background (11), C57BL/6 mice, and (C57BL/6 × Sv129)F1 mice were bred in our facility at the Federal Institute for Health Protection of Consumers and Veterinary Medicine (Berlin, Germany), and experiments were conducted according to the German animal protection law.

Mice were typed for their Nramp1 genotype with PCR as described by Weintraub et al. (18). Tail DNA was amplified using a common 3′ primer (5′-ACA GCC CGG ACA GGT GGG-3′) and either a 5′ primer specific for the Nramp1s genotype (5′-ACG CAT CCC GCT GTG GGA-3′) or a 5′ primer specific for the Nramp1r genotype (5′-ACG CAT CCC GCT GTG GGG-3′). Reaction mixtures were heated to 94°C for 5 min; this was followed by 30 cycles of 30 s at 94°C, 30 s at 60°C, and 60 s at 72°C and a final extension at 72°C for 7 min.

Two strains of S. typhimurium were used in this study: SL1344 is a wt strain of S. typhimurium (rspL, hisG), and SL7207 is an aroA strain of S. typhimurium with a block in aromatic synthesis and requires metabolites from the host organism (19). The low abundance of metabolites in mammalian tissues results in restricted growth of S. typhimurium SL7207 in infected mice. Salmonella strains were provided by Dr. B. A. D. Stocker (Department of Medical Microbiology, Stanford University, Stanford, CA). Both salmonella strains were grown overnight in LB medium, washed twice in PBS, frozen, and stored at −80°C. Aliquots were thawed, and bacterial titers were determined by plating serial dilutions on LB agar plates. For infection, aliquots were thawed and appropriately diluted in PBS. Bacteria were injected in a volume of 200 μl of PBS into the lateral tail vein of mice. For determination of bacterial burden in organs, mice were killed at the time points indicated. Liver and spleen were homogenized in PBS, serial dilutions of homogenates were plated on LB agar plates, and colonies were counted after incubation overnight at 37°C.

Spleens from infected mice were removed, and single-cell suspensions were obtained by teasing spleens through stainless steel meshes as previously described (5). Erythrocytes were lysed, and spleen cells (2 × 105/well) were cultured in 96-well plates in RPMI medium supplemented with glutamine, sodium pyruvate, 2-ME, penicillin, streptomycin, and 10% FCS. Spleen cells were stimulated by either 1 μg/ml of anti-CD3 mAb (clone 145 2C11) or 1 × 108 heat-killed S. typhimurium SL7207 (HKS). All experimental values were determined in triplicate. After 2 days, supernatants were removed and stored at −20°C. For the production of HKS an overnight culture of S. typhimurium SL7207 was washed twice and incubated at 80°C for 2 h. Bacterial number was determined by absorption at 600 nm (OD of 1 is equivalent to 1 × 109 bacteria), and effective killing was validated by plating HKS onto LB agar plates.

Cell culture supernatants were analyzed for cytokines by ELISA. Plates (96W Nunc-Immuno plate, Nunc, Roskilde, Denmark) were coated with 50 μl/well of mAb (2 μg/ml in PBS) overnight at 4°C, blocked with 200 μl/well of PBS and 1% BSA at 37°C for 2 h, and incubated with culture supernatants or serial dilutions of cytokine standards in supplemented RPMI medium overnight at 4°C. ELISA was proceeded by incubation with biotinylated mAb (2 μg/ml in PBS and 1% BSA) and with alkaline phosphatase-conjugated streptavidin (Dianova, Hamburg, Germany; 1 μg/ml in PBS and 1% BSA), each for 1 h at 37°C. After each incubation step, plates were washed four times with PBS and 0.05% Tween 20. Plates were finally developed by the addition of 50 μl of p-nitrophenyl phosphate (PNPP) substrate (1 mg/ml in diethanolamine buffer, pH 9.8; Sigma, St. Louis, MO). Enzymatic reaction was terminated by the addition of 50 μl of 0.5 M EDTA, pH 8, and absorption at 405 nm was determined. The following mAbs were used: IFN-γ, R4-6A2 (20) and biotinylated XMG1.2 (21); IL-4, 11B11 (22) and biotinylated BVD6-24G2 (23); and IL-10, JES5-2A5 (23) and biotinylated SXC-1 (PharMingen, San Diego, CA). Cytokine standards were purchased from Genzyme (Cambridge, MA).

ELISPOT assays were performed as described previously (24) with some modifications. ELISPOT plates (Millipore, Bedford, MA) were coated with anti-IFN-γ mAb (clone R4-6A2) or anti-IL-4 mAb (clone BVD4-1D11; PharMingen) at 2 μg/ml 0.05 M carbonate buffer, pH 9.6, overnight. Plates were washed twice with PBS, incubated for 2 h at 37°C with blocking buffer (PBS and 1% BSA), and washed again with PBS, and different dilutions of spleen cells (1 × 105, 3 × 104, and 1 × 104) were incubated in complete RPMI medium for 18 h at 37°C. Cells were stimulated with either HKS (1 × 108/well) or anti-CD3 mAb (1 μg/ml). To ensure uniform Ag presentation at different spleen cell dilutions, 1 × 105 cells/well of the macrophage cell line BM A3.1A7 (25) were added. Experimental values were determined in triplicate. After incubation, plates were washed five times with PBS and five times with PBS and 0.05% Tween 20. Biotinylated mAb were added (anti-IFN-γ, XMG1.2; IL-4, BVD6-24G2; 0.25 μg/ml in PBS, 0.05% Tween 20, and 0.1% BSA), and plates were incubated for 2 h at 37°C. Plates were washed 10 times with PBS and 0.05% Tween 20, and alkaline phosphatase-conjugated streptavidin (0.1 μg/ml in PBS, 0.05% Tween-20, and 0.1% BSA) was added for 1 h at 37°C. Plates were again washed 10 times and developed by the addition of 5-bromo-4-chloro-3-indolyl-phosphate substrate, prepared as indicated by the manufacturer (Sigma). Enzymatic reaction was stopped by washing the plates with distilled water. Plates were dried overnight, and spots were counted using a binocular dissecting microscope.

Mice were infected with 5 × 105 S. typhimurium strain SL7207 or 1 × 103 S. typhimurium SL1344 and bled at the time points indicated. For determination of salmonella-specific Abs, plates (Nunc-Immuno plate, Nunc) were coated with HKS (1 × 108/ml PBS) at 4°C overnight. Plates were washed with PBS and blocked with PBS and 1% BSA at 4°C. Plates were washed again, serial dilutions of sera were added, and plates were incubated overnight at 4°C. Plates were washed four times with PBS and 0.05% Tween 20 and incubated with alkaline phosphatase-conjugated mAb (1 μg/ml in PBS and 0.1% BSA) specific for total Ig (BioSource, Camarillo, CA), or the Ig isotypes IgM (BioSource), IgG1, IgG2a, and IgG3 (PharMingen). Plates were incubated for 2 h at 37°C, washed four times with PBS and 0.05% Tween 20, and developed as described for the cytokine ELISA. Data are given as the titer (dilution with half-maximal absorption). Ig isotype-specific Abs were tested against sera from both Sv129 and C57BL/6 mice to exclude failure of reactivity due to different Ab allotypes.

The statistical significance of results was determined with the statistic program included in the GraphPad Prism program (version 2.0, GraphPad Software, San Diego, CA). Survival curves were analyzed with the log-rank test. Mean bacterial titers are given as the geometric mean, and differences in titers were determined with the unpaired t test from log-transformed values.

Expression of a functional Nramp1 protein is a major component in determining the susceptibility of mice against infection with S. typhimurium (2). CD28−/− mice were generated from Sv129-derived embryonic stem cells and back-crossed onto the C57BL/6 background (11). Because Sv129 and C57BL/6 mice differ in their Nramp1 genotype, CD28−/− mice were typed by PCR and found to have the Nramp1r phenotype of Sv129 mice (data not shown). Inquiry in the Mouse Genome Database (26) revealed that the gene loci for CD28 and Nramp1 are both located on mouse chromosome 1, at 30.1 and 39.2 cM, respectively. The close proximity of the gene loci could explain why the Nramp1r genotype was conserved during back-crossing of CD28−/− mice. Because CD28−/− mice on the Sv129 background were not available to us, in all further experiments we used the back-crossed CD28−/− mice together with (C57BL/6 × Sv129)F1 control animals (CD28+/+ mice) with the dominant Nramp1r phenotype.

In a first set of experiments, CD28+/+ (C57BL/6 × Sv129, Nramp1r) and CD28−/− mice (Nramp1r) were i.v. infected with different doses of S. typhimurium strain SL1344 (Fig. 1). Although the gut represents the natural port of entry for S. typhimurium, we decided to administer the bacteria i.v. to avoid inconsistency of bacterial load due to variations in intestinal flora. In addition, the i.v. route allowed better control of the infection dose, particularly when small bacterial doses were applied. For comparison, we included Nramp1s C57BL/6 mice in the experiment shown. Despite the Nramp1r phenotype, CD28−/− mice were highly susceptible to all doses of S. typhimurium used, but compared with Nramp1s mice, the mean survival time was prolonged significantly. At different time points postinfection with 103 salmonellae (i.v.), bacterial titers in spleens and livers were determined (Fig. 2). Three days postinfection, both CD28+/+ and CD28−/− mice harbored equal bacterial titers in spleen and liver. In contrast, on day 10 postinfection, CD28−/− mice suffered from a higher bacterial burden in the liver. (Although we consistently observed slightly increased bacterial numbers in spleens of CD28−/− mice, results were only statistically significant in part of the experiments.) At this time point, both CD28+/+ and CD28−/− animals had developed splenomegaly, and histologic comparison of spleens demonstrated disorganization of the spleen architecture in both mouse strains (data not shown).

FIGURE 1.

CD28−/− mice are highly susceptible to infection with wt S. typhimurium. CD28+/+ C57BL/6 mice (Nramp1s), CD28+/+ (C57BL/6 × Sv129)F1 (Nramp1r), and CD28−/− mice (Nramp1r) were i.v. infected with the indicated inocula of S. typhimurium strain SL1344, and survival was recorded. Survival curves of the three mouse strains were significantly different from each other for all three doses of S. typhimurium (p < 0.05, with the log-rank test). Groups consisted of 5–10 mice. The experiment shown is representative for two independent experiments. Solid lines, CD28+/+ C57BL/6 mice; dotted lines, CD28+/+ (C57BL/6 × Sv129)F1 mice; dashed lines, CD28−/− mice.

FIGURE 1.

CD28−/− mice are highly susceptible to infection with wt S. typhimurium. CD28+/+ C57BL/6 mice (Nramp1s), CD28+/+ (C57BL/6 × Sv129)F1 (Nramp1r), and CD28−/− mice (Nramp1r) were i.v. infected with the indicated inocula of S. typhimurium strain SL1344, and survival was recorded. Survival curves of the three mouse strains were significantly different from each other for all three doses of S. typhimurium (p < 0.05, with the log-rank test). Groups consisted of 5–10 mice. The experiment shown is representative for two independent experiments. Solid lines, CD28+/+ C57BL/6 mice; dotted lines, CD28+/+ (C57BL/6 × Sv129)F1 mice; dashed lines, CD28−/− mice.

Close modal
FIGURE 2.

CD28−/− mice suffer from increased S. typhimurium titers at late time points of infection. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 103 bacteria of S. typhimurium strain SL1344, and bacterial burden in spleen and liver was determined on day 3 and 10 postinfection. On day 10 postinfection titers from livers of CD28−/− mice were significantly higher (p < 0.05) than titers of livers from CD28+/+ mice. The experiment shown is representative of two independent experiments.

FIGURE 2.

CD28−/− mice suffer from increased S. typhimurium titers at late time points of infection. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 103 bacteria of S. typhimurium strain SL1344, and bacterial burden in spleen and liver was determined on day 3 and 10 postinfection. On day 10 postinfection titers from livers of CD28−/− mice were significantly higher (p < 0.05) than titers of livers from CD28+/+ mice. The experiment shown is representative of two independent experiments.

Close modal

CD28−/− mice were also i.v. infected with 5 × 105 attenuated aroA S. typhimurium, and bacterial burden in liver and spleen was determined at different time points thereafter (Fig. 3). In CD28+/+ mice systemic infection with S. typhimurium was cleared within 7–10 wk. In contrast, salmonellae persisted in livers and spleens of CD28−/− mice at high titers and established chronic infection.

FIGURE 3.

CD28−/− mice fail to clear attenuated S. typhimurium infection. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207, and bacterial burden in liver and spleen was determined at the time points indicated. The experiment shown is representative of two independent experiments.

FIGURE 3.

CD28−/− mice fail to clear attenuated S. typhimurium infection. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207, and bacterial burden in liver and spleen was determined at the time points indicated. The experiment shown is representative of two independent experiments.

Close modal

S. typhimurium infection induces a potent salmonella-specific Ab response in mice, which is thought to contribute to protection (27). To analyze the role of CD28 in the generation of salmonella-specific Abs, CD28+/+ and CD28−/− mice were infected with S. typhimurium strain SL7207 and bled at different time points after inoculation, and Ab titers were determined by ELISA. On days 32 and 47 postinfection, CD28+/+ mice showed high titers of salmonella-specific Abs for all Ab isotypes tested (Fig. 4). In contrast, CD28−/− mice had reduced serum titers of salmonella-specific IgM and IgG3 and completely failed to develop detectable IgG1 and IgG2a.

FIGURE 4.

CD28−/− mice fail to mount a T-dependent Ab response against attenuated S. typhimurium. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207 and bled at the time points indicated. Ab titers were determined by ELISA.

FIGURE 4.

CD28−/− mice fail to mount a T-dependent Ab response against attenuated S. typhimurium. CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207 and bled at the time points indicated. Ab titers were determined by ELISA.

Close modal

Salmonella-specific Ab titers were also determined on day 12 postinfection with virulent S. typhimurium strain SL1344 (103 bacteria i.v.). In CD28+/+ mice we could detect specific IgM and small amounts of specific IgG2a, but no specific IgG1 or IgG3 at this time point of infection. CD28−/− mice had reduced levels of specific IgM, and specific IgG2a was absent (data not shown).

To assess whether impaired Ab production was responsible for the inability of CD28−/− mice to cope with S. typhimurium, animals received 0.5 ml of serum from mice that had been vaccinated with S. typhimurium SL7207 50 days previously (this serum had a salmonella-specific Ab titer of >10,000 for total Ig). It has been shown that when transferred together with T cells, this amount of serum protects naive mice against infection with virulent S. typhimurium (28). Mice were infected with S. typhimurium SL1344 and subsequently received normal mouse serum or serum from vaccinated mice (Fig. 5,A). Immune serum had only minimal effects. All mice succumbed to infection, and the mean survival time was only marginally prolonged compared with that in mice that had received control serum. Similarly, mice infected with attenuated S. typhimurium SL7207 received serum on day 32 postinfection, and bacterial titers in spleens and livers were determined on day 40. Comparison of bacterial titers revealed no significant difference among mice that had received no serum, control serum, or immune serum (Fig. 5,B). On day 40, CD28+/+ mice had developed high salmonella-specific Ab titers (see Fig. 4) and had almost cleared the infection with S. typhimurium SL7207. Therefore, bacterial titers after serum transfer were not determined in these mice.

FIGURE 5.

Treatment of CD28−/− mice with immune serum does not improve the response against S. typhimurium. A, CD28−/− mice were i.v. infected with 100 S. typhimurium strain SL1344. After 9 days mice were treated with 0.5 ml (i.p.) of either normal mouse serum (solid line) or serum from mice infected with 5 × 105 S. typhimurium strain SL7207 on day −50 (dotted line). Survival was recorded daily. Groups consisted of six (control serum) and eight mice (immune serum). Mean survival times were 19.5 days for mice that received control serum and 22.5 days for mice that received immune serum. Survival curves were not significant different (p > 0.05, with the log-rank test). The experiment shown is representative of two independent experiments. B, CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207. On day 32 mice received 0.5 ml of either control or immune serum. Mice were killed on day 40, and bacterial loads of spleen and liver were determined. Comparison of bacterial titers from untreated, control serum-treated, and immune serum-treated CD28−/− mice resulted in p > 0.1 for both liver and spleen. The experiment shown is representative of two independent experiments.

FIGURE 5.

Treatment of CD28−/− mice with immune serum does not improve the response against S. typhimurium. A, CD28−/− mice were i.v. infected with 100 S. typhimurium strain SL1344. After 9 days mice were treated with 0.5 ml (i.p.) of either normal mouse serum (solid line) or serum from mice infected with 5 × 105 S. typhimurium strain SL7207 on day −50 (dotted line). Survival was recorded daily. Groups consisted of six (control serum) and eight mice (immune serum). Mean survival times were 19.5 days for mice that received control serum and 22.5 days for mice that received immune serum. Survival curves were not significant different (p > 0.05, with the log-rank test). The experiment shown is representative of two independent experiments. B, CD28+/+ (⋄) and CD28−/− mice (♦) were i.v. infected with 5 × 105 S. typhimurium strain SL7207. On day 32 mice received 0.5 ml of either control or immune serum. Mice were killed on day 40, and bacterial loads of spleen and liver were determined. Comparison of bacterial titers from untreated, control serum-treated, and immune serum-treated CD28−/− mice resulted in p > 0.1 for both liver and spleen. The experiment shown is representative of two independent experiments.

Close modal

IFN-γ has been shown to be critical for host protection against S. typhimurium, whereas deficient IL-4 production does not impair acquired resistance (4, 5, 8, 9, 29). To analyze the influence of CD28 on cytokine production and Th cell differentiation, CD28+/+ and CD28−/− mice were infected with S. typhimurium SL1344. At different time points postinfection mice were killed, spleen cells were restimulated in vitro with HKS or anti-CD3 mAb, and culture supernatants were analyzed for cytokines (Table I). Spleen cells from naive CD28+/+ mice produced large amounts of IFN-γ after stimulation with anti-CD3 mAb, which were further increased after infection of mice with S. typhimurium. We detected small amounts of IFN-γ in response to HKS in cells from naive CD28+/+ mice and mice infected for 3 days with S. typhimurium. Abundant amounts of IFN-γ were produced on days 6 and 9 postinfection. In contrast, cells from CD28−/− mice produced only marginal amounts of IFN-γ after stimulation with anti-CD3 mAb or HKS. Although production of IFN-γ increased during infection, IFN-γ production by CD28−/− cells was minute compared with that by CD28+/+ cells.

Table I.

Spleen cells from CD28−/− mice secrete reduced amounts of IFN-γ after restimulation in vitroa

TreatmentMouse StrainIFN-γ (U/ml)
Day 0Day 3Day 6Day 9
Medium +/+ <10 <10 25 ± 11 25 ± 12 
 −/− <10 21 ± 5 23 ± 6 26 ± 3 
HKS +/+ 40 ± 20 51 ± 39 438 ± 212 424 ± 73 
 −/− 16 ± 8 82 ± 49 84 ± 12b 61 ± 12b 
Anti-CD3 +/+ 497 ± 88 423 ± 72 845 ± 297 689 ± 132 
 −/− 19 ± 16b 12 ± 1b 75 ± 4b 57 ± 7b 
TreatmentMouse StrainIFN-γ (U/ml)
Day 0Day 3Day 6Day 9
Medium +/+ <10 <10 25 ± 11 25 ± 12 
 −/− <10 21 ± 5 23 ± 6 26 ± 3 
HKS +/+ 40 ± 20 51 ± 39 438 ± 212 424 ± 73 
 −/− 16 ± 8 82 ± 49 84 ± 12b 61 ± 12b 
Anti-CD3 +/+ 497 ± 88 423 ± 72 845 ± 297 689 ± 132 
 −/− 19 ± 16b 12 ± 1b 75 ± 4b 57 ± 7b 
a

CD28+/+ and CD28−/− mice were i.v. infected with 103S. typhimurium strain SL1344. At the time points indicated, spleen cells (2 × 105/well) were restimulated in vitro with either anti-CD3 mAb or HKS. After 2 days, supernatants were harvested and IFN-γ was measured by ELISA. Results represent mean values ± SD from three mice. The experiment is representative for three independent experiments.

b

Value of CD28−/− mice was significantly reduced compared to value of corresponding CD28+/+ mice (p value <0.05).

Frequency determination of IFN-γ-secreting cells revealed that in the absence of stimulation or after stimulation with HKS hardly any splenocyte from naive CD28+/+ or CD28−/− mice secreted IFN-γ (Table II). After addition of anti-CD3 mAb, IFN-γ-secreting cells were detected in spleens from both mouse strains. S. typhimurium substantially increased frequencies of IFN-γ-secreting cells, even without additional in vitro stimulation. We assume that T cells, NK cells, or other undefined cells that had been activated in vivo still secreted IFN-γ in vitro, or, alternatively, that S. typhimurium-infected cells induced IFN-γ secretion by T cells in vitro. Infection increased frequencies of IFN-γ-secreting cells in spleens from CD28+/+ mice after stimulation in vitro with HKS or anti-CD3 mAb, whereas frequencies of IFN-γ-secreting cells in spleens from CD28−/− mice were far less elevated under these conditions. Thus, IFN-γ production in infected CD28−/− mice was markedly lower than that in CD28+/+ mice, and this was at least in part due to the reduced numbers of IFN-γ-secreting cells.

Table II.

Reduced frequencies of IFN-γ-secreting cells in spleens from CD28−/− mice infected with S. typhimurium SL1344a

TreatmentMouse StrainIFN-γ (spots/106 cells)
Day 0Day 3Day 6Day 9
Medium +/+ 10 ± 12 276 ± 179 478 ± 157 680 ± 332 
 −/− 20 ± 12 88 ± 67 132 ± 2 317 ± 57 
HKS +/+ 10 ± 3 591 ± 364 1,646 ± 353 4,309 ± 810 
 −/− 42 ± 16 201 ± 154 382 ± 111b 1,275 ± 511b 
Anti-CD3 +/+ 891 ± 248 1,761 ± 284 4,881 ± 1,627 18,309 ± 5,517 
 −/− 555 ± 234 981 ± 223b 945 ± 153b 3,812 ± 1,062b 
TreatmentMouse StrainIFN-γ (spots/106 cells)
Day 0Day 3Day 6Day 9
Medium +/+ 10 ± 12 276 ± 179 478 ± 157 680 ± 332 
 −/− 20 ± 12 88 ± 67 132 ± 2 317 ± 57 
HKS +/+ 10 ± 3 591 ± 364 1,646 ± 353 4,309 ± 810 
 −/− 42 ± 16 201 ± 154 382 ± 111b 1,275 ± 511b 
Anti-CD3 +/+ 891 ± 248 1,761 ± 284 4,881 ± 1,627 18,309 ± 5,517 
 −/− 555 ± 234 981 ± 223b 945 ± 153b 3,812 ± 1,062b 
a

Serial dilutions of spleen cells were incubated in anti-IFN-γ mAb-coated plates. T cells were restimulated with either anti-CD3 mAb or HKS. After incubation for 18 h cells were washed off and ELISPOT plates were developed. Results represent mean values ± SD from three mice. The experiment is representative for three independent experiments.

b

Value of CD28−/− mice was significantly reduced compared to value of corresponding CD28+/+ mice (p value <0.05).

In parallel experiments we determined IL-4 and IL-10 production after in vitro restimulation of spleen cells. In all samples analyzed from noninfected and infected CD28+/+ and CD28−/− mice, significant amounts of IL-4 were not detectable (detection limit of IL-4 ELISA, 50 pg/ml). However, IL-4-secreting cells were detected by the more sensitive ELISPOT assay (Table III). In spleens from naive CD28+/+ and CD28−/− mice, frequencies of IL-4-producing cells were 10-fold higher than those of IFN-γ-secreting cells, and stimulation of cells with anti-CD3 mAb substantially increased these frequencies in both mouse strains. S. typhimurium infection resulted in expansion of the IL-4-secreting cell population in spleens from both CD28+/+ and CD28−/− mice. However, HKS restimulation did not further enlarge the population size, and after anti-CD3 mAb restimulation, frequencies even declined in spleens from infected mice compared with those in spleens from naive mice. Spleen cells from infected CD28+/+ and CD28−/− mice did not secrete significant amounts of IL-10, and addition of anti-CD3 mAb did not change IL-10 secretion (Table IV). However, restimulation with HKS induced IL-10 secretion, and infection of mice with S. typhimurium further increased IL-10 production. Compared with spleen cells from CD28+/+ mice, there was a slight increase in the amount of IL-10 secreted by CD28−/− spleen cells.

Table III.

Frequencies of IL-4 secreting cells in spleens from CD28+/+ and CD28−/− mice infected with S. typhimurium SL1344a

TreatmentMouse StrainIL-4 (spots/106 cells)
Day 0Day 3Day 6Day 9
Medium +/+ 152 ± 54 591 ± 80 328 ± 42 801 ± 133 
 −/− 145 ± 42 373 ± 68b 232 ± 59b 423 ± 154b 
HKS +/+ 212 ± 87 576 ± 91 439 ± 91 683 ± 116 
 −/− 167 ± 41 351 ± 36b 233 ± 49b 502 ± 149 
Anti-CD3 +/+ 3,767 ± 498 3,583 ± 544 1,765 ± 252 1,177 ± 199 
 −/− 1,100 ± 428b 938 ± 220b 317 ± 135b 520 ± 100b 
TreatmentMouse StrainIL-4 (spots/106 cells)
Day 0Day 3Day 6Day 9
Medium +/+ 152 ± 54 591 ± 80 328 ± 42 801 ± 133 
 −/− 145 ± 42 373 ± 68b 232 ± 59b 423 ± 154b 
HKS +/+ 212 ± 87 576 ± 91 439 ± 91 683 ± 116 
 −/− 167 ± 41 351 ± 36b 233 ± 49b 502 ± 149 
Anti-CD3 +/+ 3,767 ± 498 3,583 ± 544 1,765 ± 252 1,177 ± 199 
 −/− 1,100 ± 428b 938 ± 220b 317 ± 135b 520 ± 100b 
a

Serial dilutions of spleen cells were incubated in anti-IL-4 mAb-coated plates. T cells were stimulated with either anti-CD3 mAb or HKS. After incubation for 18 h cells were washed off and ELISPOT plates were developed. Results represent mean values ± SD from three mice. The experiment is representative for three independent experiments.

b

Value of CD28−/− mice was significantly reduced compared to value of corresponding CD28+/+ mice (p value <0.05).

Table IV.

IL-10 secretion of spleen cells infected with S. typhimurium after restimulation in vitroa

TreatmentMouse StrainIL-10 (pg/ml)
Day 0Day 3Day 6Day 9
Medium +/+ 115 ± 24 102 ± 10 158 ± 43 184 ± 8 
 −/− 170 ± 5 138 ± 44 149 ± 53 234 ± 7 
HKS +/+ 457 ± 97 445 ± 75 451 ± 140 1,718 ± 766 
 −/− 304 ± 121 522 ± 215 1,009 ± 340 3,260 ± 604 
Anti-CD3 +/+ 206 ± 43 157 ± 48 237 ± 100 258 ± 45 
 −/− 162 ± 16 128 ± 50 145 ± 29 253 ± 7 
TreatmentMouse StrainIL-10 (pg/ml)
Day 0Day 3Day 6Day 9
Medium +/+ 115 ± 24 102 ± 10 158 ± 43 184 ± 8 
 −/− 170 ± 5 138 ± 44 149 ± 53 234 ± 7 
HKS +/+ 457 ± 97 445 ± 75 451 ± 140 1,718 ± 766 
 −/− 304 ± 121 522 ± 215 1,009 ± 340 3,260 ± 604 
Anti-CD3 +/+ 206 ± 43 157 ± 48 237 ± 100 258 ± 45 
 −/− 162 ± 16 128 ± 50 145 ± 29 253 ± 7 
a

CD28+/+ and CD28−/− mice were i.v. infected with 103S. typhimurium strain SL1344. At the time points indicated, mice were killed and spleen cells (2 × 105/well) were restimulated in vitro with either anti-CD3 mAb or HKS. After 2 days supernatants were harvested and IL-10 was measured by ELISA. Results represent mean values ± SD from three mice. The experiment is representative for three independent experiments. There were no significant differences (p value >0.05) between values of CD28+/+ and CD28−/− mice after restimulation with HKS or anti-CD3 mAb.

Our results reveal a critical role for CD28 in protective immunity against both virulent and attenuated S. typhimurium strains. CD28−/− mice were highly susceptible to systemic infection with the virulent strain SL1344 and failed to clear the attenuated aroA strain SL7207. To our knowledge, this is the first report that CD28 plays an essential role in the immune response to an intracellular microbial pathogen. Furthermore, we have identified defective mechanisms that could at least in part explain impaired resistance of CD28−/− mice against S. typhimurium.

CD28−/− mice died after 2–3 wk of systemic infection with virulent salmonellae. Thus, CD28−/− mice are still more resistant than Nramp1s mice, which died within 1 wk. Furthermore, CD28−/− and CD28+/+ mice could control salmonella infection equally well during the first 3 days, but CD28−/− mice became more susceptible thereafter, suffering from higher bacterial titers by day 10 postinfection than CD28+/+ controls. For interpretation of bacterial titers in spleen and liver, ∼108 wt S. typhimurium SL1344 organisms should be considered as a lethal dose (1). An increase in the bacterial titer by 1 order of magnitude can therefore be regarded as a sign for fatal outcome of infection. Although we cannot exclude the possibility that the lack of CD28 impairs the function of NK cells, which can express CD28 and play a role during the initial stage of the host response, both sets of results argue for a defect in the acquired T cell-dependent immune response against S. typhimurium in CD28−/− mice. Consistent with this idea, infection of CD28−/− mice with the attenuated S. typhimurium strain SL7207 caused a chronic infection. This phenotype is similar to nude mice or mice devoid of CD4+ T cells due to a deficiency in MHC class II molecules. As with CD28−/− mice, these mouse strains fail to clear aroA S. typhimurium and develop a chronic infection (5, 30).

In response to S. typhimurium, specific IgM and IgG3 were reduced, and specific IgG1 and IgG2a were absent in CD28−/− mice. In contrast to IgM and IgG3, IgG1 and IgG2a are T cell dependent, indicating that the lack of CD28 expression impaired T-B cell cooperation during infection. The role of CD28 in the generation of T-dependent Abs has been extensively analyzed in CD28−/− mice or in mice in which the CD28-B7 cooperation has been blocked with anti-B7 mAb or soluble CTLA4 molecules. Although blocking of the CD28-B7 interaction strongly impaired T-dependent Ab production in several models (14, 31, 32), there are cases where the lack of this interaction only weakly affected T-dependent Abs (11, 33, 34) or only impaired distinct IgG subclasses (35). It is not yet clear why Ab responses against certain Ags depend on CD28 whereas others do not. Furthermore, the exact role of CD28 in this process remains to be determined. An obvious explanation would be the up-regulation of the CD40 ligand on T cells through CD28 signaling. However, in vitro stimulation with increasing concentrations of anti-CD3 mAb induced equal dose-dependent surface expression of the CD40 ligand on CD4+ T cells from both CD28+/+ and CD28−/− mice (H.-W. Mittrücker, unpublished observation), and several reports have demonstrated CD28-independent expression of CD40 ligand on T cells (36, 37, 38). Other CD28-dependent mechanisms may include the expression of cytokines or surface molecules other than the CD40 ligand that are important for T-B cooperation. Finally, a prerequisite for the generation of T-dependent Abs is the formation of organized lymphoid structures that promote T-B cell interactions in vivo. CD28 is important for this process, because CD28−/− mice and mice that produce soluble CTLA4 molecules fail to generate germinal centers (14, 32). The requirement of lymphoid tissue organization for the formation of T-dependent Abs could also explain the absolute failure of CD28−/− mice to generate these Abs in response to S. typhimurium. Infection with S. typhimurium causes a high degree of inflammation and massive destruction of the lymphoid architecture in lymph nodes and spleen. As a result of this destruction, mechanisms that compensate for the lack of CD28 in response to other Ags could fail in the case of salmonella infection. Alternatively, it is known that S. typhimurium causes nitric oxide production by macrophages, leading to immunosuppression (39), which could, in turn, prevent compensatory mechanisms in CD28−/− mice.

Although our results demonstrate that CD28 is essential for the generation of T-dependent Abs against S. typhimurium, our experiments do not argue for a major contribution of these Abs in host defense against systemic S. typhimurium infection. While Abs appear to be important for protection against oral infection with S. typhimurium, their role in systemic infection remains controversial (1). S. typhimurium infects macrophages and hepatocytes, and due to its intracellular localization, the pathogen is protected from Abs. Our finding that transfer of serum from immunized competent mice to CD28−/− mice did not improve clearing of attenuated salmonellae indicates that Abs play no role or only a minor one in protection against systemic infection with attenuated S. typhimurium strain SL7207. This is in accordance with our finding that Igμ-deficient mice (cf., Nramp1s genotype), which fail to develop peripheral B cells and completely lack Abs, normally clear this strain of S. typhimurium after systemic infection (H.-W. Mittrücker, unpublished observation). Transfer of immune serum to CD28−/− mice infected with the virulent S. typhimurium strain SL1344 only marginally prolonged survival time. Although we cannot exclude the possibility that Abs play some role during systemic infection with wt S. typhimurium, our results indicate that additional CD28-dependent mechanisms exist that are essential for control of S. typhimurium infection.

IFN-γ is crucial for defense against wt and attenuated S. typhimurium, because IFN-γ receptor gene deficiency or IFN-γ neutralization with Abs results in a fatal course of infection (4, 5, 8, 9). Therefore, we analyzed whether CD28−/− mice produced IFN-γ in response to S. typhimurium infection. Spleen cells from salmonella-infected CD28−/− mice produced far less IFN-γ than those from CD28+/+ mice. Concomitantly, frequencies of IFN-γ-secreting cells were markedly reduced. We therefore assume that CD28−/− mice are highly impaired in the generation, expansion, or maintenance of Th1 cells during S. typhimurium infection, and that this impairment could explain the high susceptibility of CD28−/− animals to infection with S. typhimurium.

The generation of a Th1 response operates in fine balance with the generation of a Th2 response, and small changes during initial T cell differentiation can result in profound changes in the type of Th cell response generated, thereby strongly affecting susceptibility to infection. However, in our model reduced IFN-γ production in CD28−/− mice was not accompanied by increased IL-4 production. In all situations tested, CD28−/− spleen cells had lower frequencies of IL-4 producers than controls. In both mouse strains, infection with S. typhimurium increased frequencies of IL-4 producers 2- to 4-fold, and this was not further enhanced by HKS restimulation. We therefore assume that the majority of IL-4-secreting cells were not salmonella-specific T cells, but, rather, mast cells, basophils, or other cells that produce IL-4 in response to infection or inflammation (40, 41, 42).

IL-10 is a regulatory cytokine that can antagonize IFN-γ (43). Significant IL-10 production by spleen cells was only detected after restimulation with HKS, but not with anti-CD3 mAb, suggesting that the majority of cells that produce IL-10 under these conditions are not T cells. Cells from CD28−/− mice showed a slight increase in IL-10 production compared with cells from CD28+/+ mice. Although IL-10 production correlates with the severity of S. typhimurium infection, neutralization of IL-10 does not modify the course of infection, indicating that high levels of IL-10 are a consequence and not a cause of the high susceptibility to this pathogen (44). Hence, the slightly increased IL-10 production in infected CD28−/− mice may reflect the higher severity of infection.

The role of CD28 in Th cell differentiation has been extensively analyzed in different models. For the in vitro generation of Th2 cells CD28 costimulation is essential, whereas established Th2 cells are relatively independent of CD28 costimulation. Apparently, generation of IFN-γ-secreting Th1 cells is less dependent on CD28 costimulation. However, lack of CD28 costimulation reduces IL-2 production, resulting in impaired proliferation and expansion of Th1 cells (13, 45, 46). In vivo, Th cell differentiation has also been analyzed in CD28−/− mice and in mice in which the interactions of CD28 with its B7 ligands had been blocked by soluble CTLA4 molecules or anti-B7 mAb. Interestingly, in some experimental systems, particularly in the L. major infection model, the absence of CD28 and the blockage of CD28-B7 interactions had different effects on the generation of Th cell responses (15, 47, 48). For the correct interpretation of these results, one has to consider several issues. On the one hand, anti-B7 mAb and soluble CTLA4 molecules block not only CD28-B7 interactions but also interactions of B7 molecules with their second ligand CTLA4, and both types of interactions could influence the course of infection. On the other hand, in CD28−/− mice the immune system matures in the absence of CD28, and therefore compensatory mechanisms that bypass CD28 functions can develop.

The requirement for CD28 costimulation in the generation of Th2 cell responses in vivo was demonstrated in different mouse infection models in which the blockage of CD28 costimulation impairs generation of Th2 responses (35, 49, 50, 51). The role of CD28 in the in vivo generation of Th1 cell responses is far less clear. Infection of C57BL/6 mice with L. major results in a Th1 response, leading to parasite clearance. Blocking of CD28 costimulation in this infection model had no effect on either development of the Th1 response or parasite clearance (15, 47). In contrast, blocking of CD28 costimulation in mice infected with L. monocytogenes resulted in impaired IFN-γ production (17). However, lack of IFN-γ production had no effect on listerial growth during both primary and secondary responses (17). In terms of IFN-γ production, infection of CD28−/− mice with S. typhimurium resembled the situation in listeriosis. More importantly, however, it markedly differed in the higher bacterial load and fatal outcome of the S. typhimurium infection. The differential requirements for CD28 in these infection models are difficult to explain. They must be attributed to differential strategies for invasion and infection of the pathogens that allow for compensation for CD28 deficiency in the L. major model, partial compensation in the L. monocytogenes model, and failure to replace CD28 during S. typhimurium infection. Consequently, one has to postulate additional, IFN-γ-independent protection mechanisms that are differentially required for resistance against various intracellular pathogens and that are only in part dependent on CD28 cosignaling. The existence of such IFN-γ-independent mechanisms for resistance against S. typhimurium has been proposed based on the following observations. During S. typhimurium infection, levels of IFN-γ production in certain mouse strains do not correlate with the clearance rate of bacteria, and although neutralization of IFN-γ during the initial phase of infection is fatal, neutralization at later time points has only minor effects on the course of salmonella infection (9, 52). The identity of these IFN-γ-independent protective mechanisms is not yet clear. Although we cannot fully exclude Abs, our results argue for only a minor role in host response during systemic infection, at least in the case of attenuated salmonellae. Another possible mechanism is T cell-mediated cytotoxicity against infected cells, and deficient CD28 signaling can impair the generation of cytotoxic T cells (53). Lysis of infected cells has been considered part of the antibacterial host mechanisms against intracellular pathogens (54). Again, we cannot exclude T cell-mediated cytotoxicity as a major component in the host response against S. typhimurium, but it has been shown that CD8+ T cells are not critical for clearance of attenuated salmonellae (5).

In summary, our results demonstrate that CD28 signaling is essential for the host response against virulent and attenuated strains of S. typhimurium, and for the first time show a crucial role for CD28 in the response to an intracellular bacterium. CD28 participates in the formation of T-dependent Abs and in the generation of Th1 cells. The impaired Th1 cell generation, probably in addition to other not yet characterized mechanisms, could explain the fatal course of S. typhimurium infection in the absence of CD28 cosignaling.

We thank Manuela Stäber and Karin Bordasch for the purification and conjugation of Abs. We are grateful to Dr. B. A. D. Stocker for providing strains of S.typhimurium.

2

Abbreviations used in this paper: wt, wild type; LB, Luria Bertoni; ELISPOT, enzyme-linked immunospot; HKS, heat-killed salmonella.

1
Mäkelä, P. H., C. E. Hormaeche.
1997
. Immunity to salmonella. S. H. E. Kaufmann, ed.
Host Response to Intracellular Pathogens
143
-166. Medical Intelligence Unit, R.G. Landes Co., Austin.
2
Vidal, S., M. L. Tremblay, G. Govoni, S. Gauthier, G. Sebastiani, D. Malo, E. Skamene, M. Olivier, S. Jothy, P. Gros.
1995
. The Ity/Lsh/Bcg locus: natural resistance to infection with intracellular parasites is abrogated by disruption of the Nramp1 gene.
J. Exp. Med.
182
:
655
3
Nauciel, C..
1990
. Role of CD4+ T cells and T-independent mechanisms in acquired resistance to Salmonella typhimurium infection.
J. Immunol.
145
:
1265
4
Mastroeni, P., B. Villarreal-Ramos, C. E. Hormaeche.
1992
. Role of T cells, TNFα and IFNγ in recall of immunity to oral challenge with virulent salmonellae in mice vaccinated with live attenuated aro-Salmonella vaccines.
Microb. Pathog.
13
:
477
5
Hess, J., C. Ladel, D. Miko, S. H. E. Kaufmann.
1996
. Salmonella typhimurium aroA infection in gene-targeted mice.
J. Immunol.
156
:
3321
6
Pope, M., I. Kotlarski.
1994
. Detection of Salmonella-specific L3T4+ and Lyt-2+ T cells which can proliferate in vitro and mediate delayed-type hypersensitivity reactivity.
Immunology
81
:
183
7
Pope, M., I. Kotlarski, K. Doherty.
1994
. Induction of Lyt-2+ cytotoxic T lymphocytes following primary and secondary Salmonella infection.
Immunology
81
:
177
8
Nauciel, C., F. Espinasse-Maes.
1992
. Role of γ interferon and tumor necrosis factor α in resistance to Salmonella typhimurium infection.
Infect. Immun.
60
:
450
9
Muotiala, A., P. H. Mäkelä.
1993
. Role of γ interferon in late stages of murine salmonellosis.
Infect. Immun.
61
:
4248
10
June, C. H., J. A. Bluestone, L. M. Nadler, C. B. Thompson.
1994
. The B7 and CD28 receptor families.
Immunol. Today
15
:
321
11
Shahinian, A., K. Pfeffer, K. P. Lee, T. M. Kündig, K. Kishihara, A. Wakeham, K. Kawai, P. S. Ohashi, C. B. Thompson, T. W. Mak.
1993
. Differential T cell costimulatory requirements in CD28-deficient mice.
Science
261
:
609
12
Lucas, P. J., I. Negishi, K. Nakayama, L. E. Fields, D. Y. Loh.
1995
. Naive CD28-deficient T cells can initiate but not sustain an in vitro antigen-specific immune response.
J. Immunol.
154
:
5757
13
Thompson, C. B..
1995
. Distinct roles for the costimulatory ligands B7-1 and B7-2 in T helper cell differentiation?.
Cell
81
:
979
14
Ferguson, S. E., S. Han, G. Kelsoe, C. B. Thompson.
1996
. CD28 is required for germinal center formation.
J. Immunol.
156
:
4576
15
Brown, D. R., J. M. Green, N. H. Moskowitz, M. Davis, C. B. Thompson, S. L. Reiner.
1996
. Limited role of CD28-mediated signals in T helper subset differentiation.
J. Exp. Med.
184
:
803
16
Elloso, M. M., P. Scott.
1999
. Expression and contribution of B7-1 (CD80) and B7-2 (CD86) in the early immune response to Leishmaniamajor infection.
J. Immunol.
162
:
6708
17
Zhan, Y., C. Cheers.
1996
. Either B7-1 or B7-2 is required for Listeria monocytogenes-specific production of γ interferon and interleukin-2.
Infect. Immun.
64
:
5439
18
Weintraub, B. C., L. Eckmann, S. Okamoto, M. Hense, S. M. Hedrick, J. Fierer.
1997
. Role of αβ and γδ T cells in the host response to Salmonella infection as demonstrated in T-cell-receptor-deficient mice of defined Ity genotypes.
Infect. Immun.
65
:
2306
19
Hoiseth, S. K., B. A. D. Stocker.
1981
. Aromatic-dependent Salmonella typhimurium are non-virulent and effective as live vaccines.
Nature
291
:
238
20
Spitalny, G., E. Havell.
1984
. Monoclonal antibody to murine γ interferon inhibits lymphokine-induced antiviral and macrophage tumoricidal activities.
J. Exp. Med.
159
:
1560
21
Cherwinski, H. M., J. H. Schumacher, K. D. Brown, T. R. Mosmann.
1987
. Two types of mouse helper T cell clone. III. Further differences in lymphokine synthesis between Th1 and Th2 clones revealed by RNA hybridization, functionally monospecific bioassays, and monoclonal antibodies.
J. Exp. Med.
166
:
1229
22
Ohara, J., W. E. Paul.
1985
. Production of a monoclonal antibody to and molecular characterization of B-cell stimulatory factor-1.
Nature
315
:
333
23
Sander, B., I. Hoiden, U. Andersson, E. Moller, J. S. Abrams.
1993
. Similar frequencies and kinetics of cytokine producing cells in murine peripheral blood and spleen.
J. Immunol. Meth.
166
:
201
24
Miyahira, Y., K. Murata, D. Rodriguez, J. R. Rodriguez, M. Esteban, M. M. Rodrigues, F. Zavala.
1995
. Quantification of antigen specific CD8+ T cells using an ELISPOT assay.
J. Immunol. Methods
181
:
45
25
Hess, J., D. Miko, A. Catic, V. Lehmensiek, D. G. Russell, S. H. E. Kaufmann.
1998
.
Mycobacterium bovis bacille Calmette Guerin strains secreting listeriolysin of Listeria monocytogenes. Proc. Natl. Acad. Sci. USA
95
:
5299
26
Mouse Genome Database (MGD), Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor. World Wide Web (URL:http://www.informatics.jax.org/), January, 1999.
27
Eisenstein, T. K., L. M. Killar, B. M. Sulzer.
1984
. Immunity to infection with Salmonella typhimurium: mouse strain differences in vaccine- and serum-mediated protection.
J. Infect. Dis.
150
:
425
28
Mastroeni, P., B. Villarreal-Ramos, C. E. Hormaeche.
1993
. Adoptive transfer of immunity to oral challenge with virulent salmonellae in innately susceptible BALB/C mice requires both immune serum and T cells.
Infect. Immun.
61
:
3981
29
Everest, P., J. Allen, A. Papakonstantinopoulou, P. Mastroeni, M. Roberts, G. Dougan.
1997
. Salmonella typhimurium infection in mice deficient in interleukin-4 production: role of IL-4 in infection-associated pathology.
J. Immunol.
159
:
1820
30
Sinha, K., P. Mastroeni, J. Harrison, R. D. Hormaeche, C. E. Hormaeche.
1997
.
Salmonella typhimurium aroA, htrA and aroD htrA mutants cause progressive infections in athymic (nu/nu) BALB/c mice. Infect. Immun.
65
:
1566
31
Keane-Myers, A., W. C. Gause, P. S. Linsley, S. Chen, M. Wills-Karp.
1997
. B7-CD28/CTLA-4 costimulatory pathways are required for the development of T helper cell 2-mediated allergic airway responses to inhaled antigens.
J. Immunol.
158
:
2042
32
Lane, P., C. Burdet, S. Hubele, D. Scheidegger, U. Müller, F. McConnell, M. Kosco-Vilbois.
1994
. B cell function in mice transgenic for mCTLA4-Hγ1: lack of germinal centers correlated with poor affinity maturation and class switching despite normal priming of CD4+ T cells.
J. Exp. Med.
179
:
819
33
Wu, Y., Q. Zhou, P. Zheng, Y. Liu.
1998
. CD28-independent induction of T helper cells and immunoglobulin class switches requires costimulation by heat-stable antigen.
J. Exp. Med.
187
:
1151
34
Zimmermann, C., P. Seiler, P. Lane, R. M. Zinkernagel.
1997
. Anti-viral immune response in CTLA4-transgenic mice.
J. Virol.
71
:
1802
35
King, C. L., J. Xianli, C. H. June, R. Abe, K. P. Lee.
1996
. CD28-deficient mice generate an impaired Th2 response to Schistosoma mansoni infection.
Eur. J. Immunol.
26
:
2448
36
Roy, M., A. Aruffo, J. Ledbetter, P. Linsley, M. Kehry, R. Noelle.
1995
. Studies on the interdependence of gp39 and B7 expression and function during antigen-specific immune responses.
Eur. J. Immunol.
25
:
596
37
Jaiswal, A. I., C. Dubey, S. L. Swain, M. Croft.
1995
. Regulation of CD40 ligand expression on naive CD4 T cells: a role for TCR but not co-stimulatory signals.
Int. Immunol.
8
:
275
38
Ding, L., J. M. Green, C. B. Thompson, E. M. Shevach.
1995
. B7/CD28-dependent and -independent induction of CD40 ligand expression.
J. Immunol.
155
:
5124
39
MacFarlane, A. S., M. G. Schwacha, T. K. Eisenstein.
1999
. In vivo blockage of nitric oxide with aminoguanidine inhibits immunosuppression induced by an attenuated strain of Salmonella typhimurium, potentiates Salmonella infection, and inhibits macrophage and polymorphonuclear leukocyte influx into the spleen.
Infect. Immun.
67
:
891
40
Collins, H. L., U. E. Schaible, S. H. E. Kaufmann.
1998
. Early IL-4 induction in bone marrow lymphoid cells by mycobacterial lipoarabinomannan.
J. Immunol.
161
:
5546
41
Brown, M. A., J. H. Pierce, C. J. Watson, J. Falco, J. N. Ihle, W. E. Paul.
1987
. B cell stimulatory factor-1/interleukin-4 mRNA is expressed by normal and transformed mast cells.
Cell
50
:
809
42
Brunner, T., C. H. Heusser, C. A. Dahinden.
1993
. Human peripheral blood basophils primed by interleukin-3 produce IL-4 in response to immunoglobulin E receptor stimulation.
J. Exp. Med.
177
:
605
43
Moore, K. W., A. O’Gara, R. de Waal Malefyt, P. Vieira, T. R. Mosmann.
1993
. Interleukin-10.
Annu. Rev. Immunol.
11
:
165
44
Pie, S., P. Matsiota-Bernard, P. Truffa-Bachi, C. Nauciel.
1996
. γ Interferon and interleukin-10 gene expression in innately susceptible and resistant mice during the early phase of Salmonella typhimurium infection.
Infect. Immun.
64
:
849
45
Schweitzer, A. N., A. H. Sharpe.
1998
. Studies using antigen-presenting cells lacking expression of both B7-1 (CD80) and B7-2 (CD86) show distinct requirements for B7 molecules during priming versus restimulation of Th2 but not Th1 cytokine production.
J. Immunol.
161
:
2762
46
Dubey, C., M. Croft, S. L. Swain.
1996
. Naive and effector CD4 T cells differ in their requirements for T cell receptor versus costimulatory signals.
J. Immunol.
157
:
3280
47
Corry, D. B., S. R. Reiner, P. S. Linsley, R. M. Locksley.
1994
. Differential effects of blockade of CD28–B7 on the development of Th1 or Th2 effector cells in experimental leishmaniasis.
J. Immunol.
153
:
4142
48
Brown, J. A., R. G. Titus, N. Nabavi, L. H. Glimcher.
1996
. Blockade of CD86 ameliorates Leishmania major infection by down-regulating the Th2 response.
J. Infect. Dis.
174
:
1303
49
Greenwald, R. J., P. Lu, M. J. Halvorson, X. Zhou, S.-J. Chen, K. B. Madden, P. J. Perrin, S. C. Morris, F. D. Finkelman, R. Peach, et al
1997
. Effects of blocking B7-1 and B7-2 interaction during a type 2 in vivo immune response.
J. Immunol.
158
:
4088
50
Lu, P., X. Zhou, S.-J. Chen, M. Moorman, S. C. Morris, F. D. Finkelman, P. Linsley, J. F. Urban, W. C. Gause.
1994
. CTLA-4 ligands are required to induce an in vivo interleukin 4 response to a gastrointestinal nematode parasite.
J. Exp. Med.
180
:
693
51
Subramanian, G., J. W. Kazura, E. Pearlman, X. Jia, I. Malhotra, C. L. King.
1997
. B7-2 requirement for helminth-induced granuloma formation and CD4 type 2 helper cell cytokine expression.
J. Immunol.
158
:
5914
52
Pie, S., P. Truffa-Bachi, M. Pla, C. Nauciel.
1997
. Th1 response in Salmonella typhimurium-infected mice with a high and low rate of bacterial clearance.
Infect. Immun.
65
:
4509
53
Sigal, L. J., H. Reiser, K. L. Rock.
1998
. The role of B7-1 and B7-2 costimulation for the generation of CTL responses in vivo.
J. Immunol.
161
:
2740
54
Kaufmann, S. H. E..
1999
. Cell-mediated immunity: dealing a direct blow to pathogens.
Curr. Biol.
9
:
97