Regulatory Roles of CD1d-Restricted NKT Cells in the Induction of Toxic Shock-Like Syndrome in an Animal Model of Fatal Ehrlichiosis

Mice.BALB/c wild-type (WT) or CD1d^−/− (C.129S2-Cd1tm1Gru/J; stock number 003814) mice were used for all experiments (Jackson Laboratories, Bar Harbor, ME) except those whose results are shown in Fig. 1, in which we compared the disease characteristics in WT BALB/c mice to those in WT C57BL/6J mice. The mice were gender matched for each experiment and were 6 to 12 weeks old. Mouse handling and experimental procedures were conducted in accordance with the recommendations of the University of Texas Medical Branch Institutional Animal Care and Use Committee.

• Open in new tab
• Download powerpoint

FIG. 1.

Survival, organ bacterial burdens, serum TNF-α and ALT concentrations, and liver lesions were similar in C57BL/6 and BALB/c WT mouse strains. Both WT BALB/c and C57BL/6 mice (nine mice per group) were inoculated with a high dose of IOE (1 × 10⁵ genome copies) and monitored daily for signs of illness. The data represent the results of three independent experiments with three mice per group. (A) Survival rate of both WT C57BL/6 and BALB/c mice. (B) Bacterial burdens in the liver, lung, spleen, and kidney were not significantly (P > 0.05) different for the two mouse strains on day 7 postinfection. The bars indicate the averages and the error bars indicate the standard deviations for triplicate amplifications with three mice per group. (C and D) Serum levels of TNF-α (C) and ALT (D) were not significantly (P > 0.05) different in IOE-infected C57BL/6 and BALB/c WT mice on day 5 postinfection. The bars indicate the means and the error bars indicate the standard deviations for three mice per group. Similar results were obtained in three independent experiments.

Experimental design.For all experiments a highly virulent Ehrlichia sp. (IOE), originally isolated from I. ovatus ticks in Japan, was used (17, 35). The genome copy numbers of IOE in our inocula were determined as previously reported (37). Mice were inoculated i.p. with a high dose of IOE (∼1 × 10⁵ genome copies), and three mice per group were sacrificed on days 5 and 7 postinfection to determine the bacterial burden and the levels of cytokines and liver enzymes in the serum, as well as organ pathology. Mice were monitored daily for signs of illness, including the presence of distress, loss of appetite, roughened fur, and decreased activity. On days 0, 5, and 7 individually tagged mice were monitored for changes in temperature and weight as previously described (19).

Determination of the ehrlichial load in collected organs by quantitative real-time PCR.To determine the relative differences in the ehrlichial burden between the different groups of mice, approximately 10 mg each of liver, lung, spleen, and kidney were collected at different time points postinfection and homogenized. DNA was extracted using a DNeasy tissue kit (Qiagen, Valencia, CA) with an elution volume of deionized water of 100 μl, and ehrlichial burdens were determined using an iCycler IQ multicolor real-time detection system (Bio-Rad, Hercules, CA). The following primers (Sigma-Genosys, St. Louis, MO) and probes (Biosearch Technologies, Novato, CA) targeting both the IOE and E. muris dsb (which encodes a thio-disulfide oxidoreductase) and host glyceraldehyde-3-phosphate dehydrogenase (GAPDH) genes were used: EM/IOE dsb forward primer (5′ CAGGATGGTAAAGTACGTGTGA 3′), EM/IOE dsb reverse primer (5′ TAGCTAAYGCTGCCTGGACA 3′), EM/IOE probe (5′ 6-carboxyfluorescein-AGGGATTTCCCTATACTCGGTGAGGC-BHQ1 3′), GAPDH forward primer (5′ CAACTACATGGTCTACATGTTC 3′), GAPDH reverse primer (5′ CTCGCTCCTGGAAGATG 3′), and GAPDH probe (5′ Cy5-CGGCACAGTCAAGGCCGAGAATGGGAAGC-BHQ2 3′). The comparative cycle threshold method (4) was used to determine the ehrlichial burdens in the harvested organs.

Determination of serum cytokine and ALT concentrations.At different time points after infection, serum samples were collected from infected mice. IFN-γ, TNF-α, and IL-10 levels were determined by using a Quantikine enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN) according to the manufacturer's recommendations. The serum concentrations of the liver enzyme alanine transaminase (ALT) (an enzyme that is released from injured hepatocytes) were determined by the Clinical Chemistry Laboratory, University of Texas Medical Branch.

Histology and ultrastructural analysis.Formalin-fixed, paraffin-embedded samples from the liver, lung, and spleen were sectioned and stained with hematoxylin and eosin (H&E). Histological sections were evaluated qualitatively, and liver lesions were assessed by using four parameters, hepatocyte damage, frequency of lesions, size of inflammatory lesions, and extent of lobular and perivascular inflammation, as described previously (30).

Isolation of splenic and hepatic mononuclear cells.Spleens and livers were harvested at different time points postinfection, and mononuclear cells were purified for flow cytometric analysis. Splenocyte single-cell suspensions were prepared in RPMI 1640 medium (500 ml) supplemented with 10% heat-inactivated fetal bovine serum, 1% HEPES buffer, and 1% penicillin G (50,000 U)-streptomycin sulfate (50,000 μg). Hepatic mononuclear cells were isolated using a modified enzymatic dispersal protocol as described previously (28), which allows the maximum yield and viability of hepatic mononuclear cells. Briefly, mouse livers were perfused with cold phosphate-buffered saline through the portal vein, and liver tissues were collected in Hanks' balanced salt solution (HBSS) without Ca²⁺ and Mg²⁺ and washed. Each liver was then perfused with warm HBSS (37°C) with Ca²⁺- and Mg²⁺-containing collagenase IV (500 mg/liter; 312 U/mg), DNase I (50 mg/liter), fetal calf serum (FCS) (2%), and bovine serum albumin (0.6%). Tissue was teased apart into ∼1- to 3-mm fragments while it was submersed in the warm HBSS and then incubated at 37°C with frequent shaking. After 15 min cold HBSS without Ca²⁺ and Mg²⁺ was added to stop the enzymatic digestion. The tissue was then passed through a 30-μm cell strainer to remove any large clumps or undissociated tissue. The filtered suspension was then centrifuged (500 × g) for 10 min at 4°C, and the cell pellet was washed twice in cold HBSS. Hepatocytes were removed by differential centrifugation (36 × g) for 1 min at 4°C, and the final pellet was resuspended in RPMI 1640 medium containing 10% FCS and 1% HEPES. Hepatic mononuclear cells were further purified for intracellular cytokine staining by overlaying cells obtained from each mouse liver carefully onto 5 ml of Lympholyte M (Cedarlane Laboratories, Burlington, NC), followed by centrifugation and collection of the upper two-thirds of the resulting density gradient. Trypan blue dye exclusion was used to determine cell viability.

Flow cytometry and intracellular cytokine staining.Splenocytes and/or hepatic mononuclear cells were harvested, counted, resuspended in fluorescence-activated cell sorter staining buffer (Dulbecco's phosphate-buffered saline without Mg²⁺ or Ca²⁺ containing 1% heat-inactivated FCS and 0.09% sodium azide), and filtered (0.2-μm-pore-size membrane), the pH was adjusted to 7.4 to 7.6, and aliquots were placed into a 96-well V-bottom plate (Costar, Corning, NY) at a concentration of 10⁶ cells per well. Fc receptors were blocked with a monoclonal antibody (clone 2.4G2) against mouse CD16 and CD32 cell surface antigens for 15 min. The cells were then washed and pelleted. Optimal concentrations were determined for the following monoclonal antibodies: allophyocyanin-Cy5-conjugated CD3 (clone 145-2C11) and CD11c (clone HL3); fluorescein isothiocyanate-conjugated CD4 (clone GK1.5), B220 (clone RAB-632), and Fas (clone FAS); phycoerythrin-conjugated CD8a (clone 53-6.7), CD40 (clone 3/23), and granzyme B (eBioscience, San Diego, CA) (clone 16G6); and peridin-chlorophyll protein-conjugated CD11b (clone M1/70), CD4 (clone RM4-5), and CD8 (clone 53-6.7). Corresponding isotype controls were used. All antibodies were obtained from BD Pharmingen (San Diego, CA) unless otherwise indicated. For ex vivo intracellular granzyme B detection, splenocytes or hepatic mononuclear cells were incubated at 37°C for 4 h in complete medium supplemented with BD Golgi plug (BD Pharmingen) according to the manufacturer's recommendations. Lymphocyte populations were gated based on forward and side scatter parameters, 20,000 to 50,000 events were collected using a BD FACSCalibur (BD Immunocytometry Systems, San Jose, CA) flow cytometer with CellQuest software (Immunocytometry Systems), and data were analyzed using FlowJo (Tree Star Inc., Ashland, OR) flow cytometric analysis software.

Statistical analysis.Data were analyzed using SigmaPlot software (SPSS, Chicago, IL), and P values were determined using the Student two-tailed t test. P values less than 0.01 were considered highly significant, and P values less than 0.05 were considered significant. Mouse groups contained three mice unless otherwise indicated, and the standard deviation was determined for each group. The standard error of the mean was used for analysis of combined data from more than one experiment.

Susceptibility, in vivo bacterial burdens, TNF-α production, and pathological changes in the liver of WT BALB/c mice following lethal ehrlichial infection are similar to those of WT C57BL/6 mice.To determine whether the outcome of ehrlichial infection is influenced by the genetic background of BALB/c mice, we compared the host defenses and immune responses to Ehrlichia of these mice to those of C57BL/6 WT mice. Both groups of WT mice succumbed to i.p. inoculation with a high dose of IOE by day 8 postinfection (Fig. 1A). No significant differences in organ bacterial burdens were observed on day 5 (data not shown) or day 7 postinfection (Fig. 1B). We next examined whether BALB/c mice developed toxic shock-like syndrome following lethal ehrlichial infection like C57BL/6 mice (19). To this end, we characterized some of the main features of Ehrlichia-induced toxic shock, including high serum levels of TNF-α and hepatic injury. Similar to infected C57BL/6 mice, WT BALB/c mice infected i.p. with a high dose (∼1 × 10⁵ genome copies) of IOE had high concentrations of serum TNF-α (Fig. 1C) and serum ALT (Fig. 1D) on day 5 postinfection. Similar to analyses of IOE-infected C57BL/6 mice (19, 20, 36, 37), hepatic histopathology analysis of WT BALB/c mice showed apoptosis of both Kupffer cells and hepatocytes on day 5 postinfection and severe multifocal hepatic necrosis on day 7 postinfection (data not shown). These data suggest that lethal ehrlichial infection causes pathology and fatal disease in both WT strains of mice, indicating that IOE induction of toxic shock-like syndrome occurs independent of the genetic background of the mice.

Resistance to Ehrlichia-induced toxic shock in the absence of CD1d and CD1d-reactive NKT cells.To identify a role for CD1d-restricted NKT cells in Ehrlichia-induced fatal toxic shock syndrome, we examined the effect of ehrlichial infection on the host defenses and the pathogenesis of ehrlichiosis in WT BALB/c and NKT cell-deficient CD1d^−/− mice with a BALB/c background. Because CD1d is required for positive selection of NKT cell precursors in the thymus (in addition to stimulation of variant NKT and iNKT cells in the periphery), CD1d^−/− mice lack the CD1d-restricted iNKT cell population (27). Briefly, WT and CD1d^−/− mice were infected i.p. with a high dose of IOE (∼1 × 10⁵ genomes copies), their survival was monitored, and the host immune response was evaluated on days 5 and 7 postinfection. BALB/c CD1d^−/− mice succumbed to high-dose IOE infection on day 8 postinfection, 1 day after the death of WT BALB/c mice (Fig. 2A). However, unlike WT BALB/c mice, CD1d^−/− mice did not develop signs of toxic shock-like syndrome. Compared to WT mice on day 5 postinfection, fatal ehrlichial infection did not result in weight loss or hypothermia (data not shown) in CD1d^−/− mice. In addition, the IOE-infected CD1d^−/− mice had significantly lower serum concentrations of the liver enzyme ALT on day 5 postinfection (Fig. 2B).

• Open in new tab
• Download powerpoint

FIG. 2.

Survival, serum levels of liver enzyme and cytokines, and organ bacterial burdens in IOE-infected CD1d^−/− mice and BALB/c WT controls. (A) Survival rates of BALB/c CD1d^+/+ and CD1d^−/− mice following i.p. inoculation of a high dose of IOE (1 × 10⁵ genome copies). There was no significant difference in survival between the two groups of mice (nine mice per group). The data represent the results of one of three independent experiments with a total of nine mice per group. (B to E) Serum levels of ALT, TNF-α, IFN-γ, and IL-10 on day 5 following i.p. infection of knockout and WT BALB/c mice with a high dose of IOE. CD1d^−/− mice had significantly lower (P < 0.01) serum levels of ALT (B) and TNF-α (C) than infected WT mice. The serum levels of IFN-γ (D) and IL-10 (E) were not significantly different for the two groups. (F and G) Ehrlichial burdens in different organs of CD1d^−/− and WT mice on days 5 (F) and 7 (G) postinfection. CD1d^−/− mice had significantly (P < 0.01) greater bacterial burdens in all analyzed organs on day 5 postinfection and in only the liver on day 7 postinfection. The bars indicate the means and the error bars indicate the standard deviations for three mice per group. Similar results were obtained in three independent experiments. P values less than 0.01 were considered highly significant (filled diamond), and P values less than 0.05 were considered significant (asterisk). KO, knockout.

We next examined the effects of the CD1d molecule and NKT cells on the systemic levels of cytokines known to be associated with Ehrlichia-induced toxic shock. Similar to the results of our previous studies (37) using WT C57BL/6 mice, fatal disease in IOE-infected BALB/c mice was associated with significant high serum levels of TNF-α, IFN-γ, and IL-10 (Fig. 2 C to E). In contrast, CD1d^−/− mice had lower concentrations of TNF-α (Fig. 2C) but not of IFN-γ (Fig. 2D) or IL-10 (Fig. 2E) on day 5 postinfection.

Our previous studies demonstrated that IOE-induced tissue damage and toxic shock occur in WT C57BL/6 mice in the absence of overwhelming infection. Therefore, we first compared the differences in bacterial burden between the WT BALB/c and knockout mice using real-time PCR. The CD1d^−/− mice had significantly higher ehrlichial burdens in the liver, lung, spleen, and kidney on day 5 postinfection and in the liver on day 7 postinfection (Fig. 2F and 2G). The difference in bacterial burden was most striking in the liver on day 5 postinfection, in which the CD1d^−/− mice had a fourfold-greater ehrlichial burden than the WT mice. Examination of liver histopathology in infected CD1d^−/− mice on day 5 postinfection revealed less inflammation and very few apoptotic cells (Fig. 3A and 3C) compared to WT BALB/c mice, in which acute inflammation and a significantly higher number of apoptotic liver cells, including both hepatocytes and cells lining the sinusoidal spaces and vascular lumens (i.e., Kupffer cells, monocytes, and/or endothelial cells) (Fig. 3B and 3C), were observed. No evidence of necrosis was detected in BALB/c WT or CD1d^−/− mice on day 5 postinfection. Electron microscopic analysis of livers from infected WT and CD1d^−/− mice further confirmed the difference in apoptosis between the two groups and also demonstrated that most apoptotic cells were indeed Kupffer cells/macrophages (data not shown). Interestingly, on day 7 postinfection, WT and CD1d^−/− mice both developed multifocal hepatocellular necrosis similar to that previously reported for WT C57BL/6 mice (37). Thus, although the absence of CD1d and iNKT cells did not prevent Ehrlichia-induced mortality, the CD1d molecule and/or CD1d-restricted NKT cells mediated the early systemic overproduction of proinflammatory TNF-α, apoptotic death of host cells, liver injury, and hypothermia that collectively comprise the major manifestations of Ehrlichia-induced toxic shock syndrome.

• Open in new tab
• Download powerpoint

FIG. 3.

Differences in hepatic histopathology between WT BALB/c and CD1d^−/− mice. CD1d^−/− mice (A) had fewer apoptotic cells (arrows) in the liver than CD1d^+/+ WT BALB/c mice (B) on day 5 postinfection. H&E-stained liver sections showed similar results for three mice per group. (C) Numbers of apoptotic cells or nuclei and apoptotic bodies present in H&E-stained liver sections from each mouse in 10 medium-power fields (magnification, ×200). The bars indicate the means and the error bars indicate the standard deviations for three mice per group. Similar results were obtained in two independent experiments. P values less than 0.01 were considered highly significant (filled diamond). The cells undergoing apoptosis were similar to those reported previously (36) and included sinusoidal lining cells, as well as hepatocytes. KO, knockout.

Absence of CD1d and CD1d-reactive NKT cells restores the populations of splenic CD4⁺ and CD8⁺ T cells.We demonstrated previously (18, 19, 37) that fatal Ehrlichia-induced toxic shock-like syndrome is associated with significant apoptosis of splenic CD4⁺ T cells and decreases in the percentages and absolute numbers of CD4⁺ and CD8⁺ T cells in the spleen. Substantial lymphopenia has also been reported in patients with severe disease. To determine the role of CD1d and CD1d-restricted NKT cells in T-cell apoptosis, we examined the differences in the percentages of CD4⁺ and CD8⁺ lymphocytes in the WT and CD1d^−/− mice by flow cytometry. The IOE-infected CD1d^−/− mice had higher percentages (Fig. 4A) and absolute numbers of splenic CD4⁺ and CD8⁺ T cells than the WT mice on day 7 postinfection (the splenocyte mononuclear cell yields determined by trypan blue exclusion were similar in the two groups of mice [data not shown]). These data suggest that CD1d-restricted NKT cells mediate the significant reduction in the total number of CD4⁺ and CD8⁺ T cells.

• Open in new tab
• Download powerpoint

FIG. 4.

Absence of NKT cells in IOE-infected CD1d^−/− mice restored the percentages of both CD4⁺ and CD8⁺ T cells in the spleen and decreased Fas expression on CD4⁺ T cells and granzyme B expression in CD8⁺ T cells. BALB/c WT and CD1d^−/− mice were inoculated with a high dose of IOE. The percentages and absolute numbers of CD4⁺ and CD8⁺ T cells, as well as the expression of Fas and granzyme on splenic CD4⁺ and hepatic CD8⁺ T cells, were determined by flow cytometry on day 7 postinfection. Lymphocytes were gated based on size and granularity, and 50,000 events were studied. (A) CD1d^−/− mice had significantly higher percentages and absolute numbers of CD4⁺ and CD8⁺ T cells (splenocyte yields were not significantly different as determined by trypan blue exclusion) than CD1d^+/+ BALB/c mice on day 7 postinfection. The bars indicate the means and the error bars indicate the standard deviations for three mice per group. Similar results were obtained in two independent experiments. (B and C) CD1d^−/− mice had lower expression of Fas on splenic CD3⁺ CD4⁺ T cells and of granzyme B in hepatic CD3⁺ CD8⁺ T cells on day 7 postinfection. The dot plots shown are similar to those obtained for all three mice per group, and the data are representative of two independent experiments. The percentages below the plots are the percentages of Fas⁺ CD4⁺ T cells (upper values) and granzyme-positive CD8⁺ T cells (lower values) shown in the right upper quadrants of each dot plot. (D and E) Percentages of splenic CD3⁺ CD4⁺ T cells and hepatic CD3⁺ CD8⁺ T cells expressing Fas and granzyme B, respectively, shown graphically for statistical comparison. The bars indicate the means and the error bars indicate the standard deviations for three mice per group. P values less than 0.01 were considered highly significant (filled diamond), and P values less than 0.05 were considered significant (asterisk). KO, knockout.

Fas/FasL interaction plays a role in CD1-restricted-NKT-mediated apoptosis of CD4⁺ T cells.In several infectious and noninfectious disease models, it has been demonstrated that the induction of apoptosis in CD4⁺ T cells, as well as cytotoxic functions acquired by NKT cells, involves the Fas/FasL pathway (2, 39). We hypothesized that NKT cells mediate apoptosis of Ehrlichia-specific CD4⁺ effector cells via this mechanism. Consistent with this hypothesis, we observed that lethal ehrlichial infection resulted in up-regulation of Fas expression on splenic CD4⁺ T cells from WT and CD1d^−/− mice (Fig. 4B and 4D) on day 7 postinfection. However, the absence of CD1d-restricted NKT cells in infected CD1d^−/− mice significantly down-regulated the expression of Fas on splenic CD4⁺ T cells compared to the cells of infected WT mice (Fig. 4B and 4D).

CD1d-restricted NKT cell-mediated activation of cytotoxic CD8⁺ T cells involves granzyme B.As mentioned above, our previous studies demonstrated that there is a strong correlation between fatal disease and the expansion of TNF-α-producing CD8⁺ T cells. In addition, we recently reported that β₂m^−/− mice that lack both CD8⁺ T cells and CD1d-restricted NKT cells are more resistant to lethal ehrlichial infection than WT mice (18). To determine the contribution of CD1d-restricted NKT cells to the induction of pathogenic cytotoxic CD8⁺ T cells, we examined differences in intracellular expression of granzyme B (an important molecule involved in cytotoxic effector functions of CD8⁺ T cells) (32) in hepatic CD8⁺ T cells from CD1d^−/− and WT IOE-inoculated mice. We chose the liver since it is the predominant organ that has marked ehrlichial tropism as well as extensive Ehrlichia-induced pathology. Ehrlichial infection of the CD1d^−/− mice resulted in expression of intracellular granzyme B in CD3⁺/CD8⁺ T cells that was significantly lower than that observed in the WT mice (Fig. 4C and 4E). Taken together, our data suggest that CD1d-restricted NKT cells could potentially mediate apoptosis of CD4⁺ T cells via the Fas/FasL pathway and induction of cytotoxic CD8⁺ T cells in Ehrlichia-induced toxic shock.

IOE infection suppresses the expression of the costimulatory molecule CD40 on DCs, while NKT cells up-regulate CD40 expression on APCs.Based on the data described above, we hypothesized that NKT cells mediate induction of pathogenic CD8⁺ T cells and play a direct role in tissue injury and apoptosis of host cells during Ehrlichia-induced toxic shock via cognate interactions with APCs, including macrophages, DCs, and B cells. To test this hypothesis, we first quantified differences in CD11b⁺ cell frequency in harvested splenocytes by flow cytometry. Our data showed that CD1d^−/− mice had significantly higher percentages and absolute numbers of CD11b⁺ cells in the spleen on day 5 postinfection (Fig. 5A). These observations correlate with the decreased number of apoptotic hepatic phagocytic cells and hepatocytes in the IOE-infected CD1d^−/− mice compared to WT mice (Fig. 3A and B).

• Open in new tab
• Download powerpoint

FIG. 5.

Absence of NKT cells resulted in significantly greater numbers of splenic CD11b⁺ cells but significantly decreased the expression of the costimulatory molecule CD40 on splenic CD11b⁺, CD11c⁺, and B220⁺ cells on day 5 postinfection. BALB/c mice were inoculated with IOE (1 × 10⁵ IOE genome copies) and were sacrificed on day 5 postinfection. Splenocytes were isolated and stained with fluorescently labeled murine monoclonal antibodies, which was followed by flow cytometric analysis, and 20,000 events were collected. (A) CD1d^−/− mice had significantly higher percentages of CD11b⁺ cells than WT mice. (B) CD11b⁺, CD11c⁺, and B220⁺ populations of cells were gated within their representative dot plots, and the mean fluorescence intensity of CD40 expression was determined using FlowJo. The results show that there was significant suppression of CD40 expression on CD11c⁺ cells in WT mice during IOE infection compared to uninfected WT mice. In addition, decreased expression of CD40 on CD11c⁺, CD11b⁺, and B220⁺ cells was detected in infected CD1d^−/− mice compared to infected WT mice on day 5 after IOE infection. The data are representative data for two independent experiments. P values less that 0.01 were considered highly significant (filled diamond). KO, knockout.

We next determined the effect of IOE-activated NKT cells on the activation and/or maturation of APCs by measuring differences in the up-regulation of the costimulatory molecule CD40. CD40 was chosen since it is the main costimulatory molecule that provides strong DC maturation signals through binding of CD40L on the surface of activated conventional T cells or NKT cells (11). In addition, increasing CD40 expression on APCs is a mechanism by which a robust CTL response can be initiated in the absence of CD4 help (10), which is important during severe ehrlichiosis, in which a significant decrease in the number of CD4⁺ Th1 cells has been observed (19). Compared to uninfected controls, IOE infection of WT BALB/c mice resulted in down-regulation of CD40 expression on CD11c⁺ DCs but not on CD11b⁺ CD11c⁻ macrophages or B220⁺ cells (Fig. 5B). Further down-regulation of CD40 expression on CD11c⁺ DCs was observed in the absence of CD1d-restricted NKT cells in infected CD1d^−/− mice. Similarly, CD40 expression was significantly lower on CD11b⁺ CD11c⁻ macrophages and B220⁺ cells in the CD1d^−/− mice than in the WT mice. The difference was most significant (P < 0.001) for B220⁺ cells (Fig. 5B). These data suggest that unlike the maturation of other APCs, DC maturation is suppressed directly by IOE and that CD1d-restricted NKT cells promote up-regulation of costimulatory molecules during ehrlichial infection.