Unlike T cells restricted by major histocompatibility complex (MHC) class Ia or class II molecules, T cells restricted by MHC class I–like molecules demonstrate properties of both innate and adaptive immunity and are therefore considered innate-like lymphocytes (ILLs). ILLs are believed to have immunoregulatory functions, but their roles in autoimmunity and defense against infections remain elusive. To study the properties of ILLs, we generated mice expressing only MHC class I–like molecules by crossing CIITA−/− with Kb−/−Db−/− mice. Surprisingly, these mice developed a lymphoproliferative syndrome and autoimmunity, most notably inflammatory bowel disease (IBD) and insulitis. The CD8+ ILLs in these mice exhibit a constitutively activated phenotype, and depletion of these cells abolished the autoimmune disorders. In addition, adoptive transfer of CD8+ ILLs from Kb−/−Db−/−CIITA−/− mice to Rag-1−/−pfn−/− mice also resulted in IBD and insulitis. These findings provide direct evidence that CD8+ ILLs are sufficient to initiate and mediate autoimmune diseases.

MHC class Ia–restricted CD8+ T cells are the classic effector cells involved in cytotoxicity during adaptive immune responses. In addition to MHC class Ia, the mammalian genome encodes proteins with homology to the α chain of MHC class Ia molecules, which are often referred to as MHC class I–like molecules. These proteins are involved in the selection and maintenance of innate-like lymphocytes (ILLs), including NKT cells and MHC class Ib– restricted CD8+ T cells, which exhibit an activated memory phenotype and are thought to have critical immunomodulatory functions (15). By using genetically modified mice, recent studies have shown that NKT cells, restricted by the MHC class I–like molecule CD1d, play a central role in the pathogenesis of Th2 cell–driven immune disorders such as ulcerative colitis (6) and asthma (7). Importantly, NKT cells rapidly produce large amounts of IL-4 and IL-13 upon activation.

CD8+ T cells restricted by MHC class Ib molecules have an activated phenotype (8) and promptly produce cytokines upon activation (2), which are typical characteristics of ILLs. These cells also have a variety of immune regulatory functions, such as suppressing the activity of CD4+ T cells (9, 10), promoting differentiation of Th1 cells (2), and up-regulating immunity to infections (1114) and tumors (15). Thus, like NKT cells, these MHC class Ib–restricted CD8+ T cells have the characteristics of ILLs. These cells may serve as a bridge between innate and adaptive immunity, especially in helping the differentiation of Th1 cells by producing IFN-γ at the early stage of an immune response. Because CD8+ ILLs exhibit an activated phenotype immediately after their maturation in the thymus (8), these cells are likely primed by and specific to self-antigens, suggesting that these cells might play a role in the pathogenesis of autoimmunity. Under normal physiological conditions, the number of ILLs is limited, and these cells are relatively inert, probably because the activity of these cells is controlled by regulatory mechanisms. We therefore hypothesized that, upon release from immune regulation, ILLs can initiate inflammation and autoimmunity. This possibility is supported by earlier studies, which showed that MHC class II–deficient mice exhibit autoimmune-like syndromes (16, 17). The pathogenesis of autoimmunity in these animals is likely initiated by MHC class I–restricted T cells that are normally suppressed by CD4+ T cells. The identity of the CD8+ T cells that cause inflammation in these MHC class II– deficient mice is unclear.

Experiments regarding the immunological functions of ILLs have been hampered by the lack of specific markers for these cells. To circumvent this problem, we have used a genetic approach to generate mice that exclusively express MHC class I–like molecules. However, generating mice that are deficient in Kb, Db, and IAb genes (C57BL/6 mice do not express IE and L molecules) cannot be easily accomplished, because these genes are clustered on the H-2 locus of chromosome 17. We therefore used CIITA−/− mice, which fail to express MHC class II molecules and produce few mature CD4+ T cells (18), except for CD1d-restricted CD4+ NKT cells. Thus, we crossed Kb−/−Db−/− with CIITA−/− mice to generate mice lacking both MHC class II and Ia molecules (Kb−/−Db−/−CIITA−/−), so that they only express MHC class I–like molecules. To ensure that the T cells that developed in Kb−/−Db−/−CIITA−/− mice are indeed restricted by MHC class I–like molecules, we also generated β2m−/− CIITA−/− mice, which are devoid of all MHC class I, class I– like, and class II molecules. We found that Kb−/−Db−/− CIITA−/− mice, but not β2m−/−CIITA−/− mice, exhibited lymphoproliferative disease with an abnormal expansion of activated CD8+ T cells, spontaneous inflammatory bowel disease (IBD), and insulitis. Adoptive transfer of CD8+ T cells from Kb−/−Db−/−CIITA−/− mice to syngeneic RAG−/− pfn−/− mice induced intestinal inflammation and insulitis. Furthermore, depletion of CD8+ T cells from Kb−/−Db−/− CIITA−/− mice failed to show symptoms of IBD or insulitis, suggesting that T cells selected by MHC class I–like molecules are self-reactive, pathogenic, and capable of initiating and sustaining autoimmune inflammation.

Results

Kb−/−Db−/−CIITA−/− mice express MHC class I–like molecules but do not support the survival of MHC class Ia– or class II–restricted T cells

Double staining with 28-8-6 and 25-9-17 antibodies, which recognize MHC class Ia (both Kb and Db) and class II IAb proteins, respectively, showed that splenocytes from Kb−/−Db−/−CIITA−/− and β2m−/−CIITA−/− mice lack surface expression of MHC class Ia and class II IAb molecules (Fig. 1 a). Expression of MHC class I–like molecules, including Qa-1b, Qa-2, and H2-M3, was examined by staining with 6A8.6F10.1A, 1-1-2, and H2-M3.130 antibodies, respectively. Kb−/−Db−/−CIITA−/− mice expressed these MHC class I–like molecules at similar levels as wild-type B6 mice, whereas these molecules were undetectable in β2m−/−CIITA−/− mice (Fig. 1 b). To demonstrate that these mice were functionally devoid of MHC class Ia and class II molecules, we adoptively transferred Kb−/−Db−/−CIITA−/− mice with carboxyfluorescein diacetate succinimidyl ester (CFSE)–labeled transgenic (Tg) OT-II CD4+ or OT-I CD8+ T cells and investigated the survival of these cells 1 wk later. By tracking CFSE-labeled cells, or by using clonotypic antibodies specific for OT-II or OT-I Tg T cells, we failed to detect any of these cells in Kb−/−Db−/−CIITA−/− mice 1 wk after adoptive transfer. However, both OT-II and OT-1 cells survived and exhibited cell division in RAG−/−pfn−/− mice (Fig. 1 c). Therefore, Kb−/−Db−/−CIITA−/− mice are devoid of functional MHC class Ia and class II molecules and T cells restricted by these molecules. These mice express only MHC class I–like molecules and thus only harbor T cells restricted by MHC class I–like molecules, providing a unique opportunity to study the function of ILLs.

Kb−/−Db−/−CIITA−/− mice exhibit inflammation in the gut and pancreas

At 12 wk of age, all Kb−/−Db−/−CIITA−/− mice (n = 32) exhibited enlarged mesenteric lymph nodes, distended and highly visible Peyer's patches, and swollen intestines, as depicted in Fig. 2 a, closely resembling IBD. Histologically, there was moderate to severe inflammation characterized by infiltration of mononuclear cells (Fig. 2 b). In contrast, Kb−/−Db−/−, CIITA−/−, and β2m−/−CIITA−/− mice (n = 24), showed no signs of intestinal inflammation, both by gross examination and by histological analysis.

In humans, the histocompatibility leukocyte antigen (HLA)–E locus is associated with susceptibility to type I diabetes (19). In mice, pancreatic β cells have been found to express the HLA-E homologue Qa-1b, presenting a nonomeric peptide derived from the preproinsulin leader sequence. This complex is recognized by Qa-1b–restricted CD8+ T cells and lyses pancreatic β cells (20). Thus, CD8+ T cells restricted by MHC class I–like molecules may contribute to the inflammation in diabetes. We therefore examined Kb−/−Db−/−CIITA−/− mice for the development of insulitis. Gross examination revealed enlarged and fragile pancreati (not depicted), and histological sections revealed characteristic lymphocyte infiltration in the islets, indicative of the onset of autoimmune type I diabetes (Fig. 2 c).

Kb−/−Db−/−CIITA−/− mice exhibit lymphoproliferative disease

Lymphoid organs such as spleens and lymph nodes of Kb−/− Db−/−CIITA−/− mice were dramatically enlarged, although thymi were normal in size (unpublished data). We found that Kb−/−Db−/−CIITA−/− mice produced few CD4+ T cells but a large number of CD8+ T cells in their spleens. In contrast, β2m−/−CIITA−/− mice had virtually no mature CD4+ or CD8+ T cells (Fig. 3, a and b). Because T cells are required for most autoimmune disorders, we analyzed T cell subsets in the spleens of Kb−/−Db−/−CIITA−/− mice. The numbers of CD44hiCD8+ T cells were dramatically increased, whereas CD8+ T cells with a naive phenotype were virtually absent (Fig. 3 b). As MHC class II molecules are absent in these mice, we determined the phenotype of the few CD4+ T cells in Kb−/−Db−/−CIITA−/− mice. Staining with anti-DX5, a marker for NK and NKT cells, suggested that most CD4+ cells in Kb−/−Db−/−CIITA−/− mice are NKT cells (Fig. 3 c). Therefore, Kb−/−Db−/− CIITA−/− mice lack mature MHC class II–restricted CD4+ T cells. Collectively, these data suggest that the large number of CD8+ T cells in Kb−/−Db−/−CIITA−/− mice are selected and maintained by MHC class I–like molecules.

CD8+ T cells in Kb−/−Db−/−CIITA−/− mice are activated spontaneously

Because CD8+ T cells restricted by MHC class I–like molecules are ILLs (21), we investigated activation markers on CD8+ T cells of Kb−/−Db−/−CIITA−/− mice. The majority of CD8+ T cells have an activated phenotype (CD44hi, β7integrinlow, CD122hi, and Ly6Chi; Fig. 4 a), in agreement with previous studies indicating that CD8+ T cells are constitutively activated in naive Kb−/−Db−/− animals (8, 22). Therefore, CD8+ T cells in Kb−/−Db−/−CIITA−/− mice appear to be activated without exposure to exogenous antigens, indicating that these T cells are likely primed by self-antigens. To determine the organs where these cells are primed, we examined the thymus and found that a large proportion of the thymic single-positive CD8+ T cells already exhibited an activated phenotype (Fig. 4 a). However, no abnormal expansion of thymic single-positive CD8+ T cells was found in Kb−/−Db−/−CIITA−/− mice (Fig. 4 b). The absolute numbers of single-positive CD8 T cells in the thymus in wild-type and Kb−/−Db−/−CIITA−/− mice are not increased (Table S1). This suggests that the elevated CD8+ T cell number found in the periphery is not a result of enhanced thymic selection. Instead, it is likely that CD8+ T cells selected on MHC class I–like molecules in the thymus are activated by self-antigens and migrate to the peripheral lymphoid organs, where they expand by hyperlymphoproliferation. Interestingly, we observed that a large proportion of these cells also expresses NKG2c (Fig. 4 c), an activating NK cell receptor. Expression of NKG2c has recently been associated with inflammation and production of inflammatory cytokines by HLA-E–restricted ILLs in human celiac disease (23). Although this finding is consistent with the innate-like property of pathogenic CD8+ T cells in Kb−/−Db−/−CIITA−/− mice, NKG2c is not a marker for ILLs because it is acquired only during the progression of inflammation (23). Nevertheless, expression of NKG2c by CD8+ T cells in Kb−/−Db−/− CIITA−/− mice again confirmed the activated phenotype of these cells.

To examine the intrinsic proliferation of CD8+ T cells in thymus and peripheral lymphoid organs, we performed a BrdU incorporation experiment in vivo. We found that an increased number of CD8+ T cells in the thymus and spleen in Kb−/−Db−/−CIITA−/− mice, as compared with wild-type C57BL/6 mice, incorporated BrdU (Fig. 4 d). Therefore, the hyperproliferation of CD8+ T cells in these animals is already initiated in the thymus and continues in the periphery.

Cytokines and chemokines produced by CD8+ T cells of Kb−/−Db−/−CIITA−/− mice

To investigate cytokine and chemokine production by CD8+ T cells in Kb−/−Db−/−CIITA−/− mice, we activated sorted CD8+ T cells from B6, Kb−/−Db−/−, and Kb−/−Db−/− CIITA−/− mice with plate-bound anti-CD3 and anti-CD28 antibodies. We found a considerable increase in IFN-γ, TNF-α, IL-17, macrophage inflammatory protein 1α and 1β, and RANTES for CD8+ T cells from Kb−/−Db−/− CIITA−/− mice (Fig. 4 e) but no change in IL-4, IL-6, and IL-10 (unpublished data). Interestingly, although the levels of inflammatory cytokines produced by CD8+ T cells isolated from Kb−/−Db−/− mice are higher than C57BL/6 mice, they remain substantially lower than Kb−/−Db−/−CIITA−/− mice. It is possible that the CD8+ T cells in Kb−/−Db−/− mice are regulated by CD4+ T cells and that their activation potential is influenced by the presence of CD4+ T cells. Supplementation of IL-2 to the culture medium did not change the cytokine production pattern of CD8+ T cells from Kb−/−Db−/− mice. This observation suggests that abrogation of acquired tolerance in CD8+ T cells of Kb−/−Db−/− mice is independent of IL-2. The mechanisms controlling the tolerance of these pathogenic T cells are currently being investigated.

Surprisingly, we found that splenocytes from Kb−/−Db−/− CIITA−/− mice activated with anti-CD3 antibodies produced large amounts of IL-12 (unpublished data). T cells generally do not produce IL-12. However, it is conceivable that high levels of IFN-γ, produced by the CD8+ T cells restricted by MHC class I–like molecules, are responsible for activating IL-12–producing cells such as macrophages. To test this hypothesis, we activated splenocytes with anti-CD3 overnight and performed intracellular staining of IL-12 in combination with surface marker staining. We found that CD11b+ cells, but not CD8+ T cells, showed strong intracellular IL-12 expression (Fig. S1). Neutralization of IFN-γ with a specific antibody inhibited IL-12 production by CD11b+ cells. This is in agreement with our earlier study showing that anti-CD3–induced IL-12 production in vivo is mediated by the early burst of IFN-γ from CD8+ T cells restricted by MHC class I–like molecules (2). These findings provide direct evidence that, in the absence of conventional T cells, CD8+ T lymphocytes restricted by MHC class I–like molecules are pathogenic and sufficient to mediate autoimmune diseases. To determine whether the accumulation of peripheral CD8+ T cell is caused by clonal expansion, we tested the Vβ usage of CD8+ T cells in Kb−/−Db−/−CIITA−/− and corresponding wild-type mice. Although these cells expressed a variety of TCRβ chains, we were unable to detect T cells expressing Vβ9, Vβ10, and Vβ14. We observed a substantial increase in the usage of Vβ2, Vβ6, Vβ8.1,2, and Vβ13 (Table S2). Thus, CD8+ T cells restricted by MHC class I–like molecules bear activated phenotypes, produce increased amounts of inflammatory cytokines, and have a biased Vβ repertoire.

CD8+ T cells in Kb−/−Db−/−CIITA−/− mice are pathogenic

To verify the role of CD8+ T cells in the pathological changes in Kb−/−Db−/−CIITA−/− mice, we adoptively transferred CD8+ T cells from Kb−/−Db−/−CIITA−/− mice to RAG−/− pfn−/− mice. Within 3 wk, recipient RAG−/−pfn−/− mice developed intestinal and pancreatic inflammation (Fig. 5, a and b), whereas adoptive transfer of CD8+ T cells from normal mice did not induce such changes. Furthermore, we depleted CD8+ T cells from Kb−/−Db−/−CIITA−/− mice using monoclonal anti-CD8 antibodies (Fig. 5, c and d). These mice also failed to exhibit intestinal or pancreatic inflammation.

Discussion

Our studies have revealed that mice bearing only MHC class I–like molecules produce predominantly activated CD8+ T cells that secret considerably more inflammatory cytokines than CD8+ T cells from wild-type mice. Therefore, it is likely that CD8+ T cells restricted by MHC class I–like molecules are pathogenic and mediate the observed autoimmunity in Kb−/−Db−/−CIITA−/− mice. Adoptive transfer of CD8+ T cells from Kb−/−Db−/−CIITA−/− mice induced colonic and pancreatic inflammation, whereas Kb−/−Db−/−CIITA−/− mice depleted of CD8+ T cells failed to develop inflammation in the gut and pancreas, suggesting that CD8+ T cells restricted by MHC class I–like molecules play a critical role in the observed inflammation in Kb−/−Db−/−CIITA−/− mice. Thus, our results provide strong evidence that CD8+ T cells restricted by MHC class I–like molecules, in the absence of MHC class Ia–restricted CD8+ T cells and class II–restricted CD4+ T cells, become pathogenic, resulting in the development of the observed autoimmune diseases. The early spontaneous onset of autoimmune disease in an otherwise resistant mouse strain clearly indicates that these cells are major initiators and mediators of autoimmunity. The lack of autoimmunity in Kb−/−Db−/− and CIITA−/− mice strongly argues that CD4+ T cells and MHC class Ia–restricted CD8+ T cells are both capable of suppressing the activity of the CD8+ T cells restricted by MHC class I–like molecules. Because MHC class I–like molecules are nonpolymorphic in nature and either fail to present antigen (e.g., thymus leukemia), glycolipid (e.g., CD1d), or a very restricted set of self-antigens (e.g., Qa-1b presents a peptide derived from Hsp60), autoimmune disease in Kb−/−Db−/−CIITA−/− mice might be initiated by a small population of oligoclonal CD8+ T cells that are activated by a restricted set of self-antigens presented by MHC class I–like molecules. Although our data clearly show that T cells restricted by MHC class I–like molecules are important for the autoimmune phenotype, we do not know which antigen is responsible for eliciting the autoimmune responses.

Breaking tolerance of CD8+ T cells restricted by MHC class I–like molecules may represent a key step in the pathogenesis of autoimmune disorders. A previous report has established an etiological link between infection by gram-negative bacteria and autoimmune diseases (24). It is possible that, in mice or humans with genetic susceptibility to autoimmunity, microbial products can activate innate immune components that alter the immune tolerant status, thus allowing pathogenic CD8+ T cells restricted by MHC class I– like molecules to expand and initiate autoimmune disease. In fact, a recent report showed that a Qa-1b–restricted CD8+ T cell clone can kill macrophages that are stressed or infected with Salmonella (24). Interestingly, the killing of infected cells in this system was not guided by the harbored organism but depended on host stress proteins such as Hsp60. Furthermore, these authors showed that Qa-1 binds with an Hsp60-derived peptide, which can be recognized by a Qa-1b–restricted CD8+ T cell clone (24). Interestingly, Qa-1b can also bind a peptide derived from preproinsulin and activate the same Qa-1b–restricted CD8+ T cell clone that recognizes Hsp60. This finding suggests that self-peptides presented by Qa-1 can mediate the development of autoimmunity. Nevertheless, the contribution of Qa-1–presented peptides to autoimmunity in our model system, in which all T cells are restricted by MHC class I–like molecules, remains to be established.

CD8+ T cells restricted by MHC class I–like molecules are selected and activated in the thymus. Under normal conditions, these cells are tolerant, but when tolerance is broken these T cells rapidly expand in the peripheral lymphoid organs and initiate autoimmune disease. Our results indicate that these CD8+ T cells are highly autoreactive. Furthermore, these cells are fully capable of initiating and propagating autoimmune diseases on their own. MHC class I–like molecules are nonpolymorphic and bind to a limited set of self-antigens, resulting in a small population of selected T cells with limited diversity. As Kb−/−Db−/−CIITA−/− mice do not bear conventional CD4+CD25+ or CD8+ T reg cells, their CD8+ T cells are not subject to T reg cell–mediated tolerance and are free to initiate and propagate autoimmune diseases. Therefore, both MHC class II–restricted CD4+ T cells and MHC class Ia–restricted CD8+ T cells are required to suppress autoimmunity.

In summary, innate-like CD8+ T cells are primed and activated by self-antigens in the thymus and transported to the periphery, where they can expand exponentially and induce autoimmune diseases. These cells can induce inflammation in multiple organs. Therefore, innate-like CD8+ T cells may contribute to the generation of a variety of autoimmune diseases. These innate-like CD8+ T cells are restricted by MHC class I–like molecules, which are nonpolymorphic in nature and thus present a limited set of self and/or foreign antigens to T cells. Uncontrolled reactivity of innate-like CD8+ T cells to self or foreign antigens may play a critical role in many autoimmune diseases.

Materials And Methods

Mice and antibodies.

CIITA and β2m knockout mice on the C57BL/6 (H-2b) background were purchased from the Jackson Laboratory. Mice with phenotypes RAG-2−/−, OT-I in RAG-2−/− background, OT-II in RAG-2−/− background, and Kb−/−Db−/− double knockout were obtained from Taconic Farms. Kb−/−Db−/− double knockout mice were described previously (25). Kb−/−Db−/−CIITA−/− triple knockout and β2m−/−CIITA−/− double knockout mice were generated by breeding CIITA−/− with Kb−/− Db−/− mice or CIITA−/− with β2m−/− mice, respectively. 8–16-wk-old mice of either sex were used. Mice were housed in a specific pathogen-free colony at the Vivarium of the Robert Wood Johnson Medical School. The animal protocols for the experiments described in this paper were approved by the Institutional Animal Care and Use Committee of the Robert Wood Johnson Medical School.

Fluorochrome-conjugated antibodies against H-2Kb/Db (28-8-6), I-Ab (25-9-17), CD3 (145.2C11), TCRαβ (H-57), CD8α (53-6.7), CD4 (GK1.5), pan-NK marker CD49b (DX5), CD44 (IM7), β7 integrin (M293), CD122 (TM-β1), and Ly6C (AL-21); anti-IgG, anti-BrdU–FITC, biotin-conjugated anti–Qa-1b (6A8.6F10.1A), and anti–Qa-2 (1-1-2); and purified anti-H2-M3 (H2-M3.130) antibodies were purchased from BD Biosciences. Anti-NKG2C FITC (20d5), anti-CD11b–FITC, and anti–IL-12–PE (C17.8) for intracellular staining was purchased from eBioscience.

FACS staining and analysis.

Cells were harvested from the spleen, lymph nodes, thymus, or peripheral blood. Red blood cells were lysed by the ACK lysis buffer (Biosource International). Cells were suspended in staining buffer (PBS, 3% FCS, 0.01% Na-azide) at a concentration of 107 cells/ml. 100 μl of suspension was incubated either with directly conjugated antibodies or biotinylated antibodies for 30 min on ice. Streptavidin-PE was used as the secondary reagent with an additional incubation of 30 min on ice. Cells were washed twice with the staining buffer and fixed with 1% paraformaldehyde. Fluorescence intensity was measured by flow cytometry on a flow cytometer (FACStar; BD Immunocytometry).

T cell adoptive transfer.

CD4+ or CD8+ T cells from OT-II or OT-I TCR Tg mice and from Kb−/−Db−/−CIITA−/− mice, respectively, were purified using Dynal Beads (Dynal), according to the supplier's instructions. For T cell survival experiments, cells were stained with CFSE, and 2 × 106 cells were injected into syngeneic RAG-2−/− or Kb−/−Db−/−CIITA−/− triple knockout mice. After 7 d, cells from spleen and lymph nodes were harvested, and the presence of CFSE-labeled cells was estimated on a flow cytometer. RAG-2−/−pfn−/− mice were also adoptively transferred with CD8+ T cells isolated from Kb−/−Db−/−CIITA−/− mice and analyzed for the development of autoimmunity.

T cell activation and cytokine detection.

Purified CD8 T cells from different strains of mice were activated with plate-bound anti-CD3 and anti-CD28 antibodies for 16 h. Cytokine and chemokine levels in culture supernatants were determined by multiplexed bead array immunoassay using Luminex Technology (Bio-Plex; Bio-Rad Laboratories). Unpurified splenocytes were activated with anti-CD3 antibodies for 18 h to examine IL-12 production. IL-12 was assayed by intracellular staining after treatment with 10 μg/ml brefeldin A during the last 6 h of culture. Cells were stained for surface markers, fixed with 1% paraformaldehyde for 30 min, washed, and resuspended in a permeabilization buffer provided in the Cytofix/Cytoperm kit (BD Biosciences), then stained with PE-conjugated rat antimurine IL-12. Stained cells were then analyzed by flow cytometry on a cytometer (FACScan; Becton Dickinson).

In vivo T cell proliferation assay by BrdU.

Mice were treated with BrdU in drinking water at 0.8 mg/ml (Sigma-Aldrich) for 9 d for assessing in vivo cell proliferation. Spleen and thymus were harvested, and single-cell suspensions were made. Cells were stained for surface markers using appropriate antibodies and fluorochromes. After washing, cells were resuspended in 0.15 M NaCl and fixed by gradual addition of ice-cold 95% ethanol. Cells were washed and incubated with 1% paraformaldehyde for 30 min at room temperature. Cells were stained with anti-BrdU–FITC in the presence of DNase I. Samples were analyzed using a FACScan flow cytometer.

Gross examination of mice and histology.

Photographs of spleens, lymph nodes, or open abdominal cavities from 12-wk-old mice were taken using a digital camera (DSC V-3; Sony). For histology, pieces of intestine and pancreas were washed thoroughly in PBS and fixed in periodate-lysine-paraformaldehyde. Tissues were embedded in paraffin, and 5- to 6-μm sections were cut. Sections were stained with hematoxylin and eosin (H&E) and examined microscopically.

Online supplemental material.

Fig. S1 shows the importance of IFN-γ in anti-CD3–induced IL-12 production in splenocytes of Kb−/−Db−/− CIITA−/− mice. Table S1 demonstrates thymocyte phenotypes in different strains of mice. Table S2 reveals TCR Vβ usage in CD8+ T cells of different strains of mice.

Acknowledgments

The authors would like to thank Dr. Sidney Pestka and Arthur Roberts for critical review of this manuscript.

This work was supported in part by United States Public Health Service grants (AI43384 and AI50222), a National Space Biomedical Research Institute grant (IIH00208), which is supported by the National Aeronautics and Space Administration through Cooperative Agreement NCC 9-58, and the University Professor Fund of the University of Medicine and Dentistry of New Jersey (to Y.F. Shi).

The authors have no conflicting financial interests.

References

1
Bach, J.F., A. Bendelac, M.B. Brenner, H. Cantor, G. De Libero, M. Kronenberg, L.L. Lanier, D.H. Raulet, M.J. Shlomchik, and M.G. von Herrath.
2004
. The role of innate immunity in autoimmunity.
J. Exp. Med.
200
:
1527
–1531.
2
Das, G., S. Sheridan, and C.A. Janeway Jr.
2001
. The source of early IFN-gamma that plays a role in Th1 priming.
J. Immunol.
167
:
2004
– 2010.
3
Natarajan, K., H. Li, R.A. Mariuzza, and D.H. Margulies.
1999
. MHC class I molecules, structure and function.
Rev. Immunogenet.
1
:
32
–46.
4
Soloski, M.J., and E.S. Metcalf.
2001
. The involvement of class Ib molecules in the host response to infection with Salmonella and its relevance to autoimmunity.
Microbes Infect.
3
:
1249
–1259.
5
Taniguchi, M., and T. Nakayama.
2000
. Recognition and function of Valpha14 NKT cells.
Semin. Immunol.
12
:
543
–550.
6
Heller, F., I.J. Fuss, E.E. Nieuwenhuis, R.S. Blumberg, and W. Strober.
2002
. Oxazolone colitis, a Th2 colitis model resembling ulcerative colitis, is mediated by IL-13-producing NK-T cells.
Immunity.
17
:
629
–638.
7
Akbari, O., P. Stock, E. Meyer, M. Kronenberg, S. Sidobre, T. Nakayama, M. Taniguchi, M.J. Grusby, R.H. DeKruyff, and D.T. Umetsu.
2003
. Essential role of NKT cells producing IL-4 and IL-13 in the development of allergen-induced airway hyperreactivity.
Nat. Med.
9
:
582
–588.
8
Urdahl, K.B., J.C. Sun, and M.J. Bevan.
2002
. Positive selection of MHC class Ib-restricted CD8(+) T cells on hematopoietic cells.
Nat. Immunol.
3
:
772
–779.
9
Hu, D., K. Ikizawa, L. Lu, M.E. Sanchirico, M.L. Shinohara, and H. Cantor.
2004
. Analysis of regulatory CD8 T cells in Qa-1-deficient mice.
Nat. Immunol.
5
:
516
–523.
10
Jiang, H., and L. Chess.
2000
. The specific regulation of immune responses by CD8+ T cells restricted by the MHC class Ib molecule, Qa-1.
Annu. Rev. Immunol.
18
:
185
–216.
11
Bouwer, H.G., A. Bai, J. Forman, S.H. Gregory, E.J. Wing, R.A. Barry, and D.J. Hinrichs.
1998
. Listeria monocytogenes-infected hepatocytes are targets of major histocompatibility complex class Ib-restricted antilisterial cytotoxic T lymphocytes.
Infect. Immun.
66
:
2814
–2817.
12
Lewinsohn, D.M., A.L. Briden, S.G. Reed, K.H. Grabstein, and M.R. Alderson.
2000
. Mycobacterium tuberculosis-reactive CD8+ T lymphocytes: the relative contribution of classical versus nonclassical HLA restriction.
J. Immunol.
165
:
925
–930.
13
Lo, W.F., H. Ong, E.S. Metcalf, and M.J. Soloski.
1999
. T cell responses to Gram-negative intracellular bacterial pathogens: a role for CD8+ T cells in immunity to Salmonella infection and the involvement of MHC class Ib molecules.
J. Immunol.
162
:
5398
–5406.
14
Pamer, E.G., C.R. Wang, L. Flaherty, K.F. Lindahl, and M.J. Bevan.
1992
. H-2M3 presents a Listeria monocytogenes peptide to cytotoxic T lymphocytes.
Cell.
70
:
215
–223.
15
Chiang, E.Y., and I. Stroynowski.
2004
. A nonclassical MHC class I molecule restricts CTL-mediated rejection of a syngeneic melanoma tumor.
J. Immunol.
173
:
4394
–4401.
16
Grusby, M.J., and L.H. Glimcher.
1995
. Immune responses in MHC class II-deficient mice.
Annu. Rev. Immunol.
13
:
417
–435.
17
Mombaerts, P., E. Mizoguchi, M.J. Grusby, L.H. Glimcher, A.K. Bhan, and S. Tonegawa.
1993
. Spontaneous development of inflammatory bowel disease in T cell receptor mutant mice.
Cell.
75
:
274
–282.
18
Chang, C.H., S. Guerder, S.C. Hong, W. van Ewijk, and R.A. Flavell.
1996
. Mice lacking the MHC class II transactivator (CIITA) show tissue-specific impairment of MHC class II expression.
Immunity.
4
:
167
–178.
19
Hodgkinson, A.D., B.A. Millward, and A.G. Demaine.
2000
. The HLA-E locus is associated with age at onset and susceptibility to type 1 diabetes mellitus.
Hum. Immunol.
61
:
290
–295.
20
Chun, T., C.J. Aldrich, M.E. Baldeon, L.V. Kawczynski, M.J. Soloski, and H.R. Gaskins.
1998
. Constitutive and regulated expression of the class IB molecule Qa-1 in pancreatic beta cells.
Immunology.
94
:
64
–71.
21
Su, J., R.E. Berg, S. Murray, and J. Forman.
2005
. Thymus-dependent memory phenotype CD8 T cells in naive B6.H-2Kb−/−Db−/− animals mediate an antigen-specific response against Listeria monocytogenes.
J. Immunol.
175
:
6450
–6457.
22
Seaman, M.S., C.R. Wang, and J. Forman.
2000
. MHC class Ib-restricted CTL provide protection against primary and secondary Listeria monocytogenes infection.
J. Immunol.
165
:
5192
–5201.
23
Meresse, B., S.A. Curran, C. Ciszewski, G. Orbelyan, M. Setty, G. Bhagat, L. Lee, M. Tretiakova, C. Semrad, E. Kistner, et al.
2006
. Reprogramming of CTLs into natural killer–like cells in celiac disease.
J. Exp. Med.
203
:
1343
–1355.
24
Lo, W.F., A.S. Woods, A. DeCloux, R.J. Cotter, E.S. Metcalf, and M.J. Soloski.
2000
. Molecular mimicry mediated by MHC class Ib molecules after infection with gram-negative pathogens.
Nat. Med.
6
:
215
–218.
25
Perarnau, B., M.F. Saron, B.R. San Martin, N. Bervas, H. Ong, M.J. Soloski, A.G. Smith, J.M. Ure, J.E. Gairin, and F.A. Lemonnier.
1999
. Single H2Kb, H2Db and double H2KbDb knockout mice: peripheral CD8+ T cell repertoire and anti-lymphocytic choriomeningitis virus cytolytic responses.
Eur. J. Immunol.
29
:
1243
–1252.

Abbreviations used: CFSE, carboxyfluorescein diacetate succinimidyl ester; H&E, hematoxylin and eosin; HLA, histocompatibility leukocyte antigen; IBD, inflammatory bowel disease; ILL, innate-like lymphocyte; Tg, transgenic.