The natural killer T-cell ligand alpha-galactosylceramide prevents autoimmune diabetes in non-obese diabetic mice.

Diabetes in non-obese diabetic (NOD) mice is mediated by pathogenic T-helper type 1 (Th1) cells that arise because of a deficiency in regulatory or suppressor T cells. V alpha 14-J alpha 15 natural killer T (NKT) cells recognize lipid antigens presented by the major histocompatibility complex class I-like protein CD1d (refs. 3,4). We have previously shown that in vivo activation of V alpha 14 NKT cells by alpha-galactosylceramide (alpha-GalCer) and CD1d potentiates Th2-mediated adaptive immune responses. Here we show that alpha-GalCer prevents development of diabetes in wild-type but not CD1d-deficient NOD mice. Disease prevention correlated with the ability of alpha-GalCer to suppress interferon-gamma but not interleukin-4 production by NKT cells, to increase serum immunoglobulin E levels, and to promote the generation of islet autoantigen-specific Th2 cells. Because alpha-GalCer recognition by NKT cells is conserved among mice and humans, these findings indicate that alpha-GalCer might be useful for therapeutic intervention in human diseases characterized by Th1-mediated pathology such as Type 1 diabetes.

Activation of natural killer T cells by alpha-galactosylceramide treatment prevents the onset and recurrence of autoimmune Type 1 diabetes.

Type 1 diabetes (T1D) in non-obese diabetic (NOD) mice may be favored by immune dysregulation leading to the hyporesponsiveness of regulatory T cells and activation of effector T-helper type 1 (Th1) cells. The immunoregulatory activity of natural killer T (NKT) cells is well documented, and both interleukin (IL)-4 and IL-10 secreted by NKT cells have important roles in mediating this activity. NKT cells are less frequent and display deficient IL-4 responses in both NOD mice and individuals at risk for T1D (ref. 8), and this deficiency may lead to T1D (refs. 1,6-9). Thus, given that NKT cells respond to the alpha-galactosylceramide (alpha-GalCer) glycolipid in a CD1d-restricted manner by secretion of Th2 cytokines, we reasoned that activation of NKT cells by alpha-GalCer might prevent the onset and/or recurrence of T1D. Here we show that alpha-GalCer treatment, even when initiated after the onset of insulitis, protects female NOD mice from T1D and prolongs the survival of pancreatic islets transplanted into newly diabetic NOD mice. In addition, when administered after the onset of insulitis, alpha-GalCer and IL-7 displayed synergistic effects, possibly via the ability of IL-7 to render NKT cells fully responsive to alpha-GalCer. Protection from T1D by alpha-GalCer was associated with the suppression of both T- and B-cell autoimmunity to islet beta cells and with a polarized Th2-like response in spleen and pancreas of these mice. These findings raise the possibility that alpha-GalCer treatment might be used therapeutically to prevent the onset and recurrence of human T1D.

Binding and antigen presentation of ceramide-containing glycolipids by soluble mouse and human CD1d molecules.

We have purified soluble mouse and human CD1d molecules to assess the structural requirements for lipid antigen presentation by CD1. Plate-bound CD1d molecules from either species can present the glycolipid alpha-galactosyl ceramide (alpha-GalCer) to mouse natural killer T cells, formally demonstrating both the in vitro formation of antigenic complexes, and the presentation of alpha-GalCer by these two CD1d molecules. Using surface plasmon resonance, we show that at neutral pH, mouse CD1 and human CD1d bind to immobilized alpha-GalCer, unlike human CD1b, which requires acidic pH for lipid antigen binding. The CD1d molecules can also bind both to the nonantigenic beta-GalCer and to phosphatidylethanolamine, indicating that diverse lipids can bind to CD1d. These studies provide the first quantitative analysis of monomeric lipid antigen-CD1 interactions, and they demonstrate that the orientation of the galactose, or even the nature of the polar head group, are likely to be more important for T cell receptor contact than CD1d binding.

Tracking the response of natural killer T cells to a glycolipid antigen using CD1d tetramers.

A major group of natural killer (NK) T cells express an invariant Valpha14(+) T cell receptor (TCR) specific for the lipoglycan alpha-galactosylceramide (alpha-GalCer), which is presented by CD1d. These cells may have an important immune regulatory function, but an understanding of their biology has been hampered by the lack of suitable reagents for tracking them in vivo. Here we show that tetramers of mouse CD1d loaded with alpha-GalCer are a sensitive and highly specific reagent for identifying Valpha14(+) NK T cells. Using these tetramers, we find that alpha-GalCer-specific T lymphocytes are more widely distributed than was previously appreciated, with populations of largely NK1.1(-) but tetramer-binding T cells present in the lymph nodes and the intestine. Injection of alpha-GalCer leads to the production of both interferon gamma and interleukin 4 by nearly all NK T cells in the liver and the majority of the spleen within 2 h. These cells mostly disappear by 5 h, and they do not reappear after 1 wk. Curiously, tetramer-positive thymocytes do not rapidly synthesize cytokines, nor do they undergo decreases in cell number after lipid antigen stimulation, although they express equivalent TCR levels. In summary, the data presented here demonstrate that alpha-GalCer-specific NK T cells undergo a unique and highly compartmentalized response to antigenic stimulation.

T cell responses modulated through interaction between CD8alphaalpha and the nonclassical MHC class I molecule, TL.

The thymus leukemia antigen (TL) is a nonclassical class I molecule, expressed abundantly on intestinal epithelial cells. We show that, in contrast to other major histocompatibility complex (MHC) class I molecules that bind CD8alphabeta, TL preferentially binds the homotypic form of CD8alpha (CD8alphaalpha). Thus, TL tetramers react specifically to CD8alphaalpha-expressing cells, including most intestinal intraepithelial lymphocytes. Compared with CD8alphabeta, which recognizes the same MHC as the T cell receptor (TCR) and thus acts as a TCR coreceptor, high-affinity binding of CD8alphaalpha to TL modifies responses mediated by TCR recognition of antigen presented by distinct MHC molecules. These findings define a novel mechanism of lymphocyte regulation through CD8alphaalpha and MHC class I.

CD1-mediated antigen presentation of glycosphingolipids.

CD1 proteins are distinguished by their ability to present lipid antigens to T cells. Group II CD1 or CD1d molecules are recognized by the specialized NK T-cell subset, and this reactivity can be greatly augmented by alpha-galactosylceramide, a glycosphingolipid derived from a marine sponge. Human CD1b, which is only distantly related to the CD1d molecules, can present mammalian glycosphingolipids (gangliosides) to autoreactive T-cell clones derived from multiple sclerosis patients. Thus, CD1 responsive and glycosphingolipid-reactive cells may play an important immune regulatory role, in addition to their well-characterized role in the response to microbial lipids.

Syntheses of biotinylated alpha-galactosylceramides and their effects on the immune system and CD1 molecules.

A representative alpha-galactosylceramide (alpha-GalCer), KRN7000, has strong immunostimulatory and antitumor activity. Recent studies demonstrated that KRN7000-pulsed antigen-presenting cells (APC) can activate natural killer T (NKT) cells, a novel T-cell lineage, and CD1d molecules on APC play an important role in the activation of NKT cells. However, it remains unclear whether alpha-GalCers actually bind to CD1d molecules. To address this question, we synthesized three kinds of biotinylated alpha-GalCer and a biotinylated beta-GalCer and found that the biotinylated alpha-GalCers significantly stimulate the proliferation of murine spleen cells, but the biotinylated beta-GalCer does not and that all biotinylated compounds bind to CD1d molecules.

Natural killer T cells reactive to a single glycolipid exhibit a highly diverse T cell receptor beta repertoire and small clone size.

CD1d-restricted natural killer (NK) T cells reactive with the glycolipid alpha-galactosylceramide (alpha-GalCer) are a distinct lymphocyte sublineage. They express an invariant Valpha14-Jalpha18 T cell receptor (TcR), but the role of the beta chain has been controversial. Here, we have used CD1d tetramers to identify and isolate NK T cells based on their antigen specificity. In mice lacking germline Vbeta8, most of the alpha-GalCer-reactive T cells express either Vbeta2 or Vbeta7, strong Vbeta selection being revealed by the lack of an increase in other Vbeta regions. By contrast to the selection for complementarity determining region (CDR) 3beta sequences in some anti-peptide responses, alpha-GalCer-reactive T cells have polyclonal CDR3beta sequences. There is little CDR3beta sequence redundancy between organs or individual mice, and, surprisingly, there also is no evidence for organ-specific CDR3beta sequence motifs. These data argue against a T cell receptor-mediated self-reactivity for tissue-specific CD1d-bound ligands. Each NKT clone is represented by only 5-10 cells. This clone size is similar to naive conventional T cells, and much lower than that reported for memory T cells, although NK T cells have an activated/memory phenotype.

Human NKT cells mediate antitumor cytotoxicity directly by recognizing target cell CD1d with bound ligand or indirectly by producing IL-2 to activate NK cells.

alpha-Galactosylceramide (alphaGalCer) stimulates NKT cells and has antitumor activity in mice. Murine NKT cells may directly kill tumor cells and induce NK cell cytotoxicity, but the mechanisms are not well defined. Newly developed human CD1d/alphaGalCer tetrameric complexes were used to obtain highly purified human alphaGalCer-reactive NKT cell lines (>99%), and the mechanisms of NKT cell cytotoxicity and activation of NK cells were investigated. Human NKT cells were cytotoxic against CD1d(-) neuroblastoma cells only when they were rendered CD1d(+) by transfection and pulsed with alphaGalCer. Four other CD1d(-) tumor cell lines of diverse origin were resistant to NKT cells, whereas Jurkat and U937 leukemia cell lines, which are constitutively CD1d(+), were killed. Killing of the latter was greatly augmented in the presence of alphaGalCer. Upon human CD1d/alphaGalCer recognition, NKT cells induced potent cytotoxicity of NK cells against CD1d(-) neuroblastoma cell lines that were not killed directly by NKT cells. NK cell activation depended upon NKT cell production of IL-2, and was enhanced by secretion of IFN-gamma. These data demonstrate that cytotoxicity of human NKT cells can be CD1d and ligand dependent, and that TCR-stimulated NKT cells produce IL-2 that is required to induce NK cell cytotoxicity. Thus, NKT cells can mediate potent antitumor activity both directly by targeting CD1d and indirectly by activating NK cells.

CD1d-restricted NKT cells: an interstrain comparison.

CD1d-restricted Valpha14-Jalpha281 invariant alphabetaTCR(+) (NKT) cells are well defined in the C57BL/6 mouse strain, but they remain poorly characterized in non-NK1.1-expressing strains. Surrogate markers for NKT cells such as alphabetaTCR(+)CD4(-)CD8(-) and DX5(+)CD3(+) have been used in many studies, although their effectiveness in defining this lineage remains to be verified. Here, we compare NKT cells among C57BL/6, NK1.1-congenic BALB/c, and NK1.1-congenic nonobese diabetic mice. NKT cells were identified and compared using a range of approaches: NK1.1 expression, surrogate phenotypes used in previous studies, labeling with CD1d/alpha-galactosylceramide tetramers, and cytokine production. Our results demonstrate that NKT cells and their CD4/CD8-defined subsets are present in all three strains, and confirm that nonobese diabetic mice have a numerical and functional deficiency in these cells. We also highlight the hazards of using surrogate phenotypes, none of which accurately identify NKT cells, and one in particular (DX5(+)CD3(+)) actually excludes these cells. Finally, our results support the concept that NK1.1 expression may not be an ideal marker for CD1d-restricted NKT cells, many of which are NK1.1-negative, especially within the CD4(+) subset and particularly in NK1.1-congenic BALB/c mice.

Cytometric and functional analyses of NK and NKT cell deficiencies in NOD mice.

Defects in NK and NKT cell activities have been implicated in the etiology of type 1 (autoimmune) diabetes in NOD mice on the basis of experiments performed using surrogate phenotypes for the identification of these lymphocyte subsets. Here, we have generated a congenic line of NOD mice (NOD.b-Nkrp1(b)) which express the allelic NK1.1 marker, enabling the direct study of NK and NKT cells in NOD mice. Major deficiencies in both populations were identified when NOD.b-Nkrp1(b) mice were compared with C57BL/6 and BALB.B6-Cmv1(r) mice by flow cytometry. The decrease in numbers of peripheral NK cells was associated with an increase in their numbers in the bone marrow, suggesting that a defect in NK cell export may be involved. In contrast, the most severe deficiency of NKT cells found was in the thymus, indicating that defects in thymic production were probably responsible. The deficiencies in NK cell activity in NOD mice could only partly be accounted for by the reduced numbers of NK cells, and fewer NKT cells from NOD mice produced IL-4 following stimulation, suggesting that NK and NKT cells from NOD mice shared functional deficiencies in addition to their numerical deficiencies. Despite the relative lack of IL-4 production by NOD NKT cells, adoptive transfer of alpha beta TCR(+)NK1.1(+) syngeneic NKT cells into 3-week-old NOD recipients successfully prevented the onset of spontaneous diabetes. As both NK and NKT cells play roles in regulating immune responses, we postulate that the synergistic defects reported here contribute to the susceptibility of NOD mice to autoimmune disease.

Membrane lymphotoxin is required for the development of different subpopulations of NK T cells.

The development of lymphoid organs requires membrane-bound lymphotoxin (LT), a heterotrimer containing LTalpha and LTbeta, but the effects of LT on T cell function have not been characterized extensively. Upon TCR cross-linking in vitro, splenocytes from both LTalpha-/- and LTbeta-/- mice failed to produce IL-4 and IL-10 due to a reduction in NK T cells. Concordantly, LTalpha-/- and LTbeta-/- mice did not respond to the lipoglycan alpha-galactosylceramide, which is presented by mouse CD1 to Valpha14+ NK T cells. Interestingly, both populations of NK T cells, including those that are mouse CD1 dependent and alpha-galactosylceramide reactive and those that are not, were affected by disruption of the LTalpha and LTbeta genes. NK T cells were not affected, however, in transgenic mice in which LT signaling is blocked, beginning on day 3 after birth, by expression of a soluble decoy LTbeta receptor. This suggests that membrane-bound LT is critical for NK T cells early in ontogeny, but not for the homeostasis of mature cells.