NKT cells stimulated by long fatty acyl chain sulfatides significantly reduce the incidence of type 1 diabetes in nonobese diabetic mice [corrected]

Abstract

Sulfatide-reactive type II NKT cells have been shown to regulate autoimmunity and anti-tumor immunity. Although, two major isoforms of sulfatide, C16:0 and C24:0, are enriched in the pancreas, their relative role in autoimmune diabetes is not known. Here, we report that sulfatide/CD1d-tetramer(+) cells accumulate in the draining pancreatic lymph nodes, and that treatment of NOD mice with sulfatide or C24:0 was more efficient than C16:0 in stimulating the NKT cell-mediated transfer of a delay in onset from T1D into NOD.Scid recipients. Using NOD.CD1d(-/-) mice, we show that this delay of T1D is CD1d-dependent. Interestingly, the latter delay or protection from T1D is associated with the enhanced secretion of IL-10 rather than IFN-g by C24:0-treated CD4(+) T cells and the deviation of the islet-reactive diabetogenic T cell response. Both C16:0 and C24:0 sulfatide isoforms are unable to activate and expand type I iNKT cells. Collectively, these data suggest that C24:0 stimulated type II NKT cells may regulate protection from T1D by activating DCs to secrete IL-10 and suppress the activation and expansion of type I iNKT cells and diabetogenic T cells. Our results raise the possibility that C24:0 may be used therapeutically to delay the onset and protect from T1D in humans.

Figure 1. Structure-function analyses of sulfatides.

(A) Structures of the C16:0, C24:0 and C24:1 sulfatide isoforms. (B) Treatment with sulfatide reduces the incidence of spontaneous T1D in NOD mice. Female NOD wild type (CD1d^+/+) mice (12 week-old, n = 10–12/group) were injected once weekly for 3 weeks with 20 mg of either sulfatide, GM1 or PBS-vehicle until 15 weeks of age. The P value between the values in the control (PBS/vehicle or mGM1) versus sulfatide group was <0.0001. The occurrence of spontaneous T1D was followed for up to 30–34 weeks of age by measuring BGL. Two consecutive BGL readings of >250 mg/dl was considered diabetic. These data are representative of 4 independent experiments. (C) Treatment of NOD mice with C24:0 but not C16:0 sulfatide reduces the lymphocyte-mediated adoptive transfer of T1D. Female NOD mice (3–5 week-old, n = 10/group) were injected i.p. with either control vehicle (PBS), aGalCer (4 mg/dose) or sulfatide (C16:0 or C24:0, 100 mg/dose) on day 0 and day 4. Pooled splenocytes and PLN lymphocytes (2×10⁷) were transferred to female NOD.Scid recipients, and the recipient mice were monitored until 24 weeks of age for the development of T1D by determining their BGL. Results shown are representative of 3 independent and reproducible experiments.

Figure 2. C24:0 sulfatide-induced protection from the adoptive transfer of T1D is CD1d-dependent.

Female NOD.CD1d^−/− mice (3–5 week-old, n = 10/group) were injected i.p. with either control vehicle (PBS) or C24:0 sulfatide (100 mg/dose) on day 0 and day 4. Pooled splenocytes and PLN lymphocytes (2×10⁷) were transferred to female NOD.Scid recipients, and the recipient mice were monitored until 32 weeks of age for the development of T1D by determining their BGL. Two consecutive BGL readings of >250 mg/dl was considered diabetic. Data shown represent one of two representative and reproducible experiments.

Figure 3. Analyses of NOD T cell proliferative responses stimulated by sulfatides.

(A) Ability of C24:0 sulfatide to protect from T1D does not correlate with its capacity to stimulate a CD4⁺ T cell proliferative response. Spleen CD4⁺ T cells from female NOD mice (3–5 week-old) were co-cultured for 72 h in vitro in the presence of mitomycin C treated CD4^– cells at CD4⁺:CD4^– ratios of 1∶1, 10∶1 or 100∶1 with either control vehicle, aGalCer (100 ng/ml) or varying concentrations (5–50 mg/ml) of sulfatide (C16:0 or C24:0). [³H]-thymidine (1 mCi/well) was added for the final 18 h of culture before harvesting the cells and quantitating their thymidine incorporation. Data shown were obtained from one of two representative and reproducible experiments. (B) Proliferative responses of splenocytes from non-diabetic and diabetic NOD mice (2–4 mice/group) in response to in vitro stimulation by sulfatide or aGalCer. Splenocytes (8×10⁵) were incubated in a 96-well plate with a titrated concentration of sulfatide or aGalCer, and [³H]-thymidine incorporation in triplicate wells was determined following culture for 90 h, as previously described . No detectable response was found to another control glycolipid mGM1 in these assays. Data from one of three representative experiments are shown.

Figure 4. Analyses of NOD T cell cytokine secretion responses induced by sulfatides.

(A, B) C16:0 and C24:0 sulfatide stimulate different CD4⁺ T cell cytokine secretion profiles. NOD spleen CD4⁺ T cells were co-cultured for 72 h with mitomycin treated CD4^– cells at a CD4⁺:CD4^– ratio of 100∶1 in the presence of control vehicle or sulfatide (C16:0 or C24:0, 50 mg/ml), as described in Fig. 3A. The concentration of IFN-g (A) and IL-10 (B) secreted into cell supernatants were analyzed by ELISA. IL-2 and IL-4 were not detected in these supernatants. Data shown were obtained from one of three representative and reproducible experiments. (C, D) Treatment with sulfatide inhibits the induced anti-islet diabetogenic T cell cytokine responses. Female NOD mice (4 week-old, 4 mice/group) were injected i.p. with sulfatide (20 mg/mouse) or vehicle/PBS. One week later, the treated mice were immunized with 100 mg of either the insulin p9–23, GAD206-220, GAD524-543 or hsp277 peptide. Ten days following antigenic challenge, PLN lymphocytes were assayed for their proliferative ([³H]-thymidine incorporation) and cytokine (IFN-g, IL-4) secretion responses (Elispot assay). The average frequencies of IFN-γ (C) or IL-4 (D) secreting cells reactive to different islet antigens are shown. In comparison to control vehicle values, statistically significant reductions in the frequencies of cytokine-secreting islet antigen-reactive T cells were found for the insulin and GAD peptides but not Hsp peptide, as follows. Insulin 9–23 (P = 0.0001 for IFN-g, P = 0.0078 for IL-4); GAD 206-220 (P = 0.012 for IFN-g, P = 0.0018 for IL-4); GAD 524-543 (P<0.0001 for IFN-g, P = 0.0328 for IL-4); and Hsp 277 (Not significant, P = 0.0542 for IFN-g; Not significant, P = 0.0952 for IL-4).

Figure 5. Analyses of sulfatide-induced stimulation of the activation and expansion of type I iNKT cells.

(A) C16:0 and C24:0 do not stimulate the activation of type I iNKT cells. At 2 h after female NOD mice (3–5 week-old, n = 5/group) were injected (i.p.) with vehicle (PBS), aGalCer (5 mg) or sulfatide (C16:0 or C24:0, 100 mg), their splenocytes were analyzed by FACS for the surface expression of the early activation surface marker CD69 on TCRβ⁺a-GalCer/CD1dTet⁺ iNKT cells. (B) C16:0 and C24:0 do not stimulate the expansion of type I iNKT cells. At 2 h, 4 h and 12 h after female NOD mice (3–5 week-old, n = 5/group) were injected (i.p.) with either control vehicle (PBS), aGalCer (4 mg) or sulfatide (C16:0 or C24:0, 100 mg), splenocytes were analyzed by FACS for the expansion of TCRb⁺a-GalCer/CD1dTet⁺ iNKT cells. Data shown in (A) and (B) were obtained from one of two representative and reproducible experiments.

Figure 6. Sulfatide-reactive Type II NKT cells are enriched in the PLN during the development of T1D.

PLN lymphocytes isolated from the pancreas of female NOD mice (6–20 week-old) were stained with sulfatide/CD1d-, aGalCer/CD1d- or unfilled PBS/CD1d-tetramers and anti-TCRb and then analyzed by flow cytometry. (A) The numbers shown in each panel indicate the % tetramer⁺ cells. (B) The scatter plots show the mean values of sulfatide/CD1d-tetramer⁺ (0.17%) and αGalCer/CD1d-tetramer⁺ (0.37%) cells in diabetic NOD mice (6–20 week-old). Note that we were not able to detect any correlation between NKT cell numbers and the severity of T1D, as we did not find any significant correlation between the frequency of sulfatide-CD1d-tetramer⁺ cells in diabetic vs. non-diabetic NOD mice.