α-Galactosylceramide treatment before allergen sensitization promotes iNKT cell–mediated induction of Treg cells, preventing Th2 cell responses in murine asthma

Asthma is a common inflammatory pulmonary disorder involving a diverse array of immune cells such as proinflammatory T helper 2 (Th2) cells. We recently reported that intraperitoneal injection of α-galactosylceramide (α-GalCer) can stimulate the lung invariant natural killer T (iNKT) cells and does not lead to airway inflammation in WT mice. Other studies indicate that iNKT cells play an important role in inducing regulatory T cells (Treg cells) and peripheral tolerance. Using iNKT cell– knockout mice, functional inactivation of Treg cells, and co-culture experiments in murine asthma models, we investigated the immunoregulatory effects of α-GalCer treatment before allergen sensitization on Th2 cell responses. We also studied whether α-GalCer's effects require lung Treg cells induced by activated iNKT cells. Our results disclosed that intraperitoneal administration of α-GalCer before allergen sensitization could promote the expansion and suppressive activity of lung CD4+FoxP3+ Treg cells. These effects were accompanied by down-regulated Th2 cell responses and decreased immunogenic maturation of lung dendritic cells in WT mice. However, these changes were absent in CD1d−/− mice immunized and challenged with ovalbumin or house dust mites, indicating that the effects of α-GalCer on Treg cells mainly require iNKT cells. Moreover, functional inactivation of Treg cells could reverse the inhibitory ability of this α-GalCer therapy on Th2 cell responses in a murine asthma model. Our findings indicate that intraperitoneal administration of α-GalCer before the development of asthma symptoms induces the generation of lung Treg cells via iNKT cells and may provide a potential therapeutic strategy to prevent allergic asthma.

Bronchial asthma is a common chronic inflammatory disorder of the respiratory tracts, and a diverse array of cell types are involved in the development of airway inflammation of asthma. The predominant paradigm is that T helper 2 (Th2) 2 cells drive inflammatory responses through the release of Th2 cell cytokines, including IL-4, IL-5, and IL-13, which result in airway inflammation, B cell isotype shifting, mucus overproduction, and airway hyper-responsiveness (AHR) (1). The development of Th2 cells is modulated by regulatory T (Treg) cells, which can maintain homeostasis and inhibit aberrant immunity (2). Previous data have shown that Treg cells have a reduced number and impaired function in patients with asthma, and thereby asthma is a condition of dysregulated immune responses (3,4). Several reports indicated that CD4 ϩ CD25 ϩ Treg cells adoptively transferred into immunized mice before allergen challenge can inhibit the initiation of asthma (5,6), whereas adoptive transfer of CD4 ϩ CD25 ϩ Treg cells after acute exacerbation of asthma can reduce established allergic airway inflammation (7). These studies raise some important evidence that Treg cells can prevent initiation and progression of asthma. As a result, the induction of lung Treg cells may be a potential therapeutic protocol for the treatment of asthma (8).
Invariant natural killer T (iNKT) cells constitute a distinct and relatively minor T lymphocyte subset, which expresses a semi-invariant T-cell receptor (TCR) that can recognize glycolipid antigens presented by the nonpolymorphic major histocompatibility complex-like molecule CD1d (9,10). ␣-Galactosylceramide (␣-GalCer), a derivative from a marine sponge, can specifically and strongly activate iNKT cells to result in the secretion of both Th1 and Th2 cell cytokines, such as IFN-␥ and IL-4, which probably account for their immunomodulatory functions (11)(12)(13). Recent studies have shown that iNKT cells activated by ␣-GalCer can prevent experimental autoimmune disorders such as type 1 diabetes and myasthenia gravis through induction of CD4 ϩ CD25 ϩ Treg cells (14,15). Interestingly, Ronet et al. (16) reported that iNKT cells play a key role in the initiation of the lethal ileitis after Toxoplasma gondii infection, but activation of iNKT cells by intraperitoneal administration of ␣-GalCer before T. gondii infection can enhance the frequency of IL-10 -secreting Treg cells to control the inflamma-tory intestinal disorder. These data suggested that iNKT cells activated by ␣-GalCer play an important role in inducing the development of Treg cells and peripheral tolerance (17). In our previous study, we showed that intraperitoneal injection of ␣-GalCer has the capability to stimulate iNKT cells, but activation of iNKT cells alone does not lead to airway inflammation in wildtype (WT) mice in the absence of ovalbumin (OVA) sensitization and challenge (18). Moreover, our previous data found that immature bone marrow-derived dendritic cells (BMDCs) adopt immunogenic maturation status upon interplay with iNKT cells in the presence of OVA, but immature BMDCs undergo a tolerogenic maturation condition upon interplay with iNKT cells alone without OVA (19). Thus, it was hypothesized that ␣-GalCer treatment before antigen sensitization may prevent Th2 cell responses through the induction of lung Treg cells by iNKT cells in asthmatics.
To elucidate this hypothesis, we have investigated the immunoregulatory effects of ␣-GalCer treatment before allergen sensitization on Th2 cell responses and maturation status of lung dendritic cells (LDCs), and we further analyzed whether the influence of this ␣-GalCer therapy requires lung Treg cells induced by iNKT cells using iNKT cell-knockout mice, specific functional inactivation of Treg cells, and co-culture experiments in murine asthma models. Our findings demonstrated that ␣-GalCer treatment before allergen sensitization promotes iNKT cell-mediated induction of Treg cells, preventing Th2 cell responses in murine asthma models.

␣-GalCer treatment before allergen sensitization alleviates Th2 cell responses in an OVA-or HDM-induced murine asthma model
A recent study has shown that activation of iNKT cells by intraperitoneal administration of ␣-GalCer before T. gondii infection can enhance the frequency of IL-10 -secreting Treg cells to control the inflammatory intestinal disorder, although iNKT cells play a key role in the initiation of the lethal ileitis after T. gondii infection (16). Thus, we sought to determine the influence of ␣-GalCer treatment before allergen sensitization on Th2 cell responses of asthma. As shown in Fig. 1, ␣-GalCer treatment before OVA sensitization markedly reduced inflammatory cell infiltration in the respiratory tract (Fig. 1A), PASpositive goblet cells in the airway epithelium (Fig. 1, A and B), Figure 1. ␣-GalCer treatment before OVA sensitization alleviates Th2 cell responses in an OVA-induced murine asthma model. WT mice, intraperitoneally treated with ␣-GalCer or PBS before OVA sensitization, were sacrificed 24 h after the final OVA challenge. A, histopathologic analysis of lung tissue H&E and PAS staining. B, lung goblet cell hyperplasia is indicated as the number of PAS-positive cells per unit of length (mm) of the basement membrane. C, total and differential cell counting in BALF. Tot, total cell counts; Eos, eosinophils; Mac, macrophages; Neu, neutrophils; and Lym, lymphocytes. D, concentrations of IL-4, IL-5, IL-10, IL-13, and IFN-␥ in BALFs were analyzed by ELISA. E, OVA-specific IgE and IgG1 concentrations in the serum. Data are expressed as the means Ϯ S.D. of three independent experiments (n ϭ 5), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01; NS, not significant. F, airway response to increasing concentrations of methacholine was examined. Significant differences between asthmatic mice administered ␣-GalCer or PBS are shown as *, p Ͻ 0.05, and **, p Ͻ 0.01.

␣-GalCer induces lung Treg cells to prevent asthma
and the number of total cells, eosinophils, and macrophages in the BALF (Fig. 1C), as well as the levels of IL-4, IL-5, and IL-13 in the BALF (Fig. 1D) in the OVA-induced asthmatic mouse, compared with PBS administration (p Ͻ 0.05 or p Ͻ 0.01). However, the level of IL-10 in the BALF was significantly increased in the OVA-induced asthmatic mouse with ␣-GalCer treatment before OVA sensitization, compared with PBS administration (Fig. 1D) (p Ͻ 0.05). Furthermore, the concentrations of OVA-specific IgE and IgG1 in the serum were significantly reduced in OVA-induced asthmatic WT mice treated with ␣-GalCer before OVA sensitization, compared with PBS treatment (Fig. 1E) (p Ͻ 0.01). In addition, AHR was assessed by direct measurement of RL and Cdyn to increasing levels of Mch, ranging from 3.12 to 50 mg/ml in PBS. Our data showed that weakened RL to Mch was observed in OVA-induced asthmatic mice with ␣-GalCer treatment before OVA sensitization, compared with PBS treatment (Fig. 1F) (p Ͻ 0.05 or p Ͻ 0.01). By contrast, no difference was found in Cdyn between these two groups ( Fig. 1F) (p Ͼ 0.05). Collectively, these data suggested that ␣-GalCer treatment before OVA sensitization can alleviate Th2 cell responses in an OVA-induced murine asthma model.
The HDM is one of the most common allergens in the living environment of asthmatic patients (20). Thus, we sought to determine the influence of ␣-GalCer treatment before HDM sensitization on Th2 cell responses of asthma. As shown in Fig. 2, administration of ␣-GalCer before HDM sensitization strongly attenuated inflammatory cell infiltration ( Fig. 2A), mucus-secreting cells in the airways (Fig. 2, A and B), the number of total cells, eosinophils, and macrophages in the BALF (Fig. 2C), the release of cytokines, including IL-4, IL-5, and IL-13 in the BALF (Fig. 2D), and the levels of HDMspecific IgE and IgG1 in the serum (Fig. 2E) (p Ͻ 0.05 or p Ͻ 0.01). However, the level of IL-10 in the BALF was markedly elevated in the HDM-induced asthmatic mice with ␣-GalCer treatment before HDM sensitization, compared with PBS treatment (Fig. 2D) (p Ͻ 0.05). In addition, significantly alleviated RL to Mch was observed in HDM-induced asthmatic mice with ␣-GalCer treatment before HDM sensitization, compared with PBS treatment (Fig. 2F) (p Ͻ 0.05 or p Ͻ 0.01). However, no difference was observed in Cdyn between the two groups ( Fig. 2F) (p Ͼ 0.05). Collectively, these data suggested that ␣-GalCer treatment before HDM sensitiza- Figure 2. ␣-GalCer treatment before HDM sensitization suppresses Th2 cell responses in HDM-induced murine asthma model. WT mice, intraperitoneally treated with ␣-GalCer or PBS before HDM sensitization, were sacrificed 3 days after the final HDM challenge. A, histopathological analysis of lung tissue H&E and PAS staining. B, lung goblet cell hyperplasia is indicated as the number of PAS-positive cells per unit of length (mm) of the basement membrane. C, total and differential cell counting in BALF of mice. Tot, total cell counts; Eos, eosinophils; Mac, macrophages; Neu, neutrophils; and Lym, lymphocytes. D, BALF was collected 24 h after the final challenge, and the concentrations of IL-4, IL-5, IL-10, IL-13, and IFN-␥ were analyzed by ELISA. E, HDM-specific IgE and IgG1 concentrations in the serum. Data are expressed as the means Ϯ S.D. of three independent experiments (n ϭ 5), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01; NS, not significant. F, airway response to increasing concentrations of methacholine was examined. Significant differences between asthmatic mice treated with ␣-GalCer or PBS are shown as *, p Ͻ 0.05, and **, p Ͻ 0.01.

␣-GalCer induces lung Treg cells to prevent asthma
tion can suppress Th2 cell responses in HDM-induced asthmatic mice.

␣-GalCer-mediated inhibition of Th2 cell responses requires iNKT cells in asthmatic mice
␣-GalCer, a glycolipid ligand originally derived from marine sponges, presented by the CD1d molecule, can stimulate iNKT cells and is widely used to determine the role of iNKT cells in multiple diseases, such as asthma, tumors, and autoimmunity (21)(22)(23). We sought to determine whether the inhibition of Th2 cell responses mediated by ␣-GalCer required iNKT cells in asthmatic mice. As outlined in Fig. 3, ␣-GalCer treatment before OVA sensitization markedly reduced inflammatory cell infiltration (Fig. 3A), mucus-secreting cells in the airways (Fig.  3, A and B), the number of total cells, eosinophils, and macrophages in the BALF (Fig. 3C), the release of such cytokines as IL-4, IL-5, and IL-13 in the BALF (Fig. 3D), as well as the levels of OVA-specific IgE and IgG1 in the serum (Fig. 3E) in WT mice sensitized and challenged with OVA as compared with CD1d Ϫ/Ϫ mice sensitized and challenged with OVA (p Ͻ 0.05 or p Ͻ 0.01). However, ␣-GalCer treatment before OVA sensi-tization strongly enhanced the level of IL-10 in the BALF from WT mice sensitized and challenged with OVA, compared with CD1d Ϫ/Ϫ mice sensitized and challenged with OVA ( Fig. 3D) (p Ͻ 0.01). In addition, ␣-GalCer treatment before OVA sensitization markedly attenuated RL to Mch in WT mice sensitized and challenged with OVA, compared with CD1d Ϫ/Ϫ mice sensitized and challenged with OVA ( Fig. 3F) (p Ͻ 0.05 or p Ͻ 0.01). However, no difference was observed in Cdyn between the two groups ( Fig. 3F) (p Ͼ 0.05). Interestingly, no difference was found in airway inflammation and AHR between CD1d Ϫ/Ϫ asthmatic mice treated with ␣-GalCer and PBS (Fig. 3, A-F) (p Ͼ 0.05). Collectively, these data showed that ␣-GalCermediated inhibition of Th2 cell responses requires iNKT cells in asthmatic mice.

␣-GalCer treatment before allergen sensitization downregulates surface maturation markers and proinflammatory cytokine production of LDCs in an OVA-induced murine asthma model
The phenotypic maturation of DCs is mainly based on surface markers, including co-stimulatory molecules CD40, CD80,

␣-GalCer induces lung Treg cells to prevent asthma
and CD86 as well as MHC II molecules (24,25). To assess the modulatory effect of ␣-GalCer treatment before OVA sensitization on the phenotypic maturation of LDCs, the expression of surface maturation markers of LDCs was analyzed in asthmatic mice intraperitoneally treated with ␣-GalCer treatment before OVA sensitization. Our results indicated that LDCs from asthmatic mice treated with ␣-GalCer displayed significantly decreased expression of CD86, CD80, CD40, and MHC II, compared with PBS treatment (Fig. 4, A and B) (p Ͻ 0.05 or p Ͻ 0.01), but exhibited significantly elevated expression of CD86, CD80, CD40, and MHC II, compared with LDCs from WT mice unimmunized and unchallenged with OVA (Fig. 4, A and B) (p Ͻ 0.05 or p Ͻ 0.01).
Functionally mature DCs can secrete large quantities of proinflammatory cytokines like IL-12, IL-6, and TNF-␣ (25), but they do not secrete or they secrete negligible amounts of tolerogenic cytokine IL-10 (26). To determine the modulatory effect of ␣-GalCer treatment before OVA sensitization on the functional maturation of LDCs, LDCs sorted from WT asthmatic mice intraperitoneally treated with ␣-GalCer treatment before OVA sensitization were cultured for 72 h, and culture supernatants were harvested to measure the spontaneous secretion of IL-12p70, IL-6, TNF-␣, and IL-10. As shown in Fig. 4C, the purity of LDCs, i.e. CD11c ϩ F4/80 Ϫ cells, was over 99%. Our results showed that the concentrations of IL-12p70, IL-6, and TNF-␣ were markedly lower in the culture supernatants of LDCs from asthmatic mice treated with ␣-GalCer before OVA sensitization, compared with PBS treatment (Fig. 4D) (p Ͻ 0.05 or p Ͻ 0.01). In contrast, the concentration of IL-10 was significantly higher in the culture supernatant of LDCs from asthmatic mice treated with ␣-GalCer before OVA sensitization, compared with PBS treatment (Fig. 4D) (p Ͻ 0.05). However, the levels of IL-12p70, IL-6, TNF-␣, and IL-10 in the culture supernatant of LDCs from WT mice unimmunized and unchallenged with OVA were negligible (Fig. 4D). Furthermore, our findings showed that CD11c ϩ cells (red) in the lung tissues could be detected by immunofluorescence in OVA-induced asthmatic mice treated with ␣-GalCer or PBS before OVA sensitization and WT mice treated with ␣-GalCer or PBS (Fig. 4, E-G). However, CD11c ϩ TNF␣ ϩ cells, CD11c ϩ IL-12 ϩ cells, and CD11c ϩ IL-10 ϩ cells (yellow) in the lung tissues could be found in OVA-induced asthmatic mice treated with ␣-GalCer or PBS before OVA sensitization, but not in WT mice treated with ␣-GalCer or PBS (Fig. 4, E-G). In addition, CD11c ϩ TNF␣ ϩ cell and CD11c ϩ IL-12 ϩ cell infiltration was reduced, but CD11c ϩ IL-10 ϩ cell infiltration was increased in the lung tissues from asthmatic mice treated with ␣-GalCer before OVA sensitization (Fig. 4, E-G). Taken together, these results indicated that ␣-GalCer injection before OVA sensitization down-regulates immunogenic maturation of LDCs in an OVA-induced murine asthma model.

␣-GalCer treatment before allergen sensitization can enhance the expansion and suppressive activity of lung CD4 ؉ FoxP3 ؉ Treg cells in an OVA-induced murine asthma model
Previous data have shown that Treg cells have a reduced number and impaired function in patients with asthma (3, 4), and the induction of lung Treg cells may be a potential thera-peutic protocol for the treatment of asthma (8). Thus, we assessed whether ␣-GalCer treatment before allergen sensitization could affect the expansion and function of CD4 ϩ FoxP3 ϩ Treg cells in asthma. As shown in Fig. 5, the number of lung CD4 ϩ FoxP3 ϩ Treg cells in OVA-induced asthmatic WT mice with ␣-GalCer treatment before OVA sensitization was significantly elevated, compared with PBS treatment (Fig. 5, A and B) (p Ͻ 0.05). In addition, the IL-10 production in culture supernatants of Treg cells, whose sorted purity was ϳ98% (Fig. 5C), from OVA-induced asthmatic WT mice with ␣-GalCer treatment before OVA sensitization was markedly elevated, compared with PBS treatment (Fig. 5D) (p Ͻ 0.01). To further determine the suppressive activity of Treg cells in vitro, CD4 ϩ CD25 Ϫ T cells were co-cultured with the indicated ratio of CD4 ϩ CD25 ϩ Treg cells for 3 days, and then culture supernatants were collected to detect the levels of IL-4 and IFN-␥ by ELISA. As outlined in Fig. 5, the suppressive activity of Treg cells on cytokine production was gradually increased with increasing numbers of Treg cells. However, the lung Treg cells from asthmatic WT mice treated with ␣-GalCer before OVA sensitization were more effective in suppressing cytokine secretion of CD4 ϩ CD25 Ϫ T cells, compared with PBS treatment (Fig. 5E) (p Ͻ 0.05 or p Ͻ 0.01). Collectively, our findings showed that ␣-GalCer treatment before OVA sensitization can promote the expansion and suppressive activity of Treg cells in an OVA-induced murine asthma model.

iNKT cells contribute to the expansion and suppressive activity of Treg cells induced by ␣-GalCer treatment in asthmatic mice
Recent reports have shown that iNKT cells activated by ␣-GalCer play an important role in inducing the development of Treg cells and peripheral tolerance (14,15). Therefore, we investigated whether iNKT cells are responsible for the enhanced number and inhibitory activity of Treg cells in asthmatic mice treated with ␣-GalCer before allergen sensitization. As outlined in Fig. 6, the number of lung CD4 ϩ FoxP3 ϩ Treg cells (Fig. 6, A and B), the level of IL-10 in culture supernatant of Treg cells (Fig. 6C), as well as the suppressive capacity of Treg cells on cytokine production of CD4 ϩ CD25 Ϫ T cells (Fig. 6D) were significantly increased in OVA-immunized and OVAchallenged WT mice with ␣-GalCer treatment before OVA sensitization, compared with OVA-immunized and OVA-challenged CD1d Ϫ/Ϫ mice with same treatment (p Ͻ 0.05 or p Ͻ 0.01). Interestingly, no difference was observed in the number and suppressive capacity of Treg cells among OVA-immunized and OVA-challenged CD1d Ϫ/Ϫ mice with ␣-GalCer or PBS treatment, and OVA-immunized and OVA-challenged WT mice with PBS treatment (Fig. 6, A-D) (p Ͼ 0.05). Taken together, these data suggested that iNKT cells contribute to the expansion and suppressive activity of Treg cells induced by ␣-GalCer treatment in asthmatic mice.

Treg cells contribute to the inhibitory effects of ␣-GalCer treatment on Th2 cell responses in an OVA-induced murine asthma model
Controlling the function of Treg cells can modulate the induction of Th2 cells, which plays a vital role in suppressing aberrant immune responses and maintaining homeostasis (2).

␣-GalCer induces lung Treg cells to prevent asthma
Previous studies have shown that anti-CD25 mAb delivery can lead to functional inactivation of Treg cells in vivo (14,27). To determine whether Treg cells contributed to the inhibitory effects of ␣-GalCer on Th2 cell responses in asthma, mice were intravenously administered anti-CD25 mAb or IgG isotype mAb, rested for 1 day, followed by intraperitoneal administration of ␣-GalCer, then rested for 1 day, and subsequently sensitized and challenged with OVA. Airway inflammation and AHR were measured 24 h after the final OVA challenge. Our findings showed that anti-CD25 mAb delivery could reverse the protective effects of ␣-GalCer treatment before OVA sensitization on the level of inflammatory cell infiltration in the airways and mucus-secreting cells in the respiratory tracts (Fig. 7, A and B), the number of total cells, eosinophils, and macrophages in the BALF (Fig. 7C), the concentrations of such cytokines as IL-4, IL-5, and IL-13 in the BALF (Fig. 7D), the OVA-specific IgE and IgG1 in the serum (Fig. 7E), as well as AHR (Fig. 7F) in the OVA-induced asthmatic WT mice, compared with those treated with IgG isotype mAb (p Ͻ 0.05 or p Ͻ 0.01). Furthermore, anti-CD25 mAb administration significantly reduced the level of IL-10 in the BALF from the OVA-induced asthmatic WT mice, compared with those treated with IgG isotype mAb ( Fig. 7D) (p Ͻ 0.01). Collectively, these data showed that Treg cells contribute to the inhibitory effects of ␣-GalCer on Th2 cell responses in an OVA-induced murine model of asthma.

Treg cells contribute to the inhibitory effects of ␣-GalCer treatment on immunogenic maturation of LDCs in asthmatic mice
Furthermore, we investigated the effects of functional inactivation of Treg cells with anti-CD25 mAb delivery on surface maturation markers and proinflammatory cytokine production of LDCs in an OVA-induced murine asthma model intraperitoneally administered ␣-GalCer. Our data showed that anti-CD25 mAb administration could reverse the inhibitory influences of ␣-GalCer treatment on the expression levels of MHC II, CD80, CD86, and CD40 of LDCs (Fig. 8, A and B), as well as the capacity to secrete IL-12p70, IL-6, and TNF-␣ of LDCs ( Fig.  8C) from OVA-induced asthmatic WT mice, compared with those treated with IgG isotype mAb (p Ͻ 0.05 or p Ͻ 0.01). However, anti-CD25 mAb treatment could markedly downregulate the level of IL-10 in the culture supernatant of LDCs from OVA-induced asthmatic mice treated with ␣-GalCer, compared with IgG isotype mAb treatment (Fig. 8C) Thus, these data indicated that the suppressive capacity of ␣-GalCer on surface maturation markers and proinflammatory cytokine production of LDCs may require Treg cells in an OVA-induced murine asthma model.

␣-GalCer treatment before allergen sensitization can enhance the level of IL-2 in vivo
The generation, survival, and suppressive capacity of Treg cells are closely related to the IL-2 signal (28,29). Thus, we further determined whether ␣-GalCer treatment before allergen sensitization could promote IL-2 production in the lung by iNKT cells in OVA-induced murine asthma. The level of IL-2 in the BALF and the expression of il-2 mRNA in lung tissue were measured by ELISA and quantitative RT-PCR, respectively. Our findings showed that ␣-GalCer treatment before OVA sensitization could enhance the level of IL-2 in the BALF (Fig. 9A), as well as the expression of il-2 mRNA in the lung tissue ( Fig.  9B) in OVA-induced murine asthma, compared with PBS treatment (p Ͻ 0.01). However, no difference was detected in the level of IL-2 in the BALF and the expression of il-2 mRNA in the lung tissue between OVA-sensitized and OVA-challenged CD1d Ϫ/Ϫ mice with ␣-GalCer and PBS treatment (Fig. 9, A and B) (p Ͼ 0.05). Furthermore, the level of IL-2 in the BALF and the expression of lung il-2 mRNA from asthmatic WT mice treated with ␣-GalCer were markedly higher than asthmatic CD1d Ϫ/Ϫ mice treated with ␣-GalCer (Fig. 9, A and B) (p Ͻ 0.01). Furthermore, our data showed that the lung IL-2 ϩ iNKT cells in asthmatic mice treated with ␣-GalCer before OVA sensitization were markedly elevated, compared with asthmatic mice treated with PBS before OVA sensitization and WT mice treated with ␣-GalCer or PBS (Fig. 9, C and D) (p Ͻ 0.01). Also, our findings indicated that lung CD4 ϩ IL-2 ϩ T cells in asthmatic mice treated with ␣-GalCer before sensitization were significantly reduced compared with asthmatic mice treated with PBS before OVA sensitization, but were strongly increased compared with those in WT mice treated with ␣-GalCer or PBS (Fig. 9, E and F) (p Ͻ 0.01). Surprisingly, the number of IL-2 ϩ LDCs was very limited in asthmatic mice treated with ␣-GalCer before OVA sensitization (Fig. 4, G and H). Therefore, these results suggested that ␣-GalCer treatment before OVA sensitization can enhance the production of IL-2 in the lung through iNKT cells in OVA-induced murine asthma. . ␣-GalCer treatment before allergen sensitization down-regulates surface maturation markers and proinflammatory cytokine production of LDCs in an OVA-induced murine asthma model. WT mice, intraperitoneally treated with ␣-GalCer or PBS before OVA sensitization, were sacrificed 24 h after the final OVA challenge. Then the LDCs (CD11c ϩ F4/80 Ϫ ) were isolated using a combination of magnetic microbead selection and flow cytometry. LDCs were quantified by CD11c and F4/80 staining using flow cytometry. A, cells were stained for maturation-associated surface markers and detected by flow cytometry. Expression of MHC II, CD80, CD86, and CD40 was assessed in CD11c ϩ F4/80 Ϫ cells and isotype controls. B, expression levels of MHC II, CD80, CD86, and CD40 in CD11c ϩ F4/80 Ϫ cells. C, flow cytometry determined the purity of LDCs, which were stained with both CD11c and a mAb against F4/80 (i.e. the proportion of LDCs cells was over 99%). D, LDCs from OVA-induced asthmatic mice with PBS or ␣-GalCer treatment before OVA sensitization were cultured in vitro for 72 h. The concentrations of IL-12p70, IL-6, TNF-␣, and IL-10 in the culture supernatants were determined by ELISA. Data are represented as means Ϯ S.D. from three independent experiments (n ϭ 5), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01. E, immunofluorescence of the lung tissues was performed. Paraffin-embedded tissue sections were stained with anti-rabbit CD11c and anti-goat TNF␣ Abs followed by Alexa Fluor 594conjugated goat anti-rabbit IgG (red) or Alexa Fluor 488 -conjugated donkey anti-goat IgG (green). Red indicates CD11c ϩ cells; green indicates tnf-␣ ϩ cells, and yellow indicates CD11c ϩ tnf-␣ ϩ cells. F, paraffin-embedded lung tissue sections were stained with anti-rabbit CD11c and anti-goat il-12/il-35 p35 Abs followed by Alexa Fluor 594 -conjugated goat anti-rabbit IgG (red) or Alexa Fluor 488 -conjugated donkey anti-goat IgG (green). Red indicates CD11c ϩ cells; green indicates il-12 ϩ cells, and yellow indicates CD11c ϩ il-12 ϩ cells. G, paraffin-embedded lung tissue sections were stained with anti-rabbit CD11c and anti-goat il-10 Abs followed by Alexa Fluor 594 -conjugated goat anti-rabbit IgG (red) or Alexa Fluor 488 -conjugated donkey anti-goat IgG (green). Red indicates CD11c ϩ cells; green indicates IL-10 ϩ cells, and yellow indicates CD11c ϩ il-10 ϩ cells.

␣-GalCer induces lung Treg cells to prevent asthma Treg cells induced by ␣-GalCer-activated iNKT cells from OVAinduced murine asthma may involve in IL-2 in vitro
To directly assess the role of IL-2 in the generation of Treg cells induced by ␣-GalCer-activated iNKT cells, we co-cultured iNKT cells from OVA-induced asthmatic WT mice intraperitoneally administered ␣-GalCer before OVA sensitization and CD4 ϩ CD25 Ϫ T cells from naive WT mice in the presence of anti-IL-2 mAb or IgG isotype mAb. As shown in Fig. 10A, the purity of iNKT cells, gated as PBS-57/mCD1d tetramer ϩ TCR-␤ ϩ cells, was ϳ96%. Our results revealed that anti-IL-2 mAb treatment could significantly reduce the number of CD4 ϩ FoxP3 ϩ Treg cells and the expression of Foxp3 mRNA of culture cellular components (Fig. 10, B and D), as well as the level of IL-10 in culture supernatants (Fig. 10E), compared with IgG isotype mAb treatment (p Ͻ 0.05 or p Ͻ 0.01). Collectively, these results indicated that Treg cells induced by ␣-GalCer-activated iNKT cells from murine asthma may involve in IL-2 in vitro.

Discussion
In this study, our data indicate that intraperitoneal administration of ␣-GalCer before allergen sensitization could down- Figure 5. ␣-GalCer treatment before allergen sensitization can enhance the expansion and suppressive activity of lung CD4 ؉ FoxP3 ؉ Treg cells in an OVA-induced murine asthma model. WT mice, intraperitoneally treated with ␣-GalCer or PBS before the first OVA sensitization, were sacrificed 24 h after the final OVA challenge. Subsequently, spleen Treg cells (CD4 ϩ CD25 ϩ T cells) were isolated by magnetic microbead selection. CD4 ϩ CD25 Ϫ T cells were isolated from WT mice by magnetic microbead selection. A, lung Treg cells were confirmed by CD4 and FoxP3 staining in WT asthmatic mice intraperitoneally treated with ␣-GalCer or PBS before the first OVA sensitization using flow cytometry. The gating used for Treg cells (CD4 ϩ FoxP3 ϩ Treg cells) and the corresponding percentages are shown in each dot plot. B, percentages of lung Treg cells in WT asthmatic mice intraperitoneally treated with ␣-GalCer or PBS before first OVA sensitization. C, flow cytometry determined the purity of spleen Treg cells, which were stained with both CD4 and a mAb against CD25 (i.e. the proportion of Treg cells was over 98%). D, Treg cells were cultured in vitro for 3 days, and then the concentrations of IL-10 in supernatants were determined by ELISA. E, CD4 ϩ CD25 Ϫ T cells co-cultured with the indicated ratio of Treg cells for 3 days. The potency of the Treg-mediated suppression was expressed as the relative inhibition of cytokine (IFN-␥ and IL-4) production for each Treg/CD4 ϩ CD25 Ϫ T ratio. Data are shown as means Ϯ S.D. of three independent experiments (n ϭ 5), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01.

␣-GalCer induces lung Treg cells to prevent asthma
regulate Th2 cell responses and immunogenic maturation of LDCs in WT mice but not in CD1d Ϫ/Ϫ mice immunized and challenged with OVA or HDM. Meanwhile, intraperitoneal administration of ␣-GalCer before allergen sensitization could promote the expansion and suppressive activity of lung CD4 ϩ FoxP3 ϩ Treg cells in WT mice, but not in CD1d Ϫ/Ϫ mice immunized and challenged with OVA. Additionally, our data revealed that functional inactivation of Treg cells could reduce the inhibitory ability of ␣-GalCer on Th2 cell responses in a murine asthma model. Finally, intraperitoneal administration of ␣-GalCer could up-regulate production of IL-2 in the lung in vivo, and neutralization of IL-2 could reduce the expansion of Treg cells in vitro. Taken together, our findings provide evidence that intraperitoneal administration of ␣-GalCer before the development of asthma disorders, which can induce the generation of lung Treg cells by iNKT cells, may provide a potential therapeutic strategy to prevent allergic asthma.
␣-GalCer, a strong and specific activator for iNKT cells, can stimulate iNKT cells to promptly produce large quantities of Th1-and Th2-type cytokines, such as IL-4 and IFN-␥, and thereby regulate a diverse array of immunity (11). Our current results indicated that intraperitoneal administration of ␣-GalCer before allergen sensitization can down-regulate Th2 cell responses in an OVA-or HDM-induced murine asthma model. Interestingly, our data showed that intraperitoneal administration of ␣-GalCer before allergen sensitization can reduce the concentrations of IL-4, IL-5, and IL-13 and promote the level of IL-10 in the BALF, suggesting that this ␣-GalCer therapy can result in a switch of cytokine production profiles in asthma from Th2-like cytokines to an immunosuppressive cytokine IL-10. However, the above results were not observed in CD1d Ϫ/Ϫ mice sensitized and challenged with OVA. The CD1d Ϫ/Ϫ mice are short of the MHC class I restricting element needed by the iNKT cell and hence lack functional iNKT cells (30). Therefore, iNKT cells are mainly required for the downregulation of Th2 cell responses induced by intraperitoneal administration of ␣-GalCer before allergen sensitization in asthma. Our present results sharply differ from our previous findings showing that intraperitoneal administration of ␣-GalCer before allergen challenge can promote Th2 cell response by inducing immunogenic maturation of LDCs in murine asthma (18,19). Ronet et al. (16) reported that iNKT cells play a key role in the initiation of the lethal ileitis after T. gondii infection, but activation of iNKT cells by intraperitoneal administration of ␣-GalCer before T. gondii infection can enhance the frequency of IL-10 -secreting Treg cells to control the inflammatory intestinal disorder. These data support the idea that the contradictory role of iNKT cells activated by ␣-GalCer in different pathological conditions may be partially related to the timing of ␣-GalCer treatment (31).
Recent data have shown that Treg cells have reduced number and impaired function in patients with asthma (3,4), and the induction of lung Treg cells may be a potential therapeutic protocol for the treatment of asthma (8). IL-10 is a strong immunosuppressive cytokine that plays a fundamental role in the development of peripheral tolerance to allergen (32). It has

. iNKT cells contribute to the expansion and suppressive activity of Treg cells induced by ␣-GalCer treatment in asthmatic mice. WT mice and CD1d
Ϫ/Ϫ mice, immunized and challenged with OVA, were treated with an intraperitoneal injection of ␣-GalCer or PBS 24 h before the first sensitization. Spleen Treg cells (CD4 ϩ CD25 ϩ T cells) were isolated by magnetic microbead selection. CD4 ϩ CD25 Ϫ T cells were obtained from WT mice by magnetic microbead selection. A, lung Treg cells were confirmed by CD4 and FoxP3 staining using flow cytometry. The gating used for Treg cells (CD4 ϩ FoxP3 ϩ Treg cells) and the corresponding percentages are shown in each dot plot. B, percentages of lung Treg cells. C, spleen Treg cells were cultured in vitro for 3 days, and then the concentrations of IL-10 in supernatants were determined by ELISA. Data are shown as means Ϯ S.D. of three independent experiments (n ϭ 5), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01. NS, not significant. D, CD4 ϩ CD25 Ϫ T cells from WT mice co-cultured with the indicated ratio of Treg cells for 3 days. The potency of the Treg-mediated suppression was expressed as the relative inhibition of cytokine (IFN-␥ and IL-4) production for each Treg/CD4 ϩ CD25 Ϫ T ratio. Significant differences between the suppression ability of Treg cells from WT asthmatic mice or CD1d Ϫ/Ϫ asthmatic mice treated with ␣-GalCer are shown as **, p Ͻ 0.01.

␣-GalCer induces lung Treg cells to prevent asthma
been widely believed that IL-10 released by Treg cells can suppress the production of proinflammatory cytokines and inhibit the production of effector T-cell cytokines, and down-regulate antigen presentation and the expression of major histocompatibility complex II and costimulatory molecules of DCs (8). Actually, allergen immunotherapy can induce the generation of antigen-specific Treg cells that release IL-10 and suppress allergen-specific Th2 cell responses (33), whereas Treg cell-specific depletion of IL-10 can enhance allergic airway inflammation (34). These reports support the notion that IL-10 secreted by Treg cells plays an essential role in the development of immune tolerance in allergic airway diseases. In this study, our findings showed that ␣-GalCer treatment before allergen sensitization can promote the expansion and suppressive activity of lung Treg cells in WT mice and CD1d Ϫ/Ϫ mice sensitized and challenged with OVA, indicating that the effects of this ␣-GalCer treatment on Treg cells mainly required iNKT cells. Furthermore, our data indicated that functional inactivation of CD4 ϩ CD25 ϩ Treg cells using injection of anti-CD25 mAb can reverse the inhibitory influence of ␣-GalCer treatment before allergen sensitization on Th2 cell responses in a murine asthma model, suggesting that Treg cells contribute to the inhibitory

␣-GalCer induces lung Treg cells to prevent asthma
effects of ␣-GalCer on Th2 cell responses. Meanwhile, our findings revealed that ␣-GalCer treatment before allergen sensitization can promote the release of IL-10 from Treg cells in vitro, whereas specific functional inactivation of Treg cells by anti-CD25 mAb can abrogate the production of IL-10 in vivo. Our data suggest that Treg cells induced by this ␣-GalCer therapy may contribute to the production of IL-10 in asthma. Thus, these findings raise a potential mechanism that intraperitoneal administration of ␣-GalCer before allergen sensitization may be sufficient to protect against asthma through the induction of lung Treg cells by activated iNKT cells. Interestingly, our data showed that ␣-GalCer treatment before allergen sensitization can down-regulate immunogenic maturation of LDCs in a murine asthma model. Furthermore, in our preliminary study, our data showed that all CD11c ϩ tnf-␣ ϩ cells, CD11c ϩ il-12 ϩ cells, and CD11c ϩ il-10 ϩ cells in the lung tissues could be found by immunofluorescence in OVA-induced asthmatic mice treated with ␣-GalCer or PBS before OVA sensitization, but not in WT mice treated with ␣-GalCer or PBS. Meanwhile, CD11c ϩ TNF-␣ ϩ cell and CD11c ϩ IL-12 ϩ cell infiltration was reduced, but CD11c ϩ IL-10 ϩ cell infiltration was increased in the lung tissues from asthmatic mice treated with ␣-GalCer before OVA sensitization. Our preliminary results suggested that ␣-GalCer treatment before allergen sensitization induced the cytokine production profile switch of LDCs from TNF␣ or IL-12-producing DCs to IL-10 -producing DCs, but cannot fully demonstrate that ␣-GalCer administration prior to allergen sensitization can induce the LDC switch from TNF␣ or IL-12-producing DCs to later IL-10 -secreting DCs, or separate IL-10 only secreting LDC populations exist. Therefore, further study in this area will be required. It has been well-identified that LDCs can direct CD4 ϩ T cells differentiating into effector Th2 cells, inducing and maintaining maladaptive immunological response after allergen exposure in asthma (35,36). However, Treg cells can result in the development of tolerogenic phenotypes of dendritic cells and also suppress maturation of dendritic cells (37). As such, it is probable that intraperitoneal administration of ␣-GalCer before allergen sensitization inhibits Th2 cell response through the down-regulation of immunogenic maturation of LDCs mediated by Treg cells in asthma. Our present results partially differ from the previous reports showing that ␣-GalCer treatment can abrogate AHR and airway inflammation through activation of iNKT cells and IFN-␥ production in asthmatic mice (38,39). The

␣-GalCer induces lung Treg cells to prevent asthma
different results may be partially related to the different experimental protocols between their studies and ours, but further study is still needed.
It is widely accepted that immature or partially mature dendritic cells are able to induce the generation of Treg cells, whereas mature dendritic cells possess the capacity to polarize different subsets of effector T cells (40). This study showed that ␣-GalCer treatment before allergen sensitization can enhance the expansion and suppressive activity of lung Treg cells in asthmatic mice. However, LDCs exhibit the immunogenic maturation status in asthmatic mice treated with ␣-GalCer before allergen sensitization, albeit at a lower extent than asthmatic mice treated with PBS. Therefore, our findings provided possible evidence that the generation of lung Treg cells induced by ␣-GalCer treatment before allergen sensitization may not be related to LDCs in asthmatic mice. It is well-established that IL-2 plays a critical role in the generation, activation, and survival of Treg cells (41,42). Previous studies show some evidence that iNKT cells from murine and humans produce IL-2 to promote the development of Treg cells in the presence or absence of ␣-GalCer (17,43). In our present observation, our preliminary data showed that intraperitoneal administration of Figure 9. ␣-GalCer treatment before allergen sensitization can enhance the level of IL-2 in vivo. WT mice and CD1d Ϫ/Ϫ mice, immunized and challenged with OVA, were treated with an intraperitoneal injection of ␣-GalCer or PBS 24 h before the first sensitization. A, BALFs were collected from each group of mice, and IL-2 production was analyzed by ELISA. B, level of il-2 mRNA in the lung from each group of mice was measured by quantitative RT-PCR. C, lung iNKT cells (PBS-57/mCD1d ϩ TCR-␤ ϩ ) (gate P3) and lung IL-2 ϩ iNKT cells (PBS-57/mCD1d ϩ TCR-␤ ϩ IL-2 ϩ ) (gate P4) were determined by flow cytometry. The gating for lung iNKT cells and the corresponding percentages are indicated in each dot plot. D, percentages of IL-2 ϩ iNKT cells in lung iNKT cells were determined. E, lung CD4 ϩ T cells (CD4 ϩ ) (gate P2) and lung CD4 ϩ IL-2 ϩ T cells (CD4 ϩ IL-2 ϩ ) (gate P3) were determined by flow cytometry. The gating for lung CD4 ϩ T cells and the corresponding percentages are indicated in each dot plot. F, percentages of CD4 ϩ IL-2 ϩ T cells in lung CD4 ϩ T cells were determined. G, LDCs (CD11c ϩ F4/80 Ϫ ) (gate P2) and lung IL-2 ϩ LDCs (CD11c ϩ F4/80 Ϫ IL-2 ϩ ) (gate P3) were determined by flow cytometry. The gating for LDCs and the corresponding percentages are indicated in each dot plot. H, percentages of CD11c ϩ F4/80 Ϫ IL-2 ϩ LDCs in LDCs were determined. Data are represented as means Ϯ S.D. from three independent experiments (n ϭ 5), and one representative experiment is indicated. **, p Ͻ 0.01. NS, not significant.

␣-GalCer induces lung Treg cells to prevent asthma
␣-GalCer before allergen sensitization could promote the expression of il-2 gene and il-2 protein in the lung from asthmatic WT mice, but not asthmatic CD1d Ϫ/Ϫ mice. It has been well-demonstrated that the CD1d Ϫ/Ϫ mice are devoid of the MHC class I restricting element required by iNKT cells and thereby lack iNKT cells (30). Meanwhile, our data show that the lung IL-2 ϩ iNKT cells were markedly elevated, whereas lung CD4 ϩ IL-2 ϩ T cells were significantly reduced in asthmatic mice treated with ␣-GalCer before OVA sensitization. Also, the number of IL-2 ϩ LDCs was very limited in asthmatic mice treated with ␣-GalCer before OVA sensitization. These results provided the evidence that the elevated production of IL-2 induced by ␣-GalCer is mainly dependent on lung iNKT cells, but not lung CD4 ϩ T cells and LDCs. Furthermore, neutralization of IL-2 could reduce the production of IL-10 in culture supernatant, the expression of Foxp3 mRNA of cellular components, and the number of CD4 ϩ FoxP3 ϩ Treg cells in vitro. Taken together, our preliminary findings revealed that ␣-GalCer treatment before antigen sensitization could promote the level of IL-2 in the airways through activation of lung iNKT cells, which is likely to play a potential role in the generation of lung Treg cells in asthma.
In summary, our present findings demonstrated that ␣-GalCer treatment before antigen sensitization can promote iNKT cell-mediated induction of lung Treg cells, preventing Th2 cell responses in murine asthma. Presently, the mostly important treatments utilized to control asthma are inhaled glucocorticoids, with or without ␤2 agonists, which possess broad-scale anti-inflammatory ability (44). However, these therapy strategies merely ameliorate the symptoms, rather than abrogating the underlying mechanism, and thereby fail to control the disorders in marked proportion of patients with asthma. Treg cells can directly or indirectly suppress the initiation of allergic inflammation, mucus overproduction, IgE release, and AHR through modulating antigen-presenting cells, Th2 and Th17 cells, inflammatory cells, and B cells in asthma (8). Because CD1d molecules are extremely conserved in mammals, and TCRs expressed by murine and human iNKT cells have structural and functional similarity (45,46), ␣-GalCer can bind human CD1d molecules and activate iNKT cells in Figure 10. Treg cells induced by ␣-GalCer-activated iNKT cells from OVA-induced murine asthma may involve IL-2 in vitro. Lung iNKT cells (PBS-57/ mCD1d ϩ TCR-␤ ϩ iNKT cells) were isolated from OVA-induced asthmatic mice intraperitoneally treated with ␣-GalCer before the first OVA sensitization, and spleen CD4 ϩ CD25 Ϫ T cells were isolated from WT mice using magnetic microbead selection. iNKT cells were cultured alone or co-cultured with CD4 ϩ CD25 Ϫ T cells in the presence of anti-IL-2 mAb or IgG isotype mAb for 72 h. A, flow cytometry determined the purity of lung iNKT cells, which were stained with both PBS-57/mCD1d and a mAb against TCR-␤ (i.e. the proportion of iNKT cells was ϳ96%). B, Treg cells of cellular components from culture medium were measured by CD4 and FoxP3 staining using flow cytometry. The gating used for Treg cells (CD4 ϩ FoxP3 ϩ Treg cells) and the corresponding percentages are shown in each dot plot. C, percentages of Treg cells from cellular components from culture medium. D, expression level of Foxp3 mRNA of cellular components from culture medium was analyzed by quantitative RT-PCR. E, concentration of IL-10 in culture supernatants was determined by ELISA. Data are shown as means Ϯ S.D. of three independent experiments (n ϭ 15), and one representative experiment is indicated. *, p Ͻ 0.05; **, p Ͻ 0.01.

␣-GalCer induces lung Treg cells to prevent asthma
humans (47). Additionally, ␣-GalCer appears to be well-tolerated over a broad range of doses by human patients (48). Meanwhile, our present observations that ␣-GalCer treatment before antigen sensitization can inhibit Th2 cell responses in an OVAor HDM-induced murine asthma model indicate that this therapy may be effective for multiple antigens. Thus, our data raise a possibility that ␣-GalCer treatment before the development of asthma symptoms induces the generation of lung Treg cells via activated iNKT cells and may provide a potential therapeutic strategy to prevent allergic asthma for humans.

Mice
Female WT BALB/c mice (6 -8 weeks old) were obtained from the Animal Biosafety Level 3 Laboratory, the Center for Animal Experiment, Wuhan University (Wuhan, China). CD1d Ϫ/Ϫ mice (BALB/c background) were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice were bred under specific pathogen-free and controlled environmental conditions (22°C, 12-h light/12-h dark cycle). All procedures were approved by the Institutional Ethics Committee of Wuhan University.

Animal experimental models
WT BALB/c mice or CD1d Ϫ/Ϫ BALB/c mice were intraperitoneally immunized on days 0 and 14 with 20 g of chicken OVA (grade V, Sigma) absorbed in 2 mg of aluminum hydroxide (ThermoFisher Scientific Pierce) in 200 l of PBS. Intranasal OVA challenges (100 g/50 l in PBS) were followed on days 21-23. Mice were sacrificed 24 h after the final challenge to further study. In some experiments, WT mice were sensitized intratracheally on day 0 with 1 g of HDM extracts (Dermatophagoides pteronyssinus, Greer Laboratories) and were subsequently intranasally challenged from days 6 to 10 with 10 g of HDM extracts (49). Mice were euthanized 3 days after the final challenge to further analyze. Sex-and age-matched mice were similarly sensitized and challenged with PBS alone as control.

In vivo administration of ␣-GalCer
␣-GalCer (Enzo Life Sciences, Ann Arbor, MI) was dissolved in PBS containing 20% dimethyl sulfoxide (DMSO) at 0.01 mg/ml as a stock solution and stored at Ϫ20°C. The intraperitoneal injection was used as the route of administration of ␣-GalCer, as reported previously (50). Mice received a single intraperitoneal injection of 2 g of ␣-GalCer 1 day before the first sensitization with OVA or HDM. Control mice received intraperitoneal administration of an identical volume of vehicle solution (PBS/20% DMSO).

In vivo Ab administration
For functional inactivation of CD4 ϩ CD25 ϩ Treg cells, mice were intravenously delivered with 500 g of anti-CD25 mAb (clone PC61; BD Pharmingen) or IgG isotype mAb (Sigma), rested for 1 day, followed by in vivo administration of ␣-GalCer, then rested for 1 day, and subsequently sensitized and challenged with OVA.

Isolation of LDCs and cytokine production
LDCs were harvested by magnetic bead purification combined with fluorescence-activated cell sorting (FACS). WT mice and CD1d Ϫ/Ϫ mice were sacrificed 24 h after the last OVA challenge. Lungs were excised, and single cell suspensions were prepared, as described previously (21). Subsequently, CD11c ϩ cells were selected using CD11c microbeads (Miltenyi Biotec, Auburn, CA) and MASC columns (Miltenyi Biotec). A previous study has demonstrated that LDCs express the DC marker CD11c, but not the macrophage marker F4/80 (51). Thus, LDCs were gated as CD11c ϩ F4/80 Ϫ cells and isolated with a BD FAC-SAria III cell sorter (BD Biosciences). The purity of the sorted LDCs was measured by flow cytometry for APC-cy7conjugated anti-CD11c-positive and PE-cy5-conjugated anti-F4/80 -negative cells.
LDCs isolated from different groups of mice were cultured (37°C, 5% CO 2 ) in complete RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 1% streptomycin/penicillin, 0.1 mM nonessential amino acids, and 50 M mercaptoethanol (all from Invitrogen, Karlsruhe, Germany) in 96-well plates at 2 ϫ 10 5 cells/well for 72 h in the presence of OVA (100 g/ml) or vehicle. Supernatants were obtained to determine the concentrations of IL-10, IL-12p70, IL-6, and TNF␣ by ELISA (eBioscience, San Diego, CA) on the basis of the manufacturer's instructions.

Isolation of CD4 ؉ T cells, CD4 ؉ CD25 ؉ T cells, and CD4 ؉ CD25 ؊ T cells
CD4 ϩ CD25 ϩ T cells were isolated from spleens of WT or CD1d Ϫ/Ϫ mice treated with intraperitoneal injection of ␣-GalCer or PBS before OVA sensitization, whereas CD4 ϩ T cells and CD4 ϩ CD25 Ϫ T cells were isolated from spleens of WT mice. CD4 ϩ T cells were enriched from single spleen cell suspensions using R&D CD4 ϩ subset enrichment columns (R&D Systems, Minneapolis, MN). For obtaining CD4 ϩ CD25 Ϫ T cells and CD4 ϩ CD25 ϩ T cells, CD4 ϩ T cells were labeled with anti-CD25-PE, and then the CD4 ϩ T-cell fraction was loaded onto a MACS column placed in the magnetic field of a MACS separator (Miltenyi Biotec). Magnetically-labeled CD4 ϩ CD25 Ϫ T cells and CD4 ϩ CD25 ϩ T cells were obtained following the manufacturer's procedures. The purity of the sorted CD4 ϩ CD25 ϩ T cells was examined by flow cytometry ␣-GalCer induces lung Treg cells to prevent asthma using anti-CD4 -FITC and anti-CD25-PE (eBioscience). CD4 ϩ CD25 Ϫ T cells, used as responder T cells, and CD4 ϩ CD25 ϩ T cells were harvested for further analysis.

Co-culture of iNKT cells and CD4 ؉ CD25 ؊ T cells
To sorted lung iNKT cells, WT mice treated with intraperitoneal injection of ␣-GalCer before OVA immunization were sacrificed 24 h after the final OVA challenge. The lung single cell suspensions were obtained. iNKT cells were enriched using magnetic bead purification following the manufacturer's instruction (Miltenyi Biotec). The purity of the sorted iNKT cells was examined by flow cytometry for FITC-TCR-␤ and PE-PBS57/mCD1d tetramer double-positive cells. iNKT cells (1.5 ϫ 10 5 /well) were cultured alone or with the indicated numbers of CD4 ϩ CD25 Ϫ T cells from WT mice in round-bottom 96-well plates pre-bound with anti-CD3 (2 g/ml) in RPMI 1640 medium supplemented with 10% heat-inactivated FBS, 1% streptomycin/penicillin, 2 mM L-glutamine, 5 mM Hepes, and 1 mM sodium pyruvate and anti-CD28 (2 g/ml) (all from Invitrogen, Karlsruhe, Germany) and restimulated with ␣-GalCer (100 ng/ml) in the presence of anti-IL-2 mAb (0.2 g/ml) (IgG2a, clone S4B6; BD Pharmingen) or IgG isotype mAb (Sigma-Aldrich, Tauf-kirchen, Germany). After 3 days, the cellular components were harvested, and the frequency of CD4 ϩ FoxP3 ϩ Treg cells and the expression of Foxp3 mRNA were analyzed by flow cytometry and RT-PCR, respectively. The level of IL-10 in culture supernatants was measured by ELISA on the basis of the manufacturer's protocol (eBioscience).

Flow cytometric analysis
To examine the frequency of LDCs, lung MNCs were resuspended in FACS buffer, and then the cells were incubated with an anti-CD16/CD32 antibody (BD Biosciences) for 1 h to reduce nonspecific binding. After two washes, isotype controls or PE-cy5-conjugated anti-F4/80 and APC-cy7-conjugated anti-CD11c (eBioscience) were labeled. Forward scatter and side scatter were used to exclude dead cells. To further assess the expression of different surface markers, freshly sorted LDCs from the different experimental groups were labeled with anti-CD86 -PE, anti-CD80 -PE, anti-CD40 -PE, anti-MHC II-PE, or respective isotype controls (eBioscience).
To assess the effect of ␣-GalCer treatment before allergen sensitization on the number of lung CD4 ϩ T cells or Treg cells, lung MNCs were harvested and resuspended in the FACS buffer at a level of 1-2 ϫ 10 6 cells/ml. The lung MNCs were first blocked using an anti-CD16/CD32 antibody (BD Biosciences) to abate nonspecific binding and then labeled with isotype controls or the antibodies as follows: anti-CD4-FITC and anti-FoxP3-PeCy7 (eBioscience). Intracellular staining for FoxP3 was performed with Fix/Perm buffer reagents (eBioscience) according to the manufacturer's procedure. Lung Treg cells were identified as CD4 and Foxp3 double-positive cells.
The lung MNCs were labeled with isotype controls or PeCy5-TCR-␤ (eBioscience) and PE-PBS-57/mCD1d tetramer (gifted by the Natural Institutes of Health tetramer core facility) to determine the number of lung iNKT cells. In addition, to examine the secretion of IL-2 by lung MNCs, intracellular cytokine staining of BV421-IL-2 (BD Biosciences) was performed following the manufacturer's procedures (eBioscience).
All staining reactions were conducted at 4°C. The cells were measured by flow cytometry (Epics Altra; Beckman, Seattle, WA), and the acquired data were analyzed using FlowJo software (Tree Star, San Carlos, CA).

Examination of AHR
AHR was examined by invasive examination of airway resistance, as reported previously (53). WT mice and CD1d Ϫ/Ϫ mice intraperitoneally treated with ␣-GalCer or PBS before OVA or HDM sensitization or asthmatic WT mice intravenously injected with anti-CD25 mAb or IgG isotype mAb were anesthetized 24 or 72 h after the last allergen challenge, and then the mice were inserted with a 20-gauge polyethylene catheter. Subsequently, mechanical ventilation was performed. Dynamic compliance (Cdyn) and airway resistance (RL) to increasing concentrations of aerosolized methacholine (Mch), ranging from 3.12 to 50 mg/ml in PBS, for 3 min were measured by the FinePointe RC system (Wilmington, NC) to analyze AHR. Nonspecific airway responsiveness was measured on mice that were exposed to aerosolized PBS to decide their baseline value.

Bronchoalveolar fluid and lung histopathology
After the measurement of AHR, the lungs were lavaged using a total volume of 1.5 ml of PBS with 1 mM sodium EDTA, and subsequently, the lungs were resected, as mentioned previously (21). The supernatants of BALF were harvested and stored at Ϫ80°C to examine the levels of IL-2, IL-4, IL-5, IL-10, IL-13, and IFN-␥ by ELISA on the basis of the manufacturer's procedures (eBioscience). Cells in the BALF were collected to stain with May-Grunswald Giemsa (Jiancheng, Nanjing, China) for differential cell counting. The frequencies of eosinophils, neu-␣-GalCer induces lung Treg cells to prevent asthma trophils, lymphocytes, and macrophages were expressed as the absolute numbers on the basis of the total cell counts.
The left lungs were fixed in 4% buffered paraformaldehyde immediately after bronchoalveolar lavage. Then the samples were dehydrated and embedded in paraffin. Lung tissue sections were stained using hematoxylin-eosin (H&E) staining to evaluate airway inflammation and PAS staining (Baso, Taiwan, China) to assess goblet cell hyperplasia using light microscopy. Goblet cell hyperplasia was measured by counting the frequency of PAS-positive cells in the epithelium of the central respiratory tract with a digital image. The circumferences of the respiratory tract at the basement membrane were determined using ImageJ software (National Institutes of Health, Bethesda). Goblet cell hyperplasias were expressed as the frequency of PAS-positive cells per unit length (mm) of the basement membrane.

Analysis of OVA-or HDM-specific IgE and IgG1 in the serum
For immunoglobulin analysis, 24 h after the final OVA challenge or 3 days after the last HDM challenge, the whole blood was harvested to determine the levels of OVA-specific IgE and IgG1 or HDM-specific IgE and IgG1 by means of ELISA (eBioscience) following the manufacturer's instructions. Serum samples were diluted for total IgE or total IgG according to the instructions, respectively, in PBS containing 1% BSA (w/v) (Invitrogen, Karlsruhe, Germany). The OVA-specific IgE and IgG1 or HDM-specific IgE and IgG1 levels in serum samples were assessed by an indirect ELISA method. Briefly, the 96-well plates (ThermoFisher Scientific) were coated with 200 g/ml OVA or 10 g/ml HDM extract diluted in PBS, and blocked with 3% BSA (w/v) in PBS overnight at 4°C. Serum samples (50 l/well) were incubated overnight at 4°C. Biotin-conjugated goat anti-mouse IgE or IgG1 (Biolegend, San Diego) were added to the wells used as secondary detection antibodies and incubated for 1 h at room temperature. Avidin-horseradish peroxidase (HRP) (eBioscience) and 3,3Ј,5,5-tetramethylbenzidine (TMB) (ThermoFisher Scientific) were used to develop the colorimetric reaction. Reaction was stopped by 2 M H 2 SO 4 , and the absorbance was recorded using a BioTek Synergy 4 microplate reader (BioTek).

Statistical analyses
All data are expressed as the mean Ϯ S.D. Statistical analyses were performed with a Student's unpaired t test or one-way analysis of variance using GraphPad Prism 5 (GraphPad Software Inc., San Diego) software. p Ͻ 0.05 was considered statistically significant.