Vα14 NKT cells exhibit various immune regulatory properties in vivo, but their precise mechanisms remain to be solved. In this study, we demonstrate the mechanisms of generation of regulatory dendritic cells (DCs) by stimulation of Vα14 NKT cells in vivo. After repeated injection of α-galactosylceramide (α-GalCer) into mice, splenic DCs acquired properties of regulatory DCs in IL-10-dependent fashion, such as nonmatured phenotypes and increased IL-10 but reduced IL-12 production. The unique cytokine profile in these DCs appears to be regulated by ERK1/2 and IκBNS. These DCs also showed an ability to suppress the development of experimental allergic encephalomyelitis by generating IL-10-producing regulatory CD4 T cells in vivo. These findings contribute to explaining how Vα14 NKT cells regulate the immune responses in vivo.
The Vα14 NKT cell represents a distinct lymphocyte subpopulation characterized by the expression of an invariant Ag receptor α-chain encoded by the Vα14-Jα281 gene segments in mice. The invariant Ag receptors recognize glycolipid Ags, including α-galactosylceramide (α-GalCer)4 (1), presented by CD1d (2). It has been reported that Vα14 NKT cells play a pivotal role in influencing a very diverse group of immune responses (reviewed in Refs.3 and 4). In fact, activation of Vα14 NKT cells with α-GalCer can induce effective host defense against malignant tumor cells (5, 6, 7) or various infectious agents, such as bacteria, parasites, fungi, and viruses (8, 9, 10). In contrast, Vα14 NKT cells are also involved in immune-regulatory responses, such as the maintenance of transplantation tolerance (11, 12), anterior chamber-associated immune deviation (13), and prevention of some autoimmune diseases, i.e., experimental allergic encephalomyelitis (EAE) and type 1 diabetes in NOD mice (14, 15, 16, 17, 18). These data indicate that Vα14 NKT cells can exert both immune stimulatory and immune regulatory functions. Nonetheless, it remains enigmatic what factor dictates the direction of Vα14 NKT cell-mediated immune responses.
The mechanisms underlying the immune stimulatory responses elicited by Vα14 NKT cells are now relatively well elucidated. It is known that there is a cognate interaction between Vα14 NKT cells and dendritic cells (DCs) in their activation process, using costimulatory molecules and cytokines such as CD40L and IL-12, respectively (19, 20, 21). Activation of Vα14 NKT cells with α-GalCer induces a rapid differentiation and maturation of DCs in vivo, followed by enhancement of Th1-type immune responses (22, 23).
In contrast, the mechanisms by which Vα14 NKT cells exert immune regulatory functions are poorly understood, although some data have suggested that Th2 cytokines such as IL-4 and/or IL-10 produced by Vα14 NKT cells play some roles in the process (13, 24, 25, 26). Furthermore, a contribution of DCs in the Vα14 NKT cell-mediated regulatory immune responses has been suggested. In fact, Naumov et al. (27) have indicated that treatment of NOD mice with α-GalCer led to increased numbers of CD8α− DCs in the pancreatic lymph nodes, and adoptive transfer of such DCs could prevent the development of diabetes. Similarly, Chen et al. (28) also recently reported that activated Vα14 NKT cells inhibit autoimmune diabetes through a recruitment of tolerogenic DCs to pancreatic lymph nodes and an induction of apoptosis and anergy in diabetogenic T cells. These data strongly implicate a role of regulatory DCs acting downstream of Vα14 NKT cell activation.
Considering the ability of Vα14 NKT cells to elicit either enhancement or suppression of the immune responses, it is conceivable that distinct Vα14 NKT cell stimulations induce different properties in DCs, which subsequently alter systemic immune responses. However, little is known as to the properties of the DCs, especially regulatory DCs, induced by Vα14 NKT cell stimulation. Particularly, it is unclear which cytokines or signal transductions determine the fate and function of DCs when interacting with Vα14 NKT cells. Furthermore, it has not been fully understood how Vα14 NKT cell-induced regulatory DCs suppress the immune responses. It is thus of prime importance to elucidate the mechanism by which activated Vα14 NKT cells suppress the immune response via induction of regulatory DCs. To investigate molecular and cellular mechanisms of acquisition of regulatory properties in DCs, we used repeated injections of α-GalCer as a model of immune regulation, as it has been shown that multiple α-GalCer treatments could ameliorate autoimmune diseases.
In this study, we found that IL-10 derived from Vα14 NKT cells activates DCs to enhance IL-10 in CD8− DCs and to reduce IL-12 production in CD8+ DCs, respectively, which render DCs to be regulatory. Moreover, in the Vα14 NKT cell-induced regulatory DCs, expression of the nuclear IκB protein IκBNS and phosphorylation of ERK1/2, which seem to be responsible for high IL-10 and low IL-12 production, are augmented. These cytokine changes in DCs are responsible for induction of CD4 regulatory T cells suppressing immune responses. Therefore, IL-10 from Vα14 NKT cells triggers a regulatory cascade by modulating other immune regulatory cells including DCs and CD4 T cells.
Materials and Methods
Six- to 10-wk-old female C57BL/6 (B6) mice were purchased from Japan CLEA. Jα281−/− (Vα14 NKT) knockout (KO) mice were generated previously (5) and backcrossed >10 times to B6 mice. IL-10 KO mice with a B6 background were purchased from The Jackson Laboratory. All mice were bred and maintained in the animal facilities in RCAI, RIKEN (Kanagawa, Japan) under specific pathogen-free conditions. Animal care was in accordance with the guidelines of RIKEN.
α-GalCer was synthesized in the Laboratory for Immune Regulation (RIKEN) and also provided by Kirin Brewery. CpG oligodeoxynucleotide (ODN 1668) was synthesized by Hokkaido System Science. LPS and pertussis toxin were purchased from Sigma-Aldrich. Myelin oligodendrocyte glycoprotein (MOG)35–55, MEVGWYRSPFSRVVHLYRNGK, was synthesized by BEX. CFA and heat-killed Mycobacterium tuberculosis H37Ra were purchased from Difco. Recombinant murine CD40L was prepared in our laboratory (RIKEN). Monoclonal Abs and fluorescent reagents used for stimulation or blocking and FACS staining were obtained from BD Pharmingen. Recombinant IL-10 and IFN-γ were purchased from PeproTech. Abs for Western blot analyses were purchased from Cell Signaling Technology. A specific inhibitor for ERK1/2, U0126, was purchased from Promega.
Preparation of α-GalCer-loaded CD1d tetramers
A mouse CD1d and Ig fusion gene was created by fusing the cDNA of the extracellular domain of mouse CD1d in-frame to the CH2-CH3 portion of mutated human IgG1 to prevent binding to FcRγ. The resulting plasmid encoding the CD1d-Ig molecule along with β2-microglobulin constructed in baculoviral vector was transfected into Sf9 insect cells by Cellfectin (Invitrogen Life Technologies) according to the manufacturer’s instructions. The baculoviral stock obtained was used for viral amplification. The soluble CD1d-Ig fusion proteins were purified from the final large scale culture supernatant of baculoviral infected Sf9 cells grown in serum-free media by chromatography on a protein A-Sepharose column (Amersham Biosciences). Loading of CD1d-Ig dimers with α-GalCer was performed at neutral pH by overnight incubation at a molar ratio of 1:4 (CD1d-Ig dimer to α-GalCer) at ambient temperature. α-GalCer-loaded CD1d-Ig dimers were tetramerized with PE-conjugated polyclonal goat anti-human IgG F(ab′)2 (Beckman Coulter).
In vivo administration of α-GalCer
Mice received i.p. injections of 2 μg of α-GalCer in 200 μl of PBS. For repeated Vα14 NKT cell stimulation, mice received three injections of α-GalCer at intervals of 3–4 days.
Isolation of DCs
DCs were purified from collagenase-treated mouse spleens by magnetic cell sorting (MACS) with PE-conjugated anti-CD11c mAb and anti-PE mAb-coupled magnetic beads (Miltenyi Biotec), according to the manufacturer’s instructions. In this report, DCs from mice receiving no α-GalCer treatment (GC0) are referred to as DCGC0. DCs from mice injected once (GC1) or three (GC3) times with α-GalCer are referred to as DCGC1 or DCGC3, respectively.
Flow cytometric analysis and sorting
Splenocytes were preincubated with 2.4G2 anti-FcRγ mAb to block nonspecific binding. Then, the cells were stained with PE-conjugated anti-CD11c and a mixture of biotinylated mAbs to CD19, NK1.1, TER119, and CD3 for 30 min at 4°C. Biotinylated mAbs were detected with CyChrome-conjugated streptavidin. For surface phenotyping, cells were stained with FITC-labeled mAbs. Flow cytometric analysis was performed with a FACSCalibur flow cytometer using CellQuest software (BD Biosciences). α-GalCer-loaded CD1d tetramer-positive Vα14 NKT cells and CD8α+ or CD8α− DCs were sorted with Moflo (DAKO).
Cytokine production by DCs
CD11c+ DCs were isolated from naive or α-GalCer-treated mice, and suspended in complete RPMI 1640 medium supplemented with 10% FCS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 55 μM 2-ME. The cells were cultured in 96-well flat-bottom plates at 1 × 105 cells/well for 48 h in 200 μl of medium with or without LPS (10 μg/ml), CpG ODN (1 μM), or recombinant murine CD40L (2.5 μg/ml). In some experiments, a specific inhibitor of ERK1/2 (U0126) was added to some wells of the culture. In other experiments, purified DCs (1 × 106) were cultured with Vα14 NKT cells (1 × 105) in the presence or absence of α-GalCer (100 ng/ml), in 24-well plates with or without 50 μg/ml anti-IFN-γ (R4-6A2) or IL-10R (1B1) mAb. Three days later, DCs were isolated by depleting TCR-β-positive cells with MACS, and stimulated with CpG ODN. Concentrations of cytokines in the supernatants were measured by ELISA using OptEIA mouse cytokine detection kits (BD Pharmingen).
Western blotting, mobility shift assay, and real-time PCR
DC samples (2 × 105 cells) were suspended in 20 μl of 1× SDS sample buffer containing DTT, boiled for 15 min, and subjected to SDS-PAGE. Proteins were transferred to a polyvinylidene difluoride membrane and then subjected to immunoblotting with Abs for MAPK and IκB kinase (IKK)αβ. Nuclear extracts from the DCs were subjected to NF-κB EMSA according to the manufacturer’s instructions. Expression levels of IκBNS in cDNA samples taken from DCs were measured by real-time PCR (Applied Biosystems) after normalization for the expression of hypoxanthine-guanine phosphoribosyltransferase.
EAE induction and clinical evaluation
Before EAE induction, the host mice were pretreated with DCs isolated from naive or α-GalCer-treated mice. DCs were incubated with 100 μg/ml MOG35–55 peptide in complete medium for 2 h at 37°C. After intensive washing, the MOG35–55 peptide-pulsed DCs (5 × 105) were injected i.v. into B6 mice on days −7, −5, and −3 before EAE induction (day 0). Then, the pretreated mice were injected s.c. with 200 μg of MOG35–55 peptide in 100 μl of PBS emulsified with 100 μl of CFA and further enriched with 5 mg/ml M. tuberculosis (H37Ra). In addition, 500 ng of pertussis toxin was injected i.p. on day 0 and day 2. Clinical symptoms were monitored daily after immunization. The clinical score was graded as follows: 0, no disease; 1, tail limpness; 2, hind limb weakness; 3, hind limb paralysis; 4, fore limb weakness; 5, quadriplegia; 6, death. Cumulative disease scores were calculated by adding daily disease scores from the day after immunization until the end of the experiment.
In vitro culture of T cells
To identify Ag-specific T cells in the EAE-induced mice, splenic CD4 T cells were obtained 3 wk after EAE induction, were cultured for 4 days with MOG35–55- or OVA-pulsed DCs in the presence or absence of anti-IL-10R mAb (50 μg/ml), and were measured for their incorporation of [3H]thymidine. CD4 T cells thus cultured for 1 wk were restimulated with plate-bound anti-CD3 mAb and then assayed for their cytokine production by intracellular staining (29).
Alteration of DC phenotype and cytokine profile by in vivo α-GalCer treatment
We first assessed the alteration of surface phenotypes of splenic DCs after repeated α-GalCer treatment by FACS. As reported (22), a single injection of α-GalCer induced full maturation in splenic DCs within only 6 h, as characterized by up-regulation of costimulatory (CD40, CD80, and CD86) and MHC class II (I-Ab) molecules (DCGC1) (Fig. 1). These characteristic phenotypic changes lasted for at least 36 h and gradually returned to normal levels by 4 days after the stimulation (Fig. 1,B). Different from the findings in the case of the single α-GalCer stimulation, the maturation of DCs was not observed when α-GalCer was injected three times with intervals between injections of 3–4 days (DCGC3). Rather, expression levels of surface molecules in these DCs were significantly lower and almost similar to those of unstimulated DCs (DCGC0) (Fig. 1). These surface phenotypic alterations in DCs were not detected in Vα14 NKT cell-deficient (Vα14 NKT KO) mice, indicating that the phenotypic changes are due to the activation of Vα14 NKT cells (Fig. 1).
We next examined cytokine profiles in DCs after α-GalCer stimulation. DCs isolated from mice 0 to 5 days after the last α-GalCer injections were stimulated with CpG ODN, LPS, or recombinant murine CD40L and were measured for their cytokine production in culture supernatants by ELISA. Upon CpG ODN, LPS, or recombinant murine CD40L stimulation, DCGC1 produced large amounts of proinflammatory cytokines, such as IL-12, IL-6, and TNF-α, but only low levels of IL-10 (Fig. 2,A and data not shown). By contrast, DCGC3 showed enhanced IL-10 production but dramatically reduced levels of IL-12 (Fig. 2 A).
Vα14 NKT cells directly affect the cytokine profile of DCs
To understand the induction mechanism of the characteristic cytokine production in DCGC3, we first assessed alteration of cytokine productions of Vα14 NKT cells after in vivo α-GalCer treatment. We isolated Vα14 NKT cells from naive or in vivo α-GalCer-treated (three times) mice and restimulated them in vitro with α-GalCer-pulsed fresh DCs. As reported (30, 31), in vivo α-GalCer treatment dramatically reduced the ability of Vα14 NKT cells to produce IFN-γ but not IL-10 upon restimulation (Fig. 2,B). Then, to investigate whether the altered cytokine production in Vα14 NKT cells directly affected the high IL-10 production in DCs, we cocultured freshly isolated DCs and Vα14 NKT cells in the presence of α-GalCer, then assessed CpG ODN-induced IL-10 production by the DCs. After coculture with in vivo stimulated Vα14 NKT cells (three times), DCs obtained ability to produce approximately twice higher level of IL-10 than those cocultured with naive Vα14 NKT cells (Fig. 2,C). This up-regulation of IL-10 production by DCs was completely inhibited by addition of mAb to IL-10R but not addition of mAb to IFN-γ (Fig. 2 C) nor to IL-4 (data not shown). These results indicate that in vivo repeatedly stimulated Vα14 NKT cells can directly enhance IL-10 production of DCs in an IL-10-dependent manner.
DC subpopulations after in vivo α-GalCer treatment
To see the change of DC subpopulations after in vivo α-GalCer treatment, we examined the expression levels of CD8α/CD11b, B220/CD11c, and CD45RB/CD11c on DCs. However, there was no significant difference among DCGC0, DCGC1, and DCGC3 in the proportion of these DC subpopulations (Fig. 3,A and data not shown). We further examined the cytokine profiles in CD8α+ and CD8α− DC subpopulations. When CD8α+ DCGC0 were stimulated, they produced a higher amount of IL-12 than did CD8α− DCGC0, whereas CD8α− DCGC0 produced a higher amount of IL-10 than did CD8α+ DCs (Fig. 3,B). The IL-12 production was enhanced in CD8α+ DCGC1, but that of DCGC3 was significantly decreased (Fig. 3,B). In contrast, IL-10 production of CD8α− DCGC3 was significantly enhanced, but that of DCGC1 was hardly changed (Fig. 3,B). Taken together with FACS profiles shown in Fig. 2 A, the results strongly suggest that the α-GalCer injections substantially changed the ability to produce IL-12 and IL-10 in CD8α+ and CD8α− DCs, respectively, without significant change in the proportion of DC subpopulations.
Signal molecules responsible for changes in cytokine profiles in DCGC3
Next, we examined the intracellular signaling pathway responsible for the characteristic cytokine production in DCGC3. By Western blot analysis, we found that phosphorylation of ERK1/2, which is known to be important for IL-10 production (32), but not of p38 was more rapidly and strongly induced in DCGC3 upon CpG ODN stimulation as compared with induction in DCGC0 and DCGC1 (Fig. 4,A). Consistently, IL-10 production from DCGC3 was significantly suppressed by a specific inhibitor of ERK1/2 in a dose dependent fashion (Fig. 4 B, left), indicating that a signal pathway through ERK1/2 is responsible for the elevated level of IL-10 production in DCGC3.
We then analyzed the NF-κB signaling pathway, which is known to be important for IL-12 production (33), and found that it was also changed in DCGC3. The EMSA demonstrated that the nuclear DNA binding of NF-κB after TLR stimulation was significantly inhibited in DCGC3 (Fig. 4,C, top). However, phosphorylation levels of IKKαβ were not significantly altered (Fig. 4,C, bottom), suggesting that molecular components different from those in the authentic NF-κB signaling cascade are involved in DCGC3. As shown in Fig. 4 D, the expression of IκBNS, which directly binds NF-κB and inhibits its activity (34), was dramatically augmented in DCGC3 compared with other DCs. These results strongly suggest that down-modulation of IL-12 production in DCGC3 is due to the enhancement of IκBNS.
Prevention of EAE development by DCGC3
We used EAE to investigate whether DCGC3 have a capacity to prevent in vivo inflammatory immune responses. Syngeneic MOG35–55-pulsed DCGC0, DCGC1, or DCGC3 were i.v. injected into naive B6 mice, and then EAE was induced (Fig. 5,A). The onset and mean disease scores were not significantly different between mice pretreated with MOG35–55-pulsed DCGC0 or DCGC1 (Fig. 5,B). In striking contrast, mice pretreated with MOG35–55-pulsed DCGC3 showed delayed onset and suppressed EAE development (Fig. 5,B). These results clearly indicate that DCGC3 have the regulatory capacity to prevent EAE development. When IL-10 KO mice were used as the source of DCGC3, the suppression of EAE was abrogated (Fig. 5 B), indicating that production of IL-10 from the transferred DCs is critical for the amelioration of EAE.
To investigate the involvement of cell types other than DCGC3 in the suppression of EAE development, we isolated CD4 T cells from EAE hosts, stimulated them in vitro with MOG35–55-pulsed DCs, and analyzed their cytokine production by FACS. The intracellular staining shown in Fig. 5,C indicates that CD4 T cells from EAE-suppressed mice pretreated with DCGC3 generated more IL-10 and less IFN-γ producers compared with those from EAE-nonsuppressed mice. These results indicate that pretreatment with MOG-pulsed DCGC3 induces MOG-reactive IL-10-producing CD4 T cells. The proliferative response against MOG35–55, but not OVA, of CD4 T cells from EAE-suppressed mice was significantly lower than that of those from EAE-nonsuppressed mice (Fig. 5,D). However, the reduced proliferative response was completely abrogated by the addition of an anti-IL-10R mAb (Fig. 5,D). Therefore, IL-10 from CD4 T cells seems to act as a key molecule suppressing pathogenic CD4 T cells. Consistently, the transfer of MOG35–55-pulsed DCGC3 into IL-10 KO hosts had no suppressive effect (Fig. 5 E), indicating a requirement for IL-10-producing host cells as effector cells for EAE suppression.
Stability of the regulatory properties induced in DCGC3
Finally, we examined the stability of the observed alterations in phenotypes and cytokine profiles in DCGC3. Mice were left 10–30 days after repeated α-GalCer injections, and rechallenged with α-GalCer in vivo before measuring their phenotypes and cytokine profiles. After 10 days, the expression level of CD86 on DCGC3 was almost the same as that of DCGC0 and, at 30 days, gradually increased, but was still found to be significantly lower than that of DCGC1 (Fig. 6,A). When DCGC3 after a 30-day interval were stimulated with CpG ODN, they still showed very low IL-12 but high IL-10 production (Fig. 6 B). Therefore, changes in surface phenotypes and cytokine profiles in DCGC3 are not transient but appear to persist for at least 1 mo.
In the present study, we analyzed the cellular and molecular mechanisms of the induction of regulatory DCs by Vα14 NKT cell activation in vivo. Repeated stimulation of Vα14 NKT cells with α-GalCer led to change in their cytokine profile, producing IL-10 but not IFN-γ despite their potential ability to produce large amounts of both IL-10 and IFN-γ under naive physiological conditions (Fig. 2,B). Interestingly, these Vα14 NKT cells induced regulatory properties in DCs characterized by immature phenotypes of cell surface molecules (Fig. 1), high IL-10 and low IL-12 production (Fig. 2,A), as well as an ability to suppress in vivo immune responses (Fig. 5). We believe that these findings contribute to explaining precise cellular and molecular mechanisms for the regulatory immune responses mediated by Vα14 NKT cells.
Although it has been shown that Vα14 NKT cells did not lose their IFN-γ production even after in vivo α-GalCer treatment (35), we detected significant loss of IFN-γ in purified Vα14 NKT cells isolated from mice that received three α-GalCer injections (Fig. 2,B). We also found that in vivo-stimulated Vα14 NKT cells retained their ability to produce IL-10, which is also responsible for acquisition of regulatory properties and high IL-10 production by DCs (Fig. 2,C). In fact, the Vα14 NKT cell-induced regulatory DCs producing lower IL-12 and higher IL-10 showed reduced nuclear DNA-binding activity of NF-κB (Fig. 4 C). These data are in part in agreement with previous findings that the exposure of freshly isolated DCs to rIL-10 inhibits the nuclear DNA-binding activity of NF-κB and suppresses IL-12 production (33, 36). It is also in agreement with the findings by Sonoda et al. (13) that IL-10 but not IL-4 from Vα14 NKT cells was crucial in the induction of anterior chamber-associated immune deviation, as Vα14 NKT cells from wild-type or IL-4 KO but not IL-10 KO mice reconstituted the ability to induce anterior chamber-associated immune deviation in Vα14 NKT KO mice. Although it is still possible that IL-10 derived from other cells, including DCs themselves, modulates DC functions, the first trigger promoting the regulatory circuit after multiple α-GalCer injections seems to be provided by IL-10-producing Vα14 NKT cells. Because an importance of IL-4 in the protection of EAE has been reported (14, 15), IL-4 might play a role in the NKT cell-mediated immune regulatory cascade. However, we do not address this issue in the present experiments.
Several fractions of DC with regulatory properties expressing particular cell surface markers have been reported, such as CD8α+ or CD8α− DCs, B220+ plasmacytoid DCs, and CD45RBhigh IL-10-producing DCs (27, 36, 37, 38, 39, 40, 41). Thus, we examined the cell surface markers and whether any particular DC subsets predominantly expanded after repeated α-GalCer injections. However, in vivo α-GalCer treatment did not significantly change the proportion of subsets of DCs (Fig. 3,A), arguing that the changes in cytokine profile of CD11c+ DCs simply mirrored the changes in their ability of cytokine production. In fact, the IL-12 production of CD8α+ DC was decreased, whereas IL-10 production of CD8α− DC was intensified by in vivo α-GalCer treatments (Fig. 3 B). This is partially in line with the finding by Naumov et al. (27) that CD8α− DCs were accumulated in the pancreatic lymph nodes of NOD mice after α-GalCer treatment. Therefore, although they have shown the down-regulation of IL-12 after α-GalCer stimulation but not the up-regulation of IL-10 production by the pancreatic DCs, it is possible that the in vivo α-GalCer treatment rendered the DCs to produce IL-10 and suppress the type 1 diabetes.
Concerning the changes in cytokine profiles in DCs after α-GalCer stimulation, it is possible that after maturation of DCs with a single injection of α-GalCer, most of mature DCs undergo apoptosis, and thus newly derived immature DCs are affected by the pre-existing activated Vα14 NKT cells producing high IL-10 and low IFN-γ. In fact DCs, which receive maturation stimuli, are subject to apoptosis as reported (42), although Vα14 NKT cells are rather resistant to apoptosis upon TCR-mediated stimulation by up-regulating antiapoptotic genes (43, 44, 45). Therefore, activated Vα14 NKT cells with IL-10-shifted cytokine production survive for a long time and may affect the function of DCs to be tolerogenic. In fact, the data shown in Fig. 2 C support this notion that IL-10 from Vα14 NKT cells could change DCs to become IL-10 -producing tolerogenic cells. The regulatory DCs (DCGC3) induced by Vα14 NKT cells have different patterns of cytokine production and phosphorylation of signal molecules from those of naive DCs, thus it is unlikely that DCGC3 are merely newly derived, naive DCs even if they show similar cell surface phenotypes. However, it is still possible that survivors from the apoptosis of mature DCs become DCGC3.
The molecular mechanisms important for the acquisition of regulatory function in DCs appear to involve two different signal cascades. Phosphorylated ERK1/2 in Vα14 NKT cell-induced regulatory DCs is up-regulated, indicating that the high IL-10 production is tightly regulated by a MAPK-dependent signal (Fig. 4). Furthermore, the nuclear DNA-binding activity of NF-κB molecules important for IL-12 production is significantly reduced, whereas the expression of IκBNS mRNA is greatly augmented (Fig. 4). Because IκBNS binds directly to NF-κB, resulting in the inhibition of the translocation of NF-κB into the nucleus, and suppresses NF-κB activity (34), the down-modulation of IL-12 and enhanced IL-10 production are likely to be due in part to the augmented expression of IκBNS.
As to the effector mechanisms in Vα14 NKT cell-mediated regulation, the Vα14 NKT cell-induced regulatory DCs prevented autoimmune disease development only by the generation of IL-10-producing CD4 regulatory T cells (Fig. 5, C--E). IL-10 derived from the regulatory DCs is essential for the generation of effector type CD4 regulatory T cells because the regulatory DCs from IL-10-deficient mice failed to suppress EAE development (Fig. 5,B). IL-10 derived from CD4 regulatory T cells seems to be a final effector molecule because an anti-IL-10R mAb abrogated the inhibitory activity mediated by CD4 regulatory T cells (Fig. 5,D), and also because the regulatory DCs failed to generate functional CD4 regulatory T cells when they were transferred into IL-10 KO mice (Fig. 5 E). Thus, IL-10-producing CD4 regulatory T cells generated by the regulatory DCs are T regulatory type 1-like T cells (46, 47).
In the Vα14 NKT cell-mediated regulatory cascade, cell interactions seem to occur in three steps: IL-10 produced by Vα14 NKT cells induces regulatory DCs, which in turn generate CD4 regulatory T cells that suppress immune responses. In each step, IL-10 is essential. IL-10 from Vα14 NKT cells and from regulatory DCs is necessary to generate regulatory cell types, but does not act as a final effector molecule, suggesting IL-10 in the induction phase works as a soluble mediator modulating or affecting target cells to be regulatory. In contrast, IL-10 from CD4 regulatory T cells acts as an effector molecule mediating the suppression of T cell responses in an Ag-specific fashion. IL-10 in the final step of Vα14 NKT cell-mediated regulation renders effector T cells unresponsive, probably during Ag recognition processes in which cognate interactions among APCs, effector T cells, and CD4 regulatory T cells take place. Therefore, IL-10 from Ag-specific CD4 regulatory T cells suppresses only specific T cell responses, although IL-10 itself mediates nonspecific suppression. These findings suggest a novel regulatory mechanism mediated by Vα14 NKT cells through the control of DC function in vivo, and may shed new light of the regulation of immune responses.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by Grant-in Aid for Scientific Research from the Ministry of Education, Culture, Sport, Science and Technology of Japan.
Abbreviations used in this paper: α-GalCer, α-galactosylceramide; EAE, experimental allergic encephalomyelitis; DC, dendritic cell; IKK, IκB kinase; MOG, myelin oligodendrocyte glycoprotein; KO, knockout; ODN, oligodeoxynucleotide.