Abstract
NKT cells in donor bone marrow (BM) have been demonstrated to protect against graft-vs-host disease (GVHD) following BM transplantation. Murine NKT cells are divided into two distinct subsets based on the invariant Vα14Jα18 TCR expression. However, details of the subset and mechanisms of the BM NKT cells involved in suppressing GVHD have not been clarified. Irradiated BALB/c or C3H/HeN mice administered B6 or Jα18−/− BM cells show attenuation of GVHD, whereas recipients given CD1d−/− BM cells did not show attenuation. Moreover, coinjection of BM non-Vα14Jα18 CD1d-restricted (type II) NKT cells and CD1d−/− BM cells suppressed GVHD, whereas coinjection of BM Vα14Jα18 TCR (type I) NKT cells did not. These protective effects on GVHD depended upon IFN-γ-producing type II NKT cells, which induced the apoptosis of donor T cells. The splenocytes of mice administered BM cells from B6.IL-4−/− or Jα18−/−IL-4−/− mice produced lower levels of IL-4 and IL-10 than the splenocytes of mice transplanted with BM cells from B6, B6.IFN-γ−/−, Jα18−/−, or Jα18−/−IFN-γ−/− mice. Taken together, our results show that IFN-γ-producing BM type II NKT cells suppress GVHD by inducing the apoptosis of donor T cells, while IL-4-producing BM type II NKT cells protect against GVHD by deviating the immune system toward a Th2-type response.
Natural killer T cells are characterized by the coexpression of surface markers for both conventional αβ T cells and NK cells (1). NKT cells regulate suppressive responses in autoimmune diseases, pulmonary fibrosis, oral tolerance, transplantation rejection, and graft-vs-host disease (GVHD)3 (1, 2, 3, 4, 5, 6, 7), and they enhance immune responses to tumors, microorganisms, and inflammation (8, 9, 10). Mouse NKT cells are heterogeneous and can be divided into two distinct subsets based on the patterns of expression of the invariant Vα14Jα18 TCR (11). The first murine NKT cell subset has a rearrangement of the variable region Vα14 to the joining region Jα18, and the cells of this subset are referred to as invariant NKT (iNKT) or type I NKT cells (12). These cells are restricted for the nonclassical class I molecule CD1d and have strong responses to the α-galactosylceramide (α-GalCer) presented by CD1d (13, 14, 15). The second subset of murine NKT cells is also CD1d restricted but expresses a greater variety of TCRs; these cells referred to as type II NKT cells (11). Therefore, it has been suggested that these two subsets of NKT cells that express different TCRs may play distinct functional roles in immune responses, depending on the immunological environment. Nevertheless, most of the studies conducted to date on NKT cells have focused on investigating the biological functions of type I rather than type II NKT cells. It has been demonstrated that type I NKT cells play critical roles in regulating various immune responses, whereas the distinct functional roles of the type II NKT cell subset have not been well characterized in terms of regulating immune responses. Interestingly, several recent studies have demonstrated that type II NKT cells have critical functions in regulating tumor surveillance, diabetes, and inflammation (16, 17, 18, 19). These studies indicate that type II NKT cells have distinct functions in various immune responses, which may provide useful information for understanding the various functional roles of NKT cells in the in vivo system.
It has been demonstrated that donor-derived NKT cells in bone marrow (BM) suppress acute GVHD following BM transplantation (BMT) (6) and that BM NKT cells from IL-4−/− mice are unable to prevent against GVHD, which suggests that BM NKT cells regulate GVHD in an IL-4-dependent manner. However, it is not clear whether the two types of donor-derived BM NKT cells exert similar functional roles in protecting against GVHD. Moreover, the mechanisms by which NKT cells suppress GVDH following BMT have not been investigated. Therefore, we investigated: 1) the two types of BM NKT cells to discern which NKT cell type contributes to the protection against GVHD; and 2) how BM NKT cells attenuate GVHD following BMT. In this report we demonstrate that non-Vα14Jα18 CD1d-restricted (type II) NKT cells in donor BM cells suppress GVHD in allogeneic BMT by producing IFN-γ and IL-4, which contribute to the induction of apoptosis of donor-derived T cells and derive Th2 immune responses, respectively.
Materials and Methods
Mice
CD1d−/− (C57BL/6 background) mice were obtained from the National Institute of Allergy and Infectious Diseases/Taconic Farms facility of the National Institutes of Health (Dr. H. Gu), and Jα18−/− (H-2b) and RAG−/−Vα14tgVβ8.2tg (H-2b) (where tg is transgenic) mice were generous gifts from Dr. M. Taniguchi (Chiba University, Chiba, Japan). The B6.IFN-γ−/− (H-2b) and B6.IL-4−/− (H-2b) mice were purchased from The Jackson Laboratory, and B6.lpr.lpr (H-2b) mice were purchased from Japan SLC. C57BL/6 (H-2b), BALB/c (H-2d), and C3H/HeNCrljBgi (H-2K) mice were purchased from the Orient Company. Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−IFN-γ−/−, and CD1d−/−IL-4−/− mice were generated by crossing Jα18−/− or CD1d−/− mice, and IFN-γ−/− or IL-4−/− mice. Nine- to- 12-wk-old mice were used in all of the experiments. The mice were bred and maintained under specific-pathogen-free conditions in the Clinical Research Institute of Seoul National University Hospital, Seoul, South Korea. All animal experiments were approved by the Institutional Animal Care and Use Committee of the Clinical Research Institute of Seoul National University Hospital.
Induction of GVHD
The day before BMT, the recipient BALB/c or C3H/HeN mice received 8 gray (Gy) (BALB/c) or 12 Gy (C3H/HeN) of total body irradiation (137Cs source) split into two doses separated by 3 h and were injected i.v. with 2 × 105 (in BALB/c recipient mice) or 1 × 106 (in C3H/HeN recipient mice) nylon wool-purified splenocytes and 3 × 106 BM cells or 3 × 106 T cell-depleted (TCD) BM cells. T cell depletion was conducted using MACS (Miltenyi Biotec) separation with CD90 (Thy-1.2) MicroBeads. In some experiments the recipient mice were injected i.v. with 3 × 106 T cell-depleted BM cells from B6.IL-4−/− or B6.IFN-γ−/− mice plus 2 × 105 BM T cells enriched from B6, Jα18−/−, or CD1d−/− mice plus 2 × 105 splenocytes on day 0. For evaluation of the protective effects of BM type I NKT cells on GVHD, CD1d−/− BM plus various numbers of NKT cells enriched from RAG−/−Vα14tgVα8.2tg mice plus 2 × 105 splenocytes were transplanted into irradiated BALB/c mice on day 0. To enrich type I NKT cells in BM, BM cells from RAG−/−Vα14tgVα8.2tg mice were stained using PE-conjugated anti-NK1.1 and CyChrome-conjugated anti-TCR-β mAbs, and NK1.1+TCR-β+ NKT cells were sorted using FACStar Plus and CellQuest software (version 3.3; BD Biosciences). The purity of the sorted cells was >95%.
Assessment of GVHD
The survival rate and degree of systemic GVHD were monitored daily after BMT. Blood collected from the tail 7 and 14 days after BMT was used for measuring the serum glutamic-oxaloacetic transaminase (GOT) and glutamic-pyruvic transaminase (GPT) levels with an autoanalyzer (Antech Diagnostics). Some of the sera were used for measuring the TNF-α levels by ELISA.
ELISA
Spleens and blood samples were obtained from recipient mice 7 or 14 days post-BMT. The isolated spleen cells (2 × 105/wells) were stimulated with an anti-CD3 mAb (0.25–1.0 μg/ml) plus an anti-CD28 mAb (1.0 μg/ml) in flat-bottom 96-well plates for 48 h. The amounts of IFN-γ, IL-4, and IL-10 in the culture supernatants and the TNF-α levels in the sera were determined by ELISA (BD Pharmingen).
Abs and flow cytometry
FITC-, PE-, or allophycocyanin-conjugated mAbs against mouse CD4, CD8α, NK1.1, TCR-β, annexin V, H-2Db, H-2Dd, 7-aminoactinomycin D (7-AAD), and CD1d dimer (BD Pharmingen) were used. For flow cytometric analysis, cells were preincubated with 2.4G2 mAbs to block FcγR and then incubated with the relevant mAbs for 30 min at 4°C. The cells were then washed with PBS, fixed with 1% paraformaldehyde in PBS, and analyzed by flow cytometry in the FACSCalibur apparatus using the CellQuest software (BD Biosciences).
Real-time PCR analysis
For quantitative real-time PCR, total RNA was isolated from homogenates of the spleen, liver, and Peyer’s patches using the RNeasy kit (Qiagen) according to the manufacturer’s instructions. The RNA was reverse transcribed with Moloney murine leukemia virus reverse transcriptase Taq polymerase (Koschem) before PCR. The VIC-fluorescent GAPDH, FAM-fluorescent IFN-γ, and IL-4 primer/probe mixtures manufactured by Applied Biosystems were used for performing real-time PCR. The reactions were performed in triplicate. Briefly, 1 μg of cDNA was amplified in the presence of the TaqMan Universal Master Mix (PerkinElmer Biosystems), the gene-specific TaqMan primer/probe mixture, and water. The gene-specific PCR products were analyzed using the Applied Biosystems 7500 sequence detection system (PerkinElmer Biosystems), and the results for each cytokine were normalized in relation to GAPDH expression as an endogenous control.
Statistical analysis
Survival curves were plotted using the Kaplan-Meier estimates and compared using the log-rank test. Statistical analysis of one-way ANOVA was performed using the Prism3.0 program. Statistical significance was determined using the Tukey-Kramer test and the value of p < 0.05 was considered to be statistically significant. The results are expressed as mean ± SEM.
Results
Donor bone marrow type II NKT cells suppress GVHD in allogeneic BMT
To analyze the subpopulations of NKT cells in the BM, B6, Jα18−/−, and CD1d−/− BM cells were stained with the anti-TCRβ and anti-NK1.1 mAbs or for the α-GalCer/CD1d dimer (Fig. 1⇓A). Type I NKT cells were detected based on the expression of Vα14Jα18 TCR, which binds the α-GalCer/CD1d dimer, whereas the whole population of NKT cells (type I and type II NKT cells) was identified by the expression of NK1.1 and TCRβ. In B6 BM cells, the fraction of NK1.1+TCRβ+ NKT cells was larger (0.63%) than that of CD1d/α-GalCer dimer-positive type I NKT cells (0.35%), which indicates that B6 BM cells contain type I and type II NKT cells. In contrast, type I NKT cells were not found in the Jα18−/− BM, and neither type I nor type II NKT cells were detected in the CD1d−/− BM. Unlike the BM NKT cells, the percentages (Fig. 1⇓B) and cell numbers of CD4+ and CD8+ T cells and NK cells in the CD1d−/− or Jα18−/− BM were similar to those in the B6 BM. Donor-derived BM NKT cells, which can be divided into two distinct subsets, suppress GVHD (6). Therefore, to determine which subset of NKT cells contributes to protecting against GVHD in BMT, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into lethally irradiated BALB/c mice. All of the mice that were administered B6 splenocytes and BM cells from B6, Jα18−/−, or CD1d−/− mice showed signs of GVHD (diarrhea, weight loss, ruffled fur, and hunched posture) shortly after BMT. It is well known that the levels of hepatic enzymes and TNF-α in the serum reflect the severity of GVHD in mice (20). Therefore, to evaluate the severity of GVHD in these recipient mice, the serum hepatic enzyme and TNF-α levels in the recipient mice were measured 7 and 14 days post-BMT (Fig. 2⇓, A and B). The serum levels of hepatic enzymes (GOT and GPT) and TNF-α were elevated in recipient mice that received CD1d−/− BM cells 7 and 14 days post-BMT, while BALB/c mice administered B6 or Jα18−/− BM cells showed low levels of serum GOT, GPT, and TNF-α. (Fig. 2⇓, A and B) Moreover, BALB/c mice that were administered CD1d−/− BM cells died within 45 days, whereas most of recipient mice that were transferred with B6 or Jα18−/− BM cells survived for >60 days (Fig. 2⇓C). These findings demonstrate that B6 and Jα18−/− BM cells significantly suppress GVHD in recipient BALB/c mice, while transplanted CD1d−/− BM cells do not protect against GVHD in recipient BALB/c mice. In contrast, Jα18−/− or CD1d−/− splenocytes did not alter the survival rates of recipient BALB/c mice that were transplanted with B6 BM cells as compared with recipient mice that were coinjected with B6 splenocytes and BM cells, which suggests that donor-derived peripheral NKT cells do not contribute to the suppressive effects on GVHD (data not shown). To investigate whether B6 or Jα18−/− BM cells contribute to inhibiting GVHD in other donor recipient strain combinations, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into lethally irradiated C3H/HeN mice. Consistent with results in a B6→BALB/c mouse combination, B6 or Jα18−/− BM cells protected C3H/HeN recipient mice from GVHD whereas CD1d−/− BM cells did not. (Fig. 2⇓D) These results demonstrate that B6 or Jα18−/− BM cells inhibit GVHD in various strains of recipients. In addition, to confirm the durable engraftment of donor BM cells, splenocytes from recipient BALB/c mice transferred with B6, Jα18−/−, or CD1d−/− BM cells were taken and stained with anti-H-2Db and Dd mAbs. Most of the spleen cells (>90%) of recipient BALB/c mice were donor (B6 mice)-derived, H-2Db-positive cells 90 days after BMT (Fig. 2⇓E), indicating that donor cells were durably engrafted in recipient lymphoid tissues. Taken together, these results suggest that BM type II NKT cells contribute to the suppression of GVHD after BMT.
Subsets of NKT cells in the BM cells of B6, Jα18−/−, and CD1d−/− mice. A, B6, Jα18−/−, and CD1d−/− BM cells were stained with anti-TCR-β and anti-NK1.1 mAbs (upper panel) or anti-TCR-β and for the α-GalCer/CD1d dimer (bottom panel) and analyzed by flow cytometry. The numbers in the flow cytometry diagrams represent the percentages for each quadrant. The results shown are representative of five independent experiments. B, B6, Jα18−/−, and CD1d−/− BM cells were stained with anti-CD4, anti-CD8, anti-TCR-β, or anti-NK1.1 mAb and analyzed by flow cytometry. The data shown are the means ± SEM of five mice in each group. p > 0.05, one-way ANOVA; n.s., not statistically significant.
B6 or Jα18−/− BM cells protect against GVHD but CD1d−/− BM cells do not attenuate GVHD following allogeneic BMT. A and B, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into BALB/c or C3H/HeN mice that were irradiated lethally over the whole body (BALB/c, 400 centigray (Gy) × 2; C3H/HeN, 600 cGy × 2). Sera were taken from the recipient BALB/c mice 7 and 14 days after transplantation of the BM cells and splenocytes and analyzed for levels of GOT and GPT (A) or TNF-α (B). The filled bar in B represents the TNF-α levels 7 days post-BMT and the gray bar represents the TNF-α levels 14 days post-BMT. a*, p < 0.05 for 7 days post-BMT; b**, p < 0.001 for 14 days post-BMT; n = 6 per group; n.s. not statistically significant. C and D, The survival rates of recipient BALB/c mice (C, p < 0.001, log-rank test, n = 17 per group) and C3H/HeN (D, p < 0.05, log-rank test, n = 8 per group) were monitored after the transplantation of BM cells and splenocytes. E, Splenocytes were obtained from recipient BALB/c mice transferred with B6, Jα18−/−, or CD1d−/− BM cells and stained with anti-H-2Db and Dd mAb 90 days after BMT. The diagram presents the percentage of H-2Db+ cells among splenocytes. (n = 3). Syn., Syngeneic control.
To address this suggestion, we coinjected CD1d−/− BM cells and BM type I NKT cells into BALB/c recipient mice. The type I NKT cells were enriched from RAG−/−Vα14tgVα8.2tg BM cells that contains Vα14i TCR NKT cells in the absence of type II NKT cells and conventional αβ T cells. The administration of a large number of BM Vα14i TCR NKT cells (3 × 105 cells) did not suppress GVHD in BALB/c recipient mice that were administered CD1d−/− BM cells and B6 splenocytes (Fig. 3⇓A). In contrast, when we administered BM T cells that contained type II NKT cells enriched from Jα18−/− BM cells to BALB/c recipient mice that were transplanted with T cell-depleted CD1d−/− BM cells and B6 splenocytes, the GVHD in the recipients was suppressed (Fig. 3⇓B). These findings demonstrate that BM type II NKT cells specifically suppress GVHD in allogeneic BMT, whereas BM type I NKT cells are not able to protect against GVHD.
Donor BM type II NKT cells contribute to the attenuation of GVHD following BMT. For BMT, the recipient BALB/c mice were irradiated lethally over the whole body (400 cGy × 2). A, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into recipient BALB/c mice. In addition, CD1d−/− BM cells plus various numbers of BM type I NKT cells enriched from RAG−/−Vα14tgVα8.2tg cells were transplanted into BALB/c mice that were irradiated lethally and coinjected with B6 splenocytes. The survival rates of recipient BALB/c mice were monitored post-BMT. p < 0.0001, log-rank test, n = 12 per group. B, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into recipient BALB/c mice. CD1d−/− mouse T cell-depleted BM cells (3 × 106) plus Jα18−/− mouse BM T cells (2 × 105) were also transplanted into BALB/c mice that were irradiated lethally and coinjected with B6 mouse splenocytes. The survival rates of the recipient BALB/c mice were monitored post-BMT. p < 0.0001, log-rank test, n = 10.
Donor BM type II NKT cells contribute to the suppression of GVHD in allogeneic BMT by producing IL-4 and IFN-γ
To explore the mechanisms by which type II NKT cells contribute to the suppression of GVHD after BMT, we measured the mRNA levels of various cytokines in the spleen, liver, and Peyer’s patches from BALB/c mice that were administered B6, Jα18−/−, or CD1d−/− BM cells 7 days post-BMT. Among the cytokines tested, the mRNA levels of IFN-γ and IL-4 were significantly lower in the spleen, liver, and Peyer’s patches of BALB/c mice that had received CD1d−/− BM cells than in the organs of recipients administered B6 or Jα18−/− BM cells in the GVHD model (Fig. 4⇓A). Moreover, NKT cells secrete large amounts of IL-4 and IFN-γ upon activation, and these cytokines play critical roles in regulating the immune responses modulated by NKT cells (1, 21, 22). Therefore, these findings suggest that IFN-γ and IL-4 are involved in suppressing GVHD in recipient mice through BM type II NKT cells. To determine whether the IL-4 and IFN-γ produced by BM cells contributes to the suppression of GVHD in BMT, we coinjected B6, B6.IL-4−/−, or B6.IFN-γ−/− BM cells and/or B6, B6.IL-4−/−, or B6.IFN-γ−/− splenocytes into recipient BALB/c mice. The BALB/c mice that received B6.IFN-γ−/− BM cells and B6.IFN-γ−/− splenocytes died within 10 days post-BMT, whereas the mice that received B6.IFN-γ−/− BM cells and B6 splenocytes died within 35 days post-BMT. In contrast, all of recipients of BM cells and splenocytes from B6 mice survived for >60 days post-BMT, and most of the mice that received B6 BM cells and B6.IFN-γ−/− splenocytes also survived for >60 days post-BMT (Fig. 4⇓B, left panel). These findings suggest that IFN-γ-producing BM cells contribute to the attenuation of GVHD. The BALB/c mice that received B6.IL-4−/− BM cells and B6.IL-4−/− splenocytes died within 25 days post-BMT, and the mice that were transferred with B6.IL-4−/− BM cells and B6 splenocytes died within 35 days post-BMT. In contrast, 50% of mice that received B6 BM cells and B6.IL-4−/− splenocytes also survived >60 days post-BMT (Fig. 4⇓B, right panel), which indicates that IL-4-producing BM cells also contribute to the prevention of GVHD. Taken together, these findings suggest that the IL-4 and IFN-γ secreted from donor BM cells play more critical roles in protecting against GVHD after BMT than the IL-4 and IFN-γ produced by donor-derived splenocytes.
IL-4 and IFN-γ produced by donor BM type II NKT cells contribute to protecting against GVHD following BMT. For BMT, recipient BALB/c mice were irradiated lethally over the entire body (400 cGy × 2). A, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into recipient BALB/c mice. The spleens, livers, and Peyer’s patches were removed from these recipient BALB/c mice 7 days post-BMT. The IL-4 and IFN-γ mRNA levels were measured in these tissues using real-time PCR. The results shown are representative of five independent experiments. ∗, p < 0.05; ∗∗, p < 0.01; and ∗∗∗, p < 0.001. B, BM cells from B6, B6.IL-4−/−, or B6.IFN-γ−/− mice and splenocytes (SPC) from B6, B6.IL-4−/−, or B6.IFN-γ−/− mice were coinjected into BALB/c mice. The survival rates of the recipient BALB/c mice were monitored post-BMT. p < 0.001, log-rank test, n = 6 per group. C, B6, Jα18−/−, or CD1d−/− BM T cells and T cell-depleted (TCD) BM cells from B6, B6.IL-4−/−, or B6.IFN-γ−/− mice were coinjected into recipient BALB/c mice, which were also coinjected with B6 splenocytes. The survival rates of the recipient BALB/c mice were monitored post-BMT. p < 0.05 in left panel and p < 0.001 in right panel, log-rank test, n = 6 per group. D, BM cells from B6, B6.IFN-γ−/−, B6.IL-4−/−, Jα18−/−, CD1d−/−, Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− mice and B6 splenocytes were coinjected into recipient BALB/c mice. The survival rates of the recipient BALB/c mice were monitored post-BMT. p < 0.05, log-rank test, n = 8 per group.
To investigate the functional link between IL-4 or IFN-γ production and the GVHD-suppressive roles of BM type II NKT cells, B6, CD1d−/−, or Jα18−/− BM T cells were coadministered with T cell-depleted BM cells from B6.IL-4−/− or B6.IFN-γ−/− mice into BALB/c recipients for the GVHD model. Transplantation of B6 or Jα18−/− BM T cells increased the survival rate of the BALB/c recipients of T cell-depleted BM cells from the B6.IFN-γ−/− or B6.IL-4−/− mice in the GVHD model, whereas CD1d−/− BM T cells did not alter the survival rate of the BALB/c mice that were administered T cell-depleted BM cells from the B6.IFN-γ−/− or B6.IL-4−/− mice (Fig. 4⇑C). Moreover, the survival rates were reduced in BALB/c mice that received B6 splenocytes and BM cells from B6.IFN-γ−/−, B6.IL-4−/−, Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− mice, compared with mice that were transferred with B6 and Jα18−/− BM cells (Fig. 4⇑D). These findings suggest that the BM type II NKT cells in B6 or Jα18−/− mice play critical roles in the suppression of GVHD through the production of IFN-γ and IL-4.
Donor BM type II NKT cells induce the apoptosis of donor-derived T cells by producing IFN-γ
Several studies have demonstrated that Th1-type cytokines such as IFN-γ and IL-18, regulate acute GVHD by inducing enhanced Fas-mediated apoptosis of donor T cells soon after BMT (23, 24, 25, 26). These findings suggest that IFN-γ-producing type II NKT cells might attenuate GVHD by inducing Fas-dependent apoptosis of donor T cells. To address this suggestion, we measured the numbers of donor-derived CD4+ and CD8+ T cells in the spleens of BALB/c mice that were transplanted with B6, Jα18−/−, or CD1d−/− BM cells 7 days post-BMT. The numbers of donor-derived CD4+ and CD8+ T cells in the spleens of mice transplanted with B6 or Jα18−/− BM cells were significantly lower than those in the spleens of mice transplanted with CD1d−/− BM cells 7 days post-BMT (Table I⇓). Given the difference in donor-derived T cell numbers between these recipients, it is possible that the donor-derived CD4+ and CD8+ T cells in mice transplanted with B6 or Jα18−/− BM cells were more susceptible to apoptosis as compared with those in the mice administered CD1d−/− BM cells. Alternatively, the donor-derived T cells in mice transferred with CD1d−/− BM cells were more highly activated and proliferated to a greater extent than did those in the B6 or Jα18−/− BM cell-transplanted mice. However, the numbers of divided donor-derived T cells from mice transferred with CD1d−/− BM cells were similar to those from recipients transplanted with B6 or Jα18−/− BM cells, as evaluated by CFSE labeling at 3 and 5 days post-BMT (data not shown). These findings suggest that the reduced numbers of donor-derived T cells in mice transplanted with B6 or Jα18−/− BM cells appear to be attributable to increased susceptibility to apoptosis. To determine whether the apoptosis of donor-derived immune cells is involved in the protection against GVHD afforded by BM type II NKT cells, we measured the apoptosis of the donor-derived T cells (H-2Db+TCRβ+7AAD−) by evaluating the annexin V expression levels. The percentages of apoptotic donor-derived T cells were lower in the BALB/c mice administered CD1d−/− BM cells than in the mice administered B6 or Jα18−/− BM cells (Fig. 5⇓A). Moreover, the percentages of donor-derived Fas+ T cells were lower in BALB/c mice administered CD1d−/− BM cells than in mice administered B6 or Jα18−/− BM cells (Fig. 5⇓B). These findings suggest that the apoptosis of donor-derived T cells contributes to BM type II NKT cell-mediated protection against GVHD. To address this suggestion, we coinjected B6, Jα18−/− or CD1d−/− BM cells and/or B6 or lpr/lpr mouse splenocytes into BALB/c mice. Coinjection of lpr/lpr mouse splenocytes and B6, Jα18−/−, or CD1d−/− BM cells into BALB/c mice did not protect against GVHD after BMT, whereas coinjection of B6 splenocytes and B6 or Jα18−/− BM cells suppressed GVHD (Fig. 5⇓C). These findings suggest that the BM type II NKT cells in B6 or Jα18−/− mice suppress GVHD by enhancing the apoptosis of donor-derived immune cells in a Fas-dependent manner.
IFN-γ produced by donor BM type II NKT cells enhances Fas-dependent apoptosis of donor-derived T cells in the GVHD model. For BMT, recipient BALB/c mice were irradiated lethally over the entire body (400 cGy × 2). A and B, B6, Jα18−/−, or CD1d−/− BM cells and B6 splenocytes were coinjected into recipient BALB/c mice. The spleen cells were taken from recipient BALB/c mice 7 days post-BMT, and the expression levels of annexin V and Fas were estimated by gating on 7-AAD−H-2Db+TCRβ+ cells in the flow cytometric analysis. The numbers in the diagrams represent the positive percentages for the indicated molecules. The results shown are representative of three independent experiments. C, B6, Jα18−/−, or CD1d−/− BM cells and B6 or lpr/lpr splenocytes (SPC) were coinjected into BALB/c mice. The survival rates of the recipient BALB/c mice were monitored post-BMT. p < 0.05, log-rank test, n = 10 per group. D, The expression of annexin V was evaluated by gating on H-2Db+TCR-β+7-AAD− T cells at 7 days post-BMT from the BALB/c recipient mice that were administered B6 splenocytes and BM cells from B6, B6.IFN-γ−/−, B6.IL-4−/−, Jα18−/−, CD1d−/−, Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− mice. ∗, p < 0.05; n.s., not statistically significant.
Numbers of donor T cells in the spleen of the recipient mice 7 days after BMT (×106)a
To determine which cytokine produced by the BM type II NKT cells induces the apoptosis of donor-derived T cells in GVHD, the levels of apoptosis of donor-derived T cells were measured in the BALB/c mice administered B6 splenocytes and BM cells from B6, B6.IFN-γ−/−, B6.IL-4−/−, Jα18−/−, CD1d−/−, Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− mice. The percentages of apoptotic donor-derived T cells were increased in the spleens of BALB/c recipient mice administered BM cells from B6, B6.IL-4−/−, Jα18−/−, and Jα18−/−IL-4−/− mice, whereas the percentages were reduced in mice transferred with B6.IFN-γ−/−, Jα18−/−IFN-γ−/−, CD1d−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− BM cells (Fig. 5⇑D). These findings suggest that BM type II NKT cells induce the apoptosis of donor-derived T cells by producing IFN-γ during the protection against GVHD post-BMT.
Donor BM type II NKT cells drive Th2-type immune responses in the GVHD model by producing IL-4
It has been established that NKT cells regulate immune responses through modulation of the Th1/Th2 balance in vivo by producing IL-4 and IFN-γ. Based on these regulatory functions of NKT cells in vivo, we postulated that BM type II NKT cells protect against GVHD by modulating Th1/Th2 immune responses in vivo. To address this hypothesis, we measured Th1/Th2 cytokine production by the splenocytes of the recipients in the GVHD model when these splenocytes were stimulated with anti-CD3 and anti-CD28 mAbs in vitro. The levels of IL-4 and IL-10 production by the splenocytes of mice transplanted with CD1d−/− BM cells were low compared with those of mice administered B6 and Jα18−/− BM cells (Fig. 6⇓A). Moreover, the splenocytes of mice administered B6.IL-4−/− or Jα18−/−IL-4−/− BM cells produced less IL-4 and IL-10 and more IFN-γ than those of mice transplanted with B6, B6.IFN-γ−/−, Jα18−/−, or Jα18−/−IFN-γ−/− BM cells. To rule out the possibility that the differences in cytokine production were attributable to the proportion of donor-derived CD4+ or CD8+ T cells in this culture condition, we investigated the percentage of H-2Db+TCRβ+ CD4 and CD8 cells in the splenocytes of recipient BALB/c mice 7 days after BMT. The percentages of H-2Db+TCRβ+ CD4 and CD8 cells in the splenocytes of recipient BALB/c were similar in each group 7 days after BMT, which was not statistically significant (Fig. 6⇓B). Therefore, these findings suggest that BM type II NKT cells induce Th2-type immune responses by producing IL-4.
IL-4 produced by donor BM type II NKT cells enhances the production of Th2 cytokines by splenocytes. For BMT, recipient BALB/c mice were irradiated lethally over the entire body (400 cGy × 2). BALB/c recipient mice were administered B6 splenocytes and BM cells from B6, B6.IFN-γ−/−, B6.IL-4−/−, Jα18−/−, CD1d−/−, Jα18−/−IFN-γ−/−, Jα18−/−IL-4−/−, CD1d−/−IFN-γ−/−, or CD1d−/−IL-4−/− mice. The splenocytes taken from these mice 7 days post-BMT were stimulated with anti-CD3 and anti-CD28 mAbs for 48 h. A, The amounts of IL-4, IL-10, and IFN-γ in the culture supernatants were measured by ELISA. B, The percentages of H-2Db+TCRβ+ CD4 and CD8 cells in the splenocytes of recipient BALB/c were measured. ∗, p < 0.05; ∗∗, p < 0.001; n.s. not statistically significant. The results shown are representative of three independent experiments; n = 3 per each group.
Discussion
Jα18−/− BM cells, which contain type II NKT cells but not type I NKT cells, attenuated GVHD in recipient BALB/c mice, whereas CD1d−/− BM cells, which lack both type I and type II NKT cells, did not attenuate GVHD. The adoptive transfer of type II NKT cells enriched from Jα18−/− BM cells attenuated GVHD in BALB/c mice that were transplanted with T cell-depleted CD1d−/− BM cells, whereas BM type I NKT cells enriched from RAG−/−Vα14tgVα8.2tg mice did not attenuate GVHD in these recipients. These findings indicate that BM type II rather than BM type I NKT cells contribute to protecting against GVHD after BMT. Thus, this is the first study to clarify the distinct functions of BM type II NKT cells in regulating GVHD. A recent study (27) supports our findings by demonstrating that the absence of donor Vα14i NKT cells in the graft does not affect GVHD-associated mortality.
Crowe and colleagues (28)have reported that CD4− liver-derived NKT cells are protective against tumor cells in vivo, whereas other NKT cell subsets such as CD4+ liver-derived NKT cells and CD4+ and CD4− NKT cells from the thymus or spleen show minimal suppressive effects on tumor cells, which suggests that organ-specific NKT cell subsets may be required for the immune regulation in anti-tumor surveillance. Moreover, it has also been suggested that NKT cells have type-specific distinct functions based on the expression patterns of the invariant TCR (16, 17, 29). In BMT, it has been demonstrated that both donor-derived BM and host-residual NKT cells exert protective functions on GVHD (6, 30, 31, 32). The adoptive transfer of invariant NKT cells or the injection of α-GalCer into recipient mice suppresses GVHD in BMT (30, 32), which indicates that, unlike BM NKT cells, type I NKT cells play critical roles in host-residual NKT cell-mediated suppression of GVHD. Combining these findings with ours, it is clear that the protective effects of donor-derived BM cells on GVHD are type II NKT cell-dependent, whereas the protective effects of residual-host cells show a type I NKT cell-dependent pattern. Therefore, our findings suggest that an organ-specific mechanism and a type-specific capability of NKT cells are involved in NKT cell-mediated regulation of GVHD in BMT.
In our study, the numbers of H-2Db+TCR-β+ donor CD4+ and CD8+ T cells in mice transplanted with B6 or Jα18−/− BM cells were lower than those in mice transferred with CD1d−/− BM cells. The numbers of divided donor-derived T cells, as estimated by CFSE intensity, were similar in three recipient groups in the GVHD model (data not shown), while the levels of apoptosis and the Fas expression levels of donor T cells were higher in mice that were transferred with B6 or Jα18−/− BM cells than in mice transplanted with CD1d−/− BM cells. Moreover, the coinjection of splenocytes from Fas-deficient lpr/lpr mice and B6 or Jα18−/− BM cells into BALB/c recipient mice did not protect against GVHD. These findings indicate that the differences in donor-derived T cell numbers and the protective effects on GVHD of BM type II NKT cells are attributable to Fas-dependent apoptosis of the donor T cells in these recipients. Several studies have demonstrated that Fas-dependent signals play a dominant role in the apoptosis of CD4+ T cells in vitro, whereas the role of Fas in CD8+ T cell apoptosis is negligible (33, 34). Based on these findings, it is likely that Fas-dependent signals induce the apoptosis of donor CD4+ T cells but not of donor CD8+ T cells in the GVHD model. However, the Fas-dependent pathway has been found to play a major role in the activation-induced cell death of both CD4+ and CD8+ donor T cells by up-regulating Fas and FasL on anti-host T cells in the GVHD model (35). Consistent with these findings, both CD4+ and CD8+ donor T cells were present in lower numbers in the spleens of mice transplanted with B6 or Jα18−/− BM cells than in the spleens of mice transferred with CD1d−/− BM cells.
It has been reported that donor-derived IFN-γ attenuates GVHD by inducing Fas-mediated activation-induced cell death due to the enhancement of Fas expression and the promotion of caspase-8 production (23, 24, 25). These findings suggest that IFN-γ produced by BM type II NKT cells contributes to protecting against GVHD by inducing the apoptosis of donor T cells. Consistent with this suggestion, the apoptosis of donor-derived T cells in BALB/c recipient mice was dependent upon IFN-γ-producing BM type II NKT cells in the GVHD model. However, the protective roles of donor-derived BM or host-residual NKT cells on GVHD have been reported to be IL-4 dependent (6, 30, 31, 32, 36). Moreover, Zeng and colleagues (6) suggested, by evaluating the IFN-γ/IL-4 production ratio in BM and peripheral blood T cells in BMT, that IFN-γ secreted by NK1.1+ T cells in the absence of IL-4 worsen GVHD. However, several lines of evidence in our study clearly demonstrate that IFN-γ-producing type II NKT cells in the donor BM contribute to the attenuation of GVHD in a Fas-dependent manner. First, the transplantation of B6 IFN-γ−/− or Jα18−/−IFN-γ−/− BM cells did not attenuate GVHD, whereas B6 or Jα18−/− BM cells suppressed GVHD. Second, the percentages of apoptotic donor-derived T cells were more reduced in BALB/c mice administered BM cells from B6.IFN-γ−/− or Jα18−/−IFN-γ−/− mice than those in mice transferred with BM cells from B6, Jα18−/−, or Jα18−/−IL-4−/− mice. Thus, it is intriguing that both IFN-γ-producing and IL-4-producing BM type II NKT cells contribute to the attenuation of GVHD in BMT through different immunological mechanisms.
Baker et al. (37) have reported that the adoptive transfer of in vitro expanded CD8+ NKT cells attenuates GVHD in MHC-mismatched BMT in a process that is related to IFN-γ production. These findings suggest that NKT cells contribute to the protection against GVHD by producing IFN-γ after BMT. However, the expansion of the CD8+ NKT cells was independent of CD1d, whereas type II NKT cells are CD1d-restricted during the activation process and show the major phenotype of CD4+ or CD4−CD8−. Therefore, it seems unlikely that these CD8+ NKT cells are the same as the BM type II NKT cells that attenuated GVHD by producing IFN-γ in our study. Collectively, this is the first report to demonstrate that IFN-γ-producing type II NKT cells in donor BM inhibit GVHD in MHC-mismatched BMT.
There have been several articles reporting that the transplantation of BM cells from IL-4 knockout the or Stat6 knockout mice attenuates GVHD (38, 39). These findings suggest that IL-4 produced by donor BM cells contributes to aggravating GVHD. However, the suppressive roles on GVHD of IL-4 produced by donor BM and host NKT cells have been demonstrated previously (6, 30, 31, 32). Multiple studies have demonstrated that Th2 cytokines contribute to reducing GVHD by Th2 polarization of the donor T cells (40, 41). Moreover, it has been suggested that residual host IL-4-producing NKT cells contribute to the protective effects on GVHD following BMT by modulating and polarizing the cytokine milieu to the Th2 type in the recipient mice (30, 31, 32). Consistent with these suggestions, the splenocytes of mice that were administered B6.IL-4−/− or Jα18−/−IL-4−/− BM cells produced lower levels of IL-4 and IL-10 than those of mice that were transplanted with B6, B6.IFN-γ−/−, Jα18−/−, or Jα18−/−IFN-γ−/− BM cells. Therefore, it seems likely that donor-derived, IL-4-producing BM type II NKT cells attenuate GVHD by polarizing the immune responses to the Th2 type. In ulcerative colitis in humans, type II NKT cells that produce IL-13 mediate Th2 responses and aggravate inflammation (19), which supports the notion that BM type II NKT cells contribute to the Th2 deviation of immune responses in GVHD. However, the protective effects were not detected in recipient mice transferred with IL-4-deficient or IFN-γ-deficient BM NKT cells alone, where IFN-γ- or IL-4-producing type II NKT BM cells would be successfully engrafted in these recipient mice, respectively. Based on these findings, it is suggested that IL-4 or IFN-γ alone from type II BM NKT cells might not be enough to protect GVHD. Therefore, both IL-4 and IFN-γ from type II BM NKT cells would be required simultaneously or at different time points for attenuating GVHD. Alternatively, it is also feasible that the need for one type or the other (IFN-γ- vs IL-4-producing BM type II NKT cells) might be overcome by injecting large number of these cells into recipients for protecting GVHD.
Although GVHD is unlikely to be controllable with a single agent, NKT cell-based therapeutic approaches should be considered for the prevention of GVHD following BMT. Based on our results and previous findings, differential therapeutic strategies are required for an NKT cell-based therapy to prevent GVHD, because an organ-specific mechanism and a type-specific capability of NKT cells are involved in NKT cell-mediated regulation of GVHD following BMT. Clinically, mature T cells in donor BM cells are depleted before the transplantation of BM cells into recipients for the prevention of GVHD. However, this strategy may be to the detriment of recipients, because the mature T cells in the BM contain type II NKT cells. Moreover, α-GalCer, a potent activating agent for NKT cells, is not an appropriate therapeutic agent for BM NKT cell-based therapy for GVHD because type II NKT cells are not restricted on α-GalCer/CD1d complexes (11). Therefore, a novel approach that specifically enhances the protective functions of BM type II NKT cells in vivo may be useful in the prevention of GVHD.
In conclusion, type II NKT cells from the donor BM attenuate GVHD in allogeneic BMT by producing IFN-γ and IL-4. With respect to the attenuation of GVHD, IFN-γ-producing BM type II NKT cells induce Fas-dependent apoptosis of donor CD4+ and CD8+ T cells, while IL-4-producing type II NKT cells deviate the allogeneic immune responses into the Th2 type.
Acknowledgments
We are grateful to Dr. Seong Hoe Park for his support of this project and to Dr. Sanghee Kim for providing α-GalCer. We also thank the members of the Department of Experimental Animal Research in Clinical Research Institute (CRI) of Seoul National University Hospital (SNUH).
Disclosures
The authors have no financial conflict of interest.
Footnotes
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.
↵1 This work was supported by Korean Science and Engineering Foundation Grant R01-2001-000-00194-0 and Korean Ministry of Science and Technology Grant (M10422010004-04N2201-00410. J.H.K. was supported by the Seoul fellowship.
↵2 Address correspondence and reprint requests to Dr. Doo Hyun Chung, Department of Pathology, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul, South Korea. E-mail address: doohyun{at}plaza.snu.ac.kr
↵3 Abbreviations used in this paper: GVHD, graft-vs-host disease; 7-AAD, 7-aminoactinomycin D; α-GalCer, α-galactosylceramide; BM, bone marrow; BMT, BM transplantation; cGy, centigray; GOT, glutamic-oxaloacetic transaminase; GPT, glutamic-pyruvic transaminase; Gy, gray; tg, transgenic.
- Received August 8, 2006.
- Accepted September 11, 2007.
- Copyright © 2007 by The American Association of Immunologists