CD1d-restricted Vα14+ invariant NK T (iNKT) cells are a specialized αβ T cell subset that regulates both innate and adaptive immunity. Although costimulatory molecules are required for the activation of conventional T cells and for the development of Foxp3+ T cells, their role in iNKT cell regulation is unclear. Here we report that mice deficient in CD80/CD86 and/or B7h exhibit severe defects in thymic iNKT cell maturation, associated with largely reduced iNKT cell number in the thymus and the periphery. We show that costimulation is necessary for the optimal expansion of postselected NK1.1 immature iNKT cells in the thymus and for the proper expression of the maturation markers T-bet and CD122. Surprisingly, costimulatory molecules on both hemopoietic and nonhematopoietic cells are required for iNKT cell development. Our results thus demonstrate a previously unknown function of costimulation in the intrathymic development of iNKT cells, distinct from that of conventional T cells and regulatory T cells.

Invariant NK T (iNKT)3 cells are an evolutionarily conserved subset of T cells in vertebrates that recognize glycolipid Ags presented by the nonclassical MHC molecule, CD1d (1, 2, 3). Although α-galactosylceramide (αGC) was originally identified as an Ag for iNKT cells (4), recent studies have extended their glycolipid Ags to those derived from microbes and self sources (5, 6, 7). Since discovery, iNKT cells have quickly emerged as crucial regulators of both innate and adaptive immunity. A malfunction or defect in this population has been reported to be associated with autoimmune diseases or tumors in animals and in humans (8, 9, 10).

iNKT cells develop in the thymus via a process distinct from that of conventional αβ T cells (11). Although iNKT cells arise from double-positive thymocytes, just like conventional αβ T cells, the positive selection of this population is mediated by CD1d-expressing CD4+CD8+ thymocytes through the recognition of self-glycolipids (7, 12, 13, 14). After positive selection, iNKT cells undergo a unique maturation process; starting out with the NK1.1CD44low phenotype, then into an NK1.1CD44high transitional stage and finally becoming NK1.1+CD44high mature iNKT cells (15). Accompanying this maturation, NK1.1 iNKT cells undergo extensive clonal expansion (16). The signals regulating postselection maturation and expansion are largely unknown. Recent studies have shown that the transcription factor T-bet, which is up-regulated after positive selection, is required for the maturation and effector function of iNKT cells (17, 18). However, the factors and cell types mediating T-bet up-regulation are not clear at this point.

Although iNKT cells do not appear to be positively selected by thymic epithelial cells (12), studies of bone marrow chimera with relB−/− mice have suggested that these cells might play a role in iNKT cell development distinct from positive selection (19). Another unique feature of iNKT cell development is their signaling requirement. For instance, mice defective in Fyn, SAP, NF-κB, and T-bet are defective in iNKT cells despite normal development of their conventional αβ T cell population (9, 18, 20, 21, 22).

Costimulatory molecules in the B7 superfamily play a pivotal role in the activation and function of conventional T cells, although their function in iNKT regulation is less clear. CD28 was reported to be important in the anti-tumor activity of iNKT cells (23). Interaction between ICOS and B7h is crucial for selective effector function of conventional T cells (24, 25); a recent study demonstrated that peripheral iNKT cells constitutively express ICOS and that this costimulator is required for iNKT cell function, including cytokine production and anti-tumor activity (26). In the thymus, although not required for conventional T cell development, B7.1/B7.2 (hereafter together referred as B7) have been shown crucial for the development and homeostasis of Foxp3+ regulatory T cells (27). However, a role for these costimulatory molecules in the development of iNKT cells has not been studied.

Here, we have examined the role of CD28/B7 and ICOS/B7h molecules in the development of iNKT cells. By using gene-knockout mice and various cell transfer studies, we show that B7 and B7h costimulation plays a fundamental role in regulating iNKT cell expansion and maturation.

C57BL/6, CD80/CD86 double-deficient (B7−/−), and T-bet−/−) mice were purchased from The Jackson Laboratory. RAG-2−/− and B6.SJL (CD45.1) mice were purchased from Taconic Farms. ICOS−/−, B7h−/−, and B7B7h−/− mice were generated and subsequently back-crossed as previously described (25). CD28−/− mice were kindly provided by A. Rudensky (University of Washington, Seattle, WA). All animal experiments were approved by our institutional committee on the use and care of laboratory animals.

Frozen thymic sections from C57BL/6 mouse were stained with anti-B7h Ab (gift of Dr. William Sha, University of California, Berkeley, CA) alone or in combination with anti-CD11c Ab (eBioscience) or anti-B7.2 Ab (BD Pharmingen). Biotin conjugated anti-hamster or anti-rat Abs were used as secondary Ab and avidin-peroxidase (Vector Laboratories) was applied for the following step. The red color was used as the first staining substrate and the blue color was used as the secondary staining substrate.

Allophycocyanin-conjugated anti-CD4 Ab, FITC-conjugated anti-CD19 Ab, Alexa 488-conjugated anti-CD8, FITC- or PE-conjugated anti-NK1.1 Ab, FITC-conjugated anti-CD122 Ab, PE-conjugated anit-CD44 Ab, PerCP-Cy5.5-conjugated anti-TCRβ Ab, PE-conjugated anti-CD28 Ab, PE-conjugated anti-ICOS Ab, PE-conjugated αGC-loaded CD1d-tetramer, or allophycocyanin-conjugated PBS57-loaded CD1d-tetramer were used for phenotypic analysis. For intracellular cytokine staining, lymphoid cells from the spleens of mice that received αGC (i.p.) were stained with FITC-conjugated anti-CD19 Ab together with PE-conjugated αGC-loaded CD1d-tetramer. These cells were permeabilized and further stained with allophycocyanin-conjugated anti-IL-4 Ab or anti-IFN-γ Ab. For intracellular T-bet staining, cells were permeabilized by using fix/permeabilizing buffer (eBioscience) and further stained with PE-conjugated anti-T-bet Ab (Santa Cruz Biotechnology) or mouse IgG1 isotype control. CD1d+TCRβ+ cells were gated and analyzed.

Bone marrow cells were harvested from either wild-type (WT), B7B7h−/−, or CD1d−/− mice and Thy1.2+ cells were depleted by anti-Thy1.2 Ab and rabbit complement (Cedarlane). After washing, cells (1 × 107/transfer) were i.v. transferred into lethally irradiated (10 Gy) WT or B7B7h−/− recipient mice. Ten to twelve weeks later, lymphoid cells from the thymus or spleen of the recipients were analyzed as described above.

The BrdU incorporation experiments in thymic iNKT cells were performed by using a BrdU kit (BD Bioscience) according to the manufacturer’s instructions. Briefly, mice were i.p. injected with 1 mg of BrdU solution. Four hours later, lymphoid cells from the thymus, liver, and spleen were stained for iNKT cells as described above before permeabilzed and further stained with FITC-conjugated anti-BrdU Ab.

For intrathymic injection, mice were anesthetized and the thorax area was exposed after a small skin incision. CD4+CD8NK1.1 thymocytes from CD45.1+ mice were sorted and injected into the thymus in a suspension of 20 μl of PBS (1.5 × 108/ml). The wound was closed with surgical glue, and the recipients were warmed until fully recovered. Seven days later, thymocytes of the recipient mice were analyzed.

All statistical values were assessed by the Student t test. p values were expressed and error bars are SE.

Constitutive expression of ICOS and CD28 on hepatic and splenic iNKT cells has been described (26). Flow cytometric analysis of thymic iNKT cells, identified by the αGC-loaded CD1d tetramer, also showed a constitutive expression of CD28 and ICOS on these cells in both NK1.1+ and NK1.1 population with higher ICOS expression on NK1.1- subset in mice of 2–3 wk of age (Fig. 1,A). The constitutive expression of ICOS/CD28 on thymic iNKT cells and their ligands in the thymus suggests a potential role for these molecules in the development of iNKT cells. To address this issue, we analyzed the percentage and total number of iNKT cells in the thymus, spleen, and liver of C57BL/6 (WT), B7h−/−, B7.1−/−, B7.2−/−, (B7−/−), or B7−/−B7h−/− (B7B7h−/−) (25) mice. The development of CD4+ T, CD8+ T, and B cells appeared normal in all groups (data not shown). By contrast, a significant reduction in the percentage and total number of thymic iNKT cells was observed in B7h−/− (>50% decrease) and B7−/− mice (>70% decrease) (Fig. 1, B and C). A more profound reduction in iNKT cells was observed in B7B7h−/− mice (>85% decrease). Compared with WT mice, the percentage of iNKT cells in the liver and spleen of these knockout mice was also significantly decreased (60–70% decrease in all cases), although no significant further reduction was observed in B7B7h−/− mice (Fig. 1, B and C). Consistent with these findings, a substantial decrease of thymic and splenic iNKT cells was also found in ICOS−/− and CD28−/− mice (Fig. 1 D).

FIGURE 1.

Defects in iNKT cell population in mice deficient of B7 and/or B7h costimulation. A, Thymocytes from WT mice (3 wk) were stained with CD1d-tetramer, anti-TCRβ Ab, and anti-NK1.1 Ab in combination with anti-CD28 and anti-ICOS Abs. The expression of CD28 and ICOS was analyzed on gated TCRβ+CD1d-tetramer+ cells. B and C, Lymphoid cells from the thymus and liver of the indicated strains of mice (6 wk) were stained with CD1d-tetramer plus anti-TCRβ Ab. Splenocytes were stained with CD1d-tetramer plus anti-CD19 Ab. Dot plots in A are representatives of three mice in each group. Data in C are the mean ± SE. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT. D, Lymphoid cells from the thymus and spleen of ICOS−/− or CD28−/− mice (6 wk) were analyzed as described above. Data are representative of three mice in each group. E, Indicated strains of mice (6–7 wk) were i.p. injected with αGC (2 μg/mouse). Lymphoid cells from the spleen were isolated 2 h later and analyzed for intracellular IL-4 and IFN-γ expression by flow cytometry. CD19CD1d-tetramer+ cells were gated and analyzed. Data are representative of three independent experiments.

FIGURE 1.

Defects in iNKT cell population in mice deficient of B7 and/or B7h costimulation. A, Thymocytes from WT mice (3 wk) were stained with CD1d-tetramer, anti-TCRβ Ab, and anti-NK1.1 Ab in combination with anti-CD28 and anti-ICOS Abs. The expression of CD28 and ICOS was analyzed on gated TCRβ+CD1d-tetramer+ cells. B and C, Lymphoid cells from the thymus and liver of the indicated strains of mice (6 wk) were stained with CD1d-tetramer plus anti-TCRβ Ab. Splenocytes were stained with CD1d-tetramer plus anti-CD19 Ab. Dot plots in A are representatives of three mice in each group. Data in C are the mean ± SE. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT. D, Lymphoid cells from the thymus and spleen of ICOS−/− or CD28−/− mice (6 wk) were analyzed as described above. Data are representative of three mice in each group. E, Indicated strains of mice (6–7 wk) were i.p. injected with αGC (2 μg/mouse). Lymphoid cells from the spleen were isolated 2 h later and analyzed for intracellular IL-4 and IFN-γ expression by flow cytometry. CD19CD1d-tetramer+ cells were gated and analyzed. Data are representative of three independent experiments.

Close modal

Upon stimulation, iNKT cells promptly produce a large amount of cytokines, such as IFN-γ and IL-4 (4). Injection of αGC into WT mice efficiently induced iNKT to produce IFN-γ and IL-4 within 2 h (Fig. 1,E). A weak but consistent reduction of IFN-γ- and IL-4-producing cells was observed in B7h−/− mice and B7−/− mice; the reduction of those cytokine-producing cells was more pronounced in B7B7h−/− mice (Fig. 1 E). Thus, iNKT cell number and cytokine expression are affected in mice lacking costimulatory molecules.

The above results indicate a previously uncharacterized role of CD28 and ICOS in the development and/or homeostasis of iNKT cells. To further characterize this regulation, we examined the expression of NK1.1 and CD44 on thymic iNKT cells as indicators of iNKT cell maturation (15). As expected, most iNKT cells in the thymus of WT mice displayed the NK1.1+CD44high mature phenotype. The percentage and absolute number of this population was significantly decreased in the thymus of B7−/− mice (Fig. 2, A and B). This was associated with an increase in the percentage of immature NK1.1 cells. Such phenotype was further magnified when analyzing iNKT cell developmental intermediates from the thymus of B7B7h−/− mice (Fig. 2, A and B). Nevertheless, the absolute numbers of the NK1.1-CD44low population remained comparable between WT and all knockout mice (Fig. 2,C). The proportion of CD4+ and CD4CD8 iNKT cells of B7B7h−/− appeared to be comparable to that of WT (Fig. 2,D). Of interest, despite their severe defect in thymic maturation, the NK1.1 expression on iNKT cells in the periphery of B7B7h−/− mice was largely comparable to that of WT mice (Fig. 2 E). Thus, it appears that in the absence of B7 and B7h costimulation, the thymic maturation of iNKT cells is defective.

FIGURE 2.

Analysis of maturation stage of thymic iNKT cells in mice deficient of B7 and/or B7h costimulation. AC, Thymocytes from the indicated strains of mice (6 wk, n = 3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-CD44 and anti-NK1.1 Abs. The expression of CD44 and NK1.1 was analyzed on gated TCRβ+CD1d-tetramer+ cells (A). The composition (B) and the absolute number (C) of each population was calculated and indicated. Data are mean ± SE. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT. D, Thymocytes from the indicated strains of mice (6 wk, n = 2–3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-CD4 Ab. The expression of CD4 was analyzed on gated TCRβ+CD1d-tetramer+ cells. E, Thymocytes, hepatic mononuclear cells, and splenocytes from the indicated strains of mice (6 wk, n = 2–3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-NK1.1 Ab. The expression of NK1.1 was analyzed on gated TCRβ+CD1d-tetramer+ cells. F, Lymphoid cells from the thymus of the indicated strains of mice (6 wk) were stained with CD1d-tetramer plus anti-TCRβ Ab in combination with anti-NK1.1 or anti-CD122 Ab, followed by intracellular staining with anti-T-bet Ab. The expression of T-bet was analyzed on gated TCRβ+CD1d-tetramer+ cells.

FIGURE 2.

Analysis of maturation stage of thymic iNKT cells in mice deficient of B7 and/or B7h costimulation. AC, Thymocytes from the indicated strains of mice (6 wk, n = 3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-CD44 and anti-NK1.1 Abs. The expression of CD44 and NK1.1 was analyzed on gated TCRβ+CD1d-tetramer+ cells (A). The composition (B) and the absolute number (C) of each population was calculated and indicated. Data are mean ± SE. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT. D, Thymocytes from the indicated strains of mice (6 wk, n = 2–3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-CD4 Ab. The expression of CD4 was analyzed on gated TCRβ+CD1d-tetramer+ cells. E, Thymocytes, hepatic mononuclear cells, and splenocytes from the indicated strains of mice (6 wk, n = 2–3) were stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-NK1.1 Ab. The expression of NK1.1 was analyzed on gated TCRβ+CD1d-tetramer+ cells. F, Lymphoid cells from the thymus of the indicated strains of mice (6 wk) were stained with CD1d-tetramer plus anti-TCRβ Ab in combination with anti-NK1.1 or anti-CD122 Ab, followed by intracellular staining with anti-T-bet Ab. The expression of T-bet was analyzed on gated TCRβ+CD1d-tetramer+ cells.

Close modal

T-bet is an important regulator of iNKT cell maturation and survival; enforced expression of T-bet in immature NK1.1 iNKT cells induces the expression of maturation markers, such as CD122 (17). CD44 and NK1.1 expression profiles of iNKT cells in B7B7h−/− mice resemble that of T-bet-deficient mice (18). Furthermore, our recent study showed that costimulatory signal is essential for T-bet expression in conventional T cells (25). Therefore, we asked whether costimulation has a similar effect in the induction of T-bet in iNKT cells. To this end, we analyzed T-bet protein level in iNKT cells of WT and different knockout mice. T-bet levels were significantly reduced in iNKT cells from B7h−/−, B7−/−, and B7B7h−/− mice compared with iNKT cells from WT mice (Fig. 2,F). More than 70% of thymic iNKT cells in WT mice expressed high level of T-bet, whereas only 10–15% of iNKT cells in B7B7h−/− mice expressed T-bet. Furthermore, T-bet expression directly correlated with NK1.1 and CD122 (IL-2/15 Rβ) expression in both WT and various knockout mice (Fig. 2 F). Taken together, these results indicate that thymic iNKT cells in mice lacking the costimulators have a defect in T-bet expression that correlates with the impaired NK1.1 and CD122 expression.

Because the number of iNKT cells is largely reduced in the periphery of mice lacking B7 and/or B7h, we also tested whether these molecules might be required for the proliferation of iNKT cells. CD8-depleted thymocytes (to enrich for iNKT cells) were isolated from the thymus of WT mice, labeled with CFSE, and i.v. injected into sublethally irradiated WT or B7B7h−/− mice (28). As shown in Fig. 3 A, iNKT cells underwent one to four divisions by day 7 in WT animals, in agreement with previous results (28). We observed no apparent defect in the division of transferred iNKT cells in B7B7h−/− recipients, suggesting that B7 and B7h costimulation is not required for the homeostatic proliferation of iNKT cells in the periphery.

FIGURE 3.

Lineage expansion of iNKT cells in the absence of costimulation. A, CD8-depleted thymocytes were labeled with CFSE (10 μM) and transferred into sublethally irradiated mice (750 rad, WT or B7B7h−/−). Seven days later, splenocytes of recipient mice were stained with anti-TCRβ Abs together with CD1d-tetramer and analyzed for the dilution of CFSE. Numbers are the percentages of cells that underwent at least one division. Data are representative of two independent experiments. BD, Indicated strains of mice (6 wk, n = 3–6) were i.p. injected with BrdU (1 mg in 100 μl PBS). Four hours later, thymocytes (B and C), splenocytes, and hepatic mononuclear cells were isolated and stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-NK1.1 Ab. Anti-BrdU Ab staining was performed according to the manufacturer’s recommendation. Numbers are the mean ± SE pooled from two independent experiments. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT.

FIGURE 3.

Lineage expansion of iNKT cells in the absence of costimulation. A, CD8-depleted thymocytes were labeled with CFSE (10 μM) and transferred into sublethally irradiated mice (750 rad, WT or B7B7h−/−). Seven days later, splenocytes of recipient mice were stained with anti-TCRβ Abs together with CD1d-tetramer and analyzed for the dilution of CFSE. Numbers are the percentages of cells that underwent at least one division. Data are representative of two independent experiments. BD, Indicated strains of mice (6 wk, n = 3–6) were i.p. injected with BrdU (1 mg in 100 μl PBS). Four hours later, thymocytes (B and C), splenocytes, and hepatic mononuclear cells were isolated and stained with CD1d-tetramer and anti-TCRβ Ab in combination with anti-NK1.1 Ab. Anti-BrdU Ab staining was performed according to the manufacturer’s recommendation. Numbers are the mean ± SE pooled from two independent experiments. ∗, p < 0.05; ∗∗, p < 0.01 in comparison with WT.

Close modal

One critical step during iNKT development is the massive expansion of immature NK1.1 precursors in the thymus (16). To address the role of costimulation in this process, we injected BrdU into WT and different knockout mice and analyzed the level of BrdU incorporation in the CD1d-tetramer+ cells. As described previously (16), BrdU incorporation is largely restricted to the NK1.1 population; in these conditions ∼10% of this population was BrdU+, whereas <2% of the NK1.1+ population was BrdU+ in the thymic iNKT cells of WT mice (Fig. 3,B). We observed a remarkable reduction (∼50% decrease vs WT, p = 0.0006) of the BrdU+ population in B7h−/− mice (Fig. 3,B). B7−/− mice also showed a slight reduction (∼30% decrease vs WT, p = 0.0135) of BrdU+ cells in NK1.1 iNKT cells. However, we did not observe any synergistic or additive reduction of the BrdU+ population in B7B7h−/− mice (∼50% decrease vs WT, p = 0.0002). Of interest, we also observed a significant reduction of BrdU+ population in T-bet−/− mice (Fig. 3,C, ∼30% decrease vs WT, p = 0.024). These results indicated that costimulation and T-bet are required for efficient expansion of NK1.1 iNKT cells in the thymus. The BrdU incorporation in the peripheral iNKT cells of WT mice was substantially lower than that in the thymic iNKT cells, and we did not observe any evident difference between WT and B7B7h−/− mice, ruling out the possibility that the reduction of BrdU+ cell in the mutant mice was due to more efficient export of iNKT cells from the thymus (Fig. 3 D).

Comparable numbers of NK1.1CD44low iNKT precursors in WT and B7B7h−/− mice (Fig. 2,C) suggests that costimulation might be required for the expansion and maturation of postselected iNKT cells. To assess this, we injected CD4+CD8NK1.1 thymocytes purified from CD45.1+ congenic mice into the thymus of either WT or B7B7h−/− mice. This cell population contained ∼1.5% of immature CD1d-tetramer+ NK1.1 cells. One week later, we analyzed CD45.1+ cells in the thymus of the recipients. As shown in Fig. 4,A, ∼10% of the recovered CD45.1+ cells in the thymus of WT recipient were stained with CD1d-tetramer, indicating an extensive expansion of this population within one week after intrathymic transfer (1.5% to 10.9%). However, recovery of this population was significantly reduced in the thymus of B7B7h−/− recipient (Fig. 4,A, p = 0.0025), suggesting that the lineage expansion of this population is impaired in the absence of costimulation. To further examine the maturation of iNKT cells, we analyzed the expression of the NK1.1 marker on the recovered CD45.1+CD1d-tetramer+ cells. Compared with that of WT recipient, NK1.1+ cells in this population was much lower in B7B7h−/− recipient (Fig. 4 B, p = 0.0075). Collectively, these results demonstrated that costimulatory signals promote optimal expansion and maturation of postselected iNKT cells.

FIGURE 4.

Role of costimulation signal on the postselected immature iNKT cells. CD4+CD8NK1.1 from the thymus of CD45.1+ mice (5 wk) were sorted and injected (20 μl/mouse, 1.5 × 108/ml) into the thymus of CD45.2+ WT or B7B7h−/− mice (7 wk, n = 3–4). Seven days later, lymphoid cells from the thymuses of the recipient were costained with CD1d-tetramer and anti-CD45.1 Ab together with NK1.1 Ab. CD45.1+ (upper panels) or CD45.1+CD1d-tetramer+ cells (lower panels) were gated and analyzed. Numbers are the mean ± SE. **, p < 0.01 in comparison with WT.

FIGURE 4.

Role of costimulation signal on the postselected immature iNKT cells. CD4+CD8NK1.1 from the thymus of CD45.1+ mice (5 wk) were sorted and injected (20 μl/mouse, 1.5 × 108/ml) into the thymus of CD45.2+ WT or B7B7h−/− mice (7 wk, n = 3–4). Seven days later, lymphoid cells from the thymuses of the recipient were costained with CD1d-tetramer and anti-CD45.1 Ab together with NK1.1 Ab. CD45.1+ (upper panels) or CD45.1+CD1d-tetramer+ cells (lower panels) were gated and analyzed. Numbers are the mean ± SE. **, p < 0.01 in comparison with WT.

Close modal

The severe defect of iNKT cells in B7- and/or B7h-deficient mice prompted us to examine the expression and distribution of these costimulators in the thymus. As shown in Fig. 5, the expression of B7h is restricted to the medullar region, mainly on CD11c cells. In addition, although some medullar cells coexpressed B7h and B7.2, we observed a large population of cells that expressed either B7h or B7.2 independently (Fig. 5, lower panels). Therefore, B7 and B7h are expressed in the thymus—mainly in the medulla, presumably by different cell types.

FIGURE 5.

Expression of B7h and B7 in the thymic compartments. Frozen thymic sections from C57BL/6 mice were stained with anti-B7h Ab alone (upper panels) or in combination with anti-CD11c Ab (middle panels) or anti-B7.2 Ab (lower panels). Bars in the corner of each image are 80 μm in ×10 magnification and 60 μm in ×20 magnification. The cortex (C) and medulla (M) are marked.

FIGURE 5.

Expression of B7h and B7 in the thymic compartments. Frozen thymic sections from C57BL/6 mice were stained with anti-B7h Ab alone (upper panels) or in combination with anti-CD11c Ab (middle panels) or anti-B7.2 Ab (lower panels). Bars in the corner of each image are 80 μm in ×10 magnification and 60 μm in ×20 magnification. The cortex (C) and medulla (M) are marked.

Close modal

We next asked what cell types in the thymus are responsible for the costimulation-dependent iNKT cell development. We prepared T cell-depleted bone marrow cells (BM) from either WT or B7B7h−/− mice, and i.v. transferred them into lethally irradiated WT or B7B7h−/− mice. Twelve weeks later, the thymic and splenic iNKT cell populations of the recipient mice were analyzed. As depicted in Fig. 6,A, we observed a great reduction of iNKT cells in the thymus of WT recipients receiving B7B7h−/− BM (0.36 ± 0.02 vs 0.10 ± 0.02, respectively; p = 0.0001). We observed consistent results when we used RAG−/− mice as recipients in the same experimental settings (data not shown). Interestingly, we also observed a severe defect in the iNKT cell population in the thymus of B7B7h−/− recipients receiving WT BM compared with that of WT to WT control mice (Fig. 6,A left panel; 0.36 ± 0.02 vs 0.09 ± 0.04, p = 0.0004). Analysis of NK1.1 and CD44 expression on CD1d-tetramer+ cells revealed that thymic iNKT cells of recipient mice lacking B7 and B7h on either nonhemopoietic or BM-derived cells displayed significant defect in maturation (Fig. 6,B). In addition, splenocyte analysis showed a remarkable reduction of iNKT cells in all other groups compared with the WT recipients receiving WT BM (Fig. 6 A right panel). Altogether, these results demonstrated that the costimulatory molecules on BM-derived and nonhemopoietic cells are both required for iNKT cell development and maturation in the thymus.

FIGURE 6.

Nonredundant roles of nonhematopoietic and hemopoietic cells in the contribution of B7 and B7h during iNKT development. Bone marrow cells depleted of Thy1.2+ cells from the indicated donor mice (10 wk) were i.v. transferred into lethally irradiated recipient mice (6 wk, n = 3 per group). Twelve weeks later, the iNKT population in the thymus and spleen was analyzed. A, Lymphoid cells from the thymus were stained with CD1d-tetramer plus anti-TCRβ Ab. Splenocytes were stained with CD1d-tetramer plus anti-CD19 Ab. B, The expression of CD44 and NK1.1 on the thymic iNKT cells of recipients was analyzed on gated TCRβ+CD1d-tetramer+ cells. Dot plots in B are representatives of three mice in each group.

FIGURE 6.

Nonredundant roles of nonhematopoietic and hemopoietic cells in the contribution of B7 and B7h during iNKT development. Bone marrow cells depleted of Thy1.2+ cells from the indicated donor mice (10 wk) were i.v. transferred into lethally irradiated recipient mice (6 wk, n = 3 per group). Twelve weeks later, the iNKT population in the thymus and spleen was analyzed. A, Lymphoid cells from the thymus were stained with CD1d-tetramer plus anti-TCRβ Ab. Splenocytes were stained with CD1d-tetramer plus anti-CD19 Ab. B, The expression of CD44 and NK1.1 on the thymic iNKT cells of recipients was analyzed on gated TCRβ+CD1d-tetramer+ cells. Dot plots in B are representatives of three mice in each group.

Close modal

CD1d expression on CD4+CD8+ thymocytes is required for the thymic maturation of iNKT cells (12, 13). Results from our bone marrow chimera experiments showed that expression of costimulatory molecules on nonhemopoietic cells is also important for proper development of iNKT cells. CD1d expression on the thymocytes and splenocytes of B7B7h−/− mice appeared to be normal when compared with that of WT mice (Fig. 7,A). To determine whether normal iNKT development is dependent upon coexpression of CD1d and costimulatory molecules on the same cells, we performed mixed bone marrow chimeras. BM cells from WT, B7B7h−/−, or CD1d−/− mice were isolated and various BM combinations (WT + CD1d−/− or B7B7h−/− + CD1d−/−) injected i.v. into lethally irradiated WT mice. Mice injected with either WT or B7B7h−/− BM cells alone were used as controls. Consistent with the data in Fig. 6, mice receiving B7B7h−/− BM showed significantly reduced percentage of thymic iNKT cells along with lower expression of NK1.1 compared with mice receiving WT BM (Fig. 7,B). Mice receiving mixed BM of WT plus CD1d−/− showed comparable level of thymic iNKT cells and NK1.1 expression to those of WT control. Interestingly, mice receiving mixed BM of B7B7h−/− and CD1d−/− also had comparable number of thymic iNKT cells and NK1.1+ cells (Fig. 7 B). These data together suggest that CD1d and the costimulatory molecules can be provided by separate hemopoietic cells during thymic development of iNKT cells.

FIGURE 7.

Separate expression of CD1d and costimulatory molecules on hemopoietic cells is sufficient for normal thymic iNKT cell development. A, Thymocytes and splenocytes from WT (thick line) or B7B7h−/− mice (thin line) were stained with anti-CD1d Ab or isotype control (gray). B, BM cells from WT, B7B7h−/−, or 1:1 mixture with CD1d−/− BM were i.v. injected into lethally irradiated WT mice (n = 3 per group, 6 wk). Twelve weeks later, lymphoid cells from the thymuses of the recipient were costained with CD1d-tetramer and anti-TCRβ Ab (upper panels). The expression of NK1.1 on the thymic iNKT cells of recipients was analyzed on gated TCRβ+CD1d-tetramer+ cells (lower panels). Numbers above each box are the mean ± SE of each group.

FIGURE 7.

Separate expression of CD1d and costimulatory molecules on hemopoietic cells is sufficient for normal thymic iNKT cell development. A, Thymocytes and splenocytes from WT (thick line) or B7B7h−/− mice (thin line) were stained with anti-CD1d Ab or isotype control (gray). B, BM cells from WT, B7B7h−/−, or 1:1 mixture with CD1d−/− BM were i.v. injected into lethally irradiated WT mice (n = 3 per group, 6 wk). Twelve weeks later, lymphoid cells from the thymuses of the recipient were costained with CD1d-tetramer and anti-TCRβ Ab (upper panels). The expression of NK1.1 on the thymic iNKT cells of recipients was analyzed on gated TCRβ+CD1d-tetramer+ cells (lower panels). Numbers above each box are the mean ± SE of each group.

Close modal

Although iNKT cells have emerged as crucial regulators of immunity, the regulation of their development and function remains poorly understood. In this study, we demonstrate that mice lacking CD28 and/or ICOS signaling exhibit severe developmental defects of iNKT cells in the thymus. Costimulatory signals from hemopoietic and nonhemopoietic cells are both required for normal development of thymic iNKT cells.

Costimulatory molecules do not appear to be essential for positive selection of iNKT cells for several reasons. First, CD4+CD8+ population that mediates the positive selection of iNKT cells does not express B7.1/2 and B7h (data not shown). Second, when we analyzed Vβ usage of thymic iNKT cells, we observed no significant difference between WT and B7B7h−/− mice (data not shown). In addition to the usage of the canonical Vα14 chain, the iNKT cell repertoire is preferentially biased toward usage of the Vβ7, Vβ8.2, or Vβ2 chains (29). These TCRs combinations represent optimal solutions for the recognition of glycolipid ligand(s) involved in the positive selection of iNKT cells (30). Lastly and more directly, despite significant decrease in iNKT cells in the thymus of B7B7h−/− mice, the absolute number of NK1.1CD44low subset in these mice was comparable to that of WT.

After positive selection, iNKT precursors undergo a unique expansion and maturation process. In the current study, we found costimulation by CD28 and ICOS is required for both postselection expansion and maturation of iNKT cells. After positive selection, NK1.1 iNKT cells undergo extensive proliferation (16). Recent studies demonstrated that AP-1 and IL-7 might be involved in this process (31, 32). In our study, by performing BrdU incorporation and intrathymic transfer experiments, we demonstrated that costimulation regulates efficient lineage expansion of iNKT cells following selection. Interestingly, B7h deficiency appears to affect proliferation of NK1.1 iNKT cells more than B7 deficiency (Fig. 3,B), correlating with more severe reduction of this population in B7h knockout mice (Fig. 2 C). In addition, lack of costimulation impairs up-regulation of maturation markers such as CD44, NK1.1, T-bet, and CD122. T-bet-deficient mice also showed a similar defect in the expansion of thymic iNKT cells, suggesting that it may serve as a downstream target of costimulatory receptors. Thus, costimulation is likely to be required for least at two different stages during thymic iNKT cell development: (i) lineage expansion during NK1.1 stage, and (ii) maturation from NK1.1 to NK1.1+ stage. At this stage, it is unclear whether these two events are associated or separated. Nonetheless, this developmental regulation exhibits strong resemblance with activation and differentiation of conventional CD4+ T cells in the periphery following antigenic stimulation, in which costimulation not only mediates expansion, but also the functional differentiation (25).

Both BrdU incorporation and homeostatic proliferation studies revealed that costimulation may not be required for the proliferation of iNKT cells in the periphery. This might be partially due to the fact that homeostatic proliferation in the lymphopenic host does not depend on CD1d/TCR interaction but does depend on cytokine, especially IL-15 (28). Moreover, the proliferation of iNKT cells in the periphery in the steady state is much less efficient than that in the thymus.

The requirement of costimulation during iNKT cell activation in the periphery has been well characterized. As conventional T cells, the optimal function of iNKT cells after TCR-mediated stimulation requires costimulation such as CD40 and B7 (4, 23). These costimulatory molecules seem to be important for the expansion of iNKT cells in response to their ligand, but not for prompt cytokine expression by these cells (33). Moreover, a recent study demonstrated that ICOS costimulation is required for iNKT cell functions, such as cytokine production and anti-tumor activity (26). Despite defective thymic maturation in the absence of costimulation, the maturation status of peripheral iNKT cells was largely normal in the absence of costimulation. However, we also observed a weak but evident reduction in the IFN-γ/IL-4-producing iNKT cells in the costimulation-deficient mice, especially B7B7h−/− mice, upon αGC stimulation in vivo. Thus, the inefficient cytokine production from iNKT cell of costimulation-deficient mice may be cell intrinsic. We propose that acquisition of iNKT cell function is intrinsically linked to their development in the thymus. Therefore, costimulatory signals appear to be essential for the optimal function of iNKT cells in the periphery as well as their development in the thymus.

Although CD4+CD8+ thymocytes are the major source of CD1d to mediate positive selection of iNTK cells, results from our study demonstrate that nonhematopoietic and hemopoietic cells are both required for the development of iNKT cells. One possible explanation is that nonhematopoietic and hemopoietic cells provide different costimulation to iNKT cells in different steps of their postselection developmental process, because CD80 and CD86 molecules are expressed in both the cortex and the medulla of the thymus, mainly by CD11c+ DCs. B7h-expressing cells are more strictly restricted to the medulla, mainly by CD11c population.

Whether CD1d molecules play a role in iNKT cell development after positive selection is controversial. Using transgenic mouse with restricted CD1d expression only in thymocytes, two recent studies showed that CD1d expression on thymic T cells was sufficient for iNKT cell development and maturation (34, 35). However, it was also noted in these mice that in the absence of CD1d expression by other cell types, the proportion of NK1.1+ iNKT cells was reduced in the periphery. These results suggested a possible role for CD1d in the maturation of iNKT cells after positive selection. In agreement with this hypothesis, another study demonstrated that the efficiency of NK1.1 immature iNKT cell maturation into NK1.1+ cells is decreased in absence of CD1d expression (36). We observed similar defect in the maturation of iNKT cells when we injected NK1.1 CD4+ thymocytes into the thymus of B7B7h−/− mice and analyzed the maturation of these cells. Therefore, it is likely that costimulation and CD1d expression are both required for optimal development of iNKT cells after positive selection. Surprisingly, our mixed BM reconstitution study revealed that separate expression of the CD1d and costimulation on hemopoietic cells were sufficient to induce normal thymic iNKT cell development. It has been shown that “preactivated” CD4 T cells respond to costimulatory signals in the absence of a TCR ligand, while naive T cells do not (37). Thymic iNKT cells display an activated phenotype (38). Therefore, it is possible to surmise that costimulatory signals promote the development of thymic iNKT cells that have been activated during their positive selection.

In summary, our current study has unveiled a novel function of costimulatory molecules in the intrathymic expansion and maturation of iNKT precursors after their positive selection. This regulation appears distinct from that on the development of conventional T cells, where costimulation does not appear to have a role, or Treg cells, which are only regulated by CD28 but not by ICOS costimulation (our unpublished data). Our results provide a basis to further study iNKT cell developmental program.

We thank Dr. William Sha for anti-B7h hybridoma, Dr. Steve Reiner for suggestion, National Institutes of Health tetramer core facility for PBS57-loaded CD1d-tetramer, and the entire Dong laboratory for their help and discussion.

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.

1

This study was funded in part by grants from the National Institutes of Health (AI50746 to C.D. and AI057485 to L.G.). R.N. receives a postdoctoral fellowship from Arthritis Foundation and a Scientist Development Grant from American Heart Association. D.Z. is funded by M. D. Anderson Cancer Center and by a Developmental Award of Joe Moakley Leukemia Specialized Program of Research Excellence grant from National Cancer Institute. C.D. is a Cancer Research Institute Investigator and a Trust Fellow of M. D. Anderson Cancer Center.

3

Abbreviations used in this paper: iNKT, invariant NK T cells; αGC, α-galactosylceramide; WT, wild type; BM, bone marrow.

1
Bendelac, A., O. Lantz, M. E. Quimby, J. W. Yewdell, J. R. Bennink, R. R. Brutkiewicz.
1995
. CD1 recognition by mouse NK1+ T lymphocytes.
Science
268
:
863
-865.
2
Brigl, M., M. B. Brenner.
2004
. CD1: antigen presentation and T cell function.
Annu. Rev. Immunol.
22
:
817
-890.
3
Kronenberg, M..
2005
. Toward an understanding of NKT cell biology: progress and paradoxes.
Annu. Rev. Immunol.
23
:
877
-900.
4
Kawano, T., J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki, H. Ueno, R. Nakagawa, H. Sato, E. Kondo, et al
1997
. CD1d-restricted and TCR-mediated activation of Vα14 NKT cells by glycosylceramides.
Science
278
:
1626
-1629.
5
Kinjo, Y., E. Tupin, D. Wu, M. Fujio, R. Garcia-Navarro, M. R. Benhnia, D. M. Zajonc, G. Ben-Menachem, G. D. Ainge, G. F. Painter, et al
2006
. Natural killer T cells recognize diacylglycerol antigens from pathogenic bacteria.
Nat. Immunol.
7
:
978
-986.
6
Mattner, J., K. L. Debord, N. Ismail, R. D. Goff, C. Cantu, III, D. Zhou, P. Saint-Mezard, V. Wang, Y. Gao, N. Yin, et al
2005
. Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections.
Nature
434
:
525
-529.
7
Zhou, D., J. Mattner, C. Cantu, III, N. Schrantz, N. Yin, Y. Gao, Y. Sagiv, K. Hudspeth, Y. P. Wu, T. Yamashita, et al
2004
. Lysosomal glycosphingolipid recognition by NKT cells.
Science
306
:
1786
-1789.
8
Cui, J., T. Shin, T. Kawano, H. Sato, E. Kondo, I. Toura, Y. Kaneko, H. Koseki, M. Kanno, M. Taniguchi.
1997
. Requirement for Vα14 NKT cells in IL-12-mediated rejection of tumors.
Science
278
:
1623
-1626.
9
Nichols, K. E., J. Hom, S. Y. Gong, A. Ganguly, C. S. Ma, J. L. Cannons, S. G. Tangye, P. L. Schwartzberg, G. A. Koretzky, P. L. Stein.
2005
. Regulation of NKT cell development by SAP, the protein defective in XLP.
Nat. Med.
11
:
340
-345.
10
Akbari, O., P. Stock, E. Meyer, M. Kronenberg, S. Sidobre, T. Nakayama, M. Taniguchi, M. J. Grusby, R. H. DeKruyff, D. T. Umetsu.
2003
. Essential role of NKT cells producing IL-4 and IL-13 in the development of allergen-induced airway hyperreactivity.
Nat. Med.
9
:
582
-588.
11
MacDonald, H. R..
2002
. Development and selection of NKT cells.
Curr. Opin. Immunol.
14
:
250
-254.
12
Bendelac, A..
1995
. Positive selection of mouse NK1+ T cells by CD1-expressing cortical thymocytes.
J. Exp. Med.
182
:
2091
-2096.
13
Gapin, L., J. L. Matsuda, C. D. Surh, M. Kronenberg.
2001
. NKT cells derive from double-positive thymocytes that are positively selected by CD1d.
Nat. Immunol.
2
:
971
-978.
14
Pellicci, D. G., K. J. Hammond, A. P. Uldrich, A. G. Baxter, M. J. Smyth, D. I. Godfrey.
2002
. A natural killer T (NKT) cell developmental pathway involving a thymus-dependent NK1.1CD4+ CD1d-dependent precursor stage.
J. Exp. Med.
195
:
835
-844.
15
Matsuda, J. L., L. Gapin.
2005
. Developmental program of mouse Vα14i NKT cells.
Curr. Opin. Immunol.
17
:
122
-130.
16
Benlagha, K., T. Kyin, A. Beavis, L. Teyton, A. Bendelac.
2002
. A thymic precursor to the NK T cell lineage.
Science
296
:
553
-555.
17
Matsuda, J. L., Q. Zhang, R. Ndonye, S. K. Richardson, A. R. Howell, L. Gapin.
2006
. T-bet concomitantly controls migration, survival, and effector functions during the development of Vα14i NKT cells.
Blood
107
:
2797
-2805.
18
Townsend, M. J., A. S. Weinmann, J. L. Matsuda, R. Salomon, P. J. Farnham, C. A. Biron, L. Gapin, L. H. Glimcher.
2004
. T-bet regulates the terminal maturation and homeostasis of NK and Vα14i NKT cells.
Immunity
20
:
477
-494.
19
Sivakumar, V., K. J. Hammond, N. Howells, K. Pfeffer, F. Weih.
2003
. Differential requirement for Rel/nuclear factor κB family members in natural killer T cell development.
J. Exp. Med.
197
:
1613
-1621.
20
Gadue, P., L. Yin, S. Jain, P. L. Stein.
2004
. Restoration of NK T cell development in fyn-mutant mice by a TCR reveals a requirement for Fyn during early NK T cell ontogeny.
J. Immunol.
172
:
6093
-6100.
21
Pasquier, B., L. Yin, M. C. Fondaneche, F. Relouzat, C. Bloch-Queyrat, N. Lambert, A. Fischer, G. de Saint-Basile, S. Latour.
2005
. Defective NKT cell development in mice and humans lacking the adapter SAP, the X-linked lymphoproliferative syndrome gene product.
J. Exp. Med.
201
:
695
-701.
22
Schmidt-Supprian, M., J. Tian, E. P. Grant, M. Pasparakis, R. Maehr, H. Ovaa, H. L. Ploegh, A. J. Coyle, K. Rajewsky.
2004
. Differential dependence of CD4+CD25+ regulatory and natural killer-like T cells on signals leading to NF-κB activation.
Proc. Natl. Acad. Sci. USA
101
:
4566
-4571.
23
Hayakawa, Y., K. Takeda, H. Yagita, L. Van Kaer, I. Saiki, K. Okumura.
2001
. Differential regulation of Th1 and Th2 functions of NKT cells by CD28 and CD40 costimulatory pathways.
J. Immunol.
166
:
6012
-6018.
24
Dong, C., A. E. Juedes, U. A. Temann, S. Shresta, J. P. Allison, N. H. Ruddle, R. A. Flavell.
2001
. ICOS co-stimulatory receptor is essential for T-cell activation and function.
Nature
409
:
97
-101.
25
Nurieva, R., S. Thomas, T. Nguyen, N. Martin-Orozco, Y. Wang, M. K. Kaja, X. Z. Yu, C. Dong.
2006
. T-cell tolerance or function is determined by combinatorial costimulatory signals.
EMBO J.
25
:
2623
-2633.
26
Kaneda, H., K. Takeda, T. Ota, Y. Kaduka, H. Akiba, Y. Ikarashi, H. Wakasugi, M. Kronenberg, K. Kinoshita, H. Yagita, K. Okumura.
2005
. ICOS costimulates invariant NKT cell activation.
Biochem. Biophys. Res. Commun.
327
:
201
-207.
27
Salomon, B., D. J. Lenschow, L. Rhee, N. Ashourian, B. Singh, A. Sharpe, J. A. Bluestone.
2000
. B7/CD28 costimulation is essential for the homeostasis of the CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes.
Immunity
12
:
431
-440.
28
Matsuda, J. L., L. Gapin, S. Sidobre, W. C. Kieper, J. T. Tan, R. Ceredig, C. D. Surh, M. Kronenberg.
2002
. Homeostasis of Vα14i NKT cells.
Nat. Immunol.
3
:
966
-974.
29
Ohteki, T., H. R. MacDonald.
1996
. Stringent V β requirement for the development of NK1.1+ T cell receptor-αβ+ cells in mouse liver.
J. Exp. Med.
183
:
1277
-1282.
30
Wei, D. G., S. A. Curran, P. B. Savage, L. Teyton, A. Bendelac.
2006
. Mechanisms imposing the Vβ bias of Vα14 natural killer T cells and consequences for microbial glycolipid recognition.
J. Exp. Med.
203
:
1197
-1207.
31
Zullo, A. J., K. Benlagha, A. Bendelac, E. J. Taparowsky.
2007
. Sensitivity of NK1.1-negative NKT cells to transgenic BATF defines a role for activator protein-1 in the expansion and maturation of immature NKT cells in the thymus.
J. Immunol.
178
:
58
-66.
32
Gadue, P., P. L. Stein.
2002
. NK T cell precursors exhibit differential cytokine regulation and require Itk for efficient maturation.
J. Immunol.
169
:
2397
-2406.
33
Uldrich, A. P., N. Y. Crowe, K. Kyparissoudis, D. G. Pellicci, Y. Zhan, A. M. Lew, P. Bouillet, A. Strasser, M. J. Smyth, D. I. Godfrey.
2005
. NKT cell stimulation with glycolipid antigen in vivo: costimulation-dependent expansion, Bim-dependent contraction, and hyporesponsiveness to further antigenic challenge.
J. Immunol.
175
:
3092
-3101.
34
Zimmer, M. I., A. Colmone, K. Felio, H. Xu, A. Ma, C. R. Wang.
2006
. A cell-type specific CD1d expression program modulates invariant NKT cell development and function.
J. Immunol.
176
:
1421
-1430.
35
Wei, D. G., H. Lee, S. H. Park, L. Beaudoin, L. Teyton, A. Lehuen, A. Bendelac.
2005
. Expansion and long-range differentiation of the NKT cell lineage in mice expressing CD1d exclusively on cortical thymocytes.
J. Exp. Med.
202
:
239
-248.
36
McNab, F. W., S. P. Berzins, D. G. Pellicci, K. Kyparissoudis, K. Field, M. J. Smyth, D. I. Godfrey.
2005
. The influence of CD1d in postselection NKT cell maturation and homeostasis.
J. Immunol.
175
:
3762
-3768.
37
Liu, Y., C. A. Janeway, Jr.
1992
. Cells that present both specific ligand and costimulatory activity are the most efficient inducers of clonal expansion of normal CD4 T cells.
Proc. Natl. Acad. Sci. USA
89
:
3845
-3849.
38
Yang, Y., A. Ueno, M. Bao, Z. Wang, J. S. Im, S. Porcelli, J. W. Yoon.
2003
. Control of NKT cell differentiation by tissue-specific microenvironments.
J. Immunol.
171
:
5913
-5920.