Abstract
To elucidate the mechanisms by which glucocorticoids promote Th2-type responses, we investigated the influence of dexamethasone (DEX) on both cytokine production and viability of NK1.1+ T cells. The in vivo administration of DEX enhanced the IL-4 production of spleen cells and liver mononuclear cells in wild-type mice, but not in β2m-deficient mice. DEX reduced the cellularity of conventional T cells, but not that of NK1.1+ T cells, in both spleen and liver, suggesting an increased proportion of NK1.1+ T cells. Moreover, the proportion of IL-4-producing NK1.1+ T cells increased in the DEX-injected mice. These results suggest that DEX induced IL-4 production through the preferential survival of IL-4-producing NK1.1+ T cells. In investigating the reason for the preferential survival of NK1.1+ T cells, we found that NK1.1+ T cells were resistant to DEX-induced apoptosis and expressed a higher level of intracellular Bcl-2 compared with conventional NK1.1− T cells. In addition, splenic and hepatic NK1.1+ T cells were resistant to radiation-induced apoptosis. Collectively, our findings revealed an important role for NK1.1+ T cells in the regulation of Th1/Th2 balance by glucocorticoids and their possible functions under various apoptotic stimuli.
Glucocorticoids (GC)3 are known to be potent immunosuppressive and anti-inflammatory drugs and are used in the treatment of organ transplantation, autoimmune disease, and many types of inflammatory diseases. Although the precise mechanisms of their immunosuppressive effect have yet to be fully elucidated, accumulating reports have demonstrated an inhibitory effect of GC on cytokine production, phagocytosis, Ag presentation, migration, and cytotoxicity of immune cells (1, 2, 3, 4, 5). In addition to these pharmacologic effects, GC have also been reported to play physiologic immunoregulatory roles, such as the induction of apoptosis (6), the involvement in thymic selection (7), and the regulation of the Th1/Th2 balance (8). Moreover, disorders of such roles of GC have been reported to participate in the pathogenesis of such autoimmune diseases as arthritis, thyroiditis, and experimental allergic encephalomyelitis (EAE) (9).
It is therefore of great interest to determine how GC have opposite immunoregulatory effects on the Th1 and Th2 responses. In contrast to their suppressive effect on the production of such Th1-type cytokines as IL-2, IFN-γ, and TNF-α (10, 11), GC have been reported to augment the production of such Th2-type cytokines as IL-4, IL-10, and IL-13 both in vitro and in vivo (12, 13, 14, 15). The most implicative model of this regulation of Th1/Th2 responses by GC appears to be EAE in Lewis rats, which is a cell-mediated autoimmune disease induced by Th1-type CD4+ T cells in response to myelin basic protein (16). In these rats, the spontaneous recovery from EAE is accompanied by an elevation in the plasma levels of GC, and adrenalectomy leads to progressive disease with a fatal outcome instead of remission (17, 18). In addition, the replacement of GC in adrenalectomized rats is known to induce a recovery from EAE (17). Since EAE in Lewis rats is caused by Th1-type CD4+ T cells that produce IL-2 and IFN-γ, these lines of evidence suggest that GC may thus induce a shift from the Th1 to Th2 responses in the recovery phase of EAE. However, the mechanism by which GC preferentially induce Th2-type responses remains to be elucidated.
NK1.1+ T cells are a recently identified cell population that coexpresses TCR and NK1.1, a member of the NKR-P1 receptor family, and they have been reported to exist in the bone marrow (19), thymus (20), liver (21), and, to a lesser degree, spleen (22). Such NK1.1+ T cells are reported to be either CD4+ CD8− or CD4− CD8−, and to use a single TCR Vα-chain and skewed Vβ-chains (20, 23). Although the functions of NK1.1+ T cells have not been fully elucidated, they are suggested to play some immunoregulatory roles in the thymic selection (24), the suppression of hemopoiesis or CTL generation (25, 26), and the regulation of autoimmune responses (27, 28). More interestingly, NK1.1+ T cells are considered to influence the Th1/Th2 lineage commitment because they promptly produce IL-4 after stimulation with anti-CD3 mAb both in vivo and in vitro (22, 29). Evidence that certain mouse strains, such as SJL or β2m-deficient mice lacking NK1.1+ T cells, fail to produce IL-4 and IgE in response to anti-IgD Ab injection (30, 31) also implies that NK1.1+ T cells contribute to the development of Th2-type responses. However, there are also opposite reports that IL-4 and IgE production are normally induced in β2m-deficient mice (32, 33) or CD1-deficient mice (34, 35), which also lack NK1.1+ T cells, suggesting that the contribution of NK1.1+ T cells remains to be determined.
In this study, we investigated the participation of NK1.1+ T cells in the GC-induced IL-4 production and thus found a preferential survival of IL-4-producing NK1.1+ T cells in the spleen and liver of GC-treated mice. In addition, NK1.1+ T cells were revealed to be resistant to GC and radiation-induced apoptosis probably due to higher expression of intracellular Bcl-2 than conventional T cells. Our findings thus suggest that NK1.1+ T cells play a role in the regulation of Th1/Th2 balance by GC and in the immune responses under apoptotic stimuli.
Materials and Methods
Mice
Female C57BL/6 mice and C57BL/6-lpr/lpr mice were purchased from Japan SLC (Shizuoka, Japan), and β2m-deficient mice (H-2b) were originally supplied by The Jackson Laboratory (Bar Harbor, ME). These mice were maintained under specific pathogen-free conditions and were used for experiments at 8 to 10 wk of age.
Reagents and Abs
Dexamethasone (DEX) was purchased from Sigma (St. Louis, MO). Anti-CD3 mAb (hamster IgG) and anti-TCRαβ mAb (hamster IgG) were obtained from the supernatants of anti-CD3 mAb-producing hybridoma 145-2C11 (provided by Dr. J. A. Bluestone, Department of Pathology, University of Chicago, Chicago, IL) and anti-TCRαβ mAb-producing hybridoma H57-597 (provided by Dr. R. T. Kubo, Cytel, San Diego, CA), respectively. These Abs were purified by collecting the supernatants of the hybridoma cells growing in a serum-free medium (101, Nissui Pharmaceutical, Tokyo, Japan). Anti-murine IL-4 mAb (11B11; rat IgG1) was provided by Dr. S. Hamano (Department of Parasitology, Kyushu University, Fukuoka, Japan), and anti-murine IL-4R mAb was purchased from Genzyme (Cambridge, MA). The mAbs used in flow cytometric analysis were as follows: FITC- or PE-conjugated anti-TCRαβ mAb, PE- or biotin-conjugated anti-NK1.1 mAb, FITC-conjugated anti-IL-4 mAb, FITC-conjugated rat IgG2b, PE-conjugated anti-Fas mAb, anti-murine Bcl-2 mAb (hamster IgG), biotin-conjugated mouse anti-hamster IgG mAb, and FITC-conjugated anti-bromodeoxyuridine (BrdUrd) mAb. These mAbs were all purchased from PharMingen (San Diego, CA). Anti-CD3 mAb (rat IgG2a), purified from the supernatants of anti-CD3 mAb-producing hybridoma KT3.2 and conjugated with FITC in our laboratory, were also used in some experiments.
Preparation of liver mononuclear cells (MNC)
Isolation of liver MNC was conducted as previously reported with minor modifications (36). Briefly, the liver was pressed through a 100-gauge stainless steel mesh in a complete medium, which consisted of RPMI 1640 (Life Technologies, Grand Island, NY) supplemented with 10% heat-inactivated FCS (HyClone, Logan, UT), 5 × 10−5 M 2-ME, 20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 30 μg/ml gentacin (Schering, Kenilworth, NJ), and 0.2% sodium bicarbonate. The cells were resuspended in 5 ml of 45% Percoll (Sigma) and layered on 5 ml of 67.5% Percoll solution. The gradient was centrifuged at 2500 rpm at 20°C for 25 min. The cells at the interface were harvested, washed with the complete culture medium, and then used as liver MNC.
In vitro cytokine production
The mice were injected i.p. with either 500 μl of PBS or the indicated doses of DEX dissolved in 500 μl of PBS. After 18 h, the spleen cells and liver MNC were harvested and cultured at a dose of 1 × 106 cells/ml in flat-bottom 96-well microtiter plates (Corning, Corning, NY) precoated with 10 μg/ml of anti-TCRαβ mAb. After 2 days, the culture supernatants were harvested, and the amounts of IFN-γ and IL-4 were determined using a sandwich ELISA, as previously reported (37).
Flow cytometric analysis
The spleen cells and liver MNC prepared from either DEX-treated or untreated mice were stained with mAbs and analyzed by a FACS Calibur flow cytometer with CellQuest software (Becton Dickinson, Mountain View, CA). The supernatants of anti-FcRγ II/III mAb-producing hybridoma (2.4G2: rat IgG2b) were used to block nonspecific binding in all experiments. In some experiments the prepared cells were stained with 2 μg/ml propidium iodide (Sigma) to exclude any dead cells.
Analysis of intracellular IL-4 synthesis
The mice were injected i.p. with either PBS or 200 μg of DEX, and 18 h later they were injected i.v. with 2.5 μg of anti-CD3 mAb. After 90 min, the spleen cells and liver MNC were prepared, washed, and suspended at 1 × 106 cells/ml in the complete culture medium, and then incubated for 4 h at 37°C in the presence of 10 μg/ml brefeldin A (Wako, Osaka, Japan). These cells were harvested, washed, and incubated for 30 min at 4°C with both PE-conjugated anti-TCRαβ mAb and biotin-conjugated anti-NK1.1 mAb, followed by Red 670-conjugated streptavidin (Life Technologies, Gaithersburg, MD). The cells were washed and fixed in 100 μl of solution A (Cell Perm & Fix, Caltag, South San Francisco, CA) for 30 min at room temperature, and then washed again and resuspended in 50 μl of solution B (permeabilization buffer) containing 1 μg of FITC-conjugated anti-IL-4 mAb. The cells were incubated for 30 min at 4°C and washed, and then cell fluorescence was analyzed by a flow cytometer using the gates set by FL2 and FL3 to determine the proportion of IL-4-producing cells in TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets. As negative controls, the samples were also incubated with isotype-matched Ab (FITC-conjugated rat IgG2b) instead of FITC-conjugated anti-IL-4 mAb.
Detection of apoptotic cells
For the measurement of apoptosis, spleen cells and liver MNC were prepared from either the untreated mice or the mice injected i.p. with 200 μg of DEX either 3, 6, or 12 h previously. To analyze the apoptotic cells in conjunction with cell surface staining, we used the terminal deoxynucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate-biotin nick-end labeling (TUNEL) method, as previously reported (38). The prepared cells (1 × 106 cells) were briefly incubated with both FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb for 30 min at 4°C. For fixation and permeabilization, the cells were washed and suspended with 250 μl of 1% paraformaldehyde in PBS for 60 min at 4°C, and then stored in 200 μl of 75% ethanol at −20°C. For analysis, these cells were incubated in 50 μl of buffer containing 250 U/ml TdT and 10 μM biotin-16-dUTP (both from Boehringer Mannheim, Indianapolis, IN) for 60 min at 37°C, followed by Red 670-streptavidin staining for 30 min. Cell fluorescence was analyzed by a flow cytometer using the gates of the TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets, and the negative controls were determined by the samples incubated without TdT.
In vitro assay of cell survival
The spleen cells or liver MNC (1 × 106 cells/ml) prepared from the untreated mice were incubated at 37°C in 24-well tissue culture plates (Corning) in a volume of 1 ml/well of the complete medium in the absence or the presence of the indicated doses of DEX. In some wells, either anti-IL-4 mAb (25 μg/ml) or anti-IL-4R mAb (5 μg/ml) was added to neutralize the endogenous IL-4 production. After 8 h, these cells were harvested and stained with both FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb, and then with 2 μg/ml propidium iodide. The surviving cells in TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets were analyzed by a flow cytometer with the gate excluding dead cells that were stained with propidium iodide. The percentage of survival in each subset was calculated according to the following formula: % survival = (number of surviving cells in the presence of DEX, DEX and anti-IL-4 mAb, or DEX and anti-IL-4R mAb)/(number of surviving cells in the absence of DEX) × 100. As for radiation-induced apoptosis, the spleen cells or the liver MNC prepared from the untreated mice were irradiated with or without 150 rad by 137Cs irradiator (Gammacell 40, Atomic Energy of Canada, Ottawa, Canada) and then incubated in 24-well tissue culture plates (Corning) at 1 × 106 cells/well at 37°C. After 0, 4, 8, 12, and 20 h, the number of surviving cells of TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets were analyzed in each well by a flow cytometer.
Analysis of intracellular Bcl-2 expression
The spleen cells or liver MNC (1 × 106 cells/ml) prepared from the untreated mice were analyzed for their Bcl-2 expression either before or after incubation in the presence of 10−7 M DEX for 2 or 4 h. These cells were stained with both FITC-conjugated anti-CD3 mAb (KT3.2; rat IgG2a) and PE-conjugated anti-NK1.1 mAb (mouse IgG2a) for 30 min at 4°C. The cells were washed and fixed in 100 μl of solution A (Cell Perm & Fix, Caltag) for 30 min at room temperature, and then washed and incubated with 50 μl of solution B containing 10 μg/ml of hamster anti-murine Bcl-2 mAb for 30 min at 4°C. The cells were washed again and incubated with biotin-conjugated anti-hamster IgG mAb followed by Red 670-conjugated streptavidin. The expression level of Bcl-2 was analyzed by a flow cytometer using the gates of either the CD3+ NK1.1− or the CD3+ NK1.1+ cell subset, and the negative controls were determined by the addition of normal hamster IgG instead of anti-Bcl-2 mAb.
BrdUrd incorporation analysis
The mice were injected i.p. with 500 μl of 2 mg/ml BrdUrd (Sigma) every 12 h for 3 days and analyzed 12 h after the last injection. In a group, the mice were also injected i.p. with 200 μg of DEX 18 h before the analysis. The liver MNC prepared from these mice were stained with both PE-conjugated anti-TCRαβ mAb and biotin-conjugated anti-NK1.1 mAb, followed by Red 670-conjugated streptavidin. After surface staining, BrdUrd incorporation was analyzed as previously reported (39). The cells were washed, resuspended in cold 0.15 M NaCl, and fixed by dropwise addition of cold 75% ethanol. The cells were incubated for 30 min at 4°C, washed with PBS, then incubated with PBS containing 1% paraformaldehyde and 0.01% Tween-20 for 1 h. The cells were pelleted, then incubated with 50 Kunitz units of DNase I (Boehringer Mannheim) in 0.15 M NaCl and 4.2 mM MgCl2, pH 5, for 10 min. After washing, the cells were incubated with FITC-conjugated anti-BrdUrd mAb and analyzed by a flow cytometer with the gates of TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets. As a negative control, the mice not injected with BrdUrd were also analyzed.
Statistics
The statistical significance of the data was determined using Student’s t test. p < 0.05 was considered statistically significant.
Results
NK1.1+ T cells are necessary for the increased IL-4 production by DEX
It has been reported that GC promote the production of Th2-type cytokines such as IL-4, IL-10, and IL-13 both in vivo and in vitro (12, 13, 14, 15). To investigate the contribution of NK1.1+ T cells to the increased production of Th2-type cytokines by GC, we at first examined the effect of the in vivo administration of DEX, a synthetic GC, on the potential of the spleen cells and liver MNC to produce IFN-γ and IL-4 in wild-type mice and β2m-deficient mice that lack NK1.1+ T cells. As shown in Figure 1⇓, the in vivo treatment of wild-type mice with DEX inhibited IFN-γ production of both spleen cells and liver MNC in a dose-dependent manner. On the other hand, the IL-4 production of both the spleen cells and liver MNC in wild-type mice was augmented by the administration of DEX, and the ability of DEX to induce IL-4 production was most remarkable at a dose of 200 μg of DEX. In contrast to wild-type mice, IL-4 production was down-regulated in β2m-deficient mice, and more importantly, no enhancement of IL-4 production was induced in β2m-deficient mice even when they were injected with 200 μg of DEX. These results indicate that the DEX treatment promotes IL-4 production in both spleen cells and liver MNC and that NK1.1+ T cells would thus be necessary for such increased IL-4 production by DEX. In addition, we selected 200 μg of DEX as an administration dose in the following experiments.
DEX promotes IL-4 production in both spleen cells and liver MNC of wild-type mice, but not in β2m-deficient mice. Wild-type C57BL/6 mice and β2m-deficient mice (H-2b) were injected i.p. with either 500 μl of PBS or the indicated doses of DEX dissolved in 500 μl of PBS. After 18 h, the spleen cells and the liver MNC were harvested and cultured at a dose of 1 × 106 cells/ml in flat-bottom 96-well microtiter plates precoated with 10 μg/ml of anti-TCRαβ mAb. After 2 days, the culture supernatants were harvested, and the amounts of IFN-γ and IL-4 were determined by a sandwich ELISA. The values are the mean ± SD of triplicate wells, and the findings are representative of three independent experiments. * indicates p < 0.05 compared with the PBS-injected mice. ** indicates p < 0.01 compared with the PBS-injected mice.
DEX treatment induces the preferential survival of TCRαβ+ NK1.1+ cells in spleen and liver
To further investigate the contribution of NK1.1+ T cells to the induction of Th2-type cytokines by GC, we next examined the phenotypic changes in spleen cells and liver MNC in DEX-treated mice. Figure 2⇓ shows that DEX treatment increased the proportion of TCRαβ+ NK1.1+ cells in both spleen and liver and reduced that of TCRαβ+ NK1.1− cells in liver MNC. No definite effect on TCRαβ− NK1.1+ cells was observed. Since GC are reported to induce apoptosis in several cell types, including mature T cells (40), we further examined the absolute cell number of these subsets in the spleen and liver of DEX-treated mice. As shown in Table I⇓, the administration of DEX decreased the cellularity in the spleen but not that in the liver and also decreased the cell number of TCRαβ+ NK1.1− cells in both organs 18 h after the injection. Interestingly, the cell number of TCRαβ+ NK1.1+ cells did not decrease in the spleen and even increased in the liver. The cell number of TCRαβ− NK1.1+ cells decreased in the spleen after DEX treatment. These changes in cellularity in each subset were similar 36 h after the injection of DEX, except for a recovery of TCRαβ+ NK1.1− cells in the liver. These results suggest that TCRαβ+ NK1.1+ cells preferentially survive the in vivo treatment with DEX.
Increased proportion of TCRαβ+ NK1.1+ cells in spleen and liver of the DEX-injected mice. C57BL/6 mice were injected i.p. with either 500 μl of PBS or 200 μg of DEX dissolved in 500 μl of PBS. After 18 h, the spleen cells and the liver MNC were harvested and stained with FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb. The staining procedures were conducted after blocking the nonspecific binding with anti-FcRγ II/III mAb, and the analysis gate was set to exclude any dead cells by propidium iodide. The horizontal and vertical axes depict the intensity of fluorescence on a logarithmic scale, and the number represents the percentage of each area. These findings are representative of five independent experiments.
Preferential survival of the TCRαβ+ NK1.1+ cells in the DEX-treated mice
DEX treatment increases the proportion of IL-4-producing TCRαβ+ NK1.1+ cells in spleen and liver
To directly confirm the participation of TCRαβ+ NK1.1+ cells in GC-induced IL-4 production, we next examined the intracellular IL-4 synthesis in NK1.1+ T cells of DEX-treated mice. Since it has been reported that the i.v. injection of a low dose of anti-CD3 mAb induces IL-4 production in NK1.1+ T cells (22), we used this protocol and found that the proportion of IL-4-producing TCRαβ+ NK1.1+ cells in TCRαβ+ cells increased in both spleen and liver of DEX-treated mice (Fig. 3⇓). We also confirmed that these IL-4-positive cells were not detected by the staining with isotype-matched control Ab. These results thus indicate that IL-4 production is enhanced by DEX through the increased proportion of IL-4-producing TCRαβ+ NK1.1+ cells in DEX-treated mice.
Increased proportion of IL-4-producing TCRαβ+ NK1.1+ cells in spleen and liver of the DEX-injected mice. C57BL/6 mice were injected i.p. with either PBS or 200 μg of DEX, and 18 h later the mice were injected i.v. with 2.5 μg of anti-CD3 mAb. After 90 min, spleen cells and liver MNC were harvested and incubated at a dose of 1 × 106 cells/ml for 4 h at 37°C in the presence of 10 μg/ml brefeldin A. These cells were washed and stained with PE-conjugated anti-TCRαβ mAb and biotin-conjugated anti-NK1.1 mAb, followed by Red 670-conjugated streptavidin. These cells were then fixed, permeabilized, and stained with either FITC-conjugated anti-IL-4 mAb or isotype-matched rat IgG2b. Cell fluorescence was analyzed by a flow cytometer with the gate of TCRαβ+ cells. The numbers in the upper right quadrant and the lower right quadrant represent the percentage of IL-4-producing TCRαβ+ NK1.1+ cells and IL-4-producing TCRαβ+ NK1.1− cells in total TCRαβ+ cells, respectively. These findings are representative of four independent experiments.
Fas-mediated apoptosis is irrelevant to the preferential survival of TCRαβ+ NK1.1+ cells by DEX
To investigate the underlying mechanism by which TCRαβ+ NK1.1+ cells preferentially survived in the DEX-treated mice, we at first examined the possible role of Fas/Fas ligand interaction, because NK1.1+ T cells possess an ability to kill lymphocytes through Fas-mediated cytolysis (24). As shown in Figure 4⇓A, the expression level of Fas on TCRαβ+ NK1.1− cells and TCRαβ+ NK1.1+ cells was comparable in untreated and DEX-treated mice in both spleen cells and liver MNC. In addition, Figure 4⇓B shows that DEX treatment increased the proportion of TCRαβ+ NK1.1+ cells and reduced that of TCRαβ+ NK1.1− cells in the liver of genetically Fas-deficient C57BL/6-lpr/lpr mice. We also examined the absolute cell number of each subset in C57BL/6-lpr/lpr mice and found that the preferential survival of TCRαβ+ NK1.1+ cells, similar to that shown in Table I⇑, was also observed after treatment with DEX (data not shown). These results thus suggest that the different survival rates of TCRαβ+ NK1.1− cells and TCRαβ+ NK1.1+ cells after DEX treatment are independent of Fas-mediated apoptosis.
Fas-mediated apoptosis is irrelevant to the preferential survival of TCRαβ+ NK1.1+ cells by DEX. A, C57BL/6 mice were injected i.p. with either 500 μl of PBS or 200 μg of DEX dissolved in 500 μl of PBS. After 18 h, the spleen cells and the liver MNC were harvested and analyzed by three-color staining with FITC-conjugated anti-TCRαβ mAb, PE-conjugated anti-Fas mAb, and biotin-conjugated anti-NK1.1 mAb, followed by Red 670-conjugated streptavidin. The expression level of Fas was analyzed by a flow cytometer with the gates of TCRαβ+ NK1.1− cells or TCRαβ+ NK1.1+ cells. The dotted lines indicate the control staining with the isotype-matched mAb, and the number represents the mean fluorescence intensity of samples. B, C57BL/6-lpr/lpr mice were injected i.p. with either PBS or 200 μg of DEX, and 18 h later the liver MNC were harvested and stained with FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb. The number represents the percentage of each area, and the results are representative of three independent experiments.
TCRαβ+ NK1.1+ cells are resistant to DEX-induced apoptosis
We next compared the sensitivity to DEX-induced apoptosis between NK1.1+ T cells and conventional T cells, since GC have been reported to induce apoptotic cell death in mature lymphocytes as well as immature thymocytes (40). As shown in Figure 5⇓A, the proportion of apoptotic cells in TCRαβ+ NK1.1+ cells was significantly lower than that in TCRαβ+ NK1.1− cells after the injection of DEX. The differences were apparent in the liver, and there were small but significant differences at 3 and 6 h after the injection of DEX in the spleen cells. We further showed the representative data of the TUNEL analysis in Figure 5⇓B, which demonstrated that the percentage of apoptotic cells in hepatic TCRαβ+ NK1.1+ cells was approximately 10 times lower than that in hepatic TCRαβ+ NK1.1− cells at 6 h after the injection of DEX. These results suggest that the preferential survival of TCRαβ+ NK1.1+ cells in the DEX-treated mice is due to their resistance to DEX-induced apoptosis.
TCRαβ+ NK1.1+ cells are resistant to DEX-induced apoptosis. A, The spleen cells and liver MNC were harvested from untreated mice or mice injected i.p. with 200 μg of DEX 3, 6, or 12 h previously. These cells were stained with FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb and then applied to a TUNEL analysis as described in Materials and Methods. The percentage of apoptotic cells was analyzed by a flow cytometer with the gates of either TCRαβ+ NK1.1− cells or TCRαβ+ NK1.1+ cells. The values indicate the mean ± SD of at least five mice, and the same results were obtained in four independent experiments. * indicates p < 0.05 compared with TCRαβ+ NK1.1− cells. ** indicates p < 0.01 compared with TCRαβ+ NK1.1− cells. B, Representative data from a TUNEL analysis of the liver MNC of DEX-treated mice 6 h after the injection are shown. The dotted lines indicate the negative controls that were stained without TdT, and the number represents the percentage of TUNEL-positive apoptotic cells.
Resistance of TCRαβ+ NK1.1+ cells to DEX-induced apoptosis is independent of their IL-4 production
We further tried to clarify the reason why TCRαβ+ NK1.1+ cells were resistant to DEX-induced apoptosis, and at first examined whether the potential of NK1.1+ T cells to produce IL-4 was involved in their resistance to GC, since IL-4 protects T cells from GC-induced apoptosis (41, 42). Figure 6⇓A shows that in vivo administration of anti-IL-4 mAb before treatment with DEX resulted in changes in cellularity similar to those in the mice injected with DEX alone. We also examined the in vitro sensitivity of conventional T cells and NK1.1+ T cells to the DEX-induced apoptosis, and the influence of IL-4 on the survival of these cells in vitro. As shown in Figure 6⇓B, the percentage of survival in TCRαβ+ NK1.1+ cells was significantly higher than that of TCRαβ+ NK1.1− cells with 10−7 M DEX in spleen cells and 10−8 and 10−7 M DEX in liver MNC. Moreover, the addition of neither anti-IL-4 mAb nor anti-IL-4R mAb to the culture abrogated such preferential survival of TCRαβ+ NK1.1+ cells in the presence of DEX. These results suggest that the resistance of TCRαβ+ NK1.1+ cells to DEX-induced apoptosis is not ascribed to their potential to produce IL-4.
The preferential survival of TCRαβ+ NK1.1+ cells is independent of their potential to produce IL-4. A, C57BL/6 mice were injected i.p. with PBS (open bar) or 200 μg of DEX (closed bar), and after 18 h, total cell number and cell number of TCRαβ+ NK1.1− or TCRαβ+ NK1.1+ cell subsets in spleen and liver were analyzed. In a group, the mice were also injected i.p. with 1 mg of anti-IL-4 mAb 24 h before DEX injection (hatched bar). n.s., not significant. B, The spleen cells and liver MNC (1 × 106 cells/ml) prepared from the untreated mice were incubated in the absence or the presence of the indicated doses of DEX. During the incubation, none (circle symbols), 25 μg/ml of anti-IL-4 mAb (square symbols), or 5 μg/ml of anti-IL-4R mAb (triangle symbols) was also added. After 8 h, the percentages of cell survival in TCRαβ+ NK1.1− (closed symbols) and TCRαβ+ NK1.1+ (open symbols) cell subsets were analyzed by a flow cytometer and calculated as described in Materials and Methods. The values represent the mean ± SD at least triplicate wells, and these findings are representative of three independent experiments. ** indicates p < 0.01 compared with TCRαβ+ NK1.1− cells.
A higher expression of Bcl-2 in CD3+ NK1.1+ cells than in CD3+ NK1.1− cells
We further investigated the expression level of intracellular Bcl-2 in conventional T cells and NK1.1+ T cells, since overexpression of Bcl-2 has been reported to prevent GC-induced apoptosis in T cell hybridomas (43). As shown in Figure 7⇓, the expression level of Bcl-2 in CD3+ NK1.1+ cells was higher than that in conventional CD3+ NK1.1− cells, especially in the liver MNC. In addition, the higher expression of Bcl-2 in CD3+ NK1.1+ cells than in CD3+ NK1.1− cells was also exhibited after the in vitro culture with DEX, although the DEX treatment elevated Bcl-2 expression in both subsets. These results suggest that the resistance of NK1.1+ T cells to GC-induced apoptosis could depend on their high expression of intracellular Bcl-2.
The high expression of intracellular Bcl-2 in NK1.1+ T cells. The spleen cells or liver MNC (1 × 106 cells/ml) prepared from the untreated mice were analyzed for their Bcl-2 expression either before or after incubation in the presence of 10−7 M DEX for 2 or 4 h. These cells were harvested and stained with FITC-conjugated anti-CD3 mAb and PE-conjugated anti-NK1.1 mAb. These cells were then fixed, permeabilized, and stained with hamster anti-mouse Bcl-2 mAb, followed by biotin-conjugated anti-hamster IgG mAb and then by Red 670-conjugated streptavidin. The expression level of intracellular Bcl-2 was analyzed using a flow cytometer with the gates of either CD3+ NK1.1− cells or CD3+ NK1.1+ cells. The dotted lines indicate negative controls, which were incubated with normal hamster IgG instead of anti-Bcl-2 mAb, and the number represents the mean fluorescence intensity of the samples. These findings are representative of three independent experiments.
TCRαβ+ NK1.1+ cells are resistant to radiation-induced apoptosis
We next determined whether TCRαβ+ NK1.1+ cells could also exhibit resistance to radiation-induced apoptosis, since a high expression of Bcl-2 is known to protect lymphocytes from radiation-induced apoptosis as well as GC-induced apoptosis (44). As shown in Figure 8⇓, the in vitro exposure to radiation significantly reduced the number of surviving TCRαβ+ NK1.1− cells in both spleen cells and liver MNC from 8 to 20 h after radiation. In contrast, the numbers of surviving splenic and hepatic TCRαβ+ NK1.1+ cells after radiation were almost equal to those without radiation, although the number of hepatic TCRαβ+ NK1.1+ cells spontaneously decreased even without radiation. These results suggest that TCRαβ+ NK1.1+ cells are resistant not only to DEX-induced apoptosis but also to radiation-induced apoptosis, probably due to their high expression of intracellular Bcl-2.
TCRαβ+ NK1.1+ cell are resistant to radiation-induced apoptosis. The spleen cells and the liver MNC prepared from untreated mice were incubated at 1 × 106 cells/well in 24-well culture plates after treatment with radiation (150 rad; closed circle) or without radiation (open circle). After 0, 4, 8, 12, and 20 h, the cells were harvested and stained with FITC-conjugated anti-TCRαβ mAb and PE-conjugated anti-NK1.1 mAb and then with 2 μg/ml propidium iodide. The number of surviving cells in TCRαβ+ NK1.1− and TCRαβ+ NK1.1+ cell subsets was analyzed in each well by a flow cytometer. The values represent the mean ± SD at least triplicate wells, and these findings are representative of three independent experiments. ** indicates p < 0.01 compared with irradiated cells.
DEX treatment does not induce proliferative responses of TCRαβ+ NK1.1+ cells in liver
We lastly examined whether the in vivo treatment of DEX induced the proliferative responses of TCRαβ+ NK1.1+ cells in the liver, since the absolute cell number of hepatic TCRαβ+ NK1.1+ cells increased after DEX treatment (Table I⇑). As shown in Figure 9⇓, the incorporation of BrdUrd in hepatic TCRαβ+ NK1.1+ cells did not increase after in vivo treatment with DEX and was almost comparable to that of TCRαβ+ NK1.1− cells. These results suggest that the increased number of hepatic TCRαβ+ NK1.1+ cells after DEX treatment is independent of their proliferative activation.
The unchanged incorporation of BrdUrd in hepatic TCRαβ+ NK1.1+ cells after treatment with DEX. C57BL/6 mice were injected i.p. with 500 μl of 2 mg/ml BrdUrd (Sigma) every 12 h for 3 days and analyzed 12 h after the last injection. In a group, the mice were also injected i.p. with 200 μg of DEX 18 h before the analysis. The liver MNC prepared from these mice were stained with both PE-conjugated anti-TCRαβ mAb and biotin-conjugated anti-NK1.1 mAb, followed by Red 670-conjugated streptavidin. The cells were washed and fixed with 75% ethanol for 30 min at 4°C, and then incubated with PBS containing 1% paraformaldehyde and 0.01% Tween 20 for 1 h. The cells were washed and incubated in the presence of DNase I for 10 min. After washing, the cells were stained with FITC-conjugated anti-BrdUrd mAb and analyzed by a flow cytometer with the gates of TCRαβ+ NK1.1− cells or TCRαβ+ NK1.1+ cells. As a control, liver MNC from the mice uninjected with BrdUrd were also prepared. The number represents the percentage of BrdUrd-positive cells, and the findings are representative of three independent experiments.
Discussion
A notable finding in the present study is that the increased IL-4 production produced by DEX was accompanied by the preferential survival of IL-4-producing TCRαβ+ NK1.1+ cells in spleen and liver. More interestingly, such preferential survival of TCRαβ+ NK1.1+ cells was ascribed to their resistance to DEX-induced apoptosis, probably through a higher expression of intracellular Bcl-2 than that in conventional T cells. In addition, the splenic and hepatic TCRαβ+ NK1.1+ cells showed resistance to radiation-induced cell death in vitro. These findings therefore suggest that NK1.1+ T cells play an important role in Th1/Th2 regulation by GC and their possible functions under the conditions of various apoptotic stimuli.
Although several researchers have reported that GC promote the production of Th2-type cytokines in contrast to their inhibitory effect on Th1-type cytokines (10, 11, 12, 13, 14, 15), the underlying mechanisms of the effects have yet to be fully elucidated. In general, there appear to be at least two possible explanations. One is that GC preferentially induce differentiation in Th2-type T cells from naive CD4+ T cells, and the other is that GC enhance the proliferation and cytokine production of already differentiated Th2-type T cells. The latter possibility is supported by the reports that IL-4-dependent proliferation of Th2-type clone is augmented by GC (45) and that IL-4 rescues Th2 cells from GC-induced apoptosis (41, 42). However, this idea alone is not sufficient to explain the finding that GC increase the production of Th2-type cytokines from naive spleen and lymph node cells as well as primed T cells (13). In this study we also demonstrated that IL-4 production could be augmented by DEX in naive spleen cells and liver MNC (Fig. 1⇑). In this regard, we propose another possibility: that GC promote the development of Th2-type T cells through the preferential survival of IL-4-producing NK1.1+ T cells. We demonstrated the survival of NK1.1+ T cells after treatment with DEX both in vivo (Fig. 2⇑ and Table I⇑) and in vitro (Fig. 6⇑B). Moreover, the importance of NK1.1+ T cells in DEX-induced IL-4 production was also confirmed by the experiments using β2m-deficient mice (Fig. 1⇑) and by an analysis of intracellular IL-4 synthesis (Fig. 3⇑). Although it remains controversial that IL-4 production by NK1.1+ T cells is necessary for the development of Th2-type T cells (30, 31, 32, 33, 34, 35), our findings suggest a novel mechanism concerning the regulation of Th1/Th2 balance by GC.
To elucidate the underlying mechanism for the preferential survival of NK1.1+ T cells by DEX, we at first examined the involvement of Fas/Fas ligand interaction because NK1.1+ T cells were reported to kill lymphocytes through Fas-mediated cytolysis (24). However, this possibility could be ruled out by the comparable expression of Fas between NK1.1+ T cells and conventional T cells (Fig. 4⇑A) and the increased proportion of NK1.1+ T cells by DEX even in Fas-mutated lpr mice (Fig. 4⇑B). Furthermore, these findings also suggest that GC-induced apoptosis may be independent of the interaction of Fas/Fas ligand, in contrast to a possible role of this interaction in radiation-induced apoptosis (46). Next, we examined the sensitivity of NK1.1+ T cells to DEX-induced apoptosis and found that they were more resistant to DEX-induced apoptosis than conventional T cells in both spleen and liver (Fig. 5⇑). Based on this finding, we concluded that the preferential survival of NK1.1+ T cells would be ascribed to their resistance to GC-induced apoptosis. This explanation is plausible but not sufficient, because we cannot yet explain the result that the cellularity of hepatic TCRαβ+ NK1.1+ cells increased after treatment with DEX (Table I⇑). To explore this question, we also examined the in vivo proliferative activity of NK1.1+ T cells, but detected no increased proliferation by the treatment with DEX (Fig. 9⇑). We could not clarify the reason for the increase in hepatic NK1.1+ T cells by DEX in this study, but it is possible that the activity of migration in NK1.1+ T cells would be altered by DEX, since GC have been reported to regulate the production of several chemoattractants for leukocytes (47, 48).
The precise mechanism by which GC induce apoptosis still remains unclear, but several means to protect T cells from GC-induced apoptosis have been reported. In considering such reports, we investigated the reason why NK1.1+ T cells are resistant to GC-induced apoptosis. We first examined a possibility that IL-4 production by NK1.1+ T cells contributed to their resistance to GC-induced apoptosis, since IL-4 protects T cells from GC-induced apoptosis (41, 42) and the binding affinity of GC receptors is reduced by IL-4 and IL-2 (49). However, neither in vivo nor in vitro neutralization of IL-4 abrogated the resistance of NK1.1+ T cells to apoptosis by DEX (Fig. 6⇑), thus ruling out this possibility. Other cytokines, such as IL-6 and IL-9, and some chemokines are also reported to protect lymphocytes from GC-induced apoptosis (50, 51, 52), but we consider these factors to have only minor effects, since NK1.1+ T cells have not been shown to produce these cytokines. We next examined the expression of intracellular Bcl-2, since the transfection of this proto-oncogene into T cell hybridoma inhibits GC-induced apoptosis (43) and peripheral T cells in Bcl-2 knockout animals are sensitive to GC-induced apoptosis (53). The expression of Bcl-2 in NK1.1+ T cells was higher than that in conventional T cells regardless of culture with DEX (Fig. 7⇑), suggesting a possible contribution of Bcl-2 to the resistance of NK1.1+ T cells to GC-induced apoptosis. The expression level of Bcl-2 was less high in the splenic NK1.1+ T cells, but it may be compatible with the finding that the resistance of splenic NK1.1+ T cells to DEX-induced apoptosis was less apparent both in vivo (Fig. 5⇑A) and in vitro (Fig. 6⇑B). Interestingly, the level of Bcl-2 in both conventional T cells and NK1.1+ T cells increased after culture with DEX. With regard to the mechanism for such an increase in Bcl-2, we consider the possibility that the cells expressing high levels of Bcl-2 selectively survived the treatment with DEX or that DEX directly increased the expression level of Bcl-2, as previously reported (54). We further demonstrated the radioresistance of NK1.1+ T cells in vitro (Fig. 8⇑), which is consistent with a previous report that hepatic IL-2Rβ+ intermediate TCR cells are resistant to in vivo radiation (55). These findings may support our idea, since Bcl-2 can also inhibit radiation-induced apoptosis (44). Although the cell number of hepatic NK1.1+ T cells spontaneously decreased even without radiation, this result would depend on mechanisms irrelevant to Bcl-2, such as Bcl-2-independent apoptosis (43) or down-regulation of NK1.1 expression by the in vitro culture (56).
In this study we demonstrated the increase in IL-4 production and the survival of NK1.1+ T cells after the exogenous administration of DEX, a synthetic GC. To confirm that these findings could also be applied to physiologic conditions, we conducted experiments in which mice were burdened with repeated restraint stress. As a result, the concentration of GC in the serum was elevated in the stressed mice, and their spleen cells and liver MNC exhibited both an increase in IL-4 production and a preferential survival of IL-4-producing NK1.1+ T cells (K. Tamada, unpublished observation). These findings thus suggest that endogenously secreted GC may also potentially induce Th2-type cytokines through the survival of IL-4-producing NK1.1+ T cells.
In conclusion, we herein revealed a novel mechanism regarding GC-induced IL-4 production in which IL-4-producing NK1.1+ T cells were resistant to GC-induced apoptosis through their high expression of Bcl-2 and thereby preferentially survived in GC-treated mice. Our findings thus suggest that NK1.1+ T cells play an important role in the immune responses during therapy with GC, stressful events, and various types of apoptotic stimuli.
Acknowledgments
We thank Drs. J. A. Bluestone and R. T. Kubo for providing the hybridomas 145-2C11 and H57-597, respectively. We also thank Dr. S. Hamano for providing anti-IL-4 mAb, and the Nissui Pharmaceutical Co. for providing the serum-free medium. In addition, we thank Dr. B. T. Quinn for his comments on this manuscript.
Footnotes
-
↵1 This work was supported in part by a grant from the Ministry of Education, Science, and Culture of Japan.
-
↵2 Address correspondence and reprint requests to the current address of Dr. Koji Tamada, Department of Immunology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
-
↵3 Abbreviations used in this paper: GC, glucocorticoids; EAE, experimental allergic encephalomyelitis; DEX, dexamethasone; PE, phycoerythrin; BrdUrd, bromodeoxyuridine; MNC, mononuclear cells; TdT, terminal deoxynucleotidyl transferase; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick-end labeling.
- Received November 3, 1997.
- Accepted April 3, 1998.
- Copyright © 1998 by The American Association of Immunologists
References
In this issue
