NK T cells are a T cell subset in the human that express an invariant α-chain (Vα24invt T cells). Because of the well-described immunomodulation by glucocorticoids on activation-induced cell death (AICD), the effects of dexamethasone and anti-CD3 stimulation on Vα24invt T cell clones and CD4+ T cell clones were investigated. Dexamethasone significantly enhanced anti-CD3-mediated proliferation of Vα24invt T cells, whereas CD4+ T cells were inhibited. Addition of neutralizing IL-2 Ab partially abrogated dexamethasone-induced potentiation of Vα24invt T cell proliferation, indicating a role for autocrine IL-2 production in corticosteroid-mediated proliferative augmentation. Dexamethasone treatment of anti-CD3-stimulated Vα24invt T cells did not synergize with anti-Fas blockade in enhancing proliferation or preventing AICD. The Vα24invt T cell response to dexamethasone was dependent on the TCR signal strength. In the presence of dexamethasone, lower doses of anti-CD3 inhibited proliferation of Vα24invt T cells and CD4+ T cells; at higher doses of anti-CD3, which caused inhibition of CD4+ T cells, the Vα24invt T cell clones proliferated and were rescued from AICD. These results demonstrate significant differences in TCR signal strength required between Vα24invt T cells and CD4+ cells, and suggest important immunomodulatory consequences for endogenous and exogenous corticosteroids in immune responses.
T cells expressing the invariant Vα24JαQ (Vα24invt T cells) TCR with canonical rearrangements without N region additions are a recently described lymphocyte population in humans that are analogous to the murine Vα14Jα281 NK1 T cell (1, 2). They are restricted by CD1d, an Ag-presenting surface molecule that can present glycolipids to these invariant Vα24invt T cells. Relatively high frequencies of Vα24invt T cells, up to 1:100 of the total circulating T cell repertoire, can be observed in normal humans. They are of particular interest because of their rapid secretion of IL-4 and IFN-γ within hours of TCR engagement, suggesting an important role in immunoregulation and possibly in Th1/Th2 differentiation (3, 4, 5, 6). Although the role of NK1 T cells in Th1/Th2 differentiation is under investigation, CD1 knockout mice lacking invariant Vα14Jα281 T cells exhibit alterations in anti-CD3-induced IL-4 secretion, although they are still capable of mounting a Th2 response (7).
There are a number of observations suggesting that invariant NK T cells may function in an important regulatory role in several models of autoimmunity (8, 9, 10, 11, 12). In humans with insulin-dependent diabetes mellitus (IDDM),5 a decrease in the frequency of Vα24invt T cells in monozygotic twins discordant for IDDM was observed. Moreover, while the Vα24invt T cell clones from nondiabetic siblings secreted IL-4 and IFN-γ, the Vα24invt T cells from subjects with IDDM secreted IFN-γ, but not IL-4 (9). These alterations of Vα24invt T cells clones from identical twins discordant for IDDM suggested that environmental events contributed to the alterations in cytokine secretion. Thus, it was of interest to examine nongenetic factors regulating the function of Vα24invt, CD161a+ (NKR-P1A) T cells as compared with MHC class II-restricted CD4 T cells recognizing peptide Ags.
Among the major humoral factors that T cells encounter are endogenous glucocorticoids. Glucocorticoids have been shown to modulate Th1/Th2 cytokine secretion of T cells (13, 14, 15) and inhibit proinflammatory cytokines (16), resulting in an inhibition of proliferation (17, 18) with a wide range of effects on T cell activation. Glucocorticoids have been shown to significantly affect the disease course in several models of autoimmunity (19). For example, adrenalectomy of Lewis rats susceptible to experimental autoimmune encephalomyelitis leads to a progressive, fatal disease, whereas elevated plasma levels of glucocorticoids have been associated with spontaneous recovery from experimental autoimmune encephalomyelitis (20, 21). Thus, glucocorticoids have a clear influence on regulation of T cell function.
Unstimulated T cells normally undergo apoptosis in the presence of the synthetic glucocorticoid, dexamethasone (22). Paradoxically, AICD induced by TCR activation and dexamethasone-induced apoptosis antagonize each other in several systems (23, 24, 25, 26, 27). However, there appear to be discrepancies as to the effects of dexamethasone on activated T cells. Numerous groups have shown that dexamethasone significantly inhibits the proliferation of stimulated T cells (18, 28, 29, 30). Using T cell hybridomas, it was suggested that the reason proliferation was observed in some systems and inhibition in others was because of differences in the strength of TCR cross-linking (24). Stronger signals resulted in decreased cell death in the presence of increased amounts of dexamethasone, whereas weaker signals led to the opposite effect, proliferation.
Several mechanisms for the antagonism of dexamethasone-induced and AICD have been postulated involving both inhibition and induction of specific genes by dexamethasone. Dexamethasone has been shown to inhibit Fas ligand (31, 32), IL-2 production (17), and other activation-related genes. Additionally, GILZ, a leucine-zipper family gene, was identified as a dexamethasone-induced gene that, when transfected into T cell hybridomas, inhibited anti-CD3-induced apoptosis (33).
Finally, dexamethasone induction of cytokine receptors, notably IL-2R (26), has been proposed as an explanation for increased sensitivity to IL-2 during dexamethasone treatment, and hence increased proliferation in response to anti-CD3 in the presence of dexamethasone.
In a panel of CD4+, class II-restricted clones and Vα24invt T clones derived from twins discordant for IDDM and additional subjects, proliferation of the CD4+ clones stimulated with anti-CD3 was inhibited by dexamethasone, whereas proliferation of the Vα24invt T cell clones was augmented. This observation may be partially explained by markedly different sensitivities to AICD via strength of signal through the TCR between the two cell types, and may have implications for understanding peripheral modulation of the immune system by glucocorticoids.
Methods and Materials
Vα24invt. T cell clones (Me.10, Fc.13, Gw.4, Sb4.13) and CD4+ clones (Ob1.A12 and Ob3.E3) were generated as previously described (9, 34). CD4+ clone Sb3.D5 was generated from peripheral blood of subject Sb as a glutamic acid decarboxylase 65 (GAD)-specific clone in a manner similar to Wucherpfennig et al. (34). Briefly, PBMC purified by Ficoll-Hypaque separation were placed in media. Media was 10% autologous, heat-inactivated serum in RPMI 1640 with 10 mM HEPES buffer, 2 mM l-glutamine, and 100 U/100 μg/ml penicillin/streptomycin (all from BioWhittaker, Walkersville, MD) were pulsed with 30 μg/ml GAD for 3 h, and cells were plated at 200,000 cells/well. Whole, soluble human rGAD was generously provided by Dr. S. B. Wilson (Cancer Immunology and AIDS, Dana Farber Cancer Institute, Boston, MA). On day 7, autologous PBMC were pulsed with GAD, as described above, irradiated (5000 rad), and added at 75,000 cells/well. On day 9, 10 U/ml of human rIL-2 (Teceleukin, National Cancer Institute, Frederick, MD) was added to each well. On day 14, each well was split into four, and nonpulsed or GAD-pulsed autologous, irradiated PBMC (75,000/well) were added to two wells each of the split well. After 48 h, 1 μCi/well of [3H]thymidine was added and the wells were harvested after an additional 18 h. Wells judged positive for GAD reactivity had a stimulation index >3 and a δ cpm >1000. Positive wells were expanded on Ag, as described above for two to four more cycles. Cells were then cloned from positive wells at 0.3 cell/well with 100,000 irradiated, allogeneic feeders and 1 μg/ml PHA-P (Murex Biotech, Dartford, U.K.). Clone Sb.DP1 was generated from peripheral blood of subject Sb in an Ag-nonspecific manner by single cell-sorting CD4+ cells and grown as described (9). All clones were maintained (as described) in media that was changed every 2 to 3 days with addition of IL-2 (10 U/ml) and IL-7 (10 U/ml; Boehringer Mannheim, Indianapolis, IN). All clones were stimulated in the presence of 1 μg/ml PHA-P, IL-2, and IL-7 (10 U/ml each), irradiated, allogeneic feeders (50,000/well) and irradiated 722.221 lymphoblastoid cells (5,000/well). All experiments were performed 12 days after the last stimulation of the clones.
Clones were harvested and washed 12 days poststimulation. They were then plated at a density of 2–5 × 104 cells/well (three-six replicates of each condition, depending on the experiment) in 10% FCS in 96-well plates (Costar, Cambridge, MA) on plate-bound anti-CD3 (PharMingen, San Diego, CA) at 1 μg/ml, or in dilutions from 0.001–25 μg/ml of anti-CD3. For other experiments, cells were incubated on various concentrations of plate-bound anti-CD3 with additional factors added simultaneously to the cultures: concentrations (0.005–5000 nM in some experiments) of dexamethasone (Sigma, St. Louis, MO), anti-Fas-blocking mAb (1 μg/ml, 1/23; Boehringer Ingelheim, Ridgefield, CT), anti-IL-2-neutralizing Ab (2 μg/ml; R&D Systems, Minneapolis, MN), or IL-2 (100 U/ml). Supernatants were removed at 48 h, and cells were pulsed with 1 μ Ci/well [3H]thymidine and harvested and counted 18 h later on a beta scintillation counter (Wallace, Gaithersburg, MD).
Cell surface expression of TCR-αβ, Fas, and FasL
Vα24invt clones and CD4+ T cell clones were assayed for TCR-αβ, Fas, and FasL expression by flow cytometry 12 days after restimulation. Cells (Vα24invt clone, Gw.4 and Sb4.13; CD4+ T cell clones, Ob.3E3, Sb.DP1, and SbGAD.3D5) were stained with a pan αβ-TCR mAb or isotype-matched control Ab (Abs from Coulter-Immunotech, Miami, FL and PharMingen, respectively). Vα24invt clone, Gw.4, and CD4+ T cell clone, Ob.3E3, were examined for Fas and FasL expression by flow cytometry. Twelve days after restimulation, clones were treated with 5 μg/ml soluble anti-CD3 (PharMingen) and 5 μg/ml F(ab′)2 goat anti-mouse Ig (Sigma) with and without 5 nM dexamethasone and then stained for Fas expression (PharMingen) on day 0 (pretreatment), day 1, and day 2. For FasL expression, Gw.4 and Ob.3E3 T cell clones were treated with anti-CD3 and dexamethasone, as described for Fas expression, and with matrix metalloproteinase inhibitor KB8301 (PharMingen) at 10 μM for 2 h before staining. Cells were stained with biotin-mouse anti-human FasL and streptavidin-PE (PharMingen). Cells were analyzed by flow cytometry (FacSort; Becton Dickinson, San Jose, CA).
Cell death studies
Vα24invt clones and CD4+ T cell clones were assayed for AICD by analyzing for DNA content by flow cytometry (35, 36). Cells were incubated with 10 μg/ml plate-bound anti-CD3 at a density of 5 × 104 cells/well in 10% FCS/RPMI. Dexamethasone (5 nM) and/or 2 μg/ml anti-Fas mAb were added simultaneously. After 48 h, cells were washed and resuspended in 0.3% saponin, 50 μg/ml DNase-free RNase, 500 mM EDTA, and 5 μg/ml propidium iodide (all from Sigma) for 30 min at room temperature. Cells were washed and DNA content was analyzed using flow cytometry. Cells with a DNA content of less than that seen in the G1 cell cycle peak were considered apoptotic.
Vα24invt T cell proliferation induced by anti-CD3 is enhanced by dexamethasone, whereas CD4+ clones are inhibited
A panel of T cell clones that were either CD1d restricted and expressed the invariant Vα24JαQ TCR α-chain or were CD4+, MHC class II restricted, and expressed noninvariant TCR sequences was generated. This panel included: three IL-4-secreting Vα24invt T cell clones derived from individuals without IDDM; one non-IL-4-secreting Vα24invt T cell clone from a subject with IDDM; two MBP-reactive, CD4+ clones; one GAD-reactive CD4+ T cell clone; and one non-Ag-specific CD4+ clone. The latter two clones and one of the IL-4-secreting Vα24invt T cell clones were generated from the same patient without IDDM to show that this phenomenon was not the result of interindividual differences. The T cell clones were maintained and stimulated identically.
To examine whether culture of T cells with dexamethasone had effects on proliferation, as measured by thymidine incorporation, cultures were stimulated with 1 μg/ml anti-CD3 in the presence or absence of 5 nM dexamethasone. As expected, the proliferation of all MHC class II-restricted CD4+ clones was significantly inhibited by dexamethasone (Fig. 1,A). In marked contrast, anti-CD3-induced proliferation of all CD1d-restricted Vα24invt clones was potentiated by the dexamethasone (Fig. 1,B). There were no significant differences in response to dexamethasone between IL-4- and non-IL-4-secreting Vα24invt T cell clones. The effect of varying dexamethasone concentrations was examined on CD1d-restricted Vα24invt T cell and MHC class II-restricted T cell clones. Again, while dexamethasone significantly inhibited proliferation of the MHC class II-restricted CD4+ clone, it significantly augmented the proliferation of the CD1d-restricted Vα24invt T cell clone in a dose-dependent manner (Fig. 2). To ensure that lower levels of endogenous glucocorticoids present in serum did not contribute to the T cell response, the experiments were repeated in serum-free media with concentrations of dexamethasone as low as 5 × 10−12 M. The results were identical to that obtained with serum in the media (data not shown).
Autocrine IL-2 secretion is not the sole mechanism responsible for augmentation of proliferation of Vα24invt T cells by dexamethasone
A potential mechanism for the dexamethasone-induced augmentation of proliferation by CD1d-restricted Vα24invt T cells vs the inhibition of proliferation by MHC class II-restricted CD4+ T cells may involve autocrine IL-2 secretion (26). The rescue of MHC class II-restricted CD4+ T cells from dexamethasone-induced inhibition of proliferation by the addition of exogenous IL-2 to the culture was examined (Fig. 3). Although there was a marked enhancement in TCR-mediated proliferation without dexamethasone in the CD4+ T cell clone with addition of exogenous IL-2, significant inhibition of proliferation was observed when increasing concentrations of dexamethasone were added to the culture. In contrast, there was no significant response of CD1d-restricted Vα24invt T cells to exogenous IL-2 with increasing concentrations of dexamethasone.
Vα24invt T cells secrete large amounts of IL-2, which may protect them from dexamethasone-induced apoptosis (37, 38, 39). Anti-IL-2-neutralizing Ab was added to the culture of T cells stimulated by TCR cross-linking at either 0.1 μg/ml (Fig. 4,A) or 1 μg/ml of anti-CD3 (Fig. 4,B) with increasing concentrations of dexamethasone. The proliferation of Vα24invt T cells induced with 1 μg/ml of anti-CD3 was moderately inhibited by anti-IL-2, although dexamethasone still moderately potentiated proliferation. Interestingly, at the lower concentration of 0.1 μg/ml anti-CD3, neutralization of autocrine-secreted IL-2 resulted in suppression of proliferation at increasing concentrations of dexamethasone (Fig. 4). To determine whether increased autocrine secretion of IL-2 affected the efficiency of IL-2 neutralization, IL-2 production by Vα24invt T cell clones stimulated at both 0.1 and 1 μg/ml anti-CD3 was measured by ELISA. Although increased amounts of IL-2 produced at 1 μg/ml anti-CD3 that was not neutralized by anti-IL-2 mAb could be a potential explanation for sustained dexamethasone-mediated proliferation, the Ab concentrations used exceed the neutralization capacity of the IL-2 concentrations produced in these experiments (data not shown). These data suggest that at lower TCR signal strength, IL-2 may be required to potentiate dexamethasone-induced proliferation, whereas at higher TCR signal strength, enhancement of proliferation by dexamethasone is less IL-2 dependent.
Dexamethasone inhibited apoptosis of CD1d-restricted Vα24invt T cells and induced apoptosis of MHC class II-restricted CD4+ cells
Increases in proliferation induced by dexamethasone in the Vα24invt T cells may have been due to inhibition of cell death. This was examined by measuring the percentage of T cells undergoing cell death after TCR cross-linking, as determined by propidium iodide staining, in the presence of increasing concentrations of dexamethasone. As shown in Fig. 5, signals through the TCR with anti-CD3 mAb induced apoptosis of Vα24invt T cells, which was blocked by anti-Fas mAb blockade, indicating AICD. The inhibition of AICD by dexamethasone was examined (Fig. 5). Increasing concentrations of dexamethasone blocked the induction of AICD in the Vα24invt T cells while inducing significant levels of apoptotic cell death in the CD4+ T cell clone, and this was not affected by Fas blockade. Increasing concentrations of dexamethasone in the presence of anti-Fas blockade did not further inhibit apoptosis in the CD4+ T cell clone.
CD1d-restricted Vα24invt T cells are highly sensitive to AICD; dexamethasone does not enhance the increase in proliferation caused by Fas blockade
A potential mechanism for the antagonism of AICD by dexamethasone is the down-regulation of FasL by dexamethasone (32). If the AICD-preventative action of dexamethasone acted exclusively through a mechanism independent of the Fas-mediated pathway, one would expect to see a synergistic effect on proliferation between Fas blockade and dexamethasone. The addition of anti-Fas-blocking mAb produced a marked increase in proliferation of anti-CD3-stimulated Vα24invt T cell clones, but the combination of anti-Fas mAb and dexamethasone was not as effective as Fas blockade alone in augmenting proliferation (Fig. 6). In fact, in the presence of anti-Fas mAb, the Vα24invt T cell clone proliferation was inhibited at higher dexamethasone concentrations. Similarly, anti-Fas-blocking Ab prevented AICD of Vα24invt T cells, and this effect was not potentiated by dexamethasone (Fig. 5). The addition of anti-Fas blockade affected the proliferative capacity of the MHC class II-restricted CD4+ T cells only at lower concentrations of dexamethasone (0.05 and 0.5 nM). The proliferation of Vα24invt T cells in the presence of anti-Fas mAb alone and anti-Fas mAb and dexamethasone was dramatically augmented (Fig. 7, A and B).
We then directly examined Fas and FasL expression levels under various stimulation conditions on Vα24invt T cells and CD4+ cells (Fig. 8). Anti-CD3 stimulation in the absence or presence of dexamethasone had no effect on Fas expression of either CD4+ T cells or Vα24invt T cells (Fig. 8,A). In contrast, anti-CD3 stimulation resulted in an increase of FasL expression of both CD4+ and Vα24invt T cells, which was down-regulated by dexamethasone (Fig. 8 B). Taken together these data suggested that dexamethasone is exerting its effect at least in part by disrupting the Fas-mediated AICD pathway.
Strength of signal determines susceptibility to dexamethasone
A major functional attribute of the CD1d-restricted Vα24invt T cells is rapid secretion of cytokines with TCR cross-linking as compared with MHC class II-restricted CD4+ T cells (40). This suggested that a very strong strength of signal was mediated through the Vα24invt TCR, as compared to CD4+ T cells. One factor affecting TCR signal strength is TCR density. A panel of CD4+ clones and Vα24invt T cell clones was assayed for TCR-αβ expression on day 0 (12 days after restimulation). TCR-αβ expression on the two types of T cell clones was similar (Fig. 9). This suggested that the strength of the TCR-mediated signal of the Vα24invt T cells may be greater as compared with CD4+ T cells; this difference may explain the effect of dexamethasone on the enhancement of proliferation and protection from cell death in the Vα24invt T cells. This hypothesis was directly investigated by examining the effect of dexamethasone on proliferation with different concentrations of anti-CD3 mAb. As predicted for MHC class II-restricted CD4+ T cell clones, increasing concentrations of anti-CD3 induced greater thymidine incorporation, which was inhibited by increasing concentrations of dexamethasone (Fig. 10,A). Surprisingly, the effect of dexamethasone on proliferation of Vα24invt T cells at lower concentrations of anti-CD3 duplicated the inhibitory effects of dexamethasone observed on MHC class II-restricted CD4 T cells. Thus, in the presence of dexamethasone, at 0.01 μg/ml anti-CD3, Vα24invt T cells were inhibited. The 1.0 μg/ml of anti-CD3, which allowed dexamethasone-mediated inhibition of proliferation of the CD4+ T cell clone, was in the range of maximal proliferative response to anti-CD3 for the Vα24invt T cell clone. In general, higher doses of anti-CD3 resulted in high dose suppression of Vα24invt T cells, presumably by AICD, whereas lower doses did not induce any proliferative response (Fig. 10 B). This suggests that for a given anti-CD3 stimulus, the strength of signal via theVα24invt TCR is significantly stronger than those in CD4+ TCR and is unrelated to the cell surface density of the TCR.
CD1d-restricted Vα24invt T cells with invariant TCR α-chains represent a potentially major functional population of T cells in humans. Unlike MHC class II-restricted CD4+ T cells, Vα24invt T cells circulate in an activated state, as evidenced by medium affinity IL-2R, and rapidly secrete cytokines with TCR cross-linking (41). This suggests that CD1d-restricted Vα24invt T cells have an important early function in regulating immune responses. Corticosteroids are rapidly secreted into the circulation with infections and are a major early regulator of inflammatory responses (42). Thus, the response of Vα24invt T cells to corticosteroids may be important in directing the immune response. In this study, the biologic effects of the synthetic corticosteroid, dexamethasone, on Vα24invt T cells were compared with MHC class II-restricted CD4+ T cells. Although dexamethasone markedly inhibited proliferation by enhancing CD3-mediated AICD in CD4+ T cell clones, dexamethasone enhanced proliferation by inhibiting CD3-mediated AICD in Vα24invt T cells. The mechanism of enhanced proliferation in Vα24invt T cells was in part related to higher strengths of signals provided through the TCR.
Dexamethasone can either induce or protect from cell death, and a balance of factors is likely to dictate the response of a stimulated T cell to dexamethasone. These factors include the signal delivered through the TCR, induction of genes involved in AICD, apoptotic signals induced by dexamethasone, and interference with AICD genes by dexamethasone. These counterbalancing factors are all likely to be involved in determining the outcome of T cell activation in the presence of corticosteroids. Based on our observations, it is possible to conclude that the discrepancies in the literature regarding the effects of dexamethasone on stimulated freshly isolated T cells, hybridomas, or splenocytes may be explained by the differences in activation states, or individual responses to TCR engagement.
The augmented Vα24invt T cell proliferation induced by dexamethasone with anti-CD3 was observed only in the signal strength range of stimulation that induced TCR-mediated AICD. In contrast, CD4+ clones were inhibited by dexamethasone at concentrations of anti-CD3 250 times those that resulted in dexamethasone augmentation for Vα24invt T cell clones. Markedly increased signals through the TCR antagonize the dexamethasone-induced cell death in hybridomas (24), and this antagonism is maximal in cells undergoing vigorous AICD. In those studies, weaker signals through the TCR were incapable of overriding the dexamethasone-mediated death pathway, similar to the results of our own investigation of MHC class II-restricted CD4+ clones and in Vα24invt T cell clones at the very lowest doses of anti-CD3. It remains unclear whether increased proliferation of stimulated Vα24invt T cells in the presence of dexamethasone is directly linked to the inhibition of AICD, or whether the presence of AICD without dexamethasone is simply an accurate predictor of increased proliferation with dexamethasone addition.
Because the CD4+ and NK T cell clones were cloned, maintained, and stimulated identically and predominantly from the same individual, variations among individual sources and cloning techniques that could affect activation states were controlled. Although dexamethasone-induced enhancement of proliferation mediated by protection from AICD was demonstrated, as related to the strength of signal through the TCR, the biochemical signals regulating the protective effect are as yet unknown. Dexamethasone can indirectly inhibit AICD through induction of the protective gene GILZ, and dexamethasone can regulate FasL expression (31, 32, 33). This may occur through the transcription factors, Egr-3 (43) and Egr-2 (44); these factors may play a role in T cell survival in the presence of glucocorticoids. Future experiments will examine these pathways of activation.
Although our data are consistent with previous work demonstrating that dexamethasone acts upon the Fas-mediated death pathway (31, 32), these experiments do not exclude actions by dexamethasone on other death or antiapoptosis pathways. Dexamethasone treatment alone of anti-CD3-stimulated Vα24invt T cells does not reach the levels of proliferation or AICD inhibition induced by anti-Fas blockade. Additionally, dexamethasone treatment does not synergize with Fas blockade, in fact they appear to be somewhat antagonistic (Figs. 6 and 7). There are two possible explanations for these observations: 1) the effects of dexamethasone are exclusive to the Fas-induced AICD pathway either at the level of FasL production or further downstream and 2) dexamethasone may also be acting on other AICD pathways with its effects not as pronounced as Fas blockade; any synergy would be obscured by the effects of Fas blockade. Thus, concurrent inhibition of cellular growth or induction of other apoptotic genes by dexamethasone may explain the observed antagonism between dexamethasone and Fas blockade. Further studies examining the role of these mechanisms in Vα24invt T cell activation are important, as they may differ from those found in MHC class II-restricted CD4+ cells.
Another potential mechanism for the contrasting responses of MHC class II-restricted CD4+ T cells and CD1d-restricted Vα24invt T cells to dexamethasone was autocrine IL-2 production (26, 37, 38, 39). Neutralization of IL-2 production with anti-IL-2 mAb affected the augmentation of proliferation by dexamethasone only when the Vα24invt T cells were stimulated at low concentrations of anti-CD3. This may suggest that autocrine IL-2 production plays a role in weakly stimulated Vα24invt T cells. However, while exogenous IL-2 addition provided the CD4+ T cell clones protection from dexamethasone-induced inhibition, it did not cause dexamethasone-mediated enhancement of proliferation. Whether these observations were results of differential susceptibilities of the cell types to dexamethasone-mediated death and AICD, or due to other protective autocrine-secreted molecules is unknown.
The increased activation state of Vα24invt T cells resulting in relative ease of activation and AICD is consistent with other phenotypic characteristics described of NK T cells. Peripheral NK T cells express an array of activation markers (45), very rapidly produce cytokines when stimulated in vivo (40), and are clonally expanded, possessing a canonical TCR α rearrangement at high frequencies in the peripheral blood. This clonal expansion can occur in germfree conditions (46), consistent with the possibility that there is tonic activation of NK1.1+ cells in the periphery. It remains to be confirmed whether this higher state of activation is due to increased tonic stimulation, or to a stronger signal sent through the Vα24invt TCR as compared with other TCRs for a given stimulus. Regardless, increased susceptibility of Vα24invt T cells to dexamethasone augmentation of proliferation further suggests an in vivo mechanism for the clonal expansion of these cells.
Recently, it was demonstrated that injection of dexamethasone into mice results in increased frequency and number of NK T cells. The results were in part due to the fact that NK T cells are more resistant to dexamethasone- and radiation-induced cell death (47). In those studies, NK T cell proliferation, as measured by BrdU (5-bromo-2′-deoxyuridine) incorporation, could not explain the increased frequency and raw number of NK T cells derived from the liver after dexamethasone injection. The authors suggest this may be due to migration of cells to the liver. Proliferation may be taking place at a separate site such as the spleen or lymph nodes, resulting in increased numbers when these cells reach the liver. It was also shown that Fas-mediated apoptosis is irrelevant to the preferential survival of murine NK1 T cells by showing that Fas levels on NK1+ and NK1− cells are unchanged with dexamethasone treatment. These results are not surprising because FasL expression on lymphocytes, not Fas, is most affected in vitro studies (31, 32).
Previous investigations of T cell selection in the thymus may provide a model for the shaping of the peripheral T cell repertoire by glucocorticoids (24). Increased TCR signals result in an increased probability that glucocorticoids will enhance proliferation. Therefore, during acute stressful events such as infection when glucocorticoid levels increase, there is a positive selection for the highest affinity T cells and a negative selection for the lower affinity T cells, providing an efficient repertoire to clear infection. Tonic glucocorticoid stimulation without adequate Ag presentation would result in general T cell inhibition, which may explain why chronic infection can impair the immune response.
Our results may help further explain important aspects of immune regulation as they pertain to tolerance and autoimmunity. Increases or decreases in Vα24invt T cell frequency may have direct impact on the control of immune response and prevention of autoimmunity (9, 10, 11, 12, 45, 46, 47). Because clonal expansion of CD1d-stimulated Vα24invt T cells is augmented by glucocorticoids, glucocorticoid-mediated expansion of high affinity CD4+ T cells potentially would be accompanied by expansion of recently activated Vα24invt T cells at a sufficient rate to potentially regulate autoreactive T cell activity. In contrast, the presence of circulating corticosteroids in the absence of CD1d activation of Vα24invt T cells could potentially reduce the frequency of these regulatory T cells. Thus, circulating corticosteroids in situations of inflammation where there is increased CD1d stimulation of Vα24invt T cells may potentially regulate the nature of the T cell-mediated immune response.
We thank Dr. Betty Diamond and Dr. Byron Waksman for critical reading of the manuscript.
These studies were supported by grants (to D.A.H.) from the National Institutes of Health (R01 AI39229-01, P01 AI39671, and R01 AI44447-01) and by a grant (to D.A.H.) from the Juvenile Diabetes Foundation International (1-1998-124). J.D.M. was funded by a Howard Hughes Medical Institute Medical Student Research Training Fellowship.
Abbreviations used in this paper: IDDM, insulin-dependent diabetes mellitus; AICD, activation-induced cell death; GAD, glutamic acid decarboxylase 65.