IL-10-Producing CD4+CD25+ Regulatory T Cells Play a Critical Role in Granulocyte-Macrophage Colony-Stimulating Factor-Induced Suppression of
http://www.100md.com
免疫学杂志 2005年第11期
Abstract
Our earlier study showed that GM-CSF has the potential not only to prevent, but also to suppress, experimental autoimmune thyroiditis (EAT). GM-CSF-induced EAT suppression in mice was accompanied by an increase in the frequency of CD4+CD25+ regulatory T cells that could suppress mouse thyroglobulin (mTg)-specific T cell responses in vitro, but the underlying mechanism of this suppression was not elucidated. In this study we show that GM-CSF can induce dendritic cells (DCs) with a semimature phenotype, an important characteristic of DCs, which are known to play a critical role in the induction and maintenance of regulatory T cells. Adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated and mTg-primed donors into untreated, but mTg-primed, recipients resulted in decreased mTg-specific T cell responses. Furthermore, lymphocytes obtained from these donors and recipients after adoptive transfer produced significantly higher levels of IL-10 compared with mTg-primed, untreated, control mice. Administration of anti-IL-10R Ab into GM-CSF-treated mice abrogated GM-CSF-induced suppression of EAT, as indicated by increased mTg-specific T cell responses, thyroid lymphocyte infiltration, and follicular destruction. Interestingly, in vivo blockade of IL-10R did not affect GM-CSF-induced expansion of CD4+CD25+ T cells. However, IL-10-induced immunosuppression was due to its direct effects on mTg-specific effector T cells. Taken together, these results indicated that IL-10, produced by CD4+CD25+ T cells that were probably induced by semimature DCs, is essential for disease suppression in GM-CSF-treated mice.
Introduction
Experimental autoimmune thyroiditis (EAT) 3 is a well-established mouse model for Hashimoto’s thyroiditis (HT). HT is an organ-specific autoimmune disease characterized by lymphocyte infiltration of the thyroid that eventually leads to follicular destruction. In HT, thyroglobulin (Tg)-specific T cells are generated, and they migrate to the thyroid. These cells produce IFN-, which induces the expression of MHC class II on thyrocytes and results in further expansion and accumulation of activated mouse Tg (mTg)-specific T cells (1, 2, 3, 4, 5). The mechanism(s) of thyroid destruction, although not completely understood, appears to involve cytokine production by thyroid-infiltrating T cells that can facilitate apoptosis of thyrocytes through caspase activation (6, 7, 8, 9).
Although dendritic cells (DCs) are essential for the induction of an effective immune response against foreign Ags, they can also play a critical role in promoting and maintaining tolerance to self-Ags (10, 11, 12, 13, 14). Modulation of DC phenotype and maturation status in vitro and in vivo can have a profound effect on T cell activation and differentiation and may skew the immune response. Different subsets of DCs can preferentially influence a Th1- or a Th2-type response. Specifically, injection of CD8a+ DCs triggers the development of Th1 cells, whereas CD8a– DCs induce Th2-type responses to soluble Ags (15, 16, 17). Therefore, targeted expansion of a particular DC subset might be used to shift an immune response from one type to another and thereby prevent autoimmune disease development. In addition, DC maturation can be modulated using different cytokines to induce either regulatory T cells (Treg) or effector T cells (18, 19, 20, 21, 22).
Neither CD8a+ nor CD8a– DCs can induce optimal T cell responses when they are immature, but they become potent activators of T cells when they are mature (15, 16, 17). Although immature DCs, characterized by the expression of low levels of costimulatory molecules and proinflammatory cytokines, can promote anergy; semimature DCs that express significant levels of MHC class II and costimulatory molecules, but low levels of proinflammatory cytokines, compared with mature DCs can induce Treg (10, 21, 23). These observations clearly illustrate that modulation of functional properties of DCs can be an effective therapeutic approach for autoimmune conditions.
Our earlier studies (24, 25) showed that administration of GM-CSF or Flt3 ligand, potent DC growth factors, resulted in suppression or augmentation of EAT, respectively. Treatment with GM-CSF induced CD8a– DCs and caused a shift in the immune response against Tg from a Th1 response to a Th2 response, as seen by increased IL-4 production with a concomitant decrease in IFN- production. However, GM-CSF-induced suppression of EAT was associated not only with mere Th2 skewing but also with a selective expansion of CD4+CD25+ Treg that could suppress mTg-specific responses in vitro (24). CD4+CD25+ Treg play a critical role in the suppression of autoimmunity. Depletion or absence of CD4+CD25+ Treg has been shown to result in the development of autoimmune disease (26, 27). Although how CD4+CD25+ Treg suppress autoimmunity is not fully understood, suppressor cytokines, such as IL-10, have been implicated (28, 29, 30, 31, 32, 33). In GM-CSF-treated mice, there was a considerable increase in the levels of IL-10, and neutralization of IL-10 in lymphocyte cultures derived from GM-CSF-treated mice restored mTg-specific T cell responses. Furthermore, lymphocytes from GM-CSF-treated mice that were depleted of CD4+CD25+ T cells showed enhanced mTg-specific proliferation, with a concomitant decrease in the levels of IL-10 in vitro, suggesting that these cells were the source of IL-10 (24). These data implied a role for IL-10 in GM-CSF-induced suppression of EAT.
In the current study we investigated the direct role of CD4+CD25+ T cells and IL-10 in GM-CSF-induced suppression of EAT. We show that adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated mice into mTg-primed mice can suppress mTg-specific proliferation, and cells from recipient mice can produce higher levels of IL-10. Furthermore, in vivo blockade of IL-10R can abrogate GM-CSF-induced suppression and restore mTg-specific T cell responses, resulting in the development of EAT. Moreover, we observed an increase in DCs with a semimature phenotype in GM-CSF-treated mice, which suggested a putative mechanism for the induction of Treg. These data show the critical role that CD4+CD25+ T cells and IL-10 play in GM-CSF-induced suppression of EAT.
Materials and Methods
Statistical analysis
Mean, SD, and statistical significance were calculated using an SPSS application. Statistical significance was determined using the nonparametric Wilcoxon signed test. In most cases, values of individual treated and immunized groups were compared with those of untreated but immunized groups. A value of p 0.05 was considered significant.
Results
GM-CSF-induced DCs maintain semimature phenotype
To determine the effects of GM-CSF treatment on the maturation of DCs, we analyzed the expression of MHC class II and costimulatory molecules as well as the production of proinflammatory cytokines from DCs isolated from GM-CSF-treated and untreated mice before and after mTg immunization. Spleens from mice treated with GM-CSF showed increased numbers of CD11c+ cells (7.51%) compared with untreated controls (3.61%; Fig. 1A). Despite an increase in the number of DCs, expression levels of MHC class II, B7.1, B7.2, and CD40 were comparable in GM-CSF-treated and untreated mice after immunization with mTg (Fig. 1B). However, levels of proinflammatory cytokines, such as TNF-, IL-12, and IL-1, evaluated by RT-PCR, were significantly higher in DCs from untreated, mTg-immunized mice than in DCs from GM-CSF-treated, mTg-immunized mice (Fig. 1C). These data suggest that DCs from GM-CSF-treated, but not untreated, mice maintain a semimature status after mTg immunization.
CD4+CD25+ T cells from GM-CSF-treated mice suppress anti-mTg response in vivo
To determine whether CD4+CD25+ T cells from GM-CSF-treated mice can suppress mTg-specific autoimmune responses in vivo, purified CD4+CD25+ T cells from GM-CSF-treated and mTg-primed mice were adoptively transferred to untreated mice that were primed with mTg. As shown in Fig. 2A, mice receiving CD4+CD25+ T cells from GM-CSF-treated mice showed significantly lower mTg-specific proliferation compared with mTg control mice (p = 0.021). Analysis of mTg-induced cytokine production by spleen cells from different groups of mice showed similar levels of IFN- in both CD4+CD25+ T cell recipient and nonrecipient mTg control mice (Fig. 2B). In contrast, cells from recipient mice produced significantly higher amounts of IL-4 (p = 0.045) and IL-10 (p = 0.035) than nonrecipient, mTg-primed mice (Fig. 2, C and D, respectively).
IL-10 produced by CD4+CD25+ T cells is important for suppressing mTg-specific T cell response
Next, to determine whether IL-10 produced by GM-CSF-induced CD4+CD25+ T cells was responsible for suppressing mTg-specific T cell responses, T cells from untreated, mTg-primed mice were cocultured with CD4+CD25+ T cells from mTg-primed and GM-CSF-treated mice in the presence of anti-IL-10R mAbs or isotype control. T cell-depleted spleen cells (Fig. 3A) or isolated DCs (Fig. 3B) from naive mice were used as feeder cells. As shown in Fig. 3, mTg-primed T cells cultured with CD4+CD25+ T cells from mTg-primed and GM-CSF-treated mice in the presence of isotype control Ab showed reduced mTg-specific T cell proliferation relative to controls, as indicated by reduced CFSE dilution (0.72 and 8.18% vs 2.42 and 11.46%, respectively). However, the response was restored to the control levels or higher in the presence of anti-IL-10R mAb (i.e., 2.15 and 16.62%). These data suggested that IL-10 produced by CD4+CD25+ T cells is required to suppress mTg-specific proliferation.
Treatment with anti-IL-10R mAb abolishes GM-CSF-induced suppression of EAT
Next, we investigated the role of IL-10 in GM-CSF-induced suppression of EAT. The effects of IL-10 were blocked by the administration of saturating concentrations of anti-IL-10R mAb to GM-CSF-treated mice at various times during disease induction. Regardless of the time of administration, almost all animals that received anti-IL-10R mAb, with the exception of some mice treated with anti-IL-10R mAb immediately after GM-CSF treatment (i.e., GM-CSF/anti-IL-10R no. 2), showed increased mTg-specific proliferation compared with mice that received GM-CSF and isotype control mAb. A significant increase in proliferation was seen in mice that received anti-IL-10R mAb 5 days after GM-CSF treatment (i.e., GM-CSF/anti-IL-10R no. 3) or throughout the course of the disease (i.e., GM-CSF/anti-IL-10R no. 1; p = 0.001 and p = 0.005, respectively; Fig. 4A). Interestingly, we observed an increase in the frequency of CD4+CD25+ T cells in all GM-CSF-treated mice regardless of the time of administration of anti-IL-10R mAb (Fig. 4B), suggesting that blocking IL-10 had no effect on the expansion of these cells by GM-CSF-induced DCs.
Effect of GM-CSF treatment on thyroid microenvironment
To test the effects of GM-CSF on the target organ, we investigated the cell type and cytokine production in the thyroids of treated mice. GM-CSF treatment resulted in the expansion of CD8a– DCs in the periphery (Fig. 1A); however, this expansion was not reflected within the thyroid (data not shown). In contrast, there was an increase in the percentage of CD4+CD25+ T cells in the thyroids of GM-CSF-treated mice relative to untreated mice (24.57 and 20.06%, respectively; Fig. 5A). Previous studies had shown that MCP-1 preferentially attracts CD4+CD25+ T cells to the thyroid, whereas RANTES preferentially attracts CD4+ effector T cells (34). Therefore, we tested for the levels of these two chemokines. MCP-1 production was comparable among all experimental groups, whereas RANTES was undetectable (data not shown), suggesting that these chemokines could not account for the observed increase in CD4+CD25+ T cell frequency in GM-CSF-treated thyroids.
Next, we quantified cytokine production by thyrocytes and thyroid-resident lymphocytes. Although a slight increase in IL-10 production with a very small decrease in IFN- production were observed in GM-CSF-treated mice compared with mTg control mice (Fig. 5B), these differences were not significant.
Several studies have suggested that thyrocyte destruction in HT is due to Fas-mediated apoptosis through increased caspase expression. Therefore, we tested for the expression levels of Fas, FasL, and caspase 8 on thyrocytes by RT-PCR. Although we observed a slight increase in Fas expression in GM-CSF-treated mice compared with CFA and mTg control mice, there was no detectable FasL expression in any of the groups. Furthermore, there was no substantial difference in the expression levels of caspase 8 among the groups of mice (Fig. 5C).
Discussion
In this study we investigated mechanisms by which GM-CSF treatment can cause suppression of EAT. Our results showed that GM-CSF can expand DCs and maintain them in a semimatured status in vivo, promote expansion of CD4+CD25+ T cells, and induce higher levels of IL-10 production required for EAT suppression. These results further extend our earlier studies (24) in which we showed that GM-CSF treatment can expand CD8a– DCs and CD4+CD25+ Treg and suppress EAT.
Although DC function is traditionally associated with the induction of primary T cell responses, there is increasing evidence that they play a critical role in peripheral tolerance (10, 11, 12, 13, 14). DCs pass through several stages of maturation (10), and earlier studies have shown that semimatured DCs play a critical role in the induction and expansion of Treg (10, 18, 19, 20, 21, 22, 23). Because GM-CSF treatment led to an increase in the frequency of CD4+CD25+ T cells with regulatory properties (24), we asked whether GM-CSF exerted its effects by affecting DC maturation. We found that DCs from GM-CSF-treated mice displayed a semimature phenotype, as indicated by high levels of expression of MHC class II and B7 molecules, but low levels of expression of proinflammatory cytokines compared with untreated mTg control mice. This suggested that GM-CSF treatment most likely induced and/or promoted tolerance through the expansion of semimature DCs, which are known to aid in the generation of Treg (10, 18, 19, 20, 21, 22, 23).
In fact, an earlier study showed that DCs generated by culturing bone marrow precursor cells in low concentrations of GM-CSF are maturation resistant, and inoculation of these DCs pulsed with allopeptides could prolong allograft survival in vivo (35). Generation of tolerogenic DCs capable of preventing autoimmune diseases and allotransplant rejections have been reported extensively (10, 11, 12, 13, 14, 36, 37). One of the major properties of such DCs is their ability to induce generation of IL-10-producing type 1 Treg (Tr1) that do not express significant levels of CD25 unless they are activated (38, 39). However, other studies have clearly shown that immature and other tolerogenic DCs can help expand IL-10-producing CD4+CD25+ Treg (36, 37, 40), which may play an important role in the induction and differentiation of Tr1 cells (28, 41, 42).
Although several types of Treg have been described, each with a specific surface phenotype and a cytokine profile, naturally occurring CD4+CD25+ Treg, which constitute 5–10% of peripheral CD4+ T cells, are the predominant suppressors of autoreactive T cells that escape central tolerance (43, 44, 45). Previously we (24) demonstrated that CD4+CD25+ T cells from GM-CSF-treated mice could suppress the mTg-specific proliferative response of effector T cells in vitro. However, CD4+CD25+ T cells from untreated, but mTg-primed, mice failed to show similar suppression of mTg-specific responses. More interestingly, depletion of CD4+CD25+ T cells from in vitro cultures of lymphocytes from GM-CSF-treated mice restored mTg-specific proliferation (24). This showed that effector T cells were generated in GM-CSF-treated mice as they were in untreated, mTg-primed mice, but their function was suppressed by CD4+CD25+ T cells that were induced/expanded in GM-CSF-treated mice. In this study, adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated mice into mTg-primed mice resulted in a significant suppression of mTg-specific proliferation compared with mTg-primed nonrecipients. Although we cannot rule out the possibility that the transferred CD4+CD25+ T cell population contained some activated effector T cells, the suppressive property observed suggested that the population was primarily composed of CD4+CD25+ Treg. Furthermore, lymphocytes from recipient mice, upon in vitro stimulation with mTg, produced higher levels of IL-10 and IL-4 than mTg-primed controls. This indicated that adoptively transferred CD4+CD25+ T cells exerted suppressive effects on recipient effector T cells, as seen in GM-CSF-treated donor mice.
To explore the mechanism of suppression of mTg-specific responses by GM-CSF-induced CD4+CD25+ T cells, we conducted additional studies. Because in an earlier study we had ruled out a critical role for IL-4 in EAT suppression (24), and IL-10 is a critical mediator of Treg-induced suppression of effector T cell function (28, 29, 30, 31, 32, 33), we tested the role of IL-10 in both the expansion and the function of CD4+CD25+ T cells in GM-CSF-treated mice. Blockade of IL-10 function in vivo using anti-IL-10R Ab reversed the suppressive effects of CD4+CD25+ T cells from GM-CSF-treated mice on mTg-specific T cell responses in vitro and suggested a critical role for this cytokine in GM-CSF-induced suppression of EAT. Furthermore, we showed that blockade of IL-10 function in vivo completely abolished the disease-suppressive effects of GM-CSF and allowed development of EAT. Initiation of treatment with anti-IL-10R Ab at different time points during disease development allowed us to address two major questions; namely, whether IL-10 is required for the induction and/or expansion of CD4+CD25+ T cells in vivo, and whether it is required for merely suppressing autoreactive effector T cell function, resulting in consequent suppression of EAT. Our results showed that regardless of the time of treatment, blocking IL-10 abolished the EAT-suppressive capacity in a majority of mice. Interestingly, the number of CD4+CD25+ T cells was higher in all GM-CSF-treated mice, compared with untreated mice regardless of anti-IL-10R Ab treatment. Consistent with previous reports (46, 47), our results showed that IL-10 is not essential for the expansion of CD4+CD25+ T cells. However, IL-10 produced by these Treg is critical for the suppression of effector T cells.
IL-10 is a key regulator of inflammation, and it can inhibit both Th1- and Th2-type immune responses through the suppression of proinflammatory cytokines and T cell proliferative responses (48). One of the major mechanisms of IL-10-mediated suppression of T cells is through selective inhibition of the CD28 costimulatory pathway (46). However, in thyroiditis, alternative mechanisms of action of IL-10 have been proposed (7, 9, 49, 50, 51, 52). Injection of cDNA expression vectors encoding IL-10 into the thyroid can significantly inhibit lymphocyte infiltration and development of EAT and prevent progression of the disease (50). This suppressive effect of IL-10 is mediated either through enhancement of FasL expression on thyrocytes and induction of activation-induced cell death of thyroid-infiltrating T lymphocytes (51) or through potent up-regulation of antiapoptotic molecules, such as cellular FLIP and Bcl-xL, which can prevent CD95-induced apoptosis of thyrocytes (7, 9). Conversely, direct injection of IL-1 and TNF- into the thyroids of mTg-primed mice induced thyrocyte apoptosis, indicating that proinflammatory cytokines play a critical role in thyroid destruction (6). This raised the possibility that IL-10 might be mediating its effects through suppression of proinflammatory cytokine production.
To determine whether the increased IL-10 response in GM-CSF-treated mice had any effect on the thyroid microenvironment, we tested the levels of expression of various proapoptotic molecules in the thyroids of GM-CSF-treated mice. Although there was an increase in the frequency of CD4+CD25+ T cells and IL-10 production in the thyroids of GM-CSF-treated mice, there was no significant difference in the expression of proapoptotic molecules between the thyroids of GM-CSF-treated and untreated mTg control mice. In this context it is interesting to note that our results to date clearly show that GM-CSF-mediated suppression of EAT is primarily due to the direct effects of IL-10 on mTg-specific effector T cells. Studies using SCID and TCR transgenic mice are underway, and they should help elucidate the direct effects of GM-CSF on DCs and/or T cells.
In summary, it is likely that GM-CSF induced the expansion of semimatured DCs, and Tg peptide presentation by these DCs led to the expansion of CD4+CD25+ Treg. IL-10 produced by these Treg inhibited the autoimmune effector functions of mTg-specific T cells with consequent suppression of EAT. These results show the therapeutic potential of GM-CSF in EAT and other autoimmune diseases with pathogenesis similar to that of EAT.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant R21DK066634.
2 Address correspondence and reprint requests to Dr. Bellur S. Prabhakar, Department of Microbiology and Immunology (M/C 790), Room E-709, Building 935, 835 South Wolcott Avenue, Chicago, IL 60612. E-mail address: bprabhak{at}uic.edu
3 Abbreviations used in this paper: EAT, experimental autoimmune thyroiditis; DC, dendritic cell; FasL, Fas ligand; HT, Hashimoto’s thyroiditis; mTg, mouse Tg; Tg, thyroglobulin; Tr1, type 1 regulatory T cell; Treg, regulatory T cell.
Received for publication October 22, 2004. Accepted for publication March 23, 2005.
References
Pujol-Borrell, R., I. Todd, M. Londei, A. Foulis, M. Feldmann, G. F. Bottazzo. 1986. Inappropriate major histocompatibility complex class II expression by thyroid follicular cells in thyroid autoimmune disease and by pancreatic cells in type I diabetes. Mol. Biol. Med. 3: 159-165.
Hamilton, F., M. Black, M. A. Farquharson, C. Stewart, A. K. Foulis. 1991. Spatial correlation between thyroid epithelial cells expressing class II MHC molecules and interferon--containing lymphocytes in human thyroid autoimmune disease. Clin. Exp. Immunol. 83: 64-68.
Montani, V., M. Shong, S. I. Taniguchi, K. Suzuki, C. Giuliani, G. Napolitano, J. Saito, M. Saji, B. Fiorentino, A. M. Reimold, et al 1998. Regulation of major histocompatibility class II gene expression in FRTL-5 thyrocytes: opposite effects of interferon and methimazole. Endocrinology 139: 290-302.
Weetman, A. P., A. M. McGregor. 1994. Autoimmune thyroid disease: further developments in our understanding. Endocr. Rev. 15: 788-830.
Stafford, E. A., N. R. Rose. 2000. Newer insights into the pathogenesis of experimental autoimmune thyroiditis. Int. Rev. Immunol. 19: 501-533.
Wang, S. H., J. D. Bretz, E. Phelps, E. Mezosi, P. L. Arscott, S. Utsugi, J. R. Baker, Jr. 2002. A unique combination of inflammatory cytokines enhances apoptosis of thyroid follicular cells and transforms nondestructive to destructive thyroiditis in experimental autoimmune thyroiditis. J. Immunol. 168: 2470-2474
Stassi, G., D. Di Liberto, M. Todaro, A. Zeuner, L. Ricci-Vitiani, A. Stoppacciaro, L. Ruco, F. Farina, G. Zummo, R. De Maria. 2000. Control of target cell survival in thyroid autoimmunity by T helper cytokines via regulation of apoptotic proteins. Nat. Immunol. 1: 483-488.
Pinkoski, M. J., D. R. Green. 2000. Murder by proxy. Nat. Immunol. 1: 461-462(Eryn Gangi*, Chenthamarak)
Our earlier study showed that GM-CSF has the potential not only to prevent, but also to suppress, experimental autoimmune thyroiditis (EAT). GM-CSF-induced EAT suppression in mice was accompanied by an increase in the frequency of CD4+CD25+ regulatory T cells that could suppress mouse thyroglobulin (mTg)-specific T cell responses in vitro, but the underlying mechanism of this suppression was not elucidated. In this study we show that GM-CSF can induce dendritic cells (DCs) with a semimature phenotype, an important characteristic of DCs, which are known to play a critical role in the induction and maintenance of regulatory T cells. Adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated and mTg-primed donors into untreated, but mTg-primed, recipients resulted in decreased mTg-specific T cell responses. Furthermore, lymphocytes obtained from these donors and recipients after adoptive transfer produced significantly higher levels of IL-10 compared with mTg-primed, untreated, control mice. Administration of anti-IL-10R Ab into GM-CSF-treated mice abrogated GM-CSF-induced suppression of EAT, as indicated by increased mTg-specific T cell responses, thyroid lymphocyte infiltration, and follicular destruction. Interestingly, in vivo blockade of IL-10R did not affect GM-CSF-induced expansion of CD4+CD25+ T cells. However, IL-10-induced immunosuppression was due to its direct effects on mTg-specific effector T cells. Taken together, these results indicated that IL-10, produced by CD4+CD25+ T cells that were probably induced by semimature DCs, is essential for disease suppression in GM-CSF-treated mice.
Introduction
Experimental autoimmune thyroiditis (EAT) 3 is a well-established mouse model for Hashimoto’s thyroiditis (HT). HT is an organ-specific autoimmune disease characterized by lymphocyte infiltration of the thyroid that eventually leads to follicular destruction. In HT, thyroglobulin (Tg)-specific T cells are generated, and they migrate to the thyroid. These cells produce IFN-, which induces the expression of MHC class II on thyrocytes and results in further expansion and accumulation of activated mouse Tg (mTg)-specific T cells (1, 2, 3, 4, 5). The mechanism(s) of thyroid destruction, although not completely understood, appears to involve cytokine production by thyroid-infiltrating T cells that can facilitate apoptosis of thyrocytes through caspase activation (6, 7, 8, 9).
Although dendritic cells (DCs) are essential for the induction of an effective immune response against foreign Ags, they can also play a critical role in promoting and maintaining tolerance to self-Ags (10, 11, 12, 13, 14). Modulation of DC phenotype and maturation status in vitro and in vivo can have a profound effect on T cell activation and differentiation and may skew the immune response. Different subsets of DCs can preferentially influence a Th1- or a Th2-type response. Specifically, injection of CD8a+ DCs triggers the development of Th1 cells, whereas CD8a– DCs induce Th2-type responses to soluble Ags (15, 16, 17). Therefore, targeted expansion of a particular DC subset might be used to shift an immune response from one type to another and thereby prevent autoimmune disease development. In addition, DC maturation can be modulated using different cytokines to induce either regulatory T cells (Treg) or effector T cells (18, 19, 20, 21, 22).
Neither CD8a+ nor CD8a– DCs can induce optimal T cell responses when they are immature, but they become potent activators of T cells when they are mature (15, 16, 17). Although immature DCs, characterized by the expression of low levels of costimulatory molecules and proinflammatory cytokines, can promote anergy; semimature DCs that express significant levels of MHC class II and costimulatory molecules, but low levels of proinflammatory cytokines, compared with mature DCs can induce Treg (10, 21, 23). These observations clearly illustrate that modulation of functional properties of DCs can be an effective therapeutic approach for autoimmune conditions.
Our earlier studies (24, 25) showed that administration of GM-CSF or Flt3 ligand, potent DC growth factors, resulted in suppression or augmentation of EAT, respectively. Treatment with GM-CSF induced CD8a– DCs and caused a shift in the immune response against Tg from a Th1 response to a Th2 response, as seen by increased IL-4 production with a concomitant decrease in IFN- production. However, GM-CSF-induced suppression of EAT was associated not only with mere Th2 skewing but also with a selective expansion of CD4+CD25+ Treg that could suppress mTg-specific responses in vitro (24). CD4+CD25+ Treg play a critical role in the suppression of autoimmunity. Depletion or absence of CD4+CD25+ Treg has been shown to result in the development of autoimmune disease (26, 27). Although how CD4+CD25+ Treg suppress autoimmunity is not fully understood, suppressor cytokines, such as IL-10, have been implicated (28, 29, 30, 31, 32, 33). In GM-CSF-treated mice, there was a considerable increase in the levels of IL-10, and neutralization of IL-10 in lymphocyte cultures derived from GM-CSF-treated mice restored mTg-specific T cell responses. Furthermore, lymphocytes from GM-CSF-treated mice that were depleted of CD4+CD25+ T cells showed enhanced mTg-specific proliferation, with a concomitant decrease in the levels of IL-10 in vitro, suggesting that these cells were the source of IL-10 (24). These data implied a role for IL-10 in GM-CSF-induced suppression of EAT.
In the current study we investigated the direct role of CD4+CD25+ T cells and IL-10 in GM-CSF-induced suppression of EAT. We show that adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated mice into mTg-primed mice can suppress mTg-specific proliferation, and cells from recipient mice can produce higher levels of IL-10. Furthermore, in vivo blockade of IL-10R can abrogate GM-CSF-induced suppression and restore mTg-specific T cell responses, resulting in the development of EAT. Moreover, we observed an increase in DCs with a semimature phenotype in GM-CSF-treated mice, which suggested a putative mechanism for the induction of Treg. These data show the critical role that CD4+CD25+ T cells and IL-10 play in GM-CSF-induced suppression of EAT.
Materials and Methods
Statistical analysis
Mean, SD, and statistical significance were calculated using an SPSS application. Statistical significance was determined using the nonparametric Wilcoxon signed test. In most cases, values of individual treated and immunized groups were compared with those of untreated but immunized groups. A value of p 0.05 was considered significant.
Results
GM-CSF-induced DCs maintain semimature phenotype
To determine the effects of GM-CSF treatment on the maturation of DCs, we analyzed the expression of MHC class II and costimulatory molecules as well as the production of proinflammatory cytokines from DCs isolated from GM-CSF-treated and untreated mice before and after mTg immunization. Spleens from mice treated with GM-CSF showed increased numbers of CD11c+ cells (7.51%) compared with untreated controls (3.61%; Fig. 1A). Despite an increase in the number of DCs, expression levels of MHC class II, B7.1, B7.2, and CD40 were comparable in GM-CSF-treated and untreated mice after immunization with mTg (Fig. 1B). However, levels of proinflammatory cytokines, such as TNF-, IL-12, and IL-1, evaluated by RT-PCR, were significantly higher in DCs from untreated, mTg-immunized mice than in DCs from GM-CSF-treated, mTg-immunized mice (Fig. 1C). These data suggest that DCs from GM-CSF-treated, but not untreated, mice maintain a semimature status after mTg immunization.
CD4+CD25+ T cells from GM-CSF-treated mice suppress anti-mTg response in vivo
To determine whether CD4+CD25+ T cells from GM-CSF-treated mice can suppress mTg-specific autoimmune responses in vivo, purified CD4+CD25+ T cells from GM-CSF-treated and mTg-primed mice were adoptively transferred to untreated mice that were primed with mTg. As shown in Fig. 2A, mice receiving CD4+CD25+ T cells from GM-CSF-treated mice showed significantly lower mTg-specific proliferation compared with mTg control mice (p = 0.021). Analysis of mTg-induced cytokine production by spleen cells from different groups of mice showed similar levels of IFN- in both CD4+CD25+ T cell recipient and nonrecipient mTg control mice (Fig. 2B). In contrast, cells from recipient mice produced significantly higher amounts of IL-4 (p = 0.045) and IL-10 (p = 0.035) than nonrecipient, mTg-primed mice (Fig. 2, C and D, respectively).
IL-10 produced by CD4+CD25+ T cells is important for suppressing mTg-specific T cell response
Next, to determine whether IL-10 produced by GM-CSF-induced CD4+CD25+ T cells was responsible for suppressing mTg-specific T cell responses, T cells from untreated, mTg-primed mice were cocultured with CD4+CD25+ T cells from mTg-primed and GM-CSF-treated mice in the presence of anti-IL-10R mAbs or isotype control. T cell-depleted spleen cells (Fig. 3A) or isolated DCs (Fig. 3B) from naive mice were used as feeder cells. As shown in Fig. 3, mTg-primed T cells cultured with CD4+CD25+ T cells from mTg-primed and GM-CSF-treated mice in the presence of isotype control Ab showed reduced mTg-specific T cell proliferation relative to controls, as indicated by reduced CFSE dilution (0.72 and 8.18% vs 2.42 and 11.46%, respectively). However, the response was restored to the control levels or higher in the presence of anti-IL-10R mAb (i.e., 2.15 and 16.62%). These data suggested that IL-10 produced by CD4+CD25+ T cells is required to suppress mTg-specific proliferation.
Treatment with anti-IL-10R mAb abolishes GM-CSF-induced suppression of EAT
Next, we investigated the role of IL-10 in GM-CSF-induced suppression of EAT. The effects of IL-10 were blocked by the administration of saturating concentrations of anti-IL-10R mAb to GM-CSF-treated mice at various times during disease induction. Regardless of the time of administration, almost all animals that received anti-IL-10R mAb, with the exception of some mice treated with anti-IL-10R mAb immediately after GM-CSF treatment (i.e., GM-CSF/anti-IL-10R no. 2), showed increased mTg-specific proliferation compared with mice that received GM-CSF and isotype control mAb. A significant increase in proliferation was seen in mice that received anti-IL-10R mAb 5 days after GM-CSF treatment (i.e., GM-CSF/anti-IL-10R no. 3) or throughout the course of the disease (i.e., GM-CSF/anti-IL-10R no. 1; p = 0.001 and p = 0.005, respectively; Fig. 4A). Interestingly, we observed an increase in the frequency of CD4+CD25+ T cells in all GM-CSF-treated mice regardless of the time of administration of anti-IL-10R mAb (Fig. 4B), suggesting that blocking IL-10 had no effect on the expansion of these cells by GM-CSF-induced DCs.
Effect of GM-CSF treatment on thyroid microenvironment
To test the effects of GM-CSF on the target organ, we investigated the cell type and cytokine production in the thyroids of treated mice. GM-CSF treatment resulted in the expansion of CD8a– DCs in the periphery (Fig. 1A); however, this expansion was not reflected within the thyroid (data not shown). In contrast, there was an increase in the percentage of CD4+CD25+ T cells in the thyroids of GM-CSF-treated mice relative to untreated mice (24.57 and 20.06%, respectively; Fig. 5A). Previous studies had shown that MCP-1 preferentially attracts CD4+CD25+ T cells to the thyroid, whereas RANTES preferentially attracts CD4+ effector T cells (34). Therefore, we tested for the levels of these two chemokines. MCP-1 production was comparable among all experimental groups, whereas RANTES was undetectable (data not shown), suggesting that these chemokines could not account for the observed increase in CD4+CD25+ T cell frequency in GM-CSF-treated thyroids.
Next, we quantified cytokine production by thyrocytes and thyroid-resident lymphocytes. Although a slight increase in IL-10 production with a very small decrease in IFN- production were observed in GM-CSF-treated mice compared with mTg control mice (Fig. 5B), these differences were not significant.
Several studies have suggested that thyrocyte destruction in HT is due to Fas-mediated apoptosis through increased caspase expression. Therefore, we tested for the expression levels of Fas, FasL, and caspase 8 on thyrocytes by RT-PCR. Although we observed a slight increase in Fas expression in GM-CSF-treated mice compared with CFA and mTg control mice, there was no detectable FasL expression in any of the groups. Furthermore, there was no substantial difference in the expression levels of caspase 8 among the groups of mice (Fig. 5C).
Discussion
In this study we investigated mechanisms by which GM-CSF treatment can cause suppression of EAT. Our results showed that GM-CSF can expand DCs and maintain them in a semimatured status in vivo, promote expansion of CD4+CD25+ T cells, and induce higher levels of IL-10 production required for EAT suppression. These results further extend our earlier studies (24) in which we showed that GM-CSF treatment can expand CD8a– DCs and CD4+CD25+ Treg and suppress EAT.
Although DC function is traditionally associated with the induction of primary T cell responses, there is increasing evidence that they play a critical role in peripheral tolerance (10, 11, 12, 13, 14). DCs pass through several stages of maturation (10), and earlier studies have shown that semimatured DCs play a critical role in the induction and expansion of Treg (10, 18, 19, 20, 21, 22, 23). Because GM-CSF treatment led to an increase in the frequency of CD4+CD25+ T cells with regulatory properties (24), we asked whether GM-CSF exerted its effects by affecting DC maturation. We found that DCs from GM-CSF-treated mice displayed a semimature phenotype, as indicated by high levels of expression of MHC class II and B7 molecules, but low levels of expression of proinflammatory cytokines compared with untreated mTg control mice. This suggested that GM-CSF treatment most likely induced and/or promoted tolerance through the expansion of semimature DCs, which are known to aid in the generation of Treg (10, 18, 19, 20, 21, 22, 23).
In fact, an earlier study showed that DCs generated by culturing bone marrow precursor cells in low concentrations of GM-CSF are maturation resistant, and inoculation of these DCs pulsed with allopeptides could prolong allograft survival in vivo (35). Generation of tolerogenic DCs capable of preventing autoimmune diseases and allotransplant rejections have been reported extensively (10, 11, 12, 13, 14, 36, 37). One of the major properties of such DCs is their ability to induce generation of IL-10-producing type 1 Treg (Tr1) that do not express significant levels of CD25 unless they are activated (38, 39). However, other studies have clearly shown that immature and other tolerogenic DCs can help expand IL-10-producing CD4+CD25+ Treg (36, 37, 40), which may play an important role in the induction and differentiation of Tr1 cells (28, 41, 42).
Although several types of Treg have been described, each with a specific surface phenotype and a cytokine profile, naturally occurring CD4+CD25+ Treg, which constitute 5–10% of peripheral CD4+ T cells, are the predominant suppressors of autoreactive T cells that escape central tolerance (43, 44, 45). Previously we (24) demonstrated that CD4+CD25+ T cells from GM-CSF-treated mice could suppress the mTg-specific proliferative response of effector T cells in vitro. However, CD4+CD25+ T cells from untreated, but mTg-primed, mice failed to show similar suppression of mTg-specific responses. More interestingly, depletion of CD4+CD25+ T cells from in vitro cultures of lymphocytes from GM-CSF-treated mice restored mTg-specific proliferation (24). This showed that effector T cells were generated in GM-CSF-treated mice as they were in untreated, mTg-primed mice, but their function was suppressed by CD4+CD25+ T cells that were induced/expanded in GM-CSF-treated mice. In this study, adoptive transfer of CD4+CD25+ T cells from GM-CSF-treated mice into mTg-primed mice resulted in a significant suppression of mTg-specific proliferation compared with mTg-primed nonrecipients. Although we cannot rule out the possibility that the transferred CD4+CD25+ T cell population contained some activated effector T cells, the suppressive property observed suggested that the population was primarily composed of CD4+CD25+ Treg. Furthermore, lymphocytes from recipient mice, upon in vitro stimulation with mTg, produced higher levels of IL-10 and IL-4 than mTg-primed controls. This indicated that adoptively transferred CD4+CD25+ T cells exerted suppressive effects on recipient effector T cells, as seen in GM-CSF-treated donor mice.
To explore the mechanism of suppression of mTg-specific responses by GM-CSF-induced CD4+CD25+ T cells, we conducted additional studies. Because in an earlier study we had ruled out a critical role for IL-4 in EAT suppression (24), and IL-10 is a critical mediator of Treg-induced suppression of effector T cell function (28, 29, 30, 31, 32, 33), we tested the role of IL-10 in both the expansion and the function of CD4+CD25+ T cells in GM-CSF-treated mice. Blockade of IL-10 function in vivo using anti-IL-10R Ab reversed the suppressive effects of CD4+CD25+ T cells from GM-CSF-treated mice on mTg-specific T cell responses in vitro and suggested a critical role for this cytokine in GM-CSF-induced suppression of EAT. Furthermore, we showed that blockade of IL-10 function in vivo completely abolished the disease-suppressive effects of GM-CSF and allowed development of EAT. Initiation of treatment with anti-IL-10R Ab at different time points during disease development allowed us to address two major questions; namely, whether IL-10 is required for the induction and/or expansion of CD4+CD25+ T cells in vivo, and whether it is required for merely suppressing autoreactive effector T cell function, resulting in consequent suppression of EAT. Our results showed that regardless of the time of treatment, blocking IL-10 abolished the EAT-suppressive capacity in a majority of mice. Interestingly, the number of CD4+CD25+ T cells was higher in all GM-CSF-treated mice, compared with untreated mice regardless of anti-IL-10R Ab treatment. Consistent with previous reports (46, 47), our results showed that IL-10 is not essential for the expansion of CD4+CD25+ T cells. However, IL-10 produced by these Treg is critical for the suppression of effector T cells.
IL-10 is a key regulator of inflammation, and it can inhibit both Th1- and Th2-type immune responses through the suppression of proinflammatory cytokines and T cell proliferative responses (48). One of the major mechanisms of IL-10-mediated suppression of T cells is through selective inhibition of the CD28 costimulatory pathway (46). However, in thyroiditis, alternative mechanisms of action of IL-10 have been proposed (7, 9, 49, 50, 51, 52). Injection of cDNA expression vectors encoding IL-10 into the thyroid can significantly inhibit lymphocyte infiltration and development of EAT and prevent progression of the disease (50). This suppressive effect of IL-10 is mediated either through enhancement of FasL expression on thyrocytes and induction of activation-induced cell death of thyroid-infiltrating T lymphocytes (51) or through potent up-regulation of antiapoptotic molecules, such as cellular FLIP and Bcl-xL, which can prevent CD95-induced apoptosis of thyrocytes (7, 9). Conversely, direct injection of IL-1 and TNF- into the thyroids of mTg-primed mice induced thyrocyte apoptosis, indicating that proinflammatory cytokines play a critical role in thyroid destruction (6). This raised the possibility that IL-10 might be mediating its effects through suppression of proinflammatory cytokine production.
To determine whether the increased IL-10 response in GM-CSF-treated mice had any effect on the thyroid microenvironment, we tested the levels of expression of various proapoptotic molecules in the thyroids of GM-CSF-treated mice. Although there was an increase in the frequency of CD4+CD25+ T cells and IL-10 production in the thyroids of GM-CSF-treated mice, there was no significant difference in the expression of proapoptotic molecules between the thyroids of GM-CSF-treated and untreated mTg control mice. In this context it is interesting to note that our results to date clearly show that GM-CSF-mediated suppression of EAT is primarily due to the direct effects of IL-10 on mTg-specific effector T cells. Studies using SCID and TCR transgenic mice are underway, and they should help elucidate the direct effects of GM-CSF on DCs and/or T cells.
In summary, it is likely that GM-CSF induced the expansion of semimatured DCs, and Tg peptide presentation by these DCs led to the expansion of CD4+CD25+ Treg. IL-10 produced by these Treg inhibited the autoimmune effector functions of mTg-specific T cells with consequent suppression of EAT. These results show the therapeutic potential of GM-CSF in EAT and other autoimmune diseases with pathogenesis similar to that of EAT.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant R21DK066634.
2 Address correspondence and reprint requests to Dr. Bellur S. Prabhakar, Department of Microbiology and Immunology (M/C 790), Room E-709, Building 935, 835 South Wolcott Avenue, Chicago, IL 60612. E-mail address: bprabhak{at}uic.edu
3 Abbreviations used in this paper: EAT, experimental autoimmune thyroiditis; DC, dendritic cell; FasL, Fas ligand; HT, Hashimoto’s thyroiditis; mTg, mouse Tg; Tg, thyroglobulin; Tr1, type 1 regulatory T cell; Treg, regulatory T cell.
Received for publication October 22, 2004. Accepted for publication March 23, 2005.
References
Pujol-Borrell, R., I. Todd, M. Londei, A. Foulis, M. Feldmann, G. F. Bottazzo. 1986. Inappropriate major histocompatibility complex class II expression by thyroid follicular cells in thyroid autoimmune disease and by pancreatic cells in type I diabetes. Mol. Biol. Med. 3: 159-165.
Hamilton, F., M. Black, M. A. Farquharson, C. Stewart, A. K. Foulis. 1991. Spatial correlation between thyroid epithelial cells expressing class II MHC molecules and interferon--containing lymphocytes in human thyroid autoimmune disease. Clin. Exp. Immunol. 83: 64-68.
Montani, V., M. Shong, S. I. Taniguchi, K. Suzuki, C. Giuliani, G. Napolitano, J. Saito, M. Saji, B. Fiorentino, A. M. Reimold, et al 1998. Regulation of major histocompatibility class II gene expression in FRTL-5 thyrocytes: opposite effects of interferon and methimazole. Endocrinology 139: 290-302.
Weetman, A. P., A. M. McGregor. 1994. Autoimmune thyroid disease: further developments in our understanding. Endocr. Rev. 15: 788-830.
Stafford, E. A., N. R. Rose. 2000. Newer insights into the pathogenesis of experimental autoimmune thyroiditis. Int. Rev. Immunol. 19: 501-533.
Wang, S. H., J. D. Bretz, E. Phelps, E. Mezosi, P. L. Arscott, S. Utsugi, J. R. Baker, Jr. 2002. A unique combination of inflammatory cytokines enhances apoptosis of thyroid follicular cells and transforms nondestructive to destructive thyroiditis in experimental autoimmune thyroiditis. J. Immunol. 168: 2470-2474
Stassi, G., D. Di Liberto, M. Todaro, A. Zeuner, L. Ricci-Vitiani, A. Stoppacciaro, L. Ruco, F. Farina, G. Zummo, R. De Maria. 2000. Control of target cell survival in thyroid autoimmunity by T helper cytokines via regulation of apoptotic proteins. Nat. Immunol. 1: 483-488.
Pinkoski, M. J., D. R. Green. 2000. Murder by proxy. Nat. Immunol. 1: 461-462(Eryn Gangi*, Chenthamarak)