Peroxisome proliferator-activated receptor (PPAR)γ is a nuclear hormone receptor primarily characterized for its effect on insulin metabolism. PPARγ ligands, used to treat human type 2 diabetes, also down-regulate most immune system cells including APCs and pathogenic T cells. These effects putatively underlie the efficacy of PPARγ ligands in treating animal models of autoimmunity, leading to projections of therapeutic potential in human autoimmunity. However, the relationship between PPARγ ligands and CD4+CD25+ regulatory T cells (Tregs) has not been examined. Specifically, no studies have examined the role of Tregs in mediating the in vivo immunoregulatory effects of PPARγ ligands, and there have been no investigations of the use of PPARγ ligands to treat autoimmunity in the absence of Tregs. We now characterize the novel relationship between ciglitazone, a thiazolidinedione class of PPARγ ligand, and both murine natural Tregs (nTregs) and inducible Tregs (iTregs). In vitro, ciglitazone significantly enhances generation of iTregs in a PPARγ-independent manner. Surprisingly, and contrary to the current paradigm, we find that, in a model of graft-vs-host disease, the immunotherapeutic effect of ciglitazone requires the presence of nTregs that express PPARγ. Overall, our results indicate that, unlike its down-regulatory effect on other cells of the immune system, ciglitazone has an enhancing effect on both iTregs and nTregs, and this finding may have important implications for using PPARγ ligands in treating human autoimmune disease.
Peroxisome proliferator-activated receptor (PPARγ2) is a ligand-dependent transcription factor and member of the nuclear hormone receptor family that has been primarily characterized for its role in regulating insulin and glucose metabolism (1, 2, 3). The ability of PPARγ ligands to enhance insulin sensitivity has led to its use in humans for the treatment of type 2 diabetes. However, we and others have demonstrated that, in vitro, PPARγ ligands are also capable of down-regulating most cells of the innate and adaptive immune system (2, 4, 5, 6, 7, 8, 9). This immunoregulatory effect of PPARγ ligands has led to numerous studies demonstrating the efficacy of PPARγ ligands in treating animal models of autoimmunity including experimental allergic encephalomyelitis, asthma, arthritis, colitis, and diabetes (8, 10, 11, 12, 13, 14, 15, 16, 17, 18). The current paradigm suggests that the effectiveness of PPARγ ligands in treating models of autoimmunity results directly from down-regulation of APCs and/or T cells (8, 12, 14, 19, 20). The success of this approach has led to excitement about the potential use of PPARγ ligands as therapeutic agents in human autoimmune disease (21).
Recently, the role of CD4+CD25+ regulatory T cells (Tregs) in immune responses has been intensely investigated. At least two subtypes of Tregs have been described: thymically derived natural Tregs (nTregs) and inducible Tregs (iTregs) generated peripherally from CD4+CD25− T effector cells (Teff) (22, 23, 24). Tregs have been demonstrated to play a crucial role in the regulation of autoimmunity and immune responses (25, 26, 27, 28). Despite the great focus on both the role of Tregs in preventing autoimmune disease and the use of PPARγ ligands in treating autoimmune disease, the relationship between PPARγ ligands and Tregs has not been studied. Specifically, no studies have examined the role of Tregs in mediating the in vivo immunoregulatory effects of PPARγ ligands, and, to date, there have been no reports of PPARγ ligands treating autoimmunity in the absence of Tregs.
In these studies, we have used the PPARγ ligand ciglitazone, an example of the widely used thiazolidinedione class of synthetic PPARγ ligands, to characterize the relationship between PPARγ ligands and both iTregs and nTregs. We report for the first time that, in vitro, ciglitazone enhances conversion of Teff to iTregs, adding a new dimension to its immunotherapeutic potential. In addition, we report that, surprisingly, nTregs are required for the in vivo therapeutic effect of ciglitazone in a murine model of graft-vs-host disease (GVHD). More significantly, we have found these nTregs must express PPARγ to mediate this effect. Finally, our studies suggest that the response of Tregs to thiazolidinediones is not down-regulatory and thus differs significantly from the response seen with other cells of the immune system.
Materials and Methods
C57BL/6 mice (wild type; WT), Ppargtm2Rev (PPARgflox/flox), and B6(C)-H2-Ab1bm12/KhEgJ (referred to as bm12) mice were obtained from The Jackson Laboratory. C57BL/6NTac-TgN (CD4-Cre) mice were purchased from Taconic Farms. All mice were maintained and bred in our facilities under specific pathogen-free conditions in accordance with the guidelines and regulations of the Center for Laboratory Animal Care, University of Connecticut Health Center. T cell PPARγ-null mice (PPARγflox/flox CD4-Cre+/− termed T-PPAR mice) were generated by crossing PPARγflox/flox with CD4-Cre mice+/− and then backcrossing to PPARγflox/flox mice. Littermates (PPARγflox/flox CD4-Cre−/−) were used as WT controls in some studies.
Reagents, cell isolation, and purification
CD4+CD25+ (Tregs) and CD4+CD25− T cells (Teff) were isolated using the Murine CD4+CD25+ T Regulatory Isolation Kit (Miltenyi Biotec). T-depleted splenocytes were obtained by depleting splenocytes with murine anti-CD4 microbeads and murine anti-CD8 microbeads (Miltenyi Biotec). Where indicated, Teff were further purified by cell sorting on the FACSVantage SE (BD Biosciences) by labeling with anti-CD4-FITC (GK1.5). Cells sorted were ≥99.99% CD4+CD25− and were used as described. Ciglitazone was purchased from BIOMOL.
Generation of iTregs
CD4+CD25− Teff were isolated as described above and then stimulated (1 × 106/ml) on plate-bound anti-CD3 Ab (5 μg/ml) in the presence of soluble anti-CD28 Ab (2 μg/ml) and, where indicated, with recombinant human (rh) TGF-β1 (R&D Systems) and with ciglitazone. Cells were cultured in RPMI 1640 supplemented with 10% FCS and 5 × 10−3 M 2-ME in 10% CO2 for 3 days. At the end of 3 days, cells were stained for intracellular Foxp3 (eBioscience) per the manufacturer’s protocol.
Function of iTregs
Naive Ly5.2+CD4+CD25− “responder cells” were isolated as described above and labeled with 2.5 μM CFSE. CFSE-labeled Ly5.2+ Teff (1.5 × 104/well) were plated in round-bottom, 96-well plates with irradiated T cell-depleted spleen (5 × 104/well) in the absence or presence of soluble anti-CD3 Ab (0.7 μg/ml). DMSO/TGF-β iTregs and 40 μM ciglitazone/TGF-β iTregs cells were harvested after 3 days, washed extensively, and cocultured (1.5 × 104/well) where indicated with responder cells. Cultures were harvested at 60 h and labeled with anti-CD45.1-biotin-conjugated Ab, followed by allophycocyanin-conjugated streptavidin. CFSE dilution was assessed on a FACSCalibur (BD Biosciences), gating on Ly5.2+ cells.
C57BL/6 Teff (1 × 105), with or without C57BL/6 Tregs or T-PPAR Tregs (1.5 × 105), were adoptively transferred i.v. into sublethally irradiated (600 rad) female bm12 mice. In some studies, littermate control mice were used as a source of WT nTregs. Bm12 recipients were placed on either a control diet or a diet containing 75 ppm ciglitazone (to deliver ∼300 μg/mouse/day) (Harlan Teklad) starting on the day of injection. In some experiments, mice on the ciglitazone diet were switched to a diet containing 10 ppm ciglitazone after 14–21 days.
Total RNA was extracted using an RNeasy Mini kit (Qiagen), cleaned with Amplification Grade DNase I (Invitrogen Life Technologies), and then subjected to cDNA synthesis using the iScript cDNA Synthesis kit (Bio-Rad). PCR was then performed using primers for PPARγ (forward, TGAGGAGAAGTCACACTCTG and reverse, TGGGTCAGCTCTTGTGAATG) and hypoxanthine phosphoribosyltransferase (HPRT) as described previously (4).
Quantification of PGE2 and IL-6
PGE2 was quantified in culture supernatants using stable isotope dilution-gas chromatography-mass spectrometry as previously described by Nichols and Maraj (29). IL-6 was quantified in culture supernatants using ELISA (Pierce; Endogen) per the manufacturer’s directions.
Statistical analysis for GVHD studies
Kaplan-Meier estimates and plots of the survival curves were obtained by using SPSS version 12. The log-rank test was used to compare survival curves. The experiment-wise significance level was set at 0.05. Exact p values for the 15 pair-wise comparisons were obtained using StatXact, version 4. A Bonferroni adjustment was used for pair-wise comparisons, resulting in declaring a given comparison as statistically significant when the corresponding p value was <0.05/15 = 0.0033.
Ciglitazone enhances the conversion of Teff to iTregs
We first asked whether the PPARγ ligand ciglitazone affects the in vitro generation of iTregs. We and others have previously reported that CD4+CD25−Foxp3− Teff can be converted to CD4+CD25+Foxp3+ iTregs when stimulated in vitro in the context of TGF-β (23, 30, 31). In the present study, WT C57BL/6 Teff were stimulated using plate-bound anti-CD3 Ab and soluble anti-CD28 Ab in the absence or presence of TGF-β. Either ciglitazone or the vehicle control DMSO was added to these cultures. After 3 days, cells were harvested and analyzed for Foxp3 expression via FACS analysis. Fig. 1⇓a depicts typical results comparing iTreg generation in the context of 40 μM ciglitazone vs DMSO. In our preculture Teff populations, Foxp3 was typically expressed by <2.5% of the cells. Confirming previous results, TGF-β (2 ng/ml) and DMSO consistently resulted in Foxp3 expression in ∼50% of the cells (23, 30, 31). Importantly, the combination of ciglitazone and TGF-β significantly enhanced conversion in a dose-dependent manner. The addition of 40 μM ciglitazone typically increased the level of Foxp3-expressing cells to 80–90%. Notably, the conversion of Foxp3− Teff to Foxp3+ iTreg was not observed with ciglitazone in the absence of TGF-β.
The Foxp3-enhancing effect of ciglitazone was noted at TGF-β concentrations from 0.2 ng/ml to 2 ng/ml. This ciglitazone-enhancing response at varying concentrations of TGF-β is shown for both 40 μM (Fig. 1⇑b) and 20 μM ciglitazone (Fig. 1⇑c). In control cultures, 2 ng/ml was the optimal TGF-β concentration for generation of Foxp3+ cells; increasing the concentration to 5 ng/ml in the absence of ciglitazone did not result in enhanced conversion (data not shown). This suggests that the ciglitazone-induced enhanced conversion, clearly noted at a TGF-β concentration of 2 ng/ml, is unlikely to result from enhanced TGF-β secretion by the Teff.
Ciglitazone-induced iTreg populations have enhanced suppressor function
To determine whether ciglitazone-induced enhancement of conversion also resulted in populations with increased functionality, we used an in vitro assay of regulatory function. Naive CFSE-labeled Ly5.2+ Teff (responders) were stimulated with anti-CD3 Ab in the presence or absence of the potentially suppressive 3-day DMSO- or 40 μM ciglitazone-converted iTreg populations (Ly5.1+). Typical results are shown in Fig. 2⇓.
Anti-CD3 Ab stimulation of the responders alone led to significant CFSE dilution (31.67% of cells dividing) (Fig. 2⇑b). In initial studies using iTreg:responder ratios higher than 1:1, we found that both the DMSO- and ciglitazone-induced populations were suppressed (data not shown). We therefore titrated down the response to a 1:1 ratio to demonstrate the putative difference between the two populations. As seen in Fig. 2⇑c, coculture of the DMSO-induced iTreg population resulted in essentially no suppression of the responder cells (28.36% of cells dividing). In contrast, coculture of the ciglitazone-induced iTreg population resulted in significant suppression of the responder cells (7.69% of cells dividing) (Fig. 2⇑d). These results confirm that the enhanced Foxp3 expression induced by ciglitazone is associated with enhanced regulatory function in the resulting population.
Ciglitazone-induced enhancement of iTreg generation is PPARγ independent
Thiazolidinediones, previously believed to function solely as PPARγ ligands, are now recognized to have significant PPARγ-independent effects (32, 33, 34, 35, 36). To address whether this newly described function of ciglitazone is mediated through PPARγ, we generated mice in which the PPARγ gene is deleted specifically in T cells (T-PPAR mice). T-PPAR mice are viable and fertile and have normal numbers and percentages of splenic populations including Foxp3+CD4+CD25+ T cells (data not shown). We and others have previously demonstrated that PPARγ mRNA or protein can only be detected after T cell activation (4, 37). As seen in Fig. 3⇓a, we first confirmed that activated WT T cells expressed PPARγ mRNA, whereas activated T-PPAR T cells did not. To address the question of PPARγ dependence of the ciglitazone effect, we next examined conversion using PPARγ-null Teff derived from T-PPAR mice. Surprisingly, we found that the ciglitazone-induced enhancement of conversion using T-PPAR Teff was comparable to the enhancement seen with WT Teff over the full range of ciglitazone concentrations tested (5–20 μM). Results of a typical experiment comparing conversion of WT and T-PPAR Teff at 5, 10, and 20 μM of ciglitazone are seen in Fig. 3⇓b. In the experiment depicted in Fig. 3⇓b, the Foxp3 expression is moderately higher in T-PPAR than WT cells following treatment with TGF-β or TGF-β plus ciglitazone. Although this has not been seen in every experiment, the Foxp3 expression by TGF-β-treated or TGF-β plus ciglitazone-treated T-PPAR cells has always been found to be at least equal to that of WT-treated populations. These results document the fact that ciglitazone’s enhancement of iTreg generation is PPARγ independent and represent a new PPARγ-independent function of thiazolidinediones.
Ciglitazone-induced enhancement is not mediated through modulation of PGE2 or IL-6
To further investigate the mechanisms underlying ciglitazone-enhanced conversion, we next examined two pathways potentially involved in this effect. It has been reported that conversion of human Teff to iTregs can be mediated through PGE2 (5–10 μg/ml) (38). In addition, it has been documented that thiazolidinediones can mediate PPARγ-independent effects in keratinocytes through the expression of cyclooxygenase-2 (35). We therefore asked whether ciglitazone’s iTreg-enhancing effects were mediated through induction of PGE2 secretion. To exclude non-T cell production of PGE2 in the magnetic bead-purified Teff population, we assayed culture supernatants from FACS-sorted Teff purified to 99.9%.
Cultures initiated with FACS-sorted Teff demonstrated both normal conversion to Foxp3-expressing iTregs and the expected enhancement of 20 and 40 μM ciglitazone (data not shown). Using mass spectrometry, we quantified PGE2 in day 1 and day 3 culture supernatants. Only very low levels of PGE2, ranging from 1 to 10 ng/ml, were detected. Moreover, PGE2 levels in supernatants from both DMSO and ciglitazone cultures on both day 1 and day 3 were not elevated over background levels (data not shown). These results suggest that ciglitazone’s iTreg-enhancing effect is not mediated through PGE2 production.
Recent reports suggest that IL-6 can modulate the conversion of Teff to iTregs (39, 40). In addition, PPARγ ligands have been shown to mediate an inhibition of IL-6 function in human multiple myeloma cells via an inhibition of STAT3 function (41, 42). Therefore, ciglitazone could theoretically affect the production of or response to IL-6 in these conversion cultures. Using FACS-sorted Teff, we found that IL-6 levels in supernatants from both DMSO and ciglitazone cultures on both day 1 and day 3 were not elevated over background levels (data not shown). These results indicate that ciglitazone’s enhancing effect is not mediated through modulation of IL-6.
In vivo ciglitazone inhibits GVHD but only in the presence of nTregs
To be comprehensive in our analysis of the role of PPARγ in Tregs, we next examined the role of PPARγ in the nTreg compartment. To do this, we asked whether PPARγ ligands mediate their ameliorative effects in murine models of autoimmunity at least in part via the nTreg population. To date, no studies have examined the role of nTregs in mediating the in vivo immunoregulatory effects of PPARγ ligands, and there have been no investigations of the use of PPARγ ligands to treat autoimmunity in the absence of Tregs.
To examine this issue, we chose a model of autoimmunity in which Teff and nTregs can be examined independently. We used the bm12 GVHD model in which adoptive transfer of WT Teff into sublethally irradiated bm12 mice normally results in the death of the recipients 14–21 days later. It has previously been reported that cotransfer of ex vivo-derived nTregs has little effect in ameliorating this model of GVHD (43).
We first assessed the efficacy of dietary-administered ciglitazone in enhancing the survival of bm12 mice adoptively transferred with WT Teff but receiving no nTregs. According to the current paradigm, ciglitazone should enhance survival through its down-regulatory effect on the transferred WT Teff population or on the recipient’s APCs. Fig. 4⇓ depicts results typically seen in these studies. Mice receiving WT Teff alone and maintained on a control (nonciglitazone) diet had a median survival time of ∼16 days. Significantly, ciglitazone offered no protection to recipients receiving Teff alone. These mice, maintained on a ciglitazone diet, had a median survival time that was not significantly different from that of mice on the control diet (Fig. 4⇓). This result suggests that ciglitazone’s putative down-regulatory effect on either the transferred WT Teff or the recipient’s APCs was not sufficient to ameliorate the GVHD.
We next examined the effect of coadoptively transferring WT nTregs with WT Teff. Mice transferred with WT Teff and WT nTregs and maintained on a control diet had a median survival time not significantly different from mice receiving WT Teff alone (Fig. 4⇑). However, when WT nTregs were coadoptively transferred with WT Teff into recipients maintained on the ciglitazone diet, a strikingly different result was seen. Such mice not only demonstrated significantly enhanced survival (p = 0.0008 compared with mice on the control diet and receiving WT Teff and WT nTregs), but in addition 35% did not succumb to GVHD (Fig. 4⇑). The combination of WT nTregs and ciglitazone was the only therapeutic approach that resulted in long-term survival for a percentage of mice. These results represent the first demonstration that ciglitazone’s ability to mediate significant clinical immunoregulation in a model of autoimmunity is dependent on the presence of nTregs. This process in turn suggests that ciglitazone’s effect on nTregs, in contrast to its effect on most cells of the innate and adaptive immune system, is not inhibiting and may be enhancing.
Ciglitazone’s immunoregulatory effect requires that nTregs express PPARγ
We next examined whether ciglitazone’s in vivo dependence on nTregs requires that PPARγ be expressed by the nTregs themselves. To address this question, we compared the effect of coadoptively transferring WT vs T-PPAR nTregs. Fig. 5⇓ depicts typical results seen using T-PPAR nTregs. We again noted a consistent lack of therapeutic effect of dietary ciglitazone when administered in the absence of cotransferred nTregs. Mice cotransferred with WT Teff and T-PPAR nTregs and maintained on a control diet had a median survival time not significantly different from mice receiving WT Teff alone (Fig. 5⇓). However, in contrast to mice receiving WT nTregs and ciglitazone, mice receiving T-PPAR nTregs and ciglitazone did not demonstrate significantly enhanced survival (p = 0.4162 compared with mice on the control diet and receiving WT Teff and T-PPAR nTregs), and all mice succumbed to GVHD (Fig. 5⇓). These results indicate that nTregs must express PPARγ for ciglitazone to have a therapeutic effect. Overall, our in vivo studies represent the first demonstration that ciglitazone’s ability to mediate significant clinical immunoregulation in a model of autoimmunity is dependent on nTregs. Furthermore, whereas ciglitazone’s effect on Teff in enhancing iTreg generation is PPARγ independent, our results indicate that the requirement for nTregs in mediating ciglitazone’s in vivo immunotherapeutic effect is PPARγ dependent.
Ciglitazone does not modulate splenic nTreg numbers or phenotype
To begin to characterize mechanisms underlying the in vivo relationship between ciglitazone and nTregs, we examined nTregs from the spleens of recipient bm12 mice 7 days after adoptive transfer with WT Teff and WT nTregs. Absolute numbers of splenocytes in recipient mice did not differ significantly between mice maintained on a control vs ciglitazone diet (range, 3.0–3.5 × 106 total/spleen). Similarly, the absolute numbers of Foxp3+ cells did not differ significantly between the two groups (range 0.38–0.44 × 106 total/spleen). In addition, CD62L and CD103, two surface proteins previously associated with enhanced in vivo nTreg function (44, 45), did not differ in expression on Foxp3+ cells derived from mice maintained on control vs ciglitazone diets (average percentage of expression: control diet, CD62L-7.3% and CD103-16.6%; ciglitazone diet, CD62L-5.6% and CD103-13.5%). These results suggest that the interdependence of ciglitazone and nTregs is not mediated by modulation of either nTreg numbers or CD62L and CD103-associated functions.
We and others have previously demonstrated that, in vitro, PPARγ ligands down-regulate most cells of the innate and adaptive immune system (2, 4, 5, 6, 7, 8, 9). These findings have led to documentation of the efficacy of PPARγ ligands in treating animal models of autoimmune disease and excitement about the potential for using PPARγ ligands as therapeutic agents in human autoimmune disease (8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 21). Based on in vitro findings, the in vivo immunoregulatory effects of PPARγ ligands are believed to be mediated through down-regulation of APCs and pathogenic T cell function (8, 12, 14, 19, 20). However, there have been no reported studies to date investigating the relationship between Tregs and PPARγ ligands. Specifically, no studies have examined the role of Tregs in mediating the in vivo immunoregulatory effects of PPARγ ligands, and there have been no investigations of the use of PPARγ ligands to treat autoimmunity in the absence of Tregs. In this study, we examined the relationship between the thiazolidinedione-class PPARγ ligand ciglitazone and Tregs. It is becoming increasingly clear that both iTregs and nTregs play an important role in immunoregulation (46), and to be comprehensive, we approached this issue by examining both iTregs and nTregs.
Initially, we examined ciglitazone’s role in the in vitro generation of iTregs from Teff. We found that ciglitazone significantly enhanced TGF-β-mediated conversion of WT Teff to Foxp3+ iTregs. Consistent with an increased frequency of Foxp3+ cells, ciglitazone-induced populations also demonstrated increased regulatory function. To begin to characterize the mechanisms underlying this enhanced conversion, we first observed that ciglitazone did not mediate conversion in the absence of TGF-β. This result suggests that ciglitazone enhances the effects of TGF-β but cannot replace the requirement for TGF-β. Secondly, we found that ciglitazone’s enhancement was seen at TGF-β concentrations as high as 2 ng/ml. Conversion has been shown to plateau at TGF-β concentrations of 2 ng/ml (23, 30). Therefore, the enhancement by ciglitazone is not likely mediated via increased TGF-β production. In addition, we found that the ciglitazone-enhanced conversion was also seen in cultures initiated using PPARγ-null Teff from T-PPAR mice. This finding represents a new example of an increasing number of PPARγ-independent functions of thiazolidinediones (32, 33, 34, 35, 36), but, significantly, it is the first demonstration in T cells of such a PPARγ-independent effect.
Finally, we found that ciglitazone’s enhancing effect was not mediated through PGE2 production or modulation of IL-6 production or responsiveness. Both of these pathways have been shown to modulate iTreg generation in vitro (38, 39, 40). Although, we have not yet identified the specific mechanisms underlying ciglitazone’s enhancing effect, our results suggest that thiazolidinediones may prove to be an important adjunct in the in vitro generation of iTregs for use in adoptive immunotherapy, and thus adds a new dimension to its immunotherapeutic potential.
We next asked whether ciglitazone’s immunoregulatory effects in autoimmune diseases, presumed to be mediated via effects on Teff or APCs, could actually be dependent on nTregs. Despite documentation of the efficacy of PPARγ ligands in numerous animal models of autoimmunity, no prior studies have investigated the efficacy of thiazolidinediones in treating autoimmunity in the absence of nTregs. To address this issue, we chose a model of autoimmunity, GVHD, in which Teff and Tregs can be independently manipulated.
Using the bm12 model of GVHD, we found surprisingly that in vivo administration of ciglitazone in the absence of nTregs afforded no protection from GVHD. This study represents the first report of a PPARγ ligand tested in the absence of Tregs as well as the first report of the lack of efficacy of a PPARγ ligand in mediating immunoregulation in a model of autoimmunity. Consistent with other studies, we also found that coadoptive transfer of nTregs to mice on the control diet did not significantly enhance survival (43). In contrast, when ciglitazone was administered along with coadoptive transfer of nTregs, significant protection from GVHD and enhanced survival were noted. Thus, these results represent the first demonstration that the ability of ciglitazone to mediate significant clinical immunoregulation in a model of autoimmunity is actually dependent on the presence of nTregs. This approach suggests that the immunotherapeutic effect of PPARγ ligands in autoimmune diseases, presumed to be mediated directly through down-regulation of Teff and APC function, may also be mediated through enhancement of nTregs. This result in turn suggests that the effect of PPARγ ligands on nTregs, in contrast to their effect on most cells of the immune system, is not inhibitory. Significantly, we also demonstrated that nTregs must express PPARγ for ciglitazone’s immunoregulatory effects to be noted, because PPARγ-null Tregs were incapable of mediating this effect.
There are at least two possible interpretations of the finding that ciglitazone’s in vivo effects require nTregs that express PPARγ. First, the immunoregulatory effects of ciglitazone may be mediated directly through nTreg-PPARγ, leading to enhanced nTreg function. Second, the immunoregulatory effects of ciglitazone may occur through nTreg enhancement but be mediated indirectly through the APCs. In this scenario, ciglitazone’s major direct effect is on the APCs, which then enhance nTregs through secretion of endogenous PPARγ ligands (37).
The expression of CD62L and CD103 by nTregs has been associated in studies with increased in vivo function of nTregs (44, 45). To characterize the possible mechanisms underlying the putative enhancement of WT nTreg function by ciglitazone, 1 wk after adoptive transfer of WT Teff and WT nTregs, we compared mice that were maintained on control vs ciglitazone diet. In such mice, both the absolute numbers of splenic Foxp3+ cells and the CD62L and CD103-expression of such cells did not differ significantly between mice maintained on control vs ciglitazone diets. Therefore, our results suggest that the in vivo expansion and survival of nTregs and the expression of these phenotypic markers do not account for the differences we find in the efficacy of nTregs in the presence or absence of ciglitazone. Future studies will further characterize this functional relationship between nTregs and ciglitazone.
Overall, our studies identify for the first time the novel interrelationships between ciglitazone and both iTregs and nTregs. Future work will investigate whether other PPARγ ligands demonstrate the in vitro and in vivo effects that we report for ciglitazone. Our finding that ciglitazone mediates its effect on Teff→iTreg generation in a PPARγ-independent fashion while mediating its effect on nTregs in a PPARγ-dependent fashion is of both theoretical and practical importance. In theoretical terms, these results further highlight the dichotomy both between Teff and Tregs and between iTregs and nTregs and could prove useful in unraveling mechanisms underlying Treg physiology. In practical terms, these interrelationships could have important implications in the treatment of human autoimmune disease. At present, there is little evidence that increasing Tregs is associated with negative consequences, but it is possible that such an increase could result in problems associated with immunosuppression such as a higher incidence of malignancy. Nevertheless, if iTregs have potential use in adoptive immunotherapy for autoimmune disease, then thiazolidinediones may prove to be an important adjunct in the in vitro generation of these populations. Moreover, as the testing of the immunotherapeutic effects of thiazolidinediones broadens to include many human autoimmune diseases, the functional status of nTregs in each disease may need to be considered.
Note added in proof.
Since the acceptance of this manuscript, we have demonstrated that T-PPAR nTregs, even in the absence of synthetic ligands, demonstrate defective function in vivo when compared to WT nTregs (Wohlfert et al., manuscript in preparation).
We thank Drs. John Tsimikas and Stephen J. Walsh for expert assistance with statistical evaluation and Dr. Amanda Marzo for critically reviewing this manuscript.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 Address correspondence and reprint requests to Dr. Robert B. Clark, Department of Immunology and Department of Medicine, University of Connecticut Health Center, Farmington, CT 06032. E-mail address:
↵2 Abbreviations used in this paper: PPAR, peroxisome proliferator-activated receptor; Treg, regulatory T cell; nTreg, natural Treg; iTreg, inducible Treg; Teff, T effector cell; GVHD, graft-vs-host disease; WT, wild type; rh, recombinant human; HPRT, hypoxanthine phosphoribosyltransferase.
- Received November 2, 2006.
- Accepted January 15, 2007.
- Copyright © 2007 by The American Association of Immunologists