Foxp3 Inhibits RORγt-mediated IL-17A mRNA Transcription through Direct Interaction with RORγt*♦

The cytokine, transforming growth factor-β1 (TGF-β1), converts naive T cells into regulatory T cells that prevent autoimmunity. However, in the presence of interleukin (IL)-6, TGF-β1 has also been found to promote differentiation into IL-17-producing helper T (Th17) cells that are deeply involved in autoimmunity and inflammation. However, it has not been clarified how TGF-β1 and IL-6 determine such a distinct fate. Here we found that a master regulator for Th17, retinoic acid-related orphan receptor γt (RORγt), was rapidly induced by TGF-β1 regardless of the presence of IL-6. IL-6 reduced Foxp3 expression, and overexpression of Foxp3 in a T cell line resulted in a strong reduction of IL-17A expression. We have characterized the IL-17A promoter and found that RORγt binding is sufficient for activation of the minimum promoter in the HEK 293T cells. RORγt-mediated IL-17A promoter activation was suppressed by forced expression of Foxp3. Foxp3 directly interacted with RORγt through exon 2 region of Foxp3. The exon 2 region and forkhead (FKH) domain of Foxp3 were necessary for the suppression of RORγt-mediated IL-17A promoter activation. We propose that induction of Foxp3 is the mechanism for the suppression of Th17 and polarization into inducible Treg.

T helper (Th) 3 cells play a pivotal role in adaptive immunity. Upon stimulation by antigens, Th cells undergo distinct devel-opmental pathways, attaining specialized properties and effector functions. CD4 ϩ T cells are traditionally thought to differentiate into Th1 and Th2 cell subsets (1). Recently, a third subset of the polarized T cell subset Th17, characterized by the production of IL-17A, was identified and found to play an important role in autoimmune diseases, elimination of extracellular bacteria, and cancer (2)(3)(4). It has been shown that TGF-␤1 plays a pivotal role in the induction of Th17 (5). Th17 differentiation of naive T cells is initiated by IL-6 and TGF-␤1 (6 -8). In addition, IL-23 as well as IL-21 are thought to be a key cytokine for the maturation and/or maintenance of Th17 cells (9 -11). However, TGF-␤1 is a suppressor of Th1 and Th2 cell differentiation and drives the conversion of naive T cells to Foxp3-positive Th cells with a regulatory phenotype, so-called inducible regulatory T cells (iTreg) (12). These observations suggest that the presence or absence of proinflammatory cytokines might lead to opposing immune consequences induced by TGF-␤1.
ROR␥t has been shown to direct Th17 cell differentiation by inducing the IL-23 receptor (13). However, it has not been clarified whether ROR␥t is directly involved in IL-17A transcription. ROR␥t could be a STAT3 target since Th17 was not induced in STAT3-deficient T cells (14). STAT3 binding to the ROR␥t promoter may enhance its expression. However, it has been reported that ROR␥t can be induced in CD4 ϩ T cells in response to TGF-␤1 (15). Despite the existence of ROR␥t, naive T cells activated by TGF-␤1 alone differentiate into iTreg but not Th17. The mechanism in which Th17 development is suppressed by TGF-␤1 alone has not been clarified.
In this study, using promoter analysis, we found that ROR␥t potently up-regulates IL-17A reporter activity without any other factors such as NFAT. We identified the two conserved ROR-responsive elements (ROREs) on the IL-17A promoter, which are essential for ROR␥t-mediated IL-17A promoter activation. Interestingly, we found that IL-6 is not necessary for the induction of ROR␥t; TGF-␤1 alone can up-regulate ROR␥t in primary naive T cells as well as the EL4 T cell line. IL-6 rather suppressed Foxp3 expression in TGF-␤1-treated naive T cells. Thus, we hypothesized that Foxp3 negatively regulates IL-17A transcription. We found that overexpression of Foxp3 reduced IL-17A expression in the EL4 cells and inhibited ROR␥t-mediated IL-17A promoter activity in the HEK 293T cells. This suppression was partly dependent on the physical interaction between ROR␥t and Foxp3. Our study provides a novel mechanism for the regulation of Th17 development by Foxp3.

EXPERIMENTAL PROCEDURES
Mice-C57BL/6 mice were purchased from CLEA Japan, Inc. (Tokyo, Japan). Six-to twelve-week-old mice were used as experimental animals. All experiments were approved by the Animal Ethics Committee of Kyusyu University.
Plasmid Constructions-PCR was done to generate the IL-17A promoter plasmid by using mouse genomic DNA as a template. In the case of IL-17A promoter, an ϳ6-kb fragment corresponding to nucleotides from Ϫ6021 to ϩ37 relative to the determined transcriptional staring site of IL-17A gene was subcloned into pGV-basic 2 vector (TOYOINKI). Series of deletion mutants were generated by restriction enzyme digestion (Ϫ1812, Ϫ613, Ϫ301) or by PCR (Ϫ239, Ϫ153, Ϫ115, Ϫ94). The deletion mutants, Ϫ1812, Ϫ613, and Ϫ301, were constructed by the use of the restriction sites Bpu1102I, SpeI, and PstI, respectively. The point mutations (TGCAAT or GTT-GTA) in a putative RORE-1 (TGACCT) and RORE-2 (GTGTCA) were introduced by PCR. The point mutants were generated by using Ϫ239 plasmid as a template. The mouse ROR␥t cDNA was amplified by reverse transcription (RT)-PCR and was subcloned into pCMVT7 vector for T7 tag. The Foxp3 mutants were subcloned into pCMV14 vector (Sigma) for FLAG tag. A constitutively activated TGF-␤ receptor I (CA-T␤RI) was provided from Dr. Joan Massagué and described previously (24). Transfection into HEK 293T cells and EL4 cells and luciferase assay were described previously (15,17,18).
RT-PCR and Real-time PCR-The cells were lysed in RNAiso (Takara) for RNA preparation. RT-PCR was performed with a standard procedure. The expression level of GAPDH was first evaluated as an internal control. The primer sequences were as follows: GAPDH, 5Ј-ACC ACA GTC CAT GCC ATC AC-3Ј and 5Ј-TCC ACC ACC CTG TTG CTG TA-3Ј; TGF-␤1, 5Ј-TAA TGG TGG ACC GCA ACA ACG C-3Ј and 5Ј-GAC GGA ATA CAG GGC TTT CG-3Ј; Foxp3, 5Ј-CAC CCA GGA AAG ACA GCA ACC-3Ј and 5Ј-GCA AGA GCT CTT GTC CAT TGA-3Ј; ROR␥t, 5Ј-ACC TCC ACT GCC AGC TGT GTG CTG TC-3Ј and 5Ј-TCA TTT CTG CAC TTC TGC ATG TAG ACT GTC CC-3Ј; IL-17A, 5Ј-CAG CAG CGA TCA TCC CTC AAA G-3Ј and 5Ј-CAG GAC CAG GAT CTC TTG CTG-3Ј. Real-time PCR was performed on cDNA samples using the SYBR Green system (Applied Biosystems). The relative quantitation value is expressed as 2 Ϫ⌬Ct , where ⌬Ct is the difference between the mean Ct value of triplicates of the sample and of the endogenous GAPDH control.
Chromatin Immunoprecipitation (ChIP) Assay-The ChIP assay was performed using ChIP-IT express (Active Motif) according to the manufacturer's instructions. The cell extracts were immunoprecipitated with polyclonal rabbit anti-T7 tag Ab (Medical & Biological Laboratories) or polyclonal rabbit antiserum specific for acetylated histone H4 Ab (catalog number 06-866, Upstate Biotechnology) and protein G magnetic beads for 4 h at 4°C. The DNA-histone complexes were incubated for 15 min at 94°C for reverse cross-linking and then treated with protease K. Purified DNA fragments were subjected to PCR. The PCR addressed for the IL-17A promoter region Ϫ52 to Ϫ246 and was performed using the following primers: 5Ј-GCA GCT CTG CTC AGC TTC TA-3Ј and 5Ј-GGG CTT TTC TCC TTC TGT GG-3Ј. The PCR products were visualized using an ethidium bromide gel.
Statistical Analyses-The Student's paired two-tailed t test was used. Values of p Ͻ 0.05 were considered significant. All error bars shown in the figures in this article are S.D.

ROR␥t and Foxp3
Are Simultaneously Induced by TGF-␤1 in Naive T Cells-As reported previously, Foxp3 was induced in naive T cells in response to TGF-␤1 (iTreg condition), whereas IL-17A was induced and Foxp3 was reduced in response to TGF-␤1 plus IL-6 (Th17 condition) (Fig. 1A). However, the molecular basis in which IL-17A production is regulated has not been clarified. It has been proposed that TGF-␤1 plus IL-6 induces ROR␥t, which in turn induces IL-17A transcription. However, several reports raised the possibility that TGF-␤1 alone also induces the expression of both Foxp3 and ROR␥t in TCR-activated CD4 ϩ T cells (15). Therefore, both Foxp3 and ROR␥t are expressed simultaneously in a single cell in the early phase of CD4 ϩ T cell differentiation.
To analyze the relationship between Foxp3 and ROR␥t expression and IL-17A transcription, we investigated the timedependent expression patterns under the iTreg or Th17 condition using RT-PCR (Fig. 1B). In agreement with previous studies (15), we found that TGF-␤1 simultaneously induced the expression of ROR␥t and Foxp3 in CD4 ϩ CD25 Ϫ T cells (Fig.  1B). IL-17A mRNA was induced under both Th17 and iTreg conditions; however, the Th17 condition induced much higher levels of IL-17A than the iTreg condition. Importantly, ROR␥t mRNA was induced within 24 h of stimulation regardless of the

Foxp3 Suppresses ROR␥t-induced IL-17A Expression
presence or absence of IL-6. Foxp3 mRNA was detected 24 h after stimulation and reached a plateau at 48 h. IL-6 reduced the expression levels of Foxp3 at 48 -72 h. This is probably a direct effect of IL-6 on the Foxp3 promoter because IL-6 suppressed TGF-␤1-mediated Foxp3 promoter activity in primary T cells (data not shown). These data suggest that IL-17A levels were not simply determined by ROR␥t levels. Since Foxp3 and IL-17A expression seem to show opposing kinetics, we investigated the possibility that Foxp3 negatively regulates ROR␥tmediated IL-17A expression.
ROR␥t Up-regulates the IL-17A Transcription-To identify the regulatory mechanism of IL-17A transcription, we developed a simple reporter system with the use of the HEK 293T cell line. First, we cloned an ϳ6-kb DNA fragment upstream of the ATG initiation codon of the IL-17A gene using the mouse genomic data base. Next, this IL-17A promoter region was subcloned upstream of a luciferase reporter gene (termed mIL17p(Ϫ6021)-Luc) ( Fig. 2A). This IL-17A promoter was activated by an anti-CD3 Ab in the EL4 cells ( Fig. 2A). EL4 is a tumor cell line but maintains many T-cell properties and expresses ROR␥t constitutively (Fig. 3A). Then, IL-17A promoter activity was examined with or without an ROR␥t expression vector in the HEK 293T cells in which ROR␥t was not expressed. In the presence of ROR␥t, the promoter activity was significantly increased (Fig. 2A).
To define the ROR␥t binding sites within the IL-17A promoter, we generated a series of truncation mutants at the 5Ј end of the promoter (Fig. 2B). These constructs were assayed for promoter activity in the HEK 293T cells co-transfected with or without ROR␥t. The truncations deleting the 5Ј to Ϫ153-bp did not markedly affect on the ROR␥t-mediated IL-17A promoter activity. However, a further 38-bp deletion (mIL17p(Ϫ115)-Luc) significantly reduced IL-17A promoter activity, and an additional 21-bp deletion (ϪmIL17p(Ϫ94)-Luc) almost completely abrogated the ROR␥t-mediated IL-17A promoter activity (Fig. 2B). This suggests that the region from Ϫ153 to Ϫ94 contains ROREs. RORE has been shown to consist of the consensus core motif AGGTCA preceded by a 5-bp A/T-rich sequence (27). As shown in Fig. 2B, there are two potential ROREs (termed RORE-1 and RORE-2) within this region. These two ROREs and surrounding regions are highly conserved between rodent and human ( Fig.  2C), suggesting an important, evolutionarily conserved function of this region.
ROR␥t Directly Binds to RORE1/2 of the IL-17A Promoter-To confirm the importance of these two ROREs, we introduced mutations into these two sites. As expected, mutations in one of the two ROREs severely impaired the ROR␥t-mediated IL-17A promoter activity, and mutations in both ROREs completely abrogated the ROR␥t-mediated promoter activity (Fig. 2D). These data indicate that these two ROREs are essential for ROR␥t-mediated IL-17A promoter activity.
To verify that ROR␥t actually binds to the ROREs within the IL-17A promoter, we examined the binding of ROR␥t to this region by ChIP analysis. HEK 293T cells were transfected with ROR␥t cDNA, and then nuclear extracts were immunoprecipitated with anti-ROR␥t Ab. DNA co-precipitated with ROR␥t was amplified by PCR with primers containing the region Ϫ153 to Ϫ94. As shown in Fig. 2E, ROR␥t was shown to bind to the IL-17A promoter area. In addition, histone H4 molecules were more highly acetylated in the presence of ROR␥t than in the absence of ROR␥t. Taken together, these data indicate that ROR␥t up-regulates the IL-17A expression by directly binding to the IL-17A promoter region containing RORE-1/2.
Forced Expression of Foxp3 Suppresses the IL-17A Expression in EL4 Cells-To investigate the effect of Foxp3 on IL-17A expression, we then examined the levels of IL-17A in EL4 cells. As shown in Fig. 3A, EL4 cells constitutively expressed ROR␥t, and IL-17A mRNA was induced by anti-CD3 Ab stimulation. Stable expression of Foxp3 consistently reduced the expression levels of IL-17A. Neither expression of ROR␥t nor expression of TGF-␤1 was affected by Foxp3 overexpression. Thus, these data indicate that Foxp3 specifically inhibits IL-17A mRNA expression.
To define the suppressive mechanism of IL-17A expression by Foxp3, we investigated whether Foxp3 expression affects the ROR␥t-mediated IL-17A transcription in the HEK 293T cells.  JUNE 20, 2008 • VOLUME 283 • NUMBER 25

JOURNAL OF BIOLOGICAL CHEMISTRY 17005
As expected, Foxp3 strongly suppressed ROR␥t-mediated IL-17A promoter activity in a dose-dependent manner (Fig.  3B). However, Foxp3 did not affect CA-T␤RI-mediated Smad transcriptional activity assayed with a Smad-binding element reporter (SBE) (Fig. 3C). Similarly, leukemia inhibitory factorinduced STAT3 transcriptional activity was not affected by Foxp3 (data not shown). Then, we examined whether Foxp3 interferes with the association of ROR␥t with RORE-1/2 in the IL-17A promoter. Intriguingly, ChIP assay revealed that Foxp3 expression reduced ROR␥t binding as well as acetylation of histone H4 in the IL-17A promoter RORE-1/2 region (Fig. 3D). These data suggest that Foxp3 can specifically inhibit ROR␥t-mediated IL-17A transcription by reducing ROR␥t binding to the IL-17A promoter.
Foxp3 Suppresses ROR␥t Function by Physical Interaction-It has been shown that Foxp3 specifically represses the NFAT/ activating protein-1 (AP-1)-driven transcription and forms a cooperative complex with NFAT on DNA (22). However, we could not observe any direct binding of Foxp3 to the IL-17A promoter (data not shown). Since Foxp3 inhibited the minimum IL-17A promoter (Ϫ153-Luc) (data not shown), which does not contain any Foxp3 binding elements, we hypothesized that Foxp3 inhibits the ROR␥t function by direct physical interaction with ROR␥t.
To test this hypothesis, we carried out immunoprecipitation and reporter assays using a series of Foxp3 deletion mutants (Fig. 4A). We detected strong interaction between wild-type Foxp3 and ROR␥t. The FOXP3 protein encodes several functional domains, including a C2H2 zinc finger (ZnF), a leucine zipper (Zip), and a wingedhelix/forkhead (FKH) domain. It has been shown that the N-terminal 188 amino acids and FKH domain are essential for the suppression activity of Foxp3. The Zip motif has been shown to be necessary for homodimerization (28); however, the role of ZnF has not been clarified. Interestingly, an N-terminal region corresponding to the exon 2 region of Foxp3 (⌬2) was necessary for interaction with ROR␥t. However, the other regions, namely the ZnF, Zip, and FKH domains, were unnecessary (Fig. 4A).
A reporter assay revealed that the exon 2 region as well as the FKH domain was necessary for the suppression of ROR␥t-mediated IL-17A promoter activation (Fig.  4B). This was confirmed in endogenous IL-17A induction in EL4 cells. EL4 cells were transfected with wild-type (WT) Foxp3 or mutants cDNAs linked to internal ribosomal entry site-enhanced green fluorescent protein. After sorting the green fluorescent protein-positive cells, IL-17A expression was measured by RT-PCR and real-time PCR. As shown in Fig. 4C, transient overexpression of WT Foxp3, but not ⌬2 or ⌬FKH, inhibited induction of IL-17A transcription (Fig. 4C). Taken together, these observations suggest that Foxp3 interacts with ROR␥t in the exon 2 region and that the FKH domain plays a critical role in the suppression of ROR␥t-mediated IL-17A transcription.

DISCUSSION
IL-17A is a signature cytokine of Th17, a new subset of helper T cells. It has been demonstrated that forced expression of ROR␥t is sufficient for the expression of IL-17A (Th17 differentiation) in T cells (13). Recently, ROR␣ has also been shown  to induce IL-17A in T cells (29). However, the molecular basis for the activation of the IL-17A promoter by ROR␥t and its regulation by cytokines have not been clarified. This is the first report on characterization of the IL-17A promoter in vitro. IL-6 was necessary for the induction of IL-17A expression (Fig. 1A). It has been thought that IL-6 is necessary because it induces ROR␥t in collaboration with TGF-␤1. However, we and others found that ROR␥t is induced by TGF-␤1 alone. We noticed that reduced expression of Foxp3 was inversely correlated with IL-17A expression, whereas ROR␥t levels were not directly correlated with IL-17A expression. Therefore, we suspected that ROR␥t is essential for the induction of IL-17A, whereas Foxp3 may negatively regulate IL-17A expression in iTregs. This was confirmed by overexpression of Foxp3 in EL4 cells in which ROR␥t is constitutively expressed (Fig. 3A). Furthermore, we showed that Foxp3 overexpression suppressed ROR␥t-mediated IL-17A promoter activity (Fig. 3B).
Therefore, we propose that one of the functions of IL-6 on Th17 induction is the suppression of Foxp3 expression. So how does IL-6 suppress Foxp3 induction? We have reported that TGF-␤ can activate 10.2-kb Foxp3 promoter in EL4 cells and that IL-4/STAT6 inhibits this promoter activation (30). We showed that STAT6 directly bound to the Foxp3 promoter and suppressed the Foxp3 promoter activation. However, unfortunately, we could not see any effects of IL-6 or activated STAT3 on the Foxp3 promoter activation in EL4 cells as well as primary T cells. In addition, IL-6 as well as STAT3C (active form of STAT3) overexpression did not inhibit TGF-␤-mediated Smad transcriptional activity. Therefore, unlike STAT6, we still could not understand how STAT3 suppresses TGF-␤-mediated Foxp3 induction. There is a possibility that the STAT3 binding site was not included in our 10.2-kb promoter. Alternatively, epigenetic alternations on the promoter region may be induced by STAT3. We also could not rule out the possibility that genes induced by STAT3 are involved in STAT3 repression. Further

Foxp3 Suppresses ROR␥t-induced IL-17A Expression
study is necessary for clarifying the mechanism of Foxp3 repression by IL-6.
Although the molecular mechanism for the suppression of Foxp3 by IL-6/STAT3 remains to be clarified, we propose a novel mechanism for the regulation of IL-17A expression by IL-6. IL-6 suppresses Foxp3 expression, and thus, ROR␥t is able to activate the IL-17A promoter.
In this study, we established a simple reporter system that facilitates analysis of the regulatory mechanism of IL-17A transcription in the HEK 293T cells. This is the first demonstration that ROR␥t functions as a direct transcriptional activator of IL-17A. Since HEK 293T cells can be used for high throughput screening, this system may be useful for identifying genes that modify IL-17A promoter activity. Using these systems, we demonstrated that Foxp3 suppressed ROR␥t-mediated IL-17A promoter activation by a physical interaction. Recent reports have shown that Foxp3 interacted with NFAT and suppressed the effector functions of Th cells by inhibiting the activity of NFAT (22). Foxp3 has also been shown to inhibit NF-B transcriptional activity (21). A similar mechanism may be present for the suppression of ROR␥t by Foxp3. Interestingly, the interaction between Foxp3 and NFAT is reported to be dependent on the FKH domain (22), whereas we found that association between Foxp3 and ROR␥t was dependent on the exon 2 region. However, the FKH domain was still required for the suppression of ROR␥t activity. Thus, Foxp3 inhibits ROR␥t by direct binding through the exon 2 region and inhibits transcriptional activity of ROR␥t through the FKH domain.
It has been reported that the FKH domain is essential for nuclear localization and repressor functions as well as DNA binding ability of Foxp3 (28,31). In addition, a large cohort study revealed that mutations in patients with immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome were clustered primarily within the FKH domain and the leucine zipper but were also present within the N-terminal portion of the protein. Lopes et al. (28) identified a novel function of the N-terminal region, which is required for FOXP3mediated repression of transcription as well as repressor function. The N-terminal region may be important for interaction with transcription factors including ROR␥t. In contrast, we found that deletion of the zinc finger and leucine zipper motifs did not severely impair suppression of IL-17A transcription. Therefore, the mechanism of the transcriptional repressor function of Foxp3 may vary according to the transcription factors that interact with Foxp3.
In conclusion, we propose a novel mechanism for regulation of IL-17A expression by TGF-␤1 and IL-6. TGF-␤1 is important for the induction of ROR␥t, whereas IL-6 is necessary to suppress expression of the transcriptional repressor Foxp3. Foxp3 also may also regulate Th1 and Th2 through a similar mechanism.