Locus IL-10 Enhancer in the Through a STAT5-Responsive Intronic Regulatory T Cells Is Enhanced by IL-2 the Production of IL-10 B

Home , STAT5

The Production of IL-10 by Human Regulatory T Cells Is Enhanced by IL-2 through a STAT5-Responsive Intronic Enhancer in the IL-10 Locus This information is current as of September 26, 2021. Kazue Tsuji-Takayama, Motoyuki Suzuki, Mayuko Yamamoto, Akira Harashima, Ayumi Okochi, Takeshi Otani, Toshiya Inoue, Akira Sugimoto, Terumasa Toraya, Makoto Takeuchi, Fumiyuki Yamasaki, Shuji Nakamura and Masayoshi Kibata Downloaded from J Immunol 2008; 181:3897-3905; ; doi: 10.4049/jimmunol.181.6.3897 http://www.jimmunol.org/content/181/6/3897 http://www.jimmunol.org/

References This article cites 51 articles, 24 of which you can access for free at: http://www.jimmunol.org/content/181/6/3897.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Production of IL-10 by Human Regulatory T Cells Is Enhanced by IL-2 through a STAT5-Responsive Intronic Enhancer in the IL-10 Locus

Kazue Tsuji-Takayama,* Motoyuki Suzuki,* Mayuko Yamamoto,* Akira Harashima,* Ayumi Okochi,* Takeshi Otani,* Toshiya Inoue,* Akira Sugimoto,* Terumasa Toraya,* Makoto Takeuchi,* Fumiyuki Yamasaki,† Shuji Nakamura,1* and Masayoshi Kibata*

STAT5 molecules are key components of the IL-2 signaling pathway, the deficiency of which often results in autoimmune pa- thology due to a reduced number of CD4؉CD25؉ naturally occurring regulatory T (Treg) cells. One of the consequences of the IL-2-STAT5 signaling axis is up-regulation of FOXP3, a master control gene for naturally occurring Treg cells. However, the roles of STAT5 in other Treg subsets have not yet been elucidated. We recently demonstrated that IL-2 enhanced IL-10 production Downloaded from through STAT5 activation. This occurred in two types of human Treg cells: a novel type of umbilical cord blood-derived Treg cell, termed HOZOT, and Tr1-like Treg cells, IL-10-Treg. In this study, we examined the regulatory mechanisms of IL-10 production in these Treg cells, focusing specifically on the roles of STAT5. By performing bioinformatic analysis on the IL-10 locus, we identified one STAT-responsive element within intron 4, designated I-SRE-4, as an interspecies-conserved sequence. We found that I-SRE-4 acted as an enhancer element, and clustered CpGs around the I-SRE-4 were hypomethylated in IL-10-producing Treg cells, but not in other T cells. A gel-shift analysis using a nuclear extract from IL-2-stimulated HOZOT confirmed that CpG DNA http://www.jimmunol.org/ methylation around I-SRE-4 reduced STAT5 binding to the element. Chromatin immunoprecipitation analysis revealed the in situ binding of IL-2-activated STAT5 to I-SRE-4. Thus, we provide molecular evidence for the involvement of an IL-2-STAT5 signaling axis in the expression of IL-10 by human Treg cells, an axis that is regulated by the intronic enhancer, I-SRE-4, and epigenetic modification of this element. The Journal of Immunology, 2008, 181: 3897–3905.

he STAT molecules play key roles in intracellular signal revealed the contribution of STAT5 molecules to nTreg develop- transduction after cellular stimulation by cytokines, ment because a subset of these mice exhibited autoimmune pa- T growth factors, and hormones. Among seven mammalian thology very similar to IL-2 or IL-2R knockout mice (7). A human by guest on September 26, 2021 family members, two highly related STAT5 gene products, patient with a STAT5b mutation displayed immune dysregulation STAT5a and STAT5b, have been of particular interest because associated with decreased numbers and impaired suppressive abil- their activation can be induced by a wide spectrum of cytokines. ity of CD4ϩCD25high cells, indicating both developmental and Recently, especially in regulatory T (Treg)2 biology, STAT5 pro- functional defects of nTreg cells (8). An essential role of STAT5 teins have been recognized as particularly important due to their molecules in the IL-2/IL-2R signaling pathway was confirmed by essential downstream roles in the IL-2/IL-2R signaling pathway experiments using mice with an IL-2R␤ mutation, which exclu- (1–4). Naturally occurring Treg (nTreg) cells are generated and sively activates STAT5 (4). Therefore, the IL-2-STAT5 signaling maintained in an IL-2-dependent manner. This was demonstrated axis contributes to some of the essential properties of nTreg cells. in studies of mice deficient for IL-2, IL-2R␣, and IL-2R␤, in which Furthermore, one result of the IL-2-STAT5 signaling axis is the autoimmune diseases developed due to lower numbers of nTregs enhancement of nTreg-specific gene expression, the transcription (5, 6). Also, studies of STAT5a and STAT5b double-knockout mice factor FOXP3 (4). It has been shown that STAT5 molecules directly control FOXP3 gene expression, and STAT consensus sequences were identified within promoter or intron regions (4, 9). *Cell Biology Institute, Research Center, Hayashibara Biochemical Laboratories, Fu- jisaki, Okayama, Japan; and †Kurashiki Medical Center, Bakuro-cho, Kurashiki, Although roles of IL-2/IL-2R signaling were firmly established in Japan nTreg cells, its roles in other types of Treg cells are poorly Received for publication April 15, 2008. Accepted for publication July 10, 2008. understood. The costs of publication of this article were defrayed in part by the payment of page IL-10 is an inhibitory cytokine mediating suppression by regu- charges. This article must therefore be hereby marked advertisement in accordance latory or suppressor T cells (10). Despite its essential roles in im- with 18 U.S.C. Section 1734 solely to indicate this fact. mune responses, the mechanisms regulating IL-10 production are 1 Address correspondence and reprint requests to Dr. Shuji Nakamura, Cell Biology not well understood. We previously reported that IL-2 was effec- Institute, Research Center, Hayashibara Biochemical Laboratories, 675-1 Fujisaki, Okayama 702-8006, Japan. E-mail address: [email protected] tive for enhancement of IL-10 production in certain types of Treg 2 Abbreviations used in this paper: Treg, regulatory T; ARRE, Ag receptor-responsive cells (11). Such cells are categorized as IL-10-producing Treg elements; ChIP, chromatin immunoprecipitation; CNS, conserved noncoding se- cells, including Tr1-like cells, IL-10-Treg, and a newly character- quences; conT, conventional T; DC, dendritic cell; HS, hypersensitivity; I-SRE-4, ized Treg cell line (designated HOZOT) with a phenotype of STAT-responsive element within intron 4; nTreg, naturally occurring regulatory T; ϩ ϩ P-SRE, STAT-responsive element within proximal region; pAb, polyclonal Ab; CD4 CD8 . In that study, we analyzed the mechanisms of high rVISTA, Regulatory Visualization Tools for Alignment; SRE, STAT-responsive el- IL-10 production by these Treg cells in comparison with other T ement; TSS, transcription start site; UCB, umbilical cord blood. cells, especially Th2 cells, which are typical high producers of Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 IL-10 among non-Treg cells. For Th2 cells, both transcriptional www.jimmunol.org 3898 Treg PRODUCED IL-10 THROUGH AN INTRONIC ENHANCER and epigenetic mechanisms have been reported in the regulation of Systems) and CD28 mAb (37407.111; R&D Systems) (CD3/CD28) in the IL-10 production. At the transcriptional level, transcriptional acti- presence of 10 ng/ml IL-2. The IL-10-producing Treg (IL-10-Treg) cell line was also obtained from vators, such as c-jun, jun B (12), and NF-AT (13), have been ϩ Ϫ CD4 CD25 mononuclear cells, as previously described (27). The cells documented as essential molecules for IL-10 production. At an were cultured on immature DC derived from UCB in the presence of 1 nM epigenetic level, a Th2-specific transcription factor, GATA-3, has vitamin D3 (Sigma-Aldrich) and 50 nM dexamethasone (Sigma-Aldrich). been described as a chromatin remodeling molecule. GATA3 can The cells were expanded for another 2 wk in the presence of vitamin D3/dexamethasone and IL-2. bind to DNase I hypersensitivity (HS) sites residing on both the ϩ Ϫ Ј The Th2 cell line was obtained from CD4 CD25 mononuclear cells, 5 -proximal region and intron 4, and then induce remodeling of as described previously (28). The cells were cultured with plate-bound chromatin structure by increasing histone acetylation (14). In this CD3/CD28 in the presence of 10 ng/ml human rIL-4 (PeproTech) and 4 situation, GATA-3 acts as a stabilizing factor, and not a transcrip- ␮g/ml anti-human IFN-␥ mAb (Hayashibara Biochemical Laboratories) tional activator, keeping an open chromatin configuration for the for 48 h. Then, the cells were expanded by addition of 10 ng/ml IL-2. After ␥ IL-10 gene. Our previous study indicated that, in contrast to Th2 6 days, the production of IL-10, IFN- , and IL-4 was assessed in the su- pernatants from the CD3/CD28-treated cells. cells, GATA-3 plays a less important role in IL-10-producing Treg Naive T cells were prepared as CD4ϩCD25Ϫ mononuclear cells. The cells (IL-10-Treg and HOZOT) because GATA-3 expression in percentage of CD45RAϩ cells was Ͼ90%. Without cultivation, genomic these cells is relatively low compared with Th2 cells. We further DNA was isolated for methylation analysis. demonstrated that STAT5 molecules activated by IL-2/IL-2R sig- IL-10 measurements by ELISA naling were involved in the mechanism enhancing IL-10 produc- ϫ 5 tion in HOZOT, suggesting that, in addition to nTreg cells, the Cells (2 10 cells/ml) were cultivated for 16 h in the presence or absence of IL-2 in 24-well trays, which were precoated with CD3/CD28. The su- Downloaded from IL-2-STAT5 signaling axis defines another type of Treg cells, pernatants were collected and then used for IL-10 measurements by human IL-10-producing Treg. IL-10 ELISA kit (eBioscience). Recent studies of cytokine gene expression have focused on Bioinformatics epigenetic changes, because the DNA methylation status at CpG sites is often associated with cytokine expression (15). In partic- Alignment between the mouse and human IL-10 loci was performed, and ular, its status in the region of conserved noncoding sequences the extent of DNA sequence homology was computed with a web-based program called Regulatory Visualization Tools for Alignment (rVISTA; (CNS) is relevant for a cell’s differentiation or activation. Hypom- www.gsd.lbl.gov/vista) (29). The plot of the percentage of sequence iden- http://www.jimmunol.org/ ethylation is associated with chromatin’s open status, whereas hy- tity referred to the human sequence. Regions with a length of at least 100 permethylation corresponds to chromatin’s closed status (16). bp, which showed at least 75% sequence identity at each segment of the Therefore, the extent to which CpG sites are methylated correlates alignment between successive gaps, are identified as CNS and are shown with the level of cytokine production. Representative cases were in red in Fig. 1. reported for Th1-type cytokines, IL-2 (17, 18) and IFN-␥ (19, 20), Reporter gene construction and Th2-type cytokines, IL-4 (21, 22), IL-5, and IL-13 (23). With Starting with genomic DNA from a human T cell leukemia line, HUT-78, regard to IL-10, however, there are only a few reports of CpG the 5Ј-proximal region (1.5 kbp upstream of the transcription starting site methylation, and no significant correlation was found between (TSS) on the IL-10 locus) and a part of intron 4 (a 162-bp fragment in- IL-10 production levels and CpG methylation status. cluding STAT-responsive element (SRE) within intron 4 (I-SRE-4) on the by guest on September 26, 2021 In this study, we used HOZOT cells to determine how IL-2 IL-10 locus) were amplified by AccuPrime Tac (Invitrogen) using the following sets of specific primers. For the IL-10 5Ј-proximal region, the sense signaling enhanced the production of IL-10 by Treg cells. We first primer was 5Ј-GATGAAAACAGACACAGGGAGGATGAGTG and the investigated whether STAT5 molecules directly bound to the IL-10 antisense primer, 5Ј-GTCTGTCTTGTGGTTTGGTTTTGC. For intron 4 locus and whether binding elements acted as enhancers. We also (I-SRE-4 wild type), the sense primer was 5Ј-GATTCTCACTTAACCT examined the possibility of epigenetic regulation. GGAGTTGGTTCAA and the antisense primer was 5Ј-CATAGGCCG CACGGTTTCTGGGAAATCAG. For intron 4 (I-SRE-4 mutant), the sense primer was 5Ј-GATTCTCACTTAACCTGGAGTTGGTTCAA and Materials and Methods the antisense primer was 5Ј-CATAGGCCGCACGGTTTCTGAAGGAT Generation of HOZOTs and other types of T cell lines from CAG. Restriction endonuclease linkers were added to the amplification by umbilical cord blood (UCB) the second PCR. The amplified fragments were ligated into a pT7Blue TA vector (Merck). These fragments from the 5Ј-proximal region and intron 4 All T cells used in this study were generated from mononuclear cells of on the IL-10 locus were digested by XhoI/BglII and SalI/BamHI, respec- UCB, as previously reported (24). UCB was obtained from the Kurashiki tively, and then gel purified on ␤-agarose (Lonza). The fragment of the Medical Center in compliance with the institutional review board and with 5Ј-proximal region was ligated into the XhoI/BglII site of the 5Ј-multiclon- informed consent of the donors according to the Declaration of Helsinki. ing site in the luciferase expression vector, pGL4.10[luc2] (Promega), and The mononuclear cells from the UCB were prepared by gradient centrif- the fragment of intron 4 was ligated into the SalI/BamHI site at the 3Ј- ugation using Ficoll-Paque (GE Healthcare). multicloning site in pGL4.10[luc2] or pGL4.23[luc2/minP] (Promega). Human regulatory T cell lines, HOZOTs, were generated by coculture with murine stromal cell lines, as previously reported (24). HOZOT-1 and Transient transfection and reporter gene assays

-4, representative HOZOT cell lines, were used in this study. Briefly, to 6 Ϫ ConT cells (1 ϫ 10 cells) were transfected with 10 ␮g of a firefly lucif- generate HOZOTs, CD34 mononuclear cells were enriched by negative ␮ selection using a MACS CD34ϩ isolation kit (Miltenyi Biotec) and mini- erase vector and 0.5 gofaRenilla luciferase vector (pGL4.75; Promega) MACS column, according to the manufacturer’s instructions. The cells by electroporation using Gene Pulser Xcell (Bio-Rad). After transfection, were cultured over stromal cells in RPMI 1640 medium supplemented with cells were stimulated with or without 10 ng/ml IL-2 in the presence or 10% heat-inactivated FBS, 100 U/ml penicillin, and 50 ␮g/ml streptomycin absence of CD3/CD28 for 16 h. To circumvent the effects of endogenous at 37°C in 5% CO . HOZOT-1 and -4 were established by coculture with IL-2 production, anti-IL-2 polyclonal Ab (pAb) (AB-202-NA; R&D Sys- 2 tems) and anti-IL-2R mAb (22722.2; R&D Systems) (both at final con- murine stromal cell lines, MS-5 and ST2, respectively. Once established as ␮ cell lines, HOZOTs were expanded in medium containing 10 ng/ml IL-2 centrations of 5 g/ml) were added to the culture in the absence of exog- (PeproTech). HOZOTs were purified by Ficoll-Paque to deplete debris enous IL-2. Cell extracts were prepared, and the luciferase activity was from mouse stromal cell lines killed by HOZOTs before experiments. measured with the Dual Luciferase Reporter assay kit (Promega), accord- The nTreg cells were obtained by cultivation of CD25ϩ cells from UCB ing to the manufacturer’s instruction. All firefly luciferase activity was mononuclear cells by modification of previously reported methods (25, normalized to Renilla luciferase. ϩ 26). The CD25 cells were cultured on dendritic cells (DC), which were Bisulfite PCR for methylation analysis induced from GM-CSF- and IL-4-stimulated CD14ϩ cells. Conventional T (conT) cells were obtained by at least 1-wk cultivation Genomic DNA from T cells was isolated using DNeasy (Qiagen), accord- of CD4ϩCD25Ϫ cells on plates coated with anti-CD3 mAb (UCHT1; R&D ing to the manufacturer’s instructions. Bisulfite PCR was performed, as The Journal of Immunology 3899

FIGURE 1. Bioinformatic analysis of the conserved SRE on the IL-10 locus. The locations of the 5Ј-proximal region, exons, and introns of the IL-10 locus are illustrated at the top. Yellow boxes show exons (E) 1–5, and the blue lines show the 5Ј-proximal region and introns (I) 1–4 in the IL-10 locus. Corresponding sequences between human and mouse were aligned by rVISTA 2.0. The degree of interspecies conservation of the DNA within this segment is represented by the histogram. CNS (interspecies conservation more than 75%) are emphasized in red. SREs on the human IL-10 locus were identiﬁed using the TRANSFAC database, and are displayed with red stripes on a gray line under the histograms. The detailed sequences spanning P-SRE and

I-SRE-4 are depicted at the bottom. Blue and red squares show positions of CpG dinucleotides and STAT consensus sequences, respectively. Downloaded from described previously (30). Briefly, the genomic DNA (5 ␮g) was digested with EcoRI and then incubated in 3.2 M sodium bisulfite (Sigma-Aldrich) and 500 nM hydroquinone (Sigma-Aldrich) for 16 h at 55°C and then desalted using a DNA cleanup column (Promega). The bisulfite-converted

DNA was treated with 300 mM NaOH for 20 min at 37°C and then stored http://www.jimmunol.org/ at Ϫ80°C. Using the treated DNA, hot-start PCR for IL-10 5Ј-proximal region and intron 4 was performed with AccuPrime TaqDNA polymerase (Invitrogen). The primer sets for their ampliﬁcation are as follows: IL-10 5Ј-proximal region, 5Ј-GTATAGTTGGGGTGGGGGATAGTTGAAGAG and 5Ј-CTCTATCCCCCTTTTATATTATAAACTCAC; IL-10 intron 4, 5Ј-GGTTTAGTTAGTGAGGAGTTGTTGTTTTG and 5Ј-CTATACATA CCTTCTTTTACAAATC. The PCR products were cloned into a pT7Blue TA vector, and 20 clones from each sample were picked up for DNA sequencing.

Gel-shift and competition assays by guest on September 26, 2021 HOZOT-1 was treated with 10 ng/ml IL-2 for 2 h. Nuclear extracts were prepared from the cells, as described previously (31). The sequences of the unmethylated and methylated oligonucleotides used in this study are shown in Fig. 5, A and B. These oligonucleotides were purchased from Sigma- Aldrich. The double-stranded oligonucleotide probes were end labeled with [␥32P]ATP using T4 polynucleotide kinase (Invitrogen), according to the manufacturer’s instruction. Nuclear extracts (5–10 ␮g) were incubated with 20 fmol double-stranded probes in binding solution (1 mM EDTA, 30 mM NaCl, 5% glycerol, 0.1% Nonidet P-40, 1 mg/ml BSA, and 2 ␮g/ml poly(dI-dC)) at room temperature for 30 min. Unlabeled probes were added with a 100-fold excess as competitors. Anti-STAT5 pAb (C-17), anti- STAT3 pAb (H-190), or control rabbit pAb (DakoCytomation) were added to the binding reaction for supershift experiments. The reactions were sep- arated on 4% nondenaturing polyacrylamide gels and exposed to autora- diographic film for 24 h. Western blot analysis Western blotting was performed using whole cell extracts from HOZOT-1 FIGURE 2. Reporter assay demonstrates I-SRE-4 enhancer activity. A, cells treated with or without IL-2, as described previously (32, 33). The Reporter constructs were generated by combining a firefly luciferase gene with membranes were probed with anti-phospho-STAT5 mAb (8-5-2; Upstate a promoter from the IL-10 5Ј-proximal region or TATA-containing sequence Biotechnology), anti-phospho-STAT3 mAb (B-7; Santa Cruz Biotechnol- (TATA), and with an enhancer driven from I-SRE-4-containing sequence with 701 ogy), or anti-phospho-STAT1 pAb (Tyr ; Santa Cruz Biotechnology), or without mutation (I-SRE-4 mutant or wild, respectively). ConT cells were and then assessed with a chemiluminescent system (SuperSignal West Pico transfected with these constructs and a Renilla luciferase construct, and then Chemiluminescent Substrate; Pierce). After treatment with a deprobing treated with 10 ng/ml IL-2 and CD3/CD28 stimulation. After overnight cul- solution containing 62.5 nM Tris-HCl (pH 6.8), 2% SDS, and 100 nM tivation, cells were harvested and assayed for luciferase activity. The amount 2-ME for 60 min at 55°C, the membrane was used for second detection with anti-STAT5b pAb (C-17; Santa Cruz Biotechnology), anti-STAT3 of luciferase activity was shown by normalization to Renilla luciferase activity ,p Ͻ 0.01 compared with the value of lane 3. B ,ءء .pAb (H-190; Santa Cruz Biotechnology), or anti-STAT1 pAb (E-23; Santa for each transfection Cruz Biotechnology). All of the data were representative of more than three Effects of endogenous IL-2 were examined in the same reporter assay. To independent experiments. exclude endogenous IL-2, Abs for IL-2/IL-2R were added for 2 h before stimulation. Subsequently, conT cells were transfected with the indicated con- Chromatin immunoprecipitation (ChIP) assays structs or null vector. The transfected cells were treated with the indicated p Ͻ 0.01 compared ,ءء .HOZOT-1 cells (2 ϫ 107) were treated with IL-2 (10 ng/ml) for 4 h. The combination of IL-2 and/or CD3/CD28 stimulation cells were fixed with 1% paraformaldehyde for 10 min at room tempera- with the value of lane 3. The data are given as means Ϯ SD for triplicate ture, and then the fixation was stopped with 1.25 M glycine. Fixed cells cultures and are representative of three independent experiments. 3900 Treg PRODUCED IL-10 THROUGH AN INTRONIC ENHANCER

Table I. IL-10 production (pg/ml)a Results Bioinformatic approach to the identification of SREs in the (Ϫ) CD3/CD28 IL-10 locus (Ϫ)(ϩ)IL-2 (Ϫ)(ϩ)IL-2 We previously reported that IL-2 treatment could enhance IL-10 production in certain types of Treg cells, so-called IL-10-produc- HOZOT-1 Ͻ355Ϯ 3.2 1503 Ϯ 116 5197 Ϯ 86 HOZOT-4 Ͻ3 Ͻ3 198 Ϯ 0.4 1558 Ϯ 78 ing Treg, including IL-10-Treg and HOZOT (24). We also dem- IL-10-Treg Ͻ3 Ͻ3 120 Ϯ 10 2300 Ϯ 150 onstrated that the enhancement was mediated through STAT5 ac- Th2 Ͻ3 Ͻ3 2018 Ϯ 8 2214 Ϯ 37 tivation (11). To examine whether STAT5 molecules could Ͻ Ͻ Ϯ Ϯ nTreg 3 361.5 14 1.8 directly bind to the IL-10 locus, we performed a bioinformatic conT Ͻ3 Ͻ322Ϯ 0.8 44 Ϯ 4.5 search for SREs on the IL-10 locus using web-based software, a T cells were stimulated with CD3/CD28 in the presence or absence of 10 ng/mL rVISTA. This program can identify interspecies-conserved se- IL-2 for 16 h. The levels of IL-10 production were determined by ELISA. The data are given as means Ϯ SD for triplicate samples and are representative of three inde- quences for specific transcription factors by linking to the most pendent experiments. widely used database, TRANSFAC (29). Comparative sequence analyses among vertebrates is quite effective for finding functional coding and noncoding elements (34). As shown in Fig. 1, SREs located in noncoding or coding regions were distributed over the were washed with cold PBS. Cells were treated with lysis buffer (Santa entire human IL-10 locus, including the 5Ј-proximal region, in- Cruz Biotechnology) and sonicated five times (30 s each) to prepare chro- trons 2 and 4, and exon 5. A total of six SREs were found, five of matin extracts. The extracts were pretreated with salmon sperm DNA- Downloaded from blocked protein G beads (GE Healthcare) and then immunoprecipitated which resided in the noncoding region. Among them, the inter- with anti-acetyl histone H3 pAb (Upstate Biotechnology), anti-acetyl his- species-conserved SRE was found only within intron 4 (at position tone H4 pAb (Upstate Biotechnology), anti-STAT5 pAb (C-17; Santa Cruz ϩ3200 bp relative to the TSS), whereas the other five SREs were Biotechnology), anti-STAT3 pAb (H-190; Santa Cruz), or control rabbit pAb (DakoCytomation) for1hat4°C. The immune complexes were fur- not conserved between human and mouse. This suggested that the ther incubated with protein G beads at 4°C overnight. After washing the SRE on intron 4, designated as I-SRE-4, was particularly impor- immunoprecipitates with washing buffer (Santa Cruz Biotechnology), tant for the regulation of IL-10 gene expression. http://www.jimmunol.org/ DNA was recovered by incubation with 1% SDS and 0.1 M NaHCO3 at 67°C overnight. The DNA samples were purified and used as templates for Intronic SRE acts as an enhancer for gene expression real-time PCR, which was performed using sets of specific primers as follows: IL-10 5Ј-proximal region, 5Ј-GACCCAATTATTTCTCAATCCC To determine the functional roles of I-SRE-4, we constructed a and 5Ј-GAGCTCCTCCTTCTCTAACC; IL-10 intron 4, 5Ј-AGTCT luciferase reporter vector, consisting of a promoter sequence de- GATTTCCCAGAAACC and 5Ј-GTGCATTGACCTTCATCTCC; IL-2 Ј Ϫ Ϫ 5Ј-proximal region, 5Ј-GTTTACTCTTGCTCTTGTC and 5Ј-CCTCTTT rived from the 5 -proximal IL-10 region (from 1.5 to 0.02 kbp GTTACATTAGCCC. Real-time PCR was performed using a LightCycler relative to the TSS) and an enhancer sequence derived from I- 480 Real-Time PCR System (Roche Applied Science), and the cycle SRE-4 (from ϩ3082 to ϩ3243 bp relative to the TSS). This threshold values for each immunoprecipitate were normalized to the input 162-bp fragment with the enhancer sequence includes no other cycle threshold value as relative amounts. responsive elements for a transcription factor except for I-SRE-4. by guest on September 26, 2021 Statistics Mutation at I-SRE-4 was also introduced into the STAT consensus sequence. As target cells for the transient assay, we used conT All data are expressed as the means Ϯ SD. Statistical differences between groups were analyzed using one-way ANOVA, followed by post hoc Bon- cells because the cells are responsive to CD3/CD28 stimulation ferroni/Dunn’s tests. Values of p Ͻ 0.01 were considered to be statistically and are more readily transfected than HOZOT cells. After trans- significant. fection with the vectors, the cells were treated with IL-2 and CD3/

FIGURE 3. Analysis of the methylation patterns at CpG sites in the 5Ј-proximal and intron 4 regions among different T and B cells. Locations of CpG positions (small numbered open circles with vertical lines) in the 5Ј-proximal or intron 4 regions are indicated on the map with a base pair scale originating at the TSS. The demethylated/methylated ratio at each CpG position is depicted by a pie chart. The number of the pie charts corresponds to the number of the CpG positions on the map. Short horizontal bars above the map show positions of P-SRE and I-SRE-4. The Journal of Immunology 3901 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 4. Gel-shift analyses using oligonucleotide probes corresponding to I-SRE-4 or P-SRE. Nuclear extracts from HOZOT-1 were prepared after treatment with or without IL-2 and used for gel-shift analysis by incubating with different types of oligonucleotide probes containing SRE. The probes used for the experiments are listed below the maps of intron 4 (A) and the 5Ј-proximal (B) regions. Probes containing I-SRE-4 are either unmethylated (P1) or methylated (P2, P3, and P4). Probes with mutant (Mu) or repeated (Rep) sequences of I-SRE-4 were also used. C, Gel-shift experiments using I-SRE- 4-containing probes. Arrows indicate shifted bands. D, Competition experiments performed by incubating extracts with a 100-fold molar excess of unlabeled oligonucleotides before addition of labeled P1 probe. E, Supershift experiments performed by incubating extracts with anti-STAT5 or anti-STAT3 pAbs for 2 h before addition of labeled P1 probe. F, Gel-shift analysis performed using an unmethylated probe for P-SRE. The data are representatives of three independent experiments.

CD28 stimulated for 16 h, and then luciferase activity was mea- lane 2) compared with no stimulation. In combination with CD3/ sured. As shown in Fig. 2A, the luciferase reporter vector with CD28 stimulation, exogenous IL-2 induced a 2.8-fold increase in 5Ј-proximal region (lane 2) showed a 1.4-fold induction compared activity (lane 3). Treatment with anti-IL-2/IL-2R Abs resulted in a with a control vector without promoter/enhancer (lane 1). Addition 30% reduction (lane 4), indicating a role of endogenous IL-2 pro- of the wild-type I-SRE-4 at the 3Ј site (lane 3) resulted in a 2.6- duced by CD3/CD28 stimulation. Experiments using a vector con- fold induction, whereas addition of a mutant I-SRE-4 (lane 4) taining a mutant I-SRE-4 showed no significant effect of blocking yielded only a 1.8-fold induction. When the minimal promoter Abs (lanes 7 and 8), confirming the involvement of I-SRE-4 in the TATA was used as a promoter, no enhancer activity of I-SRE-4 endogenous IL-2 response. was observed (lane 5). These results suggested that I-SRE-4 could function as a cis-acting enhancer element on the IL-10 locus. Specific CpG demethylation around I-SRE-4 in human We further examined the effects of endogenous IL-2 on the en- IL-10-producing Treg cells hancer activity of I-SRE-4. Because conT produced endogenous Because it has been reported that differential pattern of cytokine IL-2 following CD3/CD28 stimulation, we used anti-IL-2 pAb and gene expression is also regulated by epigenetic modification as anti-IL-2R mAb to block its effect. Using the vector containing well as transcription factor binding on the locus in various T cell both the 5Ј-proximal region and I-SRE-4, the treatment with ex- subsets, we next focused on the possible epigenetic modifications. ogenous IL-2 alone did not enhance luciferase activity (Fig. 2B, As a first step, we analyzed CpG methylation patterns among a 3902 Treg PRODUCED IL-10 THROUGH AN INTRONIC ENHANCER variety of T cells in relation to their expression of the IL-10 gene. We compared a total of six normal T cell lines together with naive T cells and one leukemic B cell line, Daudi. The six T cell lines included two HOZOTs (HOZOT-1 and -4), one IL-10-Treg, one nTreg, one Th2, and one conT, which produced graded levels of IL-10, as shown in Table I. Upon CD3/CD28 stimulation, HOZOT-1, HOZOT-4, and IL-10-Treg produced IL-10 in a range of 120-1500 ng/ml in the absence of IL-2. In the presence of exogenous IL-2, production increased to a range of 1560–2300 ng/ ml. Th2 cells also produced a higher level (2020 ng/ml) of IL-10 with CD3/CD28 stimulation, but no IL-2 enhancement was observed. ConT and nTreg cells produced only low levels (14–44 ng/ml) of IL-10, even in the presence of IL-2. Next, we performed bisulfate genomic sequencing analysis over the intron 4 region spanning from ϩ2.8 to ϩ3.7 kbp downstream of TSS. CpG methylation over the 5Ј-proximal region was also examined to evaluate a human-specific SRE in this region, designated SRE within proximal region (P-SRE), an effective element for STAT3 binding in LPS/IFN-␣- or IL-10-stimulated macro- Downloaded from phages. We sequenced 20 individual DNA clones derived from each T cell line and Daudi (which failed to produce IL-10). The results of the methylation/demethylation analyses are depicted in Fig. 3 as a pie chart. We determined the methylation status of 16 CpG sites within intron 4 (labeled 1–16 in Fig. 3) and eight CpG

sites within the 5Ј-proximal region (labeled 1–8 in Fig. 3). In Fig. http://www.jimmunol.org/ 3, we indicated the IL-10-producing ability as “High” for Ͼ1000 pg/ml, “Low” for Ͻ100 pg/ml, and “No” for those below detection. In the region around I-SRE-4, five CpG sites (positions 8, 9, 10, 11, and 12) were hypomethylated in HOZOT-1, HOZOT-4, and IL-10-Treg (2, 20, and 10% of an average methylation rate of five sites, respectively). In contrast, these sites were hypermethylated in nTreg (70%), conT (40%), and naive T (52%). Th2 also showed hypermethylation of CpGs at positions 9 and 10 (60%). Daudi by guest on September 26, 2021 exhibited almost complete methylation (95%) at these sites. CpG FIGURE 5. ChIP analysis to measure acetylation levels of histone sites 1–4 within intron 4 were hypomethylated (0–42%) in all T H3/H4 at the sites of P-SRE and I-SRE-4. Locations of primers for ChIP- cells, except at position 4 of nTreg (90%), in contrast to Daudi quantitative real-time PCR (ChIP-qPCR) (arrows) at the 5Ј-proximal (A) (84%). These data suggest that the CpG hypomethylation pattern and the intron 4 (B) regions. C, ChIP experiments using IL-2-treated in the I-SRE-4 region is associated with high levels of IL-10 pro- HOZOT-1 and Th2 cells. Proteins and DNA were cross-linked with form- duction in IL-2-responsive Tregs such as HOZOTs and aldehyde, cells were lysed, and DNA was sheared. ChIP was performed using either rabbit pAb or anti-acetyl-histone H3 pAb. Quantification of IL-10-Treg. Ј immunoprecipitated DNA fragments was performed by real-time PCR us- In the 5 -proximal region, CpG site 1, located at P-SRE, was ing primer for P-SRE or I-SRE-4. Values were normalized to correspond- p Ͻ 0.01 compared with the value of lane 5. The data ,ءء .moderately methylated in HOZOT-1 (15%) and HOZOT-4 (30%), ing input control but highly (90%) methylated in IL-10-Treg. In Th2 cells, CpG site are given as means Ϯ SD for triplicate samples and are representatives of 1 was 45% methylated. The methylation level of this CpG site did three independent experiments. not correlate well with the level of IL-10 production. Therefore, the CpG hypomethylation status of I-SRE-4 is a better indicator of high IL-10 production than that of P-SRE. was used, no shifting of the band was observed (Fig. 4C). Use of In vitro methylation of SRE inhibits STAT binding a probe containing a repeated STAT consensus sequence (Rep) To examine the effects of CpG methylation of I-SRE-4 on STAT resulted in dense, double-shifted bands, which are probably STAT binding, we performed a gel-shift analysis using oligonucleotide tetramer. As shown in Fig. 4D, experiments using unmethylated probes containing either I-SRE-4 (Fig. 4A) or P-SRE (Fig. 4B). (P1), mutant (Mu), or repeat (Rep) probes as competitors revealed The nuclear extracts were prepared from IL-2-treated or untreated the specificity of factor binding to the STAT consensus sequence. HOZOT-1. As shown in Fig. 4C, when an unmethylated probe Supershift experiments with anti-STAT5 pAb or anti-STAT3 pAb (P1) was used for the analysis, the band was shifted when the showed that the factors that bound to P1 included primarily extract came from IL-2-treated cells, but not untreated cells. We STAT5 and, to a lesser extent, STAT3 (Fig. 4E). In summary, also used three types of methylated probes, as follows: one with a these in vitro experiments showed that I-SRE-4 could bind directly methylated CpG site 9 (P2); another methylated at CpG sites 9, 10, to STAT-containing nuclear factors induced by IL-2, and that and 11 (P3); and the third methylated at CpG sites 10 and 11 (P4). binding was dependent upon CpG methylation. We found that the intensity of the shifted bands with P2, P3, and In contrast to I-SRE-4, a P-SRE-containing probe (P5) produced P4 was significantly reduced (30–38% reduction in their intensity only a weak shifting of the band even under demethylated condi- determined by densitometry analysis; data not shown). Further- tions (Fig. 4F). This result indicates a lower in vitro affinity of the more, when the mutant probe (Mu) on a STAT consensus sequence STAT-containing factors for P-SRE than I-SRE-4. The Journal of Immunology 3903

IL-2 locus. These results suggest that STAT5 molecules, upon activation by IL-2, bind primarily to I-SRE-4 in situ in IL-2-responsive Treg cells.

Discussion IL-10 is an immunosuppressive cytokine regulating growth and/or differentiation of a variety of immune cells, including Th1 cells, CTLs, NK cells, B cells, and DC (10). Therefore, deficiency or dysregulated expression of IL-10 production is often a main cause of autoimmune diseases or exaggerated inflammatory responses. IL-10-deficient mice as well as patients with Crohn’s disease dis- play severe chronic enterocolitis, in which the up-regulation of inflammatory cytokines such as IFN-␥, IL-12, TNF-␣, IL-1, and IL-18 is observed, demonstrating the protective role of IL-10 in inflammatory conditions (35). Systemic lupus erythematosus is associated with the high production of IL-10, which leads to impaired T cell responses and dysregulation of B cell activation (36). Administration of anti-IL-10 Ab resulted in a long-lasting reduc-

tion of symptoms in the systemic lupus erythematosus patient. Al- Downloaded from though the control of IL-10 production is key to immune ho- meostasis in vivo, molecular mechanisms regulating IL-10 production are poorly understood. IL-10 production by human Treg cells was enhanced by exog- FIGURE 6. In situ detection of IL-2-activated STAT5 bound to enous IL-2 through the activation of STAT5 molecules, as we

I-SRE-4. A, Western blot detection of phospho-STAT5 in HOZOT-1 after previously reported (11). This response to IL-2 was found only in http://www.jimmunol.org/ IL-2 (10 ng/ml) treatment. B, ChIP analysis of STAT5 binding to SREs. IL-10-producing Treg cells, such as HOZOT and IL-10-Treg, but Cross-linked chromatin complexes from IL-2-treated or untreated not in nTreg or Th2. In the case of nTreg, the IL-2-STAT5 sig- HOZOT-1 were immunoprecipitated with anti-STAT5 pAb or a control naling axis leads to FOXP3 up-regulation. Therefore, we asked pAb. ChIP-quantitative real-time PCR (ChIP-qPCR) was performed using whether HOZOT or IL-10-Treg cells also use the IL-2-STAT5 three sets of primers for P-SRE (as in Fig. 5A), I-SRE-4 (as in Fig. 5B), signaling axis for regulation of IL-10. In this study, we first ad- and, as a negative control, ARRE from the IL-2 promoter. Relative amounts of PCR product were normalized to the amount of input chroma- dressed the question as to whether STAT5 could bind to the IL-10 p Ͻ 0.01 compared with the value in the absence of IL-2. The data locus. Bioinformatic analysis using rVISTA 2.0 (29) showed that ,ءء .tin are representative of three independent experiments. several SREs were spread across the entire IL-10 locus in both the 5Ј-proximal (promoter) and intronic regions. One SRE within the by guest on September 26, 2021 5Ј-proximal region was not conserved among five species. How- ever, it was reported to be a STAT3 binding site following expo- Histone acetylation levels at SREs sure of human B cells and macrophages to IFN-␣ (37), LPS (38), As another indicator of epigenetic control, we assessed the histone or IL-10 (39). Other SREs within intron 2, reportedly STAT1 bind- acetylation status in the regions of P-SRE and I-SRE-4, which ing sites following IFN-␥ treatment (40), were also not conserved. reflects chromatin accessibility of the genes. We performed ChIP One SRE found within intron 4, designated as I-SRE-4, was con- with anti-acetyl histone H3 pAb using HOZOT-1 and Th2 cells. served across species. We speculated that I-SRE-4 might be crit- DNA fragments in the immunoprecipitates were amplified by real- ical for IL-10 gene expression because interspecies-conserved re- time PCR with PCR primer sets corresponding to the P-SRE and gions often correspond to DNase I HS sites (41), which could I-SRE-4 region (Fig. 5, A and B). As shown in Fig. 5C, acetylation become targets for binding of transcription factors. Therefore, we of histone H3 in the I-SRE-4 regions was observed in both focused our analysis on I-SRE-4. HOZOT-1 and Th2 cells, but its acetylation level was markedly We next demonstrated that I-SRE-4 indeed acted as an enhancer higher in HOZOT-1 than in Th2 cells. In the P-SRE region, acet- element within the IL-10 locus. This enhancer worked efficiently in ylation levels were comparable between the two cells. These re- combination with a promoter derived from the 5Ј-proximal region, sults confirmed that chromatin in the I-SRE-4 region was kept but not with the minimal promoter TATA. Because the 5Ј-proxi- open in HOZOT-1, but was closed in Th2 cells. mal region includes responsive elements for NF-AT (41) and NF-␬B (42), I-SRE-4 may cooperate with these elements to exert In situ binding of activated STAT5 its enhancing effects. To further examine whether STAT5 was actually bound to SRE in One intriguing finding of our studies is that the STAT5 enhancer HOZOT-1 cells in situ, we next performed a ChIP analysis with element was localized in the intronic region on the IL-10 locus. It anti-STAT5 pAb. Because STAT5 was phosphorylated by IL-2 has been reported that STAT5 binds to intronic elements on a exposure during a 240-min treatment of HOZOT-1 (Fig. 6A), we number of other gene loci. Nelson et al. (43) performed compre- used chromatin from the cells with or without IL-2 treatment for hensive ChIP analysis with anti-STAT5 Ab using an IL-3-depen- 120 min for ChIP analysis. As a negative control, we used a set of dent cell line, BaF/3, and demonstrated that a large proportion of PCR primers for Ag receptor-responsive elements (ARRE) of IL-2 functional STAT5 binding sites was located within introns. A gene (17), in which the STAT consensus sequence is not included. number of functional intronic SREs have been found for immu- The ChIP analysis revealed that IL-2-activated STAT5 bound to nologically important genes. Among them, the FOXP3 gene con- both P-SRE and I-SRE-4 regions on the IL-10 locus (Fig. 6B). The tains a conserved SRE within intron 1. This intronic SRE is a level of binding to I-SRE-4 was higher than that to P-SRE. Almost downstream target of the IL-2-STAT5 signaling axis in nTreg no binding was observed to ARRE, a negative control region in the cells, leading to up-regulation of FOXP3 expression (4, 9, 44). Its 3904 Treg PRODUCED IL-10 THROUGH AN INTRONIC ENHANCER functional property was confirmed by luciferase assay, in which 293 cells were cotransfected with a SRE-containing reporter construct and a constitutive activated STAT5 expression construct. As another example, intron 2 of the IL-4 gene contains a SRE that bound STAT5 and enhanced IL-4 gene transcription in mast cells (45). Although accumulating evidence revealed obvious roles of intronic enhancers for STAT5, it remains unclear how these enhancers regulate cytokine gene transcription. Further studies are needed to understand the molecular interaction with the 5Ј-proximal promoter. We next asked whether epigenetic regulation was also involved in determining the level of IL-10 expression. Different types of T cells, including HOZOTs, conT, Th2, and nTreg, produce different amounts of IL-10 when the IL-2-STAT5 signaling axis is activated to the same extent. Interestingly, around I-SRE-4, there is a cluster of CpGs, suggesting the possibility of epigenetic regulation of IL-10 gene expression through CpG methylation. Many cytokine loci are targets for such control, including IL-2 (17, 18), IL-4,

IL-13 (22), and IFN-␥ (46) loci in various types of T cells. For Downloaded from example, in naive T cells, TCR stimulation causes the promoter of the IL-2 locus to undergo immediate CpG demethylation. As a consequence, the promoter region of activated T cells becomes permissive for transcriptional activation of NF-AT and AP-1 (18). CpG hypomethylation at ARRE of the IL-2 gene correlates with FIGURE 7. Hypothetical models showing the relationship between positive regulation of IL-2 expression in both stimulated and rest- transcription factor binding and CpG methylation on the IL-10 locus. A, http://www.jimmunol.org/ ing CD4ϩ T cells (17). Examples of epigenetic regulation (de- Models for high IL-10 production by HOZOTs, IL-10-Treg, and Th2 cells. methylation of specific CpG tracks) in Th1 and Th2 cells include B, A model for low IL-10 production by nTreg, conT, and naive T cells. C, the NF-AT-responsive element in the IFN-␥ loci (46) and Th2- A model for no IL-10 production by Daudi. In each, the IL-10 locus is Ј specific HS sites on IL-4 and IL-13 loci (22). shown as a bold line consisting of a 5 -proximal region, exons 1–4 (E1– E4), intron 4, and exon 5. CpG sites located in noncoding regions in po- As for IL-10 expression, Dong et al. (47) analyzed the CpG tential transcription binding sites are depicted as circles with vertical lines. methylation pattern of the IL-10 locus in different types of human ϩ Shading of a circle indicates methylation levels. STAT can bind to hy- CD4 memory T cells, which were categorized by IL-10 and pomethylated, but not hypermethylated SREs. In addition to STAT, TCR- IFN-␥ expression. They concluded that there was no good corre- stimulated transcription factors, NF-AT and NF-␬B, are shown. lation between methylation pattern and IL-10 expression. Curi- by guest on September 26, 2021 ously, they did not specify I-SRE-4 as a relevant CpG site, although they included this element in their study. In contrast, our with p300/CREB-binding protein and enhance the accessibility of ␥ study revealed an epigenetic significance of I-SRE-4 for IL-10 regulatory molecules to the TCR locus (50). Th1 and Th2 cells production. One characteristic of this element is species specific- provide other examples. During Th1 cell differentiation, IL-12- ity, namely clustered CpG dinucleotides around I-SRE-4 were activated STAT4 directly binds to the Il18r1 locus, transiently in- found in human cells, but not in mouse. There is another CpG site creases histone acetylation, and subsequently decreases DNA within the promoter region, designated P-SRE, but this CpG was methyltransferase association on the locus, resulting in higher ex- methylated even in IL-10-Treg cells, typically high producers of pression of IL-18R in Th1 cells (51). In contrast, during Th2 cell IL-10. Therefore, I-SRE-4 contributes more to IL-10 gene expres- differentiation, intronic STAT5 binding induces chromatin remod- sion than P-SRE. eling on the IL-4 locus (45). Therefore, in our IL-10 system, there In general, DNA methylation regulates gene expression through is a possibility that during induction of HOZOT in primary culture, two distinct mechanisms. First, it inhibits the binding of transcrip- STAT5 mediates chromatin alteration by binding to I-SRE-4. Fur- tion factors by imposing direct physical constraints (48). In our ther analysis is required to test this hypothesis. system, methylation-specific gel-shift analyses demonstrated that a In conclusion, our results, which are summarized in a model methylated probe containing I-SRE-4 physically inhibited STAT shown in Fig. 7, provide new insights into the importance of both binding. Second, methylated DNA recruits regulatory proteins epigenetic and transcriptional regulation for high IL-10 production (such as MBD2), which contain a methyl-CpG binding domain. in human Treg cells. Moreover, we have provided the first molec- These proteins generate an inaccessible chromatin structure by as- ular evidence for the functional relevance of the IL-2-STAT5- sembling multisubunit complexes containing corepressors and hi- IL-10 signaling axis in induced Treg cells, in contrast to the IL- stone deacetylases (49). We examined chromatin accessibility by 2-STAT5-FOXP3 signaling axis. ChIP analysis with anti-acetyl-histone H3/H4 Abs. The ChIP analysis revealed that IL-2-activated STAT5 bound to I-SRE-4, but did Acknowledgments not increase histone acetylation above that observed with the un- We express our sincere appreciation to Dr. Koichi Ikuta (Laboratory of stimulated control. Biological Protection, Department of Biological Responses, Institute for Virus Research, Kyoto University), Dr. Noriko M. Tsuji (Age Dimension One intriguing question is how the epigenetic modification ob- Research Center, National Institute for Advanced Industrial Science and served in our study occurred during IL-10-Treg cell differentiation, Technology), Dr. Kunnzo Orita (Hayashibara), and Dr. Tsunetaka Ohta in particular, the CpG demethylation around I-SRE-4. There are (Biomedical Institute, Research Center, Hayashibara Biochemical Labora- several reports that the binding of STATs acts as an inducer of tories) for helpful discussions. We also thank Dr. Kenji Akita (Biomedical chromatin remodeling in immune systems. For example, during Institute, Research Center, Hayashibara Biochemical Laboratories) for TCR␥␦ T cell differentiation, IL-7-activated STAT5 can interact technical support using radioisotopic materials. The Journal of Immunology 3905

26. Godfrey, W. R., D. J. Spoden, Y. G. Ge, S. R. Baker, B. Liu, B. L. Levine, Disclosures ϩ ϩ C. H. June, B. R. Blazar, and S. B. Porter. 2005. Cord blood CD4 CD25 - The authors have no financial conflict of interest. derived T regulatory cell lines express FoxP3 protein and manifest potent suppressor function. Blood 105: 750–758. References 27. Barrat, F. J., D. J. Cua, A. Boonstra, D. F. Richards, C. Crain, H. F. Savelkoul, R. de Waal-Malefyt, R. L. Coffman, C. M. Hawrylowicz, and A. O’Garra. 2002. 1. Beadling, C., J. Ng, J. W. Babbage, and D. A. Cantrell. 1996. Interleukin-2 In vitro generation of interleukin 10-producing regulatory CD4ϩ T cells is in- activation of STAT5 requires the convergent action of tyrosine kinases and a duced by immunosuppressive drugs and inhibited by T helper type 1 (Th1)- and serine/threonine kinase pathway distinct from the Raf1/ERK2 MAP kinase path- Th2-inducing cytokines. J. Exp. Med. 195: 603–616. way. EMBO J. 15: 1902–1913. 28. Sundrud, M. S., S. M. Grill, D. Ni, K. Nagata, S. S. Alkan, A. Subramaniam, and 2. Moriggl, R., D. J. Topham, S. Teglund, V. Sexl, C. McKay, D. Wang, D. Unutmaz. 2003. Genetic reprogramming of primary human T cells reveals A. Hoffmeyer, J. van Deursen, M. Y. Sangster, K. D. Bunting, et al. 1999. Stat5 functional plasticity in Th cell differentiation. J. Immunol. 171: 3542–3549. is required for IL-2-induced cell cycle progression of peripheral T cells. Immunity 10: 249–259. 29. Loots, G. G., and I. Ovcharenko. 2004. rVISTA 2.0: evolutionary analysis of 3. Fontenot, J. D., J. P. Rasmussen, M. A. Gavin, and A. Y. Rudensky. 2005. A transcription factor binding sites. Nucleic Acids Res. 32: W217–W221. function for interleukin 2 in Foxp3-expressing regulatory T cells. Nat. Immunol. 30. Tsuji-Takayama, K., T. Inoue, Y. Ijiri, T. Otani, R. Motoda, S. Nakamura, and 6: 1142–1151. K. Orita. 2004. Demethylating agent, 5-azacytidine, reverses differentiation of 4. Burchill, M. A., J. Yang, C. Vogtenhuber, B. R. Blazar, and M. A. Farrar. 2007. embryonic stem cells. Biochem. Biophys. Res. Commun. 323: 86–90. IL-2 receptor ␤-dependent STAT5 activation is required for the development of 31. Tsuji-Takayama, K., Y. Aizawa, I. Okamoto, H. Kojima, K. Koide, M. Takeuchi, Foxp3ϩ regulatory T cells. J. Immunol. 178: 280–290. H. Ikegami, T. Ohta, and M. Kurimoto. 1999. Interleukin-18 induces interferon-␥ 5. Nelson, B. H. 2004. IL-2, regulatory T cells, and tolerance. J. Immunol. 172: production through NF-␬B and NFAT activation in murine T helper type 1 cells. 3983–3988. Cell. Immunol. 196: 41–50. 6. Antony, P. A., C. M. Paulos, M. Ahmadzadeh, A. Akpinarli, D. C. Palmer, 32. Tsuji-Takayama, K., T. Otani, T. Inoue, S. Nakamura, R. Motoda, M. Kibata, and N. Sato, A. Kaiser, C. S. Hinrichs, C. A. Klebanoff, Y. Tagaya, and N. P. Restifo. K. Orita. 2006. Erythropoietin induces sustained phosphorylation of STAT5 in 2006. Interleukin-2-dependent mechanisms of tolerance and immunity in vivo. primitive but not definitive erythrocytes generated from mouse embryonic stem J. Immunol. 176: 5255–5266. cells. Exp. Hematol. 34: 1323–1332. Downloaded from 7. Snow, J. W., N. Abraham, M. C. Ma, B. G. Herndier, A. W. Pastuszak, and 33. Tsuji-Takayama, K., S. Matsumoto, K. Koide, M. Takeuchi, M. Ikeda, T. Ohta, and M. A. Goldsmith. 2003. Loss of tolerance and autoimmunity affecting multiple M. Kurimoto. 1997. Interleukin-18 induces activation and association of p56lck and organs in STAT5A/5B-deficient mice. J. Immunol. 171: 5042–5050. MAPK in a murine TH1 clone. Biochem. Biophys. Res. Commun. 237: 126–130. 8. Cohen, A. C., K. C. Nadeau, W. Tu, V. Hwa, K. Dionis, L. Bezrodnik, A. Teper, 34. Loots, G. G., I. Ovcharenko, L. Pachter, I. Dubchak, and E. M. Rubin. 2002. M. Gaillard, J. Heinrich, A. M. Krensky, et al. 2006. Cutting edge: decreased rVista for comparative sequence-based discovery of functional transcription fac- accumulation and regulatory function of CD4ϩCD25high T cells in human tor binding sites. Genome Res. 12: 832–839. STAT5b deficiency. J. Immunol. 177: 2770–2774. 9. Zorn, E., E. A. Nelson, M. Mohseni, F. Porcheray, H. Kim, D. Litsa, R. Bellucci, 35. Akagi, S., E. Hiyama, Y. Imamura, Y. Takesue, Y. Matsuura, and T. Yokoyama. http://www.jimmunol.org/ E. Raderschall, C. Canning, R. J. Soiffer, et al. 2006. IL-2 regulates FOXP3 expres- 2000. Interleukin-10 expression in intestine of Crohn disease. Int. J. Mol. Med. 5: sion in human CD4ϩCD25ϩ regulatory T cells through a STAT-dependent mecha- 389–395. nism and induces the expansion of these cells in vivo. Blood 108: 1571–1579. 36. Llorente, L., and Y. Richaud-Patin. 2003. The role of interleukin-10 in systemic 10. Moore, K. W., R. de Waal Malefyt, R. L. Coffman, and A. O’Garra. 2001. In- lupus erythematosus. J. Autoimmun. 20: 287–289. terleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. 19: 683–765. 37. Ziegler-Heitbrock, L., M. Lotzerich, A. Schaefer, T. Werner, M. Frankenberger, 11. Tsuji-Takayama, K., M. Suzuki, M. Yamamoto, A. Harashima, A. Okochi, and E. Benkhart. 2003. IFN-␣ induces the human IL-10 gene by recruiting both T. Otani, T. Inoue, A. Sugimoto, R. Motoda, F. Yamasaki, et al. 2008. IL-2 IFN regulatory factor 1 and Stat3. J. Immunol. 171: 285–290. activation of STAT5 enhances production of IL-10 from human cytotoxic regu- 38. Benkhart, E. M., M. Siedlar, A. Wedel, T. Werner, and H. W. Ziegler-Heitbrock. latory T cells, HOZOT. Exp. Hematol. 36: 181–192. 2000. Role of Stat3 in lipopolysaccharide-induced IL-10 gene expression. J. Im- 12. Wang, Z. Y., H. Sato, S. Kusam, S. Sehra, L. M. Toney, and A. L. Dent. 2005. munol. 165: 1612–1617. Regulation of IL-10 gene expression in Th2 cells by Jun proteins. J. Immunol. 39. Staples, K. J., T. Smallie, L. M. Williams, A. Foey, B. Burke, B. M. Foxwell, and

174: 2098–2105. L. Ziegler-Heitbrock. 2007. IL-10 induces IL-10 in primary human monocyte- by guest on September 26, 2021 13. Viola, J. P., and A. Rao. 1997. Role of the cyclosporin-sensitive transcription derived macrophages via the transcription factor Stat3. J. Immunol. 178: factor NFAT1 in the allergic response. Mem. Inst. Oswaldo Cruz 92(Suppl. 2): 4779–4785. 147–155. 40. VanDeusen, J. B., M. H. Shah, B. Becknell, B. W. Blaser, A. K. Ferketich, 14. Shoemaker, J., M. Saraiva, and A. O’Garra. 2006. GATA-3 directly remodels the ϩ G. J. Nuovo, B. M. Ahmer, J. Durbin, and M. A. Caligiuri. 2006. STAT-1- IL-10 locus independently of IL-4 in CD4 T cells. J. Immunol. 176: 3470–3479. mediated repression of monocyte interleukin-10 gene expression in vivo. Eur. 15. Fitzpatrick, D. R., and C. B. Wilson. 2003. Methylation and demethylation in the J. Immunol. 36: 623–630. regulation of genes, cells, and responses in the immune system. Clin. Immunol. 41. Kang, K. H., and S. H. Im. 2005. Differential regulation of the IL-10 gene in Th1 109: 37–45. and Th2 T cells. Ann. NY Acad. Sci. 1050: 97–107. 16. Loots, G. G., R. M. Locksley, C. M. Blankespoor, Z. E. Wang, W. Miller, E. M. Rubin, and K. A. Frazer. 2000. Identification of a coordinate regulator of 42. Saraiva, M., J. R. Christensen, A. V. Tsytsykova, A. E. Goldfeld, S. C. Ley, interleukins 4, 13, and 5 by cross-species sequence comparisons. Science 288: D. Kioussis, and A. O’Garra. 2005. Identification of a macrophage-specific chro- 136–140. matin signature in the IL-10 locus. J. Immunol. 175: 1041–1046. 17. Murayama, A., K. Sakura, M. Nakama, K. Yasuzawa-Tanaka, E. Fujita, 43. Nelson, E. A., S. R. Walker, J. V. Alvarez, and D. A. Frank. 2004. Isolation of Y. Tateishi, Y. Wang, T. Ushijima, T. Baba, K. Shibuya, et al. 2006. A specific unique STAT5 targets by chromatin immunoprecipitation-based gene identifica- CpG site demethylation in the human interleukin 2 gene promoter is an epigenetic tion. J. Biol. Chem. 279: 54724–54730. memory. EMBO J. 25: 1081–1092. 44. Yao, Z., Y. Kanno, M. Kerenyi, G. Stephens, L. Durant, W. T. Watford, 18. Bruniquel, D., and R. H. Schwartz. 2003. Selective, stable demethylation of the A. Laurence, G. W. Robinson, E. M. Shevach, R. Moriggl, et al. 2007. Nonre- interleukin-2 gene enhances transcription by an active process. Nat. Immunol. 4: dundant roles for Stat5a/b in directly regulating Foxp3. Blood 109: 4368–4375. 235–240. 45. Hural, J. A., M. Kwan, G. Henkel, M. B. Hock, and M. A. Brown. 2000. An 19. Chen, L., W. He, S. T. Kim, J. Tao, Y. Gao, H. Chi, A. M. Intlekofer, B. Harvey, intron transcriptional enhancer element regulates IL-4 gene locus accessibility in S. L. Reiner, Z. Yin, et al. 2007. Epigenetic and transcriptional programs lead to mast cells. J. Immunol. 165: 3239–3249. default IFN-␥ production by ␥␦ T cells. J. Immunol. 178: 2730–2736. 46. White, G. P., P. M. Watt, B. J. Holt, and P. G. Holt. 2002. Differential patterns 20. Northrop, J. K., R. M. Thomas, A. D. Wells, and H. Shen. 2006. Epigenetic of methylation of the IFN-␥ promoter at CpG and non-CpG sites underlie dif- remodeling of the IL-2 and IFN-␥ loci in memory CD8 T cells is influenced by ferences in IFN-␥ gene expression between human neonatal and adult CD45ROϪ CD4 T cells. J. Immunol. 177: 1062–1069. T cells. J. Immunol. 168: 2820–2827. 21. Lee, D. U., S. Agarwal, and A. Rao. 2002. Th2 lineage commitment and efficient 47. Dong, J., C. Ivascu, H. D. Chang, P. Wu, R. Angeli, L. Maggi, F. Eckhardt, IL-4 production involves extended demethylation of the IL-4 gene. Immunity 16: L. Tykocinski, C. Haefliger, B. Mowes, et al. 2007. IL-10 is excluded from the 649–660. functional cytokine memory of human CD4ϩ memory T lymphocytes. J. Immu- 22. Santangelo, S., D. J. Cousins, N. E. Winkelmann, and D. Z. Staynov. 2002. DNA nol. 179: 2389–2396. methylation changes at human Th2 cytokine genes coincide with DNase I hypersen- ϩ 48. Tang, D., G. S. Sivko, and J. W. DeWille. 2006. Promoter methylation reduces sitive site formation during CD4 T cell differentiation. J. Immunol. 169: 1893–1903. C/EBP␦ (CEBPD) gene expression in the SUM-52PE human breast cancer cell 23. Kaneko, T., H. Hosokawa, M. Yamashita, C. R. Wang, A. Hasegawa, line and in primary breast tumors. Breast Cancer Res. Treat. 95: 161–170. M. Y. Kimura, M. Kitajiama, F. Kimura, M. Miyazaki, and T. Nakayama. 2007. Chromatin remodeling at the Th2 cytokine gene loci in human type 2 helper T 49. Hutchins, A. S., A. C. Mullen, H. W. Lee, K. J. Sykes, F. A. High, cells. Mol. Immunol. 44: 2249–2256. B. D. Hendrich, A. P. Bird, and S. L. Reiner. 2002. Gene silencing quantitatively 24. Nakamura, S., M. Suzuki, A. Sugimoto, K. Tsuji-Takayama, M. Yamamoto, controls the function of a developmental trans-activator. Mol. Cell 10: 81–91. T. Otani, T. Inoue, A. Harashima, A. Okochi, R. Motoda, et al. 2007. IL-2- 50. Ye, S. K., Y. Agata, H. C. Lee, H. Kurooka, T. Kitamura, A. Shimizu, T. Honjo, independent generation of FOXP3ϩCD4ϩCD8ϩCD25ϩ cytotoxic regulatory T and K. Ikuta. 2001. The IL-7 receptor controls the accessibility of the TCR␥ locus cell lines from human umbilical cord blood. Exp. Hematol. 35: 287–296. by Stat5 and histone acetylation. Immunity 15: 813–823. 25. Yamazaki, S., K. Inaba, K. V. Tarbell, and R. M. Steinman. 2006. Dendritic cells 51. Yu, Q., V. T. Thieu, and M. H. Kaplan. 2007. Stat4 limits DNA methyltransferase expand antigen-specific Foxp3ϩ CD25ϩ CD4ϩ regulatory T cells including sup- recruitment and DNA methylation of the IL-18R␣ gene during Th1 differentia- pressors of alloreactivity. Immunol. Rev. 212: 314–329. tion. EMBO J. 26: 2052–2060.