The SWI/SNF Chromatin-remodeling Complex Modulates Peripheral T Cell Activation and Proliferation by Controlling AP-1 Expression*

The SWI/SNF chromatin-remodeling complex has been implicated in the activation and proliferation of T cells. After T cell receptor signaling, the SWI/SNF complex rapidly associates with chromatin and controls gene expression in T cells. However, the process by which the SWI/SNF complex regulates peripheral T cell activation has not been elucidated. In this study, we show that the SWI/SNF complex regulates cytokine production and proliferation of T cells. During T cell activation, the SWI/SNF complex is recruited to the promoter of the transcription factor AP-1, and it increases the expression of AP-1. Increased expression of the SWI/SNF complex resulted in enhanced AP-1 activity, cytokine production, and proliferation of peripheral T cells, whereas knockdown of the SWI/SNF complex expression impaired the AP-1 expression and reduced the activation and proliferation of T cells. Moreover, mice that constitutively expressed the SWI/SNF complex in T cells were much more susceptible to experimentally induced autoimmune encephalomyelitis than the normal mice were. These results suggest that the SWI/SNF complex plays a critical role during T cell activation and subsequent immune responses.

The chromatin structure of lymphocytes undergoes a dramatic change when the lymphocytes receive an activating antigenic signal. Naïve resting cells contain small and compact nuclei with dense heterochromatin. In contrast, activated lymphocytes contain enlarged nuclei that are largely composed of euchromatin. After activation, T lymphocytes express immediate early genes, including the genes for transcription factors such as c-Fos, c-Jun, NFAT, c-Myc, and NF-B (1)(2)(3). The expression of these transcription factors induces the expression of additional genes that ultimately lead to the proliferation, differentiation, and effector function of T lymphocytes. Although the underlying mechanisms of these lymphocyte activation-induced rapid and dramatic changes in chromatin structure have not been elucidated, it is possible that several chromatin-remodeling complexes are involved in this process. Notably, the SWI/ SNF chromatin-remodeling complex has been shown to rapidly associate with chromatin after T cell receptor signaling (4).
The SWI/SNF chromatin-remodeling complex is an evolutionarily conserved multisubunit complex that uses the energy derived from ATP hydrolysis to modify nucleosomes and remodel chromatin structures. The SWI/SNF complex consists of 9 -11 subunits; the actual number of subunits depends on the tissue and cell type. BRG1, BRM, BAF47/SNF5, BAF155/SRG3, and BAF170 are known to be the core components of the mammalian SWI/SNF complex (5). Previous studies have demonstrated that the SWI/SNF complex plays an important role in the proliferation of peripheral T lymphocytes. In vitro antigen stimulation induces the binding of SWI/SNF complexes to the chromatin in lymphocytes (4). Proliferative defects have been observed in BRG1-deficient mature peripheral T cells (6,7). Similarly, the concanavalin A (ConA) 4 -activated T lymphocytes from lymphoid-specific helicase (Lsh)-deficient mice showed a rapid decrease in proliferation (8). Lsh is a member of the SNF2 subfamily of helicases, which also includes BRG1. Moreover, effector CD4 ϩ T cells that show changes in the chromatin structure at the cytokine gene loci show higher rates of cytokine gene transcription in comparison with the transcription in naïve cells (9,10). When naïve CD4 ϩ T cells differentiate into either T h 1 or T h 2 effector cells, the chromatin structures of the interferon (IFN)-␥ locus and the interleukin (IL)-4 locus, which are expressed in T h 1 or T h 2 cells, respectively, undergo epigenetic modifications (11).
The transcription factor activator protein 1 (AP-1) is one of the earliest transcription factors induced in T cell activation. AP-1 is a dimer of the Fos and/or Jun proteins, and it regulates the expression of a broad variety of genes (12). The target genes of AP-1 play important roles in proliferation, differentiation, and apoptosis (13). The expression of AP-1 is tightly linked to mitogenic stimulation, and this finding suggests that AP-1 performs essential functions in cell proliferation. Indeed, fibroblasts from c-Jun Ϫ/Ϫ embryos show completely defective proliferation (14). AP-1 regulates cell proliferation by inducing the transcription of the genes involved in cell cycle progression, such as cyclin D1, or by repressing the expression of negative regulators of the cell cycle, such as p53 and p16 INK4A (13). AP-1 is essential for the expression of the IL-2 gene in response to TCR/CD28 stimulation; therefore, AP-1 is intimately involved in T cell proliferation (15,16).
AP-1 has been also shown to be involved in the inflammatory response and the expression of many cytokines, including IL-1, IL-2, IL-4, and IFN-␥, is activated by AP-1 (13,17,18). Recently, it was demonstrated that the expression of IL-17, a cytokine implicated in the development of autoimmune diseases, is also regulated by AP-1 (18). AP-1 and other transcription factors such as NF-B and NFAT have been implicated in several T cell-related chronic inflammatory diseases. Indeed, patients suffering from rheumatoid arthritis (RA) show increased AP-1 activity, and the mononuclear cells of lupus patients showed abnormal expressions of AP-1 and NF-B (19,20). Moreover, AP-1 and NF-B were up-regulated in the immune cells of multiple sclerosis (MS) patients (21,22).
Whereas previous studies have implicated the SWI/SNF complex in the activation and proliferation of T cells, the direct target or the precise mechanism of the SWI/SNF complex-mediated regulation of peripheral T cell activation are not known. In this study, we demonstrate that the SWI/SNF complex controls the expression of the transcription factor AP-1 during peripheral T cell activation. Thus, the SWI/SNF complex may regulate cytokine production and proliferation of activated T cells. Our results also suggest that the complex is associated with the development of autoimmune diseases.

EXPERIMENTAL PROCEDURES
Mice and Cells-C57BL/6 SRG3 transgenic (CD2-SRG3) mice were obtained by using previously described methods (23). C57BL/6 mice were purchased from Charles River Laboratories. The mice were bred and maintained under pathogenfree conditions, and the experiments were performed in accordance with institutional and national guidelines.
Cell Culture, Transfection, and Reporter Assay-The EL4 cell line was cultured in Iscove's modified Eagle's medium supplemented with 10% fetal bovine serum. The L7 cell line was cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Transfections were performed using the Gene Pulse-II RF electroporator (Bio-Rad). Reporter assays were performed using the luciferase assay system (Promega), and ␤-galactosidase activity was used to normalize the luciferase results.
FACS Analysis and Cell Purification-Lymphocytes were isolated from the spleen and the lymph nodes of 4 -8-week-old mice. For flow cytometry, the cells were stained with antibodies in phosphate-buffered saline (PBS) containing 2% bovine serum albumin. Routine fluorescence-activated cell sorter (FACS) analysis was performed on a FACSCantoII (Becton Dickinson) system. Intracellular staining and BrdU incorporation were performed according to the manufacturer's protocols (BD Pharmingen). Antibodies for flow cytometry were purchased from BD Pharmingen. To obtain peripheral CD4 ϩ T cells (Ͼ90% pure) for proliferation assays and biochemical analyses, we performed negative selection on an AutoMACS (Miltenyi Biotech). The GFP ϩ -infected cells were sorted using FACSAria (Becton Dickinson).
Retroviral Transduction-SRG3 short hairpin RNA (shRNA) (sequence, 5Ј-CATCCTGGTTTGATTATAA-3Ј) was cloned into an MDH1 retroviral vector. We prepared similar MDH1-shRNA constructs for BRG1 (sequence 5Ј-CCGTCAAGGT-GAAGATCAA-3Ј) and Jun (sequence 5Ј-CAGCTTCCTGC-CTTTGTAA-3Ј). Phoenix-eco cells were transfected with an empty MDH1 vector and either MDH1-shSRG, MDH1-shBrg1, or MDH1-shJun cDNA, and the culture supernatant containing high titers of retrovirus was collected 48 h after the transfection. Negatively purified CD4 ϩ T cells were activated with plate-bound anti-CD3 (2 g/ml) and anti-CD28 (1 g/ml) for 24 h. The activated T cells were cultured in 1 ml of a retroviral supernatant containing 5 mg/ml polybrene (Sigma), centrifuged at 2600 rpm for 90 min at room temperature, and incubated for 48 h. The infected T cells were allowed to proliferate for 2 days in the presence of recombinant IL-2 (20 units/ml); then, they were restimulated for intracellular cytokine analysis or BrdU incorporation. In some experiments, the transduced cells were purified by cell sorting with a FACSAriaII (BD Biosciences).
Quantitative Real-time PCR-Total RNA from the cultured cells was purified by TRIzol (Invitrogen) according to the manufacturer's instructions. Reverse transcription was performed using 1 g of RNA and the SuperScriptIII RT System (Invitrogen). Polymerase chain reaction (PCR) was performed using ABI Prism 7700 (Applied Biosystems) with the 2ϫ SYBR Green MasterMix (Applied Biosystems). The cDNA (4 l) was amplified in a 25 l of PCR mixture containing 0.25 M of the respective primers by using the following parameters: denaturation for 5 min at 94°C and 40 cycles of denaturation at 94°C for 15 s annealing for 1 min, and extension at 60°C. The sequences of the primer pairs are shown in supplemental Table S1.
Proliferation and Suppression Assay-Purified T cells (10 5 cells/well in 96-well round bottom plates) were incubated either with medium alone, with plate-bound anti-CD3 (0.5-5 g/ml) and soluble anti-CD28 (1 g/ml), or with phorbol myristate acetate (PMA; 10 ng/ml) and ionomycin (125 ng/ml). The cells were cultured for 48 h and labeled with 1 Ci/well of [ 3 H]thymidine for 16 h. The incorporation of [ 3 H]thymidine was measured using a MicroBeta JET (PerkinElmer). For the suppression assays, 10 5 responder CD4 ϩ CD25 Ϫ T cells from each mouse were plated in 96-well plates. The indicated ratios of the sorted CD4 ϩ CD25 ϩ Treg cells were added to each plate. The APCs consisted of splenocytes depleted of T cells and treated with mitomycin C. The cells were activated by ConA treatment for 48 h, labeled with 1 Ci/well of [ 3 H]thymidine for 16 h, and proliferation was detected using the technique described above.
Induction and Assessment of EAE-Myelin oligodendrocyte glycoprotein (MOG)-(35-55) (MEVGWYRSPFSRVVHLYRNGK) was synthesized by Peptron (Seoul, Korea). Female C57BL/6 SRG3 transgenic (CD2-SRG3) and WT littermate mice (age 8 -12 weeks) received subcutaneous injections of 200 g of the peptide in 100 l of sterile PBS emulsified with an equal volume of complete Freund's adjuvant (Difco) containing 4 mg/ml of Mycobacterium tuberculosis H37Ra (Difco). The mice also received an intravenous injection of 200 ng of Bordetella pertussis toxin (Sigma) in 100 l of sterile PBS at the time of immunization, and another injection was administered 48 h later. Each day, the mice were scored for signs of clinical disease by using the following scale: 0, no signs of disease; 1, weakness of tail; 2, partial hind limb paralysis; 3, complete hind limb paralysis; 4, forelimb paralysis; 5, death. Mouse cages were coded, and individual mice were scored blindly.
Electrophoretic Mobility Shift Assay-Primary CD4 ϩ T cells were either kept unstimulated or stimulated with PMA and ionomycin. The nuclear extracts were prepared by using a previously described protocol (24). A specific amount (2 g) of each extract was used in each reaction. A 32 P-labeled oligonucleotide containing the AP-1 consensus site (5Ј-CGCTTGAT-GACTCAGCCGGAA-3Ј) was used as a probe. For the detection of specific protein-DNA complexes, antibodies to c-Fos (sc-253X) and c-Jun (sc-1694X) were obtained from Santa Cruz Biotechnology.

Constitutive Expression of SRG3 Increases AP-1 Expression in
Peripheral T Cells-SRG3 (SWI3-related gene), a murine homolog of human BAF155, is a core component of the murine SWI/SNF complex (25). SRG3 functions as a scaffolding protein that controls the stability of the SWI/SNF complex by directly interacting with the other major components of the complex (26). Therefore, any alteration in the expression of SRG3 causes changes in the overall level of the SWI/SNF com-plex and the activity of the complex (26,27). In the transgenic mice that constitutively expressed SRG3 in the T lineage cells under the control of the human CD2 promoter (CD2-SRG3 mice), the expression of SRG3 in the peripheral T cells was 2-fold up-regulated in comparison with the expression in wildtype T cells (23). Furthermore, in these mice, the levels of the other SWI/SNF components, including BRG1, SNF5, and BAF60a, were up-regulated, and the chromatin-remodeling activity had increased (26).
We observed that the levels of c-Fos and c-Jun transcripts in the naïve CD4 ϩ T cells from the CD2-SRG3 transgenic mice were higher than the levels of these transcripts in the CD4 ϩ T cells from the littermate control mice (Fig. 1A). When the T cells were stimulated with PMA and ionomycin, we observed a significant increase in the mRNA-(c-Fos, 2.5-fold increase; c-Jun, 2.3-fold increase) and protein level (c-Fos, 2.1-fold increase; c-Jun, 2.4-fold increase) expressions of these transcription factors in the CD4 ϩ T cells from CD2-SRG3 mice, in comparison with the corresponding expressions in the cells from control mice (Fig. 1, A and B and supplemental Fig. S1). The expression of NF-B (p50) also slightly increased in the CD2-SRG3 CD4 ϩ T cells (1.7-fold increase). In contrast, the expressions of NFATc and NFATp were similar in the CD4 ϩ T cells from transgenic and control mice. To determine whether FIGURE 1. Constitutive expression of the SWI/SNF complex increases AP-1 expression. A, CD4 ϩ T cells from each mouse were stimulated with PMA (10 ng/ml) and ionomycin (125 ng/ml) for 2 h. Total RNA from the cultured cells was used for reverse transcription. The cDNA samples and the primers specific for each transcription factor were used for real-time PCR. B, CD4 ϩ T cells were incubated for 2 h with or without PMA and ionomycin, as indicated, and the total extracts were prepared. The extracts were immunoblotted with antibodies specific to SRG3, c-Fos, c-Jun, NF-B, or NFATc. The same extracts were analyzed with an antiserum against actin as the control. C, purified CD4 ϩ T cells were stimulated with medium alone or with PMA and ionomycin for 16 h as indicated, and the nuclear extracts were prepared. Gel mobility shift assays were performed using 32 P-labeled probes containing the AP-1 binding sequence.
the increased expression of AP-1 resulted in an increase in its transcriptional activity, we performed a gel mobility-shift assay. As seen in Fig. 1C, stimulation with PMA and ionomycin induce a significant increase in the AP-1 activity in the CD4 ϩ T cells from CD2-SRG3 mice, in comparison with the corresponding levels in control cells.
The SWI/SNF Complex Controls AP-1 Expression-To confirm the role of the SWI/SNF complex in AP-1 expression, the SRG3 expression was decreased by transduction of retroviral vectors carrying a SRG3-targeting short hairpin sequence (shSRG3). Transduction of a shSRG3-encoding retrovirus into thymoma cell lines led to substantial depletion of SRG3 expression, in comparison with the expression in the controls (EL4 cells, 70% decrease; L7 cells, 75% decrease; data not shown). The expression of c-Fos and c-Jun in the EL4 and L7 thymoma cells that were transduced with the silencing construct were markedly lower than the expression in the cells transduced with a control vector ( Fig. 2A). Moreover, the DNA binding activity of AP-1 in the shSRG3-expressing EL4 and L7 thymoma cells was lower than that in the control cells (Fig. 2B). Furthermore, when purified normal CD4 ϩ T cells were infected with the retroviral construct, shSRG3 expression induced a reduction in the expression of c-Fos and c-Jun transcripts, in comparison with the expression in the control cells (Fig. 2C). Additionally, down-regulation of Brg1, the core ATPase component of the SWI/SNF complex, by RNA interference (shBRG1) also induced substantial depletion of c-Fos and c-Jun transcripts in the CD4 ϩ T cells (Fig. 2C). Furthermore, we observed lower levels of NF-B transcripts in the shSRG3-or shBRG1-expressing cells. Next, we investigated whether the SWI/SNF complex is required for AP-1 expression in other cell types. NIH-3T3 fibroblast cells were transduced with shSRG3 or shBRG1, and FIGURE 2. The SWI/SNF complex controls AP-1 expression and DNA binding activity. A, EL4 and L7 cells were transduced with empty vector or shSRG3encoding retrovirus encoding. The GFP ϩ EL4 or L7 cells were sorted and stimulated for 2 h with medium alone or with PMA and ionomycin for 2 h, as indicated. Whole cell extracts were prepared and analyzed by Western blotting for c-Fos, c-Jun, and actin. B, sorted GFP ϩ EL4 or L7 cells were stimulated for 2 h with medium alone or with PMA and ionomycin, as indicated. Nuclear extracts were prepared from each sample. Gel mobility shift assays were performed using 32 P-labeled probes containing the AP-1 binding sequences. Results are representative of three independent experiments. C, CD4 ϩ T cells were transduced with an empty vector or an shSRG3-or shBRG1-encoding retrovirus. The GFP ϩ cells were sorted and stimulated with PMA and ionomycin for 2 h. Then, total RNA was isolated and reverse transcribed. Real-time PCR was performed for each cDNA sample, and primers specific to each transcription factor were used. D, NIH3T3 cells were infected with the vector or a retrovirus encoding a short hairpin sequence targeting SRG3 or Brg1. The infected GFP ϩ cells were cultured without serum for 24 h. Serum was added for various durations, as indicated. Purified RNA samples were used for reverse transcription. The cDNA samples were used for real-time PCR with primers specific for the c-Fos transcription factor. *, p Ͻ 0.05 and **, p Ͻ 0.03 for Student's t test.
the cells were cultured without serum for 24 h. Then, serum was added, and AP-1 expression was measured. Similar to the results shown in Fig. 2D, we observed a reduction in the expression of c-Fos and c-Jun transcripts in the cells that were transduced with shSRG3 or shBRG1. Collectively, these results suggest that the SWI/SNF complex regulates the expression and DNA binding activity of the AP-1 transcription factor.
The SWI/SNF Complex Is Recruited to the AP-1 Promoter-It is known that the SWI/SNF complex regulates the expression of target genes by altering their chromatin structure. Therefore, to determine whether the SWI/SNF complex controls AP-1 expression by modifying the chromatin structures of c-Fos and c-Jun promoters, we performed ChIP assays using EL4 and L7 cells. Notably, SRG3 was found to associate with the c-Fos and c-Jun promoters in unstimulated cells, and although the total SRG3 protein level remained the same after stimulation, the recruitment of SRG3 to the promoters significantly increased (Fig. 3A and data not shown). The c-Fos and c-Jun promoters were also found to associate with Brg1, and this association substantially increased after stimulation. Furthermore, similar recruitment patterns were observed when chromatin from naïve CD4 ϩ T cells was compared with samples from activated cells (Fig. 3B). These results indicate that the SWI/SNF complex regulates AP-1 expression by modifying the chromatin structure of the AP-1 promoter region.
Constitutive Expression of SRG3 Enhances Cytokine Production and Proliferation of Peripheral T Cells-AP-1 and NF-B are inducible transcription factors involved in cytokine production and proliferation of T cells. We observed that the SWI/SNF complex controls the expression of AP-1 in the peripheral T cells; therefore, it is likely that the complex affects proliferation and cytokine production in T cells. To test this hypothesis, we first examined the T cell proliferations in 6 -8week-old SRG3 transgenic mice in response to various stimulations. Peripheral T cells from CD2-SRG3 mice and wild-type littermates were purified and stimulated with ConA in the presence of APCs. The proliferation in the T cells derived from CD2-SRG3 mice was higher than that in the cells derived from wildtype mice (Fig. 4A). The CD4 ϩ and CD4 ϩ CD25 Ϫ T cells purified from the CD2-SRG3 mice proliferated to a greater extent than those from the control mice. Even in the absence of APCs, CD4 ϩ T cells from the CD2-SRG3 mice displayed increased proliferation in response to either anti-CD3 alone or anti-CD3 and anti-CD28 (Fig. 4B).
PMA and ionomycin stimulate the pathways mediated by protein kinase C (PKC) and calcium signaling, and can bypass the requirements for early TCR signaling. To test whether the SWI/SNF complex regulates T cell activation by affecting the downstream factors of PKC and the calcium signaling pathways, we analyzed the proliferation of T cells in response to PMA and ionomycin. Activation by treatment with PMA and ionomycin also resulted in increased [ 3 H]thymidine incorporation in the CD4 ϩ T cells from CD2-SRG3 mice (Fig. 4C). IL-2 is a key growth factor for the proliferation of T cells. The IL-2 expression in the T cells from CD2-SRG3 mice was significantly higher than that in the T cells from control mice (Fig. 4D). To exclude the possibility that the increased thymidine uptake by the T cells from CD2-SRG3 mice was due to reduction in cell death as opposed to accelerated proliferation, we investigated the level of cell death induced by the anti-CD3 treatment. We found that the percentage of annexin V-expressing T cells in CD2-SRG3 mice was only slightly higher than that in control mice (Fig. 4E).
Next, we examined the effects of constitutive expression of the SWI/SNF complex on cytokine production in peripheral CD4 ϩ T cells. Purified CD4 ϩ T cells from CD2-SRG3 transgenic mice and littermate control mice were stimulated with anti-CD3 and anti-CD28 antibodies or PMA and ionomycin in the presence of Golgistop, and the activated cells were stained with antibodies against various cytokines. The IFN-␥and IL-4producing cell fractions in the CD4 ϩ T cells from CD2-SRG3 mice were significantly higher than those in the cells from wildtype mice (Fig. 4F). Thus, these results show that the increased expression of the SWI/SNF complex amplified peripheral T cell proliferation and cytokine production.

. The SWI/SNF complex associates with the c-Fos and c-Jun promoters.
A, EL4 and L7 cells were analyzed directly or stimulated with PMA (10 ng/ml) and ionomycin (125 ng/ml), which was followed by ChIP using normal mouse IgG, anti-SRG3, or anti-Brg1 antibodies as indicated. Then, c-Fos and c-Jun promoter DNA were detected in the precipitate or in the non-fractionated sample (input) by performing PCR within the linear range of amplification. The results are representative of three independent experiments. B, CD4 ϩ T cells were purified and analyzed directly or stimulated with PMA and ionomycin and analyzed by ChIPs using normal mouse IgG, anti-SRG3, or anti-Brg1 antibodies as indicated. A representative autoradiograph from 1 of 3 independent experiments is shown.
We next examined the activation status of CD4 ϩ T cells in wild-type and CD2-SRG3 mice by analyzing the expression of activation markers. In older (5 months or more) mice, CD4 ϩ T cells from CD2-SRG3 showed an increased expression of CD25, CD69, and CD44 compared with those from wild-type mice. These changes were not apparent in younger (4 -6-week-old) mice (supplemental Fig. S2). These data suggest that there are increased numbers of activated or memory CD4 ϩ T cells in aged CD2-SRG3 mice compared with normal mice.
The SWI/SNF Complex Regulates Peripheral T Cell Proliferation and Cytokine Production-To further confirm the role of the SWI/SNF complex in peripheral T cell activation, we analyzed the T cell proliferation and cytokine production after the depletion of the SWI/SNF complex. Purified normal CD4 ϩ T cells were activated to allow retrovirus-mediated transduction, and the expression of SRG3 was decreased by the transduction of shSRG3-carrying retroviral vectors. Then, the transduced CD4 ϩ T cells were restimulated with anti-CD3 and anti-CD28 for 18 h and pulse-labeled with BrdU for 1 h. The population of the BrdUincorporating cells among the CD4 ϩ T cells transduced with the shSRG3-retrovirus was lower than that among the control cells (Fig.  5A, upper panel). BAF57 is the DNA-binding subunit of the SWI/ SNF complex. BAF57⌬N, which is a dominant-negative form of BAF57, lacks the DNA-binding domain of BAF57 (28). When BAF57⌬N was used in transduction experiments, the T cell proliferation was lower than that in the control (Fig. 5A, low  panel). Similarly, EL4 thymoma cells transduced with shSRG3 grew more slowly than the control cells (data not shown). To explore the role of the SWI/SNF complex on IL-2 expression, purified CD4 ϩ T cells were transduced with retroviral vectors targeting SRG3 or Brg1. The transduction of shSRG3 or shBRG1 resulted in a 30% reduction in the expression of the IL-2 transcript in peripheral T cells (Fig. 5B). Furthermore, when we used a luciferase reporter construct driven by the IL-2 promoter (positions Ϫ310 to ϩ10), we observed lower IL-2 promoter activity in the cells that were transduced with shSRG3 or shBRG1 (Fig. 5C).
We also examined cytokine production in T cells after the depletion of SRG3. When CD4 ϩ T cells were transduced with the shSRG3-encoding retrovirus, the proportions of the IFN-␥ and IL-4 producing cells obtained after restimulation were significantly lower than the corresponding proportions for cells infected with the control vector (Fig. 5D). Although the precise in vivo function of T h 17 cells has not been elucidated, T h 17 cells have been shown to regulate autoimmune diseases, including experimental autoimmune encephalomyelitis (EAE) and collagen-induced arthritis (29,30). We also observed that the IL-17-producing cell fraction in the CD4 ϩ T cells transduced with shSRG3 was lower than that in the cells transduced with empty vector (Fig. 1D). Collectively, these results suggest that the SWI/SNF complex regulates T cell proliferation and cytokine production in CD4 ϩ T cells.
Next, we investigated whether the enhanced activation of CD2-SRG3 CD4 ϩ T was caused by the increased AP-1 activity. Because c-Jun is a central component of AP-1 complexes (31), we reduced the expression of c-Jun by using shRNAs targeting murine c-Jun. The transduction of a shc-Jun-encoding retrovi- B, purified CD4 ϩ T cells were stimulated with medium alone or with the indicated dose of plate-bound anti-CD3 either alone or with soluble anti-CD28 (1 g/ml). C, CD4 ϩ T cells were stimulated with medium alone or with PMA (10 ng/ml) and ionomycin (125 ng/ml). D, quantitative real-time PCR was performed using mRNA from the CD4 ϩ T cells. Isolated CD4 ϩ T cells were stimulated with medium alone or with PMA (10 ng/ml) and ionomycin (125 ng/ml) for 2 h, and then, total RNA was purified. PCR was performed using IL-2-specific primers. E, peripheral T cells from wild-type, and CD2-SRG3 mice were stimulated with the indicated doses of plate-bound anti-CD3 for 48 h. Then, the cells were stained with annexin V-PE antibody. F, purified CD4 ϩ T cells were stimulated with plate-bound anti-CD3 and anti-CD28 or with PMA (10 ng/ml) and ionomycin (125 ng/ml). After stimulation, the production of IFN-␥ and IL-4 was detected by intracellular cytokine staining.
rus into a thymoma cell line reduced the expression of c-Jun by more than 70% (data not shown).
Although the transduction of the retrovirus caused a mild decrease in the proportion of IFN-␥-producing cells in the CD4 ϩ T cells from littermate mice, the IFN-␥-producing cell fraction in CD4 ϩ T cells from CD2-SRG3 mice reduced by more than 30%, in comparison with the cells infected with the control vector (Fig. 5C). In addition, knockdown of c-Jun decreased the proportion of INF-␥-producing cells in the CD4 ϩ T cells from CD2-SRG3 mice enough to resemble to the result FIGURE 5. The SWI/SNF complex controls proliferation and cytokine production in peripheral T cells. A, control vector-, shSRG3-, or the dominantnegative BAF57⌬N-transduced CD4 ϩ T cells were activated with anti-CD3 and anti-CD28, and pulsed with BrdU and the proliferating cells were detected. Data are representative of three independent experiments. The BrdU signal of cells in the GFP ϩ gate is shown. B, quantitative real-time PCR was performed using mRNA from CD4 ϩ T cells. The isolated CD4 ϩ T cells were transduced with control, shSRG3, and shBRG1 retrovirus. The infected GFP ϩ cells were sorted and stimulated with PMA and ionomycin for 2 h, and then, total RNA was purified. PCR was performed with IL-2-specific primers. C, infected L7 cells were transfected with the pIL-2-Luc reporter construct. The luciferase activity was measured in cells stimulated with medium alone or with PMA (10 ng/ml) and ionomycin (125 ng/ml) for 2 h. ␤-Galactosidase activity was measured to normalize the transfection efficiency. The unstimulated control was considered to be 1. **, p Ͻ 0.03 for Student's t test. D, purified CD4 ϩ T cells were transduced with empty vector or shSRG3-encoding retrovirus. Infected populations were restimulated with PMA and ionomycin for 5 h. The cells were analyzed by flow cytometry after staining for IFN-␥, IL-4, and IL-17. The cells are gated on GFP ϩ , CD4 ϩ cells. E, purified CD4 ϩ T cells from littermate and CD2-SRG3 mice were transduced with an empty vector or shJun-encoding retrovirus. The infected cells were stimulated with PMA (5 ng/ml) and ionomycin (50 ng/ml) for 4 h, and the IFN-␥ production was detected by intracellular cytokine staining. A representative autoradiograph from 1 of 3 independent experiments is shown. of WT CD4 ϩ T cells infected with the control vector. Therefore, these results suggest that the increased AP-1 expression due to constitutive expression of the SWI/SNF complex may be at least partially responsible for the enhanced activation of CD2-SRG3 T cells.
CD2-SRG3 Mice Are Highly Susceptible to EAE-AP-1 and NF-B activities are linked to various autoimmune diseases involving T lymphocytes, including systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and multiple sclerosis (MS) (19,20,32). EAE, a murine model of the human autoimmune disease MS, is believed to be caused, in part, by alterations in the activation threshold of peripheral T cells (33)(34)(35). In this study, we demonstrated that the CD4 ϩ T cells that overexpressed the SWI/SNF complex are hypersensitive to stimulation and express increased levels of AP-1 and NF-B. In addition, IL-17-producing T cells, which are considered to be involved in the induction of autoimmune diseases, were also affected by the SWI/SNF complex. Therefore, we hypothesized that CD2-SRG3 mice may be more susceptible to the induction of EAE. To test this hypothesis, we immunized CD2-SRG3 (8 -10 week-old) mice and littermate control mice with pertussis toxin and an MOG peptide that was emulsified in complete Freund's, and we scored the signs of clinical disease over a period of 45 days. As shown in Fig. 6A and Table 1, the incidence and severity of EAE in the CD2-SRG3 mice was significantly higher than that in control mice. Notably, the disease occurred much earlier in the CD2-SRG3 mice than in the control mice. To further confirm the increased susceptibility of CD2-SRG3 mice to EAE, we examined the MOG-specific proliferative response of the CD4 ϩ T cells from each mouse. At 15 days post-EAE induction, the CD4 ϩ T cells were purified from the spleen and the draining lymph nodes of the EAE-induced mice. Then, these cells were restimulated with various concentrations of MOG peptides. The peptide-specific proliferative response of the CD4 ϩ T cells isolated from the draining lymph node of CD2-SRG3 mice was higher than that observed in the cells isolated from control mice (Fig. 6B).
Regulatory T cells, which express the transcription factor Foxp3, are a subset of CD4 ϩ T lymphocytes. These cells, which are characterized by high surface expression of the IL-2 receptor ␣-chain (CD25), develop in the thymus and perform suppressor functions (36). Studies in animals and humans suggest that the developmental or functional failure of regulatory T cells results in the development of autoimmune diseases (37). To determine whether the high susceptibility of CD2-SRG3 mice to EAE is due to decreased functioning of regulatory T cells, we investigated the suppression activity of regulatory T cells from CD2-SRG3 mice. In vitro, regulatory T cells (CD4 ϩ CD25 ϩ ) from control and CD2-SRG3 mice were equally capable of suppressing the proliferation of CD4 ϩ CD25 Ϫ responder T cells (Fig. 6C). We found that the Foxp3-expressing cell fraction was rather higher in the CD2-SRG3 mice (data not shown). These results suggest that the increased susceptibility of CD2-SRG3 mice to EAE is not due to reduction of the regulatory T cell function in CD2-SRG3 mice. Thus, mice expressing increased levels of the SWI/SNF complex are much more susceptible to the induction of autoimmune diseases such as EAE.

DISCUSSION
In this study, we have shown that T cells from transgenic mice overexpressing the SWI/SNF complex in the T cell lineage displayed enhanced proliferation and cytokine production, FIGURE 6. SRG3 overexpression exacerbates EAE. A, CD2-SRG3 and wildtype mice were immunized with pertussis toxin (on days 0 and 2) and MOG-(35-55) emulsified in complete Freund's adjuvant, as described under "Experimental Procedures." Disease progression over a period of 45 days was monitored by assigning clinical scores to the mice. The severity of EAE is presented as the mean clinical score in each group. B6, n ϭ 6; CD2, n ϭ 11. *, p Ͻ 0.05; **, p Ͻ 0.03; and †, p Ͻ 0.01 for Student's t test. B, 15 days after EAE induction, CD4 ϩ T cells were purified from the spleen and the draining lymph nodes of EAE-induced CD2-SRG3 and wild-type mice. Then 10 5 CD4 ϩ T cells from each mouse were plated in 96-well plates and restimulated with the indicated concentrations of MOG- . Cell proliferation was measured by monitoring the [ 3 H]thymidine uptake. C, CD4 ϩ CD25 ϩ regulatory T cells and CD4 ϩ CD25 Ϫ responder T cells were sorted from the spleen or the lymph nodes of 4 -6-week-old CD2-SRG3 and wild-type mice. Then, 10 5 CD4 ϩ CD25 Ϫ responder T cells from each mouse were plated in 96-well plates. The indicated ratios of sorted CD4 ϩ CD25 ϩ regulatory T cells were added to each plate. APCs consisted of splenocytes depleted of T cells and treated with mitomycin C. The mixed cells were activated with Con A for 48  which coincided with the enhanced expression of the transcription factor AP-1. The SWI/SNF complex bound to the c-Fos and c-Jun promoters in resting CD4 ϩ T cells, and these bindings were enhanced upon activation. In contrast, knockdown of the SWI/SNF complex in normal CD4 ϩ T cells impaired AP-1 expression, T cell proliferation and cytokine production. Moreover, in comparison with normal mice, SRG3 transgenic mice were much more susceptible to a mouse model of multiple sclerosis. Thus, our results strongly suggest a link between the regulation of AP-1 by the SWI/SNF complex and the regulation of inflammatory responses and autoimmune diseases.
It has been reported that the SWI/SNF complex may control the expression of some cytokine genes by directly acting on their promoters. IFN-␥ promoter-specific recruitment of BRG1 and nucleosomal remodeling was observed during T h 1-specific events (38). In contrast, we found that the SWI/SNF complex regulates the expression of AP-1 and NF-B, which are involved in cytokine expression. Furthermore, in addition to IFN-␥, the expression of other cytokines such as IL-4, IL-2, and IL-17 was also affected by the SWI/SNF complex. Thus, it appears that the SWI/SNF complex regulates the expression of cytokine genes through the induction of AP-1 and NF-B, and, at least in some cases, through direct recruitment of the complex to the regulatory regions of their genes.
When we examined the post-stimulation pattern of tyrosine phosphorylation in the CD4 ϩ T cells, we could not identify significant differences between the cells from wild-type and CD2-SRG3 mice (data not shown). In addition, we found that the SWI/SNF complex regulates T cell activation and cytokine production in response to either antibodies or PMA and ionomycin (Figs. 4 and 5). These results indicate that the SWI/SNF complex regulate T cell activation by affecting the downstream factors of PKC and the calcium signaling pathways. The SWI/ SNF complex regulates the expression of target genes by altering their chromatin structure. Thus, we surmised that the SWI/ SNF complex might control T cell activation by regulating the expression of the downstream target genes of PKC and the calcium signaling. We observed that the expression of AP-1 and NF-B in the T cells derived from CD2-SRG3 mice was significantly higher than that in the cells derived from wild-type littermate mice. The AP-1 and NF-B genes are the immediate early genes induced by PKC and calcium signaling, and these genes are intimately involved in T cell proliferation and cytokine production. Therefore, these findings are consistent with our observations that the SWI/SNF complex affects T cell activation and proliferation by regulating AP-1 expression. This inference is further supported by results demonstrating that the introduction of shJun into CD4 ϩ T cells from CD2-SRG3 mice resulted in significantly reduced IFN-␥ expression, in comparison with cells from wild-type mice (Fig. 5E). The SWI/SNF complex is the general transcription machinery, and the chromatin structure of T cells undergoes a dramatic change after receiving the activation signal. In addition, the SWI/SNF complex appears to show a genome-wide recruitment to chromatin when the T cell is activated. Therefore, it is still possible that the SWI/SNF complex may control activation and proliferation of T cells via additional mechanisms.
AP-1 and NF-B are the primary targets of protein kinase C (PKC) signaling, and PKC-deficient T cells show dramatic defects in AP-1 and NF-B activation (39,40). PKC is the predominant type of PKC in T lymphocytes (32). It has been shown that mature T cells from PKC-deficient (PKC Ϫ/Ϫ ) mice show defects in TCR-induced proliferation, cytokine production, and differentiation (41,42), and PKC Ϫ/Ϫ T cells show impairment in the activation of several transcription factors essential for IL-2 production and T cell proliferation (32,39,43). Interestingly, recent studies suggest that PKC Ϫ/Ϫ mice were resistant to EAE, and that IL-17 expression was impaired in these mice (18). Our results show that the SWI/SNF complex controls the expression of AP-1, NF-B, and IL-17 and that constitutive expression of SRG3 in T cells rendered the mice more susceptible to EAE. Therefore, it is possible that the SWI/ SNF complex controls the activation and proliferation of CD4 ϩ T cells by affecting the PKC signal transduction pathway. However, the detailed relationship between the SWI/SNF complex and PKC in T lymphocytes needs to be investigated.
Previously, we reported that the expression of transgenic SRG3 in the thymus resulted in minor changes in the T cell population after thymic selection (27). Therefore, these results raise a possibility that the enhanced susceptibility of CD2-SRG3 mice to EAE induction was due to autoreactive T cells that escaped from negative selection. Even though this possibility cannot be completely excluded, it is less likely, because the negative selection process is still operational in CD2-SRG3 mice, which is evidenced by the intact endogenous superantigen-mediated negative selection (27). In addition, the TCR repertoires in CD2-SRG3 transgenic mice are very similar to those in control mice, suggesting there were no such dramatic changes in the T cell population in the transgenic mice. However, several studies showed that enhanced T cell activation causes an increase in the incidence and severity of EAE (44,45). Our results showed that the expression level of the SWI/SNF complex affected T cell activation (Figs. 4 and 5). Moreover, the expression of AP-1, NF-B, and IL-17, which has been reported to play an important role in EAE, was also regulated by the SWI/SNF complex in normal T cells (Figs. 2 and 5). Thus, these results suggest that the enhanced EAE induction in CD2-SRG3 mice is at least to some extent due to increased T cell activation and AP-1 activity, which are caused by an increase in the SWI/ SNF complex.
We found that the RNA and protein level expression of SRG3 in a thymoma cell line (23,46) and in peripheral T cells (data not shown) was down-regulated after stimulation through TCR signaling. Because SRG3 controls the stability of other components of the SWI/SNF complex, the down-regulation of SRG3 after TCR signaling induces the reduction of other components of the SWI/SNF complex (26,27). In this study, we have shown that the SWI/SNF complex plays an essential

of MOG-(35-55)-induced EAE
The incidence of disease reflects the number of mice that scored a one or higher. The maximum clinical score is the average of the highest score reached by each mouse. role in T cell activation and proliferation by regulating AP-1 expression. Because the SWI/SNF complex overexpressing T cells are hyperresponsive to TCR stimulation and the CD2-SRG3 transgenic mice are highly susceptible to EAE, the down-regulation of the SWI/SNF complex after activation may be responsible for preventing exaggerated immune responses by the activated T cells. AP-1 and NF-B are the two principal transcription factors involved in many autoimmune diseases. Increased expression of these transcription factors was seen in the T cells from patients with systemic lupus erythematosus (19). In addition, transgenic mice that constitutively expressed a trans-dominant form of IB␣ displayed lower incidence and severity of collagen-induced arthritis, an animal model of human rheumatoid arthritis (47). AP-1 and/or NF-B induce the expression of cytokines and chemokines involved in EAE and MS. Tumor necrosis factor ␣ (TNF␣) has been shown to activate both AP-1 and NF-B, and TNF␣ is a potent pro-inflammatory cytokine that plays an important role in MS (48,49). Matrix metalloproteinase-9 (MMP-9), which is regulated by AP-1 and NF-B, also plays a crucial role in lymphocyte recruitment to the central nervous system in MS, and patients with MS show high levels of MMP-9 in the cerebrospinal fluid (50,51). Up-regulation of AP-1 and NF-B has also been reported in oligodendrocytes and macrophages in MS lesions (21,22). Overall, these studies indicate a strong association between the activities of AP-1 and NF-B and the incidence of autoimmune diseases. In this report, we demonstrated that the mice that constitutively express the SWI/SNF complex in T lineage cells displayed increased levels of AP-1 and NF-B and that these mice were much more susceptible to EAE. Therefore, our results strongly suggest a possible link between the dysregulation of the SWI/ SNF complex and the induction of autoimmune diseases. The detailed relationship between the abnormal activity of the SWI/ SNF chromatin-remodeling complex and the incidence of human autoimmune diseases needs to be elucidated.