Transcriptional repressor DREAM regulates T-lymphocyte proliferation and cytokine gene expression
2005; Springer Nature; Volume: 24; Issue: 20 Linguagem: Inglês
10.1038/sj.emboj.7600810
ISSN1460-2075
AutoresMagali Savignac, Belén Pintado, Alfonso Gutiérrez‐Adán, Małgorzata Palczewska, Britt Mellström, José R. Naranjo,
Tópico(s)T-cell and B-cell Immunology
ResumoArticle22 September 2005free access Transcriptional repressor DREAM regulates T-lymphocyte proliferation and cytokine gene expression Magali Savignac Magali Savignac Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Belen Pintado Belen Pintado Departamento de Reproducción Animal, Instituto Nacional de Investigaciones Agrarias, Madrid, Spain Search for more papers by this author Alfonso Gutierrez-Adan Alfonso Gutierrez-Adan Departamento de Reproducción Animal, Instituto Nacional de Investigaciones Agrarias, Madrid, Spain Search for more papers by this author Malgorzata Palczewska Malgorzata Palczewska Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Britt Mellström Britt Mellström Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Jose R Naranjo Corresponding Author Jose R Naranjo Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Magali Savignac Magali Savignac Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Belen Pintado Belen Pintado Departamento de Reproducción Animal, Instituto Nacional de Investigaciones Agrarias, Madrid, Spain Search for more papers by this author Alfonso Gutierrez-Adan Alfonso Gutierrez-Adan Departamento de Reproducción Animal, Instituto Nacional de Investigaciones Agrarias, Madrid, Spain Search for more papers by this author Malgorzata Palczewska Malgorzata Palczewska Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Britt Mellström Britt Mellström Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Jose R Naranjo Corresponding Author Jose R Naranjo Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain Search for more papers by this author Author Information Magali Savignac1, Belen Pintado2, Alfonso Gutierrez-Adan2, Malgorzata Palczewska1, Britt Mellström1 and Jose R Naranjo 1 1Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, Madrid, Spain 2Departamento de Reproducción Animal, Instituto Nacional de Investigaciones Agrarias, Madrid, Spain *Corresponding author. Departamento de Biología Molecular y Celular, Centro Nacional de Biotecnología, Consejo Superior de Investigaciones Cientificas, 28049 Madrid, Spain. Tel.: +34 91 585 4682; Fax: +34 91 585 4506; E-mail: E-mail: [email protected] The EMBO Journal (2005)24:3555-3564https://doi.org/10.1038/sj.emboj.7600810 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Downstream Regulatory Element Antagonist Modulator (DREAM) is a Ca2+-dependent transcriptional repressor expressed in the brain, thyroid gland and thymus. Here, we analyzed the function of DREAM and the related protein KChIP-2 in the immune system using transgenic (tg) mice expressing a cross-dominant active mutant (EFmDREAM) for DREAM and KChIPs Ca2+-dependent transcriptional derepression. EFmDREAM tg mice showed reduced T-cell proliferation. Tg T cells exhibited decreased interleukin (IL)-2, -4 and interferon (IFN)γ production after polyclonal activation and following antigen-specific response. Chromatin immunoprecipitation and transfection assays showed that DREAM binds to and represses transcription from these cytokine promoters. Importantly, specific transient knockdown of DREAM or KChIP-2 induced basal expression of IL-2 and IFNγ in wild-type splenocytes. These data propose DREAM and KChIP-2 as Ca2+-dependent repressors of the immune response. Introduction The transcriptional repressor Downstream Regulatory Element Antagonist Modulator (DREAM) is a Ca2+-binding protein that binds specifically to DRE sequences in DNA (Carrion et al, 1999). Binding of DREAM to DREs is regulated by the level of nuclear Ca2+, by the interaction with other nucleoproteins like αCREM and CREB and by the PI3 kinase pathway (Carrion et al, 1999; Ledo et al, 2000a, 2002; Sanz et al, 2001). Mutation of two key amino acids within any of the functional EF hands of DREAM (EFmDREAM) results in a protein insensitive to Ca2+ that remains bound to DNA during Ca2+ stimulation (Carrion et al, 1999). Since DREAM binds to DRE sites as a tetramer, EFmDREAM was proposed to function as a dominant active mutant in a background of wild-type (wt) DREAM (Ledo et al, 2000b). DREAM, also known as KChIP-3, belongs to a group of structurally and functionally related Ca2+-binding proteins (KChIP-1 to -4) that interact with Kv4 potassium channels (An et al, 2000) and presenilins (Lilliehook et al, 2002) to regulate channel gating and Ca2+ release from the endoplasmic reticulum, respectively. Engagement of the TCR/CD3 complex triggers a signal transduction cascade that involves early increases in intracellular Ca2+ and diacylglycerol (Cantrell, 1996), the activation of various kinases and phosphatases, and changes in gene expression (Feske et al, 2001). Activation of Ca2+-dependent pathways is required to induce expression of the cytokine genes interleukin (IL)-2, -4 and interferon (IFN)γ (Rao et al, 1997; Macian et al, 2001). The first and best characterized Ca2+-dependent event in the transcriptional control of cytokine genes is the dephosphorylation of NFAT by the Ca2+/calmodulin-dependent protein phosphatase calcineurin (Rao et al, 1997). This modification allows NFAT to translocate from the cytoplasm to the nucleus and activate transcription (Clipstone and Crabtree, 1992; Kubo et al, 1994; Campbell et al, 1996; Rincon and Flavell, 1997). Expression of IL-2, -4 and IFNγ is the signature for different types of effector T helper (Th) cells after differentiation. In response to antigen stimulation, naive Th cells differentiate into different types of effector cells that are classified based on their distinct cytokine profiles and immune-regulatory functions (Mosmann et al, 1986; Paul and Seder, 1994). Th1 cells produce IL-2 and IFNγ, and play an important role in cell-mediated immune response against intracellular pathogens. Th2 cells produce IL-4, -5 and -10, and are involved in humoral immunity and in the allergic response. DREAM is highly expressed in many brain areas and in some peripheral organs, including the thyroid gland, testis and the thymus. While specific target genes for DREAM repression have been described in the brain (Carrion et al, 1999) and in the thyroid gland (Rivas et al, 2004), the role of DREAM in the immune response has not been previously investigated. Here we show that DREAM and the related protein KChIP-2 are expressed in all subpopulations of T cells and that DREAM expression is downregulated after TCR engagement. Expression of the Ca2+-insensitive DREAM mutant, EFmDREAM, in transgenic (tg) lymphocytes reduced proliferation and decreased IL-2, -4 and IFNγ production following TCR engagement and during primary antigen-specific T-cell responses. Importantly, DREAM binds to and represses the transcriptional activity of these cytokine promoters, suggesting a direct repression. Furthermore, specific transient knockdown of DREAM or KChIP-2 induced basal expression of IL-2 and IFNγ mRNAs in wt splenocytes. Results DREAM is downregulated upon TCR stimulation DREAM mRNA is abundantly expressed in the thymus (Carrion et al, 1999). To understand its function in the immune system, we first wanted to know the expression pattern of DREAM and related KChIP genes in the various lymphoid compartments. Qualitative reverse transcriptase (RT)–PCR showed, in addition to DREAM, expression of KChIP-2 in thymus and spleen, while KChIP-1 was barely detectable and KChIP-4 was not present (data not shown). We then quantified DREAM and KChIP-2 expression in different T-cell populations by real-time RT–PCR. DREAM and, to a lower extent, KChIP-2 were present in thymocytes, splenocytes and in purified T, CD4+ and CD8+ cells from spleen (Figure 1A). Figure 1.DREAM expression is downregulated after stimulation in lymphoid cells. (A) Quantitative real-time RT–PCR analysis of DREAM and KChIP-2 mRNA levels in thymocytes, splenocytes and purified T, CD4+ and CD8+ cells. Results are representative of two experiments. (B) DREAM mRNA levels in thymocytes or splenocytes at different times after stimulation with αCD3 (4 μg/ml). Results are expressed as fold induction compared to nonstimulated cells and are representative of four experiments. Download figure Download PowerPoint To investigate a potential role for DREAM and KChIP-2 in the immune response, we analyzed changes in both transcripts upon TCR activation. Primary cultured thymocytes and splenocytes were stimulated for different times with plate-bound αCD3. The levels of DREAM mRNA were decreased as early as 30 min following TCR engagement and the decrease was still noticeable 24 h later (Figure 1B), while KChIP-2 expression was unaffected (data not shown). The downregulation of DREAM after TCR engagement indicates a potential role for DREAM in T-cell function. EFmDREAM tg mice, a tool to study DREAM function in vivo Since genetic ablation of DREAM did not result in any noticeable phenotype in the immune system (Cheng et al, 2002), we aimed to analyze the function of DREAM and KChIP-2 in T cells by developing a tg mouse model using a DREAM mutant insensitive to Ca2+, EFmDREAM, that could block the Ca2+-dependent derepression function of DREAM and other KChIP proteins endogenously expressed. Conceptually, the strategy was based on the known property of DREAM to bind DNA and repress transcription when forming a tetramer (Carrion et al, 1999; Osawa et al, 2001). Since the other KChIP proteins are structurally and functionally related to DREAM, we analyzed whether DREAM and KChIP-2 are able to physically interact and form functional heterotetramers. We performed yeast two-hybrid assays using EF-mutated proteins, since oligomerization has been shown to be regulated by Ca2+ (Osawa et al, 2005). The DREAM–DREAM interaction was confirmed and also a specific interaction between DREAM and KChIP-2 could be detected by this assay (Figure 2A). In addition, we performed coimmunoprecipitation experiments after transient cotransfection of Flag–DREAM and Myc–KChIP-2 expression vectors in Hela cells. An anti-Flag antibody successfully immunoprecipitated Myc–KChIP-2 (Figure 2B), confirming the DREAM/KChIP-2 interaction. Figure 2.EFmDREAM interacts with DREAM and KChIP-2 and blocks Ca2+-mediated derepression. (A) Yeast two-hybrid assay showing the specific interactions between EFmDREAM and DREAM or KChIP-2. pAS2-1 is the empty vector used as control. Use of low (DDO) or high (QDO) selective medium is indicated. (B) Coimmunoprecipitation analysis of Flag–DREAM and Myc–KChIP-2 after transient transfection in HeLa cells. (C) Repression of the DRE-containing pHD3CAT reporter by DREAM, KChIP-2, EFmDREAM or the combinations of EFmDREAM with DREAM or KChIP-2 (in a 1:3 ratio) after transient cotransfection in HEK293 cells. Treatment with caffeine was started 6 h before harvesting the cells. Results are the mean±s.d. of triplicates and are representative of three independent experiments. Download figure Download PowerPoint To ascertain the formation of functional heterotetramers and, more importantly, to investigate whether the EFmDREAM protein could act as a cross-dominant mutant for the DREAM transcriptional function, we performed transient transfections with a DRE-reporter (pHD3CAT) and DREAM or KChIP-2 expression vectors, alone or in combination with EFmDREAM in HEK293 cells. As previously reported (Carrion et al, 1999; Link et al, 2004), DREAM, EFmDREAM and KChIP-2 repressed basal transcription from the pHD3CAT reporter (Figure 2C). Stimulation by caffeine relieved DREAM- and KChIP-2-induced repression, transactivating five- to six-fold the expression from the reporter (Figure 2C) by a mechanism that involves Ca2+ release from intracellular stores (Hernandez-Cruz et al, 1990) and increased CREB-dependent transcription (Carrion et al, 1999; Ledo et al, 2002). However, repression by the EFmDREAM was only slightly reversed by caffeine through an EF-independent mechanism not yet clarified (Figure 2C). Cotransfection of EFmDREAM with DREAM or KChIP-2 in a ratio 1 to 3 repressed basal transcription to a similar extent as each expression vector alone; however, the presence of EFmDREAM blocked the five- to six-fold derepression following caffeine stimulation, indicating that heterotetramers between EFmDREAM and DREAM or KChIP-2 remained bound to DNA in the presence of high Ca2+ concentrations (Figure 2C). Block of the DRE-dependent caffeine-induced derepression by EFmDREAM has been previously shown to be still noticeable at ratios of up to 1:6 of EFmDREAM with DREAM or KChIP-2 (Ledo et al, 2000b). Thus, since EFmDREAM blocks the functional response of DREAM and KChIP-2 to [Ca2+]i increase, preventing derepression, hereafter, we will refer to EFmDREAM as a cross-dominant active mutant for the Ca2+-mediated derepression function, though EFmDREAM is still a DRE-dependent repressor. Normal development but impaired proliferation in EFmDREAM tg T-cells In order to characterize the role of DREAM and KChIP-2 in the immune function in vivo, we used tg mice that ubiquitously express the EFmDREAM mutant. Two lines of tg mice that showed different levels of transgene expression in lymphoid organs were used. As quantified by real-time RT–PCR (Figure 3A), expression of the transgene in the thymus and spleen of line 33 was seven- and six-fold, respectively, higher than in line 1, in which the levels of EFmDREAM and endogenous DREAM mRNAs were at ratios 1:1 and 1:5 in thymus and spleen, respectively. Importantly, in spite of the different expression of the EFmDREAM transgene in the two lines, the results reported hereafter correspond to experiments performed with line 1 and were completely comparable to data from line 33, indicating that a low level of expression of EFmDREAM, in the same range as endogenous DREAM or lower, is sufficient to disturb DREAM and KChIP-2 function in T cells. Figure 3.The T-cell proliferative response is impaired in EFmDREAM tg mice. (A) Quantitative real-time RT–PCR analysis of the expression of EFmDREAM in the thymus and spleen of trangenic mice from lines 1 and 33. Results are the mean±s.d. of 4–8 mice. (B) Flow-cytometric analysis of cells from thymus and spleen. Single-cell suspensions were stained with the indicated antibody combination. Positive cells (%) within a quadrant are indicated. The results are representative of six wt and six tg mice. (C) Cell proliferation was quantified in wt and tg thymocytes stimulated with the indicated amount of plate-bound αCD3 alone or αCD3 (6 μg/ml) in combination with αCD28 for 48 h. Results are the mean±s.d. of four mice per group and are representative of three independent experiments. (D) Cell proliferation was analyzed in wt and tg thymocytes 48 h after stimulation with plate-bound αCD3 (6 μg/ml) in the presence or absence of recombinant murine IL-2 (20 ng/ml). Results are the mean±s.d. of four mice per group and are representative of two independent experiments. Asterisks represent the statistical significance versus the appropriate control in each case. *P<0.05, **P<0.01 and ***P<0.001. Download figure Download PowerPoint Analysis of 6–8-weeks-old tg mice showed that expression of EFmDREAM in lymphoid compartments did neither affect total cellularity nor the distribution of single- and double-positive CD4 and CD8T lymphocytes compared to wt thymus or peripheral lymphoid organs (Figure 3B). Furthermore, evaluation of several lymphoid markers in the thymus, spleen and lymph nodes of tg mice did not reveal major abnormalities in the development of the immune system (data not shown). Thus, overexpression of EFmDREAM does not affect the normal development of T lymphocytes. Since TCR activation results in T-cell proliferation and sustained downregulation of DREAM mRNA levels, we wanted to check whether DREAM could function as a negative regulator of lymphocyte proliferation. Proliferation of thymocytes stimulated with αCD3 alone or in combination with αCD28 was significantly lower in EFmDREAM expressing T cells than in wt cells (Figure 3C), without change in apoptosis (Supplementary Figure S1). The reduced proliferative response after polyclonal stimulation was still noticeable even when the experiment was performed in the presence of saturating concentrations of recombinant IL-2, though in these conditions both wt and tg T cells showed an increased response (Figure 3D). Given that the proliferative response in tg cells was not restored in the presence of IL-2, we analyzed the expression of IL-2Rα chain (CD25) and CD69, well-characterized markers of the early phase of T-cell activation. Interestingly, the induction of CD25 and CD69 was smaller in both tg CD4+CD8− and CD4−CD8+ T cells as compared to the corresponding wt cells (Figure 4A and B). Thus, EFmDREAM is able to function as a negative regulator of TCR-induced lymphocyte proliferation. Figure 4.Reduced expression of activation markers in tg T cells. (A, B) Thymocytes were stimulated with plate-bound αCD3 (6 μg/ml) and analyzed by cytometry for CD25, CD69, CD4 and CD8 expression after 24 h. (A) shows a representative example and (B) shows the %±s.d. of CD25+ and CD69+ cells of the gated CD4+CD8− (left panel) or of the gated CD4−CD8+ (right panel) after stimulation of four mice per group. Asterisks represent statistical significance versus the appropriate control in each case. *P<0.05 and ***P<0.001. Download figure Download PowerPoint Reduced cytokine expression in tg lymphoid cells T-cell proliferation is dependent on cytokine gene expression, which has been shown to be Ca2+-dependent (Rao et al, 1997); hence, we analyzed the cytokine response in tg T cells. Importantly, induction of both IL-2 and IFNγ in tg thymocytes (Figure 5A and B) or purified CD4+ tg T cells from the spleen (Figure 5C and D) was weaker than in wt cells. The reduction was observed following αCD3 (Figure 5A–D) or αCD3 plus αCD28 (data not shown) stimulation and irrespectively of the antibody concentrations used, and remained significantly diminished up to 72 h after the stimulation (data not shown). Furthermore, the polarized effector T-cell tg populations Th1 and Th2 showed a significant reduction in IFNγ and IL-4 production, respectively (Figure 5E and F). Thus, in addition to IL-2 and IFNγ, expression of the Th2-specific cytokine IL-4 is as well repressed in EFmDREAM tg mice. Figure 5.Decreased IL-2 and IFNγ protein in tg T cells. Thymocytes (A, B) or CD4+ T (C, D) from wt and tg mice were stimulated for 48 h with different concentrations of αCD3. Levels of IL-2 (A, C) or IFNγ (B, D) were quantified by ELISA. In (A, B), results are the mean±s.d. of five mice per group and are representative of four independent experiments. In (C, D), results are the mean±s.d. of triplicates and are representative of two experiments. (E, F) CD4+ T cells from wt and EFmDREAM mice were cultured under Th1 or Th2 differentiating conditions for 2 weeks and cytokine levels analyzed. Results are representative of three experiments. Asterisks represent statistical significance versus the appropriate control in each case *P<0.05, **P<0.01 and ***P<0.001. Download figure Download PowerPoint Moreover, we examined the effect of EFmDREAM on the primary antigen-specific T CD4+ cell response. Mice were immunized with ovalbumin (OVA)–complete Freund's adjuvant (CFA) and the polarization of the antigen-specific T-cell response in the draining lymph nodes was tested 9 days later. The number and the repartition of CD4+ and CD8+ cells in the lymph node were not different between wt and tg mice (data not shown). Thus, EFmDREAM does not modify lymphocyte recruitment in lymph nodes. Exposure to OVA induced IL-2 and IFNγ production (Figure 6A and B) and T-cell proliferation (Figure 6C) that were significantly reduced in EFmDREAM as compared to wt lymphocytes. These data indicate that expression of EFmDREAM downregulates cytokine production both after polyclonal activation and during antigen-specific T-cell response. Figure 6.EFmDREAM decreases cytokine production following antigen-specific stimulation. Wt and tg mice were immunized s.c. with 50 μg of OVA in CFA. Draining lymph node cells were harvested 9 days later and cultured in the presence of the indicated concentration of OVA. (A) IL-2, (B) IFNγ secretion and (C) antigen-induced proliferation were quantified after 48 h. Data are expressed as mean±s.d. of five mice per group. Results are representative of two experiments. Asterisks represent statistical significance versus the appropriate control in each case. *P<0.05, **P<0.01 and ***P<0.001. Download figure Download PowerPoint DREAM binds to and represses the IL-2, -4 and IFNγ promoters Since DREAM and KChIP-2 proteins display different functions in different cellular compartments (Carrion et al, 1999; An et al, 2000; Ledo et al, 2002; Lilliehook et al, 2002), we first excluded that expression of EFmDREAM could modify the early and late signaling pathways that activate cytokine gene expression (Supplementary Figure S2) to then explore a direct repressor effect by DREAM and KChIP-2 at the transcriptional level. We searched for potential DRE sites (Carrion et al, 1998; Ledo et al, 2000b) in the promoter regions of IL-2, -4 and IFNγ genes. One direct DRE site (GTCA) and one inverted DRE (TGAC) are present in the mouse IL-2 and IL-4 genes at positions +74 and +39 downstream from the transcription initiation site, respectively, while the mouse IFNγ gene contains a DRE site at position +20 (Figure 7A). EMSA using thymocyte extracts from wt mice and oligonucleotide probes encompassing these DRE sites revealed specific retarded bands. As shown for DREIL-2, protein–DNA complexes could be competed by an excess of cold DREIL-2 and DREDyn, but were unaffected by competition with a cold unrelated oligonucleotide (Sp1). Incubation with a specific αDREAM antibody substantially reduced the protein–DNA complex of the DREIL-2, indicating that endogenous DREAM is participating in the retarded bands (Figure 7B). Similar results were observed with oligonucleotide probes encompassing DREIFNγ and DREIL-4 sites (data not shown). These results indicate that the DRE sites identified in the promoters of IL-2, IL-4 and IFNγ genes can bind DREAM in vitro. To show the dynamic binding of DREAM and KChIP-2 to DRE sites, we performed EMSA using extracts from HEK293 cells transiently transfected with DREAM or KChIP-2 and prepared in the presence of EGTA or CaCl2. Binding of DREAM or KChIP-2 to the IL-2 probe was reduced in the presence of Ca2+ (Figure 7C). These data show that DREAM and KChIP-2 are able to bind DRE sites in a Ca2+-dependent manner. Figure 7.DREAM regulates the transcriptional activity of IL-2, -4 and IFNγ promoters in vivo. (A) Putative DRE sites present in IL-2, IFNγ and IL-4 regulatory regions. Arrows indicate the orientation of the DRE sequence. (B) EMSA using extracts from wt thymocytes and the DREIL-2 probe. Competitions with DREIL-2, DREDyn and SP1 are shown. The arrow indicates the specific DRE retarded band eliminated by preincubation with αDREAM (Ab1014). The asterisks mark the appearance of nonspecific bands after serum incubation. (C) EMSA using extracts from HEK293 cells transiently transfected with DREAM or KChIP-2 and the DREIL-2 probe. The arrow indicates the specific retarded band that is competed with Ca2+. (D) DRE-dependent DREAM repression of IL-2, -4 and IFNγ promoters. Plasmid pRL-CMV was used to correct for transfection efficiency. The transfection experiments were repeated twice in quadruplicates. Asterisks represent statistical significance versus the appropriate control in each case. *P<0.05 and ***P<0.001. (E) ChIP using tg thymocytes or (F) Jurkat cells nonstimulated cells (−) or cells after 2 h of PMA/Iono stimulation (+). Download figure Download PowerPoint To investigate whether the binding of DREAM and KChIP-2 to the DRE sites in cytokine promoters is functional and affects their basal transcription, we performed transient transfection experiments in HEK293 cells. Expression of DREAM or KChIP-2 significantly repressed basal transcription from the pIL2CAT, pIFNγCAT or pIL4Luc reporters (Figure 7D and Supplementary Figure S3), but failed to repress transcription from reporter plasmids pIL2mDRECAT, pIFNγmDRECAT and pIL4mDRELuc, in which the DRE sites are mutated (Figure 7D and Supplementary Figure S3). To test whether the binding of EFmDREAM on cytokine promoters also occurs in vivo, we performed chromatin immunoprecipitation using a specific αDREAM antibody and genomic DNA from acutely dissociated tg thymocytes. αDREAM specifically immunoprecitated chromatin fragments containing promoter regions for the three cytokine genes (Figure 7E), while an irrelevant antibody failed to do so (Figure 7E). Furthermore, the αDREAM antibody did not immunoprecipitate chromatin containing the β-actin promoter that was used as a negative control (Figure 7E). Moreover, using genomic DNA from Jurkat T cells, the binding of DREAM to the IL-2 promoter was reduced by 65% after 2 h of Ca2+ stimulation (Figure 7F). Real-time quantitative PCR revealed reduced levels of IL-2 and IFNγ both in tg thymocytes and splenocytes in basal conditions and at different times after stimulation, supporting a transcriptional effect of DREAM on cytokine gene expression (Supplementary Figure S4). Taken together, the results from EMSA, transient transfections and ChIP assays support a Ca2+-dependent binding of DREAM and KChIP-2 to cytokine promoters. Specific transient knockdown of endogenous DREAM or KChIP-2 upregulates IL-2 and IFNγ expression To substantiate the role of endogenous DREAM and KChIP-2 proteins as Ca2+-dependent repressors of the cytokine response, we performed specific transient knockdown of DREAM or KChIP-2 expression and analyzed the effect on basal levels of IL-2 and IFNγ mRNA. In these experiments, we used an antisense DREAM expression vector (AS-DREAM) previously shown to dramatically reduce basal levels of endogenous DREAM protein in PC12 cells and to increase basal DRE-dependent transcription (Ledo et al, 2002), and a similar antisense KChIP-2 expression vector (AS-KChIP-2). Since DREAM and KChIP-2 show a high homology in their nucleotide sequence, we first characterized the specificity of each antisense vector using transient cotransfection of AS-DREAM or AS-KChIP-2 and HA-tagged DREAM or KChIP-2 expression vectors in HEK293 cells. Importantly, AS-DREAM substantially reduced HA-DREAM protein levels (40% decrease) without affecting the level of HA-KChIP-2 protein or Myc-KIAA1007, an unrelated corepressor protein used as a loading control (Figure 8A, upper panels). Conversely, AS-KChIP-2 reduced HA-KChIP-2 protein levels (32% decrease) without affecting the level of HA-DREAM or Myc-KIAA1007 (Figure 8A, lower panels). Given that TCR engagement downmodulates DREAM expression, we analyzed the effect of specific transient knockdown of DREAM or KChIP-2 in basal condition. The specific transient knockdown of DREAM or KChIP-2 in wt splenocytes showed a significant increase in IL-2 mRNA, and a modest but reproducible increase in IFNγ mRNA (Figure 8B). Interestingly, cotransfection with both antisense vectors resulted in a six- and two-fold increase in IL-2 and IFNγ mRNA, respectively (Figure 8B). This represents 10 and 28% of the maximum inducible level after 4 h of αCD3 stimulation for IL-2 and IFNγ, respectively. Thus, by only removing the DREAM- and/or KChIP-2-mediated repression, we observed a substantial induction of the basal expression without activating signaling pathways, that is, NFAT and MEF-2, known to participate in the up-regulation of cytokines. The level of IL-4 mRNA was below the detection limit both in vector- and in AS-DREAM-transfected cells (data not shown), which agrees with the Th1-prone cytokine response reported for B6 mice. In conclusion, the transcriptional repressors DREAM and KChIP-2 control basal expression of IL-2 and IFNγ genes. Figure 8.Specific transient knockdown of DREAM or KChIP-2 induces cytokine expression in splenocytes. (A) Western blot analysis of HEK293 cells transiently cotransfected with AS-DREAM or AS-KChIP-2 together with HA-DREAM or HA-KChIP-2 and Myc-KIAA1007 (loading control). The upper panel shows the specific knockdown of the HA-DREAM protein by AS-DREAM (arrowhead), and the lack of effect on HA-KChIP-2 (asterisk) or Myc-KIAA1007. The lower panel shows the specific knockdown of the HA-KChIP-2 protein by AS-KChIP-2 (asterisk) and the lack of effect on HA-DREAM (arrowhead) or Myc-KIAA1007. (B) Quantitative analysis of IL-2 and IFNγ mRNA ex
Referência(s)