Systemic lupus erythematosus T cells display decreased amounts of TCR ζ mRNA that results in part from limited binding of the transcriptional enhancer Elf-1 to the TCR ζ promoter. We have identified a new cis-binding site for the cAMP response element (CRE) modulator (CREM) on the TCR ζ promoter, centered on the −390 nucleotide. Transfection of T cells with an antisense CREM α plasmid reduced the binding of CREM to the TCR ζ promoter, as shown by chromatin and reporter chromatin immunoprecipitation assays, and enhanced the production of TCR ζ mRNA and protein. Mutagenesis of the −390 CRE site prevented the binding of CREM to the TCR ζ promoter. The mechanism of CREM-mediated repression appears to be chromatin dependent, because antisense CREM promotes the acetylation of histones on the TCR ζ promoter. Finally, we established an enhanced binding of CREM to the TCR ζ-chain promoter in systemic lupus erythematosus cells compared with control T cells. Our studies demonstrate that CREM α binds to the TCR ζ promoter and repress its activity.
Systemic lupus erythematosus (SLE)3 is an autoimmune disease of unknown etiology that is characterized by multiple T cell signaling abnormalities (1). T cells from patients with SLE have been shown to have decreased levels of TCR ζ protein and mRNA (2). The decreased levels of TCR ζ mRNA have been shown to be the result of decreased TCR ζ promoter activity (3). In normal T cells, the transcriptional activity of the TCR ζ promoter is regulated through the binding of Elf-1, a member of the Ets family of transcription factors, to two distinct binding motifs defined by the TCR ζ promoter (4). Mutagenesis of these motifs greatly diminishes the promoter activity of this TATA-less promoter (4, 5). No other obvious cis sites have been noted on the proximal TCR ζ promoter.
While studying the causes for the transcriptional repression of the TCR ζ promoter in SLE T cells, we identified two groups of patients. In the first group the 98-kDa form of Elf-1 that binds to the TCR ζ promoter and accounts for its activity was absent despite the fact that the 80-kDa form was present at levels comparable to those in normal T cells. In the second group, the 98-kDa form was present, but it was not able to bind to the TCR ζ-chain promoter. In these patients the 98-kDa form resolved into two bands in isoelectric focusing gels, whereas the DNA binding 98-kDa form resolved into three bands (3). In the same experiments we noticed that although TCR ζ promoter activity in SLE T cells was decreased, mutagenesis of the 2 Elf-1 cis sites of the TCR ζ promoter further decreased its activity (3). This stepwise decrease in promoter activity suggested the presence of additional transcription factors involved in the transcriptional repression of the TCR ζ promoter. The presumed additional transcriptional repression could reflect the result of transcriptional enhancers that were either absent or defective in SLE T cells or the presence of transcriptional repressors. In addition, inspection of the proximal TCR ζ-chain promoter revealed seven cAMP response element (CRE) half-sites (TGAC or GTCA) to which members of the CRE-binding protein/CRE modulator (CREM) family of transcription factors could bind.
Because SLE T cells have been shown to express increased amounts of CREM α that binds to the −180 site of the IL-2 promoter and represses its activity (6, 7), we considered that CREM α may bind to an as yet unidentified site on the TCR ζ promoter and repress its activity. In this study we demonstrate that indeed CREM α binds to a palindromic CRE half-site (TGAC or GTCA) centered at the −390 nt of the TCR ζ promoter and represses its activity. In addition, we show increased binding of CREM α to the TCR ζ promoter in SLE T cells.
Materials and Methods
Heparinized peripheral venous blood was obtained from healthy volunteers, patients with rheumatoid arthritis, and five patients with the diagnosis of SLE according to American College of Rheumatology criteria (8). Non-T cells were selected by a tetrameric Ab mixture against CD14, CD16, CD19, CD56, and glyA (StemCell Technologies) and bound to erythrocytes. These complexes were separated from the T cells by a Lymphoprep gradient (Nycomed). The purified T cells were >98% positive for CD3, as tested by flow cytometry.
Abs against CREM and TCR ζ-chain were purchased from Santa Cruz Biotechnology.
Preparation of mRNA and cDNA, PCR, and real-time PCR
One million T cells were used for extracting RNA (RNA Easy Mini kit; Qiagen). RNA was quantified, and 500 ng of total RNA was used for cDNA synthesis by RT (RT-PCR kit; Promega). A total of 50 ng of cDNA was used for each PCR. PCR primers were synthesized by Sigma Genosys.
PCR beads were used for amplification (BD Biosciences). Real-time PCR was conducted with a Cepheid Smart Thermocycler by adding SYBR Green to the reaction mixture. Primers used for PCR were as follows: ζ-chain, 5′ 3′; reverse, 5′ 3′; CREM, 5′-GAA ACA GTT GAA TCC CAG CAT GAT GGA AGT-3′; and reverse, 5′-TGC CCC GTG CTA GTC TGA TAT ATG-3′. PCR products of semiquantitative RT-PCR were separated on a 1.5% agarose gel, and the OD was quantified using QuantityOne software (Bio-Rad) after background subtraction from each band.
Chromatin immunoprecipitation (ChIP) analysis
Five million T cells were used per investigated Ab. The cells were treated with formalin (1%) for 10 min, washed, lysed, and sonicated. The DNA-protein complexes were immunoprecipitated with an Ab and extracted by protein A/G-Sepharose beads (Santa Cruz Biotechnology). After several washing steps, the protein was digested with proteinase K, the cross-link between DNA and protein was reversed at 65°C overnight, and the DNA was extracted (QiaAmp DNA extraction kit; Qiagen). The DNA was amplified with primers flanking the TCR ζ promoter, including the −390 site (forward, 5′-CAACGCCACACAGCAAGTAGG-3′; reverse, 5′-GGCTGGCAAAATGACAGAACC-3′). DNA from ∼1 million cells was used per PCR. PCR products were run on a 1.5% agarose gel and quantified with QuantityOne software.
Transfection of T cells or Jurkat cells
Approximately 5–10 × 106 freshly isolated T cells or PBMCs were transfected by electroporation with an AMAXA electroporator. Plasmids encoding CREM α antisense (a gift from Dr. P. Sassone-Corsi, Institut de Genetique et de Biologie Moleculaire et Cellulaire, Centre National de la Recherche Scientifique, Institut National de la Sante et de la Recherche Medicale, Universite Louis Pasteur, Strasbourg, France) or corresponding empty vector plasmid were used for transfection. One microgram of each plasmid was used per transfection. Alternatively, Jurkat cells were used and transfected with a Bio-Rad electroporator (250 V, 975 μF). After 18 h, T cells were harvested for different purposes, such as FACS analysis, Western blot, ChIP assay, or RT-PCR.
This technique enables study of the in vivo binding of transcription factors to a specific binding site on a promoter of a reporter construct. The TCR ζ promoter-luciferase construct used in this study has been described previously (4). Mutagenesis of the −390 site into luciferase was performed with a site-directed mutagenesis kit (Stratagene) and primers including the mutagenesis site (5′-GGACTCTAATCCAGGGTTCTGGTATTTTGCCAGCCACCAT-3′) (mutated nucleotides are underlined). The mutated construct was transfected into 5 million Jurkat cells, and the nonmutated construct was transfected into another 5 million cells for control purposes. Eighteen hours after transfection, ChIP of the cells was performed with an anti-CREM Ab (Santa Cruz Biotechnology). PCR was performed with a primer specific for the TCR ζ-chain promoter and another one specific for the luciferase plasmid to ensure that only reporter-specific DNA, and not genomic DNA, is amplified.
Preparation of nuclear extracts and EMSA
Five to 10 million T cells were used for preparation of nuclear extracts as previously described (9). The dsDNA probe of the −390 site on the TCR ζ-chain promoter was 5′-CAGGGTTCTGTCATTTTGCCAGCCA-3′ in shift assays as previously described (9). Recombinant GST-tagged CREM (plasmid was a gift from M. Montminy, Peptide Biology Laboratories, The Salk Institute for Biological Studies, La Jolla, CA) and recombinant human ribonucleoprotein-D (hnRNP-D) were purified as previously described (10).
Flow cytometric analysis
Cells were transfected with antisense CREM or an empty vector (psGV) and a plasmid containing GFP. Twenty-four hours after transfection, 1 million cells/assay were permeabilized with a BD Cytofix/Cytoperm Plus kit (BD Biosciences) according to the manufacturer’s instructions and stained with PE-labeled anti-TCR ζ-chain Ab or an isotype-specific control (Santa Cruz Biotechnology). A minimum of 10,000 events were collected for evaluation on a FACSCalibur (BD Biosciences), and data were analyzed using CellQuestPro software. The mean fluorescence intensity was calculated, and the samples were compared by paired t test.
CREM binds to the TCR ζ promoter in vivo
Because CREM α is up-regulated in SLE T cells (6, 7) and suppresses IL-2 production in these cells, we were interested to find out whether CREM binds to the TCR ζ promoter that also displays decreased activity in SLE (3, 11). To investigate whether CREM binds to the TCR ζ promoter in live cells, unstimulated normal T cells were fixed with formalin to cross-link DNA to proteins, then sonicated, and the DNA-protein complexes were immunoprecipitated. The DNA was extracted and amplified with sets of primers specific to the ζ promoter. As shown in Fig. 1 a (representative of four experiments), we detected enhanced binding of CREM to the TCR ζ promoter, although minimal binding of c-Fos was detected.
We have previously reported that the activity of a luciferase reporter construct driven by the IL-2 promoter is reduced by a CREM α plasmid, whereas it is enhanced by an antisense CREM plasmid in T cells (12). We thus transfected normal T cells with the antisense CREM α or control (PSG5) plasmids. In Fig. 1 b (left panel, representative experiment; right panel, cumulative data from four experiments), we show that transfection of T cells with the antisense CREM plasmid prevents the binding of CREM to the TCR ζ promoter.
CREM binds to the −390 site of the TCR ζ promoter
We searched the proximal TCR ζ promoter for CRE binding motifs to which CREM could bind. As shown in Fig. 2, seven such sites were identified, representing one part of the canonical CRE palindrome. To identify which of the seven sites served as a binding sites(s) of CREM, we performed EMSA using seven oligonucleotides defining the palindromic half-sides within −1000 bp of the TCR ζ promoter and recombinant CREM α protein. CREM α was found to bind only to the CRE site centered on the −390 nt site (Fig. 3,a), but not to sites centered on the −219, −352, −439, −483, and −631 sites (Fig. 3,b). The binding was specific, because it was not inhibited by a 50-fold increase in nonspecific oligonucleotides (Elf-1; Fig. 3,a). In addition, the −390 site did not bind hnRNP protein, which, in contrast, bound to the a previously identified hnRNP binding site (13) (Fig. 3 c).
Using a modified reporter construct-based ChIP method, we explored whether the −390 site was also important for in vivo binding. The −390 site of a TCR ζ promoter-luciferase construct was mutagenized and transfected into Jurkat T cells. As a control, we transfected the wild-type construct into the same number of cells. Subsequently, we performed ChIP assays using an anti-CREM α-specific Ab. The PCR was performed using primers specific to the reporter construct to ensure that only plasmid DNA is amplified. As shown in Fig. 4 (upper panel, representative experiment; lower panel, cumulative data from three experiments), mutagenesis of the −390 site prevented the binding of CREM to the reporter construct in vivo. Mutagenesis of the −219-centered TGAC half-site did not affect the binding of CREM α to the TCR ζ promoter (Fig. 4).
Increased CREM binding to the TCR ζ promoter in live SLE T cells
Because CREM is known to be up-regulated in SLE T cells (7), we asked whether CREM binding to the TCR ζ promoter was also increased in SLE T cells. We purified T cells from the peripheral blood of patients with SLE, performed ChIP assays using an anti-CREM Ab, and compared the binding of CREM to the TCR ζ promoter in SLE T cells (n = 5) with that in normal T cells (n = 7) and that in patients with rheumatoid arthritis (n = 3). The PCR was conducted in real-time mode. A representative experiment is shown in Fig. 5,a, and cumulative data are shown in Fig. 5 b. The data show that more CREM binds to the TCR ζ promoter in SLE T cells compared with normal T cells and T cells form patients with rheumatoid arthritis (mean real-time PCR cycle threshold, 34 (SLE) vs 36.7 (normal) vs 38.1 (rheumatoid arthritis)).
CREM binding to the TCR ζ promoter is up-regulated after T cell stimulation and enhances chromatin remodeling
We have previously shown that CREM protein is up-regulated after T cell stimulation and binds to the IL-2 promoter in vivo to terminate IL-2 production by a chromatin-dependent mechanism (14). We therefore conducted experiments to determine whether CREM uses a similar mechanism to regulate the activity of the TCR ζ promoter. We first stimulated T cells with CD3 and CD28 Abs and determined TCR ζ-chain mRNA and protein expression as well as CREM mRNA expression. As shown in Fig. 6,a (representative; n = 3), ζ-chain protein expression decreased dramatically after 24 h. Furthermore, although CREM mRNA expression was strongly enhanced (∼4.4-fold, representing approximately two cycles difference), ζ-chain mRNA synthesis decreased by ∼50% (Fig. 6,b; n = 8). In addition, we observed an enhanced binding of CREM protein to the TCR ζ promoter by ChIP assay (Fig. 6 c, representative experiment).
Next we transfected Jurkat T cells or naive T cells with either an antisense CREM or an empty vector and determined whether it affected the binding of acetylated histone 4 to the TCR ζ promoter 24 h after transfection. As shown in Fig. 7 (representative of four experiments), antisense CREM enhanced the acetylation of histone 4 next to the TCR ζ promoter, although in the empty vector, which does not prevent the binding of CREM, enhanced deacetylation of histone 4 occurred. These data suggest a chromatin-dependent mechanism by which CREM influences TCR ζ promoter activity.
Antisense CREM enhances TCR ζ-chain expression
After showing that CREM binds to the TCR ζ promoter and influences chromatin remodeling, we asked whether CREM is able to regulate the expression of the endogenous TCR ζ gene. To this end we transfected T cells with antisense CREM or empty vector constructs and determined the levels of cell surface expression of TCR ζ protein in gated GFP-positive cells and the levels of TCR ζ-chain mRNA by PCR. We performed flow cytometric analysis on transfected cells and gated on the GFP-positive cells. As shown in Fig. 8,a (representative) and Fig. 8,b (cumulative data; n = 12; p = 0.0015), transfection of T cells with an antisense CREM construct resulted in a 26% up-regulation of TCR ζ-chain expression on the cell surface membrane, which was associated with parallel increase in the levels of TCR ζ-chain mRNA (n = 5; Fig. 8 c).
In this communication we present evidence that besides Elf-1, CREM α is involved in regulation of the activity of the TCR ζ promoter. Rellahan et al. (4) have established that the proximal TCR ζ promoter defines two Elf-1 sites that are important regulators of TCR ζ gene expression. CREM is a transcriptional repressor known to be of importance in SLE T cell pathophysiology and to repress the transcription of both IL-2 (6, 7) and c-fos genes (15). TCR ζ mRNA and protein expression are diminished in patients with SLE (2), and we have shown that decreased DNA binding of Elf-1 to contribute to the decreased activity of the TCR ζ promoter (3). In this study we show that CREM α binds to the −390 site of the TCR ζ promoter using shift, ChIP, and reporter construct-based ChIP assays. We have also shown that this binding represses transcription of the TCR ζ-chain gene and expression of the ζ-chain.
It is unclear at this point whether and how CREMα interacts with Elf-1 before or after binding to the TCR ζ-chain promoter. Elf-1 undergoes phosphorylation and O-GlcNAc glycosylation in two phases: during the first phase, an 80-kDa form is produced that does not bind to DNA, whereas a 98-kDa form is produced after T cell activation (3). Elf-1 defines at least 13 phosphorylation and O-GlcNAc glycosylation sites, and the order and significance of each modification are unknown. More importantly, posttranscriptional modification defects in SLE T cells lead to either defective production of the DNA-binding 98-kDa form or a wrongly phosphorylated 98-kDa form, which, again, does not bind to DNA. Although it is possible that in SLE T cells the mere absence of Elf-1 binding to the TCR ζ promoter permits the unbalanced repressive action of CREM, it is unclear though how Elf-1 and CREM interact in normal T cells.
This study demonstrates CREM to be involved in the transcription repression of a third gene in immune cells. Previously, we had demonstrated that CREM binds to the IL-2 promoter (6) and represses its activity. Indeed, antisense CREM can effectively augment the transcriptional activity of the IL-2 promoter (7). Similarly, CREM binds to the c-fos promoter and represses its activity, and a decrease in the levels of CREM by siRNA CREM enhances the expression of c-fos mRNA and protein and the formation of AP1 heterodimers (15).
We have shown previously that CREM is induced in normal human T cells after stimulation, and we have proposed that it is involved in the termination of IL-2 production (14). The data presented in this study demonstrate that CREM can exert the same effect on TCR ζ gene transcription, a mechanism that could help terminate the T cell response. Under physiologic conditions, the T cell response after stimulation of TCR leads to the up-regulation of cytokines, such as IL-2. At the same time, mechanisms that will eventually silence the expression of these proinflammatory genes are activated, thus making sure that the T cell activation will be terminated in a timely fashion. Along the same lines, the expression of TCR ζ mRNA and protein is down-regulated shortly after activation of the T cell, a mechanism that limits further TCR-mediated stimulating signals.
How does binding of CREM to the −390 site of the TCR ζ-chain promoter enable the repression of its activity? CREM α exerts its transcriptional repression activity by the mere fact that it misses the Q1 and Q2 domains that are responsible for its association with the transcription coactivators, CREB binding protein and p300 (16). We have shown previously that binding of CREM to the IL-2 promoter enhances the deacetylation of histones associated with the promoter, an effect that limits the accessibility of the promoter to endonucleases (14). In this communication we show that similar to the IL-2 promoter, binding of CREM to the −390 site of the TCR ζ promoter facilitates the deacetylation of histones next to the TCR ζ promoter, rendering the promoter inaccessible to basal transcription machinery and thus terminating transcription.
In conclusion, we have demonstrated the involvement of a second transcription factor in regulation of the activity of the TCR ζ promoter. The enhancing effect of Elf-1 is apparently balanced, or even negated, after binding of the repressor, CREM. Because CREM is induced after T cell activation, its binding to the TCR ζ promoter may contribute to termination of the T cell response.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Grants RO1AI42269 and RO149954 (to G.C.T.) and Deutsche Forschungsgemeinschaft Grant TE 339/1-1. The opinions expressed in this study are those of the authors and do not represent those of the Department of Defense.
Abbreviations used in this paper: SLE, systemic lupus erythematosus; ChIP, chromatin immunoprecipitation analysis; CRE, cAMP response element; CREM, CRE modulator; hnRNP-D, heterogeneous nuclear ribonucleoprotein-D.