Patients with Systemic Lupus Erythematosus Activator Protein-1 ...

The Journal of Immunology

Cyclic Adenosine 5ⴕ-Monophosphate Response Element Modulator Is Responsible for the Decreased Expression of c-fos and Activator Protein-1 Binding in T Cells from Patients with Systemic Lupus Erythematosus Vasileios C. Kyttaris,*† Yuang-Taung Juang,* Klaus Tenbrock,*‡ Arthur Weinstein,† and George C. Tsokos1* T cells from patients with systemic lupus erythematosus express increased levels of the cAMP response element modulator (CREM) that has been shown to bind to the IL-2 promoter and suppress its activity. In this study, we demonstrate that CREM binds to the proximal promoter of the c-fos proto-oncogene in live systemic lupus erythematosus T cells and represses its expression following stimulation in vitro. Decreased levels of c-fos protein result in decreased AP-1 activity, as determined in shift assays. Blockade of the translation of CREM mRNA with an antisense CREM vector increases the expression of c-fos and the AP-1 activity. The levels of c-fos mRNA vary with disease activity. We conclude that CREM represses the expression of c-fos and limits the activity of the enhancer AP-1. Thus, CREM is involved indirectly in the modulation of transcriptional regulation of multiple genes including IL-2. The Journal of Immunology, 2004, 173: 3557–3563.

A

berrant T cell signaling is thought to be central in the pathogenesis of systemic lupus erythematosus (SLE),2 a complex autoimmune disease affecting multiple organs (1, 2). T cells from patients with SLE produce decreased amounts of IL-2, a critical molecule for their maturation, proliferation, and eventual activation-induced cell death (AICD) (3, 4). Accordingly, limited production of IL-2 by SLE T cells has been linked to increased susceptibility to infections (5) and persistence of autoreactive T and B cells (6) in SLE patients. Decreased transcription of the IL-2 gene upon activation of the T cell has been shown to be responsible for the decreased production of IL-2 by SLE T cells (7). Multiple transcription factors including NF-␬B, Oct, and CREB have been shown to bind to the proximal 300 bp of the IL-2 promoter (8). Initial studies from our group in SLE T cells showed a decrease in the activating transcription factor NF-␬B p65 subunit (9), and gene transfer of the p65 subunit into SLE T cells increased the activity of the IL-2 promoter (10). Furthermore, we have shown that the transcription repressor cAMP response element (CRE) modulator (CREM) binds the ⫺180 site of the IL-2 promoter in SLE T cells and directly inhibits IL-2 production in these cells (7). Transfection of *Department of Cellular Injury, Walter Reed Army Institute of Research, Silver Spring, MD 20910, and Department of Medicine, Uniformed Services University of the Health Sciences, Bethesda, MD 20814; †Department of Medicine, Section of Rheumatology, Washington Hospital Center, Washington, DC 20010; and ‡Department of Pediatrics, University of Muenster, Muenster, Germany Received for publication February 5, 2004. Accepted for publication June 21, 2004. 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. 1 Address correspondence and reprint requests to Dr. George C. Tsokos, Department of Cellular Injury, Walter Reed Army Institute of Research, Building 503, Room 1A32, 503 Robert Grant Avenue, Silver Spring, MD 20910-7500.

SLE T cells with a plasmid encoding antisense CREM limited the expression of CREM and resulted in increased production of IL-2 (11). These experiments lead us to propose that CREM is a key repressor of the IL-2 transcription in SLE T cells. CREM is suspected to bind to the promoters of at least 3000 genes, including the proto-oncogene c-fos. Indeed, CREM binds to the ⫺57 CRE site of the c-fos promoter and blocks the cAMPinduced expression of the c-fos proto-oncogene (12), a prototype of the early response genes. c-fos belongs to the AP-1 family of transcription factors that consists of a mixture of heterodimers and homodimers of jun (v-jun, c-jun, junB, junD) and fos (v-fos, c-fos, fosB, fra1, fra2) proteins (13, 14). Each protein contains a leucine zipper region that enables its dimerization with other members of the fos/jun family. The jun proteins can form homodimers among themselves, while fos proteins form only heterodimers with jun proteins. Both jun-jun and jun-fos dimers bind to the 12-O-tetradecanoate-13-acetate response element (TRE) that contains the TGACTCAA motif. The binding though of the jun-jun homodimers to DNA is weak compared with jun-fos heterodimers (15, 16). Soon after antigenic stimulation, jun and fos proteins are expressed, and subsequently, AP-1 (particularly the c-jun/c-fos dimer) binds to the IL-2 promoter. This in turn induces the expression of the IL-2 gene (8, 17). Because CREM is transcriptionally up-regulated in SLE T cells (11), we conducted experiments to investigate the effect of CREM on the expression of the c-fos gene and the binding of AP-1 in SLE T cells. We demonstrate that CREM indeed binds to the c-fos promoter in SLE T cells, and it is directly responsible for the decreased production of c-fos and decreased AP-1 binding.

Materials and Methods Study subjects

2

Abbreviations used in this paper: SLE, systemic lupus erythematosus; AICD, activation-induced cell death; CBP, CREB-binding protein; ChIP, chromatin immunoprecipitation analysis; CRE, cAMP response element; CREM, CRE modulator; GKLF, gut-enriched Kruppel-like factor; hnRNP, heterogeneous ribonucleoprotein; RA, rheumatoid arthritis; SLEDAI, SLE disease activity index; TRE, 12-O-tetradecanoate-13-acetate response element. Copyright © 2004 by The American Association of Immunologists, Inc.

Twenty-five female patients diagnosed with SLE according to the American College of Rheumatology criteria (18) donated blood for these studies. The mean age of the patients was 38 (21– 80) years old, and the mean SLE disease activity index (SLEDAI) was 3.7 (0 –12) (19). Eighty-four percent of the patients were taking oral prednisone at a mean dose of 8.3 mg 0022-1767/04/$02.00

3558

CREM DOWN-REGULATES c-fos AND AP-1 BINDING IN SLE T CELLS

(2.5–30). Other immunosuppressive medications that the patients were on included hydroxychloroquine (70.6% of the patients), azathioprine (14.7%), and mycophenalate mofetil (17.1%). One patient was receiving monthly i.v. cyclophosphamide, with the last dose given 13 days before the blood draw for the study. The institutional review boards of all involved institutions approved the study protocol, and informed consents were obtained from all of the study subjects. Prednisone was held for 24 h before the blood draw. Five of the study patients donated blood twice, and two patients three times each. Each patient was matched for age with a healthy female volunteer. Four patients with rheumatoid arthritis (RA) and matched controls were also analyzed as part of the study.

T lymphocyte isolation and stimulation Peripheral venous blood was obtained from each study subject, in heparinlithium tubes. The blood was incubated for 20 min with a tetrameric Ab mixture against CD14, CD16, CD19, CD56, and glyA that attaches non-T cells to erythrocytes. Ficoll containing Lymphoprep gradient (Nycomed, Oslo, Norway) was subsequently used to separate these complexes from T cells. Using flow cytometry, we established that the purified cells were ⬎98% positive for CD3. Where mentioned, the T cells were stimulated with either PMA (10 ng/ml) and calcium ionophore A23187 (0.5 ␮g/ml) or anti-CD3 Ab (10 ␮g/ml), CD28 Ab (2.5 ␮g/ml), and goat anti-mouse cross-linking Ab (25 ␮g/ml).

Antibodies Anti-CREM, anti-fos, anti-jun, gut-enriched Kruppel-like factor (GKLF), anti-CREB-binding protein (CBP), anti-E47, anti-rabbit HRP, and antigoat HRP Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-CREB Ab was purchased from Upstate Biotechnology (Lake Placid, NY). Rabbit anti-heterogeneous ribonucleoprotein (hnRNP) serum was used in Western blots, as previously described (20). Anti-CD3 Ab was purchased from Ortho-McNeil Pharmaceuticals (Raritan, NJ). The anti-CD28 Ab was purchased from BD Biosciences (Franklin Lakes, NJ), and the goat anti-mouse cross-link Ab was purchased from Sigma Genosys (The Woodlands, TX).

Protein purification and Western blotting Blood from eight patients and eight controls was used for cytoplasmic protein and from another eight patients and controls for nuclear protein extraction. To prepare cytoplasmic protein extracts, cells were treated on ice in 400 ␮l with a lysis buffer (10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA supplemented with freshly added 1 mM DTT, 0.5 mM PMSF, 2 mM aprotinin, 1 mM leupeptin, 10 mM NaF, and 2 mM Na3VO4) for 15 min. At the end of the incubation, Nonidet P-40 was added to the reaction mixture at a concentration of 0.6%. The reaction mixture was vortexed for 10 s and then subjected to centrifugation at 13,000 rpm for 15 s. The supernatant was saved as cytoplasmic extract. The pellet was resuspended in 40 ␮l of buffer (20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, 1 mM Na3VO4, 1 mM PMSF, 2 mM aprotinin, and 1 mM leupeptin) and then shaked for 15 min at 4°C. After centrifugation for 5 min at 13,000 rpm, the supernatant was stored as nuclear extract. We followed the manufacturer’s instructions (ECL; Amersham, Piscataway, NJ) for the Western blotting. The film was scanned, and the density of each band was calculated with QuantityOne software (Bio-Rad, Hercules, CA).

Chromatin immunoprecipitation analysis (ChIP) Five million T cells were used per investigated Ab. The cells were treated with formalin (1% final concentration) for 10 min, washed, lysed, and sonicated. The DNA-protein complexes were immunoprecipitated with a desired Ab and extracted by protein A/G-Sepharose beads (Santa Cruz Biotechnology). After several washing steps, the cross-link between DNA and protein was reversed at 65°C, followed by protein digestion with proteinase K, and the DNA was extracted (QiaAmp DNA extraction kit; Qiagen, Valencia, CA). The DNA was amplified with primers flanking the c-fos promoter, including the ⫺57 site. The sequences are: forward, 5⬘ATGCTCACGAGATTAGGACACG-3⬘; reverse, 5⬘-GCTGCAGATGCG GTTGGAGT-3⬘. DNA from 1 million cells was used for each PCR. PCR products were separated on a 1.0% agarose gel. Using QuantityOne software (Bio-Rad), the OD of each band was quantitated after background subtraction.

Preparation of mRNA, RT-PCR, and real-time PCR One million T cells were used for extracting RNA (RNA Easy Mini kit; Qiagen). The RNA was treated with DNase I (Qiagen) and quantitated; 250 ng of total RNA was used for cDNA synthesis by reverse transcription (RT-PCR kit; Promega, Madison, WI). Sigma Genosys synthesized the PCR primers. PCR beads were used for amplification (Pharmacia, Piscataway, NJ) PCR was conducted in a conventional thermocycler. Real-time PCR was conducted with a Cepheid Smart Thermocycler (Cepheid, Sunnyvale, CA) by adding SYBR green to the reaction mixture. Primers used for PCR were as follows: GAPDH forward, 5⬘-CAACTACATGGTT TACATGTTCC-3⬘; reverse, 5⬘-GGACTGTGGTCATGAGTCCT-3⬘. c-fos forward, 5⬘-CCGAAGGGAAAGGAATAAG-3⬘; reverse, 5⬘-CGGGG TAGGTGAAGACGAA-3⬘. PCR products were separated on a 1.0% agarose gel, and the OD was quantitated with the QuantityOne software (Bio-Rad) after background subtraction from each band. The real-time PCR products were quantitated, as previously described (21). Briefly, a cDNA specimen was used to create serially diluted samples that were subsequently amplified using different primers. A standard curve was generated for each set of primers, by plotting the threshold cycles of the real-time PCR against the logarithm of the relative concentrations of the diluted samples. Thereafter, the relative mRNA copies of the experimental samples were calculated after real-time amplification, using the linear regression of the standard curve.

Transfection of Jurkat cells Jurkat cells were cultured in RPMI 1640 and 10% FBS without stimulation. Plasmids encoding CREM␣ antisense (a kind gift from P. Sassone-Corsi, Laboratoire de Ge´ ne´ tique Moleculaire des Eucaryotes, Institute National de la Sante´ et de la Recherche Me´ dical, Strasbourg, France) and corresponding empty vector plasmid were used for transfection. Five micrograms of each plasmid were used per transfection. Approximately 5–10 ⫻ 106 T cells were transfected by electroporation with an AMAXA electroporator (AMAXA, Cologne, Germany).

Statistical analysis The paired two-tailed Student’s t test and the Pearson product moment correlation coefficient were used for statistical analysis. Statistical significance was defined as p ⬍ 0.05.

Results CREM binding to the c-fos promoter is increased in SLE T cells

Preparation of nuclear extracts and EMSAs Five million T cells were used for preparation of nuclear extracts, as described above, from 10 patients and 10 controls. Nuclear extracts (1 ␮g) were incubated with a radiolabeled dsDNA probe, 1 ␮g of poly(dI/dC) in the binding buffer for 15 min at room temperature. The reaction mixture was then subjected to separation in 6% nondenaturing gel (Invitrogen Life Technologies, Carlsbad, CA). The dried gel was then autoradiographed. For supershift assays, the nuclear proteins were incubated with specific Abs at 4°C for 10 min before the probe, poly(dI/dC), and binding buffer were added. The reaction was further conducted for another 15 min at room temperature. The sequences of the probes used are as follows: c-fos promoter containing the ⫺57 site: forward, 5⬘-GAGCCCGTGACGTTTA CACTCA-3⬘; reverse, 5⬘-TGAGTGTAAACGTCACGGGCTC-3⬘. AP-1 consensus-binding sequence: forward, 5⬘-CGCTTGATGACTCAGCCGGAA-3⬘; reverse, 5⬘-TTCCGGCTGAGTCATCAAGCG-3⬘. Elf-1 binding site at the TCR ␨-chain promoter: forward, 5⬘-TCGAGAACCTCCAGGGCTTCCT GCCTGTGAACCA-3⬘; reverse, 5⬘-TGGTTCACAGGCAGGAAGCCC TGGAGGTTCTCGA-3⬘.

Antigenic stimulation of T cells leads to cAMP-dependent protein kinase A-mediated phosphorylation of the CREB that in turn induces the expression of the c-fos gene. The site that CREB binds to the c-fos promoter is a CRE centered on the ⫺57 nt. The same CRE site is the target of CREM binding to the promoter, as shown in JEG-3 cells (12). To determine whether CREM that is overexpressed in unstimulated SLE T cells binds to the c-fos promoter in live SLE T cells, we performed ChIP assays using anti-CREM Ab (Fig. 1). Eight of 10 patients with SLE had increased CREM binding to the c-fos promoter (mean relative density of immunoprecipitated DNA levels as assessed by PCR: SLE, 2046 ⫾ 564; control, 801.8 ⫾ 207.2; p ⫽ 0.05). As control, ChIP using an unrelated anti-E47 Ab showed no binding to the c-fos promoter. These results demonstrate that CREM binding to the c-fos promoter is significantly


FIGURE 1. Increased CREM binding to the c-fos promoter in live SLE T cells. T cells from SLE patients and controls were formalin fixed, lysed, washed, and sonicated, according to the ChIP protocol outlined in Materials and Methods. The anti-CREM Ab immunoprecipitates were subjected to PCR using primers specific for the c-fos promoter. L stands for SLE, and N for control specimens. a, Representative ChIP experiments using SLE and normal T cells. b, The density of the bands was measured by densitometry, the background was subtracted, and the means and SEM were calculated from 10 patients and their controls. c, T cells from an SLE patient were treated according to the ChIP protocol outlined in Materials and Methods. Anti-CREM and anti-E47 immunoprecipitates were subjected to PCR using the same as above c-fos promoter-specific primers. d, T cells from SLE patients (L) and controls (N) were formalin fixed, lysed, washed, and sonicated, according to the ChIP protocol outlined in Materials and Methods. The anti-CREB Ab immunoprecipitates were subjected to PCR using primers specific for the c-fos promoter.

increased in live unstimulated SLE compared with control T cells. Furthermore, we performed ChIP with anti-CREB Ab and found no specific pattern difference between SLE and control T cells in CREB binding to the c-fos promoter (Fig. 1d). To ascertain that CREM indeed binds to the ⫺57 site of the c-fos promoter in T cells, we performed EMSA using a 32P-labeled oligonucleotide containing the ⫺57 CRE sequence and nuclear protein extracted from unstimulated SLE T cells (Fig. 2a). The EMSA indeed showed a band that can be shifted with anti-CREM Ab, but not an unrelated anti-CBP Ab. Supershift was also seen with anti-CREB Ab, indicating that CREB and CREM can bind to the c-fos ⫺57 site as a heterodimer. Similar results were seen with extracts from control T cells (Fig. 2b).

3559

FIGURE 2. CREM and CREB bind to the ⫺57 site of the c-fos promoter in T cells. a, T cell nuclear protein extracts from an SLE patient were incubated with a radiolabeled oligonucleotide spanning the ⫺57 CRE site of the c-fos promoter in the absence or presence of anti-CREM, antiCREB, or anti-CBP Ab, as described in Materials and Methods. b, T cell nuclear protein extracts from an SLE patient (L) and control (N) were incubated with a radiolabeled oligonucleotide spanning the ⫺57 CRE site of the c-fos promoter in the absence or presence of anti-CREM or antiCREB, as described in Materials and Methods.

PMA and calcium ionophore. GAPDH mRNA levels were determined also, and the ratio of c-fos to GAPDH mRNA was compared between SLE and control T cells. As shown in Fig. 4, c-fos mRNA production in SLE T cells is markedly decreased when compared with controls (mean ratio of c-fos-GAPDH mRNA: SLE, 0.30 ⫾ 0.09; control, 0.54 ⫾ 0.11; p ⫽ 0.02). We also analyzed mRNA levels in unstimulated T cells from SLE patients and controls (data not shown). Because the mRNA levels were very low, it was impossible to reach a conclusion regarding true differences between SLE T cells and controls, although there was a trend for higher basal c-fos mRNA levels in controls (basal c-fos

c-fos up-regulation after stimulation in T cells is impaired in SLE patients To determine whether the increased CREM binding to the c-fos promoter is associated with impaired c-fos production after stimulation, we incubated T cells from SLE patients and control individuals with PMA and calcium ionophore. Real-time RT-PCR was used to assess c-fos mRNA levels, and Western blot to assess c-fos protein levels. Kinetic studies (Fig. 3a) show that c-fos mRNA in both SLE and control T cells is up-regulated within 10 min, reaching its peak at 30 min, and returns to baseline levels at 1 h after stimulation. Similarly, Fig. 3b shows that following cell stimulation c-fos protein production is quickly up-regulated, reaching a plateau at 1 h. These experiments show that the kinetics of c-fos production upon stimulation in our system is similar to previously published data. Using RT-PCR, we then calculated the peak levels of c-fos mRNA in SLE and control T cells 30 min after stimulation with

FIGURE 3. Similar induction kinetics of c-fos mRNA and protein in normal and SLE T cells. a, T cells from an SLE patient (L9) and a matched control (N30) were stimulated with PMA and calcium ionophore for different time points. The mRNA was isolated, reverse transcribed, and subjected to real-time PCR using specific primers for the c-fos gene, as described in Materials and Methods. The relative number of mRNA copies was calculated, as described in Materials and Methods. b, Primary T cells from a healthy volunteer were stimulated with PMA and calcium ionophore for different time points. The nuclear protein was isolated and used for Western blot analysis, as described in Materials and Methods.

3560


FIGURE 4. c-fos mRNA following stimulation is decreased in SLE T cells compared with controls. T cells from SLE patients (L) and matched controls (N) were stimulated with PMA and calcium ionophore for 30 min. The mRNA was isolated, reverse transcribed, and subjected to PCR using specific primers for the c-fos and the GAPDH genes, as described in Materials and Methods. a, Representative experiments showing the c-fos and GAPDH mRNA levels as measured by RT-PCR in SLE and control T cells. The density of the bands was measured by densitometry, and the background was subtracted; subsequently, the ratio of c-fos-GAPDH mRNA was calculated. b, Means and SEM of c-fos-GAPDH mRNA from 13 patients and controls are shown here. The results are statistically significant (p ⫽ 0.02).

mRNA copies as assessed by real-time PCR: SLE, 17.9 ⫾ 15.5; control, 48.7 ⫾ 35.7; p ⫽ 0.16). Next, we determined the levels of c-fos protein in nuclear extracts from SLE and control T cells stimulated with PMA and calcium ionophore for 60 min using Western blotting. hnRNP nuclear protein levels were used as control, and the ratio of c-fos to hnRNP was compared between SLE and control T cells. As shown in Fig. 5a, SLE T cells had significantly decreased c-fos protein levels in their nucleus (ratio of c-fos-hnRNP nuclear protein levels: SLE, 1.37 ⫾ 0.25; control, 1.87 ⫾ 0.35; p ⫽ 0.04). Along the same lines, when we used anti-CD3/CD28 and anti-mouse cross-linking Ab to stimulate the cells, T cells from three SLE patients had decreased c-fos protein levels in their nucleus relative to controls (one of three representative experiments in T cells from the three SLE/control pairs is shown in Fig. 5a). Similar results were observed when we compared the cytoplasmic levels of c-fos protein in SLE and control T cells (Fig. 5c). ␤ actin protein levels were used as control, and the c-fos to ␤ actin ratio was calculated. Fig. 5d depicts the significant decrease in the ratio of c-fos to ␤ actin protein levels in the cytoplasm of SLE when compared with control T cells (mean ratio of c-fos-actin cytoplasmic protein levels: SLE, 3.52 ⫾ 0.44; control, 5.94 ⫾ 0.47; p ⫽ 0.01). Taken together, these results demonstrate that following activation of the T cell, the expected early up-regulation of c-fos expression is significantly impaired in SLE. On the contrary, c-fos protein levels in T cells from four RA patients did not significantly differ from the controls, as shown in Fig. 5e (mean ratio of c-fos-actin cytoplasmic protein levels: RA, 2.31 ⫾ 0.39; control, 1.68 ⫾ 0.34; p ⫽ 0.13; mean ratio of c-fos-actin nuclear protein levels: RA, 12.08 ⫾ 3.93; control, 12.79 ⫾ 5.09; p ⫽ 0.92). AP-1 binding in SLE T cells is lower than controls As mentioned above, early after T cell stimulation, the c-jun/c-fos dimer forms and binds strongly to the TRE motif containing AP-1

FIGURE 5. c-fos protein levels following stimulation are decreased in SLE T cells compared with controls. T cells from SLE patients (L), RA (R), and matched controls (N) were stimulated with PMA and calcium ionophore or anti-CD3/anti-CD28 and goat anti-mouse cross-linking Ab for 60 min. The nuclear and cytoplasmic proteins were isolated and used for Western blot analysis, as described in Materials and Methods. a, Representative experiment of PMA and calcium ionophore-treated T cell nuclear protein extracts from an SLE patient (L13) and a matched control (N14) that were immunoblotted using anti-c-fos and anti-hnRNP Ab. The intensity of the bands was measured by densitometry, and the background was subtracted; subsequently, the ratio of c-fos-hnRNP nuclear protein was calculated. A representative experiment of anti-CD3/anti-CD28 and goat anti-mouse cross-linking Ab-treated T cell nuclear protein extracts from an SLE patient (L15) and a matched control (N11) that were immunoblotted using anti-c-fos and anti-hnRNP Ab is also shown in the same panel. b, Means and SEM of c-fos-hnRNP nuclear protein from eight patients and controls are shown here. The T cells were stimulated with PMA and calcium ionophore, as described in Materials and Methods. The results are statistically significant (p ⫽ 0.04). c, Representative experiment of T cell cytoplasmic protein extracts from an SLE patient (L2) and a matched control (N5) that were immunoblotted using anti-c-fos and anti-␤ actin Ab. The density of the bands was measured by densitometry, and the background was subtracted; subsequently, the ratio of c-fos-␤ actin nuclear protein was calculated. d, Means and SEM of c-fos-␤ actin nuclear protein from eight patients and controls are shown here. The results are statistically significant (p ⫽ 0.01). e, Representative experiment of T cell cytoplasmic protein from RA patients (R) and matched controls (N) that were immunoblotted with anti-c-fos and anti-␤ actin Ab. The ratios of c-fos-␤ actin are also shown under each specimen’s blot.

binding sites. Because c-fos expression upon stimulation was found to be deficient in SLE T cells, we hypothesized that this would result in decreased overall AP-1 transcription factor binding. Accordingly, we constructed an oligonucleotide containing the consensus AP-1-binding TRE motif. EMSA using this oligonucleotide (Fig. 6a) and nuclear protein extracts showed a band that was shifted with both anti-c-jun and anti-c-fos Abs, but not an unrelated anti-GKLF Ab. This experiment demonstrates that our construct binds specifically to the c-fos/c-jun heterodimer, the main AP-1 complex. Fig. 6, b and c, shows the results of EMSA using this AP-1 consensus oligonucleotide labeled with 32P and nuclear extracts from SLE and control T cells that have been stimulated for 1 h with PMA and calcium ionophore. The SLE T cells had a significantly lower poststimulation AP-1 binding when compared with controls (mean AP-1 binding as assessed by the EMSA


FIGURE 6. AP-1 binding is decreased in stimulated SLE compared with control T cells. a, T cell nuclear protein extracts from an SLE patient were incubated with a radiolabeled oligonucleotide encoding two TRE sites (AP-1 binding sites) in the absence or presence of anti-c-fos, anti-cjun, or anti-GKLF Abs, as described in Materials and Methods. The same was performed in the presence of the same oligonucleotide in excess as well as the presence of an unrelated elf-1-binding oligonucleotide in excess. b, T cell nuclear protein extracts from SLE patients (L), and controls (N) were incubated with the same radiolabeled oligonucleotide encoding two TRE sites, as described in Materials and Methods. A representative experiment is shown. The density of the bands was measured by densitometry, and the background was subtracted. c, Means and SEM of AP-1 binding from 10 SLE patients and 10 controls are shown. The results are statistically significant (p ⫽ 0.03).

band density: SLE, 2168 ⫾ 430.7; control, 3590 ⫾ 421; p ⫽ 0.03). AP-1 binding in T cells from patients with RA showed no clear difference from controls (mean AP-1 binding as assessed by the EMSA band density: RA, 1570 ⫾ 644; control, 1721 ⫾ 262; p ⫽ 0.8). Thus, the decreased expression of c-fos in stimulated SLE T cells results in decreased AP-1 binding. Antisense CREM up-regulates c-fos mRNA and AP-1 binding To determine whether the increased binding of CREM to the c-fos promoter and the decreased production of c-fos and AP-1 binding upon stimulation are directly related, we transfected Jurkat cells with either a control plasmid or a plasmid encoding antisense CREM␣. Fig. 7 shows an increase in the c-fos mRNA level and the

FIGURE 7. Jurkat cells were transfected with plasmids encoding antisense CREM or empty vector. The cells were cultured for 24 h and then subjected to stimulation with PMA and calcium ionophore, as described in Materials and Methods. a, Jurkat cells transfected with control plasmid or antisense CREM␣ were stimulated with PMA and calcium ionophore for 30 min. The mRNA was isolated, reverse transcribed, and subjected to PCR using specific primers for the c-fos and the GAPDH genes, as described in Materials and Methods. The ratios of c-fos-GAPDH mRNA are shown under each specimen’s blot. b, T cell nuclear protein extracts from Jurkat cells transfected with either control plasmid or antisense CREM␣ and stimulated with PMA and calcium ionophore for 60 min were incubated with a radiolabeled oligonucleotide encoding two TRE sites (AP-1 binding sites), as described in Materials and Methods. The density of the bands was measured by densitometry, and the background was subtracted.

3561

FIGURE 8. Stimulation-induced c-fos mRNA levels in SLE T cells correlate with disease activity. T cells from SLE patients were stimulated with PMA and calcium ionophore. The mRNA was isolated, reverse transcribed, and subjected to PCR using specific primers for the c-fos and the GAPDH genes, as described in Materials and Methods. The density of the bands was measured by densitometry, and the background was subtracted; subsequently, the ratio of c-fos-GAPDH mRNA was calculated. Means and SEM of c-fos-GAPDH mRNA ratios from 13 controls, 7 nonactive (SLEDAI: 0 –3), and 6 active (SLEDAI ⬎3) patients with SLE are shown here. The results are statistically significant (p ⫽ 0.03).

AP-1 binding in cells that were transfected with the antisense CREM␣-encoding plasmid compared with cells transfected with a control vector and stimulated with PMA and calcium ionophore (ratio of c-fos-GAPDH mRNA: transfected with antisense CREM, 0.70; control, 0.51; AP-1 binding as assessed by the EMSA band density: transfected with antisense CREM, 3212; control, 1149). Because antisense CREM␣ causes a decrease in the levels of CREM␣ protein (11), these data indicate that CREM␣ is involved directly in the transcriptional repression of c-fos gene that results in decreased c-fos protein and AP-1 binding. c-fos mRNA is decreased in active SLE patients Because SLE patients in this cohort were at different stages of their disease, we evaluated the effect of disease activity on the expression of c-fos. Fig. 8 shows that the poststimulation c-fos mRNA levels are significantly decreased in active SLE patients (patients with SLEDAI of 4 or more) compared with both inactive patients (SLEDAI ⬍4) and controls (mean ratio of c-fos-GAPDH mRNA levels: control, 55.39 ⫾ 12.99 (n ⫽ 13); inactive SLE patients, 44.16 ⫾ 15.50 (n ⫽ 7); active SLE patients, 13.35 ⫾.4.98 (n ⫽ 6); p ⫽ 0.03). Furthermore, a decrease in the SLEDAI was observed in two patients who were studied twice over time, and this was associated with an increase in the levels of c-fos mRNA. The first patient (L13) had a decrease of SLEDAI from 8 to 4 with a concomitant increase in c-fos-GAPDH mRNA ratio from 3.95 to 35.85, and the second patient (L10) had a decrease of SLEDAI from 2 to 0 with a concomitant increase in c-fos-GAPDH mRNA ratio from 17.43 to 89.16. On the contrary, there was no association between the levels of c-fos mRNA and the type and dose of medications that the patients were receiving at the time of the study. Specifically, the prednisone dose did not correlate with the ratio of c-fos-GAPDH mRNA levels (Pearson product moment correlation coefficient: ⫺0.271, p ⫽ 0.37) or AP-1-binding activity (Pearson product moment correlation coefficient: ⫺0.373, p ⫽ 0.288).

Discussion The present study shows that CREM binding to the c-fos promoter, in unstimulated live T cells from SLE patients, is significantly increased. Increased binding results in decreased transcription of c-fos gene, decreased production of c-fos protein, and AP-1 binding in the nuclei of SLE T cells. Of importance is the finding that

3562


interference with the translation of the c-fos mRNA using antisense CREM vector resulted in increased c-fos protein levels and AP-1 binding. Besides reaffirming that CREM can suppress the transcription of c-fos gene in human T cells, as is the case with JEG-3 cells (12), our experiments demonstrate a mechanism whereby CREM contributes to decreased binding of an important transcriptional enhancer AP-1. This decreased induction of c-fos was mainly observed in active SLE patients, underscoring the fact that this is probably a secondary phenomenon directly related to the activity of the disease. Klinman et al. (22) have shown decreased c-fos mRNA in PBMC from SLE patients 90 min after mitogenic stimulation. Other investigators have looked also at the levels of c-fos mRNA in unstimulated T cells from patients with SLE. Although the spontaneous expression of c-fos was lower in bone marrow extracts from patients with SLE when compared with controls (23), there was no difference in the levels of c-fos mRNA in unstimulated peripheral blood mononuclear cells (24). In contrast to SLE patients, patients with RA have significantly higher expression of c-fos (25, 26) and AP-1 (27) than normal individuals. Multiple signaling abnormalities have been recognized in SLE. Upon Ag binding, the TCR/CD3 complex triggers a rapid rise in intracellular calcium (2). This rise in calcium is significantly higher and accelerated in SLE T cells as compared with controls (28). The substitution of TCR ␨-chain by the FcR␥ chain in SLE T cells (29) contributes to the increased calcium response in SLE T cells. In T cells, calcium acts as a second messenger, leading to the activation of NF-AT-responsive genes, including the IL-2 gene. But, despite the fact that the SLE T cells show this particularly robust calcium response, they have a phenotype similar to anergic cells with deficient in vivo response to new and recall Ags as well as in vitro responses to recall Ags and mitogens (28) and deficient production of IL-2 upon activation (4). It is well recognized now that a second stimulus initiated by the CD28-B7 interaction and resulting in the activation of the transcription factors AP-1, CREB, and NF-␬B is essential for the induction of IL-2 and the propagation of the immune response. In SLE T cells, the decrease of NF-␬B and the increase of CREM are responsible for the decreased expression of IL-2. CREM further affects a second stimulatory pathway by blocking the expression of c-fos and the activation of AP-1, as shown in this study. The imbalance between an overactive Ca2⫹-mediated signaling and impaired AP-1 activation has broader consequences. As mentioned earlier, Ca2⫹ influx in the T cell results in the activation of NF-AT. AP-1 interacts with NF-AT, forming complexes that upregulate several cytokine gene promoters, including IL-2, IL-3, IL-4, GM-CSF, Fas ligand, and MIP1␣ (30). The up-regulation of such genes is important for the productive immune response. In the absence or decrease of AP-1 (particularly the inducible c-fos-jun dimer), the NF-AT:AP-1 complexes are not formed. Thus, NF-AT can only up-regulate genes such as TNF-␣ and IL-13 (30, 31) that do not require the NF-AT:AP-1 complex to form. Along the same lines, T cells that have activated NF-AT while lacking sufficient AP-1 do not undergo AICD as readily (30). Thus, the activation of NF-AT with impaired AP-1 activation results in a nonproductive immune response. In the SLE T cell paradigm, a relative imbalance between NF-AT and AP-1 activation can explain the dichotomy between high calcium responses and deficient production of IL-2. It can also help explain the in vivo and in vitro abnormal T cell responses to Ags (28) and the deficient AICD (3, 32). In the same context, Th2 cell cytokine production would be skewed toward IL-10, a cytokine that has been found to be elevated in SLE patients (33). Interestingly, IL-10 is associated with in vivo spontaneous apo-

ptosis of T cells (34), a mechanism that can explain the lymphopenia seen in SLE patients. It has also been proposed that this increased lymphocytic apoptosis can flood the circulation with nuclear and cytoplasmic autoantigens (28, 34), leading to perpetuation of the abnormal immune response. The study of the molecular aberrations in SLE T cells has suggested approaches to correct the production of IL-2. We have shown that replenishing the TCR ␨-chain of the TCR (35) in SLE T cells results not only in normalization of the calcium responses, but also in the restoration of IL-2 production. Also, replenishment of the missing NF-␬B p65 subunit in SLE T cells (10) results in increased IL-2 production. Along the same lines, the elimination of CREM by introducing an antisense CREM plasmid in SLE T cells (11) resulted in the increase of IL-2 production. In conclusion, this study has shown that the increased CREM expression in SLE T cells can limit the expression of c-fos that is a key component of the transcriptional enhancer AP-1. Therefore, CREM contributes to the decreased expression of IL-2 by both binding directly to the IL-2 promoter (7) and limiting the AP-1 binding by binding to the c-fos promoter.

References 1. Tsokos, G. C., and D. T. Boumpas. 2001. Systemic lupus erythematosus. In Molecular Pathology of Autoimmunity. C. Bona, ed. Gordon and Breach, Harwood Academic Publishers, New York, p. 255. 2. Tsokos, G. C., and G. M. Kammer. 2000. Molecular aberrations in human lupus. Mol. Med. Today 6:418. 3. Refaeli, Y., L. Van Parijs, C. A. London, J. Tschopp, and A. K. Abbas. 1998. Biochemical mechanisms of IL-2 regulated Fas-mediated T cell apoptosis. Immunity 8:615. 4. Linker-Israeli, M., A. C. Bakke, R. C. Kitridou, S. Gendler, S. Gillis, and D. A. Horwitz. 1983. Defective production of interleukin 1 and interleukin 2 in patients with systemic lupus erythematosus (SLE). J. Immunol. 130:2651. 5. Iliopoulos, A. G., and G. C. Tsokos. 1996. Immunopathogenesis and spectrum of infections in systemic lupus erythematosus. Semin. Arthritis Rheum. 25:318. 6. Tsokos, G. C. 1999. Overview of cellular immune function in systemic lupus erythematosus. In Systemic Lupus Erythematosus. R. G. Lahita, ed. Academic Press, New York, p. 17. 7. Solomou, E. E., Y.-T. Juang, M. F. Gourley, G. M. Kammer, and G. C. Tsokos. 2001. Molecular basis of deficient IL-2 production in T cells from patients with systemic lupus erythematosus. J. Immunol. 166:4216. 8. Jain, J., C. Loh, and A. Rao. 1995. Transcriptional regulation of the IL-2 gene. Curr. Opin. Immunol. 7:333. 9. Wong, H. K., G. M. Kammer, G. Dennis, and G. C. Tsokos. 1999. Abnormal NF-␬B activity in T lymphocytes from patients with systemic lupus erythematosus is associated with decreased p65-RelA protein expression. J. Immunol. 163:1682. 10. Herndon, T. M., Y.-T. Juang, E. Solomou, S. W. Rothwell, W. R. Gilliland, and G. C. Tsokos. 2002. Electroporation-mediated gene transfer of p65 into T lymphocytes from systemic lupus erythematosus patients leads to increased levels of interleukin-2 promoter activity. Clin. Immunol. 103:145. 11. Tenbrock, K., Y.-T. Juang, M. Gourley, M. Nambiar, and G. C. Tsokos. 2002. Antisense cyclic adenosine 5⬘-monophosphate response element modulator upregulates IL-2 in T cells from patients with systemic lupus erythematosus. J. Immunol. 169:4147. 12. Foulkes, N., B. Laoide, F. Schlotter, and P. Sassone-Corsi. 1991. Transcriptional antagonist cAMP-responsive element modulator (CREM) down-regulates c-fos cAMP-induced expression. Proc. Natl. Acad. Sci. USA 88:5448. 13. Wong, H. K. 2000. Transcription factors. In Principles of Molecular Rheumatology. G. C. Tsokos, ed. Humana Press, Totowa, p. 118. 14. Karin, M., Z.-G. Liu, and E. Zandi. 1997. AP-1 function and regulation. Curr. Opin. Cell Biol. 9:240. 15. Nicklin, M. J., and G. Casari. 1991. A single site mutation in a truncated Fos protein allows it to interact with the TRE in vitro. Oncogene 6:173. 16. Nakabeppu, Y., K. Ryder, and D. Nathans. 1988. DNA binding activities of three murine Jun proteins: stimulation by Fos. Cell 55:907. 17. Jain, J., V. E. Valge-Archer, and A. Rao. 1992. Analysis of the AP-1 sites in the IL-2 promoter. J. Immunol. 148:1240. 18. Hochberg, M. 1997. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 40:1725. 19. Bombardier, C., D. D. Gladman, M. B. Urowitz, D. Caron, and C. H. Chang. 1992. Derivation of the SLEDAI. Arthritis Rheum. 6:630. 20. Tolnay, M., L. Baranyi, and G. C. Tsokos. 2000. Heterogeneous nuclear ribonucleoprotein D0 contains transactivator and DNA-binding domains. Biochem. J. 348:151. 21. Bustin, S. A. 2000. Absolute quantification using real-time reverse transcription polymerase chain reaction assays. J. Mol. Endocrinol. 25:169.

The Journal of Immunology 22. Klinman, D. M., J. F. Mushinski, M. Honda, Y. Ishigatsubo, J. D. Mountz, E. S. Ravenche, and A. D. Steinberg. 1986. Oncogene expression in autoimmune and normal peripheral blood mononuclear cells. J. Exp. Med. 163:1292. 23. Alvarado-de la Barrera, C., J. Alcocer-Varela, Y. Richaud-Patin, D. Alarcon-Segovia, and L. Llorente. 1998. Differential oncogene and TNF-␣ mRNA expression in bone marrow cells from systemic lupus erythematosus patients. Scand. J. Immunol. 48:551. 24. Deguchi, Y., H. Hara, S. Negoro, T. Kakunaga, and S. Kishimoto. 1987. Protooncogene expression in peripheral blood mononuclear cells from patients with systemic lupus erythematosus as an indicator of the disease activity. Clin. Immunol. Immunopathol. 45:424. 25. Trabandt, A., W. K. Aicher, R. E. Gay, V. P. Sukhatme, H. G. Fassbender, and S. Gay. 1992. Spontaneous expression of immediately-early response genes c-fos and egr-1 in collagenase producing rheumatoid synovial fibroblasts. Rheumatol. Int. 12:53. 26. Hikasa, M., E. Yamamoto, H. Kawasaki, K. Komai, K. Shiozawa, A. Hashiramoto, Y. Miura, and S. Shiozawa. 2003. p21waf1/cip1 is down-regulated in conjuction with up-regulation of c-fos in the lymphocytes of rheumatoid arthritis patients. Biochem. Biophys. Res. Commun. 304:143. 27. Han, Z., D. L. Boyle, A. M. Manning, and G. S. Firestein. 1998. AP-1 and NF-␬B regulation in rheumatoid arthritis and murine collagen-induced arthritis. Autoimmunity 28:197. 28. Tsokos, G. C., and S. N. Liossis. 1999. Immune cell signaling defects in lupus: activation, anergy and death. Immunol. Today 20:119.

3563 29. Krishnan, S., V. G. Warke, M. P. Nambiar, G. C. Tsokos, and D. L. Farber. 2003. The FcR␥ subunit and Syk kinase replace the CD3 ␨-chain and ZAP-70 kinase in the TCR signaling complex of human effector CD4 T cells. J. Immunol. 170:4189. 30. Macian, F., C. Garcia-Rodriguez, and A. Rao. 2000. Gene expression elicited by NFAT in the presence or absence of cooperative recruitment of Fos and Jun. EMBO J. 19:4783. 31. Macian, F., F. Garcia-Cozar, S. H. Im, H. F. Horton, M. C. Byrne, and A. Rao. 2002. Transcriptional mechanisms underlying lymphocyte tolerance. Cell 109:719. 32. Kovacs, B., D. Vassilopoulos, S. A. Vogelgesang, and G. C. Tsokos. 1996. Defective CD3-mediated cell death in activated T cells from patients with systemic lupus erythematosus: role of decreased intracellular TNF-␣. Clin. Immunol. Immunopathol. 81:293. 33. Houssiau, F. A., C. Lefebvre, M. Vanden Berghe, M. Lambert, J. P. Devogelaer, and J. C. Renauld. 1995. Serum interleukin 10 titers in systemic lupus erythematosus reflect disease activity. Lupus 4:393. 34. Georgescu, L., R. K. Vakkalanka, K. B. Elkon, and M. K. Crow. 1997. Interleukin-10 promotes activation-induced cell death of SLE lymphocytes mediated by Fas ligand. J. Clin. Invest. 100:2622. 35. Nambiar, M. P., C. U. Fisher, V. G. Warke, S. Krishnan, J. P. Mitchell, N. Delaney, and G. C. Tsokos. 2003. Reconstitution of deficient T cell receptor ␨ chain restores T cell signaling and augments T cell receptor/CD3-induced interleukin-2 production in patients with systemic lupus erythematosus. Arthritis Rheum. 48:1948.