Activation of Nuclear Orphan Receptor NURR1 Transcription by NF-κB and Cyclic Adenosine 5′-Monophosphate Response Element-Binding Protein in Rheumatoid Arthritis Synovial Tissue

Tissue collection

All synovial biopsies were obtained from the knee by arthroscopy following informed consent from patients diagnosed with RA (34). All patients attended the Early Arthritis Clinic at St. Vincent’s University Hospital (Dublin, Ireland). At the time of biopsy, all patients (n = 10) had active synovitis of recent onset (median 4 mo; range 0–12 mo). Patients may have been on nonsteroidal antiinflammatory medication but had received no disease-modifying agents, including prednisone, before the time of biopsy. Histologically normal synovium (n = 2) was obtained from patients undergoing lower limb amputation.

Synoviocyte cell culture

Synovial tissue obtained from the knee by arthroscopy was treated for 4 h with 1 mg/ml collagenase (type I; Worthington Biochemical, Freehold, NJ) in RPMI 1640 at 37°C in 5% CO₂. Dissociated cells were plated in RPMI 1640 supplemented with 10% FCS (Life Technologies, Paisley, U.K.), penicillin (100 U/ml), and streptomycin (100 U/ml) (31). To eliminate nonadherent mononuclear cells from synovial cell preparations, the plated cells were cultured for at least 24 h. The cells were then washed extensively with PBS. Primary RA synoviocytes were used between the third and seventh passage for subsequent experiments (35). Primary synoviocyte cells were found to be morphologically homogeneous fibroblast-like cells and did not react with Abs to the macrophage/monocyte Ag CD14. The immortalized normal human K41 M synoviocyte cell line was grown, as described previously (36). All synoviocytes were grown in serum-free medium for 24 h before stimulation. Recombinant human TNF-α, IL-1β (Calbiochem, Darmstadt, Germany), TGF-β, platelet-derived growth factor (PDGF), basic fibroblast factor (bFGF) (R&D Systems, Minneapolis, MN), forskolin, and PGE₂ (Sigma-Aldrich, Dorset, U.K.) were included, as indicated.

Transient transfection

Twenty-four hours before transfection, 1 × 10⁵ synoviocyte cells were plated in 6 × 6-cm² dishes and allowed to attach. Endotoxin-free DNA, 0.1 μg CMV-β-galactosidase (Promega, Madison, WI), or 0.1 μg −1329/+132 or −396/+373 NURR1-β-galactosidase was added to 35 μl Effectine reagent (Qiagen, Cambridge, U.K.) in a volume of 900 μl RPMI 1640 (serum free) and incubated at room temperature for 30 min. RPMI 1640 (0.7 ml) was added to the transfection mixture, transferred onto the cells, and incubated for 6 h at 37°C in 5% CO₂. An equal volume of medium was added and the cells were incubated overnight at 37°C in 5% CO₂. The DNA-Effectine-containing medium were replaced, and the cells were left untreated or treated with 10 ng/ml IL-1β, TNF-α, 1 μM PGE₂, or 25 μM forskolin, and incubated for an additional 24 h. Protein concentrations were determined by the Bradford assay (Bio-Rad, Richmond, CA). For β-galactosidase staining of cells growing in monolayer, cells were washed with cold PBS, fixed with cold 0.5% glutaraldehyde, and washed twice with PBS before incubation with staining solution (1 M MgCL₂, 5 M NaCl, 0.5 M HEPES (pH 7.3), 30 mM potassium ferricyanide, 30 mM potassium ferrocyanide, and 2% 5-bromo-4-chloro-3-indoyl-β-galactopyranoside) for 12 h at 37°C (31). β-Galactosidase levels in cell extracts were measured using 30 μl cell extract incubated for 1–4 h in 0.1 M sodium phosphate containing 2.5 mg/ml O-nitrophenyl-B-d-galactopyranoside, 0.3 mM MgCl₂, and 15 mM 2-ME. The reaction was stopped by the addition of 1 M Na₂CO₃, and the OD of each reaction was read at 410 nm.

Northern blot analysis

Cell cultures grown in 25-cm² tissue culture dishes were maintained in serum-free RPMI 1640 for 24 h before treatment. Total RNA was isolated using RNeasy (Qiagen) at specific times after treatment. RNA was quantitated by UV absorption, and 10 μg total RNA was electrophoresed on a standard Northern gel and transferred to nylon membrane (Bio-Rad). NURR1, NUR77, and NOR-1 cDNA probes, spanning the amino-terminal region to avoid cross-hybridization, were radiolabeled to a high sp. act. using [α-³²P]dCTP and a random primer labeling system (Promega) (21, 31). Blots were exposed to film at −80°C using intensifying screens, and autoradiographic intensity was quantitated using an imaging densitometer.

Immunohistochemistry

Synovial tissue sections (7 μm) were placed on glass slides coated with 2% aminopropyl-triethoxy-silane. Synoviocytes grown on apyrogenic glass coverslips were treated with methanol for 15 min before staining. The primary Ab (1/100 dilution) for NURR1 (Santa Cruz Biotechnology, Santa Cruz, CA) was a rabbit polyclonal Ab mapping to the amino terminus of human and rat NURR1 (31). Following 2-h incubation with the primary Ab, a biotinylated secondary Ab (1/500; Vector Laboratories, Burlingame, CA) was spotted on sections, followed by the avidin-biotin-peroxidase complex (Vectastain Elite ABC kit; Vector Laboratories).

Immunofluorescence microscopy on K41 M synoviocytes was conducted using the similar procedures as that used for immunohistochemistry. Synoviocytes were incubated in diluted normal goat serum (Vector Laboratories). The primary polyclonal Ab for NURR1 was diluted 1/10 in 10% normal human serum and incubated on the sections for 90 min. Sections were washed in PBS and incubated for 30 min in biotinylated anti-rabbit secondary Ab (1/500; Vector Laboratories). A Cy3 or FITC fluorochrome-conjugated anti-biotin Ab (BN-34, 1:100; Sigma-Aldrich, St. Louis, MO) was added to bind the biotinylated anti-rabbit added previously. Following final washing in PBS, slides were mounted in DAKO fluorescent mounting medium (DAKO, Carpinteria, CA). For negative controls, isotype-matched nonimmune IgG were included and the primary Ab was preabsorbed with its specific synthetic peptide (blocking peptide (BP); 200 μg/ml; Santa Cruz Biotechnology).

Preparation of nuclear extracts and EMSAs

Nuclear protein extracts were prepared as previously described (21). Briefly, 1 × 10⁶ primary synoviocytes or 100 mg synovial tissue were homogenized with 1 ml TBS (25 mM Tris-HCl (pH 7.4), 130 mM NaCl, and 5 mM KCl). The homogenate was centrifuged at 7,000 × g for 30 s and resuspended in 0.25 ml buffer A (10 mM HEPES-KOH (pH 7.9), 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM DTT, 0.1 mM EGTA, 0.5 mM PMSF, and 2 μg each of the protease inhibitors antipain, pepstatin A, and aprotinin per milliliter). Then 1.25 μl 10% Nonidet P-40 was added and mixed, and the cell suspension was incubated for 2 min on ice. The cells were centrifuged at low speed (1,700 rpm), and the supernatant was removed (cytosolic fraction). To the pellet, 0.125 ml buffer B (0.4 M NaCl, 10 mM HEPES-KOH (pH 7.9), 1.5 mM MgCl₂, 0.1 mM EGTA, 0.5 mM DTT, 5% glycerol, and 0.5 mM PMSF) was added. The mixture was placed on ice for 2 h, with frequent agitation. Supernatants of the nuclear extracts were obtained by centrifugation at 15,000 × g for 5 min and then stored in aliquots at −80°C. The protein concentration was estimated with the Bradford protein assay kit (Bio-Rad).

For the EMSAs, 1 μg nuclear extract was incubated for 20 min in the presence of 20 mM HEPES (pH 7.9), 5 mM MgCl₂, 20% glycerol, 100 mM KCl, 0.2 mM EDTA, 8% Ficoll, 600 mM KCl, 500 ng/μl poly(deoxyinosinic-deoxycytidylic) acid, 50 mM DTT, and [α-³²P]dCTP-labeled double-stranded oligonucleotide. The following oligonucleotides were used in the mobility shift assays: κB wild type, 5′- TAGGGGAATCCCAT-3′; κB mutant, 5′-TAGGTTAATCCCAT-3′; CRE wild type, 5′-TCGTGACGTCAGGT-3′; and CRE mutant, 5′-TCGTGTGGTCAGGT-3′ (the underlined sequences indicate consensus sequence of the binding sites). The oligonucleotide probes were annealed with counter oligonucleotides, end labeled with [α-³²P]dCTP using DNA polymerase I (Klenow fragment), and then purified by Sephadex G-50 column chromatography. The DNA-protein samples were electrophoresed at 11 V/cm for 2.5 h through a 5.5% nondenaturing polyacrylamide gel in 0.5× Tris-borate-EDTA buffer. After gel electrophoresis, the gel was dried and exposed to x-ray film (Kodak X-OMAT; Kodak, Rochester, NY) with intensifying screens for 24 h at −80°C.

The specificity of the protein-DNA complexes was tested by immunodepletion and competition assays. For immunodepletion assays (supershift experiments), Abs to NF-κB/p50, NF-κB/p65, or CREB-1 (Santa Cruz Biotechnology) were incubated with nuclear extracts for 20 min at room temperature before EMSA. For competition studies, the EMSA reaction was performed with excess concentrations of nonradiolabeled oligonucleotide probe, as indicated.

Statistical analysis

Data are expressed as mean values ± SEM. Comparisons between normal synovium and synovium obtained from patients diagnosed with RA were made using the Student’s t test for unpaired values. Comparisons between treatments were made using Student’s t test for paired values.

Expression of NURR1 in normal and RA synovial tissue

Specific NURR1 immunostaining was extensive in all samples from patients with RA (n = 6) and significantly less in samples from normal patients (n = 2) (Fig. 1⇓). In normal synovial tissue, the localization of NURR1 was most intense in the synovial vasculature, including endothelial cells (Fig. 1⇓A). Occasional synoviocyte cells in the hypocellular normal synovial lining layer and sublining regions were also NURR1 positive. In RA, strong NURR1 staining was observed in synoviocyte and macrophage cells of the synovial lining layer and the subsynovial regions (Fig. 1⇓C). Synovial vascular endothelium was also a site of increased NURR1 staining in all inflamed tissue studied. Importantly, NURR1 staining in RA cells was predominately nuclear compared with dominant cytoplasmic localization in normal synovium (Fig. 1⇓, B and D).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Immunohistochemical staining for NURR1 in synovial tissue sections. Representative normal (NL; A and B) and RA synovial tissue (C and D) were stained with NURR1-specific antiserum. Positive immunoreactivity appears as brown. Nuclei are counterstained with hematoxylin (blue/purple). In normal synovium, specific NURR1 staining was seen in the synovial vasculature (arrow) and localized predominantly to the cytoplasm (double arrow). In RA synovial lining cells (LL) and sublining cells (SL) within the tissue stained for NURR1, with specific staining observed predominantly in the nucleus and to a lesser extent in the cytoplasm. Original magnification ×200 in A and B; ×450 in C and D.

PGE₂, IL-1β, and TNF-α promote NURR1 expression in RA and normal synoviocytes

Cytokine-stimulated release of resorptive agents such as MMPs and PGE₂ by synoviocytes occurs in association with a change from fibroblast-like to stellate morphology (37). Our immunohistochemical staining analysis indicates that TNF-α (Fig. 2⇓A), IL-1β, and PGE₂ (data not shown) induction of NURR1 also occurs in parallel with a transformation of RA synoviocytes (n = 3) from the fibroblast-like to stellate shape. Similar to the nuclear staining seen in vivo (Fig. 1⇑), NURR1 is primarily localized to the nucleus in primary RA synoviocytes under both basal and stimulated conditions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Regulation of NURR1 expression in primary RA synoviocytes. A, Immunohistochemical staining of cultured RA synoviocytes with anti-NURR1 immune serum. Primary RA synoviocytes were left untreated (NT) or stimulated for 1 h with TNF-α (10 ng/ml). Specific NURR1 staining is seen in the cytoplasm and nucleus of untreated cells. Following TNF-α stimulation, synoviocytes transform from a fibroblast-like to a stellate shape with a concomitant increase in NURR1. B, Time course of cytokine induction of NURR1 mRNA levels in RA synoviocytes. Cells were left untreated or treated with TNF-α (10 ng/ml), IL-1β (10 ng/ml), or PGE₂ (1 μM) for the indicated times. Total RNA was isolated and assayed by Northern analysis for expression of NURR1 and GAPDH. Mean NURR1 mRNA levels are expressed as arbitrary densitometry units of NURR1/GAPDH ratios. Representative results are shown. The experiment was performed using three individual RA synoviocyte cultures with similar findings. C, Northern analysis of NURR subfamily members in primary RA synoviocytes. Total RNA was isolated from RA synoviocytes untreated (NT) or treated for 1 h with TNF-α (10 ng/ml), IL-1β (10 ng/ml), or PGE₂ (1 μM) and assayed for expression of NURR1, NUR77, NOR-1, and GAPDH, as described in Materials and Methods.

We extended this analysis to investigate the relative effects of TNF-α, IL-1β, and PGE₂ on the induction of NURR1 gene expression in RA synoviocytes (n = 3) by Northern blot analysis (Fig. 2⇑, B and C). Basal NURR1 mRNA levels in primary RA synoviocytes varied between synovial tissue samples, reflecting differences in the grade of tissue inflammation and patient to patient variability. Importantly, induction of NURR1 mRNA levels by all inflammatory mediators tested peaked at 1–2 h and returned to basal levels within 4 h of treatment. TNF-α (6.87 ± 2.1-fold) and IL-1β (5.55 ± 1.9-fold) significantly up-regulated NURR1 mRNA (p < 0.05); consistently, PGE₂ had the most potent and sustained effect (31.3 ± 12.6-fold) in stimulating NURR1 mRNA levels in these cells. Coadministration of inflammatory stimuli augmented the effects on NURR1 transcription additively.

To elucidate the relative contribution of the three NURR subfamily members in the mediation of cytokine-induced inflammatory responses in synovium, we compared the induction of NURR1 (Fig. 2⇑B), NUR77, and NOR-1 gene expression in primary RA synoviocytes (Fig. 2⇑C). NUR77 and NOR-1 transcript levels were modestly and differentially regulated by each of the inflammatory agonists studied. Inflammatory treatment resulted in a rapid but modest induction of NUR77 and NOR-1 mRNA levels that also peaked at 1 h (Fig. 2⇑C) and declined thereafter. In contrast, NURR1 mRNA levels were significantly altered, indicating that this is the major cytokine-regulated member of the NURR subfamily.

To assess the spectrum of stimuli that activate NURR1 in these cells, we examined the ability of growth factors to regulate NURR1 expression (Fig. 3⇓). In contrast to the prominent effects of PGE₂, primary synoviocytes stimulated with bFGF and PDGF modestly up-regulated (up to 3.1- to 3.6-fold) and TGF-β had little effect (1.1-fold) on NURR1 mRNA levels. Immunohistochemical studies of NURR1 expression in primary synoviocytes confirmed that treatment with PDGF and bFGF induced a moderate increase in endogenous NURR1 protein (Fig. 3⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Induction of NURR1 expression in primary synoviocytes by growth factors. A, Primary synoviocytes were left untreated or treated with TGF-β (10 nM), PDGF (10 nM), bFGF (10 nM), or PGE₂ (1 μM) for 1 h. Following treatment, total RNA was isolated and assayed by Northern analysis for expression of NURR1. B, Immunohistochemical staining of cultured primary synoviocytes with anti-NURR1 immune serum or isotype-matched IgG. Synoviocytes were left untreated (NT) or stimulated for 1 h with PDGF or bFGF for 1 h.

Because proinflammatory cytokines and PGE₂ consistently appeared to be the more potent inducers of NURR1, we further tested the ability of these mediators to regulate NURR1 expression in synoviocytes from nonarthritic joints. A human synoviocyte line (K41 M) from a healthy donor has been established and well characterized (36). K41 M synoviocyte cells maintain the ability to respond to TNF-α and PGE₂ stimulation; however, these cells fail to respond to IL-1β due to loss in expression of IL-1R (36). Similar to the temporal responses observed in primary RA synoviocytes, a rapid and transient increase in NURR1 mRNA levels was observed following TNF-α and PGE₂ stimulation. In accordance with pathways shown to date to regulate NURR1, forskolin activation of cAMP/protein kinase A (PKA) pathways rapidly induced NURR1 in these cells (Fig. 4⇓A). Immunohistochemical studies of NURR1 expression in K41 M synoviocytes revealed a similar pattern of staining to that seen in primary RA synoviocytes (Fig. 2⇑). Treatment with TNF-α and forskolin induced a rapid and marked increase in endogenous NURR1 protein (Fig. 4⇓B). The specificity of NURR1 staining was verified by the significant reduction in staining seen when the NURR1 Ab was preincubated with an excess of Ag (BP).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Induction of NURR1 expression in normal human synoviocyte K41 M cell line. A, K41 M cells were left untreated (NT) or treated with TNF-α (10 ng/ml), IL-1β (10 ng/ml), PGE₂ (1 μM), or forskolin (25 μM) for 1 h. Following treatment, total RNA was extracted and Northern blots were performed, as described in Materials and Methods. NURR1 mRNA levels were measured using 20 μg total RNA. B, Immunohistochemical staining of cultured K41 M cells with anti-NURR1 immune serum. K41 M cells were left untreated (NT) or stimulated for 1 h with TNF-α (10 ng/ml) and forskolin (25 μM). Specificity of staining was determined by comparison with cells stained with NURR1 Ab that was preabsorbed with a NURR1-specific BP.

Inflammatory mediators enhance the transcriptional activity of the NURR1 promoter in primary RA and normal synoviocytes

To determine whether inflammatory mediators are capable of regulating the expression of the NURR1 promoter, we generated β-galactosidase reporter plasmids containing proximal promoter regions of the NURR1 gene and tested these in transfection experiments (Fig. 5⇓A). Transcriptional regulation of the NURR1 promoter was measured by transient transfection in primary RA synoviocytes. Individual RA synoviocyte cell lines were transiently transfected with each construct and incubated with or without cytokine. These experiments showed that in primary RA synoviocytes NURR1 basal activity for both the −1329/+132 and −396/+373 promoter regions were extremely high and variable. The autocrine action of PGE₂ produced by primary RA synoviocytes (3) may explain the high transcriptional activity of the NURR promoter.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Activation of transcription from the NURR1 promoter in cultured K41 M synoviocytes. A, To monitor NURR1 promoter activity, K41 M cells were transfected with −1329/+132 or −396/+373 NURR1 promoter constructs fused to a β-galactosidase reporter plasmid and maintained under normal conditions (NT), or exposed to TNF-α (10 ng/ml), PGE₂ (1 μM), or forskolin (FOR) (25 μM) for 12 h. Values (mean ± SEM) are the results from five individual experiments. B, Nucleotide sequences of the human and mouse NURR1 promoter. Positions of the κB and CRE consensus sequences are highlighted (underlined). ∗, Conserved nucleotides.

Having established that the K41 M cells produce endogenous NURR1 and respond to TNF-α and PGE₂ in a manner similar to primary RA synoviocytes, we assayed the transcriptional activity of the NURR1 promoter constructs using this synoviocyte cell line (Fig. 5⇑A). Consistent with the results of Northern blot analysis, TNF-α (4 ± 1.9-fold), PGE₂ (9.8 ± 1.7-fold), and forskolin (6.2 ± 1.6) significantly (p < 0.01) enhanced the transcriptional activity of the −1329/+132 NURR1 promoter region.

Similar to the larger NURR1 promoter region, the −396/+373 region was efficiently induced by PGE₂ and forskolin, but did not respond significantly to TNF-α treatment (Fig. 5⇑A). These results suggest that cis-acting elements mapping between −1329/−396 appear to be required for the TNF-α-mediated activation, while sequences within the proximal promoter −396 region can confer PGE₂ and forskolin responsiveness to the NURR1 promoter. It is well established that IL-1β and TNF-α mediate transcription coupling through NF-κB, and PGE₂ may signal by activation of CREB-dependent pathways. Examination of the 5′ flanking region of the NURR1 gene (38, 39) reveals potential consensus sequences for the binding of both NF-κB (−585/−576) and CREB (−171/−163) (Fig. 5⇑B). Moreover, the CRE and κB consensus sequences are highly conserved across the human and mouse NURR1 promoter sequence (Fig. 5⇑B). This very high evolutionary conservation suggests that these binding sites may have critical functional roles.

IL-1β and TNF-α induce NF-κB-binding activity to the NURR1 promoter

We have identified a potential consensus NF-κB binding site located within the −585- and −576-bp region of the NURR1 promoter (Fig. 5⇑B). Because of the role of this region in the inducibility of the promoter by TNF-α, we predicted that NF-κB activation may be necessary for cytokine-mediated effects on NURR1 expression in synovial tissue. To test the ability of this site to bind NF-κB, we designed a synthetic oligonucleotide that included the −585/−576 NURR1 κB site and used EMSA to examine DNA-binding properties of NF-κB in primary RA synoviocytes treated with IL-1β and TNF-α for 1 h (Fig. 6⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

EMSA of NF-κB binding to the NURR1 promotor in RA synoviocytes and RA synovial tissue. A, Nuclear extracts from primary RA synoviocytes, untreated (NT) or treated for 1 h with TNF-α (10 ng/ml) or IL-1β (10 ng/ml), were prepared and used in EMSA with ³²P-labeled (10 ng/ml) NURR1 oligonucleotide centered on the −584/−576 κB site. TNF-α- and IL-1β-inducible (complexes 1 and 2) and constitutive (complexes 3 and 4) DNA-protein complexes are indicated. For competition analysis, 50× molar excess of homologous oligonucleotide (κBwt) and mutant oligonucleotide (κBmt) was used. Nuclear extracts were examined in the presence of NF-κB subunit-specific antisera (p50Ab and p65Ab). The p65-p50 heterodimer was identified as forming the upper complex, because preincubation with antisera to both these subunits yielded a supershift (complexes 1 and 2) and a decrease in intensity of the upper band (complex 1). The lower band (complex 2) was formed by binding of a p50 homodimer, as indicated by loss of the lower complex using antisera to the p50 subunit alone. B, Binding activity to the −585/−576 NURR1 κB site in nuclear extracts of RA synoviocytes stimulated with TNF-α for 1 h and in nuclear extracts of synovial tissue from patients with RA (n = 6). ns, Nonspecific complex; ∗, the supershifted complex.

The −585/−576 NURR1 κB probe efficiently and specifically bound two inducible nuclear protein complexes (complexes 1 and 2) present in TNF-α- and IL-1β-treated primary RA synoviocytes but absent in unstimulated cells (Fig. 6⇑A, NT). As observed during a previous study of NF-κB regulation of cyclooxygenase-2 and MMP-1 expression by IL-1β (40, 41), we also detected two additional bands (complexes 3 and 4) not significantly regulated by IL-1β or TNF-α also bind to the κB probe. The identity of the proteins in these complexes and their significance are not known. Competition electrophoretic mobility shift experiments using 50-fold molar excess of unlabeled oligonucleotide revealed all complexes were specifically competed with self oligonucleotide (κBwt), but not with an oligonucleotide containing a mutated recognition site (κBmt), demonstrating that protein binding to the DNA fragment was specific and competitive.

We next investigated which NF-κB family members were responsible for the observed TNF-α- and IL-1β-induced binding activity present in complexes 1 and 2 (Fig. 6⇑A). TNF-α- and IL-1β-stimulated RA synoviocyte nuclear extracts were preincubated with Abs specific for NF-κB family members (p50 and p65). Addition of the NF-κB/p50 antiserum resulted in both a prominent supershift and reduction in complexes 1 and 2, while addition of NF-κB/p65 antiserum caused a significant reduction in only complex 1, strongly indicating that complex 1 contains a p65-p50 heterodimer and that complex 2 is comprised of a p50 homodimer. Taken together, these data confirm that the temporal IL-1β- and TNF-α-mediated increase in specific NF-κB protein subunit binding to the NURR1 promoter in these cells correlates with IL-1β- and TNF-α-mediated NURR1 transcription.

High DNA-binding activity of NF-κB to the NURR1 promoter in RA synovium

To determine the levels of NF-κB binding to the NURR1 promoter in RA synovial tissue, equal amounts of fresh nuclear extracts obtained from RA synovium were analyzed by EMSA using labeled −585/−576 NURR1 κB probe (Fig. 6⇑B). A marked and variable binding activity was detected in all RA samples (n = 6), indicating constitutive binding of NF-κB to the NURR1 promoter in RA synovium. The DNA-protein complexes detected correspond to the NF-κB/p50- and NF-κB/p65-binding activity observed in TNF-α-stimulated RA synoviocytes (Fig. 6⇑B). To further confirm the specificity of NF-κB subunit binding, we performed cold competition with self −585/−576 κB probe and immunodepletion assays (Fig. 6⇑B).

Constitutively CREB-1 binding to the NURR1 promoter in RA synoviocytes and synovial tissue

Using EMSA and nuclear extracts from primary RA synoviocytes (n = 3), we analyzed protein binding to the NURR1 CRE consensus binding site (CREwt), which contains the −171/−163 nucleotide sequence of the NURR1 promoter. Results shown in Fig. 7⇓ demonstrate that in unstimulated RA synoviocytes, nuclear proteins formed a DNA-protein complex that was specifically competed with self oligonucleotide (CREwt), but not with an oligonucleotide containing a mutated recognition site (CREmt). Stimulation of the cells with PGE₂ resulted in significantly increased binding of protein to this DNA fragment (Fig. 7⇓A). The increased DNA binding could be effectively competed with 50-fold molar excess of CREwt, but not with a mutated consensus sequence (CREmt). Supershift assays using anti-CREB-1 antiserum resulted in a shift of both the constitutive and PGE₂-inducible protein complex (Fig. 7⇓A). Thus, these results confirm that the CRE binding site plays an important role in the induction of this promoter by PGE₂ pathways and also contributes to the basal activity of the NURR1 promoter in synoviocyte cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

EMSA of protein binding to the NURR1 CRE promotor sequence in RA synoviocytes and synovial tissue. A, Nuclear extracts from primary RA synoviocytes, untreated (NT) or treated for 1 h with PGE₂ (1 μM), were prepared and used in EMSA with ³²P-labeled NURR1 oligonucleotide centered on the −171/−163 CRE site. For competition analysis, 50× molar excess of homologous oligonucleotide (CREwt) and mutant oligonucleotide (CREmt) was used. Composition of DNA-protein complexes in untreated and PGE₂-treated synoviocytes was examined in the presence of CREB-1-specific antiserum. B, Binding activity to the −171/−163 NURR1 CRE site in nuclear extracts of synovial tissue from patients with RA (n = 6). For competition analysis, 50× molar excess of homologous oligonucleotide (CREwt) was used. ∗, The supershifted complex.

To determine the extent of CREB binding to the NURR1 promoter in RA synovium (n = 6), equal amounts of fresh nuclear extracts obtained from RA synovium were analyzed by EMSA using the labeled −170/−163 NURR1 CRE probe (Fig. 7⇑B). A significant and variable protein-binding activity was detected in all RA samples (n = 6), indicating that CREB-1 binding to the NURR1 promoter is constitutive in RA synovium. The DNA-protein complexes detected correspond to the CREB-1-binding activity observed in PGE₂-stimulated RA synoviocytes (Fig. 7⇑). To further confirm the specificity of CREB-1 binding, we performed cold competition with self oligonucleotide-mutated oligonucleotide and CREB-1 immunodepletion assays (Fig. 7⇑).