To address the role of the nuclear receptor 4A (NR4A) family of orphan nuclear receptors in synoviocyte transformation, hyperplasia, and regulation of matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) in models of inflammatory arthritis.
NR4A messenger RNA levels in synovial tissue and primary synoviocytes were measured by quantitative reverse transcription–polymerase chain reaction (RT-PCR). NR4A2 was stably overexpressed in normal synoviocytes, and cell proliferation, survival, anchorage-independent growth, migration, and invasion were monitored in vitro. MMP and TIMP expression levels were analyzed by quantitative RT-PCR, and MMP-13 promoter activity was measured using reporter assays. Stable depletion of endogenous NR4A levels was achieved by lentiviral transduction of NR4A short hairpin RNA (shRNA), and the effects on proliferation, migration, and MMP-13 expression were analyzed.
NR4A2 was expressed at elevated levels in normal, OA, and RA synovial tissue and in primary RA synoviocytes. Tumor necrosis factor α (TNFα) rapidly and selectively induced expression of NR4A2 in synoviocytes. Ectopic expression of NR4A2 in normal synoviocytes significantly increased proliferation and survival, promoted anchorage-independent growth, and induced migration and invasion. MMP-13 gene expression was synergistically induced by NR4A2 and TNFα, while expression of TIMP-2 was antagonized. NR4A2 directly transactivated the proximal MMP-13 promoter, and a point mutation in the DNA binding domain of NR4A2 abolished transcriptional activation. Depletion of endogenous NR4A receptors with shRNA reduced synoviocyte proliferation, migration, and MMP-13 expression.
The orphan nuclear receptor NR4A2 is a downstream mediator of TNFα signaling in synovial tissue. NR4A2 transcriptional activity contributes to the hyperplastic and invasive phenotype of synoviocytes that leads to cartilage destruction, suggesting that this receptor may show promise as a therapeutic target in inflammatory arthritis.
Synovial hyperplasia and angiogenesis are prominent hallmarks of inflammatory arthritis. Inflammatory signals promote the recruitment of leukocytes into synovial tissue, increase the proliferation of resident synovial cells, and trigger the degradation of tissue via the activity of matrix-degrading enzymes (1–3). Although the extent of synovitis ranges from an aggressive pannus structure in rheumatoid arthritis (RA) to a mild thickening of the synovial layer in osteoarthritis (OA), these changes exacerbate inflammation within joints and ultimately contribute to the erosion of cartilage and bone. The proinflammatory cytokine tumor necrosis factor α (TNFα) is a major regulator of synovial hyperplasia and angiogenesis (2). Given the clinical benefits of targeting TNFα, it is critical to elucidate the molecular targets downstream of this cytokine pathway.
During the progression of arthritis, inflammation mediators such as TNFα transform synovial fibroblasts (synoviocytes) into aggressive cells that invade surrounding cartilage and bone. Much like tumor cells, transformed synoviocytes demonstrate increased proliferation, loss of contact inhibition, and anchorage-independent cell growth (1–3). A number of genes promoting inflammation, angiogenesis, and tissue degradation are expressed at elevated levels in transformed synoviocytes. TNFα induces factors such as vascular endothelial growth factor and interleukin-8 (IL-8), which promote blood vessel formation and leukocyte recruitment (2). TNFα is also a potent inducer of matrix metalloproteinases (MMPs), which are enzymes that degrade the extracellular matrix and mediate cartilage and bone erosion (4). Synoviocyte transformation is largely regulated by transcription factors activated by inflammatory pathways. For example, NF-κB, c-Myc, CREB, and p53 can modulate synoviocyte proliferation and invasion and MMP gene expression (5–10).
The NR4A subfamily of orphan nuclear receptors is another group of transcription factors with potential regulatory roles in the onset and progression of arthritis. These receptors consist of nerve growth factor–induced protein B (NGFI-B)/Nur-related protein 77 (NUR77, NR4A1), nuclear receptor related 1 protein (NURR1, NR4A2), and neuron-derived orphan receptor 1 (NOR-1, NR4A3). In contrast to other members of the nuclear receptor superfamily, NR4A receptors are ligand-independent transcription factors, and their activity is tightly controlled at the level of expression and posttranslational modifications (11). NR4A receptors induce transcription of target genes by binding to NGFI-B–responsive elements (AAAGGTCA motif) in promoter regions (12). These receptors can also act as transcriptional repressors through mechanisms involving protein–protein interactions with other transcription factors or coregulatory molecules. Although the NR4A receptors have a high degree of sequence homology and use similar transcription mechanisms, distinct roles for these 3 receptors have been described in a variety of cell and tissue types (11).
We previously observed elevated levels of NR4A receptors in synovial tissue and cartilage from patients with RA, patients with psoriatic arthritis, and patients with OA (13–17). Inflammatory signals activating the NF-κB and CREB pathways rapidly and potently induce NR4A expression in cells derived from inflamed joints (13). However, the transcriptional targets of these receptors and their impact on cellular activities associated with arthritis have not been thoroughly explored. In synoviocytes, NR4A2 induces IL-8 transcription (18, 19), potentially linking this transcription factor to cell migration and angiogenesis. In chondrocytes, NR4A2 antagonizes cytokine-induced MMP-1, MMP-3, and MMP-9 expression (14), suggesting a protective role of the receptor within cartilage. In endothelial cells, the NR4A receptors regulate cell survival, proliferation, migration, and angiogenesis (20–24). In some types of cancer, the NR4A receptors contribute to cellular transformation, increased proliferation, and cell survival (25–27). In light of emerging roles information on roles of the NR4A receptors in these processes, we hypothesized that the receptors may also promote synovial hyperplasia and tissue damage during the progression of inflammatory arthritis.
In this study, we investigated the function of the NR4A receptors in human synovial tissue and synoviocytes. NR4A2 is expressed at elevated levels in vivo and in synoviocytes stimulated with TNFα and prostaglandin E2 (PGE2). Overexpression of NR4A2 enhances synoviocyte proliferation, survival, anchorage-independent growth, migration, and cellular invasion. At the molecular level, NR4A2 induces MMP-13 transcription through a mechanism that requires the DNA binding domain of the receptor. Depletion of endogenous NR4A receptors attenuates synoviocyte proliferation, reduces migration, and suppresses MMP-13 expression. Taken together, these results show that NR4A2 promotes an aggressive phenotype in synoviocytes, suggesting that this receptor may have promise as a therapeutic target in inflammatory arthritis.
MATERIALS AND METHODS
Human synovial tissue and cell culture.
Human synovium used for quantitative reverse transcription–polymerase chain reaction (RT-PCR) was obtained from patients undergoing total knee replacement surgery at New York University Hospital for Joint Diseases. Nonarthritic normal synovium was obtained from patients with fractures or from accident victims (National Disease Research Interchange, Philadelphia, PA). RA synovial tissue specimens for histologic analysis were obtained from patients undergoing arthroscopic surgery at St. Vincent's University Hospital. Primary human synoviocytes were derived from patients with RA and were cultured as described previously (13). Activated and quiescent synoviocytes were used at passages 1–4 and 5–7, respectively. Human K4IM synoviocytes were cultured as described previously (28). Ethics approval was obtained from the local hospital ethics committee.
RNA extraction, reverse transcription, and quantitative PCR.
Total RNA from synovium and synoviocytes was isolated using TRIzol reagent (Sigma) and Qiagen RNeasy columns. RNA was reverse-transcribed into complementary DNA (cDNA) using an oligo(dT) primer and Moloney murine leukemia virus reverse transcriptase (Invitrogen). Quantitative RT-PCR was performed using SYBR Green Master Mix and primers spanning previously described exon junctions (29, 30). Relative expression levels were calculated using the 2 method, with normalization to GAPDH. Gene expression was measured with TaqMan Master Mix and gene expression assays (Applied Biosystems). Absolute expression levels of the NR4A receptors were determined using standard curves of cDNA plasmids, and results were presented as NR4A messenger RNA (mRNA) copies/GAPDH mRNA copies.
Immunofluorescence and immunohistochemistry.
Arthroscopic synovial tissue cryosections (7 μm) were fixed and blocked with goat serum (Vector). NR4A2 localization was detected by immunofluorescence using a primary anti-NR4A2 rabbit polyclonal antibody (N-20) (200 μg/ml; Santa Cruz Biotechnology). Following a second blocking step, sections were incubated with mouse anti–Ki-67 antibody (Dako). Each primary antibody was sequentially fluorescence labeled with fluorescein isothiocyanate (FITC)–conjugated goat anti-mouse IgG antibodies (Southern Biotechnology) and mouse anti-biotin–Cy3 antibodies (Sigma). Sections were counterstained with Hoechst (Invitrogen), mounted, and analyzed using a multiphoton microscope (Zeiss LSM 510 META, 200×). For immunohistochemical analysis, cryosections were fixed, blocked with horse serum, and incubated with mouse anti-CD68 (Dako). Sections were incubated with horse anti-mouse biotinylated secondary antibody, avidin–biotin–peroxidase solution (Vector), and 3,3′-diaminobenzidine (Sigma) and counterstained in Mayer's hemalum. Isotype-matched nonimmune IgG was included as a control for each of the primary antibodies.
Generation of stable K4IM transfectants.
Stable K4IM clones expressing control LacZ or NR4A2 cDNA were generated by Cell Trends. NR4A2 expression was confirmed by Western blotting and quantitative RT-PCR, and 3 clones were analyzed. Lentiviral transductions were conducted with NR4A2 cDNA or empty-vector control (kindly provided by Dr. C. J. M. de Vries, University of Amsterdam, Amsterdam, The Netherlands). Stable NR4A-depleted K4IM cells were generated by lentiviral transduction of control scrambled short hairpin RNA (shRNA) and shRNA specific for NR4A1, NR4A2, and NR4A3 receptors (Sigma-Aldrich). Puromycin-resistant colonies were selected, and NR4A knockdown was confirmed by quantitative RT-PCR and Western blot analysis. A population of mixed clones demonstrating NR4A knockdown was used in this study.
Cell surface area measurements.
Equivalent numbers (1 × 105) of stable K4IM clones expressing control LacZ cDNA or NR4A2 cDNA were cultured in 6-well plates. Surface area measurements were performed by staining cells with toluidine blue, capturing images (200× magnification), and measuring 50 cells/group using Zeiss LSM 5 Image Browser software.
Stable K4IM clones were seeded at a density of 0.1 × 106 (clones expressing LacZ or NR4A2 cDNA) or 0.5 × 105 (scrambled or NR4A1–3 shRNA cells) in 6-well plates. Cells were cultured in complete growth medium, and proliferation was measured using a Beckman Coulter Vi-CELL XR 2.01 system. Results were consistent in 3 independent experiments conducted in triplicate.
Cells (1 × 105) were grown in complete medium for 24–96 hours and were left untreated or were treated with CdCl2 (30 μM) for a further 24 hours. Cells were resuspended in calcium-binding buffer containing FITC-labeled annexin V (IQ products) for 10 minutes. Propidium iodide (1 mg/ml 3.4 mM sodium citrate, 1 mM Tris base, 0.08 mM EDTA, and 0.005× Triton) was added, and cells were analyzed using a flow cytometer (CyAn ADP 9).
Soft agar colony formation.
Cells (1 × 104) were seeded in 1 ml of 0.35% agarose/complete growth medium in 24-well plates and grown for 7 days. Viable cells were stained with alamarBlue (Invitrogen), and the optical density of each well was measured at 570 nm (25). Colonies were stained with 0.005% crystal violet, and the diameter of 20 colonies per well stained was measured using Image-Pro Plus software version 4.5. Each clone was assayed in triplicate, and results were consistent in 3 independent experiments.
Migration and invasion assays.
Cells were seeded at 4 × 104 in 24-well plates containing Millicell 8.0-μm hanging PET (polyethylene terephthalate) membrane inserts (Millipore). They were grown in complete growth medium for 48 hours, followed by removal of nonmigrating cells on the apical side of the insert. Cells that migrated to the basolateral side of the insert were stained with 0.25% crystal violet and counted. For invasion assays, hanging inserts were coated with type II collagen (150 μg/cm2; Sigma). Cells were seeded at 1 × 105 and grown in complete growth medium for 72 hours; those that invaded to the basolateral side of the membrane were fixed with 2% paraformaldehyde and stained with 0.1% crystal violet. Each clone was assayed in triplicate, and results were consistent in 3 independent experiments.
Cells were transiently transfected using GenePORTER 2 reagents (Gene Therapy Systems). MMP-13 promoter–luciferase constructs were kindly provided by Dr. Constance Brinckerhoff (The Geisel School of Medicine at Dartmouth, Hanover, NH). The putative NBRE site at −28 bp (AAAGGTAA) was mutated to AAAAACAA in the −181-bp construct, using site-directed mutagenesis. CMX-NURR1 expression vector and NBRE3-tk-Luc plasmid were kindly provided by Dr. Thomas Perlmann (Karolinska Institutet, Stockholm, Sweden). CMX-NURR1 C283G was generated by site-directed mutagenesis, as described previously (31). Reporter assays were repeated multiple times in triplicate, and consistent results were obtained.
K4IM stable shRNA cells were seeded at 2 × 105 in 6-well plates and grown to confluence. Wounds were generated by scratching with a pipette tip. After 0 or 12 hours, cells were fixed and stained with toluidine blue. Images were obtained at 40× magnification. Results were consistent in 3 independent experiments.
NR4A mRNA levels in human synovial tissue derived from normal, OA, and RA knee joints were measured by quantitative RT-PCR. All 3 NR4A transcripts were detected, with mean expression levels of 0.4 (NR4A1), 350 (NR4A2), and 1.5 (NR4A3) copies per 1,000 copies of GAPDH. In normal synovial tissue, absolute levels of NR4A2 greatly exceeded the levels of NR4A1 and NR4A3 (P < 0.001) (Figure 1A). Results consistent with this finding were observed in OA and RA tissue, with NR4A2 levels elevated to the same extent above NR4A1 and NR4A3 levels in normal, OA, and RA tissue.
To further examine expression patterns of the NR4A receptors, pooled RNA from another set of patients was hybridized to high-density microarrays. NR4A2 levels were 2-fold higher in OA and RA tissue compared with normal control tissue, while the levels of NR4A1 and NR4A3 in OA and RA tissue were not different from the level in normal tissue (data not shown). Thus, NR4A2 exhibited a unique expression pattern in synovium, where it was the most highly expressed of the NR4A receptors and, as demonstrated by microarray analysis, was present at elevated levels in advanced OA and RA. Consistent with this expression pattern and with previous results (13, 15, 17), the expression of NR4A2 protein was increased in inflamed synovium, with abundant nuclear expression in synoviocytes from the sublining and lining layers (Figure 1B and results not shown).
To examine potential links between NR4A2 expression and synovial hyperplasia, colocalization of NR4A2 and a cell cycle–associated protein, Ki-67, was examined in RA synovial tissue. Ki-67 expression was low or absent in tissue with little or no inflammation, as measured by CD68 staining (n = 4; results not shown). However, inflamed tissue stained positive for both NR4A2 and Ki-67 expression in the synovial sublining and lining layers (n = 6) (Figure 1B and results not shown). Nuclear colocalization of these factors suggested an in vivo relationship between NR4A2 expression, active proliferation, and inflammation. To investigate this further, NR4A mRNA expression was measured in activated and quiescent synoviocytes derived from patients with RA. NR4A2 levels were elevated in activated synoviocytes undergoing rapid proliferation (Figure 1C). However, they were sharply decreased in quiescent cells (P < 0.05). NR4A1 expression followed a similar trend, while NR4A3 levels remained constant (data not shown). The results for RA tissue and primary synoviocytes highlighted a correlation between NR4A2 levels and proliferation, suggesting that NR4A2 may promote cellular changes associated with synovial hyperplasia.
To address the cellular functions of NR4A2 in vitro, we used the normal human synoviocyte cell line, K4IM. This cell line has been used as a model to study synoviocyte activation and responses to inflammation (18, 19, 28). Consistent with the findings of our analysis of intact synovial tissue, absolute levels of NR4A2 mRNA in K4IM cells exceeded the levels of NR4A1 and NR4A3 (Figure 2A). Previous studies demonstrated that the NR4A receptors are rapidly and transiently induced by signals that promote synovial inflammation and tissue damage (13, 16, 18). To confirm this, we stimulated K4IM cells with physiologic concentrations of TNFα or PGE2 for 1 hour. TNFα selectively induced NR4A2 expression by 2-fold (P < 0.005) (Figure 2A), while the expression of NR4A1 and NR4A3 was not significantly altered. PGE2 potently induced expression of all 3 receptors and resulted in the highest absolute level of NR4A2 mRNA (P < 0.005) (Figure 2A). NR4A2 expression patterns in the K4IM cells paralleled those in inflamed synovial tissue, where NR4A2 was the most highly expressed member of the NR4A receptor subfamily and was potently regulated by inflammatory mediators.
To study the impact of NR4A2 on cellular processes linked to synovial hyperplasia and tissue degradation, NR4A2 was overexpressed in K4IM cells. Ectopic expression was confirmed by quantitative RT-PCR, and transcriptional activity of the receptor was validated using a consensus NBRE reporter construct (results not shown). Robust nuclear localization of ectopic NR4A2 protein paralleled localization of the endogenous receptor induced by PGE2 (Figure 2B).
TNFα-stimulated release of resorptive agents, such as MMPs, from synoviocytes occurs in association with a change from a fibroblast-like morphology to a stellate morphology (32). We previously observed that synoviocyte transformation following TNFα stimulation occurred with a concomitant increase in the expression of NR4A2 protein (13). In the absence of TNFα, stable NR4A2 clones exhibited a distinctive stellate morphology and a 50% reduction in cell surface area and nuclei circumference (Figure 2C and results not shown), suggesting that the cellular changes induced by TNFα may be mediated in part by NR4A2. Furthermore, the proliferation rates of NR4A2 clones exceeded those of control cells, reaching a 5-fold difference after 6 days (P < 0.005) (Figure 2D). In support of this finding, an increased percentage of NR4A2-overexpressing cells was observed in the S phase of the cell cycle by flow cytometry (data not shown). The cell viability of control cells and NR4A2 clones remained constant (>90%) during the course of the proliferation assays. Furthermore, NR4A2 clones exhibited resistance to cadmium chloride–induced apoptosis, suggesting that this receptor promotes cell-survival pathways (Figure 2E).
Next, we investigated the impact of NR4A2 on the processes of anchorage-independent growth, migration, and invasion. NR4A2 overexpression promoted formation of synoviocyte colonies in soft agar, and NR4A2 colonies were 3-fold larger than controls (P < 0.005) (Figures 3A and B), indicating persistent survival and growth in a detached state. In addition, the NR4A2 clones showed increased migration through Transwell filters relative to control cells (P < 0.005) (Figure 3C). NR4A2 potently increased synoviocyte invasion through Transwell filters coated with type II collagen or Matrigel (P < 0.005) (Figure 3D and data not shown). Taken together, these results showed that NR4A2 induced a transformed phenotype in normal synoviocytes, potentially by activating transcriptional pathways involved in hyperplasia, cell survival, and extracellular matrix degradation.
We hypothesized that NR4A2 may mediate the invasive potential of synoviocytes by regulating the expression of MMPs and TIMPs. K4IM cells express MMP-1, MMP-2, MMP-9, and MMP-13, and we investigated the effects of TNFα and NR4A2 on these genes, using quantitative RT-PCR. Stimulation with TNFα and NR4A2 induced MMP-13 expression and resulted in the synergistic induction of mRNA levels by 40-fold (P < 0.05) (Figure 4A). MMP-1 and MMP-9 were significantly induced by TNFα, but NR4A2 did not regulate these genes alone and, when administered with TNFα, did not increase their induction beyond that observed with the cytokine alone. MMP-2 expression was constitutively low and was not regulated by TNFα or NR4A2. TIMP-2 levels were suppressed by 70% by the combination of NR4A2 and TNFα (P < 0.01), while TIMP-1 was unaffected) (Figure 4B). In support of the results of our mRNA analysis, NR4A2 and TNFα synergistically induced MMP-13 protein secretion by 3-fold (P < 0.01) (Figure 4C). Taken together, these results showed that NR4A2 induced MMP-13 expression and antagonized TIMP-2 expression in K4IM cells. The net impact of these changes could ultimately promote extracellular matrix degradation and cell invasion.
We focused on the mechanism of NR4A2-induced regulation of MMP-13, because this enzyme plays a critical role in the degradation of type II collagen molecules in articular cartilage (4). NR4A2 transactivated human MMP-13 promoter constructs in transiently transfected K4IM cells (Figure 5A), indicating that regulation of this gene occurs at least in part at the level of transcription. The full-length MMP-13 promoter (−3.4 kb) was induced 2-fold by NR4A2, consistent with the effects of NR4A2 on endogenous MMP-13 mRNA and protein shown in Figure 4. The −405-bp and −181-bp promoter constructs were induced 4-fold by NR4A2, suggesting that NR4A2 targets the proximal region of the MMP-13 promoter. A putative NBRE site exists within the proximal promoter at position −28 bp; however, mutation of the core nucleotides in this sequence did not abrogate induction by NR4A2 (Figure 5B). Next, we tested a point mutation in the DNA binding domain of NR4A2, C283G, for its ability to regulate the MMP-13 promoter. This mutation disrupts a zinc finger motif in the receptor that is critical for DNA interactions (14, 18, 31). The C283G mutant failed to transactivate the −181-bp promoter (Figure 5C), suggesting that regulation of MMP-13 by NR4A2 requires a functional DNA binding domain.
To investigate the roles of the endogenous NR4A receptors in K4IM synoviocytes, levels of NR4A1, NR4A2, and NR4A3 were reduced by transduction with shRNA. We used this strategy because all 3 receptors were expressed in these cells (Figure 2), and the high degree of homology among these receptors could allow for compensatory functions (11). Endogenous NR4A1, NR4A2, and NR4A3 mRNA levels were reduced by 50–65% in K4IM cells stably transduced with specific shRNA molecules (Figure 6A). Depletion of the receptors attenuated proliferation by 35% at time points as early as 36 hours (P < 0.001) (Figure 6B), indicating that the endogenous receptors intersect with cell proliferation pathways. Furthermore, NR4A-depleted synoviocytes demonstrated a defect in wound-healing assays, in which they failed to migrate into monolayer scratches after 12 hours (Figure 6C). We also investigated regulation of MMP-13 by the endogenous receptors and observed a 50% reduction in TNFα-induced levels of MMP-13 in NR4A-depleted cells (Figure 6D). Taken together, these results suggest that the endogenous NR4A receptors are critical regulators of synoviocyte proliferation, migration, and MMP-13 gene expression.
In this study, we documented expression of the NR4A orphan receptors in synovial tissue and specifically explored the cellular and molecular functions of NR4A2 in synoviocytes. Our results highlight a potential pathogenic role of NR4A2, because overexpression of this receptor induces a phenotypic shift in normal synoviocytes that parallels the cellular transformation and hyperplasia observed during the progression of inflammatory arthritis. This study is the first to demonstrate that NR4A2 increases synoviocyte proliferation, survival, and anchorage-independent growth while enhancing cellular migration and invasion. Furthermore, we identified MMP-13 and TIMP-2 as transcriptional targets of NR4A2 in synoviocytes. These results complement recent studies proposing proinflammatory roles for the NR4A receptors in synoviocytes (18, 19).
During the progression of arthritis, synovial hyperplasia and angiogenesis promote inflammation and degradation of articular cartilage and bone (1–3). Transcription factors such as NF-κB and CREB regulate the expression of genes involved in these pathologic changes (6, 9). We previously demonstrated that NF-κB and CREB induce expression of the NR4A receptors in inflamed synovial and cartilage tissue (13–16). However, the roles of these receptors in synoviocyte transformation and hyperplasia have not yet been established. In addition to MMP-13 and TIMP-2 (identified in this study), other transcriptional targets of NR4A2 in synoviocytes include IL-8, amphiregulin, and Kit ligand (18, 19). Collectively, these gene products promote cell recruitment, proliferation, and survival and tissue damage in inflamed joints, suggesting that NR4A2 controls proinflammatory pathways in vivo.
The enhanced migration and invasion of synoviocytes in response to NR4A2 are likely mediated in part by the up-regulation of MMP-13. This collagenase has been linked to the invasion and metastasis of numerous types of cancer due to its ability to degrade extracellular matrix components and facilitate cell motility (4). Of particular importance in arthritic joints, MMP-13 has the highest catalytic activity against type II collagen, the major protein component of articular cartilage (4). Intraarticular expression of MMP-13 in mice results in synovial hyperplasia and inflammation (33), indicating that this enzyme may also contribute to the onset of inflammatory arthritis. NR4A2 also down-regulates TIMP-2, an endogenous inhibitor of MMPs, which could in turn promote cell migration and invasion. Furthermore, TIMP-2 functions as an inhibitor of angiogenesis by reducing the mitogenic response of microvascular endothelial cells to growth factors independently of MMPs (34). Within inflamed synovium, down-regulation of TIMP-2 could enhance angiogenesis and exacerbate inflammation and tissue damage.
NR4A2 induces MMP-13 transcription through a mechanism that targets the proximal promoter and requires a functional DNA binding domain. The C283G mutation disrupts a zinc finger motif in NR4A2 that is critical for DNA interactions (31), suggesting that NR4A2 induces MMP-13 transcription via direct interactions with the promoter. Several NR4A target genes are induced by receptor binding to promoter NBRE sequences (AAAGGTCA) (11). We identified a putative NBRE site at position −28 bp in the human MMP-13 promoter; however, mutation of the core residues in this sequence did not prevent induction by NR4A2. This site is immediately adjacent to the TATA box at position −37 bp and contains a single nucleotide difference from the consensus element (AAAGGTAA).
Our data suggest that NR4A2 utilizes an NBRE-independent mechanism to regulate the MMP-13 promoter. NR4A2 induces IL-8 transcription through an NBRE-independent mechanism that involves receptor interactions with p65 on the proximal promoter (18). Similar to IL-8, MMP-13 is synergistically induced by NR4A2 and TNFα, suggesting that cross-talk between NR4A2 and p65 may also contribute to the induction of MMP-13. Protein–protein interactions have also been implicated in the repression of NR4A target genes in other cell types. For example, NR4A1 interacts with p65 to antagonize the expression of steroidogenic enzymes, and NR4A2 uses similar mechanisms to attenuate inflammatory gene transcription (35, 36). In chondrocytes, NR4A2 represses MMP-1 expression via negative interactions with Ets transcription factors (14). Thus, NR4A2 may repress TIMP-2 expression through similar protein–protein interactions that converge on the promoter.
Although our study is the first to define functions for the NR4A receptors in synoviocyte proliferation, survival, anchorage-independent growth, and migration, these activities have been investigated in other systems. As with several members of the nuclear receptor family, cell- and tissue-specific functions for the NR4A receptors have been demonstrated (11). Differential roles for the receptors have been proposed in angiogenesis, a process that is regulated in part by endothelial cell proliferation and migration. NR4A3 promotes endothelial cell proliferation (22), while NR4A1 mediates cell cycle arrest by regulating p27Kip1 and cyclin A protein levels (37). In vivo, NR4A1 and NR4A2 positively impact angiogenesis by increasing endothelial proliferation, survival, and migration (20, 24). During vascular remodeling, NR4A1 blocks smooth muscle cell proliferation and macrophage recruitment (21, 38), while NR4A3 induces mitogenic activity (39, 40). The NR4A receptors have been implicated in several types of cancer, in which they enhance cell proliferation, survival, and migration (25–27, 41, 42). In colorectal cancer, decreasing NR4A2 levels may promote antitumor activities downstream of cyclooxygenase 2 inhibition (43). In melanoma cells, NR4A1 and NR4A2 have recently been linked to tumorigenicity and regulation of the Wnt/β-catenin pathway (44). Consistent with our findings in synoviocytes, NR4A2 expression is associated with anchorage-independent growth of cervical, prostate, and colon carcinoma cells in soft agar (25).
The NR4A2-dependent cellular activities we documented in synoviocytes may occur in part via changes in previously described target genes (18, 19). Furthermore, cyclin D2, a critical regulator of cell cycle progression and a transcriptional target of NR4A receptors in monocytes and smooth muscle cells (40, 45), may also mediate changes in synoviocyte proliferation. We are currently investigating additional NR4A2 target genes that may impact these activities in synoviocytes.
Because the NR4A receptors exhibit a high degree of sequence homology in their DNA binding domains (11), these receptors may have overlapping and redundant molecular functions. For example, mice deficient in NR4A1 or NR4A3 exhibit mild abnormalities (46, 47), while deletion of both genes leads to lethal acute myeloid leukemia (48). Our investigation focused on the role of NR4A2 in synoviocytes; however, we cannot exclude the possibility that NR4A1 and NR4A3 have similar functions in these cells. We detected expression of all 3 receptors in synovial tissue and synoviocytes, and it is possible that all of the NR4A receptors converge on similar transcriptional targets. To account for this possibility and to avoid potential redundancy issues, we used an shRNA knockdown strategy to reduce endogenous levels of all 3 receptors. Similar approaches have been used to establish roles of the NR4A receptors in DNA damage responses in melanocytes and regulation of the Wnt/β-catenin pathway in melanoma cells (44, 49). Depleting the NR4A receptors in synoviocytes attenuated proliferation and migration and also reduced expression of MMP-13. Importantly, these findings highlight a role of the endogenous receptors in regulating basal activities in synoviocytes.
In response to chronic inflammatory signals, elevated levels of the NR4A receptors may contribute to synovial hyperplasia and tissue damage in vivo. Our results suggest that these orphan nuclear receptors may show promise as therapeutic targets in inflammatory arthritis.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Mix had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Mix, Smyth, Fearon, Veale, Murphy.