SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To elucidate whether the microRNA (miRNA) cluster miR-17–92 contributes to the activated phenotype of rheumatoid arthritis synovial fibroblasts (RASFs).

Methods

RASFs were stimulated with tumor necrosis factor α (TNFα), and the expression and regulation of the miR-17–92 cluster were studied using real-time quantitative PCR (PCR) and promoter activity assays. RASFs were transfected with single precursor molecules of miRNAs from miR-17–92 and the expression of matrix-degrading enzymes and cytokines was measured by quantitative PCR and enzyme-linked immunosorbent assay. Potential miRNA targets were identified by computational prediction and were validated using reporter gene assays and Western blotting. The activity of NF-κB signaling was determined by reporter gene assays.

Results

We found that TNFα induces the expression of miR-17–92 in RASFs in an NF-κB–dependent manner. Transfection of RASFs with precursor molecules of single members of miR-17–92 revealed significantly increased expression levels of matrix-degrading enzymes, proinflammatory cytokines, and chemokines in precursor miR-18a (pre-miR-18a)–transfected RASFs. Using reporter gene assays, we identified the NF-κB pathway inhibitor TNFα-induced protein 3 as a new target of miR-18a. In addition, pre-miR-18a–transfected RASFs showed stronger activation of NF-κB signaling, both constitutively and in response to TNFα stimulation.

Conclusion

Our data suggest that the miR-17–92–derived miR-18a contributes to cartilage destruction and chronic inflammation in the joint through a positive feedback loop in NF-κB signaling, with concomitant up-regulation of matrix-degrading enzymes and mediators of inflammation in RASFs.

The microRNA-17–92 (miR-17–92) cluster has been associated with the maturation of the immune system and the development of hematopoietic tumors and other malignancies (1–3). The polycistronic primary transcript C13orf25 encodes 6 distinct members of the cluster: miR-17, 18a, 19a, 19b, 20a, and 92a (2, 4). Amplification of miR-17–92 has been observed in malignant diseases, leading to advantages in growth and survival due to the silencing of tumor-suppressor genes, such as cyclin-dependent kinase inhibitor 1A (p21), PTEN, or Bcl-2–interacting protein Bim, by individual miR-17–92 members (2, 3, 5–7). In miR-17–92–knockout mice, skeletal defects during embryogenesis were noted (8), and the animals died directly after birth due to severe developmental defects of the lung and the heart (9). In contrast, moderate overexpression of miR-17–92 in the lymphoid lineage in mice was found to lead to lymphoproliferative disease and autoimmunity (5). Interestingly, the expression of miR-17–92 is triggered by inflammatory cytokines, such as interleukin-6 (IL-6) (10, 11).

Rheumatoid arthritis (RA) is a chronic inflammatory disease that leads ultimately to the destruction of joints and bones. The cause of RA is still unknown, but it is clear that the disease is driven by infiltrating immune cells and by resident cells within the joint, as well as the mutual interaction between these cells (12). In this regard, RA synovial fibroblasts (RASFs) have been identified as key players in disease pathogenesis. Within the inflamed synovium, RASFs are stimulated by cytokines and other molecules to activate major signaling pathways, which leads to overproduction and secretion of matrix-degrading enzymes and mediators of inflammation. Immune cells are attracted to the synovium and are activated by RASF-secreted cytokines, and in turn, they stimulate RASFs, resulting in a self-sustaining cycle of chronic joint inflammation (13). Due to the disturbed expression pattern of various protooncogenes and tumor suppressors, the aggressive phenotype of the RASFs has been compared with the behavior of a locally invading tumor (14), and as such, these cells promote adhesion to, and destruction of, articular cartilage (12, 15). Similar to the pathogenesis of malignant tumors, aberrant expression of several miRNAs has recently been associated with the pathogenesis of RA (16).

MicroRNAs are characterized by their ability to target different messenger RNAs (mRNAs) simultaneously, which might influence several signaling pathways at once. MicroRNAs are considered to act by fine-tuning gene expression. The effects of a single miRNA on a specific target might thus be rather weak; however, by repressing many targets at the same time, a significant alteration of gene expression can occur (17–18).

We have previously shown that IL-6 induces miR-17–92, both in pulmonary arterial endothelial cells and in human hepatocytes, enhancing the acute-phase response in the latter cells (10, 11). Thus, to further elucidate the role of miR-17–92 in inflammatory processes, we examined the expression and function of miR-17–92, in particular, miR-18a, in RASFs in relationship to tumor necrosis factor α (TNFα), one of the major proinflammatory cytokines involved in the pathogenesis of RA (19, 20).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Patient samples/cell culture.

Synovial fibroblast cultures were established by collagenolytic digestion of synovial tissue specimens obtained from RA patients during joint surgery (kindly provided by Dr. C. Kolling, Schulthess Clinic, Zurich, Switzerland). Patients fulfilled the American College of Rheumatology 1987 criteria for RA (21). Fibroblast cultures were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (FCS) or in RPMI 1640 medium (only for the leukocyte Transwell migration assays). For stimulation experiments, cells were serum-starved (0.5% FCS) and stimulated with 10 ng/ml of TNFα (R&D Systems). To inhibit NF-κB signaling, RASFs were treated with 50 μM SC514 (EMD Millipore) or with DMSO (as control) for 1 hour before TNFα was added.

Real-time quantitative polymerase chain reaction (PCR).

We quantified mRNA and miRNA expression by SYBR Green real-time PCR using an ABI Prism 7900HT Sequence Detection System. Primer sequences for mRNA were as follows: for GAPDH, 5′-GGGAAGCTTGTCATCAATGGA-3′ (forward) and 5′-TCTCGCTCCTGGAAGATGGT-3′ (reverse); for C13orf25, 5′-TTGCTAAGTGGAAGCCAGAAG-3′ (forward) and 5′-CATCCACGTGGCAAAACAT-3′ (reverse) (as described by O'Donnell et al [22]); for matrix metalloproteinase 1 (MMP-1), 5′-GCTAACAAATACTGGAGGTATGATG-3′ (forward) and 5′-ATTTTGGGATAACCTGGATCCATAG-3′ (reverse); for IL-6, 5′-CCCTGAGAAAGGAGACATGTAAC-3′ (forward) and 5′-CCTCTTTGCTGCTTTCACACATG-3′ (reverse); for IL-8, 5′-TTGGCAGCCTTCCTGATTTC-3′ (forward) and 5′-TGGCAAAACTGCACCTTCAC-3′ (reverse); for monocyte chemoattractant protein 1 (MCP-1), 5′-CTCGCTCAGCCAGATGCAATC-3′ (forward) and 5′-AAGTTATAACAGCAGGTGACTGG-3′ (reverse); and for RANTES, 5′-CCAACCCAGCAGTCGTCTTTG-3′ (forward) and 5′-TGGCACACACTTGGCGGTTC-3′ (reverse).

MicroRNAs were quantified by stem–loop reverse transcription (RT)–PCR according to the method described by Chen et al (23), using the following primers: for miR-18a, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACCTATCT-3′ (RT) and 5′-GGCGGTAAGGTGCATCTAGT-3′ (forward); for miR-19, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCAGTT-3′ (RT) and 5′-CGGCGGTGTGCAAATCTATGC-3′ (forward); for miR-92a, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACACAGGC-3′ (RT) and 5′-CGGCGGTATTGCACTTGTCCC-3′ (forward); for universal miRNA reverse primer, 5′-GTGCAGGGTCCGAGGT-3′ (for miR-18a, miR-19, and miR-92a); and for miR-20a, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTACCTG-3′ (RT), 5′-CCTTAAAGTGCTTATAGTGCAG-3′ (forward), and 5′-TGCAGGGTCCGAGGTAT-3′ (reverse). Data were analyzed by the comparative Ct method. GAPDH mRNA was used as the endogenous control.

Plasmid construction.

Part of the C13orf25 promoter (–1,341 to +11, as described elsewhere [24]) was cloned into pGL3basic, and the GAPDH promoter (–1,087 to –24) was cloned into pRL (Promega). The 3′-untranslated regions (3′-UTRs) of dual-specificity phosphatase 16 (DUSP-16; 1–936 bp), protein tyrosine phosphatase 4A3 (PTP4A3; 1–476 bp), and TNFα-induced protein 3 (TNFAIP-3; 1–1,996 bp) were cloned into pmirGLO (Promega). Site-directed mutagenesis PCR was used to mutate the NF-κB binding site (–959 to –950 bp) in the C13orf25 promoter construct and the miR-18a seed matches in the 3′-UTR constructs (392–398 bp in DUSP-16, 405–412 bp in PTP4A3, and 671–677 bp in TNFAIP-3).

Transfection of small RNAs.

RASFs were transfected with 25 nM precursor miRNAs (pre-miRs) or 100 nM anti–miR-18a (both from Ambion) using Lipofectamine 2000 (Invitrogen). After 6 hours or 24 hours, the transfection mixture was replaced with fresh medium. Small interfering RNAs (siRNAs) (100 nM; Qiagen) were transfected using a Primary Mammalian Fibroblast Nucleofection kit (Amaxa/Lonza). Pre-miR Negative Control #1 (Ambion), Anti-miR Negative Control #1 (Ambion), or AllStars Negative Control siRNA (Qiagen) served as scrambled transfection controls. At 24 hours or 48 hours after transfection, cells were serum-starved for another 24 hours and then stimulated with TNFα as indicated.

Reporter gene assay.

Using Nucleofector technology (Amaxa/Lonza), RASFs were transfected with 1.2 μg of pRL_GAPDH plus 1.8 μg of pGL3basic_C13orf25prom (wild-type or mutated). After 24 hours, RASFs were serum-starved for another 24 hours and then stimulated for 8 hours with TNFα (10 ng/ml). To measure NF-κB activity, RASFs were transfected with 1.2 μg of pRL_GAPDH plus 1.8 μg of pGL4.32 (5 NF-κB response elements in row; Promega) by Amaxa Nucleofection. After 24 hours, RASFs were transfected with Pre-miR Negative Control #1 or with pre-miR-18a using Lipofectamine 2000 as described above. After 48 hours, cells were serum-starved for another 24 hours and stimulated for 5 hours with TNFα (10 ng/ml).

Using Lipofectamine 2000, HEK 293 cells were transfected with 180 ng of pRL_GAPDH plus 360 ng of pGL3basic_C13orf25prom (wild-type or mutated) and stimulated for 4 hours or 8 hours with TNFα (10 ng/ml). For the validation of miR-18a targets, HEK 293 cells were transfected with 230 ng of pmirGLO_DUSP-16, pmirGLO_TNFAIP-3 or pmirGLO_PTP4A3 (wild-type or mutated miR-18a seed match) together with 25 nM Pre-miR Negative Control #1 or with pre-miR-18a and harvested after 24 hours.

Firefly luciferase activity was measured with a Dual Luciferase Reporter Assay System (Promega), and the results were normalized to the activity of Renilla luciferase. An arbitrary cutoff of 1,000 relative luciferase units was used for data analysis.

Enzyme-linked immunosorbent assays (ELISAs) for MMP-1, IL-6, IL-8, MCP-1, and RANTES.

To measure MMP-1, IL-6, IL-8, MCP-1, and RANTES in cell culture supernatants, we used commercially available ELISA development kits from R&D Systems (MMP-1 and RANTES) or BD Biosciences (IL-6, IL-8, and MCP-1).

Isolation of peripheral blood leukocytes.

Fresh blood from healthy donors was diluted with 1 volume of isotonic NaCl solution and treated with 8 volumes of hypotonic erythrocyte lysis buffer (155 mM NH4Cl, 10 mM NaHCO3, 0.1 mM EDTA, pH 7.4) for 5 minutes on ice. After centrifugation, lysis was repeated with 4 volumes and then with 1.5 volumes of erythrocyte lysis buffer in isotonic NaCl. Cells were washed twice with ice-cold phosphate buffered saline (PBS), counted, and resuspended in RPMI 1640 medium at a concentration of 107 cells/ml.

Leukocyte Transwell migration assay.

In a 100-μl volume, 1 × 106 leukocytes per insert were seeded in a 96-well Transwell plate (5-μm polycarbonate membrane; Corning Costar). Medium (RPMI 1640 supplemented with 0.5% FCS) or conditioned medium (150 μl/well) from pre-miR-18a– or control-transfected RASFs was added to the feeder tray beneath the Transwell. Leukocytes were allowed to migrate for 6 hours. To facilitate detachment of adherent cells, 50 μl of ice-cold 20 mM EDTA/0.5% FCS in PBS was added to the lower wells, and the plates were incubated on ice for 15 minutes. Migrated cells were counted with a CASY Cell Counter (Schärfe Systems). Leukocytes from 2 different healthy donors were used for analysis of each conditioned medium preparation and vice versa. All conditions were measured in duplicate.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and Western blotting.

For Western blotting, we used rabbit anti–TNFAIP-3 (Abcam), mouse anti-PTP4A3 (R&D Systems), and mouse anti–α-tubulin (Sigma-Aldrich and Abcam) antibodies. Bands were detected using horseradish peroxidase–labeled species-specific secondary antibodies (Jackson ImmunoResearch) and enhanced chemiluminescence analysis (GE Healthcare). Protein expression was quantified by spot densitometry with Alpha Imager software (Alpha Innotech) or Bio-1D software (Vilber Lourmat).

Statistical analysis.

For statistical analysis, GraphPad Prism 5.0 software was used. Data were tested for Gaussian distribution using the Shapiro-Wilk normality test and were analyzed by parametric (paired 2-tailed t-test) or nonparametric (Wilcoxon's matched pairs signed rank test) statistical tests as appropriate. Values are presented as mean ± SD. P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

TNFα-induced expression of C13orf25 and mature miRNAs from the miR-17–92 cluster in RASFs.

TNFα is one of the major cytokines involved in the pathogenesis of RA (19, 20). We therefore investigated whether TNFα could influence the expression of miR-17–92 in RASFs. As shown in Figure 1A, the primary transcript of miR-17–92, C13orf25, was induced by TNFα in a time-dependent manner. Accordingly, mature miRNAs derived from miR-17–92 (i.e., miR-18a, 19a, 20a, and 92a) were up-regulated at 8 hours and 16 hours of TNFα stimulation (Figure 1B).

thumbnail image

Figure 1. Tumor necrosis factor α (TNFα) induces the expression of the microRNA-17–92 (miR-17–92) cluster via an NF-κB binding site in the C13orf25 promoter. A, Rheumatoid arthritis synovial fibroblasts (RASFs) (n = 4–16 samples) were left unstimulated (control) or were stimulated for the indicated times with TNFα. The C13orf25 primary transcript was significantly induced at 8, 16, and 24 hours of stimulation. B, TNFα stimulation for 8 hours and 16 hours increased the levels of mature miRNAs (i.e., miR-18a, 19a, 20a, and 92a) from miR-17–92 in RASFs (n = 15–16 samples). C and D, RASFs (n = 5 samples) were incubated in the presence of DMSO or SC514 for 1 hour and then were left untreated (⊘) or were stimulated for 8 or 16 hours with TNFα. TNFα-induced induction of C13orf25 (at 16 hours) (C) and of TNFα-induced protein 3 (TNFAIP-3) (D) was reduced by treatment with SC514. E and F, RASFs (n = 6 samples) (E) and HEK 293 cells (n = 5 samples) (F) were transfected with reporter gene constructs encompassing the C13orf25 promoter with wild-type (WT) or mutated (Δ) NF-κB binding site and stimulated for the indicated times with TNFα or left untreated (control). C13orf25 promoter activity was induced in the wild-type construct, whereas mutation of the NF-κB binding site inhibited the TNFα-mediated increase in C13orf25 promoter activity. Broken horizontal line represents the extension of the control culture value (set at 1.0) from the y-axis. Values are the mean ± SD. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 versus control in A and B and for the indicated comparisons in C–F, by paired t-test or Wilcoxon's matched pairs signed rank test. NS = not significant.

Download figure to PowerPoint

Dependence of the TNFα-induced C13orf25 expression on an NF-κB–binding site in its promoter.

TNFα is an important inducer of NF-κB and JNK signaling. To study whether C13orf25 induction by TNFα depends on these pathways, we treated RASFs with SC514, a chemical inhibitor of IκB kinase β, which is upstream of NF-κB nuclear translocation. We found that SC514 partly inhibited TNFα-induced C13orf25 expression (Figure 1C). Figure 1D shows the expression of TNFAIP-3, an established NF-κB target gene and negative regulator of NF-κB signaling in RASFs (25, 26), which was used as positive control for effective inhibition of NF-κB signaling by SC514. In contrast, treating RASFs with the JNK inhibitor SP600125 did not affect TNFα-induced C13orf25 expression (data not shown).

Zhou et al (27) previously demonstrated that the C13orf25 promoter contains 3 NF-κB binding sites; the one lying most proximal to the transcription start site (at –827 bp) is responsible for lipopolysaccharide-induced C13orf25 transactivation in human biliary epithelial cells. We used a reporter gene construct encompassing the C13orf25 promoter with a wild-type or mutated NF-κB binding site controlling the firefly luciferase gene to address whether the TNFα-induced transcription of C13orf25 in RASFs depends on this NF-κB binding site. The C13orf25 promoter activity was enhanced in RASFs stimulated for 8 hours with TNFα, and this induction was diminished when the C13orf25 promoter construct in which the NF-κB binding site had been mutated was used (Figure 1E). Similar results were obtained when HEK 293 cells were used for the reporter gene assay (Figure 1F), which suggests that NF-κB–dependent induction of C13orf25 by TNFα is a general mechanism that might be applicable to cell types other than RASFs. Since blockade of NF-κB and mutation of the NF-κB binding site did not completely abolish TNFα-dependent C13orf25 induction, other pathways might also be involved in this process.

Up-regulation of MMP-1, IL-6, IL-8, MCP-1, and RANTES and increased chemoattractive potential of RASFs following enforced expression of miR-18a.

A major hallmark of RASFs is the production of matrix-degrading enzymes and mediators of inflammation (i.e., cytokines and chemokines), which contribute to cartilage degradation and joint infiltration by immune cells (13). To elucidate which roles miRNAs from the miR-17–92 cluster might play in the activation of RASFs, we chose the expression of MMP-1 as a functional readout for the effect of miR-17–92 transfection on RASFs. We transfected RASFs with precursor molecules of miR-18a, 19a, 20a, and 92a (representing the 4 different miRNA families of the cluster) and subsequently stimulated the transfected RASFs with TNFα for 6 hours. Only pre-miR-18a transfection led to a significant increase in both the constitutive and the TNFα-induced expression of MMP-1 mRNA (Figure 2A). Our further experiments were thus focused on this miRNA.

thumbnail image

Figure 2. Promotion of the expression of matrix-degrading enzymes and mediators of inflammation in rheumatoid arthritis synovial fibroblasts (RASFs) by microRNA-17–92 (miR-17–92)–derived miR-18a. A, RASFs (n = 5 samples) were transfected with scrambled pre-miRNA (scr) or with precursor molecules of single miRNAs (pre-miRNAs) and were left unstimulated (control) or were stimulated for 6 hours with tumor necrosis factor α (TNFα), after which the fold change in matrix metalloproteinase 1 (MMP-1) levels was measured. B, RASFs (n = 8 samples) were transfected with scrambled miRNA or with pre-miR-18a, and after 48 hours, were left unstimulated (left) or were stimulated for 24 hours with TNFα (right). Levels of mRNA for MMP-1, interleukin-6 (IL-6), monocyte chemoattractant protein 1 (MCP-1), and RANTES, but not IL-8, were increased in pre-miR-18a–transfected RASFs under unstimulated conditions, and levels of all genes examined were increased following TNFα stimulation. C, RASFs (n = 8 samples) were transfected with scrambled anti-miRNA or with anti–miR-18a, and after 24 hours, were left unstimulated (left) or were stimulated for 24 hours with TNFα (right). Levels of mRNA for IL-8 and MCP-1 were reduced in unstimulated anti–miR-18a–transfected RASFs. D, MicroRNA-18a was overexpressed (left) or silenced (right) in the RASFs examined in B and C, respectively, and the fold change in miR-18a was determined in the untreated (⊘) or the TNFα-treated cells. Broken horizontal line represents the extension of the control culture value (set at 1.0) from the y-axis. Values are the mean ± SD. = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001; ∗∗∗∗ = P < 0.0001 versus scrambled pre-miRNA–transfected cells, by paired t-test or Wilcoxon's matched pairs signed rank test. NS = not significant.

Download figure to PowerPoint

To gain deeper insight into the role of miR-18a in the activation of RASFs, we measured MMP-1, IL-6, IL-8, MCP-1, and RANTES in pre-miR-18a– and anti–miR-18a–transfected RASFs at the mRNA level (Figures 2B and C). The constitutive expression of MMP-1, IL-6, MCP-1, and RANTES was increased by pre-miR-18a transfection, and TNFα-induced up-regulation of MMP-1, IL-6, IL-8, MCP-1, and RANTES (at 24 hours of stimulation) was further enhanced by pre-miR-18a transfection. Conversely, inhibition of miR-18a using anti–miR-18a significantly reduced the expression of IL-8 and MCP-1 under unstimulated conditions. In TNFα-treated cells, antagonizing miR-18a showed a trend toward reduced cytokine expression. These results, however, did not reach statistical significance. One can speculate that in TNFα-stimulated cells, anti–miR-18a blocks miR-18a in an inefficient manner (Figure 2D). In this regard, TNFα seemed to counteract the anti–miR-18–mediated inhibition of miR-18a by inducing the expression of miR-18a in RASFs. Our data indicate that miR-18a contributes to the activation of RASFs by up-regulating the expression of important mediators of matrix infiltration and degradation.

We further corroborated our findings by measuring secreted protein levels of MMP-1, IL-6, IL-8, MCP-1, and RANTES in supernatants from RASFs that had been transfected with pre-miR-18a. Analogous to the mRNA data (Figure 2B), miR-18a–overexpressing RASFs secreted higher levels of MMP-1 and the cytokines we investigated than did the control cells (Figure 3A).

thumbnail image

Figure 3. Increased chemoattractive potential of RASFs conferred by miRNA-18a. A, Levels of MMP-1, IL-6, and the chemokines IL-8, MCP-1, and RANTES were increased in the supernatants of pre-miR-18a–transfected RASFs (n = 8 samples; same as in Figure 2B). # = MMP-1 levels in unstimulated scrambled RASFs were noted below the lowest standard value (156 pg/ml). B, In a Transwell migration assay, conditioned medium (CM) from TNFα-stimulated RASFs (n = 8 samples) that had been transfected with pre-miR-18a attracted increased numbers of leukocytes toward the lower chamber as compared with scrambled-transfected RASFs (scr). Leukocyte migration toward conditioned medium from unstimulated RASFs (⊘) was similar to background migration (medium alone). Values are the mean ± SD. = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001; ∗∗∗∗ = P < 0.0001 for the indicated comparisons, by paired t-test or Wilcoxon's matched pairs signed rank test. See Figure 2 for other definitions.

Download figure to PowerPoint

To determine whether the effects of miR-18a on gene expression in RASFs, in particular, the increased secretion of chemokines IL-8, MCP-1, and RANTES, might indeed play a functional role in the pathogenesis of RA, we performed a Transwell migration assay using peripheral blood leukocytes from healthy donors and conditioned medium from RASFs transfected with pre-miR-18a and stimulated with TNFα placed in the lower compartment as a chemoattractant. Figure 3B demonstrates that conditioned medium from pre-miR-18a–transfected and TNFα-stimulated RASFs attracted more leukocytes toward the lower chamber than did conditioned medium from scrambled control RASFs. These experiments show that miR-18a up-regulation in RASFs increases the chemoattractive potential of these cells and, thus, might contribute to the infiltration of the inflamed joint with immune cells.

Direct targeting of the NF-κB signaling inhibitor TNFAIP-3 by miR-18a in RASFs.

In RASFs, the expression of MMPs, cytokines, and chemokines depends on the activation of important signaling pathways, such as NF-κB, p38, and JNK signaling (28, 29). Based on the fact that miRNAs are negative regulators of gene expression, whereas miR-18a was found to up-regulate downstream targets of TNFα signaling in RASFs, we hypothesized that miR-18a might repress inhibitors of signal transduction pathways. Using the TargetScan database (www.targetscan.org) for computational prediction of miRNA targets, we selected the NF-κB pathway inhibitor TNFAIP-3 (26), the p38 phosphatase PTP4A3 (30), and the JNK phosphatase DUSP-16 (31) as potential candidates for target validation.

Since miRNAs repress gene expression by binding to the 3′-UTRs of mRNAs, leading to mRNA degradation and/or inhibition of translation (17, 18), we cloned the 3′-UTRs of the 3 potential miR-18a targets into a reporter vector downstream of the luciferase gene. Cotransfection of these reporter vectors with pre-miR-18a into HEK 293 cells reduced the luciferase activity when it was under control of the TNFAIP-3 and PTP4A3 3′-UTRs. Pre-miR-18a also reduced DUSP-16 3′-UTR–controlled luciferase activity; however, the results did not reach statistical significance. Mutation of the respective miR-18 seed matches in those 3′-UTRs rescued luciferase activity in miR-18a–transfected cells (Figure 4A). These data indicate that miR-18a directly represses TNFAIP-3 and PTP4A3 via the miR-18 seed matches in their 3′-UTRs.

thumbnail image

Figure 4. The NF-κB signaling pathway inhibitor tumor necrosis factor α–induced protein 3 (TNFAIP-3), a direct target of miR-18a in RASFs. A, Using reporter vectors encompassing the 3′-untranslated regions (3′-UTRs) of TNFAIP-3, protein tyrosine phosphatase 4A3 (PTP4A3), and dual-specificity phosphatase 16 (DUSP-16), transfection with pre-miR-18a was found to significantly reduce the luciferase activity of the wild-type (WT) TNFAIP-3 and PTP4A3 reporter constructs. Mutated (Δ) miR-18a seed matches returned the luciferase activity to the levels in cells transfected with scrambled miRNA (scr) (n = 4 samples each). B, Transfection with pre-miR-18a increased levels of mRNA for TNFAIP-3 (left), but not PTP4A3 (right), in RASFs (n = 8 samples; same as in Figure 2B). C and D, Expression of TNFAIP-3 and PTP4A3 proteins was determined by Western blotting (C) and then quantified by densitometry (D). TNFAIP-3 protein expression was reduced in RASFs (n = 7 samples) transfected with pre-miR-18a for 48 hours, whereas PTP4A3 levels were not affected. α-tubulin was used as a loading control. Values in A, B, and D are the mean ± SD. = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 for the indicated comparisons, by paired t-test or Wilcoxon's matched pairs signed rank test. See Figure 2 for other definitions.

Download figure to PowerPoint

Subsequently, we analyzed the expression of TNFAIP-3 and PTP4A3 in RASFs transfected with pre-miR-18a. Whereas PTP4A3 mRNA levels were not affected, TNFAIP-3 mRNA expression was increased in RASFs overexpressing miR-18a (Figure 4B). Since miRNAs often inhibit only translation and do not mediate degradation of the target mRNA, we also measured protein levels of TNFAIP-3 and PTP4A3 in miR-18a transfected RASFs (Figures 4C and D). Despite directly targeting the PTP4A3 3′-UTR in the reporter gene assay, miR-18a transfection failed to reduce PTP4A3 levels in RASFs. In contrast, TNFAIP-3 protein levels were reduced at 48 hours after transfection. This discrepancy compared with the increased mRNA levels of TNFAIP-3 might be explained by the fact that TNFAIP-3 itself is a negative feedback regulator of NF-κB and, thus, translational repression of TNFAIP-3 by miR-18a in RASFs may induce the transcriptional activation of TNFAIP-3 via enhancement of NF-κB signaling.

Concomitant up-regulation of C13orf25 and TNFAIP-3 in TNFα-stimulated RASFs.

We next examined the expression of TNFAIP-3 and C13orf25 in matched TNFα-stimulated RASF samples. Transcription of both TNFAIP-3 and C13orf25 was induced by TNFα (Figures 5A and B). Interestingly, the significant up-regulation of TNFAIP-3 protein (Figures 5C and D) was much weaker than that at the mRNA level, suggesting that the efficiency of translation was reduced under these conditions. This may be partly due to the concomitant up-regulation of miR-17–92, in particular of miR-18a, and thus the simultaneous translational repression of TNFAIP-3. A similar scenario was previously suggested for the relationship of c-Myc–induced miR-17–92 and its target E2F, leading to diminished induction of E2F protein as compared with the corresponding mRNA (32).

thumbnail image

Figure 5. Expression kinetics of tumor necrosis factor α–induced protein 3 (TNFAIP-3) and C13orf25 in TNFα-stimulated rheumatoid arthritis synovial fibroblasts (RASFs). RASFs (n = 8 samples) were left unstimulated (control) or were stimulated for 8, 16, or 24 hours with TNFα, and levels of mRNA for TNFAIP-3 (A), mRNA for C13orf25 (B), and TNFAIP-3 protein (C) induced by TNFα were measured. A representative Western blot image is also shown (D). α-tubulin was used as a loading control. Broken horizontal line represents the extension of the control culture value (set at 1.0) from the y-axis. Values in A–C are the mean ± SD. ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 versus control, by paired t-test or Wilcoxon's matched pairs signed rank test.

Download figure to PowerPoint

Mimicking miR-18a–mediated effects in RASFs by silencing of TNFAIP-3.

Having demonstrated that TNFAIP-3 is a direct target of miR-18a in RASFs, we addressed whether silencing of this NF-κB inhibitory gene could mimic the effects of miR-18a on MMP-1 and cytokine expression. Therefore, we transfected RASFs with siRNAs targeting TNFAIP-3, stimulated the cells with TNFα, and measured the target genes by real-time PCR. As shown in Figure 6A, TNFAIP-3 silencing was efficiently achieved. Under unstimulated conditions, silencing of TNFAIP-3 increased the mRNA levels of IL-6, IL-8, and RANTES, whereas MMP-1 and MCP-1 mRNA levels were not significantly affected. In contrast, following TNFα stimulation, silencing of TNFAIP-3 increased the expression of all 5 genes studied (Figure 6B).

thumbnail image

Figure 6. Enhanced activation of RASFs by TNFα-induced miR-18a through a positive feedback loop in NF-κB signaling via TNFAIP-3. A and B, RASFs (n = 8 samples) were transfected with scrambled (scr) small interfering RNA (siRNA) or with siRNA targeting TNFAIP-3 and were stimulated for 24 hours with TNFα. Expression of TNFAIP-3 mRNA (left) and protein (right) was efficiently silenced, as demonstrated quantitatively (top), as well as by Western blotting (bottom) of 4 samples (RA1–RA4) (A). α-tubulin was used as a loading control for the Western blots. A representative Western blot image is shown. Silencing of TNFAIP-3 increased constitutive levels of mRNA for IL-6, IL-8, and RANTES (left), while the TNFα-induced expression of all 5 genes studied was enhanced (right) (B). Broken horizontal line represents the extension of the control culture value (set at 1.0) from the y-axis. C, RASFs (n = 8 samples) transfected with an NF-κB–responsive reporter vector and pre-miR-18a showed stronger activation of NF-κB signal transduction than did scrambled miRNA–transfected RASFs. Values in A–C are the mean ± SD. = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001; ∗∗∗∗ = P < 0.0001 versus scrambled (scr) siRNA–transfected RASFs in B and for the indicated comparisons in A and C, by paired t-test or Wilcoxon's matched pairs signed rank test. D, Shown is a model of the signaling events in RASFs involving miR-18. See Figure 2 for other definitions.

Download figure to PowerPoint

Enhancement of NF-κB signaling by miR-18a.

Finally, we sought to investigate whether miR-18a could influence NF-κB signaling in RASFs, as suggested by the repression of the NF-κB pathway inhibitor TNFAIP-3. As expected, RASFs transfected with miR-18a displayed a stronger constitutive, as well as TNFα-induced, NF-κB activity (Figure 6C).

Taken together, we suggest the following model of the signaling events involving miR-18a in RASFs (Figure 6D). Chronic synovial inflammation, as simulated in vitro by TNFα, induces the activation of NF-κB and, in turn, the up-regulation of miR-18a and TNFAIP-3. In a negative feedback loop, TNFAIP-3 terminates the action of NF-κB. Repression of TNFAIP-3 by miR-18a interferes with this feedback loop, resulting in enhanced expression of downstream effector genes.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

In this study, we have shown that the miR-17–92 cluster is induced in RASFs by TNFα through activation of NF-κB and that miR-18a is involved in the up-regulation of both constitutive and TNFα-induced secretion of MMP-1 and inflammatory cytokines and chemokines and, thus, increases the chemoattractive potential of RASFs. Moreover, using reporter gene assays, we identified TNFAIP-3 as a novel direct target of miR-18a, and we demonstrated that by repressing TNFAIP-3 expression, miR-18a enhances NF-κB signaling in RASFs. Thus, miR-18a constitutes a novel positive feedback loop of the NF-κB signaling pathway via translational repression of TNFAIP-3 (Figure 6D), which might further aggravate the activated phenotype of synovial fibroblasts in the pathogenesis of RA.

In recent years, it has become clear that miRNAs are dysregulated in RA and that, acting as repressors of gene expression, they contribute to autoimmunity and joint destruction (for review, see ref.16). The miR-17–92 cluster has been associated with important inflammatory processes, such as the acute-phase response (11), and maturation of the immune system. Xiao and colleagues (5), for example, have shown that in mice, overexpression of miR-17–92 in lymphocytes led to the development of autoimmunity, as reflected by the production of autoantibodies and the infiltration of lymphocytic cells into nonlymphoid tissues. In a mouse model of systemic lupus erythematosus, miR-17–92 was found to be specifically up-regulated in splenic T cells (32), and another study showed that miR-17 and miR-19b are critical regulators of Th1 cell responses and Treg cell differentiation (33).

We found in our study an up-regulation of the miR-17–92 primary transcript and its mature miRNAs in response to TNFα, which in turn, repressed the NF-κB pathway inhibitor TNFAIP-3, thus mediating proinflammatory functions in our experiments. A recent study by Philippe et al (34), on the other hand, found miR-19a and miR-19b to be repressed in RASFs by stimulation with Toll-like receptor 2 (TLR-2) and TLR-4 ligands. Moreover, TLR-2 was identified as a direct target of the miR-19 family (34). Thus, it may be assumed that different signals connected to the pathogenesis of RA might lead to differential expression and function of miR-17–92 in RASFs. Interestingly, it has been predicted that miR-19a, similar to miR-18a, targets TNFAIP-3 (www.targetscan.org), and our own unpublished data using reporter gene assays with the 3′-UTR of TNFAIP-3 confirmed this prediction. Furthermore, miR-19 was shown to repress the NF-κB inhibitor CYLD (35). Gantier and colleagues (36) very recently found that miR-19b targets not only TNFAIP-3, but also other negative regulators of NF-κB signaling. Based on these findings, we strongly propose a proinflammatory role of miR-17–92 in the pathogenesis of RA.

In our screening to detect a functional role of miR-17–92 in RASFs, we chose MMP-1 expression as readout and found that miR-18a had significant effects on mRNA levels of MMP-1. Further analysis showed that miR-18a also enhanced the expression of other important mediators involved in the pathogenesis of RA, including IL-6, IL-8, MCP-1, and RANTES. These findings, together with data from our previous work showing that miR-18a targets the protein inhibitor of activated STAT-3 (PIAS-3), an inhibitor of STAT-3 signaling (11), we postulated that miR-18a might similarly target inhibitors of TNFα-induced signaling pathways, and we found by a computational approach using TargetScan and subsequent validation experiments, that miR-18a acts as a repressor of TNFAIP-3. Interestingly, it has been shown that knockout of TNFAIP-3 in myeloid cells triggers the development of an erosive polyarthritis (37); conversely, adenoviral delivery of TNFAIP-3 was shown to improve inflammation and bone destruction in collagen-induced arthritis (25). These data show that the NF-κB signaling pathway is a crucial mediator of arthritis, and they imply that controlling NF-κB activity (directly via TNFAIP-3 or indirectly through miR-18a) in the different cell types involved in the pathogenesis of RA may be a promising therapeutic approach.

In addition to TNFAIP-3, we identified PTP4A3 as a new target of miR-18a by use of a reporter gene assay. However, transfection with pre-miR-18a did not reduce PTP4A3 levels in RASFs. In general, target repression by miRNAs through the miRNA-induced silencing complex may be additionally regulated by RNA binding proteins that facilitate or inhibit the interaction of miRNA with mRNA, depending on different cellular and environmental conditions (38). One might speculate that the functional discrepancy observed between 3′-UTR targeting of PTP4A3 by miR-18a in HEK 293 cells and in RASFs result from the different cellular environments. It thus appears that repression of TNFAIP-3 is the major contributor to the miR-18a–mediated up-regulation of MMP-1 and inflammatory cytokines.

Silencing of TNFAIP-3, however, did not entirely duplicate the effects of miR-18a transfection, as shown strikingly by experiments involving IL-8. MicroRNA-18a increased the expression of IL-8 only when cells were stimulated with TNFα, whereas silencing of TNFAIP-3 additionally up-regulated the constitutive expression of IL-8 mRNA. These results indicate that additional, as-yet-unidentified, targets of miR-18a may be responsible for the final gene expression observed after pre-miR-18a transfection. One such target may be connective tissue growth factor, which was shown to be de-repressed by anti–miR-18a treatment in glioblastoma spheroid cultures (39) and which has been connected to the expression of IL-6 and IL-8 in tendon fibroblasts (40). The involvement of connective tissue growth factor in MMP-1 and cytokine expression in RASFs remains elusive for the moment and warrants further experimental analysis.

This study is the first to show that miR-18a is part of a positive regulatory loop in NF-κB signaling in RASFs. Studies from recent years have established a complex network of miRNAs that positively and negatively regulate NF-κB signaling at different levels of the signaling cascade (for review, see ref.41). Two miRNAs that have been associated with the pathogenesis of RA are miR-146a and miR-155, the latter of which is involved in the production of proinflammatory cytokines and the development of collagen-induced arthritis (42–45). Both of them are induced through NF-κB–dependent pathways but seem to be engaged in negative feedback loops. MicroRNA-155, for example, may repress inflammatory signaling by targeting the inhibitor of κB kinases β and ε (46). Together, these data support the importance of intact NF-κB signaling. Different levels of regulation appear to include the involvement of various miRNAs to result in a finely balanced activation pattern that becomes disturbed under pathologic conditions such as RA.

In conclusion, the results of this study, together with our previous data (11), underscore the role of miR-18a as a regulator of intracellular signaling pathways that acts as an endogenous amplifier of extracellular signals. More work is needed to complete the picture of miR-18a functions in cell signaling and to further unravel the role of miR-17–92 in autoimmune diseases, particularly in RA. Nevertheless, we have shown that miR-18a is induced by TNFα and is part of a positive feedback loop in the NF-κB signaling pathway in RASFs. MicroRNA-18a thus increases the expression of MMP-1 and cytokines and enhances the chemoattractive potential of RASFs by potentiating the effect of TNFα. These data suggest that miR-17–92 plays an important role in TNFα-mediated signaling mechanisms, which although not directly evidenced by our experiments, might result in RASF-mediated cartilage destruction and immune cell infiltration of the joint in the pathogenesis of RA.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

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. Trenkmann 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. Trenkmann, Brock, R. E. Gay, Michel, S. Gay, Huber.

Acquisition of data. Trenkmann, Brock.

Analysis and interpretation of data. Trenkmann, Brock, S. Gay, Huber.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Synovial tissue specimens were kindly provided by Dr. C. Kolling (Schulthess Clinic, Zurich, Switzerland). We thank Drs. M. Connolly, M. Frank, A. Kopitar, and C. Ospelt for valuable suggestions. We are grateful for the excellent technical assistance of M. Comazzi, B. Henriques Campos, P. Künzler, S. Mettler, and U. Treder.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES