MicroRNA-29, a Key Regulator of Collagen Expression in Systemic Sclerosis

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 62, No. 6, June 2010, pp DOI /art , American College of Rheumatology MicroRNA-29, a Key Regulator of Collagen Expression in Systemic Sclerosis Britta Maurer, 1 Joanna Stanczyk, 1 Astrid Jüngel, 1 Alfiya Akhmetshina, 2 Michelle Trenkmann, 1 Matthias Brock, 1 Otylia Kowal-Bielecka, 3 Renate E. Gay, 1 Beat A. Michel, 1 Jörg H. W. Distler, 2 Steffen Gay, 1 and Oliver Distler 1 Objective. To investigate the role of microrna (mirna) as posttranscriptional regulators of profibrotic genes in systemic sclerosis (SSc). Methods. MicroRNA, which target collagens, were identified by in silico analysis. Expression of mirna-29 (mir-29) was determined by TaqMan real-time polymerase chain reaction analysis of skin biopsy and fibroblast samples from SSc patients and healthy controls as well as in the mouse model of bleomycininduced skin fibrosis. Cells were transfected with precursor mirna (pre-mirna)/anti-mirna of mir-29 Dr. Maurer s work was supported by the Olga Mayenfisch Foundation. Dr. Stanczyk s work was supported by the Foundation for the Medical Research University of Zurich and the Swiss National Science Foundation. Dr. Jüngel s work was supported by the European Union Sixth Framework Programme (project AutoCure) and Seventh Framework Programme (project Masterswitch). Mr. Brock s work was supported by the Zurich Center of Integrative Human Physiology. Dr. J. H. W. Distler s work was supported by the DFG (grant DI 1537/2-1), the Interdisciplinary Center of Clinical Research in Erlangen (grant A20), and a Career Support Award of Medicine from the Ernst Jung Foundation. Dr. O. Distler s work was supported by a EULAR Orphan Disease Programme grant. 1 Britta Maurer, MD, Joanna Stanczyk, MD, PhD, Astrid Jüngel, PhD, Michelle Trenkmann, MSc, Matthias Brock, MSc, Renate E. Gay, MD, Beat A. Michel, MD, Steffen Gay, MD, Oliver Distler, MD: University Hospital Zurich and Zurich Center of Integrative Human Physiology, Zurich, Switzerland; 2 Alfiya Akhmetshina, PhD, Jörg H. W. Distler, MD: University of Erlangen Nuremberg, Erlangen, Germany; 3 Otylia Kowal-Bielecka, MD: Medical University of Bialystok, Bialystok, Poland. Drs. Maurer, S. Gay, and O. Distler have filed an application for a patent for the use of mir-29 as a novel pharmacologic treatment of scleroderma. Dr. J. H. W. Distler has received consulting fees, speaking fees, and/or honoraria from Actelion, Pfizer, and Glaxo- SmithKline (less than $10,000 each). Dr. O. Distler has received consulting fees, speaking fees, and/or honoraria from Actelion, Pfizer, Encysive, FibroGen, ErgoNex, NicOx, Biovitrum, and Bristol-Myers Squibb (less than $10,000 each). Address correspondence and reprint requests to Oliver Distler, MD, Center of Experimental Rheumatology, University Hospital Zurich, Gloriastrasse 25, 8091 Zurich, Switzerland. oliver. distler@usz.ch. Submitted for publication March 30, 2009; accepted in revised form February 23, using Lipofectamine. Collagen gene expression was also studied in luciferase reporter gene assays. For stimulation, recombinant transforming growth factor (TGF ), platelet-derived growth factor B (PDGF-B), or interleukin-4 (IL-4) was used. The effects of inhibiting PDGF-B and TGF signaling on the levels of mir-29 were studied in vitro and in the bleomycin model. Results. We found that mir-29a was strongly down-regulated in SSc fibroblasts and skin sections as compared with the healthy controls. Overexpression in SSc fibroblasts significantly decreased, and accordingly, knockdown in normal fibroblasts increased, the levels of messenger RNA and protein for type I and type III collagen. In the reporter gene assay, cotransfection with pre-mir-29a significantly decreased the relative luciferase activity, which suggests a direct regulation of collagen by mir-29a. TGF, PDGF-B, or IL-4 reduced the levels of mir-29a in normal fibroblasts to those seen in SSc fibroblasts. Similar to human SSc, the expression of mir-29a was reduced in the bleomycin model of skin fibrosis. Inhibition of PDGF-B and TGF pathways by treatment with imatinib restored the levels of mir-29a in vitro and in the bleomycin model in vivo. Conclusion. These data add the posttranscriptional regulation of collagens by mir-29a as a novel aspect to the fibrogenesis of SSc and suggest mir-29a as a potential therapeutic target. Systemic sclerosis (SSc) is a multisystemic fibrotic disorder with high morbidity and mortality rates (1). The progressive replacement of normal tissue by collagenrich extracellular matrix leads to impairment and, ultimately, to failure of affected organs. Fibroblasts, the cellular key players of this process, are activated by profibrotic cytokines and growth factors, such as interleukin-4 (IL-4), transforming growth factor (TGF ), and platelet-derived growth factor B (PDGF- 1733

2 1734 MAURER ET AL B). Despite the ongoing research in identifying stimuli that contribute to the continuous activation of fibroblasts in SSc, the intracellular molecular events that induce and sustain the status of activation have yet to be elucidated. MicroRNA (mirna) have recently been discovered to be a new class of posttranscriptional repressors of gene expression (2). Up to date, more than 700 of these small, noncoding, evolutionarily conserved RNAs have been experimentally validated in humans. In silico analyses estimate that mirna account for 2 3% of the human genome (3). The term mirna refers to the mature molecule of nucleotides in length, which is generated in a multistep process in the nucleus and in the cytoplasm. The mirna mediate the posttranscriptional regulation of gene expression by binding to partially complementary sites in the 3 -untranslated region (3 -UTR) of target messenger RNA (mrna). The formation of mirna mrna duplexes leads to mrna degradation or translational repression (4). Since mirna are predicted to regulate 30% of the genes in humans (5), dysregulation of mirna seems a likely scenario for the development of diseases such as cancer (6) or neurologic (7), cardiovascular (8), or autoimmune (9) disorders. There is rapidly growing evidence from in vitro studies and studies of animal models that the modification of altered mirna levels with chemically modified anti-mir or mirna mimics may provide a novel, specific, molecular-based therapy (10 13). The role of mirna in the pathogenesis of SSc has not been addressed so far. Thus, the objectives of our present study were to identify differentially expressed mirna in patients with SSc, to investigate the regulation of selected mirna by profibrotic mediators, and to characterize the therapeutic effects of mirna modulation on collagen production. PATIENTS AND METHODS Patients, biopsy specimens, and cell culture. Skin biopsy specimens were obtained by punch biopsy from 12 SSc patients and 11 healthy donors. Skin fibroblasts were expanded by outgrowth culture in Dulbecco s modified Eagle s medium (DMEM; Gibco Invitrogen) as described previously (14). Cells from passages 3 8 were used for the experiments. All patients fulfilled the criteria for SSc as described by LeRoy et al (15), and the study was approved by the local ethics committee. (Detailed characteristics of the 12 SSc patients, as well as additional information concerning the methodology used in these studies, are available upon request from the authors.) Reagents and stimulation assays. After 24 hours in serum-reduced DMEM (0.5% fetal calf serum), cells were stimulated for 24 and 48 hours with TGF (10 ng/ml), PDGF-B (40 ng/ml), or IL-4 (10 ng/ml) (R&D Systems). In subsets of experiments, dermal fibroblasts were incubated for 24 hours with imatinib mesylate that had been dissolved in 0.9% NaCl to a final concentration of 1.0 g/ml (stock concentration 10 mg/ml; Novartis) (14). Myofibroblast differentiation was induced by stimulating SSc fibroblasts with TGF (2 ng/ml) for 4 days (16). Transfection experiments. Cells were transfected with synthetic precursor mirna (pre-mir), with inhibitors of mir-29 (anti-mir), or with negative controls (Pre-miR/AntimiR Negative Control #1; Ambion/Applied Biosystems) at a final concentration of 100 nm with the use of Lipofectamine 2000 reagent (Invitrogen). At 72 hours after transfection, cellular lysates were collected to analyze the expression of types I and III collagen. Transfection efficiency was controlled by TaqMan-based real-time polymerase chain reaction (PCR). Additionally, normal skin fibroblasts were treated with TGF, PDGF-B, or IL-4 for 48 hours after overnight serum starvation, then transfected with pre-mir-29a or negative control, and lysed after 24 hours to measure COL1A1 and COL3A1 mrna. In another subset of experiments, SSc skin fibroblasts were transfected with small interfering RNA (sirna) against c-abl, PDGF receptor (PDGFR ), and c-kit (Qiagen) at a final concentration of 50 nm and lysed after 24 hours to examine the expression of mir-29. Computational prediction analysis of mirna targeting collagen genes. Screening for mirna targeting collagen genes as major components of the secreted extracellular proteins in SSc was performed by merging the results of several computational prediction algorithms provided at the mirgen database ( (17). RNA isolation and quantitative real-time PCR analysis of mirna. Skin biopsy specimens (0.5 cm 2 ) were homogenized with TissueLyser (Qiagen). A mirvana mirna Isolation kit and a RecoverAll Total Nucleic Acid Isolation kit were used for isolation of total RNA (Ambion/Applied Biosystems). Specific single TaqMan mirna assays (Ambion/Applied Biosystems) were used to measure the expression levels of selected mirna in a model 7500 real-time PCR system analyzer (Applied Biosystems). Expression of the U6B small nuclear RNA (RNU6B) or snorna-202 was used as endogenous control to normalize the data for human and murine samples, respectively. For relative quantification, the comparative threshold cycle (C t ) method was used (18). To measure the expression of collagen genes, total RNA was isolated using an RNeasy Mini kit (Qiagen). Gene expression was quantified using SYBR Green real-time PCR, as previously described (14). Western blot analysis for type I 2 and type III collagen. Equivalent amounts of protein from each cell lysate were separated by sodium dodecyl sulfate 10% polyacrylamide gel electrophoresis and electrotransferred onto nitrocellulose membranes, as described previously (14). Polyclonal antibodies against type I 2 and type III collagen (Santa Cruz Biotechnology) and horseradish peroxidase conjugated goat anti-rabbit secondary antibodies (Jackson ImmunoResearch) were used. Semiquantitative analysis based on densitometry was performed using AlphaImager software.

3 mir-29 REGULATION OF COLLAGEN EXPRESSION IN SSc 1735 Luciferase reporter assay for targeting the 3 -UTR of COL3A1. To validate the idea that mir-29a/29b/29c directly regulate the expression of type III collagen by binding to the complementary seed matches within the 3 -UTR of collagen genes, we performed a luciferase reporter gene assay. The human embryonic kidney cell line HEK 293 was used in these experiments, and the results were confirmed with primary dermal fibroblasts. (Data available upon request from the authors.) For the experiment, a COL3A1 3 -UTR segment of 973 bp that contained 2 conserved binding sites for mir-29 was amplified by PCR from human genomic DNA and inserted into the pgl3 control vector (Promega) by using the Xba I site immediately downstream of the stop codon of the luciferase. HEK 293 cells were transfected with the pgl3 control 3 -UTR of the COL3A1 construct (0.15 g/well) using Lipofectamine (Invitrogen). For normalization, a control vector containing Renilla luciferase (prl-sv40 vector, 0.8 g/well; Promega) was added. For each well, a 100 nm concentration of pre-mir- 29a, pre-mir-29b, pre-mir-29c, or scrambled control (Ambion/Applied Biosystems) was used. Firefly and Renilla luciferase activities were measured in the cell lysates by the Dual-Luciferase Reporter Assay system (Promega) 24 hours after transfection. For the 3 -UTR of COL1A1 and COL1A2, conserved seed matches for mir-29 were likewise predicted (19). Bleomycin-induced skin fibrosis in mice. Skin fibrosis was induced in 6-week-old female DBA mice by local intracutaneous injection of bleomycin as described previously (14,20,21). The animal protocol was approved by the local animal care and use committee. Briefly, the mice were challenged every other day for 6 weeks with intracutaneous injections of bleomycin (100 l dissolved in 0.9% NaCl of a 0.5 mg/ml concentration) in defined areas of 1.5 cm 2 on the upper back. One group of mice was treated with intraperitoneal injections of imatinib 150 mg/kg/day for the last 3 weeks to assess the impact of imatinib on mir-29 levels in established Figure 1. Expression of microrna-29 (mir-29) in systemic sclerosis (SSc) patients and healthy controls, as determined by TaqMan-based real-time polymerase chain reaction analysis. Expression of mir-29 family member genes in SSc patient samples was determined relative to the healthy controls, which was defined as 100%. A, All members of the mir-29 family (mir-29a, mir-29b, and mir-29c) were expressed in SSc dermal fibroblasts. Lower C t values indicate higher levels of expression. MicroRNA-29a showed the highest baseline expression (n 8), whereas mir-29b expression was near the detection limit. Horizontal lines show the mean. B, Levels of all mir-29 family members, in particular, mir-29a, were markedly reduced in SSc skin fibroblasts (n 6) as compared with healthy controls (n 6). C, Similar to the in vitro findings, the expression of mir-29a in paraffin-embedded skin sections from SSc patients (n 12) was down-regulated as compared with healthy controls (n 11). D, Expression of mir-29a in fresh skin biopsy samples from SSc patients (n 5) was down-regulated as compared with healthy donors (n 5). Values in B D are the mean and SEM. P 0.05 versus healthy controls.

4 1736 MAURER ET AL fibrosis. Another group was treated with 0.9% NaCl instead of imatinib for the last 3 weeks. The respective control group for bleomycin exposure received intracutaneous injections of 100 l of 0.9% NaCl for 6 weeks instead of bleomycin and additionally received 0.9% NaCl for the last 3 weeks. After 6 weeks, animals were asphyxiated with CO 2. The skin was removed and processed for further analysis. Statistical analysis. GraphPad Prism software was used for statistical analyses. Normal distribution of the data was confirmed using the Kolmogorov-Smirnov test. Student s unpaired t-test was used for unrelated parametric data. Data are expressed as the mean SEM. P values less than 0.05 were considered statistically significant. RESULTS Identifying mirna-targeting collagen genes. Biocomputational prediction algorithms are wellestablished screening tools for mirna profiling (22). Among other mirna, such as the let-7 family, mir-26, and mir-155, the mir-29 family members (mir-29a/ 29b/29c) were identified as potential posttranscriptional regulators of collagen genes. The mir-29 family was selected for further analysis because it was predicted to have the best seed matches with the collagens ( and because it is broadly conserved among most vertebrates (23). MicroRNA-29a and mir- 29b1 are clustered together and are located in chromosome 7 in humans, whereas the mir-29b2/mir-29c cluster is found on chromosome 1 ( mirbase.org). Other predicted profibrotic targets of mir-29a included, for example, PDGF-B, PDGFR, Sp-1, thrombospondin, and secreted protein, acidic and rich in cysteine (SPARC). Down-regulation of baseline mir-29 expression in SSc. Given that most mirna act as posttranscriptional repressors and that in SSc, the expression of collagen genes is remarkably up-regulated, we expected a down-regulation of the mir-29 family in SSc patients as compared with healthy controls. Therefore, we analyzed the expression of the mir-29 family and of type I and type III collagen in skin fibroblasts and biopsy samples from SSc patients and healthy controls. We found that all members of the mir-29 family were expressed in skin fibroblasts from SSc patients, with no differences in disease subsets (limited versus diffuse SSc) or disease stages (early [ 2 years] versus late [ 2 years]). (Data on the expression of mir-29 in different disease subsets and stages of SSc are available upon request from the authors.) Within the mir-29 family, mir-29a showed the highest level of expression (mean SEM C t ), followed by mir-29c ( C t ), and mir-29b Figure 2. Expression of type I and type III collagen in systemic sclerosis (SSc) patients and healthy controls, as determined by SYBR Green real-time polymerase chain reaction analysis. Whereas expression of the microrna-29 family members was found to be upregulated in SSc patients, the expression of type I and type III collagen in these same patients was increased as compared with healthy controls. Expression of collagen genes in SSc patient samples was determined relative to the healthy controls, which was defined as 1. A, Levels of mrna for COL1A1 and COL3A1 were increased in SSc fibroblasts (n 6) as compared with normal skin fibroblasts (n 6). B, Levels of mrna for COL1A1 and COL3A1 were also increased in fresh skin biopsy samples from SSc patients (n 5) as compared with healthy controls (n 5). Values are the mean and SEM. P 0.05 versus healthy controls. ( C t ) or mir-29c ( C t ) (Figure 1A). However, compared with normal skin fibroblasts, the baseline levels of mir-29 were consistently downregulated in SSc fibroblasts: by a mean SEM of 60 15% for mir-29a (P 0.004), 45 15% for mir-29b (P 0.001), and 58 7% for mir-29c (P ) (Figure 1B). In the same SSc fibroblasts, this was paralleled by increased levels of COL1A1 and COL3A1 mrna (mean SEM fold and fold, respectively; P 0.06) in SSc fibroblasts as compared with healthy control fibroblasts (Figure 2A). Next, we investigated the expression of the mir-29 family in skin samples from SSc patients. First, the expression of mir-29 in tissue extracts isolated from paraffin-embedded skin sections was analyzed, and similar to the findings in SSc fibroblasts, decreased levels of the mir-29 family members were observed. In particular, mir-29a was down-regulated in the dermis of SSc patients as compared with healthy controls (by 79 9%; P ) (Figure 1C). To exclude the possibility that paraffin processing influenced the content of mirna in the tissue, we additionally investigated the expression of the mir-29 family in fresh skin biopsy samples. Consistent with the

5 mir-29 REGULATION OF COLLAGEN EXPRESSION IN SSc 1737 Figure 3. Influence of overexpression and knockdown of microrna-29a (mir-29a) on the expression of collagen. Expression of type I and type III collagen in systemic sclerosis (SSc) patient samples was determined relative to the controls transfected with scrambled RNA, which was defined as 100% or as 1. A, At the mrna level, transfection of SSc fibroblasts (n 5) with precursor mir-29a (pre-mir-29a) as compared with scrambled RNA transfected controls (n 5) decreased the levels of type I and type III collagen, as determined by SYBR Green real-time polymerase chain reaction (PCR) analysis. B, At the protein level, overexpression of mir-29a in SSc fibroblasts (n 6) also decreased the expression of type I 2 and type III collagen as compared with scrambled RNA transfected controls (n 6), as determined by Western blotting (top). Representative duplicate results are shown. The semiquantitative analysis was performed by densitometry (bottom). C, At the mrna level, knockdown of mir-29a in fibroblasts from healthy donors (n 5) as compared with scrambled RNA transfected controls (n 5) increased the levels of type I and type III collagen, as determined by SYBR Green real-time PCR. D, At the protein level, knockdown of mir-29a in normal skin fibroblasts (n 3) similarly increased the expression of type I 2 and type III collagen as compared with scrambled RNA transfected controls (n 5), as determined by Western blotting (top). Representative duplicate results are shown. The semiquantitative analysis was performed by densitometry (bottom). Values are the mean and SEM. In the Western blots, -tubulin was used as a loading control. P 0.05 versus scrambled RNA transfected controls. previous results, mir-29a was found to be downregulated by 42 8% (P ) in fresh skin biopsy samples from SSc patients as compared with healthy controls (Figure 1D). In the same SSc skin biopsy samples, this was again paralleled by increased levels of COL1A1 and COL3A1 mrna ( fold [P 0.09] and fold [P 0.03], respectively) as compared with healthy control skin in the same samples (Figure 2B). Taken together, the mir-29 family members and mir-29a in particular were consistently expressed at a remarkable down-regulation in skin fibroblasts and skin tissue sections from SSc patients as compared with healthy controls. Altered collagen expression following overexpression and knockdown of mir-29. We next aimed to analyze the functional effects of altered mir-29 expression on fibrogenesis. According to our hypothesis, the restoration of mir-29 levels would decrease collagen levels in SSc fibroblasts, whereas down-regulation of

6 1738 MAURER ET AL Figure 4. Direct regulation of collagen expression by members of the microrna-29 (mir-29) family. Luciferase activity in HEK 293 cells transfected with precursor mir-29 (pre-mir-29) or with anti-mir-29 was determined relative to controls transfected with scrambled RNA, which was defined as 100%. A, HEK 293 cells (n 8) were transfected with the pgl3 luciferase reporter containing a segment of the 3 -untranslated region (3 -UTR) sequence of COL3A1 that was predicted to have 2 harbor binding sites for the mir-29 family: and B, Cotransfection of HEK 293 cells with pre-mir-29a, pre-mir-29b, or pre-mir-29c, which mimics up-regulation of the respective mirna, significantly reduced the relative luciferase activity as compared with the scrambled RNA transfected controls (n 8). C, Cotransfection of HEK 293 cells with the pgl3 luciferase vector and inhibitory molecules of mir-29a, mir-29b, or mir-29c (n 6) significantly increased the relative luciferase activity as compared with the scrambled RNA transfected controls (n 8). All experiments were performed at least in duplicate. Values are the mean and SEM. P 0.05 versus scrambled RNA transfected controls. mir-29 in normal skin fibroblasts would induce an SSc-like phenotype, with an increase in the expression of collagen. In SSc fibroblasts, transfection with pre-mir-29a/ 29b/29c increased the levels of the respective mirna by a mean SEM of fold, 1,625 1,271 fold, and fold. Transfection with the respective pre-mir did not substantially alter the levels of the other mir-29 family members. Knockdown with antimir reduced the expression of the respective mirna to a mean SEM of fold for mir-29a, fold for mir-29b, and fold for mir-29c. Except for knockdown of mir-29a, which also affected the levels of mir-29b, the effects were specific. In all gain-and-loss of function experiments, mir- 29a, as compared with mir-29b/29c, showed the strongest effects. The expression of types I and III collagen decreased by a mean SEM of 66 10% (P ) and 65 8% (P ), respectively, on the mrna level, as analyzed by TaqMan-based real-time PCR (Figure 3A), and by 37 5% (P ) and 45 9% (P ), respectively, on the protein level, as analyzed by Western blotting (Figure 3B). Knockdown of mir-29a in normal skin fibroblasts increased the expression of types I and III collagen by fold and fold, respectively, on the mrna level (Figure 3C) and by fold (P 0.03) and fold, respectively, on the protein level (Figure 3D). Overexpression of mir-29b in SSc fibroblasts down-regulated the expression of mrna for type I and type III collagen by 35 9% (P 0.01) and 40 15% (P 0.03), respectively, and reduced the level of type I 2 collagen protein by 44 22% (P 0.05). The expression of type III collagen protein did not change significantly. Similar to the results with mir-29a, knockdown of mir-29b in healthy dermal fibroblasts increased the expression of types I and III collagen mrna, but the effects at the protein level were less distinct. (Data showing the effects of mir-29b on the expression of collagen are available upon request from the authors.)

7 mir-29 REGULATION OF COLLAGEN EXPRESSION IN SSc 1739 Figure 5. MicroRNA-29a (mir-29a), a downstream mediator of profibrotic molecules. Expression of mir-29a, mir-29b, or mir-29c in normal or systemic sclerosis (SSc) skin fibroblasts was determined relative to the respective controls, which was defined as 100% or 1. A and B, Normal skin fibroblasts obtained from healthy donors were treated with recombinant transforming growth factor (TGF ), platelet-derived growth factor B (PDGF-B), or interleukin-4 (IL-4) to explore the effects on mir-29 expression, as measured by TaqMan-based real-time polymerase chain reaction. Stimulation of normal skin fibroblasts (n 6) reduced the expression of mir-29a as compared with untreated controls (n 6) (A). Compared with unstimulated normal skin fibroblasts, treatment with TGF, PDGF-B, or IL-4 increased the expression of mrna for COL1A1 and COL3A1. However, overexpression of mir-29a by using precursor mir-29a (pre-mir-29a) in cytokine-stimulated normal skin fibroblasts decreased the expression of mrna for COL1A1 and COL3A1 as compared with stimulated and scrambled small interfering RNA (sirna) transfected cells (n 5 for each condition) (B). C and D, In skin fibroblasts from SSc patients, PDGF-B and TGF signaling pathways were inhibited, and the effects on the expression of mir-29 were analyzed. Treatment of SSc fibroblasts (n 6) with imatinib increased the expression of mir-29 family members as compared with untreated controls (n 6) (C). SSc skin fibroblasts were transfected with sirna against c-abl, PDGF receptor (PDGFR ), or c-kit to explore potential mechanisms of how imatinib restores the levels of mir-29a. Inhibition of c-abl and PDGFR, but not c-kit signaling, increased the expression of mir-29a in SSc skin fibroblasts (n 4) as compared with scrambled sirna-transfected controls (n 4) (D). Enforced expression of mir-29c in SSc fibroblasts decreased type I collagen mrna by 62 14% (P 0.001) and type III collagen mrna by 61 14% (P 0.002), whereas on the protein level, expression of types I 2 and III collagen was reduced by 14 7% (P 0.05) and 30 5% (P ), respectively. The effects of mir-29c knockdown in healthy control fibroblasts on the expression of types I and III collagen were similar to the effects of the other mir-29 family members examined. (Data showing the effects of overexpression and knockdown of mir-29c on collagen expression are available upon request from the authors.) COL3A1, a direct target of mir-29. The observed reduction in collagen expression by mir-29 family members could be caused indirectly, for example, by a modification of profibrotic cytokines and growth factors, or directly, by direct posttranscriptional regulation of collagen gene expression. To assess whether there was direct regulation of collagen genes, we created a luciferase reporter gene system by cloning a part of the 3 -UTR of COL3A1 with respective binding sites for mir-29a, mir-29b, and mir-29c into the pgl3 control vector (Figure 4A). All mir-29 family members were expressed in HEK 293 cells, but at lower baseline levels as compared with the expression in skin fibroblasts (mean SEM C t for mir-29a, for mir-29b, and for mir-29c). Transfection of HEK 293

8 1740 MAURER ET AL Figure 6. MicroRNA-29 (mir-29) in the murine model of bleomycininduced skin fibrosis and in vivo effects of imatinib treatment, as determined by TaqMan-based real-time polymerase chain reaction analysis. Expression of mir-29 family members in mice treated with bleomycin (bleo) only or with bleomycin plus imatinib (ima) was determined relative to the saline-treated controls, which was defined as 100% or 1. A, Mice were challenged with bleomycin or saline for 6 weeks (n 10 mice per group). Similar to the findings in systemic sclerosis patients, the expression of mir-29a, mir-29b, and mir-29c was significantly down-regulated in bleomycin-treated mice as compared with saline-treated controls. B, All mice were challenged with bleomycin for 6 weeks. Then, one group received imatinib 150 mg/kg/day during the last 3 weeks, and the control group received NaCl during the last 3 weeks (n 8 mice per group). This is a biologically relevant concentration that corresponds to mean plasma peak and trough concentrations in humans. The expression of mir-29 family members was increased in imatinib-treated mice as compared with control mice (only challenged with bleomycin). Values are the mean and SEM. P 0.05 versus saline-treated controls. cells with pre-mir was specific and had no substantial effects on the other family members. In HEK 293 cells, transfection with pre-mir-29a/29b/29c increased the levels of the respective mirna by a mean SEM of fold, 5,517 1,597 fold, and fold. Knockdown of the mir-29 decreased their expression to fold for mir-29a, fold for mir-29b, and fold for mir-29c. Except for knockdown of mir-29c, which also affected the levels of mir-29b and, to a lesser extent, mir-29a, the effects were specific. In the reporter gene assay, cotransfection of HEK 293 cells with pre-mir-29a and the luciferase reporter plasmid containing the 3 -UTR of COL3A1 (Figure 4A) decreased the relative luciferase activity by a mean SEM of 47 10% (P 0.007) compared with cells transfected with the plasmid and scrambled controls (Figure 4B). Accordingly, cotransfection with antimir-29a increased the relative enzyme activity by 20 6% (P 0.01) (Figure 4C). Similar effects were observed for mir-29b and mir-29c. Cotransfection with pre-mir-29b reduced the relative luciferase activity by 67 10% (P ) (Figure 4B), and cotransfection with anti-mir-29b increased the relative enzyme activity by 30 7% (P 0.008) (Figure 4C). Cotransfection with pre-mir-29c decreased the relative luciferase activity by 70 6% (P ) (Figure 4B), whereas cotransfection with anti-mir-29c did not show significant effects (Figure 4C). In summary, these findings prove that mir-29 directly regulates collagen production. MicroRNA-29, a downstream mediator of profibrotic molecules. To explore factors that potentially down-regulate mir-29 levels in SSc, we analyzed the effects of major profibrotic cytokines and growth factors. Treatment of normal skin fibroblasts with profibrotic cytokines induced an SSc-like phenotype, with downregulation of mir-29a to an extent similar to that seen in SSc fibroblasts. The most pronounced effects were observed for mir-29a after 48 hours of treatment, with a reduction in the levels of mrna for TGF by 80 8% (P ), for PDGF-B by 70 20% (P 0.003), and for IL-4 by 74 11% (P ) (Figure 5A). The levels of mir-29b and mir-29c were also reduced, but to a lesser extent. (Data showing mir-29b and mir-29c as downstream mediators of profibrotic molecules are available upon request from the authors.) To confirm mir-29a as a downstream mediator of the profibrotic effects of these cytokines, we investigated whether rescuing cytokine-treated normal fibroblasts by overexpression of mir-29a could decrease the levels of collagens induced by these cytokines. Stimulation with TGF, PDGF-B, or IL-4 increased the levels of COL1A1 and COL3A1 mrna in nonrescued scrambled controls, but this effect was abolished by overexpression of mir-29a (Figure 5B). Thus, it seems likely that mir-29 is indeed a downstream mediator of the profibrotic effects of the cytokines we examined. We next analyzed whether treatment with TGF, PDGF-B, or IL-4 could further decrease the reduced levels of mir-29 in SSc fibroblasts. In contrast to healthy fibroblasts, no additional decrease in mir-29a/29b/29c levels was observed. Accordingly, transdifferentiation of SSc fibroblasts into myofibroblasts by long-term stimulation with TGF for 4 days did not further reduce the levels of mir-29 (data not shown). The recombinant proteins used in these experiments were biologically active, as proven by the strong decrease in mir-29 in healthy fibroblasts and by an induction in downstream molecules, such as type I collagen, in both healthy and

9 mir-29 REGULATION OF COLLAGEN EXPRESSION IN SSc 1741 SSc fibroblasts. A potential explanation for this finding was that mir-29 levels in SSc fibroblasts were already maximally suppressed and therefore could not be further decreased by profibrotic cytokines. To further analyze this hypothesis, SSc fibroblasts were incubated with the tyrosine-kinase inhibitor imatinib, which abrogates TGF and PDGF-B signaling (14,21). Indeed, treatment of SSc fibroblasts with imatinib increased the levels of mir-29a by a mean SEM of 40 7% (P 0.002), mir-29b by 60 3% (P 0.05), and mir-29c by 30 6% (Figure 5C). To further investigate which of the imatinib targets mediates the observed effects, we performed selective knockdown experiments for c-abl, PDGFR, and c-kit in SSc fibroblasts and analyzed the levels of mir-29. Inhibition of c-abl and PDGFR, but not c-kit, increased the levels of mir-29a by a mean SEM of fold and fold (P 0.03), respectively, as compared with the scrambled controls (Figure 5D). Similar effects were observed for mir-29c and for mir-29b (data available upon request from the authors). These results suggest abrogation of PDGFR and c-abl signaling as the main mechanism of how imatinib restores the levels of mir-29 in SSc fibroblasts. MicroRNA-29 in the murine model of bleomycininduced skin fibrosis. Finally, we wanted to investigate whether the reduced levels of mir-29 in human SSc are mimicked in the SSc model of bleomycin-induced skin fibrosis in mice. Similar to the findings in human SSc (Figure 1B), the expression of all mir-29 family members in the skin of mice with bleomycin-induced fibrosis was down-regulated after 6 weeks of bleomycin challenge: mir-29a was reduced by a mean SEM of 34 19% (P 0.02), mir-29b by 50 27% (P 0.01), and mir-29c by 69 11% (P ) (Figure 6A). We next used imatinib to analyze the effects of impaired PDGF-B and TGF signaling on the levels of mir-29 in vivo. To mimic the clinical situation, we used a modified bleomycin model in which fibrosis is established before treatment with imatinib is started (24). Remarkably, imatinib treatment of mice with bleomycininduced fibrosis increased the expression of mir-29a by a mean SEM of fold (P 0.003) as compared with NaCl control treatment. The levels of the other mir-29 family members showed a similar, but not significant, up-regulation (Figure 6B). These results suggest that even in preestablished fibrosis, abrogation of PDGFR and c-abl/tgf signaling could increase the levels of mir-29a. As compared with normal controls (i.e., mice treated with saline instead of bleomycin for 6 weeks and then treated with NaCl instead of imatinib during the last 3 weeks), the levels of mir-29a, but not mir-29b or mir-29c, were normalized (data not shown). DISCUSSION The direct posttranscriptional regulation of collagens by mir-29 adds a novel aspect to the complex network of factors that regulate the expression of collagen genes in SSc at the level of transcription. This includes deregulated stimulatory (e.g., Sp-1, Smad3, p300/creb binding protein) and inhibitory transcription factors and cofactors (e.g., Smad7, Fli-1, peroxisome proliferator activated receptor ), as well as several stimulatory cytokines and growth factors (25). The relative importance of these factors in the pathogenesis of SSc is unclear. Similarly, there is limited knowledge about the extent to which down-regulated repressors and up-regulated stimulators of collagen transcription interplay. Of note, mir-29a as a posttranscriptional repressor, which is induced by the stimulators mentioned above, acts downstream of most of the previously identified profibrotic molecules. Thus, the main hypothesis derived from the present study is that targeting mir-29 family members as posttranscriptional regulators could be a potent antifibrotic approach. This needs to be tested in future studies. Specifically, the effects of the modification of mir-29 levels on the synthesis of extracellular matrix should be analyzed in several animal models of SSc covering different aspects of the disease. Optimally, mir-29 knockout mice could be created and potentially crossed with genetic fibrotic animal models or challenged with profibrotic stimuli. If the findings of these experiments turned out to be promising, modification of mir-29 levels could be adapted to humans and the first clinical proof-of-concept studies and studies on toxicity could be performed. So far, mirna-based therapeutic approaches in animal studies of other diseases have been promising, since toxic effects were not observed (10 13). Accordingly, our experiments with the restoration of mir-29 in skin fibroblasts did not have overt toxic effects on cell survival and proliferation (data not shown). Our results also demonstrated that the bleomycin model of skin fibrosis, although it does not cover all of the features of human SSc, represents one of the potential in vivo models in which to test the different strategies of mir-29 restoration in SSc because it displayed the reduced levels of mir-29 that are seen in humans with SSc. Abrogation of PDGFR and, eventually, TGF signaling by admin-

10 1742 MAURER ET AL istration of imatinib increased the levels of mir-29 in vivo. The direct therapeutic modulation of mir-29 might have even more pronounced effects. Whether imatinib also restores mir-29 levels in human SSc should be analyzed in the ongoing imatinib trials. Interestingly, mir-29 might additionally influence the production of other profibrotic molecules, such as PDGF-B, PDGFR, thrombospondin, and SPARC, since in silico analyses have shown these molecules to be predicted targets of mir-29. The ability of mir-29 to coordinate such a broad spectrum of profibrotic molecules illustrates the potential of mirna to control complex molecular interactions. Further research should include the analysis of mirna expression profiles in SSc biologic samples, as well as the identification of mirna that regulate other key molecules, such as growth factors, profibrotic cytokines, and matrix metalloproteinases. The precise mechanisms that lead to the downregulation of mir-29, especially mir-29a, in SSc remain to be determined, but our study showed that TGF, PDGF-B, and IL-4, all of which are major profibrotic mediators in SSc, decreased the expression of mir-29a in normal skin fibroblasts. Moreover, the stimulatory effects of these cytokines on collagen synthesis could be reduced by rescuing cells with mir-29a, suggesting that the reduced levels of mir-29a occur downstream of these cytokines and that mir-29 at least partially mediates their profibrotic effects. Accordingly, inhibition of TGF and PDGF-B signaling by imatinib or, more specifically, by knockdown of c-abl or PDGFR restored the decreased levels of mir-29 in SSc fibroblasts. Since these major profibrotic factors are not specific for SSc, the mir-29 family could also play a role in other fibrotic diseases. Indeed, there is evidence that mir-29b plays a role in cardiac fibrosis (26), and mir-29c was recently found to be involved in the mechanisms of fibrosis during the development of nasopharyngeal carcinomas (24). Additional factors potentially involved in the down-regulation of mir-29 in SSc fibroblasts may include intrinsic mechanisms, because the low expression of mir-29 family members was maintained from early to late cell culture (passages 3 8) without additional profibrotic stimuli and because stimulation with the profibrotic cytokines did not further decrease the levels of mir-29 in SSc fibroblasts. Epigenetic dysregulation of gene expression may also be part of this scenario. Fli-1, a negative regulator of collagen synthesis, has been proven to be epigenetically silenced in SSc fibroblasts by DNA methylation and histone deacetylation (27). Changes in the expression of histone deacetylases have also been found to play a role in the profibrotic mechanisms of SSc (28,29). Thus, epigenetic silencing of mir-29 might also contribute to the activated state of SSc fibroblasts. Recently, a study of skeletal myogenesis and the pathogenesis of rhabdomyosarcoma demonstrated that in rhabdomyosarcoma cells and primary tumors with impaired differentiation, mir-29 was epigenetically silenced by an activated NF- B/YY-1 pathway (30). On the other hand, DNA methyl transferases are experimentally validated targets of mir-29 (31), and thus, it might also be hypothesized that mir-29 plays a role in the epigenetic regulation of profibrotic genes in SSc. So far, the available data on mir-29 indicate a general role in the development of fibrosis not limited to SSc. To evaluate in SSc whether low mir-29 levels correlate with individual profibrotic activity, as assessed by the modified Rodnan skin thickness score (MRSS) or by disease stage (early versus late), longitudinal, rather than cross-sectional, studies with larger cohorts of patients will be needed, since in our study, patients with high MRSS scores and/or early disease were underrepresented and, in general, the patients were only assessed at a single time point. In addition, it seems that there is some organ specificity in terms of which member of the mir-29 family has a predominant action. This fits very well with the findings of previous studies demonstrating that fibroblasts from different anatomic sites showed remarkable differences in their patterns of gene expression (32). In conclusion, our study shows that in SSc mir- 29a directly regulates collagen expression at the posttranscriptional level, which is a novel aspect of the complex regulatory network of fibrosis in SSc. Furthermore, our data indicate that the dysregulation of mir- 29a is mediated, at least in part, by profibrotic cytokines and growth factors. Based on the encouraging results from mirna-based therapeutic approaches in animal studies (10 12,33) and given the fact that the first human trial targeting a liver-specific mirna involved in the replication of the hepatitis C virus has been launched (34), the development of strategies to maintain the expression or to prevent the repression of mir-29a in SSc appears to be a promising future therapeutic strategy. ACKNOWLEDGMENTS We thank Ferenc Pataky and Peter Künzler for excellent technical support.

11 mir-29 REGULATION OF COLLAGEN EXPRESSION IN SSc 1743 AUTHOR CONTRIBUTIONS 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. O. Distler 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. Maurer, Stanczyk, J. H. W. Distler, S. Gay, O. Distler. Acquisition of data. Maurer, Akhmetshina, Trenkmann, Brock. Analysis and interpretation of data. Maurer, Stanczyk, Jüngel, Akhmetshina, Kowal-Bielecka, R. E. Gay, Michel, S. Gay, O. Distler. REFERENCES 1. Mayes MD, Lacey JV Jr, Beebe-Dimmer J, Gillespie BW, Cooper B, Laing TJ, et al. Prevalence, incidence, survival, and disease characteristics of systemic sclerosis in a large US population. Arthritis Rheum 2003;48: Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie AE, et al. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 2000;403: Berezikov E, Plasterk RH. Camels and zebrafish, viruses and cancer: a microrna update. Hum Mol Genet 2005;14:R Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 2004;116: Lewis BP, Shih IH, Jones-Rhoades MW, Bartel DP, Burge CB. Prediction of mammalian microrna targets. Cell 2003;115: Wiemer EA. The role of micrornas in cancer: no small matter. Eur J Cancer 2007;43: Lukiw WJ, Zhao Y, Cui JG. An NF- B-sensitive micro RNA- 146a-mediated inflammatory circuit in Alzheimer disease and in stressed human brain cells. J Biol Chem 2008;283: Care A, Catalucci D, Felicetti F, Bonci D, Addario A, Gallo P, et al. MicroRNA-133 controls cardiac hypertrophy. Nat Med 2007; 13: Stanczyk J, Pedrioli DM, Brentano F, Sanchez-Pernaute O, Kolling C, Gay RE, et al. Altered expression of microrna in synovial fibroblasts and synovial tissue in rheumatoid arthritis. Arthritis Rheum 2008;58: Elmen J, Lindow M, Schutz S, Lawrence M, Petri A, Obad S, et al. LNA-mediated microrna silencing in non-human primates. Nature 2008;452: Elmen J, Lindow M, Silahtaroglu A, Bak M, Christensen M, Lind-Thomsen A, et al. Antagonism of microrna-122 in mice by systemically administered LNA-antimiR leads to up-regulation of a large set of predicted target mrnas in the liver. Nucleic Acids Res 2008;36: Esau C, Davis S, Murray SF, Yu XX, Pandey SK, Pear M, et al. mir-122 regulation of lipid metabolism revealed by in vivo antisense targeting. Cell Metab 2006;3: Krutzfeld J, Rajewsky N, Braich R, Rajeev KG, Tuschl T, Manoharan M, et al. Silencing of micrornas in vivo with antagomirs. Nature 2005;438: Distler JH, Jungel A, Huber LC, Schulze-Horsel U, Zwerina J, Gay RE, et al. Imatinib mesylate reduces production of extracellular matrix and prevents development of experimental dermal fibrosis. Arthritis Rheum 2007;56: LeRoy EC, Black C, Fleischmajer R, Jablonska S, Krieg T, Medsger TA Jr, et al. Scleroderma (systemic sclerosis): classification, subsets, and pathogenesis. J Rheumatol 1988;15: Rajkumar VS, Shiwen X, Bostrom M, Leoni P, Muddle J, Ivarsson M, et al. Platelet-derived growth factor- receptor activation is essential for fibroblast and pericyte recruitment during cutaneous wound healing. Am J Pathol 2006;169: Megraw M, Sethupathy P, Corda B, Hatzigeorgiou AG. mirgen: a database for the study of animal microrna genomic organization and function. Nucl Acids Res 2007;35:D Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C t method. Nat Protoc 2008;3: John B, Enright AJ, Aravin A, Tuschl T, Sander C, Marks DS. Human microrna targets. PLoS Biol 2004;2:e Yamamoto T, Nishioka K. Cellular and molecular mechanisms of bleomycine-induced murine scleroderma: current update and future perspective. J Exp Dermatol 2005;14: Akhmetshina A, Venalis P, Dees C, Busch N, Zwerina J, Schett G, et al. Treatment with imatinib prevents fibrosis in different preclinical models of systemic sclerosis and induces regression of established fibrosis. Arthritis Rheum 2009;60: Sethupathy P, Megraw M, Hatzigeorgiou AG. A guide through present computational approaches for the identification of mammalian microrna targets. Nat Methods 2006;3: Landgraf P, Rusu M, Sheridan R, Sewer A, Iovino N, Aravin A, et al. A mammalian microrna expression atlas based on small RNA library sequencing. Cell 2007;129: Sengupta S, den Boon JA, Chen IH, Newton MA, Stanhope SA, Cheng YJ, et al. MicroRNA 29c is down-regulated in nasopharyngeal carcinomas, up-regulating mrnas encoding extracellular matrix proteins. Proc Natl Acad Sci U S A 2008;105: Varga J, Abraham D. Systemic sclerosis: a prototypic multisystem fibrotic disorder. J Clin Invest 2007;117: Van Rooij E, Sutherland LB, Thatcher JE, DiMaio JM, Naseem RH, Marshall WS, et al. Dysregulation of micrornas after myocardial infarction reveals a role of mir-29 in cardiac fibrosis. Proc Natl Acad Sci U S A 2008;105: Wang Y, Fan PS, Kahaleh B. Association between enhanced type I collagen expression and epigenetic repression of the FLI1 gene in scleroderma fibroblasts. Arthritis Rheum 2006;54: Huber LC, Distler JH, Moritz F, Hemmatazad H, Hauser T, Michel BA, et al. Trichostatin A prevents the accumulation of extracellular matrix in a mouse model of bleomycin-induced skin fibrosis. Arthritis Rheum 2007;56: Hemmatazad H, Maciejewska Rodrigues H, Maurer B, Brentano F, Pileckyte M, Distler JH, et al. Histone deacetylase 7, a potential target for the antifibrotic treatment of systemic sclerosis. Arthritis Rheum 2009;60: Wang H, Garzon R, Sun H, Ladner KJ, Singh R, Dahlman J, et al. NF- B YY1 mir-29 regulatory circuitry in skeletal myogenesis and rhabdomyosarcoma. Cancer Cell 2008;14: Fabbri M, Garzon R, Cimmino A, Liu Z, Zanesi N, Callegari E, et al. MicroRNA-29 family reverts aberrant methylation in lung cancer by targeting DNA methyltransferases 3A and 3B. Proc Natl Acad Sci U S A 2007;104: Chang HY, Chi JT, Dudoit S, Bondre C, van de Rijn M, Botstein D, et al. Diversity, topographic differentiation, and positional memory in human fibroblasts. Proc Natl Acad Sci U S A 2002;99: Esau CC, Monia BP. Therapeutic potential for micrornas. Adv Drug Deliv Rev 2007;59: Jopling CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus RNA abundance by a liver-specific microrna. Science 2005;309: