Smad1 pathway is activated in systemic sclerosis fibroblasts and is targeted by imatinib mesylate.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 58, No. 8, August 2008, pp 2528–2537 DOI 10.1002/art.23698 © 2008, American College of Rheumatology Smad1 Pathway Is Activated in Systemic Sclerosis Fibroblasts and Is Targeted by Imatinib Mesylate Jaspreet Pannu,1 Yoshihide Asano,1 Sashidhar Nakerakanti,1 Edwin Smith,1 Stefania Jablonska,2 Maria Blaszczyk,2 Peter ten Dijke,3 and Maria Trojanowska1 normalized the production of CCN2 and collagen. Imatinib mesylate blocked activation of the Smad1 pathway in transforming growth factor ␤–stimulated control fibroblasts and reversed activation of this pathway in SSc fibroblasts. Likewise, blockade of c-Abl abrogated activation of the Smad1 pathway in SSc fibroblasts. Conclusion. Our findings demonstrate that activation of Smad1 signaling occurs in a subset of SSc patients and contributes to persistent activation of SSc fibroblasts. Demonstration that the Smad1/CCN2 pathway is blocked by imatinib mesylate further clarifies the mechanism of the antifibrotic effects of this compound. This study suggests that SSc patients with activated Smad1 signaling may benefit from imatinib mesylate treatment. Objective. Activation of Smad1 signaling has recently been implicated in the development of fibrosis. The goal of the present study was to gain further insights into activation of the Smad1 pathway in fibrosis in systemic sclerosis (SSc) and to determine whether this pathway is targeted by the antifibrotic drug imatinib mesylate. Methods. Levels of phosphorylated Smad1 and total Smad1 were examined in SSc and control skin biopsy samples by immunohistochemistry and in cultured fibroblasts by Western blotting. Activity of the CCN2 promoter was examined by a luciferase reporter gene assay. Interactions of Smad1 with the CCN2 promoter were examined by in vitro and in vivo DNA binding assays. Expression of the nonreceptor tyrosine kinase c-Abl and Smad1 was blocked using respective small interfering RNA. Results. Total and phosphorylated Smad1 levels were significantly elevated in SSc skin biopsy samples and in cultured SSc fibroblasts and correlated with elevated CCN2 protein and CCN2 promoter activity. DNA binding assays demonstrated that Smad1 was a direct activator of the CCN2 gene. Small interfering RNA–mediated depletion of Smad1 in SSc fibroblasts Fibrosis, affecting skin, lungs, and other organs, is the most prominent feature of systemic sclerosis (SSc). While many factors are known to contribute to fibrosis, activation of the transforming growth factor ␤ (TGF␤) signaling pathway is considered to play a central role in this process (1). TGF␤ is a pleiotropic cytokine with diverse functions in many cell types (2). The Smad pathway is a primary mediator of TGF␤ signaling (3). The canonical Smad pathway is activated upon binding of TGF␤ to the heteromeric serine/threonine kinase receptors TGF␤ receptor type II (TGF␤RII) and TGF␤RI (activin receptor–like kinase 5 [ALK-5]), which leads to phosphorylation of Smads 2 and 3, oligomerization with Smad4, nuclear translocation, and activation of various transcription programs (2). There is also increasing evidence for the activation of other (noncanonical) signaling pathways downstream of TGF␤, but the specific mechanisms involved in activation of Smadindependent pathways have not been fully elucidated (4). With respect to fibrosis, the nonreceptor tyrosine kinase c-Abl is of special interest, because of the potent Dr. Pannu’s work was supported by the National Scleroderma Foundation (New Investigator grant). Dr. ten Dijke’s work was supported by the Dutch Cancer Society. Dr. Trojanowska’s work was supported by the NIH (grants AR-42334 and AR-44883). 1 Jaspreet Pannu, PhD (current address: University of Maryland School of Medicine, Baltimore), Yoshihide Asano, MD, PhD, Sashidhar Nakerakanti, PhD, Edwin Smith, MD, Maria Trojanowska, PhD: Medical University of South Carolina, Charleston; 2Stefania Jablonska, MD, Maria Blaszczyk, MD: Warsaw Medical Academy, Warsaw, Poland; 3Peter ten Dijke, PhD: Leiden University Medical Center, Leiden, The Netherlands. Address correspondence and reprint requests to Maria Trojanowska, PhD, Division of Rheumatology and Immunology, Medical University of South Carolina, 96 Jonathan Lucas Street, Suite 912, Mail Stop Code 637, Charleston, SC 29425-6370. E-mail: trojanme@ musc.edu. Submitted for publication August 8, 2007; accepted in revised form April 18, 2008. 2528 Smad1 SIGNALING PATHWAY IN SSc antifibrotic effects of the pharmacologic inhibitor of c-Abl, imatinib mesylate. Imatinib mesylate has been shown to inhibit profibrotic effects of TGF␤ in cultured cells, including SSc fibroblasts (5), and to prevent lung and kidney fibrosis in experimental mouse models (6,7). TGF␤ activation of c-Abl occurs through phosphatidylinositol 3-kinase and p21-activated kinase 2 and is independent of the Smad2/3 pathway (8). The role of TGF␤ signaling in the process of fibrosis in SSc has been mainly studied using fibroblasts explanted from SSc skin. During early passages in culture, these cells demonstrate an “activated phenotype” characterized by elevated synthesis of extracellular matrix (ECM) proteins (9). Accumulating evidence suggests that alterations of TGF␤ signaling may play an essential role in the persistent activation of SSc fibroblasts. The documented changes include up-regulation of ␣v␤5 and ␣v␤3 integrins, which were shown to contribute to activation of latent TGF␤ and establishment of the autocrine TGF␤ loop (10,11). Additional changes include alterations of the TGF␤ receptor ratio (12) and the presence of persistently phosphorylated Smad3 (13). Recent studies from our laboratory using an in vitro model of SSc based on the altered ratio of TGF␤ receptors have also established that activation of the Smad1 pathway may represent a novel aspect of profibrotic TGF␤ signaling that functions independently of activation of the canonical Smad2/3 pathway (14). Activation of Smad1 downstream of TGF␤ was first described in endothelial cells, where TGF␤ signals through 2 distinct type I receptors, ALK-5 and ALK-1, and their respective signal transducers (Smads 2 and 3 for ALK-5 and Smads 1 and 5 for ALK-1) (15). Based on our recent study, it appears that this mode of signaling is also operational in fibroblasts, but the contribution of Smad1 signaling to SSc fibrosis has not been examined. The present study was undertaken to further evaluate the role of Smad1 signaling in SSc fibrosis. Given that the inhibition of the TGF␤-mediated fibrotic response by imatinib mesylate is independent of Smads 2 and 3, we also sought to determine whether Smad1 is a downstream target of imatinib mesylate. Our results show that Smad1 signaling is activated in a subset of SSc fibroblasts. We also demonstrate that imatinib mesylate blocks TGF␤-induced activation of Smad1 and ERK-1/2 pathways in control fibroblasts and induces phosphorylation of Smad1 and ERK-1/2 in SSc fibroblasts. The present study provides novel insights into the mechanism of fibrosis in SSc. 2529 PATIENTS AND METHODS Immunohistochemistry. The study group consisted of 8 patients with diffuse cutaneous SSc (dcSSc), 5 patients with limited cutaneous SSc (lcSSc), and 8 healthy volunteers (Table 1). All patients fulfilled the criteria of the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) for SSc (16). Upon obtaining informed consent and in compliance with the Institutional Review Board for Human Studies, skin biopsy samples were obtained from the affected areas (dorsal forearm) of the SSc patients. Skin biopsy samples were embedded in paraffin and used for immunohistochemistry. Immunohistochemical staining of paraffin-embedded sections was performed using a Vectastain ABC kit (Vector, Burlingame, CA) according to the manufacturer’s instructions. Fourmicrometer–thick sections were mounted on silane-coated slides and were then deparaffinized with Histoclear and rehydrated through a graded series of solutions of ethyl alcohol and phosphate buffered saline (PBS). Skin sections were treated with hydrogen peroxide for 30 minutes to block endogenous peroxidase activity and then subjected to a 45-minute antigenretrieval treatment with antigen unmasking solution (Vector). For immunostainings, we used primary antibodies against total Smad1 (Abcam, Cambridge, MA) and phospho-Smad1 (17). Incubations with the primary antibodies diluted in horse serum in PBS to the indicated concentrations were performed overnight in a humidified chamber at 4°C as previously described (18). After 3 rinses in PBS, binding sites of the primary antibodies were detected with biotinylated IgG, and the sites of peroxidase activity were visualized by using diaminobenzidine. The sections were then counterstained with hematoxylin. Immunostaining was detected by light microscopy. Normal rabbit IgG was used as a negative control (data not shown). Cell culture. Skin biopsy samples were obtained from the affected areas of the dorsal forearm of 7 patients with dcSSc, all of whom fulfilled the ACR criteria for dcSSc (16). Normal adult dermal fibroblasts derived from 5 healthy donors matched for age, sex, and race with each SSc patient served as controls. These fibroblasts were analyzed in parallel with the respective SSc cell lines. Pairs of SSc and normal fibroblast samples were processed as they were obtained from the donors to avoid freezing the cell cultures and to ensure that all cells in passages 2–5 were used. Dermal fibroblasts were obtained from healthy donor skin biopsy samples by enzymatically dissociating tissue specimens in 0.25% type I collagenase (Sigma, St. Louis, MO) and 0.05% DNase (Sigma) in Dulbecco’s modified Eagle’s medium (DMEM) with 20% fetal bovine serum (FBS). Fibroblasts were maintained in DMEM supplemented with 10% FBS for all experiments. Before treatment with TGF␤ or imatinib mesylate and infection with adenoviruses, fibroblasts were incubated in serum-free medium (DMEM/0.1% bovine serum albumin) for 24 hours. Western blotting. Confluent healthy dermal fibroblasts were lysed, and protein content was determined as described previously (19). Protein (40–50 g) was separated via sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDSPAGE) and transferred to a nitrocellulose membrane (BioRad, Hercules, CA). The membrane was then blocked at room temperature using 3% milk/Tris buffered saline–Tween (TBST) for 1 hour. The blots were probed overnight with a 2530 PANNU ET AL Table 1. Total Smad1 and phospho-Smad1 staining of fibroblasts in the skin of SSc patients and healthy subjects* Subject/age/sex Patients with dcSSc 1/47/F 2/42/F 3/36/F 4/61/F 5/52/F 6/27/F 7/41/F 8/60/F Patients with lcSSc 9/72/F 10/52/F 11/77/F 12/37/F 13/63/F Healthy subjects 1/63/F 2/55/M 3/42/F 4/54/F 5/40/M 6/38/F 7/52/F 8/27/F Disease duration 1 8 7 5 4 4 4 7 Rodnan skin thickness score (range 0–51) No. of total Smad1–positive fibroblasts† No. of phosphoSmad1–positive fibroblasts† 15 18 35 44 25 14 10 41 92 73 65 80 38 52 4 35 87 20 21 42 18 55 7 78 8 16 4 2 5 32 45 68 50 16 11 17 33 40 5 – – – – – – – – 8 40 47 4 19 15 17 29 0 22 9 0 2 3 10 0 year years years years years months years months 18 years 5 years 2 years 1 year 21 years – – – – – – – – * SSc ⫽ systemic sclerosis; dcSSc ⫽ diffuse cutaneous SSc; lcSSc ⫽ limited cutaneous SSc. † At least 100 fibroblasts were counted in each specimen. 1:1,000 dilution of appropriate primary antibodies against Smad1 and phospho-Smad1 (both from Cell Signaling Technology, Beverly, MA) and against TGF␤RI, TGF␤RII, and CCN2 (all from Santa Cruz Biotechnology, Santa Cruz, CA). Following washes with TBST, blots were incubated with an appropriate horseradish peroxidase–conjugated secondary antibody. As a control for equal protein loading, membranes were stripped and reprobed for ␤-actin using a monoclonal antibody to ␤-actin (Sigma). Protein levels were visualized using ECL reagent (Pierce, Rockford, IL) and quantitated using densitometry software (NIH Image, National Institutes of Health, Bethesda, MD; online at: http://rsb.info.nih.gov/ nih-image/). Transient transfection, luciferase assay, and RNA interference. SSc and normal dermal fibroblasts were seeded on 6-well plates (105 cells/well) and transfected 24 hours later. Transient transfections with the previously described wild-type (WT) CCN2 promoter construct (14) (a gift from Dr. Gary Grotendorst) were performed in duplicate using FuGene 6 reagent (Roche, Indianapolis, IN) according to the manufacturer’s specifications. Two SSc and healthy control fibroblast strains were transiently transfected with the CCN2-Luc promoter construct. Transfections were normalized using the pSV-␤-galactosidase control vector. Luciferase activities in aliquots normalized for equal protein concentrations were determined 24 hours after transfection using the Luciferase Assay System (Promega, Madison, WI). Small interfering RNA (siRNA) directed against c-Abl was purchased from Santa Cruz Biotechnology, along with the specific primers for real-time quantitative polymerase chain reaction (PCR). Nonsilencing siRNA used for this experiment was purchased from Qiagen (Chatsworth, CA). Confluent cultures of normal and SSc fibroblasts were transfected with the indicated siRNA using RNAiFect reagent (Qiagen) for 48 hours. The cells were then lysed in radioimmunoprecipitation assay (RIPA) buffer, and the amount of protein was estimated in each sample as described above. Western blot analysis was performed for phospho-Smad1 and total Smad1 as well as for type I collagen protein levels. The experiments using siRNA directed against Smad1 were performed as previously described (14). Quantitative real-time PCR was performed to determine the levels of c-Abl and Smad1 depletion (14). DNA affinity precipitation assay. Confluent dishes of dermal fibroblasts were either left untreated or were treated with TGF␤ (1 ng/ml) for 90 minutes. Cell pellets were lysed in RIPA buffer, and 500 mg of whole cell extract was incubated with 20 g of poly(dI-dC) in 1⫻ DNA affinity precipitation assay binding buffer (10 mM Tris HCl [pH 8.0], 40 mM KCl, 1 mM dithiothreitol, 6% glycerol, and 0.05% Nonidet P40) for 10 minutes at 4°C. The extracts were precleared with streptavidincoated agarose beads (Pierce), and the supernatants were incubated with 1.5 g of double-stranded WT connective tissue growth factor (CTGF) (WT CCN2) or CCN2 oligonucleotide mutated at the Sp-1 site (mutated Sp-1 CTGF) (both labeled with biotin at the 5⬘ end) for 4 hours at 4°C with constant rotation. The sequences of the forward oligonucleotides were Smad1 SIGNALING PATHWAY IN SSc 2531 Figure 1. Immunodetection of Smad1 and phospho-Smad1 in skin samples from healthy individuals and from patients with systemic sclerosis (SSc). A–D, Sections of normal skin (A and C) and SSc skin (B and D) were stained for total Smad1 (A and B) and phospho-Smad1 (C and D). Thick arrows indicate positively stained fibroblasts; thin arrows indicate positively stained endothelial cells (original magnification ⫻ 400). E, Percentage of Smad1-positive fibroblasts and phospho-Smad1–positive fibroblasts in SSc and normal healthy skin. Each symbol represents an individual subject. At least 100 fibroblasts were counted for each specimen. The percentage of Smad1-positive fibroblasts and phospho-Smad1–positive fibroblasts was higher in skin from patients with diffuse cutaneous SSc (dcSSc) than in skin from healthy individuals (P ⬍ 0.01 and P ⬍ 0.006, respectively). lcSSc ⫽ limited cutaneous SSc. 5⬘-GCC-TCG-GCC-GCC-CGC-CCC-AAA-CTC-ACA-CAAC-3⬘ for WT CTGF and 5⬘-GCC-TCG-GGC-CGC-TTC-CTCAAA-CTC-ACA-CAA-C-3⬘ for mutated Sp-1 CTGF. Poly(dIdC) (3 g) was used as a negative control. Streptavidin-coated agarose beads were added to each tube, and the samples were rotated for an additional 2 hours at 4°C. Beads were recovered by centrifugation and washed twice with Tris–EDTA followed by 2 washes with 1⫻ DNA affinity precipitation assay binding buffer and 2 washes with 1⫻ PBS. Proteins were eluted by heating the beads with 2⫻ sample loading buffer at 95°C for 2532 PANNU ET AL 5 minutes. The supernatants were separated by 10% SDSPAGE and transferred onto nitrocellulose membrane. Western blotting was performed with antibody against Smad1 (Santa Cruz Biotechnology), and the bands were visualized using enhanced chemiluminescence reagent. Chromatin immunoprecipitation (ChIP). ChIP was performed as described previously (20). Adult dermal fibroblasts were treated with either AdGFP or AdTGF␤RI (AdALK-5) for 18 hours. Half of the AdGFP-treated cells were also treated with TGF␤ (1 ng/ml) for 90 minutes. The cells were crosslinked and lysed, and nuclei were isolated as described previously (20). Chromatin was sheared by sonication to yield an average fragment size of 400 bp and microcentrifuged at 14,000 revolutions per minute, and supernatant containing soluble chromatin was collected. The chromatin solution was precleared with protein G–Sepharose for 30 minutes at 4°C and subsequently incubated with 4 g of antibody against Smad1 at 4°C overnight with constant rotation. Recovery of immunoprecipitated chromatin was done using protein G–Sepharose beads for 2 hours at 4°C. The chromatin–antibody–protein G–Sepharose complexes were recovered by centrifugation and washed, and the DNA was eluted. DNA was purified using the QIAquick PCR purification kit (Qiagen). One microliter of the purified DNA was used in each PCR. The PCR reaction was carried out for either 28 or 33 cycles, and the product was visualized by agarose gel electrophoresis. A 213-bp region (–194 to ⫹19) of the CCN2 promoter encompassing the Sp-1 site was amplified using the primers described previously (20). RESULTS Activation of the Smad1 pathway in SSc skin biopsy samples. To investigate the involvement of the Smad1 pathway in SSc fibrosis, we performed immunostaining for total Smad1 and phosphorylated Smad1 in skin specimens from 8 patients with dcSSc, 5 patients with lcSSc, and 8 healthy controls. Representative staining of SSc and healthy control skin samples is shown in Figures 1A–D. Smad1- and phospho-Smad1–positive fibroblasts were counted in each specimen, and the summary of the results is shown in Figure 1E and Table 1. The analysis revealed heterogeneity of Smad1 and phospho-Smad1 expression among SSc and control skin sections; however, a significantly higher proportion of positive fibroblasts was found in the skin of dcSSc patients. The mean ⫾ SEM percentage of Smad1positive fibroblasts was 54.8 ⫾ 10% in dcSSc skin compared with 22 ⫾ 5% in healthy skin (P ⬍ 0.01), and the mean ⫾ SEM percentage of phospho-Smad1– positive fibroblasts was 41 ⫾ 11% in dcSSc skin compared with 6 ⫾ 3% in healthy skin (P ⬍ 0.006) (Figure 1E). The expression of Smad1 and phospho-Smad1 in patients with lcSSc was moderately elevated, but the differences were not statistically significant. These find- ings strongly suggest that activation of the Smad1 pathway contributes to the development of fibrosis in dcSSc. Activation of the Smad1 pathway in cultured SSc fibroblasts. Cultured SSc fibroblasts at early passages retain some of the characteristics of the activated phenotype, including elevated levels of collagen and CCN2 genes (9). We examined the Smad1 phosphorylation status and total Smad1 levels in 7 pairs of SSc and closely matched control dermal fibroblast samples. Results from 3 pairs of representative SSc and control fibroblast samples are shown in Figure 2A. Phospho-Smad1 was not detectable in control dermal fibroblasts, but was present in 6 of 7 SSc fibroblast strains (results for the remaining SSc and control strains are shown below). The level of total Smad1 was markedly increased in all SSc strains compared with that in control fibroblasts (summarized in Figure 2B). The elevated levels of Smad1 correlated with increased expression of CCN2 (Figure 2A). Consistent with a previous report (21), SSc fibroblasts exhibited elevated CCN2 promoter activity, which was resistant to further stimulation with TGF␤ (Figure 2C). CTGF/CCN2 gene is a direct effector of the Smad1 pathway. We have recently shown that overexpression of TGF␤RI (AdALK-5) in dermal fibroblasts leads to up-regulation of CCN2 promoter activity through the GC-rich motif, which was previously shown to mediate Sp-1–dependent stimulation (14). Since this response required Smad1, we wanted to determine whether Smad1 binds directly to the proximal CCN2 promoter region. We first assessed Smad1 binding using the DNA affinity precipitation assay. Fragments (30-bp) of the CCN2 promoter corresponding to the WT and the GC-rich motif promoter mutant were labeled with biotin at the 5⬘ end and used in binding reactions with cell extracts from untreated and TGF␤-treated fibroblasts (Figure 3A). Smad1 bound to the CCN2 promoter in untreated cells, and the binding was further increased after TGF␤ treatment. Mutating the GC-rich motif almost completely abrogated Smad1 binding. Since a previous study using electrophoretic mobility shift assay analyses suggested that Sp-1 interacts with the proximal GC-rich motif (21), we also investigated the binding of Sp-1 using the DNA affinity precipitation assay; however, binding of Sp-1 to this promoter region was not detected under our experimental condition (data not shown). The reason for this discrepancy is currently not known. To assess the interaction of Smad1 with the endogenous CCN2 promoter, we used a ChIP assay. In untreated cells, low levels of Smad1 were detectable at the CCN2 promoter (Figure 3B); however, promoter occupancy by Smad1 was increased after TGF␤ stimu- Smad1 SIGNALING PATHWAY IN SSc 2533 pathways but is independent of the activation of canonical Smad2/3 signaling (14). Constitutive activation of ERK-1/2 signaling is also observed in cultured SSc fibroblasts (22,23). Since imatinib mesylate blocks fibrosis through a mechanism that is also independent of Smad2/3, we wished to determine whether imatinib mesylate affects activation of the Smad1 and/or ERK-1/2 pathways. The effects of imatinib mesylate were first tested in control fibroblasts stimulated with TGF␤ and in our Figure 2. Correlation of Smad1 and CCN2 expression in systemic sclerosis (SSc) fibroblasts. A, Three pairs of fibroblast samples from SSc patients and closely matched normal subjects (NS) were analyzed for the presence of phospho-Smad1, total Smad1, and CCN2 by Western blotting. Beta-actin was used as loading control. B, PhosphoSmad1 (pSmad1) and total Smad1 (TSmad1) protein levels were determined in 7 pairs of normal and SSc fibroblast cultures. ⴱ ⫽ P ⬍ 0.0018; ⴱⴱ ⫽ P ⬍ 0.0078, versus normal fibroblast cultures. C, CCN2 promoter activity in SSc fibroblasts was measured. Two SSc and healthy control fibroblast strains were transiently transfected with the CCN2-Luc promoter construct. Transfections were normalized using the pSV-␤-galactosidase control vector. The activity of the CCN2 promoter in unstimulated control fibroblasts was arbitrarily set at 1. The graph represents fold change in promoter activity in unstimulated SSc fibroblasts and in response to stimulation with transforming growth factor ␤ (TGF␤). Values in B and C are the mean and SEM. lation. Constitutively elevated Smad1 binding was observed in cells transduced with AdALK-5, which is consistent with the activation of the Smad1 pathway in these cells (14). These data indicate that Smad1 is a direct activator of the CCN2 gene in dermal fibroblasts. ERK-1/2–Smad1 pathway is targeted by imatinib mesylate. We have previously demonstrated that in our experimental model of SSc based on an altered ratio of TGF␤ receptors, the profibrotic response is mediated by a persistent activation of Smad1 and ERK-1/2 signaling Figure 3. Smad1 binds to the GC-rich motif in the proximal CCN2 promoter region in vitro and in vivo. A, DNA affinity precipitation assay showing the binding of Smad1 to the 5⬘ end–labeled wild-type CCN2 oligonucleotide probe (CCN2-oligo WT), but not the 5⬘ end– labeled CCN2 oligonucleotide mutated at the Sp-1 site (CCN2-oligo Mut.), in both the absence and presence of transforming growth factor ␤ (TGF␤) (top). C ⫽ control. Levels of total Smad1 in the samples before the DNA affinity precipitation assay (input) are also shown. Fifteen percent of the input samples were loaded. Mean and SEM relative Smad1 levels are at the bottom. B, Chromatin immunoprecipitation assay of CCN2 promoter after stimulation for 90 minutes with TGF␤ (1 ng/ml) of cells transduced with AdGFP or cells transduced with AdALK-5 with no stimulation. A polyclonal antibody to Smad1 was used for immunoprecipitation. Shown are polymerase chain reaction (PCR) analyses of the input (control), in the absence of antibody (No Ab.), and after immunoprecipitation with antibody using primers corresponding to the proximal region of the CCN2 promoter encoding the Smad1 site. An increased number of PCR cycles (33) was used to visualize Smad1 binding in control cells in the absence of TGF␤. 2534 Figure 4. Imatinib mesylate blocks transforming growth factor ␤ (TGF␤)–induced phosphorylation of ERK-1/2–Smad1 and upregulation of the CCN2 gene in dermal fibroblasts. Confluent dishes of adult fibroblasts were transduced with AdALK-5 (lanes 3 and 4) or control adenovirus (AdGFP) (lanes 1 and 2) as previously described (14). AdGFP-transduced cells were also stimulated with TGF␤ (1 ng/ml) for 45 minutes (lanes 5 and 6) or 24 hours (lanes 7 and 8). Where indicated, imatinib mesylate (2 g/ml) was added to cells 1 hour prior to TGF␤ stimulation or 6 hours after transduction with AdALK-5. Cell lysates were obtained as described in Patients and Methods, and Western blot analysis was carried out for CCN2, phospho-Smad1, total Smad1, phospho–ERK-1/2, and total ERK-1/2 protein levels. The blots were stripped and reprobed with antibody to ␤-actin for normalization. p44/42 MAPK ⫽ ERK-1/2. ALK-5–based experimental model of SSc (Figure 4). Consistent with the previous study, overexpression of ALK-5 resulted in up-regulation of CCN2 protein levels, increased phospho-Smad1 and total Smad1 levels, and increased phospho–ERK-1/2 levels (compare lanes 1 and 3 in Figure 4). Treatment with imatinib mesylate abrogated the up-regulation of CCN2, phospho-Smad1, and phospho–ERK-1/2 (compare lanes 3 and 4), while elevated total Smad1 levels were not affected by this treatment. We next examined the effects of imatinib mesylate on TGF␤-induced phosphorylation of Smad1 and ERK-1/2 and up-regulation of the CCN2 gene. Two time points, 45 minutes and 24 hours, were examined in order to evaluate both short- and long-term responses. Treatment with imatinib mesylate blocked induction of both phospho-Smad1 and phospho–ERK-1/2 (compare lanes 5 and 6). Up-regulation of the CCN2 gene observed 24 hours after stimulation was also blocked by imatinib mesylate (compare lanes 7 and 8). Contribution of c-Abl to activation of the Smad1 pathway in SSc fibroblasts. The effects of imatinib mesylate were next tested in 2 pairs of SSc and control fibroblasts. Consistent with our previous reports, SSc fibroblasts had elevated levels of TGF␤RI compared PANNU ET AL with control fibroblasts (Figure 5A). Increased levels of phosphorylated Smad1 and ERK-1/2, as well as increased total Smad1 and CCN2 levels, were present in both SSc strains (Figure 5B). Treatment with imatinib mesylate reversed the phosphorylation of Smad1 and ERK-1/2 but did not normalize the elevated total Smad1 levels. Up-regulation of CCN2 was completely normalized in one of the pairs and partially inhibited in the other pair. The heterogeneous response of SSc fibroblasts may reflect multiple mechanisms involved in upregulation of CCN2 in SSc fibroblasts. Because imatinib mesylate has an inhibitory activity toward 3 kinases, c-Abl, c-Kit, and platelet-derived growth factor receptor (PDGFR) and because both c-Abl and PDGFR have been linked to SSc fibrosis, we next determined whether blockade of c-Abl affects Smad1 phosphorylation. As shown in Figure 6A, downregulation of c-Abl expression levels using specific siRNA completely abolished Smad1 phosphorylation and normalized collagen production in SSc fibroblasts. These data demonstrate that c-Abl is required for the activation of Smad1 in SSc fibroblasts. To confirm the contribution of the Smad1 pathway to SSc fibrosis, Smad1 was blocked in SSc and control fibroblasts using Figure 5. Imatinib mesylate inhibits activation of the ERK-1/2– Smad1 pathway in systemic sclerosis (SSc) fibroblasts. A, Expression of transforming growth factor ␤ receptor I (TGF␤R I) and TGF␤RII was analyzed by Western blotting in 2 pairs (pairs 4 and 5) of fibroblast samples from SSc patients and normal subjects (NS). B, Cells were treated with 2 g/ml of imatinib mesylate 24 hours prior to cell lysis. Western blot analysis was carried out for CCN2, phospho-Smad1, total Smad1, phospho–ERK-1/2, and total ERK-1/2 protein levels. The blots were stripped and reprobed with antibody to ␤-actin for normalization. p44/42 MAPK ⫽ ERK-1/2. Smad1 SIGNALING PATHWAY IN SSc 2535 Figure 6. A, The nonreceptor tyrosine kinase c-Abl contributes to activation of the Smad1 pathway in systemic sclerosis (SSc) fibroblasts. Two pairs of fibroblast samples (pairs 6 and 7) from SSc patients and normal subjects (NS) were treated either with control nonsilencing small interfering RNA (siRNA) (Scr.) or with siRNA specifically directed against c-Abl (siAbl) for 48 hours prior to cell lysis. Western blot analysis was performed to determine phospho-Smad1 (pSmad1) and total Smad1 (TSmad1) levels as well as type I collagen levels. The blots were stripped and reprobed with antibody to ␤-actin for normalization (top). The down-regulation of c-Abl in fibroblasts in response to siAbl was quantified by real-time quantitative polymerase chain reaction (PCR) (bottom) as described in Patients and Methods. B, Smad1 contributes directly to regulation of CCN2 and type I collagen protein levels. Two pairs of normal and SSc fibroblasts were treated either with nonsilencing siRNA (Scr.) or with siRNA directed against Smad1 (siS1) as described previously (14). Western blot analysis was performed to determine CCN2 and type I collagen levels. The blots were stripped and reprobed with antibody to ␤-actin for normalization (top). Smad1 levels were quantified by real-time quantitative PCR (bottom) using specific primers as described previously (14). Values in the lower panels in A and B are the mean and SEM. previously developed Smad1 siRNA adenoviruses. Blockade of Smad1 normalized the elevated CCN2 and collagen levels in SSc fibroblasts to levels observed in the corresponding healthy controls (Figure 6B). Together, these data support the conclusion that activation of the c-Abl/Smad1 signaling pathway contributes to SSc fibrosis. DISCUSSION TGF␤ is considered a central mediator of the excessive collagen deposition in SSc, but the molecular mechanisms underlying the chronic fibrotic response are not fully elucidated. This study provides further support for the contribution of a novel fibrogenic pathway mediated by the activation of the ERK-1/2–Smad1– CCN2 cascade to SSc fibrosis. Analyses of SSc skin samples revealed elevated expression of total and phospho-Smad1 protein in a subset of patients with active disease. Fibroblasts cultured from SSc biopsy samples demonstrated elevated levels of Smad1 protein, and in 6 of 7 strains tested also showed the presence of phosphorylated Smad1, indicating that activation of the Smad1 pathway persists in cultured cells. Since several previous studies have shown elevated levels of phosphoSmad3 in SSc fibroblasts (13,24), these data suggest that distinct Smad pathways may play complementary roles in the activation of cultured SSc fibroblasts. Surprisingly, however, comparison of SSc and healthy skin tissues showed similar levels of phosphorylated Smad3 (24). The reason for the different expression of phosphorylated Smad3 in cells in vivo and in cultured cells remains unknown. In vitro and in vivo DNA binding assays have demonstrated that Smad1 binds to the GC-rich motif in the proximal CCN2 promoter region. Although this site was previously characterized as an Sp-1 binding site (21), our data strongly suggest that this GC-rich motif represents a Smad1 binding site. This finding is also consistent with the previous studies that showed binding of Smad1 to GC-rich motifs in several other promoters (25). Our recent studies have shown that siRNA-mediated inhibition of Smad1 abrogated TGF␤-induced up-regulation of CCN2 messenger RNA and protein levels as well as AdTGF␤RI-induced CCN2 promoter activity (14). In the present study, we extended these analyses to SSc fibroblasts and demonstrated that the Smad1 pathway 2536 also contributes to activation of profibrotic genes in these cells. Together with the results of the DNA binding assays, these studies indicate that the CCN2 gene is a direct effector of the Smad1 pathway. CCN2 is required for the profibrotic effects of TGF␤ in cultured cells and is considered to be a primary mediator of chronic fibrotic processes in SSc skin (26). Thus, activation of the Smad1 pathway may contribute to maintaining constitutively elevated expression levels of CCN2 in SSc fibroblasts in vivo and in vitro. TGF␤-dependent activation of the Smad1 pathway was first described in endothelial cells (27). In these cells, TGF␤ signals through a heteromeric receptor complex that includes TGF␤RII, ALK-5, and ALK-1 (28). While the specific interactions between ALK-5 and ALK-1 are still not fully elucidated, there is evidence that the relative ratio of these 2 receptor subtypes regulates the switch between proliferation and differentiation of endothelial cells (15). However, recent studies have also shown that the ALK-1/Smad1 pathway may play a role in kidney fibrosis (29–31). Furthermore, TGF␤ signaling via activation of the ALK-1/Smad1 pathway and subsequent up-regulation of the Id1 gene have been shown to contribute to transdifferentiation of hepatic stellate cells into myofibroblasts (32). These recent studies strongly suggest that activation of the ALK-1/Smad1 pathway may represent a common feature of various fibrotic diseases and that this pathway may play an important role in the development of organ fibrosis, including SSc. Imatinib mesylate is a tyrosine kinase inhibitor that inhibits a wide range of kinases, including c-Abl (or, Bcr-Abl), c-Kit, and PDGFR. Imatinib mesylate has been effective in preventing the development of organ fibrosis in the kidney, lung, liver, and skin in several animal models (5–7); however, it was significantly less effective in ameliorating an established fibrosis (33,34). The antifibrotic effects of imatinib mesylate include inhibition of fibroblast proliferation via blockade of PDGFR kinase and inhibition of collagen production in response to TGF␤. While it has been established that the antifibrotic effects of imatinib mesylate are mediated through inhibition of c-Abl, the role of this nonreceptor tyrosine kinase in collagen regulation is currently unknown. Our findings suggest that Smad1 may represent a downstream effector of the c-Abl signaling cascade in dermal fibroblasts. Further studies are needed to clarify the mechanism by which c-Abl exerts its profibrotic effects. Excessive deposition of collagen and other ECM components is the hallmark of SSc, occurring in all PANNU ET AL target organs and progressively interrupting circulation. Effective antifibrotic treatments for SSc and other fibrotic diseases are not available at present. Due to its prominent role in matrix regulation, TGF␤ signaling is considered an attractive therapeutic target for controlling fibrosis. However, the complexity of the profibrotic effects of TGF␤ that involve both Smad3-dependent and -independent mechanisms presents a challenge in selecting appropriate therapeutic targets. For example, pharmacologic inhibitors of TGF␤RI kinase that result in blockade of Smad3 signaling (ALK-5 kinase inhibitors) did normalize elevated CCN2 levels in SSc fibroblasts and, in general, ameliorated only selected aspects of SSc fibrosis in vitro (35,36). Our study suggests that due to the pleiotropic nature of TGF␤ signaling, targeting a single component of the TGF␤ signaling pathway may not be effective in ameliorating SSc fibrosis. The Smad1 pathway may represent, in addition to Smad3, a potential target for antifibrotic therapy in SSc and other fibrotic diseases. AUTHOR CONTRIBUTIONS Dr. Trojanowska 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 design. Pannu, Smith, Jablonska, Blaszczyk, Trojanowska. Acquisition of data. Pannu, Asano, Nakerakanti, Smith, Jablonska, Blaszczyk. Analysis and interpretation of data. Pannu, Asano, Nakerakanti, ten Dijke, Trojanowska. Manuscript preparation. Pannu, ten Dijke, Trojanowska. Statistical analysis. Pannu, Asano. Supply and characterization of key reagents. Ten Dijke. REFERENCES 1. Pannu J, Trojanowska M. Recent advances in fibroblast signaling and biology in scleroderma. Curr Opin Rheumatol 2004;16: 739–45. 2. Massague J. TGF-␤ signal transduction. Annu Rev Biochem 1998;67:753–91. 3. Miyazono K, ten Dijke P, Heldin CH. TGF-␤ signaling by Smad proteins. Adv Immunol 2000;75:115–57. 4. Moustakas A, Heldin CH. Non-Smad TGF-␤ signals. J Cell Sci 2005;118:3573–84. 5. 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;56:311–22. 6. Daniels CE, Wilkes MC, Edens M, Kottom TJ, Murphy SJ, Limper AH, et al. Imatinib mesylate inhibits the profibrogenic activity of TGF-␤ and prevents bleomycin-mediated lung fibrosis. J Clin Invest 2004;114:1308–16. 7. Wang S, Wilkes MC, Leof EB, Hirschberg R. Imatinib mesylate blocks a non-Smad TGF-␤ pathway and reduces renal fibrogenesis in vivo. FASEB J 2005;19:1–11. 8. Wilkes MC, Leof EB. Transforming growth factor ␤ activation of c-Abl is independent of receptor internalization and regulated by Smad1 SIGNALING PATHWAY IN SSc 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. phosphatidylinositol 3-kinase and PAK2 in mesenchymal cultures. J Biol Chem 2006;281:27846–54. Trojanowska M. What did we learn by studying scleroderma fibroblasts? Clin Exp Rheumatol 2004;22 Suppl 33:S59–63. Asano Y, Ihn H, Yamane K, Jinnin M, Mimura Y, Tamaki K. Increased expression of integrin ␣v␤3 contributes to the establishment of autocrine TGF-␤ signaling in scleroderma fibroblasts. J Immunol 2005;175:7708–18. Asano Y, Ihn H, Yamane K, Kubo M, Tamaki K. Increased expression levels of integrin ␣v␤5 on scleroderma fibroblasts. Am J Pathol 2004;164:1275–92. Pannu J, Gardner H, Shearstone JR, Smith E, Trojanowska M. Increased levels of transforming growth factor ␤ receptor type I and up-regulation of matrix gene program: a model of scleroderma. Arthritis Rheum 2006;54:3011–21. Mori Y, Chen SJ, Varga J. Expression and regulation of intracellular SMAD signaling in scleroderma skin fibroblasts. Arthritis Rheum 2003;48:1964–78. Pannu J, Nakerakanti S, Smith E, ten Dijke P, Trojanowska M. Transforming growth factor-␤ receptor type I-dependent fibrogenic gene program is mediated via activation of Smad1 and ERK1/2 pathways. J Biol Chem 2007;282:10405–13. Goumans MJ, Lebrin F, Valdimarsdottir G. Controlling the angiogenic switch: a balance between two distinct TGF-␤ receptor signaling pathways. Trends Cardiovasc Med 2003;13:301–7. Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581–90. Persson U, Izumi H, Souchelnytskyi S, Itoh S, Grimsby S, Engstrom U, et al. The L45 loop in type I receptors for TGF-␤ family members is a critical determinant in specifying Smad isoform activation. FEBS Lett 1998;434:83–7. Kubo M, Czuwara-Ladykowska J, Moussa O, Markiewicz M, Smith E, Silver RM, et al. Persistent down-regulation of Fli1, a suppressor of collagen transcription, in fibrotic scleroderma skin. Am J Pathol 2003;163:571–81. Gore-Hyer E, Pannu J, Smith EA, Grotendorst G, Trojanowska M. Selective stimulation of collagen synthesis in the presence of costimulatory insulin signaling by connective tissue growth factor in scleroderma fibroblasts. Arthritis Rheum 2003;48:798–806. Nakerakanti SS, Kapanadze B, Yamasaki M, Markiewicz M, Trojanowska M. Fli1 and Ets1 have distinct roles in connective tissue growth factor/CCN2 gene regulation and induction of the profibrotic gene program. J Biol Chem 2006;281:25259–69. Holmes A, Abraham DJ, Chen Y, Denton C, Shi-wen X, Black CM, et al. Constitutive connective tissue growth factor expression in scleroderma fibroblasts is dependent on Sp1. J Biol Chem 2003;278:41728–33. Svegliati S, Cancello R, Sambo P, Luchetti M, Paroncini P, Orlandini G, et al. Platelet-derived growth factor and reactive oxygen species (ROS) regulate Ras protein levels in primary human fibroblasts via ERK1/2: amplification of ROS and Ras in systemic sclerosis fibroblasts. J Biol Chem 2005;280:36474–82. Tourkina E, Gooz P, Pannu J, Bonner M, Scholz D, Hacker S, et 2537 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. al. Opposing effects of protein kinase C␣ and protein kinase C⑀ on collagen expression by human lung fibroblasts are mediated via MEK/ERK and caveolin-1 signaling. J Biol Chem 2005;280: 13879–87. Verrecchia F, Laboureau J, Verola O, Roos N, Porcher R, Bruneval P, et al. Skin involvement in scleroderma—where histological and clinical scores meet. Rheumatology (Oxford) 2007;46: 833–41. Massague J, Seoane J, Wotton D. Smad transcription factors. Genes Dev 2005;19:2783–810. Shi-wen X, Pennington D, Holmes A, Leask A, Bradham D, Beauchamp JR, et al. Autocrine overexpression of CTGF maintains fibrosis: RDA analysis of fibrosis genes in systemic sclerosis. Exp Cell Res 2000;259:213–24. Goumans MJ, Valdimarsdottir G, Itoh S, Rosendahl A, Sideras P, ten Dijke P. Balancing the activation state of the endothelium via two distinct TGF-␤ type I receptors. EMBO J 2002;21:1743–53. Goumans MJ, Valdimarsdottir G, Itoh S, Lebrin F, Larsson J, Mummery C, et al. Activin receptor-like kinase (ALK)1 is an antagonistic mediator of lateral TGF␤/ALK5 signaling. Mol Cell 2003;12:817–28. Abe H, Matsubara T, Iehara N, Nagai K, Takahashi T, Arai H, et al. Type IV collagen is transcriptionally regulated by Smad1 under advanced glycation end product (AGE) stimulation. J Biol Chem 2004;279:14201–6. Takahashi T, Abe H, Arai H, Matsubara T, Nagai K, Matsuura M, et al. Activation of STAT3/Smad1 is a key signaling pathway for progression to glomerulosclerosis in experimental glomerulonephritis. J Biol Chem 2005;280:7100–6. Matsubara T, Abe H, Arai H, Nagai K, Mima A, Kanamori H, et al. Expression of Smad1 is directly associated with mesangial matrix expansion in rat diabetic nephropathy. Lab Invest 2006;86: 357–68. Wiercinska E, Wickert L, Denecke B, Said HM, Hamzavi J, Gressner AM, et al. Id1 is a critical mediator in TGF-␤-induced transdifferentiation of rat hepatic stellate cells. Hepatology 2006; 43:1032–41. Neef M, Ledermann M, Saegesser H, Schneider V, Widmer N, Decosterd LA, et al. Oral imatinib treatment reduces early fibrogenesis but does not prevent progression in the long term. J Hepatol 2006;44:167–75. Vittal R, Zhang H, Han MK, Moore BB, Horowitz JC, Thannickal VJ. Effects of the protein kinase inhibitor, imatinib mesylate, on epithelial/mesenchymal phenotypes: implications for treatment of fibrotic diseases. J Pharmacol Exp Ther 2007;321:35–44. Chen Y, Shi-wen X, Eastwood M, Black CM, Denton CP, Leask A, et al. Contribution of activin receptor–like kinase 5 (transforming growth factor ␤ receptor type I) signaling to the fibrotic phenotype of scleroderma fibroblasts. Arthritis Rheum 2006;54: 1309–16. Ishida W, Mori Y, Lakos G, Sun L, Shan F, Bowes S, et al. Intracellular TGF-␤ receptor blockade abrogates Smad-dependent fibroblast activation in vitro and in vivo. J Invest Dermatol 2006;126:1733–44.