Transforming Growth Factor-β Receptor Type I-dependent Fibrogenic Gene Program Is Mediated via Activation of Smad1 and ERK1/2 Pathways*

The transforming growth factor (TGF)-β/Smad3 signaling pathway is considered a central mediator of pathological organ fibrosis; however, contribution of Smad2/3-independent TGF-β signaling has not been fully explored. The present study utilized previously a described model of scleroderma (SSc) fibrosis based on forced expression of the TGF-βRI (ALK5) (Pannu, J., Gardner, H., Shearstone, J. R., Smith, E., and Trojanowska, M. (2006) Arthritis Rheum. 54, 3011–3021). This study was aimed at determining the molecular mechanisms underlying the profibrotic program in this model. We demonstrate that the TGF-βRI-dependent up-regulation of collagen and CCN2 (CTGF) does not involve Smad2/3 activation but is mediated by ALK1/Smad1 and ERK1/2 pathways. The following findings support this conclusion: (i) Smad2 and -3 were not phosphorylated in response to TGF-βRI, (ii) a TGF-βRI mutant defective in Smad2/3 activation, ALK5(3A), potently stimulated collagen production, (iii) elevation of TGF-βRI triggered sustained association of ALK5 with ALK1 and high levels of Smad1 phosphorylation, (iv) blockade of Smad1 via small interfering RNA abrogated collagen and CCN2 up-regulation in this model, (v) elevated TGF-βRI led to a prolonged activation of ERK1/2, (vi) the pharmacologic inhibitor of ERK1/2 inhibited Smad1 phosphorylation and abrogated profibrotic effects of elevated TGFβ-RI. Additional experiments demonstrated that a GC-rich response element located -6 to -16 (upstream of the transcription start site) in the CCN2 promoter mediated Smad1-dependent increased promoter activity in this model. This element was shown previously to mediate up-regulation of the CCN2 promoter in SSc fibroblasts. In conclusion, this study defines a novel ALK1/Smad1- and ERK1/2-dependent, Smad3-independent mode of TGF-β signaling that may operate during chronic stages of fibrosis in SSc.

Transforming growth factor-␤ (TGF-␤) 2 is a pleiotropic cytokine with diverse functions in many cell types (1). The Smad pathway is a central mediator of TGF-␤ signaling (2). The canonical Smad pathway is activated upon binding of TGF-␤ to the heteromeric serine-threonine kinase receptors, TGF-␤RII and -RI (ALK5), which leads to phosphorylation of Smad2 and -3, oligomerization with Smad4, nuclear translocation, and activation of various transcription programs (1). Furthermore, a more complex TGF-␤ signaling has recently been described in endothelial cells, in which TGF-␤ signals through two distinct type I receptors, ALK5 and ALK1, and their respective signal transducers: Smad2 and Ϫ3 for ALK5 and Smad1 and Ϫ5 for ALK1 (3). There is also evidence for various modifications of basic TGF-␤/Smad signaling, especially in pathological conditions such as cancer or fibrosis (2). Whereas the canonical TGF-␤/Smad pathway has been studied in detail, other modes of TGF-␤ signaling are still poorly understood.
Because TGF-␤ is a potent inducer of extracellular matrix (ECM) synthesis, it has long been considered a primary mediator of organ fibrosis (4). Numerous studies have demonstrated that blockade of TGF-␤ signaling ameliorated fibrotic response in various experimental models of fibrosis (5)(6)(7)(8)(9). Moreover, a specific role of Smad3 in fibrosis was demonstrated in mice lacking Smad3 (10). Smad3-null mice showed reduction of fibrosis in different organs, including lung, skin, liver, and kidney (10). Surprisingly, however, production of ECM was unaffected during cutaneous wound healing in Smad3-null mice, suggesting the existence of Smad3-independent mechanisms involved in physiologic tissue repair by dermal fibroblasts (11). Animal models, at best, represent only certain aspects of human fibrotic diseases, which are chronic and progressive and may take many years to develop. Despite numerous studies, the specific contribution of TGF-␤ signaling to the human fibrosis has not been fully elucidated and, in particular, the basis for the chronic nature of TGF-␤ signaling in fibrosis remains unknown. One of the characteristics of fibrotic lesions is the aberrant expression of TGF-␤ receptors. For example, an increased ratio of TGF-␤RI to RII expression levels have been * This work was supported by National Institutes of Health Grants AR-42334 and AR-44883 (to M. T.), the National Scleroderma Foundation (to J. P.), and the Dutch Cancer Society (to P. t. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1-S3. 1 To whom correspondence should be addressed: CSB 912, 96 Jonathan Lucas St., Charleston, SC 29425. Tel.: 843-792-7921; Fax: 843-792-7121; E-mail: trojanme@musc.edu. 2 The abbreviations used are: TGF-␤, transforming growth factor-␤; qRT, quantitative reverse transcription; ECM, extracellular matrix; SSc, scleroderma; ERK, extracellular signal-regulated kinase; DMEM, Dulbecco's modified Eagle's medium; siRNA, small interfering RNA; m.o.i., multiplicity of infection; TBST, Tris-buffered saline-Tween; co-IP, co-immunoprecipitation; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; MAPK, mitogen-activated protein kinase; T␤RE, transforming growth factor-␤ response element.
observed in liver fibrosis (12), atherosclerosis (13), and scleroderma (SSc) (14), whereas an increase in RII has been observed in kidney fibrosis (15,16). Furthermore, constitutive nuclear localization of Smad2 unresponsive to the blockade of the endogenous TGF-␤ signaling has been shown in scleroderma fibroblasts and in activated hepatic stellate cells (17,18). In addition, previous studies have demonstrated that constitutively elevated expression levels of CCN2 (CTGF) in scleroderma fibroblasts is Smad2/3-independent (19). Consistent with this observation, blockade of Smad signaling by a pharmacologic inhibitor of ALK5 kinase did not normalize elevated CCN2 levels in scleroderma fibroblasts (20). Whereas collectively these findings suggest that overproduction of matrix in fibrosis may be, at least in part, due to the alternative activation of TGF-␤ signaling, the specific mechanisms involved in such activation are still poorly understood. Based on the observation that SSc fibroblasts exhibit an elevated ratio of TGF-␤ type I to type II receptors we have established an experimental model that mimicked this condition in control dermal fibroblasts by carefully titrating the dose of adenoviral vector expressing TGF-␤RI (14). We found considerable overlap between the genes elevated in this experimental model and those previously reported to be overexpressed by lesional SSc fibroblasts and in scleroderma biopsies in vivo (21,22). Among such genes were collagens type I, III, and V, fibronectin, proline 4-hydroxylase, CTGF, SPARC, IGFBP-3, and other matrix protein-related genes. Furthermore, collagen protein production in this model was resistant to the inhibition by ALK5 kinase inhibitors, suggesting that this in vitro experimental model reproduced to a large extent the fibrotic program characteristic of lesional SSc fibroblasts. The present study was undertaken to delineate the signaling molecules that mediate ECM up-regulation in response to the elevated TGF-␤RI levels. The results of this study demonstrate that TGF-␤RI-dependent induction of the fibrotic program does not involve Smad2/3 and is mediated by activation of ALK1/Smad1 and ERK1/2 pathways.

EXPERIMENTAL PROCEDURES
Cell Culture-Upon informed consent and in compliance with the Institutional Review Board of Human Studies, skin biopsies were taken from the dorsal forearm of healthy donors. Dermal fibroblasts were obtained from the biopsies by enzymatically dissociating tissue specimens in 0.25% collagenase type I (Sigma) and 0.05% DNase (Sigma) in Dulbecco's modified Eagle's medium (DMEM) with 20% fetal bovine serum. Fibroblasts were maintained in DMEM supplemented with 10% fetal bovine serum for all experiments. Before stimulation with cytokines and infection with adenoviruses, fibroblasts were incubated in serum-free medium (DMEM, 0.1% bovine serum albumin) for 24 h.
Adenoviral Constructs and siRNA Vectors-Replication-incompetent adenoviral vector expressing rat full-length ALK5/ TGF-␤RI (abbreviated as AdTGF␤RI or AdRI) and control green fluorescent protein (AdGo) were generated as described earlier (14). In experiments using AdTGF-␤RI, dermal fibroblasts were transduced with 25-50 m.o.i. of the adenovirus. Multiplicity of infection is calculated as (total number of cells per well ϫ desired m.o.i.) divided by the titer of the virus. AdALK5(3A) was previously described (23). Adenoviral Smad1 siRNA vector with the target sequence 5Ј-CCTGTCATTATT-GCTTACT-3Ј was prepared as follows. Annealed doublestranded oligonucleotides were cloned into MluI and XhoI sites of the pRNAT-H1.1 shuttle vector (Genescript). The shuttle vector with target siRNA sequence was linearized with PmeI and then electroporated into BJ5183-AD-1 (Stratagene) to generate recombinant Adenoeasy vector. The recombinant Adenoeasy plasmid after linearization with the PacI enzyme was transfected into QBI-293 cells using Transfectin (Bio-Rad) for generation of adenovirus. The shuttle vector plasmid and Adenoeasy vector plasmid were sequenced to confirm the cloning. The primary adenoviral stock was then amplified and concentrated by cesium chloride density gradient centrifugation. Typical viral titer was 1 ϫ 10 10 plaque forming units/ml. In addition, another adenoviral Smad1 siRNA targeted to the 5Ј-CACACACCTTGGTAACATA-3Ј sequence of the Smad1 gene was also prepared as described above, to validate the effect of Smad1 suppression. The sequence of Scrambled siRNA used as a non-silencing control for this experiment is 5Ј-TTCTC-CGAACGTGTCACGT-3Ј and was prepared in the same manner. Adenoviral vector expressing siRNA directed toward TGF-␤RII were provided by Dr. Peter ten Dijke. siRNA directed against human Smad2 and Smad3 were purchased from Dharmacon RNA Technologies, CO. The sequence of the sense oligo for siRNA directed against Smad2 is 5Ј-GUCCCAUGAAAA-GACUUAATT-3Ј and for Smad3 is 5Ј-GGAGAAAUGGUGC-GAGAAGTT-3Ј.
Immunoprecipitation (IP) and Co-immunoprecipitation (Co-IP)-For IP, 400 g of total cell protein were used. Adult dermal fibroblasts were grown to confluence after which they were made quiescent by treatment with DMEM, 0.1% bovine serum albumin. The cells were treated with increasing doses of either AdGFP or AdTGF-␤RI. TGF-␤ (5 ng/ml) treatment for 30 min was used as a positive control for this experiment. Cells were washed twice with cold phosphate-buffered saline and lysed in buffer A containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.02% sodium azide, 0.1% SDS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 50 mM sodium fluoride, 0.5 mM dithiothreitol, 2 mM sodium orthovanadate, and 1 mM phenylmethylsulfonyl fluoride with protease inhibitors (Mixture Set III, Calbiochem). Protein concentrations were determined by BCA Protein Assay (Pierce). Complex formation was performed overnight at 4°C using 5 g of rabbit polyclonal Smad3 or rabbit Sp1 antibody (sc-8332, Santa Cruz, Biotechnology) followed by precipitation with protein G-Sepharose (Amersham Biosciences) for 2 h in 4°C. Negative controls were performed using normal rabbit IgG. The immunoprecipitates were washed four times in buffer A, eluted by boiling for 5 min in 2ϫ SDS sample buffer, and analyzed by Western blot. Samples were subjected to electrophoresis in a 12% SDS-polyacrylamide gel and trans-blotted onto polyvinylidene difluoride membranes (Millipore). After blocking with 3% bovine serum albumin, the membranes were incubated with 1:1000 dilution of mouse monoclonal anti-phosphoserine antibody (ab6639-100, Abcam) overnight, followed by incubation with horseradish peroxidase-conjugated secondary antibody, and washing with Tween/Tris-buffered saline solution. Proteins were detected using enhanced chemiluminescence (Amersham Biosciences). Total Smad3 protein was detected by Western blotting on the same membranes after stripping and re-probing the blot with mouse monoclonal anti-Smad1/2/3 (sc-7960, Santa Cruz Biotechnology) and appropriate secondary antibody.
For co-IP, 1000 g of total cell protein was used. Confluent dishes of adult dermal fibroblasts and human umbilical vein endothelial cells were serum-starved for 24 h and treated with 1 ng/ml TGF-␤ for 45 min where indicated. The cells were scraped and lysed in co-IP buffer (1% Triton X-100, 20 mM Tris, pH 7.0, 125 mM NaCl, 1 mM EDTA) for 30 min and centrifuged to separate the cell debris. Protein concentration was determined as described above. Complex formation was performed overnight at 4°C with rabbit polyclonal ALK5 antibody (Santa Cruz Biotechnology) and processed as described above. Western blot analysis was carried out with goat polyclonal ALK-1 antibody and subsequently with goal polyclonal ALK5 antibody (both from Santa Cruz Biotechnology).
Northern Blotting-Fibroblasts were grown to confluence in 60-mm 2 dishes, incubated in serum-free DMEM, and transduced with control AdGo or AdTGF-␤RI. UO126 was added 24 h before harvesting the cells. 48 h later, total RNA was extracted and analyzed by Northern blotting as previously described (25). TGF-␤ (5 ng/ml) treatment for 24 h was used as a positive control. Filters were hybridized sequentially with radioactive probes for COL1A2, COL1A1, CTGF, and 18 S rRNA. The signal was quantified using NIH Image densitometry software (National Institutes of Health).
Site-directed Mutagenesis and RT-PCR-Point mutations of the Sp1 site (proximal to the start site) in the CTGF promoter were introduced using the Exsite PCR-based site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. The sequence of the oligonucleotides used for mutation is: forward, 5Ј-GGCTGTCTCCTCTCAGCGGG-3Ј and reverse, 5Ј-TTTTATACGCTCCGGGCGGCCGCGGC-3Ј. The mutation of the Smad binding site and the putative TGF-␤ response element (T␤RE) was carried out using the QuikChange site-directed mutagenesis kit. The sequences of the oligonucleotide used to mutate the Smad binding site were: forward, 5Ј-CTGGAGTGTGCCAGCTTTTTTGGCTGGAG-GAATGCTGAGTGTCAAG-3Ј and reverse, 5Ј-CTTGACAC-TCAGCATTCCTCCAGCCAAAAAAGCTGGCACACTCC-AG-3Ј, and for T␤RE are: forward 5Ј-TCAGACGGAGGAA-TGCTGAACATCCCGGAGCCAGGATCAATCGC-3Ј and reverse-5Ј-GCGATTGATCCTGGCTCCGGGATGTTCAG-CATTCCTCCGTCTGA-3Ј. All mutations were confirmed by sequencing. RT-PCR for TGF-␤RII mRNA levels was carried out as described earlier (14).
Quantitative RT-PCR (qRT-PCR)-For validation of genes at the mRNA level, total RNA was isolated from dermal fibroblasts using TriReagent (MRC Inc.) according to the manufacturer's instructions. 2 g of RNA was reverse transcribed in a 20-l reaction using random primers and Transcriptor First Strand synthesis kit (Roche Applied Sciences). qPCR was carried out using IQ SYBR Green mixture (Bio-Rad) on an iCycler PCR machine (Bio-Rad) using 1 l of cDNA in triplicate with ␤-actin as the internal control. The primers are listed in Table 1.
Transient Transfection and Luciferase Assay-Adult dermal fibroblasts were seeded on 6-well plates (10 5 cells/well) and transfected 24 h later. Transient transfections with the indicated reporter, expression, and control constructs were performed in duplicate using FuGENE 6 reagent (Roche Applied Sciences) according to the manufacturer's specifications. After overnight incubation at 37°C, cells were stimulated with TGF-␤1 (5 ng/ml) for an additional 24 h. Luciferase activities in aliquots normalized for equal protein concentrations were determined 24 h after TGF-␤1 stimulation using the Luciferase Assay System (Promega). Transfections were repeated at least four times using two different plasmid preparations.
In some experiments dermal fibroblasts were first transduced with AdRI (m.o.i. 25) and 8 h later transfected with CTGF-Luc or pSBE4-luc promoter using FuGENE 6 reagent as described above. Cells were harvested 48 h later and luciferase activity was measured in aliquots normalized for equal protein concentrations.
In some experiments dermal fibroblasts were first transduced with Ad siRNA directed against Smad1 for 24 h followed by transduction with AdRI (m.o.i. 25) for an additional 8 h. Cells were then transfected with CTGF-luc promoter using FuGENE Forward AATGTCGCGGAGGACTTTGATTGC Reverse AGGATGGCAAGGGACTTCCTGTAA 6 reagent as described above and collected 48 h later. Luciferase activity was determined as described above.
Statistical Analysis-Student's t test analysis using GraphPad InStat statistics software (version 1.12) was performed to determine statistical significance. Values less than or equal to 0.05 were considered statistically significant.

TGF-␤RI-dependent Induction of Collagen Does Not
Depend on Activation of Smad2/3 Signaling-We have previously established an experimental model of fibrosis using adenoviralmediated TGF-␤RI overexpression in human dermal fibroblasts (21). To begin the investigation of the molecular mechanism underlying the profibrotic effects of TGF-␤RI, the kinetics of the TGF-␤RI protein expression levels after adenoviral transduction were determined. Confluent cells were transduced with the previously established optimal dose of AdTGF-␤RI for 6, 12, 18, 24, 48, and 72 h. The TGF-␤RI protein levels were consistently elevated at 12 h, and reached plateau at 18 h posttransduction. There was no further increase in TGF␤-RI protein up to 72 h post-transduction (Fig. 1A). The collagen type I protein levels were similarly increased starting at 12-18 h posttransduction reaching steady-state expression levels at 24 h and remaining at this level for up to 72 h. Thus, under these experimental conditions, the steady-state TGF-␤RI signaling activity is maintained for the duration of the experiment. We have previously shown that the endogenously produced TGF-␤ is required for the observed profibrotic response in this model (21).
Smad2 and -3 are the best characterized downstream effectors of TGF-␤ and are phosphorylated in response to phosphorylation/activation of TGF-␤RI (1). To determine the signaling molecules that mediate ECM up-regulation in response to the elevated TGF-␤RI levels, we examined the phosphorylation status of Smad2 and Smad3. In contrast to the potent, acute Smad2 phosphorylation observed in response to TGF-␤ (1 ng/ml) treatment for 45 min, no considerable Smad activation was seen in dermal fibroblasts that were transduced with the AdTGF-␤RI at any of the time points examined (Fig. 1A). Likewise, Smad3 was not phosphorylated (Fig. 1B). These results are consistent with our previous observation demonstrating that the specific TGF-␤RI kinase inhibitors, SB431542 and SD-208 failed to inhibit stimulation of collagen in this experimental model (20,21). Together, these data support the conclusion that the observed effects of TGF-␤RI overexpression may not involve activation of the Smad2/3 pathway.
To directly test this possibility, we utilized the TGF-␤RI mutant defective in Smad activation, ALK5(3A) (23). Fig. 1C shows the basal expression level of ALK5(3A) at increasing doses of virus. Collagen protein levels were analyzed in normal dermal fibroblasts transduced with increasing doses of AdALK5(3A). Overexpression of ALK5(3A) potently stimulated basal collagen type I protein levels at all doses tested, with m.o.i. 25-50 being the most effective. Together, these data support the conclusion that simulation of collagen type I production by overexpression of TGF-␤RI may not involve activation of Smad2/3 pathway.
The lack of involvement of the Smad2/3 pathway in profibrogenic effects of TGF-␤RI was unexpected and, therefore, to exclude the possibility that low levels of Smad2/3 signaling were contributing to matrix gene expression in our model, additional experiments were carried out. First, we demonstrated that whereas the siRNA-mediated blockade of either Smad2 or -3 completely prevented TGF-␤-induced collagen up-regulation, AdTGF-␤RI dependent up-regulation of collagen remained unaffected by this treatment (supplemental materials Fig. S1). In addition, we utilized the pSBE4-luc promoter that consists of four tandem copies of Smad3 binding sites (26) and represents a very sensitive tool to measure activation of Smad3-dependent signaling. This promoter is highly responsive to TGF-␤ stimulation in control dermal fibroblasts, but there was no detectable activation of this promoter in cells transduced with AdTGF-␤RI (supplemental materials Fig. S2). Together, these data indi- cate that stimulation of collagen type I production by overexpression of TGF-␤RI does not require activation of the Smad2/3 pathway.

TGF-␤RI-dependent Induction of the Profibrotic Gene Program Is Mediated through Activation of Smad1
Pathway-Smad1 is a downstream mediator of BMP signaling, but more recent studies using endothelial cells have also placed Smad1 in the TGF-␤ signaling pathway (27). We, therefore, examined the phosphorylation status of Smad1 in cells transduced with AdTGF-␤RI. As shown in Fig. 2A, Smad1 phosphorylation was significantly increased from 18 to 72 h post-transduction. Interestingly, phosphorylation of Smad1 was also seen in response to TGF-␤, suggesting that this mode of TGF-␤ signaling may not be restricted to endothelial cells. Furthermore, total Smad1 mRNA and protein levels were also elevated in cells overexpressing TGF-␤RI in a time-dependent manner (Fig. 2, A and B).
To determine whether activation of the Smad1 pathway contributes to the profibrotic effects of AdTGF-␤RI, we utilized an RNA interference approach to block expression of Smad1. As shown in Fig. 2C, efficient suppression of Smad1 protein levels were achieved using adenovirally expressed Smad1-specific siRNA. siRNA directed against Smad1 did not affect the Smad2/3 protein levels (supplemental materials Fig. S3). Suppression of Smad1 abrogated basal and TGF-␤-induced collagen protein expression levels and prevented elevated production of collagen in response to AdTGF-␤RI overexpression. These findings were confirmed at the mRNA level (Fig. 2C,  lower panel). Up-regulation of COL1A1, COL1A2, and CTGF mRNAs in response to either TGF-␤ stimulation or TGF-␤RI overexpression were abrogated by Smad1 siRNA. Similar results were obtained with two different Smad1-specific targeting sequences. These results suggested that activation of the ALK1/Smad1 signaling pathway in response to TGF-␤ stimulation or TGF-␤RI overexpression is involved in regulation of collagen synthesis. To further test this possibility, we examined formation of ALK5/ALK1 complexes. As shown in Fig. 2D, ALK5 does not associate with ALK1 in unstimulated fibroblasts, however, ALK5/ALK1 complexes form in response to TGF-␤ stimulation. In cells overexpressing TGF-␤RI, ALK5 constitutively associates with ALK1 with no further increase after TGF-␤ stimulation. Human umbilical vein endothelial cells were used as positive control. Together, these data strongly suggest that Smad1 is a critical mediator of collagen up-regulation in response to TGF-␤. Furthermore, Smad1, in the absence of Smad3 activation, mediates profibrotic response in our experimental model.
MEK/ERK Pathway Is Activated in Response to TGF␤RI Overexpression and Is Required for Its Profibrotic Effects-It was recently shown that activation of the ERK1/2 (p44/42) MAPK pathway contributes to collagen up-regulation in skin and lung fibroblasts from scleroderma patients (28,29). However, the mechanism responsible for activation of ERK1/2 in dermal fibroblasts has not been investigated. We asked whether overexpression of TGF-␤RI affects activation of this pathway. The levels of phosphorylated ERK1/2 were markedly elevated in cells overexpressing TGF-␤RI (Fig. 3A). Because activation of p44/42 MAPK has earlier been reported to be involved in upregulation of CTGF in lung fibroblasts and mesangial cells (30,31), we also tested the effect of UO126 on CTGF mRNA expression in normal dermal fibroblasts transduced with AdTGF-␤RI at m.o.i. 25 and 50. As shown in Fig. 3B, UO126 effectively blocked the stimulatory effect of TGF-␤RI overexpression on CTGF mRNA, thus confirming the involvement of the p44/42 MAPK pathway in CTGF gene expression. TGF-␤ (5 ng/ml) treatment was used as a positive control for these experiments. Interestingly, UO126 could only partially reverse the potent stimulation of CTGF in response to TGF-␤. These data suggest that induction of CTGF in response to TGF-␤ involves ERK1/ 2-dependent and ERK1/2-independent mechanisms, whereas activation of the ERK1/2 pathway is necessary in AdTGF-␤RIinduced expression of CTGF. We have also examined the effects of TGF␤-RI/ALK5 kinase inhibitor, SD-208, on TGF-␤RI dependent activation of ERK1/2. As shown in Fig. 3C, SD-208 was ineffective in reversing phosphorylation of ERK1/2, consistent with our previous observation demonstrating the inability of SD-208 to prevent collagen up-regulation in response to TGF-␤RI overexpression (21).
To examine the possible cross-talk between ERK1/2 and Smad1 signaling pathways, TGF-␤RI-transduced dermal fibroblasts were treated with a specific MEK inhibitor UO126 (10 M). As expected, UO126 inhibited phosphorylation of p44/42 MAPK (Fig. 3D, left panel). Whereas up-regulation of Smad1 expression levels were not affected by this treatment, UO126 abrogated phosphorylation of Smad1. Furthermore, both basal and AdTGF-␤RI-induced collagen protein levels were greatly reduced. In contrast, treatment with siRNA directed against Smad1 had no effect on p44/42 MAPK activation (Fig. 3D, right  panel). Together, these data strongly suggest that activation of the ERK1/2 pathway is required for the sustained activation of Smad1 signaling.
TGF-␤RI Overexpression Stimulates CTGF Promoter Activity in an ERK1/2 and Smad1-dependent Manner-To further dissect the molecular mechanism involved in profibrotic effects of AdTGF-␤RI we utilized CTGF promoter. We first examined the effects of TGF-␤RI overexpression on CTGF promoter activity. Normal adult dermal fibroblasts were transduced with AdTGF-␤RI (m.o.i. 25) followed by transient transfection with 1 g of wild type CTGF-luc promoter and assayed for luciferase activity 48 h later. Elevated TGF-␤RI levels potently stimulated CTGF promoter activity. TGF-␤ was included in our experiment as a positive control (Fig.  4A). A marked stimulation of CTGF promoter activity was also observed with the ALK5(3A) mutant. Furthermore, consistent with mRNA and protein data, blockade of either ERK1/2 using UO126 or Smad1 using AdSmad1 siRNA com-  25 and 50) and were subsequently treated with 400 nM of a specific RI kinase inhibitor, SD-208 (Scios, CA), where indicated. Samples were analyzed for phospho-and total p44/42 MAPK levels. D, ERK1/2 activation is responsible for the Smad1-mediated stimulatory effect of AdRI on collagen and CTGF. Dermal fibroblasts were treated with either AdGo or AdRI and then with either UO126 (10 M) (left panel) or siRNA against Smad1 (right panel). The cells were harvested after 48 h, lysed, and separated via SDS-PAGE. Western blot analysis was carried out to determine phospho-and total p44/p42 MAPK, phospho-and total Smad1, and collagen type I levels. The protein levels were normalized by probing for ␤-actin.
pletely abrogated stimulation of the CTGF promoter in response to TGF-␤RI overexpression (Fig. 4, B and C).
CTGF promoter has been partially characterized and several response elements have already been mapped, including Smad3 and Sp1 binding sites. The Smad binding site mediates up-regulation of the CTGF promoter activity by TGF-␤ (32). In scleroderma fibroblasts, the up-regulation of CTGF gene expression has been reported to be independent of the Smad binding site (19). Moreover, of the two putative Sp1 binding sites in the CTGF promoter, only one was reported to be "functional" in the up-regulation of CTGF in scleroderma fibroblasts. To examine the contribution of the specific response element in the CTGF promoter to the TGF-␤RI-dependent up-regulation, we generated site-specific mutants of the known response elements in this promoter. As shown in Fig. 4C, mutation of the functional Sp1 site completely abrogated the potent stimulatory effect of TGF-␤RI overexpression on CTGF promoter activity. In addition, mutation in the T␤RE element (33) partially abrogated TGF-␤RI-dependent activa-tion of the CTGF promoter. The transcription factors that bind to the T␤RE response element are currently unknown. Mutation of the CAGA motif (Smad3 binding site) had no effect (Fig. 4C). Together, these promoter analyses have shown that the proximal GC-rich motif mediates TGF-␤RI stimulation of the CTGF promoter. Because previous studies have shown that in selected promoters Smad1 mediates its response via GC-rich motifs (34), we investigated whether the proximal GC-rich motif mediates Smad1 stimulation of the CTGF/CCN2 promoter. Co-transfection of the wild type promoter with Smad1 resulted in increased promoter activity, whereas mutation of the GC-rich proximal motif completely abolished stimulatory effects of Smad1 (Fig. 4D). These data demonstrate that Smad1 through the GC-rich motif mediates transcription activation of the CTGF gene downstream from TGF-␤RI.

Reduction of TGF-␤RII Expression Levels Leads to Activation of Profibrotic Response in Dermal
Fibroblasts-Previous studies have shown that the increased ratio of TGF-␤RI:RII levels in scleroderma fibroblasts correlates with increased collagen production (14,21). Likewise, in our experimental model used in this study, a 2-3-fold increase of TGF-␤RI levels over the endogenous RI levels triggered the profibrotic gene program in healthy dermal fibroblasts (21). The remaining unresolved issue is whether the primary determinant of these effects is the overall elevated expression of TGF-␤RI or the change in the ratio of type I to type II receptor subunits. To answer this question, a second approach was used to modulate the receptor subunit ratio. The expression levels of TGF-␤RII were down-regulated using adenoviral vector expressing siRNA directed against TGF-␤RII. As shown in Fig. 5, A and B, AdRII siRNA inhibited TGF-␤RII mRNA and protein levels in a dose-dependent manner. An increase in collagen production was observed within a narrow range of RII down-regulation. Similar to the results obtained with overexpression of the RI receptor we also observed activation of ERK1/2 and Smad1 pathways (Fig. 5C). However, the response was not as robust as that observed with TGF-␤RI overexpression. These experiments suggest that most likely both the ratio of the RI to RII receptor as well as the overall expression levels of the receptors, contribute to the observed profibrotic effects.

DISCUSSION
TGF-␤ is considered a central mediator of the excessive collagen deposition, a key feature of organ fibrosis. The molecular mechanisms underlying this process are not fully elucidated. We utilized an in vitro model of scleroderma based on overexpression of TGF-␤RI to gain additional insights into the mechanism of pathologic fibrosis. The findings of this study provide evidence for the pro-fibrogenic role of Smad1 and ERK1/2 pathways in this experimental model. Using several independent approaches, we determined that activation of the Smad2/3 pathway is not required for the AdTGF-␤RI stimulation of collagen and CCN2 expression. Specifically, we have shown that in contrast to stimulation of cells with TGF-␤, neither Smad2 nor Smad3 became phosphorylated in response to AdTGF-␤RI. We have also shown that the TGF-␤RI (ALK5) mutant that does not associate with Smad3, potently stimulated collagen protein levels and CCN2 promoter activity. Consistent with these findings, pharmacological ALK5 kinase inhibitors were ineffective in blocking profibrotic response in this model (21). The following findings supported the involvement of the ALK1/Smad1 pathway in this model. TGF-␤RI overexpression resulted in constitutive ALK5/ALK1 complex formation and persistent phosphorylation of Smad1. Furthermore, Smad1-specific siRNA abolished elevated type I collagen and CCN2 expression in this model. Importantly, a previously characterized GC-rich response element, which mediated up-regulation of the CCN2 promoter in scleroderma fibroblasts, was also shown to be required for Smad1 stimulation of the CCN2 gene in response to AdTGF-␤RI overexpression. Together, these data demonstrate that activation of the ALK1/Smad1 signaling pathway plays a central role in regulation of the profibrotic gene program in this model of fibrosis.
TGF-␤-dependent activation of the Smad1 pathway has been first described in endothelial cells (35). In these cells TGF-␤ signals through the heteromeric receptor complex that includes TGF-␤RII, ALK5, and ALK1 (36). Whereas the specific interactions between ALK5 and ALK1 are still not fully elucidated, there is evidence that the relative ratio of these two receptor subtypes regulate the switch between proliferation and differentiation of endothelial cells (3). Interestingly, recent studies have also shown involvement of ALK1/Smad1 pathways in kidney and liver fibrosis. Elevated expression of Smad1 protein was observed in human diabetic kidney and in animal models of diabetic nephropathy (37)(38)(39). In addition, Smad1 was shown to directly up-regulate expression of the collagen type IV gene in mesangial cells (37). Similarly, TGF-␤ signaling via activation of the ALK1/Smad1 pathway and subsequent up-regulation of Id1 gene has been shown to contribute to transdifferentiation of hepatic stellate cells into myofibroblasts (40). Smad1 phosphorylation and Id1 expression correlated with severity of bile duct ligation-induced liver fibrosis (40). These recent studies strongly suggest that activation of ALK1/Smad1 may play an important role in development of organ fibrosis.
In our study we focused on transcriptional regulation of CCN2, a profibrogenic cytokine up-regulated in various fibrotic diseases. In agreement with earlier studies, we showed that TGF-␤-dependent CCN2 stimulation is mediated through a Smad3 response element, whereas in fibroblasts overexpressing TGF-␤RI, CCN2 is primarily regulated via a GC-rich motif that mediates Smad1 response. In addition, we demonstrated that activation of the ERK1/2 pathway, which plays a relatively small role in the TGF-␤-dependent CCN2 stimulation in dermal fibroblasts, was required for AdTGF␤-RI-dependent persistent phosphorylation of Smad1 and a subsequent up-regulation of CTGF and collagen genes. Similarly, activation of ERK1/2 was required for the sustained phosphorylation of Smad1 in cardiomyocytes (41). Furthermore, concomitant activation of Smad1 and ERK1/2 pathways was observed in hepatic stellate cells in a bile duct ligation-induced liver fibrosis, suggesting that interaction between Smad1 on ERK pathways may be a general phenomenon (42).
In conclusion, this study provides evidence for the existence of the alternative TGF-␤-dependent, Smad3-independent signaling pathway that may operate during chronic stages of SSc fibrosis. Fibroblasts from Smad3 null mice also show a compensatory increase in activated MAPK levels in response to TGF-␤ (43), thus raising the possibility of a similar Smad3-independent pathway functioning in vivo during wound healing in this mouse model. As shown by our study activation of this pathway is triggered in SSc fibroblasts in the absence of exogenous ligand (21), indicating that SSc fibroblasts are self-sufficient in establishing this profibrotic cascade. This hypothesis is consistent with previous in vivo findings that showed the elevated presence of TGF-␤ only in the early SSc lesions and lack of detectable TGF-␤ in subsequent chronic stages, which are characterized by elevated ECM production (44).
Altered ratio of TGF-␤ receptor subunits have also been observed in other fibrotic diseases. Importantly, modulation of the TGF-␤RI/RII ratio also takes place in fibroblasts during wound healing. Whereas TGF␤RII levels are more potently induced earlier during the initial proliferative stages of wound healing and subsequently decline, TGF␤RI levels increase later and remain elevated at the end of the proliferative stage with the onset of the synthetic phase during which matrix deposition increases (45). In addition, human dermal fibroblasts cultured in hypoxic conditions exhibit an increased TGF-␤RI to RII ratio. Therefore, the increased RI/RII ratio that we observe in the majority of SSc fibroblasts derived from patients with early disease may be a manifestation of the hypoxic environment that is present in SSc skin due to the vascular damage associated with this disease (46). The observed alterations of TGF-␤ receptor levels, together with the mechanistic insights described in this study, suggest that the modulation of the RI to RII ratio may represent a novel mode of diversifying TGF-␤ signaling and its biologic effects, which occurs both in physiologic and pathologic conditions. Furthermore, the ALK1/Smad1 pathway may represent, in addition to Smad3, a potential target for the antifibrotic therapy in SSc.