Expression and regulation of intracellular SMAD signaling in scleroderma skin fibroblasts.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 48, No. 7, July 2003, pp 1964–1978 DOI 10.1002/art.11157 © 2003, American College of Rheumatology Expression and Regulation of Intracellular SMAD Signaling in Scleroderma Skin Fibroblasts Yasuji Mori, Shu-Jen Chen, and John Varga Objective. Scleroderma is characterized by excessive synthesis and accumulation of matrix proteins in lesional tissues. Transforming growth factor ␤ (TGF␤) plays a central role in the pathogenesis of fibrosis by inducing and sustaining activation of fibroblasts; however, the underlying mechanisms are poorly understood. We undertook this study to examine the expression and function of SMADs, recently characterized intracellular effectors of TGF␤ signaling, in scleroderma fibroblasts. Methods. Primary dermal fibroblasts obtained from 14 patients with scleroderma and from 4 healthy adult volunteers were studied. Northern analysis was used to determine the expression of endogenous SMAD messenger RNA (mRNA), and Western analysis was used to determine SMAD protein expression. Intracellular compartmentalization of cellular SMAD proteins in the presence and absence of TGF␤ was studied by antibody-mediated immunofluorescence confocal microscopy. The effect of TGF␤ blockade on SMAD subcellular distribution was determined using anti-TGF␤ antibodies as well as a dominant-negative TGF␤ receptor type II (TGF␤RII) vector to disrupt TGF␤ responses. SMAD-regulated luciferase reporter expression was examined to investigate the potential functional significance of activation and nuclear accumulation of endogenous SMADs in scleroderma fibroblasts. Results. Protein and mRNA levels of SMAD3, but not of SMAD4 or SMAD7, were variably elevated in scleroderma fibroblasts compared with those from healthy controls. In sharp contrast to control fibroblasts, which displayed predominantly cytoplasmic localization of SMAD3/4 in the absence of exogenous TGF␤, in scleroderma fibroblasts SMAD3 and SMAD4 consistently showed elevated nuclear localization. Furthermore, phosphorylated SMAD2/3 levels were elevated and nuclear localization of phosphorylated SMAD2/3 was increased, suggesting activation of the SMAD pathway in scleroderma fibroblasts. Blockade of autocrine TGF␤ signaling with antibodies or by expression of dominant-negative TGF␤RII failed to normalize SMAD subcellular distribution, suggesting that elevated nuclear SMAD import was due to alterations downstream of the TGF␤ receptors. The activity of a SMADresponsive minimal promoter–reporter construct was enhanced in transiently transfected scleroderma fibroblasts. Conclusion. This study is the first to demonstrate apparently ligand-independent constitutive activation of the intracellular TGF␤/SMAD signaling axis in scleroderma fibroblasts. SMAD signaling may be a mechanism contributing to the characteristic phenotype of scleroderma fibroblasts and playing a role in the pathogenesis of fibrosis. Scleroderma is associated with fibrosis of affected tissues due to excessive synthesis and progressive accumulation of connective tissue macromolecules (1). Scleroderma fibroblasts display features of sustained activation in vitro and in vivo. In culture, these cells show elevated synthesis of collagens, fibronectin, proteoglycans, and tissue inhibitors of metalloproteinases, constitutive secretion of interleukin-1␣ (IL-1␣) and IL-6, and reduced production of collagenase (for review, see ref. 2). These phenotypic alterations closely correspond to changes in normal fibroblasts stimulated with transforming growth factor ␤ (TGF␤) and persist during serial passage for a limited number of replications in vitro. The mechanisms responsible for the activated phenotype of scleroderma fibroblasts remain unknown. Supported by NIH grant AR-42309 and by the Scleroderma Foundation. Yasuji Mori, MD, PhD, Shu-Jen Chen, PhD, John Varga, MD: University of Illinois at Chicago College of Medicine. Address correspondence and reprint requests to John Varga, MD, Section of Rheumatology (M/C 733), University of Illinois College of Medicine, Room 1158 Molecular Biology Research Building, 900 South Ashland Avenue, Chicago, IL 60607. E-mail: email@example.com. Submitted for publication October 28, 2002; accepted in revised form March 12, 2003. 1964 SMAD SIGNALING IN SCLERODERMA FIBROBLASTS TGF␤, the principal mediator of physiologic tissue remodeling, is strongly implicated in the pathogenesis of scleroderma (3). The expression of TGF␤ is elevated in scleroderma dermal fibroblasts, as well as in monocyte/macrophages and other infiltrating inflammatory cells, even prior to the appearance of fibrosis (4–11). Levels of TGF␤ receptor type I (TGF␤RI) and TGF␤RII are also elevated on scleroderma fibroblasts (12–14). Tissue expression of TGF␤ is prominent in animal models of scleroderma, and blockade of TGF␤ signaling with neutralizing antibodies or naturally occurring TGF␤ antagonists is effective in reducing collagen accumulation and ameliorating experimental fibrosis (15,16). Furthermore, TSK mice that have been genetically engineered to be deficient for one TGF␤ allele show marked reduction in the thickness of the dermis and accumulation of matrix (17). In fibroblasts, TGF␤ stimulates matrix production and induces its own synthesis (autoinduction) as well as that of connective tissue growth factor (CTGF) (18–20). Recent studies have shed light on the intracellular signaling mechanisms that mediate relevant profibrotic TGF␤ responses. Binding of TGF␤ to the ubiquitous type II serine/threonine kinase receptor (TGF␤RII) triggers heterodimerization with and activation of TGF␤RI. The signal is then propagated downstream through SMADs, a family of evolutionarily conserved intracellular mediators that convey information from the cell membrane into the nucleus (21). In vertebrates, members of the SMAD family segregate into 3 functionally distinct groups: SMADs 2 and 3 are direct substrates for activated TGF␤RI, SMAD4 is the common signaling partner, and SMAD7 is inhibitory. Upon their phosphorylation and activation by the TGF␤ receptor kinase at the cell membrane, SMAD2 and SMAD3 associate with SMAD4. The heteromeric SMAD complex then translocates into the nucleus, where, in conjunction with other DNA binding factors or with transcriptional coactivators or corepressors, it regulates target gene transcription. While SMAD2 and SMAD3 work synergistically and in tandem with SMAD4 in transducing TGF␤ signals, the antagonistic SMAD7 binds to the TGF␤ receptor complex and prevents SMAD2/3 access, thereby blocking SMAD phosphorylation. Because its expression is rapidly induced by TGF␤ (22–24), SMAD7 serves a critical negative feedback function by limiting the amplitude and duration of cellular responses to TGF␤. In addition to inhibitory SMAD7, multiple other intracellular mechanisms to modulate the intensity of SMAD signaling have been characterized (25–28). In normal fibroblasts, 1965 SMAD3 and SMAD4 are predominantly cytosolic; upon stimulation, these SMADs reversibly translocate into the nucleus (24). Nuclear import is required for SMAD-dependent transcriptional responses. The relative subcellular compartmentalization of SMAD3 and SMAD4 is thus an important factor in setting the level of TGF␤ signaling, but the mechanisms governing SMAD nuclear–cytoplasmic shuttling in physiologic TGF␤ responses are poorly understood. The SMAD pathway plays a fundamental role in regulation of collagen synthesis, and SMADs are necessary to mediate TGF␤-dependent stimulation (29–33). In light of the singular importance of TGF␤ in both initiating and sustaining fibroblast activation, and given the pivotal role of SMADs as major intracellular effectors of TGF␤-induced profibrotic responses, we undertook extensive characterization of SMAD signaling in a large panel of scleroderma fibroblasts. The present results indicate that messenger RNA (mRNA) and protein expression of the positive signal mediator SMAD3, but not those of inhibitory SMAD7, are variably elevated in scleroderma fibroblasts. Treatment with TGF␤ caused repression of SMAD3 expression and stimulation of SMAD7 expression in both healthy control and scleroderma fibroblasts. In sharp contrast to control fibroblasts, however, scleroderma fibroblasts were characterized by substantial increase in endogenous SMAD2/3 phosphorylation and nuclear accumulation in the absence of TGF␤ stimulation. Nuclear redistribution did not appear to be driven by autocrine TGF␤, because blockade of TGF␤ signaling failed to diminish the proportion of scleroderma fibroblasts showing nuclear SMAD localization. Collectively, these results demonstrate ligand-independent constitutive activation of the SMAD signaling pathway downstream of the TGF␤ receptors in scleroderma fibroblasts, suggesting a novel mechanism that may contribute to the pathogenesis of fibrosis. PATIENTS AND METHODS Cell culture. Primary dermal fibroblasts were established from excisional skin biopsy samples using previously described procedures (29). In these experiments, 14 patients with scleroderma were studied. The diagnosis of scleroderma was established according to the classification criteria of the American College of Rheumatology (formerly, the American Rheumatism Association) (34). The mean duration of disease at the time of biopsy was 12 months, and 3 patients were receiving disease-modifying therapy (methotrexate, prednisone, or D-penicillamine). Biopsy samples were obtained from the leading edge of clinically apparent “lesional” skin. 1966 MORI ET AL Table 1. Clinical characteristics of patients with scleroderma Patient/age/sex 1/56/F 2/35/F 3/46/F 4/59/F 5/39/F 6/61/M 7/50/M 8/38/F 9/47/F 10/43/F 11/19/F 12/19/F 13/30/F 14/44/M Race* Disease duration, years† Skin involved, %‡ B H B O W W H W W W H B W W 1 1 1 1 1 1 2 1 1 1 0.5 1 1 1 20 10 5 5 50 5 50 36 40 5 40 15 10 20 * B ⫽ black; H ⫽ Hispanic; O ⫽ Oriental; W ⫽ white. † Interval from first non-Raynaud symptom of scleroderma to time of skin biopsy. ‡ Extent determined by clinical examination and recorded utilizing a body surface diagram. The characteristics of the patients are shown in Table 1. As a control, skin biopsy samples were obtained from 4 healthy adult volunteers (1 man and 3 women, mean age 45 years). Biopsies were performed with written informed consent, and the protocol was approved by the Institutional Review Board of the University of Illinois. Culture media were from BioWhittaker (Walkersville, MD); all other tissue culture reagents were from Gibco BRL (Grand Island, NY). For these experiments, control and scleroderma fibroblasts were grown simultaneously and studied between passages 4 and 8. Cells were grown at 37°C in a 5% CO2 atmosphere in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS), 1% vitamin solution, 100 units/ml penicillin/streptomycin, and 2 mM L-glutamine. When the fibroblasts reached confluence, fresh medium with indicated concentrations of TGF␤1 (Amgen, Thousand Oaks, CA) was added. In some experiments, cultures were incubated with panspecific neutralizing antibody against TGF␤ (R&D Systems, Minneapolis, MN) or nonimmune rabbit IgG as control, alone or in combination with recombinant latency-associated peptide (LAP), a naturally occurring TGF␤ antagonist (R&D Systems), for 0.5 hours prior to addition of fresh medium. In other experiments, scleroderma or control fibroblasts were transduced with adenoviral vector expressing green fluorescent protein (GFP) as control, or with dominant-negative kinase-deficient TGF␤RII to inhibit signal transduction from all 3 mammalian TGF␤ isoforms (35). Transduction was carried out by incubation of confluent fibroblasts for 48 hours at room temperature with 3 ⫻ 109 plaque-forming units/ml of virus. Efficient transduction was confirmed by high levels of GFP expression in all fibroblasts. Extraction and analysis of RNA. At the end of each experiment, total RNA was isolated from fibroblasts with TRIzol reagent (Gibco BRL). Relative levels of mRNA were examined by Northern analysis using ␥32P-dCTP–labeled com- plementary DNA (cDNA) probes, as described (29). Following extensive washing of the nitrocellulose membranes, the cDNA–mRNA hybrids were visualized by autoradiography on X-AR5 film (Eastman Kodak, Rochester, NY) exposed for 24–72 hours with intensifying screens. The following human cDNA probes were used for hybridization: a 1.4-kb restriction fragment that included the entire coding region of human SMAD3, a 1.6-kb restriction fragment that included the entire coding region of human SMAD4, a 1.9-kb restriction fragment that included the entire coding region of human SMAD7, GAPDH, and 18S. Signal intensities were quantitated by densitometry, and results were normalized to the levels of GAPDH or 18S mRNA in each sample. Western analysis of cellular SMADs. Immunoblot of whole cell lysates was performed as previously described (29). To prevent protein dephosphorylation, a Phosphatase Inhibitor Mix (Sigma, St. Louis, MO) was added. Nuclear extracts were prepared as previously described (31). Protein concentrations in the cytosolic, nuclear, or whole cell fractions were determined by Bradford assay (Bio-Rad, Hercules, CA). Equal aliquots of whole cell lysates (15 g/lane) or cytosolic or nuclear extracts (20 g/lane) were separated by reducing electrophoresis in 10% sodium dodecyl sulfate (SDS)– polyacrylamide gels. Proteins in the gels were transferred onto Immobilon-P (polyvinylidene difluoride) membranes (Millipore, Bedford, MA). Following blocking with 5% nonfat dry milk in Tris buffered saline–Tween at room temperature for 1 hour, membranes were incubated for 2 hours with antibodies against SMAD3 (Zymed, South San Francisco, CA) or against SMAD2/3, phospho-SMAD2/3, SMAD4, SMAD7, TGF␤RI, histone H3, or actin (all from Santa Cruz Biotechnology, Santa Cruz, CA), followed by incubation with horseradish peroxidase (HRP)–conjugated secondary antibodies. Antibody specificities were confirmed using blocking peptides supplied by the manufacturer. After washing, immunoblots were developed with chemiluminescence reagents (Pierce, Rockford, IL) according to the manufacturer’s protocol. In order to detect phosphorylated TGF␤RI or SMAD2/3 in fibroblasts, anti-TGF␤RI or anti-SMAD2/3 antibodies with protein G–Sepharose were used for immunoprecipitation. The immunoprecipitates from lysates of fibroblasts left untreated or treated either with TGF␤ (500 pM) or with anti-TGF␤ antibody were eluted by boiling, electrophoresed through 8% SDS–polyacrylamide gels, and processed as described above. Following overnight blocking, membranes were incubated with antiphosphoserine antibody (Zymed). Signal intensities were quantitated by densitometry, and results were normalized to the intensities of IgG or histone H3 bands in each sample. Cellular immunofluorescence imaging. The expression and intracellular localization of endogenous SMAD proteins were studied by immunocytochemistry and fluorescence confocal microscopy. For this purpose, fibroblasts (10,000/well) from healthy controls or patients with scleroderma were seeded into 8-well Lab-Tek II chamber glass slides (Nalge Nunc International, Naperville, IL). The next day, fresh media with 0.1% FCS and indicated concentrations of TGF␤1 were added for 2 hours. Cells were washed and fixed with 100% methanol for labeling of SMADs. Cells were then incubated with antibodies (10 g/ml) against SMAD2/3 (from Santa Cruz Biotechnology or Zymed; both yielding identical results), SMAD SIGNALING IN SCLERODERMA FIBROBLASTS 1967 SMAD4 or phospho-SMAD2/3 (both from Santa Cruz Biotechnology), or TGF␤RII (Santa Cruz Biotechnology) for 1 hour and then washed, followed by incubation with HRPconjugated secondary antibodies and staining with fluorescein isothiocyanate or rhodamine according to the manufacturer’s protocol (Tyramide Signal Amplification; NEN Life Science Products, Boston, MA). To stain the nuclei, chambers were mounted with Vectashield plus 4⬘,6-diamidino-2-phenylindole (Vector, Burlingame, CA). Nonimmunized IgG was used as negative control. Following stringent washing of the slides, the pattern and subcellular distribution of fluorescence were evaluated by immunofluorescence or laser scanning confocal microscopy. Each experiment was repeated at least 3 times with consistent results. Quantitative analysis was performed by scoring 100 individual fibroblasts from different microscopic fields as showing a predominantly nuclear or predominantly cytoplasmic distribution of immunofluorescence. The observer was blinded to the identity of each section. Transient transfection. Subconfluent cultures of fibroblasts were transfected using Superfect reagent (Qiagen, Valencia, CA), as described previously (35). The plasmid pSBE4luc (from Dr. B. Vogelstein), containing 4 tandem repeats of an 8-bp palindromic consensus SMAD binding element (SBE) that is specifically recognized by SMAD2 or SMAD3 (36), was used to measure SMAD-mediated transcriptional responses. pGL3-Luc, the empty vector of pSBE4-luc and pRL-TK (both from Promega, Madison, WI) were used as internal controls to correct for variations in transfection efficiency between the samples. Following their recovery overnight in DMEM with 10% FCS, fibroblasts were incubated in fresh media containing 0.1% FCS. After 48 hours, cells were harvested and lysates were prepared using passive lysis buffer (Promega). Luciferase activity in equal aliquots was determined by Dual-Luciferase Reporter Assay System (Promega). All experiments were performed in triplicate and repeated at least 3 times. Significance of differences between experimental groups was determined by Mann-Whitney U test. P values less than 0.05 were considered significant. RESULTS Expression and regulation of endogenous SMAD mRNA in scleroderma fibroblasts. We previously demonstrated that in normal skin fibroblasts, SMAD3 and SMAD4 played fundamental roles in mediating TGF␤ responses, including stimulation of collagen synthesis (29,30). To begin to elucidate their role in the pathogenesis of fibrosis, we examined the expression of endogenous SMADs in scleroderma fibroblasts. Total RNA was isolated from low-passage confluent fibroblasts and examined by Northern analysis. The results showed that control and scleroderma fibroblasts similarly expressed two specific SMAD3 mRNA transcripts of 7.0 kb and 3.0 kb (Figure 1A). Levels of SMAD3 mRNA were moderately elevated in scleroderma fibroblasts (mean ⫾ SEM 167 ⫾ 51% of those in control Figure 1. SMAD mRNA expression in fibroblasts. Fibroblasts derived from lesional skin of 14 patients with scleroderma (S) or 4 healthy controls (N) were cultured in parallel. At early confluence, cells were harvested and total RNA was isolated. Levels of SMAD3, SMAD7, 18S, and GAPDH mRNA were determined by Northern blot analysis. A, Representative Northern blot. Asterisks indicate the positions of 18S and 28S ribosomal RNA. B, SMAD3 and SMAD7 mRNA levels were quantitated by densitometric scanning of the autoradiographs. Results in the upper panel, corrected for the levels of GAPDH mRNA in each sample, are expressed in arbitrary units. Each data point represents the mean of several independent experiments for each individual scleroderma (solid bars) and control (open bars) fibroblast line. Lower panel shows the ratio of SMAD3 mRNA level:SMAD7 mRNA level calculated for each fibroblast line. Middle and upper/lower lines represent the mean ⫾ SEM. fibroblasts), although the difference did not reach significance; in 11 of 14 scleroderma fibroblast lines, SMAD3 mRNA levels were ⬎1 SEM higher than the mean in controls (Figure 1B). 1968 In contrast to SMAD3, the expression of SMAD4 mRNA (9.5 kb and 3.9 kb) and SMAD7 mRNA (4.6 kb) showed no consistent differences between control and scleroderma fibroblasts, although SMAD7 mRNA levels in 5 scleroderma fibroblast lines were ⬎1 SEM lower than the mean in controls (Figure 1B). The ratio of the mRNA levels of positive (SMAD3) and negative (SMAD7) mediators of SMAD signaling (SMAD3 mRNA:SMAD7 mRNA ratio) was increased in scleroderma fibroblasts (mean ⫾ SEM 160 ⫾ 46% of the ratio in control fibroblasts), although the difference was not statistically significant. These results were reproducible in several independent experiments, and they were consistent when individual fibroblast lines were examined at different passages (passages 4 and 7). In contrast to catalytic intracellular second messenger systems, the SMADs are devoid of intrinsic enzymatic activity, and therefore the SMAD pathway has no potential for amplifying input signals. Consequently, TGF␤-induced cellular responses are sensitive to small changes in SMAD steady-state levels (37). We therefore examined the regulation of cellular SMAD mRNA expression by TGF␤. The results showed that treatment with TGF␤ (500 pM) caused a selective reduction in SMAD3 mRNA expression in normal fibroblasts. After 48 hours of treatment, the SMAD3 mRNA steady-state levels were decreased to 38 ⫾ 6% (mean ⫾ SEM) of those in untreated fibroblasts. Failure to suppress SMAD3 expression could potentially result in uncontrolled TGF␤ signaling due to a relative increase in SMAD3 levels and persistent availability of SMAD3 for activation by the TGF␤ receptor kinase. We therefore examined whether the regulation of SMAD3 mRNA expression by TGF␤ was altered in scleroderma fibroblasts. As shown in a representative Northern blot (Figure 2A), SMAD3 steady-state mRNA levels in scleroderma fibroblasts were markedly reduced by TGF␤ (to 34 ⫾ 9% of the levels in untreated fibroblasts). An essentially identical magnitude of TGF␤induced decrease in SMAD3 mRNA was observed in 4 individual scleroderma fibroblast lines examined. Inhibitory SMAD7 competes with SMAD3 for access to the activated TGF␤ receptor, thereby preventing SMAD3 phosphorylation and blocking TGF␤induced responses in fibroblasts (29). Negative feedback inhibition of TGF␤/SMAD signaling mediated through endogenous SMAD7 limits the magnitude and/or duration of cellular responses to TGF␤, and rapid induction of SMAD7 is therefore essential for preventing excessive TGF␤ stimulation. Defective induction of cellular SMAD7 results in unopposed SMAD signaling, leading MORI ET AL Figure 2. Regulation of SMAD mRNA expression by transforming growth factor ␤ (TGF␤). Fibroblast lines obtained from lesional skin of scleroderma patients or from skin of healthy controls were cultured in parallel. When fibroblasts reached confluence, TGF␤1 (500 pM) was added to the cultures. After the indicated period of incubation, cultures were harvested and relative mRNA levels were determined by Northern analysis. A, SMAD3 and SMAD7 mRNA levels determined in fibroblasts incubated (solid bars in lower panel) or not incubated (open bars in lower panel) with TGF␤ for 24 hours (SMAD3) or 90 minutes (SMAD7). An autoradiograph representative of multiple independent experiments is shown in the upper panel. Levels of mRNA were quantitated by densitometric scanning of the autoradiographs (lower panel). Results represent the mean and SEM from 4 pairs of scleroderma (S) and normal (N) fibroblast lines corrected for the levels of GAPDH mRNA in each sample. B, Kinetics of SMAD7 mRNA induction by TGF␤ in control (left panel) and scleroderma (right panel) fibroblasts. Note the difference in scales between the two panels. Results for 7 individual scleroderma fibroblast lines are shown. to exaggerated or sustained TGF␤ responses. We showed previously that transient disruption of endogenous SMAD7 function using antisense oligonucleotides SMAD SIGNALING IN SCLERODERMA FIBROBLASTS resulted in increased TGF␤ stimulation of COL1A2 transcription in normal fibroblasts (38). We therefore examined the possibility that defective SMAD7 induction could account for the activated phenotype of scleroderma fibroblasts. As shown in Figure 2A, TGF␤ induced a significant early rise in SMAD7 mRNA levels. Quantitation of results from several independent experiments showed that in scleroderma and control fibroblast lines treated with TGF␤ for 90 minutes, SMAD7 mRNA levels were increased to 261 ⫾ 42% and 256 ⫾ 35% (mean ⫾ SEM), respectively, of those in untreated fibroblasts. Next, in a separate set of experiments, we carefully compared the kinetics of SMAD7 induction in control (n ⫽ 4) and scleroderma (n ⫽ 7) fibroblast lines in parallel. As shown in Figure 2B, there was a similar time course of SMAD7 induction, with maximal TGF␤ response at 120 minutes both in control and in scleroderma fibroblast lines. Significantly, none of the examined scleroderma fibroblast lines failed to respond to TGF␤ with a ⬎2-fold increase in SMAD7 mRNA expression. Taken together, these results clearly indicate that control and scleroderma fibroblasts are characterized by comparable sensitivity to TGF␤, with similar magnitude and kinetics of delayed down-regulation of SMAD3 mRNA expression and rapid up-regulation of SMAD7 mRNA expression. SMAD protein expression and activation in scleroderma fibroblasts. The cellular steady-state levels of SMAD proteins are determined by the rates of transcription of the corresponding genes, as well as by posttranscriptional regulatory mechanisms. SMADs are subject to intracellular proteolysis mediated through the ubiquitin ligase proteosome pathways in both liganddependent and ligand-independent manners (28). Therefore, it was important to examine the levels of SMAD proteins in control and scleroderma fibroblasts. To this end, whole cell lysates were prepared from confluent control and scleroderma fibroblasts in parallel and subjected to immunoblot analysis. In Figure 3A, a Western blot from a representative experiment demonstrates bands at ⬃52 kd (the expected size of SMAD3) and at ⬃45 kd (the expected size of SMAD7). Equal loading and transfer of protein in each lane of the Western blots was confirmed using antiactin antibody. Densitometric analysis of the results demonstrated that 5 of 11 separate scleroderma fibroblast lines consistently displayed elevated SMAD3 protein levels compared with control fibroblasts. Generally, a good correlation between relative levels of SMAD3 protein and mRNA expression was found in individual fibroblast lines. In contrast to SMAD3, the levels of 1969 Figure 3. SMAD protein expression in fibroblasts. Fibroblasts from normal control or lesional scleroderma skin were cultured in parallel. At early confluence, cells were harvested and identical amounts of whole cell lysates were electrophoresed and subjected to immunoblotting. A, A representative Western blot. Antibodies against SMAD2/3, SMAD7, and actin were used. Numbers at the bottom indicate control (N) or scleroderma (S) fibroblast lines. B, Increased SMAD phosphorylation in scleroderma fibroblasts. Whole cell lysates were prepared from confluent control (lanes 1–3 [N3] and 4 [N1]) or scleroderma (lanes 5 [S3], 6 [S5], 7 [S11], and 8 [S12]) fibroblasts in parallel. In lanes 2 and 3, fibroblasts were incubated with transforming growth factor ␤ for 30 minutes and 60 minutes, respectively. Proteins were immunoprecipitated with antibody against SMAD2/3. The complexes were electrophoresed through sodium dodecyl sulfate–polyacrylamide gel, followed by immunoblotting with antiphosphoserine antibody (p-Serine). Whole cell lysates were analyzed by immunoblotting for IgG. A representative blot is shown. Numbers at the bottom indicate the ratio of phospho-SMAD2/3 to total SMAD2/3 in each lane, determined by quantitating the signal intensities by densitometry and normalizing results to IgG to correct for small variations in protein loading and transfer. SMAD4 and SMAD7 proteins were similar in control and scleroderma fibroblasts (Figure 3A and results not shown). TGF␤ treatment resulted in a marked decrease of SMAD3 protein at 48 hours and in an elevation of SMAD7 protein at 4 hours both in control and in scleroderma fibroblasts, whereas SMAD4 expression remained unchanged. These changes in protein closely paralleled the TGF␤-induced changes in SMAD mRNA levels shown in Figure 2. Upon TGF␤ stimulation, cellular SMAD2 and SMAD3 become rapidly phosphorylated by the activated TGF␤ receptor serine/threonine kinase. The phosphorylation state of endogenous SMAD2/3 thus serves 1970 as a marker for the activation of the TGF␤/SMAD pathway. We therefore sought to determine the relative phosphorylation state of receptor-activated SMADs in scleroderma fibroblasts. For this purpose, unstimulated quiescent control and scleroderma fibroblasts in parallel were lysed, and equal aliquots were immunoprecipitated with anti-SMAD2/3 antibody, followed by electrophoresis and immunoblotting with antiphosphoserine antibody. As shown in a representative immunoblot in Figure 3B, only low levels of SMAD2/3 phosphorylation could be detected in untreated control fibroblasts. Treatment of the fibroblasts with TGF␤ rapidly induced phosphorylation of SMAD2/3, as expected. The expression levels of total SMAD2/3 proteins were unaffected by treatment with TGF␤ for 60 minutes, as indicated by immunoblotting with the anti-SMAD2/3 antibody. In each of the scleroderma fibroblast lines studied, phosphorylated SMAD2/3 could be clearly detected in the absence of TGF␤ stimulation (Figure 3B). Quantitation of the immunoblot results from several independent experiments indicated that the ratio of phosphorylated SMAD2/3 to total SMAD2/3 in lysates from scleroderma fibroblasts (n ⫽ 4) was 1.62 ⫾ 0.47, compared with 0.19 ⫾ 0.05 in control fibroblasts (n ⫽ 2) (Figure 3B and data not shown). Nuclear accumulation of SMAD3 and SMAD4. Because SMAD-mediated TGF␤ signal transduction is critically dependent on translocation of SMAD3 and SMAD4 from the cytoplasm into the nucleus, the activity of the SMAD signaling pathway can be regulated by the spatial compartmentalization of endogenous SMADs. In order to examine subcellular SMAD distribution in scleroderma, SMAD3/4 nuclear accumulation was analyzed using antibody-mediated immunofluorescence with confocal microscopy. As shown in Figure 4A, SMAD3 showed predominantly nuclear distribution in ⬍10% of control fibroblasts in the absence of exogenous TGF␤. Treatment with TGF␤ resulted in a rapid increase in SMAD nuclear accumulation, which was maximal by 120 minutes (Figure 4A). Interestingly, we found that TGF␤ concentrations substantially below the threshold required for induction of transcriptional responses (such as stimulation of COL1A2 and plasminogen activator inhibitor 1) were capable of inducing SMAD3/4 nuclear translocation (e.g., a concentration of TGF␤ as low as 10 pM was sufficient to induce nuclear accumulation of endogenous SMAD3/4 in ⬎60% of normal fibroblasts ). MORI ET AL Figure 4. Increased nuclear SMAD accumulation in scleroderma fibroblasts. Control or scleroderma fibroblasts in parallel were incubated with transforming growth factor ␤1 (TGF␤1), fixed, and stained with fluorescein isothiocyanate (FITC)–conjugated SMAD3- or SMAD4-specific antibodies. A, Immunofluorescence confocal microscopic images from a representative experiment. Green color (FITC) indicates SMAD4; blue color (4⬘,6-diamidino2-phenylindole) indicates the nucleus. a, Untreated control fibroblasts; b, fibroblasts treated with TGF␤; c, untreated scleroderma fibroblasts; d, scleroderma fibroblasts treated with TGF␤ (original magnification ⫻ 100 in upper panels; ⫻ 630 in lower panels). B, The proportion of scleroderma or control fibroblasts unambiguously displaying a predominantly nuclear pattern of SMAD3 or SMAD4 immunofluorescence in the absence of added TGF␤ was quantitated by scoring 100 individual cells in each culture in a blinded manner. Results shown represent the mean and SEM from 5 individual scleroderma (solid bars) and 3 individual control (open bars) fibroblast lines. ⴱ ⫽ P ⬍ 0.05. C, Nuclear accumulation of SMAD2/3. Nuclear extracts were isolated from confluent scleroderma fibroblasts (S5, S12) or untreated or TGF␤-treated control fibroblasts (N1) and examined by immunoblot using antibodies to SMAD2/3 and histone H3. SMAD SIGNALING IN SCLERODERMA FIBROBLASTS Figure 5. Increased nuclear phospho-SMAD accumulation in scleroderma fibroblasts. Control or scleroderma fibroblasts incubated with or without antibody (Ab) to TGF␤ in parallel were fixed and stained with FITC-conjugated phospho-SMAD2/3 antibodies. A, Immunofluorescence confocal microscopic images from a representative experiment. Green color indicates phospho-SMAD2/3; blue indicates nuclei. a, Control fibroblasts; b, scleroderma fibroblasts; c, scleroderma fibroblasts incubated with anti-TGF␤ antibody (original magnification ⫻ 100 in upper panels; ⫻ 400 in lower panels). B, The proportion of fibroblasts displaying predominantly nuclear localization of phospho-SMAD2/3 in the absence of added TGF␤ was quantitated by scoring 100 individual cells in each culture in a blinded manner. Results shown represent the mean and SEM from 3 independent experiments with 4 individual scleroderma (solid bars) and 3 individual control (open bars) fibroblast lines. ⴱ ⫽ P ⬍ 0.05. See Figure 4 for other definitions. In contrast to control fibroblasts, 30–40% of unstimulated scleroderma fibroblasts showed predominantly nuclear localization of SMAD3 in the absence of exogenous TGF␤. The intensity of nuclear SMAD3 1971 immunofluorescence showed some variability among individual fibroblasts. Elevated nuclear accumulation of SMAD3 persisted during serial in vitro passages of the scleroderma fibroblasts. Results were essentially identical when subcellular distribution of SMAD4 was compared between control and scleroderma fibroblasts. Incubation of fibroblasts with irrelevant IgG as negative control resulted in minimal fluorescence. Quantitation of results from multiple experiments with several fibroblast lines indicated that the proportion of fibroblasts displaying predominantly nuclear SMAD3 localization was 3.3-fold greater in scleroderma cultures (n ⫽ 6) than in control cultures (n ⫽ 3); for SMAD4, this proportion was 3.7-fold greater in scleroderma cultures (Figure 4B). To confirm enhanced nuclear accumulation of SMAD3 in scleroderma fibroblasts, nuclear extracts were prepared from confluent cultures of control or scleroderma fibroblasts in parallel and examined by Western analysis. As shown in a representative immunoblot (Figure 4C), the accumulation of SMAD2/3 in the nucleus was ⬃7-fold greater in scleroderma fibroblasts than in control fibroblasts. As expected, treatment of control fibroblasts with TGF␤ rapidly induced an ⬃10fold increase in nuclear SMAD accumulation. We next compared the nuclear accumulation of phosphoSMAD2/3 in control and scleroderma fibroblasts. In the absence of TGF␤, only relatively low levels of phosphoSMAD2/3 could be detected in control fibroblasts, largely localized in the cytoplasm. In contrast, ⬃30% of examined scleroderma fibroblasts showed strong expression and dominant nuclear localization of phosphoSMAD2/3 (Figures 5A and B). SMAD7 acts as a component in negative feedback regulation of the fibroblast SMAD signaling pathway. By stably binding to the activated TGF␤ receptor, SMAD7 competes with SMAD2/3 for receptor interaction (23). In previous studies, SMAD7 was shown to be located predominantly in the nucleus and to translocate into the cytoplasm upon stimulation by TGF␤. In those experiments, transformed cell lines and recombinant SMAD7 protein were used. We examined the subcellular localization of native SMAD7 in normal fibroblasts by confocal microscopy. The results showed that in contrast to transformed cells such as COS7, in normal fibroblasts SMAD7 was located predominantly in the cytoplasm, both in the absence and in the presence of exogenous TGF␤ (results not shown). Therefore, SMAD7 subcellular localization appeared to be cell type dependent. The level of SMAD7 expression visualized 1972 by immunocytochemistry and its ligand-independent cytoplasmic localization were similar in control and scleroderma fibroblasts. Effect of TGF␤ blockade on SMAD subcellular distribution. Because TGF␤ is capable of inducing its own production in fibroblasts, autocrine signaling may contribute to sustaining of cellular responses to TGF␤. Autocrine stimulation by endogenous TGF␤ has been implicated in the constitutively activated phenotype of scleroderma fibroblasts (12,13). Because autocrine TGF␤ stimulation could account for SMAD activation and nuclear accumulation observed in scleroderma fibroblasts, we examined the effect of blocking TGF␤mediated signaling by two complementary approaches. For this purpose, first a neutralizing antibody recognizing all 3 isotypes of TGF␤ was used, either alone or in combination with the naturally occurring TGF␤ antagonist LAP. Confluent fibroblasts were washed extensively to remove secreted TGF␤, fresh media containing indicated concentrations of anti-TGF␤ antibody (1–20 g/ml) or nonimmune IgG were added, and the subcellular distribution of SMAD3/4 was then determined by confocal microscopy. The results showed that exposure of scleroderma fibroblasts to the antibody for up to 48 hours, alone or in combination with the TGF␤ antagonist LAP (5 g/ml), failed to down-regulate the level of nuclear SMAD4 or SMAD3 (Figure 6A and data not shown). As an important control, ligand-induced SMAD nuclear import was examined. As expected, addition of anti-TGF␤ antibody resulted in a dose-dependent inhibition of SMAD4 nuclear translocation induced by TGF␤ in control fibroblasts. Anti-TGF␤ antibody (10 g/ml) was capable of reducing by ⬎70% the SMAD nuclear translocation induced by TGF␤ at concentrations of up to 100 pM (Figure 6A and data not shown). Therefore, failure to reverse enhanced SMAD nuclear accumulation in scleroderma fibroblasts by partial blockade of the autocrine TGF␤ activation loop suggests that a mechanism independent of ligand-mediated TGF␤ receptor stimulation may have been responsible. The effect of anti-TGF␤ antibodies on subcellular distribution of phospho-SMAD2/3 was characterized. As shown in Figures 5A and B, ⬃30% of scleroderma fibroblasts and ⬍5% of control fibroblasts showed a predominantly nuclear localization of phosphoSMAD2/3, and anti-TGF␤ antibody failed to significantly reduce the proportion of scleroderma fibroblasts showing nuclear phospho-SMAD2/3. Immunoblots of nuclear extracts demonstrated elevated levels of phospho-SMAD2/3 in scleroderma fibroblasts and fur- MORI ET AL Figure 6. The effect of blockade of transforming growth factor ␤ (TGF␤) signaling on SMAD nuclear accumulation. Confluent cultures of fibroblasts were incubated with neutralizing antibody (Ab) to TGF␤ (␣-TGF-␤; 1–20 g/ml) or with nonimmune IgG, followed by addition of TGF␤ (10 pM) for 90 minutes. A, Cells were stained with SMAD4specific antibodies and examined by confocal fluorescence microscopy. The proportion of control (open bars) or scleroderma (solid bars) fibroblasts displaying a predominantly nuclear pattern of immunofluorescence was quantitated by scoring 100 individual cells in each culture in a blinded manner. Results shown represent the mean and SEM of duplicate determinations from 3 independent experiments. B, Serum-starved normal (N1) or scleroderma (S5, S12) fibroblasts were incubated with anti-TGF␤ antibody (10 g/ml). Cells were harvested and whole cell lysates (200 g) were immunoprecipitated with antibody to TGF␤ receptor type I, followed by immunoblotting with antiphosphoserine antibody (p-Serine). Membranes were stripped and reprobed with antibody to IgG to confirm equal protein loading in each lane. ther confirmed the failure of anti-TGF␤ antibody to “normalize” phospho-SMAD2/3 distribution (results not shown). SMAD SIGNALING IN SCLERODERMA FIBROBLASTS In order to confirm that neutralizing anti-TGF␤ antibody was capable of blocking autocrine TGF␤ signaling, we examined the phosphorylation state of endogenous TGF␤RI (which is primarily phosphorylated on serine residues in its GS domain upon activation by the ligand) on untreated control and scleroderma fibroblasts. For this purpose, TGF␤RI was immunoprecipitated from whole cell lysates, followed by electrophoresis and immunoblotting with antiphosphoserine antibodies. The results showed that while TGF␤RI was constitutively phosphorylated in both control and scleroderma fibroblasts due to autocrine signaling, scleroderma fibroblasts exhibited ⬎2-fold higher TGF␤RI phosphorylation than did control fibroblasts (Figure 6B). The levels of TGF␤RI were similar in scleroderma and control fibroblasts (data not shown). Importantly, treatment with anti-TGF␤ antibody caused a substantial reduction in TGF␤RI phosphorylation in scleroderma fibroblasts, but not in control fibroblasts. These results indicate that under the conditions employed in these experiments, anti-TGF␤ antibody effectively disrupted the autocrine TGF␤ activation loop in scleroderma fibroblasts. To further determine the potential contribution of autocrine TGF␤ signaling in enhanced SMAD nuclear localization, we used a dominant-negative TGF␤RII vector to disrupt TGF␤ responses. The dominant-negative TGF␤ receptor is unable to phosphorylate the type I receptor in response to any of the 3 TGF␤ isoforms, thus interrupting intracellular TGF␤ signaling. Normal and scleroderma fibroblasts in parallel were transduced with the dominant-negative TGF␤RII or empty vector, followed 48 hours later by incubation with TGF␤ for 90 minutes. As shown in a representative experiment in Figure 7A, in normal fibroblasts transduced with dominantnegative TGF␤RII, TGF␤-stimulated SMAD nuclear accumulation was markedly reduced (compare panels b and d). As expected, scleroderma fibroblasts showed a markedly higher level of basal nuclear SMAD accumulation in the absence of added TGF␤ (panel e), and this was not reduced by the dominant-negative TGF␤RII (panel f). A high level of dominant-negative TGF␤RII expression in the same cells was confirmed by immunocytochemistry (lower panels). Quantitation of the results from several independent experiments confirmed the ability of adenovirus-mediated dominant-negative TGF␤RII overexpression to reduce by ⬎80% the TGF␤-induced nuclear migration of SMAD2/3 in normal fibroblasts, as well as its failure to “normalize” 1973 Figure 7. The effect of adenovirus-mediated expression of dominantnegative transforming growth factor ␤ receptor type II (TGF␤RII) on SMAD nuclear accumulation in scleroderma fibroblasts. A, Confluent cultures of normal (a–d) or scleroderma (e and f) fibroblasts were transduced with empty vector (a, b, and e) or adenovirus expressing dominant-negative TGF␤RII (c, d, and f), followed 48 hours later by addition of 10 pM TGF␤ (b and d). Following 90 minutes of incubation, fibroblasts were stained with antibodies to SMAD2/3 (upper panels) or TGF␤RII (lower panels) and examined by confocal fluorescence microscopy. B, The proportion of control (open bars) or scleroderma (solid bars) fibroblasts displaying a predominantly nuclear pattern of SMAD3 immunofluorescence was quantitated by scoring 100 individual cells in each culture in a blinded manner. Results shown represent the mean and SEM of duplicate determinations from several independent experiments with 3 separate scleroderma cell lines. C, Control (a) or scleroderma (b and c) fibroblasts were transduced with empty vector (a and b) or dominant-negative TGF␤RII (c) and were stained 48 hours later with antibodies to SMAD2/3 or TGF␤RII. Immunofluorescence confocal microscopic images from a representative experiment are shown. In c, note that immunostaining confirms substantial expression of dominant-negative TGF␤RII (green) in scleroderma fibroblasts showing a high level of nuclear SMAD2/3 accumulation (pink). dnT␤RII ⫽ dominant-negative TGF␤RII. (Original magnification ⫻ 400.) 1974 MORI ET AL activity of pGL3-luc was similar in lesional and control fibroblasts. DISCUSSION Figure 8. SMAD-regulated transcriptional activity in fibroblasts. Subconfluent scleroderma (S) and control (N) fibroblasts in parallel were transiently transfected with pSBE4-luc and appropriate control plasmids, and luciferase (Luc) activities were determined after 48 hours of incubation. Results, normalized by renilla luciferase activity to correct for small variations in transfection efficiency between samples, represent the means of triplicate samples from 3 independent experiments. Bars show the group mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05. SMAD subcellular distribution in scleroderma fibroblasts (Figure 7B). Essentially identical results were noted with 3 separate scleroderma cell lines. Transduction with GFP vector had no effect on either TGF␤induced SMAD nuclear migration in control fibroblasts or constitutive nuclear accumulation in scleroderma fibroblasts. Note the elevated nuclear accumulation of SMAD2/3 in scleroderma fibroblasts showing high levels of cellular dominant-negative TGF␤RII expression 48 hours posttransduction in the representative images shown in Figure 7C. SMAD-mediated transcriptional activity in scleroderma fibroblasts. In order to investigate the potential functional significance of ligand-independent activation and nuclear accumulation of endogenous SMAD3 and SMAD4 in scleroderma fibroblasts, we examined SMAD-regulated luciferase reporter expression. The SMAD-responsive minimal pSBE4-luc construct was transiently transfected in control and scleroderma fibroblasts in parallel, and luciferase activity was determined after 48 hours. As shown in Figure 8, SMAD-inducible promoter activity was elevated in untreated lesional fibroblasts compared with controls. The The recent discovery and characterization of the SMAD protein family has provided novel opportunities for delineating the involvement of TGF␤ in fibrosis (39). In normal fibroblasts, SMADs function as essential cytoplasmic effectors for TGF␤ stimulation of collagen synthesis, as well as that of CTGF (40) and of TGF␤ itself (41). Genetic targeting of the SMAD3 locus protected null mice from the development of radiationinduced fibrosis, indicating the fundamental physiologic role of SMAD3 in processes of tissue repair and fibrosis (42). Because the activated scleroderma fibroblast phenotype is thought to reflect enhanced stimulation by, or intrinsic responsiveness to, TGF␤, in the present study we examined the expression and regulation of the SMAD signaling pathway in these cells. The results indicated that in a majority of scleroderma fibroblasts, SMAD3 protein and mRNA expression were elevated in comparison with control fibroblasts. Elevated levels of SMAD3 have also been observed in hepatic myofibroblasts (43) and in a murine model of renal fibrosis (44). Elevated SMAD3 expression could result in enhanced formation and nuclear import of transcriptionally active SMAD heterocomplexes, with increased transcription of TGF␤-regulated target genes. However, in light of the relatively modest and variable increase in SMAD3 mRNA levels in scleroderma versus control fibroblasts (⬃60%), we consider it unlikely that elevated SMAD3 expression plays a major role in altered TGF␤ signaling in scleroderma fibroblasts. Steady-state levels of SMAD3 are determined in part by the rate of transcription of the SMAD3 gene and the intracellular degradation of SMAD3 protein. The factors determining the rate of SMAD3 transcription are currently unknown. Because we found in the present experiments that the increase in SMAD3 protein levels in scleroderma fibroblasts compared with control fibroblasts was of relatively greater magnitude than the increase in mRNA levels, it is possible that SMAD3 protein degradation was impaired; this is currently under investigation. Importantly, the expression of SMAD3 mRNA in both normal and scleroderma fibroblasts was selectively repressed by TGF␤. Inhibition of SMAD3 mRNA expression was a delayed response, with maximal effect after 48 hours of exposure. These results clearly indicate that TGF␤ can regulate the steady-state levels SMAD SIGNALING IN SCLERODERMA FIBROBLASTS of its own second messengers by pretranslational modulation, in addition to the well-documented proteasomal degradation pathways. Repression of SMAD3 mRNA levels in response to TGF␤ has been noted in epithelial and mesangial cells as well as in lung fibroblasts (45–47). In light of the dependence of TGF␤ signaling intensity on the steadystate levels of its intracellular second messengers, inhibition of SMAD3 expression may thus be a negative feedback mechanism with autoregulatory function to prevent continuous SMAD signaling in response to TGF␤ stimulation. The present results indicate that elevated SMAD3 steady-state levels seen in some scleroderma fibroblasts cannot be attributed to failure of TGF␤ to repress SMAD3 gene expression in these cells. In normal cells, TGF␤ directly stimulates transcription of inhibitory SMAD7. Thus, SMAD7 induction serves as a critical intracellular “brake” on TGF␤induced responses. The physiologic importance of this SMAD7-mediated autoinhibitory feedback loop is highlighted by recent observations with cultured cells in in vivo animal models. For example, tumor necrosis factor ␣–induced abrogation of TGF␤-dependent profibrotic responses was linked to the induction of endogenous SMAD7 (48). On the other hand, defective SMAD7 induction is implicated in the exaggerated TGF␤ responsiveness characteristic of hepatic cells and myofibroblasts from chronically injured livers (49,50). While the highest level of SMAD7 expression is normally found in the kidneys (23), in spontaneous renal fibrosis in TGF␤1-transgenic mice, and in renal fibrosis induced by anti–Thy-1 antibody, excessive matrix accumulation occurs in the setting of reduced SMAD7, resulting in exaggerated or sustained local TGF␤ response (51,52). Furthermore, a recent study demonstrated that SMAD7 basal expression levels and TGF␤-dependent inducibility were both markedly impaired in scleroderma fibroblasts in vivo and in vitro (53). These observations are intriguing, because SMAD7 deficiency could provide a mechanistic explanation for elevated TGF␤ receptor expression and relative resistance to apoptosis that have been reported in scleroderma fibroblasts. In the present studies, we failed to detect consistent intrinsic alterations in basal levels of SMAD7 expression or SMAD7 inducibility. Protein and mRNA expression of SMAD7, its regulation by TGF␤, and its subcellular localization in scleroderma fibroblasts were all comparable with those in normal fibroblasts. Therefore, our results do not support the hypothesis that in scleroderma, fibroblasts may be “sensitized” to the profibrotic effects of TGF␤ because of inability to 1975 “apply the brakes” on TGF␤-induced cellular responses due to impaired SMAD7 regulation. It remains possible, however, that despite its apparently normal inducibility by TGF␤ in vitro, SMAD7 in scleroderma fibroblasts somehow fails to inhibit receptor-mediated activation of SMAD2/3. In the absence of ligand stimulation, receptoractivated SMADs and SMAD4 reside mostly in the cytoplasm, and their accumulation within the nucleus is a key event in the intracellular propagation of TGF␤ signaling. Due to the presence of an N-terminal nuclear localization signal sequence, SMAD3 displays intrinsic nuclear import activity (54). In contrast, SMAD4 requires association with activated SMAD3 in order to accumulate in the nucleus (55). In normal fibroblasts, only a low level of nuclear SMAD accumulation could be detected in the absence of exogenous TGF␤. In contrast, the present studies revealed a substantial degree of endogenous SMAD phosphorylation and nuclear accumulation in scleroderma fibroblasts. These findings were highly consistent in each of the scleroderma fibroblast lines studied. Similar ligand-independent SMAD activation has been described in hepatic stellate cells derived from fibrotic livers (56) and in dermal fibroblasts derived from keloid lesions (57). Constitutive SMAD activation is therefore likely to contribute to the profibrotic phenotype of scleroderma fibroblasts. One of the significant profibrotic effects of TGF␤ is its ability to inhibit production of the matrix-degrading enzyme collagenase 1. Repression of collagenase in normal fibroblasts by TGF␤ is mediated through SMAD3 (58). Therefore, the present results demonstrating signal-independent activation of SMAD3 in scleroderma fibroblasts may provide a mechanistic explanation to account for the reduced collagenase expression previously described in scleroderma (59). Recent reports indicate that optimal induction of COL1A2 transcription by TGF␤ involves an interaction between activated SMAD3 and the ubiquitous DNA binding transcription factor Sp1 (32,33). In this regard, it is of interest that Sp1 has been shown to be constitutively phosphorylated and activated in scleroderma fibroblasts (60). These findings suggest that for maximal transactivation of COL1A2 in scleroderma fibroblasts, SMAD activation and Sp1 phosphorylation may both be required. The mechanisms responsible for constitutive SMAD activation in scleroderma fibroblasts are currently unknown. Autocrine stimulation by endogenous TGF␤ may provide a possible explanation. Indeed, a key role for autocrine stimulation by TGF␤ in the activated 1976 phenotype of scleroderma fibroblasts has been suggested, and elevated collagen synthesis was reduced by disrupting endogenous TGF␤ signaling (12,13). It has been recently demonstrated that SMAD3/4 undergo continuous nucleocytoplasmic shuttling, and their relative levels within the nucleus are directly dictated by the level of TGF␤ receptor activity (61). However, in the present studies, we found that neutralizing anti-TGF␤ antibody and LAP failed to normalize SMAD subcellular distribution in scleroderma fibroblasts despite a decrease in TGF␤RI phosphorylation. Furthermore, transduction of dominant-negative TGF␤ receptor that prevented TGF␤-induced SMAD activation in normal fibroblasts failed to reduce SMAD nuclear accumulation in scleroderma fibroblasts, suggesting that constitutive SMAD nuclear accumulation was not due to autocrine signaling by endogenous TGF␤. It is conceivable that in the present experiments, neither neutralizing antibody nor transduction of dominant-negative TGF␤RII was able to completely abrogate fibroblast activation mediated through endogenous TGF␤. In that case, the “ligand-independent” SMAD activation we observed in scleroderma fibroblasts may in fact reflect autocrine TGF␤ stimulation. This possibility cannot be ruled out, given that scleroderma fibroblasts in some studies (13,14), although not in others (62), were shown to display elevated expression of TGF␤ receptors and may thus have been “sensitized” to TGF␤ (13,14); SMAD nuclear migration in fibroblasts can be induced by very low concentrations of TGF␤. Alternatively, it is possible that non-TGF␤ ligands trigger cellular SMAD activation and nuclear migration. For example, a recent report indicated that insulin-like growth factor binding protein 3 (IGFBP-3) could by itself induce SMAD2/3 phosphorylation and potentiate TGF␤-stimulated cellular responses (63). The ability of IGFBPs to induce TGF␤-independent SMAD phosphorylation/activation may be particularly relevant in the pathogenesis of scleroderma, since IGFBP gene expression is increased in scleroderma fibroblasts (64). In summary, the present results provide evidence for consistent alterations in the activation states of intracellular TGF␤/SMAD signaling components in the absence of exogenous TGF␤ in scleroderma fibroblasts. These alterations may contribute to sensitizing scleroderma fibroblasts to TGF␤ or other stimuli that utilize the SMAD pathway, resulting in enhanced fibrotic responses elicited by extracellular signals in pathologic fibrosis. MORI ET AL ACKNOWLEDGMENTS We are grateful to Dr. R. Derynck (University of California, San Francisco), Dr. P. ten Dijke (Ludwig Institute for Cancer Research, Uppsala, Sweden), Dr. J. Massague (Memorial Sloan-Kettering Cancer Center, New York, NY), and Dr. B. Vogelstein (Johns Hopkins University, Baltimore, MD) for providing us with reagents, and we thank members of our laboratory staff for helpful discussions. REFERENCES 1. Varga J, Mori Y, Takagawa S. Molecular and cellular basis of fibrosis. Semin Clin Immunol 2000;2:15–29. 2. Widom RL. Regulation of matrix biosynthesis and degradation in systemic sclerosis. Curr Opin Rheumatol 2000;12:534–9. 3. Blobe GC, Schiemann WP, Lodish HF. Role of transforming growth factor beta in human disease. N Engl J Med 2000;342: 1350–8. 4. Kulozik M, Hogg A, Lankat-Buttgereit B, Krieg T. Co-localization of transforming growth factor beta 2 with alpha 1(I) procollagen mRNA in tissue sections of patients with systemic sclerosis. J Clin Invest 1990;86:917–22. 5. Gabrielli A, Di Loreto C, Taborro R, Candela M, Sambo P, Nitti C, et al. Immunohistochemical localization of intracellular and extracellular associated TGF beta in the skin of patients with systemic sclerosis (scleroderma) and primary Raynaud’s phenomenon. Clin Immunol Immunopathol 1993;68:340–9. 6. Gruschwitz M, Muller PU, Sepp N, Hofer E, Fontana A, Wick G. Transcription and expression of transforming growth factor type beta in the skin of progressive systemic sclerosis: a mediator of fibrosis? J Invest Dermatol 1990;94:197–203. 7. Sfikakis PP, McCune BK, Tsokos M, Aroni K, Vayiopoulos G, Tsokos GC. Immunohistological demonstration of transforming growth factor-beta isoforms in the skin of patients with systemic sclerosis. Clin Immunol Immunopathol 1993;69:199–204. 8. Rudnicka L, Varga J, Christiano AM, Iozzo RV, Jimenez SA, Uitto J. Elevated expression of type VII collagen in the skin of patients with systemic sclerosis: regulation by transforming growth factor-beta. J Clin Invest 1996;93:1709–15. 9. Mason GI, Hamburger J, Matthews JB. Mast cells, extracellular matrix components, TGF␤ isoforms and TGF␤ receptor expression in labial salivary glands in systemic sclerosis. Ann Rheum Dis 2000;59:183–9. 10. Higley H, Persichitte K, Chu S, Waegell W, Vancheeswaran R, Black C. Immunocytochemical localization and serologic detection of transforming growth factor ␤1: association with type I procollagen and inflammatory cell markers in diffuse and limited systemic sclerosis, morphea, and Raynaud’s phenomenon. Arthritis Rheum 1994;37:278–88. 11. Querfeld C, Eckes B, Huerkamp C, Krieg T, Sollberg S. Expression of TGF-beta 1, -beta 2 and -beta 3 in localized and systemic scleroderma. J Dermatol Sci 1999;21:13–22. 12. Kawakami T, Ihn H, Xu W, Smith E, LeRoy C, Trojanowska M. Increased expression of TGF-beta receptors by scleroderma fibroblasts: evidence for contribution of autocrine TGF-beta signaling to scleroderma phenotype. J Invest Dermatol 1998;110:47–51. 13. Ihn H, Yamane K, Kubo M, Tamaki K. Blockade of endogenous transforming growth factor ␤ signaling prevents up-regulated collagen synthesis in scleroderma fibroblasts: association with increased expression of transforming growth factor ␤ receptors. Arthritis Rheum 2001;44:474–80. 14. Kubo M, Ihn H, Yamane K, Tamaki K. Up-regulated expression of transforming growth factor ␤ receptors in dermal fibroblasts in skin sections from patients with localized scleroderma. Arthritis Rheum 2001;44:731–4. SMAD SIGNALING IN SCLERODERMA FIBROBLASTS 15. McCormick LL, Zhang Y, Tootell E, Gilliam AC. Anti-TGF-beta treatment prevents skin and lung fibrosis in murine sclerodermatous graft-versus-host disease: a model for human scleroderma. J Immunol 1999;163:5693–9. 16. Yamamoto T, Takagawa S, Katayama I, Nishioka K. Anti-sclerotic effect of transforming growth factor-beta antibody in a mouse model of bleomycin-induced scleroderma. Clin Immunol 1999;92: 6–13. 17. McGaha T, Saito S, Phelps RG, Gordon R, Noben-Trauth N, Paul WE, et al. Lack of skin fibrosis in tight skin (TSK) mice with targeted mutation in the interleukin-4R alpha and transforming growth factor-beta genes. J Invest Dermatol 2001;116:136–43. 18. Verrecchia F, Chu ML, Mauviel A. Identification of novel TGFbeta/Smad gene targets in dermal fibroblasts using a combined cDNA microarray/promoter transactivation approach. J Biol Chem 2001;276:17058–62. 19. Van Obberghen-Schilling E, Roche NS, Flanders KC, Sporn MB, Roberts AB. Transforming growth factor beta 1 positively regulates its own expression in normal and transformed cells. J Biol Chem 1988;263:7741–6. 20. Grotendorst GR, Okochi H, Hayashi N. A novel transforming growth factor beta response element controls the expression of the connective tissue growth factor gene. Cell Growth Differ 1996;7: 469–80. 21. Massague J, Wotton D. Transcriptional control by the TGF-beta/ Smad signaling system. EMBO J 2000;19:1745–54. 22. Topper JN, Cai J, Qiu Y, Anderson KR, Xu YY, Deeds JD, et al. Vascular MADs: two novel MAD-related genes selectively inducible by flow in human vascular endothelium. Proc Natl Acad Sci U S A 1997;94:9314–9. 23. Nakao A, Afrakhte M, Moren A, Nakayama T, Christian JL, Heuchel R, et al. Identification of Smad7, a TGFbeta-inducible antagonist of TGF-beta signalling. Nature 1997;389:631–5. 24. Mori Y, Chen SJ, Varga J. Modulation of endogenous Smad expression in normal skin fibroblasts by transforming growth factor-beta. Exp Cell Res 2000;258:374–83. 25. Moustakas A, Souchelnytskyi S, Heldin CH. Smad regulation in TGF-beta signal transduction. J Cell Sci 2001;114:4359–69. 26. Tsukazaki T, Chiang TA, Davison AF, Attisano L, Wrana JL. SARA, a FYVE domain protein that recruits Smad2 to the TGFbeta receptor. Cell 1998;95:779–91. 27. Dong C, Li Z, Alvarez R Jr, Feng XH, Goldschmidt-Clermont PJ. Microtubule binding to Smads may regulate TGF beta activity. Mol Cell 2000;5:27–34. 28. Lo RS, Massague J. Ubiquitin-dependent degradation of TGFbeta-activated Smad2. Nat Cell Biol 1999;1:472–8. 29. Chen SJ, Yuan W, Mori Y, Levenson A, Trojanowska M, Varga J. Stimulation of type I collagen transcription in human skin fibroblasts by TGF-beta: involvement of Smad 3. J Invest Dermatol 1999;112:49–57. 30. Chen SJ, Yuan W, Lo S, Trojanowska M, Varga J. Interaction of Smad3 with a proximal Smad-binding element of the human alpha2(I) procollagen gene promoter required for transcriptional activation by TGF-beta. J Cell Physiol 2000;183:381–92. 31. Ghosh AK, Yuan W, Mori Y, Varga J. Smad-dependent stimulation of type I collagen gene expression in human skin fibroblasts by TGF-beta involves functional cooperation with p300/CBP transcriptional coactivators. Oncogene 2000;19:3546–55. 32. Poncelet AC, Schnaper HW. Sp1 and Smad proteins cooperate to mediate TGF-beta1-induced alpha2(I) collagen expression in human glomerular mesangial cells. J Biol Chem 2001;276:6983–92. 33. Zhang W, Ou J, Inagaki Y, Greenwel P, Ramirez F. Synergistic cooperation between Sp1 and Smad3/Smad4 mediates transforming growth factor beta1 stimulation of alpha 2(I) collagen (COL1A2) transcription. J Biol Chem 2000;275:39237–45. 34. Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Commit- 1977 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. tee. Preliminary criteria for the classification of systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581–90. Brand T, MacLellan WR, Schneider MD. A dominant-negative receptor for type beta transforming growth factors created by deletion of the kinase domain. J Biol Chem 1993;268:11500–3. Zawel L, Dai JL, Buckhaults P, Zhou S, Kinzler KW, Vogelstein B, et al. Human Smad3 and Smad4 are sequence-specific transcription activators. Mol Cell 1998;1:611–7. Wakefield LM, Roberts AB. TGF-beta signaling: positive and negative effects on tumorigenesis. Curr Opin Genet Dev 2002;12: 22–9. Ghosh AK, Yuan W, Mori Y, Chen S, Varga J. Antagonistic regulation of type I collagen gene expression by interferon-gamma and transforming growth factor-beta: integration at the level of p300/CBP transcriptional coactivators. J Biol Chem 2001;276: 11041–8. Varga J. Scleroderma and Smads: dysfunctional Smad family dynamics culminating in fibrosis. Arthritis Rheum 2002;46: 1703–13. Holmes A, Abraham DJ, Sa S, Shiwen X, Black CM, Leask A. CTGF and Smads, maintenance of scleroderma phenotype is independent of Smad signaling. J Biol Chem 2001;276:10594–601. Piek E, Ju WJ, Heyer J, Escalante-Alcalde D, Stewart CL, Weinstein M, et al. Functional characterization of transforming growth factor beta signaling in Smad2- and Smad3-deficient fibroblasts. J Biol Chem 2001;276:19945–53. Roberts AB, Piek E, Bottinger EP, Ashcroft G, Mitchell JB, Flanders KC. Is Smad3 a major player in signal transduction pathways leading to fibrogenesis? Chest 2001;120:43S–47S. Dooley S, Streckert M, Delvoux B, Gressner AM. Expression of Smads during in vitro transdifferentiation of hepatic stellate cells to myofibroblasts. Biochem Biophys Res Commun 2001;283: 554–62. Hong S, Isono M, Chen S, Iglesias-De La Cruz MC, Han DC, Ziyadeh FN. Increased glomerular and tubular expression of transforming growth factor-beta1, its type II receptor, and activation of the Smad signaling pathway in the db/db mouse. Am J Pathol 2001;158:1653–63. Yanagisawa K, Osada H, Masuda A, Kondo M, Saito T, Yatabe Y, et al. Induction of apoptosis by Smad3 and down-regulation of Smad3 expression in response to TGF-beta in human normal lung epithelial cells. Oncogene 1998;17:1743–7. Poncelet AC, de Caestecker MP, Schnaper HW. The transforming growth factor-beta/SMAD signaling pathway is present and functional in human mesangial cells. Kidney Int 1999;56:1354–65. Zhao Y, Geverd DA. Regulation of Smad3 expression in bleomycin-induced pulmonary fibrosis: a negative feedback loop of TGF-beta signaling. Biochem Biophys Res Commun 2002;294: 319–23. Bitzer M, von Gersdorff G, Liang D, Dominguez-Rosales A, Beg AA, Rojkind M, et al. A mechanism of suppression of TGF-beta/ SMAD signaling by NF-kappa B/RelA. Genes Dev 2000;14: 187–97. Tahashi Y, Matsuzaki K, Date M, Yoshida K, Furukawa F, Sugano Y, et al. Differential regulation of TGF-beta signal in hepatic stellate cells between acute and chronic rat liver injury. Hepatology 2002;35:49–61. Dooley S, Delvoux B, Lahme B, Mangasser-Stephan K, Gressner AM. Modulation of transforming growth factor beta response and signaling during transdifferentiation of rat hepatic stellate cells to myofibroblasts. Hepatology 2000;31:1094–106. Uchida K, Nitta K, Kobayashi H, Kawachi H, Shimizu F, Yumura W, et al. Localization of Smad6 and Smad7 in the rat kidney and their regulated expression in the anti-Thy-1 nephritis. Mol Cell Biol Res Commun 2000;4:98–105. Schiffer M, Bitzer M, Roberts IS, Kopp JB, ten Dijke P, Mundel 1978 53. 54. 55. 56. 57. 58. P, et al. Apoptosis in podocytes induced by TGF-beta and Smad7. J Clin Invest 2001;108:807–16. Dong C, Zhu S, Wang T, Yoon W, Li Z, Alvarez RJ, et al. Deficient Smad7 expression: a putative molecular defect in scleroderma. Proc Natl Acad Sci U S A 2002;99:3908–13. Xiao Z, Liu X, Henis YI, Lodish HF. A distinct nuclear localization signal in the N terminus of Smad 3 determines its ligandinduced nuclear translocation. Proc Natl Acad Sci U S A 2000;97: 7853–5. Liu F, Pouponnot C, Massague J. Dual role of the Smad4/DPC4 tumor suppressor in TGF-beta-inducible transcriptional complexes. Genes Dev 1997;11:3157–67. Inagaki Y, Mamura M, Kanamaru Y, Greenwel P, Nemoto T, Takehara K, et al. Constitutive phosphorylation and nuclear localization of Smad3 are correlated with increased collagen gene transcription in activated hepatic stellate cells. J Cell Physiol 2001;187:117–23. Chin GS, Liu W, Peled Z, Lee TY, Steinbrech DS, Hsu M, et al. Differential expression of transforming growth factor-beta receptors I and II and activation of Smad 3 in keloid fibroblasts. Plast Reconstr Surg 2001;108:423–9. Yuan W, Varga J. Transforming growth factor-beta repression of MORI ET AL 59. 60. 61. 62. 63. 64. matrix metalloproteinase-1 in dermal fibroblasts involves Smad3. J Biol Chem 2001;276:38502–10. Takeda K, Hatamochi A, Ueki H, Nakata M, Oishi Y. Decreased collagenase expression in cultured systemic sclerosis fibroblasts. J Invest Dermatol 1994;103:359–63. Ihn H, Tamaki K. Increased phosphorylation of transcription factor Sp1 in scleroderma fibroblasts: association with increased expression of the type I collagen gene. Arthritis Rheum 2000;43: 2240–7. Inman GJ, Nicolas FJ, Hill CS. Nucleocytoplasmic shuttling of Smads 2, 3, and 4 permits sensing of TGF-beta receptor activity. Mol Cell 2002;10:283–94. Leask A, Abraham DJ, Finlay DR, Holmes A, Pennington D, Shi-Wen X, et al. Dysregulation of transforming growth factor ␤ signaling in scleroderma: overexpression of endoglin in cutaneous scleroderma fibroblasts. Arthritis Rheum 2002;46:1857–65. Fanayan S, Firth SM, Baxter RC. Signaling through the Smad pathway by insulin-like growth factor binding protein-3 in breast cancer cells: relationship to transforming growth factor-1 signaling. J Biol Chem 2002;277:7255–61. Feghali CA, Wright TM. Identification of multiple, differentially expressed messenger RNAs in dermal fibroblasts from patients with systemic sclerosis. Arthritis Rheum 1999;42:1451–7.