Requirement of transforming growth factor activated kinase 1 for transforming growth factor induced ╨Ю┬▒-smooth muscle actin expression and extracellular matrix contraction in fibroblasts.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 60, No. 1, January 2009, pp 234–241 DOI 10.1002/art.24223 © 2009, American College of Rheumatology Requirement of Transforming Growth Factor ␤–Activated Kinase 1 for Transforming Growth Factor ␤–Induced ␣-Smooth Muscle Actin Expression and Extracellular Matrix Contraction in Fibroblasts Xu Shi-wen,1 Sunil K. Parapuram,2 Daphne Pala,2 Yunliang Chen,3 David E. Carter,4 Mark Eastwood,3 Christopher P. Denton,1 David J. Abraham,1 and Andrew Leask2 Objective. Fibrosis is believed to occur through normal tissue remodeling failing to terminate. Tissue repair intimately involves the ability of fibroblasts to contract extracellular matrix (ECM), and enhanced ECM contraction is a hallmark of fibrotic cells in various conditions, including scleroderma. Some fibrogenic transcriptional responses to transforming growth factor ␤ (TGF␤), including ␣-smooth muscle actin (␣-SMA) expression and ECM contraction, require focal adhesion kinase/Src (FAK/Src). The present study was undertaken to assess whether TGF␤-activated ki- nase 1 (TAK1) acts downstream of FAK/Src to mediate fibrogenic responses in fibroblasts. Methods. We used microarray, real-time polymerase chain reaction, Western blot, and collagen gel contraction assays to assess the ability of wild-type and TAK1-knockout fibroblasts to respond to TGF␤1. Results. The ability of TGF to induce TAK1 was blocked by the FAK/Src inhibitor PP2. JNK phosphorylation in response to TGF␤1 was impaired in the absence of TAK1. TGF␤ could not induce matrix contraction or expression of a group of fibrotic genes, including ␣-SMA, in the absence of TAK1. Conclusion. These results suggest that TAK1 operates downstream of FAK/Src in mediating fibrogenic responses and that targeting of TAK1 may be a viable antifibrotic strategy in the treatment of certain disorders, including scleroderma. Supported by the Canadian Institutes of Health Research, the Canadian Foundation for Innovation, the Arthritis Research Campaign, the Raynaud’s and Scleroderma Association, and the Scleroderma Society. Dr. Parapuram is recipient of a postdoctoral fellowship from the Canadian Institutes of Health Research, Canadian Scleroderma Research Group. Ms Pala was recipient of a Network of Oral Research Training and Health scholarship for undergraduate dental research and is recipient of an Ontario Graduate Scholarship in Science and Technology at the University of Western Ontario. Dr. Leask is a New Investigator of the Arthritis Society (Scleroderma Society of Ontario) and recipient of an Early Researcher Award from the Ontario Ministry of Research and Innovation. 1 Xu Shi-wen, PhD, Christopher P. Denton, PhD, FRCP, David J. Abraham, PhD: University College London (Royal Free Campus), London, UK; 2Sunil K. Parapuram, PhD, Daphne Pala, MSc, Andrew Leask, PhD: University of Western Ontario, London, Ontario, Canada; 3Yunliang Chen, PhD, Mark Eastwood, PhD: University of Westminster, London, UK; 4David E. Carter, MSc: London Regional Genomics Centre, London, Ontario, Canada. Dr. Denton has received consulting fees, speaking fees, and/or honoraria from Actelion and Encysive (less than $10,000 each). Dr. Abraham has received consulting fees, speaking fees, and/or honoraria from Actelion and Encysive (less than $10,000 each). Address correspondence and reprint requests to Andrew Leask, PhD, CIHR Group in Skeletal Development and Remodeling, Department of Physiology and Pharmacology, Schulich School of Medicine and Dentistry, Dental Sciences Building, University of Western Ontario, London, Ontario N6A 5C1, Canada. E-mail: Andrew.Leask@schulich.uwo.ca. Submitted for publication April 30, 2008; accepted in revised form October 3, 2008. Normal tissue repair involves migration of fibroblasts into the wound. The fibroblasts then synthesize and remodel extracellular matrix (ECM), resulting in wound closure. These events involve the ability of fibroblasts to attach to ECM through specialized cell surface structures called focal adhesions (1). Persistently activated adhesion and adhesive signaling is a hallmark of fibrotic cells, including those cultured from lesional areas in patients with the fibrotic autoimmune disease systemic sclerosis (SSc; scleroderma) (2,3). Adhesion to ECM involves integrins, whose signals are transmitted by focal adhesion kinase (FAK), which is present at focal adhesions and is phosphorylated after integrin-mediated cell attachment (4). FAK has been classically considered to mediate fibroblast attachment to ECM; however, a priori it is also possible that it may be involved in transducing signals from growth factors (5). Indeed, we have recently found that the ability of transforming 234 NECESSITY OF TAK1 FOR FIBROBLAST RESPONSES TO TGF␤ 235 TAK1 to investigate the contribution of TAK1 to mediation of transcriptional responses to TGF␤ in fibroblasts. Our results provide new insights into the contribution of TAK1 to tissue repair and remodeling responses in fibroblasts and suggest that targeting of TAK1 may be a viable antifibrotic treatment approach. MATERIALS AND METHODS Figure 1. Requirement for focal adhesion kinase in transforming growth factor ␤ (TGF␤) induction of TGF␤-activated kinase 1 (TAK1) phosphorylation in fibroblasts. A, Fibroblasts from lesional areas of skin from patients with systemic sclerosis (SSc) and age-, sex-, and site-matched controls were serum starved for 24 hours and subjected to Western blot analysis with anti-TAK1 and anti–phospho-TAK1 antibodies. Cells from 4 normal subjects and 4 SSc patients were used. B, Fibroblasts from tak⫹/⫹ mice were serum starved and left untreated or treated with TGF␤1 (4 ng/ml) for 30 minutes. Cells were pretreated with PP2 (10 M) for 45 minutes prior to addition of TGF␤ or were not pretreated. In parallel experiments, fibroblasts from fak⫹/⫹ and fak⫺/⫺ mice were similarly tested. Cell extracts were prepared and subjected to Western blot analysis with anti–phospho-TAK1 and anti-TAK1 antibodies as indicated. A total of 4 independent experiments were performed, and representative blots are shown. growth factor ␤ (TGF␤) to induce expression of messenger RNA (mRNA) for profibrotic genes involves FAK (6). FAK was required in order for TGF␤ to induce JNK, and the FAK/MEKK1/JNK pathway was required for TGF␤ to induce myofibroblast formation, as identified by ␣-smooth muscle actin (␣-SMA) expression and the ability of fibroblasts to contract ECM (6). TGF␤-activated kinase 1 (TAK1), a member of the MAPKKK family, is thought to be a key modulator of the inducible transcription factors NF-B and activator protein 1, and therefore plays a crucial role in regulating the genes that mediate inflammation. In fibroblasts, TAK1 deletion does not impair the ability of TGF␤ to activate the Smad pathway (7). However, Smad-independent pathways are known to mediate the action of TGF␤; thus, the effect of loss of TAK1 on transcriptional responses in fibroblasts is unclear. We used wild-type fibroblasts and fibroblasts deficient in Cell culture and Western blot analysis. Fibroblasts derived from E9.75 wild-type (tak⫹/⫹) and homozygous mutant (tak⫺/⫺) mouse embryos were immortalized by stably transfecting linearized SV40 large T antigen plasmid (7). For some experiments, fibroblasts obtained from E8.5 FAK wildtype and FAK-knockout mouse embryos (American Type Culture Collection, Rockville, MD) were used. In addition, dermal fibroblasts from affected areas in patients with SSc and from age-, sex- and site-matched controls were used. Cells were grown in Dulbecco’s modified Eagle’s medium containing 5% fetal calf serum (FCS), 2 mM L-glutamine, antibiotics (100 units/ml penicillin and 100 g/ml streptomycin), and 1 mM sodium pyruvate (Invitrogen, Burlington, Ontario, Canada). Cells were serum starved for 24 hours and then left untreated or treated with TGF␤1 (4 ng/ml; R & D Systems, Minneapolis, MN). Whole cell protein extracts were produced and subjected to Western blot analysis using anti-TAK, anti–phospho-TAK, anti-JNK, anti–phospho-JNK1 (all from Cell Signaling Technology, Vancouver, British Columbia, Canada), anti-CCN2 (Abcam, Cambridge, UK), anti–␣-SMA (Sigma, St. Louis, MO), or anti-GAPDH (Abcam). Intensity of signals relative to GAPDH controls was calculated using densitometry, and averages and standard deviations of data obtained from 3 independent experiments were calculated. RNA quality assessment, probe preparation, and GeneChip hybridization and analysis. Microarray analysis was performed as previously described (6,8,9). GeneChips were processed at the London Regional Genomics Centre (Robarts Research Institute, London, Ontario, Canada [http:// Figure 2. Impairment of transforming growth factor ␤ (TGF␤)– induced JNK phosphorylation in the absence of TGF␤-activated kinase 1. Fibroblasts from tak⫹/⫹ mice (wild-type [WT]) or tak⫺/⫺ mice (knockout [KO]) were serum starved and left untreated or treated with TGF␤1 (4 ng/ml) for 1 hour. Cell extracts were prepared and subjected to Western blot analysis with anti-JNK1 or anti– phospho-JNK1 antibodies. A total of 4 independent experiments were performed, and representative blots are shown. 236 SHI-WEN ET AL Figure 3. Necessity of TGF␤-activated kinase 1 for TGF␤ induction of ␣-smooth muscle actin (␣-SMA) protein and mRNA in fibroblasts. A, Fibroblasts from tak⫹/⫹ and tak⫺/⫺ mice were serum starved and left untreated or treated with TGF␤1 (4 ng/ml) for 24 hours. Proteins were subjected to Western blot analysis with anti–␣-SMA, anti-CCN2 (anti–connective tissue growth factor [anti-CTGF]), or anti-GAPDH antibodies. In tak⫺/⫺ mouse cells, ␣-SMA protein was not induced, whereas CCN2 induction was observed. Densitometry of the blots resulting from 4 separate experiments was performed, and quantitative results for ␣-SMA (upper graph) and CCN2 (lower graph) are shown. Bars from left to right show the mean and SD results in untreated tak⫹/⫹ mouse fibroblasts, untreated tak⫺/⫺ mouse fibroblasts, TGF␤-treated tak⫹/⫹ mouse fibroblasts, and TGF␤-treated tak⫺/⫺ mouse fibroblasts, respectively. ⴱ ⫽ P ⬍ 0.05 versus TGF␤-treated tak⫺/⫺ mouse fibroblasts (upper graph) or versus untreated tak⫹/⫹ and tak⫺/⫺ mouse fibroblasts (lower graph). B, Fibroblasts from tak⫹/⫹ and tak⫺/⫺ mice were serum starved and left untreated or treated with TGF␤1 (4 ng/ml) for 6 hours. Real-time polymerase chain reaction analysis was performed with primers detecting ␣-SMA, CCN2, and ribosomal RNA (rRNA); samples were standardized to rRNA. Values are the mean ⫾ SD fold increase in response to TGF␤ (in relation to basal expression in WT and KO cells [set at 1]); SDs within groups were ⬍10%. Induction of mRNA for ␣-SMA, but not CCN2, was impaired in tak⫺/⫺ cells. ⴱ ⫽ P ⬍ 0.05 versus TGF␤-treated tak⫹/⫹ cells. See Figure 2 for other definitions. www.lrgc.ca]). Briefly, RNA was harvested using TRIzol (Invitrogen, Carlsbad, CA), quantified, and quality was assessed with a Degradometer (www.dnaarrays.org) (Agilent, Palo Alto, CA). Biotinylated complementary RNA (cRNA) was prepared from 10 g of total RNA, according to instructions in the Affymetrix GeneChip Technical Analysis Manual (Affymetrix, Santa Clara, CA). Double-stranded cDNA was synthesized using Superscript II (Invitrogen) and oligo(dT) 24 primers. Biotin-labeled cRNA was prepared by in vitro transcription using the Bizarre High-Yield RNA Transcript Labeling kit (Enzo Brioche, New York, NY) with biotinylated UTP and CTP. Fifteen micrograms of labeled cRNA was hybridized to Mouse Genome 430 2.0 GeneChips for 16 hours at 45°C as described (Affymetrix). GeneChips were stained with streptavidin–phycoerythrin, an antibody solution, and a second streptavidin–phycoerythrin solution. Liquid handling was performed using a GeneChip Fluidics Station 450. GeneChips were scanned with a GeneChip Scanner 3000 (Affymetrix). Signal intensities of genes were determined with GCOS1.2 (Affymetrix), using default values for the statistical NECESSITY OF TAK1 FOR FIBROBLAST RESPONSES TO TGF␤ 237 Figure 4. Necessity of TGF␤-activated kinase 1 (TAK1) for TGF␤ induction of collagen gel contraction in fibroblasts. A, Floating gel contraction assay. Fibroblasts from tak⫹/⫹ and tak⫺/⫺ mice were placed in collagen gel lattices. After polymerization, lattices were detached from tissue culture plates and left untreated or treated with TGF␤1 (4 ng/ml) for 24 hours. Contraction was monitored by measuring gel weight. Values are the mean and SD from 3 independent experiments. ⴱ ⫽ P ⬍ 0.05 versus untreated control WT cells. B, Fibroblast-populated collagen lattice assay. Fibroblasts from tak⫹/⫹ and tak⫺/⫺ mice were placed in collagen gel lattices. After polymerization, lattices were left untreated or treated with TGF␤1 (4 ng/ml) for 24 hours. Contractile forces were monitored over 24 hours as previously described (13). Experiments were performed 3 times; results of a representative experiment are shown. 3-D ⫽ 3-dimensional (see Figure 2 for other definitions). expression algorithm parameters, a target signal of 150 for all probe sets, and a normalization value of 1. Normalization was performed with GeneSpring 7.2 (Agilent). An RMA preprocessor was used to import data from the .cel files. Data were first transformed (measurements ⬍0.01 set to 0.01) and then normalized per chip to the fiftieth percentile, and per gene to wild-type control samples. Experiments were performed twice, and fold changes were identified using the GeneSpring filter. The change between treated and untreated samples had to be at least 2-fold for a transcript to be identified as being altered. These criteria had to be met in both sets of experiments; the variation between the data points in each set was ⬍25%. Probe sets that were up-regulated by ⱖ2 fold, as well as 52 probe sets that were down-regulated (0.5-fold and below), were further analyzed using the above-mentioned software packages and DAVID (http://david.abcc.ncifcrf.gov/home.jsp) (1,2). The analysis was carried out at high stringency and generated 12 functional clusters. Real-time polymerase chain reaction (PCR). Cells were serum starved for 24 hours and treated with 4 ng TGF␤ for 6 hours. Total RNA was isolated using TRIzol, and the integrity of the RNA was verified by gel electrophoresis or with an Agilent bioanalyzer. For initial time course analysis, total RNA (25 ng) was reverse transcribed, and amplified using TaqMan Assays-on-Demand (Applied Biosystems, Foster City, CA) in a 15-l reaction volume containing 2 unlabeled primers and 6-carboxyfluorescein–labeled TaqMan MGB probe. Samples were combined with TaqMan One-step Master Mix (Applied Biosystems). Amplified sequences were detected using the ABI Prism 7900 HT sequence detector according to the instructions of the manufac- turer (PerkinElmer Cetus, Vaudreuil, Quebec, Canada). Triplicate samples were run, and transcripts and expression values were standardized to values obtained with control 28S RNA primers using the ⌬⌬Ct method as previously described (9–11). Variation within samples was ⬍10%. Floating collagen gel cultures and quantitation of gel contraction. Collagen gel experiments were performed essentially as previously described (12). Briefly, 24-well tissue culture plates were precoated with bovine serum albumin. Trypsinized fibroblasts were suspended in MCDB medium and mixed with collagen solution (1 part 0.2M HEPES [pH 8.0], 4 parts collagen [Vitrogen-100, 3 mg/ml] [Cohesion Technologies, Palo Alto, CA], and 5 parts MCDB X2), yielding a final concentration of 80,000 cells per ml and 1.2 mg/ml collagen. Collagen/cell suspension (1 ml) was added to each well. After polymerization, gels were detached from wells by adding 1 ml MCDB medium. Contraction of the gel was quantified based on loss of gel weight and decrease in gel diameter over a 24-hour period. For inhibition experiments, cells were preincubated in the presence of inhibitor for 30 minutes prior to initiation of the assay. Fibroblast-populated collagen lattices (FPCLs). Measurement of contractile force generated within a 3-dimensional, tethered floating fibroblast–populated collagen lattice was performed as described previously (13,14). Using 1 ⫻ 106 cells/ml of collagen gel (First Link, Birmingham, UK), we determined the force generated across the collagen lattice, with a culture force monitor. This instrument is capable of measuring the minute forces exerted by cells within a collagen lattice (13,14) over 24 hours as fibroblasts attach, spread, 238 SHI-WEN ET AL Table 1. Genes whose expression was increased ⬎2-fold in wild-type (WT) but not in tak-knockout (KO) mouse embryonic fibroblasts* Fold increase Affymetrix ID WT KO Cell motility genes 1428650_at 1426955_at 1452587_at 2.1 2.4 2.2 1.7 0.9 0.8 1416157_at 1417122_at 1450413_at 2.2 2.3 2.9 0.8 0.9 1.8 2.4 2.0 1.2 1.5 2.4 2.1 2.2 3.0 2.2 0.9 0.8 1.0 1.9 1.8 1448416_at 1423606_at 1417455_at 2.4 2.8 2.1 1.0 1.3 1.5 1448228_at 1448123_s_at 1416121_at 1419089_at 3.9 3.8 3.6 3.3 0.7 0.5 0.5 1.0 1449335_at 2.9 1.0 1452445_at 1452595_at 2.2 2.2 1.7 1.1 1416786_at 1421198_at Extracellular matrix genes 1426955_at 1418599_at 1416741_at 1460302_at 1451527_at Gene name Tensin 1 Procollagen type XVIII ␣1 Actin-related protein 2 homolog Vinculin Vav 3 Platelet-derived growth factor, B Activin A receptor type 1 ␣V integrin Procollagen type XVIII ␣1 Procollagen type XI ␣1 Procollagen type V ␣1 Thrombospondin 1 Procollagen c-endopeptidase enhancer 2 Matrix Gla protein Periostin Transforming growth factor ␤3 Lysyl oxidase Beta_IG Lysyl oxidase Tissue inhibitor of metalloproteinases 3 Tissue inhibitor of metalloproteinases 3 ADAM-28 ADAMTS-4 * Values are the mean from 2 independent experiments. migrate, and differentiate into myofibroblasts. Briefly, a rectangular fibroblast-seeded collagen gel was cast and floated in medium with 2% FCS in the presence or absence of TGF␤1 (4 ng/ml, 24 hours), tethered to 2 flotation bars on either side of the long edges, which were in turn attached to a ground point at one end and a force transducer at the other. Cell-generated tensional forces in the collagen gel were detected by the force transducer and the data logged into a personal computer. Graphic readings were produced every 15 seconds, providing a continuous output of force (dynes; 1 ⫻ 10⫺5N) generated (13,14). Statistical analysis. Student’s paired and unpaired t-tests were used to assess the significance of differences between treatment groups. P values less than 0.05 were considered significant. RESULTS Necessity of FAK/Src for TGF␤ induction of TAK phosphorylation in fibroblasts. We previously showed that the profibrotic protein endothelin 1 (ET-1) induces myofibroblast formation through a JNK-dependent mechanism, and that constitutive ET-1–dependent JNK activation is a key feature of fibrotic fibroblasts in SSc (15). TGF␤ can further induce JNK via a FAKdependent mechanism, leading to myofibroblast formation (6). Moreover, we found that TAK was constitutively phosphorylated in SSc fibroblasts (Figure 1A). To investigate the control mechanisms through which TGF␤ can induce a fibrotic phenotype in fibroblasts, we evaluated the role of TAK1 in this pathway. We first used fibroblasts cultured from wild-type mice to verify, using Western blot analysis with anti– phospho-TAK1 and anti-TAK1 antibodies, that TGF␤ could induce TAK in fibroblasts (Figure 1); enhanced adhesive signaling is a feature of fibrotic cells (3,16). To assess whether FAK was required in order for TGF␤ to induce JNK, fibroblasts were pretreated with the FAK/ Src inhibitor PP2 prior to addition of TGF␤. The resultant protein extracts were subjected to Western blot analysis with anti–phospho-JNK and anti-JNK antibodies. TGF␤ induced TAK1 phosphorylation in fibroblasts, and this was significantly impaired by PP2 (Figure 1B). Moreover, TAK phosphorylation was reduced in response to TGF␤ in FAK-knockout mouse fibroblasts (Figure 1B). These results suggested that TGF␤ induces TAK1 in fibroblasts, in a FAK-dependent manner. Impairment of TGF␤-induced JNK phosphorylation in the absence of TAK1. The ability of TGF␤ to induce the Smad pathway does not involve TAK1 (7). Since FAK/Src is required for JNK activation in fibroblasts (6), we assessed whether the ability of TGF␤ to induce JNK phosphorylation was impaired in the absence of TAK1. Cells were serum starved for 24 hours and then treated with TGF␤ for 1 hour. Protein was then harvested from the cell layers and subjected to Western blot analysis with anti-JNK1 and anti–phospho-JNK1 antibodies. The results demonstrated that the ability of TGF␤ to induce JNK1 phosphorylation was impaired in TAK1-deficient mouse fibroblasts (Figure 2). Impairment of TGF␤-induced ␣-SMA expression and ECM contraction in the absence of TAK1. We then wished to examine the contribution of FAK/Srcdependent TAK1 activation to the ability of TGF␤ to induce a tissue remodeling phenotype in fibroblasts. To address this, we first assessed whether TGF␤-induced myofibroblast formation was impaired in TAK1deficient fibroblasts. Wild-type mouse fibroblasts were left untreated or treated with TGF␤ for 24 hours, and cells were then subjected to Western blot analysis with an anti–␣-SMA antibody. TGF␤ potently increased anti–␣-SMA protein expression in fibroblasts from NECESSITY OF TAK1 FOR FIBROBLAST RESPONSES TO TGF␤ 239 Figure 5. Inability of TGF␤ to induce expression of mRNA for various tissue remodeling molecules in tak⫺/⫺ fibroblasts. Fibroblasts from tak⫹/⫹ and tak⫺/⫺ mice were serum starved and left untreated or treated with TGF␤1 (4 ng/ml) for 6 hours. Real-time polymerase chain reaction analysis was performed with primers detecting vinculin, tissue inhibitor of metalloproteinases 3 (TIMP-3), thrombospondin 1 (TSP-1), and lysyl oxidase (LOX). Samples were standardized to ribosomal RNA. Values are the mean ⫾ SD fold increase in response to TGF␤ (in relation to basal expression in WT and KO cells [set at 1]) in 1 representative experiment (of 3 performed); SDs within groups were ⬍10%. ⴱ ⫽ P ⬍ 0.05 versus TGF␤-treated tak⫹/⫹ cells. See Figure 2 for other definitions. tak⫹/⫹ mice, but not in fibroblasts from tak⫺/⫺ mice (Figure 3A). Results were confirmed by real-time PCR analysis of mRNA harvested from cells that were untreated or were treated with TGF␤ for 6 hours (Figure 3B). Intriguingly, the ability of TGF␤ to induce CCN2 protein and mRNA, an action not involving JNK (17), was not impaired in tak⫺/⫺ mouse embryonic fibroblasts. Moreover, TGF␤ was able to induce tak⫹/⫹ mouse fibroblasts, but not tak⫺/⫺ mouse fibroblasts, to contract in a floating collagen gel matrix (Figure 4A). Similar results were obtained using an FPCL assay, in which force generation across a floating collagen gel matrix fixed at one end was examined (Figure 4B). Collectively, these results are consistent with the notion that FAK/TAK-dependent JNK activation is essential to the ability of TGF␤ to signal in fibroblasts and, in particular, in TGF␤-induced ␣-SMA expression and matrix contraction. Impairment of TGF␤-induced expression of fibrogenic genes in takⴚ/ⴚ mouse fibroblasts. To evaluate the contribution of TAK1 to the ability of TGF␤ to induce gene expression in fibroblasts, we cultured tak⫹/⫹ and tak⫺/⫺ mouse fibroblasts to 80% confluence and serum starved the cells for 24 hours. Cells were then left untreated or treated with TGF␤ (4 ng/ml) for an additional 6 hours. Total RNA was prepared from these cells, reverse transcribed, and applied to Affymetrix MOE430 arrays. Experiments were performed twice, and average induction values were obtained. Analysis of the data by GeneSpring revealed 265 transcripts that were increased ⬎2-fold by TGF␤ in tak⫹/⫹ mouse fibroblasts. Of these, 194 were not induced ⬎2-fold in tak⫺/⫺ mouse fibroblasts. Cluster analysis using DAVID revealed that expression of profibrotic transcripts involved in cell motility and ECM was reduced in response to TGF␤ in TAK1-deficient cells. These transcripts included genes for thrombospondin 1, tissue inhibitor of metalloproteinases 3, vinculin, lysyl oxidase, and several collagens (Table 1). Results obtained using microarray analysis for these transcripts were verified by real-time PCR analysis of RNA isolated from tak⫹/⫹ and tak⫺/⫺ mouse embryonic fibroblasts that were left untreated or treated with TGF␤ for 6 hours (Figure 5). Collectively, these findings 240 SHI-WEN ET AL suggest that TAK1 is required for a subset of TGF␤ responses in fibroblasts, including the induction of certain profibrotic genes. DISCUSSION In the present study, we examined the signaling events downstream of TGF␤ and FAK/Src that lead to induction of a tissue remodeling phenotype. TGF␤ promotes ECM production and contraction in fibroblasts, leading to deposition of granulated tissue (18). Although most studies have focused on the general TGF␤ signaling pathway, including TGF␤ receptors and Smads, in controlling ECM production and remodeling, it is now understood that additional pathways mediate the action of TGF␤ (18). In the potential development of therapies that target these pathways, it is hoped that wound healing and scarring responses could be affected while other, perhaps beneficial, effects of TGF␤ are left intact (18). MAP kinase cascades play key roles in driving tissue repair and fibrogenesis in response to TGF␤; for example, both the Ras/MEK/ERK and FAK/JNK cascades control the expression of TGF␤ target genes in fibroblasts, in a gene-specific manner (6,17,19). Both of these pathways appear to work independent of the Smad pathway (6,17,19). Previously, we have shown that JNK mediates TGF␤ induction of ET-1 and that constitutive JNK activation, mediated by endogenous ET-1 production, contributes at least partially to the persistent fibrotic phenotype of scleroderma lung fibroblasts (15). Moreover, we have demonstrated that JNK activation in response to TGF␤ in normal fibroblasts depends on FAK/Src (6). In this study, we showed that TGF␤induced JNK phosphorylation is impaired in the absence of TAK1. Moreover, deletion of TAK1 resulted in a failure of fibroblasts to support not only JNK activation in response to TGF␤, but also the expression of mRNA for ␣-SMA and a group of profibrotic TGF␤-induced genes, and ECM contraction. All of these responses required JNK and FAK in wild-type mouse fibroblasts (6). It was of interest that, in the FPCL assay, TAK1deficient cells contracted less well than did wild-type cells. This suggests that TAK1 is required for basal fibroblast function. Indeed, microarray analysis demonstrated that TAK1-deficient fibroblasts have reduced gene expression even in the absence of added TGF␤ (results not shown). It should also be noted that the FPCL assay measures myofibroblast induction in response to mechanical tension as well as to factors such as cell migration (13). Thus, it is indeed possible that TAK1 deficiency may also affect basal fibroblast function. Exploring this possibility was beyond the scope of the current study. Since TAK1-deficient mice died early in development, our studies with TAK1-knockout mouse cells required use of a transformed line derived from multiple embryos, in order to obtain sufficient cells for analysis (7). It is therefore theoretically possible that biases in clonal selection may have contributed to the present results. Overall, however, our findings suggest that, in fibroblasts, TAK is required downstream of FAK/Src to mediate a subset of profibrotic responses to TGF␤. TAK1 was constitutively phosphorylated in fibroblasts isolated from 4 SSc patients, relative to healthy control fibroblasts. This raises the intriguing possibility that TAK1 inhibitors (although we are aware of none that are readily commercially available) might be used to combat the fibrosis observed in SSc. It is important to note that since TAK1-deficient mice die in utero (7), it is possible that TAK1 is responsible for functions other than those described herein. Further evaluation of the suitability of TAK1 as a therapeutic target for fibrosis must await the development of conditionally deleted mouse strains. In conclusion, the findings reported herein provide evidence that the ability of TGF␤ to induce ␣-SMA expression and myofibroblast formation in fibroblasts depends on TAK1. Compared with general blockade of TGF␤ signaling by antagonizing TGF␤ receptors or Smads (18,20), it is likely that modifying TAK1 activity may be more suitable for control of tissue remodeling and development of selective antifibrotic therapies for disorders such as scleroderma. AUTHOR CONTRIBUTIONS Dr. Leask had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Shi-wen, Pala, Denton, Abraham, Leask. Acquisition of data. Shi-wen, Parapuram, Pala, Chen, Carter, Eastwood. Analysis and interpretation of data. Shi-wen, Pala, Carter, Eastwood, Denton, Abraham, Leask. Manuscript preparation. Shi-wen, Denton, Abraham, Leask. Statistical analysis. Shi-wen. REFERENCES 1. Hinz B, Gabbiani G. Mechanisms of force generation and transmission by myofibroblasts. Curr Opin Biotechnol 2003;14:538–46. 2. Desmouliere A, Chaponnier C, Gabbiani G. Tissue repair, contraction, and the myofibroblast. Wound Repair Regen 2005;13: 7–12. NECESSITY OF TAK1 FOR FIBROBLAST RESPONSES TO TGF␤ 3. Chen Y, Shi-wen X, van Beek J, Kennedy L, McLeod M, Renzoni EA, et al. Matrix contraction by dermal fibroblasts requires transforming growth factor-␤/activin-linked kinase 5, heparan sulfate-containing proteoglycans, and MEK/ERK: insights into pathological scarring in chronic fibrotic disease. Am J Pathol 2005;67:1699–711. 4. Parsons JT. Focal adhesion kinase: the first ten years. J Cell Sci 2003:116:1409–16. 5. Cox BD, Natarajan M, Stettner MR, Gladson CL. New concepts regarding focal adhesion kinase promotion of cell migration and proliferation. J Cell Biochem 2006;99:35–52. 6. Liu S, Shi-wen X, Kennedy L, Pala D, Carter DE, Black CM, et al. FAK is required for TGF␤-induced JNK phosphorylation in fibroblasts: implications for acquisition of a matrix remodeling phenotype. Mol Biol Cell 2007;18:2169–78. 7. Shim JH, Xiao C, Paschal AE, Bailey ST, Rao P, Hayden MS, et al. TAK1, but not TAB1 or TAB2, plays an essential role in multiple signaling pathways in vivo. Genes Dev 2005;19:2668–81. 8. Kennedy L, Liu S, Shi-wen X, Chen Y, Eastwood M, Carter DE, et al. CCN2 is necessary for the function of mouse embryonic fibroblasts. Exp Cell Res 2007;313:952–64. 9. Shi-wen X, Stanton LA, Kennedy L, Pala D, Chen Y, Howat SL, et al. CCN2 is necessary for adhesive responses to transforming growth factor-␤1 in embryonic fibroblasts. J Biol Chem 2006;281: 10715–26. 10. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-⌬⌬CT method. Methods 2001;25:402–8. 11. Pala D, Kapoor M, Woods A, Kennedy L, Liu S, Chen S, et al. FAK/src suppresses early chondrogenesis: central role of CCN2. J Biol Chem 2008;283:9239–47. 12. Shi-Wen X, Chen Y, Denton CP, Eastwood M, Renzoni EA, 241 13. 14. 15. 16. 17. 18. 19. 20. Bou-Gharios G, et al. Endothelin-1 promotes myofibroblast induction through the ETA receptor via a rac/phosphoinositide 3-kinase/Akt-dependent pathway and is essential for the enhanced contractile phenotype of fibrotic fibroblasts. Mol Biol Cell 2004; 15:2707–19. Eastwood M, McGrouther DA, Brown RA. Fibroblast responses to mechanical forces. Biochim Biophys Acta 1994;1201:186–92. Grinnell F. Fibroblast biology in three-dimensional collagen matrices. Trends Cell Biol 2003;13:264–9. Shi-wen X, Rodriguez-Pascual F, Lamas S, Holmes A, Howat S, Pearson JD, et al. Constitutive ALK5-independent c-Jun Nterminal kinase activation contributes to endothelin-1 overexpression in pulmonary fibrosis: evidence of an autocrine endothelin loop operating through the endothelin A and B receptors. Mol Cell Biol 2006;26:5518–27. Mimura Y, Ihn H, Jinnin M, Asano Y, Yamane K, Tamaki K. Constitutive phosphorylation of focal adhesion kinase is involved in the myofibroblast differentiation of scleroderma fibroblasts. J Invest Dermatol 2005;124:886–92. Leask A, Holmes A, Black CM, Abraham DJ. Connective tissue growth factor gene regulation: requirements for its induction by transforming growth factor-␤2 in fibroblasts. J Biol Chem 2003; 278:13008–15. Leask A, Abraham DJ. TGF-␤ signaling and the fibrotic response. FASEB J 2004;18:816–27. Stratton R, Rajkumar V, Ponticos M, Nichols B, Shiwen X, Black CM, et al. Prostacyclin derivatives prevent the fibrotic response to TGF-␤ by inhibiting the Ras/MEK/ERK pathway. FASEB J 2002;16:1949–51. Leask A. Scar wars: is TGF␤ the phantom menace in scleroderma? [review]. Arthritis Res Ther 2006;8:213.