Endothelin-1 induces expression of matrix-associated genes in lung fibroblasts through MEK/ERK.

The endothelins are a family of endothelium-derived peptides that possess a variety of biological activities, including potent vasoconstriction. Endothelin-1 (ET-1) is up-regulated during tissue repair and pulmonary fibrosis. Here, we use genome-wide expression array analysis to show that the addition of ET-1 (100 nm, 4 h) to normal lung fibroblasts directly induces expression of matrix and matrix-associated genes, including the profibrotic protein CCN2 (connective tissue growth factor, or CTGF). ET-1 induces the MEK/ERK MAP kinase pathway in fibroblasts. Blockade of the MEK/ERK kinase pathway with U0126 abrogates the ability of ET-1 to induce expression of matrix and matrix-associated mRNAs and the CCN2 protein. The CCN2 promoter possesses an ET-1 response element, which maps to the previously identified basal control element-1 (BCE-1) site. Our results suggest that ET-1 induces a program of matrix synthesis in lung fibroblasts and that ET-1 may play a key role in connective tissue deposition during wound repair and in pulmonary fibrosis.

The endothelins are a family of endothelium-derived peptides that possess a variety of biological activities, including potent vasoconstriction. Endothelin-1 (ET-1) is up-regulated during tissue repair and pulmonary fibrosis. Here, we use genome-wide expression array analysis to show that the addition of ET-1 (100 nM, 4 h) to normal lung fibroblasts directly induces expression of matrix and matrix-associated genes, including the profibrotic protein CCN2 (connective tissue growth factor, or CTGF). ET-1 induces the MEK/ERK MAP kinase pathway in fibroblasts. Blockade of the MEK/ERK kinase pathway with U0126 abrogates the ability of ET-1 to induce expression of matrix and matrix-associated mRNAs and the CCN2 protein. The CCN2 promoter possesses an ET-1 response element, which maps to the previously identified basal control element-1 (BCE-1) site. Our results suggest that ET-1 induces a program of matrix synthesis in lung fibroblasts and that ET-1 may play a key role in connective tissue deposition during wound repair and in pulmonary fibrosis.
As a response to environmental insults or a consequence of local inflammatory processes, structural damage to tissue can occur, triggering a wound-healing response. This response consists of an integrated series of biochemical, immunological, and structural changes that result in the de novo synthesis of a new epithelium, blood vessels, and connective tissue (1). The proper repair of connective tissue requires the synthesis and organization of new ECM 1 components, such as collagen and fibronectin (2).
A growing body of evidence implicates the vasoconstrictive peptide endothelin-1 (ET-1) as a key mediator of tissue repair (3). Each of the three known endothelin isoforms (-1, -2, and -3) arise by proteolytic processing of large precursors (ϳ200 amino acid residues). Intermediates, termed big ET-1, -2, and -3 (38 -41 amino acids), are excised from pre-propeptides at sites containing paired basic amino acids. Big endothelins, which have little or no biological activity (4), are cleaved at Trp-21-Val/Ile-22 to produce mature, 21-residue, biologically active peptides (5,6). The enzyme responsible for the specific cleavage at Trp-21 has been termed the endothelin-converting enzyme (7,8). Injury and the wound-healing response lead to stabilization of endothelin-converting enzyme-1 mRNA and to the generation of bioactive endothelin (9). Elevated levels of ET-1 have been shown in patients with fibrotic disease, suggesting that ET-1 may play a key role not only in normal wound repair but also in the pathogenesis of fibrosis (10 -14).
ET-1 demonstrates a wide range of biological properties, including significant mitogenic activity toward a number of cell types such as smooth muscle cells and fibroblasts (15). ET-1 also promotes the contractile ability of normal fibroblasts (16), which is essential for wound closure and reconstitution of the dermis (17). In addition, ET-1 modifies extracellular matrix metabolism (15, 18 -20). For example, ET-1 enhances collagen types I and III and decreases matrix metalloproteinase-1 (MMP-1) mRNA and protein expression in dermal fibroblasts (18,20). However, the signal transduction mechanism and transcription factors through which ET-1 affects gene expression is largely unknown. Furthermore, it is unclear to what extent ET-1 might contribute to wound repair or fibrogenic responses.
In this report, we investigate the functional and mechanistic contribution of ET-1 to matrix synthesis in primary human lung fibroblasts. Our results provide new insights into ET-1 biology and suggest a role for ET-1 in enhancing matrix expression and organization during tissue repair and fibrogenesis.

MATERIALS AND METHODS
Cell Culture-Primary human lung fibroblasts were grown from macroscopically and histologically normal lung resection specimens by explant culture as described previously (21) using DMEM with 10% fetal bovine serum (Invitrogen). Cells were used between passages 2 and 5.
Gene Array Analysis-Lung fibroblasts were serum starved for 18 h and treated with 100 nM ET-1 for 4 h. At the end of the treatment period, total RNA was harvested (Trizol; Invitrogen) and quantified, and integrity was verified by denaturing gel electrophoresis. Equal amounts of identically treated RNA were pooled and reverse tran-scribed (Invitrogen) into cDNA, which was then in vitro transcribed into biotinylated cRNA. The target cRNA was then fragmented and hybridized to the Affymetrix human U133A array following Affymetrix (Santa Clara, CA) protocol. Hybridization of cRNA to Affymetrix human U133A chips, signal amplification, and data collection were performed using an Affymetrix fluidics station and chip reader. Chip files were scaled to an average intensity of 100 per gene and analyzed using the Affymetrix version 5.0 (MAS5) comparison analysis software. Experiments were performed twice, and the fold changes presented in Table I are an average of these independent studies. Criteria indicated by Affymetrix were used to determine robust changes in gene expression. Briefly, transcripts were defined as up-regulated by ET-1 only when identified as "present" (ET-1-treated chips) by the Affymetrix detection algorithm and "significantly increased," as determined by the Affymetrix change algorithm, with a change p value of Ͻ0.01. The fold change between treated and untreated samples had to be at least 1.5-fold to identify a transcript as being altered. These criteria had to be met in both sets of experiments. We have entered one experiment (ET-1 untreated/treated) in the NCBI Gene Expression Omnibus (GEO) as samples GSM15500 and GSM15501, and the other experiment has been entered as GSM17262 and GSM17263. The GEO series, including all samples, has been entered as GSE1081 (www.ncbi.nlm.nih.gov/geo).
Western Blot Analysis-Lung fibroblasts were grown to confluence in DMEM with 10% fetal bovine serum and cultured in DMEM and 0.5% bovine serum albumin for 24 h. To ascertain the effect of ET-1 on CCN2 (connective tissue growth factor, or CTGF) expression, cells were stimulated with 100 nM ET-1 for 48 h with 0.5% bovine serum albumin treatment. To assess the effect of ET-1 on the stimulation of signaling pathways, ET-1 was added to cells for up to 16 h. When appropriate, cells were incubated with an inhibitor for 45 min prior to the addition of ET-1. The activities of inhibitors were verified by examining their ability to block phosphorylation of Akt or by assessing their ability to block induction of promoters responsive to particular signaling cascades (Clontech). Cell layer lysates were subjected to SDS-PAGE and electroblotted onto nitrocellulose membranes. CCN2 protein was detected using an anti-CCN2 (CTGF) antibody (Santa Cruz Biotechnology), and the glyceraldehyde-3-phosphate (GAPDH) protein was detected using an anti-GAPDH antibody (Santa Cruz Biotechnology). Akt, phospho-Akt, p38, phospho-p38, p42/44 MAPK, and phospho-p42/44 MAPK were detected using antibodies as described by the manufacturer (Cell Signaling Technology, Beverly, MA). Blots were then hybridized with an appropriate horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch Laboratories), and protein was detected using chemiluminescence (Amersham Biosciences).
Promoter Assays-CCN2 promoter/secreted enhanced alkaline phosphatase (SEAP) reporter constructs were as described previously (25,26). Promoter/reporter constructs were transfected into lung fibroblasts using FuGENE6 transfection reagent (Roche Applied Science) according to the manufacturer's instructions. Promoter/reporter plasmids were cotransfected with pCMV-␤Gal (Clontech), which was used to adjust for differences in transfection efficiencies between samples. Following transfection, cells were incubated in DMEM with 0.5% fetal bovine serum for 18 h. Media were changed, and cells were incubated in the presence or absence of inhibitors for 45 min and cultured for an additional 24 h in the presence or absence of ET-1 (100 nM). Media were taken for SEAP assays. Fibroblasts were rinsed once with phosphatebuffered saline, and cellular protein was extracted using 200 l of reporter lysis buffer (Promega Corp, Madison, WI). Reporter gene activity was measured by luminometry (Turner Designs, Sunnyvale, CA) using SEAP and ␤-galactosidase assays (Tropix Inc. Bedford, MA) according to the manufacturers' instructions. Values given are mean Ϯ S.E. of triplicate assays from three individual experiments.

RESULTS
Global Expression Profiling in Response to ET-1-To identify transcripts induced by ET-1, we treated primary lung fibroblasts for 4 h in the presence or absence of 100 nM ET-1. After this period of incubation, we harvested total RNA. Affymetrix U133A gene arrays were hybridized to these labeled cRNAs derived from the total RNA. Gene array analysis revealed that a 4-h treatment of primary lung fibroblasts with ET-1-modified expression of a cluster of transcripts encoding proteins that would be expected to promote matrix deposition and remodeling (Table I). ET-1 treatment induced expression of the matrix protein collagen types IV, V, and VII (27). ET-1 treatment also induced expression of the collagen-modifying enzyme lysyl ox-
ET-1 Enhances CCN2 Expression through a MEK/ERK-dependent Mechanism-CCN2 (CTGF) is an important profibrotic protein that induces collagen and acts with TGF␤ to promote sustained fibrosis in vivo (29,34). Because our array analysis showed that CCN2 mRNA was induced by ET-1 treatment, we sought to explore the notion that ET-1 might contribute to fibrotic responses in vivo through the induction of CCN2. Thus, we treated normal lung fibroblasts for 24 h with 100 nM ET-1. Cell layers were harvested, and equal amounts of protein (20 g) were subjected to Western blot analysis with an anti-CCN2 antibody. Confirming our microarray data, we found that 100 nM ET-1 induced expression of CCN2 protein (Fig. 1).
To begin to probe the signaling mechanism through which ET-1 induced CCN2, we performed Western blot analysis of protein extracts prepared from lung fibroblasts that had been treated with ET-1 for different lengths of time. Using antiphospho-ERK, anti-phospho-p38, and anti-phospho-Akt antibodies, we showed that ET-1 activated ERK (p42/p44), p38, and PI3-kinase/Akt pathways in lung fibroblasts (Fig. 2). To ad-dress the contribution of these pathways to the ET-1 induction of CCN2, we used Western blot analysis with an anti-CCN2 antibody to show that a 45-min pre-incubation with the MEK/ ERK inhibitors PD98059 and U0126 (35,36), prior to the addition of ET-1, blocked the ability of ET-1 to induce CCN2 protein (Fig. 1, bottom panel). Conversely, the addition of the p38 inhibitor SB203580 (37) or the Akt/PI-3-kinase inhibitors wortmannin and LY294002 (38,39) had no discernible effect on the ability of ET-1 to modulate CCN2 production. In addition, a 45-min preincubation of lung fibroblasts with the dual specificity ETA/B receptor antagonist bosentan (40) blocked the ability of ET-1 to induce CCN2 (Fig. 1). Thus, ET-1 stimulated CCN2 protein expression in normal human lung fibroblasts through a MEK/ERK-dependent mechanism that required either the ETA or the ETB receptor.
ET-1 Modifies Expression of Fibrogenic Genes in a MEK/ERKdependent Mechanism-After determining that ET-1 induced CCN2 protein through a MEK/ERK-dependent mechanism, we decided to investigate further the contribution of MEK/ERK to ET-1 induced gene expression. Using real time polymerase chain reaction analysis, we confirmed that ET-1 induced MMP-1, CCN2, collagen IV, TSP-1, and TIMP-3 mRNAs (Fig.  3). Induction of MMP-1, CCN2, TSP-1, and TIMP-3 seemed to be direct, showing potent induction 2 h after the application of ET-1. To determine whether ET-1 induced these transcripts via a MEK/ERK-dependent mechanism, cells were pre-incubated for 45 min with or without the MEK/ERK MAP kinase inhibitor U0126 (10 M) prior to the addition of ET-1. We found that pre-incubation of cells with U0126 potently inhibited the ability of ET-1 to induce MMP-1, CCN2, collagen IV, TSP-1, and TIMP-3 mRNA expression. Thus, ET-1 is a potent fibrogenic peptide, able to induce mRNA expression of matrix genes in lung fibroblasts via MEK/ERK.
ET-1 Activates the CCN2 Promoter through the BCE-1 Element-After establishing that ET-1 induced the appearance of CCN2 protein and mRNA in lung fibroblasts, we sought to investigate further the mechanism through which ET-1 performs this function. For these experiments, we assessed whether an ET-1 response element existed in the CCN2 promoter. To perform this experiment, we transfected a full-length CCN2 promoter/SEAP reporter construct, spanning nucleotides Ϫ805 to ϩ17 of the CCN2 promoter (Fig. 4, Ϫ805; Ref. 26), into lung fibroblasts. After an 18-h serum starvation step, cells were cultured in the presence or absence of 100 nM ET-1 for an additional 24 h. Consistent with our data showing that ET-1 induced CCN2 protein and mRNA expression, we found that ET-1 induced CCN2 promoter activity (Fig. 4). To further delineate the ET-1 response element in the CCN2 promoter, we transfected two additional CCN2 promoter/SEAP reporter constructs into fibroblasts. These new constructs contained segments of the CCN2 promoter spanning nucleotides Ϫ244 to ϩ17 (Fig. 4, Ϫ244) or Ϫ86 to ϩ17 (Fig. 4, Ϫ86; Ref. 26). We found that, although the fragment of the CCN2 promoter between Ϫ244 to ϩ17 could respond to ET-1, the removal of nucleotides Ϫ244 to Ϫ86 abolished the ability of the CCN2 promoter to respond to ET-1 (Fig. 4). Thus, the ET-1 response element in the CCN2 promoter lay between nucleotides Ϫ244 and Ϫ86.
Previously, we had analyzed the transcription factor binding sites lying between nucleotides Ϫ244 and Ϫ86 of the CCN2 promoter and identified a Smad binding element, which is required for the TGF␤-induction of CCN2, and a basal control element-1 (BCE-1) site, which is required for basal CCN2 expression but is not involved with the TGF␤-induction of CCN2 (25). To assess the roles of Smad and BCE-1 elements in the ET-1 induction of the CCN2 promoter, we transfected into lung fibroblasts constructs that contained mutations in the Smad and BCE-1 elements of the CCN2 promoter (Smad and BCE-1, respectively, Fig. 4) but were otherwise identical to the fulllength (Ϫ805 to ϩ17) CCN2 promoter construct. We found that removal of the Smad element had no significant impact on the ability of the CCN2 reporter to respond to ET-1 (Fig. 4, Ϫ805  Smad). Conversely, removal of the BCE-1 element abolished the response of the CCN2 promoter to ET-1 (Fig. 4, Ϫ805  BCE-1). Thus, ET-1 promotes CCN2 expression in a fashion that is independent of Smad/TGF␤ signaling but dependent on BCE-1. These results suggest that ET-1 promotes gene expression in a manner that is independent of TGF␤.

DISCUSSION
In this study, we used primary human lung fibroblasts as an in vitro model with which to probe the role of ET-1 in ECM production. Previously, we have shown that ET-1 enhances collagen types I and III and decreases MMP-1 mRNA and protein expression in dermal fibroblasts (18,20). In this report, we extend these results by showing that ET-1 in itself is capable of inducing the expression of ECM or ECM-associated genes. ET-1 induced collagen type IV, a key basement membrane component (41). ET-1 induced collagen type VII, the major component of anchoring fibrils that attach epidermis and dermis; mutations in the COL7A1 gene encoding type VII collagen cause a severe blistering disease, dystrophic epidermolysis bullosa (42). In addition, ET-1 induces collagen type V, which interacts with collagen type I to regulate the diameters of type I collagen fibrils (43). Thus, by the ability of ET-1 to induce collagens IV, V, and VII, ET-1 promotes the woundhealing response by enhancing reconstitution of the basement membrane and wound closure.
ET-1 induces CCN2, a member of the CCN family of proteins (44,45). CCN2 is a cysteine-rich, pro-adhesive, matricellular protein that plays an essential role in the formation of blood vessels, bone, and connective tissue (46). Because the expression of this protein is potently induced by TGF␤ in a Smad-dependent fashion (22), it has been hypothesized that CCN2 mediates several of the downstream actions of TGF␤ (26). Ours is the first report to show that ET-1 induces CCN2 and, thus, that CCN2 could also be a downstream mediator of ET-1 activity. The ET-1 response element of the CCN2 promoter is independent of its Smad element, which is required for the TGF␤induction of CCN2 (22,47). Instead, the ET-1 induction of CCN2 occurs through the BCE-1 element of the CCN2 promoter, which is not required for the TGF␤-induction of CCN2 (22). These results suggest that TGF␤ and ET-1 might work Ϫ805, a construct containing nucleotides Ϫ805 to ϩ 17 of the CCN2 promoter; Ϫ244, a construct containing nucleotides Ϫ244 to ϩ17 of the CCN2 promoter; Ϫ86, a construct containing nucleotides Ϫ86 to ϩ17 of the CCN2 promoter; Ϫ805 Smad, a construct containing a mutation in the Smad binding element of the CCN2 promoter; Ϫ805 BCE-1, a construct containing a mutation in the BCE-1 binding element of the CCN2 promoter. These latter two mutations were made in the context of the larger Ϫ805 construct.

ET-1 Induces Matrix Synthesis in Fibroblasts via MEK/ERK
independently on the CCN2 promoter to elevate CCN2 expression in vivo and, thus, that TGF␤ and ET-1 might cooperate to produce fibrogenic responses in vivo thorough the coordinate induction of CCN2 (Fig. 5). The identity of the factor(s) binding to the BCE-1 site of the CCN2 promoter is not known but is currently under investigation.
It is interesting to note that we have shown previously that a 48-h treatment of normal dermal fibroblasts with ET-1 caused the induction of collagen type I mRNA (20). In this report, using normal lung fibroblasts, our transcriptional profiling using the Affymetrix system showed that ET-1 did not induce collagen type I mRNA after 4 h. Using real time PCR, we confirmed that ET-1 induced type I collagen mRNA commencing 6 h after ET-1 treatment (not shown). Collectively, these results suggest that the ability of ET-1 to induce type I collagen might be indirect. However, in this report we showed appreciable induction of CCN2 and thrombospondin-1 mRNA 2 h after ET-1 treatment. CCN2 induces type I collagen (29,48,49), and thrombospondin-1 activates latent TGF␤ (30,50). Thus, the ability of ET-1 to induce collagen type I expression may be indirect, possibly via the induction of CCN2 or by the activation of latent TGF␤ (Fig. 5). It is interesting to note that, in endothelial cells, TGF␤ induces ET-1 expression via Smad and Ap-1 sites in the ET-1 promoter (51). Although TGF␤ has been shown to be a key pro-fibrotic cytokine, its action can be amplified or suppressed by several other cytokines (52). Understanding the interactions among these profibrotic and antifibrotic cytokines is likely to have a major impact in understanding wound-healing and fibrotic responses in vivo. Thus, further research elucidating the mechanism by which ET-1, TGF␤ and CTGF might interact to promote pro-fibrotic responses is of major importance.
Because ET-1 is markedly up-regulated during tissue repair (9) and in patients with fibrotic disease, these results suggest that ET-1 may play a key role not only in normal wound repair but also in the pathogenesis of fibrosis (10 -14). Elevated levels of circulating ET-1 occurred in patients with skin and lung fibrosis, a finding that correlated with the severity of the fibrotic phenotype (53)(54)(55). This increase in circulating ET-1 was paralleled by an increase in ET-1 synthesis in vivo (53)(54)(55). Thus, ET-1 may be an important therapeutic target in the modulation of fibrogenesis. Indeed, our data suggest that ET-1 is a potent fibrogenic peptide for lung fibroblasts, able to induce matrix production by lung fibroblasts via a MEK/ERK-depend-ent mechanism requiring either the ETA or the ETB receptors (Fig. 5).
The results presented in this report are consistent with several studies published recently that have linked the p42/p44 (MEK/ERK) MAP kinase signaling cascade and fibrosis. For example, the TGF␤ induction of CCN2 requires the ras/MEK/ ERK signaling cascade (47,56,57). Recently, we showed that the synthetic prostacyclin iloprost, which alleviates symptoms of fibrosis in patients with the fibrotic disease systemic sclerosis (49), works, at least in part, through the antagonism of the MEK/ERK pathway (57). Collectively, our results suggest that the antagonism of MEK/ERK might be an effective anti-fibrotic approach.
In summary, the results presented in our current report emphasize that ET-1 is an important regulator of extracellular matrix biosynthesis by lung fibroblasts and further emphasize the key role of the MEK/ERK cascade in tissue injury and wound healing. Our results further suggest that there may be a potential therapeutic advantage in using MEK/ERK kinase inhibitors or endothelin antagonists to ameliorate the pathological scarring observed in pulmonary fibrosis. FIG. 5. Schematic diagram showing that ET-1 contributes to ECM synthesis in lung fibroblasts. ET-1 promotes ECM synthesis in normal and lung fibroblasts through MEK/ERK kinase. ET-1 also induces expression of the profibrotic protein CCN2, which induces matrix synthesis (49,50). ET-1 also induces thrombospondin-1 (30), a potent activator of TGF␤, a key inducer of ECM production and fibrotic responses (61, 62).