Transforming Growth Factor-β1-induced Transcript 1 Protein, a Novel Marker for Smooth Muscle Contractile Phenotype, Is Regulated by Serum Response Factor/Myocardin Protein

Background: TGFB1I1 is involved in vascular injury, but the regulation of TGFB1I1 expression is unknown.

Results: TGFB1I1 is specifically expressed in smooth muscle cells and regulated by SRF/myocardin.

Conclusion: TGFB1I1 is a novel marker for the smooth muscle contractile phenotype.

Significance: This study provides the first evidence that TGFB1I1 is an SRF/myocardin-regulated smooth muscle marker and critical for maintaining the smooth muscle contractile phenotype.

Abstract

Serum response factor (SRF) plays a central role in regulating expression of smooth muscle-specific genes partly by associating with the potent tissue-specific cofactor myocardin. Previous studies have shown that transforming growth factor-β1-induced transcript 1 (TGFB1I1, also known as Hic-5) is a TGF-β-responsive gene and is involved in the cellular response to vascular injury, but the regulation of TGFB1I1 expression remains elusive. In this report, we demonstrated that TGFB1I1 is a novel marker for the smooth muscle contractile phenotype and is regulated by SRF/myocardin. We found that TGFB1I1 is specifically expressed in smooth muscle cells (SMCs) and in smooth muscle-rich tissues. Furthermore, TGFB1I1 expression is significantly down-regulated in a variety of models for smooth muscle phenotypic modulation. The TGFB1I1 promoter contains an evolutionarily conserved CArG element, and this element is indispensible for myocardin-induced transactivation of TGFB1I1 promoter. By oligonucleotide pulldown and chromatin immunoprecipitation assays, we found that SRF binds to this CArG element in vitro and in vivo. Ectopic expression of myocardin is sufficient to induce endogenous TGFB1I1 expression in multiple cell lines whereas knocking-down myocardin or SRF significantly attenuated TGFB1I1 expression in SMCs. Furthermore, our data demonstrated that SRF is essential for TGF-β-mediated induction of TGFB1I1. Finally, silencing of TGFB1I1 expression significantly promotes SMC proliferation. Collectively, this study provides the first evidence that TGFB1I1 is not only an SRF/myocardin-regulated smooth muscle marker but also critical for maintaining smooth muscle contractile phenotype by inhibiting smooth muscle proliferation.

Introduction

Smooth muscle cells (SMCs)² are the major contractile component of most hollow organs, such as esophagus, colon, bladder, and blood vessels. Unlike striated cardiac and skeletal muscles that are terminally differentiated, under many pathological conditions SMCs maintain the ability to modulate their phenotype from contractile to synthetic in response to environmental cues or growth factors (1).

Transforming growth factor-β (TGF-β) is a potent regulatory growth factor with diverse functions on vascular SMCs (2). It has been shown to play important roles in vascular development and in pathogenesis of vascular diseases, including atherosclerosis and restenosis (2). TGFB1I1 was originally identified as a TGF-β-regulated gene in mouse osteoblasts (3, 4). TGFB1I1 is a focal adhesion protein with four terminal LIM (Lin11, lsl, and Mec-3) domains that share extensive homology with paxillin, a well studied focal adhesion protein (5). TGFB1I1 can translocate directly from focal adhesions to actin stress fibers upon mechanical stress thereby regulating the contractile capability of a cell (6). In addition, TGFB1I1 can shuttle between the nucleus and cytoplasm through an oxidant-sensitive nuclear export sequence (7). Furthermore, TGFB1I1 can function as a transcription cofactor in conjunction with nuclear receptors (8, 9) or with a number of other transcription factors including LEF/TCF, Sp1, Smad3, and Smad7 to modulate gene expression (10–13). During vascular injury TGFB1I1 expression is decreased, and overexpression of exogenous TGFB1I1 using adenoviral delivery attenuates injury-induced neointimal expansion in vivo (14). Consistent with this, neointimal formation was enhanced after wire injury of the femoral artery in TGFB1I1-deficient mice (15), suggesting that TGFB1I1 plays an important role in the progression of vascular diseases. Additionally, TGFB1I1 also plays a critical role in myofibroblast persistence during wound healing (16, 17). Because TGFB1I1 plays important roles in cell senescence, tumorigenesis, wound healing, vascular injury, and fetal gene expression during cardiac hypertrophy (4, 15, 17–19), we sought to determine the mechanisms by which TGFB1I1 is regulated under physiological and pathological conditions.

Serum response factor (SRF) plays a critical role in regulating expression of many smooth muscle-specific genes. SRF is an evolutionarily conserved MADS (MDM1, agamous, deficiens, SRF) box family transcription factor that binds to a highly conserved cis-regulatory element referred to as the CArG (CC(A/T)₆GG) box (20). At least one or more conserved CArG elements can be found in the majority of SMC-restricted genes (20). SRF is a weak activator of gene expression, requiring physical interactions with other cofactors to fully direct smooth muscle gene transcription. Of the SRF associated proteins, myocardin has been identified as the most potent activator of CArG-dependent SMC marker genes (21). In this report, we provide the first evidence that TGFB1I1 is an SRF/myocardin-regulated marker of the smooth muscle contractile phenotype, and the high level of TGFB1I1 expression is critical for maintaining the smooth muscle contractile phenotype.

EXPERIMENTAL PROCEDURES

Cell Culture and PDGF-BB, TGF-β Treatment

Mouse 10T1/2 fibroblast cells, green monkey kidney fibroblast COS-7 cells, HEK293, rat A10, and A7r5 aortic SMCs were purchased from ATCC and cultured in growth medium (DMEM containing 10% FBS and antibiotics). PAC1 SMCs were a gift of Dr. Li Li (Wayne State University). HUVECs, C6 glioma, and MCF-7 cells were a gift from Dr. Peter A. Vincent, Dr. Yunfei Huang, and Dr. Chunhong Yan, respectively (Albany Medical College). Rat primary aortic SMCs were prepared by enzymatic dispersion from descending aorta of adult Sprague-Dawley rats as described previously (22). For PDGF-BB treatment experiments, rat primary SMCs were grown to 80–90% confluence and serum-starved for 2 days and then treated with recombinant rat PDGF-BB (50 ng/ml; Calbiochem) for 48 h. Cells treated with vehicle served as control. For TGF-β treatment, 10T1/2 cells were grown to 80–90% confluence and serum-starved for 2 days and then treated with 10 ng/ml recombinant human TGF-β (BD Biosciences) overnight. For the smooth muscle calcification model, primary rat SMCs were treated every 3 days with differentiation medium (50 ng/ml human recombinant BMP-2 (Calbiochem), 10 mm β-glycerol phosphate (Sigma), and 100 μg/ml ascorbic acid (Sigma)). Cells that were cultured in DMEM containing 10% FBS growth medium served as control. After 10 days, the cultured cells were rinsed with PBS, and protein or RNA was harvested for measuring gene expression by Western blotting or real-time PCR.

Aortic Tissue Preparation and Aortic Organ Culture

Rat aorta was dissected to remove adhering tissue and denuded with a catheter. The aortic tissue was then cut into 2-mm cylindrical segments and cultured in 10% FBS DMEM for 60 h at 37 °C in a humidified chamber (5% CO₂). The fresh isolated tissues or culture vessels were then harvested with TRIzol for total RNA to evaluate gene expression by real-time PCR.

Mouse Carotid Artery Ligation Model

All animal studies were approved by the Institutional Animal Care and Use Committee. Carotid artery ligation was performed on 3-month-old C57BL/6 mice, and gene expression was analyzed as described previously (23).

Western Blotting

Western blot analysis was carried out as described previously (22–24). Total protein (30 μg) was fractionated on 5 or 15% SDS-polyacrylamide gels and transferred electrophoretically to a nitrocellulose membrane. The membrane was then probed with a series of antibodies: β-actin (AC-74, 1:2000; Sigma), calponin (1:5000; Sigma), GAPDH (1:2000; Santa Cruz Biotechnology), TGFB1I1 (1:5000; BD Biosciences), myocardin (M16, 1:2000; Santa Cruz Biotechnology), NMHCII-B (1:2000; Covance), paxillin (1:5000; BD Biosciences), proliferating cell nuclear antigen (1:500; Santa Cruz Biotechnology), SM α-actin (1:10,000; Sigma), SM22α (1:5000; Sigma), SRF (1:5000; Santa Cruz Biotechnology), and vinculin (1:5000; Sigma).

Immunohistochemistry

To detect TGFB1I1 expression, tissues from Sprague-Dawley rats were fixed in 4% paraformaldehyde and embedded in paraffin for all histological assays. Sections were then stained with TGFB1I1 (1:200; BD Biosciences) or SM α-actin (1:1200; Sigma) antibody. For detection of primary antibodies, we used avidin-biotin method with diaminobenzidine substrate as chromogen (brown; Vector Laboratories). Sagittal sections from control wild-type or SRF cardiovasculature-specific knock-out E11.5 embryos were prepared as described previously (25) and stained with the M.O.M. immunodetection kit (Vector Laboratories). Sections were lightly counterstained using hematoxylin to visualize cell nuclei (blue). Control sections were processed identically as experimental samples except no primary antibody was applied. In some experiments, hematoxylin and eosin staining was performed for general morphology by a standard protocol.

Immunocytochemistry

Mouse 10T1/2 cells were grown on coverslips and infected with GFP control or myocardin/GFP adenovirus for 48 h and then fixed, permeabilized, and incubated with monoclonal anti-TGFB1I1 antibody (1:300; BD Biosciences), followed by incubation with rhodamine-conjugated anti-mouse IgG (1:400) secondary antibodies (Jackson ImmunoResearch Laboratories). Cells were counterstained with Hoechst (1:5000) to visualize nuclei.

Quantitative Real-time RT-PCR (qRT-PCR) Analysis and MicroRNA (miR) qPCR

Total RNA was isolated with TRIzol reagent, and qRT-PCR was performed with respective gene-specific primers as we reported previously (22–24) or otherwise listed in supplemental Table 1. All samples were amplified in duplicate, and every experiment was repeated twice independently. Relative gene expression was converted using the 2^ΔΔCt method against the internal control acidic ribosomal phosphoprotein P0 housekeeping gene. For measuring miR-145 expression, total RNA was isolated with TRIzol, and then small noncoding RNAs were converted into quantifiable cDNAs with a QuantiMir RT kit (System Biosciences). MiR-145 was then evaluated as described in our recent report (22).

Adenoviral Construction and Cell Infection

Adenovirus encoding TGFB1I1 was generated as reported previously (16). Adenovirus encoding the full-length cardiac form of myocardin was generated, and cell infection was performed as described previously (22).

Luciferase Reporter Assays

By using mouse BAC (RP23, 144N15; Invitrogen) DNA as template, a 2.5-kb fragment spanning TGFB1I1 proximal promoter was amplified by PCR with primers harboring KpnI and XhoI restriction enzyme sites (primer sequences are listed in supplemental Table 1). The PCR products were then cloned into pGL2-Basic (Promega) luciferase reporter vector. Mutation of the CArG box in the TGFB1I1 promoter was carried out with QuikChange Site-directed Mutagenesis kit (Stratagene). The mouse TGFB1I1 promoter-luciferase reporter included nucleotides −1630 to +930 of the mouse TGFB1I1 gene relative to the transcription start site as determined by DBTSS. A conserved CArG box within the TGFB1I1 gene promoter was identified by a sequence alignment among multiple vertebrate species. All plasmids were sequenced to verify the integrity of the insert. Mammalian expression plasmids for myocardin, MRTF-A and MRTF-B, were described in our previous report (24). Transfection was carried out with FuGENE6 transfection reagent (Roche Applied Science) as described previously (24). The level of promoter activity was evaluated by measurement of the firefly luciferase activity relative to the internal control TK-Renilla luciferase activity using the Dual Luciferase Assay System as described by the manufacturer (Promega). A minimum of six independent transfections was performed, and all assays were replicated at least twice. Results are reported as the mean ± S.E.

siRNA Transfection

Control siRNA and siRNA SMARTpool against rat myocardin and SRF were purchased from Dharmacon. PAC1 cells were transfected either with control siRNA or siRNAs against myocardin or SRF using Lipofectamine 2000 transfection reagent (Invitrogen) following the manufacturer's protocol. After 48 h, total RNA or protein was harvested for qRT-PCR or Western blot analysis. For silencing TGFB1I1 in rat primary aortic SMCs, siRNA duplex against TGFB1I1 purchased from Ambion and transfection was performed using the Neon transfection system (Invitrogen) essentially following the manufacturer's protocol.

Oligonucleotide Pulldown Assays

Bacterially expressed SRF or myocardin was harvested in binding buffer (20 mm HEPES (pH 7.9), 80 mm KCl, 1 mm MgCl₂, 0.2 mm EDTA (pH 8.0), 10% glycerol, 1 mm DTT, 0.1% Triton X-100, 1% protease inhibitor mixture) and incubated with biotinylated double-stranded oligonucleotides containing the TGFB1I1 CArG box at 4 °C overnight. Then, 50 μl of streptavidin-agarose beads were added for 1 h at 4 °C to precipitate the oligonucleotide-protein complex. The beads were collected by centrifuging and washed with binding buffer three times. Subsequently, the precipitated protein was eluted from the beads by adding SDS sample buffer, and SRF or myocardin protein was detected by Western blotting. For competition assays, 100× excess nonbiotinylated oligonucleotides were used. The wild-type and mutant oligonucleotide sequences are shown in supplemental Table 1.

Quantitative Chromatin Immunoprecipitation (ChIP) Assays

PAC1 SMCs were fixed with formaldehyde, and ChIP was performed by using an anti-SRF antibody (G20X; Santa Cruz Biotechnology) or IgG control as described by the manufacturer (Upstate) and in our previous report (24). Primers for quantitative evaluation of enrichment of the telokin promoter CArG region, TGFB1I1 gene CArG region, and exon 10 are listed in supplemental Table 1 or described in our recent report (24).

SMC Proliferation Assay

The proliferation of rat primary aortic SMCs were measured by a cell proliferation WST-1 kit (Roche Applied Science) as described in our previous report (24). For the cell number counting, following transfection of the TGFB1I1-silencing duplex, rat primary aortic SMCs were serum-starved in DMEM for 48 h and then cultured in 10% FBS medium. Cell numbers were counted manually at the indicated time points.

RESULTS

TGFB1I1 Expression Is Restricted to SMCs

TGFB1I1 is a focal adhesion protein that is homologous to paxillin and was originally identified as a gene induced by TGF-β and hydrogen peroxide (4). By immunohistochemistry, TGFB1I1 has been found to be highly expressed in both mouse and human smooth muscle tissues (6, 14, 26). To extend these studies, protein lysates were harvested from SMC lines (A10 and A7r5 are SMC lines that are derived from the rat thoracic aorta (27), PAC1 cells are established from the rat pulmonary artery (28)) and non-SMC lines (HUVEC, COS-7, HEK293, 10T1/2, MCF-7, C6 glioma), and Western blotting was carried out to detect TGFB1I1 and paxillin expression (Fig. 1A). Unlike paxillin, TGFB1I1 resembles the expression pattern of other well recognized smooth muscle-specific genes, calponin and SM α-actin, in that TGFB1I1 is expressed exclusively in SMC lines (Fig. 1A). We also found that TGFB1I1 was expressed abundantly in all SMC-rich tissues including the aorta, bladder, colon, esophagus, lung, spleen, stomach, and trachea of the adult mouse (Fig. 1B). To characterize the TGFB1I1 expression pattern further, rat tissues were harvested, and immunohistochemistry staining was carried out with a TGFB1I1 antibody. Data from these experiments revealed that TGFB1I1 expression resembles SM α-actin tissue distribution that is present exclusively in smooth muscles but not in cardiac and skeletal muscles (Fig. 1, C and D, and supplemental Fig. 1). The weak TGFB1I1 expression detected in heart, kidney, and liver by Western blotting (Fig. 1B) is due to abundant TGFB1I1 expression in the vasculature of these tissues, which was confirmed by immunohistochemistry staining of TGFB1I1 (Fig. 1C and supplemental Fig. 1).

FIGURE 1.

TGFB1I1 expression is restricted to SMCs. A, TGFB1I1 expression is restricted to SMC lines. Protein lysates were harvested from multiple cell lines as indicated, and Western blot analyses were performed to detect expression of TGFB1I1, the TGFB1I1 homolog paxillin, as well as smooth muscle-specific genes calponin and SM α-actin. β-Actin served as loading control. B, TGFB1I1 expression is restricted in SMC-rich tissues in mouse. Mouse tissues were harvested for Western blotting and expression of TGFB1I1, and smooth muscle-specific genes SM α-actin and SM22α were detected. SK, skeletal. C, TGFB1I1 expression is restricted in SMCs in rat. Serial sections were prepared from adult rat heart, and immunohistochemistry staining was used to detect the expression of TGFB1I1 or SM α-actin as indicated. A section stained with the secondary antibody alone served as a negative control. TGFB1I1, similar to SM α-actin, is exclusively expressed in the arterial smooth muscle layer, as indicated by an arrow, but not in cardiomyocytes. Original magnification, ×20. D, immunohistochemistry staining was performed in contiguous sections from the rat upper esophagus using antibody against TGFB1I1 or SM α-actin, as indicated. TGFB1I1 expression is localized specifically in smooth muscles of perivasculature and muscularis mucosae as indicated by a thin or heavy arrow, respectively. Epithelium (E) and skeletal muscles (M) have no TGFB1I1 expression. L, lumen. Original magnification, ×20.

TGFB1I1 Expression Is Down-regulated during Phenotypic Modulation of Smooth Muscle

Previous studies have suggested that TGFB1I1 is down-regulated following vascular injury (14). To examine the expression of TGFB1I1 during smooth muscle phenotypic modulation further, we examined TGFB1I1 expression in a variety of models in which SMCs are induced to modulate their phenotype toward a synthetic state. First, we found that TGFB1I1 expression is dramatically attenuated following passage of primary rat aortic SMCs in culture (P2 and P6), whereas the smooth muscle “synthetic” marker NMHCII-B is induced (29) (Fig. 2A). In contrast, expression of paxillin did not change under these conditions (Fig. 2A). PDGF-BB is known as a potent growth factor that inhibits smooth muscle marker expression while promoting SMC proliferation (30). In response to PDGF-BB treatment, TGFB1I1 expression was significantly down-regulated along with other smooth muscle markers, including calponin, SM22α, and the cardiovascular-specific transcription factor myocardin, at the protein and/or mRNA level, whereas the marker proliferating cell nuclear antigen is significantly induced in rat primary aortic SMCs (Fig. 2B). Aortic organ culture has been widely used as an ex vivo model in which SMCs undergo profound phenotypic modulation with down-regulation of smooth muscle markers and induction of matrix gene expression (31). In this model, expression of TGFB1I1, SM22α, and myocardin mRNAs were dramatically attenuated, whereas the matrix gene versican was significantly increased (6-fold; Fig. 2C). MiR-145, a miR that is considered a novel marker for the smooth muscle contractile phenotype (32), was also significantly down-regulated during organ culture (Fig. 2C). Smooth muscle marker gene expression has previously been shown to be attenuated when SMCs are cultured in bone differentiation medium to induce SMC calcification (33). Similarly, we found that TGFB1I1 expression is also attenuated at both mRNA and protein levels under these conditions (Fig. 2D). We further confirmed that TGFB1I1 expression is down-regulated along with other smooth muscle contractile genes in vivo in a mouse carotid artery ligation model (34). In this arterial injury model, Western blot and real-time PCR data revealed that expression of TGFB1I1, similar to another smooth muscle marker SM α-actin, was significantly down-regulated in injured blood vessels from 1 to 14 days following injury (Fig. 2E). Taken together, these data demonstrate that expression of TGFB1I1 is attenuated in tandem with other smooth muscle markers following phenotypic modulation of vascular SMCs.

FIGURE 2.

Down-regulation of TGFB1I1 expression during smooth muscle phenotypic modulation. A, TGFB1I1 expression is down-regulated during SMC culture. Aortic arteries were dissected from rat and either harvested directly or cultured to passage 2 or 6 following enzymatic digestion. Subsequently, the protein extracts of tissue or cultured SMCs were analyzed by Western blotting as indicated. B, PDGF-BB treatment inhibits TGFB1I1 expression in SMCs. Rat primary aortic SMCs were treated with 50 ng/ml PDGF-BB or vehicle for 48 h. Cells were then harvested, and Western blotting was performed to detect smooth muscle genes and TGFB1I1 expression as indicated. Vinculin served as loading control. Cells treated with PDGF-BB were also harvested for total RNA, and real-time PCR was performed to measure TGFB1I1, SM22α, and myocardin mRNA expression (right). The relative expression level of genes in vehicle-treated cells served as control and was normalized to a value of 1. *, p ≤ 0.05. C, expression of TGFB1I1 is attenuated during arterial organ culture. Rat aortic artery was dissected and either harvested immediately for total RNA or cultured in 10% FBS DMEM for 3 days and then harvested for total RNA. The changes in gene or miR expression were measured by qRT-PCR. The relative expression of genes in intact tissue was set to 1. D, down-regulation of TGFB1I1 expression during SMC calcification is shown. Rat aortic primary SMCs were cultured in growth medium (GM) or differentiation medium (DM) to induce SMC calcification for 10 days. Protein from these cells was then harvested for Western blotting (left), or total RNA was extracted for qRT-PCR (right) to measure TGFB1I1, smooth muscle genes calponin and SM22α, and miR-145 expression. E, left, expression of TGFB1I1 is down-regulated in mouse carotid artery following ligation injury. Carotid artery ligation was performed in 12-week-old C57BL/6 mice. Protein was harvested from uninjured control (con) or injured carotid artery (lig) at postsurgery day 1, 3, 7, and 14, and Western blotting was carried out to assess TGFB1I1 expression as indicated. Right, total RNA from 5 or 10 days postoperation injured and contralateral arterial tissues were harvested for qRT-PCR to measure TGFB1I1 and other smooth muscle gene expression. The relative gene expression in uninjured control vessels was normalized to a value of 1. n = 4.

TGFB1I1 Gene Promoter Contains a Functional CArG Element Responsive to SRF/Myocardin

Previous studies demonstrated that the complex of SRF/myocardin is a critical determinant for smooth muscle-specific gene expression (35). Data described above revealed that TGFB1I1 has a restricted pattern of expression in SMCs and TGFB1I1 expression is down-regulated in a manner similar to other known smooth muscle markers during smooth muscle phenotypic switch. This raised the possibility that TGFB1I1 transcription is under the control of an SRF/myocardin transcriptional complex. To begin exploring this possibility, we carefully analyzed the TGFB1I1 gene proximal promoter. In agreement with a previous report (36) a putative SRF-binding element (CArG) was identified. This CArG box is evolutionarily conserved among the species of vertebrates and harbors a C substitution in the A/T-rich core (Fig. 3A). To validate experimentally that TGFB1I1 is a myocardin or myocardin family protein target gene, transient transfection experiments were performed using the luciferase reporters containing TGFB1I1 promoters with either wild-type or mutated CArG sequences. These assays revealed that luciferase activity of the wild-type TGFB1I1 promoter reporter was significantly activated (7–9-fold) in the presence of myocardin or its family member MRTF-A. In contrast, mutating the CArG box in the TGFB1I1 promoter completely abolished myocardin or MRTF-A-induced transactivation (Fig. 3B). Consistent with previous studies showing that MRTF-B is a weak transcription coactivator for smooth muscle genes (37), MRTF-B had a moderate effect on the wild-type TGFB1I1 promoter reporter gene (Fig. 3B). To confirm that TGFB1I1 promoter activity is SRF/myocardin-dependent in SMCs, SRF or myocardin expression was reduced in PAC1 SMCs using siRNA, and the subsequent effects on TGFB1I1 promoter activity were determined. As shown in Fig. 3C, the activity of the wild-type TGFB1I1 promoter but not the mutated CArG TGFB1I1 promoter was significantly reduced (40–50% of control levels) in PAC1 SMCs following knockdown of SRF or myocardin, resembling basal CArG mutant levels. Furthermore, CArG mutation in the TGFB1I1 promoter resulted in a 50% reduction in basal activity of TGFB1I1 promoter in PAC1 SMCs, suggesting the intact CArG box in the TGFB1I1 promoter is critical for TGFB1I1 promoter activity in SMCs (Fig. 3C).

FIGURE 3.

The TGFB1I1 gene promoter contains a functional CArG element responsive to SRF/myocardin. A, schematic diagrams a luciferase reporter of the mouse TGFB1I1 promoter (−1630 to +930) containing an evolutionarily conserved CArG box among vertebrate species. The position of CArG element is relative to the annotated transcription start site (+1). A mutation placed in the CArG box is shown at the bottom. B, mouse myocardin and myocardin family proteins MRTF-A and B, or pcDNA empty expression plasmid, were co-transfected together with a TGFB1I1 gene promoter luciferase reporter containing either wild-type or mutated CArG box, into 10T1/2 cells. The level of promoter activity was determined by measuring the firefly luciferase activity relative to the control Renilla luciferase. Fold activation of promoter activity relative to control vector transfections (normalized to 1) is presented as mean ± S.E. (error bars). n = 6. C, PAC1 SMCs were transfected with either 100 pmol of an RNA duplex against myocardin (siMyocardin), SRF (siSRF), or scrambled control RNA duplex (siControl). After 12 h they were then transfected with wild-type or CarG-mutated TGFB1I1 promoter luciferase reporter construct. 24 h following plasmid transfection cells were harvested, and promoter activity was measured by dual luciferase assay as described in B. Reporter activity is normalized to a Renilla luciferase internal control and expressed relative to siRNA control transfections (normalized to 1). Knocking down endogenous SRF or myocardin in SMCs decreases wild-type TGFB1I1 promoter activity but not the activity of the CArG mutation reporter.

SRF/Myocardin Binds to the TGFB1I1 Gene Promoter in Vitro and in Vivo

The data described above demonstrate that the TGFB1I1 promoter contains a functional CArG box responsive to SRF/myocardin. We next sought to determine whether SRF/myocardin can directly bind to the TGFB1I1 promoter CArG element. Oligonucleotide pulldown assays showed that SRF binds directly to wild-type but not to the mutated TGFB1I1 promoter CArG box. Myocardin alone did not bind to the probe, but it could be precipitated in the presence of SRF, indicating that myocardin binds indirectly to the CArG box through SRF (Fig. 4A). The specificity of SRF binding to the CArG box was demonstrated by a competition assay in which excess unlabeled wild-type CArG oligonucleotides but not mutated oligonucleotides could compete for SRF binding to the biotinylated probe (Fig. 4A). To confirm SRF binding to the TGFB1I1 promoter CArG box in vivo further, ChIP assays were performed in PAC1 SMCs. In agreement with the oligonucleotide pulldown data, ChIP assays unveiled approximately 2-fold enrichment of DNA fragment containing the TGFB1I1 CArG box region following immunoprecipitation with a SRF antibody, but not TGFB1I1-coding region exon 10 (Fig. 4B). As a positive control, ChIP assays showed an approximately 4-fold enrichment of DNA containing telokin CArG box region in PAC1 SMCs. Taken together, these data suggest that SRF binds to the TGFB1I1 promoter CArG box in vitro and in cells.

FIGURE 4.

SRF/myocardin binds to TGFB1I1 gene promoter in vitro and in vivo. A, SRF/myocardin binds to the CArG box within the TGFB1I1 promoter in vitro. Bacterially expressed SRF or myocardin was incubated with biotinylated double-stranded oligonucleotides harboring TGFB1I1 promoter CArG box, and then streptavidin-agarose beads were added to precipitate the oligonucleotide-protein complex. Subsequently, the beads were collected, and protein was detected by Western blotting as indicated. For competition assays, 100 × excess nonbiotinylated wild-type or mutated CArG oligonucleotides were added. B, SRF specifically binds to the CArG box region of TGFB1I1 promoter in vivo. Cross-linked chromatin from PAC1 SMCs was immunoprecipitated with anti-SRF antibody or control IgG, and the precipitated DNA was amplified by real-time PCR with TGFB1I1 gene-specific primers spanning CArG box region or exon (E) 10. A known CArG region within the smooth muscle-specific gene telokin served as a positive control. The enrichment of SRF binding is indicated relative to IgG control (normalized to 1).

Myocardin Induces TGFB1I1 Expression

Data described above demonstrate that myocardin can significantly activate the TGFB1I1 promoter. To test whether myocardin is sufficient to induce TGFB1I1 expression, adenovirus encoding myocardin or GFP control were transduced into rat primary vascular SMCs and non-SMC lines, including HUVECs and COS-7 and 10T1/2 cells, and the expression of TGFB1I1 protein and mRNA was analyzed by Western blotting and qRT-PCR, respectively. Data from these experiments revealed that forced expression of myocardin dramatically induces endogenous TGFB1I1 protein and mRNA expression in both non-SMCs (Fig. 5, A–C and E) and SMCs (Fig. 5, D and E). Furthermore, immunofluorescence staining revealed that forced expression of myocardin dramatically induces endogenous TGFB1I1 expression at focal adhesions in 10T1/2 cells (Fig. 5F). Overexpression of myocardin in HUVECs had no effect on endogenous TGFB1I1 family protein paxillin expression (Fig. 5A).

FIGURE 5.

Overexpression of myocardin induces endogenous TGFB1I1 expression in multiple cell lines. Adenovirus encoding myocardin or control GFP virus was transduced into HUVECs (A), COS-7 cells (B), 10T1/2 cells (C), and rat primary aortic vascular SMCs (D) for 48 h, and subsequently protein was harvested for Western blotting to detect TGFB1I1 gene expression at protein level, or total RNA was extracted for qRT-PCR to detect TGFB1I1 expression at mRNA level as indicated (E). C, *, nonspecific signal with anti-myocardin antibody. F, 10T1/2 cells were plated on coverslips and transduced with the GFP control or GFP/myocardin adenovirus for 48 h and then stained with monoclonal anti-TGFB1I1 antibody to detect endogenous TGFB1I1 expression (red). Immunofluorescence data are shown with a representative picture. Overexpression of myocardin in 10T1/2 cells results in a dramatic induction of TGFB1I1 localized in focal adhesions.

Expression of TGFB1I1 Is Dependent on SRF/Myocardin

To determine the role of endogenous SRF and myocardin in regulating TGFB1I1 expression, SRF or myocardin levels were decreased by siRNA. Knocking down endogenous myocardin or SRF with SMARTpool siRNA duplexes (Dharmacon) in A7r5 and PAC1 SMCs resulted in an approximate 50% reduction of basal TGFB1I1 mRNA and protein expression (Fig. 6, A–G). To determine further whether SRF is required for TGFB1I1 expression in vivo, we detected TGFB1I1 expression in SRF cardiovasculature-specific knock-out mouse embryos (25) by immunohistochemistry staining. Data from this experiment revealed that TGFB1I1 expression is reduced dramatically in the dorsal aorta of SRF knock-out embryo compared with control wild-type counterpart (Fig. 6H). Taken together, these results demonstrate that SRF and myocardin are necessary for TGFB1I1 expression.

FIGURE 6.

Silencing SRF or myocardin down-regulates endogenous TGFB1I1 expression in SMCs. A–D, A7r5 SMCs were transfected with silencing RNA duplexes against SRF (A and B) or myocardin (C and D) for 48 h, and the cells were harvested for Western blotting (A and C) or qRT-PCR (B and D) to detect TGFB1I1 expression. The cells transfected with scrambled silencing RNA duplex served as control. Blots in A and C are representative of three independent experiments. Vinculin served as loading control. E and F, PAC1 cells were transfected with silencing myocardin or SRF duplex, and cells were harvested for measuring TGFB1I1 expression at protein (E) and mRNA level (F) as described in A–D. GAPDH served as loading control. G, densitometric quantification of TGFB1I1 expression in A, C, and E is shown. TGFB1I1 signals were normalized to vinculin or GAPDH signal to account for variability in loading and expressed as relative to silencing control samples (normalized to 1). *, p ≤ 0.05. Error bars, S.E. H, sagittal sections from control wild-type (WT) or SRF cardiovasculature-specific knock-out mouse E11.5 embryos (SRF KO) were stained with TGFB1I1 antibody (brown). Staining with secondary antibody alone served as negative control. All sections were lightly counterstained with hematoxylin to visualize cell nuclei (blue). SRF deficiency resulted in a significant reduction of TGFB1I1 staining in dorsal aorta (DA) as indicated by an arrow. *, heart.

SRF Is Required for TGF-β-induced TGFB1I1 Expression

Previous studies have shown that TGF-β is a potent cytokine that induces TGFB1I1 expression (4, 17). In response to TGF-β treatment, 10T1/2 fibroblast cells differentiate toward a smooth muscle- or myofibroblast-type fate with induction of an array of smooth muscle-specific genes (38). To assess the role of SRF in TGF-β-mediated induction of TGFB1I1, 10T1/2 cells were transfected with SRF siRNA overnight and then treated with TGF-β. After 48 h, cells were harvested, and Western blotting was performed to evaluate TGFB1I1 expression. Data from this experiment revealed that depletion of SRF significantly diminished TGF-β-induced TGFB1I1 expression to approximately 50% in a manner similar to other smooth muscle genes, including SM22α and SM α-actin (Fig. 7, A and B). Consistent with the previous study (39), we also observed that TGF-β treatment increased endogenous SRF expression (Fig. 7A). Furthermore, luciferase reporter assays demonstrated that TGF-β significantly increases the WT TGFB1I1 promoter activity 2-fold. In contrast, a TGFB1I1 promoter in which the CArG box was mutated not only exhibited a 50% reduction in basal activity but also completely lost the ability to be activated by TGF-β (Fig. 7C). These data demonstrated that SRF and an intact CArG box in the TGFB1I1 gene promoter are required for the TGF-β-mediated induction of TGFB1I1.

FIGURE 7.

SRF is required for TGF-β-induced TGFB1I1 expression. A, 10T1/2 fibroblast cells were transfected with scrambled control silencing RNA duplex or RNA duplex against SRF for 12 h, and the cells were then treated with 10 ng/ml TGF-β for 48 h. Subsequently, protein lysates were harvested for Western blotting to detect TGFB1I1 expression, as indicated. B, densitometric quantification of immuno bands of TGFB1I1 and other smooth muscle markers in A is shown. The level of protein expressed was determined by measurement of the protein absorbance units relative to the respective vinculin expression as an internal control. A value of 1 was arbitrarily assigned to the protein expression level in cells transfected with siRNA control and treated with TGF-β. *, p ≤ 0.05. C, 10T1/2 cells were transfected with wild-type or mutated CArG TGFB1I1 promoter reporter for 12 h and then treated with TGF-β (10 ng/ml) or vehicle control for 24 h. Subsequently, cells were harvested for dual luciferase assay to measure TGFB1I1 promoter activity in the presence or absence of TGF-β. The luciferase activity of the wild-type TGFB1I1 promoter treated with vehicle served as control (normalized to 1). Data were presented as mean ± S.E. (error bars). n = 6.

Depletion of TGFB1I1 Expression Promotes SMC Proliferation

Data described above demonstrated that TGFB1I1 is a novel marker for smooth muscle contractile phenotype. We next sought to determine TGFB1I1 function in vascular SMCs. To test whether TGFB1I1 is able to affect SMC proliferation, a silencing duplex against TGFB1I1 was transfected into rat primary aortic SMCs by electroporation. This transfection resulted in a significant decrease of endogenous TGFB1I1 protein expression (Fig. 8A), and SMC proliferation in all culture media examined was significantly promoted compared with the control cells transduced with scrambled control siRNA (Fig. 8B). Furthermore, silencing endogenous TGFB1I1 expression significantly increased rat primary aortic SMC growth (Fig. 8C). Taken together, these data suggest that expression of TGFB1I1 is critical for maintaining the SMC contractile phenotype by inhibiting SMC proliferation. Overexpression or silencing TGFB1I1 in rat primary aortic SMCs does not promote or attenuate smooth muscle differentiation (supplemental Fig. 2).

FIGURE 8.

Depletion of TGFB1I1 expression promotes SMC proliferation. A, rat primary aortic SMCs were transfected with control or TGFB1I1 silencing RNA duplex, and protein was then harvested for Western blotting to assess TGFB1I1 expression. B, following transfection with silencing RNA duplex, rat primary aortic SMCs were plated at equal density either in DMEM with supplement of 50 ng/ml PDGF-BB, 10% FBS medium, or 10% FBS supplemented with 50 ng/ml PDGF-BB, and then the proliferation of SMCs was measured using the cell proliferation WST-1 kit (Roche Applied Science). *, p < 0.05. C, following silencing TGFB1I1 rat aortic primary SMCs were seeded at equal density in 10% FBS medium, and cell numbers were counted at each time point as indicated. *, p < 0.05. Error bars, S.E.

DISCUSSION

In this study we demonstrated that expression of TGFB1I1 is largely restricted to smooth muscle-enriched tissues (Fig. 1). TGFB1I1 expression is significantly down-regulated during smooth muscle phenotypic modulation in a manner parallel to other well documented smooth muscle marker genes (Fig. 2). These findings suggested that TGFB1I1 is a marker for differentiated contractile SMCs. Furthermore, consistent with our previous finding that TGFB1I1 regulates myofibroblast proliferation (16), silencing TGFB1I1 in SMCs significantly increases SMC proliferation (Fig. 8), suggesting a novel role of TGFB1I1 in maintaining the smooth muscle contractile phenotype.

TGFB1I1 was originally described as a marker for the early developing heart in mouse (40). At E9.5, TGFB1I1 is detected in the atria and ventricles of developing heart and in the somites (40). Later, TGFB1I1 expression is diminished in the heart and by E16.5 becomes restricted in smooth muscle layers of multiple organs, including bladder and intestine (40). Consistent with this finding, we found that the expression of TGFB1I1 is enriched predominantly in the medial smooth muscle layer of arterial tissues and in other smooth muscle-rich tissues, whereas it is barely detectable in postnatal rat heart and skeletal muscles (Fig. 1 and supplemental Fig. 1). Interestingly, the spatial and temporal regulation of TGFB1I1 expression during mouse development resembles that of several other SMC-specific genes including SM α-actin and SM22α that display transient expression and promoter activity in the developing myocardium (1). All of these genes contain conserved SRF-binding CArG elements within their promoter regions or adjacent introns (20). A previous study suggested that TGFB1I1 is a SRF target gene (36), and TGFB1I1 expression was reduced in a SRF-null mouse embryonic heart (41). In this study we extended these findings and identified an evolutionarily conserved CArG box in the TGFB1I1 gene promoter (Fig. 3). In agreement with these studies, TGFB1I1 expression is significantly attenuated in the dorsal aorta of SRF cardiovascular knock-out mouse embryos (Fig. 6H). We showed that this CArG box in TGFB1I1 gene promoter is vital for myocardin/SRF-dependent transactivation (Fig. 3) and for activation of the promoter by TGF-β (Fig. 7). We also confirmed that SRF binds to the TGFB1I1 promoter CArG sequence in vitro and in vivo (Fig. 4). Interestingly, this CArG box we identified within the TGFB1I1 gene promoter harbors a conserved C substitution in the central A/T-rich region (Fig. 3). Previous studies have shown that this type of CArG degeneracy diminishes binding affinity for ubiquitously expressed SRF (42). In agreement with this, quantitative ChIP assays demonstrated relatively weak binding of SRF to the TGFB1I1 CArG box compared with the canonical CArG box in the telokin promoter in SMCs (Fig. 4B). Previous studies demonstrated that this reduced SRF binding affinity is critical for controlling gene expression in response to pathophysiological stimuli such as vascular injury (42). Although the TGFB1I1 promoter reporter gene we cloned can be readily activated by myocardin and TGF-β treatment, further work is needed to test directly whether this fragment of the TGFB1I1 promoter can recapitulate the expression patterns of the endogenous gene in transgenic mice and whether the degenerated CArG box is required for changes in promoter activity following vascular injury.

The effects of TGF-β on smooth muscle marker gene expression are thought to be mediated by activation of the SMAD family of proteins (43–45). Previous studies have shown that TGF-β activates the SM22α promoter via either a CArG box-dependent manner by direct association of Smad3 with SRF or by a CArG box-independent pathway involving direct interaction of Smad-binding element-bound Smad3 and myocardin (45, 46). Although we found that TGF-β stimulation of TGFB1I1 requires SRF, and the intact CArG box in the TGFB1I1 promoter is similar to the CArG-dependent activation of the SM22α promoter, we cannot rule out the possibility that an alternative CArG-independent mechanism is involved in TGF-β-induced TGFB1I1 expression. In future studies it will be interesting to examine whether the TGFB1I1 promoter contains any putative Smad-binding element sites that are important for TGFB1I1 expression.

Unlike myocardin whose expression is restricted to cardiac and smooth muscle and localizes exclusively within nucleus, MRTF-A is broadly expressed, and the function of MRTF-A can be regulated by interactions with the actin cytoskeleton, resulting in subcellular distribution either in cytoplasm or nucleus (47). In response to Rho/ROCK signaling that can also be activated by TGF-β treatment, MRTF-A translocates into the nucleus where it binds to SRF, thereby activating a subset of CArG box-containing genes, including SM α-actin (48). Similar to SM α-actin, TGFB1I1 is also highly expressed in myofibroblasts and induced by TGF-β during wound healing (16, 17). Previous studies demonstrated that up-regulation of TGFB1I1 expression following TGF-β-induced epithelial-mesenchymal transition is both RhoA- and ROCK-dependent (49). In the current study we demonstrated that MRTF-A strongly activated a TGFB1I1 promoter reporter gene to an extent similar to myocardin. Consistently, in our accompanying study³ we found that forced expression of MRTF-A can dramatically induce endogenous TGFB1I1 expression in human fibroblast cells. However, myocardin is almost undetectable in myofibroblasts compared with abundant expression of MRTF-A. Consistent with this finding, silencing MRTF-A attenuated TGF-β-induced TGFB1I1 expression in fibroblasts. Taken together, these studies demonstrated that myocardin family proteins play important roles for TGFB1I1 expression through various mechanisms whereby using the distinct availability of myocardin family proteins in SMCs versus myofibroblasts.

In summary, this study identified TGFB1I1 as a novel marker for the contractile phenotype of vascular SMCs that is regulated by SRF/myocardin and the expression of TGFB1I1 is critical for maintaining smooth muscle contractile phenotype.