Induction of Tenascin-C in Cardiac Myocytes by Mechanical Deformation

Mechanical overload may change cardiac structure through angiotensin II-dependent and angiotensin II-independent mechanisms. We investigated the effects of mechanical strain on the gene expression of tenascin-C, a prominent extracellular molecule in actively remodeling tissues, in neonatal rat cardiac myocytes. Mechanical strain induced tenascin-C mRNA (3.9 ± 0.5-fold, p < 0.01, n = 13) and tenascin-C protein in an amplitude-dependent manner but did not induce secreted protein acidic and rich in cysteine nor fibronectin. RNase protection assay demonstrated that mechanical strain induced all three alternatively spliced isoforms of tenascin-C. An angiotensin II receptor type 1 antagonist inhibited mechanical induction of brain natriuretic peptide but not tenascin-C. Antioxidants such as N-acetyl-l-cysteine, catalase, and 1,2-dihydroxy-benzene-3,5-disulfonate significantly inhibited induction of tenascin-C. Truncated tenascin-C promoter-reporter assays using dominant negative mutants of IκBα and IκB kinase β and electrophoretic mobility shift assays indicated that mechanical strain increases tenascin-C gene transcription by activating nuclear factor-κB through reactive oxygen species. Our findings demonstrate that mechanical strain induces tenascin-C in cardiac myocytes through a nuclear factor-κB-dependent and angiotensin II-independent mechanism. These data also suggest that reactive oxygen species may participate in mechanically induced left ventricular remodeling.

Cardiac hypertrophy is an independent risk factor of cardiac morbidity and mortality (1) and is characterized by an increase in myocyte mass and volume, as well as an increase of extracellular matrix proteins such as collagen (2). Angiotensin II is a potent stimulator of cardiac hypertrophy (3), and angiotensin-converting enzyme inhibitors prevent left ventricular hypertrophy in hypertensive animals and humans. For example, Kojima et al. (4) reported that treatment with TCV-116, an angiotensin II receptor type 1 (AT 1 ) 1 antagonist, decreased left ventricular weight, left ventricular wall thickness, and the transverse diameter of cardiac myocytes in spontaneously hypertensive rats.
Recent studies (5,6) indicate that angiotensin II-independent mechanisms may also mediate cardiac hypertrophy. Harada et al. (5) demonstrated that acute pressure overload could induce hypertrophic responses such as induction of c-fos, c-jun, and brain natriuretic peptide (BNP) gene expression, mitogen-activated protein (MAP) kinase activation, and increased heart weight/body weight, in the hearts of AT 1A knockout mice. Harada et al. (6) also reported that there were no significant differences between wild-type mice and AT 1A knockout mice in expression levels of fetal-type cardiac genes, in left ventricular wall thickness and systolic function, or in histological changes such as myocyte hypertrophy and fibrosis.
Many intracellular signaling pathways are thought to play important roles in mechanotransduction. Recent studies of myocardial hypertrophy have focused on activation of protein kinases including protein kinase C, Raf-1 kinase, S6 peptide kinase, and MAP kinases, which precede an increase in specific gene expression and protein synthesis (7)(8)(9)(10)(11). The MAP kinase signaling pathways consist of three major phosphorylation cascades as follows: the extracellular signaling-regulated protein kinases, the c-Jun NH 2 -terminal kinases (JNK), and the p38 MAP kinases (12,13). JNK and p38 pathways are collectively termed stress-activated protein kinases because they are activated by various stress-related stimuli (14).
Intracellular reactive oxygen species (ROS) may participate in cellular responses to various stimuli including hemodynamic forces and act as signal transduction messengers. ROS in the heart may play a role in pathophysiological conditions such as myocardial ischemia, reperfusion, apoptosis, and heart failure (15). Mechanical stimuli can modulate intracellular ROS in endothelial cells (16 -18), vascular smooth muscle cells (19), and cardiac muscle (20). In addition, Kheradmand et al. (21) reported that ROS were essential for nuclear factor (NF)-Bdependent transcriptional regulation of collagenase-1 gene expression induced by cell shape change.
Ultimately, mechanotransduction events lead to increased cell size and cardiac remodeling. Several extracellular matrix * This work was supported in part by a Grant-in-Aid from the American Heart Association and by NHLBI Grant HL-54759 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.   1 The abbreviations used are: AT 1 , angiotensin II receptor type 1; NRVM, neonatal rat ventricular myocytes; BNP, brain natriuretic peptide; MAP, mitogen-activated protein; JNK, c-Jun NH 2 -terminal kinases; ROS, reactive oxygen species; NF, nuclear factor; HBSS, Hanks' balanced salt solution; DMEM, Dulbecco's modified Eagle's medium; FCS, fetal calf serum; SPARC, secreted protein acidic and rich in cysteine; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; CAT, chloramphenicol acetyltransferase; PDGF-BB, platelet-derived growth factor-BB; IKK, IB kinase. molecules, including tenascin-C and fibronectin, are induced in remodeling tissues. Tenascin-C is a disulfide-linked hexameric extracellular matrix protein with subunit molecular masses of 190 -300 kDa depending on the species and on alternative splicing within fibronectin type III repeats (22). Tenascin-C is able to interact with cell surface receptors including integrins (23)(24)(25) and also binds to extracellular matrix proteins such as fibronectin (26). Tenascin-C inhibits the attachment of fibroblasts and endothelial cells to various adhesive proteins (27). Thus, tenascin-C may be important in regulating cell-extracellular matrix interactions by promoting cell rounding, migration, and/or differentiation.
By using a mechanical deformation device that applies a highly uniform biaxial strain field over the culture substrate, we investigated the effects of mechanical strain on tenascin-C gene expression in cultured neonatal rat cardiac myocytes. We found that tenascin-C is mechanically induced in cardiac myocytes through the activation of NF-B, suggesting that tenascin-C can be an angiotensin II-independent early marker for cardiac remodeling. In addition, our data suggest that ROS may participate in mechanically induced cardiomyocyte responses. Culture of Neonatal Rat Ventricular Myocytes (NRVM)-NRVM from 1-day-old Harlan Sprague-Dawley rats were isolated by previously described methods (28). The ventricles were excised, cut into several pieces, and incubated overnight at 4°C in 1 mg/ml 1:300 trypsin (Life Technologies, Inc.) in Hanks' balanced salt solution (HBSS, Life Technologies, Inc.). The ventricular tissue was then digested with 1 mg/ml collagenase type II (239 units/mg; Worthington) in HBSS, centrifuged twice at 50 ϫ g to remove less dense cells such as fibroblasts, and then plated. The cells were cultured at 37°C, 5% CO 2 in Dulbecco's modified Eagle's medium (DMEM, BioWhittaker, Walkersville, MD) containing 7% bovine fetal calf serum (FCS, Life Technologies, Inc.), 50 units/ml penicillin, and 50 g/ml streptomycin (Life Technologies, Inc.).

Materials
Mechanical Strain Device and Preparation of Cells-Mechanical deformation was applied to a thin and transparent membrane on which cells were cultured, an approach that produces controlled cellular strain as well as visualization of cells. The device used in this study provides a nearly homogeneous and uniform biaxial strain profile, i.e. strains that are equal at all locations on the membrane and in all directions. This approach eliminates locations on the substrate that have very high strains (20 -30%) in one direction. We have previously measured membrane strains with a high resolution video device (29); the cams used for this study gave strains of 1, 4, 9, and 14%. We have observed that strains larger than 14% lead to cell injury (data not shown), and the design of this device is limited to a maximum strain of 14%.
NRVM were plated on a fibronectin-coated (2 g/ml) membrane dish at a density of 2,000,000 cells/dish in 13 ml of DMEM containing 7% FCS and incubated 24 h. Approximate cell confluence was 85-90%. NRVM were then made quiescent by washing with 10 ml of HBSS twice and incubating with 10 ml of DMEM containing 1% insulin, transferrin, selenium media supplement (Sigma), 50 units/ml penicillin, and 50 g/ml streptomycin. All experiments were performed with NRVM that had been serum-starved for 24 h. We routinely obtained primary cultures with Ͼ95% myocytes, as assessed by microscopic observation of spontaneous contraction and by immunofluorescence staining with a monoclonal human ventricular myosin heavy chain antibody (Biogenesis, Poole, UK). 90 -95% of these cells also stained positive by a monoclonal human tenascin-C antibody (Locus Genex Oy, Helsinki, Finland).
Northern Analysis-Total RNA was isolated by the guanidinium thiocyanate and phenol chloroform method (30). Purified RNA (1 g) was used for the synthesis of cDNA with a reverse transcriptasepolymerase chain reaction system (Stratagene, La Jolla, CA). Synthesis of the cDNAs was performed by polymerase chain reaction with Taq polymerase (Perkin-Elmer). The primer set for the synthesis of the 450-base pair tenascin-C cDNA probe contained the 5Ј-TCT-GTC-CTG-GAC-TGC-TGA-TG-3Ј sense and 5Ј-TCT-TCA-AAT-CCC-TTC-ATG-GC-3Ј antisense oligonucleotides. The primer set for the synthesis of the 585-base pair secreted protein acidic and rich in cysteine (SPARC) cDNA probe contained the 5Ј-AAA-CAT-GGC-AAG-GTG-TGT-GA-3Ј sense and 5Ј-GGT-CTC-AAA-GAA-GCG-AGT-GG-3Ј antisense oligonucleotides. The primer set for the synthesis of the 546-base pair fibronectin cDNA probe contained the 5Ј-GGG-AGA-AGT-TTG-TGC-ATG-GT-3Ј sense and 5Ј-CCT-CGC-TCA-GTT-CGT-ACT-CC-3Ј antisense oligonucleotides. Rat BNP cDNA was kindly provided by Dr. David G. Gardner, University of California, San Francisco. These cDNA were radiolabeled by the random priming method with [␣-32 P]dCTP and the Klenow fragment of DNA polymerase (Stratagene). For Northern blotting, 15 g of total RNA was loaded on a 1.0% formaldehyde gel (2.0 M), transferred to a nylon membrane (Stratagene), and UV cross-linked with a UV Stratalinker (Stratagene). The probe was hybridized with QuikHyb solution (Stratagene) at 68°C for 1 h. The membrane was exposed to x-ray film overnight at Ϫ80°C with one intensifying screen. Normalization of RNA for equal loading was carried out by hybridizing the blots with a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probe (CLONTECH, Palo Alto, CA). Levels of tenascin-C and GAPDH mRNA were quantitated by densitometry of the Northern blot autoradiographs using Optimas 5.0 densitometry software (Optimas, Bothell, WA). RNase Protection Assay-RNase protection assay was performed using the RPA II kit (Ambion, Austin, TX). A 409-base pair rat tenascin-C cDNA for RNase protection assay was used as probe. Synthesis of the tenascin-C cDNA was performed by polymerase chain reaction with Taq polymerase (Perkin-Elmer). The primer set for the synthesis of the tenascin cDNA probe contained the 5Ј-GCA-TCC-GTA-CCA-AAA-CCA-TC-3Ј sense and 5Ј-CGG-AAA-TTC-TCC-ACT-TGA-GC-3Ј antisense oligonucleotides. The tenascin-C cDNA was subcloned into the pCR®II vector (Invitrogen, Carlsbad, CA), and the sequence of tenascin-C cDNA was confirmed. The vector was linealized with HindIII restriction endonuclease, and a 32 P-labeled antisense riboprobe was made using T7 RNA polymerase. A 32 P-labeled riboprobe for 18 S was also transcribed with T7 RNA polymerase. Twenty-microgram samples were then hybridized with both probes (tenascin-C, 300,000 cpm/sample; 18 S, 20,000 cpm/sample) overnight at 45°C, followed by RNase A/RNase T1 digestion. The protected fragments were separated on a 5% polyacrylamide, 8 M urea gel. The gel was exposed to x-ray film at Ϫ80°C overnight. The amount of RNA was internally standardized using 18 S mRNA levels.
Western Analysis of Tenascin-C Protein-Conditioned media were concentrated by Centricon-10 miniconcentrators (Amicon, Inc., Beverly, MA). Total protein concentration was measured by the Bradford method (Bio-Rad), and equal quantities of total protein were loaded on a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane in 25 mM Tris base (pH 8.5), 0.2 M glycine, and 20% methanol. The nitrocellulose membrane was blocked by 5% nonfat dried milk in TBS washing buffer containing 20 mM Tris base (pH 7.6), 137 mM NaCl, and 0.1% Tween 20 for 2 h. For the detection of tenascin-C, the membrane was incubated with 1:1000 diluted human monoclonal anti-tenascin-C antibody (Locus Genex Oy) for 1 h at 37°C and washed with TBS washing buffer for 30 min. After application of goat anti-mouse IgG coupled to peroxidase (Bio-Rad), the membrane was developed with the enhanced chemiluminescent (ECL) method (Amersham Pharmacia Biotech).
Transfection-CAT Assays-The human tenascin-C promoter constructs containing the chloramphenicol acetyltransferase (CAT) reporter were kindly provided by Dr. Bernard Binétruy, Institut de Recherche sur le Cancer, France (31). Cardiac myocytes were transfected with each reporter plasmid (10 g) using the calcium phosphate precipitation method (32). Dominant negative mutants of IB␣ pCMV4-IB␣⌬N and IB kinase (IKK) ␤ pCDNA3-IKK␤⌬34 (5 g), which were kindly provided by Dr. Dean W. Ballard, Vanderbilt University School of Medicine, Nashville, TN (33), and Dr. George Mosialos, Brigham and Women's Hospital, Boston, MA (34), respectively, were cotransfected. As an internal control for transfection efficiency, CMV.␤-gal plasmid (3 g) was cotransfected in all experiments; ␤-galactosidase staining indicated that cellular transfection efficiency was 10%. Cells were subjected to mechanical strain 48 h after transfection, and lysates were prepared 24 h later for CAT and ␤-galactosidase assays (Promega, Madison, WI). CAT assays were performed as described by the manufacturer (Promega). The relative CAT activity was calculated as the ratio of CAT to ␤-galactosidase activity and standardized to unstimulated Ϫ220-CAT promoter activity (fold induction).
Electrophoretic Mobility Shift Assay-Nuclear extracts were prepared as described (35), and the protein concentration of each extract was determined by the Bradford method (Bio-Rad). Oligonucleotide probes for the NF-B consensus sequences (Promega) were end-labeled with [␥-32 P]dATP by incubating with T 4 polynucleotide kinase at 37°C for 10 min. The labeled probe was separated from unincorporated nucleotide using a Sephadex G-50 column (Amersham Pharmacia Biotech). Ten g of nuclear extract was incubated in 10 l of binding buffer containing 5 mM MgCl 2 , 2.5 mM EDTA, 2.5 mM dithiothreitol, 250 mM NaCl, 50 mM Tris-HCl (pH 7.5), 0.25 mg/ml poly(dI-dC), and 20% glycerol for 10 min at room temperature. 32 P-Labeled oligonucleotide probe (50,000 -200,000 cpm) was then added, and the reaction mixture was incubated for 20 min at room temperature. DNA-protein complexes were resolved with 6% nondenaturing polyacrylamide gel electrophoresis at 12 V/cm in 0.5ϫ TBE. Specificity was determined by the addition of p50 antibody (15 g IgG/ml; Santa Cruz Biotechnology Inc., Santa Cruz, CA) or excess unlabeled (cold) NF-B oligonucleotide (1.75 pM) to the nuclear extracts for 10 min before addition of radiolabeled probe.
Statistical Analysis-Data are expressed as the mean Ϯ S.D. A paired two-tailed Student's t test was used to analyze differences between two groups, and p values of Ͻ 0.05 were considered significant.

Effects of Mechanical Strain on
Tenascin-C mRNA Accumulation-We first investigated whether mechanical strain modulated gene expression of tenascin-C as well as SPARC, a remodeling related matrix protein, and fibronectin in NRVM. As shown in Fig. 1, 9% cyclic mechanical strain at 1 Hz induced tenascin-C mRNA accumulation with a maximum peak at 6 h (3.9 Ϯ 0.5-fold by densitometry, p Ͻ 0.01, n ϭ 13), whereas mechanical strain did not induce SPARC nor fibronectin mRNA. In addition, when cyclic biaxial strains of 1, 4, 9, and 14% at 1 Hz for 6 h were imposed, induction of tenascin-C mRNA expression in NRVM was amplitude-dependent (Fig. 2). In these studies, no morphologic changes in cardiac myocytes were detected following strains of 1-14%.
LaFleur et al. (36) reported that there are alternatively spliced isoforms of rat tenascin-C, and our Northern analysis suggested that at least two of these species are induced by deformation of cardiac myocytes. We then performed an RNase protection assay with a probe designed to distinguish the three rat tenascin-C cDNA species. We observed three bands of 107, 375, and 409 base pairs in size, corresponding to the three alternatively spliced messages (36); 9% mechanical strain and PDGF-BB (1 ng/ml) induced all three alternatively spliced isoforms of tenascin-C (Fig. 3).
Effects of Mechanical Strain on the Stability of Tenascin-C mRNA-To determine whether mechanical strain increased tenascin-C mRNA accumulation by increasing the rate of synthesis or by decreasing the rate of degradation, NRVM were incubated in the presence or absence of 9% strain for 6 h and then incubated further with actinomycin D (5 g/ml) to inhibit transcriptional activity. The half-life (t1 ⁄2 Ͻ3 h) of tenascin-C mRNA was not affected by mechanical strain (Fig. 4). This experiment suggested that mechanical strain increases the rate of synthesis of tenascin C mRNA.
Effects of Mechanical Strain on Tenascin-C Protein Accumulation-We next investigated whether the increase in tenascin-C mRNA by mechanical strain was accompanied by an increase in tenascin-C protein. The expression of tenascin-C protein was analyzed by immunoblotting with anti-tenascin-C antibody. The tenascin-C protein bands at 220 and 250 kDa were increased in an amplitude-dependent manner (Fig. 5).
Effects of Cycloheximide and Angiotensin II Receptor Type 1 Antagonist on the Induction of Tenascin-C mRNA-Mechanical strain may induce proteins including growth factors and vasoconstrictors in cardiac myocytes. We studied whether new protein synthesis was required for the induction of tenascin-C mRNA expression by strain. The addition of cycloheximide (10 M; 0% inhibition), a protein synthesis inhibitor, did not affect the induction of tenascin-C mRNA by strain in NRVM (Fig.  6A). This indicates that cyclic mechanical strain induces tenascin-C mRNA expression in NRVM without new protein synthesis. In addition, we investigated whether the effect of cyclic mechanical strain on tenascin-C mRNA expression in NRVM is angiotensin II-dependent. NRVM were subjected to 9% cyclic strain at 1 Hz for 6 h in the presence or absence of an AT 1 antagonist. As shown in Fig. 6B, CP191,166 (0.1 M), an AT 1 antagonist, inhibited the induction of BNP mRNA (52% inhibition) but not tenascin-C mRNA (0% inhibition) by cyclic mechanical strain. Therefore, in contrast to the previously reported effect of angiotensin II on strain-mediated BNP induction (37), it is unlikely that angiotensin II is involved in the effect of cyclic mechanical strain on tenascin-C mRNA expression in NRVM.
Involvement of Protein Kinase C, Tyrosine Kinase, and MAP Kinase-Mechanical strain activates protein kinase C, tyrosine kinases, and MAP kinases in cardiac myocytes (7-11). We examined the role of three pathways on the induction of tena- NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed to 0 or 9% cyclic mechanical strain (1 Hz) for 24 h. Total RNA was isolated and analyzed by Northern blotting with 32 P-labeled tenascin-C, SPARC, fibronectin, and GAPDH cDNA probes. Data are representative of six experiments that gave nearly identical results. At least two tenascin-C mRNA species are increased at 3, 6, and 12 h.
FIG. 2. The dependence of tenascin-C mRNA expression in NRVM on the amplitude of mechanical strain. NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed for 6 h to 0, 1, 4, 9, and 14% cyclic mechanical strain (1 Hz). Total RNA was isolated and analyzed by Northern blotting with 32 P-labeled tenascin-C (upper panel) and GAPDH (lower panel) cDNA probes. Data are representative of two experiments that gave nearly identical results.
Effects of Antioxidants on the Induction of Tenascin-C mRNA Expression and Tenascin-C Promoter Activity-ROS may participate in mechanotransduction; therefore, we investigated the effects of antioxidants on the induction of tenascin-C mRNA by strain. Antioxidants, such as N-acetyl-L-cysteine (10 mM; 90% inhibition), catalase (500 units/ml; 93% inhibition), and Tiron (10 mM; 95% inhibition) inhibited the effect of strain, whereas inhibition of reactive nitrogen species by 7-nitroindazole (100 M; 3% inhibition) had no effect (Fig. 8). Furthermore, the addition of H 2 O 2 (100 M) induced tenascin-C mRNA in the absence of strain.
To identify possible sites of induction of the tenascin-C promoter by mechanical strain, we performed transfection studies with promoter reporter constructs containing functional cisactivating elements for activated protein (AP)-1, NF-B, and GCN4 (AP-1-like site) (Fig. 9A). The tenascin-C promoter contains a consensus AP-1 site at position Ϫ875, the function of which has FIG. 3. Regulation of alternative tenascin-C mRNA transcripts in NRVM by mechanical strain or PDGF-BB. NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed to 0 or 9% cyclic mechanical strain (1 Hz) or PDGF-BB (1 ng/ml) for 6 h. Total RNA was isolated, and this RNA (20 g) was hybridized with 300,000 cpm tenascin-C and 20,000 cpm 18 S riboprobes at 45°C overnight. Hybrids were treated with RNase solution and resolved 5% polyacrylamide gel. Yeast RNA (20 g) was used as a negative control. All three predicted tenascin-C species (107, 375, and 409 base pairs) are increased in the presence of strain.
FIG. 4. Effects of mechanical strain on the stability of tenascin-C mRNA. NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed to 0 (E) or 9% (q) cyclic mechanical strain (1 Hz) for 6 h and were further incubated with actinomycin D (5 g/ml) for the indicated times. For each time point, total RNA (10 g) was prepared and analyzed by Northern blotting. Each point is the mean of two separate experiments.

FIG. 5. Effects of mechanical strain on tenascin-C protein ex-
pression. NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed to 0, 4, 9, or 14% cyclic mechanical strain (1 Hz) for 24 h. Media were then analyzed by Western analysis. The molecular mass of tenascin C protein is 220 and 250 kDa. Data are representative of two experiments that gave nearly identical results.

FIG. 6. Effects of cycloheximide (A) and angiotensin II receptor type 1 antagonist (B) on induction of tenascin-C mRNA expression by mechanical strain.
NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed for 6 h to 0 or 9% cyclic mechanical strain (1 Hz) in the presence or absence of cycloheximide (10 M) or CP191,166 (0.1 M). Cycloheximide or CP191,166 was applied to the myocytes 30 min before mechanical strain. Total RNA was isolated and analyzed by Northern blotting with 32 P-labeled tenascin-C (upper panel), brain natriuretic peptide (BNP, middle panel), and GAPDH (lower panel) cDNA probes. Data are representative of two experiments that gave nearly identical results. not been demonstrated (38). The Ϫ1175-CAT reporter construct, containing the consensus AP-1 site, displayed a low basal activity and was stimulated 2.6-fold by mechanical strain compared with no strain. A shorter promoter construct, Ϫ220-CAT, was stimulated 6.6 Ϯ 0.9-fold by mechanical strain, whereas promoter constructs without the NF-B site were not activated by strain (Fig. 9A). To investigate the role of NF-B in strain-induced tenascin-C promoter activity, dominant negative mutants of IB␣ (pCMV4-IB␣⌬N) or IKK␤ (pCDNA3-IKK␤⌬34) were cotransfected. Both mutants blocked Ϫ220-CAT promoter activation by mechanical strain, whereas control vectors pCMV4 and pCDNA3 had no effect (Fig. 9B).
In addition, N-acetyl-L-cysteine (10 mM), catalase (500 units/ ml), and Tiron (10 mM) inhibited the effect of strain, whereas inhibition of reactive nitrogen species by 7-nitroindazole (100 M) had no effect (Fig. 10). The addition of H 2 O 2 (100 M) also increased Ϫ220-CAT promoter activity in the absence of strain. These findings suggest that activation of tenascin-C promoter regions by mechanical strain through the NF-B-binding site involves ROS.
Effects of Antioxidant on NF-B Activation by Mechanical Strain-Since it is likely that the activation of NF-B is required for induction of tenascin-C expression by mechanical strain, electrophoretic mobility shift assays were performed using radiolabeled NF-B oligonucleotide. As shown in Fig. 11, 9% mechanical strain increased the amount of shifted complex. N-Acetyl-L-cysteine (10 mM) significantly inhibited the activation of NF-B by mechanical strain. The shifted complexes were specific for NF-B since they were supershifted in the presence of antibody to NF-B subunit and disappeared with excess unlabeled oligonucleotide. Furthermore, the addition of H 2 O 2 (100 M) induced the activation of NF-B in the absence of strain. DISCUSSION Tenascin-C is prominent in embryonic and adult tissues that are actively remodeling (39). In cardiovascular tissues, tenas-  8. Effects of antioxidants on induction of tenascin-C mRNA expression by mechanical strain. NRVM were plated on 2 g/ml fibronectin in DMEM containing 7% FCS for 24 h. After serum deprivation for 24 h, myocytes were exposed for 6 h to 0 or 9% cyclic mechanical strain (1 Hz) in the presence or absence of N-acetyl-Lcysteine (NAC, 10 mM), catalase (Cat, 500 units/ml), Tiron (10 mM), 7-nitroindazole (7-NI, 100 M), or hydrogen peroxide (H 2 O 2 , 100 M). N-Acetyl-L-cysteine, catalase, Tiron, or 7-nitroindazole was applied to the myocytes 1 h before mechanical strain. Total RNA was isolated and analyzed by Northern blotting with 32 P-labeled tenascin-C (upper panel) and GAPDH (lower panel) cDNA probes. Data are representative of two experiments that gave nearly identical results.

FIG. 9. Role of NF-B in induction of tenascin-C promoter activity by mechanical strain.
A, tenascin-C promoter-CAT gene reporter constructs were used in transfection experiments. The length of the promoter upstream from the initiation start site is donated in base pairs along with putative nuclear protein binding domains. Cardiac myocytes were transfected with each reporter plasmid (10 g) and CMV.␤-gal plasmid (3 g). B, dominant negative mutants of IB␣ pCMV4-IB␣⌬N or IKK␤ pCDNA3-IKK␤⌬34 (5 g) were cotransfected with Ϫ220-CAT construct (10 g) and CMV.␤-gal plasmid (3 g). Cells were subject to 0 (open column) or 9% (closed column) mechanical strain 48 h after transfection, and lysates were prepared 24 h later for CAT assays and ␤-galactosidase assays. The relative CAT activity was calculated as the ratio of CAT to ␤-galactosidase activity and standardized to unstimulated Ϫ220-CAT promoter activity (fold induction). Each experiment was performed three times with triplicate data measurements. Results are presented as mean Ϯ S.D. * represents a significant change from no exposure to mechanical strain (p Ͻ 0.01). ** represents a significant change from exposure to mechanical strain (p Ͻ 0.01).
cin-C expression has been described in the developing heart (40), in normal blood vessels (41), and in carotid arteries after experimental balloon injury (42). Previous studies reported that mechanical strain may induce tenascin-C in fibroblasts (43) and in pulmonary arteries (44). Chiquet-Ehrismann et al. (43) reported that tenascin-C promoter expression was directly or indirectly activated in fibroblasts generating and experiencing mechanical stress within a restricted collagen matrix. Jones and Rabinovitch (44) demonstrated that the induction of tenascin-C accompanied progressive pulmonary vascular changes. In the present study, cyclic mechanical strain induced tenascin-C mRNA and protein accumulations in cardiac myocytes. These findings suggest that tenascin-C may participate in cardiac remodeling induced by mechanical overload, such as hypertensive heart disease.
The local renin-angiotensin system may play an important role in the adaptation of the heart to pressure and volume overload (45,46). Sadoshima et al. (47) reported that [Sar-1,Ile-8]angiotensin II, a specific angiotensin II receptor antagonist, completely inhibited the stretch-induced c-fos expression in neonatal rat cardiac myocytes. Although angiotensin II secreted from cardiac myocytes is an important mediator of the strain-induced hypertrophic response in vitro, angiotensin II does not appear necessary for cardiac hypertrophy induced by mechanical strain in vivo. Thienelt et al. (48) demonstrated that the acute growth responses induced by systolic pressure overload in rat did not depend on AT 1 activation, and Dell'Italia et al. (49) reported that volume overload cardiac hypertrophy in dogs with chronic mitral regurgitation was unaffected by angiotensin-converting enzyme inhibitor treatment. Harada et al. (5,6) proposed that AT 1 -mediated angiotensin II signaling is not essential for the development of pressure overload-induced cardiac hypertrophy. In the present study, AT 1 blockade did not affect the induction of tenascin-C mRNA expression in NRVM, whereas AT 1 blockade partially inhibited BNP mRNA induction by strain. This is consistent with the hypothesis that factors other than angiotensin II are involved in cardiac remodeling induced by mechanical strain.
Mechanical strain rapidly increases protein kinase C and MAP kinase activity in cardiac myocytes (8,11). Stretch-induced c-fos expression is inhibited by both protein kinase C inhibitors and down-regulation of protein kinase C (7,11). In addition, Schwachtgen et al. (50) reported that treatment of human endothelial cells with PD98059 (MAP kinase kinase inhibitor) inhibited shear stress activation of egr-1. In the present study, staurosporine and calphostin C (protein kinase C inhibitors), PD98059, and SB203580 (a p38-MAP kinase inhibitor) did not significantly inhibit the stimulatory effect of mechanical strain on tenascin-C mRNA expression in cardiac myocytes, suggesting that this effect is not mediated via activation of protein kinase C nor MAP kinase. However, we cannot exclude the possibility that the JNK pathway, one of three major MAP kinase signaling pathways, might be involved in the effect of mechanical strain on tenascin-C mRNA induction. Furthermore, it is important to recognize that in vitro biaxial strain amplitudes cannot be directly extrapolated to in vivo myocardial deformations. In vivo, the cardiac myocyte is in a three-dimensional tissue with complex active and passive loads, as well as extracellular matrix components that contribute to mechanical behavior.
Many studies have reported that superoxide or its derivative radicals can be demonstrated in reperfused isolated hearts (51-53). Nakamura et al. (54) recently reported that cardiac hypertrophy induced by tumor necrosis factor-␣ and angiotensin II was inhibited by antioxidants in rat cardiac myocytes and concluded that tumor necrosis factor-␣ and angiotensin II cause hypertrophy in part via the generation of ROS. Peng et al. (55) demonstrated that nuclear proteins induced by H 2 O 2 in rat cardiac myocytes are capable of binding to a DNA probe containing the NF-B elements. Cheng et al. (20) demonstrated that overstretching produced a 2.4-and a 21-fold increase in the generation of ROS and apoptotic myocytes, respectively. Their findings suggest that ROS production by mechanical deformation may lead to the impairment in force development of the myocardium.
In porcine aortic endothelial cells, cyclic strain induces an oxidant stress, and NADH/NADPH oxidase is a potential source of H 2 O 2 release in cyclically strained cells (56). An interesting observation is that many activators of NF-B have FIG. 10. Effects of antioxidants on induction of tenascin-C promoter activity by mechanical strain. Cardiac myocytes were transfected with Ϫ220-CAT construct (10 g) and CMV.␤-gal plasmid (3 g). After transfection for 48 h, myocytes were exposed for 24 h to 0 (open column) or 9% (closed column) cyclic mechanical strain (1 Hz) in the presence or absence of N-acetyl-L-cysteine (NAC, 10 mM), catalase (Cat, 500 units/ml), Tiron (10 mM), 7-nitroindazole (7-NI, 100 M), or hydrogen peroxide (H 2 O 2 , 100 M). N-Acetyl-L-cysteine, catalase, Tiron, or 7-nitroindazole was applied to the myocytes 1 h before mechanical strain. Lysates were prepared for CAT assays and ␤-galactosidase assays. The relative CAT activity was calculated as the ratio of CAT to ␤-galactosidase activity and standardized to unstimulated Ϫ220-CAT promoter activity (fold induction). Each experiment was performed three times with triplicate data measurements. Results are presented as mean Ϯ S.D. * represents a significant change from exposure to mechanical strain (p Ͻ 0.01).
FIG. 11. Electrophoretic mobility shift assay showing the effects of N-acetyl-L-cysteine (NAC) on activation of NF-B by mechanical strain. Myocytes were exposed for 1 h to 0 or 9% cyclic mechanical strain (1 Hz) or hydrogen peroxide (H 2 O 2 , 100 M) in the presence or absence of N-acetyl-L-cysteine (NAC, 10 mM). N-Acetyl-Lcysteine was applied to the myocytes 1 h before mechanical strain. Specificity was determined by addition of p50 antibody (Ab) (supershift) or unlabeled (cold) NF-B oligonucleotide (1.75 pM) to the nuclear extracts. Two separate experiments yielded similar results. been reported to also increase oxidative stress. In cardiac myocytes, the mechanisms by which mechanical strain generates ROS and by which generated ROS activates NF-B remain unknown. Activation of transcriptional factors NF-B and AP-1 by ROS during gene induction have been reported (57,58). However, in this study, the function of a consensus AP-1 site at position Ϫ875 was not demonstrated, as described previously (31,38). NF-B is retained in the cytoplasm by IB, which comprises a distinct family of proteins that bind to NF-B and inhibit nuclear translocation and DNA binding. IKK activation is followed by IB degradation and NF-B translocation into the nucleus, resulting in activation of NF-B. IB␣ is the principal regulator of NF-B activity and can remove bound NF-B from the nucleus, whereas IB␤ cannot (59,60). Two closely related kinases designated IKK␣ and IKK␤ have been identified as components of the multiprotein IKK complex (500 -900 kDa) that directly phosphorylates the critical Ser residues of IB proteins (61,62). In the present study, dominant negative mutants of IB␣ and IKK␤ significantly inhibited tenascin-C promoter activation by mechanical strain, indicating that NF-B pathway is involved in the induction of tenascin-C by strain. In addition, antioxidants inhibited the induction of tenascin-C mRNA and gene transcription and NF-B activation by cyclic mechanical strain in cardiac myocytes. These findings support the premise that mechanical strain induces the generation of ROS, which activates NF-B in rat cardiac myocytes, and that ROS may be a second messenger in cardiac remodeling.