INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CARDIAC CELLS RESPOND to mechanical forces in various ways, including activation of gene expression, protein synthesis, and mitogenesis (13). Mechanical loading induces the production and release of a number of trophic factors that act in an autocrine or paracrine fashion to regulate the proliferative response and phenotypes of cardiac fibroblasts, myocytes, and neighboring cells (21). The activation of cardiac fibroblasts by pressure overload is also crucial in the fibrotic response after hypertensive heart disease, dilated cardiomyopathy, and postmyocardial infarction because of the central role of these cells in the synthesis of extracellular matrix proteins (3). During the development of cardiac hypertrophy in response to chronic hemodynamic overload, cardiac fibroblasts undergo a phenotypic switch to become contractile myofibroblasts, cells that express abundant -smooth muscle actin (SMA; 2, 14). The increased expression of SMA is strongly upregulated in cells expressing angiotensin II receptors (23) and may be mediated by angiotensin II and other cytokines that are induced by mechanical load (3, 21). Notably, after adaptation to pressure overload, SMA-expressing cells regress (14). Currently, it is unknown if continuous, static force application alone directly regulates actin gene expression in cardiac fibroblasts or if trophic factors like angiotensin II interact with force to regulate SMA expression.

Some of the signaling pathways that mediate mechanically induced biological effects have been identified in vitro and include the extracellular signal-regulated kinase (ERK) pathways (19) and c-Jun NH₂-terminal kinase (JNK; 12). In rat cardiac fibroblasts subjected to passive biaxial stretch, the ERK and JNK pathways are rapidly activated, whereas the p38 kinase is unaffected (16). However, in other fibroblast model systems in which perpendicular forces are applied through integrins, p38 is activated, whereas ERK and JNK are unaffected (15). As cardiac fibroblasts adhere to extracellular matrix proteins through integrins (10), which provide sites for force transfer to the actin cytoskeleton, we examined whether SMA mRNA and protein are regulated by force applied through integrins. Magnetite beads coated with collagen were incubated with cells and subjected to controlled perpendicular forces generated by a magnetic field (8, 9). We asked whether regulation of SMA expression requires an intact actin cytoskeleton and whether mechanical signal transduction requires p38, ERK, and JNK kinases. We report here that, in cardiac fibroblasts induced to express SMA by culture on rigid plastic substrates (1), static force application through collagen receptors reduces constitutive SMA mRNA and protein content that is mediated through p38. This force-dependent reduction is not only abrogated by angiotensin II but is increased by this important prohypertensive peptide.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Reagents. Anti--SMA (clone no. 1A4), anti- actin (clone no. AC-15), anti-talin (clone no. 8D4), anti-vinculin (clone no. HVIN-1), anti-desmin (clone no. DE-U-10), anti-vimentin (clone no. VIM-13.2), cytochalasin, cycloheximide, and collagenase (C5138) were used. BSA, fibronectin, poly-L-lysine (PL), and angiotensin II were purchased from Sigma (St. Louis, MO). Goat anti-mouse IgG_2a and goat anti-mouse IgG₁ were purchased from Caltag (Burlinghame, CA). Cytochalasin D and SB-203580 were purchased from Calbiochem (San Diego, CA). Anti-p38, anti-ERK 1/2, and anti-JNK antibodies as well as the phosphospecific antibodies to each of these kinases were purchased from New England BioLabs (Beverly, MA). Losartan was a kind gift from Merck (Whitehouse Station, NJ). Collagen was obtained from Collagen Corporation.

Cell culture. Primary cardiac fibroblasts were obtained from 125-g adult Sprague-Dawley rats as previously described (17). In brief, rats were killed by CO₂ asphyxiation, and the hearts were quickly removed under sterile conditions. Ventricular tissue was excised, minced, and digested with 0.3% collagenase containing (wt/vol) 1.8% sorbitol, 0.05% DNase, 6.25 U/ml elastase, and 0.05% trypsin in Krebs buffer with Zn²⁺. Nonadherent cells (primarily myocytes, leukocytes, and endothelial cells) were washed away. The cardiac fibroblasts attached and proliferated much more rapidly than other cardiac cell types, properties that enabled us to obtain virtually pure cultures by the first passage. Cells were maintained in HG-DME containing 10% fetal bovine serum and a 1:10 dilution of an antibiotic solution [0.17% (wt/vol) penicillin V, 0.1% gentamycin sulfate, and 0.01 µg/ml amphotericin; Sigma] at 37°C in a humidified incubator gassed with 95% O₂-5%CO₂. Cells were passaged with 0.01% trypsin (GIBCO, Burlington, ON). Studies were performed on cells at passages 1-3 in HG-DME serum-free medium.

Bead coating. As described earlier (8, 9), 0.4 g of magnetite beads (Sigma-Aldrich) were incubated for 1 h with 1 ml of an acidic bovine collagen solution (Vitrogen, Collagen; >95% type I collagen) at 37°C and neutralized to pH 7.4 with 100 µl of 1 N NaOH. Under these conditions, collagen polymerizes and forms fibrils around the beads within 30 min. In some experiments, beads were coated with plasma fibronectin (Sigma) or with the fibronectin IIIA domain polypeptide (obtained from Dr. Lloyd Culp, Cleveland Clinic). The beads were sonicated to eliminate clumps. BSA, PL, or fibronectin beads were prepared in a similar fashion by incubating beads in solutions of 1 mg/ml BSA, 1 mg/ml PL, or 1 mg/ml fibronectin and then dispersed. Analysis of bead size was performed by electronic particle counting (Coulter Channelyzer, Coulter Electronics, Hialeah, FL). Particles tended to exhibit a heterogeneous size distribution with a pronounced modal peak at 5 µm, although there were many particles with smaller diameters. Beads were rinsed in PBS, washed three times, and resuspended in Ca²⁺-Mg²⁺-free PBS.

Force generation. A ceramic permanent magnet (Gr. 8, 2.2 cm × 9.6 cm × 11 cm; Jobmaster, Mississauga, ON) was used to generate perpendicular forces on beads attached to the dorsal surface of cells. For all experiments, the pole face was parallel to and 2 cm from the cell culture dish surface. At this distance, the force on a single fibroblast with ~750-µm² area of dorsal bead coverage was 480 pN or 0.65 pN/µm². Because the surface area of the magnet was larger than the culture dishes and because bead covering was relatively uniform for all cells, the forces applied to cells across the width of the culture dish were relatively uniform (7). A constant force of varying duration was used for all experiments. Before incubation with cells, beads were rinsed in PBS, washed three times, resuspended in calcium-free buffer, and added to attached cells in full medium for 10 min. Cells were washed three times to remove unbound beads and exposed to force in a PBS (pH 7.4) containing calcium and magnesium ions.

Isolation of focal adhesions. Proteins enriched in bead-associated focal adhesion complexes were assessed with previously described methods (18). Cells and attached beads were collected by scraping cells into ice-cold cytoskeleton extraction buffer (CSKB; Triton-X-100, 50 mM NaCl, 300 mM sucrose, 3 mM MgCl₂, 20 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 mM PMSF, and 10 mM PIPES; pH 6.8). The isolation procedure was carried out at 4°C using a side-pull magnetic isolation apparatus (Dynal, Lake Placid, NY). The cell-bead suspension was sonicated for 10 s (output setting 3, power 15%; Sonifier 185, Branson) and homogenized in a 2-ml Dounce homogenizer (20 strokes). The magnetic beads were pelleted and washed three times with CSKB before protein analysis.

Immunofluorescence and immunoblotting. We assessed SMA content in early passage cultures by immunostaining for SMA with the SMA antibody followed by fluorescein isothiocyanate-conjugated goat anti-mouse IgG. Cells were examined in an epifluorescence microscope and photographed. For immunoblots, protein from beads or cell lysates prepared from cell cultures (60-mm dishes) that had been subjected to an applied force for specific time intervals were analyzed. Cells were rinsed with PBS, lysed by adding 200 µl of SDS sample buffer (62.5 mM Tris · HCl, pH 6.8, 2% SDS, 10% glycerol, 50 mM dithiothreitol, and 0.1% wt/vol bromophenol blue), and transferred to a microfuge tube. The samples were kept on ice and then boiled for 5 min. Protein concentration was assessed by the Bio-Rad assay, and equal amounts of protein were loaded in each lane. Isolated proteins were separated by SDS-PAGE (10% acrylamide) and transferred to nitrocellulose. Actin, vinculin, talin, JNK, ERK 1/2, and p38 proteins were identified by immunoblotting. Blots were blocked for 1 h with 5% skim milk in PBS and incubated in the indicated antibody (diluted 1:1,000 in 0.5% Tween-PBS) for 1 h at room temperature. Blots were washed with 0.5% Tween-PBS for 10 min, incubated with appropriate second antibodies for 1 h, washed four times in Tween-PBS, and developed by chemiluminescence (Amersham). X-OMAT Kodak films were exposed to the blots, and the density of the bands was analysis by IP Lab Gel Scientific Image Processing (Signal Analytics, Vienna, VA).

Northern analysis. Total RNA was isolated from cells by the QIAGEN RNAeasy Total RNA kit according to the manufacturer's instructions and quantified by spectrophotometry (Ultrospec 3000; Pharmacia Biotech; Montreal, Quebec). RNA samples (15 µg) were separated in a 1.2% denaturing agarose gel containing 2.2 M formaldehyde in MOPS running buffer, transferred to a nitrocellulose membrane (OPTITRAN; Schleicher & Schuell), cross-linked by ultraviolet light treatment, and hybridized with ³²P-labeled oligonucleotide probes. These probes were designed from portions of the sequences of the rat -SMA mRNA 5'-untranslated region (5'-GAAAAGAACTGAAGGCGCTGATCCACAAAACATTCACAGTTG-3') and from the rat -actin mRNA 3'-untranslated region (5'-CGCCTTCACCGTTCCAGTTTTTAAATCCTTGAGTCAAAAGCGCCA-3').

Statistical analysis. For all assays, three or more separate experiments were performed. Means ± SE were calculated for continuous variables and, when appropriate, comparisons between two groups were analyzed by unpaired t-tests.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

On the basis of previous experiences with gingival fibroblasts that normally synthesize minimal SMA in vivo but can be induced to synthesize SMA in culture (1), we induced SMA expression in cardiac fibroblasts by plating cells on rigid tissue culture plastic and culturing in 10% fetal bovine serum for up to 5 days. Cells were evaluated by immunofluorescence at passage levels 1-3. Under these conditions, staining for SMA was very low on the first day of culture in first passage cells (Fig. 1A) but within 48 h of plating, there was abundant SMA. Similarly, in cells at passages 2-3, cells were brightly stained for SMA, which was localized to stress fibers (Fig. 1, B-E). For all further experiments, cells from passages 2 and 3 were analyzed, and these cells uniformly expressed abundant SMA as determined by immunoblotting and Northern blotting (see below). The cells used in these experiments were desmin negative and vimentin positive (Fig. 1F) and consistently showed the morphology of fibroblastic cells in phase contrast microscopy. Collectively, these findings indicated that the cells under study were indeed fibroblasts.

View larger version (94K): [in this window] [in a new window]   Fig. 1.   Phase contrast and immunofluorescence micrographs of -smooth muscle actin (SMA) staining in primary (1st passage) rat cardiac fibroblast cultures (A-C) or in 3rd passage cultures (D, E). Cells were cultured for the times indicated in corner, photographed in a phase contrast microscope, fixed, and immunostained for SMA. Note that culture of 1st passage cells on tissue culture plastic within 48 h induces abundant SMA expression in cardiac fibroblasts. Staining for vimentin and absence of staining for desmin (F) confirm the fibroblastic identify of these cells.

In cells exposed to force applied through collagen-coated magnetite beads, SMA was reduced by a minimum of 50% after 4-6 h of constant force application (Fig. 2A; P < 0.05). In five other independent experiments, force reduced SMA to 30-50% of baseline levels, and at least 4-6 h was required before this reduction was observed. This force-induced reduction of SMA did not affect total -actin content, and, indeed, when the ratio of SMA to -actin protein was calculated, there was also a 50% reduction of this ratio (P < 0.02; Fig. 2A).

View larger version (43K): [in this window] [in a new window]   Fig. 2.   A: -SMA content by force application. Immunoblots of SMA and -actin in cells at various times after physical stretching with magnetite beads. Note the reduction of SMA content compared with -actin content at 4 h, which is reflected quantitatively in the ratios of densitometric scans of the blots (n = 5 blots per sample; each sample was taken from a separate culture dish). SMA was significantly reduced at 4 and 6 h compared with controls (P < 0.05). Equal amounts of protein (by Bio-Rad assay) were loaded on each lane. B: requirement for actin filaments and microtubules. Cells were subjected to force applied through beads coated with either collagen or BSA as indicated. Cells were either treated with beads and no force (No Force) or with force for 4 h (Force) or were preincubated with either cytochalasin D (Cyto D) or colchicine (Colc) as indicated. BSA-coated beads and cells preincubated with cytochalasin D or colchicine showed no reduction of SMA content. For immunostaining of SMA, cells were treated with vehicle (no colchicine) or with colchicine and immunostained for SMA.

The force-induced reduction required intact actin filaments, because cells pretreated with cytochalasin D (1 µM; 20 min) showed no reduction of SMA after force application (Fig. 2B). Cytochalasin D treatment alone had no significant effect on SMA content. Pretreatment with 10 µM colchicine also prevented the force-induced reduction of SMA but did not destroy stress fibers (Fig. 2B), indicating that microtubules and stress fibers per se were not directly involved in the force regulation of SMA. Furthermore, immunoblotting showed that colchicine by itself did not change SMA levels (Fig. 2B). Application of force through BSA-coated beads also caused no reduction of SMA, indicating that force required transmission through collagen receptors for the effect on SMA (Fig. 2B). This difference between BSA and collagen-coated beads was not due to failure of the BSA beads to remain attached to the cells. After force application followed by washing away of unbound beads, image analysis measurements of the area of cells covered by beads showed no statistically significant difference between collagen and BSA beads (for collagen beads mean = 53,594 ± 2,923 µm²/1.43 mm², for BSA beads mean = 47,600 ± 6,348 µm²/1.43 mm²; P > 0.2).

We found that the effect of SMA reduction by force application was apparently reversible, because removal of force (for 24 h) led to nearly complete restoration of SMA levels (Fig. 2A), a result caused presumably by an adaptation to the applied force (i.e., increased SMA synthesis and reduced leakage; see below). In the presence of constant force application, SMA was reproducibly decreased after 4 h and then returned to near baseline levels by 8 h. Consequently, we determined if the force-induced reduction could be repeated by subjecting cells to a second round of force application 24 h after the removal of force. Under these conditions, force again reduced SMA content but only to ~60% of baseline levels (Fig. 3).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   -SMA content after repeated force applications. Ratios of SMA and -actin content in cardiac fibroblasts after various stretching regimens (top). Stretching protocols over time are as indicated in diagram at bottom. Note that force reduces SMA content by ~50% and that removal of force leads to a return of normal SMA levels. A second application of force (F) 24 h later again induces a reduction of SMA. For each treatment, n = 5 samples. Stretching for 4 h induced a significant reduction of SMA (P < 0.05).

We measured SMA and -actin mRNA content after force application. Although 18S and 28S RNA were unchanged by force application, the ratio of SMA to -actin mRNA was reduced ~40% at 4 h and returned to baseline values by 6-8 h (Fig. 4). The force-induced reduction was not detectable if cells were preincubated with 1 µM cytochalasin D. 

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Effect of force on SMA mRNA. Top: Northern blots for SMA and -actin were performed with oligonucleotide probes specific for these isoforms. The relative abundance of SMA and -actin mRNAs was reduced 40% at 4 h (P < 0.05; n = 3). Although there was some slight reduction of -actin mRNA content, there was a relatively larger reduction of SMA, which is shown by the ratio data (bottom).

Angiotensin II has been strongly implicated in the force-induced cardiac hypertrophic response (20) and in the phenotypic switch of cardiac fibroblasts to myofibroblasts (22). As myofibroblast metabolism may be regulated through angiotensin type II receptors (23), we examined whether angiotensin II may be involved in the regulation of SMA by applied force. Incubation of cells with angiotensin II (10 nM) but without force application for up to 4 h caused an increase (25%; P < 0.05) of SMA content with respect to -actin by 2 h (Fig. 5A). This increase was blocked by 100 nM of losartan (SMA:-actin = 0 ± 5%), as is shown by the immunostaining in Fig. 5B. Notably, losartan greatly reduced SMA staining in cells with or without prior angiotensin II treatment. Most of the losartan-induced reduction of SMA was the staining in stress fibers. If cells were pretreated with angiotensin II before force application, there was no reduction of SMA (Fig. 5A) over a 4-h time period. Thus angiotensin II neutralized the reduction of SMA mediated by force. Notably, when losartan (100 nM) was administered concurrently with stretch, the reduction of SMA was increased (to 35% of baseline levels).

View larger version (46K): [in this window] [in a new window]   Fig. 5.   A: combined treatment with ANG II and force. Cells were treated with either ANG II (10 nM) for the indicated times or with ANG II and force (0-4 h). Note that ANG II increases SMA protein content (P < 0.05 at 2 and 4 h) but that force and ANG II together produce no change (n = 5 samples per treatment). B: immunofluorescence staining for SMA in cells treated with ANG II and losartan as indicated. Note that losartan reduces SMA staining, particularly in stress fibers.

Because collagen but not BSA-coated beads were required for the reduction of SMA, we asked if the force transmission was mediated through focal adhesions, specialized adhesive domains in fibroblasts that are enriched with integrins, vinculin, and other actin binding proteins (4). Immunoblotting of proteins associated with magnetite beads showed abundant vinculin, but there was no change of vinculin levels after 4 h of force application (Fig. 6A). This indicated that the reduction of SMA induced by force application was probably not due to a change in the relative numbers of focal adhesions that were binding beads. Indeed, beads coated with PL or BSA showed only very little bead-associated vinculin in immunoblots, despite equal protein loading (by Bio-Rad assay) on the gels for all types of bead coatings. Immunostaining for vinculin showed no change in the staining intensity or distribution of staining in focal adhesions after force application (Fig. 6B).

View larger version (42K): [in this window] [in a new window]   Fig. 6.   A: immunoblot of focal adhesion-related proteins obtained from magnetite bead preparations. Collagen- or poly-L-lysine-coated magnetite beads were incubated with cells and were either untreated (C) or were subjected to force for 4 h (4 h). The bead-associated proteins were normalized by Bio-Rad assay and immunoblotted for the focal adhesion protein vinculin. Note that 4 h of force does not increase the amount of vinculin, indicating that the number of focal adhesions was not influenced by the force. Poly-L-lysine-coated beads recruit very low levels of vinculin, indicating that the collagen-coated beads were much more efficient at binding to focal adhesions. B: cells were incubated with collagen-coated beads, and force was applied for 4 h as indicated. Immunostain for vinculin in focal adhesions shows no change after 4 h of force application. C: in control or 4 h-treated cells, SMA and -actin were immunoblotted from collagen-bead associated proteins equilibrated for equal loading. Note that force causes a very large reduction of SMA compared with -actin in the bead-associated proteins. D: comparison of bead coatings (Coll, collagen; Poly, poly-L-lysine) and effect on SMA after 4 h of force application (F). Note that only collagen-coated beads can induce reductions of SMA. E: cell culture medium from cells incubated with beads only (C) or beads and 4 h of stretch (F) was concentrated 5 times by Amicon filtration and immunoblotted for SMA and -actin. Note the marked increase of SMA compared with -actin in the sample that was subjected to force. F: effect of protein synthesis inhibition on force-induced SMA reduction. Cells were preincubated with cycloheximide (CHX) as described in the RESULTS and in Ref. 15. One sample was treated with CHX and with force (CHX+F). Note that inhibition of protein synthesis by CHX abrogated the stretch-induced reduction of SMA.

Immunoblotting of bead-associated proteins for SMA and -actin showed a 10-fold reduction of SMA (but not -actin) after force application (Fig. 6C), indicating that force exerted a more pronounced effect on SMA associated with focal adhesions than with global cellular SMA. We also tested whether cells incubated with PL-coated beads would show reductions of SMA after force application. Similar to BSA-coated beads, when force was applied to PL-coated beads, there was no reduction of SMA (Fig. 6D).

The reduction of SMA by force suggested the possibility that cells may selectively leak cytoskeletal proteins in response to the force. We immunoblotted five times Amicon-concentrated cell culture medium after 4 h of force application and compared the relative abundance of SMA and -actin (Fig. 6E). Force application caused a marked increase in the abundance of SMA but not -actin in the cell culture medium, and thus the ratio of SMA to -actin was also substantially increased by force.

As shown above (Fig. 2), immunoblotting of the cell pellets (without protein concentration) showed a substantial reduction of the ratio of SMA to -actin. We asked if the force-induced reduction required protein synthesis. In cells pretreated with cycloheximide (70 µM for 30 min; Ref. 15) and then subjected to force, there was no change in the ratio of SMA to -actin (Fig. 6F), indicating that the force-induced reduction of SMA required protein synthesis.

Force-mediated regulation of gene expression in many cell types is thought to be mediated in part through the mitogen-activated protein kinase pathway (20). We determined if a single, 4-h application of force through collagen-coated beads would activate the ERK, JNK, and p38 pathways. Cell lysates were immunoblotted for phosphorylated forms of these proteins and for total protein with specific antibodies. Force application caused a marked, time-dependent activation of p38, which was detectable at 15 min and persisted for up to 2 h. Treatment of cells with SB-203580 (10 µm; previously optimized for fibroblasts; Ref. 15) showed complete inhibition of p38 kinase activity (Fig. 7A). Ultraviolet light treatment also caused strong activation of p38 and of JNK; however, neither JNK or ERK 1/2 were detectably activated by force application. We tested whether ERK 1/2 could be activated by phorbol ester (PMA) and, as expected, PMA caused a rapid activation of ERK 1/2. These results were not due to uneven protein loading on the gels because reprobing the blots for -actin and for the unphosphorylated forms of the kinases showed equal protein loading. Application of force through BSA-coated beads (data not shown), plasma fibronectin-coated beads (Fig. 7B), or beads coated with the cellular E-IIIA fibronectin domain (Fig. 7C) caused no increase of p38 activation above baseline levels nor any change in SMA content. Furthermore, neither JNK nor ERK 1/2 was activated by force applied through fibronectin-coated beads.

View larger version (41K): [in this window] [in a new window]   Fig. 7.   A: immunoblots of phosphorylated and unphosphorylated forms of p38, c-Jun NH2-terminal kinase (JNK), and extracellular signal-related kinase (ERK) 1/2 after stimulation with either force, phorbol ester (PMA), or ultraviolet light (UV). Force induced a time-dependent increase of phosphorylated p38, which was completely blocked by the p38 kinase inhibitor SB-203580. UV light also induced p38 kinase activation. C6 shows the positive controls for phosphorylated p38. Force did not affect ERK 1/2 activation, although PMA induced a strong activation of ERK 1/2. Similarly, force had no effect on JNK, whereas stimulation with UV light also activated JNK (positive control). Equal protein loading was performed by equilibration with the Bio-Rad assay and by immunoblotting for -actin. B,C: immunoblots show that plasma fibronectin-coated beads (B) or beads coated with the E-IIIA splice variant of fibronectin (C) and subjected to force induced no reduction of SMA over a 4-h time course and did not substantially activate p38. Similarly, ERK 1/2 and JNK were not activated by force in this model. D: immunoblots show that ANG II did not activate p38. In cells treated with force and ANG II, there was evidence of p38 activation but only after prolonged exposure of immunoblots (10× normal exposures). Simultaneous treatment of cells with force, ANG II, and the p38 kinase inhibitor SB-203586 caused a nearly complete blockade of force-induced p38 activation.

We showed above (Fig. 5) that force and angiotensin II may interact to regulate SMA content. Consequently, we determined if they might also differentially activate p38. Over a 4-h time course with continuous incubation, angiotensin II alone caused a very slight and transient activation of p38 at 2 h if the film exposure time for immunoblotting was increased by 10-fold. Force plus angiotensin II caused a prolonged activation of p38 (from 15 min to 2 h; Fig. 7D), an effect that was similar to force alone (Fig. 7A). If cells were treated with force, angiotensin II, and the p38 kinase inhibitor SB-203586 (10 µm), there was virtually no detectable p38 activation, even with 10-fold longer film exposures (Fig. 7D).

Because p38 appeared to be important in regulating SMA content in response to force, we preincubated cells with the p38 inhibitor SB-203580 and measured SMA. As shown before, in cells treated with vehicle and force the ratio of SMA to -actin content was reduced ~50% by 4 h. If, however, cells were preincubated with SB-203580, force caused a nearly 50% increase of the ratio of SMA to -actin content (P < 0.05; Fig. 8).

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Inhibition of p38 kinase blocks force-induced reduction of SMA. Cells were treated with either force alone or with force and the p38 kinase inhibitor SB-203580 (10 µm). Immunoblots of SMA and -actin show that inhibition of p38 kinase not only abrogated the force-induced SMA reduction but led to an increase of SMA (P < 0.05 at 2 and 4 h).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the pathogenesis of pressure overload hypertrophy, cardiac fibroblasts are of central importance because they synthesize excessive amounts of extracellular matrix proteins (3). Notably, pressure overload induces cardiac fibroblasts to transiently express SMA (2, 14), an actin isoform that is a marker for myofibroblasts in the diseased heart (23). The principal finding in this report is that mechanical forces, when applied through collagen receptors in focal adhesions, can selectively reduce SMA content in cardiac fibroblasts, whereas the prohypertensive peptide, angiotensin II, increases SMA content. These data indicate that the increase of SMA in cardiac fibroblasts seen in pressure overload conditions (14) is more likely a response to angiotensin II than to altered force levels alone. Indeed, force and angiotensin II may act antagonistically to regulate SMA in pathological hypertrophic situations.

Up to the present, the regulatory mechanisms of mechanically induced SMA expression in cardiac myofibroblasts have not been examined. In this study, we used a well-characterized model (7, 8, 15) to deliver perpendicular forces of consistent and physiologically meaningful magnitudes (8, 16a) through integrins to actin. Unlike previous studies that have used deformable substrates (16, 20) to generate highly variable levels of passive whole cell stretching through integrins (25), the model used here creates tensile forces of a consistent magnitude at focal adhesions, sites that are enriched with actin filaments. The importance of actin filaments and microtubule arrays in the stretch response was shown by cytochalasin D and colchicine treatments that abrogated the force-induced reduction of SMA, an effect that is similar to that of force on -skeletal actin (15). In addition, SMA was enriched in focal adhesions. For myofibroblasts derived from a heavily force-loaded tissue such as the myocardium, this finding underlines the structural importance of the SMA isoform in the stretch response mediated through collagen receptors.

The collagen-magnetite bead model is well suited to the study of force-induced changes of SMA in fibroblasts from cardiac tissues. Fibrillar collagen is the most abundant cardiac extracellular matrix protein, and the greatly increased synthesis of this protein is central to the pathogenesis of the hypertrophic response (3). We used collagen to coat the beads and thereby provided a physiologically meaningful "connector" by which force could be applied to collagen receptors in focal adhesions (8, 18). As beads coated with PL did not induce a stretch response and were inefficient at recruiting focal adhesion proteins to beads, we conclude that the aggregation of collagen receptors in focal adhesions is a requirement for the force-induced regulation of SMA. The force effect was ligand specific: application of force through collagen but not fibronectin could regulate SMA levels, although either of these proteins when coated on beads would induce focal adhesion formation. The lack of effect with either the plasma fibronectin or the cellular fibronectin IIIA domain polypeptide is notable in that previous studies have shown an important role for this domain in fibronectin for the induction of SMA in vitro (21a). In a previous report using matrix ligand-coated deformable surfaces, cardiac fibroblasts were activated by biaxial stretch (16), and this activation was integrin specific. For the focal adhesion model reported here, we surmise that integrin specificity (e.g., ₂₁-integrin but not ₅₁) may be a restriction factor in stretch-induced SMA regulation.

In the context of SMA, hypertrophy, and cardiac overload (2), we induced SMA before experimentation to model the situation that exists when myofibroblasts are already induced. We used static force application protocols that are intentionally simplified (i.e., single amplitude, single stroke) because we had also used a very simple protocol to induce SMA in the first instance (i.e., culturing on rigid substrates). Thus, in contrast to the functional myocardium in which multiple cycles of complex and highly variable force magnitudes and directions are created, only a single, unidirectional continuous stretch was applied (15), and the stretch was maintained for relatively long periods of time. Nevertheless, removal of the force or prolonged force application was associated with a return to baseline SMA levels, and a repeated application 24 h later also induced a reduction of SMA content.

The force-induced reduction of SMA content was mediated partly by selective leakage of SMA from the cell. We found that SMA was selectively enriched in cell culture media after stretching, a phenomenon that has been previously reported (15). This leakage may account for the early loss of SMA (i.e., at 4 h); however, it is unlikely that leakage was the only reason: the reduction of SMA could be mediated also by inhibition of SMA translation or transcription. As treatment with cycloheximide abrogated the reduction of SMA, it is likely that the later stretch effect (i.e., after 4 h) was mediated at the level of SMA transcription and not translation. Indeed, Northern blots showed a reduction of SMA mRNA within 4 h after stretch. We surmise that the early loss of SMA was related to leakage and that the later effects were due to effects on transcription. In view of these considerations and because the MAP kinases have been implicated in force-induced activation of rat cardiac fibroblasts (16) and can regulate SMA promoter activity (11), we measured activation of mitogen-activated protein kinases in response to stretch. In contrast to passive biaxial stretch response using deformable substrates that showed activation of ERKs and JNK but not p38 (16), we found that only p38 was activated by perpendicular stretch through focal adhesions. Notably, maximal activation of p38 occurred at 30 min after stretch, well before the maximal reduction of SMA mRNA. This finding suggests that p38 may mediate the synthesis of another inhibitory protein, a notion that is consistent with the observation that cycloheximide blocked force-induced SMA inhibition, perhaps by blocking the synthesis of the transcriptional inhibitory protein. Consistent with the notion that p38 activation might be important in stretch-induced regulation of SMA, incubation of cells with the p38 inhibitor SB-203586 blocked the reduction of SMA. This result is similar to that obtained for the -skeletal actin gene when transfected in fibroblastic cells (15).

Cardiac fibroblasts synthesize and secrete angiotensin II (6) and also express angiotensin II receptors (5). The expression of these receptors is spatially associated with expression of SMA in myofibroblasts in vivo (22). Because angiotensin II induces SMA transcription (7) and because mechanical stretch induces autocrine release of angiotensin II (21), we considered that mechanical stretch and angiotensin II might antagonistically regulate SMA. Measurements of SMA after combined treatments with force and angiotensin II showed that the force-induced reduction of SMA was blocked by angiotensin II at 10 nM, despite p38 activation. Angiotensin II induces SMA through the serum response factor and the homeodomain transcription factor MHox (7), and, conceivably, the conflicting stimulatory and inhibitory signals induced by angiotensin II and force are mediated by different transcription factor binding elements on the SMA promoter. Notably, because treatment of stretched cardiac fibroblasts with the p38 kinase inhibitor SB-302586 caused a large increase of SMA, we surmise that stretch may release angiotensin II from cells (21). Without the p38-mediated inhibition of SMA, the released angiotensin II may have induced an autocrine increase of SMA (2). Indeed, the greater reduction of SMA in losartan-treated cells after stretch is consistent with a previous report that stretch may induce release of small amounts of angiotensin II from cardiac myocytes (21). Thus the autocrine release of angiotensin II may modulate the p38-dependent reduction of SMA induced by force.

SMA is not normally expressed in cardiac fibroblasts but is seen after induction of pressure overload (2), a time when angiotensin II is also released at very high levels from myocytes and other neighboring cells. We intentionally induced SMA in cardiac fibroblasts before application of force to model how force application in the absence of high levels of angiotensin II from neighboring cells may regulate SMA. In conclusion, this study demonstrates a direct regulatory effect of force on SMA content that antagonizes the induction of SMA by angiotensin II. Force-induced inhibition appears to be mediated in part through the p38 kinase and competes with angiotensin II to regulate SMA.