INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MECHANICAL LOAD is a major cause of cardiac growth and hypertrophy. However, whether the primary stimulus for growth is mechanical stress itself or an accompanying increase in neural or humoral factors is still controversial. Rat hearts hypertrophy when they are perfused with extracts of hypertrophying dog hearts (12, 13), indicating that diffusible factors of unknown cell origin may be responsible. Thus signals other than load itself might contribute to growth responses in overloaded tissue through paracrine actions. The intimate contact between microvascular endothelial cells (MVEC) and cardiac myocytes suggests that some of these regulatory and trophic factors are secreted by endothelial cells. A recent review (30) points out that the contractile function of myocytes can also be modulated by endothelial factors. Cardioactive substances released by endothelial cells include nitric oxide, endothelin-1, prostanoids, natriuretic peptides (30), and growth factors such as transforming growth factor-, basic fibroblast growth factor (bFGF), insulin-like growth factor (IGF)-1 (3), and IGF-2 (18, 19).

Although these and many other previous reports support the idea of MVEC as a regulator of adaptive and deficient myocardial growth, in vitro data that demonstrate growth effects of MVEC on adult myocytes are still lacking. Neither microvascular nor aortic endothelial cells induced any observable morphological change in cultured adult myocytes (7). Investigations of adult myocytes in coculture with cardiac MVEC did not show any marked changes in the myocyte phenotype (24), or cells began to increase in surface area after a 12-day coculture period (24). Paracrine-acting diffusible factors could not be identified in culture medium conditioned by nonmyocyte fractions that included endothelial cells (5). The reasons for the discrepancies between in vivo and in vitro expectations and results are difficult to unravel, but the fact that all myocytes in these adult cultures were exposed to serum certainly complicates the analysis of endothelial and myocyte interactions.

In this study, we show that in the absence of mechanical load heart MVEC produce trophic factors that are highly effective in the reorganization of the contractile apparatus, the induction of protein synthesis, and the increase of survival in adult cardiac myocytes. Moreover, we demonstrate that MVEC were also able to induce sarcomerogenesis and the reexpression of fetal -smooth muscle actin, which is usually not expressed in adult myocytes. We compared these results by showing essential differences with the effects of serum and IGF because these are produced by endothelial cells and have been demonstrated to induce protein synthesis in cultured adult rat cardiomyocytes (5, 9). Furthermore, we discuss the possible role of the endothelium in hypertrophy during mechanical strain.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Isolation and culture of adult cardiac myocytes. Ventricular cardiac myocytes of 3-mo-old male Wistar rats were isolated as described elsewhere (26). Briefly, hearts were perfused for 8 min with Ca²⁺-free Krebs-Henseleit bicarbonate buffer (KHB, pH 7.4) containing (in mM) 110 NaCl, 2.6 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 11 glucose, and 10 HEPES, which was gassed with 95% O₂-5% CO₂ at 37°C. These hearts were then perfused for 30 min with KHB solution containing 0.04% collagenase (Worthington) and 40 µM calcium. Ventricles were minced in the same collagenase solution containing 1.25% fatty acid-free BSA (Sigma). After two washing steps at 25 g for 3 min with increasing calcium concentrations of 0.2 and 0.5 mM, myocytes were layered over a 4% BSA gradient containing 1 mM calcium and centrifuged for 1 min at 15 g. The cell pellet was suspended in basic medium consisting of medium 199 (Sigma) with Earle's balanced salts without L-glutamine including 25 mM HEPES, 25 mM NaHCO₃, 100 IU/ml penicillin, and 100 µg/ml streptomycin and supplemented with 2 mM L-carnitine, 5 mM creatine, and 5 mM taurine (Sigma) as described elsewhere (26, 32). Myocytes were plated at a high density of 1.5 × 10⁴ cells/cm² on laminin (10 µg/ml; Sigma)-coated chamber slides (Nunc) for fluorescence microscopy analysis and on six-well culture dishes (Falcon) for Western blots, protein synthesis determination, and interference-contrast microscopy. A cell density of 0.5 × 10⁴ cells/cm² was used to determine the survival rate. Two hours after plating the medium was changed, and cells were allowed to recover for 1 day. Thereafter, basic medium, conditioned medium (CM), fetal calf serum (Sigma), and recombinant human growth factors (PromoCell) were added as indicated. Media were replaced every other day. All experiments were performed in duplicate and constantly treated with 10 µM 1-(-D-arabinofuranosyl)cytosine to prevent nonmyocyte growth.

Isolation and culture of MVEC and preparation of CM. Porcine cardiac MVEC and aortic endothelial cells were isolated as described previously (1). Pig MVEC instead of rat endothelial cells were used because of higher purity, lower passage number, the reduced number of macrovascular endothelial cells, and the exclusion of endocardial endothelial cells (1). Porcine brain MVEC were isolated and characterized as described previously (8). Confluent MVEC of the third and fourth passages in Nunc Cell Factories (Nunc) were washed three times with serum-free basic medium and then cultured for 2 days in the same medium to obtain CM. CM was diluted with basic medium or fetal calf serum (Sigma) in a proportion of 4:1 (vol/vol) before application because this dilution gave the highest activity.

Protein synthesis, protein, and DNA content. Myocytes were radiolabeled with 2.5 µCi/ml of [³H]phenylalanine (Amersham) during the last 5 h of culture. Cells were washed twice with Hanks' balanced salt solution, fixed for 1 h in 10% TCA, washed twice with 10% TCA and three times with 95% ethanol, air dried, and extracted in 0.3 M NaOH. Aliquots were removed for counting [³H]phenylalanine incorporation and for measuring total protein content by the detergent-compatible protein assay (Bio-Rad). For measuring DNA content, the pH was adjusted to 12.3 with 10 mM EDTA and determination was carried out as described elsewhere (33). Three representative experiments are pooled, and data are means ± SE. Statistical significance was assessed by a two-tailed, paired Student's t-test, with P values <0.05 taken as being statistically significant.

Electron microscopy and interference-contrast microscopy. Myocytes were fixed overnight with 4% glutaraldehyde in phosphate-buffered saline. Cells were then postfixed in 1% osmium tetroxide, dehydrated in ethanol, and embedded in Epon. Semithin (0.5 µm) sections were stained with toluidine blue and viewed in a Leica DM microscope. Ultrathin sections were stained with uranyl acetate and lead citrate and then viewed and photographically recorded in a Philips CM 10 electron microscope. For interference-contrast microscopy, cells were fixed in 4% formaldehyde.

Immunocytochemistry and fluorescence microscopy. Myocytes were fixed in 4% formaldehyde, permeabilized in 0.05% Triton X-100 for 20 min, and blocked in PBS containing 0.1% BSA (BSA-C, Aurion) for 10 min followed by the incubation of the fixed cells with the corresponding primary antibodies for 12 h at 4°C. F-actin was fluorescently stained using rhodamine-phalloidin (Molecular Probes). The other proteins were stained using indirect immunofluorescence with the following primary antibodies: monoclonal antibody against -actinin (Sigma), -smooth muscle actin (Sigma), or myomesin (kindly provided by Dr. H. M. Eppenberger). Myocytes were then incubated for 1 h at room temperature with biotinylated donkey anti-mouse IgG antibodies (Dianova) followed by an incubation with streptavidin-Cy2 (Biotrend). After fixation, cells were washed four times for 3 min between all steps. Preparations with primary antibody omitted served as a negative control. Myocytes were viewed in a Leica DM microscope and photographed in a Leica TCS laser confocal microscope with corresponding filters. Micrographs were taken with Kodak 100 ASA color slide film. All pictures presented here are taken from the same representative of three independent experiments.

Immunoblotting. Myocytes were washed twice with Hanks' balanced salt solution and then directly lysed in Laemmli gel running buffer (Bio-Rad). Protein content was determined as mentioned before. Protein (7.5 µg) of each sample was electrophoresed on a SDS-10% denaturating polyacrylamide gel. Transfers were carried out at 30 V overnight. Nitrocellulose membranes (Amersham) were incubated with monoclonal antibodies against -actinin (Sigma), pan-actin (Boehringer Mannheim), and -smooth muscle actin (Sigma). The corresponding bands were detected by the enhanced chemiluminescence method (Amersham).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of CM and serum on morphology of cardiomyocytes. We compared the effect of CM with that of serum because serum and serum growth factors were shown to have profound effects on morphology, growth, and survival of adult cardiac myocytes in culture (31, 32). Serum-stimulated adult cardiomyocytes underwent morphological changes that were most obvious in the spreading pattern (pattern of increase in surface area) from the rod-shaped phenotype (Fig. 1A). The spread configuration (Fig. 1B) was achieved either through a spherical intermediate or as the result of spreading directly from the rodlike shape (31, 32) and therefore displayed a heterogeneous morphology. Differences between different serum concentrations only affected the speed of these processes. The phenotype of myocytes exposed to CM was different from that of serum-stimulated groups, being characterized by the formation of multiple pseudopodia-like extensions as early as 36 h from the rod-shaped phenotype, which increased with time. Through the rapid establishment of new cell-cell contacts, myocytes showed the strong tendency to form tissuelike structures (Fig. 1C) within an 8-day culture period. After 10-12 days, a complete monolayer of adult cardiac myocytes was visible.

View larger version (108K): [in this window] [in a new window]   Fig. 1.   Phase-contrast micrographs of adult cultured cardiac myocytes (ARC). Freshly isolated (A; magnification ×60), 5% serum-stimulated (B; magnification ×120) and conditioned medium (CM)-stimulated (C; magnification ×120) ARC at day 8 are shown.

Effect of CM and serum on protein synthesis, protein accumulation, and nucleolar size. We examined the effect of CM and fetal calf serum on protein synthesis (Fig. 2A) by measuring the total [³H]phenylalanine incorporation-to-DNA ratio and their potency in protein accumulation (Fig. 2B) by determination of the protein-to-DNA ratio. The mass of DNA is considered to be a reliable measure of the relative cell number in terminally differentiated myocytes because there is no change in DNA content during growth, maturation, and aging (10). Values of unstimulated control cells at day 2 served as control values, and they were set at a value of 100%. CM caused significant increase in protein synthesis from 142% at day 2 to 600% at day 8. Serum (5%)-stimulated cells showed an induction of 147% on day 2 and increased to 289% on day 8. The rate of protein synthesis after 20% serum stimulation was similar (539%) to that after CM, and supplementation of serum to CM was additive by increasing synthesis to >1,340% (Fig. 2A). Unstimulated cells decreased steadily, both in protein synthesis to a level of 50% (Fig. 2A) as well as in the protein content to 60%, showing a clear atrophy (Fig. 2B). The elevation of protein synthesis correlated well with protein accumulation in serum-stimulated cells. Surprisingly, the protein-to-DNA ratio in CM-stimulated cells decreased to 87% of control, whereas on the other hand CM enhanced the accumulation of protein in serum stimulated cells from 130 to 170% (Fig. 2B). In comparison to 5% serum-treated control cells, CM-stimulated cells showed a twofold higher rate of protein synthesis (Fig. 2A) but lower protein content of 26% of control (Fig. 2B) at day 8. The level of protein synthesis, which is also regulated at the level of ribosome formation (for references see Ref. 15), correlated well with the size of the nucleoli. Nucleolus size appeared increased at day 8 in cell cultures stimulated with CM (Fig. 3A) compared with 5% serum-treated controls (Fig. 3B).

View larger version (51K): [in this window] [in a new window]   Fig. 2.   A: Time course of total [3H]phenylalanine incorporation in cultured myocytes harvested 2-8 days after various treatments. Values are expressed relative to DNA content in percentage of untreated control value at day 2. Control cells (Con) were kept in basic medium or were supplemented with 5% FCS (Con 5%). Myocytes were stimulated with 20% FCS (20%), CM, and CM + 20% FCS (CM+20). B: protein-to-DNA ratio of these samples is expressed in percentage of untreated control value at day 2.

View larger version (128K): [in this window] [in a new window]   Fig. 3.   Electron microscopy of ARC at day 8. A: typical cell nucleus with 2 nucleoli of a 5% FCS control myocyte. B: characteristic nucleus with a nucleolus of CM-stimulated cells. Bar, 2 µm; magnification, ×4,400.

Influence of other cardioactive substances on myocyte morphology and protein synthesis. Known growth factors showed minor effects on the cell morphology of adult cardiomyocytes. Minimal or no spreading activities (data not shown) were produced by various combinations and concentrations up to 100 ng/ml of acidic FGF, 100 ng/ml bFGF, 100 nM endothelin, 1 µM angiotensin, 10 ng/ml transforming growth factor-, 100 ng/ml leukemia-inhibitory factor, and 100 ng/ml interleukin 11. The latter substances are known to potently induce hypertrophy in neonatal myocytes, and they belong to the same cytokine family as cardiotrophin (25). IGF-1 at 500 ng/ml produced the strongest effect on the cell morphology of adult cardiomyocytes by inducing a limited flattening and a limited increase in the surface area (not shown). Antibodies directed against the above-mentioned growth factors did not influence the activity of CM.

View larger version (80K): [in this window] [in a new window]   Fig. 4.   A: total [3H]phenylalanine incorporation in cultured myocytes after various stimulations at day 8. Values are expressed relative to DNA content in percentage of untreated control value of day 8. Control cells were kept in basic medium. Myocytes were stimulated with CM, 500 ng/ml recombinant human (rh) insulin-like growth factor (IGF)-1, 500 ng/ml rhIGF-2, 500 ng/ml rhIGF-1 + 500 ng/ml rhIGF-2 (IGF-1+IGF-2), CM + 500 ng/ml rhIGF-1 (CM+IGF-1), and CM + 500 ng/ml rhIGF-2 (CM+IGF-2). B: protein-to-DNA ratio of these samples is expressed in percentage of untreated control value at day 8.

Effect of CM and serum on survival of cardiomyocytes. Myocytes plated at low density were used to exclude the influence of cell-cell interactions for analysis of the survival rate. One day after plating, the number of attached myocytes was determined and set as 100%. Nonstimulated control myocytes started to decrease steadily from day 4 to <25% at day 8 (Fig. 5). Most of the detached cells showed typical deterioration such as hypercontraction and blebbed cell remnants. At day 8 >90% of 20% serum-treated and 79% of CM-stimulated myocytes were still attached (Fig. 5), whereas under 5% serum treatment the number of attached cells decreased to 68%. Analysis by electron microscopy demonstrating intact sarcolemma and undamaged mitochondria and nuclear shapes (Fig. 3) served routinely as indicators for the viability of attached cells. Approximately 40% of CM-stimulated myocytes at low density and >60% at high density survived for >6 wk (not shown). Furthermore, at high cell density fractions of CM-stimulated myocytes were beating spontaneously from day 4. From day 7 to 10, >95% of myocytes were beating and beating was synchronous where cells formed contacts (not shown), indicating high viability. Serial dilutions of CM decreased spreading and the rate of surviving cells, indicating an absolute dependency on diffusible factors in CM.

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Time course of survival of cultured myocytes at low density after various treatments. Cell number was determined at indicated time points by counting. Values are expressed relative to percentage of cells attached 1 day (~2 × 103 cells/cm2) after plating (day 0). Untreated control cells (Con) were kept in basic medium throughout culture. Myocytes were stimulated with 5% (5%) or 20% FCS (20%), CM, and CM + 20% FCS (CM+20). Three representative experiments are pooled, and data are means ± SD. P values indicate statistical significance for differences between Con vs. CM-stimulated cells.

Effect of CM and serum on myofibrillogenesis. In freshly isolated rod-shaped myocytes, longitudinally oriented myofibrils are in register. During the process of spreading, typical sarcomeric structures disappear rapidly, and new filaments are preferentially formed in the perinuclear region of CM-treated cells (Fig. 3B). Sarcomeres were visualized by double staining for actin and for myomesin. Cells stimulated with 5% serum showed myomesin staining limited to the center in the hexagonal or the smoothly spread shape (Fig. 6A). Serum (20%)-treated cultures showed a larger hexagonal shape and a conspicuous increase in number of newly built sarcomeres. Myomesin staining was visible until the end of the cell border in a cross-striated pattern (Fig. 6B). The phenotype of myocytes exposed to CM was different from that of serum-stimulated cells. Multiple long extensions were filled with actin filaments, and fully developed sarcomeres were evident by the incorporation of myomesin (Fig. 6C). Supplementation of CM with 20% serum resulted in a hybrid phenotype, with a marked increase in cell surface area similar to serum induction and the formation of extensions as observed after CM stimulation (Fig. 6D). Furthermore, simultaneous stimulation of myocytes by CM and 20% serum potentiated their effects in the formation of newly built sarcomeres (Fig. 6D).

View larger version (59K): [in this window] [in a new window]   Fig. 6.   Fluorescence labeling for actin (red) and myomesin (green-yellow) of adult rat cardiomyocytes after various stimulations at day 8. Cardiomyocytes were treated with 5% FCS as control (A), 20% FCS (B), CM (C), or CM + 20% FCS (D).

Influence on -smooth muscle actin accumulation by CM, serum, and IGF. The accumulation of -smooth muscle actin, a fetal gene product, could be clearly detected by immunoblotting at day 6 after CM stimulation and increased further until day 8 (Fig. 7). In contrast, the amount of pan-actin was unchanged and also no increase in total -actinin accumulation (Fig. 7) could be detected. Serum alone was not sufficient to induce -smooth muscle actin accumulation in the 8-day period, but the combination of serum and CM enhanced -smooth muscle actin accumulation (Fig. 7). The effect of CM in the induction of -smooth muscle actin could be mimicked neither by any other cardioactive growth factors nor by various combinations that are known to play a role in the induction of hypertrophy of neonatal myocytes, including IGF (Fig. 7B) and FGF. The addition of IGF-1 or IGF-2 was additive to CM in protein synthesis and protein accumulation as described in Influence of other cardioactive substances on myocyte morphology and protein synthesis but did not enhance the accumulation of -smooth muscle actin in CM-stimulated cells. The data obtained by immunoblotting were verified by immunohistochemistry. -Smooth muscle actin-positive cells were hardly detectable in serum-stimulated cells (Fig. 8A), whereas approximately one-third of CM-stimulated myocytes expressed -smooth muscle actin (Fig. 8B). The combination of CM with 20% serum only slightly increased the number of -smooth muscle actin-positive cells; however, the relative amount of -smooth muscle actin per myocyte increased markedly (Fig. 8C). The number of -smooth muscle actin-positive cells could be increased up to >50% by concentrating CM five times using a filter with a molecular mass cutoff of 50 kDa. -Smooth muscle actin was hardly detectable in adult myocytes stimulated by using a 5× concentrate of molecules of <50 kDa. Furthermore, the activity of CM was sensitive to proteases (Pronase), ammonium sulfate precipitation, and suramin (50 µg/ml), a nonspecific receptor blocker of a variety of growth factors (23).

View larger version (91K): [in this window] [in a new window]   Fig. 7.   Immunoblots of proteins from cultured cardiomyocytes with antibodies against -actinin, -smooth muscle actin (-sm-actin), and pan-actin. Myocytes were untreated (Con) or treated with 20% FCS (20%), CM, CM + 20% FCS (CM + 20%), 500 ng/ml rhIGF-1, 500 ng/ml rhIGF-2, or 500 ng/ml rhIGF-1 + 500 ng/ml rhIGF-2. Control cells were kept in untreated basic medium. Extracts from day 6 (A) and day 8 (B) were electrophoresed and blotted as described. One representative out of three experiments is shown. Detection was performed by exposing film for ~2 s after pan-actin labeling and 1 min after -actinin and -smooth muscle actin simultaneous labeling.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Fluorescence labeling for actin (red) and -smooth muscle actin (green) of adult rat cardiomyocytes after various stimulations at day 8. Cardiomyocytes were treated with 20% FCS (A), CM (B), or CM +20% FCS (C).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cardiac MVEC attracted our interest because they may, by their intimate contact with cardiac myocytes, function as a source of trophic factors under physiological and pathophysiological conditions and may act also as a sensor of hemodynamic stress. Cultured MVEC are to some extent "stress activated," because they lack their natural environment and have to adapt to serum-free culture during the conditioning process. A recent report (15) demonstrates an early transcriptional activity of endocardial endothelial cells by pressure overload and adds to the view that endothelial cells may serve as a mechanotransducer. The paracrine effects of CM that we described in this report showed similarities with results obtained in in vivo models of mechanical stress, contributing to the view of the endothelium as a mechanotransducer of hemodynamic changes. We therefore discuss and compare our data summarized in Table 1 with criteria that appear to be analogous to mechanical load. These criteria were hypertrophic responses (17, 20, 21), survival (4, 21, 22), (re)organization of the contractile apparatus including sarcomerogenesis (2, 16, 29), and the reexpression of fetal genes (2, 16, 29) in cultured adult cardiomyocytes.

MVEC markedly stimulate protein synthesis in adult cardiac myocytes. Concomitantly with the reorganization of the contractile apparatus, a modulation of the hypertrophic response occurred during CM stimulation. CM increased the rate of protein synthesis up to 6-fold until day 8 and, in combination with serum, a 13-fold increase was detectable. Despite the stimulation of protein synthesis, CM failed to accumulate protein, indicating that the process of remodeling involved protein synthesis and degradation to an equal extent. On the other hand, CM enhanced protein accumulation in serum-stimulated cells. The modulatory activity of CM could be caused by the lack of some growth supplements (5) in the culture medium, such as insulin, which leads to an accelerated proteolysis. Therefore, the presence of degradation-inhibitory factors in serum might be the reason for the additive effect of CM in protein accumulation in serum-stimulated cells.

MVEC enhance survival of adult myocytes. The enhanced viability of CM-stimulated cardiomyocytes could also play a role during mechanical stress. Terminally differentiated myocytes have irreversibly lost the capacity to undergo mitosis and division. In all our cell cultures it was quite obvious that myocytes that did not spread on the substratum tended to lose viability. A previous review (27) describes that cell spreading controls proper growth and survival in vitro as well as in vivo. Furthermore, cell-cell contact of surviving myocytes, which is accelerated during CM stimulation, must be reestablished to maintain contractile function. Beating as an additional criterion for viability was observed for at least 3 wk. Usually, stable cultures are achieved by including serum at 10-20% concentration in culture media (31, 32). In our culture system, CM without the addition of any growth supplements or serum was able to produce stable cultures judged by all methods presented here. Our data provide evidence for the existence of antiapoptotic factors and also for substances protective against necrosis in CM.

MVEC induce myofibrillogenesis in adult cardiac myocytes. The most surprising observation was that MVEC were able to induce myofibrillogenesis. The myofibrillar apparatus of myocytes degenerates during CM stimulation, and new myofibrillar structures grow out from the perinuclear region (Fig. 3B) into the cell periphery. The appearance of myomesin demonstrates that some myofibrils were already transformed into sarcomeres (Fig. 6, C and D) after 8 days, because myomesin appears late during myofibrillogenesis and is a characteristic marker for mature sarcomeres (28). Our observations demonstrate that MVEC not only enhance the "loss of the preexisting myofibrillar structure" but also induce a "gain of myofibrillar structure" that results in a remodeling of the adult cardiac myocyte with maintained function.

MVEC induce reexpression of fetal -smooth muscle actin in adult cardiac myocytes. The speed of the de novo formation of filaments in the pseudopodia is reflected in the dramatic speed of resynthesis of the fetal -smooth muscle actin that was easily detectable at day 6. It has been postulated that stress fiberlike structures containing -smooth muscle actin serve as a scaffold for the formation of new myofibrils in long-term cultures of cardiac myocytes (6, 14). A reactivation of this gene in adult rat hearts by mechanical load possessing features in common with growth factor signal transduction was also reported, and -smooth muscle actin was described as a molecular marker of the presence and extent of pressure-overload hypertrophy (2). Moreover, it is generally accepted that cardiac-specific gene expression in overloaded myocardium recapitulates a fetal program (2, 16, 29). Therefore, the rapid resynthesis of the fetal -smooth muscle actin suggests that MVEC are involved in the transduction of growth signals of cardiac overload.

Effects of known trophic (growth) factors on adult cardiac myocytes. The available evidence indicates that the trophic factor(s) in CM showed an activity different from all known substances. Under a variety of growth factors tested, IGF-1 and IGF-2 showed the strongest effect on protein synthesis and to some extent on the increase in surface area in the 8-day period. IGF in combination with CM were additive in regard to protein synthesis, protein content, and spreading. However, they neither induced -smooth muscle actin synthesis when applied solely nor enhanced -smooth muscle actin synthesis during CM stimulation, indicating that these growth factors may play a minor role under the criteria presented here. Moreover, IGF-1 downregulated -smooth muscle actin (6) in cultured adult cardiomyocytes. A recent report (14) showed that under serum conditions bFGF enhanced the expression of -smooth muscle actin in adult cultured myocytes. The same group (11) demonstrated that the effect of serum was reduced after triiodothyronine (T₃) stripping and that T₃-containing serum was permissive for the action of bFGF. We failed to detect any -smooth muscle actin expression induced by bFGF alone or in combination with various growth supplements when serum was absent. Furthermore, CM that was stripped of proteins >50 kDa was completely without effect on -smooth muscle actin expression, whereas fractions of protein >50 kDa enhanced the number of -smooth muscle actin positive cells to >50%. In addition, CM showed a fivefold higher rate of protein synthesis in comparison to a high dose of 50 ng of bFGF.