INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

DEPENDING ON THE NATURE of the initiating stimulus, morphologically distinct forms of cardiac hypertrophy have been identified. Concentric cardiac hypertrophy, which develops in response to chronic systolic overload (e.g., hypertension), is characterized by a parallel pattern of sarcomere replication, leading to an increase in myocyte cell width (2, 11). Morphologically, this cellular adaptation causes ventricular wall thickening and a modest reduction in chamber radius (2 11). By contrast, a chronic diastolic overload (e.g., aortic regurgitation or secondary to arteriovenous shunting) promotes a series pattern of sarcomere replication, leading to an increase in myocyte cell length (2, 11). Morphologically, this cellular adaptation causes a disproportionate increase in chamber radius with relatively little increase in wall thickness, leading to an eccentric pattern of cardiac hypertrophy (2, 11). It has been shown that chronic systolic and diastolic overload inevitably leads to contractile dysfunction (7, 17). By contrast, the intermittent diastolic load associated with a physiological stimulus such as exercise training leads to an eccentric pattern of cardiac hypertrophy with enhanced cardiac function (1, 2, 16, 19).

At the molecular level, it has long been recognized that the development of concentric or eccentric pathological cardiac hypertrophy in the adult rat was associated with the recapitulation of the cardiac fetal gene prepro-atrial natriuretic peptide (prepro-ANP) (5, 8). By contrast, a modest increase or no change in ventricular prepro-ANP was observed in the hypertrophied heart of exercise-trained animals (e.g., swimming, treadmill running) (4, 10, 17). Thus the disparate pattern of prepro-ANP mRNA regulation appears to represent at least a molecular marker differentiating physiological from pathological cardiac hypertrophy.

Among the plethora of identified hypertrophic stimuli, recent studies have focused on the potential role of peptide growth factors. In vitro studies have reported that the exogenous administration of peptide growth factors promoted cardiac myocyte hypertrophy and the expression of cardiac fetal genes (21). Moreover, a dynamic regulation of several peptide growth factor mRNAs [e.g., transforming growth factor (TGF)-₁ and -₃, acidic and basic fibroblast growth factor, and insulin-like growth factor] has been observed in the myocardium of various rat models of cardiac hypertrophy, and their increased expression has been documented in isolated cardiac myocytes in response to putative hypertrophic stimuli (e.g., ₁-adrenergic agonists, angiotensin II) (5, 23, 25, 27). These data suggest that peptide growth factors could represent an important autocrine/paracrine mechanism influencing myocardial remodeling associated with pathological cardiac hypertrophy. By contrast, it remains to be shown whether peptide growth factor mRNAs are induced during the development of physiological hypertrophy.

Although the recapitulation of prepro-ANP and the induction of peptide growth factor mRNAs are prevalent molecular events among diverse rat models of pathological cardiac hypertrophy, the pattern of gene expression associated with physiological cardiac hypertrophy has not been fully elucidated. Moreover, it remains to be established whether the molecular phenotype commonly associated with pathological cardiac hypertrophy is exclusively dependent on the nature of the stimulus or reflects a conserved feature of cardiac hypertrophy regardless of the nature of the initiating stimulus (e.g., physiological vs. pathological). To address this issue, the expression of the cardiac fetal gene prepro-ANP and the peptide growth factors TGF-₁ and TGF-₃ were examined in the left ventricle of a voluntary exercise rat model of physiological cardiac hypertrophy.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal models. Experiments were performed in accordance with the principles of the Canadian Council on Animal Care and approved by the Animal Ethics Committee of the Université de Montréal. Female Sprague-Dawley rats weighing 150-175 g (Charles River Canada, St. Constant, PQ, Canada) were housed individually in voluntary exercise wheels (10 × 21 cm) and permitted to run freely. The distance ran was measured by a magnetic counter connected to a personal computer. Sedentary rats were also individually housed in exercise wheels, but the wheel was prevented from rotating. After 3 or 6 wk of voluntary exercise, the rats were anesthetized with pentobarbital sodium (60 mg/kg), and the heart was rapidly excised, separated into the right and left ventricles and atria, weighed, and immersed in liquid nitrogen. Subsequently, the triceps muscles were excised, weighed, and immersed in liquid nitrogen. Tissue samples were subsequently stored at 80°C. In a second series of experiments, female Sprague-Dawley rats weighing 180-200 g received subcutaneous infusion of drugs with Alzet osmotic minipumps (model 1003D, Alza, Palo Alto, CA) implanted subcutaneously on the back slightly posterior to the scapulae. Minipumps delivered l-norepinephrine in ascorbic acid (2 mg/ml) at a mean flow rate of 1 µl/h at a rate of 500 µg · kg¹ · h¹ for 36-48 h. Sham rats were implanted with minipumps containing ascorbic acid. Previous studies demonstrated that norepinephrine infusion promotes systemic hypertension (14). After the experimental protocol, rats were killed as described above, and the heart was rapidly excised, separated into the right and left ventricles and atria, weighed, and immersed in liquid nitrogen.

Cytochrome c oxidase assay. Cytochrome c oxidase activity in the left and right triceps muscles was measured by a spectrophotometric assay. Each muscle was homogenized separately and processed according to the method of Smith (24).

Measurement of cardiac catecholamine level and hydroxyproline content. After the homogenization of left ventricular tissue as described by Eldrup and Richter (9), catecholamines were extracted using alumina adsorption and separated by high-performance liquid chromatography as previously described (13). Hydroxyproline content of left ventricular tissue was measured as previously described (18, 28). Collagen content was estimated by multiplying the hydroxyproline content by a factor of 8.2 (18).

Northern hybridization. Total RNA was isolated by a modification of the technique of Chomczynski and Sacchi (6). A 0.7-kb fragment of rat prepro-ANP (courtesy of Dr. M. Boluyt), a 0.985-kb fragment of rat TGF-₁ [American Type Culture Collection (ATCC), Rockville, MD], a 1.2-kb fragment of mouse TGF-₃ (ATCC), a 0.6-kb fragment of rat fibronectin (courtesy of Dr. R. O. Hynes), a 0.3-kb fragment of type I preprocollagen-₁ (ATCC), and a 1.2-kb fragment of rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH; ATCC) were labeled with [³²P]dCTP (NEN) to a specific activity of 1-2 × 10⁶ counts · min¹ · ng¹ cDNA by the random hexamer (Pharmacia) priming method and hybridized to nylon membranes (Dupont) for 18-24 h at 42°C as previously described (5). The filters exposed to the cDNA probes were washed twice (15 min at room temperature) with 300 mmol/l NaCl-30 mmol/l trisodium citrate and 0.1% SDS and twice (15 min at 45°C) with 30 mmol/l NaCl-3 mmol/l trisodium citrate and 0.1% SDS. Nylon membranes were subsequently exposed to Kodak XAR film with an intensifying screen at 70°C, and films were scanned with a laser densitometer (Chemilmager 4000 I version 4.04 software, Alphan Innotech). All levels of mRNA are normalized to the level of GAPDH mRNA.

Statistical methods. Values are means ± SE. Morphological heart measurements were assessed by ANOVA, and a significant value was determined by the Newman-Keuls test. Changes in mRNA expression between sedentary and exercise-trained rats at each time point were assessed by a Student's unpaired t-test (2-tailed). P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of voluntary exercise on cardiac hypertrophy, skeletal muscle adaptation, and ventricular catecholamine levels. Female rats housed individually in exercise wheels ran a total distance of 126 ± 5 (n = 11) and 508 ± 29 km (n = 13) at 3 and 6 wk, respectively (Table 1). The training index, as reflected by the kilometers ran per day during the last week of exercise, was 17 ± 1 and 11 ± 0.8 km/day at 3 and 6 wk of exercise, respectively. After 3 and 6 wk of voluntary exercise, left ventricular weight-to-body weight ratio increased by 20 and 22% (Table 1), respectively, compared with the sedentary rats. By contrast, a nonsignificant increase (16%) in right ventricular weight-to-body weight ratio was observed in the 3-wk exercise-trained rats, whereas a significant increase (27%) was observed in the 6-wk exercise-trained rats compared with sedentary rats (Table 1). Interestingly, the training index calculated during the last week of voluntary exercise in the 3-wk exercise-trained rats did not correlate with the magnitude of left (r = 0.47) or right ventricular hypertrophy (r = 0.0015). Likewise, at 6 wk, the training index did not correlate with the induction of left (r = 0.22) or right ventricular hypertrophy (r = 0.35).

Pattern of mRNA expression in the hypertrophied left ventricle after voluntary exercise. The progression of physiological cardiac hypertrophy was not associated with an increased mRNA expression of the extracellular matrix proteins fibronectin or preprocollagen-₁ at 3 or 6 wk of voluntary exercise compared with sedentary rats (Fig. 1, Table 2). Consistent with the latter data, the collagen concentration, as estimated by the hydroxyproline content, was similar in the left ventricle of rats voluntarily exercised for 3 wk (21 ± 1.9 mg hydroxyproline/g left ventricular tissue, n = 7, P = 0.07 vs. sedentary) and sedentary rats (28 ± 3.9 mg/g, n = 5). Despite the development of left ventricular hypertrophy, a modest nonsignificant increase in the steady-state mRNA level of the cardiac fetal gene prepro-ANP was observed at 3 and 6 wk of voluntary exercise compared with the sedentary rats (Figs. 2 and 3, Table 2). By contrast, in response to 3 wk of voluntary exercise, the steady-state mRNA level of TGF-₁ was significantly increased (n = 9, P < 0.001) compared with sedentary rats (Figs. 1 and 2, Table 2). However, the induction of left ventricular TGF-₁ mRNA in the rats exercised for 3 wk weakly correlated with the training index (in km/day; r = 0.51, P = 0.15). By contrast, a significant correlation was observed between left ventricular TGF-₁ mRNA expression and the development of left ventricular hypertrophy (left ventricular weight-to-body weight ratio; r = 0.76, P = 0.01). Indeed, the exercise-trained rats, which developed a modest increase in left ventricular hypertrophy (11 ± 2% increase vs. sedentary, n = 3), were associated with a 51 ± 9% induction of TGF-₁ mRNA (n = 3, P < 0.01 vs. sedentary). By contrast, the voluntarily exercised rats, which exhibited a 24 ± 3% increase in left ventricular weight-to-body weight ratio compared with sedentary rats, were associated with a 135 ± 24% increase (n = 6, P < 0.001 vs. sedentary) in the steady-state mRNA level of TGF-₁ in the left ventricle. Although the training index of exercise-trained rats with a modest hypertrophic response (16 ± 1.7 km/day) was slightly lower than that of exercise-trained rats with a significantly greater left ventricular weight-to-body weight ratio (18 ± 1.3 km/day), this difference was not statistically different (P = 0.35). Likewise, at the end of 6 wk of voluntary exercise, TGF-₁ mRNA level remained elevated and was significantly higher (n = 7, P < 0.01) than in the sedentary rats (Fig. 3, Table 2). Moreover, as observed at 3 wk, a correlation between the induction of left ventricular TGF-₁ mRNA and the development of left ventricular hypertrophy (left ventricular weight-to-body weight ratio; r = 0.67, P = 0.07) was observed in rats that voluntarily exercised for 6 wk. By contrast, regardless of the magnitude of left ventricular hypertrophy at 3 or 6 wk of voluntary exercise, the steady-state mRNA level of a second member of the TGF family, TGF-₃, was not significantly increased (Fig. 3, Table 2).

View larger version (67K): [in this window] [in a new window]   Fig. 1.   A representative autoradiograph of a Northern hybridization of total RNA from the left ventricle of sedentary (Sed) and exercise-trained (Exe) female rats showing changes in left ventricular preprocollagen-1, fibronectin, and transforming growth factor (TGF)-1 mRNA expression after 3 wk of voluntary exercise. Changes in mRNA levels are normalized to the level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA.

View larger version (52K): [in this window] [in a new window]   Fig. 2.   Autoradiograph of a Northern hybridization of total RNA from the left ventricle of sedentary and exercise-trained female rats showing changes in left ventricular prepro-atrial natriuretic peptide (prepro-ANP) and TGF-1 mRNA expression after 3 wk of voluntary exercise. Changes in mRNA levels are normalized to the level of GAPDH mRNA.

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Autoradiograph of a Northern hybridization of total RNA from the left ventricle of sedentary and exercise-trained female rats showing changes in left ventricular prepro-ANP, TGF-1, and TGF-3 mRNA expression after 6 wk of voluntary exercise. Changes in mRNA levels are normalized to the level of GAPDH mRNA.

Norepinephrine infusion induces a pathological phenotype of cardiac hypertrophy. Previous studies have demonstrated that norepinephrine infusion in the rat promoted cardiac hypertrophy and the induction of TGF-₁ mRNA. Consistent with these data, 36-48 h of infusion of norepinephrine (500 µg · kg¹ · h¹) in adult female Sprague-Dawley rats resulted in a 34% increase in left ventricular weight-to-body weight ratio (Table 3). Concomitant with left ventricular hypertrophy, the steady-state mRNA level of TGF-₁ was increased (2.3 ± 0.12-fold, n = 2) compared with sham rats (Fig. 4). By contrast, a modest induction of TGF-₃ mRNA level (43 ± 11% increase vs. sham, n = 2) was observed in the left ventricle of norepinephrine-treated rats (Fig. 4). Consistent with a molecular phenotype of pathological cardiac hypertrophy, the steady-state mRNA levels of prepro-ANP (12 ± 2-fold, n = 5; Fig. 4) and preprocollagen-₁ (29 ± 7-fold, n = 3) were significantly increased (P < 0.01 vs. sham) in the left ventricle of the norepinephrine-treated rats compared with sham rats.

View larger version (71K): [in this window] [in a new window]   Fig. 4.   Autoradiograph of a Northern hybridization of total RNA from the left ventricle of sham and norepinephrine-treated (NE) female rats showing changes in left ventricular prepro-ANP, TGF-1, and TGF-3 mRNA expression after norepinephrine infusion. Changes in mRNA levels are normalized to the level of GAPDH mRNA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Treadmill exercise and swimming represent the prevalent forms of physical training, and depending on the training regimen and gender employed, an inconsistent hypertrophic response (e.g., 0-30%) has been documented in the rat heart (1, 3, 4, 10, 16, 17, 20). By contrast, voluntary exercise permits the animal to exercise in a pattern closer to its normal activity than either swimming or treadmill exercise. In the present study, adult female Sprague-Dawley rats permitted to run freely in exercise wheels were associated with biventricular hypertrophy, normal left ventricular catecholamine content, and skeletal muscle adaptation, as reflected by the significant increase in cytochrome c oxidase activity in the triceps (4, 12, 15, 22). However, a dissociation between the hypertrophic response of the heart and the training index was observed at 3 and 6 wk of voluntary exercise. Thus, although exercise represents the initiating stimulus, the development of cardiac hypertrophy, at least in the rat model of voluntary exercise, did not correlate with the degree of training.

The increased expression of ventricular prepro-ANP mRNA and the extracellular matrix protein collagen represent conserved features of several distinct models of pathological cardiac hypertrophy (5, 17, 27, 28). In the present study, and consistent with previous reports, the increase in systolic load imposed on the heart after norepinephrine infusion of female rats induced cardiac hypertrophy and the upregulation of left ventricular prepro-ANP and preprocollagen-₁ mRNA levels (14, 27). In response to physical training (e.g., swimming and treadmill exercise), previous studies found no change or a modest increase in prepro-ANP mRNA in the myocardium (4, 17). In the present study, left ventricular hypertrophy was associated with a modest nonsignificant increase in prepro-ANP mRNA after 3 wk of voluntary exercise, whereas no difference was observed at 6 wk of exercise compared with sedentary rats. Likewise, the mRNA levels of fibronectin and preprocollagen-₁ and total collagen concentration in the left ventricle were similar between exercise-trained and sedentary rats. Thus the disparate pattern of ventricular prepro-ANP and extracellular matrix protein mRNA observed between exercise- and norepinephrine-induced cardiac hypertrophy reaffirms that this pattern of gene expression distinguishes between physiological and pathological hypertrophic stimuli. Moreover, these data further support a heterogeneity in the signaling events coupled to the development of physiological and pathological cardiac hypertrophy in the rat.

The increased expression of ventricular TGF-₁ mRNA in numerous models of pathological cardiac hypertrophy, as well as in cardiac cells in response to putative hypertrophic factors, suggests that this peptide growth factor may play a critical role in myocardial remodeling (5, 21, 23, 25-27). In the norepinephrine infusion model of cardiac hypertrophy, TGF-₁ mRNA levels were increased in the left ventricle, as previously demonstrated (25). Likewise, the steady-state levels of TGF-₁ mRNA were increased at 3 wk and remained elevated at 6 wk in voluntarily exercised rats. Collectively, these data suggest that the induction of TGF-₁ mRNA may represent a conserved feature of cardiac hypertrophy, regardless of the initiating stimulus. Interestingly, the expression of TGF-₁ mRNA did not correlate with the training index of voluntarily exercised rats, but rather a correlation was observed with the magnitude of left ventricular hypertrophy. Thus this latter observation suggests that the hypertrophic growth response associated with exercise, rather than the degree of exercise per se, represents the primary stimulus of TGF-₁ mRNA expression in the rat model of voluntary exercise. In contrast to TGF-₁, TGF-₃ mRNA levels were not significantly increased in the hypertrophied left ventricle after 3 or 6 wk of voluntary exercise. Moreover, a modest induction in the steady-state mRNA level of TGF-₃ was observed in the left ventricle of norepinephrine-infused rats. Indeed, important quantitative differences in ventricular TGF-₁ and TGF-₃ mRNA expression have been previously observed after chronic pressure overload (5). Moreover, in response to chronic volume overload, the increase in ventricular TGF-₁ mRNA was associated with a concomitant decrease in the mRNA levels of TGF-₃ (5). Collectively, these data support a heterogeneity in the signaling pathways coupled to the regulation of ventricular TGF-₁ and TGF-₃ mRNA.

In summary, the present study has demonstrated that voluntary exercise promoted cardiac hypertrophy in female rats and the upregulation of ventricular TGF-₁ mRNA. By contrast, the hypertrophied left ventricle of the rats that voluntarily exercised was not associated with an increased expression of prepro-ANP, the peptide growth factor TGF-₃, or extracellular matrix protein mRNAs. These latter data highlight a molecular phenotype that distinguishes physiological and pathological cardiac hypertrophy. The underlying mechanism(s) responsible for the heterogeneous pattern of gene expression among diverse rat models of cardiac hypertrophy remains undefined. By contrast, the upregulation of ventricular TGF-₁ mRNA appears to represent a conserved molecular event of cardiac hypertrophy. This finding suggests that TGF-₁ may constitute an essential feature of hypertrophic growth, regardless of the nature of the initiating stimulus or the subsequent pattern of cardiac growth.