INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

APOPTOSIS, OR PROGRAMMED cell death, is observed in experimental heart failure due to multiple intracoronary embolization (26), tachycardia (18), postmyocardial infarction (22), or left ventricular hypertrophy (1, 10, 17) and is seen in human hearts with myocardial infarction (13, 24), ischemic cardiomyopathy (20, 21), and idiopathic dilated cardiomyopathy (20, 21). These findings support a role for apoptosis in the pathogenesis of heart failure.

Heightened adrenergic drive, as well as activation of the renin-angiotensin system, is the hallmark of the neurohumoral changes in heart failure, and the extent of elevation of plasma norepinephrine correlates with the severity of the left ventricular dysfunction (28) and with mortality (4). In addition, chronic adrenergic activation is considered to be deleterious in heart failure because of its hemodynamic consequences and direct cardiocyte toxicity.

The purpose of this study was to determine whether continuous -adrenergic stimulation alone can induce apoptosis in rat myocardium in vivo without other intercurrent alterations in systolic stress and, if so, whether pathologically hypertrophied cardiocytes have a greater susceptibility to catecholamine-induced apoptosis.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Male Wistar rats weighing 250-330 g were used for this study. The animals were maintained in accordance with the guidelines of the National Research Council and the standing committee on Animal Care of the Albert Einstein College of Medicine. Catecholamine was infused continuously, as previously reported (2, 11). Briefly, Alzet minipumps (model 2001, Alza Pharmaceutical, Palo Alto, CA) were implanted subcutaneously in the neck, under methoxyflurane anesthesia. The minipump was filled with l-isoproterenol HCl dissolved in acidified isotonic saline (0.001 N HCl) or acidified isotonic saline alone (vehicle). The pump rate was 1 µl/h, and the drug was delivered at 40 or 400 µg · kg¹ · h¹. The higher dose has been previously shown to significantly increase heart rate without any change in systolic blood pressure and to cause significant desensitization to catecholamines after 2 h of infusion (2, 11). In initial studies, control and isoproterenol treatment was for 12 h, 24 h, or 7 days. On the basis of the results of this time-course study, the 24-h infusion was chosen for the subsequent pacing and hypertrophy protocols. Heart rate and systolic blood pressure were measured by the tail-cuff technique under light methoxyflurane anesthesia before implantation, 24 h, 4 days, and 7 days after implantation, and at the time the animals were killed.

Pacing protocol. To determine whether the effects of isoproterenol on cardiac apoptosis were related to elevated heart rate alone, a separate cohort of rats was subjected to cardiac pacing. Rats were anesthetized with chloral hydrate (400 mg/kg ip) and ventilated with positive pressure. A left thoracotomy was performed, and Biomed wires (Cooner Wire, Chatsworth, CA) were sutured on the epicardial surface of the apex and base of the heart with 9-0 sutures (Ethicon, Somerville, NJ). Wires were connected to a stimulator (model SD9, Grass Medical Instruments, Quincy, MA), and the heart was paced at 500 beats/min. This heart rate was chosen to approximate the mean of isoproterenol-induced heart rates [517 ± 21 (SE) beats/min, n = 4]. The pacing was confirmed by visual inspection of the heart in the open chest when the stimulator was turned on. The chest was closed, and wires were passed through the skin, tunneled to the neck, and exteriorized. The wires were protected by a custom-made rat vest jacket and a flexible coiled metal spring tube. Rats were able to freely ambulate in the cage and were paced continuously for 24 h.

Aortic banding. Descending aortic coarctation was created using techniques that are standard in our laboratory. Briefly, animals were anesthetized with methoxyflurane, the abdominal aorta just above the renal artery bifurcation was isolated, and a 20-gauge needle was placed along the side of the abdominal aorta and tied with a 5-0 silk thread. After the suture was secured, the needle was removed, creating partial banding of the abdominal aorta. Rats were cage confined for 2 wk to allow ventricular hypertrophy to develop, then they were treated with vehicle or with 400 µg · kg¹ · h¹ isoproterenol for 24 h via Alzet pump, as described above. Additional age- and weight-matched sham-operated rats were treated for 24 h with vehicle only or 400 µg · kg¹ · h¹ isoproterenol and served as controls.

Staining and morphometric analysis. Hearts were excised quickly after animals were anesthetized with methoxyflurane. The aorta was cannulated and perfused with 10 ml of chilled normal saline, then with 10 ml of chilled 10% neutral buffered Formalin. The heart without atria was weighed and then sliced into three sections perpendicular to the long axis of the ventricle. Sections were stored in 10% neutral buffered Formalin for 24 h and then embedded in paraffin. Slide sections (5 µm) from midventricular level were stained by in situ terminal deoxynucleotide transferase-mediated dUTP nick end labeling (TUNEL) using an in situ apoptosis detection kit [TACS 2TdT (TBL), Trevigen, Gaithersburg, MD]. The details of TUNEL assay have been described elsewhere (7). Positive-stained controls were prepared by incubating serial sections of each paraffin block with 10 U/ml DNase I for 20 min at 37°C before treatment with terminal transferase. Negative controls were prepared by staining serial slides without terminal deoxynucleotide transferase. Red counterstain C (Trevigen) was used to stain cytoplasm. TUNEL-positive nuclei stained blue in this assay. Serial sections were also used for hematoxylin-eosin staining for standard histological evaluation. The entire stained section was scanned at ×100 under light microscopy. When TUNEL-positive cells were detected, magnification was changed to ×400 to assess whether staining occurred in cardiocytes or noncardiocytes. Only TUNEL-positive cardiocytes were counted in each slide. We classified cells as TUNEL-positive cardiocytes according to the following criterion: nuclei were inside cardiocyte contour. Any nuclei that were ambiguous were not counted. Round homogenous dark dots (considered to be artifact) and stained cell debris were carefully eliminated from counting. Brownish nuclei-like deposits, considered to be Formalin crystals, were not counted. To assess whether tissue damage was present in locations of TUNEL-positive cardiocytes, the corresponding hematoxylin-eosin-stained serial sections were examined. The area of the heart in the slide preparation was measured by a modified cut-and-weigh technique (23). Briefly, the area of the heart slice was enlarged to ×64 by a photo enlarger. The contour of the heart section was traced on high-quality paper. The paper was cut as traced and weighed. Areas were computed as the ratio of the weight of the paper to the weight of a unit of known area. The numbers of apoptotic cardiocytes were normalized to areas of tissue (cm²) and per 10,000 cardiocyte nuclei estimated by the morphometric method. From the positive control of each section, 12 light-microscopic fields (×400) were chosen to count the number of cardiocyte nuclei. Endocardial, midmyocardial, and epicardial areas were chosen from septum and lateral, anterior, and posterior walls of the left ventricle. The number of cardiocyte nuclei in each microscopic field was counted under a grid covering a total area of 0.0625 mm². Only nuclei that were completely contained within cardiocyte contours were counted. The number of cardiocyte nuclei in 12 fields was averaged. Estimated total cardiocyte nuclei per slide were calculated as the area of tissue of each slide (mm²) × [the average number of cardiocyte nuclei in 12 light-microscopic fields/0.0625 (mm²)]. The number of TUNEL-positive cells per 10,000 cardiocyte nuclei in each slide was calculated as total positive cells in each slide divided by estimated total cardiocyte nuclei in each slide multiplied by 10,000. The morphometry results of two sections from each heart that were >3 mm apart were averaged for statistical comparisons.

Ligase reaction. To verify the results of TUNEL assay, midventricular sections from 9 rats (3 from 12-h vehicle-treated rats, 4 from 12-h isoproterenol-treated rats, 2 from aortic-banded rats) were subjected to this in situ apoptosis staining. We used the technique published by Didenko and Hornby (5) with minor modifications. Probes for this reaction were prepared by the following methods. Briefly, a 168-bp double-stranded DNA fragment was prepared using primers 5'-AATTAACCCTCACTAAAGGG-3' and 5'-GCAATTAACCCTCACTAAAG-3' complementary to plasmid pBluescript II KS (Stratagene, La Jolla, CA). To prepare fragments by PCR, the following reaction mix was set up: 100 µl of 10 mM Tris · HCl, pH 8.3, 50 mM KCl, 1.5 mM MgCl₂, 70 µM digoxigenin-11-dUTP (Boehringer Mannheim, Indianapolis, IN), 130 µM dTTP, 200 µM dATP, 200 µM dGTP, 200 µM dCTP (other nucleotides from Sigma Chemical, St. Louis, MO), 200 nM each primer, and 10 ng of plasmid. Taq polymerase (5 U; PE Applied Biosystems, Foster City, CA) was added. After the mixture was heated to 94°C for 5 min, PCR was performed with 35 cycles of 1 min at 94°C, 1 min at 50°C, and 2 min at 74°C, with the final cycle having an extension time of 4 min. The resultant probe was purified by spin columns using high pure PCR products purification kit (Boehringer Mannheim) according to the company's instruction. The Formalin-embedded sections were deparaffinized, rehydrated, and treated with proteinase K, as previously described (5). Sections were incubated with a mixture of 50 mM Tris · HCl, pH 7.8, 10 mM MgCl₂, 10 mM dithiothreitol, 1 mM ATP, and 50 µg/ml BSA, with digoxigenin-DNA fragment at 3 µg/ml and DNA T4 ligase (New England BioLabs, Beverly, MA) at 32 U/ml in a humidified box for 1 h. Then sections were preblocked and incubated with sheep anti-digoxigenin Fab fragment-alkaline phosphatase conjugate in blocking buffer (Boehringer Mannheim), as previously described (5). For color development, sections were then placed in the solution recommended by the manufacturer (0.1 M Tris, pH 9.5, 50 mM MgSO₄, 0.4 mg/ml nitro blue tetrazolium, 0.19 mg/ml 5-bromo-4-chloro-3-indolyl phosphate), and color development was monitored under the microscope. Positive and negative controls were stained simultaneously. The positive controls were treated with DNase I as described above, and negative controls were stained without DNA T4 ligase. The reaction was terminated by washing sections in water when color of nuclei of positive controls developed. The wet sections were mounted by Fluoromount-G (Southern Biotechnology Associates, Birmingham, AL) and covered by cover glasses without counterstain. Morphometric analyses were performed on these slides as described above.

Statistics. Values are means ± SE. The statistical significance of normalized positive cells, body weight, heart weight, and hemodynamic variables between the control and isoproterenol-treated group was evaluated using Student's nonpaired t-test. The statistical differences of apoptotic cardiocyte numbers, body weight, heart weight, and hemodynamic variables between the different time periods in the same group and between more than three groups were tested by ANOVA with Bonferroni's post hoc test. P < 0.05 was considered to be significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

General characteristics of animal models. All the control animals survived the vehicle treatment. However, two of the six animals in the 24-h isoproterenol-treated group and two of the six animals in the 7-day isoproterenol-treated group died before the animals were to be killed.

1

1

1

1

1

1

1

1

Histology of the heart. Hearts from rats treated with 400 µg · kg¹ · h¹ isoproterenol showed diffuse interstitial edema that was more prominent in the endocardial layer at 12 h and interstitial edema with infiltration of inflammatory cells including mast cells in the perivascular areas and occasionally at cardiocyte layers at 24 h of treatment. At 7 days, focal fibrosis associated with cell necrosis was seen predominantly in endocardial layers but was also scattered throughout the midmyocardial and epicardial layers with significant thickening of the arterial walls. No remarkable histological changes were seen in hearts treated with 40 µg · kg¹ · h¹ isoproterenol and in paced hearts. No qualitative increase in fibrosis was seen in control or aortic-banded control groups.

Apoptotic cardiocytes. Figure 1 illustrates TUNEL staining of apoptotic cells. Positive control staining is shown in Fig. 1A, in which almost all cell nuclei were stained. Negative control staining is shown in Fig. 1B. Apoptotic cardiocytes were delineated well under microscopic observation (Fig. 1C). Cardiocyte apoptosis is unusual in control hearts (Fig. 1D).

View larger version (128K): [in this window] [in a new window]   Fig. 1.   Staining of myocardial sections by terminal deoxynucleotide transferase-mediated dUTP nick end labeling (TUNEL). A: positive control staining from 12-h vehicle-treated animal. B: negative control staining from 12-h vehicle-treated animal. C: typical TUNEL-positive cardiocyte (arrow) in 12-h isoproterenol-treated heart. D: typical control heart. No cardiocyte apoptosis is seen. Original magnification ×400.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Apoptotic cardiocytes in experimental animals normalized to areas (per cm2, A) and cardiocyte nuclei estimated from positive control slides of serial sections (per 10,000 cardiocyte nuclei, B). * P < 0.05; ** P < 0.01 vs. control.

View larger version (81K): [in this window] [in a new window]   Fig. 3.   Percentages of apoptotic cardiocytes in histologically intact areas and damaged areas from 12 h to 7 days of treatment.

Effects of heart rate on cardiocyte apoptosis. Because isoproterenol increases heart rate, the increase in apoptosis might have been due to this effect rather than a direct effect of -adrenergic stimulation on cardiocytes. To differentiate these possibilities, we studied a cohort of rats subjected to cardiac pacing or to the lower isoproterenol infusion rate (40 µg · kg¹ · h¹). Although 40 µg · kg¹ · h¹ isoproterenol elicited an increase in heart rate similar to that elicited by 400 µg · kg¹ · h¹ [497 ± 8 (n = 5) vs. 517 ± 21 (n = 4) beats/min, P > 0.05], cardiocyte apoptosis detected by TUNEL assay in the group treated with the lower dose was not significantly different from that in the control group. Furthermore, pacing at a rate analogous to that seen with isoproterenol (at either dose) did not increase cardiocyte apoptosis relative to nonpaced control animals (Fig. 4).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Relationship between heart rate and extent of cardiocyte apoptosis. Cardiocyte apoptosis is significantly greater in rats treated with 400 µg · kg1 · h1 isoproterenol than in rats treated with 40 µg · kg1 · h1 isoproterenol (P < 0.05), although heart rate is not significantly different between groups. Mechanical pacing at 500 beats/min also does not increase cardiocyte apoptosis with increase in heart rate similar to that in rats treated with 400 µg · kg1 · h1 isoproterenol. * P < 0.05 vs. cardiocyte apoptosis of control (cont); # P < 0.05 vs. heart rate of control.

Effects of left ventricular hypertrophy on cardiocyte apoptosis. Because the transition from left ventricular hypertrophy to heart failure may be partially mediated by apoptotic cell loss, we evaluated rats with left ventricular hypertrophy secondary to abdominal aortic banding to establish whether hypertrophied hearts manifest cardiocyte apoptosis and whether they are susceptible to additional damage by -adrenergic stimulation. Aortic-banded rats without isoproterenol treatment showed a significant increase in cardiocyte apoptosis compared with nonbanded age- and weight-matched control rats [7.9 ± 1.5 vs. 1.1 ± 0.5 cm² (P < 0.05) and 1.08 ± 0.23 vs. 0.14 ± 0.07 per 10,000 cardiocytes (P < 0.05)]. In these rats, most of cardiocyte apoptosis was seen in the left ventricle; however, 11% of cardiocyte apoptosis was observed in the right ventricle. When isoproterenol was superimposed on established ventricular hypertrophy, the extent of cardiocyte apoptosis was not significantly different from that seen with hypertrophy alone (Fig. 5). Isoproterenol induced a degree of cardiocyte apoptosis in control, non-aortic-banded rats similar to that induced in the slightly younger rats used for time-course study (Fig. 5).

View larger version (34K): [in this window] [in a new window]   Fig. 5.   Effects of isoproterenol (400 µg · kg1 · h1) on cardiocyte apoptosis in hypertrophied hearts induced by suprarenal abdominal aortic banding. Left ventricular hypertrophy per se increases cardiocyte apoptosis. Isoproterenol does not increase cardiocyte apoptosis of hypertrophied hearts further. * P < 0.05; ** P < 0.01 vs. non-aortic-banded control rats.

Ligase reaction. Single-base 3' overhangs have been reported to be highly specific for the DNA damage associated with apoptosis as opposed to a variety of nonapoptotic processes (5). The ligase reaction can stain single 3' overhangs in situ. Thus we used this assay to verify the results of our TUNEL staining. As seen in Fig. 6, most nuclei were stained in a positive control (Fig. 6A), and no stained nuclei were seen in a negative control (Fig. 6B). Ligase reaction-positive cardiocyte nuclei were seen in isoproterenol-treated hearts (Fig. 6C) and aortic-banded animals (Fig. 6D). Cardiocyte apoptosis was seen at 17.5 ± 1.2 cm² and 2.32 ± 0.08 per 10,000 cardiocyte nuclei in 12-h isoproterenol-treated rats (n = 4), 2.4 ± 1.2 cm² and 0.38 ± 0.19 per 10,000 cardiocyte nuclei in control rats treated with vehicle for 12 h (n = 3), and 13.7 ± 2.7 cm² and 1.53 ± 0.19 per 10,000 cardiocyte nuclei in aortic-banded rats (n = 2), which were comparable to the results seen in our TUNEL assay.

View larger version (102K): [in this window] [in a new window]   Fig. 6.   Staining of myocardial sections by ligase reaction. A: positive control treated with DNase I. B: negative control. Ligase-positive cardiocyte nuclei are seen in isoproterenol (400 µg · kg1 · h1)-treated hearts (C) and hearts of aortic-banded rats (D). Original magnification ×400.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In this study we demonstrate that 1) continuous -adrenergic stimulation by isoproterenol induces consistent apoptosis of cardiocytes throughout a study period of 7 days in rat myocardium, 2) cardiocyte apopotosis induced by -adrenergic stimulation is not mediated solely by increased heart rate, 3) cardiac hypertrophy by abdominal aortic banding leads to an increase in cardiocyte apoptosis, and 4) no synergistic effect on cardiocyte apoptosis is seen when hypertrophy and -adrenergic stimulation are superimposed.

Apoptosis is a type of cell death associated with chromatin condensation, endonuclease digestion of internucleosomal DNA into 180- to 200-bp segments, and cell fragmentation into small membrane-bound vesicles (6, 16). An apoptotic body is a rounded, eosinophilic body with highly condensed nuclear chromatin, broken into an aggregate of small dense fragments seen in routine histology (12). In the present study we did not use routine histological analysis alone to assess apoptosis, since previous reports suggest that typical apoptotic bodies are rarely seen in myocardium, even in studies of myocardial ischemia and reperfusion (8) and myocardial infarction (24). The amount of apoptosis was insufficient for quantitation by DNA ladder assay, which is a highly specific marker of apoptosis, although this would not distinguish the particular cell type involved. We therefore employed the TUNEL method, which is a sensitive assay that also allows morphological assessment of cell types to detect apoptotic cell bodies in the myocardium. The specificity of the TUNEL method has been questioned, and the technique may well overestimate the number of apoptotic cells (25). We have addressed this in two ways. First, we used very strict histological criteria to eliminate noncardiocytes and contamination with nonspecific staining from analysis and, as a result, we observed relatively small numbers of apoptotic cells. Second, we employed an independent method (the ligase reaction) to define apoptotic nuclei. Didenko and Hornby (5) recently reported that the ligase reaction is more specific to apoptotic cell death than blunt-ended DNA damage detected by TUNEL, which can be seen in a variety of cases of nonapoptotic cell death in a low level. In our study the amount of cardiocyte apoptosis in control, isoproterenol-treated, and aortic-banded hearts seen with the ligase reaction was quite similar to that seen using the TUNEL assay, which provides independent confirmation that the extent of apoptosis is accurately reflected in our data.

Few TUNEL-positive cells were observed in control hearts, which is consistent with previous studies (6, 17, 18, 21, 24). Although the number of TUNEL-positive cells was significantly greater in the hearts infused with isoproterenol than in control hearts, we still observed relatively small numbers of cardiocytes undergoing apoptosis at 12 h, 24 h, and 7 days of -adrenergic stimulation. Because apoptosis occurs over a very short time in many types of cells (16), the accumulating cardiocyte loss due to apoptosis in the in vivo setting may be much more substantial. For example, in neonatal rat hearts during adaptation, 10.4, 6.1, and 2.5 cardiocyte nuclei per 10,000 cardiocyte nuclei undergo apoptotic changes on the day of birth in right ventricle, interventricular septum, and left ventricle, respectively (15). Corresponding values at 5 days after birth are 3.7, 3.5, and 2.0 cardiocyte nuclei per 10,000 cardiocyte nuclei. These numbers are qualitatively similar to those we observed in -adrenergic-induced apoptosis. Because significant anatomic changes occur in neonatal hearts as they adjust from the fetal circulation to the adult circulation, the apoptotic cardiocyte loss in our model may represent substantial damage to the heart even in the short time period. In clinical autopsy study, as many as 0.8% of cells have been identified as apoptotic in the border zones of acute myocardial infarction (24). This may reflect the relative severity and transient nature of the stimulus that differs from our model, in which the damage and apoptotic cardiocyte loss occur less acutely. In pacing-induced heart failure in dogs, 37 cardiocytes per 10,000 total nuclei undergo apoptotic changes (18).

Direct toxic effects of catecholamines on the myocardium have been known for many years since Ziegler (29) reported that catecholamines induced myocardial damage. High doses of catecholamines also cause diffuse necrotic lesions in the myocardium with accompanying inflammation and subsequent fibrous scarring (9). Recently, Matsui et al. (19) reported induction of apoptosis of cultured neonatal rat cardiocytes by norepinephrine and suggested that catecholamines may trigger cardiocyte apoptosis in hearts in vivo. Our report extends the observation of Matsui et al. in cultured neonatal rat cardiocytes and implies that a part of catecholamine-induced myocardial injury in vivo may be mediated by cell loss due to apoptosis. We chose isoproterenol, because it lacks a hypertensive effect that might contribute to cardiocyte apoptosis due to increased systolic wall stress (3).

Previously, selective desensitization of cardiac -adrenoceptors in rats by prolonged infusion of isoproterenol at the same infusion rate that we used has been reported (2, 11). In those studies, desensitization occurs as early as 2 h after infusion starts and lasts for at least 7 days. This degree of -adrenergic desensitization, which is considered to be an adaptive mechanism to -adrenergic stimulation, is probably not sufficient to protect hearts from -adrenergic stimulation-induced apoptosis in our model, inasmuch as the extent of cell death was similar and sustained throughout the time course of our analysis.

The mechanism(s) by which -adrenergic simulation induces cardiocyte apoptosis is unclear. Tachycardia is a hemodynamic consequence secondary to -adrenergic simulation. As previously reported, apoptosis is seen in pacing-induced heart failure (18). In our model, heart rates increased by 20-33% from baseline with isoproterenol treatment, which is less than that seen in the pacing-induced canine heart failure model (73%) (18). We tested rats paced for 24 h to assess whether an increase in heart rate alone can induce cardiocyte apoptosis. In this protocol the increase in the heart rate by pacing is not sufficient to induce cardiocyte apoptosis. We also observed that low-dose isoproterenol-treated rats failed to show an increase in cardiocyte apoptosis, even though an increase in heart rate similar to that seen in rats treated with 400 µg · kg¹ · h¹ isoproterenol is observed.

Cardiac muscle stretch has been associated with myocyte apoptosis (3), but no significant changes in systolic blood pressure were seen in our model, which suggests that major changes in systolic load did not occur. This makes stretch per se unlikely to play a major role in catecholamine-induced apoptosis. Myocardial ischemia has also been reported to cause apoptosis (7, 8, 14). -Adrenergic stimulation increases cardiac contractility and oxygen consumption and is known to cause vascular damage, which can contribute to myocardial ischemia.

Recently, it was reported that cardiocytes undergoing necrosis also manifest the typical DNA fragmentation that is usually seen in apoptosis (13). We have observed that some TUNEL-positive cardiocytes are located in the layers of normal-appearing cardiocytes without vascular injury and cardiocyte necrosis by light-microscopic observation (Fig. 3). Although it would seem likely that these apoptotic cardiocytes are nonischemic and not vulnerable to secondary necrosis, this is unproven. The results in neonatal cardiocyte culture suggest that a direct effect of catecholamines not mediated by ischemia can induce cardiocyte apoptosis (19).

In our study we show that left ventricular hypertrophy by abdominal aortic banding increases cardiocyte apoptosis. Previous reports suggest that cardiocyte apoptosis is increased by cardiac hypertrophy (1, 10, 17, 27). Teiger et al. (27) reported that apoptotic activity peaks at 4 days and lasts for >30 days after the ascending thoracic aorta is banded in a rat model. They observed a 50% increase in left ventricular weight by day 4 of ascending aortic banding, which is correlated with peak apoptotic activity (27). On the other hand, using TUNEL assay, Li et al. (17) reported 3.89 ± 1.28 apoptotic cells per 10,000 nuclei in spontaneous hypertensive rats between 18 and 24 mo of age (17).

In our study, abdominal aortic banding was used to induce cardiac hypertrophy, and the number of apoptotic cardiocytes was normalized to the area and the total cardiocyte nuclei. We observed a relative increase of 22% in the heart weight-to-body weight ratio over 2 wk of abdominal aortic banding with increased apoptotic cardiocytes. The extent of left ventricular hypertrophy was milder in our model than in the study by Teiger et al. (27). Our results suggest that milder and chronic hypertrophy is associated with cardiocyte apopotosis in rat hearts.

We also observed no synergistic effect on cardiocyte apoptosis between hypertrophy and -adrenergic stimulation in our experimental setting. Whether this result is due to the true adaptive mechanism or selection process, namely, elimination of susceptible cardiocyte to catecholamine-induced cardiac hypertrophy during development of hypertrophy, is unclear. Also, it is possible that the effects of -adrenergic stimulation and hypertrophy overlap, so no synergy is seen.

In conclusion, we have shown that continuous -adrenergic stimulation induces significant cardiocyte apoptosis constantly over 12 h to 7 days in rat myocardium. Given the high load of catecholamines in heart failure, this may be an important cause for the progression of heart failure. The exact mechanisms of -adrenergic stimulation-induced apoptosis are still not known. Left ventricular hypertrophy per se increases apoptotic cardiocytes, and no synergistic effect on cardiocyte apoptosis was seen between ventricular hypertrophy and -adrenergic stimulation. The mechanisms to explain these observations need to be elucidated.