Myocyte Response to β-Adrenergic Stimulation Is Preserved in the Noninfarcted Myocardium of Globally Dysfunctional Rat Hearts After Myocardial Infarction
Background—Cellular mechanisms underlying the diminished inotropic response of remodeled hearts after myocardial infarction (MI) remain unclear.
Methods and Results—Left ventricular (LV) remodeling and function were assessed by 2D echocardiography and isolated perfused heart studies in 6-week post-MI and sham-operated rats. LV myocytes from sham and noninfarcted MI hearts were used for morphometric and functional studies. β-Adrenergic receptor (β-AR) agonist isoproterenol (ISO)-induced contractile response was measured in isolated hearts. The effects of ISO and forskolin on contractile function and calcium transients of isolated myocytes were recorded. ISO-induced cAMP generation was compared in sham and MI myocytes. β-AR density was measured by radioligand binding. MI hearts were remodeled (LV diameter 8.5±0.3 versus 5.7±0.3 mm, P<0.001) and showed global (% fractional shortening 19.1±2.5 versus 55.3±2.2, P<0.01) and regional contractile dysfunction of noninfarcted myocardium (% systolic posterior wall thickening 37±4 versus 62±10, P<0.01). Isolated heart function (LV developed pressure 58±2 versus 72±3 mm Hg, P=0.004) and ISO concentration response were reduced in MI hearts. Myocytes from the noninfarcted LV were structurally remodeled (32% longer and 18% wider), but their contractile response and intracellular calcium kinetics to ISO and forskolin were not diminished. β-AR receptor density (Bmax 24±1.5 versus 22.4±1.6 fmol/mg protein) and β-AR agonist–stimulated cAMP were similar in both groups.
Conclusions—Isolated myocytes from the remodeled and dysfunctional myocardium are structurally modified but contract normally under basal conditions and in response to β-AR stimulation. β-AR density is preserved in remodeled myocytes. Nonmyocyte factors may be more important in the genesis of contractile dysfunction in the remodeled rat heart up to 6 weeks after MI.
After a large myocardial infarction (MI), the heart undergoes a series of structural and functional changes known as ventricular remodeling.1 2 The cellular and molecular changes contributing to the global and regional contractile dysfunction of the remodeled myocardium are poorly understood. Of particular intrigue is the lack of strong evidence confirming the presence of contractile abnormalities of individual myocytes from the remote remodeled myocardium. The prevalent hypothesis is that contractile dysfunction is either caused by or at least contributed to by contractile dysfunction of the underlying cardiomyocytes. Abnormalities may involve intrinsic as well as modulated function of these myocytes.3
The β-adrenergic (β-AR) system is the predominant neurohormonal modulator of cardiac contractility and may be of greater significance in states of diminished contractile reserve. Since the original description of β-AR desensitization in the failing human myocardium,4 various abnormalities of its signal transduction pathway have been shown in different models of heart failure. These changes may not be present uniformly, across different species, and in different models of cardiomyopathy.5 Studies of altered β-AR transduction in the post-MI model have been few, and their findings are contradictory.6 7
In the rat infarct model, we had previously shown normal intrinsic myocyte contractile function.8 The current study was designed to evaluate whether abnormalities of β-AR–modulated function were present in ventricular myocytes from the remodeled myocardium and if these could be explained by alterations in the β-AR quantity and agonist-induced cAMP activity. Our findings suggest the presence of a unique state in the evolution of post-MI heart failure, exhibited by marked global and regional contractile dysfunction despite the presence of intact contractility at the myocyte level.
Myocardial infarction was produced in male Sprague-Dawley rats (weight 150 to 200 g) by ligation of the left coronary artery as previously described.8 In sham-operated rats (sham), the ligature was passed around the coronary artery but not tied. Surviving animals (sham, 35; MI, 50) were studied in 3 groups, 6 weeks later. The experimental protocol was approved by our institutional review board.
Echocardiography (2D) was performed in all animals 1 and 5 weeks after surgery2 with a 7.5-MHz, phased-array transducer (SONOS 2500 Hewlett Packard). M-mode measurements of cavity size and anterior and posterior left ventricular (LV) wall thickness in systole and diastole were made (AWTs or d and PWTs or d). Global LV function was assessed by calculating LV percent fractional shortening. The percent posterior wall thickening in systole (% PWT=[PWTs−PWTd]/PWTd×100) and relative wall thickness (RWT=2×PWTd/LV internal dimension [ID]d) were calculated.2
Isolated Heart Function
Isolated heart function was studied in 6 MI and 5 sham group 1 rats. After pentobarbital anesthesia (50 mg/kg intraperitoneal), hearts were removed and perfused in a modified “work-performing mode.”9 Hearts were paced at 5.5 Hz, and left atrial (LA), LV, and aortic pressures were monitored simultaneously. LA inflow and aortic outflow were measured by electromagnetic flow probes (T206, Transonic Systems). Baseline LV systolic, diastolic, and end-diastolic pressures (LVEDP) and peak LV dP/dt maximum and minimum were recorded (LA inflow 10 mL · g−1 · min−1; aortic pressure 50 mm Hg). To investigate the Frank-Starling relation, LA inflow was increased in increments of 2 mL · g−1 · min−1 to 20 mL · g−1 · min−1. Response to β-AR stimulation was studied with the nonselective agonist isoproterenol (ISO, 10−10 to 3×10−7 mol/L). ISO was infused into the LA cannula, and hemodynamics were measured after a 3-minute stabilization at each concentration. Hearts were removed and fixed for measurement of infarct size.
Infarct Size Determination
Isolated hearts used for perfusion studies (group 1) were fixed in 10% phosphate buffered formaldehyde for 24 hours. Hearts were sectioned (2-mm thickness) from apex to base, and 5-μm representative sections were prepared from the basal surface of each slice for Masson’s trichrome staining. Images of serial sections were captured on a digital camera, and planimetric measurements were performed with a computerized imaging program (MetaMorph, Universal Imaging). Infarct size was calculated from 3 mid-wall sections of each heart by the formula [(infarct endocardial+epicardial circumference)/(LV endocardial+epicardial circumferences)]×100 and averaged.
Isolated Myocyte Studies
Ventricular myocytes were isolated from 15 sham and 25 MI hearts (group 2) by retrograde enzymatic perfusion.8 In MI hearts, the infarcted area and a 2-mm rim of peri-infarct tissue were carefully removed and discarded. Myocytes were isolated from the remaining noninfarcted myocardium. Myocyte viability (trypan blue exclusion) of sham (75±4%) and MI cells (70±4%) were similar. Cells were suspended in a HEPES buffer containing 200 μmol/L extracellular calcium ([Ca2+]o) and divided into aliquots for functional and morphometric studies.
Myocyte dimensions were measured in 100 randomly chosen LV myocytes by phase-contrast microscopy8 (sham, 1500 cells; MI, 2500 cells).
Measurement of Myocyte Contractile and Intracellular Calcium Response
Myocytes loaded with a fluorescent dye Fura-2 AM (Fura-2 acetoxymethyl ester) were randomly selected, and simultaneous measurements of contraction and intracellular calcium ([Ca2+]i) were made with a video edge detector and a high-speed camera coupled to a dual-excitation fluorescence system.8 Myocytes were perfused with HEPES buffer (2 mL/min, 30°C) and stimulated at 0.5 Hz. After recording basal function at 1 mmol/L [Ca2+]o, myocytes were exposed to either ISO (10−9, 3×10−9, and 10−8 mol/L) or the adenylate cyclase agonist forskolin (FSK, 10−7, 10−6, and 10−5 mol/L) for 5 minutes before recording the effect.
Membrane fractions were prepared from myocytes from the noninfarcted myocardium of 13 MI and 9 sham hearts.10 Briefly, membranes were extracted in 0.5 mol/L KCl and stored at − 70°C.125Iodocyanopindolol (CYP) (New England Nuclear; 5 to 320 pmol/L) was used to label membrane β-AR. The assay was performed with 50 μg of membrane protein at 37°C for 60 minutes. The nonselective β-AR agonist l-isoproterenol (5 μmol/L) was used to calculate nonspecific binding. Competitive binding curves were derived by means of a selective β-AR antagonist, ICI-118 551 (3×10−9 to 3×10−4). Data were analyzed with Graphpad Prism v 2.0.
Agonist-stimulated cAMP generation was measured in myocytes from 6 MI and 6 sham hearts by radioimmunoassay (Amersham Corp).11
All data are expressed as mean±SEM. Comparison between 2 groups was made by a 2-tailed Student’s t test. Repeated variables were analyzed by a repeated-measures ANOVA. Between-group (sham versus MI) differences in repeated variables were considered significant if the interaction term P value was <0.05. This was followed by a post hoc analysis to identify pairwise differences at each point, only if the interaction term P value was <0.05. Statistical analysis was performed by a software package (SPSS 7.5, SPSS Inc). For all tests, a P value of ≤0.05 was considered significant.
Body Weight, Heart Weight, and Infarct Size
MI rats weighed less than sham rats (372±16 versus 428±13 g, P=0.01). The heart weight (1.92±0.07 versus 1.54±0.06, sham) and heart weight (g)–to–body weight (g) ratio (5.49±0.29 versus 3.7±0.2, sham) were significantly increased in MI rats (both P<0.001). Mean infarct size was 38.9±2.4%.
Results of echocardiography, done on all 35 sham and 50 MI rats, are shown in Table 1⇓. LV dimensions in systole and diastole were significantly increased in MI hearts. Most of the increase was seen 1 week after MI, with only a minimal change thereafter. LV fractional shortening was reduced by half 1 week after MI, with no further progression at 5 weeks. Diastolic posterior wall thickness (PWTd) was similar in MI and sham hearts throughout the study. However, the systolic posterior wall thickness (PWTs) decreased in 5-week post-MI hearts, resulting in a significant reduction in percent systolic thickening of the posterior wall. Relative posterior wall thickness was reduced in MI hearts at both 1 and 5 weeks.
Isolated Work-Performing Heart Function
Infarcted hearts had significantly lower LV developed pressure and peak positive dP/dt and higher LVEDP at similar LA inflow (10 mL · g−1 · min−1) and aortic resistance (50 mm Hg; Table 2⇓). The Frank-Starling relations showed a preload-dependent increase in LV dP/dt in both groups that was significantly blunted in infarcted hearts (Figure 1⇓). ISO caused an increase in peak LV dP/dtmax and dP/dtmin that was significantly reduced in MI hearts (Figure 2⇓).
Isolated Myocyte Structure and Function
Myocytes from MI hearts were significantly longer (148±5 versus 112±2 μm versus 109±8, P<0.01) and wider (24±1 versus 20±1 μm, P<0.001) than sham hearts. Similar changes were noted in myocytes used for functional study (143±4 versus 119±2 μm; P<0.001).
Myocyte Contractile Response
Effect of ISO on myocyte contractility and [Ca2+]i was studied in 81 MI and 80 sham myocytes (Table 3⇓). In the basal state ([Ca2+]o 1 mmol/L, 0.2 Hz, 30°C), there was no difference in percent myocyte shortening between sham and MI myocytes. However, MI myocytes had a significantly greater mean velocity of shortening and shorter time to 70% relengthening. ISO caused a significant concentration-dependent increase in percent myocyte shortening and mean velocity of shortening and a decrease in time to 70% relengthening in both groups that was greater in MI as compared with sham myocytes (P<0.05, interaction term; ANOVA).
Myocyte Intracellular Calcium Kinetics
At 1 mmol/L [Ca2+]o, MI myocytes had a significantly lower resting [Ca2+]i and greater [Ca2+]i transient amplitude. There was no difference in the mean velocity of [Ca2+]i rise and time to 70% decline in [Ca2+]i transient between the groups. No difference was seen in the basal calcium sensitivity (change in length/change in Fura-2 ratio) between sham and MI cells. ISO caused a concentration-dependent increase in resting [Ca2+]i, amplitude of [Ca2+]i, mean velocity of rise in [Ca2+]i, and a reduction in the time to 70% decline in [Ca2+]i in both groups. ISO also caused a similar increase in the calcium sensitivity in both groups. The ISO concentration-dependent increase in amplitude of [Ca2+]i and mean velocity of [Ca2+]i rise were greater in MI as compared with sham myocytes. However, there was no difference in the response of resting [Ca2+]i, time to 70% decline in [Ca2+]i, and calcium sensitivity ratio between the groups.
Myocyte contractile and [Ca2+]i response to FSK were studied in 34 sham and 65 MI myocytes. Like ISO, FSK also caused a concentration-dependent increase in percent myocyte shortening and Fura-2 amplitude (Figure 3⇓). These responses tended to be greater in MI as compared with sham myocytes.
Linear plots of bound/free versus bound (Figure 4⇓ insert) demonstrated a saturable, highly specific binding to a single class of sites. There was no difference in the equilibrium dissociation constant (KD) of β-AR radioligand binding between MI and sham myocytes (MI, 18.7±5.2 versus sham, 20.6±4.6 nmol/L). The maximal number of radioligand binding sites was similar in the two groups (Bmax, MI, 24±1.5 versus sham, 22.4±1.6 fmol/mg protein).
The proportion of β1-AR and β2-AR was determined by plotting competitive binding curves with CYP and the highly selective β1-AR antagonist ICI 118551. Binding data indicated that a 2-site model was the perfect fit for ICI 118551. There was no difference in β1:β2 AR subtype ratio in myocytes from sham (β1:β2 68±2.6%: 32±3.4%) and MI hearts (β1:β2 64±3.9%: 36±2.2%). The Ki for β1- and β2-AR were 1.24±0.1 and 269±18 nmol/L for sham and 1.56±0.08 and 281.6±20 nmol/L for MI. None of these differences were statistically significant.
l-Isoproterenol caused a similar concentration-dependent increased cAMP production in both groups of myocytes (Figure 5⇓, P=0.9, ANOVA interaction term).
In this study we have shown that the inotropic response to the β-AR stimulation was reduced in 6-week post-MI hearts but not in ventricular myocytes isolated from their remodeled myocardium. Interestingly, the contractile function and [Ca2+]i kinetics were, in fact, slightly increased in the remodeled myocytes. We also found no differences in the total β-AR density, β1:β2 AR ratio, or agonist-induced cAMP generation in myocytes isolated from the noninfarcted myocardium.
Ventricular Remodeling in Rat Infarct Model
Serial echocardiography confirmed the presence of LV remodeling and contractile dysfunction. Marked enlargement of LV cavity was evident by 1 week, accompanied by reduction in global LV function (percent fractional shortening). By 5 weeks, although there was only a minimal further increase in LV dimension, the remote myocardium developed contractile dysfunction with reduction in percent systolic thickening of the posterior wall. Global LV systolic and diastolic dysfunction was also confirmed in isolated heart function studies. The blunted Frank-Starling relation seen in infarcted rat hearts is similar to that reported in humans and supports the observation that contractile reserve, albeit diminished, is still preserved in failing myocardium.
An important finding in this model is that compensatory hypertrophy of the remote noninfarcted myocardium does not increase wall thickness.2 12 The PWTd at 5 weeks did not differ between sham and MI hearts. This “inadequate hypertrophy” of the surviving myocardium could contribute to the global and regional contractile dysfunction by worsening wall stress.
Myocyte Function in Remodeled Myocardium
It is reasonable to suspect that intrinsic contractile abnormalities of surviving myocytes may contribute to contractile dysfunction seen in the noninfarcted myocardium. Significant structural remodeling was indeed noted in myocytes isolated from these regions, similar to those described by us previously.8 However, these remodeled myocytes did not show any abnormalities of contractile function or [Ca2+]i kinetics in the basal state. Whether abnormalities of myocyte contractile function develop in the noninfarcted myocardium remains controversial. Studies on myocytes specifically isolated from the remote regions have found no evidence of contractile dysfunction,8 13 14 in contrast to others where a clear distinction between the remote and peri-infarct regions was not made.15 16 It is important to emphasize that in post-MI remodeling, myocyte hypertrophy can be differentially regulated across peri-infarct and remote myocardium.14 17 This distinction may be useful in interpreting results of myocyte function in other comparable studies reporting conflicting results.
The decrease in global inotropic response to β-AR stimulation in isolated infarcted hearts observed in our study is comparable to previous studies in the rat model.6 18 Myocyte loss as a result of MI and consequent ventricular dilation and increased wall stress could explain some of the diminished contractile response in isolated heart preparations. This, however, cannot explain the attenuated contractile response to ISO reported in noninfarcted isolated papillary muscle.19 20 A major limitation of multicellular preparations is that significant changes occur in the extracellular matrix in this model that may alter mechanical properties of the noninfarcted myocardium and affect its contractile response.21 22 Litwin and Morgan19 studied papillary muscles of 6-week post-MI rats and found that ISO caused a concentration-dependent increase in [Ca2+]i similar to that seen in our isolated myocytes. However, unlike the preserved inotropic response in the latter, agonist-induced increase in papillary muscle tension was significantly blunted. A blunted mechanical response in the presence of a normal [Ca2+]i in papillary muscle could be explained by a defect in mechanical coupling of myocytes resulting from extracellular matrix abnormalities.
Contractile response to β-AR stimulation has been reported to be diminished in myocytes from other models of heart failure.23 24 25 26 It is important to note that in these studies, even basal myocyte contractile function was significantly depressed. Moreover, these studies only tested a single concentration of ISO, making direct comparisons difficult. The finding of normal or exaggerated inotropic response in our myocytes is not unique among published literature. Recently, McIntosh et al27 found that ventricular myocytes isolated from the subendocardial region in a rabbit model of post-MI heart failure had shorter action potentials and greater [Ca2+]i amplitudes. These findings were in stark contrast to the cells from the subepicardial region that demonstrated blunted peak [Ca2+]i and longer action potentials.
β-AR Modulation in Postinfarction Remodeling
Since β-AR downregulation may be dependent on the cause of heart failure,5 variations in receptor modulation in post-MI heart failure could aid in understanding the beneficial mechanisms of clinical β-AR blockade. In the rat infarct model, myocardial β-AR have been found to be preserved,10 28 29 30 31 downregulated,7 or even upregulated.18 Unlike Sethi et al,7 who found that β-AR downregulation correlated well with the extent of hemodynamic dysfunction, we could not confirm these findings despite profound global and regional contractile dysfunction in our rats. It is unclear whether they studied only the remote noninfarcted myocardium. The differences reported in the literature may be explained on whether the remote or entire LV was studied. Recent studies30 32 show that β-AR density is preserved in the remote noninfarcted myocardium of 5-week post-MI hearts, whereas significant reduction is seen in the peri-infarct zone.30 These data further support the emerging hypothesis that surviving ventricular myocytes may be differentially regulated after a large MI.17 33 The apparently preserved and indeed mildly enhanced modulated function in myocytes from the remodeled myocardium is further strengthened by our finding of normal agonist-induced cAMP generation. These findings are internally consistent with the results of our functional studies with the adenylate cyclase agonist FSK. Whereas investigation of the β-AR transduction pathway distal to cAMP is beyond the scope of the present study, it is possible that abnormalities involving the G proteins and cAMP-independent mechanisms could explain the diminished modulated function in failing hearts.
Although our studies were limited to an early stage of post-MI heart failure, it is possible that myocyte contractile dysfunction and abnormalities of β-AR signal transduction appear later in the natural history of this model. Another significant limitation common to all isolated myocyte studies is that cells studied may not be representative of the myocytes in the intact remodeled myocardium. The isolation procedure could inadvertently affect the yield of viable but dysfunctional myocytes. Despite this possibility, the finding of supranormal function in a significant proportion of myocytes is in itself interesting.
The rat infarct model appears to have significant differences from other models of heart failure. In this model, LV chamber dilation and global LV dysfunction occur early after MI (1 week), with a gradual development of regional contractile dysfunction in the remote noninfarcted myocardium. Myocytes from the remote noninfarcted regions of LV myocardium are structurally remodeled similar to that seen in other models of dilated cardiomyopathy. However, unlike other models, both basal and modulated functions of ventricular myocytes remain preserved for up to 6 weeks after MI. Contractile response to nonselective β-AR agonist may even be increased in these ventricular myocytes. These results demonstrate that early in the course of global myocardial dysfunction, unitary contractile machinery could still be functionally intact. In addition, β-AR density, β1:β2 subtype ratio, and agonist-induced cAMP production are also preserved in these myocytes. Nonmyocyte factors such as increased wall stress and extracellular matrix abnormalities probably make significant contributions to global and regional contractile dysfunction in this model. β-AR downregulation may not always be present in remodeled ventricular myocardium despite the presence of an attenuated global adrenergic inotropic response.
Presented in part at the 47th Annual Scientific Sessions of the American College of Cardiology, Atlanta, Ga, March 29 to April 1, 1998, and 71st Scientific Sessions of the American Heart Association, Dallas, Tex, November 8 to 11, 1998.
- Received March 6, 2000.
- Revision received May 12, 2000.
- Accepted May 15, 2000.
- Copyright © 2000 by American Heart Association
Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial infarction: experimental observations and clinical implications. Circulation. 1990;81:1161–1172.
Litwin SE, Katz SE, Morgan JP, et al. Serial echocardiographic assessment of left ventricular geometry and function after large myocardial infarction in the rat. Circulation. 1994;89:345–354.
Bristow MR. Why does the myocardium fail? Insights from basic science. Lancet. 1998;352(suppl 1):SI8-SI14.
Bristow MR, Feldman AM. Changes in the receptor-G protein-adenylyl cyclase system in heart failure from various types of heart muscle disease. Basic Res Cardiol. 1992;87:15–35.
Sethi R, Dhalla KS, Beamish RE, et al. Differential changes in left and right ventricular adenylyl cyclase activities in congestive heart failure. Am J Physiol. 1997;272:H884–H893.
Anand IS, Liu D, Chugh SS, et al. Isolated myocyte contractile function is normal in postinfarct remodeled rat heart with systolic dysfunction. Circulation. 1997;96:3974–3984.
Grupp IL, Subramaniam A, Hewett TE, et al. Comparison of normal, hypodynamic, and hyperdynamic mouse hearts using isolated work-performing heart preparations. Am J Physiol. 1993;265:H1401–H1410.
Chasteney EA, Liang CS, Hood WB Jr. Beta-adrenoceptor and adenylate cyclase function in the infarct model of rat heart failure. Proc Soc Exp Biol Med. 1992;200:90–94.
Steinberg SF, Robinson RB, Lieberman HB, et al. Thrombin modulates phosphoinositide metabolism, cytosolic calcium and impulse initiation in the heart. Circ Res. 1991;68:1216–1229.
Lefroy DC, Crake T, Del Monte F, et al. Angiotensin II and contraction of isolated myocytes from human, guinea pig, and infarcted rat hearts. Am J Physiol. 1996;270:H2060–H2069.
Melillo G, Lima JA, Judd RM, et al. Intrinsic myocyte dysfunction and tyrosine kinase pathway activation underlie the impaired wall thickening of adjacent regions during postinfarct left ventricular remodeling. Circulation. 1996;93:1447–1458.
Cheung JY, Musch TI, Misawa H, et al. Impaired cardiac function in rats with healed myocardial infarction: cellular vs myocardial mechanisms. Am J Physiol. 1994;266:C29–C36.
Capasso J, Anversa P. Mechanical performance of spared myocytes after myocardial infarction in rats: effects of captopril treatment. Am J Physiol. 1992;263:H841–H849.
Litwin SE, Morgan JP. Captopril enhances intracellular calcium handling and beta-adrenergic responsiveness of myocardium from rats with postinfarction failure. Circ Res. 1992;71:797–807.
Litwin SE, Litwin CM, Raya TE, et al. Contractility and stiffness of noninfarcted myocardium after coronary ligation in rats: effects of chronic angiotensin converting enzyme inhibition. Circulation. 1991;83:1028–1037.
Zhang XQ, Moore RL, Tillotson DL, et al. Calcium currents in postinfarction rat cardiac myocytes. Am J Physiol. 1995;269:C1464–C1473.
Yamamoto S, Tsutsui H, Tagawa H, et al. Role of myocyte nitric oxide in beta-adrenergic hyporesponsiveness in heart failure. Circulation. 1997;95:1111–1114.
McIntosh MA, Cobbe SM, Smith GL. Heterogeneous changes in action potential and intracellular Ca2+ in left ventricular myocyte sub-types from rabbits with heart failure. Cardiovasc Res. 2000;45:397–409.
Cherng WJ, Liang CS, Hood WB Jr. Effects of metoprolol on left ventricular function in rats with myocardial infarction. Am J Physiol. 1994;266:H787–H794.
Huang B, Wang S, Qin D, et al. Diminished basal phosphorylation level of phospholamban in the postinfarction remodeled rat ventricle: role of beta-adrenergic pathway, Gi protein, phosphodiesterase, and phosphatases. Circ Res. 1999;85:848–855.