We studied the role of β-adrenergic receptor subtype signaling to cAMP and calcium in the genesis of catecholamine-dependent arrhythmias in German shepherd dogs that develop lethal arrhythmias at ∼5 mo of age. There were three major findings in this study: 1) isoproterenol induces similar increases in cAMP in afflicted and control dogs exclusively through β1-receptors (not β2), 2) cells from afflicted dogs display prolonged relaxation kinetics at long cycle lengths and large frequent spontaneous calcium oscillations (and aftercontractions) with little increase in calcium transient amplitude in response to β1-receptor agonists, and 3) β2-receptor agonists induce a similar marked increases in calcium transient and twitch amplitude, with only rare spontaneous calcium oscillations in afflicted and control cells. These results indicate that catecholamines provide inotropic support to canine cardiomyocytes through distinct β1- and β2-receptor pathways with differing requirements for cAMP. The propensity to develop arrhythmias is not induced by β2-receptors (or a rise in calcium alone), but rather occurs in the context of β1-receptor activation of the cAMP-dependent pathway.
- cardiac arrhythmia
- β-adrenergic receptors
german shepherd dogsthat develop lethal ventricular tachycardias from 4 to 6 mo of age have become a useful tool to explore mechanisms that mediate catecholamine-dependent arrhythmias in the young (3, 10, 11,15). These animals manifest delayed maturation of sympathetic innervation of the anterospetal left ventricle (2) and two types of arrhythmias: those which are pause dependent and potentiated by α-adrenergic receptor stimulation of the Purkinje system (3,15) and those which are tachycardia initiated and secondary to delayed afterdepolarization-induced triggered activity in the myocardium (15). The latter mechanism is mediated by β1- (but not β2) receptor-dependent mechanisms (16). In view of recent evidence that cardiomyocyte β1- and β2-adrenergic receptors can exhibit distinct signaling properties (17), the present study examines β-adrenergic receptor subtype linkage to downstream effectors. The goal is to determine the extent to which differences in the propensity of the afflicted animals to develop arrhythmias can be attributed to differences in β1- versus β2-adrenergic receptor signaling to the cAMP pathway and/or the modulation of intracellular calcium.
We studied 11 German shepherd dogs from an inbred colony at 20–25 wk of age that were identified as having lethal arrhythmias via previously described electrocardiogram monitoring techniques (10, 11). Eight unrelated German shepherd dogs of the same age without arrhythmias served as controls. Animals were anesthetized with pentobarbital sodium (30 mg/kg), and the hearts were excised via a thoracotomy and placed in ice-cold Tyrode's solution. The anteroseptal and posterobasal regions of the left ventricle were identified as previously (15) and were cut away from the remaining myocardium. Cell preparation was as described below.
Reagents were purchased commercially from the following sources: fura 2-acetoxymethyl ester (AM) (Molecular Probes), ICI-188551 (Cambridge Research Biochemicals), forskolin (Sigma). CGP-20712A was purchased from Ciba-Geigy (Berne, Switzerland) and zinterol was a generous gift from Bristol-Myers Squibb (Wallingford, CT). All other chemicals were reagent grade.
Anteroseptal (AS) and posterobasal (PB) regions of the left ventricle were dissected and perfused through the left anterior descending or left circumflex coronary artery, respectively. Collagenase dissociation was conducted as originally described by Hewett et al. (4) and subsequently modified according to the method of Hoffman et al. (5, 6). It employs a variation on the Langendorff perfusion where only a perfused wedge of left ventricular free wall (with cut vessels that were tied off or stapled closed) is used, rather than the entire heart, and perfusion is maintained with a roller pump rather than gravity. The roller pump maintains flow through the heart wedge at about 10 ml/min. The wedge is perfused with a collagenase, the epicardial and endocardial surfaces are then trimmed away, and the remaining tissue minced and triturated in an additional series of collagenase solutions to produce isolated myocytes.
Cells from the AS or PB regions of the ventricle were preincubated for 60 min at room temperature with 10 mM theophylline. Assays were performed in the absence or presence of agonist for 5 min at room temperature and were terminated by removal of the incubation buffer and the addition of 1 ml of ethanol. Each condition was assayed in duplicate, with cAMP measured in quadruplicate. Aliquots of the alcohol-fixed cell extract were dried under a stream of nitrogen, and cAMP in the residue was determined using a radioimmunoassay (DuPont-New England Nuclear; Wilmington, DE).
Measurements of calcium transients and twitches.
Myocytes were loaded with fura 2-acetoxymethyl ester (AM) by incubation for 20 min at 37°C with 3 mM fura 2-AM and 1.5 ml of 25% (wt/wt in dimethyl sulfoxide) Pluronic F-127 (BASF Wyandotte) dissolved in 1.0 ml Tyrode's solution. Myocytes were rinsed with fresh Tyrode's solution and maintained for at least 15 min at room temperature to permit deesterification of the dye before protocols. Intracellular fura 2 fluorescence was monitored (at a sampling rate of 100 Hz) with a device that alternately illuminates the cells with 340 and 380 nm light while it measures emission at 520 nm (Photon Technologies; Princeton, NJ). Myocytes were superfused with room temperature Tyrode's solution gassed with 95% O2-5% CO2 at a rate of 1 ml/min. The experimental protocol consisted of successive 1-min intervals of electrical field stimulation at 1, 0.5, 0.2, and 0.1 Hz, followed by pacing at 0.5 Hz during superfusion with β-receptor agonist. Intracellular calcium is reported as the fura 2 fluorescence ratio because of the uncertainties inherent in any attempt to calibrate these signals (7).
To monitor cell motion, myocytes were simultaneously illuminated with red light, and a dichroic mirror (630 nm cutoff) in the emission path deflected the cell image to a video optical system (Crescent Electronics), which tracked motion of the cell edges along a raster line segment of the image during electrically stimulated contractions. The analog voltage output from the motion detector was calibrated to convert to microns of motion. The motion signal was obtained at a rate of 60 Hz and reflected the motion of the same myocyte simultaneously monitored with fura 2 for calcium. The digitized signal was stored along with the fluorescence data.
Data are presented as means ± SE. Statistical significance was determined by ANOVA, Student's t-test, or Fisher's exact test, as appropriate. P < 0.05 was regarded as significant.
β-Adrenergic receptor activation of cAMP.
Figure 1 shows the results of cAMP measurements on cells prepared from the AS and PB regions of control and afflicted German shepherd dogs. Basal cAMP levels do not significantly differ across control and afflicted groups and, in each case, cAMP accumulation is markedly enhanced by exposure to isoproterenol and forskolin (Fig. 1 A). The isoproterenol-dependent cAMP accumulation is blocked by the β1-adrenergic receptor antagonist CGP-20712A, but not by the β2-adrenergic receptor antagonist ICI-188551. Similarly, zinterol, which is selective for the β2-adrenergic receptor at the concentration used (7), induces no significant increase in cAMP accumulation. Although basal and stimulated levels of cAMP tend to be lower in the AS region of afflicted dogs, the values do not differ across groups (P > 0.05). Figure 1 B displays the concentration-response relationships for isoproterenol-dependent cAMP accumulation; 50% effective concentrations for isoproterenol are similar in each preparation.
Characteristics of the calcium transient and twitch.
Regulation of intracellular calcium depends on the integration of processes regulating calcium entry and extrusion at the cell surface as well as calcium release and reuptake at the sarcoplasmic reticulum. Figure 2 compares amplitudes and kinetics of calcium transients (Fig. 2 A) and twitches (Fig.2 B). Cells isolated from both regions of control and afflicted dogs display rate-dependent increases in amplitude and a hastening of the kinetics of the calcium transient and the twitch. The amplitudes of the calcium transients and twitches are similar across all preparations [at all cycle lengths (CL)]. At 1 and 0.5 Hz, the kinetics of the calcium transients and twitches also are not distinguishable. However, at long CL (0.2 and 0.1 Hz), the calcium transient and twitch kinetics of afflicted cells were slower than the controls (P < 0.05).
β-Adrenergic receptor subtype modulation of intracellular calcium.
To examine the functional consequences of cellular activation by β1- and β2-adrenergic receptor subtypes, cardiomyocytes were exposed to progressively increasing concentrations of isoproterenol (10−9–10−7 M) or to 10−7 M zinterol during continuous electrical stimulation at 0.5 Hz. In control cells, the effect of isoproterenol on promoting cAMP accumulation (see above) was accompanied by a marked increase in the amplitude of the calcium transient and the twitch. Three cells from the AS and PB regions of control dogs displayed spontaneous impulse initiation and severe calcium overload at 10−7 M isoproterenol. The remaining 14 control cells tolerated the study protocol. Occasional oscillations of calcium accompanied by aftercontractions that delayed the return of calcium to baseline values and relaxation to resting cell length were observed during superfusion with 10−7 M isoproterenol in 6 of 14 cells in the control group (4 cells from the AS and 2 cells from the PB regions). However, these were isolated events with amplitudes that never exceeded 10% of the amplitude of the electrically stimulated twitch (as shown in Fig.3).
The study of calcium transients and motion had to be aborted in 7 of 66 cells from afflicted dogs (P > 0.05 of control) due to failure to sustain stable baseline conditions (i.e., spontaneous calcium transients and aftercontractions were prominent before the initiation of superfusion with isoproterenol). Four cells that sustained stable baseline electrically driven contractions (2 cells from the AS region and 2 cells from the PB region) were exposed to 5 × 10−9 M isoproterenol. In marked contrast to cells from control dogs, this concentration of isoproterenol induced a 60–100% increase in the amplitude of the electrically driven twitch with little to no increase in the amplitude of the associated calcium transient. Rather, isoproterenol induced frequent large spontaneous calcium oscillations (accompanied by large aftercontractions) that rapidly culminated in cell death terminating each of these experiments (Fig. 4).
Given the toxicity of isoproterenol to cells from afflicted ventricles, the protocol was revised to span a lower range of agonist concentrations (5 × 10−10 to 5 × 10−9 M). Isoproterenol 5 × 10−10 M evoked a substantial increment in the amplitude of the twitch (61 ± 4%) despite only a minor increase in the amplitude of the associated calcium transient (6 ± 2% n = 26, Fig. 5). This concentration of isoproterenol generally was tolerated. However, at 5 × 10−9 M isoproterenol, all 15 AS and 11 PB cells developed prominent calcium oscillations and aftercontractions, with no significant additional increment in calcium transient or twitch amplitudes. No regional differences in calcium and inotropic responses (or cardiomyocyte susceptibility to the arrhythmogenic properties of isoproterenol) were identified between AS and PB cells.
To distinguish potential contributions of β-adrenergic receptor subtypes to catecholamine-dependent changes in calcium and cell motion, eight cells were stimulated with 10−7 M zinterol to selectively activate β2-adrenergic receptors. The response of a cell isolated from the AS region of an afflicted dog is depicted in Fig. 6. This result is typical of data obtained in myocytes from both regions of afflicted and control dogs. Two points are noteworthy. First, zinterol elicited a 151 ± 32% increase in the magnitude of the twitch in association with a 132 ± 22% increase in the amplitude of the calcium transient. Interestingly, a similar level of inotropic support via β1-adrenergic receptors was unaccompanied by any change in calcium transient amplitude in afflicted dogs (compare Figs. 5 and6). Second, the zinterol-dependent increase in calcium only rarely evoked oscillations of calcium in afflicted dogs, and these were self-limited (in 2 of 8 cells). This indicates that the rise in calcium alone is not sufficient to induce arrhythmias in cardiomyocytes from afflicted dogs.
These results identify catecholamine-dependent inotropic support in cardiomyocytes from control and afflicted dogs through distinct β1- and β2-adrenergic receptor pathways (with differing requirements for cAMP); only catecholamine stimulation of the cAMP-dependent β1-adrenergic receptor pathway (not the β2-receptor pathway) leads to calcium oscillations that are large, frequent, and associated with large frequent aftercontractions in afflicted dogs. These abnormalities are likely to contribute to the mechanism for catecholamine-induced ventricular arrhythmias in this German shepherd model of spontaneous lethal arrhythmias that develop during 4–6 mo of age.
As previously demonstrated (15), two types of arrhythmias occur in this model: pause-dependent arrhythmias are enhanced by α-adrenergic stimulation and arrhythmias induced by rapid heart rates are facilitated by β1-adrenergic stimulation. The latter have been attributed to triggered activity induced by delayed afterdepolarizations (15). Yet early afterdepolarization (EAD)-dependent arrhythmias also may be facilitated by β-adrenergic mechanisms, such that there is an increase in amplitude of EADs and likelihood of triggered activity (20). This type of effect is demonstrated in Figs. 4 and 5.
This study identifies two abnormalities in calcium cycling that may contribute to the genesis of arrhythmias in afflicted animals. First, cardiomyocytes from afflicted dogs are severely impaired in their ability to “tolerate” β1-receptor stimulation; this distinct biological property is manifest as the development of calcium oscillations, severe calcium overload, and cell death during exposure to typical concentrations of β1-receptor agonists (Fig.4). Second, cardiomyocytes from afflicted dogs display increased duration of the calcium transient and prolonged twitch relaxation rates at long CL (Fig. 2). Impaired calcium extrusion mechanisms are predicted to contribute to arrhythmias that develop during the frequent pauses that interrupt tachycardias or at times of marked variability in heart rate both of which characteristics are seen frequently during sinus rhythm in these animals (10, 11, 15). Candidate loci for these defects in calcium handling include calcium regulatory proteins in the sarcoplasmic reticulum, i.e., sarco(endo)plasmic reticulum Ca2+-ATPase, phospholamban, and ryanodine receptor, or plasma membrane (the Na/Ca exchanger). Many of these proteins undergo developmental maturational processes that would be susceptible to developmental dysregulation and would be pertinent to a syndrome that displays an age-dependent phenotype. However, the slower calcium transient relaxation kinetics of ventricular myocytes from afflicted animals, that are manifest only at long CLs, may also be the consequence of the longer action potential duration at slow drive rates. Previous studies (15, 16) attributed this defect to prolongation of the plateau. Whereas abnormalities of several channels are possible, preliminary data suggest that the pathological phenotype may be attributed at least in part to increased L-type calcium current in ventricular myocytes from afflicted but not control animals.
As noted in methods, the Ca signaling experiments were performed at room temperature. It is true that action potential configuration at room temperature and physiological temperature will differ. However, it is important to note that the starting point for the present study was our previous report that in ventricular tissue at physiological temperature, isoproterenol but not zinterol induced more triggered action potentials in afflicted than control animals (16). The present study employs isolated myocytes rather than intact tissue, and measures Ca transients and cell shortening rather than electrical activity. However, the basic phenomenon is clearly reproduced in the single cells even when studied at room temperature (see Figs. 3 and 4). Thus the aftercontractions observed in afflicted cells during isoproterenol exposure at room temperature are completely compatible with results from the earlier study (16) performed at physiological temperature.
The electrophysiological abnormalities in this model have been attributed to β1- rather than β2-receptor mechanisms (16). Although there is general agreement that β1-adrenergic receptors couple to the activation of adenylyl cyclase and the generation of cAMP, the relative role/importance of cardiomyocyte β2-adrenergic receptors as an entrée into the cAMP-generating pathway (and as a mechanism for inotropic support), has been the subject of recent controversy (17). Solid evidence links β2-adrenergic receptors to the promotion of cAMP accumulation in certain cardiomyocyte preparations, but the linkage of β2-adrenergic receptors to global increases in cAMP appears to be highly contextual (17). Our present results indicate that β2-adrenergic receptors markedly elevate intracellular calcium and provide inotropic support in cardiomyocytes from normal and afflicted German shepherd dogs. However, this occurs via a pathway that is not associated with a detectable elevation of cAMP. Although a localized increase in cAMP (confined to the membrane compartment) in response to β2-receptor agonists cannot be excluded, these experiments demonstrate that the β2-adrenergic receptor pathway is distinct from the pathway emanating from β1-adrenergic receptors. Only β1-receptors induce a marked global increase in cAMP formation and lead to pronounced abnormalities of cardiac rhythm. Afflicted German shepherd dogs display excessive sensitivity to β1-adrenergic receptor activation; the β2-adrenergic receptor pathway is not measurably different in tissues from afflicted and control animals.
We previously (15) performed biochemical measurements of β-adrenergic receptor density and activation of adenylyl cyclase in membranes prepared from control and afflicted dogs in an effort to identify the mechanism(s) that mediate enhanced sensitivity to β-adrenergic receptor stimulation. In that study, β-adrenergic receptor density was higher and basal isoproterenol-activated adenylyl cyclase activity was greater in membranes prepared from the vulnerable anteroseptal region of left ventricle of afflicted than the corresponding region of control dogs, suggesting that a defect in β-adrenergic receptor signaling to cAMP formation may contribute to the genesis of arrhythmias in this model (15). An abnormality of rhythm and calcium handling attributable to β1-receptors (not β2) suggests the presence of differences in β1- and β2-receptors signaling to cAMP accumulation. Hence, the studies were extended to discriminate the individual signaling phenotypes of cardiomyocyte β1- and β2-receptors. Because β2-adrenergic receptors constitute the minor subtype in cardiomyocytes, but are the predominant β-receptor subtype in cardiac fibroblasts that contaminate intact tissue preparations, the studies were pursued in isolated cardiomyocyte preparations. This model, which provides the optimal strategy to evaluate integrated signaling from β-adrenergic receptors to adenylyl cyclase, provided evidence that cAMP accumulates to high levels in response to β1-agonists and that the response is equivalent across regions of the ventricle and between control and afflicted dogs. Although cells from the vulnerable region of afflicted dogs display a subtle generalized reduction in cAMP accumulation, including to forskolin (compared with other regions of afflicted or control dogs; see Fig. 1), this biochemical difference is not likely to constitute a mechanism that contributes to the higher incidence of arrhythmias that are initiated in this region. Rather, the lower cAMP levels would more likely result from an associated (or compensatory) abnormality in this region of the ventricle. Whereas these measurements exclude a defect in the pathway for cAMP formation, the measurements of cAMP accumulation were performed in the presence of a phosphodiesterase inhibitor and do not entirely exclude an abnormality in the integration of the cAMP signal in the intact ventricle due to defective cAMP breakdown. Future experiments that consider a potential defect in cAMP catabolism due to reduced phosphodiesterase expression as well as interrogate other molecular targets (ion channels, Ca regulatory proteins) will be key to our understanding of the action potential changes and arrhythmias that develop in afflicted animals. The present observation that β1-receptors (not β2) induce abnormalities of rhythm and calcium handling suggested that β1- and β2-receptors may differ in their signaling to the cAMP pathway and that only the β1-receptor pathway differs between cardiomyocytes from control and afflicted tissues.
Clinical implications and future directions.
This study has likely implications for the catecholaminergic ventricular tachycardias that have been described in human subjects (1, 8). These tend primarily to affect children and adolescents, an age distribution that is similar to that for the occurrence of tachycardias and death in the canine model studied here. The tachycardias are pleomorphic in human subjects and primarily pleomorphic but occasionally monomorphic in dogs (15). In neither human nor canine is there a prolongation of the QT interval on electrocardiogram and neither manifests the pattern typical of torsades de pointes.
A genetic linkage of the catecholaminergic tachycardias has been described in human subjects, involving the ryanodine receptor gene, hRyR2 (13). The linkage of catecholaminergic ventricular tachycardias in humans is to lq42-q43 (18), to which RyR2 maps as well (12, 19). The mechanism for the tachycardia in humans apparently involves hyperphosphorylation of RyR2 via β-adrenergic receptor-dependent stimulation of protein kinase A (9). It has been suggested that the mutation of the RyR2 gene increases its sensitivity to calcium, such that β-adrenergic receptor stimulation leads to calcium overload and tachyarrhythmias (13). Although the arrhythmia in dogs also is inherited, the responsible gene or genes have not been identified. Nonetheless, the involvement of a cAMP/protein kinase A-dependent β1-adrenergic receptor pathway leading to abnormalities in calcium cycling in afflicted animals is consistent with the defect described in human subjects.
This study was supported by National Heart, Lung, and Blood Institute Grant HL-28958.
First published December 13, 2001;10.1152/ajpheart.00871.2001
Address for reprint requests and other correspondence: M. R. Rosen, Center for Molecular Therapeutics, College of Physicians and Surgeons of Columbia University, Dept. of Pharmacology, 630 W. 168 St., PH7W-321, New York, NY 10032 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society