INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

ALTHOUGH THE ORIGINAL Lands classification of -adrenergic receptors (ARs) attributed the regulation of cardiac function to the ₁-AR (21, 22), physiological and radioligand binding studies have since indicated that functional ₁- and ₂-ARs coexist in mammalian heart (2). In human myocardium, ₂-ARs are thought to represent a significant fraction of the total myocardial pool of -ARs, but in other species the contribution of this subtype remains highly controversial. Particularly in tissue such as rodent heart, in which the fraction of ₂-ARs may be comparatively small, methodological considerations can greatly complicate the interpretation of radioligand binding studies. Contamination with nonmyocyte cells and the selection of ligands with inappropriate subtype selectivity can both skew the apparent distribution.

Interest in understanding the potential role of the cardiac ₂-AR in species other than humans is not purely academic; suggestions of signaling responses unique to individual -AR subtypes have relied heavily on work in model systems, including rodent heart. Indeed, although both receptor subtypes are generally thought to share a common signal transduction scheme linked to a rise in intracellular adenosine 3',5'-cyclic monophosphate (cAMP), some important differences have been noted in their respective ability to activate downstream responses. For instance, in dog and human myocardial tissue, the ₁-AR-mediated increase in cAMP has been shown to be much more tightly linked to contractility changes than that mediated by ₂-ARs (17, 30). Although such discrepancies have often been attributed to differential compartmentalization of cAMP accumulated by each subtype, various investigators have also invoked the possible involvement of novel cAMP-independent mechanisms particular to one subtype or another. In isolated rat ventricular myocytes, Xiao et al. (28) asserted that application of the ₂-AR agonist zinterol causes robust inotropic changes that are further potentiated when inhibition by a coactivated pertussis toxin (PTX)-sensitive G protein is masked. No such potentiation by PTX treatment was observed with norepinephrine (NE), an agonist with a higher selectivity for ₁-ARs than ₂-ARs. In earlier work (29), these investigators had reported important differences in the cellular response to subtype-selective stimulation; in particular, a paradoxical absence of lusitropic changes with ₂-AR stimulation. If indeed present, a novel inhibitory signal transduction arm activated by ₂-ARs, but not ₁-ARs, might account for such discrepancies.

Intrigued by these reports, we initially intended to further elucidate the mechanisms by which ₂-ARs apparently target different Ca²⁺ homeostatic pathways in isolated rat ventricular myocytes. However, this effort was frustrated by our inability to detect any ₂-AR-mediated effect. We found that application of the ₂-AR agonist zinterol had only minimal effects on the Ca²⁺ current and intracellular Ca²⁺ concentration ([Ca²⁺]_i) dynamics compared with the effects of the nonselective -AR agonist isoproterenol (Iso). Moreover, even this action was effectively blocked by the addition of the ₁-AR antagonist CGP-20712A. We also found no evidence for the involvement of a PTX-sensitive inhibitory pathway in ₂-AR signaling. Finally, although we could detect a small cAMP rise likely mediated by ₂-ARs, it was more than an order of magnitude smaller than the cAMP rise induced by the nonselective -AR agonist Iso. Thus we found no evidence that ₂-ARs exert a physiologically relevant action on Ca²⁺ homeostasis in rat ventricular myocytes.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Ventricular cardiomyocyte isolation. For electrophysiological and [Ca²⁺]_i studies, ventricular myocytes were isolated from 150- to 250-g Sprague-Dawley rats by an enzymatic dispersion method previously described (11, 25). For cAMP assays, this protocol was modified slightly. After digestion, the tissue was placed in 1 mM Ca²⁺-Tyrode and gently triturated with a wide-bore pipette. Suspended cells were filtered and then allowed two gravity-sedimentation steps (15-20 min each) to separate the denser rod-shaped myocytes from hypercontracted myocytes, other cell types, and debris. All procedures performed on animals were in accordance with institutional guidelines.

Standard extracellular and intracellular solutions. For cardiomyocyte isolation and cAMP assays, a modified Tyrode solution A [in mM: 135.0 NaCl, 5.4 KCl, 1.0 MgCl₂, 0.33 NaH₂PO₄, 10.0 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), and 10.0 glucose] was prepared with a variety of different [Ca²⁺] by the addition of CaCl₂. For electrophysiological experiments, cells were placed in an extracellular solution B (in mM: 140.0 NaCl, 5.0 KCl, 1.0 CaCl₂, 1.0 MgCl₂, 0.33 NaH₂PO₄, 10.0 HEPES, 3.0 CsCl, 5.0 tetraethylammonium chloride, 3.0 4-aminopyridine, and 10.0 glucose) intended to block potassium conductances. Intracellular solutions were also composed to reduce contaminating potassium currents as well as Ca²⁺ current rundown. For those experiments involving Ca²⁺ imaging, the pipette contained intracellular solution A (in mM: 25.0 CsCl, 10.0 NaCl, 120.0 Cs-aspartate, 3.0 Mg-ATP) with 150 µM K₅-fura 2. For those experiments in which [Ca²⁺] measurements were not made, intracellular solution B [in mM: 25.0 CsCl, 95.0 Cs-aspartate, 5.0 Mg-ATP, 0.4 GTP, 10.0 ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid, 5.0 tris(hydroxymethyl)aminomethane (Tris)₂-phosphocreatine, 5.0 pyruvic acid, 2.0, Na₃-fructose-1,6-diphosphate, 1.0 NaH₂PO₄, and 0.5 K-ADP] was used in the pipette.

Fluorescent [Ca²⁺]_i measurements. Fura 2 fluorescence was measured with a high-temporal-resolution microfluorimeter described previously (11, 20). Briefly, the excitation light from a xenon lamp source was rapidly alternated (~110 Hz) between 340 and 380 nm, and the fluorescence emission light (510-550 nm) was quantified with a photomultiplier tube-photon counter interfaced to an IBM-compatible computer. The custom control and data acquisition software stored fluorescence data for each wavelength once per cycle. Raw 340- and 380-nm fluorescence counts were corrected for background, and autofluorescence was measured before the membrane patch was ruptured. The 340/380 excitation fluorescence ratio was used as an index of the cytosolic [Ca²⁺].

Electrophysiological recording. Cells were patch-clamped in the ruptured-patch whole cell configuration (12) with Corning 7052 borosilicate glass electrodes (2-3 M) filled with one of the pipette solutions described in Standard extracellular and intracellular solutions. During experimental protocols, cells were maintained in extracellular solution B at room temperature (~22°C). Voltage clamp was achieved with an AxoPatch 1D or AxoClamp 2A amplifier (Axon Instruments, Foster City, CA). The fast sodium current was inactivated by holding the cell at a resting potential of 40 mV. The inward calcium current was activated by repeated 300-ms depolarizations from 40 to +10 mV (0.5 Hz). Consistent with reported characteristics for this current, it could be eliminated by application of nifedipine or by lowering the extracellular [Ca²⁺] to micromolar levels (not shown). When fura 2 fluorescence was measured simultaneously, membrane current was sampled at 330 Hz (3× acquisition rate of fluorescence). At all other times, membrane current was sampled at 666 Hz.

cAMP measurements. After sedimentation, aliquots of cardiomyocytes were placed in separate tubes containing extracellular solution A and 1 mM Ca²⁺ at room temperature. If cells were to be treated with any receptor antagonist, it was added at this time. After 5 min, the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX; 0.5 mM) was added to each tube. After an additional 5-min incubation, the agonist regimen (if any) was added for the final 5-min incubation. Subsequently, cells were pelleted (<500 revolutions/min) and the buffer was rapidly removed. Cold ethanol (65% vol/vol) was added to stop the reaction. Cells were rigorously vortexed and centrifuged at 4°C for 15 min. The pellet from each tube was collected and solubilized in sodium dodecyl sulfate buffer (10% wt/vol in Tris-buffered saline) for subsequent protein determination with the bicinchoninic acid procedure (Pierce). The supernatant fraction was dried under warm nitrogen (50°C) in borosilicate glass tubes and then stored at 20°C for subsequent cAMP determination. After serial dilution, the quantity of cAMP in each tube was measured with an enzyme immunoassay system (Biotrak/Amersham). The resultant product of the cAMP-linked peroxidase reaction was read at 450 nm on a 96-well microtiter plate spectrophotometer (Bio-Rad). The assay proved to be reasonably robust over a range of 25-3,200 fmol/well.

Reagents. CGP-20712A was kindly donated by Ciba-Geigy (Summit, NJ), and zinterol was kindly provided by Dr. Kenneth Minneman. Fura 2 was obtained from Molecular Probes (Eugene, OR). All other reagents and chemicals were obtained from Sigma (St. Louis, MO).

Statistics. Values are reported as means ± SE. Statistical significance was determined by the Student's t-test.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Zinterol has only modest effects on calcium homeostasis. Recent reports have suggested that activation of ₂-ARs can have significant inotropic effects in isolated rat ventricular myocytes (6, 29) without eliciting accompanying lusitropic effects (29). To better resolve these effects, we compared the acute effects of the ₂-AR agonist zinterol on I_Ca and [Ca²⁺]_i to those elicited by application of the nonselective -AR agonist Iso. Myocytes were voltage-clamped at 40 mV and dialyzed with a pipette solution containing the calcium indicator fura 2 (intracellular solution A). I_Ca was activated by depolarization to +10 mV, and the resultant [Ca²⁺]_i transient was recorded for examination of its amplitude and kinetics.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Isoproterenol (Iso; 1 µM) substantially enhanced L-type calcium current (ICa), intracellular Ca2+ concentration ([Ca2+]i) transient amplitude (as indexed by fura 2 340 nm/380 nm fluorescence ratio), and rate of [Ca2+]i decline. Calcium currents were elicited by a 300-ms depolarization to +10 mV from a holding potential of 40 mV. A: representative current traces before and during isoproterenol application. B: corresponding 340/380 fluorescence ratio.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Zinterol (Zint; 10 µM) modestly enhanced L-type ICa and rate of [Ca2+]i decline but had no appreciable effect on [Ca2+]i transient amplitude. Calcium currents were elicited by depolarization as described in Fig. 1. A: representative current traces before and during Zint application. B: corresponding 340/380 fluorescence ratio measurements.

Blockade of ₁-adrenergic receptors prevents enhancement of I_Ca by zinterol. Although zinterol is known to be selective for ₂- over ₁-ARs, all experiments to this point have employed particularly high concentrations of this agonist. The published affinity of zinterol for ₂-ARs is ~40 nM; its affinity for ₁-ARs is ~1,000 nM (24). Hence, although 10 µM zinterol has been shown to be necessary to elicit a maximal response (29), this concentration of agonist would be expected to result in full occupancy at both receptor subtypes. Because the rat heart has a large ₁-AR reserve, where full activation of only a few percent of the receptor pool may be sufficient to produce a full functional response (3), an appropriate subtype-selective antagonist should be used. We therefore examined the ability of 10 µM zinterol to alter the I_Ca in the presence of 1 µM CGP-20712A, a highly selective ₁-AR antagonist with an affinity of ~3 nM for that subtype (23). This concentration of CGP-20712A should block virtually all ₁-ARs even if, as has been reported by Kitagawa et al. (18), there exists a subpopulation of ₁-ARs with a reduced (~20 nM) affinity for CGP-20712A. On the other hand, blockade of ₂-ARs at this concentration should not be a concern, because the affinity of CGP-20712A at that subtype has been reported to be only ~5-10 µM (18, 23).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Zint (10 µM) had no appreciable effect on ICa after preincubation with 1 µM CGP-20712A (CGP). In this representative example, a myocyte was repetitively depolarized to elicit L-type ICa while bathing solution was rapidly switched from control to drug-containing solutions. A: changes in magnitude of resulting ICa with time; individual points reflect box average of 5 current traces. B: contains representative currents elicited at times a-c indicated in A.

PTX pretreatment does not potentiate enhancement of calcium curent by zinterol. Recently, Xiao et al. (28) proposed that stimulation of ₂-ARs activates a novel signal transduction cascade involving a pertussis toxin-sensitive G protein, an effect not mediated by ₁-AR activation. If so, the small effect of zinterol in our preparation, and its complete sensitivity to CGP-20712A, might be accounted for if this inhibitory pathway was more active in our cell preparation. Because our previous experiments had indicated the enhancement of the I_Ca to be the most evident inotropic action of zinterol, we investigated whether pretreatment of cells with PTX would unmask a distinct ₂-AR mediated action on this current. Myocytes were exposed to PTX (7.5 µg/ml) for at least 2.5 h at 37°C. However, when these cells were challenged with 10 µM zinterol, no potentiation of the response was observed (Fig. 4). In four cells, zinterol increased the magnitude of the I_Ca by 27 ± 3%, an effect virtually identical to that previously observed in non-PTX-treated cells. In contrast, PTX treatment did abolish muscarinic receptor-mediated effects. Figure 5A shows that, as expected, the enhancement of I_Ca by 1 µM forskolin in an untreated cell could be eliminated with the subsequent addition of 10 µM acetylcholine. However, as shown in Fig. 5B, in an otherwise identical cell treated with PTX, the attenuation by acetylcholine was abolished. Thus the failure of PTX to potentiate the enhancement of the I_Ca by zinterol cannot be explained by postulating that our PTX-treatment protocol was ineffective.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Responsiveness to 10 µM Zint was not enhanced by pretreatment with pertussis toxin (PTX). In this representative example, a myocyte was repetitively depolarized to elicit L-type ICa while bathing solution was rapidly switched from control to a Zint-containing solution. A: changes in magnitude of ICa with time; individual points reflect box average of 5 current traces. B: representative currents elicited at times a and b indicated in A.

Intracellular cAMP concentration responses to zinterol with and without ₁-AR blockade. Although the application of 10 µM zinterol produced modest electrophysiological and [Ca²⁺]_i transient changes in rat cardiomyocytes, these responses were prevented by the ₁-AR antagonist CGP-20712A. At this point, therefore, it seemed reasonable to ask whether functional ₂-ARs were even expressed in these cells. We chose to further examine this issue using the change in total cellular cAMP accumulation as a more proximal measure of responsiveness. Recent reports have disagreed as to what degree of intracellular cAMP concentration elevation in rat ventricular myocytes is elicited by ₂-AR activation. For instance, although Kitagawa et al. (18) could not detect any increase in cAMP production with ₂-AR stimulation, Kuznetsov et al. (19) found that, even with blockade of ₁-ARs with 100 nM CGP-20712A, a modest, presumably ₂-AR-mediated rise occurred (to ~2-fold over control) with 100 nM zinterol. Xiao et al. (27) also observed an ~50% increase in cAMP accumulation at 100 nM zinterol (a concentration at which ₁-AR binding should be less than ~10%) and as much as a 70% increase at 10 µM zinterol.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   PTX treatment was sufficient to remove Gi-coupled signal transduction pathways. Success in ablating Gi was tested by examining ability of acetylcholine to attenuate forskolin-stimulation of ICa. A: individual ICa traces in a normal myocyte while in control (C) solution, 1 µM forskolin (F), and 1 µM forskolin with 10 µM acetylcholine (F + A). B: corresponding responses to these drugs in a myocyte pretreated with PTX.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Although radioligand binding studies have long suggested the presence of a heterogeneous population of -ARs in whole rat myocardium (see Ref. 24), Buxton and Brunton (4) found only a homogeneous ₁-AR population in enzymatically isolated rat ventricular myocytes. Later, using markers for endothelial and myocyte plasmalemma, Freissmuth et al. (9) extended this work to conclude that most of the ₂-ARs of rat and guinea pig heart were localized to the coronary vessels. Unfortunately, the possibility of a small pool of ₂-ARs below the limit of detection (perhaps <10%) could not be totally eliminated in the latter two studies because of the unavailability of sufficiently subtype-selective ligands. A more recent study by Kitagawa et al. (18) employed better pharmacological tools, in particular, the ₁-AR antagonist CGP-20712A, which is at least 1,000× more selective for ₁-ARs than for ₂-ARs. They concluded that ₂-ARs could not be detected on rat ventricular myocytes, either by quantitative radioligand binding or by cAMP assay. These conclusions are in disagreement with other recent reports (6, 7, 19; CGP-20712A was used in Refs. 6 and 19) that estimated ₂-ARs to comprise 8-20% of the total -AR population of isolated rat myocytes. We note that, although all of these studies used enzymatically dispersed cardiomyocytes, Kitagawa et al. (18) performed density gradient centrifugation after cell isolation to remove nonmyocyte cells, a more effective selection procedure than the more commonly employed gravity sedimentation procedure.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Comparison of cAMP accumulation over control in myocytes exposed to various -adrenergic receptor ligands: 10 µM Zint, 1-AR antagonist CGP, Zint + CGP, 1 µM Iso, or 1 µM norepinephrine (NE). Each value is mean ± SE of n experiments. Difference from control (*) or CGP () statistically significant (P < 0.05).

Although we obviously cannot resolve the issue of whether ₂-ARs are present in isolated rat ventricular sarcolemma, our findings are not consistent with the presence of physiologically relevant ₂-ARs capable of modulating [Ca²⁺]_i homeostasis. In contrast to results with the nonselective -AR agonist Iso, we found that the application of 10 µM zinterol, a concentration that should bind to >99% of all ₂-ARs, induced only a minimal enhancement of the I_Ca and left the [Ca²⁺]_i transient amplitude essentially unaffected. Furthermore, we observed no potentiation of the zinterol-mediated effect on the I_Ca after treatment with PTX, leading us to conclude that there was no coactivation of a G_i- or G_o-linked inhibitory signal transduction pathway by zinterol via any receptor type. It should be noted that this failure of PTX to potentiate the I_Ca response to zinterol is somewhat surprising, even if the agonist was not directly activating a G_i-linked pathway. Indeed, other investigators (14) have noted some potentiation in the enhancement of the I_Ca by Iso, an effect generally attributed to inhibition of tonically active G_i and not to direct activation of PTX-sensitive G proteins by -ARs. Perhaps the basal response to zinterol is just too small to permit any appreciable potentiation on removal of tonic inhibition by G_i. Finally and most significantly, we found the very modest I_Ca modulation by zinterol to be entirely sensitive to CGP-20712A, demonstrating that the cellular responses we observed were in fact mediated by the ₁-AR subtype. Indeed, our own conclusions are entirely consistent with the "classical" view of adrenergic pharmacology in which the ₂-AR plays, at best, only a minimal direct role in the regulation of cardiac contractility.

As noted previously, zinterol can bind to ₁-ARs, although with a 25-fold lower affinity than for ₂-ARs (24). Additionally, numerous studies have also indicated that zinterol (as well as many other agents classified as ₂-AR agonists) can activate cardiac ₁-ARs, albeit less strongly than full ₁-AR agonists. The earliest such report was by Freyss-Beguin et al. (10). These investigators examined the stimulation of intrinsic beating rate and cAMP production in cultured rat heart cells with a battery of presumptively ₂-AR- selective agonists, including zinterol, and found that their dose-response relations corresponded better with expectations for activation of the ₁-AR than the ₂-AR subtype. More recently, Kuznetsov et al. (19) concluded that high concentrations of zinterol were capable of accumulating cAMP in adult rat cardiomyocytes through both subtypes. A number of investigators have also reached a similar conclusion regarding the ability of zinterol to act as an agonist at ₁-ARs in other systems (13, 15, 16).

In the presence of 10 µM zinterol, the concentration used by Xiao and Lakatta (29), Cerbai et al. (6), Kuznetsov et al. (19), and ourselves, essentially every ₁- and ₂-AR should be occupied by the agonist. Especially given the very large ₁-AR-receptor reserve in rat myocardium, one might reasonably expect at least a component of the response to this agonist to be mediated by the ₁-AR subtype. Indeed, it is hard not to notice that the dose-response relation for zinterol enhancement of rat cardiomyocyte twitch amplitude reported by Xiao and Lakatta (Fig. 2 of Ref. 29) closely matches its predicted binding to ₁-ARs rather than to ₂-ARs. It is therefore perplexing that these investigators reported that the enhancement of twitch and [Ca²⁺]_i amplitude elicited by 10 µM zinterol were completely resistant to antagonism by 300 nM CGP-20712A. We are at a loss to explain this observation, given that the zinterol-mediated response in our hands was entirely sensitive to 1 µM CGP-20712A, a concentration of the antagonist that should reduce occupancy of ₁-ARs by zinterol to <3% although leaving >99% of ₂-ARs still bound to zinterol.

Although we did find that a component of the zinterol-mediated cAMP increase was not CGP-20712A sensitive, this observation should be kept in perspective. The 10 µM zinterol-stimulated cAMP accumulation was more than 1 order of magnitude smaller than that stimulated by 1 µM Iso, a finding that confirms quantitatively similar observations by Kuznetsov et al. (19). Thus, even allowing for a ₂ component of the Iso response, we are confident in concluding that ₁-ARs are clearly much more effective than ₂-ARs receptors in stimulating cAMP production. In contrast, Xiao et al. (27) concluded that these two receptor subtypes were equally effective in raising total cAMP levels (by ~50-70%) on the basis of a comparison of the response to zinterol with that to NE, which they asserted was acting almost exclusively via ₁-ARs. In part, the difference between their results and ours with respect to both the absolute magnitude of the cAMP change and the magnitude of the -AR subtype differential can be accounted for by differences in technique, particularly our use of a phosphodiesterase inhibitor (0.5 mM IBMX). However, an additional key difference was the method of ₁-AR activation, in particular, the assumption by Xiao et al. (27) that NE was a pure ₁-AR agonist. Although NE has a higher affinity for ₁- than ₂-ARs, it also activates -ARs. It has previously been shown that the cAMP response to NE in rat heart is substantially attenuated by its coactivation of ₁-ARs (1, 5).

Therefore, although we did observe a modest CGP-20712A-resistant cAMP elevation to zinterol, we conclude that this extremely small effect was simply insufficient to mediate any significant effect on Ca²⁺ homeostasis. Indeed, we were unable to identify a specifically ₂-AR-mediated enhancement of any [Ca²⁺]_i homeostatic mechanism, either in control or PTX-treated cardiomyocytes. Hence, although other investigators have established that activation of myocyte ₂-ARs does affect contractility by altering [Ca²⁺] dynamics in humans (8), we cannot confirm that this subtype plays a similar role in rat ventricular myocytes.