Abstract
A receptor can be activated either by specific ligand-directed changes in conformation or by intrinsic, spontaneous conformational change. In the β2-adrenergic receptor (AR) overexpression transgenic (TG4) murine heart, spontaneously activated β2AR (β2-R*) in the absence of ligands has been evidenced by elevated basal adenylyl cyclase activity and cardiac function. In the present study, we determined whether the signaling mediated by β2-R* differs from that of a ligand-elicited β2AR activation (β2-LR*). In ventricular myocytes from TG4 mice, the properties of L-type Ca2+ current (ICa), a major effector of β2-LR* signaling, was unaltered, despite a 2.5-fold increase in the basal cAMP level and a 1.9-fold increase in baseline contraction amplitude as compared with that of wild-type (WT) cells. Although the contractile response to β2-R* in TG4 cells was abolished by a β2AR inverse agonist, ICI118,551 (5 × 10−7 M), or an inhibitory cAMP analog, Rp-CPT-cAMPS (10−4 M), no change was detected in the simultaneously recorded ICa. These results suggest that the increase in basal cAMP due to β2-R*, while increasing contraction amplitude, does not affect ICa characteristics. In contrast, the β2AR agonist, zinterol elicited a substantial augmentation of ICa in both TG4 and WT cells (pertussis toxin-treated), indicating that L-type Ca2+channel in these cells can respond to ligand-directed signaling. Furthermore, forskolin, an adenylyl cyclase activator, elicited similar dose-dependent increase in ICa amplitude in WT and TG4 cells, suggesting that the sensitivity of L-type Ca2+channel to cAMP-dependent modulation remains intact in TG4 cells. Thus, we conclude that β2-R* bypasses ICa to modulate contraction, and that β2-LR* and β2-R* exhibit different intracellular signaling and target protein specificity.
β-adrenergic receptor (AR) stimulation plays a prominent role in modulation of cardiac myocyte performance in response to an increased peripheral demand. Driven by sympathetic neurotransmitters and adrenal hormones, βAR activation regulates virtually all major steps of the cardiac cell excitation-contraction (E-C) coupling cascade, e.g., the sarcolemmal L-type Ca2+ current (ICa), sarcoplasmic reticulum (SR) Ca2+ release and reuptake, and the responsiveness of contractile myofilaments to cytosolic Ca2+. Because ICa provides the trigger for SR Ca2+ release, and is a major determinant of intracellular calcium homeostasis, modulation of this current by βAR system has been extensively studied over the last two decades. It has been demonstrated that both β1AR and β2AR subtypes coexist in cardiac myocytes in many mammalian species, and that stimulation of each of these receptor subtypes increases cardiac ICa (Xiao and Lakatta, 1993; Cerbai et al., 1995) through the classic stimulatory G protein (Gs)-adenylyl cyclase-cAMP-protein kinase A (PKA) signaling cascade (Hartzell et al., 1991; Zhou et al., 1997; Skeberdis et al., 1997; Xiao et al., 1999). The existence and functional importance of a more rapid, direct interaction of the βAR-activated Gs and L-type Ca2+ channel remain controversial (Yatani and Brown, 1989; Hartzell et al., 1991;Zhou et al., 1997; Skeberdis et al., 1997).
A prevailing receptor theory (two-state model) states that a G protein-coupled receptor, such as β1AR or β2AR, exists in an equilibrium between two conformational states: an inactive (R) state and an active (R*) state, the latter having high affinity for G proteins (Bond et al., 1995). In the absence of a receptor agonist, spontaneous transition between the R* and R states results in a constitutive or intrinsic activation of only minority of receptors (Chidiac et al., 1994; Bond et al., 1995) and thus the functional significance of R* is not always evident. The presence of a large number of spontaneously activated β2ARs (β2-R*s), which alter basal function, has been experimentally demonstrated in a transgenic (TG) murine model, the TG4 mouse (Milano et al., 1994; Bond et al., 1995; Xiao et al., 1999), in which the human β2AR is overexpressed by ∼200-fold in a cardiac-specific manner. Hence, this transgenic model provides a unique opportunity to study the transmembrane signal transduction originating from unliganded β2-R* in comparison with that from the ligand-activated β2AR (β2-LR*). According to the two-state receptor model, β2-R* ought to be identical with β2-LR*, because there is only a single active conformational state. However, there is no a priori reason that this has to be the case. By analogy to ionic channels and enzymes, it is more plausible that a receptor may possess multiple, distinct active conformations (Perez et al., 1996; Gurdal et al., 1997). If β2-R* and β2-LR* differ in their active conformational states, spontaneous and agonist-induced β2-adrenergic signaling may not be functionally equivalent, e.g., in modulating their target proteins, such as L-type Ca2+ channels.
In the present study, we examined the possible modulatory effects of β2-R* on basal ICa and cell contraction in single ventricular myocytes and on basal cAMP in myocardium from TG4 mice and wild-type (WT) littermates. Surprisingly, we found no evidence that ICa was regulated by β2-R* in TG4 heart cells. In contrast, both β2-LR* signaling in the presence of pertussis toxin (PTX) and direct adenylyl cyclase activation by forskolin augmented ICa to an extent similar to that observed in WT cells. Our results support the idea that despite many similarities, β2-R* and β2-LR* may represent distinct functional conformation states of the receptor, eliciting different intracellular signaling patterns, and having differential effects on target proteins. These findings require an extension of the current model of β2AR to encompass multiple active conformational states.
Experimental Procedures
Cell Isolation and Measurement of Contraction.
Single murine cardiac myocytes were isolated from the hearts of 2- to 3- month-old mice via a standard enzymatic technique (Korzick et al., 1997). Briefly, hearts were retrogradely perfused with collagenase B and protease using the Langendorff method. Cells were shaken loose from the heart after this perfusion and then suspended in HEPES buffer solution consisting of: 1 mM CaCl2, 137 mM NaCl, 5.4 mM KCl, 15 mM dextrose, 1.3 mM MgSO4, 1.2 mM NaH2PO4, and 20 mM HEPES, pH 7.4, adjusted with NaOH. Ca2+ tolerant cells were kept at 37°C, with or without incubation with 1.5 μg/ml PTX for at least 3 h, as described previously (Xiao et al., 1995).
Cells were placed on the stage of an inverted microscope (Zeiss, model IM-35; Carl Zeiss, Thornwood, NY) and superfused with HEPES-buffered solution at a flow rate of 1.8 ml/min. Each cell was illuminated with red (650–750 nm) light through the normal brightfield path of the microscope and field stimulated at 0.5 Hz at 23°C. Cell length was monitored from the brightfield image by an optical edge tracking method using a photodiode array (model 1024 SAQ;, Reticon) with a 3-ms time resolution (Spurgeon et al., 1990).
Criteria for viable mouse myocytes have been described in a previous report (Korzick et al., 1997), i.e., 1) rod shape; 2) clearly defined sarcomeric striations; 3) a clear negative staircase after rest for a period of ∼1 min; and 4) a stable steady-state contraction amplitude for at least 5 min before drug administration.
Ca2+ Current Measurement.
ICa was measured via the whole-cell patch clamp technique using an Axopatch 1D amplifier (Axon Instruments Inc., Foster City, CA). Low-resistance (1–2 MΩ) micropipettes were pulled via a two-stage micropipette puller (model P-97; Sutter Instrument Co., Novato, CA). The average series resistance (Rs) in whole-cell configuration was 5.71 ± 0.28 MΩ for TG4 cells (n = 34) and 5.99 ± 0.39 MΩ for WT cells (n = 25), and routinely compensated ∼70% in our experiments. To selectively examine ICa, cells were voltage-clamped at −40 mV to inactivate the sodium and T-type Ca2+ channels. Potassium currents were inhibited by appropriate blockers in the extracellular HEPES buffer solution (4 mM 4-aminopyridine, 5.4 mM CsCl substituted for KCl in standard HEPES buffer solution) and in the pipette solution containing: 100 mM CsCl, 10 mM NaCl, 20 mM tetraethylammonium chloride 20, 10 mM HEPES, 5 mM MgATP, and 5 mM EGTA; pH was adjusted to 7.2 with CsOH. In some experiments to simultaneously record ICa and cell contraction, EGTA was omitted from the pipette solution and normal HEPES buffer constituted the extracellular solution. ICa was elicited by 300-ms pulses from a holding potential of −40 mV to test potentials from −30 to +50 mV in 10-mV increments at 0.1 Hz at 23°C. To monitor drug effects, ICa elicited by a depolarization from −40 to 0 mV was continuously recorded. The amplitude of ICa was measured as the difference between the peak inward current and that at the end of 300-ms pulse. The decay of ICa was fitted to a biexponential function:
To determine whether there is a current-voltage (I-V) shift, the voltage-dependence of ICa steady-state activation was calculated from the equation:
Measurement of cAMP Accumulation.
Cardiac membranes were prepared as previously described (Xiao et al., 1998). cAMP levels were assayed by the radioimmunoassay. Briefly, 10 μl of membrane vesicles (20 μg total protein) was added to a 40-μl reaction mixture to make a final concentration of 4 mM Tris-EDTA and 10 μM Ro 20–1724 (an inhibitor of phosphodiesterase IV) with or without 0.5 μM ICI 118,551 (ICI is a β2AR inverse agonist). The reaction was performed for 15 min at 37°C and 25 μl of supernatant was assayed using a cAMP 3H assay kit obtained from Amersham (Arlington Heights, IL). Protein was measured using the Bradford method (Bio-Rad, Richmond, CA) with BSA as the standard.
Materials.
PTX, tetrodotoxin, forskolin, isoproterenol hydrochloride, norepinephrine (NE), prazosin, and Ro 20–1724 were purchased from Sigma Chemical Co. (St. Louis, MO). Rp diastereomers of 8-(4-chlorophenylthio)-cAMP (Rp-CPT-cAMPS) was purchased from Biolog Life Science Institute (La Jolla, CA). cAMP assay kits were purchased from Amersham. Zinterol was kindly supplied by Bristol-Myers (Evansville, IN); ICI was kindly supplied by Imperial Chemical Industry (London, United Kingdom). CGP20712A (CGP) was kindly supplied by Ciba-Geigy Corp. (Basel, Switzerland).
Data Analysis.
Data are reported as mean ± S.E.M. Student’s t test was used to test for differences between TG4 and WT groups and for PTX-treated and nontreated groups; a pairedt test was used for assessing the significance of drug effects. A value of P < .05 was considered to be statistically significant.
Results
In the absence of exogenous β2AR agonists, the basal cAMP level was increased by 2.5-fold in TG4 relative to WT hearts (Fig. 1A). Concomitantly, basal contraction amplitude was enhanced by 1.9-fold in single ventricular myocytes isolated from TG4 mice (Fig. 1B). A β2AR inverse agonist, ICI (5 × 10−7 M), which had no significant effect on either basal cAMP or contractility in WT mice, reduced the baseline cAMP (Fig. 1A) and contractility of TG4 cells (Fig. 1B) to levels similar to those of WT littermates. These data are in agreement with previous observations that ICI depresses the elevated basal adenylyl cyclase activity, heart rate and cardiac contractility in vivo and in isolated atria (Milano et al., 1994; Bond et al., 1995; Du et al., 1996). Therefore, the results so far support the notion of spontaneous β2AR activation in the absence of an agonist (Chidiac et al., 1994; Milano et al., 1994; Bond et al., 1995; Xiao et al., 1999) and indicate that β2-R* augments cAMP production and cardiac contractility, as is the case for ligand-induced β2AR stimulation (Xiao and Lakatta, 1993; Xiao et al., 1994, 1995; Altschuld et al., 1995; Zhou et al., 1997). If β2-R* and β2-LR* were functionally equivalent, as predicted by the two-state model, the L-type Ca2+ channel, a key target effector of β2-LR* signaling, would be modulated by β2-R* in a similar fashion, i.e., baseline ICa in TG4 cells would be expected to be tonically elevated and sensitive to ICI. To our surprise, basal ICa was not elevated in TG4 cells (see below). Furthermore, although ICI (5 × 10−7 M) rapidly and reversibly attenuated the augmented baseline contraction amplitude in TG4 ventricular myocytes (Fig.2A), it had virtually no effect on the amplitude (Fig. 2B; 97.2 ± 3.4% of control, n = 9) and time course (Fig. 2C) of ICa in TG4 cells. This result was further confirmed by the simultaneous recording of ICa and contraction using the EGTA-free pipette solution. As shown in Fig. 3, ICI induced a marked decrease in cell contraction amplitude without any change of ICa in the same TG4 cell.
The differential effects of β2-R* on ICa and contractility are in sharp contrast to the traditional views that the L-type Ca2+channel is an obligatory effector of β2AR signaling (Xiao and Lakatta, 1993; Cerbai et al., 1995; Altschuld et al., 1995; Zhou et al., 1997). The results also raise doubts as to whether the β2-R* effect to augment contractility in TG4 myocytes even requires the classical cyclase-cAMP-PKA signaling. To directly address this issue, we used an inhibitory cAMP analog, Rp-CPT-cAMPS, to specifically block PKA activation. As shown in Fig. 3, similar to the effect of the inverse agonist ICI, Rp-CPT-cAMPS reversed the β2-R* effect on contraction without affecting the simultaneously recorded ICa. This observation indicates that the β2-R*-stimulated inotropic effect in TG4 cells depends largely on β2-R*-elicited cAMP signaling, as does β2-LR* (Zhou et al., 1997; Skeberdis et al., 1997; Xiao et al., 1999). Thus, the inability of β2-R* to modulate L-type Ca2+ channels may be attributed to either a qualitative difference between β2-R* and β2-LR*, or to an alteration in L-type Ca2+ channels of TG4 cells (see below).
To further characterize the L-type Ca2+ channel properties in TG4 cells, whole-cell ICaamplitude, current-voltage relation, and inactivation kinetics were systematically examined in both TG4 and WT ventricular myocytes. Figure4A shows typical traces of ICa elicited by a depolarization from −40 to 0 mV in a WT and a TG4 myocyte in the absence of any β2AR ligands. The baseline ICa in TG4 and WT cells are virtually indistinguishable in amplitude and time course (Fig. 4A), consistent with the absence of ICI-sensitive (β2-R*) component of ICa described above. The average amplitude of ICa at 0 mV was 1.01 ± 0.05 nA in TG4 (n = 34) and 1.03 ± 0.07 nA in WT cells (n = 38). Rundown of ICa was not significantly different between these two groups (12.4 ± 4.9 and 14.1 ± 6.2% at 10 min for TG4 and WT cells, respectively;n = 3 for both groups). Because there was no significant difference in cell membrane capacitance (166 ± 10 pF,n = 34, in TG4 cells versus 161 ± 12 pF,n = 38, in WT cells), the density of ICa (i.e., ICa normalized by capacitance) was also similar in TG4 and WT groups (6.73 ± 0.43 pA/pF, n = 34 and 6.86 ± 0.49 pA/pF,n = 38, respectively). The similarity in membrane capacitance between TG4 and WT cells is consistent with a previous report that no cellular hypertrophy occurs in 2- to 4-month-old TG4 hearts (Milano et al., 1994; Xiao et al., 1999).
Next, we determined the current-voltage relation of ICa in both TG4 and WT myocytes. Cells were depolarized from a holding potential of −40 mV to various test potentials from −30 to +50 mV in 10-mV increments. Over the entire voltage range examined, the ICa density-voltage relations in TG4 and WT cells overlapped (Fig. 4B), indicating that voltage-dependent activation of L-type Ca2+channel in TG4 cells was unchanged as compared with WT controls. Furthermore, ICa inactivation time constants (τf and τs) and the voltage-dependence of τf or τs of WT cells were similar to those of TG4 cells (Fig. 4C); likewise, there is no difference in the amplitude proportion of the two exponential components between these two groups (Af/As = 1.24 ± 0.08 at 0 mV, n = 20, in TG4 versus 1.19 ± 0.16,n = 19, in WT). Therefore, no measured parameters of ICa, including amplitude, voltage-dependence, and inactivation kinetics were altered by spontaneous β2AR activation in TG4 cardiac myocytes.
If L-type Ca2+ channels in TG4 cells were somehow modified via compensatory mechanisms so that ICacould no longer respond to β2-R*-mediated cAMP signaling, the ICa response to any other cAMP signaling should be similarly blunted. However, forskolin, an activator of adenylyl cyclase, induced a robust increase in the Cd2+-sensitive ICa in TG4 cells (Fig. 5, A and B). More importantly, the dose-response curves of ICa to forskolin in TG4 and WT cells virtually overlapped, with no significant difference in EC50 (3.97 × 10−7 M for WT and 5.96 × 10−7 M for TG4; P > .05, Fig.5C). Thus, the sensitivity of cardiac L-type Ca2+channel to cAMP-PKA modulation remains intact in TG4 mice.
Our recent studies have shown that cardiac β2AR couples to the PTX-sensitive inhibition proteins, (Gi) Gi2 and Gi3 (Xiao et al., 1995, 1999), and that this coupling partially offsets the β2AR agonist-mediated contractile response in rat myocytes (Xiao et al., 1995) and completely negates the β2AR agonist-mediated contractile (Xiao et al., 1999) and ICa responses (Fig. 5, A and B) in TG4 and WT murine ventricular myocytes. Therefore, it is reasonable to assume that an excessive Gi coupling to β2-R* could be involved in the inability of β2-R* to modulate ICa. To test this hypothesis, baseline ICa was re-examined in PTX-treated cells and compared with that in PTX-untreated cells. Figure 6B shows that in TG4 cells, PTX treatment had no significant effect on the baseline ICa amplitude or its current-voltage relation. Similar results were also obtained in WT cells (Fig. 6A). Moreover, even in PTX-treated TG4 cells, neither the amplitude nor the kinetics of the basal ICa were affected by ICI (data not shown). These results suggest that Gi proteins are not involved in the unresponsiveness of ICato β2-R*.
Although Gi inhibition failed to rescue ICa response to β2-R*, in the same TG4 cells, PTX permitted β2-LR* induced by zinterol to significantly enhance ICa (Fig. 6, C and D). The PTX rescued ICa response to β2-LR* in TG4 cells (149 ± 12% of control, at 0 mV, n = 8) was comparable with that of WT cells (153 ± 11% of control, at 0 mV,n = 4). In addition, the ICa-voltage relation was shifted leftward by zinterol (V1/2 was −16.58 ± 1.33 and −23.03 ± 1.67 mV in the absence and presence of zinterol, respectively, P = .01, Fig. 6C), in agreement with previous observations in rat ventricular myocytes (Xiao and Lakatta, 1993). However, neither the inactivation kinetics (τf, 101 ± 8% of control, τs, 108 ± 3% of control), nor the ratio of Af/As (95 ± 19% of control, n = 5) were significantly altered by zinterol in PTX-treated TG4 cells. Figure 6D shows that the ICa response to zinterol in a PTX-treated TG4 cell was completely blocked by the β2AR-selective antagonist, ICI at 5 × 10−7 M (96.2 ± 6.2% of control,n = 5, P > .05 versus control). Thus, PTX treatment permits β2-LR*, but not β2-R*, to modulate L-type Ca2+ channel activity in TG4 heart.
Although in mouse cardiac myocytes β1-AR is unable to couple to Gi proteins, as manifested by the G protein photoaffinity labeling profile (Xiao et al., 1999), previous studies in guinea pig (Hool and Harvey, 1997) raised doubt as to whether the PTX rescued effect of zinterol is related to the activation of β1AR. We therefore examined the effect of β1AR stimulation in the presence and absence of PTX treatment in TG4 myocytes. Interestingly, β1AR agonist NE even at maximal concentration (NE 10−7 M) plus prazosin 10−6 M (Korzick et al., 1997) did not induce a discernible increase in ICa of TG4 cells, whereas it markedly increased ICa in WT myocytes (Fig.7, A and B). The absence of ICa response to β1AR stimulation is consistent with previous observations on the loss of contractile response to β1AR stimulation by either NE plus prazosin or isoproterenol plus the β2AR blocker, ICI (Bond et al., 1995; Du et al., 1996). Whereas PTX treatment fully rescued the contractile (Xiao et al., 1999, also see Fig. 7C) and ICa (Fig. 6) response to β2AR agonist stimulation, it was unable to restore contractile and ICa response to β1AR stimulation (Fig. 7). In addition, in TG4 cells, the PTX-restored contractile response to a mixed βAR agonist, isoproterenol 10−6 M, was specifically inhibited by a β2AR antagonist, ICI 10−7 M, but not by a β1AR antagonist, CGP 3 × 10−7 M (Fig. 7C). This further corroborates our previous notions that, unlike β2AR, β1AR does not couple to Gi protein(s) in mouse myocardium (Xiao et al., 1999).
Discussion
β2-R* Does Not Regulate ICa.
The presence of β2-R* in the TG4 heart is evidenced by the elevated basal adenylyl cyclase activity (Milano et al., 1994) and cAMP production (Fig. 1A), the enhanced cardiac contractility (Milano et al., 1994; Bond et al., 1995; Du et al., 1996; Rockman et al., 1996; Xiao et al., 1999) (Fig. 1B), and the blockade of these augmentations by the inverse β2AR agonist, ICI (Milano et al., 1994; Bond et al., 1995; Du et al., 1996; Xiao et al., 1999) (Figs. 2 and 3). In the present study, we have provided direct evidence that β2-R*-mediated modulation of cardiac contractility is largely cAMP-PKA-dependent, because it is sensitive to the PKA inhibitor Rp-CPT-cAMPS (Fig. 3). The most surprising and unexpected finding of this study is that baseline ICa in TG4 cardiac myocytes is not increased or altered by β2-R* (Fig. 4). The simplest explanation for this observation would be that β2-R*-directed signaling is totally diverted from the L-type Ca2+ channels. However, the interpretation for the results obtained from the transgenic model may not be so straightforward, because compensatory changes have been documented in TG4 hearts, e.g., down-regulation of the SR protein phospholamban (PLB) (Rockman et al., 1996) and up-regulation of Gi proteins (R-P.X., unpublished data). Several additional experiments have therefore been undertaken to explore alternative possibilities.
If the L-type Ca2+ channel protein expression were reduced in TG4 heart cells so that ICadensity in these cells was lower than normal in the absence of β2-R*, it could mask a β2-R*-mediated stimulatory effect on ICa. In other words, an adaptive “down-regulation” of ICa might offset an increase in this current induced by β2-R*. This possibility was tested by using the inverse β2AR agonist, ICI. Because ICI inactivates β2-R* and prevents spontaneous β2AR activation (Bond et al., 1995), the ICI-sensitive component would thus reflect the magnitude of the β2-R* effect. We have found that ICI has no detectable effect on ICa, although it markedly reduces basal cell contractility and cAMP content (Figs. 1-3). Thus, our results do not support an adaptive reduction in L-type Ca2+ channel number in TG4 mice.
A second possible explanation for the absence of enhancement of ICa in TG4 cells is that L-type Ca2+ channels might be somehow modified, thereby losing their sensitivity to cAMP-dependent modulation. If this were the case, ICa should no longer respond to any other cAMP-dependent stimulation, or the responses should be markedly attenuated. This possibility, however, have also been excluded on the basis that agonist-elicited β2AR stimulation enhances ICa (in PTX-treated TG4 myocytes) to an extent similar to that in (PTX-treated) WT cells; and that the ICa (in TG4 cells) dose-response curve to the adenylyl cyclase activator forskolin overlaps with that in WT cells (Fig. 5C), indicating that the responsiveness of L-type Ca2+ channels to cAMP-PKA-dependent regulation in TG4 cells is not significantly altered. Thus, the unresponsiveness of ICa to β2-R* is not caused by the changes in the channel proteins.
In mammalian hearts, agonist-elicited β2AR stimulation evokes bifurcated Gs- and Gi-mediated signaling cascades: the β2AR-Gi pathway exerts a negative feedback control of the β2AR-Gs effects (Xiao et al., 1995, 1999; Zhou et al., 1997). The Gi-mediated inhibition of Gs signaling could account for the apparent uncoupling of β2-LR* to L-type Ca2+ channel in non-PTX-treated WT and TG4 cells, because PTX unmasks a de novo ICa response to β2AR agonist zinterol (Fig. 6, C and D), and the β2AR agonist zinterol enhances the photoaffinity labeling of the α subunits of the Gi proteins, Gi2 and Gi3 (Xiao et al., 1999). However, Gi-mediated inhibition cannot explain the inability of β2-R* to augment ICa in TG4 cells, because PTX fails to potentiate basal ICa(Fig. 6B), and ICI has no effect on the baseline ICa regardless of PTX (Figs. 2 and 3). These functional results suggest that β2-R* does not couple to Gi proteins as efficiently as does β2-LR*. This is in good agreement with the fact that in transgenic mice with high or medium levels of β2AR overexpression, β2AR in the absence of an agonist, coprecipitates with Gs but barely with Gi/Go (Gurdal et al., 1997). Taken together, we conclude that spontaneous β2AR activation in TG4 cells, whereas increasing cell contractility, does not regulate ICa, a key effector of β2-LR*.
Differences between β2-R*- and β2-LR*-Mediated Signaling.
In contrast to the prediction of the two-state receptor model, the differential regulation of ICa by β2-R* and β2-LR* suggests that the liganded and unliganded active β2ARs are different active receptor species, likely having different conformations and initiating distinct postreceptor signaling pathways. Several lines of additional evidence support this hypothesis. First of all, whereas β2-R* in TG4 heart significantly increases the baseline contractility, β2-LR* induced by zinterol or isoproterenol at maximal concentrations are unable to further increase contraction amplitude (Milano et al., 1994;Du et al., 1996; Xiao et al., 1999), even though the basal contractility is not at the maximum contractile state yet (Du et al., 1996; Xiao et al., 1999). Secondly, β2-R*, unlike β2-LR*, does not couple to Gi proteins, as reflected by the lack of a PTX effect on the basal ICa (Fig. 6) and by immunoprecipitation data on receptor-G protein interaction (Gurdal et al., 1997). Finally, it has recently been shown that in rat and mouse cardiac myocytes, multiple active conformational states of β2AR can be induced by different β2AR ligands (R-P.X., unpublished data). Similar observations have been reported previously for β2AR and other G protein-coupled receptors in transfected cells (e.g., Eason et al., 1994) or artificial lipid vesicles (Gether et al., 1997). The present finding that β2-R* differs from β2-LR* is in general agreement with the emerging concept of multiple active receptor states for a given receptor.
Another intriguing difference between β2-R* and β2-LR* is manifested by their chronic noncontractile effect. Agonist-induced, chronic, mixed βAR or β2AR stimulation has been shown to enhance cardiac cell growth in vitro (Boluyt et al., 1995; Zhou et al., 1996) and cause cardiac hypertrophy in vivo (Kudej et al., 1997). Cardiac hypertrophy also occurs in other transgenic murine models in which Gs or the cAMP signaling cascade has been genetically up-regulated (Iwase et al., 1996). In contrast, the TG4 model is exceptional in that it has tonically elevated cardiac contractile function and cAMP signaling without evident cardiac and cellular hypertrophy as shown in the present and previous studies (Milano et al., 1994; Xiao et al., 1999, Heubach et al., 1999). Given the central role of sarcolemmal ICa in intracellular Ca2+ homeostasis, and given the role of Ca2+ signaling in cell hypertrophy in vivo and in vitro (Molkentin et al., 1998), it is tempting to speculate that the lack of L-type Ca2+ current response to β2-R*, as demonstrated here, may be of particular relevance to the lack of cardiac hypertrophy and cardiomyopathy in the TG4 model.
The present results also illustrate that, although both β2-LR* (Xiao et al., 1999) and β2-R* (Fig. 3) couple to cAMP-dependent signal transduction pathway, their cAMP signaling may be differentially compartmentalized. Specifically, the cyclase activity or cAMP-PKA signal due to β2-R* must be somehow shielded from L-type Ca2+ channels, but is readily accessible to other E-C coupling machineries. In contrast to β2-R*, previous studies in many species (rat, mouse, and dog) have shown that, L-type Ca2+channel is the major target protein of β2-LR*, whereas the SR and other cytosolic proteins do not always respond to β2-LR*-stimulated cAMP-PKA signaling (Xiao et al., 1994; Altschuld et al., 1995; Kuschel et al., 1999b). Thus, β2-R* differs qualitatively from β2-LR*; this difference might not be simply explained by different coupling efficiency to various targets. Taken together, not only the receptor type or subtype (e.g., Zhou et al., 1997), but also the conformational state of the same receptor is an important determinant of intracellular sorting of cAMP signaling. Selective shielding of cAMP signaling from a subset of target proteins implies that an additional counteracting mechanism(s) must be simultaneously engaged. In this respect, we have shown, in rat and dog, that the β2-LR*-Gisignaling pathway can fully antagonize the β2-LR*-Gs- cAMP-mediated effects in the bulk cytosolic compartment (Xiao et al., 1994; Altschuld et al., 1995; Kuschel et al., 1999a); but not in the vicinity of L-type Ca2+ channel (Xiao and Lakatta, 1993; Altschuld et al., 1995; Xiao et al., 1995; Zhou et al., 1997; Kuschel et al., 1999b). In the mouse, β2-LR*-Gisignaling dominates, negating β2-LR*-Gseffects in both sarcolemmal and cytosolic compartments (Xiao et al., 1999; also see Fig. 5, A and B). Hence, activation of Gi is involved in the intracellular sorting of β2-LR*-Gs-cAMP signal. However, the same mechanism cannot explain the inability of β2-R* to modulate the L-type Ca2+ channel because there is little β2-R*-Gi coupling (Gurdal et al., 1997), and in the present study, PTX treatment cannot potentiate the basal ICa in TG4 cells (Fig. 6B). Thus, some unidentified mechanisms must be involved in the differential cAMP signaling induced by β2-R* versus β2-LR*. For example, β2-R* and β2-LR* could couple to different isoforms of Gs (Seifert et al., 1998) or adenylyl cyclase (for review see Tang and Hurley, 1998), or to distinctively localized components of the cAMP signaling cascade, such as cAMP (Hohl and Li, 1991) or PKA (Buxton and Brunton, 1983). In addition, localized activation of phosphodiesterase (Jurevicius and Fischmeister, 1996), protein phosphatase (Kuschel et al., 1999a), or specific anchoring proteins of PKA (Gray et al., 1998) may also contribute to subcellular compartmentalization of cAMP or PKA during β2-R* or β2-LR* stimulation. The exact mechanism underlying the inability of β2-R*-cAMP signaling to regulate ICa remains to be elucidated in future studies.
Possible Mechanism for β2-R* to Augment Cardiac Contractility.
Cardiac contractility is an integrated parameter determined by several effectors involved in the E-C coupling cascade. Although ICa is unaffected by β2-R*, the increase in the adenylyl cyclase activity and cAMP production may modulate the E-C coupling cascade by PKA-dependent phosphorylation of target proteins downstream of L-type Ca2+ channels, e.g., the SR Ca2+ release channels, SR membrane protein PLB, and some contractile proteins. Indeed, our preliminary observations have shown that in TG4 ventricular myocytes, the frequency of “Ca2+ sparks” (i.e., the elementary SR Ca2+ release events) and the amplitude of whole cell Ca2+ transients are markedly increased in TG4 cells, and that both are sensitive to ICI. In addition, there is an adaptive down-regulation of PLB expression in TG4 hearts (Rockman et al., 1996) and thereby less basal inhibition of the SR Ca2+ pump in cardiac cells from these transgenic animals. Thus, the enhanced SR Ca2+ recycling may be sufficient to account for the augmentation of baseline contractility in TG4 heart. Regardless of the specific mechanism, the suppression of the enhanced basal contractility by Rp-CPT-cAMPS (Fig. 3) indicates that the β2-R*-elicited contractile effect is largely cAMP/PKA dependent.
Loss of β1AR Function Associated with β2AR Overexpression.
Although both β1AR and β2AR coexist in mouse ventricular myocyte, the function of β1AR is undetectable in β2AR overexpression transgenic (TG4) murine heart, as shown by the absence of ICa (Fig. 7, A and B) or contractile response (Fig. 7C; also see Bond et al., 1995; Du et al., 1996) to β1AR stimulation by either NE plus prazosin or isoproterenol plus the β2AR blocker, ICI. In contrast, in WT mouse ventricular myocyte, β1AR stimulation produced a dose-dependent increase in contraction amplitude (Korzick et al., 1997) and ICa (Fig. 7, A and B). In TG4 myocytes, PTX treatment only rescues the contractile and ICaresponses to β2AR agonists, but not to β1AR agonists (Fig. 7; also see Xiao et al., 1999). Although the exact mechanism for the loss of β1AR function in TG4 heart is unknown, this phenotype seems to be linked to the overexpression of β2AR, because the β1AR function also disappeared in rat C6 glioma cells overexpressed β2AR (Zhong et al., 1996). These results indicate a complex interaction between βAR subtypes (Zhong et al., 1996).
β2-AR Stimulation in TG4 Hearts at Different Ages.
Recent studies have shown that ICadensity is increased in embryonic/neonatal TG4 myocytes (An et al., 1999), but decreased in 3- to 8-month old TG4 mouse heart cells (Heubach et al., 1999) as compared with age-matched controls. In the present study, we found no evidence for any difference in ICa characteristics between transgenic and WT cells from young adult animals (2–3 months old). This apparent discrepancy may reflect an age-related change in βAR signaling cascade. In nontransgenic rat, there are striking developmental changes with respect to β2AR agonist sensitivity and functions (Kuznetsov et al., 1995, Xiao et al., 1998), perhaps due to a developmental changes in β2AR-Gi coupling. In this scenario, it is not surprising that spontaneous β2AR activation may exhibit differential functions at different stages of development. Alternatively, it is possible that some compensatory changes (e.g., expression of L-type Ca2+ channel) may occur progressively as a result of the receptor overexpression, rendering divergent and even conflicting phenotypes at different ages. Nevertheless, as discussed above, a compensatory change in Ca2+ channel sensitivity to cAMP-PKA signaling cannot account for the inability of β2-R*s to regulate ICa in the young mouse heart.
Additionally, it is noteworthy that there is a common thread among these reports: the effect of β2-R*s in TG4 cardiac myocytes is highly compartmentalized and target protein-specific. In embryonic/neonatal TG4 cells, β2-R*s augment ICa but not cAMP-sensitive potassium currents (IK) (An et al., 1999). In young adult TG4 cells (2–3 months), baseline contraction is increased but ICa is unchanged (this study); whereas in older (3–8 months) TG4 cells, ICa is down-regulated without changing baseline contractility (Heubach et al., 1999). The results in adult TG4 cells also suggest a general pattern for dissociation between alterations in baseline contractility and ICa in this transgenic model.
In summary, we have provided several lines of evidence that in TG4 cardiac myocytes, ligand-independent, spontaneously activated β2ARs, in contrast to the ligand-activated β2ARs, do not regulate the L-type Ca2+ channel, despite the fact that both β2-R* and β2-LR* can increase cAMP and contractility. However, salient properties of L-type channels in TG4 cells are unaltered and ICa response to β2-LR* (in PTX-treated cells) or forskolin remains intact. These results suggest that β2-R* may differ from β2-LR*, and thereby the two-state receptor model apparently needs to be expanded to accommodate additional active receptor species. These novel findings of the present study also raise many important unsolved questions. 1) What is the mechanism controlling the sorting of intracellular signals en route from the same receptor at different active states? 2) What are the effectors via which β2-R* produce a positive inotropic effect? 3) Why are L-type Ca2+ channels inaccessible to β2-R*-stimulated cAMP yet receptive to β2-LR*- and adenylyl cyclase-elicited cAMP signaling? 4) What is the mechanism underlying the development- and age-associated differences in β2-AR signaling? Future studies are required to further understand these detailed aspects of β2-R* and β2-LR* signaling.
Acknowledgments
We thank Drs. Walter J. Koch and Robert J. Lefkowitz for kindly providing the β2AR overexpression transgenic (TG4) mice, and Dr. Harold A. Spurgeon and Bruce Ziman for their excellent technical support.
Footnotes
- Received March 1, 1999.
- Accepted May 14, 1999.
-
Send reprint requests to: Rui-Ping Xiao, M.D., Ph.D., Laboratory of Cardiovascular Science, Gerontology Research Center, National Institute on Aging, National Institutes of Health, 5600 Nathan Shock Dr., Baltimore, MD 21224. E-mail:xiaor{at}grc.nia.nih.gov
Abbreviations
- βAR, β-adrenergic receptor
- β-R*, spontaneously activated βAR
- β-LR*
- ligand activated βAR
- CGP
- CGP20712A
- E-C
- excitation-contraction
- Gi and Gs, inhibitory and stimulatory G protein(s)
- respectively
- ICa
- L-type Ca2+ current
- ICI
- ICI118,551
- NE
- norepinephrine
- PKA
- cAMP-dependent protein kinase A
- PLB
- phospholamban
- PTX
- pertussis toxin
- R and R*
- inactive and active receptor conformational states, respectively
- Rp-CPT-cAMPS
- Rp diastereomers of 8-(4-chlorophenylthio)-cAMP
- SR
- sarcoplasmic reticulum
- TG4 mice
- transgenic mice overexpressing human β2AR
- WT mice
- wild-type mice.
- The American Society for Pharmacology and Experimental Therapeutics