Review

β-Adrenergic Signaling in the Heart: Dual Coupling of the β2-Adrenergic Receptor to Gs and Gi Proteins

See allHide authors and affiliations

Science's STKE  16 Oct 2001:
Vol. 2001, Issue 104, pp. re15
DOI: 10.1126/stke.2001.104.re15

Abstract

β-adrenergic receptor (AR) subtypes are archetypical members of the G protein-coupled receptor (GPCR) superfamily. Whereas both β1AR and β2AR stimulate the classic Gs-adenylyl cyclase-3′,5′-adenosine monophosphate (cAMP)-protein kinase A (PKA) signaling cascade, β2AR couples to both Gs and Gi proteins, activating bifurcated signaling pathways. In the heart, dual coupling of the β2AR to Gs and Gi results in compartmentalization of the Gs-stimulated cAMP signal, thus selectively affecting plasma membrane effectors (such as L-type Ca2+ channels) and bypassing cytoplasmic target proteins (such as phospholamban and myofilament contractile proteins). More important, the β2AR-to-Gi branch delivers a powerful cell survival signal that counters apoptosis induced by the concurrent Gs-mediated signal or by a wide range of assaulting factors. This survival pathway sequentially involves Gi, Gβγ, phosphoinositide 3-kinase, and Akt. Furthermore, cardiac-specific transgenic overexpression of βAR subtypes in mice results in distinctly different phenotypes in terms of the likelihood of cardiac hypertrophy and heart failure. These findings indicate that stimulation of the two βAR subtypes activates overlapping, but different, sets of signal transduction mechanisms, and fulfills distinct or even opposing physiological and pathophysiological roles. Because of these differences, selective activation of cardiac β2AR may provide catecholamine-dependent inotropic support without cardiotoxic consequences, which might have beneficial effects in the failing heart.

Overview

G protein-coupled receptors (GPCRs) constitute the largest class of cell surface signaling molecules in eukaryotes and some prokaryotes. They share a common overall structure feature: seven hydrophobic transmembrane helical domains. In the worm Caenorhabditis elegans, GPCR-encoding genes constitute 5% of the genome with ~1100 members (1), whereas there are more than 700 GPCRs in the human genome (2). By activating their cognate heterotrimeric guanosine triphosphate (GTP)-binding proteins (G proteins), GPCRs transduce stimulatory or inhibitory signals for a wide array of endogenous hormones and neurotransmitters, and ambient physical and chemical stimuli, as well as exogenous therapeutic reagents. Occupation of these receptors with agonists promotes guanosine diphosphate--guanosine triphosphate (GDP-GTP) exchange on the Gα subunit and subsequent dissociation of Gα from Gβγ, leading to activation of Gα and release of free Gβγ heterodimers (3-6). Both Gα and Gβγ then serve as signaling mediators to directly interact with a variety of effector proteins, including enzymes and ionic channels (7, 8).

Intracellular propagation of GPCR signaling is an intricate process orchestrated by a myriad of G proteins. In mammals, there are at least 27 Gα, 5 Gβ, and 13 Gγ subtypes (9). On the basis of the primary sequences of the Gα subunits, G proteins can be divided into four families: Gs, Gi, Gq, and G12 (10). Historically, specificity and selectivity in GPCR signaling was thought to be achieved by coupling of a given GPCR to a single class of G proteins. As the prototypical GPCR, β-adrenergic receptor (AR) was found to interact exclusively with Gs, which in turn activated adenylyl cyclase (AC), catalyzing 3′,5′-adenosine monophosphate (cAMP) formation. Subsequently, activation of cAMP-dependent protein kinase (PKA, also known as A-kinase) would lead to phosphorylation of target proteins. This paradigm, however, has been shifted with the finding that multiple GPCRs can couple to more than one G protein pathway. Whereas the β1AR subtype appears to stimulate solely the Gs pathway, compelling evidence indicates that β2AR conducts a duet of signaling that includes both Gs and Gi. The additional Gi pathway not only reshapes the spatiotemporal pattern of the Gs-AC-cAMP signaling, but also delivers Gs-independent signals. These βAR subtypes are currently believed to fulfill distinct, sometimes even opposite, physiological and pathological roles.

This review will highlight recent advances in our understanding of signal transduction by these βAR subtypes, particularly β2AR, in the heart. It will attempt to unravel the intricacies of how multiple signals are compartmentalized and integrated in space and time to achieve diversity and specificity for GPCR signaling. Furthermore, a conceptual framework for understanding the physiological and pathophysiological relevance of the coupling of GPCRs to multiple G proteins and of the coexistence of receptor subtypes will be provided.

Dichotomous Coupling of Native β2AR to Gs and Gi in the Heart

Cardiac tissue expresses at least two subtypes of βAR: β1AR and β2AR. In the heart, nonselective βAR stimulation activates the Gs-AC-cAMP cascade, leading to PKA-dependent phosphorylation of a set of regulatory proteins involved in cardiac excitation-contraction coupling and energy metabolism, including L-type Ca2+ channels, the sarcoplasmic reticulum (SR) membrane protein phospholamban (PLB), myofilament proteins, and glycogen phosphorylase kinase. A hallmark of β1AR or mixed βAR stimulation is to increase cardiac contractility (positive inotropic effect), accelerate cardiac relaxation (positive lusitropic effect), and increase heart rate (positive chronotropic effect). However, in adult rat ventricular myocytes, although stimulation of both βAR subtypes increases the amplitudes of L-type Ca2+ currents (ICa), intracellular Ca2+ transient, and contraction strength, β2AR stimulation fails to accelerate the decay of the intracellular Ca2+ transient and the contractile relaxation (11, 12). The absence of β2AR-mediated relaxation occurs in many other mammalian species, including cats and sheep (13, 14), but not in humans and dogs (15-18). These observations provide the first clue that there can be substantial differences in intracellular signal transduction pathways initiated by β2AR, compared to those activated by β1AR.

In the search for the answer to the "anomalous" behavior of cardiac β2AR stimulation, studies over the past decade have provided evidence for the coupling of native β2AR to at least two pathways under physiological conditions. Studies using physiological conditions at the single-cell level demonstrated that disrupting Gi signaling by pertussis toxin (PTX)-mediated Gi ribosylation markedly enhances β2AR-induced contractile response in rat and mouse ventricular myocytes (19, 20). These results suggest that PTX-sensitive Gi proteins may partially negate the Gs-mediated contractile response in cardiac myocytes. Moreover, photoaffinity labeling of G proteins with [32P]azidoanilide-GTP in conjunction with immunoprecipitation of endogenous G proteins with antibodies specific for Gαs and Gαi provided direct biochemical evidence that native β2AR interacts with both Gs and Gi (specifically, Gi2 and Gi3) signaling pathways in freshly isolated adult mouse cardiac myocytes (20). The maximal effect of β2AR stimulation on the Gi proteins in the mouse cardiomyocytes is comparable to that induced by carbachol, a muscarinic acetylcholine M2 receptor agonist (20). PTX treatment or application of a β2AR antagonist (ICI 118,551) prevents the β2AR-mediated activation of Gi. Under the same experimental conditions, β1AR stimulation does not increase Gi activity; thus, the coupling to Gi is specific for β2AR (20). Similarly, in human myocardium, cardiac Gi is activated by stimulation of β2AR; this property is not shared by β1AR (21). Thus, whereas β1AR activates only the Gs pathway, β2AR can activate both Gs and Gi signaling pathways.

Several other GPCRs, including histamine, serotonin, and glucagon receptors, stimulate both Gs and Gi proteins in human heart (21). Thus, multiple G protein coupling appears to be rather common, albeit not universal, among GPCRs. These findings raise important questions regarding the consequences of coupling one receptor to multiple G proteins in physiological and pathophysiological contexts.

β2AR-to-Gi Signals Compartmentalize Gs-Mediated cAMP Signaling

A species-dependent diversity has been documented with respect to β2AR-stimulated cAMP accumulation and PKA activation. In the human heart, β2AR stimulation efficiently increases cellular cAMP and PKA-dependent phosphorylation of intracellular regulatory proteins [PLB, troponin-I (TnI), and C protein], similar to β1AR stimulation (15, 16, 22). In freshly isolated canine ventricular myocytes and cultured 18-day embryonic mouse cardiomyocytes, however, β2AR elevates neither total cellular cAMP nor PKA activity, whereas β1AR induces a robust increase in cAMP accumulation under the same experimental conditions (17, 18, 23). Between these extremes, in freshly isolated rat ventricular myocytes, the dose-response of cAMP to β2AR overlaps that to β1AR stimulation (12, 24, 25). Nevertheless, both biochemical evidence and biophysical evidence indicate that β1AR-generated cAMP signaling can broadcast throughout the cell, whereas β2AR-initiated cAMP signaling is confined to subsarcolemmal microdomains (26). Specifically, in adult rat and canine hearts, β1AR stimulation increases phosphorylation of PLB, which accelerates Ca2+ sequestration into SR, resulting in accelerated cardiac relaxation (11, 17-19, 24, 27, 28). β1AR stimulation also promotes phosphorylation of TnI and C protein (18), which reduces myofilament sensitivity to Ca2+. In contrast, β2AR stimulation modulates specifically sarcolemmal L-type Ca2+ channels without affecting the aforementioned intracellular regulatory proteins in these species (11, 17-19, 24, 27, 28). Furthermore, experiments with patch-clamp single-channel recordings showed that in rat cardiomyocytes, β2AR stimulation modulated single L-type Ca2+ channel activity only in a local mode (agonist included within the patch pipette with tip diameter ~1.0 μm) and not in a remote mode (agonist perfused outside the patch), whereas β1AR stimulation acted in either mode (29). These results are in general agreement with the observation that in frog cardiomyocytes, in which the β2AR subtype predominates (30), local βAR stimulation by isoproterenol applied to one end of the cell has little stimulatory effect on L-type Ca2+ channels residing on the other end (31).

These studies initially evoked doubts as to whether the β2AR cardiac response is mediated by a cAMP-dependent signaling pathway. One theory proposed is that the cardiac effects of βAR (subtype not specified) might be, in part, mediated by a direct interaction between Gαs and L-type Ca2+ channels (32, 33). However, in other studies, except one in rat ventricular myocytes (34), specific PKA inhibitors, including a peptide inhibitor (PKI), an inactive cAMP analog (RP-cAMP), and a synthetic compound (H-89), not only blocked the effects of β1AR stimulation, but also completely inhibited the effects of β2AR stimulation (18, 20, 27, 29). [Although H-89 has been widely used as a PKA inhibitor, recent studies indicate that H-89 is also a potent blocker of both βAR subtypes (35)]. The results obtained through PKA inhibition corroborate the notion that the effect of nonselective βAR stimulation by isoproterenol on cardiac ICa is mediated exclusively by a cAMP-dependent mechanism. Specifically, the ICa response to isoproterenol is ablated by PKI (36). Hence, the modulation of ICa by β2AR should require cAMP-dependent PKA activation, but this β2AR-stimulated cAMP-to-PKA signaling appears to be tightly localized to the surface membrane in the vicinity of L-type Ca2+ channels and cannot be transmitted to nonsarcolemmal proteins (Fig. 1).

Fig. 1.

Dual coupling of β2AR to Gs and Gi proteins in cardiac myocytes. The activation of β2AR-coupled Gi proteins functionally localizes the concurrent Gs-mediated cAMP-to-PKA signaling to the subsarcolemmal microdomain. The Gi coupling also delivers cell survival signals through a Gi-Gβγ-PI3K-Akt pathway (PTX, pertussis toxin; βARK-ct, a peptide inhibitor of Gβγ signaling; LY, a PI3K inhibitor; Akt, protein kinase B). The arrow from G to global cAMP (but not to the local cAMP) indicates that the Gi coupling functionally localizes the Gs-stimulated cAMP signaling. The local modulation of the sarcolemmal L-type Ca2+ channel by β2AR constitutes the major mechanism for the receptor-mediated positive contractile response. In contrast, β1AR couples exclusively to Gs, which induces a global cAMP signal. AC, adenylyl cyclase.

Several lines of evidence indicate that activation of the β2AR-to-Gi signaling pathway is essential for the spatial localization and effector selectivity of the Gs-stimulated cAMP-to-PKA signaling. First, disrupting Gi function with PTX permits β2AR to stimulate remote L-type Ca2+ channels (29). Second, PTX treatment leads to a robust β2AR-mediated phospholamban phosphorylation and a positive relaxant effect not normally present in β2AR cardiac signaling (19, 28). Thus, coupling of the cardiac β2AR to multiple G proteins can paradoxically enhance, rather than compromise, the spatial and temporal specificity of the receptor signaling.

A challenging question is how β2AR-to-Gi signaling results in the compartmentalization of β2AR-to-Gs-to-cAMP signaling. Possible mechanisms for limiting the cAMP signaling pathway include physical restriction of cAMP diffusion, a local imbalance between AC and phosphodiesterase activities, restriction of the diffusion of PKA, or regulation of the pathway downstream of PKA activity. The diffusible second messenger cAMP can traverse a micrometer-scale distance on a millisecond time scale; hence, it seems unlikely that limiting cAMP diffusion is the mechanism. There is evidence that compartmentalization is a consequence of regulation of the pathway downstream of PKA. PTX treatment, which abrogates the functional compartmentalization of β2AR-to-cAMP signaling in freshly isolated rat ventricular myocytes, has no significant effect on the β2AR-mediated global cAMP accumulation or PKA activation (27, 28). Other Gi-coupled receptors, such as the muscarinic receptor M2 or adenosine receptor A1, counteract the effect of PKA, in part, through activation of protein phosphatases (37, 38). Emerging evidence suggests that inhibition of protein phosphatases with calyculin A, an inhibitor of phosphatases 1 (PP1) and 2A (PP2A), mimics the effects of PTX treatment and enhances β2AR-mediated positive contractile response (28). Because the effects of PTX treatment and calyculin A are not additive, the mechanism by which β2AR-coupled Gi signaling compartmentalizes the concurrent Gs signaling may be through activation of protein phosphatase(s) (28). Activation of the β2AR-coupled Gi proteins stimulates a phosphoinositide 3-kinase (PI3K)-Akt (also known as protein kinase B) cell survival signaling pathway in rat and mouse cardiac myocytes (39, 40) (see below). A question to be answered is whether PI3K signaling also contributes to the Gi-dependent localization of β2AR-to-cAMP signaling; if so, it will be necessary to determine the relation of the PI3K signaling to the Gi-activated protein phosphatases.

Another candidate mechanism underlying compartmentalization of cAMP signaling is the structural restriction of PKA diffusion by specific A-kinase anchoring proteins (AKAPs) (41, 42). For example, a peptide inhibitor of AKAP can inhibit the modulation of L-type Ca2+ channels by PKA, which suggests that AKAPs are necessary for targeting PKA to this substrate (43). Interestingly, AKAPs not only traffic the bound PKA and other enzymes (such as protein phosphatases) to specific compartments, but also functionally modulate the activity of the bound enzymes. This is clearly demonstrated by the inhibition of PKA and stimulation of PP1 by certain AKAPs (44-46). Increasing evidence indicates that direct interaction of β2AR with some AKAPs (such as gravin and AKAP79/150) is essential for agonist-induced β2AR trafficking and desensitization (47-51). A potentially interesting question to be examined is whether AKAPs participate in the Gi-dependent compartmentalization of the β2AR to Gs-mediated cAMP signaling.

β2AR-to-Gi Coupling Delivers Cell Survival Signals

In addition to the modulation of cardiac excitation-contraction coupling by acute βAR stimulation, as discussed above, both in vivo and in vitro studies have shown that prolonged βAR signaling stimulates cardiac myocyte apoptosis (52-54). Apoptosis has been implicated in cardiac ischemic and reperfusion injury and is involved in the transition from cardiac hypertrophy to decompensated heart failure (55-59). Pharmacological evidence suggests that β1AR and β2AR stimulation may exert different effects on cardiac cell survival (60,61). To avoid complicated interactions between βAR subtypes, we created a genetically "pure" β1AR or β2AR experimental system by individually expressing either βAR subtype in the null background of β1AR and β2AR double-knockout adult mouse cardiac myocytes in culture (40, 62). These studies provided evidence that stimulation of β1AR leads to cardiac apoptosis, whereas stimulation of β2AR activates concurrent proapoptotic and antiapoptotic signals, with the net effect being cell survival (40). The distinct effects of β1AR and β2AR on cardiac cell survival and cell death have been further confirmed using gene-targeted mice lacking either βAR subtype or in cultured wild-type adult mouse ventricular myocytes using βAR subtype-selective agonists and antagonists (63).

These differences between the two βAR subtypes might be simply explained by their differential coupling to cAMP. However, this is unlikely because cardiac-specific overexpression of adenylyl cyclase types V or VI in transgenic mouse models markedly increases cAMP and cardiac contractility without apoptotic effects (64, 65). In transgenic mouse hearts or cultured adult mouse cardiomyocytes, overexpression of human β2AR significantly elevates basal cAMP level but is not associated with myocyte apoptosis (66-68). In addition, it has been suggested that myocyte apoptosis induced by β1AR stimulation is independent of cAMP signaling (69).

Alternatively, the differential regulation of cardiac cell survival and cell death by these βAR subtypes can be explained by the additional coupling of β2AR to PTX-sensitive Gi proteins. This conclusion is supported by several independent lines of evidence. First, β2AR stimulation leads to myocyte apoptosis only under conditions in which Gi is inhibited with PTX (40). Second, β2AR, but not β1AR, activates a Gi-Gβγ-PI3K-Akt signaling pathway. Inhibition of Gi-to-Gβγ signaling with PTX or βARK-ct (a peptide inhibitor of Gβγ signaling), or inhibition of PI3K activity with LY294002, completely abolishes β2AR-stimulated Akt activation; more important, it converts β2AR signaling from survival to apoptotic (40). Therefore, PI3K constitutes an intracellular messenger of the β2AR-to-Gi pathway, which protects myocytes against Gs-mediated apoptosis through activation of the survival factor Akt (Fig. 1). Third, pretreatment of cultured neonatal rat cardiac myocytes with β2AR agonists (zinterol or isoproterolol plus a β1AR antagonist, CGP20712A) protects these myocytes from a range of apoptotic assaults, including hypoxia or reactive oxygen species (ROS), through the Gi-dependent, PI3K-mediated mechanism (39).

In addition to the PI3K survival pathway, it has been suggested that in cultured adult rat cardiac myocytes, both βAR subtypes activate p38 mitogen-activated protein kinase (MAPK) in a Gi-dependent manner, and that the activated p38 MAPK results in an antiapoptotic effect (70). However, this finding contradicts earlier observations from the same laboratory that β1AR and β2AR exhibit opposing effects on cardiac myocyte apoptosis because of the specific Gi coupling to β2AR, but not β1AR (60). In fact, evidence obtained from the mouse β1AR-β2AR knockout system argues against the possibility that p38 MAPK is involved in β2AR-mediated cardiac myocyte survival. This is because both β1AR and β2AR increase p38 MAPK activation through a cAMP-to-PKA signaling pathway, but not by a Gi-dependent mechanism (40, 71), and because pharmacological inhibition of p38 by SB 203580 (10 μM) cannot block the β2AR survival effect. These studies indicate that p38 MAPK activation is not related to the β2AR-stimulated, Gi-mediated antiapoptotic effect in adult mouse cardiac myocytes (40). Furthermore, in vivo activation of p38 MAPKs using transgenic overexpression of activated mutants of upstream kinases MKK3bE and MKK6bE neither induces nor suppresses cardiomyocyte apoptosis or hypertrophy in mice (72). Thus, it appears unlikely that p38 MAPK plays an essential role in β2AR-induced antiapoptotic effect.

Other members of the MAPK family, particularly the extracellular signal-regulated protein kinases (ERK1 and ERK2), can also protect cells from apoptosis (73, 74). Stimulation of β1AR or β2AR is able to activate ERK1 and ERK2 in multiple cell types, including cardiac myocytes (39, 75, 76). Interestingly, the effect of β2AR, but not β1AR, on ERK is markedly attenuated by PTX treatment, which suggests that ERK is a downstream target of β2AR-coupled Gi signaling (39). However, inhibition of ERK activation with the inhibitor PD98059, which inhibits the upstream kinase MEK1, cannot prevent a β2AR-mediated antiapoptotic effect (39).

Molecular and Cellular Mechanisms Underlying β2AR-to-Gi Coupling

The mechanisms underlying the differential coupling of βAR subtypes to G proteins are not well understood. Multiple hierarchical mechanisms may act in concert to render the subtype-specific (AR-to-G protein interaction. At the molecular level, β1AR and β2AR are genetically distinct entities. The human β1AR gene is located at chromosome 10 and encodes a protein of 477 amino acids (77), whereas the β2AR gene is located on chromosome 5 and encodes a protein of 413 amino acids (78). The sequences of β1AR and β2AR share 71% and 54% amino acid identity in the seven transmembrane spanning domains and in overall sequence, respectively (77-79). Studies on chimeric or mutated G protein-coupled receptors (including the major subtypes of α- and β-adrenergic receptors) have shown that the third intracellular loop of these receptors is an important structural determinant for G protein coupling (80-82). The third intracellular loop of β1AR is considerably longer than its β2AR counterpart because of the presence of a proline-rich motif that has been implicated as a negative modulator of βAR-Gs coupling. This may, at least in part, explain the difference in the efficacy of β1AR and β2AR coupling to Gs and AC (83, 84). Our preliminary results suggest that replacement of the third intracellular loop and the COOH-terminal tail of β1AR with those of β2AR allows the chimeric receptor to activate both Gi and Gs signaling pathways (85). Thus, the distinct G protein coupling of β1AR and β2AR could eventually be ascribed to some critical differences in the sequences of the third intracellular loops and the COOH-terminal tails of the receptors. Furthermore, a potential contribution of receptor posttranslational modifications to receptor-G protein selectivity has been demonstrated in HEK 293 cells, in which PKA-mediated phosphorylation of β2AR switches the receptor coupling preference from Gs to Gi (75).

The distinct G protein coupling of these βAR subtypes might, to some extent, be attributable to differential subcellular localization of the receptor subtypes and G proteins. In the absence of agonist stimulation, β1ARs are enriched in noncaveolar cell surface membranes, whereas β2ARs are located predominantly in the caveolar membrane fraction of cardiac myocytes (86). The difference in the subcellular distribution of βAR subtypes suggests that β2AR might physically colocalize with Gi proteins, so that Gi proteins are preferentially accessible to β2ARs. This hypothesis is further supported by the fact that Gαi proteins are most abundant in caveolae, whereas Gαs and Gβγ subunits are distributed in both caveolar and noncaveolar cell surface membranes in cardiac myocytes (86).

β2AR-to-Gi Signaling in Developing Hearts

In contrast to the situation in adult cardiac myocytes, the β2AR-mediated contractile response is insensitive to PTX treatment in neonatal rat cardiac myocytes (87). In those cells, β2AR stimulation, like β1AR, induces phosphorylation of PLB and TnI, and accelerates contractile relaxation (12). The dose-response curve of contraction in response to the β2AR agonist zinterol is shifted ~2 orders of magnitude leftward in neonatal myocytes, as compared to that of adult myocytes. Thus, β2AR may play a more important role in mediating the contractile response to catecholamines in the noninnervated neonatal heart than in the innervated adult heart. This developmental change in cardiac β2AR responsiveness appears not to be caused by a difference in the amount of receptor expression, because there is no postnatal change in β2AR density (12). The contraction dose-response to zinterol in neonatal rat myocytes (12) is similar to that in PTX-treated adult rat myocytes (19). Thus, β2AR coupling to Gi proteins might be acquired or reinforced by the onset of innervation during development or by agonist stimulation.

The lack of PTX sensitivity of β2AR contractile response in neonatal rat cardiac myocytes appears to contradict the fact that simultaneous β2AR stimulation and β1AR blockade results in an antiapoptotic effect through a Gi-dependent survival pathway (39), similar to the case in adult myocytes (40). These studies suggest that in neonatal cardiac myocytes, β2AR-to-Gi coupling is rather effective in regulating certain vital cellular processes such as cell survival, whereas it is relatively weak in terms of inhibiting the Gs-mediated positive inotropic effect and phosphorylation of intracellular target proteins that control contraction. Thus, it is possible that the aforementioned compartmentalization mechanisms may not yet be in place in the developing heart.

Interaction Between β2AR and Other Gi-Coupled Receptors

In cardiac myocytes, β2AR differs from β1AR regarding their interaction with several cardiac Gi-coupled receptors. In neonatal rat ventricular myocytes, the β1AR-mediated cAMP accumulation and its inotropic and lusitropic effects are all prevented by M2-muscarinic acetylcholine receptor stimulation with carbachol. In contrast, the β2AR-induced cAMP accumulation and the inotropic effect persist in the presence of carbachol, although β2AR-stimulated phosphorylation of PLB and TnI and the lusitropic response are abolished by carbachol treatment (87). Interestingly, in the absence of agonist stimulation, M2-muscarinic receptors colocalize with β1ARs, but not β2ARs, in noncaveolar cell surface membranes (88). This may explain, in part, the differential interactions of M2 receptors with β1ARs versus β2ARs.

In adult rat myocardium, there is also a striking difference between these βAR subtypes with respect to their cross-talk with Gi- and Go-coupled δ-opioid receptors. A δ-opioid receptor agonist, leucine enkephalin, markedly inhibits β1AR-mediated positive inotropy (89,90). In contrast, leucine enkephalin has no effect on β2AR-mediated increase in cardiac contractility (89), indicating that δ-opioid receptor signaling selectively interacts with cardiac β1AR, but not β2AR, signaling. The exact mechanism underlying the differential interaction of the βAR subtypes and Gi-coupled receptors, both in neonatal and adult rat cardiomyocytes, merits further study.

β2AR, but Not β1AR, Undergoes Spontaneous Activation

According to the extended ternary complex model (91, 92) and the cubic ternary complex model (93), GPCRs, including βARs, exist in an equilibrium of states, including two functionally and conformationally distinct states: an inactive conformation (R) and an active conformation capable of activating G proteins (R*) (66, 94, 95). In the absence of a receptor ligand, the receptor can undergo a spontaneous transition to the activated state; the equilibrium between R and R* sets the level of basal receptor activation. Thus, the overexpression of a given receptor would be expected to proportionally increase the number of R* state receptors. Indeed, in a transgenic mouse model (TG4), cardiac-specific overexpression of β2AR by a factor of ~200 leads to an agonist-independent enhancement in both the baseline AC activity and myocardial contractility (20, 66, 95). These results from the transgenic animals are corroborated by acute gene manipulation in cultured wild-type or β1AR-β2AR double-knockout adult mouse ventricular myocytes, in which adenovirus-directed overexpression of the human β2AR also results in agonist-independent increases in cellular cAMP production and in contractility (67, 68). These studies suggest that cardiac contractility can be enhanced through genetically manipulating the β2AR system, which might hold therapeutic promise for improving the function of the failing heart.

Studies on constitutively active GPCRs have suggested the concept of inverse agonists, that is, drugs that preferentially bind to R and inhibit basal receptor activity (66, 94, 95). In this regard, ICI 118,551 has been identified as an inverse agonist of the β2AR. Although the two-state ternary complex model for the βAR is sufficient to explain many aspects of β2AR activation, there are several important differences between spontaneously activated β2ARs and agonist-stimulated β2ARs in terms of their effector selectivity. In TG4 ventricular myocytes, β2AR agonists produce a marked increase in ICa, whereas ligand-independent constitutive β2AR activation increases cardiac contractility without affecting ICa (96). Hence, spontaneously activated β2AR and agonist-activated β2AR may represent functionally distinct conformational states of the receptor. This is in agreement with recent reports that β2ARs exhibit multiple active states (97, 98).

The property of spontaneous activation is not shared by the β1AR, the predominant (75 to 85%) cardiac βAR subtype. In the mouse β1AR-β2AR null background, overexpression of β1AR to similar levels, or even greater levels, relative to overexpression of β2AR has virtually no effect on cAMP accumulation, contraction amplitude, or contractile kinetics (67). These observations are consistent with the results from transgenic mice overexpressing β1AR (by a factor of 5 to 15 relative to the wild type) (99). Apparently, β1AR, unlike β2AR, does not readily undergo spontaneous activation. Similarly, dopamine receptor subtypes 1A and 1B exhibit strikingly different constitutive activities (100). Thus, not all GPCRs appear to undergo spontaneous activation.

Pathophysiological Relevance of β1AR Versus β2AR Signaling

β1AR and β2AR manifest strikingly different or opposing effects on gene expression, cell growth, and cell death. Specifically, stimulation of β1ARs can produce hypertrophy in cultured neonatal rat cardiac myocytes through activation of a PI3K-Akt-glycogen synthase kinase-3β (GSK-3β) -GATA4 (a member of zinc finger transcription factor family) signaling pathway (101,102). However, this appears to be independent of PTX-sensitive Gi signaling (103) or ERK activation (101). In addition, β1AR activation can exhibit robust apoptotic effects in vivo and in cultured adult myocytes (40, 60, 61, 104). In sharp contrast, β2AR stimulation does not cause cardiomyocyte hypertrophy or apoptosis. Instead, β2AR activation protects myocytes against apoptosis induced by a wide array of assaulting factors, including enhanced β1AR signaling, hypoxia, and ROS (39, 40, 60). Furthermore, chronic stimulation of each βAR subtype in the heart elicits distinctly different phenotypes and results in differences in prognosis in terms of cardiac hypertrophy and heart failure in transgenic mouse models. Overexpression of cardiac β1AR by a factor of 5 to 40 leads to cardiac hypertrophy, myocyte apoptosis, and fibrosis within a few weeks after birth, and heart failure within several months (99, 104). Overexpression of cardiac β2AR by a factor of 100 to 200 does not produce hypertrophy or heart failure, at least up to the age of 1 year (66, 105, 106). However, higher levels of expression of β2AR (such as 350 to 1000 times the normal levels) result in pathological phenotypes (105, 106), perhaps caused by a mechanical and metabolic overload due to spontaneous β2AR activation. The opposing effects of βAR subtypes on cardiac myocyte growth and cell death may explain, at least in part, the inverse relationship between the plasma concentration of norepinephrine (with higher affinity for β1AR than for β2AR) and survival in patients with chronic heart failure (107) and the salutary effects of βAR blockade on morbidity and mortality in heart failure patients (108). These insights into the cellular responses to βAR subtype stimulation also imply that the selective down-regulation of β1ARs in the failing heart (109-112) may represent a protective mechanism to slow the progression of cardiomyopathy and myocyte apoptosis. This idea is further supported by the fact that the second and third generations of relatively β1AR-selective blockers used clinically (such as metoprolol, bisoprolol, and carvedilol) effectively reduce mortality and morbidity of heart failure patients (113), whereas for the first generation of nonselective βAR antagonists (such as propranolol), the drug intolerability rate is prohibitively high because of myocardial depression and worsening of cardiac contractile dysfunction (114).

Although activation of β2AR-coupled Gi protects cardiac myocytes against apoptosis, an imbalance of β2AR-initiated Gs and Gi signaling pathways may have pathological consequences. Chronic heart failure in human and animal models is characterized by a diminished contractile response to βAR stimulation (109-112, 115, 116) and is accompanied by an increase in the amount or activity of Gi proteins (112, 115-118) and a selective down-regulation of β1AR, leading to a higher β21 ratio (109-112). In light of the Gs and Gi dichotomy, the up-regulation of Gi may participate in the reduced βAR inotropic effect in the decompensated failing heart. This idea is supported by the fact that PTX treatment restores the diminished βAR inotropic response in a rat myocardial infarction heart failure model (119) and in myocytes from failing human hearts (120). On the basis of these findings, it is speculated that the selective down-regulation of β1AR and the up-regulation of β2AR to Gi signaling in the functionally compensated hypertrophic heart or in the early stages of heart failure may represent a cardiac protective mechanism. This change in the balance of β1AR and β2AR signaling may protect against myocyte apoptosis and consequently slow the progression of cardiomyopathy and contractile dysfunction. However, exaggerated β2AR-to-Gi signaling may blunt Gs-mediated contractile support, contributing to the phenotype of decompensated heart failure.

Therapeutic Implications of Cardiac β2AR Signaling

Whether enhancing βAR signaling is beneficial or deleterious for the failing heart has been a matter of much controversy. The prevalent view is that chronically increasing nonselective βAR stimulation is toxic to the heart. However, the discovery of (i) the new paradigm of β2AR signaling (dual G protein coupling), (ii) the opposing effects of stimulation of these βAR subtypes on cardiomyocyte apoptosis, and (iii) the distinct phenotypes of cardiac-specific overexpression of β1AR versus β2AR underscore the necessity and importance of distinguishing β2AR signaling from that of β1AR in terms of their cardiac functional roles and therapeutic implications.

Selective enhancement of β2AR signaling may provide a therapeutic strategy for the prevention and treatment of chronic heart failure because of its evident antiapoptotic and positive inotropic effects. Indeed, crossing transgenic mice overexpressing moderate amounts of cardiac β2AR with transgenic mice overexpressing Gαq not only improves cardiac performance, but also reverses hypertrophy in the Gαq overexpression heart failure model (105). Because extremely high levels of β2AR overexpression fail to rescue the genetic mouse heart failure model and can be detrimental (105, 106), caution must be exercised when designing therapies to enhance β2AR signaling so that the beneficial levels of activity are not exceeded. The beneficial effect of β2AR stimulation in the context of heart failure is clearly supported by the analysis of β2AR polymorphisms in chronic heart failure patients. The prognosis of heart failure patients with Ile164 polymorphism (a Thr-to-Ile switch at amino acid 164 with reduced β2AR signaling efficacy) is much worse than the prognosis of patients without the β2AR variant (121). Thus, moderate selective activation of the β2AR subtype may have beneficial effects in the failing heart. Given that epinephrine is a potent β2AR agonist, it would be interesting and informative to determine whether the beneficial effects of exercise might be, in part, attributable to increased cardiac β2AR stimulation by epinephrine.

Other βAR Subtypes in the Heart

The third class of βARs, β3AR, was previously named an "atypical βAR" and was considered genetically and pharmacologically different from either β1AR or β2AR (122, 123). Recent studies provide strong evidence that β3ARs, important regulators of the physiologic properties of adipose tissue and the gastrointestinal tract (and thus the target for antiobesity and antidiabetic drugs), are also present in human cardiomyocytes. In contrast to β1AR and β2AR, they have been implicated as inhibitors of contractile function (124), apparently through a PTX-sensitive G protein-dependent activation of a nitric oxide synthase pathway (125). β3AR also plays an important role in regulating smooth muscle relaxation, which could reflexively influence cardiac contractility. It is noteworthy that β3AR function is up-regulated in the failing heart (126), suggesting that enhanced β3AR signaling may contribute to the phenotype of chronic heart failure. Stimulation of β3AR activates both Gs and Gi signaling pathways in cultured neonatal cardiomyocytes from β1AR-β2AR double-knockout mice. In the absence of PTX, β3AR stimulation has a small and relatively brief inhibitory effect on the spontaneous cell contraction rate, whereas inhibition of Gi with PTX unmasks a positive chronotropic effect (127). In addition, a fourth βAR subtype has been reported to mediate positive chronotropic and inotropic effects in the human heart (128). This "receptor" is now described as a low-affinity state of the β1AR (129, 130), although its genetic identity and pharmacological properties await confirmation.

Beyond G Protein Doctrine: G Protein-Independent βAR Signaling

Possible G protein-independent mechanisms underlying βAR-mediated cellular responses have also been demonstrated. For instance, physical binding of Na+/H+ exchange regulatory factor (NHERF), an inhibitor of Na+/H+ exchanger type 3 (NHE3), to a PDZ domain at the β2AR COOH-terminus relieves the NHERF inhibitory effect on NHE3 (131). The relevance of this phenomenon to β2AR signaling in cardiomyocytes, however, has not yet been explored. Another observation in HEK 293 cells demonstrates that the COOH-terminal SH3 domain of the endophilin SH3p4 specifically binds to the proline-rich motif of the β1AR third intracellular loop (132). This protein-protein interaction is implicated in promoting agonist-induced internalization and in decreasing the Gs coupling efficacy of β1ARs (132). Thus, βAR signaling is highly diversified not only through coupling to multiple G proteins, but also through G protein-independent protein-protein interactions between βARs and various effector proteins.

Concluding Remarks

The discovery of the dichotomous coupling of β2AR to Gs and Gi has challenged the linear one receptor-one G protein paradigm of GPCR signaling in physiological systems. The additional Gi coupling of β2AR creates a functional barrier that localizes the concurrent Gs-mediated cAMP signaling, thus enhancing receptor signaling specificity and effector selectivity. The Gi branch also delivers a Gs-independent cardioprotective signal through the Gi-Gβγ-PI3K-Akt pathway, which not only counteracts the Gs-mediated apoptotic effect but also protects cells from a variety of apoptosis-triggering assaults. Further, the differential G protein coupling, to a large extent, accounts for the distinctly different physiological and pathological roles in the heart for β2AR versus those of β1AR. The delicate balance of Gs and Gi signaling in space and time might be crucial to normal cellular functions, whereas an imbalance may have important pathophysiological relevance and clinical implications. Thus, selectively targeting βAR signaling pathways might afford novel therapeutic strategies for improving the function of the failing heart. Furthermore, these advances in understanding the signaling pathways begin to unravel the logic of multiple G protein coupling of GPCRs and the coexistence of GPCR subtypes in a single cell.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.
  99. 99.
  100. 100.
  101. 101.
  102. 102.
  103. 103.
  104. 104.
  105. 105.
  106. 106.
  107. 107.
  108. 108.
  109. 109.
  110. 110.
  111. 111.
  112. 112.
  113. 113.
  114. 114.
  115. 115.
  116. 116.
  117. 117.
  118. 118.
  119. 119.
  120. 120.
  121. 121.
  122. 122.
  123. 123.
  124. 124.
  125. 125.
  126. 126.
  127. 127.
  128. 128.
  129. 129.
  130. 130.
  131. 131.
  132. 132.
  133. 133.
View Abstract

Navigate This Article