Alpha₁-adrenenoceptor stimulation inhibits cardiac excitation–contraction coupling through tyrosine phosphorylation of beta₁-adrenoceptor

Abstract

Adrenoceptor stimulation is a key determinant of cardiac excitation–contraction coupling mainly through the activation of serine/threonine kinases. However, little is known about the role of protein tyrosine kinases (PTKs) activated by adrenergic signaling on cardiac excitation–contraction coupling. A cytoplasmic tyrosine residue in β₁-adrenoceptor is estimated to regulate G_s-protein binding affinity from crystal structure studies, but the signaling pathway leading to the phosphorylation of these residues is unknown. Here we show α₁-adrenergic signaling inhibits b-adrenergically activated Ca²⁺ current, Ca²⁺ transients and contractile force through phosphorylation of tyrosine residues in β₁-adrenoceptor by PTK. Our results indicate that inhibition of b-adrenoceptor-mediated Ca²⁺ elevation by α₁-adrenocep tor-PTK signaling serves as an important regulatory feedback mechanism when the catecholamine level increases to protect cardiomyocytes from cytosolic Ca²⁺ overload.

1. Introduction

Catecholamines released from the sympathetic nervous system act on adrenoceptors (ARs) in heart to regulate heart inotropy and chronotropy [1–6]. In mammalian heart β₁-, β₂-and α₁-AR are the main AR subtypes expressed, with β₁-AR being the most abundant [4,5]. Acute activation of β-AR signaling controls cardiac excitation–contraction coupling through phosphorylation of various Ca²⁺-handing proteins mainly by serine/threonine kinases such as protein kinase A (PKA), PKB, PKC, PKD and Ca²⁺/calmodulin-dependent PK II [2,3,6], while chronic β-AR stimulation in response to stress strongly contributes to changes in contractility and to pathological remodeling, leading to hypertrophy and/or heart failure [7,8]. In cardiomyocytes, another AR subtype, α₁-AR is also expressed. Activation of α₁-AR also shows positive inotropic effects and contributes to pathological remodeling [2,5]. Although stimulation of the heart by either α₁ or β-ARs is associated with increases of ventricular contractile force, the molecular mechanisms underlying the mechanical response to each receptor stimulation are quite different [1,2,4]; β-AR stimulation induces the major contractile effect in the heart [1,3,9], while α₁-AR alone shows a moderate and delayed inotropic response [1,4,10,11].Although modulation of excitation–contraction coupling and contribution to cardiac pathogenesis through both β-AR and α₁-AR has been extensively studied, little is known about the interactions (or cross talk) between these AR-subtypes’ downstream signaling. Early studies from the 1960s to 1990s reported that an α₁-AR stimulation can inhibit the positive inotropic and chronotropic effects of β-AR stimulation [12–14]. However, the detailed mechanism underlying cardiac α₁-AR inhibition of β-AR signaling is still not well understood.

Here we showed that α₁-AR stimulation inhibits β-adrenergically activated Ca²⁺ transients (CaT) and contractile force through the inhibition of the Ca²⁺ current through L-type channels (I_Ca) by protein tyrosine kinase (PTK). We found that I_Ca inhibition is mediated through a decrease in cyclic adenosine monophosphate (cAMP) production concomitant with the phosphorylation of cytoplasmic tyrosine residues in β₁-AR by α₁-AR-PTK signaling. Interestingly, only a single cytoplasmic tyrosine residue exists in the cytoplasmic structures of β₁-AR, which is located in the 2nd cytoplasmic loop (C-loop) [15]. Because this single tyrosine residue in the 2nd C-loop is predicted to regulate G_s-protein binding affinity from observation in crystal structures [15], these results indicate that α₁-AR-mediated tyrosine phosphorylation of β₁-AR may reduce the binding affinity with G_s protein, followed by the inactivation of downstream signaling and the inhibition of cardiac excitation–contraction coupling. Our findings suggest that α₁-AR-PTK signaling serves as an important regulatory feedback mechanism when catecholamine levels increase to protect cardiomyocytes from cytosolic Ca²⁺ overload.

2. Material and methods

An expanded Section 2 is in the online data supplement. The papillary muscles obtained from adult rat heart ventricles [9,11,16] were used for measuring CaT and tension with the injection of Ca²⁺-sensitive photoprotein, aequorin [9,11,16]. Single adult rat ventricular myocytes [17–19] were used for measuring I_Ca by perforated patch-clamp [18,19] and measuring cAMP levels by competitive enzyme immunoassay [19]. HEK293T cells were used for detecting tyrosine phosphorylation of β₁-AR by co-transfection of HA-tagged human α_1A-AR and Flag-tagged β₁-AR. All results are shown as mean ± standard error (SE) or otherwise indicated. For multiple comparisons, one-way ANOVA or one-way repeated measure ANOVA followed by Bonferroni post hoc test was used with the significance level set at P < 0.05.

3. Results

3.1. α₁-AR stimulation inhibits CaT and contraction during β-AR stimulation

At first, we investigated whether α₁-AR stimulation changes Ca²⁺ handing during β-AR stimulation in tissue/cell levels. We examined intracellular Ca²⁺ kinetics and contractile properties simultaneously using aequorin-injected rat papillary muscle as we previously described [1,9,11,20] (Fig. 1). At first, papillary muscle was treated with b-AR agonist, 10 lM isoproterenol (Iso) for 5 min. Iso dramatically increased both CaT and tension [1,9] (Fig. 1A). Then, 100 lM phenylephrine (Phe) was applied in the continued presence of Iso. Indeed, Phe significantly inhibited Isostimulated peak tension and also decreased peak CaT (Fig. 1A) and this effect was abolished by removing Phe from the extracellular solution (Fig. 1). Thus, we confirmed that α₁-AR stimulation decreases b-adrenergically activated contractile force with decreasing peak CaT.

Fig. 1.

α₁-AR stimulation inhibits β-adrenergically activated CaT and contractile force in rat papillary muscle. (A) Representative traces of CaT (upper traces) and isometric tension (lower traces) in rat papillary muscle. (B) Summarized data showing the effect of Phe in the presence of Iso on CaT (upper traces) and tension (lower traces).

3.2. α₁-AR stimulation inhibits I_Ca during β-AR stimulation

Next, we hypothesized that the decrease in CaT by α₁-AR stimulation is due to the inhibition of I_Ca. Therefore, we next examined the effect of a₁-AR stimulation on I_Ca when β-AR is stimulated using perforated patch clamp. At first, isolated adult rat ventricular myocytes were treated with 100 nM Iso for 10 min and then, 100 lM Phe was applied in the continued presence of Iso (Fig. 2A–C). Indeed, Phe significantly inhibited the peak of I_Ca in the presence of Iso and this effect was abolished by removing Phe from the extracellular solution (Fig. 2A–C). This effect was not observed in the presence of an α₁-AR antagonist, 1 lM prazosin, confirming that this effect is mediated through α₁-AR stimulation (Fig. 2F and Supplementary Fig. 1A and D). In the heart β₁and β₂-ARs are expressed with β₁-AR being the most abundant [4]. To address the question of which b-AR subtype is responsible for the increase of I_Ca by Iso, we investigated the effect of Iso in the presence of a β₂-AR selective antagonist, 100 nM ICI 118,551, for evaluating the effect of selective β₁-AR stimulation. Iso still increased I_Ca about 1.84-fold even in the presence of the β₂-AR selective antagonist (Supplementary Fig. 2). Next we tested the effect of β₂-AR selective stimulation on I_Ca by the combination of a β₂-AR relatively-selective agonist, 10 lM zinterol with β₁-AR selective antagonist, 300 nM CGP 20712A. Zinterol did not increase I^ca amplitude in the presence of CGP 20712A, confirming that β₂-AR G_s signaling is not functionally coupled to I_Ca in our preparations as previously reported (n = 4) (data not shown) [21]. Thus, β₁-AR is the dominant subtype responsible for the increase of I_Ca by Iso in our preparation. We next examined the effect of α₁-AR stimulation on I_Ca when β₁-AR is selectively stimulated (Supplementary Fig. 2). Phe significantly inhibited I_Ca in the presence of Iso and a β₂-AR selective antagonist, ICI 118,551, suggesting that α₁-AR signaling can inhibit β₁-adrenergically activated I_Ca.

Fig. 2.

α₁-AR signaling inhibits β-adrenergically activated I_Ca through PTK in rat isolated ventricular myocytes. (A) Typical time course of I_Ca during 100 lM Phe stimulation in the presence of 100 nM Iso. Holding potential was set at 40 mV and I_Ca was evoked by 0 mV depolarization pulse at 1 Hz. (B) Summarized data of the effect of Phe in the presence of Iso on voltage-current (I–V) relationships. (C) Summarized data of the effect of Phe on Iso-activated I_Ca at 0 mV depolarization pulse. *P < 0.05. (D) Typical time course of I_Ca during 100 lM Phe stimulation in the presence of 100 nM Iso and PTK inhibitor 5 lM lavendustin A (LavA). (E) Summary data of the effect of Phe in the presence of Iso and LavA on I–V relationships. (F) Summarized data of the inhibitory effect of Phe on Iso- or 10 lM forskolin (Forsk)-activated I_Ca in the absence or presence of various inhibitors. *P < 0.05.

3.3. Phenylephrine-induced I_Ca inhibition during β-AR stimulation is mediated through α_1A-AR subtype

In rat cardiomyocytes, two subtypes of α₁-AR, α_1A-and α_1B-AR are functionally expressed and each subtype has a distinct downstream signaling pathway [10,19]. Next we examined which subtype of a₁-AR is responsible for this mechanism. The α_1A-AR selective antagonist, 2 lM WB4101 [10,19], blocked this inhibitory effect by Phe (Supplementary Fig. 1C and D), but α_1B-AR selective antagonist, 100 nM L-765,314 [10,19] did not (Supplementary Fig. 1B and D). We also tested the effect of selective α_1A-AR agonist, 1 lM A61603 [10,19]. A61603 significantly inhibited I_Ca in the presence of Iso as in the case of Phe (Supplementary Fig. 1D). These results indicated that Phe-induced I_Ca inhibition during b-AR stimulation is mediated through α_1A-AR subtype, but not α_1B-AR. In adult rat cardiomyocytes, 5 subtypes of G proteins (G_S, G_q, G_i-2, G_i-3, G_o) are expressed [10,19,22]. We reported that α_1A-AR is mainly coupled to G_qα-phospholipase C-PKC signaling and regulates I_Ca in the absence of β-AR stimulation [19] (Supplementary Fig. 3). However, inhibition of phospholipase C by 1 lM U73122 [19] did not block the Phe-induced inhibition of I_Ca during β-AR stimulation (Fig. 2F), indicating that other G-protein signaling (not G_qα) is involved in this mechanism (Supplementary Fig. 4A).

3.4. Inhibitory effect of α₁-AR stimulation is acting upstream of adenylate cyclase activation

α₁-ARs are also known to couple with pertussis toxin (PTX)-sensitive G proteins (G_i/o), which inhibits cAMP production [23,24] (Supplementary Fig. 3). It is reported that α₁-AR activation antagonizes β-adrenergically stimulated cAMP levels through an increase in cAMP breakdown [25] or inhibition of cAMP synthesis in cardiac muscle [26]. Indeed, treatment with 100 nM Iso for 15 min strongly increased cAMP levels about 3-fold in isolated adult rat ventricular myocytes as we previously reported [19] and 100 lM Phe inhibited the β-adrenergically stimulated cAMP level (32.4 ± 2.58%, n = 3) as observed by a competitive enzyme immunoassay (see online Material and Methods). Therefore, we next hypothesized that a decrease in I_Ca might be due to the inhibition of β-adrenergically stimulated cAMP levels through the G_i/o pathway. However, we found that inhibition of G_i/o by PTX [19] did not block the Phe-induced inhibition of I_Ca during b-AR stimulation (Fig. 2F), indicating that (1) other signaling (not G_qα or G_i/o) may be involved and (2) this signaling might exert its effect through inhibiting upstream of cAMP production (Supplementary Fig. 4A). Therefore, we directly stimulated adenylate cyclase (AC) with 10 lM forskolin and observed the effect of Phe on I_Ca (Fig. 2F and Supplementary Fig. 4A–C). Phe was unable to inhibit the I_Ca activated by forskolin, suggesting that α₁-AR stimulation exerts its effect by inhibiting upstream of cAMP production through a non-G_qa and non-G_i/o pathway.

3.5. PTK inhibition antagonizes the inhibitory effect of α₁-AR stimulation

Because Harvey’s group previously reported that α₁-AR stimulation did not inhibit the I_Ca and I_Cl in guinea-pig cardiomyocytes activated by histamine (which is another receptor coupled with G_s like b-ARs) [27,28], we hypothesized that α₁-AR signaling may act directly at the level of the β₁-AR rather than G_s protein (Supplementary Fig. 5). In addition, our data clearly showed that α₁-AR signaling downstream of G_qα and G_i/o are not involved in the inhibitory effect on β-adrenergically stimulated I_Ca (Supplementary Fig. 4A). Therefore, another signaling pathway associated with α₁-AR might be involved such as PTK activity, which is activated through βƔ subunits of G_q [29] (Supplementary Fig. 5). Therefore, we next tested whether PTK activity is involved in the signaling pathway for the I_Ca inhibition. In the presence of a general PTK inhibitor, 5 lM lavendustin A, Phe-induced inhibition of I_Ca during Iso stimulation was completely abolished (Fig. 2D–F). We also tested another general PTK inhibitor, 50 lM genistein, and found that this also abolished the inhibitory effect of α₁-AR stimulation on β-adrenergically stimulated I_Ca (data not shown). We concluded that α_1A-AR signaling inhibits β₁-adrenergically activated I_Ca through PTK activity, which possibly inhibits the β₁-AR signaling at the level of receptor (Supplementary Fig. 5).

3.6. α₁-AR stimulation phosphorylates tyrosine residues in β₁-AR

We next addressed whether β₁-AR can be directly phosphorylated by PTKs upon α₁-AR stimulation because a cytoplasmic tyrosine residue in β₁-AR is predicted to regulate G_s protein binding affinity from the observation in crystal structures [15]. We overexpressed human HA-tagged α_1A-AR and Flag-tagged β₁-AR into HEK293T cells and stimulated the cells by 100 nM Iso for 5 min and then added 100 lM Phe in the continued presence of Iso for an additional 15 min. We found that β₁-AR showed tyrosine phosphorylation by α₁-AR stimulation using general anti-phosphotyrosine antibody [30] (Fig. 3A). Collectively, these data indicate that α₁-AR stimulation signals for phosphorylation of tyrosine residues in β₁-AR, which may reduce the binding affinity of G_s to β₁AR and decrease the cAMP levels.

Fig. 3.

α₁-AR stimulation phosphorylates tyrosine residues in β₁-AR. (A) HEK293T cells transfected with human HA-tagged α_1A-AR and Flag-tagged β₁-AR were stimulated with or without 100 lM Phe in the presence of 100 nM Iso. Western blot was performed on same membrane with anti-Flag (left) and aniti-phosphotyrosine (right) antibodies. (B) Schematic structure of human β₁-AR showing the location of a cytoplasmic tyrosine residue (Y166). Y377 is located at the boarder of the transmembrane domain and cytoplasmic C-term tail. 389R>G is the location of a polymorphism found in human.

4. Discussion

In this study, we identified a detailed molecular mechanism of α₁-adrenergic modulation of β-adrenergic signaling focusing on excitation–contraction coupling regulation by PTK (Fig. 4). Our data shows that α₁-AR signaling inhibits b-adrenergically activated CaT and contractility (Fig. 1) through inhibition of L-type Ca²⁺ channel activity (I_Ca) (Fig. 2). This effect requires PTK activity activated by α_1A-AR-G_qbc pathway, and is mediated through inhibition at the upstream portion of β₁-AR signaling (Fig. 2). Moreover, we found that PTK activated by α₁-AR signaling phosphorylates the tyrosine residue of β₁-AR, which may desensitize downstream signaling by decreasing the binding affinity of G_s (Fig. 3). Inhibition of b-AR signaling by α₁-AR stimulation may serve as an important regulatory feedback mechanism when catecholamine levels increase under pathophysiological conditions.

Fig. 4.

Working model: α_1A-AR stimulation inhibits cardiac excitation–contraction coupling through tyrosine phosphorylation of β₁-AR. Schematic diagram of the proposed intracellular mechanism underlying the inhibitory effect of α_1A-AR signaling on β₁-AR signaling.

4.1. Crosstalk signaling between α₁and β-adrenergic stimulation in cardiac excitation–contraction coupling

It is well reported from 1960s that α₁-AR agonist can inhibit the positive inotropic and chronotropic effects of β-ARs stimulation [12–14,31,32]. From the 1990s, the α₁-AR inhibitory effect was found not only in β-AR-mediated increase in contractile force [14], but also I_Ca [26,27,33] and I_Cl [28,34,35]. A suggested mechanism for this inhibition was α₁-AR activation antagonizing β-AR stimulated cAMP levels in cardiac muscle [25] possibly through PTX-sensitive G-proteins [23,24], but Hool et al. showed that G_i protein is not involved in this mechanism [34]. Chen et al. reported that PKC is involved in α₁-AR inhibitory effect on β-adrenergically stimulated I_Ca [33], but Oleksa et al. showed that PKC activation does not mimic this effect [28]. Belevych et al. first reported that broad PTK inhibitors could antagonize the α₁-AR inhibitory effect [27]. Overall, the information from prior research related to the mechanism underlying cardiac α₁-AR inhibition of β-AR signaling is widely scattered and lacking continuity to understand the whole picture of excitation–contraction coupling under adrenergic stimulation. In this study, we confirmed that α₁-AR signaling inhibits both CaT and contractility during β-AR stimulation through I_Ca inhibition (Fig. 4). Furthermore, we clearly identified a detailed signaling pathway in which the α_1A-AR-PTK pathway phosphorylates β₁-AR, which inhibits downstream signaling (G_s-AC-cAMP-PKA), followed by a decrease in I_Ca, CaT and contractility (Fig. 4).

4.2. Tyrosine phosphorylation of β₁-AR by α₁-AR stimulation and its possible role in adrenergic signaling

β₁-AR signaling is strongly down-regulated in pathophysiological conditions [5] through enhanced activity of a serine/threonine kinase such as G-protein coupling receptor (GPCR) kinase-2, which phosphorylates agonist-occupied b-ARs, leading to homologous desensitization and impaired signaling through these receptors despite the continued presence of agonist [36]. β₂-AR can be phosphorylated by PTKs on tyrosine residues located in 2nd C-loop and the C-terminal tail leading to receptor internalization in noncardiac cells [37,38]. However, there are no reports on β₁-AR desensitization by tyrosine residue phosphorylation and the upstream signaling responsible for the tyrosine phosphorylation. In human β₁-AR, there are 9 tyrosine residues. Interestingly human β₁-AR has only one tyrosine at the cytoplasmic region based on recently reported crystal structure, which is Y166 (Y149 in turkey) in the 2nd C-loop [15] (Fig. 3B). It is reported that this single tyrosine residue in the 2nd C-loop is likely to have a key role in binding with G_s proteins [15]. It is also shown in another GPCR, muscarinic M5 receptor that 2nd C-loop has a key role in G-protein coupling[39]. Human Y377 is located at the end of the 5th transmembrane domain and just before the C-term tail (the boarder of transmembrane domain and cytoplasmic C-term). This tyrosine residue might also be phosphorylated by cytosolic PTKs (Fig. 3B). Interestingly this Y377 is located in close proximity to a polymorphism (389R>G) (Fig. 3B) shown to have major functional effects in humans, due to a reduction of basal cAMP production [40]. This polymorphic variant of β₁-AR is found in 25–40% of the human population and has been shown to influence the risk of ventricular arrhythmia and survival rate after heart failure in patients [40]. Therefore, these studies supports our idea that α₁-AR-mediated tyrosine phosphorylation of Y166/Y377 of β₁-AR mediates decreased G_s binding affinity, thereby down-regulating β₁-AR-G_sAC-cAMP-PKA signaling. Future studies will address which tyrosine residue (Y166/Y377) is the α₁-AR signaling-specific phosphorylation site and ultimately responsible for the antagonism of ß₁-AR downstream signaling.

4.3. α₁-AR-mediated PTK activations and their role on cardiac excitation–contraction coupling

Serine/threonine protein kinases are activated by either β-or α₁-AR stimulation. Upon activation they phosphorylate various Ca²⁺-handling proteins and regulate cardiac functions including excitation–contraction coupling [2–5]. The effect of serine/threonine protein kinases on excitation–contraction coupling is widely studied, however little is known about the role of PTKs on excitation–contraction coupling. G_qPCRs including α₁-AR are known to activate serine/threonine protein kinases and various non-receptor or receptor-type PTKs [29,41,42]. The PTKs downstream of G_qPCRs regulate transcription factors, leading to cardiac hypertrophy [29,43]. PTKs activated by G_qPCRs are well studied as components for nuclear signal transduction, but little information is available about the effect of PTKs on cardiac Ca²⁺ handling. Only several reports including this current report showed that general PTK inhibitors increase L-type Ca²⁺ channel activity and cardiomyocyte contractility, consequently showing a gain of excitation–contraction coupling [27,44]. Further studies are needed (1) to specify the type of PTK that contributes to excitation–contraction coupling and (2) to determine target proteins of this PTK in cardiomyocytes other than β₁-AR.

In summary, the current results show that α₁-AR stimulation inhibits cardiac excitation–contraction coupling during β₁-AR stimulation. This effect is mediated through tyrosine phosphorylation of β₁-AR by α_1A-AR-PTK signaling, which may reduce the binding affinity with G_s protein, followed by the uncoupling of downstream β₁-AR signaling. Our findings suggest that inhibition of β₁-AR signaling by α_1A-AR-PTK signaling serves as an important regulatory feedback mechanism for protecting the cardiomyocytes from cytosolic Ca²⁺ overload when catecholamine levels increase. Furthermore, inhibition of α_1A-AR-PTK signaling may lead to the development of novel pharmacological managements for the treatment of cardiac dysfunction in heart failure patients, where β₁-AR signaling is down regulated.

Abbreviations:

AR

adrenoceptor

PTK

protein tyrosine kinase

C-loop

cytoplasmic loop

CaT

Ca²⁺ transients

I_Ca

Ca²⁺ Ca current through L-type channels

cAMP

cyclic adenosine monophosphate

Phe

phenylephrine

Iso

Isoproterenol

AC

adenylate cyclase

GPCR

G-protein coupled receptor

PTX

pertussis toxin