RGS2 Is a Primary Terminator of β₂-Adrenergic Receptor-Mediated G_i Signaling

Abstract

Two major β-adrenergic receptor (βAR) subtypes, β₁AR and β₂AR, are expressed in mammalian heart with β₁AR coupling to G_s and β₂AR dually coupling to G_s and G_i proteins. In many types of chronic heart failure, myocardial contractile response to both β₁AR and β₂AR stimulation is severely impaired. The dysfunction of βAR signaling in failing hearts is largely attributable to an increase in G_i signaling, because disruption of the G_i signaling restores myocardial contractile response to β₁AR as well as β₂AR stimulation. However, the mechanism terminating the β₂AR-G_i signaling remains elusive, while it has been shown activation of the G_i signaling is dependent on agonist stimulation and subsequent PKA-mediated phosphorylation of the receptor. Here we demonstrate that regulator of G protein signaling 2 (RGS2) is a primary terminator of the β₂AR-G_i signaling. Specifically, prolonged absence of agonist stimulation for 24h impairs the β₂AR-G_i signaling, resulting in enhanced β₂AR- but not β₁AR-mediated contractile response in cultured adult mouse cardiomyocytes. Increased β₂AR contractile response is accompanied by a selective upregulation of RGS2 in the absence of alterations in other major cardiac RGS proteins (RGS3-5) or G_s, G_i or βAR subtypes. Administration of a βAR agonist, isoproterenol (ISO, 1.0 nM), prevents RGS2 upregulation and restores the β₂AR-G_i signaling in cultured cells. Furthermore, RGS2 ablation, similar to βAR agonist stimulation, sustains the β₂AR-G_i signaling in cultured cells, whereas adenoviral overexpression of RGS2 suppresses agonist-activated β₂AR-G_i signaling in cardiomyocytes and HEK293 cells. These findings not only define RGS2 as a novel negative regulator of the β₂AR-G_i signaling, but also provide a potential novel target for the treatment of chronic heart failure.

1. Introduction

Stimulation of β-adrenergic receptor (βAR), a prototypical member of G protein-coupled receptors (GPCR) superfamily, is broadly involved in the regulation of energy metabolism, growth control, and cardiac function [1]. In mammalian myocardium, the closely related βAR subtypes, β₁AR and β₂AR, activate subtype-specific G protein signaling pathways and elicit distinct actions[2]. While β₁AR couples to the classic G_s-adenylyl cyclase (AC)-cAMP-PKA signaling cascade, β₂AR activates pertussis toxin (PTX)-sensitive G_i proteins in addition to the G_s pathway, with the G_i signaling negating the G_s-mediated positive inotropic effect in the heart[3-5] and promoting cardiomyocyte survival[6]. However, exaggerated β₂AR-G_i signaling is a hallmark of the failing heart of mammalian species, including rat, canine, and human[7-9]. In the failing heart, the enhanced G_i signaling negates β₁AR- as well as β₂AR-mediated contractile response[10-12], thus contributing to the pathogenesis of heart failure. This perception has been recently corroborated by the fact that selective β₂AR-G_s stimulation with fenoterol, bypassing the G_i signaling, markedly improves cardiac structure and function in an ischemic heart failure model[7, 13]. Thus, delineating the mechanism that regulates the β₂AR-G_i signaling will not only deepen our understanding of βAR signal transduction but also bear important pathological and therapeutic implications.

The classic view on βAR signaling involves an agonist-induced change in the receptor conformation that causes the activation of the G_s protein, leading the formation of the second messenger, cAMP, which activates PKA and downstream signaling. The termination of this cascade, a process known as desensitization, occurs when G protein-coupled receptor kinases (GRKs) and the second messenger kinase, PKA, phosphorylate the activated receptor and promote the binding of β-arrestins which sterically block the coupling of G_s to the receptor.

In contrast to the βAR-G_s signaling, the β₂AR-G_i signaling is enhanced by PKA-mediated phosphorylation of the receptor [14, 15]. The fundamental question is what is the mechanism underlying the termination of the β₂AR-G_i signaling. In this regard, it has been shown that upon GPCR activation, GDP is exchanged for GTP on the Gα subunit, resulting dissociation of the Gα from Gβγ subunits and the activation of downstream effortors. The intrinsic GTPase activity of the α subunit of G proteins serves as a molecular clock, turning down GPCR signaling via returning G proteins to the GDP-bond heterotrimeric form. Regulator of G protein signaling (RGS) proteins are GTPase-activating proteins (GAPs) which accelerate GTPase-mediated hydrolysis of GTP to GDP on Gα, thus reconstituting the heterotrimeric G protein complex and terminating G protein signaling[16, 17]. Different RGS proteins have GAP activity towards specific G protein families and, in some cases, even for particular receptors, although the mechanisms for such selectivity remain unknown.

In the heart of mammalian species, major RGS proteins include RGS2-5[18]. In particular, RGS2 can negatively regulate the signaling of G_q-coupled receptors, including α_1A-adrenergic receptor, angiotensin II receptor 1A, and interleukin receptor[19, 20]. Deregulation of RGS2 has been implicated in the pathogenesis of cardiac hypertrophy and hypertension[21]. Interestingly, it has been recently reported that RGS2 can physically interact with β₂AR[22], in addition to G_q-coupled GPCRs such as M₁ muscarinic receptor[23] and α_1AAR[24]. However, it is unknown whether RGS2 regulates G_i signaling in general and the β₂AR-coupled G_i signaling in particular.

In this study, we seek to determine the potential role of RGS proteins, particularly RGS2 in regulating the β₂AR-coupled G_i signaling in a physiologically relevant setting, adult mouse cardiomyocytes. We provide multiple lines of evidence to define RGS2 as a primary terminator of the β₂AR-G_i signaling. First, an elevation in RGS2 expression impairs the β₂AR-G_i signaling, as manifested by the loss of PTX sensitivity of β₂AR contractile response in cultured adult mouse cardiomyocytes and the failure of β₂AR stimulation to increase PTX-sensitive ERK1/2 activation in HEK293 cells. In contrast, RGS2 deficiency, similar to βAR agonist stimulation, sustains the G_i signaling in cultured cardiomyocytes even in the prolonged absence of agonist stimulation. These data have revealed a previously unappreciated novel negative regulation of β₂AR-activated G_i signaling by RGS2 in mammalian hearts.

2. Materials & Methods

2.1. Isolation, Culture and Adenoviral Infection of Adult Cardiomyocytes Isolated from Wild-type or RGS2 Knockout Mice

Single myocytes were isolated from the hearts of 2-3 month-old Black C57 or RGS2 Knockout mice using a standard enzymatic technique, as described previously[25]. In a subset of experiments, isolated myocytes were cultured and infected with target gene-carrying adenoviral vectors in the presence or absence of isoproterenol (ISO, 10 nM). Adenovirus-mediated gene transfer was implemented by adding adenoviral vectors carrying β-galactosidase (Adv-β-gal, as a control virus) or N-terminally Flag-tagged RGS2 (Adv-RGS2), RGS3 (Adv-RGS3), RGS4 (Adv-RGS4) or RGS5 (Adv-RGS5) (each at a multiplicity of infection, M.O.I., of 100). At the M.O.I. used, almost 100% myocytes were positively infected, as evidenced by β-gal staining[25] or GFP fluorescent signal (Adv-GFP infection, unpublished data). All experiments were performed after cells were cultured for 24 h, unless indicated otherwise.

2.2. Cell Contraction Measurements

Cells were perfused with a HEPES-buffered solution, and electrically paced at 0.5 Hz at 23°C. To inhibit G_i signaling, freshly isolated or cultured myocytes were incubated with PTX (1.5 μg/ml, at 37°C for at least 3 h), prior to contraction measurements. Cell sarcomere length was monitored by an optical tracking method using a photodiode array (Model 1024 SAQ, Reticon) as described previously [25].

2.3. Radioligand-Binding Assay

A βAR-specific ligand [¹²⁵I]-iodocyanopindolol ([¹²⁵I]-CYP) binding assay was performed, as described previously[26]. Competition experiments were carried out at 50 pM [¹²⁵I]-CYP in the presence of a highly selective β₂AR antagonist, ICI 118.551, or a highly selective β₁AR antagonist, CGP 20712A. Non-specific binding was determined in the presence of 10 μM propranolol and was about 10-30% of total binding of [¹²⁵I]-CYP (50 pM).

2.4. Western Blotting of G_s and G_i Proteins and RGS proteins

The expression of either G_s or G_i proteins was measured by Western blotting with specific antibodies, as previous described[7]. Endogenous RGS2 protein abundance was assayed by an affinity-purified antibody from GenWay Biotech[26]. Endogenous RGS3, RGS4 and RGS5 protein levels were detected with specific antibodies from Santa Cruz, whereas Flag-tagged RGS2-5 was detected by an antibody specifically reacting with Flag.

2.5 cAMP Measurement

Intracellular cAMP levels were assayed by radioimmunoassay, as previously described [7]. Briefly, cultured mouse cardiomyocytes or HEK 293 cells were treated with the phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine (IBMX, 1 mM), for 30 min at 37°C. The cells then were incubated with zinterol (10 μM, a β₂AR agonist) in the presence of CGP27013A (0.3 μM, a β₁AR selective antagonist) for 10 min. cAMP levels were assayed with the cAMP assay kit (Parameter TM, R& D Systems, # KGE002B) with a duplicate in each experiment. Protein content was measured using the Piece BCA (Thermo Scientific # 23228).

2.5. Materials

Zinterol was kindly supplied by Bristol-Myers, Evansville, IN. Antibodies recognizing the α-subunits of G_s, G_o, G_i1, G_i2, and G_i3, were purchased from Santa-Cruz, USA. The antibody recognizing RGS2 was obtained from GenWay Biotech, while antibodies for other RGS proteins (RGS3, RGS4, RGS5) and secondary antibodies were purchased from Santa Cruz. The antibody reacting with Flag was from Affinity Bioreagents. Pertussis toxin (PTX), isoproterenol (ISO), propranolol and norepinephrine (NE), CGP20712A (CGP), and minimal essential medium (MEM) were purchased from Sigma (St. Louis, MO). ICI 118,551 (ICI) was kindly supplied by ICI Pharmaceutic Group (Wilmington, DE). FBS and Penicillin-Streptomycin were purchased from Gibco (Gaithersburg, MD). Laminin was purchased from Upstate. [¹²⁵I]-CYP was purchased from NEN Life Science Products, Inc. (Boston, MA).

2.6. Statistics

Data reported are mean ± standard error of mean (SEM). Statistical comparisons were made by student's t test or paired t test when appropriate. Two-factor analysis of variance, ANOVA, was used to analyze the overall drug dose-response. A P value of < 0.05 was considered to be statistically significant.

3. Results

3.1. Prolonged Absence of Agonist Stimulation Desensitizes β₂AR-Mediated G_i Signaling and Selectively Elevates RGS2 Expression in Adult Mouse Cardiomyocytes

In adult mouse cardiac myocytes, β₂AR stimulation activates both G_s and G_i proteins with the G_i signaling substantially negating the G_s-mediated contractile response[27]. Consistent with the previous notion[27], the contractile response to a selective β₂AR agonist, zinterol (10^-8 to 10^-5 M), was minor and profoundly enhanced by disrupting G_i signaling with PTX in freshly isolated adult mouse ventricular myocytes (Fig. 1A), without altering cell basal contractility (Table A1 of the online-only Data Supplement). Interestingly, prolonged absence of agonist stimulation markedly augmented zinterol-induced contractile responses which could not be further enhanced by PTX treatment in myocytes cultured for 24 h (Fig. 1B). However, the contractile response to β₁AR stimulation by norepinephrine (NE, plus an α₁AR antagonist prazosin at 10^-6M) remained intact in cultured myocytes (Fig. 1C), suggesting that the excitation-contraction machinery and the G_s-AC-cAMP-PKA signaling cascade are unaltered during 24 h cell culture. These results strongly suggest that the β₂AR-coupled G_i signaling is severely desensitized, while the G_s pathway is intact during 24h absence of agonist stimulation in cultured rodent cardiomyocytes.

Fig. 1. Prolonged absence of agonist stimulation selectively enhances β₂AR-, but not β₁AR-, mediated contractile response in cultured adult mouse cardiomyocytes.

The contractile response of freshly isolated adult mouse ventricular myocytes (Fresh) (Panel A) or those cultured for 24 h (Culture) (Panel B) to a β₂AR selective agonist zinterol or a β₁AR agonist norepinephrine (NE) in the presence of an αAR antagonist (prazosin, 10^-6 M) (Panel C) (n = 8∼14 cells from at least 6 hearts for each data point). Note that PTX markedly enhances β₂AR-mediated contractile response in freshly isolated, but not in cultured, cardiomyocytes. There is no significant difference between freshly isolated cells and cultured cells in their basal contraction amplitudes in the presence or absence of PTX treatment (Table 1S, data supplements).

Because many RGS proteins are GAPs for Gα_i/o family and negatively regulate G_i/o signaling in in vitro experimental systems[28, 29], we hypothesized that RGS proteins may be involved in the desensitization of the β₂AR-G_i signaling. To determine which particular RGS proteins regulate the β₂AR-G_i signaling, we examined β₂AR-G_i-dependent activation of ERK1/2 MAPK in HEK 293 cells in the presence or absence of adenoviral overexpression of RGS2, RGS3, RGS4 and RGS5. β₂AR stimulation with ISO (1.0 μM) increased ERK1/2 phosphorylation level by 3-4 folds in a PTX-sensitive manner in uninfected HEK 293 cells or those infected with a control adenovirus, Adv-β-gal (100 M.O.I..), (Fig. 2 A&B), consistent with previous report[14]. Importantly, adenoviral overexpression of Flag-tagged RGS2 or RGS5 fully abolished the β₂AR-G_i-mediated activation of ERK1/2 (Fig. 2A), indicating that both RGS2 and RGS5 negatively regulate the β₂AR-mediated G_i signaling. In contrast, overexpression of RGS3 or RGS4 to a greater protein level had no significant effect on β₂AR-induced, G_i-dependent ERK1/2 activation (Fig. 2A&B).

Fig. 2. Effects of overexpression of RGS2-5 on the β2AR-Gi-dependent activation of ERK1/2 in HEK293 cells.

Cells were transferred with an adenovirus (at 10 M.O.I.) expressing β-gal or Flag-tagged RGS2, RGS3, RGS4, and RGS5 for 24 h and subjected to serum starvation overnight. A. Activation of ERK1/2 by β2AR stimulation with isoproterenol (ISO) was inhibited by overexpression of RGS2 or RGS5 but not by overexpression of RGS3 and RGS4. B. Disrupting G_i signaling with PTX treatment suppressed ISO-induced activation of ERK1/2 in HEK293 infected with an adenovirus (at 10 M.O.I.) expressing β-gal or Flag-tagged RGS3 or RGS4. Cells were stimulated with ISO (1 μM) for 5 min with or without PTX (0.5 μg/ml) pretreatment and reactions were stopped by adding 100 μl of cell lysis buffer. Phosphorylation of ERK1/2 was assayed by Western blotting using a site-specific antibody reacting with phosphorylated ERK1/2. * P<0.01 versus respective untreated group; †P<0.01 versus β-gal group in the absence of PTX (n=3-4).

Next, we examined the expression profile of a panel of major cardiac RGS proteins (RGS 2, RGS3, RGS4, and RGS5) at mRNA and protein levels using real-time PCR and Western blotting, respectively, in freshly isolated or cultured cardiomyocytes. To our surprise, among examined RGS proteins, only RGS2 expression was increased at both mRNA (Fig. 3A) and protein levels (Fig. 3 B), in the absence of alterations in other major cardiac RGS proteins (Fig. 3A&B), in cultured cardiomyocytes compared to freshly isolated cells. Thus, RGS2 was selectively upregulated in response to the prolonged absence of agonist stimulation (Fig. 3), although when overexpressed both RGS2 and RGS5 blocked the β₂AR-G_i-dependent ERK1/2 activation (Fig. 2A). RGS2 protein level was similarly elevated in cultured adult rat and mouse cardiac myocytes, but basal abundance of RGS2 was relatively lower in mouse myocytes (Fig. 4A&B). Confocal immunostaining revealed that RGS2 was largely located along cell subsurface membranes in both freshly isolated and cultured myocytes (Fig. 4B).

Fig. 3. Expression of RGS2-5 at mRNA and protein levels in freshly isolated or cultured cardiomyocytes.

A. mRNA levels of RGS2-5 assayed by real-time PCR (n= 3 independent experiments; * P< 0.01 vs. RGS3, RGS4 and RGS5). B. Representative Western blots probed with an antibody reacting with RGS2, RGS3, RGS4, or RGS5 or β-actin in freshly isolated or cultured cardiomyocytes.

Fig. 4. Expression and intracellular distribution of RGS2 in freshly isolated or cultured adult rat or mouse cardiomyocytes.

A. A representative Western blot and average data of RGS2 probed with an antibody reacting with RGS2 (* P< 0.05 vs. freshly isolated mouse cardiomyocytes; †P<0.01 vs. respective freshly isolated myocytes; n=6 hearts for each group). B. Typical confocal immunacytochemical imaging and average data of RGS2 in freshly isolated or cultured cardiomyocytes (n= 41 and 43 cells from 6 hearts, respectively; *P<0.001 vs. the value of freshly isolated cells)

3.2. Agonist Stimulation Prevents RGS2 upregulation and Restores β₂AR-G_i Signaling in Cultured Cardiomyocytes

Next, we determined whether the selective upregulation of RGS2 can be prevented by adding a βAR agonist during cell culture. Indeed, administration of a βAR agonist, ISO (1.0 nM for 24 h), blocked the upregulation of RGS2 in cultured adult mouse cardiomyocytes (Fig. 5 A&B). Previous studies have reported that activation of PKA with βAR stimulation or forskolin increases rather than decreases the expression of RGS2 in cardiac myocytes and other cell type[30-32] However, in those previous studies, cells were subjected to a short-term (2-3h) instead of 24h treatment of βAR agonists or forskolin. We then compared the expression of RGS2 in cardiomyocytes subjected to either a short-term (2 h) or a prolonged (24 h) treatment with ISO (1 nM) or forskolin (1 μM). Consistent with previous studies, incubation of cells with ISO (1 nM) or forskolin (1 μM) for 2h significantly elevated RGS2 protein abundance by 2-3 folds (Fig. 5 A&B). In sharp contrast, treatment of myocytes from the same hearts with the same agents for 24h reduced RGS2 expression as compared to the baseline of cultured cells in the absence of ISO (Fig. 5), suggesting there is a biphasic regulation of RGS2 expression at protein level by either ISO or forskolin. Together, our data indicate that the upregulation of RGS2 was largely prevented by the presence of the βAR agonist, ISO (1.0 nM for 24 h), during cell culture.

Fig. 5. Biphasic regulation of RGS2 expression by βAR stimulation with ISO (1.0 nM) or adenylate cyclase activation with forskolin (1.0 μM).

Panels A and B show, respectively, a representative Western blot and the average data probed with an antibody reacting with RGS2 in freshly isolated adult mouse cardiomyocytes or those cultured for 2h or 24 h in the presence or absence of ISO (1.0 nM) or forskolin (1.0 μM). The same membrane was striped and re-probed with anti-β-actin (the top panel). * P<0.01 vs. fresh mouse cardiomyocytes; †P<0.01 vs. as indicated; at least three independent experiments for each group.

Next, we investigated whether ISO can restore β₂AR-mediated G_i signaling as indexed by the PTX sensitivity of the receptor-induced contractile response in cultured adult mouse cardiomyocytes. Remarkably, ISO (1.0 nM for 24 h) fully prevented the augmentation of β₂AR-mediated contractile response, and restored its PTX sensitivity (Fig. 6), implying a full restoration of the receptor-coupled G_i signaling by agonist stimulation. These results suggest that RGS2 negatively regulated the β₂AR- G_i signaling, and that agonist -induced β₂AR-G_i signaling is likely mediated by suppressing RGS2 expression.

Fig. 6. Agonist stimulation restores β₂AR-mediated G_i signaling in cultured myocytes.

Addition of 1.0 nM ISO into the culture medium fully restores the PTX sensitivity of β₂AR contractile response. Treatment with ISO or PTX did not alter the baseline contraction (see Table S1, online-only data supplement).

3.3. Overexpression of RGS2 Suppresses Agonist-Restored β₂AR-G_i Signaling in Cultured Mouse Cardiomyocytes

To determine whether there is a causative relationship between the upregulation of RGS2 and the impairment of β₂AR-activated G_i signaling in cultured myocytes, we overexpressed RGS2 protein by approximately 2-fold over baseline using adenoviral gene transfer (Fig. 7A). The increase in the contraction amplitude induced by the β₂AR agonist, zinterol (5×10^-7 M, a concentration close to EC₅₀), was clearly augmented in cultured adult mouse cardiomyocytes infected with Adv-RGS2, regardless of the presence of ISO (1.0 nM for 24 h) and the absence of PTX (Fig. 7B). Importantly, overexpression of RGS2 diminished PTX-sensitive component of β₂AR-mediated contractile response (Fig. 7B). These results indicate that similar to the situation in HEK293 cells (Fig. 2A), an elevation of RGS2 abundance is sufficient to suppress agonist-induced β₂AR-G_i signaling in cardiac myocytes and subsequently augments β₂AR-induced contractile response.

Fig. 7. Adenoviral gene transfer-mediated overexpression of RGS2 abolishes agonist-restored β₂AR-G_i coupling in cultured adult mouse cardiomyocytes.

Panel A shows a typical western blot of RGS2 in cultured adult mouse cardiomyocytes infected by Adv-β-gal or Adv-RGS2 (both at m.o.i. 100 for 24 h). Panel B the average contractile response to the β₂AR agonist, zinterol (5×10^-7 M) in cultured adult mouse cardiomyocytes infected by Adv-β-gal or Adv-RGS2 (both at m.o.i. 100 for 24 h) (*P<0.01 vs. groups without PTX; †P<0.01 as indicated; n = 6-13 cells from 3 mouse hearts). Note that RGS2 overexpression led to augmented β₂AR contractile response with reduced PTX-sensitive component, indicating that RSG2 overexpression suppresses ISO-restored β₂AR-G_i signaling in cultured myocytes. Panel C Overexpression of RGS2 with adenoviral gene transfer increased β₂AR-induced cAMP formation in cultured adult mouse cardiomyocytes infected by Adv-β-gal or Adv-RGS2 (both at m.o.i. 100 for 24 h). Cultured myocytes were stimulated with a β₂AR agonist, zinterol (10 μM) in the presence of β₁AR selective antagonist, CGP27013A (0.3 μM) and IBMX (1 mM) for 10 min (*P<0.01 vs. respective control for each group; †P<0.01 as indicated; n = 4-6 independent experiments).

3.4 RGS2 Ablation Sustains the b₂AR-G_i Signaling in Cultured Cardiomyocytes in the Absence of Agonist Stimulation

To further define the role of RGS2 protein in regulating the β₂AR-coupled G_i signaling, we took advantage of a gene-targeted RGS2 knockout (KO) mouse model[33]. Western blotting confirmed the absence of RGS2 protein in myocytes isolated from RGS2 KO mice (Fig. 8A). The most important difference between RGS2 KO and WT groups is that β₂AR-mediated contractile response is sensitive to PTX treatment in cultured RGS2 KO cardiomyocytes but not in cultured WT cells (Fig. 8B), indicating that RGS2 ablation is able to retain the β₂AR-G_i signaling even in the prolonged absence of βAR agonist stimulation. Thus, RGS2 deficiency leads to constitutive β₂AR-G_i signaling which is independent of agonist stimulation, whereas upregulation of RGS2 suppresses the β₂AR-mediated G_i signaling and subsequently augments β₂AR-induced contractile response.

Fig. 8. RGS2 ablation retains the β₂AR-G_i signaling in cultured mouse cardiomyocytes in the absence of agonist stimulation.

The cell contractile response to zinterol (5×10^-7 M, a dose close to the EC₅₀) in freshly isolated or cultured myocytes from WT or RGS2 KO mice (n = 10∼16 cells from 8 hearts for each data point; *, P<0.01 vs. all groups with PTX and cultured WT without PTX; †P< 0.05 as indicated). Note that in cultured cardiomyocytes, 24h-culture impaired the β₂AR-G_i signaling in WT but not RGS2-deficient myocytes.

3.5. Neither the Expression Level of G_s and G_i Proteins Nor the Expression and Ligand Binding Properties of β₁AR and β₂AR Are Altered in Cultured Cardiomyocytes

In principle, prolonged absence of agonist stimulation-caused desensitization of the β₂AR-G_i signaling could be attributable to a downregulation of G_i proteins, in addition to upregulation of RGS2. To test this possibility, we first examined the protein level of both G_s and PTX-sensitive G proteins (Gα_o, Gα_i1, Gα_i2, Gα_i3) in freshly isolated or cultured cells, and found no significant difference between the two groups (Fig. A1 of the online-only Data Supplement).

Table A2 (online-only Data Supplement) summarizes β₁AR and β₂AR densities and ligand binding properties assayed with a β₁AR or a β₂AR selective antagonist, CGP20712A or ICI 118,551, respectively. The density of total βARs, the β₁AR and β₂AR subpopulations, and their ratio in freshly isolated cardiomyocytes were similar to those of cultured cells. In addition, there was no significant difference between the two groups in β₁AR or β₂AR binding properties for a radioligand, [¹²⁵I]-CYP. Thus, the failure of β₂AR to activate G_i signaling in cultured myocytes is not caused by alterations in the expression of G proteins or β₂AR density or ligand binding properties.

4. Discussion

4.1. RGS2 Constitutively Inhibits β₂AR-mediated G_i Signaling in Cardiac Myocytes

Multiple RGS proteins, including RGS2-5, are expressed in the heart of mammalian species[18, 19]. In the present study, we have defined RGS2 as a powerful endogenous terminator of β₂AR-mediated G_i signaling in the physiologically relevant setting, adult mouse cardiomyocytes. Upregulation of RGS2 caused by the lack of βAR agonist stimulation suppresses the β₂AR-G_i signaling in cultured cardiomyocytes, as evidenced by the lack of PTX sensitivity of β₂AR contractile response. Similarly, overexpression of RGS2 with adenoviral gene transfer markedly diminished agonist-restored PTX sensitivity of β₂AR contractile response in cultured cardiomyocytes, and fully blocked β₂AR-G_i-dependent activation of ERK1/2 in HEK293 cells. In contrast, RGS2 ablation is able to sustain the β₂AR-G_i signaling even in the prolonged absence of βAR agonist stimulation in cells cultured for 24 h. To our knowledge, this is a previously unappreciated intrinsic cellular mechanism responsible for the termination of β₂AR-activated G_i signaling in a physiologically relevant system, adult mouse cardiomyocytes.

Since RGS2 constitutively binds to β₂AR[22], it is reasonable to assume that the β₂AR-coupled G_i signaling is constitutively suppressed by RGS2 under physiological conditions, leading to apparent G_s-predominant β₂AR signaling in cardiac myocytes of most mammalian species except mouse[2, 27]. The high efficiency of β₂AR-mediated G_i signaling in mouse cardiomyocytes is likely attributed to the relatively low basal level of cardiac RGS2 expression in this particular species (Fig. 4A). While these present findings seem to be contradicting the previous notion that RGS2 selectively regulates the signaling of G_q but not G_i/o proteins in cultured cardiac myocytes or in the heart[19] and in vitro experimental systems[34-36], emerging evidence suggests that RGS2 regulates G_i/o-, but not G_q-, mediated signaling pathway in hippocampal neurons[37] and in cardiomyocytes derived from embryonic stem cells[38].

In addition to its GAP activity, RGS2 has been shown to inhibit adenylyl cyclase (AC) activity by directly binding to Gα_s (without displaying GAP activity for this Gα subunit)[19] or AC[39]. But we have demonstrated that the upregulation of RGS2 in cultured cardiomyocytes has no detectable effect on β₁AR-G_s-AC-mediated contractile response (Fig. 1C), suggesting RGS2 does not directly regulate AC activity in rodent cardiac myocytes, consistent with our previous notion[19].

4.2. Dynamic Regulation of RGS2 Expression by G Protein-mediated Signaling

Recent studies have shown that mRNA and protein levels of RGS2, but not RGS3-5, are increased in response to enhanced signaling of G_q and G_s (βAR or forskolin stimulation for 3 h) in adult rat ventricular cardiomyocytes[19] and osteoblasts[31, 39]. Interestingly, we have demonstrated, for the first time, that there is a biphasic regulation of RGS2 expression by enhanced G_s signaling induced by ISO or AC activation with forskolin. Short-term (2 h) treatment of cells with ISO or forskolin increases RGS2 protein level by 2-fold, whereas prolonged treatment (24 h) with the same stimuli significantly reduces RGS2 expression relative to untreated cultured myocytes (Fig. 5). Although the underlying mechanism of the biphasic regulation of RGS2 expression by G protein-mediated signaling remains to be explored, it might explain the opposing regulation of RGS2 expression by acute and chronic augmentation of G_q signaling. For instance, acute activation of G_q induced by α₁AR stimulation increases RGS2 expression[19], whereas sustained enhancement of G_q signaling, as is the case in cardiac hypertrophy models, selectively suppresses RGS2 expression[40]. Taken together, the present and previous findings suggest that RGS2 is dynamically regulated by GPCR-mediated signaling, highlighting important physiological and pathological relevance of the fine tune of this particular RGS protein.

4.3. Receptor Activation-Dependent β₂AR-G_i signaling Is Attributable to Downregulation of RGS2

In this study, we have shown that prolonged absence of agonist stimulation causes the loss of β₂AR-activated G_i signaling, resulting in enhanced β₂AR contractile response in cultured cardiomyocytes. Chronic agonist stimulation (ISO 1.0 nM included in the culture medium) rescues the G_i signaling. Consistent with the perception that phosphorylation of the receptor by PKA is required for the G_i signaling[14], these present data indicates that the β₂AR-G_i signaling is dependent on the receptor activation. Most importantly, the present study has revealed, for the first time, that the switch of the receptor signaling from G_s to G_i pathway is mediated, at least in part, by βAR-induced downregulation of RGS2 protein (Figs. 5&6).

4.4. Potential Clinical Implications of Cardiac RGS2 Deregulation

The selective upregulation of RGS2 and the resultant augmentation of β₂AR contractile response induced by the lack of βAR stimulation in cultured myocytes are intriguing given the fact that β-blocker therapy can resensitize βAR-mediated contractile support and improve cardiac function in patients with congestive heart failure. The beneficial effects of β-blockers might be mediated, in part, by increasing RGS2 expression. In contrast, adrenergic overdriving, as is the case in hypertension and cardiac hypertrophy in different animal models, is accompanied by a selective downregulation of RGS2[40-42]. Under those pathological circumstances, the reduction in RGS2 expression is expected to enhance G_q- and G_i-mediated signaling. In this regard, recent studies have shown that RGS2 gene silencing blocks α₁AR-induced cardiac myocyte hypertrophy [19], and that pressure overload by trans-aortic constriction results in enhanced G_q signaling, exacerbated cardiac hypertrophy, heart failure and premature death in RGS2-deficient mice as compared to wild type counterparts[33], implying that RGS2 may play a central role in protecting the heart against stress-induced maladaptive remodeling.

In addition to the well-established G_q signaling, the current study has demonstrated that the lack of this RGS protein promotes β₂AR-mediated G_i signaling, whereas overexpression of RGS2 suppresses the G_i signaling in cardiomyocytes. In the failing heart, both catecholamine levels and G_i expression are elevated, the G_i signaling is exaggerated[7-9], leading to defects of both β₁AR- and β₂AR-mediated positive inotropic effects[10-12]. The downregulation of RGS2 likely contributes to the enhancement of the β₂AR-coupled G_i signaling, which, in turn, may accelerate the development of hypertension and cardiac hypertrophy, eventually resulting in heart failure.

5. Conclusion

In the present study, we have demonstrated, for the first time, that RGS2 confers a powerful negative regulation of β₂AR-activated G_i signaling in mammalian cardiac myocytes. Deregulation or malfunction of RGS2 may constitute a pathogenic element for the development of heart failure in addition to its known role in the pathogenesis of hypertension. Thus, the current findings not only define RGS2 as a novel cellular mechanism responsible for the termination of β₂AR-mediated G_i signaling, but also bear important pathogenic and therapeutic implications.

Appendix A

Supplementary data associated with this article can be found in the online version.

Abbreviations

AC

adenylyl cyclase

Adv

adenovirus

GPCR

G protein-coupled receptor

βAR

β-adrenergic receptor

WT

wild type

ISO

isoproterenol

CGP

CGP20712A

ICI

ICI 118,551

Tpeak

the time from stimulation to peak shortening

T₅₀

the time from the peak to 50% relaxation

[¹²⁵I]-CYP

[¹²⁵I]- iodocyanopindolol

m.o.i

multiplicity of infection

PKA

protein kinase A

PKI

a peptide inhibitor of PKA

RGS

regulator of G-protein signaling

PTX

pertussis toxin

GAP

GTPase-activating protein

ERK

Extracellular signal-regulated kinases