Cardiac‐Restricted Overexpression of the A_2A‐Adenosine Receptor in FVB Mice Transiently Increases Contractile Performance and Rescues the Heart Failure Phenotype in Mice Overexpressing the A₁‐Adenosine Receptor

Abstract

In the heart, adenosine binds to pharmacologically distinct G‐protein‐coupled receptors (A₁‐R, A_2A‐R, and A₃‐R). While the role of A₁‐and A₃‐Rs in the heart has been clarifed, the effect of genetically manipulating the A_2A‐R has not been defned. Thus, we created mice overexpressing a cardiac‐restricted A_2A‐R transgene. Mice with both low (Lo) and high (Hi) levels of A_2A‐R overexpression demonstrated an increase in cardiac contractility at 12 weeks. These changes were associated with a signifcantly higher systolic but not diastolic [Ca²⁺]ⁱ, higher maximal contraction amplitudes, and a signifcantly enhanced sarcoplasmic reticulum Ca²⁺ uptake activity. At 20 weeks, the effects of A_2A‐R overexpression on cardiac contractility diminished. The positive effects elicited by A_2A‐R overexpression differ from the heart failure phenotype we observed with A₁‐R overexpresson. Interestingly, coexpression of A_2A‐R TG_Hi, but not A_2A‐R TGLo, enhanced survival, prevented the development of left ventricular dysfunction and heart failure, and improved Ca²⁺ handling in mice overexpressing the A₁‐R. These results suggest that adenosine‐mediated signaling in the heart requires a balance between A₁‐ and A_2A‐Rs—a fnding that may have important implications for the ongoing clinical evaluation of adenosine receptor subtype‐specifc agonists and antagonists for the treatment of cardiovascular diseases.

Introduction

The ubiquitous purine nucleotide, adenosine, regulates a variety of cardiac pathways. including growth and differentiation, angiogenesis, coronary blood fow, cardiac conduction, and heart rate, as well as mediating the cardioprotective process of ischemic preconditioning. These biologic actions of adenosine are mediated by a family of G‐protein‐coupled receptors found on the sarcolemmal surface of cardiac myocytes, including the A₁‐, A_2A‐, a n d A ₃‐ adenosine receptors (R) and the A_2B‐R, which is expressed only in the cardiac vasculature. ¹ In contrast to other G‐protein‐coupled receptors in the heart, activation of select adenosine receptor subtypes can result in diametrically opposite physiologic efects. For example, activation of the A₁‐R or A₃‐R pathways inhibits adenylyl cyclase due to coupling with the inhibitory guanine nucleotide binding protein G_i. ² By contrast, activation of the A_2A‐R enhances the production of the intracellular second messenger cAMP through G_s‐mediated activation of adenylyl cyclase. Thus, these two receptor‐mediated signaling pathways subserve very different physiologic functions, leading investigators to propose that the role of A_2A‐R in the normal heart is to reduce A₁‐R mediated antiadrenergic actions. ³

Studies using transgenic mouse models harboring a transgene overexpressing the A₁‐R or A₃‐R ⁴ , ⁵ , ⁶ as well as mice in which the A₁‐R has been ablated ⁷ have helped to clarify the role of these receptor subtypes and have confirmed their importance in cardiac protection during ischemia‐reperfusion and myocardial infarction. ⁸ Interestingly, recent studies using transgenic models have suggested that robust activation of A₁‐R and A₃‐R can have adverse consequences on cardiac morphology and function, albeit in a strain‐ and dose‐specifc manner. ⁹ , ¹⁰ , ¹¹

While transgenic models have helped to clarify the role of A₁/A₃‐R pathways in cardiac physiology, the role of A_2A‐R pathway in the heart has remained controversial. For example, studies have shown disparate effects of A_2A‐R activation on cardiac contractility. ¹² , ¹³ , ¹⁴ , ¹⁵ This controversy is due in large part to the fact that subtype‐selective adenosine receptor agonists and antagonists lack receptor‐subtype selectivity and to species‐specifc responses to pharmacologic agents. ¹⁶ Therefore, to more efectively evaluate the role of A_2A‐R in cardiac physiology, we created transgenic mice with cardiac‐restricted overexpression of the A_2A‐R. In contrast with mice overexpressing the A₁‐R, overexpression of the A_2A‐R markedly enhanced cardiac function. These salutary efects of A_2A‐R activation on myocardium were due at least in part to enhanced sarcoplasmic reticulum Ca²⁺ uptake, resulting in higher systolic but not diastolic [Ca²⁺]_i. Furthermore, overexpression of the A_2A‐R prevented the cardiodepressant efects of A₁‐R activation. These results provide important insights regarding the use of adenosine receptor subtype‐selective agonists and antagonists in the treatment of patients with cardiovascular diseases.

Materials and Methods

Transgenic mouse generation

Experiments were carried out in transgenic mice with cardiac‐restricted constitutive overexpression of the human A_2A‐R engineered on an FVB background as previously described. ⁹ , ¹⁷ All protocols were approved by the Institutional Animal Care and Use Committee of Thomas Jeferson University.

Methods: Calculating Copy Number of Inserted Transgene

To quantify the number of transgenes inserted into the genome, genomic DNA from mouse‐tail was isolated using the Qiagen DNAeasy kit. 100 ng of genomic DNA were used for Real‐time PCR using the A_2A‐AR‐specifc primer set (5′‐CAG CTG AAG CAG ATG GAG AGC‐3′; 5′‐GAT GGC CAG GTA CAT GAG CCA‐3′) and actin set (5′ AGG ACC TGT ACG CCA ACA AC 3′, 5′ACA TCT GCT GGA AGG TGG AC 3′). A_2A‐R‐specifc primers were conserved between mouse and human A_2A‐AR and detected human and mouse A_2A gene with equal efciency (data not shown).

Real time PCR was performed in a 30 μL reaction (100 ng of genomic DNA; 250 nM each primer; 1X SYBRE Green Master Mix). Each experimental sample was performed in triplicate. The ACT method was used to quantify the results, which are presented as relative fold changes to the actin gene. Each primer set was designed to assume a Tm of 60°C and 50% GC content.

Echocardiography

Echocardiographic studies were performed using an ultrasonographic system (ACUSON Sequoia C256, Siemens, Malvern, PA, USA) as described. ¹⁸ , ¹⁹ Age matching, nontransgenic mice in FVB background served as controls. Mice were anesthetized with 2.5% Avertin (10 μL/g body weight, IP, Aldrich Chemical Co) and placed in the supine position. A 14‐MHz transducer was applied to the left hemithorax. Two‐dimensional targeted M‐mode imaging was obtained from the short‐axis view at the level of the greatest lef ventricular dimension at baseline. M‐mode measurements of left ventricular end‐diastolic and end‐systolic diameter and left ventricular anterior‐ and posterior‐wall thickness were made using the leading‐edge convention of the American Society of Echocardiography. End diastole was determined at the maximal left ventricular diastolic dimension, and end systole was taken at the peak of posterior‐wall motion.

Left ventricular hemodynamics measurement

After anesthetization with 2.5% Avertin (10 μL/g body weight, IP, Aldrich Chemical Co), mice were placed in the supine position. A 1.4 F micromanometer catheter (Millar Instruments) was inserted into the left ventricle through the right carotid artery. ¹⁸ , ¹⁹ Left ventricular pressure and heart rate were then recorded.

Immunoblotting and histopathology of myocardium

Frozen ventricular tissues were homogenized on ice using a nonionic detergent‐based lysis buffer (25 mM Tris‐HCl pH 7.6, 137 mM NaCl, 10% glycerol, 1% NP40, or IGEPAL CA‐630, 10 mM NaF) freshly supplemented with 1 mM sodium pyrophosphate, 5 μg/mL leupeptin, 5 μg/mL aprotinin, 1 mM EDTA, 10 mM PMSF, and 1 mM NaVO4. After electrophoresis in SDS‐PAGE and transfer onto nitrocellulose membranes, the blots were probed with anti‐A₁‐R (Affinity BioReagents), anti‐A_2A‐R (Millipore), anti‐actin (Sigma), anti‐Gαi (Abcam), anti‐NCX1 (Swant), anti‐SERCA2 (Bethyl lab), anti‐pT308‐Akt (Cell Signaling), anti‐total Akt (BD Biosciences), anti‐calsequestrin (Swant) and anti‐Na⁺‐K⁺‐ATPase (Gif from Dr. R. Levenson, Pennsylvania State University) as previously described. ⁹ , ²⁰ For anti‐A₁‐R and anti‐A_2A‐R blots, special precaution was taken to denature protein extract in the absence of reducing agent at 70°C for 5 minutes. All blots were incubated with either IRDye 700 or 800 secondary antibodies and visualized using the Odyssey Infrared Imaging System software (Li‐Cor, Lincoln, NE, USA). Histopathology staining was performed by RADIL, University of Missouri and imaged at TJU Pathology Imaging Facility.

Isolation of adult murine cardiac myocytes

Cardiac myocytes were isolated from the septum and left ventricular free wall of wild‐type and transgenic mice (male, 8–9‐week old) as recently described. ²⁰ Briefy, mice were heparinized (1,500 U/kg ip) and anesthetized (pentobarbital sodium, 50 mg/kg ip). Excised heart was mounted on a steel cannula and retrograde perfused (100 cmH₂O, 37°C) with Ca²⁺‐free bicarbonate bufer followed by enzymatic digestion (collagenases B and D, protease XIV). Isolated myocytes were cultured on laminin‐coated glass cover slips and the Ca²⁺ concentration of the buffer was progressively increased from 0.05 to 0.125 to 0.25 to 0.5 mM in three steps (10‐minute interval each). The 0.5 mM Ca²⁺ buffer was then aspirated and replaced with minimal essential medium (MEM, Sigma M1018) containing 1.2 mM Ca²⁺, 2.5% FBS, and antibiotics (1% penicillin/streptomycin). After 1 hour (5% CO₂, 37°), media were replaced with FBS‐free MEM. Myocytes were used within 2–8 hours of isolation.

Myocyte shortening measurements

Myocytes adherent to cover slips were bathed in 0.6 mL of air‐ and temperature‐equilibrated (37°), HEPES‐bufered (20 mM, pH 7.4) medium 199 containing 0.6, 1.8, or 5.0 mM [Ca²⁺]_o. Measurements of myocyte contraction (1 Hz) were performed using a charge‐coupled device video camera and edge‐detection software (Ionoptix, Milton, MA, USA) as previously described. ²⁰ , ²¹ , ²²

[Ca²⁺]_i transient measurements

Myocytes were exposed to 0.67 μM of fura‐2 AM for 15 minutes at 37°C Fura‐2‐loaded myocytes were feld‐stimulated to contract (1 Hz, 37°C) in medium 199 containing 0.6, 1.8, or 5.0 mM [Ca²⁺]_o. Fura‐2‐loaded myocytes mounted on [Ca²⁺]_i transient measurements using a Dvorak‐Stotler chamber situated in a temperature‐controlled stage (37°C) of a Zeiss IM 35 inverted microscope were performed as previously described. ²⁰ , ²¹ , ²²

Immunofuorescence microscopy

Freshly isolated myocytes were fxed in 2 mM EGTA/phosphate‐buffers saline (PBS‐EGTA) containing 4% paraformaldehyde, 20 minutes at room temperature. Cells were washed with PBS‐EGTA and pelleted by centrifugation (Beckman GS6, 600 rpm, 5 minutes). Approximately 3,000 rod‐shaped myocytes suspended in 600 μL PBS‐EGTA were mounted onto polylysine‐coated slides by cytospin (400 rpm, 2 minutes). The mounted myocytes were made permeable by incubating with blocking buffer (0.01 M Tris‐HCl, pH7.5, 0.15 M NaCl, 0.1% BSA and 0.002% NaN3) containing 0.2% Triton X‐100 for 5 minutes at room temperature. Subsequently, the slides were incubated with anti‐A₁‐R (Affinity BioReagents), anti‐A_2A‐R (Millipore and Affinity BioReagents) and anti‐α‐actinin (Sigma) antibodies in a 37°C humidified incubator for 45 minutes. After rinsing with blocking buffer, the myocytes were incubated with Alexa‐594‐conjugated anti‐rabbit and Alexa‐488‐conjugated anti‐mouse secondary antibodies (1:2,000) and 350 nM DAPI nuclear stain. After incubation and washing, the myocyte slides were mounted and sealed with ProLong® Gold reagent (Invitrogen) and stored in the dark at 2–6°C Myocyte images were taken with a Confocal Microscope (Zeiss LSM 510 META Confocal, KCC Bioimaging facility).

Statistics

All results are expressed as means ± SE. Kaplan‐Meier survival curves were compared between groups using log‐rank tests. Two‐way analysis of variance was used to analyze the calcium transient and contraction results. Commercial sofware package were used for all statistical analysis (JMP version 4.05; SAS Institute, Cary NC, USA) or SPSS for Windows (version 11.5). Categorical differences were analyzed using the Mann‐Whitney test. In all analyses, p < 0.05 was taken to be statistically significant.

Results

Creation of transgenic mice overexpress ing the A_2A‐adenosine receptor

We placed the human A_2A‐ adenosine receptor (A_2A‐R) cDNA under the control of a cardiac‐specifc promoter as described previously. ⁹ Using an anti‐A_2A‐R antibody, we analyzed levels of A_2A‐R expression in both wild‐type and 15 lines of transgenic mice. Based on these measurements, the transgenic lines were classified as low expression (2–5×, A_2A‐TG_Lo) or high expression (more than 50×, A_2A‐TG_Hi) ( Figure 1A ). Transgenic expression was confirmed at the mRNA level by real‐time polymerase chain reaction RT‐PCR (Data not shown). Consistently, RT‐PCR assay showed that the A_2A‐TG_Lo line has seven genomic copies of the transgene, whereas, the A_2A‐TG_Hi line has over 25 copies ( Figure 2 ). Immunofuorescence staining of permeabilized myocytes showed that the overexpressed A_2A‐R receptors were mostly localized at the sarcolemma ( Figure 1B ).

Figure 1.

Creation of transgenic mice expressing cardiac‐specific a2a‐r. (a) 
A_2a‐R transgenic founder lines expressing low and high levels of A_2a‐R. Total ventricular protein extracts from 6‐week‐old male mice were immunoblotted with anti‐A_2a‐R antibody.
(B) Immunofuorescence staining of isolated myocytes showing A_2a‐R and α‐actinin. 60× whole cell and enlarged images are shown.
(C) Morphology of isolated myocytes and their length and width measurements. 20× images are shown. Values are means +/‐SE. n= 38–56 myocytes pooled from four to five mice.

Figure 2.

Genomic copies of inserted A_2a‐R transgene. Copy number was calculated by real‐time pcr using 100 ng of genomic dna and mouse/human specifc A_2a‐R primers. Data were normalized to actin and wild‐type A_2a‐R gene values. Each sample was measured in triplicate and 4–6 samples were used for each genotype. Graph is shown with standard error.

A_2A‐R overexpression increased cardiac contractility

As seen in Table 1 , overex pression of the A_2A‐R resulted in a small but significant increase in left ventricular mass (ventricular wt/body wt) at 8–12 weeks in mice with high levels of A_2A‐R overexpression but not in those with low levels of A_2A‐R overexpression. Interestingly, heart rates and wall thickness were significantly increased when compared with WT mice, but they were not significantly different between A_2A‐TG_Hi and A_2A‐TG_Lo transgenic lines ( Table 1 ). Furthermore, measurements of single cell morphology showed no increase in either ventricular myocyte width or length ( Figure 1C ). Echocardiography demonstrated an increase in heart rate and fractional shortening as well as a signifcant decrease in left ventricular end‐systolic dimension (LVESD) in mice with both high and low levels of A_2A‐R overexpression ( Table 1 ).

Table 1.

Basal physiology of high and low expression lines of A_2a‐TG mice.

	WT	A2a‐TGHi	p value vs. Wt	A2a ‐TGlo	p value vs. Wt
N =	15	17		12
VW/BW	3.86	4.57	<0.001	4.04	NS
	+/−0.04	+/−0.2		+/−0.09
Heart Rate	372	457	<0.001	494	<0.001
	+/−13	+/−13		+/−11
LVEDD, mm	3.28	3.04	NS	2.94	<0.005
	+/−0.06	+/−0.01		+/−0.08
LVESD, mm	1.86	1.35	<0.005	1.33	<0.005
	+/−0.09	+/−0.12		+/−0.09
FS,%	43.6	56.6	<0.005	55.05	<0.005
	+/−Ί.9	+/−2.5		+/−2.4
AWT	0.70	0.90	<0.005	0.90	<0.005
	+/−0.02	+/−0.03		+/−0.03
PWT	0.73	0.85	<0.005	0.95	<0.005
	+/−0.02	+/−0.03		+/‐0.03

Organ weights and echocardiography data in mice expressing high levels (A_2a‐TG_Hi) or low levels (A_2a‐TG_Lo) of A_2a‐R. Mice were examined at 8–12 weeks. VW/BW = ventricular weight/body weight. LVEDD and LVESD = left ventricular end‐diastolic and ‐systolic dimensions respectively. FS =% of Fractional Shortening; AWT = anterior wall thickness; PWT = posterior wall thickness.

Overexpression of the A_2A‐R prevents the development of the cardiomyopathic phenotype in mice overexpressing the A₁‐R

In a previous study, we demonstrated that constitutive overexpression of the A₁‐R decreased cardiac contractility in FVB mice. ⁹ Te present finding that A_2A‐R increased cardiac contractility led us to hypothesize that the adverse effects of A₁‐R overexpression were due at least in part to a lack of balance in the endogenous stoichiometry of the A₁‐ and A_2A‐Rs. To test this hypothesis, we created double recombinants harboring both the A₁‐R and A_2A‐R_Hi transgenes (A₁/A_2A‐TG_Hi). As seen in Figure 3A , coexpression of the A₁‐R and A_2A‐R transgenes did not infuence the expression of either the A₁‐R or A_2A‐R receptors when compared with mice in which either the A₁‐R‐ or A_2A‐R transgenes were expressed alone. Also, immunofuorescence microscopy of isolated myocytes confirmed that both receptors colocalized to the sarcolemmal surface ( Figure 3B ). Importantly, while mice overexpressing A₁‐R alone developed profound cardiac dilatation, A_2A‐R coexpression prevented the profound cardiac dilatation in A₁‐TG mice ( Figure 3C ). Furthermore, as seen in Figure 4 , coexpression of the A₁‐R and A_2A‐R transgenes significantly improved cardiac hemodynamics when compared with mice overexpressing the A₁‐R. Indeed, fractional shortening, left ventricular end‐diastolic pressures, +dp/dt, and ‐dp/dt were similar in A₁/A_2A‐TG mice and wild‐type controls. However, heart rate remained depressed in the A₁/A_2A‐TG mice, albeit at a rate signifcantly higher than that seen in the A₁‐TG mice. Furthermore, as seen in Figure 5A , coexpression of the A₁‐R and A_2A‐R_Hi transgenes significantly improved survival compared to mice overexpressing A₁‐R alone. By contrast, coexpression of the A₁‐R with the transgenic line expressing lower levels of A_2A‐R (A_2A‐R TG_Lo) had no effect on the cardiac phenotype when compared with A₁‐TG mice ( Figure 5B and data not shown).

Figure 3.

Creation of transgenic mice overexpressing both A₁‐R and A_2A‐R. (A) A₁‐R and A_2A‐R expression in WT, A1‐TG, A_2A‐TG and A1/A_2A‐TG mice. Ventricular extracts from 8‐ week‐old male mice were probed with indicated antibodies. (B) A₁‐R and A_2A‐R colocalized on cardiac myocyte membrane. 60× confocal images are shown. (C) Horizontal sections of mouse hearts stained with hematoxylin‐eosin (HE). Minimum two animals (8‐week‐old male) of each genotype were stained and representative scaled photographic images (1×) are shown.

Figure 4.

A_2a‐R expression improved cardiac function and hemodynamics without affecting heart rate in A₁‐TG mice. (A) Percent fractional shortening of indicated mouse groups; (B) Heart rate of indicated mouse groups; (C) and (D) Hemodynamic functions of transgenic mice: maximum rate of ventricular pressure rise and decline (+ DP/DT and ‐DP/DT.) *p < 0.001 versus WT, †p < 0.001 versus A₁TG. (n= 15–21, SE, male 8–12‐week‐old mice).

Figure 5.

Animal survival, calcium handling, and gene expression in A₁TG and A_2a‐TG mice. (a) Kaplan‐Meier survival curve for A₁‐R transgenic lines coexpressing high levels of A_2a‐R (A_2a‐TGHi). A1 versus A₁/A_2a p < 0.001. (B) Kaplan‐Meier survival curve for A₁‐R transgenic lines coexpressing low levels of A_2a‐R (A_2a‐TGlo). A₁ versus A₁/A_2a NS. (C) Representative tracings of myocyte Ca²⁺ transients and contractions. Detailed calculations are shown in Table 2. (D) SERCA2 and Gαi2 expression. Ventricular extracts from 8‐week‐old male mice were probed with indicated antibodies. SERCA2 and Gαi2 signals were normalized to average intensity in WT hearts (n= 4–7) and were analyzed by one‐way analysis of variance followed by Dunnett's test. Analyzed values were mean +/‐ SE, *p < 0.05 versus WT.

Overexpression of A_2A‐R enhances myocyte [Ca²⁺]_i transport

In view of the marked increase in contractility in the A_2A‐TG_Hi mice and their ability to prevent heart failure in the A₁‐TG mice, we hypothesized that mice harboring the A_2A transgene might have enhanced Ca²⁺ handling. To test this hypothesis, we first compared Ca²⁺ handling in myocytes isolated from A_2A‐TG_Hi, A₁‐TG, and wild‐type nontransgenic littermates. As seen in Figure 5C and Table 2 , when compared to WT myocytes, constitutive overexpression of A_2A‐R resulted in significant higher systolic (p < 0.0001) but not diastolic [Ca²⁺]_i (p < 0.60). Maximal contraction amplitudes (p < 0.0001), maximal shortening (p < 0.0003), and relengthening velocities (p < 0.006) were all higher in myoctyes overexpressing A_2A‐R ( Table 2 ). Sarcoplasmic reticulum (SR) Ca²⁺ uptake activity was significantly faster (p < 0.0001).

Table 2.

Effects of adenosine receptor overexpression on [Ca2+]i transients, myocyte contraction, and calcium handling protein expression.

	WT	A2a‐TCHi	Ai‐TC	Ai‐TC A2a‐TCHi
Calcium transients [Ca2+]0 mM
Systolic [Ca2+]I, nM
0.6	167 ± 10	205 ± 13*	110 ± 9*	181 ± 14*$
	(20)	(28)	(20)	(12)
1.8	215 ± 10	295 ± 14*	172 ± 12*	271 ± 22*$
	(39)	(32)	(38)	(32)
5.0	336 ± 11	488 ± 18*	304 ± 30*	480 ± 80*$
	(41)	(35)	(15)	(19)
Diastolic [Ca2*], nM
0.6	118 ± 7	108 ± 6+	73 ± 9*	94 ± 13*
1.8	111 ± 7	116 ± 6+	95 ± 7*	86 ± 11*
5.0	131 ± 5	125 ± 5+	113 ± 11*	85 ± 17*
t1/2[Ca2+]| transient decline, ms
0.6	216 ± 13	175 ± 11*+	307 ± 32*	231 ± 22*$
1.8	174 ± 7	124 ± 7*+	243 ± 13*	143 ± 8*$
5.0	151 ± 5	100 ± 4*+	200 ± 8*	133 ± 8*$
Myocyte contraction [Ca2+]0
Maximal contraction amplitude,% cell length
1.8	6.80 ± 0.43	9.07 ± 0.42*+	4.99 ± 0.45*	7.85 ± 0.41 *$
	(56)	(38)	(36)	(36)
5.0	10.59 ± 0.44	12.79 ± 0.57*+	7.17 ± 0.82*	11.92 ± 0.35*$
	(38)	(49)	(11)	(22)
Maximal shortening velocity, ce1l length/s
1.8	1.05 ± 0.07	1.39 ± 0.06*+	0.63 ± 0.06*	1.00 ± 0.05$
5.0	1.58 ± 0.10	1.84 ± 0.09*+	0.90 ± 0.13*	1.53 ± 0.07$
Maximal relengthening velocity,cell length/s
1.8	0.92 ± 0.07	1.17 + 0.07*+	0.42 ± 0.04*	0.75 ± 0.05$
5.0	1.29 ± 0.10	1.49 ± 0.09*+	0.68 ± 0.13	1.17 ± 0.05$
Selected calcium handling protein expression
SERCA2	1.00 ± 0.07	1.42 ± 0.05*$	0.63 ± 0.11*	1.37 ± 0.14*$
	(7)	(4)	(7)	(4)
Na+ pump	1.00 ± 0.03	0.78 ± 0.09$	0.43 ± 0.11*	0.71 ± 0.04*$
	(4)	(4)	(4)	(4)
NCX1	1.00 ± 0.12	1.45 ± 0.15$	0.83 ± 0.19	1.32 ± 0.17$
	(4)	(4)	(4)	(4)

For calcium transient and contraction measurements, numbers in parentheses are myocytes pooled from 4–5 mice/genotype group. Mouse groups: wild‐type (WT), A_2a‐TG_Hi, A₁‐TG) and A₁/A_2a‐TG. Values are means ± SE. Two‐way analysis of variance was used for analysis. *p < 0.05, compared to WT; ^$p < 0.0001, A₁‐TG vs. A₁‐TG/A_2a‐TG_Hi; + p < 0.05, A_2a‐TG vs. A₁‐TG/A_2a‐TGHi. Protein expression was determined by immunoblotting. Numbers in parentheses are number of hearts from each mouse group. Signal band intensities on each blot were normalized to the average intensity of that protein measured in wild‐type hearts. One‐way analysis of variance followed by Dunnett's test was used to analyze the results. *p < 0.05, compared to WT; ^$A₁‐TG vs, A_2a‐TG_Hi or A₁‐TG vs. A₁/A_2a‐TG_Hi.

By contrast, myocytes in which A₁‐R was constitutively overexpressed had significantly lower systolic (p < 0.0005) and diastolic [Ca²⁺]_i (p < 0.0007) at all [Ca²⁺]_o examined ( Table 2 ). Altered [Ca²⁺]_i transients in myocytes with constitutively overexpressed A₁‐R resulted in decreased maximal contraction amplitudes as well as maximal shortening and relengthening velocities (p < 0.0001 for all three parameters; Table 2 ). Furthermore, the t_1/2 of [Ca²⁺]_i transient decline, an estimate of sarcoplasmic reticulum Ca²⁺ uptake, ²³ was significantly prolonged in A₁‐R TG myocytes when compared with WT (p < 0.0001) myocytes.

Co‐overexpression of A₁‐ and A_2A‐R ameliorates the abnormal [Ca²⁺]_i handling seen in the A₁‐R TG myocytes

When analyzing myocyte function in our dual A₁‐/A_2A‐R mice, we found that the overexpression of the A_2A‐R significantly ameliorated the contractile abnormalities observed in A₁‐R TG myocytes (p < 0.0001 for maximal contraction amplitude, maximal shortening, and relengthening velocities, Table 2 ). In addition, maximal contraction amplitude of A_1/A_2A–R TG myocytes was significantly higher (p < 0.015) than that observed in WT myocytes but lower (p < 0.05) than that measured in A_2A‐R overexpressed myocytes ( Table 2 ). With respect to [Ca²⁺]_i homeostasis, SR Ca²⁺ uptake activity (p < 0.0001) and systolic (p < 0.0001) but not diastolic [Ca²⁺]_i (p < 0.36) was improved by co‐overexpression of A₁ and A_2A‐R when compared with constitutive overexpression of A₁‐R alone ( Table 2 ). Diastolic [Ca²⁺]_i was significantly lower in A₁/A_2A myocytes when compared to WT (p < 0.0001) or A_2A‐R overexpressed (p < 0.0001) myocytes. Te t_1/2 of [Ca²⁺]_i transient decline was significantly shorter in A₁/A_2A myocytes when compared to WT myocytes (p < 0.009) but longer when compared to A_2A‐R overexpressed myocytes (p < 0.002).

Efects of enhanced A_2A‐R signaling on biochemical pathways in the heart

In order to understand the mechanisms responsible for the enhanced contractility effected by increased A_2A‐R signaling, we assessed the levels of proteins involved in Ca²⁺ homeostasis and G‐protein‐coupled receptor signaling in the heart. As seen in Figure 5D , the amount of SERCA2 was significantly increased in hearts from A_2A‐R TG_Hi mice when compared with wild‐type controls. The level of SERCA2 protein was also significantly elevated in A₁/A_2A‐TG_Hi mice but by contrast was significantly lower than WT controls in mice overexpressing the A₁‐R. Te levels of the Na⁺ pump and NCX1 in the A_2A‐TG mice had changing trends, but they were not statistically different from the WT mice. By contrast, mice overexpressing the A₁‐R had decreased Na⁺ pump protein levels. It is interesting to note that co‐overexpression of A_2A‐R in A₁‐TG mice enhanced the expression of all three proteins involved in Ca²⁺ homeostasis when compared to mice expressing only A ₁‐R ( Figure 5D and Table 2 ). However, the change in contractility seen in the A_2A‐R TG_Hi mice as well as in the A_2A‐R TG_Lo mice was not associated with an increase in steady‐state adenylyl cyclase activity (data not shown). By contrast, levels of the G‐inhibitory protein, Gαi‐2, as well as the levels of mRNA encoding Gαi‐2 were increased in all three experimental models (A_2a‐TG_Hi, A₁‐TG, a1/a_2a‐tg; Figure 5D and data not shown).

Long‐term effects of enhanced A_2A‐R signaling

Although overexpression of the A_2A‐R markedly enhanced cardiac contractility in mice up to 12 weeks of age, long‐term overexpression was not associated with an increase in cardiac contractility. Indeed, cardiac function was identical in 20‐week‐old transgenic and nontransgenic controls as seen by measurement of fractional shortening (42 ± 2.0% WT, n = 7 vs. 40.3 ± 3.4% A_2A‐TG_Hi, n = 13, p= NS). At 20 weeks, A_2A‐TG mice remained hypertrophied, but without dilation (data not shown). However, compared to younger mice, the 20‐week‐old A_2A‐TG hearts developed significant fibrosis ( Figure 6 ). Furthermore, the fibrosis progressively worsened at 30 weeks (data not shown). Also, SERCA2 levels were not elevated in 20‐week‐old A_2A‐TG_Hi mice (1.1 ± 0.1, n= 7) when compared with wild‐type controls (1.0 + 0.1,H = 13, p= NS). Similarly, while phosphorylation (i.e., activation) of Akt was signifcantly enhanced in young mice (12 weeks old) overexpressing the A_2A‐R transgene (1.03 ± 0.2 WT, n= 3 vs. 1.8 ± 0.16, A_2A‐TG_Hi, n = 6, p < 0.01), pAkt levels were not elevated in 20‐week‐old A_2A‐R TG_Hi mice (0.91 ± 0.12, n = 9) when compared with wild‐type controls (1.12 ± 0.15, n = 7, p= NS). However, levels of Gαi remained elevated in the hearts of the 20‐week‐old A_2A‐TG_Hi mice (1.9 ± 0.3 A_2A‐R TG_Hi, n= 13 vs. 1.0 ± 0.1 WT, n = 7, p < 0.01).

Figure 6.

Pico Sirius Red staining of 12‐week and 20‐week‐old A_2a‐TG_hi myocardium (collagen‐stained purple). Representative 10× objective images are shown. Two independent observers evaluated fbrosis based on 4–7 samples/group.

Discussion

In the present study, we have utilized cardiac‐restricted overexpression of the A_2A‐R to assess for the first time the effects of A_2A‐R‐signaling on cardiac morphology and function independent of the limitations imposed by experiments utilizing receptor subtype “selective” agonists and antagonists. Constitutive overexpression of the A_2A‐R in young mice resulted in super normal contractility, which was associated with a modest increase in heart rate and a small but significant increase in LV mass. In myocytes isolated from hearts overexpressing the A_2A‐R, systolic but not diastolic [Ca²⁺]_i was elevated and t^1/2 of [Ca²⁺]_i transient decline was much shorter when compared with wild‐type controls and myocyte size did not change. These observations are consistent with enhanced SR Ca²⁺ uptake, resulting in increased SR Ca²⁺ content, more Ca²⁺ available for release per beat, and larger systolic [Ca²⁺]_i values and twitch amplitudes in A_2A‐R overexpressed myocytes. In support of this mechanism of action is the observation that SERCA2 but not Na⁺‐K⁺‐ATPase or NCX1 expression was increased in A_2A‐R overexpressed myocytes. Thus, viewed in terms of excitation‐contraction coupling, the major alteration induced by A_2A‐R overexpression was increased SERCA2 expression and SR Ca²⁺ uptake activity.

These findings are disparate from previous studies evaluating the role of adenosine receptor subtype‐specific agonists and antagonists, which suggested that A_2A‐R‐mediated inotropy was accompanied by only a small increase in Ca²⁺ transients leading investigators to suggest that the A_2A‐R‐induced contractile effects were mediated by Ca²⁺‐independent inotropic mechanisms. ²⁴ Furthermore, these earlier studies also suggested that A_2A‐R activation could increase shortening and the rate of maximal shortening in isolated myocytes without an effect on maximal rate of relaxation ²⁵ while in isolated hearts, A_2A‐R activation could increase left ventricular pressure and the maximal rate of pressure development (+dP/dt/max) without infuencing cardiac relaxation. ¹⁴ By contrast, in this study we demonstrate enhanced contractility as well as enhanced relaxation, these effects being associated with increased expression of SERCA2 and robust changes in Ca²⁺ handling. Importantly, the changes in cardiac function after A_2A‐R overexpression were not due to an increase in heart rate as we were able to see enhanced contractility both invivo as well as in isolated and paced myocytes. The disparity between the present results and earlier studies may be due to the lack of “specifcity” of pharmacologic agonists and antagonists. However, we cannot exclude the possibility of species‐related differences as earlier studies were performed in rats.

Because of the contrasting effects of A₁‐ and A_2A‐R overexpression on cardiac function, we wondered whether abnormalities in calcium homeostasis might also be responsible for the adverse effects of A₁‐R overexpression. Indeed, left ventricular myocyte contractility was severely depressed in A₁‐TG myocytes. In addition, systolic [Ca²⁺]_i in A₁‐R myocytes was lower at all 3 [Ca²⁺]_o examined, suggesting that SR Ca²⁺ uptake was diminished. ²⁶ The decreased SERCA2 expression and prolonged t^1/2 of [Ca²⁺]_i transient decline in A₁‐TG myocytes are consistent with deffective SR Ca²⁺ uptake activity, resulting in decreased SR Ca²⁺ content and diminished twitch amplitudes. Diastolic [Ca²⁺]_i was also lower with A₁‐R overexpression, suggesting that accelerated Ca²⁺ effux may further contribute to reduced SR Ca²⁺ content and twitch amplitude. As one of the major alterations in excitation‐contraction coupling in A_2A‐TG myocytes was enhanced SERCA2 expression and SR Ca ²⁺ uptake, we hypothesized that enhanced A_2A‐R signaling might ameliorate the negative inotropic effects of enhanced A₁‐R signaling. Indeed, co‐overexpression of the A_2A‐R improved cardiac contractility, decreased end‐diastolic pressure, enhanced SERCA2 expression, and markedly improved survival when compared with mice overexpressing the A₁‐R. These salutary benefits were also seen at the single cell level as coexpression of the A₁‐R and A_2A‐R ameliorated the marked cellular abnormalities found in the A₁‐TG mice. Because systolic [Ca²⁺]_i and t^1/2 of [Ca²⁺]_i transient decline improved without a change in diastolic [Ca²⁺]_i in the A₁‐/A_2A‐TG mice when compared with A₁‐TG mice, it is likely that defects in SR Ca²⁺ uptake but not Ca²⁺ effux pathways induced by constitutive A₁‐R overexpression were influenced by co‐overexpression of A_2A‐R. However, it is also noteworthy that the ability of A_2A‐R signaling to ameliorate the adverse efects of A₁‐R overexpression was dose related as crossing the A₁‐TG with A_2A‐R TG_Lo mice had no effect on cardiac hemodynamics or outcomes despite an increase in contractility in the A_2A‐TG_Lo mice.

Despite the salutary benefits of short‐term overexpression of the A_2A‐R, long‐term benefits were not obvious as by 20 weeks of age, contractility decreases substantially to levels less similar to those seen in WT controls. Interestingly, this fnding was consistent with results seen after overexpression of other G‐protein‐coupled receptors. For example, overexpression of the β₂‐adrenergic receptor (AR) results in left ventricular dysfiunction in older mice while mice overexpressing the β₁‐AR develop signifcant LV dysfunction and early death. The markedly different phenotype in young mice overexpressing the β₂‐AR and those overexpressing the β₁‐AR has been attributed to the cardioprotective effects of enhanced Gαi signaling in the β₂‐AR overexpressing mice. ²⁷ The increase in Gαi we observed in mice overexpressing the A_2A‐R may have provided similar beneficial effects in young mice by limiting or opposing the effects of activation of Gas. However, it is interesting to note that in contrast with β‐AR overexpression, both mice overexpressing the A_2A‐R and mice overexpressing Gαs ²⁸ did not demonstrate an increase in steady‐state adenylyl cyclase activity. Thus, the enhanced inotropy secondary to increased A_2A‐R signaling in the mouse may be mediated through adenylyl cyclase dependent and independent pathways. Regardless, increased cardiac fbrosis in older A_2A‐R overexpressing mice suggested a potential mechanism for diminishing contractile benefits derived from A_2A‐R overexpression. Since A_2A‐R, through G_s coupling, increased cAMP‐dependent transsarcolemmal calcium fux and contraction, it is highly possible that sustained contractile overdrive in A_2A‐TG mice may ultimately lead to “calcium‐overload” leading to dysfunction and structural injury. However, similar to the β‐adrenergic receptor transgenic mice, ²⁷ other pathways may also participate in the late myopathy.

Another possible explanation for the finding that early activation of A_2A‐R signaling did not cause left ventricular dysfunction or pathology is the signifcant increase in Akt levels in young mice. Phosphorylation (activation) of Akt is known to have cardioprotective effects. ²⁹ Indeed, while phosphorylation of Akt was enhanced in young mice overexpressing the A_2A‐R, it was not increased in older A_2A‐TG mice with “normalization” of left ventricular function. These results were consistent with our recent fnding that Akt activity was substantially decreased in mice with left ventricular dysfunction secondary to overexpression of the A₁‐R. ⁹ Finally, our results suggest that the therapeutic usefulness of A_2A‐R overexpression might be predicated on “dose” as despite an inability to reverse left ventricular dysfunction caused by A₁‐R overexpression, low levels of A_2A‐R overexpression significantly increased cardiac contractility without causing an increase in either ventricular or atrial hypertrophy. Nonetheless, additional studies will be required to identify the role of low levels of A_2A‐R overexpression in preventing or rescuing left ventricular dysfunction in other experimental models. Furthermore, additional studies using controlled overexpression of the A_2A‐R receptor may shed additional light on the important downstream signaling pathways that play a role in the development of the early and late A_2A‐R phenotype.

Taken together, our results suggest that despite the fact that the A₁‐R and A_2A‐R signaling pathways affect different downstream events, the physiologic integrity of the heart, at least in a chronic sense, requires an ongoing balance between the activation of these two pathways. Indeed, the family of adenosine receptor subtypes is an anomaly because it is the only group of G‐protein‐coupled, seven‐transmembrane‐spanning receptors in which a single ligand can bind to multiple receptor subtypes in the same tissue and in so doing mediate opposing signaling pathways—that is, adenylylcyclase activation in the case of the A_2A‐R and adenylyl cyclase inhibition viathe A₁‐R. Thus, it would appear axiomatic that the chronic use of subtype selective adenosine receptor agonists or antagonists might have unexpected consequences. These fndings should be considered when designing adenosine agonists and/or antagonists for chronic use in patients with cardiovascular diseases.