β-Adrenergic Receptor-PI3K Signaling Crosstalk in Mouse Heart: Elucidation of Immediate Downstream Signaling Cascades

Sustained β-adrenergic receptors (βAR) activation leads to cardiac hypertrophy and prevents left ventricular (LV) atrophy during LV unloading. The immediate signaling pathways downstream from βAR stimulation, however, have not been well investigated. The current study was to examine the early cardiac signaling mechanism(s) following βAR stimulation. In adult C57BL/6 mice, acute βAR stimulation induced significant increases in PI3K activity and activation of Akt and ERK1/2 in the heart, but not in lungs or livers. In contrast, the same treatment did not elicit these changes in β1/β2AR double knockout mice. We further showed the specificity of β2AR in this crosstalk as treatment with formoterol, a β2AR-selective agonist, but not dobutamine, a predominantly β1AR agonist, activated cardiac Akt and ERK1/2. Acute βAR stimulation also significantly increased the phosphorylation of mTOR (the mammalian target of rapamycin), P70S6K, ribosomal protein S6, GSK-3α/β (glycogen synthase kinase-3α/β), and FOXO1/3a (the forkhead box family of transcription factors 1 and 3a). Moreover, acute βAR stimulation time-dependently decreased the mRNA levels of the muscle-specific E3 ligases atrogin-1 and muscle ring finger protein-1 (MuRF1) in mouse heart. Our results indicate that acute βAR stimulation in vivo affects multiple cardiac signaling cascades, including the PI3K signaling pathway, ERK1/2, atrogin-1 and MuRF1. These data 1) provide convincing evidence for the crosstalk between βAR and PI3K signaling pathways; 2) confirm the β2AR specificity in this crosstalk in vivo; and 3) identify novel signaling factors involved in cardiac hypertrophy and LV unloading. Understanding of the intricate interplay between β2AR activation and these signaling cascades should provide critical clues to the pathogenesis of cardiac hypertrophy and enable identification of targets for early clinical interaction of cardiac lesions.

Introduction b-adrenergic receptors (bAR) are members of the G proteincoupled receptor family [1]. Cardiomyocytes express all three bAR subtypes-b 1 AR, b 2 AR, and, in some species, b 3 AR [2,3,4]. Different subtypes of bAR have the capacity to activate multiple intracellular signaling pathways through different G proteins and second messengers, thereby mediating diverse heart functions under various conditions [5,6,7]. There is accumulating evidence that sustained bAR activation leads to cardiac hypertrophy, which represents an independent risk factor for cardiovascular morbidity and mortality [8]. Thus, bAR blockers have become an established treatment of chronic heart failure. On the other hand, several lines of evidence suggest that selective bAR stimulation may be beneficial in the setting of heart failure [9,10]. More recently, clinical and experimental studies have shown that bAR activation may prevent left ventricular (LV) atrophy during LV unloading [11,12]. Hence, understanding the signaling mechanisms under which bAR exert diverse roles in different cardiac conditions is likely to provide a better therapeutic approach towards treatment of heart failure. Our knowledge about the mechanisms by which bAR activation promotes myocardial growth was mostly acquired from in vitro systems or based on studies using animal models of chronic systemic bAR agonist administration [8]. The immediate signaling pathway(s) following acute bAR stimulation have not been well investigated.
PI3K signaling contributes to many important cellular processes, including regulation of cell cycle, growth, survival and migration [13]. Several genetic models with alterations of PI3K and other signaling molecules, including phosphatase and tensin homologue deleted from chromosome 10 (PTEN) and Akt, have demonstrated altered cardiac phenotypes [14,15,16,17]. Moreover, we have previously demonstrated that the activity of the PI3K signaling pathway during cardiac development is highly regulated with the highest levels found during the fetal-neonatal transition period and the lowest levels in the adult [18].
bAR stimulation has been shown to affect insulin signaling via PI3K/Akt in cardiomyocytes [19]. In cultured cardiomyocytes, bAR stimulation increases PI3K activity [20,21]. Both the b 1 AR and b 2 AR have been reported to transactivate PI3K in vitro [22,23]. Moreover, bAR stimulation-induced increases in heart weight, contractile abnormalities, and myocardial fibrosis, and cardiac ''fetal'' genes were markedly attenuated in PI3Kcknockout mice [24]. Our laboratory has shown that bAR-induced transactivation of PI3K signaling pathway plays important roles in the regulation of cell proliferation and cellular protection against apoptosis in neonatal cardiomyocytes [25,26]. Recently, we further provided in vitro evidence for such bAR-PI3K crosstalk and identified some critical signaling molecules involved [7]. The crosstalk between bAR and PI3K signaling pathway in vivo, however, has not been well studied, especially in a cardiac context. Moreover, the early signaling events following bAR stimulation preceding the onset of cardiac hypertrophy has not been demonstrated. A more detailed picture of these early signaling events would enable identification of targets for early clinical interaction for the bAR-related cardiac lesions.
To elucidate the cardiac signaling mechanisms responsible for the crosstalk between bAR and PI3K signaling pathway after acute bAR agonist administration, wild type adult male C57BL/6 and b 1 /b 2 AR double knockout (b 1 /b 2 DKO) mice were employed. We demonstrated that multiple signaling molecules downstream of PI3K are affected following acute bAR stimulation, including Akt, P70S6/mTOR/S6 axis, GSK3a/b and FOXOs. More important, two novel players, atrigin-1 and MuRF1, were shown to be part of the early signaling cascade.

Ethics Statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of the Lifespan (Assurance Number: A3922-01; Protocol Number: 002809). All efforts were made to minimize suffering.

Materials
Isoproterenol (ISO), a nonspecific bAR agonist, and dobutamine, a predominantly b 1 AR agonist [27], were obtained from Hospira Inc., Lake Forest, IL. Formoterol, a b 2 AR-selective agonist (displays a 330-fold selectivity for b 2 AR over b 1 AR, pK d values are 8.12 and 5.58 respectively), was obtained from Tocris Bioscience, Ellisville, MO. LY294002, a potent inhibitor of PI3K was purchased from Santa Cruz Biotechnology, inc. (sc-201426). All other chemicals and reagents were obtained from Sigma unless stated otherwise.

Experimental animals
Three months old male C57BL/6 wild-type mice (The Jackson Laboratory, Bar Harbor, ME) and b 1 /b 2 DKO mice (Strain Name: STOCK Adrb1 tm1Bkk Adrb2 tm1Bkk /J, Stock number: 003810, The Jackson Laboratory) were used. Animals were fed standard laboratory chow ad libitum. Mice were injected intraperitoneally (i.p.) with ISO (1.25 mg/kg), dobutamine (1.7 mg/kg, the same molar equivalent with ISO), formoterol (2.1 mg/kg, the same molar equivalent with ISO) or saline for 30 min (unless otherwise specified) before euthanized as previously described [7]. In some experiments, mice were pretreated with DMSO (5%), LY294002 (1.4 mg/kg, i.p.) or H-89 (20 mg/kg, i.p.), a PKA inhibitor, for 30 min before the ISO treatment described above. Whole hearts were harvested and LV was identified, dissected, flash frozen in liquid nitrogen, and stored at 280uC. Left ventricular tissue lysates were prepared as described [7].

Real Time Quantitative-PCR
Total RNA was extracted from ventricular tissue using TRIzol reagent (Invitrogen, Carlsbad, CA), and further purified using RNeasy MinElute Cleanup Kit (Qiagen, Hilden, Germany). RNA concentration and purity were determined by spectrophotometry (Thermo Scientific, Wilmington, DE). The reverse transcription was performed using the Superscript III (Invitrogen) and Oligo (dT) primers from total RNA (1 mg). Quantitative determination of muscle-specific E3 ligases atrogin-1 (atrogin-1) and muscle ring finger protein-1 (MuRF1) mRNA levels was performed by realtime quantitative-PCR using specific TaqMan probes and GAPDH as the endogenous control. The following primer/probe pairs were used: atrogin-1, forward, 59 [28]. Real Time Quantitative PCR was performed on a 7500 Fast RT-PCR System (Applied Biosystems) using the TaqMan Gene Expression Master Mix (Applied Biosystems). The mRNA level was determined with the comparative Ct method (2 -DDCt ). Each sample was analyzed in triplicate.

Statistical Analysis
All data were expressed as mean 6 S.E. based on data derived from multiple independent experiments. The intensity of bands from Western blots was scanned with densitometry and digitally analyzed using the Image J software (NIH). The statistical bAR-PI3K Crosstalk in Mouse Heart significance was tested by Student's t-test. A probability of p,0.05 was considered statistically significant.

Results
Acute bAR stimulation induces increases in PI3K activity and phosphorylation of Akt and ERK1/2 in adult mouse heart C57BL/6 mice were treated with saline or ISO for 30 min. Left ventricular tissue lysates were subject to PI3K assay and Western blotting. Acute ISO treatment increased phosphotyrosine-associated PI3K activity, Akt phosphorylation (Thr308 and Ser473), and ERK1/2 phosphorylation in adult mice (Fig. 1A,B,C). Pretreatment with LY294002, a potent PI3K inhibitor, abolished bAR stimulation-induced activation of PI3K and Akt, confirming the direct involvement of PI3K (Fig. 1D,E). In a subset of experiments, mice were treated with ISO for various time periods shorter (15 min) or longer than 30 min (1, 2, 4, or 6 hr). At all time points, ISO treatment significantly increased Akt phosphorylation (Fig. 2). The biggest increase, however, was found within the first hour following ISO treatment and tapered off with time. The 30-mim ISO treatment time point was then used for the experiments described below to examine the acute effect of bAR stimulation. We next further established the role of bAR in the transactivation of PI3K/Akt. The effects of ISO on PI3K/Akt were completely abrogated when the same treatment was performed on age-and gender-matched b 1 /b 2 DKO mice ( Fig. 3A/B). Acute bAR stimulation induced increase in ERK1/2 phosphorylation was also abrogated in b 1 /b 2 DKO mice (Fig. 3C). To determine if the crosstalk is cardiac-specific, Western blotting was also performed  using lung, kidney and liver tissue lysates from control or ISOtreated mice. Acute bAR stimulation did not increase Akt phosphorylation in the lung, kidney (Fig. 4A,B) or liver (data not shown). Interestingly, acute bAR stimulation significantly increased the phosphorylation of ERK1/2 in the kidney but not in the lung (Fig. 4A,C). It is well established that bAR stimulation can induce cellular effects via the activation of G proteins/cAMP/ PKA signaling pathway. It is not known, however, if the cAMP/ PKA cascade is required for the crosstalk between acute bAR stimulation and PI3K in vivo. Our previous study based on in vitro model suggests that cAMP/PKA pathway may not be needed for such crosstalk [7]. To investigate this, mice were pretreated with H-89, a PKA inhibitor, for 30 min before injection with ISO for 30 min. As expected, ISO treatment significantly increased cardiac PI3K activity and phosphorylation of Akt and FOX-O1(Thr24) which were not affected by pretreatment with H-89 (Fig. 5).This suggests that the cAMP/PKA signaling pathway is not directly involved with bAR stimulation-mediated activation of PI3K signaling pathway. Taken together, these data provide solid evidence for a crosstalk between bAR and PI3K signaling pathway in adult mouse heart.
Acute bAR stimulation-induced increase in phosphorylation of Akt and ERK1/2 is b 2 AR-specific As mentioned earlier, both b 1 AR and b 2 AR are expressed in the heart. To assess which subtype of bAR is responsible for ISOinduced increase in phosphorylation of Akt and ERK1/2, mice were treated with saline, dobutamine (a predominantly b 1 AR agonist) or formoterol (a b 2 AR-selective agonist) for 30 min. Acute b 2 AR stimulation with formoterol induced significant increases in PI3K activity (Fig. 6A) and phosphorylation of both Akt (Thr308, Ser473) and ERK1/2 (Fig. 6). In contrast, acute b 1 AR stimulation with dobutamine had no effect (Fig. 6B). In a subset of experiments, we further examined changes in phosphorylation in other downstream signaling factors. Consistent with results shown in Fig. 6, acute treatment with formoterol, but not dobutamine, increased phosphorylation of P70S6 (Thr389 and T421/S424), S6 (S235/236 and S240/244), GSK3a (Ser21), GSK3b (Ser9) and FOXO3a (S318/321) (Fig. 7). These results suggest that the effects of acute bAR stimulation on PI3K signaling pathway and ERK1/2 in adult mouse heart are b 2 ARspecific.

Acute bAR stimulation increases phosphorylation of other signaling factors downstream of PI3K
We next examined other signaling molecules downstream of PI3K following the same ISO treatment. The mammalian target of rapamycin (mTOR) is an important downstream effector of PI3K/Akt pathway [29]. Following acute ISO treatment, phosphorylation of mTOR (Ser2448) was markedly increased (Fig. 8A). Phosphorylation of P70S6K (ribosomal protein S6 kinase 1), immediately downstream of mTOR [30], was significantly increased at both the Thr389 and Thr421/Ser424 sites (Fig. 8B). Furthermore, phosphorylation of ribosomal protein S6 (Ser235/ 236, S240/244), the effector of p70S6K, was also significantly increased after acute bAR stimulation in mouse heart (Fig. 8C).
Glycogen synthase kinase-3 (GSK-3) is another critical downstream element of the PI3K/Akt signaling pathway. These kinases are key signaling mediators in myocardial hypertrophy and progressive heart failure [31]. As shown in Fig. 9, acute bAR stimulation induced a significant increase in the phosphorylation of both GSK-3a (Ser21) and GSK-3b (Ser9).
The forkhead box family of transcription factors (FOXOs) is implicated in the regulation of a variety of cellular processes, including the cell cycle, apoptosis, DNA repair, stress resistance, and metabolism. FOXOs proteins are negatively regulated by the PI3K/Akt signaling pathway [32]. We have demonstrated, for the first time, that acute bAR stimulation induced significant increases in the phosphorylation of both FOXO1 (Thr24) and FOXO3a (Thr32, Ser318/321, Ser253) in vivo (Fig. 10). Taken together, these data indicate that acute bAR stimulation-induced signaling cascade involves multiple factors downstream of Akt, including mTOR/P70S6K/S6, GSK-3a/b and FOXO1/3a.
Acute bAR stimulation reduces the mRNA levels of atrgin-1 and MuRF1 It has been shown that atrogin-1 and MuRF1 are associated with hypertrophy and atrophy processes in skeletal muscle through the PI3K/Akt signaling pathway [33,34]. To investigate whether cardiac atrogin-1 and MuRF1 are affected by acute bAR stimulation, we examined their mRNA levels by real-time quantitative PCR. As shown in Fig. 11A, bAR stimulation induced a time-dependent reduction of both atrogin-1 and MuRF1 mRNA levels. The trend in reduction in the mRNA levels can be seen as early as one hour following ISO treatment. To confirm these results, mice were treated with saline or ISO for 4 hours. A similar reduction (50%,60%) in the mRNA levels of both atrogin-1 and MuRF1 was seen in the ISO-treated mice ( Fig. 11B/C). This is the first demonstration in mouse heart that bAR stimulation reduces the mRNA levels of atrogin-1 and MuRF1.

Discussion
In the present study, we set out to elucidate the early signaling events that mediate the effect of acute bAR stimulation in mouse heart. Our results provide convincing in vivo evidence for the crosstalk between bAR and PI3K signaling pathway. We demonstrate that the bAR transactivation of PI3K/Akt pathway is b 2 AR-specific and confines to the heart. Many known signaling factors downstream of PI3K, including the mTOR/P70S6K/S6 axis, GSK-3a/b, and FOXO1/FOXO3a, were shown to be involved. Although these cascades have been well investigated, the link between bAR and these signaling pathways has not been systematically characterized in vivo, especially in the cardiac context. More important, we demonstrate, for the first time, that  The levels and activity of the cardiac PI3K signaling pathway are tightly controlled during development. The highest activity is found in late gestation and early postnatal life and then decreases dramatically in older animals [18]. We have in vitro and in vivo evidence showing that bAR stimulation induces activation of cardiac PI3Ka but not the other isoforms of PI3K [7]. However, the responsible signaling mediators downstream of PI3K after acute bAR stimulation have not been well established. The results of the current study are the first systematical identification of multiple important signaling factors for such bAR-PI3K crosstalk.
Akt is a serine/threonine protein kinase downstream of PI3K. Full activation of Akt requires phosphorylation at both Thr308 [by 3-phosphoinositide-dependent protein kinase 1 (PDPK1)] and Ser473 (by a putative PDPK2) [35,36]. In mouse heart, acute bAR stimulation increases the phosphorylation of Akt at both the Thr308 and Ser473 sites (Fig. 1B). ERK1/2 is another important signaling effector involved in development of bAR-mediated cardiac hypertrophy [37,38]. Following acute bAR stimulation, ERK1/2 is also activated (Fig. 1C). It is likely, that the signaling cascades responsible for bAR-induced cardiac hypertrophy are not confined to PI3K and ERK1/2. Moreover, it is possible that there are different mechanisms underlying induction of hypertrophy, depending on the pathological stimuli [39]. b 1 /b 2 DKO mice are viable, fertile, normal in size and do not display any gross physical or behavioral abnormalities [40]. It has been shown that the impact of b 1 /b 2 AR loss on basal physiological functions such as heart rate, blood pressure, or metabolic rate is remarkably modest. Hence, b 1 /b 2 DKO mice represent a useful model for the study of bAR-modulated cardiac functions in vivo [40]. We showed that transactivation of PI3K and the subsequent activation of Akt and ERK1/2 following acute bAR stimulation  were completely abrogated in b 1 /b 2 DKO mice, providing further evidence for the cardiac bAR-PI3K crosstalk in vivo.
There is increasing evidence to support a differential role of b 1 AR and b 2 AR in terms of regulation of cardiac function and alterations in cardiac structure [6,26,41,42,43]. For example, b 1 AR has been shown to mediate pro-apoptotic responses while b 2 AR mediates anti-apoptotic responses in cardiomyocytes [26,43]. b 2 AR signaling has a lower overall effect on contractility than b 1 AR signaling does [44]. Coupled to Ga i , b 2 AR signaling has been demonstrated to induce the recruitment and activation of other signaling molecules, including PI3K/Akt and ERK [5,22,44,45]. We have previously shown bAR stimulation-induced activation of PI3K signaling pathway is b 2 AR-specific in H9c2 cardiomyoblasts, a b 2 AR predominant cell line [7]. In comparison, b 1 AR comprises approximately 70-80% of the total bAR population in healthy heart [42]. Despite these differences of bAR subtype abundance in H9c2 cells and in mouse heart, acute bAR stimulation-induced activation of PI3K signaling pathway appears to be b 2 AR-specific. It is unknown, however, whether or not chronic bAR stimulation-induced cardiac hypertrophy is solely mediated by b 2 AR. It has been reported in cultured rat neonatal cardiomyocytes that bAR-induced hypertrophy is mediated primarily by the b 1 AR subtype [46]. This discrepancy may reflect the fundamental differences between in vitro and in vivo systems. The in vivo data shown here are novel and add new insights to our understanding regarding the characteristics among bAR subtypes.
This crosstalk may be tissue-specific since the same bAR stimulation did not activate Akt in the lungs, kidneys (Fig. 4) or liver. Acute bAR stimulation, however, did activate ERK1/2 in  Activation of Akt and mTOR plays a crucial part in the regulation of cell growth and proliferation by monitoring nutrient availability, cellular energy levels, oxygen levels and mitogenic signals [29,47,48]. One of the most extensively characterized downstream targets of mTOR is P70S6K, which is pivotal to the regulation of protein synthesis, via the phosphorylation and activation of its major effector, ribosomal protein S6 [29,30]. In the present study, we demonstrated that acute bAR stimulation in adult mouse heart increased the phosphorylation of the whole mTOR/P70S6K/S6 signaling axis at multiple sites. Considering the important roles of growth promotion by the mTOR/P70S6K/ S6 signaling axis, it is plausible to reason that chronic bAR activation may lead to cardiac hypertrophy via this signaling pathway.
Akt-mediated phosphorylation of GSK-3 is inhibitory [49]. It is believed that GSK-3 inhibition induced-cardiac hypertrophy is independent of the mTOR pathway [31,50]. It has been shown that GSK-3b inactivation occurs in response to bAR agonists [20].
More recent studies also indicate that expression of inactivationresistant GSK-3a/b isoforms avert the pathological hypertrophy caused by chronic ISO stress [51]. In isolated C57BL/6 mice hearts, acute ISO exposure resulted in significant GSK-3b, but not GSK-3a, phosphorylation. In most previous studies, particular attention has focused on GSK-3b because it is regarded as the predominant isoform, but increasingly GSK-3a has been considered to play equally important roles. For example, it has been shown that up-regulation of GSK-3a inhibited cardiac growth and pressure overload-induced cardiac hypertrophy which is mediated through inhibition of ERK [52]. Our data provide strong evidence that acute bAR stimulation induces phosphorylation (inhibition) of both GSK-3a and GSK-3b in vivo.
Phosphorylation of FOXOs by Akt is inhibitory and leads to nuclear exclusion and the inhibition of the forkhead transcriptional program [53]. Three members of the FOXOs, FOXO1, FOXO3a, and FOXO4, have been implicated in regulating diverse cellular functions including differentiation, metabolism, proliferation, and survival [32,54]. In cardiomyocytes, it has been established that FOXO1 and FOXO3a regulates apoptotic responses [55] and cell size [56], respectively. Recently, FOXOs have also been shown to promote atrophy in skeletal muscle [34,57]. It is, however, not known if the regulation and function of FOXOs in the heart are under the control of bAR. Our study provides evidence, for the first time, that acute bAR stimulation  It has been extensively investigated that atrogin-1 and MuRF1 are common mediators in several models of skeletal muscle hypertrophy and atrophy. The levels of atrogin-1 and MuRF1 are regulated by the growth factor/Akt signaling axis through direct transcriptional regulation by FOXOs [33,34,57]. b 2 AR-agonists has also been shown to increase skeletal muscle strength in normal volunteers [58] and in a small number of patients [59]. In a cardiac context, atrogin-1 and MuRF1 have been demonstrated to inhibit cardiac hypertrophy [56,60,61]. It was also shown that atrogin-1 and MuRF1 regulate cardiac myosin-binding protein C which is related to familial hypertrophic cardiomyopathy [28]. Moreover, recent clinical and experimental studies have suggested that bAR activation may prevent LV atrophy during LV unloading [11,12], with the mechanism unknown. However, it has not been reported whether or not atrogin-1 and MuRF1 are involved in the bAR stimulation-induced cardiac hypertrophy and/or the anti-atrophic effect. Our finding is the first to demonstrate the link from bAR to atrogin-1 and MuRF1 in the heart, presumably via PI3K/Akt/FOXOs. Further study of this crosstalk should yield useful insights that may have important clinical implications.
In summary, our data suggest that crosstalk between bAR and PI3K signaling pathway activates multiple downstream signaling effectors and cascades in vivo. Our data support a model depicted in Fig. 12, where acute stimulation of the b 2 AR activates PI3K, which affects the activities of ERK1/2, Akt and its downstream signaling cascades, including mTOR/P70S6K/S6 axis, GSK-3a/ b, FOXOs, atrogin-1 and MuRF1. These data provide mechanistic and signaling insights for the pathophysiological consequences induced by chronic bAR stimulation. Figure 12. A hypothetical signaling cascades for the crosstalk between b2AR and PI3K/Akt pathway in vivo. Activation of ERK1/ 2-related signaling and the mTOR/P70S6K/S6 axis favors the progression of cardiac hypertrophy. Akt-induced phosphyrylation of GSK-3a/b inhibits its activity, thereby nullifying its anti-hypertrophic effect. Aktinduced phosphorylation of FOXOs also inhibits ist activity, negating its effects on activation of atrogin-1 and MuRF1, which results in decreased protein breakdown (atrophy). These multiple signaling scenarios likely contribute to the pathogenesis of cardiac hypertrophy and/or the antiatrophic effect following sustained bAR stimulation. ''R'' represents activation, ''x'' represents inhibition. Solid lines represent data supplied by our study, and dotted lines represent hyperthetical schematics. doi:10.1371/journal.pone.0026581.g012