Journal of the American College of Cardiology Vol. 45, No. 11, by the American College of Cardiology Foundation ISSN /05/$30.

Transcription

1 Journal of the American College of Cardiology Vol. 45, No. 11, by the American College of Cardiology Foundation ISSN /05/$30.00 Published by Elsevier Inc. doi: /j.jacc Targeted Inhibition of -Adrenergic Receptor Kinase-1 Associated Phosphoinositide-3 Kinase Activity Preserves -Adrenergic Receptor Signaling and Prolongs Survival in Heart Failure Induced by Calsequestrin Overexpression Cinzia Perrino, MD,* Sathyamangla V. Naga Prasad, PHD,* Mrinali Patel*, Matthew J. Wolf, MD, PHD,* Howard A. Rockman, MD* Durham, North Carolina OBJECTIVES BACKGROUND METHODS RESULTS CONCLUSIONS Desensitization and down-regulation of -adrenergic receptors ( ARs) are prominent features of heart failure largely mediated by increased levels of AR kinase-1 ( ARK1). -adrenergic receptor kinase 1 interacts with phosphoinositide-3 kinase (PI3K), and upon agonist stimulation, the ARK1/PI3K complex is recruited to agonist-stimulated ARs. Here we tested the hypothesis that in vivo selective inhibition of ARK1-associated PI3K activity would preserve AR signaling and, therefore, improve cardiac function and survival in experimental heart failure. We used a murine model of heart failure induced by calsequestrin (CSQ) cardiac-specific overexpression; CSQ mice were crossed with mice overexpressing in the heart a catalytically inactive PI3K (PI3K inact ) to competitively displace endogenous PI3K from ARK1. Catalytically inactive PI3K PI3K overexpression in CSQ mice inhibited ARK1-associated PI3K activity, normalized AR levels, and preserved AR responsiveness to isoproterenol (ISO). Restoration of AR signaling via PI3K inact overexpression resulted in marked improvement of cardiac function and a significant prolongation of survival. Importantly, the effects of PI3K inact overexpression were restricted to AR signaling, because cellular PI3K signaling was unaltered, as shown by the similar activation of multiple downstream signaling pathways in both CSQ and CSQ/PI3K inact mice. These data in the CSQ model of cardiac dysfunction indicate that membrane-targeted PI3K activity plays a detrimental role in heart failure, and its inhibition represents a novel therapeutic approach to ameliorate cardiac dysfunction and improve survival. (J Am Coll Cardiol 2005;45: ) 2005 by the American College of Cardiology Foundation Heart failure is a complex disorder whose prognosis remains poor, despite continuous advances in therapy (1). Clarification of the signaling pathways that promote deterioration of cardiac function is an intense area of research because inhibition of these deleterious signals could serve as a novel therapeutic approach. Failing human hearts are characterized by extensive abnormalities of the -adrenergic receptor ( AR) system, including down-regulation and desensitization of ARs resulting in a state of reduced responsiveness to agonist stimulation (2). Interestingly, whether changes in AR signaling represent an adaptive and protective process, as some postulate (3), or whether AR dysregulation is actually detrimental (4) is still controversial. Results from our previous studies suggest that chronic AR dysfunction in the failing heart is maladaptive and contributes to the deterioration in cardiac function (5 7). Previous studies have shown that agonist-induced AR dysfunction begins with receptor phosphorylation, followed From the Departments of *Medicine, Cell Biology and Molecular Genetics, Duke University Medical Center, Durham, North Carolina. This work was supported by a National Institutes of Health grant to Dr. Rockman (HL-61558). Manuscript received December 13, 2004; revised manuscript received January 28, 2005, accepted February 14, by rapid uncoupling of the receptor from its cognate G protein (desensitization), and subsequent targeting of the phosphorylated receptor for endocytosis, a process linked to chronic receptor down-regulation (4). Phosphorylation can be mediated by second messenger kinases (for example protein kinase A or protein kinase C), or by a specialized family of G protein-coupled receptor kinases (GRKs) (7,8); GRK2, known commonly as AR kinase-1 ( ARK1), is the most abundant isoform expressed in the heart (9,10). Because ARK1 levels are consistently and markedly elevated in both human (11) and experimental heart failure (5,7,12), up-regulation of ARK1 has been postulated to play an important role in the reduced AR responsiveness associated with cardiac dysfunction. Upon agonist stimulation, binding of cytosolic ARK1 to liberated G subunits facilitates its translocation to the plasma membrane, resulting in the phosphorylation of agonistoccupied receptors (13). We have recently demonstrated that ARK1 forms a cytosolic complex with phosphoinositide-3 kinase (PI3K) (14). Upon agonist stimulation, the ARK1/PI3K complex is recruited to activated ARs (15), wherein generation of D-3 phosphoinositides by PI3K is required for receptor endocytosis (14). In vivo, inhibition of receptor-localized PI3K activity prevents AR abnormalities

2 JACC Vol. 45, No. 11, 2005 June 7, 2005: Perrino et al. Targeted Inhibition of PI3K in Heart Failure 1863 Abbreviations and Acronyms AR -adrenergic receptor ARK1 AR kinase-1 camp cyclic adenosine monophosphate CSQ calsequestrin GSK glycogen synthase kinase 3 ISO isoproterenol MAPKs mitogen-activated protein kinases PKB protein kinase B PI3K phosphoinositide-3 kinase PI3K inact catalytically inactive PI3K WT wild type %FS percent fractional shortening after catecholamine stimulation or pressure overload, resulting in less deterioration in cardiac function (7). However, whether targeted inhibition of PI3K can rescue AR abnormalities under conditions of chronic heart failure, and ultimately improve cardiac function and survival, is not known. To test this, we used a murine model of heart failure induced by cardiac-specific overexpression of the calcium-binding protein calsequestrin (CSQ) (16). This model recapitulates human heart failure in many aspects, including the morphological progression from cardiac hypertrophy to dilated cardiomyopathy (16), the high rate of premature death (6), and the sensitivity to beta-blocker therapy (6). In order to inhibit AR-localized PI3K activity in failing hearts, we bred the CSQ mice with transgenic mice overexpressing a catalytically inactive PI3K (PI3K inact ) that competitively displaces the endogenous enzyme from ARK1, therefore preventing its recruitment to agonist-stimulated ARs. We determined the effects of PI3K inact overexpression on AR function and PI3K signaling, and monitored the progression of cardiac dysfunction and the rates of survival of single and binary transgenic mice. METHODS Experimental animals. Mice overexpressing CSQ or PI3K inact were engineered as previously described (7,16). Briefly, the alpha-myosin heavy chain promoter was used to drive cardiac-targeted overexpression of canine cardiac CSQ or a PI3K mutant lacking the adenosine triphosphatebinding domain (PI3K inact ). To ensure identical genetic backgrounds, both CSQ and PI3K inact strains were backcrossed 10 generations onto a dilute brown non-agouti genetic background. F1 pups were generated from the crossbreeding of CSQ transgenic mice with PI3K inact transgenic mice. Wild type (WT), CSQ transgenic, or CSQ/PI3K inact transgenic littermates of either gender were used for this study and were handled according to the approved protocols and the animal welfare regulations at Duke University Medical Center. Membrane fractionation, AR radioligand binding, and adenylyl cyclase activity. Membrane and cytosolic fractions from left ventricles flash-frozen in liquid N 2 were prepared as described previously (8). Because we have previously shown the K d of 125 I-CYP binding to the ARs in membranes prepared from mouse hearts is 30 pmol/l (17), we used a single saturating concentration of 125 I-CYP (250 pmol/l) to assess the total AR density in the mouse cardiac membranes. Receptor density (fmol) was normalized to milligrams of membrane protein. Adenylyl cyclase assays were performed as described previously (8), using 20 g of the membrane fraction. Generated cyclic adenosine monophosphate (camp) was quantified using a liquid scintillation counter (MINAXI-4000, Packard Instrument Co., PerkinElmer, Boston, Massachusetts). Immunoprecipitation and immunoblotting. Immunoprecipitation and immunoblotting were performed as described previously (7,8). Detection was carried out using ECL (Amersham Biosciences Corp., Piscataway, New Jersey), and bands were quantified using Bio-Rad (Hercules, California) Flouro-S Multimage software densitometry. PI3K. Phosphoinositide 3-kinase activity assays were carried out by immunoprecipitation of the PI3K and isoforms from the cytosolic fraction as described previously (15). ARK1-associated PI3K activity was measured after immunoprecipitation of 400 g of proteins from the membrane fraction with polyclonal antibody directed against ARK1 (Santa Cruz Biotechnology Inc., Santa Cruz, California). Mitogen-activated protein kinase activity. Mitogenactivated protein kinase activities were assessed from left ventricular cytosolic extracts as the capacity of immunoprecipitated extracellular signal-regulated kinase (ERK)2- p42/erk1-p44, c-jun N-terminal kinase (JNK)1, p38, and p38-beta (Santa Cruz) to phosphorylate in vitro substrates (myelin basic protein and glutathione S transferase [GST]- proto-oncogene c-jun [cjun]) as previously described (18). RNA isolation and Northern blotting. RNA was isolated from left ventricles using the RNA-Bee-RNA isolation reagent (Tel-Test Temco Inc., Friendswood, Texas) according to the manufacturer s instructions; 10 g ofrna were size-fractionated by denaturing formaldehyde gel electrophoresis, transferred to nylon membrane by capillary action, cross-linked with ultraviolet light, and hybridized with a mouse 1 AR 32 P-labeled cdna probe. After hybridization, filters were washed under stringent conditions, and transcripts were detected by autoradiography. Transthoracic echocardiography. Serial echocardiography was performed on conscious mice with an HDI 5000 echocardiograph (Philips, Böblingen, Germany) at 8, 12, and 16 weeks of age, as previously described (19). Statistical analysis. Data are expressed as mean values SEM. Multigroup comparisons were performed using oneway analysis of variance (ANOVA) with Neuman-Keuls correction. Serial echocardiography results were analyzed by repeated-measures ANOVA. Survival was analyzed by Kaplan-Meier analysis. Hazard ratios for the presence of PI3K inact transgene and gender were derived from the Cox proportional hazard model using SAS software (SAS Insti-

3 1864 Perrino et al. JACC Vol. 45, No. 11, 2005 Targeted Inhibition of PI3K in Heart Failure June 7, 2005: Figure 1. Catalytically inactive phosphoinositide-3 kinase (PI3K inact ) overexpression normalizes AR kinase-1 ( ARK1)-associated PI3K activity in failing hearts from calsequestrin (CSQ) transgenic mice. (a) ARK1- associated PI3K activity in cardiac membranes from wild type (WT) (n 8), CSQ (n 8), and CSQ/PI3K inact (n 8) mouse hearts. *p CSQ vs. WT or CSQ/PI3K inact (analysis of variance with Neuman-Keuls correction); PI3K (b) or (c) activities were assayed from the cytosolic extracts of the same hearts. Left panels show representative PI3K assays; right panels show summary data. There was no significant increase in the activity of PI3K over WT levels in both CSQ or CSQ/PI3K inact (in CSQ fold, in CSQ/PI3K inact fold compared with WT). Ori origin; PIP phosphatidylinositol-mono-phosphate; PIP2 phosphatidylinositol-bis-phosphate. tute, Cary, North Carolina). For all analyses, p 0.05 was considered significant. RESULTS Competitive displacement of endogenous PI3K from ARK1 preserves AR signaling in heart failure. To test the hypothesis that cardiac inhibition of receptortargeted PI3K activity would ameliorate AR signaling and, in turn, alter the progression of heart failure, we compared the molecular phenotype of single CSQ transgenic mice and binary CSQ/PI3K inact mice (7). In order to assess receptor-localized PI3K activity in failing hearts, ARK1 was immunoprecipitated and assayed for associated PI3K activity in the cardiac membranes of 12-week-old mice; CSQ hearts displayed a marked increase of ARK1-associated PI3K activity compared with WT controls (Fig. 1A, fold induction in CSQ over WT, *p CSQ vs. WT, ANOVA with Neuman-Keuls correction), consistent with our previous studies showing increased ARK1 levels (7,20). Importantly, PI3K inact overexpression completely displaced endogenous PI3K from ARK1, reducing ARK1 associated PI3K activity in CSQ/PI3K inact mice to control levels (Fig. 1A, *p CSQ vs. CSQ/ PI3K inact ). Because PI3K and PI3K are known to play important roles in cardiac growth and function (7,21,22) and to interact with ARK1 (7), we sought to determine the relative activities of the PI3K isoforms under conditions of heart failure. PI3K and PI3K were immunoprecipitated from the cytosolic fraction of the same hearts and assayed for PI3K activity. Compared with WT controls, failing CSQ hearts were characterized by a selective significant increase in the activity of PI3K (Fig. 1B, *p CSQ vs. WT), which was markedly reduced by the overexpression of the PI3K inact transgene (in CSQ/PI3K inact fold compared with WT, *p CSQ vs. CSQ/PI3K inact, ANOVA with Neuman-Keuls correction). In contrast, no significant changes in PI3K activity were detected among the different groups (Fig. 1C). We next tested whether targeted inhibition of PI3K activity within the receptor complex would correct abnormalities in AR signaling in CSQ mice by directly measuring AR density and adenylyl cyclase activity in membrane fractions from WT, CSQ, and CSQ/ PI3K inact hearts; CSQ overexpressing hearts were characterized by a significant 25% reduction in AR density and depressed isoproterenol (ISO)-stimulated cyclase activity (Table 1). Interestingly, inhibition of ARK1 associated PI3K activity in CSQ/PI3K inact mice completely normalized AR density to WT levels (Table 1) without having any effect on the transcriptional regulation of ARs (Fig. 2A). Increased AR levels in CSQ/ PI3K inact mice were also accompanied by a significant improvement in AR responsiveness to agonist, as measured by enhanced response in membrane adenylyl cyclase activity and increased camp generation (Table 1); ISOstimulated adenylyl cyclase activity exhibited a nonsignificant lower trend in CSQ/PI3K inact -overexpressing mice compared to WT, despite the complete normalization of AR density. These results are likely attributed to the persistent increase in the membrane levels of protein G inhibitory (Gi) in both single CSQ and binary CSQ/ PI3K inact transgenic hearts (data not shown). Targeted inhibition of PI3K at the site of activated ARs does not affect downstream PI3K cellular signaling. Because PI3K plays an important role in regulating cellular signaling, we wanted to exclude the possibility that overexpression of PI3K inact altered the activity of downstream kinases such as the mitogen-activated protein kinases (MAPKs), protein kinase B (PKB), and glycogen synthase kinase 3 (GSK) in transgenic hearts. Compared with WT, we found a significant increase in the activity of all MAPKs analyzed (ERK, JNK, and p38/p38 ) in CSQ mouse hearts

4 JACC Vol. 45, No. 11, 2005 June 7, 2005: Perrino et al. Targeted Inhibition of PI3K in Heart Failure 1865 Table 1. AR Signaling in WT, CSQ, and CSQ/PI3K inact Mice AR Density fmol/mg protein (n 16) Adenylyl Cyclase Activity pmol/mg Protein/min of camp (n 12) Basal Iso 10 4 M WT CSQ * * CSQ/PI3K inact *p 0.05 CSQ vs. WT or CSQ/PI3K inact ; p 0.05 WT, CSQ, or CSQ/PI3K inact vs. respective basal (analysis of variance with Neuman-Keuls correction); sodium fluoride-induced camp generation: WT ; CSQ ; CSQ/PI3K inact (pmol/mg/min). AR -adrenergic receptor; camp cyclic adenosine monophosphate; CSQ calsequestrin; PI3K inact catalytically inactive phosphoinositide 3-kinase; WT wild type. (Figs. 2B to 2E). Overexpression of PI3K inact in CSQ mice did not affect the activity of any of the three major MAPKs pathways (Figs. 2B to 2E), and, importantly, it did not change the phosphorylation status of PKB or of GSK, known downstream targets of PI3K (Fig. 2F). Taken together, these studies show that overexpression of Figure 2. Inhibition of receptor-localized phosphoinositide 3-kinase (PI3K) activity does not affect downstream signaling pathways. (a) Northern blotting analysis showing -adrenergic receptor ( 1 AR) levels in 12-week-old wild type (WT), calsequestrin (CSQ), and CSQ/catalytically inactive PI3K (CSQ/PI3K inact ) hearts (upper panel); equal loading of the different RNA samples was confirmed by methylene blue staining of nylon membranes (bottom panel). (b to e) Mitogen-activated protein kinase activation was determined in WT, CSQ, and CSQ/PI3K inact hearts by the ability to in vitro phosphorylate myelin binding protein (MBP) or recombinant glutathione S transferase-proto-oncogene c-jun (GST-cJun). Representative kinase assays and relative densitometric evaluation of at least six independent experiments are shown for extracellular signal-related kinase (ERK) (b), c-jun N-terminal kinase (JNK) (c), p38 (d), and p38 (e). Western blotting was carried out to evaluate total protein levels of each kinase (b to d). *p 0.01 for CSQ or CSQ/PI3K inact versus WT (analysis of variance with Neuman-Keuls correction). (f) Western blotting analysis showing similar activation of protein kinase B (PKB) and glycogen synthase kinase (GSK) under basal conditions in single CSQ and binary CSQ/PI3K inact mice. IB immunoblotting; IP immunoprecipitation.

5 1866 Perrino et al. JACC Vol. 45, No. 11, 2005 Targeted Inhibition of PI3K in Heart Failure June 7, 2005: Table 2. Echocardiographic Measurements in WT, CSQ, and CSQ/PI3K inact Mice WT (12 Weeks) (n 20) CSQ (n 16) 8 Weeks 12 Weeks 16 Weeks CSQ/PI3K inact (n 17) CSQ (n 15) CSQ/PI3K inact (n 16) CSQ (n 13) CSQ/PI3K inact (n 15) Serial echocardiography LVEDD, mm * LVESD, mm * IVS, mm PW, mm FS, % * Mean Vcfc, circ/s * Heart rate, beats/min * *p WT vs. CSQ or CSQ/PI3K inact at all time points; p 0.05 CSQ or CSQ/PI3K inact vs. respective 8-week value; p 0.01 CSQ/PI3K inact vs. age-matched CSQ (repeated measures ANOVA). FS fractional shortening; IVS interventricular septum; LVEDD left ventricular end-diastolic diameter; LVESD LV end-systolic diameter; PW posterior wall; Vcfc ventricular circumferential shortening, corrected by heart rate. Other abbreviations as in Table 1. PI3K inact primarily functions to competitively displace endogenous PI3K from ARK1, thereby normalizing AR signaling in failing hearts without altering other known PI3K downstream pathways. Preservation of AR signaling by targeted inhibition of PI3K improves cardiac function and delays the progression of cardiomyopathy in CSQ transgenic mice. To evaluate the effects of improved AR signaling on the progression of cardiac dysfunction due to CSQ overexpression, serial echocardiography was performed on both groups of heart failure mice (CSQ and CSQ/PI3K inact )at8,12, and 16 weeks of age and on WT mice (Table 2, Fig. 3); CSQ hearts underwent a progressive enlargement of the left ventricle and deterioration of cardiac function over time, expressed by reduced fractional shortening (%FS) (Table 2, Fig. 3). Calsequestrin/PI3K inact mice showed a smaller left ventricular end-diastolic diameter compared with agematched CSQ mice (Fig. 3B, left panel), but exhibited a similar rate of enlargement of the left ventricular enddiastolic diameter over time (Fig. 3B, right panel). In contrast, PI3K inact overexpression in CSQ mice not only reduced left ventricular end-systolic diameter, but it also significantly reduced the rate of left ventricular end-systolic diameter enlargement over time in CSQ/PI3K inact mice (Fig. 3C, left and right panels). At all time points, CSQ/ PI3K inact mice had significantly improved %FS compared to CSQ mice (Fig. 3D, left panel), and, importantly, CSQ/PI3K inact mice had a slower progression of cardiac dysfunction over time (Fig. 3D, right panel). The echocardiographic results for cardiac size were independently confirmed in another set of animals that underwent dissection and measurement of left ventricular weight, heart weight, and body weigh at 12 weeks of age (Fig. 4). Overexpression of the CSQ transgene markedly increased left ventricular weight/body weight and heart weight/body weight ratios compared with WT levels (Fig. 4, *p WT vs. CSQ or CSQ/PI3K inact, ANOVA with Neuman-Keuls correction). Importantly, PI3K inact overexpression in CSQ mice significantly reduced left atrial weight as well as left ventricular weight/body weight and heart weight/body Figure 3. Cardiac specific overexpression of catalytically inactive phosphoinositide-3 kinase (PI3K inact ) improves cardiac function of calsequestrin (CSQ) transgenic mice. (a) Representative echocardiograms from CSQ or CSQ/PI3K inact mice at 8, 12, and 16 weeks of age. Absolute values of left ventricular end-diastolic diameter (LVEDD, b), left ventricular end-systolic diameter (LVESD, c), and % fractional shortening (%FS, d) in age-matched wild type (WT) (striped circle), CSQ (solid circle), and CSQ/PI3K inact (open circle) mice are shown in the left panels. Intheright panels percent change variation over time is also shown for CSQ and CSQ/PI3K inact mice. *p 0.05 CSQ or CSQ/PI3K inact vs. respective eight weeks measurement; p 0.05 CSQ/PI3K inact vs. CSQ age-matched mice (repeated measures analysis of variance); p WT vs. CSQ or CSQ/PI3K inact at all time points.

6 JACC Vol. 45, No. 11, 2005 June 7, 2005: Perrino et al. Targeted Inhibition of PI3K in Heart Failure 1867 indicating that the survival advantage of female CSQ/ PI3K inact mice is not related to differential regulation of the transgenes according to gender or age (Fig. 6). Figure 4. Targeted inhibition of phosphoinositide-3 kinase (PI3K) reduces cardiac chamber size. Bar graphs showing (a) body weight (BW), (b) left atrium (LA) weight, (c) left ventricular weight/body weight (LV/BW), (d) heart weight/body weight (H/BW) in wild type (WT) mice, single calsequestrin (CSQ) transgenic, and binary CSQ/catalytically inactive phosphoinositide 3-kinase (PI3K inact ) transgenic mice. *p WT vs. CSQ or CSQ/PI3K inact ; p 0.01 CSQ/PI3K inact vs. CSQ (analysis of variance with Neuman-Keuls correction). weight ratios compared with single CSQ mice (Fig. 4, p 0.01 CSQ/PI3K inact vs. CSQ). Selective inhibition of AR-targeted PI3K activity prolongs survival of heart failure in CSQ transgenic mice. To test whether PI3K inact -mediated improvement in cardiac function would translate to a benefit on lifespan, survival analysis was conducted in CSQ mice, binary transgenic CSQ/PI3K inact mice, and their WT littermates. As expected, cardiac overexpression of the CSQ gene resulted in premature death (Fig. 5A). However, PI3K inact overexpression markedly prolonged the lifespan of CSQ mice as shown by Kaplan-Meier analysis (Fig. 5A). Remarkably, inhibition of AR-targeted PI3K activity completely inhibited death events in CSQ mice of both genders within the first five months of life (Cox proportional hazard analysis censored at 150 days of survival: *p 0.05 for PI3K inact ;p 0.3 for gender). Interestingly, this beneficial effect was maintained for a longer period of time in CSQ/PI3K inact mice of female gender (Fig. 5B, Cox proportional hazard analysis: p 0.01 for PI3K inact and female gender). The gender-related difference in survival is likely not attributed to a greater effect of the transgene in female mice, because PI3K inact overexpression resulted at 12 weeks of age in similar improvements for both genders in AR signaling ( AR density, fmol/mg of protein: CSQ/ PI3K inact males ; CSQ/PI3K inact females ), ISO-stimulated adenylyl cyclase activity (pmol of camp/mg/min: CSQ/PI3K inact males ; CSQ/PI3K inact females ), and cardiac function (%FS: CSQ/PI3K inact males ; CSQ/PI3K inact females ). Moreover, no differences in the expression levels of the two transgenes were observed between males and females in either young (12-week-old) or older mice (age 160 to 252 days), DISCUSSION In the present study, we report that targeted inhibition of ARK1-associated PI3K activity preserves AR signaling and improves cardiac function and survival in the murine model of severe heart failure induced by CSQ overexpression. Importantly, the effects of targeted inhibition of PI3K activity are specific to AR signaling because MAPKs and other PI3K downstream pathways are unaffected. Our findings in this genetic model of heart failure are consistent with our previous reports in the mouse models of chronic catecholamine stimulation and pressure overload (7), suggesting that receptor-targeted inhibition of PI3K activity might represent a novel therapeutic approach in the treatment of the failing heart. Figure 5. Restoration of -adrenergic receptor function improves survival in mice with heart failure. (a) Kaplan-Meier survival curves of wild type (WT) (n 18), calsequestrin (CSQ) (n 19), and CSQ/catalytically inactive phosphoinositide 3-kinase (PI3K inact )(n 20) mice; CSQ mice had a mean survival age of days , whereas CSQ/ PI3K inact mice reached a mean survival age of days. *p for WT vs. CSQ or CSQ/PI3K inact ; p CSQ/PI3K inact vs. CSQ, Kaplan-Meier analysis. (b) Kaplan-Meier survival curves according to gender in CSQ (females, n 11; males, n 8) and CSQ/PI3K inact mice (females, n 11; males, n 9). In the analysis censored at 150 days, *p 0.05 for PI3K inact,p 0.3 for gender; in the analysis censored at 300 days, p 0.01 for PI3K inact and female gender (Cox proportional hazard analysis).

7 1868 Perrino et al. JACC Vol. 45, No. 11, 2005 Targeted Inhibition of PI3K in Heart Failure June 7, 2005: Figure 6. Equal expression levels of calsequestrin (CSQ) and catalytically inactive phosphoinositide 3-kinase (PI3K inact ) transgenes in young and old male and female mice. Representative immunoblottings (two animals/group) and relative quantitative densitometric analysis (three to four animals/group) showing similar expression levels of the alpha-myosin-heavy-chain-driven transgenes CSQ and PI3K inact in the cytosolic fraction of hearts from female and male mice at 12 weeks of age (a, 84 days, 12 weeks) or older (b, 160 to 252 days, 22 weeks). Expression levels of -adrenergic receptor kinase 1 ( ARK1) in the cytoplasm of the same animals were also unchanged. -adrenergic receptors are the most important regulators of cardiac function in vivo by interacting with endogenous catecholamines epinephrine and norepinephrine. In the failing human heart, increased levels of circulating catecholamines are associated with a reduction of cardiac AR density and desensitization of the remaining receptors (2,23). Our study shows that translocation of PI3K to activated receptors is an important step in the series of events that ultimately lead to chronic receptor downregulation in heart failure. Moreover, similar to previous observations in different animal models, our data shows that strategies that restore AR signaling consistently ameliorate cardiac dysfunction and improve survival in heart failure (5 7,12). Our data are also consistent with clinical studies showing that the treatment of patients with heart failure with beta-blockers actually results in the restoration of AR function, thereby improving cardiac function as well as survival (24 27). In addition to highlighting the important role of ARlocalized PI3K activity in failing hearts, our data also suggest that preferential activation of PI3K isoform might play a role in the pathogenesis of heart failure. A number of recent studies have demonstrated an important role for PI3K in the regulation of cardiac function (7,15,22,28,29). Recently, it has been shown that PI3K can regulate intracellular camp levels after pressure overload by interacting with the phosphodiesterase PDE3B (29). This interaction, likely based on the formation of a large multiprotein complex, does not require the kinase activity of PI3K because it was preserved in mice with a knock-in of a PI3K kinase-dead mutant gene (29). These interesting observations suggest that the regulation of intracellular camp levels is a complex process that results from a balance between camp generating and degradating pathways (29). Although it is not known whether the knock-in of the kinase-dead PI3K would actually normalize AR levels and function (29), our data from this study would predict that the kinase-dead PI3K would displace the active PI3K enzyme from ARK1 and lead to the normalization of AR signaling under pathological conditions. In fact, a large body of data using different in vitro and in vivo approaches, including the present study, indicates that inhibition of AR-targeted kinase activity of PI3K preserves AR function under pathological conditions

8 JACC Vol. 45, No. 11, 2005 June 7, 2005: Perrino et al. Targeted Inhibition of PI3K in Heart Failure 1869 (7,14,15,28). Distinctively, unlike strategies that utilize chronic AR activation as a therapy, such as chronic catecholamine infusion, overexpression of ARs or Gs (30 33), our approach of displacing the kinase activity of PI3K from ARK1 selectively restores agonist-dependent AR signaling without inducing chronic AR stimulation, a process we believe inherently maladaptive (4). Moreover, evidence is accumulating to support the concept that internalization of ARs per se is pathological, because the internalizing receptor can directly activate maladaptive signaling pathways in a G-protein independent fashion (34). Therefore, inhibition of AR down-regulation might actually prevent the activation of pathological pathways because it blocks maladaptive signals triggered by AR internalization (34). Importantly, we have previously shown that the absence of PI3K is not sufficient to prevent the increase in ARK1-associated PI3K activity on chronic catecholamine stimulation, because other isoforms of PI3K can eventually interact with ARK1 through the conserved phosphoinositide-kinase domain of PI3K and be targeted to activated receptors (7). Indeed, it has been recently shown that PI3K knockout mice undergo deterioration of cardiac function with necrosis and fibrosis when exposed to pressure overload (29). Therefore, a successful therapeutic strategy would require the displacement of all PI3K isoforms from ARK1, rather than the inhibition of a single isoform. Because the region responsible for the interaction with ARK1 is common to all PI3K isoforms, this therapeutic aim could be accomplished by the overexpression of small molecules that mimic this region and competitively displace the endogenous enzymes from the complex with ARK1. Undoubtedly, the most appealing aspect of this novel strategy is its specificity for AR-associated PI3K signaling; PI3K inact overexpression in transgenic mice reduces the amount of active PI3K enzyme targeted to agoniststimulated receptors but, importantly, does not affect other PI3K cytosolic downstream-signaling pathways. Indeed, in contrast to other PI3K mutants (29), PI3K inact overexpression does not alter PKB/GSK and MAPKs pathways, nor does it affect basal camp levels. Although it is conceivable that overexpression of PI3K inact might interact in the cytosol with phosphodiesterases (29) or directly affect adenylyl cyclase (7,21,22), we believe it is unlikely because only agonist-stimulated, and not basal cyclase activity, is affected in the binary CSQ/PI3K inact, which indicates the restoration of normal AR-Gs coupling. In contrast to the overexpression of the C-terminus of ARK1, which we have shown to ameliorate cardiac dysfunction in a number of heart failure models by disrupting the ARK1/G - interaction (5,6,12), selective inhibition of receptor-localized PI3K activity does not interfere with G - -mediated translocation of ARK1 to the receptor complex, or with other G - -mediated signals. Interestingly, because PI3K inact preserves AR responsiveness to catecholamine stimulation without inhibiting receptor phosphorylation by ARK1, these results suggest that PI3K inact overexpression might possibly preserve AR signaling by promoting a rapid sequence of receptor dephosphorylation and recycling to the plasma membrane. An additional unanticipated finding in our study was that the PI3K inact -mediated preservation of AR signaling resulted in a prolonged beneficial effect on survival in female mice. This result is congruent with data from clinical trials with heart failure patients indicating the existence of genderrelated differences in heart failure for presentation, etiology, and long-term survival (35,36). Moreover, our results are consistent with previous data from experimental studies showing reduced susceptibility of female mice to ischemia/ reperfusion injury (37,38), confirming the presence of a gender-related advantage for females in response to cardiovascular injury. These results are likely to depend on hormonal differences between genders, and future studies will be needed to specifically address the role of estrogens/ androgens in this murine model of heart failure. In conclusion, our study demonstrates that membranetargeting of PI3K at the site of activated ARs exerts an important role in the processes that lead to AR downregulation. Competitive displacement of PI3K from ARK1 preserves AR signaling in the failing heart, delaying the progression of cardiac dysfunction and prolonging the lifespan in this murine model of genetic heart failure. Hence, inhibition of AR-localized PI3K activity represents a novel therapeutic approach for patients with heart failure. Acknowledgment The authors gratefully thank Kristine Hesser Porter for her excellent technical assistance. Reprint requests and correspondence: Dr. Howard A. Rockman, Departments of Medicine, Cell Biology, and Molecular Genetics DUMC 3104, Room 226 CARL Building, Durham, North Carolina h.rockman@duke.edu. REFERENCES 1. Levy D, Kenchaiah S, Larson MG, et al. Long-term trends in the incidence of and survival with heart failure. N Engl J Med 2002;347: Bristow MR, Ginsburg R, Minobe W, et al. Decreased catecholamine sensitivity and beta-adrenergic-receptor density in failing human hearts. N Engl J Med 1982;307: Lohse MJ, Engelhardt S, Eschenhagen T. What is the role of beta-adrenergic signaling in heart failure? Circ Res 2003;93: Rockman HA, Koch WJ, Lefkowitz RJ. Seven-transmembranespanning receptors and heart function. Nature 2002;415: Rockman HA, Chien KR, Choi DJ, et al. Expression of a betaadrenergic receptor kinase 1 inhibitor prevents the development of myocardial failure in gene-targeted mice. Proc Natl Acad Sci U S A 1998;95: Harding VB, Jones LR, Lefkowitz RJ, Koch WJ, Rockman HA. Cardiac beta ARK1 inhibition prolongs survival and augments beta blocker therapy in a mouse model of severe heart failure. Proc Natl Acad Sci U S A 2001;98: Nienaber JJ, Tachibana H, Naga Prasad SV, et al. Inhibition of

9 1870 Perrino et al. JACC Vol. 45, No. 11, 2005 Targeted Inhibition of PI3K in Heart Failure June 7, 2005: receptor-localized PI3K preserves cardiac beta-adrenergic receptor function and ameliorates pressure overload heart failure. J Clin Invest 2003;112: Esposito G, Rapacciuolo A, Naga Prasad SV, et al. Genetic alterations that inhibit in vivo pressure-overload hypertrophy prevent cardiac dysfunction despite increased wall stress. Circulation 2002;105: Inglese J, Freedman NJ, Koch WJ, Lefkowitz RJ. Structure and mechanism of the G protein-coupled receptor kinases. J Biol Chem 1993;268: Akhter SA, Luttrell LM, Rockman HA, Iaccarino G, Lefkowitz RJ, Koch WJ. Targeting the receptor-gq interface to inhibit in vivo pressure overload myocardial hypertrophy. Science 1998;280: Ungerer M, Parruti G, Bohm M, et al. Expression of beta-arrestins and beta-adrenergic receptor kinases in the failing human heart. Circ Res 1994;74: White DC, Hata JA, Shah AS, Glower DD, Lefkowitz RJ, Koch WJ. Preservation of myocardial beta-adrenergic receptor signaling delays the development of heart failure after myocardial infarction. Proc Natl Acad Sci U S A 2000;97: Perry SJ, Lefkowitz RJ. Arresting developments in heptahelical receptor signaling and regulation. Trends Cell Biol 2002;12: Naga Prasad SV, Laporte SA, Chamberlain D, Caron MG, Barak L, Rockman HA. Phosphoinositide 3-kinase regulates beta2-adrenergic receptor endocytosis by AP-2 recruitment to the receptor/beta- Arrestin complex. J Cell Biol 2002;158: Naga Prasad SV, Barak LS, Rapacciuolo A, Caron MG, Rockman HA. Agonist-dependent recruitment of phosphoinositide 3-kinase to the membrane by beta-adrenergic receptor kinase 1. A role in receptor sequestration. J Biol Chem 2001;276: Jones LR, Suzuki YJ, Wang W, et al. Regulation of Ca2 signaling in transgenic mouse cardiac myocytes overexpressing calsequestrin. J Clin Invest 1998;101: Kohout TA, Takaoka H, McDonald PH, et al. Augmentation of cardiac contractility mediated by the human beta(3)-adrenergic receptor overexpressed in the hearts of transgenic mice. Circulation 2001; 104: Esposito G, Prasad SV, Rapacciuolo A, Mao L, Koch WJ, Rockman HA. Cardiac overexpression of a G(q) inhibitor blocks induction of extracellular signal-regulated kinase and c-jun NH(2)- terminal kinase activity in in vivo pressure overload. Circulation 2001;103: Esposito G, Santana LF, Dilly K, et al. Cellular and functional defects in a mouse model of heart failure. Am J Physiol Heart Circ Physiol 2000;279:H Cho MC, Rapacciuolo A, Koch WJ, Kobayashi Y, Jones LR, Rockman HA. Defective beta-adrenergic receptor signaling precedes the development of dilated cardiomyopathy in transgenic mice with calsequestrin overexpression. J Biol Chem 1999;274: Shioi T, Kang PM, Douglas PS, et al. The conserved phosphoinositide 3-kinase pathway determines heart size in mice. EMBO J 2000;19: Crackower MA, Oudit GY, Kozieradzki I, et al. Regulation of myocardial contractility and cell size by distinct PI3K-PTEN signaling pathways. Cell 2002;110: Bristow MR, Minobe WA, Raynolds MV, et al. Reduced beta 1 receptor messenger RNA abundance in the failing human heart. J Clin Invest 1993;92: Gilbert EM, Abraham WT, Olsen S, et al. Comparative hemodynamic, left ventricular functional, and antiadrenergic effects of chronic treatment with metoprolol versus carvedilol in the failing heart. Circulation 1996;94: Lowes BD, Gilbert EM, Abraham WT, et al. Myocardial gene expression in dilated cardiomyopathy treated with beta-blocking agents. N Engl J Med 2002;346: Sigmund M, Jakob H, Becker H, et al. Effects of metoprolol on myocardial beta-adrenoceptors and Gi alpha-proteins in patients with congestive heart failure. Eur J Clin Pharmacol 1996;51: Krum H, Roecker EB, Mohacsi P, et al. Effects of initiating carvedilol in patients with severe chronic heart failure: results from the COPERNICUS Study. JAMA 2003;289: Naga Prasad SV, Esposito G, Mao L, Koch WJ, Rockman HA. Gbetagamma-dependent phosphoinositide 3-kinase activation in hearts with in vivo pressure overload hypertrophy. J Biol Chem 2000;275: Patrucco E, Notte A, Barberis L, et al. PI3Kgamma modulates the cardiac response to chronic pressure overload by distinct kinasedependent and -independent effects. Cell 2004;118: Remme WJ. Inodilator therapy for heart failure. Early, late, or not at all? Circulation 1993;IV : 31. Engelhardt S, Hein L, Wiesmann F, Lohse MJ. Progressive hypertrophy and heart failure in beta1-adrenergic receptor transgenic mice. Proc Natl Acad Sci U S A 1999;96: Bisognano JD, Weinberger HD, Bohlmeyer TJ, et al. Myocardialdirected overexpression of the human beta(1)-adrenergic receptor in transgenic mice. J Mol Cell Cardiol 2000;32: Iwase M, Bishop SP, Uechi M, et al. Adverse effects of chronic endogenous sympathetic drive induced by cardiac GS alpha overexpression. Circ Res 1996;78: Lefkowitz RJ, Whalen EJ. beta-arrestins: traffic cops of cell signaling. Curr Opin Cell Biol 2004;16: Adams KF Jr., Sueta CA, Gheorghiade M, et al. Gender differences in survival in advanced heart failure. Insights from the FIRST study. Circulation 1999;99: Ghali JK, Krause-Steinrauf HJ, Adams KF, et al. Gender differences in advanced heart failure: insights from the BEST study. J Am Coll Cardiol 2003;42: Cross HR, Lu L, Steenbergen C, Philipson KD, Murphy E. Overexpression of the cardiac Na /Ca2 exchanger increases susceptibility to ischemia/reperfusion injury in male, but not female, transgenic mice. Circ Res 1998;83: Cross HR, Murphy E, Koch WJ, Steenbergen C. Male and female mice overexpressing the beta(2)-adrenergic receptor exhibit differences in ischemia/reperfusion injury: role of nitric oxide. Cardiovasc Res 2002;53: