Dynamic Regulation of Phosphoinositide 3-Kinase-γ Activity and β-Adrenergic Receptor Trafficking in End-Stage Human Heart Failure
Background— Downregulation of β-adrenergic receptors (βARs) under conditions of heart failure requires receptor targeting of phosphoinositide 3-kinase (PI3K)–γ and redistribution of βARs into endosomal compartments. Because support with a left ventricular assist device (LVAD) results in significant improvement of cardiac function in humans, we investigated the effects of mechanical unloading on regulation of PI3Kγ activity and intracellular distribution of βARs. Additionally, we tested whether displacement of PI3Kγ from activated βARs would restore agonist responsiveness in failing human cardiomyocytes.
Methods and Results— To test the role of PI3K on βAR endocytosis in failing human hearts, we assayed for PI3K activity in human left ventricular samples before and after mechanical unloading (LVAD). Before LVAD, failing human hearts displayed a marked increase in βAR kinase 1 (βARK1)–associated PI3K activity that was attributed exclusively to enhanced activity of the PI3Kγ isoform. Increased βARK1-coupled PI3K activity in the failing hearts was associated with downregulation of βARs from the plasma membrane and enhanced sequestration into early and late endosomes compared with unmatched nonfailing controls. Importantly, LVAD support reversed PI3Kγ activation, normalized the levels of agonist-responsive βARs at the plasma membrane, and depleted the βARs from the endosomal compartments without changing the total number of receptors (sum of plasma membrane and early and late endosome receptors). To test whether the competitive displacement of PI3K from the βAR complex restored receptor responsiveness, we overexpressed the phosphoinositide kinase domain of PI3K (which disrupts βARK1/PI3K interaction) in primary cultures of failing human cardiomyocytes. Adenoviral-mediated phosphoinositide kinase overexpression significantly increased basal contractility and rapidly reconstituted responsiveness to β-agonist.
Conclusions— These results suggest a novel paradigm in which human βARs undergo a process of intracellular sequestration that is dynamically reversed after LVAD support. Importantly, mechanical unloading leads to complete reversal in PI3Kγ and βARK1-associated PI3K activation. Furthermore, displacement of active PI3K from βARK1 restores βAR responsiveness in failing myocytes.
Received April 1, 2007; accepted September 21, 2007.
Heart failure is the common clinical syndrome that results from virtually all forms of cardiac disease and is consistently characterized by extensive abnormalities in the β-adrenergic receptor (βAR) system.1,2 A vast body of evidence indicates that βAR dysfunction is an early event in the transition of the overloaded heart toward failure, triggered by a combination of excessive neurohumoral stimulation and mechanical stress.3 Indeed, either mechanical unloading of the left ventricle through the use of left ventricular assist devices (LVADs) or β-blocker therapy consistently ameliorates left ventricular dysfunction and restores βAR signaling in human heart failure.4–8
Clinical Perspective p 2579
βAR desensitization and downregulation are the classic receptor abnormalities associated with human heart failure. Desensitization of βARs is initiated by phosphorylation of the receptor by βAR kinase 1 (βARK1).9,10 β-Arrestin binds to the phosphorylated βAR sterically, which hinders further G-protein coupling (desensitization) and commences the process of receptor internalization.10 Internalized βARs are trafficked through specialized endosomal compartments based on the fate of the receptor. Receptors trafficking through early endosomes are thought to undergo dephosphorylation before being recycled back to the plasma membrane.11 Receptors trafficking through late endosomes are targeted for the lysosomal degradation pathway.11 According to the current paradigm, a combination of increased rates of β1AR lysosomal degradation and reduced receptor transcript determines selective β1AR downregulation at the plasma membrane under conditions of heart failure.12,13 However, the durations for which the internalized receptors reside in each of the endosomal compartments are currently unknown. Moreover, it is not known whether the path toward degradation can be blocked or even reversed. Indeed, accumulating evidence now suggests that the process of internalization itself may be pathological, because the internalization of receptors can activate detrimental signaling pathways directly.14 Therefore, strategies that might alter receptor endocytosis may exert a beneficial effect in heart failure.
The intracellular route of internalized βARs in end-stage human heart failure and the molecules involved in their trafficking toward lysosomal degradation are not clearly known. Our previous studies have shown that efficient βAR internalization requires the recruitment of phosphoinositide 3-kinase (PI3K) to agonist-stimulated βARs.15,16 The recruitment of PI3K to the βAR is brought about by βARK1, which associates with PI3K to form a cytosolic complex through the phosphoinositide kinase (PIK) domain of PI3K.15,16 Indeed, cardiac-specific overexpression of the PIK domain competitively displaces all of the endogenous isoforms of PI3K from βARK1, which results in reduced βAR-localized PI3K activity.17 Importantly, targeted PI3K inhibition prevents βAR sequestration into endosomal compartments despite chronic agonist stimulation and reverses βAR abnormalities in a large-animal model of heart failure.17
Because very little is known about βAR trafficking in the heart, we hypothesized that the dynamic sequestration of βARs into endosomal compartments may represent an important mechanism to regulate adrenergic responsiveness in the failing human heart. Furthermore, we tested the hypothesis that βAR-targeted PI3K activity plays a pivotal role in receptor sequestration using a gene transfer approach. This approach allowed us to determine whether competitive displacement of PI3K from the receptor complex would reconstitute βAR responsiveness to β-agonist stimulation in failing human cardiomyocytes.
Paired samples of left ventricular tissue were obtained from 18 patients at a single institution at the time of LVAD implantation and at subsequent orthotopic heart transplantation. The indication for LVAD implantation in all patients was end-stage heart failure with deterioration of cardiac function despite maximal medical therapy (New York Heart Association class IV). Patients were enrolled between May 2000 and April 2006. After a period of LVAD support (Table), these patients had their LVAD explanted and a heart transplant performed. Nonfailing control myocardium was obtained from 10 unmatched organ donors whose hearts were not suitable for transplantation in spite of normal ventricular structure and function. All procedures for tissue procurement were performed in compliance with institutional guidelines for human research and an approved institutional review board protocol at Duke University Medical Center and The Cleveland Clinic Foundation.
A transmural apical core of the left ventricle was excised during LVAD implantation. The specimen was examined, and any visible scar tissue was excised and discarded. The remaining myocardium was snap-frozen in liquid nitrogen and stored at −80°C. A second sample of myocardium was taken from the left ventricular free wall at the time of cardiac transplantation and stored at −80°C. All samples were paired from LVAD implantation to explantation (n=18). Nonfailing left ventricular tissue was obtained in a similar fashion from organ donor hearts deemed unsuitable for transplantation but demonstrated no evidence of functional or structural heart disease.
Purification of Plasma Membrane, Early Endosomes, and Late Endosomes
To separate the plasma membrane and cytosolic fractions from the left ventricles, cardiac samples were homogenized in ice-cold lysis buffer containing 5 mmol/L Tris-Cl pH 7.5, 5 mmol/L EDTA, 1 mmol/L PMSF, and 2 μg/mL leupeptin and aprotinin. Intact cell debris and nuclei were removed by centrifugation at 1000g for 5 minutes. The collected supernatant was subjected to centrifugation at 37 000g for 20 minutes. The crude supernatant cytosolic fraction was transferred to a new tube, and the pellet of cardiac membranes was resuspended in 75 mmol/L Tris-HCl pH 7.5, 2 mmol/L EDTA, 12.5 mmol/L MgCl2. Early and late endosomal fractions were recovered after ultracentrifugation of the crude cytosolic fraction for 1 hour at 300 000g and 200 000g, respectively. The pellets corresponding to early and late endosomes were resuspended in 75 mmol/L Tris-Cl pH 7.5, 2 mmol/L EDTA, 12.5 mmol/L MgCl2, and fractions were immunoblotted for the early and late endosomal markers Rab5 and Lampl, respectively.17
βAR Radioligand Binding and Adenylyl Cyclase Activity
Receptor binding with 20 μg of protein from the plasma membrane and early/late endosomal fractions was performed as described previously17 with βAR ligand [125I]-cyanopindolol (250 pmol/L). All assays were performed in triplicate, and receptor density (in femtomoles) was normalized to milligrams of membrane protein. Adenylyl cyclase assays were performed as described previously17 with 20 μg of the membrane fraction under basal conditions and after stimulation with isoproterenol (10−4 mol/L). Generated cAMP was quantified with a liquid scintillation counter (MINAXI-4000, Packard Instrument Co/PerkinElmer, Waltham, Mass).
PI3K assays were performed by immunoprecipitation of the PI3K α- and γ-isoforms from the cytosolic fraction, as described previously.15 βARK1-associated PI3K activity was measured after immunoprecipitation of 400 μg of proteins from the membrane fraction with a polyclonal antibody directed against βARK1 (Santa Cruz Biotechnology, Santa Cruz, Calif). After the lipid kinase assay, lipids were extracted with chloroform/methanol (ratio of 1:1). The organic phase was spotted on thin-layer chromatography plates and resolved chromatographically with 2N glacial acetic acid/1-propanol (35:65). Dried plates were exposed for autoradiography. These signals were further quantified with Bio-Rad Fluoro-S MultiImager software (Bio-Rad, Hercules, Calif).
Primary Culture of Failing Human Cardiac Myocytes and Cell Contractility Studies
Cell contractility studies were performed on human cardiac myocytes isolated as described previously.18 The myocytes were infected with AdEV (empty virus) or AdPIK (which encoded the PIK domain of PI3K) viruses at a multiplicity of infection (MOI) of 100, as described previously.18 Uninfected cells from the same preparations were used as controls. Thirty-six to 48 hours after infection, single-cell contractions were measured in rod-shaped cells by video edge detection (Crescent Electronics, Sandy, Utah). The recordings were made during electric field stimulation in basal conditions and after administration of isoproterenol 1 μmol/L, as described previously.18
Confocal Microscopy in Fixed Cells
Confocal microscopy was performed as described previously.17 Briefly, failing human cardiac myocytes were plated onto glass-bottomed dishes and infected with an MOI of 100 for adenoviruses encoding either green fluorescent protein (AdGFP) or FLAG-PIK (AdPIK). Cells infected with AdEV were used as control cells for visualization. Cells were fixed in 8% paraformaldehyde and made permeable with a solution that contained 0.5% Triton X-100 in PBS. FLAG-PIK expression was visualized by incubation of cells with anti-FLAG monoclonal antibody (1:500, Sigma, St. Louis, Mo) followed by goat anti-mouse IgG conjugated with Texas Red (1:500, Molecular Probes, Carlsbad, Calif). Samples were all visualized with single sequential line excitation filters at 488 and 568 nm and emission filters set at 505 to 550 nm for GFP detection and 585 for Texas Red detection.
RNA Isolation and Northern Blotting
RNA was isolated from left ventricles with the RNA-Bee-RNA isolation reagent (Tel-Test, Friendswood, Tex). Briefly, samples were homogenized in 1.0 mL of RNA-Bee reagent with a polytron homogenizer. Chloroform (0.2 mL) was added to the homogenate and mixed vigorously for 15 to 30 seconds. The samples were incubated on ice for 5 minutes and centrifuged at 12 000g for 15 minutes at 4°C. The aqueous phase was transferred to a new tube, and 0.5 mL isopropanol was added. The tubes were incubated for 10 minutes at room temperature. RNA was precipitated by centrifugation at 12 000g for 5 minutes at 4°C. The RNA pellet was washed with 75% ethanol, air-dried, and resuspended in RNase-free water. Ten micrograms of RNA was size-fractionated by denaturing formaldehyde gel electrophoresis, transferred to nylon membrane by capillary action, and cross-linked with ultraviolet light. The membrane was stained with 0.5% methylene blue to visualize the transferred RNA onto the nylon membrane and also to check for equal loading across lanes. The membrane was destained with diethyl pyrocarbonate–treated sterile water. After destaining, the membrane was hybridized with a human β1AR [32P]-labeled cDNA probe. The nylon membrane was stripped and reprobed with β2AR [32P]-labeled cDNA probe. After hybridization, filters were washed under stringent conditions, and transcripts were detected by autoradiography.
Data are expressed as mean±SEM repeated-measures ANOVA. Changes in the same human samples before and after LVAD were evaluated by repeated-measures ANOVA. One-way ANOVA was used when single values were compared, and 2-way ANOVA with Bonferroni correction was used when multiple comparisons were performed. For all analyses, P<0.05 was considered significant, and in case of multiple comparisons with Bonferroni correction, P<0.01 was considered significant.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Patient characteristics are summarized in the Table. Eighteen patients, 15 men and 3 women, were included in the biochemical analysis, and the average age of the patients was 54.9±8.7 years. Twelve of the patients were white, and 6 were black. Nine of the patients had an ischemic cardiomyopathy, 1 had a viral cardiomyopathy, and the remaining 8 had idiopathic dilated cardiomyopathies. The mean duration of LVAD support was 78.5±55.3 days (range 9 to 190 days). The majority of patients were supported with a HeartMate XVE (n=10). Five patients had a Thoratec LVAD, whereas 1 patient each had a HeartMate II, Thoratec BIVAD, or Abiomed BIVAD. The average pre-LVAD ejection fraction was 20%. When available, the post-LVAD/pretransplant ejection fraction was typically unchanged, although the duration of the time of the study after LVAD implantation varied considerably. The average age of the 4 patients in the isolated myocyte study was 50 years. All patients were male, 3 were white, and 1 was black. Three were supported by a HeartMate XVE LVAD for durations of 8, 51, and 279 days, respectively, and 1 patient did not have an LVAD. The average ejection fraction was 17.5%, and no patients had a post-LVAD ejection fraction available. The majority of patients were given some form of inotropic and/or vasopressor support at the time of LVAD implantation (Table).
Load-Dependent Trafficking of βARs in End-Stage Human Heart Failure
Previous studies have shown that mechanical unloading of the failing human heart ameliorates βAR abnormalities and improves cardiac dysfunction5,6; however, the mechanism responsible for this beneficial effect is currently unknown. To test whether endosomal sequestration of βARs might play a role in regulating adrenergic responsiveness in the failing human heart, we fractionated by ultracentrifugation the plasma membrane and early and late endosomal compartments from human heart samples before and after implantation of an LVAD. Control heart samples from unmatched nonfailing human hearts were obtained from organ donors whose hearts were unsuitable for transplantation and who had no prior history of cardiac disease.
Human heart failure (HF) was characterized by marked downregulation of βARs from the plasma membrane (Figure 1A; P<0.01 versus corresponding controls), as expected.12,13 Interestingly, the downregulation of βARs at the plasma membrane was associated with a marked increase in the levels of the receptor in the early and late endosomal fractions compared with control hearts (Figure 1A; P<0.01 versus corresponding control hearts). Consistent with previous studies,6,19 mechanical unloading of the left ventricle through an LVAD significantly increased plasma membrane levels of βARs (Figure 1A). Importantly, replenishment of plasma membrane βARs after LVAD appeared to be associated with a marked reduction of “sequestered” endosomal pools of βARs (Figure 1A), which suggests a redistribution of the receptors from the endosomal fractions to the plasma membrane. This replenishment of the receptors in the plasma membrane after LVAD appeared to restore βAR densities to control patient levels (Figure 1A).
Despite the marked differences in the intracellular distribution of βARs among the different groups (Figure 1A), total βAR levels (after addition of all the fractions) were not statistically different in control and HF samples. Furthermore, the total number of receptors was not affected by mechanical unloading of the failing ventricle (Figure 1B). To test whether unloading of the heart after LVAD results in transcriptional modulation of β1ARs or β2ARs, resulting in replenishment of receptor at the plasma membrane, we performed Northern blot analysis. RNA was isolated from control, HF, and HF/LVAD patient samples, and the blots were probed with radiolabeled β1AR or β2AR probes. Significant transcriptional downregulation of β1ARs was observed in HF and HF/LVAD samples compared with control samples (Figure 1C), which indicates that mechanical unloading by LVAD implantation does not modulate β1AR transcription. Interestingly, we did not observe significant differences in the β2AR transcript among the samples. Equal loading of RNA for each of the samples was confirmed by methylene blue staining of ribosomal RNA 28S and 16S on the nylon membrane after the transfer of RNA to the membrane (Figure 1C).
To test whether increased βAR levels at the plasma membrane after LVAD would translate into restoration of βAR signaling, we measured adenylyl cyclase activity in plasma membrane fractions derived from hearts belonging to the different groups. As expected, HF samples displayed reduced basal and isoproterenol-stimulated adenylyl cyclase activity compared with normal hearts (Figure 1D; P<0.01 versus control isoproterenol stimulation). Interestingly, LVAD support restored βAR/Gs protein coupling, as measured by a significant improvement in basal and isoproterenol-stimulated adenylyl cyclase activity in HF/LVAD human hearts (Figure 1D; P<0.01 versus basal levels). Furthermore, direct activation of G protein by sodium fluoride showed significant reduction in adenylyl cyclase activity (data not shown) consistent with upregulation in the inhibitory G protein, Gαi.20 Taken together, these results suggest that βAR downregulation in human heart failure results from a combination of reduced receptor synthesis and enhanced intracellular receptor sequestration in various endosomal compartments and that βAR sequestration is a dynamic process that can be reversed by mechanical unloading of the failing human heart.
Load-Dependent Increase in βAR-Targeted PI3K Activity in Failing Human Hearts
Agonist-dependent sequestration of βARs requires translocation and activation of βARK1 by liberated Gβγ-subunits.9 Therefore, we measured βARK1 protein levels in myocardial lysates from the different patient groups. A marked increase in βARK1 levels was seen in HF hearts compared with unmatched controls, which is consistent with previous studies (Figure 2A). Importantly, a complete reversal of βARK1 protein levels occurred in hearts after LVAD support (Figure 2A). We have previously shown that βARK1 associates with PI3K to form a stable cytosolic complex that is recruited to the activated βARs after receptor stimulation.15 Furthermore, enhanced membrane-targeted βARK1-associated PI3K activity is invariably associated with βAR dysfunction in different animal models of heart failure.17,20 To test whether an increase in βARK1-associated PI3K activity occurs in human heart failure, βARK1 was immunoprecipitated from the cardiac plasma membranes, and associated PI3K activity was measured. A marked increase in βARK1-associated PI3K activity was seen in cardiac membranes before LVAD (Figure 2B). Importantly, mechanical unloading resulted in a dramatic reduction of βARK1-associated PI3K activity to levels similar to the controls (Figure 2B).
Selective Increase of PI3Kγ Activity in End-Stage Human Heart Failure
βARK1-associated PI3K activity is enhanced in human heart failure, and because βARK1 can interact with either PI3K γ- or α-isoforms, we attempted to identify the PI3K isoform that contributes to the increase in PI3K activity. PI3K γ- and PI3K α-isoforms were immunoprecipitated from the myocardial lysates, and lipid kinase activity was measured in the immunoprecipitates. PI3Kγ activity was increased significantly in HF hearts, which showed that the PI3K γ-isoform was responsible for the significant changes in βARK1-associated PI3K activity (Figure 3A and 3B). Importantly, mechanical unloading after LVAD support resulted in a significant reduction of PI3Kγ activity (Figure 3A and 3B). In contrast, no appreciable change in PI3Kα activity was seen in human heart failure samples compared with control (Figure 3C and 3D), which suggests that PI3Kα may not play a key role in end-stage heart failure. Taken together, these data demonstrate that the βAR abnormalities seen in human heart failure are associated with a significant increase in βARK1-associated PI3K activity, with a selective increase in the activity of the PI3K γ-isoform.
Overexpression of PIK Domain Peptide Restores Cardiac Contractility in Myocytes Isolated From Failing Human Hearts
Since failing human hearts displayed increased βARK1-associated PI3K activity and profound βAR abnormalities, we tested whether targeted inhibition of PI3K at the receptor might reconstitute βAR responsiveness in failing human cardiomyocytes. To this end, we infected failing cardiomyocytes with recombinant adenoviruses containing a FLAG-tagged PIK domain of PI3K.17 Human cardiac myocytes were infected at an MOI of 100 to result in significant expression of PIK domain peptide (AdPIK). Myocytes were also infected with adenoviruses encoding only GFP (AdGFP) or empty vector (AdEV) as controls. To test expression of the PIK domain peptide, we performed confocal microscopy on infected cardiomyocytes. The myocytes infected with empty vector (AdEV) showed minimal staining with Texas Red (Figure 4A). In contrast, intense Texas Red staining was seen in cardiomyocytes on AdPIK infection, which showed expression of PIK domain peptide (AdPIK; Figure 4A). Myocytes infected with only AdGFP showed strong GFP immunofluorescence (Figure 4A).
To test whether PIK overexpression could rapidly reconstitute βAR responsiveness to agonist in failing cardiomyocytes, contractility studies were performed on cells isolated from failing human hearts at the time of cardiac transplantation. The isolated cardiac myocytes were infected with either empty virus AdEV (HF-AdEV) or AdPIK (HF-AdPIK), and contractility studies were performed on uninfected (HF) or infected (36 to 48 hours after infection) cardiomyocytes. In multiple sets of cells, we measured basal and isoproterenol-stimulated percentage of cell shortening, velocity of shortening (dL/dt−), and velocity of relaxation (dL/dt+). PIK overexpression markedly normalized the rate of cell contraction and relaxation under basal conditions and after stimulation with isoproterenol (Figure 4B; P<0.001 for isoproterenol versus basal levels, P<0.005 AdPIK isoproterenol versus HF isoproterenol and AdEV isoproterenol, ANOVA with Bonferroni correction). In addition, human cardiomyocytes infected with AdPIK displayed improved cellular contractility compared with AdEV or HF, both under basal conditions and on agonist stimulation (Figure 4C; P<0.001 for all isoproterenol versus basal, P<0.005 AdPIK isoproterenol versus HF isoproterenol and AdEV isoproterenol, ANOVA with Bonferroni correction). Taken together, these studies demonstrate that disruption of the βARK1/PI3K complex improves failing myocyte function by reversing already established abnormalities in βAR function.
The failing human heart is characterized by significant downregulation of βARs at the plasma membrane and sequestration of receptors into stable intracellular pools, consistent with animal models of hypertrophy and heart failure.17 In the present study, we demonstrate that mechanical unloading of the failing human heart depletes the stable intracellular reservoirs of βARs, potentially promoting redistribution of the receptors to the plasma membrane. The concept of redistribution of receptors is consistent with restoration of βAR responsiveness after LVAD support, because we showed that no change in the β1AR or β2AR transcript levels occurred after mechanical unloading. Furthermore, mechanical unloading is associated with significant reduction in PI3Kγ activity and βARK1-associated PI3K activity that is markedly elevated in failing human hearts. Importantly, competitive displacement of active PI3K from the βARK1 complex by the PIK domain peptide of PI3K results in remarkable reversal of established βAR abnormalities in failing cardiomyocytes. These studies highlight a novel regulatory mechanism of βAR trafficking in conditions of heart failure that could potentially be regulated by PI3K activity.
βARs are the interface between the heart and the continuously changing environment.10 Sudden cardiac overload rapidly induces βAR desensitization and downregulation,3 and these abnormalities are believed to be intrinsically linked to the deleterious progression of cardiac dysfunction.3,10 According to the current paradigm, βAR internalization/degradation along with a significant dampening of βAR transcription determines the downregulation of the βAR system in failing human hearts.13,21 In contrast to the current paradigm, the present studies show that end-stage failing human hearts are characterized by large intracellular pools of βARs within endosomal compartments, consistent with animal models of heart failure.17 Furthermore, we show that the change in the environmental conditions (ie, change in cardiac load or neurohumoral stimulation) might indeed promote the reverse trafficking of “sequestered” βARs back to normal function at the plasma membrane, because normalized receptor numbers are not mirrored by transcriptional activation. This is an important concept and a shift from the current understanding wherein it is thought that receptors targeted for degradation are trafficked to the late endosomes.11 Although it is not known how long βARs reside within the late endosomal compartment before degradation, the reduction of receptor densities from the late endosomal component of failing human hearts on mechanical unloading supports the concept of reverse trafficking. It is proposed that mechanical unloading of the heart by an LVAD appears to initiate a process of reverse remodeling22 that is associated with significant recovery of βAR myocardial responsiveness and βAR densities and reversal of ryanodine receptor hyperphosphorylation.6,19,23,24 This reverse remodeling in response to LVAD therapy supports the contention that the failing human heart retains some capacity for recovery. Furthermore, because the downregulation of βARs is reversible, it suggests that reverse trafficking of receptors could be integral to the remodeling process.
Evidence from previous studies has demonstrated that LVAD support results in significant improvement in βAR responsiveness and restoration of the inotropic response to sympathetic stimulation.6 It has been shown that the amelioration of cardiac function after LVAD support is associated with recovery in receptor density at the cardiac membrane,19 consistent with the present studies. Importantly, the present studies demonstrate that the recovery of receptor density at the plasma membrane after LVAD could be due to redistribution of the sequestered βARs from various endosomal compartments to the plasma membrane. This concept is supported by the RNA analysis, which showed no significant changes at the transcriptional level in the paired samples after mechanical unloading by LVAD. The lack of change in total βAR densities among nonfailing and failing human hearts suggests that βAR endosomal sequestration might be an important mechanism governing the process of receptor downregulation from the cell.
The mechanisms by which mechanical unloading of the left ventricle might promote the translocation of βARs toward the plasma membrane are still unknown. Besides reducing mechanical stress, LVAD support has been shown to decrease adrenergic hyperactivation and the local release of cytokines that might be involved in the regulation of cardiac function,25,26 and therefore, it might exert beneficial effects on βAR signaling either directly or indirectly. The partial recovery of ventricular function by reverse remodeling after LVAD could likely involve regulation of multiple signaling pathways,22 and we believe that the dynamic regulation of PI3Kγ activity seen in the present studies would be integral to this process.
The present study shows that end-stage failing human hearts are characterized by selective augmentation of PI3Kγ activity, and these data are indeed consistent with our previous observations in animal models of hypertrophy and failure.17 Surprisingly, PI3Kγ activation can be completely reversed by mechanical unloading of the failing ventricle. In contrast, we did not observe any significant changes in PI3Kα activity among the groups. These results from end-stage failing human hearts are consistent with recent studies in mice that showed significant activation of PI3Kα on physical training3 or PI3Kγ activation in response to pathological stresses.3 Activation of PI3Kγ is thought to be involved in deleterious signaling, because overexpression of catalytically inactive PI3Kγ results in amelioration of cardiac function in mouse models of heart failure.20,27 The reversal in the activation of PI3Kγ after LVAD suggests that this enzyme might be integral to the beneficial reverse remodeling that occurs after mechanical unloading of the failing ventricle. Indeed, PI3K regulates multiple signaling pathways, including Akt/GSK and p70S6K, and dynamic regulation of AKT, p70S6K, and mitogen-activated protein kinase/extracellular signal-regulated kinase is seen before and after LVAD implantation,28 which suggests that targeting PI3Kγ could be a potential therapeutic strategy in conditions of heart failure.
It is known that βARK1 interacts with PI3K to form a cytosolic complex and mediates PI3K recruitment to the agonist-activated βAR.15 Furthermore, βARK1-mediated PI3K activity at the receptor complex plays a critical role in receptor downregulation (internalization) in mice.20 The present studies in the human heart show that a significant increase in βARK1-associated PI3K activity occurs in cardiac membranes from failing left ventricles, which suggests a critical role for the βARK1/PI3K complex in the process of receptor downregulation seen in heart failure. Displacement of active PI3K from the βARK/PI3K complex by a dominant-negative strategy results in inhibition of receptor internalization and significant amelioration of cardiac dysfunction in animal models of heart failure.20,27 Therefore, the displacement of active PI3K from the βARK/PI3K complex would be an attractive strategy for novel therapeutic interventions in heart failure. Consistent with this idea, our in vitro gene therapy approach indicates that the competitive displacement of PI3K from βARK1 markedly improves multiple parameters of contractility in failing human cardiomyocytes. One of the potential mechanisms by which the PIK domain peptide could provide these beneficial effects may involve changing the fate of agonist-stimulated βARs and redirecting sequestered pools of βARs toward plasma membrane, which would result in functional receptors and enhanced contractility. Because of the limited amount of available primary human cardiomyocytes, we were not able to perform molecular/biochemical analyses to address βAR compartmentalization after PI3K inhibition in human cardiac cells.
On the basis of our observations, we believe that inhibition of PI3K by a dominant-negative approach could be a novel therapeutic strategy. This approach is critically different from the strategy of using the inhibitor of βARK1, βARKct (the carboxyl-terminus of βARK1), which has been shown to ameliorate contractility of failing human cardiomyocytes.18,29 βARKct sequesters Gβγ-subunits, which leads to inhibition of receptor phosphorylation by βARK1, and this results in the amelioration of cardiac function.18
The activation and subsequent reversal of βARK1-associated PI3K activity after mechanical unloading reflects the fact that βARK1-associated PI3K activity is part of the deleterious signal activated in conditions of heart failure that is reversed as part of the beneficial remodeling that occurs after LVAD support. The reduction in βARK1-associated PI3K activity could be due to lower PI3Kγ activity or reduced levels of βARK1. Dynamic regulation of βARK1 protein levels before and after LVAD is consistent with previous studies18,29 and may be one of the key components in the formation of the βARK1/PI3K complex and initiation of deleterious signaling.
In conclusion, the present studies demonstrate that end-stage failing human hearts are characterized by a significant increase in βARK1-associated PI3K activity. The augmentation of βARK1-associated PI3K activity is specifically due to the increase in PI3Kγ isoform activity. Mechanical unloading of the heart by an LVAD reduces βARK1-associated PI3K activity and results in restoration of plasma membrane βAR densities to densities similar to those of the nonfailing heart. The restoration of plasma membrane βAR densities potentially could occur by redistribution of the receptors from the sequestered endosomal compartments. Finally, a dominant-negative strategy to inhibit the recruitment of active PI3K to the βAR complex improves multiple parameters of contractility in failing human cardiomyocytes, which suggests that targeting the βARK1/PI3K complex could be a novel, attractive therapeutic strategy to normalize βAR function in human heart failure.
The authors would like to acknowledge the valuable support and help of Drs Randall Starling and Christine S. Moravec, Division of Cardiovascular Medicine, Cleveland Clinic Foundation, for some of the nonfailing and failing human heart samples. The authors also would like to thank Weili Zou and Anita Shukla for their excellent technical assistance.
Sources of Funding
This work was supported in part by a Beginning Grant-in-Aid (046580U) from the American Heart Association Mid-Atlantic Affiliate to Dr Naga Prasad and a National Institutes of Health grant HL072183 to Dr Milano.
Barry WH, Gilbert EM. How do beta-blockers improve ventricular function in patients with congestive heart failure? Circulation. 2003; 107: 2395–2397.
Lohse MJ, Engelhardt S, Eschenhagen T. What is the role of beta-adrenergic signaling in heart failure? Circ Res. 2003; 93: 896–906.
Levin HR, Burkhoff D, Oz MC, Chen JM, Rose EA, Packer M. Reversal of ventricular remodeling in patients with end-stage heart failure by prolonged mechanical unloading. Circulation. 1995; 91: 2717–2720.
Ogletree-Hughes ML, Stull LB, Sweet WE, Smedira NG, McMarthy PM, Moravec CS. Mechanical unloading restores beta-adrenergic responsiveness and reverses receptor downregulation in the failing human heart. Circulation. 2001; 104: 881–886.
Bristow M. Antiadrenergic therapy of chronic heart failure: surprises and new opportunities. Circulation. 2003; 107: 1100–1102.
Gilbert EM, Abraham WT, Olsen S, Hattler B, White M, Mealy P, Larrabee P, Bristow MR. Comparative hemodynamic, left ventricular functional, and antiadrenergic effects of chronic treatment with metoprolol versus carvedilol in the failing heart. Circulation. 1996; 94: 2817–2825.
Lefkowitz RJ. G protein-coupled receptors, III: new roles for receptor kinases and beta-arrestins in receptor signaling and desensitization. J Biol Chem. 1998; 273: 18677–18680.
Drake MT, Shenoy SK, Lefkowitz RJ. Trafficking of G protein-coupled receptors. Circ Res. 2006; 99: 570–582.
Bristow MR, Ginsburg R, Umans V, Fowler M, Minobe W, Rasmussen R, Zera P, Menlove R, Shah P, Jamieson S. Beta 1- and beta 2-adrenergic-receptor subpopulations in nonfailing and failing human ventricular myocardium: coupling of both receptor subtypes to muscle contraction and selective beta 1-receptor down-regulation in heart failure. Circ Res. 1986; 59: 297–309.
Lefkowitz RJ, Shenoy SK. Transduction of receptor signals by beta-arrestins. Science. 2005; 308: 512–517.
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: 18953–18959.
Naga Prasad SV, Laporte SA, Chamberlain D, Caron MG, Barak LS, 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: 563–575.
Perrino C, Naga Prasad SV, Schroder JN, Hata JA, Milano CA, Rockman HA. Restoration of beta-adrenergic receptor signaling and contractile function in heart failure by disruption of the betaARK1/phosphoinositide 3-kinase complex. Circulation. 2005; 111: 2579–2587.
Williams ML, Hata JA, Schroder J, Rampersaud E, Petrofski J, Milano CA, Koch WJ. Targeted beta-adrenergic receptor kinase (betaARK1) inhibition by gene transfer in failing human hearts. Circulation. 2004; 109: 1590–1593.
Wohlschlaeger J, Schmitz KJ, Schmid C, Schmid KW, Keul P, Takeda A, Weis S, Levkau B, Baba HA. Reverse remodeling after insertion of left ventricular assist devices (LVAD): a review of the morphological and molecular changes. Cardiovasc Res. 2005; 68: 376–386.
Barbone A, Holmes JW, Heerdt PM, The’ AH, Naka Y, Joshi N, Daines M, Marks AR, Oz MC, Burkhoff D. Comparison of right and left ventricular responses to left ventricular assist device support in patients with severe heart failure: a primary role of mechanical unloading underlying reverse remodeling. Circulation. 2001; 104: 670–675.
Perrino C, Naga Prasad SV, Patel M, Wolf MJ, Rockman HA. Targeted inhibition of beta-adrenergic receptor kinase-1-associated phosphoinositide-3 kinase activity preserves beta-adrenergic receptor signaling and prolongs survival in heart failure induced by calsequestrin overexpression. J Am Coll Cardiol. 2005; 45: 1862–1870.
Baba HA, Stypmann J, Grabellus F, Kirchhof P, Sokoll A, Schafers M, Takeda A, Wilhelm MJ, Scheld HH, Takeda N, Breithardt G, Levkau B. Dynamic regulation of MEK/Erks and Akt/GSK-3beta in human end-stage heart failure after left ventricular mechanical support: myocardial mechanotransduction-sensitivity as a possible molecular mechanism. Cardiovasc Res. 2003; 59: 390–399.
Hata JA, Williams ML, Schroder JN, Lima B, Keys JR, Blaxall BC, Petrofski JA, Jakol A, Milano CA, Koch WJ. Lymphocyte levels of GRK2 (betaARK1) mirror changes in the LVAD-supported failing human heart: lower GRK2 associated with improved beta-adrenergic signaling after mechanical unloading. J Card Fail. 2006; 12: 360–368.
The β-adrenergic receptor (βAR) signaling system plays an important role in regulating Ca2+ homeostasis that mediates positive inotropic effects in the heart via adenylyl cyclase activation. Abnormalities in the βAR system, including reduced catecholamine sensitivity and decreased βAR density, are hallmarks of failing human hearts. Consequently, βAR density, sensitivity, and distribution are irrevocably linked to heart failure treatment and cardiac recovery. Indeed, βAR-blocker therapy is currently the most effective available treatment for cardiac dysfunction. According to the current paradigm, internalization/degradation of βARs along with a significant dampening of βAR transcription determines the downregulation of βARs in the failing human heart. Conversely, our studies show that end-stage failing human hearts are characterized by large intracellular pools of “sequestered” βARs within endosomal compartments that could be redistributed dynamically after changes in cardiac load or neurohumoral stimulation, resulting in increased plasma membrane receptors. This dynamic redistribution of βARs could account in part for the significant recovery of cardiac function seen with mechanical unloading after use of a left ventricular assist device. Increased βAR sequestration in heart failure is accompanied by selective activation of phosphoinositide 3-kinase-γ that is reversed after mechanical unloading, which is associated with restoration of plasma membrane βAR density. Importantly, inhibition of phosphoinositide 3-kinase recruitment to the βAR by a dominant-negative strategy rapidly reconstitutes βAR responsiveness in failing human cardiomyocytes, thus demonstrating a novel regulation of βAR responsiveness by phosphoinositide 3-kinase-γ. Therefore, targeting phosphoinositide 3-kinase recruitment to the βAR complex could be a novel therapeutic strategy to normalize βAR function in human heart failure.