Determining the Absolute Requirement of G Protein-Coupled Receptor Kinase 5 for Pathological Cardiac Hypertrophy

Abstract

Rationale

Heart failure (HF) is often the end-phase of maladaptive cardiac hypertrophy. A contributing factor is activation of a hypertrophic gene expression program controlled by decreased class II histone deacetylase (HDAC) transcriptional repression via HDAC phosphorylation. Cardiac-specific overexpression of G protein-coupled receptor kinase-5 (GRK5), has previously been shown to possess nuclear activity as a HDAC5 kinase, promoting an intolerance to in vivo ventricular pressure-overload, however, its endogenous requirement in adaptive and maladaptive hypertrophy remains unknown.

Objective

We used mouse models with global or cardiomyocyte-specific GRK5 gene deletion to determine the absolute requirement of endogenous GRK5 for cardiac hypertrophy and HF development following chronic hypertrophic stimuli.

Methods and Results

Mice with global deletion of GRK5 were subjected to transverse aortic constriction (TAC). At 12 weeks, these mice showed attenuated hypertrophy, remodeling, and hypertrophic gene transcription along with preserved cardiac function. Global GRK5 deletion also diminished hypertrophy and related gene expression due to chronic phenylephrine infusion. We then generated mice with conditional, cardiac-specific deletion of GRK5 that also demonstrated similar protection from pathological cardiac hypertrophy and HF following TAC.

Conclusions

These results define myocyte GRK5 as a critical regulator of pathological cardiac growth following ventricular pressure-overload, supporting its role as an endogenous (patho)-physiological HDAC kinase. Further, these results define GRK5 as a potential therapeutic target to limit HF development after hypertrophic stress.

INTRODUCTION

Heart failure (HF), a leading cause of death in the western world, often occurs as an end-phase of pathological myocardial hypertrophy¹. Hypertrophy is initially an adaptive response to stresses ranging from hypertension, valve disease, or cardiac injury^{2, 3}. In an attempt to normalize wall stress, cardiomyocytes enlarge, sarcomeres reorganize, fibroblasts proliferate, and hypertrophic genes, including the so-called fetal gene program, are up-regulated. If prolonged, these responses lead to chamber dilation, myocardial apoptosis, and HF^{2, 3}.

In pathological cardiac growth, the molecular pathways affecting transcription lie downstream of the nodal hypertrophic signal transducer, Gq¹. Among these complex pathways are the class II histone deacetylases (HDACs). Physiologically opposed to cardiac growth, the HDACs repress expression of key hypertrophic genes, primarily through inhibition of myocyte enhancer factor 2 (MEF2)³. Genetic deletion of the class II HDAC, HDAC5, sensitizes mice to cardiac stress, while murine deletion of MEF2 confers cardioprotection, decreasing pathological hypertrophy, and attenuating up-regulation of the fetal gene program⁴. HDAC kinases control nuclear HDAC activity as phosphorylation of HDAC5 induces its nuclear export and MEF2 derepression^3–5.

We recently identified G protein-coupled receptor (GPCR) kinase-5 (GRK5) as a nuclear HDAC kinase, joining Protein Kinase D (PKD) and Calmodulin-dependent kinase-II (CaMKII) as HDAC-mediated facilitators of cardiac growth after hypertrophic stimuli^5–7. This represents a novel, non-GPCR cardiac role for this GRK. We showed that mice with cardiac overexpression of GRK5 demonstrate Gq-dependent nuclear translocation of GRK5 where it can phosphorylate HDAC5 and induce MEF2 activity⁶. Moreover, cardiac GRK5-overexpressing mice displayed an intolerance to ventricular pressure-overload with potentiated maladaptive hypertrophy and accelerated HF following transverse aortic constriction (TAC)⁶. Importantly, these transgenic studies did not address whether endogenous GRK5 plays a role in the hypertrophic response.

In this study we used global and cardiomyocyte-specific GRK5 knockout (KO) mice to directly address the importance of endogenous GRK5 in cardiac hypertrophy. We found that GRK5 in cardiomyocytes is absolutely required for hypertrophic responses after stress. Further, our data indicate that limiting GRK5 expression in the heart can protect against maladaptive cardiac growth and HF development.

METHODS

Generation of GRK5cKO mice

Detailed mouse protocols are described in Supporting Information (SI) Materials and Methods.

Surgical Procedures and echocardiography

TAC methodology, minipump implantation and determination of in vivo cardiac function and morphology are described in SI Materials and Methods

RNA analysis

Methods for RNA analysis are described in SI Materials and Methods.

Immunoblot analysis

Detailed methods for heart subfractionation and immunoblotting are described in SI Materials and Methods.

RESULTS

Global GRK5 deletion diminishes in vivo cardiac hypertrophy

We subjected male global GRK5 gene knockout mice (GRK5gKO)⁸ and littermate WT control mice to TAC. Constitutive GRK5 deletion attenuated pressure-overload-induced cardiac growth seen in WT mice at 12-weeks post-TAC (Fig. 1A). Cardiac dimensions were measured serially by echocardiography to track development of hypertrophy and LV dilatation over 12 weeks. WT mice showed a quick rise in LV posterior wall thickness (LVPWT) with a peak thickness of 1.99±0.05mm at 4-weeks post-TAC, which then decreased rapidly, indicative of adverse remodeling (Fig 1B). Interestingly, global GRK5 deletion significantly delays the initiation of cardiac hypertrophy following TAC as GRK5gKO mice do reach a similar LVPWT (1.89±0.09mm) as WT mice but not until 12 weeks post-TAC (Fig. 1B). At the end of 12 weeks, GRK5gKO mice had significantly less cardiac hypertrophy as determined by smaller heart weight-to-body weight (HW/BW) ratios (Fig. 1C). Importantly, GRK5gKO mice show no signs of LV dilatation via echocardiograph measurements of systolic LV interior diameter (LVIDs), which were significantly increased in post-TAC WT mice (Fig. 1D and online Fig. I). Further, post-TAC GRK5gKO mice showed preserved cardiac function as determined by ejection fraction (%EF) compared to post-TAC WT mice which had significant LV dysfunction (Fig. 1E and online Fig. I).

Fig. 1. Attenuated hypertrophy seen in GRK5gKO mice post-TAC.

(A) Hearts from WT and GRKgKO mice subjected to a sham operation or TAC (Top). Histological sections stained with Masson’s Trichome for fibrosis (Middle; Bottom). WT Sham n=6, WT TAC n=9, gKO Sham = 6, gKO TAC = 11. (B) Systolic LVPWT measured serially by echocardiogram following sham or TAC operations. *p<0.05 (C) HW/BW ratios were measured 12-weeks post-TAC. *p<0.01. (D-E) LVIDs (D) or %EF (E) as measured by echocardiogram at 12-weeks post-TAC. *p<0.01. (F-G) RT-PCR was used to measure mRNA expression of known markers of cardiac hypertrophy (F), and genes directly regulated by MEF2 (G) n=8, *p<0.05.

We have previously shown that elevated GRK5 plays a critical role in myocytes as an HDAC5 kinase, derepressing hypertrophy-related transcription⁶. Up-regulation of generalized hypertrophy markers, including those of the fetal gene program - atrial natriuretic factor (ANF), β-myosin heavy chain (βMHC), and procollagen, type Iα2 (Col1a2) – were significantly attenuated in the GRK5gKO mice compared WT mice 12-weeks post-TAC (Fig. 1F). Loss of GRK5 expression globally also prevented the post-TAC up-regulation of hypertrophic genes directly regulated by MEF2⁹: brain natriuretic peptide (BNP), Actinα1 (Acta-1) and Connective Tissue Growth Factor (CTGF) (Fig. 1G). Thus, deletion of GRK5 is protective against hypertrophy at the molecular level, consistent with the in vivo phenotype.

GRK5gKO mice are resistant to phenylephrine-dependent hypertrophy

Phenylephrine (PE), acting through α-adrenergic receptors (αARs), induces cardiomyocyte hypertrophy in vitro and in vivo. Previously, we have found that PE causes GRK5 nuclear translocation and increased MEF2 activity in myocytes⁶. Therefore, we tested whether endogenous GRK5 was necessary for development of PE-induced cardiac hypertrophy. Male mice (WT and GRK5gKO) were treated with a subpressor dose of PE (35mg/kg/day) or phospho-buffered saline (PBS) for 14 days via osmotic minipumps (Fig. 2A). This period covers only an initial hypertrophy stage without decompensation thereby testing the requirement of endogenous GRK5 in myocardial αARs-mediated hypertrophy. PE treatment caused cardiac growth in WT mice with a significant 26.9±8% increase in HW/BW ratio at 2-weeks while GRK5gKO mice subjected to PE had only an 11.2±7% increase in HW/BW, insignificant compared to PBS-treated GRK5gKO mice (Fig. 2B). Two weeks of PE treatment does not change cardiac function, which is what we found for %EF (Fig. 2C), although PE significantly altered morphology in WT mice (Fig. 2D, 2E). Importantly, PE-treated GRK5gKO mice showed no changes in cardiac dimensions (Fig. 2D, 2E).

Fig. 2. PE induces less hypertrophy in GRK5gKO mice.

(A) Hearts from WT and GRK5 gKO mice subjected to 2 weeks of chronic infusion of PBS or subpressor PE (35mg/kg/day) (Top). Histological sections stained with Masson’s Trichome for fibrosis (Middle; Bottom). WT PBS n=8, WT PE n=10, gKO PBS = 9, gKO PE = 11. (B) HW/BW ratios for WT and GRK5gKO mice were measured after 2 weeks of chronic PBS or PE infusion. *p<0.05. (C-E) %EF (C), LVIDs (D), and LVPWT (E) as measured by echocardiogram at 2 weeks. *p<0.05. (F-G) RT-PCR was used to measure mRNA expression of known markers of cardiac hypertrophy (F) or MEF-2 regulated genes (G) n=8, *p<0.05.

As with post-TAC, GRK5 deletion significantly decreased PE-mediated up-regulation of hypertrophy markers. Indeed, while PE lead to robustly increased expression of our panel of hypertrophic genes in WT mice, no such up-regulation was seen in GRK5gKO mice (Fig. 2F, 2G). Overall, these data demonstrate that GRK5 plays a key role in cardiac hypertrophy downstream of PE, identical to post-TAC phenotype.

Global GRK5 ablation decreases nuclear HDAC5 export following a hypertrophic stimulus

The diminished up-regulation of genetic hypertrophy markers seen in GRK5gKO mice post-TAC suggests that GRK5 can regulate cardiac gene transcription. This is likely due to GRK5’s ability to phosphorylate HDAC5, inducing its export from the nucleus⁶. Hence, we examined location and phosphorylation of HDAC5 following hypertrophic stress in GRK5gKO and WT mice. Hearts from the above experiments were subjected to subcellular fractionation. Following TAC, WT mice showed a significantly increased amount of phosphorylated HDAC5 in the non-nuclear subcellular fraction compared to GRK5gKO TAC mice and WT sham mice (WT TAC: 1.94±0.13; GRK5gKO TAC 0.91±0.14, WT sham 1.02±0.17) (Fig. 3A, B). Similar results were seen in the non-nuclear fraction from WT mice that had received chronic infusion of PE. These mice showed significantly greater non-nuclear phosphorylated HDAC5 than PE-infused GRK5gKO mice and PBS-infused WT mice (Fig. 3C), further reinforcing the role of endogenous GRK5 as an HDAC5 kinase.

Fig. 3. Global GRK5 ablation decreases phosphorylated HDAC5 in the cytoplasm.

(A) Twelve weeks following TAC or a sham operation, hearts from WT and GRK5gKO mice were fractionated into non-nuclear and nuclear fractions. HDAC5 was immunoprecipitated from the non-nuclear fraction and immunoblotted for phosphorylated and total amounts (pHDAC5 and tHDAC5, respectively). (B) Denistometric quantification of pHDAC5 normalized to tHDAC. *p<0.01 vs. all groups, n=6. (C) Denistometric quantification of pHDAC5 normalized to tHDAC5 following immunoprecipitation and immunoblotting for pHDAC5 in WT and GRK5gKO mouse hearts following 2 weeks of chronic PBS or PE infusion. *p<0.01 vs. all groups, n=6.

Cardiac-specific deletion of GRK5 attenuates hypertrophy following TAC

The above results show that a complete GRK5 ablation attenuates the cardiac hypertrophic response but don’t address the specific role of myocyte GRK5 in maladaptation and post-TAC HF. Therefore, we developed conditional GRK5KO mice where GRK5 deletion was cardiac-specific. We bred floxedGRK5 mice with transgenic mice expressing Cre-Recombinase under control of the αMHC promoter¹⁰. These conditional GRK5 KO (GRK5cKO) mice had greater than 50% loss of cardiac GRK5 determined by either protein immunoblotting or RT-PCR (Online Fig. II). As above, we stressed GRK5cKO mice and WT control mice (GRK5floxed) via TAC and studied these groups alongside Sham-operated mice for 12 weeks (Fig. 4A). Serial echocardiography showed significantly attenuated hypertrophy in GRK5cKO mice with maximally increased cardiac mass at 12-weeks when WT mice are clearly decompensated following their peak hypertrophy 4-weeks post-TAC (Fig. 4B). This delayed compensatory hypertrophy led to a trend towards lower HW/BW ratio in GRK5cKO mice at 12-weeks post-TAC compared to WT mice (Fig. 4C). This effect is less robust than the GRK5gKO mice but no doubt due to incomplete GRK5 ablation. However, from in vivo functional studies, it is clear that the myocyte GRK5 loss protects these hearts from adverse LV remodeling and HF as GRK5cKO mice displayed no increased LV dilatation at the study’s end, compared to significantly increased LVIDs in WT mice (Fig. 4D, Online Fig. SIII). This protection against HF was also evident in global in vivo cardiac function as 12-week post-TAC WT mice had a significant loss of LV %EF compared to sham WT mice, while there was absolutely no drop in %EF in 12-week post-TAC GRK5cKO mice (Fig. 4E).

Fig. 4. Attenuated hypertrophy seen in GRK5cKO mice post-TAC.

(A) Hearts from WT and GRK5cKO mice subjected to a sham operation or TAC (Top). Histological sections stained with Masson’s Trichome for fibrosis (Middle; Bottom). WT Sham n=9, WT TAC n=14, cKO Sham = 8, cKO TAC = 9. (B) Systolic LVPWT measured serially by echocardiogram following sham or TAC operations. *p<0.05 (C) HW/BW ratios for WT and GRK5cKO mice were measured 12 weeks post-TAC. (D-E) LVIDs (D) or %EF (E) as measured by echocardiogram at 12 weeks post-TAC. *p<0.01. (F-G) RT-PCR was used to measure mRNA expression of known markers of cardiac hypertrophy (F), and MEF2-regulated genes (G) n=8, *p<0.05.

Again, we examined our panel of hypertrophy-related genes. At 12-weeks post-TAC, hearts from GRK5cKO mice showed significantly diminished up-regulation of these common cardiac hypertrophy markers, including those of the fetal gene program (Fig. 4F), and specific MEF2-regulated genes (Fig. 4G). Overall, these data demonstrate that GRK5 expression in cardiomyocytes alone is required for WT molecular, functional and morphological responses post-TAC.

DISCUSSION

Our results indicate an absolute requirement of cardiomyocyte GRK5 for normal hypertrophic responses and that this kinase plays a critical pathological role in ventricular decompensation and transition to HF following ventricular pressure-overload. Importantly, merely decreasing cardiomyocyte GRK5 in mice blunts hypertrophic myocardial growth and prevents HF post-TAC. Importantly, these data demonstrate that endogenous myocyte GRK5 plays a crucial role in adaptive and maladaptive hypertrophy and is required for WT response to stress. Thus, increased GRK5 in the failing human heart¹¹ has pathological significance since lowering GRK5 or inhibiting its activity appears to offer novel beneficial effects against maladaptive cardiac growth.

These results in global and cardiac-specific GRK5KO mice, coupled with our previous results showing that GPCR-independent nuclear activity of GRK5 can facilitate hypertrophy⁶, indicate that nuclear targeting of endogenous GRK5 as a class II HDAC kinase is physiologically significant in normal and abnormal cardiac growth after stress. Interestingly, loss of nuclear GRK5 activity can delay HF onset through slower cardiac growth, although its deletion does not completely ameliorate hypertrophy. This may be due to compensatory and discrete roles of the other known HDAC5 kinases, CAMKII and PKD, acting downstream of hypertrophic signaling^{5, 7}. Each kinase has been shown to cause HDAC5 nuclear export following select receptor activation such as endothelin-1 receptor for CAMKII, and αAR for PKD^{5, 12}. We now can add GRK5 to the list of physiological HDAC kinases downstream of TAC and αAR stimulation. These distinct activators of HDAC kinases appear to underlie a complex network of parallel signaling converging on the same target, HDAC5. Overall, our data demonstrates that targeting only one HDAC kinase can delay HF and may advantageous in therapeutic intervention by allowing some signaling.

Since GRK5 plays a dual role in the cardiomyocyte – membrane-associated GPCR desensitizing kinase and nuclear kinase facilitating transcription -- the most critical activities for cardiac signaling and function are uncertain. Classical GRK5 activity towards β-adrenergic receptors (βARs) has been shown to induce trans-activation of the cardioprotective epidermal growth factor receptor¹³. A human polymorphism of GRK5 (Q41L) appears to increase βAR desensitization, protecting some HF patients chronically¹⁴. Additionally, transgenic overexpression of Gαq caused a slight dilatation in GRK5gKO mice compared to WT controls¹⁵. Therefore, there is some question whether the increased GRK5 in is protective or injurious.

In this study, simply decreasing myocyte GRK5, either completely, in GRK5gKO mice, or significantly, in GRK5cKO mice, attenuated cardiac hypertrophy and prevented pathogenesis of HF. We show here that GRK5 ablation does not completely prevent hypertrophy but significantly delays it.Therefore, despite potential beneficial activities at the plasma membrane, removing cardiomyocyte GRK5 has a profound positive effect on outcomes after pressure-overload and chronic α-adrenergic stress. Clearly, our data now indicates that GRK5, and, most likely, its nuclear HDAC kinase activity, represent a novel target to prevent maladaptive cardiac hypertrophy and protect against ventricular decompensation and HF.

Novelty and Significance.

What Is Known?

• G protein-coupled receptor kinase 5 (GRK5) is up-regulated in numerous models of heart failure (HF), as well as in the failing human heart.

• GRK5 enters the nucleus and acts as a Histone Deacetylase 5 (HDAC5) kinase, increasing transcription of cardiac hypertrophy genes.

• Increased nuclear GRK5 is pathological in the setting of chronic pressure overload.

What New Information Does This Article Contribute?

• Ablation of GRK5 significantly delays maladaptive cardiac remodeling and HF following chronic pressure overload or α-adrenergic receptor (αAR) stimulation.

• Removing nuclear GRK5 by global ablation decreases HDAC5 export.

• Deletion of GRK5 in cardiomyocytes alone significantly delays the onset of HF.

Pathological cardiac hypertrophy, a process commonly ending in HF, occurs through activation of nodal signal transducer Gq. Downstream of Gq, class II HDACs represses hypertrophic gene transcription. GRK5, a recently identified HDAC kinase, has been shown to be up-regulated in human HF, although the role of endogenous GRK5 in maladaptive cardiac remodeling is unknown. We investigated the role of endogenous GRK5 in maladaptive cardiac remodeling. Our results show that global and cardiomyocyte-specific ablation of GRK5 significantly attenuates pathological cardiac hypertrophy, delaying HF onset. This cardioprotection may be attributed to decreased nuclear HDAC5 export following GRK5 deletion. Overall, this study suggests that endogenous nuclear GRK5 plays an injurious role in maladaptive hypertrophy and may represent a novel therapeutic target.

Non-Standard Abbreviations

HF

Heart Failure

HDAC

Histone Deacetylase

GRK

G protein-coupled Receptor Kinase

TAC

Transverse Aortic Constriction

MEF2

Myocyte Enhancer Factor 2

GPCR

G protein-coupled Receptor

PKD

Protein Kinase D

CaMKII

Calmodulin-dependent Kinase II

LVPWT

Left Ventricular Posterior Wall Thickness

HW/BW

Heart Weight-to-Body Weight Ratio

LVIDs

Systolic Left Ventricular Internal Diameter

%EF

Ejection Fraction

PE

Phenylephrine

AR

Adrenergic Receptor

PBS

Phospho-buffered Saline