GRK2-Mediated Inhibition of Adrenergic and Dopaminergic Signaling in Right Ventricular Hypertrophy

Experimental Protocols

The University of Chicago Institutional Animal Care and Use Committee approved all protocols. Three RVH models were created in adult male Sprague-Dawley rats: (1) PAB-RVH, (2) PAH-RVH, induced by SU5416+chronic hypoxia (CH + SU) or Monocrotaline (MCT). End points were studied after 4 weeks in each model (n=8–13). The infusion of inotropes was primarily performed in additional cohorts of Control and MCT (n=7–12/group), although results were confirmed in CH + SU (n=5) and PAB (n=2). In additional cohorts, gallein (1.8 mg/kg/day, Tocris Bioscience, Ellisville, MO) was injected intraperitoneally for 2 weeks, beginning 2 weeks after Monocrotaline injection or PAB surgery (n=6–9).

Experimental Models

The PAB model has been described previously¹² (see online-only Data Supplement). In CH + SU model, rats (260–280 g) were injected with the VEGF receptor antagonist SU5416 (20 mg/kg, subcutaneously) and then transferred to hypoxic cages (≈10% oxygen, Biospherix, Lacon, NY) for 4-weeks. In the MCT model, rats (260–280 g) were injected with monocrotaline (60 mg/kg, subcutaneously; Sigma, St. Louis, MO).

Treadmill Distance

Exercise capacity was tested by measuring maximal distance run on a motorized treadmill, as described¹² (see online-only Data Supplement).

Echocardiography

A Vevo 2100 (Visual Sonics, Ontario, Canada) was used to assess CO, stroke volume (SV), and RV function, as described¹³ (see online-only Data Supplement).

Right Ventricular Hypertrophy

RVH was measured postmortem as the ratio of RV/(LV+septum) weight.

RV and LV Langendorff Models

The Langendorff models were performed as previously described¹² (see online-only Data Supplement).

Thermodilution Cardiac Output

Thermodilution CO was measured as previously described¹³ (see online-only Data Supplement).

Right Heart Catheterization With Infusion of Dopamine and Dobutamine

Rats were anesthetized (3% of isoflurane with 95% oxygen), intubated, and placed on a heated surgical table (37°C). A 1.9F pressure–volume catheter (Scisense Inc, London, Ontario, Canada) was inserted into RV via the right jugular vein to monitor the RV systolic pressure (RVSP) and volume. After stabilization, a pressure–volume signal was continuously recorded at sampling rate of 1000/s using an MPVS-300 (ADInstruments; Colorado Springs, CO) coupled to a PowerLab8/30 converter (ADInstruments). Dopamine or dobutamine was infused via the left jugular vein in 1 mL over 5-minute at clinically relevant doses¹⁴ (11 and 22 μg/kg/min), using a syringe pump (Cole-Parmer, Vernon Hills, IL). Heart rate (HR), RVSP, CO, ejection fraction, and SV were computed using pressure–volume analysis software (Labchart7.2; ADInstruments).

Laser Capture Microdissection

RV tissue was embedded in OCT compound (Sakura Finetek, Torrance, CA) and stored at −80°C. RV sections 7 μm thick were stained with the Arcturus HistoGene frozen-section staining kit (Applied Biosystems, Carlsbad, CA). RV myocytes and intramyocardial RV coronary arteries were harvested from fresh cryosections using the Palm MicroBeam system (Carl Zeiss, Thornwood, NY), and mRNA was isolated immediately. Each sample consisted of 20 to 30 RV myocytes or intramyocardial coronary arteries. Tissue localization of mRNA expression was guided visually and confirmed by enriched expression of tissue-specific reporters (α-myosin heavy chain for cardiomyocytes versus CD31 for vasculature).

qRT-PCR, Immunoblot, and Immunofluorescence

These techniques were performed as described previously (see online-only Data Supplement).¹²

Human Tissue Microarrays

Immunohistochemistry was performed on 1-mm RV cores from formalin-fixed, paraffin-embedded archival material of autopsied PAH patients (n=10) or age-matched non-PAH patients (n=8). Briefly, samples were arranged on a microscope slide, creating a tissue microarray (see text and Table I in the online-only Data Supplement for immunofluorescence technique and patient demographics). Consent was obtained for each autopsy, and the institutional review board was notified of the research performed on anonymized specimens from deceased individuals.

GRK2 Activity (Rhodopsin Phosphorylation) and cAMP levels (ELISA)

The heart was isolated from anesthetized rats, mounted in the Langendorff apparatus, and perfused with oxygenated Krebs–H-enseleit buffer for 10 minutes. Hearts were then perfused for 2 minutes with vehicle or dobutamine or dopamine (10 μmol/L each). The RV was isolated and stored at −80°C for measurement of cAMP levels and GRK2 activity. These assays were performed as previously described¹⁵ (see online-only Data Supplement).

Statistics and Sample Size

Values were expressed as mean±SEM. Sample sizes are shown in each figure. Prism 5 (GraphPad Software, La Jolla, CA) was used for data analysis. Comparisons between groups used an ANOVA or unpaired Student t test, as appropriate. Post hoc testing used a Bonferroni correction for multiple comparisons. If the test for normality failed or if the sample was <5, a Fisher exact test was used. A P<0.05 was considered statistically significant.

Depressed RV Contractile Reserve in RVH

RV mass was similarly increased in all 3 RVH models. In the RV Langendorff model, basal RVSP was greater in PAB than CH + SU or Monocrotaline (98 ± 26 versus 71 ± 7 and 59 ± 4 mm Hg, Figure 1). All RVH models had higher RVSP than control (29 ± 6 mm Hg, P<0.01). The inotropic reserve (whether defined as the fold- or absolute-increase in RVSP caused by 10 nmol/L dobutamine; Figure 1D and Figure I in the online-only Data Supplement) was lower in all RVH groups versus Control and was the lowest in the PAH-RVH models (Figure 1D).

Although basal LVSP remained unchanged in PAB, CH+SU and Monocrotaline (95±4, 96±2, 117±9 mm Hg) versus Control (105±5 mm Hg; Figure IA in the online-only Data Supplement), the LV inotropic reserve in response to dobutamine (10 nmol/L) was reduced in the PAH-RVH models (P>0.05; Figure IB in the online-only Data Supplement).

CO was reduced in PAB, CH+SU, and MCT (82.9±4.8, 71.5±8.0; 58.3±4.9 mL/min) versus control (154.2±16.2 mL/min; Figure 1E). Likewise, treadmill walking distance was significantly decreased in PAB, CH+SU, and MCT (244.1±12.3, 180.0±25.7, 41.5±11.6) versus control (406.3±54.4 m; Figure 1F). MCT rats had the lowest CO and the shortest treadmill distance.

Comparative Inotropic Potency

Whereas the inotropic response of the RV to dobutamine and dopamine was reduced in all RVH groups versus control (Figure 2A–2D), the inotropic reserve of LV was only depressed in the PAH-RVH models (Figure 2E–2H). The dose-response to dobutamine was left-shifted versus dopamine in both the LV and RV in all RVH models (Figure 2B–2D and 2F–2H), reflecting the superior potency of dobutamine. A similar maximal pressure could be obtained with dopamine (albeit at ≈1-log higher dose).

In the RV Langendorff, 10 nmol/L dobutamine, a dose in the clinically-relevant range, increased RVSP more than equimolar dopamine (Figure IIA in the online-only Data Supplement). Consistent with this, the EC50 of dobutamine was significantly lower than the EC50 of dopamine in all RVH groups (Figure IIB in the online-only Data Supplement). The EC50s were highest in PAH-RVH models, consistent with reduced potency. In the LV, the EC50 of dobutamine was also lower than that of dopamine, but it was unaltered by RVH (Figure IIC and IID in the online-only Data Supplement).

A dose–response curve to isoproterenol and phenylephrine in control and MCT (Figure III in the online-only Data Supplement) confirmed that contractile reserve to these inotropes was also reduced in RVH. The potency of inotropes in the RV Langendorff model in MCT hearts was as follows (in descending order): dobutamine=isoproterenol>dopamine>phenylephrine.

Dobutamine Versus Dopamine In Vivo

Baseline HR, CO, and SV were each reduced in MCT versus Control (HR, 294±5 versus 320±7 bpm, P<0.01; CO, 75±5 versus 110±5 mL/min, P<0.001; SV, 0.30±0.02 versus 0.37±0.03 mL, P<0.05; Figure 3A). In contrast, RVSP was increased in MCT versus control (RVSP, 61±3 versus 27±1 mm Hg, P<0.001). Dobutamine (22 μg/kg/min) caused a greater fold-increase in HR, CO, and SV than dopamine (22 μg/kg/min) in MTC (Fold increase: HR, 1.30±0.01 versus 1.22±0.05; CO, 1.7±0.1 versus 1.2±0.1; SV, 1.4±0.1 versus 1.2±0.1; Figure 3B–3D), suggesting greater efficacy of dobutamine versus dopamine. There were no differences in baseline HR, RVSP, CO, dP/dt, or SV in the 2 groups of MTC rats before infusion of dobutamine versus dopamine (Figure IVA–IVF in the online-only Data Supplement). Similar results, showing the superiority of dobutamine, were obtained at 11 μg/kg/min (data not shown). Detailed hemodynamics, obtained using an RV conductance catheter, are summarized in Table II (in the online-only Data Supplement). MTC rats had a markedly elevated RV end diastolic pressure versus control (13±2 versus 3±1 mm Hg; P<0.01). However, RV dP/dt was not different in MTC versus control (1221±72 versus 1557±134 mm Hg/s), and increased further with inotrope infusion (Table II in the online-only Data Supplement).

As in the MTC group, dobutamine increased CO and SV more than dopamine in the PAB and CH+SU groups, although some values were not statistically significant because of the smaller sample sizes (Figure V in the online-only Data Supplement).

The D1-R Antagonist SCH23390 Selectively Reduces the RV Inotropic Reserve in Response to Dopamine in RVH

SCH23390¹⁶ (10 μmol/L) did not change basal RV contractility or the contractile response to dobutamine in either control or MTC (P>0.05 versus baseline; Figure 4). However, SCH23390 inhibited dopamine-induced RV inotropy in MTC without altering the response of control RV to dopamine (Figure 4). These findings are consistent with a role for the D1-R in dopamine-induced RV inotropy in RVH.

Down-regulation of β1-AR, Dopamine Receptors, and α1-AR in RVH

RV β1-AR mRNA expression was significantly decreased in all RVH models versus control (Figure 5A), whereas RV β2-AR expression was only downregulated in CH + SU (Figure VIA in the online-only Data Supplement). LV β1-AR levels were unaltered in PAB and MTC (Figure 5B) but were decreased in CH+SU. LV β2-AR mRNA levels were unaltered in any RVH models (Figure VIF in the online-only Data Supplement). Laser capture microdissection confirmed that the down-regulation of β1-AR and D1-R occurred within RV myocytes, rather than intramyocardial coronary arteries (Figure 5C–5D and Figure VII in the online-only Data Supplement).

Qualitative immunofluorescence and quantitative immunoblot analysis of membrane proteins confirmed decreased β1-AR and D1-R expression in all RVH models (Figures 5E–5F and 6A–6B). Receptor downregulation was most profound in PAH-RVH models (Figure 5E–5F).

RV D2–5R mRNA expression was downregulated in CH+SU and MTC versus control but remained unchanged in PAB (Figure VIB–VIE in the online-only Data Supplement). In the LV, D2–5R mRNA was unaltered in any model (Figure 6G–6J in the online-only Data Supplement). RV α1-AR mRNA (α1a, α1b and α1 day) and protein (Figure VIII in the online-only Data Supplement) were decreased in all RVH groups.

Human RV Tissue Microarray

The expression of β1-AR was downregulated in the hypertrophied RV myocytes in RV tissue microarray specimens from PAH patients versus age- and sex-matched controls (Figure 6C and 6D and Table I in the online-only Data Supplement).

Uncoupling of Adrenergic Receptor Signaling in RVH

Basal RV intracellular cAMP production was similarly decreased in all RVH groups (Figure 7A). Increases in cAMP concentrations in response to dobutamine and dopamine (10 nmol/L) were reduced in MTC versus control (Figure 7B). Dobutamine caused significantly greater increases in cAMP than dopamine in control and MTC RVs, consistent with better coupling to adenylyl cyclase (Figure 7B).

RV GRK2 activity increased in all RVH groups (Figure 7C). In control, both inotropes increased GRK2 activity (Figure 7D), whereas in MTC, neither inotrope increased GRK2 activity beyond its already elevated basal level, suggesting maximal basal desensitization of the β1-AR signaling pathway in this model (Figure 7D).

Serial measurements revealed that RV GRK2 activity and expression increased by week 3 and 5 after monocrotaline injection, respectively (Figure IXA–IXC in the online-only Data Supplement). RV expression of β1-AR and D1-R mRNA began to decrease at week 3 and continued to decline thereafter (Figure IXE and IXF in the online-only Data Supplement). Thus, GRK2 activation and receptor downregulation occur concomitantly, consistent with the proposed role for GRK2 in adrenergic remodeling.

In Vivo Gallein Therapy

In PAB, 2 weeks of gallein treatment increased treadmill distance and cardiac index. Gallein caused a statistically insignificant trend toward increased tricuspid annular plane systolic excursion, a measure of RV function (P>0.05; Figure 8A–8D). In MTC, gallein significantly increased CO and tricuspid annular plane systolic excursion on echocardiography (P<0.01; Figure 8C and 8D); however, RV/LV+septum ratio, the catheterization and treadmill data, though trending toward benefit, were not statistically significant (P>0.05; Figure 8A and 8B and Figure XA and XB in the online-only Data Supplement). To confirm that gallein indeed had an effect on GRK2, we measured phospho-670 GRK2 and total GRK2 protein. RV expression of both proteins was reduced by gallein in both PAB and MTC (Figure 8E–8H and Figure XC–XF in the online-only Data Supplement). Phospho-ERK1/2 and total ERK1/2 expression, which were increased in MTC, were decreased by gallein (P<0.01, P<0.05 versus MTC; Figure 8I and 8J and Figure XG and XH in the online-only Data Supplement). Gallein treatment did not restore β1-AR protein expression in MTC (P>0.05 versus untreated MTC; Figure XIA and XIB in the online-only Data Supplement).

Molecular Basis and Chamber Specificity of Impaired Adrenergic and Dopaminergic Signaling in RVH

In the current study, both mRNA and protein expression of adrenergic and dopaminergic receptors were downregulated in all RVH models (Figures 5 and 6). Down-regulation and desensitization of adrenergic (β1- and α1-adrenoreceptors) and dopaminergic receptors (D1-R) primarily occurred in RV myocytes (Figures 5 and 6 and Figures VI and VIII in the online-only Data Supplement). Although it may seem intuitive that the down-regulation of β1-AR and D1-R would occur in RV myocytes, there are other candidate cells in the RV (vascular cells, fibroblasts, and inflammatory cells). Laser capture microdissection provided the precision to identify the affected cell type as the RV myocyte (Figure 5C and 5D and Figure VII in the online-only Data Supplement). Decreased expression of the β1-AR is also evident in the RV myocytes of humans with PAH (Figure 6C and 6D), confirming previous studies in rodents¹⁷ and human.¹⁸ The pathological relevance of this adrenergic remodeling is supported by the observation that the magnitude of the reduction in β1-AR expression is greatest in PAH-RVH models, and these models have the greatest impairment of contractile reserve (Figures 5 and 6).

Bristow et al¹⁸ examined the β-adrenergic system in patients with idiopathic PAH who either had RVF or were well compensated. Only those with RVF had decreased RV β-AR density. This PAH-RVF subset displayed depletion of norepinephrine and decreased adenylyl cyclase responsiveness to β-agonists, consistent with our findings in rats with PAH-RVH. Noontens et al¹⁹ demonstrated high circulating catecholamine levels in PAH-RVF patients, noting the loss of the normal ability to augment catecholamine levels with exercise. Likewise, in PAH-RVH rats, we found that inotropic stimulation could not further augment RV GRK2 activity beyond the elevated basal levels (Figure 7). Bristow et al reported that in humans with PAH, adrenergic impairment affected only the RV (not the LV).¹⁸ A similar finding was made in a canine RVF model (induced by severe PAB plus tricuspid insufficiency).²⁰ Consistent with these studies, we report that in PAB-RVH, adrenergic and dopaminergic receptor downregulation is confined to the RV and the depression of RV inotropic reserve is mild. However, in rodent PAH-RVH models, down-regulation of adrenergic and dopaminergic receptors was more severe and occurred in both the RV and LV, resulting in biventricular impairment of contractile reserve (Figure 2).

By evaluating the inotropic responses in both RV and LV Langendorff preparations we were able to assess contractility in each chamber, without the confounding effects of heart rate and afterload. This allowed us to address a longstanding controversy regarding the functional consequences of β1-AR down-regulation to LV function in PAH. The reduced LV response to dobutamine in MTC and CH+SU (Figure 2E–2H) confirms the physiological relevance of observed down-regulation and uncoupling of adrenergic receptors in the LV in these models. The finding of decreased β-AR expression in the LV in the PAH-RVH models is consistent with previous studies of monocrotaline rats.^17,21 LV involvement in PAH-RVH, but not PAB-RVH, likely reflects fundamental differences in how the models were created rather than differences in RVH severity. PAH-RVH models were induced by endothelial toxins (SU5416 or monocrotaline), which likely have vasculotoxic effects beyond the pulmonary circulation, per-haps including the RV vasculature. The concept that the stimuli that initiate PAH may also induce the RV coronary vascular injury is intriguing, particularly as many forms of PAH have evidence of systemic arterial endothelial damage. For example, in scleroderma-associated PAH, patients have circulating cytotoxic autoantibodies that are directed against epitopes in endothelial cells. Interestingly, compared with patients with idiopathic PAH, these scleroderma patients have more severe RV dysfunction,²² potentially suggesting an effect of the endothelial-targeted antibodies on the RV circulation.

In addition to β-1 and α-AR down-regulation, most dopamine receptor isoforms are also reduced in RVH (Figures 5 and 6 and Figure VI and VIII in the online-only Data Supplement). Dopamine receptors have been intensively studied in blood vessels and neural and renal tissues²³; however, only recently was the expression of D1, D2, D4, and D5 receptors reported in human hearts.²⁴ Their function remained uncertain, although Li et al showed that D1-R activation modulates ischemia/reperfusion-induced apoptosis in neonatal rat cardiomyocytes.²⁵ Dopamine receptors are pharmacologically classified into two classes: D1-like (which stimulate adenylyl cyclase) and D2-like (which inhibit adenylyl cyclase). The pattern of depressed dopamine receptor expression observed in RVH would be predicted to reduce D1-signaling and might contribute to the observed impairment in cAMP production. To test the relevance of D1-R down-regulation to impaired contractile reserve we administered a selective D1-R antagonist, SCH23390. This inhibitor selectively reduced dopamine contractility in RVH without affecting the response of dopamine in control hearts or altering the response to dobutamine (Figure 4). This reveals for the first time that D1-R contributes to dopamine-induced RV inotropy. Thus, the loss of dopamine receptors in experimental RVH likely contributed to the reduced contractile response to dopamine, both in vitro and in vivo (Figures 2 and 3 and Figures I, II, and IV in the online-only Data Supplement). As with the β1-AR, D1-R expression is most depressed in PAH-RVH models, although in contrast D1-R expression is not reduced in the LV in any model (Figures 5 and 6).

Optimal Inotropic Support in RVH

Dobutamine was shown to be more effective in increasing CO than dopamine in all RVH models, as a result of both superior coupling to adenylyl cyclase (as indicated by the greater evoked increase in cAMP) and the loss of D1-R expression in RVH, which exclusively impairs dopamine contractility (Figures 2–4 and 7). In RV and LV Langendorff models, the dose–response curves of dobutamine were left-shifted compared with those of dopamine (Figure 2). At equimolar doses (10 nmol/L), dobutamine induced greater contractility than dopamine in all RVH models (and even in the normal RV; Figure II in the online-only Data Supplement). These data indicate that dobutamine is superior to dopamine as an RV inotrope. Consistent with observations in left heart failure,²⁶ dobutamine caused a greater increase in CO (resulting from greater increases in SV and HR) than dopamine (Figure 3). However, dopamine is better at increasing dP/dt in vivo (Table II in the online-only Data Supplement). Ejection fraction is severely reduced in MTC-RVH, and dobutamine tends to increase ejection fraction more than dopamine (P>0.05; Table II in the online-only Data Supplement). Although dobutamine is a better inotrope than dopamine, this comes at some cost of increased heart rate (Figure 3). However, a significant portion of the net effect of dobutamine on CO is related to its ability to increase stroke volume. The greater increase in RV contractility in the RV Langendorff preparation also proves that increased heart rate is not the only reason for the superior augmentation of CO by dobutamine. The reduced RV response to dopamine reflects the combined consequence of alterations in adrenergic signaling in RVH, some of which uniquely or differentially impair the response to dopamine versus dobutamine. Impaired RV responses to dopamine in PAH-RVH reflect the combination of decreased β1-AR expression (which also effect dobutamine; Figures 5 and 6), inferior coupling of dopamine to adenylyl cyclase (less cAMP production; Figure 7), and the loss of D1-R–dependent contractility (which only affects dopamine; Figure 4). The greater elevation of cAMP concentrations in response to dobutamine versus dopamine, which we note, has also been reported in human heart failure.²⁷

Role of GRK2 in Mediating Receptor downregulation in RVH

We identified a key role for GRK2 in the adrenergic and dopaminergic receptor downregulation in RVH. GRK2 activity was markedly increased in RVH and was maximal in MTC (Figure 7 and Figure IV in the online-only Data Supplement), consistent with the severe RVF noted in this model (RVEDP 13±2 mmHg; Table II in the online-only Data Supplement). Our findings support those of Leineweber et al,²⁸ who reported increased GRK2 activity in monocrotaline-RVF.

The Reduced RV Response to Dopamine Reflects Therapeutic Value of Interrupting Gβγ–GRK2 Interaction In Vivo

We explored the therapeutic potential of correcting the adrenergic remodeling. One possible intervention would be the use of an inhibitor of β1-AR. The α- and β-AR inhibitor, carvedilol, improves LV function and survival in patients with LVF. Although β-AR antagonists are not approved for treatment of PAH, carvedilol improves RV function in the CH+SU5416 model.⁹ Because of the central role of GRK2 we chose to inhibit Gβγ–GRK2 interaction, using gallein, as an alternative to a β-AR antagonist. Gallein inhibits Gβγ subunit–dependent signaling and has been used for interrupting the interaction between GRK and the G/βγ subunit of activated G proteins in vitro.³ Gallein and other peptide Gβγ inhibitors have been shown to reduce GRK2 expression and improve cardiac function in experimental LVF.³ In the current study, gallein improved cardiac function, as evidenced by improved treadmill distance, tricuspid annular plane systolic excursion, and CO in both PAB-RVH and PAH-RVH (Figure 8). Consistent with its proposed mechanism of action, gallein decreased RV GRK2 expression. Further evidence that the beneficial effects of gallein related to its actions on the GRK2 pathway came from the demonstration that it decreased expression of activated (phosphorylated) ERK1/2, a kinase that regulates GRK2 activity (Figure 8I and 8J and Figure XG and XH in the online-only Data Supplement). The relationship between ERK1/2 and GRK is complex. Some studies suggest that ERK is upstream and phosphorylates β-arrestin and GRK,²⁹ whereas others place ERK downstream from GRK.³⁰

Limitations

We did not explore the mechanism for GRK2 activation. However, previous studies suggest that in states of hypertrophy and autonomic activation, protein kinase C phosphorylates and activates GRK2.¹⁵

There are some limitations to the gallein experiments. First, gallein undoubtedly has effects in addition to Gβγ inhibition. Second, at the doses used, gallein did not restore β1-AR protein expression, although it did inhibit the expression of GRK2. The regimen that restored β1-AR expression in LV failure (30 mg/kg/d for 3–4 weeks³) was more intense than what we used (1.8 mg/kg/d for 2 weeks). However, whereas gallein (0.1 μmol/L) acutely increases contractility in Control and PAB, it slightly decreased contractility in MTC (Figure XIC and XID in the online-only Data Supplement), suggesting lower doses may be required in PAH-RVH. Third, although gallein increased CO in PAB rats (a model devoid of pulmonary or systemic vascular disease), studies are needed to assess possible effects of gallein on the pulmonary and systemic vasculature.

Conclusion

GRK2-mediated adrenergic remodeling of the RV and LV contributes to impaired cardiac function in PAH-RVH. Acute RV inotropic support in PAH-RVH is best accomplished with dobutamine. Inhibition of Gβγ–GRK2 interaction may have promise as a therapy in RVH.