Therapeutic Efficacy of a Novel Pharmacologic GRK2 Inhibitor in Multiple Animal Models of Heart Failure
Original Research - Preclinical
Visual Abstract
Highlights
| • | Inhibition of GRK2 is an area of interest for the development of a potential new class of HF drugs that is underexplored compared with other HF drug classes. | ||||
| • | We demonstrate that a novel GRK2 inhibitor derived from paroxetine, CCG258208, can improve LV contractile function and limit adverse remodeling in 2 clinically relevant mouse models of HF as well as act therapeutically acutely in a post-MI chronic HF model in mini-swine. | ||||
| • | As a GRK2 inhibitor, CCG258208 can potentially be considered a lead molecule for further evaluation for HF translation in future preclinical and clinical studies. | ||||
Summary
GRK2 is the most prominent G protein-coupled receptor kinase that is upregulated in heart failure (HF), and inhibiting GRK2 has improved cardiac function in mice. CCG258208, generated from the paroxetine scaffold, which has GRK2 inhibitory properties, has a 50-fold higher selectivity for GRK2 at 100-fold lower doses. We evaluated CCG258208 in 2 mice HF models and found that CCG258208 has robust therapeutic effects. In a chronic mini-swine HF model, acute administration of CCG258208 enhanced dobutamine inotropic responses. Our results indicate that CCG258208 has robust cardioprotective and HF-reversing effects in different HF models and it stands as a promising lead for HF therapy.
Introduction
Despite pharmacological and surgical advancements, heart failure (HF) continues to be on the rise and in the United States alone—about 6.2 million adults have HF.1 Although the development and use of emerging drug classes such as vasodilators,2 diuretics,3 angiotensin-converting enzyme inhibitors,4 β-blockers,4 and the most recent sodium-glucose cotransporter 2 inhibitors5 have improved outcomes, there is still a need for targeting critical regulatory cell signaling components in the failing myocardium. One underexplored component in HF is G protein-coupled receptor kinases (GRKs).6 In the heart, GRKs function by phosphorylating activated G protein–coupled receptors (GPCRs) such as β-adrenergic receptors (β-ARs), leading to the recruitment of β-arrestin and clathrin-mediated receptor internalization.7 Among GRKs, GRK2 is abundantly expressed in the heart and plays a crucial role in the pathogenesis of HF. In response to the increased sympathetic drive postinjury, GRK2 is up-regulated and remains elevated in mouse as well as human HF samples.6,7
Because cardiac function is tightly regulated by catecholamine-mediated stimulation of β-ARs and GRK2 plays an important role in their desensitization, GRK2 is an ideal target for therapeutic intervention. Targeting GRK2 in HF has been previously reported in mouse models by expression of βARKct, a C-terminal fragment of GRK2 that blocks Gβγ-mediated GRK2 membrane translocation, which is required for GRK2 phosphorylation of agonist-occupied GPCRs. βARKct expression in the heart has resulted in reduced deterioration of cardiac function and prolonged survival in HF mice.8-10 In addition, different levels of myocardial βARKct expression have shown that a dose-dependent decrease in GRK2 activity positively correlates with an increase in cardiac contractile function in a mouse HF model.8 Further, Casey et al11 have recently demonstrated that the small molecules M119 and Gallein halt HF progression and improve cardiac function in vitro and in vivo likely by competing with GRK2 for Gβγ in cardiomyocytes. Thereafter, Thal et al12 discovered that paroxetine, a selective serotonin reuptake inhibitor (SSRI), binds to GRK2 and inhibits its catalytic activity at its active site. In vitro, paroxetine increased contractility in isolated adult ventricular myocytes treated with isoproterenol.12 Our lab also reported that mice in HF as a result of myocardial infarction (MI) had a significant reversal of declining cardiac function and limited adverse left ventricular (LV) remodeling after treatment with paroxetine.13 This study also highlighted that GRK2 inhibition with paroxetine conferred additional benefits compared with using a β-blocker (metroprolol) alone, and paroxetine even reversed HF in mice with forced overexpression of myocyte GRK2.13
To further increase the potency of GRK2 inhibition while still retaining the selectivity for GRK2, Waldschmidt et al14 generated a library of compounds based on the paroxetine scaffold. From this library of compounds, one such derivative, CCG258208 (referred to as 14as in that paper), was found to have at least 50-fold higher potency for GRK2 inhibition.14 In vitro, this compound increased mouse cardiomyocyte contractility in response to isoproterenol stimulation similar to that of paroxetine but at a 100-fold lower dose.14 Such profound effects of CCG258208 led us to investigate the role of this compound on GRK2 inhibition in AC16 cells in vitro; in vivo in 2 different mouse models of HF, post-MI, and post–transverse aortic constriction (TAC); and on cardiac inotropic reserve with an acute dose, in mini-swine with chronic post-MI HF, and finally evaluate its overall toxicity profile.
Methods
A detailed description of methods is available in the Supplemental Appendix.
Cyclic adenosine monophosphate assay
AC16 cells were used to perform a cyclic adenosine monophosphate (cAMP) assay, to determine β-AR stimulation in the presence and absence of GRK2 kinase activity using 1 μM paroxetine and 0.1 μM CCG258208.
Pharmacokinetics
The mouse pharmacokinetics study protocol was approved by Purdue Institutional Animal Care and Use Committee (PACUC) (Approval No. 1112000342 expires January 16, 2025) and conforms to National Institutes of Health animal care guidelines. The pharmacokinetics of CCG258208 was evaluated following peroral administration in female C57BL/6 mice (8-12 weeks old, ∼20 g body weight) at dose 20 mg/kg of mouse weight. Mice were humanely sacrificed at 1, 2, 4, and 7 hours after oral administration or 30 minutes after intraperitoneal injection. As a control for the peroral route of administration, the same compound dose was administered by intraperitoneal injection and then the samples were collected 30 minutes after injection. Organs and blood specimens were collected from euthanized mice and analyzed for the CCG258208 content by liquid chromatography/mass spectrometry. Liquid chromatography/mass spectrometry analysis was done using an Agilent 1200 Rapid Resolution liquid chromatography system coupled to an Agilent 6470 series QQQ mass spectrometer. Agilent Mass Hunter Quantitative Analysis version 10.1 was used for all data analysis.
Mouse models of HF
All animal procedures were approved by the Institutional Animal Care and Use Committee of Temple University and were performed in accordance with the National Institutes of Health guidelines for the care and use of laboratory animals. We first performed a dose-response study of CCG258208 in a post-MI HF model in mice to demonstrate pharmacologic efficacy and found an optimal dose to use for subsequent studies. Two weeks after MI (or sham), mice were administered either paroxetine, vehicle, or CCG258208 at 3 different doses: high (2 mg/kg/day), medium (0.5 mg/kg/day), and low (0.1 mg/kg/day). Subsequently, the high dosage of CCG258208 was used in the mouse model of post-TAC HF. In the TAC model, fluoxetine and paroxetine were used at 5 mg/kg/day alongside CCG258208 at 2 mg/kg/day.
Mouse echocardiography
Cardiac function was measured at baseline and 2, 4, and 6 weeks post-MI and at baseline and 2, 4, 6, 8, 10, and 12 weeks post-TAC by transthoracic echocardiography using an MS400 linear array transducer attached to Fujifilm VisualSonics Vevo 2100. All echocardiography was done blinded and data were analyzed using VevoLAB software (version 1.7.0) (Fujifilm).
Mouse tissue collection and tissue processing
For histology, hearts were perfused with KCl to arrest them in diastole, followed by perfusion with phosphate-buffered saline and 4% paraformaldehyde. Heart, lungs, and tibia bones were dissected, and weights and lengths were recorded.
Mouse tissue histology
Paraffin-embedded post-TAC and post-MI heart samples were sectioned at a plane in which all 4 cardiac chambers and the aortic outflow tract were visible. Sections were stained with Masson’s trichrome, wheat germ agglutinin, and CD31.
Mini-swine model of HF and hemodynamic analysis
MI was induced in Göttingen mini-swine 9 to 12 months of age. Baseline measurements were collected before any drug administration. After baseline measurements of LV pressure, 10 mg/kg/min of dobutamine was administered continuously using an Alaris PC Guardian Infusion System Model 8110 series (Cardinal Health). Hemodynamic measurements were taken 5 minutes after every dobutamine, CCG258208 (2 mg/kg), and fluoxetine (5 mg/kg) administration to measure the inotropic reserve after GRK2 inhibition. Differences between the 2 dobutamine responses (after 2 infusions of CCG258208/fluoxetine) were quantified.
Statistical analysis
All data are presented as mean ± SEM. For measures such as histological and gravimetric data, a 1-way analysis of variance was used to compare group differences. For repeated measures such as echocardiographic parameters post-TAC, linear mixed-effects models were used to determine predicted mean values at each assessment point (baseline and 2, 4, 6, 8, 10, and 12 weeks post-TAC) and to test group differences at a given time point. In each linear mixed-effects model, time was included as a fixed effect. Statistical analyses were performed using SAS 9.4 (SAS Institute). Two-way analysis of variance followed by Tukey's multiple comparisons test was used for the cell culture cAMP assay (GraphPad Prism; GraphPad Software; version 10.2.3 [347]). A P value of <0.05 was used to define the significance of all statistical tests.
Results
CCG258208 inhibits GRK2 kinase activity
β-AR stimulation and subsequent cAMP production are typically increased with GRK2 inhibition. Isoproterenol caused an increase in cAMP levels in untreated cells (negative control), in cells treated with paroxetine (positive control), and in cells treated with CCG258208 (Figure 1A). However, the difference in the increased levels of cAMP between the untreated control cells and paroxetine and CCG258208 were significantly different (Figure 1B). No differences were observed between cAMP levels in cells treated with 1 μM paroxetine and 0.1 μM CCG258208. This indicates that CCG258208 is a more potent GRK2 inhibitor than paroxetine, as it has identical GRK2 inhibitory properties at a 10-fold lower concentration.
cAMP Levels in Cells With Control, Paroxetine, and CCG258208 Pretreatment
(A) Cyclic adenosine monophosphate (cAMP) readouts in untreated and isoproterenol-treated cells. n = 3. #P < 0.001 compared with samples without isoproterenol. (B) Differences in cAMP readouts between control cells and cells pretreated with paroxetine and CCG258208. Data are presented as mean ± SEM, n = 3, and data are compared using 2-way analysis of variance followed by Tukey's multiple comparisons test. ∗∗P = 0.002 compared with control condition, ∗∗∗P < 0.001 compared with control condition. w = with; w/o = without.
CCG258208 distributes to the plasma and heart but not to the brain
The pharmacokinetics of CCG258208 was first evaluated to be sure that it would hit the target organs, namely the heart and plasma in this study. It was also desirable that the compound would not cross the blood-brain barrier like the parent compound paroxetine so that serotonin-related effects could largely be avoided. Peroral administration demonstrated that up to 3 hours, concentrations of the compound remained above its half-maximal inhibitory concentration in plasma, heart, and liver (Figure 2A). The amount found in the brain was negligible, even when administered via intraperitoneal injection (Figure 2B). Importantly, heart tissue seems to be one of the major sites of CCG258208 distribution with either route of administration, supporting the subsequent mouse studies.
Time Course of CCG258208 Distribution in Mouse Plasma and Organs
(A) Distribution after peroral administration. Inhibitory concentrations (>30 nM) of the compound remain in both plasma and heart at 10× above IC50 level after 2 hours and at IC50 level after 4 hours. (B) Distribution 30 minutes after intraperitoneal administration. In either route of administration, no compound reaches the brain, eliminating serotonin reuptake inhibition associated off-target effects in that tissue. Data are presented as mean ± SD.
Effect of CCG258208 treatment on LV contractile function in post-MI HF mice
Contractile function sharply declined in response to MI injury after 2 weeks and before any treatment. As observed from echocardiography data, the 3 doses of CCG258208 had a dose-dependent effect on LV contractile function reflected by changes in ejection fraction (EF) and fractional shortening (FS) over the 4-week treatment (Figures 3B and 3C). Of the 3 doses evaluated, the high dose of CCG258208 (2 mg/kg/day) showed the most improvement in contractile function compared with the vehicle-treated group at the study endpoint, as evidenced by higher EF% and FS%, and this was indistinguishable from the effects of paroxetine (Figures 3B and 3C). Importantly, the medium dose of CCG258208 (0.5 mg/kg/day) also significantly improved cardiac function compared with the vehicle-treated group (Figures 3B and 3C). The low dose of CCG258208 (0.1 mg/kg/day) did not provide any significant improvement in cardiac contractility (Figures 3B and 3C). Supplemental Table 1 provides individual group functional measurements, remodeling values, detailed group comparisons, and statistical significance values.
Dose-Dependent Effects of CCG258208 on Cardiac Function in Post-MI Mice
Cardiac function was assessed in all mice by serial echocardiography using M-mode tracing. (A) Experimental design in which there was an even distribution of mice randomly assigned to all treatment and vehicle groups based on left ventricular (LV) ejection fraction and fractional shortening measurements at 2 weeks post–myocardial infarction (MI) to ensure similar functional starting points. Treatment groups were CCG258208 high (2 mg/kg/day), medium (0.5 mg/kg/day), and low (0.1 mg/kg/day); paroxetine (5 mg/kg/day); vehicle; and a vehicle sham group (n = 6-17 per group). (B) LV ejection fraction; (C) LV fractional shortening; (D) calculated LV mass; (E) left ventricular inner diameter at systole (LVID;s); (F) LV volume at systole (Vol;s). ∗P < 0.05 compared with paroxetine, ∗∗∗P < 0.001 compared with paroxetine, ˆˆP < 0.01 compared with vehicle, ˆˆˆP < 0.001 compared with vehicle. (B-F) High-dose CCG258208 was not significantly different from paroxetine. (D-F) Medium-dose CCG258208 was not significantly different from paroxetine. Data are presented as mean ± SEM. Linear mixed-effects models were used to determine predicted mean values at each assessment point and to test group differences at a given time point. In each linear mixed-effects model, time was included as a fixed effect. Vehicle sham: n = 17; vehicle: n = 13; CCG258208 high: n = 12; CCG258208 medium: n = 13; CCG258208 low: n = 6; paroxetine: n = 11. echo = echocardiography.
Consistent with changes in contractile function, LV dilation, and adverse remodeling were most attenuated in post-MI mice treated with the high dose of CCG258208 compared with the vehicle group, although the medium dose of CCG258208 also significantly reduced cardiac hypertrophy as measured by LV mass after 4 weeks of treatment (Figure 3D). The increase in LV chamber dimensions post-MI and the subsequent effect of CCG258208 treatment was also reflected in LV inner diameter at systole (LVID;s) (Figure 3E). The high dose–treated hearts had the smallest LVID;s followed by the medium dose–treated hearts, with both of these doses being significantly therapeutic approaching statistical significance compared with the vehicle group. Changes in EF, FS, and LVID;s also positively correlated with LV volume at systole (Figure 3F). The high-dose CCG258208-treated hearts had the smallest residual LV volumes at systole compared with the vehicle group. The medium dose of CCG258208 was also able to significantly reduce residual LV blood volume compared with the vehicle, and the volumes were similar to those measured with paroxetine treatment (Figure 3F). Consistent with other readouts, low-dose CCG258208 was unable to significantly reduce residual LV volume at systole compared with the vehicle group (Figure 3F). This indicates that compromised contractile function led to insufficient blood being pumped out of the LV, resulting in a higher residual LV blood volume. Treating mice with high and medium doses of CCG258208 prevented this cascade of pathological events. Overall, CCG258208 had robust therapeutic effects on post-MI contractile and reverse remodeling effects at doses lower than paroxetine, consistent with its higher potency against GRK2.14 No significant changes in survival rate were observed with any of the 3 compounds post-MI (Supplemental Table 2). While a higher mortality rate is associated with MI surgery, reduced post-MI mortality in our study is most likely due to a different less invasive, nonventilated surgical technique (Supplemental Ref. 1).
Effect of CCG258208 treatment on heart and lung weight in post-MI HF mice
Gravimetric analysis indicated that all post-MI hearts were larger compared with sham hearts as a result of LV dilation and compensatory LV hypertrophy (Figures 4A and 4B). Consistent with LV mass data from echocardiography analysis, mice treated with the medium and high doses of CCG258208 had significantly lower heart weight (HW)-to-body weight (BW) ratios compared with mice treated with vehicle (Figure 4A). Low-dose CCG258208 did not reduce MI-associated increases in HW compared with vehicle (Figure 4A). Because the length of the tibia bone is a more reliable indicator of age-appropriate growth than an increase in body weight, which can vary, HWs were also normalized to tibia lengths (TLs) to eliminate differences in BW gain throughout the study, and both the high and medium doses of CCG258208 significantly reduced the post-MI increases in HW/TL ratios similar to the levels seen with paroxetine (Figure 4B). Similar to the HW/BW ratios, low-dose CCG258208 did not reduce the HW/TL ratio compared with vehicle (Figure 4B). CCG258208 did not have an adverse effect on mice growth over time (Supplemental Figure 2). An increase in lung weight (LW) has been attributed to both fluid congestion and lung fibrosis, and vascular remodeling seen during HF.15 Although assessing specific pulmonary changes was beyond the scope of this study, medium and high doses of CCG258208 were also able to rescue the increase in post-MI LW/BW ratios (Figure 4C). The ability of the high-dose CCG258208 treatment (similar to paroxetine) to lower LW/BW ratio to near sham levels indicates that GRK2 inhibition does play a role in modulating cardiopulmonary circulation post-MI. Supplemental Table 3 contains specific values for the gravimetric and statistical data of CCG258208 compared with vehicle- and paroxetine-treated mice.
Gravimetric Analysis of Healthy and Post-MI Hearts Treated With CCG258208
Heart weight (HW) was normalized to (A) body weight (BW) and (B) tibia length (TL) to account for changes in growth rate in mice during the 6-week study period. (C) An increase in lung weight (LW) after myocardial infarction (MI) injury was also recorded and normalized to body weight. CCG258208 high was 2 mg/kg/day, CCG258208 medium was 0.5 mg/kg/day, and CCG258208 low was 0.1 mg/kg/day; n = 6 to 17 per group. Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗∗P < 0.01 compared with vehicle, ∗∗∗P < 0.001 compared with vehicle. n.s. = not significant.
Effect of CCG258208 treatment on cardiac fibrosis and cellular hypertrophy post-MI
For histological evaluation of post-MI hearts, we limited our studies to the vehicle-treated (sham and MI) and paroxetine and high-dose CCG258208 groups. Consistent with cardiac physiology results, histology results also indicated that CCG258208-treated hearts had a significantly lower degree of fibrosis and LV dilation, resulting in a smaller LV scar size and dilation compared with the vehicle-treated post-MI mice, and were similar to paroxetine treatment (Figure 5). This was evident in whole sections (Figures 5A to 5D) as well as quantification of the LV circumference and LV infarct length (Figures 5E and 5F). The smaller infarct length (Figure 5F) at the study endpoint as a result of GRK2 inhibition is most likely due to restricted expansion of the infarcted myocardium over time, which typically occurs in an MI injury model. Cardiomyocyte cross-sectional area was significantly lower in CCG258208-treated mice compared with vehicle-treated mice (Figure 6). Reduction in cellular hypertrophy was also similar between CCG258208 and paroxetine-treated hearts (Figure 6). Specific values from all histological measurements for LV scar size and myocyte cross-sectional area are provided in Supplemental Tables 4 and 5, respectively.
Effect of CCG258208 Treatment on LV Circumference at 6 Weeks Post-MI
Mice treated with high-dose CCG258208 had smaller infarcts and were less dilated compared with vehicle-treated mice, reflected by a smaller LV circumference at 6 weeks post-MI. Representative images of (A) sham vehicle–treated heart, (B) vehicle-treated post-MI heart, (C) CCG258208 high dose–treated post-MI heart, (D) paroxetine-treated post-MI heart. Magnification ×0.8; scale bar = 2,000 μm. (E) Quantification and group-wise comparisons of LV circumference. (F) Quantification and group-wise comparisons of infarct length. n = 3 to 4 per group. Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗∗P < 0.01 compared with vehicle. Abbreviations as in Figure 3.
Effect of CCG258208 Treatment on Post-MI Cardiomyocyte Cellular Hypertrophy
Six weeks after myocardial infarction (MI) (4 weeks after drug treatment), hearts were fixed, sectioned, and stained with wheat germ agglutinin for measurements of myocyte cross-sectional area (CSA) as an assessment of cellular hypertrophy. Shown are representative sections and quantification. (A) Vehicle sham, (B) post-MI vehicle, (C) post-MI CCG258208 high-dose treatment, and (D) post-MI paroxetine treatment (×25.2 magnification; scale bar = 20 μm). (E) Quantification and group-wise comparisons (n = 3-4 per group). Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗P < 0.05 compared with vehicle. n.s. = not significant.
Changes in contractile function and adverse remodeling post-TAC after treatment with CCG258208
From the previous post-MI study, we concluded that the 2-mg/kg/day dose of CCG258208 was the most effective in HF and accordingly used that dose to evaluate this novel GRK2 inhibitor in a post-TAC model of HF due to LV pressure overload. Because TAC is a different form of HF than MI and takes more time for severe LV dysfunction, we allowed decompensation to occur for 6 weeks after injury and then treated mice from 6 to 10 weeks post-TAC with CCG258208 and paroxetine. Fluoxetine was used as a control against potential SSRI activity interference (Figure 7A). We followed the function of all mice via echocardiography for 12 weeks. Interestingly, all mice had a slow decline in cardiac function as determined by EF and FS from weeks 0 to 6, but mice treated with CCG258208 and paroxetine showed marked and significant improvement in contractile function beginning at 8 weeks that persisted throughout the rest of the study compared with fluoxetine-treated mice (Figures 7B and 7C). The continued therapeutic effect even 2 weeks after paroxetine and CCG258208 were stopped is consistent with what we previously found with paroxetine in the mouse MI model of HF.13 Importantly, CCG258208 or paroxetine did not affect contractile function in sham animals (Figures 7B and 7C). A sham group with fluoxetine was not included in the current study, as our lab has previously shown that fluoxetine does not affect cardiac function in sham animals.13 Exact values for all echocardiographic parameters described previously and subsequently are provided in Supplemental Table 6.
Effect of CCG258208 on Cardiac Function in Heart Failure Mice after TAC
All measurements were made from serial M-mode transthoracic echocardiography. Two-dimensional images were acquired before (at baseline and 2, 4, and 6 weeks) and after (at 8, 10, and 12 weeks) treatment with CCG258208 or control groups. (A) Study design in which mice were treated from 6 weeks post–transverse aortic constriction (TAC) to 10 weeks post-TAC with CCG258208 (2 mg/kg/day), paroxetine (5 mg/kg/day), and fluoxetine 5 mg/kg/day, along with 2 sham groups (paroxetine and CCG258208 treated). (B) LV fractional shortening; (C) LV ejection fraction; (D) LV mass; (E) LVID;s; (F) LV diameter at systole; (G) LV volume at systole. Data are presented as mean ± SEM. Linear mixed-effects models were used to determine predicted mean values at each assessment point and to test group differences at a given time point. In each linear mixed-effects model, time was included as a fixed effect. ∗∗P < 0.01 compared with fluoxetine, ∗∗∗P < 0.001 compared with fluoxetine. There was no difference between CCG258208 and paroxetine in any of the measured functional parameters. Sham paroxetine: n = 5; sham CCG258208: n = 5; TAC CCG258208: n = 8; TAC paroxetine: n = 8; TAC fluoxetine: n = 10. Abbreviations as in Figure 3.
As expected, TAC led to increased LV mass compared with sham hearts (Figure 7D). This increase in hypertrophy was consistent with the decline in LV contractile function. At 6 weeks post-TAC, when mice were treated with paroxetine, CCG258208, or fluoxetine, hearts in all groups had a similar degree of hypertrophy, as indicated by similar LV mass (Figure 7D). The consistent increase in hypertrophy continued up to 8 weeks post-TAC. However, at 8 weeks and beyond, LV mass in mice treated with fluoxetine continued to increase, while with paroxetine or CCG258208, hypertrophy was arrested. At the study endpoint (12 weeks post-TAC), CCG258208-treated mice had significantly smaller hearts than fluoxetine-treated mice, and these were similar to paroxetine-treated mice (Figure 7D). This was also reflected by gravimetric data, as HW/BW ratios and HW/TL ratios at 12 weeks post-TAC showed CCG258208 treated mice to have significantly smaller hearts than fluoxetine-treated mice, with a similar degree of attenuated hypertrophy as paroxetine treated (Figures 8A and 8B). CCG258208 did not have an adverse effect on mice growth over time (Supplemental Figure 3). Specific gravimetric data are found in Supplemental Table 8.
Effect of CCG258208 on Cardiac Hypertrophy Post-TAC
Following 12 weeks post-TAC, mice were sacrificed and HW was compared with BW and TL. Shown are the quantification of (A) HW/BW ratios and (B) HW/TL ratios. n = 5 to 10 per group. Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗∗P < 0.01 compared with TAC fluoxetine. Abbreviations as in Figures 4 and 7.
Differences in hypertrophic responses from LV mass readouts were also reflected in LV chamber dimensions and consecutive LV volume measurements. Mice treated with CCG258208 had a smaller LV inner diameter at systole (Figure 7E) and an overall smaller LV diameter at systole (Figure 7F), which included the LV inner cavity and the anterior and posterior walls, compared with fluoxetine-treated hearts. These measurements in mice treated with CCG258208 were similar and nondistinct from those treated with paroxetine (Figures 7E and 7F, Supplemental Table 6). The larger LV inner diameter in fluoxetine-treated mice also resulted in the hearts having a higher residual LV blood volume at systole compared with the CCG258208 group (Figure 7G). Treatment of sham mice with CCG258208 did not induce LV structural or functional changes in hearts throughout the 12-week experimental time (Figure 7). No significant difference in survival rate was observed between the 3 compounds post-TAC (Supplemental Table 7).
Changes in cardiomyocyte growth with CCG258208 post-TAC
In our TAC model of pressure-overload hypertrophy and HF, an increase in myocyte cell size was observed in all 3 post-TAC groups at 12 weeks as determined by wheat germ agglutinin staining of cardiac sections (Figure 9). In line with LV mass data obtained from echocardiography and gravimetric data found at dissection, CCG258208-treated mouse hearts had a significantly smaller cross-sectional area of myocytes compared with myocytes from post-TAC fluoxetine-treated mice (Figure 9B). This was evident in all chambers of the heart measured including the LV, right ventricle, and septum. Importantly, myocyte sizes from CCG258208-treated mice post-TAC were not different from paroxetine-treated mice. Specific group-wise comparisons are outlined in Supplemental Table 9.
Effect of CCG258208 Treatment on Post-TAC Cardiomyocyte Cross-Sectional Area
Twelve weeks after TAC, hearts were fixed, sectioned, and stained with wheat germ agglutinin for measurements of myocyte CSA as an assessment of cellular hypertrophy. Shown are representative sections and quantification. (A) Representative images of stained heart sections from TAC and sham hearts treated with paroxetine, CCG258208, or fluoxetine. (B) Quantification of myocytes (in arbitrary units [AU]) from groups from left ventricular, right ventricular, and septum areas of the hearts. Red indicates wheat germ agglutinin, blue indicates DAPI. Image magnification ×20, n = 3 per group. Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗P < 0.05 compared with TAC fluoxetine. Abbreviations as in Figures 6 and 7.
Effect of CCG258208 treatment on cardiac microvasculature and fibrosis post-TAC
Besides ischemic conditions in the heart, GRK2 has also been shown to be upregulated in skeletal muscles in rat models of hind limb ischemia and in restricted endothelial cell function in vitro.16 In the TAC model of HF, although the degree of myocardial ischemia is not comparable to hind limb ischemia or MI, an increase in maladaptive cardiomyocyte hypertrophy has been shown to positively correlate with a decrease in microvessel density resulting in a predisposition to myocardial ischemia.17 Indeed, in our study, we found that after 12 weeks of TAC, cardiac hypertrophy was associated with a decrease in microvessel density, and importantly, CCG258208-treated mice had hearts with significantly increased microvessel density compared with fluoxetine-treated hearts in the LV, septum, and right ventricle (Figures 10A and 10B). Similar to changes in hypertrophy, CCG258208 inhibition of GRK2 was most pronounced in the LV and effects were similar to paroxetine-treated mice. Specific values are outlined in Supplemental Table 10.
Effect of CCG258208 Treatment on Left Ventricular Microvasculature 12 Weeks Post-TAC
Twelve weeks after transverse aortic constriction (TAC), hearts were fixed, sectioned, and stained with anti-CD31, an endothelial marker as a measure of altered microvasculature after cardiac injury and heart failure. Shown are representative sections and quantification. (A) Representative images of stained heart sections from TAC and sham hearts treated with paroxetine, CCG258208, or fluoxetine (as labeled). (B) Quantification of microvasculature in the left ventricle, septum, and right ventricle. Red indicates CD31, Blue indicates DAPI, Image magnification ×40, n = 3 per group. Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗P < 0.05 compared with TAC fluoxetine. HPF = high-power field.
Although there is minimal cellular loss post-TAC, compared with post-MI, TAC injury also leads to the deposition of excess extracellular matrix leading to the formation of interstitial and perivascular fibrotic tissue in the myocardium. The presence of interstitial fibrosis leads to the stiffening of the ventricular wall and over time ultimately affects contractile function. In line with echocardiography data, histological staining indicated that hearts from CCG258208-treated mice had significantly less interstitial fibrosis compared with hearts from fluoxetine-treated mice, and this decreased fibrosis was similar to that seen in mice treated with paroxetine (Figures 11A-11C). The extent of perivascular fibrosis was also less in CCG258208-treated hearts compared with fluoxetine, although this was not statistically significant (Figures 11D and 11E). Specific values are outlined in Supplemental Tables 11 and 12.
Effect of CCG258208 Treatment on Cardiac Fibrosis Post-TAC
Twelve weeks after transverse aortic constriction (TAC), hearts were fixed, sectioned, and stained with Masson’s trichrome to assess and quantitate interstitial and perivascular fibrotic areas of the post-TAC hearts. (A) Representative whole-mount sections of 12 weeks post-TAC hearts from groups treated with fluoxetine, paroxetine, or CCG258208. Image magnification ×0.8; scale bar = 2,000 μm. (B) Representative sections showing interstitial fibrosis. Image magnification ×20; scale bar = 20 μm. (C, E) Quantification of groups for interstitial and perivascular fibrosis (n = 3 each), Data are presented as mean ± SEM, and 1-way analysis of variance was used to compare group differences. ∗P < 0.05 compared with fluoxetine. (D) Representative sections showing perivascular fibrosis around the left ventricular vasculature. Image magnification ×20; scale bar = 20 μm. A.U. = arbitrary units; n.s. = not significant.
Effect of β-adrenergic responsiveness with CCG258208 in a swine model of ischemic HF
In this model, mini-swine were 3 months post-MI and in HF. We measured hemodynamics and initial dobutamine responses followed by a second challenge with this β-AR agonist after pretreatment with 2 boluses of CCG258208 or fluoxetine as a control condition (see Methods and schematic in Figure 12A). We found that treatment of these HF pigs acutely with CCG258208 significantly increased dobutamine responsiveness to systolic pressure and inotropic reserve compared with fluoxetine pretreatment (Figures 12B and 12C). Exact values are outlined in Supplemental Table 13. The percentage change in end-systolic pressure between the first and second dobutamine response to CCG258208 administration was higher compared with fluoxetine (Figure 12B). This improved adrenergic response to CCG258208 was also reflected in a significantly higher percentage change in +dP/dtmax between the first and second dobutamine response in CCG258208-treated swine compared with fluoxetine-treated swine (Figure 12C). Although the response to the dobutamine challenge was assessed using acute administration of 2 boluses of either CCG258208 or fluoxetine, it does indicate that in a clinically relevant large animal model of MI, used at the same dose as in mice (2 mg/kg), CCG258208 was able to acutely inhibit GRK2 function sufficiently that resulted in significantly improved adrenergic responses.
Effect of Acute CCG258208 Administration on Systolic Pressure and β-Adrenergic Inotropic Reserve in Mini-Pigs With Chronic Heart Failure
(A) Göttingen mini-swine that had chronic heart failure 3 months after MI was studied experimentally as shown with a baseline hemodynamic assessment (see Methods for details), first dobutamine infusion (10 mg/kg/min) followed by 2 boluses of CCG258208 (2 mg/kg each bolus) or fluoxetine (5 mg/kg each bolus) (n = 3 each), and a second dobutamine infusion. ˆ indicates all points in which pressure recordings were made. (B) Change in end-systolic pressure (ESP) in CCG258208 and fluoxetine groups between the first and second dobutamine administrations. (C) LV +dP/dtmax % change in CCG258208 and fluoxetine-treated pigs between the first and second dobutamine administration. Data are presented as mean ± SEM. ∗P < 0.05 vs fluoxetine-treated swine. Abbreviations as in Figure 3.
Discussion
In this study, we sought to determine if CCG258208, a novel small molecule selective inhibitor of GRK2, is effective in treating HF in 2 different clinically relevant postinjury models in mice and to test its acute effect on β-adrenergic responsiveness in a mini-swine model of chronic HF. Indeed, we found this was robustly therapeutic in the 2 small animal models of HF, and CCG258208 was also able to improve inotropic reserve compared with control conditions in a chronic post-MI mini-swine model of HF. Further, in the post-MI HF model in mice, CCG258208 displayed a dose-dependent therapeutic response showing its future pharmacologic potential. The highest dose utilized (2 mg/kg/day) was equivalent to the therapeutic effect achieved by 5 mg/kg/day of paroxetine in both models (post-MI and post-TAC) of murine HF. Because CCG258208 retains paroxetine’s GRK2 selectivity (up to 60-fold over other GRKs) but has a higher potency for GRK2 inhibition,14 CCG258208 can be used at lower doses and represents a lead compound for future development as a selective GRK2 inhibitor and a first in class drug molecule.
Our study led to the following important findings: 1) despite being a paroxetine derivative, CCG258208 does not cross the blood-brain barrier, eliminating SSRI-related adverse effects; 2) GRK2 inhibition by CCG258208 affects not only contractile function, but also hypertrophy progression and adverse LV chamber dilation over time; 3) CCG258208-mediated GRK2 inhibition is similar to paroxetine-mediated GRK2 inhibition, as they produce near-identical effects; 4) improvement in contractile function and attenuation of adverse remodeling, including fibrosis, mediated by GRK2 inhibition via CCG258208 treatment is relevant in 2 fundamentally different mice HF models; and 5) CCG258208 did not cause any noticeable adverse effects after 4 weeks of treatment in sham or HF mice, reflecting its safety. Multiorgan toxicity profiling further reflected its overall safety profile.
The underlying mechanism by which paroxetine inhibits GRK2 activity is by binding to and reorganizing its catalytic site.12 This is also true for CCG258208 based on crystallographic analysis.14 Inhibiting the catalytic activity of GRK2 prevents it from phosphorylating GPCRs and thereby preventing desensitization. It is unclear whether it also prevents the GRK2-Gβγ complex from binding to GPCRs, but if it does not, it could also block heterotrimeric G protein coupling in a dominant negative manner. In the failing heart, the most critical GPCR supporting contractile function is the β-AR, and CCG258208 action was directly demonstrated to improve signaling and functionality through the β-AR in the pig model of HF in our study. Besides this canonical function, GRK2 has an interactome that includes noncanonical functions that are GPCR independent.18 The potential mechanisms by which GRK2 inhibition by CCG258208 mediates HF reversal and cardioprotection could therefore be a combination of both canonical and noncanonical functions.
Some of the noncanonical functions of systemic GRK2 inhibition by CCG258208 that might potentially play a role in cardioprotection are: 1) cardiac metabolism involving improved glucose uptake and insulin sensitivity19,20; 2) mitochondrial translocation of GRK2 postinjury; and 3) attenuation of GRK2-mediated catecholamine release from the adrenal medulla.21
The novel aspects of our study are: 1) biodistribution assessment of CCG258208 after oral administration, which is more relevant from a clinical standpoint; and 2) data obtained in a preclinical, large animal model in which GRK2 inhibition via acute CCG258208 administration was demonstrated in mini-swine in HF 3 months after MI. We found increased systolic pressure and a significant improvement in β-AR inotropic reserve that was not evident with fluoxetine treatment, showing that this acute contractility benefit was due to the GRK2 inhibition properties of CCG258208 and not any SSRI effects that the agent may have. A chronic HF effect of CCG258208 in a large animal model is certainly warranted and will be the next step in a potential clinical usage evaluation.
Interestingly, based on our previous success with reversing HF in mice with paroxetine, 2 small clinical studies in human patients have occurred. Results of the CARE-AMI (Paroxetine-Mediated GRK2 Inhibition to Reduce Cardiac Remodeling After Acute Myocardial Infarction) trial22 indicated that there were no differences in contractile function between the paroxetine-treated and control group. However, in another trial paroxetine treatment in patients experiencing depression along with hypertension showed attenuated cardiac hypertrophy.23 Because the dose range of paroxetine used for its SSRI-related function ranges from 12.5 to 75 mg/day in patients, a dose of 20 mg/day (used in the CARE-AMI trial) might not have been sufficient for non–SSRI-related cardiovascular benefits. Selection of a paroxetine dose to achieve clinically relevant cardiovascular benefits without undesired SSRI activity would be challenging, a bottleneck that can be resolved by using CCG258208, because of its biodistribution profile.
In evaluating CCG258208 from a potential HF drug standpoint, its effects on HR, blood pressure, and renal function will be important parameters to consider and evaluate. In hemodynamic assessments in mice, acute administration (1 hour before assessment) of paroxetine led to an initial increase in HR.12 However, sustained administration of paroxetine13 and CCG258208 for 4 weeks did not affect HR in our mouse injury models. Paroxetine has previously been evaluated in patients experiencing ischemic heart diseases accompanied by depression, to evaluate its cardiovascular safety profile. The study concluded that paroxetine did not affect HR, heart rhythm, blood pressure, or cardiac conduction.24 In a separate study administering paroxetine to chronic kidney disease patients and end-stage renal disease patients accompanied by depression, this also did not result in adverse effects, indicating its renal safety profile.25 Although structurally and functionally similar, drawing parallels between the clinical safety profile of paroxetine and CCG258208 is only speculative at this point and requires clinical evaluation.
Study limitations
One of the limitations of this study is that mini-swine in HF could not be chronically treated with CCG258208 similar to mice in HF because of a lack of a sufficient amount of synthesized CCG258208. Only acute effects of CCG258208 could be assessed. Another limitation of the study is that although HF is a common cause of mortality in all age groups, it is particularly high in aged individuals. Because of the limited availability of the CCG258208 compound, we were unable to determine if CCG258208 confers the same degree of therapeutic effect in reversing HF in aged HF mice as in the young adult HF mice used in the present study. We also did not address gender-specific differences in HF onset and progression in adults, but these are beyond the scope of this study and are obvious future extensions for study. For statistical analysis, adjustments for multiple testing were performed in some experiments and omitted in others. However, we do not expect the group-wise differences or overall conclusions to change in the experiments in which the adjustment was omitted.
Conclusions
Over the last few decades, in vitro and in vivo studies have suggested a pivotal role of GRK2 inhibition in modulating different aspects of cardiac dysfunction.7,8,13,26,27 Within the scope of our experimental results and previous data generated by our lab, we believe that CCG258208 is a viable candidate to move forward to final preclinical studies and perhaps phase 1 studies. Importantly, CCG258208 has been derived from and is similar to paroxetine in structure, has strong cardioprotective and therapeutic effects, and paves the way for GRK2 inhibitors as a new class of drugs that can be targeted to ameliorate the detrimental effects of GRK2 upregulation in pathological conditions including HF.
Perspectives
COMPETENCY IN MEDICAL KNOWLEDGE: Inhibition of GRK2 as a prospective avenue for drug targeting for HF has not been adequately explored as compared with other currently clinically prescribed HF drugs, although our previous study using paroxetine in the post-MI HF model showed advantages of GRK2 inhibition over β-AR blockade.13 Interestingly, this is gaining traction with clinical trials investigating the cardioprotective effects of paroxetine in HF patients. To that end, CCG258208 is a much more potent inhibitor of GRK2 at a much lower dose than paroxetine.
TRANSLATIONAL OUTLOOK: The 2 mouse models of HF used in this study to assess the efficacy of CCG258208 adequately represent clinical situations (MI indicates major coronary artery blockage, TAC indicates aortic stenosis, aortic coarctation). Being structurally similar to paroxetine, which has no serious adverse effects that outweigh the beneficial effects, the possibility of CCG258208 having serious adverse effects in patients is low, although this is speculative at this time. This along with the efficacy of CCG258208 in 2 clinically relevant models of HF, and the robust effects seen acutely in the pig model of HF, make it a candidate that can potentially be considered as a new class of HF drug for evaluation in future clinical trials.
Funding Support and Author Disclosures
This work was funded by National Institutes of Health grants R01 HL061690 and R01 HL157151 (to Dr Koch), AHA Merit Award (to Dr Koch) R01 HL071818 (to Drs Koch and Tesmer), and National Institutes of Health RO1 grants CA254402 and CA221289 (to Dr Tesmer). Funding for pharmacokinetics was provided by the PIDD (Purdue) FY20 Drug Evaluation Committee Project “Selective Inhibition of G Protein-Coupled Receptor Kinases for the Treatment of Heart Failure.” The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Abbreviations and Acronyms
| AR | adrenergic receptor |
| BW | body weight |
| cAMP | cyclic adenosine monophosphate |
| EF | ejection fraction |
| FS | fractional shortening |
| GPCR | G protein–coupled receptor |
| GRK | G protein-coupled receptor kinase |
| HF | heart failure |
| HW | heart weight |
| LV | left ventricle/ventricular |
| LVID;s | left ventricular inner diameter at systole |
| LW | lung weight |
| MI | myocardial infarction |
| SSRI | selective serotonin reuptake inhibitor |
| TAC | transverse aortic constriction |
| TL | tibia length |
References
1. "Heart Disease and Stroke Statistics-2020 Update: a report from the American Heart Association". Circulation . 2020;141:e139-e596.
2. "Effect of vasodilator therapy on mortality in chronic congestive heart failure. Results of a Veterans Administration Cooperative Study". N Engl J Med . 1986;314:1547-1552.
3. "Diuretic therapy in heart failure - current approaches". Eur Cardiol . 2015;10:42-47.
4. "15 years of heart-failure trials: what have we learned?"Lancet . 1998;352:Suppl 1: SI29-SI33.
5. "Effect of empagliflozin on the clinical stability of patients with heart failure and a reduced ejection fraction: the EMPEROR-Reduced trial". Circulation . 2021;143:326-336.
6. "G protein-coupled receptor kinases as therapeutic targets in the heart". Nat Rev Cardiol . 2019;16:612-622.
7. "Cardiac function in mice overexpressing the beta-adrenergic receptor kinase or a beta ARK inhibitor". Science . 1995;268:1350-1353.
8. "Level of beta-adrenergic receptor kinase 1 inhibition determines degree of cardiac dysfunction after chronic pressure overload-induced heart failure". Circulation . 2005;111:591-597.
9. "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;98:5809-5814.
10. "Gene-mediated inhibition of the b-adrenergic receptor kinase: a new therapeutic strategy for heart failure". Minerva Cardioangiol . 2001;49:389-394.
11. "Small molecule disruption of G beta gamma signaling inhibits the progression of heart failure". Circ Res . 2010;107:532-539.
12. "Paroxetine is a direct inhibitor of g protein-coupled receptor kinase 2 and increases myocardial contractility". ACS Chem Biol . 2012;7:1830-1839.
13. "Paroxetine-mediated GRK2 inhibition reverses cardiac dysfunction and remodeling after myocardial infarction". Sci Transl Med . 2015;7:277ra31.
14. "Structure-based design of highly selective and potent G protein-coupled receptor kinase 2 inhibitors based on paroxetine". J Med Chem . 2017;60:3052-3069.
15. "Left ventricular failure produces profound lung remodeling and pulmonary hypertension in mice: heart failure causes severe lung disease". Hypertension . 2012;59:1170-1178.
16. "β adrenergic receptor kinase C-terminal peptide gene-therapy improves β2-adrenergic receptor-dependent neoangiogenesis after hindlimb ischemia". J Pharmacol Exp Ther . 2016;356:503-513.
17. "Alterations of the architecture of subendocardial arterioles in patients with hypertrophic cardiomyopathy and impaired coronary vasodilator reserve: a possible cause for myocardial ischemia". J Am Coll Cardiol . 1998;31:1089-1096.
18. "Noncanonical roles of G protein-coupled receptor kinases in cardiovascular signaling". J Cardiovasc Pharmacol . 2017;70:129-141.
19. "G protein-coupled receptor kinase 2 activity impairs cardiac glucose uptake and promotes insulin resistance after myocardial ischemia". Circulation . 2011;123:1953-1962.
20. "Antidiabetic and cardioprotective effects of pharmacological inhibition of GRK2 in db/db mice". Int J Mol Sci . 2019;20:1492.
21. "Reduction of sympathetic activity via adrenal-targeted GRK2 gene deletion attenuates heart failure progression and improves cardiac function after myocardial infarction". J Biol Chem . 2010;285:16378-16386.
22. "Paroxetine-mediated G-protein receptor kinase 2 inhibition in patients with acute anterior myocardial infarction: final 1-year outcomes of the randomized CARE-AMI trial". J Am Heart Assoc . 2022;11:e026362.
23. "Paroxetine attenuates cardiac hypertrophy via blocking GRK2 and ADRB1 interaction in hypertension". J Am Heart Assoc . 2021;10:e016364.
24. "Comparison of paroxetine and nortriptyline in depressed patients with ischemic heart disease". JAMA . 1998;279:287-291.
25. "A practical approach to the treatment of depression in patients with chronic kidney disease and end-stage renal disease". Kidney Int . 2012;81:247-255.
26. "Expression of a beta-adrenergic receptor kinase 1 inhibitor prevents the development of myocardial failure in gene-targeted mice". Proc Natl Acad Sci U S A . 1998;95:7000-7005.
27. "Cardiac fibroblast GRK2 deletion enhances contractility and remodeling following ischemia/reperfusion injury". Circ Res . 2016;119:1116-1127.
Appendix
Supplemental MaterialFootnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
