Essential Role of the RIα Subunit of cAMP-Dependent Protein Kinase in Regulating Cardiac Contractility and Heart Failure Development

Abstract

Background: The heart expresses 2 main subtypes of cAMP-dependent protein kinase (PKA; type I and II) that differ in their regulatory subunits, RIα and RIIα. Embryonic lethality of RIα knockout mice limits the current understanding of type I PKA function in the myocardium. The objective of this study was to test the role of RIα in adult heart contractility and pathological remodeling.

Methods: We measured PKA subunit expression in human heart and developed a conditional mouse model with cardiomyocyte-specific knockout of RIα (RIα-icKO). Myocardial structure and function were evaluated by echocardiography, histology, and ECG and in Langendorff-perfused hearts. PKA activity and cAMP levels were determined by immunoassay, and phosphorylation of PKA targets was assessed by Western blot. L-type Ca²⁺ current (I_Ca,L), sarcomere shortening, Ca²⁺ transients, Ca²⁺ sparks and waves, and subcellular cAMP were recorded in isolated ventricular myocytes (VMs).

Results: RIα protein was decreased by 50% in failing human heart with ischemic cardiomyopathy and by 75% in the ventricles and in VMs from RIα-icKO mice but not in atria or sinoatrial node. Basal PKA activity was increased ≈3-fold in RIα-icKO VMs. In young RIα-icKO mice, left ventricular ejection fraction was increased and the negative inotropic effect of propranolol was prevented, whereas heart rate and the negative chronotropic effect of propranolol were not modified. Phosphorylation of phospholamban, ryanodine receptor, troponin I, and cardiac myosin-binding protein C at PKA sites was increased in propranolol-treated RIα-icKO mice. Hearts from RIα-icKO mice were hypercontractile, associated with increased I_Ca,L, and [Ca²⁺]_i transients and sarcomere shortening in VMs. These effects were suppressed by the PKA inhibitor, H89. Global cAMP content was decreased in RIα-icKO hearts, whereas local cAMP at the phospholamban/sarcoplasmic reticulum Ca²⁺ ATPase complex was unchanged in RIα-icKO VMs. RIα-icKO VMs had an increased frequency of Ca²⁺ sparks and proarrhythmic Ca²⁺ waves, and RIα-icKO mice had an increased susceptibility to ventricular tachycardia. On aging, RIα-icKO mice showed progressive contractile dysfunction, cardiac hypertrophy, and fibrosis, culminating in congestive heart failure with reduced ejection fraction that caused 50% mortality at 1 year.

Conclusions: These results identify RIα as a key negative regulator of cardiac contractile function, arrhythmia, and pathological remodeling.

Keywords: arrhythmias, cardiac; calcium; cyclic AMP-dependent protein kinases; excitation contraction coupling; heart; heart failure; protein kinases.

Figure 1:. Decreased expression of the PKA regulatory subunit RIα in human failing hearts and characterization of a mouse model of conditional and cardiomyocyte-specific knockout of Prkar1A encoding RIα (RIα-icKO).

A, Western blot analysis of RIα protein expression in total protein extracts from non-failing (NF, N=9) and failing human hearts of ischemic origin (ICM, N=10). Actin was used as a loading control. B, Western blot analysis of RIα protein expression in left ventricle (LV), right ventricle (RV), left atrium (LA), right atrium (RA) and sino-atrial node (SAN) from Prkar1a^loxP/loxP (CT, N=6) and Prkar1a^loxP/loxP:α-MHC-MerCreMer (KO, N=6) mice two weeks after tamoxifen injection. GAPDH was used as a loading control. C, Representative Western blot and quantification of the PKA subunits, RIα, RIIα, and Cα in ventricular myocytes (VMs) from CT (N=6) and RIα-icKO (N=8) mice two weeks after tamoxifen injection. GAPDH was used as a loading control. D, Basal (no cAMP) and maximal (cAMP, 25 μmol/L) PKA activity measured in total protein extracts from CT (N=4) and RIα-icKO (N=4) VMs. Data are normalized to the basal CT PKA activity. *P<0.05, Mann-Whitney test (A); ^{$ $ $}P<0.001 between CT and RIα-icKO (B-D); ***P<0.001 compared to CT (no cAMP) (D). Nonparametric (B) or ordinary (C,D) two-way ANOVA followed by Sidak's multiple comparisons test.

Figure 2:. Cardiac structure and function of RIα-icKO mice.

Cardiac structure and function were evaluated by echocardiography in anesthetized CT (N=14) and RIα-icKO (N=14) mice 2-3 weeks after tamoxifen injection. A, Ejection fraction (EF) of the left ventricle (LV). B, LV internal diameter in systole (LVIDs). C, LV internal diameter in diastole (LVIDd). D, Heart rate (HR) in beats per min (bpm). E, Interventricular septum thickness in diastole (IVSd). F, Left ventricular mass to body weight ratio (LVM/BW). G, Gravimetric analysis of heart weight to tibia length ratio (HW/TL) in CT (N=18) and RIα-icKO (N=17) mice two weeks after tamoxifen injection. H, Blood flow peak velocity from left ventricular relaxation in early diastole (E wave). I, Deceleration time (DT) of the E wave. J, Blood flow peak velocity in late diastole (A wave). K, Ratio of E wave to A wave. L, Isovolumic relaxation time (IVRT). M-O, Cardiac function measured in CT (N=5) and RIα-icKO (N=5) mice at baseline, 15 min after atropine (1 mg/kg, IP) and 15 min after propranolol (2 mg/kg, IP) injections. M, Time-Motion mode views of LV function. N, Average HR. O, Average EF of the left ventricle. **P<0.01, ***P<0.001 Mann-Whitney test or Student t-test as appropriate. (A-L); *P<0.05, ***P<0.001 between treatments; ^{$ $ $}P<0.001 between genotypes. Two-way, repeated measures ANOVA with Sidak's multiple comparisons test (N, O).

Figure 3:. Phosphorylation of major ECC proteins is increased in RIα-icKO mice.

Representative Western blot and quantification of the phosphorylation status of major ECC proteins in total ventricular extracts from CT and RIα-icKO mice. Mice were treated with propranolol (2 mg/kg, IP) 10 min. before injection of sodium pentobarbital (150 mg/kg, IP). A, Phosphorylation of phospholamban (PLN) at Ser16 and total protein (N=6/group). B, Phosphorylation of ryanodine receptor 2 (RyR2) at Ser2808 and total protein (N=10/group). C, Phosphorylation of troponin I (TnI) at Ser22/23 and total protein (N=12/group) D, Phosphorylation of myosin-binding protein C (cMyBP-C) at Ser282/283 and total protein (N=10/group). E, Phosphorylation of PLN at Thr17 and total protein (N=6/group). F, Phosphorylation of RyR2 at S2814 and total protein (N=13 /group). *P<0.05, **P<0.01, **P<0.001. Student t-test or Mann-Whitney test as appropriate.

Figure 4:. Isolated perfused hearts from RIα-icKO mice are hypercontractile.

Hearts from CT (N=10) and RIα-icKO (N=9) mice were mounted on a Langendorff apparatus and the following parameters were measured in spontaneously beating hearts in basal conditions and upon 100 nmol/L Isoproterenol (Iso) perfusion. A, Developed pressure (DP) of the left ventricle. B, Maximal positive first derivative of developed pressure (+dP/dt_max). C, Maximal negative first derivative of the pressure (-dP/dt_max). D, Heart rate. ^$P<0.05 between CT and RIα-icKO mice; ***P<0.001 between Basal and Iso. Two-way ANOVA with Sidak’s multiple comparisons test. E, F, Concentration-response curves of developed pressure (DP) and maximal negative first derivative of the pressure (-dP/dt_max) to Iso from 0.03 to 100 nmol/L in Langendorff-perfused hearts from CT and RIα-icKO mice paced at 650 bpm. Both curves were different (extra sum-of-squares F test) with statistically different bottom (^$P<0.001, Student t-test).

Figure 5:. Increased basal L-type Ca²⁺ current (I_Ca,L) in VMs from RIα-icKO mice.

A, Typical time course of I_Ca,L in a CT VM. The cell was first superfused with external Cs⁺-Ringer solution during ~70 s. and then challenged with 30 nmol/L isoproterenol (Iso) during 15 s. as indicated by the black bar. B, Same experiment as in A, in a VM from a RIα-icKO mouse. Each symbol represents I_Ca,L density (dI_Ca,L, in pA/pF) recorded every 8 seconds during a depolarization to 0 mV from a resting potential of −50 mV. Individual current traces shown on top of the graphs were recorded at the time indicated by the corresponding lower case letters on the graphs. C, Average I_Ca,L density at baseline and at the peak of Iso (30 nmol/L, 15 s) stimulation and D, cell capacitance (in picofarad, pF) in CT (n=22 VMs from 4 mice) and RIα-icKO (n=19 VMs from 4 mice). ^$P<0.05; ***P<0.001. C, Nested ANOVA of ART data with Sidak’s multiple comparisons test. D, Nested t-test.

Figure 6:. Ca²⁺ homeostasis in VMs from CT and RIα-icKO mice.

A, Line-scan confocal images recorded during field stimulation at 2 Hz of a CT (left) and a RIα-icKO (right) VM in control external solution (Basal) and upon 100 nmol/L Isoproterenol (Iso) perfusion. Above each image, the fluorescence trace expressed as F/F₀ is shown. B, Amplitude of the [Ca²⁺]_i transient expressed as the peak F/F₀ in CT (n=18 from 5 mice) and RIα-icKO (n=16 from 6 mice) VMs in control external solution (Basal) and with Iso 100 nmol/L. C, Decay time constant (Tau) of [Ca²⁺]_i transients obtained by fitting the fluorescence trace to a single exponential function in the same cells and experimental conditions as in B. D, Line-scan confocal images recorded in a quiescent CT (left) and RIα-icKO (right) VM in basal conditions. E, Ca²⁺ sparks frequency in CT (n=17 from 5 mice) and RIα-icKO (n=13 from 5 mice) VMs perfused with control external solution (Basal) and with Iso 100 nmol/L. F, Ca²⁺ waves frequency in CT (n=17 from 5 mice) and RIα-icKO (n=13 from 5 mice) VMs perfused with control external solution (Basal) and with Iso 100 nmol/L. Data are reported as mean±SEM. *P<0.05, **P<0.01, ***P<0.001, Iso versus basal; ^$P<0.05, ^{$ $}P<0.01, CT versus RIα-icKO. Nested ANOVA of ART data followed by Sidak’s multiple comparisons test in B, C, F. In E, Nested ANOVA of ART data indicated that interaction between genotype and treatment was not significant, however genotype had a statistically significant effect on Ca²⁺ sparks frequency (^$P<0.05).

Figure 7:. Increased susceptibility to ventricular tachycardia in RIα-icKO mice.

A, Example of an ECG trace with an episode of ventricular tachycardia (VT) recorded in a RIα-icKO mouse after isoproterenol injection (1.5 mg/kg, IP). B, Number of VT episodes recorded during 1h after isoproterenol injection in CT and RIα-icKO mice. C, Number of mice with VT (in red) or without VT (in white) after isoproterenol injection CT, N=10; RIα-icKO, N=11. **P<0.01, Mann-Whitney test (B), Fisher’s exact test (C).

Figure 8:. RIα-icKO mice develop HF upon aging.

A, Kaplan-Meier plot of survival in CT (N=13) and RIα-icKO mice (N=14). B, Examples of hearts (top) and transversal sections stained with hematoxylin-eosin (bottom) from a CT and a RIα-icKO mouse at 1 year of age (45 weeks post-tamoxifen injection). C-H, Echocardiographic parameters in CT (N=10) and a RIα-icKO mice (N=10) at 1 year of age (45 weeks post-tamoxifen injection). C, LV ejection fraction (EF). D, Ratio of E wave to A wave. E, LV internal diameter in diastole (LVIDd). F, LV internal diameter in systole (LVIDs). G, LV mass to tibia length ratio. H, Heart rate (HR) in beats per minutes (bpm) I, Heart weight to tibia length ratio in CT (N=10) and RIα-icKO (N=10) mice. J, Representative images of cardiac transversal sections from CT and RIα-icKO mice stained with fluorescein-conjugated wheat germ agglutinin to label cell membranes. Scale bar 50 μm. Quantification of cardiomyocyte cross-sectional area (CSA) in CT (N=6) and RIα-icKO mice (N=6). K, Lung weight to tibia length ratio in CT (N=10) and RIα-icKO (N=10) mice. L, Representative images of cardiac transversal sections from CT and RIα-icKO mice stained with Masson’s trichrome (scale bar 200 μm) and quantification of interstitial fibrosis area in CT (N=5) and RIα-icKO mice (N=7). **P<0.01, log rank test (A); *P<0.05, **P<0.01, Student t-test or Mann-Whitney test as appropriate (C-L).