Inhibition of the upregulated phosphodiesterase 4D isoforms improves SERCA2a function in diabetic cardiomyopathy

Abstract

Background and purpose: Sarcoplasmic reticulum Ca²⁺-ATPase (SERCA2a) is impaired in heart failure. Phosphodiesterases (PDEs) are implicated in the modulation of local cAMP signals and protein kinase A (PKA) activity essential for cardiac function. We characterise PDE isoforms that underlie decreased activities of SERCA2a and reduced cardiac contractile function in diabetic cardiomyopathy.

Experimental approach: Wild type mice were fed with either normal chow or a high-fat diet (HFD). Cardiomyocytes were isolated for excitation-contraction coupling (ECC), fluorescence resonant energy transfer PKA biosensor and proximity ligation assays.

Key results: The upregulated PDE4D3 and PDE4D9 isoforms in HFD cardiomyocytes specifically bound to SERCA2a but not ryanodine receptor 2 (RyR2) on the sarcoplasmic reticulum (SR). The increased association of PDE4D isoforms with SERCA2a in HFD cardiomyocytes led to reduced local PKA activities and phosphorylation of phospholamban (PLB) but minimally effected the PKA activities and phosphorylation of RyR2. These changes correlate with slower calcium decay tau in the SR and attenuation of ECC in HFD cardiomyocytes. Selective inhibition of PDE4D3 or PDE4D9 restored PKA activities and phosphorylation of PLB at the SERCA2a complex, recovered calcium decay tau, and increased ECC in HFD cardiomyocytes. Therapies with PDE4 inhibitor roflumilast, PDE4D inhibitor BPN14770 or genetical deletion of PDE4D restored PKA phosphorylation of PLB and cardiac contractile function.

Conclusion and implications: The current study identifies upregulation of specific PDE4D isoforms that selectively inhibit SERCA2a function in HFD-induced cardiomyopathy, indicating that this remodelling can be targeted to restore cardiac contractility in diabetic cardiomyopathy.

Keywords: PDE4D; SERCA2a; diabetic cardiomyopathy; excitation contraction coupling; myocytes.

FIGURE 1. High-fat diet (HFD) feeding induces changes in PDE expression associated with subcellular nanodomains in cardiomyocytes.

(a and b) WT mice were fed with a HFD or normal chow (NC) diet for 4.5 months. The heart tissues were harvested for western blot analysis to detect the expression of β1-adrenoceptor, PDE3A, 4A, 4B, 4D and γ-tubulin. The quantification of western blots is presented in dot plots. (c–f) Myocytes were isolated and infected with adenoviruses expressing different AKAR3 biosensors (PM, SR, TNT, Reg). Myocytes were treated with EHNA (10 mM), cilostamide (Cilo) (1 mM), rolipram (Roli) (10 mM) and sildenafil (Sild) (1 mM) as indicated. The changes in FRET ratio were recorded, and the maximal changes in FRET ratio were plotted. * P < 0.05 by Student t test in Figure 1b and by two-way ANOVA followed by Tukey’s test in other panels.

FIGURE 2. Inhibition of PDE4 rescues PKA phosphorylation of phospholamban and SR calcium load in high-fat diet (HFD)-fed cardiomyocytes.

(a, b, f, g) Cardiac myocytes isolated from normal chow (NC) or HFD mice were treated with Roli (10 μM) for 10 min. Cells were lysed for western blots to detect the levels of phosphorylation of PLB at Ser16 and RYR2 at Ser2808. The quantification of western blot is shown in dot plots. (c–e) Cardiac myocytes loaded with calcium dye Fluo-4AM and paced at 1 Hz. Myocytes were treated with EHNA (10 μM), cilostamide (Cilo) (1 μM), rolipram (Roli) (10 μM) and sildenafil (Sild) (1 μM) as indicated. Calcium signals and SS were recorded. The maximal changes in sarcomere shortening (SS), calcium decay tau and transient amplitude were plotted. *P < 0.05 by two-way ANOVA followed by Tukey’s test.

FIGURE 3. The β₁-adrenoceptor-induced PKA activity of the SERCA2a nanodomains is reduced in high-fat diet (HFD)-fed cardiomyocytes.

(a–j) Isolated myocytes from normal chow (NC) and HFD fed mice were infected with adenoviruses expressing RyR2 and SERCA2a associated FKBP and SR AKAR3 biosensors. Myocytes were pretreated with CGP20712a (300 nM) or ICI11811 (100 nM) before stimulation with isoproterenol (ISO, 100 nM) and followed by addition of forskolin (FSK) (10 μM) and IBMX (100 μM). The time courses and the maximal changes in FRET ratio over the baseline were calculated and plotted. *P < 0.05 by two-way ANOVA followed by Tukey’s test. Ctrl, control.

FIGURE 4. Inhibition of PDE4 restores β₁-adrenoceptor-induced PKA activity at the SERCA2a nanodomains and improves calcium recycling and contractile function in high-fat diet (HFD)-fed myocytes.

(a–f) Isolated myocytes from normal chow (NC) and HFD mice were infected with adenoviruses expressing RyR2 and SERCA2a associated FKBP and SR AKAR3 biosensors. After treatment with PDE4 inhibitor Roli (100 nM) or Veh, myocytes were stimulated with the β-adrenoceptor agonist ISO (100 nM). The time courses and maximal changes in FRET ratio was presented. (g–i) Isolated myocytes were stimulated with ISO (100 nM, 10 min) in the presence of rolipram (Roli) (100 nM) or Veh as indicated. Cell lysates were subjected to detection of phosphorylation of PLB at Ser16 and RyR2 at Ser2808. The western blots were quantified in dot plots. (j–l) Isolated myocytes were loaded with calcium dye Fluo-4 AM and paced at 1 Hz. Myocytes were stimulated with ISO (100 nM) in the presence and absence of Roli (100 nM) or Veh and followed by addition of IBMX (100 μM)/FSK (10 μM). Calcium signals and sarcomere shortening (SS) were recorded. Sarcomere shortening and calcium decay tau and transient amplitude were calculated and plotted. *P < 0.05 by two-way ANOVA followed by Tukey’s test.

FIGURE 5. Inhibition of PDE4 improves diastolic function of high-fat diet (HFD)-fed mice.

(a) Diagram shows that WT mice were fed wi7th HFD for 4.5 months and then treated with saline or roflumilast (Rofl) (10 mg kg⁻¹) for 4 weeks. (b–i) Echocardiography was used to detect diastolic function parameters E′/A′ and isovolumic relaxation time (IVRT) and systolic function parameter ejection fraction (EF) in HFD mice after treatments. (j and k) After treatments, heart tissues from HFD mice were harvested for western blot analysis to detect the expression of PLB, SERCA2a, RyR2 and γ-tubulin, as well as phosphorylation of PLB at Ser16 and phosphorylation of RyR2 at Ser2808. The western blots were quantified and presented in dot plots. *P < 0.05 by one-way ANOVA followed by Tukey’s test in Figure 5d, f, i, and by Student t test in other panels.

FIGURE 6. The upregulated PDE4D3 and PDE4D9 selectively associates with SERCA2a in high-fat diet (HFD)-fed myocytes.

(a and b) Heart tissues from normal chow (NC) and HFD mice were subjected to western blot analysis of protein levels of PDE4D isoforms. The levels of protein expression were quantified in dot plots. (c–e) Isolated myocytes were subjected to PLA analysis with rabbit antibodies against PDE4D3, 4D5 and 4D9 together with an antibody agonist SERCA2a or RYR2, respectively. The PLA signals were detected with a Zeiss confocal microscope and quantified. (g and h) Isolated myocytes were infected with adenoviruses expressing SR AKAR3 biosensors. After treatment with membrane permeable dominant negative N-terminal (NT) PDE4D peptides (1 μM), myocytes were stimulated with isoproterenol (ISO) (100 nM), followed by addition of Roli (100 nM) and IBMX (100 mM) as indicated. The changes in FRET ratio were recorded. The maximal changes in the FRET ratio induced by ISO and other inhibitors were analysed and plotted. *P < 0.05 by Student t test in Figure 6b and e and by one-way ANOVA followed by Tukey’s test in Figure 6g.

FIGURE 7. Inhibition of PDE4D3 or PDE4D9 restores subcellular β-adrenoceptor signalling and ECC in high-fat diet (HFD)-fed myocytes.

Isolated myocytes were loaded with calcium dye Fluo-4AM and paced at 1 Hz. (a–c) Myocytes were treated with ISO, forskolin (FSK) (10 μM) and IBMX (100 μM) as indicated. The maximal changes in sarcomere shortening (SS) and calcium decay tau and transient amplitude were analysed and plotted. (d–f) Myocytes were pretreated with membrane permeable dominant negative N-terminal (NT) PDE4D isoform peptides (1 μM) before stimulation with ISO (100 nM). The maximal changes in sarcomere shortening and calcium decay tau and transient amplitude were plotted. (g–i) Cardiac myocytes were isolated from normal chow (NC) and HFD mice and treated with ISO (100 nM) in the presence of NT PDE4D isoform peptide (1 μM). Cell lysates were applied to western blots to detect the levels of total and phosphorylation of PLB at Ser16 and RyR2 at Ser2808. *P < 0.05 by two-way ANOVA followed by Tukey’s test. Con, control.

FIGURE 8. Selectively inhibition of PDE4D effectively ameliorates cardiac dysfunction of high-fat diet (HFD)-fed mice.

(a) Diagram shows that WT mice were fed with HFD for 4.5 months and then treated with saline or BPN (0.03 mg kg⁻¹) for 4 weeks. (b–i) Echocardiography was used to detect diastolic function parameters E′/A′ and isovolumic relaxation time (IVRT) and systolic function parameter ejection fraction (EF) in HFD mice after treatments. (j and k) After treatments, heart tissues from HFD mice were harvested for western blot analysis to detect the expression of PLB, SERCA2a, RyR2 and GAPDH, as well as phosphorylation of PLB at Ser16 and phosphorylation of RyR2 at Ser2808. The western blots were quantified and presented in dot plots. *P < 0.05 by one-way ANOVA followed by Tukey’s test and Student t test.