Melatonin inhibits nitric oxide signaling by increasing PDE5 phosphorylation in coronary arteries

Abstract

Melatonin inhibits nitric oxide (NO)-induced relaxation of coronary arteries. We tested the hypothesis that melatonin increases the phosphorylation of phosphodiesterase 5 (PDE5), which increases the activity of the enzyme and thereby decreases intracellular cGMP accumulation in response to NO and inhibits NO-induced relaxation. Sodium nitroprusside (SNP) and 8-Br-cGMP caused concentration-dependent relaxation of isolated coronary arteries suspended in organ chambers for isometric tension recording. In the presence of melatonin, the concentration-response curve to SNP, but not 8-Br-cGMP, was shifted to the right. The effect of melatonin on SNP-induced relaxation was abolished in the presence of the PDE5 inhibitors zaprinast and sildenafil. Melatonin markedly inhibited the SNP-induced increase in intracellular cGMP in coronary arteries, an effect that was also abolished by zaprinast. Treatment of coronary arteries with melatonin caused a nearly fourfold increase in the phosphorylation of PDE5, which increased the catalytic activity of the enzyme and thereby increased the degradation of cGMP to inactive 5'-GMP. Melatonin-induced PDE5 phosphorylation was markedly attenuated in the presence of the PKG1 inhibitors DT-2 or Rp-8-Br-PET-cGMPS and in those arteries in which PKG1 expression was first downregulated by 24-h incubation with SNP before exposure to melatonin. The selective MT(2) receptor antagonist 4-phenyl-2-propionamidotetralin completely blocked the stimulatory effect of melatonin on PDE5 phosphorylation as well as the inhibitory effect of melatonin on SNP-induced relaxation and intracellular cGMP. Thus, in coronary arteries, melatonin acts via MT(2) receptors and PKG1 to increase PDE5 phosphorylation, resulting in decreased cGMP accumulation in response to NO and impaired NO-induced vasorelaxation.

Fig. 1.

Log concentration-response curves for sodium nitroprusside (SNP; A) or 8-Br-cGMP (B) in producing relaxations of isolated porcine coronary arteries (without endothelium) in the absence and presence of melatonin (Mel). Data are expressed as percentages of the U-46619 (3 × 10⁻⁹ M)-induced increase in tension, which averaged 5.83 ± 0.6 g in control rings and did not differ significantly in rings incubated with Mel. Each point represents the mean ± SE; n = 9. *Statistically significant difference from control in the presence of Mel (P < 0.05).

Fig. 2.

A: effect of Mel (10⁻⁷ M) on SNP-induced relaxation of isolated porcine coronary artery arteries (without endothelium) in the absence and presence of the selective MT₂ receptor antagonist 4-phenyl-2-propionamidotetralin (4P-PDOT; 10⁻⁷ M). B: effect of Mel (10⁻⁷ M) on SNP-induced relaxation of isolated porcine coronary arteries in the presence of the phosphodiesterase 5 (PDE5) inhibitor zaprinast (Zap; 10⁻⁵ M). Data are expressed as percentages of the U-46619 (3 × 10⁻⁹ M)-induced increase in tension, which averaged 2.65 ± 0.3 g in control rings and did not differ significantly in rings incubated with either 4P-PDOT or Zap. Each point represents the mean ± SE; n = 6–8. *Statistically significant difference in the presence of Mel (P < 0.05).

Fig. 3.

Effect of the PDE5 inhibitor Zap (10⁻⁵ M) and the selective MT₂ receptor antagonist 4P-PDOT (10⁻⁷ M) on the SNP (10⁻⁵ M)-induced increase in cGMP levels in the presence and absence of Mel (10⁻⁷ M). cGMP levels are expressed as picomoles per microgram of protein; n = 6–7. *Statistically significant difference between groups (P < 0.05).

Fig. 4.

A: PDE5 protein detected as an intense immunoreactive band at 95 kDa in porcine coronary artery homogenates from four different animals (lanes 1–4). B and C: effect of Mel on PDE5 phosphorylation in the presence and absence of the selective MT₂ receptor antagonist 4P-PDOT (10⁻⁷ M, n = 6). C shows the immunodensity of phospho-PDE5 normalized to the corresponding PDE5 protein band immunodensity. *Statistically significant difference between groups (P < 0.05).

Fig. 5.

A: effect of SNP on PKG1 protein expression in porcine coronary arteries in the presence and absence of a soluble guanylyl cyclase inhibitor, 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ; 10⁻⁵ M, n = 5). Double-distilled water (DDW) and DMSO were the solvents for SNP and ODQ, respectively. B: immunodensity of PKG1 normalized to the corresponding β-tubulin protein band immunodensity. *Statistically significant difference between groups (P < 0.05).

Fig. 6.

A: Mel-induced (10⁻⁷ M) PDE5 phosphorylation was significantly attenuated in porcine coronary arteries that had been first treated with SNP (10⁻⁵ M, 24 h) to induce the downregulation of PKG1 (n = 5). B: immunodensity of phospho-PDE5 normalized to the corresponding PDE5 protein band immunodensity. *Statistically significant difference between groups (P < 0.05).

Fig. 7.

A: Mel-induced (10⁻⁷ M) PDE5 phosphorylation was significantly attenuated in porcine coronary arteries that had been first treated with Rp-8-Br-PET-cGMPS (3 × 10⁻⁵ M, 30 min) or DT-2 (10⁻⁵ M, 30 min) to inhibit PKG1 activity in arterial rings (n = 4). B: immunodensity of phospho-PDE5 normalized to the corresponding PDE5 protein band immunodensity. *Statistically significant difference between groups (P < 0.05).