Modification of the hydantoin ring in the previously described lead compound 2a has led to the discovery of compound 12a, tadalafil, a highly potent and highly selective PDE5 inhibitor. The replacement of the hydantoin in compound 2a by a piperazinedione ring led to compound cis-11a which showed similar PDE5 inhibitory potency. Introduction of a 3,4-methylenedioxy substitution on the phenyl ring in position 6 led to a potent PDE5 inhibitor cis-11c with increased cellular potency. Optimization of the chain on the piperazinedione ring led to the identification of the racemic cis-N-methyl derivative 11i. High diastereospecificity for PDE5 inhibition was observed in the piperazinedione series with the cis-(6R,12aR) enantiomer displaying the highest PDE5 inhibitory activity. The piperazinedione 12a, tadalafil (GF196960), has been identified as a highly potent PDE5 inhibitor (IC(50) = 5 nM) with high selectivity for PDE5 vs PDE1-4 and PDE6. Compound 12a displays 85-fold greater selectivity vs PDE6 than sildenafil 1. 12a showed profound and long-lasting blood pressure lowering activity (30 mmHg/>7 h) in the spontaneously hypertensive rat model after oral administration (5 mg/kg).

Starting from ethyl beta-carboline-3-carboxylate (beta-CCE), 1, a modest inhibitor of type 5 phosphodiesterase (PDE5), a series of functionalized tetrahydro-beta-carboline derivatives has been identified as a novel chemical class of potent and selective PDE5 inhibitors. Optimization of the side chain on the hydantoin ring of initial lead compound 2 and of the aromatic ring on position 5 led to the identification of compound 6e, a highly potent and selective PDE5 inhibitor, with greater selectivity for PDE5 vs PDE1-4 than sildenafil. Compound 6e demonstrated a long-lasting and significant blood pressure lowering effect after iv administration in the spontaneously hypertensive rat model but showed only moderate oral in vivo efficacy.

1In rat aortic rings precontracted with phenylephrine, the P-adrenoceptor agonist isoprenaline (10 nM to 30 gM) produces greater relaxant effects in preparations with endothelium than in endotheliumdenuded preparations. The aim of this study was to determine the mechanisms involved in this effect and in particular investigate the possibility of a synergistic action between adenosine 3': 5'-cyclic monophosphate (cyclic AMP) and guanosine 3': 5'-cyclic monophosphate (cyclic GMP). 2 Isoprenaline-induced relaxation of rat aortic rings precontracted with phenylephrine was greatly reduced by the nitric oxide (NO) synthase inhibitor N--nitro-L-arginine methyl ester (L-NAME, 300 giM) or the soluble guanylate cyclase inhibitors methylene blue (10 gM) or lH-[l,2,4]oxadiazolo[4,3-a]quinoxalin-l-one (ODQ, 10 giM) but unaffected by indomethacin (10 uM), a cyclo-oxygenase inhibitor.Similarly, in intact rings, the concentration-response curve of forskolin (10 nM to 1 jiM) was shifted to the right upon endothelium removal or treatment with methylene blue. 3 In endothelium-denuded rat aortic rings, isoprenaline-induced relaxation was potentiated by the guanylate cyclase activators atrial natriuretic factor (ANF, 1 to 10 nM) and sodium nitroprusside (SNP, 1 to 10 nM), and to a greater extent in the presence of the cyclic GMP-specific phosphodiesterase (PDE 5) inhibitor, 1,3dimethyl-6-(2-propoxy-5-methane sulphonylamidophenyl) pyrazolo [3,4-d] pyrimidin-4-(5H)-one (DMPPO, 30 nM). Relaxation induced by isoprenaline was also potentiated by the cyclic GMP-inhibited PDE (PDE 3) inhibitor cilostamide (100 nM). 4 Intracellular cyclic nucleotide levels were measured either in rat cultured aortic smooth muscle cells or in de-endothelialized aortic rings. In both types of preparation, isoprenaline (5 nM and 10 jiM) increased cyclic AMP levels and this effect was potentiated by cilostamide (10 pM), by rolipram, a cyclic AMP-specific PDE (PDE 4) inhibitor (10 jM) and by cyclic GMP-elevating agents (50 nM ANF or 30 nM SNP plus 100 nM DMPPO). In isoprenaline-stimulated conditions, the increase in cyclic AMP induced by rolipram was further potentiated by cilostamide and by cyclic GMP-elevating agents. Cilostamide and cyclic GMP-elevating agents did not potentiate each other, suggesting a similar mechanism of action. 5 We conclude that in vascular smooth muscle (VSM) cells an increase in cyclic GMP levels may inhibit PDE 3 and, thereby, cyclic AMP catabolism. Under physiological conditions of constitutive NO release, and to a greater extent in the presence of the PDE 5 inhibitor DMPPO, cyclic GMP should act synergistically with adenylate cyclase activators to relax VSM.

1 Cyclic guanosine 3' ± 5'-monophosphate (cyclic GMP) is the second messenger of important physiologically active mediators controlling the pulmonary vascular tone. To potentiate the e ects of cyclic GMP on the pulmonary vasculature, we used DMPPO, a new selective PDE-5 inhibitor, and examined its action in a rat model of hypoxic pulmonary hypertension. 2 Levels of cyclic GMP measured during baseline conditions at 5 and 60 min of perfusion were similar in the perfusate of isolated lungs from normoxic and chronically hypoxic rats and did not di er with time. Pretreatment with DMPPO (1 mM) induced a larger increase in cyclic GMP concentration in the perfusate from chronically hypoxic rat lungs (319+36 at 5 min to 1821+83 pmol ml 71 at 60 min) than in normoxic rat lungs (329+20 to 1281+127 pmol ml 71 , P50.05). 3 In isolated lungs preconstricted with U-46619, pretreatment with DMPPO (1 mM) potentiated the vasodilator e ects of atrial natriuretic peptide (100 pM ± 10 nM) and sodium nitroprusside (1 pM ± 10 nM), but did not alter vasodilation to isoproterenol. 4 In conscious rats previously exposed to 15 days hypoxia and studied under 10% O 2 , DMPPO (0.01, 0.05 and 0.1 mg kg 71 , i.v. bolus) caused a dose-dependent decrease in pulmonary arterial pressure (Pap) with no change in systemic artery pressure (Sap) and cardiac output. 5 Continuous infusion of DMPPO (0.1 mg kg 71 h 71 i.v. by osmotic pumps) in rats exposed to 10% O 2 during 2-weeks reduced the Pap (P50.05) and the degree of muscularization of pulmonary vessels at the alveolar wall (P50.01) and alveolar duct levels (P50.05) despite no signi®cant change in right ventricular hypertrophy. 6 These results suggest that cyclic GMP phosphodiesterase inhibition may selectively dilate pulmonary circulation during chronic hypoxia.

The aim of this study was to investigate the cardiovascular effects of a novel, potent and specific phosphodiesterase 5 (PDE 5) inhibitor, 1,3 dimethyl‐6‐(2‐propoxy‐5‐methane sulphonylamidophenyl)‐pyrazolo[3,4‐d]pyrimidin‐4‐(5H)‐one (DMPPO) in phenylephrine‐precontracted rat aortic rings and different in vivo rat preparations. DMPPO elicited a concentration‐dependent relaxation of rat aortic rings with functional endothelium. DMPPO‐induced relaxation was abolished by endothelium removal or pretreatment with the soluble guanylate cyclase inhibitor, methylene blue (10 μm). In aortic rings without endothelium, the potency (pD2 = ‐log10 EC50) of atrial natriuretic peptide (ANP) to induce relaxation increased from 8.13 ± 0.05 in the absence of DMPPO to 8.32 ± 0.05 and 8.52 ± 0.08 in the presence of 30 nM and 100 nM DMPPO, respectively. Similarly, the potency of sodium nitroprusside (SNP) in inducing relaxation increased from 7.38 ± 0.07 in the absence of the PDE 5 inhibitor to 8.07 ± 0.11 and 8.15 ± 0.08 in the presence of 30 nM and 100 nM DMPPO, respectively. In contrast, relaxation to the adenylate cyclase activator, forskolin, was unchanged by DMPPO (100μm). In rings without endothelium, DMPPO (100 nM) increased by 2.5 fold intracellular levels of guanosine 3′:5′‐cyclic monophosphate (cyclic GMP). Moreover, DMPPO (100 nM) potentiated the increases in cyclic GMP levels induced by ANP (30 nM) by 3 fold and SNP (30 nM) by 2.7 fold. Adenosine 3′:5′‐cyclic monophosphate (cyclic AMP) levels were not modified by DMPPO. In anaesthetized normotensive or spontaneously hypertensive rats (SHR), DMPPO (2 and 5 mg kg−1, i.v.) lowered blood pressure without affecting heart rate. Similarly, in conscious SHR, orally administered DMPPO (5 mg kg−1) induced a 25 mmHg decrease in blood pressure for at least 7 h without modifying heart rate. Meanwhile, urinary cyclic GMP was increased by 50% whereas cyclic AMP remained unchanged. In normotensive anaesthetized rats, sodium nitroprusside (SNP) (i.v. bolus) induced a decrease in blood pressure which rapidly returned to baseline. In DMPPO (1 mg kg−1, i.v.)‐treated rats, the hypotensive effects of SNP (10 to 100 μg kg−1) were prolonged over time whereas the peak effect was unchanged. In pithed rats, phenylephrine (i.v. bolus) induced dose‐dependent increases in blood pressure. Pretreatment with DMPPO (5 mg kg−1, i.v.) partially inhibited the pressor response to phenylephrine (0.3 to 100 μg kg−1). In conclusion, the potent and selective PDE 5 inhibitor, DMPPO, produces relaxation in isolated vessels in the presence of a cyclic GMP drive and reduces blood pressure of intact animals. Its high oral bioavailability and long duration of action should make it a useful tool to study the role of cyclic GMP in various biological systems.