International Journal of Impotence Research (2001) 13, 240±246 ß 2001 Nature Publishing Group All rights reserved 0955-9930/01 $15.00 www.nature.com/ijir Dual mechanism of action of nicorandil on rabbit corpus cavernosal smooth muscle tone GC Hsieh1*, T Kolasa1, JP Sullivan1 and JD Brioni1 1Neurological and Urological Diseases Research, Abbott Laboratories, Illinois, USA The potential of ATP-sensitive potassium channel openers (KCOs) for the treatment of male erectile dysfunction has recently been suggested based on positive clinical outcomes following intra-cavernosal administration of pinacidil. Agents that increase the levels of cGMP via elevation of nitric oxide (NO) nitroglycerin, for example, are also effective in improving erectile function preclinically and clinically. The aim of the present study was to determine the effects and mechanism of the action of nicorandil on rabbit corpus cavernosum. The in vitro regulation of smooth muscle tone was assessed in isolated cavernosal tissues pre-contracted with phenyl- ephrine. Nicorandil, but not its major metabolite, relaxed phenylephrine-precontracted caverno- sum smooth muscle with an EC50 of 15 mM. The effects of nicorandil were only partially reversed by the KATP channel blocker glyburide (10 mM) or by a soluble guanylate cyclase (sGC) inhibitor 1H-[1,2,4] oxadiazole [4,3-a] quinoxalin-1-one (ODQ, 3 mM). However, a combination of ODQ and glyburide completely blocked the relaxant effects of nicorandil. The results of the present study indicate that nicorandil can relax rabbit cavernosal tissue in vitro via a mechanism that involves activation of KATP channels and stimulation of soluble guanylate cyclase. International Journal of Impotence Research (2001) 13, 240±246. Keywords: nicorandil; erectile dysfunction; KATP channel; guanylate cyclase Introduction and NO donors) have shown limited ef®cacy in clinical trials for male erectile dysfunction.4 How- ever, since these agents relax the peripheral vascu- Corpus cavernosal smooth muscle tone is controlled lature via a similar mechanism and are prone to by a number of discrete signaling pathways co- systemic side-effects,5 they must be administered ordinated at the level of the peripheral and central locally by intracavernosal injection or topically. nervous system. This smooth muscle tone ulti- ATP-sensitive potassium channels (KATP) are an mately regulates penile ¯accidity and erection.1 important family of potassium channels that couple Contraction is primarily mediated by alpha- cellular energy metabolism to membrane electrical adrenoceptor stimulation1,2 while relaxation is 6,7 activity. At physiologic ATP levels, KATP chan- mediated by the interaction of several types of nels remain mostly in their closed state while at low neurotransmission, including nonadrenergic ± non- ATP levels, the channels open. Potassium channel cholinergic (NANC) neural input and the release of openers (KCOs) such as L-cromakalim, diazoxide, nitric oxide,1,3 Nitric oxide, released from NANC pinacidil and its analogue P1075 activate KATP nerve terminals and from cholinergically-activated channels producing hyperpolarization of vascular endothelial cells, diffuses into adjacent smooth and non-vascular smooth muscle membranes which muscle cells where it activates soluble guanylate reduce Ca2 in¯ux through voltage-activated L-type cyclase. This increases the intracellular levels of Ca2 channels, thus resulting in smooth muscle cyclic guanosine monophosphate (cGMP) that leads relaxation.6 Recently, the prototypic potassium to the relaxation of cavernosal smooth muscle. The channel openers such as L-cromakalim and pinaci- increased blood ¯ow and engorgement of the dil, when administered as intracavernosal injec- trabecular spaces results in veno-occlusion and tions, have been found effective in the initiation penile erection. Nitrovasodilators (organic nitrates and maintenance of penile erection in experi- mental animal models8±10 and in clinical study in humans.11 *Correspondence: G Hsieh, Neurological and Urological Nicorandil (N-(2-hydroxyethyl)nicotinamide ni- Diseases Research, D-4ND, AP-9, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064-6119, USA. trate ester) is an orally ef®cacious anti-anginal E-mail: [email protected] drug.12 While this agent has been shown to relax Received 21 November 2000; accepted 16 April 2001 human corpus cavernosum,13 the mechanism under- KATP and guanylate cyclase activation by nicorandil GC Hsieh et al 241 lying this effect remains to be fully elucidated. The dissected free of the tunica albuginea and surround- present study was designed to investigate the ing connective tissues). Each tissue strip was mechanism of action of nicorandil in rabbit corpus longitudinally cut into three strip preparations cavernosum. Speci®cally, the role of the KCO and measuring approximately 26267 mm (unstretched nitrate activities associated with this compound was length, mean weight 30 mg). investigated. Additionally, the possibility that the relaxant effects of nicorandil are mediated via its major metabolite was also assessed. Measurement of isometric tension Materials and methods The strips were transferred and mounted to organ baths (10 ml) containing Kreb ± Henseleit buffer solution (pH 7.4) maintained at 37C by a thermo- Drugs regulated water circuit. The buffer solution con- tained D-glucose 2.0 g=l, MgSO4 0.141 g=l, KH2PO4 Nicorandil (N-(2-hydroxyethyl) nicotinamide nitrate 0.16 g=l, KCl 0.35 g=l, NaCl 6.9 g=l (Sigma Chemical ester) and its metabolite N-(2-hydroxyethyl) nicotin- Co) plus CaCl2.2H2O 0.373 g=l and NaHCO3 2.1 g=l. The buffer was continuously aerated with a mixture amide (Figure 1), KATP channels openers L-croma- of 95% O =5% CO . The tissue strips were loaded kalim, P1075, and KATP channel blocker glyburide 2 2 were synthesized at Abbott Laboratories (Abbott with a resting tension of 2 g and equilibrated for Park, IL). Guanylate cyclase inhibitor ODQ (1H- 90 min. During equilibration, the bath solution was [1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one), diazoxide, replaced every 10 ± 15 min. Changes in isometric sodium nitroprusside, and phenylephrine were tension of muscle tissues were measured using a purchased from Sigma Chemical Co. (St Louis, force-displacement transducer (Grass FT03 model, MO). Nicorandil was dissolved in dimethyl sulf- Grass Instruments Co., Quincy, MA) and recorded oxide (DMSO) to make a 100 mM stock solution. All on a Grass 7D model polygraph. Repeated adjust- drugs, unless indicated otherwise, were further ment of tension was performed if necessary. No diluted with ethanol before adding into the buffer changes in tension were made after the experiment solutions in tissue bath chamber so that the ®nal was started. total solvent (DMSO plus ethanol) concentrations in the assays was less than 0.5% (v=v) which itself did not alter the smooth muscle tone. Stock solutions of Drug treatment phenylephrine prepared in saline included 0.1% ascorbic acid as an antioxidant. The relaxation effects of nicorandil, nicorandil metabolite, L-cromakalim, P1075, diazoxide, and Tissue preparation sodium nitroprusside were assessed in rabbit corpus cavernosum preparations precontracted with phe- nylephrine (1.5 mM). At stable tension, compounds The in vitro regulation of smooth muscle tone was to be tested were added cumulatively at half log unit assessed using isolated rabbit cavernosum smooth increments, and a new concentration was not added muscle strips mounted in organ bath chamber and until the response to the previous one had attained a pre-contracted with phenylephrine. Corpus caver- steady state tension. nosum tissues were prepared from adult male New In another set of experiments, tissue preparations Zealand white rabbits (Covance Research Produc- were contracted with phenylephrine 1.5 mM and tion, Kalamazoo, MI), weighing 3.0 ± 3.5 kg, after subsequently treated with 3 ± 10 mM ODQ (a guany- euthanasia by intravenous pentobarbital sodium late cyclase inhibitor), 10 mM glyburide (for compe- injection. The two corpus cavernosum strips were titive blockade of ATP-dependent K channel), carefully dissected under a illuminated magni®er (a or combination of ODQ (3 mM) and glyburide slit was made in the proximal end of the tunica and (10 mM) for 15 min. The inhibitors were added after extended distally and the corpus cavernosum was the phenylephrine-induced contraction reached O O O OH N N NO2 H H N N Figure 1 Structure of nicorandil and its metabolite N-(2-hydroxyethyl) nicotinamide. International Journal of Impotence Research KATP and guanylate cyclase activation by nicorandil GC Hsieh et al 242 steady-state. A cumulative concentration ± response Effects of KATP channel openers P1075, curve was then constructed for the relaxant agents L-cromakalim, and diazoxide on phenylephrine above described. contracted corpus cavernosal strips In order to determine the effect of other KCOs Data analysis lacking nitrate activity on rabbit corpus cavernosal smooth muscle tone, three reference openers, P1075, Sigmoid curves were ®tted to concentration- L-cromakalim, and diazoxide, were evaluated. All response data by nonlinear regression analysis three KCOs relaxed the tissue preparation in a concentration-dependent manner. The EC values (GraphPAD, San Diego, CA) to obtain EC50 values 50 as appropriate. Data are expressed as the mean for P1075, L-cromakalim and diazoxide were 0.064, Æ s.e.m. 0.75, and 85 mM, respectively (Figure 3). The rank order of potency of these compounds in this tissue