Effects of Various Selective Phosphodiesterase Inhibitors on Relaxation and Cyclic Nucleotide Contents in Porcine Iris Sphincter

FULL PAPER Pharmacology

Takuya YOGO1), Takeharu KANEDA2), Yoshinori NEZU1), Yasuji HARADA1), Yasushi HARA1), Masahiro TAGAWA1), Norimoto URAKAWA2) and Kazumasa SHIMIZU2)

1)Laboratories of Veterinary Surgery and 2)Veterinary Pharmacology, Nippon Veterinary and Life Science University, 7–1 Kyonan-cho 1– chome, Musashino, Tokyo 180–8602, Japan

(Received 26 March 2009/Accepted 1 July 2009)

ABSTRACT. The effects of various selective phosphodiesterase (PDE) inhibitors on muscle contractility and cyclic nucleotide contents in porcine iris sphincter were investigated. Forskolin and sodium nitroprusside inhibited carbachol (CCh)-induced contraction in a concentration-dependent manner. Various selective PDE inhibitors, vinpocetine (type 1), erythro -9-(2-hydroxy-3-nonyl)adenine (EHNA, type 2), milrinone (type 3), Ro20–1724 (type 4) and zaprinast (type 5), also inhibited CCh-induced contraction in a concentration-dependent manner. The rank order of potency of IC50 was zaprinast > Ro20–1724 > EHNA  milrinone > vinpocetine. In the presence of CCh (0.3 M), vinpocetine, milrinone and Ro20–1724 increased cAMP, but not cGMP, contents. In contrast, zaprinast and EHNA both increased cGMP, but not cAMP, contents. This indicates that vinpocetine-, milrinone- and Ro20–1724-induced relaxation is correlated with cAMP, while EHNA- and zaprinast- induced relaxation is correlated with cGMP in porcine iris sphincter. KEY WORDS: cAMP, cGMP, iris sphincter, PDE inhibitor, smooth muscle. J. Vet. Med. Sci. 71(11): 1449–1453, 2009

Cyclic nucleotides are important secondary messengers, chol (CCh)-induced contraction of porcine iris sphincter and are associated with smooth muscle relaxation [7]. induced by selective PDE (type1–5) inhibitors. Cyclic nucleotides are synthesized by adenyl cyclase or guanylyl cyclase and degraded by phosphodiesterases (PDEs). MATERIALS AND METHODS In the smooth muscles of the iris-ciliary body, it has been suggested that -adrenoceptor agonists [22, 24] and Muscle preparations and tension measurement: Eyes adrenomedullin [26] increase cAMP accumulation by stim- from adult pigs of either sex were obtained from a local ulating adenyl cyclase, and that atrial natriuretic peptide [9] abattoir. Two strips of iris sphincter muscle were cut from and nitric oxide [16] increase cGMP accumulation by stim- each eye (ciliary margin removed). The muscle strips were ulating guanylyl cyclase. incubated with physiological salt solution (PSS) containing Currently, PDEs are classified into 11 families [3, 10, 27] (in mM): NaCl, 136.8; KCl, 5.4; CaCl2, 2.5; MgCl2, 1.0; and selective PDE inhibitors have been identified [4]. It has NaHCO3, 11.9; and glucose, 5.6. PSS was aerated with 95% been reported that M&B 22948 (zaprinast, type5 inhibitor) O2 and 5% CO2 to adjust pH to 7.2 at 37C. Muscle tension inhibited the carbachol-induced contraction in bovine iris was recorded isometrically. One end of each strip was sphincter muscle [2]. However, to our knowledge, there bound to a glass holder and the other end was connected by have been no reports that compared the potency of maximal a silk thread to a strain-gauge transducer (TB-611T; Nihon relaxation of various selective PDE inhibitors-induced- Kohden, Tokyo, Japan) in an organ bath containing PSS relaxation in iris smooth muscle. with a resting tension of 0.5 g. Muscle strips were equili- The angle closure glaucoma is a main pathologic condi- brated for 30 min in order to obtain stable contractility tion with the elevation in intraocular pressure because of iri- induced by hyperosmotic 65 mM KCl. docorneal angle obstraction. When iridocorneal angle Assay of cGMP and cAMP contents: The cGMP and obstraction, the play of iris and ciliary muscle is also cAMP contents of muscle strips were measured by enzyme changed. To study the influence of PDE (type1–5) inhibi- immunoassay. After incubation of the strips with vinpocet- tors in the iris and ciliary muscle contraction may not only ine, EHNA, milrinone, Ro20–1724 or zaprinast for 10 min clarify the mechanism of cAMP or cGMP with PDE in the presence of CCh (0.3 M), the strips were rapidly fro- isozyme, but also contributes to the development of novel zen in liquid nitrogen and stored at –80C until homogeniza- antiglaucoma drug. tion in 6% trichloroacetic acid (0.4 ml). Homogenates were In the present study, we investigated the relationship centrifuged at 3000  g for 15 min and the supernatant was between relaxation and cyclic nucleotide contents in carba- washed with 1.5 ml of water-saturated diethylether 4 times; the cGMP and cAMP contents of the strips were then *CORRESPONDENCE TO: YOGO, T., Laboratory of Veterinary Surgery (e-mail: [email protected]), KANEDA, T., Laboratory of Veteri- assayed using enzyme immunoassay kits (GE Healthcare, nary Pharmacology, (e-mail: [email protected]), Nippon Vet- Buckinghamshire, UK). The cGMP and cAMP contents erinary and Life Science University, 7–1 Kyonan-cho 1-chome, were expressed in terms of picomole per gram wet weight of Musashino, Tokyo 180–8602, Japan. 1450 T. YOGO ET AL. tissue. Chemicals: The chemicals used were forskolin, sodium nitroprusside (SNP), milrinone, zaprinast, CCh (Sigma, St. Louis, MO, U.S.A.), vinpocetine, erythro-9-(2-hydroxy-3- nonyl)adenine•HCl (EHNA) (BIOMOL Research Laborato- ries, Plymouth Meeting, PA, U.S.A.) and Ro20–1724 (LC Laboratories, New Boston, MA, U.S.A.). Vinpocetine crys- tallizes if the concentration exceeds 30 M in PSS; therefore, we used vinpocetine at below 30 M in the present experiments. Statistics: Concentration-response curve for each inhibitor was fitted to a logistic equation using MS Excel; the fit- ting parameter was IC50. Values are expressed as means  S.E.M. and statistical analyses were performed by Student’s t-test.

RESULTS Fig. 1. Effects of forskolin and sodium nitroprusside on the Effects of forskolin, and SNP on CCh-induced contrac- contraction induced by 0.3 M carbachol (CCh) in porcine iris tion: When the contractile response induced by 0.3 M CCh sphincter. Preparations were precontracted with CCh, and the reached a steady level at about 15–20 min after addition, indicated agents were added cumulatively. The maximum con- forskolin, an adenyl cyclase activator, or SNP, a soluble traction induced by 0.3 M CCh in the absence of these agents was taken as 100%. Each point represents the mean of 6–7 guanylyl cyclase activator, was applied cumulatively. The preparations. Vertical bars indicate S.E.M. threshold concentrations of forskolin and SNP to induce the relaxation of the iris sphincter were 0.1 M and 0.03 M, respectively, and both inhibited CCh-induced contraction in a concentration-dependent manner (Fig. 1). The concentrations of forskolin and SNP producing 50% relaxation (IC50) of CCh-induced contraction were 0.68 (0.51–0.89) or 0.053 (0.045–0.061) M, respectively. Based on IC50 and maximum inhibition, SNP is more potent than forskolin in the porcine iris sphincter (P<0.01). Effects of various selective PDE inhibitors on CCh- induced contraction: When the contractile response induced by 0.3 M CCh reached a steady level at about 15–20 min after application, each PDE inhibitor was added cumulatively. Vinpocetine (type 1), EHNA (type 2), milrinone (type 3), Ro20–1724 (type 4) and zaprinast (type 5) inhibited CCh-induced contraction in a concentration-dependent manner (Fig. 2). The IC50 and maximum relaxation concentrations of these inhibitors are shown in Table 1. The rank orders for the relaxant effects of IC50 were zap- Fig. 2. Effects of vinpocetine, EHNA, milrinone, Ro20–1724 rinast (type 5) > Ro20–1724 (type 4) > EHNA (type 2)  and zaprinast on the contraction induced by 0.3 M CCh in milrinone (type 3) > vinpocetine (type 1) (P<0.05). porcine iris sphincter. Preparations were precontracted with Effects of various PDE inhibitors on cGMP and cAMP CCh and the indicated agents were added cumulatively. The contents: In the presence of CCh (0.3 M), vinpocetine, mil- maximum contraction induced by 0.3 M CCh in the absence rinone and Ro20–1724 increased cAMP contents. With of these agents was taken as 100%. Each point represents the these drugs, cAMP contents increased by 1.47 (vinpocet- mean of 5–8 preparations. Vertical bars indicate S.E.M. ine), 1.6 (milrinone) and 1.5 (Ro20–1724) times above resting levels. cAMP contents were not affected by EHNA and DISCUSSION zaprinast (Fig. 3A). In contrast, zaprinast and EHNA both increased cGMP contents by 1.57 (EHNA) and 2.15 (zapri- Physiological iris contractions are mediated by the nona- nast) times above resting levels. cGMP contents were not drenergic, noncholinergic nervous system in addition to the affected by vinpocetine, milrinone or Ro20–1724 (Fig. 3B). adrenergic, cholinergic nervous system. The stimulators of adenyl cyclase, -adrenoceptor agonists [22, 24] and adrenomedullin [26], induce relaxation by increasing cAMP contents, and the stimulators of guanylyl cyclase, atrial PHOSPHODIESTERASE INHIBITORS IN IRIS SPHINCTER 1451

natriuretic peptide [9] and nitric oxide [16], induce relax- Table 1. IC50 and maximum relaxation concentrations for various selective PDE inhibitors in porcine iris sphincter treated with 0.3 ation via cGMP in the iris sphincter muscle. In the present M CCh study, forskolin and SNP inhibited the muscarinic agonist carbachol (CCh)-induced contraction in a concentration- Agent IC50 (M) Maximum relaxation (%) n dependent manner (Fig. 1). These results suggest that the Vinpocetine >30 43.9  4.6 5 cAMP/adenyl cyclase and cGMP/guanylyl cyclase path- EHNA 92.4 (89.4 –98.6) 53.9  1.6 6 ways are both involved in the relaxation of porcine iris Milrinone 99.4 (54.0 – >100) 52.8  4.0 8 Ro20–1724 58.3 (36.9 – 95.0) 58.4  3.4 6 sphincter muscle. However, the activitys of PDE isozymes Zaprinast 4.5 (2.8–7.2) 88.4  4.8 5 which hydrolysis cAMP and/or cGMP in smooth muscle of iris-ciliary body have not been still clear. Therefore, we Numbers in parentheses indicate 95% confidence limits. Maximum relaxation refers to resting tension after washing, which was arbitrarily investigated the effects of selective PDE inhibitors on CCh- set to 100%. induced contraction in order to confirm the role of PDE isozymes in the mediation of iris contraction. PDEs, which play a major role in the regulation of intrac- ellular concentrations of cAMP and cGMP, are classified into 11 families [3, 10, 27]. Of these, PDE1–5 isozymes have selective inhibitors. It has been shown that the potency of relaxation induced by these PDE inhibitors different in vascular [17], tracheal [19, 20], urinary [18, 25] and intesti- nal smooth muscle [13, 14, 23]. It has been reported that vinpocetine (type 1 inhibitor) induces relaxation in rabbit aorta by increasing cGMP contents [1, 11], but in rat urinary bladder [18] and guinea pig taenia coli [14], vinpocetine acts by increasing cAMP contents. In the present study, vinpocetine inhibited CCh- induced contraction and increased cAMP, but not cGMP, contents in porcine iris sphincter muscle. These results suggest that the relaxation induced by PDE type 1 inhibitors involves increased cAMP contents in porcine iris sphincter muscle. The PDE1 family includes PDE1A, 1B and 1C. PDE1A and 1B have high affinity for cGMP and PDE1C has similar affinity for cAMP and cGMP [3]. It is known that PDE1 hydrolyzes cGMP in some smooth muscle cells [20], and in the porcine iris sphincter muscle, vinpocetine appears to inhibit PDE1C, but not PDE1A and B. It has been suggested that EHNA (type 2 inhibitor) relaxes rat pulmonary artery tone by increasing cAMP and cGMP contents [12]. It has also been reported that the relaxation induced by EHNA involves the increase in cGMP contents in rat urinary bladder [18] and guinea pig taenia coli [14]. In the present study, EHNA inhibited CCh-induced contraction and increased cGMP, but not cAMP, contents in porcine iris sphincter muscle. These results suggest that the relaxation induced by PDE type 2 inhibitors involves increased cGMP contents. The present data showed that the relationship between type 1 and type 2 inhibitor-induced relaxation, and the increase in cyclic nucleotides in porcine iris sphincter muscle, is similar to that in rat urinary bladder [18] and guinea pig taenia coli [14]. Milrinone (type 3 inhibitor) increased cAMP contents and inhibited phenylephrine-induced contraction in a con- Fig. 3. Effects of vinpocetine (Vin, 30 M), EHNA (100 M), centration-dependent manner in guinea pig aorta [21]. In milrinone (Mil, 100 M), Ro20–1724 (Ro, 100 M) and zapri- the present study, milrinone inhibited CCh-induced contrac- nast (Zap, 100 M) on cAMP contents of porcine iris sphincter. Preparations were contracted with 0.3 M CCh and were tion and increased cAMP, but not cGMP, contents in por- treated with the indicated agents for 10 min. Vertical bars indi- cine iris sphincter muscle. These results suggest that the cate S.E.M. *, ** Significant difference from each respective characterization of PDE isozymes varies with organ, and control at P<0.05, P<0.01. that activity in iris sphincter muscle is smaller that in other 1452 T. YOGO ET AL. smooth muscle types. sphincter muscle. It is known that PDE4 is specific for the hydrolysis of cAMP [3], and it has been reported that rolipram (type 4 REFERENCES inhibitor) relaxes bovine trachea by increasing cAMP contents [20]. Moreover, rolipram inhibits lipopolysaccharide- 1. Ahn, H.S., Crim, W., Romano M., Sybertz, E. and Pitts, B. induced uvetitis in rats [5]. In the present study, Ro20–1724 1989. Effects of selective inhibitors on cyclic nucleotide phos- inhibited CCh-induced contraction (Fig. 2) and increased phodiesterases of rabbit aorta. Biochem. Pharmacol. 38: 3331– cAMP, but not cGMP, contents (Fig. 3) in porcine iris 3339. 2. Akhtar, R.A. and Abdel-Latif, A.A. 1991. The effect of M&B sphincter muscle. These data suggest that Ro20–1724 (type 22948 on carbachol-indced insitol triphosphate accumulation 4 inhibitor) induces relaxation by increasing cAMP con- and contraction in iris sphincter smooth muscle. Eur. J. Phar- tents. macol. 206: 291–295. In the present study, relaxation induced by zaprinast was 3. Beavo, J.A. 1995. Cyclic nucleotide phosphodiesterase : func- more potent than that with other PDE inhibitors, and zapri- tional implications of multiple isoforms. Physiol. Rev. 75: 725– nast increased cGMP contents. Zaprinast (M&B22948) also 748. inhibited CCh-induced contraction in bovine iris sphincter 4. Beavo, J.A. and Reifsnyder, D.H. 1990. Primary Sequence of muscle [2]. These data are consistent with the report that cyclic nucleotide phosphodiesterase isozymes and the design zaprinast relaxed guinea pig taenia coli and ileum rat aorta of selective inhibitors. Trend. Pharmacol. Sci. 11: 150–155. by increasing cGMP contents [13, 14]. 5. Chi, Z.L., Hayasaka, S., Zhang, X.Y., Hayasaka, Y. and Cui, H.S. 2004. Effects of rolipram, a selective inhibitor of type 4 In guinea pig tracheal smooth muscle, selective PDE phosphodiesterase, on lipopolysaccharide-induced uveitis in inhibitors reduced histamine-induced contraction and the rats. Invest. Ophthalmol. Vis. Sci. 45: 2497–2502. rank order of inhibitory potency was rolipram (type 4) > 6. Chuman, H., Chuman, T., Nao-i, N., Sawada, A., Yamamoto, siguazodan (type 3) > zaprinast (type 5) > vinpocetine (type R., Kobayashi, H. and Wada, A. 1997. Different responsive- 1) [19]. In rat urinary bladder, selective PDE inhibitors ness to nitric oxide-cyclic guanosine monophosphate pathway reduced CCh-induced contraction and the rank order of in cholinergic and tachykinergic contractions of the rabbit iris inhibitory potency was vinpocetine (type 1), zaprinast (type sphincter muscle. Invest. Ophthalmol. Vis. Sci. 38: 1719–1725. 5) > EHNA (type 2) > rolipram (type 4) > milrinone (type 3) 7. Diamond, J. 1978. Role of cyclic nucleotides in control of [18]. In guinea pig taenia coli, selective PDE inhibitors smooth muscle contraction. pp. 27–340. In: Advances in reduced CCh-induced contraction and the rank order of Cyclic Nucleotide Research vol. 9 (George, W.J. and Ignarro, L.J. eds.), Raven Press, New York. 3. inhibitory potency was zaprinast (type 5) > vinpocetine 8. Ding, K.H., Ali, N. and Abdel-Latif, A.A. 1999. Atrial natri- (type 1) > EHNA (type2) > Ro20–1724 (type 4) > milrinone uretic peptide provokes a dramatic increase in cyclic GMP for- (type 3). In the present study, selective PDE inhibitors mation and markedly inhibits muscarinic-stimulated Ca2+ reduced CCh-induced contraction in porcine iris sphincter mobilisation in SV-40 transformed cat iris sphincter smooth muscle and the rank order of inhibitory potency of IC50 was muscle (SV-CISM-2) cells. Cell. Signal. 11: 87–94. zaprinast (type 5) > Ro20–1724 (type 4) > EHNA (type 2)  9. Ding, K.H. and Abdel-Latif, AA. 1997. Actions of C-type milrinone (type 3) > vinpocetine (type 1). As the rank order natriuretic peptide and sodium nitroprusside on carbachol- of inhibitory potency varies with organ, the characterization stimulated inositol phosphate formation and contraction in cili- of PDE isozymes is also apparently different in each organ. ary and iris sphincter smooth muscles. Invest. Ophthalmol. Vis. It is known that the motility of iris and ciliary muscle Sci. 38: 2629–2638. 10. Fawcett, L., Baxendale, R., Stacey, P., McGrother, C., Harrow, relate to mediate intraocular pressure. NO donors reduced I., Soderling, S.H. and Beavo, J.A. 2000. Molecular cloning intraocular pressure in normal rabbit eyes without systemic and characterization of distinct human phsphodiesterase gene effects [15], and relax iris and ciliary muscle contraction [6, family: PDE11A. Proc. Natl. Acad. Sci. U.S.A. 97: 3702–3707. 8]. In the present study, SNP inhibited CCh-induced con- 11. Hagiwara, M., Endo, T. and Hidaka, H. 1984. Effects of vinpo- traction in porcine iris sphincter muscle, too. Moreover, the cetine on cyclic nucleotide metabolism in vascular smooth type 5 PDE inhibitor, zaprinast was the most potent in inhi- muscle. Biochem. Pharmacol. 33: 453–457. bition of CCh-induced contraction. These results imply that 12. Haynes, J. Jr., Killiea, D.W., Peterson, P.D. and Thompsom, J. type 5 PDE is important to mediate the motility of iris and 1996. Erthro-9-(2-hydroxy-3-nonyl)adenine inhibits cyclic- ciliary muscle and intraocular pressure. However, it 3’,5’-guanosine monophosphate-stimulated phosphodiesterase remains to clear the mediation of other type PDEs in these to reverse hypoxic pulmonary vasoconstriction in the perfused rat lung. J. Pharmacol. Exp. Ther. 276: 752–756. functions. 13. Kaneda, T., Shimizu, K., Nakajyo, S. and Urakawa, N. 1997. In conclusion, selective PDE inhibitors reduced CCh- Effects of various selective phosphodiesterase inhibitors on induced contraction in porcine iris sphincter muscle and the muscle contractility in guinea pig ileal longitudinal smooth rank order of inhibitory potency of IC50 was zaprinast (type muscle. Jpn. J. Pharmacol. 75: 77–85. 5) > Ro20–1724 (type 4) > EHNA (type 2)  milrinone (type 14. Kaneda, T., Shimizu, K., Urakawa, N. and Nakajyo, S. 2004. 3) > vinpocetine (type 1). Moreover, it was suggested that Effects of various selective phosphodiesterase inhibitors on vinpocetine-, milrinone- and Ro20–1724-induced relaxation carbachol-induced contraction and cyclic nucleotide contents was correlated with cAMP, and that EHNA- and zaprinast- in guinea pig taenia coli. J. Vet. Med. Sci. 66: 1047–1052. induced relaxation was correlated with cGMP in the iris 15. Nathanson, J.A. 1992. Nitrovasodilators as a new class of ocu- lar hypotensive agents. J. Pharmacol. Exp. Ther. 260: 956– PHOSPHODIESTERASE INHIBITORS IN IRIS SPHINCTER 1453

965. bovine iris sphincter smooth muscle. Biochemical and func- 16. Pianka, P., Oron, Y., Lazar, M. and Geyer, O. 2000. Nonadren- tional interactions between the cyclic AMP and calcium signal- ergic, noncholinergic relaxation of bovine iris sphincter: role of ling systems. Invest. Ophthalmol. Vis. Sci. 30: 2232–2239. endogenous nitric oxide. Invest. Ophthalmol. Vis. Sci. 41: 880– 23. Tomkinson, A. and Raeburn, D. 1996. The effect of isoen- 886. zyme-selective PDE inhibitors on methacholine-induced con- 17. Polson, J.B. and Strada, S.J. 1996. Cyclic nucleotide phos- traction of guinea-pig and rat ileum. Br. J. Pharmacol. 118: phodiesterases and vascular smooth muscle. Ann. Rev. Phra- 2131–2139. macol. Toxicol. 36: 403–427. 24. Topalkara, A., Karadas, B., Toker, M.I., Kaya, T., Durmus, N. 18. Qiu, Y., Kraft, P., Craig, E.C., Liu, X. and Haynes-Johnson, D. and Turgut, B. 2006. Relaxant effects of beta-adrenoceptor 2001. Identification and functional study of phosphodiesterases agonist formoterol and BRL 37344 on bovine iris sphincter and in rat urinary bladder. Urol. Res. 29: 388–392. ciliary muscle. Eur. J. Pharmacol. 548: 144–149. 19. Raeburn, D. and Advenier, C. 1995. Isoenzyme-selective 25. Truss, M.C., Ückert, S., Stief, C.G., Kuczyk, M., Schulz- cyclic nucleotide phosphodiesterase inhibitors: Effects on air- Knappe, P., Forssmann, W.-G. and Jonasw, U. 1996. Effects of ways smooth muscle. Int. J. Biochem. Cell Biol. 27: 29–37. various phosphodiesterase-inhibitors, forskolin, and sodium 20. Shahid, M, van Amsterdam, R.G., de Boer, J., ten Berge, R.E., nitroprosside on porcine detrusor smooth muscle tonic Nicholson, C.D. and Zaagama, J. 1991. The presence of five responses to muscarinergic stimulation and cyclic nucleotide cyclic nucleotide phosphodiesterase isoenzyme activities in levels in vitro. Neurourol. Urodyn. 15: 59–70. bovine tracheal smooth muscle and the functional effects of 26. Yousufzai, S.Y., Ali, N. and Abdel-Latif, A.A. 1999. Effects of selective inhibitors. Br. J. Pharmacol. 104: 471–477. adrenomedullin on cyclic AMP formation and on relaxation in 21. Silver PJ, Lepore RE, O’Connor B., Lemp, B.M., Hamel, L.T., iris sphincter smooth muscle. Invest. Ophthalmol. Vis. Sci. 40: Bentley, R.G. and Harris, AL. 1988. Inhibition of the low Km 3245–3253. cyclic AMP phosphodiesterase and activation of the cyclic 27. Yuasa, K., Kotera, J., Fujishige, K., Michibata, H., Sasaki, T. AMP system in vascular smooth muscle by milrinone. J. Phar- and Omori, K. 2000. Isolation and characterization of two macol. Exp. Ther. 247: 34–42. novel phosphodiesterase PDE11A variants showing unique 22. Tachado, S.D., Akhtar, R.A. and Abdel-Latif, A.A. 1989. Acti- structure and tissue-specific expression. J. Biol. Chem. 275: vation of beta-adrenergic receptors causes stimulation of cyclic 31469–31479. AMP, inhibition of inositol trisphosphate, and relaxation of