Smooth Muscle Relaxing and Hypotensive Activities of Synthetic Calciseptine and the Homologous Snake Peptide FS2

Takushi X. Watanabe, Yukako Itahara, Hisaya Kuroda, Yun-Neng Chen, Terutoshi Kimura and Shumpei Sakakibara Peptide Institute, Inc., Protein Research Foundation, 4-1-2 Ina, Minoh, Osaka 562, Japan D..,.,.:..,.,11,.,....,. C 1000 A--I-4 A-;1 '7G 1005

ABSTRACT-The biological activities of synthetic calciseptine and FS2, a homologous peptide from , were determined using in vitro and in vivo preparations. Calciseptine and FS2 produced dose- dependent relaxation in pre-constricted rat aorta, pulmonary artery and . The onset and duration pattern of these relaxing effects were similar to those caused by , an L-type Ca 21 channel blocker. Calciseptine relaxed the contraction of rat aorta provoked by an L-type channel agonist, Bay K 8644. This relaxation was not affected by 1V~-nitro-L-arginine, indomethacin or propranolol. Calciseptine and FS2 in- hibited the contraction caused by acetylcholine in guinea pig ileal longitudinal muscle. In case of in vivo study using anesthetized rats, calciseptine, FS2 and nifedipine showed depressor effects. The hypotensive effects of the two peptides were more potent and sustained than that of nifedipine. These findings show that both synthetic calciseptine and FS2 have similar biological activities like nifedipine, an L-type Ca 21 channel blocker. In addition, these two peptides with large molecular weights may be unique and useful tools for studying the Ca 21 channel.

Keywords: Calciseptine, FS2 (peptide venom), Ca" channel blocker (L-type), Spasmolytic activity, Hypotensive activity

Voltage-dependent Ca 21 channels play important roles should be the extracellular surface for the Ca 21 channel in the transmitting of signals by neurons and in the con- because of its large molecular weight. This is useful for traction of cardiac or smooth muscle cells. There are four studying the Ca 21 channel because the binding site for the types (L-, N-, T- and P-types) of voltage-dependent Ca 21 Ca 2+ blockers that are small organic molecules is still con- channels (1). The L-type channel is widely distributed in troversial. The receptor site for the 1,4-dihydropyridines excitable cells and has the largest conductance. L-Type is proposed to be in the extracellular surface domain of channel blockers that are organic molecules, such as 1,4- the a,-subunit of the channel (6). In contrast, Regulla et dihydropyridines and phenylalkylamines, have been deve- al. (7) reported that the 1,4-dihydropyridines bind to the loped and used for the treatment of cardiovascular dis- putative cytosolic domain of the Ca2+ channel. orders and other diseases (2). We searched for peptides homologous to calciseptine The selective blockers are very important as pharma- in the data base PRF/SEQDB of the Protein Research cological tools for studying the physiological role of the Foundation. The search identified snake venom peptide subtypes of Ca2+ channels. The N-type Ca 21 channel in FS2 as a homologous peptide (8) (Fig. 1); This peptide is neural cells is blocked by a natural peptide , w- a three substituted homolog with much less GVIA (3). A peptide toxin from the funnel- toxicity than other snake (9). Although FS2 is the web , w-agatoxin IVA, has been recently reported third most abundant component isolated from black to block the P-type Ca 21 channel present in cerebellar venom, the pharmacological properties of FS2 Purkinje cells (4). had not been studied precisely so far. Although there are many L-type blockers as mentioned We have synthesized calciseptine and FS2 by the solu- above, no natural substance had been known until the re- tion procedure. In this study, we examined its biological cent discovery of calciseptine in venom (5). activities in vitro and in vivo. The activities of synthetic This peptide consists of 60 amino acid residues with eight calciseptine and FS2 were determined in rat aorta, pulmo- cysteine residues (Fig. 1). The binding site of calciseptine nary artery and trachea and also in guinea pig ileum. We v

Fig. 1. Amino acid (single-letter code) sequences of calciseptine and FS2. Each peptide consists of 60 amino acid residues with eight cysteine residues, respectively. The differences of the amino acid residues are at positions 5, 30 and 32. The intramolecular disulfide bonds of calciseptine and FS2 were the same compared to other short snake (18) and the details of disulfide bond will be described elsewhere.

also studied the effect of these peptides on blood pressure tide. in anesthetized rats. The activities of the two peptides were compared with those of a control drug, nifedipine, Relaxing activities on vasculatures and trachea as a common L-type Ca" blocker. The relaxing activities on precontracted smooth mus- A part of the preliminary account of this work about cles were determined in rat thoracic aorta, pulmonary calciseptine has been given in abstracts (10). artery and trachea preparations. Smooth muscle contrac- tions were determined as reported previously (14). Tho- racic aortas, pulmonary arteries and were removed from decapitated male Sprague-Dawley rats (260 - 290 g) Synthesis of peptides and placed in Krebs-Henseleit solution of the following Calciseptine and FS2 were synthesized as reported previ- composition: 118 mM NaCI, 4.7 mM KCI, 2.5 mM CaC12 ously (11). Briefly, all peptides were synthesized by means •2H2O , 0.6 mM MgSO4.7H2O, 1.2 mM KH2PO4, 25 mM of the solution procedure applying the maximum protec- NaHCO3 and 11 mM glucose. The smooth muscles were tion strategy (12) and recently deprotected solvent systems cleaned of adipose and connective tissue and cut into ring that are useful for the synthesis of slightly soluble pep- preparations (about 3 mm). The pulmonary arterial rings tides (13). The fully protected peptides were unblocked by were isolated from the right branch. The trachea ring anhydrous hydrogen fluoride and oxidized for disulfide (two to three cartilages) was isolated from the portion just bond formation in ammonium acetate buffer in the above the bronchus. These ring segments were mounted presence of reduced and oxidized glutathione on stainless steel wires and suspended from force displace- (GSH/GSSG). The crude products with disulfide bonds ment transducers in tissue baths with a water-jacket were purified by ion-exchange high performance liquid (371C) and containing 10 ml Krebs-Henseleit medium, chromatography (HPLC), followed by preparative gassed with (95% 02-5% C02). The tissues were reversed-phase HPLC. The homogeneity of the final equilibrated for 1 hr under an optimal resting load of 0.5 product was confirmed by amino acid analysis, thin layer g (aortas) or 0.3 g (pulmonary arteries and tracheas), maxi- chromatography, reversed-phase HPLC, ion-exchange mally contracted with 80 mM KCI, and then washed and HPLC and capillary zone electrophoresis. The purity of allowed to dilate to baseline tension. The changes of mus- each preparation was over 99% of the final purified pep- cle tension were measured with an isometric recording sys- tern (Star-Medical SUL2GR, PA-011; San-Ei 8K21; Krebs-Henseleit medium. The preparations were Tokyo). Before the determination of muscle relaxation, equilibrated for 1 hr under an optimal resting load of 0.5 smooth muscles were precontracted with 40 mM KC1 as g, maximally constricted with 1 ,uM ACh, then washed reported previously (5). In these experiments, 100070re- and allowed to dilate to resting tension. Muscle tension laxation represents the resting tension before the muscle was measured with an isotonic recording system (Star- contraction. Medical IA-001, San-Ei 8K21). Contracting responses by ACh (1 pM) were determined 10 min after calciceptine Inhibiting activities on ileum contraction and FS2 treatment. Inhibiting activities were expressed as The inhibiting activities on acetylcholine (ACh) con- a percentage of the responses compared to ACh (1 pM) traction were determined in guinea pig Real longitudinal contraction before the peptide treatments. muscle preparations (15). Briefly, ilea were removed from decapitated male guinea pigs (600 - 800 g). Segments Hypotensive activities (approximately 20 mm) were cut at the portion about 100 Male Sprague-Dawley rats (260 - 290 g) were anesthe- mm proximal to the ileo-caecal junction and placed in tized with pentobarbital (50 mg/kg, i.p.), and a Krebs-Henseleit solution. These segments were cleaned of tracheotomy was performed. Arterial blood pressure was extraneous connective tissue and mesenteric arteries. The determined directly through a femoral arterial cannula. ileal longitudinal muscle preparations were suspended on Heart rate was continuously monitored by counting the isotonic transducers in tissue baths (10 ml) containing pulse rate of this aortic pressure. Blood pressure and

Fig. 2. Vasorelaxing effects of calciseptine (CaS) and FS2 on precontracted rat thoracic aorta induced by 40 mM KC1. heart rate were monitored with an electronic system (Century Technology CP-01, Inglewood, CA, USA; Star Medical PA-011, Star Medical HR-001; San-Ei 8K21). Peptides were injected as a bolus (0.5 ml/kg) through a femoral venous cannula.

Statistical analyses The results are given as the mean±S.E. with the number of experiments in parentheses. The statistical significance of the results was evaluated by the unpaired Student's t-test. P<0.05 was taken as the statistically sig- nificant value.

Drugs The drugs used and their sources were acetylcholine chloride, indomethacin and nifedipine (Sigma Chemical Co., St. Louis, MO, USA); Bay K 8644 (Research Bio- chemicals Incorporated, Natick, MA, USA); 1V'-nitro-L- Fig. 3. Dose-response curves for the relaxation of rat thoracic arginine (Peptide Institute, Osaka); pentobarbital sodium aorta rings by calciseptine, FS2 and nifedipine. The relaxing (Abbott Laboratories, North Chicago, IL, USA); responses were elicited by cumulative administration of calciseptine and propranolol hydrochloride (ICI-Pharma, Plankstadt, (•, n=9), FS2 (A, n=6) and nifedipine (0, n=7) to aortas precon- Germany). tracted with 40 mM KCI. Each point represents the mean and the bar shows S.E. In the experiments, 100% relaxation represents the rest- ing tension before the muscle contraction. *Significantly different (P<0.05) from the relaxation observed with nifedipine.

Relaxing activities in vasculatures The biological activities of synthesized calciseptine and the natural product as reported previously (5). The onset FS2 were measured in precontracted rat thoracic aorta was within two minutes, and it took 40 - 50 min to reach (Fig. 2), by the same method as used for natural calcisep- the plateau. The common L-type Cat' blocker nifedipine tine (5). The onset and duration pattern of the relaxing showed a similar pattern of activity (data not shown). The effect of synthetic calciseptine were the same as those of cumulative dose-response curves of calciseptine, FS2 and

Fig. 4. Dose-response curves for the relaxation of rat pulmonary artery rings by calciseptine (0, n=6), FS2 (A, n=6) and nifedipine (0, n=7). Details are the same as in Fig. 3. nifedipine were obtained (Fig. 3). At 10 nM-1 1M, cal- rat aorta and pulmonary artery precontracted with 0.1 ciseptine and FS2 showed dose-related relaxation on teM norepinephrine (data not shown). The patterns of precontracted rat aortas. The average ED50 values in this relaxation were the same as those observed in the same experiment were 74.9:L- 13.9 nM (n = 9) and 93.6:L 8.0 nM vasculature precontracted by 40 mM KCl as shown in (n=6) for calciseptine and FS2, respectively. Nifedipine Fig. 2. The onset was within two min, and it took 40-50 had a stronger potency than these peptides, with an ED50 min to reach the plateau (approximately 90% relaxation). value of 47.7.± 8.8 nM (n = 7) in this preparation. These results indicate that the Ca 21 channel is activated Calciseptine and FS2 demonstrated vasorelaxing activ- by adrenergic receptor as reported previously (16), and ity in the pulmonary artery. The cumulative dose- calciseptine blocks this activated Ca 21 channel. response curves are shown in Fig. 4. Calciseptine caused dose-dependent relaxation with an ED50 value of Relaxing activity in Bay K 8644 contracted aorta 62.5 ± 9.2 nM (n=6). FS2 showed a stronger tendency We observed the relaxing effect of synthetic calciseptine than calciseptine, with an ED50 value of 44.5±2.9 nM on precontracted rat aorta by Bay K 8644, a specific L- (n=6). The ED50 value of nifedipine was 79.4 ± 19.1 nM type Ca 2+ channel agonist (Fig. 5). The rat aorta was (n = 7). pretreated with 8 mM KC1 to induce contractility. Other Synthetic calciseptine (0.1 1M) produced relaxation in vasodilatation was prevented by pretreating (approximate-

Fig. 5. Vasorelaxing effect of calciseptine (CaS) on rat thoracic aorta ring preparation precontracted with Bay K 8644 (Bay). The aorta was pretreated with 8 mM KCl for inducing the contractility of Bay K 8644. Relaxation other than that involving the L-type Ca 2+ channel was prevented by treatment with propranolol (PPL), indomethacin (IM) and M -nitro-L-arginine (NO2- Arg). This typical data represents 6 similar responses. ly 20 min before) the preparation with three blockers: pro- The ED50 values were 32.6 ± 7.5 nM (n = 5), 51.7::L 13.7 pranolol (10 rM) to block the f3-adrenergic receptor, indo- nM (n=6) and 15.4± 1.8 nM (n=6) for nifedipine, FS2 methacin (5 rM) to block the synthesis of prostaglandins and calciseptine, respectively. and 1VG-nitro-L-arginine (100 p M) to block the nitric oxide (NO) synthase. The upper panel of Fig. 5 shows the con- Hypotensive activities trol response of Bay K 8644 contraction in rat aorta. Con- Calciseptine, FS2 and nifedipine showed depressor traction was sustained until the preparation was washed. activity in anesthetized rats. The mean blood pressure Calciseptine (0.1 1M) completely blocked the contraction level of anesthetized rats before peptide administration produced by Bay K 8644 (lower panel of Fig. 5). These was 90-100 mmHg. The dose-response curves of acute findings indicate that calciseptine specifically blocks the hypotensive effects are shown in Fig. 8. The responses L-type Ca" channel. FS2 showed the same result in rat represented changes in blood pressure within 5 min after aorta (data not shown). administration of the peptides and nifedipine. The hypotensive potencies of calciseptine and FS2 were greater Relaxing activities in trachea than that of nifedipine. At a higher dose, FS2 showed In the trachea ring precontracted with 40 mM KCI, stronger activity than calciseptine. calciseptine at 0.1 pM also showed relaxing activity. The The time course study of the hypotensive effect is pattern of relaxation was similar to that observed with the shown in Fig. 9. The mean blood pressure level of the vasculature preparation. The cumulative dose-response anesthetized rats before peptide administration was curves of calciseptine, FS2 and nifedipine are shown in 90-100 mmHg. The doses used in this study were equipo- Fig. 6. Calciseptine, FS2 and nifedipine relaxed precon- tent with those in Fig. 8. The durations of the hypotensive tracted tracheas dose-dependently. The order of relative activities of calciseptine and FS2 were longer than that of potency in the trachea was FS2 > calciseptine > nifedi- nifedipine. In addition, these peptides showed biphasic pine. The ED50 values were 6.94-±:1.08 nM (n=6), depressor effects; acute within 5 min, followed by a long 9.75 ± 2.93 nM (n = 7) and 24.3 ± 3.3 nM (n = 6) for FS2, sustained phase. Nifedipine showed only the acute calciseptine and nifedipine, respectively. hypotensive phase. The duration times were 20, 80 and 120 min or more for nifedipine, FS2 and calciseptine, Inhibiting activities on ACh contraction in ileum respectively. The second phase of depressor activity was Calciseptine inhibited the contraction with ACh (0.1 stronger with calciseptine than FS2. The hypotensive ac- ,uM) in guinea pig ileal longitudinal muscle. The dose- tivity was not inhibited by propranolol (1 pmol/kg, i.v.), response curves of calciseptine, FSZ and nifedipine for the indomethacin (5 mg/kg, i.v.) or IV'-nitro-L-arginine (10 inhibiting effects on ACh contraction are shown in Fig. 7. pmol/kg, i.v.) (data not shown).

Fig. 6. Dose-response curves for the relaxation of rat trachea rings by calciseptine (0, n=7), FS2 (A, n=6) and nifedipine (O, n=6). Details are the same as in Fig. 3. Fig. 7. Dose-response curves for the inhibiting activities of calciseptine (41, n=6), FS2 (A, n=6) and nifedipine (0, n=5) on the contraction with ACh in guinea pig ileal longitudinal muscle. Details are the same as in Fig. 3.

Calciseptine caused only a small increase in heart rate dose-dependent. Nifedipine showed a similar effect on (approximately 10-40 beats/min), although it had a heart rate, like calciseptine. strong hypotensive effect. This finding agrees with the comment of De Weille et al. (5) that calciseptine leads to very significant decreases of arterial pressure with only small effects on cardiac rhythm. This increase in heart We have demonstrated that our synthetic calciseptine rate was not dose-dependent (data not shown). FS2 showed the same relaxing activity on precontracted rat caused a small heart rate change, and the effect was not aorta as the natural calciseptine. The synthetic calcisep- tine relaxed Bay K 8644 contracted rat aorta. Its relaxing activities were similar to those of nifedipine in various in vitro and in vivo preparations. Our synthetic calciseptine binds to a 1,4-dihydropyridine recognition site of the L- type Ca" channel in rat synaptosomal membranes (17). These data demonstrated that synthetic calciseptine is a specific inhibitor of L-type Ca" channels. The biological activities of synthetic FS2 were very simi- lar to those of the calciseptine both in vitro and in vivo. In addition, FS2 bound to a 1,4-dihydropyridine recogni- tion site of the L-type Ca 21 channel in rat synaptosomal membranes (0. Yasuda et al., personal communication), like calciceptine. These findings indicate that FS2 is a specific inhibitor of the L-type Ca 21 channel. Calciseptine and FS2 caused a sustained decrease in blood pressure, with the effect being biphasic: short-last- ing, followed by a long-lasting phase. In contrast, nifedi- pine showed only a short-lasting decrease, and the decreased response recovered to the blood pressure level before the administration of nifedipine within 20 min. Fig. 8. Dose-response curves of acute hypotensive effects of These biphasic decreases in blood pressure by the two pep- calciseptine (0, n=7), FS2 (0, n=6) and nifedipine (0, n=6) in anesthetized rats. Mean blood pressure levels of anesthetized rats tides were not abolished by pre-administered proprano- before peptide administration were 90-100 mmHg. Significant lol, indomethacin or N -nitro-L-arginine. The major differences are the same as in Fig. 3. mechanism for the hypotensive effects of these peptides Fig. 9. Time course study of the hypotensive effects of calciseptine (4b, 20 nmol/kg, n=7), FS2 0, 20 nmol/kg, n=5) and nifedipine (0, 200 nmol/kg, n = 6) in anesthetized rats. Other details are the same as in Fig. 8.

may be caused by the peripheral vasorelaxing activities of the potencies of the peptides in the different tissues. The these Ca 21 channel inhibitions, because changes in heart differences in the mechanisms of action of these two rate were small. The first phase may be caused by the peptides will be clarified by using patch clamp or other main site of the Ca 21 channel inhibition, because the pat- methods. terns of the depressor effects of the two peptides are simi- In conclusion, the studies in this report show the lar to that of nifedipine. The second phase of calciseptine synthetic calciseptine and FS2 to be specific inhibitors of or FS2 may be caused by its interaction with a binding site the L-type Ca 21 channel. These two peptides, both large different from the Ca2+ channel and/or different sensitiv- molecules, would be useful tools for studying the binding ity of the vasculature to the relaxing activity. The precise site of the L-type Ca2+ channel. Furthermore, in the site of actions of these two phases of hypotensive effects future, we must perform a structure-activity relationship remains to be investigated. study on calciseptine by substituting amino acid residues The pharmacological profiles of FS2 in different assay around positions 30 and 32. preparations were similar to those of calciseptine. These results may be due to the sequence homology of the two Acknowledgments We are grateful to Drs. Osamu Yasuda, Shigeto Morimoto and peptides. However, there are differences in the charged Toshio Ogihara of Osaka University Medical School for determining group of the side chain of these two peptides at positions the binding activity of FS2 to rat synaptosomal membranes. 30-32 (Fig. 1). FS2 has a more positive charge than cal- ciseptine around this area. Calciseptine has the acidic side chain of glutamic acid with a negative charge at position 32. In contrast, FS2 has the basic side chain of histidine 1 Spedding M and Paoletti R: Classification of calcium channels with a positive charge at position 30. These differences of and the sites of action of drugs modifying channel function. the charged group may influence the potency of the two Pharmacol Rev 44, 363 - 376 (1992) peptides. The potencies of FS2 differed from those of cal- 2 Janis RA, Silver PJ and Triggle DJ: Drug action and cellular cal- ciseptine both in vitro and in vivo. In pulmonary artery, cium regulation. Adv Drug Res 16, 309-591 (1987) 3 Gray WR, Olivera BM and Cruz LJ: Peptide toxins from veno- trachea and ileum, FS2 showed stronger potency than that mous Conus snails. Annu Rev Biochem 57, 665-700 (1988) of calciseptine. In hypotensive activities, FS2 showed 4 Mintz IM, Venema VJ, Swiderek KM, Lee TD, Bean BP and stronger activity than calciseptine at a high dose. These Adams ME: P-type calcium channels blocked by the spider results indicate that the conformation of these two pep- toxin w-Aga-IVA. Nature 355, 827-829 (1992) tides might be different and/or the effect of charged 5 De Weille JR, Schweitz H, Maes P, Tartar A and Lazdunski M: groups on the binding sites of L-type Ca 21 channels might Calciseptine, a peptide isolated from black mamba venom, is a be slightly different, and these differences may influence specific blocker of the L-type calcium channel. Proc Natl Acad Sci USA 88, 2437-2440 (1991) [hPTH(1 - 84)]. Biopolymers 20, 1823-1832 (1981) 6 Nakayama H, Taki M, Striessnig J, Glossmann H, Catterall 13 Kuroda H, Chen Y-N, Kimura T and Sakakibara S: Powerful WA and Kanaoka Y: Identification of 1,4-dihydropyridine bind- solvent systems useful for synthesis of sparingly-soluble pep- ing regions within the a, subunit of the skeletal muscle Ca 21 tides in solution. Int J Peptide Protein Res 40, 294-299 (1992) channels by photoaffinity labeling with diazipine. Proc Natl 14 Watanabe TX, Itahara Y, Nakajima K, Kumagaye S, Kimura T Acad Sci USA 88, 9203 - 9207 (1991) and Sakakibara S: The biological activity of endothelin-1 ana- 7 Regulla S, Schneider T, Nastainczyk W, Meyer HE and logues in three different assay systems. J Cardiovasc Pharmacol Hofmann F: Identification of the site of interaction of the 17, Supp 7, S5 - S9 (1991) dihydropyridine channel blockers nitrendipine and azidopine 15 Kosterlitz HW, Lydon RJ and Watt AJ: The effects of adrena- with the calcium-channel a, subunit. EMBO J 10, 45-49 (1991) line, noradrenaline and isoprenaline on inhibitory a- and (3- 8 Strydom DJ: Snake venom toxins: The amino-acid sequence of adrenoceptors in the longitudinal muscle of the guinea-pig a short- homologue from Dendroaspis polylepis poly- ileum. Br J Pharmacol 39, 398-413 (1970) lepis (black mamba) venom. Eur J Biochem 76, 99-106 (1977) 16 Goto T, Satoh K and Taira N: Bay K 8644, a dihydropyridine 9 Strydom DJ: Snake venom toxins: Purification and properties calcium agonist, auguments vasoconstrictor responses to endog- of low-molecular-weight polypeptides of Dendroaspis polylepis enous and exogenous noradrenaline in the peripheral vascu- polylepis (black mamba) venom. Eur J Biochem 69, 169-176 lature of the dog. Br J Pharmacol 85, 913-916 (1985) (1976) 17 Yasuda 0, Morimoto S, Chen Y, Jiang B, Kimura T, 10 Watanabe TX, Itahara Y, Kuroda H, Kimura T and Sakakibara Sakakibara S, Koh E, Fukuo K, Kitano S and Ogihara T: Cal- S: Biological activities of calciseptine, a specific L-type calcium ciseptine binding to a 1,4-dihydropyridine recognition site of channel blocker. Jpn J Pharmacol 61, Supp I, 89P (1993) the L-type calcium channel of rat synaptosomal membranes. 11 Kuroda H, Chen Y-N, Watanabe TX, Kimura T and Biochem Biophys Res Commun 194, 587-594 (1993) Sakakibara S: Solution synthesis of calciseptine, an L-type 18 Endo T and Tamiya N: Current view on the structure-function specific calcium channel blocker. Peptide Res 5, 265-268 (1992) relationship of postsynaptic from snake venoms. 12 Kimura T, Takai M, Masui Y, Morikawa T and Sakakibara S: Pharmacol Ther 34, 403-451 (1987) Strategy for the synthesis of human parathyroid hormone