European Journal of Pharmacoh)gy, 215 (1992) 253-257 ~ 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00
253
EJP 52444
Effects of imidazoline-related compounds on the mechanical response to nicorandil in the rat portal vein Keiichi Okumura, Kenji Ichihara and Mitsuaki Nagasaka Research Laboratories, Maruko Pharmaceutical Co., Ltd., Kasugai, Aichi 486, Japan Received 11 February. 1992, accepted 25 February 1992
The purpose of this study was to investigate the interactions of compounds structurally related to imidazoline at K + channels located in the rat portal vein. Nicorandil, a K + channel activator, dose dependently inhibited spontaneous contractions of the isolated rat portal vein. Glibenclamide (0.1-1 /.tM), an ATP-sensitive K + channel blocker, competitively antagonized the response 1o nicorandil, whereas methylene blue (10 #M), a guanylate cyclase inhibitor, did not. Phentolamine, antazoline, tolazoline, and midaglizole also shifted the dose-response curve for nicorandil to the right in the dose range of 1-100/.tM. The rank order of potency was gfibenclamide >> phentolamine = antazolinc = midaglizole > tolazoline. In contrast, clonidinc, idazoxan, imidazole, 1-benzylimidazole, and yohimbine were ineffective. In addition, cromakalim (1-100 nM), a selective K + channel activator, also inhibited spontaneous contractions of the rat portal vein, and this effect was antagonized by phentolamine in a similar way to that found with nicorandil. These results suggest that some 2-substituted imidazolines, including phentolamine, possibly act as K + channel blockers, like glibenclamide, in vascular smooth muscle.
Nicorandil; Imidazoline; Portal vein (rat); K + channels
1. Introduction Phentolamine, a non-selective a - a d r e n o c e p t o r antagonist, is known to potentiate the insulin response to glucose (Ahr6n and Lundquist, 1985; A h r d n et al., 1984; O d a et al., 1986). Schulz and Hasselblatt (1989) argued that the insulin-releasing action of phentolamine is related to its imidazoline structure, but not to its a - a d r e n o c e p t o r - b l o c k i n g property. This n o n - a d r e n ergic action of p h e n t o l a m i n e was evidenced by a subsequent study showing that p h e n t o l a m i n e directly inhibits ATP-sensitive K + currents in pancreatic /3-cells (Plant and Henquin, 1990). C r o m a k a l i m and nicorandil activate K + channels located in vascular smooth muscle, resulting in vasodilation (Cook, 1988; H a m i l t o n and Weston, 1989). Some imidazolines, including phentolamine, have also been r e p o r t e d to antagonize the vascular action of cromakalim ( M c P h e r s o n and Angus, 1989). These findings suggest that K + channels are important in the physiological action of imidazolines. Moreover, since imidazo-
Correspondence to: K. Ichihara, Research Laboratories, Maruko Pharmaceutical Co., Ltd., 1212 Teramae, Gejo-cho, Kasugai, Aichi 486, Japan. Tel. 81.568.81 4467, fax 81.568.84 8831.
lines such as clonidine and idazoxan bind to nonadrenergic imidazole binding sites (Bousquet and Schwartz, 1983; B o u s q u e t et al., 1984), the K + channels may be one of these binding sites. To obtain further information about the relation between the imidazoline structure and the K + channels in vascular smooth muscle, the present study was designed to investigate w h e t h e r several imidazoline derivatives, including p h e n t o l a m i n e and idazoxan, antagonize the effect of nicorandil on the isolated rat portal vein (fig. 1).
2. Materials and methods
2.1. Experimental procedure Male Wistar rats (SLC, Japan), weighing 200-250 g, were anesthetized with ether and the portal vein was carefully dissected. Segments of about 9 m m length were set up longitudinally in organ baths (20 ml) containing K r e b s - H e n s e l e i t solution kept at 37°C and aerated with a gas mixture of 95% 0 2 and 5% C O : . The composition in m M of the K r e b s - H e n s e l e i t solution was as follows: NaCI 119; KCI 4.7; CaCl 2 2.5; K H 2 P O 4 1.2; M g S O 4 1.2; N a H C O ~ 25; glucose 11. The
254
Inactive
Active
-- CHz
HN~'~
\
Phentolamine
Midaglizole
Antazoline
Tolazoline
/
Imidazole
1-Benzylimidazole
Idazoxan
Clonidine
Fig. 1. The structures of the imidazolines tested for their ability to antagonize the response elicited by nicorandil.
preparation was attached to an isometric force-displacement transducer (Nihon Kohden, Japan) to measure the contractile force of spontaneous contractions. Preparations were equilibrated for 30 rain under a resting tension of 0.3 g. Cumulative dose-response curves for nicorandil were determined in the presence and absence of test drugs. Each drug was applied to the tissue for 20 min before obtaining dose-response curves for nicorandil. In another experiment, the effect of nicorandil was examined against responses evoked by 20 and 80 mM KC1 (no equimolar replacement of NaC1) in the rat portal vein. Several different concentrations of nicorandiI were applied 20 min before the K+-induced contraction was evoked.
2.2. Chemicals The following drugs were used: antazoline hydrochloride (Sigma, USA); 1-benzylimidazole (Aldrich, FRG); clonidine hydrochloride (Sigma, USA); cromakalim; glibenclamide (Yamanouchi Pharmaceutical Co., Japan); idazoxan hydrochloride (Sigma, USA); imidazole (Aldrich, FRG); methylene blue (Wako, Japan); midaglizole; nicorandil; phentolamine methanesulfonate (Ciba-Geigy, FRG); tolazoline hydrochloride (Sigma, USA); yohimbine hydrochloride (Sigma, USA). Nicorandil, cromakalim and midaglizole were synthesized in our chemical laboratory. Glibenclamide and cromakalim were initially dissolved in a minimum volume of dimethylsulfoxide and then diluted with distilled water. Other drugs were dissolved in distilled water.
2.3. Data analysis The antagonist potencies were quantified by calculating pA 2 values according to the method of Arunlakshana and Schild (1959). The results are presented as means _+ S.E.
3. Results
3.1. Effect of nicorandil on the rat portal t.,ein Nicorandil inhibited spontaneous contractions of the rat portal vein (table 1). Nicorandil reduced both the amplitude and frequency of spontaneous contractions in a dose-related way, whereas the basal tone remained unaltered. At intermediate concentrations, nicorandil inhibited the amplitude more than it inhibited the frequency. In the following experiment, we used the
TABLE 1 Effect of nicorandil on the amplitude and the frequency of spontaneous contractions in the isolated rat portal vein. Mean basal values of amplitude and frequency were 0.35 _+0.02 g and 4.1 _+0.4 cycles/min, respectively. Each value represents the mean _+ S.E. of four animals. Nicorandil
Inhibition (%)
(#M)
Amplitude
Frequency
0.3 1 3 10
7.8 _+ 1.0 26.2 _+ 1.7 54.7_+2.7 100
1.4 +_0.8 9.4 _+3.4 28.9_+8.5 100
255
100
GIIbenclamlde -0- : Control ~) # / ~ y -.-: o.t..
i / , y
/J/
-*-:0.3.M
/
100
.
. 6
.
.
.
,o.o
.
5
~
_
/j/l
Nlcorandll (-IogM)
Phentolamine -0-- :Control p
100
--,,- : 1 . .
),
+:"°'171t .
Methyleneblue -C~ : Control ~,~/ID
:3uM
~1) / j . ~ B
I/~,,.
"~ i- 50 _c
0 4
Nicorandll (-IogM)
Fig. 2. Effects of glibenclamide (left penel) and methylene blue (right panel) on the response to nicorandil of spontaneous contractions of the rat portal vein. Glibenclamide and methylene blue were applied 20 rain before the nicorandil d o s e - r e s p o n s e curve was made. Each value represents the mean +_S.E. of four animals.
amplitude of contractions as an indicator of spontaneous activity of the portal vein. Figure 2 shows the effects of glibenclamide and methylene blue on the response of spontaneous contractions of the rat portal vein to nicorandil. Nicorandil dose dependently inhibited the amplitude of spontaneous contractions in the dose range of 0.3-10 /~M. Glibenclamide (0.1-1 /~M) shifted the d o s e - r e s p o n s e curve for nicorandil to the right in a parallel manner. Methylene blue (10 ~ M ) did not affect the response to nicorandil. Figure 3 shows the effect of nicorandil on single contractions elicited by K+-induced depolarization in the rat portal vein. Nicorandil (1-100 # M ) inhibited the force development of the contraction induced by a lower K + concentration (20 mM) to a greater extent than when the contraction was elicited by a higher concentration (80 raM). 3.2. Effects of imidazolines on the response to nicorandil
A number of the imidazoline-related compounds were tested for their ability to antagonize the inhibitory
oo] K¢,
/o
-0- :20raM I ~
_=.5-'-7 oi NIcorandll (-IogM) Fig. 3. Inhibitory effects o f nicorandil on KCI (20 or 80 mM)-induced contractions in the rat portal vein. Nicorandil was applied 20 rain before the start of the KCI-induced contractions. Maximum responses of control tissue to 20 and 80 mm KCI were 0.24_+ 0.05 and 0.59 + 0.08 g, respectively. Each value represents the mean _+S.E. of four animals.
o
~
~
4
Nicorandil (-IogM) Fig. 4. Effect of phentolamine on the response to nicorandil of spontaneous contractions of the rat portal vein. Phentolamine was applied 20 min before the nicorandil d o s e - r e s p o n s e curve was made. Each value represents the mean + S.E. of four animals.
effect of nicorandil on spontaneous contractions of the isolated rat portal vein. The chemical structures of the imidazolines used in this study are illustrated in fig. 1. Phentolamine shifted the d o s e - r e s p o n s e curve for nicorandil to the right in a parallel manner (fig. 4). Phentolamine also dose dependently shifted the inhibition curve of cromakalim on spontaneous contractions to the right, with a pA 2 value of 6.01 _+ 0.15 (fig. 5). Antazoline, tolazoline, and midaglizole also shifted the d o s e - r e s p o n s e curve for nicorandil to the right, whereas clonidine, idazoxan, imidazole, and 1-benzylimidazole failed to show enough effect to evaluate their antagonist potency even at high concentrations. Idazoxan was also ineffective on the response to cromakalim (data not shown). Table 2 summarizes the pA 2 values and the slopes of the Schild regression lines for these compounds on nicorandil-induced responses. The pA 2 values of glibenclamide, phentolamine, antazoline, tolazoline, and midaglizole were 7.25, 5.83, 5.53, 4.45, and 5.30, respectively. None of the slopes were significantly different from unity. Yohimbine, an az-adrenoceptor antagonist, did not affect the d o s e - r e s p o n s e curve for nicorandil at the concentration of 10/~M.
Phentolamlne 100, -o-_~__e_ ::31JM: C ~ o 1n t r0o l /u M
0
i
8 7 6 5 Cromakalim (-IogM)
Fig. 5. Effect of phe nt ol a mi ne on the response to cromakalim of spontaneous contractions of the rat portal vein. Phentolamine was applied 20 min before the cromakalim d o s e - r e s p o n s e curve was made. Each value represents the mean + S.E. of four animals.
256 TABLE 2 Antagonistic effects on the nicorandil-induced inhibition of spontaneous contractions in the isolated rat portal vein. Antagonist potencies are given as pA z values calculated according to Arunlakshana and Schild (1959). Each value represents the mean + S.E. of the number of animals indicated under N. Compound
N
Glibenclamide Phentolamine Antazoline Tolazoline Midaglizole ldazoxan CIonidine lmidazole l-Benzylimidazole Yohimbine
4 4 5 5 5 4 4 4 4 4
pA 2
< < < <
253
EJP 52444
Effects of imidazoline-related compounds on the mechanical response to nicorandil in the rat portal vein Keiichi Okumura, Kenji Ichihara and Mitsuaki Nagasaka Research Laboratories, Maruko Pharmaceutical Co., Ltd., Kasugai, Aichi 486, Japan Received 11 February. 1992, accepted 25 February 1992
The purpose of this study was to investigate the interactions of compounds structurally related to imidazoline at K + channels located in the rat portal vein. Nicorandil, a K + channel activator, dose dependently inhibited spontaneous contractions of the isolated rat portal vein. Glibenclamide (0.1-1 /.tM), an ATP-sensitive K + channel blocker, competitively antagonized the response 1o nicorandil, whereas methylene blue (10 #M), a guanylate cyclase inhibitor, did not. Phentolamine, antazoline, tolazoline, and midaglizole also shifted the dose-response curve for nicorandil to the right in the dose range of 1-100/.tM. The rank order of potency was gfibenclamide >> phentolamine = antazolinc = midaglizole > tolazoline. In contrast, clonidinc, idazoxan, imidazole, 1-benzylimidazole, and yohimbine were ineffective. In addition, cromakalim (1-100 nM), a selective K + channel activator, also inhibited spontaneous contractions of the rat portal vein, and this effect was antagonized by phentolamine in a similar way to that found with nicorandil. These results suggest that some 2-substituted imidazolines, including phentolamine, possibly act as K + channel blockers, like glibenclamide, in vascular smooth muscle.
Nicorandil; Imidazoline; Portal vein (rat); K + channels
1. Introduction Phentolamine, a non-selective a - a d r e n o c e p t o r antagonist, is known to potentiate the insulin response to glucose (Ahr6n and Lundquist, 1985; A h r d n et al., 1984; O d a et al., 1986). Schulz and Hasselblatt (1989) argued that the insulin-releasing action of phentolamine is related to its imidazoline structure, but not to its a - a d r e n o c e p t o r - b l o c k i n g property. This n o n - a d r e n ergic action of p h e n t o l a m i n e was evidenced by a subsequent study showing that p h e n t o l a m i n e directly inhibits ATP-sensitive K + currents in pancreatic /3-cells (Plant and Henquin, 1990). C r o m a k a l i m and nicorandil activate K + channels located in vascular smooth muscle, resulting in vasodilation (Cook, 1988; H a m i l t o n and Weston, 1989). Some imidazolines, including phentolamine, have also been r e p o r t e d to antagonize the vascular action of cromakalim ( M c P h e r s o n and Angus, 1989). These findings suggest that K + channels are important in the physiological action of imidazolines. Moreover, since imidazo-
Correspondence to: K. Ichihara, Research Laboratories, Maruko Pharmaceutical Co., Ltd., 1212 Teramae, Gejo-cho, Kasugai, Aichi 486, Japan. Tel. 81.568.81 4467, fax 81.568.84 8831.
lines such as clonidine and idazoxan bind to nonadrenergic imidazole binding sites (Bousquet and Schwartz, 1983; B o u s q u e t et al., 1984), the K + channels may be one of these binding sites. To obtain further information about the relation between the imidazoline structure and the K + channels in vascular smooth muscle, the present study was designed to investigate w h e t h e r several imidazoline derivatives, including p h e n t o l a m i n e and idazoxan, antagonize the effect of nicorandil on the isolated rat portal vein (fig. 1).
2. Materials and methods
2.1. Experimental procedure Male Wistar rats (SLC, Japan), weighing 200-250 g, were anesthetized with ether and the portal vein was carefully dissected. Segments of about 9 m m length were set up longitudinally in organ baths (20 ml) containing K r e b s - H e n s e l e i t solution kept at 37°C and aerated with a gas mixture of 95% 0 2 and 5% C O : . The composition in m M of the K r e b s - H e n s e l e i t solution was as follows: NaCI 119; KCI 4.7; CaCl 2 2.5; K H 2 P O 4 1.2; M g S O 4 1.2; N a H C O ~ 25; glucose 11. The
254
Inactive
Active
-- CHz
HN~'~
\
Phentolamine
Midaglizole
Antazoline
Tolazoline
/
Imidazole
1-Benzylimidazole
Idazoxan
Clonidine
Fig. 1. The structures of the imidazolines tested for their ability to antagonize the response elicited by nicorandil.
preparation was attached to an isometric force-displacement transducer (Nihon Kohden, Japan) to measure the contractile force of spontaneous contractions. Preparations were equilibrated for 30 rain under a resting tension of 0.3 g. Cumulative dose-response curves for nicorandil were determined in the presence and absence of test drugs. Each drug was applied to the tissue for 20 min before obtaining dose-response curves for nicorandil. In another experiment, the effect of nicorandil was examined against responses evoked by 20 and 80 mM KC1 (no equimolar replacement of NaC1) in the rat portal vein. Several different concentrations of nicorandiI were applied 20 min before the K+-induced contraction was evoked.
2.2. Chemicals The following drugs were used: antazoline hydrochloride (Sigma, USA); 1-benzylimidazole (Aldrich, FRG); clonidine hydrochloride (Sigma, USA); cromakalim; glibenclamide (Yamanouchi Pharmaceutical Co., Japan); idazoxan hydrochloride (Sigma, USA); imidazole (Aldrich, FRG); methylene blue (Wako, Japan); midaglizole; nicorandil; phentolamine methanesulfonate (Ciba-Geigy, FRG); tolazoline hydrochloride (Sigma, USA); yohimbine hydrochloride (Sigma, USA). Nicorandil, cromakalim and midaglizole were synthesized in our chemical laboratory. Glibenclamide and cromakalim were initially dissolved in a minimum volume of dimethylsulfoxide and then diluted with distilled water. Other drugs were dissolved in distilled water.
2.3. Data analysis The antagonist potencies were quantified by calculating pA 2 values according to the method of Arunlakshana and Schild (1959). The results are presented as means _+ S.E.
3. Results
3.1. Effect of nicorandil on the rat portal t.,ein Nicorandil inhibited spontaneous contractions of the rat portal vein (table 1). Nicorandil reduced both the amplitude and frequency of spontaneous contractions in a dose-related way, whereas the basal tone remained unaltered. At intermediate concentrations, nicorandil inhibited the amplitude more than it inhibited the frequency. In the following experiment, we used the
TABLE 1 Effect of nicorandil on the amplitude and the frequency of spontaneous contractions in the isolated rat portal vein. Mean basal values of amplitude and frequency were 0.35 _+0.02 g and 4.1 _+0.4 cycles/min, respectively. Each value represents the mean _+ S.E. of four animals. Nicorandil
Inhibition (%)
(#M)
Amplitude
Frequency
0.3 1 3 10
7.8 _+ 1.0 26.2 _+ 1.7 54.7_+2.7 100
1.4 +_0.8 9.4 _+3.4 28.9_+8.5 100
255
100
GIIbenclamlde -0- : Control ~) # / ~ y -.-: o.t..
i / , y
/J/
-*-:0.3.M
/
100
.
. 6
.
.
.
,o.o
.
5
~
_
/j/l
Nlcorandll (-IogM)
Phentolamine -0-- :Control p
100
--,,- : 1 . .
),
+:"°'171t .
Methyleneblue -C~ : Control ~,~/ID
:3uM
~1) / j . ~ B
I/~,,.
"~ i- 50 _c
0 4
Nicorandll (-IogM)
Fig. 2. Effects of glibenclamide (left penel) and methylene blue (right panel) on the response to nicorandil of spontaneous contractions of the rat portal vein. Glibenclamide and methylene blue were applied 20 rain before the nicorandil d o s e - r e s p o n s e curve was made. Each value represents the mean +_S.E. of four animals.
amplitude of contractions as an indicator of spontaneous activity of the portal vein. Figure 2 shows the effects of glibenclamide and methylene blue on the response of spontaneous contractions of the rat portal vein to nicorandil. Nicorandil dose dependently inhibited the amplitude of spontaneous contractions in the dose range of 0.3-10 /~M. Glibenclamide (0.1-1 /~M) shifted the d o s e - r e s p o n s e curve for nicorandil to the right in a parallel manner. Methylene blue (10 ~ M ) did not affect the response to nicorandil. Figure 3 shows the effect of nicorandil on single contractions elicited by K+-induced depolarization in the rat portal vein. Nicorandil (1-100 # M ) inhibited the force development of the contraction induced by a lower K + concentration (20 mM) to a greater extent than when the contraction was elicited by a higher concentration (80 raM). 3.2. Effects of imidazolines on the response to nicorandil
A number of the imidazoline-related compounds were tested for their ability to antagonize the inhibitory
oo] K¢,
/o
-0- :20raM I ~
_=.5-'-7 oi NIcorandll (-IogM) Fig. 3. Inhibitory effects o f nicorandil on KCI (20 or 80 mM)-induced contractions in the rat portal vein. Nicorandil was applied 20 rain before the start of the KCI-induced contractions. Maximum responses of control tissue to 20 and 80 mm KCI were 0.24_+ 0.05 and 0.59 + 0.08 g, respectively. Each value represents the mean _+S.E. of four animals.
o
~
~
4
Nicorandil (-IogM) Fig. 4. Effect of phentolamine on the response to nicorandil of spontaneous contractions of the rat portal vein. Phentolamine was applied 20 min before the nicorandil d o s e - r e s p o n s e curve was made. Each value represents the mean + S.E. of four animals.
effect of nicorandil on spontaneous contractions of the isolated rat portal vein. The chemical structures of the imidazolines used in this study are illustrated in fig. 1. Phentolamine shifted the d o s e - r e s p o n s e curve for nicorandil to the right in a parallel manner (fig. 4). Phentolamine also dose dependently shifted the inhibition curve of cromakalim on spontaneous contractions to the right, with a pA 2 value of 6.01 _+ 0.15 (fig. 5). Antazoline, tolazoline, and midaglizole also shifted the d o s e - r e s p o n s e curve for nicorandil to the right, whereas clonidine, idazoxan, imidazole, and 1-benzylimidazole failed to show enough effect to evaluate their antagonist potency even at high concentrations. Idazoxan was also ineffective on the response to cromakalim (data not shown). Table 2 summarizes the pA 2 values and the slopes of the Schild regression lines for these compounds on nicorandil-induced responses. The pA 2 values of glibenclamide, phentolamine, antazoline, tolazoline, and midaglizole were 7.25, 5.83, 5.53, 4.45, and 5.30, respectively. None of the slopes were significantly different from unity. Yohimbine, an az-adrenoceptor antagonist, did not affect the d o s e - r e s p o n s e curve for nicorandil at the concentration of 10/~M.
Phentolamlne 100, -o-_~__e_ ::31JM: C ~ o 1n t r0o l /u M
0
i
8 7 6 5 Cromakalim (-IogM)
Fig. 5. Effect of phe nt ol a mi ne on the response to cromakalim of spontaneous contractions of the rat portal vein. Phentolamine was applied 20 min before the cromakalim d o s e - r e s p o n s e curve was made. Each value represents the mean + S.E. of four animals.
256 TABLE 2 Antagonistic effects on the nicorandil-induced inhibition of spontaneous contractions in the isolated rat portal vein. Antagonist potencies are given as pA z values calculated according to Arunlakshana and Schild (1959). Each value represents the mean + S.E. of the number of animals indicated under N. Compound
N
Glibenclamide Phentolamine Antazoline Tolazoline Midaglizole ldazoxan CIonidine lmidazole l-Benzylimidazole Yohimbine
4 4 5 5 5 4 4 4 4 4
pA 2
< < < <
