European Journal of Pharmaco!ogy, 216 (1992) 235-242
235
© 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00
EJP 52478
Relative agonist potencies of C2-substituted analogues of adenosine: evidence for adenosine A2B receptors in the guinea pig aorta P a u l i n e L. M a r t i n Department of Pharmacology, Whitby Research, bw., 2801 Reserve Street, Richmond, VA 23220, USA Received 17 January 1992, accepted 17 March 1992
Nine C2-substituted adenosine analogues that are potent and selective for the A2-adenosine receptor were tested for their ability to induce relaxations of the guinea pig aorta. Compounds tested were 2-phenylethoxyadenosine (PEA), 2-phenylethoxy-5'N-ethylcarboxamidoadenosine (PENECA), 2-cyclohexylethoxyadenosine (CEA), 2-fluorophenylethoxyadenosine (FPEA), 2methoxyphenylethoxyadenosine (MPEA), 2-naphthylethoxyadenosine (NEA), 2-phenylaminoadenosine (CV-1808), 2-phenylethylaminoadenosine (PEAA) and 2-carboxyethylphenethylamino-5'-N-ethylcarboxamidoadenosine (CGS21680). The responses to these agents were compared to those of three standard adenosine receptor agonists, 5'-N-ethylcarboxamidoadenosine (NECA), N6-cyclohexyladenosine (CHA) and R-N6-phenylisopropyladenosine (R-PIA). The C2-ethoxyadenosine analogues were 30- to 140-fold less potent than NECA and the C2-amino-substituted analogues were 250 to 1000-fold less potent than NECA at inducing relaxations of the guinea pig aorta. All of the analogues were also less potent than the Al-selective agonist R-PIA. However, only responses to NECA were competitively antagonized by the non-selective adenosine receptor antagonist 8-phenyltheophylline (8-PT), pK~ = 6.83 _+0.05. The results suggest that the C2-substituted analogues produce relaxations of the guinea pig aorta through a combination of actions at AE-adenosine receptors and at xanthine resistant sites. The lack of potency of these analogues at activating the xanthine sensitive A2-receptors in the guinea pig aorta suggests that these adenosine receptors may be of the A2b-subtype. Adenosine receptors; Adenosine analogues; Aorta (guinea-pig); 8-Phenyltheophylline
I. Introduction
Adenosine exerts its physiological actions through activation of two distinct subtypes of cell surface receptors designated A 1 and A 2 (Van Calker et al., 1979). In the cardiovascular system adenosine A2-receptors mediate relaxation of vascular smooth muscle and A l-receptors mediate cardiac depression (Olsson and Pearson, 1990). A number of compounds have been described recently that are capable of activating selectively the adenosine Az-receptor (Ueeda et al., 1991a,b; Hutchison et al., 1989; Abiru et al., 1991). A series of 2-(ar)alkoxyadenosines were shown to be very potent at inducing coronary vasodilation in the guinea pig Langendorff heart but to have very low potency at inducing negative dromotropic responses (Ueeda et al., 1991a,b). A similar group of compounds in which the bulky substituents are linked to the C2-position of the adeno-
Correspondence to: P.L. Martin, Department of Pharmacology, Whitby Research, Inc., 2801 Reserve Street, Richmond, VA 23220, U.S.A. 1.804.254 4400, fax 1.804.254 4035.
sine molecule via an amino linkage bind with high affinity to rat brain A2-receptors, and in the rat isolated working heart were shown to be potent agonists at inducing coronary vasodilation but to have low potency for inducing negative chronotropic responses (Hutchison et al., 1990). The most potent and selective of these compounds 2[4-(2-carboxyethyl)phenylamino]5'-N-ethylcarboxamidoadenosine (CGS21680) was additionally shown to potently decrease blood pressure in the spontaneously hypertensive rat but to have very little effect on heart rate (Hutchison et al., 1989). However, although CGS21680 was shown to be approximately 50-fold more potent than 2-phenylaminoadenosine (CV-1808) at inducing coronary vasodilation (Oei et al., 1988) it was not more potent at reducing blood pressure. It is therefore possible that the A 2adenosine receptors in the coronary vasculature may not be identical to the A2-adenosine receptors in other vascular smooth muscle. Alternatively, the differences in the relative potencies of these two agonists in vivo and in vitro may be the consequence of different degrees of metabolism or distribution of the two drugs in the whole animal. In order to test for possible receptor heterogeneity, a n u m b e r of A2-substituted
236
adenosine analogues that have been shown to be A zselective and potent coronary vasodilators were tested for their ability to induce A2-receptor-mediated relaxations of the guinea pig aorta, a tissue where metabolic influences are minimal. 2. Materials and methods
2.1. Guinea pig thoracic aorta
Where a is the asymptote, [As0] is the concentration of [A] for half-maximal response, and n is the midpoint slope parameter. In order to test for parallelism between concentration-effect curves, a one-way analysis of variance was performed on n values and a values between treatment groups. If no significant differences were detected, the computed [As0] values were fitted by an iterative procedure to the following form of the Schild equation: log[As0 ] = 1 + log[A~50 ] + n log[B] - log K ~
Thoracic aortas were removed from male guinea pigs weighing approximately 350 g that had been anaesthetized with halothane and killed by cervical dislocation. The vessels were cleared of any adhering tissue and cut into ring segments approximately 4 mm long. Tissues were placed in 30 ml organ baths containing Krebs solution (composition: (mM) NaC1 117.1, KC1 5.9, N a H 2 P O 4 1.2, N a H C O 3 24.7, MgSO 4 1.2, CaC12 2.5 and glucose 11), gassed with 95% 0 2 and 5% CO 2 and maintained at a temperature of 35°C. In all experiments, except those specifically designed to study the endothelium, the endothelium was removed by rubbing the intimal surface of the vessels. Aortic rings were placed under an initial resting tension of 2 g and responses to agonists were measured isometrically. An equilibration period of 1 h was allowed with exchange of bath fluid contents every 15 min. At the end of this period, phenylephrine (3 X 10 - 6 M) was added to each of the organ baths. Once a stable contracture had been obtained to this a-adrenoceptor agonist (approximately 1 h), antagonist or vehicle was added to the tissues and allowed to equilibrate for a further 1 h. At the end of the incubation period, concentration-effect curves to the adenosine receptor agonists were constructed cumulatively. Responses to adenosine analogues were expressed as the percentage relaxation of the initial phenylephrine contraction. In experiments designed to study the influence of the endothelium, tissues were set up as described above but with care being taken to maintain t~te integrity of the endothelial layer. Once a stable contraction to phenylephrine had been obtained, tissues were challenged with carbachol (10 -5 M). Tissues that relaxed in response to this acetylcholine receptor agonist were assumed to contain a functional endothelial cell layer. The phenylephrine and carbachol were subsequently washed out of the baths and experiments conducted in an identical fashion to that described above.
2.2. Data analysis Individual concentration-effect curve data were fitted by means of a least-square iterative computer program to a logistic function of the form: a[A] n E
[As0]n+ [A]n
Where [ACs0] is the midpoint location of the control concentration-effect (E/[A]) curve and [As0] is the midpoint location of the curve in the presence of antagonist [B]. K B is the antagonist-dissociation constant, and n is the order of [B] corresponding to the Schild plot slope parameter. If n was found to be not significantly different from unity, then the data were re-analyzed with n constrained to one, and a pK B value ( - log dissociation constant) quoted. Data are expressed as the means _+ S.E.M. Significant differences (P < 0.05) were calculated by Student's t-test for comparison of paired data. A N O V A followed by Bonferroni multiple comparisons procedure was used to analyze differences between mean values in an experiment with several treatment groups.
2.3. Drugs and solutions N6-Cyclohexyladenosine (CHA), 5'-N-ethylcarboxamidoadenosine (NECA), N6-R-phenylisopropyladen osine (R-PIA), 2-phenylaminoadenosine (CV-1808), and 2-[4-(2-carboxyethyl)phenylamino]-5'-N-ethylcarboxamidoadenosine (CGS21680) were obtained from Research Biochemicals Inc., Natick, MA, USA. 8Phenyltheophylline (8-PT) and phenylephrine HCI were obtained from Sigma Chemical Co., St. Louis, MO, USA. The C2-substituted adenosine analogues 2-(2-phenylethoxy)adenosine (PEA), 2-(2-phenylethoxy)-5'-Nethylcarboxamidoadenosine (PENECA), 2-(2-cyclohexylethoxy)adenosine (CEA), 2-[2(4-fluorophenyl)ethoxy]adenosine (FPEA), 2-[2-(4-methoxyphenyl)ethoxy]adenosine (MPEA), 2-[2(2-naphthyl)ethoxy]a d e n o s i n e ( N E A ) , and 2 - ( 2 - p h e n y l e t h y l a m i n o ) adenosine (PEAA) were synthesized by Drs. R.A. Olsson and R. Thompson at the Department of Internal Medicine University of South Florida, Tampa, FL 33612. 8-PT was dissolved to give a concentration of 10 -2 M in 100% DMSO. All other drugs were dissolved to give a concentration of 10 -2 M in 10% DMSO. Serial dilutions were made in distilled water. Preliminary experiments demonstrated that the maximum concentrations of DMSO added to the organ baths during the course of these experiments did not have any direct effects on the tissues (results not shown).
237 TABLE 1 Effect of 8-PT (3 × 10 -6 M) on the responses of guinea pig aorta to adenosine analogs.
100
Dose ratios (DR) were calculated by dividing the [As0] value in the presence of 8-PT by the [As0] value in the absence of 8-PT.
Z SCHILD PtOT
y.
50 1,1
o
NECA R-PIA CHA PEA CEA NEA MPEA FPEA PEAA CGS21680
[8-PT] : lo~d H i
i
-8
-7
i
[
I
-6 -5 -4 -3 [NECA] : log M
-2
-1
Fig. 1. Antagonism by 8-PT of the responses of the guinea pig aorta to NECA. E / [ A ] curves were constructed in the absence (e), and in the presence of 3 x 1 0 - 7 M (©); 10 -6 M ( I ) ; 3)
235
© 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00
EJP 52478
Relative agonist potencies of C2-substituted analogues of adenosine: evidence for adenosine A2B receptors in the guinea pig aorta P a u l i n e L. M a r t i n Department of Pharmacology, Whitby Research, bw., 2801 Reserve Street, Richmond, VA 23220, USA Received 17 January 1992, accepted 17 March 1992
Nine C2-substituted adenosine analogues that are potent and selective for the A2-adenosine receptor were tested for their ability to induce relaxations of the guinea pig aorta. Compounds tested were 2-phenylethoxyadenosine (PEA), 2-phenylethoxy-5'N-ethylcarboxamidoadenosine (PENECA), 2-cyclohexylethoxyadenosine (CEA), 2-fluorophenylethoxyadenosine (FPEA), 2methoxyphenylethoxyadenosine (MPEA), 2-naphthylethoxyadenosine (NEA), 2-phenylaminoadenosine (CV-1808), 2-phenylethylaminoadenosine (PEAA) and 2-carboxyethylphenethylamino-5'-N-ethylcarboxamidoadenosine (CGS21680). The responses to these agents were compared to those of three standard adenosine receptor agonists, 5'-N-ethylcarboxamidoadenosine (NECA), N6-cyclohexyladenosine (CHA) and R-N6-phenylisopropyladenosine (R-PIA). The C2-ethoxyadenosine analogues were 30- to 140-fold less potent than NECA and the C2-amino-substituted analogues were 250 to 1000-fold less potent than NECA at inducing relaxations of the guinea pig aorta. All of the analogues were also less potent than the Al-selective agonist R-PIA. However, only responses to NECA were competitively antagonized by the non-selective adenosine receptor antagonist 8-phenyltheophylline (8-PT), pK~ = 6.83 _+0.05. The results suggest that the C2-substituted analogues produce relaxations of the guinea pig aorta through a combination of actions at AE-adenosine receptors and at xanthine resistant sites. The lack of potency of these analogues at activating the xanthine sensitive A2-receptors in the guinea pig aorta suggests that these adenosine receptors may be of the A2b-subtype. Adenosine receptors; Adenosine analogues; Aorta (guinea-pig); 8-Phenyltheophylline
I. Introduction
Adenosine exerts its physiological actions through activation of two distinct subtypes of cell surface receptors designated A 1 and A 2 (Van Calker et al., 1979). In the cardiovascular system adenosine A2-receptors mediate relaxation of vascular smooth muscle and A l-receptors mediate cardiac depression (Olsson and Pearson, 1990). A number of compounds have been described recently that are capable of activating selectively the adenosine Az-receptor (Ueeda et al., 1991a,b; Hutchison et al., 1989; Abiru et al., 1991). A series of 2-(ar)alkoxyadenosines were shown to be very potent at inducing coronary vasodilation in the guinea pig Langendorff heart but to have very low potency at inducing negative dromotropic responses (Ueeda et al., 1991a,b). A similar group of compounds in which the bulky substituents are linked to the C2-position of the adeno-
Correspondence to: P.L. Martin, Department of Pharmacology, Whitby Research, Inc., 2801 Reserve Street, Richmond, VA 23220, U.S.A. 1.804.254 4400, fax 1.804.254 4035.
sine molecule via an amino linkage bind with high affinity to rat brain A2-receptors, and in the rat isolated working heart were shown to be potent agonists at inducing coronary vasodilation but to have low potency for inducing negative chronotropic responses (Hutchison et al., 1990). The most potent and selective of these compounds 2[4-(2-carboxyethyl)phenylamino]5'-N-ethylcarboxamidoadenosine (CGS21680) was additionally shown to potently decrease blood pressure in the spontaneously hypertensive rat but to have very little effect on heart rate (Hutchison et al., 1989). However, although CGS21680 was shown to be approximately 50-fold more potent than 2-phenylaminoadenosine (CV-1808) at inducing coronary vasodilation (Oei et al., 1988) it was not more potent at reducing blood pressure. It is therefore possible that the A 2adenosine receptors in the coronary vasculature may not be identical to the A2-adenosine receptors in other vascular smooth muscle. Alternatively, the differences in the relative potencies of these two agonists in vivo and in vitro may be the consequence of different degrees of metabolism or distribution of the two drugs in the whole animal. In order to test for possible receptor heterogeneity, a n u m b e r of A2-substituted
236
adenosine analogues that have been shown to be A zselective and potent coronary vasodilators were tested for their ability to induce A2-receptor-mediated relaxations of the guinea pig aorta, a tissue where metabolic influences are minimal. 2. Materials and methods
2.1. Guinea pig thoracic aorta
Where a is the asymptote, [As0] is the concentration of [A] for half-maximal response, and n is the midpoint slope parameter. In order to test for parallelism between concentration-effect curves, a one-way analysis of variance was performed on n values and a values between treatment groups. If no significant differences were detected, the computed [As0] values were fitted by an iterative procedure to the following form of the Schild equation: log[As0 ] = 1 + log[A~50 ] + n log[B] - log K ~
Thoracic aortas were removed from male guinea pigs weighing approximately 350 g that had been anaesthetized with halothane and killed by cervical dislocation. The vessels were cleared of any adhering tissue and cut into ring segments approximately 4 mm long. Tissues were placed in 30 ml organ baths containing Krebs solution (composition: (mM) NaC1 117.1, KC1 5.9, N a H 2 P O 4 1.2, N a H C O 3 24.7, MgSO 4 1.2, CaC12 2.5 and glucose 11), gassed with 95% 0 2 and 5% CO 2 and maintained at a temperature of 35°C. In all experiments, except those specifically designed to study the endothelium, the endothelium was removed by rubbing the intimal surface of the vessels. Aortic rings were placed under an initial resting tension of 2 g and responses to agonists were measured isometrically. An equilibration period of 1 h was allowed with exchange of bath fluid contents every 15 min. At the end of this period, phenylephrine (3 X 10 - 6 M) was added to each of the organ baths. Once a stable contracture had been obtained to this a-adrenoceptor agonist (approximately 1 h), antagonist or vehicle was added to the tissues and allowed to equilibrate for a further 1 h. At the end of the incubation period, concentration-effect curves to the adenosine receptor agonists were constructed cumulatively. Responses to adenosine analogues were expressed as the percentage relaxation of the initial phenylephrine contraction. In experiments designed to study the influence of the endothelium, tissues were set up as described above but with care being taken to maintain t~te integrity of the endothelial layer. Once a stable contraction to phenylephrine had been obtained, tissues were challenged with carbachol (10 -5 M). Tissues that relaxed in response to this acetylcholine receptor agonist were assumed to contain a functional endothelial cell layer. The phenylephrine and carbachol were subsequently washed out of the baths and experiments conducted in an identical fashion to that described above.
2.2. Data analysis Individual concentration-effect curve data were fitted by means of a least-square iterative computer program to a logistic function of the form: a[A] n E
[As0]n+ [A]n
Where [ACs0] is the midpoint location of the control concentration-effect (E/[A]) curve and [As0] is the midpoint location of the curve in the presence of antagonist [B]. K B is the antagonist-dissociation constant, and n is the order of [B] corresponding to the Schild plot slope parameter. If n was found to be not significantly different from unity, then the data were re-analyzed with n constrained to one, and a pK B value ( - log dissociation constant) quoted. Data are expressed as the means _+ S.E.M. Significant differences (P < 0.05) were calculated by Student's t-test for comparison of paired data. A N O V A followed by Bonferroni multiple comparisons procedure was used to analyze differences between mean values in an experiment with several treatment groups.
2.3. Drugs and solutions N6-Cyclohexyladenosine (CHA), 5'-N-ethylcarboxamidoadenosine (NECA), N6-R-phenylisopropyladen osine (R-PIA), 2-phenylaminoadenosine (CV-1808), and 2-[4-(2-carboxyethyl)phenylamino]-5'-N-ethylcarboxamidoadenosine (CGS21680) were obtained from Research Biochemicals Inc., Natick, MA, USA. 8Phenyltheophylline (8-PT) and phenylephrine HCI were obtained from Sigma Chemical Co., St. Louis, MO, USA. The C2-substituted adenosine analogues 2-(2-phenylethoxy)adenosine (PEA), 2-(2-phenylethoxy)-5'-Nethylcarboxamidoadenosine (PENECA), 2-(2-cyclohexylethoxy)adenosine (CEA), 2-[2(4-fluorophenyl)ethoxy]adenosine (FPEA), 2-[2-(4-methoxyphenyl)ethoxy]adenosine (MPEA), 2-[2(2-naphthyl)ethoxy]a d e n o s i n e ( N E A ) , and 2 - ( 2 - p h e n y l e t h y l a m i n o ) adenosine (PEAA) were synthesized by Drs. R.A. Olsson and R. Thompson at the Department of Internal Medicine University of South Florida, Tampa, FL 33612. 8-PT was dissolved to give a concentration of 10 -2 M in 100% DMSO. All other drugs were dissolved to give a concentration of 10 -2 M in 10% DMSO. Serial dilutions were made in distilled water. Preliminary experiments demonstrated that the maximum concentrations of DMSO added to the organ baths during the course of these experiments did not have any direct effects on the tissues (results not shown).
237 TABLE 1 Effect of 8-PT (3 × 10 -6 M) on the responses of guinea pig aorta to adenosine analogs.
100
Dose ratios (DR) were calculated by dividing the [As0] value in the presence of 8-PT by the [As0] value in the absence of 8-PT.
Z SCHILD PtOT
y.
50 1,1
o
NECA R-PIA CHA PEA CEA NEA MPEA FPEA PEAA CGS21680
[8-PT] : lo~d H i
i
-8
-7
i
[
I
-6 -5 -4 -3 [NECA] : log M
-2
-1
Fig. 1. Antagonism by 8-PT of the responses of the guinea pig aorta to NECA. E / [ A ] curves were constructed in the absence (e), and in the presence of 3 x 1 0 - 7 M (©); 10 -6 M ( I ) ; 3)
