Kinetic characterization of adenosine A2 receptor-mediated relaxation in isolated rabbit aorta

European Journal of Pharmacology, 238 (1993) 65-74

65

© 1993 Elsevier Science Publishers B.V. All rights reserved 0014-2999/93/$06.00

EJP 53156

Kinetic characterization of adenosine A 2 receptor-mediated relaxation in isolated rabbit aorta H a r v e y L. W i e n e r a n d G e o r g e P. T h a l o d y 1 Department of Pharmaceutical Sciences, College of Pharmacy and Allied Health Professions, St. John's Uni~'ersity,Jamaica, NY 11439, USA

Received 31 December 1992, revised MS received 31 March 1993, accepted 13 April 1993

Previous studies in our laboratory (Wiener et al., 1991, Soc. Neurosci. Abstr. 17, 989) have addrcssed aspects of the functional antagonism between the responses mediated by activated adenosine A 2 receptors and a~-adrenoceptors in adventitia- and endothelium-denuded rabbit thoracic aortic rings by steady-state protocols which ignore the time course of response generation. In the present communication we describe aspects of the time-dependent kinetics of relaxation responses to adenosine A 2 receptor agonists in tissues pre-contracted with the al-adrenoceptor agonist phenylephrine. The results were analyzed by application of the model originally developed by Keitz et al. (1990, J. Pharmacol. Exp. Ther. 255, 650) to describe the relaxation response, to a/3-adrenoceptor agonist, as a first-order exponential decrease in tissue tension over time to estimate the apparent rate constant for relaxation (k~el) and the magnitude of relaxation at equilibrium. The magnitude of the relaxation responses to adenosine, N6-cyelohexyladenosine, N6-methyladenosine, 5'-N-ethylcarboxamidoadenosine, and R(-)-N6-(2-phenylisopropyl)adenosine were agonist concentration-dependent and saturable, as were the apparent rate constants for relaxation. In addition, the magnitude of the apparent rate constants for relaxation and the relaxation responses were inversely proportional to the fractional occupancy of the ax-adrenoceptor. The hypothesis put forth by Keitz et al. that the maximal value of the apparent rate constant for relaxation may serve as the kinetic definition of agonist efficacy was also tested and found to be invalid for the adenosine A 2 receptor. We propose that this pair of activated receptors and tissue preparation is a good model to study quantitative aspects of functional antagonism by kinetic paradigms. Kinetics; Antagonism, functional; Adenosine A 2 receptors; al-Adrenoceptor; Thoracic aorta (rabbit)

1. Introduction Time-dependent kinetic models provide useful information in studying responses of tissues to various pharmacological agents. Maayani and Osman and their colleagues (Cory et al., 1984, 1986; Clancy et ai., 1987; Keitz et al., 1990; Osman et al., 1990; Ben-Harari et al., 1991) have developed several kinetic models to study the responses of the adventitia- and endothelium-denuded isolated rabbit thoracic aorta to biogenic amines. Others have also utilized their models to study kinetic responses in other blood vessels (Onaran and B6kesoy, 1990) or tissues (Tagliente et al., 1992). The kinetic model utilized in the present study was

Correspondence to: H.L. Wiener, Department of Pharmaceutical Sciences, College of Pharmacy and Allied Health Professions, St. John's University, Grand Central and Utopia Parkways,Jamaica, NY 11439, USA. Tel. (718) 990-1678, fax (718) 969-0753. Present address: Central Nervous System Neuropharmacology, Bristol-Myers Squibb Co., Pharmaceutical Research Institute, 5 Research Parkway, Wallingford, CT 06492, USA.

originally described by Keitz et al. (1990) and Osman et al. (1990) to characterize the relaxation response of the denuded rabbit aorta to the /3-adrenoceptor agonist isoproterenol. In this tissue the response to isoproterenol is biphasic, consisting of a rapid relaxation followed by a slower regaining of tissue tension (Keitz et al., 1990). Since these two responses were temporally separated, Keitz et al. (1990) described each process individually; the relaxation response was adequately described as a first-order exponential decrease in tissue tension, while the regaining of tissue tension or desensitization was described as a first-order increase in tissue tension. Use of their kinetic model to analyze relaxation responses allows the investigator to estimate two parameters, the apparent rate constant for relaxation, k rcl, and the magnitude of the relaxation response. In analyzing the responses to isoproterenol, Keitz et al. (1990) noted that the value of kre ~ was isoproterenol concentration-dependent and saturable; furthermore, they suggested that the maximal value of krel, (krel)max, may serve as a kinetic definition of drug efficacy. A recent report from our laboratory (Wiener

66

ct al., 1992) extended the use of this rowel kinetic model to characterize aspects of the relaxation responses to two nitrovasodilators whose effects are mediated by the production of 3',5'-GMP (lgnarro and Kadowitz, 1985), namely nitroglycerin and sodium nitroprusside, and to an adenosine A 2 receptor agonist, 5'-N-ethylcarboxamidoadenosine (NECA), whose response is mediated by the production of 3',5'-AMP. Adenosine analogues relax blood vessels by interacting with the adenosine A 2 receptor which is linked to the stimulation of adenylyl cyclase through a cholera toxin-sensitive stimulatory guanine nucleotide binding protein (Sabouni et al., 1989, 1991). For recent reviews on adenosine receptors see Olah and Stiles (1992) and Stiles (1992). Although the effects of adenosine receptor agonists on various isolated blood vessels have been studied by traditional steady-state methods, which ignore the time-dependent aspects of response generation (Herlihy et al., 1976; Ghai and Mustafa, 1982; Collis and Brown, 1983; McCormack et al., 1989; Headrick and Berne, 1990; Sabouni et al., 1990a; Balwierczak et al., 1991; Cushing et al., 1991), a thorough kinetic characterization of adenosine receptor-mediated relaxation has not, to the best of our knowledge, been reported. The present communication describes the application of the kinetic model of Keitz et al. (1990) to study the relaxation responses to five adenosine analogues (adenosine, N6-cyclohexyladenosine, N6-methyladenosine, R(-)-N6-(2-phenylisopropyl)adenosine (R-PIA), and NECA; fig. I) in the adventitia- and endotheliumdenuded isolated rabbit aorta. The agonist concentration dependence and saturability of the values of the apparent rate constant for relaxation and the magnitude of relaxation are described, as is the inverse dependence of these values on the contractile stimulus in the denuded aorta. The hypothesis put forth by Keitz et al. (1990) that the maximal value of the apparent rate constant for relaxation may serve as the kinetic definition of drug efficacy was also tested.

2. Materials and methods

2.1. Reagents N6-Cyclohexyladenosine, N°-methyladenosine, and R-PIA were purchased from Research Biochemicals (Natick, MA, USA). All other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO, USA) and were of the highest grade available.

2.2. Animals and tissue preparation Adventitia- and endothelium-denuded rings were prepared from the isolated thoracic aortae of male

New Zealand white rabbits (1.5-2.0 kg; Buckshire Corp., Perkasic, PA, USA) as described previously (Wiener et al., 1992).

2.3. Experimental design All experiments began with the construction of a cumulative concentration-response curve to phcnylephrine to estimate its ECs0 and subsequently, the concentration of phenylephrine utilized was expressed in relation to its ECs0. Two different kinetic experiments were performed for each adenosine A 2 receptor agonist. First, a sequential concentration-response curve was constructed for each agonist in adventitiaand endothelium-denuded rings pre-contracted with 10 ECs0 of phenylephrine. Secondly, a maximum concentration of adenosine A 2 receptor agonist was used to relax aortic rings pre-contracted with different concentrations of phenylephrine (i.e., 1, 3, 10, 30, and 100 ECs0). The former protocol allowed for a comparison of relaxation rate constants (see below) as well as relaxation responses as a function of adenosine A 2 receptor fractional occupancy. The latter protocol provided the same values as a function of al-adrenoceptor fractional occupancy. During each experimental trial one ring from each rabbit was utilized as a paired-control to verify that the contractile response to phenylephrine was stable during the time required to assay the relaxation response to adenosine.

2.4. Collection of data The tension developed was measured using isometric force displacement transducers (Grass Instrument Co., Quincy, MA, USA, model FF-03c) connected to a multichannel polygraph. Analog data from the original polygraph tracings (recorded using a chart speed > 25 mm/min) were transferred to an IBM compatible computer using a digitizing tablet and the SigmaScan (Jandel Scientific, Corte Madera, CA, USA) software package. Digital values of muscle tension were analyzed using non-linear regression analysis using SigmaPlot (Jandel Scientific) and the equations described below.

2.5. Kinetic model and data analysis The kinetic model of Keitz et al. (1990), as recently described by us (Wiener et al., 1992), was applied in the present study to characterize the time-dependent kinetics of relaxation responses to adenosine A 2 receptor agonists. Briefly, relaxation responses described were as a first-order exponential decay that began at the observed peak level of the contractile response (C; fig. 2) and reached a limiting value of R at steady-state: T = C-

R(I -e

- krt~t )

(1)

67 where T was the observed tissue tension (g), C was the initial contraction (g), R was the adenosine A2 receptor agonist-induced relaxation (g), k re, was the observed rate constant for relaxation ( s - ' ) , and t was the elapsed time since the adenosine A a receptor agonist was added (s). C was measured from the experimental data, while the values of R and krd were estimated by fitting the experimental data to the model presented in equation 1 by non-linear regression analysis. During curve fitting, R was constrained to < C. In order to compare different aortic rings, the extent of relaxation was described as the fractional relaxation ( R / C ) response. Concentration-response curves were analyzed using non-linear regression analysis by fitting the observed response and drug concentration to the following three-parameter logistic equations modified from De Lean et al. (1978): (R/C),,,** ( R ) = 1+ (ECs0 / A)n

(2)

(k ~,)r.~ k,e~= 1 + (EC50/A) n

(3)

where R / C was the observed fractional relaxation response, (R/C)ma x was the maximal fractional relaxation response, k re, was the observed relaxation rate constant, (krel)max was the maximal relaxation rate constant, ECs0 was the concentration producing one-half maximal response, A was the agonist concentration, and n was the slope index.

2.6. Statistical evaluation All numerical values reported are the arithmetic mean ± S.E.M., except for EC50 values which are the geometric means (Gaddum, 1945; Fleming et al., 1972). Statistical evaluation was by one-way analysis of variance followed by Duncan's multiple-range test. The accepted level of significance was P < 0.05.

3. Results

3.1. Concentration dependence and saturability of adenosine A 2 receptor-mediated relaxation responses All adenosine analogues (fig. 1) tested did not alter basal tissue tone in naive adventitia- and endotheliumdenuded isolated rabbit thoracic aortic rings (data not shown); the failure of R - P I A or N6-cyclohexyladeno sine to elicit a contractile response suggested that this preparation did not contain adenosine A , receptors, in contrast to the isolated guinea pig aorta (Stoggall and Shaw, 1990). Therefore, in order to visualize the responses to adenosine analogues we employed func-

NH~

N

HO

HO

OH

OH

N6-Cyclohexyladenosine

Adenosine .CH3 HN

Nil2

C2HsNH~ HO

___. HO OH

Oft

N6-Methyladenosine

H(~

5'-N-Ethylcarboxamidoadenosine (NECA)

OH

R(-)-N6-(2-Phenylisopropyl)adenosine (R-PIA) Fig. 1. Adenosine analogues. tional antagonism of phenylephrine-pre-contracted rings using the protocol illustrated in fig. 2. The contractile response to phenylephrine was stable during the time required to assay the relaxation response to AD

C O O3 C 09

~E Time Fig. 2. Schematic depiction of the time-dependent responses to phenylephrine (PE) and adenosine (AID) in the adventitia- and endothelium-denuded isolated rabbit thoracic aorta. Phenylepbrine was added to the organ bath (where indicated by the dot) to elicit the steady-state contractile response, C. Addition of adenosine (where indicated by the dot) elicited a monotonic relaxation response, R. The relaxation response data were fitted to equation 1 (see Materials and methods) to estimate the values of R and the apparent rate constant for relaxation, kre 1.

68 an a d e n o s i n e a n a l o g u e (data not shown); it was also r e p r o d u c i b l e over the course of e x p e r i m e n t a l trials (up to seven assays over a 12 h period). T h e values of the ECs0 a n d agonist dissociation constant, K A ( e s t i m a t e d by the m e t h o d of Stollack a n d F u r c h g o t t (1983) following partial irreversible occlusion with the alkylating agent d i b e n a m i n e ) , for p h e n y l e p h r i n c d e t e r m i n e d by a s e p a r a t e series of e x p e r i m e n t s were 0.21 (pECs0 = 6.67 + 0.04) a n d 2.3 /zM (pK A = 5.64 + 0.12), respectively, for 12 rings from three rabbits. T h e s e ECso a n d K A values were similar to those r e p o r t e d by Christ et al. (1990) and, more recently, by us ( W i e n e r et al., in press) in the same tissue p r e p a r a t i o n . T h a t the K A / E C s 0 was approximately 11 suggested that this tissue p r e p a r a t i o n had s u b s t a n t i a l receptor reserve for phcnylephrine. A d e n o s i n e (figs. 3A and 4A), N E C A (fig. 5A), N 6m c t h y l a d e n o s i n e (fig. 6A), R - P I A (fig. 7A) a n d N °cyclohexyladenosine (fig. 8A) relaxed rabbit aortic rings p r e - c o n t r a c t e d with 10 times the ECs0 of phenyle p h r i n e in a c o n c e n t r a t i o n - d e p e n d e n t a n d s a t u r a b l e m a n n e r . T h e m o n o t o n i c relaxation responses to these a n a l o g u e s were stable for over 30 m i n a n d reproducible t h r o u g h o u t the course of the e x p e r i m e n t (data not shown). T h e fractional relaxation ( R / C ) response c o n c e n t r a t i o n - r e s p o n s e p a r a m e t e r s for each a n a l o g u e , in rings p r e - c o n t r a c t e d with 10 times the ECs0 of p h e n y l e p h r i n e , e s t i m a t e d using e q u a t i o n 2 ( ( R / C ) m a x, ECs0, a n d n, the slope index; table 1) were c o m p a r e d by one-way analysis of variance followed by D u n c a n ' s m u l t i p l e - r a n g e tcst. For (R/C)max: F(4,38) = 15.2, P < 0.0001; the rank o r d e r of intrinsic activity, in rings

c'© 9 C (D 6

3

0 0

"20

240

Time (s)

360

p r o - c o n t r a c t e d with 10 times the ECs~ ~ of phenyle p h r i n c , was a d e n o s i n e = N % m c t h y l a d e n o s i n e = N E C A > R - P I A = N~'-cyclohexyladenosinc. This rank o r d e r of intrinsic activity was similar to the rank o r d e r of intrinsic activity, in rings p r e - c o n t r a c t e d with the ECs0 of p h e n y l e p h r i n e , r e p o r t e d recently by us ( W i e n e r et al., in press), which was a d e n o s i n e = Nt'-methyl a d e n o s i n e > N E C A > R - P I A > N6-cyclohexyladenosine For the slope index: F ( 4 , 3 8 ) = 28.4, P < 0.0001; N 6cyclohexyladenosine had the highest slope index (1.02 _+ 0.02) a n d its value was significantly different than that of all o t h e r a d e n o s i n e agonists tested in rings p r e - c o n t r a c t e d with 10 times the ECs0 of phenyle p h r i n e . A l t h o u g h the values of the slope indices for a d e n o s i n e , N 6 - m e t h y l a d e n o s i n e , N E C A , and R - P I A were significantly less t h a n unity, they were, however,

Fig. 3. Sequential reslxmses to adenosine in the adventitia- and endothelium-denuded isolated rabbit thoracic aorta as a function of adenosine (panel A) or phenylephrine (panel B) concentration. Panel A illustrates the sequential responses of a single aortic ring to 1 (G), 3 (o), 10 ( v' ), 30 ( • ), 100 ( [] ), 300 ( • ), and 1000 ( ~ ) p-M adenosine in rings pre-contracted with 10 times the ECsu of phenylephrine. A cumulative concentration-response curve to phenylephrine was first constructed to estimate the ECs0 of phenylephrine (0.2 p.M). The ring was then contracted with 2 p-M phenylephrine (contraction not shown) and when the contraction plateaued the indicated concentration of adenosine was added to the organ bath to elicit the relaxation response. One hour was allowed to elapse between experimental trials. The symbols represent the digitized data and the smooth curves were obtained by fitting the tension and time data to equation 1 (Materials and methods) to estimate the values of the relaxation response, R, and the apparent rate constant for relaxation, kre I. For clarity, every other data point was illustrated. The values of the fractional relaxation (relaxation/contraction; R/C) response, k~ l (s-t), number of data points, and r 2 were as fl~llows: for 1 /~M adenosine: 0.05, 0.003, 45, and 0.94, respectively; for 3 p-M adenosine: 0.15, 0.005, 48, and 0.94, respectively; for 10 p-M adenosine: 0.37, 0.008, 53, and 0.97, respectively; for 30 p-M adenosine: 0.50, 0.012, 48, and 0.96, respectively; for 100 p-M adenosine: 0.61, 0.016, 48, and 0.96, respectively; for 300 p-M adenosine: 0.74, 0.017, 51, and 0.95, respectively; for 1000 p-M adenosine: 0.81, 0.018, 69, and 0.96, respectively. This experiment was repeated using nine rings from three rabbits with similar results. Panel B illustrates the relaxation responses of a single aortic ring to 200 ~M adenosine following pre-contraction with 1 (o), 3 (o), 10 (~71, 30 (•), and IlK}([]) times the ECso of phenylephrine. A cumulative concentration-resl~mse curve to phenylephrine was first constructed to estimate the ECso of phenylephrine. The ring was then contracted with the indicated concentration of phenylephrine (contraction not shown) and when the contraction plateaued 200 p-M adenosine was added to the organ bath to elicit the relaxation response. One hour was allowed to elapse between experimental trials. For clarity, every other data point was illustrated. The values of the fractional relaxation (relaxation/contraction; R/C) response, kre I (s-1), number of data points, and r e were as follows: for 1 ECs0 of phenylephrine: 0.83, 0.024, 50, and 0.95, respectively; for 3 ECso of phenylephrine: 0.37, 0.022, 36, and 0.98, respectively; for 10 ECs0 of phenylephrine: 0.21, 0.016, 46, and 0.97, respectively; for 30 ECso of phenylephrine: 0.14, 0.012, 24, and 0.93, respectively; for 100 ECs0 of phenylephrine: 0.11, 0.011, 20, and 0.94, respectively. This experiment was repeated using three rings from three rabbits with similar results.

not different than each other in rings pre-contracted with 10 times the ECs0 of phenylephrine. For the pECs0: F(4,38)= 100, P < 0.0001; the rank order of agonist potency, in rings pre-contracted with 10 times the ECs0 of phenylephrine, was NECA > R-PIA > adenosine = N6-cyclohexyladenosine > N6-methylade nosine, and this was consistent with an adenosine A 2 receptor subtype (Williams, 1990); furthermore, this rank order of agonist potency was consonant with the results observed by us in a previous study using rings pre-contracted with the ECs0 of phenylephrine (Wiener et al., in press). The apparent rate constant for relaxation, krel, for adenosine (fig. 4A), NECA (fig. 5A), N6-methyladeno sine (fig. 6A), R-PIA (fig. 7A) and N6-cyclohexyladenosine (fig. 8A) was also agonist concentration-dependent and saturable. The k rd concentration-response parameters for each analogue estimated using equation 3 ((krel)max, ECs0, and n; table 1) were also compared by one-way analysis of variance followed by Duncan's multiple-range test. For (kr~l)ma~: F(4,38)= 156, P < 0.8

0.6

g • --

0.020

~

0,015

0.4

0.010

0.2

0,005

(3 0.0 '

IJO

1001

0.000

10100

Adenosine ( ,u,M)

-6 C 0 0 0 L. L._

IL

--~, 0.024 ~_~

1.0

0,6 0.6

2~-

~

0.016

0.4. 0.008

0.2 0.0

[] I

1

, 3

L 10

, 30

I

100

I 0.000

[Phenylephrine]/EC5o Fig. 4. The fractional relaxation response ( R / C , o ) and apparent rate constant for relaxation (krc p e) were adenosine concentrationdependent and saturable (panel A), but inversely proportional to the phenylephrine concentration (panel B). Illustrated in panel A are sequential concentration-response curves to adenosine constructed in rings pre-contracted with 10 times the ECso of phenylephrine using the protocol illustrated in fig. 3A. The results are the mean + S.E.M. for nine rings from three rabbits. The ( R / C ) and krc mdata were fitted to equations 2 and 3 (Materials and methods), respectively, and the results are listed in table 1. Illustrated in panel B is the inverse dependence of both the fractional relaxation response to 200 ~ M adenosine and kre I for this response on the phenylephrine concentration constructed using the protocol illustrated in fig. 36. The results are the mean 5: S.E.M. for three rings from three rabbits.

o608i 0.0,5

69

0.010

0.4 ._

0.2

(3

0.0 I

0.005

t

I

o.ol

ol

I

1

lo

I

10o

0.000

~-

rY

-6

C 0

NECA

(/J.M)

1.0

0.03

I

0.8

LL

0.02

0.6 0.4

0 0 i m

1

,

3

10

'

.30

100

0.00

[Pheny!ephrine]/ECs0 Fig. 5. The fractional relaxation response ( R / C , O) and apparent rate constant for relaxation (krc=, e) were NECA concentration-dependent and saturable (panel A), but inversely proportional to the phenylephrine concentration (panel B). Illustrated in panel A are sequential concentration-response curves to NECA constructed in rings pre-contracted with 10 times the ECs0 of phenylephrine using a protocol similar to that illustrated in fig. 3A. The results are the mean _+S.E.M. for nine rings from three rabbits. The ( R / C ) and kr¢ I data were fitted to equations 2 and 3 (Materials and methods), respectively, and the results are listed in table 1. Illustrated in panel B is the inverse dependence of both the fractional relaxation response to 60 /zM NECA and kre I for this response on the phenylephrine concentration constructed using a protocol similar to that illustrated in fig. 3B. The results, which are from Wiener et al. (1992), are the mean 5: S.E.M. for three rings from three rabbits.

0.0001; the rank order for (krel)ma x was adenosine > NECA > R-PIA = N6-cyclohexyladenosine = N6-methyladenosine in rings pre-contracted with 10 times the ECs0 of phenylephrine. For the slope index: F(4,38) = 15.6, P < 0.0001; although N6-cyclohexyladenosine had the highest slope index (1.06 + 0.04) and its value was significantly different than that of all other agonists tested, it was not statistically different than unity in rings pre-contracted with 10 times the ECs0 of phenylephrine. Although the values of the slope indices for adenosine, N6-methyladenosine, NECA, and R-PIA were significantly less than unity, they were, however, not different than each other in rings pre-contracted with 10 times the ECs0 of phenylephrine. For the pECs0: F(4,38) = 143, P < 0.0001; as was observed for fractional relaxation concentrationresponse curves, the rank order of agonist potency for k,e t concentration-response curves was also NECA > R-PIA > adenosine = N6-cyclohexyladenosine > N 6methyladenosine in rings pre-contracted with 10 times

70 09

[Tq L..~

0 6

c0 I~ X (D

0.0

75

f

/

03

0005

,~

0004

0 003 0.002

0,001

~

~

10

'00

n__ 0000 1000

Methylocenosi'le

7~-

(/zM)

-

- 0.006

0

10

(3

0,8

[--

:

I

0004

0.6

o,l ,.

o

0.0 1

3

I

,

~

10

30

100

002

OCO0

[Phenylepqriqe~/ECso Fig. 6. The fractional relaxation response (R/C, o) and apparent rate constant for relaxation (krel, O) were N6-methyladenosine concentration-dependent and saturable (panel A), but inversely proportional to the phenylephrine concentration (panel B). Illustrated in panel A are sequential concentration-response curves to N6-methyladenosine constructed in rings pre-contracted with 10 times the EC~) of phenylephrine using a protocol similar to that illustrated in fig. 3A. The results are the mean±S.E.M, for seven rings from three rabbits. The (R/C) and kr~n data were fitted to equations 2 and 3 (Materials and methods), respectively, and the results are listed in table 1. Illustrated in panel B is the inverse dependence of both the fractional relaxation response to 1000/zM N~'-methyladenosineand kr~n for this response on the phenylephrine concentration constructed using a protocol similar to that illustrated in fig. 3B. The results are the mean + S.E.M. for three rings from three rabbits. the ECs0 of p h e n y l e p h r i n e . T h e pECs0 values for both the fractional relaxation a n d k,~j c o n c e n t r a t i o n - r e sponse curves were also c o m p a r e d . While they were similar for a d e n o s i n e (F(1,16) = 2.50, P = 0.13), R - P I A (F(1,16) = 0.49, P = 0.50), a n d N6-cyclohexyladenosine (F(1,16) = 3.62, P = 0.08), the pEC50 values of the fractional relaxation a n d kr~ c o n c e n t r a t i o n - r e s p o n s e curves were different for N E C A (F(1,16) = 8.73, P < 0.01) a n d N 6 - m e t h y l a d e n o s i n e (F(I,16) = 38.2, P < 0.0001).

3.2. lnuerse dependence of adenosine A 2 receptor-mediated relaxation responses on the phenylephrine concentration T o test the d e p e n d e n c e of relaxation response par a m e t e r s on the contractile stimulus, the c o n c e n t r a t i o n of p h e n y l e p h r i n e was varied from 1 to 100 times its EC~ 0 ( c o n c e n t r a t i o n s c o r r e s p o n d i n g to 8 - 9 0 % a l a d r e n o c e p t o r fractional occupancy) a n d aortic rings were relaxed with equiactive c o n c e n t r a t i o n s of a d e n o sine A 2 r e c e p t o r agonists using the protocol illustrated

in fig. 3B. T h e fractional relaxation response to a d e n o sine (figs. 3B and 4B), N E C A (fig. 5B), N"-methyla d e n o s i n e (fig. 6B), R - P I A (fig. 7B), or N~-cyclohexyla d e n o s i n e (fig. 8B) was inversely p r o p o r t i o n a l to the p h e n y l c p h r i n e c o n c e n t r a t i o n . Similarly, the a p p a r e n t rate c o n s t a n t s for relaxation were also inversely proportiona[ to the p h e n y l e p h r i n e c o n c e n t r a t i o n , indicating that the rate of g e n e r a t i o n of the relaxation response was i n f l u e n c e d by the a l - a d r e n o c e p t o r fractional occupancy and, therefore, the extent of the contractile stimulus. It was also interesting to note that the relaxation responses to the a d e n o s i n e A 2 receptor partial agonists, R-P1A (fig. 7B) a n d N6-cyclohcxyl a d e n o s i n e (fig. 8B), were abolished at p h e n y l e p h r i n c c o n c e n t r a t i o n s _> 30 times its ECs0, c o n f i r m i n g a recent report by us ( W i e n e r et al., in press). T o test if s t i m u l a t i o n of /3-adrenoceptors by high c o n c e n t r a t i o n s of p h e n y l c p h r i n e affected the a d e n o sine c o n c e n t r a t i o n - r e s p o n s e curvc the following experim e n t was p e r f o r m e d : c u m u l a t i v e c o n c e n t r a t i o n - r e sponse curves to a d e n o s i n e were c o n s t r u c t e d at differ0.5

- 0.0o5

0.4.

I 0004

05

I 0.003 i

g o2

"~

C.] (1)

t) (:3 I,

- 0.001

O0

~

0 1

10

I

100

0 000

~K-

I

R - P I A (uM)

75 E 0

to

01

0.9

½

O,OC6

0.6 ~ 0.3[ O0

~--!

~-~

J 0.004

0.002

. 3

--

x 10

0.000

[ Pb enyleph fine J/.~[CSC Fig. 7. The fractional relaxation response (R/C, O) and apparent rate constant for relaxation (kr¢ b O)were R-PIA concentration-dependent and saturable (panel A), but inversely proportional to the phenylephrine concentration (panel B). Illustrated in panel A are sequential concentration-response curves to R-PIA constructed in rings pre-contraeted with 10 times the ECs0 of pbenylephrine using a protocol similar to that illustrated in fig. 3A. The results are the mean ± S.E.M. for nine rings from three rabbits. The (R/C) and kreI data were fitted to equations 2 and 3 (Materials and methods), respectively, and the results are listed in table 1. Illustrated in panel B is the inverse dependence of both the fractional relaxation response to 2(X) ~M R-PIA and kre~ for this response on the phenylephrine concentration constructed using a protocol similar to that illustrated in fig. 3B. The results are the mean±S.E.M, for three rings from three rabbits.

71 0.5

0.004

0.40.20.3

(~

[ ~ ~

TABLE 1 Fractional relaxation ( R / C ) and apparent rate constant for relaxation (krcI) concentration-response curve parameters for adenosine analogues in adventitia- and endothelium-denuded isolated rabbit aortic rings pre-contracted with 10 times the EC.so of phenylephrine.

0.003 0.002 0.001

0.1

Analogue

Parameter

R/C a

kret a

Adenosine

Maximum Slope index pECs0 E C ~ (#M)

0.76±0.03 0.73+0.01 4.79±0.02 16

0.017 ± 0.001 0.77 +0.04 4.82 ±0.02 15

Maximum Slope index pEC,,0 ECs0 (txM)

0.47±0.05 1.02 + 0.02 4.87±0.01 14

0.0036±0.001 1.06 _+0.04 4.90 +0.01 13

Maximum Slope index pECs0 ECso (~M) Maximum Slope index pEC50 ECs0 (v,M) Maximum Slope index pECs0 ECs0 (,~M)

0.81 +0.01 0.84+0.03 4.55 +0.06 28 0.76+0.04 0.57+0.07 6.44+0.15 0.37 0.50 f 0.08 0.62±0.01 5.25 ±0.02 5.6

0.0053+0.0001 0.78 + 0.04 4.16 + 0.02 68 0.014 ± 0.001 0.57 ±0.08 5.89 +0.10 1.3 0.0050 ± 0.0001 0.59 ±0.01 5.27 +_0.02 5.4

.

0.001--

i

J

1

10

L

100

£E

Cyclohexylc]denosine -6t- 0.8

(,u,M)

~_

"G"

i 0.008 ~j

O

tD l,

0.000

N 6-Cyclohexyladenosine

0.006

0.6

t

0.004

0.4

N6-Methyl adenosine

0.002

0.2

NECA 0.0

J !

" 3

I 10

0.00C

[Phenylephr:ne]/ECs0 Fig. 8. The fractional relaxation response (R/C, o) and apparent rate constant for relaxation (kre I, e) were N6-cyclohexyladenosine concentration-dependent and saturable (panel A), but inversely proportional to the phenylephrine concentration (panel B). Illustrated in panel A are sequential concentration-response curves to N 6cyclohexyladenosine constructed in rings pre-contracted with 10 times the EC50 of phenylephrine using a protocol similar to that illustrated in fig. 3A, The results are the mean±S.E.M, for nine rings from three rabbits. The ( R / C ) and kreI data were fitted to equations 2 and 3 (Materials and methods), respectively, and the results are listed in table 1. Illustrated in panel B is the inverse dependence of both the fractional relaxation response to 100 v.M N6-cyclohexyladenosine and kreI for this response on the phenylephrine concentration constructed using a protocol similar to that illustrated in fig. 3B. The results are the mean ± S.E.M. for three rings from three rabbits.

ent phenylephrine concentrations in the absence and presence of 1 tzM (+)-propranolol to block the /3adrenoceptor. Adenosine concentration-response curves constructed in the presence and absence of propranolol were superimposible at all phenylephrine concentrations tested (data not shown) suggesting that there was no /3-adrenoceptor component in the response to phenylephrine.

4. Discussion

4.1. Kinetic characterization of adenosine Z 2 receptors in the isolated rabbit aorta: agonist concentration dependence and saturability of the apparent rate constant for relaxation and the fractional relaxation response The present communication describes the use of the kinetic model of Keitz et al. (1990) to characterize relaxation responses mediated by the adenosine A 2 receptor in the adventitia- and endothelium-denuded

R-PIA

a The values of the maximum, slope index, and pECso are the arithmetic means±S.E.M, for nine rings from three rabbits for all analogues except for N6-methyladenosine, were they are for seven rings from three rabbits. The ECso values are the geometric means. For statistical evaluation see Results.

isolated rabbit thoracic aorta. The denuded preparation was chosen since it provides several advantages over the more traditional intact ring; removing the adventitia eliminated nerve endings along with endogenous substances that can affect the observed response (Maxwell et al., 1968), and removing the endothelium eliminated any contribution it may have by releasing a variety of factors that can also affect the observed response (Furchgott, 1983). Removal of these two tissue components also eliminated any contribution that they may have as a diffusion barrier slowing the access of the vasodilator to the site of action (Wiener and Sehba, 1993). In contrast to the guinea pig aorta which contains both the A n and A 2 subtypes of adenosine receptors (Stoggali and Shaw, 1990), the denuded rabbit thoracic aorta appeared to contain only the adenosine A x receptor subtype since the traditional adenosine A n receptor subtype agonists, R-PIA and N6-cyclohexyladenosine, failed to elicit a contractile response. These analogues, however, relaxed phenylephrine-pre-contracted aortic rings and were less efficacious than adenosine, NECA, or N%methyladenosine. The rank order of adenosine A 2 receptor agonist potency observed in denuded rings pre-contracted with 10 times the EC50 of phenylephrine reported in the present

72

study was similar to the rank order of agonist potency reported by the following researchers: (1) Collis and Brown (1983) in rings prepared from isolated intact guinea pig aortae pre-contracted with the ECs0 of norepinephrinc; (2) Sabouni et al. (1990b) in rings prepared from isolated intact human left anterior descending coronary arteries prc-contracted with either 25 or 50 mM KCI; and (3) Headrick and Bernc (1990) in endothelium-intact and -denuded rings prepared from isolated guinea pig aortae pre-contracted with 2 /zM prostaglandin F2,, which was a concentration reported to produce 75% of maximal contraction. Furthermore, the rank order of adenosine A 2 receptor agonist intrinsic activity reported in the present study was similar to that reported by the aforementioned researchers. The kinetic model utilized in the present study allowed for the estimation of two parameters used in characterizing the relaxation response, the apparent rate constant for the relaxation response, k ~ , and the fractional relaxation response, R / C . Although the value of the fractional relaxation response for adenosine analogues could be obtained from steady-state studies, the value of k~ t can only be obtained from a kinetic evaluation. In addition for vasorelaxants, such as isoproterenol (Kcitz et al., 1990) and nitroglycerin (Wiener et al., 1992), whose relaxation responses are biphasic in the denuded aorta, estimation of the correct value of the fractional relaxation response cannot bc achieved by traditional steady-state analyses therefore necessitating the usage of such a kinetic model. Kinetic analysis of the digitized data allowed us to construct two sequential concentration-response curves for each agonist: a fractional relaxation response concentration-response curve and an apparent rate constant for relaxation concentration-response curve. While the ECs0 values of these two concentration-response curves were similar for adenosine, R-PIA, and N~'-cyclohcxyladenosine, they were different for NECA and N6-mcthyladenosine (table 1). For NECA and N6-methyladenosine the ratio of the k,~j ECs0 to the fractional relaxation response ECs0 were 3.5 and 2.4, respectively. This was similar to the ratio observed by us previously (Wiener ct al., 1992) for sodium nitroprusside (2.7), but different than that observed by Kcitz et al. (1990) for isoproterenol (0.25). These differences are intriguing in light of the fact that the two concentration-response curves constructed for adenosine, R-PIA, and N6-c3'clohexyladenosine had similar ECs0 values (table 1). 4.2. Is ( k r d ) m a x a suitable k&etic definition of drug efficacy? Keitz et al. (1990), based on their kinetic studies with a single ,t3-adrenoceptor agonist (isoproterenol),

suggested that the value of (krel)ma x m a y provide the kinetic definition of agonist efficacy. Their suggestion was based on the following: (1) there was a saturable dependence of krc ~ on the isoprotercnol conccntration; (2) there was a saturable dependence of the observed rate constant for the onsct of the tonic contractile responses (kob~) to an al-adrcnoceptor agonist and to 5-HT 2 receptor agonists in the denuded aorta (Cory et al., 1984, 1986); and (3) that the value of (k~,b~)m,X had already bccn suggested as a kinetic definition of drug efficacy at the 5-HT z receptor (Cory ct al., 1986). The adenosine A 2 receptor was chosen as a model system to test the hypothesis put forth by Kcitz et al. (1990) since adenosine analogues elicited a monophasic and stable relaxation response devoid of the inherent complication observed by Keitz et al. (1990) for isoproterenol, namely fade or desensitization of the relaxation response. Since the adenosine A 2 receptor did not appear to desensitize under the conditions employed in the present study, and since its responses are mediated by the activation of adenylyl cyclase, like the /3-adrcnoceptor, it was the receptor system of choice to probe the hypothesis of Keitz et al. (1990). Construction of sequential fractional relaxation concentration-response curves also allowed us to estimate the rank order of intrinsic activity (adenosine = N6-methyladcnosine = NECA > R-PIA = N6-cyclohexyladenosinc; table 1). The rank order of (krd)m~ x values was adenosine > NECA > R-PIA = N°-cyclohexyladenosine = N6-meth yladcnosine (table 1); this did not correlate with the rank order of intrinsic activity, suggesting that for adenosine A 2 receptor agonists, in the adventitia- and cndothelium-denuded rabbit aorta, the value of (krcl)ma x should not bc viewed as the kinetic definition of drug efficacy. It was also interesting to note that the three agonists with the lowest (krcJ)m~ values, R-PIA, N6-cyclohexyladenosine and N6-methyladcnosine, arc N6-substituted adenosine analogues, while NECA is a 5'-substituted analogue (see fig. 1). While it is therefore tempting to conclude that the value of (krel)max does not appear to be the kinetic definition of agonist efficacy for adenosine A 2 receptor agonists, a similar conclusion should not necessarily be extended to the /3-adrenoceptor until the proper kinetic experiments can be performed for a serics of /3-adrenoceptor agonists. 4.3. Inverse dependence of the fractional relaxation response and apparent rate constant for relaxation on the contractile stimulus The fractional relaxation responses to adenosine A 2 receptor agonists were inversely proportional to the phenylephrine concentration (figs. 4B, 5B, 6B, 7B, and 8B), confirming a recent report from our laboratory (Wiener et al., in press). Additionally, as the phenyl-

73 ephrine concentration was raised, the rate of relaxation declined as was evident from the inverse dependence of the apparent rate constants of relaxation on the phenylephrine concentration. The inverse dependence of these parameters on the phenylephrine concentration was also reported for the relaxation responses to isoproterenol (Osman et al., 1990) and nitroglycerin (Wiener et al., 1992). Interestingly, while the fractional relaxation response to sodium nitroprusside was inversely proportional to the phenylephrine concentration, the apparent rate constant for relaxation showed no d e p e n d e n c e on phenylephrine concentration (Wiener et al., 1992). As the phenylephrine concentration was raised to >_ 30 times its EC50, the fractional relaxation responses to the adenosine A 2 receptor partial agonists, R - P I A and N6-cyclohexyladenosine, were abolished, confirming our earlier reports (Wiener et al., 1991, and in press). This was also consonant with the report of Sabouni et al. (1990b) who stated that R-PIA failed to relax rings prepared from isolated intact human left anterior descending coronary arteries which were pre-contracted with 100 mM KCI. Analysis of a family of cumulative steady-state concentration-response curves, for the same adenosine analogues used in the present study, by the Black and Left (1983) operational model of agonism (Wiener et al., in press) revealed that the Black and Left efficacy parameter, ~-, was inversely proportional to the phenylephrine concentration and therefore to the a t - a d r e n o c e p t o r fractional occupancy and stimulus. The elimination of relaxation responses to the partial agonists observed in the present study was also similar to the results of Buckner and Saini (1975), and, more recently, Lemoine and Overlack (1992) who demonstrated that the steady-state relaxation responses to partial /3-adrenoceptor agonists in guinea pig trachea are attenuated more than those to full/3-adrenoceptor agonists as the muscarinic receptor contractile stimulus is increased. The question remaining to be answered is why does the rate of relaxation decrease as the contractile stimulus is increased? If we assume that the affinity of the adenosine agonists for their receptors and that the number of adenosine A 2 receptors are not affected as the a l - a d r e n o c e p t o r fractional occupancy is increased, then any change in either the rate or magnitude of response must be related to a change in the efficacy of the vasorelaxant agonist and its response. In their recent study of the functional antagonism between responses mediated by /32-adrenoceptors and muscarinic cholinergic receptors in guinea pig lung and trachea, Lemoine and Overlack (1992) addressed some of these concerns. They demonstrated that the binding of isoproterenol to membranes prepared from guinea pig lung was not affected by the addition of a maximal concentration of carbachol; furthermore they also reported that the coupling between the /32-adrenoceptor

and adenylyl cyclase was not affected by muscarinic receptor stimulation. The results of their study suggest that muscarinic receptor stimulation may alter /32adrenoceptor agonist efficacy at a step distal to the production of 3',5'-AMP by adenylyl cyclase. In summary, the present communication describes the use of the kinetic model developed by Keitz et al. (1990) to characterize the relaxation responses to adenosine A 2 receptor agonists in the adventitia- and endothelium-denuded isolated rabbit thoracic aorta. The fractional relaxation responses and their apparent rate constants were agonist concentration-dependent and saturable and inversely proportional to the contractile stimulus. The hypothesis put forth by Keitz et al. (1990) that the value of (krel)ma x m a y serve as a kinetic definition of efficacy was probed, but found to be apparently invalid for the five adenosine A 2 receptor agonists tested in the present study using adventitia- and endothelium-denuded isolated rabbit aortic rings pre-contracted with 10 times the ECs0 of phenylephrine. Of course an elaborate analysis would be required before a similar statement could be made for other receptors a n d / o r tissue preparations. We propose that the present pair of activated receptors and tissue preparation is a good model to study quantitative aspects of functional antagonism by kinetic paradigms.

Acknowledgement Supported by a grant from St. John's University (Award No. 050-92-002-2/6).

References Balwierczak, J.L., R. Sharif, C.M. Krulan, F.P. Field, G.B. Weiss and M.J.S. Miller, 1991, Comparative effects of a selective adenosine A 2 receptor agonist, CGS 21680, and nitroprusside in vascular smooth muscle, Eur. J. Pharmacol. 196, 117. Ben-Harari, R.R., B.A. Dalton, R. Osman and S. Maayani, 1991, Kinetic characterization of 5-hydroxytryptaminereceptor desensitization in guinea-pig trachea and rabbit aorta, J. Pharmacol. Exp. Ther. 257, 416. Black, J.W. and P. Left, 1983, Operational models of pharmacological agonism, Proc. R. Soc. London B 220, 141. Buckner, C.K. and R.K. Saini, 1975, On the use of functional antagonism to estimate dissociation constants for beta adrenergic receptor agonists in isolated guinea-pig trachea, J. Pharmacol. Exp. Ther. 194, 565. Christ, G.J,, J. Goldfarb and S. Maayani, 1990, Receptor-mediated mutual-effect amplification elicited by phenylephrine and serotonin in isolated rabbit aorta, J. Pharmacol. Exp. Ther. 252, 500. Clancy, B.M., R. Osman and S. Maayani, 1987, Kinetics of competitive drug action at 5-hydroxytryptamine2 receptor in isolated rabbit aorta, J. Pharmacol. Exp. Ther. 242, 108. Collis, M.G. and C.M. Brown, 1983, Adenosine relaxes the aorta by interacting with an A 2 receptor and an intracellular site, Eur. J. Pharmacol. 96, 61.

74 Cory, R.N., R. Osman and S. Maayani. 1984, Kinetic characterization of the rabbit aorta contractile reslxmse to an alpha adrenergic agonist, J. Pharmacol. Exp. Ther. 230, 162. Cory, R.N., R. Osman and S. Maayani, 1986, Kinetic definition of agonist efficacy at a 5-hydroxytry.ptamine (5-HT 2) receptor in the isolated rabbit aorta, J. Pharmacol. Exp. Ther. 236, 48. Cushing, D.J., G.L. Brown, M.II. Sabouni and S.J. Mustafa, 1991, Adenosine receptor-mediated coronary artery relaxation and cyclic nucleotide prt~uction, Am. J. Physiol. 261, 11343. De Lean, A., P.J. Munson and D. Rodbard, 1978, Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay, and physiological dose-response curves, Am. J. Physiol. 235, E97. Fleming, W.W., D.P. Westfall, I.S. De La Lande and L.B. Jellett, 1972. Log-normal distribution of equieffective doses of norepinephrine and acetylcholine in several tissues, J. Pharmacol. Exp. Ther. 181,339. Furchgott, R.F., 1983, Role of endothelium in responses of vascular smooth muscle, Circ. Res. 53, 557. Gaddum, J.]t., 1945, Lognormal distributions, Nature 156, 463. Ghai, G. and S.J. Mustafa, 1982, Demonstration of a putative adenosine receptor in rabbit aorta, Blood Vessels 19, 117. Headrick, J.P. and R.M. Berne, 1990, Endothelium-dependent and -independent relaxations to adenosine in guinea pig aorta, Am. J. Physiol. 259, H62. Herlihy, J.T., E.L. Bockman, R.M. Berne and R. Rubio, 1976, Adenosine relaxation of isolated vascular smooth muscle, Am. J. Physiol. 230, 1239. Ignarro, L.J. and P.J. Kadowitz, 1985, The pharmacological and physiological role of cyclic GMP in vascular smooth muscle relaxation, Annu. Rev. Pharmacol. Toxicol. 25, 171. Keitz, S.A., R. Osman, W.P. Clarke, J. Goldfarb and S. Maayani, 1990, Functional interactions in smooth muscle: kinetic characterization of the relaxation and desensitization responses to a beta adrenergic agonist in the rabbit aorta, J. Pharmacol. Exp. Ther. 255, 650. Lemoine, 11. and C. Overlack, 1992, ttighly potent beta-2 sympathomimetics convert to less potent partial agonists as relaxants of guinea pig tracheae maximally contracted by carbachol. Comparison of relaxation with receptor binding and adenylate cyclase stimulation, J. Pharmacol. Exp. Ther. 261,258. Maxwell, R.A., S.B. Eckhardt and W.B. Wastila, 1968, Concerning the distribution of endogenous norepinephrine in the adventitial and mcdia-intimal layers of the rabbit aorta and the capacity of these layers to bind tritiated norepinephrine, J. Pharmacol. Exp. Ther. 161, 34. McCormack, D.G., B. Clarke and P.J. Barnes, 1989, Characterization of adenosine receptors in human pulmonary arteries, Am. J. Physiol. 256, I-t41. Olah, M.E. and G.L. Stiles, 1992, Adenosine receptors, Annu. Rev. Physiol. 54, 211.

Onaran, It.O. and T.A. B6kesoy, 1990, Kinetics of antagonism at histamine-tl n receptors in isolated rabbit arteries, NaunynSchmied. Arch. Pharmacol. 341,316. Osman, R., S.A. Keitz, J. Goldfarb and S. Maayani, 1990, Kinetics of response and drug action, Neuropsychopharmacology 3, 417. Sabouni, M.H., D.J. Cushing and S.J. Mustafa, 1989. Adenosine receptor-mediated relaxation in coronary artery: evidence for a guanyl nucleotide-binding regulatory protein involvement, J. Pharmacol. Exp. Ther. 251,943. Sabouni, M.II., G.L. Brown, A.N. Kotake, F.I,. Douglas and S.J. Mustafa, 1990a, Effects of CGS-15943A on the relaxations produced by adenosine analogs in human blood vessels, Eur. J. Pharmacol. 187, 525. Sabouni, M.IT., M.V. Ramagopal and S.J. Mustafa, 1990b, Relaxation by adenosine and its analogs of potassium-contracted human coronary arteries, Naunyn-Schmied. Arch. Pharmacol. 341, 388. Sabouni, M.H., T. Hussain, D.J. Cushing and S.J. Mustafa, 1991, G proteins subserve relaxations mediated by adenosine receptors in human coronary artery, J. Cardiovasc. Pharmacol. 18, 696. Stiles, G.L., 1992, Adenosine receptors, J. Biol. Chem. 267, 6451. Stoggall, S.M. and J.S. Shaw, 1990, The coexistence of adenosine AI and A2 receptors in guinea-pig aorta, Eur. J. Pharmacol. 190, 329. Stollack, J.S. and R.F. Furchgott, 1983, Use of selective antagonists for determining the types of receptors mediating the actions of 5-hydroxytryptamine and tryptamine in the isolated rabbit aorta, J. Pharmacol. Exp. Ther. 224, 215. Tagliente, T.M., B.A. Dalton and R.R. Ben-Harari, 1992, Phasic responses to carbachol in isolated guinea pig trachea arc augmented by cooling and inhibited by nifedipine, J. Pharmacol. Exp. Ther. 261,755. Wiener, H.L. and F.A. Sehba, 1993, Kinetics of relaxation responses to vasorelaxants in intact and adventitia- and endothelium-denuded isolated rabbit thoracic aorta, Gen. Pharmacol. 24.43. Wiener. H.L., G.P. Thalody and S. Maayani, 1991, Interactions between responses mediated by simultaneous activation of adenosine A , and al-adrenergic receptors, Soc. Neurosci. Abstr. 17, 989. Wiener, H.L., J.M. Murray, G.P. Thalody and S. Maayani, 1992, Kinetics of relaxation responses to vasorelaxants in isolated rabbit aorta, Eur. J. Pharmacol. 220, 131. Wiener, H.L., G.P. Thalody and S. Maayani, in press, Interactions between responses mediated by activation of adenosine A 2 receptors and a~-adrenoceptors in the rabbit isolated aorta, Br. J. Pharmacol. Williams, M., 1990, Purine receptors, in: Receptor Pharmacology and Function, eds. M. Williams, R.A. Glennon and P.B.M.W.M. Timmermans (Marcel Dekker, New York) p. 503.