Investigations of prejunctional α2-adrenoceptors in rat atrium, vas deferens and submandibular gland

European Journal of Pharmacology, 211 (1992) 251-256

251

© 1992 Elsevier Science Pubhshers B.V. All rights reserved 0014-2999/92/$05.00

EJP 52284

Investigations of prejunctional a2-adrenoceptors in rat atrium, vas deferens and submandibular gland K a r e n Smith

1 Sonia

C o n n a u g h t o n 2 a n d J a m e s R. D o c h e r t y 1

Department of Chmcal Pharmacology, Royal College of Surgeons m Ireland, 123 St Stephen's Green, Dubhn 2, Ireland Recewed 29 July 1991, revmed MS recewed 5 November 1991, accepted 19 November 1991

We have examined the effects of a series of a2-adrenoceptor antagonists on the stimulation-evoked release of tritium from rat atrium, vas deferens and submandibular gland pre-incubated with [3H]noradrenaline, and correlated these potencies with affinities for the a2A-ligand binding site of human platelet and the a2B-ligand binding site of rat kidney. The am-selective adrenoceptor antagonists prazosin and ARC 239 showed significantly higher, and the a2A-selective antagonist BRL 44408 showed significantly lower, potency in atrium than in vas deferens and submandibular gland. Yohimbine and BDF 8933 failed to distinguish between prejunctional a2-adrenoceptors or between ligand binding sites. It is concluded that the preiunctlonal a2-adrenoceptor of rat atrium resembles the a2B-ligand binding site, and differs from the prejunctional a2-adrenoceptors of rat vas deferens and submandibular gland, which resemble the azA-ligand binding site a2-Adrenoceptors; a2A-Adrenoceptors; a2B-Adrenoceptors; Prazosin; ARC 239

I. Introduction

We have previously reported that the functional prejunctional a2-adrenoceptors of rat vas deferens and rat atrium differ, and resemble the a2A-ligand binding site of human platelet and the a2B-ligand binding site of rat kidney, respectively (Connaughton and Docherty, 1990). We have also reported that, based on the low prejunctional potency of the a2B-selective ligand ARC 239 (see Bylund et al., 1988), the functional prejunctional a2-adrenoceptor of rat submandibular gland resembles the a2A-ligand binding site (Docherty and Connaughton, 1991). However, other authors have suggested that the prejunctional a2-adrenoceptor of rat submandibular gland resembles the a2B-ligand binding site, based on a high prejunctional potency of prazosin (Turner et al., 1984). The purpose of the present study was to confirm our identification of the prejunctional a2-adrenoceptors in the three tissues by expanding our results in two ways.

Correspondence to: J.R. Docherty, Department of Clinical Pharmacology, Royal College of Surgeons in Ireland, 123 St. Stephen's Green, Dublin 2, Ireland. Tel. 353.1.780 200. 1 Present address Department of Physiology, Royal College of Surgeons in Ireland, 123 St Stephen's Green, Dubhn 2, Ireland. 2 Present address: Department of Pharmacology, St. Lores University School of Medicine, St Lores, MO 63104, U.S.A

Firstly, we have investigated further antagonists in rat submandibular gland so that our identification does not rely solely on the potency of a single agent (ARC 239). Secondly, since all our previous results were obtained using only non-selective or am-selective antagonists, we have obtained results in all three tissues employing the a2A-selective antagonist BRL 44408 (see Young e t a . , 1989). Some of these results have been published in shortened form (Connaughton and Docherty, 1991; Smith et al. 1991).

2. Materials and methods

Male Wistar rats (250-350 g) were obtained from Trinity College Dublin, Ireland and a variety of isolated tissues were used.

2.1. Radioactive overflow experiments Rat isolated atria, vas deferens or submandibular glands were preincubated for 1 h in 1 ml medium containing [3H]noradrenaline (0.5 /zmol 1-1, specific activity 39 Ci mmo1-1) before being superfused with [ 3H]noradrenaline-free Krebs-Henseleit solution of the following composition (mmol 1-t): NaCI 119, NaHCO 3 25, (+)-glucose 11.1, KCI 4.7, CaCI 2 2.5, K H 2 P O 4 1.2, MgSO 4 1.0, ascorbic acid 0.28, tetrasodium EDTA

252

0.03, corticosterone 0.03, propranolol 0.001. In addition, cocaine (3 /zmol 1-1) was present after pre-incubation with [3H]noradrenaline. Tissues were placed between platinum electrodes in organ baths, and superfused at a rate of 2 ml min-1 at 37°C. In all experiments, tissues were stimulated (supramaximal voltage, 0.5 ms pulses) 4 or 5 times (S0-S 4) for 3 min at a frequency of 2 Hz (atria) or 5 Hz (vas deferens and submandibular gland) at intervals of 27 min, beginning after 99 min of superfusion. Effluent samples were collected in 6 ml aliquots beginning after 120 min of superfusion so that sampling began 6 min before S r Antagonist drugs or distilled water vehicle were added to the superfusion stream at a rate of 50 ~1 min-1 in two or three cumulative concentrations beginning 12 min before S 2. Two tissues were obtained from each animal, one of which received the antagonist, the other received vehicle. Effects of antagonists on stimulation evoked release of tritium were corrected for changes occurring in vehicle experiments. At the end of the experiment, tissues were made soluble in 1 ml of tissue solubiliser.

2.2. Rat vas deferens isometric contraction Prostatic portions of rat vas deferens (four segments per animal) were placed between platinum electrodes in organ baths and bathed in Krebs-Henseleit solution of the same composition as listed above except that it did not contain ascorbic acid, EDTA, cocaine, corticosterone or propranolol. Tissues were attached to myograph transducers for recording of isometric contractions. Responses to single pulse field stimulation (supramaximal voltage, 0.5 ms pulses) were obtained at intervals of 5 min. When consistent control responses had been obtained, antagonist or vehicle was administered and 15 min later the a2-adrenoceptor agonist xylazine was administered cumulatively in 0.5 log unit increments at intervals of 5 min. Agonist IC50 values (concentration producing 50% of maximum inhibition of the contraction to a single stimulus) were obtained from individual experiments in which one tissue received vehicle and the others received antagonists in various concentrations. Antagonist potency was expressed as the dissociation constant K a from the equation K B = [ B ] / ( D R - 1), where [B] is the concentration of antagonist and D R is the agonist dose ratio produced by the antagonist, or as a pA 2 value. Antagonist pA 2 values were obtained from the x-intercept of the plot of log (agonist D R - 1) against log antagonist concentration, where the slope was not significantly different from negative unity (Arunlakshana and Schild, 1959).

2.3. Radioligand bmding studies Preparation of human platelet and rat kidney membranes was carried out exactly as described by Connaughton and Docherty (1990). The resultant pellets were used immediately or stored at - 2 0 ° C for later use.

In saturation experiments, aliquots of membrane suspension were incubated with various concentrations of [3H]yohimbine at 37°C (0.3-20 nmol l - l ; human platelet; incubation buffer: Tris-HC1 50 mmol 1-1, MgC12 8 mmol 1-1, E G T A 5 mmol 1- l, pH 7.4 at 37°C) or 25°C (0.5-30 nmol 1-1; rat kidney; incubation buffer: Tris-HC1 50 mmol 1-1, E D T A 5 mmol 1-l, pH 7.4 at 25°C) for 30 min. In competition studies, [3H]yohimbine (5 nmol 1-l) was incubated with displacing ligands in concentrations from 0.1 nmol 1-1 in 0.5 log unit increments for 30 min. Non-specific binding was determined in the presence of phentolamine 10 ~mol 1-1. Assays were terminated by washing with ice-cold incubation buffer, followed by rapid vacuum filtration through Whatman G F / C filters, using a Brandel Cell Harvester. Radioactivity retained on filters was determined by liquid scintillation spectroscopy. The inhibition constant (K,) for displacement of radiolabelled ligand was determined from the formula: K, = IC50/(1 + [3H]/KD) where IC50 is the concentration of competing ligand that inhibits radioligand specific binding by 50%, K D is the dissociation constant for the radioligand (platelet: 5.75 + 0.78 nmol 1-1, n = 6; kidney: 8.80 + 0.62 nmol 1-1, n = 7), and 3H is the concentration of tritiated yohimbine employed (5 nmol 1-1).

2.4. Statistical evaluation Results are expressed as means + S.E.M. or mean and 95% confidence limits. Antagonist EC30 values (concentration producing a 30% increase in stimulation-evoked overflow of tritium) were compared between tissues, and effects of antagonist on stimulationevoked release of tritium or on basal overflow of tritium were compared with the effects of vehicle using a Student's t-test for unpaired or paired data, where appropriate. Linear regression analysis was used to determine correlations between parameters, and P values of less than 0.05 were considered statistically significant (t-test and ANOVA). The non-parametric Mann-Whitney test was used where data were not normally distributed (see Snedecor and Cochran, 1980).

2.5. Drugs The following drugs were used: ARC 239 (2-(2,4-(0methoxyphenyl)-piperazin- 1-yl)-ethyl-4,4 dimethyl-l,3(2H,4H)-isoquinolindine chloride (gift: Karl Thomae,

253 Biberach, F.R.G.), B D F 8933 (4-fluoro-2-(imidazolin2-ylamino)-isindoline maleate; gift: Beiersdorf, H a m burg, F.R.G.); B R L 44408 (2-((4,5-dihydro-lH-imidazol-2-yl)methyl)-2,3-dihydro-l-methyl-lH-isoindole; gift: Smith Kline Beecham, Epsom, U.K.); chlorpromazine hydrochloride (Sigma, Poole, U.K.); cocaine hydrochloride (Sigma); corticosterone (Sigma); H V 723 (a-ethyl-3,4,5-trimethoxy-a-(3-((2-(2-methoxyphenoxy)ethyl)-amino)-propyl)-benzene acetonitrile fumarate; gift: Hokurika, Seiyaku, Katsuyama, Japan); prazosin hydrochloride (gift: Pfizer, Sandwich, U.K.); (+_)-propranolol hydrochloride (Sigma); yohimbine hydrochloride (Sigma). D r u g stocks were dissolved in distilled water and dilutions m a d e up in distilled water, with the exception of corticosterone which was m a d e up in 100% ethanol.

chlorpromazine in vas deferens, nor for prazosin in submandibular gland; in these cases EC30 is q u o t e d as less than a certain concentration ( - l o g M). T h e EC30 values obtained are shown in table 1. Yohimbine and B D F 8933 had similar potencies in all three tissues, but A R C 239 and H V 723 showed significantly higher potencies, and B R L 44408 showed significantly lower potency, in rat atrium than in the other tissues (Student's t-test; P < 0.05). Prazosin and chlorpromazine were also m o r e potent in atrium than in the other tissues even t h o u g h absolute EC30 values were only available for the atrium. Using the non-parametric M a n n - W h i t n e y test, prazosin and chlorpromazine were significantly m o r e potent in the atrium than in the vas deferens, and prazosin was significantly m o r e potent in the atrium than submandibular gland (P <

0.05). 3. Results

3.2. Vas deferens isometric contraction

3.1. Radioactive overflow experiments

In prostatic portions of rat vas deferens, single pulse electrical stimulation p r o d u c e d an isometric contraction of 1.11 _+ 0.07 g (n = 53). In vehicle experiments, xylazine p r o d u c e d c o n c e n t r a t i o n - d e p e n d e n t inhibition of the stimulation-evoked contraction with an ICs0 of 63.1 nmol 1-1 (95% confidence limits of 51.3-77.6 nmol l-1). Antagonists in the concentrations employed did not significantly affect the stimulation-evoked isometric contractions except for chlorpromazine (30 ~ m o l 1-1: r e d u c e d contraction to 63.9 _+ 16.8%). Potency of chlorpromazine was expressed as a K B from the effects of 3 and 10 tzmol l -~. F o r prazosin and A R C 239, since slope of Schild plots were significantly different from negative unity, p o t e n c y was expressed as a K B from the effects of the two lowest concentrations to p r o d u c e shifts in agonist potency. T h e p A 2 or K B values are shown in table 1.

Field stimulation of rat isolated tissues at a frequetlcy of 2 H z (atria) or 5 H z (vas deferens, submandibular gland) for 3 min p r o d u c e d an evoked overflow of tritium at S~ of 1.76 + 0.18% (n = 62), 1.22 + 0.10% (n = 32) and 2.35 + 0.10% (n = 24) of tissue tritium in atrium, vas deferens and submandibular gland, respectively. Basal outflow of tritium was 0.15 + 0.016, 0.13 + 0.007 and 0.17 + 0.010% of tissue tritium per min, respectively, before S r In the concentrations employed to increase stimulation-evoked overflow of tritium, antagonists did not significantly alter the basal outflow of tritium with the exception of prazosin (1 /~mol 1-1." all three tissues), chlorpromazine 3 /xmol 1-1 (vas deferens). Hence, it was not possible to calculate an EC30 for prazosin or

TABLE 1 EC30 values (concentration producing a 30% increase in stimulation-evoked overflow of tritium) obtained for a-adrenoceptor antagonists m rat

atrium, vas deferens and submandibular gland, and pA 2 values obtained in rat vas deferens Values are means and 95% confidence limits (-log M) from at least four experiments (except for BDF 8933 m submandibular gland, where n = 3) Some of these results have been pubhshed previously (Connaughton and Docherty, 1990)

Chlorpromazlne BDF 8933 Prazosin ARC 239 Yohimbine BRL 44408 HV 723

Atrmm EC 30 6.53 + 0 36 9.18 + 0.24 6.97 + 0.26 7.40 + 0.63 7 89 + 0.14 6 49+0.37 7 15+0.24

Vas deferens EC 30 < 6.0 c 9.26 + 0.32 < 6.5 c 6 17 + 0.71 b 7.60 + 0.04 7 24+0.64 b 6 54+0.53 b

Submandlbular EC 3o _ 8.88 + 1.22 < 6.5 c 5.95 _+0.75 b 7.35 + 0.88 7.17+0.28 b 6 13+0.14 b

Vas deferens pA 2 5.65 _+0.18 a.b 8.92 5.88 + 0 20 a.b 5.71 + 0.24 a.b 7.50 7.63 b 6 22 b

K8 b.c Denote potency in rat vas deferens or submandlbular gland sigmficantly different from potency in rat atrium (b Student's t-test, P < 0 05, c Mann-Whitney test, P < 0 05). a

254 TABLE 2

than with the 0/2A-ligand binding site, but the reverse was true for vas deferens and submandibular gland (table 3).

K~ values obtained for a-adrenoceptor antagomsts at displacing [3H]yohlmbme binding m human platelet (a2A) and rat kidney (Ot2B) membranes Values are means and 95% confidence hmlts ( - l o g M) from at least three experiments. Some of these results have been pubhshed previously (Connaughton and Docherty, 1990). Antagomst

Platelet K,

Kidney K,

Chlorpromazme BDF 8933 Prazosm ARC 239 Yohlmbme BRL 44408 HV 723

6 48 _+0 29 8.72 _ 0.25 5 62 + 0.40 5 45 _+0 04 8.04 + 0.25 7.28 _+0 42 6.58 + 0.42

6 29 _+0 08 8.91 _+0.60 7 12 _+0 17 6.76 _+0.13 7.93 _+0.07 6.11 + 0.32 7 22 _+0.67

4. Discussion

In this study, we have confirmed our previous suggestion that the prejunctional a2-adrenoceptor of rat atrium resembles the a2B-ligand binding site of rat kidney, and that the prejunctional a2-adrenoceptors of rat vas deferens and rat submandibular gland resemble each other and the a2A-ligand binding site of human platelet (Connaughton and Docherty, 1990; Docherty and Connaughton, 1991). Our previous identification of prejunctional a 2adrenoceptors in rat atrium and vas deferens was carried out using non-selective or a2B-selective ligands. However, we have found that the test compound BRL 44408, which shows selectivity for a2A-ligand binding sites (see also Young et al., 1989), also showed higher potency at the prejunctional a2-receptors of rat vas deferens and submandibular gland than at the prejunctional a2-adrenoceptors of rat atrium. BRL 44408 had 15 (present study) or 85 (Young et al., 1989) times higher affinity for the a2A- over the a2B-site in ligand binding studies, and showed 6 and 5 times, respectively, higher potency at prejunctional a2-adrenoce ptors in rat vas deferens and submandibular gland than in rat atrium in functional studies. In contrast, ARC 239, prazosin and, to a lesser extent, HV 723 showed selectivity for the a2B- over the a2A-ligand binding site, having 20, 32 and 4 times higher affinity for 0/2B, respectively. In studies from other laboratories, prazosin and ARC 239 were reported to have 50 and 55 times higher affinity for 0/2B than 0/2A (Bylund et al., 1988) and prazosin was reported to have 78 times higher affinity for 0/2B than 0/2A (Michel et al., 1989). HV 723 is a potent ai-adrenoceptor antagonist which has been reported to distinguish between subtypes of al-adrenoceptor in functional studies (Muramatsu et al., 1990), but, as far as we are aware, has not been examined in other studies of az-adrenoceptors. ARC

3.3. Radioligand binding studtes K, values for the displacement by antagonists of [3H]yohimbine binding to human platelet and rat kidney membranes were obtained (see table 2). Hill slopes for all antagonists were close to unity, so that it was assumed that a single homogeneous population of ligand binding sites was present in both tissues.

3.4. Correlation between tissues Correlations among functional prejunctional o/2adrenoceptors and correlations between functional prejunctional 0/2-adrenoceptors and ligand binding sites are shown in table 3. Potencies of antagonists at functional prejunctional az-adrenoceptors were taken as the EC30 values shown in table 1, except that the potencies of prazosin and chlorpromazine in vas deferens were taken as the K B values obtained in twitch experiments. The prejunctional az-adrenoceptor of atrium showed a poor correlation with the prejunctional 0/2-adrenoceptors of vas deferens or submandibular gland, whereas the prejunctional 0/2-adrenoceptor of rat vas deferens showed a good correlation with the prejunctional 0/2-adrenoceptor of submandibular gland (table 3). The prejunctional 0/2adrenoceptor of atrium correlated better with the 0/2B" TABLE 3

Correlation of antagomst potenoes (EC30) and affimtles (K t) between tissues. Correlation coefficient (r) and slgmficance of correlation are shown Correlations were obtained w,th seven antagonists except those involving the submandxbular gland where six antagomsts were used (see table 1). Vas deferens

Platelet 2A (K,) I(adney 2B (K~) Vas deferens (EC30) Atrmm (EC30)

Submand,bular

Atrmm

r

P<

r

P<

r

P<

0 90 a 0 72 a -

0.01 NS -

0.89 0.71 0 95 a 0 76

0.05 NS 0.01 NS

0 66 0.95 0.79 a -

NS 0 001 0.05 -

a Correlanon obtained including vas deferens K n values for chlorpromazme and prazosm.

255 239, prazosin and HV 723 showed 17, > 3 and 4 times lower potency at prejunctional a2-adrenoceptors in rat vas deferens, and 28, > 3 and 10 times lower potency at prejunctional a2-adrenoceptors in rat submandibular gland, than at prejunctional a2-adrenoceptors in rat atrium, respectively. The above results comparing relative potencies of prazosin, ARC 239, H V 723 and BRL 44408, suggest that the prejunctional a2-adrenoceptors of rat vas deferens and submandibular gland resemble each other but differ from the prejunctional a2-adrenoceptor of rat atrium. Furthermore, the prejunctional a2-adrenoceptor of rat vas deferens and rat submandibular gland resemble the a2A-ligand binding site and the prejunctional a2-adrenoceptor of rat atrium resembles the a2B-ligand binding site. These conclusions are supported by the results of correlation analysis (see Resuits). Other authors have investigated subtypes of functional a2-adrenoceptors. Prazosin was found to have higher potency at the prejunctional a2-adrenoceptors of rat submandibular gland (Turner et al., 1984), rat cerebral cortex (Nasseri and Minneman, 1987) and rabbit pulmonary artery (Kapocsi et a1.,1987), suggesting, based on the relatively high potency of prazosin, that prejunctional a2-adrenoceptors in these tissues may resemble the a2B-ligand binding site. However, the present results, which show that the prejunctional a2-adrenoceptor of rat submandibular gland resembles the a2A-ligand binding site based on the use of four antagonists which show subtype selectivity, demonstrate the unreliability of identifying subtypes of a 2adrenoceptor based on the potency of only one antagonist. Admittedly, the present study also differed methodologically from that of Turner et al. (1984) in that the latter authors measured release of endogenous noradrenaline. Likewise, the low potency of prazosin prejunctionally at the cholinergic nerves of guinea-pig ileum (Kapocsi et al., 1987) and at the noradrenergic and serotonergic nerve terminals of rat brain synaptosomes (Gobbi et al., 1991) may suggest that these receptors resemble the a2A-ligand binding site. In rat cerebral cortex, other authors have found evidence that two subtypes of adrenoceptor modulate noradrenaline release: the effects of exogenous noradrenaline were prazosin-sensitive whereas the effects of endogenous noradrenaline were prazosin resistant (Harsing and Vizi, 1991). In a study employing several subtype selective antagonists including prazosin, ARC 239 and BRL 44408, the prejunctional a2-autoreceptor of rabbit brain cortex has been identified as resembling the a2A-ligand binding site (Limberger et al., 1991). Molecular cloning techniques have demonstrated at least three genes which code for subtypes of a2-adrenoceptor on chromosomes 2, 4 and 10 in man and in rat (Regan et al., 1988; Lorenz et al., 1990; Flordellis et

al., 1991). The a2-C10 receptor corresponds to the azA-ligand binding site, while the a2-C2 receptor corresponds to the a2a-ligand binding site (Lorenz et al., 1990; Weinshank et al., 1990; Zeng et al., 1990), but the a2-C4 may correspond to the aEc-ligand binding site of an opossum kidney cell line (Bylund et al., 1991) or to an unrecognised subtype of the a2B-receptor (Lomasney et al., 1991). A fourth subtype of ligand binding site has been identified in the bovine pineal gland (a2D: Bylund et al., 1991), and perhaps also in the rat submandibular gland (Michel et al., 1989), but may be a homologue of the human a2A-receptor (Lanier et al., 1991). Hence, due to species variations, ligand binding techniques may produce differences between homologues of the same gene-coded receptor. In this regard, it must be remembered that the model for the a2A-ligand binding site employed in this study was the human platelet a2A , SO that some differences from the rat OIZA (which may be the azD-ligand binding site: Lanier et al., 1991) might be expected: this may be reflected in the fact that the potency of chlorpromazine in rat vas deferens prejunctionally was lower than would be expected from its binding affinity in human platelet. In conclusion, the prejunctional az-adrenoceptors of rat vas deferens and rat submandibular gland resemble each other and differ from the prejunctional a2-adrenoceptor of rat atrium. These two subtypes of functional az-adrenoceptor resemble the a2A- and azB-ligand binding sites, respectively.

Acknowledgements Supported by the Irish Heart Foundahon, the Health Research Board (Ireland) and by the Royal College of Surgeons m Ireland.

References Arunlakshana, O and H.O. Schild, 1959, Some quantitative uses of drug antagonists, Br. J Pharmacol. Chemother. 14, 45 Bylund, D B., C. Ray-Prenger and T.J. Murphy, 1988, Ot2A- and otaB-adrenerglc receptor subtypes antagomst binding in tissues and cell hnes containing only one subtype, J. Pharmacol. Exp. Ther. 245, 600. Bylund, D B., H.S. Blaxall, T.J. Murphy and V Slmmoneaux, 1991, Pharmacological evidence for a2c- and a2o-adrenergic receptor subtypes, m. Adrenoceptors. Structure, Mechanisms, Function, Advances m PharmacologicalSciences, eds. E. Szabadl and C.M. Bradshaw (Blrkhauser Verlag, Basel) p. 27 Connaughton, S and J.R. Docherty, 1990, Functional evidence for heterogeneity of peripheral prejunctlonal a2-adrenoceptors, Br. J. Pharmacol 101, 285. Docherty, J.R and S Connaughton, 1991, Role and pharmacologyof vascular pre- and postjunctlonal a2-adrenoceptors, m Aspects of Synaptlc Transmission, ed. T Stone (Taylor and Francis, London) p 224.

256 Flordelhs, C.S., D.E Handy, M.R. Bresnahan, V I Zannls and H Gavras, 1991, Cloning and expression of a rat brain alpha2Badrenerglc receptor, Proc. Natl. Acad. Scl. U.S.A 88, 1019. Gobbl, M , E Fnttoh and T Mennml, 1990, The modulation of [3H]noradrenahne and [3H]serotonm release from rat brain synaptosomes is not mediated by the ot2n-adrenoceptor subtype, Naunyn-Schmledeb. Arch Pharmacol. 342, 382. Harsmg, L.G. and E.S. Vlzl, 1991, Evidence that two stereochemlcally different a2-adrenoceptors modulate norepmephrme release m rat cerebral cortex, J Pharmacol. Exp Ther. 256, 44 Kapocsl, J , G.T Somogyl, N. Ludvlg, P Serfozo, L G Harsmg, R.J. Woods and E S. Vlzl, 1987, Neurochemlcal evidence for two types of presynaptlc ot2-adrenoceptors , Neurochem Res. 12, 141. Lanier, S.M, S Downing, E. Duzlc and C.J Homey, 1991, Isolation of rat genomlc clones encoding subtypes of the a2-adrenerglc receptor, J. Biol. Chem. 266, 10470 Lorenz, W., J W Lomasney, S Collins, J.W. Regan, M.G. Caron and R.J. Lefkowltz, 1990, Expression of three a2-adrenerglc receptor subtypes in rat tissues: Imphcatlons for a2-receptor classification, Mol Pharmacol 38, 599 Limberger, N , L. Spath and K Starke, 1991, Subclasslficatlon of the presynaptlc ot2-autoreceptors in rabbit brain cortex, Br J Pharmacol 103, 1251. Lomasney, J.W., S. Cotecchm, W. Lorenz, D A Schwmn, M.G. Caron and R.J. Lefkowitz,1 991, Structure and molecular characterization of a-adrenerglc receptors, in. Adrenoceptors: Structure, Mechanisms, Function, Advances m Pharmacological Sciences, eds. E Szabadl and C M. Bradshaw (Birkhauser Verlag, Basel) p. 67. Michel, A.D., D.N. Loury and R.L. Whiting, 1989, Differences

between the c~2-adrenoceptor m rat submaxillary gland and the Ot2A- and t~2B-adrenoceptor subtypes, Br. J. Pharmacol. 98, 890 Muramatsu, I, T. Ohmura, S K~goshl, S Hasblmoto and M Oshlta, 1990, Pharmacological subclassificatlon of ~tl-adrenoceptors in vascular smooth muscle, Br. J Pharmacol. 99, 197 Nasserl, A and K.P Mmneman, 1987, Relationship between a 2adrenerglc receptors mhlbmng norepmephrme release m rat cerebral cortex Mol Pharmacol 32, 655 Regan, J.W, T S Kobdka, T L Yang-Feng, M G Caron, R J Lefkowltz and B.K. Kobllka, 1988, Cloning and expression of a human kidney cDNA for an a2-adrenergic receptor subtype, Proc Natl Acad. Sc. U.S A. 85, 6301. Smith, K., S. Connaughton and J.R. Docherty, 1991, Subtypes of functional a2-adrenoceptors m the rat, Fundam Chn. Pharmacol 5, 453. Snedecor, G W. and W G Cochran, 1980, Statistical Methods (Iowa State University Press, Ames) Turner, J T., D.L. Pierce and D.P. Bylund, 1984, a2-Adrenerglc regulation of norepinephrlne release m the rat submandlbular gland as measured by HPLC-EC, Life Scl. 35, 1385. Wemshank, R.L, J M Zgomblck, M Macchl, N. Adham, H Lichtblau, T.A. Branchek and P R. Hartlg, 1990, Cloning, expression, and pharmacological characterization of a human a2B-adrenerglc receptor, Mol. Pharmacol 38, 681. Young, P., J Berge, H. Chapman and M.A. Cawthorne, 1989, Novel ae-adrenoceptor antagonists show selectivity for ~2A" and a2Badrenoceptor subtypes, Eur. J. Pharmacol. 168, 381 Zeng, D , J.K. Harrison, D D D'Angelo, C.M. Barber A.L. Tucker, Z Lu and K.R Lynch, 1990, Molecular characterization of a rat c~2B-adrenerglc receptor, Proc Natl Acad Scl. U.S.A 85, 6301