CCB, a novel specific κ opioid agonist, which discriminates between opioid and σ1 recognition sites

Life Sciences. Vol. 57, No. 16, pp. 1487-1495, 1995 Copyright 0 1995 Elsevier Science Inc. Printed in the USA. All rights reserved 0024-3205/95 $9.50 + .OO

Pergamon

0024-3205(95)02120-S

CCB, A NOVEL SPECIFIC K OPIOID AGONIST, WHICH DISCRIMINATES BETWEEN OPIOID AND 01 RECOGNITION SITES G.Ronsisvalle’,

O.Prezzavento’, J.A.Gomez-Vidal’,

L.Pasquinucci’, A.Marrazzo’, L.Carboni’, S.Spampinato2

F.Vittorio’,

‘Institute of Pharmaceutical Chemistry, University of Catania, Italy, *Department of Pharmacology, University of Bologna, Italy. (Received

in final form August

3, 1995)

Summary 6,ll -Dimethyl-1,2,3,4,5,6-hexahydro-3-{[2’-methoxycarbonyl-2’-(4CCB, chlorophenyl)cyclopropyl]methyl}-2,6-methano-3-benzazocin-8-ol, displays specificity and very high affinity for K opioid receptor types (K, = 0.41+0.19 nM). In contrast to other K opioid agonists, CCB is also selective with respect to o, sites (K, = 1,050+55 nM). CCB displays antinociceptive and sedative effects in the mouse comparable to those of U50,488H and morphine. Naltrexone fully antagonizes these effects. The sedative effects of CCB are blocked in mice pretreated with naltrexone or nor-BNI. CCB and U50,488H produce a superimposable diuresis in rats. Naltrexone and nor-BNI, both are effective in antagonizing the effect. CCB does not produce any stereotyped behavior or ataxia in the behavioral assay in doses up to 40 mg/kg s.c.. These findings suggest that CCB might be a useful tool to investigate the physiological role of K opioid receptors. Key Words: CCB, specific K opioid agonist, /J binding, K binding, 6 binding, aI binding Benzomorphan derivatives represent invaluable tools to characterize K opioid receptors (1) and have also allowed the recognition of o binding sites (2). However, these compounds display heterogeneity of binding affinities towards different recognition sites (3,4). In fact, benzomorphans reported as K agonists, such as bremazocine, ethylketocyclazocine, pentazocine, and N-allyl-normethazocine (SKF10,047) may also interact, either in vitro or in viva, with other opioid receptor types and/or with o sites (5-8). In other words, (+)-pentazocine and (+)-SKF10,047 highly recognize the o1 sites (9,10), however, their (-)-isomers, which preferentially bind to K opioid receptors, afso display a moderate affinity for o sites or for other opioid receptor types (5,6,8,11). In order to identify a benzomorphan-derived pharmacophore required for selective binding to the K opioid receptor and capable of discriminating between K Corresponding Author: Giuseppe Ronsisvalle, Institute of Pharmaceutical Chemistry, University of Catania, V.le A.Doria, 6 - 95125 Catania - Italy. Phone/Fax: +39 95 336722.

1488

CCB: A Novel Specific

K

Opioid Agonist

opioid receptors and o1 sites, we investigated new series of (-)-benzomorphan derivatives.

Vol. 57, No. 16, 1995

the binding profile and in vivo activity of a

We recently reported the opioid agonist activity of a series of carboxy esters derived from normetazocine (12). Among the synthesized ligands, MPCB [(-)-R,S-6,1 lDimethyl-l ,2,3,4,5,6-hexahydro-3-[(2’-methoxycarbonyl-2’-phenylcyclopropyl)methyl]2,6-methano-3-benzazocin-8-011 (Fig.1) was found to be a specific K opioid agonist. Although displaying a lower affinity towards K receptors than U50,488H, MPCB was active in the mouse tail flick assay in doses comparable to U50,488H after subcutaneous administration. MPCB differs from bremazocine for the contemporary presence of a phenyl ring and an H-bonding group on the cyclopropyl ring, and it has been proposed that it might bind the same receptor site of arylacetamide analogues (12,13). The aim of this study was to investigate a series of derivatives of MPCB with chloro or methoxy substituents in the phenyl ring (Fig.1) and to increase the affinity towards the K opioid receptor while maintaining selectivity. We also addressed our experiments to a key question of growing importance in recent years, i.e. the selectivity of opioid ligands with respect to o receptors due to the fact that many psichotomimetic and disphoric side effects exhibited by K ligands might be at least in part attributable to interactions with these receptors (2).

HO

Fig.1

MPCB: R=Me, R’=H CCB: R=Me, R’=4-Cl 1 R=Me, R’=3,4-Cl, 2 R=Me, R’=2-OMe 3 R=Me, R’=3-Cl

Materials and Methods

Chew. The series of derivatives of MPCB with chloro or methoxy substituents in the phenyl ring (Fig.1) was prepared by an alkylation of (-)nor-metazocine following a procedure previously reported (12). Elemental analysis, IR, nmr and mass spectra were performed to confirm these structures.

Vol. 57, No. 16, 1995

CCB: A Novel Specific K Opioid Agonist

1489

m. The following drugs were used: morphine sulphate, U50,488H (Upjohn), (DAla’, D-Leu’)enkephalin (DADLE), D-Ala2,N-Me-Phe-4,Gly-ol)enkephalin (DAMGO) (Peninsula, England), (+)-N-allyl-normetazocine (SKF10,047), naltrexone and norbinaltorphimine (nor-BNI; RBI, Natik, MA, U.S.A.), [3H]diprenorphine (Amersham Italia), [3H]pentazocine (NEN, Boston, MA, U.S.A.). . . . mdrutudres to oprord recw. To evaluate the specific binding to K sites, a crude membrane fraction from male Hartley guinea pig cerebella (300-350 g, Charles River, Calco, Como, Italy) was prepared as described (14,15). Aliquots (100 pg / ml protein) were incubated at 25°C for 60 min with 1 nM [“Hldiprenorphine (30 Ci/mmol, K, = 0.55kO.7 nM; n = 5) and various concentrations of test compounds. Binding studies were performed in the presence of DAMGO (100 nM) and DADLE (100 nM) to eliminate the interaction with f,~and 6 receptors, respectively. Non specific binding was determined by the addition of MR2266 (100 nM, a purported K receptor antagonist). Binding to ~1and 6 sites was carried out on crude membrane fraction obtained from the whole rat brain minus cerebellum as previously reported (12). To evaluate the specific binding to u sites, displacement of the binding of [3H]diprenorphine (1 nM) was measured by the addition of U50,488H (100 nM) and of DADLE (100 nM) added to saturate the K and 6 receptors, respectively (the K, of the radioligand was 0.2250.03 nM; n = 5). Non specific binding was determined by the addition of DAMGO (100 nM). To evaluate the specific binding to 6 sites, the displacement of the binding of [3H]diprenorphine (the K, of the radioligand was 0.44+0.03 nM; n = 5) was measured in the presence of U50,488H (100 nM) and DAMGO (100 nM) added to saturate the K and u receptors, respectively. Binding assays were carried out as previously referred. Non specific binding was determined in the presence of DADLE (100 nM). a to IS receotprs Brains minus cerebella were dissected from male Hartley Guinea pigs. Membrane fraction homogenate was prepared following the method of Bowen et a/. (16) and protein content was evaluated (17). Binding assay was performed as described by DeHaven et al. (10). Briefly, each assay tube contained 500 pg of membrane proteins, 3 nM [3H]pentazocine (31.6 Ci/mmol, K, 4.3kO.7 nM; n = 3) and the test compound in a concentration range from 10.” to 1O-5 M. Non-specific binding was determined by adding 1 f.rM haloperidol. The reaction was performed for 150 min at 37 o C and terminated by rapid filtration over Whatman GF/B glass fiber filters that were presoaked in a 0.5% polyethylenimine solution. Radioactivity of the filters was measured by liquid scintillation spectrometry using a Beckman LS 1701 counter after overnight incubation in scintillation cocktail. Scatchard parameters and inhibition constants (K, values) were calculated using EBDA-LIGAND program (18). Anim&+ Male Sprague-Dawley rats and male Swiss mice were obtained from Charles River (Calco, Como, Italy) and housed in plasic cages in a temperature-controlled room (2252°C) with a 12h light, 12h dark cycle and 60% humidity. They received food and water ad libitum for at least 10 days before beeing used experimentally. . Male mice (20-25 g) were used. Agonists were administered subcutaneously (s.c.) in a dose-volume of 10 ml/kg 30 min before testing. Opioid antagonists were administered S.C. 10 min before agonists. Antinociception was measured by the acetylcholine-induced abdominal constriction test (19). Acetylcholine chloride (10 mg/kg, dissolved in 10 ml of saline) was injected intraperitoneally and mice were observed for 5 min and the number of abdominal constrictions was counted and

1490

CCB: A Novel Specific K Opioid Agonist

Vol. 57, No. 16,199.5

taken as an index of nociception. The ED, was defined as the dose that reduced the mean number of constrictions to 50% of the saline-treated mice. m.Thirty min after opioid agonist treatment mice were placed individually on an accelerated rotarod (Ugo Basile, Milan, Italy). The ED, value was defined as the dose that halved the saline-treated latency for falling off the rotarod. of urine Q&IL& Male rats (250-300 g) were starved overnight on the eve of an experiment and were water-loaded (25 mg/kg) 10 min before S.C. administration of the drug under test. Rats were placed in individual metabolism cages and urine was collected via a funnel into a graduated tube for 2 h. During this test period rats were deprived of food and water. The mean urine output (ml) of vehicle-treated rats was considered a 100% response and the changes of urine output observed in treated rats were reported as A% of this volume. vroral a8$ay. Male rats (200-250 g) were placed individually in plastic cages and allowed at least 1 h to acclimatize. CCB was administered, dissolved in saline containing 0.5% carboxymethylcellulose. Rating for stereotyped behavior and ataxia were taken at 10 min intervals and lasted for 120 min as described by Contreras et al. (20). Briefly, the stereotyped behavioral rating scale was: (0) inactive; (1) sniffing, grooming or rearing more frequently than control; (2) non-directional movements, occasional reciprocal forepaw treading, frequency of sniffing, rearing and grooming greater than (1); (3) turning or backpedding; (4) rapid and continuous turning, backpedding; (5) dyskinetic extension and flexion of limbs, head and neck. The ataxia rating scale was: (0) inactive or coordinate movements; (1) awkward or jerkly movements or loss of balance while rearing; (2) moderate rate of falling; (3) frequent falling or partial impairment of antigravity reflexes; (4) cannot move beyond a small area, may support weight on stomach or haunches; (5) unable to move except twitching movements.

Results Table I shows the binding profiles of the synthesized benzomorphans in comparison with selective ligands (U50,488H, (&)pentazocine, (-)SKF10,047) for the CL, 6, K and cl sites. In consideration of the paucity of data on o affinity of very selective and/or high affinity K binding agonists and to better evaluate the profile of synthesized compounds Ki of some BRL derivatives taken from literature or obtained from personal communication were also included in Table I. CCB, among the synthesized compounds, was found to have the higher K opioid receptor affinity and it was 2.57 times more potent than the prototype K agonist U50,488H. It was not possible to evaluate the l.r./rcor the S/K ratio for CCB, since it did not displace the binding of [3H]diprenorphine from rat brain membranes up to the doses of 25,000 nM. Noticeable is the profile of CCB with respect to the other benzomorphan derivatives, none of which significantly discriminate among opioid sites and only (-)SKFlO,O47 well discriminate between opioid and (T sites. The other synthesized derivatives displayed somewhat lower affinity for K receptors. As shown in Table I, none of the compounds in the series significantly binds to (+)-[3H]-pentazocine sites (K, > 1,000 nM).

CCB: A Novel Specific

Vol. 51, No. 16, 1995

K

1491

Opioid Agonist

In viva studies. CCB was effective in rising the nociceptive threshold in the mouse abdominal constriction test, being as potent as U50,488H and morphine (Table II). The antinociceptive effect was fully antagonized by naltrexone at a dose (5.0 mglkg,

TABLE I Binding Affinity to

K, p

, 6 and (TReceptors (K,+SEM, nM).

Compound

CCB

0.41+0.19

>25,000

>25,000

1,050&55

1

80.5k4.3

>25,000

>25,000

>l,OOO

2

388k16.8

>25,000

>25,000

>I ,000

3

>25,000

>25,000

>25,000

>1 ,000

MPCB

240f39

>25,000

>25,000

>l ,000

U50,488H

1.07kO.66

2,128fl60

nd.

>l ,000

BRL 525~37~,~

0.31 kO.05

1,559+146

>10,000(2)

28.6k6.4”

BRL 53001 qd

47.Ok4.5

6,790f639

>10,000(4)

1,213+365

BRL 52580a,b

0.20f0.02

30.2f6.4

113f5

>l,OOO(l)’

BRL 52762”,“,’

0.22f0.03

0.69f0.10

8.69&l .57

>l,OOO(l)

(-)SKF10,047

4.7g

3.0g

15g

1,22ah

(+)Pentazocine

87’

3gg

20.83’

2.83h

and 6 receptor binding assays were performed as described in (22). o receptor binding assay was performed according to M.Sbacchi and G.D.Clarke (23). T,he radioligand employed in the binding assays are as follows: for the K sates, [ HI-Bl$L 52537 423) (0.44nM), In the presence of haloperidol (100 nM); for the p srtesd [ HI-[D-Ala ,MePhe ,Gly-ol5]-Enkephalin (DAMGO) (1 nM); for 6 site, [3H]-[D-Ala ,D-Leu5]-Enkephalin (DADLE) (1 nM), in the presence of DAMGO (40nM); for o sites, [3H]-BRL 52537 (0.8 nM), in the presence of Naloxone (10 uM). Each value represents the mean from the concentration-response curves performed inb triplicate (n = 3 exdperiments) unless otherwise indicated in parentheses. See (22). ‘See (23). See (24) and (25). “See (26). ‘Unpublished &pj2provrded by SmrthK!rne Beecham Farmaceutrcr S.p.A., Milan (Italy); BRL 1-[(l -pyrrolldlnyl)methyl]-2-[(3,4-dlchlorophenyl)acetyl]-5-hydroxy1,2,3,4-tetyahydroiso-quinoline. ‘Data obtained from (4). hData obtained from (9). n.d. = not determined. ‘K, CL,

1492

CCB: A Novel Specific

K

Opioid Agonist

Vol. 57, No. 16, 1995

s.c.) similar to that antagonizing the effect of U50,488H and considerably higher than that needed to block the effect of the u agonist morphine (0.1 mg/kg, s.c.). The selective K opioid receptor antagonist nor-BNI (21) inhibited the antinociceptive effect of CCB and U50,488H, but not that of morphine (Table II). CCB, U50,488H and morphine all produced sedation (locomotor incapacitation) as noted by visual observation and determined by the ability of mice to maintain their position in the rotarod test (Table II). The order of potency was U50,488H 2 CCB > morphine. The sedative effects of U50,488H and CCB were significatively blocked in mice pretreated with naltrexone (5.0 mg/kg, s.c.) and nor-BNI while the lower dose of naltrexone (0.1 mglkg) was capable of blocking morphine-induced sedation. CCB, U50,488H and morphine were also tested for their effects on urine output in the rat. As shown in Table II, 0.3 mg/kg of CCB or of U50,488H produced a superimposable diuresis whereas morphine (2.0 mg/kg, s.c.) reduced the rate of urine production. Naltrexone (5.0 mg/kg, s.c.) or nor-BNI were effective in antagonizing the diuretic effect of CCB and U50,488H. Morphine-induced antidiuresis was blocked by naltrexone (0.1 and 5.0 mg/kg, s.c.) but not by nor-BNI.

TABLE II In viva Activity of CCB in Comparison with U50,488H and Morphine. Antagonism Naltrexone (NLT) or by Norbinaltorphimine (nor-BNI). Compound”

CCB CCB+NLT (0.1 mg/kg) CCB+NLT (5.0 mglkg) CCB+nor-BNI (10 mg/kg) U50,488H U50,488H+NLT (0.1 mg/kg) U50,488H+NLT (5.0 mg/kg) U50,488H+nor_BNI

Abdominal cramping test ED, and 95% confidence limit (mg/kg) 0.50 (0.24-0.86) 0.48 (0.26-0.98) b b 0.45 (0.21-0.93) 0.40 (0.18-0.71) b

Rotarod test

by

ED, and 95% confidence limit

Cumulative urine output A% vs vehicle-treated rats

7.53 (2.80-l 7.30) 6.34 (2.05-20.55) b b 6.42 (1.33-25.50) 6.50 (2.24-21.32) b

+44% at 0.3 mg/kg +48% at 0.3 mglkg +3% at 0.3 mgfkg b +52% at 0.3 mglkg +54% at 0.3 mg/kg +2% at 0.3 mg/kg

b +6% at 0.3 mglkg b (10 mg/kg) -60% at 2.0 mg/kg 14.38 (6.72-3188)’ 0.54 (0.30-0.70) Morphine b b -2% at 2.0 mglkg Morphine+NLT (0.1 mg/kg) +4% at 2.0 mglkg b b Morphine+NLT (5.0 mg/kg) Morphine+nor-BNI 0.57 (0.28-0.88) 16.72 (8.74-8.54)” -58% at 2.0 mg/kg I(10 mg/kg) “Test compounds were administered subcutaneously. The antagonist was administered 10 min before the agonist. In each experiment 5 animals per dose level were used. bFull antagonism of the pharmacological effect produced by the ED dose of the agonist. “The potency ratio of morphine verssus U50,488H or versus CC8 was significant (27).

Vol. 57, No. 16, 1995

CCB: A Novel Spexific

K

1493

Opioid Agonist

CC6 in doses up to 40 mg/kg S.C. did not produce any stereotyped ataxia in the behavioral assay (data not shown).

behavior

or

Discussion CCB is among the first non-peptide opioid agonists displaying both specificity and high affinity for K opioid receptors. CCB also showed a very low affinity for CI~sites. In fact, the most potent K agonists reported to date are not capable of discriminating equally well among different opioid receptor types and c binding sites. Some of them, although showing high affinity for the K opioid receptors, have a relatively lower selectivity with respect to other opioid receptors (i.e. U50,488H, BRL 52580 or BRL 52762) than CCB (28), and others, although selective with respect to u and 6 opioid receptors are not capable of discriminating between K and o sites (i.e. BRL 52537 or BRL 53001) (see references in Tab.1).

K agonists derived from normetazocine, such as (f)-pentazocine, low (T/K selectivity ratio, while (-)-SKF10,047 is not able to discriminate receptor types (Tab. I).

show a very among opioid

As Q receptors are very permissive with regard to the chemical structure of ligands, we hypothesize that only the contemporary presence of normetazocine in the (-)-form supporting the H-bonding group and the 4-cloro phenyl ring simoultaneously present on the nitrogen substituent might be responsible for the unique binding profile of CCB. Binding data of this series of compounds indicate that 4-chloro substitution is responsible of the high (nanomolar) affinity with respect to K sites, while 3-chloro substitution is detrimental. The 3,4-dichloro, 2-methoxy and the unsubstituted derivatives, although specific for K receptors, show a 200, 900 and 600 times lower affinity, respectively. The novel compound described in this paper displayed the typical in viva profile of K opioid receptor agonists. In fact, CCB and the reference agonist U50,488H were both effective in increasing the nociceptive threshold to a chemical stimulus in our experiments. They were also depressant in the mouse and both induced diuresis in the rat. Such diuretic effects are well known for K agonists (29) and are thought to be due primarily to the presence of K opioid receptors in the posterior pituitary gland (30). On the contrary, in agreement with reported data, morphine produced only a decrease in urine output in the water-loaded rat; an effect antagonized by a low dose of naltrexone. Moreover, Contreras et al. (20) have reported that several benzomorphan derivatives such as (-)-cyclazocine and SKF10,047, but not ethylketocyclazocine, display stereotyped behavior in rat and ataxia interacting with binding sites other than opioid receptors. On the contrary, CCB did not produce any ataxia or stereotyped behavior in rodents; therefore, it is unlikely that this compound may bind to ts or phencyclidine sites to cause psycotomimetic effects. These findings suggest that CCB might be a new useful tool to investigate physiological role of K opioid receptors.

the

1494

CCB: A Novel Specific I Opioid Agonist

Vol. 57, No. 16, 1995

Acknowledaments This study was financially supported by M.U.R.S.T. (Italy). We acknowledge the provision of radioligand binding data on BRL 52580 and BRL 52762 by Drs.M.Sbacchi and G.D.Clarke (SmithKline Beecham Farrnaceutici S.p.A., Milan, Italy). We gratefully acknowledge Fabbrica ltaliana Sintetici, Italy, for providing (-)-normetazocine.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

M.WOLLEMANN, S.BENYHE AND JSIMON, Life Sci. 2 599-611 (1993). J.M. MUSACCHIO, Neuropsychopharmacol. 3 191-200 (1990). S.L.TODD, R.L.BALSTER and B.R.MARTIN, Life Sci. & 895-901 (1990). S.W.TAM, Eur.J.Pharmacol. j_Q$f33-41 (1985). J.MAGNAN, S.J.PATERSON, A.TAVANI and H.W.KOSTERLITZ, Arch.Pharmacol. m 197-205 (1982). J.A. WEYHENMEYER and K.J. MACK, Neuropharmacol. 24 111-l 15 (1985). B.L.LARGENT, H.WfKSTROM, A.L.GUNDLACK and S.H.SNYDER, Molecular Pharmacol. 2 772-784 (1987). D.L.DeHAVEN-HUDKINS and L.C.FLEISSNER, Life Sci. a PL65-PL70 (1992). A. CAGNOTTO, A.BASTONE and T.MENNINI, Eur.J.Pharmacol. - Molecular Pharmacol. Section m 131-l 38 (1994). D.L.DeHAVEN-HUDKINS, L.C. FLEISSNER and F.Y. FORD-RICE, Eur.J.Pharmacol. 222 371-378 (1992). R.QUIRION, W.D.BOWEN, Y.HZHAK, J.L.JUNIEN, J.M.MUSACCHIO, R.B.ROTHMAN, T.-S.SU, S.W.TAM and D.P.TAYLOR, TIPS a 85-86 (1992). G. RONSISVALLE, L. PASQUINUCCI, M.S. PAPPALARDO, F. VITTORIO, G. FRONZA, C. ROMAGNOLI, E. PISTACCHIO, S. SPAMPINATO and S. FERRI, J.Med.Chem. s 1860-1865 (1993). G. RONSISVALLE, 0. PREZZAVENTO, L. PASQUINUCCI, L. CARBONI, E. PISTACCHIO and S. SPAMPINATO, Reg.Peptides, SUDDI.1 S31-S32 (1994). M.G.C. GILLAN, H.W. KOSTERLITZ and S.J. PATERSON, Br.J.Parmacol. ZQ 481-90 (1980). J.A. WEYHENMEYER and K.J. MACK, Neuropharmacol. 24 11 l-1 15 (1985). W.D. BOWEN, S.B. HELLEWELL, M. KELEMEN, R. HUEY and D. STEWART, J.Biol. Chem. m 13434-13439 (1987). G.L.PETERSON, Anal.Biochem. &8 346-356 (1977). G.A. MCPHERSON, J. Pharmacol.Methods, 14 213-228 (1985). A.G.HAYES, M.J.SHEEHAN and M.B.TYERS, Br.J.Pharmac. 91823-832 (1987). P.C. CONTRERAS, K.C. RICE, A.E. JACOBSON, T.L. O’DONOHUE, Eur.J.Pharmacol. j_Zl9-18 (1986). A.E.TAKEMORI, M.M.SCHWARTZ and P.S.PORTOGHESE, J.Pharmacol.Exp.Ther. 242 971-974 (1988). V.VECCHIETTI, G.D.CLARKE, R.COLLE, G.GIARDINA, G.PETRONE and M.SBACCHI, J.Med.Chem. a 2624-2633 (1991). M.SBACCHI and G.D.CLARKE, Excerpta Medica 914 211-212 (1990). P.ZARATIN, G.PETRONE, A.PIZZI, M.SBACCHI and G.D.CLARKE, 269-270 (1992). Pharmacol.Res. m

Vol. 57, No. 16, 1995

25. 26. 27. 28. 29. 30.

CCB: A Novel Specific K Opioid Agonist

1495

G.GIARDINA, G.D.CLARKE, G.DONDIO, G.PETRONE, MSBACCHI and V.VECCHIElTI, J.Med.Chem. Z 3482-3491 (1994). V.VECCHIETTI, G.D.CLARKE, R.COLLE, G.DONDIO, G.GIARDINA G.PETRONE and M.SBACCHI, J.Med.Chem. 3 2970-2978 (1992). . . R.J.TALLARIDA and R.B.MURRAY, &nual of Pharm%olocjcal Calmth Corn-r Prom, Springer-Verlag, New York (1987). P.F.VONVOIGTLANDER, R.A.LAHTl and J.H.LUDENS, J.Pharmacol.Exp.Ther. 224 7-12 (1983). J.D.LAENDER, J.Pharmacol.Exp.Ther. 224 89-94 (1983). D.P.BROOKS, G.GIARDINA, M.GELLAI, G.DONDIO, R.M.EDWARDS, G.PETRONE, P.D.DePALMA,M.SBACCHI, M.JUGUS, P.MISIANO, Y.-X.WANG and G.D.CLARKE, J.Pharmacol.Exp.Ther. 2& 164-171 (1993).