Tolerance to N6-(l-phenylisopropyl ) adenosine Contribution of behavioral mechanisms and cross-tolerance profile

Neuropharmacology Vol. 23, No. 6, pp. 671-676, Printed in Great Britain. All rights reserved

TOLERANCE

1984 Copyright 0

TO

002%3908/84 $3.00 + 0.00 1984 Pergamon Press Ltd

N6-(L-PHENYLISOPROPYL)

ADENOSINE CONTRIBUTION OF BEHAVIORAL MECHANISMS CROSS-TOLERANCE PROFILE

AND

D. G. SPENCER JR*, P. CALDWELL and M. W. EMMETT-OGLESBY Department of Pharmacology, Texas College of Osteopathic Medicine, Fort Worth, TX 76107, U.S.A. (Accepted

3 September

1983)

Summary-The contribution of behavioral mechanisms to tolerance to N6-(L-phenylisopropyI)adenosine (L-PIA) was studied, along with the degree of cross-tolerance to other drugs active in the CNS. Rats were stabilized on a fixed-ratio of a 20 lever-pressing schedule for food reward and were then assigned to three daily-treatment groups. One group (saline-behavior associated) was injected with saline 15 min before the session, another (L-PIA-behavior associated) was injected with L-PIA (0.08 mg/kg) 15 min before the session and the last (L-PIA-behavior dissociated) was injected with L-PIA (0.08 mg/kg) immediately after the session. Tolerance developed to the decreasing effects of L-PIA on response rate in both groups, L-PIA-behavior associated and L-PIA-behavior dissociated. Behavioral mechanisms were thus not important in tolerance to L-PIA. In subsequent cross-tolerance tests, L-PIA-tolerant rats were crossthe adenosine Al receptor agonist, N6-cyclohexyladenosine. The drugs tolerant to 4,5,6,7-tetrahydroisoxazolo[5,4-clpyridin-3-01 (THIP), diazepam, pentobarbital, ketamine, clonidine, d-amphetamine and caffeine did not produce differential effects in L-PIA-tolerant and non-tolerant subjects; however, L-PIA-tolerant subjects were more sensitive to the suppressive effects of chlorpromazine on the response-rate. Key words: N6+phenylisopropyl)adenosine,

N6-cyclohexyl adenosine, PIA, tolerance, behavior, rats.

Analogs of adenosine have been developed that are more metabolically stable than the parent compound and that have a greater affinity for the Al adenosine receptor that inhibits the activity of adenylate cyclase (Londos and Wolff, 1977; Daly, Bnms and Snyder, 1981). Exogenous application of adenosine or Al agonist adenosine analogs is known to produce hypotension (Maitre, Ciesielski, Lehmann, Kempf and Mandel, 1974), sedation (Snyder, Katims, Annau, Bruns and Daly, 1981) and anticonvulsant effects (Dunwiddie and Worth, 1982). Levels of endogenous adenosine rise in sleep-deprived rats and intraventricularly-administered adenosine induces behavioral and electrophysiological signs of sleep in dogs (Haulica, Ababei, Branisteanu and Topoliceanu, 1973). Since it has also been found that one pool of adenosine in synaptosomes of rat cortex undergoes calcium-dependent rapid uptake and release (Bender, Wu and Phillis, 1981) adenosine has been proposed to have a neurotransmitter or neuromodulator function, in addition to its metabolic role. The behavioral effects of adenosine and its analogs have consequently become of interest. Systemic administration of Al agonists such as N6-(L-phenylisopropyl)adenosine (L-PIA) and

N6-cyclohexyladenosine (CHA) decrease foodrewarded response-rate on such schedules of reinforcement as differential reinforcement of low rate (Carney and Coffin-Sirochman, 1981) and fixed ratio (Sirochman and Carney, 1981; Spencer and Lal, 1983). The development of tolerance to the decreasing effects of L-PIA on response-rate was reported in these studies, but two questions remain. First, tolerance to the effects of some sedative-hypnotic drugs, such as ethanol (Mansfield, Benedict and Woods, 1983) and pentobarbital (Branch, 1983), involves behavioral mechanisms as well as pharmacokinetic and pharmacodynamic changes. That is, a greater degree of tolerance to these drugs can occur if the drug-intoxicated subject is allowed to practice the behavior in which tolerance is being assessed. Since L-PIA also has sedative-hypnotic properties, the present study examined the contribution of behavioral mechanisms of tolerance to L-PIA. Second, regardless of the exact mechanism of tolerance to L-PIA, the relative specificity of tolerance to L-PIA was assessed in cross-tolerance tests with respresentatives of other psychoactive classes of drug. METHODS

*Address reprint requests to: D. G. Spencer, Jr, Neurobiology Department, Troponwerke, Neurather Ring 1 220-232, 5000 Cologne 80, FRG.

Forty male Long-Evans rats were maintained 320 g weight by deprivation of food and served 671

at as

612

D. G.

SPENCER JR et al.

subjects. Rats were trained in conventional operant chambers to press a single lever for a reward of food on a fixed ratio of 20 schedule (FR2wvery twentieth lever press was reinforced). Sessions were controlled and data recorded with a TRS-80 microcomputer, using a modification of the program described by Emmett-Oglesby, Spencer and Arnoult (1982). Twenty-minute operant sessions were conducted 7 days a week until each rate of response of each subject stabilized. Criteria of stability were at least 60 operant sessions and three consecutive sessions in which the response-rate neither increased nor decreased consistently, and during which the responserate did not vary more than 10% from the mean. At this time, a dose-response relationship for the suppressive effects of L-PIA on the response-rate was determined. Subjects were first stabilized to an injection of saline (1 ml/kg, i.p.), given 15 min before operant sessions. Thereafter, L-PIA (0.02, 0.0,4 or 0.08 mg/kg, i.p.) was injected 15min before every fourth session. Each subject received all doses. Once the initial dose-response determination to L-PIA had been completed, subjects were assigned to three chronic treatment groups of 12 to 13 rats each: L-PIA-behavior associated, L-PIA-behavior dissociated and saline-behavior associated. Fifteen minutes before each daily session, the L-PIA-behavior L-PIA injected with associated group was (0.08 mg/kg) and the saline-behavior associated group was injected with saline. The L-PIA-behavior dissociated group was injected with L-PIA (0.08 mg/kg) immediately after each session. After the subjects in the L-PIA-behavior associated group had reached an asymptotic level of tolerance to the suppressive effects on the response-rate of the pre-session injections of L-PIA, the dose-response function of L-PIA was re-determined in all three groups. Chronic treatment conditions continued during sessions intervening between L-PIA-dose tests. After re-determination of the L-PIA dose-response relationship, the L-PIA-behavior dissociated group was discontinued and cross-tolerance tests began in the L-PIA-behavior associated and the salinebehavior associated groups. Every fourth session was preceded by an intraperitoneal injection of a test drug, with chronic treatment conditions (PIA, O.O8mg/kg, or saline) continuing on intervening sessions. Each subject in both the L-PIA-behavior associated and the saline-behavior associated groups received all doses of all the test drugs. The order in which the different test doses of the drug were administered was randomized for each subject using numbers table. The test drugs a random 4,5,6,7-tetrahydroisoxazolo[5,4-clpyridin-3-01 (THIP -5, 10 and 20mg/kg), pentobarbital (5, 10 and 20mg/kg), ketamine (5, 10 and 20mg/kg), chlorpromazine (1.25, 2.5 and 5 mg/kg), d-amphetamine (1.25, 2.5, 5 and 10 mg/kg), N6-cyclohexyladenosine (CHA-0.04, 0.08 and 0.16 mg/kg) and caffeine (10, 20, 40 and 80mg/kg) were all injected intra-

peritoneally 15 min before each session. Clonidine (0.01, 0.02 and O.O4mg/kg) was injected 60min before the session and diazepam (5, 10 and 20mg/kg) was injected 30 min before the session. The effects of treatment on response-rate were calculated for each subject as a percentage of the response-rate of that subject following an injection of saline. The responserate to saline for each rat was taken as the mean response rate of three sessions with saline pretreatment. All effects of drugs were analyzed by repeated measures analysis of variance. RESULTS The drug L-PIA produced a dose-dependent reduction in response-rate, with an ED,, of 0.045 mg/kg, as calculated by probit analysis. Prior to the onset of chronic treatment conditions, the dose-effect functions of L-PIA on the saline-behavior associated, the L-PIA-behavior associated and the L-PIA-behavior dissociated groups did not differ significantly from each other. Asymptotic levels of tolerance to the decreasing effects on the response-rate of L-PIA (0.08 mg/kg) administered prior to each session (the L-PIA-behavior associated group) were reached Re-determination of the within 20 sessions. dose-effect curve of L-PIA for the suppression of response-rate at this time showed that the three significantly groups differed [F(2,33) = 9.05, P < 0.001; Fig. 11. Whether L-PIA was administered before (L-PIA-behavior associated) or after (L-PIA-behavior dissociated) daily sessions during chronic treatment conditions, the re-determined dose-effect curves for these two groups were significantly different from the chronic saline-injected group [F(1,22) = 24.2, P < 0.001; F(1,22) = 7.35, P < 0.0251, but not from each other. A comparison H o--J

100

T

PRE-TREATMENT. ED50 = 0 045 SALINE-BEHAVIOR ASSOCIATED. ED50 = 0 070

-

PIA-BEHAVIOR ASSOCIATED. ED60 = 0 155

P-v,

PIA-BEHAVIOR DISSOCIATED. ED50=0116

T

PIA

DOSE

(mglkg)

Fig. 1. Effects of acute administration of L-PIA (doses represented on log scale) on a fixed ratio 20 lever-pressing rate. Rates are expressed as a percentage of the response rate of each group following injection of saline. Data are based on the 13, 12 and 12 subjects in groups L-PIA-behavior associated, L-PIA-behavior dissociated and saline-behavior associated, respectively.

N6-(L-phenylisopropyl)adenosine

613

tolerance

DIAZEPAM

PENTOBAREITAL

KETAMINE

CLONIDINE 100

60

‘.%,

60

\. -... 60 40

60 -. T 40

20

20

100 11 5

10

20

CHLORP~~OMAZINE

125

25

d-AMPHETAMINE

50

125

CHA

25

50

10

CAFFEINE

60 60

'\i

JO

I-.,

20

r-4

100 I,, 004

I 006

I 016

I,!

10

I

I

20

40

do

DOSE (mg/kg) Fig. 2. Effects of acute administration of various drugs (doses shown on log scale) on the lever-pressing rate in the L-PIA-behavior associated group and the saline-behavior associated group. All data are based on response to drug of 13 subjects in the L-PIA-behavior associated group and 12 subjects in the

saline-behavior associated group, with the exception of data for response to ketamine, for which only 6 subjects from each group were used. of the ED,,+ in the saline-behavior associated and the L-PIA-behavior associated groups resulting from probit analysis revealed that chronic pre-session treatment with L-PIA resulted in a 2.2-fold rightward shift in the dose-response function of L-PIA. Tolerance to L-PIA conferred tolerance to N6-cyclohexyladenosine [F( 1,22) = 13.4, P < 0.01, Fig. 21. However, no cross-tolerance was observed to the rate-suppressing effects of THIP, diazepam, pentobarbital, ketamine, clonidine, d-amphetamine and caffeine (Fig. 2). Tolerance to L-PIA was accompanied by enhanced sensitivity to the ratesuppressant effects of chlorpromazine [F( 1,20) = 7.37, P < 0.0251. DISCUSSION chronically L-PIA Both groups given (L-PIA-behavior associated and L-PIA-behavior dis-

sociated) developed tolerance to the rate-suppressant effects of the drug. Since the degree of tolerance in these groups was not significantly different, behavioral mechanisms do not appear to contribute markedly to the tolerance observed to L-PIA in a food-reinforced, fixed ratio task. That is, the opportunity to practice the lever-pressing task while under the influence of L-PIA was not a critical factor in determining the development of tolerance to this drug. It is true, however, that at all doses of L-PIA, the L-PIA-behavior associated group exceeded the L-PIA-behavior dissociated group in the percentage of the response-rate to saline. It is thus possible that some behavioral influence exists. Conversely, it has typically been the case that when behavioral tolerance to a drug is found, it is the only contributing factor; subjects receiving the drug after the behavioral session are no more tolerant than those treated with

614

D. Cl. SPENCERJR et al.

saline (Branch, 1983; Demellweek and Goudie, 1983; Emmett-Oglesby and Taylor, 1981). Cross-tolerance revealed that only tests N6-cyclohexyladenosine, another adenosine Al receptor agonist, had less effect in L-PIA-tolerant rats than in the saline-behavior associated control group. Furthermore, N6cyclohexyladenosine affected behavior in the two groups at doses of 0.04, 0.08 and 0.16 mg/kg to about the same extent as L-PIA itself, agreeing with previous reports of the equal behavioral potency of the two compounds (Sirochman and Carney, 1981). Although several investigators have reported interactions between adenosine, GABA and benzodiazepines in vitro (Phillis, Bender and Wu, 1980; Skolnick, Paul and Marangos, 1980; Ticku and Birch, 1980; Marangos, Martino, Paul and Skolnick, 1981; Phillis, Wu and Bender, 1981; Williams, Risley and Huff, 1981), in vivo, inosine and Al agonists do not share the anxiolytic effects of diazepam and pentobarbital (Gherezghiher, Spencer, Elmesallamy and Lal, 1982). In the present study, no differential effects were seen in the L-PIA-behavior associated and the saline-behavior associated groups to THIP (a GABA agonist), diazepam or pentobarbital. This finding is of additional interest because all three drugs, as well as ketamine, are very effective sedative-hypnotics, producing similar gross behavioral signs to L-PIA. Lack of cross-tolerance between these compounds and L-PIA suggests that tolerance to L-PIA is not due to interactions at the GABA-benzodiazepine-ionophore receptor complex or to a general adaptation to the sedative properties of the drug. The cross-tolerance data for N6-cyclohexyladenosine and the various sedative-hypnotics are compatable with either a pharmacokinetic or a pharmacodynamic basis for tolerance to L-PIA. That is, cross-tolerance was only observed with another adenosine analog. Thus, tolerance to L-PIA may be the result of increased specific metabolism of adenosine analogs. Alternatively, tolerance may be the result of a pharmacodynamic adaptation to drugs that interact with adenosine receptors. Although neither mechanism was tested directly in the present study, the observation that tolerance to L-PIA resulted in sensitization to the effects of chlorpromazine is most readily interpreted as the result of a pharmacodynamic adaptation to L-PIA that modified the normal response to chlorpromazine. Stone and Taylor (1978a, b, c, 1979, 1980) have shown that in the cerebral cortex of the rat, adenosine and norepinephrine have similar and additive inhibitory effects on electrophysiological activity. More recent studies have shown that adenosine may also act presynaptically to inhibit the release of norepinephrine (Taylor and Stone, 1980). In addition, Watanabe, Lai and Yoshida (1983) found that incubation with either adenosine or 2-chloroadenosine (an adenosine Al receptor agonist) selectively in-

creased noradrenergic a2 receptor sites (as measured by binding of tritiated clonidine) in the rat vas deferens, indicating a direct interaction between u2 and Al receptors. However, in the present study, the noradrenergic t(2 agonist clonidine had no differential effect in L-PIA-tolerant and non-tolerant rats. Further, d-amphetamine, which indirectly releases norepinephrine as well as dopamine (Scheel-Kruger, 1972), also had no differential effect on the t-PIA-behavior associated and the saline-behavior associated groups. These data suggest that tolerance can occur to the behavioral suppressant effects of L-PIA without modifying the sensitivity of catecholaminergic receptors. Phenothiazines have been reported to inhibit the uptake of adenosine in synaptosomes in rat brain (Phillis, Wu and Bender, 1980). Adenosine also strongly modulates the function of dopamine in the striatum (Green, Proudfit and Yeung, 1982). In addition, Geyer, Wilker, Spaulding, Cornfeldt, Brugger, Huger and Novick (1982) have shown that the Al agonist N6-cyclohexyladenosine has behavioral effects indicative of antipsychotic action in animals. It would, therefore, be expected that cross-tolerance might exist between L-PIA and the phenothiazine, However, chlorpromazine was chlorpromazine. found to reduce the response-rate more in the L-PIA-behavior associated group than in the salinebehavior associated group; sub-acute treatment with L-PIA sensitized the L-PIA-behavior associated group to the disruptive effects of chlorpromazine. It is unknown, though, whether this unexpected sensitization was due to blockade of dopamine receptors, or some other property of phenothiazines. Methylxanthines, such as caffeine and theophylline, have been characterized as Al receptor antagonists (Daly et al., 1981; Synder et al., 1981). Indeed, caffeine at 0.32 and 3.2 mg/kg antagonizes the L-PIA-induced decrease in the rate of leverpressing (Sirochman and Carney, 1981) and at 2.5 and 5.0 mg/kg, blocks the L-PIA-stimulus in animals trained to discriminate between L-PIA (0.08 mg/kg) and saline (Spencer and Lal, 1983). Glowa and Spealman (1983) showed that although caffeine produced opposite effects on the response-rate depending on the reinforcement schedule, doses of caffeine as small as 1.0 mg/kg (having little behavioral effect of their own) consistently antagonized the uniformly rate-decreasing effects of L-PIA on all reinforcement schedules. However, L-PIA did not antagonize either the rate-increasing or rate-decreasing effects of caffeine. Holloway and Carney (personal communication) also found that L-PIA was ineffective in blocking the stimulus of caffeine in rats trained to discriminate between caffeine (32 mg/kg) and saline. Therefore, it would seem that while caffeine is an effective Al antagonist at systemic doses of at least 0.32 mg/kg, there is no evidence that blockade of Al receptors mediates the acute behavioral effects of caffeine.

N6-(L-phenylisopropyI)adenosine

In the present study, no sensitization was seen to the rate-suppressant effects of substantial doses of caffeine (40 and 80mg/kg) in L-PIA-tolerant rats. However, in a previous study of tolerance to L-PIA, Nakamura, Carney and Christensen (1982) found that rats given 1.0 mg/kg of L-PIA chronically were selectively sensitized to the increasing effects of small doses of caffeine (e.g. 3.2mg/kg) on the responserate. Although Glowa and Spealman (1983) found the rate-increasing effects of caffeine were resistant to blockade by L-PIA, it is conceivable that Al blockade could contribute to, or indirectly promote this effect of caffeine. Thus, a decrease in function of Al receptors produced by chronic administration of L-PIA could selectively magnify the contribution of Al blockade to the group of neuropharmacological effects underlying rate-increasing effects of caffeine at small doses. Since the large doses of caffeine used in the present study to suppress the response-rate are even less specific for blockade of Al receptors, the hypothesis that tolerance to L-PIA is based on a pharmacodynamic modification in A 1 receptors remains the most likely possibility. However, much remains that is unclear about interactions between adenosine analogs and caffeine, especially as regards states of tolerance. Acknowledgements-We thank Dr Harbans La1 and Dr John Lane for their conceptual and editorial contributions. This study was partially supported by grant number 82-l I-045 from the American Osteopathic Association.

REFERENCES Bender A. S., Wu P. H. and Phillis J. W. (1981) The rapid uptake and release of ‘H adenosine by rat cerebral cortical synaptosomes. J. Neurochem. 36: 651-660. Branch M. N. (1983) Behavioral tolerance to stimulating effects of pentobarbital: a within-subject determination. Pharmac. Biochem. Behab. 18: 25-30. Carney J. M. and Coffin-Sirochman V. L. (1981) Methylxanthines antagonize the effects of phenylisopropyl adenosine on operant behavior. The Pharmacologist 23: 151. Daly J. W., Bruns R. F. and Snyder S. H. (1981) Adenosine receptors in the central nervous system: relationship to the central actions of methylxanthines Life Sci. 28: 2083-2097. Demellweek C. and Goudie A. J. (1984) An analysis of behavioral mechanisms involved in the acquisition of amphetamine anorectic tolerance. Psychopharmacology (In press). Dunwiddie T. V. and Worth T. (1982) Sedative and anticonvulsant effects of adenosine analogs in mouse and rat. J. Pharmac. exp. Ther. 220: J&76. Emmett-Oglesby M. W., Spencer D. G., Jr and Arnoult D. (1982) A TRS-80-based system for the control of behavioral experiments. Pharmac. Biochem. Behav. 17: 583-587. Emmett-Oglesby M. W. and Taylor K. E. (1981) Role of dose-interval in the acquisition of tolerance to methylphenidate. Neuropharmacology 20: 995-1002. Geyer H. M., Wilker J., Spaulding T., Cornfeldt M., Brugger S., Huger F. and Novick W. (1982) N6-cyclohexyladenosine (CHA): Some in uiuo pharmacology. Fedn Proc. Fedn Am. Sots exp. Biol. 41: 10781. Gherezghiher T., Spencer D. G., Jr, Elmesallamy F. and Lal H. (1982) Failure of adenosine analogs to produce

tolerance

675

benzodiazepine-like specific behavioral effects. Sot. Neurosci. Abs. 8: 121. Glowa J. R. and Spealman R. D. (1983) Behavioral effects of caffeine, (-)-N6-((R)-l-methyl-2-phenylethyl)adenosine (PIA) and their combination. Fedn Proc. Fedn Am. Sots exp. Biol. 42: 1362. Green R. D., Proudfit H. K. and Yeung S. H. (1982) Modulations of striatal dopaminergic function by local injection of 5-N-ethylcarboxamide adenosine. Science 218: 58-60. Haulica I., Ababei L., Branisteanu D. and Topoliceanu F. (1973) Preliminary data on the possible hypnogenic role of adenosine. J. Neurochem. 21: 1019-1020. Londos C. and Wolff J. (1977) Two distinct adenosinesensitive sites on adenylate cyclase. Proc. natn. Acad. Sci U.S.A. 74: 5482-5486. Maitre M., Ciesielski L., Lehmann A., Kempf E. and Mandell P. (1974) Protective effect of adenosine and nicotinamide against audiogenic seizures. Biochem. Pharmat. 23: 2807-2816. Mansfield J. G., Benedict R. S. and Woods S. C. (1983) Response specificity of behaviorally augmented tolerance to ethanol supports a learning interpretation, Psychopharmacology 79: 94-98.

Marangos P. J., Martin0 A. M., Paul S. M. and Skolnick P. (1981) The benzodiazepines and inosine antagonize caffeine-induced seizures. Psychopharmacology 72: 269-273. Nakamura M., Carney J. M. and Christensen H. D. (1982) An animal model for CNS caffeine intolerance. Sot. Neurosci. Abs. 8: 5709. Phillis J. W., Bender A. S. and Wu P. H. (1980) Benzodiazepines inhibit adenosine uptake into rat brain synaptosomes. Brain Res. 195: 494498. Phillis J. W., Wu P. H. and Bender A. S. (1981) Inhibition of adenosine uptake into rat brain synaptosomes by the benzodiazepines. Gen. Pharmac. 12: 67-70. Scheel-Kruger J. (1972) Behavioral and biochemical comparison of amphetamine derivatives, cocaine, benztropine and tricyclic antidepressant drugs. Eur. J. Pharmac. 18: 63-73. Sirochman V. and Carney J. M. (1981) Behavioral pharmacology of adenosine analogs. Fedn Proc. Fedn Am. Sots exp. Biol. 40: 294. Skolnick P., Paul S. M. and Marangos P. J. (1980) Purines as endogenous ligands of the benzodiazepine receptor. Fedn Proc. Fedn Am. Sots exp. Biol. 39: 3050-3055. Snyder S. H., Katims J. J., Annau Z., Bruns R. F. and Daly J. W. (1981) Adenosine receptors and behavioral actions of methylxanthines. Proc. natn. Acad. Sei. U.S.A. 78: 326&3264. Spencer D. G., Jr and La1 H. (1983) Discrimination stimulus properties of L-phenylisopropyl adenosine: Blockade by caffeine and generalization of 2-chloroadenosine. Life Sci. 32: 2329-2333. Stone T. W. and Taylor D. A. (1978a) An electrophysiological demonstration of a synergistic interaction between norepinephrine and adenosine in the cerebral cortex. Brain Res. 147: 396400. Stone T. W. and Taylor D. A. (1978b) Antagonism by clonidine of neuronal depressent responses to adenosine, adenosine-5’-monophosphate and adenosine triphosphate. Br. J. Pharmac. 64: 369-374. Stone T. W. and Taylor D. A. (1978~) Clonidine as an adenosine antagonist. J. Pharm. Pharmac. 30: 792-793. Stone T. W. and Taylor D. A. (1979) Antidepressent drugs potentiate suppression by adenosine of neuronal firing in rat cerebral cortex. Neurosci. Lett. 11: 93-97. Stone T. W. and Taylor D. A. (1980) Effects of adenosine and related compounds on an inhibitory process in rat cerebral cortex. Expl Neurol. 70: 556566. Taylor D. A. and Stone T. W. (1980) The action of adenosine on noradrenergic neuronal inhibition induced

676

D. G. SPENCERJR et al.

by stimulation of locus coeruleus. Brain Res. 183: 367-376. Ticku M. K. and Burch T. (1980) Purine inhibition of ‘H-y-aminobutyric acid receptor binding to rat brain membranes. Biochem. Pharmac. 29: 1217-1220. Watanabe Y., Lai R. and Yoshida H. (1983) Increase of )H

clonidine binding sites induced by adenosine receptor agonists in rat vas deferens in vitro. Eur. J. Pharmac. 86: 265-269.

Williams M., Risley E. A. and Huff J. R. (1981) Interaction of putative anxiolytic agents with central adenosine receptars. Can. J. Physiol. Pharmac. 59: 897-900.