- Email: [email protected]
Effects of cocaine alone and in combination with haloperidol and SCH 23390 on cardiovascular function in squirrel monkeys
Life Sciences, Vol. 48, pp. 1547-1554 Printed in the U.S.A.
Pergamon Press
EFFECTS OF COCAINE AIZ)NE AND IN COMBINATION WITH HALOPERIDOL AND SCH 23390 ON CARDIOVASCULAR FUNCTION IN SQUIRREL MONKEYS C. W. Schindlerl, 2, S. R. Tellal, 3, J. M. Witldn 1, and S. R. Goldbergl, 2 1preclinical Pharmacology Branch, NIDA Addiction Research Center PO Box 5180, Baltimore, MD 21224 2Department of Pharmacology and Experimental Therapeutics University of Maryland, Baltimore, MD 21201 and 3Department of Pharmacology, Georgetown University School of Medicine 3900 Reservoir Road, Washington, DC 20007 (Received in final form February 12, 1991)
SUMMARY The potential involvement of D 1 and D2 dopamine receptors in the effects of cocaine on cardiovascular function in squirrel monkeys was evaluated. A low dose of cocaine (0.1 mg/kg i.v.) produced increases in both blood pressure and heart rate. At the higher doses of cocaine (1.0-3.0 mg/kg) the heart rate response was biphasic, consisting of an early decrease followed by an increase in heart rate 10-20 min following injection. The dopamine D2 antagonist haloperidol (0.01 mg/kg i.m.) attenuated the heart rate increasing effect of cocaine, but doses as high as 0.03 mg/kg did not alter the blood pressure increase. The D1 antagonist SCH 23390 (0.01-0.03 mg/kg i.m.) did not attenuate either the blood pressure or heart rate increasing effects of cocaine. The D2 agonist quinpirole (1.0 mg/kg i.v.) produced increases in heart rate similar to cocaine, with little effect on blood pressure. Although effective against the heart rate increasing effect of cocaine, haloperidol (0.01 mg/kg) did not antagonize the heart rate increasing effects of quinpirole. The D1 agonist SKF 38393 (3.0 mg/kg i.v.) decreased heart rate and increased blood pressure. The blood pressure increasing effect of SKF 38393 was antagonized by 0.01 mg/kg SCH 23390. Haloperidol's ability to partially antagonize the tachycardiac response to cocaine suggests the involvement of D2 receptors in that response. However, the failure of haloperidol to antagonize quinpirole's tachycardiac effect suggests that non-dopaminergic mechanisms may also be involved in haloperidors antagonism of cocaine's tachycardiac effect. The pressor effects of cocaine do not appear to be controlled by selective dopamine receptors. Continued high levels of cocaine abuse with accompanying behavioral and physiological toxicity has increased efforts to identify the mechanisms associated with that toxicity. However, as cocaine has diverse pharmacological actions, a number of factors may be relevant to the toxicity associated with this compound. For example, local anesthetic actions and the involvement of both catecholamine and indolamine neurotransmission have all been implicated in the effects of cocaine (1). One of the major actions of cocaine is blockade of the reuptake of dopamine with an accompanying increase in the synaptic availability of this neurotransmitter (2). The action of cocaine at dopamine synapses may be further influenced by the existence of multiple dopamine 0024-3205/91 $3.00 +.00
1548
Cocaine and Dopamlne
Vol. 48, No. 16, 1991
receptor subtypes (3). While the involvement of dopamine in the behavioral effects of cocaine has been studied extensively (4), there has only been limited study of the involvement of dopamine receptor subtypes in the toxic effects of cocaine. Witkin et al. (5,6) demonstrated in rodents that the lethal effects of cocaine could be attenuated by pretreatment with SCH 23390 and related D1 antagonists, but not by the inactive enantiomer of SCH 23390. The D2 antagonist haloperidol was inactive against the lethal effects of cocaine, but conferred protection against methamphetamineinduced lethality. The cardiovascular system has often been suggested as a primary target for cocaine toxicity. While the mechanisms by which cocaine produces its cardiovascular effects are not completely understood, adrenergic mechanisms are certainly involved (7). Nevertheless, the potential for dopaminergic systems to be involved in cardiovascular functioning is also evident. It has recently been established that both peripheral and central dopaminergic systems can be involved in cardiovascular regulation (8-10). In hypertensive patients (11) and experimental hypertensive rats (12) central dopaminergic activity has been shown to be altered as reflected in elevation of plasma prolactin levels. Dopaminergic receptors have also been identified in the periphery (13) and action at those receptors can modify cardiovascular function (14). Further, two distinct receptor subtypes for dopamine have been identified in the periphery. These two receptor subtypes appear to be pharmacologically similar to those described for the central nervous system (13). Thus, an action of cocaine at dopamine synapses may be expected to contribute to its cardiovascular effects. In support of this notion, a recent study by Sherer et al. (15) showed that haloperidol could partially attenuate the cardiovascular effects of cocaine in humans. Therefore, the purpose of the present study was to determine whether selective dopamine antagonists could modify cardiovascular effects of cocaine. Both the D2 antagonist haloperidol and the D1 antagonist SCH 23390 were studied in the squirrel monkey. For comparison, the dopamine D1 agonist SKF 38393 and the D2 agonist quinpirole were also studied. METHQD$ Subiects. The subjects were 11 adult male squirrel monkeys (Saimiri sciureus) housed in individual cages in rooms in which light; temperature and humidity were controlled. Fresh water was continuously available. The monkey's daily food intake was restricted to maintain their body weights between 800-1000 gms. The subjects were first implanted with both a venous catheter for the delivery of drug and an arterial catheter for the measurement of blood pressure during a single sterile surgery. The general surgical procedure has been described in detail elsewhere (16). In brief, polyvinyl chloride catheters were implanted in the external iliac vein and the internal iliac artery during anesthesia with halothane-oxygen mixtures. The distal ends of the catheters were passed s.c. out through the skin in the middle of the back. Monkeys wore nylon jackets at all times to protect the catheters. Following a 2 week recovery period, experiments were begun. Catheters were flushed with heparinized saline at least twice weekly and sealed with stainless steal obturators when not in use. ADtmratus. During experimental sessions, the monkeys sat in Plexiglas chairs similar to the one descrll~ed by Hake and Azrin (17) and were loosely restrained in the seated position by a waist lock. The chairs were enclosed in ventilated, sound-attenuating chambers (model AC-3; Industrial Acoustics Co., Bronx NY) that were provided with continuous white noise to mask extraneous sounds. The distal end of the arterial catheter, was connected via polyethylene tubing to a blood pressure transducer (no. T42-20, Coulboum Instruments, Lehigh Valley, PA). The arterial catheter was continuously flushed with heparinized saline (20 units/ml) at a rate of 0.13 ml/min. The transducer was connected to an associated amplifier (no. $72-25, Coulbourn Instruments) and blood pressure processor (no. $77-34, Coulbourn Instruments) outside the experimental chamber. The blood pressure processor analyzed the raw transducer signal, giving analog outputs of systolic (SP), diastolic (DP) and mean pressure (DP + [(SP - DP)/3]) at the end of each cardiac cycle. The signal for the end of the cardiac cycle was fed into an Apple IIe computer. For each cycle, the computer measured the time between cycles with a resolution of 1 msec and read the analog signals for pressure from the blood pressure processor with a resolution of 1 mm of Hg. These values were summed and averaged over periods of 30 secs for subsequent analysis. Only mean pressure and heart rate (HR) were used for statistical analysis. The distal end of the venous catheter was
Vol. 48, No. 16, 1991
Cocaine and Dopamine
1549
6O
~25 O~ "r
E
g2O
*
W r~
I¢1
n"
I,U 15 IX (3 o. ¢D
m n"
10
i SAL
I
,
I
.1
1
10
~"
2O
X
10 SAL
.1
1
10
DOSE (mg/kg)
DOSE (mg/kg)
FIG. 1 Dose-effect functions for the effects of cocaine on blood pressure and heart rate. The maximum blood pressure (mm Hg) and heart rate (BPM) increases are presented. These values were based on the means of each 5-min period following the injection of cocaine or saline, with the 5-min period just prior to the injection used as baseline. Each point is the mean ~ 1 SEM) of 4 monkeys, * indicates difference from saline control (p < .05). • 60.
40.
&
SALINE COC 0.1 COC 1.0 C ( ~ 3.0
m W
0 z < .1 I,I.-
,,~ .~. z
S~.
.1
1
10
5-MIN MEANS
FIG. 2 Time course for the effects of cocaine heart rate (BPM). Only selected doses of cocaine are presented. Each point is the mean of a 5-min period, with the 5-min period just prior to the cocaine or saline injection serving as baseline from which the change scores were calculated. The insert presents the maximum heart rate decrease that was observed for each dose of cocaine tested. Each point is the mean of 4 monkeys. The error bars are + 1 SEM, * indicates difference from saline control (p < .05). passed outside the experimental chamber via polyethylene tubing and connected to a syringe for the injection of either saline, cocaine, quinpirole or SKF 38393.
1550
Cocaine and Dopamine
Vol. 48, No. 16, 1991
Procedure. The subjects were placed in the chamber every weekday. Typically, injections of drugs were given on Tuesdays and Fridays and saline was given on Thursdays. The order of doses was non-systematic. For all sessions, the subjects were placed in the chamber for a period of 90 min, with the i.v. (1.0 ml/kg) injection of cocaine, quinpirole, SKF 38393 or saline given 30 min after initiation of the session. An i.m. injection (1.0 ml/kg) of saline, haloperidol or SCH 23390 was given 5 min preceding those sessions in which an i.v. injection of cocaine, quinpirole, SKF 38393 or saline was given 30 min into the session. There were typically two determinations of each drug condition, and multiple determinations of saline control. Dru~s. Cocaine hydrochloride (Mallinkrodt, St. Louis, MO) was dissolved in sterile physiologic~ saline. Quinpirole hydrochloride (Research Biochemicals Inc., RBI Natick, MA) and SKF 38393 hydrochloride (RBI) were dissolved in sterile water. SCH 23390 (RBI) and haloperidol (McNeil, Spring House, NJ) were prepared in sterile water, with slight acidification and mild heat. Doses are expressed as the salt (cocaine, quinpirole, SKF 38393, SCH 23390) or the base for haloperidol. The i.v. injections were given over a period not exceeding 30 see. Because of the length of the catheters, each i.v. injection was followed by a 1 ml flush of saline to insure all the drug was delivered to the animal. Data Analysis. Heart rate and blood pressure data were averaged over 5-rain periods from the 30-see periods averaged by the computer. Change scores were calculated using the 5 min period prior to the i.v. injection of saline, cocaine, quinpirole or SKF 38393 as the baseline. In addition, maximum increases and decreases from baseline were also determined across the entire period following the i.v. injection. These maximums were derived from the 5-rain means. Data were subjected to repeated measures analysis of variance, with follow-up tests performed by the method of Tukey, or Dunnett when comparisons were made to control values (18). RESULTS Figure 1 presents dose-effect functions for cocaine on blood pressure and heart rate. The measures presented are the maximum increase in blood pressure and heart rate. Cocaine produced clear, dose-dependent increases in blood pressure, F(5,15) = 4.7, p < .05. There was also a tendency for heart rate to be increased by cocaine, although this effect did not reach significance, F(5,15) = 2.0, p = .17. Figure 2 shows the time course for the cardiovascular effects of selected doses of cocaine on heart rate. Heart rate following saline injections remained fairy constant, with a small tendency to increase later in the session. Higher doses of cocaine (1.0-3.0 mg/kg) produced initial decreases in heart rate. The decrease in heart rate was followed by a return to baseline and a subsequent elevation above baseline later in the session. A lower dose of cocaine (0.1 mg/kg) produced only an increase in heart rate. The insert shows the dose-effect function for the maximum heart rate decrease. There was a clear, dose-dependent decrease in heart rate at the higher doses, F(5,15) = 3.5, p < .05. When given alone neither haloperidol (0.01 and 0.03 mg/kg) nor SCH 23390 (0.01 and 0.03 mg/kg) produced any effects which were significantly different from saline (data not shown). There was a clear trend for heart rate to increase by approximately 20 bpm towards the end of the session, which was significant for saline (see Fig. 2), haloperidol and SCH 23390 (p < .005). Figure 3, top panel, presents the time course for 3.0 mg/kg cocaine alone and in combination with 0.01 and 0.03 mg/kg haloperidol on heart rate. While neither dose of haloperidol attenuated the initial heart rate decreasing effect of this dose of cocaine, the heart rate increasing effect seen later in the session was reduced by the lower dose of haloperidol, F(22,88) = 2.34, p < .005. The middle panel of Figure 3 shows the time course for heart rate for 3.0 mg/kg cocaine alone and in combination with 0.01 and 0.03 mg/kg of SCH 23390. In contrast to haloperidol, SCH 23390 did not antagonize the heart rate increasing effects of cocaine, F(22,165) < 1.2, p > .29.
Vol. 48, No. 16, 1991
Cocaine and Dopamlne
1551
e0. 60. 40. 20.
0. -~O, -40. • -60 &
COG 3.0 +OCC HAL .(~ + CCC
-80 120. 100-
~~
o.
~.
10
~,
HAL .01 ÷COC * , ~ H .01 +COG
0
5-MIN MEANS
FIG. 3 Time course for the effects of cocaine (COC) alone and in combination with haiopeddol (HAL) and SCH 23390 (SCH) on heart rate. Details of the analysis are as in Fig. 2. The top panel is for 3.0 mg/kg cocaine and haloperidol (n = 4), the middle panel is for the same dose of cocaine and SCH 23390 (n = 6) and the bottom panel is for 0.3 mg/kg cocaine alone and in combination with both haloperidol and SCH 23390 (n = 4). The doses for haloperidol and SCH 23390 are given in mg/kg. The error bars are + 1 SEM, * indicates difference from cocaine (p < .05). As the effect of haloperidol on cocaine appeared to be restricted primarily to the heart rate increasing effect of cocaine, both haloperidol and SCH 23390 were tested in combination with a lower dose of cocaine (0.3 mg/kg) which produced only heart rate increasing effects. The bottom panel of Figure 3 presents the results of that experiment on the heart rate measure. SCH 23390 (0.01 mg/kg) failed to block the effect of this dose of cocaine on heart rate; however, haloperidol(0.01 mg/kg) clearly attenuated the heart rate increase, F(1,6) = 14.0, p < .01.
1552
Cocaine and Dopamine
120
•
Q4JIN1.0
Vol. 48, No. 16, 1991
40
100
T
,
[]
HAL .01
I~ SCH .01
;.
-
~
i!1
q $KF ,
1
,
2
i
3
,
4
,
5
,
6
,
7
,
8
,
9
,
,
,
10 11 12
,
~
,
COC3.0
COC0.3
QUINI,0
.
SK3.0
5-MIN M E A N S
FIG. 4 Time course (left panel) for the effects of 1.0 mg/kg quinpirole (QUIN) and 3.0 mg/kg SKF 38393 (SKF) alone and in combination with 0.01 mg/kg haloperidol (HAL) and 0.01 mg/kg SCH 23390 (SCH) on heart rate. The fight panel shows the effect of cocaine (COC), quinpirole (QUIN) and SKF 38393 (SKF) alone and in combination with haloperidol (HAL) or SCH 23390 (SCH) on blood pressure. The measure presented is a change score for the five min period immediately following the injection of cocaine, quinpirole or SKF 38393. The doses for haloperidol and SCH 23390 are given in mg/kg. The error bars are + 1 SEM, * indicates difference from cocaine (p < .05). Figure 4 (left panel) shows that 1.0 mg/kg quinpirole increased heart rate to a degree comparable to 0.3 mg/kg cocaine. Haloperidol (0.01 mg/kg) failed to antagonize the quinpiroleinduced tachyeardia. In contrast to quinpirole's effect, SKF 38393 (3.0 mg/kg) decreased heart rate. Like haloperidol against quinpirole, SCH 23390 failed to significantly antagonize that effect. While the effects of cocaine, quinpirole and SKF 38393 on heart rate varied over time, the peak effects observed on blood pressure primarily occurred immediately following the injection. Although the data analysis included all time points, the change scores for all 3 drugs are presented only for the first 5 min following the injection in Figure 4 (fight-hand panel). Both doses of cocaine increased blood pressure, with the higher dose producing a slightly larger effect. Neither haloperidol nor SCH 23390 (0.01 mg/kg) antagonized this effect. Similar results were observed at the 0.03 mg/kg dose of SCH 23390 (data not shown). Quinpirole (1.0 mg/kg) did not significantly affect blood pressure. Like cocaine, SKF 38393 increased blood pressure. However, in contrast to the results for cocaine, SCH 23390 antagonized the effect of SKF 38393 on blood pressure at some time periods, F(11,66) = 3.4, p < .001. Follow-up tests revealed that SCH 23390 antagonized the effect of SKF 38393 at the 5 min time period shown, p < .05. DISCUSSION Cocaine produced clear increases in both blood pressure and heart rate in squirrel monkeys. These results are consistent with previous reports of the effects of cocaine on cardiovascular function in squirrel monkeys (7,19), and are in agreement with results from both human(20) and animal (21) studies. At higher doses the heart rate response was biphasic, with an initial bradycardia followed by an increase in heart rate. This bradycardiac effect of cocaine is not typically observed in other species, but it may be related to the method of drug delivery. Tella et al. (7) showed that the bradycardiac effect of 3.0 mg/kg cocaine was not typically seen when the i.v.
Vol. 48, No. 16, 1991
Cocaine and Dopamine
1553
cocaine was infused slowly over a 10 rain period suggesting that direct depressant actions of cocaine on pacemaker function (22) may have slowed the heart. In addition, individual differences may play an important role in the heart rate decreasing effects of cocaine. Of the 11 subjects studied, only 6 showed a clear heart rate decreasing effect and 2 showed no evidence that cocaine would slow the heart. Reflex mechanisms are probably not involved in cocaine's bradycardiac effect as the ganglionic blocker hexamethonium does not attenuate the cocaine-induced bradycardia in most monkeys (7). The primary purpose of the current study was to determine whether the dopamine antagonists haloperidol or SCH 23390 would modify cocaine's effects on cardiovascular function. A number of previous reports have implicated dopamine in a number of cocaine's effects (23-25). For example, both SCH 23390 and haloperidol have been reported to attenuate the operant response rate increasing effects of cocaine (26). However, the potential contribution of dopamine systems to the cardiovascular effects of cocaine has not been investigated. The results of the present experiments suggest that dopamine may play at best a minor role in the cardiovascular effects of cocaine. At doses which by themselves had no effect on cardiovascular function, haloperidol attenuated the heart rate increasing effects of cocaine, while SCH 23390 had no effect. Neither haloperidol nor SCH 23390 affected the cocaine induced pressor response. The fact that neither haloperidol nor SCH 23390 affected blood pressure or heart rate when given alone, suggests that dopamine does not play a significant role in the tonic control of cardiovascular function in squirrel monkeys. These results suggest that dopamine D2 mechanisms may be involved in the positive chronotropic effect of cocaine on the cardiovascular system. The dopamine D2 agonist quinpirole increased heart rate similarly to cocaine. Further, dopamine D2 agonists have also been previously reported to produce excitatory effects on the cardiovascular system in conscious rodents (8-10). However, as haloperidol did not completely antagonize either cocaine's or quinpirole's cardiovascular effects and the antagonism that was seen was not dose-dependent, actions other than dopamine D2 receptor-mediated effects may also be involved. In particular, it has previously been reported that quinpirole's actions on cardiovascular function may be mediated through vasopressin (10). Whether vasopressin may also play a role in cocaine's effects is not clear. Further, the high dose of quinpirole required to produce cardiovascular effects may have influenced haloperidors ability to block those effects. Whereas 0.03 mg/kg quinpirole significantly affects the behavior of squirrel monkeys (27), 1.0 mg/kg was required to induce cardiovascular effects. Thus, an action of haloperidol through non-dopaminergic mechanisms cannot be ruled out. Nevertheless, the fact that a low dose of haloperidol (0.01 mg/kg) attenuated the tachycardiac effect of both 0.3 and 3.0 mg/kg cocaine does suggest the potential involvement of dopamine D2 mechanisms in the cardiovascular effects of cocaine. The dopamine D1 agonist SKF 38393 produced only decreases in heart rate and an increase in pressure. Unlike cocaine, the pressor effect of SKF 38393 was antagonized by SCH 23390. In addition, SCH 23390 attenuates the behavioral effects of cocaine (27). These results demonstrate that the doses and presession times for SCH 23390 should have been sufficient to block any cardiovascular effects of cocaine involving D1 receptors. Thus, the failure of SCH 23390 to attenuate the effects of cocaine on cardiovascular systems appears to rule out any D1 effects of cocaine on cardiovascular function. It should be noted that while the D1 agonist SKF 38393 appears to have high specificity in vitro, this may not be the case in vivo (28). Thus, we cannot rule out the possibility that SKF 38393 was having D2 agonist effects as well. Thus, while adrenergic mechanisms clearly play a crucial role in the cardiovascular effects of cocaine in squirrel monkey (7), dopaminergic mechanisms may also be involved. In particular, dopamine D2 actions may play some role in the positive chronotropic effect of cocaine. However, this relatively minor role for dopaminergic mechanisms in the cardiovascular effects of cocaine stands in marked contrast to the role of dopaminergic mechanisms in cocaine's behavioral effects.
1554
Cocaine and Dopamlne
Vol. 48, No. 16, 1991
A(~.NOWLEDGEMENTS The authors thank Eric Thorndike, Curt Vinyard, Scott Berger and Debra Boan for their technical assistance. A preliminary report appeared in FASEB J. 3 A296 (1989). REFERENCES 1.
.
3. 4. 5. 6. 7. 8. 9. 10. 11 12. 13.
14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
R.L. BALSTER, Mechanisms of cocaine abuse and toxicity. NIDA Research MonoaraDh 88, D. Clouet, K. Asghar and R. Brown (eds), 1-13, US Government Printing Office, Washington, DC (1988) R. E. HEIKKILA, H. ORLANSKY and G. COHEN, Pharmacol. Biochem. Pharmacol. 24 847-852 (1975). D. CLARK and F. J. WHITE, Synapse 1 347-388 (1987). C.-E. JOHANSON and M. M. FISCHMAN, Pharmacol. Rev. 41 3-52 (1989). J. M. W1TKIN, S. R. GOLDBERG and J. L. KATZ, Life Sci. 44 1285-1291 (1989). J. M. WITKIN, S. R. GOLDBERG, J. H. JAFFE and J. L. KATZ, Soc. Neurosci. Abst. 15 253 (1989). S. R. TELLA, C. W. SCHINDLER and S. R. GOLDBERG, J. Pharmacol. Exp. Ther. 252 491-499 (1990). Y.-F. CHEN, H. JIN, R. PAUL and S. NAGAHAMA, J. Pharmacol. Exp. Ther. 246 485492 (1988). S. NAGAHAMA, H. S. ANN, Y.-F. CHEN, M. D. LINDHEIMER and S. OPARIL, J. Pharmacol. Exp. Ther. 242 143-151 (1987). S. NAGAHAMA, Y.-F. CHEN, M. D. LINDHEIMER and S. OPARIL, J. Pharmacol. Exp. Ther. 2~(i 735-742 (1986). R.E. KOLLOCH, K. O. STUMPE, U. ISMER, O. KLETZKY and V. DEQUATTRO, Clin. Sci. 61 231s-234s (1981). J. R. SOWERS, G. RESCH, G. TEMPEL, J. HERZOG and M. COLANTINO, Acta Endocrino190 1-7 (1979). 0.-E. BRODDE, Cardiovascular function of oerivheral dopamine receptors, J. P. Hieble (ed.), 27-74, Marcel Dekker, Inc., New York (1990). J. D. KOHLI and L. I. GOLDBERG, Cardiovascular function of peripheral dopamine receptors, J. P. Hieble (ed.), 3-25, Marcel Dekker, Inc., New York (1990). M. ,~. SHERER, K. M. KUMOR and J. H. JAFFE, Psychiatry Res. 27 117-125 (1989). J. A. HERD, W. H. MORSE, R. T. KELLEHER and L. G. JONES, Am. J. Physiol. 217 2429 (1969). D.F. HAKE and N. AZRIN, J. Exp. Anal. Behav. 6 297-298 (1963). G. KEPPEL, Design and Analysis: A Researcher'8 Handbook, Prentice-Hall, Englewood Cliffs, NJ (1973). F. A. GONZALEZ and L. D. BYRD, Life Sci. 21 1417-1424 (1977). M. W. FISCHMAN, C. R. SCHUSTER, L. RESNEKOV, J. F. E. SHICK, N. A. KRASNEGOR, W. FENNEL and D.X. FREEDMAN, Arch. Gen. Psychiatry 33 983-989 (1976). R. D. WlLKERSON, J. Pharmacol. Exp. Ther. 246 466-471 (1988). U. TRENDELENBURG, J. Pharmacol. Exp. Ther. 161 222-231 (1968). J. BERGMAN, B. K. MADRAS, S. E. JOHNSON and R. D. SPEALMAN, J. Pharmacol. Exp. Ther. 2~1 150-155 (1989). S. FISHER, A. RASKIN and E. H. UHLENHUTH, (~o~aine: Clinical and Biobehavioral Aspects. Oxford University Press, New York (1987). R.]). SPEALMAN, B. K. MADRAS and J. BERGMAN, J. Pharmacol. Exp. Ther. 251 142149 (1989). J. L. KATZ and J. M. WlTKIN, Psychopharmacology 103 33-40 (1991). J. M. WlTKIN, C. W. SCHINDLER, S. R. TELLA and S. R. GOLDBERG, Psychopharmacology (in press). P. H. ANDERSON, Eur. J. Pharmacol. 146 113-120 (1988).
Pergamon Press
EFFECTS OF COCAINE AIZ)NE AND IN COMBINATION WITH HALOPERIDOL AND SCH 23390 ON CARDIOVASCULAR FUNCTION IN SQUIRREL MONKEYS C. W. Schindlerl, 2, S. R. Tellal, 3, J. M. Witldn 1, and S. R. Goldbergl, 2 1preclinical Pharmacology Branch, NIDA Addiction Research Center PO Box 5180, Baltimore, MD 21224 2Department of Pharmacology and Experimental Therapeutics University of Maryland, Baltimore, MD 21201 and 3Department of Pharmacology, Georgetown University School of Medicine 3900 Reservoir Road, Washington, DC 20007 (Received in final form February 12, 1991)
SUMMARY The potential involvement of D 1 and D2 dopamine receptors in the effects of cocaine on cardiovascular function in squirrel monkeys was evaluated. A low dose of cocaine (0.1 mg/kg i.v.) produced increases in both blood pressure and heart rate. At the higher doses of cocaine (1.0-3.0 mg/kg) the heart rate response was biphasic, consisting of an early decrease followed by an increase in heart rate 10-20 min following injection. The dopamine D2 antagonist haloperidol (0.01 mg/kg i.m.) attenuated the heart rate increasing effect of cocaine, but doses as high as 0.03 mg/kg did not alter the blood pressure increase. The D1 antagonist SCH 23390 (0.01-0.03 mg/kg i.m.) did not attenuate either the blood pressure or heart rate increasing effects of cocaine. The D2 agonist quinpirole (1.0 mg/kg i.v.) produced increases in heart rate similar to cocaine, with little effect on blood pressure. Although effective against the heart rate increasing effect of cocaine, haloperidol (0.01 mg/kg) did not antagonize the heart rate increasing effects of quinpirole. The D1 agonist SKF 38393 (3.0 mg/kg i.v.) decreased heart rate and increased blood pressure. The blood pressure increasing effect of SKF 38393 was antagonized by 0.01 mg/kg SCH 23390. Haloperidol's ability to partially antagonize the tachycardiac response to cocaine suggests the involvement of D2 receptors in that response. However, the failure of haloperidol to antagonize quinpirole's tachycardiac effect suggests that non-dopaminergic mechanisms may also be involved in haloperidors antagonism of cocaine's tachycardiac effect. The pressor effects of cocaine do not appear to be controlled by selective dopamine receptors. Continued high levels of cocaine abuse with accompanying behavioral and physiological toxicity has increased efforts to identify the mechanisms associated with that toxicity. However, as cocaine has diverse pharmacological actions, a number of factors may be relevant to the toxicity associated with this compound. For example, local anesthetic actions and the involvement of both catecholamine and indolamine neurotransmission have all been implicated in the effects of cocaine (1). One of the major actions of cocaine is blockade of the reuptake of dopamine with an accompanying increase in the synaptic availability of this neurotransmitter (2). The action of cocaine at dopamine synapses may be further influenced by the existence of multiple dopamine 0024-3205/91 $3.00 +.00
1548
Cocaine and Dopamlne
Vol. 48, No. 16, 1991
receptor subtypes (3). While the involvement of dopamine in the behavioral effects of cocaine has been studied extensively (4), there has only been limited study of the involvement of dopamine receptor subtypes in the toxic effects of cocaine. Witkin et al. (5,6) demonstrated in rodents that the lethal effects of cocaine could be attenuated by pretreatment with SCH 23390 and related D1 antagonists, but not by the inactive enantiomer of SCH 23390. The D2 antagonist haloperidol was inactive against the lethal effects of cocaine, but conferred protection against methamphetamineinduced lethality. The cardiovascular system has often been suggested as a primary target for cocaine toxicity. While the mechanisms by which cocaine produces its cardiovascular effects are not completely understood, adrenergic mechanisms are certainly involved (7). Nevertheless, the potential for dopaminergic systems to be involved in cardiovascular functioning is also evident. It has recently been established that both peripheral and central dopaminergic systems can be involved in cardiovascular regulation (8-10). In hypertensive patients (11) and experimental hypertensive rats (12) central dopaminergic activity has been shown to be altered as reflected in elevation of plasma prolactin levels. Dopaminergic receptors have also been identified in the periphery (13) and action at those receptors can modify cardiovascular function (14). Further, two distinct receptor subtypes for dopamine have been identified in the periphery. These two receptor subtypes appear to be pharmacologically similar to those described for the central nervous system (13). Thus, an action of cocaine at dopamine synapses may be expected to contribute to its cardiovascular effects. In support of this notion, a recent study by Sherer et al. (15) showed that haloperidol could partially attenuate the cardiovascular effects of cocaine in humans. Therefore, the purpose of the present study was to determine whether selective dopamine antagonists could modify cardiovascular effects of cocaine. Both the D2 antagonist haloperidol and the D1 antagonist SCH 23390 were studied in the squirrel monkey. For comparison, the dopamine D1 agonist SKF 38393 and the D2 agonist quinpirole were also studied. METHQD$ Subiects. The subjects were 11 adult male squirrel monkeys (Saimiri sciureus) housed in individual cages in rooms in which light; temperature and humidity were controlled. Fresh water was continuously available. The monkey's daily food intake was restricted to maintain their body weights between 800-1000 gms. The subjects were first implanted with both a venous catheter for the delivery of drug and an arterial catheter for the measurement of blood pressure during a single sterile surgery. The general surgical procedure has been described in detail elsewhere (16). In brief, polyvinyl chloride catheters were implanted in the external iliac vein and the internal iliac artery during anesthesia with halothane-oxygen mixtures. The distal ends of the catheters were passed s.c. out through the skin in the middle of the back. Monkeys wore nylon jackets at all times to protect the catheters. Following a 2 week recovery period, experiments were begun. Catheters were flushed with heparinized saline at least twice weekly and sealed with stainless steal obturators when not in use. ADtmratus. During experimental sessions, the monkeys sat in Plexiglas chairs similar to the one descrll~ed by Hake and Azrin (17) and were loosely restrained in the seated position by a waist lock. The chairs were enclosed in ventilated, sound-attenuating chambers (model AC-3; Industrial Acoustics Co., Bronx NY) that were provided with continuous white noise to mask extraneous sounds. The distal end of the arterial catheter, was connected via polyethylene tubing to a blood pressure transducer (no. T42-20, Coulboum Instruments, Lehigh Valley, PA). The arterial catheter was continuously flushed with heparinized saline (20 units/ml) at a rate of 0.13 ml/min. The transducer was connected to an associated amplifier (no. $72-25, Coulbourn Instruments) and blood pressure processor (no. $77-34, Coulbourn Instruments) outside the experimental chamber. The blood pressure processor analyzed the raw transducer signal, giving analog outputs of systolic (SP), diastolic (DP) and mean pressure (DP + [(SP - DP)/3]) at the end of each cardiac cycle. The signal for the end of the cardiac cycle was fed into an Apple IIe computer. For each cycle, the computer measured the time between cycles with a resolution of 1 msec and read the analog signals for pressure from the blood pressure processor with a resolution of 1 mm of Hg. These values were summed and averaged over periods of 30 secs for subsequent analysis. Only mean pressure and heart rate (HR) were used for statistical analysis. The distal end of the venous catheter was
Vol. 48, No. 16, 1991
Cocaine and Dopamine
1549
6O
~25 O~ "r
E
g2O
*
W r~
I¢1
n"
I,U 15 IX (3 o. ¢D
m n"
10
i SAL
I
,
I
.1
1
10
~"
2O
X
10 SAL
.1
1
10
DOSE (mg/kg)
DOSE (mg/kg)
FIG. 1 Dose-effect functions for the effects of cocaine on blood pressure and heart rate. The maximum blood pressure (mm Hg) and heart rate (BPM) increases are presented. These values were based on the means of each 5-min period following the injection of cocaine or saline, with the 5-min period just prior to the injection used as baseline. Each point is the mean ~ 1 SEM) of 4 monkeys, * indicates difference from saline control (p < .05). • 60.
40.
&
SALINE COC 0.1 COC 1.0 C ( ~ 3.0
m W
0 z < .1 I,I.-
,,~ .~. z
S~.
.1
1
10
5-MIN MEANS
FIG. 2 Time course for the effects of cocaine heart rate (BPM). Only selected doses of cocaine are presented. Each point is the mean of a 5-min period, with the 5-min period just prior to the cocaine or saline injection serving as baseline from which the change scores were calculated. The insert presents the maximum heart rate decrease that was observed for each dose of cocaine tested. Each point is the mean of 4 monkeys. The error bars are + 1 SEM, * indicates difference from saline control (p < .05). passed outside the experimental chamber via polyethylene tubing and connected to a syringe for the injection of either saline, cocaine, quinpirole or SKF 38393.
1550
Cocaine and Dopamine
Vol. 48, No. 16, 1991
Procedure. The subjects were placed in the chamber every weekday. Typically, injections of drugs were given on Tuesdays and Fridays and saline was given on Thursdays. The order of doses was non-systematic. For all sessions, the subjects were placed in the chamber for a period of 90 min, with the i.v. (1.0 ml/kg) injection of cocaine, quinpirole, SKF 38393 or saline given 30 min after initiation of the session. An i.m. injection (1.0 ml/kg) of saline, haloperidol or SCH 23390 was given 5 min preceding those sessions in which an i.v. injection of cocaine, quinpirole, SKF 38393 or saline was given 30 min into the session. There were typically two determinations of each drug condition, and multiple determinations of saline control. Dru~s. Cocaine hydrochloride (Mallinkrodt, St. Louis, MO) was dissolved in sterile physiologic~ saline. Quinpirole hydrochloride (Research Biochemicals Inc., RBI Natick, MA) and SKF 38393 hydrochloride (RBI) were dissolved in sterile water. SCH 23390 (RBI) and haloperidol (McNeil, Spring House, NJ) were prepared in sterile water, with slight acidification and mild heat. Doses are expressed as the salt (cocaine, quinpirole, SKF 38393, SCH 23390) or the base for haloperidol. The i.v. injections were given over a period not exceeding 30 see. Because of the length of the catheters, each i.v. injection was followed by a 1 ml flush of saline to insure all the drug was delivered to the animal. Data Analysis. Heart rate and blood pressure data were averaged over 5-rain periods from the 30-see periods averaged by the computer. Change scores were calculated using the 5 min period prior to the i.v. injection of saline, cocaine, quinpirole or SKF 38393 as the baseline. In addition, maximum increases and decreases from baseline were also determined across the entire period following the i.v. injection. These maximums were derived from the 5-rain means. Data were subjected to repeated measures analysis of variance, with follow-up tests performed by the method of Tukey, or Dunnett when comparisons were made to control values (18). RESULTS Figure 1 presents dose-effect functions for cocaine on blood pressure and heart rate. The measures presented are the maximum increase in blood pressure and heart rate. Cocaine produced clear, dose-dependent increases in blood pressure, F(5,15) = 4.7, p < .05. There was also a tendency for heart rate to be increased by cocaine, although this effect did not reach significance, F(5,15) = 2.0, p = .17. Figure 2 shows the time course for the cardiovascular effects of selected doses of cocaine on heart rate. Heart rate following saline injections remained fairy constant, with a small tendency to increase later in the session. Higher doses of cocaine (1.0-3.0 mg/kg) produced initial decreases in heart rate. The decrease in heart rate was followed by a return to baseline and a subsequent elevation above baseline later in the session. A lower dose of cocaine (0.1 mg/kg) produced only an increase in heart rate. The insert shows the dose-effect function for the maximum heart rate decrease. There was a clear, dose-dependent decrease in heart rate at the higher doses, F(5,15) = 3.5, p < .05. When given alone neither haloperidol (0.01 and 0.03 mg/kg) nor SCH 23390 (0.01 and 0.03 mg/kg) produced any effects which were significantly different from saline (data not shown). There was a clear trend for heart rate to increase by approximately 20 bpm towards the end of the session, which was significant for saline (see Fig. 2), haloperidol and SCH 23390 (p < .005). Figure 3, top panel, presents the time course for 3.0 mg/kg cocaine alone and in combination with 0.01 and 0.03 mg/kg haloperidol on heart rate. While neither dose of haloperidol attenuated the initial heart rate decreasing effect of this dose of cocaine, the heart rate increasing effect seen later in the session was reduced by the lower dose of haloperidol, F(22,88) = 2.34, p < .005. The middle panel of Figure 3 shows the time course for heart rate for 3.0 mg/kg cocaine alone and in combination with 0.01 and 0.03 mg/kg of SCH 23390. In contrast to haloperidol, SCH 23390 did not antagonize the heart rate increasing effects of cocaine, F(22,165) < 1.2, p > .29.
Vol. 48, No. 16, 1991
Cocaine and Dopamlne
1551
e0. 60. 40. 20.
0. -~O, -40. • -60 &
COG 3.0 +OCC HAL .(~ + CCC
-80 120. 100-
~~
o.
~.
10
~,
HAL .01 ÷COC * , ~ H .01 +COG
0
5-MIN MEANS
FIG. 3 Time course for the effects of cocaine (COC) alone and in combination with haiopeddol (HAL) and SCH 23390 (SCH) on heart rate. Details of the analysis are as in Fig. 2. The top panel is for 3.0 mg/kg cocaine and haloperidol (n = 4), the middle panel is for the same dose of cocaine and SCH 23390 (n = 6) and the bottom panel is for 0.3 mg/kg cocaine alone and in combination with both haloperidol and SCH 23390 (n = 4). The doses for haloperidol and SCH 23390 are given in mg/kg. The error bars are + 1 SEM, * indicates difference from cocaine (p < .05). As the effect of haloperidol on cocaine appeared to be restricted primarily to the heart rate increasing effect of cocaine, both haloperidol and SCH 23390 were tested in combination with a lower dose of cocaine (0.3 mg/kg) which produced only heart rate increasing effects. The bottom panel of Figure 3 presents the results of that experiment on the heart rate measure. SCH 23390 (0.01 mg/kg) failed to block the effect of this dose of cocaine on heart rate; however, haloperidol(0.01 mg/kg) clearly attenuated the heart rate increase, F(1,6) = 14.0, p < .01.
1552
Cocaine and Dopamine
120
•
Q4JIN1.0
Vol. 48, No. 16, 1991
40
100
T
,
[]
HAL .01
I~ SCH .01
;.
-
~
i!1
q $KF ,
1
,
2
i
3
,
4
,
5
,
6
,
7
,
8
,
9
,
,
,
10 11 12
,
~
,
COC3.0
COC0.3
QUINI,0
.
SK3.0
5-MIN M E A N S
FIG. 4 Time course (left panel) for the effects of 1.0 mg/kg quinpirole (QUIN) and 3.0 mg/kg SKF 38393 (SKF) alone and in combination with 0.01 mg/kg haloperidol (HAL) and 0.01 mg/kg SCH 23390 (SCH) on heart rate. The fight panel shows the effect of cocaine (COC), quinpirole (QUIN) and SKF 38393 (SKF) alone and in combination with haloperidol (HAL) or SCH 23390 (SCH) on blood pressure. The measure presented is a change score for the five min period immediately following the injection of cocaine, quinpirole or SKF 38393. The doses for haloperidol and SCH 23390 are given in mg/kg. The error bars are + 1 SEM, * indicates difference from cocaine (p < .05). Figure 4 (left panel) shows that 1.0 mg/kg quinpirole increased heart rate to a degree comparable to 0.3 mg/kg cocaine. Haloperidol (0.01 mg/kg) failed to antagonize the quinpiroleinduced tachyeardia. In contrast to quinpirole's effect, SKF 38393 (3.0 mg/kg) decreased heart rate. Like haloperidol against quinpirole, SCH 23390 failed to significantly antagonize that effect. While the effects of cocaine, quinpirole and SKF 38393 on heart rate varied over time, the peak effects observed on blood pressure primarily occurred immediately following the injection. Although the data analysis included all time points, the change scores for all 3 drugs are presented only for the first 5 min following the injection in Figure 4 (fight-hand panel). Both doses of cocaine increased blood pressure, with the higher dose producing a slightly larger effect. Neither haloperidol nor SCH 23390 (0.01 mg/kg) antagonized this effect. Similar results were observed at the 0.03 mg/kg dose of SCH 23390 (data not shown). Quinpirole (1.0 mg/kg) did not significantly affect blood pressure. Like cocaine, SKF 38393 increased blood pressure. However, in contrast to the results for cocaine, SCH 23390 antagonized the effect of SKF 38393 on blood pressure at some time periods, F(11,66) = 3.4, p < .001. Follow-up tests revealed that SCH 23390 antagonized the effect of SKF 38393 at the 5 min time period shown, p < .05. DISCUSSION Cocaine produced clear increases in both blood pressure and heart rate in squirrel monkeys. These results are consistent with previous reports of the effects of cocaine on cardiovascular function in squirrel monkeys (7,19), and are in agreement with results from both human(20) and animal (21) studies. At higher doses the heart rate response was biphasic, with an initial bradycardia followed by an increase in heart rate. This bradycardiac effect of cocaine is not typically observed in other species, but it may be related to the method of drug delivery. Tella et al. (7) showed that the bradycardiac effect of 3.0 mg/kg cocaine was not typically seen when the i.v.
Vol. 48, No. 16, 1991
Cocaine and Dopamine
1553
cocaine was infused slowly over a 10 rain period suggesting that direct depressant actions of cocaine on pacemaker function (22) may have slowed the heart. In addition, individual differences may play an important role in the heart rate decreasing effects of cocaine. Of the 11 subjects studied, only 6 showed a clear heart rate decreasing effect and 2 showed no evidence that cocaine would slow the heart. Reflex mechanisms are probably not involved in cocaine's bradycardiac effect as the ganglionic blocker hexamethonium does not attenuate the cocaine-induced bradycardia in most monkeys (7). The primary purpose of the current study was to determine whether the dopamine antagonists haloperidol or SCH 23390 would modify cocaine's effects on cardiovascular function. A number of previous reports have implicated dopamine in a number of cocaine's effects (23-25). For example, both SCH 23390 and haloperidol have been reported to attenuate the operant response rate increasing effects of cocaine (26). However, the potential contribution of dopamine systems to the cardiovascular effects of cocaine has not been investigated. The results of the present experiments suggest that dopamine may play at best a minor role in the cardiovascular effects of cocaine. At doses which by themselves had no effect on cardiovascular function, haloperidol attenuated the heart rate increasing effects of cocaine, while SCH 23390 had no effect. Neither haloperidol nor SCH 23390 affected the cocaine induced pressor response. The fact that neither haloperidol nor SCH 23390 affected blood pressure or heart rate when given alone, suggests that dopamine does not play a significant role in the tonic control of cardiovascular function in squirrel monkeys. These results suggest that dopamine D2 mechanisms may be involved in the positive chronotropic effect of cocaine on the cardiovascular system. The dopamine D2 agonist quinpirole increased heart rate similarly to cocaine. Further, dopamine D2 agonists have also been previously reported to produce excitatory effects on the cardiovascular system in conscious rodents (8-10). However, as haloperidol did not completely antagonize either cocaine's or quinpirole's cardiovascular effects and the antagonism that was seen was not dose-dependent, actions other than dopamine D2 receptor-mediated effects may also be involved. In particular, it has previously been reported that quinpirole's actions on cardiovascular function may be mediated through vasopressin (10). Whether vasopressin may also play a role in cocaine's effects is not clear. Further, the high dose of quinpirole required to produce cardiovascular effects may have influenced haloperidors ability to block those effects. Whereas 0.03 mg/kg quinpirole significantly affects the behavior of squirrel monkeys (27), 1.0 mg/kg was required to induce cardiovascular effects. Thus, an action of haloperidol through non-dopaminergic mechanisms cannot be ruled out. Nevertheless, the fact that a low dose of haloperidol (0.01 mg/kg) attenuated the tachycardiac effect of both 0.3 and 3.0 mg/kg cocaine does suggest the potential involvement of dopamine D2 mechanisms in the cardiovascular effects of cocaine. The dopamine D1 agonist SKF 38393 produced only decreases in heart rate and an increase in pressure. Unlike cocaine, the pressor effect of SKF 38393 was antagonized by SCH 23390. In addition, SCH 23390 attenuates the behavioral effects of cocaine (27). These results demonstrate that the doses and presession times for SCH 23390 should have been sufficient to block any cardiovascular effects of cocaine involving D1 receptors. Thus, the failure of SCH 23390 to attenuate the effects of cocaine on cardiovascular systems appears to rule out any D1 effects of cocaine on cardiovascular function. It should be noted that while the D1 agonist SKF 38393 appears to have high specificity in vitro, this may not be the case in vivo (28). Thus, we cannot rule out the possibility that SKF 38393 was having D2 agonist effects as well. Thus, while adrenergic mechanisms clearly play a crucial role in the cardiovascular effects of cocaine in squirrel monkey (7), dopaminergic mechanisms may also be involved. In particular, dopamine D2 actions may play some role in the positive chronotropic effect of cocaine. However, this relatively minor role for dopaminergic mechanisms in the cardiovascular effects of cocaine stands in marked contrast to the role of dopaminergic mechanisms in cocaine's behavioral effects.
1554
Cocaine and Dopamlne
Vol. 48, No. 16, 1991
A(~.NOWLEDGEMENTS The authors thank Eric Thorndike, Curt Vinyard, Scott Berger and Debra Boan for their technical assistance. A preliminary report appeared in FASEB J. 3 A296 (1989). REFERENCES 1.
.
3. 4. 5. 6. 7. 8. 9. 10. 11 12. 13.
14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
R.L. BALSTER, Mechanisms of cocaine abuse and toxicity. NIDA Research MonoaraDh 88, D. Clouet, K. Asghar and R. Brown (eds), 1-13, US Government Printing Office, Washington, DC (1988) R. E. HEIKKILA, H. ORLANSKY and G. COHEN, Pharmacol. Biochem. Pharmacol. 24 847-852 (1975). D. CLARK and F. J. WHITE, Synapse 1 347-388 (1987). C.-E. JOHANSON and M. M. FISCHMAN, Pharmacol. Rev. 41 3-52 (1989). J. M. W1TKIN, S. R. GOLDBERG and J. L. KATZ, Life Sci. 44 1285-1291 (1989). J. M. WITKIN, S. R. GOLDBERG, J. H. JAFFE and J. L. KATZ, Soc. Neurosci. Abst. 15 253 (1989). S. R. TELLA, C. W. SCHINDLER and S. R. GOLDBERG, J. Pharmacol. Exp. Ther. 252 491-499 (1990). Y.-F. CHEN, H. JIN, R. PAUL and S. NAGAHAMA, J. Pharmacol. Exp. Ther. 246 485492 (1988). S. NAGAHAMA, H. S. ANN, Y.-F. CHEN, M. D. LINDHEIMER and S. OPARIL, J. Pharmacol. Exp. Ther. 242 143-151 (1987). S. NAGAHAMA, Y.-F. CHEN, M. D. LINDHEIMER and S. OPARIL, J. Pharmacol. Exp. Ther. 2~(i 735-742 (1986). R.E. KOLLOCH, K. O. STUMPE, U. ISMER, O. KLETZKY and V. DEQUATTRO, Clin. Sci. 61 231s-234s (1981). J. R. SOWERS, G. RESCH, G. TEMPEL, J. HERZOG and M. COLANTINO, Acta Endocrino190 1-7 (1979). 0.-E. BRODDE, Cardiovascular function of oerivheral dopamine receptors, J. P. Hieble (ed.), 27-74, Marcel Dekker, Inc., New York (1990). J. D. KOHLI and L. I. GOLDBERG, Cardiovascular function of peripheral dopamine receptors, J. P. Hieble (ed.), 3-25, Marcel Dekker, Inc., New York (1990). M. ,~. SHERER, K. M. KUMOR and J. H. JAFFE, Psychiatry Res. 27 117-125 (1989). J. A. HERD, W. H. MORSE, R. T. KELLEHER and L. G. JONES, Am. J. Physiol. 217 2429 (1969). D.F. HAKE and N. AZRIN, J. Exp. Anal. Behav. 6 297-298 (1963). G. KEPPEL, Design and Analysis: A Researcher'8 Handbook, Prentice-Hall, Englewood Cliffs, NJ (1973). F. A. GONZALEZ and L. D. BYRD, Life Sci. 21 1417-1424 (1977). M. W. FISCHMAN, C. R. SCHUSTER, L. RESNEKOV, J. F. E. SHICK, N. A. KRASNEGOR, W. FENNEL and D.X. FREEDMAN, Arch. Gen. Psychiatry 33 983-989 (1976). R. D. WlLKERSON, J. Pharmacol. Exp. Ther. 246 466-471 (1988). U. TRENDELENBURG, J. Pharmacol. Exp. Ther. 161 222-231 (1968). J. BERGMAN, B. K. MADRAS, S. E. JOHNSON and R. D. SPEALMAN, J. Pharmacol. Exp. Ther. 2~1 150-155 (1989). S. FISHER, A. RASKIN and E. H. UHLENHUTH, (~o~aine: Clinical and Biobehavioral Aspects. Oxford University Press, New York (1987). R.]). SPEALMAN, B. K. MADRAS and J. BERGMAN, J. Pharmacol. Exp. Ther. 251 142149 (1989). J. L. KATZ and J. M. WlTKIN, Psychopharmacology 103 33-40 (1991). J. M. WlTKIN, C. W. SCHINDLER, S. R. TELLA and S. R. GOLDBERG, Psychopharmacology (in press). P. H. ANDERSON, Eur. J. Pharmacol. 146 113-120 (1988).