Brain Research, 594 (1992) 327-330
327
© 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
BRES 25408
N M D A receptors in the nucleus accumbens modulate intravenous cocaine but not heroin self-administration in the rat Luigi Pulvirenti, R a f a e l M a l d o n a d o - L o p e z a n d G e o r g e F. K o o b Department of Neuropharmacology, The Scripps Research Institute, La Jolla, CA 92037 (USA) (Accepted 28 July 1992)
Key words: Cocaine; Intravenous self-administration; Glutamate; Nucleus accumbens; Reward
The role of endogenous glutamate neurotransmission within the nucleus accumbens in the modulation of intravenous (i.e.) cocaine aad heroin self-administration in rats was analyzed. APV (2-amino-5-phosphonovaleric acid), a blocker of glutamate receptors of the N-methyI-D-aspartate (NMDA) type, was microinfused within the nucleus accumbens of the ventral striatum of rats trained to lever press for i.e. cocaine or heroin self-administration. APV, at the dose of 1.5 and 3.0/zg/side, reduced the rewarding value of cocaine while it left heroin self-administration unaffected. These results suggest that integrity of NMDA receptor function within the nucleus accumbens may be of importance for the maintenance of i.e. cocaine, but not heroin self-administration in rats.
The nucleus accumbens of the ventral striatum (AAC) is part of the limbic striatum and has been suggested as a critical substrate for drug reinforcement and drug dependence m. The maintenance of intravenous (i.e.) cocaine self-administration by rats seems to depend upon activation of dopamine receptors within the NAC "~'1"~+2°.Heroin self-administration, in contrast, seems to depend upon activation of opiate receptors within the NAC 23, while dopamine neurotransmission appears to play a less important role in this respect. Indeed, lesions of NAC dopamine terminals induced by the selective neurotoxin 6-hydroxydopamine abolished cocaine self-administration, but left heroin selfadministration relatively unaffected ~5. The efferent projections of the NAC involved in the stimulatory and rewarding effects of drugs seem to include the ventral pallidum and the dorsal medial thalamus 5'22. No information, in contrast, is available on the functional importance of the afferent systems to the NAC involved in the regulation of drug reinforcement. Such primary afferents arise from a number of limbic structures, of which the amygdaloid complex and the hippocampal formation provide the densest
input 4's'9 and seem to be glutamatergic in nature. Behavioral studies suggest that these pathways contribute to the modulation of locomotor activity t3. The aim of this study was therefore to investigate the role of glutamate within the NAC in the modulation of responding of i.e. self-administration of cocaine and heroin. The effects of microinjections of 2-amino5.phosphonovaleric acid (APV), a selective N-methylv-aspartate (NMDA) receptor antagonist, into the NAC of rats trained to respond for i.e. cocaine or heroin self-administration were examined. Male Wistar rats weighing 200-220 g at the start of the experiment were group-housed (2-3 per cage) and maintained on a 12-h light-dark cycle (lights on 05.00 h to 17.00 h) with food and water ad libitum. Rats were surgically prepared under halothane anesthesia with a chronic polyethylene catheter implanted into the external jugular vein as described by Hubner and Koob 5. At the same time, rats were stereotaxically implanted (Kopf Instruments, Tijunga, CA) with bilateral 23gauge, 10-ram steel intracranial cannulae aimed 3 mm above each NAC at coordinates AP + 3.0, L _+ 1.7, DV -4.8 (skull surface), with the tooth bar 5 mm above
Correspondence: L. Pulvirenti, Department of Neuropharmacology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037 USA. Fax: (I) (619) 554-6480.
328 ear bar zero. The cannulae were then fastened to the skull with dental cement and sealed with a 10-ram stylet wire. For self-administration testing, a cannula connector assembly was connected to a swivel and syringe pump 5, and was attached to the polyethylene assembly mounted on the animal's back immediately prior to the start of each session. The cannula connector was removed following the completion of a self-administration session and replaced with a guide cannula stylet.
iZ
',~.'~':~ ,". ,'~",
..
, ..: .Q .
\
\:...
/
J
o',,,ll;.'V ..
' "'..,~..
1[
\'~ ~
/
Fig. I, tti~tological localization of cannula placement, For coordinates, see text.
Four days after surgery, cocaine self-administration training was initiated. Rats were allowed 3-h access, 6 days per week, to a metal lever mounted on the wall of the self-administration chambers. Lever pressing resulted in i.e. injections of 0.1 ml cocaine (cocaine hydrochloride, 0.75 mg/kg/injection) or heroin (0.06 mg/kg/injection) (NIDA, Baltimore, MD) dissolved in 0.9% physiologic saline and administered over 4 s. A white signal light mounted above the lever indicated the onset of an injection and remained lit for 20 s, during which time the lever was inactive. Lever pressing was reinforced under a fixed ratio (FR) 1 (continuous reinforcement) schedule. After stable baseline responding for cocaine or heroin was obtained ( + 10% of the average for 2 consecutive days), rats were pretreated immediately before the beginning of the session with either APV or vehicle as described below. At least 48 h and 1 day of baseline self-administration separated all testing days. Twenty-three rats were divided into two separate groups. For the cocaine study, 12 rats were treated with either saline or two doses of APV; for the heroin study 11 rats were treated with either saline or APV. For intra-NAC injection, rats received an infusion (1 ~! over 2 min) of either saline (0.9% NACl) or APV (1.5 and 3.0 ~g/side) through 30-gauge injectors fashioned to extend 3 mm beyond the ventral tip of the cannula. These doses were chosen based on behavioral observations in previous studies ~s and given in random order. Data were analyzed using one-way ANOVA with repeated measures on dose. Individual mean comparisons were made with Duncan's multiple range test. Following completion of behavioral testing, all animals were sacrificed by overdose of pentobarbital and per. fused through the heart with cold 10% formalin/saline. The brains were then removed and 30 ~m frozen sections were cut in a frontal plane using ,', freezing microtome, Canula¢ sites were assessed without knowledge of behavioral results. Histological examination revealed that the injection sites were located within the NAC, 2,4-3,6 anterior to bregma (Fig, 1), Most cannulae fell within a plane of the NAC, between 3,0 and 3,2 mm anterior to bregma. Rats trained to respond for cocaine maintained a stable level of drug intake (baseline 34.3 ± 3,4 infusions/3 h; 11,8 ± 1,55 infus~ons/lst hour). The effect of intra-NAC injection of APV is shown in Fig. 2, left. APV treatment caused a dose-dependent increase in cocaine intake, both during the first 60 min (F2,2z = 5,852, P < 0.01) and during the entire 180-min session (F2,22 -" 5,805, P < 0.01). Similarly, rats trained to respond for self-adminis-
329 0'-60'
aoo
0'-60' NS
-
B
0
0'-180'
20o
0'-180'
| Q.
q)m n-
100
o
Saline
1.S
Dose (pg/side)
3.0
S
Saline
1.5
3.0
Dose (pg/side)
Fig. 2. Effect of intra-NAC microinfusion of APV on lever pressing for i.v. cocaine (left) or heroin (right) self-administration. Top panels: first 60 rain. Bottom panels: 180-rain session. * P < 0.05 and • * P < 0.01 compared to saline. NS, non-significant.
tration of heroin showed a stable level of basal intake (baseline 14.1 + 1.6 infusions/3 h; 5.1 + 0.7 infusions/ 1st hour). Intra-NAC microinjection of APV at either dose (Fig. 2, right) did not affect responding for heroin during the first 60 min (F2,20< 1, non-significant) or during the entire 180-rain session (F2.z0 < 1, non-significant). These results indicate that NMDA receptors in the NAC modulate cocaine, but not heroin self-administration. In a procedure of limited daily access to i.v. self-administration, an increase in drug intake can be observed following reduction of the dose dispensed per injection or after specific pharmacological blockade such as administration of dopamine receptor antagonists for cocaine or opiate receptor antagonists for heroin ~,~,t2'1~. This has been interpreted as a compensatory mechanism by the animal to restore the original rewarding value of the drug, reduced by competitive antagonism at the receptor site. It is therefore reasonable to consider the increase in cocaine self-administration observed in the present study as a reflection of the decreased effectiveness of cocaine as a reinforcer following intra-NAC injection of APV. There are several reasons to hypothesize that a glutamate-dopamine interaction within the NAC may represent the neurochemical basis of these effects. First, anatomical evidence suggests that afferent fibers to the NAC originate within the amygdala and the hippocampus and they appear to be glutamatergic4,s'~. Electron microscopy studies also showed that, within the NAC, these fibers of hippocampal origin lay in close apposition with tyrosine hydroxylase-positive nerve terminals, originating in the ventral tegmental area (VTA), which represents the main source of dopamine innervation of the NAC el. Second, both in vitro and in vivo neurochemical studies suggest that
infusion of glutamate or glutamate agonists within the NAC induces release of dopamine 7 ,4. Third, electrophysiological evidence indicates that NAC neurons show excitatory response to hippocampal stimulation 24. Finally, behavioral evidence suggests that microinfusion of glutamate within the NAC induces locomotor hyperactivity, which is blocked by administration of dopamine receptor antagonists 2. Furthermore, intraNAC injection of various glutamate receptor antagonists appears to reduce the activating properties of psychostimulant drugs 17'~8 and the rewarding properties of ethanol ~9. The NAC is a limbic structure critical for the expression of locomotor activity, the psychostimulant effect of drugs of abuse and drug reward ~°. Separate neurochemical components, however, underlie the rewarding and psychomotor activating properties of different drugs of abuse such as psychostimulants and opiates 3'~. It is known that the maintenance of cocaine self-administration requires the integrity of the NAC dopamine system, while heroin self-administration can be independent from activation of dopamine terminals in the NAC. Both neurotoxic lesions of the NAC dopamine terminals is and administration of dopamine receptor antagonists alter responding for cocaine but can leave heroia self-administration relatively unaffected 3'~s. Lesions of the efferent neuronal system of the NAC, in contrast, reduce the reinforcing properties of both cocaine and heroin 5'25. The lack of effect of glutamate antagonists on heroin self.administration appears, therefore, to support the hypothesis that dopamine interactions within the NAC aye not critical for the rewarding properties of opiate drugs. The release of dopamine from slices of the NAC is stimulated primarily by NMDA, with the activation of 2.amino.3.hydroxy-5-methylisoxazole-4-propionic acid (AMPA) and kainate receptors (two other subtypes of glutamate receptors) less potent ~. This relative AMPA insensitivity is further supported by the observation that intra-NAC administration of AMPA receptor antagonists did not modify i.v. cocaine self.administration (Pulvirenti and Koob, unpublished results). If a glutamate-dopamine interaction is the basis of the effects reported here, then it is not surprising that blockade of NMDA receptors produces more sustained effects compared with blockade of AMPA receptors. Taken together, these data suggest that endogenous glutamate neurotransmission within the NAC, possibly through a preferential activation of NMDA receptors, may play a 'permissive' role on the expression of at least some of the integrated functions of the NAC. The VTA-NAC-ventral pallidal system has been proposed as a major functional system for motivated
330 behavior t°. An allocorticaI-NAC circuitry may then be one of the functional substrates for iimbic motor integration 3 and a gating mechanism translating motivation into action, such as in goal-directed psychomotor responses. The present and other ,9 results suggest that NMDA receptors with the NAC may also modulate drug-seeking behavior. Evaluation at other levels of analysis will reveal the exact nature of the role of glutamate in this interface and will clarify the modulatory role of allocortical structures on subcortical function at the level of the NAC. This ts Publication NP-6289 of The Scripps Research Institute. This work was partially supported by NIDA Grants DA 04398 and DA 04043. We thank Dia,..l Smith for excellent technical assistance. We thank Dr. F.E. Bloom for critical discussion. REFERENCES I Bergman, J Kamien, J.B. and Spealman, R.D., Antagonism of cocaine self-administration by selective dopamine D! and D2 antagonist, Behae'. Pharmacol., 1 (1990)355-363. 2 Donzanti, B.A. and Uretsky, N.J., Effects of excitatory amino acids on locomotor activity after bilateral microinjec~ion into the rat nucleus accumbens: possible dependence on dopaminergic mechanisms, Neuropharmacology, 22 (1983) 971-98 I. 3 Ettenberg, A., Pettit, H.O., Bloom, F.E. and Kqob, G.F., Heroin and cocaine intravenous self-administration in rats: mediation by separate neural systems, P,sychopharmacology, 78 (1982)204-209. 4 Fuller, T.A., Ruschen, F.T. and Price, J.L., Source of presump. tire glutamatergic/aspartergic afferents to the rats ventral stri. atopallidal region, J. Comp. ~'~arol., 258 (1987) 317-338. 5 Hubner, C.B, and Koob, G.F., The ventral pallidum plays a role in mediating cocaine and heroin self.administration in the rat, Brain Rt,s., 508 (1990) 20-29. 6 l-hbner, C.B, and Moreton, J,E,, Effects of selective DI and D2 dopamine antagonists on cocaine self-administration in the rat, P~ychopharmacology, 105 ( 1991) I51 = 156, 7 Jones, S.M., Snell, L,D, and Johnson, K,M., Inhibition by phency. clidine of excitatory amino acid-stimulated release of neurotrans. mitter in the nucleus aceumbens, Net~ropharmacol~gy, 26 (1987) 173-179. 8 Kelley,A.E. and Domesick, V.B., The distribution of the projection from the hippocampal formation to the nucleus accumbens in the rat: an anterograde and retrograde horseradish peroxidas¢ study, Net~roscience, 7 (1982) 2321-2335. 9 Kelley, A.E., Domesick, V.B. and Nauta, W.J.H., The amygdalo. striatal projection in the rat. An anatomical study by anterograde and retrograde tracing methods, Neuroscience, 7 (1982)615-630. 10 Koob, G.F. and Bloom, F,E., Cellular and molecular mechanisms of drug dependence, Science, 242 (1988) 715-723.
I 1 Koob, G.F. and Goeders, N., Neuroanatomical substrates of drug self-administration. In J.M. Liebman and S.J. Cooper (Eds.), Neuropharmacologlcal Basis of Reward, Oxford University Press, Oxford, 1989, pp. 214-263. 12 Maldonado, R., Robledo, P., Chover, J. and Koob, G.F., Effects on cocaine self-administration of different dopamme antagonists injected into the nucleus accumbens m the rat, Soc. Neurosci. Abstr., 20 (1991) 683. 13 Mogenson, G.J., Limbic-motor integration. In: Progress m Psychobtology and Physiological ~,ychology, Vol. 12, Academic Press, New York, 1987, pp. 117-170. 14 Payson, M.M. and Donzanti, B.A., Effect of excitatory amino acids on mvivo dopamine release and metabolism in the nucleus accumbens, Soc. Neuroscl. Abstr., 584 (1989). 15 Pettit, H.O., Ettenberg, A., Bloom, F.E. and Koob, G.F., Destruction of dopamine in the nucleus accumbens selectively attenuates cocaine but not heroin self-administration in rats, Psy. chopharmacolo~, 84 (1984) 167-173. 16 Phillips, A.G., Broekkamp, C,L. and Fibiger, H.C., Strategies for studying the neurochemical substrates of drug reinforcement in rodents, Prec. Neuropsychopharmacol. Biol. P~ychiatry, 7 (1983) 585-590. 17 Pulvirenti, L., Swerdlow, N.R. and Koob, G.F., Microinje:tion of a glutamate antagonist into the nucleus accumbens reduces psychostimulant locomotion in rats, Neurosci. Lett., 103 (1989) 197245. 18 Pulvirenti, L., Swerdlow, N.R. and Koob, G.F., Nucleus accumhens NMDA antagonist decreases locomotor activity produced by cocaine, heroin, or accumbens dopamine, but not caffeine, Pharmacol. Biochem. Behav., in press. 19 Rassnick, S., Pulvirenti, L. and Koob, G.F., Oral ethanol self-administration is reduced by the administration of dopamine and glutamate receptor antagonists into the nucleus accumbens, Psy. chopharmacology (1992)in press. 20 Roberts, D.C.S., Corcoran, M.E. and Fibiger, H.C., On the role of ascending noradrenergic systems in intravenous self-administration of cocaine, Pharmacol. Biochem. Behav., 6 (1977)615-620. 21 Sesack, S,R. and Pickel, V.M., In the rat model nucleus accumbens, hippocampal and catecholaminergic terminals converge on spiny neurons and are in apposition to each other, Brain Res., 527 (1990) 266-279. 22 Swerdlow, N.R, and Koob, G,F,, Lesions of the dorsomedial nucleus of the thalamus, medial prefrontal cortex and pedunculopontine nucleus: effects on locomotor activity, mediated by nucleus accumbens-ventral pallidal circuitry, Brain R~:~,, 412 (1987) 233-243, 23 Vaccarino, FJ., Bloom, F.E. and Koob, G.F., Blockade of nucleus aecumbens opiate receptors attenuates intravenous heroin reward in the rat, Psychopharmacology, 85 (1985) 37-42. 24 Yang, C.R. and Mogenson, G J., Electrophysiological response of neurones in the nucleus accumbens to hippocampal stimulation and the attenuation of excitatory responses by the mesolimbic dopaminergic system, Brain Res,, 324 (1984) 69-84. 25 Zito, K.A., Viekers, G. and Roberts, D.C.S., Disruption of cocaine and heroin self-administration following i ainic acid lesions of the nucleus aecumbens, Pharmacol. Biochem. Behav., 23 (1985) 1029-1036.
327
© 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
BRES 25408
N M D A receptors in the nucleus accumbens modulate intravenous cocaine but not heroin self-administration in the rat Luigi Pulvirenti, R a f a e l M a l d o n a d o - L o p e z a n d G e o r g e F. K o o b Department of Neuropharmacology, The Scripps Research Institute, La Jolla, CA 92037 (USA) (Accepted 28 July 1992)
Key words: Cocaine; Intravenous self-administration; Glutamate; Nucleus accumbens; Reward
The role of endogenous glutamate neurotransmission within the nucleus accumbens in the modulation of intravenous (i.e.) cocaine aad heroin self-administration in rats was analyzed. APV (2-amino-5-phosphonovaleric acid), a blocker of glutamate receptors of the N-methyI-D-aspartate (NMDA) type, was microinfused within the nucleus accumbens of the ventral striatum of rats trained to lever press for i.e. cocaine or heroin self-administration. APV, at the dose of 1.5 and 3.0/zg/side, reduced the rewarding value of cocaine while it left heroin self-administration unaffected. These results suggest that integrity of NMDA receptor function within the nucleus accumbens may be of importance for the maintenance of i.e. cocaine, but not heroin self-administration in rats.
The nucleus accumbens of the ventral striatum (AAC) is part of the limbic striatum and has been suggested as a critical substrate for drug reinforcement and drug dependence m. The maintenance of intravenous (i.e.) cocaine self-administration by rats seems to depend upon activation of dopamine receptors within the NAC "~'1"~+2°.Heroin self-administration, in contrast, seems to depend upon activation of opiate receptors within the NAC 23, while dopamine neurotransmission appears to play a less important role in this respect. Indeed, lesions of NAC dopamine terminals induced by the selective neurotoxin 6-hydroxydopamine abolished cocaine self-administration, but left heroin selfadministration relatively unaffected ~5. The efferent projections of the NAC involved in the stimulatory and rewarding effects of drugs seem to include the ventral pallidum and the dorsal medial thalamus 5'22. No information, in contrast, is available on the functional importance of the afferent systems to the NAC involved in the regulation of drug reinforcement. Such primary afferents arise from a number of limbic structures, of which the amygdaloid complex and the hippocampal formation provide the densest
input 4's'9 and seem to be glutamatergic in nature. Behavioral studies suggest that these pathways contribute to the modulation of locomotor activity t3. The aim of this study was therefore to investigate the role of glutamate within the NAC in the modulation of responding of i.e. self-administration of cocaine and heroin. The effects of microinjections of 2-amino5.phosphonovaleric acid (APV), a selective N-methylv-aspartate (NMDA) receptor antagonist, into the NAC of rats trained to respond for i.e. cocaine or heroin self-administration were examined. Male Wistar rats weighing 200-220 g at the start of the experiment were group-housed (2-3 per cage) and maintained on a 12-h light-dark cycle (lights on 05.00 h to 17.00 h) with food and water ad libitum. Rats were surgically prepared under halothane anesthesia with a chronic polyethylene catheter implanted into the external jugular vein as described by Hubner and Koob 5. At the same time, rats were stereotaxically implanted (Kopf Instruments, Tijunga, CA) with bilateral 23gauge, 10-ram steel intracranial cannulae aimed 3 mm above each NAC at coordinates AP + 3.0, L _+ 1.7, DV -4.8 (skull surface), with the tooth bar 5 mm above
Correspondence: L. Pulvirenti, Department of Neuropharmacology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037 USA. Fax: (I) (619) 554-6480.
328 ear bar zero. The cannulae were then fastened to the skull with dental cement and sealed with a 10-ram stylet wire. For self-administration testing, a cannula connector assembly was connected to a swivel and syringe pump 5, and was attached to the polyethylene assembly mounted on the animal's back immediately prior to the start of each session. The cannula connector was removed following the completion of a self-administration session and replaced with a guide cannula stylet.
iZ
',~.'~':~ ,". ,'~",
..
, ..: .Q .
\
\:...
/
J
o',,,ll;.'V ..
' "'..,~..
1[
\'~ ~
/
Fig. I, tti~tological localization of cannula placement, For coordinates, see text.
Four days after surgery, cocaine self-administration training was initiated. Rats were allowed 3-h access, 6 days per week, to a metal lever mounted on the wall of the self-administration chambers. Lever pressing resulted in i.e. injections of 0.1 ml cocaine (cocaine hydrochloride, 0.75 mg/kg/injection) or heroin (0.06 mg/kg/injection) (NIDA, Baltimore, MD) dissolved in 0.9% physiologic saline and administered over 4 s. A white signal light mounted above the lever indicated the onset of an injection and remained lit for 20 s, during which time the lever was inactive. Lever pressing was reinforced under a fixed ratio (FR) 1 (continuous reinforcement) schedule. After stable baseline responding for cocaine or heroin was obtained ( + 10% of the average for 2 consecutive days), rats were pretreated immediately before the beginning of the session with either APV or vehicle as described below. At least 48 h and 1 day of baseline self-administration separated all testing days. Twenty-three rats were divided into two separate groups. For the cocaine study, 12 rats were treated with either saline or two doses of APV; for the heroin study 11 rats were treated with either saline or APV. For intra-NAC injection, rats received an infusion (1 ~! over 2 min) of either saline (0.9% NACl) or APV (1.5 and 3.0 ~g/side) through 30-gauge injectors fashioned to extend 3 mm beyond the ventral tip of the cannula. These doses were chosen based on behavioral observations in previous studies ~s and given in random order. Data were analyzed using one-way ANOVA with repeated measures on dose. Individual mean comparisons were made with Duncan's multiple range test. Following completion of behavioral testing, all animals were sacrificed by overdose of pentobarbital and per. fused through the heart with cold 10% formalin/saline. The brains were then removed and 30 ~m frozen sections were cut in a frontal plane using ,', freezing microtome, Canula¢ sites were assessed without knowledge of behavioral results. Histological examination revealed that the injection sites were located within the NAC, 2,4-3,6 anterior to bregma (Fig, 1), Most cannulae fell within a plane of the NAC, between 3,0 and 3,2 mm anterior to bregma. Rats trained to respond for cocaine maintained a stable level of drug intake (baseline 34.3 ± 3,4 infusions/3 h; 11,8 ± 1,55 infus~ons/lst hour). The effect of intra-NAC injection of APV is shown in Fig. 2, left. APV treatment caused a dose-dependent increase in cocaine intake, both during the first 60 min (F2,2z = 5,852, P < 0.01) and during the entire 180-min session (F2,22 -" 5,805, P < 0.01). Similarly, rats trained to respond for self-adminis-
329 0'-60'
aoo
0'-60' NS
-
B
0
0'-180'
20o
0'-180'
| Q.
q)m n-
100
o
Saline
1.S
Dose (pg/side)
3.0
S
Saline
1.5
3.0
Dose (pg/side)
Fig. 2. Effect of intra-NAC microinfusion of APV on lever pressing for i.v. cocaine (left) or heroin (right) self-administration. Top panels: first 60 rain. Bottom panels: 180-rain session. * P < 0.05 and • * P < 0.01 compared to saline. NS, non-significant.
tration of heroin showed a stable level of basal intake (baseline 14.1 + 1.6 infusions/3 h; 5.1 + 0.7 infusions/ 1st hour). Intra-NAC microinjection of APV at either dose (Fig. 2, right) did not affect responding for heroin during the first 60 min (F2,20< 1, non-significant) or during the entire 180-rain session (F2.z0 < 1, non-significant). These results indicate that NMDA receptors in the NAC modulate cocaine, but not heroin self-administration. In a procedure of limited daily access to i.v. self-administration, an increase in drug intake can be observed following reduction of the dose dispensed per injection or after specific pharmacological blockade such as administration of dopamine receptor antagonists for cocaine or opiate receptor antagonists for heroin ~,~,t2'1~. This has been interpreted as a compensatory mechanism by the animal to restore the original rewarding value of the drug, reduced by competitive antagonism at the receptor site. It is therefore reasonable to consider the increase in cocaine self-administration observed in the present study as a reflection of the decreased effectiveness of cocaine as a reinforcer following intra-NAC injection of APV. There are several reasons to hypothesize that a glutamate-dopamine interaction within the NAC may represent the neurochemical basis of these effects. First, anatomical evidence suggests that afferent fibers to the NAC originate within the amygdala and the hippocampus and they appear to be glutamatergic4,s'~. Electron microscopy studies also showed that, within the NAC, these fibers of hippocampal origin lay in close apposition with tyrosine hydroxylase-positive nerve terminals, originating in the ventral tegmental area (VTA), which represents the main source of dopamine innervation of the NAC el. Second, both in vitro and in vivo neurochemical studies suggest that
infusion of glutamate or glutamate agonists within the NAC induces release of dopamine 7 ,4. Third, electrophysiological evidence indicates that NAC neurons show excitatory response to hippocampal stimulation 24. Finally, behavioral evidence suggests that microinfusion of glutamate within the NAC induces locomotor hyperactivity, which is blocked by administration of dopamine receptor antagonists 2. Furthermore, intraNAC injection of various glutamate receptor antagonists appears to reduce the activating properties of psychostimulant drugs 17'~8 and the rewarding properties of ethanol ~9. The NAC is a limbic structure critical for the expression of locomotor activity, the psychostimulant effect of drugs of abuse and drug reward ~°. Separate neurochemical components, however, underlie the rewarding and psychomotor activating properties of different drugs of abuse such as psychostimulants and opiates 3'~. It is known that the maintenance of cocaine self-administration requires the integrity of the NAC dopamine system, while heroin self-administration can be independent from activation of dopamine terminals in the NAC. Both neurotoxic lesions of the NAC dopamine terminals is and administration of dopamine receptor antagonists alter responding for cocaine but can leave heroia self-administration relatively unaffected 3'~s. Lesions of the efferent neuronal system of the NAC, in contrast, reduce the reinforcing properties of both cocaine and heroin 5'25. The lack of effect of glutamate antagonists on heroin self.administration appears, therefore, to support the hypothesis that dopamine interactions within the NAC aye not critical for the rewarding properties of opiate drugs. The release of dopamine from slices of the NAC is stimulated primarily by NMDA, with the activation of 2.amino.3.hydroxy-5-methylisoxazole-4-propionic acid (AMPA) and kainate receptors (two other subtypes of glutamate receptors) less potent ~. This relative AMPA insensitivity is further supported by the observation that intra-NAC administration of AMPA receptor antagonists did not modify i.v. cocaine self.administration (Pulvirenti and Koob, unpublished results). If a glutamate-dopamine interaction is the basis of the effects reported here, then it is not surprising that blockade of NMDA receptors produces more sustained effects compared with blockade of AMPA receptors. Taken together, these data suggest that endogenous glutamate neurotransmission within the NAC, possibly through a preferential activation of NMDA receptors, may play a 'permissive' role on the expression of at least some of the integrated functions of the NAC. The VTA-NAC-ventral pallidal system has been proposed as a major functional system for motivated
330 behavior t°. An allocorticaI-NAC circuitry may then be one of the functional substrates for iimbic motor integration 3 and a gating mechanism translating motivation into action, such as in goal-directed psychomotor responses. The present and other ,9 results suggest that NMDA receptors with the NAC may also modulate drug-seeking behavior. Evaluation at other levels of analysis will reveal the exact nature of the role of glutamate in this interface and will clarify the modulatory role of allocortical structures on subcortical function at the level of the NAC. This ts Publication NP-6289 of The Scripps Research Institute. This work was partially supported by NIDA Grants DA 04398 and DA 04043. We thank Dia,..l Smith for excellent technical assistance. We thank Dr. F.E. Bloom for critical discussion. REFERENCES I Bergman, J Kamien, J.B. and Spealman, R.D., Antagonism of cocaine self-administration by selective dopamine D! and D2 antagonist, Behae'. Pharmacol., 1 (1990)355-363. 2 Donzanti, B.A. and Uretsky, N.J., Effects of excitatory amino acids on locomotor activity after bilateral microinjec~ion into the rat nucleus accumbens: possible dependence on dopaminergic mechanisms, Neuropharmacology, 22 (1983) 971-98 I. 3 Ettenberg, A., Pettit, H.O., Bloom, F.E. and Kqob, G.F., Heroin and cocaine intravenous self-administration in rats: mediation by separate neural systems, P,sychopharmacology, 78 (1982)204-209. 4 Fuller, T.A., Ruschen, F.T. and Price, J.L., Source of presump. tire glutamatergic/aspartergic afferents to the rats ventral stri. atopallidal region, J. Comp. ~'~arol., 258 (1987) 317-338. 5 Hubner, C.B, and Koob, G.F., The ventral pallidum plays a role in mediating cocaine and heroin self.administration in the rat, Brain Rt,s., 508 (1990) 20-29. 6 l-hbner, C.B, and Moreton, J,E,, Effects of selective DI and D2 dopamine antagonists on cocaine self-administration in the rat, P~ychopharmacology, 105 ( 1991) I51 = 156, 7 Jones, S.M., Snell, L,D, and Johnson, K,M., Inhibition by phency. clidine of excitatory amino acid-stimulated release of neurotrans. mitter in the nucleus aceumbens, Net~ropharmacol~gy, 26 (1987) 173-179. 8 Kelley,A.E. and Domesick, V.B., The distribution of the projection from the hippocampal formation to the nucleus accumbens in the rat: an anterograde and retrograde horseradish peroxidas¢ study, Net~roscience, 7 (1982) 2321-2335. 9 Kelley, A.E., Domesick, V.B. and Nauta, W.J.H., The amygdalo. striatal projection in the rat. An anatomical study by anterograde and retrograde tracing methods, Neuroscience, 7 (1982)615-630. 10 Koob, G.F. and Bloom, F,E., Cellular and molecular mechanisms of drug dependence, Science, 242 (1988) 715-723.
I 1 Koob, G.F. and Goeders, N., Neuroanatomical substrates of drug self-administration. In J.M. Liebman and S.J. Cooper (Eds.), Neuropharmacologlcal Basis of Reward, Oxford University Press, Oxford, 1989, pp. 214-263. 12 Maldonado, R., Robledo, P., Chover, J. and Koob, G.F., Effects on cocaine self-administration of different dopamme antagonists injected into the nucleus accumbens m the rat, Soc. Neurosci. Abstr., 20 (1991) 683. 13 Mogenson, G.J., Limbic-motor integration. In: Progress m Psychobtology and Physiological ~,ychology, Vol. 12, Academic Press, New York, 1987, pp. 117-170. 14 Payson, M.M. and Donzanti, B.A., Effect of excitatory amino acids on mvivo dopamine release and metabolism in the nucleus accumbens, Soc. Neuroscl. Abstr., 584 (1989). 15 Pettit, H.O., Ettenberg, A., Bloom, F.E. and Koob, G.F., Destruction of dopamine in the nucleus accumbens selectively attenuates cocaine but not heroin self-administration in rats, Psy. chopharmacolo~, 84 (1984) 167-173. 16 Phillips, A.G., Broekkamp, C,L. and Fibiger, H.C., Strategies for studying the neurochemical substrates of drug reinforcement in rodents, Prec. Neuropsychopharmacol. Biol. P~ychiatry, 7 (1983) 585-590. 17 Pulvirenti, L., Swerdlow, N.R. and Koob, G.F., Microinje:tion of a glutamate antagonist into the nucleus accumbens reduces psychostimulant locomotion in rats, Neurosci. Lett., 103 (1989) 197245. 18 Pulvirenti, L., Swerdlow, N.R. and Koob, G.F., Nucleus accumhens NMDA antagonist decreases locomotor activity produced by cocaine, heroin, or accumbens dopamine, but not caffeine, Pharmacol. Biochem. Behav., in press. 19 Rassnick, S., Pulvirenti, L. and Koob, G.F., Oral ethanol self-administration is reduced by the administration of dopamine and glutamate receptor antagonists into the nucleus accumbens, Psy. chopharmacology (1992)in press. 20 Roberts, D.C.S., Corcoran, M.E. and Fibiger, H.C., On the role of ascending noradrenergic systems in intravenous self-administration of cocaine, Pharmacol. Biochem. Behav., 6 (1977)615-620. 21 Sesack, S,R. and Pickel, V.M., In the rat model nucleus accumbens, hippocampal and catecholaminergic terminals converge on spiny neurons and are in apposition to each other, Brain Res., 527 (1990) 266-279. 22 Swerdlow, N.R, and Koob, G,F,, Lesions of the dorsomedial nucleus of the thalamus, medial prefrontal cortex and pedunculopontine nucleus: effects on locomotor activity, mediated by nucleus accumbens-ventral pallidal circuitry, Brain R~:~,, 412 (1987) 233-243, 23 Vaccarino, FJ., Bloom, F.E. and Koob, G.F., Blockade of nucleus aecumbens opiate receptors attenuates intravenous heroin reward in the rat, Psychopharmacology, 85 (1985) 37-42. 24 Yang, C.R. and Mogenson, G J., Electrophysiological response of neurones in the nucleus accumbens to hippocampal stimulation and the attenuation of excitatory responses by the mesolimbic dopaminergic system, Brain Res,, 324 (1984) 69-84. 25 Zito, K.A., Viekers, G. and Roberts, D.C.S., Disruption of cocaine and heroin self-administration following i ainic acid lesions of the nucleus aecumbens, Pharmacol. Biochem. Behav., 23 (1985) 1029-1036.
