From a functional standpoint, excitatory amino acid neurotransmission within the limbic system appears to play an important role in the modulation of various behavioral consequences of exposure to psychostimulant drugs. Earlier studies had proposed a behavioral stimulatory role for excitatory amino acids in the nucleus accumbens on the basis of the increase in locomotor activity produced by local infusion of glutamate agonists, an effect accompanied by an increase in dopamine turnover
(46,47). Also, the locomotor hyperactivity produced by intranucleus accumbens infusion of glutamate agonists was blocked by a dopamine receptor antagonist (46), suggesting that pharmacological activation of nucleus accumbens excitatory amino acid receptors produced psychomotor stimulation, probably by facilitating dopamine neurotransmission at the same site, in accordance with neuro-chemical findings (31,32).
To investigate more closely the physiological significance of these observations, the reinforcing and locomotor-activating effects of psychostimulant drugs were tested after temporary reduction of nucleus accumbens excitatory amino acid neurotransmission obtained through local infusion of glutamate receptor antagonists. Microinfusion of both NMDA and non-NMDA receptor antagonists within the nucleus accumbens reduced the expression of the acute psychomotor-activating properties of cocaine and amphetamine administered systemically or amphetamine and dopamine infused into the nucleus accumbens (48-50) as well as novelty-induced locomotion in rats (51,52). These findings suggest that intact glutamate neurotransmission within the nucleus accumbens may be essential for the full expression of the acute locomotor-activating properties of psychostimulant drugs (53). Similarly, earlier studies pointed toward a tonic "permissive" role of excitatory amino acid in the expression of the acute reinforcing properties of cocaine. In fact, microinfusion of AP-5 within the core of the nucleus accumbens reduced the interreinforcement interval of cocaine self-administered intravenously by rats (54).
Anatomically, two subregions of the nucleus accumbens, the core and the shell, have recently been characterized on the basis of histochemical differences and separate neural connections. Investigation of the dopamine system within the core and the shell at the neurochemical and behavioral levels has revealed important differences (for a review, see ref. 55). The modulation of the nucleus accumbens function within the core and shell also appears to be differentially modulated by glutamate receptors. Local blockade of NMDA receptors within the nucleus accumbens core significantly reduced cocaine-induced locomotor activity while leaving spontaneous locomotion unaffected. In contrast, the blockade of NMDA receptors within the nucleus accumbens shell did not significantly modify cocaine-induced locomotor activity but increased spontaneous locomotion (56). This finding is also in accordance with previous observations indicating that intranucleus accumbens infusion of the glutamate receptor agonists NMDA and AMPA or the glutamate receptor antagonists CNQX and AP-5 all increased spontaneous locomotion. By comparison, the microinfusion of both antagonists decreased amphetamine-induced locomotor activity (57), although no differentiation was made in that study regarding core/shell infusions. These findings are in accordance with the observation that activity of nucleus accumbens neurons is associated with a high level of cortical arousal (58) and, therefore, it is likely that glutamate receptors may play a more substantial role in the regulation of nucleus accumbens neurons during periods of behaviorally relevant neural inputs of limbic origin.
The regulation of the integrated function of the nucleus accumbens by glutamate neurotransmission is therefore likely to depend on (1) the behavioral state of the organism and (2) the specific substructure where glutamate acts within the nucleus accumbens. The relative strength of each of these components is likely to determine the behavioral outcome achieved by increased or decreased excitatory amino acid neurotransmission within this structure.
Within the context of cocaine addiction, much evidence has been accumulated that antagonism of glutamate receptors affects various measures of cocaine self-administration. Support for the role of nucleus accumbens NMDA receptors in reinforcement comes from studies showing that the noncom-petitive NMDA receptor antagonists dizocilpine (MK-801), phencyclidine, and the competitive NMDA receptor antagonist 3-(2-carboxypiperazin-4-yl)propyl-1-phosphonic acid (CPP) support intracranial self-administration within the frontal cortex and the shell of the nucleus accumbens of the rat (59). The fact that microinfusion of the NMDA receptor antagonist AP-5 within the core of the nucleus accumbens appeared to reduce the interreinforcement interval of cocaine self-administration in a fixed ratio schedule (54) suggests the possibility that a dissociation between the core and the shell of the nucleus accumbens may exist with regard to the modulation of reinforcement, as it appears to exist with regard to the modulation of spontaneous and cocaine-induced locomotor activity (56).
Systemic pretreatment with dizocilpine within the dose range of 0.1-0.15 mg/kg, reduced responding for cocaine self-administration in a fixed ratio schedule (60), increased the breaking point of progressive ratio responding for cocaine (61), and failed to reinstate cocaine-seeking behavior after extinction (62). In contrast, within the dose range of 0.2-0.3 mg/kg, dizocilpine decreased responding for cocaine maintained both on a fixed ratio and on a progressive ratio schedule and reinstated cocaine-seeking behavior following extinction (60-62) Therefore it has been suggested that dizocilpine's ability to enhance the reinforcing effects of cocaine may lie within a narrow dose range on an inverted-U function and suggests that discrepancies in the outcome of dizocilpine's effects in the literature may result from using different doses in differently sensitive paradigms (61). Dizocilpine has also been reported to produce variable effects on dopamine release (60,63,64).
In a series of electrophysiological studies, French and co-workers (65-69) have shown that dizocilpine and phencycline are potent activators of ventral tegmental area A 10 neurons and this effect is not shared and is even blocked by competitive NMDA antagonists. More recent results have also shown that centrally acting competitive NMDA receptor antagonists, including CGS 19755, significantly modified the firing pattern of dopamine neurons, reducing the incidence and intensity of burst firing (70). These observatons suggest that competitive and non-competitive NMDA receptor antagonists may produce different effects on mesolimbic dopamine neurotransmission.
Altogether the use of dizocilpine as well as phencyclidine has provided valuable initial relevant information that have prompted much research on the role of excitatory amino acids on drug addiction and sensitization. However, the steepness of the dose-response curve produced by dizocilpine is associated with profound and bizarre behavioral effects, including motor impairment and ataxia, which have been reported at the dose of 0.25 mg/kg (63,71). In addition, the peculiar stimulating effects of dizocilpine on the firing of mesolimbic dopamine neurons are not shared with and are even reversed by other NMDA antagonists. These observations, together with the lack of potential for clinical application of dizocilpine, have prompted call for caution (63) in the interpretation of the effects produced by dizocilpine in the context of the investigation for the physiological role of NMDA neurotransmission.
Recently, a comparative investigation of the effects of systemically administered site-specific NMDA receptor antagonists has examined the effects produced by competitive and noncompetitive NMDA antagonists (72) in rats self-administering cocaine. The noncompetitive NMDA antagonist memantine was found to reduce responding for cocaine on a fixed-ratio schedule and produce a sizable (although reportedly nonsignificant) reduction of the breaking point in a progressive-ratio schedule. In addition, the competitive NMDA receptor antagonist CGP 39551 and the NMDA/glycine recognition-site antagonist L-701,324 did not modify responding for cocaine self-administration on a fixed-ratio schedule (72).
These results are similar to those obtained with the noncompetitive NMDA antagonist dex-tromethorphan (73). Dextromethorphan is a widely used antitussive agent that has been shown to act within the central nervous system as a noncompetitive antagonist at the NMDA receptor complex. In rats trained to self-administer cocaine, dextromethorphan reduced the maintenance of cocaine self-administration and significantly reduced the maximum number of responses performed by rats to obtain a dose of cocaine in a progressive-ratio schedule ("breaking point") The effects produced by dextromethorphan are similar to those produced by memantine, another noncompetitive NMDA antagonist (72) and suggest that these drugs may effectively modulate cocaine self-administration by suppressing the motivational strength to obtain the drug.
Comparatively less experimental evidence is available on the role of non-NMDA receptors in animals self-administering cocaine. The non-NMDA receptor antagonist DNQX reduced lever pressing for cocaine self-administration, but the effects appeared to be the result of general suppression of operant behavior because responding for food reinforcement was also reduced by DNQX (60) More recently, the role of glutamate, in the cocaine-dependence cycle has been investigated across multiple behavioral measures using a number of different behavioral paradigms of cocaine self-administration in order to extend previous findings to the various phases of the course of cocaine addiction.
Was this article helpful?