The N-methyl-D-aspartate-evoked release of [3H]acetylcholine previously formed from [3H]choline was estimated in striosome- (identified by [3H]naloxone binding) or matrix-enriched areas of the rat striatum using an in vitro microsuperfusion procedure. Experiments were performed in either the absence or the presence of dopaminergic and/or GABAergic receptor antagonists. Although the cell bodies of the cholinergic interneurons were mainly found in the matrix, in the absence of magnesium, N-methyl-D-aspartate (50 microM) stimulated the release of [3H]acetylcholine in both striatal compartments. These responses were blocked by either magnesium, dizocilpine maleate, 7-chlorokynurenate or tetrodotoxin. N-Methyl-D-aspartate responses were concentration-dependent, but the 1 mM N-methyl-D-aspartate response was higher in striosomes than in the matrix. The co-application of D-serine (10 microM) enhanced the 10 microM N-methyl-D-aspartate response in both compartments, but reduced those induced by 1 mM N-methyl-D-aspartate, this reduction being higher in striosomes. The blockade of dopaminergic transmission with the D2 and D1 dopaminergic receptor antagonists, (-)-sulpiride (1 microM) and SCH23390 (1 microM), was without effect on the 50 microM N-methyl-D-aspartate-evoked release of [3H]acetylcholine, but markedly enhanced the 1 mM N-methyl-D-aspartate+D-serine-evoked response in striosomes and to a lesser extent in the matrix. Disinhibitory responses of similar amplitude were observed not only in striosomes but also in the matrix when (-)-sulpiride was used alone, while SCH23390 alone enhanced the 1 mM N-methyl-D-aspartate+D-serine response only in striosomes and to a lower extent than (-)-sulpiride. These results indicate that D2 receptors are mainly involved in the inhibitory effect of dopamine on the 1 mM N-methyl-D-aspartate+D-serine-evoked release of [3H]acetylcholine. They also show that the stimulation of D1 receptors can either reduce (striosomes) or enhance (matrix) this response, since in the latter case the effect induced by the combined application of the D1 and D2 receptor antagonists was smaller than that observed with the D2 receptor antagonist alone. Indicating that released GABA facilitates N-methyl-D-aspartate responses, the blockade of GABAA receptors with bicuculline (5 microM) reduced the 50 microM N-methyl-D-aspartate-evoked release of [3H]acetylcholine in both striatal compartments and the 1 mM N-methyl-D-aspartate+D-serine response in the matrix. These effects result from an inhibition by GABA of the evoked release of dopamine, since the reducing effects of bicuculline on N-methyl-D-aspartate responses were not observed under the complete blockade of dopaminergic transmission by the D1 and D2 receptor antagonists. Further demonstrating a facilitatory role of GABA in the control of N-methyl-D-aspartate-evoked release of [3H]acetylcholine, in the presence of bicuculline, (-)-sulpiride and SCH23390 alone or in combination enhanced, in both compartments, the responses induced not only by 1 mM N-methyl-D-aspartate+D-serine, but also by 50 microM N-methyl-D-aspartate.