5-Hydroxytryptamine-mediated effects of nicotine on endogenous GABA efflux from guinea-pig cortical slices

1. The effect of nicotine on endogenous basal GABA outflow was studied in guinea-pig cerebral cortex slices. 2. Nicotine 1.86-18.6 mumol l-1 significantly decreased the basal, tetrodotoxin-sensitive GABA efflux, whereas at higher concentrations (186-620 mumol l-1) nicotine increased it. The inhibition was prevented by mecamylamine while the facilitation was blocked by mecamylamine, (+)-tubocurarine and tetrodotoxin. 3. The effect of nicotine was due to an indirect 5-hydroxytryptaminergic action. In fact, MDL 72222 (1 mumol l-1) completely prevented the alkaloid inhibition and methysergide (1 mumol l-1) reversed the facilitation into inhibition; concomitant treatment with methysergide and MDL 72222 antagonized the effect of nicotine at 186 mumol l-1 4. Lower concentrations of 5-HT (3-10 mumol l-1) decreased, whereas higher concentrations (30-100 mumol l-1) increased, spontaneous GABA outflow. The inhibition of GABA efflux was prevented by MDL 72222 whereas the facilitation was reversed by methysergide (1 mumol l-1) into inhibition, and prevented by MDL 72222 1 mumol l-11. 5. These results suggest that, by activating nicotinic receptors present on 5-hydroxytryptaminergic terminals, nicotine releases 5-HT which, in turn, inhibits or increases the secretory activity of cortical GABA interneurones via 5-HT3 and methysergide-sensitive receptors, respectively.

Synthesis and photobiological properties of 3-acylangelicins, 3-alkoxycarbonylangelicins and related derivatives

Convenient synthesis of 3-acyl-2H-furo[2,3-h]-1-benzopyran-2-ones, esters of 2-oxo-2H-furo[2,3-h]-1-benzopyran-3-carboxylic acid and 2H-furo[2,3-h]-1-benzopyran-3-carboxamides was accomplished via aromatization of the adducts obtained by a reaction between (E)-5-dimethyl-aminomethylene-6,7-dihydrobenzofuran-4(5H)-one and the appropriate acylacetate or dialkyl malonate. These compounds are angelicin derivatives which were prepared with the aim of obtaining intrinsically monofunctional drugs for photochemotherapy, with only one photoreactive site in their molecule. The new angelicins appear to be free of the known phototoxicity of furocoumarins on the skin and at a genetic level. The 3-carboxylic esters showed significant antiproliferative activity in Ehrlich ascites cells and T2 bacteriophage; the other derivatives were only slightly effective. The features of these compounds are such that they represent a new model for non-toxic agents for photochemotherapy.

Early changes in prodynorphin mRNA and ir-dynorphin A levels after kindled seizures in the rat

Prodynorphin mRNA and immunoreactive dynorphin A (ir-dynorphin A) levels were measured in different brain areas at various time points after amygdala kindled seizures. In the hippocampus, striatum and hypothalamus, prodynorphin mRNA levels were not significantly changed in kindled rats (killed 1 week after the last stimulus-evoked seizure), but they were significantly increased 1 h after seizures. The relative increase was the highest in the hippocampus (approximately 3-fold). In the brainstem, midbrain and cerebral cortex no changes in prodynorphin mRNA were detected in kindled rats, 1 h or 1 week after a kindled seizure. ir-Dynorphin A levels were significantly reduced in the hippocampus and in the striatum of kindled rats, as well as 5 and 60 min after kindled seizures, but they were increased back to control levels after 120 min. In the hypothalamus, ir-dynorphin A levels were significantly increased 120 min after a kindled seizure. ir-Dynorphin A levels were also significantly reduced in the brainstem and in the frontal, parietal and temporal cortex 120 min, but not 5 or 60 min, after a kindled seizure. Taken together, these data support the hypothesis that the dynorphinergic system is activated after amygdala kindled seizures, with different kinetics in different brain areas.

Differences in the anatomic distribution of immediate-early gene expression in amygdala and angular bundle kindling development

Kindling is a model in which fleeting changes of neuronal activity produce a lifelong modification of neuronal structure and function in the mature nervous system. Immediate-early genes (IEGs) such as c-fos have been implicated as a causal link in the chain of molecular events coupling fleeting pathologic activity to lasting hyperexcitability. Identification of the brain structures exhibiting IEG expression during the evolution of kindling is necessary to guide investigations of the phenotypic consequences. We used in situ hybridization histochemistry to identify the structures exhibiting expression of multiple IEGs during the evolution of amygdala kindling and compared this to the pattern following angular bundle kindling. The principal findings included that: (1) generalized limbic and clonic motor (class 5) kindled seizures evoked by stimulation of one amygdala induced the expression of IEGs in a small subset of limbic structures with remarkable symmetry between the two hemispheres; (2) the anatomic extent of seizure-evoked expression of c-fos mRNA expanded progressively following focal limbic and motor (classes 0-3) seizures during the development of amygdala kindling; c-fos mRNA was detected first ipsilaterally in AM, ACO, and PC and with higher-class seizures in hippocampal formation and homologous structures contralaterally, and (3) class 5 seizures evoked by stimulation of two different sites in the limbic system (amygdala or angular bundle) induced IEG expression in distinct but partially overlapping anatomic structures. We propose that synaptic activation of glutamate receptors contributes to the expression of these diverse IEGs throughout the forebrain. The findings provide a constellation of anatomic structures in which to investigate the structural and functional consequences of IEG expression.

Adenosine A1 receptors in the rat brain in the kindling model of epilepsy

Adenosine and adenosine analogues have potent anticonvulsant effects on various seizure models, including kindling, an animal model of temporal lobe epilepsy. It is now reported that binding of a specific ligand (cyclohexyladenosine) to adenosine A1 receptors is not changed in the cerebral cortex of kindled rats. However, the affinity of cyclohexyladenosine to adenosine receptors is significantly increased in the hippocampus. In addition, cyclohexyladenosine is slightly more potent to inhibit [3H]D-aspartate outflow from hippocampal synaptosomes taken from kindled than from control rats. Taken together, these data suggest that an increased affinity of adenosine to A1 receptors may play a role in the anticonvulsant effect of adenosine A1 analogues in the kindling model.

Effect of arachidonic acid on [3H]D-aspartate outflow in the rat hippocampus

The aim of this study was to investigate the effect of arachidonic acid on [3H]d-aspartate outflow in rat hippocampus synaptosomes and slices. Arachidonic acid 1) increased basal outflow of [3H]d-aspartate in both synaptosomes and slices, and 2) increased K(+)-evoked overflow in slices but not in synaptosomes. The latter effect was dependent (at least in part) on arachidonic acid metabolism, most likely mediated by lipo-oxygenase metabolites and free radical production. It was prevented by nordihydroguairetic acid but not by indomethacin, and was significantly reduced by free radical scavengers (superoxide-dismutase and catalase). This effect was dependent upon stimulation since it could not be observed after a continuous perfusion of arachidonic acid in the absence of stimulation. Furthermore, it was long-lasting since a 30 min perfusion of arachidonic acid was sufficient to exert a significant effect on a stimulation following termination of the application.

Characterization of K(+)-evoked [3H]D-aspartate outflow in the rat hippocampus in vitro

The characteristics of K(+)-evoked outflow of [3H]D-aspartate, a glutamate release marker, were systematically investigated in the rat hippocampus, using 35 mM K(+)-evoked [3H]noradrenaline outflow as a reference. Elevation of external K+ concentrations increased [3H]D-aspartate outflow in a concentration-dependent manner both in slices and synaptosomes. In the absence of external Ca2+, K(+)-evoked [3H]D-aspartate outflow was decreased by approx 60% in synaptosomes and 80% in slices. However, elimination of external Ca2+ in the presence of 2 mM EGTA significantly reduced only 100 mM K(+)-evoked outflow, both in slices and synaptosomes. In the absence of external Ca2+, 35 mM K(+)-evoked [3H]noradrenaline outflow was abolished even when EGTA was present in the solution. Furthermore, the Ca(2+)-channel blockers omega-conotoxin (10 nM) and nifedipine (0.5 microM) did not significantly reduce K(+)-evoked [3H]D-aspartate outflow; [3H]noradrenaline outflow, however, was reduced by more than one third by omega-conotoxin. Finally [3H]D-aspartate overflow was insensitive to tetrodotoxin (0.5 microM) both in synaptosomes and in slices, while that of [3H]noradrenaline was significantly reduced in slices. It is concluded that (1) [3H]D-aspartate outflow is partly Ca(2+)-dependent; (2) differences between K(+)-evoked [3H]D-aspartate and [3H]noradrenaline outflow include sensitivity to stimulation by EGTA, to Ca(2+)-channel blockers and to tetrodotoxin. Some of these discrepancies may be ascribed to the existence of a cytosolic, Ca(2+)-independent pool of releasable glutamate and [3H]D-aspartate. These observations pose some problems as to the experimental approach for the study of Ca(2+)-dependent [3H]D-aspartate release.

Noradrenergic modulation of gamma-aminobutyric acid outflow from the human cerebral cortex

The noradrenergic modulation of endogenous gamma-aminobutyric acid (GABA) outflow from slices and synaptosomes prepared from human cerebral cortex biopsies has been studied. GABA outflow was responsive to depolarizing stimuli such as ouabain and high potassium. Basal GABA outflow in slices, but not in synaptosomes, appeared to be largely dependent upon neuronal activity, being prevented by tetrodotoxin (TTX). 10 mM K(+)-evoked outflow in synaptosomes also proved to be TTX sensitive. Norepinephrine (NE) concentration dependently increased basal GABA outflow both in slices and synaptosomes. This effect was alpha 1-adrenoreceptor-mediated because it was prevented by a selective antagonist of the alpha 1-adrenoreceptor class (prazosin) but not by the alpha 2 antagonist idazoxan. However, an alpha 2-mediated inhibitory modulation was also present in the preparations used, since (1) in slices, NE significantly inhibited GABA outflow in the presence of prazosin; (2) in synaptosomes, NE significantly inhibited 10 mM K(+)-evoked outflow in the presence of prazosin. Both of these effects were prevented by idazoxan. No beta-adrenoreceptor modulation could be demonstrated. A comparison between species was also conducted. The response to ouabain and to TTX proved similar in human, rat and guinea-pig cerebral cortex. In the most simple tissue preparation used (synaptosomes), a close similarity between the three species could be observed. In all species, NE stimulated basal GABA outflow, an effect prevented by prazosin. This suggests a predominant alpha 1-adrenoreceptor-mediated stimulatory effect. In a more complex preparation (slices), differences between species could be demonstrated.(ABSTRACT TRUNCATED AT 250 WORDS)

A pathogenetic hypothesis of temporal lobe epilepsy

Temporal lobe epilepsy is the most common type of epilepsy in adults. It frequently develops in previously normal nervous tissue, secondary to trauma, tumour or stroke. The disease has a tendency to progress toward generalization and neurologic deficits. The pathogenesis of temporal lobe epilepsy is still unclear. In this article, a hypothesis is proposed suggesting that a cascade of biological events may underlie its development and progression. These include increased excitatory amino acid release, NMDA receptor activation, influx of calcium into neurones, activation of calcium-dependent enzymes (including phospholipase A2), immediate early gene expression, and synthesis of new proteins. Positive and negative feedback loops as well as other events, taking place in parallel, are also hypothesized. The clinical and pharmacological ramifications of this working hypothesis are discussed.

Photobiological activity of certain new methylazapsoralens

The photobiological activity of a series of psoralen isosters carrying a nitrogen atom at 8 position, new potential drugs for the photochemotherapy of hyperproliferative skin diseases, have been studied; the more active derivatives appeared to be 5,4'-dimethyl-8-azapsoralen and 3,4,4'-trimethyl-8-azapsoralen which induced a strong inhibition of DNA synthesis in Ehrlich ascites cells, very similar to that provoked by 8-methoxypsoralen, the furocoumarin at present used in photochemotherapy. Such compounds induced a small amount of inter-strand DNA cross-links and were non phototoxic when assayed on guinea-pig skin; however, both derivatives appeared to be highly mutagenic in E. coli WP2 TM6. This strain contains the plasmid R46 and it is proficient in DNA repair, and therefore monoadducts do not should be mutagenic in such a strain. Because the first steps of excision, which remove monoadducts, and of the main cross-link repair use the same enzymes (produced by the uvrABC complex), in the presence of a great number of monofunctional lesions, it is possible that there are not sufficient enzyme molecules for removing cross-links according this pathway, which could be repaired by a second one, uvrABC independent and based on glycosilase activity, which works at reduced levels and is much less accurate.

Differential expression of immediate early genes in the hippocampus in the kindling model of epilepsy

Kindling is a phenomenon in which brief afterdischarges (ADs) evoked by periodic electrical stimulation of the brain eventually result in generalized clonic motor seizures. Once present, the enhanced sensitivity to electrical stimulation is lifelong. The mechanism by which brief ADs produce this long-lasting effect may involve a change in gene expression. To begin to investigate changes in gene expression that occur during kindling, we used in situ hybridization histochemistry to examine the time course of expression of mRNAs of the immediate early genes (IEGs) c-fos, c-jun, NGFI-A, and c-myc within the dorsal hippocampus of rats following a kindling AD. Three principal findings resulted from this study. First, the expression of all mRNAs except c-myc was significantly increased (P less than 0.05) within discrete neuronal populations. Second, the time course of expression of the IEGs differed markedly within the same neuronal population. Third, for a given IEG, the time course and anatomic pattern of expression were strikingly different among different neuronal populations of the hippocampus. The prolonged and distinctly different patterns of IEG expression suggest that target genes are differentially regulated in these neuronal populations for prolonged periods following a kindling AD.

Alpha-2 adrenoreceptor-mediated decrease in gamma-aminobutyric acid outflow in cortical slices and synaptosomes during morphine tolerance

Morphine tolerance has proven to be accompanied by alterations in the efficiency of many neuronal signals, as well as by an inversion of the noradrenergic signal response of cortical acetylcholine terminals and gamma-aminobutyric acid (GABA) neurons in vivo (decreased acetylcholine and increased GABA release in normal animals, vice versa in tolerant). The latter observation may be relevant in interpreting morphine withdrawal, because the noradrenergic neuron firing rate increases dramatically during its course. This study was designed as an attempt to anatomically localize the inversion of the GABA response to norepinephrine. Because this phenomenon is observed in cortical slices and synaptosomes, it can be postulated that it occurs in the neocortex at the level of the intracortical GABA nerve terminals. Pharmacological analysis demonstrates that although the stimulation observed in controls is alpha-1 adrenoreceptor-mediated, the inhibition in tolerant animals is exerted via alpha-2 adrenoreceptors. Therefore, an increase in number or an improved coupling to the transduction system of alpha-2 adrenoreceptors is hypothesized. This observation gives a clue to a molecular interpretation of the inversion of GABA response to norepinephrine in tolerant animals, which may be of heuristic value in terms of biological interpretation of morphine tolerance.

Alpha 1-adrenoreceptor-mediated increase in acetylcholine release in brain slices during morphine tolerance

Norepinephrine, clonidine, and phenylephrine increased the electrically evoked release of endogenous acetylcholine in cortical slices taken from morphine-tolerant guinea pigs. This effect was alpha 1-adrenoreceptor mediated and was opposite to the alpha 2-adrenoreceptor-mediated inhibition of acetylcholine release, normally elicited by norepinephrine and clonidine. In the presence of prazosin, clonidine recovered its normal inhibitory properties, suggesting that morphine tolerance induced the appearance of an alpha 1-adrenoreceptor-mediated response that overshadowed, but did not cancel, the still present alpha 2-adrenoreceptor inhibitory control. The attempt to prove the presence of alpha-adrenoreceptors on the nerve endings by testing the effect of norepinephrine in synaptosomal preparations (preloaded with [3H]choline and depolarized with KCl and veratridine) was unsuccessful. Therefore the problem of the exact location of this excitatory input remains to be solved. These results confirm previous findings reporting the increase in cortical acetylcholine release induced by the alpha-adrenoreceptor agonists in morphine-tolerant, freely moving guinea pigs and demonstrate that opiate tolerance inverts the direction of the noradrenergic modulation even in the isolated intracortical cholinergic structures.

Changes in cortical acetylcholine and gamma-aminobutyric acid outflow during morphine withdrawal involve alpha-1 and alpha-2 receptors

Naloxone (0.3-9 mumol kg-1), electrical stimulation of locus ceruleus or clonidine at low doses (7.5-112 nmol kg-1) increased the release of acetylcholine from the exposed parietal cortex of freely moving, morphine-tolerant guinea pigs. This increase was not additive and was prevented by prazosin (35.8 nmol kg-1), suggesting the involvement of alpha-1 receptors. At high doses (374 nmol kg-1 or more) clonidine inhibited acetylcholine release through alpha-2 receptors, as it did in naive animals at 7.5 nmol kg-1. Clonidine (374 nmol kg-1) and prazosin (35.8 nmol kg-1) reduced the objective signs of naloxone-precipitated withdrawal. Electrical stimulation of the locus ceruleus or naloxone treatment reduced the release of gamma-aminobutyric acid (GABA) from the exposed parietal cortex of morphine-tolerant guinea pigs. This reduction was not additive and was prevented by idazoxan (84 nmol kg-1), suggesting the involvement of alpha-2 receptors. Clonidine (7.5 nmol kg-1), too, reduced the release of GABA in morphine-tolerant animals. However, when tested jointly with naloxone, clonidine (7.5-112 nmol kg-1) induced alpha-1-mediated facilitation of GABA release (like that elicited in naive animals at 112-374 nmol kg-1) leaving the signs of withdrawal unchanged. This points to the stimulation of alpha-1 receptors highly responsive to this agonist (but not to locus ceruleus stimulation) during naloxone-precipitated withdrawal. In conclusion, chronic morphine treatment modifies the alpha-1- and alpha-2-mediated control of GABA and acetylcholine neurons.(ABSTRACT TRUNCATED AT 250 WORDS)

Lack of excitatory amino acid-induced effects on calcium fluxes measured with 45Ca2+ in rat cerebral cortex synaptosomes

Ca2+ uptake was measured in purified rat cerebral cortex synaptosomes (P3 pellets) using 45Ca2+ as a tracer. Ca2+ influx increased in time, and with an increase in external K+ concentration and temperature. The net (external K+-induced, depolarization-dependent) uptake follows a two-component course. The exponential term, due to the opening of voltage-operated calcium channels (VOC), has a rate constant which increases with an increase in the depolarization level (1.04 versus 0.54 nmol/s/mg protein for 50 mM - versus 15 mM [K+]-dependent net influx). The linear term, due to the Na+/Ca2+ exchange system, has a similar rate constant at all depolarization levels (0.16 +/- 0.05 and 0.11 +/- 0.02 nmol/s/mg protein). Excitatory amino acids (glutamate, kainate and n-methyl-d-aspartate-NMDA-) were tested on this preparation at doses ranging between 5 x 10(-5) M and 5 x 10(-3) M and at multiple incubation times, under resting conditions and under two depolarizing conditions (partial depolarization: 15 mM external K+ and maximal depolarization: 50 mM external K+). NMDA was also tested in the absence of Mg2+. No effect was detectable under any of these experimental conditions. Hypotheses to interpret these data are discussed. Further studies on other preparations are needed in order to directly investigate the presynaptic effects of excitatory amino acids.

Inversion of the alpha-2 and alpha-1 noradrenergic control of the cortical release of acetylcholine and gamma-aminobutyric acid in morphine-tolerant guinea pigs

In normal guinea pigs the adrenergic agonists clonidine and norepinephrine are known to inhibit directly the cortical outflow of acetylcholine (ACh) through alpha-2 receptors and to increase the cortical outflow of gamma-aminobutyric acid (GABA) through alpha-1 receptors. GABA, in turn, contributes to inhibit ACh through GABAA receptors. This scheme is changed drastically by morphine tolerance. In morphine-tolerant guinea pigs, clonidine at 7.5, 18.7 and 112 nmol/kg i.p. stimulates the cortical release of ACh through alpha-1 receptors. This effect is prevented by prazosin, 35.8 nmol/kg i.p. Clonidine reduces ACh release at high doses only (374 and 1122 nmol/kg i.p.). Furthermore, electrical stimulation of locus ceruleus also gives rise to a prazosin-sensitive increase in ACh release. In addition, locus ceruleus stimulation often causes behavioral activation rather than sedation. In morphine-tolerant guinea pigs, clonidine at 7.5 and 18.7 nmol/kg i.p. reduces GABA efflux through alpha-2 receptors, as the drug effect is prevented by idazoxan, 84 nmol/kg i.p. Clonidine increases GABA efflux at high doses only (112 and 374 nmol/kg i.p.). Locus ceruleus stimulation also gives rise to an idazoxan-sensitive reduction in GABA outflow. This new condition, evident after 7 days of morphine treatment, can be defined as inversion of the physiological norepinephrine control over ACh and GABA outflow and can represent a major part of the neurochemical derangement associated with opioid tolerance.

Dose- and time-dependent hippocampal cholinergic lesions induced by ethylcholine mustard aziridinium ion: effects of nerve growth factor, GM1 ganglioside, and vitamin E

Ethylcholine mustard aziridinium ion (ECMA) was infused intracerebroventricularly (icv) to rats followed by measurement of two markers of presynaptic cholinergic neurons, choline acetyltransferase (ChAT) activity and high affinity choline transport (HAChT), in the hippocampus and cortex. Bilateral icv administration of 1, 2, or 3 nmol of ECMA per side produced dose-dependent reductions in each marker in the hippocampus, but not in the cortex, one week after treatment. Reductions of 52% and 46% for ChAT activity and HAChT, respectively, were produced in the hippocampus by 3 nmol ECMA. Measurement of these two markers at different times after icv infusion of 2 nmol ECMA/ventricle revealed that the activity of ChAT was reduced to a greater extent than was HAChT in the hippocampus 1 day and 1, 2, 4, and 6 weeks after treatment. The maximal reductions of ChAT activity and HAChT (61% and 53%, respectively) were reached between 1 and 2 weeks after ECMA administration. There was no evidence of regeneration of either marker at 4 or 6 weeks posttreatment. HAChT and ChAT activity in the cortex were not altered at any of the posttreatment times examined. ECMA-induced deficits in hippocampal ChAT activity and HAChT were not counteracted by the following treatments: (i) daily administration of GM1 ganglioside (10 mg/kg, intraperitoneally (ip)) from the day prior to infusion of ECMA until 2 weeks later; (ii) daily administration of GM1 ganglioside between 2 and 6 weeks after infusion of ECMA; and (iii) icv administration of nerve growth factor (NGF) twice per week for 2 weeks after ECMA treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

Acetylcholine content in rat brain is elevated by status epilepticus induced by lithium and pilocarpine

The effects of status epilepticus on the concentration, synthesis, release, and subcellular localization of acetylcholine, the concentration of choline, and the activity of acetylcholinesterase in rat brain regions were studied. Generalized convulsive status epilepticus was induced by the administration of pilocarpine to lithium-treated rats. The concentration of acetylcholine in the cortex, hippocampus, and striatum decreased prior to the onset of spike activity or status epilepticus. Once status epilepticus began, the concentration of acetylcholine increased over time in the cortex and hippocampus, reaching peak levels that were 461% and 304% of control levels, respectively, after 2 h of seizures. Such high in vivo levels of acetylcholine had not been reported previously following any treatment. During status epilepticus, the concentration of acetylcholine in the striatum returned to control levels after the initial depression, but did not accumulate to high levels as it did in the other two regions. The in vivo cortical efflux of acetylcholine was also increased during the seizures. Choline levels were increased by status epilepticus in all three brain regions. Inhibition of seizures by pretreatment with atropine blocked the increases of acetylcholine and choline. Synaptosomes prepared from the cortex and from the hippocampus of rats with status epilepticus had elevated concentrations of acetylcholine: in the hippocampus the acetylcholine was principally in the cytoplasmic fraction, whereas in the cortex the acetylcholine was elevated in both the cytoplasmic and the vesicular fractions. The extra acetylcholine was in a releasable compartment, since increased K+ in the media or ouabain increased the release of acetylcholine from cortical slices to a greater extent in tissue from seized rats than from controls.(ABSTRACT TRUNCATED AT 250 WORDS)

Cortical acetylcholine release is increased and gamma-aminobutyric acid outflow is reduced during morphine withdrawal

The effects of naloxone on acetylcholine (ACh) and gamma-aminobutyric acid (GABA) outflow from the cerebral cortex of freely moving, morphine-dependent guinea-pigs was studied. The cortical efflux of ACh in chronically-treated guinea-pigs was about half of that of normal animals. GABA efflux was unaffected. During opioid withdrawal precipitated by naloxone (0.1-10 mg kg-1, i.p.) the guinea-pigs showed jumping, hyperactivity and wet dog shakes, the intensity of which was directly related to the dose of naloxone. The withdrawal syndrome was accompanied by a dose-dependent increase in ACh release and reduction in GABA outflow; ACh release was increased by naloxone at doses lower (0.1-3 mg kg-1) than those acting on GABA efflux (1-10 mg kg-1). Atropine (10 mg kg-1) and diazepam (5 mg kg-1) did not prevent GABA and ACh changes.