Glutamate agonists and [3H]GABA release from rat hippocampal slices: Involvement of metabotropic glutamate receptors in the quisqualate-evoked release

The effects of glutamate agonists and their selective antagonists on the Ca²⁺-dependent and independent releases of [³H]GABA from rat coronal hippocampal slices were studied in a superfusion system. The Ca²⁺-dependent release evoked by glutamate, kainate and N-methyl-D-aspartate (NMDA) gradually declined with time despite the continuous presence of the agonists. Quisqualate (QA) caused a sustained release which exhibited no tendency to decline within the 20-min period of stimulation. This release was enhanced in Ca²⁺-free medium. The release evoked by QA in Ca²⁺-containing medium was significantly inhibited by (+)-5-methyl-10,11-dihydro-5H-dibenzo(a,d)cyclohept-5,10-imine hydrogen maleate (MK-801) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), showing that QA activates NMDA receptors directly or indirectly through (RS)--amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors. The inhibition of MK-801 was slightly diminished and that of CNQX totally abolished in Ca²⁺-free medium. Verapamil inhibited the QA-activated release in both Ca²⁺-containing and Ca²⁺-free media. The effect of QA but not that of AMPA was blocked in Ca²⁺-free medium by L(+)-2-amino-3-phosphonopropionate (L-AP3), a selective antagonist of the metabotropic glutamate receptor. It is suggested that the sustained release of GABA is also mediated partly by activation of metabotropic receptors and mobilization of Ca²⁺ from intracellular stores.     

The release of [3H]GABA formed from [3H]glutamate in rat hippocampal slices

to compare the storage and release of endogenous GABA, of [3H]GABA formed endogenously from glutamate, and of exogenous [14C]GABA, hippocampal slices were incubated with 5 microCi/ml [3,4-3H]1-glutamate and 0.5 microCi/ml [U-14C]GABA and then were superfused in the presence or absence of Ca+ with either 50 mM K+ or 50 microM veratridine. Endogenous GABA was determined by high performance liquid chromatography which separated labeled GABA from its precursors and metabolites. Exogenous [14C]GABA content of the slices declined spontaneously while endogenous GABA and endogenously formed [3H]GABA stayed constant over a 48 min period. In the presence of Ca+ 50 mM K+ and in the presence or absence of Ca2+ veratridine released exogenous [14C]GABA more rapidly than endogenous or endogenously formed [3H]GABA, the release of the latter two occurring always in parallel. The initial specific activity of released exogenous [14C]GABA was three times, while that of endogenously formed [3H]GABA was only 50% higher than that in the slices. There was an excess of endogenous GABA content following superfusion with 50 mM K+ and Ca2+, which did not occur in the absence of Ca2+ or after veratridine. The observation that endogenous GABA and [3H]GABA formed endogenously from glutamate are stored and released in parallel but differently from exogenous labelled GABA, suggests that exogenous [3H] glutamate can enter a glutamate pool that normally serves as precursor of GABA.     

Electrically evoked release of [(3)H]noradrenaline from mouse cultured sympathetic neurons: release-modulating heteroreceptors

Cultured neurons from the thoracolumbar sympathetic chain of newborn mice are known to possess release-inhibiting alpha(2)-autoreceptors. The present study was carried out in a search for release-modulating heteroreceptors on these neurons. Primary cultures were preincubated with [(3)H]noradrenaline and then superfused and stimulated by single pulses, trains of 8 pulses at 100 Hz, or trains of 36 pulses at 3 Hz. The cholinergic agonist carbachol reduced the evoked overflow of tritium. Experiments with antagonists indicated that the inhibition was mediated by M(2) muscarinic receptors. The cannabinoid agonist WIN 55,212-2 reduced the evoked overflow of tritium through CB(1) receptors. Prostaglandin E(2), sulprostone, and somatostatin also caused presynaptic inhibition. The inhibitory effects of carbachol, WIN 55,212-2, prostaglandin E(2), and somatostatin were abolished (at the highest concentration of WIN 55, 212-2 almost abolished) by pretreatment of the cultures with pertussis toxin (250 ng/ml). Several drugs, including the beta(2)-adrenoceptor agonist salbutamol, opioid receptor agonists, neuropeptide Y, angiotensin II, and bradykinin, failed to change the evoked overflow of tritium. These results demonstrate a distinct pattern of presynaptic inhibitory heteroreceptors, all coupled to pertussis toxin-sensitive G proteins. The lack of operation of several presynaptic receptors known to exist in adult mice in situ may be due to the age of the (newborn) donor animals or to the culture conditions.     

Characterization of electrically evoked [3H]-D-aspartate release from hippocampal slices

Electrical stimulation has certain advantages over chemical stimulation methods for the study of neurotransmitter release in brain slices. However, measuring detectable quantities of electrically evoked release of endogenous or radiolabeled markers of excitatory amino acid neurotransmitters has required current intensities or frequencies much higher than those usually required to study other transmitter systems. We demonstrate here that [3H]-D-aspartate (D-ASP) release can be detected from hippocampal slices at lower stimulation intensities in the presence of a glutamate reuptake inhibitor. Subsequently, we optimized the electrical stimulus parameters for characterizing electrically evoked D-ASP release. Under the experimental conditions described, greater than 90% of electrically evoked D-ASP release is calcium-dependent. Evoked D-ASP release is markedly reduced by pre-treating slices with the synaptic vesicle toxin bafilomycin A1 (BAF A1) or in the presence of 10-mM magnesium. Evoked D-ASP release is also reduced to variable degrees by N- and P/Q type voltage-sensitive calcium channel antagonists. Neither spontaneous efflux nor evoked D-ASP release were affected by NMDA, AMPA or group I metabotropic glutamate receptor (mGluR) antagonists. Evoked D-ASP release was reduced in the presence of an adenosine A1 receptor agonist and potentiated by treatment with a group I mGluR5 agonist. Evoked [3H]-D-ASP release was similar in magnitude to evoked [3H]-L-glutamate (L-GLU) release. Finally, in separate experiments using the same electrical stimulus parameters, more than 90% of electrically evoked endogenous L-GLU release was calcium dependent, a pattern similar to that observed for evoked [3H]-D-ASP release. Taken together, these results indicate that electrically evoked [3H]-D-ASP release mimics evoked glutamate release in brain slices under the experimental conditions employed in these studies.     

Haloperidol reduces K+-evoked Ca2+-dependentd-[3H]aspartate release from rat hippocampal slices

Rat hippocampal slices preloaded withd-[³H]aspartate, a non metabolizable analogue ofl-glutamate, were superfused with artifical CSF. Depolarization was induced by 53.5 mM K⁺, in the presence of Ca²⁺ (1.3 mM) or Mg²⁺ (5 mM) to determine the Ca²⁺ dependent release. Haloperidol added in the superfusion medium at 100 M reduced by about 60% the Ca²⁺ dependent release ofd-[³H]aspartate. This drug at 20 M or 100 M inhibited the non-activated glutamate dehydrogenase (GDH) but had no effect on GDH activated by ADP (2 mM) or leucine (5 mM). In addition no effect was observed on phosphate activated glutaminase (PAG) in the presence either of 20 mM or 5 mM phosphate. These results indicate that the effect of haloperidol is exerted on presynaptic mechanisms regulating neurotransmitter release.     

Effect of selective noradrenergic denervation on noradrenaline content and [3H]dopamine release in rat nucleus accumbens slices

DSP4 (N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine) treatment (50 mg/kg i.p., 10 days previously) significantly decreased the noradrenaline (NA) content of the rostral part of the nucleus accumbens. The medial and caudal areas were not affected. The nucleus accumbens appears to receive noradrenergic innervation predominantly from subcoeruleus nuclei of the pons-medulla while the locus coeruleus neurons project to the rostral area. The isoproterenol-induced enhancement of the K⁺-evoked release of [³H]dopamine (DA) was not affected by DSP4 treatment. Noradrenergic denervation does not appear to have been sufficient to cause up-regulation of postsynaptic -adrenoceptors.     

An investigation of noradrenaline uptake and release by the CATH.a cell line

The cell bodies of ascending noradrenergic neurons in the brain are located predominantly in the locus coeruleus. An in vitro model of locus coeruleus neurons could prove to be a useful tool in the investigation of noradrenergic neural networks and their associated pathophysiologies. The CATH.a cell line demonstrates some of the properties expected of locus coeruleus neurons, and the present study investigated the neurotransmitter uptake and release properties of the CATH.a cells. It was surprising that the CATH.a cells failed to accumulate [3H]noradrenaline ([3H]NA), suggesting the lack of a functional NA transporter. RT-PCR supported this finding by demonstrating the absence of NA transporter mRNA. Treatment of CATH.a cells with various differentiating agents failed to increase the [3H]NA uptake. Endogenous NA release was studied using HPLC detection, which revealed a lack of depolarisation-induced increases in endogenous NA release. A human NA transporter-transfected CATH.a cell line was generated (termed RUNT), and a study of the [3H]NA uptake revealed that the RUNT cells displayed significant uptake that could be blocked by cocaine (10 microM). Furthermore, the uptake capacity could be dramatically increased by differentiation of the cells with dibutyryl cyclic AMP (1 mM) for 24 h. Using dibutyryl cyclic AMP-differentiated RUNT cells, high K+ concentrations (50 mM) significantly increased [3H]NA release above basal levels.     

Pharmacological characterization of the effects of methylmercury and mercuric chloride on spontaneous noradrenaline release from rat hippocampal slices

The environmental contaminants methylmercury (MeHg) and mercuric chloride (HgCl2) stimulated the spontaneous release of [3H]noradrenaline ([3H]NA) from hippocampal slices in a time- and concentration-dependent manner. Both MeHg and HgCl2 were similarly potent, with an EC50 of 88.4 microM and 75.9 microM, respectively. The releasing effects of MeHg and HgCl2 increased in the presence of desipramine, showing that the mechanism does not involve reversal of the transmitter transporter, and were completely blocked by reserpine preincubation, indicating a vesicular origin of [3H]NA release. The voltage-gated Na+ channel blocker tetrodotoxin (TTX) did not affect the response to mercury compounds. [3H]NA release elicited by MeHg was partially dependent on extracellular Ca2+, since it decreased significantly in a Ca2+-free EGTA-containing medium whereas HgCl2 induced a release of [3H]NA independent of extracellular Ca2+. Neither Ca2+-channels blockers, cobalt chloride (CoCl2) and (omega-conotoxin-GVIA, nor the Na+/Ca2+-exchanger inhibitor benzamil reduced MeHg-evoked [3H]NA release. Moreover, thapsigargin or caffeine, endoplasmic reticulum Ca2+-depletors, did not modify metal-evoked [3H]NA release, whereas ruthenium red, which inhibits the mitochondrial Ca2+ transport, decreased the effect of both MeHg and HgCl2. All these data indicate that, in hippocampal slices, mercury compounds release [3H]NA from the vesicular pool by a mechanism involving Ca2+ mobilization from mitochondrial stores.     

Influence of pipecolic acid on the release and uptake of [3H]GABA from brain slices of mouse cerebral cortex

Recently, pipecolic acid (PA) has been involved in the functioning of the GABAergic system. In the present work we have studied the effect of PA on GABA uptake and release in cerebral cortex slices. PA (100 M) was able to increase the release of [³H]GABA (90%) stimulated by mild depolarization with 15 mM potassium. If during the labeling of the tissue with [³H]GABA, -alanine was present, PA also enhanced the release (42%). However, when nipecotic acid was present instead -alanine, no stimulation of [³H]GABA release by potassium was observed neither in the control nor in the presence of PA. Spontaneous release was not affected by PA in any of the experimental conditions tested. In uptake experiments, only when -alanine was present in the medium PA significantly diminished the uptake (36%) of [³H]GABA. These results suggest that the effect of PA is mostly at the presynaptic level, inhibiting the neuronal GABA uptake and/or enhancing its release.     

Characterization of divalent cation-induced [3H]Acetylcholine release from EGTA-treated rat hippocampal synaptosomes

Calcium-naive synaptosomes were used to assess the effects of divalent cations on [³H]acetylcholine release from rat hippocampal homogenates. Following equilibration with calcium-free buffer (containing 10M EGTA), calcium reversibly increased [³H]acetylcholine efflux (up to five-fold) while causing no measurable efflux of lactate dehydrogenase. When substituted for calcium, strongtium and barium behaved similarly although barium exhibited three-fold greater efficacy. In the presence of elevated potassium, 4-aminopyridine or tetraethylammonium, the secretagogue efficacy of calcium (but not barium) was markedly increased. The release-promoting effects of both cations were inhibited by lanthanum, magnesium, cadmium, and -conotoxin but were insensitive to nifedipine and cobalt (both 10 M). In addition, stimulation of muscarnic cholinergic autoreceptors substantially inhibited both calcium and barium-evoked [³H]acetylcholine release. Taken together, these results indicate that cation-evoked transmitter release from calcium-naive synaptosomes is subject to normal neuroregulatory mechanisms and therefore should be useful for investigating presynaptic modulation of neuronal exocytosis.     

Potassium and veratrine-stimulated l-[3H]cysteine sulfinate and l-[3H]glutamate release from rat brain slices

The release of l-[³H]cysteine sulfinic acid, l-[³H]glutamatic acid and [³H]GABA from preloaded slices of various rat brain regions in response to either 30 mM K⁺ or veratrin was investigated. All these aminoacids were released by both depolarizing agents, which did not produce any changes in the spontaneous efflux of [³H]lysine. The K⁺ stimulated cysteine sulfinate release from superfused slices was found partly Ca²⁺-dependent in the subiculum, and mainly Ca²⁺-independent in the hippocampus whereas the K⁺-elicited glutamate release was partly Ca²⁺-dependent in both regions. The veratrine-induced release of both cysteine sulfinate and glutamate was blocked by verapamil in a dose-dependent way, although a small verapamil concentration independent release remained. The release pattern of both amino acids was heterogeneous, but roughly correlated among brain regions, except in the subiculum and hypothalamus.These findings demonstrate the releasability of both substances from various brain regions and suggest that those releases occur from different pools, being probably mainly of neuronal origin. They give further evidence that cysteine sulfinate as well as glutamate may serve a neurotransmitter role in the CNS.     

Histochemical demonstration of transmitter release from noradrenaline,dopamine and 5-hydroxytryptamine nerve terminals in field stimulated rat brain slices

Summary The effect of electrical field stimulation on noradrenaline (NA), dopamine (DA) and 5-hydroxytryptamine (5-HT) nerve terminals in rat brain slicesin vitro was investigated. Slices prepared from the cerebral cortex or the neostriatum were incubated in physiologic buffer for 30 min and then superfused by buffer and stimulated by an electrical field (biphasic pulses, 10 Hz, 12 mA, 2 ms) for various time periods. The effect of the stimulation was studied at the cellular level with the histochemical fluorescence technique of Falck and Hillarp. The transmitter overflow into the superfusing buffer caused by the stimulation was studied with isotope technique. Cerebral Cortex NA Nerve Terminals. Stimulation caused release of NA from cortical NA nerve terminals. Already after 2 min stimulation a slight decrease of the fluorescence intensity of the nerve terminals could be found. Stimulation for 15 to 30 min greatly reduced the fluorescence intensity. In slices preincubated with³H-NA the stimulation-induced overflow of tritium during 2 min stimulation was about 15% (i.e. 15% of the tissue tritium content was overflowing into the superfusing buffer in response to stimulation for 2 min). During prolonged stimulation there was a considerable decline of the tritium efflux. Cerebral Cortex 5-HT Nerve Terminals. The 5-HT-analogue 6-hydroxytryptamine (6-HT) which is readily taken up into 5-HT nerve terminals was used to permit good visualization of these nerve terminals. Uptake of 6-HT into cortical NA nerve terminals was prevented by preincubation with 6-hydroxydopamine (6-OH-DA) or protriptyline. Stimulation for 15 to 30 min reduced the fluorescence intensity of the 5-HT nerve terminals. In slices preincubated with³H-5-HT the stimulation-induced overflow of tritium during 2 min stimulation was about 5%. The tritium efflux slowly decreased during continuous stimulation. Neostriatal DA Nerve Terminals. In slices frozen directly after preparation an intense diffuse fluorescence could be seen. After incubation in drug-free buffer at 37° C the fluorescence was localized in the varicosities of the DA nerve terminals. The central parts of the slices almost completely lacked specific fluorescence, while the outer zones were brightly fluorescent. No clear reduction of the fluorescence intensity of the DA nerve terminals in the outer zone could be observed after stimulation for 30 min. In slices preincubated with³H-DA the stimulation-induced overflow of tritium during 2 min stimulation was about 2%. The tritium efflux slowly decreased during continuous stimulation.It is suggested that the differences in release between the various nerve terminal systems foundin vitro reflect differences in transmitter release occurringin vivo. The comparably high release of NA per impulse from the cortical NA nerve terminals implicate that the discharge rate of these neuronsin vivo is very low.This investigation has been supported by grants from the Swedish Medical Research Council (B72-14X-2330-05A) and Magnus Bergwall's Foundation.The author is greatly indebted to Mrs. Annika Hamberger for her skillful technical assistance. For generous supplies of drugs thanks are due to Hässle, Göteborg, Sweden, through Dr. H. Corrodi (6-HT, 6-OH-DA and H44/68).     

[3H]Acetylcholine and [3H]5-hydroxytryptamine release from rat midbrain slices and the effects of calcium and phenobarbital

The K-stimulated release of [³H]ACh from rat midbrain slices prelabeled by incubation with [³H]choline was dependent on extracellular Ca. Phenobarbital inhibited the K-stimulated [³H]ACh release and the IC₅₀ was equal to that found for K-stimulated endogenous ACh release. These results support the suggestion that barbiturates primarily inhibit the Ca-dependent stimulated release of ACh and affect ACh synthesis only indirectly. K-Stimulated release of [³H]5-HT was also inhibited by removing Ca from the medium or by adding phenobarbital which further supports the effects of barbiturates on the depolarization-induced release process. Fluoxetine, an inhibitor of 5-HT uptake, increased the amount of [³H]5-HT found in the medium but did not fully block the uptake of [³H]5-HT in this slice preparation.     

Effect of corticosterone on noradrenergic nuclei in the pons-medulla and [3H]NA release from terminals in hippocampal slices

The aim of the present study was to investigate possible membrane and genomic effects of corticosterone on the noradrenergic system of the rat brain. Corticosterone effects were studied in vivo by treating rats s.c. with 10 mg/kg corticosterone for 7 or 14 days. In the first two experiments corticosterone significantly decreased th noradrenaline (NA) and dopamine (DA) levels in the pons-medulla, an area which contains the A1-A7 noradrenergic cell groups, while the NA and DA levels in the dorsal hippocampus remained unchanged. In a third experiment where the locus coeruleus (LC) and the A1 and A2 nuclei (A1,A2) were analysed separately, NA levels were unchanged but total MHPG levels and the total MHPG/NA ratio were decreased in the A1,A2 area. Chronic corticosterone treatment (14 days) did not alter the ₂-adrenoceptor-mediated modulation of [³H]NA release from dorsal hippocampal slices. Neither the spontaneous outflow nor the electrically stimulated release of [³H]NA from dorsal hippocampal slices of untreated rats was affected by exposure of the slices to corticosterone (10^–7 M–10^–4 M) in the superfusion buffer. Thus, chronic corticosterone treatment of rats altered the noradrenergic system of the pons-medulla, but did not change the ₂-adrenoceptor-mediated modulation of NA release in the dorsal hippocampus, a major terminal area of the LC neurons. Corticosterone also did not appear to have a direct membrane effect on the NA terminals in the dorsal hippocampus of the rat.     

Modulation of [H]noradrenaline release by endothelin-1 in the rat tail artery

The effect of endothelin-1 (ET-1) on the increase in perfusion pressure and the release of noradrenaline produced by electrical field stimulation were examined in isolated perfused/superfused rat tail arteries. ET-1 (1–30 nM) increased, in an identical concentration-dependent manner, the basal perfusion pressure and the stimulation-evoked tritum overflow, whereas the basal outflow of noradrenaline was not changed by the peptide. These results show that, besides its postjunctional vasoconstrictor effect, ET-1 exerts in the rat tail artery a prejunctional action which might be involved in the modulation of stimulation-evoked noradrenaline release from postganglionic nerves.     

Presynaptic modulation by noradrenaline and an opioid of the substance P-induced release of [3H]acetylcholine from the myenteric plexus

Substance P (7.5-750 nM) applied in superfusion dose-dependently released 3H from isolated strips of myenteric plexus-longitudinal muscle of the guinea-pig ileum loaded with [3H]choline. Separation of the [3H]acetylcholine and [3H]choline components of the released radioactivity revealed that in response to substance P (SP) administration only the release of [3H]acetylcholine increased above resting level. A slowly developing tachyphylaxis to the effect of SP was observed. Evidence has been obtained that the slow tachyphylaxis developed to the acetylcholine-releasing effect of SP was not due to the exhaustion of releasable acetylcholine pool. Release of acetylcholine by 150 nM SP was completely prevented by tetrodotoxin or in a Ca2+-free medium and greatly reduced in the presence of noradrenaline or the opioid receptor agonist (D-Met2,Pro5)-enkephalinamide. The effect of noradrenaline and the opioid peptide was apparently prevented by yohimbine and naloxone, respectively.     

Spermine Inhibits [3H]Glycine Binding at the NMDA Receptors from Plexiform Layers of Chick Retina

Saturable specific binding of glycine to synaptosomal membranes from plexiform layers of the retina has been described, which seems to correspond to the modulatory site on NMDA-receptors (26). Spermine inhibited specific [³H]glycine binding to membranes from synaptosomal fractions from the outer (P₁) and the inner (P₂) plexiform layers of 1–3 day-old chick retinas in a dose-dependent manner with an IC₅₀ = 35 M for the P₁ fraction and 32 M for the P₂ fraction. Kinetic experiments and non-linear regression analysis of [³H]glycine-specific binding showed a K_d ~ 100–150 nM in both fractions, and a higher B_max (4.11 ± 0.47 pmol/mg protein) for the inner plexiform layer compared to the outer plexiform layer (B_max = 2.76 ± 0.25 pmol/mg protein). Strychnine-insensitive [³H]glycine binding was inhibited by 100 M spermine, due to a reduction in B_max (P₁ = 0.84 ± 0.16 pmol/mg protein; P₂ = 0.81 ± 0.16 pmol/mg protein) without affecting the K_d. Association and dissociation constants in the absence and presence of 50 M spermine remained unchanged. Results demonstrate the presence of a single modulatory site for spermine on NMDA receptors, in both synaptic layers of the chick retina.     

HP 749 enhances calcium-independent release of [3H]norepinephrine from rat cortical slices and synaptosomes

Previous studies have shown that, at concentrations of 1 M and 10 M, HP 749 increased electrically-stimulated release of [³H]norepinephrine (NE) from rat cortical slices. These effects were Ca²⁺-dependent, indicating an effect on release from vesicular stores. At 100 M, HP 749 had two effects. In addition to enhancing the Ca²⁺-dependent electrically-evoked release, it also induced a rise in the basal efflux (spontaneous release) of [³H]NE, which was observed in both cortical slices and synaptosomes. The spontaneous release effect was (1) not blocked by the reuptake inhibitor nomifensine, (2) not affected by removal of external calcium, (3) not blocked by vesicular depletion with reserpine, and (4) not inhibited by the sodium channel blocker tetrodotoxin (TTX). As would be expected, the spontaneous [³H]NE release induced by the cytoplasmic releaser tyramine and the sodium channel activator veratridine were blocked by nomifensine and TTX, respectively. Notably, however, the Ca²⁺-independent veratridine-induced release was completely blocked by 100 M HP 749. The mechanism of spontaneous release of [³H]NE caused by 100 M HP 749 is unresolved at present; however, the data are consistent with this release originating from a cytoplasmic source.     

Loss of [3H]kainate and of NMDA-displaceable [3H]glutamate binding sites in brain in thiamine deficiency: Results of a quantitative autoradiographic study

Previous studies suggest that alterations of brain glutamate synthesis and release occur in experimental thiamine deficiency. In order to assess the integrity of post-synaptic glutamatergic receptors in thiamine deficiency, binding sites for [³H]glutamate (displaced by NMDA), [³H]-kainate, and [³H]quisqualate (AMPA sites) were evaluated using Quantitative Receptor Autoradiography in rat brain following 14 days of treatment with the central thiamine antagonist pyrithiamine. Compared to pair-fed controls, brains of symptomatic thiamine-deficient animals contained significantly fewer NMDA-displaceable binding sites in cerebral cortex, medial septum and hippocampus. It has been suggested that NMDA-receptor mediated glutamate excitotoxicity plays a role in the pathogenesis of neuronal loss in thiamine deficiency. If such is the case, the selective loss of NMDA binding sites in cerebral cortex and hippocampus offers a possible explanation for the relative nonvulnerability of these brain regions to pyrithiamine-induced thiamine deficiency. [³H]quisqualate (AMPA) binding sites were unchanged in all brain regions of pyrithiamine-treated rats whereas [³H]kainate sites were significantly reduced in density in medial and lateral thalamus. The decline in these binding sites may be due to neuronal loss in pyrithiamine-induced thiamine deficiency. Alterations of glutamatergic synaptic function involving both NMDA and kainate receptor subclasses could contribute to the pathogenesis of neurological dysfunction in Wernicke's Encephalopathy in humans.