Excitatory amino acid receptors in the brain: membrane binding and receptor autoradiographic approaches.

In last month's article in this series, Lodge and Johnson discussed the contribution of noncompetitive excitatory amino acid antagonists to understanding of these receptors. In this third article, Anne Young and Graham Fagg describe how radioligand binding experiments have helped to fuel the recent burst of progress in understanding excitatory amino acid receptors in the brain. New and selective radioligands have facilitated mapping the distributions of the major excitatory receptor subtypes in normal and diseased brain, examining allosteric interactions within the NMDA receptor, searching for novel therapeutic agents and determining drug mechanisms, and making first steps along the path to defining receptor structure at the molecular level.

Characterization of the cerebroprotective efficacy of the competitive NMDA receptor antagonist CGP40116 in a rat model of focal cerebral ischemia: an in vivo magnetic resonance imaging study.

The cerebroprotective properties of the competitive NMDA antagonist D-(E)-2-amino-4-methyl-5-phosphono-3-pentenoic acid (CGP 40116) were evaluated in a rat model of focal cerebral ischemia. CGP 40116 (5-40 mg/kg i.v.) was injected immediately following permanent occlusion of the left middle cerebral artery (MCA). MK 801 (1 or 3 mg/kg i.v.), D-CPPene (20 mg/kg i.v.), and CGS 19755 (40 mg/kg i.v.) were used for comparison. Lesion volume was assessed using in vivo magnetic resonance imaging, which in initial experiments with parallel histological determinations proved to be an accurate method for the measurement of brain infarction and the determination of a cerebroprotective drug effect. CGP 40116 dose-dependently reduced the volume of cortical infarction, with an ED50 of 11 mg/kg i.v. and a maximal effect equivalent to a 62% reduction in cortical edema volume. Its cerebroprotective efficacy was thus comparable to that of MK 801. The rank order of potency for the NMDA antagonists was MK 801 > CGP 40116 approximately D-CPPene > CGS 19755. Neuroprotection by CGP 40116 was still apparent when treatment was started 30 min after MCA occlusion. It is concluded that CGP 40116 is an effective cerebroprotectant with potential clinical utility for amelioration of focal cerebral ischemic damage.

Gostatin blocks physiological actions and binding of acidic amino acids in rat brain.

Gostatin, an inhibitor of aspartate aminotransferases isolated from Streptomyces sumanensis NK-23, was tested as an antagonist of acidic amino acid-mediated responses in the in vitro hippocampal slice and of acidic amino acid binding. Gostatin blocked responses to N-methyl-aspartate (AMPA) or L-glutamate. Gostatin also displaced N-methyl-D-aspartate-sensitive [3H]-L-glutamate binding (Ki = 22.0 microM) more potently than [3H]-kainate binding. Gostatin appears to be a relatively nonselective acidic amino acid antagonist in the mammalian central nervous system.

A comparative analysis of the neuroprotective properties of competitive and uncompetitive N-methyl-D-aspartate receptor antagonists in vivo: implications for the process of excitotoxic degeneration and its therapy.

Injection of the N-methyl-D-aspartate receptor agonist, quinolinic acid, into the rat striatum in vivo results in the degeneration of cholinergic and GABAergic neurons, as determined seven days later using the marker enzymes, choline acetyltransferase and glutamate decarboxylase, respectively. Such damage was dose-dependently prevented by CGP 37849 or MK-801 (competitive and uncompetitive N-methyl-D-aspartate receptor antagonists, respectively) administered systemically or intrastriatally at the same time as quinolinic acid. The neuroprotective activity of CGP 37849 was associated with the D-enantiomer, CGP 40116 (ED50 7.5 mg/kg i.p.), which was approximately 1.5-fold and 3.5-fold more potent than the related compounds, D-CPPene and CGS 19755, respectively. CGP 37849 was a weaker neuroprotectant than MK-801 (ED50 0.8 mg/kg i.p) when administered systemically, but was dramatically more potent following coinjection with quinolinic acid (ED50's 0.2 and 117 nmol, respectively). When injected intrastriatally 0.5-2 h post-quinolinic acid, CGP 37849 was protective over the entire period studied, whereas MK-801 was less effective at all post-quinolinic acid injection times. The finding that CGP 37849 is neuroprotective when administered intrastriatally 1-2 h post-quinolinic acid supports the hypothesis that a period exists following excitotoxic insult in which neurons are not committed to die, and can be rescued by blockade of ongoing N-methyl-D-aspartate receptor activation. Competition studies indicated that, when coinjected with 100-400 nmol quinolinic acid into the striatum, CGP 37849 exhibited kinetics predicted of a competitive N-methyl-D-aspartate receptor antagonist (declining neuroprotective potency with increasing doses of agonist), whereas MK-801 displayed a complex picture, with weak protective activity at low doses of quinolinic acid. Following systemic administration, neither antagonist was markedly affected by the dose of excitotoxin. When given i.p. at up to 6 h post-quinolinic acid, CGP 37849 and MK-801 showed essentially identical profiles of post-insult protection; degeneration of cholinergic neurons was reduced significantly throughout the entire post-insult period, whereas GABAergic neurons were protected only when drugs were administered 2 h or earlier post-quinolinic acid. The data indicate that competitive and uncompetitive N-methyl-D-aspartate receptor antagonists are effective neuroprotectants in vivo, and that parameters such as drug lipophilicity or mechanism of action at the receptor do not impinge upon their properties as systemically active cerebroprotectants.

Cl-/Ca2+-dependent L-glutamate binding sites do not correspond to 2-amino-4-phosphonobutanoate-sensitive excitatory amino acid receptors.

A series of phosphono and phosphino analogues of glutamate were used to compare the pharmacological properties of (a) Cl-/Ca2+-dependent, 2-amino-4-phosphonobutanoate (AP4)-sensitive L-[3H]-glutamate binding sites in rat brain synaptic plasma membranes (SPMs) and (b) AP4-sensitive excitatory synaptic responses by use of electrophysiological techniques. In the presence of Cl- and Ca2+, L-[3H]-glutamate bound to SPMs with Kd 804 nM and Bmax 53 pmol mg-1 protein. The AP4-sensitive (Ki 7.3 microM) population of binding sites represented 61% of L-glutamate specifically bound. omega-Substituted analogues of AP4 were potent inhibitors of L-[3H]-glutamate binding (Ki values 2.4-38 microM), whereas N-substituted compounds or propionic acid derivatives were inactive. Experiments with AP4 alone and in combination with other analogues demonstrated that the primary target of all substances was the AP4-sensitive population of L-glutamate binding sites. In the hippocampal slice in vitro, AP4 antagonized lateral perforant path-evoked field potentials with an IC50 of 2.7 microM. In contrast to their actions at AP4-sensitive L-glutamate binding sites, all other compounds (except for the omega-carboxymethylphosphino analogue, IC50 19 microM) were weak or inactive as antagonists of this synaptic response (IC50 values greater than 100 microM). Inactive compounds which exhibited activity in the binding assay did not reverse the synaptic depressant effects of AP4, indicating that they were neither agonists nor antagonists at AP4-sensitive synapses. 4 The lack of correspondence between (a) the Cl- /Ca2 -dependent, AP4-sensitive population of L- [3H]-glutamate binding sites and (b) AP4-sensitive synaptic responses indicates that these binding sites are not the receptors through which AP4 exerts its neuropharmacological effects. The possibility that Cl- /Ca2+-dependent 'binding sites' represent transport into resealed SPM vesicles is discussed. 5 Electrophysiological data demonstrate that AP4-sensitive synaptic receptors display a high degree of ligand selectivity. High antagonist potency is shown only by glutamate analogues with unmodified alpha-amino and alpha-carboxyl groups, and with a bifunctional (dianionic) omega-terminal.

CGP 37849 and CGP 39551: novel and potent competitive N-methyl-D-aspartate receptor antagonists with oral activity.

1. The pharmacological properties of CGP 37849 (DL-(E)-2-amino-4-methyl-5-phosphono-3-pentenoic acid; 4-methyl-APPA) and its carboxyethylester, CGP 39551, novel unsaturated analogues of the N-methyl-D-aspartate (NMDA) receptor antagonist, 2-amino-5-phosphonopentanoate (AP5), were evaluated in rodent brain in vitro and in vivo. 2. Radioligand binding experiments demonstrated that CGP 37849 potently (Ki 220 nM) and competitively inhibited NMDA-sensitive L-[3H]-glutamate binding to postsynaptic density (PSD) fractions from rat brain. It inhibited the binding of the selective NMDA receptor antagonist, [3H]-((+/-)-3-(2-carboxypiperazin-4-yl)propyl-1-phosphonate (CPP), with a Ki of 35 nM, and was 4, 5 and 7 fold more potent than the antagonists [+/-)-cis-4-phosphonomethylpiperidine-2-carboxylic acid) (CGS 19755), CPP and D-AP5, respectively. Inhibitory activity was associated exclusively with the trans configuration of the APPA molecule and with the D-stereoisomer. CGP 39551 showed weaker activity at NMDA receptor recognition sites and both compounds were weak or inactive at 18 other receptor binding sites. 3. CGP 37849 and CGP 39551 were inactive as inhibitors of L-[3H]-glutamate uptake into rat brain synaptosomes and had no effect on the release of endogenous glutamate from rat hippocampal slices evoked by electrical field stimulation. 4. In the hippocampal slice in vitro, CGP 37849 selectively and reversibly antagonized NMDA-evoked increases in CA1 pyramidal cell firing rate. In slices bathed in medium containing low Mg2+ levels, concentrations of CGP 37849 up to 10 microM suppressed burst firing evoked in CAl neurones by stimulation of Schaffer collateral-commissural fibres without affecting the magnitude of the initial population spike; CGP 39551 exerted the same effect but was weaker. In vivo, oral administration to rats of either CGP 37849 or CGP 39551 selectively blocked firing in hippocampal neurones induced by ionophoreticallyapplied NMDA, without affecting the responses to quisqualate or kainate. 5. CGP 37849 and CGP 39551 suppressed maximal electroshock-induced seizures in mice with ED50 s of 21 and 4 mg kg'- p.o., respectively. 6. CGP 37849 and CGP 39551 are potent and competitive NMDA receptor antagonists which show significant central effects following oral administration to animals. As such, they may find value as tools to elucidate the roles of NMDA receptors in brain function, and potentially as therapeutic agents for the treatment of neurological disorders such as epilepsy and ischaemic brain damage in man.

Evaluation of quinolinic acid induced excitotoxic neurodegeneration in rat striatum by quantitative magnetic resonance imaging in vivo.

Excitotoxic neurodegeneration in the rat striatum was induced by direct injection of quinolinic acid. The degree of damage was evaluated in vivo 1 day later by quantitative magnetic resonance imaging (MRI) and 7 days later in the same animals by measuring the activities of the neuronal marker enzymes choline acetyltransferase and glutamic acid decarboxylase. Striatal damage assessed using the two approaches was highly correlated. Moreover the cerebroprotective efficacy of the N-methyl-D-aspartate receptor antagonist CGP 40116 was indistinguishable based on all analytical parameters. MRI, however, was more reproducible than the enzymatic methods and was faster and simpler for routine analyses of excitotoxic damage and cerebroprotection in vivo.

Acidic amino acid binding sites in mammalian neuronal membranes: their characteristics and relationship to synaptic receptors.

This review summarizes studies designed to label and characterize mammalian synaptic receptors for glutamate, aspartate and related acidic amino acids using in vitro ligand binding techniques. The binding properties of the 3 major ligands employed--L-[3H]glutamate, L-[3H]aspartate and [3H]kainate--are described in terms of their kinetics, the influence of ions, pharmacology, molecular nature, localization and physiological/pharmacological function. In addition, the binding characteristics are described of some new radioligands--[3H]AMPA, L-[3H]cysteine sulphinate, L-[35S]cysteate, D-[3H]aspartate, D,L-[3H]APB, D-[3H]APV and D,L-[3H]APH. Special emphasis is placed on recent findings which allow a unification of the existing binding data, and detailed comparisons are made between binding site characteristics and the known properties of the physiological/pharmacological receptors for acidic amino acids. Through these considerations, a binding site classification is suggested which differentiates 5 different sites. Four of the binding site subtypes are proposed to correspond to the individual receptor classes identified in electrophysiological experiments; thus, A1 = NMDA receptors; A2 = quisqualate receptors; A3 = kainate receptors; A4 = L-APB receptors; the fifth site is proposed to be the recognition site for a Na+-dependent acidic amino acid membrane transport process. An evaluation of investigations designed to elucidate regulatory mechanisms at acidic amino acid binding sites is made; hypotheses such as the Ca2+-activated protease hypothesis of long-term potentiation are assessed in terms of the new binding site/receptor classification scheme, and experiments are suggested which will clarify and expand this exciting area in the future.

Myelin protein composition in the rat spinal cord in culture and in vivo: a developmental comparison.

Biochemical characterization of the development of myelin in vitro was extended to an analysis of myelin protein composition in cultures of explanted foetal rat spinal cord. Myelin fractions were isolated from pooled explants after 12-30 days in vitro and, for comparison, from the spinal cords of rats of equivalent developmental ages. Electron microscopic examination of the culture myelin fractions revealed the presence of multilamellar myelin fragments and some single membranes. All fractions were analyzed using a micro-polyacrylamide gel electrophoresis technique. Qualitatively similar protein profiles were observed for myelin isolated from either cultures or from spinal cords. Fractions from cultures contained a greater proportion of high molecular weight proteins than those from spinal cords, although with respect to the 'major' myelin proteins, a quantitatively similar developmental pattern was observed both in vivo and in vitro.

Descending fibre-mediated release of endogenous glutamate and glycine from the perfused cat spinal cord in vivo.

The release of endogenous amino acids into the perfused central canal of the cat spinal cord in vivo was studied using a [3H]dansyl chloride assay procedure. Perfusion of 0.1 mM p-chloromercuriphenyl-sulphonate (pCMS), an uptake inhibitor, augmented the efflux of most amino acids studied. The large increases in glycine, glutamate and GABA efflux (9.9, 5.3 and 4.1-fold, respectively) suggest that the membrane fluxes of these amino acids were particularly high, a finding consistent with their putative neurotransmitter functions. Subsequent electrical stimulation of descending spinal tracts increased the efflux of glycine, glutamate, GABA and leucine (1.8, 2.2, 1.9 and 1.3-fold, respectively), although only the release of glycine and glutamate was clearly independent of the concomitant rise in blood pressure. The evoked amino acid release, which was accompanied by a substantial ventral root discharge, was not observed in the absence of pCMS, nor in the presence of a low concentration (0.01 mM) of this compound. These results have been discussed in relation to the putative neurotransmitter roles for amine acids in the spinal cord.

Chloride ions enhance L-glutamate binding to rat brain synaptic membranes.

Chloride ions increased L-glutamate (L-Glu) binding to synaptic membranes. The binding was saturable and resulted in a 2.5-fold enhancement at concentrations of 20--40 mM chloride. Sodium and potassium ions inhibited only chloride stimulated L-Glu binding. Calcium ions also increased L-Glu binding but this was observed only in the presence of chloride. The anion selectivity of the enhancement of L-Glu binding was similar to that reported for the membrane chloride channel, suggesting that some L-Glu binding sites may be associated with this channel.

Regulation of glutamate receptors: possible role of phosphatidylserine.

The polar head group of the phospholipid phosphatidylserine is similar in structure to the glutamate analogue 2-amino-4-phosphonobutyric acid (APB), an antagonist of excitatory transmission in the brain. When tested in ligand binding assays phosphatidylserine and its polar head group components O-phosphoserine and L-alpha-glycerophosphoserine were good displacers of APB-sensitive L-glutamate binding. The polar head group components were also antagonists of synaptic field potentials in the rat dentate gyrus evoked by stimulation of a presumed glutamate-using pathway. These results suggest that phosphatidylserine may regulate the activity of synaptic L-glutamate receptors.

L-glutamate and L-aspartate bind to separate sites in rat brain synaptic membranes.

The specific binding of L-glutamate (L-Glu) and L-aspartate (L-Asp) was measured in rat brain synaptic plasma membranes (SPMs). A distinction between the binding sites for these amino acids was made on the basis of the kinetics, ion effects, pharmacology and chemical susceptibility of the binding. The existence of distinct binding sites for L-Glu and L-Asp is consistent with electrophysiological data that mammalian neurons possess separate receptors for these amino acids.

Distribution of PNS myelin proteins and membrane enzymes in fractions isolated by continuous gradient zonal centrifugation.

Myelin was purified from adult rabbit sciatic nerve by two procedures: discontinuous gradient centrifugation and continuous gradient zonal centrifugation. Two fractions were obtained from the discontinuous gradient. The fraction floating on 0.32 M sucrose and the fraction recovered from the 0.32/0.85 M sucrose interface showed typical myelin membranes by electron microscopy and typical myelin proteins by gel electrophoresis. The specific activity of 2',3'-cyclic nucleotide 3'-phosphodiesterase (CNP) decreased from the top to the bottom of the discontinuous gradient. The myelin separated by zonal centrifugation on a continuous sucrose gradient showed three distinct peaks (on monitoring optical density) at 0.10, 0.30 and 0.57 M sucrose. The latter peak yielded 92% of the material applied. The two minor peaks of low density exhibited high CNP and acetylcholinesterase (AChE) activities but the specific activity of both enzymes increased markedly at the heavy end of the gradient. The zonal fractions showed typical myelin proteins in all fractions by polyacrylamide gel electrophoresis but with important quantitative differences. These results indicate that PNS myelin shows significant heterogeneity.

Comparison of L-[3H]glutamate, D-[3H]aspartate, DL-[3H]AP5 and [3H]NMDA as ligands for NMDA receptors in crude postsynaptic densities from rat brain.

L-[3H]Glutamate, D-[3H]aspartate, DL-[3H]2-amino-5-phosphonovalerate (AP5) and [3H]N-methyl-D-aspartate (NMDA) were evaluated as radioligands using postsynaptic density (PSD) preparations from rat brain. L[3H]Glutamate had the highest affinity and greatest percentage specific binding, followed by D-[3H]aspartate greater than DL-[3H]AP5 greater than [3H]NMDA. The pharmacological specificity of the binding of each radioligand indicated an interaction with NMDA-preferring receptors, the order of potency of displacing compounds tested being L-glutamate greater than D-aspartate greater than D-AP5 greater than DL-AP5 greater than ibotenate greater than NMDLA greater than quisqualate. For L-[3H]glutamate, the data revealed an interaction with two sites, the major one having NMDA receptor characteristics and the minor one resembling the quisqualate-preferring receptor. Against L-[3H]glutamate binding, quisqualate showed a two-component inhibition profile with an affinity of 25 microM at the NMDA site and 0.19 microM at the quisqualate site. Thus, by using several radioligands possessing activity at NMDA receptors, we confirm that an NMDA receptor binding site is present in crude PSDs. Although it is less selective than D-[3H]aspartate, DL-[3H]AP5 and [3H]NMDA, L-[3H]glutamate remains, by virtue of its high affinity, the ligand of choice for the study of NMDA receptors in preparations where such sites predominate.

Electrophysiological characterization of a novel potent and orally active NMDA receptor antagonist: CGP 37849 and its ethylester CGP 39551.

The selectivity and potency of the novel competitive N-methyl-D-aspartate (NMDA) receptor antagonists, CGP 37849 and CGP 39551, were investigated in vitro and in vivo using electrophysiological approaches. Like the reference blocker DL-AP5, both compounds acted in vitro (hippocampus, substantia nigra, spinal cord) to antagonize the excitatory actions of exogenously administered NMDA as well as the synaptically elicited, physiological NMDA receptor responses in hippocampus and spinal cord. In all isolated preparations CGP 37849 was more potent than CGP 39551, and 5- to 10-fold more potent than DL-AP5. Neither compound showed any marked effect on responses evoked by quisqualate and kainate. NMDA excited dopaminergic cells in the pars compacta region of the substantia nigra in a concentration-dependent manner. This effect also could be selectively antagonized by CGP 37849 and CGP 39551. In the anaesthetized rat, excitatory responses of hippocampal pyramidal cells evoked by iontophoretic application of NMDA were antagonized by CGP 37849 and CGP 39551 following their oral administration without reducing quisqualate or kainate responses. In contrast to the in vitro situation, CGP 39551 was more potent than CGP 37849 in vivo. Effective doses were 30 mg/kg p.o. for CGP 39551 and 100 mg/kg p.o. for CGP 37849. In conclusion, it is demonstrated that CGP 37849 and CGP 39551 selectively antagonize NMDA evoked neuronal responses in vivo and in vitro and that the drugs are centrally active following their oral administration.

Modulation of the NMDA receptor by D-serine in the cortex and the spinal cord, in vitro.

We present a comparative study of the modulation of the N-methyl-D-aspartate (NMDA) receptor at the strychnine-insensitive glycine site in the spinal cord and in the cortex. The excitatory effect of NMDA was potentiated by D-serine (a glycine mimetic) in the hemisected rat spinal cord. The non-competitive NMDA antagonists 7-chlorokynurenic acid (7-Cl KYNA; 10 microM) and 3-amino-1-hydroxypyrrolid-2-one (HA-966; 100 or 200 microM) antagonized the effect of NMDA in the spinal cord and cortical wedge preparation. The antagonism was reversed by the addition of D-serine. This effect was strychnine-insensitive and hence not related to the inhibitory glycine receptor known to be present in the spinal cord. Our results suggest strongly that glycine positively modulates the NMDA system not only at a supraspinal level but also at the spinal level. As the positive modulation of NMDA responses by D-serine was also seen in the presence of tetrodotoxin, we conclude that the NMDA/glycine complex is (also) located on motoneurones in addition to the known glycine-mediated inhibitory system.

Chloride and calcium ions separate L-glutamate receptor populations in synaptic membranes.

Cl-/Ca2+-dependent and Cl-/Ca2+-independent L-[3H]glutamate binding sites in rat brain synaptic membranes showed marked differences in their pharmacological properties. One site resembled L-2-amino-4-phosphonobutyrate (L-APB)-sensitive receptors and the other N-methyl-D-aspartate (NMDA) receptors. Inhibition studies demonstrated that L-aspartate was more potent at Cl-/Ca2+-independent than at Cl-/Ca2+-dependent sites although L-glutamate was of similar potency at both sites; the D-isomers of aspartate, glutamate and alpha-aminoadipate exhibited the opposite trend. Quisqualate and ibotenate showed high and low affinity inhibition components in the presence of Cl- and Ca2+, and only low affinity inhibition at Cl-/Ca2+-independent sites. For a series of alpha-amino-omega-phosphono carboxylic acids (propionate-heptanoate), peaks of inhibitory activity in the presence of Cl- and Ca2+ were shifted to l-carbon shorter homologues than in the absence of these ions. These data indicate that the ionic environment is of critical importance for the activity of different physiological receptor populations in vitro.

CGP 35348: a centrally active blocker of GABAB receptors.

The biochemical, electrophysiological and pharmacological properties of the new GABAB receptor blocker CGP 35348 are described. In a variety of receptor binding assays CGP 35348 showed affinity for the GABAB receptor only. CGP 35348 had an IC50 of 34 microM at the GABAB receptor. The compound antagonized (100, 300, 1000 microM) the potentiating effect of L-baclofen on noradrenaline-induced stimulation of adenylate cyclase in rat cortex slices. In electrophysiological studies CGP 35348 (10, 100 microM) antagonized the effect of L-baclofen in the isolated rat spinal cord. In the hippocampal slice preparation CGP 35348 (10, 30, 100 microM) blocked the membrane hyperpolarization induced by D/L-baclofen (10 microM) and the late inhibitory postsynaptic potential. CGP 35348 appeared to be 10-30 times more potent than the GABAB receptor blocker phaclofen. Ionophoretic and behavioural experiments showed that GABAB receptors in the brain were blocked after i.p. administration of CGP 35348. This compound may be of considerable value in elucidating the roles of brain GABAB receptors.

[3H]CGP 39653: a new N-methyl-D-aspartate antagonist radioligand with low nanomolar affinity in rat brain.

CGP 39653 (D,L-(E)-2-amino-4-propyl-5-phosphono-3-pentenoic acid) was initially discovered to inhibit the binding of [3H]L-glutamate and [3H]3-[+/-)2-carboxypiperazin-4-yl)-propyl-1- phosphonic acid [( 3H]CPP) with Ki values of 230 and 5 nM, respectively. The radiolabeled compound [3H]CGP 39653 binds to rat frontal cortical membranes in a saturable and reversible manner. Analysis of saturation experiments revealed that the ligand labels one binding site with a Kd value of 6 nM. Competition experiments indicated that the order of potency of a number of competitive excitatory amino acid agonist and antagonist compounds was similar to that found previously for other N-methyl-D-aspartate (NMDA) receptor ligands. In contrast to these competitive inhibitors, which produced steep inhibition curves, glycine inhibited binding in a complex manner. When the functional activity of the unlabeled compound was explored, CGP 39653 blocked NMDA-evoked depolarizations in the rat cortical wedge in vitro and inhibited L-glutamate stimulated [3H]N(1-[2-thienyl]cyclohexyl)3,4-piperidine [( 3H]TCP) binding in cortical membranes. These results suggest that [3H]CGP 39653 selectively binds to the NMDA receptor as an antagonist with high affinity and is currently the ligand of choice for labeling the NMDA receptor.