A family of glutamate receptor genes: evidence for the formation of heteromultimeric receptors with distinct channel properties

Thiocyanate ions inhibit AMPA-activated currents in recombinant non-NMDA receptors expressed in Xenopus laevis oocytes: the role of the GluR2 subunit

The functional interaction of thiocyanate (SCN-) ions with recombinant non-N-methyl-D-aspartate receptors was examined by studying alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)- and kainic acid (KA)-activated currents in Xenopus laevis oocytes. Recombinant receptors were expressed after microinjection of oocytes with combinations of cDNAs or cRNAs encoding for GluR1, GluR2, GluR2(R586Q), GluR3 or GluR6 subunits. When the GluR2 subunit was expressed with GluR1, SCN- (2 mM) inhibited the responses to 50 microM AMPA, whereas responses to 100 microM KA were slightly increased and responses to 200 microM L-glutamate were unaffected. Equilibrium concentration-response curves for AMPA were antagonized in a non-competitive manner by SCN- with a reduction in the EC50. The inhibitory effects of SCN- were unaffected by prior reduction of receptor desensitization with either 10 microM Concanavalin-A or 0.5 mM diazoxide. AMPA-activated currents recorded from homomeric GluR1 or Glur3 receptors were not affected by SCN-, and GluR6 homomeric receptors, which are sensitive to KA but not to AMPA, were also unaffected. In contrast, AMPA activation of homomeric GluR2(R586Q) subunit receptors, or combinations of GluR1 or GluR3 + GluR2(R586Q) subunits, were markedly inhibited by SCN-. In addition, the inhibitory effect of AMPA on KA-activated responses on these heteromeric receptors, was enhanced by SCN-. These results indicate that SCN- exert an inhibitory effect on 'AMPA receptors' but only when the recombinant non-NMDA receptor is a GluR2 homomer, or when GluR2 subunits are present as part of a heteromeric combination. Moreover, this inhibitory effect was unaffected by the 'Q/R' site in the presumed second transmembrane domain, since currents mediated by the GluR2(R586Q) subunit were also susceptible to inhibition by SCN-. Thus the inhibition was not related to the rectification properties or calcium permeability of the non-NMDA receptors. It is suggested that the GluR2 subunit may have a specific binding site for anions which could modulate the function of non-NMDA receptors and that SCN- may be a useful probe for the detection of these subunits in native neurons.

Organization of AMPA receptor subunits at a glutamate synapse: a quantitative immunogold analysis of hair cell synapses in the rat organ of Corti

Sensitive and high-resolution immunocytochemical procedures were used to investigate the spatial organization of AMPA receptor subunits (GluR1-4) at the synapse between the inner hair cells and the afferent dendrites in the rat organ of Corti. This is a synapse with special functional properties and with a presynaptic dense body that defines the center of the synapse and facilitates its morphometric analysis. A quantitative postembedding immunocytochemical analysis was performed on specimens that had been embedded in a metachrylate resin at low temperature after freeze substitution. Single- and double-labeling procedures indicated that GluR2/3 and GluR4 subunits were colocalized throughout the postsynaptic density, with a maximum distance of 300 nm from the presynaptic body and with higher concentrations peripherally than centrally. No receptor immunolabeling was found at extrasynaptic membranes, but some GluR4 subunits appeared to be expressed presynaptically. The synapses between outer hair cells and afferent dendrites were devoid of labeling. The present data indicate that AMPA receptor subunits are inserted into the postsynaptic membrane in a very precise manner and that their density increases on moving away from the center of the synapse.

A light- and electron-microscopic study of GluR4-positive cells in cerebral cortex, subcortical white matter and corpus callosum of neonatal, immature and adult rats

The distribution of the [3H]alpha-amino-3-hydroxy-5-methylisoxzalepropionic acid (AMPA) receptor subunit GluR4 was studied in frontal, parietal and temporal cerebral cortex, subcortical white matter and corpus callosum of neonatal, immature and mature rats. In 1- to 2-day-old rats, a few oligodendrocyte progenitors and amoeboid microglia in the supraventricular part of the corpus callosum were immunolabelled for GluR4. At 7 to 10 days, the number of amoeboid microglia and oligodendrocyte progenitors in white matter increased; many neurons in cortex, including pyramidal neurons, were also moderately labelled for GluR4. The pattern of GluR4 immunostaining in 14-day-old rats was different from that in 7- to 10-day-old rats, but similar to the adult, in that there was no immunoreactivity in microglia and oligodendrocyte progenitors in subcortical white matter. A proportion of non-pyramidal neurons in cortex were moderately labelled, while some pyramidal neurons were lightly labelled. A population of small glial cells with features of oligodendrocyte progenitors were densely labelled in cortex.

Expression of AMPA-glutamate receptor B subunit in rat hippocampal GABAergic neurons

The physiological properties of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-type glutamate receptor (GluR) channels are determined by their subunit composition. In particular, the expression of the GluR-B subunit determines the divalent ion selectivity of the channel. We studied the distribution of GluR-B mRNA and protein in adult rat hippocampal GABAergic neurons combining non-radioactive in situ hybridization with immunocytochemistry. The majority of GABAergic hippocampal neurons were GluR-B mRNA-positive, but overall the levels of GluR-B transcript were lower than in pyramidal and granule cells, which showed the highest hybridization signal. The different GluR-B mRNA expression levels in GABAergic versus principal neurons were also observed at the protein level. There was a paucity of GluR-B/C immunoreactivity in the vast majority of somata of the GABAergic neurons studied, which contrasted with the strong expression of GluR-B/C proteins in the hippocampal principal neurons. Our results demonstrate the general low expression of the GluR-B subunit mRNA and GluR-B/C proteins in GABAergic hippocampal neurons. Considering the dominant role of the GluR-B subunit in determining the divalent ion permeability of the receptor, it is likely that most GABAergic neurons express AMPA receptor channels with different calcium permeabilities.

Different characteristics of AMPA receptor agonists acting at AMPA receptors expressed in Xenopus oocytes

A series of (RS)-2-amino-3-(3-hydroxy-5-methyl-4-isoxazolyl)propionic acid (AMPA) analogues were evaluated for activity at homo-oligomeric glutamate1-flop (Glu1-flop) receptors expressed in Xenopus oocytes, using the two-electrode voltage clamp technique. (RS)-2-Amino-3-(3-carboxy-5-methyl-4-isoxazolyl)propionic acid (ACPA) (EC50, 2.4 microM), a homologue of AMPA having a carboxyl group as the terminal acidic functionality, was five times more potent than AMPA (EC50, 12 microM) and 20 times more potent than kainate (EC50, 46 microM). (RS)-2-Amino-3(3-hydroxy-5-trifluoromethyl-4-isoxazolyl)propionic acid (Tri-F-AMPA), in which an electronegative trifluoromethyl group is substituted for the methyl group on the isoxazole ring in the AMPA structure, was three times more potent than AMPA, whereas (RS)-3-hydroxy-4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridine-5-carboxylic acid (5-HPCA), a bicyclic analogue of AMPA with highly restricted conformational flexibility was 10 times less potent than AMPA. The limiting slope of log-log plots of Glu1-flop receptor currents versus low agonist concentrations had a value of 1.7 for ACPA and kainate compared to 1.5 for Tri-F-AMPA and 1.3 for 5-HPCA and AMPA. The amplitude of responses evoked by near saturating concentrations of the agonists varied more than 7-fold. The sequence of efficacy was ACPA = kainate > Tri-F-AMPA > AMPA > 5-HPCA. Moreover, when saturating concentrations of Tri-F-AMPA and kainate were co-applied, the response was significantly greater than when each of the agonists was applied separately. The potency of the antagonist 2,3-dihydroxy-6-nitro-7-sulfamoylbenzo(f)quinoxaline (NBQX) (estimated KB, approximately 200 nM), to block currents mediated by Glu1-flop receptors was similar for all of the agonists tested in this study. These results indicate that relatively minor changes in the molecular structure of AMPA are associated with marked effects on potency and efficacy. In particular, it is suggested that the acidity of the terminal group plays a major role in determining the degree of receptor activation in the steady state.

Organization of AMPA receptor subunits at a glutamate synapse: a quantitative immunogold analysis of hair cell synapses in the rat organ of Corti

Sensitive and high-resolution immunocytochemical procedures were used to investigate the spatial organization of AMPA receptor subunits (GluR1-4) at the synapse between the inner hair cells and the afferent dendrites in the rat organ of Corti. This is a synapse with special functional properties and with a presynaptic dense body that defines the center of the synapse and facilitates its morphometric analysis. A quantitative postembedding immunocytochemical analysis was performed on specimens that had been embedded in a metachrylate resin at low temperature after freeze substitution. Single- and double-labeling procedures indicated that GluR2/3 and GluR4 subunits were colocalized throughout the postsynaptic density, with a maximum distance of 300 nm from the presynaptic body and with higher concentrations peripherally than centrally. No receptor immunolabeling was found at extrasynaptic membranes, but some GluR4 subunits appeared to be expressed presynaptically. The synapses between outer hair cells and afferent dendrites were devoid of labeling. The present data indicate that AMPA receptor subunits are inserted into the postsynaptic membrane in a very precise manner and that their density increases on moving away from the center of the synapse.

Distribution of the N-methyl-D-aspartate glutamate receptor subunit NR2A in control and amyotrophic lateral sclerosis spinal cord

The distribution of the different glutamate receptor subunits in human spinal cord has yet to be fully elucidated. The aim of this study was to examine the distribution of the N-methyl-D-aspartate (NMDA) glutamate receptor modulatory subunit NR2A, in control human spinal cord and to examine in parallel the expression of the mRNA in amyotrophic lateral sclerosis (ALS). The aetiology of ALS is poorly understood, although abnormalities in glutamate and glycine transport have been reported as well as alterations in NMDA receptors including the NR1 subunit; suggesting a role for glutamate in the disease process. We have used the technique of in situ hybridisation to localise this receptor subunit to the laminae of human spinal cord and have found that it shows a widespread distribution similar to that previously reported for the universal NMDA receptor subunit NR1. Quantitation of mRNA expression in control and ALS cases showed a significant widespread loss of NR2A from both dorsal and ventral horns with losses of 55% and 78%, respectively, in ALS as compared to control. These results were substantiated by analysis of spinal cord homogenates, which showed a significant total decrease of 50% in ALS spinal cord as compared to control.

Characterization of the ligand-binding domains of glutamate receptor (GluR)-B and GluR-D subunits expressed in Escherichia coli as periplasmic proteins

We recently reported that a functional ligand-binding site of an alpha-amino-5-methyl-3-hydroxy-4-isoxazole propionate (AMPA)-selective glutamate receptor (GluR)-D subunit can be expressed in insect cells as a soluble, N-glycosylated fusion protein consisting of two segments (S1 and S2) that are related by amino acid sequence to bacterial periplasmic binding proteins (Kuusinen, A., Arvola, M., and Keinänen, K., EMBO J. 14, 6327-6332). In an attempt to further characterize the structural determinants for ligand binding, we have now expressed the ligand-binding sites of GluR-B and GluR-D subunits in Escherichia coli as soluble periplasmic proteins. The bacterially expressed S1-S2 fusion proteins bound [3H]AMPA with a high affinity (Kd of 12 nM for GluR-B, Kd of 60 nM for GluR-D) and with a ligand pharmacology typical of native AMPA receptors, indicating that N-linked glycosylation is not required for the formation or the maintenance of the ligand-binding site. The flip and flop splice variants of the GluR-D S1-S2 fusion protein bound [3H]AMPA with equal affinities, whereas deletion of the C-terminal one-third of the S2 segment including the flip/flop sequence resulted in a loss of binding activity. Our results highlight the potential of bacterial expression for the analysis of the binding site and support a close structural similarity between glutamate receptors and bacterial proteins.

Structure and pharmacological properties of a molluscan glutamate-gated cation channel and its likely role in feeding behavior

We describe the isolation of a molluscan (Lymnaea stagnalis) full-length complementary DNA that encodes a mature polypeptide (which we have named Lym-eGluR2) with a predicted molecular weight of 105 kDa that exhibits 44-48% identity to the mammalian kainate-selective glutamate receptor GluR5, GluR6, and GluR7 subunits. Injection of in vitro-transcribed RNA from this clone into Xenopus laevis oocytes results in the robust expression of homo-oligomeric cation channels that can be gated by L-glutamate (EC50 = 1.2 +/- 0.3 micron) and several other glutamate receptor agonists; rank order of potency: glutamate >> kainate > ibotenate > AMPA. These currents can be blocked by the mammalian non-NMDA receptor antagonists 6,7-dinitroquinoxaline-2,3-dione, 6-cyano-7-nitroquinoxaline-2,3-dione, and 1-(4-chlorobenzoyl)piperazine-2,3-dicarboxylic acid. Ionic-replacement experiments have shown that the agonist-induced current is carried entirely by sodium and potassium ions. In situ hybridization has revealed that the Lym-eGluR2 transcript is present in all 11 ganglia of the Lymnaea CNS, including the 4-cluster motorneurons within the paired buccal ganglia. The pharmacological properties and deduced location of Lym-eGluR2 are entirely consistent with it being (a component of) the receptor, which has been identified previously on buccal motorneurons, that mediates the excitatory effects of glutamate released from neurons within the feeding central pattern generator.

The glycine binding site of the N-methyl-D-aspartate receptor subunit NR1: identification of novel determinants of co-agonist potentiation in the extracellular M3-M4 loop region

The N-methyl-D-aspartate (NMDA) subtype of ionotropic glutamate receptors is a heterooligomeric membrane protein composed of homologous subunits. Here, the contribution of the M3-M4 loop of the NR1 subunit to the binding of glutamate and the co-agonist glycine was investigated by site-directed mutagenesis. Substitution of the phenylalanine residues at positions 735 or 736 of the M3-M4 loop produced a 15- to 30-fold reduction in apparent glycine affinity without affecting the binding of glutamate and the competitive glycine antagonist 7-chlorokynurenic acid; mutation of both residues caused a >100-fold decrease in glycine affinity. These residues are found in a C-terminal region of the M3-M4 loop that shows significant sequence similarity to bacterial amino acid-binding proteins. Epitope tagging revealed both the N-terminus and the M3-M4 loop to be exposed extracellularly, whereas a C-terminal epitope was localized intracellularly. These results indicate that the M3-M4 loop is part of the ligand-binding pocket of the NR1 subunit and provide the basis for a refined model of the glycine-binding site of the NMDA receptor.

Patch-clamp study on membrane properties and transmitter activated currents of rabbit area postrema neurons

Using the patch-clamp technique in combination wit sliced tissue preparation the membrane properties of newborn rabbit area postrema neurons were investigated. The neurons responded upon depolarization with a fast Na+-current followed by an inactivating and non-inactivating K+-current. GABA-activated currents were investigated resulting in a large C1-(-)conductance, indicating the expression of GABAA-receptors. The expression of glutamate receptor mRNA was studied by in situ hybridization and electrophysiological measurements of these receptors by means of the patch-clamp technique. As a main result it was found that ionotropic glutamate receptors in the area postrema are composed of "flop" variants of the GluA-, GluB- and GluC-subunits.

Cytotoxic effects of kainate ligands on HEK cell lines expressing recombinant kainate receptors

Exposure of neurons either for prolonged periods of time or to high concentrations of excitatory amino acids (EAA), such as glutamate, results in neuronal death. Kainate also causes cell toxicity through the glutamate receptors. However, it is unclear whether the kainate receptor itself mediates any of the toxic responses. In the present study, HEK cells expressing the GluR6 +/- KA2 receptor subunit(s) were studied for their susceptibility to toxicity through the kainate receptor by kainate ligands. The natural ligand, glutamate, did not result in toxicity to the recombinant cell lines over that observed with the untransfected HEK cells, whereas kainate produced a 2-3-fold increase in LDH in both the HEK/GluR6 (ANOVA, P = 0.0001) and HEK/GluR6 + KA2 (ANOVA, P = 0.0002) cell lines following treatment with various dosages, but did not affect the HEK cells. Similar 2-3-fold increases in LDH activity were detected in both recombinant cell lines following treatment with 100 nM of SYM2081 ((2S,4R)-4-methylglutamic acid), a dose at which agonistic activity is elicited. The rank order potencies for eliciting toxicity are consistent with the previously reported EC50 values (SYM2081 > kainate > > > glutamate). Surprisingly, the kainate antagonist, NBQX, was the most toxic of the compounds tested although it had an affinity for the kainate receptor similar to glutamate. Treatment with as little as 10 nM elicited a dramatic increase in toxicity (6-10-fold) in the recombinant cell lines. At 1 microM, NBQX was significantly more toxic (Fisher PLSD, P < 0.05) than any of the other compounds tested. Thus, it appears that cell toxicity can be mediated via kainate receptor through two independent mechanisms: activation and blockage of the kainate receptor.

Co-expression of receptor mRNA and protein: striatal dopamine and excitatory amino acid subtypes

Dopamine (DA) is known to modulate the post-synaptic response of the excitatory amino acid (EAA) neurotransmitters in the striatum. Thus the intrinsic neurons in this nucleus are potential sites of cross-interaction between these two systems. The recent isolation of 5 different DA receptor subtypes and more than 20 EAA subunits argues for a complicated functional role for the protein products encoded by these transcripts. The simultaneous detection of cellular mRNA distributions and translated protein products was an initial step to determine differences in post-translational expression at the cellular level of resolution for two of these receptors. The cloned D2 DA receptor subtype and the ionotropic GluR1 EAA receptor subunit were examined by fluorescence in situ transcription (FIST) following hybridization of specific cDNA primers, complementary to the mRNA transcripts encoding these receptors. Nascent extension of the annealed primer using reverse transcriptase was detected after incorporation of fluorescently labeled dUTP. Protein products were visualized by standard immunofluorescence after incubation with anti-peptide antisera that were selective for each receptor protein. The experimental data corroborate previous work describing the regional expression of ligand binding and in situ hybridization detected with radiolabeled probes for the DA and EAA receptor systems in the striatum. The dual fluorescence method can be completed within 2 days and may be adapted to cellular localization of many novel mRNA/protein combinations to examine post-translational processing within thin tissue slices.

Origin of aspartate-induced responses in rat cerebellar Purkinje cells

The effects of glutamate, aspartate and N-methyl-D-aspartate (NMDA) on Purkinje cells and interneurons were investigated in cerebellar slice cultures using the whole-cell configuration of the patch-clamp technique. L-Glutamate and L-aspartate induced inward currents in Purkinje cells voltage-clamped at -60 mV. In standard external solution, the amplitude of the responses induced by the two amino-acids was a linear function of the membrane potential. L-Aspartate-induced currents were inhibited by 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), a selective antagonist of non-NMDA receptors. NMDA, a selective agonist of NMDA receptors, had no effect of its own on the excitability of Purkinje cells, but was effective in blocking the responses induced by aspartate in Purkinje cells in a voltage-independent manner. In contrast, D-(-)-2-amino-5-phosphonovaleric acid (D-APV), a selective antagonist of NMDA receptors, had no effect on aspartate-induced responses. D-Aspartate also induced responses in Purkinje cells, and the amplitude of these responses was a linear function of the membrane potential. Currents induced by L- and D-aspartate were inhibited by dihydrokainate, a glutamate uptake blocker. In sodium-free external solution, glutamate still induced outward currents in Purkinje cells, whereas L- and D-aspartate no longer evoked any current. When sodium was replaced by lithium in the external medium, no change in the holding current could be detected in Purkinje cells maintained at -60 mV; moreover, in this bathing medium L-aspartate no longer evoked any current whereas glutamate-induced responses were still present. In contrast, interneurons were sensitive to both NMDA and aspartate applications, and these responses were antagonized by D-APV. In addition, aspartate still induced an outward current in sodium-free external solution. This study presents rather direct evidence in favour of L-aspartate as being a very selective NMDA receptor agonist in the cerebellum. L-Aspartate-induced currents in Purkinje cells are not due to activation of mixed NMDA/non-NMDA receptors, but are probably due to the release of L-glutamate induced by aspartate through glutamate uptake.

An immunocytochemical study of glutamate receptors and glutamine synthetase in the hippocampus of rats injected with kainate

Immunocytochemistry was used to study the distribution of the kainate receptors GluR1, GluR2/3 and GluR4 and of the N-methyl-D-aspartate (NMDA) receptor NMDAR1 as well as the astrocyte markers glutamine synthetase (GS) and glial fibrillary acidic protein (GFAP) in the hippocampus of normal and kainate-lesioned rats. Hippocampal pyramidal neurons and dentate granule neurons were labelled heavily for GluR1 and GluR2/3, but only lightly for GluR4. Dense GluR4 immunopositivity was, however, observed in oligodendrocyte-like glial cells. Hippocampal pyramidal neurons and dentate granule neurons were moderately labelled for NMDAR1. Intravenous kainate injections resulted in a decrease in GluR1 and GluR2/3 immunoreactivity on the apical dendrites of pyramidal neurons as early as 7 h postinjection. At 18 h, there was a marked reduction in GluR1 and GluR2/3 receptors in the terminal tuft of dendrites of most hippocampal pyramidal neurons in the affected area, although some cells showed labelling in other portions of the apical dendrites and in basal dendrites. Immunostaining for GluR4 and NMDAR1 was also reduced at this time. At postinjection day 3, only the cell bodies and the basal dendrites of a few scattered pyramidal cells were labelled. Taken together, these results indicate a progressive loss of glutamate receptors, which affects the apical dendritic tree before the basal dendritic tree. The decrease in receptor immunoreactivity could be due to a downregulation of the receptors, since it occurred as early as 7 h postlesion, before cell death was evident in Nissl-stained sections. At long intervals after kainate injection, all pyramidal cells at the centre of the lesion showed a lack of glutamate receptor staining, and no partially labelled pyramidal cells were observed. The periphery of the lesion, however, contained many partially labelled pyramidal neurons among the unlabelled cells and had features of early lesions. The present study also showed an early decrease in GS immunoreactivity in the affected CA fields of the hippocampus (18 h to 3 days postinjection), followed by a medium-term increase (5-68 days) and a late decrease in GS immunoreactivity (81 days). The decrease in GS immunoreactivity at 81 days is not due to an absence of astrocytes, since GFAP staining showed many densely labelled astrocytes in the affected CA field.

Structure and pharmacological properties of a molluscan glutamate-gated cation channel and its likely role in feeding behavior

We describe the isolation of a molluscan (Lymnaea stagnalis) full-length complementary DNA that encodes a mature polypeptide (which we have named Lym-eGluR2) with a predicted molecular weight of 105 kDa that exhibits 44-48% identity to the mammalian kainate-selective glutamate receptor GluR5, GluR6, and GluR7 subunits. Injection of in vitro-transcribed RNA from this clone into Xenopus laevis oocytes results in the robust expression of homo-oligomeric cation channels that can be gated by L-glutamate (EC50 = 1.2 +/- 0.3 micron) and several other glutamate receptor agonists; rank order of potency: glutamate >> kainate > ibotenate > AMPA. These currents can be blocked by the mammalian non-NMDA receptor antagonists 6,7-dinitroquinoxaline-2,3-dione, 6-cyano-7-nitroquinoxaline-2,3-dione, and 1-(4-chlorobenzoyl)piperazine-2,3-dicarboxylic acid. Ionic-replacement experiments have shown that the agonist-induced current is carried entirely by sodium and potassium ions. In situ hybridization has revealed that the Lym-eGluR2 transcript is present in all 11 ganglia of the Lymnaea CNS, including the 4-cluster motorneurons within the paired buccal ganglia. The pharmacological properties and deduced location of Lym-eGluR2 are entirely consistent with it being (a component of) the receptor, which has been identified previously on buccal motorneurons, that mediates the excitatory effects of glutamate released from neurons within the feeding central pattern generator.

Response to repetitive stimulation of AMPA receptors in patches excised from fields CA1 and CA3 of the hippocampus

Patches excised from fields CA1 and CA3 of the hippocampus were used to study AMPA receptor responses to repetitive application of brief pulses of agonist. AMPA receptors in field CA1 exhibited a larger degree of paired-pulse depression and a slower recovery rate from desensitization than those from CA3 patches. Responses became progressively smaller during a train of short pulses (4 at 100 Hz), an effect that was greater in CA1 than CA3 patches and that was blocked by a drug (cyclothiazide) that slows desensitization. Cyclothiazide also increased steady state currents during a long pulse of glutamate and had a lower EC50 for CA3 than CA1 AMPA receptors. These results suggest (i) that differences in the relative balance of flip vs. flop subunits affect the kinetic and pharmacological properties of AMPA receptors in vivo, and (ii) that the rate of recovery from desensitization may influence responses to the bursts of afferent activity commonly used to induce long-term potentiation.

Distribution of NMDA and AMPA receptors in the cerebellar cortex of rhesus macaques

The distribution of glutamate receptors in the cerebellar cortex of the rhesus macaque was examined by light microscopic immunocytochemistry using an antibody specific to the N-methyl-D-aspartate (NMDA) R1 receptor subunit (i.e. NMDAR1) as well as antibodies specific to alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor subunits (i.e. GluR1, GluR2/3, and GluR4). NMDAR1 immunolabeling was most prevalent in the Purkinje cell perikarya and dendrities, but was also significant in the stellate and basket cells of the granular layer and Golgi cells of the molecular layer. On the other hand, GluRl and GluR4 immunolabeling was concentrated principally in the processes of the Bergmann glia located in the vicinity of the Purkinje cell perikarya. Although GluR2/3 immunolabeling also occurred in these Bergmann glia processes as well as in the Bergmann fibers, it was more pronounced in the Purkinje cell perikarya and dendrites; additionally, significant GluR2/3 labeling was evident in the stellate and basket cells of the molecular layer and medium-size soma of the granular layer (most likely Golgi cells). In situ hybridization histochemistry (ISHH), using cRNA probes to NMDAR1. GluR1.GluR2, and GluR3, showed glutamate receptor mRNA distribution patterns consistent with those disclosed in the immunocytochemical study. Furthermore, the ISHH findings suggest that the positive immunocytochemical labeling of Purkinje cells with the GluR2/3 antibody is most likely due to the gene expression of both GluR2 and GluR3 AMPA receptor subtypes. Taken together, the results are potentially important for the elucidation of mechanisms that control aspects of cerebellar function, such as long-term depression.

Three-dimensional models of non-NMDA glutamate receptors

Structural models have been produced for three types of non-NMDA inotropic glutamate receptors: an AMPA receptor, GluR1, a kainate receptor, GluR6; and a low-molecular-weight kainate receptor from goldfish, GFKAR alpha. Modeling was restricted to the domains of the proteins that bind the neurotransmitter glutamate and that form the ion channel. Model building combined homology modeling, distance geometry, molecular mechanics, interactive modeling, and known constraints. The models indicate new potential interactions in the extracellular domain between protein and agonists, and suggest that the transition from the "closed" to the "open" state involves the movement of a conserved positive residue away from, and two conserved negative residues into, the extracellular entrance to the pore upon binding. As a first approximation, the ion channel domain was modeled with a structure comprising a central antiparallel beta-barrel that partially crosses the membrane, and against which alpha-helices from each subunit are packed; a third alpha-helix packs against these two helices in each subunit. Much, but not all, of the available data were consistent with this structure. Modifying the beta-barrel to a loop-like topology produced a model consistent with available data.

Relative concentrations and seizure-induced changes in mRNAs encoding three AMPA receptor subunits in hippocampus and cortex

In situ hybridization was used to determine 1) the relative concentrations of mRNAs encoding different subunits of the alpha-amino 3-hydroxy-5-methyl-4- isoxazolepropionate receptor family in select regions of rat forebrain and 2) whether limbic seizures alter the balances of the subunit mRNAs. GluR1 and GluR2 mRNA levels were about equal and were much greater than GluR3 mRNA levels in the principal neurons of each hippocampal subdivision. Probable interneurons in hippocampal molecular layers had much higher levels of GluR1 mRNA than of either GluR2 or GluR3 mRNA. Pyramidal cell layers in neo- and paleocortex had a balance of mRNAs that was significantly different from the balance in hippocampus: GluR1 mRNA and GluR3 mRNA levels were about equal and were substantially lower than those of GluR2 mRNA. Lesion-induced limbic seizures caused transient changes in mRNA levels that were differentiated with regard to subunit and brain region. All three mRNAs were decreased in the pyramidal layers of cortex, and changes in hippocampal pyramidal cells were smaller. Seizure-induced changes in granule cells of the dentate gyrus differed from all other regions examined: GluR1 mRNA was reduced to a greater degree than GluR2 mRNA, whereas GluR3 mRNA content was markedly increased. These data strongly suggest that the subunit composition of alpha-amino 3-hydroxy-5-methyl-4-isoxazolepropionate receptors differs significantly between areas of the cortical telencephalon. Furthermore, the data indicate that aberrant patterns of physiological activity differentially influence the expression of subunit mRNAs in a region-specific and/or cell-type-specific manner.

CDNA cloning of chick brain alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors reveals conservation of structure, function and post-transcriptional processes with mammalian receptors

Several types of functional ionotropic glutamate receptor have been cloned in the recent years from the mammalian central nervous system, but till now, none from other vertebrate species. Here, we report the cloning and functional analysis of four chick brain cDNAs, coding for members of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor subtype of glutamate receptors. These receptors are highly homologous to the mammalian GluR1-4 (A-D) receptors ( > 90%), and conserve their post-transcriptional modifications. The flip/flop exons are conserved not only at the amino acid level but also at the nucleotide level, and the intron of GluR4 involved in the RNA editing of the R/G site displays a rat-chick sequence conservation of 95%. Significant sequence differences are found only in the region containing the immunogenic epitope of neuroactive anti-GluR3 antibodies. Chick AMPA receptors are expressed in both the cerebrum and cerebellum. The ion channel activities of chick GluR1-4 were analyzed in Xenopus oocytes and found to be similar to those of mammalian AMPA receptors. Though their contribution to kainate binding activity in the cerebellum is minor, the profile of channel activity of the chick GluR1-4 suggests that they account for the kainatergic channel activity expressed by total chick cerebellar mRNAs.

Cell surface expression of the human N-methyl-D-aspartate receptor subunit 1a requires the co-expression of the NR2A subunit in transfected cells

Cell surface expression of the NR1a subunit has been examined in mouse L cell lines permanently transfected with the complementary DNA for human NR1a or with the complementary DNAs for NR1a and NR2A. The expression of the subunits was under the control of the murine mammary tumour virus promoter and following induction of expression by dexamethazone both cell lines expressed high levels of the NR1a subunit as determined by immunofluorescence using permeabilized cells and immunoblotting of cell membranes with subunit specific antibodies. However, cell surface expression of the NR1a subunit was found only in the cells expressing both the NR1a and NR2A subunits. This was confirmed by cell surface biotinylation of the two cell lines and affinity isolation of the receptor subunits. To determine if this result was solely due to the use of a particular cell line and or the choice of expression vector, Cos-7 cells were transiently transfected with either NR1a or NR1a plus NR2A. Here too cell surface expression was only found following co-transfection of both subunits.

Functional and molecular analysis of glutamate-gated channels by patch-clamp and RT-PCR at the single cell level

In the central nervous system (CNS) rapid excitatory neurotransmission is mainly mediated by ligand gated, cationic channels activated by glutamate. Three main subtypes of glutamate-gated channels have been characterized by pharmacological studies. They have been named according to their preferred agonist, N-methyl-D-aspartate (NMDA), high affinity kainate and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA). Furthermore, a large diversity within each class of glutamate-gated channels has been revealed by the molecular cloning of multiple subunits and their spliced and edited variants (for review see Wisden and Seeburg, 1993). These subunits can potentially form different oligomeric complexes with diverging properties. A crucial question is therefore to determine the actual subunit composition of naturally occurring glutamate receptors. We have combined patch-clamp recording, reverse transcription (RT) and PCR to correlate, at the single cell level, the pattern of subunits expression with the functional properties of native glutamate receptors. We describe here results obtained on the AMPA receptors of hippocampal neurones and on the NMDA receptors of cerebellar granule cells which show that the subunit composition of these two types of receptors explains some of their functional properties. Furthermore, our data also indicate that the expression of NMDA receptor subunits during the postnatal development of cerebellar granule cells is regulated by an activity-dependent mechanism.

New developments in the molecular pharmacology of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate and kainate receptors

Separation of non-N-methyl-D-aspartate subtypes of glutamate receptors, known as alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA) and kainate receptors, is traced through conventional pharmacology to molecular biology. The physiology and pharmacology of recombinant receptor subtypes (GluR1-7 and KA1-2) are described. Competitive antagonists, e.g., the quinoxalinedione, 2,3-dihyroxy-6-nitro-7-sulphamoyl-benz(F)quinoxaline, and the decahydroisoquinoline, 3S,4aR,6R, 8aR-6-[2-(1(2)H-tetrazol-5-yl)ethyl]-decahydroisoquinolin e-3-carboxylate, have a broad antagonist spectrum, except that the latter is inactive on GluR6. The 2,3-benzodiazepines noncompetitively antagonise the AMPA receptor GluR1-4. Desensitisation of AMPA (GluR1-4) and kainate (GluR5-7 and KA1-2) receptors is blocked by cyclothiazide and concanavalin A, respectively. Polyamine toxins block AMPA receptors not containing GluR2 and unedited kainate receptors (GluR5-6). These data correlate well with results on native neurons characterised by techniques such as in situ hybridisation.

Differential localization of AMPA glutamate receptor subunits in the two segments of the globus pallidus and the substantia nigra pars reticulata in the squirrel monkey

The subthalamic nucleus has long been known as the main source of glutamatergic afferents to the pallidum and the substantia nigra in primates. Recent findings showed that the excitatory effects induced by the subthalamic nucleus in pallidal cells are mediated through the activation of non-NMDA receptors in the rat. The objective of the present study was to analyse the distribution of alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA) glutamate receptor subunits in the external pallidum (GPe), the internal pallidum (GPi) and the substantia nigra pars reticulata (SNr) in squirrel monkeys (Saimiri sciureus). This was achieved by means of immunohistochemistry using antibodies raised against the GluR1 and the GluR2/3 subunits of the AMPA receptor. Our results show that all neuronal perikarya in GPe and GPi display immunoreactivity for GluR2/3 subunits whereas GluR1 is confined exclusively to cells in the GPe. The proportion of GluR1-immunoreactive neurons is not uniform throughout the rostrocaudal extent of GPe; in the rostral third all GPe cells display GluR1 immunoreactivity, whereas in the caudal third the proportion of GluR1-positive cells decreases to 50%. The intensity of GluR1 immunostaining associated with GPe cells is lower than that associated with neighbouring large-sized neurons in the nucleus basalis of Meynert. In contrast to GPi cells, the neurons in the SNr display immunoreactivity for both GluR1 and GluR2/3 subunits. In conclusion, our results provide the first evidence for a different distribution of the GluR1 subunit of the AMPA receptors in the two segments of the globus pallidus in monkeys. These findings imply that the control of the basal activity of GPe and GPi cells by the subthalamic nucleus is exerted via the activation of AMPA receptors composed of different subunits. These data reinforce the view that the two segments of the globus pallidus are different entities that possess their own functional characteristics in primates.

High-resolution immunogold localization of AMPA type glutamate receptor subunits at synaptic and non-synaptic sites in rat hippocampus

The cellular and subcellular localization of the GluRA, GluRB/C and GluRD subunits of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) type glutamate receptor was determined in the rat hippocampus using polyclonal antipeptide antibodies in immunoperoxidase and immunogold procedures. For the localization of the GluRD subunit a new polyclonal antiserum was developed using the C-terminal sequence of the protein (residues 869-881), conjugated to carrier protein and absorbed to colloidal gold for immunization. The purified antibodies immunoprecipitated about 25% of 3[H]AMPA binding activity from the hippocampus, cerebellum or whole brain, but very little from neocortex. These antibodies did not precipitate a significant amount of 3[H]kainate binding activity. The antibodies also recognize the GluRD subunit, but not the other AMPA receptor subunits, when expressed in transfected COS-7 cells and only when permeabilized with detergent, indicating an intracellular epitope. All subunits were enriched in the neuropil of the dendritic layers of the hippocampus and in the molecular layer of the dentate gyrus. The cellular distribution of the GluRD subunit was studied more extensively. The strata radiatum, oriens and the dentate molecular layer were more strongly immunoreactive than the stratum lacunosum moleculare, the stratum lucidum and the hilus. However, in the stratum lucidum of the CA3 area and in the hilus the weakly reacting dendrites were surrounded by immunopositive rosettes, shown in subsequent electron microscopic studies to correspond to complex dendritic spines. In the stratum radiatum, the weakly reacting apical dendrites contrasted with the surrounding intensely stained neuropil. The cell bodies of pyramidal and granule cells were moderately reactive. Some non-principal cells and their dendrites in the pyramidal cell layer and in the alveus also reacted very strongly for the GluRD subunit. At the subcellular level, silver intensified immunogold particles for the GluRA, GluRB/C and GluRD subunits were present at type 1 synaptic membrane specializations on dendritic spines of pyramidal cells throughout all layers of the CA1 and CA3 areas. The most densely labelled synapses tended to be on the largest spines and many smaller spines remained unlabelled. Immunoparticle density at type 1 synapses on dendritic shafts of some non-principal cells was consistently higher than at labelled synapses of dendritic spines of pyramidal cells. Synapses established between dendritic spines and mossy fibre terminals, were immunoreactive for all studied subunits in stratum lucidum of the CA3 area. The postembedding immunogold method revealed that the AMPA type receptors are concentrated within the main body of the anatomically defined type 1 (asymmetrical) synaptic junction. Often only a part of the membrane specialization showed clustered immunoparticles. There was a sharp decrease in immunoreactive receptor density at the edge of the synaptic specialization. Immunolabelling was consistently demonstrated at extrasynaptic sites on dendrites, dendritic spines and somata. The results demonstrate that the GluRA, B/C and D subunits of the AMPA type glutamate receptor are present in many of the glutamatergic synapses formed by the entorhinal, CA3 pyramidal and mossy fibre terminals. Some interneurons have a higher density of AMPA type receptors in their asymmetrical afferent synapses than pyramidal cells. This may contribute to a lower activation threshold of interneurons as compared to principal cells by the same afferents in the hippocampal formation.

Purification of recombinant GluR-D glutamate receptor produced in Sf21 insect cells

GluR-D glutamate receptors carrying FLAG and polyHis affinity tags at the N-terminus and C-terminus, respectively, were expressed in recombinant baculovirus-infected Spodoptera frugiperda Sf21 cells. Affinity-tagged receptors displayed ligand-binding affinity (Kd = 40 nM) and an expression level (Bmax 10-30 pmol/mg protein) similar to that of insect-cell-expressed wild-type GluR-D, as determined by [3H]-alpha-amino-5-hydroxy-3-methyl-4-isoxazole propionic acid ([3H]AMPA) binding. The receptor was solubilized in Triton X-100, and purified using a two-step protocol consisting of immobilized metal-chelation affinity chromatography followed by immunoaffinity chromatography. The purified receptor preparation contained over 2000 pmol high-affinity [3H]AMPA-binding sites/mg protein, and migrated as a single 110-kDa species in SDS/PAGE. Peptide:N-glycosidase F treatment reduced the size of GluR-D from 110 kDa to 100 kDa, indicating the presence of N-linked glycans. Up to 100 micrograms purified GluR-D was obtained from 1 l Sf21 suspension culture (2-3 x 10(6) cells/ml). High-level expression of affinity-tagged GluRs in insect cells should be an efficient strategy to produce GluR subtypes for biochemical and structural studies.

Synaptic expression of the high-affinity kainate receptor subunit KA2 in hippocampal cultures

Non-N-methyl-D-aspartate glutamate receptors are responsible for fast excitatory neurotransmission in the mammalian CNS. These receptors are rapidly activated and desensitized in the presence of glutamate, and are often further subdivided into alpha-amino-3-hydroxy-5-methyl-4-isoxazole proprionic acid and kainate receptors based on their selective agonists. Non-NMDA glutamate receptors are composed of multiple subunits which recently have been cloned, and studies on the recombinant glutamate receptors have helped clarify the distinctions between AMPA and kainate-preferring glutamate receptors. Although the subunits which make up both AMPA and kainate receptors have a widespread distribution, most currents recorded in vivo are characteristic of recombinant AMPA receptors. To help clarify the functional role of high-affinity kainate receptors, we have characterized the expression of a high-affinity kainate receptor subunit, KA2, in cultured hippocampal neurons. Using immunocytochemistry, we found that KA2 was expressed in hippocampal neurons at all times during the development of the cells in culture, and the subunit was enriched in dendritic spines after about 14 days. The subcellular distribution of KA2 paralleled that of the AMPA receptor subunit GluR1, with the AMPA and kainate subunits being colocalized at all times in culture. The enriched KA2 immunoreactivity co-localized with the synaptic vesicle protein synaptophysin at the resolution of light microscopy, indicating synaptic localization of KA2. Although the kainate subunit KA2 co-localized with the AMPA subunit GluR1, co-immunoprecipitation experiments demonstrated a direct interaction between the AMPA receptor subunits GluR1 and GluR2/3, but not between GluR1 and the kainate subunits GluR6/7 or KA2. We therefore, conclude that both AMPA and kainate receptor subunits are enriched in the same dendritic spines, yet do not combine to form receptor complexes.