Increased [³H]quisqualic acid binding density in the dorsal striatum and anterior insula of alcoholics: A post-mortem whole-hemisphere autoradiography study.

The function of group I metabotropic glutamate receptors mGluR1 and mGluR5 is involved in the hyperglutamatergic state caused by chronic alcohol. Preclinical studies suggest that group I mGluR modulation could serve as a novel treatment of alcoholism. Considering the wide role of glutamatergic neurochemistry in addiction, group I mGluR binding was studied in brain areas involved in decision-making, learning and memory. Post-mortem whole hemisphere autoradiography was used to study the binding density of [³H]quisqualic acid, a potent group I mGluR agonist, in 9 Cloninger type 1 alcoholics, 8 Cloninger type 2 alcoholics and 10 controls. Binding was studied in the dorsal striatum, hippocampus and cortex. Alcoholics displayed a trend towards increased [³H]quisqualic acid binding in all brain areas. The most robust findings were in the putamen (p = 0.006) and anterior insula (p = 0.005), where both alcoholic subtypes displayed increased binding compared to the controls. These findings suggest altered group I mGluR function in alcoholic subjects in the dorsal striatum, which is involved in habitual learning, and in the anterior insula, which has a pivotal role in the perception of bodily sensations. Increased [³H]quisqualic acid binding might suggest a beneficial impact of mGluR1/5 modulators in the treatment of alcoholism.

Orthosteric and allosteric drug binding sites in the Caenorhabditis elegans mgl-2 metabotropic glutamate receptor.

The metabotropic glutamate receptors (mGluRs) are evolutionarily conserved from nematodes to vertebrates. The Caenorhabditis elegans (C. elegans) genome contains three mGluR genes referred to as mgl-1, mgl-2, and mgl-3. The aim of this study was to characterize the pharmacological profiles of orthosteric and allosteric mGluR ligands on mgl-2. A phylogenetic analysis revealed that mgl-2 is closely associated with the mammalian Group 1 mGluRs (mGluR1 and mGluR5) and is distinct from Group 2 and 3 mGluRs. The ligand binding domain of mgl-2 displayed higher homology to the rat Group 1 mGluRs binding domains compared to the level of homology in the heptahelical transmembrane domain regions. We found that, when transiently expressed in human embryonic kidney 293 cells, mgl-2 can be activated by glutamate and couples to human G-proteins to induce the release of intracellular calcium. Dose-response analyses revealed that mgl-2 has approximately a 15-20-fold lower affinity for glutamate and quisqualate compared to rat mGluR5. In contrast to orthosteric agonists, Group 1 negative allosteric modulators that target the transmembrane domain were ineffective at mgl-2. Surprisingly, CDPPB, an mGluR5 positive allosteric modulator, potentiated glutamate mediated activation of mgl-2, although MPEP and fenobam, two mGluR5 antagonists that share similar binding residues with CDPPB were ineffective at mgl-2. These findings indicate that selective pressures on mGluR protein structures have resulted in conservation of the glutamate binding site, whereas the allosteric modulator sites have been subjected to greater divergent evolutionary changes.

The loss of an electrostatic contact unique to AMPA receptor ligand binding domain 2 slows channel activation.

Ligand-gated ion channels undergo conformational changes that transfer the energy of agonist binding to channel opening. Within ionotropic glutamate receptor (iGluR) subunits, this process is initiated in their bilobate ligand binding domain (LBD) where agonist binding to lobe 1 favors closure of lobe 2 around the agonist and allows formation of interlobe hydrogen bonds. AMPA receptors (GluAs) differ from other iGluRs because glutamate binding causes an aspartate-serine peptide bond in a flexible part of lobe 2 to rotate 180° (flipped conformation), allowing these residues to form cross-cleft H-bonds with tyrosine and glycine in lobe 1. This aspartate also contacts the side chain of a lysine residue in the hydrophobic core of lobe 2 by a salt bridge. We investigated how the peptide flip and electrostatic contact (D655-K660) in GluA3 contribute to receptor function by examining pharmacological and structural properties with an antagonist (CNQX), a partial agonist (kainate), and two full agonists (glutamate and quisqualate) in the wildtype and two mutant receptors. Alanine substitution decreased the agonist potency of GluA3(i)-D655A and GluA3(i)-K660A receptor channels expressed in HEK293 cells and differentially affected agonist binding affinity for isolated LBDs without changing CNQX affinity. Correlations observed in the crystal structures of the mutant LBDs included the loss of the D655-K660 electrostatic contact, agonist-dependent differences in lobe 1 and lobe 2 closure, and unflipped D(A)655-S656 bonds. Glutamate-stimulated activation was slower for both mutants, suggesting that efficient energy transfer of agonist binding within the LBD of AMPA receptors requires an intact tether between the flexible peptide flip domain and the rigid hydrophobic core of lobe 2.

Pyk2 uncouples metabotropic glutamate receptor G protein signaling but facilitates ERK1/2 activation.

Group I metabotropic glutamate receptors (mGluRs) are coupled via Galphaq/11 to the activation of phospholipase Cbeta, which hydrolyzes membrane phospholipids to form inositol 1,4,5 trisphosphate and diacylglycerol. This results in the release of Ca2+ from intracellular stores and the activation of protein kinase C. The activation of Group I mGluRs also results in ERK1/2 phosphorylation. We show here, that the proline-rich tyrosine kinase 2 (Pyk2) interacts with both mGluR1 and mGluR5 and is precipitated with both receptors from rat brain. Pyk2 also interacts with GST-fusion proteins corresponding to the second intracellular loop and the distal carboxyl-terminal tail domains of mGluR1a. Pyk2 colocalizes with mGluR1a at the plasma membrane in human embryonic kidney (HEK293) cells and with endogenous mGluR5 in cortical neurons. Pyk2 overexpression in HEK293 results in attenuated basal and agonist-stimulated inositol phosphate formation in mGluR1 expressing cells and involves a mechanism whereby Pyk2 displaces Galphaq/11 from the receptor. The activation of endogenous mGluR1 in primary mouse cortical neuron stimulates ERK1/2 phosphorylation. Treatments that prevent Pyk2 phosphorylation in cortical neurons, and the overexpression of Pyk2 dominant-negative and catalytically inactive Pyk2 mutants in HEK293 cells, prevent ERK1/2 phosphorylation. The Pyk2 mediated activation of ERK1/2 phosphorylation is also Src-, calmodulin- and protein kinase C-dependent. Our data reveal that Pyk2 couples the activation mGluRs to the mitogen-activated protein kinase pathway even though it attenuates mGluR1-dependent G protein signaling.

Microsecond-to-millisecond conformational dynamics demarcate the GluR2 glutamate receptor bound to agonists glutamate, quisqualate, and AMPA.

Chemical shift changes and internal motions on microsecond-to-millisecond time scales of the S1S2 ligand-binding domain of the GluR2 ionotropic glutamate receptor have been studied by NMR spectroscopy in the presence of the agonists glutamic acid (glutamate), quisqualic acid (quisqualate), and alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA). Although the crystal structures of the three agonist-bound forms of GluR2 S1S2 ligand-binding domain are very similar, chemical shift changes imply that AMPA-bound GluR2 S1S2 is conformationally distinct from glutamate- and quisqualate-bound forms of GluR2 S1S2. NMR spin relaxation measurements for backbone amide (15)N nuclei reveal that GluR2 S1S2 exhibits reduced chemical exchange line broadening, resulting from microsecond-to-millisecond conformational dynamics, in AMPA-bound compared to glutamate- and quisqualate-bound states. The largest changes in line broadening are observed for two regions of GluR2 S1S2: Val683 and the segment around Lys716-Cys718. The differences in binding affinity of these agonists do not explain the differences in microsecond-to-millisecond conformational dynamics because quisqualate and AMPA bind with similar affinities that are 10-fold greater than the affinity of glutamate. Differences in conformational mobility may reflect differences in the binding mode of AMPA in the GluR2 S1S2 active site compared to the other two ligands. The sites of conformational mobility in GluR2 S1S2 imply that subtle differences exist between the agonists glutamate, quisqualate, and AMPA in modulating glutamate receptor function.

Blockade of excitatory amino acid transporters in the rat hippocampus results in enhanced activation of group I and group III metabotropic glutamate receptors.

The idea that excitatory amino acid transporters (EAATs) can control the activation of specific metabotropic glutamate receptors (mGluRs) was investigated in rat hippocampal slices. Using the accumulation of inositol phosphates as a measure of group I mGluR activity, we have shown that the broad spectrum, non-transportable EAAT blocker, TBOA, produces a significant shift to the left of agonist concentration-response curves. Moreover, this increase in potency did not occur if endogenous glutamate was enzymatically removed, suggesting a glutamate-dependent mechanism. This shift in potency was shown to be NMDA and group II mGlu receptor independent. Additionally, experiments with selective antagonists indicated that the group I receptor responsible for the stimulation of inositol phosphate production in this preparation is likely to be mGluR5. Inhibition of forskolin-stimulated cyclic AMP (cAMP) production was used as an index of group II/III mGluR activity. TBOA produced a rightward shift of the forskolin concentration-response curve. A group III, but not a group II, mGluR agonist also produced this effect, suggesting that the TBOA-mediated increase in glutamate activates a receptor, which appears to be a member of the group III mGluR subset. This was confirmed by the observation that an antagonist of group III mGluRs, prevented the TBOA-induced rightward shift in forskolin potency. These results provide evidence of a role for EAATs in the regulation of mGluR5 and group III mGluRs in the rat hippocampus, which may have therapeutic implications.

Amino acid mutagenesis of the ligand binding site and the dimer interface of the metabotropic glutamate receptor 1. Identification of crucial residues for setting the activated state.

Previously, we determined the crystal structures of the dimeric ligand binding region of the metabotropic glutamate receptor subtype 1. Each protomer binds l-glutamate within the crevice between the LB1 and LB2 domains. We proposed that the two different conformations of the dimer interface between the two LB1 domains define the activated and resting states of the receptor protein. In this study, the residues in the ligand-binding site and the dimer interface were mutated, and the effects were analyzed in the full-length and truncated soluble receptor forms. The variations in the ligand binding activities of the purified truncated receptors are comparable with those of the full-length form. The mutated full-length receptors were also analyzed by inositol phosphate production and Ca(2+) response. The magnitude of the ligand binding capacities and the amplitude of the intracellular signaling were almost correlated. Alanine substitutions of four residues, Thr(188), Asp(208), Tyr(236), and Asp(318), which interact with the alpha-amino group of glutamate in the crystal, abolished their responses both to glutamate and quisqualate. The mutations of the Tyr(74), Arg(78), and Gly(293) residues, which interact with the gamma-carboxyl group of glutamate, lost their responsiveness to glutamate but not to quisqualate. Furthermore, a mutant receptor containing alanine instead of isoleucine at position 120 located within an alpha helix constituting the dimer interface showed no intracellular response to ligand stimulation. The results demonstrate the crucial role of the dimer interface in receptor activation.

Characterization of mGluR5R, a novel, metabotropic glutamate receptor 5-related gene.

We report here the isolation of a novel gene termed mGluR5R (mGluR5-related). The N-terminus of mGluR5R is highly similar to the extracellular domain of metabotropic glutamate receptor 5 (mGluR5) whereas the C-terminus bears similarity to the testis-specific gene, RNF18. mGluR5R is expressed in the human CNS in a coordinate fashion with mGluR5. Although the sequence suggests that mGluR5R may be a secreted glutamate binding protein, we found that when expressed in HEK293 cells it was membrane associated and not secreted. Furthermore, mGluR5R was incapable of binding the metabotropic glutamate receptor class I selective agonist, quisqualate. Although mGluR5R could not form disulfide-mediated covalent homodimers, it was able to form a homomeric complex, presumably through noncovalent interactions. mGluR5R also formed noncovalent heteromeric associations with an engineered construct of the extracellular domain of mGluR5 as well as with full-length mGluR5 and mGluR1alpha. The ability of mGluR5R to associate with mGluR1alpha and mGluR5 suggests that it may be a modulator of class I metabotropic glutamate receptor function.

Identification and characterization of a novel splice variant of the metabotropic glutamate receptor 5 gene in human hippocampus and cerebellum.

The G-protein coupled metabotropic glutamate receptor mGlu5 plays a pivotal role as a modulator of synaptic plasticity, ion channel activity and excitotoxicity. Two splice variants, hmGlu5a and -5b have been reported previously. During screening of a human brain cDNA library for hmGlu5a, we identified a novel variant (hmGlu5d) generated by alternative splicing at the C-terminal domain. The predicted hmGlu5d protein has a C-terminal 267 amino acid shorter than that of hmGlu5a. The pattern of mRNA expression of mGluR5 variants in human brain were analyzed by RT-PCR and in situ hybridization histochemistry. RT-PCR analysis demonstrated the presence of the hmGlu5d transcript, although at low level, in human whole brain, cerebellum, cerebral cortex and hippocampus. [3H]Quisqualate displayed similar affinity at the hmGlu5 splice variants (K(D) values of 80+/-8 and 54+/-17 nM for hmGlu5a and -5d receptors, respectively). For the five mGlu agonists studied, a similar rank order of potency was observed on both hmGlu5a and -5d receptors: quisqualate>glutamate>DHPG>L-CCGI approximately ACPD. MPEP inhibited the glutamate (2 microM)-induced [Ca(2+)](i) response in hmGlu5a and -5d-HEK293 cells also with similar potency (IC(50) values 25+/-1.5 and 20+/-1.4 nM, respectively). Therefore, the large truncation of the C-terminal tail of mGlu5 does not have any apparent major effect on the potency and efficacy of agonists as measured by the [Ca(2+)](i) responses or by activation of recombinant G-protein coupled inwardly rectifying K(+) (GIRK) channel currents. The only major functional difference is the increased sensitivity of hmGlu5d to protein kinase C (PKC)-mediated desensitization, relative to hmGlu5a.

Positive allosteric modulators of metabotropic glutamate 1 receptor: characterization, mechanism of action, and binding site.

We have identified two chemical series of compounds acting as selective positive allosteric modulators (enhancers) of native and recombinant metabotropic glutamate 1 (mGlu1) receptors. These compounds did not directly activate mGlu1 receptors but markedly potentiated agonist-stimulated responses, increasing potency and maximum efficacy. Binding of these compounds increased the affinity of a radiolabeled glutamate-site agonist at its extracellular N-terminal binding site. Chimeric and mutated receptors were used to localize amino acids in the receptor transmembrane region critical for these enhancing properties. Finally, the compounds potentiated synaptically evoked mGlu1 receptor responses in rat brain slices. The discovery of selective positive allosteric modulators of mGlu1 receptors opens up the possibility to develop a similar class of compounds for other family 3 G protein-coupled receptors.

Characterization with [3H]quisqualate of group I metabotropic glutamate receptor subtype in rat central and peripheral excitable tissues.

Radioligand binding studies were performed to label metabotropic glutamate receptor (mGluR) in rat brain synaptic membranes using [3H]quisqualic acid (QA) synthesized in our laboratory as a radioligand. In the presence of ionotropic glutamate receptor (iGluR) agonists, including N-methyl-D-aspartic (NMDA), DL-alpha-amino-3-hydroxy-5-methylisoxasole-4-propionic (AMPA) and kainic acids (KA), at concentrations maximally effective in displacing each receptor binding, the agonists for group I mGluR subtype (+/-)-1-aminocyclopentane-trans-1,3-dicarboxylic acid (trans-ACPD) and (S)-3,5-dihydroxyphenylglycine ((S)-3,5-DHPG) more potently displaced [3H]QA binding in a concentration-dependent manner than their absence. The addition of these three iGluR agonists did not significantly affect potencies of (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV) and L-(+)-2-amino-4-phosphonobutyric acid (L-AP4) to displace [3H]QA binding. Scatchard analysis revealed that [3H]QA binding consisted of a single component with a maximal number of binding sites (B(max)) of 431.6 fmol/mg protein and a dissociation constant (K(d)) of 50.9 nM, in the presence of the three iGluR agonists. [3H]QA binding was markedly inhibited by GTP and its analogues; but not by GDP, GMP and ATP, under these conditions. Inhibition by GTP was seen in all central structures examined, but [3H]QA binding was not detectable in peripheral tissues, such as pituitary and adrenal glands. Neither reverses transcription polymerase chain reaction nor immunoblotting analysis demonstrated the expression of mGluR1 and mGluR5 subunits in the aforementioned two peripheral tissues. These results suggest that [3H]QA indeed labels group I mGluR subtype functionally coupled to GTP binding protein in rat brain synaptic membranes under the experimental conditions employed. Group I mGluR subtype seems to be selectively distributed in central structures but not in pituitary and adrenal glands.

Direct radiolabeling by [3H]quisqualic acid of group I metabotropic glutamate receptor in rat brain synaptic membranes.

[3H]Quisqualic acid (QA) was synthesized and used to label metabotropic glutamate receptor (mGluR) in rat brain synaptic membranes in the presence of three different ionotropic glutamate receptor agonists at respective saturating concentrations. Of several mGluR agonists tested, group I agonists were more potent in displacing [3H]QA binding than group II and group III agonists in the presence of the three ionotropic agonists. [3H]QA binding was markedly inhibited by guanine nucleotide analogues in a concentration-dependent manner at a concentration range of 10 nM to 1 mM. Scatchard analysis revealed that [3H]QA binding consisted of a single component with a K(d) of 50.9+/-5.3 nM and a B(max) of 431. 6+/-18.3 fmol/mg protein. These results suggest that [3H]QA indeed labels group I mGluR functionally coupled to GTP binding protein in rat brain synaptic membranes when determined under the experimental conditions employed.

The tetrameric structure of a glutamate receptor channel.

The subunit stoichiometry of several ligand-gated ion channel receptors is still unknown. A counting method was developed to determine the number of subunits in one family of brain glutamate receptors. Successful application of this method in an HEK cell line provides evidence that ionotropic glutamate receptors share a tetrameric structure with the voltage-gated potassium channels. The average conductance of these channels depends on how many subunits are occupied by an agonist.

Specific [3H]glutamate binding in the cerebral cortex and hippocampus of rats during development: effect of homocysteine-induced seizures.

Specific [3H]glutamate binding to synaptic membranes from the cerebral cortex and hippocampus of 7-, 12- and 18-day-old rats was examined, both in control animals and during seizures induced by homocysteine. In the cerebral cortex a transient peak of glutamate binding was observed in 7-day-old group, whereas in the hippocampus it occurred in 12-day-old animals. Total specific [3H]glutamate binding was not influenced by preceding seizure activity in either of the age groups and both the studied regions. NMDA- and QA-sensitive glutamate bindings represent the highest portion of the total binding. Moreover, NMDA-sensitive binding in the cerebral cortex of 7-day-old rats is significantly higher as compared to the two more mature groups. The proportion of individual receptor subtypes on total binding in each age group was not influenced by preceding seizure activity. However, NMDA-sensitive binding in the hippocampus of 12-day-old rats, sacrificed during homocysteine-induced seizures, was significantly increased as compared to corresponding controls. In contrast to the effect of NMDA, AMPA, kainate and quisqualate which displaced to a different extent [3H]glutamate binding, homocysteine had no effect when added to membrane preparations. Similarly, [3H]CPP and [3H]AMPA bindings were not affected in the presence of homocysteine. It thus seems unlikely that homocysteine is an effective agonist for conventional ionotropic glutamate receptors. Its potential activity at some of the modulatory sites at the NMDA receptor channel complex or at metabotropic receptors has to be clarified in further experiments.

Homology modeling of metabotropic glutamate receptors. (mGluRs) structural motifs affecting binding modes and pharmacological profile of mGluR1 agonists and competitive antagonists.

A three-dimensional model of the amino terminal domain (ATD) of the mGluR1 receptor subtype was constructed on the basis of the previously reported sequence homology with bacterial periplasmic proteins. The model was utilized for revealing structural motifs affecting the interaction with mGluR1 agonists and competitive antagonists. The agonist binding site region, identified on the basis of published site-directed mutagenesis experiments, is located on the surface of one of the two lobes constituting the mGluR1 ATD. A number of electrostatic and hydrogen-bonding interactions can be detected between mGluR1 agonists such as L-Glu (1), Quis (2), and (1S,3R)-ACPD (4) and binding site residues. A different binding mode was proposed for mGluR1 competitive antagonists such as 4CPG (5), 4C3HPG (6), and UPF523 (10). Interactions with both lobes of the ATD of mGluR1 and the lack of a specific role for the phenyl moiety of mGluR1 antagonists are important features of the proposed antagonist binding mode. The correspondence of the molecular modeling results with the pharmacological data of mGluR1 agonists and competitive antagonists is a confirmation of the plausibility of the model.

A venus flytrap mechanism for activation and desensitization of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptors.

Desensitization of the alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) subtype of glutamate receptor channels is an important process shaping the time course of synaptic excitation. Upon desensitization, the receptor channel closes and the agonist affinity increases. So far, the nature of the structural rearrangements leading to these events was unknown. On the basis of the structural homology of the ligand binding domains of AMPA receptors and of the bilobated bacterial periplasmic proteins, we now show that agonist interaction with one lobe of the GluR1 subunit of homomeric AMPA receptors controls channel activation while additional interactions with the other lobe cause channel desensitization. Accordingly, we suggest that the transition of the AMPA receptor channel to the desensitized state involves the agonist-mediated stabilization of the closed lobe conformation of its binding domain and is a process akin to that used by the venus flytrap.

Subcellular localization and molecular pharmacology of distinct populations of [3H]-AMPA binding sites in rat hippocampus.

1. The subcellular distributions of [3H]-alpha-amino-3-hydroxy-5- methylisoxazolepropionate ([3H]-AMPA) and [3H]-kainate binding sites in rat hippocampus were investigated by cell fractionation techniques. 2. Two major populations of [3H]-AMPA sites were detected with the majority of binding located intracellularly in the microsomal (P3) fraction. Most of the remaining sites were in the synaptosomal membrane fraction but some were also present in the nuclear fraction. In contrast, essentially all of the [3H]-kainate binding sites were in the synaptosomal membrane fraction. 3. Saturation binding analysis yielded KD and Bmax values for [3H]-AMPA of 147 nM and 2.6 pmol mg-1 protein respectively for the synaptosomal membrane-associated sites and 129 nM and 5.3 pmol mg-1 protein respectively for the microsomal sites. 4. Both main populations of [3H]-AMPA sites displayed the same rank order of inhibition by competitive ligands, the apparent Mr values of GluR1 subunits were equivalent, suggesting the same degree of post-translational modification and the hydrodynamic properties of the receptor complexes were identical. 5. These data are consistent with the hypothesis that the movement of AMPA receptors between cellular compartments in the postsynaptic neurone could constitute one mechanism underlying long-term potentiation in the hippocampus.

Modulation of alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) binding sites by nitric oxide.

Nitric oxide release is reported to be involved in physiological processes associated with altered sensitivity of the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) class of glutamate receptor. A series of compounds liberating nitric oxide were therefore tested for their ability to modulate in vitro the characteristics of [3H]AMPA binding to sections of rat brain. Pretreatment of forebrain or cerebellar sections with sodium nitroprusside (1 mM), S-nitroso-N-acetylpenicillamine (SNAP, 200 microM), glyceryl trinitrate (1 microM), or isosorbide dinitrate (0.5 mM) all increased the binding of 3 nM [3H]AMPA by 15-30%. These actions were reproduced by 8-bromo-cyclic GMP (200 microM) in the cerebellum but not in the forebrain. In a similar manner, the effect of SNAP was attenuated by an inhibitor of cyclic GMP-dependent protein kinase in the cerebellum but not in the forebrain. The elevated [3H]AMPA binding observed after pretreatment with SNAP was caused by an increase in binding affinity, but the capacity of the sites was unchanged. Autoradiographic analysis showed that forebrain binding was enhanced in the cerebral cortex and hippocampus but not in the striatum. Nitric oxide therefore appears to be able to increase the affinity of AMPA binding sites via two distinct mechanisms in different brain areas. This action may contribute to synaptic plasticity associated with nitric oxide release.

Glutamate receptor binding in juvenile and adult Rana pipiens CNS.

Autoradiographic methods were used to map NMDA- and quisqualate-sensitive glutamate binding sites in the brain of mature and juvenile Rana pipiens frogs. NMDA- and quisqualate-sensitive sites were consistently co-localized in the CNS. The highest glutamate binding occurred in the telencephalon, hypothalamus, and cerebellum. Glutamate binding sites were also specifically localized in visual pathways, including the superficial neuropil of the optic tectum, consistent with glutamate being the retinal ganglion cell neurotransmitter. The distribution of glutamate binding sites in the brain of juvenile postmetamorphic frogs was similar to that in adults. In general, Quis binding increased about twofold in adults compared to juveniles, whereas NMDA binding did not show a comparable developmental increase. To test whether glutamate binding sites are located on retinal axon terminals or on tectal cell dendrites in the optic tectum, juvenile postmetamorphic frogs were enucleated unilaterally, and receptor binding was performed following 1, 3, 7, and 14 days survival. The denervated tectal neuropil showed a delayed decrease in NMDA- and quisqualate-sensitive binding, consistent with the receptors being located on postsynaptic tectal cell dendrites.