Biochemical evidence for the co-association of three N-methyl-D-aspartate (NMDA) R2 subunits in recombinant NMDA receptors.

Functional characterization of wild-type and mutant cloned N-methyl-D-aspartate (NMDA) receptors has been used to deduce their subunit stoichiometry and quaternary structure. However, the results reported from different groups have been at variance and are thus inconclusive. This study has employed a biochemical approach to determine the number of NMDA R2 (NR2) subunits/receptor together with the NMDA R1 (NR1)/NR2 subunit ratio of both cloned and native NMDA receptors. Thus, human embryonic kidney 293 cells were transfected with the NR1-1a and NR2A NMDA receptor subunits in combination with both FLAG- and c-Myc epitope-tagged NR2B subunits. The expressed receptors were detergent-extracted and subjected to double immunoaffinity purification using anti-NR2A and anti-FLAG antibody immunoaffinity columns in series. Immunoblotting of the double immunopurified NR2A/NR2B(FLAG)-containing material demonstrated the presence of anti-NR1, anti-NR2A, anti-FLAG, and, more important, anti-c-Myc antibody immunoreactivities. The presence of anti-c-Myc antibody immunoreactivity in the double immunoaffinity-purified material showed the co-assembly of three NR2 subunits, i.e. NR2A/NR2B(FLAG)/NR2B(c-Myc), within the same NMDA receptor complex. Control experiments excluded the possibility that the co-immunopurification of the three NR2 subunits was an artifact of the solubilization procedure. These results, taken together with those previously described that showed two NR1 subunits/oligomer, suggest that the NMDA receptor is at least pentameric.

Pharmacological differentiation of kainate receptors on neonatal rat spinal motoneurones and dorsal roots.

The objectives of this study, conducted on neonatal rat spinal cord and dorsal roots in vitro, were to characterise the actions of a range of willardiine analogues on GluR5-containing kainate receptors present in dorsal roots, to determine whether GluR5-containing receptors are also present on motoneurones, and to differentiate responses mediated by kainate receptors from those mediated by AMPA receptors on motoneurones. (S)-5-Trifluoromethyl-willardiine, (S)-5-iodowillardiine, (S)-5-iodo-6-azawillardiine and ATPA were found to be potent agonists of kainate receptors on dorsal roots (EC50 values 0.108 +/- 0.002, 0.127 +/- 0.010, 0.685 +/- 0.141 and 1.3 +/- 0.3 microM, respectively) being more potent but of lower efficacy than kainate (EC50 value 14.8 +/- 1.8 microM). (S)-5-Iodo-6-azawillardiine blocked kainate-induced depolarisations of the dorsal root, probably via its desensitising action. Kainate-induced responses of dorsal roots were weakly antagonised by (RS)-3,5-dicarboxyphenylglycine (DCPG) (apparent KD 1.5 +/- 0.4 mM). Kainate receptors containing GluR5 subunits do not appear to be present on motoneurones since (RS)-3,5-DCPG (1 mM) potentiated rather than antagonised kainate-induced depolarisations of motoneurones. Although (S)-5-iodowillardiine (a potent and selective agonist at GluR5-containing kainate receptors) depolarised motoneurones (EC50 value 5.8 +/- 0.6 microM), such depolarisations were antagonised by both (RS)-3,4- and (RS)-3,5-DCPG, which are selective AMPA receptor antagonists at motoneurones, showing a KD value of 73 microM (Schild slope, 0.96 +/- 0.09) and an apparent KD value of 123 +/- 38 microM, respectively. This accords with the previously reported activity of willardiine analogues at AMPA receptors. Since neither (RS)-3,4- nor (RS)-3,5-DCPG antagonised kainate-induced motoneuronal depolarisations but cyclothiazide enhanced and GYK153655 blocked these responses it is possible that a component of the kainate response may be mediated by a population of DCPG-insensitive AMPA receptors on motoneurones. However, it is also possible that a population of kainate receptors other than those containing GluR5 subunits, are responsible for these effects. The new compounds introduced in this study are likely to be useful tools for studying the physiological role of kainate receptors in CNS function.

Inhibition of AMPA receptor-stimulated 57Co2+ influx by D- and L-2-amino-4-phosphonobutanoic acid (D- and L-AP4) and L-serine-O-phosphate (L-SOP) in cultured cerebellar granule cells.

This study describes the inhibition of 57Co2+ influx through Ca2+-permeable alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, consequent to the application of L-2-amino-4-phosphonobutanoic acid (L-AP4), D-AP4 and L-serine-O-phosphate (L-SOP) in cultured cerebellar granule cells. The forskolin-stimulated accumulation of cyclic AMP was inhibited by (2S,1'S,2'S)-2-(carboxycyclopropyl)glycine (L-CCG-1) with an IC50 = 491 +/- 135 nM and by L-AP4 in a biphasic manner (IC50(1) = 232 +/- 61 nM and IC50(2) = >300 microM), confirming the presence of group II and group III mGlu receptors, respectively. 57Co2+ influx was stimulated by kainate (EC50 = 42.2 +/- 11.3 microM) and, in the presence of 30 microM cyclothiazide, by (S)-5-fluorowillardiine (EC50 = 0.7 +/- 0.1 microM) and (S)-AMPA (EC50 = 2.8 +/- 0.5 microM). The effects of the latter were abolished by 10 microM 6-nitro-7-sulphamoylbenzo[f]quinoxaline-2,3-dione (NBQX). L-AP4 (IC50 = >300 microM), D-AP4 (IC50 = >100 microM) and L-SOP (IC50 = 199 +/- 6 microM) inhibited 6 microM (S)-AMPA-stimulated 57Co2+ influx, whereas L-CCG-1 (up to 10 microM), 300 microM (RS)-3,5-dihydroxyphenylglycine, 300 microM (+/-)-baclofen and 1 mM carbachol were ineffective. Pre-incubation with either pertussis toxin (250 ng/ml, 48 hr), 1 mM dibutyryl cyclic AMP, or the potent group III mGlu receptor antagonist (RS)-alpha-cyclopropyl-4-phosphonophenylglycine ((RS)-CPPG), tested at 400 microM, failed to alter the inhibition of AMPA receptor activity by 300 microM L-SOP. Unlike 10 microM NBQX, neither L-AP4, D-AP4 or L-SOP (tested at 1 mM) inhibited the binding of 10 nM (S)-[3H]5-fluorowillardiine (a selective AMPA receptor ligand) to granule cell membranes. Therefore, in these neurones, high concentrations (>100 microM) of L-AP4, L-SOP and D-AP4 inhibit Ca2+-permeable AMPA receptors by a mechanism distinct from known mGlu receptor action and at a site independent from that for AMPA receptor agonists.

Characterization of the pharmacology and regional distribution of (S)-[3H]-5-fluorowillardiine binding in rat brain.

1. This study examined the binding of the new radioligand (S)-[3H]-5-fluorowillardiine to rat brain synaptic membranes. Specific binding represented greater than 80% of the total binding and was increased by 10% in the presence of 100mM potassium thiocyanate (KSCN). 2. In the absence of KSCN, (S)-[3H]-5-fluorowillardiine identified two binding sites with KD1=22.5 nM, Bmax1=1.4 pmol mg(-1) protein and KD2=1.5 microM, Bmax2=10.8 pmol mg(-1) protein. In the presence of 100 mM KSCN the affinities of both the binding sites were increased, yielding values of KD1=6.9 nM and KD2=0.4 microM KSCN was without effect on the Bmax values. 3. (S)-[3H]-5-fluorowillardiine binding was displaced by non-NMDA receptor ligands with the rank order of potency: 2,3-dihydroxy-6-nitro-7-sulphamoyl-benzo(F)quinoxaline (NBQX) > domoate > (S)-alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionice acid (AMPA) = L-glutamate > 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) > kainate >> (R)-5-fluorowellardiine. In contrast, both N-methyl-D-aspartate (NMDA) and the metabotropic glutamate receptor agonist, (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (ACPD) were inactive. 4. By use of quantitative autoradiography the regional distribution of (S)-[3H]-5-fluorowillardiine binding in rat brain was assessed. The highest levels of binding were in the dentate gyrus and the CA1 region of the hippocampus. Lower levels of binding were detected in the cerebral cortex, olfactory system, lateral septum, caudate putamen and nucleus accumbens. 5. We conclude that the pharmacological profile and regional distribution of (S)-[3H]-5-fluorowillardiine binding is consistent with its specific interaction with AMPA receptors.

Binding of the new radioligand (S)-[3H]AMPA to rat brain synaptic membranes: effects of a series of structural analogues of the non-NMDA receptor agonist willardiine.

This study examined the binding of (S)-[3H]AMPA, the radiolabelled active isomer of AMPA, to rat brain synaptic membranes. Under non-chaotropic conditions specific binding of 10 nM (S)-[3H]AMPA represented 33 +/- 2% of the total; this increased to 74 +/- 1% in the presence of 100 mM KSCN. (S)-[3H]AMPA binding was inhibited by non-NMDA receptor agonists and the antagonists NBQX and CNQX, with the following rank order of potency: NBQX > (S)-AMPA > or = quisqualate > CNQX > L-glutamate > domoate > or = kainate > (R)-AMPA. NMDA, and the metabotropic glutamate receptor agonist (1S,3R)-ACPD, up to 100 microM, did not inhibit (S)-[3H]AMPA binding. A number of willardiine analogues all effectively inhibited (S)-[3H]AMPA binding with the rank order of potency: (S)-5-fluorowillardiine > (S)-5-nitrowillardiine > (S)-5-trifluoromethylwillardiine > (S)-5-bromowillardiine approximately (S)-5-chlorowillardiine > (S)-5-cyanowillardiine > (S)-willardiine > (S)-5-iodowillardiine > (S)-6-methylwillardiine > (S)-5-methylwillardiine. This rank order closely reflects data from equilibrium measurements made, under voltage clamp, on cultured hippocampal neurons. In contrast the respective (R)-enantiomers and the racemate mixtures of (R,S)-3, 5 and 6-isowillardiine were relatively inactive. Similar IC50 values and thus rank orders of potency for the willardiines were observed in the presence of 100 mM KSCN.

5-Carboxamidotryptamine-insensitive 5-HT1-like receptors are concentrated in guinea pig but not rat, claustrum.

5-CT (5-carboxamidotryptamine)-insensitive (5-HT1E/5-HT1F) 5-HT1-like recognition sites have been mapped autoradiographically in rat and guinea pig brain using [3H]5-HT in the presence of 5-CT and mesulergine to mask 5-HT1A, 5-HT1B, 5-HT1D and 5-HT2C binding sites. Binding was more dense in the guinea pig but in both species 5-CT-insensitive 5-HT1-like sites were located in the olfactory tubercle, interpeduncular nucleus, caudate putamen, nucleus accumbens, substantia nigra, frontal cortex and hippocampus. These receptors were particularly marked in the claustrum of the guinea pig but not the rat.

Stereoselective actions of the isomers of metitepine at 5-HT1D receptors in the guinea pig brain.

The present studies examined the relative antagonist potencies of the optical isomers of the 5-HT receptor antagonist metitepine at the 5-HT1D binding site labelled by the novel radioligand serotonin-O-carboxymethylglycyl [125]iodotyrosinamide ([125I]GTI), and at the terminal 5-HT autoreceptor in guinea pig frontal cortex, a proposed model of 5-HT1D receptor activation. The pharmacological specificity of the [125I]GTI binding site in guinea pig frontal cortex was similar to previously published studies in the bovine cortical 5-HT1D recognition site labelled with [3H]5-HT. The (+) isomer of metitepine displaced [125I]GTI binding with a lower affinity (64 nM) than did the (-) isomer (18 nM), which was equiactive with the racemic mixture. The (-) isomer of metitepine was more effective than the (+) isomer at attenuating the inhibitory effects of 5-HT and sumatriptan at the guinea pig terminal 5-HT autoreceptor; the apparent pA2 of the (-) isomer was 8.0 (sumatriptan) and 7.7 (5-HT) while the apparent pA2 of the (+) isomer was 7.1 (sumatriptan) and 6.8 (5-HT). The (-) isomer was more effective than the (+) isomer at enhancing stimulated [3H]5-HT release. These findings support the identification of the guinea pig 5-HT terminal autoreceptor as a 5-HT1D receptor and reinforce the species homology between the 5-HT1B and 5-HT1D receptors.