Evidence for heterogenous glycine domains but conserved multiple states of the excitatory amino acid recognition site of the NMDA receptor: regional binding studies with [3H]glycine and [3H]L-glutamate

Summary The possible heterogeneity of the agonist and glycine sites of the N-methyl-D-aspartate (NMDA) receptor-complex was examined using receptor binding techniques. Binding of ³H]L-glutamate (³H]GLU) and ³H]glycine to synaptic membranes of cerebral and cerebellar cortices, and membranes of a granule cell preparation of rat cerebellum, was characterized. ³H]Glycine always labelled a single population of sites; densities of binding sites (B_max) in cortical, cerebellar and granule membranes were 3.1, 0.87 and 3.6 pmol/mg protein, respectively. Dissociation constants (K_d) in the same three preparations were 0.13, 0.31 and 1.9 M, respectively. In competition studies, D-cycloserine, but not D-serine and 7-chlorokynurenate, showed varying potency between the membrane preparations, and analysis of variance (ANOVA) revealed a significant interaction between ligands and membrane fractions. Binding of ³H]GLU was saturable and to a single population of sites: K_d 0.5–0.9 M and B_max 3.2–3.6 pmol/mg protein. In all three membrane preparations the rank order of potency of NMDA agonists as inhibitors of the binding of ³H]GLU was always L-aspartate>L-cysteate>L-cysteinesulphinate>L-serine-O-sulphate>ibotenate>L-homocysteate. NMDA, quinolinate and competitive NMDA antagonists were only weak inhibitors of the binding of ³H]GLU and never fully inhibited specific binding. Other subtype-selective excitatory amino acids were very weak or ineffective inhibitors of binding. Binding of NMDA agonists was better described by a two site model whereby the proportion of high affinity sites did not vary significantly across the three membrane preparations. Although the binding of ³H]GLU was relatively insensitive to NMDA itself and competitive NMDA antagonists, binding may be to a recognition site for NMDA-like agonists, since they fully inhibited specific binding. This excitatory amino acid recognition for NMDA agonists was conserved in the three membrane preparations. In cortical and granule membranes the B_max values for the binding of ³H]GLU and ³H]glycine had a stoichiometry of 1:1, whilst in cerebellar synaptic membranes this ratio was 4:1. Receptor autoradiography of NMDA-related ³H]GLU and ³H]glycine binding in tissue sections failed to reveal any differential labelling patterns in cerebral cortex and cerebellum. In the cerebellum, densities of silver grains found with both ³H]ligands were concentrated in the granule cell layer relative to the molecular layer, but the differences detected in membrane binding studies were not observed in cerebellum. Our findings suggest the existence of three types of heterogeneity for the glycine domain of the NMDA receptor: (1) differing affinities for glycine, (2) differing pharmacological profiles, and (3) differing stoichiometry in relation to the putative NMDA-like agonist site. Our evidence supports an hypothesis for the existence of multiple glycine domains which might differentially modulate NMDA-mediated neurotransmission.