Native N-methyl-D-aspartate receptors containing NR2A and NR2B subunits have pharmacologically distinct competitive antagonist binding sites

The pharmacological properties of native N-methyl-D-aspartate (NMDA) receptors were determined in rat brain sections with quantitative autoradiography of [³H](E)-2-amino-4-propyl-5-phosphono-3-pentenoic acid (CGP39653) binding. With five competitive antagonists as displacers, two subpopulations of binding sites were observed in the horizontal plane of section examined. These two populations corresponded anatomically to NR2A and NR2B subunits. Quantitative analysis of NR2A-like and NR2B-like binding sites was enabled by examining the cerebellar granule cell layer, which expresses NR2A and NR2C subunits, and the medial striatum, which predominately expresses NR2B subunits. The antagonists (R)-(E)-4-(3-phosphonoprop-2- enyl)piperazine-2-carboxylic acid and (R)-2-amino-5-phosphonopentanoate (D- AP5) displayed similar affinities at cerebellar NMDA receptors and medial striatal NMDA receptors. In contrast, the NMDA receptor antagonists (±)-6- (1H-Tetrazol-5-ylmethyl)decahydroisoquinoline-3-carboxylic acid, (S)-α- amino-5-(phosphonomethyl)[1,1'-biphenyl]-3-propanoic acid, and (±)-cis-4-(4- phenylbenzoyl) piperazine-2,3-dicarboxylic acid displayed varied, higher affinities at medial striatal NMDA receptors than at cerebellar NMDA receptors. For the five antagonists, there was a strong correlation (r = 0.9) between the cerebellar K(i)/medial striatum K(i) ratio and the NR2A K(i)/NR2B K(i) ratio for recombinant receptors. Thus, [³H]CGP39653 labels two pharmacologically distinct populations of NMDA receptors that have pharmacological and anatomical properties consistent with NR2A and NR2B subunits. Because native NR2A- and NR2B-containing receptors are pharmacologically distinct, it should be possible to develop NR2A- and NR2B- selective glutamate site antagonists.