Distribution and synaptic localization of immunocytochemically identified NMDA receptor subunit proteins in sensory-motor and visual cortices of monkey and human

NMDA receptors are composed of multiple receptor subunit proteins, of which NMDAR1 appears to be a critical component for normal receptor function (Nakanishi, 1992). In this study, quantitative immunocytochemical methods were used at the light and electron microscopic levels to localize NMDAR1 subunits in the primary motor (M1) and somatic sensory (S1) cortex of monkeys, and in the primary visual cortices (V1) of monkey and human. Three principal features of NMDAR1 subunit organization were examined in detail in the monkey cortex: (1) the laminar and cellular distribution patterns, relying in part on double-labeling paradigms with the calcium-binding proteins parvalbumin (PV) and calretinin (CR) as markers for discrete subpopulations of GABAergic interneurons; (2) the codistribution of NMDAR1 subunits with non-NMDA ionotropic receptor subunits; (3) a quantitative assessment of the percentages of asymmetrical synapses in layers II/III, IV, and V/VI that were NMDAR1 immunoreactive. In monkey M1, S1, and V1, NMDAR1 immunoreactivity was present in all layers, localized primarily to large numbers of pyramidal cell somata and proximal apical dendrites, to presumptive spiny stellate cells in layer IV of VI, and to the vast majority (~80-90%) of PV-immunoreactive cells. By contrast, NMDAR1 immunoreactivity was present in only a very small percentage of the CR-immunoreactive cells (~6-9%). Colocalization with non-NMDA receptor subunits showed that all cells (100%) that contained GluR2/3 subunits were also NMDAR1 immunoreactive. In addition, the complete codistribution of GluR5/6/7 subunits with GluR2/3 subunits suggests, indirectly, that all GluR5/6/7-immunoreactive cells are also NMDAR1 immunoreactive. The laminar and cellular distribution patterns of immunostaining in human V1 were very similar to those in monkey V1. Electron microscopy of monkey sections confirmed an extensive dendritic and synaptic localization of NMDAR1 subunits. Labeling of synapses was present on asymmetrical postsynaptic densities associated with both dendritic shafts and spines. In supragranular layers of V1, a greater percentage of asymmetrical synapses were NMDAR1 immunopositive (44%) in comparison to layer IVCβ (34%) or deep layers (19%). In contrast, in area 3b of S1, the percentage of labeled synapses was greatest in layer IV (45%) in comparison to superficial (26%) and deep (37%) layers, while in M1, the percentages of labeled synapses were similar between superficial (46%) and deep (40%) layers. Taken together, these data indicate that NMDAR1-immunoreactive cells in neocortex represent a morphologically, functionally, and neurochemically heterogeneous population. In addition, the NMDAR1-immunopositive synapses represent a major proportion of the asymmetrical synapses in primary sensory and motor cortex, and have a distribution suggesting that in primates, they play a major role in mediating a diverse set of excitatory afferents.