Cellular localization and laminar distribution of NMDAR1 mRNA in the rat cerebral cortex.

N-methyl-D-aspartate (NMDA) receptors, which play a critical role in many cortical functions, are composed of a heteromeric assembly of different subunits: of these, the NMDA receptor subunit 1 (NMDAR1) is a constant component of, and thus an excellent marker for, NMDA receptors. In this study, we have investigated the cellular localization and laminar distribution of NMDAR1 mRNA in the cerebral cortex of adult rats by in situ hybridization histochemistry with a S-35-labeled cRNA probe. Specificity and background levels were determined in adjacent sections incubated with a S-35-labeled sense RNA. In sections incubated with the antisense RNA probe, specific hybridization signal was observed in a large number of cells. Some cells, however, did not appear to contain NMDAR1 mRNA. The vast majority of these unlabeled cells were small, suggesting that they are astrocytes or other small nonneuronal cells. Double-labeling studies with in situ hybridization histochemistry and immunocytochemistry with antibodies to glial fibrillary acidic protein (GFAP) showed that about 95.7% of the GFAP-positive cells did not express NMDAR1 mRNA, indicating that virtually all astrocytes do not contain this transcript. A semiquantitative evaluation of cortical neurons, defined as those cells larger than the GFAP-positive astrocytes, revealed that about 80% were associated with silver grains. The number of silver grains associated with every neuron was determined from sections exposed for 15 days, the background level was subtracted, and all labeled neurons were grouped into five groups: A (less than or equal to 10 grains), B (11-20 grains), C (21-30 grains), D (31-40 grains), and E (>40 grains). The number of neurons belonging to each group was then evaluated according to their occurrence in each cortical layer. In layer I all labeled neurons were in group A, whereas in layers II-III and V-VI positive neurons were in groups A-E. In layer ni most neurons were in groups A and B, whereas only a few were in group E. These observations indicate that 1) virtually all cortical cells containing NMDAR1 mRNA in adult rats are neurons; 2) about 80% of all cortical neurons express NMDAR1 mRNA; and 3) labeled neurons can be divided into several groups on the basis of NMDAR1 mRNA levels expressed, which presumably reflect the number of NMDA receptors. The existence of neurons with a different number of receptors may be a critical factor for determining the physiological effect of NMDA receptor activation. In addition, the differential laminar distribution of neurons with different intensities of hybridization signal suggests that NMDA receptors have a different influence at different stages of cortical processing. (C) 1994 Wiley-Liss, Inc.