N-methyl-D-aspartate receptors at parallel fiber synapses in the dorsal cochlear nucleus.

1. N-methyl-D-aspartate (NMDA) binding and NMDA-receptors immunolocalization experiments have revealed an enhanced expression of these receptors in the outer two layers of the dorsal cochlear nucleus (DCN). The distribution of the receptors is congruent with the distribution of synapses produced by the granule cell-parallel fiber system. To determine the functional distribution and contribution of NMDA receptors at parallel fiber synapses, synaptic responses to parallel fiber stimulation were studied in in vitro brain slice preparations of the guinea pig and rat dorsal cochlear nucleus. 2. The field potential response to parallel fiber stimulation in guinea pigs includes three postsynaptic components. The short latency components (the P3(2) and N2(2)) are blocked by general excitatory receptor antagonists, including the non-NMDA-receptor blockers 6,7-dinitroquinoxaline-2,3-dione (DNQX) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), but are insensitive to NMDA-receptor antagonists. 3. A slower component (P4(2)) is revealed when the slices are washed with a low magnesium solution to eliminate the magnesium block of currents through NMDA receptors. This slow component is reduced by D- or DL-2-amino-5-phosphonovaleric acid (D-APV, DL-APV) and 3-[(+/-)-2-carboxypiperazine-4-yl] propyl-1-phosphonate, but is not blocked by DNQX or CNQX. Eliminating the voltage dependence of the NMDA receptors also results in a complex oscillatory response in some slices. This response exhibits the same pharmacological sensitivity as the slow potential. The pharmacologic sensitivity to NMDA-receptor antagonists suggest that the slow component (P4(2)) and the associated oscillatory response are mediated through activation of NMDA receptors. 4. Current source-density analysis of the parallel fiber-evoked field potentials was carried out to determine the relative spatial distributions of the fast and slow synaptic currents. Both synaptic components were associated with a superficial current sink and a deeper current source, localized within the superficial 250 microM of the nucleus. The slow (APV-sensitive) current was slightly shifted in depth relative to the fast (DNQX-sensitive) current in three of five slices with the maximum current sink and source occurring approximately 16 microns further from the surface of the DCN. These data suggest that either the NMDA receptors are not present at all of the synapses that generate the fast non-NMDA currents or that postsynaptic cells with different dendritic distributions have different densities of NMDA receptors. 5. The types of cells in layers 1 and 2 exhibiting NMDA-receptor-mediated synaptic potentials were investigated. Intracellular recordings with sharp electrodes in guinea pig slices showed that eliminating the voltage dependence of the NMDA receptors in low magnesium revealed a slow excitatory postsynaptic potential (EPSP) in both simple and complex spiking cells. The late phase of the EPSP could be reduced by APV in both cell types. These results could be explained by NMDA receptors on the postsynaptic cells or by NMDA receptors on excitatory interneurons. Attempts to demonstrate an appropriate voltage dependence of the parallel fiber synaptic response in normal magnesium medium under current clamp were confounded by the intrinsic voltage-dependent conductances of the cells. 6. To determine whether NMDA receptors were present on postsynaptic cells, the direct sensitivity of DCN cells to NMDA application was examined during intracellular recording. Both simple spiking and complex spiking cells responded to NMDA with depolarization. The response to NMDA persisted when non-NMDA receptors were blocked with CNQX or DNQX. However in all cells tested, the response to NMDA was blocked by APV. These experiments further support the postsynaptic localization of NMDA receptors on both simple and complex spiking cells. (ABSTRACT TRUNCATED)