Our previous studies on the survival-promoting influence of elevated concentrations of extracellular K⁺ ([K⁺]_e) on cultured cerebellar granule cells led to the proposal that depolarization in vitro mimics the effect of the earliest afferent inputs received by the granule cells in vivo. This, in turn, might be mediated through the stimulation of excitatory amino acid receptors, in particular the N-methyl-d-aspartate-preferring subtype gating ion channels which are also permeable to Ca²⁺. Here we report that N-methyl-d-aspartate indeed has a dramatic effect on the survival in culture of cells derived from dissociated cerebella of 7–8-day-old rats and cultured in media containing ‘low’ [K⁺]_e (5–15 mM). In addition to the visual inspection of the cultures, the effect of N-methyl-d-aspartate was quantitatively evaluated, using estimates related to the number of viable cells (determination of DNA and of reduction rate of a tetrazolium salt). Furthermore, proteins which are relatively enriched in either nerve cells (neuronal cell adhesion molecule, D3-protein and synaptin) or in glia (glutamine synthetase) were also measured. The findings showed that the rescue of cells by N-methyl-d-aspartate involved primarily nerve cells and that the survival requirement for N-methyl-d-aspartate, as for high K⁺, developed between 2 and 4 days in vitro. The effect depended on both the concentration of N-methyl-d-aspartate and the degree of depolarization of the cells: both the potency and the efficacy of N-methyl-d-aspartate were increased as [K⁺]_e was raised from 5 to 15 mM, at which range K⁺ on its own has little if any influence on granule cell survival. These characteristics are consistent with the voltage-dependence of ion conductance through the N-methyl-d-aspartate receptor-linked channel. The most pronounced effect of N-methyl-d-aspartate was obtained in the presence of 15 mM K⁺, when cell survival approached that obtained in ‘control’ cultures (grown in 25 mM K⁺-containing media without N-methyl-d-aspartate), and the potency of N-methyl-d-aspartate (half-maximal effective concentration, ec₅₀, about 20μM) was similar to its known affinity in binding to cerebral membranes. The effect of N-methyl-d-aspartate was blocked by the specific receptor antagonist 2-amino-5-phosphonovalerate, which also reduced the limited survival of cells in cultures grown in ‘low’ K⁺ in the absence of N-methyl-d-aspartate. However, 2-amino-5-phosphonovalerate did not interfere with granule cell survival in 25 mM K⁺-containing medium, indicating that, under the latter conditions, Ca²⁺ entry through voltage-sensitive channels activated by depolarization, rather than the N-methyl-d-aspartate receptor-linked ion channels, is involved in the rescue of the cells. In a medium containing 25 mM K⁺, no toxic effects of N-methyl-d-aspartate were detectable on granule cells cultured for about a week.

Comparison of the influence of N-methyl-d-aspartate on cultured granule cells with observations either on cerebellar slices or in vivo indicates that the consequences of the N-methyl-d-aspartate receptor stimulation depends upon the stage of maturation of the granule cells. We suggest that trophic influences mediated through the stimulation of N-methyl-d-aspartate receptors may not be unique to cerebellar granule cells, but may also be manifested—depending on the developmental stage—in other types of cells through the nervous system.