In a neural cell line, the secretion of excitatory amino acids in response to a depolarizing stimulus is potentiated by the addition of serotonin. The duration of this potentiation is dependent on the strength of the stimulus. Persistent secretory potentiation induced by a strong stimulus requires the activation of both serotonin and NMDA receptors. Inhibiting the NMDA receptor during serotonin presentation prevented the induction of potentiation. The temporal characteristic of the potentiation is correlated with the elevation of CAMP levels. Serotonin exposure while inhibiting NMDA receptors results in a transient elevation of CAMP levels, whereas coactivation with NMDA and serotonin results in a persistent elevation of CAMP. Thus, it is possible to obtain potentiation of secretion in a single cell either transiently or persistently. The timing of potentiated responses in this system is of the same magnitude as that in similar systems used as models for short-term and long-term memory. tntroduction

Many experimental model systems have been used to study memory. Aplysia siphon withdrawal(Kandel and Schwartz, 1982), hippocampal long-term potentiation (Teyler and DiScenna, 1987; Brown et al., 1988), animal behavior(Kapp and Pascoe, 1986), and even bacterial chemotaxis(Koshland, 1983) have provided critical information about aspects of memory. Hebb (1949) postulated that memory can be stored in the brain through permanent changes in synaptic strength. This attractive model for mammalian memory is often referred to as synaptic plasticity. Long-term potentiation(LTP), the increase in synaptic efficacy, has served as an effective model for mammalian memory. An enormous amount of information has been obtained using hippocampal slices. In particular, electrophysiological and pharmacological studies of brain slices have identified a postsynaptic role in LTP for N-methyl-o-aspartate(NMDA) receptors (Cotman et al., 1988; Kauer et al., 1988), calcium (Lynch et al., 1983; Malenka et al., 1988), protein kinase C (Akers et al., 1986; Malinow et al., 1988), calcium/calmodium-dependent kinase (Malenka et al., 1989; Malinow et al., 1989), and calpain I (Siman et al., 1987). Recently, presynaptic changes have been found to contribute importantly to LTP (Barinaga, 1990; Zalutsky and Nicoll, 1990).