Comparison of CaMKinase II activation in a dendritic spine computed with deterministic and stochastic models of the NMDA synaptic conductance

In models of long-term potentiation the NMDA conductance is usually computed deterministically. However, the actual number of open NMDA receptor channels at a synapse is small, so a deterministic representation may not be valid. Here NMDA synaptic conductances computed stochastically with MCell were used in a dentate granule cell model that computed calcium influx and subsequent CaMKinase II activation in a dendritic spine following LTP induction conditions. Spine head calcium concentration and levels of CaMKinase II activation were highly variable with different stochastic simulations of NMDA channel openings. This variability in CaMKinase II activity levels due to stochastic NMDA channel openings may play an important role in LTP induction in individual spines. (C) 2000 Elsevier Science B.V. All rights reserved.