A role of synaptic variation depending on precise timings of pre- and postsynaptic depolarization in suppression of disturbance due to tail bending for accurate electrolocation

In order to clarify a role of depolarization timing-dependent learning in s uppression of reafferent signals of weakly electric fish, we made a neural network model of electrosensory lateral-line lobe (ELL) which consists of b asilar pyramidal (bp) cells. We propose a learning rule depending on tempor al correlation between pre- and post-synaptic inputs. It is shown that the learning depending on the timings of pre- and postsynaptic depolarization p lays an essential role in the discrimination of the synchronous signals fro m noisy background descending signals. The active property of the apical de ndrite, by which the depolarization of soma reliably propagates backward fr om soma to dendrite, is necessary to detect the temporal correlation betwee n the ascending and descending signals. (C) 2000 Elsevier Science B.V. All rights reserved.