"Seeing" network activity through membrane potential fluctuations

Alain Destexhe
CNRS, Gif
France

Abstract

Neocortical neurons intracellularly recorded during activated states in vivo show persistent voltage fluctuations, a depolarized level and a high membrane conductance. We investigated the computational properties of cortical neurons in such "high-conductance" states. We first show that the combination of voltage-dependent currents and high-conductance fluctuations induces a stochastic integrative mode in which the global efficacy, as well as the timing, of individual synaptic events become approximately independent of their position in the dendrites. As a consequence, cortical neurons are able to process synaptic inputs at high temporal resolution with little dependence on where these inputs are located in the dendritic tree. We illustrate this aspect by showing the detection of small correlation changes within a seemingly random afferent activity. We next introduce simplified models of this activity based on random processes on two global conductances (excitatory, inhibitory). This approach allows one to derive analytic expressions for the steady-state voltage distribution and by fitting these distributions to experimental data, it is possible to determine the average and variance of the conductances from current-clamp recordings. We show that these variables can be related to the mean activity and correlation between synaptic release events, or equivalently, between afferent neurons. Thus, this method allows one to detect important statistical properties of distributed network activity through analysis of the membrane potential fluctuations of single cells.

Return to the Theoretical Neuroscience Symposium Main Page