Dopamine-mediated stabilization of delay-period activity in a network model of prefrontal cortex.

Microelectrode studies in monkeys performing short-term memory tasks show the sustained elevated discharge of cortical neurons during the retention of recalled sensory information. Cortical cells that are part of memory networks are assumed to receive numerous inputs of excitatory as well as inhibitory nature and local as well as remote. Thus it is reasonable to postulate that the temporal and spatial summation of diverse inputs on any cell in an activated network will result in temporally discrete groups of spikes in its firing. The activation of a network in active memory supposedly increases the magnitude and diversity of those inputs and thus increases the discontinuities and frequency fluctuations in the firing of cells in the network. In this study we use a new method of analysis that allows the quantification of firing discontinuities in a spike train. We apply it to parietal cells recorded from monkeys during the performance of a tactile short-term memory task. In our method, time is divided into bins of equal duration and the measure of discontinuities is the total count of the number of transitions between consecutive time bins with and without spikes. The results of the analysis show that in many of the cells studied, discontinuities (transitions between spiking and nonspiking) reflect memory-related activity obscured in the measures of raw spike frequency over a wide range of frequencies. These cells show more firing transitions in active short-term memory than in baseline (intertrial) conditions.