Recovery time after inhibition in pyramidal neurons controls network frequency.

<p><b>A</b> Recovery time after inhibition was determined by simulating an inhibitory synaptic conductance waveform in isolated model pyramidal neurons and in quiescent biological neurons <i>in vitro</i> via dynamic clamp. The example illustrated was taken from a representative dynamic clamp experiment. These simulated synaptic conductances varied in their magnitude, decay kinetics, and in the amount of tonic background conductance present. <b>B</b> Predictions of network frequency (lines), taken as the inverse of measured recovery time after inhibition, match the frequency of network oscillations determined in full simulations (solid dots) when RSP neurons are in either a low conductance state (gray) or high conductance state (blue). <b>C</b> The derivative of membrane voltage following inhibition in model pyramidal neurons in the oscillating model network depends upon the conductance state of those cells. RSP neurons in a low-conductance state (gray, green) recover slowly following inhibition, while RSP neurons in a high-conductance state (blue, red) recover relatively quickly. <b>D</b> Mean values of the inverse of recovery time after inhibition in model RSP neurons (<i>i</i>) and biological layer II/III pyramidal neurons (<i>ii</i>) as a function of decay time constant and synapse magnitude (axes) and background conductance (surfaces). Recovery time is controlled by the decay time constant of inhibition and total level of membrane conductance but not the magnitude of phasic conductance in model and biological neurons.