Cell-type specific sensing and control of firing rate statistics via channel dynamics

Abstract

Neurons maintain functionality through homeostatic regulation of spiking activity over extended timescales. Calcium dependent conductance expression is known to regulate mean firing rate, but this is not sufficient to ensure dynamic range in spiking activity and sensitivity to input. This raises the question of whether firing rate variance can be sensed and controlled intracellularly. Using conductance-based models, we demonstrate that time-averaged intracellular calcium dynamics inherently provide a direct readout of both the mean and variance of spiking activity. We show that calcium-based feedback regulation of membrane conductance density can therefore jointly stabilize firing rate mean and variance against input disturbances. Because tuning maximal conductances modulates the underlying relationship between rate statistics, a cells homeostatic response is statedependent rather than fixed. As a consequence, cell-type-specific mixtures of ionic conductances yield distinct homeostatic modalities, implying that cell type dictates homeostatic behaviour as well as spiking and integrative properties.