LAUSR

Converging experimental reports have shown that the firing dynamics of neural networks in several cortical brain areas can exhibit Up-Down synchronization regimes, spontaneously alternating between long periods of high collective firing activity (Up state) and long periods of relative silence (Down state). The molecular or cellular mechanisms that support the emergence of these reversible transitions are still uncertain. In addition to intrinsic mechanisms supported by the local neurons of the network, recent experimental studies have suggested that the astrocytes of the local network can actually control the emergence of Up-Down regimes. Here we propose and study a neural network model to explore the implication of astrocytes in this dynamical phenomenon. We consider three populations of cells: excitatory neurons, inhibitory neurons and astrocytes, interconnected by gliotransmission events, from neurons to astrocytes and back. We derive two models for this three-population system: a rate model and a stochastic spiking neural network with thousands of neurons and astrocytes. In numerical simulations of these three-population models, the presence of astrocytes is indeed observed to promote the emergence of Up-Down regimes with realistic characteristics. Linear stability analysis reveals that astrocytes in these models do not change the bifurcation structure of these systems, but change the localization of the bifurcations in the parameter space. Accordingly, with the addition of astrocytes, the network can enter a bistability region of the dynamics, where the Up-Down dynamical regime emerges. Simulations of the stochastic network model further evidence that astrocytes provide a stationary and stable background of gliotransmission events to the neurons, that triggers spontaneous transitions between synchronized Up and Down phases of neuronal firing. Taken together, our work provides a theoretical framework to test scenarios and hypotheses on the modulation of Up-Down dynamics by gliotransmission from astrocytes.