Synaptic inhibition controls network synchrony during thalamic oscillations
Oscillations are ubiquitous in neural systems, and inhibitory synapses often regulate their properties. This regulation commonly manifests in two ways: (1) inhibitory neurons form interconnected networks, and (2) these networks rhythmically inhibit excitatory neurons. Both inhibitory motifs are present in the thalamus, where inhibitory neurons of the thalamic reticular (RE) nucleus inhibit each other and thalamocortical (TC) relay neurons during oscillations, such as spindles, characteristic of slow-wave-sleep, and spike-wave discharge (SWD), which occurs in the context of generalized absence epilepsy. In thalamic slices, pharmacologic and genetic manipulations of inhibitory synaptic transmission can transform spindle-like oscillations into hypersynchronous, spike-wave-like epileptiform discharges. Therefore, we used thalamic slices to study how synaptic inhibition alters the activity of individual neurons, and how these cellular effects ultimately reshape network activity. We found that: (1) Using mutant mice in which the thalamic effects of the anti-absence drug clonazepam are restricted to either RE or TC neurons, enhancing intra-RE inhibition is necessary and sufficient to suppress synchronized, rhythmic activity in thalamic slices. (2) In intracellular recordings from RE neurons during thalamic oscillations, intra-RE inhibition limits the number and synchrony, but not the duration, of RE cell bursts. (3) In simulated thalamic networks, simply by vetoing occasional RE neuron bursts, intra-RE inhibition can prevent GABABR-activation and emergent network synchronization, which both drive epileptiform activity. (4) In recursive inhibitory networks, inhibitory coupling generated emergent gamma-frequency oscillations in multiple cell types, and desynchronized spindle-frequency activity in RE cells, but not in other cell types. (5) Modest increases in the amplitude or duration of IPSCs in TC neurons accelerate and desynchronize rebound bursts, and reduce the period and synchrony of oscillations in thalamic slices. These findings elucidate cellular and network-level mechanisms underlying potentially anti-epileptic effects of intra-RE inhibition and, to a lesser extent, of connections from RE to TC neurons. They also demonstrate that fully characterizing inhibitory connections requires an understanding not only of synaptic dynamics, but of dynamics at the level of neurons and networks. Finally, our results may be relevant to oscillations outside of the thalamus, since inhibitory motifs similar to those we have studied, occur throughout the brain.