Networks of inhibitory interneurons could be largely responsible for the propagation of higher-frequency activity in the cortex, concludes a new report from David McCormick's laboratory.

Most synapses onto neurons in the neocortex are supplied by other cortical neurons, forming 'recurrent' or 'feedback' networks. The encoding of information in cortical networks depends on the firing rate of individual neurons and on the temporal precision and relative timing of action potentials between neurons. Cortical networks can generate activity at a wide range of frequencies, but although slow, sleep-related oscillations have been studied in some detail, little is known about the production of higher-frequency cortical oscillations.

Hasenstaub et al. looked at spontaneously occurring recurrent network activity in the ferret dorsal prefrontal cortex. They recorded postsynaptic potentials or currents from cortical pyramidal neurons, showing that GABAA (γ-aminobutyric acid type A) receptor-mediated inhibition is required for the higher-frequency components of network-driven synaptic activity.

During periods of recurrent network activity (so-called UP states), inhibitory postsynaptic potentials carried more power than excitatory potentials at frequencies above 10 Hz, particularly in the gamma (30–80 Hz) frequency range, and were more synchronized. Fast-spiking inhibitory interneurons discharged strongly during UP states, and a robust relationship was detected between the discharge probability of these cells and the phase of the gamma oscillation in the local field potential. By injecting repeating patterns of excitatory and inhibitory conductances into pyramidal neurons — matching the characteristics of excitation and inhibition during actual UP states — the authors showed that inhibition is important in determining the timing and probability of action potential generation.

This study builds on previous findings to show that inhibitory interneurons can control the precise timing of action potentials in postsynaptic cells, thereby driving cortical network synchronization. In this way, inhibitory networks direct the flow of information in the neocortex.