activity was captured for 2:12 minutes (1 frame captured every 0.11 seconds), after which the AMPA receptor blocker CNQX was added to the chamber at a final concentration of 20 μM to suppress excitatory synaptic currents and an additional 3:51 minutes were recorded. During the recording of basal activity, neurons in both chambers demonstrated isolated spontaneous activity, as well as coordinated network activity between and within compartments (Fig. 5 and Video S1†). Importantly, this circuit activity was often coordinated between chambers, indicating that axons traversed the microchannels and formed synaptic contacts with neurons within the opposing chamber (Fig. 5A and B). Excitatory neurons demonstrated low frequency circuit bursts that influenced inhibitory neuron firing. Inhibitory neurons displayed frequent network bursts, with many correlated to the activity of the excitatory neurons. After CNQX application, excitatory neuronal firing was no longer able to coordinate large circuit bursting activity (Fig. 5C), with only one exception, and only after 3 minutes of no bursting activity. To confirm that the effect was due to CNQX and not mechanical disruption of the cells from reagent dispensing, we added 10 μl HEPES (vehicle) as a negative control. When vehicle-only was added, the burst calcium spikes likely driven by
