A biophysical model of cortical up and down states: roles of excitatory and inhibitory balance and H current.
1. ARL-NSMA, Univ. Arizona, Tucson, AZ;
2. Psychol, Univ. Arizona, Tucson, AZ
During slow-wave sleep and under anesthesia, cortical neurons oscillate between two states: an up state during which the membrane potential is high and spiking activity occurs, and a down state during which the membrane potential is low and virtually no spikes are generated. A similar oscillation between these two states also occurs in cortical brain slices where it is generated locally and propagates between regions of the cortex. It has been suggested that the transitions between these two states are generated through a balance between excitatory and inhibitory inputs. To test this hypothesis, a network model of excitatory and inhibitory cortical neurons was created.
Excitatory neurons were interconnected with AMPA/NMDA synapses to allow for reverberations and were also connected to inhibitory neurons which had GABA synapses back onto the excitatory neurons to create feedback inhibition. Random synaptic noise was added to all neurons to mimic the input from neurons outside of this network during the down state. Up states that resemble those measured in vivo were generated by injecting a current pulse to a small subset of the excitatory neurons. The balance of excitation and inhibition was adjusted so that firing rates of the excitatory neurons were in the range of 15-20 Hz, as seen in vivo. Under these conditions, up states were generated 70 percent of the time and terminated after 500-1500 ms of activity. The average membrane potential of excitatory neurons during the up state was around -57 mV, with a standard deviation of 5.5 mV. A precise balance between excitation and inhibition was necessary to obtain these statistics, and to successfully turn off these up states.
Additionally, an H current was added to the excitatory cells in the network. Because this current is activated during hyperpolarizations, an activation of inhibitory neurons just prior to the excitation of excitatory neurons could generate an up state 60 percent of the time with fewer neurons, while with no prior hyperpolarization, up states were generated only 30 percent of the time. This current may thus promote the propagation of the slow oscillations in vivo, if feed-forward inhibition arrives just prior to feed-forward excitation from a nearby cortical area. It also may act as a synchrony detector and allow for the generation of up states by activating a very small number of excitatory neurons synchronously. The propagation of up state activity will be further studied by creating multiple networks connected via excitatory and inhibitory connections. The effects of the modulation of the H-current by neuromodulators will also be studied.
Keyword (Complete): simulations; network; sleep; oscillations