The proposed experiments of this Merit Review competitive renewal are part of the effort by many labs to determine how central nervous system neurons encode their synaptic inputs. In particular, the experiments are designed to determine how specific neurons (or abnormal function of those neurons) contribute to synchronous network activity that occurs during both normal signal processing and during seizure activity. The major questions ask 'What temporal aspects of natural patterns of synaptic activity are most effective for driving a neuron's action potential firing?' and 'What are the governing underlying mechanisms?'. The focus is on the responses of principal cells in neocortex (pyramidal neurons) to local circuit activity. The convergent synaptic activity generated by cortical networks has a frequency spectrum similar to broad- band noise. We have preliminary data that the pyramidal neurons show resonant firing to this type of input. That is, specific frequency components of their input drive firing better then others. Two input frequencies cause the resonance: theta (~7 Hz) and fast ripple (~250 Hz). We also have preliminary evidence that interactions between frequency components in the input cause differences in the resonance (i.e. the resonance is nonlinear). Additional preliminary studies point to certain potassium conductance mechanisms as being critical for the type and amount of resonance a pyramidal neuron has. A mutation in the gene (KCNQ) for one of these potassium channels (Kv7) is known to cause some forms of human epilepsy. There are three sets of experiments. In the first we will determine the basic interactions of the input frequencies that result in resonant firing. In the second set we will determine the influence of dendritic filtering on the resonant firing. In the third set we will determine how the resonant firing is transmitted across synapses to the different targets of the pyramidal neurons. In all three sets of experiments we will determine the role of specific potassium channels on the resonance. The methods employ patch pipettes (or sharp electrodes in some experiments) to record from visualized neurons in brain slices. Neurons are stimulated in current clamp and dynamic clamp using computer-generated waves that simulate the statistics of ongoing synaptic activity generated by the local cortical circuit (focusing mostly on excitatory inputs). Potassium channel blockers and openers and receptor agonists and antagonists are applied to the neurons during the electrical recording. We also add artificial potassium conductances to neurons with a method we devised called spike-triggered dynamic clamp. Anatomical classification of neurons is done by intracellular staining with biocytin. The second set of experiments uses photolysis to uncage glutamate to stimulate the dendrites. In the third set of experiments we simultaneously record from synaptically connected pairs of neurons.
Potential Impact on Veterans Health Care Traumatic brain injury is a major cause of epilepsy in our veteran patient population. Epilepsy caused by injury to the neocortex is hard to manage medically and the surgical options are more restricted than for temporal lobe epilepsy. So we need to find better ways to treat this problem. Particular frequency components of neocortical network activity are specifically associated with seizures and promote their initiation and propagation. Other frequencies may be suppressive. We expect that the proposed studies will close major gaps in our basic understanding of both normal and pathological neocortical oscillations and provide a rational basis for the design of new treatments for epilepsy.