Most experimental and theoretical work on synaptic plasticity in the brain has focused on long-term changes in synaptic strength. However, synaptic function is also influenced by activity over time scales of milliseconds to seconds. Synapses that exhibit such short-term plasticity are powerful computational elements that can have profound impact on cortical circuits. We propose a combined theoretical and experimental program to study the functional implication of cortical short-term synaptic plasticity for transmission of information across single synapses, synaptic integration by a postsynaptic neuron, and multi-neuron network dynamics. Using a combination of correlation and information theoretic techniques, we will explore the idea that synapses with short- term plasticity act as stochastic nonlinear temporal filters that transmit information efficiently and compactly by decorrelating presynaptic spike sequences. Theoretical studies will be combined with direct experimental tests to determine whether short-term plasticity is tuned to the statistics of presynaptic spike trains to maximize information transfer while conserving synaptic resources. We will study feedforward, recurrent and inhibitory pathways using mathematical models and dual intracellular recordings from slices of rat visual cortex. Preliminary data show significant differences in short-term plasticity at these three classes of synapses, and we will determine how this affects the information they transmit. Response-weighted averaging methods will be used to characterize the temporal selectivity of neuronal responses to afferent activity. Modeling studies and experiments using simultaneous intra- and extracellular stimulation will explore synaptic integration of multiple inputs to determine how short-term plasticity interacts with voltage-dependent conductances to affect selectivity. Finally, using mathematical descriptions developed from the data, we will build cortical circuit models to investigate the impact of short-term plasticity on network function. We will study what happens to the effects of short- term plasticity at single synapses when they are iterated across the layers of a feedforward network. Using network simulations, we will study what the experimental results on short-term plasticity imply about network dynamics, including stabilization of, and switching between, silent, oscillatory, and self- sustained activity states.
Showing the most recent 10 out of 19 publications
