Project Details



Denis Pare

Understanding memory, that is, how the brain stores information, is a major challenge of contemporary neuroscience. Indeed, the brain contains an astronomical number of nerve cells that communicate by specialized structures called synapses. Most neurons make synapses on hundreds to thousands of other neurons and reciprocally. Much evidence suggests that memory depends on changes in the strength or efficacy of individual synapses distributed across a large population of synapses. It was shown that when a neuron contributes to excite another nerve cell beyond a certain level, the synapse between these two cells becomes more efficient (or stronger). However, when synapses with such properties are introduced in computer models of neuronal networks, problems of stability develop because the reinforcement of synapses increases the likelihood that they will be further reinforced, leading the network into unchecked excitation. Thus, the question is how does the brain prevent runaway increases in the strength of synapses? This proposal tests the possibility that when particular synapses are strengthened, other synapses to the same cells are depressed. Thus, experience would modify the relative strength of synapses, but the total strength of synapses to any given neuron would remain stable. The proposed work will examine the intracellular mechanisms that allow the strength of individual synapses to change while keeping the total impact of synapses to target cells within normal bounds. This will be achieved by recording neurons in brain slices kept alive in a dish. Understanding how the brain keeps the weight of plastic synapses within normal bounds would have important implications for artificial intelligence and robotics where adaptable computer programs simulating neuronal networks constitute the most promising approach toward progress

Effective start/end date7/15/026/30/07


  • National Science Foundation: $649,903.00


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.