The fascinating capabilities of neural networks arise from the interplay between the intrinsic properties of neurons and their interconnections. Synapses are the first place where the neuron receives information. In order to understand information processing and storage in the brain, it is important to first understand the properties of synapses that define their response to stimulation and how these properties are altered during development or by experience. Recent modeling work by the PI has shown that the microscopic details of glutamate receptor activation have important implications for synaptic plasticity. The studies conducted in the present project will extend this research in several important directions. First, a combination of experiment (focal glutamate uncaging using two-photon microscopy) and modeling (Monte Carlo models of stochastic receptor activation) will be used to assess how the differences in the biophysical properties of the NMDA subtype of glutamate receptor shape the synaptic response. Second, a large scale computational model of signal transduction networks involved in synaptic plasticity will be developed to parse out the differential contributions of different glutamate receptor subtypes in long-term modifications of synaptic strength. The models will be based on detailed anatomical and physiological data, and the results will be compared with experiments. Results of the proposed research will shed light on some of the complex features of signaling events that drive bidirectional modification of synaptic strength. The investigators include a theorist and experimental neuroscientist and the project thus serves as a model for how distinct approaches can be brought together into a cohesive effort to address a general problem in neuroscience. The results of this project are also expected to generate interest from a diverse community of researchers ranging from mathematics to molecular, cellular and systems neuroscience. The research is part of an initiative at Duke to promote multidisciplinary approaches to multi-scale modeling, computation and analysis. The funding will also support the PI in the development of a course on theoretical neuroscience at Duke for students from neurobiology, biomedical engineering, and mathematics.
