Kim "Avrama" Blackwell
Speaker of Workshop 2
Will talk about: Simulating the long time scales and large molecules numbers involved in synaptic plasticity
Kim Blackwell received a PhD in Bioengineering and a V.M.D. in veterinary medicine from the University of Pennsylvania. Her research interests have revolved around memory storage in neurons and neuronal networks. Her professional career began at the not-for-profit Environmental Research Institute of Michigan, where she began developing artificial neural networks for pattern recognition. Unsatisfied with the non-biophysical learning rules used in ANN, she began investigating the biophysical and biochemical mechanisms of long term memory storage in neurons using computational and electrophysiological techniques. In 1996 Blackwell moved to the Krasnow Institute at George Mason University, where she is now a professor in the Molecular Neuroscience Department. Her research presently focuses on calcium dynamics and signaling pathways underlying memory storage, synaptic plasticity, and the ability of hippocampal and striatal neurons to discriminate spatio-temporal patterns of inputs. Because of the importance of dynamics, she has developed the software tools Chemesis and NeuroRD for large scale dynamical modeling of the signaling pathways in neurons underlying memory storage. She continues to integrate computational modeling with experiments, both in her own lab and through collaborations.
The mechanisms underlying discrimination of temporal pattern and spatial specificity are some of the most critical, yet least understood aspects of synaptic plasticity, memory storage and information processing in neurons. Computer simulation of these mechanisms require tens of minutes of simulation time, because stimulation patterns for induction of synaptic plasticity often span minutes and the time course of activation of critical kinases range from seconds to 10s of minutes. Spatial specificity of various molecules has an equally large range, and investigations require simulation of large reaction-diffusion systems with many molecular populations in spiny dendrites of 10 to 100 or more microns in length. For example, though calcium elevations are limited to stimulated spines; other molecules, such as Ras, diffuse several microns, or even further. In order to simulate systems with these diverse spatial and temporal scales, we created NeuroRD, software that extends the Gillespie tau-leap algorithm for stochastic reactions into the diffusion domain. Using NeuroRD to efficiently simulate stochastic interactions both within spines and between spines arranged along a dendrite, we investigate the mechanisms controlling spatial specificity of diffusible second messengers and protein kinases. We address the role of buffers, both diffusible and immobile, as well as enzymatic degradation. These large scale reaction-diffusion models have potential utility multi-scale modeling that interfaces signaling pathways with models of neuronal electrical activity.