Modeling of Calcium Dynamics in Spines

California Institute Of Technology, Pasadena CA

Investigators

Linked publications & trials

Abstract

The long-term goal of this project is to develop a model to explain the critical time window involved in the pairing protocol used for the induction of long-term potentiation (LTP) and long-term depression (LTD). This will require a detailed analysis of dendritic Ca 2+ dynamics in dendritic spines and an investigation of the Ca2+-mediated signal transduction cascades initiated by neural activity. The model will draw on experimental data from the other projects in this proposal on the structure of the spine and the location of important molecules (Project 2: Weinberg), the biochemical reactions that occur in the spine following the influx of Ca 2+ (Project 3: Kennedy) and the direct measurements of Ca 2+ dynamics using 2-photon microscopy (Project 4: Svoboda). These measurements will be incorporated into MCelI, a Monte Carlo computer program that simulates subcellular signaling by following the random walk and interactions between diffusible molecules. Three specific aims will be pursued in parallel. First a 5 _mx 5 _m ? 5 _m volume of hippocampal area CA1 neuropil from mouse will be reconstructed to serve as the anatomical substrate of the simulations. This will allow simulations of neurotransmitter release and diffusion in the extracellular space to be accurately modeled. Second, we will use the model to estimate [Ca 2+] in small functional microdomains, such as in the postsynaptic density. Since the reconstruction will likely contain around 100 spines of varying shapes and sizes, we will also obtain estimates of variability in the system. The model will be able to measure the activation of calmodulin (CAM) following an EPSP, an action potential, or both occurring with a temporal offset, and to then allow activated CaM to bind to, and activate CaMKII. Third, we will examine theoretically how much variability is expected to result from the stochastic schemes implemented by MCell to determine how many simulation runs will be needed to obtain accurate estimates. We will develop an analytical approximation to the stochastic model to determine the statistical distribution of system dynamics. In preliminary studies, we used computer simulation of receptor activation and calcium dynamics at glutamatergic synapses in a simplified model of a segment of dendrite with dendritic spine and PSD. Since this signal transduction cascade implies interactions between multiple diffusible species, the new modeling capabilities for MCell proposed in Core Facility 1 are prerequisite to this project.