PhD Proposal: Nathan Clark

Computational Microscopy Applied to Determine Spatial Distribution of Membrane-Associated Proteins in Cells

Abstract

Modern cell biology relies heavily on fluorescent tags and signals to understand the spatial and temporal response of cells. Precisely knowing the spatial fluorophore distribution is critical. However, in fluorescence microscopy, the optical system images both light in the focal plan and from out-of-focus. The localization of labeled protein is degraded or lost because of the out-of-focus light. Our novel optical approach is to use analytic and numerical methods to properly account for out-of-focus light and reconstruct fluorophore distribution from a series of images. This reconstruction is obtained by modeling a cell and the optical characteristics of the microscope, synthesizing images of the cell that are representative of the microscope, and using nonlinear optimization to determine a model of the fluorophore distribution that produces images similar to a measured, through-focus image series. Our computational microscopy method will be applied in studying the erythrocyte cytoskeleton as well as the spatial distribution of leukocyte components relative to the inflammation process. Results of this project will enable researchers to use conventional fluorescence microscopy and common computer hardware to study cell physiology, signaling, and interaction with a resolution that was previously possible only through more complicated and expensive imaging modalities.