Date of Award

12-2016

Level of Access Assigned by Author

Open-Access Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Physics

Advisor

Samuel Hess

Second Committee Member

Raymond Astumian

Third Committee Member

Robert Meulenberg

Additional Committee Members

George Bernhardt

Paul Millard

Abstract

Fluorescence Photoactivation Localization Microscopy(FPALM) and other super resolution localization microscopy techniques can resolve structures with nanoscale resolution. Unlike techniques of electron microscopy, they are also compatible with live cell and live animal studies, making FPALM and related techniques ideal for answering questions about the dynamic nature of molecular biology in living systems. Many processes in biology occur on rapid sub second time scales requiring the imaging technique to be capable of resolving these processes not just with a high enough spatial resolution, but with an appropriate temporal resolution. To that end, this Dissertation in part investigates high speed FPALM as an experimental technique showing images can be reconstructed with effective temporal resolutions of 0.1s. Using fluorescent proteins attached to an influenza viral protein, hemagglutinin(HA), questions of protein clustering and cluster dynamics on the host cell membrane are explored. The results indicate that these HA clusters may be more dynamic than previously thought. The principle disadvantage of the increased speed of imaging is the reduction in information that comes through collecting fewer photons to localize each molecule, and fewer molecules overall. As the molecules become dimmer, they also become harder to identify using conventional identification algorithms. Tools from machine learning and computer vision such as artificial neural networks(ANNs) have been shown to be adept at object identification. Here a method for repeatedly training an ANN is investigated. This method is shown to have exceptional performance on simulations indicating that it can be regarded as a method of high fidelity, even in the presence of weakly fluorescent molecules. Development of this technique can be used to recover more molecules from data sets with weaker molecular fluorescence, such as those obtained with high speed imaging, allowing for higher sampling, and overall higher spatial resolution of the final image. The combination of a high speed experimental technique coupled with a sensitive and robust identification algorithm allow FPALM and related techniques to probe questions of fast biological processes while limiting the sacrifice to spatial resolution inherent in high speed techniques.

Share