Development of screening systems for enhanced fluorescent protein engineering

Abstract

Fluorescent proteins (FPs) have dramatically advanced life science research. Since their discovery they have become invaluable tools for imaging living systems and have been developed into precision instruments for measuring normally invisible events such as fluctuations in calcium ion concentration, protein-protein interactions (PPi), and even membrane voltage. More surprisingly, FPs have now been converted into optogenetic actuators capable of manipulating the biochemistry of the cell. There is substantial interest in improving both the diversity and quality of available FPs because of their great utility and potential. In this thesis, I describe my efforts to design better methods to easily improve FPs and I use these methods to create a variety of new FP variants. First, I explored the potential of developing FPs as tandem dimers. I used directed evolution to create a series of heterodimeric FPs called the vine Tomatoes (vTs). Specifically, I created green-green (GGvT), green-red (GRvT), and red-red (RRvT) heterodimers by genetically fusing two tightly dimerizing FP domains and then evolving as a pair. This allows the two monomers to differentiate, creating tandem heterodimers with advanced characteristics such as exceptionally high FRET efficiency in GRvT of 99%, and the brightest red fluorescent protein to date with RRvT at 120. Next I developed a robot-assisted screening system for photostability screening. This simplified screening process led to the development of Citrine2, a variant with 9-fold improvement in photostability relative to its precursor, mCitrine. I also observed that concentration plays a significant role in photostability, and I so I attempted to modify this property. With only five mutations, the concentration dependence of photostability switched from being an inverse relationship in mCitrine, to a direct relationship in Citrine2. From there I switched to developing new screening systems for photocleavable proteins. Several different screening systems based on bimolecular fluorescence complementation and FRET were developed to screen directly for the photoinduced dissociation of a photocleavable protein (PhoCl) developed in our lab. A second photocleavable protein called SplitOr was created from PSmOrange2. The goal was to develop a spectrally orthogonal photocleavable protein that could be photocleaved with wavelengths of light that do not cause PhoCl photocleavage. PSmOrange2 was circularly permuted and its fluorescence and photoconversion properties rescued, creating three new versions of SplitOr with photophysical characteristics that suggest photocleavage is occurring. Altogether, these projects have advanced the field of FP development. I created three research-ready FPs, RRvT, GRvT, and Citrine2; I developed and characterized a rapid, low cost, robot-assisted illumination system which will aid our lab and others in the development of photostable FP variants; I advanced our knowledge of photocleavable proteins by developing an evolution system for the photocleavable protein PhoCl; and finally I created an orthogonal photocleavable protein that may find use as an optogenetic actuator.