Fluorescent Protein Fret as a Probe of Protein Conformation in Vivo and in Vitro

Abstract

While methods for the determination of protein structure in vitro are highly advanced, the question always remains as to whether the observations are representative of the molecule in its native environment. Fluorescence resonance energy transfer (FRET) is widely used to probe protein structure both in vitro and in vivo. Because of their ease of use, in vivo FRET measurements frequently rely on the cyan and yellow fluorescent proteins (i.e. CFP donor and YFP acceptor). The severe spectral overlap and significant direct excitation of the acceptor complicate extraction of true FRET efficiency. Furthermore, the large size and poor photophysical properties of fluorescent proteins (FPs) make them a less than ideal FRET pair. To benchmark the reliability of FP-FRET as a probe of protein conformation in vivo, we characterized a series of CFP-YFP tandems with varying linker lengths. Ensemble and single molecule fluorescence were used to measure FRET in vitro. We then measured FRET in live cells with total internal reflection fluorescence microscopy. Only by using a single filter cube with a ratiometric image splitter, were we able to get agreement between in vitro and in vivo measurements. However, measurements of the radius of hydration, using analytical ultracentrifugation and chromatography, revealed a much different size distribution than was suggested by the FRET dependence on the linker length between CFP and YFP. Our studies suggest that the conformation of our protein test set is the same in live cells and in solution. In vitro and in vivo measurements for each construct were in good agreement. Unfortunately, distances derived from FP-FRET appear to misrepresent the actual interchromophore distance.