“Decorating” Cells with Genetically Encoded Fluorescent Proteins - What Color Suits Your Experiment Best?

Abstract

A key aspect of simultaneous multicolor single molecule real-time (SMRT) imaging in the living cell is choosing the appropriate combinations of fluorescent proteins (FPs). Although ample information is available on fluorescent brightness, photo-stability, pH sensitivity and natural oligomerization state of FPs, interpolating published data into performance in live human cell experiments is difficult. Ultimately the effort of re-cloning of genetic constructs and advanced quantitative analysis to determine the most suitable FP favor the use of suboptimal label combinations. However, the use of high power to excite fluoresphores often results in shorter observation times, higher photo-damage and possibly the induction of ‘stress’ reaction by the cell. To determine the best FP combination for multicolor imaging relies on not only obvious factors like optimal spectral separation, best photostability and highest brightness but also the amount of excitation power needed to generate sufficient signal at minimal background noise level which could come from live cell autofluorescence, cross-talk between different channels and cross-excitation of FP.We tested 36 different FPs in a parameter space specific to their suitability in live cell multicolor single molecule imaging. Particular attention was paid to biased subcellular localization effects and performance of the FPs for labeling of dynamic (soluble FP & diffusive fusion protein), labeling of confined molecules (nuclear pore marker on nuclear envelope), the amount of fluorescence signal under low excitation power imaging conditions and cross excitation and cross emission at non-maximal photon output.We present different sets of FPs for different labeling purposes and discuss possible protocols for testing and specifying FP performance in low light live cell imaging conditions.