Impact of photobleaching of fluorescent proteins on FRET measurements under two-photon excitation

Förster resonance energy transfer (FRET) is a widely used technique for nanoscale molecular distance measurements, which makes FRET ideal for studying protein interactions and quaternary structure of protein complexes. In this work, we were interested in how photobleaching of donor and acceptor molecules affects the FRET results under various excitation conditions. We conducted a systematic study, under two-photon excitation, of the effects of the excitation power and the choice of excitation wavelengths upon the measured FRET efficiencies of multiplex protein constructs, consisting of one donor (D) and two acceptors (A) or one acceptor and a non-fluorescent tag (N), using both the kinetic theory of FRET and numerical simulations under given excitation conditions. We found that under low excitation power and properly chosen excitation wavelengths the relationship between the FRET efficiency of a trimeric construct ADA agrees within 2% with the FRET efficiency computed (via the kinetic theory of FRET in the absence of photobleaching) from two dimeric constructs ADN and NDA. By contrast, at higher excitation powers the FRET efficiencies changed significantly due to the photobleaching of both the donor (through direct excitation) and the acceptor (mostly through FRET-induced excitation). Based on these results and numerical simulations using a simple but competent algorithm, we developed guidelines for choosing appropriate experimental conditions for reliable FRET measurements, as well as for interpreting the results of existing experiments using different excitation schemes.