Fast and Accurate FRET Quantification through Computation of the Minimal Fraction of Donor (mfD) from TCSPC Data

The study of the spatial and temporal dynamics of molecular interactions using FRET-FLIM is often compromised by the large number of photons required to fit the multiple-lifetime decay of the donor population in each pixel of an image. Long acquisitions prevent interacting dynamics to be detected in an image, while the use of high excitation intensities causes bleaching or unexpected cell responses. The computation of the minimal fraction of donor molecules (mfD) undergoing FRET is a non-fitting approach that allows quantification of molecular interactions where the complexity associated to fitting a fluorescent decay with fretting and non-fretting components hampers quantification.We have established the experimental conditions on which quantitative FRET analysis based on mfD computation is preferable to fitting the fluorescent decay to a model with a weighted non-linear least-squares algorithm. The accuracy of the quantifying parameters as a function of the fraction of donor molecules involved in the interaction and the number of photons was studied for both fitting and non-fitting strategies using simulated TCSPC data that had been validated against actual experiments with FRETting constructs of fluorescent proteins (mTFP1-YFP) expressed in living cells.In summary, the validity of the non-fitting minimal fraction of donor computation (mfD) strategy for quantification of TCSPC FRET experiments is demonstrated not only for cases when fitting strategies fail due to the complexity of the decay, but also for simpler models when the number of detected photons is small. The conditions on which quantitative mfD analysis of FRET experiments allows faster acquisitions than fitting strategies have been established. This approach is well suited for imaging protein interactions in living cells as faster acquisitions result in better resolved spatio-temporal dynamics.