Accounting for dye diffusion and orientation when relating FRET measurements to distances: three simple computational methods.

Förster resonance energy transfer (FRET) allows in principal for the structural changes of biological systems to be revealed by monitoring distributions and distance fluctuations between parts of individual molecules. However, because flexible probes usually have to be attached to the macromolecule to conduct these experiments, they suffer from uncertainty in probe positions and orientations. One of the way to address this issue is to use molecular dynamics simulations to explicitly model the likely positions of the probes, but, this is still not widely accessible because of the large computational effort required. Here we compare three simpler methods that can potentially replace MD simulations in FRET data interpretation. In the first, the volume accessible for dye movement is calculated using a fast, geometrical algorithm. The next method, adapted from the analysis of electron paramagnetic studies, utilises a library of rotamers describing probe conformations. The last method uses preliminary MD simulations of fluorescent dyes in solution, to identify all conformational states of dyes and overlays this on the macromolecular system. A comparison of these methods in the simple system of dye-labelled polyproline, shows that in the case of lack of interaction between the dye and host, all give results comparable with MD simulations but require much less time. Differences between these three methods and their ability to compete with MD simulations in the analysis of real experiment are demonstrated and discussed using the examples of cold shock protein and leucine transporter systems.