An integrative approach to molecular dynamics and single molecule FRET techniques.

Abstract

Molecular dynamics (MD) simulation and single-molecule fluorescence resonance energy transfer (smFRET) spectroscopy are two techniques that can be used to study conformational heterogeneity of proteins. However, significant challenges exist in these techniques that limit their effectiveness. All-atom MD is currently too slow to adequately sample conformational space of proteins. From an experimental perspective, deciding where to place the fluorescent dye reporters for smFRET measurements can lead to a great deal of wasted time, effort, and expense in producing labeled proteins that may or may not provide useful information. Even if the dyes are suitably placed, interpretation of smFRET data is also challenging since many conformations can lead to the same FRET efficiency for a single smFRET-labeled protein. In this work, we have addressed these problems using a highly integrative approach to smFRET and MD. We have developed an iterative method to characterize the conformational ensemble of proteins by combining MD-based enhanced sampling techniques and smFRET spectroscopy, where the computational and experimental approaches can benefit mutually from each other. We first employed an enhanced sampling based MD method (ES-MD) using system-specific collective variables for the system under study to sample various conformations. Since these simulations are biased, the relative weight of different populations of samples may not be accurate. In the next step, we predicted the FRET efficiency distributions for a large number of residue pairs and from the imperfect results determined the residues that are predicted to provide the most informative FRET distributions. The measured smFRET data was then used within a resampling scheme to correct the original computed set of conformations, which was then used to predict additional smFRET distributions for other labeled residues. We have examined this methodology on intrinsically-disordered/structured and membrane/soluble proteins.