Protein structure elucidation by combining computational and experimental methods

Abstract

Membrane proteins remain a particular challenge in structural biology. Only approximately 1.5% of reported tertiary structures and around 100 unique polytopic membrane proteins are represented in the Protein Data Bank (PDB). Site-directed spin labeling electron paramagnetic resonance spectroscopy (SDSL-EPR) is often used for the structural characterization of proteins that elude other techniques, such as X-ray crystallography and NMR. RosettaEPR combines SDSL-EPR distance data with computational methods to improve high-resolution protein structure prediction. We demonstrated the feasibility of using RosettaEPR with soluble proteins by benchmarking the method on T4-lysozyme, for which a final model of 1.7Å accuracy was obtained. RosettaTMH, a new membrane protein de novo folding method that employs rigid body sampling, is also introduced. It expands upon RosettaEPR to cover the important class of membrane proteins. RosettaTMH was benchmarked on 34 membrane proteins of known structure using simulated EPR distance restraints. It was able to sample the correct topology for 33 of 34 proteins and improves Rosetta’s ability to predict the three-dimensional structure of large membrane proteins, such as transporters and receptors.