Calculation of Membrane Protein Structures in Oriented-Sample NMR and Their Validation using Rosetta

Membrane proteins encode about 30% of all eukaryotic genomes and constitute the majority of current drug targets. However, only 2% of solved protein structures in Protein Data Bank are membrane proteins. To contribute to this outstanding problem, we have developed a computational methodology for calculating membrane protein structures from oriented sample solid-state NMR. We have demonstrated the feasibility of obtaining atomicresolution three-dimensional backbone structures of membrane proteins solely from the heteronuclear spin-spin dipolar couplings. We have further validated the calculated structures and determined their optimal immersion depth in the heterogeneous membrane-aqueous environment by combining the above structural fitting method and the Rosetta structure predicting package. Extension of this research include elucidation of the oligomerization states of homomeric protein assemblies by simultaneously utilizing solid-state NMR spectroscopy and Rosetta symmetry-docking protocols.