Structure refinement and membrane positioning of selectively labeled OmpX in phospholipid nanodiscs.

Abstract

NMR structural studies on membrane proteins are often complicated by their large size, taking into account the contribution of the membrane mimetic. Therefore, classical resonance assignment approaches often fail. The large size of phospholipid nanodiscs, a detergent-free phospholipid bilayer mimetic, prevented their use in high-resolution solution-state NMR spectroscopy so far. We recently introduced smaller nanodiscs that are suitable for NMR structure determination. However, side-chain assignments of a membrane protein in nanodiscs still remain elusive. Here, we utilized a NOE-based approach to assign (stereo-) specifically labeled Ile, Leu, Val and Ala methyl labeled and uniformly (15)N-Phe and (15)N-Tyr labeled OmpX and calculated a refined high-resolution structure. In addition, we were able to obtain residual dipolar couplings (RDCs) of OmpX in nanodiscs using Pf1 phage medium for the induction of weak alignment. Back-calculated NOESY spectra of the obtained NMR structures were compared to experimental NOESYs in order to validate the quality of these structures. We further used NOE information between protonated lipid head groups and side-chain methyls to determine the position of OmpX in the phospholipid bilayer. These data were verified by paramagnetic relaxation enhancement (PRE) experiments obtained with Gd(3+)-modified lipids. Taken together, this study emphasizes the need for the (stereo-) specific labeling of membrane proteins in a highly deuterated background for high-resolution structure determination, particularly in large membrane mimicking systems like phospholipid nanodiscs. Structure validation by NOESY back-calculation will be helpful for the structure determination and validation of membrane proteins where NOE assignment is often difficult. The use of protein to lipid NOEs will be beneficial for the positioning of a membrane protein in the lipid bilayer without the need for preparing multiple protein samples.