“Gaining structural information on membrane proteins in their native lipid environment is a long‐standing challenge in molecular biology. Instead, it is common to employ membrane mimetics, which has been shown to affect protein structure, dynamics, and function severely. Here, we describe the incorporation of a bacterial outer membrane protein (OmpW) into natively excreted membrane vesicles for solution nuclear magnetic resonance (NMR) spectroscopy using a mutant Escherichia coli strain with a high outer membrane vesicle (OMV) production rate. We collected NMR spectra from both vesicles containing overexpressed OmpW and vesicles from a control strain to account for the presence of physiologically relevant outer membrane proteins in vesicles and observed distinct resonance signals from OmpW. Due to the increased production of OMVs and the use of non‐uniform sampling techniques we were able to obtain high‐resolution 2D (HSQC) and 3D (HNCO) NMR spectra of our target protein inside its native lipid environment. While this workflow is not yet sufficient to achieve in situ structure determination, our results pave the way for further research on vesicle‐based solution NMR spectroscopy.”
”
2.3. NMR spectroscopy
All NMR experiments were performed at 310 K on a Bruker NeoAdvance III 600 MHz spectrometer equipped with a cryogenic triple resonance probe head. Before performing the NMR experiments, all samples were dialyzed into a buffer containing 20 mM sodium phosphate pH 6.5, to which 0.1% NaN3 and 10% D2O were added. Samples were then transferred to 5 mm Shigemi tubes for NMR spectroscopy. The 1H‐15N correlation spectra for OmpW‐OMV were recorded using fast NMR methods (Bruker pulse program b_trosyetf3gpsi.2) with a nonuniform sampling rate of 50%, 128 hyper‐complex points in the indirect dimension, and processed with Topspin. For OmpW in nanodiscs, standard 2D 1H‐15N transverse relaxation optimized spectroscopy‐heteronuclear single quantum coherence (TROSY‐HSQC) with uniform sampling was recorded. The final size of the matrix processed was 4096 (F2) X 2048 (F1) points. The number of scans were varied between 1 k and 256 depending on the protein concentration as determined from absorbance measurements at 280 nm.
“