Comparative analysis of the orientation of transmembrane peptides using solid-state 2H- and 15N-NMR: mobility matters

Abstract

Many solid-state nuclear magnetic resonance (NMR) approaches for membrane proteins rely on orientation-dependent parameters, from which the alignment of peptide segments in the lipid bilayer can be calculated. Molecules embedded in liquid-crystalline membranes, such as monomeric helices, are highly mobile, leading to partial averaging of the measured NMR parameters. These dynamic effects need to be taken into account to avoid misinterpretation of NMR data. Here, we compare two common NMR approaches: (2)H-NMR quadrupolar waves, and separated local field (15)N-(1)H polarization inversion spin exchange at magic angle (PISEMA) spectra, in order to identify their strengths and drawbacks for correctly determining the orientation and mobility of α-helical transmembrane peptides. We first analyzed the model peptide WLP23 in oriented dimyristoylphosphatidylcholine (DMPC) membranes and then contrasted it with published data on GWALP23 in dilauroylphosphatidylcholine (DLPC). We only obtained consistent tilt angles from the two methods when taking dynamics into account. Interestingly, the two related peptides differ fundamentally in their mobility. Although both helices adopt the same tilt in their respective bilayers (~20°), WLP23 undergoes extensive fluctuations in its azimuthal rotation angle, whereas GWALP23 is much less dynamic. Both alternative NMR methods are suitable for characterizing orientation and dynamics, yet they can be optimally used to address different aspects. PISEMA spectra immediately reveal the presence of large-amplitude rotational fluctuations, which are not directly seen by (2)H-NMR. On the other hand, PISEMA was unable to define the azimuthal rotation angle in the case of the highly dynamic WLP23, though the helix tilt could still be determined, irrespective of any dynamics parameters.