Immersion Depths of Lipid Carbons in Bicelles Measured by Paramagnetic Relaxation Enhancement

Myriads of biological processes occur in or at cellular lipid membranes. Knowledge about the localization of proteins, lipids, and other molecules within biological membranes is thus crucial for the understanding of such processes. Here, we present a method to determine the immersion depths of lipid carbon atoms in membranes by paramagnetic relaxation enhancement (PRE) caused by the presence of doxylated lipids. As membrane mimetics, we employ small isotropic bicelles made of synthetic lipids and of natural Escherichia coli phospholipid extract. Bicelles are particularly suitable for solution state NMR since they maintain a lipid bilayer while they are at the same time amenable to solution state NMR experiments. PREs were measured in the presence of different doxylated lipids with the nitroxide radical located in the headgroup and at various positions in the acyl chain. Theoretical PREs were calculated assuming a simple bicelle model using the Solomon-Bloembergen equations. Immersion depths of the lipid carbon atoms were obtained by a least-squares fit of the theoretical to the experimental PREs. The carbon immersion depths correspond well to results obtained by other methods and differences do not exceed 3-5 Å. This means that the method presented here provides sufficient resolution to distinguish the localization of carbons in different regions of the lipid bilayer, despite considerable simplifications of the underlying theory. These simplifications include a simple form of the spectral density function, which we find is sufficient to reliably determine immersion depths. A more complicated spectral density function that includes bicelle, lipid, and local motions may only improve the results if its parametrization is good enough. The approach presented here may be extended to the determination of protein localization in membranes employing realistic membrane mimetics like the bicelles made of E. coli phospholipid extract used here.