Lipid-gramicidin interactions using two-dimensional Fourier-transform electron spin resonance.

The application of two-dimensional Fourier-transform electron-spin-resonance (2D-FT-ESR) to the study of lipid/gramicidin A (GA) interactions is reported. It is shown that 2D-FT-ESR spectra provide substantially enhanced spectral resolution to changes in the dynamics and ordering of the bulk lipids (as compared with cw-ESR spectra), that result from addition of GA to membrane vesicles of dipalmitoylphosphatidylcholine (DPPC) in excess water containing 16-PC as the lipid spin label. The agreement between the theory of Lee, Budil, and Freed and experimental results is very good in the liquid crystalline phase. Both the rotational and translational diffusion rates of the bulk lipid are substantially decreased by addition of GA, whereas the ordering is only slightly increased, for a 1:5 ratio of GA to lipid. The slowing effect on the diffusive rates of adding GA in the gel phase is less pronounced. It is suggested that the spectral fits in this phase would be improved with a more detailed dynamic model. No significant evidence is found in the 2D-FT-ESR spectra for a second immobilized component upon addition of GA, which is in contrast to cw-ESR. It is shown from simulations of the observed 2D-FT-ESR spectra that the additional component seen in cw-ESR spectra, and usually attributed to "immobilized" lipid, is inconsistent with its being characterized by increased ordering, according to a model proposed by Ge and Freed, but it would be consistent with the more conventional model of a significantly reduced diffusional rate. This is because the 2D-FT-ESR spectra exhibit a selectivity, favoring components with longer homogeneous relaxation times, T2. The homogeneous linewidths of the 2D-FT-ESR autopeaks appear to broaden as a function of mixing time. This apparent broadening is very likely due to the process of cooperative order director fluctuations (ODF) of the lipids in the vesicle. This real-time observation of ODF is distinct from, but appears in reasonable agreement with, NMR results. It is found that addition of GA to give the 1:5 ratio has only a small effect on the ODF, but there is a significant temperature dependence.