No abstract

We have so far demonstrated that well-resolved and site-specifically assigned (13)C peaks as recorded by site-directed NMR study on (13)C-labeled membrane proteins can serve as a convenient probe to reveal their local conformation and dynamics. We attempted here to clarify the extent to which (13)C NMR spectra of (13)C-labeled fully hydrated bacteriorhodopsin (bR) as a typical membrane protein are visible or well resolved in the presence of inherent fluctuation motions with frequency of 10(2)-10(8) Hz, especially at the membrane surfaces. Accordingly, we estimated the relative proportion of (13)C NMR signals from the surface areas with and without peak suppression by the accelerated transverse relaxation effect by surface-bound Mn(2+) ions, which could be effective for residues within 8.7 angstroms of the membrane surface. It turned out that the experimental findings are consistent with the predicted amount of amino acid residues under consideration located within 8.7 angstroms of the surface for [1-(13)C]Val- and Ile-labeled bR and also [3-(13)C]Ala-bR. In contrast, (13)C NMR peaks from such surfaces area are almost completely or partially suppressed for [1-(13)C]Gly-, Ala-, Leu-, Phe- and Trp-labeled bR, as a result of plausible interference of the fluctuation frequency with frequency of magic angle spinning (10(4) Hz). We further assigned several (13)C NMR signals of [1-(13)C] Val-, Trp- and Ile-labeled bR on the basis of a variety of site-directed mutants with reference to those of the wild type. Further, we recorded the (13)C NMR of bR in lipid bilayers to search for the optimal conditions to be able to obtain signals with the highest peak intensities and spectral resolution. Backbone dynamics turn out to be essential for recording (13)C NMR spectra so as to escape from motional frequencies of the order of 10(4)-10(5) Hz, either in the direction of accelerated fluctuation or slowed motions in the direction of forming the 2D array.

No abstract