Thirty single cysteine substitution mutants of T4 lysozyme have been prepared and spin-labeled with a sulfhydryl-specific nitroxide reagent in order to systematically investigate the relationship between nitroxide side-chain mobility and protein structure. The perturbation caused by replacement of a native residue with a nitroxide amino acid was assessed from the resulting changes in biological activity, circular dichroism, and free energy of folding. The nitroxide produced context-dependent changes in stability and activity similar to those observed for substitution with natural amino acids at the same site but had little effect on the circular dichroism spectra. At solvent-exposed sites, the structural perturbation appears to be small at the level of the backbone fold. Nitroxide side-chain mobility faithfully reflects the protein tertiary fold at all sites investigated. The primary determinants of nitroxide side-chain mobility are tertiary interactions and backbone dynamics. Tertiary interactions constrain the side-chain mobility to an extent closely correlated with the degree of interaction. At interhelical loop sites, the side chains have a high mobility, consistent with high crystallographic thermal factors. On the exposed surfaces of alpha-helices, the side-chain mobility is not restricted by interactions with nearest neighbor side chains but appears to be determined by backbone dynamics. An unexpected result is a striking difference between the mobility of residues near the C- and N-termini of helices. These results provide the foundation for another dimension of information in site-directed spin-labeling experiments that can be interpreted in terms of the protein tertiary fold, its equilibrium dynamics and time-dependent conformational changes.

No abstract

Site-directed spin labeling of proteins has proven to be a practical means for determining secondary structure and its orientation; surfaces of tertiary interactions; inter-residue distances; chain topology and depth of a given side chain from the membrane/aqueous surface in membrane proteins; and local electrostatic potentials at solvent-exposed sites. Moreover, the mobility of a side chain together with its solvent-accessibility may serve to uniquely identify the topographical location of specific residues in the protein fold. Future spectral analysis should permit a quantitative estimation of the contribution of backbone flexibility to the overall side-chain dynamics. The ability to time-resolve the structural features mentioned above makes SDSL a powerful approach for exploring the evolution of structure on the millisecond time scale. We anticipate future applications to the study of protein folding both in solution and in chaperone-mediated systems.