Characterization of long-range structure in the denatured state of staphylococcal nuclease. II. Distance restraints from paramagnetic relaxation and calculation of an ensemble of structures

Structural analysis of Δ131Δ, a fragment model of the denatured state of staphylococcal nuclease, has been extended by obtaining long-range distance restraints between chain segments by paramagnetic relaxation enhancement. Fourteen unique PROXYL spin labels were introduced at sites that are solvent-exposed in the native state, and the resulting enhancements of T₂ for the amide protons were measured by NMR spectroscopy. When these data were combined with either measured or estimated correlation times τ(c), the r^-6-weighted, time and ensemble-averaged distance between the spin label and 30 to 60 amide protons could be calculated for each spin-labeled protein. On the basis of approximately 700 such loose distance restraints, ensembles of compatible structures were generated by a combined distance geometry/molecular dynamics approach. Because of the large uncertainty in the physical basis of these distance restraints, a number of calculations were carried out to establish the sensitivity of the calculated structures to systematic errors in these restraints. Overall, the structural features reflected in the paramagnetic relaxation data were robust; large variations in τ(c), in the bounds window of allowed distances, or in the number of restraint distances used had small effects on the general features common to all calculated structures. The global topology of this denatured form of staphylococcal nuclease, as described by an ensemble of conformations consistent with the data, is strikingly similar to that of the native state, the major difference being the segregation of two hydrophobic segments that form a beta hairpin in the native state. These findings suggest that the topology of a protein's fold is established in the denatured state in the absence of cooperative interactions involving tight packing or stable hydrogen bonding. Hydrophobic interactions alone may encode global topology.