Global Structure Analysis of Acid-Unfolded Myoglobin with Consideration to Effects of Intermolecular Coulomb Repulsion on Solution X-ray Scattering

Abstract

To obtain information on the global structure of protein in the acid-unfolded (AU) state, the structure of apomyoglobin (apoMb) was analyzed by using the solution X-ray scattering (SXS) method. SXS profiles were obtained over a wide range of protein concentrations, 1-18 mg mL-1, under strongly acidic conditions. From analysis of the SXS profile extrapolated to a zero protein concentration, the mean square radius, Rsq, of AU-apoMb at 20 mM HCl was estimated to be 4.81 +/- 0.31 nm. This estimate is more than 1.3 nm larger than those of 3.0-3.5 nm reported thus far. The difference originates from the fact that effects of Coulomb repulsive forces acting between AU-apoMb molecules have not been correctly taken into account in the conventional analysis. In fact, even at a low protein concentration of 1 mg mL-1 close to the limit of measurement in the present SXS method, the solution condition applicable to estimating accurately structural parameters of AU-apoMb is very limited. At HCl concentrations lower than 10 mM, the scattering intensity at a small scattering vector decreases remarkably through the effect of intermolecular repulsive forces and the forward scattering intensity is significantly lower than the estimate from the partial specific volume of protein. On the other hand, at HCl concentrations higher than 50 mM, some compact molten-globule-like structures emerge. As a result, the intermediate concentration of 20 mM HCl is the best choice of the solution condition for determining Rsq of AU-apoMb. The effect of intermolecular Coulomb repulsion on the SXS profile of AU-apoMb is at its maximum for forward scattering and decreases monotonously with an increase in the scattering angle to be virtually negligible at K approximately 0.63 nm(-1). Whereas urea-denatured apoMb shows a SXS profile typical of Gaussian chains, the intrinsic SXS profile of AU-apoMb differs significantly from those of Gaussian chains.