MODELLING THE UNFOLDED STATE OF PROTEINS

点击翻译按钮获取中文摘要

The denatured state of proteins has been a topic of riveting interest and great perplexity for almost six decades, following the proposal of Mirsky and Pauling that a theory of protein structure is precisely a theory of protein denaturation. Pauling's early work was followed by seminal contributions from Flory, Tanford and many others. Yet, to date there is still no model of the denatured state that is sufficiently detailed for practical use in computational studies of folding and binding. We seek a reliable minded of the denatured state that can be used to calculate changes in the solvent accessible surface area upon folding and/or binding. Most literature values take the surface area of a residue, , in the denatured state as being its area in an extended Gly-X Gly tripeptide or, occasionally, in some other closely related compound. While any well-defined state is formally suitable for use as a standard state, the denatured state has all too often been taken to the equivalent to this standard state, resulting typically ina large over-estimate of the area lost on folding or binding. One obvious approach to this problem would be to stimulate athe behavior of a peptide of desired sequence under denaturing conditions. However, the result of such a simulation would be highly sensitive to the choice of force field. We propose to circumvent this force field-dependent problem by bracketing the answer between reliable extremes, thus providing an upper and lower limit. At one extreme is the hard sphere model which has only repulsive forces (i.e. excluded volume), and which therefore exposes more surface than would be expected in a molecule that can self-associate under denaturing conditions. At the other extreme, one can use segments of limited length taken from folded proteins. Such segments are devoid of long range interactions, although it is certainly true that long range interactions may have been involved in determining the conformation of the segment. Nevertheless, folded segments still serve as a useful extreme in that the corresponding peptide under denaturing conditions is unlikely to be better shielded from the solvent. Preliminary results suggest that the interval between the two limits - hard sphere and folded protein-may be small enough to be applied usefully in the structural and thermodynamic studies proposed in this Program Project.