A theoretical pathway for the folding of RNase into its native conformation is derived from the contact map computed from crystallographic coordinates. The pathway is based on the hypothesis of Tanaka and Scheraga, according to which localized conformations stabilized by short-and medium-range interactions form before those conformational features that are stabilize primarily by long-range interactions. The pathway deduced from the contact map agrees with experimental information on intermediates detected in the thermal unfolding of RNase and in immunochemical studies on the formation of stable antigenic sites when reduced RNase is oxidized with glutathione. Ambiguities in the interpretation of the contact map are resolved by the combination of structural information contained in the cotitact map and experimental information.Contact maps (1)(2)(3)(4)(5) [and related, more quantitative representations, distance maps (6-10)] have been used to analyze the structures of native proteins (2,(6)(7)(8) and to compare the results of simulations of protein folding to the native structure (1, 3-5, 9). Some studies have dealt with the description of ordered structures, such as a-helices and parallel or antiparallel pleated sheets on the contact map (7,8) or with the recognition of large structural domains (2, 10). Although most of these studies concerned the native structure, we recently exploited the results of Monte Carlo calculations (1) and introduced a hypothetical mechanism of protein folding, based on a qualitative visual analysis of contact maps (2). It was shown that a self-consistent pathway of folding can be postulated for several proteins on the basis of the contact map alone.In this communication, we show that the pathway for the folding of RNase, derived from the contact map, is consistent with the one proposed on the basis of experimental information on intermediates during both thermal unfolding (with intact disulfide bonds) and glutathione oxidation of the reduced protein. Theoretical analysis and the interpretation of experimental data complement each other by narrowing the possible choices of pathways. Experimental data on the unfolding of RNase and the arguments that support the concept of a preferred pathway have been reviewed elsewhere (11-13).
METHODSThe contact map shown in Fig. 1 (Fig. 1) Fig. 1 by using the hypothesis of Tanaka and Scheraga (2). This hypothesis assumes that short-range interactions take precedence over mediumrange interactions, and medium-range interactions take precedence over long-range interactions. This criterion is satisfied by pathways in which contacts near the diagonal of the contact map are formed first, and contact regions are then formed in the order of increasing distance from the diagonal. The contact regions can be ordered by using the point with the smallest values of Ii -j