Kinetic studies show that most small proteins unfold and refold in at least two distinct steps, with the slower step having a relaxation time of several minutes or longer at 10 C. Many proteins show multiple relaxations in the slow step. Although these slow phases have frequently been attributed to the accumulation of structural intermediates, the present work has lead to an alternative interpretation which attributes these to cis-trans isomerization about peptide bonds which occur N-terminal to proline residues in the chain sequence. The continuing controversy over the origin of the slow phases is due to the lack of any suitable experimental technique which is capable of """"""""seeding"""""""" proline isomerization as it occurs during protein folding. A direct technique for following isomerization has now been developed, based on isomer-specific proteolysis (ISP). Many proteases have been shown to be specific for trans peptide bonds: the cis form must first isomerize before it can be hydrolyzed. These include prolidase and aminopeptidase P (where the active bond must be trans), trypsin chymotrypsin, and papain (where the bond following the active bond must be trans), and post-proline cleaving enzyme (where the bond preceding the active bond must be trans). The ISP method has been applied successfully to proline 93 in RNase A, where it was shown that folding into the native structure can only occur when this residue is cis. Other proline residues in RNase A and in other small proteins will be examined, so that the mechanism of folding can be determined and the role of isomerization can be clarified. The ISP technique will also be applied to another very important type of reaction, termed subtle conformational changes, which occur between two native-like conformations of the same protein and which frequently cause large changes in activity. These reactions usually occur in the slow-to-very-slow time range, and our working hypothesis is that the rate-limiting step may in some cases be the isomerization of peptide bonds of proline residues, as has already been suggested for several proteins including concanavalin A, fibrinogen fragment 1, and superoxide dismutase.
