Experimental studies of pathways of protein folding.

Studies of a stable molten globule intermediate (I) of apomyoglobin have shown that: (1) the A, G and H helices, but not the B and E helices, of myoglobin are stabilized in I, (2) individual peptides containing the G and H sequences do not show stable helix formation, although the H peptide shows partial (30%) helix formation, and (3) the mechanism by which the A, G and H helices are stabilized in I is not side chain interdigitation between helices at the helix contact sites of myoglobin. Consequently, the molten globule intermediate confers stability on the A, G and H helices, and the mechanism of stabilization is not the direct interaction found in native myoglobin. Kinetic studies of the folding reactions of small proteins have shown folding intermediates that could be either framework intermediates or molten globule intermediates, but a clear distinction between these two classes of kinetic intermediates has been lacking. An operational distinction is proposed here: molten globule intermediates are not stabilized by side chain interdigitation at contact sites between units of secondary structure found in the corresponding native protein, whereas framework intermediates are stabilized in this way. Site-directed mutagenesis experiments can distinguish between the two classes of intermediate. On the basis of this definition, the kinetic folding intermediates that are detected by far-UV circular dichroism can be molten globule intermediates, and when both a molten globule and a framework intermediate occur on the same folding pathway, the molten globule intermediate should precede the framework intermediate. Protection of individual amide protons against exchange has given the most detailed information thus far about the structures of folding intermediates in non-covalent folding reactions. It is possible that amide proton protection might occur during folding either by a non-specific mechanism, such as a hydrophobic collapse, or by the formation and later breakdown of non-native secondary structure; either event would pose a serious problem for interpretation of the results. Tests are available for assessing whether either event occurs, and they are discussed here.