Sub-millisecond studies of protein folding

The primary sequence of a protein encodes the information that is required for its structure formation. How a protein folds into its well-defined three dimensional structure is an intriguing question in biological science. In this thesis, cytochrome c (cyt c) and lipid binding proteins (LBP) were used as model proteins to study the factors affecting the efficiency and fidelity of a protein folding reaction. In order to monitor the early folding events, our home-built sub-millisecond solution mixer (dead time ∼100 musec) was used to initiate the reaction. Cyt c is an alpha-helical protein with a covalently linked heme group. I demonstrated that the early collapse of polypeptide chain of cyt c does not facilitate the subsequent folding process. Furthermore, the molten globule state (MG), a native-like partly folded state, which has been widely believed to be a transient folding intermediate in many proteins, is not an obligatory folding intermediate of cyt c; on the contrary, the folding efficiency is greatly impaired if it goes through the MG state. I also found that the slower folding of yeast iso-2 cyt c as compared to horse heart cyt c is mainly due to the slow ligation of Met80 to the heme although the two proteins have high sequence homology and structural similarity. LBP is a family of proteins with primarily beta-sheet structure. Mutagenesis studies showed that the turn regions of intestinal fatty acid binding protein (IFABP)---a member of the LBP family---does not affect the early stage of folding. One the other hand, significant changes were observed in the later stages of folding upon mutations within turn regions. The folding kinetics of several members of the LBP family were compared. It was found that they exhibit distinct folding kinetics despite their high structural similarities. These results give us a glimpse of how the folding pathway of a protein is encoded in its primary sequence.