Analysis of Protein Folding by Protein Engineering

Abstract Denatured proteins can in many cases refold to the native structure. It is generally accepted that the folding has to progress along a specific pathway (1), although Harrison and Durbin (2) have argued that folding could follow many different parallel pathways (jigsaw puzzle model). Several different models of protein folding have been proposed. The framework model (3) suggests that the earliest steps are the formation of native secondary structure elements, while an alternative model proposes that folding begins with a rapid hydrophobic collapse (4-7). A somewhat different view is reflected in the micro-domain folding models (8-10) which propose that small structural subdomains of the protein assume their native structure early during folding. These initiation sites either grow directly by extension or by diffusion and collision with other subdomains. If there are several initiation sites it is easily imaginable that many different, parallel, folding pathways exist. The molten globule hypothesis of protein folding (11, 12) suggests that a 'molten globule' intermediate occurs late on the folding pathway of all proteins. This intermediate is compact and has native-like secondary structure but fluctuating tertiary interactions. Its conversion into the native state is the rate-determining step of protein folding.