More than a third of human proteins contain disulfide bonds, yet our understanding of how these are formed in vivo has remained rudimentary. Formation of disulfides takes place during oxidative folding and is catalyzed by the enzyme Protein Disulfide Isomerase (PDI). We have developed a single molecule approach to study this process in real-time, using force-clamp AFM. We investigated the oxidative folding of human titin domains, in the presence and absence of oxidase enzymes such as human PDI. Our approach enables, for the first time, independent kinetic measurements of folding and disulfide formation in individual proteins. Strikingly, we discovered a critical period for disulfide formation in the titin I1 domain. After protein folding was completed, disulfide formation could no longer be catalyzed by PDI. Thus, the competing kinetics of PDI-mediated oxidation and protein folding determines the outcome of oxidative folding. The involvement of glutathione in oxidative folding has long been under speculation. We show that glutathionylation can retard the kinetics of protein folding by more than 10-fold, revealing a possible novel role of this ubiquitous small molecule. In protein substrates with more than two cysteines, our data show that correct pairing is achieved through substrate folding rather than through the action of the oxidase enzyme. Therefore, the accuracy of cysteine pairing increases along the folding pathway of the substrate. These unprecedented observations shed new light on how disulfide formation relates to protein folding in vivo. Taken together, we are able to construct a complete kinetic model of oxidative folding, incorporating experimentally measured rates for the individual reaction steps. The technique presented here can be used to elucidate the pathway of oxidative folding for a wide variety of proteins, and can also be used to identify mechanisms that trigger misfolding and disease.
Read full abstract