Non-canonical disulfide bond formation in the model system Escherichia coli: Implications for protein evolution and pathway design

Disulfide bonds are an important post-translational modification that provides stability for many proteins. In the bacterium Escherichia coli, oxidative protein folding is carried out in the periplasm by the thiol-disulfide oxidoreductase proteins. The thiol disulfide oxidoreductase DsbA is responsible for the initial generation of disulfide bonds into oxidatively folding proteins. DsbA is maintained in a catalytic oxidized state by the inner membrane protein DsbB, which channels electrons from DsbA to the electron transport chain through quinone reduction. My plan was to investigate if mutation and selection/screening could generate possible substitutes for the normal disulfide formation pathway. This may provide insight into what minimal requirements define a thiol disulfide oxidase. First, we took the disulfide reductase thioredoxin, a distantly related homologue to DsbA, and selected for mutants in the active site CXXC motif that when exported to the periplasm through the twin-arginine pathway could complement a DeltadsbB strain. We were able to show that mutation of the active site from CGPC to either CACC or CPCC resulted in variants that were capable of partial complementation of the defect in disulfide formation seen in the DeltadsbB strain; surprisingly, the mutant thioredoxin had acquired an iron sulfur cluster. These results demonstrated that an iron sulfur-containing thioredoxin could function to complement the entire disulfide formation pathway. Second, I discovered mutations in DsbB that bypassed the need for DsbA by utilizing DsbC as a catalyst for forming disulfide bonds. DsbC is one of the two putative disulfide isomerases in the periplasm. This work suggests that in addition to functioning as an isomerase, DsbC can also function as an oxidase. Finally, we have been able to isolate a chromosomal mutant that is able to bypass the need for both DsbA and DsbB. The strain suppresses all tested disulfide phenotypes, and is not dependent on any Dsb protein except DsbC. Initial mapping locates the source of suppression to an area on the chromosome corresponding to Complex I of the electron transport chain. These results reveal a surprising variety of ways that the cell can compensate for the loss of the normal disulfide catalytic machinery. In addition, this work may point to novel functions for the thiol disulfide isomerase DsbC.