The engineering of de novo pathways for oxidative protein folding in Escherichia coli

Access full-text files




Masip, Lluis

Journal Title

Journal ISSN

Volume Title



Many commercially important proteins contain disulfide bonds, i.e. covalent linkages joining the thiol groups in cysteine amino acids. The formation of disulfide bonds is required in some proteins in order to attain their biologically active threedimensional conformation and for stability to temperature extremes or the action of protease enzymes. While E. coli is used extensively for recombinant protein manufacturing at an industrial scale, its ability to produce proteins containing multiple disulfide bonds is limited by biological constraints. This is because the cellular machinery that catalyzes the oxidation of protein cysteine residues resulting in the formation of disulfide bonds has evolved to cope with the relatively simple proteins of this organism which typically contain no more than 2-3 disulfides. The objective of this work was to re-engineer the disulfide bond formation machinery of the organism and thus synthesize pathways that have no known biological equivalent. As the results of this thesis demonstrate, the synthetic approach provided important and unanticipated insights into the mechanism of protein disulfide bonds, the vi evolution of the machinery responsible for oxidative protein folding and the physiology of the secretory compartment, the periplasmic space, of gram-negative bacteria. Furthermore, the engineered pathways described here may prove useful for protein manufacturing purposes. Initially, a de novo disulfide bond formation pathway was created through a combination of rational design and directed evolution which was mechanistically different and independent of the endogenous pathway in E. coli. Additionally, development of this pathway resulted in the isolation of mutations that transformed thioredoxin 1, which is a monomeric disulfide reductase, into a bridged [Fe2S2] dimer capable of catalyzing oxygen dependent sulfhydryl oxidation in vitro. The second periplasmic oxidative pathway developed, involved the use of directed evolution to improve the disulfide bond formation properties of thioredoxin 1. Specifically, thioredoxin 1 was optimized to function as an oxidase when exported to the periplasm of strains that lack the soluble component (DsbA) of the endogenous periplasmic disulfide bond formation pathway in E. coli. Finally, the last oxidative pathway developed was the result of the periplasmic localization of glutaredoxin 3, which surprisingly resulted in the promotion of disulfide bond formation even in the absence of the complete periplasmic oxidative machinery (DsbA and DsbB). Further characterization of this process suggests the presence and active participation of glutathione in the periplasmic space in a novel disulfide bond formation mechanism. This diverse group of oxidative pathways highlights the plasticity of the different components of the disulfide bond exchange pathways in E. coli, which can interact and complement each other rather effortlessly. From the transformation of cytoplasmic disulfide reductases into periplasmic oxidases to the unexpected role of the cytoplasmic redox buffer glutathione in a periplasmic disulfide bond formation processes.