M.Berkmen; "Disulfide bond formation in Escherichia coli", 11.7, 13:40
  • FENS
  • M.Berkmen; "Disulfide bond formation in Escherichia coli", 11.7, 13:40

You are here

Faculty of Engineering and Natural Sciences

Disulfide bond formation in Escherichia coli.

Mehmet Berkmen
New England Biolabs, ABD

 The formation of disulfide bonds in the E. coli periplasm by the oxidizing enzyme DsbA is not an accurate process. For example, the expression in E. coli of many eukaryotic proteins with non-consecutive disulfide bonds requires an additional protein, the disulfide bond isomerase DsbC for their full activity. The importance of non-consecutive disulfide bonds lead us to propose that proteins with non-consecutive disulfide bonds require DsbC for full activity and that disulfide bonds are formed predominantly during translocation in a consecutive fashion. To this end, based on its disulfide bond connectivity, E. coli phytase AppA was selected to be a candidate substrate for DsbC. As predicted, AppA depends on DsbC for its full activity. However, the activity of an AppA mutant lacking its non-consecutive disulfide bond is DsbC independent. The AppA homolog Agp is another periplasmic acid phosphatase with similar structure. It lacks the non-consecutive disulfide bond, but has the three consecutive disulfide bonds found in AppA. The consecutive disulfide bonded Agp is not dependent on DsbC, but is rendered dependent by the engineering into it AppA’s conserved non-consecutive disulfide bond.
The current dogma dictates that mis-oxidized and therefore mis-folded proteins are either degraded by periplasmic proteases or their disulfide bonds are shuffled into their active correctly-oxidized states by the disulfide bond isomerase, DsbC. Intriguingly, the detailed mechanism of disulfide bond isomerization by DsbC in vivo remains to be understood.
We propose that disulfide bond isomerization by DsbC in vivo is at least partly the selective reduction of disulfide bonds in mis-oxidized proteins, allowing DsbA a second chance to correctly oxidize the reduced protein. Our model would therefore predict that a periplasmic reductase could complement a DdsbC phenotype.
In order to identify a naturally occurring periplasmic reductase we analyzed the cysteine content of exported proteins in genomes of prokaryotes. Our analysis revealed that the periplasm of Bacteroides fragilis could be reducing and should therefore have periplasmic reductases which maintain the cysteines in their reduced state. Candidate periplasmic reductases were cloned and their capacity to complement disulfide bond isomerization was investigated in E. coli. The E. coli phytase AppA is dependent on DsbC for its full activity. Expression of the B. fragilis thioredoxin ortholog trxP completely alleviated the dependence of AppA on DsbC.
We present detailed characterization of the in vivo properties of TrxP which supports the model that a reductase could functionally complement the correct oxidation of a protein in the absence of DsbC.

July 11, 2007, 13:40, FENS 2019