Construction and characterization of a single polypeptide chain containing two enzymatically active dihydrofolate reductase domains.

A single polypeptide chain containing two dihydrofolate reductase (DHFR) sequences from Escherichia coli was constructed to determine if a repeat sequence fusion protein could be expressed in an active form. The possibility that intersequence interactions could play a significant role for this enzyme is suggested by the results of Hall and Frieden (1989, Proc. Natl Acad. Sci. USA, 86, 3060-3064) who observed a substantial decrease in the yield of active enzyme when folded in the presence of a large C-terminal fragment. The fusion protein [DHFR(Cys152Glu)--Ile--DHFR (Met1Gln)] was efficiently expressed in E. coli cells and has an activity which is twice that of the wild-type enzyme in the standard assay. The Michaelis constants of the fusion protein for the substrate, dihydrofolate and the cofactor, NADPH, are essentially unchanged from those of the wild-type protein. The urea-induced in vitro unfolding reaction of the fusion protein at low concentrations was found to be fully reversible and follow a three state model, suggesting that the two domains unfold independently. At higher protein concentrations the unfolding transition broadened and shifted to a higher urea concentration. Size-exclusion chromatography results are consistent with the formation of aggregates at the higher protein concentration, even in the absence of denaturant.