Abstract
Fusarium graminearum produces an α-l-fucosidase, FgFCO1, which so far appears to be the only known fungal GH29 α-l-fucosidase that catalyzes the release of fucose from fucosylated xyloglucan. In our quest to synthesize bioactive glycans by enzymatic catalysis, we observed that FgFCO1 is able to catalyze a transglycosylation reaction involving transfer of fucose from citrus peel xyloglucan to lactose to produce 2'-fucosyllactose, an important human milk oligosaccharide. In addition to achieving maximal yields, control of the regioselectivity is an important issue in exploiting such a transglycosylation ability successfully for glycan synthesis. In the present study, we aimed to improve the transglycosylation efficiency of FgFCO1 through protein engineering by transferring successful mutations from other GH29 α-l-fucosidases. We investigated several such mutation transfers by structural alignment, and report that transfer of the mutation F34I from BiAfcB originating from Bifidobacterium longum subsp. infantis to Y32I in FgFCO1 and mutation of D286, near the catalytic acid/base residue in FgFCO1, especially a D286M mutation, have a positive effect on FgFCO1 transfucosylation regioselectivity. We also found that enzymatic depolymerization of the xyloglucan substrate increases substrate accessibility and in turn transglycosylation (i.e., transfucosylation) efficiency. The data include analysis of the active site amino acids and the active site topology of FgFCO1 and show that transfer of point mutations across GH29 subfamilies is a rational strategy for targeted protein engineering of a xyloglucan-active fungal α-l-fucosidase.