Abstract
α-L-Fucosidases are attractive biocatalysts for the production of bioactive fucosylated oligosaccharides, however, poor regioselectivity and activity for transglycosylation have significantly limited their applications. We have recently derived an α-L-Fucosidase, BF3242, from Bacteroides fragilis NCTC9343, which could efficiently synthesize a mixture of Fuc-α-1,3/1,6-GlcNAc, but its 1,3/1,6-regioselectivity was observably affected by reaction temperature. Here, we integrated loop-targeted random mutagenesis and site-directed mutagenesis to engineer the regioselectivity and transglycosylation activity of BF3242. Loop-targeted random mutagenesis revealed that L266 in the loop-4 (H242-S267) within the model of BF3242 was a key residue for the regioselectivity for transglycosylation, and the saturation mutagenesis at residue L266 uncovered a mutant L266H with a significantly increased 1,3-regioselectivity of 97 % from 69 % of WT BF3242. Subsequently, five designed single-site mutations at the putative aglycone subsites were performed, resulting in a double-site mutant L266H/M285C that increased the overall yield of Fuc-α-1,3/1,6-GlcNAc to 76 % from 68 % of WT BF3242. The saturation mutagenesis at residue M285 finally generated a double-site mutant L266H/M285T with the maximal overall yield of Fuc-α-1,3/1,6-GlcNAc of 85 % and 1,3-regioselectivity of 98 %. The R (T/H) of L266H/M285T was approximately 2.7-fold higher than that of the WT BF3242. Molecular dynamics simulations revealed that the structural flexibility of the loop-4 was substantially reduced in mutant L266H, and the hydrogen bond formation and binding affinity between mutant L266H/M285T and Fuc-α-1,3-GlcNAc was significantly enhanced. The semi-rationally engineered enzyme L266H/M285T would be a promising biocatalyst for highly 1,3-regioselective synthesis of fucosyl-N-acetylglucosamine disaccharide.