How to predict high fructosyltransferase activity in fungal GH32 enzymes

What is it about?

Sucrolytic enzymes catalyse sucrose hydrolysis or the synthesis of fructooligosaccharides (FOSs), a prebiotic in human and animal nutrition. FOS synthesis capacity differs between sucrolytic enzymes. Amino-acid-sequence-based classification of FOS synthesizing enzymes would greatly facilitate the in silico identification of novel catalysts, as large amounts of sequence data lie untapped. The development of a bioinformatics tool to rapidly distinguish between high-level FOSs synthesizing predominantly sucrose hydrolysing enzymes from fungal genomic data is presented. Sequence comparison of functionally characterized enzymes displaying low- and high-level FOS synthesis revealed conserved motifs unique to each group. New light is shed on the sequence context of active site residues in three previously identified conserved motifs. We characterized two enzymes predicted to possess low- and high-level FOS synthesis activities based on their conserved motif sequences. FOS data for the enzymes confirmed our successful prediction of their FOS synthesis capacity. Structural comparison of enzymes displaying low- and high-level FOS synthesis identified steric hindrance between nystose and a long loop region present only in low-level FOS synthesizers. This loop is proposed to limit the synthesis of FOS species with higher degrees of polymerization, a phenomenon observed among enzymes displaying low-level FOS synthesis. Conserved sequence motifs surrounding catalytic residues and a distant structural determinant were identifiers of FOS synthesis capacity and allow for functional annotation of sucrolytic enzymes directly from amino acid sequence. The tool presented may also be useful to study the structure–function relationships of β-fructofuranosidases by identifying mutations present in a group of closely related enzymes displaying similar function.

Featured Image

Why is it important?

The ability to identify enzymes with a preferred activity directly from protein sequence would greatly facilitate in silico bioprospecting for novel biocatalysts. Here we report on the development of a bioinformatics tool to rapidly distinguish between high-level FOSs synthesizing vs. predominantly sucrose hydrolysing enzymes from fungal genomic data. Furthermore, by investigation of the active pocket geometries of enzymes displaying low-level and high-level FOS synthesis and superimposing three-dimensional crystal structure models we also identified a shorter loop region in high-level FOS producing beta-fructofuranosidases, which we hypothesize creates space for FOS species with high degrees of polymerization.