Common ancestors of GMC oxidoreductases

The glucose-methanol-choline (GMC) superfamily of FAD-dependent oxidoreductases comprises a range of oxidases and dehydrogenases acting on various substrates such as sugars or alcohols. This large superfamily of oxidoreductases is defined by a common fold and by a close phylogenetic relationship, and includes industrially relevant enzymes such as aryl-alcohol oxidases (AAO) and pyranose dehydrogenases (PDH), forming one clade of this superfamily, or glucose oxidases (GOx) glucose dehydrogenases (GDH) constituting another GMC clade. These enzymes are among others of importance for the deconstruction of lignocellulose since they can increase the yield of fermentable sugars that can be obtained by enzymatic hydrolysis. By applying the method of ancestral sequence reconstruction we want to gain a better understanding how these enzymes evolved over time. In addition we aim at obtaining enzymes that show a higher thermostability, which will be useful for various applications.

The clade of pyranose oxidases (POx) is a distant branch in the AA3 family oxidoreductases, and bacterial members have hardly been studied in the past. We could show that the sequence space of bacterial pyranose oxidases contains two different activities, POx-like activity (oxidation of monosaccharides) and C-glycoside 3-oxidase activity (CGOx; oxidation of the glucose moiety in C- and O-glycosides). These glycosides are frequently found in plants and play an important role in human nutrition and health, hence knowledge on their breakdown is important. By studying this sequence space in more detail, we showed that actinobacterial sequences form 4 distinct clades, one containing (dimeric) POx activities, two containing monomeric CGOx activities, and one unstudied. We then used ancestral sequence reconstruction to follow the shift of a presumed ancestral (C-glycoside) 3-oxidase to extant monosaccharide-oxidizing POx and bacterial CGOx. We could thus show a shift from the oldest ancestor, which was a generalist, to the different specific activities in individual clades during evolution. Structural models of ancestors confirmed a possible change of substrate preferences based on the differences in regions that guide and tune up substrate intake in the active site - substrate loop, flavinylation motif, or residues that interact with the substrate, as well as the state of oligomerization. Based on the hypothesis that oligomerisation affects substrate preference we engineered a dimeric monosaccharide-specific POx to a monomer, which indeed showed significantly increased activity with C-glycosides and decreased activity with monosaccharides. Furthermore, we studied the sequence space of the AA3 enzymes glucose oxidases and glucose dehydrogenases. We could identify enzymes that preferentially oxidized oligosaccharides such as maltose. These could be of interest for applications in maltose-specific biosensors.