To support the central role that carbohydrates and their derivatives play in the living world, Nature has evolved a large collection of diverse enzymes that are dedicated to their synthesis, degradation, or modification. These carbohydrate-active enzymes (CAZymes) have found their way into countless economical applications in the food, feed, chemical, bioenergy, and other industries. However, they still hold plenty of untapped biotechnological potential for the production of valuable sugars, glycans, or glycosides with useful properties.

The database of CAZymes continues to grow at a strong pace, with numerous new activities and families being discovered every year [1]. However, this rich pool of natural enzymes still cannot fulfill the wild dreams of academics and industry scientists, who are typically searching for robust proteins that can catalyze innovative conversions with superior efficiency and selectivity, often involving challenging unnatural substrates. Hence, protein engineers have a long and successful history of designing tailored CAZymes of their own.

The greatest accomplishments in this area over the past few years follow a clear trend: they were fueled by creative engineering strategies that paid special attention to structural rationalization, mechanistic details, or sequence patterns left behind by natural protein evolution. In this review, we zoom in on the most important jumps that were made across the fitness landscape of glycoside hydrolases (GH), glycosyltransferases (GT), and glycoside phosphorylases (GP), showing that we are slowly mastering the art of building mutational bridges that connect the different peaks of the landscape.