Starch-derived oligosaccharides as prebiotics in innovative functional foods.

Driven by the escalating demand for innovative functional foods, starch-derived oligosaccharides (SDOs), particularly low-degree-of-polymerization fractions, have emerged as prominent functional carbohydrates, with selected SDO fractions exhibiting prebiotic properties. Their digestive behavior and gut microbial utilization depend strongly on degree of polymerization, glycosidic linkage, and molecular structure. Findings from in vitro and cell-based studies, animal models, and selected human intervention trials further indicate structure-dependent microbial and physiological responses. Herein, the structural characteristics, underlying prebiotic mechanisms, and traditional biocatalytic synthesis strategies of diverse SDOs are systematically summarized and discussed. Currently, enzymatic transglycosylation dominates the industrial production of SDOs; however, this process is fundamentally bottlenecked by the inherent hydrolytic propensity and poor regioselectivity of wild-type glycosidases or transglycosidases, leading to heterogeneous product profiles and limited thermodynamic yields. To overcome these natural biocatalytic constraints, customizing robust enzymes via artificial intelligence (AI) and computational structural redesign has become an important research direction. By integrating deep learning algorithms with structure-based de novo design and multidimensional kinetic prediction, the architecture of native catalytic clefts can be rationally remodeled. Such computational strategies facilitate favorable orientation of carbohydrate acceptor substrates and enhance transglycosylation selectivity toward targeted SDOs promoting innovative functional food applications.