Human milk oligosaccharides (HMOs), the third most abundant solid component in breast milk after lactose and lipids, represent a distinctive compositional difference between human and bovine milk 1, 2. These oligosaccharides act as natural prebiotics, influencing the infant gut microbiota, while also serving as bioactive molecules that support brain development and exhibit antimicrobial properties [3]. Several HMOs have been classified as generally recognized as safe and are increasingly incorporated into infant formulas, underscoring their growing importance in infant nutrition [4]. To date, over 200 distinct HMO structures have been identified, with more than 180 of them characterized in detail 5, 6. The structural diversity of HMOs arises from variations in glycosylation patterns, glycosidic linkages, and the specific monosaccharide units involved [7]. Among them, the fucosylation of the lactose core via α-1,2 or α-1,3 linkages generates fucosylated HMOs such as 2′-fucosyllactose (2′-FL) and 3-fucosyllactose (3-FL), while sialylation with N-acetylneuraminic acid (Neu5Ac) through α-2,3 or α-2,6 linkages results in sialylated HMOs 4, 8, 9. Additionally, the incorporation of other monosaccharides, such as galactose or N-acetylglucosamine (GlcNAc), leads to extended structures like lacto-N-neotetraose (LNnT), which feature β-1,3 linkages between galactose and GlcNAc [10]. These modifications generate linear and branched configurations, significantly enhancing the complexity and structural diversity of HMOs in human milk (Figure 1).

Traditionally, HMO extraction has relied on direct isolation from breast milk [11]. With the rapid development in synthetic biology and metabolic engineering, microbial synthesis using engineered cell factories has emerged as a more efficient and scalable approach for HMO production [7]. Recent studies have notably increased HMO yields by optimizing microbial metabolic pathways and supplementing key substrate materials, such as lactose and glycosylation precursors. These advancements have been thoroughly discussed in recent reviews 1, 4, 7, 12. Another key aspect of HMO biosynthesis is glycosyltransferases (GTs), which catalyze the glycosylation of lactose to form various oligosaccharide structures 6, 13. The main GTs involved in HMO synthesis include β-N-acetylglucosamine transferases (GlcNAcTs), β-galactosyltransferases (GalTs), α-fucosyltransferases (FucTs), and α-sialyltransferases (SiaTs), which respectively catalyze the addition of GlcNAc, β-galactose, α-fucose, and α-sialic acid to the lactose core (Figure 1) 6, 14. However, insufficient enzymatic capabilities of these GTs, such as low catalytic efficiency and limited substrate specificity, pose significant bottlenecks that impede HMO production [14].

Although research on the GTs involved in HMO synthesis has garnered increasing attention in recent years, an overview of enzyme engineering based on structural and catalytic mechanisms is still lacking. This review initially focuses on the structural and mechanistic understanding of GTs, complemented by a summary of pivotal studies on rational modification from the past 5 years. Furthermore, this review offers valuable insights into future directions for GTs in HMO biosynthesis.