Plasmid- and Inducer-Free Biosynthesis of Difucosyllactose through Rational Fusion Enzyme Engineering and Systematic Pathway Optimization

Difucosyllactose (DFL), a fucosylated human milk oligosaccharide, exhibits significant application potential in the food and pharmaceutical industries. However, the microbial production of DFL faces challenges such as the accumulation of the intermediate 2'-fucosyllactose (2'-FL) and the high costs associated with the use of antibiotics and inducers. In this study, we constructed a plasmid- and inducer-free Escherichia coli MG1655 strain to enhance the biosynthesis of DFL while minimizing 2'-FL accumulation. Initially, the de novo DFL biosynthetic pathway was established by introducing different α-1,3-fucosyltransferases (α-1,3-FucT) into a 2'-FL-producing strain. Combinatorial metabolic engineering strategies were then employed to improve DFL accumulation. Subsequently, guided by the in silico multienzyme assembling by reshaping space (iMARS) framework, a fusion enzyme, FucTaY218K-L60-FutC, was rationally designed to effectively alleviate 2'-FL accumulation. Furthermore, modulation of the hydrophobic microenvironment in the substrate-binding pocket of FucTaY218K yielded the beneficial mutant FucTaY218K/W31R. Finally, modifications to the lactose operon significantly improved strain growth while concurrently enhancing DFL production. The best-performing strain achieved a DFL titer of 74 g/L in a 5-L bioreactor, corresponding to a productivity of 0.96 g/L/h, representing the highest productivity reported to date. This study establishes a robust and environmentally friendly platform for the industrial-scale production of DFL and provides a strategy for the microbial synthesis of other high-value human milk oligosaccharides.