Engineering Escherichia coli for high-level production of lacto-N-fucopentaose I by stepwise de novo pathway construction

Engineering Escherichia coli for Highly Efficient Biosynthesis of Lacto-N-difucohexaose II through De Novo GDP-l-fucose Pathway

Lacto-N-difucohexaose II (LNDFH II) is a typical fucosylated human milk oligosaccharide and can be enzymatically produced from lacto-N-tetraose (LNT) by a specific α1,3/4-fucosyltransferase from Helicobacter pylori DMS 6709, referred to as FucT14. Previously, we constructed an engineered Escherichia coli BL21(DE3) with a single plasmid for highly efficient biosynthesis of LNT. In this study, two additional plasmids harboring the de novo GDP-L-fucose pathway module and FucT14, respectively, were further introduced to construct the strain for successful biosynthesis of LNDFH II. FucT14 was actively expressed, and the engineered strain produced LNDFH II as the major product, lacto-N-fucopentaose (LNFP) V as the minor product, and a trace amount of LNFP II and 3-fucosyllactose as very minor products. Additional expression of the α1,3-fucosyltransferase FutM1 from a Bacteroidaceae bacterium from the gut metagenome could obviously enhance the LNDFH II biosynthesis. After optimization of induction conditions, the maximum titer reached 3.011 g/L by shake-flask cultivation. During the fed-batch cultivation, LNDFH II was highly efficiently produced with the highest titer of 18.062 g/L and the productivity yield of 0.301 g/L·h.

Highly-efficient in vivo production of lacto-N-fucopentaose V by a regio-specific α1,3/4-fucosyltransferase from Bacteroides fragilis NCTC 9343

Lacto-N-fucopentaose V (LNFP V) is a typical human milk pentasaccharide. Multi-enzymatic in vitro synthesis of LNFP V from lactose was reported, however, microbial cell factory approach to LNFP V production has not been reported yet. In this study, the biosynthetic pathway of LNFP V was examined in Escherichia coli. The previously constructed E. coli efficiently producing lacto-N-tetraose was used as the starting strain. GDP-fucose pathway module and a regio-specific glycosyltransferase with α1,3-fucosylation activity were introduced to realize the efficient synthesis of LNFP V. The α1,3/4-fucosyltransferase from Bacteroides fragilis was selected as the best enzyme for in vivo biosynthesis of LNFP V from nine candidates, with the highest titer and the lowest by-product accumulation. A beneficial variant K128D was obtained to further enhance LNFP V titer using computer-assisted site-directed mutagenesis. The final strain EW10 could produce 25.68 g/L LNFP V by fed-batch cultivation, with the productivity of 0.56 g/L·h.

De Novo Biosynthesis of Difucosyllactose by Artificial Pathway Construction and α1,3/4-Fucosyltransferase Rational Design in Escherichia coli

Difucosyllactose (DFL) is a significant and plentiful oligosaccharide found in human breast milk. In this study, an artificial metabolic pathway of DFL was designed, focusing on the de novo biosynthesis of GDP-fucose from only glycerol. This was achieved by engineering Escherichia coli to endogenously overexpress genes manB, manC, gmd, and wcaG and heterologously overexpress a pair of fucosyltransferases to produce DFL from lactose. The introduction of α-1,2-fucosyltransferase from Helicobacter pylori (FucT2) along with α-1,3/4-fucosyltransferase (HP3/4FT) addressed rate-limiting challenges in enzymatic catalysis and allowed for highly efficient conversion of lactose into DFL. Based on these results, molecular modification of HP3/4FT was performed based on computer-assisted screening and structure-based rational design. The best-performing mutant, MH5, containing a combination of five mutated sites (F49K/Y131D/Y197N/E338D/R369A) of HP3/4FT was obtained. The best strain BLC09-58 harboring MH5 yielded 45.81 g/L of extracellular DFL in 5-L fed-batch cultures, which was the highest titer reported to date.

Efficient Biosynthesis of Lacto-N-Biose I, a Building Block of Type I Human Milk Oligosaccharides, by a Metabolically Engineered Escherichia coli

Lacto-N-biose I (LNB), termed a Type 1 disaccharide, is an important building block of human milk oligosaccharides. It shows promising prebiotic activity by stimulating the proliferation of many gut-associated bifidobacteria and thus displays good potential in infant foods or supplements. Enzymatic and microbial approaches to LNB synthesis have been studied, almost all of which involve glycosylation of LNB phosphorylase as the final step. Herein, we report a new and easier microbial LNB synthesis strategy through the route "lactose → lacto-N-triose II (LNTri II) → lacto-N-tetraose (LNT) → LNB". A previously constructed LNT-producing Escherichia coli BL21(DE3) strain was engineered for LNB biosynthesis by introducing Bifidobacterium bifidum LnbB. LNB was efficiently produced, accompanied by lactose regeneration. Genomic integration of key pathway genes related to LNTri II and LNT synthesis was performed to enhance LNB titers. The final engineered strain produced 3.54 and 26.88 g/L LNB by shake-flask and fed-batch cultivation, respectively.

Efficient Biosynthesis of Difucosyllactose via De Novo GDP-l-Fucose Pathway in Metabolically Engineered Escherichia coli

Difucosyllactose (DFL) is an important component of human milk oligosaccharides (HMOs) and has significant benefits for the growth and development of infants. So far, a few microbial cell factories have been constructed for the production of DFL, which still have problems of low production and high cost. Herein, a high-level de novo pathway DFL-producing strain was constructed by multistep optimization strategies in Escherichia coli BL21star(DE3). We first efficiently synthesized the intermediate 2'-fucosyllactose (2'-FL) in E. coli BL21star(DE3) by the advisable stepwise strategy. The truncated α-1,3/4-fucosyltransferase (Hp3/4FT) was then introduced into the engineered strain to achieve de novo biosynthesis of DFL. ATP-dependent protease (Lon) and GDP-mannose hydrolase (NudK) were deleted, and mannose-6-phosphate isomerase (ManA) was overexpressed to improve GDP-l-fucose accumulation. The regulator RcsA was overexpressed to fine-tune the expression level of pathway genes, thereby increasing the synthesis of DFL. The final strain produced 6.19 g/L of DFL in the shake flask and 33.45 g/L of DFL in the 5 L fermenter, which were the highest reported titers so far. This study provides a more economical, sustainable, and effective strategy to produce the fucosylated human milk oligosaccharides (HMOs).

Microbial Synthesis of Lacto-N-fucopentaose I with High Titer and Purity by Screening of Specific Glycosyltransferase and Elimination of Residual Lacto-N-triose II and Lacto-N-tetraose

Lacto-N-fucopentaose I (LNFP I) has recently been approved as generally recognized as safe, demonstrating its great commercial potential in the food industry. Microbial synthesis through metabolic engineering strategies is an effective approach for large-scale production of LNFP I. Biosynthesis of LNFP I requires consideration of two key points: high titer with low byproduct 2'-fucosyllactose (2'-FL) generation and high purity with low lacto-N-triose II (LNTri II) and lacto-N-tetraose (LNT) residues. Herein, α1,2-fucosyltransferase from Thermoanaerobacterium sp. RBIITD was screened from 16 selected LNFP I-producing glycosyltransferase candidates, showing the highest in vivo LNFP I productivity. Chromosomal integration of wbgO enhanced the LNFP I production by improving the precursor conversion from LNTri II to LNT. The best engineered strain produced 4.42 and 35.1 g/L LNFP I in shake-flask and fed-batch cultivation, respectively. The residual LNTri II and LNT were eliminated by further cultivation with a recombinant strain coexpressing Bifidobacterium bifidum β-N-acetylhexosaminidase and lacto-N-biosidase. A strategy for LNFP I biosynthesis with high yield and purity was finally realized, providing support for its practical application in large-scale production.

Elimination of Residual Lacto-N-triose II for Lacto-N-tetraose Biosynthesis in Engineered Escherichia coli

Lacto-N-tetraose (LNT) is an important neutral human milk oligosaccharide (HMO) and acts as a significant core structure for complex HMO biosynthesis. We previously achieved high-yield LNT biosynthesis (57.5 g/L) using fed-batch fermentation; however, residual lacto-N-triose II (LNTri II) was also found (21.58 g/L). Here, we re-engineered an efficient LNT-producing Escherichia coli with low LNTri II accumulation using genetically stable LNTri II-producing strains with a genomic insertion of lgtA (encoding β1,3-N-acetylglucosaminyltransferase). Comparable and low titers of LNT (3.73-4.61 g/L) and LNTri II (0.33-0.63 g/L), respectively, were obtained by introducing β1,3-galactosyltransferase. To reduce residual LNTri II, the E. coli transporter gene setA was disrupted, obviously reducing the accumulation of LNTri II and LNT. Next, the gene encoding β-N-acetylhexosaminidase (BbhI) was introduced into LNT-producing strains or E. coli BL21(DE3) for single- or mixed-strain cultivation, respectively. Finally, LNT was obtained (30.13 g/L) in a cocultivation system of mixed engineered strains without undesired LNTri II.