α1,3-Fucosyltransferase (α1,3-FucT) is essential for the biosynthesis of biologically active α1,3-fucosyloligosacchairdes (3-FOs) from human milk oligosaccharides (HMO), particularly 3-fucosyllactose (3-FL) trisaccharide. α1,3-FucT from Helicobacter pylori 26695 (FutA) accepts lactose and LacNAc as glycan acceptors and has a very low level of expression in Escherichia coli, and it shows a low catalytic activity for lactose in the large-scale synthesis of 3-FL. To overcome the poor solubility of FutA, codon optimization, and systematic truncation of the protein at the C-terminus with only one heptad repeat remaining (Δ52 FutA) were conducted to yield 150-200 mg/L of soluble protein of FutA and resulting in more than an 18-fold increase in the 3-FL yield. To improve the low level of enzyme catalytic activity for lactose, focused directed evolution was attempted using a semi-rational approach that combines structure-guided computational analysis and subsequent iterative saturation mutagenesis (ISM). In order to select the functional residues in active site/substrate binding site, docking simulation was used together with HotSpot Wizard to target evolutionarily variable amino acid positions. A128 site was selected from the key residue located in the active site, and A128N mutant displayed a 3.4-fold higher catalytic activity than wild-type Δ52 FutA. Considering that the A128N mutation is located in the deep cleft of the lactose binding site, the residues within the substrate binding sites, especially on the two α-helices for lactose and one α-helix for GDP-fucose, were subjected to structure-guided ISM. The selected residues from each helix were clustered, and ISM was performed for each cluster in parallel. In particular, the mutant with triple mutations (A128N/H129E/Y132I) located on the α5 helix exhibited a 9.6-fold improvement in specific activity when compared to wild-type Δ52 FutA. When such clustered mutations on two α-helices (α5 and α2/loop) were combined, mutants with triple (A128N/H129E/S46F) and quadruple mutations (A128N/H129E/Y132I/S46F) were generated, which showed the synergistic effects, that is 14.5- and 15.5-fold improvement in specific activity relative to wild-type Δ52 FutA, respectively. The mutations increased their binding affinity for lactose and kcat values for lactose and GDP-fucose. The Δ52 FutA quadruple mutant (A128N/H129E/Y132I/S46F) was successfully applied to in vitro synthesis of 3-FL with an improved yield and productivity (>96% yield based on 5 mM of GDP-Fuc within 1 h). Biotechnol. Bioeng. 2016;113: 1666-1675. © 2016 Wiley Periodicals, Inc.

3-Fucosyllactose (3-FL) is one of the major fucosylated oligosaccharides in human milk. Along with 2′-fucosyllactose (2′-FL), it is known for its prebiotic, immunomodulator, neonatal brain development, and antimicrobial function.Whereas the biological production of 2′-FL has been widely studied and made significant progress over the years, the biological production of 3-FL has been hampered by the low activity and insoluble expression of α-1,3-fucosyltransferase (FutA), relatively low abundance in human milk oligosaccharides compared with 2′-FL, and lower digestibility of 3-FL than 2′-FL by bifidobacteria.In this study, we report the gram-scale production of 3-FL using E. coli BL21(DE3). We previously generated the FutA quadruple mutant (mFutA) with four site mutations at S46F, A128N, H129E, Y132I, and its specific activity was increased by nearly 15 times compared with that of wild-type FutA owing to the increase in k cat and the decrease in K m . We overexpressed mFutA in its maximum expression level, which was achieved by the optimization of yeast extract concentration in culture media. We also overexpressed L-fucokinase/GDP-L-fucose pyrophosphorylase to increase the supply of GDP-fucose in the cytoplasm. To increase the mass of recombinant whole-cell catalysts, the host E. coli BW25113 was switched to E. coli BL21(DE3) because of the lower acetate accumulation of E. coli BL21(DE3) than that of E. coli BW25113. Finally, the lactose operon was modified by partially deleting the sequence of LacZ (lacZΔm15) for better utilization of D-lactose. Production using the lacZΔm15 mutant yielded 3-FL concentration of 4.6 g/L with the productivity of 0.076 g·L −1 ·hr −1 and the specific 3-FL yield of 0.5 g/g dry cell weight. K E Y W O R D S fucosyllactose, fucosyltransferase, human milk, oligosaccharide, metabolic engineering *Yun Hee Choi and Bum Seok Park contributed equally to this study.

2'-Fucosyllactose (2'-FL) is the most abundant oligosaccharide in human milk and one of the most actively studied human milk oligosaccharides (HMOs). When 2'-FL is produced through biological production using a microorganism, like Escherichia coli, D-lactose is often externally fed as an acceptor substrate for fucosyltransferase (FT). When D-glucose is used as a carbon source for the cell growth and D-lactose is transported by lactose permease (LacY) in lac operon, D-lactose transport is under the control of catabolite repression (CR), limiting the supply of D-lactose for FT reaction in the cell, hence decreasing the production of 2'-FL. In this study, a remarkable increase of 2'-FL production was achieved by relieving the CR from the lac operon of the host E. coli BL21 and introducing adequate site-specific mutations into α-1,2-FT (FutC) for enhancement of catalytic activity and solubility. For the host engineering, the native lac promoter (P lac ) was substituted for tac promoter (P tac ), so that the lac operon could be turned on, but not subjected to CR by high D-glucose concentration. Next, for protein engineering of FutC, family multiple sequence analysis for conserved amino acid sequences and protein-ligand substrate docking analysis led us to find several mutation sites, which could increase the solubility of FutC and its activity. As a result, a combination of four mutation sites (F40S/ Q150H/C151R/Q239S) was identified as the best candidate, and the quadruple mutant of FutC enhanced 2'-FL titer by 2.4-fold. When the above-mentioned E. coli mutant host transformed with the quadruple mutant of futC was subjected to fedbatch culture, 40 g l −1 of 2'-FL titer was achieved with the productivity of 0.55 g l −1 h −1 and the specific 2'-FL yield of 1.0 g g −1 dry cell weight.

Although recent advances in deep learning approaches for protein engineering have enabled quick prediction of hot spot residues improving protein solubility, the predictions do not always correspond to an actual increase in solubility under experimental conditions. Therefore, developing methods that rapidly confirm the linkage between computational predictions and empirical results is essential to the success of improving protein solubility of target proteins. Here, we present a simple hybrid approach to computationally predict hot spots possibly improving protein solubility by sequence-based analysis and empirically explore valuable mutants using split GFP as a reporter system. Our approach, Consensus design Soluble Mutant Screening (ConsenSing), utilizes consensus sequence prediction to find hot spots for improvement of protein solubility and constructs a mutant library using Darwin assembly to cover all possible mutations in one pot but still keeps the library as compact as possible. This approach allowed us to identify multiple mutants of Escherichia coli lysine decarboxylase, LdcC, with substantial increases in soluble expression. Further investigation led us to pinpoint a single critical residue for the soluble expression of LdcC and unveiled its mechanism for such improvement. Our approach demonstrated that following a protein’s natural evolutionary path provides insights to improve protein solubility and/or increase protein expression by a single residue mutation, which can significantly change the profile of protein solubility.