Biotechnological production of human milk oligosaccharides

Biotechnological production of human milk oligosaccharides

Biotechnology Advances 30 (2012) 1268–1278 Contents lists available at SciVerse ScienceDirect Biotechnology Advances journal homepage: www.elsevier...

700KB Sizes 0 Downloads 26 Views

Biotechnology Advances 30 (2012) 1268–1278

Contents lists available at SciVerse ScienceDirect

Biotechnology Advances journal homepage: www.elsevier.com/locate/biotechadv

Research review paper

Biotechnological production of human milk oligosaccharides Nam Soo Han a, Tae-Jip Kim a, Yong-Cheol Park b, Jaehan Kim c, Jin-Ho Seo d,⁎ a

Department of Food Science and Technology, Chungbuk National University, Cheongju 361-763, Republic of Korea Department of Advanced Fermentation Fusion Science and Technology, Kookmin University, Seoul 136-702, Republic of Korea c Department of Food Nutrition, Chungnam National University, Daejeon 305-764, Republic of Korea d Department of Agriculture Biotechnology, Seoul National University, Seoul 151-742, Republic of Korea b

a r t i c l e

i n f o

Available online 15 November 2011 Keywords: Human milk oligosaccharides HMOs Sialyllactose Fucosyllactose Lacto-N-biose

a b s t r a c t Human milk contains a large variety of oligosaccharides (HMOs) that have the potential to modulate the gut flora, affect different gastrointestinal functions, and influence inflammatory processes. This review introduces the recent advances in the microbial and coupled enzymatic methods to produce HMOs with grouping them into trisaccharides (sialyllactose and fucosyllactose) and complex oligosaccharides (lacto-N-biose derivatives). The high purity and low cost of HMOs should make their use possible in new fields such as the food or pharmaceutical industries. © 2011 Elsevier Inc. All rights reserved.

Contents 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . 1.1. Structure of HMOs . . . . . . . . . . . . . . . . 1.2. Physiological properties of HMOs . . . . . . . . . 1.3. Recovery from natural resources . . . . . . . . . 1.4. Chemical synthesis of HMOs . . . . . . . . . . . 2. Biological production of sialyllactose . . . . . . . . . . . 2.1. Coupled microbial method . . . . . . . . . . . . 2.2. Single-cell methods . . . . . . . . . . . . . . . 2.3. Coupled enzymatic methods . . . . . . . . . . . 3. Biological production of fucosyllactose . . . . . . . . . . 3.1. GDP-fucose synthetic pathway . . . . . . . . . . 3.2. Enzymatic and microbial synthesis of GDP-L-fucose. 3.3. Production of fucosyllactose . . . . . . . . . . . 4. Production of complex oligosaccharides . . . . . . . . . 4.1. Building block (lactose-N-biose) . . . . . . . . . 4.2. Lacto-N-oligosaccharide derivatives . . . . . . . . 4.3. Fucosylated-complex oligosaccharides . . . . . . . 5. Potential and future works of HMOs . . . . . . . . . . . Competing interests . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

1. Introduction During the first several months of life, people rely on milk, whether formula or human milk. Despite the balanced nutritional

⁎ Corresponding author. Tel.: + 82 2 880 4855; fax: + 82 2 873 5095. E-mail address: [email protected] (J.-H. Seo). 0734-9750/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.biotechadv.2011.11.003

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

1268 1269 1269 1270 1270 1270 1270 1270 1272 1272 1272 1273 1273 1274 1274 1274 1275 1275 1276 1276 1276

supplementation of formula milk, human milk is yet considered as the best and gold standard for the purpose of achieving the wellbeing of infants (Bode, 2009). In addition to the traditional nutrients, human milk contains essential components that promote infant health. A part of these functional ingredients is human milk oligosaccharides (HMOs) (Bode, 2006, 2009). Oligosaccharides in human milk carry very distinctive features that those in other mammalian milk do not have. The concentration of HMOs is up to 15 g/L which is as high

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

as proteins in milk (Zivkovic et al., 2011). Only trace amounts of oligosaccharides are present in mature milk of domestic animals such as cows, goats, pigs, and sheep (Lane et al., 2010) and, as a consequence, in cow's milk-based infant formula. The structure of HMOs is very diverse showing more than 200 compositional and structural isomers in pooled milk (German et al., 2008). The potential health benefits of HMOs have been studied over the years with an emphasis on their prebiotic effects. Despite the fact that the HMOs play a key role for the beneficiary impact of breast milk on infants (Kunz and Rudloff, 2006), HMOs are mostly absent from infant formula due to the lack of industrial production methods. Instead, galactooligosaccharides and/or fructooligosaccharides are often added to infant formulas as an inexpensive alternative expecting the prebiotic effect to promote a bacterial microflora that closely resembles that of breastfed infants. Owing to their important biological functions that were recently discovered (Bode, 2009; Kunz et al., 2000; Zivkovic et al., 2011), HMOs have attracted substantial interest and many synthesis methods for HMOs or mimics have been studied. Chemical synthesis requires the repetitive and multiple protection and deprotection steps for each HMO molecule which decrease the yields and productivity, thus a great effort has been put into biological synthesis such as microbial and enzymatic methods (Koizumi et al., 1998). For example, the synthesis of sialylated oligosaccharides has been reported using bacterial sialyltransferase which was expressed in Escherichia coli in conjunction with the multiple microbial or enzymatic synthesis of cytidine monophosphate (CMP)-Neu5Ac (Endo et al., 1999; Endo et al., 2000; Gilbert et al., 1998). In this respect, this review introduces the biotechnological synthesis methods and the recent scientific achievement in mass production of HMOs. 1.1. Structure of HMOs One liter of mature human milk contains about 5–15 g of free oligosaccharides, which is similar to the amount of milk proteins depending on the lactation periods (Bode, 2006). HMOs are comprised by the five monosaccharides: D-glucose (Glc), D-galactose (Gal), N-acetylglucosamine (GlcNAc), L-fucose (Fuc), and sialic acid (Sia; N-acetyl neuraminic acid [Neu5Ac]). The structures of HMOs are very diverse and complicated. Having different compositions and glycosyl linkages, more than 200 isomers were found with different degrees of polymerization (DP 3 to 20). Despite their structural complexity, HMOs share some common backbones. Most HMOs have the lactose (Galβ1-4Glc) residue at the reducing end (Fig. 1). Gal in lactose can be sialylated in α-(2,3)- and/or α-(2,6)-linkages to form 3′-sialyllactose and 6′-sialyllactose, respectively. Lactose can also be fucosylated in α-(1,2)- and α-(1,3)-linkages to form 2′-fucosyllactose

1269

and 3′-fucosyllactose, respectively. These trisaccharides are called the short-chain milk oligosaccharides. To form the complex milk oligosaccharides, N-acetyllactosamine (Galβ1-3/4GlcNAc) repeating units are propagated in a linear or branched pattern. Finally the lactose or polylactosamine backbone can be sialylated in α-(2,3)- and/or α-(2,6)-linkages and/or fucosylated in α-(1,2)-, α-(1,3)-, and/or α-(1,4)-linkages. Approximately 200 different complex oligosaccharides have been identified in human milk (Table 1). In contrast, infant formula contains only trace amounts of oligosaccharides that are less structurally complex (Zivkovic et al., 2011). The 3 most abundant oligosaccharides were fucosylated and the proportion of fucosylated ones was approximately 77% (137 out of 183) (Ninonuevo et al., 2006). Of the remaining oligosaccharides those that have been proposed to be biologically interesting in particular were the sialylated ones which were approximately 28% of HMOs (39 out of 183) (Kobata, 2010; Ninonuevo et al., 2006). 1.2. Physiological properties of HMOs The prebiotic effect of human milk has been studied from the middle of the 20th century. György et al. (1954) reported that the components of human milk have been known to promote the growth of Bifidobacterium bifidum by their prebiotic effect. Recent studies showed that this prebiotic effect (also known as Bifidogenic effect) is linked to the oligosaccharide in human milk. It was reported that the infant-borne bifidobacteria preferentially consume small mass HMOs initially then consume completely in a late stage of cell growth (LoCascio et al., 2007). Furthermore, most intestinal bacteria including food-borne pathogens cannot grow using HMOs as a sole carbon source (Mills et al., 2011). With these regards, Zivkovic et al. (2011) suggested a model that explains how HMOs influence the human health and gastrointestinal microbiota. The carbohydrate moiety that binds lectin of most pathogens often shares the similar structure with parts of HMOs. It suggests that HMOs possibly serve as soluble ligand analogs and block pathogen adhesion protecting breast-fed infants against infections and diarrhea (Newburg et al., 2005). It has been shown that the fucosylated HMOs inhibit the binding of Campylobacter to human intestinal mucosa ex vivo (Ruiz-Palacios et al., 2003). The diarrhea in breast-fed infants caused by Campylobacter is inversely related to the amount of 2′-fucosyllactose in the mother's milk (Morrow et al., 2004). It was also reported that the human milk inhibits adhesion of Streptococcus pneumoniae and Haemophilus influenzae to human pharyngeal and buccal epithelial cells (Andersson et al., 1986). HMOs can influence the intestinal environment as well. It has been reported that one of the sialylated oligosaccharides in human milk can change the glycome of intestinal epithelial cells. Microarray

Fig. 1. Structural composition of human milk oligosaccharides (HMOs) (Bode, 2009). HMOs can be grouped into short-chain trisaccharides (sialyllactose or fucosyllactose) and complex high-molecular-weight oligosaccharides.

1270

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

1.4. Chemical synthesis of HMOs

Table 1 Composition of human milk oligosaccharides. Adapted from Ninonuevo et al., 2006. Sugar composition Hex

HexNA

Fuc

4 3 4 4 5 5 5 4 6 3 6 4 5 4 6 3 6 3 4 5 3 5 5 9

2 1 2 2 3 3 3 2 4 1 4 2 3 2 4 2 4 2 1 3 1 3 2 1

1

NeuAc

2 2 1 3 1 2 1 1 1

1

3 3 1

4 1 1 1 1

1

Molecular weight

Relative amounts (%)

Accumulated amounts (%)

1220.454 709.264 1366.512 1074.396 1731.644 1585.586 1877.702 1511.549 2096.776 1000.359 1950.718 1365.492 1439.528 1512.570 2242.834 1058.401 1804.661 912.343 871.317 2023.760 855.322 1876.682 1382.507 1915.655

20.35% 16.90% 10.95% 7.92% 5.84% 4.79% 4.01% 3.06% 2.93% 2.50% 2.35% 2.17% 1.61% 1.51% 1.44% 1.16% 1.12% 1.08% 0.77% 0.64% 0.57% 0.54% 0.42% 0.41%

20.35% 37.25% 48.20% 56.12% 61.96% 66.75% 70.77% 73.83% 76.76% 79.26% 81.61% 83.78% 85.39% 86.90% 88.34% 89.50% 90.62% 91.70% 92.47% 93.11% 93.68% 94.22% 94.63% 95.04%

glycan profiling of Caco-2 intestinal epithelial cells reveals a reduction in specific glycan epitopes upon exposure to 3′-sialyllactose, one of the sialylated HMOs (Angeloni et al., 2005). In this study, the gene expression levels of various glycosyltransferases including α-2,3sialyltransferases were reduced, suggesting that exogenous 3′sialyllactose modifies the cell surface glycome profile by regulating the expression of genes encoding the enzymes involved in glycan assembly. Although humans can carry the hydrolyzing enzymes, HMOs are partially absorbed intact in the infant's intestine and subsequently appear in the urine of breast-fed infants (Gnoth et al., 2001). Assuming that HMOs reach the systemic circulation, they may alter protein–carbohydrate interactions at a systemic level. In vitro and ex vivo assays have shown that leukocyte rolling and adhesion as well as platelet–neutrophil formation and activation are reduced in the presence of physiologically relevant HMO concentrations (Bode et al., 2004a, 2004b). However, the biochemical underpinnings are not clear and HMOs appear to compete with physiological ligands for selectin binding that have a glycan moiety similar to HMOs, consequently resulting in modulation of the immune system.

The chemical synthesis of long chain glycans is still challenging. The development of automated solid-phase oligosaccharide synthesis (Plante et al., 2001) enables the fast and efficient production of HMOlike sugars. In solid-phase synthesis, the oligosaccharides being synthesized are linked to an insoluble material (beads or resins) that allows the fast separation of reaction products from excess reagents, soluble reaction by-products, and solvents (Seeberger and Werz, 2005). The synthesis of complex oligosaccharides such as the Lewis X–Lewis Y required more than 1 year using solution-phase synthesis; it can now be accomplished in 1 day using automated solid-phase synthesis (Routenberg Love and Seeberger, 2004). Recently, a crystalline intermediate technology was developed to assist a purification step in protection and deprotection of saccharides and by using this method fast synthesis of 2′-O-fucosyllactose was possible (Dékany et al., 2010). However, despite the development of chemical synthesis methods, there are several hurdles for the industry-scale synthesis of HMOs and they are low stereoselectivity, low overall yields, and the use of toxic reagents not suitable for food products. 2. Biological production of sialyllactose The development of efficient systems for biological production of HMOs started with the synthesis of short-chain trisaccharides, particularly sialyllactose, through the identification of sialyltransferase genes that are expressed functionally in E. coli (Gilbert et al., 1998). To date, various methods have been developed that are mainly based on the microbial and enzymatic systems described below. 2.1. Coupled microbial method For the industrial production of sialyllactose, a coupled microbial method has been developed by Kyowa Hakko Kogyo company (Endo et al., 2000). In this system, Corynebacterium ammoniagenes was employed for UTP synthesis and 2 recombinant E. coli strains were used for the CMP-NeuAc biosynthesis and sialic acid transfer reaction. After a 27-h reaction starting with orotic acid and N-acetylneuraminic acid, CMP-NeuAc accumulated at 27 mM (17 g/L) in C. ammoniagenes. Consecutively, E. coli cells expressing the α-2,3-sialyltransferase gene of Neisseria gonorrhoeae produced sialyllactose at 52 mM (33 g/L) after an 11-h reaction starting with orotic acid, N-acetylneuraminic acid, and lactose. To facilitate passive circulation of the substrates between the coupled microbial cells, xylene was added and the reaction was carried out in non-growing cells. This method had also been used for synthesis of other oligosaccharides such as globotriose (Koizumi et al., 1998) and lactosamine (Endo et al., 1999). 2.2. Single-cell methods

1.3. Recovery from natural resources Naturally occurring oligosaccharides can be isolated from the milk of other species. Although the milk oligosaccharides are uniquely abundant and diverse in humans, they are also found in the milk of most mammals, i.e. primates, cows, pigs, goats, sheep, and elephants (Kunz et al., 1999). It has been already shown that HMO-like oligosaccharides were separated by membrane technology on a large scale from by-products in goat cheese production (Martinez-Ferez et al., 2006). In this study, a two-stage tangential ultrafiltration–nanofiltration system was used for oligosaccharide separation from goat milk. A virtually lactose and salt-free product was obtained containing more than 80% of the original oligosaccharide content. In addition, separation processes of sialyloligosaccharides in waste streams from cheese processing and from other dairy sources using α-2,3-trans-sialidase enzymes have been described (Pelletier et al., 2004).

Because of the high complexity and low productivity of the coupled microbial method, single-cell methods have been developed that use recombinant E. coli systems. Ishikawa and Koizumi (2010) produced NeuAc by using recombinant E. coli cells by expression of GlcNAc 2-epimerase encoded by slr1975 from Synechocystis sp. and NeuAc synthetase encoded by neuB. In this system, an acetateresistant mutant strain was used for high-density cell culture and the NeuAc aldolase gene (nanA) was disrupted to prevent NeuAc digestion. The recombinant E. coli produced 172 mM (53 g/L) for a 22 h reaction with 540 mM (120 g/L) GlcNAc in a 5 L jar fermenter. As shown in Fig. 2, sialyllactose or complex sialylated oligosaccharides can be produced in living E. coli cells either with synthesis of NeuAc from glucose via the de novo synthetic pathway (Fierfort and Samain, 2008) or with its uptake from medium via the salvage method (Dékany et al., 2010). In this system, lactose was internalized by LacY permease and

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

1271

Fig. 2. Engineered metabolic pathways for the production of sialyllactose and GD3 ganglioside either with an endogenous synthesis of NeuAc from glucose (A) (Fierfort and Samain, 2008) or with its uptake from medium (B) (adapted from Drouillard et al., 2010). Over-expressed genes are in bold. Discontinued arrows represent the enzymatic activities that have been eliminated. Lactose and Neu5AC are transported into the cell by specific permeases LacY and NanT, and they are protected from degradation because of βgalactosidase (lacZ) and aldolase (nanA) inactivation. The expression of both cytidine monophosphate (CMP)-NeuAc synthetase and sialyltransferase allows the intracellular activation of NeuAc into CMP-NeuAc and its subsequent transfer on lactose and sialyllactose. Glycerol provides the carbon and energy source required for bacterial growth, lactose transport, and sialyllactose synthesis.

sialylated by glycosyltransferase, using CMP-Neu5Ac, which was constantly regenerated by the enzymatic machinery of the living cells (Priem et al., 2002). For the salvage synthesis method, Neu5Ac was transported into the cell by NanT permease and it was protected from degradation because of aldolase (nanA) inactivation. For the de novo synthesis of sialyllactose, the pathway for the synthesis of CMP-Neu5Ac had to be introduced into E. coli strain K12 derivatives, since the E. coli strain able to produce CMP-Neu5Ac in nature is a pathogenic strain K1 that is unsuitable for biotechnological use. To accomplish this, the α-2,3-sialyltransferase gene from Neisseria meningitidis was coexpressed

with the neuC, neuB and neuA Campylobacter jejuni genes encoding N-acetylglucosamine-6-phosphate-epimerase, sialic acid synthase and CMP-Neu5Ac synthetase, respectively. Simultaneously, Neu5Ac aldolase, ManNAc kinase, and β-galactosidase activities were eliminated to ensure a higher yield of sialyllactose. Sialyllactose concentration of 25 g/L was obtained with continuous lactose feed in the high cell density culture of the recombinant E. coli cell (Fierfort and Samain, 2008). This system was further extended to the production of the gangliosides GD3 (Dékany et al., 2010), GM2 and GM1 (Antoine et al., 2003) by additionally expressing the appropriate glycosyltransferase genes.

Fig. 3. Scheme for the coupled enzymatic synthesis of sialyllactose. Substrate molecules are in the box. The oval frame represents the intermediate molecules to be recycled. Discontinuing arrows represent the alternative cytidine triphosphate (CTP) recycle pathway. PPK, polyphosphate kinase; PolyPn, polyphosphate.

1272

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

2.3. Coupled enzymatic methods For the enzymatic synthesis of sialyllactose, sequential enzyme reactions and a nucleotides recycle system should be linked with substrates in a one-pot reaction (Fig. 3). For the production of Neu5Ac from GlcNAc, GlcNAc 2-epimerase (EC 5.1.3.8) and Neu5Ac aldolase (EC 4.1.3.3) were commonly used in sequence (Lee et al., 2004, 2007). GlcNAc 2-epimerase converts GlcNAc to N-acetyl-D-mannosamine (ManNAc), which then reacts with pyruvate to form Neu5Ac through the enzymatic action of Neu5Ac aldolase. The mammalian GlcNAc 2epimerase, a known renin-binding protein (Maru et al., 1996; Takahashi et al., 2001), has been used to produce Neu5Ac (Lee et al., 2004). GlcNAc 2-epimerase from Synechocystis sp. has been also used for practical applications (Tabata et al., 2002). The second enzyme, Neu5Ac aldolase (previously named N-acetylneuraminate lyase) was found in E. coli K12 (Ohta et al., 1985) and many pathogenic bacteria (Li et al., 2008). The E. coli gene for Neu5Ac has been overexpressed and used for Neu5Ac production from ManNAc and pyruvate (Mahmoudian et al., 1997) and for the production of 2-keto-3-deoxyD-glycero-D-galacto-nonopyranulosonic acid (KDN) from D-mannose and pyruvate (Wang and Lee, 2006). To establish an efficient synthetic procedure for Neu5Ac, Neu5Ac aldolase has been immobilized on Eupergit-C (Mahmoudian et al., 1997) and has also been retained in membrane reactors using ultrafiltration membranes (Salagnad et al., 1997). Furthermore, a gene fusion technology using GlcNAc 2-epimerase and Neu5Ac aldolase was used to efficiently synthesize of sialic acid (Wang et al., 2009). Bacterial CMP-Neu5Ac synthetases have been used in biotechnology applications, while the enzymes were found in both eukaryotic and prokaryotic organisms (Table 2). The enzyme from N. meningitidis has frequently been used and comprehensive enzyme information is available in the BRENDA database (http://www.brenda-enzymes.org) where a reader can select specific organisms to obtain a full description of the enzyme. As shown in the upper part of Fig. 3, CMP-NeuAc can be produced via 2 different cytidine triphosphate (CTP) regeneration systems using acetate kinase, cytidine monophosphate (CMP) kinase, and polyphosphate kinase enzymes. CTP was synthesized from an inexpensive substrate CMP in high yield of more than 97% via cytidine diphosphate (CDP) by CMP-kinase and acetate kinase (Woo et al., 2008). According to the study, a mixture of various substrates and enzymes was subjected to a one-pot reaction, thus, the reactivity of each enzyme could be maximized, and CMP-Neu5Nac could be synthesized at a high yield of more than 90%, even if small amounts of enzymes and substrates were used. As a phosphate donor for CDP, a very small amount of adenosine triphosphate (ATP) was used, and the use of adenosine diphosphate (ADP) was economical in that it was reused by excess acetyl Pi and acetate kinase. Particularly, in this method, a complicated process for removing ADP remaining as an impurity after reaction completion can be eliminated through the use of CTP instead of ATP as the phosphate donor for CDP. For the conversion of CDP to CTP as well as for the ATP regeneration, an acetate kinase was used. There are reports executing ATP regeneration by using endogenous acetate kinase in E. coli extract without

any exogenous enzyme addition (Ryabova et al., 1995). Endogenous inorganic pyrophosphatase in E. coli was also utilized to degrade the inorganic pyrophosphate (Lee et al., 2002). Alternatively, the enzymatic method using inorganic polyphosphate (polyP) as the sole phosphorus donor has been developed to regenerate CTP from CMP (Lee et al., 2002). The combined activity of polyP kinase (PPK) and CMP kinase from E. coli converted CMP into CDP, while PPK by itself further phosphorylated CDP. This regeneration system was coupled with CMP-NeuAc synthetase of H. influenzae to synthesize CMP-NeuAc. ATP or CTP can also be synthesized from cognate NMPs in a polyP-dependent manner. It has been known that PPK is responsible for the synthesis of polyP from ATP in E. coli and that it catalyzes polyP-dependent phosphorylation of all 4 NDPs. It was demonstrated that polyP and PPK function as an ATP substitute for adenylate kinase or CMP kinase to give polyP: AMP or CMP phosphotransferase activities (Ishige et al., 2001). 3. Biological production of fucosyllactose 3.1. GDP-fucose synthetic pathway L-Fucose is a special mono-sugar present in HMOs and glycosylation residues in eukaryotic glycoproteins and glycolipids (Becker and Lowe, 2003; Ma et al., 2006). In animals, for example, L-fucose is a part of glycoconjugates such as Lewis group antigens, which have been demonstrated to have a role in various types of biochemical recognition processes (Tonetti et al., 1998). It is also a component of colonic acid, an extracellular polysaccharide in enteric bacteria such as E. coli and Salmonella enterica. Recently, fucosylated conjugates have gained an increasing interest from many pharmaceutical companies to ensure that GDP-L-fucose, an activated form of L-fucose, is stably and sufficiently supplied. GDP-L-fucose production procedures have been developed in chemical and biological manners. In the chemical synthesis of GDPL-fucose, L-fucopyranosyl tetraacetate was used as the starting material, and toxic chemicals of HBr, Ag2CO3 and tetra-n-butylammonium dibenzylphosphate were added to initiate the chemical reactions (Gokhale et al., 1990; Nunez et al., 1981). An efficient supply of guanosine monophosphate (GMP) morpholdate and multiple purification steps should be established. In biological and especially microbial methods, two different metabolic pathways are identified; the salvage pathway and the de novo pathway. In the salvage pathway found in humans, extracellular L-fucose is transferred into the cells and phosphorylated by L-fucose kinase (EC 2.7.1.52) at the expense of ATP. L-Fucose-1-phosphate is combined with guanosine triphosphate (GTP) by L-fucose-1-phosphate guanylyltransferase (EC 2.7.7.30) to produce GDP-L-fucose (Becker and Lowe, 2003). The de novo pathway was identified in bacteria, mammals, and plants (Albermann et al., 2000; Stevenson et al., 1996), of which a representative reaction scheme is shown in Fig. 4. Fructose-6phosphate, an intermediate in the glycolysis from glucose, is metabolized into mannose-1-phosphate by mannose-6-phosphate isomerase (ManA, E.C. 5.3.1.8) and phosphomannomutase (ManB, E.C. 5.4.2.8). Mannose-1-phosphate is combined with GTP by mannose-1phosphate guanyltransferase (ManC, E.C. 2.7.7.22), resulting in the formation of GDP-D-mannose. GDP-D-mannose 4,6-dehydratase (Gmd,

Table 2 Cytidine monophosphate (CMP)-sialic acid synthetase: source and characteristics (Mizanur and Pohl, 2008). Enzyme source (MW)

E. coli (49,000) N. meningitidis (24,800) Group B streptococci (45,500) H. ducreyi (25,400) P. haemolytica (43,000) C. thermocellum (26,000)

Temp. optima

pH optima

Cation

37 37 25–37 37 37 50

9.0 8.5 8.3–9.4 8–9.5 9.0 9.5

Mg2 + Mg2 + or Mn2 + Mg2 + Mg2 + Mg2 + Mg2 +

KM (mM) Neu5Ac

CTP

4 0.34 7.6 0.26 1.82 0.13

0.31 0.31 1.4 0.035 1.77 0.24

Sulfhydryl requirement

References

DTT (stimulatory) DTT (required) DTT (stimulatory) No No DTT (stimulatory)

Vann et al. (1987) Warren and Blacklow (1962) Haft and Wessels (1994) Tullius et al. (1996) Bravo et al. (2001) Mizanur and Pohl (2007)

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

Fig. 4. The de novo pathway for GDP-L-fucose production from glucose. The names of enzymes are abbreviated as follows; ManA, mannose-6-phosphate isomerase; ManB, phosphomannomutase; ManC, mannose-1-phosphate guanylyltransferase; Gmd, GDP-D-mannose-4,6-dehydratase; WcaG, GDP-L-fucose synthase. PPP, ATP, GTP and P denote the pentose phosphate pathway, adenosine-triphosphate, guanosinetriphosphate and phosphate, respectively.

E.C. 4.2.1.47) then removed a water molecule from GDP-D-mannose. GDP-L-fucose synthase (WcaG, E.C. 1.1.1.271) catalyzes the reduction of the keto group at the C4 position of GDP-4-keto-6-deoxymannose to synthesize GDP-L-fucose, where reduced nicotinamide dinucleotide phosphate (NADPH) is supplied as a reducing power (Albermann et al., 2000; Becker and Lowe, 2003; Jang et al., 2010). In the overall reaction, 1 mol of glucose is converted into 1 mol of GDP-L-fucose at the expense of 1 mol each of ATP, GTP and NADPH as cofactors. 3.2. Enzymatic and microbial synthesis of GDP-L-fucose For enzymatic production of GDP-L-fucose, the genes encoding Gmd and WcaG enzymes from E. coli K12 were overexpressed under the E. coli-T7 promoter system and purified using the His-tag system (Albermann et al., 2000). GDP-D-mannose used as a substrate was successfully converted to GDP-L-fucose by a 2-step enzymatic reaction, of which production yield was achieved at 78%. This enzymatic process has the hurdles of strong product inhibition of GDP-L-fucose and a supply of expensive NADPH cofactor. Microbial production of GDP-L-fucose has been undertaken in Saccharomyces cerevisiae, E. coli and C. ammoniagenes. Since S. cerevisiae is known to have a rich pool of GDP-D-mannose, the E. coli genes coding for Gmd and WcaG were transformed into S. cerevisiae, of which expression was controlled by the galactose-inducible promoters of GAL1 and GAL10, respectively (Mattila et al., 2000). In simple batch fermentation using galactose as a carbon source, recombinant S. cerevisiae produced 0.2 mg/L GDP-L-fucose. In vitro biosynthesis using yeast hydrolysates showed that an increase in GDP-L-fucose production was achieved using an external supply of GDP-D-mannose and not NADPH, suggesting that reducing power was not a limiting factor in this system (Mattila et al., 2000). S. cerevisiae was engineered to coexpress gmd and wcaG from Arabidopsis thaliana and in vitro enzymatic assay provided that WcaG might help the structural maintenance of the active Gmd (Nakayama et al., 2003). Meanwhile, the salvage synthetic method using L-fucose as a starting material was employed for GDP-L-fucose production in recombinant S. cerevisiae

1273

overexpressing the Bacteroides fragilis fkp gene coding for both fucose kinase and fucose-1-phosphate guanylyltransferase (Liu et al., 2011). By using this recombinant system, 230 nmol of GDP-fucose was produced by addition of 15 mM L-fucose into 50 mL of culture medium. Besides, E. coli is a well-known workhorse for production of proteins and metabolites and harbors the endogenous manA, manB and manC genes. The gmd and wcaG genes from E. coli K12 were transformed into an enzyme-deficient E. coli BL21(DE) strain and overexpressed by isopropylthio-β-galactoside (IPTG) induction (Byun et al., 2007). Simple batch fermentation with glucose as a starting substance resulted in 3.8 mg/L GDP-L-fucose and a further improvement of a fermentation strategy by adapting a glucose-limited fed-batch fermentation enhanced GDP-L-fucose concentration by 10 times (38.9 mg/L). As mentioned above, NADPH is a cofactor involved in the WcaGmediated reaction. An NADPH-regenerating metabolic enzyme, glucose-6-phosphate dehydrogenase (G6PDH), which is encoded by the zwf1 gene, was co-expressed in recombinant E. coli strains producing GDP-L-fucose. Finally 55 mg/L GDP-L-fucose was obtained by fed-batch fermentation (Byun et al., 2007). Combinatorial expression of 5 enzymes was carried out to determine the rate-limiting metabolic step in GDP-L-fucose production from glucose (Lee et al., 2009). Coexpression of ManB, ManC, Gmd and WcaG gave the highest titer of GDP-L-fucose (0.17 g/L) in fed-batch fermentation using glucose. ManA expression did not influence GDP-L-fucose production. The final step catalyzed by WcaG requires NADPH as the reducing power, sufficient supply of which was assessed by overexpression of NADPH-dependent metabolic enzymes such as isocitrate dehydrogenase (Icd), malate dehydrogenase (MaeB) and G6PDH. The optimized strain of E. coli BL21star (DE3) overexpressing ManB, ManC, Gmd, WcaG and G6PDH produced 0.24 g/L GDP-L-fucose (Lee et al., 2011). Using GMP and mannose as starting materials, GDP-L-fucose (18.4 g/L) was produced in a bioconversion process with a dual microbial system composed of recombinant E. coli cells overexpressing the GDP-L-fucose biosynthetic enzymes and C. ammoniagenes cells producing GTP (Koizumi et al., 2000). 3.3. Production of fucosyllactose The availability of large amounts of fucose-containing oligosaccharides would make them useful as precursors or ready-to-use drugs for fundamental investigation and therapeutic purposes. The chemical synthesis of fucosyloligosaccharides including the Lewis blood group antigen has long been achieved (Gokhale et al., 1990; Kameyama et al., 1991; Kretzschmar and Stahl, 1998). However, this method is expensive and time consuming and requires multiple protection and deprotection steps (Kameyama et al., 1991; Kretzschmar and Stahl, 1998). Instead of chemical synthesis, biocatalytic strategies using enzymes or whole microbial cells may be more efficient for fucosylated oligosaccharide synthesis. Enzymatic fucosylation of sugars requires GDP-L-fucose as a fucose donor and fucosyltransferase transferring L-fucose to an acceptor molecule (Albermann et al., 2000; Bülter and Elling, 1999). On the basis of the L-fucose binding site, fucosyltransferases are classified into α-(1,2)-, α-(1,3/4)-, α-(1,6)- and O-fucosyltransferases (Ma et al., 2006). Various fucosyltransferases have been identified in mammals, most of which are located in the Golgi apparatus except for O-fucosyltransferase present in the endoplasmic reticulum. The 13 fucosyltransferase genes have been isolated in the human genome and well summarized in a review (Becker and Lowe, 2003). For example, FUT1 and FUT2 are responsible for the H blood group antigen synthesis, while FUT3-7 and FUT9 are involved in the synthesis of the Lewis X [Galβ1-4(Fucα1-3)GlcNAc] and sialyl Lewis X antigens. Various types of bacterial fucosyltransferases with α-(1,2)- and α-(1,3/4)- fucosylation activity were identified in E. coli K12, E. coli O128 strains, S. enterica LT2, Yersinia enerocolitica, Helicobacter pylori

1274

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

and Vibrio cholerae, of which characteristics is well documented in a recent review (Ma et al., 2006). In addition to the characterization of fucosyltransferases at the gene and protein levels, this review focuses on the biological production of fucosyl oligosaccharides by fucosyltransferases only or by whole cell catalysts expressing fucosyltransferases. The enzymatic synthesis of fucosyloligosaccharide was examined by using α-1,2-fucosyltransferase from H. pylori (FucT2), GDP-L-fucose and lactose as the starting materials to produce 2′-fucosyllactose with 65% yield (Albermann et al., 2001). Several fucose-containing lactooligosaccharides were produced in recombinant E. coli by overexpression of the positive regulator protein in the colanic acid biosynthesis, RcsA, and by inactivation of the colanic acid biosynthetic UDP-glucose lipid carrier transferase (WcaJ), which is involved in GDP-L-fucose consumption in the colanic acid synthesis (Dumon et al., 2004, 2006). α-1,2-Fucosyltransferase from H. pylori (FutC), β-1,3-N-acetylglucosaminyltransferase and β-1,4-galactosyltransferase from N. meningitidis (LgtA and LgtB, respectively) were co-expressed in a recombinant E. coli strain overexpressing the rcsA gene and deficient in the wcaJ and lacZ genes (Drouillard et al., 2006). By sequential feeding of glucose and lactose in fed-batch fermentation, 3 g/L of oligosaccharide mixture was produced that was composed of 57% Fucα-2Galβ-4Glc (fucosyllactose) and 23% Fucα-2Galβ-4GlcNAcβ-3Galβ-4Glc. Without the lgtA and lgtB expression and with the addition of more lactose, 11 g/L fucosyllactose was produced extracellularly (Drouillard et al., 2006). A dual microbial system consisting of recombinant E. coli expressing the GDP-L-fucose synthetic enzymes and recombinant C. ammoniagenes expressing α-(1,3)-fucosyltransferase from H. pylori was developed to produce 21 g/L Lewis X from GMP, mannose and N-acetyl lactosamine as the substrates (Koizumi et al., 2000). Even though L-fucose is a more expensive source for fucosyloligosaccharide production than glucose, the alternative salvage synthetic method using L-fucose was employed for fucosyllactose production. An E. coli strain was engineered to overexpress E. coli K12 L-fucose transporter (FucP) and Bacteroides fragilis fucose kinase/fucose-1-phosphate guanylyltransferase (Fkp) under the control of the T7 promoter (Hüfner et al., 2010). To block the L-fucose metabolism for cell growth, its chromosome was more modified by disruption of the fucA gene encoding fuculose-1-phosphate aldolase. This recombinant E. coli strain was able to convert L-fucose to 369 μM GDP-fucose. In recombinant E. coli expressing the fucP and fkp genes, and losing the chromosomal fucA gene, additional overexpression of the H. pylori fucT2 or codon-optimized futAco gene allowed the production of 2′- or 3′-fucosyllactose. Besides the fucosyltransferase processes, it was reported that Thermotoga maritima α-L-fucosidase was evolved into α-Ltransfucosidase with high transglycosylation activity to generate pNP-Gal-α-1,2-Fuc from a pNP-fucoside donor (Osanjo et al., 2006). However, this enzymatic synthesis has critical problems of low product selectivity and acceptor specificity. As an alternative way, the glycosynthase technology was applied to the enzymatic production of 2′-fucosyllactose (Wada et al., 2008). An inverting 1,2-α-L-fucosidase from Bif. bifidum was converted to a 1,2-α-Lfucosynthase via D766G mutation, which could synthesize 2′fucosyllactose from a β-L-fucosyl fluoride triacetate donor and lactose acceptor. Even though these approaches might be possible, the chemical instability of the donor molecules should be solved for the enhanced conversion yield.

building block of HMOs and one of the bifidus growth factor candidates (Urashima et al., 2009). Lactic acid bacteria (LABs), including bifidobacteria, have received much focus for their beneficial effects on human health due to their rapid growth in human intestine resulting in the prevention of pathogenic infection. It was found that the bifidobacteria possess a unique metabolic pathway utilizing galacto-N-biose (GNB; Galβ1-3GalNAc) and lacto-N-biose I (Kitaoka et al., 2005). Nevertheless, insufficient amounts of most HMOs in nature were a barrier to elucidate their physiological roles in the human health in relation to LABs. An industrial scale and cost-effective process for LNB-I production from sucrose and GlcNAc was developed via a one-pot enzymatic reaction (Nishimoto and Kitaoka, 2007). Four kinds of enzymes including sucrose phosphorylase (SP; EC 2.4.1.7) from Bifidobacterium longum, UDP-glucose-hexose-1-phosphate uridylyltransferase (UGH1PUT; EC 2.7.7.12), UDP-glucose-4-epimerase (UG4E; EC 5.1.3.2), and lacto-Nbiose phosphorylase (LNBP; EC 2.4.1.211) from B. bifidum were applied to the LNB production and kilograms of LNB could be obtained with an approximate 83% reaction yield in the presence of UDP-glucose and inorganic phosphate (Fig. 5). In the same way, GNB can be easily produced using a one-pot enzymatic reaction (Nishimoto and Kitaoka, 2009). An alternative process to produce GNB or LNB was suggested using E. coli K12 galactokinase and Bifidobacterium infantis D-galactosyl-β1-3-N-acetyl-D-hexosamine phosphorylase (Yu et al., 2010). Recently, in vivo studies on the prebiotic effects of LNB indicated that some Bifidobacterium strains predominantly found in infant intestines can utilize LNB as a sole carbon source (Kiyohara et al., 2009; Xiao et al., 2010). 4.2. Lacto-N-oligosaccharide derivatives Human milk contains a variety of oligosaccharide derivatives that consist of lactose and several repeating lacto-N-biose units modified by terminal fucosylation with α-(1,2)-, α-(1,3)-, and α-(1,4)-linkages, and/or sialylation with α-(2,3)- and α-(2,6)-linkages. Among these oligosaccharide derivatives in HMOs, 2′-fucosyllactose (Fucα12Galβ1-4Glc), lacto-N-tetraose (LNT; Galβ1-3GlcNAcβ1-3Galβ1-4Glc), lacto-N-fucopentaose I (LNFP-I; Fucα1-2Galβ1-3GlcNAcβ1-3Galβ14Glc), lacto-N-fucopentaose II (LNFP-II; Galβ1-3[Fucα1-4]GlcNAcβ13Galβ1-4Glc), lacto-N-fucopentaose III (LNFP-III; Galβ1-4[Fucα1-3] GlcNAcβ1-3Galβ1-4Glc), and lacto-N-difucohexaose I (LNDFH-I; Fucα1-2Galβ1-3[Fucα1-4]GlcNAcβ1-3Galβ1-4Glc) were identified as the major oligosaccharide types (Kunz et al., 2000). Lacto-N-tetraose and lacto-N-neotetraose (LNnT; Galβ1-4GlcNAcβ13Galβ1-4Glc) were synthesized by sequential conversion from lactose as Pi

Sucrose

Fructose

Glucose-1-P

SP UDP-Galactose

UG4E

UGH1PUT UDP-Glucose

Pi

Lacto-N-biose

GlcNAc

Galactose-1-P

LNBP ADP

4. Production of complex oligosaccharides 4.1. Building block (lactose-N-biose) In common HMOs, lactose (Galβ1-4Glc) at the reducing end is linked with several repeating units such as lacto-N-biose I (LNB-I; Galβ1-3GlcNAc) or lacto-N-biose II (LNB-II; Galβ1-4GlcNAc) units via β-(1,3)-linkages. In particular, LNB-I is known as a dominant

Mg2+ GK ATP

Galactose Fig. 5. The reaction scheme of one-pot enzymatic production of lacto-N-biose (LNB)-I (Nishimoto and Kitaoka, 2007). N-Acetylglucosamine (GlcNAc), sucrose phosphorylase (SP), UDP-glucose-hexose-1-phosphate uridylyltransferase (UGH1PUT), UDP-glucose4-epimerase (UG4E), lacto-N-biose phosphorylase (LNBP), and galactokinase (GK).

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

1275

In vivo production of lacto-N-fucooligosaccharides was carried out using metabolically engineered E. coli (Dumon et al., 2001; Priem et al., 2002). A mixture of LNT and LNnT was obtained from high cell density culture of an E. coli strain co-overexpressing Neisseria β-(1,3)N-acetyl glucosaminyltransferase, β-(1,4)-galactosyltransferase, and Helicobacter α-(1,3)-fucosyltransferase (FucT) genes. Liu et al. (2011) developed sequential 2-step synthesis of sialyl T-antigens and derivatives, including α-2,3-sialylated LNB. During the first step, galactose was enzymatically converted to LNB via a one-pot reaction with galactokinase and LNB phosphorylase. The resulting LNB was successively transformed to an α-2,3-sialylated LNB using a one-pot 3-enzyme sialylation process employing Pasteurella multocida sialic acid aldolase, N. meningitidis CMP-sialic acid synthetase, and P. multocida α-2,3-sialyltransferase. For the controlled enzymatic production of complex HMOs, a variety of transferases with highly specific donor/acceptor preference should be obtained and utilized (Table 3).

a starting material (Fig. 6) (Murata et al., 1999). Firstly, lacto-N-triose II (GlcNAcβ1-3Galβ1-4Glc) was readily prepared by transferring the N-acetylglucosamine residue from UDP-GlcNAc to lactose using bovine serum β-(1,3)-N-acetylglucosaminyltransferase (β-1,3GnT). In the next step, the resulting lacto-N-triose II was converted into LNT by Bacillus circulans β-D-galactosidase-mediated transglycosylation with orthonitrophenyl β-D-galactopyranoside (Galβ-oNP) donor, while the lactose donor was utilized for LNnT synthesis. In addition, it was reported that large-scale production of lacto-N-triose II and LNnT was performed using bacterial glycosyltransferases, β-(1,3)-N-acetyl-glucosaminyltransferase and β-(1,4)-D-galactosyltransferase from N. gonorrhoeae (Johnson, 1999). Using lacto-N-biosidase from Aureobacterium species, conversion of lactose into lacto-N-tetraose was also tried (Murata et al., 1999). Lacto-N-biosidase is an endo-glycosidase which can catalyze the transfer of the β-lacto-N-biosyl residue from Galβ1-3GlcNAc-pNP to the 3′-hydroxyl group of lactose. However, its conversion efficiency was too low to be applicable for further studies.

5. Potential and future works of HMOs 4.3. Fucosylated-complex oligosaccharides Studies of human milk oligosaccharides are expected to be very useful in the development of new antibacterial or antiviral compounds. Some carbohydrate epitopes from HMOs might circulate in the infant's blood for a period of time before they are excreted. Hence, prevention of inflammatory bowel diseases or their chronic manifestation in breast-fed infants by carbohydrate interactions with mucosal leukocytes might be possible. Today, extensive efforts are being made to develop new drugs based on carbohydrates with nonimmunological functions. To pick up minor but useful oligosaccharides from the mixture of a large number of different milk oligosaccharides, the key to success is how one will be able to develop an effective method for isolation. To

In order to obtain various lacto-N-fucooligosaccharides, terminal fucosylation might be essential in the final step. Human α-(1,3)and α-(1,4)-fucosyltransferases were applied to produce a variety of lacto-N-fucooligosaccharides (Nimtz et al., 1998). Meanwhile, LNnT was converted into LNFP-III by α-(1,3)-fucosyltransferase from chicken serum (Totani et al., 2002). LNDFH-I was also enzymatically synthesized via fucosylations of LNT (Miyazaki et al., 2010). Human fucosyltransferase I (FUT1) transferred L-fucose to the D-galactose residue of LNT with an α-(1,2)-linkage (Fig. 6). The resulting LNFP-I was subsequently reacted with GDP-β-L-fucose and commercial fucosyltransferase III (FUT3) to generate LNDFH-I.

HO

UDP-GlcNAc

OH O

OH O

HO

O HO

OH

UDP

OH

HO HO

-1,3-GnT

OH

OH O

Lactose

HO

O NHAc

OH O

O HO

OH

OH

OH

Lacto-N-triose II

Gal β- NP

NP

OH O

Lactose -Gal

OH O HO

HO HO

OH O

HO

O NHAc

O

OH O OH

OH

O HO

OH O

Glucose OH

HO

OH

Lacto-N-tetraose

HO

GDP-Fuc

FUT1

OH O OH

OH O

O HO

HO

O NHAc

Lacto-N-neotetraose

GDP

OH O HO

HO HO H3C HO

OH O O

HO O

OH

NHAc

O

OH O

O HO

OH O

OH

HO

OH O

HO

OH

Lacto-N-fucopentaose I

OH

OH O

O HO

HO

O NHAc

Lacto-N-neofucopentaose

GDP-Fuc

OH

OH O

HO H3C HO

O O

H3C

OH O

O O

OH

OH OH O

HO

O

NHAc

OH O OH

OH

OH

O

OH

OH

GDP

O HO

OH O OH

O

HO OH

OH O

HO H3C HO

OH

OH O

GDP-Fuc

O OH

GDP-Fuc

FucT

H3C OH

OH

OH

OH

FucT

GDP

HO

OH

HO

FUT3 O

OH O

O HO

GDP

O OH

OH O

Lacto-N-difucohexaose I

OH O

O

O NHAc

O

OH

HO

OH O H3C

O OH

OH

HO

OH O

O O

OH

OH

OH

O OH

OH

Lacto-N-neodifucohexaose

Fig. 6. Enzymatic synthesis of various lacto-N-oligosaccharide derivatives and complex oligosaccharides (Miyazaki et al., 2010; Murata et al., 1999). UDP-N-acetylglucosamine (UDPGlcNAc), GDP-N-fucose (GDP-Fuc), ortho-nitrophenyl β-D-galactopyranoside (Galβ-oNP), β-(1,3)-N-acetyl-glucosaminyltransferase (β-1,3-GnT), and β-(1,4)-D-galactosidase (β-Gal), FUT1, FUT2, and FucT (fucosyltransferases).

1276

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

Table 3 Productivities of human milk oligosaccharides (HMOs) and their intermediates depending on substrates and synthetic methods. Product

Synthetic methods

Enzyme and/or microbial systems

Substrates

Productivity (yield)

Ref.

NeuAc CMP-NeuAc GDP-fucose

Microbial Microbial Microbial

Sialyllactose

Microbial

GlcNAc Orotic acid, NeuAc Glc Glc Glycerol, lactose Orotic acid, NeuAc, lactose GlcNAc, Pyruvate Lactose, CTP, Acetyl Pi (polyP), ATP

53 g/L 17 g/L 0.24 g/L 18.4 g/L 25 g/L 33 g/L >100 g/L (> 90%) (personal communication)

Ishikawa and Koizumi (2010) Endo et al. (2000) Lee et al. (2011) Koizumi et al. (2000) Fierfort and Samain (2008) Endo et al. (2000) Woo et al. (2008)

Glc, Lactose Sucrose, GlcNAc, Pi, UDP-Glc Galactose, GlcNAc, ATP Lactose, UDP-GlcNAc Lactose, UDP-GlcNAc Lacto-N-triose II, Galβ-oNP Lacto-N-triose II, Lactose

14 g/L 140 g/L (83%) (95%) 2.5 g/L (85%) b μM (26%) (20%) (19%)

Drouillard et al. (2006) Nishimoto and Kitaoka (2007) Yu et al. (2010) Yu et al. (2010) Murata et al. (1999) Murata et al. (1999) Murata et al. (1999)

Fucosyllactose Lacto-N-biose (LNB)

Microbial Enzymatic

Lacto-N-triose II

Enzymatic

Lacto-N-tetraose (LNT) Lacto-N-neotetraose (LNnT) Lacto-N-fucopentaose I (LNFP-I) Lacto-N-difucohexaose I (LNDFH-I) Lacto-N-neofucopentaose (LNnFP)

Enzymatic Enzymatic

E. coli (slr1975, neuB, nanA−) E. coli (neuA, pyrG) + C. ammoniagenes E. coli (gmd, wcaG, zwf1) E. coli + C. ammoniagenes E. coli (LacZ−, nanA−, neuABC) C. ammoniagenes + E. coli (2,3ST) Acetate kinase, PPK, CMP kinase, GlcNAc epimerase, NeuAc aldolase, CMP-NeuAc synthetase E. coli (futC, rcsA, wcaJ−, lacZ−) SP, LNBP, UG4E, UGH1PUT GK, GalHexNAcP β-1,3-GnT β-1,3-GnT β-galactosidase β-galactosidase

Enzymatic

1,2FT

LNT, GDP-Fuc

3.0 mg/3.51 μmol (71%)

Miyazaki et al. (2010)

Enzymatic

1,4FT

LNFP-I, GDP-Fuc

1.7 mg (85%)

Miyazaki et al. (2010)

Microbial

E. coli (lgtB, lgtA, fucT, lacZ−, wcaJ−)

Lactose, glucose

3.0 g/L

Dumon et al. (2001)

Enzymatic

SP, sucrose phosphorylase; LNBP, lacto-N-biose phosphorylase, UG4E, UDP-Glc-4-epimerase; UGH1PUT, UDP-glucose-hexose-1-phosphate uridylyltransferase; GK, galactokinase; GalHexNAcP, galactosyl-β1-3-N-acetyl-hexosamine phosphorylase; β-1,3-GnT, β-1,3-N-acetylglucosaminetransferase; 2,3ST, α-(2,3)-sialyltransferase; 1,2FT, α-(1,2)-fucosyltransferase; 1,4FT, α-(1,4)-fucosyltransferase.

screen for potentially bioactive HMOs, at least 2 approaches are applicable: oligosaccharides can be passed over a column-bound lectin or HMOs can be immobilized on glyco-chips and the binding specificities of lectins can be determined. For further development of the production methods, elucidation of the structures of all human milk oligosaccharides is also indispensable. As described above, different core oligosaccharides were found of human milk oligosaccharides. Whether additional larger cores exist and the locations of additional fucoses and sialic acid residues remain to be elucidated. Understanding how HMOs are synthesized in the human mammary gland could guide us in producing HMOs. HMOs carry lactose at the reducing end, which is synthesized in the Golgi of mammary gland epithelial cells. In the presence of α-lactalbumin, which is specifically expressed during lactation, the substrate specificity of β-(1,4)galactosyltransferase shifts from N-acetylglucosamine to galactose linked to glucose to form lactose. This review introduced the 2 biotechnological methods for HMOs production that were established using microbial and coupled enzymatic methods. Comparing the enzymatic and microbial processes for fucosyloligosaccharide production, the enzymatic approach may

be less desirable for large-scale synthesis due to the number of required enzymes and cofactors involved in the synthesis of GDP-Lfucose. The whole-cell approach may be more realistic for industrial applications since it does not require enzyme isolation or expensive starting materials and there is limited availability of necessary elements for oligosaccharide synthesis. However, as shown in Table 4, the coupled enzymatic method in the one-pot reaction shows rather promising results for the synthesis of sialyllactose, lactose-N-biose, and complex oligosaccharides with high yield, even if small amounts of enzymes and substrate are used. These achievements resulted from the development of innovative nucleotide-reuse technology and the discovery of novel and specific enzymes for glycosyl transfer reactions. Economical production of various HMOs will provide us better opportunities for a variety of future studies on the physiological functions of HMOs.

Table 4 Commercially available human milk oligosaccharides (HMOs) and their manufacturers.

This research was supported by World Class University (WCU) program (R32-2008-000-10183-0) and the Advanced Biomass R&D Center (ABC) (2010-0029799) both funded by the Korean Ministry of Education, Science and Technology.

HMOs

Manufacturers

Sialyllactose (3′-sialyllactose, 6′-sialyllactose)

Carbosynth (Compton, Berkshire, UK) Dextra (Reading, UK) Elicityl (Crolles, France) Genechem (Daejeon, Korea) Kyowa Hakko (Kogyo, Japan) Prozyme (Hayward, USA) Carbosynth (Compton, Berkshire, UK) Dextra (Reading, UK) Elicityl (Crolles, France) Jennewein (Rheinbreitbach, Germany) Prozyme (Hayward, USA) Carbosynth (Compton, Berkshire, UK) Dextra (Reading, UK) Elicityl (Crolles, France)

Fucosyllactose (2′-fucosyllactose 3′-fucosyllactose)

Lacto-N-biose Lacto-N-tetraose Other HMOs

Competing interests The authors have no competing interests to declare. Acknowledgments

References Albermann C, Distler J, Piepersberg W. Preparative synthesis of GDP-β-L-fucose by recombinant enzymes from enterobacterial sources. Glycobiology 2000;10: 875–81. Albermann C, Piepersberg W, Wehmeier UF. Synthesis of the milk oligosaccharide 2′-fucosyllactose using recombinant bacterial enzymes. Carbohydr Res 2001;334: 97-103. Andersson B, Porras O, Hanson LÅ, Lagergård T, Svanborg-Edén C. Inhibition of attachment of Streptococcus pneumoniae and Haemophilus influenzae by human milk and receptor oligosaccharides. J Infect Dis 1986;153:232–7. Angeloni S, Ridet JL, Kusy N, Gao H, Crevoisier F, Guinchard S, et al. Glycoprofiling with micro-arrays of glycoconjugates and lectins. Glycobiology 2005;15:31–41.

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278 Antoine T, Priem B, Heyraud A, Greffe L, Gilbert M, Wakarchuk WW, et al. Large-scale in vivo synthesis of the carbohydrate moieties of gangliosides GM1 and GM2 by metabolically engineered Escherichia coli. Chembiochem 2003;4:406–12. Bülter T, Elling L. Enzymatic synthesis of nucleotide sugars. Glycoconj J 1999;16:147–59. Becker DJ, Lowe JB. Fucose: biosynthesis and biological function in mammals. Glycobiology 2003;13:41R–53R. Bode L. Recent advances on structure, metabolism, and function of human milk oligosaccharides. J Nutr 2006;136:2127–30. Bode L. Human milk oligosaccharides: prebiotics and beyond. Nutr Rev 2009;67: S183–91. Bode L, Kunz C, Muhly-Reinholz M, Mayer K, Seeger W, Rudloff S. Inhibition of monocyte, lymphocyte, and neutrophil adhesion to endothelial cells by human milk oligosaccharides. Thromb Haemost 2004a;92:1402–10. Bode L, Rudloff S, Kunz C, Strobel S, Klein N. Human milk oligosaccharides reduce platelet–neutrophil complex formation leading to a decrease in neutrophil ß 2 integrin expression. J Leukoc Biol 2004b;76:820–6. Bravo IG, Barrallo S, Ferrero MA, Rodríguez-Aparicio LB, Martínez-Blanco H, Reglero A. Kinetic properties of the acylneuraminate cytidylyltransferase from Pasteurella haemolytica A2. Biochem J 2001;358:585–98. Byun SG, Kim MD, Lee WH, Lee KJ, Han NS, Seo JH. Production of GDP-L-fucose, L-fucose donor for fucosyloligosaccharide synthesis, in recombinant Escherichia coli. Appl Microbiol Biotechnol 2007;74:768–75. Dékany G, Bajza I, Boutet J, Pérez IF, Hederos M, Horvath F, et al., inventor. Synthesis of 2′-o-fucosyllactose. WIPO patent 2010115935. 2010 Apr 7. Drouillard S, Driguez H, Samain E. Large-scale synthesis of H-antigen oligosaccharides by expressing Helicobacter pylori α1,2-fucosyltransferase in metabolically engineered Escherichia coli cells. Angew Chem 2006;118:1810–2. Drouillard S, Mine T, Kajiwara H, Yamamoto T, Samain E. Efficient synthesis of 6'-sialyllactose, 6,6'-disialyllactose, and 6'-KDO-lactose by metabolically engineered E. coli expressing a multifunctional sialyltransferase from the Photobacterium sp. JT-ISH-224. Carbohydr Res 2010;345:1394–9. Dumon C, Bosso C, Utille JP, Heyraud A, Samain E. Production of Lewis x tetrasaccharides by metabolically engineered Escherichia coli. Chembiochem 2006;7:359–65. Dumon C, Priem B, Martin SL, Heyraud A, Bosso C, Samain E. In vivo fucosylation of lacto-N-neotetraose and lacto-N-neohexaose by heterologous expression of Helicobacter pylori α-1,3 fucosyltransferase in engineered Escherichia coli. Glycoconj J 2001;18:465–74. Dumon C, Samain E, Priem B. Assessment of the two Helicobacter pylori α-1,3-fucosyltransferase ortholog genes for the large-scale synthesis of LewisX human milk oligosaccharides by metabolically engineered Escherichia coli. Biotechnol Prog 2004;20:412–9. Endo T, Koizumi S, Tabata K, Kakita S, Ozaki A. Large-scale production of N-acetyllactosamine through bacterial coupling. Carbohydr Res 1999;316:179–83. Endo T, Koizumi S, Tabata K, Ozaki A. Large-scale production of CMP-NeuAc and sialylated oligosaccharides through bacterial coupling. Appl Microbiol Biotechnol 2000;53:257–61. Fierfort N, Samain E. Genetic engineering of Escherichia coli for the economical production of sialylated oligosaccharides. J Biotechnol 2008;134:261–5. German JB, Freeman SL, Lebrilla CB, Mills DA. Human milk oligosaccharides: evolution, structures and bioselectivity as substrates for intestinal bacteria. Nestle Nutr Workshop Ser Pediatr Program 2008;62:205–22. Gilbert M, Bayer R, Cunningham A-M, DeFrees S, Gao Y, Watson DC, et al. The synthesis of sialylated oligosaccharides using a CMP-Neu5Ac synthetase/sialyltransferase fusion. Nat Biotechnol 1998;16:769–72. Gnoth MJ, Rudloff S, Kunz C, Kinne RKH. Investigations of the in vitro transport of human milk oligosaccharides by a Caco-2 monolayer using a novel high performance liquid chromatography–mass spectrometry technique. J Biol Chem 2001;276:34363–70. Gokhale UB, Hindsgaul O, Palcic MM. Chemical synthesis of GDP-fucose analogs and their utilization by the Lewis α (1–>4) fucosyltransferase. Can J Chem 1990;68: 1063–71. György P, Norris RF, Rose CS. Bifidus factor I. A variant of Lactobacillus bifidus requiring a special growth factor. Arch Biochem Biophys 1954;48:193–01. Hüfner E, Parkot J, Jennewein S, inventor. Synthesis of fucosylated compounds. WIPO patent 2010070104. 2010 Jun 24. Haft RF, Wessels MR. Characterization of CMP-N-acetylneuraminic acid synthetase of group B streptococci. J Bacteriol 1994;176:7372–4. Ishige K, Hamamoto T, Shiba T, Noguchi T. Novel method for enzymatic synthesis of CMP-NeuAc. Biosci Biotechnol Biochem 2001;65:1736–40. Ishikawa M, Koizumi S. Microbial production of N-acetylneuraminic acid by genetically engineered Escherichia coli. Carbohydr Res 2010;345:2605–9. Jang MH, Lee WH, Shin SY, Han NS, Seo JH, Kim MD. Molecular cloning of the genes for GDP-mannose 4, 6-dehydratase and GDP-L-fucose synthetase from Bacteroides thetaiotaomicron. Food Sci Biotechnol 2010;19:849–55. Johnson KF. Synthesis of oligosaccharides by bacterial enzymes. Glycoconj J 1999;16: 141–6. Kameyama A, Ishida H, Kiso M, Hasegawa A. Total synthesis of sialyl Lewis X. Carbohydr Res 1991;209:C1–4. Kitaoka M, Tian J, Nishimoto M. Novel putative galactose operon involving lacto-N-biose phosphorylase in Bifidobacterium longum. Appl Environ Microbiol 2005;71:3158–62. Kiyohara M, Tachizawa A, Nishimoto M, Kitaoka M, Ashida H, Yamamoto K. Prebiotic effect of lacto-N-biose I on bifidobacterial growth. Biosci Biotechnol Biochem 2009;73:1175–9. Kobata A. Structures and application of oligosaccharides in human milk. Proc Jpn Acad Ser B Phys Biol Sci 2010;86:731–47.

1277

Koizumi S, Endo T, Tabata K, Nagano H, Ohnishi J, Ozaki A. Large-scale production of GDPfucose and Lewis X by bacterial coupling. J Ind Microbiol Biotechnol 2000;25:213–7. Koizumi S, Endo T, Tabata K, Ozaki A. Large-scale production of UDP-galactose and globotriose by coupling metabolically engineered bacteria. Nat Biotechnol 1998;16: 847–50. Kretzschmar G, Stahl W. Large scale synthesis of linker-modified sialyl LewisX, LewisX and N-acetyllactosamine. Tetrahedron 1998;54:6341–58. Kunz C, Rudloff S. Health promoting aspects of milk oligosaccharides. Int Dairy J 2006;16:1341–6. Kunz C, Rudloff S, Baier W, Klein N, Strobel S. Oligosaccharides in human milk: structural, functional, and metabolic aspects. Annu Rev Nutr 2000;20:699–722. Kunz C, Rudloff S, Schad W, Braun D. Lactose-derived oligosaccharides in the milk of elephants: comparison with human milk. Br J Nutr 1999;82:391–9. Lane JA, Mehra RK, Carrington SD, Hickey RM. The food glycome: a source of protection against pathogen colonization in the gastrointestinal tract. Int J Food Microbiol 2010;142:1-13. Lee JO, Yi JK, Lee SG, Takahashi S, Kim BG. Production of N-acetylneuraminic acid from N-acetylglucosamine and pyruvate using recombinant human renin binding protein and sialic acid aldolase in one pot. Enzyme Microb Technol 2004;35:121–5. Lee SG, Lee JO, Yi JK, Kim BG. Production of cytidine 5′-monophosphate N-acetylneuraminic acid using recombinant Escherichia coli as a biocatalyst. Biotechnol Bioeng 2002;80: 516–24. Lee WH, Chin YW, Han NS, Kim MD, Seo JH. Enhanced production of GDP-L-fucose by overexpression of NADPH regenerator in recombinant Escherichia coli. Appl Microbiol Biotechnol 2011:1-10. Lee WH, Han NS, Park YC, Seo JH. Modulation of guanosine 5′-diphosphate-D-mannose metabolism in recombinant Escherichia coli for production of guanosine 5′-diphosphate-Lfucose. Bioresour Technol 2009;100:6143–8. Lee YC, Chien HCR, Hsu WH. Production of N-acetyl-D-neuraminic acid by recombinant whole cells expressing Anabaena sp. CH1 N-acetyl-D-glucosamine 2-epimerase and Escherichia coli N-acetyl-D-neuraminic acid lyase. J Biotechnol 2007;129:453–60. Li Y, Yu H, Cao H, Lau K, Muthana S, Tiwari V, et al. Pasteurella multocida sialic acid aldolase: a promising biocatalyst. Appl Microbiol Biotechnol 2008;79:963–70. Liu TW, Ito H, Chiba Y, Kubota T, Sato T, Narimatsu H. Functional expression of Lfucokinase/guanosine 5′-diphosphate-L-fucose pyrophosphorylase from Bacteroides fragilis in Saccharomyces cerevisiae for the production of nucleotide sugars from exogenous monosaccharides. Glycobiology 2011;21:1228–36. LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, Lebrilla CB, et al. Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates strain specific, preferential consumption of small chain glycans secreted in early human lactation. J Agric Food Chem 2007;55:8914–9. Ma B, Simala-Grant JL, Taylor DE. Fucosylation in prokaryotes and eukaryotes. Glycobiology 2006;16:158R–84R. Mahmoudian M, Noble D, Drake CS, Middleton RF, Montgomery DS, Piercey JE, et al. An efficient process for production of N-acetylneuraminic acid using Nacetylneuraminic acid aldolase. Enzyme Microb Technol 1997;20:393–400. Martinez-Ferez A, Rudloff S, Guadix A, Henkel CA, Pohlentz G, Boza JJ, et al. Goats' milk as a natural source of lactose-derived oligosaccharides: isolation by membrane technology. Int Dairy J 2006;16:173–81. Maru I, Ohta Y, Murata K, Tsukada Y. Molecular cloning and identification of N-acyl-D-glucosamine 2-epimerase from porcine kidney as a renin-binding protein. J Biol Chem 1996;271:16294–9. Mattila P, Räbinä J, Hortling S, Helin J, Renkonen R. Functional expression of Escherichia coli enzymes synthesizing GDP-L-fucose from inherent GDP-D-mannose in Saccharomyces cerevisiae. Glycobiology 2000;10:1041–7. Mills S, Ross RP, Hill C, Fitzgerald GF, Stanton C. Milk intelligence: mining milk for bioactive substances associated with human health. Int Dairy J 2011;21:377–401. Miyazaki T, Sato T, Furukawa K, Ajisaka K. Enzymatic synthesis of lacto-N-difucohexaose I which binds to Helicobacter pylori. In: Minoru F, editor. Methods in enzymology. Academic Press; 2010. p. 511–24. Mizanur R, Pohl N. Cloning and characterization of a heat-stable CMP-N-acylneuraminic acid synthetase from Clostridium thermocellum. Appl Microbiol Biotechnol 2007;76: 827–34. Mizanur R, Pohl N. Bacterial CMP-sialic acid synthetases: production, properties, and applications. Appl Microbiol Biotechnol 2008;80:757–65. Morrow AL, Ruiz-Palacios GM, Altaye M, Jiang X, Lourdes Guerrero M, Meinzen-Derr JK, et al. Human milk oligosaccharides are associated with protection against diarrhea in breast-fed infants. J Pediatr 2004;145:297–303. Murata T, Inukai T, Suzuki M, Yamagishi M, Usui T. Facile enzymatic conversion of lactose into lacto-N-tetraose and lacto-N-neotetraose. Glycoconj J 1999;16:189–95. Nakayama K, Maeda Y, Jigami Y. Interaction of GDP-4-keto-6-deoxymannose-3,5epimerase-4-reductase with GDP-mannose-4,6-dehydratase stabilizes the enzyme activity for formation of GDP-fucose from GDP-mannose. Glycobiology 2003;13:673–80. Newburg DS, Ruiz-Palacios GM, Morrow AL. Human milk glycans protect infants against enteric pathogens. Annu Rev Nutr 2005;25:37–58. Nimtz M, Grabenhorst E, Gambert U, Costa J, Wray V, Morr M, et al. In vitro α1-3 or α1-4 fucosylation of type I and II oligosaccharides with secreted forms of recombinant human fucosyltransferases III and VI. Glycoconj J 1998;15:873–83. Ninonuevo MR, Park Y, Yin H, Zhang J, Ward RE, Clowers BH, et al. A strategy for annotating the human milk glycome. J Agric Food Chem 2006;54:7471–80. Nishimoto M, Kitaoka M. Practical preparation of lacto-N-biose I, a candidate for the bifidus factor in human milk. Biosci Biotechnol Biochem 2007;71:2101–4. Nishimoto M, Kitaoka M. One-pot enzymatic production of β-D-galactopyranosyl-(1–>3)2-acetamido-2-deoxy-D-galactose (galacto-N-biose) from sucrose and 2-acetamido2-deoxy-D-galactose (N-acetylgalactosamine). Carbohydr Res 2009;344:2573–6.

1278

N.S. Han et al. / Biotechnology Advances 30 (2012) 1268–1278

Nunez HA, O'Connor JV, Rosevear PR, Barker R. The synthesis and characterization of αand β-L-fucopyranosyl phosphates and GDP fucose. Can J Chem 1981;59:2086–95. Ohta Y, Watanabe K, Kimura A. Complete nucleotide sequence of the E. coli N-acetylneuraminate lyase. Nucleic Acids Res 1985;13:8843–52. Osanjo G, Dion M, Drone J, Solleux C, Tran V, Rabiller C, et al. Directed evolution of the α-L-fucosidase from Thermotoga maritima into an α-L-transfucosidase. Biochemistry 2006;46:1022–33. Pelletier M, Barker WA, Hakes DJ, Zopf DA, inventor. Methods for producing sialyloligosaccharides in a dairy source. US patent 6706497. 2004 Mar 16. Plante OJ, Palmacci ER, Seeberger PH. Automated solid-phase synthesis of oligosaccharides. Science 2001;291:1523–7. Priem B, Gilbert M, Wakarchuk WW, Heyraud A, Samain E. A new fermentation process allows large-scale production of human milk oligosaccharides by metabolically engineered bacteria. Glycobiology 2002;12:235–40. Routenberg Love K, Seeberger PH. Automated solid-phase synthesis of protected tumor-associated antigen and blood group determinant oligosaccharides. Angew Chem Int Ed 2004;116:612–5. Ruiz-Palacios GM, Cervantes LE, Ramos P, Chavez-Munguia B, Newburg DS. Campylobacter jejuni binds intestinal H(O) antigen (Fucα1, 2Galβ1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J Biol Chem 2003;278: 14112–20. Ryabova LA, Vinokurov LM, Shekhovtsova EA, Alakhov YB, Spirin AS. Acetyl phosphate as an energy source for bacterial cell-free translation systems. Anal Biochem 1995;226:184–6. Salagnad C, Gödde A, Ernst B, Kragl U. Enzymatic large-scale production of 2-keto-3-deoxy-D-glycero-D-galacto-nonopyranulosonic acid in enzyme membrane reactors. Biotechnol Prog 1997;13:810–3. Seeberger PH, Werz DB. Automated synthesis of oligosaccharides as a basis for drug discovery. Nat Rev Drug Discov 2005;4:751–63. Stevenson G, Andrianopoulos K, Hobbs M, Reeves P. Organization of the Escherichia coli K-12 gene cluster responsible for production of the extracellular polysaccharide colanic acid. J Bacteriol 1996;178:4885–93. Tabata K, Koizumi S, Endo T, Ozaki A. Production of N-acetyl-D-neuraminic acid by coupling bacteria expressing N-acetyl-D-glucosamine 2-epimerase and N-acetyl-D-neuraminic acid synthetase. Enzyme Microb Technol 2002;30:327–33. Takahashi S, Hori K, Takahashi K, Ogasawara H, Tomatsu M, Saito K. Effects of nucleotides on N-acetyl-D-glucosamine 2-epimerases (renin-binding proteins): comparative biochemical studies. J Biochem 2001;130:815–21.

Tonetti M, Sturla L, Bisso A, Zanardi D, Benatti U, De Flora A. The metabolism of 6-deoxyhexoses in bacterial and animal cells. Biochimie 1998;80:923–31. Totani K, Shimizu K, Harada Y, Murata T, Usui T. Enzymatic synthesis of oligosaccharide containing Le(x) unit by using partially purified chicken serum. Biosci Biotechnol Biochem 2002;66:636–40. Tullius MV, Munson RS, Wang J, Gibson BW. Purification, cloning, and expression of a cytidine 5′-monophosphate N-acetylneuraminic acid synthetase from Haemophilus ducreyi. J Biol Chem 1996;271:15373–80. Urashima T, Odaka G, Asakuma S, Uemura Y, Goto K, Senda A, et al. Chemical characterization of oligosaccharides in chimpanzee, bonobo, gorilla, orangutan, and siamang milk or colostrum. Glycobiology 2009;19:499–508. Vann WF, Silver RP, Abeijon C, Chang K, Aaronson W, Sutton A, et al. Purification, properties, and genetic location of Escherichia coli cytidine 5′-monophosphate N-acetylneuraminic acid synthetase. J Biol Chem 1987;262:17556–62. Wada J, Honda Y, Nagae M, Kato R, Wakatsuki S, Katayama T, et al. 1,2-α-L-Fucosynthase: a glycosynthase derived from an inverting α-glycosidase with an unusual reaction mechanism. FEBS Lett 2008;582:3739–43. Wang FL, Cui SX, Sun LP, Qu XJ, Xie YY, Zhou L, et al. High expression of α 2, 3-linked sialic acid residues is associated with the metastatic potential of human gastric cancer. Cancer Detect Prev 2009;32:437–43. Wang TH, Lee WC. Production of 2-keto-3-deoxy-D-glycero-D-galacto-nonopyranulosonic acid (KDN) using fusion protein of N-acetyl-D-neuraminic acid aldolase. Biochem Eng J 2006;29:75–80. Warren L, Blacklow RS. The biosynthesis of cytidine 5′-monophospho-N-acetylneuraminic acid by an enzyme from Neisseria meningitidis. J Biol Chem 1962;237:3527–34. Woo JS, Sohng JK, Kim BG, Kang SY, Kim DH, Jang KS, et al., inventor. Novel N-acetylglucosamine-2-epimerase and method for producing CMP-neuraminic acid using the same. WIPO patent 2008072861. 2008 Jun 19. Xiao JZ, Takahashi S, Nishimoto M, Odamaki T, Yaeshima T, Iwatsuki K, et al. Distribution of in vitro fermentation ability of lacto-N-biose I, a major building block of human milk oligosaccharides, in bifidobacterial strains. Appl Environ Microbiol 2010;76:54–9. Yu H, Thon V, Lau K, Cai L, Chen Y, Mu S, et al. Highly efficient chemoenzymatic synthesis of β1-3-linked galactosides. Chem Commun 2010;46:7507–9. Zivkovic AM, German JB, Lebrilla CB, Mills DA. Human milk glycobiome and its impact on the infant gastrointestinal microbiota. Proc Natl Acad Sci U S A 2011;108: 4653–8.