Enzymatic Synthesis of Human Milk Fucosides α1,2-Fucosyl para-Lacto-N-Hexaose and its Isomeric Derivatives

Article abstract of DOI:10.1002/adsc.201800518

Enzymatic synthesis of para-lacto-N-hexaose and its isomeric structures as well as those α1,2-fucosylated variants naturally occurring in human milk oligosaccharide (HMOs) was achieved using a sequential one-pot enzymatic system. Three glycosylation routes comprising bacterial glycosyltransferases and corresponding sugar-nucleotide-generating enzymes were developed to facilitate efficient production of extended type-1 and type-2 N-acetyllactosamine (LacNAc) backbones and hybrid chains. Further fucosylation efficiency of two α1,2-fucosyltransferases on both type-1 and type-2 chains of the hexasaccharide was investigated to achieve practical synthesis of the fucosylated glycans. The availability of structurally defined HMOs offers a practical approach for investigating future biological applications. (Figure presented.).

Full text of DOI:10.1002/adsc.201800518

Title: Enzymatic Synthesis of Human Milk Fucosides α1,2-Fucosyl
para-Lacto-N-Hexaose and its Isomeric Derivatives
Authors: Jia-Lin Fang, Teng-Wei Tsai, Chin-Yu Liang, Jyun-Yi Li, and
Ching-Ching Yu
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Adv. Synth. Catal. 10.1002/adsc.201800518
Link to VoR: http://dx.doi.org/10.1002/adsc.201800518

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
COMMUNICATION
DOI: 10.1002/adsc.201((will be filled in by the editorial staff))
Enzymatic Synthesis of Human Milk Fucosides α1,2-Fucosyl
para-Lacto-N-Hexaose and its Isomeric Derivatives
Jia-Lin Fang, Teng-Wei Tsai, Chin-Yu Liang, Jyun-Yi Li and Ching-Ching Yu*
a
Department of Chemistry and Biochemistry, National Chung Cheng University, 168 University Road, Min-Hsiung,
Chiayi 62102, Taiwan
Fax: (+886)-5-2721040; phone: (+886)-5-2729287; e-mail: checcyu@ccu.edu.tw
Received: ((will be filled in by the editorial staff))
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201######.((Please
delete if not appropriate))
structure of HMOs is a lactose moiety at the reducing
and its isomeric structures as well as those α1,2-fucosylated end, which can be extended by disaccharide units and
Abstract. Enzymatic synthesis of para-lacto-N-hexaose
variants naturally occurring in human milk oligosaccharide
(HMOs) was achieved using sequential one-pot
rarely found monosaccharide, for example, GlcNAc
serves as the terminal unit at the non-reducing end of
a
enzymatic system. Three glycosylation routes comprising
bacterial glycosyltransferases and corresponding sugar-
nucleotide–generating enzymes were developed to facilitate
efficient production of extended type-1 and type-2 N-
acetyllactosamine (LacNAc) backbones and hybrid chains.
the
lacto-N-triose.
The
N-acetyllactosamine
(Galβ1,4GlcNAc, type-2 LacNAc) can serve as both
internal and terminal disaccharide units, while the
lacto-N-biose (Galβ1,3GlcNAc, type-1 LacNAc) can
only serve as the non-reducing end terminal
disaccharide. By assembling these building blocks
into linear and/or branched structures, HMOs with
large structural complexity can be produced (>200
discovered in nature).[11-12] Although most of the
functions of HMOs have not yet been elucidated,[11]
HMOs with high molecular weight are expected to be
biologically important. The fucosylated portion of
HMOs containing the H blood group epitope
Further
fucosylation
efficiency
of
two
α1,2-
fucosyltransferases on both type-1 and type-2 chains of the
hexasaccharide was investigated to achieve practical
synthesis of the fucosylated glycans. The availability of
structurally defined HMOs offers a practical approach for
investigating future biological applications.
Keywords: Human milk oligosaccharides;
Chemoenzymatic synthesis; Glycosyltransferases;
Fucosylation; Human microbiota
(Fucα1,2Gal)
has
antiadhesive
antimicrobial
functions, including antibacterial, antiyeast, and
antiviral activities.[13-15] The α1,2-linked fucosyl
HMOs are therefore considered attractive targets for
drug development or infant food additives. For
instance, the European Food Safety Authority has
reported that 2’-O-fucosyllactose (2’-FL) is safe to be
used in human food additives.[16]
Human milk, which contains a variety of bioactive
compounds such as proteins, fat globules, and
oligosaccharides, is an essential source of nutrition
for infants.[1-2] Human milk oligosaccharides (HMOs)
are a family of structurally diverse unconjugated
glycans and are the third most abundant solid
component of human milk following lactose and
lipids. Instead of being digested by infants, HMOs
serve as prebiotics for beneficial bacteria, thus
shaping the intestinal microbiota.[3-5] Because of their
structural similarity to some intestinal mucosal cell
surface glycans, HMOs also act as decoys and disrupt
the binding of microbial or viral lectins to host cell
receptors, thereby preventing infections of the host by
these organisms.[6-7] Increasing evidence has
demonstrated that HMOs modulate epithelial and
immune responses and reduce excessive mucosal
leukocyte infiltration and activation, decreasing the
risk of necrotizing enterocolitis.[8-10]
Linear HMO hexasaccharides with type-2
disaccharide repeat units (para-lacto-N-neohexaose,
Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAcβ1-3Galβ1-
4Glc) and an extended type-1–type-2 chain (para-
lacto-N-hexaose,
Galβ1-3GlcNAcβ1-3Galβ1-
4GlcNAcβ1-3Galβ1-4Glc) were first identified in
human milk in 1977.[17] In addition to human milk,
these hexasaccharides with fucosyl extensions are
found acting as glycan antigens in erythrocyte
membranes and are considered to have essential
biological functions.[18-19] In addition to the extended
type-2 and type-1–type-2 hybrid chain structures, the
isomeric form of hexasaccharides with extended
type-1 disaccharide units (Galβ1-3GlcNAcβ1-
3Galβ1-3GlcNAcβ1-3Galβ1-4Glc) and its fucosyl
derivatives have been observed in colonic and gastric
cancer cell lines.[20-21] Identification of the linkages
with internal LacNAc (β1-3- or β1-4-linked) through
mass spectrometry is challenging, and the glycan
HMOs comprise five monosaccharides, namely
glucose (Glc), galactose (Gal), N-acetyl glucosamine
(GlcNAc), fucose (Fuc), and sialic acid. The core
1
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
standard is usually required to obtain unambiguous
results.[22]
Because of the biological significance and clinical
importance of HMOs, several elegant approaches
have been explored for the synthesis of HMOs with
strain 26695 (HP1105),[31] β1,4-galactosyltransferase
from H. pylori strain NCTC11637 (HP0826),[32] and
β1,3-galactosyltransferase from Escherichia coli
O55:H7 (WbgO),[33] overexpressed in E. coli (Figure
S1 and Table S1). The sugar donors required for the
GT-catalyzed reaction, uridine diphosphate galactose
(UDP-Gal) and uridine diphosphate N-acetyl
glucosamine (UDP-GlcNAc), were generated through
sequential one-pot enzymatic reaction with slight
modification.[34] UDP-GlcNAc was prepared by the
action of N-acetyl hexosamine kinase from
Bifidobacterium longum (NahK)[34] and glucose-1-
hexasaccharide
cores,
including
chemical
synthesis,[23-25] enzymatic synthesis,[22][26] and whole-
cell fermentation.[27] Despite the numerous synthetic
approaches reported to date, the collective synthesis
of high-molecular-weight HMOs including linear
hexasaccharides
straightforward
with
mammalian
fucosylation
through
glycosyltransferase
(GT)-based multienzyme synthesis has only recently
been achieved.[26] As an alternative, bacterial GTs
with high-expression yields and broad substrate
specificity are known as a versatile tool for complex
oligosaccharide preparation.[28-30] Herein, we report
an effective enzyme-catalyzed system using bacterial
GTs for the synthesis of nonfucosylated HMO-type
phosphate
thymidylyltransferase
from
Aneurinibacillus thermoaerophilus (RmlA);[35] in
addition, UDP-Gal was generated using galactokinase
from Meiothermus taiwanensis (MtGalK)[36] and
UDP-sugar pyrophosphorylase from Arabidopsis
thaliana (AtUSP).[37] By combining HP1105 and
WbgO with the corresponding UDP-sugar-generating
enzymes in the sequential one-pot system, lactose 1
was smoothly converted into defined lengths of HMO
backbone containing lacto-N-biose (type-1 LacNAc)
repeats (2-5), as shown in Scheme 1.
gylcan
backbones
from
disaccharide
to
hexasaccharide (para-lacto-N-hexaose) and their
isomeric
structures.
Subsequently,
backbone
fucosylation was performed to yield α1,2-fucosylated
HMO-type glycan products, as shown in Figure 1.
Scheme 1. Enzymatic synthesis of human milk
oligosaccharides with type-1 LacNAc backbones. The
circle containing HP1105, RmlA, and NahK represents
sequential one-pot β1,3-N-acetyl glucosaminylation; the
circle containing WbgO, AtUSP, and MtGalK represents
β1,3-galactosylation.
First, 6-azidohexyl lactoside acceptor (Lac, 1) was
prepared, and the azide group should facilitate
compound immobilization onto glass slides for
further glycan array development. Trisaccharide
lacto-N-triose (Lc3) 2 was prepared by treatment of 1
with HP1105 in the presence of UDP-GlcNAc
through sequential one-pot synthesis (see Supporting
Information). By combining the use of an SPE tC18
cartridge and size-exclusion chromatography, the
desired trisaccharide (2) was obtained in 82% yield
(165 mg).
Figure 1. Fifteen human milk oligosaccharide (HMO)-type
gly-cans and related isomeric glycans synthesized in this
study. aNot reported in human milk, but found as the
backbones of the human tumor-associated antigen.
The lacto-N-tetraose (LNT, 3) was synthesized
To efficiently and quickly produce hexasaccharide
backbones with different linkages, we used three
recombinant bacterial GTs, including β1,3-N-acetyl-
glucosaminyltransferase from Helicobacter pylori
through
WbgO-catalyzed
sequential
one-pot
galactosylation with in situ UDP-Gal preparation
(Scheme 1); the desired LNT 3 was obtained in 85%
yield (106 mg). As previously reported, MtGalK
2
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
exhibited excellent activity at a steady working
temperature ranging from 40 to 65 °C.36 Therefore, in
the in situ one-pot UDP-Gal preparation, the reaction
temperature was set at 42 °C to avoid the additional
step of temperature elevation. In addition, the UDP-
Gal could be simply purified from the reaction
mixture through recrystallization using ethanol and
water in gram-scale amounts (see Supporting
Information). Following the sequential one-pot
procedure, type-1 pentasaccharide (4, 88%, 44 mg)
and type-1 hexasaccharide (5, 85%, 22 mg) were
readily prepared through the subsequent treatment of
3 with HP1105 in the presence of UDP-GlcNAc,
followed by WbgO and UDP-Gal (Scheme 1). It is
worth mentioned that the glycan structures of type-1
pentasaccharide 4 and type-1 hexasaccharide 5 have
not been reported in human milk, but found as the
backbones of the human tumor-associated antigen.[20]
substrate for HP1105 and reacted with UDP-GlcNAc
for 63 h at 25 °C to obtain type-2 pentasaccharide 7
in 80% yield (23 mg). The resulting product was
treated with UDP-Gal under catalysis with HP0826
for 30 min at 37 °C to obtain para-lacto-N-
neohexaose (p-LNnH) 8 in 97% yield (59 mg). In
addation, hundreds of mg scale syntheses of both
LNnT 6 (517 mg) and p-LNnH 8 (137 mg) could be
achieved from 1 and 6, respectively, after two cycles
of enzymatic glycosylation reactions (see Supporting
Information). To prepare para-lacto-N-hexaose (p-
LNH, 9), 7 was used in the WbgO-catalyzed reaction
and incubated with UDP-Gal at 37 °C for 14 h to
produce 9 in 74% yield (45 mg), as illustrated in
Scheme 2. These results indicate that HP1105
exhibited excellent activity on both type-1 and type-2
acceptors (3 and 6), regardless of their penultimate
glycan sequence. The results are similar to those of
another β1,3GlcNAcTs from H. pylori strain J99
(JHP1032), which had a very similar amino acid
sequence to HP1105.[38]
To determine the suitable α1,2-fucosyltransferase
(α1,2-FucT) for the preparation of fucosylated HMOs,
two bacterial α1,2-FucTs, FutC from H. pylori strain
26695 and WbsJ from E. coli O128:B12 were
recombinant-overexpressed in E. coli for testing
(Figure S1). FutC[39] was employed in whole-cell
fermentation to produce 2’-FL and lacto-N-
neofucopentaose-I (LNnFP-I).[40] Although the
substrate specificity of FutC from H. pylori strain
NCTC 11639 was higher for type-1 than type-2
acceptors,[41] the substrate tolerance of FutC from
strain 26695 for in vitro glycan synthesis, especially
in HMO production, was slightly unclear. WbsJ was
reported to show broad acceptor specificity with
β1,3-linked acceptors (Galβ1,3GalNAc) and β1,4-
linked acceptors (lactose);[42] however, synthetic
application to HMO synthesis was not available. We
employed a simple thin-layer chromatography assay
using guanosine diphosphate (GDP)-fucose as a
donor to access the ability of these two α1,2-FucTs
(FutC and WbsJ) to transfer fucose to 9 HMO
backbone types for the production of potential H-
antigen-containing HMOs. The bifunctional L-
fucokinase/GDP-fucose pyrophosphorylase from
Bacteroides fragilis (FKP)[43] was responsible for the
preparation of GDP-fucose from L-fucose in the
presence of ATP, guanosine triphosphate, and MnCl2
in Tris–HCl buffer (100 mM, pH 7.5).
Scheme 2. Enzymatic synthesis of human milk
oligosaccharides containing type-2 LacNAc backbones.
The circle containing HP1105, RmlA, and NahK
represents
sequential
one-pot
β1,3-N-acetyl
glucosaminylation; the circle containing WbgO, AtUSP,
and MtGalK represents β1,3-galactosylation; the circle
containing HP0826, AtUSP, and MtGalK represents β1,4-
galactosylation.
The percentage conversion of a series of HMOs
containing di-, tri-, tetra-, penta-, and hexasaccharides
upon enzymatic fucosylation is illustrated in Figure 2.
The degree of α1,2-fucosylation of 1 to form 2’-FL
10 was 71% and 17% for FutC and WbsJ,
respectively. The formation of LNFP-I 11 from 3 was
conducted by FutC and WbsJ in excellent conversion
rates (92%–96%); the reaction rate of FutC was
slightly higher than that of WbsJ. Notably, no
significant activity in the difucosylation of LNT 3
was detected by WbsJ and FutC in the presence of
more GDP-fucose (2.5 eq.); however, increased
monofucosyl transfer yields were observed (97%–
With a similar procedure to the sequential one-pot
synthesis, the type-2 HMOs with N-acetyl
lactosamine (LacNAc) repeats (6, 7, and 8) were
synthesized in good to excellent yields (80%–97%).
HP0826 was the enzyme that efficiently catalyzed the
formation of β1,4-galactosidic linkages during the
type-2 HMO synthesis. As illustrated in Scheme 2,
Lc3 2 was galactosylated by HP0826 in the presence
of UDP-Gal, with incubation for 30 min at 37 °C,
using the sequential one-pot enzymatic synthesis
procedure to produce lacto-N-neotetraose (LNnT, 6)
in 94% yield (58 mg). The LNnT 6 was used as the
3
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
98%). In addition, the use of excess GDP-fucose (10
eq.) resulted in only terminal monofucosylated
products, indicating the infeasibility of internal α1,2-
fucosylation on HMOs by these two bacterial
enzymes (Figure S2 and S3). The conversion of the
FutC-catalyzed α1,2-fucosylation reaction on LNnT 6
to form LNnFP-I 12 was moderate (54%), while
complete conversion was achieved with longer
incubation (95% for 24 h). No significant
improvement in the conversion was observed on
LNnT 6 when more GDP-fucose was employed. The
lower conversion for FutC on type-2 glycans showed
a similar tendency to the reported FutC from strain
11639.[41][44] In addition, WbsJ was almost inactive on
LNnT 6. The α1,2-fucosylation conversion on
hexasaccharides 5 and 9 by FutC were excellent
(86% and 88%). In comparison, WbsJ-catalyzed
α1,2-fucosylation resulted in a moderate conversion
on 5 (68%) and 9 (80%). Interestingly, increasing the
amount of GDP-fucose to 2.5 eq. resulted in excellent
conversion for both the WbsJ and FutC reactions
(80%–99%). The formation of 15 (α1,2-fucosyl para-
lacto-N-neohexaose) from 8 catalyzed by WbsJ in the
presence of GDP-fucose (1.2 and 2.5 eq.) was not
observed. Notably, FutC could α1,2-fucosylate the
type-2 glycan pLNnH 8 with moderate conversion
(58%–61%). These results indicate the preference of
FutC for the type-1 HMOs, with WbsJ only active on
type-1 HMOs. In addition, the elongation of the type-
1 chain and the presence of internal type-2 linkage
did not considerably affect WbsJ- or FutC- catalyzed
α1,2-fucosyl transfer. Wang and co-workers recently
reported the use of α1,2-FucT from Helicobacter
mustelae (Hmα1,2-FT) that could fucosylate the
branched HMOs with terminal type-2 LacNAcs in
excellent yields.[45] Sequence alignment of the these
α1,2-FucTs reveals the differences in the conserved
motif II and III, both are expected as the acceptor
binding domains (Figure S4).[46]
Table 1. Fucosylation of type-1 and type-2 glycans by
fucosyltransferases FutC or WbsJ.
Yield (%)
Entry Acceptor Product FutC[a] WbsJ[b]
1
2
3
4
5
31
n.r.[c]
1
3
5
6
8
9
10
11
13
12
15
14
71
76
98
39[d]
82
n.r.
n.r.
55
45
6
80
[a]
Conditions: HEPES 50 mM, pH 5.0, MnCl2 5 mM,
acceptor 10 mM, GDP-fucose 25 mM, FutC 0.35 mg/mL,
o
[b]
alkaline phosphatase (AP) 5 U/mL, 25 C.
Conditions:
Tris-HCl 50 mM, pH 7.4, MnCl2 5 mM, acceptor 10 mM,
GDP-fucose 25 mM, WbsJ 0.35 mg/mL, AP 5 U/mL, 25
[c]
[d]
oC.
n.r. = no reaction.
Yield could be improved to
98% (123 mg) in the presence of high concentration of
FutC (1 mg/mL).
Encouraged by these results, preparative synthesis
of fucosylated HMOs (10–15) was performed with
the same enzyme concentration for comparison
(Table 1). FutC-catalyzed α1,2-fucosylations on β1,3-
linked galactosides (3, 5, and 9; each 10 mM) were
completed within 3.5 h at 25 °C in HEPES buffer (50
mM, pH 5.0) in the presence of GDP-fucose (25 mM),
MnCl2 (5 mM), and alkaline phosphatase (AP) to
produce 11, 13 (α1,2-fucosyl type-1 hexasaccharide),
and 14 (α1,2-fucosyl para-lacto-N-hexaose) with
isolation yields of 71%, 98% and 80%, respectively.
Longer incubation (24 h) was required to α1,2-
fucosylate the β1,4-linked acceptors (1, 6, and 8) to
produce 10 (31% yield), 12 (39% yield), and 15 (45%
yield). Notably, although the β1,4-linked glycan was
not the best substrate for FutC, the complete α1,2-
fucosylation on type-2 glycan 6 to form 12 could still
be achieved in the presence of high enzyme
concentration. The WbsJ-catalyzed reactions were
only active on terminal β1,3-linked galactosides (3, 5,
and 9), producing α1,2-fucosylatd HMOs 11 (76%
yield), 13 (82% yield), and 14 (55% yield) in Tris–
HCl (50 mM, pH 7.4) in the presence of GDP fucose
(25 mM), MgCl2 (5 mM), 1,4-dithiothreitol (1 mM),
and AP and incubated at 37 °C for 24 h. The products
were analyzed in terms of their purity and structure
through electrospray ionization mass spectroscopy
and nuclear magnetic resonance (NMR) spectroscopy.
The structures of the compounds were confirmed
through 1D and 2D NMR analysis with full
assignments (Table S2 and S3).
Figure 2. Enzymatic fucosylation of human milk
oligosaccharides using α1,2-fucosyltransferases FutC and
WbsJ. Percentage conversion was determined through thin-
layer chromatography and using the ImageJ software after
staining and is illustrated in a vertical bar chart. Data were
averaged from three independent experiments. *Reaction
was conducted in the presence of 2.5-eq. of GDP-fucose.
The structural analysis of these compounds using
heteronuclear multiple bond correction (HMBC) 2D
NMR spectroscopy is exemplified for compound 13
in Figure 3. A cross-signal for the anomeric proton of
residue F and the C3 carbon atom of residue E was
observed (blue double arrow in Figure 3); similarly, a
cross-signal was observed for the H3 proton of
residue E and anomeric carbon atom of residue F (red
4
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
Experimental Section
double arrow in Figure 3). These cross-peaks suggest
that the glycosidic linkage was a 1→3 link. Using a
similar approach, we confirmed the 1→3 glycosidic
linkage with the backbone structure. In addition, the
terminal fucosylation was determined by the cross-
peak signal between the F and G residues. The
anomeric proton–proton coupling constants (3JH,H) of
residues B, C, D, E, F, and G were determined
through 1D NMR analysis to be 8.1, 8.5, 8.1, 8.3, 7.5,
Fucosyl type 1 hexasaccharide-C6N3 (13). To a buffered
(50 mM HEPES, pH 5.0) solution (835 μL) of 5 (10 mg,
8.3 μmol, 10 mM), GDP-fucose (13.2 mg, 21 μmol, 25
mM), manganese(II) chloride (5 mM) and alkaline
phosphatase (5 U/mL) was incubated at 25 °C in the
presence of FutC (0.35 mg/mL) with agitation at 200 rpm
for 24 h. The product formation was monitored by TLC (n-
butanol/ water/ acetic acid = 2:1:1 (v/v/v), 2 runs, Rf =
0.41). The reaction solution was quenched by heating at
100 °C for 10 min as TLC analysis indicated the
completion of the reaction. The resulting mixture was
centrifuged (4 °C, 10,000 x g, 10 min) to remove the
proteins and insoluble precipitates. The supernatant was
concentrated, purified by Sep-Pak® Vac tC18 cartridges
(Waters) followed by passing through a TOYOPEAL®
HW-40F gel (packed in BioRad column, 1.5 cm x 120 cm)
with water (10 mL/h) to obtain the desired product with an
3
and 3.9 Hz, respectively. The measured JH,H values
suggest that the glycosidic linkages of the backbone
in compound 13 were in the β-form, whereas the
terminal fucose unit was in the α-form.
isolation yield of 98% (11 mg) as the white powder. [α]28
D
= -15.4° (c 1.0, H2O); 1H NMR (700 MHz, D2O) δ 5.20 (d,
J = 3.9 Hz, 1H), 4.73 (d, J = 8.5 Hz, 1H), 4.65 (d, J = 7.5
Hz, 1H ), 4.64 (d, J = 8.3 Hz, 1H), 4.49 (d, J = 8.0 Hz, 1H),
4.44 (d, J = 8.1 Hz, 1H), 4.43 (d, J = 8.1 Hz, 1H), 4.30 (q,
J = 6.5 Hz, 1H), 4.16 (d, J = 2.5 Hz, 1H), 4.13 (d, J = 2.6
Hz, 1H), 4.02-3.87 (m, 6H), 3.86-3.62 (m, 24H), 3.62-3.46
(m, 8H), 3.36-3.27 (m, 3H), 2.06 (s, 3H), 2.04 (s, 3H),
1.69-158 (m, 4H), 1.45-1.36 (m, 4H), 1.24 (d, J = 6.6 Hz,
3H); 13C NMR (175 MHz, D2O) δ 174.93, 174.17, 103.22,
102.99, 102.83, 102.51, 101.90, 100.13, 99.41, 81.92,
81.30, 81.10, 78.27, 77.05, 76.55, 75.15, 75.05, 74.96,
74.79, 74.65, 74.60, 74.34, 73.38, 72.72, 71.75, 70.44,
70.03, 69.90, 69.31, 69.02, 68.49, 68.41, 68.32, 68.20,
67.93, 66.38, 61.05, 60.85(x2), 60.39, 60.32, 60.00, 54.90,
54.67, 51.03 28.47, 27.77, 25.55, 24.49, 22.16, 22.04,
15.16; HRMS (ESI-TOF) m/z calcd for C52H89N5O35
[M+Na]+: 1366.5236; found 1366.5189.
Acknowledgements
Figure 3. The structure of fucosylated type-1
hexasaccharide 13 and its HMBC spectrum. Anticipated
cross-peaks are indicated by either blue or red double
arrows. The results of HMBC analysis confirmed that the
linkages of the HMO backbone were β1→3 links and the
α1,2-fucosylation on terminal galactose. No resonances
were detected from 85 to 95 ppm, and thus, this region was
truncated. Residues represented are: A, Glc; B, Gal; C,
GlcNAc; D, Gal; E, GlcNAc; F, Gal; G, Fuc.
We are grateful to Prof. Warren W. Wakarchuk of Ryerson
University for FutC and WbsJ plasmids, and Prof. David H.
Kwan of Concordia University for WbgO plasmid. This research
was supported by National Chung Cheng University and the
Ministry of Science and Technology of Taiwan (MOST 105-2113-
M-194-001-MY2 and MOST 106-2113-M-194-004-MY2).
References
[1] K. G. Eriksen, S. H. Christensen, M. V. Lind, K. F.
Michaelsen, Curr. Opin. Clin. Nutr. Metab. Care. 2018,
21, 200–206.
We developed an effective method for synthesizing
15 fucosylated and nonfucosylated HMO-type
glycans and the isomeric glycans by using a few
bacterial glycosyltransferases and sugar-nucleotide–
generating enzymes. The fucosylation on HMOs from
lactose to para-lacto-N-hexaose and its isomeric
structures were examined using two α1,2-
fucosyltransferases, FutC and WbsJ. FutC exhibited
favorable catalytic properties over WbsJ in terms of
catalytic efficiency and substrate tolerance.
Additionally, both enzymes performed the α1,2-
fucosyl transfer only on the terminal Gal but not on
the internal ones. This approach is highly divergent
and can be used to prepare several oligosaccharide
linkage isomers in the scale of tens to hundreds of
[2] O. Ballard, A. L. Morrow, Pediatr. Clin. North Am.
2013, 60, 49–74.
[3] A. R. Pacheco, D. Barile, M. A. Underwood, D. A.
Mills, Annu. Rev. Anim. Biosci. 2015, 3, 419–445.
[4] M. R. Charbonneau, L. V. Blanton, D. B. DiGiulio, D.
A. Relman, C. B. Lebrilla, D. A. Mills, J. I. Gordon,
Nature 2016, 535, 48–55.
[5] L. Fernández, S. Langa, V. Martín, A. Maldonado, E.
Jiménez, R. Martín, J. M. Rodríguez, Pharmacol. Res.
2013, 69, 1–10.
[6] S. Etzold, L. Bode, Curr. Opin. Virol. 2014, 7, 101–107.
milligrams,
enabling
complete
structural
[7] J. T. Smilowitz, C. B. Lebrilla, D. A. Mills, J. B.
German, S. L. Freeman, Annu. Rev. Nutr. 2014, 34,
143–169.
characterization through NMR spectroscopy and
supplemental biological applications.
5
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
[8] L. Bode, Glycobiology 2012, 2, 1147–1162.
[29] Schmaltz, R. M.; Hanson, S. R.; Wong, C.-H. Chem.
Rev. 2011, 111, 4259–4307.
[9] A. Kulinich, L. Liu, Carbohydr. Res. 2016, 432, 62–70.
[30] L. L. Lairson, B. Henrissat, G. J. Davies, S. G.
[10] L. Bode, E. Jantscher-Krenn, Adv. Nutr. 2012, 3,
383S–391S.
Withers, Annu. Rev. Biochem. 2008, 77, 521–555.
[31] S. M. Logan, E. Altman, O. Mykytczuk, J.-R. Brisson,
V. Chandan, F. S. Michael, A. Masson, S. Leclerc, K.
Hiratsuka, N. Smirnova, J. Li, Y. Wu, W. W.
Wakarchuk, Glycobiology 2005, 15, 721–733.
[11] A. Kobata, Proc. Jpn. Acad. Ser. B Phys. Biol. Sci.
2010, 86, 731–747.
[12] X. Chen, Adv. Carbohydr. Chem. Biochem. 2015, 72,
113–190.
[32] D.-J. Namdjou, H.-M. Chen, E. Vinogradov, D.
Brochu, S. G. Withers, W. W. Wakarchuk,
ChemBioChem 2008, 9, 1632–1640.
[13] E. Castanys-Muñoz, M. J. Martin, P. A. Prieto, Nutr.
Rev. 2013, 71, 773–789.
[33] X.-w. Liu, C. Xia, L. Li, W.-y. Guan, N. Pettit, H.-c.
Zhang, M. Chen, P. G. Wang, Bioorg. Med. Chem.
2009, 17, 4910–4915.
[14] A. L. Morrow, G. M. Ruiz-Palacios, M. Altaye, X.
Jiang, M. Lourdes Guerrero, J. K. Meinzen-Derr, T.
Farkas, P. Chaturvedi, L. K. Pickering, D. S. Newburg,
J. Pediatr. 2004, 145, 297–303.
[34] W.-T. Chien, C.-F. Liang, C.-C. Yu, C.-H. Lin, S.-P.
Li, I. Primadona, Y.-J. Chen, K. K. T. Mong, C.-C. Lin,
Chem. Commun. 2014, 50, 5786–5789.
[15] J. E. Heggelund, A. Varrot, A. Imberty, U. Krengel,
Curr. Opin. Struct. Biol. 2017, 44, 190–200.
[35] W.-T. Chien, C.-F. Liang, C.-C. Yu, J.-H. Lin, H.-T.
[16] EFSA NDA Panel (EFSA Panel on Dietetic Products,
Nutrition and Allergies), 2015. Scientific opinion on
the safety of 2’-O-fucosyllactose as a novel food
ingredient pursuant to Regulation (EC) No 258/97.
EFSA J. 2015, 13, 4184.
Wu, C.-C. Lin, Adv. Synth. Catal. 2012, 354, 123–132.
[36] S.-P. Li, W.-C. Hsiao, C.-C. Yu, W.-T. Chien, H.-J.
Lin, L.-D. Huang, C.-H. Lin, W.-L. Wu, S.-H. Wu, C.-
C. Lin, Adv. Synth. Catal. 2014, 356, 3199–3213.
[17] K. Yamashita, Y. Tachibana, A. Kobata, J. Biol.
Chem. 1977, 252, 5408–5411.
[37] T.-I. Tsai, H.-Y. Lee, S.-H. Chang, C.-H. Wang, Y.-C.
Tu, Y.-C. Lin, D.-R. Hwang, C.-Y. Wu, C.-H. Wong, J.
Am. Chem. Soc. 2013, 135, 14831–14839.
[18] S.-i. Hakomori, Biochim. Biophys. Acta 1999, 1473,
247–266.
[38] W. Peng, J. Pranskevich, C. Nycholat, M. Gilbert, W.
Wakarchuk, J. C. Paulson, N. Razi, Glycobiology 2012,
22, 1453–1464.
[19] A. Pfenninger, M. Karas, B. Finke, B. Stahl, J. Am.
Soc. Mass Spectr. 2002, 13, 1341–1348.
[39] G. Wang, D. A. Rasko, R. Sherburne, D. E. Taylor,
[20] M. R. Stroud, S. B. Levery, E. D. Nudelman, M. E.
Salyan, J. A. Towell, C. E. Roberts, M. Watanabe, S.
Hakomori, J. Biol. Chem. 1991, 266, 8439–8446.
Mol. Microbiol. 1999, 31, 1265–1274.
[40] S. Drouillard, H. Driguez, E. Samain, Angew. Chem.
2006, 118, 1810–1812; Angew. Chem. Int. Ed. 2006, 45,
1778–1780.
[21] M. R. Stroud, S. B. Levery, M. E. K. Salyan, C. E.
Roberts, S. Hakomori, Eur. J. Biochem. 1992, 203,
577–586.
[41] D. B. Stein, Y.-N. Lin, C.-H. Lin, Adv. Synth. Catal.
2008, 350, 2313–2321.
[22] Y.-Y. Fan, S.-Y. Yu, H. Ito, A. Kameyama, T. Sato,
C.-H. Lin, L.-C. Yu, H. Narimatsu, K.-H. Khoo, J. Biol.
Chem. 2008, 283, 16455–16468.
[42] M. Li, X.-W. Liu, J. Shao, J. Shen, Q. Jia, W. Yi, J. K.
Song, R. Woodward, C. S. Chow, P. G. Wang,
Biochemistry 2008, 47, 378–387.
[23] A. Miermont, Y. Zeng, Y. Jing, X.-s. Ye, X. Huang, J.
Org. Chem. 2007, 72, 8958–8961.
[43] W. Wang, T. Hu, P. A. Frantom, T. Zheng, B. Gerwe,
D. S. del Amo, S. Garret, R. D. Seidel, P. Wu, Proc.
Natl. Acad. Sci. USA 2009, 106, 16096–16101.
[24] K. R. Love, P. H. Seeberger, Angew. Chem. 2004, 116,
612–615; Angew. Chem. Int. Ed. 2004, 43, 602–605.
[44] J. Ye, X.-w. Liu, P. Peng, W. Yi, X. Chen, F. Wang,
[25] S. Nilsson, H. Lönn, T. Norberg, J. Carbohydr. Chem.
1991, 10, 1023–1048.
H. Cao, ACS Catal. 2016, 6, 8140–8144.
[45] Z. Xiao, Y. Guo, Y. Liu, L. Li, Q. Zhang, L. Wen, X.
Wang, S. M. Kondengaden, Z. Wu, J. Zhou, X. Cao, X.
Li, C. Ma, P. G. Wang, J. Org. Chem. 2016, 81, 5851–
5865.
[26] A. R. Prudden, L. Liu, C. J. Capicciotti, M. A. Wolfert,
S. Wang, Z. Gao, L. Meng, K. W. Moremen, G.-J.
Boons, Proc. Natl. Acad. Sci. USA 2017, 114, 6954–
6959.
[46] B. Petschacher, B. Nidetzky, J. Biotechnol. 2016, 235,
61–83.
[27] C. Dumon, B. Priem, S. L. Martin, A. Heyraud, C.
Bosso, E. Samain, Glycoconj. J. 2001, 18, 465–474.
[28] C.-C. Yu, S. G. Withers, Adv. Synth. Catal. 2015, 357,
1633–1654.
6
This article is protected by copyright. All rights reserved.

10.1002/adsc.201800518
Advanced Synthesis & Catalysis
COMMUNICATION
Enzymatic Synthesis of Human Milk Fucosides
α1,2-Fucosyl para-Lacto-N-Hexaose and its
Isomeric Derivatives
Adv. Synth. Catal. Year, Volume, Page – Page
Jia-Lin Fang, Teng-Wei Tsai, Chin-Yu Liang,
Jyun-Yi Li and Ching-Ching Yu*
7
This article is protected by copyright. All rights reserved.