α-Thioglycoligase-based synthesis of O-aryl α-glycosides as chromogenic substrates for α-glycosidases

Article abstract of DOI:10.1016/j.molcatb.2012.10.008

α-Thioglycoligases are retaining α-glycosidase mutants, with modification of their general acid/base catalytic residue to an inactive amino acid residue, catalyzing the formation of S-glycosidic linkages using a sugar donor with an excellent leaving group and suitable sugar acceptors with a thiol group as the substrate. In this study, we describe the enzymatic synthesis of O-aryl α-glycosides catalyzed by α-thioglycoligases. An α-xylosidase mutant (YicI-D482A) efficiently catalyzed the synthesis of O-aryl α-xylosides in near-quantitative yields (up to 99%) using 4-methylumbelliferone and nitrophenols. Synthesis did not occur with those acceptors having a nitro group at the ortho-position. The conversion yields of 3-nitrophenol markedly increased at pH 8.0, whereas those of other aryl compounds were nearly independent of pH, ranging from pH 6.0 to 8.0. The O-aryl α-xylosides were prepared on a preparative scale with yields of up to 96%. Upon employing the O-aryl α-xylosides as the substrate for the wild-type YicI, Br?nsted relationships of log k_cat versus pK_a and log (k_cat/K_M) versus pK_a both showed a linear monotonic dependence on the leaving group pK_a with low β_lg values of 0.39 and 0.38, respectively. In addition, synthesis of O-aryl α-glucosides was successfully conducted by an α-glucosidase mutant (MalA-D416A) in the same fashion with high yields. Therefore, this strategy can be used for the synthesis of O-aryl α-glycosides using an acid/base mutant of retaining α-glycosidases that hydrolyze the glycosides.

Full text of DOI:10.1016/j.molcatb.2012.10.008

Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
␣-Thioglycoligase-based synthesis of O-aryl ␣-glycosides as chromogenic
substrates for ␣-glycosidases
Chao Lia, Jin-Hyo Kimb, Young-Wan Kima,∗
a Department of Food and Biotechnology, Korea University, Sejong 339-700, Republic of Korea
b Chemical Safety Division, National Academy of Agricultural Sciences, Rural Development Administration, Suwon 441-701, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 August 2012
Received in revised form 16 October 2012
Accepted 16 October 2012
Available online 26 October 2012
␣-Thioglycoligases are retaining ␣-glycosidase mutants, with modification of their general acid/base
catalytic residue to an inactive amino acid residue, catalyzing the formation of S-glycosidic linkages
using a sugar donor with an excellent leaving group and suitable sugar acceptors with a thiol group
as the substrate. In this study, we describe the enzymatic synthesis of O-aryl ␣-glycosides catalyzed
by ␣-thioglycoligases. An ␣-xylosidase mutant (YicI-D482A) efficiently catalyzed the synthesis of O-
aryl ␣-xylosides in near-quantitative yields (up to 99%) using 4-methylumbelliferone and nitrophenols.
Synthesis did not occur with those acceptors having a nitro group at the ortho-position. The conversion
yields of 3-nitrophenol markedly increased at pH 8.0, whereas those of other aryl compounds were nearly
independent of pH, ranging from pH 6.0 to 8.0. The O-aryl ␣-xylosides were prepared on a preparative
scale with yields of up to 96%. Upon employing the O-aryl ␣-xylosides as the substrate for the wild-
type YicI, Brønsted relationships of log kcat versus pKa and log (kcat/KM) versus pKa both showed a linear
monotonic dependence on the leaving group pKa with low ˇlg values of 0.39 and 0.38, respectively. In
addition, synthesis of O-aryl ␣-glucosides was successfully conducted by an ␣-glucosidase mutant (MalA-
D416A) in the same fashion with high yields. Therefore, this strategy can be used for the synthesis of O-aryl
␣-glycosides using an acid/base mutant of retaining ␣-glycosidases that hydrolyze the glycosides.
© 2012 Elsevier B.V. All rights reserved.
Keywords:
O-Aryl ␣-glycoside
Chemoenzymatic synthesis
Retaining ␣-glycosidase
Thioglycoligase
Transglycosylation
1. Introduction
syntheses of 1,2-cis-glycosides, such as aryl ␣-glucosides and aryl
␣-galactosides, have been developed, but these reactions require
A wide variety of natural products, such as antibiotics and
functional glycosides, consist of phenols and sugars that are
connected through a glycosidic bond [1]. In addition, synthetic
O-aryl glycosides with chromogenic or fluorogenic aglycons are
invaluable substrates for the study of enzyme kinetics [2–4].
Therefore, the chemical synthesis of O-aryl glycosides is an inter-
esting topic in carbohydrate chemistry. However, in many cases,
these chemical reactions lead to a mixture of two stereoisomers
that differ in the configuration of the anomeric center. Of the
stereoisomers, 1,2-trans-glycosides (for example: ␤-glucosides, ␤-
xylosides, and ␣-mannoside) were obtained almost exclusively
through glycosylation procedures based on the neighboring group
participation by a 2-O-acyl functionality [5]. In contrast, the
introduction of 1,2-cis-glycosidic linkages using activated glycosyl
donors with a nonassisting functionality at C-2 leads to the for-
mation of an anomer mixture [6]. Recently, several stereoselective
laborious protection/deprotection processes and sometimes costly
reaction conditions, including low temperatures (<0 ◦C) or reflux for
several hours [7–9]. In the case of aryl ␣-xylosides, the general syn-
thesis pathway has not yet been reported, and only 4-nitrophenyl
␣-d-xylopyranoside (4-NP␣Xyl) is commercially available.
On the other hand, enzymatic synthesis has emerged as a
potent alternative to conventional chemical methods due to
its high stereoselectivity and generally mild reaction conditions
[10–13]. Although glycosyltransferases are necessary for the nat-
urally occurring production of oligosaccharides and glycosides,
glycosidases have attracted much attention in oligosaccharide
synthesis due to the utilization of cheaper substrates than glyco-
syltransferases, which use expensive activated nucleotide sugars
as substrates. Thus far, various glycosidases have been applied
to the synthesis of oligosaccharides, glycosides and glycoconju-
gates [14–16], and several strategies for glycosidase engineering
have addressed the problem of rehydrolysis of the glycosylated
products by wild-type glycosidases [12,17–19]. The first success-
ful engineered glycosidases termed glycosynthases are retaining
glycosidase mutants in which the catalytic nucleophile has been
converted to a non-nucleophilic residue [12,20]. These mutants
catalyze the formation of a new O-glycosidic bond with the same
∗
Corresponding author at: Department of Food and Biotechnology, Korea Univer-
sity, 2511 Sejong-Ro, Sejong 339-700, Republic of Korea. Tel.: +82 44 860 1436;
fax: +82 44 860 1586.
E-mail address: ywankim@korea.ac.kr (Y.-W. Kim).
1381-1177/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2012.10.008

C. Li et al. / Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
25
Fig. 1. Mechanism of transglycosylation catalyzed by a ␣-glycosynthase derived from a ␣-glucosidase (A [20]), YicI-D482A as a ␣-thioglycoligase (B [18]), and YicI-D482A
for synthesis of O-aryl ␣-xylosides (C).
stereochemistry as the original substrates when glycosyl fluorides
with the anomeric configuration opposite to that of the natural
substrate, thereby mimicking the glycosyl enzyme intermediate,
are employed as donor substrates (Fig. 1A). In contrast, thiogly-
coligases catalyze the formation of the S-glycosidic bond (Fig. 1B)
[18,21]. The glycosidase mutants whose general acid/base catalyst
is mutated to an inactive amino acid residue are capable of form-
ing glycosyl enzyme intermediates using substrates with a good
leaving group, such as dinitrophenol and fluoride. Upon employing
suitable sugar acceptors with a more nucleophilic thiol group than
the hydroxyl group, the intermediate turnover to free enzyme by
nucleophilic attack of the thio-sugar acceptor results in the forma-
tion of S-glycosidic linkages.
Given the transglycosylation mechanism of thioglycoligases,
aryl compounds possessing a hydroxyl group with a pKa value
lower than normal sugars would be used as the sugar accep-
tor, resulting in the formation of O-aryl glycosides instead of the
thio-glycosides. Recently, two ␣-thioglycoligases (YicI-D482A and
MalA-D416A) derived from a ␣-xylosidase from Escherichia coli
(YicI) and a ␣-glucosdiase from Sulfolobus solfataricus (MalA),
respectively, have been reported [18], raising the question of
whether such mutants will function as described for the synthesis
of O-aryl ␣-glycosides. This report describes the facile enzymatic
synthesis of O-aryl ␣-glycosides using ␣-thioglycoligases (Fig. 1C).
Two ␣-thioglycoligases (YicI-D482A and MalA-D416A) synthesized
O-aryl ␣-glycosides with excellent yields, and kinetic analysis was
conducted using O-aryl ␣-xylosides as a substrate for the wild-type
enzyme.
(4-NP), 4-methylumbelliferone (MU), 4-NP␣Xyl, 4-nitrophenyl
␣-d-glucopyranoside (4-NP␣Glc), 4-methylumbelliferyl ␣-d-
glucopyranoside (MU␣Glc), an ␣-glucosidase from Saccharomyces
cerevisiae, and silica gel (200–425 mesh) for flash chromatography
were purchased from Sigma–Aldrich Co., Ltd. High-performance
liquid chromatography (HPLC) was carried out using a YL9100
HPLC system (Younglin, Anyang, Korea) equipped with a Nova-
Pak® C18 column (3.9 mm × 150 mm, Waters). The products were
eluted at 0.8 mL/min with acetonitrile–water (20:80, v/v) and
analyzed using an ultraviolet (UV) detector at 310 nm. 1H nuclear
magnetic resonance (NMR) spectra were recorded on a 300 MHz
spectrometer from Varian Technologies Ltd. Mass spectra for aryl
xylosides were recorded using an API 3200 triple quadrupole mass
spectrometer (AB Sciex, Singapore). Thin-layer chromatography
(TLC) was performed on aluminum-backed sheets of silica gel
60F254 (Whatmann, USA) of 0.2-mm thickness using 17.5:2:0.5
(v/v/v) ethyl acetate/methanol/water. The plates were visualized
using UV light (254 nm) and/or by exposure to 10% sulfuric acid in
methanol followed by charring. Wild-type YicI and two thioglycol-
igases (YicI-D482A and MalA-D416A) were obtained as described
previously [18,23]. The protein concentration was determined by
a Bradford method using serum albumin as a standard [24].
2.2. Transglycosylation of aryl compounds using mutant
glycosidases
The transglycosylation reaction with 1 mM ␣XylF and an aryl
compound (1 mM, 2 mM and 3 mM) as the donor and acceptor,
respectively, and YicI-D482A (final concentration 5 mg/mL) was
conducted in 200 mM sodium phosphate buffer at pH 6.0, 7.0 and
8.0. The reactions were carried out at 25 ◦C for 4 h. After monitoring
by TLC, the reactions were terminated by adding a 19-fold volume
of acetonitrile (v/v) followed by centrifugation of the mixtures at
10,000 rpm for 15 min to remove precipitated enzymes and buffer
salts. To calculate the reaction yield, the samples were subjected
to HPLC analysis. Standard curves for the aryl glycosides were
determined by HPLC using the corresponding aryl glycosides
2. Materials and methods
2.1. Materials and general analysis
␣-d-Xylosyl fluoride (␣XylF) and ␣-d-glucosyl fluoride (␣GlcF)
were synthesized as described previously [22]. 2-Nitrophenol
(2-NP), 2,4-dinitrophenol (2,4-DNP), 2,5-dinitrophenol (2,5-DNP),
3-nitrophenol (3-NP), 3,4-dinitrophenol (3,4-DNP), 4-nitrophenol

26
C. Li et al. / Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
purified from the enzymatic reactions on a preparative scale as the
standards. The reaction conditions for MalA-D416A (final concen-
tration 5 mg/mL) were conducted in 200 mM sodium phosphate
buffer (pH 8.0) containing ␣GlcF (1 mM) and an aryl compound
(2 mM) as the donor and acceptor, respectively. In order to confirm
the anomeric configuration of the aryl glucosides, a ␣-glucosidase
from S. cerevisiae (Sigma) was added in the reaction mixture (final
2 U/mL). After 2 h incubation at 37 ◦C, the reaction mixtures were
analyzed by TLC.
3.80–3.60 (m, 5H, the rest protons of sugar ring). ESI-MS: calc. for
C11H12N2O9 + Na+ = 339.2; found 339.2.
2.4.4. 4-Methylumbelliferyl-˛-d-xylopyranoside (MU˛Xyl)
Yellow powder (38.2 mg, 96%). 1H NMR (CD3OD, 300 MHz): 7.73
(m, 1H, Ph-H), 7.15 (m, 1 H, Ph-H), 7.11 (m, 1H, Ph-H), 5.60 (d, 1H,
J = 3.9 Hz, H-1), 3.70–3.60 (m, 5H, the rest protons of sugar ring).
ESI-MS: calc. for C15H16O7 + Na+ = 331.2; found 331.2.
3. Results and discussion
2.3. Kinetic analysis of hydrolysis catalyzed by wild-type YicI
3.1. Transglycosylation of 4-nitrophenol by YicI-D482A
Kinetic analysis was performed by measuring the enzymatic
activity using the aryl xylosides as the substrates at 30 ◦C in
50 mM phosphate buffer (pH 7.0). Fifty microliters of wild-type
YicI, preincubated at 30 ◦C, were added to 50-␮L buffer solution
containing the aryl xylosides ranging from 0.2 to 2.5 mM, which
were also preincubated at the corresponding temperature. The
release of phenolic compounds was monitored at 400 nm using a
plate reader (SPECTRAMax, Molecular Devices Corporation, USA).
Standard curves for the aryl compounds were determined by mea-
suring the absorbance of the corresponding free aryl compounds
at 400 nm at pH 7.0 and 30 ◦C. In the case of MU␣Xyl, the amount
of released MU was measured using a spectrofluorimeter (VICTOR3
1420 Multilabel plate readers; Perkin-Elmer, USA) with excitation
and emission wavelengths of 365 and 450 nm, respectively. The
standard curve for MU was determined by measuring the fluores-
cence signal using the spectrofluorimeter at pH 7.0 and 30 ◦C using
MU as the standard. Kinetic parameters were obtained by a direct fit
of the data to the Michaelis–Menten equation using GraFit version
7.0 software (Leatherbarrow, R.J., Erithacus Software Ltd., Staines,
UK).
First, the transglycosylation activity of wild-type YicI and YicI-
D482A were compared using ␣XylF and 4-NP as the donor and
acceptor, respectively. ␣XylF rapidly disappeared in the reaction
catalyzed by YicI, and the transglycosylation product was rarely
detected by TLC analysis. In contrast, the reaction catalyzed by YicI-
D482A yielded one transglycosylation product (4-NP␣Xyl), which
was detected by TLC analysis. Given the previous kinetic analysis
of YicI-D482A, the YicI mutant should be capable of hydrolyzing
␣-XylF with the catalytic efficiency (kcat/KM) with three orders of
magnitude lower than the wild-type enzyme, whereas 4-NP␣Xyl
was an inert substrate for YicI-D482A [19]. YicI-D482A formed a
covalent xylosyl-enzyme intermediate when ␣XylF was used as the
substrate. This was followed by a transfer of the xylose moiety to
the hydroxyl group of 4-NP, and the transfer product then accumu-
lated in the reaction mixture due to an absence of hydrolysis of the
expected product. Due to the spontaneous and enzymatic hydrol-
ysis of ␣XylF, use of an equimolar ratio of ␣XylF and 4-NP was
not sufficient to achieve the maximum yield of the transfer prod-
uct. To improve the transglycosylation yield, the amount of 4-NP
was increased to overcome the competition with water molecules.
Use of a 1:2 ratio of ␣XylF to 4-NP resulted in improved produc-
tion of the product, whereas the addition of more 4-NP (molar
ratio of ␣XylF to 4-NP up to 1:3) decreased the quantity of product
(Fig. 2). Previously, it was reported that YicI-D482A possessed not
only thioglycoligase activity but also highly regiospecific transg-
lycosylation activity, termed O-glycoligase activity, to transfer the
xylose moiety from ␣XylF to the 6-OH group of both glucoside and
mannoside but not to that of galactoside and xyloside lacking the 6-
OH group [19]. Given the sugar-acceptor specificity of YicI-D482A
no second transfer product was observed by TLC or HPLC analysis.
The transfer product was purified using flash column chromatogra-
phy and subjected to detailed structural investigation by 1H NMR
analysis. The NMR data described in Section 2 revealed a chem-
ical shift from 9 to 7 ppm, which represents the phenolic group,
2.4. Production of aryl xylosides using YicI-D482A on a
preparative scale
A mixture of ␣XylF (0.13 mmole) and an aryl compound
(0.26 mmole) in phosphate buffer (5 mL of 0.2 M, pH 8.0) was
treated with YicI-D482A (5 mg/mL). The mixture was then
incubated at 25 ◦C for 4 h. To purify the aryl xylosides, the
reaction mixtures were subjected to a C18 SEP PAK cartridge
(Waters) to remove free sugars, enzyme and salts, and the
solvent was then evaporated under reduced pressure. Trans-
fer products were purified by silica-gel flash chromatography
(EtOAc/MeOH/H2O = 17.5:2:0.5). The isolated products were ana-
lyzed and identified by NMR and ESI-MS.
2.4.1. 4-Nitrophenyl-˛-d-xylopyranoside (4-NP˛Xyl)
White powder (32.7 mg, 93%). 1H NMR (CD3OD, 300 MHz): 8.25
(m, 2H, Ph-H), 7.30 (m, 2H, Ph-H), 5.63 (d, 1H, J = 4.5 Hz, H-1),
3.80–3.50 (m, 5H, the rest protons of sugar ring). ESI-MS: calc. for
C
11H13NO7 + Na+ = 294.2; found 294.2.
2.4.2. 3-Nitrophenyl-˛-d-xylopyranoside (3-NP˛Xyl)
White powder (31.5 mg, 89%). 1H NMR (CD3OD, 300 MHz): 7.98
(m, 1H, Ph-H), 7.90 (m, 1H, Ph-H), 7.58 (m, 1H, Ph-H), 7.53 (m,
1H, Ph-H), 5.58 (d, 1H, J = 3.9 Hz, H-1), 3.70–3.50 (m, 5H, the rest
protons of sugar ring). ESI-MS: calc. for C11H13NO7 + Na+ = 294.2;
found 294.2.
2.4.3. 3,4-Dinitrophenyl-˛-d-xylopyranoside (3,4-DNP˛Xyl)
Yellow powder (36.4 mg, 89%). 1H NMR (CD3OD, 300 MHz):
8.14 (d, 1H, J = 9.0 Hz, Ph-H), 7.66 (d, 1H, J = 2.4 Hz, Ph-H), 7.49
(dd, 1H, J = 2.7 Hz, J = 9.0 Hz, Ph-H), 5.70 (d, 1H, J = 3.9 Hz, H-1),
Fig. 2. Effect of the ␣XylF to 4-NP ratio on the 4-NP␣Xyl yield using YicI-D482A.

C. Li et al. / Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
27
Table 1
3.4. Kinetic analysis of aryl ˛-xylosides as wild-type YicI
substrates
Transglycosylation catalyzed by YicI-D482A and yields of the transfer products at
different pH.a
Aryl compound
pKa
Product
Conversion yield (%)b
The aryl glycosides bearing different leaving groups are chro-
mogenic substrates that can be used for detailed kinetic analysis
of retaining glycosides [2–4]. The kinetic parameters of YicI
were measured for the hydrolysis of the aryl ␣-xyloside sub-
strates. Hydrolysis rates over substrate concentrations were fit to
the Michaelis–Menten equation (Fig. S2, Supplementary Informa-
tion). From this, the kinetic parameters for these substrates were
determined (Table 2). Comparison of the KM values for the aryl
␣-xylosides revealed that there was no significant difference in
the apparent binding of these substrates (0.99–1.5 mM). The main
difference between these substrates was in their kcat values, with
3,4-DNP␣Xyl being ∼7-fold greater than that of 4-NP␣Xyl, which
was 11-fold greater than that of MU␣Xyl. A Brønsted plot repre-
senting the relationship of the rate of cleavage of a series of aryl
xylosides with the pKa of the corresponding phenol allows us to
understand the mechanism of glycosidases. The kinetic parameters
of the aryl xylosides were plotted in the form of Brønsted relation-
ships (Fig. 3). The log kcat versus pKa and log (kcat/KM) versus pKa
plots clearly revealed a dependence, albeit shallow (ˇlg = −0.39 and
−0.38, respectively) rate on the leaving group ability, suggesting
that the rate-limiting step involves breakage of the aryl glycoside
bond to form the covalent glycosyl enzyme intermediate.
pH 6
pH 7
pH 8
4-NP
3-NP
3,4-DNP
MU
7.15
8.39
5.36
7.79
4-NP␣Xyl
3-NP␣Xyl
3,4-DNP␣Xyl
MU␣Xyl
92
23
95
95
93
28
95
97
99
94
97
99
a
Reaction condition: 1 mM ␣XylF, 2 mM aryl compound, 5 mg/mL YicI-D482A in
200 mM phosphate buffer at 25 ◦C for 4 h.
b
The conversion yields were calculated using the amount of the product deter-
mined by HPLC analysis based on the amount of the employed ␣XylF.
while the xylose moiety showed a chemical shift at high-field from
6 to 3 ppm. The chemical shift of H-1 harbored by 4NP␣Xyl showed
typical double peaks around 5.63 ppm. It is worth noting that the
1H NMR spectra exhibited a small coupling constant for H-1 sig-
nal (J = 4.5), revealing that the aryl product is a 1,2-cis-xyloside, as
expected for an ␣-anomer.
3.2. Acceptor specificity of YicI-D482A for the synthesis of aryl
xylosides
To investigate sugar-acceptor specificity, various aryl com-
pounds were tested at pH 7.0. Given the results from the reaction
using 4-NP as the sugar acceptor, the ratio of ␣XylF to an aryl
compound was fixed to 1:2. TLC analysis revealed that 3,4-DNP, 4-
NP, and a structurally different acceptor, MU, were excellent sugar
acceptors with yields up to 97%, while 3-NP was a weaker accep-
tor with a poor yield (<30%, Table 1). Unexpectedly, three kinds
of ortho-substituted nitrophenols, such as 2-NP, 2,4-DNP and 2,5-
DNP, did not function as the sugar acceptor (data not shown). This
was likely caused by the steric hindrance between the ortho-nitro
group and the active site of YicI-D482A. Such acceptor specificity
can be properly explained by previous analysis of the complex
structure of YicI with a thio-linked substrate analog [18]. Given
the structure of the +1 subsite, YicI-D482A is not expected accom-
modate enough space for binding of the ortho-nitro group of the
unsuccessful nitrophenols in the +1 subsite. The ortho-nitro group
would clash with either Asp185 and Arg466 or Phe277 and Trp380
of YicI (Fig. S1, Supplementary Information).
In the case of retaining ␤-glycosidases, a biphasic relationship
is quite common in Brønsted plots, that is one of evidences on the
mechanism of retaining ␤-glycosidases via formation of the cova-
lent glycosyl enzyme intermediate [3]. Although any break in these
plots for YicI, one of retaining ␣-glycosidases was not observed,
such linear monotonic dependence on leaving group pKa with low
ˇlg values of ∼−0.4 is consistent with previous studies on ␣-1,4-
glucan lyase belonging to the same glycoside hydrolase family as
YicI [2].
3.3. pH effect on aryl xyloside synthesis and preparative
production
To determine the optimum pH conditions, reactions using the
four aryl compounds as acceptors were conducted at pH 6.0, 7.0
and 8.0. HPLC analysis revealed that the yields of transfer products
from the reaction using 3,4-DNP, 4-NP and MU slightly increased
with pH. Interestingly, the yield of 3-NP␣Xyl markedly increased at
pH 8.0 (94%, Table 1). Of the aryl compounds that yielded transfer
products, 3-NP had the highest pKa value (8.39), suggesting that the
improved yield for 3-NP␣Xyl at pH 8.0 was due to enhancement of
the nucleophilicity of the hydroxyl group of 3-NP at a basic pH.
Based on the optimization results, a 5-mL scale reaction was
conducted with 26 mM ␣XylF and 52 mM each of the aryl com-
pounds (3-NP, 4-NP, 3,4-DNP, and MU) as the substrates at pH
8.0. The reactions were completed within 4 h. The transfer product
was purified using flash column chromatography and subjected to
detailed structural investigation by 1H NMR analysis. As described
in Section 2, the 1H NMR spectra of all transfer products showed a
typical double peak of H-1 with the chemical shift around 5.6 ppm
and a small coupling constant (J = 3.9 or 4.5), indicating that these
aryl xylosides were ␣-anomers.
Fig. 3. Brønsted plot constructed from the data in Table 2 showing the relation-
ship between the rate of cleavage of a series of aryl xylosides with the pKa of the
corresponding phenol. (a) Log kcat vs. pKa and (b) log (kcat/KM) vs pKa.

28
C. Li et al. / Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
Table 2
Kinetic parameters for hydrolysis of aryl ␣-xylosides by YicI.a
b
pKa
kcat (s−1
)
KM (mM)
kcat/KM (s−1 mM−1
)
Ref.
␣XylF
4-NP␣Xyl
3-NP␣Xyl
3,4-DNP␣Xyl
MU␣Xyl
–
75
3
0.97 0.13
1.08 0.04
0.99 0.06
1.20 0.14
1.51 0.09
77
[23]
7.15
8.39
5.36
7.79
0.16 0.006
0.084 0.003
1.1 0.034
0.15
0.085
0.92
0.063
This study
This study
This study
This study
0.095 0.006
a
Reaction condition: 50 mM phosphate buffer (pH 7.0) at 30 ◦C.
pKa value of the aglycon of the corresponding xyloside.
b
Fig. 4. TLC analysis of MalA-D416A-catalyzed reactions. The TLC plate was visualized by UV at 254 nm (left) and by sulfuric acid followed by charring (right). Transfer products
are indicated by dotted circles. Lane S: standard including ␣GlcF and glucose; lane 1: blank reaction without MalA-D416A; lane 2: reaction mixture with MalA-D416A; lane
3: hydrolysis reaction mixture with ␣-glucosidase.
3.5. Synthesis of O-aryl ˛-glucosides using MalA-D416A
4. Conclusions
The successful synthesis of O-aryl ␣-xylosides using YicI-D482A
raised the question as to whether such a strategy can be expanded
to the synthesis of other O-aryl ␣-glycosides. Therefore, trans-
glycosylation reactions catalyzed by another ␣-thioglycoligase,
MalA-D416A, derived from an ␣-glucosidase [18], were carried out
with ␣GlcF and aryl compounds as the sugar donors and acceptors,
respectively, in 0.2 M phosphate buffer (pH 8.0). As shown in Fig. 4,
MalA-D416A also catalyzed the synthesis of aryl ␣-glucosides. The
aryl compounds that were transformed by MalA-D416A were 3,4-
DNP, 3-NP, 4-NP, and MU as sugar acceptors (Fig. 4, lane 2 of
each aryl compound), whereas the ortho-substituted nitrophenols,
including 2-NP, 2,4-DNP and 2,5-DNP did not function as the sugar
acceptors (data not shown). These results were consistent with
those of YicI-D482A. From this, YicI and MalA, which belong to
the same glycoside hydrolase family 31, share the same active
site architecture [25] that would lead to such acceptor specificity.
Given the mechanism of retaining ␣-glucosidase and the results
of YicI-D482A experiments, the synthesized aryl glucosides were
expected to be ␣-anomers. Although NMR analysis can be used
to determine the anomeric configuration of these aryl glucosides,
the direct evidence can arise from the result of the hydrolysis of
the products by wild-type ␣-glucosidase. Upon employment of ␣-
glucosidase (2 U/mL) in the reaction mixture, TLC results showed
that the spots of each product disappeared completely, while those
corresponding to glucose were detected (Fig. 4, lane 3 for each aryl
compounds). This suggests that the anomeric configuration of all
aryl glucosides synthesized by MalA-D416A was that of ␣-anomer.
Using standard curves of the glucosides purchased from Sigma, the
yields of the 4-NP␣Glc and MU␣Glc, as determined by HPLC, were
83% and 75%, respectively, based on the donor sugar employed.
We demonstrate here for the first time the facile enzymatic
synthesis of O-aryl ␣-glycosides, which was developed using
␣-thioglycoligases derived from two retaining ␣-glycosidases.
These ␣-glycosidase mutants convert non-ortho-substituted nitro-
phenols and MU to the corresponding aryl sugars with near
quantitative yields. Glycosylation of the aryl compounds catalyzed
by the engineered glycosidases was dependent of the nucleophilic-
ity of the hydroxyl group to which the glycosyl moiety would
be transferred. Such thioglycoligase-based aryl glycoside synthe-
sis strategies can be used to synthesize O-aryl glycosides using an
acid/base mutant of retaining glycosides that hydrolyze the glyco-
sides.
Acknowledgements
This work was supported by in part National Research Founda-
tion of Korea Grant (2011-0012467) and in part the 21st Century
Frontier R&D Program in Microbial Genomics & Applications (11-
2008-18-004-00) funded by the Ministry of Education, Science and
Technology. We thank Professor Cheol Min Yoon, Korea University
for NMR analysis.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.
molcatb.2012.10.008.

C. Li et al. / Journal of Molecular Catalysis B: Enzymatic 87 (2013) 24–29
29
References
[14] H.S. Lee, J.H. Auh, H.G. Yoon, M.J. Kim, J.H. Park, S.S. Hong, M.H. Kang, T.J. Kim,
T.W. Moon, J.W. Kim, K.H. Park, J. Agric. Food Chem. 50 (2002) 2812–2817.
[15] C.H. Choi, S.H. Kim, J.H. Jang, J.T. Park, J.H. Shim, Y.W. Kim, K.H. Park, J. Sci. Food
Agric. 9 (2010) 1179–1184.
[16] L.X. Wang, W. Huang, Curr. Opin. Chem. Biol. 13 (2009) 592–600.
[17] G. Osanjo, M. Dion, J. Drone, C. Solleux, V. Tran, C. Rabiller, C. Tellier, Biochem-
istry 46 (2007) 1022–1033.
[18] Y.W. Kim, A.L. Lovering, H. Chen, T. Kantner, L.P. McIntosh, N.C. Strynadka, S.G.
Withers, J. Am. Chem. Soc. 128 (2006) 2202–2203.
[19] Y.W. Kim, R. Zhang, H. Chen, S.G. Withers, Chem. Commun. 46 (2010)
8725–8727.
[20] M. Okuyama, H. Mori, K. Watanabe, A. Kimura, S. Chiba, Biosci. Biotechnol.
Biochem. 66 (2002) 928–933.
[21] M. Jahn, J. Marles, R.A.J. Warren, S.G. Withers, Angew. Chem. Int. Ed. 42 (2003)
352–354.
[22] M. Hayashi, S. Hashimoto, R. Noyori, Chem. Lett. 10 (1984) 1747–1750.
[23] A.L. Lovering, S.S. Lee, Y.W. Kim, S.G. Withers, N.C.J. Strynadka, J. Biol. Chem.
280 (2005) 2102–2125.
[24] M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[25] H.A. Ernst, L. Lo Leggio, M. Willemoës, G. Leonard, P. Blum, S. Larsen, J. Mol. Biol.
358 (2006) 1106–1124.
[1] S.H. Kim, C. Naveen Kumar, H.J. Kim, D.H. Kim, J. Cho, C. Jin, Y.S. Lee, Bioorg.
Med. Chem. Lett. 19 (2009) 6009–6013.
[2] S.S. Lee, S. Yu, S.G. Withers, Biochemistry 42 (2003) 13081–13090.
[3] T. Bravman, G. Zolotnitsky, V. Belakhov, G. Shoham, B. Henrissat, T. Baasov, Y.
Shoham, Biochemistry 42 (2003) 10528–10536.
[4] F.S. Shidmoossavee, L. Cheng, J.N. Watson, A.J. Bennet, Biochemistry 49 (2010)
6473–6484.
[5] G.J. Boons, Contemp. Org. Synth. 3 (1996) 173–200.
[6] A.V. Demchenko, Synlett (2003) 1225–1240.
[7] S.S. Chang, C.C. Lin, Y.K. Li, K.K. Mong, Carbohydr. Res. 344 (2009)
432–438.
[8] J. Kadokawa, T. Nagaoka, J. Ebana, H. Tagaya, K. Chiba, Carbohydr. Res. 327
(2000) 341–344.
[9] H.M. Chen, S.G. Withers, ChemBioChem 8 (2007) 719–722.
[10] A. Trincone, Recent Pat. Biotechnol. 4 (2010) 30–47.
[11] A. Trincone, A. Giordano, Curr. Org. Chem. 10 (2006) 1163–1193.
[12] S.M. Hancock, M.D. Vaughan, S.G. Withers, Curr. Opin. Chem. Biol. 10 (2006)
509–519.
[13] P. Bojarová, V. Kren, Trends Biotechnol. 27 (2009) 199–209.