Acceptor-dependent regioselectivity of glycosynthase reactions by Streptomyces E383A β-glucosidase

Article abstract of DOI:10.1016/j.carres.2006.04.049

The nonnucleophilic mutant E383A β-glucosidase from Streptomyces sp. has proven to be an efficient glycosynthase enzyme, catalyzing the condensation of α-glucosyl and α-galactosyl fluoride donors to a variety of acceptors. The enzyme has maximal activity

Full text of DOI:10.1016/j.carres.2006.04.049

Carbohydrate Research 341 (2006) 2055–2065
Acceptor-dependent regioselectivity of glycosynthase reactions by
Streptomyces E383A b-glucosidase
*
´
Magda Faijes, Marc Saura-Valls, Xavi Perez, Marta Conti and Antoni Planas
`
Laboratory of Biochemistry, Institut Quı´mic de Sarria, Universitat Ramon Llull, Via Augusta 390, E-08017 Barcelona, Spain
Received 3 February 2006; received in revised form 27 April 2006; accepted 30 April 2006
Available online 23 May 2006
Abstract—The nonnucleophilic mutant E383A b-glucosidase from Streptomyces sp. has proven to be an efficient glycosynthase
enzyme, catalyzing the condensation of a-glucosyl and a-galactosyl fluoride donors to a variety of acceptors. The enzyme has max-
imal activity at 45 ꢀC, and a pH-dependence reflecting general base catalysis with an apparent kinetic pKa of 7.2. The regioselectivity
of the new glycosidic linkage depends unexpectedly on the acceptor substrate. With aryl monosaccharide acceptors, b-(1!3) disac-
charides are obtained in good to excellent yields, thus expanding the synthetic products available with current exo-glycosynthases.
With xylopyranosyl acceptor, regioselectivity is poorer and results in the formation of a mixture of b-(1!3) and b-(1!4) linkages.
In contrast, disaccharide acceptors produce exclusively b-(1!4) linkages. Therefore, the presence of a glycosyl unit in subsite +II
redirects regioselectivity from b-(1!3) to b-(1!4). To improve operational performance, the E383A mutant was immobilized on a
Ni2+-chelating Sepharose resin. Immobilization did not increase stability to pH and organic solvents, but the operational stability
and storage stability were clearly enhanced for recycling and scaling-up.
ꢁ 2006 Elsevier Ltd. All rights reserved.
Keywords: Glycosynthase; Oligosaccharide synthesis, enzymatic; Glycosylation; Immobilization; b-Glucosidase; Streptomyces sp; Enzyme specificity
1. Introduction
The glycosynthase methodology has been explored
for several (but yet limited number) retaining glycosi-
dases belonging to different families (as classified on
the bases of sequence similarities in the CAZY database
(http://afmb.cnrs-mrs.fr/CAZY/)).6 For endo-glycosi-
dases, their derived glycosynthases are highly regiospec-
ific, catalyzing the formation of a unique glycosidic
bond matching the specificity shown by the parental
wt-hydrolase enzyme.2,7–11 exo-Glycosidases have a
more relaxed specificity. Several Family 1 glycosidases
have shown to function as glycosynthases upon muta-
tion of the catalytic nucleophile, but their regioselectiv-
ity is substantially different.1,12–14 This approach will be
considered as a general synthetic methodology when an
extensive repertoire of glycosynthases covering different
regio and stereoselectivities will become available.
Engineering a retaining glycosidase into a glycosynthase
involves mutation of the catalytic nucleophile to a smal-
ler nonnucleophilic amino acid residue and the use of
activated glycosyl donors (i.e., glycosyl fluorides) with
the opposite anomeric configuration than that of the
normal substrate of the parental wild type enzyme
(Scheme 1).1,2 Because the mutant is hydrolytically inac-
tive, the condensation product is not hydrolyzed and
accumulates to achieve high yields. Therefore, glyco-
synthases share the advantages of glycosidases in trans-
glycosylation reactions (availability, stability, and use of
simple glycoside donors), but overcome their major
drawback (low synthetic yields due to product hydroly-
sis).3–5
b-Glucosidase from Streptomyces sp. (a Family 1 gly-
cosyl hydrolase) has a broad substrate specificity,15–17
and it is a new candidate to explore its abilities as glyco-
synthase. The wt enzyme is highly active on glucosyl and
*
Corresponding author. Tel.: +34 932672000; fax: +34 932056266;
e-mail: antoni.planas@iqs.es
0008-6215/$ - see front matter ꢁ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.04.049

2056
M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
O
O
O
O
H
OH
F
OH
OH
OH
O
HO
O
O
O
O
HO
O
HO
HO
HO
HO
H
O
R
R
O
OH
OH
OH
OH
F
CH3
CH3
nucleophile mutant
Scheme 1. Glycosynthase reaction catalyzed by a nucleophile mutant of retaining glycosidases.
fucosyl glycosides and, to a lesser extent, on galactosyl
and xylosyl derivatives. Besides, it hydrolyzes either
b-(1!2) (sophorose), b-(1!4) (cellobiose) or b-(1!3)
(laminaribiose) linked disaccharides with similar activ-
ity. Under kinetically controlled transglycosylation con-
ditions, these three glycosidic linkages are also formed.18
Since ꢀI and +I subsites accept a variety of glycosidic
units and linkages, the glycosynthase mutants could in
principle couple a wide combination of donors and
acceptors and broaden the spectrum of di- and trisac-
charides already obtained with currently existing
glycosynthases.
The temperature profile shows a maximal activity at
45 ꢀC. But at this temperature, spontaneous hydrolysis
of the fluoride donor measured by capillary electropho-
resis (data not shown) is significant. The ratio between
glycosynthase activity and spontaneous hydrolysis was
maximal at 35 ꢀC, so this temperature was chosen for
all coming experiments.
The plot vo versus pH (Fig. 1B) corresponds to gen-
eral base catalysis with residual glycosynthase activity
(Ares) at pH 5.5, and maximum (Bmax) activity above
1.0
On the basis of sequence alignment, site-directed
mutagenesis, and chemical rescue experiments, Glu383
was identified as the catalytic nucleophile of the b-gluco-
sidase from Streptomyces sp.16 In this paper, we explore
the properties of the E383A mutant as a glycosynthase
to define its activity and specificity for a wide combina-
tion of donors and acceptors. Additionally, the enzyme
was immobilized to increase its stability and improve
operational conditions for recycling and scaling-up.
A
0.8
0.6
0.4
0.2
0
20
30
40
50
60
2. Results and discussion
T (ºC)
Glycosynthase activity of the E383A mutant b-glucosi-
dase from Streptomyces sp. was first assayed by reaction
between a-glucosyl fluoride (1) and p-nitrophenyl gluco-
side (4), which produced a disaccharide (m/z 486.1
[M+Na]+). It was chosen as the model reaction to set
optimal glycosynthase conditions and further explore
the specificity with different donor and acceptor
saccharides.
1.2
1.0
0.8
0.6
0.4
0.2
0
B
2.1. Enzyme kinetics: pH and temperature dependence
of the glycosynthase activity
5
6
7
8
9
10
Initial rates of product formation (vo) between donor 1
and acceptor 4 were determined by HPLC analysis.
The reactions were done at constant acceptor (5 mM)
and donor (0.9 mM) concentrations in phosphate buffer
either at pH 7 over the temperature range from 25 to
55 ꢀC, or at 35 ꢀC over the pH range from 5.5 to 10.
The results are presented in Figure 1.
pH
Figure 1. Temperature and pH dependencies of glycosynthase-cata-
lyzed condensation of donor 1 and acceptor 4 by E383A b-glucosidase
in 50 mM phosphate buffer and constant ionic strength (0.15 M): (A)
temperature profile, [enzyme] = 8.5 nM, [donor (1)] = 0.8 mM, [accep-
tor (4)] = 5.7 mM, pH 7.0; (B) pH profile, [enzyme] = 8.1–14.8 nM,
[donor (1)] = 0.9 mM, [acceptor (4)] = 5.0 mM, 35 ꢀC.

M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
2057
1.2
1.0
0.8
0.6
0.4
0.2
0
a-galactosyl and a-xylosyl fluorides (1–3), were assayed,
and a set of different monosaccharides (4–9) and disac-
charides (10–11) were evaluated as acceptors (Tables 1
and 2). Donor and acceptor at 1–5 M ratio were incu-
bated with the enzyme (0.01–35 lM, depending on sub-
strates reactivity) at pH 7.0 and 35 ꢀC. The excess of
acceptor minimizes elongation reactions where the gly-
cosyl donor may react with the condensation product
leading to oligomers. Two sets of reactions were carried
out: analytical scale reactions at low enzyme concentra-
tion to determine initial rates of product formation by
HPLC, and preparative reactions at higher enzyme con-
centration to evaluate product yields after longer reac-
tion times and to determine the regioselectivity of the
newly formed glycosidic bond by NMR spectroscopy.
As shown in Tables 1 and 2, a-glucosyl and galactosyl
fluorides are good donors, but not a-xylosyl fluoride,
where the absence of the 5-hydroxymethyl group has a
dramatic effect and no glycosynthase activity with any
acceptor was detected. The importance of this substitu-
ent in subsite ꢀI of the enzyme binding site has been ob-
served in the wt hydrolase activity of this and other
Family 1 b-glycosidases, with 15–19 kJ molꢀ1 transition
state destabilization on C-6-deoxygenated substrates as
compared to their oxygenated homologues.16,19–21 Like-
wise, a-xylosyl fluoride was not a donor for the glyco-
synthase derived from Agrobacterium b-glucosidase,1
but remarkably, directed evolution has recently over-
come this limitation to afford an efficient catalyst for
xylooligosaccharides synthesis.22
Relative reactivities were evaluated from initial rates
of condensation (Tables 1 and 2). The glucosyl donor
1 reacts five-times faster than the galactosyl donor 2
with the p-nitrophenyl glucoside acceptor, in agreement
with kinetics of the wild type hydrolase activity.15 For
the acceptors, the sequence of reactivities is Glu >
Gal ꢃ Xyl ꢃ Man ꢄ Fuc. The fucoside acceptor reacts
slowly with a-glucosyl fluoride (1), and no reaction
was detected with the galactosyl donor 2 under the same
conditions. p-Nitrophenyl cellobioside (10) is a better
acceptor than the laminaribioside 11, but both gave
lower transglycosylation initial rates than the glucoside
acceptor 4. This is probably the consequence of compe-
tition of the disaccharide acceptors for the donor sub-
site. However, direct comparison to monosaccharides
cannot be done because of the different regioselectivity
of the reactions, as discussed below.
Donor
Acceptor
0
1
2
3
4
5
6
Concentration (mM)
Figure 2. Glycosynthase kinetics of E383A b-glucosidase with donor 1
and acceptor 4 in 50 mM phosphate buffer, pH 7 and 35 ꢀC: (d)
kinetics for the acceptor at fixed 0.9 mM donor concentration, (s)
kinetics for the donor at fixed 4.9 mM acceptor concentration.
[Enzyme] = 4.8–7.6 nM.
pH > 8. Data were fitted to Eq. 1 to give an apparent
kinetic pKa value of 7.24.
v
½Eꢁ
Bmax
¼ Ares
þ
ð1Þ
1 þ 10ðpK ꢀpHÞ
a
This pKa reflects ionization of Glu178 as the general
base in the glycosynthase mechanism (Scheme 1). It
has been proved for another glycosynthase, the E134A
mutant (1!3)-(1!4)-b-glucanase from Bacillus licheni-
formis, that the same residue acting as the general
acid/base catalyst in the wild type hydrolase activity is
indeed the general base in the glycosynthase reaction.7
p-Nitrophenyl glucoside (4) showed significant sub-
strate inhibition of the wt hydrolase activity (KI
1.4 mM, KM 0.15 mM).16 To evaluate the potential
inhibitory effect on the glycosynthase reaction, kinetics
with varying acceptor concentration and fixed a-glucosyl
fluoride (1) donor (0.9 mM) at pH 7.0 and 35 ꢀC were
analyzed. The Michaelis–Menten curve (Fig. 2) shows
no inhibition up to 6 mM, with apparent kinetic para-
meters kapp ¼ 0:72 ꢂ 0:08 sꢀ1 and KaMpp ¼ 3:0 ꢂ 0:8 mM.
cat
Kinetic parameters for the donor were also determined
at constant (4.9 mM) acceptor concentration (Fig. 2):
app
cat
k
¼ 1:53 ꢂ 0:06 sꢀ1 and KaMpp ¼ 1:7 ꢂ 0:2 mM.
2.2. Donor and acceptor specificity
The wt enzyme has a broad substrate specificity, being
able to hydrolyze aryl b-glucosides, fucosides, galacto-
sides and xylosides, as well as b-(1!4), b-(1!3) and
b-(1!2) glycosidic linkages.15,16 The transglycosylation
activity of the wild type enzyme with p-nitrophenyl b-
D-galactoside as donor and D-xylose as acceptor yielded
b-(1!3) (40%), b-(1!2) (26%) and b-(1!4) (4%)
disaccharides.18
Preparative reactions were finalized when no increase
of product concentration was detected by HPLC. Reac-
tion mixtures were directly purified by reverse-phase
chromatography to isolate the condensation products.
Good yields were obtained with some donor–acceptor
combinations. With a-glucosyl fluoride donor yields
varied from 30% to 100% depending on the acceptor,
and with a-galactosyl fluoride they were slightly lower
(14–77%). All compounds were characterized by MS
To define donor and acceptor glycosynthase specific-
ity of the E383A b-glucosidase, three donors, a-glucosyl,

2058
M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
Table 1. Glycosynthase reactions of E383A b-glucosidase using a-glucosyl fluoride (1) donor and different acceptors
Acceptor
Disaccharide product (% reaction yield)a
v/[E] · 102 (sꢀ1 b
)
Reactivityc (%)
OH
O
HO
HO
4
5
b-(1!3) (96)
46.14
1.80
100
O
OH
NO2
OH
OH
HO
O
b-(1!3) (100)
3.9
0.6
O
OH
NO2
O
HO
HO
O
O
6
b-(1!3) (54)
b-(1!4) (28)
0.28
OH
NO2
NO2
OH
OH
O
HO
HO
7
8
b-(1!3) (53)
b-(1!3) (30)
0.45
0.02
1
OH
CH3
O
0.04
O
HO
OH
NO2
OH
O
HO
HO
9
Nr
—
—
21
1.5
O
NHAc
NO2
OH
OH
O
O
HO
O
HO
HO
10
11
b-(1!4) (80)
b-(1!4)
9.72
0.70
O
OH
OH
OH
NO2
OH
O
O
O
HO
HO
HO
O
OH
OH
NO2
a Preparative reactions in 50 mM phosphate buffer, pH 7.0, 35 ꢀC, at 1 to 5 donor:acceptor molar ratio, and 2–28 lM enzyme concentration. Yields
in isolated product after purification.
b Initial rates of product formation (by HPLC). Conditions: 50 mM phosphate buffer, pH 7.0, 35 ꢀC. [donor] = 0.9 mM, [acceptor] = 5 mM.
c Reactivity expressed as percentage of initial rates relative to acceptor 4. Nr: no reaction.
and NMR spectroscopy. Connectivities of the new gly-
cosidic linkages were assigned by comparison to pub-
lished 13C NMR spectra,23,24 by calculating the spectra
taking into account the effect of O-glycosidation on
b-(1!3) disaccharides in 50% and 60%, respectively.
Lower yields were obtained for p-nitrophenyl galactos-
yl-b-(1!3)-galactoside and glucosyl-b-(1!3)-fucoside
products (<30%).
The b-(1!3) regioselectivity of the E383A glucosyn-
thase was partially lost with the xyloside acceptor 6.
Condensation of glucosyl and galactosyl donors with
p-nitrophenyl xyloside yielded b-(1!3) and b-(1!4)
regioisomers. It suggests that the 5-hydroxymethyl
group of the saccharide acceptor in subsite +I directs
regioselectivity to b-(1!3) linkages, whereas its absence
results in a relaxed selectivity. A similar but inverse
behaviour was observed with the glycosynthase mutants
1
13C chemical shifts,25,26 or by two-dimensional H and
13C NMR experiments (see Section 3).
With monosaccharide acceptors, the E383A glyco-
synthase gives specifically b-(1!3) disaccharides (Tables
1 and 2). p-Nitrophenyl b-laminaribioside, glucosyl-
b-(1!3)-galactoside and galactosyl-b-(1!3)-glucoside
were obtained with nearly quantitative yields. Conden-
sation between glucosyl and galactosyl donors and p-
nitrophenyl mannoside acceptor gave the corresponding

M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
2059
Table 2. Glycosynthase reactions of E383A b-glucosidase using a-galactosyl fluoride (2) donor and different acceptors
Acceptor
Disaccharide product (% reaction yield)a
v/[E] · 102 (sꢀ1 b
)
Reactivityc (%)
OH
O
HO
HO
O
4
5
b-(1!3) (77)
9.44
100
OH
OH
NO2
OH
HO
O
b-(1!3) (14)
0.017
0.2
O
OH
NO2
O
HO
HO
O
O
6
b-(1!3) (25)
b-(1!4) (12)
0.034
0.3
4.0
OH
NO2
NO2
OH
OH
O
HO
HO
7
8
b-(1!3) (47)
0.38
—
OH
CH3
O
OH
OH
Nr
—
—
O
HO
NO2
O
HO
HO
9
Nr
—
O
NHAc
NO2
a Preparative reactions in 50 mM phosphate buffer, pH 7.0, 35 ꢀC, at 1 to 5 donor:acceptor molar ratio, and 2–28 lM enzyme concentration. Yields in
isolated product after purification.
b Initial rates of product formation (by HPLC). Conditions: 50 mM phosphate buffer, pH 7.0, 35 ꢀC. [donor] = 0.9 mM, [acceptor] = 5 mM.
c Reactivity expressed as percentage of initial rates relative to acceptor 4. Nr: no reaction.
of b-glucosidase from Agrobacterium.1,22 Aryl glycoside
acceptors were condensed via b-(1!4) linkages in all
cases, except for the b-xyloside acceptor, which gave
b-(1!3) products.
Therefore, the regioselectivity of glycosidic bond for-
mation is switched from b-(1!3) with monosaccharide
acceptors to b-(1!4) with disaccharides. It suggests that
enzyme–ligand interactions in subsite +II have a deci-
sive effect on the orientation of the acceptor. For 4-
nitrophenyl b-D-glucopyranoside, the aryl group occu-
pies subsite +II and the glycosyl unit in subsite +I pre-
sents the 3-OH group to be glycosylated (Scheme 2).
In contrast, when a glucosyl unit occupies subsite +II
(cellobiose and laminaribiose), the glucosyl residue in
subsite +I presumably binds in an alternative mode,
presenting the 4-OH group for glycosylation. Although
the orientation of the pyranose ring in subsite +II for
the cellobioside and laminaribioside acceptors differs,
the presence of this glucose residue instead of the aryl
group favours formation of a b-(1!4) linkage in both
cases.
Compared to the other exo-glycosynthases, the
E383A glycosynthase has a different specificity than
other Family 1 glycosynthases. b-Glucosidase from
Agrobacterium1 produces exclusively b-(1!4) linkages
with a variety of mono- and disaccharide acceptors, with
the exception of a xylosyl acceptor that gives a b-(1!3)
bond. Recently, it has been shown that the regioselectiv-
ity of this glycosynthase can be modified by changes in
In contrast to monosaccharide, condensation with
disaccharide acceptors resulted in the formation of b-
(1!4) glycosidic linkage. p-Nitrophenyl cellotrioside
was obtained in 80% yield by reaction between a-gluco-
syl fluoride (1) and the cellobioside acceptor 10 as
confirmed by NMR spectroscopy. In the same way,
transfer to the laminaribioside acceptor 11 yielded the
trisaccharide p-nitrophenyl glucosyl-b-(1!4)-laminari-
bioside, which was identified by coinjection with a syn-
thetic sample27 in HPLC. In both reactions, a second
minor compound appeared after six hours in 3% yield.
For the cellobioside acceptor 10, it was corrobora-
ted not to be the regioisomer of the cellotrio-
side Glcb3Glcb4Glc-pNP, by coinjection with an
independent standard.7 And for the laminaribioside
acceptor 11, this second product was identified as
Glcb4Glcb4Glcb3Glc-pNP. These minor compounds
are tetrasaccharides proposed to be elongation products
coming from a second coupling of the a-glucosyl fluo-
ride donor with the first trisaccharide product via b-
(1!4) linkage.

2060
M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
NO
2
OH
O
OH
O
HO
HO
O
HO
O
OH
OH
OH
OH
OH
O
O
HO
HO
O
HO
HO
O
HO
O
O
O
OH
OH
NO
NO
2
OH
NO
2
O
HO
HO
NO
2
OH
O
HO
HO
O
OH
2
+I
+II
+I
+II
Scheme 2. Binding modes of monosaccharide and disaccharide acceptors according to the regioselectivity observed for the E383A b-glucosidase. The
hydroxyl group to be glycosylated is indicated.
the acceptor; that is, 6-benzyl or 6-benzoyl glucopyran-
ose as acceptor yielded b-(1!2) disaccharides as a con-
sequence of rotation of the glycosyl moiety into subsite
+1 because subsite +II accommodates the aromatic sub-
stituent in O-6.28 On the other hand, b-(1!3) glycosyl-
ation is mainly obtained with glycosynthases from
Sulfolobus solfataricus,12 Pyrococcus furiosus, Thermos-
phaera aggregans13 and Thermus thermophilus.14
stability) and after multiple cycles of conversion (opera-
tional stability). Several immobilized enzyme samples
were stored at 4 ꢀC for different time periods, before
being incubated with donor 1 and acceptor 4 at 35 ꢀC.
The reaction was stopped after 5 h by centrifugation
and the condensation product was quantified by HPLC.
Table 3 presents the % of activity relative to the initial
activity. The yield in product formation is the same as
that obtained with the soluble enzyme but the immobi-
lized form retains at least 80% activity after 55 days of
storage. Moreover, when recycled and reused, complete
operational stability was found after nine cycles (Table
3).
2.3. Immobilized glycosynthase enzyme
Since yields obtained in the preparative glycosynthase
reactions looked promising (Table 1), the E383A b-glu-
cosidase was immobilized to improve its operational
performance for large-scale oligosaccharide synthesis.
Taking advantage of the N-terminal His-tag fused to
the enzyme to simplify the protein purification proto-
col,16 immobilization was performed with Chelating
Glycosynthase activities of soluble and immobilized
enzymes were comparable in the presence of organic
co-solvents. Both maintained 100% activity up to 10%
Sepharose resin previously charged with Ni2+
.
Table 3. Storage and operational stabilities of the immobilized E383A
b-glucosidasea
Quantitative immobilization was assured using a four-
fold excess of resin relative to its nominal loading
capacity.
Operational stability
Storage stability
Cycle
Activity
Cycle
Activity
Day
Activity
1
2
3
4
100
96
103
95
5
6
7
8
9
104
92
102
100
93
0
1
12
54
100
93
94
Reaction between a-glucosyl fluoride (1) and p-nitro-
phenyl glucoside (4) using the immobilized enzyme was
completed after 5 h. The corresponding b-(1!3) disac-
charide was obtained in 95% yield.
83
The soluble enzyme is rather unstable at 4 ꢀC and
should be stored at ꢀ20 ꢀC in 45% glycerol for long
periods.15 The glycosynthase activity of the immobilized
enzyme was checked after long-term storage (storage
a Transglycosylation activities of the immobilized glycosynthase
(10 lM) with donor 1 and acceptor 4 during nine cycles of reuse
(operational stability) and after 54 days of storage (storage stability).
Activities are expressed as percentage of the initial rate.

M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
2061
3. Experimental
100
80
60
40
20
0
A
3.1. General
Unidimensional 1H and 13C NMR spectra were re-
corded on a Varian Gemini-300 spectrometer, and
two-dimensional spectra (COSY, TOCSY, NOESY,
HSQC and HMBC) were recorded on a Inova-500 spec-
trometer. Proton chemical shifts (d in ppm) were refer-
enced to internal Me4Si for solns in CDCl3 and to an
external reference for solns in D2O or Me2SO-d6. Mass
spectra were determined on a VG Platform spectrometer
in the FAB ionization mode: m/z for the peaks (100%)
corresponding to [M+Na]+ are given.
0
10
20
30
40
ACN (%)
100
80
60
40
20
0
B
3.2. Enzyme
E383A mutant b-glucosidase from Streptomyces was ex-
pressed in Escherichia coli and purified as published.16
Enzyme concentrations were determined by UV spectro-
photometry using e280 = 107,000 Mꢀ1 cmꢀ1
.
3.3. Glycosyl fluoride donors 1–3
a-Glycosyl fluorides were prepared according to stan-
dard procedures29,30 by treatment of the corresponding
peracetylated monosaccharide with hydrogen fluoride
in pyridine, purification of the peracetylated a-glycosyl
fluorides by flash chromatography and O-deacetylation
with NaOMe in MeOH to afford compounds 1–3 in
70–80% yield.
0
10
20
30
40
DMSO (%)
Figure 3. Glycosynthase yields of immobilized E383A b-glucosidase
(d) and soluble E383A b-glucosidase (s) with donor 1 (1 mM) and
acceptor 4 (5 mM) in phosphate buffer (50 mM, pH 7) and MeCN (A)
or Me2SO (B) after 20 h of reaction.
The donors were dissolved in deionized D2O, filtered
and stored at ꢀ20 ꢀC. Concentrations of the stock solns
were determined by measuring the total amount of
fluoride ion after hydrolysis with 20 mM H2SO4 at
100 ꢀC for 10 min and neutralization with NaOH using
a selective fluoride electrode (Sentek, UK interfaced
with a CiberScan Bench pH/Ion meter from Eutech
Instruments).
and 30% of MeCN and Me2SO, respectively (Fig. 3).
Likewise, the pH profiles were comparable (data not
shown).
2.4. Conclusions
The E383A glycosynthase from Streptomyces sp. offers
an unexpected regioselectivity compared to the other
glycosynthases so far reported. For monosaccharide
acceptors, the enzyme produces b-(1!3) disaccharides
in excellent yields. Particularly, the disaccharides
Glcb3Man and Galb3Man here obtained have not been
reported for any other glycosynthase. In contrast, b-
(1!4) linkages are obtained with disaccharide accep-
tors. Therefore, interactions in subsite +II of the enzyme
seem to play a major role in directing regioselectivity.
Although immobilization does not increase stability in
organic solvents, the enzyme is remarkably more stable
than the soluble form when stored at 4 ꢀC in phosphate
buffer at pH 7.0, and can be reused for several cycles.
Therefore, the immobilized form guarantees the avail-
ability of this enzyme for the scale-up synthesis of oligo-
saccharides and glycoconjugates.
3.4. Glycosyl acceptors
4-Nitrophenyl b-D-glucopyranoside (4), 4-nitrophenyl
b-D-galactopyranoside (5), 4-nitrophenyl b-D-xylopyr-
anoside (6), 4-nitrophenyl b-D-mannopyranoside (7), 4-
nitrophenyl b-D-fucopyranoside (8) and 4-nitrophenyl
2-acetamido-2-deoxy-glucose (9) were from Fluka.
4-Nitrophenyl b-cellobioside (10) was prepared as
described,31 and 4-nitrophenyl b-laminaribioside (11)
was obtained by glycosynthase reaction of E383A b-
glucosidase as described below.
3.5. Glycosynthase kinetics
Glycosyl donor and acceptor at appropriate concentra-
tions in phosphate buffer (50 mM, ionic strength

2062
M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.27 (d,
0
0
0
0
0.15 M with added KCl, pH 7.0) were pre-incubated at
35 ꢀC for 5 min. After addition of E383A b-glucosidase
(0.01–35 lM, final vol of 0.3 mL), the reactions were
kept at 35 ꢀC. Aliquots were withdrawn at regular time
intervals, diluted with deionized water (1:10), and ana-
lyzed by HPLC (NovaPak C18 column, 4 lm,
3.9 · 150 mm from Waters, 1 mL minꢀ1 or Zorbax
Eclypse XDB C-18 column, 1.8 lm, 4.6 · 50 mm, from
Agilent, 1.2 mL minꢀ1, 5–16% MeOH in water, UV
detector at 300 nm). Chromatographic peaks were iden-
tified by co-injection with independent standards or by
spectroscopic analysis after preparative synthesis. Initial
rates (vo) were calculated from the linear progress curve
of product formation (normalized area vs time) and
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.28 (d, 1H, J1,2
7.5 Hz, H-1I), 4.74 (d, 1H, J1,2 7.5 Hz, H-1II), 4.31 (d,
1H, J4,3 2.7 Hz, H-4I), 4.08–3.40 (m, 11H, H-2I,II, H-
3I,II, H-4II, H-5I,II, H-6aI,II, H-6bI,II). 13C NMR (D2O,
25 ꢀC): d 162.4 (C-10), 143.2 (C-40), 126.7 (C-30, C-50),
117.1 (C-20, C-60), 104.4 (C-1II), 100.3 (C-1I), 82.6 (C-
3I), 76.5–76.0 (C-3II, C-5I,II), 74.0 (C-2II), 70.1, 70.1 (C-
2I, C-4II), 68.7 (C-4I), 61.3, 61.2 (C-6I,II). COSY (D2O,
25 ꢀC) selected data: 4.15, 3.82 (H-4I, H-3I); NOESY
(D2O, 25 ꢀC) selected data: 4.59, 3.82 (H-1II, H-3I);
TOCSY (D2O, 25 ꢀC) selected data: 5.13, 3.82 (H-1I,
H-3I); HSQC (D2O, 25 ꢀC) selected data: 3.82, 82.6 (H-
3I, C-3I); HMBC (D2O, 25 ꢀC) selected data: 4.59, 82.6
(H-1II, C-3I). HPLC (10% MeOH): tR = 10.85 min.
b-(1!3) Connectivity: confirmed by the presence of a
correlation of H-1II and C-3I (4.59, 82.6) in the HMBC
spectrum.
0
0
0
0
expressed as vo/[E] in sꢀ1
.
3.6. Preparative glycosynthase-catalyzed syntheses
A soln of donor (5–20 mg, 1 equiv), acceptor (5 equiv)
and E383A b-glucosidase (2–28 lM) was incubated at
35 ꢀC in phosphate buffer (50 mM, pH 7.0, 1–6 mL).
The reactions were monitored by HPLC as indicated
above. After 24–48 h, the reaction mixture was filtered
and directly loaded onto a Lichroprep RP-18 Lobar-A
column (E. Merck), which was eluted with a gradient
MeOH–water from 0% to 3–20% (v/v). Unreacted
acceptor eluted first, followed by the condensation
product. It was freeze dried and lyophilized for storage.
The results are summarized in Tables 1 and 2.
Yields and spectral data of the condensation products
are:
3.6.3. 4-Nitrophenyl b-D-glucopyranosyl-(1!4)-b-D-xylo-
pyranoside. From donor 1 (5.2 mg) and acceptor 6
(37.6 mg); 3.1 mg (28%); FABMS: m/z 456.15
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.29 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.27 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.27 (d, 1H, J1,2
7.5 Hz, H-1I), 4.57 (d, 1H, J1,2 8.0 Hz, H-1II), 4.23–
3.27 (11H, H-2I,II, H-3I,II, H-4I,II, H-5aI, H-5bI, H-5II,
H-6aII, H-6bII). 13C NMR (D2O, 25 ꢀC) d: 162.2 (C-
10), 143.3 (C-40), 126.8 (C-30, C-50), 117.1 (C-20, C-60),
101.9 (C-1II), 100.5 (C-1I), 76.8–76.2 (C-3II, C-4I, C-
5II), 74.2–73.2 (C-2I,II, C-3I), 70.3 (C-4II), 63.8 (C-5I),
61.4 (C-6II). HPLC (16% MeOH): tR = 7.77 min.
b-(1!4) Connectivity: dH at 4.57 ppm characteristic
of H-1 in a b-(1!4)-linked glucose, and dC at 76 ppm
corresponding to O-glycosylated C-4.23,24,18
0
0
0
0
3.6.1. 4-Nitrophenyl b-D-glucopyranosyl-(1!3)-b-D-
glucopyranoside (11). From donor 1 (5.1 mg) and
acceptor 4 (43 mg); 12.4 mg (96%); FABMS: m/z
486.12 [M+Na]+; 1H NMR (D2O, 25 ꢀC): d 8.29 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.27 (d, 2H, H-20,
H-60), 5.31 (d, 1H, J1,2 7.5 Hz, H-1I), 4.80 (d, 1H, J1,2
8.0 Hz, H-1II), 3.95–3.35 (m, 12H, H-2I,II, H-3I,II, H-
4I,II, H-5I,II, H-6aI,II, H-6bI,II); 13C NMR (D2O,
25 ꢀC): d 162.4 (C-10), 143.3 (C-40), 126.8 (C-30, C-50),
117.2 (C-20, C-60), 103.5 (C-1II), 99.9 (C-1I), 84.6 (C-
3I), 76.7–76.2 (C-3II, C-5I,II), 74.1, 73.2 (C-2I,II), 70.3
(C-4II), 68.5 (C-4I), 61.4, 61.1 (C-6I,II). HPLC (16%
MeOH): tR 5.14 min.
0
0
0
0
3.6.4. 4-Nitrophenyl b-D-glucopyranosyl-(1!3)-b-D-xylo-
pyranoside. From donor 1 (5.2 mg) and acceptor 6
(37.6 mg); 6.7 mg (54%); FABMS: m/z 456.11
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.28 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.26 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.26 (d, 1H, J1,2
7.0 Hz, H-1I), 4.8 (d, 1H, J1,2 8.0 Hz, H-1II), 4.13–3.4
(m, 11H, H-2I,II, H-3I,II, H-4I,II, H-5aI, H-5bI, H-5II,
H-6aII, H-6bII). 13C NMR (D2O, 25 ꢀC): d 162.2 (C-
10), 143.3 (C-40), 126.8 (C-30, C-50), 117.1 (C-20, C-60),
103.4 (C-1II), 100.4 (C-1I), 84.1 (C-3I), 76.7, 76.2 (C-
3II, C-5II), 74.1, 72.9 (C-2I,II), 70.3 (C-4II), 68.3 (C-4I),
65.6 (C-5I), 61.4 (C-6II). HPLC (16% MeOH):
tR = 14.37 min.
0
0
0
0
b-(1!3) Connectivity: dH at 4.80 ppm characteristic
of H-1 in a b-(1!3)-linked glucose, and dC at
84.6 ppm corresponding to O-glycosylated C-3 and at
68.5 ppm characteristic of unsubstituted C-4 next to
O-glycosylated C-3.23,25
b-(1!3) Connectivity: dH at 4.8 ppm characteristic of
H-1 in a b-(1!3)-linked glucose, and dC at 84.1 ppm
corresponding to O-glycosylated C-3 and at 68.3 ppm
assigned to unsubstituted C-4 next to O-glycosylated
C-3.23,24,18
3.6.2. 4-Nitrophenyl b-D-glucopyranosyl-(1!3)-b-D-
galactopyranoside. From donor 1 (5.1 mg) and accep-
tor 5 (40 mg); 13.1 mg (100%); FABMS: m/z 486.12
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.28 (d,

M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
2063
3.6.5. 4-Nitrophenyl b-D-glucopyranosyl-(1!3)-b-D-
mannopyranoside. From donor 1 (5.2 mg) and accep-
tor 7 (43.2 mg); 7.1 mg (53%); FABMS: m/z 486.14
b-(1!3) Connectivity: dH at 4.73 ppm characteristic
of H-1 in a b-(1!3) linkage, and dC at 84.6 ppm corre-
sponding to O-glycosylated C-3 and at 68.6 ppm charac-
teristic of unsubstituted C-4 next to O-glycosylated C-
3.23–25
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.28 (d,
J3 ;2 ¼ J5 ;6 9:3 Hz, 2H, H-30, H-50), 7.24 (d,
0
0
0
0
I
J2 ;3 ¼ J6 ;5 9:3 Hz, 2H, H-20, H-60), 5.53 (s, 1H, H-1 ),
0
0
0
0
4.65 (d, 1H, J1,2 7.8 Hz, H-1II), 4.47 (d, 1H, J2,3
3.6.8. 4-Nitrophenyl b-D-galactopyranosyl-(1!3)-b-D-
galactopyranoside. From donor 2 (10.8 mg) and accep-
tor 5 (89 mg); 3.6 mg (14%); FABMS: m/z 486.123
2.5 Hz, H-2I), 4.10–3.40 (11H, H-2II, H-3I,II, H-4I,II
,
H-5I,II, H-6aI,II, H-6bI,II). 13C NMR (D2O, 25 ꢀC): d
162.0 (C-10), 143.1 (C-40), 126.7 (C-30, C-50), 116.9 (C-
20, C-60), 100.9 (C-1II), 97.6 (C-1I), 80.2 (C-3I), 77.0–
76.2 (C-3II, C-5I,II), 73.6 (C-2II), 70.2 (C-4II), 68.4 (C-
2I), 65.5 (C-4I), 61.5, 61.4 (C-6I,II). COSY (D2O,
25 ꢀC) selected data: 4.31, 3.83 (H-2I, H-3I); NOESY
(D2O, 25 ꢀC) selected data: 4.50, 3.83 (H-1II, H-3I);
HSQC (D2O, 25 ꢀC) selected data: 3.83, 80.2 (H-3I, C-
3I); HMBC (D2O, 25 ꢀC) selected data: 4.50, 80.2 (H-
1II, C-3I). HPLC (5% MeOH): tR = 12.97 min.
b-(1!3) Connectivity: confirmed by the presence of a
correlation of H-1II and C-3I (4.50, 80.2) in the HMBC
spectrum.
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.13 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.13 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.14 (d, 1H, J1,2
8.0 Hz, H-1I), 4.53 (d, 1H, J1,2 7.5 Hz, H-1II), 4.17 (dd,
1H, J4,3 = J4,5 3.5 Hz, H-4I), 3.92–3.39 (11H, H-2I,II
H-3I,II, H-4I,II, H-5I,II, H-6aI,II, H-6bI,II). 13C NMR
(D2O, 25 ꢀC): d 162.5 (C-10), 143.2 (C-40), 126.8 (C-30,
C-50), 117.2 (C-20, C-60), 105.0 (C-1II), 100.3 (C-1I),
82.5 (C-3I), 75.9–70.1 (C-2I,II, C-3II, C-5I,II), 69.2 (C-
4II), 68.8 (C-4I), 61.6, 61.3 (C-6I,II). NOESY (D2O,
25 ꢀC) selected data: 5.14, 3.89 (H-1I, H-3I), 4.53,
3.89 (H-1II, H-3I). HMBC (D2O, 25 ꢀC) selected data:
4.53, 82.5 (H-1II, C-3I). HPLC (5% MeOH): tR =
11.12 min.
0
0
0
0
3.6.6. 4-Nitrophenyl b-D-glucopyranosyl-(1!3)-b-D-fuco-
pyranoside. From donor 1 (5.1 mg) and acceptor 8
(41 mg); 3.8 mg (30%); FABMS: m/z 470.13 [M+Na]+.
b-(1!3) Connectivity: confirmed by the presence of a
correlation of H-1II and C-3I (4.53, 82.5) in the HMBC
spectrum.
1H NMR (D2O, 25 ꢀC): d 8.13 (d, J3 ;2 ¼ J5 ;6 9:0 Hz,
0
0
0
0
2H, H-30, H-50), 7.12 (d, J2 ;3 ¼ J6 ;5 9:0 Hz, 2H, H-20,
H-60), 5.12 (d, 1H, J1,2 7.0 Hz, H-1I), 4.58 (d, 1H, J1,2
7.5 Hz, H-1II), 3.97–3.26 (m, 10H, H-2I,II, H-3I,II, H-
4I,II, H-5I,II, H-6aII, H-6bII), 1.32 (d, 3H, J6,5 6.5 Hz,
0
0
0
0
3.6.9. 4-Nitrophenyl b-D-galactopyranosyl-(1!4)-b-D-
xylopyranoside. From donor 2 (12.6 mg) and acceptor
6 (79.4 mg); 3.6 mg (12%); FABMS: m/z: 456.112
13
H-6I). C NMR (D2O, 25 ꢀC): d 162.4 (C-10), 143.2
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.29 (d,
(C-40), 126.7 (C-30, C-50), 117.0 (C-20, C-60), 104.5 (C-
1II), 100.0 (C-1I), 82.7 (C-3I), 76.5, 76.2 (C-3II, C-5II),
74.0–69.8 (C-2I,II, C-4I,II, C-5I), 61.2 (C-6II), 16.0 (C-
6I). COSY (D2O, 25 ꢀC) selected data: 3.97, 3.83 (H-4I,
H-3I). HSQC (D2O, 25 ꢀC) selected data: 3.83, 82.7
(H-3I, C-3I). HMBC (D2O, 25 ꢀC) selected data: 4.58,
82.7 (H-1II, C-3I). HPLC (16% MeOH): tR = 11.37 min.
b-(1!3) Connectivity: confirmed by the presence of a
correlation of H-1II and C-3I (4.58, 82.7) in the HMBC
spectrum.
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.27 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.26 (d, 1H, J1,2
7.5 Hz, H-1I), 4.49 (d, 1H, J1,2 7.5 Hz, H-1II), 4.22–3.5
(m, 11H, H-2I,II, H-3I,II, H-4I,II, H-5aI, H-5bI, H-5II,
H-6aII, H-6bII). 13C NMR (D2O, 25 ꢀC): d 162.2 (C-
10), 143.3 (C-40), 126.8 (C-30, C-50), 117.1 (C-20, C-60),
102.5 (C-1II), 100.5 (C-1I), 76.8–74.3 (C-3II, C-4I, C-
5II), 73.3–71.3 (C-2I,II, C-3I), 69.3 (C-4II), 63.8 (C-5I),
61.8 (C-6II). HPLC (16% MeOH): tR = 6.60 min.
0
0
0
0
b-(1!4) Connectivity: dH at 4.49 ppm characteristic
of H-1 in a b-(1!4) glycosidic linkage, and dC at
76 ppm corresponding to O-glycosylated C-4.23,24,18
3.6.7. 4-Nitrophenyl b-D-galactopyranosyl-(1!3)-b-D-
glucopyranoside. From donor 2 (10.2 mg) and acceptor
4 (85.7 mg); 19.9 mg (77%); FABMS: m/z 486.123
3.6.10. 4-Nitrophenyl b-D-galactopyranosyl-(1!3)-b-D-
xylopyranoside. From donor 2 (12.6 mg) and acceptor
6 (79.4 mg); 7.5 mg (25%); FABMS: m/z 456.112
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.27 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.25 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.31 (d, 1H, J1,2
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.29 (d,
0
0
0
0
6.0 Hz, H-1I), 4.73 (d, 1H, J1,2 7.0 Hz, H-1II), 4.00–
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.27 (d,
0
0
0
0
3.60 (m, 12H, H-2I,II, H-3I,II, H-4I,II, H-5I,II, H-6aI,II
,
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.28 (d, 1H, J1,2
5.5 Hz, H-1I), 4.71 (d, 1H, J1,2 6.5 Hz, H-1II), 4.12–
3.55 (m, 11H, H-2I, II, H-3I,II, H-4I,II, H-5aI, H-5bI, H-
5II, H-6aII, H-6bII). 13C NMR (D2O, 25 ꢀC): d 162.2
(C-10), 143.3 (C-40), 126.8 (C-30, C-50), 117.2 (C-20, C-
60), 103.9 (C-1II), 100.4 (C-1I), 84.3 (C-3I), 76.0–71.9
0
0
0
0
13
H-6bI,II). C NMR (D2O, 25 ꢀC): d = 162.3 (C-10),
143.3 (C-40), 126.7 (C-30, C-50), 117.1 (C-20, C-60),
104.0 (C-1II), 99.9 (C-1I), 84.6 (C-3I), 76.5–71.9 (C-
2I,II, C-3II, C-5I,II), 69.3 (C-4II), 68.6 (C-4I), 61.8, 61.1
(C-6I,II). HPLC (10% MeOH): tR = 8.13 min.

2064
M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
(C-2I,II, C-3II, C-5II), 69.2 (C-4II), 68.4 (C-4I), 65.5 (C-
5I), 61.7 (C-6II). HPLC (16% MeOH): tR = 11.06 min.
b-(1!3) Connectivity: dH at 4.71 ppm characteristic
of H-1 in a b-(1!3)-linked glucose, and dC at
84.3 ppm corresponding to O-glycosylated C-3 and at
68.4 ppm assigned to unsubstituted C-4 next to O-gly-
cosylated C-3.23,24,18
in phosphate buffer (50 mM, pH 7) and E383A mutant
soln (40–170 lg) was added. After 15 min, the suspension
was centrifuged. Immobilization yields were quantified
determining unbound enzyme by spectrophotometry.
3.8. Immobilized glycosynthase reactions
a-Glucosyl fluoride (1) as donor and p-nitrophenyl glu-
coside (2) as acceptor at 1–5 M ratio were added to a
suspension containing immobilized E383A (40–200 lg)
in phosphate buffer (50 mM, pH 7, final vol 300 lL).
The reactions were kept at 35 ꢀC, terminated by filtra-
tion or centrifugation, diluted with deionized H2O
(1:10) and analyzed by HPLC as before.
In order to determine the storage stability, several
samples of immobilized E383A (162 lg enzyme/20 mg
resin) were stored at 4 ꢀC. After different times of
storage, the remaining glycosynthase activity was
determined.
With respect to the operational stability, three sam-
ples of immobilized E383A (108 lg enzyme/20 mg resin)
were prepared and the glycosynthase activity was mea-
sured after nine cycles of substrates conversion. Between
reactions, the immobilized enzyme was kept at 4 ꢀC sus-
pended in phosphate buffer for 19 h.
3.6.11. 4-Nitrophenyl b-D-galactopyranosyl-(1!3)-b-D-
mannopyranoside. From donor 2 (10.3 mg) and accep-
tor 7 (89 mg); 12.3 mg (47%); FABMS: m/z 486.123
[M+Na]+. 1H NMR (D2O, 25 ꢀC):
d 8.28 (d,
J3 ;2 ¼ J5 ;6 9:5 Hz, 2H, H-30, H-50), 7.24 (d,
0
0
0
0
J2 ;3 ¼ J6 ;5 9:5 Hz, 2H, H-20, H-60), 5.53 (d, 1H, H-
1I), 4.60 (d, 1H, J1,2 7.0 Hz, H-1II), 4.45 (d 1H, J2,3
2.5 Hz, H-2I), 4.08 (dd, 1H, J3,2 2.5 Hz, J3,4 9.5 Hz,
0
0
0
0
H-3I), 4.02–3.6 (m, 10H, H-2II, H-3II, H-4I,II, H-5I,II
,
H-6aI,II, H-6bI,II). 13C NMR (D2O, 25 ꢀC): d 162.1
(C-10), 143.1 (C-40), 126.7 (C-30, C-50), 116.9 (C-20,
C-60), 101.5 (C-1II), 97.6 (C-1I), 80.2 (C-3I) 77.0–73.3
(C-3II, C-5I,II), 71.4 (C-2II), 69.3 (C-4II), 68.6 (C-2I),
65.6 (C-4I), 61.8, 61.5 (C-6I,II). COSY (D2O, 25 ꢀC)
selected data: 4.31, 3.83 (H-2I, H-3I); NOESY (D2O,
25 ꢀC) selected data: 4.45, 3.83 (H-1II, H-3I); HMBC
(D2O, 25 ꢀC) selected data: 4.45, 80.2 (H-1II, C-3I).
HPLC (16% MeOH): tR = 4.12 min.
b-(1!3) Connectivity: confirmed by the presence of a
correlation of H-1II and C-3I (4.45, 80.2) in the HMBC
spectrum.
Acknowledgements
This work was supported in part by Grants BIO2001-
2064-C02-02 and BFU2004-06377-C02-02 from the
3.6.12. 4-Nitrophenyl b-cellotrioside. From donor 1
(45 mg) and acceptor 10 (263 mg); 125 mg (80%);
FABMS: m/z 648.18 [M+Na]+. 1H NMR (D2O,
´
Ministerio de Ciencia y Tecnologıa, Spain. M.F. and
X.P. acknowledge the doctoral fellowships from Insti-
´
`
tuto Danone and Institut Quımic de Sarria, respectively.
25 ꢀC): d 8.28 (d, J3 ;2 ¼ J5 ;6 9:3 Hz, 2H, H-30, H-50),
0
0
0
0
´
We thank Dr. J. A. Perez Pons, M. Vallmitjana and
7.26 (d, J2 ;3 ¼ J6 ;5 9:3 Hz, 2H, H-20, H-60), 5.31 (d,
1H, J1,2 7.5 Hz, H-1I), [4.58 (d, 1H, J1,2 7.8 Hz), 4.53
(d, 1H, J1,2 7.8 Hz) H-1II,III], 4.00–3.20 (m, 18H,
H-2I–III, H-3I–III, H-4I–III, H-5I–III, H-6aI–III, H-6bI–III).
13C NMR (DMSO, 25 ꢀC): 162.3 (C-10), 143.3 (C-40),
126.8 (C-30, C-50), 117.1 (C-20, C-60), 103.2–99.9
0
0
0
0
E. Querol, for providing the gene of the Streptomyces
b-glucosidase and for helpful collaboration.
References
(C-1I–III), 79.0, 78.7 (C-4I,II), 76.6–73.2 (C-2I–III
,
1. Mackenzie, L. F.; Wang, Q.; Warren, R. A. J.; Withers,
S. G. J. Am. Chem. Soc. 1998, 120, 5583–5584.
2. Malet, C.; Planas, A. FEBS Lett. 1998, 440, 208–212.
3. Williams, S. J.; Withers, S. G. Aust. J. Chem. 2002, 55, 3–12.
4. Perugino, G.; Trincone, A.; Rossi, M.; Moracci, M.
Trends Biotechnol. 2004, 22, 31–37.
5. Planas, A.; Faijes, M. Afinidad 2002, 59, 295–313.
6. Coutinho, P. M.; Henrissat, B. Recent Advances in
Carbohydrate Bioengineering; The Royal Society of Chem-
istry: Cambridge, 1999; Chapter 1, pp 3–12.
C-3I–III, C-5I–III), 70.1 (C-4III), 61.2–60.3 (C-6I–III).
HPLC (14% MeOH): tR = 6.44 min.
b-(1!4) Connectivity: dH at 4.53 ppm characteristic
of H-1 in b-(1!4) glycosidic linkages, and dC at
76 ppm corresponding to O-glycosylated C-4.23,24,31
3.7. Enzyme immobilization
´
´
7. Faijes, M.; Perez, X.; Perez, O.; Planas, A. Biochemistry
2003, 42, 13304–13318.
Chelating Sepharose (500 lL, Amersham Biosciences 17-
0575-01) was washed with 2 vol of deionized water
(2 · 5 mL) and then saturated with 0.8 M NiSO4 soln.
The resin was washed with deionized water (5 · 5 mL)
to remove unbounded Ni2+ and dried under diminished
pressure for 5 min. Dried resin (20 mg) was resuspended
8. Fort, S.; Boyer, V.; Grefee, L.; Davies, G. J.; Moroz, O.;
Christiansen, L.; Schulein, M.; Cottaz, S.; Driguez, H. J.
¨
Am. Chem. Soc. 2000, 122, 5429–5437.
9. Hrmova, M.; Imai, T.; Rutten, S. J.; Fairweather, J. K.;
Pelosi, L.; Bulone, V.; Driguez, H. J. Biol. Chem. 2002,
277, 30102–30111.

M. Faijes et al. / Carbohydrate Research 341 (2006) 2055–2065
2065
10. Jahn, M.; Stoll, D.; Warren, R. A. J.; Szabo, L.; Singh, P.;
Gilbert, H. J.; Ducros, V. M.; Davies, G. J.; Withers, S. G.
Chem. Commun. 2003, 1327–1329.
11. van Lieshout, J.; Faijes, M.; Nieto, J.; van der Oost, J.;
Planas, A. Archaea 2004, 1, 285–292.
12. Trincone, A.; Perugino, G.; Rossi, M.; Moracci, M.
Bioorg. Med. Chem. Lett. 2000, 10, 365–368.
13. Perugino, G.; Trincone, A.; Giordano, A.; van der Oost,
J.; Kaper, T.; Rossi, M.; Moracci, M. Biochemistry 2003,
42, 8484–8493.
20. Roth, N. J.; Huber, R. E. J. Biol. Chem. 1996, 271, 14296–
14301.
21. Bauer, M. W.; Kelly, R. M. Biochemistry 1998, 37, 17170–
17178.
22. Kim, Y.-W.; Chen, H.; Withers, S. G. Carbohydr. Res.
2005, 340, 2735–2741.
23. Bock, K.; Pedersen, C.; Pedersen, H. Adv. Carbohydr.
Chem. Biochem. 1984, 42, 193–225.
24. Bradbury, J. H.; Jenkins, G. A. Carbohydr. Res. 1984, 126,
125–156.
14. Drone, J.; Feng, H.; Tellier, C.; Hoffmann, L.; Tran, V.;
Rabiller, C.; Dion, M. Eur. J. Org. Chem. 2005, 1977–1983.
15. Perez-Pons, J. A.; Cayetano, A.; Rebordosa, X.; Lloberas,
J.; Guasch, A.; Querol, E. Eur. J. Biochem. 1994, 223, 557–
565.
25. Kochetkov, N. K.; Vinogradov, E. V.; Knirel, Y. A.;
Shashkov, A. S.; Lipkind, G. M. Bioorg. Khim. 1992, 18,
116–125.
26. Jansson, P. E.; Kenne, L.; Widmalm, G. Carbohydr. Res.
1987, 168, 67–77.
´
´
16. Vallmitjana, M.; Ferrer-Navarro, M.; Planell, R.; Abel,
M.; Ausin, C.; Querol, E.; Planas, A.; Perez-Pons, J. A.
Biochemistry 2001, 40, 5975–5982.
27. Planas, A.; Millet, O.; Palası, J.; Pallares, C.; Abel, M.;
Viladot, J. L. Carbohydr. Res. 1998, 310, 53–64.
28. Stick, R. V.; Stubbs, K. A.; Watts, A. G. Aust. J. Chem.
2004, 57, 779–786.
´
17. Montero, E.; Vallmitjana, M.; Perez-Pons, J. A.; Querol,
´
E.; Jimenez-Barbero, J.; Canada, F. J. FEBS Lett. 1998,
29. Hayashi, M.; Hashimoto, S. I.; Noyori, R. Chem. Lett.
1984, 1747–1750.
˜
421, 243–248.
´
18. Montero, E. Ph.D. Thesis, Universidad Autonoma de
30. Yokoyama, M. Carbohydr. Res. 2000, 327, 5–14.
31. van Tilbeurgh, H.; Loontiens, F. G.; Engelborgs, Y.;
Claeyssens, M. Eur. J. Biochem. 1989, 184, 553–
559.
Madrid, 1998.
19. Kempton, J. B.; Withers, S. G. Biochemistry 1992, 31,
9961–9969.