Examining the role of phosphate in glycosyl transfer reactions of Cellulomonas uda cellobiose phosphorylase using d-glucal as donor substrate

Article abstract of DOI:10.1016/j.carres.2012.04.003

Cellobiose phosphorylase from Cellulomonas uda (CuCPase) is shown to utilize d-glucal as slow alternative donor substrate for stereospecific glycosyl transfer to inorganic phosphate, giving 2-deoxy-α-d-glucose 1-phosphate as the product. When performed in D₂O, enzymatic phosphorolysis of d-glucal proceeds with incorporation of deuterium in equatorial position at C-2, implying a stereochemical course of reaction where substrate becomes protonated from below its six-membered ring through stereoselective re side attack at C-2. The proposed catalytic mechanism, which is supported by results of docking studies, involves direct protonation of d-glucal by the enzyme-bound phosphate, which then performs nucleophilic attack on the reactive C-1 of donor substrate. When offered d-glucose next to d-glucal and phosphate, CuCPase produces 2-deoxy-β-d-glucosyl-(1→4)-d-glucose and 2-deoxy-α-d-glucose 1-phosphate in a ratio governed by mass action of the two acceptor substrates present. Enzymatic synthesis of 2-deoxy-β-d-glucosyl-(1→4)-d-glucose is effectively promoted by catalytic concentrations of phosphate, suggesting that catalytic reaction proceeds through a quaternary complex of CuCPase, d-glucal, phosphate, and d-glucose. Conversion of d-glucal and phosphate presents a convenient single-step synthesis of 2-deoxy-α-d-glucose 1-phosphate that is difficult to prepare chemically.

Full text of DOI:10.1016/j.carres.2012.04.003

Carbohydrate Research 356 (2012) 224–232
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Examining the role of phosphate in glycosyl transfer reactions of Cellulomonas
uda cellobiose phosphorylase using D-glucal as donor substrate
Patricia Wildberger a, Lothar Brecker b, Bernd Nidetzky a,
⇑
a Institute of Biotechnology and Biochemical Engineering, Graz University of Technology, Petersgasse 12/1, A-8010 Graz, Austria
b Institute of Organic Chemistry, University of Vienna, Währingerstraße 38, A-1090 Vienna, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Cellobiose phosphorylase from Cellulomonas uda (CuCPase) is shown to utilize
donor substrate for stereospecific glycosyl transfer to inorganic phosphate, giving 2-deoxy-
phosphate as the product. When performed in D2O, enzymatic phosphorolysis of
incorporation of deuterium in equatorial position at C-2, implying a stereochemical course of reaction
where substrate becomes protonated from below its six-membered ring through stereoselective re side
attack at C-2. The proposed catalytic mechanism, which is supported by results of docking studies,
D
-glucal as slow alternative
-glucose 1-
-glucal proceeds with
Received 1 March 2012
Received in revised form 3 April 2012
Accepted 4 April 2012
a-D
D
Available online 13 April 2012
Keywords:
Cellobiose phosphorylase
Glycoside hydrolase family GH-94
involves direct protonation of
philic attack on the reactive C-1 of donor substrate. When offered
phate, CuCPase produces 2-deoxy-b- -glucosyl-(1?4)- -glucose and 2-deoxy-
in a ratio governed by mass action of the two acceptor substrates present. Enzymatic synthesis of 2-
deoxy-b- -glucosyl-(1?4)- -glucose is effectively promoted by catalytic concentrations of phosphate,
suggesting that catalytic reaction proceeds through a quaternary complex of CuCPase, -glucal, phos-
phate, and -glucose. Conversion of -glucal and phosphate presents a convenient single-step synthesis
of 2-deoxy- -glucose 1-phosphate that is difficult to prepare chemically.
Ó 2012 Elsevier Ltd. All rights reserved.
D
-glucal by the enzyme-bound phosphate, which then performs nucleo-
-glucose next to -glucal and phos-
-glucose 1-phosphate
D
D
D
D
a-D
D-Glucal
Catalytic mechanism
Stereospecific protonation
2-Deoxy-D-glucosyl transfer
D
D
D
D
D
a
-D
1. Introduction
residue in the CPase active site. CPase crystal structures reveal a
distinct phosphate recognition site in the enzyme, rich in positively
charged amino acids (Fig. 1b). Once accommodated at this site,
phosphate is optimally positioned for direct nucleophilic attack
on the glycosidic carbon of cellobiose (Scheme 1).
Cellobiose phosphorylase (CPase; EC 2.4.1.20) catalyzes revers-
ible phosphorolysis of cellobiose into -glucose 1-phosphate and
-glucose.1 Like the majority of known disaccharide phosphoryl-
a
-D
D
ases, CPase is an inverting enzyme that forms a glycosidic product
In an original classification based on sequence analysis, CPase
was classified as a member of the glycosyltransferase families
and further categorized into family GT-36. This classification,
which is in accordance with CPase catalytic function of glucosyl
transfer to and from phosphate, as also reflected in the EC number,
needed to be revised when the first crystal structure of an enzyme
from the CPase group, Vibrio proteolyticus chitobiose phosphorylase
(VbChPase), became available.6 Evidence that VbChPase was struc-
turally similar to certain glycoside hydrolases, in particular to
maltose phosphorylase and glucoamylase from glycoside hydro-
lase family GH-65 and GH-15, respectively, necessitated reclassifi-
cation of CPase as a glycoside hydrolase, forming a new family GH-
94.6 It is worth noting therefore that CPase enzymes usually show
no or just scant glycoside hydrolase activity. As of today, family
GH-94 consists of four principal members, CPase (EC 2.4.1.20), cel-
lodextrin phosphorylase (EC 2.4.1.49), ChPase (EC 2.4.1.-), and cyc-
lic b-1,2-glucan synthase (EC 2.4.1.-) (http://www.cazy.org/CAZY).
Among individual enzymes belonging to family GH-94, CPase from
Cellvibrio gilvus (CgCPase) is the one most extensively studied.7–9
Two crystal structures of CgCPase are available, one containing
having the opposite anomeric configuration as the substrate.2 The
stereochemical course of cellobiose conversion involves a b-to-
a
change in anomeric configuration, proposed to result from a single
displacement-like reaction that produces an equatorial-to-axial
substitution at the anomeric carbon.3 The canonical catalytic
mechanism of enzymatic glycosyl transfer with inversion (used
by glycoside hydrolases, for example) involves a pair of acidic res-
idues, typically Asp or Glu, that are positioned in the active site so
that one can function as the general catalytic acid and the other as
the general catalytic base.4,5 CPase is thought to employ a modified
version of this mechanism, in which only one catalytic residue
(Asp) is used, and this has the role of a general acid.2 Since the
acceptor substrate, phosphate, is already ionized at physiological
pH, it does not need to be activated by deprotonation from a cata-
lytic base, therefore explaining the absence of a corresponding
⇑
Corresponding author. Tel.: +43 316 873 8400; fax: +43 316 873 8434.
E-mail address: bernd.nidetzky@tugraz.at (B. Nidetzky).
0008-6215/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carres.2012.04.003

P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
225
shown in Figure 1a and consists of four distinct regions: an N-
a
terminal domain, a helical linker region, an
and a C-terminal domain.
a-helix barrel domain,
The substrate specificity of CPase has been examined in a num-
ber of biochemical studies using enzymes from different
sources.1,10–12 It was convenient to analyze the reverse direction
of reaction where a glycosyl residue is transferred from sugar 1-
phosphate donor to acceptor. The various CPases are similar in that
relatively relaxed specificity for the acceptor substrate is con-
trasted with highly restricted use of glycosyl donors.7,13 Two rele-
vant examples of active donor substrate analogues are
a-D-glucose
1-fluoride14 and -galactose 1-phosphate.15 Note that ChPase
a
-D
has of course different structural requirements to its donor
substrate as compared to CPase where in particular 2-hydroxyl in
a-D-glucose 1-phosphate is replaced by 2-acetamido in a-D-N-ace-
tylglucosamine 1-phosphate.6 In a recent paper, Kitaoka and co-
workers reported the interesting observation that the endocyclic
enitol
deoxy-glucosyl transfer to
(e.g. -xylose, -mannose) used normally in the reaction with
-glucose 1-phosphate.16 They also demonstrated the transfer
products to have b-(1?4) glycosidic structure. Using Cellulomonas
uda CPase (CuCPase), Nidetzky and co-workers had earlier exam-
D
-glucal was accepted by CgCPase as donor substrate for 2-
b
D
-glucose and other sugar acceptors
W488
2.5 Å
D
D
a-
D
D490
2.9 Å
ined reactivity toward
transfer to phosphate or
any.14 Inspired by Kitaoka’s findings, the topic of utilization of
D-glucal was therefore revisited herein, and we would like to com-
municate results of a comprehensive analysis that provide deep-
ened insight into the underlying reaction mechanism. We in
particular clarify the role of phosphate in promoting the enzymatic
D
-glucal as donor substrate for glycosyl
E649
Y653
E659
D
-glucose but were unable to detect
2.7 Å
2.6 Å
3.8 Å
R362
H666
2.6 Å
3.2 Å
3.6 Å
K658
D368
G732
reaction with
tions used, -glucal serves as donor substrate for 2-deoxy-
syl transfer to phosphate or sugar acceptor. Requirement for
phosphate to essentially ‘activate’ CuCPase reaction with -glucal
D-glucal and show that depending on reaction condi-
G712
D
D-gluco-
R351
D
T731
explains negative findings of our previous experiments that in
direction of disaccharide synthesis were all performed in the ab-
sence of phosphate. In addition, the amount of native enzyme (iso-
lated from the natural host C. uda) and the analytical method used
(TLC) may have been insufficient for identification of
donor substrate for phosphorolysis. We herein also show that con-
version of -glucal in the presence of phosphate constitutes a con-
venient single-step method for preparation of 2-deoxy- -glucose
1-phosphate (1) that is difficult to synthesize chemically.
In chemical terms, -glucal and related glycals are characterized
as 1,2-unsaturated derivatives of pentoses or hexoses. These com-
pounds have been widely used as mechanistic probes for study of
glycoside hydrolases and phosphorylases.17–23 Their unique fea-
ture of structure and function lies in the half-chair conformation
of the typically six-membered ring that contains a highly reactive
double bond between C-1 and C-2. It is known from seminal work
D-glucal as
Figure 1. (a) Three-dimensional fold of the protein subunit of CgCPase (PDB-entry
2cqt, molecule B). CgCPase consists of an N-terminal domain (blue), linker helices
(gray), an
up view of the predicted binding of
entry 2cqt, molecule B) complexed with phosphate (orange) and
a
-helix barrel domain (red), and a C-terminal domain (green). (b) Close-
-glucal (gray) in subsite -1 of CgCPase (PDB-
-glucose (purple)
D
D
a-D
D
in subsite +1. Interactions potentially relevant for catalysis and the reaction are
shown as gray-dashed lines, and strong hydrogen bonds stabilizing the quaternary
complex are shown as black-dashed lines. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)
D
sulfate (PDB-entry 2cqs) and one containing phosphate (PDB-entry
2cqt) at the catalytic site.2 The structure of CgCPase monomer is
Asp-490
Asp-490
Asp-490
+
+
δ-
-
O
O
O
OH
OH
HO
O
δ+
O
O
OH
H
OH
OH
δ-
OH
H
O
H
H
O
OH
O
O
OH
O
OH
HO
HO
HO
O
-
O
O
H
O
H
O
δ+
O
O
δ-
O
HO
HO
HO
O
H
O
OH
-
HO
HO
OH
-
OH
-
O
O
O
HO
P
P
O
O
P
HO
O
O
Scheme 1. Proposed catalytic mechanism of CuCPase. Asp490 is thought to adopt the catalytic role of a general acid and base in direction of cellobiose phosphorolysis and
synthesis, respectively. See Ref. 2 for a detailed discussion.

226
P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
of Hehre, Lehmann, Legler and their co-workers performed as early
as the 1970s that glycals are substrates to become hydrated by
CuCPase in a previous paper from this laboratory.31 It should be
noted that CuCPase isolation was possible with only a single puri-
fication step via the enzyme’s His-tag, using Ni-chelating Sephar-
ose Fast Flow chromatography with a His-Trap HP column. In
previous work using Cu-loaded chelating Sepharose CL-4B, a sec-
ond chromatographic step on hydroxyapatite resin was necessary
to obtain highly purified enzyme. Elution of CuCPase occurred at
a higher concentration of imidazole (250 mM) using the ‘new’
method as compared to the published protocol (120 mM imidaz-
ole), suggesting that interaction of the His-tagged protein with
the chromatographic material was enhanced under the conditions
used herein. Overall, the purification of CuCPase was made simpler
and more time-efficient as compared to the reported protocol. In
some batches of enzyme purified according to literature, we noted
the problem of co-purification of phosphorylase and phosphatase
many glycoside hydrolases.18,24,25
D-Glucal is utilized by a-glucosi-
dase from Candida tropicalis and b-glucosidase from
sweet almond.18 Disaccharide analogues of maltose (maltal) and
cellobiose (cellobial) in which
ing-end’ -glucose moiety are hydrated by, respectively, sweet po-
tato b-amylase19 and exo- as well as endo-type cellulases.26 Work
of Klein et al. on
-glucan phosphorylase27 and glycogen phos-
phorylase28 provided a first case for reaction of a phosphorylase
with -glucal. Unlike the hydrolases that promoted double bond
hydration only, the phosphorylase catalyzed transfer of the 2-
deoxy-glucosyl-moiety generated from -glucal upon protonation
D-glucal replaces the original ‘reduc-
D
a
D
D
at C-2. Depending on relative abundance of acceptor substrates
present in the reaction, transfer occurred preferentially to
phosphate or
a
-(1?4)-
D
-glucosyl oligo/polysaccharide.27,28 The
activities. The phosphatase hydrolyzed
a-D-glucose 1-phosphate
exocyclic enitol heptenitol (2,6-anhydro-1-deoxy-
D
-gluco-hept-1-
into -glucose and phosphate. It was important to confirm there-
D
enitol) was applied in the study of different glycoside hydrolases
and glycogen phosphorylase. It is particularly relevant for this
work that the phosphorylase catalyzed glycosyl transfer to phos-
phate, forming heptulose 2-phosphate as product.29
fore that the purified CuCPase used herein was absolutely devoid
of such phosphatase activity.
2.2. Identification of
D-glucal as CuCPase substrate for 2-deoxy-
Mechanistic proposals for enzymatic utilization of glycals share
a common initial step that is chemical ‘activation’ of the glycal by
protonation at C-2. Analysis of stereospecifity of this protonation
event has therefore been an important component of the mechanis-
D
-glucosyl transfer to phosphate
When offered
D-glucal (50 mM) and phosphate (50 mM) as sub-
strates, CuCPase (5
l
M) catalyzed formation of a single main prod-
tic characterization of different enzymes, including
a
- and b-gluco-
uct that eluted in analytical high performance anion exchange
chromatography with pulsed amperometric detection (HPAED-
PAD) at a position comparable to that of elution of an authentic
sidases18, exo- -glucanases,30 and glycogen phosphorylase.27 To
a
examine the stereochemical course of the transformation, reactions
were performed in D2O and their progress analyzed by NMR spec-
troscopy. The same methodology was applied in this work to mon-
a-D
-glucose 1-phosphate standard. We determined that product
formation took place at the expense of -glucal (HPAED-PAD).
D
itor CuCPase-catalyzed 2-deoxy-
We show that deuterium from solvent is incorporated stereospecif-
ically in equatorial position of 1, revealing that proton attack occurs
on the re side of C-2, that is from below the ring of D-glucal.
D
-glucosyl-transfer to phosphate.
NMR analysis from a reaction mixture in which about 20% of the
original substrate had been converted identified the product to
be a phosphorylated sugar. Results depicted in Figure 2a establish
identity of the phospho-sugar product as 1. Presence of the phos-
phate group was indicated by a 31P signal at 2.4 ppm, which
3
showed a JP-H coupling in 1H/31P-HSQC to proton H-1 with a cou-
2. Results and discussion
3
pling constant of 7.6 Hz. The JH-H coupling of 3.0 Hz between H-1
and H2a pointed to
a-anomeric configuration at C-1. In Figure 2a
2.1. Single-step purification of recombinant CuCPase
the TOCSY trace of protons H-1 is shown, which indicates all pro-
tons in the respective spin system. Several scalar couplings can
be deduced from additionally shown separated proton signals of
1. NMR spectroscopic data are summarized in Supplementary
Table 1.
Using the procedures described under 3.2 enzyme preparation,
we obtained CuCPase in a highly purified form. The isolated en-
zyme migrated as single protein band in SDS PAGE (Supplementary
Fig. 1), and its specific activity of 14.4 U/mg was in useful agree-
ment with the value of 19.2 U/mg reported for the same His-tagged
Figure 2. NMR spectroscopy based identification of 2-deoxy-a-D-glucose 1-phosphate (1) (panel a) and 2-deoxy-a-D-[(2e)-D]-glucose 1-phosphate (4) (panel b). Shown are
the 2D TOCSY traces of protons H-1 in f2-dimension. Additionally, the separated and not overlapped 1H NMR signals are attached to the axes, respectively. Indicative shifts
and coupling constants are listed in Supplementary Table 1.

P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
227
NMR analysis revealed that 2-deoxy-
D
-glucose (2) was a minor
HPAED-PAD analysis to consist of 2 next to 2-deoxy-b-
syl-(1?4)-
nated from hydrolysis of the trace amounts of 3 formed under
these conditions. Direct hydration of -glucal by CuCPase is there-
fore supported. The hydration product is assumed to be 2-deoxy-
D
-gluco-
additional product of the conversion of -glucal that was formed in
D
D-glucose (3). Product 2 is quite unlikely to have origi-
low abundance (62%) relative to 1. Careful controls performed un-
der otherwise identical reaction conditions in the absence of en-
D
zyme showed that
D
-glucal was stable over the entire timespan
a-
of the experiment, eliminating the possibility that spontaneous,
non-enzymatic reactions are responsible for the observed product
pattern. These results indicate that CuCPase promotes conversion
D
-glucose, which then undergoes mutarotation in solution, yielding
the a- and b-anomers of 2 in the reaction mixture, as observed in
the NMR analysis.
of
-glucal serves as donor substrate for phosphorolysis. From time-
course data presented in Figure 3 we estimated that the reactivity
of -glucal as substrate for phosphorolysis by CuCPase is roughly
D-glucal via 2-deoxy-D-glucosyl transfer to phosphate whereby
D
2.3. Enzymatic synthesis of 1 from D-glucal
D
Figure 3a shows time courses of conversion of 50 mM D-glucal
three orders of magnitude lower than reactivity of the native donor
substrate cellobiose used under otherwise identical conditions
(50 mM of each cellobiose and phosphate). A plausible explanation
for formation of 2 during phosphorolysis of D-glucal is that slow
non-enzymatic hydrolysis of the relatively unstable 1 takes place
in the presence of the same molar concentration of phosphate
and a fivefold molar excess of phosphate. Reaction rate was en-
hanced (twofold) by increasing the initial phosphate concentration
from 50 mM to 250 mM. The final concentration of 1 obtained in
the enzymatic conversion was doubled from 10 mM to around
20 mM by the same increase in phosphate concentration, indicat-
as a secondary reaction of the overall transformation. The alterna-
tive possibility is that CuCPase catalyzes direct hydration of
D
-glu-
ing a mass action effect. The maximum conversion of D-glucal un-
cal at a rate much slower than the phosphorolysis rate. In
experiments described later where CuCPase was incubated with
der the conditions used was around 40%. Despite this rather
modest yield, single-step production by CuCPase appears to be
the currently most convenient synthesis of 1. We will show later
D
-glucal and D-glucose in the absence of phosphate, the enzyme
synthesized tiny amounts of a product that was shown by
under 2.5 enzymatic reaction with
D-glucal in the presence of
D-
glucose and phosphate, that the presence of
D
-glucose further stim-
ulates catalytic formation of 1, making the synthesis even more
efficient.
60
a
2.4. Stereochemical course of enzymatic reaction with
D-glucal
50
40
30
20
10
0
A plausible mechanism of utilization of -glucal for 2-deoxy-D-
D
glucosyl transfer to phosphate involves substrate activation by
protonation at C-2 prior to or concomitant with nucleophilic attack
of phosphate oxygen on C-1. To probe the stereoselectivity of the
protonation event, we performed enzymatic reaction in D2O using
50 mM of each D-glucal and phosphate, and analyzed with NMR
the incorporation of deuterium in the resulting 1 obtained after
24 h of incubation. The obtained NMR spectrum identified the
formed product as 2-deoxy-a-D-[(2e)-D]-glucose 1-phosphate (4).
This compound has the same anomeric configuration as 1 and
shows a 31P signal at 2.5 ppm. It shows the proton signal of H-2a,
while no signal of H-2e is detectable (Fig. 2b). Concomitant the sig-
nal of H-1 and H-3 show only scalar couplings to H-2a as well as to
0
5
10
15
20
25
30
35
1
Time [h]
H-2a and H-4, respectively. Furthermore, C-2 has a typical JC-D
coupling (data not shown). NMR spectroscopic 1H and 13C data of
4 are summarized in Supplementary Table 1.
60
50
40
30
20
10
0
b
Our results indicate that protonation of D-glucal by CuCPase is
highly stereoselective. They also suggest a stereochemical course
of the protonation where the proton is delivered from below the
six-membered ring of substrate in
a re side attack on C-2
(Scheme 2B). We will discuss later with the help of molecular
docking that the most plausible mechanistic scenario is that en-
zyme-bound phosphate serves as proton donor in the reaction.
2.5. Enzymatic reaction with
D-glucal and D-glucose in the
absence and presence of phosphate
Kitaoka et al. have shown exploitation of CgCPase-catalyzed 2-
deoxy-b- -glucosyl transfer from -glucal to different acceptors
-glucose, -xylose, -mannose, and 2-deoxy- -arabino-hexose)
for synthesis of rare disaccharide analogs of cellobiose.16 We used
CuCPase to perform the same synthesis using -glucal and -glu-
D
D
(D
D
D
D
0
5
10
15
20
25
30
35
D
D
Time [h]
cose (50 mM each), but obtained poor yields (64%). Analysis by
HPAED-PAD of a sample taken after 30 h of incubation confirmed
formation of a disaccharide product that eluted at a position simi-
lar, yet not identical to that of elution of authentic cellobiose. How-
ever, this product was present in a concentration of only about
Figure 3. Time-courses for 2-deoxy-
-glucose (2) (x) production from 50 mM
phosphate (a) and 50 mM phosphate (b). The enzyme concentration was 5
Lines show trend of the data.
a-
D
-glucose 1-phosphate (1) (s) and 2-deoxy-
D
D-glucal (d) in the presence of 250 mM
lM.

228
P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
a
+
+
OH
OH
OH
O
O
HO
O
δ+
HO
O
H
O
H
HO
δ-
O
H
O
O
O
-
P
O
HO
P
HO
O
HO
O
O
HO
P
O
+
OD
OD
+
b
OD
O
O
DO
DO
O
δ+
D
O
D
O
DO
δ-
D
O
D
D
O
O
O
P
-
O
O
DO
P
DO
DO
O
O
DO
P
O
Scheme 2. Reaction course for 2-deoxy-D-glucosyl transfer from D-glucal to phosphate catalyzed by CuCPase in H2O (a) and D2O (b). In D2O, all acidic protons are assumed to
be exchanged by deuterium.
2 mM (calibrated as cellobiose). We also noticed formation of 2,
presumably resulting from hydration of -glucal (see Fig. 4f).
the time courses in Figure 4 (panels a–e) at 10 h and above, the
reaction rate became stalled and the concentrations of substrates
and products remained constant over time. Therefore, this sug-
gested that an apparent equilibrium had been reached in each of
the reactions. The results in Figure 4 are interesting as they hint
D
The reaction was then repeated in the presence of phosphate.
Interestingly and at first unexpectedly, disaccharide synthesis
was substantially enhanced under these conditions as compared
to the control lacking phosphate. Two main products were
formed: a disaccharide whose identity was revealed by NMR as
3, and 1, re-confirmed by the NMR data. The molar ratio of syn-
thesis and phosphorolysis products was dependent on the initial
phosphate concentration used; it decreased as the phosphate
concentration was raised (see Fig. 4). NMR data for the disaccha-
a catalytic role of phosphate and
-glucal into 3 (panels d and e compared to panel f) and 1 (panel
a compared to Fig. 3), respectively. A plausible scenario is that qua-
ternary complex between CuCPase, phosphate, -glucal, and -glu-
cose shows substantially enhanced activity as compared to the
corresponding ternary complexes (enzyme/ -glucal/ -glucose; en-
zyme/ -glucal/phosphate). Preferred direction of reaction, 2-
deoxy- -glucosyl transfer to -glucose or phosphate, is determined
by prevalence of acceptor substrate for synthesis or phosphoroly-
sis. Reaction of CuCPase with -glucal therefore appears to occur
D-glucose during conversion of
D
D
D
D
D
ride product identify the constituent 2-deoxy-b-
D-glucosyl group
D
from the typical large coupling constants between the axial pro-
tons present in each position, including C-1. The additional equa-
torial oriented proton on position C-2 shows concomitant smaller
3JH-H couplings and leads to a high field shift of the C-2 13C sig-
nal. The terminal saccharide is a glucose moiety, present in a typ-
D
D
D
under thermodynamic control (see later under 2.6 molecular dock-
ing studies).
ical 2:3 mixture of
has been determined by a weak NOE between the H-1 of the 2-
deoxy-b- -glucosyl moiety and H-4 of the terminal glucose. We
a
- and b-anomers. The interglycosidic linkage
Kinetic analysis of the data in Figure 4 reveals that CuCPase
activity was lowest for reaction with
D-glucal and D-glucose in
D
the absence of phosphate. An approximate turnover number (kapp)
isolated 3 from the product mixture, however, only in small
of 3.5 ꢀ 10ꢁ3 sꢁ1 was calculated from data in Figure 4f. Using
3
amounts that were not sufficient to determine JC-H couplings be-
D-glucal and phosphate in the absence of D-glucose (Fig. 3b), kapp
tween the two sugar units. In spite of that, all NMR spectroscopic
data (Supplementary Table 2) indicate the disaccharide to be a 3
and are in very good accordance with an earlier reported analysis
of this compound.16
was approximately 1.8 ꢀ 10ꢁ2 sꢁ1
.
Presence of phosphate
(150 mM) and
D-glucose next to D-glucal (Fig. 4f compared to
Fig. 3b) resulted in a roughly 15-fold increase in kapp to a value
of 0.27 sꢁ1. At 1.00 mM phosphate, kapp was still 0.2 sꢁ1
Reaction of CPase with -glucal is of interest to be applied for
synthesis of 2-deoxy- -glucosyl saccharides, as already demon-
.
Figure 4 summarizes time-course data for conversion of
D-glu-
D
cal and -glucose (50 mM each) in the presence of different phos-
D
D
phate concentrations in the range 1.00–250 mM. A control lacking
phosphate is also shown (panel f). Each time course in Figure 4
strated by Kitaoka and co-workers. This work adds to their pre-
vious study showing that 1 can also be synthesized. Generally,
deoxy-sugars in which a particular hydroxy group has been
substituted by hydrogen have proven to be valuable tools in
the biochemical study of substrate binding recognition and catal-
ysis by carbohydrate-active enzymes.32,33 The OH?H replace-
ment can be viewed as chemical mutagenesis of substrate,
resulting in selective removal of hydrogen-bonding interactions
that the original hydroxy group had with the enzyme. Consider-
(panels a–e) was characterized by relatively fast drop of
concentration in the first 5 h of reaction. Depending on the phos-
phate concentration used, -glucose was consumed at roughly
the same rate (1 and 5 mM phosphate) or at a rate much lower
(P50 mM phosphate) than the -glucal utilization rate. The prod-
D-glucal
D
D
ucts 3 and 1 were formed according to mass balance. At low phos-
phate concentration, 3 was essentially the only product derived
from
D
-glucal. At phosphate concentrations of 50 mM and greater,
ing the central physiological role of
the interface of cellular catabolism and anabolism, 1 presents
an interesting analog of -glucose 1-phosphate that could be
used in the examination of various enzymes utilizing -glucose
1-phosphate as their native substrate (e.g., phosphoglucomutase,
a-D-glucose 1-phosphate at
the distribution of transfer products varied in dependence of the
phosphate concentration. While at 50 mM phosphate the products
were formed in almost identical amounts, use of 250 mM phos-
phate rendered 1 as the predominant reaction product. In each of
a
-D
a-D

P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
229
60
50
40
30
20
10
0
a
60
50
40
30
20
10
0
d
0
0
0
5
5
5
10
10
10
15
20
20
20
25
25
25
30
30
30
35
35
35
0
5
10
15
20
25
30
35
Time [h]
Time [h]
60
50
40
30
20
10
0
60
50
40
30
20
10
0
b
e
0
5
10
15
20
25
30
35
15
Time [h]
Time [h]
c
60
50
40
30
20
10
0
f
5
4
3
2
1
0
60
50
40
30
20
10
0
15
0
5
10
15
Time [h]
20
25
30
35
Time [h]
Figure 4. Time-courses for 2-deoxy-b-
glucose (j) in the presence of 250 mM phosphate (a), 150 mM phosphate (b), 50 mM phosphate (c), 5 mM phosphate (d) and 1 mM phosphate (e), respectively. (f) shows the
time-course of 2-deoxy-b- -glucosyl-(1?4)- -glucose (2) (x) production from 50 mM -glucal (d) and 50 mM -glucose (j) in the absence of
-glucose (3) (4) and 2-deoxy-
phosphate. Lines show trend of the data.
D-glucosyl-(1?4)-D-glucose (3) (4) and 2-deoxy-a-D-glucose 1-phosphate (1) (s) production from 50 mM D-glucal (d) and 50 mM D-
D
D
D
D
D
NDP-glucose pyrophosphorylase, phosphatase, phosphorylase).
However, applying purely chemical approaches, synthesis of 1
has turned out to be remarkably difficult due to compound
instability.34–38 Percival and Withers used an enzymatic ap-
proach that involved five enzymes and proceeded via 2-deoxy-
and starch were used. Applying maltotetraose, authors reported
synthesis of 1 in two steps: a 2-deoxy- -glucosyl polysaccharide
having an average degree of polymerization of 20 was prepared
first and separated from the reaction mixture; polysaccharide
was then used for phosphorolysis, giving 1 in a yield of 50%.
An issue of using starch, which is cheaper than maltotetraose,
D
D
-glucose 6-phosphate and 2-deoxy-UDP-
and co-workers exploited reaction of glycogen phosphorylase
with
-glucal.40 The enzyme used is special in that its activity
is dependent on an -1,4-oligo- -glucoside primer. Maltotetra-
ose, which is the smallest primer utilized by the phosphorylase,
D
-glucose.39 Thiem
was formation of
a-D-glucose 1-phosphate next to 1 and diffi-
D
culty of separation of the two sugar phosphates. Therefore, syn-
thesis of 1 using CuCPase under conditions as in Figure 4 (panels
a and b) is promising.
a
D

230
P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
2.6. Molecular docking studies and proposal for the enzymatic
reaction mechanism
The crystal structure of CgCPase (PDB-entry 2cqt) in complex
with phosphate and
D-glucose (bound at subsite +1) was used for
anenergy-minimized dockingexperiment inwhich
D
-glucalwas rig-
idly placed into subsite ꢁ1 of the enzyme. Inherent ambiguity of the
used docking procedure notwithstanding, predicted accommoda-
tion of
D
-glucal at the catalytic subsite ꢁ1 of CgCPase appears to be
highly plausible. Compared to the PDB structure, no conformational
changes were observed in the docked structure. Figure 1b shows the
resulting enzyme–ligand interaction map. The docking provides
useful insight into formation of and possible reactive group arrange-
Scheme 4. Proposed kinetic mechanism for reaction of CuCPase with
CuCPase; D: -glucal; PH: phosphate, protonated; A: -glucose; DG: transient 2-
deoxyglucosyl oxocarbenium ion (2); DG-P: 2-deoxy- -glucose 1-phosphate (1);
DG-A: 2-deoxy-b- -glucose-(14)- -glucose (3); P: deprotonated phosphate.
[E.DG.P.A] is the quaternary transition state complex. The parameter is an
D-glucal. E:
D
D
a-D
ment in a quaternary complex, enzyme/phosphate/
cose, that might be catalytically competent. -Glucal is positioned
through a network of hydrogen bonds that involves protein residues
(D368; W488, D490; H666), but also phosphate and -glucose.
D-glucal/D-glu-
D
D
D
a
interaction coefficient describing the effect of binding of one substrate to the
enzyme on the binding of the respective other substrate.
D
D-
Glucal makes contacts from each of its hydroxy groups, whereby
the 4-OH and 6-OH show strong interactions with, respectively,
phosphate and the main-chain NH of D490 in particular. Both phos-
ble explanation for switch from thermodynamic to kinetic control
in two otherwise similar phosphorylase-catalyzed transformations
of
number for reaction with the natural donor substrate
1-phosphate, glycogen phosphorylase (relative kcat: 0.1–0.4)27 is
by far (P100-fold) more reactive toward -glucal than CuCPase
D-glucal could be that, when compared on the basis of turnover
phate and
contribute to binding affinity of
Figure 1b supports a catalytic mechanism for reaction of
CuCPase with -glucal. The proposed mechanism is shown in
Scheme 3. -Glucal is accommodated in a position that allows
D
-glucose (indirectly via D490) are therefore expected to
a-D-glucose
D
-glucal in the quaternary complex.
D
D
(relative kcat: 0.001). Mechanistic differences between the two phos-
phorylases, reflected in a different stereochemical course of the
reaction catalyzed, could also be responsible. Glycogen phosphory-
lase is a retaining enzyme that promotes an axial-to-axial substitu-
tion at C-1 of the transferred glycosyl residue. Unlike CuCPase that is
D
the bound phosphate molecule to serve as proton donor for stereo-
specific re-face protonation at C-2. From previous studies of invert-
ing glycoside hydrolases and glycosyltransferases,41–43 the
subsequent glycosyl transfer is expected to proceed through an
oxocarbenium ion-like transition state. Partial negative charges
at the reactive O atoms of phosphate and D-glucose will stabilize
this transition state. The ionized side chain of D490 will provide
general base catalytic assistance to the reaction through partial
reactive with
acceptor, glycogen phosphorylase must form a quaternary complex
of enzyme, phosphate, -glucal, and oligosaccharide acceptor
D-glucal and phosphate in the absence of sugar
D
(primer) to become detectably active. Arrangement of reactive
groups in this complex may be such that phosphate, which is
thought to serve a proton donor function in the reaction with
deprotonation of the glucose 4-OH. Regioselective 2-deoxy-
D-glu-
cosyl transfer to -glucose is therefore explained. The proposed
D
D
-glucal analogous to that proposed for CuCPase,27 cannot directly
role of D490 during conversion of D-glucal is analogous to its role
in the natural reaction when cellobiose is synthesized.2
participate as nucleophilic acceptor of the transferred 2-deoxy-
glucosyl moiety. Initial formation of oligosaccharide elongated by
2-deoxy- -glucosyl residues may thus be a necessity of the mecha-
D-
A kinetic mechanism for reaction of CuCPase with D-glucal is
D
proposed in Scheme 4. ‘Activation’ of donor substrate to formally
generate a 2-deoxy- -glucosyl cation is thought to be completely
rate-limiting and therefore, 2-deoxy- -glucosyl transfer between
phosphate and -glucose comes to equilibrium at all times during
nismutilized. Itis interestingthatglycogenphosphorylasepromotes
conversion of heptenitol to heptulose-2-phosphate in the absence of
primer molecule,29 using a reaction course highly similar to that of
D
D
D
direct 1 formation by CuCPase from D-glucal and phosphate.
the reaction. The thermodynamic equilibrium constant for conver-
sion of 2-deoxy-cellobiose into 1 at given pH (Keq) is defined in
kinetic terms as the ratio of third-order rate constants in direction
of phosphorolysis and synthesis (see Scheme 4). It is interesting that
glycogen phosphorylase differs from CuCPase in that its reaction
3. Experimental
3.1. Materials
with
D-glucal is kinetically controlled, where initial 2-deoxy-
D
-glucosyl transfer occurs exclusively on the carbohydrate acceptor
Unless otherwise stated, all chemicals used have been described
elsewhere.44
-Glucal was purchased from Sigma, 2-deoxy-
glucose (2) from Roth.
substrate (e.g., maltotetraose).27,40 Compound 1 is formed only in a
subsequent step, as required by reaction thermodynamics. A possi-
D
a-D-
Asp-490
Asp-490
Asp-490
+
+
OH
δ-
-
O
O
O
O
OH
O
HO
O
δ+
HO
O
OH
HO
OH
H
H
O
H
OH
O
δ-
O
H
OH
HO
HO
OH
OH
O
HO
O
O
O
O
H
O
δ+
O
O
HO
O
H
δ-
O
O
OH
OH
HO
OH
-
O
HO
P
O
HO
P
HO
P
O
O
O
Scheme 3. Proposed catalytic mechanism for reaction of CuCPase with D-glucal, D-glucose and phosphate.

P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
231
3.2. Enzyme preparation
a linear gradient from 100 mM NaOAc to 400 mM NaOAc applied
within 15 min in an isocratic background of 100 mM NaOH. The
column was washed 10 min with 52 mM NaOH. Under the condi-
tions applied, the following retention times were obtained:
Recombinant Cellulomonas uda CPase (CuCPase) was prepared
by a modification of a procedure reported elsewhere.31 Details of
the protocol were as follows. Escherichia coli BL21-Gold(DE3)
expression cells harboring the pQE30-vector encoding CuCPase
were cultivated in 1-L baffled shaken flasks at 37 °C and 110 rpm
using LB-media and 115 mg/L ampicillin. When OD600 reached
0.9–1.0, temperature was decreased to 22 °C and gene expression
was induced with 0.1 mM IPTG for 20 h. Cells were harvested by
centrifugation at 4 °C and 5000 rpm for 30 min in a Sorvall RC-5B
Refrigerated Superspeed centrifuge. Resuspended cells were frozen
at ꢁ20 °C, thawed and the suspension was passed twice through a
French pressure cell press (Aminco) at 150 bar. Cell debris was re-
moved by centrifugation at 4 °C, 14,000 rpm for 45 min. The result-
ing supernatant was used for further enzyme purification after it
2.6 min for
D
-glucal, 5.8 min for 2, 7.9 for
D
-glucose, 11.2 min for
3, and 23.0 min for 1.
3.5. Reactions with
D
-glucal and phosphate
D
-Glucal was tested as a possible glycosyl donor in phosphorol-
ysis direction catalyzed by CuCPase. 5 M CuCPase were incubated
with 50 mM -glucal and different concentrations of phosphate
l
D
(50 mM, 250 mM) in 50 mM MES buffer, pH 6.6. The formation of
a new product for CuCPase, 1, was monitored by HPAED-PAD after
the reaction mixture was incubated at 30 °C, 550 rpm for up to
30 h.
was passed through a 1.2
l
m cellulose-acetate filter (Sartorius).
To determine the stereochemical course of the reaction, we per-
As the pQE30-vector encoding CuCPase has a His-tag fused to the
N-terminus of the enzyme, CuCPase was purified by using a Ni-
chelating Sepharose Fast Flow chromatography (His-Trap HP, GE
Healthcare) at 4 °C. All buffers were degassed and filtered using
formed a reaction using 50 mM
D-glucal, 50 mM phosphate, 50 mM
MES (pH 6.6), and 5 M CuCPase in 99% (v/v) D2O. The mixture was
l
incubated at 30 °C, 550 rpm for 24 h and the content of the mixture
was analyzed by NMR analysis (see 3.8).
0.45 lm cellulose-acetate filters (Sartorius). The column was equil-
ibrated with buffer A (20 mM Tris–HCl, 20 mM imidazole, 500 mM
NaCl, pH 7.4) at a flow rate of 5 mL/min (5 column volumes). The
crude cell extract was applied at a flow rate of 1 mL/min. After
washing with buffer A at a flow rate of 5 mL/min (5 column vol-
umes), fractions containing the enzyme (91-kDa band in SDS–
PAGE) were eluted with buffer E (20 mM Tris–HCl, 250 mM imid-
azole, 500 mM NaCl, pH 7.4) at a flow rate of 2 mL/min (6 column
volumes). Pooled fractions were concentrated and desalted using a
Vivaspin ultrafiltration tube (10 kDa cutoff; Vivascience) and again
concentrated to about 40 mg/mL in 50 mM MES buffer, pH 6.6.
Stock solutions of CuCPase were stored at ꢁ20 °C. The yield of pure
protein from one liter culture was about 50 mg. Purification was
monitored by SDS–PAGE and the measurement of the specific
activity (see Supplementary Fig. 1). To further check the prepara-
tion’s purity and specifically the absence of a phosphatase related
3.6. Reactions with
absence of phosphate
D-glucal and D-glucose in presence and
To test whether CuCPase still utilizes
and -glucose is present, 5 M CuCPase were incubated with
50 mM -glucal, 50 mM -glucose, and different concentrations
D-glucal when phosphate
D
l
D
D
of phosphate (1 mM, 5 mM, 50 mM, 150 mM, 250 mM) in 50 mM
MES buffer, pH 6.6, at 30 °C for 30 h. The reaction mixture was ana-
lyzed by HPAED-PAD and the synthesis products were identified by
NMR-analysis as 3 and 1. To test the role of phosphate in the reac-
tion, CuCPase was incubated with only
(without phosphate). 5 M CuCPase were incubated with 50 mM
-glucal and 50 mM -glucose in 50 mM MES buffer, pH 6.6, at
D-glucal and D-glucose
l
D
D
30 °C and after 24 h the reaction was stopped and the reaction
mixture was analyzed by HPAED-PAD.
impurity, CuCPase (1 lM) was incubated with 50 mM a-D-glucose
1-phosphate in 50 mM MES buffer, pH 6.6 for 24 h and the reaction
mixture was analyzed by high performance anion exchange chro-
matography with pulsed amperometric detection (HPAED-PAD)
3.7. Energy-minimized molecular docking
Yasara V 11.11.21 was used for the determination of energy-
minimized structure and the enzyme–ligand docking. PyMOL
(http://pymol.sourceforge.net) was used for visualization. The X-
(see below). No hydrolytic activity of
a-D-glucose 1-phosphate into
D
-glucose and phosphate could be detected with purified CuCPase
prepared with the protocol just described.
ray crystal structure of CgCPase having a D-glucose and phosphate
bound in subsite +1 and phosphate binding site, respectively (PDB-
entry 2cqt) was used as macromolecule in a rigid molecular dock-
3.3. Assays
ing experiment that employed
D-glucal as the ligand. A search
Enzyme activity in direction of phosphorolysis of cellobiose was
determined at 30 °C using a continuous coupled enzymatic assay
with phosphoglucomutase and glucose-6-phosphate dehydroge-
nase.14 Protein concentration was measured with the BioRad
dye-binding method referenced against BSA.
space of 15 ꢀ 12 ꢀ 12 Å was used, spanning the enzyme’s active
site. The docking algorithm resulted in 4 binding modes with com-
parable energies. From these, the most accurate model was se-
lected on the basis of the binding mode of N-acetylglucosamine
in subsite ꢁ1 of VpChPase.6
3.4. Analytical methods
3.8. NMR measurements
All synthesis products were analyzed by high performance an-
ion exchange chromatography with HPAED-PAD. The analysis
was performed using a Dionex BioLC system equipped with a Carb-
oPac PA10 column (4 ꢀ 250 mm) and an Amino Trap guard column
NMR spectroscopic measurements were made from crude or
partially purified reaction mixtures. Lyophilized samples were dis-
solved in D2O (99.95%, 0.7 mL) and transferred into 5 mm NMR
sample tubes (Promochem, Wesel, Germany). All spectra were re-
corded on a Bruker DRX-400 AVANCE spectrometer (Bruker, Rhein-
stetten, Germany) at 400.13 MHz (1H), 100.61 MHz (13C), and
161.98 MHz (31P) using the Bruker Topspin 1.3 software. Measure-
ments temperature was 300 0.1 K. For 1D-spectra, 32k data
points were acquired using a relaxation delay of 1.0 s and appropri-
ate number of scans. After zero filling to 64k data points and Fou-
(4 ꢀ 50 mm) thermostated at 30 °C.
D-Glucal, 2, D-glucose, 3, and 1
were detected with an ED50A electrochemical detector using a
gold working electrode and a silver/silver chloride reference elec-
trode by applying the predefined waveform for carbohydrates. Elu-
tion was carried out at a flow rate of 1 mL/min with the following
method: isocratic flow of 52 mM NaOH for 20 min, followed by an
isocratic flow of 100 mM NaOH and 100 mM NaOAc for 5 min, and
rier transformation, spectra with
a
range of 7200 Hz (1H),

232
P. Wildberger et al. / Carbohydrate Research 356 (2012) 224–232
20,000 Hz (13C), and 16,200 Hz (31P) were obtained, respectively.
To determine the 2D COSY, TOCSY, NOESY, HMQC, and HMBC as
well as 1H/31P-HSQC spectra, 128 experiments with 8 scans and
1024 data points each were performed. After linear forward predic-
tion to 256 data points in the f2-dimension and appropriate sinu-
soidal multiplication in both dimensions, the data were Fourier
transformed to obtain 2D-spectra with ranges of 4000 Hz (1H),
20,000 Hz (13C), and 8100 Hz (31P). Chemical 1H and 13C shifts were
referenced to external acetone (dH 2.225 ppm; dC 31.45 ppm) and
phosphorus shifts were referenced to external aq 85% phosphoric
acid (31P: dP 0.00 ppm).
11. Sih, C. J.; McBee, R. H. Proc. Montana Acad. Sci. 1955, 15, 21–22.
12. Sih, C. J.; McBee, R. H. Bacteriol. Proc. 1955, 126.
13. Alexander, J. K. Arch. Biochem. Biophys. 1968, 123, 240–246.
14. Nidetzky, B.; Eis, C.; Albert, M. Biochem. J. 2000, 351, 649–659.
15. De Groeve, M. R. M.; De Baere, M.; Hoflack, L.; Desmet, T.; Vandamme, E. J.;
Soetaert, W. Protein Eng. Des. Sel. 2009, 22, 393–399.
16. Kitaoka, M.; Nomura, S.; Yoshida, M.; Hayashi, K. Carbohydr. Res. 2006, 341,
545–549.
17. Fritz, H.; Lehmann, J.; Schlesselmann, P. Carbohydr. Res. 1983, 113, 71–92.
18. Hehre, E. J.; Genghof, D. S.; Sternlicht, H.; Brewer, C. F. Biochemistry 1977, 16,
1780–1787.
19. Hehre, E. J.; Kitahata, S.; Brewer, C. F. J. Biol. Chem. 1986, 261, 2147–21453.
20. Kasumi, T.; Tsumuraya, Y.; Brewer, C. F.; Kerstershilderson, H.; Claeyssens, M.;
Hehre, E. J. Biochemistry 1987, 26, 3010–3016.
21. Kitahata, S.; Brewer, C. F.; Genghof, D. S.; Sawai, T.; Hehre, E. J. J. Biol. Chem.
1981, 256, 6017–6026.
Acknowledgements
22. Lai, E. C. K.; Morris, S. A.; Street, I. P.; Withers, S. G. Bioorg. Med. Chem. 1996, 4,
1929–1937.
23. Lehmann, J.; Zieger, B. Carbohydr. Res. 1977, 58, 73–78.
24. Legler, G.; Roeser, K. R.; Illig, H. K. Eur. J. Biochem. 1979, 101, 85–92.
25. Lehmann, J.; Schroter, E. Carbohydr. Res. 1972, 23, 359–368.
26. Kanda, T.; Brewer, C. F.; Okada, G.; Hehre, E. J. Biochemistry 1986, 25, 1159–
1165.
Financial support from the Austrian Science Funds FWF (project
L586-B03) is gratefully acknowledged. We thank Susanne Felsinger
for kind help with measurement of NMR spectra.
27. Klein, H. W.; Palm, D.; Helmreich, E. J. Biochemistry 1982, 21, 6675–6684.
28. Palm, D.; Klein, H. W.; Schinzel, R.; Buehner, M.; Helmreich, E. J. Biochemistry
1990, 29, 1099–1107.
Supplementary data
29. Klein, H. W.; Im, M. J.; Palm, D.; Helmreich, E. J. M. Biochemistry 1984, 23, 5853–
5861.
30. Chiba, S.; Brewer, C. F.; Okada, G.; Matsui, H.; Hehre, E. J. Biochemistry 1988, 27,
1464–1469.
31. Nidetzky, B.; Griessler, R.; Schwarz, A.; Splechtna, B. J. Mol. Catal. B: Enzym.
2004, 29, 241–248.
32. Niggemann, J.; Lindhorst, T. K.; Walfort, M.; Laupichler, L.; Sajus, H.; Thiem, J.
Carbohydr. Res. 1993, 246, 173–183.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.carres.2012.04.
003.
References
33. Nikaido, H.; Hassid, W. Z. Adv. Carbohydr. Chem. Biochem. 1971, 26, 351–483.
34. McDonald, D. L. J. Org. Chem. 1962, 27, 1107–1109.
35. Sabesan, S.; Neira, S. Carbohydr. Res. 1992, 223, 169–185.
36. Schmidt, R. R.; Stumpp, M. Liebigs Ann. Chem. 1984, 680–691.
37. Westerduin, P.; Veeneman, G. H.; Vandermarel, G. A.; Vanboom, J. H.
Tetrahedron Lett. 1986, 27, 6271–6274.
38. Withers, S. G.; Percival, M. D.; Street, I. P. Carbohydr. Res. 1989, 187, 43–66.
39. Percival, M. D.; Withers, S. G. Can. J. Biochem. 1988, 66, 1970–1972.
40. Evers, B.; Mischnick, P.; Thiem, J. Carbohydr. Res. 1994, 262, 335–341.
41. Brüx, C.; Ben-David, A.; Shallom-Shezifi, D.; Leon, M.; Niefind, K.; Shoham, G.;
Shoham, Y.; Schomburg, D. J. Mol. Biol. 2006, 359, 97–109.
42. Vasella, A.; Davies, G. J.; Bohm, M. Curr. Opin. Chem. Biol. 2002, 6, 619–629.
43. Withers, S. G. Carbohydr. Polym. 2001, 44, 325–337.
44. Goedl, C.; Schwarz, A.; Minani, A.; Nidetzky, B. J. Biotechnol. 2007, 129, 77–86.
1. Alexander, J. K. J. Bacteriol. 1961, 81, 903–910.
2. Hidaka, M.; Kitaoka, M.; Hayashi, K.; Wakagi, T.; Shoun, H.; Fushinobu, S.
Biochem. J. 2006, 398, 37–43.
3. Stam, M. R.; Blanc, E.; Coutinho, P. M.; Henrissat, B. Carbohydr. Res. 2005, 340,
2728–2734.
4. Henrissat, B.; Davies, G. Curr. Opin. Struct. Biol. 1997, 7, 637–644.
5. McCarter, J. D.; Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885–892.
6. Hidaka, M.; Honda, Y.; Kitaoka, M.; Nirasawa, S.; Hayashi, K.; Wakagi, T.; Shoun,
H.; Fushinobu, S. Structure 2004, 12, 937–947.
7. Kitaoka, M.; Sasaki, T.; Taniguchi, H. J. Biochem. 1992, 112, 40–44.
8. Liu, A. M.; Tomita, H.; Li, H. B.; Miyaki, H.; Aoyagi, C.; Kaneko, S.; Hayashi, K. J.
Ferment. Bioeng. 1998, 85, 511–513.
9. Percy, A.; Ono, H.; Hayashi, K. Carbohydr. Res. 1998, 308, 423–429.
10. Alexander, J. K. J. Biol. Chem. 1968, 243, 2899–2904.