Binding pattern of intermediate UDP-4-keto-xylose to human UDP-xylose synthase: Synthesis and STD NMR of model keto-saccharides

Article abstract of DOI:10.1016/j.carres.2016.11.001

Human UDP-xylose synthase (hUXS1) exclusively converts UDP-glucuronic acid to UDP-xylose via intermediate UDP-4-keto-xylose (UDP-Xyl-4O). Synthesis of model compounds like methyl-4-keto-xylose (Me-Xyl-4O) is reported to investigate the binding pattern thereof to hUXS1. Hence, selective oxidation of the desired hydroxyl function required employment of protecting group chemistry. Solution behavior of synthesized keto-saccharides was studied without enzyme via¹H and¹³C NMR spectroscopy with respect to existent forms in deuterated potassium phosphate buffer. Keto-enol tautomerism was observed for all investigated keto-saccharides, while gem-diol hydrate forms were only observed for 4-keto-xylose derivatives. Saturation transfer difference (STD) NMR was used to study binding of synthesized keto-gylcosides to wild type hUXS1. Resulting epitope maps were correlated to earlier published molecular modeling studies of UDP-Xyl-4O. STD NMR results of Me-Xyl-4O are in good agreement with simulations of the intermediate UDP-Xyl-4O indicating a strong interaction of proton H3 with the enzyme, potentially caused by active site residue Ala⁷⁹. In contrast, pyranoside binding pattern studies of methyl uronic acids showed some differences compared to previously published STD NMR results of UDP-glycosides. In general, obtained results can contribute to a better understanding in binding of UDP-glycosides to other UXS enzyme family members, which have high structural similarities in the active site.

Full text of DOI:10.1016/j.carres.2016.11.001

Carbohydrate Research 437 (2017) 50e58
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Binding pattern of intermediate UDP-4-keto-xylose to human
UDP-xylose synthase: Synthesis and STD NMR of model
keto-saccharides
,
*
Claudia Puchner a, Thomas Eixelsberger b, Bernd Nidetzky b, Lothar Brecker a
a
€
University of Vienna, Institute of Organic Chemistry, Wahringerstrasse 38, A-1090 Vienna, Austria
b Graz University of Technology, Institute of Biotechnology and Biochemical Engineering, Petersgasse 12/1, A-8010 Graz, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Human UDP-xylose synthase (hUXS1) exclusively converts UDP-glucuronic acid to UDP-xylose via in-
termediate UDP-4-keto-xylose (UDP-Xyl-4O). Synthesis of model compounds like methyl-4-keto-xylose
(Me-Xyl-4O) is reported to investigate the binding pattern thereof to hUXS1. Hence, selective oxidation of
the desired hydroxyl function required employment of protecting group chemistry. Solution behavior of
synthesized keto-saccharides was studied without enzyme via 1H and 13C NMR spectroscopy with
respect to existent forms in deuterated potassium phosphate buffer. Keto-enol tautomerism was
observed for all investigated keto-saccharides, while gem-diol hydrate forms were only observed for 4-
keto-xylose derivatives. Saturation transfer difference (STD) NMR was used to study binding of syn-
thesized keto-gylcosides to wild type hUXS1. Resulting epitope maps were correlated to earlier published
molecular modeling studies of UDP-Xyl-4O. STD NMR results of Me-Xyl-4O are in good agreement with
simulations of the intermediate UDP-Xyl-4O indicating a strong interaction of proton H3 with the
enzyme, potentially caused by active site residue Ala79. In contrast, pyranoside binding pattern studies of
methyl uronic acids showed some differences compared to previously published STD NMR results of
UDP-glycosides. In general, obtained results can contribute to a better understanding in binding of UDP-
glycosides to other UXS enzyme family members, which have high structural similarities in the active
site.
Received 3 August 2016
Received in revised form
30 October 2016
Accepted 1 November 2016
Available online 5 November 2016
Keywords:
4-Keto-saccharides
Human UDP-xylose synthase
Saturation transfer difference NMR
Binding pattern
© 2016 Published by Elsevier Ltd.
1. Introduction
nicotinamide adenine dinucleotide (NADþ or NADPþ) as co-factor,
which gets regenerated after each catalytic cycle. Moreover, sugar
Nucleotide activated 4-keto-hexose derivatives are central pre-
cursors in the biosynthesis of various deoxygenated saccharides.
From these intermediates, subsequent combinations of different
enzymatic transformations like oxidation, reduction, epimerization
and isomerization provide a broad diversity of sugar structures.
Amongst others, seldom occurring carbohydrates like nucleoside
transformation is induced by NAD(P)þ dependent oxidation of the
C4 sugar hydroxyl group [3]. However, conversion of UDP-
a-D-
glucuronic acid (UDP-GlcUA) to UDP- -xylose (UDP-Xyl) is cata-
a-D
lyzed by the enzyme UDP-xylose synthase (UXS), which is found in
numerous organisms. In detail, this sugar conversion includes for-
mation of the intermediate UDP-4-keto-a-D-xylose (UDP-Xyl-4O)
diphosphate-6-deoxy-
a-
D-glucose (NDP-6-deoxy-Glc) and uridine
[4e7].
diphosphate- -apiose (UDP-Api) are formed [1,2]. Activated 4-
a
-
D
The human UXS (hUXS1) belongs to the extended SDR enzyme
family and contains six various active site residues, which promote
catalysis through heavily pyranoside ring distortion of enzyme
bound substrate. Thr118, Tyr147 and Lys151 are often found in the SDR
group, while Ser119, Glu120 and Arg277 are typical members of the
UXS enzyme class. As a result, the sugar ring passes through various
conformations during the catalytic event (Scheme 1) [8]. The UXS
induced conversion of UDP-GlcUA starts with oxidation of the C4
pyranosyl hydroxyl function using NADþ as co-factor. The resulting
keto-saccharides are often used as intermediates by the short-
chain dehydrogenase/reductase enzyme superfamily (SDR). This
group of biocatalysts has a significant role, inter alia, in carbohy-
drate and xenobiotic metabolism. In general, SDR enzymes require
* Corresponding author.
E-mail address: lothar.brecker@univie.ac.at (L. Brecker).
http://dx.doi.org/10.1016/j.carres.2016.11.001
0008-6215/© 2016 Published by Elsevier Ltd.

C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
51
Scheme 1. hUXS1 catalyzed conversion of substrate UDP-GlcUA to UDP-Xyl via intermediate UDP-Xyl-4O. The carboxylate group of UDP-GlcUA is brought into a nearly axial
position by active site residues promoting subsequent decarboxylation. Given scheme is related to published catalytic mechanism of hUXS1 [8].
unstable
b
-keto-acid intermediate decarboxylates and forms UDP-
part presumably does not lead to entire changes in pyranoside
binding pattern due to its small size [21]. Therefore, synthesis of
Xyl-4O, which finally gets reduced to UDP-Xyl by previously
generated NADH [8e12].
methyl-4-keto-
we disposed to prepare methyl-4-keto-
ester (Me-GlcUA-4O-Me) according to intermediate UDP-4-keto-
-glucuronic acid (UDP-GlcUA-4O). At this point, esterification of
D-xylose (Me-Xyl-4O) was accomplished. Further,
However, UDP-Xyl displays an essential part in formation of
mammalian extracellular matrix proteoglycans due to covalent
attachment of their glycosaminoglycan side chain to a protein core
via xylosyl moiety. Hence, decreased UXS activity leads to a
defective extracellular matrix and deformity of diverse tissues,
which can promote tumor progression and metastasis [7,8]. Thus,
studying and fully understanding the mechanism of UXS can
redound to develop various strategies that will reduce or prevent
metastasis. Furthermore, binding pattern studies of various car-
bohydrates to hUXS1 allow useful insights into the function of this
enzyme. Presented results can contribute to a better understanding
in binding of UDP-glycosides to other UXS enzyme family mem-
bers, which have high structural similarities in the active site [12].
For this purpose, saturation transfer difference (STD) NMR displays
a popular application to investigate biological questions, involving
molecular interactions between binding ligands and proteins. This
robust analytical method provides reliable information about
binding pattern of various substrates to proteins [13e20].
D-glucuronic acid methyl
a
-
D
UDP-GlcUA-4O model compound is necessary to prevent undesired
spontaneous decarboxylation. Last, we synthesized methyl-4-keto-
D-glucose (Me-Glc-4O) due to same glycoside stereochemistry.
Pentoses and Hexoses usually contain up to five different hy-
droxyl groups with varying reactivities. In general, even small
regioselective changes of a desired hydroxyl moiety often requires
implementation of protecting groups, which leads to multi-step
protocols in synthesis. Nevertheless, these functionalities can be
differentiated by their subtle reactivity differences and their rela-
tive orientation [22e24]. However, no regioselective discrimination
can be afforded in
tions of hydroxyl groups in this monosaccharide with respect to the
anomeric center. This circumstance led us to C4 epimer of -xylose
as suitable starting material for the synthesis of Me-Xyl-4O. Hence,
the presence of an axial OH group at C4 in -arabinose (1) enabled
D-xylo configuration, due to all equatorial posi-
D
L
In our previous work we reported STD NMR studies of some
UDP-carbohydrates, which are in good agreement with the crystal
structure of hUXS1 and docking studies of UDP-GlcUA [8,21].
Hence, we are strongly interested in synthesis of various UDP-Xyl-
4O model compounds and molecular interactions thereof with
hUXS1. Furthermore, STD NMR results of prepared 4-keto-sugars
are correlated to published MD simulations of UDP-Xyl-4O [8] to
offer insights into UDP-Xyl-4O intermediate binding pattern.
us to mask the vicinal diequatorial diol system at C2 and C3 via
cyclic protecting group. Thus, the C4 hydroxyl residue was oxidized
to obtain Me-Xyl-4O, which is outlined in Scheme 2.
Starting from inexpensive -arabinose (1), this compound was
L
converted quantitatively to its corresponding methyl glycoside by
treatment with an acidic resin (Amberite IR 120 H) in dry methanol
[25]. Crystallization in isopropanol yielded methyl-b-L-arapyrano-
side (2). Next, the 1,2-trans diol system at C2 and C3 of 2 was
blocked under acidic conditions with 1,2-butandione in methanol
to give butane 2,3-bisacetal 3 in good yields [26]. Then, hydroxyl
group at C4 of 3 was oxidized via Dess-Martin periodinane to obtain
ketone 4 [27]. Finally, the trans-acetal moiety of protected 4-oxo-
saccharide 4 was removed in aqueous trifluoroacetic acid to give
target compound 5 in excellent yields and good purity [26].
Synthesis of Me-Glc-4O (9) and Me-GlcUA-4O-Me (15) were
prepared following the reaction sequence shown in Scheme 3. Fully
2. Results and discussion
2.1. Preparation of methyl-4-keto-glycosides and uronic acids
The synthetic approach focused on preparation of three
different 4-keto-glycosides containing a small agylcone moiety,
which locks the anomeric center in desired
a-position. Hence, an
protected
D-glucose 6 was synthesized as starting material
easily accessible methyl group at C1 was introduced. This agylcone

52
C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
Scheme 2. Synthetic route of Me-Xyl-4O (5). Reagents and conditions: a) Amberlite IR 120 H, MeOH, reflux, 7 h, 99% (
CSA, MeOH, reflux, 20 h, 85%; d) DMP, DCM, rt, 24 h, 72%; e) aq. TFA, rt, 2 min, 99%.
b/a 3:1); b) i-PrOH, reflux, 20 min; c) 2,3-butandione, TMOF,
Scheme 3. Synthetic approach of Me-Glc-4O (9) and Me-GlcUA-4O-Me (15). Reagents and conditions: a) Et3SiH, TFA, DCM, rt, 3 h, 74%; b) DMP, DCM, rt, 24 h, 89%; c) NaBrO3,
Na2S2O4, EtOAc/H2O, rt, 22 h, 80% (C4/C3 ketone 4:1); d) TFA, MeOH, reflux, 4 h, 91%; e) TEMPO, NaOCl, NaHCO3, NaBr, TBAF, EtOAc/H2O, 0 ꢀC, 2.5 h, 68%; f) Pd/C, H2, MeOH, rt, 48 h,
91%; g) SOCl2, MeOH, 0 ꢀC, 4 h, 71%; h) DMP, DCM, rt, 24 h, 52%.
according to Matwiejuk and Thiem [28]. Next, 4,6-O-benzylidene
acetal of compound 6 was opened reductive via triethyl silane and
trifluoroacetic acid in dichloromethane to give desired 6-O-benzyl-
4-hydroxy saccharide 7 in good yield and high selectivity [29]. Next,
free alcohol of 7 was converted to its corresponding ketone 8 in
excellent yield [27]. Finally, removal of benzyl protecting groups
was achieved to receive target compound 9 using a modified pro-
tocol by Adinolfi et al. [30] In this experimental approach sodium
bromate and sodium dithionite were used in a two-phase system of
methyl-
acturonic acid (Me-GalUA) 20 [26,28]. In addition, we prepared 1-
phosphate- -glucuronic acid (1-P-GlcUA) 22 from 1-phosphate
glucose employing TEMPO-oxidation [32,34]. Hence, compound
13, 19 and 22 were studied via STD NMR. In particular, we were
interested in potential varying binding pattern [21], which may is
caused through stereo-chemical variations at C4 and presence of an
anomeric phosphate group, respectively. Because former STD NMR
studies of UDP-glycosides indicated changes in binding pattern,
which resulted from varying stereochemistry at C4 [21].
a-D-glucuronic acid (Me-GlcUA) 13 and methyl-a-D-gal-
a-D
ethyl acetate and water, which led to formation of an unstable a-
bromo ether. Subsequent hydrolysis of the former one gave final
substrate 9. However, NMR spectra of keto compound 9 also
2.2. Solution behavior of keto-saccharides
showed signals of methyl-3-keto- -glucose (Me-Glc-3O) 10. The
D
ratio of C4/C3 ketone was 4:1 after reaction work-up and changed
within some days to C4/C3 ketone 1:5.
In solution, keto function of UDP-Xyl-4O is primary present as
gem-diol hydrate, which can be determined by 13C NMR spectros-
copy. The keto group shows a13C signal at ca. 100 ppm, which is
characteristic for a hydrated keto form [35]. In addition, we also
observed keto-enol tautomerism of intermediate UDP-Xyl-4O.
Within 7 d noticeable deuterium incorporation at C3 was ach-
ieved in deuterated potassium phosphate buffer (Fig. 1). Decreasing
H3 signal intensity was obtained followed by changes in signal
shape of H2 due to deuterium insertion at C3. Further, singlet
appearance next to H5a possibly was provoked by keto-enol
tautomerism between C4 and C5. However, we definitely can
conclude that migration of the C4 ketone in UDP-Xyl-4O is a none
enzymatic catalyzed event since deuterium incorporation was
observed after mutant separation.
Last, we tried to prepare Me-GlcUA-4O-Me (15) starting from
fully protected sugar 6. First, benzylidene acetal cleavage of 6 fol-
lowed by TEMPO-Oxidation yielded 2,3-O-benzyl protected glu-
curonic acid 12 [31,32]. Then, sugar 12 was esterified via Schotten-
Baumann esterification to give compound 14 [33]. However, final
oxidation of 14 did not yield Me-GlcUA-4O-Me (15), instead furanon
16 was obtained. Mass and NMR analysis of compound 16 indicated
loss of the ester group and presence of a C4 keto function. In general,
oxidation of compound 14 via various oxidation protocols, e.g.
Swern, onlyled to decompositionproducts. Unfortunately, oxidation
of (2'R,3'R)-methyl-2,3-O-(2',3'-dimethoxybutane-2',3'-diyl)-a-D-
galacturonic acid methyl ester also did not yield the desired 4-keto-
ester [26,32]. Nevertheless, simple protecting groups cleavage of
glucuronic acid 12 and galacturonic acid 19 (Scheme 4) provided us
For this purpose, we wanted to examine existing forms of syn-
thesized oxo-saccharides in deuterated potassium phosphate

C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
53
Scheme 4. Synthesis of Me-GalUA (20) and 1-P-GlcUA (22). Reagents and conditions: a) 2,3-butandione, TMOF, CSA, MeOH, reflux, 20 h, 64%; b) TEMPO, NaOCl, NaHCO3, NaBr, TBAF,
EtOAc/H2O, 0 ꢀC, 2.5 h, 96%; c) aq. TFA, rt, 2 min, 98%; d) TEMPO, NaOCl, NaHCO3, NaBr, TBAF, EtOAc/H2O, 0 ꢀC, 2.5 h, 69%.
methyl-a-D-glucopyranoside and methyl-a-D-galactopyranoside.
However, this oxidation procedure only yielded a mixture of C3/C4
ketones in our case. We then investigated the solution behavior of
compound 10 in deuterated potassium phosphate buffer. Interest-
ingly, keto function of 10 migrated between C3/C2 and C2/C1
leading to decreasing signal intensities of corresponding protons.
Nevertheless, migration of keto function was clearly favored be-
tween C3 and C2 due to fast decreasing signal intensity of H2
compared to the anomeric proton. Within 36 h complete 1H-NMR
signal loss of proton H2 was observed, which resulted in a singlet
signal of H1.
In addition, no gem-diol hydrate form was received of ketone 10.
However, lacking hydrate form can lead to a potentially different
sugar ring conformation and absence of ketone at C4 likely leads to
modified binding to enzyme hUXS1 compared to compound 5.
2.3. STD NMR of keto-saccharides and uronic acids
Fig. 1. 1H-NMR spectra cut out of UDP-Xyl-4O. Mutante Y147 catalyzed conversion of
UDP-GlcUA to UDP-Xyl-4O. a) Spectrum of non deuterated UDP-Xyl-4O in D2O. This
spectrum was slightly shifted up field (ꢁ0.035 ppm) for a better overview. b) Deute-
rium insertion of H5a was caused by enzymatic transformation in deuterated phos-
phate buffer at pD 7.1. Slow none enzymatic catalyzed keto-enol tautomerism was
observed between C3 and C4 after 7 d, which led to deuterium incorporation at C3.
Hence, decreasing H3 signal intensity and changes in signal shape of H2 were obtained.
c) Spectrum of intermediate showed sophisticated deuterium insertion at C3 after one
month. In addition, arising singlet next to H5a potentially was caused by keto-enol
tautomerism between C4 and C5.
Binding pattern studies of synthesized keto-sugars Me-Xyl-4O
(5) and Me-Glc-3O (10) were performed to receive more insights
into UDP-Xyl-4O binding and substrate specificity. Presented
epitope maps of keto-saccharides display relative saturation
transfer of enzyme to investigated substrates and are given in Fig. 2.
STD NMR results of saccharide 5 demonstrated close contact of
H3 with groups on the protein compared to remaining pyranoside
ring protons. In general, binding pattern of ketone 5 was similar to
our previous published STD NMR results of methyl-a-D-xylopyr-
buffer via 1H and 13C NMR, which may be relevant for binding to
enzyme hUXS1. Thus, Me-Xyl-4O (5) was dissolved in buffer and
showed a complex 1H spectrum due to the presence of dimeric and
hydrate forms. Nevertheless, exclusive monomeric gem-diol hy-
drate form of 5 was obtained after standing at rt for 24 h. Constant
1H NMR spectra were measured within a month to determine po-
tential keto-enol tautomerism of ketone 5. All in all, we only
observed relatively slow deuterium insertion at C3 compared to
UDP-Xyl-4O.
anoside (Me-Xyl) [21]. Further, one proton of the C5 methylene
group showed a slightly higher STD effect. Unfortunately, deter-
mination of pro-R H5 and pro-S H5 in Me-Xyl-4O (5) and thus clear
assignment of STD signals were not feasible due to lacking nuclear
Overhauser effects between C5 methylene group and remaining
ring protons. However, based on our previous published STD
investigation [21], slightly higher STD response likely was achieved
On the other hand, investigation of Me-Glc-4O (9) showed fast
keto-enol tautomerism. To our surprise, Me-Glc-4O (9) rapidly
converted to Me-Glc-3O (10) shortly after deprotection reaction
work-up. In general, observation of 9 was given as an inseparable
mixture of C3/C4 ketone immediately after protecting group
cleavage. Within a few days, compound 10 mainly was present.
Several attempts like carefully heating up and/or further standing
at rt in solution did not cause isomerization to the C4 ketone 9.
Another synthetic option to receive 4-oxo-glucose 9 was in situ
bromolysis of sugar-bis-tributyltin oxide intermediates [36,37].
This approach offers regioselective C4 oxidation of unprotected
Fig. 2. Relative proton saturation of synthesized Me-Xyl-4O (5) and Me-Glc-3O (10)
are displayed as epitope maps and expressed in %. The most intensive STD response
was set to 100% followed by identification of remaining STD signals, which were
referenced to the former one. Protons close to water signal were not determined and
marked as “nd”.

54
C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
by the C5 equatorial proton of Me-Xyl-4O (5). In general, distinct
STD effects of H3 and H5 are supported by MD simulations of UDP-
Xyl-4O [8]. These MD simulations show strong hydrogen bond
formation between residue Ala79 and the C3 hydroxyl moiety and
moderate interaction of Asn176 with the C5 ring oxygen, respec-
tively. Hence, binding pattern studies of 5 point to a similar binding
mode as UDP-Xyl-4O. Thus, these findings indicate an important
role of the C4 keto function and its corresponding gem-diol hydrate
form, which optimally locates the 4-keto-sugar at the active site. In
addition, these results are supported by former mutant preparation
in which Tyr147 was replaced by Phe. This change in active site
resulted in release of keto-xylose intermediate [8]. Further, the
aglycone moiety of 5 also received high saturation. This presumably
was caused by the small size of the anomeric methyl group, which
has more conformational flexibility compared to a space-filling and
clearly hindered uridine-di-phosphate residue in UDP-saccharides.
Keto-saccharide Me-Glc-3O (10) presented a suboptimal sub-
strate to gain further knowledge into UDP-Xyl-4O intermediate
binding. Nevertheless, the binding behavior of ketone 10 was
studied. In particular, we were interested in a potential hUXS1
enzymatic influence concerning keto-enol tautomerism towards
desired keto function at C4. Hence, a series of alternating 1H and
STD NMR spectra was recorded to observe a feasible increase of Me-
Glc-4O 9 and thus a possible change in binding behavior. However,
recorded spectra indicated no appreciable generation of 9 or an
endiol intermediate. Further, pyranoside binding pattern of 10
displayed a distinct saturation transfer of H1 and H4. Strong STD
response of proton H4 may be explained through interaction with
active site residues Tyr147 and Thr118. Nevertheless, further inves-
tigation of 10 is necessary to confirm this presumption due to less
structural similarity compared to UDP-Xyl-4O. Quantification of
saturation transfer via STD amplification factor displayed a higher
absolute STD response of ketone 10 than 5 (Fig. 3). This circum-
stance initially let suggest that 10 is a better substrate than Me-Xyl-
4O (5). At this point, it must be considered that the KD value has a
significant influence in absolute magnetization transfer. A fast off
rate of substrate leads to more saturated molecules in solution and
thus higher absolute STD responses, which indicates loose ligand
binding. However, comparison and statements of substrate affinity
and binding strength were not viable due to missing kon/koff rates of
investigated ketones.
Fig. 4. Relative STD effects of saccharides 13, 20 and 22. In each hexuronic acid proton
H5 indicated a strong interaction with binding site of hUXS1. Anomeric proton of both
methyl-glycosides could not be calculated due to water suppression and thus distin-
guished as “nd”.
Fig. 5. Absolute magnetization transfer of investigated compounds Me-GlcUA (13),
Me-GalUA (20) and 1-P-GlcUA (22). Given amplification factor for each proton resulted
from the ratio of peak intensity in the on-resonance and off-resonance spectrum.
Finally, we investigated the binding pattern of three different
uronic acids. Resulting epitope maps and absolute STD responses of
Me-GlcUA (13), Me-GalUA (20) and 1-P-GlcUA (22) are given in
Fig. 4 and Fig. 5. STD NMR investigations of 13 and 22 indicated
similar binding to hUXS1. In both cases, protons H2, H3 and H5
possessed strong interactions with groups on the protein. However,
pyranoside binding pattern of 13, 20 and 22 showed some differ-
ences compared to previously published STD NMR results of UDP-
glycosides. In particular, STD effects of Me-GlcUA (13) and 1-P-
GlcUA (22) indicated less interaction of proton H4 with the binding
site, whereas reduced STD NMR response has been observed for
pyranoside proton H3 in UDP-glycosides [21]. In general, binding
pattern of compounds 13 and 22 were not in common with MD
simulations of substrate UDP-GlcUA [8]. Hence, these results point
to a different binding of synthesized uronic acids. Further, glycone
moiety of 20 clearly received less absolute saturation. All in all,
pyranoside binding pattern of Me-GalUA (20) only showed two
positive STD responses (H4 and H5). Obtained negative saturation
of protons H2 and H3 likely was caused by trapped D2O molecules
or water coordination to active site residues. The D2O interferes
with the studied substrate during saturation and spin lock event
and, hence, causes negative STD effects [21].
3. Conclusion
Fig. 3. Absolute saturation transfer of oxo-saccharides 5 and 10 are presented as
amplification factors due to varying substrate concentration. Absolute values were
calculated according to (I0-ISTD)/I0.
Taken together, synthesis of keto-saccharides yielded Me-Xyl-

C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
55
4O (5) and Me-Glc-4O (9) via protecting group chemistry. However,
unstable compound 9 rapidly converted to Me-Glc-3O (10). Solu-
tion investigation of UDP-Xyl-4O and synthesized ketones in
deuterated potassium phosphate buffer indicated keto-enol
tautomerism. Deuterium insertion of all compounds was moni-
tored via decreasing peak intensity and thus changes in signal
shape of neighbor protons. Further, gem-diol hydrate species of
Me-Xyl-4O (5) and UDP-Xyl-4O were observed. Binding pattern
study of prepared oxo-glycoside 5 showed reasonable binding to
active site of hUXS1, which is supported by MD simulations of UDP-
Xyl-4O. Glycoside proton H3 of compound 5 showed a strong
interaction with the binding site, which potentially is caused by
amino acid residue Ala79. Further, presence of keto function at C4
seems to be decisive in optimally locating the pyranoside ring of
intermediate in the binding pocket.
sample was irradiated at ꢁ1 ppm (on-resonance) and 30 ppm (off-
resonance). Corresponding 1H NMR spectra of saccharides and
control experiments in absence of protein were recorded [15].
STD effects of investigated compounds were determined by
subtraction of on-resonance from off-resonance spectra. The frac-
tional STD was obtained by (I0-ISTD)/I0. The term (I0-ISTD) represents
the peak intensity in the STD spectrum and I0 the signal intensity in
the corresponding off-resonance spectrum. The largest STD
response in each sample was thus (I0-ISTD)/I0 x 100. Relative signal
intensities in the same spectrum were determined according to the
most intense STD signal [15]. Peak intensities of protons close to
water signal ( 60 Hz) were not determined due to water sup-
pression, which influences the signal intensity.
Absolute STD responses were obtained by calculation of the STD
amplification factor ASTD ¼ (I0-ISTD)/I0 x ligand excess [14]. Deter-
mination of corrected absolute STD effects using CORCEMA analysis
[41] and thus detailed comparison of binding strength of investi-
gated saccharides were not enable due to, inter alia, lacking kon/koff
rates.
4. Experimental
4.1. Materials
In situ STD NMR - a 5 mM solution of Me-Glc-3O (10) was pre-
pared in deuterated phosphate buffer (50 mM, pD 7.1 [38]) to
investigate the binding pattern and keto-enol tautomerism. The
All chemicals and solvents were acquired from Sigma-Aldrich
(Taufkirchen, Germany), Alfa Aesar (Karlsruhe, Germany) and Car-
bosynth (Compton, United Kingdom). Methanol (p.a.) and DCM
(p.a.) were dried by standing over activated 3 Å molecular sieves for
several days [38]. Deuterated solvents were purchased from Eur-
isotop (Saint-Aubin, France).
sample contained 6.6 mM hUXS1 leading to a 1000 fold excess. Over
a time period of 12 h, STD and 1H NMR spectra were measured in
altering order under above conditions described.
Purification via column chromatography was performed with
4.2.1. 1-O-Methyl- -arabinopyranoside (2)
b
-
L
Machery-Nagel silica gel 60 (63-20
mm). Thin layer chromatography
A suspension of L-arabinopyranoside (5 g, 33.30 mmol) and 5 g
was done with Machery-Nagel silica gel 60 F254 plates. Spot visu-
alization was carried out by UV and/or treatment with a vanillin
solution (27 g vanillin in 380 mL EtOH, 20 mL H2SO4 and
50 mL H2O) followed by carbonizing with a heat gun.
Amberlite resin IR-120 H in dry MeOH (50 mL) was heated to reflux
for 7 h. After cooling to rt, the mixture was filtered and concen-
trated under reduced pressure to give a quantitative mixture of
methyl-glycosides. The oily residue was crystallized in i-PrOH to
MS experiments were carried out in ESI mode on a Bruker HTC
Plus and maXis mass spectrometer. IR measurements were ob-
tained as ATR spectra using a Bruker VERTEX 70 IR spectrometer.
Optical rotations of final compounds were measured on a Perkin-
Elmer Polarimeter 341. Enzyme preparation and purification of
wild type hUXS1 and mutant Y147F were obtained via reported
procedures [8].
yield desired methyl-
b
-
L
-glycoside 2 as a white solid (1.42 g) in an
anomeric ratio of
b/
a
20:1. 1H NMR (400 MHz, D2O)
d
3.41 (s, 3 H,
OMe), 3.66 (dd, J ¼ 2.2/12.8 Hz, 1 H, H5eq), 3.84 (m, 2 H, H2/H3),
3.88 (dd, J ¼ 1.2/12.8 Hz, 1 H, H5ax), 4.00 (m, 1 H, H4), 4.83 (d,
J ¼ 3.0,1 H, H1); 13C NMR (400 MHz, D2O)
d 56.07 (OMe), 63.36 (C5),
68.99 (C2), 69.65 (C3/C4), 69.75 (C3/C4), 100.74 (C1); HRMS (m/z):
[M þ Na]þ calcd for C6H12O5Na: 187.0582; found: 187.0572.
4.2. NMR experiments
4.2.2. (2'R,3'R)-Methyl-2,3-O-(2',3'-dimethoxybutane-2',3'-diyl)-
-arabinopyranoside (3)
To a stirred solution of 2 (500 mg, 3.03 mmol) in dry MeOH
(86 mL) under Ar atmosphere were added trimethyl orthoformate
(2.72 mL, 24.82 mmol), butane-2,3-dione (680 L, 7.75 mmol) and
CSA (91 mg, 0.39 mmol). The reaction was heated to reflux for 20 h.
Then, the mixture was neutralized with Et3N (55 L, 0.39 mmol)
b-
L
1H and 13C (J modulated) spectra were recorded on Bruker
Avance (Rheinstetten, Germany) DRX 400 or AV III 600 spectrom-
eters. All measurements were performed in 5 mm high precision
NMR sample tubes (Promochem, Wesel, Germany). Chemical shifts
were referenced to CD3OD (3.31 ppm), CDCl3 (7.26 ppm) or D2O
(external calibration with acetone at 2.225 ppm). Chemical shifts
m
m
and concentrated in vacuum. The residue was purified by flash
column chromatography (hexanes/EtOAc 1:2) to yield 3 as a white
(d) of synthesized compounds are quoted in part per million [ppm]
and coupling constants (J) in [Hz].
solid (719 mg, 85%). 1H NMR (400 MHz, CD3OD)
d 1.28 (s, 3 H, Me),
STD NMR - Samples were prepared in 0.70 mL deuterated buffer
(99.96%, potassium phosphate, 50 mM, pH 7.1 [39]). Each sample
1.29 (s, 3 H, Me), 3.24 (s, 3 H, OMe), 3.25 (s, 3 H, OMe), 3.39 (s, 3 H,
OMe), 3.60 (dd, J ¼ 1.9/12.4 Hz, 1 H, H5ax), 3.83 (dd, J ¼ 1.0/12.4 Hz,
1 H, H5eq), 3.87 (m, 1 H, H4), 4.01 (dd, J ¼ 3.2/10.6, 1 H, H3), 4.15
(dd, J ¼ 3.6/10.6 Hz, 1 H, H2), 4.70 (d, J ¼ 3.6 Hz, 1 H, H1); 13C NMR
contained 10 mM ligands and 6.6
fold excess of investigated saccharides. Spectra were recorded on a
Bruker AV III 600 AVANCE spectrometer at 298.15 and
mM enzyme, which led to a 2000
K
(400 MHz, CD3OD) d 17.99 (Me), 18.11 (Me), 48.13 (OMe), 48.18
600.13 MHz (1H), respectively at 600.25 MHz (1H), with a 5 mm
PABBO BB or 5 mm CPPBBO BB probe head. The pulse program
stddiffgp19.3 and the standard program of Topspin version 3.0 and
3.2 were used for all binding pattern studies. All STD NMR mea-
surements included WATERGATE suppression [40]. Each experi-
ment was obtained with a spectral width of 11 ppm and 39,612 data
points. Selective protein saturation was performed via saturation
cascades of Gaussian shaped pulses of 50 ms length with 1 ms
delay, resulting in a total saturation time of 2.0 s. All NMR spectra
were measured with 768 scans. For binding pattern studies each
(OMe), 55.54 (OMe), 64.79 (C5), 66.58 (C2), 67.22 (C3), 68.91 (C4),
100.08 (C1), 101.31 (Cq), 101.40 (Cq); HRMS (m/z): [M þ Na]þ calcd
for C12H22O7Na: 301.1263; found: 301.1258.
4.2.3. (2'R,3'R)-Methyl-2,3-O-(2',3'-dimethoxybutane-2',3'-diyl)-4-
keto-b-L-arabinopyranoside (4)
A suspension of 3 (600 mg, 2.16 mmol), NaHCO3 (720 mg,
8.58 mmol) and Dess-Martin periodinane (1.82 g, 4.31 mmol) in
DCM (22 mL) was stirred at rt for 24 h. The reaction mixture was
diluted with DCM (140 mL) and transferred into a separatory funnel

56
C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
containing a 1:1 mixture of sat. aq. Na2S2O3 (140 mL) and sat. aq.
NaHCO3 (140 mL). The aqueous layer was extracted with EtOAc
(3 ꢂ 180 mL). The combined organic phases were washed with
brine (150 mL), dried over Na2SO4 and reduced to dryness. Ob-
tained residue was subsequently purified by a short flash column
chromatography (hexanes/EtOAc 1:1) to yield 4 as white solid
4.2.7. 1-O-Methyl-4-keto-
a-D-glucopyranoside (9)
To solution of protected keto-saccharide
a
8 (185 mg,
0.39 mmol) in EtOAc (5.5 mL) was added an aq. solution of NaBrO3
(543 mg, 3.60 mmol; dissolved in 12 mL H2O). The two-phase
system was vigorously stirred followed by slow dropwise addi-
tion of aq. Na2S2O4 (627 mg, 3.60 mmol; dissolved in 24 mL H2O).
The reaction mixture was stirred at rt for 22 h. After phase sepa-
ration, the aqueous phase was reduced to dryness. The resulting
residue was washed with EtOAc and a small amount of MeOH to
give an inseparable mixture of ketones 9 and 10 (C4/C3 ketone 4:1)
(440 mg, 72%). 1H NMR (400 MHz, CDCl3)
d 1.35 (s, 3 H, Me), 1.37 (s,
3 H, Me), 3.24 (s, 3 H, OMe), 3.28 (s, 3 H, OMe), 3.52 (s, 3 H, OMe),
3.93 (d, J ¼ 14.8 Hz, 1 H, H5eq), 4.09 (dd, J ¼ 3.1/11.1 Hz, 1 H, H2),
4.18 (d, J ¼ 14.8 Hz, 1 H, H5ax), 4.84 (d, J ¼ 11.1 Hz, 1 H, H3), 4.90 (d,
J ¼ 3.1 Hz, 1 H, H1); 13C NMR (400 MHz, CDCl3)
d
17.68 (Me), 17.75
as yellow oil (60 mg, 80%). 1H NMR (400 MHz, D2O)
d 3.42 (s, OMe),
(Me), 48.21 (OMe), 48.51 (OMe), 56.16 (OMe), 66.80 (C5), 70.00
(C2), 71.02 (C3), 98.31 (C1), 100.05 (Cq), 100.60 (Cq), 199.84 (C4);
HRMS (m/z): [M þ Na]þ calcd for C12H20O7Na: 299.1107; found:
299.1107.
3.67 (d, J ¼ 10.0 Hz, 1 H, H2), 3.74 (m, 1 H, H5), 3.78 (m, 1 H, H6a),
3.99 (dd, J ¼ 1.9/11.8 Hz, 1 H, H6b), 4.39 (d, J ¼ 10.0 Hz, 1 H, H3), 4.84
(d, J ¼ 3.2 Hz, 1 H, H1); 13C NMR (400 MHz, D2O)
d 55.71 (OMe),
59.73 (C6), 70.86 (C2), 73.19 (C3), 73.85 (C5), 99.75 (C1), 205.55
(C4).
4.2.4. 1-O-Methyl-4-keto-a-D-xylopyranoside (5)
4.2.8. 1-O-Methyl-3-keto-a-D-glucopyranoside (10)
Compound 4 (440 mg,1.59 mmol) was stirred in a 9:1 mixture of
TFA/H2O (11.37 mL) at rt for 2 min. The solvent was removed
directly under reduced pressure to yield 5 as a white solid (259 mg,
IR (ATR):
n
¼ 3388, 1733, 2924 cmꢁ1
;
1H NMR (600 MHz, D2O)
d
3.45 (s, OMe), 3.83 (m, 1 H, H5), 3.91 (dd, J ¼ 4.7/12.5 Hz,1 H, H6a),
99%) in good purity. IR (ATR):
(600 MHz, D2O)
n
¼ 3355, 1761, 2936 cmꢁ1
;
1H NMR
3.98 (dd, J ¼ 2.1/12.5 Hz, 1 H, H6b), 4.43 (dd, J ¼ 1.5/9.9 Hz, 1 H, H4),
4.66 (dd, J ¼ 1.5/4.4, 1 H, H2) 5.05 (d, J ¼ 4.4 Hz, 1 H, H1); 13C NMR
d
3.40 (s, 3 H, OMe), 3.47 (d, J ¼ 12.0 Hz, 1 H, H5a),
(600 MHz, D2O)
d 55.73 (OMe), 62.50 (C6), 73.34 (C4), 76.04 (C2),
3.66 (dd, J ¼ 3.7/9.8 Hz, 1 H, H2), 3.67 (d, J ¼ 12.0 Hz, 1 H, H5b), 3.70
76.72 (C5), 103.70 (C1), 207.03 (C3); HRMS (m/z): [M þ Na]þ calcd
(d, J ¼ 9.8 Hz, 1 H, H3), 4.79 (d, J ¼ 3.7, 1 H, H1); 13C NMR (600 MHz,
for C7H12O6Na: 215.0532; found: 215.0525.
D2O)
d 55.79 (OMe), 64.63 (C5), 70.67 (C2), 73.18 (C3), 93.22 (C4),
100.28 (C1); HRMS (m/z): [M (Hydrate)
þ
Na]þ calcd for
C6H12O6Na: 203.0532; found: 203.0525; a2D0 ¼ þ201.3 (c ¼ 10 mg/
4.2.9. 1-O-Methyl-2,3-di-O-benzyl-a-D-glucopyranoside (11)
A solution of fully protected -glucose 6 (2.0 g, 4.32 mmol) in
D
mL H2O).
MeOH (20 mL) was treated with TFA (0.10 mL) and heated to reflux
for 4 h. After cooling to rt, the reaction mixture was neutralized by
addition of Et3N (0.25 mL) and concentrated in vacuum. The oily
product was purified via gradient column chromatography (EtOAc/
hexanes 1:2e2:1) to give desired product 11 as colorless oil (1.47 g,
4.2.5. 1-O-Methyl-2,3,6-tri-O-benzyl-a-D-glucopyranoside (7)
TFA (0.80 mL,11.77 mmol) was added dropwise to a solution of 6
(1.00 g, 1.07 mmol) and Et3SiH (1.70 mL, 11.77 mmol) in dry DCM
(5.00 mL) at 0 ꢀC under Ar atmosphere. After complete addition, the
reaction mixture was allowed to come to rt and stirred for 3 h. The
mixture was diluted with EtOAc, washed with sat. aq. NaHCO3
(20 mL) and brine (20 mL). The solution was dried over Na2SO4 and
concentrated under reduced pressure. The crude product was pu-
rified via column chromatography (EtOAc/hexanes 1:5) to yield
final product 7 as colorless oil (74%). 1H NMR (400 MHz, CDCl3)
91%). 1H NMR (400 MHz, CD3OD)
d 3.39 (s, 3 H, OMe), 3.45 (m, 1 H,
H2), 3.47 (m, 1 H, H4), 3.54 (ddd, J ¼ 2.2/5.3/9.9 Hz, 1 H, H5), 3.67
(dd, 1 H, J ¼ 5.3/11.8 Hz, 1 H, H6a), 3.73 (dd, 1 H, J ¼ 8.8/9.4 Hz, 1 H,
H3), 3.81 (dd,1 H, J ¼ 2.2/11.8 Hz,1 H, H6b), 4.64 (d, J ¼ 11.7 Hz,1 H)/
4.70 (d, J ¼ 11.7 Hz, 1 H) (benzylic), 4.72 (d, J ¼ 3.6 Hz, 1 H, H1), 4.83
(d, J ¼ 11.2 Hz,1 H)/4.87 (d, J ¼ 11.2 Hz,1 H) (benzylic), 7.23e7.40 (m,
10 H, aromatic); 13C NMR (400 MHz, CD3OD)
d 55.82 (OMe), 62.93
d
3.40 (s, 3 H, OMe), 3.55 (dd, J ¼ 3.2/9.2 Hz, 1 H, H2), 3.61 (m, 1 H,
(C6), 72.08 (C2), 73.98 (C5), 74.45/76.84 (benzylic), 81.56 (C4), 83.37
(C3), 99.58 (C1), 128.90e129.83/140.01/140.82 (aromatic); HRMS
(m/z): [M þ Na]þ calcd for C21H26O6Na: 397.1627; found: 397.1617.
H4), 3.68e3.75 (m, 3 H, H5, H6a, H6b), 3.80 (t, J ¼ 9.2 Hz, 1 H, H3),
4.55 (d, J ¼ 12.1, 1 H)/4.60 (d, J ¼ 12.1, 1 H) (benzylic), 4.65 (d, J ¼ 3.2,
1 H, H1), 4.67 (d, J ¼ 11.3, 1 H)/4.75 (d, J ¼ 11.5, 1 H)/4.78 (d, J ¼ 11.3,
1 H)/5.01 (d, J ¼ 11.5, 1 H) (benzylic), 7.27e7.40 (m, 15 H, aromatic);
4.2.10. 1-O-Methyl-2,3-di-O-benzyl-a-D-glucopyranuronic acid
(12)
13C NMR (400 MHz, CDCl3)
d 55.18 (OMe), 69.47 (C6), 69.86 (C5),
70.71 (C4), 73.10e75.36 (benzylic), 79.57 (C2), 81.42 (C3), 98.15
(C1), 127.57e128.54/137.98e138.78 (aromatic); HRMS (m/z): [M þ
Na]þ calcd for C28H32O6Na: 487.2097; found: 487.3901.
To a 0 ꢀC cooled solution of protected sugar 11 (2.0 g, 5.34 mmol)
in EtOAc (26 mL) were added NaBr (65 mg, 0.63 mmol), TBAFx3H2O
(114 mg, 0.36 mmol), TEMPO (21 mg, 0.13 mmol) and sat. NaHCO3
(15 mL). Next, a cooled mixture of NaOCl (6% w/v, 26 mL), sat.
NaHCO3 (13 mL) and brine (26 mL) was added dropwise to the
latter one. After stirring at 0 ꢀC for 2.5 h, the reaction mixture was
diluted with H2O and EtOAc. The organic phase was washed with
half-sat. NaHCO3 (3ꢂ). Then, the aqueous phases were combined
and acidified to pH 2 using 2 M HCl. The resulting aqueous layer
was extracted with EtOAc (5ꢂ) and DCM (3ꢂ), dried over Mg2SO4
and reduced to dryness to give a yellow oil (1.41 g, 68%). 1H NMR
4.2.6. 1-O-Methyl-2,3,6-tri-O-benzyl-4-keto-a-D-glucopyranoside
(8)
Ketone 8 was prepared from compound 7 (413 mg, 0.89 mmol)
in the same fashion as described for 4. Yield: 371 mg (89%); 1H NMR
(400 MHz, CDCl3)
d
3.49 (s, 3 H, OMe), 3.68 (dd, J ¼ 6.2/10.9 Hz, 1 H,
H6a), 3.80 (dd, J ¼ 3.5/10.0 Hz, 1 H, H2), 3.91 (dd, J ¼ 3.5/10.9 Hz,
1 H, H6b), 4.29 (dd, J ¼ 3.5/6.2 Hz, 1 H, H5), 4.43 (d, J ¼ 10.0 Hz, 1 H,
H3), 4.55 (d, J ¼ 12.1 Hz, 1 H)/4.61 (d, J ¼ 12.1 Hz, 1 H)/4.67 (d,
J ¼ 12.2 Hz, 1 H)/4.68 (d, J ¼ 11.3 Hz, 1 H) (benzylic), 4.80 (d,
J ¼ 3.5 Hz, 1 H, H1), 4.86 (d, J ¼ 12.2 Hz, 1 H)/4.95 (d, J ¼ 11.3 Hz, 1 H)
(benzylic), 7.26e7.45 (m, 15 H, aromatic); 13C NMR (400 MHz,
(400 MHz, CD3OD)
d
3.41 (s, 3 H, OMe), 3.52 (dd, J ¼ 3.6/8.9 Hz, 1 H,
H2), 3.71 (m, 2 H, H3 and H4), 4.01 (d, J ¼ 9.4 Hz 1 H, H5), 4.64 (d,
J ¼ 11.8 Hz, 1 H)/4.71 (d, J ¼ 11.8 Hz, 1 H) (benzylic), 4.76 (d,
J ¼ 3.6 Hz, 1 H, H1), 4.82 (d, J ¼ 11.1 Hz, 1 H)/4.88 (d, J ¼ 11.1 Hz, 1 H)
(benzylic), 7.25e7.41 (m, 10 H, aromatic); 13C NMR (400 MHz,
CDCl3)
80.08 (C2), 82.56 (C3), 98.44 (C1), 127.66e128.50/137.73e137.86
(aromatic), 202.01 (C4); HRMS (m/z): [M
Na]þ calcd for
28H30O6Na: 485.1940; found: 485.1932.
d 56.16 (OMe), 67.61 (C6), 72.72 (C5), 73.66e74.41 (benzylic),
CD3OD)
d 55.90 (OMe), 72.54 (C5), 73.40 (C4), 74.23/76.51
þ
(benzylic), 80.62 (C2), 82.38 (C3), 99.78 (C1), 128.51e129.42/
C
139.97/140.70 (aromatic), 173.10 (C6); HRMS (m/z): [M þ Na]þ

C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
57
calcd for C21H24O7Na: 411.1420; found: 411.1411.
4.2.15. (2'R,3'R)-Methyl-2,3-O-(2',3'-dimethoxybutane-2',3'-diyl)-
-galactopyranuronic acid (19)
Compound 19 (2.5 g, 8.11 mmol) was prepared under the same
a-D
4.2.11. 1-O-Methyl- -glucopyranuronic acid (13)
a-D
reaction conditions used in the synthesis of protected glucuronic
Partial protected sugar 12 (200 mg, 0.51 mmol) was dissolved in
dry MeOH (10 mL). To this solution Pd(10%)/C (15 mg) was added
and the suspension was stirred under H2 atmosphere at rt for 48 h.
The catalyst was filtered off and the solvent was removed in vacuo
to give product 13 as yellow oil (97 mg, 91%). 1H NMR (400 MHz,
acid 12. The product 19 was obtained as yellow oil (2.53 g, 96%). 1H
NMR (400 MHz, CD3OD) d 1.28 (s, 3 H, Me), 1.29 (s, 3 H, Me), 3.25 (s,
3 H, OMe), 3.26 (s, 3 H, OMe), 3.42 (s, 3 H, OMe), 4.06 (dd, J ¼ 3.1/
10.6 Hz, 1 H, H3), 4.18 (dd, J ¼ 3.6/10.6 Hz, 1 H, H2), 4.26 (dd, J ¼ 1.3/
3.1 Hz, 1 H, H4), 4.37 (d, J ¼ 1.3 Hz, 1 H, H5), 4.88 (m, 1 H, H1); 13
C
CD3OD)
d
3.43 (s, OMe), 3.44 (dd, J ¼ 3.7/9.6 Hz, 1 H, H2), 3.51 (dd,
NMR (400 MHz, CD3OD)
d 17.97 (Me), 18.08 (Me), 48.19 (OMe),
J ¼ 9.9/9.0 Hz 1 H, H4), 3.63 (dd, J ¼ 9.0/9.6 Hz, 1 H, H3), 4.00 (d,
J ¼ 9.9 Hz, 1 H, H5), 4.72 (d, J ¼ 3.7 Hz, 1 H, H1); 13C NMR (400 MHz,
48.20 (OMe), 56.02 (OMe), 66.13 (C2), 67.51 (C3), 70.04 (C4), 72.69
(C5), 99.97 (C1), 101.33 (Cq), 101.41 (Cq), 173.20 (C6); HRMS (m/z):
[M - H]- calcd for C13H21O9: 321.1191; found: 321.1172.
CD3OD) d 56.02 (OMe), 72.55 (C5), 73.11 (C2), 73.35 (C4), 74.58 (C3),
101.78 (C1), 173.28 (C6); HRMS (m/z): [M - H]- calcd for C7H11O7:
207.0510; found: 206.9899; a2D0 ¼ þ82.4 (c ¼ 10 mg/mL H2O).
4.2.16. 1-O-Methyl-a-D-galactopyranuronic acid (20)
Uronic acid 21 (200 mg, 0.62 mmol) was stirred in a 9:1 mixture
of TFA/H2O (4.44 mL) at rt for 2 min. The mixture was reduced to
dryness, which gave compound 20 as off-white solid (128 mg, 99%).
4.2.12. 1-O-Methyl-2,3-di-O-benzyl-a-D-glucopyranuronic acid
methyl ester (14)
To a 0 ꢀC cooled solution of glucuronic acid derivative 12
1H NMR (400 MHz, CD3OD)
d 3.41 (s, 3 H, OMe), 3.79 (m, 2 H, H2
(200 mg, 0.51 mmol) in dry MeOH was slowly added SOCl2 (112 mL,
and H3), 4.22 (m, 1 H, H4), 4.38 (d, J ¼ 1.4 Hz, 1 H, H5), 4.79 (d,
15.44 mmol) under Ar atmosphere. The cooled mixture was stirred
for 4 h. The solvent was removed followed by addition of
1.28 mL H2O. The pH was adjusted to 9e10 with sat. NaOH and was
then extracted with EtOAc. The resulting organic layer was washed
with brine and dried over MgSO4. The solvent was removed under
reduced pressure yielding the crude product. Purification via flash
chromatography (hexanes/EtOAc 1:2) gave final product as yellow
J ¼ 2.3 Hz, 1 H, H1); 13C NMR (400 MHz, CD3OD)
d 56.18 (OMe),
69.59 (C2), 70.90 (C3), 71.57 (C5), 71.80 (C4), 101.78 (C1), 172.53
(C6); HRMS (m/z): [M - H]- calcd for C7H11O7: 207.0510; found:
207.0510; a2D0 ¼ þ68.1 (c ¼ 10 mg/mL H2O).
4.2.17. 1-O-phosphate-a-D-glucopyranuronic acid sodium salt (22)
1-O-Phosphate- -glucose disodium salt (850 mg, 2.80 mmol)
a-D
oil (147 mg, 71%). 1H NMR (400 MHz, CD3OD)
d 3.39 (s, 3 H, OMe),
was dissolved in a mixture of EtOAc (14 mL) and sat. NaHCO3
(8 mL). Then, NaBr (35 mg, 0.34 mmol), TBAFx3H2O (61 mg,
0.19 mmol) and TEMPO (11 mg, 0.07 mmol) were added. The
resulting mixture was cooled to 0 ꢀC followed by dropwise addition
of a cooled solution of NaOCl (6% w/v, 14 mL), sat. NaHCO3 (7 mL)
and brine (14 mL). The reaction was stirred at 0 ꢀC for 2.5 h. After
removing EtOAc under reduced pressure, MeOH was added to
precipitate the product. The white crystals were filtered and
washed with MeOH to give compound 22 (657 mg, 69%). 1H NMR
3.51 (m, 1 H, H2), 3.74 (m, 2 H, H3 and H4), 3.76 (s, 3 H, OMe), 4.05
(d, J ¼ 9.5 Hz, 1 H, H5), 4.64 (d, J ¼ 11.8 Hz, 1 H)/4.69 (d, J ¼ 11.8 Hz,
1 H) (benzylic), 4.75 (d, J ¼ 3.6 Hz, 1 H, H1), 4.81 (d, J ¼ 11.2 Hz, 1 H)/
4.87 (d, J ¼ 11.2 Hz, 1 H) (benzylic), 7.22e7.51 (m, 10 H, aromatic);
13C NMR (400 MHz, CD3OD)
d 52.83 (OMe), 55.92 (OMe), 74.22/
76.52 (benzylic) 72.72 (C5), 73.24 (C4) 80.60 (C2), 82.23 (C3), 99.87
(C1), 128.53e129.43/139.56/140.22 (aromatic), 171.76 (C6); HRMS
(m/z): [M þ Na]þ calcd for C22H26O7Na: 425.1576; found: 425.1568.
(600 MHz, D2O)
d
3.51 (dd, J ¼ 9.3/10.2 Hz, 1 H, H4), 3.54 (ddd,
4.2.13. 1-O-Methyl-2,3-di-O-benzyl-4-keto-furanose (16)
J ¼ 1.8/3.6/9.7 Hz, 1 H, H2), 3.81 (t, J ¼ 9.3, 1 H, H3), 4.04 (d, J ¼ 10.2,
1 H, H5), 5.49 (dd, J ¼ 3.6/7.7 Hz, 1 H, H1); 13C NMR (600 MHz, D2O)
Compound 16 (100 mg, 0.25 mmol) was prepared according to
the synthesis of Me-Xyl-4O 5. Flash chromatography (hexanes/
EtOAc 5:1) yielded the product (42 mg, 52%) as yellow oil. 1H NMR
d
72.73 (C2), 72.82 (C4/C5), 72.83 (C4/C5), 73.57 (C3), 94.16 (C1),
177.94 (C6); a2D0 ¼ þ206.3 (c ¼ 10 mg/mL H2O).
(400 MHz, CD3OD)
d
3.48 (s, 3 H, OMe), 4.26 (dd, J ¼ 4.7/9.0 Hz, 1 H,
H2), 4.45 (d, J ¼ 9.0 Hz, H3), 4.65 (s, 2 H) (benzylic), 4.75 (d,
Acknowledgments
J ¼ 11.8 Hz, 1 H)/4.95 (d, J ¼ 11.8 Hz, 1 H) (benzylic), 5.33 (d, J ¼ 4.7,
1 H, H1), 7.01e7.89 (m, 10 H); 13C NMR (400 MHz, CD3OD)
d 57.50
T. Eixelsberger and B. Nidetzky thank the Austrian Science Fund
FWF (project DK Molecular Enzymology; W901-B05) for financial
support.
(OMe), 73.32/73.86 (benzylic), 76.87 (C3), 80.65 (C2), 101.93 (C1),
128.77e141.04/138.62/138.79 (aromatic), 174.00 (C4); HRMS (m/z):
[M þ Na]þ calcd for C19H20O5Na: 351.1208; found: 351.1198.
Appendix A. Supplementary data
4.2.14. (2'R,3'R)-Methyl-2,3-O-(2',3'-dimethoxybutane-2',3'-diyl)-
a-
D
-galactopyranoside (18)
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.carres.2016.11.001.
Protected sugar 18 was prepared from 1-O-Methyl-a-D-gal-
actopyranoside 17 (2.5 g, 12.87 mmol) in the same fashion as
described for 3. Purification via gradient column chromatography
(EtOAc/heptanes 2:1- EtOAc) yielded sugar 18 (2.54 g, 64%) as
References
yellow solid. 1H NMR (600 MHz, CDCl3)
d 1.30 (s, 3 H, Me), 1.33 (s,
[1] C.J. Thibodeaux, C.E. Melançon, H.-W. Liu, Nature 446 (2007) 1008.
[2] A. Stein, M.-R. Kula, L. Elling, S. Verseck, W. Klaffke, Angew. Chem. Int. Ed. 34
(1995) 1748.
[3] K.L. Kavanagh, H. Jornvall, B. Persson, U. Oppermann, Cell Mol. Life Sci. 65
(2008) 3895.
[4] T. Oka, J. Yigami, FEBS 273 (2006) 2645.
[5] G.J. Seifert, Curr. Opin. Plant Biol. 7 (2004) 277.
[6] S.J. Polizzi, R.M. Walsh, P. Le Magueres, A.R. Criswell, Z.A. Wood, Biochemistry
52 (2013) 3888.
[7] J.L. Moriarity, K.J. Hurt, A.C. Resnick, P.B. Storm, W. Laroy, R.L. Schnaar,
S.H. Snyder, J. Biol. Chem. 277 (2002) 16968.
3 H, Me), 3.25 (s, 3 H, OMe), 3.26 (s, 3 H, OMe), 3.42 (s, 3 H, OMe),
3.83 (dd, J ¼ 4.5/10.3 Hz, 1 H, H6a), 3.84 (m, 1 H, H5), 3.95 (dd,
J ¼ 4.4/10.3 Hz, 1 H, H6b), 4.05 (d, J ¼ 3.2 Hz, 1 H, H4), 4.08 (dd,
J ¼ 3.2/10.4 Hz, 1 H, H3), 4.20 (dd, J ¼ 3.6/10.4 Hz, 1 H, H2), 4.85 (d,
€
J ¼ 3.6 Hz, 1 H, H1); 13C NMR (600 MHz, CDCl3)
d 17.69 (Me), 17.75
(Me), 47.94 (OMe), 47.95 (OMe), 55.21 (OMe), 62.92 (C6), 65.03 (C2),
66.10 (C3), 69.24 (C4), 70.01 (C5), 98.34 (C1), 100.13 (Cq), 100.15
(Cq); HRMS (m/z): [M þ Na]þ calcd for C13H24O8Na: 331.1369;
found: 331.3117.
[8] T. Eixelsberger, S. Sykora, S. Egger, M. Brunsteiner, K.L. Kavanagh,
U. Oppermann, L. Brecker, B. Nidetzky, J. Biol. Chem. 287 (2012) 31349.

58
C. Puchner et al. / Carbohydrate Research 437 (2017) 50e58
K.E. Wesson, J. Chem. Soc. Perkin Trans. (1997) 2023.
[9] J.S. Schutzbach, D.S. Feingold, J. Biol. Chem. 245 (1970) 2476.
[10] H. Ankel, D.S. Feingold, Biochemistry 4 (1965) 2468.
[27] S.L. Bartlett, C.M. Beaudry, J. Org. Chem. 76 (2011) 9852.
[28] M. Matwiejuk, J. Thiem, Eur. J. Org. Chem. (2011) 5860.
[29] D.N. Kursanov, Z.N. Parnes, N.M. Loim, Synthesis 9 (1974) 633.
[30] M. Adinolfi, G. Barone, L. Guariniello, A. Iadonisi, Tetrahedron Lett. 40 (1999)
8439.
[31] M. Matwiejuk, J. Thiem, Eur. J. Org. Chem. (2012) 2180.
[32] C.S. Rye, S.G. Withers, J. Org. Chem. 67 (2002) 4505.
[33] J.-G. Tang, Y.-H. Wang, R.-R. Wang, Z.-J. Dong, L.-M. Yang, Y.-T. Zheng, J.-K. Liu,
Chem. Biodiv. 5 (2008) 447.
[11] S.D. Breazeale, A.A. Ribeiro, C.R.H. Raetz, J. Biol. Chem. 277 (2002) 2886.
[12] T. Eixelsberger, H. Weber, B. Nidetzky, Carbohydr. Res. 416 (2015) 1.
[13] J.L. Wagstaff, S.L. Taylor, M.J. Howard, Mol. Biosyst. 9 (2013) 571.
[14] B. Meyer, T. Peters, Angew. Chem. Int. Ed. 42 (2003) 864.
[15] B. Meyer, M. Mayer, Angew. Chem. Int. Ed. 38 (1999) 1784.
[16] M. Mayer, B. Meyer, J. Am. Chem. Soc. 123 (2001) 6108.
€
[17] L. Brecker, A. Schwarz, C. Godl, R. Kratzer, C.E. Tyl, B. Nidetzky, Carbohydr. Res.
343 (2008) 2153.
[18] Y. Yuan, X. Wen, D.A.R. Sanders, B.M. Pinto, Biochemistry 44 (2005) 14080.
[19] T. Biet, T. Peters, Angew. Chem. Int. Ed. 40 (2001) 4189.
[20] L.C.R. Carvalho, D. Ribeiro, R.S.G.R. Seixas, A.M.S. Silva, M. Nave, A.C. Martins,
S. Erhardt, E. Fernandes, E.J. Cabrita, M.M.B. Marques, RSC Adv. 5 (2015)
49098.
[34] A. Heeres, H.A. van Doren, K.F. Gotlieb, I.P. Bleeker, Carbohydr. Res. 299 (1997)
221.
[35] X. Gu, J. Glushka, Y. Yin, Y. Xu, T. Denny, J. Smith, Y. Jiang, M. Bar-Peled, J. Biol.
Chem. 285 (2010) 9030.
[36] Y. Tsuda, M. Hanajima, N. Matshuhira, Y. Okuno, K. Kanemitsu, Chem. Pharm.
Bull. 37 (1989) 2344.
[21] C. Puchner, T. Eixelsberger, B. Nidetzky, L. Brecker, RSC Adv. 5 (2015) 86919.
ꢀ
[22] J.D.C. Codee, A. Ali, H.S. Overkleeft, G.A. van der Marel, C. R. Chim. 14 (2011)
[37] Y. Tsuda, N. Matsuhira, K. Kanemitsu, Chem. Pharm. Bull. 33 (1985) 4095.
[38] D.B.G. Williams, M. Lawton, J. Org. Chem. 75 (2010) 8351.
[39] P.R. Mussini, T. Mussini, S. Rondinini, pH values were calculated according to:
pD ¼ pH þ 0.4, Pure Appl. Chem. 69 (1997) 1007.
178.
[23] S.V. Ley, D.R. Owen, K.E. Wesson, J. Chem. Soc. Perkin Trans. 1 (1997) 2805.
[24] J.-L. Montchamp, F. Tian, M.E. Hart, J.W. Frost, J. Org. Chem. 61 (1996) 3897.
[25] L.F. Bornaghi, S.-A. Poulsen, Tetrahedron Lett. 46 (2005) 3485.
[26] A. Hense, S.V. Ley, H.M.I. Osborn, D.R. Owen, J.-F. Poisson, S.L. Warriner,
[40] M. Piotto, V. Saudek, V. Sklenar, J. Biomol. NMR 2 (1992) 661.
[41] V. Jayalakshmi, N.R. Krishna, J. Am. Chem. Soc. 127 (2005) 14080.