Biochemical Characterization and Mechanistic Analysis of the Levoglucosan Kinase from Lipomyces starkeyi

Article abstract of DOI:10.1002/cbic.201700587

Levoglucosan kinase (LGK) catalyzes the simultaneous hydrolysis and phosphorylation of levoglucosan (1,6-anhydro-β-d-glucopyranose) in the presence of Mg²⁺–ATP. For the Lipomyces starkeyi LGK, we show here with real-time in situ NMR spectroscopy at 10 °C and pH 7.0 that the enzymatic reaction proceeds with inversion of anomeric stereochemistry, resulting in the formation of α-d-glucose-6-phosphate in a manner reminiscent of an inverting β-glycoside hydrolase. Kinetic characterization revealed the Mg²⁺ concentration for optimum activity (20–50 mm), the apparent binding of levoglucosan (K_m=180 mm) and ATP (K_m=1.0 mm), as well as the inhibition by ADP (K_i=0.45 mm) and d-glucose-6-phosphate (IC₅₀=56 mm). The enzyme was highly specific for levoglucosan and exhibited weak ATPase activity in the absence of substrate. The equilibrium conversion of levoglucosan and ATP lay far on the product side, and no enzymatic back reaction from d-glucose-6-phosphate and ADP was observed under a broad range of conditions. 6-Phospho-α-d-glucopyranosyl fluoride and 6-phospho-1,5-anhydro-2-deoxy-d-arabino-hex-1-enitol (6-phospho-d-glucal) were synthesized as probes for the enzymatic mechanism but proved inactive with the enzyme in the presence of ADP. The pyranose ring flip ⁴C₁→¹C₄ required for 1,6-anhydro-product synthesis from d-glucose-6-phosphate probably presents a major thermodynamic restriction to the back reaction of the enzyme.

Full text of DOI:10.1002/cbic.201700587

Accepted Article
Title: Biochemical characterization and mechanistic analysis of the
levoglucosan kinase from Lipomyces starkeyi
Authors: Christina Rother, Alexander Gutmann, Ramakrishna
Gudiminchi, Hansjörg Weber, Alexander Lepak, and Bernd
Nidetzky
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemBioChem 10.1002/cbic.201700587
Link to VoR: http://dx.doi.org/10.1002/cbic.201700587
A Journal of
www.chembiochem.org

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
Biochemical characterization and mechanistic analysis of the
levoglucosan kinase from Lipomyces starkeyi
Christina Rother+,[a] Alexander Gutmann+,[a] Ramakrishna Gudiminchi,[a, b] Hansjörg Weber,[c] Alexander
Lepak,[a] and Bernd Nidetzky*[a, b]
Abstract: Levoglucosan kinase (LGK) catalyzes the simultaneous
hydrolysis and phosphorylation of levoglucosan (1,6-anhydro-β-D-
glucopyranose) in the presence of Mg2+-ATP. For the Lipomyces
starkeyi LGK, we show here with real-time in situ NMR spectroscopy
at 10°C and pH 7.0 that the enzymatic reaction proceeds with
inversion of anomeric stereochemistry, resulting in the formation of
α-D-glucose-6-phosphate, in a manner reminiscent of an inverting β-
glycoside hydrolase. Kinetic characterization revealed the Mg2+
concentration for optimum activity (20-50 mM), the apparent binding
of levoglucosan (Km = 180 mM) and ATP (Km = 1.0 mM) as well as
the inhibition by ADP (Ki = 0.45 mM) and D-glucose-6-phosphate
(IC50 = 56 mM). The enzyme was highly specific for levoglucosan
and exhibited weak ATPase activity in the absence of substrate. The
equilibrium conversion of levoglucosan and ATP lay far on the
product side and no enzymatic back reaction from D-glucose-6-
phosphate and ADP was observed in a broad range of conditions. 6-
Phospho-α-D-glucopyranosyl fluoride and 6-phospho-1,5-anhydro-2-
deoxy-D-arabino-hex-1-enitol (6-phospho-D-glucal) were synthesized
as probes of the enzymatic mechanism but proved inactive with the
enzyme in the presence of ADP. The pyranose ring-
flip 4C1 à 1C4 required for 1,6-anhydro-product synthesis from D-
glucose-6-phosphate probably presents a major thermodynamic
restriction to the back reaction of the enzyme.
acetylmuramic acid kinase), LGK forms
a sub-family of
anhydrosugar kinases within the large hexokinase protein
2]
3]
family.[1a,
LGKs from Lipomyces starkeyi,[1a,
Aspergillus
niger[4] and Sporobolomyces salmonicolor[5] have been purified
and characterized. Furthermore, LGK activity was detected in
the cell extracts of various yeasts and filamentous fungi.[1b, 6]
Because of the anhydro linkage the pyranose ring of 1 is
locked in the otherwise energetically highly unfavorable 1C4 chair
conformation.[7] 1 is a major product of lignocellulose pyrolysis,
accounting for up to 30 wt % of the liquid fraction obtained in the
process.[8] Therefore, the use of 1 as substrate for bioconversion
by microbial fermentation is of a rapidly growing interest.[8a, 9]
Introduction
Levoglucosan kinase (LGK) catalyzes conversion of the
anhydrosugar levoglucosan (1,6-anhydro-β-D-glucopyranose, 1)
in the presence of ATP and Mg2+.[1] The enzymatic reaction
involves a combined hydrolysis and phosphorylation of the
substrate to yield D-glucose-6-phosphate (2) and ADP as the
products (Scheme 1). Together with AnmK (1,6-anhydro-N-
Scheme 1. Mechanistic proposals suggest formation of α-2 (A) or an open-
chain form of 2 (B) as products of the LGK reaction.[1a] Anomerization of 2
(shown in grey) most probably occurs in solution. C) Compounds 3 and 4 were
synthesized as mechanistic probes of the enzymatic reaction in the reverse
direction.
[a]
C. Rother, Dr. A. Gutmann, Dr. R. Gudiminchi, A. Lepak, Prof. B.
Nidetzky
Institute of Biotechnology and Biochemical Engineering
Graz University of Technology, NAWI Graz
Petersgasse 12, A-8010 Graz, Austria
E-mail: bernd.nidetzky@tugraz.at
Besides the biocatalytic application of LGK, the enzyme's
catalytic mechanism has also drawn considerable attention. How
[b]
[b]
[+]
Dr. R. Gudiminchi, Prof. Dr. B. Nidetzky
Austrian Centre of Industrial Biotechnology
Petersgasse 14, A-8010 Graz, Austria
LGK couples hydrolysis of the anhydro-substrate
1 with
phosphorylation from ATP is problem of fundamental
a
Prof. H. Weber
significance. LGK and AnmK uniquely combine catalytic features
of a glycoside hydrolase (GH) with those of an ATP-dependent
sugar kinase.[10] Structures of the LGK from L. starkeyi
(LsLGK)[1a] and the AnmK from Pseudomonas aeruginosa[10a, 11]
have been determined. In both enzymes, a conserved Asp was
Graz University of Technology, NAWI Graz, A-8010 Graz, Austria
Stremayrgasse 9, A-8010 Graz, Austria
These authors contributed equally to this work.
Supporting information for this article is given via a link at the end of
the document.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
suggested to serve as catalytic base for attack of water on the
substrate's anomeric carbon. On hydrolytic opening of the 1,6-
Purified LsLGK proved to be a highly unstable protein.
Differential scanning fluorimetry (DSF) revealed
a melting
anhydro-ring, phosphorylation at C6 also occurs.
A
temperature (Tm) of just 37°C. The enzyme was not stabilized by
one of its substrates or products, except for ATP (2 mM) together
with MgCl2 (10 mM) which increased Tm by 3°C. Consistent with
the low Tm, LsLGK was quickly inactivated under typical reaction
conditions (Figure S4A). At 30°C and pH 7.5, all activity was lost
within just 2 h. Temperature decrease to 25°C and addition of 1
mg mL-1 BSA enhanced the enzyme's half-life to around 4 h.
Raising the BSA concentration to 5 mg mL-1 did not further
enhance the stability (Figure S4B). Stability was pH dependent,
with an optimum at around pH 7.9 (Figure S4C). Limited stability
of LsLGK was noted before,[14] but in another study[3] the
enzyme was found to be stable (for 30 min) at 30°C in the pH
range 7-9. With the main factors of LsLGK stability unknown at
this time, we proceeded taking carefully into account that fast
activity loss restricted the usable incubation times in biochemical
assays for enzyme characterization.
stereochemical reaction course was proposed for LGK (Scheme
1A) in which the α-anomer of 2 is the product. Accordingly, the
anomeric configuration changes between β in 1 and α in 2,
which is reminiscent of that of an inverting β-glycoside
hydrolase.[12] A NMR time course experiment was used to
analyze the formation of 2 by LsLGK.[1a] However, product 2 was
detected as an equilibrium mixture of anomers (38% α, 62%
β).[13] It was therefore suggested that the LGK reaction might
involve a pyranose ring-opening step, as shown in Scheme
1B.[1a] This might facilitate release of ring strain present in 1.
Ring closure would then yield the mixture of anomers of 2, at
equilibrium. However, the very fast mutarotation of
2
[13]
complicates this analysis.
Interestingly, NMR analysis of the
reaciton of AnmK revealed formation of the α-anomer in a
stereochemically inverting reaction. [10a]
The current study was performed to advance mechanistic
understanding of LsLGK based on three lines of experiment.
The problem of the catalytic reaction's stereochemical course
was re-assessed by real-time NMR under conditions that reduce
the rate of mutarotation. 6-Phospho-α-D-glucopyranosyl fluoride
(3) and 6-phospho-1,5-anhydro-2-deoxy-D-arabino-hex-1-enitol
(4) were synthesized and evaluated as mechanistic probes of
LGK, potentially functioning as analogues of 2 in the reverse
reaction of the enzyme. Compounds 3 and 4 were of particular
interest because unlike 2, which must release water to enable
1,6-anhydro-ring formation (Scheme 1A), they could become
"activated" much more easily by expulsion of fluoride or
protonation, respectively. It was also considered that
thermodynamically (due to the exergonic release of fluoride), the
conversion of 3 into 1 might be much more favored than the
conversion of 2. Finally, although LsLGK was characterized
biochemically in some detail before,[1a, 3] certain properties of the
enzyme and the reaction catalyzed by it were not clear. This
required a systematic re-investigation and new evidence was
obtained in its course.
Kinetic characterization and substrate specificity of LsLGK
The pH-activity profile of LsLGK had a sharp optimum at pH 8
(Figure S5), corresponding well to the optimum of stability at pH
7.9 (Figure S4C). Therefore, HEPES buffer (pH 7.8) was used in
all further studies.
The active site of LsLGK contains two Mg2+ ions which
interact with the phosphate groups of ATP.[1a] The ratio of Mg2+
/
ATP strongly influences the activity of LsLGK. Using 4 and 8 mM
Mg2+ the activity was highest at a Mg2+ / ATP ratio of 2 and
drastically dropped when the ratio was less than 1.[1a] We
considered that the ideal concentration of Mg2+ might also
depend on the concentrations of ATP and LsLGK used.
Employing a constant ATP concentration (2.5 mM), the enzyme
activity increased with the Mg2+ concentration until a Mg2+ / ATP
ratio of 8 was reached at 20 mM MgCl2 (Figure 1A). A further
increase of Mg2+ led to a moderate decrease in activity. It was
previously shown for a cyclin-dependent kinase, which like
LsLGK requires the binding of two magnesium ions for
activity,[15] that the complex of 2 Mg2+ and ADP binds more
tightly to the enzyme than does the corresponding ATP complex.
The same effect might also lead to "Mg2+ substrate" inhibition of
LsLGK.[1a, 15]
Results and Discussion
The influence of Mg2+ was considered in the determination of
the Km for ATP (Figure 1B). Adjusted to the ATP concentration
used, the MgCl2 concentration was varied from 10 to 50 mM to
avoid conditions too high or too low in Mg2+. The resulting Km
was 1.0 ± 0.2 mM, as shown in Figure 1B. Somewhat lower Km
values of 0.20 ± 0.02 mM[3a] and 0.68 ± 0.06 mM[3b] were
previously obtained using a constant MgCl2 concentration of 10
mM. At the time of these earlier measurements, rate reduction
due to Mg2+ limitation was not yet reported and so might have
been mistaken for substrate saturation by ATP. The Km for 1 was
177 ± 36 mM (Figure S6). It was again slightly higher but
Purification and stabilization of LsLGK
LsLGK equipped with N- or C-terminal His-tag was produced in
Escherichia coli. The N-terminally tagged protein accumulated
mostly in aggregates and was hardly active. The N-termini in
dimeric assembly of LsLGK are in close proximity with each
other (Figure S1). Steric conflict may therefore prevent
accommodation of the His-tag in a still native enzyme structure.
In contrast, the C-terminally tagged protein was expressed well
and could be purified conveniently (Figure S2). A specific activity
of 16.9 ± 2.2 U mg-1 was determined at pH 7.8 under optimized
assay conditions. The far-UV circular dichroism (CD) spectrum
of the purified LsLGK (Figure S3) indicated a properly folded
protein. The relative composition of secondary structural
elements in LsLGK calculated from the CD data was in
agreement with that found in the enzyme's crystal structure.[1a]
3]
comparable with earlier reported values (69-119 mM).[1a,
Product inihibiton of LGK was tested in reactions containing 300
mM 1 and 2.5 mM ATP. ADP was a strong competitive inhibitor
against ATP with a Ki of 0.45 ± 0.03 mM (Figure S7A). A similarly
strong inhibition by ADP was reported for the LGKs from
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
Sporobolomyces salmonicolor (Ki = 0.15 mM)[5] and Aspergillus
niger (Ki = 0.20 mM).[4a] Product 2 also inhibited the LsLGK,
however only weakly (IC50 = 56 ± 10 mM, Figure S7B). D-
Glucose, D-mannose and D-fructose hardly inhibited the enzyme
(Figure S8).
presence of ATP (Figure S10). In all cases, no conversion of the
putative substrate was detected. Consumption of ATP was
identical to the control comprising only enzyme and ATP.
Formation of ADP was always closely matched to the release of
phosphate. High substrate specificity of LsLGK is supported.
Figure 1. The dependence of the LsLGK reaction rate on Mg2+ and ATP
concentration was determined (500 mM 1, 0.02 mg mL-1 LsLGK, pH 7.8, 30°C).
A) Reactions in the presence of 2.5 mM ATP are shown. B) To determine the
Km for ATP, the Mg2+ concentration was adjusted to the ATP concentration
used: 10 mM (light grey), 20 mM (black) or 50 mM (dark grey) MgCl2 were used.
The solid line shows a hyperbolic fit of the data.
Figure 2. A) The active site of LsLGK is shown (PDB ID: 4ZLU, chain A).[1a]
Only the catalytic base Asp-212 forms a hydrogen bond with C4 of 1, which is
surrounded by water molecules. The bound
1 appears to adopt a
1C4
pyranosyl ring conformation. B) ATP consumption (square, solid line) and
release of free phosphate (circle, dashed line) were measured to study the
canonical LsLGK reaction and error hydrolysis of ATP. Reactions in absence
of LsLGK (black) or 1 (grey) were compared to a standard conversion (green).
Reaction conditions: 0/100 mM 1, 2 mM ATP, 0/1 mg mL-1 LsLGK, 10 mM
MgCl2, pH 7.8, 30°C.
The crystal structure of LsLGK revealed a relatively "open"
substrate-binding site (Figure 2A).[1a] Somewhat surprisingly for
a carbohydrate-active enzyme, 1 forms only a single hydrogen
bond with LsLGK (Asp-212). Otherwise 1 is surrounded by
numerous water molecules. This raised concern that hydrolysis
might compete effectively with the canonical use of ATP for
phosphorylation of 1. Hexokinases are generally known to
exhibit ATPase activity in the absence of substrate.[16]
Indications of minor "error" hydrolysis of ATP were also found
with the LGK from S. salmonicolor.[5] To distinguish between
synthesis of 2 and hydrolysis of ATP in the enzymatic reaction,
we analyzed the release of ADP and inorganic phosphate on
conversion of ATP. We compared utilization of ATP in the
presence of 1 to its utilization when 1 or LsLGK was lacking
(Figures 2B and S9). When 1 was present, all ATP (2 mM) was
rapidly converted to ADP and no hydrolysis took place. In the
absence of 1, however, ATP was degraded by hydrolysis.
Activity of LsLGK for ATP hydrolysis (~4 mU mg-1) was about
400-fold slower than activity for conversion of 1 to 2. Gradual
slowdown of ATP conversion is explainable by the limited
stability of LsLGK. The ADP produced initially was not
hydrolyzed further. A control lacking LsLGK showed that ATP
was stable under the applied conditions (Figure 2B).
In situ NMR supports an inverting mechanism of LsLGK
The use of real-time proton NMR to study the stereochemical
course of the LGK reaction is complicated by the fast
mutarotation of the released 2. At 25°C and pH 7, the half-lives
of α- and β-2 were reported to be just 5 and 10 s, respectively.[13]
On lowering the temperature to 0°C the half-life of β-2 was
increased to 2.1 min.[17] Change of solvent from H2O to D2O also
enhanced anomer stability by about 1.7-fold.[18] Conditions
suitable for in situ 1H NMR analysis therefore involve an
enzymatic conversion of 1 into 2 that proceeds faster than the
subsequent mutarotation. We carefully optimized temperature
(10°C), pD (7.0), LsLGK (2 mg mL-1) and substrate
concentrations (100 mM 1, 10 mM ATP) to this end.
Results of real-time monitoring of the enzymatic reaction are
shown in Figure 3A. The first NMR spectrum recorded 5 min
after the start reveals a clear excess of α-2 (56%) over β-2
(44%). The relative abundance of β-2 subsequently increased
until after about 25 min the mutarotation equilibrium was
reached (Figure 3B). The seemingly conflicting observation of an
anomeric mixture of 2 by Bacik et al. is probably due to
increased mutarotation at higher reaction temperature (20°C
instead of 10°C) and longer reaction time (4 h instead of 40
min).[1a] In fact, during the less emphasized first minutes of the
conversion some accumulation of α-2 was also visible in the
previous in-situ NMR study.[1a] The distinct patterns of α-2 and β-
The binding of 1 by LsLGK (Figure 2A) inspires the belief
that the enzyme might utilize anhydrosugars other than 1 as its
substrates. However, biochemical evidence suggests LGKs to
4a, 5]
be rather specific for 1.[3b,
We tested a series of
anhydrosugars and also D-glucose for utilization by LsLGK in the
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
2 formation in dependence of time (Figure 3B) support a
reaction scenario in which LsLGK is specific for formation of the
α-anomer of 2. Because glucopyranose ring opening (Scheme
1B) is not in line with initial accumulation of α-2, we propose an
inverting (β→α) stereochemical reaction course of LsLGK,
similar to that of AnmK (Scheme 1A). Accordingly, LsLGK
(HPAEC-PAD) was used and samples were pretreated to
decrease the large excess of 2 and ADP. Sample preparation
was validated with 0.3 mM 1 as an internal standard (Figure
S11). The detection limit of 1 was indicated to be about 50 µM.
Analysis of mechanistic probes of the LGK reaction
combines mechanistic features of an ATP-dependent kinase and
Compounds 3 and 4 (Scheme 1C) were synthesized as potential
probes of the reverse reaction of LsLGK. Their use was inspired
by the fact that glycosyl fluorides and glycals have proved useful
as alternative substrates of different glycoside hydrolases from
both the retaining and the inverting mechanistic type.[21]
However, enzymatic reactivity with these analogues can vary in
a broad range and in various glycoside hydrolases, especially
the glycals proved to be potent inhibitors.[21g, 22] We nonetheless
considered that formation of the intramolecular β-1,6-glycosidic
bond could be facilitated by replacing the difficult water release
step of the canonical reaction of 2 by a different, probably more
easily achieved chemical step: release of fluoride from 3 and
protonation of the double bond at C2 in 4 (Scheme 2). Both 3
and 4 do not rely in principle on the protonation of Asp-212 for
reactivity. Although conversion of 4 also involves a protonation
step, it occurs at C2 and not at the OH-leaving group at C1
which forms a hydrogen bond with Asp-212 (Schemes 1A and
2B). Therefore, we envisioned that the numerous water
molecules in the active site of LsLGK (Figure 2A) could act as
shuttles for long range proton transfer from an alternative
general acid to the C2 of 4.
[10c, 10d]
an inverting glycoside hydrolase.
From the enzyme's
structure,[1a] Asp-212 might function as general base facilitating
the attack of water in the catalytic mechanism (Scheme 1A,
Figure 2A).
To examine the reactivity of 3, LsLGK (1.5 mg mL-1) was
added directly to a reaction mixture obtained by phosphorylation
of α-D-glucopyranosyl fluoride from ATP using hexokinase.
Figure 3. The stereochemical course of the conversion of 1 by LsLGK was
analyzed by using real-time in situ NMR (100 mM 1, 10 mM ATP, 20 mM MgCl2,
2 mg mL-1 LsLGK, pD 7.0, 10°C). A) Superposition of 1H NMR spectra from
conversion of 1 in 2 is shown. Characteristic signals for α-2 and β-2 are
indicated in the spectra. B) The relative portion of α- and β-anomer of 2
formed during conversion of 1 is shown in dependence of reaction time.
While the formation of 3 was confirmed unambiguously using 19
NMR analysis (Figure S12), H NMR analysis revealed that no
F
1
further conversion of 3 into 1 occurred over 24 h (Figure S13).
Also no transformation of
detectable.
3 into any other product was
The LsLGK reaction is experimentally quasi-irreversible
Conversion of 1 into 2 is expected to release a large amount of
free energy, resulting from ATP conversion and from relaxation
of conformational strain in 1 upon hydrolysis of the β-1,6-
glycosidic bond. Just the phosphorylation of D-glucose from ATP
releases ~20 kJ mol-1 of energy when performed at neutral pH in
the presence of Mg2+.[19] Enzymatic reactions at varied substrate
concentrations (ATP: 2-10 mM; 1: 2-100 mM; pH 7.8) confirmed
that the equilibrium of the LsLGK reaction lay far on the side of
product 2. However, despite high enzyme concentrations used
in the experiments (3 mg mL-1), the time to reaction equilibrium
exceeded the lifetime of the enzyme. Determination of
equilibrium constant (Keq) was therefore not pursued. To
investigate formation of 1 via enzymatic back reaction, a broad
range of ADP (0.2-20 mM) and 2 (1-150 mM) concentrations was
tested. The pH was also varied in the range 6 - 9, considering its
possible effect on Keq in phosphoryl transfer reactions.[20] No 1
was detected under the different conditions used. Detection of 1
was performed routinely by HPLC using refractive index
detection. To enhance sensitivity, high-performance anion
exchange chromatography with pulsed amperometric detection
Scheme 2. Proposed mechanisms of LsLGK-catalyzed conversion of 3 (A)
and 4 (B) in the presence of ADP are shown.
The reactivity of 4 was examined similarly as described above
for 3. Compound
4
was synthesized by enzymatic
phosphorylation of "D-glucal" (1,5-anhydro-2-deoxy-D-arabino-
hex-1-enitol) and isolated by anion-exchange chromatography
and barium acetate precipitation. To avoid thermodynamic
restrictions during conversion of 4, ATP was removed in situ by
adenylate kinase from Escherichia coli (EcADK).[23] The enzyme
catalyzes the reversible formation of two ADP molecules from
ATP and AMP. Therefore, the reaction mixture contained
besides 160 mM 4 and 10 mM ADP also 90 mM AMP. To
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
compensate potentially low reaction rates with the substrate
analogue, 10 mg mL-1 LsLGK were used and the reaction was
1
followed for 24 h by in situ H-NMR (Figure S14). However, no
2-deoxy-1 was detectable and also no conversion of 4 to
alternative products was observed.
Generally, these results suggest that "activation" of substrate
in 3 and 4 compared to 2 could not be exploited by the enzyme
to overcome an apparently very large energy barrier to β-1,6-
anhydro-ring formation. Note that release of fluoride from 3 is
expected to release a large amount of free energy,[21b] which in
principle could be utilized to promote the formation of 1.
Interestingly, therefore, uncoupling of hydrolysis from phosphoryl
transfer, to form 1 from 3 or 2-deoxy-1 from 4, was also not
observed in the reactions of LsLGK.
Mechanistic proposal: a hypothetical catalytic itinerary of
sugar puckering conformations
Considering Hammond's postulate,[24] the large decrease in
standard Gibbs free energy (ΔG0) on conversion of 1 into 2
would imply an early transition state, whose structure resembles
1 more than it resembles 2. Besides cleavage of the anhydro
linkage and transfer of a phosphate group from ATP, conversion
4
of 1 by LsLGK also has to involve a 1C4 → C1 pyranose ring flip.
This rearrangement must happen via an itinerary of sugar ring
conformations, the variety of which is summarized in the
Cremer-Pople spherical coordinate system (Figure 4A).[25]
Further point of consideration was that a glycoside hydrolase-
like reaction mechanism of LsLGK would plausibly involve an
oxocarbenium ion-like transition state.[26] Such transition states
are characterized by partial change in hybridization (sp3→sp2) at
the anomeric carbon.[27] The implied coplanar orientation of C5,
O5, C1 and C2 can only be adopted by two boat (B2,5 and 2,5B),
two half-chair (4H3 and 3H4) and four envelope conformations (3E,
Figure 4. A mechanistic proposal for the LsLGK reaction is shown. A) The
suggested catalytic itinerary of sugar puckering conformations is displayed on
the Cremer-Pople spherical coordinate system. Conformations in bold can
adopt oxocarbenium ion-like transition states. The proposed 3H4 transition
state involving cleavage of the anhydro-bond with coincident change from β-
(blue) to α-configuration (red) is shown in purple. The conformational itinerary
by the dotted line is expected to happen on the enzyme. The additional
conformational rearrangements could also occur in solution. A tentative route
is indicated. A mechanistic proposal for anhydro-bond cleavage is shown in
(B). The 1C4 chair conformation is observed for 1 bound to LsLGK.[1a]
4
E3, E and E4).[26b, 27a] Considering the conformational itineraries
utilized by glycoside hydrolases, reviewed in References 26 and
27, we suggest a catalytic reaction coordinate for LGK, as
shown in Figure 4.
Conclusions
Among the sugar ring conformations directly accessible from
the 1C4 chair (Figure 4A), only the 3H4 and the closely related 3E
and E4 conformations can accommodate oxocarbenium ion
Conversion of 1 into 2 by LsLGK involves configurational
inversion at the anomeric carbon, β (equatorial) in substrate and
α (axial) in product. This stereochemical course is consistent
with the hybrid kinase-hydrolase mechanism depicted in Figure
4. In this mechanism, opening of the β-1,6-anhydro-ring is
promoted by nucleophilic attack of O6 on the terminal γ-
phosphorus of ATP. The water nucleophile reacts under general
base catalytic assistance from Asp-212. The transition state is
expected to have substantial oxocarbenium ion-like character. A
hypothetical conformational itinerary for conversion of the 1C4
3
character. A H4 transition state is tentatively suggested for the
LGK reaction (Figure 4B). The 3H4 half-chair limits ring strain
from the 1,6-anhydro linkage and has typically a somewhat
lower energy than the 3E and the E4 conformation.[26b] It is
structurally close to the 1C4 chair of 1 observed in the substrate-
bound structure of LsLGK.[1a] Its formation in the transition state
would be in line with the principle of least nuclear motion in
catalysis. The 1C4
à
3H4‡ à 3S1 itinerary suggested for LsLGK
‡
chair of 1 into the 4C1 chair of α-2 involves a 3H4 half-chair
(Figure 4B) has precedence from inverting mannosidases of
glycoside hydrolase families GH 47 and GH 134.[28] It is also
found in retaining α-fucosidases and sialidases.[26b, 29] In the LGK
reaction, conversion of α-2 from the post-catalytic 3S1
transition state and a 3S1 skew-boat conformation. Absence of
detectable reaction of 2, or its "activated" analogue 3, to yield 1
4
involves the large energy barrier for converting the C1 chair to
the 3S1 conformation. Lack of reactivity of 4 might reflect inability
to adopt the 3S1 conformation. Besides the mechanistic analysis,
new biochemical data for LsLGK provide an advanced
knowledge basis for this interesting class of enzymes.
4
conformation to the most stable C1 conformation might happen
on the enzyme or in solution. An itinerary via E1, or alternatively
4
OE, is feasible.[25b] The large barrier for conversion of the C1 to
the 3S1 conformation (~32 kJ mol-1 for α-D-glucose)[25b] would
restrict enzymatic reaction of 2 or 3 to 1.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
5.6-6.8; HEPES pH 6.6-8.3; glycine pH 8.7-10.5). The influence of the
MgCl2 concentration (1-75 mM) was tested in reactions containing 500
mM 1 and 2.5 mM ATP. The Km of LsLGK for ATP was determined in the
presence of 500 mM 1 and 1 mg mL-1 BSA. The MgCl2 concentration was
adjusted to the ATP level. 0.02-1 mM ATP: 10 mM MgCl2; 2 mM ATP: 20
mM MgCl2; 5-20 mM ATP: 50 mM MgCl2. To determine the Km for 1 in the
presence of 2.5 mM ATP and 10 mM MgCl2, the concentration of 1 was
varied from 10 to 750 mM. Product inhibition by 0.1-5 mM ADP or 10-190
mM 2 was studied with 2.5 mM ATP and 300 mM 1. The inhibiting effect of
monosaccharides was studied with the continuous G6P-DH assay (300
mM 1, 2.5 mM ATP, 1 mg mL-1 BSA). Sugar concentrations were 10 times
higher than the corresponding Km of S. cerevisiae hexokinase:[31] 3.3 mM
fructose, 1.2 mM glucose or 0.5 mM mannose.
Experimental Section
Materials: ATP, ADP, (anhydro)-sugars and 2,3,4,6-tetra-O-acetyl-α-D-
glucopyranosyl fluoride were from Carbosynth (Berkshire, UK).
Phosphoenolpyruvic acid (PEP), D-glucal, pyruvate kinase from rabbit
muscle, hexokinase type III from S. cerevisiae (25 U mg-1) and glucose-
6-phosphate
dehydrogenase
(G6P-DH)
from
Leuconostoc
mesenteroides were from Sigma-Aldrich (Vienna, Austria).
Preparation of LsLGK: Detailed procedures for creation of LsLGK
expression strains, recombinant production in Escherichia coli and
protein purification are found in the Supporting Information. In short, for
expression with N- and C-terminally fused His-tag the codon optimized
LsLGK gene (GenBank EU751287) was inserted into the pQE-30 and
pET-41b(+) vector, respectively. LsLGK from overnight expression in E.
coli BL21-Gold (DE3) was affinity purified by Ni-NTA chromatography.
Substrate specificity and error hydrolysis of ATP: Phosphorylation of
various (anhydro)-sugars by LsLGK was evaluated: 1, D-glucose, 1,6-
anhydro-β-D-mannopyranose, 1,6-anhydro-β-D-galactopyranose, 1,6-
anhydro-β-D-cellobiose, 1,6-anhydro-β-D-maltose. At the same time the
influence of these sugars on ATP hydrolysis was tested. Reaction
mixtures contained 100 mM sugar, 2 mM ATP, 10 mM MgCl2, and 50 mM
HEPES, pH 7.8. 24 h long conversions at 30°C were started by addition
of 1 mg mL-1 LsLGK. Control reactions contained either no sugar or no
LsLGK. Conversion of ATP to ADP was quantified by HPLC and release
of inorganic phosphate was measured spectrophotometrically at 850 nm
as reported by Saheki et al.[32]
LsLGK activity assays: Either conversion of ATP to ADP was
monitored by HPLC or formation of 2 was quantified photometrically.
Aliquots withdrawn for HPLC analysis were stopped by addition of an
equal volume of acetonitrile. ADP and ATP separation on a Kinetex™
C18 column (5µm, 100 Å, 50 x 4.6 mm; Phenomenex, Aschaffenburg,
Germany) was described elsewhere in detail.[30] A flow rate of 2 mL min-1
(40 mM TBAB, 20 mM KH2PO4, pH 5.9) was applied and nucleotides
were quantified by UV-detection at 259 nm. G6P-DH was used to couple
synthesis of
2
to equimolar formation of NADH which was
Real-time in situ NMR: Reaction mixtures contained 100 mM 1, 10 mM
ATP, 20 mM MgCl2, 2 mg mL-1 LsLGK and 1 mg mL-1 BSA. The pD was
set to 7.0 and the temperature was 10°C. Spectra were recorded on a
Varian (Agilent) INOVA 500-MHz spectrometer (Agilent Technologies,
Santa Clara, United States) using the VNMRJ 2.2D software. 1H NMR
spectra were measured at 499.98 MHz on a 5 mm indirect detection
PFG-probe. The water signal was pre-saturated by a shaped pulse, using
a standard pre-saturation sequence: relaxation delay 2 s; 90° proton
pulse; acquisition time 2.048 s; spectral width 8 kHz; number of points 32
k. For integration of the signal of the β-anomer of 2, the clearly separated
half of the doublet signal was integrated and the resulting area was
multiplied by two.
spectrophotometrically quantified at 340 nm (ε = 6.22 mM-1 cm-1). Unless
mentioned otherwise, reaction mixtures for continuous measurement of 2
formation contained 300 mM 1, 2.5 mM ATP, 1.4 mM NAD+, 0.02 mg mL-1
LsLGK, 10 U mL-1 G6P-DH, 20 mM MgCl2, 1 mg mL-1 BSA and 50 mM
HEPES, pH 7.8. Conversions were performed in microplates at 30°C and
started by addition of 1. Alternatively the assay was used discontinuously.
To quantify 2 aliquots were withdrawn from LsLGK reactions and the
enzyme was inactivated by heating to 95°C for 5 min. After removal of
precipitated protein by centrifugation, samples were 10-fold diluted with
50 mM phosphate buffer, pH 7.0 containing 1.4 mM NAD+, 10 mM MgCl2
and 10 mM EDTA. The concentration of 2 was inferred from differences
in absorption at 340 nm before and after incubation with 10 U mL-1 G6P-
DH (30 min at 30°C).
Reaction equilibrium: Concentrations of 1 (2-100 mM) and ATP (2-10
mM) were varied, resulting in ratios of 1/ATP ranging from 0.7 to 16.7.
Reaction mixtures contained 3 mg mL-1 LsLGK, 1 mg mL-1 BSA, 20 mM
MgCl2 and 100 mM HEPES, pH 7.8. Over the 24-h long conversions at
30°C aliquots were withdrawn at certain times. The concentrations of
Stability of LsLGK: Verification of proper folding of LsLGK by far-UV CD
spectroscopy and evaluation of enzyme stabilization by DSF are
described in detail in the Supporting Information. Time-dependent
inactivation was tested by measuring LsLGK activity with the continuous
G6P-DH assay after various incubation times ranging from 1 to 24 h.
Unless mentioned otherwise, 1 mg mL-1 LsLGK was incubated at 25°C in
20 mM HEPES buffer, pH 7.5 containing 50 mM NaCl and 0.5 mM DTT.
An increase of temperature from 25 to 30°C and addition of 1 and 2 mg
mL-1 BSA were evaluated. In presence of 1 mg mL-1 BSA the effect of pH
was tested. 50 mM MES (pH 6.4), HEPES (pH 7.9) or glycine (pH 7.0 and
9.6) were used. To determine the operational stability of LsLGK, time
courses of 2 formation were analyzed in presence of 0 to 5 mg mL-1 BSA.
300 mM 1 and 2.5 mM ATP were converted by 0.01 mg mL-1 LsLGK at
25°C. Reactions were buffered to pH 7.8 by 50 mM HEPES containing 20
mM MgCl2. Throughout the conversions samples were withdrawn and the
concentration of 2 was determined by the discontinuous G6P-DH assay.
ADP and ATP were determined by HPLC and
2 was quantified
photometrically by the discontinuous G6P-DH assay. The residual 1 was
calculated from the initially applied 1 and the amount of 2 formed.
Reversal of the LsLGK reaction: To identify possible conditions for
synthesis of 1, various concentrations of 2 (1-150 mM) and ADP (0.2-20
mM) were tested. Also the Mg2+ concentration was varied from 1 to 50
mM and a pH range from 6 to 9 was investigated. Reactions were carried
out at 25°C and 300 rpm in presence of 1 mg mL-1 BSA. Conversions
were started by addition of 0.5-1.1 mg mL-1 LsLGK and followed for up to
24 h. HPLC analysis was used to detect 1. A HPX-87H column (Bio-Rad,
Vienna, Austria) was operated at a flow rate of 0.8 mL min-1 (5 mM H2SO4,
65°C) and refractive index detection was employed. For a more sensitive
detection of 1 selected samples were also analyzed by HPAEC-PAD as
described in detail in the Supporting Information.
Kinetic characterization of LsLGK: Unless indicated otherwise,
reactions were performed at 30°C and pH 7.8. Reaction mixtures
contained 0.02 mg mL-1 LsLGK, 50 mM HEPES, 20 mM MgCl2 and 5 mg
mL-1 BSA. If not mentioned elsewise, ion-pairing HPLC was used to
monitor conversion of ATP to ADP. Initial reaction rates were determined
from measurements at four distinct points in time.
A pH-activity profile was recorded by following conversion of 100 mM 1
and 2 mM ATP by 0.06 mg mL-1 LsLGK with the discontinuous G6P-DH
assay. Reactions contained 10 mM MgCl2 and 50 mM buffer (MES pH
Synthesis and conversion of 6-phospho-α-D-glucopyranosyl
fluoride (3): Preparation of α-D-glucopyranosyl fluoride from 2,3,4,6-
tetra-O-acetyl-α-D-glucopyranosyl fluoride is described in the Supporting
Information. About 100 mM α-D-glucopyranosyl fluoride were used in a
reaction with 100 mM ATP, 100 mM MgCl2 and 4 mg mL-1 hexokinase in
D2O. The pD was adjusted to 7.0 with concentrated NaOH. 19F-NMR
analysis showed that 23 mM 3 were formed after 24 h of incubation at
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
30°C. LsLGK (1.5 mg mL-1) was added and after another 24 h at 25°C
the reaction mixture was analyzed by 1H-NMR.
12283-12291; b) J. P. Bacik, L. R. Jarboe, IUBMB life 2016, 68, 700-
708; c) G. Davies, B. Henrissat, Structure 1995, 3, 853-859; d) D.
Pollard-Knight, A. Cornish-Bowden, Mol Cell Biochem 1982, 44, 71-80.
[11] J.-P. Bacik, M. Tavassoli, T. R. Patel, S. A. McKenna, D. J. Vocadlo, M.
Khajehpour, B. L. Mark, J. Biol. Chem. 2014, 289, 4504-4514.
[12] a) M. L. Sinnott, Chem. Rev. 1990, 90, 1171-1202; b) D. L. Zechel, S.
G. Withers, Curr. Opin. Chem. Biol. 2001, 5, 643-649.
Synthesis and conversion of 6-phospho-D-glucal (4): Synthesis of 4
by hexokinase catalyzed phosphorylation of D-glucal was adopted from
Chenault et al.[33] Anion-exchange chromatography and barium
precipitation were applied to isolate 4. Preparation of 4 is described in
detail in the Supporting Information.
[13] K. J. Schray, S. J. Benkovic, Acc. Chem. Res. 1978, 11, 136-141.
[14] J. R. Klesmith, J.-P. Bacik, R. Michalczyk, T. A. Whitehead, ACS Synth.
Biol. 2015, 4, 1235-1243.
Conversion of 160 mM 4 and 10 mM ADP by 10 mg mL-1 LsLGK in D2O
1
was followed by in situ H-NMR. The reaction was buffered to pD 7.8 by
10 mM phosphate buffer containing 20 mM MgCl2 and 1 mg mL-1 BSA.
For ATP removal 90 mM AMP and 1 mg mL-1 EcADK were added (see
the Supporting Information for details on EcADK preparation).
Conversion at 30°C was monitored for 24 h on a Varian INOVA 500-MHz
spectrometer as described for the canonical LsLGK reaction by recording
one spectrum per hour.
[15] D. M. Jacobsen, Z.-Q. Bao, P. O’Brien, C. L. Brooks III, M. A. Young, J.
Am. Chem. Soc. 2012, 134, 15357-15370.
[16] A. Kaji, S. P. Colowick, J. Biol. Chem. 1965, 240, 4454-4462.
[17] E. Gernert, A. S. Keston, Arch. Biochem. Biophys. 1974, 161, 420-425.
[18] K. J. Schray, E. E. Howell, Arch. Biochem. Biophys. 1978, 189, 102-
105.
[19] E. A. Robbins, P. D. Boyer, J. Biol. Chem. 1957, 224, 121-135.
[20] a) J. W. Lawson, R. L. Veech, J. Biol. Chem. 1979, 254, 6528-6537; b)
A. Gutmann, B. Nidetzky, Adv. Synth. Catal. 2016, 358, 3600-3609.
[21] a) Y. Tanaka, W. Tao, J. S. Blanchard, E. J. Hehre, J. Biol. Chem. 1994,
269, 32306-32312; b) S. J. Williams, S. G. Withers, Carbohydr. Res.
2000, 327, 27-46; c) E. J. Hehre, in Enzymatic Degradation of Insoluble
Carbohydrates, Vol. 618 (Eds.: J. N. Saddler, M. H. Penner), American
Chemical Society, 1996, pp. 66-78; d) P. Wildberger, L. Brecker, B.
Nidetzky, Carbohydr. Res. 2012, 356, 224-232; e) G. Legler, K. R.
Roeser, H. K. Illig, Eur. J. Biochem. 1979, 101, 85-92; f) B. Lougheed,
H. D. Ly, W. W. Wakarchuk, S. G. Withers, J. Biol. Chem. 1999, 274,
37717-37722; g) J. C. Díaz Arribas, A. G. Herrero, M. Martín-Lomas, F.
J. Cañada, S. He, S. G. Withers, Eur. J. Biochem. 2000, 267, 6996-
7005.
Acknowledgements
The authors thank Tea Pavkov-Keller (University of Graz,
Austria) for performing far-UV CD spectroscopy experiments
and Sandra Kulmer for preparing EcADK. This work has been
financially supported by the Austrian BMWFW, BMVIT, SFG,
Standortagentur Tirol, Government of Lower Austria and
Business Agency Vienna through the Austrian FFG-COMET-
Funding Program.
Keywords: Carbohydrates • Conformational analyses • Enzyme
[22] a) A. G. Santana, G. Vadlamani, B. L. Mark, S. G. Withers, Chem.
Commun. 2016, 52, 7943-7946; b) M. Petricevic, L. F. Sobala, P. Z.
Fernandes, L. Raich, A. J. Thompson, G. Bernardo-Seisdedos, O.
Millet, S. Zhu, M. Sollogoub, J. Jiménez-Barbero, C. Rovira, G. J.
Davies, S. J. Williams, J. Am. Chem. Soc. 2017, 139, 1089-1097.
[23] M. Brune, R. Schumann, F. Wittinghofer, Nucleic Acids Res. 1985, 13,
7139-7151.
catalysis • Levoglucosan kinase • Reaction mechanisms
[1]
a) J.-P. Bacik, J. R. Klesmith, T. A. Whitehead, L. R. Jarboe, C. J.
Unkefer, B. L. Mark, R. Michalczyk, J. Biol. Chem. 2015, 290, 26638-
26648; b) Y. Kitamura, Y. Abe, T. Yasui, Agric. Biol. Chem. 1991, 55,
515-521.
[24] G. S. Hammond, J. Am. Chem. Soc. 1955, 77, 334-338.
[25] a) D. Cremer, J. A. Pople, J. Am. Chem. Soc. 1975, 97, 1354-1358; b)
H. B. Mayes, L. J. Broadbelt, G. T. Beckham, J. Am. Chem. Soc. 2014,
136, 1008-1022.
[2]
[3]
T. Uehara, K. Suefuji, N. Valbuena, B. Meehan, M. Donegan, J. T. Park,
J. Bacteriol. 2005, 187, 3643-3649.
a) J. Dai, Z. Yu, Y. He, L. Zhang, Z. Bai, Z. Dong, Y. Du, H. Zhang,
World J. Microbiol. Biotechnol. 2009, 25, 1589-1595; b) J. Ning, Z. Yu,
H. Xie, H. Zhang, G. Zhuang, Z. Bai, S. Yang, Y. Jiang, World J.
Microbiol. Biotechnol. 2008, 24, 15-22.
[26] a) G. J. Davies, V. M.-A. Ducros, A. Varrot, D. L. Zechel, Biochem. Soc.
Trans. 2003, 523-527; b) G. Speciale, A. J. Thompson, G. J. Davies, S.
J. Williams, Curr. Opin. Struct. Biol. 2014, 28, 1-13.
[4]
a) X. Zhuang, H. Zhang, Protein Expression Purif. 2002, 26, 71-81; b)
H.-j. Xie, X.-l. Zhuang, H.-x. Zhang, Z.-h. Bai, H.-y. Qi, FEMS Microbiol.
Lett. 2005, 251, 313-319.
[27] a) A. Ardèvol, C. Rovira, J. Am. Chem. Soc. 2015, 137, 7528-7547; b)
G. J. Davies, A. Planas, C. Rovira, Acc. Chem. Res. 2012, 45, 308-316.
[28] a) K. Karaveg, A. Siriwardena, W. Tempel, Z.-J. Liu, J. Glushka, B.-C.
Wang, K. W. Moremen, J. Biol. Chem. 2005, 280, 16197-16207; b) A. J.
Thompson, J. Dabin, J. Iglesias-Fernández, A. Ardèvol, Z. Dinev, S. J.
Williams, O. Bande, A. Siriwardena, C. Moreland, T. C. Hu, D. K. Smith,
H. J. Gilbert, C. Rovira, G. J. Davies, Angew. Chem. 2012, 124, 11159-
11163; Angew. Chem. Int. Ed. 2012, 51, 10997-11001; c) Y. Jin, M.
Petricevic, A. John, L. Raich, H. Jenkins, L. Portela De Souza, F.
Cuskin, H. J. Gilbert, C. Rovira, E. D. Goddard-Borger, S. J. Williams,
G. J. Davies, ACS Cent. Sci. 2016, 2, 896-903.
[5]
[6]
Y. Kitamura, T. Yasui, Agric. Biol. Chem. 1991, 55, 523-529.
H. Xie, X. Zhuang, Z. Bai, H. Qi, H. Zhang, World J. Microbiol.
Biotechnol. 2006, 22, 887-892.
[7]
[8]
a) M. K. Dowd, A. D. French, P. J. Reilly, Carbohydr. Res. 1994, 264,
1-19; b) L. Smrčok, M. Sládkovičová, V. Langer, C. C. Wilson, M. Koóš,
Acta Crystallogr., Sect. B: Struct. Sci. 2006, 62, 912-918.
a) L. R. Jarboe, Z. Wen, D. Choi, R. C. Brown, Appl. Microbiol.
Biotechnol. 2011, 91, 1519-1523; b) Z. U. Islam, Y. Zhisheng, E. B.
Hassan, C. Dongdong, Z. Hongxun, J. Ind. Microbiol. Biotechnol. 2015,
42, 1557-1579; c) H. B. Mayes, M. W. Nolte, G. T. Beckham, B. H.
Shanks, L. J. Broadbelt, ACS Sustainable Chem. Eng. 2014, 2, 1461-
1473.
[29] a) A. L. van Bueren, A. Ardèvol, J. Fayers-Kerr, B. Luo, Y. Zhang, M.
Sollogoub, Y. Blériot, C. Rovira, G. J. Davies, J. Am. Chem. Soc. 2010,
132, 1804-1806; b) G. Sulzenbacher, C. Bignon, T. Nishimura, C. A.
Tarling, S. G. Withers, B. Henrissat, Y. Bourne, J. Biol. Chem. 2004,
279, 13119-13128; c) M. a. F. Amaya, A. G. Watts, I. Damager, A.
Wehenkel, T. Nguyen, A. Buschiazzo, G. Paris, A. C. Frasch, S. G.
Withers, P. M. Alzari, Structure 2004, 12, 775-784.
[9]
a) J. Lian, M. Garcia-Perez, S. Chen, Bioresour. Technol. 2013, 133,
183-189; b) Z. Yang, Z. Bai, H. Sun, Z. Yu, X. Li, Y. Guo, H. Zhang,
Microb. Cell Fact. 2014, 13, 182; c) X. L. Zhuang, H. X. Zhang, J. Z.
Yang, H. Y. Qi, Bioresour. Technol. 2001, 79, 63-66.
[30] S. T. Kulmer, A. Gutmann, M. Lemmerer, B. Nidetzky, Adv. Synth.
Catal. 2017, 359, 292-301.
[10] a) J.-P. Bacik, G. E. Whitworth, K. A. Stubbs, A. K. Yadav, D. R. Martin,
B. A. Bailey-Elkin, D. J. Vocadlo, B. L. Mark, J. Biol. Chem. 2011, 286,
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
[31] R. Fernandez, P. Herrero, F. Moreno, J. Gen. Microbiol. 1985, 131,
2705-2709.
[33] H. K. Chenault, R. F. Mandes, Bioorg. Med. Chem. 1994, 2, 627-629.
[32] S. Saheki, A. Takeda, T. Shimazu, Anal. Biochem. 1985, 148, 277-281.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/cbic.201700587
ChemBioChem
FULL PAPER
FULL PAPER
A
kinase-inverting
β-
Christina Rother, Alexander
glycoside hydrolase hybrid
reaction going one way: The
Gutmann, Ramakrishna
Gudiminchi, Hansjörg Weber,
Alexander Lepak, Bernd
Nidetzky*
levoglucosan
kinase
from
Lipomyces starkeyi converts 1
with inversion of anomeric
configuration to yield α-2.
Backwards reaction of 2, or of
the "activated" analogues (3,
Page No. – Page No.
Biochemical characterization
and mechanistic analysis of
the levoglucosan kinase from
Lipomyces starkeyi
4),
probably reflecting
energy barrier comprised of
was
not
detectable,
a
large
sugar ring
restrictions.
conformational
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.