Thermostable transketolase from Geobacillus stearothermophilus: Characterization and catalytic properties

Article abstract of DOI:10.1002/adsc.201200590

Here we have characterized the first transketolase (TK) from a thermophilic microorganism, Geobacillus stearothermophilus, which was expressed from a synthetic gene in Escherichia coli. The G. stearothermophilus TK (mTK _gst) retained 100% activity for one week at 50 °C and for 3 days at 65 °C, and has an optimum temperature range around 60-70 °C, which will be useful for preparative applications and for future biocatalyst development. The thermostability of the mTK_gst allowed us to carry out an easy, one-step purification by heat shock treatment of crude cell extracts at 65 °C for 45min, directly yielding 132mg of pure mTK _gst from 1L of culture. The reaction rate of mTK_gst with glycolaldehyde was 14 times higher at 70 °C than at 20 °C, and 4 times higher at 50 °C when compared to E. coli TK under identical conditions. When tested at 50 °C with other aldehydes as acceptors, mTK_gst activity was approximately 3 times higher than those obtained at 20 °C. Applications of this new TK in biocatalysis were performed with hydroxypyruvate as donor and three different aldehydes as acceptors-glycolaldehyde, D-glyceraldehyde and butyraldehyde-from which the corresponding products L-erythrulose 1, D-xylulose 2 and 1,3-dihydroxyhexan-2-one 3 were obtained, respectively. The optical rotations for products 1 and 2 indicate that the stereospecificity of mTK_gst is identical to that of other TK sources, leading to a (3S) configuration. With the non-hydroxylated substrate, butanal, the ee value was 85% (3S), showing higher enantioselectivity than the E. coli TK (75% ee, 3S). Processes at elevated temperatures could offer opportunities to extend the applications of TK biocatalysis, by favoring hydrophobic aldehyde acceptor substrate solubility and tolerance towards non-conventional media. Copyright

Full text of DOI:10.1002/adsc.201200590

FULL PAPERS
DOI: 10.1002/adsc.201200590
Thermostable Transketolase from Geobacillus stearothermophi-
lus: Characterization and Catalytic Properties
Juliane Abdoul-Zabar,a,b Isabelle Sorel,a,b,c Virgil Hꢀlaine,a,b
Franck Charmantray,a,b Titu Devamani,e Dong Yi,e Vꢀronique de Berardinis,c
Dominique Louis,c Philippe Marliꢁre,d Wolf-Dieter Fessner,e
and Laurence Hecqueta,b,*
a
Clermont Universitꢀ, Universitꢀ Blaise Pascal, Institut de Chimie de Clermont-Ferrand, BP 10448, F-63000
Clermont-Ferrand, France
Fax : (+33)-4-7340-7717; phone: (+33)-4-7340-787; e-mail: laurence.hecquet@univ-bpclermont.fr
CNRS, UMR 6296, ICCF, F-63177 Aubiꢁre, France
CEA, DSV, IG, Genoscope, 2 rue Gaston-Crꢀmieux, F-91057 Evry, France
Isthmus SARL, 81 rue Rꢀaumur, F-75002 Paris, France
b
c
d
e
Institut fꢂr Organische Chemie und Biochemie, Technische Universitꢃt Darmstadt, Petersenstraße 22,d-64287 Darmstadt,
Germany
Received: July 6, 2012; Revised: October 12, 2012; Published online: December 28, 2012
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201200590.
Abstract: Here we have characterized the first trans- were performed with hydroxypyruvate as donor and
ketolase (TK) from a thermophilic microorganism, three different aldehydes as acceptors – glycolalde-
Geobacillus stearothermophilus, which was expressed hyde, d-glyceraldehyde and butyraldehyde – from
from a synthetic gene in Escherichia coli. The G. which the corresponding products l-erythrulose 1, d-
stearothermophilus TK (mTKgst) retained 100% ac- xylulose 2 and 1,3-dihydroxyhexan-2-one 3 were ob-
tivity for one week at 508C and for 3 days at 658C, tained, respectively. The optical rotations for prod-
and has an optimum temperature range around 60– ucts 1 and 2 indicate that the stereospecificity of
708C, which will be useful for preparative applica- mTKgst is identical to that of other TK sources, lead-
tions and for future biocatalyst development. The ing to a (3S) configuration. With the non-hydroxylat-
thermostability of the mTKgst allowed us to carry out ed substrate, butanal, the ee value was 85% (3S),
an easy, one-step purification by heat shock treat- showing higher enantioselectivity than the E. coli TK
ment of crude cell extracts at 658C for 45 min, di- (75% ee, 3S). Processes at elevated temperatures
rectly yielding 132 mg of pure mTKgst from 1 L of could offer opportunities to extend the applications
culture. The reaction rate of mTKgst with glycolalde- of TK biocatalysis, by favoring hydrophobic alde-
hyde was 14 times higher at 708C than at 208C, and hyde acceptor substrate solubility and tolerance to-
4 times higher at 508C when compared to E. coli TK wards non-conventional media.
under identical conditions. When tested at 508C with
other aldehydes as acceptors, mTKgst activity was ap-
proximately 3 times higher than those obtained at Keywords: biocatalysis; C C bond formation; mono-
208C. Applications of this new TK in biocatalysis saccharides; thermostability; transketolases
À
Introduction
preferentially used with hydroxypyruvate (HPA)[1] as
the donor substrate, thereby rendering the ketol
This paper focuses on a thiamine-dependent enzyme, transfer irreversible, and making TK an ideal tool for
transketolase (TK, EC 2.2.1.1.), which in vivo catalyz- ketose synthesis (Scheme 1).[2–9] Previous studies
À
es the stereospecific formation of a C C bond by a re- showed that this enzyme type preferentially accepts
versible transfer of the C-1 C-2 ketol unit from a-hydroxylated aldehydes with (2R) configuration, al-
À
a ketose phosphate to an aldose phosphate. The new though activity is lower when aldehydes are non-phos-
asymmetric center is stereospecifically formed in the phorylated and less functionalized (alkyl, allyl, aro-
(S) configuration. For synthetic applications, TK is matic or heterocyclic moieties).[2b,3] The final ketose
116
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
Results and Discussion
Analysis of the tk Gene from G. stearothermophilus
for Heterologous Expression
The protein sequences and solved dimeric 3D struc-
tures of microbial TKs from E.coli,[20a] S. cerevisiae[21a]
and B. anthracis[22] show high structural homologies.
These micro-organisms contain two TK isoenzymes:
TKTA/TKTB in E.coli, TKL1/TKL2 in yeast, and
TK1/TK2 in B. anthracis. E. coli TKTA[20b] and yeast
Scheme 1. Irreversible reaction catalyzed by TK in the pres-
ence of hydroxypyruvate (HPA) as donor substrate and an
a-hydroxy aldehyde as acceptor substrate.
products invariably have the d-threo configuration TKL1[21b].are the most active and most frequently
(3S,4R).
used for biocatalysis.
To advance this research, TKs from yeast[10]
Since no G. stearothermophilus TK (TKgst) was
(TKyeast) or from E. coli[11] (TKeco) have been cloned hitherto known in the public databases, we used the
and overexpressed, giving sufficient amounts of pro- draft genomic sequence of G. stearothermophilus
tein with high specific activity. To improve synthetic strain 10 to find a gene orthologous to that of the
processes based on TK-catalyzed reactions, some known Bacillus anthracis tk1 gene (belonging to the
studies have described TK immobilization method- same family). Analysis of the G. stearothermophilus
s.[12,13a] and modular microfluidic reactor design.[13b]
genome revealed only one tk gene, suggesting that
From the industrial point of view, enzymatic pro- this organism contains only a single TK enzyme (55%
cesses at elevated temperature offer many advantages, identity with TK1 and 72% with TK2 of B. anthracis)
such as higher reaction rate, improved solubility of or- (Figure 1).
ganic substrates, and higher tolerance towards non-
However, since the publicly available G. stearother-
conventional media. In addition, for TK improvement mophilus genome sequence is an incomplete draft,
by mutagenesis,[14,15] a thermostable enzyme also a second TK might have been missed. We further
offers better resistance against protein destabilizing used the TK protein sequence of G. stearothermophi-
factors.
lus DSM 13240 to search for homologous TK proteins
We have chosen Geobacillus stearothermophilus in the other available draft genomes of Geobacillus
(formerly Bacillus stearothermophilus) as a potential species (six Geobacillus sp., G. kaustophilus, G. ther-
source of a process and thermally stable TK enzyme. modenitrificans and G. thermoglucosidasius). For all
This Gram-positive thermophilic bacterium is widely these available draft genomes, also only a single tk
distributed in soil, hot springs and ocean sediments, gene homologous to that of the G. stearothermophilus
and grows over a temperature range of 30–758C. was found, suggesting that generally only one tk gene
Many heat-stable enzymes, such as xylanase,[16] l-ara- occurs in Geobacillus species.
binose isomerase,[17] glycosidases,[18] and a-amylases,[19]
have been isolated from this thermophilic bacterium.
As yet, no complete genome sequence is available in Cloning, Expression, and Purification of G.
public databases for any G. stearothermophilus strain stearothermophilus TK from Recombinant E. coli
and to the best of our knowledge no transketolase
enzyme has hitherto been described.
The coding sequence for the wild-type TK of G. stear-
Here we report the first characterization of a Geo- othermophilus, amplified by PCR, was cloned into ex-
bacillus TK expressed from the G. stearothermophilus pression vector pGEN452, derived from pET47b (see
gene tk, which was identified from the total known Experimental Section). Thus the expression of the
genome analysis, in recombinant E. coli cells. In line cloned gene is under control of the T7 promotor, al-
with the thermophilic property of the donor microor- lowing the production of large enzyme quantities. In
ganism, we planned herein to study the thermostabili- view of intended further studies we modified the
ty of the TK enzyme and the optimal conditions for coding sequence of G. stearothermophilus wild-type
enzyme purification and biocatalysis. Moreover, we TK to remove some restriction sites and introduce
considered its behavior towards various acceptor alde- new unique restriction sites for cassette mutagenesis
hydes for a comparison to TKeco and TKyeast, with re- purposes. This slightly modifies the sequence of the
spect to kinetic parameters versus temperature. Final- gene and changes three amino acids in the protein se-
ly, to determine its stereoselectivity, the G. stearother- quence: L397F, D399G and H479Q. Because these
mophilus TK was applied for reactions performed on residues are located at positions distant from the
a preparative scale at 508C in the presence of Li- active site, the changes were expected not to disturb
HPA as donor and some typical acceptor substrates.
the enzyme activity. The gene sequence was synthe-
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
117

Juliane Abdoul-Zabar et al.
FULL PAPERS
Figure 2. SDS/PAGE analysis of the TKgst and mTKgst purifi-
cation (for conditions, see Experimental. Section). Lane 1:
SeeBlueꢅ Plus2 Pre-Stained Standard, lane 2: crude extract
mTKgst (20 mg), lane 3: mTKgst purified by IMAC (5 mg),),
lane 4: mTKgst purified by heat shock (5 mg), lane 5: See-
Blueꢅ Plus2 Pre-Stained Standard, lane 6: crude extract
TKgst (20 mg), lane 7: TKgst purified by IMAC (5 mg).
by IMAC. The expressed His6-tagged fusion proteins
were applied to an Ni2+ chelating affinity column for
purification. By this procedure, 160 mg each of the
purified enzymes TKgst or mTKgst (0.6 U·mgÀ1) were
effectively obtained from the cells grown in 1 L cul-
ture (5 g·LÀ1 of wet cells) (Table 1).
Apparently, the sequence modifications introduced
into the wild-type tkgst gene and causing small varia-
tions in protein sequence did not affect the overall
specific mTKgst activity. Thus, for all further studies
we exclusively used the protein expressed from the
synthetic gene, and all the results presented here were
Figure 1. Sequence alignment of G. stearothermophilus
transketolase (GstTK) and B. anthracis transketolase 1 and
2 (BanTK1 and BanTK2) using the program CLUSTALW.
Identical residues are indicated by grey background.
sized commercially (see Experimental Section) and obtained from the corresponding protein variant
cloned into the expression vector pGEN452. mTKgst. Given the thermostability of the mTKgst (vide
The high rates of overexpression of G. stearother- infra), we performed an alternative purification pro-
mophilus wild-type TK (TKgst) and G. stearothermo- cedure by heat shock treatment (Figure 2 and
philus TK expressed from the modified synthetic gene Table 1). After cell lysis, crude extract was heated at
sequence (mTKgst) were confirmed by SDS-PAGE 658C for 45 min, then centrifuged at 6,000 g for
and found to be very similar (Figure 2). A hexa-histi- 25 min. This easy and rapid purification method yield-
dine-tag was added at the N-terminus of the enzyme ed 132 mg of purified enzyme from the cells grown in
to facilitate purification.
1 L of culture. The specific activity and the SDS/
After expression in E. coli BL21 (DE3), TKgst and PAGE profile for mTKgst were similar to that ob-
mTKgst were purified using affinity chromatography tained with the affinity chromatography method.
118
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
Table 1. Purification of the TKgst and mTKgst from recombinant E. coli cells.
Purification step
Protein[a]
[mg]
Total activity[b]
[U]
Specific activity[b]
[U·mgÀ1]
Yield
[%]
Purification
(fold)
Crude extract
Purified TKgst or mTKgst by
IMAC
600
160
180
140
0.3
0.9
–
78
3
3
Purified mTKgst by heat shock
132
112
0.9
62
[a]
Crude extract obtained after sonication of E. coli cells overexpressing TKgst or mTKgst harvested from 1 L of culture.
Activity measured in the presence of l-erythrulose and d-ribose-5-phosphate as donor and acceptor substrates, respec-
tively, in buffered medium at 208C. Glycolaldehyde released was measured spectrophotometrically at 340 nm via NADH
consumption by alcohol dehydrogenase.
[b]
Determination of mTKgst Activity
from all microbial TKs characterized so far. The E.
coli enzyme displays maximum activity in glycylgly-
Specific activity of mTKgst was determined in the cine buffer at pH 8.0 to 8.5,[11a] between pH 7.0 and
presence of l-erythrulose as donor and d-ribose-5- 7.5 in phosphate buffer,[24] and 7.0 in Tris buffer.[25a]
phosphate (d-R5P) as acceptor substrates.[23] l-Eryth- Reported optimum pH values for bakersꢆs and brew-
rulose, a non-natural substrate, was used in order to erꢆs yeast TK are 7.5 and 7.6, respectively.[26,27]
replace d-xylulose-5-phosphate,[8b] a natural donor
The influence of temperature on mTKgst activity
substrate which is not continuously commercially was studied between 20–758C in Tris buffered
available (Scheme 2). This assay was chosen because medium (Figure 3, B). mTKgst has an optimum tem-
it allows a direct, continuous measurement of TK ini- perature range of 60–708C, leading to an improve-
tial velocity at different temperatures (208C–708C; ment of up to 10-fold when compared to the activity
using a Peltier-effect controller) by following the for- measured at 208C. This temperature profile was iden-
mation of glycolaldehyde (GA) spectrophotometrical- tical to that obtained for wild-type TKgst (data not
ly at 340 nm via consumption of NADH in a coupled shown). The optimum temperature of mTKgst was
enzymatic reaction using yeast alcohol dehydrogenase much higher than that of other characterized TK
(ADH).[23] We checked that ADH was stable in the from microbial sources such as TKeco which has
range from 208C to 758C at least for the time re- a broad optimum activity at 20–408C.[11a] In contrast,
quired for the measurement of initial TK velocity erythrocyte TK exhibited an optimal pH range at
(5 min.). This method was used throughout for TK ac- 508C with a sharp drop in activity when temperature
tivity determinations.
was increased to 608C.[11e]
pH and Temperature Profiles
Thermostability
The effect of pH on mTKgst activity was determined Thermostability was studied in the range 20–858C by
in the range of pH 5–9 at 208C (Figure 3, A). measuring the residual enzymatic activity after ex-
The enzyme showed the highest activity at pH 7–8, tended incubation at certain temperatures. mTKgst re-
which is similar to the optimum pH for TK activity tained 100% activity for one week at 508C and 3 days
Scheme 2. Assay for activity determination of TK in the presence of l-erythrulose and d-ribose-5-phosphate (d-R5P) as sub-
strates leading to d-sedoheptulose-7-phosphate (d-S7P) and glycolaldehyde (GA). Glycolaldehyde released is measured
spectrophotometrically at 340 nm using alcohol dehydrogenase (ADH) and NADH.
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
119

Juliane Abdoul-Zabar et al.
FULL PAPERS
Figure 3. mTKgst activities measured in buffered medium in
the presence of l-erythrulose and d-ribose-5-phosphate as
donor and acceptor substrates, respectively. Glycolaldehyde
released was measured spectrophotometrically at 340 nm via
NADH consumption with alcohol dehydrogenase. (A)
Effect of pH on the mTKgst activity. Buffers (50 mM) used
Figure 4. Thermostability of different TK sources against
&
^
time. (A) Thermostability of mTKgst at 508C , at 708C
,
~
at 758C and at 858C ꢀ. (B) Thermostability for other mi-
~
&
crobial TKs; TKeco at 508C , and at 708C , Maximum ac-
tivity corresponds to a specific activity of 0.86 mgÀ1. TKyeast
^
at 508C , and at 708C ꢀ. Maximum activity corresponds to
~
&
were glycylglycine (pKa 8.17) , MES (pKa 6.15) , and Tris
a specific activity of 0.8 U·mgÀ1 for TKyeast and of 12 U·mgÀ1
for TKeco. Following incubation of TK at the indicated tem-
perature in glycylglycine buffer at pH 7.5, TK activities were
measured from aliquots at 208C in presence of l-erythrulose
and d-ribose-5-phosphate as donor and acceptor substrates,
respectively. Glycolaldehyde released was measured with al-
cohol dehydrogenase and NADH by spectrophotometry at
340 nm.
^
(pKa 7.2) . (B) Effect of temperature on the mTKgst activi-
ty. TK was incubated at the indicated temperature; the max-
imum activity corresponds to
a specific activity of
11 U·mgÀ1. Activity of the enzyme measured at 208C: maxi-
mum activity corresponds to
0.86 U·mgÀ1.
a
specific activity of
at 708C (data not shown). The results obtained from
two-hour incubations at 508C, 708C, 758C and 858C showed a rapid thermoinactivation of TKyeast at
are indicated in Figure 4, A. The mTKgst activity de- 608C.[25b]
creased rapidly at 758C (half-life was about 15 min),
Our study shows that mTKgst exhibits high thermo-
and at 858C mTKgst lost its activity immediately. The stability as expected from the property of the micro-
results were compared with other microbial TK sour- organism, extending the potential of this catalyst to
ces, E. coli and S. cerevisiae (Figure 4, B). At 508C, a simple method of heat shock purification and possi-
yeast TK activity decreased gradually (half-life ble improvements for demanding applications in syn-
35 min) and activity loss was immediate at 708C. thetic biocatalysis.
TKeco was more resistant, retaining 100% of its activi-
ty at 508C for 90 min, while its activity completely
vanished during 5 min exposure at 708C. This finding Determination of mTKgst Kinetic Constants
was corroborated by a recent study[25a] of the thermal
inactivation of TKeco at above 408C for incubation The kinetic parameters KM and vmax of mTKgst were
times of up to 1 h. At 40–558C the residual activity in- determined using the standard Michaelis–Menten
creased with incubation time and temperature when model and compared with those obtained for TKeco
measured after recooling samples to 258C. TKeco lost and TKyeast. The results obtained by the Hanes–Woolf
its activity at above 558C. In addition, Kochetov method are summarized in Table 2.
120
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
Table 2. Temperature dependence of kinetic constants for mTKgst and for TKeco
.
TK sources
T [8C]
l-Erythrulose
vmax [U·mgÀ1][a]
d-Ribose 5-P
vmax [U·mgÀ1][a]
KM [mM][a]
KM [mM][a]
mTKgst
25
50
25
5.2Æ0.2
8.0Æ0.7
18Æ2
0.97Æ0.02
6.6Æ0.1
12Æ1
0.20Æ0.03
0.13Æ0.09
1.4[11a]
1.0Æ0.03
8.4Æ0.2
50.4[11a]
TKeco
KM (mM), vmax (U·mgÀ1) and kcat values were measured by varying l-erythrulose and d-R5P concentrations in the pres-
ence of NADH and YADH by spectrophotometry at 340 nm. One unit (U) of TK releases 1 mmol of glycolaldehyde per
minute at pH 7.5 (glycylglycine buffer, 50 mM) and at the specified temperatures. Hanes–Woolf plots were used to deter-
mine vmax and KM values. For KM, the cosubstrate was applied in excess.
[a]
Substrate affinities of mTKgst for l-erythrulose and
d-ribose-5-phosphate were found to be similar inde-
pendent of the temperature and were better than
those obtained with TKeco and TKyeast. The vmax values
of mTKgst increased at 508C (6 times higher with l-er-
ythrulose and 8 times higher with d-ribose-5-phos-
phate than at 258C).
Substrate Specificity of mTKgst
Li-HPA Stability against Temperature
For synthetic purposes, we used Li-HPA as donor sub-
strate, making the reaction irreversible due to the re-
lease of carbon dioxide (Scheme 1). We studied its
stability at temperatures of 208C, 508C and 708C for
60 min in different buffers at pH 7.5 (Figure 5). At
208C (Figure 5, A) Li-HPA was stable in water and in
buffered medium. At 508C (Figure 5, B) and 708C
(Figure 5, C), Li-HPA appeared to be stable in plain
water, but gradually decomposed in buffered medium.
HPA was shown to be more stable in glycylglycine
buffer than in TEA, Hepes or phosphate buffers.
We further compared the stability of Li-HPA with
and without mTKgst in glycylglycine buffer (Figure 6).
Li-HPA disappeared much faster in the presence of
TK than in its absence. As reported in previous stud-
ies,[2b] Li-HPA decarboxylation in buffered solution
yields glycolaldehyde as the primary product, which
then reacts as an acceptor substrate for TK, giving
rise to the formation of l-erythrulose by subsequent
condensation with residual Li-HPA. The velocity of
this side reaction paralleled the increase in TK activi-
ty with increasing temperature.
Given the instability of Li-HPA in buffered solution
at temperatures above 208C, determination of TK
specific rates towards aldehydes were performed at
508C in glycylglycine buffer, in which Li-HPA degra-
dation is insignificant during the first minutes of the
reaction. In all cases, the TK rate values take into ac-
count the velocity of Li-HPA degradation in the ab-
sence of added aldehydes. Preparative scale syntheses
were performed in water at 508C, where Li-HPA is
Figure 5. Li-HPA stability against time and temperature in
~
^
the absence of TK in H2O , in glycylglycine buffer , in
&
*
TEA buffer , in KH2PO4 buffer ꢇ, and in Hepes buffer
.
(A) Stability of Li-HPA at 208C, (B) at 508C and (C) at
708C. Li-HPA concentrations (50 mM=100%) determined
against time from aliquots of medium in the presence of l-
lactate dehydrogenase as auxiliary enzyme and NADH by
spectrophotometry at 340 nm.
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
121

Juliane Abdoul-Zabar et al.
FULL PAPERS
tive to glycolaldehyde (best reference substrate for
E.coli and yeast TK[2,3]) (Table 3).
Initial rates were determined in glycylglycine buffer
at different temperatures in the presence of Li-HPA
(taking into account its limited stability) by measuring
residual Li-HPA concentrations from aliquots of the
TK-catalyzed reaction, which were analyzed spectro-
photometrically at 340 nm by the standard l-lactate
dehydrogenase (LDH) assay in the presence of
NADH (Scheme 3).
With glycolaldehyde, the mTKgst relative velocity at
508C and 658C (data not shown) was 8 and 14 times
higher, respectively, than at 208C. The value of the
relative velocity of mTKgst at 508C was 5.5 times
higher than that obtained with TKeco at 208C. The de-
termination of relative velocities for mTKgst with
other aldehydes as acceptor substrates were investi-
gated at 508C only owing to the instability of Li-HPA
Figure 6. Li-HPA stability (50 mM=100%) against time and
temperature with mTKgst, ThDP and MgCl2 in glycylglycine
~
&
^
buffer. at 208C , 508C and 708C
.
sufficiently stable. The pH can be easily adjusted con- at higher temperature. For these other substrates, the
tinuously by pH-stat control.
relative activity of mTKgst was approximately 3 times
higher than those obtained at 208C. We noted that at
208C and 508C, the mTKgst relative activities with
d,l-glyceraldehyde were 4 times lower than with
enantiomerically pure d-glyceraldehyde, which sug-
Substrate Specificity for Different Aldehydes
We determined initial rates of mTKgst for some stan- gests partial inhibition by l-glyceraldehyde as already
dard aldehydes as acceptor substrates having varied described in the literature by using other TK
carbon chain lengths and hydrophobic properties rela- sources.[2a,3b,9b,26]
Table 3. Substrate range and relative velocities[a] of mTKgst with Li-HPA as a function of temperature (normalized to TKeco
with glycolaldehyde at 208C).
Acceptor substrates
mM
208C Other TKs
208C mTKgst
508C mTKgst
Glycolaldehyde
d-Glyceraldehyde
d,l-Glyceraldehyde
Butyraldehyde
d-Ribose
50
50
100
200
100
1[a,b,c]
0.7
0.9
0.2
0.07
0.04
5.5
2.8
0.7
0.2
0.12
0.77[b]
0.55,[b] 0.37[c]
0.11[b]
0.3[b]
[a]
Own values obtained with TKeco
Literature values obtained with TKyeast
Literature values obtained with TKeco
.
[b]
[c]
[2a,3a]
.
[28]
.
Scheme 3. Irreversible reaction catalyzed by TK in the presence of hydroxypyruvate (Li-HPA) as donor and various acceptor
substrates.
122
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
The relative velocities of mTKgst increased with ing was corroborated by derivatization to l-erythru-
temperature and were significantly higher than those lose triacetate, which furnished an ee of ꢀ 95% (S) by
obtained with other TK sources (except for d-ribose) chiral HPLC.[13a,14]
showing the great utility of this enzyme for further
Using d-glyceraldehyde as acceptor complete con-
applications in biocatalysis, particularly with hydro- version occurred rapidly, and d-xylulose 2 was ob-
phobic aldehyde acceptor substrates. It is tempting to tained with an isolated yield of 38%. The optical rota-
speculate that the mTKgst enzyme should also display tion of the isolated product with an [a]2D5 of À30.9 was
a higher tolerance towards non-conventional media, comparable to that reported in the literature.[29] Cou-
including the application of larger fractions of organic pound 2 was obtained with high diastereoselectivity
1
cosolvents to aid in increasing the solubility of more (de >95%, determined by H NMR). This indicates
hydrophobic substrate analogues.
that the high stereoselectivity of mTKgst led to the for-
mation of a d-threo ketose (3S,4R) as described for
other TK enzymes with a-hydroxylated aldehydes.[2–9]
With butanal as acceptor, reaction proceeded more
slowly than with 2-hydroxylated substrates, and the
Synthetic Applications using mTKgst
To examine the synthetic potential, in particular the expected 1,3-dihydroxyhexan-2-one 3 was obtained
efficiency and the stereoselectivity of mTKgst, we per- with a 43% isolated yield. The enantiomeric excess of
formed the preparative-scale syntheses of different the product was 85% with absolute (3S)-configura-
ketoses and analogues with Li-HPA as donor and tion, determined by chiral HPLC after derivatization
with different aldehydes as acceptors. We chose three of 1 with benzoyl chloride.[15a] We conclude that
acceptor substrates tested previously at an analytical mTKgst shows higher stereoselectivity with butanal
scale (Scheme 4): two a-hydroxylated substrates, gly- than the TKeco (75% ee, 3S).[14]
colaldehyde (R=CH2OH), the best unphosphorylated
TK acceptor substrate, and d-glyceraldehyde [R=
CHOH(R)-CH2OH] and a generic non-hydroxylated Conclusions
substrate, butyraldehyde (R=CH2CH2CH3).
All reactions were performed at elevated tempera- In this work, we characterized the first TK from
ture (508C), in water (pH adjusted by addition of a thermostable Geobacillus organism expressed in E.
HCl), using equimolar concentrations of donor and coli, using a synthetic tk gene of G. stearothermophi-
acceptor in the presence of mTKgst. The conversion of lus.
Li-HPA was monitored by TLC and by enzymatic
assays on aliquots taken from the reaction mixture in the optimum temperature range of 60–708C, allowing
certain intervals. an easy purification by heat shock treatment. The
For future biocatalyst development, we identified
With glycolaldehyde as acceptor, the substrate con- thermostability of this new TK source can improve
sumption proceeded rapidly and completely to furnish the TK catalytic activity for practical applications:
the product l-erythrulose 1 with an isolated yield of mTKgst retained 100% activity for one week at 508C
36%. The optical rotation of the isolated product, and 3 days at 708C, whereas TKeco, commonly used in
[a]2D5: +6.3, was comparable to published values.[13a,14] biocatalysis, showed a gradual decrease in its activity
The product was characterized unambiguously by all already at 508C (half-life 35 min) and a total loss of
analytical data, which demonstrates that mTKgst activity within 5 min of exposure to 708C. The toler-
highly stereoselectively creates a new asymmetric ance to elevated temperature also offers higher reac-
centre with absolute (3S) configuration, in corre- tion rate: the relative velocity of mTKgst catalysis at
spondence to the well-established behaviour of TKeco 508C and 708C was 8 and 14 times higher, respective-
and TKyeast when using the same substrate. This find- ly, than at 208C, and at 508C was 5.5 times higher
than obtained with TKeco at 208C. Finally, to deter-
mine the stereoselectivity, the mTKgst was applied for
reactions performed on a preparative scale at 508C in
plain water to avoid LiHPA decomposition. Glycolal-
dehyde and d-glyceraldehyde as typical a-hydroxylat-
ed acceptor substrates, as well as n-butanal as a gener-
ic, non-hydroxylated substrate analogue, gave the cor-
responding expected products l-erythrulose 1, d-xylu-
lose 2 and 1,3-dihydroxyhexan-2-one 3, respectively,
in acceptable yields similar to those reported in the
Scheme 4. Irreversible reaction catalyzed by mTKgst in the
[2a,14,15a]
literature using TKeco
.
The major advantage of
presence of hydroxypyruvate lithium salt (Li-HPA) as donor
substrate and a-hydroxy aldehydes (glycolaldehyde, d-glyc-
eraldehyde) and generic butanal as acceptor substrates.
the mTKgst-catalyzed reactions at 508C was the short-
ened reaction time for the syntheses. The optical rota-
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
123

Juliane Abdoul-Zabar et al.
FULL PAPERS
plasmid named pGEN717 was verified. A hexa-histidine-tag
was added at the N-terminus of the enzyme for purification
purposes. The overexpression of G. stearothermophilus wild-
type TK was checked by SDS-PAGE (Figure 1). Then we
modified the coding sequence of G. stearothermophilus
wild-type TK to remove some restriction sites and introduce
new unique restriction sites for cassette mutagenesis purpos-
es. The modified nucleotides are capitalized: atggcgcatt
caatcgaggagttggcgattacgacgattcgaacgctgtcgatt
gacgcgatcgaaaaagcgaaatccgggcaTccgggcatgcc
gatgggcgcggcgccaatggcgtacacgctttggacgaaatt
tatgaatcataacccggcgaatccaaactggttcaaccgc
gaccgTtttgtcttgtcagccgggcacgggtcgatgttattgta
cagcttgcttcatttaagcggctacgacgtatcgatggat
gatttgaaacaattccgtcaatggggaagcaaaacgccgggc
catccggaatacggccatacgccgggcgtggaagcgac
gaccggcccactcggccaagggattgcgatggcggtcgg
catggcgatggcggaacggcatttggccgctacatacaaccgc
gacgggtttgagattatcaatcattatacgtacgc
tions determined for products 1 and 2 indicate that
the mTKgst shows apparent diastereospecificity with
a-hydroxylated substrates, similar to other known TK
sources; also, the enzyme retains high stereoselectiv-
ity with aliphatic analogues, leading to ketose-type
products such as 3 with absolute (3S) configuration.
Processes at elevated temperature can offer many
opportunities to improve and extend TK biocatalysis
applications, such as higher substrate solubility of hy-
drophobic aldehyde acceptors and higher tolerance
towards non-conventional media, which we will be
continuing to investigate using free and immobilized
enzyme. Other avenues of research will be the 3D
structure determination of this new TK protein and
the potential improvement of its biocatalytic proper-
ties by mutagenesis.
catttgcggcgatggcgatttgatggaaggagtggcgagc
gaagctgcgtcactcgccggccacttgaagctcggTcgAct
gatcgtcctgtatgactcgaacgacatttcgctggacggg
gagctcaacctgtcgttctcggaaaacgtcgcccaacgtttc
caagcAtacggctggcaatatttgcgcgttgaggacggcaacaa
tattgaagaaatcgccaaagcgctGgaggaggcgcgggcg
gacctcagccggccgacgctcattgaagtaaaaacgac
gattggctacggcgcgccaaataaagcgggcacgtccggcgtc
cacggtgctccgctcggcgcccaagaggcgaagctgacgaaa
gaggcgtatcgttggacatttgcggaagatttttacgtgcca
gaagaagtgtacgcccacttccgtgcgacggtgcaagagccgg
gagcgaaaaaagaggcgaaatggaatgagcagctcgccgcctat
gaacaggcccatccggaactggccgcccaattgaagcgagc
gatcgaaggcaaacttccagatggatgggaagcttctttgccgg
tatacgaagcaggcaaaagcttggcaacccgctcatcgtccggg
gaagtgatcaacgccatcgccaaagcggtAccg
Experimental Section
Strains and Chemicals
G. stearothermophilus strain 10 was purchased from DSMZ,
E. coli BL21 (DE3) Gold strain from Stratagene. Chemicals
were purchased from Sigma–Aldrich. Reagents for molecu-
lar biology were purchased from Life Technologies. En-
zymes were purchased from Thermo Fischer Scientific. Oli-
gonucleotides were purchased from Eurofins-MWG-
Operon. Hydroxypyruvate lithium salt was synthesized
chemically by a literature method.[32]
Sequence Analysis
The draft genome sequence of G. stearothermophilus is
caattgtttggcggttcggcggacttggcaagctcgaa
taaaacgctcatcaaaggcggcggcaacttcttCccggGcagc
tacgaagggcgcaacgtttggtttggcgtgcgcgagtttgc
catgggggcggcgctgaacggcatggcgcttcacggcgggct
gaaagtgttcggcggcacgttcttcgtgttctctgac
tatttgcgtccggcgatccgcttggcggcgctgatgggct
tAccggtcatctacgtcttgacgcacgacagcatcgccgtcggc
gaagacgggccgacgcacgagccgatcgaacaGc
available through a BLAST server on the University of
Oklahoma
website
(http://www.genome.ou.edu/bstear-
o.html), and no annotation was available. TK1 of B. anthra-
cis [Q81MY7] was chosen as tblastn.[33]The detected genom-
ic region of G. stearothermophilus was then analyzed to de-
termine the tk nucleotide sequence. The translation of tk se-
quence was done to provide the G. stearothermophilus TK
protein sequence. The genome analysis using G. stearother-
mophilus TK against other Geobacillus species draft ge-
nomes was done through IMG Blast server (http://img.jgi.-
doe.gov/cgi-bin/w/main.cgi).
tAgcttcgcttcgggcgatgccgaacttgtcggt
catccgtccggctgacgcaaacgaaacggcggcag
catggcggctggcgctcgaatcgacggacaagcc
gactgcgctcgtcttgacgcgtcaagatgtgccgacgttggcgg
caacagctgagttggcgtatgaaggcgtgaaaaaaggtgcA
tacgtcgtttcaccggcgaaaaacggcgctccg
gaggcgctgttgttggcgactggctcggaagtcggtctggccg
taaaagcgcaagaagcgctcgccgctgagggcatc
Cloning of the G. stearothermophilus Wild-Type and
Synthetic tk Genes
The coding sequence of G. stearothermophilus wild-type TK
was amplified by PCR with the following primers: forward
5’-AGGCGTTAATTAATGGCGCATTCAATCGAGGAGTTGG-3’
catgtctccgtcatcagcatgccatcgtgggaccgcttc
gaagcgcagccaaaatcgtaccgcgat
gaagtgcttccgccggccgtgacgaagcggctcgccatt
gaaatgggcgcgtcgctcggttgggagcgctacgtcggcgcc
gagggcgacattttggccatcgacc
gattcggtgcttccgctccgggagagaaaatcatggccgag
tatggctttacggttgacaacgtcgtccgccgca
and
reverse
5’-ATAGTTTAGCGGCCGCTTACTTGCCGAG
CAGCGCTTTTGTGC-3’ including PacI/NotI restriction sites.
This sequence was cloned into expression vector pGEN452,
derived from pET47b by SacII/AvrII digestion and introduc-
tion of the following sequence GGCCCGATCGCCGCGCGGCCG
CAAGCTTC with PvuI and NotI restriction sites for cloning
purposes. Thus the expression of the cloned gene is under
the control of T7 promotor, allowing the production of
a large quantity of enzyme. The sequence of the resulting
caaaagcgctgctcggcaagtaa.
The G. stearothermophilus tk nucleotide sequence was
used to design a synthetic G. stearothermophilus tk gene
purchased from Invitrogen. The synthetic G. stearothermo-
124
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
philus tk gene was cloned into pGEN452 by digestion of the
plasmid received from Invitrogen and subsequent ligation.
The sequence of the resulting plasmid named pGEN768 was
verified.
Determination of TK Activity
One unit of TK activity is defined as the amount of TK that
catalyzes the formation of 1 mmol of ketose product per
minute at 208C in glycylglycine buffer (50 mM, pH 7.5). TK
enzymatic assay was performed in the presence of l-erythru-
lose and d-ribose-5-phosphate (d-R5P) leading to d-sedo-
heptulose-7-phosphate (d-S7P) and glycolaldehyde (GA).[25]
The GA formed is reduced by yeast alcohol dehydrogenase
(ADH) to glycol in the presence of nicotine adenine dinu-
cleotide, reduced form (NADH). l-Erythrulose (88 mM,
10 mg·mLÀ1), d-R5P (9.0 mm, 2.4 mg·mLÀ1), ThDP (2.4 mM,
1.1 mg·mLÀ1), MgCl2 (0.5 mM, 0.1 mg·mLÀ1), NADH
(0.28 mM, 0.2 mg·mLÀ1), ADH (50 U·mLÀ1, 0.15 mg·mLÀ1)
and the TK suspension (10 mL) were added to disposable
plastic cuvettes and completed to 1 mL with glycylglycine
buffer (50 mM, pH 7.5). The disappearance of NADH was
Expression and Purification of G. stearothermophilus
TK (TKgst and mTKgst) from E. coli
Expression: Expression was carried out in E. coli BL21
(DE3) Gold strain. This strain was first transformed by elec-
troporation with pGEN717 for the expression of the wild-
type TK (TKgst) and plated on LB medium with 30 mg·mLÀ1
kanamycin. The E. coli BL21 (DE3) strain was then trans-
formed by electroporation with pGEN768 for the expression
of the mTKgst from the synthetic tk gene. A 100 mL pre-cul-
ture was grown at 378C in LB medium with 30 mg·mLÀ1 ka-
namycine overnight. The pre-culture was used to inoculate
1 liter of culture medium containing 30 mg·mLÀ1 kanamy-
cine. Isopropyl b-d-thiogalactoside (IPTG) at 1 mM was in-
troduced when the OD range was 0.7–0.8. The cells were
then grown overnight at 308C and harvested (5 g·LÀ1 of bio-
mass).
Purification by chromatography affinity: The recombi-
nant TKgst or mTKgst was purified as a 6ꢇHis-tagged fusion.
The harvested recombinant cells from 1 L culture were sus-
pended in 50 mL of 50 mM phosphate buffer containing
300 mM NaCl (pH 8.0). The cells were disrupted by sonica-
tion for 15 min on ice, and the insoluble pellet was discarded
after centrifugation (10,000 g, 20 min, 48C). The crude ex-
tract was applied to a Ni-NTA (QIAGEN) Ni2+ chelating af-
finity column equilibrated with the phosphate buffer men-
tioned above. After washing the column with the same
buffer, the 6ꢇHis-tagged recombinant protein was finally
eluted with 50 mM phosphate buffer containing 300 mM
NaCl and 250 mM imidazole (pH 8.0). The fractions con-
taining TK (estimated Bradford test and activity assay) were
collected and concentrated using an Amicon Ultra-15 centri-
fugal filter device (Millipore). Protein concentration was de-
termined by the method of Bradford, and bovine serum al-
bumin (BSA) was used as the standard. The enzyme activity
was determined by the ADH activity assay described. After
purification from 1 L culture, 160 mg of total proteins was
obtained with a total TKgst or mTKgst activity of 140 U, cor-
responding to a specific activity of 0.9 U·mgÀ1.
followed by spectrophotometry at 340 nm (value eNADH 340
=
6220MÀ1·cmÀ1).
Determination of TK activity as a function of the temper-
ature: The temperature of the assay was controlled by a Pelt-
ier effect controller. The TK sample was first incubated at
the chosen temperature for 10 min in the presence of ThDP,
MgCl2, NADH, l-erythrulose and buffer in the above con-
centrations. The reaction was started by the addition of a so-
lution of YADH and d-ribose-5-phosphate and the disap-
pearance of NADH was followed by spectrophotometry at
340 nm. At temperatures above 508C, a larger excess of the
auxiliary enzyme ADH (100 U·mlÀ1) was used to deal with
its quick loss of activity. Glycylglycine was chosen as
a buffer since l-erythrulose was instable in other buffers
such as HEPES or Tris at higher temperatures.
Determination of the Li-HPA Concentration
The Li-HPA concentration was determined by an enzymatic
assay using l-lactate dehydrogenase. Sample of Li-HPA
(20 mL) was introduced in a cuvette containing 20 mL of
NADH (10 mg·mLÀ1), 20 mL of l-lactate dehydrogenase
(100 U·mLÀ1) and 960 mL of TEA buffer 0.1M, pH 7.5. The
disappearance of NADH was followed by spectrophotome-
try at 340 nm and difference between initial absorbance and
final absorbance gave the Li-HPA concentration using the
Beer–Lambert law (value eNADH 340 =6220MÀ1·cmÀ1).
Determination of the TK Temperature, pH and
Purification of TK by heat shock: The harvested re- Thermostability Profiles
combinant cells from 1 L of culture were suspended in
The optimum temperature for TK activity was examined on
50 mL of 50 mM glycylglycine buffer (pH 7.0). The cells
were disrupted by sonication for 15 min on ice, and the in-
soluble pellet was discarded after centrifugation
(10,000 rpm, 20 min, 48C). The supernatant crude extract
was heated at 658C for 45 min. and then centrifuged at
6,000 g for 25 min. The insoluble pellet was discarded. After
this treatment of 1 L culture, 132 mg of total proteins was
obtained with a total activity of 112 U, corresponding to
a specific activity of 0.9 U·mgÀ1 at 208C. Sodium dodecyl sul-
fate-polyacrylamide gel electrophoresis (SDS-PAGE)1 was
performed on a 10% polyacrylamide gel with SeeBlueꢅ
Plus2 Pre-Stained Standard in NuPAGEꢅ MOPS SDS run-
ning buffer. Proteins were visualized by staining with Coo-
massieꢅ blue G-250. Protein determination was performed
with the Bradford test[34] using BSA as a standard.
the range of 20–758C in 50 mM Tris-HCl buffer (pH 7.5)
with l-erythrulose and d-ribose-5-phosphate as substrates.
The effect of pH on the TK activity was examined using
50 mM MES buffer (pH 5.0–6.5), HEPES buffer (pH 6.5–
7.5), and glycylglycine buffer (pH 7.0–9.0). The thermosta-
bility was studied at the range 50–658C with l-erythrulose
and R-5-P as substrates. Enzyme solutions (4.5 mg·mLÀ1)
were incubated at various temperatures in 500 mL tubes for
up to 150 h in a sand bath. The residual activities were mea-
sured at different times by the standard assay described
above.
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
125

Juliane Abdoul-Zabar et al.
FULL PAPERS
supernatant and the suspension was concentrated to dryness
under vacuum. The residue was dry-loaded onto a flash
silica column and purified by flash liquid chromatography
using dichloromethane:methanol v/v 9:1 as eluent. The frac-
tions containing the product were combined, concentrated
under vacuum, dissolved in 20 mL of water and chromato-
graphed on cation exchange resin (Dowex-50W). The frac-
tions containing the product were combined and their pH
was adjusted to 6.5 using anion exchange resin (Dowex Mar-
athon 11). The solution was filtered and concentrated under
vacuum to give d-xylulose as a colourless oil; yield: 115 mg
(38%); [a]2D5: À30.9 (c 2.0, H2O), lit. [a]2D6: À30.0 (c 2.4,
H2O).[29] NMR spectra were identical to those previously de-
scribed.[35,36] The b-anomer was the preponderant form over
the a-anomer and the keto isomer. 1H NMR (400 MHz,
D2O): d (keto isomer)=4.55 (d, J=19.4 Hz, 1H, 1-HA), 4.43
(d, J=19.4 Hz, 1H, 1-HB), 3.98 (m, 1H, 4-H), 3.98 (m, 1H,
3-H), 3.58 (m, 2H, 5-H); d (a-anomer)=4.36 (m, 1H, 4-H),
4.21–4.16 (m, 1H, 3-H), 4.14 (m, J=9.4 Hz, 1H, 5-HA), 3.79
(dd, J=9.5, 4.1 Hz, 1H, 5-HB), 3.59 (m, 2H, 1-H2); d (b-
anomer)=4.28 (dd, J=11.8, 5.6 Hz, 1H, 4-H), 4.09 (dd, J=
9.5, 6.5 Hz, 1H, 5-HA), 3.95 (d, J=5.6 Hz, 1H, 3-H), 3.55
(m, 1H, 5-HB), 3.49 (q, J=12.1 Hz, 2H, 1-H2); 13C NMR
(101 MHz, D2O): d (keto isomer)=215.70 (C-2), 83.25 (C-
3), 74.71 (C-4), 68.79 (C-1), 64.65 (C-5); d (a-anomer)=
108.49 (C-2), 78.60 (C-3), 78.06 (C-4), 74.71 (C-5), 65.18 (C-
1); d (b-anomer)=105.70 (C-2), 78.99 (C-3), 77.65 (C-4),
72.60 (C-5), 65.78 (C-1); MS (HR-ESI): m/z=149.0446,
calcd. for C5H10O5: 149.0450.
Determination of TK Kinetic Constants
KM (mM) vmax (U·mgÀ1) and kcat values were determined for
l-erythrulose and d-ribose-5-phosphate using the enzyme
assay described above. Initial velocities of TK were mea-
sured by varying l-erythrulose and d-R-5-P concentrations
in the presence of NADH and YADH by spectrophotome-
try at 340 nm. One unit (U) of TK leads to 1 mmol of glyco-
laldehyde per minute at pH 7.5 (Glycylglycine buffer
50 mM) and at various temperatures.
Hanes–Woolf plots were used to determine vmax and KM
values. For KM, the co-substrate was present in excess. All
measurements were performed in triplicate.
Synthesis of Ketoses and Analogues Catalyzed by
mTKgst
Synthesis of l-erythrulose 1: The synthesis was performed
from glycolaldehyde (GA) and hydroxypyruvate lithium salt
(Li-HPA). ThDP (66 mg, 140 mmol) and MgCl2·6H2O
(117 mg, 580 mmol) were dissolved in H2O (29 mL) and the
pH adjusted to 7.5 with 0.1M NaOH. To this stirred solution
was added mTKgst (1 mL, 9 mg of TK) and the mixture was
stirred for 20 min at 508C. In another flask, Li-HPA
(330 mg, 3 mmol) and glycolaldehyde (180 mg, 3 mmol)
were dissolved in H2O (30 mL) and the pH adjusted to 7.5
with 0.1M NaOH. Following the 20 min enzyme/cofactor
preincubation, the Li-HPA/aldehyde mixture was added to
the enzyme solution and the mixture was stirred at 508C.
The pH was maintained at 7.5 by adding 0.1M HCl using
a pH stat (Schott, Titroline alpha). The reaction was moni-
tored by titrating Li-HPA at 340 nm in the presence of l-lac-
tate dehydrogenase (3 U·mLÀ1) and NADH (0.4 mg·mLÀ1).
After total conversion of Li-HPA (1 h), silica was added and
the reaction mixture concentrated to dryness under vacuum
then dry loaded onto a flash silica column. After flash liquid
chromatography using ethyl acetate:methanol v/v 9:1 as
eluent, l-erythrulose was isolated as a yellow oil; yield:
160 mg (44%). [a]D25: +6.3 (c 1.48, H2O), lit. [a]2D5: +6.3 (c
1.1, H2O),[13a] [a]D25: +6.7 (c 1.48, H2O).[14] NMR spectra
were identical to those previously described.[13a] 1H NMR
(400 MHz, MeOD): d=4.52 (d, J=19.3 Hz, 1H, 1-HA), 4.45
(d, J=19.3 Hz, 1H, 1-HB), 4.26 (t, J=4.2 Hz, 1H, 3-H), 3.80
(dd, J=11.6, 4.5 Hz, 1H, 4-H), 3.76 (dd, J=11.6, 4.0 Hz,
1H, 4-H); 13C NMR (101 MHz, MeOD): d=212.98 (C-2),
77.82 (C-3), 67.75 (C-1), 64.86 (C-4).
Synthesis of d-xylulose 2: ThDP (44 mg, 100 mmol) and
MgCl2·6H2O (78 mg, 380 mmol) were dissolved in H2O
(19 mL) and the pH adjusted to 7.5 with 0.1M NaOH. To
this stirred solution was added mTKgst (1 mL, 9 mg of TK),
and the mixture was stirred for 20 min at 508C. In another
flask, Li-HPA (220 mg, 2 mmol) and d-glyceraldehyde
(188 mg, 2 mmol) were dissolved in H2O (20 mL) and the
pH adjusted to 7.5 with 0.1M NaOH. Following the 20 min
enzyme/cofactor preincubation, the Li-HPA/aldehyde mix-
ture was added to the enzyme solution and the mixture was
stirred at 508C. The pH was maintained at 7.5 by adding
0.1M HCl using a pH stat (Schott, Titroline alpha). The re-
action was monitored by titrating Li-HPA at 340 nm in the
presence of l-lactate dehydrogenase (3 U·mLÀ1) and NADH
(0.4 mg·mLÀ1). After total conversion of Li-HPA (3 h), pro-
teins were precipitated by adding three volumes of metha-
nol, discarded after centrifugation. Silica was added to the
Synthesis of 1,3-dihydroxyhexan-2-one 3:[15a] ThDP
(28 mg, 2.4 mM) and MgCl2·6H2O (48 mg, 9.4 mM) were
dissolved in H2O (25 mL, total volume) and the pH adjusted
to 7.5 using 0.1M NaOH. To this stirred solution was added
butyraldehyde (90 mg, 50 mm). After adjusting the pH to
7.5, mTKgst (15 mg) and Li-HPA (138 mg, 50 mm) were
added and the resulting mixture was stirred at 408C for
24 h. The pH was maintained at 7.5 throughout by adding
0.1M HCl using a pH stat. Reactions were monitored by
TLC, and after 24 h the reaction mixture was concentrated
under vacuum. Purified 1,3-dihydroxyhexan-2-one was ob-
tained as a white powder by flash column chromatography
using hexane:ethyl acetate (1:1) as eluent; yield: 72 mg
(43%). NMR spectroscopic data were identical to those de-
scribed previously.[15a] 1H NMR (300 MHz, CDCl3): d=4.48
(d, J=20.1 Hz, 1H, 1-HA), 4.42 (d, J=20.1 Hz, 1H, 1-HB),
4.32 (dd, J=8.0, 3.9 Hz, 1H, 3-H), 2.73 (br, 2H, OH), 1.75
(m, 1H, 4-H), 1.61 (m, 1H, 4-H) 1.47 (m, 2H, 5-H), 0.95 (t,
J=7.3 Hz, 3H, 6-H); 13C NMR (75 MHz, CDCl3): d=211.7
(C-2), 74.6 (C-3), 65.4 (C-1), 36.2 (C-4), 18.0 (C-5), 13.7 (C-
6).
Synthesis of 1,3-dibenzoyloxy-hexane-2-one:[15a] Diben-
zoylation
of
1,3-dihydroxyoxy-hexane-2-one
(70 mg,
1 equiv.) was carried out with benzoyl chloride (0.3 mL,
5 equiv.) and triethylamine (0.36 mL, 5 equiv.) in CH2Cl2
(5 mL) for 2 h at room temperature. After 2 h the reaction
was quenched by adding saturated NaHCO3 solution
(5 mL), and the organic phase was dried and concentrated
under vacuum. Purified 1,3-dibenzoyloxyhexane-2-one was
obtained by flash silica gel chromatography using hexa-
ne:ethyl acetate (1:1) as eluent; yield: 102 mg (56%). Enan-
tiomeric excess of the product was found to be 85%, deter-
mined by chiral HPLC. Analysis was performed on a Shi-
126
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128

Thermostable Transketolase from Geobacillus stearothermophilus: Characterization and Catalytic Properties
madzu instrument equipped with a Chiralcel IB column
(Diacel, 250ꢇ4.6 mm) using hexane:2-propanol (90:10) as
eluent. H NMR (300 MHz, CDCl3): d=8.01 (m, 4H, Ar-H),
7.40 (m, 2H, Ar-H), 7.40 (m, 4H, Ar-H), 5.4 (t, J=6.3 Hz,
1H, 3-H), 5.1 (d, J=17.4 Hz, 1H, 1-HA) 5.0 (d, J=17.4 Hz,
1H, 1-HB), 1.90 (m, 2H, 4-H), 1.50 (m, 2H, 5-H), 0.93 (t,
J=7.5 Hz, 3H, 6-H); 13C NMR (75 MHz, CDCl3): d=200.8
(C-2), 166.03 (C-1’), 165.78 (C-3’), 133.63, 133.43, 130.17,
129.92, 129.87, 129.13, 128.59, 128.45 (Ar-C), 77.11 (C-3),
66.58 (C-1), 32.89 (C-4), 18.43 (C-5), 13.72 (C-6).
[11] a) G. A. Sprenger, U. Scçrken, G. Sprenger, H. Sahm,
Eur. J. Biochem. 1995, 230, 525–532; b) G. R. Hobbs,
R. K. Mitra, R. P. Chauhan, J. M. Woodley, M. D. Lilly,
J. Biotechnol. 1996, 45, 173–179; c) E. G. Hibbert, T.
Senussi, S. J. Costelloe, W. L. Lei, M. E. B. Smith, J. M.
Ward, H. C. Hailes, P. A. Dalby, J. Biotechnol. 2007,
131, 425–432; d) S. D. Himmo, M. Thomson, C. J.
Gubler, Prep. Biochem. 1988, 18, 261–276.
1
[12] a) G. A. Kochetov, R. A. Usmanov, A. T. Mevkh, Anal.
Biochem. 1978, 88, 296–301; b) L. M. Reimer, D. L.
Conley, D. L. Pompliano, J. W. Frost, J. Am. Chem. Soc.
1986, 108, 8010–8015; c) S. Bhattacharya, M. Schiavone,
J. Gomes, S. K. Bhattacharya, J. Biotechnol. 2004, 111,
203–217; d) S. Matosevic, G. J. Lye, F. Baganz, Biotech-
nol. Prog. 2010, 26, 118–126.
Acknowledgements
[13] a) K. Benaissi, V. Hꢀlaine, V. Prevot, C. Forano, L.
Hecquet Adv. Synth. Catal. 2011, 353, 1497–1509; b) B.
OꢆSullivan, H. Al-Bahrani, J. Lawrence, M. Camposa,
A. Cazares, F. Baganz, R. Wohlgemuth, H. C. Hailes,
N. Szita, J. Mol. Catal. B: Enzymatic 2012, 77, 1–8.
[14] M. E. B. Smith, E. G Hibbert, A. B. Jones, P. A. Dalby,
H. C. Hailes, Adv. Synth. Catal. 2008, 350, 2631–2638.
[15] a) A. Cꢈzares, J. L. Galman, L. G. Crago, M. E. Smith,
J. Strafford, L. Rꢉos-Solꢉs, G. J. Lye, P. A. Dalby, H. C.
Hailes, Org. Biomol. Chem. 2010, 8, 1301–1309;
b) E. G. Hibbert, T. Senussi, M. E. B. Smith, S. J. Cost-
elloe, J. M. Ward, H. C. Hailes, P. A. Dalby, J. Biotech-
nol. 2008, 134, 240–245.
The authors thank the French National Research Agency
(grant ANR-09-BLAN-0424-CSD3 to LH) and Deutsche
Forschungsgemeinschaft (grant Fe244/9-1 to WDF) for finan-
cial support under the DeoTK research program, as well as
the ESF for support via COST CM0701. DY gratefully ac-
knowledges a PhD scholarship administered by DAAD.
References
[1] E. Racker, G. de La Haba, I. G. Leder, J. Am. Chem.
Soc. 1953, 75, 1010.
[2] a) J. Bolte, C. Demuynck, H. Samaki, Tetrahedron Lett.
1987, 28, 5525–5528; b) Y. Kobori, D. C. Myles, G. M.
Whitesides, J. Org. Chem. 1992, 57, 5899–5907; c) W.-D.
Fessner, V. Helaine, Curr. Opin. Biotechnol. 2001, 12,
574–586.
[3] a) C. Demuynck, J. Bolte, L. Hecquet, V. Dalmas, Tet-
rahedron Lett. 1991, 32, 5085–5088; b) N. J. Turner,
Curr. Opin. Biotechnol. 2000, 11, 527–531.
[16] G. Garg, S. S. Dhiman, R. Mahajan, A. Kaur, J.
Sharma, J. Biotechnol. 2011, 1, 58–64.
[17] L Cheng, W Mu, B Jiang, J. Sci. Food Agric. 2010, 8,
1327–1333.
[18] S. Ferdjani, M. Ionita, B Roy, M. Dion, Z. Djeghaba,
C. Rabiller, C. Tellier, Biotechnol. Lett. 2011, 3, 1215–
1219.
[19] W. Cuevas, D. A. Estell, S. H. Hadi, H. Sura, S. K. Lee,
S. Ramer, W. A. Shaw, A. R. Toppozada, W. Weyler,
2009, Int. Patent PCT/US2009/046034, 2009.
[20] a) J. Littlechild, N. Turner, G. Hobbs, M. Lilly, A.
Rawas, H. Watson, Acta Crystallogr. Sect. D: Biol.
Crystallogr. 1995, 51, 1074–1076; b) G. A. Sprenger,
Biochim. Biophys. Acta 1993, 1216, 307–310.
[21] a) M. Sundstrçm, Y. Lindqvist G. Schneider U. Hell-
man H. Ronne, J. Biol. Chem. 1993, 268, 24346–24352;
b) I. Schaff-Gerstenschlager, G. Mannhauft, I. Vet-
ter. F. K. Zimmermann, H. Feldmann, Eur. J. Biochem.
1993, 217, 487–492.
[22] N. Maltseva, Y. Kim, K. Kwon, A. Joachimiak, W. F.
Anderson, RCSB protein data bank (PDB entries 3H-
YL and 3M49, submitted 06/2009 and 04/2010).
[23] L. Hecquet, J. Bolte, C. Demuynck, Biosci. Biotechnol.
Biochem. 1993, 57, 2174–2176.
[4] D. C. Myles, P. J. Andrulis, G. M. Whitesides, Tetrahe-
dron Lett. 1991, 32, 4835–4838.
[5] L. Hecquet, J. Bolte, C. Demuynck, Tetrahedron 1996,
52, 8223–8232.
[6] T. Ziegler, A. Straub, F. Effenberger, Angew. Chem.
1988, 100, 737–738; Angew. Chem. Int. Ed. Engl. 1988,
27, 716–717.
[7] F. Charmantray, P. Dellis, V. Helaine, S. Samreth, L.
Hecquet, Eur. J. Org. Chem. 2006, 5526–5532.
[8] a) C. Guerard, V. Alphand, A. Archelas, C. Demuynck,
L. Hecquet, R. Furstoss, J. Bolte, Eur. J. Org. Chem.
1999, 3399–3402; b) F. T. Zimmermann, A. Schneider,
U. Schorken, G. A. Sprenger, W. D. Fessner, Tetrahe-
dron: Asymmetry 1999, 10, 1643–1646; c) O. N. Sloveje-
va, G. A. Kochetov, J. Mol. Catal. B: Enzymatic 2008,
54, 90–92; d) F. Charmantray, V. Helaine, B. Legeret, L.
Hecquet, J. Mol. Catal. B: Enzymatic 2009, 57, 6–9.
[9] a) C. U. Ingram, M. Bommer, M. E. B. Smith, P. A.
Dalby, J. M. Ward, H. C. Hailes, G. J. Lye, Biotechnol.
Bioeng. 2007, 96, 559–569; b) D. Yi, T. Devamani, J.
Abdoul-Zabar, F. Charmantray, V. Helaine, L. Hec-
quet, W. D. Fessner, ChemBioChem 2012, 13, 2290–
2300.
[24] R. K. Mitra, J. M. Woodley, M. D. Lilly, Enzyme
Microb. Technol. 1998, 22, 64–70.
[25] a) R. R. F. Jahromi, P. Morris, R. J. Martinez-Torres,
P. A. Dalby, J. Biotechnol. 2011, 155, 209–216; b) A.
Kuimov, M. Filippov, G. Kochetov, Biochem. Int. 1990,
21, 1081–1087.
[26] A. G. Datta, E. Racker, J. Biol. Chem. 1961, 236, 617–
623.
[10] C. Wikner, L. Meshalkina, U. Nilsson, M. Nikkola, Y.
Lindqvist, M. Sundstrom, G. Schneider, J. Biol. Chem.
1994, 269, 32144–32150.
[27] S. Saitou, T. Ozawa, I. Tomita, FEBS Lett. 1974, 40,
114–118.
Adv. Synth. Catal. 2013, 355, 116 – 128
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
127

Juliane Abdoul-Zabar et al.
FULL PAPERS
[28] G. R. Hobbs, M. D. Lilly, N. J. Turner, J. M. Ward, A. J.
Willets, J. M. Woodley, J. Chem. Soc. Perkin Trans.
1 1993, 54, 165–166.
[33] S. F. Altschu, T. L. Madden, A. A. Schaffer, J. Zhang,
Z. Zhang, W. Miller, D. J. Lipman, Nucleic Acids Res.
1997, 25, 3389–3402.
[34] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[35] C. Demuynck, J. Bolte, L. Hecquet, H. Samaki, Carbo-
hydr. Res. 1990, 206, 79–85.
[36] Reference spectrum from Biological Magnetic Reso-
nance Data Bank: http://www.bmrb.wisc.edu/metabolo-
mics/.
[29] H. Habe, T. Fukuoka, D. Kitamoto, K. Sakaki, J. Oleo
Sci. 2009, 58; 595–600.
[30] F. Effenberger, V. Null, T. Ziegler Tetrahedron Lett.
1992, 33. 5157–5160.
[31] V. Dalmas, C. Demuynck, Tetrahedron: Asymmetry
1993, 4, 1169–1172.
[32] F. Dickens, D. H. Williamson, Biochem. Prep. 1962, 9,
86–90.
128
asc.wiley-vch.de
ꢄ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 116 – 128