Asymmetric reduction of α-keto esters with Thermus thermophilus NADH-dependent carbonyl reductase using glucose dehydrogenase and alcohol dehydrogenase for cofactor regeneration

Article abstract of DOI:10.1002/ejoc.201100107

The enantioselective synthesis of methyl (R)-mandelate and methyl (R)-o-chloromandelate was investigated using an NADH-dependent carbonyl reductase from Thermus thermophilus (TtADH) and, separately, archaeal glucose dehydrogenase and Bacillus stearothermo

Full text of DOI:10.1002/ejoc.201100107

FULL PAPER
DOI: 10.1002/ejoc.201100107
Asymmetric Reduction of α-Keto Esters with Thermus thermophilus NADH-
Dependent Carbonyl Reductase using Glucose Dehydrogenase and Alcohol
Dehydrogenase for Cofactor Regeneration
[a]
[b]
[a]
[a]
Angela Pennacchio, Assunta Giordano, Mosè Rossi, and Carlo A. Raia*
Keywords: Asymmetric catalysis / Enzymes / Biotransformations / Ketones / Reduction
The enantioselective synthesis of methyl (R)-mandelate and
methyl (R)-o-chloromandelate was investigated using an
NADH-dependent carbonyl reductase from Thermus thermo-
the hydrophobicity of the short-chain linear alcohols em-
ployed as co-substrates of the bacillar ADH. The bioreduc-
tion of methyl benzoylformate yielded the (R)-alcohol with
philus (TtADH) and, separately, archaeal glucose dehydroge- a 77% yield (ee = 96%) using glucose dehydrogenase and
nase and Bacillus stearothermophilus alcohol dehydrogenase
BsADH) for NADH regeneration. Optimal reaction times
and substrate concentrations in the absence and presence of
organic solvents were determined. The enantiofacial selec-
tivity of TtADH was shown to be inversely proportional to
glucose, and 81% yield (ee = 94%) applying BsADH and
ethanol. The bioreduction of methyl o-chlorobenzoylformate
yielded the halogenated (R)-alcohol with 95% and 92% ee,
and 62% and 78% yield using glucose dehydrogenase and
BsADH, respectively.
(
drogenation.[7] A process has also been reported for resolu-
tion of racemic methyl o-chloromandelate using Candida
antarctica lipase A-mediated transesterification which pro-
Introduction
Optically active hydroxy esters provide very versatile
building blocks widely used as chiral intermediates for the
synthesis of pharmaceuticals and other fine chemicals.
[8]
duces 2a in 99% ee and 41% yield. More recently, Ema
[
1]
[9]
and co-workers reported the efficient and environmentally
Among them, methyl (R)-mandelate (1a) and methyl (R)-o-
chloromandelate (2a) are valuable synthons used in organic
synthesis (Scheme 1). O-protected 1a is used as an interme-
diate for the synthesis of pharmaceuticals[ and 2a is an
intermediate for the anti-thrombotic agent, (S)-clopidogrel,
commercialized under the brand name Plavix (clopidogrel
friendly chemoenzymatic synthesis of 2a from 2 in Ͼ99%
ee on a 15-g scale using recombinant Escherichia coli over-
producing a carbonyl reductase from baker’s yeast. More-
over, Jeong and co-workers have also reported the reduction
of 2a from 2 in 96.1% ee and 100% conversion using whole
2]
[10]
cells of baker’s yeast.
[
3]
sulfate). Compound 1a has been obtained by reduction of
methyl benzoylformate (1) using resting cells of Rhodotorula
[
4a]
sp. AS2.2241
and more recently on lab-scale using Sac-
charomyces cerevisiae AS2.1392 whole cells with 85.8% iso-
lated yield and ee = 96.6%.[ However, several methods
for the preparation of the key enantiomer for clopidogrel,
4b]
2a have been developed, including (i) preparation of the
corresponding (R)-carboxylic acid through fractional
crystallization after diastereomeric salt formation of (R,S)-
[
3]
o-chloromandelic acid, (ii) asymmetric reduction of the
Scheme 1. Reduction of methyl benzoylformate (1) and methyl o-
corresponding α-keto acid,[ (iii) asymmetric hydrocyan- chlorobenzoylformate (2) catalyzed by T. thermophilus ADH
5]
ation of the corresponding o-chlorobenzaldehyde using (R)- (TtADH) coupled with T. acidophilum glucose dehydrogenase
selective oxynitrilase,[6] and (iv) direct reduction of methyl
(TaGDH) for NADH regeneration.
o-chlorobenzoylformate (2) by Ru-catalyzed asymmetric hy-
A number of carbonyl reductases from different sources
have been described which catalyze the direct asymmetric
reduction of 1 to 1a with high yields and high ee, such as
NADH-dependent alcohol dehydrogenase (ADH) from
[
[
a] Istituto di Biochimica delle Proteine, CNR,
Via P. Castellino 111, 80131 Naples, Italy
Fax: +39-081-613-2277
[
11]
E-mail: ca.raia@ibp.cnr.it
Pyrococcus furiosus,
NADH-dependent α-keto ester
b] Istituto di Chimica Biomolecolare, CNR,
Via Campi Flegrei 34, 80078 Pozzuoli, Naples, Italy
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201100107.
reductase from the actinomycete Streptomices coelicolor
[
12]
A3(2),
S749.
and NADH-dependent ADH from Leifsonia sp.
[
13]
Eur. J. Org. Chem. 2011, 4361–4366
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4361

A. Pennacchio, A. Giordano, M. Rossi, C. A. Raia
FULL PAPER
More recently, an NADH-dependent, highly enantiose- of 1 proceeded in 99% after 6 h, affording the (R)-α-hy-
lective ADH (TtADH), identified from the thermophilic, droxy ester 1a with 95% ee; the yield and ee obtained using
halotolerant gram-negative eubacterium Thermus thermo- BsADH were 99% and 91%, respectively (see the Support-
[
14]
philus HB27, has been purified and characterized in our ing Information). The degree of conversion and ee of the
[
14]
laboratory.
The thermophilic enzyme catalyses the re- biotransformation was unchanged at reaction times as long
duction of α-methyl and α-ethyl benzoylformate to methyl as 24 h, as already observed for the BsADH/2-propanol sys-
[
14]
(
R)-mandelate (ee = 91%) and ethyl (R)-mandelate (ee = tem. The high conversion is noteworthy, considering that
9
5%), respectively, by way of an in situ NADH-recycling TaGDH has a marked preference for NADP(H) over
[
15]
system involving 2-propanol and Bacillus stearothermo- NAD(H).
philus ADH (BsADH).[
14]
Bioconversions were then carried out using increasingly
This paper reports the enzymatic synthesis of 1a and 2a higher concentrations of 1 for a reaction time of 24 h. The
by TtADH and the evaluation of glucose dehydrogenase as conversion proceeded with Ͼ99% at 1 to 10 gL– 1, but
well as BsADH for achieving cofactor regeneration. The decreased to ca. 10% and ca. 1% yields at 30 and 50 gL–1
work describes the determination of the optimal reaction substrate, respectively. However, the ee of 1a was 95% over
time and substrate concentration in the absence and pres- the whole range of concentrations for 1 that were examined
ence of organic solvents and the choice of the alcohol which (see Supporting Information for more details). The decrease
is most suitable co-substrate for the TtADH/BsADH sys- in conversion could be due to a decrease in solubility of the
1
tem.
substrate, enzyme inactivation by substrate or product, or
both factors. The next step was to develop a suitable solvent
system to improve bioconversions at higher substrate con-
centrations. TtADH possesses a remarkable tolerance to
common organic solvents. Indeed, it even showed a signifi-
cant increase in activity in the presence of various organic
solvents, reaching 150% and 180% of the initial value with
Results and Discussion
Process Development using Glucose Dehydrogenase for
Cofactor Regeneration
5
% v/v acetonitrile and 10% v/v 2-propanol, respectively,
Previous studies showed that the E. coli expression sys- after 24 h of incubation at 50 °C and 182% with 10% v/v
[
14]
tem developed for TtADH is quite efficient, producing ca. hexane after 65 h at 25 °C. The stability of the archaea-
0 mg protein per litre of culture, and that this enzyme bacterial glucose dehydrogenase in the presence of organic
3
shows potential for applications involving bioconversions of solvents was investigated to evaluate its tolerance to organic
substituted acetophenones and aromatic α-keto esters such solvents (see the Supporting Information). TaGDH was in-
[
14]
as 1.
Kinetic studies showed that this latter compound activated by 34% and 28% following 24 h incubation at
was a better substrate when halogenated at the ortho posi- 50 °C, in the absence and presence of 20% v/v hexane,
tion of the benzene ring. The kcat, Km and kcat/Km values respectively. Moreover, the enzyme was inactivated by over
–1
determined for the reduction of 1 were 38.1Ϯ3.7 s , 50% in the presence of 5% v/v and 10% v/v DMSO, 10%
–
1
–1
2
2
.7Ϯ0.6 mm, and 14.1 s mm , respectively and those for v/v acetonitrile or 10% v/v methanol. However, incubation
–
1
–1
–1
were 7.1Ϯ0.2 s , 0.32Ϯ0.1 mm, and 22.2 s mm , in the presence of 5% and 10% v/v 2-propanol, 5% v/v
respectively. This indicates that the electronic properties of methanol or 5% v/v acetonitrile resulted in activities that
the halogen rather than its steric effects determine the effi- were similar to those measured in aqueous buffer after 6 h
ciency of the reaction.
and 24 h. Therefore, bioconversions were carried out at dif-
Glucose dehydrogenase from Thermoplasma acidophilum ferent concentrations of 1 in the presence of acetonitrile or
(
TaGDH), a thermophilic and stable enzyme possessing hexane at concentrations of 5% and 20% v/v, respectively;
[15]
dual cofactor-specificity,
was chosen for the in situ these concentrations fulfilled the majority of the criteria for
NADH-regeneration system utilizing glucose as the hydride a solvent system in which both enzymes retained activity.
source (Scheme 1). Experimental conditions including As shown in the Supporting Information, the presence of
buffer, pH, temperature, reaction time and organic solvent water-miscible or immiscible organic solvents did not im-
were considered in order to optimise bioconversion of 1. prove conversion at concentrations of 1 higher than
–1
The optimal pH for the reduction reaction catalyzed by 10 gL , since the profiles of the conversion with acetoni-
[
14]
TtADH is approximately 6.0
and that for glucose oxi- trile and hexane were similar to those obtained in aqueous
dation catalysed by TaGDH is 7.0.[ Due to the instability buffer (Figure S2). Nevertheless, the good levels of conver-
16]
–1
of the reduced cofactor under acidic conditions, neutral pH sion obtained between 1 and 10 gL substrate was ac-
was chosen as a compromise taking into account cofactor companied by a slight decrease in enantioselectivity in the
stability and TtADH activity at suboptimal pHs. Neverthe- presence of acetonitrile (ee = 85%–89%) as well as in the
less, TaGDH retains full catalytic activity after 9 h at presence of hexane (ee = 92–93%) compared to the 95% ee
while TtADH is highly efficient and selective at obtained in aqueous buffer. The decrease in enantio-
0 °C. To establish the optimal reaction time the conver- selectivity could be related to the solvent-mediated en-
[
17]
5
5
5 °C,
[
14]
sion of 1 was carried out at 50 °C, at low substrate concen- hancement of catalysis as a result of increased flexibility of
–
1
[14]
tration (1 gL , 6.2 mm) in aqueous solution and by letting the enzyme active site.
However, an increase in ee from
the reactions proceed for 1, 3, 6 and 24 h. The conversion 37% in phosphate buffer up to 43% in the presence of ace-
4362
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4361–4366

Enzymatic Asymmetric Reduction of α-Keto Esters
tonitrile was observed for the 2-butanone to (R)-2-butanol and in the presence of 2% v/v 2-propanol.[14] The scaled-up
reduction catalyzed by L. brevis ADH.[
18]
bioreduction of 1 involving 100 mg of substrate was per-
In light of our findings related to enantioselectivity and formed by increasing the amount of 2-propanol to 4% v/v
concentration, optimum conditions for the bioconversion of and using different enzyme concentrations. The data sum-
1
5
as well as 2 were determined to involve a temperature of marized in Table 2 indicate that the improved conversion
–
1
0 °C, substrate concentrations up to 10 gL , in the pres- was obtained either in a short reaction time (6 h) using
+
ence of 1 mm NAD in pH 7.0 aqueous buffer. Table 1 sum- higher amounts of the two enzymes (Table 2, Entry 3) or
marizes the results of a study carried out for the two α- using lower amounts of enzymes at a reaction time which
hydroxy esters. The bioreduction of 100 mg of 1 at 50 °C was four times longer (Table 2, Entry 2). Notably, the yields
for 24 h gave 74 mg of 1a with ee = 95% (Table 1, Entry 1). of 96–97% were similar to those obtained from bioreduc-
[
14]
The same reaction carried out in 6 h with a 5-fold lower tions using 3 mg substrate,
but with a lower optical pu-
substrate and 2-fold higher TtADH concentrations yielded rity (86–87% ee relative to 91% ee), suggesting that the
slightly better results (Table 1, Entry 2). However, the con- change in TtADH selectivity could be related to 2-propanol
version and optical purity of 2a were somewhat lower when concentrations.
–
1
the concentration of 2 was 10 gL (Table 1, Entry 3). Nev-
ertheless, the bioreduction of 100 mg of 2 gave 86 and
Table 2. Asymmetric reduction of 1 with TtADH by the BsADH/
-propanol-NADH regeneration system.[
a]
2
6
2 mg of 2a with similar ee (Table 1, Entries 4 and 5) at
–
1
Conv.[b]
[%]
ee [b]
[%]
substrate concentrations of 2 and 5 gL , respectively. This Entry TtADH
BsADH Time 2-Propanol
–1
–1
[
mg mL ] [mgmL ]
[h]
(% v/v)
suggests that greater insolubility of the halogenated keto
ester and/or enzyme inactivation alters the efficiency of the
biotransformation. It is noteworthy that, although the affin-
ities of the two substrates are quite different, the presence
of the halogen in substrate 2 did not affect TtADH enantio- [a] Conditions: concentration of 1: 10 gL ; TtADH and BsADH
selectivity.
1
2
3
0.050
0.050
0.265
0.010
0.010
0.024
6
24
6
4
4
4
81
97
96
84
86
87
–
1
+
at the indicated concentrations; 1 mm NAD ; buffer, 0.1 m sodium
phosphate, pH 7.0, 0.1 m KCl, 5 mm 2-mercaptoethanol. Reaction
volume, 10 mL; T = 50 °C. [b] Determined by HPLC {Chiralcel
Table 1. Reduction of 1 and 2 to 1a and 2a with TtADH by the
OD-H, hexane/iPrOH [(9:1)]}.
TaGDH/glucose-NADH regeneration system.[
a]
Entry
1
mg]
[1]
[gL ]
TtADH
[mgmL ] [mgmL ]
TaGDH
Conv.[b]
[%]
ee[b]
[%]
In addition to 2-propanol (kcat/Km = 9 s–1 mm–1),
BsADH oxidizes other alcohols with even greater efficiency.
–
1
–1
–1
[
1
2
100
100
10
2
0.05
0.10
0.010
0.010
100 (74) [
d]
95
96
Examples include ethanol (k /K = 64 s mm ), 1-propa-
–1
–1
cat m
[c]
[d]
78 (77)
–1
–1
–1
–1
nol (286 s mm ), 1-butanol (437 s mm ), 1-pentanol
(64 s–1 mm ), and 1-hexanol (64 s mm ) (Raia, unpub-
lished data) and these were therefore tested as alternative
hydride sources for NADH recycling. Moreover, these
alcohols and their respective aldehydes are not substrates
–1
–1
–1
2
mg]
[2a]
–
1
[
[gL ]
[
d]
3
4
5
100
100
100
10
2
5
0.125
0.100
0.125
0.0275
0.020
0.0275
28 (n.d.)
96 (86)
100 (62)
88
93
95
[
d]
[
d]
[14]
of TtADH. Thus, only the cofactor is the co-substrate of
[
a] Reactions were carried out for 24 h at 50 °C. [b] Determined by TtADH and BsADH.
HPLC {Chiralcel OD-H, hexane/iPrOH [(9:1)]}. [c] The reaction
was carried out in 6 h. [d] Isolated yield in parentheses; n.d.: not
determined.
Bioreductions of 100 mg of 1 were carried out in the
presence of each alcohol at a concentration of 4% v/v,
which corresponds to 700 mm for ethanol down to 310 mm
for 1-hexanol. All values were over-saturating for the bacil-
lar ADH. The data obtained were plotted against the re-
spective logP values to correlate conversions and the
TtADH enantioselectivity with the hydrophobicity of the
alcohol added (see Supporting Information). The TtADH
Process Development using BsADH for Cofactor
Regeneration
Early bioconversion processes performed to establish enantioselectivity decreased as the hydrophobicity of the
TtADH enantioselectivity were carried out on analytical medium increased. However, the level of conversion re-
scale by way of an in situ NADH-recycling system involving mained high and near constant for all of the alcohols tested,
a second thermophilic NAD-dependent ADH, the recombi- emphasizing the versatility and high efficiency of the
[
14]
nant ADH from Bacillus stearothermophilus (BsADH).
BsADH/alcohol substrate system in recycling the reduced
The bacillary enzyme is active with 2-propanol which was cofactor. An increase in conversion rate with increasing
therefore used as both co-solvent and substrate. The en- logP with no effect on enantioselectivity was recently re-
zyme is inactive on aliphatic and aromatic ketones, includ- ported for the 2-octanone to (R)-2-octanol reduction with
[
19]
ing the carbonyl substrates of TtADH and the correspond- Oenococcu oeni cells in a biphasic system.
Moreover,
ing alcohols. However, TtADH does not accept 2-propanol Lactobacillus brevis ADH enantioselectivity was recently
or acetone as substrates. These features allowed the reaction shown not to be altered by the presence of an organic
[
20]
to proceed almost to completion for reduction of 3 mg of 1 phase. Instead, Thermoanaerobium brockii ADH enantio-
to 1a with 99% conversion and ee = 91%, in 24 h at 50 °C, selectivity was seen to increase proportionally to increasing
Eur. J. Org. Chem. 2011, 4361–4366
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4363

A. Pennacchio, A. Giordano, M. Rossi, C. A. Raia
Table 3. Asymmetric reduction of 1 with the TtADH/BsADH system at different ethanol and cofactor concentrations.[a]
FULL PAPER
Entry
Ethanol (% v/v)
NAD+ [mm] TtADH [mg mL–1]
BsADH
mg mL ]
Conversion [%]
ee [%]
Isolated yield
[%]
–
1
[
1
2
3
4
5
6
0.6
1
2
3
4
1 (2)
1 (2)
1 (2)
1 (2)
1 (2)
2
0.05
0.05
0.05
0.05
0.05
0.10
0.01
0.01
0.01
0.01
0.01
0.02
60 (89)
79 (98)
92 (98)
91 (98)
91 (99)
99
90 (90)
89 (92)
90 (93)
92 (92)
91 (93)
94
n.d.
n.d.
n.d.
n.d.
n.d.
81
4
–
1
–1
[
a] Conditions reaction and chiral analysis as in Table 1. Amount of 1: 100 mg. 1 concentration, 10 gL , except for #6: 2 gL . Values
+
in parentheses refer to data obtained with 2 mm NAD . n.d.: not determined.
water composition.[21] In general terms, the stereoselectivity with good productivity (Table 4). Using relatively low sub-
of enzymes decreased as solvent hydrophobicity in- strate concentrations 78 mg of 2a (ee = 92%) was obtained
[
22]
creased
and enantiofacial selectivity was also signifi- from 100 mg of 2 in 24 h at 50 °C (Table 3, Entry 3). As in
[23]
cantly affected by the reaction medium.
Although the the case for the TtADH/TaGDH system (Table 1, Entry 3)
examples mentioned describe biphasic systems, including the substrate concentration of 2 gL–1 was found to be opti-
high concentrations of an apolar solvent, the effect of rela- mal for scaling up production of 2a. However, the forma-
tively low concentrations of polar protic solvents on the tion of the more reactive acetaldehyde could limit the use
TtADH enantiofacial selectivity is rather remarkable. of ethanol as a regeneration reductant. Nevertheless, new
Docking calculations using the TtADH structure explain applications for ethanol as a tunable nicotinamide reduct-
[
24]
the selective formation of the methyl (R)-mandelate 1a
by showing that the methyl benzoylformate molecule as- zymatic synthesis of important synthons,
sumes the lowest energy orientation by fitting the phenyl vent and reductant in the enantioselective synthesis of (S)-
ant under four-electron redox conditions for the chemoen-
[
25a]
as well as sol-
[
25b]
ring into a hydrophobic pocket, with the two carbonyl profens using Sulfolobus solfataricus ADH
and (2S)-2-
[
25c]
groups staggered by about 78° and the methoxy group arylpropanols using horse liver and yeast ADHs
pointing toward the carboxyamide group of the cofactor. been recently reported.
Moreover, the docking analysis showed that the keto ester
molecule can also assume another conformation, less ener-
getically favourable compared to that described above,
which has the opposite face of the carbonyl group directed
to the nicotinamide ring, and therefore leading to the (S)-
enantiomer of the alcohol product. Importantly, there is no
major steric constraint preventing the positioning of this
have
Table 4. Reduction of 2 to 2a with TtADH by the BsADH/ethanol-
[a]
NADH regeneration system.
Entry
2
[2]
TtADH
BsADH
Conv.[b]
[%]
ee[b]
[%]
–
1
–1
–1
[
mg] [gL ] [mg mL ] [mg mL ]
1
2
3
1
2
100
1
2
2
0.05
0.05
0.10
0.01
0.01
0.02
100
100
98 (78)
91
92
92
[c]
[
24]
alternative conformation in the enzyme active site.
A
[
a] The reactions were carried out for 24 h at 50 °C. Ethanol, 4%
v/v. [b] Determined by HPLC {Chiralcel OD-H, hexane/iPrOH
(9:1)]}. [c] Isolated yield in parentheses.
change in polarity occurring in the active site may induce a
structural rearrangement that facilitates the positioning of
this alternative conformation.
[
Since the TtADH/BsADH system showed higher
enantioselectivity with ethanol than with 2-propanol, the
next step was to examine the synthesis of 1a using concen-
trations of ethanol lower than 4% v/v (Table 3). The keto
Conclusions
The enzymatic synthesis of methyl (R)-mandelate (1a)
ester reduction was limited by the cofactor recycling rate and methyl (R)-o-chloromandelate (2a), two important
when the ethanol percentage was 0.6% or 1% v/v and the pharmaceutical building blocks, has been developed in a
cofactor concentration was 1 mm (Table 3, Entries 1 and 2). one-phase system, using a carbonyl reductase and two dif-
This concentration is much lower than the cofactor Km ferent dehydrogenases to recycle the NADH. The two re-
(13Ϯ2 mm) and is therefore far below that required for sat- generation modes gave similar yields and optical purities.
uration of the BsADH active sites. However, by doubling BsADH displayed two distinct advantages: feasibility of pu-
the cofactor concentration, the conversion increased from rification and high efficiency at suboptimal pH. For the
60 to 89% (Table 3, Entry 1) and from 79 to 98% (Table 3, TaGDH method, the hydride source was glucose, which
Entry 2) in the presence of 0.6% and 1% v/v ethanol, constitutes cheap biomass whereas for the BsADH method,
respectively. Conditions employing 4% v/v ethanol and ethanol was the cheap sacrificial substrate. Both methods
2
mm NAD (Table 3, Entries 5 and 6) and a substrate con- are characterized by favourable thermodynamics since
–
1
centration of 2 gL (as in Table 1, Entry 2) were shown to alcohol/aldehyde, and glucose/gluconic acid do not interfere
be ideal for achieving asymmetric reduction of 100 mg of 1 with the synthesis reaction. Only a few examples are known
with 81% isolated yield and ee = 94% (Table 3, Entry 6). for enzyme-coupled cofactor regeneration involving a sec-
[
26]
The bioconversion of 2 to 2a occurred with the same ond ADH
and the present study represents a successful
enantioselectivity as the non-halogenated compound and application of the bacillary ADH.
4364
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4361–4366

Enzymatic Asymmetric Reduction of α-Keto Esters
TtADH/TaGDH and TtADH/BsADH system, respectively. For de-
tails on the NMR and optical rotation analyses, see the Supporting
Information.
Experimental Section
Chemicals: NAD(P)+ and NAD(P)H were obtained from Ap-
pliChem. (Darmstadt, Germany). Methyl benzoylformate 1 and
methyl (S)- and (R)-mandelate were obtained from Sigma–Aldrich.
Methyl o-chlorobenzoylformate 2 was obtained from Ricci Chimica
The degree of conversion and enantiomeric purity of the products
were determined on the basis of the peak areas of ketone substrates
and alcohol products separated and visualized by HPLC, on a Chi-
ralcel OD-H column (Daicel Chemical Industries, Ltd., Osaka, Ja-
pan). The absolute configuration of product alcohols was deter-
mined by comparing the HPLC data with standard samples. Prod-
ucts were analyzed with isocratic elution under the following condi-
(Perugia, Italy). Other chemicals were A grade substances from Ap-
+
plichem. Solutions of NAD(P)H and NAD were prepared as pre-
[
14]
viously reported. All solutions were made up with MilliQ water.
Enzymes and Kinetic Assays: Recombinant Thermus thermophilus
ADH (TtADH) and recombinant Bacillus stearothermophilus
LLD-R strain (BsADH) were prepared as described previously.[
Glucose dehydrogenase from Thermoplasma acidophilum (TaGDH)
was from Sigma, St. Luis, MO. TtADH activity was assayed spec-
trophotometrically at 65 °C by measuring the change in absorbance
of NADH at 340 nm using a Cary 1E spectrophotometer equipped
with a Peltier effect-controlled temperature cuvette holder. The ki-
netic parameters of TtADH for α-keto esters were determined as
tions: hexane/2-propanol (9:1) (mobile phase), flow rate of
14]
–1
1
mLmin , detection for bioconversions of 1 and 2 at 210 nm. At
this wavelength, both 1 and 1a as well as 2 and 2a have the same
molar extinction coefficient values, so that areas of substrates and
products are equally proportional to concentrations. Retention
times were as follows: 6.14, 8.82 and 14.15 min for 1, methyl (S)-
mandelate and methyl (R)-mandelate, respectively; 6.92, 10.12 and
17.16 min for 2, methyl (S)-o-chloromandelate and methyl (R)-o-
[
14]
described previously. TaGDH was assayed at 50 °C by measur-
ing the change in absorbance of NADPH at 340 nm. The standard
assay was performed by adding 0.25 μg of enzyme to 1 mL of pre-
heated assay mixture containing 50 mm glucose and 0.4 mm NADP
in 50 mm sodium phosphate, pH 7.0.
chloromandelate, respectively. The absolute stereochemistry of the
two halogenated alcohol enantiomers were assigned by comparison
to the values described in the literature for methyl o-chloromandel-
[9]
ates. The logP values were obtained from Laane and co-
[27]
workers.
The effect of organic solvents on TaGDH was investigated by incu-
–1
Supporting Information (see footnote on the first page of this arti-
cle): Time course of the bioreduction of 1 by TtADH. Production
of methyl (R)-mandelate 1a at different concentrations of 1. Effects
of organic solvents on TaGDH. Production of methyl (R)-mandel-
ate (1a) at different concentrations of 1 in the presence of the water-
miscible or immiscible organic solvents. Effect of the BsADH
alcohol substrates on the enantioselectivity and efficiency of the
reduction catalysed by TtADH. NMR spectra and spectroscopic
analyses. Optical rotation analyses.
bating 0.025 mgmL protein in 50 mm sodium phosphate, pH 7.0,
at 50 °C, in the absence and presence of organic solvents. At spe-
cific time intervals, the samples were centrifuged and small aliquots
were withdrawn and assayed. The volume of solution in the tight
capped test tube did not change during incubation.
Procedure for Bioreduction: Bioreduction of α-keto esters was per-
formed at 50 °C using two NADH regeneration systems. The first
consisted of TaGDH and glucose. For analytical biotransformation
the reaction mixture contained 6.2 mm carbonyl compound, 1 mm
+
NAD , 50 mm glucose, 25 μg TtADH, and 5.5 μg TaGDH in
0
.2 mL of 100 mm sodium phosphate, pH 7.0. Semi-preparative re- Acknowledgments
actions were performed on a 0.1 g scale in a reaction volume up to
0 mL, using 0.05 mgmL–1 TtADH, and 0.01 mgmL TaGDH.
–1
1
This work was funded by Fondo per gli Investimenti della Ricerca
di Base (FIRB) n. RBNE034XSW, and by the Agenzia Spaziale
Italiana (ASI), project “From Molecules to Man” (MoMan), 1/014/
06/0. We thank Prof. Tadashi Ema for the generous gift of methyl o-
chlorobenzoylformate at an early stage of the work.
During the reaction the pH was maintained at pH 6.5–7.0 with the
addition of a 2 m NaOH solution.
The second NADH regeneration system consisted of BsADH and
14]
2
-propanol or different linear alcohols as described previously[
with some modifications. The reaction mixture contained 6.2 mm
+
carbonyl compound, 1 mm NAD , 15 μg of BsADH, 0.6 to 4% v/
[
1] a) D. Zhu, Y. Yang, L. Hua, J. Org. Chem. 2006, 71, 4202–
205; b) T. Ema, H. Yagasaki, N. Okita, K. Nishikawa, T. Kor-
v alcohol substrate, and 125 μg TtADH in 1 mL of 100 mm sodium
phosphate, pH 7.0, 5 mm 2-mercaptoethanol and 100 mm KCl.
Semi-preparative reactions were performed on a 0.1 g scale in a
4
enaga, T. Sakai, Tetrahedron: Asymmetry 2005, 16, 1075–1078;
c) C. He, D. Chang, J. Zhang, Tetrahedron: Asymmetry 2008,
19, 1347–1351; d) R. Kratzer, M. Pukl, S. Egger, B. Nidetzky,
Microb. Cell Fact. 2008, 7, 37; e) K. Nakamura, R. Yamanaka,
T. Matsuda, T. Harada, Tetrahedron: Asymmetry 2003, 14,
–1
reaction volume up to 10 mL, using 0.05 mgmL TtADH, and
0
.01 mgmL–1 BsADH. In both systems the mixtures were shaken
at 160 rpm for different reaction times in a temperature-controlled
water bath. Upon termination of the reaction, the mixtures were
extracted twice with ethyl acetate, dried with anhydrous Na SO
2 4
and concentrated under reduced pressure. The samples were pre-
pared in hexane/2-propanol (9:1) for HPLC analysis.
2659–2681.
[
2] a) Y. Kobayashi, Y. Takemoto, Y. Ito, S. Terashima, Tetrahe-
dron Lett. 1990, 31, 3031–3034; b) C. Pousset, M. Haddad, M.
Larchevêque, Tetrahedron 2001, 57, 7163–7167.
[
3] A. Bousquet, A. Musolino, PCT Int. Appl., WO9918110 A1;
Chem. Abstr. 1999, 130, 296510e.
To determine the isolated yield of the semi-preparative reactions
carried out using the two NADH regeneration systems, the ethyl
acetate extracts were dried with anhydrous Na SO , filtered, and
2 4
concentrated under reduced pressure. Purification by silica gel col-
umn chromatography (hexane/ethyl acetate, 9:1) gave 1a or 2a as
colourless oils; yield: 77 mg (77% isolated yield; conversion: 78%)
and 81 mg (81%; 99%) of 1a produced by the TtADH/TaGDH and
TtADH/BsADH system, respectively; 86 mg (86% isolated yield;
conversion: 96%) and 78 mg (78%; 98%) of 2a produced by
[
4] a) W. Yang, J.-H. Xu, Y. Xie, Y. Xu, Zh. Gang, G.-Q. Lin,
Tetrahedron: Asymmetry 2006, 17, 1769–1774; b) J. L. Guo,
X. Q. Mu, Y. Xu, Bioprocess Biosyst. Eng. 2010, 33, 797–804.
5] K. Mitsuhashi, H. Yamamoto, Eur. Pat. Appl. 2003,
EP1316613 A2.
6] L. M. van Langen, R. P. Selassa, F. van Rantwijk, R. A. Shel-
don, Org. Lett. 2005, 7, 327–329.
[
[
[7] a) Y. Sun, X. Wan, J. Wang, Q. Meng, H. Zhang, I. Jiang, Z.
Zhang, Org. Lett. 2005, 7, 5425–5427; b) L. Yin, W. Shan, X.
Eur. J. Org. Chem. 2011, 4361–4366
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4365

A. Pennacchio, A. Giordano, M. Rossi, C. A. Raia
FULL PAPER
Jia, X. Li, A. S. C. Chan, J. Organomet. Chem. 2009, 694,
2092–2095.
[18] J. Shumacher, M. Eckstein, U. Kragl, Biotechnol. J. 2006, 1,
574–581.
[
[
[
[
[
8] K.-N. Uhm, S.-J. Lee, H.-K. Kim, H.-Y. Kang, Y. Lee, J. Mol. [19] J. Hu, Y. Xu, Biotechnol. Lett. 2006, 28, 1115–1119.
Catal. B 2007, 45, 34–38.
[20] M. Villela Filho, T. Stilliger, M. Müller, A. Liese, C. Wandrey,
Angew. Chem. Int. Ed. 2003, 4, 2993–2996.
[21] Å. Jönsson, E. Wehtje, P. Adlercreutz, B. Mattiasson, Biochim.
Biophys. Acta 1999, 1430, 313–322.
[22] G. Carrea, S. Riva, Angew. Chem. 2000, 39, 2226–2254.
[23] A. M. Klibanov, Nature 2001, 409, 241–246.
[24] A. Pennacchio, A. Giordano, L. Esposito, E. Langella, M.
Rossi, C. A. Raia, Protein Pept. Lett. 2010, 17, 437–444.
[25] a) S. Broussy, R. W. Cheloha, D. B. Berkowitz, Org. Lett. 2009,
11, 305–308; b) J. A. Friest, Y. Maezato, S. Broussy, P. Blum,
D. B. Berkowitz, J. Am. Chem. Soc. 2010, 132, 5930–5931; c)
P. Galletti, E. Emer, G. Gucciardo, A. Quintavalla, M. Pori, D.
Giacomini, Org. Biomol. Chem. 2010, 8, 4117–4123.
[26] K. Goldberg, K. Schroer, S. Lütz, A. Liese, Appl. Microbiol.
Biotechnol. 2007, 76, 237–248.
9] T. Ema, S. Ide, N. Okita, T. Sakai, Adv. Synth. Catal. 2008,
350, 2039–2044.
10] M. Jeong, Y. M. Lee, S. H. Hong, S. Y. Park, I. K. Yoo, M. J.
Han, Biotechnol. Lett. 2010, 32, 1529–1531.
11] D. Zhu, H. T. Malik, L. Hua, Tetrahedron: Asymmetry 2006,
17, 3010–3014.
12] K. Ishihara, H. Yamaguchi, T. Omori, T. Uemura, N. Naka-
jima, N. Esaki, Biosci. Biotechnol. Biochem. 2004, 68, 2120–
2127.
[
[
[
[
13] K. Inoue, Y. Makino, N. Itoh, Tetrahedron: Asymmetry 2005,
16, 2539–2549.
14] A. Pennacchio, B. Pucci, F. Secundo, F. La Cara, M. Rossi,
C. A. Raia, Appl. Environ. Microbiol. 2008, 74, 3949–3958.
15] J. John, S. J. Crennell, D. W. Hough, M. J. Danson, G. L. Tay-
lor, Structure 1994, 2, 385–393.
16] J. Woodward, K. A. Cordray, R. J. Edmonston, M. Blanco-Ri-
vera, S. M. Mattingly, B. R. Evans, Energy Fuels 2000, 14, 197–
[27] C. Laane, S. Boeren, R. Hilhorst, C. Veeger, Optimization of
Biocatalysis in Organic Media, in: Biocatalysis in Organic Me-
dia (Eds.: C. Laane, J. Tramper, M. D. Lilly), Elsevier, Amster-
dam, 1987, vol. 29, pp. 65–84.
201.
[17] L. D. Smith, N. Budgen, S. J. Bungard, M. J. Danson, D. W.
Received: January 25, 2011
Hough, Biochem. J. 1989, 261, 973–977.
Published Online: June 1, 2011
4366
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 4361–4366