An enzymatic toolbox for the kinetic resolution of 2-(Pyridin-x-yl)but-3- yn-2-ols and tertiary cyanohydrins

Article abstract of DOI:10.1002/ejoc.201000193

The kinetic resolution of a series of acetates of tertiary alcohols bearing a nitrogen substituent has been studied by using several recombinant carboxyl esterases and variants thereof expressed in E. coli. Most of the enzymes were active in the conversion of these tertiary alcohols and excellent enantioselectivities were achieved in the synthesis of three 2-(pyridin-x-yl)but-3-yn-2-ols with the nitrogen atom in the pyridine ring in the 2′-, 3′-, and 4′-positions. The resolution of tertiary cyanohydrins proved to be more difficult as the enantioselectivity of the enzymes was generally lower. Nevertheless, (S)-1-cyano-2,2,2-trifluoro-1- phenylethyl acetate was obtained with 99% ee. The results show that the limited substrate range of the individual enzymes in the synthesis of a series of tertiary alcohols can be efficiently overcome by using a combination of different enzymes.

Full text of DOI:10.1002/ejoc.201000193

FULL PAPER
DOI: 10.1002/ejoc.201000193
An Enzymatic Toolbox for the Kinetic Resolution of 2-(Pyridin-x-yl)but-3-yn-
2-ols and Tertiary Cyanohydrins
Giang-Son Nguyen,[a,b] Robert Kourist,[a] Monica Paravidino,[b] Anke Hummel,[a]
Jessica Rehdorf,[a] Romano V. A. Orru,[c] Ulf Hanefeld,*[b] and Uwe T. Bornscheuer*[a]
Keywords: Alcohols / Kinetic resolution / Cyanohydrins / Enzymes / Enantioselectivity
The kinetic resolution of a series of acetates of tertiary
alcohols bearing a nitrogen substituent has been studied by
using several recombinant carboxyl esterases and variants
thereof expressed in E. coli. Most of the enzymes were active
in the conversion of these tertiary alcohols and excellent
enantioselectivities were achieved in the synthesis of three
2-(pyridin-x-yl)but-3-yn-2-ols with the nitrogen atom in the
pyridine ring in the 2Ј-, 3Ј-, and 4Ј-positions. The resolution
of tertiary cyanohydrins proved to be more difficult as the
enantioselectivity of the enzymes was generally lower. Nev-
ertheless, (S)-1-cyano-2,2,2-trifluoro-1-phenylethyl acetate
was obtained with 99% ee. The results show that the limited
substrate range of the individual enzymes in the synthesis of
a series of tertiary alcohols can be efficiently overcome by
using a combination of different enzymes.
necessary. In addition, the stereodifferentiation becomes
considerably more difficult when ketones are the sub-
strates.[12,13] The alternative chemical asymmetric syntheses
only give reasonable results with methyl ketones and are
thus rather limited.[6,9] Ketone-derived cyanohydrins are
tertiary alcohols. The enantioselective hydrolysis of their es-
ters by carboxyl esterases is a promising alternative to the
addition of HCN to ketones as the starting compounds are
easy to prepare.[14,15] When a neutral pH is used in the reac-
tion, the cyanohydrin formed decomposes into the corre-
sponding ketone and HCN. The resulting ketones can be
reused for the synthesis of the corresponding racemic esters.
Therefore no racemization is needed in the kinetic resolu-
tion. Previous attempts with commercially available lipases,
proteases,[15] and whole-cell catalysts[16–18] have met only
moderate success so far or have been limited to aliphatic
tertiary cyanohydrins.
In the last few years, the kinetic resolution of racemic
tertiary alcohol acetates catalyzed by hydrolases has been
established as an attractive method for the synthesis of terti-
ary alcohols.[8,15,19,20] In particular, the esterase-catalyzed
hydrolysis of esters of tertiary alcohols is a highly enantio-
selective route to optically pure building blocks.[8,15] Only a
very few commercially available enzymes, however, convert
these sterically very demanding substrates. We have iden-
tified a series of esterases with confirmed activity towards
tertiary alcohols, some of them showing excellent enantio-
selectivity towards a range of compounds.[20–22] All these
esterases share a motif of three glycines or two glycines and
an alanine in their active sites. This so-called GGG(A)X
motif has been related to the ability of the esterases to con-
vert tertiary alcohols.[22] The substrate specificity of the es-
terases, however, is very high and seems to be affected by
Introduction
Optically pure tertiary alcohols and tertiary cyanohyd-
rins are versatile building blocks in organic synthesis. For
instance, compound 1c has been used in the synthesis of an
A2A receptor antagonist that was shown to be orally active
in a mouse catalepsy model[1] or the acetylene moiety of 1a
can be used to synthesize useful heterocyclic structures (see
Scheme 1).[2] The precursor for the synthesis of anticonvul-
sants can also be obtained from 1i.[3] Moreover, optically
pure tertiary alcohols have recently received attention as
potential inhibitors of HIV protease[4] and reverse tran-
scriptase.[5] The synthesis of these compounds by both
chemical and biocatalytic routes is challenging and no stan-
dard method has been developed so far.[5–9]
The multitude of applications for aldehyde-derived cya-
nohydrins has been reviewed recently.[10,11] A straightfor-
ward biocatalytic approach for their preparation involves
the addition of HCN to aldehydes catalyzed by hydroxyni-
trile lyases. Ketone-derived cyanohydrins have fewer appli-
cations owing to their difficult synthesis. Indeed, in the
analogous synthesis of tertiary cyanohydrins by the ad-
dition of HCN to ketones, the equilibrium is on the side of
the substrates. This makes the use of a large excess of HCN
[a] Institute of Biochemistry, Department of Biotechnology
& Enzyme Catalysis,
Felix-Hausdorff-Str. 4, 17487, Greifswald, Germany
E-mail: uwe.bornscheuer@uni-greifswald.de
[b] Gebouw voor Scheikunde, Afdeling Biotechnologie,
Technische Universiteit Delft,
Julianalaan 136, 2628 BL, Delft, The Netherlands
[c] Department of Chemistry, Vrije Universiteit,
De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201000193.
Eur. J. Org. Chem. 2010, 2753–2758
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2753

U. Hanefeld, U. T. Bornscheuer et al.
FULL PAPER
Scheme 1. Kinetic resolution of tertiary alcohol acetates 1a–j. Compounds 1k and 1l are shown for comparison.
steric and electronic effects caused by minor changes in the sponding racemic acetates 1a–j. For all compounds a
substrate structure.[20,21] We attempted to overcome this method for the separation of their enantiomers by GC
limitation by providing a wide diversity of different analysis was established (see the Supporting Information).
GGG(A)X esterases consisting of novel biocatalysts from The absolute configurations of the cyanohydrins 1f–h were
the metagenome[21] and highly enantioselective esterase recently determined and the configurations of 1i and 1j
variants created by protein design.[19,23,24] Esterase BS2 were assigned on the basis of their elution order in chiral
from Bacillus subtilis has previously been shown to be GC analysis.[15,18]
highly enantioselective towards tertiary alcohol acetates.[20]
The acetates were subjected to analytical-scale kinetic
The enzyme variants BS2 G105A with increased enantio- resolution using nine esterases and variants thereof. The re-
selectivity[19] and BS2 E188W/M193C with inverse enantio- actions were carried out at pH 7.5 and 37 °C in the presence
preference[24] were created by protein design. The metagen- of an appropriate cosolvent (Scheme 1). Aryl-substituted al-
ome-derived esterase Est8[25] and esterase PEstE from the iphatic tertiary alcohol acetate 1k undergoes a nonenzy-
thermophilic archeon Pyrobaculum calidifontis[26] are both matic SN1-type hydrolysis in aqueous systems.[28] Even
highly enantioselective in the kinetic resolution of tertiary though this can be suppressed to a certain degree by the
alcohols bearing a tert-butylcarbamoyl substituent. A novel addition of water-miscible cosolvents such as DMSO,[23] it
esterase from Pseudomonas putida has recently been cloned still lowers the effective enantioselectivity of the kinetic res-
and overexpressed.[37] It was identified by the presence of olution. The trifluorinated analogue 1l is stable in buffer
the GGG(A)X motif as being active towards tertiary because of the electron-withdrawing effect of the trifluoro-
alcohol acetates. Recombinantly expressed isoenzymes of methyl group.[28] Pyridyl-substituted aliphatic tertiary
pig liver esterase (PLE) offer several advantages over the alcohols can also be expected to have a higher stability in
use of the commercially available crude extract, which con- aqueous systems due to a lower electron density in the pyr-
sists of a mixture of different isoenzymes, such as a higher idine ring. Indeed, none of the acetates 1a–d underwent
enantioselectivity, a higher reproducibility, and fewer regu- measurable autohydrolysis during the reaction; neither did
latory issues.[27] The sequences of the two isoenzymes γPLE the cyanohydrins 1f–j. As expected, the products of the ki-
and PLE5 differ only in 21 amino acids, but this leads to a netic resolutions of 1f–j decomposed to the corresponding
striking change in enantioselectivity and enantiopreference ketones under the screening conditions.
towards the acetates of secondary alcohols.
In the work reported herein, we prepared a series of terti-
Activity of the Enzymes
ary alcohols containing a nitrogen atom in various posi-
tions, including pyridine-substituted aryl-substituted ali-
phatic tertiary alcohols and tertiary cyanohydrins
(Scheme 1). The effect of variation of the position of a ni-
trogen heteroatom on the enantioselectivity of the different
enzymes is discussed.
The kinetic resolution of 1a–d was investigated to study
the effect of the pyridine substituent (R1, Scheme 1) on the
activity and enantioselectivity of the esterases. Acetate 1e
was included for a better understanding of the influence of
R2 and to facilitate a comparison with 1d and 1l.[20] The
screening of 1f–j allowed the performance of the enzymes to
be determined in the conversion of tertiary alcohol acetates
bearing a CN substituent in position R3. Table 1 summa-
rizes the enantioselectivity of the esterases towards 1a–1j.
For details, the reader is referred to the Supporting Infor-
Results and Discussion
Substrate Synthesis, Analysis, and Stability
The racemic tertiary alcohols 2a–j were prepared from mation. All wild-type enzymes showed activity towards at
the corresponding ketones by Grignard reaction or addition least some of the substrates (Table 1). The enzyme activity
of TMSCN. Subsequent acetylation yielded the corre- towards tertiary alcohol acetates was determined relative to
2754
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2753–2758

Kinetic Resolution of Pyridinylbutynols and Cyanohydrins
Table 1. Enantioselectivity of GGG(A)X esterases towards tertiary alcohol acetates 1a–1j.
the activity towards p-nitrophenyl acetate (pNPA), for ex- The double bond CH2=CH in 1d is less rigid than the triple
ample, the number of units necessary to reach 50% conver- bond CHϵC in 1c and this can lead to a different behavior
sion in 1 h. The differences between the activities of the of the compounds in the active site.
GGG(A)X esterases towards the structurally related com-
Previously, we modeled the substrate recognition of BS2
pounds were striking. For instance, application of 0.5 U of esterase in molecular dynamics studies.[23,24] The active site
PPE led to 58% conversion in the resolution of 1b whereas of this esterase is large and allows for the accommodation
126 U of BS2 were necessary to achieve a similar conver- of the substrate in several binding modes, for which one
sion. Wild-type esterase BS2 and its mutant BS2 G105A productive configuration for each enantiomer could be
converted the pyridyl aliphatic tertiary alcohol acetates 1a– identified. Given the large size of the active center, it is sur-
d, whereas double mutant E188W/M193C showed no ac- prising that a subtle electronic change, such as the differ-
tivity. A ten-fold greater amount of mutant BS2 G105A was ence between a 2- and 3-pyridyl moiety in 1a and 1b, in the
necessary to resolve the pyridyl aliphatic compounds 1a– substrate structure might have such a strong impact on the
c (around 60 U) in comparison to the phenyl-substituted activity and enantioselectivity of the enzyme. Several other
aliphatic 1l (6 U).[20,24] The same holds true for the vinyl esterases of the series also show remarkable differences in
analogue of 1l, 1e (56 U). Both effects are combined in 1d, enantioselectivity between 1a and 1b. This led us to con-
with 160 U being necessary to achieve 50% conversion. In clude that the substrate recognition of the esterases is
the case of the aryl-substituted aliphatic tertiary alcohol strongly influenced by electronic interactions rather than
acetates 1a–1e, the position of the nitrogen atom has a mere steric effects.
strong effect on the enantioselectivity of the esterases. Ester-
The effect of electronic interactions on the enantio-
ase variant BS2 G105A resolved 1a–1c and 1e with moder- selectivity of carboxyl hydrolases was observed by Rotticci
ate-to-excellent enantioselectivity. Est8, PPE, and Est56 et al.[29] when they studied the kinetic resolution of second-
showed moderate enantioselectivity towards 1a, 1b, and 1e, ary alcohols bearing methyl and bromo substituents cata-
respectively (Table 1). With substrates 1a–1c, the difference lyzed by lipase B from Candida antarctica. Similar catalytic
in enantioselectivities is assumed to reside not in the size behavior has previously been observed with GGG(A)X hy-
effect of the substituents but in the electronic effects as the drolases[14,20,21] and commercially available lipases,[29] pro-
three compounds share a pyridine substituent of nearly teases,[15] and whole-cell biocatalysts[16–18] and seems to be
identical size. This phenomenon has also been observed be- the general rule in the conversion of these very difficult sub-
fore with lipases in the kinetic resolution of secondary strates.
alcohols.[29,30] Furthermore, the effect could arise from dif-
Interactions with the solvent may also account for the
ferences in the interaction of the nitrogen atom of the pyr- differences in substrate recognition in the cases of both pyr-
idine ring with residues or water molecules in the active site idine/phenyl and ethynyl/nitrile substituents. Savile and
of enzymes. Zhu et al.[31] observed this kind of effect in the Kazlauskas[32] demonstrated that the solvation of the sub-
reduction of aryl ketones with ketoreductases. Also, in the strate contributes to the enantiorecognition of the protease
revised rule to predict the enantioselectivity in the kinetic subtilisin. They showed an inversion of the enantioprefer-
resolution of secondary alcohols, Savile and Kazlauskas[32] ence towards some aryl-substituted aliphatic secondary
described how substrate solvation could strongly effect the alcohols in the hydrolysis in buffer and in the esterification
enantioselectivity.
in organic solvent. This switch was not observed for sub-
Surprisingly, the enantioselectivity dropped dramatically strates bearing a pyridine N-oxide substituent (Figure 1).
in the conversion of 1d in comparison to its analogue 1c
The active site of GGG(A)X esterases is generally open
despite the small difference in the R3 substituents. This can so that the tetrahedral intermediate is exposed to the sol-
be explained by the difference in the nature of bond: vent. Molecular modeling[23,24] suggested that both enantio-
Eur. J. Org. Chem. 2010, 2753–2758
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2755

U. Hanefeld, U. T. Bornscheuer et al.
FULL PAPER
alcohol acetates 1a–e but only Est8 and PEstE were enan-
tioselective towards the cyanohydrins 1f–j. Pig liver esterase
did not show any activity towards the tertiary cyanohydrins.
Interestingly, Est8 and PEstE displayed an inverse enantio-
preference. Small changes in substrate structure had a con-
siderable effect on the enantioselectivity of the esterases.
For instance, by changing the p-chloro substituent on the
phenyl ring (1g) to p-methoxy (1h), the eeS value from
PEstE decreased significantly from 91% ee (at 67% conver-
sion) to 7% ee (at 71% conversion), respectively. Est8 also
behaved differently in terms of enantioselectivity with 1f
and 1h. Adding a p-methoxy group to the phenyl ring in-
creased the enantioselectivity from 78% ee for 1f to 91% ee
for 1h at similar conversions. This narrow substrate scope
has also been observed with other enzymes used in the
screening (Table 1). Est8 showed a promising E value (E =
10) in the kinetic resolution of 1i at 37 °C, which could be
increased to E = 18 by lowering the temperature to 4 °C
(see the Supporting Information). The enantioselectivity of
Est8 was strongly influenced by the choice of cosolvent
[10% (v/v) of 2-propanol, toluene, dimethylformamide
(DMF), or dioxane]. The addition of DMF led to an in-
crease in the E value from 18 [10% (v/v) DMSO] to 116
[10% (v/v) DMF] at 4 °C. The influence of water-miscible
cosolvents on the enantioselectivity of esterases has been
known for a long time.[33] It is a complex phenomenon and
has been attributed to conformational changes in the bio-
catalyst but also to an improved solubility of the sub-
strate[33] and hence the effect is difficult to predict. In some
cases, adding[23,30] or changing a cosolvent[34] could lead to
dramatic changes (even inversion) in the enantioselectivity.
An increase in the enantioselectivity of Est8 towards ter-
tiary alcohols after addition of DMF has been observed
previously.[25] Surprisingly, the application of the optimized
conditions for 1i to other cyanohydrins did not show any
improvement in enantioselectivity. This might be explained
by the overall different reactivity of the CF3-substituted cy-
anohydrin ester 1i. Indeed, all the enzymes tested in this
study had a remarkably lower activity towards 1i, which
confirms the results of a previous study.[15] The strong effect
of the CN substituent on the enantioselectivity of esterases
has already been observed for esterase BS2. A molecular
dynamics study indicated that the main difference between
tertiary alcohols with an ethynyl substituent and tertiary
cyanohydrins lies in the ability of the nitrile group to form
hydrogen bonds with water molecules in the active site of
the enzyme.[14]
Figure 1. Accommodation of (R)-1l and (S)-1l in the active site of
BS2 esterase.[24] A similar accommodation of pyridine-substituted
substrates such as (S)-1c would result in an entirely different sol-
vation of the substrate.
mers have an inverse orientation and that the phenyl ring
of the S enantiomer is oriented towards the entrance of the
active site. It is easy to imagine that the interactions of the
pyridine ring of 1a–c with the solvent would be quite dif-
ferent to that of a phenyl ring. Also, the position of the
nitrogen atom in the ring might influence the solvation.
Furthermore, molecular modeling indicated that the main
difference in substrate recognition of esterase BS2 between
tertiary alcohols with an ethynyl substituent and tertiary
cyanohydrins lies in the ability of the nitrile group to form
hydrogen bonds with water molecules in the active site of
the enzyme.[14] These observations suggest that the sol-
vation of solvent-exposed groups may be an important fac-
tor in the enantiorecognition of esterases.
The isoenzymes of pig liver esterase differ in their
enantioselectivities towards 1a–1c, whereas with 1d and 1e
there was no conversion. Interestingly, an inverse enantio-
preference of the two PLE isoenzymes was observed in the
case of 1c. This demonstrates, as has already been shown
before for the resolution of secondary alcohol acetates,[27]
how isoenzymes with high sequence similarity can vary in
their properties thus broadening the substrate scope. More-
over, the isoenzyme γPLE showed excellent and good
enantioselectivity towards compounds 1a and 1c, respec-
tively.
The effect of the CϵCH or CN substituent on the
enantioselectivity was strong. This has also been observed
previously with commercially available lipases, proteases,[15]
and microbial whole-cell catalysts.[16,17] Most of the en-
zymes showed at least some selectivity towards the tertiary
Table 2. Kinetic resolution of tertiary alcohol acetates on a preparative scale.
Entry
Compound
Enzyme
Conv. [%]
% eeS[a]
Yield [%]
% eeP[a]
Yield [%]
E[b]
1
2
3
4
5
1a[c]
1b[c]
1c[c]
1e[c]
1i[d]
γ PLE
BS2 G105A
BS2 G105A
BS2 G105A
Est8
46
55
52
65
52
98
Ͼ99
Ͼ99
90
31
35
36
36
38
97
86
90
95
n.d
30
34
30
31
n.d
Ͼ100
51
99
Ͼ100
Ͼ100
Ͼ99
[a] Determined by chiral GC. [b] Calculated from eeS and eeP. [c] Conditions: 10% DMSO, 37 °C, 1 h reaction time. [d] Conditions:
10% DMF, 4 °C, 4 h reaction time.
2756
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2753–2758

Kinetic Resolution of Pyridinylbutynols and Cyanohydrins
added to acetate 1a–i (12 µmol) dissolved in cosolvent (100 µL) to
form a total volume of 1 mL. The reaction mixture was shaken in
a Thermoshaker (Eppendorf, Germany) at 37 °C for a certain time
(0.5, 1, and 4 h) after which a sample (300 µL) was removed. The
sample was extracted twice with dichloromethane (2ϫ400 µL). The
combined organic layers were dried with Na2SO4, filtered, and
transferred to a GC vial. The enantioselectivities and conversions
were calculated according to Chen et al.[36] The E values of com-
pounds 1a–d were calculated based on values of eeS and eeP. In
the case of compounds 1e–i, the products released by the enzymes
The kinetic resolution of 1a, 1b, 1c, 1e, and 1i was also
performed in experiments on the 100 mg scale. The reaction
products were isolated in moderate-to-good yields and with
excellent enantiopurity (Table 2). Owing to the low activity
of the esterases towards the tertiary alcohol acetates, con-
siderable amounts of biocatalyst were needed. This was par-
ticularly the case with Est8, for which the low reaction tem-
perature made a greater amount of enzyme necessary. Note,
however, that these esterases can be easily synthesized by
recombinant expression and consequently sufficient decomposed immediately to the corresponding ketone and HCN;
therefore their E values were calculated from eeS values and the
conversions.
amounts can be produced easily.
Preparative-Scale Esterase-Catalyzed Resolution: An appropriate
amount of esterase dissolved in phosphate buffer (29.7 mL,
100 m, pH 7.5) was added to a solution of the acetate (0.4 mmol)
in DMSO or DMF (3.3 mL) to give a total volume of 33 mL. The
solution was stirred for 4 h at 37 °C and the product was extracted
three times with dichloromethane (3ϫ300 mL). After drying with
Na2SO4, the solvent was removed by distillation under reduced
pressure and the products were purified by column chromatog-
raphy.
Conclusions
Several aryl-substituted aliphatic tertiary alcohols, in-
cluding tertiary cyanohydrins, were resolved in good yields
and with very high enantiopurity by GGG(A)X esterases
although the resolution of the cyanohydrins proved to be
more difficult. The substrate range of the individual en-
zymes in the synthesis of a given series of tertiary alcohols
is very narrow. This study shows how this can be efficiently
overcome by a combination of different enzymes. In ad-
dition, nonsteric interactions, especially electronic effects,
can have a strong influence on the enantioselectivity of en-
zymes. A possible explanation for this observation is the
interaction of enzyme-bound substrates with the solvent.
2-(2-Pyridyl)but-3-yn-2-yl Acetate (1a): The reaction was carried
out by following the general procedure (46% conversion) using
4700 U of esterase γPLE and 10% (v/v) DMSO. Acetate (+)-1a
was obtained as a yellow liquid after column chromatography using
aluminum oxide and n-hexane/ethyl acetate (15:1) as eluent (32 mg,
0.17 mmol, 31%, 98% ee). [α]2D0 = +92.9 (c = 1.25, CHCl3). The
GC–MS and NMR analyses matched the data given in the Sup-
porting Information. Alcohol (+)-2a was obtained as a brown li-
quid (24 mg, 0.16 mmol, 30%, 97% ee). [α]2D0 = +119.9 (c = 1.35,
CHCl3). The GC–MS and NMR analyses matched the above given
data.
Experimental Section
Characterization data for all of the compounds, including chiral
analysis, spectroscopic, and analytical data, and general details of
the equipment used are provided in the Supporting Information
2-(3-Pyridyl)but-3-yn-2-yl Acetate (1b): The reaction was carried
out by following the general procedure (55% conversion) using
4700 U of esterase BS2 G105A and 10% (v/v) DMSO. Acetate
(–)-1b was obtained as a brown liquid after column chromatog-
raphy using aluminum oxide and n-hexane/ethyl acetate (3:1) as
eluent (35 mg, 0.18 mmol, 35%, 99% ee). [α]2D0 = –53.8 (c = 1.00,
CHCl3). The GC–MS and NMR analyses matched the data given
in the Supporting Information. Alcohol (+)-2b was obtained as a
yellow liquid (27 mg, 0.18 mmol, 34%, 86% ee). [α]2D0 = +5.2 (c =
0.92, CHCl3). The GC–MS and NMR analyses matched the above
given data.
General Procedure for Expression of Recombinant Esterases:[35] E.
coli strain BL21 was used as host for the transformation of the
appropriate plasmid DNA.[19,23] The strains were grown in 400 mL
LB liquid media supplemented with the appropriate antibiotics
(100 µgmL–1) ampicillin for BS2 variants[24] and PestE,[26] ad-
ditionally 50 µgmL–1 kanamycin for PPE and the PLE variants[27]
at 37 °C and supplemented with the appropriate inductor to induce
esterase production at an optical density of 0.5 at 600 nm. Cultiva-
tion continued for 4 h at the appropriate cultivation temperature.
Cells were collected by centrifugation (15 min, 4 °C, 4000 g) and
washed twice with sodium phosphate buffer (100 m, pH 7.5,
4 °C). Cells were disrupted by using a French press (French Pres-
sure cell press, Thermo spectronic, USA). Production of BS2 vari-
ants gave after induction with rhamnose (0.2% v/v) and 4 h growth
at 37 °C 32000 units of wild-type BS2, 6000 units (pNPA assay) of
BS2 G105A, and 5400 units of BS2 E188W/M193C. Production of
PestE gave after induction with IPTG (1 m) and 4 h growth at
37 °C and purification by heat-precipitation after cell disruption
8 000 units. Production of PPE gave after induction with IPTG
(1 m) and 4 h growth at 20 °C 3900 units. Production of PLE
isoenzymes gave after induction with IPTG (1 m) and 4 h growth
at 30 °C 63 units of isoenzyme PLE5 and 77 units of the γ-isoen-
zyme. The metagenomic esterases were produced by BRAIN AG
and used as glycerol-stabilized crude cell extracts or lyophylisate.
2-(4-Pyridyl)but-3-yn-2-yl Acetate (1c): The reaction was carried
out by following the general procedure (52% conversion) using
3000 U of esterase BS2 G105A and 10% (v/v) DMSO. Acetate (–)-
1c was obtained as a brown liquid after column chromatography
using aluminum oxide and n-hexane/ethyl acetate (3:1) as eluent
(36 mg, 0.19 mmol, 36%, 99% ee). [α]2D0 = –40.1 (c = 0.95, CHCl3).
The GC–MS and NMR analyses matched the data given in the
Supporting Information. Alcohol (+)-2c was obtained was ob-
tained as a white solid (24 mg, 0.16 mmol, 30%, 90% ee). [α]2D0
=
+6.4 (c = 1.15, EtOH). The GC–MS and NMR analyses matched
the above given data.
1,1,1-Trifluoro-2-phenylbut-3-en-2-yl Acetate (1e): The reaction was
carried out by following the general procedure (65% conversion)
using 2700 U of esterase BS2 G105A and 10% (v/v) DMSO. Acet-
ate (–)-1e was obtained as a white solid after column chromatog-
raphy using aluminum oxide and n-hexane/ethyl acetate (10:1) as
eluent (36 mg, 0.15 mmol, 36%, 90% ee). [α]2D0 = –38.9 (c = 1.25,
CHCl3). The GC–MS and NMR analyses matched literature
General Procedure for Small-Scale Esterase-Catalyzed Resolutions:
Esterase solution [900 µL, with a crude esterase concentration of
1–3 mgmL–1 dissolved in phosphate buffer (100 m, pH 7.5)] was
Eur. J. Org. Chem. 2010, 2753–2758
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2757

U. Hanefeld, U. T. Bornscheuer et al.
FULL PAPER
data.[20] Alcohol (+)-2c was obtained as a transparent liquid
(24 mg, 12 mmol, 31%, 95% ee). [α]2D0 = +21.8 (c = 1.75, CHCl3).
The GC–MS and NMR analyses matched the above given data.
[12] F. L. Cabirol, P. L. Tan, B. Tay, S. Cheng, U. Hanefeld, R. A.
Sheldon, Adv. Synth. Catal. 2008, 350, 2329–2338.
[13] T. Purkarthofer, W. Skranc, H. Weber, H. Griengl, M. Wub-
bolts, G. Scholz, P. Pöchlauer, Tetrahedron 2004, 60, 735–739.
[14] M. Wiggers, J. Holt, R. Kourist, S. Bartsch, I. W. C. E. Arends,
A. J. Minnaard, U. T. Bornscheuer, U. Hanefeld, J. Mol. Catal.
Enz. B 2009, 60, 82–86.
[15] J. Holt, I. W. C. E. Arends, A. J. Minnaard, U. Hanefeld, Adv.
Synth. Catal. 2007, 349, 1341–1344.
[16] H. Ohta, Y. Kimura, Y. Sugano, Tetrahedron Lett. 1988, 29,
6957–6960.
(S)-1-Cyano-2,2,2-trifluoro-1-phenylethyl Acetate (1i): The reaction
was carried out by following the general procedure (52% conver-
sion) using 20550 U of esterase Est8 and 10% (v/v) DMF. Acetate
(S)-1i was obtained as a colorless liquid at room temperature after
column chromatography using silica gel and petroleum ether/ethyl
acetate (10:1) as eluent (38 mg, 0.16 mmol, 38%, 99% ee). The
GC–MS and NMR analyses matched literature data.[15]
[17] H. Ohta, Y. Kimura, Y. Sugano, T. Sugai, Tetrahedron 1989,
45, 5469–5476.
Supporting Information (see also the footnote on the first page of
this article): Chemical syntheses and spectroscopic data of alcohols
and acetates are provided together with conditions for GC analysis.
[18] M. Paravidino, J. Holt, D. Romano, N. Singh, I. W. C. E. Ar-
ends, A. J. Minnaard, R. V. A. Orru, F. Molinari, U. Hanefeld,
J. Mol. Catal. Enz. B 2010, 63, 87–92.
[19] E. Henke, U. Bornscheuer, R. Schmid, J. Pleiss, ChemBioChem
2003, 4, 485–493.
[20] R. Kourist, S. Bartsch, U. Bornscheuer, Adv. Synth. Catal.
2007, 349, 1393–1398.
Acknowledgments
R. K. and G. S. N. thank the European Social Funds and the Ven-
tureCup M-V (UG09005) for financial support. G. S. N. also
thanks the Deutscher Akademischer Austauschdienst (DAAD)
(Grant A/07/95194), the Vietnamese Ministry of Education and
Training (Grant 3413/QDBGDDT-VP), and Integrated Design of
Catalytic Nanomaterials for a Sustainable Production (IDECAT)
for financial support. M. P. gratefully acknowledges financial sup-
port from the IBOS program of NWO/ACTS. The authors grate-
fully thank Dr. Radka Snajdrova, Dr. Michael Lalk, and Hien
Nguyen (all from Greifswald University) for interesting scientific
discussions. Prof. H. Atomi (Kyoto University, Kyoto, Japan) is
gratefully acknowledged for providing the gene encoding PEstE.
We are also grateful to BRAIN AG (Zwingenberg, Germany) for
provision of Est8 and Est56.
[21] R. Kourist, S. Krishna, J. Patel, F. Bartnek, T. Hitchman, D.
Weiner, U. Bornscheuer, Org. Biomol. Chem. 2007, 5, 3310–
3313.
[22] E. Henke, E. Pleiss, U. Bornscheuer, Angew. Chem. Int. Ed.
2002, 41, 3211–3213.
[23] B. Heinze, R. Kourist, L. Fransson, K. Hult, U. Bornscheuer,
Protein Eng. Des. Sel. 2007, 20, 125–131.
[24] S. Bartsch, R. Kourist, U. Bornscheuer, Angew. Chem. Int. Ed.
2008, 47, 1508–1511.
[25] R. Kourist, G. Nguyen, D. Strübing, D. Böttcher, K. Liebeton,
C. Naumer, J. Eck, U. Bornscheuer, Tetrahedron: Asymmetry
2008, 19, 1839–1843.
[26] Y. Hotta, S. Ezaki, H. Atomi, T. Imanaka, Appl. Environ.
Microbiol. 2002, 68, 3925–3931.
[27] A. Hummel, E. Brusehaber, D. Bottcher, H. Trauthwein, K.
Doderer, U. Bornscheuer, Angew. Chem. Int. Ed. 2007, 46,
8492–8494.
[1] G. Yao, S. Haque, L. Sha, G. Kumaravel, J. Wang, T. Engber,
E. Whalley, P. Conlon, H. Chang, W. Kiesman, R. Petter, Bio-
org. Med. Chem. Lett. 2005, 15, 511–515.
[28] D. O’Hagan, N. Zaidi, J. Chem. Soc. Perkin Trans. 1 1992, 947–
949.
[29] D. Rotticci, C. Orrenius, K. Hult, T. Norin, Tetrahedron:
Asymmetry 1997, 8, 359–362.
[30] K. Watanabe, S. Ueji, J. Chem. Soc. Perkin Trans. 1 2001, 1386–
1390.
[2] D. Friel, M. Snapper, A. Hoveyda, J. Am. Chem. Soc. 2008,
130, 9942–9951.
[3] H. Schenck, P. Lenkowski, I. Choudhury-Mukherjee, S. Ko, J.
Stables, M. Patel, M. Brown, Bioorg. Med. Chem. 2004, 12,
979–993.
[4] J. Ekegren, J. Gising, H. Wallberg, M. Larhed, B. Samuelsson,
A. Hallberg, Org. Biomol. Chem. 2006, 4, 3040–3043.
[5] P. Cozzi, R. Hilgraf, N. Zimmermann, Eur. J. Org. Chem. 2007,
5969–5994.
[31] D. Zhu, B. Rios, J. Rozzell, L. Hua, Tetrahedron: Asymmetry
2005, 16, 1541–1546.
[32] C. Savile, R. Kazlauskas, J. Am. Chem. Soc. 2005, 127, 12228–
12229.
[33] K. Faber, Biotransformations in Organic Chemistry, Springer,
Berlin, 2004
[34] S. Tawaki, A. Klibanov, J. Am. Chem. Soc. 1992, 114, 1882–
1884.
[6] M. North, D. Usanov, C. Young, Chem. Rev. 2008, 108, 5146–
5226.
[7] C. Moberg, E. Wingstrand, Synlett 2010, 3, 355–367.
[8] R. Kourist, P. de Maria, U. Bornscheuer, ChemBioChem 2008,
9, 491–498.
[35] N. Krebsfänger, F. Zocher, J. Altenbuchner, U. Bornscheuer,
Enzyme. Microb. Technol. 1998, 22, 641–646.
[36] C. Chen, Y. Fujimoto, G. Girdaukas, C. Sih, J. Am. Chem. Soc.
1982, 104, 7294–7299.
[37] J. Rehdorf, G. Behrens, G. S. Nguyen, R. Kourist, U. T.
Bornscheuer, submitted.
[9] J. Brunel, I. Holmes, Angew. Chem. Int. Ed. 2004, 43, 2752–
2778.
[10] T. Purkarthofer, W. Skranc, C. Schuster, H. Griengl, Appl.
Microbiol. Biotechnol. 2007, 76, 309–320.
[11] J. Holt, U. Hanefeld, Curr. Org. Synth. 2009, 6, 15–37.
Received: February 12, 2010
Published Online: April 6, 2010
2758
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 2753–2758