Is the ring conformation the most critical parameter in lipase-catalysed acylation of cycloalkanols?

Article abstract of DOI:10.1039/b408861a

CAL-B catalysed the resolution of several five and six-membered cyclic β-hydroxy esters efficiently with the exception of the cis-cyclohexanol (±)-4. When employing molecular modelling techniques the conformation turned out to be the most important determ

Full text of DOI:10.1039/b408861a

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Is the ring conformation the most critical parameter in lipase-
catalysed acylation of cycloalkanols?†
Laura M. Levy, Iván Lavandera and Vicente Gotor*
Departamento de Química Orgánica e Inorgánica, Facultad de Química, Universidad de Oviedo,
33006 Oviedo, Spain. E-mail: vgs@fq.uniovi.es; Fax: +34-985103448; Tel: +34-985103448
Received 11th June 2004, Accepted 27th July 2004
First published as an Advance Article on the web 23rd August 2004
CAL-B catalysed the resolution of several five and six-membered cyclic b-hydroxy esters efficiently with the
exception of the cis-cyclohexanol (±)-4. When employing molecular modelling techniques the conformation turned
out to be the most important determinant for their reactivity towards O-acetylation. In all cases, the R enantiomers
reacted faster than the S enantiomers since the reactive intermediates of the former can adopt more favourable ring
conformations and thus experience less steric hindrance in the active site. Furthermore, the minimised structure for
the main conformer of R-4 showed that the axial hydrogens in the 3 and 5-positions with respect to the hydroxyl
group prevent the enzymatic reaction.
Computer-aided molecular modelling is a potent tool to
1. Introduction
explain many experimental results in biotransformations.13
The force field methods relate the geometry with the potential
energy of a molecule using an analytical function. Amongst
them, AMBER force field is one of the most commonly used
for protein structures.14 Conformations can be either randomly
or manually generated by rotation of the dihedral angles within
the molecule. Crystal structures of the enzyme can be studied
with its ligand bound, and both the electrostatic and steric
interactions are considered. Thus, modelling of these transition
states helps to understand the mechanism,15 and the selectivity16
of the enzymes.
The kinetic resolution of racemic mixtures with lipases is an
attractive means for the preparation of enantiomerically pure
molecules1 and is used as a major method for their industrial
scale production.2
Several 5 and 6-membered cyclic alcohols have been resolved
by lipase-catalysed acylation reactions.3,4 Amongst them,
different 1,2-disubstituted cycloalkanols, which are important
chiral building blocks for the asymmetric synthesis of bio-
logically active compounds, have been obtained enantio-
merically pure using this tool.5,6 These resolutions show some
common trends, such as, first, the R enantiomer is always
preferred by the lipase, secondly, cyclopentanols react more
rapidly than cyclohexanols, and thirdly, the cis isomers react
more slowly than their trans analogues. During the development
of the resolution of several five, six and seven-membered cyclic
cis and trans-b-hydroxy esters (Fig. 1) with lipases A and B
from Candida antarctica (CAL-A and CAL-B), we noticed that
CAL-B reacted very efficiently with all R enantiomers except
in the case of the cis-cyclohexanol.7 There are other examples
known in the literature in which the cis-2-substituted cyclo-
hexanols do not react with lipases, even when the corresponding
cis-2-substituted cyclopentanols do.3b,8 However, in some cases
it has been shown that the cis-2-substituted cycloheptanol
derivatives react under the same conditions.7,9
In this work, we have modelled the intermediates in the
CAL-B-catalysed acetylation of five and six-membered cyclic cis
and trans-b-hydroxy esters (Fig. 1, n = 1, 2), in order to explain
the excellent selectivity seen with the R-enantiomers. Together
with NMR spectroscopy, the lack of reactivity of the cis-cyclo-
hexanol derivative compared to the five-membered analogue
was also studied.
2. Results and discussion
2.1. Resolution of cyclic b-hydroxy esters by lipase-catalysed
acetylation
As mentioned in the introduction, we have developed the
resolution of several five, six, and seven-membered cyclic
b-hydroxy esters via CAL-B-catalysed acetylation with vinyl
acetate in tert-butyl methyl ether at 30 °C (Scheme 1).7 Under
these reaction conditions (Table 1), the acylation of the
trans isomers (±)-1 and (±)-2 took place smoothly, yielding
both the S-substrates and the R products enantiomerically
pure. When the same methodology was applied to the cis-
configured b-hydroxy esters, the R enantiomer of the five-
membered derivative (±)-3 was rapidly acetylated. However,
CAL-B showed very low activity towards cis-cyclohexanol
(±)-4, obtaining only 5% conversion after 2 days, despite
exhibiting again complete enantiodiscrimination (E > 200).
Fig. 1
Conformational effects could account for this tendency, since
in general hydroxyl groups react more slowly in axial than in
equatorial position in lipase-catalysed resolutions.3b,10 The same
is true for non-biocatalysed transformations. For example, Eliel
and Biros showed that the acylation of 2-alkylcyclohexanols
occurs faster with the trans isomers (OH equatorial) than with
the cis isomers (OH axial) due to steric and polar effects.11
Pasto and Rao also recorded a similar trend in 2,5-di-tert-
butylcyclohexanol derivatives.12
Scheme 1
2.2. Molecular modelling
† Electronic supplementary information (ESI) available: complete 1H-
NMR spectral data are shown. Furthermore, some extensive modelling
figures are included. See http://www.rsc.org/suppdata/ob/b4/b408861a/
Starting with the X-ray crystal structure of CAL-B (1LBS),18 we
modelled the analogues of the second tetrahedral intermediate
2 5 7 2
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
Table 1 Resolution of cyclic b-hydroxy esters with CAL-Ba
nb
Substrate
t/h
cc
Ec
1 (trans)
2 (trans)
1 (cis)
(±)-1
(±)-2
(±)-3
(±)-4
1.3
1.7
2.7
50
50
50
5
>200
>200
>200
>200
2 (cis)
48
a See Ref. 7. b See Scheme 1. c Calculated from the e.e. of the substrate
and the product. See Ref. 17.
in the acetylation of the cycloalkanols, since this transition state
determines the rate and selectivity in a kinetic resolution of
racemic alcohols (Fig. 2).
Fig. 2 Phosphonate intermediates used for molecular modelling
in the acetylation of cyclopentanols (n = 1) or cyclohexanols (n = 2)
catalysed by CAL-B. Key hydrogen bonds between the lipase and the
intermediate are displayed as dashed lines (a–f).
In order to simplify the modelling process using molecular
mechanics force fields, we chose the phosphonate analogues of
these tetrahedral intermediates for minimisation calculations.18
Geometry optimisation of the phosphonate containing a
methoxy group attached to the phosphorus atom afforded a
structure in which all key hydrogen bonds required for catalysis
were present (Fig. 2).19 Next, the whole molecule was added and
minimised. Different conformations of these structures were
systematically searched for and focused on those most likely
to mimic catalytically productive transition states (see under
Computational details). The following two criteria had to be
fulfilled: the presence of all key hydrogen bonds within the active
site and the lack of steric impediments between the phosphonate
and the lipase.
Fig. 3 Phosphonate models for the O-acetylation of cyclohexanol 2:
A) R enantiomer; and B) S enantiomer. Red represents oxygen atoms,
blue represents nitrogen atoms, and yellow represents phosphorus
atoms. Above images display some amino acids in a wireframe
representation to allow a better view of the structures.
The active site of CAL-B contains a large hydrophobic
pocket (or acyl pocket) above the Asp-His-Ser catalytic triad
and a medium-sized pocket (or nucleophile pocket) below it.18
The acyl moiety of the intermediate lies in the large subsite
(the methyl of the acetate in our case), while the carbocycle
of the cycloalkanol binds to the medium-sized pocket. In
general, the active site below the catalytic triad restricts the
binding of the alcohol substrate to such an extent that only a
few conformations were obtained with all the catalytic hydrogen
bonds present.19 Conformation and ring size proved to be most
crucial when determining reactive intermediates.
moiety lies in the large pocket, with the oxygen of the carbonyl
group pointing out towards the solvent.
In the case of the phosphonate analogue of S-2, the ring also
binds to the medium-sized pocket (Fig. 3B). This time however,
H-2 and eq-H-3, which are trans to the phosphonate group, are
very close to the catalytic histidine despite rotation of both the
cyclohexane moiety and the histidine group. As a consequence,
three key hydrogen bonds are lost (Row 2 in Table 2).
The different orientations of the substrates in the active
site, and the steric hindrance between the cyclohexane and
His224 residue, appeared to be key determinants of the enzyme
selectivity. This was also confirmed by the energies obtained
for both structures, since the R intermediate was 4.2 kcal mol−1
more stable than the S-enantiomer.21
2.2.1. Enzymatic acylation of trans-cyclohexanol (±)-2.
First, we studied the CAL-B-catalysed acetylation of (±)-2.
Cyclohexane derivatives are present in a two chair-equilibrium
in solution, which lies on the side of the thermodynamically
more stable one. trans-1,2-Disubstituted cyclohexanes virtually
exist only in the conformation that places both substituents in
equatorial positions.20 Thus, only those structures with both the
alcohol and the ester group placed in the less hindered positions
were taken into account during the molecular modelling.
The best structure of the tetrahedral intermediate for the
O-acetylation of R-2 is shown in Fig. 3A. The cyclohexane ring
fits perfectly into the medium-sized pocket with all key hydrogen
bonds present (Row 1 in Table 2) and with no steric hindrance.
Both substituents occupy equatorial positions and the ester
2.2.2. Enzymatic acylation of trans-cyclopentanol (±)-1.
Next, we modelled the enzymatic acylation of trans-cyclo-
pentanol (±)-1. Conformations of cyclopentanes have been
much less studied. Nonetheless, it has been established that
there are two symmetrical puckered conformations, the
envelope (Cs) and the half-chair (C2), with the energy barrier
very low between both (Fig. 4). The most stable conformer of
trans-cyclopentane-1,2-diol is shown in Fig. 4A (X = OH).22
Normally, five-membered ring derivatives react enzymatically
faster than cyclohexanes.3 Accelerative factors could be, first,
the smaller size of the cycle, which would produce less steric
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7
2 5 7 3

View Article Online
Table 2 Key hydrogen bonds and energy in CAL-B-catalysed acetylation of cyclic trans- and cis-b-hydroxy esters of five and six membersa
H-bond distance/Å (angle)c
Row
Compound
Fig.b
a
b
c
d
e
f
E/kcal mol−1
1
2
3
4
5
6
7
8
9
(2R)-2
3A
3B
7A
7B
S4
S1
S2
5
3.02 (134°)
3.25 (124°)
3.06 (132°)
3.07 (150°)
3.48 (126°)
3.03 (135°)
3.23 (127°)
3.01 (134°)
3.00 (133°)
3.14 (131°)
3.31 (134°)
3.01 (146°)
2.96 (128°)
2.95 (127°)
3.16 (129°)
3.32 (139°)
2.91 (131°)
3.15 (132°)
2.91 (147°)
2.91 (133°)
3.40 (139°)
2.95 (139°)
3.27 (113°)
2.83 (144°)
2.85 (139°)
3.06 (149°)
2.93 (155°)
3.03 (160°)
3.39 (162°)
2.97 (161°)
3.16 (155°)
3.18 (160°)
3.01 (157°)
3.55 (165°)
2.85 (154°)
3.09 (161°)
2.75 (168°)
2.76 (168°)
2.78 (172°)
2.80 (165°)
2.72 (167°)
2.77 (172°)
2.76 (161°)
2.77 (174°)
2.78 (168°)
2.82 (170°)
2.85 (168°)
2.83 (166°)
2.78 (166°)
3.03 (165°)
2.80 (172°)
2.79 (159°)
2.85 (175°)
3.16 (128°)
0
(2S)-2
+4.2d
+6.4d
+9.3d
+5.9d
0
(2R)-4 (OH ax)
(2R)-4 (OH eq)
(2S)-4
(2R)-1
(2S)-1
+4.4e
+2.4e
+4.1e
(2R)-3
(2S)-3
S3
c
a Lost hydrogen bonds appear in bold. b See text. Fig. 2 defines the hydrogen bonds. d Compared with the energy of Fig. 3A. e Compared with the
energy of Fig. S1 (ESI).
Fig. 4 Main conformation for: A) trans-1,2-disubstituted cyclo-
pentanols; and B) cis-1,2-disubstituted cyclopentanols.
hindrance in the active site, and secondly, the greater ease of
cyclopentanes to convert from one conformation into another,
by this increasing the possibility of finding an acceptable
disposition in the active site.
On the other hand, different conformations of cyclohexanes
are more difficult to interconvert. Thus, if the lowest energy
chair conformer does not fit into the active site in a successful
orientation, it is less probable that the reaction takes place.
When computer modelling was used to minimise the
intermediates of the trans-cyclopentane derivatives (±)-1, similar
results to the six-membered analogues (±)-2 were obtained.
Fig. 5 Model for the O-acetylation of the R enantiomer of cyclo-
For R-1, the cycle fits into the medium subsite in a half-chair
pentanol 3. The above image displays some amino acids in a wireframe
representation to allow a better view of the structure.
conformation without any steric impediments (Fig. 4A,
X = CO2Et), with all key hydrogen bonds present (Row 6 in
Table 2). In contrast, for S-1, the ring lies in the medium-sized
pocket in a less stable envelope conformer. Furthermore, H-2
and H-3c are too close to His224, destabilising the intermediate
by 4.4 kcal mol−1 with respect to the R enantiomer (Row 7
in Table 2). Models are shown in electronic supplementary
information (ESI)† as Fig. S1 and S2.
amino acid and the oxyanion takes place (Row 9 in Table 2).
This is also confirmed by the energies obtained for both struc-
tures, since the R intermediate is 1.7 kcal mol−1 more stable than
the S analogue.
2.2.4. Enzymatic acylation of cis-cyclohexanol (±)-4. Next,
we studied cis-cyclohexanol (±)-4. This substrate proved to be
poor for CAL-B, observing almost no product formation even
after 2 days. In this case, the ring conformation could be unclear
due to the existence of two different chairs which would present
one group in an axial position, and the other in an equatorial
position.
2.2.3. Enzymatic acylation of cis-cyclopentanol (±)-3. The
most stable conformation of cis-cyclopentane-1,2-diol is a
slightly distorted envelope, shown in Fig. 4B (X = OH).22
The phosphonate intermediate shown in Fig. 5 was chosen
as the best model for the acetylation of R-3. As with the trans
isomer, the ring binds perfectly to the medium-sized pocket,
with the hydroxyl group in an axial position at the fold of the
envelope and the ester located in a less hindered equatorial
position at the flap of the envelope (Fig. 4B, X = CO2Et). This
group lies in the large hydrophobic subsite with the oxygen of
the carbonyl pointing out towards the solvent. All key hydrogen
bonds are present in this structure (Row 8 in Table 2). There
are no significant steric interactions in the active site, only H-4c
and H-5c being slightly closer to the catalytic histidine (Fig. 8A),
but not enough to disrupt the interaction between this amino
acid and the nucleophile. This may be the cause for the slower
reaction with respect to the trans analogue, since the energy of
the cis intermediate is 2.4 kcal mol−1 higher.
1H-NMR spectroscopy was used to discern this equilibrium
(Fig. 6). Several spectra were recorded in three solvents, i.e.
CDCl3, CD3CN, and acetone-d6, the latter two were used as
reaction solvents in the enzymatic resolution of (±)-4.7 Spectra
at 30 °C and at low temperature were also obtained to study
the main conformation. The peak of H-1 (H a to the ester) was
used in the analysis since it is well resolved as a ddd without any
overlapping at 2.4–2.5 ppm. In all the spectra, no signal splitting
was observed at low temperature, instead a similar pattern was
3
always seen: one coupling constant of 11–12 Hz (typical JHH
3
3
ax–ax), and two of 3–4 Hz (typical JHH ax–eq or JHH eq–eq).
This unequivocally confirms H-1 in an axial position with a large
3
ax–ax coupling constant with trans-H-6, and two small JHH
When the phosphonate analogue of S-3 was modelled, a
structure in which both groups occupied equatorial positions
at the fold of the envelope was obtained, with a slightly higher
energy (Fig. S3 in ESI†). Furthermore, the ester moiety lies
too close to the His224 residue, moving away the cycle towards
Thr40. As a result of this, a weaker interaction between this
ax–eq with H-2 and cis-H-6. Thus, in the major conformation
of (±)-4 the hydroxyl group lies in an axial position. Hence, this
more stable conformer of R-4 was investigated first (Fig. 7A).
The cyclohexane lies in the medium-sized pocket of the active
site and the ester group in the large pocket with the oxygen of
the carbonyl pointing out towards the solvent. In this structure,
2 5 7 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7

View Article Online
Fig. 6 1H-NMR spectra of compound (±)-4 (below) and expanded signal of H-1 (above) in: A) CDCl3; B) CD3CN; and C) acetone-d6.
Fig. 7 Models for the O-acetylation of the R enantiomer of cyclohexanol 4 in two different conformations: A) OH in axial position; B) OH in
equatorial position. Above images display some amino acids in a wireframe representation to allow a better view of the structure.
cis-hydrogen atoms in 3,5-diaxial positions with respect to the
hydroxyl group clash with the catalytic histidine (Fig. 8B). These
steric hindrances in the active site were enough to disrupt the
necessary interaction with the nucleophile, and therefore, the
energy of the intermediate is destabilised by 6.4 kcal.mol−1
in comparison with the trans isomer R-2 (Row 3 in Table 2).
These observations can explain the lower reactivity of the
cis-R-isomer.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7
2 5 7 5

View Article Online
4. Experimental section
4.1. General
Compounds (±)-2, (±)-3 and (±)-4 are commercially available
from Aldrich and were used without further purification.
Compound (±)-1 was synthesised using a procedure similar to
that reported by Nakata and Oishi.7,26 1H-NMR spectra were
obtained with a Bruker AMX 400 (400.13 MHz) spectrometer,
using CDCl3, CD3CN and acetone-d6 as solvents.
4.2. Computational details
The program Insight II, version 2000.1, was used for viewing the
structures. The geometric optimisations were performed using
Discover, version 2.9.7 (Accelrys, San Diego, CA, USA), using
the AMBER force field.14 The distance dependent dielectric
constant was set to 4.0 and the 1–4 van der Waals interactions
were scaled to 50%. The crystal structure of CAL-B (1LBS)18 was
obtained from the Protein Data Bank (www.rcsb.org/pdb/).
Hydrogen atoms were added to the structure, tested for
partial charge balance and corrected for atom types within
the AMBER force field. The pH of the catalytic histidine was
adjusted to pH 4 and the structure was then relaxed. Geometry
optimisations were carried out in two steps. First, the steepest
descent algorithm got the molecule to a local minimum with a
RMS deviation of about 0.02 Å mol−1. Secondly, a conjugate
gradient algorithm gave a more precise minimisation near the
local minimum, obtaining an RMS value of 0.005 Å mol−1.
Next, a phosphonate core of the tetrahedral intermediate
mimicking methylation of methanol was assembled in the
crystal structure. Partial charges on the intermediate atoms
were set to those of the tetrahedral intermediate (carbon, not
phosphorus), obtaining the values by semi-empirical calculation
performed using Chem3D using the AM1 parameters.
Fig. 8 Schematic showing the steric hindrance between His224
and: A) R-3; B) R-4. Hydrogen bond lost is marked with a red cross.
Subsequently, the phosphonate analogue with the hydroxyl
group in an equatorial position was built and minimised
(Fig. 7B). We obtained a structure in which all key hydrogen
bonds are present, but the energy of the system is almost
3 kcal mol−1 higher than its conformer in Fig. 7A (Row 4 in
Table 2). This fact could be due to two main factors: first, a
higher-energy conformation of the cyclohexane, and secondly,
a weaker binding of the substrate within the active site. The
ring does not lie in the medium-sized pocket because of steric
hindrance between His224 and the ester in the axial position,
resulting in the loss of stabilising van der Waals interactions.
Lastly, the intermediate of S-4 was minimised (Fig. S4
in ESI†). Although the best structure had a similar energy
compared with the one obtained for the R enantiomer, some
crucial hydrogen bonds were lost (Row 5 in Table 2), explaining
the even lower reactivity of CAL-B towards this enantiomer.
For the optimisation of this phosphonate core a systematic
approach was applied. Initially, the intermediate was allowed
to adjust to the enzyme active site by keeping the entire enzyme
fixed during the geometry optimisation. The side chains of the
lipase were then released and allowed to adjust. Finally, the
entire complex was allowed to adjust. This approach avoided
drastic changes in the lipase structure caused by non-optimal
conformations of the intermediate until an RMS maximum
value of 0.005 Å mol−1 was obtained. A hydrogen bond
calculation was performed after each minimisation in order to
assure the presence of the critical hydrogen bonds around the
catalytic site.19
The rest of the molecule was added and different
conformations were produced by manual adjustment of
dihedral angles specified in Fig. 9. Geometry optimisations
were again performed using the same approach described
above. Due to the flexible structure of these intermediates,
small adjustments of the dihedral angle a caused large
changes in the orientation of the ester moiety. Adjustments of
approximately 10–20° were performed for both dihedral angles
adjacent to that position. Structures causing obvious steric
hindrance were ignored.
3. Conclusions
Although the results of biotransformations are more predict-
able today than ten years ago, still some trends remain which
cannot be explained satisfactorily; e.g. the lack of reactivity of
cis-1,2-disubstituted cyclohexanols in lipase-catalysed acyla-
tion processes. Using computer-aided molecular modelling, a
molecular basis for the acylation of different cyclic b-hydroxy
esters has been proposed in this paper.
The R-enantiopreference for these compounds shown
by CAL-B is in good agreement with Kazlauskas’ rule.23
The preferred enantiomers adopt a more favourable ring
conformation, in which their binding into the active site is
improved. In contrast, S enantiomers had a worse interaction
with the catalytic triad, and crucial hydrogen bonds were lost
in some cases.
Furthermore, we have explained the low reactivity of the
cis-cyclohexanol derivative. In its major conformation with the
hydroxyl group in an axial position, the cis-hydrogens in 3,5-
diaxial positions impeded the reaction due to steric hindrance.
However, the five-membered analogue can react due to the
smaller size and the better ring conformation.
The results shown can also explain related experimental facts,
such as the lack of reactivity of several trans-1,3-disubstituted
cyclohexanes.24 In this class, the group in the 3 position would
be preferentially placed in an equatorial disposition while
the hydroxyl group would be in an axial position. The model
proposed by us is in accordance with the good reactivity of
cis-2-cyanocyclohexanol with lipases,3a since small groups
in the 2 position would favour a ring conformation with an
equatorial hydroxyl group. This model can explain the reactivity
of cis-1,2-cyclopentanols and of cycloheptanols. The seven-
membered analogues could adopt several energetically similar
conformers,25 thus allowing for many different dispositions
within the active site of the enzyme, and by this being able to
react more easily, despite their bigger size.
Fig. 9 Manipulated dihedral angles for the intermediates.
Acknowledgements
Financial support of this work by the Spanish Ministerio
de Ciencia y Tecnología (Project No. PPQ-2001-2683) and
by Principado de Asturias (Project No. GE-EXP01-03)
is gratefully acknowledged. I.L. thanks MCYT for a pre-
doctoral fellowship.
2 5 7 6
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7

View Article Online
14 (a) S. J. Weiner, P. A. Kollman, D. A. Case, U. C. Singh, C. Ghio,
G. Alagona, S. Profeta and P. Weiner, J. Am. Chem. Soc., 1984, 106,
765; (b) W. D. Cornell, P. Cieplak, C. I. Bayly and P. A. Kollmann,
J. Am. Chem. Soc., 1993, 115, 9620.
15 S. Park, E. Forró, H. Grewal, F. Fülöp and R. J. Kazlauskas, Adv.
Synth. Catal., 2003, 345, 986.
16 (a) F. Hæffner and T. Norin, Chem. Pharm. Bull., 1999, 47, 591;
(b) K. Hult and P. Berglund, Curr. Opin. Biotechnol., 2003, 14, 395.
17 C. S. Chen, Y. Fujimoto, G. Girdaukas and C. J. Sih, J. Am. Chem.
Soc., 1982, 104, 7294.
18 J. Uppenberg, N. Öhrner, M. Norin, K. Hult, G. J. Kleywegt,
S. Patkar, V. Waagen, T. Anthonsen and T. A. Jones, Biochemistry,
1995, 34, 16838.
References
1 (a) C. J. Sih and S.-H. Wu, Top. Stereochem., 1989, 19, 63;
(b) U. T. Bornscheuer and R. J. Kazlauskas, Hydrolases in Organic
Synthesis: Regio- and Stereoselective Biotransformations, Wiley-
VCH, Weinheim, 1999.
2 A. Liese, K. Seelbach and C. Wandrey, Industrial Biotransforma-
tions, Wiley-VCH, Weinheim, 2000.
3 (a) E. Forró, K. Lundell, F. Fülöp and L. T. Kanerva, Tetrahedron:
Asymmetry, 1997, 8, 3095; (b) A. Maestro, C. Astorga and V. Gotor,
Tetrahedron: Asymmetry, 1997, 8, 3153; (c) A. Luna, C. Astorga,
F. Fülöp and V. Gotor, Tetrahedron: Asymmetry, 1998, 9, 4483;
(d) E. Forrö, Z. Szakonyi and F. Fülöp, Tetrahedron: Asymmetry,
1999, 10, 4619; (e) L. M. Levy and V. Gotor, J. Org. Chem., 2004,
69, 2601.
4 J. González-Sabín, V. Gotor and F. Rebolledo, Tetrahedron:
Asymmetry, 2004, 15, 1335.
5 J. K. Whitesell, Chem. Rev., 1992, 92, 953.
6 (a) D. H. Apella, L. A. Christianson, D. A. Klein, D. R. Powell,
X. Huang, J. Barchi, Jr. and S. H. Gellman, Nature (London),
1997, 387, 381; (b) A. Pecunioso, M. Maffeis, C. Marxhioro,
L. Rossi and B. Tamburini, Tetrahedron: Asymmetry, 1997, 8, 775;
(c) E. Racanska and F. Gregan, Pharmazie, 1999, 54, 68.
7 L. M. Levy, J. R. Dehli and V. Gotor, Tetrahedron: Asymmetry,
2003, 14, 2053.
19 To identify hydrogen bonds, a donor atom to acceptor atom
distance of less than 3.20 Å and a donor atom–hydrogen–acceptor
atom angle of 120° or greater are required.
20 Several 1H-NMR spectral data of (±)-2 were recorded in
different solvents, and signals of H-1 (a to the ester) and H-2 (a to
the hydroxyl group), showed two 3JHH ax–ax (between 10–14 Hz) and
one 3JHH between 3–5 Hz, typically ax–eq. This indicates that both
hydrogens must be in axial positions.
21 The relation between the difference in the Gibbs energy of the
transition states of the enantiomers and the enantiomeric ratio can
‡
be shown with the transition state theory to be DR − SDG‡ = DGR
−
DGS‡ = −RTlnE.
8 (a) E. Forró, L. T. Kanerva and F. Fülöp, Tetrahedron: Asymmetry,
1998, 9, 513–520; (b) C. Brunet, M. Zarevucka, Z. Wimmer and
M.-D. Legoy, Enzyme Microb. Technol., 2002, 31, 609.
9 E. Forró and F. Fülöp, Tetrahedron: Asymmetry, 1999, 10, 1985.
10 R. Tanikaga, Y. Matsumoto, M. Sakaguchi, Y. Koyama and K. Ono,
Tetrahedron Lett., 2003, 44, 6781.
11 E. L. Eliel and F. J. Biros, J. Am. Chem. Soc., 1966, 88, 3334.
12 D. J. Pasto and D. R. Rao, J. Am. Chem. Soc., 1970, 92, 5151.
13 (a) R. J. Kazlauskas, Curr. Opin. Chem. Biol., 2000, 4, 81;
(b) R. Kazlauskas, Science, 2001, 293, 2277.
22 R. J. Abraham, R. Koniotou and F. Sancassan, J. Chem. Soc., Perkin.
Trans. 2, 2002, 2025.
23 R. J. Kazlauskas, A. N. E. Weissfloch, A. T. Rappaport and L. A.
Cuccia, J. Org. Chem., 1991, 56, 2656.
24 (a) R. Tanikaga and A. Morita, Tetrahedron Lett., 1998, 39,
635; (b) L. M. Levy, G. de Gonzalo and V. Gotor, Tetrahedron:
Asymmetry, 2004, 15, 2051.
25 (a) D. F. Bocian and H. L. Strauss, J. Am. Chem. Soc., 1977, 99,
2876; (b) K. B. Wiberg, J. Org. Chem., 2003, 68, 9322.
26 T. Nakata and T. Oishi, Tetrahedron Lett., 1980, 21, 1641.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 7 2 – 2 5 7 7
2 5 7 7