The importance of ester and alkoxy type functionalities for the chemo- And enantio-recognition of substrates by hydrolysis with Candida rugosa lipase

Article abstract of DOI:10.1039/b005512n

Racemic esters of 1-phenyl- and 1-(pyridyl)ethyl acetates 1a-c (R = Me, Ph) were subjected to hydrolysis in water in the presence of Candida rugosa lipase (CRL). 1-Pyridylethyl acetates (1, R = Me) are hydrolyzed by crude CRL (C-CRL) and isopropanol (isopropyl alcohol) treated CRL (PT-CRL) at very low rates, and the enantiorecognition was disappointing. By using 1-pyridylethyl benzoates (1, R = Ph) the results differed greatly: hydrolysis occurred much faster, and the enantiorecognition was good for 3- and 4-pyridyl derivatives and excellent for 2-pyridyl derivative. Analogous results were obtained by submitting the 1-phenylethanol esters 4 to enzymatic hydrolysis under the same experimental conditions. The hydrolysis of methyl o-acetoxybenzoates 7 (R = Me) gave quantitatively the deacetylated methyl o-hydroxybenzoates 9 (R = Me) by using either C-CRL or PT-CRL. A complete reversed selectivity is observed in the hydrolysis of phenyl o-acetoxybenzoates 7 (R = Ph) catalyzed by PT-CRL. Molecular modeling studies, aimed at probing the substrate specificity and the enantioselectivity of enzyme in terms of its three-dimensional structure is reported. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/b005512n

View Article Online / Journal Homepage / Table of Contents for this issue
The importance of ester and alkoxy type functionalities for the
chemo- and enantio-recognition of substrates by hydrolysis with
Candida rugosa lipase
1
Francesca Bellezza, Antonio Cipiciani,* Gabriele Cruciani and Francesco Fringuelli
Dipartimento di Chimica, Università di Perugia, Via Elce di Sotto 8, 06100 Perugia, Italy
Received (in Cambridge, UK) 10th July 2000, Accepted 30th October 2000
First published as an Advance Article on the web 30th November 2000
Racemic esters of 1-phenyl- and 1-(pyridyl)ethyl acetates 1a–c (R = Me, Ph) were subjected to hydrolysis in water in
the presence of Candida rugosa lipase (CRL). 1-Pyridylethyl acetates (1, R = Me) are hydrolyzed by crude CRL (C-
CRL) and isopropanol (isopropyl alcohol) treated CRL (PT-CRL) at very low rates, and the enantiorecognition was
disappointing. By using 1-pyridylethyl benzoates (1, R = Ph) the results differed greatly: hydrolysis occurred much
faster, and the enantiorecognition was good for 3- and 4-pyridyl derivatives and excellent for 2-pyridyl derivative.
Analogous results were obtained by submitting the 1-phenylethanol esters 4 to enzymatic hydrolysis under the
same experimental conditions. The hydrolysis of methyl o-acetoxybenzoates 7 (R = Me) gave quantitatively the
deacetylated methyl o-hydroxybenzoates 9 (R = Me) by using either C-CRL or PT-CRL. A complete reversed
selectivity is observed in the hydrolysis of phenyl o-acetoxybenzoates 7 (R = Ph) catalyzed by PT-CRL. Molecular
modeling studies, aimed at probing the substrate specificity and the enantioselectivity of enzyme in terms of its
three-dimensional structure is reported.
Although different strategies have been adopted to obtain
chiral pyridyl alcohols with satisfactory enantioselectivities
Introduction
An enzyme is generally useful if it is stable, inexpensive and
and yields it is still necessary to find new, general and
accepts a broad range of substrates while retaining high
reliable methods. Optically active 1-(pyridyl)ethanols have
selectivity for each one.
Lipase from Candida rugosa (CRL) has these qualities and
it is extensively used for the hydrolysis and esterification of
been prepared by lipase-catalyzed asymmetric acetylation of
individual alcohols and currently the reaction has been limited
to some 1-(2-pyridyl)ethanols.18 Among the lipases tested,
organic compounds1 and the knowledge2 of its crystal struc-
C-CRL is reported18b to be undesirable for the enantio-
ture has greatly contributed to understanding the mechanism
selective transesterification of racemic 1-(pyridyl)ethanols with
vinyl acetate.
We have submitted racemic 1-(pyridyl)ethyl acetates (1, R =
of its selective recognition of substrates.3 In order to improve
the chemo-, regio- and diastereoselectivity of the enzyme,
commercial CRL (C-CRL), as well as conventional purific-
Me) to enzymatic hydrolysis with C-CRL and with isopropanol
ations,4 have been subjected to several treatments that cause
treated CRL (PT-CRL)19 at room temperature in aqueous
molecular (covalent modifications) or conformational (non-
covalent modifications) changes. Covalent modifications are
medium at pH 7.2 (Table 1). The hydrolysis was very slow,
the 4-pyridyl-derivatives were not accepted and the enantio-
caused by immobilizing C-CRL on several supports5 or by the
recognition was disappointing.20 By using 1-(pyridyl)ethyl
nitration of tyrosyl residues.6 Non-covalent modifications are
benzoates (1, R = Ph), the results were very greatly different;
brought about by changing the reaction temperature,4 pH7
the hydrolysis occurred much faster, the 4-pyridyl-derivatives
and solvent,8,9 by introducing additives such as CaCl2,10
were accepted and the enantiorecognition was good for 3- and
surfactants,11 microemulsing agents12 and carbohydrates13 and
4-pyridyl-derivatives and excellent for the 2-pyridyl-derivative
by treating with organic solvents.14,15
Surprisingly when the C-CRL does not accept a substrate,
no attempt to modify the compound is generally made but the
C-CRL is rejected and other lipases are tested.
when hydrolyzed in the presence of C-CRL.
Analogous results were obtained by submitting the 1-phenyl-
ethanol esters 4 to enzymatic hydrolysis under the same experi-
mental conditions (Table 2). The C-CRL and PT-CRL do not
show enantiopreference for the acetate 4 (R = Me) while they
preferentially hydrolize the (ϩ)-R benzoate 4 (R = Ph).
o-Acetoxybenzoates 7 are potential precursor of aspirin.
To reduce or block the side effects of aspirin (gastric irritation
Following our studies on the use of CRL in organic syn-
thesis,16 this paper gives three examples that demonstrate
that the optical resolution of alcohols via hydrolysis of esters
and the chemoselective hydrolysis of an alkoxy group in the
presence of an ester functionality, depend greatly on the acid
and bleeding) numerous chemical derivatives of acetylsalicylic
and alcoholic residues of the ester function and on whether the
acid (precursor of aspirin) have been tested22 but it is still an
CRL is crude or treated.
open problem.
Moreover hydrolases play an important role in the meta-
bolism of many xenobiotics, and humans have been shown to
express hydrolases in the liver, plasma, intestine, brain, stomach
Results and discussion
Optically active pyridyl-derivatives have received increasing
attention as chiral auxiliaries and chiral ligands17 but the use of
systems with a chiral center on the pyridine side is actually
limited because the introduction of a chiral center directly
attached to the pyridyl ring is difficult to achieve.
colon, macrophage and monocytes. Although hydrolases
activities vary quite strongly between species, the active site
region of the enzymes is sometimes conserved and the mechan-
ism of hydrolysis shows a similar pattern. Thus the results
of this study can be easily extended to check the role in the
DOI: 10.1039/b005512n
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444
This journal is © The Royal Society of Chemistry 2000
4439

View Article Online
Table 1 C-CRL and PT-CRL catalyzed hydrolysis of 1-(pyridyl)ethanol esters
Entry
Enzyme
Substrate
t/h
Conv. (%)a
2 (ee)b
3 (ee)b
Ec
1
2
3
4
5
6
7
8
9
C-CRL
PT-CRL
C-CRL
PT-CRL
C-CRL
PT-CRL
C-CRL
PT-CRL
C-CRL
PT-CRL
1a R = Me
1a R = Me
1a R = Ph
1a R = Ph
1b R = Me
1b R = Me
1b R = Ph
1b R = Ph
1c R = Ph
1c R = Ph
120
76
7
70
33
48
39
35
41
49
44
43
45
27
24
91
75
49
46
83
83
84
81
64
29
97
68
36
61
80
65
65
75
3
2
100
15
9
2
408
309
22
10
54
29
5
27
21
26
22
10
a Conversion of reaction after consumption of the required amount of NaOH 0.2 M. b Enantiomeric excess (%) determined by measuring the specific
optical rotation. c Enantioselectivity factor.
Table 2 C-CRL and PT-CRL catalyzed hydrolysis of 1-phenylethanol esters
Entry
Enzyme
R
t/h
Conv. (%)a
5 (ee)b
6 (ee)b
Ec
1
2
3
4
C-CRL
PT-CRL
C-CRL
PT-CRL
Me
Me
Ph
31
22
36
21
46
51
45
47
44
29
81
70
39
43
77
58
4
3
20
10
Ph
a For footnotes a,b,c see Table 1.
Table 3 C-CRL and PT-CRL catalyzed hydrolysis of o-acetoxybenzoates
Entry
Enzyme
R
t/h
Conv. (%)a
8b
9b
10b
1
2
3
4
C-CRL
PT-CRL
C-CRL
PT-CRL
Me
Me
Ph
3
3
60
11
100
100
85
—
—
50
100
100
100
—
—
—
50
—
Ph
100
—
a See Table 1. b Determined by GC.
pharmacokinetic behaviour of most therapeutic agents con-
taining acetoxybenzoate-like compounds.
Using C-CRL or PT-CRL, the hydrolysis of methyl o-
acetoxybenzoates 7 (R = Me) at room temperature in aqueous
medium at pH 7.2 gave quantitatively the deacetylated methyl
o-acetoxybenzoates 7 (R = Ph) catalyzed by PT-CRL. Only
the hydrolysis of carboxylate function was observed and the
o-acetoxybenzoic acid 8 was obtained with high yield and
high chemoselectivity. Phenyl o-acetoxybenzoate 7 (R = Ph)
is therefore an excellent precursor of aspirin in the presence
of PT-CRL, while less than 50% of the compound acts as
precursor of aspirin in the presence of C-CRL.
o-hydroxybenzoates
9 (R = Me) (Table 3). A completely
reversed selectivity is observed in the hydrolysis of phenyl
4440
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444

View Article Online
To probe the substrate specificity and the enantioselectivity
of C-CRL we carried out molecular modeling studies related to
the hydrolysis of 1-(pyridyl)ethyl acetates and benzoates 1a–c in
terms of their three-dimensional structure.
Fig. 1 reports an example of the structure of the complex
between R- and S-enantiomers of compound 1a (R = Ph) and
the CRL structure obtained by docking experiments.
Ser-450 plays a fundamental role in the enantioselection
of the enzyme. Scheme 1 illustrates the complexes (A and B)
of C-CRL with the two enantiomers (ϩ)-R-1a and (Ϫ)-
S-1a. Serine 450 is able to form a strong hydrogen bond
(about Ϫ4 kcal molϪ1) with the pyridine nitrogen of the R-
enantiomer (complex A). The distance between the nitrogen
and serine hydroxy is 3.2 Å and the orientation of the serine
hydrogen and pyridine nitrogen lone pair are optimal for
interaction.
Conversely, the distance between the nitrogen and hydroxy
group of serine for the S-enantiomer is about 4.0 Å (complex B)
and the pyridyl ring is not optimally oriented for hydrogen
bonding, thus leading to a very weak bond (about 0.3 Kcal
molϪ1). The hydroxy group of Ser-209 therefore, preferentially
attacks the carbonyl of (ϩ)-R 1a (R = Ph) in agreement with
experimental findings. The pyridyl rings of both enantiomers
of 1b (R = Ph) and 1c (R = Ph) are not oriented so as to form
strong hydrogen bonding with Ser-450 and, consequently,
a minor enantioselection is observed. This also accounts for
the low enantioselection observed in the hydrolysis of 1-phenyl-
ethyl benzoate 4, (R = Ph) which cannot form a hydrogen bond
with Ser-450.
The results can be explained considering that: (i) the phenyl
of the benzoyl group [compounds 1a–c (R = Ph); 4 (R = Ph)]
interacts with hydrophobic region 1 more favourably than
the methyl group of the acetatye [compounds 1a–c (R = Me);
4 (R = Me)]; (ii) the pyridyl group can form a hydrogen bond
with Ser-450 and this is particularly advantageous when the
heterocyclic ring is α-substituted; and (iii) the enantioselection
depends on the orientation of the methyl of the chiral center
towards a close hydrophobic region 2 of the active site of the
enzyme.
Amino acids forming the catalytic triad of the enzyme, as
well as those amino acids in contact with 1a are highlighted.
The hydrophobic regions calculated by GRID are shown in
grey. Three hydrophobic regions are present in the active site
of the enzyme. Region 1 generated by Phe-296 and Phe-345
(not shown) is used by both enantiomers to interact with the
enzyme. This interaction is particularly important for both
enantiomers of all the compounds 1a–c (R = Ph) because the
phenyl ligand of the benzoyl group is positioned exactly in the
above mentioned region. Conversely, enantiomers of com-
pounds 1a–c (R = Me) only partially interact with the methyl
moiety of acetyl groups in this region, thereby recovering a little
stabilization energy. The hydrophobic region 2 is generated
by Val-127 and Phe-296 and is important for discriminating
between the selective interaction with the R- and S-enantiomers.
In fact, compound R-1a (R = Ph) can favourably position the
methyl moiety of the chiral center, in this region (see Fig. 1-A),
while the S-1a enantiomer tends to orient the methyl group out
of this hydrophobic region and toward the catalytic histidine
side-chain His-449 (see Fig. 1-B). The latter orientation tends to
disfavour the binding energy as well as the catalytic activity
thereby disrupting the exact orientation of the catalytic triad.
Hydrophobic region 3, close to Leu-302 (not shown), is
not used by either of the two enantiomers of 1a and therefore
cannot be responsible for the observed enantioselectivity.
It has been reported14,15 that the purification of crude CRL
(C-CRL) by treatment with simple aliphatic alcohols play an
important role in the increase of enantioselection of esters con-
taining stereocenters in the acid portion of the ester. In the
present paper we put in evidence that propan-2-ol-treated CRL
(PT-CRL) is less enantioselective than C-CRL toward esters
containing stereocenters in the alcohol portion of the ester.
This result is in agreement with the finding of Kazlauskas
and co-workers14b in the hydrolysis of ( )-menthyl acetate and
( )-(2-acetoxy-1-naphthyl)methylphenylphosphine oxide. The
results from a broader spectrum of compounds are necessary to
determine the general applicability of this rule.
In conclusion, CRL is not always able to accept and
selectively recognize the methyl ester function derived from
a carboxylic group or the acetoxy function derived from a
hydroxy group of alcohols and phenols (this is the common
Fig. 1
Scheme 1
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444
4441

View Article Online
way to functionalize the substrate submitted to enzymatic
hydrolysis); no improvement is observed by using non-covalent
modification of the enzyme (PT-CRL). The problem can be
overcome by derivatising the carboxylic acid by phenylation
instead of methylation and by using the benzyloxy derivative
of an alcohol instead of its acetate.
1-(2-Pyridyl)ethanol 2a. Yield 70%, colourless crystals, mp
27–28 ЊC (from n-hexane), MS (EI) m/z (rel. intensity) 52 (33),
80 (50), 108 (100), 121 (70); 1H-NMR δ 1.15 (d, 3H, CH3,
J = 6.3 Hz), 4.32 (br s, 1H, OH), 4.90 (q, 1H, CH, J = 6.3 Hz),
7.19–7.23 (m, 1H, Harom), 7.29 (d, 1H, Harom, J = 7.9 Hz), 7.70
(ddd, 1H, Harom, J = 7.7, 7.7, 1.7 Hz), 8.55 (d, 1H, Harom, J = 4.9
Hz).
In accord with the molecular modeling related to the
hydrolysis of 1-(pyridyl)ethyl acetates and benzoates a quali-
tative suggestion can be advanced to explain the complete
chemoselectivity of PT-CRL observed in the hydrolysis of
phenyl o-acetoxybenzoate 7 (R = Ph) with respect to methyl
o-acetoxybenzoate 7 (R = Me). The phenyl (ligand) of the
benzoyl group of 7 (R = Ph), and 1a–c (R = Ph) compounds,
likes to be positioned in the region 1 and the phenyl of phenoxy
group prefers to be positioned in the hydrophobic region 2
so that only the benzoate function is favourably oriented for
hydrolysis. Conversely, methyl o-acetoxybenzoate 7 (R = Me)
only partially interacts with the methyl moiety of the acetyl
group in the region 1, thereby recovering a little stabilization
energy. Moreover the substrate may change the orientation and
the position in the active site thus favoring the hydrolysis of
the acetate function.
1-(3-Pyridyl)ethanol 2b. Yield 97%, oil, MS (EI) m/z (rel.
1
intensity) 53 (20), 80 (80), 108 (100), 123 (Mϩ, 35); H-NMR
δ 1.52 (d, 3H, CH3, J = 6.4 Hz), 2.17 (br s, 1H, OH), 4.97
(q, 1H, CH, J = 6.4 Hz), 7.29 (dd, 1H, Harom, J = 8.0, 5.0 Hz),
7.74 (ddd, 1H, Harom, J = 7.9, 1.7, 1.7 Hz), 8.52 (dd, 1H, Harom
J = 4.9, 1.3 Hz), 8.60 (s, 1H, Harom).
,
1-(4-Pyridyl)ethanol 2c. Yield 83%, colourless crystals,
mp 55–58 ЊC (from n-hexane), MS (EI) m/z (rel. intensity)
43 (60), 51 (70), 78 (100), 106 (90), 121 (75); 1H-NMR δ 1.48 (d,
3H, CH3, J = 6.5 Hz), 3.5 (br s, 1H, OH), 4.88 (q, 1H, CH,
J = 6.5 Hz), 7.29 (dd, 2H, Harom, J = 4.5, 1.6 Hz), 8.45 (dd, 2H,
Harom, J = 4.5, 1.6 Hz).
The higher activity and chemoselectivity of PT-CRL,
compared to C-CRL, in the hydrolysis of phenyl o-acetoxy-
benzoate 7 (R = Ph) probably occurs because the propan-2-ol
treatment causes a conformational change14,23 of the lipase
which influences the structure of the active site and the flexi-
bility of the enzyme. It is therefore possible that the enzyme
becomes more flexible15 and the substrate 7 (R = Ph) can be
positioned favourably in the active site (induced fit enzyme24)
for the hydrolysis.
The use of CRL has been extended which is of practical
interest because it is commercially available, stable and inexpen-
sive. Moreover the proposed model should also be useful in
estimating the mechanism of hydrolysis in the presence of CRL
for pro-drug or soft-drug candidates ahead of their synthesis.
Preparation of racemic esters 1a–c (R ؍ Ph, Me) and 4 (R ؍ Ph,
Me)
The alcohol (40 mmol) and dry pyridine (8 ml) were mixed
at 0 ЊC under N2 atmosphere. The acetyl or benzoyl chloride
(60 mmol) was then added dropwise during 5–10 min. The mix-
ture was left at room temperature for 1 h and then quenched
with water and extracted with EtOAc. The organic layer
was washed with water (3 × 20 ml), dried over Na2SO4 and
evaporated at reduced pressure. The residue was purified by
column chromatography on silica gel eluted with ethyl ether
to give acetates and with ethyl ether–petroleum ether 1:1 to
give benzoates. Chemical yields and spectroscopic data are
described as follows:
1-(2-Pyridyl)ethyl acetate 1a, (R ؍ Me). Yield 95%, oil, MS
(EI) m/z (rel. intensity) 43 (50), 78 (33), 105 (80), 123 (100),
Experimental section
General
1
165 (Mϩ, 30); H-NMR δ 1.60 (d, 3H, CH3, J = 6.7 Hz), 2.12
(s, 3H, COCH3), 5.92 (q, 1H, CH, J = 6.7 Hz), 7.23–7.27 (m,
1H, Harom), 7.35 (dm, 1H, Harom, J = 7.8 Hz), 7.68 (ddd, 1H,
Harom, J = 7.8, 7.8, 1.8 Hz), 8.60 (dm, 1H, Harom, J = 4.5 Hz).
Candida rugosa lipase (crude CRL E.C.3.1.1.13 type VII)
was purchased from Sigma. PT-CRL was purified as recently
described.15a 2-, 3-, 4-Acetylpyridine and 1-phenylethanol were
obtained from Aldrich Chemical Co. Compound 7 was pre-
pared according to the literature.16a Compounds 2a–c are not
well characterized in the literature; compounds 3a–c (R = Ph)
and 4 (R = Ph) are new compounds. The optical rotations of
2a–c, 3a–c 5 and 6 are known.18a,24 All the organic solvents
were of reagent grade and used without further purification
1-(3-Pyridyl)ethyl acetate 1b, (R ؍ Me). Yield 85%, oil, MS
(EI) m/z (rel. intensity) 43 (50), 78 (45), 105 (85), 123 (100),
1
165 (Mϩ, 20); H-NMR δ 1.57 (d, 3H, CH3, J = 6.6 Hz), 2.08
(s, 3H, COCH3), 5.90 (q, 1H, CH, J = 6.6 Hz), 7.25–7.32 (m,
1H, Harom), 7.66 (dm, 1H, Harom, J = 7.9 Hz), 8.55 (dd, 1H,
Harom, J = 4.8, 1.7 Hz), 8.61–8.62 (m, 1H, Harom).
1
(except for pyridine which was distilled over KOH). H-NMR
1-(4-Pyridyl)ethyl acetate 1c, (R ؍ Me). Yield 80%, oil, MS
(EI) m/z (rel. intensity) 43 (50), 78 (35), 106 (50), 123 (100),
spectra were recorded in CDCl3 solution on a Bruker AC 200
MHz spectrometer. GC analyses were performed on a Hewlett-
Packard 5890 chromatograph with HP-5-fused silica capillary
column (30 m, 0.25 internal diameter, 0.25 µm film thickness),
an “on-column” injector system, a FID detector and H2 as
gas carrier. Mass spectra were recorded at 70 eV on a GC-MS
apparatus. The specific optical rotations were measured on a
JASCO-DIP 360 polarimeter in CHCl3 or MeOH solution.
1
165 (Mϩ, 20); H-NMR δ 1.50 (d, 3H, CH3, J = 6.7 Hz), 2.11
(s, 3H, COCH3), 5.83 (q, 1H, CH, J = 6.7 Hz), 7.24 (d, 2H,
Harom, J = 5.7 Hz), 8.58 (d, 2H, Harom, J = 5.7 Hz).
1-(2-Pyridyl)ethyl benzoate 1a, (R ؍ Ph). Yield 82%, oil, MS
(EI) m/z (rel. intensity) 51 (20), 77 (45), 106 (65), 122 (100),
182 (10); 1H-NMR δ 1.75 (d, 3H, CH3, J = 6.65 Hz), 6.19
(q, 1H, CH, J = 6.65 Hz), 7.15–7.25 (m, 1H, Harom), 7.42–7.49
Preparation of racemic pyridylethanols 2a–c
(m, 3H, Harom), 7.54–7.58 (m, 1H, Harom), 7.68 (ddd, 1H, Harom
J = 7.8, 7.8, 1.8 Hz), 8.12 (dm, 2H, Harom, J = 6.9 Hz), 8.66 (dm,
1H, Harom, J = 5.0 Hz).
,
NaBH4 (52 mmol) was added to a stirred solution of acetyl-
pyridine (41 mmol) in MeOH (20 ml) at room temperature.
After being stirred for 5 h, the mixture was poured into water
and extracted with EtOAc. The organic layer was dried with
Na2SO4 and then evaporated at reduced pressure. The alcohol
was obtained with GC purity 95% and then used without
further purification. Chemical yields and spectroscopic data are
described as follows:
1-(3-Pyridyl)ethyl benzoate 1b, (R ؍ Ph). Yield 72%, oil, MS
(EI) m/z (rel. intensity) = 51 (35), 77 (65), 105 (100), 122
1
(40), 227 (Mϩ, 65); H-NMR δ 1.7 (d, 3H, CH3, J = 6.64 Hz),
6.15 (q, 1H, CH, J = 6.64 Hz), 7.25–7.57 (m, 4H, Harom), 7.76
4442
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444

View Article Online
(dm, 1H, Harom, J = 7.9 Hz), 8.06 (dm, 2H, Harom, J = 6.9 Hz),
8.62–8.63 (m, 1H, Harom), 8.72 (s, 1H, Harom).
Filmap27 were then used in tandem. Minim finds and lists
all the energy minima in the grid map for the potassium
counterions. Filmap then used a simulated annealing procedure
to postprocess the list of the minima in order to select the
minima locations subset that gives the most favourable overall
interaction energy with the counterions layer. Hydrophobic
interaction in the active site of the protein was computed by the
GRID program. GRID basically is a computational procedure
for detecting energetically favourable binding sites on macro-
molecules. The program works by defining a 3D grid of points
that contains the chosen substrate-binding site. At each node of
the grid, the energy between the probe and the target is calcu-
lated as the electrostatic, hydrogen bond and Lennard–Jones
interactions of chemically selective probes with the chosen
target structure. For hydrophobic interactions a methyl probe
group interacting with the enzyme was used. Ligand structures
(compound 1a (R = Ph, R- and S-enantiomers)) were generated
with the SYBYL28 program and energy-minimised using the
Amsol29 force field to which solvation energy was added. Con-
formational analysis was not required. After conversion to the
appropriate format, the output conformers of compounds 1a
and 2a and the enzyme structure were used by the program
Autodock to find the docked conformations of compounds 1a
and 2a in the active site of CRL enzyme. Autodock is a state of
the art tool for identifying possible binding interactions
between macromolecules one of which is considered rigid
(the protein) and the other (the ligand) is considered to have a
flexible structure. Autodock automatically performs conforma-
tional analysis of the ligand when docking procedure is applied.
During the docking operations, Autodock30 moves the ligand
into the site, modifies the conformation of the ligand molecule
and evaluates the energy of the various complexes as generated.
In our work, only the minimum energy complexes were selected
and reported for interpretation purposes. It is interesting to
note that this procedure is relatively independent of the initial
orientation and conformation of the modeled docked ligands.
1-(4-Pyridyl)ethyl benzoate 1c, (R ؍ Ph). Yield 77%, oil, MS
(EI) m/z (rel. intensity) 51 (35), 77 (65), 105 (100), 122 (40), 227
(Mϩ, 65); 1H-NMR δ 1.65 (d, 3H, CH3, J = 6.65 Hz), 6.45
(q, 1H, CH, J = 6.65 Hz), 7.30 (dd, 2H, Harom, J = 4.5, 1.6 Hz),
7.40–7.54 (m, 3H, Harom), 8.08 (dm, 2H, Harom, J = 6.9 Hz), 8.59
(dd, 2H, Harom, J = 4.5, 1.6 Hz).
1-Phenylethyl acetate 4, (R ؍ Me). Yield 88%, oil, MS (EI)
m/z (rel. intensity) 43 (82), 51 (31), 77 (54), 104 (88), 122 (100),
1
165 (Mϩ, 0.5); H-NMR δ 1.56 (d, 3H, CH3, J = 6.6 Hz), 2.09
(s, 3H, OCH3), 5.91 (q, 1H, CH, J = 6.6 Hz), 7.30–7.39 (m, 5H,
Harom).
1-Phenylethyl benzoate 4, (R ؍ Ph). Yield 86%, oil, MS (EI)
m/z (rel. intensity) = 51 (25), 77 (55), 105 (100), 226 (Mϩ,
35); 1H-NMR δ 1.67 (d, 3H, CH3, J = 6.6 Hz), 6.14 (q, 1H, CH,
J = 6.6 Hz), 7.32–7.56 (m, 8H, Harom), 8.08 (dm, 2H, Harom
,
J = 6.9 Hz).
Enzymatic hydrolysis in presence of crude CRL
In a standard experiment, crude CRL (120 mg) was suspended
in water (12 ml) and phosphate buffer (NaH2PO4–Na2PO4,
0.1 M, pH 7.2, 2 ml), stirred at room temperature and the
pH adjusted to 7.2. Racemic ester (1 mmol) was added and the
mixture was maintained at pH 7.2 by automatic titration with
NaOH 0.2 M using a Mettler DK pH-stat under vigorous
stirring. When the hydrolysis reached the conversion indicated
in the tables, the reaction was stopped adding a saturated solu-
tion of NaCl (10 ml), the solution was filtered through Celite
and the mixture was extracted with EtOAc (3 × 20 ml). The
organic layer was dried with Na2SO4 and concentrated in vacuo.
Alcohol and ester were separated by column chromatography
on silica gel using eluents described above.
Acknowledgements
Enzymatic hydrolysis in presence of PT-CRL
The Ministero dell’Università e della Ricerca Scientifica e
Tecnologica (MURST) Consiglio Nazionale delle ricerche
(CNR, Rome) and Università degli Studi di Perugia are
thanked for financial support.
12 ml of solution of PT-CRL (225 units with p-NPA assay) and
phosphate buffer (NaH2PO4–Na2PO4, 0.1 M, pH 7.2, 2 ml)
were stirred at room temperature and the pH adjusted to
7.2. Racemic ester (1 mmol) was added and the mixture was
maintained at pH 7.2 by automatic titration with NaOH 0.2 M
under vigorous stirring. When the hydrolysis reached the con-
version indicated in the tables, the reaction mixture was worked
up as described above.
References
1 C. H. Wong and G.-M. Whitesides, Enzymes in Organic Chemistry,
Pergamon, New York, 1994.
2 P. Grochulski, Y. Li, J. D. Schrag, F. Bouthillier, P. Smith,
D. Harrison, B. Rubin and M. Cygler, J. Biol. Chem., 1993, 268,
12843; P. Grochulski, Y. Li, J. D. Schrag and M. Cygler, Protein
Sci., 1994, 3, 82.
3 M. Cygler, P. Grochulski, R. J. Kazlauskas, J. D. Schrag,
F. Bouthillier, B. Rubin, A. N. Serreqi and A. K. Gupta, J. Am.
Chem. Soc., 1994, 116, 3180.
4 M. L. Rua, T. Diaz-Maurino V. M. Fernandez, C. Otero and
A. Ballesteros, Biochim. Biophys. Acta, 1993, 1156, 181.
5 J. V. Sinisterra, E. F. Llama, C. Del Campo, M. J. Cabezas,
J. M. Moreno and M. J. Arroyo, J. Chem. Soc., Perkin Trans. 2,
1994, 1330; B. Berger and K. Faber, J. Chem. Soc., Chem. Commun.,
1991, 1198.
Determination of the enantiomeric excess
The ee of all alcohols (ϩ)-2a–c and acetates (Ϫ)-3a–c (R = Me)
obtained in the enzymatic hydrolysis was determined by known
specific optical rotation.18a,25 The acetate (Ϫ)-6 (R = Me)
and the benzoates (Ϫ)-3a–c (R = Ph) and 6 (R = Ph) were first
hydrolyzed to the corresponding alcohols (ϩ)-5 and (ϩ)-2
respectively by treatment with NaOH 0.5 M for 3 h at reflux
and the ee were determined by comparing them with known
specific optical rotation.18a,25
6 Q. M. Gu and C. J. Sih, Biocatalysis, 1992, 6, 115.
7 L. Lam, R. Hui and J. B. Jones, Eur. J. Org. Chem., 1998, 1441.
8 T. Miyazawa, S. Kurita, S. Ueji, T. Yamada and S. Kuwata,
Biotechnol. Lett., 1992, 14, 941.
9 S. Parida and J. S. Dordick, J. Am. Chem. Soc., 1991, 113, 2253.
10 E. Holmberg, M. Holmquist, E. Hedenstrom, P. Berglund, T. Norin,
H. E. Hogberg and K. Hult, Appl. Microbiol. Biotechnol., 1992,
35, 572.
11 A. Bhaskar Rao, H. Rehaman, B. Krishnakumari and J. S. Yadav,
Tetrahedron Lett., 1994, 35, 2611.
12 G. Hedstrom, M. Backlund and J. P. Slotte, Biotechnol. Bio-
engineer., 1993, 42, 618.
Molecular modeling
Atomic coordinates of CRL protein were obtained from
Brookhaven National Laboratories Protein Databank PDB
(structure code 1CRL at 2.0 Å resolution). Water molecules
were deleted from the model. The global charge of the enzyme
target was Ϫ4. Since uncharged systems are preferable for
the analysis, potassium counterions were added to the CRL
protein. To perform this task a GRID26 map with potassium
cations was prepared, and the utility programs Minim and
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444
4443

View Article Online
13 J. M. Sànchez-Montero, V. Hamon, D. Thomas and M. D. Legoy,
Biochim. Biophys. Acta, 1991, 1078, 345.
14 (a) S. H. Wu, Z. H. Guo and C. J. Sih, J. Am. Chem. Soc., 1990, 112,
1990; (b) I. J. Colton, N. A. Sharmin and R. J. Kazlauskas, J. Org.
Chem., 1995, 60, 212.
15 (a) A. Cipiciani, M. Cittadini and F. Fringuelli, Tetrahedron,
1998, 54, 7890; (b) A. Cipiciani, F. Bellezza, F. Fringuelli and
M. Stillitano, Tetrahedron: Asymmetry, 1999, 10, 4599.
16 (a) A. Cipiciani, F. Fringuelli and A. M. Scappini, Tetrahedron,
1996, 52, 9869; (b) A. Cipiciani, F. Fringuelli, V. Mancini,
O. Piermatti, A. M. Scappini and R. Ruzziconi, Tetrahedron, 1997,
53, 11853.
(I. J. Colton, N. A. Sharmin and R. J. Kazlauskas, J. Org. Chem.,
1995, 60, 212).
20 Experiments at 40 ЊC in aqueous medium at pH 7.2 and, for
substrate 1a (R = Me), also in water–organic solvent 10:1 v/v
(organic solvent: acetone, acetonitrile, benzene, cyclohexane,
n-hexane), give very low E values.
21 C. S. Chen, Y. Fujimoto, G. Girdauskas and C. J. Sih, J. Am. Chem.
Soc., 1982, 104, 7294.
22 N. M. Nielsen and H. Bunbgaard, J. Pharm. Sci., 1988, 77, 285;
M. Ankersen, K. K. Nielsen and A. Senning, Acta Chem. Scand.,
1989, 43, 213; M. Hundewadt and A. Senning, J. Pharm. Sci., 1991,
80, 545.
17 M. R. Uskokovic, R. L. Lewis, J. J. Partridge, C. W. Despreaux
and D. L. Pruess, J. Am. Chem. Soc., 1979, 101, 6742; A. Franck
and C. Ruchardt, Chem. Lett., 1984, 1431; M. Amat, M. D. Coll,
D. Passarella and J. Bosh, Tetrahedron: Asymmetry, 1996, 7,
2775.
18 (a) M. Seemayer and M. P. Schneider, Tetrahedron: Asymmetry,
1992, 3, 827; (b) K. Kano, S. Negi, A. Kawashima and K.
Nakamura, Enantiomer, 1997, 2, 261; (c) J. Uenishi, T. Hiraoka,
S. Hata, K. Nishiwaki, O. Yonemitsu, K. Nakamura and H.
Tsukube, J. Org. Chem., 1998, 63, 2481.
23 M. Botta, E. Cernia, F. Corelli, F. Manetti and S. Soro, Biochim.
Biophys. Acta, 1997, 1337, 302.
24 F. Theil, Chem. Rev., 1995, 95, 2203.
25 D. Battail-Robert and D. Ganaire, Bull. Soc. Chim. France, 1965, 208.
26 P. J. Goodford, J. Med. Chem., 1985, 28, 849; D. N. A. Bobbyer, P. J.
Goodford and P. M. McWhinnie, J. Med. Chem., 1989, 32, 1083.
27 GRID v. 17, Molecular Discovery Ltd., Oxford, G.B., 1999.
28 SYBYL, Tripos Associates, St. Louis, Missouri, USA.
29 C. J. Cramer, G. D. Hawikins, G. C. Lynch, D. J. Giesen, I. Rossi,
W. J. Storer, D. G. Thrular and D. A. Liotard, QCPE Bull., 1995,
15, 41.
19 PT-CRL was selected for this study because the isopropanol
treatment is easy, gives synthetically useful amounts of enzyme
and increases the enantiopreference of enzyme towards esters.
30 G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Hvey, W. E. Hart,
R. K. Betew and A. J. Olson, J. Comput. Chem., 1998, 19, 1639.
4444
J. Chem. Soc., Perkin Trans. 1, 2000, 4439–4444