Absolute configurations of monoesters produced by enzyme catalyzed hydrolysis of diethyl 3-hydroxyglutarate

Article abstract of DOI:10.1016/j.tetasy.2004.04.007

Biocatalytic asymmetrizations of diethyl 3-hydroxyglutarate furnish a route to the enantiomers of ethyl 4-cyano-3-hydroxybutanoate. The enantiopreference of different enzymes has been established by chiral chromatography. Conclusive evidence for absolute configurations has been provided by X-ray crystallographic structure determination of co-crystals of the predominant monoester with (R)-phenylethylamine. The predominant enantiopure monoester produced by ammonolysis of diethyl 3-hydroxyglutarate catalyzed by immobilized lipase B from Candida antarctica (Novozym 435) was ethyl (3S)-4-carbamoyl-3-hydroxybutanoate. This was converted to ethyl (3S)-4-cyano-3-hydroxybutanoate in high yield and enantiomeric excess. Growing cells of Acinetobacter lwoffii gave low ee and predominance of the (S)-enantiomer when used for hydrolysis of diethyl 3-hydroxyglutarate as opposed to previous reports. When Novozym 435 was used for hydrolysis it could be re-used 10 times without loss of activity and selectivity.

Full text of DOI:10.1016/j.tetasy.2004.04.007

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 1551–1554
Absolute configurations of monoesters produced by enzyme
catalyzed hydrolysis of diethyl 3-hydroxyglutarate
c
a
b
c
ꢀ
Anders Riise Moen, Bard Helge Hoff, Lars Kristian Hansen, Thorleif Anthonsen and
Elisabeth Egholm Jacobsenc,*
aBorregaard Synthesis, PO Box 162, N-1701 Sarpsborg, Norway
bDepartment of Chemistry, University of Tromsø, N-9037 Tromsø, Norway
cDepartment of Chemistry, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
Received 17 February 2004; accepted 5 April 2004
Abstract—Biocatalytic asymmetrizations of diethyl 3-hydroxyglutarate furnish a route to the enantiomers of ethyl 4-cyano-3-
hydroxybutanoate. The enantiopreference of different enzymes has been established by chiral chromatography. Conclusive evidence
for absolute configurations has been provided by X-ray crystallographic structure determination of co-crystals of the predominant
monoester with (R)-phenylethylamine. The predominant enantiopure monoester produced by ammonolysis of diethyl 3-hydroxy-
glutarate catalyzed by immobilized lipase B from Candida antarctica (Novozym 435) was ethyl (3S)-4-carbamoyl-3-hydroxybuta-
noate. This was converted to ethyl (3S)-4-cyano-3-hydroxybutanoate in high yield and enantiomeric excess. Growing cells of
Acinetobacter lwoffii gave low ee and predominance of the (S)-enantiomer when used for hydrolysis of diethyl 3-hydroxyglutarate as
opposed to previous reports. When Novozym 435 was used for hydrolysis it could be re-used 10 times without loss of activity and
selectivity.
Ó 2004 Published by Elsevier Ltd.
F
1. Introduction
Lipitor has for several years been one of the best selling
drugs. It is used to reduce the natural synthesis of cho-
lesterol in patients who suffer from high cholesterol level
in blood. Lipitor acts by inhibiting the enzyme
hydroxymethyl co-enzyme A reductase, which is essen-
tial for the biosynthesis of cholesterol.1 The active
ingredient in lipitor, atorvastatin (Fig. 1), is a member of
a family of statins, which also comprises lovastatin,2
compactin, simvastatin, fluvastatin, cerivastatin and
rosuvastatin.1 These all contain a C-7 side chain with
two stereocentres. This C-7 moiety has, for the synthesis
of atorvastatin, been made from ethyl (3R)-4-cyano-3-
hydroxybutanoate 1 and has been the focus of devel-
oping efficient syntheses for this chiral building block in
its enantiopure (R)-form. A frequently used way of
producing chiral building blocks is starting with enan-
tiopure natural products. Starting with iso-vitamin C,3
OH OH
CO2H
N
H
N
O
Figure 1. Atorvastatin.
enantiopure nitrile 1 was synthesized using a series of
non-‘Green chemistry’ steps. Based on lactose as the
starting material a range of chiral building blocks are
commercially available; among them starting materials
for 1.4 Biocatalysis may also offer a greener way to
many, in particular chiral, building blocks. This has
been demonstrated in recent years both in the produc-
tion of bulk chemicals such as acryl amide and in the
fine chemicals and drug industry.5
* Corresponding author. Tel.: +47-73596212; fax: +47-73550877;
e-mail: elisabeth.jacobsen@chem.ntnu.no
URL: http://Bendik.chembio.ntnu.no
0957-4166/$ - see front matter Ó 2004 Published by Elsevier Ltd.
doi:10.1016/j.tetasy.2004.04.007

1552
A. Riise Moen et al. / Tetrahedron: Asymmetry 15 (2004) 1551–1554
OH
O
O
OH
O
Cl
OR
HO
OEt
2
4
A
C
OH
O
N
C
B
OEt
E
1
O
OH
O
D
O
OH
O
H2N
OEt
EtO
OEt
5
3
Figure 2. Retrosynthetic analysis for the synthesis of enantiopure target molecule 1 by biocatalysis. Route A, resolution of racemic chlorohydroxy
ester 2; route B, enzyme catalyzed asymmetrization by ‘half’-hydrolysis of diethyl ester 3, followed by route C, conversion of the enantiopure
monoester 4 to nitrile 1; route D ammonolysis of prochiral diester 3 to give enantiopure amide 5, followed by route E, conversion of amide 5 to target
nitrile 1.
2. Results and discussion
values (ꢀ50%). It has previously been reported that the
(R)-monoethyl ester is predominant in hydrolysis cata-
lyzed by chymotrypsin, and moreover, that the reaction
gives the product with very high ee.12 This reported high
ee-value and configuration is based on comparison with
a previously reported specific rotation value for the
monomethyl ester produced by classical resolution via
the cinchonidine salt.13 It has been reported that grow-
ing cells of Acinetobacter lwoffii gives predominance of
the (R)-enantiomer when used for hydrolysis of diethyl
3-hydroxyglutarate with an ee of 80%.14 In our hands
the same organism (ATCC-17925) gave the (S)-enan-
tiomer and with low ee (56%).
We analyzed the possibilities of making 1 by biocatalysis
based on the retrosynthetic scheme shown in Figure 2.
Our first approach was to make 1 from racemic chlo-
rohydroxy ester 2 by enzyme catalyzed kinetic resolution
(route A).6 The drawback with this resolution is that
only 50% of the right enantiomer could be obtained.
This may not be serious since quantitative conversion to
one single enantiomer is possible using the technique of
stereoinversion.7;8 We started a systematic approach by
varying the R-groups, solvents, acyl donors and en-
zymes. The highest E-value of >100 was obtained for the
tert-butyl ester with vinyl propanoate as the acyl donor
and lipase from Rhizomucor miehei as the catalyst.
However, the synthesis of the ester substrate was not
trivial starting with the diketene. Moreover, target ni-
trile 1 is an ethyl ester and a transesterification would be
necessary in order to convert the tert-butyl ester to the
ethyl ester. Hence we looked for an alternative method.
By asymmetric synthesis instead of resolution, a theo-
retical yield of 100% of one enantiomer may be obtained
directly, however, with enzyme catalysis, it is not guar-
anteed that this will be the wanted enantiomer.
Since the specific rotation of the monoethyl ester 4 is a
very small numerical value, we wanted to check the
absolute configuration of the monoester produced by
CALB not only by comparison of the ½aꢁ values, but
D
with a more reliable method. The monoethyl ester pro-
duced by CALB was co-crystallized with (R)-phenyl-
ethylamine with the crystal structure determined to be as
shown in Figure 3. This structure determination shows
conclusively that the monoester produced by CALB is
the (S)-enantiomer and so we consider it safe to assume
that amide 5 has the same configuration. Based on very
reliable chiral GLC analyses we also find that chymo-
trypsin gives a predominance of the (R)-enantiomer 4.
However, in our hands the ee values obtained were
much lower than previously reported12 both for the
diethyl and dimethyl esters.11 Recently, a patent for a
process for producing the monoethyl ester of diethyl 3-
acetoxyglutarate using chymotrypsin, has been reported.
The ee-values are very high and the configuration (R) as
wanted for synthesis of 1.15
An obvious starting material for the asymmetric syn-
thesis would be diethyl 3-hydroxyglutarate 3, a prochiral
substrate. Enantiopure monoester 4 could be obtained
by route B using enzymatic hydrolysis and this in turn
could be converted to the nitrile using chlorosulfony-
lisocyanate (route C).9 An even simpler method would
be enzyme catalyzed ammonolysis via route D to the
enantiopure monoamide 5, which in turn could be
converted to the target nitrile 1 using 1-(3-dimethyla-
minopropyl)-3-ethylcarbodiimide (route E).10
After use in the first hydrolysis of diethyl and dimethyl
3-hydroxyglutarate, the immobilized catalyst CALB was
re-used more than 10 times with retention of high
activity and enantioselectivity. In fact, the activity of the
enzyme increased significantly from the first to the sec-
ond batch of hydrolysis of 3 (Fig. 4).
All of the mentioned strategies were tried using a range
of enzymes.11 By far the highest ee-values were obtained
using lipase B from Candida antarctica (CALB). Indeed,
the ammonolysis reaction gave ee ¼ 98% and 95% yield.
Chymotrypsin catalyzed hydrolysis gave much lower ee-

A. Riise Moen et al. / Tetrahedron: Asymmetry 15 (2004) 1551–1554
1553
H1C
H1B
H1A
N1
C14
H9
H14
H15C
C8
C9
C10
H15A
H13
C15
H10
H11
C13
C12
H15B
C11
H12
H2B
O3
H7A
H4B
H3
C3
H7C
O2
O5
C2
H6A
C5
C4
O1
C7
C1
C6
H7B
H2A
H4A
H3A
H6B
O4
Figure 3. X-ray crystal structure of the co-crystal of (R)-phenylethylamine and the monoethyl ester produced by hydrolysis catalyzed by CALB. The
stereocentre of (R)-phenylethylamine is labelled C-14 and it is inferred that it has (R)-configuration. The carboxylate group of the monoester is
labelled C-1 and O-1, O-2; the ethyl ester group is C-6, C-7. From the X-ray structure it is clear that the stereocentre at C-3 has an (S)-configuration.
61.53, 14.5; enantiomeric excess by chiral GLC 98% (see
100
% Conv.
80
20
D
below), same as amide 5, ½aꢁ ¼þ31:5 (c 1.0, CHCl3),
20
for the (R)-enantiomer ½aꢁ ¼ ꢂ31:3 (c 1.0, CHCl3).16
D
1
2
3
4
5
6
60
40
3.2. Determination of enantiomeric excesses
The product from the above reaction, predominantly
ethyl (3S)-4-cyano-3-hydroxybutanoate 1, was deriva-
tized with trifluoroacetic anhydride and analyzed using
Varian 3400 gas chromatograph equipped with a chiral
CP-Chirsil Dex CB column (25 m, 0.32 mm i.d., 0.25 lm
film thickness), temperature program 90–95 °C, 0.2 °C/
min, 95–180 °C, 15 °C/min, column pressure 7.5 psi and
split flow 60 mL/min. Retention times (R) ¼ 15.4 min,
(S) ¼ 15.9 min, resolution RS ¼ 1:6. Chiral GLC of 4-
ethoxycarbonyl-3-hydroxybutanoic acid and ethyl 4-
carbamoyl-3-hydroxybutanoate have been described
earlier.11
20
0.0
0.0
10.0 20.0
30.0
40.0
50.0 60.0
Reaction time (min)
70.0
Figure 4. Conversion in six subsequent CALB catalyzed hydrolyses of
diethyl 3-hydroxyglutarate 3. The reaction time increased from reac-
tion 1 to 6.
3. Experimental
Cells of A. lwoffii (ATCC-17925) were grown and used
as described previously.14 The resulting monoester was
analyzed as described earlier11 with the results show-
ing that the (S)-enantiomer was formed with an ee of
56%.
3.1. Ethyl (3S)-4-cyano-3-hydroxybutanoate 1
Ethyl (3S)-4-carbamoyl-3-hydroxybutanoate (1.0 g,
5.7 mmol) was added to CH2Cl2 (15 mL), pyridine
(0.89 g, 11.4 mmol) and 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide, EDCI (1.78 g, 11.4 mmol). The mix-
ture was stirred at room temperature for 4 days and the
reaction monitored using GLC. After the addition of
water (10 mL) and shaking, the water layer was removed
and the solution dried over MgSO4. The solvent was
removed under vacuum, yield: 0.81 g (90%) MS: M+1
3.3. Determination of the absolute configuration of
monoethyl ester
Monoethyl ester produced by CALB hydrolysis (0.5 g,
2.8 mmol) was dissolved in Et2O (15 mL) and (R)-
phenylethylamine (0.34 g, 2.8 mmol) was added. After
cooling on ice, crystals were formed. The X-ray crys-
tallographic structure is shown in Figure 3 and proves
the identity as being (3S)-3-hydroxypentanedioic mono-
ethyl ester.
1
158, m=z 130, 112, 88, 84, 43; H NMR (Bruker DPX
300, 300 MHz; CDCl3): 1.22 (t, 3H), 4.15 (q, 2H), 4.30
(m, 1H), 2.60 (m, 4H), 13C NMR (Bruker DPX 300,
75 MHz, CDCl3): 174.71, 41.40. 64.41, 25.65, 117.73,

1554
A. Riise Moen et al. / Tetrahedron: Asymmetry 15 (2004) 1551–1554
3.4. X-ray crystallography
References and notes
1. Istvan, E. S.; Deisenhofer, J. Science 2002, 292, 1160–
1164.
Diffraction data were collected on a CAD-4 diffracto-
meter (hmax ¼ 23°) using graphite monochromated
2. Henwood, J. M.; Heel, R. C. Drugs 1988, 36, 429–454.
3. Andelic, I.; Fure, R.; Leivestad, T.; Weiseth, M.; Daniel-
sen, K.; Aarum, P.; Josefsen, S. Abstract 15th Norwegian
Organic Wintermeeting, Norwegian Chemical Society,
Geilo, Norway, 2000; p 46.
ꢀ
Mo-Ka radiation (k ¼ 0:71069 A). The structure was
solved by direct methods and refined using the inte-
grated program package OSCAIL.17
3.5. X-ray data
4. Synthon Chiragenics Corp., New Jersey, USA.
5. Schmid, A.; Dordick, J. S.; Hauer, B.; Kiener, A.;
Wubbolts, M. C.; Witholt, B. Nature 2001, 409, 258–268.
6. Hoff, B. H.; Anthonsen, T. Tetrahedron: Asymmetry 1999,
10, 1401–1412.
Orthorhombic
space
b ¼ 14:546ð5Þ,
group
P212121
c ¼ 19:477ð6Þ A,
with
ꢀ
a ¼ 5:9796ð16Þ,
3
ꢀ
V ¼ 1694:1ð9Þ A and one molecule in the asymmetric
unit. The structure was refined to R ¼ 0:086 and
Rw ¼ 0:27 (the crystal was of poor quality) using 1394
unique reflections. H-atom parameters were not refined.
CCDC 235996.18
€
7. Vanttinen, E.; Kanerva, L. T. Tetrahedron: Asymmetry
1995, 6, 1779–1786.
8. Liu, H.-L.; Anthonsen, T. Chirality 2002, 14, 25–27.
9. Luhaus, G. Org. Synth. 1988, 6, 304–305.
10. Bose, D. S.; Sunder, K. S. Synth. Commun. 1999, 29, 4235–
4239.
11. Jacobsen, E. E.; Hoff, B. H.; Riise Moen, A.; Anthonsen,
T. J. Mol. Catal. B 2003, 21, 55–58.
3.6. Re-use of Novozym 435
12. Cohen, S. G.; Khedouri, E. J. Am. Chem. Soc. 1961, 83,
4228–4233.
To diethyl 3-hydroxyglutarate 3 (3.0 g, 14.7 mmol) was
added buffer (15 mL) and CALB (0.5 and 0.05 g) and the
reaction monitored using a pH-stat. The amount of
added base (1 M NaOH) expressed the rate of reaction.
The catalyst was separated by filtration and re-used 10
times under the same conditions.
13. Serck-Hanssen, K. Arkiv Kemi 1956, 10, 135–149.
14. Gopalan, A. S.; Sih, C. J. Tetrahedron Lett. 1984, 25,
5235–5238.
€
15. Ohrlein, R.; Baisch, G.; End, N.; Burkhardt, S.; Studer,
M. Patent WO 03/004450 A, 1-55, Germany, 2003.
16. Wang, G.; Hollingsworth, R. I. Tetrahedron: Asymmetry
1999, 10, 1895–1901.
17. McArdle, P. J. Appl. Cryst. 1993, 26, 752.
18. Crystallographic data have been deposited with the
Cambridge Crystallographic Data Centre. Copies of the
data can be obtained free of charge on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ [fax:
+44(0)-1223-336033 or e-mail: deposit@ccdc.cam.ac.uk].
Acknowledgements
We are grateful to Novozymes, Bagsværd, Denmark for
a gift of Novozym 435.