Lipase-catalyzed synthesis of a tri-substituted cyclopropyl chiral synthon: A practical method for preparation of chiral 1-alkoxycarbonyl-2-oxo- 3-oxabicyclo[3.1.0]hexane

Article abstract of DOI:10.1016/S0957-4166(99)00392-4

An efficient method for the preparation of optically active enantiomers of 1-ethoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane 1 has been developed. Treatment of 1 with lipase Amano PS gave (1S,5R)-1-carboxy-2-oxo-3- oxabicyclo[3.1.0]hexane 4a which was converted to (1S,5R)-1-ethoxycarbonyl-2- oxo-3-oxabicyclo[3.1.0]hexane 1a with high enantiomeric purity (98.0% ee, 75% yield), while the (1R,5S)-lactone ester 1b remained intact. A simple procedure for the recovery of 4a from the reaction mixture was also established.

Full text of DOI:10.1016/S0957-4166(99)00392-4

Pergamon
Tetrahedron: Asymmetry 10 (1999) 3819–3825
Lipase-catalyzed synthesis of a tri-substituted cyclopropyl chiral
synthon: a practical method for preparation of chiral
1-alkoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane
Takashi Tsuji,∗ Tomoyuku Onishi and Katsutoshi Sakata
Pharmaceutical Research Laboratories, Ajinomoto Co. Inc., 1-1 Suzuki-cho, Kawasaki, 210-8681 Japan
Received 3 August 1999; accepted 31 August 1999
Abstract
An efficient method for the preparation of optically active enantiomers of 1-ethoxycarbonyl-2-oxo-3-oxa-
bicyclo[3.1.0]hexane 1 has been developed. Treatment of 1 with lipase Amano PS gave (1S,5R)-1-carboxy-2-oxo-
3-oxabicyclo[3.1.0]hexane4a which was converted to (1S,5R)-1-ethoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane
1a with high enantiomeric purity (98.0% ee, 75% yield), while the (1R,5S)-lactone ester 1b remained intact. A
simple procedure for the recovery of 4a from the reaction mixture was also established. © 1999 Elsevier Science
Ltd. All rights reserved.
1. Introduction
1-Alkoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexanes such as 1 are useful chiral synthons with a
trisubstituted cyclopropane ring, which have been used for the synthesis of conformationally constrained
amino acids1–4 and, recently, for the synthesis of nucleoside analogues5 with potent antiherpetic activity.
Several synthetic methods for the enantiomers of 1-alkoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane
have been reported. Cycloalkylation of dimethyl malonate with chiral epichlorohydrin gave the methyl
ester of 1 in 93% ee.3 Chiral glycidol triflate4 and chiral cyclic sulfate of glycerol6 were also used for the
same purpose. The disadvantage of these methods is the use of expensive chiral sources, though these
methods gave good enantiomeric excesses.6 An alternative method was reported starting with prochiral
2-(tetrahydropyranyl)hydroxymethyl dimethyl cyclopropane-1,1-dicarboxylate utilizing enantioselective
hydrolysis by an industrial esterase followed by lactonization. About 90% ee was achieved by this method
after two crystallizations; however, the overall yield is quite low (<20%).7 In the course of research using
an enantiomer of 1 as a key intermediate, we attempted to develop an economical enzymatic process
∗
Corresponding author. Tel: 044-210-5822; fax: 044-210-5871; e-mail: plr_tuji@te10.ajinomoto.co.jp
0957-4166/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00392-4

3820
T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
starting with racemic 1 which is easily prepared from epichlorohydrin and diethyl malonate in a single
step.
2. Results and discussion
2.1. Screening of enzymes
To obtain an enzyme to achieve enantioselective hydrolysis, commercial lipases and esterases of
microbial and mammalian origins were screened. From preliminary experiments in an aqueous medium,
some lipases showed slow, but enantioselective hydrolysis of 1. Esterases showed higher hydrolysis
rates, but no stereoselectivity. It was also observed that hydrolysis by lipases was accelerated in the
presence of organic solvents and a two-phase system consisting of phosphate buffer and diisobutylether
was employed for screening. Compound 1 was treated with each enzyme, the reaction was terminated by
addition of ethyl acetate, and the amount of remaining 1a and 1b in the organic layer was quantified by
chiral HPLC. Under these conditions, consumption of some of 1 was observed in the non-enzymatic
control, indicating that 1 was slowly hydrolyzed non-enzymatically. As described later, the lactone
moiety of 1 is easily hydrolyzed under basic conditions and hydroxycarboxylic acid ester 2 was possibly
formed under these conditions (Scheme 1). Thus, the amount of each enantiomer consumed in the
enzymatic reaction was estimated by subtracting that in the non-enzymatic control as a reference, and
the enantiomeric value (E)8 was calculated as an index of selectivity of the enzymatic reaction as shown
in Table 1. Among the lipases and esterases, lipases from Pseudomonas sp. such as lipase PS and lipase
CES showed preferred consumption of (1S,5R)-1-ethoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane 1a
over (1R,5S)-lactone ester 1b. Other lipases from yeast, fungi, and pancreas showed no stereoselective
hydrolysis. Pig liver esterase showed some preference for 1a. From these results, lipase PS was selected
for the further optimization of reaction conditions.
Scheme 1.
2.2. Identification of the reaction product
There are two possible hydrolysis sites in 1, namely the lactone ring and the ethyl ester. Since
the hydroxycarboxylic acid formed by hydrolysis of the lactone moiety easily forms a lactone ring

T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
3821
Table 1
Enzymatic hydrolysis of 1 by lipases and esterasea
a25 mg of 1 was treated with 5 mg of enzyme in 1 ml of 0.1 M phosphate buffer (pH 7.0) and 3 ml diiodobutylether,
and the mixture was stirred for 18 h at 37°C. Compounds 1a and 1b were extracted with 3 ml of ethyl acetate and
b
analyzed by chiral HPLC. See Section 3.2 for details. % of hydrolysis of 1a is calculated as follows: 100×[(1a
in non-enzymatic control)−(1a in enzymatic reaction)]/(1a in non-enzymatic control). % of hydrolysis of 1b is
c
determined in the same way. E is calculated from the extent of enzymatic conversion of the substrate and the
enantiomeric excess of the remaining substrate as described.8
during isolation, the reference compounds were prepared in situ to assign the structure of the enzymatic
hydrolysis product of 1a. Hydroxycarboxylic acid 2 was prepared by treatment of 1 with an equimolar
amount of NaOH. When 1 was treated with an equimolar amount of NaOD in C2D5OD in an NMR tube,
complete disappearance of NMR signals of 1 and appearance of new signals of 2 were observed within
15 minutes at room temperature.9 Compound 1 was treated with two equimolar amounts of alkali to form
dicarboxylic acid 3, which was then cyclized to form lactone carboxylic acid 4 as shown in Scheme 1.
These reference compounds are separable by reversed-phase HPLC and the major hydrolysis product
of 1 by lipase PS was proved to be identical to 4 by reversed-phase HPLC analysis of the products in
the aqueous phase. Small amounts of 2 and 3 were also formed; however, 2 was also formed in the
non-enzymatic control in the same magnitude. On the other hand, neither 3 nor 4 was formed in the
non-enzymatic control. Thus, it was proved that the lipases hydrolyzed the ethyl ester moiety of 1a while
keeping the more reactive lactone moiety intact.
2.3. Optimization of the reaction conditions
Since it was observed that the presence of organic solvents enhanced the hydrolytic activity of lipases
and it is known that the choice of solvents affects the enantiomeric ratio,10,11 further optimization of
the reaction conditions was carried out using two-phase systems. The enzymatic reaction was carried
out in a 1:1 mixture of phosphate buffer (pH 7.0) and various solvents. The reaction was terminated by
addition of CHCl3 and the remaining 1a and 1b in the organic phase was analyzed by chiral HPLC. In
this case, the amount of 1a and 1b hydrolyzed was calculated from a difference between the initial and

3822
T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
remaining amounts as shown in Table 2. The ratio of hydrolyzed 1a and 1b is an index of selectivity of
the enzymatic (the ethyl ester moiety) and non-enzymatic (the lactone moiety) hydrolysis. Use of water
immiscible aliphatic hydrocarbons and higher dialkylethers resulted in high conversion and selectivity.
Aromatic hydrocarbons such as toluene, which is often used as a co-solvent in lipase reactions,12,13 gave
inferior conversion.
Table 2
Total hydrolysis of 1 in various solventsa
a25 mg of 1 was treated with 5 mg of lipase PS in 1 ml each of 0.1 M phosphate buffer (pH 7.0) and organic
solvent, and the mixture was stirred for 21 h at 30°C. Compounds 1a and 1b were extracted with 2 ml of CHCl3
and analyzed by chiral HPLC. b% of total hydrolysis of 1a is calculated as follows: 100×[(1a added)−(1a in
enzymatic reaction)]/(1a added). % of total hydrolysis of 1b is determined in the same way.
The hydrolysis of 1b is mainly caused by a non-enzymatic reaction at the lactone moiety, since the
enzymatic reaction is highly stereospecific as shown in Table 1. The lactone moiety of 1 is slowly
hydrolyzed in buffer and this spontaneous hydrolysis causes a decrease in selectivity. The selectivity is
affected by the rate of the enzymatic hydrolysis at the ethyl ester moiety and the rate of the spontaneous
hydrolysis at the lactone moiety. The latter may depend on the solubility of the substrate and water in
the co-solvents. Generally, a water immiscible solvent is suitable as a co-solvent in lipase reactions.14 In
this case, with the presence of the lactone moiety in the substrate, such co-solvents are effective not only
for the acceleration of the lipase reaction but also for protection from spontaneous hydrolysis. Once 2
is formed it may be possible that 2a is a substrate of lipase; however, from separate experiments it was
proved that 2 is not a substrate for hydrolysis by lipase and the formation of 2 disturbs the progress of
the enzymatic reaction resulting in poor optical purity and yield. On the other hand, 3a is recovered as 4a
by acid treatment. Thus, the key to achieve high efficiency is to accelerate the enzymatic reaction while
minimizing spontaneous hydrolysis of 1. As estimated, hydrolysis of the lactone ring is pH dependent,
and the pH should be kept as low as possible. There is no difference in the rate of enzymatic reaction
between pH 6 and 7 (data not shown). Since the progress of hydrolysis generates acid and makes the
reaction conditions acidic, control of pH is required by using concd buffer or pH control in the actual
process.

T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
3823
2.4. Process for the preparation of 1a
After further studies on several factors including effects of buffer/solvent ratio, reaction pH, ways of
controlling pH, concentration of substrate and buffer, etc., the following process was established. Two
grams of 1 were treated with 50 mg of lipase PS in 10 ml of 1:1 mixture of 0.5 M phosphate buffer (pH
7.0) and petroleum benzin for 23 h at 30°C with pH maintained over 5 by addition of 0.5N NaOH. In
this process, starting the reaction at pH 7 resulted in better enantiomeric excess than starting at lower pH.
Most of the reaction proceeds in the early stages of the reaction time, and the pH is kept as low as 5 in
the later stage of the reaction to avoid hydrolysis of the lactone moiety.
After the enzymatic reaction, intact 1b was recovered by CHCl3 extraction (49.6% recovery, 83% ee)
and 4a was recovered from the acidified aqueous layer by absorption onto a column of hydrophobic
absorbent SP-207 at pH 1 followed by elution with 20% methanol. Compound 3a was converted to 4a by
acidification and also recovered in the same fraction. Compound 1 regenerated from 2 by acid treatment
was eluted with 100% MeOH. Compound 4a was then re-esterified with SOCl2 and EtOH to give 1a in
37.7% yield (75.4% based on 1a) with 98.0% ee. The absolute stereochemistry of 1a was supported by
the previous report3 as well as by X-ray crystallography of the 4-bromophenacyl ester of 4a (data not
shown).
This procedure gave 1a with a higher enantiomeric excess than that expected from the small scale
experiments such as those shown in Table 2. Under these reaction conditions, 1 shows limited solubility
both in water and light petroleum. Most of the substrate remains insoluble and the substrate at the
water/organic solvent interface might be more preferable for lipase hydrolysis as shown in the other
experiments,14 and the release of the product into the aqueous layer manipulates the equilibrium to
promote hydrolysis. A high concentration of the substrate gives better results, minimizing spontaneous
hydrolysis of the lactone moiety.
3. Experimental
3.1. Materials and methods
Reagents and solvents used were the highest quality available commercially. (R)- and (S)-
Epichlorohydrin (>98% ee) were obtained from Daiso (Osaka). Lipase types II and VII were obtained
from Sigma. Other enzymes were obtained from Amano. SEPABEADS SP-207 was purchased from
Mitsubishi Chemicals. Racemic 1-ethoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane 1 was prepared
from racemic epichlorohydrin and diethyl malonate as described previously.5 The enantiomers were
prepared from (R)- and (S)-epichlorohydrin.5 1H NMR spectra were recorded with a Varian XL-300 300
MHz spectrometer, using tetramethylsilane as an internal standard. Mass spectra were recorded on a Jeol
JMS-D300 spectrophotometer. Optical rotations were determined with a Jasco DIP370 polarimeter.
3.2. Analysis of hydrolysis products
The enantiomers of 1-ethoxycarbonyl-2-oxo-3-oxabicyclo[3.1.0]hexane 1 were separated in a Chiral-
pack AD column (250 mm×4.6 mm Φ, Daicel) using n-hexane and ethanol (80:20) as eluent at a flow
rate of 1 ml/min, monitoring UV absorbance at 230 nm. The enantiomeric excess of 1 was determined by
this method. For the screening of enzymes, 5 mg of enzymes and 25 mg of 1 in 50 µl acetonitrile were
incubated at 37°C for 18 h in a mixture of 1 ml of 0.1 M phosphate buffer and 3 ml of diisobutylether.

3824
T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
The reaction was terminated by an addition of 3 ml of ethyl acetate, and 5 µl each of the organic layer
was subjected for the analysis of the enantiomers of 1.
The reaction products in the aqueous phase were analyzed by reversed-phase HPLC using YMC-
Pack ODS-AM302 (150 mm×4.6 mm Φ, YMC) with a gradient elution of 0.1% trifluoroacetic acid
and acetonitrile at a flow rate of 1 ml/min, monitoring UV absorbance at 230 nm. Compound 2 was
prepared by an addition of an equimolar amount of NaOH to 1 in EtOH. Compound 3 was prepared by
treating 1 with two equimolar amount of NaOH. Since isolation of 2 and 3 was unsuccessful because
of spontaneous formation of the lactone ring, their structures were confirmed by conversion to 1 and 4,
respectively, under acidic conditions. Purification of 2 and 3 by reversed-phase HPLC under the condition
as above gave 1 and 4 quantitatively.
3.3. Enzymatic process for preparation of 1a and 1b
Two grams of 1 and 50 mg of lipase Amano PS were suspended in 5 ml of 0.5 M phosphate buffer
(pH 7.0) and 5 ml of petroleum benzin. The mixture was stirred for 23 h at 30°C with pH controlled at
5.0 by addition of 0.5N NaOH. The pH of the reaction mixture was adjusted to 7.0 by adding saturated
NaHCO3, and the products were extracted twice with 30 ml and 15 ml of CHCl3. The organic layer was
concentrated in vacuo to give 992 mg (49.6%) of 1b with 82.9% ee.
The aqueous layer was acidified to pH 1.0 by addition of 3 ml of concd HCl and then applied onto
a column containing 80 ml of SEPABEADS SP-207 previously treated with MeOH overnight, washed
with water, and equilibrated with 0.1N HCl. The column was washed with 120 ml of 0.1N HCl and 80 ml
of water, and 4a was eluted with 300 ml of 20% MeOH. The fractions containing 4a were concentrated
in vacuo and the residue was dissolved in 20 ml of EtOH and treated with 1.0 ml of thionyl chloride at
−10°C, and the mixture was held overnight at room temperature. After concentration in vacuo, the residue
was dissolved in 20 ml of CHCl3, washed with saturated NaHCO3, dried over anhydrous MgSO4, and
concentrated in vacuo to give 754 mg (37.7%) of 1a with 98.0% ee.
Compound 1 derived from 2 in the reaction mixture was eluted from the column with 100% MeOH.
After concentration in vacuo, 91.1 mg (4.6%) of 1b (48.1% ee) was recovered.
Compound 1a: colorless oil; 1H NMR (CDCl3) δ 1.31 (t, J=7.1 Hz, 3H, CH3CH2), 1.37 (dd, J=4.8, 5.4
Hz, 1H, CH2 of cyclopropane), 2.08 (dd, J=4.8, 8.0 Hz, 1H, CH2 of cyclopropane), 2.72 (m, 1H, CH of
cyclopropane), 4.18 (d, J=9.6 Hz, 1H, CH2 of lactone), 4.27 (q, J=7.1 Hz, 2H, CH3CH2), 4.36 (dd, J=4.5,
9.6 Hz, 1H, CH2 of lactone); FAB MS 170 (M+); [α]D25=−147.6 (c=1.20, EtOH) (lit.5 [α]25D=−146.6
(c=1.22, EtOH)).
1
Compound 4a: waxy oil; H NMR (CDCl3) δ 1.56 (dd, J=4.5, 5.6 Hz, 1H, CH2 of cyclopropane),
2.15 (dd, J=4.5, 7.8 Hz, 1H, CH2 of cyclopropane), 2.96–3.30 (m, 1H, CH of cyclopropane), 4.31 (d,
J=9.6 Hz, 1H, CH2 of lactone), 4.47 (dd, J=5.0, 9.6 Hz, 1H, CH2 of lactone); FAB MS 143 (MH+);
[α]D25=−159.3 (c=1.22, MeOH). Anal. calcd for C6H6O4: C, 50.71; H, 4.26; found: C, 50.35; H, 4.50.
References
1. Bland, J.; Shah, A.; Bartolussi, A.; Stammer, C. H. J. Org. Chem. 1988, 53, 992–995.
2. Stammer, C. H. Tetrahedron 1990, 46, 2231–2254.
3. Pirrung, M. C.; Dunlap, S. E.; Trinks, U. P. Helv. Chim. Acta 1989, 72, 1301–1310.
4. Burgess, K.; Ho, K.-K.; Ke, C.-Y. J. Org. Chem. 1993, 58, 3767–3768.
5. Sekiyama, T.; Hatsuya, S.; Tanaka, Y.; Uchiyama, M.; Ono, N.; Iwayama, S.; Oikawa, M.; Suzuki, K.; Okunishi, M.; Tsuji,
T. J. Med. Chem. 1998, 41, 1284–1298.
6. Burgess, K.; Ho, K.-K. Tetrahedron Lett. 1992, 33, 5677–5680.

T. Tsuji et al. / Tetrahedron: Asymmetry 10 (1999) 3819–3825
3825
7. Fliche, C.; Braun, J.; Goffic, F. L. Synth. Commun. 1991, 21, 1429–1432.
8. Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc. 1982, 104, 7294–7299.
1
9. H NMR (C2D5OD) of sodium salt of 2 is as follows: δ 1.19–1.38 (m, 2H, CH2 of cyclopropane), 1.27 (t, J=7.2 Hz, 3H,
CH3CH2), 1.91–2.02 (m, 1H, CH of cyclopropane), 3.44 (dd, J=9.4, 12.3 Hz, 1H, CH2OD), 3.86 (dd, J=5.8, 12.3 Hz, 1H,
CH2OD), 4.16 (q, J=7.2 Hz, 2H, CH3CH2).
10. Anthonsen, T.; Jongejan, J. A. Meth. Enzymol. 1997, 286, 473–495.
11. Anthonsen, T.; Hoff, B. H. Chem. Phys. Lipids 1998, 93, 199–207.
12. Matsumae, H.; Furui, M.; Shibatani, T. J. Ferment. Bioeng. 1993, 75, 93–98.
13. Patel, R. N.; Robson, R. S.; Szarka, L. J. Appl. Microbiol. Biotechnol. 1990, 34, 10–14.
14. Laane, C.; Boeren, S.; Vos, K.; Veeger, C. Biotechnol. Bioeng. 1987, 30, 81–87.