Lipase-catalyzed highly enantioselective kinetic resolution of racemic α-hydroxy butenolides

Article abstract of DOI:10.1016/j.tetasy.2004.06.047

The cyclization of 1,3-bis(trimethylsilyloxy)-1,3-butadienes 2a-e with oxalyl chloride afforded functionalized γ-alkylidene butenolides 3a-e while their subsequent hydrogenation gave racemic α-hydroxy butenolides 4a-e. These were subjected to lipase-catalyzed kinetic resolutions in a biphasic mixture composed of phosphate buffer and 20% toluene. The highest enantioselectivities (E = 67-100) were achieved using lipase from Burkholderia cepacia (Amano PS).

Full text of DOI:10.1016/j.tetasy.2004.06.047

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 2871–2874
Lipase-catalyzed highly enantioselective kinetic resolution of
racemic a-hydroxy butenolides
a,
Anett Kirschner,a,b Peter Langerb, and Uwe T. Bornscheuer
*
*
aDepartment of Technical Chemistry and Biotechnology, Institute of Chemistry and Biochemistry, Greifswald University,
Soldmannstr. 16, D-17487 Greifswald, Germany
bDepartment of Bioorganic Chemistry, Institute of Chemistry and Biochemistry, Greifswald University, Soldmannstr. 16,
D-17487 Greifswald, Germany
Received 28 May 2004; revised 16 June 2004; accepted 21 June 2004
Available online 11 September 2004
Abstract—The cyclization of 1,3-bis(trimethylsilyloxy)-1,3-butadienes 2a–e with oxalyl chloride afforded functionalized c-alkylidene
butenolides 3a–e while their subsequent hydrogenation gave racemic a-hydroxy butenolides 4a–e. These were subjected to lipase-
catalyzed kinetic resolutions in a biphasic mixture composed of phosphate buffer and 20% toluene. The highest enantioselectivities
(E = 67–100) were achieved using lipase from Burkholderia cepacia (Amano PS).
ꢀ 2004 Elsevier Ltd. All rights reserved.
1.Introduction
We have recently reported a new synthesis of a-hydroxy-
c-alkylidenebutenolides by Lewis acid catalyzed cycliza-
tion of 1,3-bis(trimethylsilyloxy)-1,3-butadienes with
oxalyl chloride.4 This method was applied to the synthe-
sis of racemic a-hydroxy butenolides based on diastereo-
selective hydrogenations.5 Herein, we report the
enzymatic resolution of these compounds, which pro-
vides a convenient access to enantiomerically pure
a-hydroxy butenolides.6
Butenolides represent pharmacophoric groups of a vari-
ety of natural products. For example, drugs such as (+)-
leptosphaerin1 or compound WF-3681, which represents
an aldose reductase inhibitor produced by Chaetomella
raphigera2 (Scheme 1), possess a butenolide core struc-
ture. Butenolides are also important building blocks
for the synthesis of natural products. For example,
they have been used during the synthesis of (+)- and
(ꢀ)-eldanolide,3a the antileukaemic lignans (+)-trans-
burseran,3b (ꢀ)-isostegane,3b (+)- and (ꢀ)-steganacin,3c
(ꢀ)-verrucarinolactone3d and chrysanthemic acid
analogues.3e
2.Results and discussion
Racemic butenolides 4a–e were synthesized according to
the literature:4b,5 The cyclization of 1,3-bis-silyl enol
ethers 2a–e prepared from the b-keto esters 1a–e with
oxalyl chloride afforded the c-alkylidene butenolides
3a–e. Hydrogenation (H2, Pd/C)7 of the latter gave race-
mic a-hydroxy butenolides 4a–e (Scheme 2).
Me
HN
Ph
OH
O
H
O
HO
O
O
O
O
Different lipases and esterases8 were investigated for
their activity and selectivity on racemic 4a (Scheme 3).
Out of these, lipase from Burkholderia cepacia (Amano
PS, Amano) was found to hydrolyze the substrate with
high enantioselectivity (E >100, Table 1) in a biphasic
mixture composed of phosphate buffer (10mM,
pH7.5) and 20% (v/v) toluene at 30ꢁC. For the isolation
and determination of the enantiomeric excess, the
non-converted ester 4a was recovered from the reaction
HO
HO
(+)-Leptosphaerin
WF-3681
Scheme 1. Examples of butenolides.
*
Corresponding authors. Tel.: +49 3834 86 4461; fax: +49 3834 86
4373 (P.L.); tel.: +49 3834 86 4367; fax: +49 3834 86 80066 (U.T.B.);
e-mail addresses: peter.langer@uni-greifswald.de; uwe. bornscheuer@
uni-greifswald.de
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.06.047

2872
A. Kirschner et al. / Tetrahedron: Asymmetry 15 (2004) 2871–2874
O
O
1) NEt3, Me3SiCl,
C6H6
R2O
(COCl)2
O
O
H2, Pd/C
R2OH
R2O
Me3SiO
R2O
OSiMe3
R1
R1
R1
R1
O
O
2) LDA
3) Me3SiCl, THF,
-78 ˚C
R2O
+ Me3SiOTf,
CH2Cl2, -78 ˚C
OH
OH
O
O
1a-e
2a-e
3a-e
4a-e
Scheme 2. Synthesis of butenolides 4a–e.
O
O
O
R2O
R1
R2O
HO
R1
R1
a: R1 = OMe, R2 = Me
b: R1 = OEt, R2 = Et
c: R1 = Me, R2 = Me
d: R1 = Et, R2 = Et
e: R1 = Pr, R2 = Et
Amano PS
O
O
O
+
phosphate buffer,
toluene
OH
OH
OH
O
O
O
(+)- 4a-e
(+)- 7a-e
4a-e
(–)- 5a-e
(–)- 6a-e
CH2N2
CH2N2
Scheme 3. Lipase-catalyzed kinetic resolution of 4a–e and derivatization of the acids 5a–e produced. Please note that compounds 6a and 7a, as well
as 6c and 7c, are identical.
Table 1. Kinetic resolution of racemic butenolides 4a–e with lipase from Burkholderia cepacia (Amano PS) in small-scale reactions
Compound
Time (h)
Conversiona (%)
Enantiomeric excess
Ea
b
EeS (%)
b,c
EeP (%)
4a
4b
4c
4d
4e
6
5098
51
97
94
94
94
97
0
>10
90
8
67
>100
509530
41
26
93
49
5098
99
0
>10
a Calculated according to Chen et al.9
b Determined by GC with chiral column.
c Of methyl esters 6a–e.
mixture by extraction with methylene chloride so that its
enantiomeric excess could be determined directly by GC
analysis. For the analysis of the formed free acid 5a, the
mixture was acidified and the acid extracted with diethyl
ether. The free acid was then converted to the corre-
sponding methyl ester 6a with diazomethane, whereby
the hydroxy group at the a-position was also methylated
(Scheme 3). Using this protocol, butenolides 4b–e were
also subjected to lipase-catalyzed resolutions and ana-
lyzed in a similar manner (Table 1). In the cases of non-
hydrolyzed 4b–4e, these were also methylated with
diazomethane on an analytical scale yielding 7b–7e in
order to facilitate GC separation of the enantiomers.
tioselectivities similar to those observed in the small-
scale reactions (Table 2). Specific rotations were deter-
mined for the remaining substrates 4a and 4d and the
free acids 5a and 5d, which were synthesized for the first
time. The determination of the absolute configurations
of all optically active compounds is currently under
investigation.
3.Conclusions
In this paper, we have demonstrated that an efficient
kinetic resolution of a-hydroxy butenolides can be
achieved using lipase catalysis. Thus, these important
building blocks are now available in their enantiomeri-
cally pure forms and can serve as a basis for natural
product synthesis.
This revealed that the butenolides with an alkoxy side
chain (4a–b) were hydrolyzed much faster (6–8h) than
the compounds bearing an alkyl side chain (4c–e, 26–
50h). Enantioselectivities were excellent (E = 94 to
>100) in all cases except for 4b (E = 67). In this case,
slow autohydrolysis was observed, which may explain
the decrease in enantiomeric excesses.
4.Experimental
4.1. General
Preparative-scale experiments using 0.5mmol of racemic
butenolide had been carried out with 4a and 4d so far.
The isolated yields were moderate to good and the enan-
Commercial enzyme preparations were used as supplied.
Specific rotations were recorded on a Perkin–Elmer

A. Kirschner et al. / Tetrahedron: Asymmetry 15 (2004) 2871–2874
2873
Table 2. Kinetic resolution of racemic butenolides 4a and 4d with lipase from Burkholderia cepacia (Amano PS) in preparative scale
Compound
Remaining substrate
Product
Ed
21
D
21
D
a
EeS (%)
b
a,c
EeP (%)
b
Yield (%)
½aŠ
Yield (%)
½aŠ
4a
4d
6095
9097
+40
+21
43
92
6097
ꢀ28
90
>100
ꢀ26
a Determined by GC with chiral column.
b Specific rotation10 recorded at 21ꢁC in MeOH or EtOH, c = 1 (see Experimental).
c Of methyl esters 6a and 6d.
d Calculated according to Chen et al.9
polarimeter 241 at 589nm in MeOH (free acids and
methyl esters) or EtOH (ethyl esters) at c = 1. H and
4.3.3. (+)-(3-Ethyl-4-hydroxy-5-oxo-2,5-dihydrofuran-2-
yl)-acetic acid ethyl ester 4d. Yield: 90%; 97% ee;
½aŠ ¼ þ21 (c 1, EtOH). For spectroscopic data, see
1
21
D
Ref. 5.
13C NMR-spectra were recorded in d6-acetone or
CD3OD on a 300 or 600MHz (Bruker) instrument.
4.2. Synthesis of the a-hydroxy butenolides
4.3.4. (ꢀ)-(Ethyl-4-hydroxy-5-oxo-2,5-dihydrofuran-2-
21
D
yl)-acetic acid 5d. Yield: 60%; 97% ee; ½aŠ ¼ ꢀ26 (c
1
The a-hydroxy butenolides 4a–e were prepared accord-
1, MeOH); H NMR (600MHz, d6-acetone): d = 1.16
(3H, t, J = 7.8Hz, CH3), 2.26 (1H, m, CH2CH3), 2.45
(1H, dd, J = 16.2, 8.4Hz, CH2CH), 2.54 (1H, m,
CH2CH3), 2.98 (1H, dd, J = 16.2, 3.0Hz, CH2CH),
5.24 (1H, dd, J = 8.4, 3.0Hz, CH); 13C NMR
(150MHz, d6-acetone): d = 12.1 (CH3), 18.2 (CH2),
38.3 (CH2), 77.0( CH), 135.5 and 138.8 (C@C), 160.1
(CH2C), 171.1 (CO2H).
ing to literature.4b,5
4.3. General procedure for the enzyme-catalyzed kinetic
resolutions
For small-scale reactions substrates 4a–e (0.025mmol)
and lipase Amano PS from Burkholderia cepacia
(300U, 10mg) were dissolved in phosphate buffer
(400lL, 10mM, pH7.5) and toluene (100 lL). The mix-
tures were shaken in a thermoshaker (Eppendorf, Ham-
burg, Germany) at 30ꢁC and 1000rpm. After certain
time intervals, samples (100lL volume) were taken,
and the same amount of distilled water was added.
The samples were extracted, vortexed and centrifuged
(1min, 13,000rpm) before GC analysis (see below).
4.4. Chiral GC analysis
For GC analysis, free acids 5a–e were converted into the
corresponding methyl esters 6a–e by the addition of an
etheric solution of diazomethane. Excess reagent and
solvent were removed under nitrogen and the residue
dissolved in 10lL methylene chloride for GC analysis.
Enzymatic reactions on a preparative scale were per-
formed in 50mL flasks with 0.5mmol substrate,
6000U of lipase Amano PS dissolved in 8mL phosphate
buffer (10mM, pH7.5) and 2mL toluene. The mixtures
were stirred in a water bath at 30ꢁC until 50% conver-
sion was reached. After addition of the same amount
(10mL) of distilled water, substrate and product were
isolated and separated by extraction. The remaining
ester was extracted with methylene chloride (5 · 5mL)
and after acidification with 2M HCl [(2mL, 10%
(v/v)], the free acid was extracted from the mixture with
diethyl ether (5 · 5mL). The organic phases were dried,
filtrated and the excess solvent removed in vacuo.
GC analyses were carried out on a Shimadzu GC-14A
gas chromatograph with a chiral column (heptakis-
(2,3-di-O-acetyl-6-O-tertbutyldimethylsilyl)-b-cyclodex-
trin). Separation protocols and retention times for the
compounds 4a–e and 5a–e had to be determined first
(Table 3). Esters 4b–e were also reacted with diazometh-
ane in order to methylate the hydroxyl group at the
a-position 7b–e resulting in a better separation. Injection
and detection temperature were always set to 220ꢁC.
Table 3. Chiral GC analysis
4.3.1. (+)-(4-Hydroxy-3-methoxy-5-oxo-2,5-dihydrofuran-
2-yl)-acetic acid methyl ester 4a. Yield: 60%; 95%
ee; ½aŠ ¼ þ40( c 1, MeOH). For spectroscopic data,
Compound
TColumn (ꢁC)
16018.5/21.1
Retention timea (min)
21
D
see Ref. 5.
4a
6a
6b
6cb
6d
6e
15016.5/18.5
145
145
21.8/24.2
15.7/17.5
4.3.2. (ꢀ)-(4-Hydroxy-3-methoxy-5-oxo-2,5-dihydrofu-
.6
14019.6/20
135
145
ran-2-yl)-acetic acid 5a. Yield: 43%; 92% ee;
35.2/36.5
27.4/30.1
15.7/17.5
21
D
½aŠ ¼ ꢀ28 (c 1, MeOH); 1H NMR (300MHz,
7b
7cb
7d
7e
145
14024.1/25.1
135
CD3OD): d = 2.44 (1H, dd, J = 16.3, 8.7Hz, CH2),
2.87 (1H, dd, J = 16.3, 3.6Hz, CH2), 4.17 (3H, s,
CH3O), 5.01 (1H, dd, J = 8.7, 3.6Hz, CH). A small
amount of hydroxy acid resulting from lactone cleavage
was present.
43.9/45.1
a Retention times of the two enantiomers.
b Compounds 6c and 7c are of identical structure.

2874
A. Kirschner et al. / Tetrahedron: Asymmetry 15 (2004) 2871–2874
4. (a) Langer, P.; Stoll, M. Angew. Chem., Int. Ed. 1999, 38,
Acknowledgements
1803; (b) Langer, P.; Schneider, T.; Stoll, M. Chem. Eur. J.
2000, 6, 3204.
5. Langer, P.; Saleh, N. N. R.; Ko¨hler, V. Eur. J. Org. Chem.
2002, 1566.
6. For the syntheses of enantiomerically pure butenolides,
We thank the German National Academic Foundation
(Bonn, Germany) for a scholarship to Anett Kirschner
and the Deutsche Bundesstiftung Umwelt (DBU,
Osnabruck, Germany, AZ13071) and the Fonds der
¨
Chemischen Industrie (Frankfurt, Germany) for finan-
cial support.
see for example: Harcken, C.; Bruckner, R. Angew. Chem.,
¨
Int. Ed. Engl. 1997, 36, 2750, and references cited therein.
7. (a) For related hydrogenations of butenolides, see:
Marino, C.; Varela, O.; de Lederkremer, R. M. Carbo-
hydr. Res. 1990, 220, 145; (b) Jerocic, L. O.; Varela, O. J.;
Cirelli, A. F.; de Lederkremer, R. M. Tetrahedron 1984,
40, 1425.
8. Lipase B from Candida antarctica, Candida rugosa lipase,
pig liver esterase and recombinant p-nitrobenzylesterase
from Bacillus subtilis were also investigated, but little to no
enantioselectivity was found.
References
1. Pallenberg, A. J.; White, J. D. Tetrahedron Lett. 1986,
5591.
2. Namiki, T.; Nishikawa, M.; Itoh, Y.; Uchida, I.; Hashi-
moto, M. Tetrahedron Lett. 1986, 27, 2015.
3. (a) Vigneron, J. P.; Meric, R.; Larcheveque, M.; Debal,
A.; Lallemand, J. Y.; Kunesch, G.; Zagetti, P.; Gallois, M.
Tetrahedron 1984, 40, 3521; (b) Tomioka, K.; Ishiguro, T.;
Koga, K. J. Chem. Soc., Chem. Commun. 1979, 652; (c)
Tomioka, K.; Ishiguro, T.; Iitaka, Y.; Koga, K. Tetrahe-
dron 1984, 40, 1303; (d) Tomioka, K.; Sato, F.; Koga, K.
Heterocycles 1982, 17, 311; (e) Mann, J.; Thomas, A. J.
Chem. Soc., Chem. Commun. 1985, 737.
9. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J.
J. Am. Chem. Soc. 1982, 104, 7294–7299.
10. The specific rotation values for the acid produced should
be regarded with care, as we cannot exclude that during
sample preparation and handling small amounts of by-
products (e.g. due to ring opening of the lactone) might
have formed.