Nitrile biotransformations for highly enantioselective synthesis of oxiranecarboxamides with tertiary and quaternary stereocenters; efficient chemoenzymatic approaches to enantiopure α-methylated serine and isoserine derivatives

Article abstract of DOI:10.1021/jo0482615

(Chemical Equation Presented) Biotransformations of a number of differently substituted and configured oxiranecarbonitriles using Rhodococcus sp. AJ270, a microbial whole-cell catalyst that contains nitrile hydratase/amidase, were studied. While almost al

Full text of DOI:10.1021/jo0482615

Nitrile Biotransformations for Highly Enantioselective Synthesis
of Oxiranecarboxamides with Tertiary and Quaternary
Stereocenters; Efficient Chemoenzymatic Approaches to
Enantiopure r-Methylated Serine and Isoserine Derivatives
Mei-Xiang Wang,* Gang Deng, De-Xian Wang, and Qi-Yu Zheng
Laboratory for Chemical Biology, Center for Molecular Science, Institute of Chemistry,
Chinese Academy of Sciences, Beijing 100080, People’s Republic of China
mxwang@iccas.ac.cn
Received October 4, 2004
Biotransformations of a number of differently substituted and configured oxiranecarbonitriles using
Rhodococcus sp. AJ270, a microbial whole-cell catalyst that contains nitrile hydratase/amidase,
were studied. While almost all trans-configured 3-aryl-2-methyloxiranecarbonitriles and 2,3-
dimethyl-3-phenyloxiranecarbonitrile were efficiently hydrated by the action of the less enantio-
selective nitrile hydratase, the amidase exhibited excellent 2S,3R-enantioselectivity against
2-methyl-3-(para-substituted-phenyl)oxiranecarboxamides. Under very mild conditions, biotrans-
formations of nitriles provided an efficient and practical synthesis of 2R,3S-(-)-3-aryl-2-methyl-
oxiranecarboxamides, electrophilic epoxides with tertiary and quaternary stereocenters, in excellent
yield with enantiomeric excess greater than 99.5%. The synthetic applications of the resulting
enantiomerically pure epoxides were demonstrated by convenient and straightforward syntheses
of polyfunctionalized chiral molecules possessing a quaternary stereocenter such as R-(+)-2-hydroxy-
2-methyl-3-phenylpropionic acid, 2R,3R-(-)-3-amino-2-hydroxy-2-methyl-3-phenylpropionic acid,
and 2S,3S-(+)-2-amino-3-hydroxy-2-methyl-3-phenylpropionic acid, employing the regio- and
stereospecific epoxide ring opening reactions of 2R,3S-(-)-2-methyl-3-phenyloxiranecarboxamide
as the key steps.
Introduction
and others, no general and single approach stands out
for the synthesis of optically active electrophilic epoxides.4
It is even more challenging to synthesize chiral electro-
Enantiomerically pure electrophilic epoxides with ter-
tiary and quaternary stereocenters are versatile and
powerful intermediates in the synthesis of a wide range
of chiral molecules bearing both tertiary and quaternary
stereocenters in vicinal positions upon the regio- and
stereoselective ring opening reactions of epoxides by
various nucleophiles. The resulting highly functionalized
organic compounds, which are hard to obtain by other
synthetic methods, are not only useful in synthetic
chemistry but are also valuable entities in medicinal
chemistry.1-3 Although the preparation of enantiopure
epoxide compounds has been well developed by Sharpless
(2) Paul, P. K. C.; Sukumar, M.; Bardi, R.; Piazzesi, A. M.; Valle,
G.; Toniolo, C.; Balaram, P. J. Am. Chem. Soc. 1986, 108, 6363 and
references therein.
(3) (a) Ojima, I.; Wang, T.; Delaloge, F. Tetrahedron Lett. 1998, 39,
3663. (b) Denis, J.-N.; Fkyerat, A.; Gimbert, Y.; Coutterez, C.; Man-
tellier, P.; Jost, S.; Greene, A. E. J. Chem. Soc., Perkin Trans. 1 1995,
1811.
(4) For a recent review, see: (a) Porter, M. J.; Skidmore, J. Chem.
Commun. 2000, 1215. For examples of asymmetric epoxidation of R,â-
unsaturated ketones, see: (b) Enders, D.; Zhu, J.; Raabe, G. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1725. (c) Yu, H.-B.; Zheng, X.-F.; Lin,
Z.-M.; Hu, Q.-S.; Huang, W.-S.; Pu, L. J. Org. Chem. 1999, 64, 8149.
(d) Watanabe, S.; Kobayashi, Y.; Arai, T.; Sasai, H.; Bougauchi, M.;
Shibasaki, M. Tetrahedron Lett. 1998, 39, 7353. (e) Watanabe, S.; Arai,
T.; Sasai, H.; Bougauchi, M.; Shibasaki, M. J. Org. Chem. 1998, 63,
8090. (f) Elston, C. L.; Jackson, R. F. W.; MacDonald, S. J. F.; Murray,
P. J. Angew. Chem., Int. Ed. Engl. 1997, 36, 410. (g) Wang, Z.-X.;
Miller, S. M.; Anderson, O. P.; Shi, Y. J. Org. Chem. 1999, 64, 6443
and references therein.
* To whom correspondence should be addressed. Telephone: +86-
10-62565610. Fax: +86-10-62564723.
(1) For a useful review, see: Jung, M. In Chemistry and Biochem-
istry of Amino Acids; Barret, G. C., Ed.; Chapman and Hall: New York,
1985; pp 22-296.
10.1021/jo0482615 CCC: $30.25 © 2005 American Chemical Society
Published on Web 03/02/2005
J. Org. Chem. 2005, 70, 2439-2444
2439

Wang et al.
philic epoxides with tertiary and quaternary stereo-
centers.5-9 Most of the syntheses reported to date are
multistep ones either using a chiral auxiliary5 or starting
from Sharpless asymmetric epoxidation of allylic alcohols
followed by oxidation of the hydroxy group to the car-
bonyl.6 Highly enantioselective Darzens reaction of a
camphor-derived sulfonium amide7 and catalytic asym-
metric epoxidation of R,â-unsaturated amides8 have been
reported very recently to provide highly enantiopure
oxiranecarboxamides, but neither method gave quater-
nary carbon-centered epoxide analogues. In the presence
of benzylquininium chloride, epoxidation of 2-substituted
1,4-naphthoquinones yielded optically active 2,3-epoxides
with enantiomeric excess values less than 45%.9
Biotransformations of nitriles, either through a direct
conversion from a nitrile to a carboxylic acid catalyzed
by a nitrilase10 or through the nitrile hydratase catalyzed
hydration of a nitrile followed by amide hydrolysis
catalyzed by amidase,11 are an effective and environmen-
tally benign method for the production of carboxylic acids
and their amide derivatives.12 Recent studies have dem-
onstrated that biotransformations of nitriles complement
the existing asymmetric chemical and enzymatic methods
for the synthesis of chiral carboxylic acids and their
derivatives.13,14 The distinct features of enzymatic trans-
formations of nitriles are the formation of enantiopure
carboxylic acids and the straightforward generation of
enantiopure amides, which are valuable organonitrogen
compounds in synthetic chemistry. Very recently, we
have shown that Rhodococcus sp. AJ270,15 a whole-cell
catalyst that contains nitrile hydratase/amidase, is able
to efficiently and enantioselectively transform cyclo-
propanecarbonitriles16 and oxiranecarbonitriles17 into the
corresponding carboxylic acids and amides. A prediction
model for reaction efficiency and enantioselectivity has
also been proposed.16f To further explore the synthetic
potential of the nitrile biotransformations catalyzed by
Rhodococcus sp. AJ270 and to validate the prediction
model for the three-membered substrates, we undertook
the current study. In this paper we report an efficient
and convenient synthesis of enantiopure oxiranecarbox-
amides with tertiary and quaternary stereocenters and
their applications in the synthesis of R-methylated R-
hydroxycarboxylic acid and R-methylated serine and
isoserine derivatives.
Results and Discussion
We first examined the reaction of racemic trans-2-
methyl-3-phenyloxiranecarbonitrile 1a. Catalyzed by the
Rhodococcus sp. AJ270 microbial whole-cell catalyst
under very mild conditions, nitrile 1a was very rapidly
and effectively hydrolyzed. For example, more than 50%
of the nitrile 1a was hydrated within 5 min and a
complete hydration was effected in about 30 min (entries
1 and 2 in Table 1). The enantiomeric excess (ee) values
obtained for both the amide 2a and the recovered nitrile
1a were extremely low (5%) after 50% hydration (entry
1 in Table 1), indicating that the nitrile hydratase
involved in this microbial cell catalyst shows very low
enantioselectivity against trans-2-methyl-3-phenyl-
oxiranecarbonitrile. Although the subsequent amide hy-
drolysis was slower than the nitrile hydration, the
amidase involved in Rhodococcus sp. AJ270 cells cata-
lyzed the biohydrolysis of the resulting amide in a few
hours to produce the corresponding enantiomerically pure
2R,3S-2-methyl-3-phenyloxiranecarboxamide (-)-2a in
excellent yield (entry 3 in Table 1) and 2S,3R-2-methyl-
3-phenyloxiranecarboxylic acid 3a, with the latter being
not isolable because it underwent a spontaneous decom-
position similar to that of its 2S,3R-2-phenylglycidic acid
analogue17 to form benzyl methyl ketone under the
reaction conditions (Scheme 1). To shed further light on
the stereochemistry of the reaction, we then investi-
gated the biotransformation of racemic trans-2-methyl-
3-phenyloxiranecarboxamide 2a under the identical con-
ditions. It was found that (()-2a was resolved after 7.5
h into optically active 2R,3S-2-methyl-3-phenyloxirane-
carboxamide (-)-2a in 44% yield with 81% ee. Again, no
(14) For recent examples, see: (a) Wang, M.-X.; Lin, S.-J.; Liu, J.;
Zheng, Q.-Y. Adv. Synth. Catal. 2004, 346, 439. (b) Wang, M.-X.; Lin,
S.-J. J. Org. Chem. 2002, 67, 6542. (c) Wang, M.-X.; Zhao, S.-M.
Tetrahedron Lett. 2002, 43, 6617. (d) Wang, M.-X.; Zhao, S.-M.
Tetrahedron: Asymmetry 2002, 13, 1695. (e) Wang, M.-X.; Li, J.-J.;
Ji, G.-J.; Li, J.-S. J. Mol. Catal. B: Enzymol. 2001, 14, 77. (f) Wang,
M.-X.; Lin, S.-J. Tetrahedron Lett. 2001, 42, 6925. (g) Wang, M.-X.;
Liu, C.-S.; Li, J.-S. Tetrahedron: Asymmetry 2001, 12, 3367. (h) Wang,
M.-X.; Lu, G.; Ji, G.-J.; Huang, Z.-T.; Meth-Cohn, O.; Colby, J.
Tetrahedron: Asymmetry 2000, 11, 1123. (i) Wang, M.-X.; Liu, C.-S.;
Li, J.-S.; Meth-Cohn, O. Tetrahedron Lett. 2000, 41, 8549. (j) DeSantis,
G.; Zhu, Z.; Greenberg, W. A.; Wong, K.; Chaplin, J.; Hanson, S. R.;
Farwell, B.; Nicholson, L. W.; Rand, C. L.; Weiner, D. P.; Robertson,
D. E.; Burk, M. J. J. Am. Chem. Soc. 2002, 124, 9024. (k) Wu, Z.-L.;
Li, Z.-Y. Chem. Commun. 2003, 386. (l) Effenberger, F.; Osswald, S.
Tetrahedron: Asymmetry 2001, 12, 279. (m) Hann, E. C.; Sigmund,
A. E.; Fager, S. K.; Cooling, F. B.; Gavagan, J. E.; Ben-Bassat, A.;
Chauhan, S.; Payne, M. S.; Hennessey, S. M.; DiCosimo, R. Adv. Synth.
Catal. 2003, 345, 775. (n) Wu, Z.-L.; Li, Z.-Y. J. Mol. Catal. B: Enzymol.
2003, 22, 105. (o) Preiml, M.; Hillmayer, K.; Klempier, N. Tetrahedron
Lett. 2003, 44, 5057. (p) Yokoyama, M.; Kashiwagi, M.; Iwasaki, M.;
Fushuku, K.; Ohta, H.; Sugai, T. Tetrahedron: Asymmetry 2004, 15,
2817.
(5) (a) Adam, W.; Pastor, A.; Peters, K.; Peters, E.-M. Org. Lett. 2000,
8, 1019. (b) Ruano, J. L. G.; Castro, A. M. M.; Ramos, J. H. R.;
Flamarique, A. C. R. Tetrahedron: Asymmetry 1997, 8, 3503. (c) Torres-
Valencia, J. M.; Cerda-Garcia-Rojas, C. M.; Joseph-Nathan, P. Tetra-
hedron: Asymmetry 1998, 9, 757. (d) Hayashi, M.; Terashima, S.; Koga,
K. Tetrahedron 1981, 37, 2797.
(6) For representative examples, see: (a) Jung, M. E.; Anderson,
K. L. Tetrahedron Lett. 1997, 38, 2605. (b) Jung, M. E.; D’Amico, D.
C. J. Am. Chem. Soc. 1995, 117, 7379. (c) Marshall, J. A.; Trometer,
J. D.; Blough, B. E.; Crute, T. D. J. Org. Chem. 1988, 53, 4274. (d)
Molander, G. A.; Shubert, D. C. J. Am. Chem. Soc. 1987, 109, 576 and
references therein.
(7) Aggarwal, V. K.; Hynd, G.; Picoul, W.; Vasse, J.-L. J. Am. Chem.
Soc. 2002, 124, 9964.
(8) (a) Nemoto, T.; Kakei, H.; Gnanadesikan, V.; Tosaki, S.-y.;
Ohshima, T.; Shibasaki, M. J. Am. Chem. Soc. 2002, 124, 14544. (b)
Kakei, H.; Nemoto, T.; Ohshima, T.; Shibasaki, M. Angew. Chem., Int.
Ed. 2004, 43, 317.
(9) Pluim, H.; Wynberg, H. J. Org. Chem. 1980, 45, 2498.
(10) Kobayashi, M.; Shimizu, S. FEMS Microbiol. Lett. 1994, 120,
217 and references therein.
(15) Blakey, A. J.; Colby, J.; Williams, E.; O’Reilly, C. FEMS
Microbiol. Lett. 1995, 129, 57.
(16) (a) Wang, M.-X.; Feng, G.-Q. Tetrahedron Lett. 2000, 41, 6501.
(b) Wang, M.-X.; Feng, G.-Q. New J. Chem. 2002, 1575. (c) Wang, M.-
X.; Feng, G.-Q. J. Org. Chem. 2003, 68, 621-624. (d) Wang, M.-X.;
Feng, G. Q. J. Mol. Catal. B: Enzymol. 2002, 18, 267. (e) Wang, M.-
X.; Feng, G.-Q.; Zheng, Q.-Y. Adv. Synth. Catal. 2003, 345, 695. (f)
Wang, M.-X.; Feng, G.-Q.; Zheng, Q.-Y. Tetrahedron: Asymmetry 2004,
15, 347.
(11) (a) Meth-Cohn, O.; Wang, M.-X. J. Chem. Soc., Perkin Trans.
1 1997, 1099. (b) Meth-Cohn, O.; Wang, M.-X. J. Chem. Soc., Perkin
Trans. 1 1997, 3197 and references therein.
(12) Nagasawa, T.; Schimizu, H.; Yamada, H. Appl. Microbiol.
Biotechnol. 1993, 40, 189.
(13) For reviews, see: (a) Sugai, T.; Yamazaki, T.; Yokoyama, M.;
Ohta, H. Biosci. Biotechnol. Biochem. 1997, 61, 1419 and references
therein. (b) Martinkova, L.; Kren, V. Biocatal. Biotrans. 2002, 20, 73.
(17) Wang, M.-X.; Lin, S.-J.; Liu, C.-S.; Zheng, Q.-Y.; Li, J.-S. J. Org.
Chem. 2003, 68, 4570.
2440 J. Org. Chem., Vol. 70, No. 7, 2005

Synthesis of Oxiranecarboxamides
TABLE 1. Biotransformations of Racemic trans-3-Aryl-2-methyloxiranecarbonitriles 1
entry
1
Ar
reaction conditionsa
2 yieldb (%)
2 eec (%)
1
2
1a
1a
1a
1a
1b
1c
1d
1d
1e
1e
1f
C6H5
C6H5
C6H5
C6H5
2 mmol, 5 min
50d
89
45
46e
31
49
32
52
40f
21g
48
31
44
32
45
8
8
2 mmol, 20 min
3
2 mmol, 7.5 h
>99.5
>99.5
99
4
13.1 mmol, 4 dayse
2 mmol, 10 h
5
4-F-C6H4
6
4-Cl-C6H4
2 mmol, 7.5 h
2 mmol, 5.5 days
>99.5
20
7
3-Cl-C6H4
8
3-Cl-C6H4
2 mmol, acetone (2.5 mL), 6 days
41
9
2-Cl-C6H4
2 mmol, 11 hf
<5
10
11
12
13
14
15
2-Cl-C6H4
2 mmol, 7 daysg
<5
4-Br-C6H4
2 mmol, acetone (2.5 mL), 8.5 h
2 mmol, 11.5 h
>99.5
>99.5
<5
1g
1h
1i
4-Me-C6H4
2-Me-C6H4
3,4-OCH2O-C6H3
3,4-OCH2O-C6H3
1 mmol, 7 days
2 mmol, 1 day
2 mmol, acetone (2.5 mL), 1 day
50
32
1i
a Biotransformation was carried out in a suspension of Rhodococcus sp. AJ270 cells (2 g wet weight) in phosphate buffer (50 mL, pH
7.25) at 30 °C. b Isolated yield. c Determined by HPLC analysis using a Chiralcel OD or OJ column (see Supporting Information). d Nitrile
(36%, ee < 5%) was recovered. e Nitrile was added portionwise during 4 days, and a small amount of nitrile (394.3 mg) was recovered.f A
mixture of trans and cis isomers (1:1) was used, and quantitative racemic cis-nitrile (50%) was recovered. g A mixture of trans and cis
isomers (1:1) was used, and almost all racemic cis-nitrile (45%) was recovered.
SCHEME 1. Biotransformations of Racemic
1). In all cases, the nitrile hydratase catalyzed hydration
reaction proceeded very rapidly; all nitriles tested were
found to undergo a complete but low enantioselective
hydration reaction within a few hours. The amide hy-
drolysis, however, was strongly dependent upon the
structure of the substrate. More noticeably, it is the
substitution pattern rather than the nature of the sub-
stituent on the benzene ring of the substrate that plays
a crucial role in determining both the rate and the
enantioselectivity of an amidase-catalyzed reaction, and
therefore the overall reaction. As illustrated in Table 1,
nitrile 1a and all its para-substituted analogues 1b, 1c,
1f, and 1g underwent a rapid hydrolysis to give enan-
tiomerically pure amide products in excellent yields
(entries 3, 5, 6, 11, and 12). With a substituent at the
meta position, substrate 1d took a long incubation time
to effect ca.50% conversion of amide, giving optically
active amide (-)-2d with low enantiomeric excess (entry
7 in Table 1). The hydrolysis of 1i, a substrate derived
from piperonal, proceeded rapidly to afford amide (-)-2i
with 50% ee (entry 14 in Table 1). The slowest reaction
and lowest enantioselection were observed for the reac-
tion of substrates 1e and 1h, which contain an ortho-
substituted benzene ring. In both cases, a long incubation
time such as 7 days and a lower substrate concentration
were required to achieve 50% conversion of the amide
(entries 10 and 13 in Table 1). Unlike the desymmetri-
zation of 3-susbtituted glutaronitriles,14g addition of
acetone (2.5 mL) as an additive or a cosolvent did not
lead to the improvement of enantioselectivity of bio-
transformations of nitriles (entries 8 and 15 in Table 1).
In contrast to the racemic trans-nitrile and amide
isomers 1a and 2a, the biotransformations of racemic cis-
2-methyl-3-phenyloxiranecarbonitrile substrate 4a and
of amide 5a proceeded sluggishly. Under same reaction
conditions as those for 1a, for example, oxiranecarbo-
nitrile (()-4a could not be completely hydrated after 7
days interaction with the whole-cell catalyst, with 66%
of the optically inactive nitrile 4a being recovered. The
subsequent amide hydrolysis was also very slow, and only
a small amount of the resulting amide (<9%) was
trans-3-Aryl-2-methyloxiranecarbonitriles 1
acid product 3a was obtained due to its instability under
the incubation conditions. The use of acetone (2.5 mL)
as a cosolvent to increase the solubility of the amide
substrate 2a in the aqueous buffer gave rise to the
improved conversion rate and enantioselectivity of the
reaction, yielding enantiopure 2R,3S-2-methyl-3-phenyl-
oxiranecarboxamide (-)-2a (Scheme 2). The aforemen-
tioned results indicated clearly that the amidase of
Rhodococcus sp. AJ270 displays high 2S,3R enantio-
selectivity against trans-2-methyl-3-phenyloxiranecar-
boxamide, while the nitrile hydratase shows low enantio-
control against trans-2-methyl-3-phenyloxiranecarbo-
nitrile. The excellent enantioselection of biotransforma-
tion of nitrile 1a originated from the combined effects of
enantioselective nitrile hydratase and amidase, with the
latter being a dominant force (Scheme 1). It is worth
noting that, to prepare enantiomerically pure 2R,3S-2-
methyl-3-phenyloxiranecarboxamide (-)-2a, it is advan-
tageous to employ biotransformation of nitrile rather
than amide. To demonstrate practical application, the
preparative biotransformation was performed under
identical conditions by adding substrate (()-2a (13.1
mmol) portionwise during a period of 4 days, and enantio-
pure (-)-2a was obtained on a gram scale (entry 4 in
Table 1).
To examine the scope of the reaction and the influence
of substituent on the efficiency and enantioselectivity of
biotransformations, a number of racemic trans-3-aryl-2-
methyloxiranecarbonitriles 1 were prepared18 and sub-
jected to incubation with Rhodococcus sp. AJ270 (Scheme
(18) Svoboda, J.; Kocfeldova´, Z.; Palecˇek, J. Collect. Czech. Chem.
Commun. 1988, 53, 822.
J. Org. Chem, Vol. 70, No. 7, 2005 2441

Wang et al.
SCHEME 2. Kinetic Resolution of Racemic trans-2-Methyl-3-phenyloxiranecarboxamide 2a
SCHEME 3. Biotransformations of Racemic
cis-2-Methyl-3-phenyloxiranecarbonitrile 4a
SCHEME 4. Biotransformations of a Mixture of
Racemic trans- and cis-3-Aryloxiranecarbonitriles
SCHEME 5. Biotransformations of
2,3-Dimethyl-3-phenyloxiranecarbonitriles 6 and 7
consumed by the amidase, leaving 25% of amide (-)-5a
in low enantiomeric purity (28% ee) (Scheme 3).
In our previous studies of the biotransformations of
cyclopropanecarbonitriles16 and oxiranecarbonitriles,17 we
have found that the highly efficient and enantioselective
biotransformations of nitriles originate from the com-
bined effects of the enantioselective nitrile hydratase and
amidase, which in turn are strongly dependent upon the
substrate structures. The match of the steric bulkiness
of the substituents on the three-membered ring can lead
to efficient biocatalysis of both trans- and cis-configured
cyclopropanecarbonitriles and amides with remarkable
enantiocontrol. For racemic trans-3-aryloxiranecarbo-
nitrile analogues, the nitrile hydratase of Rhodococcus
sp. AJ270 is very active against both enantiomers of
racemic trans isomers but not cis-isomeric analogues. On
the other hand, the amidase displays high activity with
varied enantioselectivity against trans-3-aryloxirane-
carboxamides, depending on the substitution pattern of
a substituent on the benzene ring. Our current study
revealed that the biotransformations of racemic 3-aryl-
2-methyloxiranecarbonitriles and amides followed the
same general reaction pattern as that of 3-aryloxirane-
carbonitriles and amides. The introduction of one more
methyl group into the oxirane ring, however, resulted in
a more salient steric effect on the reaction efficiency and
enantioselectivity. For example, biocatalytic hydrolysis
of all of the methylated 3-aryloxiranecarbonitriles 1 and
amides 2 proceeded slower than their nonmethylated
analogues, 3-aryloxiranecarbonitriles and amides.17 More
noticeably, in the case of 3-aryl-2-methyloxiranecarbox-
amides, the amidase exhibited much more pronounced
enantio-discriminating power over the substitution pat-
tern of the substituent on the benzene ring of the a
substrate. This was exemplified by the observation of
enantioselectivity ranging from excellent to low and to
eventually naught with a substituent on the benzene ring
of the substrate 2 moving from para to meta and to ortho
position. The outcomes of our study were in agreement
with the conclusions that the nitrile hydratase in Rhodo-
coccus sp. AJ270 is an efficient and less enantioselective
biocatalyst for many nitriles except for sterically heavily
hindered or folded ones, while the amidase is more size-
limited and is more sensitive toward the steric effect of
the substrates even when a substituent is remote from
the amido functional group.16,17
To take advantage of a huge difference in hydration
reaction rate between trans- and cis-nitrile isomers, a
mixture of trans- and cis-2-methyl-3-phenyloxirane-
carbonitriles (()-1a and (()-4a, obtained directly from
the Darzens reaction, was subjected to biocatalysis, with
the hope of achieving asymmetric synthesis of enantio-
pure oxiranecarboxamide and the easy separation of pure
cis-nitrile in one operation. Gratifyingly, biotransforma-
tion did produce optically active amide (-)-2a in both
excellent chemical yield (92% based on theoretical yield)
and enantiomeric purity (ee > 99.5%) after quenching of
the reaction in 5 h. Pure cis-nitrile 4a was also readily
isolated in high yield (96% based on theoretical yield)
(Scheme 4). The successful reaction allowed us to prepare
enantiomerically pure 2R,3S-amides (-)-2 while avoiding
somewhat the tedious chromatographic separation18 of
trans and cis isomers of nitriles prior to biotransforma-
tions.
Biotransformations of racemic 3-aryl-3-methyloxirane-
carbonitriles18 were attempted. The substrates were
found to be unstable, however, undergoing simultaneous
decomposition under the reaction conditions. We then
studied the biocatalytic hydrolysis of 2,3-dimethyl-3-
phenyloxiranecarbonitriles 6 and 7 (Scheme 5). While
trans-2,3-dimethyl-3-phenyloxiranecarbonitrile 6 under-
went a rapid hydration reaction, the biocatalytic hydra-
tion of cis-2,3-dimethyl-3-phenyloxiranecarbonitrile 7
appeared to be extremely slow. The enantiomeric excess
values for both nitriles recovered and amides formed were
very disappointing, reflecting again the low enantio-
selectivity of the nitrile hydratase against oxiranecarbo-
nitriles (Table 2). Hydrolysis of both amides 8 and 9 were
found to be inefficient, most probably due to the increased
steric hindrance of the substrates.
Optically active 3-aryl-2-methyloxiranecarboxamides
(-)-2 are valuable intermediates for the synthesis of a
2442 J. Org. Chem., Vol. 70, No. 7, 2005

Synthesis of Oxiranecarboxamides
TABLE 2. Biotransformations of 2,3-Dimethyl-3-phenyloxiranecarbonitriles 6 and 7
6 or 7
6 or 7
8 or 9
8 or 9
entry
substrate
reaction conditionsa
yieldb (%)
eec (%)
yieldb (%)
eec (%)
1
2
3
4
5
6
7
6 (trans)
6 (trans)
6 (trans)
7 (cis)
2 mmol, 8 min
2 mmol, 7 days
0.2 mmol, 7 days
2 mmol, 7 days
1 mmol, 7 days
0.5 mmol, 5 days
0.2 mmol, 7 days
6 (48)
6 (14)
8 (44)
8 (92)
8 (64)
9 (15)
9 (35)
9 (63)
9 (88)
8 (9)
8 (<5)
8 (35)
9 (19)
9 (17)
9 (17)
9 (<5)
7 (71)
7 (65)
7 (35)
7 (5)
7 (11)
7 (35)
7 (cis)
7 (cis)
7 (cis)
a Biotransformation was carried out in a suspension of Rhodococcus sp. AJ270 cells (2 g wet weight) in phosphate buffer (50 mL, pH
7.25) at 30 °C. b Isolated yield. c Determined by HPLC analysis using a Chiralcel OJ column (see Supporting Information). Absolute
configurations of the products were not determined.
SCHEME 6. Synthesis of
attempt was made to open the epoxide ring of 2a at the
2-position using a number of nitrogen nucleophiles under
different conditions.22 Unfortunately, no satisfactory
results were obtained. In almost all cases, the ring
opening reaction preferentially occurred at the 3-position
(data not shown). Although the strong tendency of the
ring opening reaction at the 3-position of 2R,3S-2-methyl-
3-phenyloxiranecarboxamide 2a requires detailed inves-
tigation, it might be most probably due to the electronic
effect of the phenyl and amido substituents that lead the
3-carbon to be more electron positive than the 2-carbon
and therefore more susceptible to nucleophilic attack. The
steric hindrance of the quaternary carbon might pose
further inhibition of substitution reaction at the 2-carbon.
To synthesize R-methylated serine derivative, an alterna-
tive approach was tried. It was reported by Zwanenburg
and co-workers23 that diethyl 3-azido-2-hydroxysuccinate
can undergo intramolecular aziridination reaction upon
the treatment of Ph3P in DMF to afford diethyl aziridine-
2,3-dicarboxylate, while Davis and Zhou24 showed an
efficient hydrolytic ring opening reaction of 3-arylaziri-
dine-2-methanol using p-toluenesulfonic acid (TsOH)
under very mild conditions. Thus, on the treatment of
PPh3 followed by heating, azide 12 was transformed into
aziridine 2S,3R-15 in 65% yield. Following Davis and
Zhou’s procedure,24 however, no effective hydrolytic ring
opening reaction of 13 was observed. Only at an elevated
temperature (100 °C) did the reaction proceed. Subse-
quent hydrolysis of the resulting amide 16 under acidic
conditions furnished enantiopure 2S,3S-(+)-2-amino-3-
hydroxy-2-methyl-3-phenylpropionic acid 1725 in good
yield (Scheme 7). Without isolation of 15, the overall yield
of 17 was improved to 79.5%.
2R-2-Hydroxy-2-methyl-3-phenylpropionic Acid 11
and Amide 10
range of diversified chiral compounds. To demonstrate
its versatility in organic synthesis, and also to determine
the absolute configuration of biocatalytic products (-)-
2, we first converted 2R,3S-2-methyl-3-phenyloxirane-
carboxamide (-)-2a into R-(+)-2-hydroxy-2-methyl-3-
phenylpropionic acid, a useful building block in natural
products synthesis. Thus, regiospecific hydrogenation of
2a catalyzed by Pd/C in the presence of molecular sieves
(4 Å) gave R-(+)-2-hydroxy-2-methyl-3-phenylpropion-
amide 10 in an almost quantitative yield. Chemical
hydrolysis of amide 10 in refluxing hydrochloric acid (6
N) furnished R-(+)-2-hydroxy-2-methyl-3-phenylpropi-
onic acid 11 in 92% yield (Scheme 6). The optical rotation
of 11 is identical to that of an authentic R-(+)-2-hydroxy-
2-methyl-3-phenylpropionic acid sample,19 suggesting
that the configuration of 11 is R, and therefore the
biotransformation product (-)-2a is 2R,3S configured.
Optically active 2-hydroxy-2-methyl-3-phenylpropionic
acid had been obtained from either a tedious optical
resolution20 or lengthy multistep syntheses using chiral
auxiliaries.19,21
Ring opening reaction of 2R,3S-2-methyl-3-phenyl-
oxiranecarboxamide (-)-2a by sodium azide proceeded
very efficiently and diastereospecifically under mild
conditions to afford exclusively 2R,3R-(-)-3-azido-2-
hydroxy-2-methyl-3-phenylpropionamide 12 in 90% yield.
Catalytic hydrogenation of azide 12 led to the formation
of 2R,3R-(-)-3-amino-2-hydroxy-2-methyl-3-phenylpro-
pionamide 13, which was readily hydrolyzed in hydro-
chloric acid (6 N) to produce 2R,3R-(-)-3-amino-2-hydroxy-
2-methyl-3-phenylpropionic acid 14, an R-methylated
isoserine derivative, in excellent yield (Scheme 7). An
Conclusion
In summary, we have shown that Rhodococcus sp.
AJ270 whole cells can catalyze the hydrolysis of a number
of differently substituted and configured oxiranecarbo-
nitriles under very mild conditions. Both the efficiency
SCHEME 7. Synthesis of r-Methylated Serine and Isoserine Derivatives
J. Org. Chem, Vol. 70, No. 7, 2005 2443

Wang et al.
and enantioselectivity of biocatalysis, however, were
strongly dependent upon the structures of both nitrile
and amide substrates. While almost all trans-configured
3-aryl-2-methyl-oxiranecarbonitriles and trans-2,3-di-
methyl-3-phenyloxiranecarbonitrile were efficiently hy-
drated with the aid of the low enantioselective nitrile
hydratase, the amidase exhibited excellent enantio-
selectivity against 2S,3R-2-methyl-3-(para-substituted
phenyl)oxiranecarboxamides. The biocatalytic reaction of
oxiranecarbonitriles has provided a highly efficient and
practical synthesis of enantiopure 2R,3S-(-)-3-aryl-2-
methyloxiranecarboxamides, electrophilic epoxides with
tertiary and quaternary stereocenters. The resulting
2R,3S-(-)-3-aryl-2-methyloxiranecarboxamides, which are
hard to prepare by other methods, can serve as versatile
chiral synthons in the synthesis of polyfunctionalized
chiral molecules bearing a quaternary carbon. This has
been exemplified by a straightforward synthesis of R-(+)-
2-hydroxy-2-methyl-3-phenylpropionic acid through the
regio- and stereospecific hydrogenation of the epoxide
ring of 2R,3S-(-)-2-methyl-3-phenyloxiranecarboxamide
followed by hydrolysis of amide. We have also demon-
strated that 2R,3S-(-)-2-methyl-3-phenyloxiranecarbox-
amide is able to undergo regio- and diastereospecific ring
opening reaction with sodium azide to afford exclu-
sively 2R,3R-(-)-3-azido-2-hydroxy-2-methyl-3-phenyl-
propionamide, an powerful intermediate that has been
further transformed readily into 2R,3R-(-)-3-amino-2-
hydroxy-2-methyl-3-phenylpropionic acid and 2S,3S-(+)-
2-amino-3-hydroxy-2-methyl-3-phenylpropionic acid, the
enantiomerically pure R-methylated isoserine and serine
derivatives, respectively.
Experimental Section
General Procedure for the Biotransformations of
Nitriles or Amides. To an Erlenmeyer flask (150 mL) with
a screw cap was added Rhodococcus sp. AJ270 cells15 (2 g wet
weight) and potassium phosphate buffer (0.1 M, pH 7.25, 50
mL), and the resting cells were activated at 30 °C for 0.5 h
with orbital shaking. Racemic nitrile 1, 4, 6, or 7 or amide 2a
as fine powder was added in one portion to the flask and the
mixture was incubated at 30 °C using an orbital shaker (200
rpm). The reaction, monitored by TLC and HPLC, was
quenched after a specified period of time (see Tables 1 and 2
and text) by removing the biomass through a Celite pad
filtration. The resulting aqueous solution was extracted with
ethyl acetate (60 mL × 3). After drying (MgSO4) and removing
solvent under vacuum, the residue was chromatographed on
a silica gel column using a mixture of petroleum ether and
ethyl acetate (from 2:1 to 1:5) as the mobile phase to give pure
product. All products were characterized by their spectra data
and comparison of the melting points and optical rotary power
with that of the known compounds, or by full characterization
(see Supporting Information). Enantiomeric excess values were
obtained from HPLC analysis (see Supporting Information).
(19) Jew, S.-s.; Terashima, S.; Koga, K. Tetrahedron 1979, 35, 2337
and references therein.
Acknowledgment. We thank the State Major
Fundamental Research Program (2003CB716005,
G2000077506), the Ministry of Science and Technology
(2002CCA03100), the National Natural Science Foun-
dation of China, and the Chinese Academy of Sciences
for financial support.
(20) Davies, A. G.; Ebeid, F. M.; Kenyon, J. J. Chem. Soc. 1957, 3154.
(21) Frye, S. V.; Eliel, E. L. Tetrahedron Lett. 1985, 26, 3907.
(22) For various examples of ring opening reactions of electrophilic
epoxides by nitrogen nucleophiles, see: (a) Chong, J. M.; Sharpless,
K. B. J. Org. Chem. 1985, 50, 1560. (b) Behrens, C. H.; Sharpless, K.
B. J. Org. Chem. 1985, 50, 5696. (c) Saito, S.; Takahashi, N.; Ishikawa,
T.; Moriwake, T. Tetrahedron Lett. 1991, 32, 667. (d) Cacciola, J.;
Alexander, R. S.; Fevig, J. M.; Stouten, P. F. W. Tetrahedron Lett. 1997,
38, 5741. (e) Fringuelli, F.; Pizzo, F.; Rucci, M.; Vaccaro, L. J. Org.
Chem. 2003, 68, 7041 and references therein.
Supporting Information Available: General experimen-
tal. Spectroscopic data of all compounds prepared. 1H and 13
C
(23) Legters, J.; Thijs, L.; Zwanenburg, B. Tetrahedron 1991, 47,
5287.
(24) Davis, F.; Zhou, P. Tetrahedron Lett. 1994, 35, 7525.
(25) Avenoza, A.; Cativiela, C.; Corzana, F.; Peregrina, J. M.;
Zurbano, M. M. Tetrahedron: Asymmetry 2000, 11, 2195.
(26) Jew, S.-s.; Kim, H.-a.; Kim, J.-h. Heterocycles, 1997, 46, 65.
NMR spectra of 2, 1a, 4a, 4e, 5a, 6-15, and 17. HPLC analysis
of all chiral products. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO0482615
2444 J. Org. Chem., Vol. 70, No. 7, 2005