Enantioselectivity of a recombinant epoxide hydrolase from Agrobacterium radiobacter

Article abstract of DOI:10.1016/S0957-4166(98)00003-2

The recombinant epoxide hydrolase from Agrobacterium radiobacter AD1 was used to obtain enantiomerically pure epoxides by means of a kinetic resolution. Epoxides such as styrene oxide and various derivatives thereof and phenyl glycidyl ether were obtained in high enantiomeric excess and in reasonable yield. The enantioselectivity (E-value) of the resolution was calculated from progress curves for styrene oxide (E= 16.2) and para- chlorostyrene oxide (E=32.2).

Full text of DOI:10.1016/S0957-4166(98)00003-2

Pergamon
Tetrahedron: Asymmetry 9 (1998) 459–466
Enantioselectivity of a recombinant epoxide hydrolase from
Agrobacterium radiobacter
a
a
b
a,∗
Jeffrey H. Lutje Spelberg, Rick Rink, Richard M. Kellogg and Dick B. Janssen
a
Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen,
Nijenborgh 4, 9747 AG, Groningen, The Netherlands
Department of Organic and Molecular Inorganic Chemistry, University of Groningen, Nijenborgh 4, 9747 AG, Groningen,
The Netherlands
b
Received 18 November 1997; accepted 29 December 1997
Abstract
The recombinant epoxide hydrolase from Agrobacterium radiobacter AD1 was used to obtain enantiomerically
pure epoxides by means of a kinetic resolution. Epoxides such as styrene oxide and various derivatives thereof and
phenyl glycidyl ether were obtained in high enantiomeric excess and in reasonable yield. The enantioselectivity
(
E-value) of the resolution was calculated from progress curves for styrene oxide (E=16.2) and para-chlorostyrene
oxide (E=32.2). © 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction
Enantiomerically pure epoxides are important building blocks for the production of a wide range
of pharmaceutical products. Especially during the past decade, much effort has been devoted to the
development of biocatalytic methods for the production of these compounds.1
Epoxide hydrolases are cofactor-independent enzymes that hydrolyze epoxides to diols. If these
enzymes display enantioselectivity, they can be used as chiral catalysts for the production of enantiomer-
2
ically pure epoxides or diols. The enzymes from mammalian sources have been studied extensively
3
because of their involvement in the metabolism of toxic xenobiotics. Styrene oxide, for example, is
hydrolyzed to the less toxic product phenylethanediol. The potential of mammalian epoxide hydrolases
as chiral catalysts on a preparative scale is limited due to the low availability of these enzymes. Several
4
epoxide hydrolases from microbial sources have been discovered recently. High enantioselectivity
5
was obtained with the hydrolysis of (substituted) styrene oxides using fungal cells, and with a broad
6
variety of aryl and aliphatic epoxides with the yeast Rhodotorula glutinis. During the last few years
∗
Corresponding author. Tel.: 31-50-363-4208; fax: 31-50-363-4165; e-mail: d.b.janssen@chem.rug.nl
0957-4166/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00003-2

460
J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
Fig. 1. Epoxides used as substrates by the epoxide hydrolase from A. radiobacter AD1
7
8
enantioselective bacterial epoxide hydrolases have been detected in Rhodococcus sp., Nocardia sp.,
9
10
Corynebacterium sp. and some other genera. The availability of the bacterial enzymes is considerably
higher compared to that of microsomal epoxide hydrolase. However, unlike for the mammalian enzymes,
very little biochemical information is available for the above-mentioned epoxide hydrolases from micro-
bial sources. The genes have not been cloned and their structure and mechanism is unknown.
Recently we have characterized a bacterial epoxide hydrolase from Agrobacterium radiobacter AD1.1
The bacterium was initially isolated for environmental reasons because of its ability to degrade epichloro-
hydrin. Such xenobiotic-degrading enzymes must have high turnover numbers, since they are involved
in the assimilation of a primary substrate for growth. The epoxide hydrolase gene was cloned and
brought to expression in E. coli and the mechanism was characterized.12 The epoxide ring is opened by
a nucleophilic attack of an aspartic acid residue yielding a covalent ester intermediate. This intermediate
is subsequently hydrolyzed and the diol is released. This mechanism is similar to that which has been
1
13
proposed for mammalian epoxide hydrolase. In a recombinant E. coli strain containing the epoxide
hydrolase gene, up to 40% of the total cellular protein content consisted of the epoxide hydrolase,
allowing the isolation of up to 200 mg of pure enzyme from a one liter culture.
In this paper we describe the biocatalytic resolution of a number of chiral epoxides by the epoxide
hydrolase from A. radiobacter AD1. Since this recombinant epoxide hydrolase showed good enantiose-
lectivity towards several substrates and can be produced on a multi-gram scale, it has the potential to be
used as an industrial biocatalyst for the production of enantiomerically pure epoxides.
2. Results and discussion
The epoxide hydrolase gene from A. radiobacter AD1 was cloned and brought to overexpression
12
in E. coli as described previously. The recombinant enzyme was produced by fermentation and
lyophilized partially purified enzyme was prepared. The epoxides 1–7 (Fig. 1) were subjected to
enzymatic hydrolysis by the epoxide hydrolase. The results are shown in Table 1. With all the kinetic
resolutions an enantiomeric excess (ee) of the remaining epoxide of higher then 99% was reached.
The yields of the remaining enantiomerically pure epoxides varied between 27% and 36%. A factor
contributing to the lower yield was some instability of the epoxides in the reaction medium, which is
caused by chemical hydrolysis yielding the racemic diol.
The remaining enantiomerically pure epoxides 1–6 were all of the (S)-configuration, indicating that

J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
461
Table 1
Hydrolysis of epoxides 1–7 by the cloned epoxide hydrolase from A. radiobacter AD1
a
Determined by an additional injection of enantiomerically pure epoxide.
b
c
d
e
24
Determined by specific rotation.
Deduced from a previously described elution order on chiral GLC.
25
21
Determined by reduction of the corresponding diol to 1-phenyl-1,2-ethanediol.
In buffer containing 10% DMSO.
the substituents do not influence the positioning of the substrate in the active site of the enzyme. In the
case of epoxide 3 the absolute configuration was determined by comparison with a previously reported
14
optical rotation, although a contradictory optical rotation was also reported in literature. The assignment
coincided with the fact that with the kinetic resolutions of all the styrene oxide derivatives the remaining
enantiomer was of the (S)-configuration. The absolute configuration of epoxide 4 was determined by
the product of the kinetic resolution, para-chloro-1-phenyl-1,2-ethanediol. Direct determination of the
epoxide was not possible because no literature value was found. Therefore, the para-chloro-1-phenyl-
1,2-ethanediol was reduced to 1-phenyl-1,2-ethanediol which was analysed by chiral HPLC using a
Chiralcell OB column. The phenyl glycidyl ether 7 yielded the (R)-enantiomer which has the same
relative configuration at the stereogenic carbon atom as the styrene oxide derivatives. The opposite
stereochemical designation is due to a change in priority.
The enantioselective hydrolysis of racemic styrene oxide was followed with time (Fig. 2). After
the start of the reaction, the (R)-enantiomer of styrene oxide was hydrolyzed to (R)-phenylethanediol,
indicating that the attack was preferentially at the terminal carbon atom. During the first 40% of the total
conversion, the ee of the (R)-1-phenyl-1,2-ethanediol was 73%. When the (R)-enantiomer was completely
hydrolysed (ee>99%) the ee of the diol had decreased to 49%. Similar results were obtained with the
hydrolysis of epoxide 7. Because of the moderate enantioselectivity, the ee of the remaining (R)-diol was
38% at the timepoint at which the ee of the (S)-enantiomer of epoxide 7 was higher then 99%. These
low enantiomeric purities make the application of this epoxide hydrolase less interesting for preparing
enantiomerically pure diols.
After conversion of the (R)-enantiomer, the (S)-enantiomer of the styrene oxide started to be hy-
drolyzed too, and at a much faster rate than the (R)-enantiomer. A possible explanation for this kinetic
behavior is that the (R)-enantiomer has a higher affinity for the active site, thereby inhibiting the
binding of the (S)-enantiomer. Once 90% of the (R)-enantiomer has been hydrolyzed, the (S)-enantiomer

462
J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
Fig. 2. Kinetic resolution of A; styrene oxide ((ꢀ), (R)-enantiomer; (ꢁ), (S)-enantiomer) and B; para-chlorostyrene oxide ((ꢀ),
R)-enantiomer, (ꢁ), (S)-enantiomer)
(
reacts with a higher rate due to faster chemical reaction steps at the enzyme active site. A similar
1
5
16
kinetic behavior has been observed previously with styrene oxide, para-nitrostyrene oxide and tert-
17
butyloxirane by microsomal epoxide hydrolases. To investigate these results further, the conversion of
both enantiomers was studied separately. The specific activity of the (R)-enantiomer of styrene oxide
−
1
−1
−1
−1
was 1.82 µmol min mg and for the (S)-enantiomer 5.14 µmol min mg . Both activities were
measured for 25% pure enzyme. Initial experiments to determine the K value failed because the value
m
was below the detection limit of 50 µM for both enantiomers.
Epoxides 2–7 showed a more classical kinetic resolution. As an example, the enantioselective hy-
drolysis of epoxide 6 is shown in Fig. 2b. During the hydrolysis of the faster reacting (R)-enantiomer,
the hydrolysis of the (S)-enantiomer was inhibited. After more then 90% of the (R)-enantiomer was
hydrolyzed, the (S)-enantiomer was hydrolyzed at a lower rate. This is in contrast to the hydrolysis rate

J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
463
of the (S)-enantiomer of epoxide 1. The same kinetic resolution pattern was observed for the epoxides
2–5 and 7.
The enantioselectivity of the reaction can be described by using the enantiomeric ratio18 which is
defined by Eq. 1, where Vmax and K represent the Michaelis–Menten parameters.
m
S
S
)/(K )
m
E = (
V
(
(1)
max
R
R
V
)/K )
max
m
The conversion of both enantiomers with time can be described by competitive Michaelis–Menten
19,20
kinetics,
using Eqs 2 and 3 which also account for chemical hydrolysis.
S
dS
dt
V
R
K
S
max
=
−
ꢀ
ꢁ
− kcS
m
(2)
(3)
S
S +
R +
R
+ 1 K
m
R
dR
dt
V
R
max
=
−
ꢀ
ꢁ
− k R
c
S
+ 1 KR
S
m
K
m
In these equations R represents the concentration of the (R)-enantiomer, S the concentration of the (S)-
enantiomer and k the first order chemical hydrolysis rate constant. The constant k was determined under
c
c
the same conditions as the enzymatic conversion, but in the absence of the enzyme. The k for epoxide
c
−6
−1
−6 −1
1
was found to be 8.63×10
s
and for epoxide 6 6.2×10 s . To estimate the kinetic parameters,
Eqs 2 and 3 were fitted by numerical integration to the data shown in Fig. 2, and a good fit was obtained.
With the obtained kinetic parameters, the E-value was calculated using Eq. 1. For epoxide 1 a value of
16.2 and for epoxide 6 a value of 32.2 was calculated.
This method could not be used to obtain unique solutions for all of the kinetic parameters of epoxides 1
and 6. For epoxide 6 a unique solution could only be calculated for the Vmax of the (R)-enantiomer. In the
case of styrene oxide unique solutions were obtained for the values of Vmax and for the ratio between the
K of the (S)-enantiomer and that of the (R)-enantiomer (57.9). The resulting E-value was independent
m
of the K value for each enantiomer. The estimated K of the (S)-enantiomer is lower than 50 µM and
m
m
that of the (R)-enantiomer lower than 1 µM, but to obtain the exact data a more sensitive analysis method
is needed. These results show that the progress of the kinetic resolution of styrene oxide, with a fast
conversion of the remaining (S)-enantiomer after the depletion of the (R)-enantiomer, can be explained
by the higher K and V
of the (S)-enantiomer compared to that of the (R)-enantiomer.
max
m
According to the method described by Chen et al. (Eq. 4), the enantiomeric ratio is derived from the
degree of conversion (c) and the enantiomeric excess of the remaining enantiomer of the substrate (ees).
Because this method is time independent it does not take into account the influence of the chemical
hydrolysis.
ln[(1 − c)(1 − ee )]
s
E =
(4)
ln[(1 − c)(1 + ee )]
s
Using the data points from each curve (Fig. 2), the enantiomeric ratios determined with Eq. 4 were
13.8 for epoxide 1 and 28.2 for epoxide 6, which is significantly lower than when a correction was made
for chemical hydrolysis.
The above results demonstrate that the cloned epoxide hydrolase from A. radiobacter AD1, which can
be produced in large amounts, is enantioselective for a number of epoxides. This makes the production
of enantiomerically pure epoxides on an industrial scale feasible. Since the mechanism of this epoxide
hydrolase has been elucidated and the active site is known, it becomes possible to construct mutants of

464
J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
the enzyme with altered substrate specificity and enantioselectivity. Activities towards the creation of
such (improved) enzymes are currently underway in our laboratory.
3. Experimental
3.1. General
The enantiomeric excess (% ee) and the yield of the epoxides was determined with a Hewlett–Packard
5890 gas chromatograph equipped with a FID detector, using a Chiraldex G-TA capillary column (col.
I) or a CP-cyclodextrin-b-2,3,6-M-19 capillary column (col. II), both of 50 m length and 0.25 mm inside
diameter. The absolute configuration of the phenylethanediol was determined by HPLC using a Chiralcell
OB column. Optical rotations were determined on a Perkin–Elmer 241 MC polarimeter at 589 nm.
NMR spectra were recorded in CDCl using TMS as an internal standard. Purification of the remaining
3
epoxides was performed by flash chromatography using silica 60 H.
3.2. Enzyme preparation
For the described enzymatic resolutions a partially purified enzyme was used. Initially the epoxide
hydrolase gene was cloned by means of polymerase chain reaction and brought to expression in E. coli
12
as described previously. Briefly, plasmid DNA was transformed by electroporation to competent E.
coli BL21 (DE3) cells, which were then plated out on LB plates containing ampicillin and incubated
overnight at 30°C. A preculture was started by inoculating 100 ml of LB containing ampicillin with the
transformants from a plate to a starting OD600 of 0.1. The culture was incubated at 30°C until an OD600
of 1–2 was reached, and then diluted in 1 liter of LB medium containing ampicillin. The culture was
incubated overnight at 20°C. The cells were subsequently centrifuged, washed and resuspended. A crude
extract was prepared by ultrasonic disruption and centrifugation of the cells. This was followed by a
purification step with a DEAE–cellulose anion exchange column. Finally, the partially purified enzyme
was dialyzed against a Tris-SO buffer, lyophilized and stored at 4°C. Prior to use the lyophilized enzyme
4
preparation was suspended in Tris-SO buffer (pH 9.0, 50 mM). From a 1 liter culture, 300 to 400 mg of
4
lyophilized partially purified enzyme could be prepared with a purity of 25 to 40%.
3.3. Kinetic resolution with epoxide hydrolase
A reaction vessel containing 20 to 100 ml Tris buffer (50 mM, pH 9) was incubated in a shaking
waterbath at 30°C. In the case of the epoxides 2, 4, 5 and 6 the buffer contained 10% DMSO. The
epoxide was added to a final concentration of 5 mM. The reaction was started by the addition of the
enzyme solution to a final concentration of 25 to 125 µg/ml. The reaction was monitored by periodically
taking samples from the reaction mixture. The samples were extracted with diethyl ether containing
mesitylene as an internal standard. Prior to analysis by chiral GLC, the samples were dried on a short
column containing MgSO4.
For the determination of the specifc rotation of the remaining enantiomers, the reaction was terminated
at the point where an ee of >99% was reached. The water phase was extracted twice with hexane. The
combined organic phases were dried (MgSO ) and evaporated. The residue was further purified by flash
4
chromatography using hexane–ether as eluent. In the case of epoxide 4, the water phase was extracted
with hexane, saturated with NaCl and extracted twice with ethyl acetate. The combined organic phases

J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
465
were dried (MgSO ) and evaporated. The obtained para-chloro-1-phenyl-1,2-ethanediol was converted
4
21
to 1-phenyl-1,2-ethanediol, by reduction with sodium metal in ethanol. For this, the para-chloro-1-
phenyl-1,2-ethanediol (50 mg) was dissolved in absolute ethanol (3 ml) and refluxed under argon for
30 min. Sodium (150 mg) was added in 25 mg portions to the stirring solution. After 2.5 hours the
suspension was evaporated to near dryness and aqueous HCl was added. The aqueous phase was extracted
with ethyl acetate, dried (MgSO ) and the 1-phenyl-1,2-ethanediol was found to be the (R)-enantiomer
4
as determined by HPLC using a Chiralcell OB column. According to the enzyme mechanism, by which
the (R)-enantiomer of the epoxide is hydrolysed with retention of configuration, it was concluded that
with the kinetic resolution the remaining epoxide was of the (S)-configuration.
The data from the chiral GLC and the obtained specific rotations are as follows: (1) col. I: t (S)=12.4
r
2
5
24
min and t (R)=14.8 min at 110°C; (2) (S) [α]
+19.5 (c 1.97, CHCl ), col. II: t (R)=52.8 min and t
3 r r
r
D
25
24
(
S)=54.4 min at 90°C; (3) [α]
−7.8 (c 3.84, CHCl ), col. I: t (S)=27.9 min and t (R)=30.3 min at
D
3
r
r
25
90°C; (4) [α]
+32.2 (c 1.19, CHCl ), col. I: t (R)=15.6 min and t (S)=20.6 min at 120°C; (5) col. I:
D 3 r r
t (S)=20.2 min and t (R)=31.1 min at 120°C; (6) col. I: t (S)=40.7 min and t (R)=43.0 min at 100°C;
r
r
r
r
(
7) col. I: t (S)=83.3 min and t (R)=84.5 min at 95°C.
r r
The absolute configuration of the enantiomers of phenylethanediol was determined by HPLC by
coinjection of both enantiomers on a Chiralcell OB column using hexane–2-propanol as the eluent (ratio
:1, 0.5 ml/min, retention times 14.7 min and 17.9 min). The formation of 3-phenoxy-1,2-propanediol
9
during the hydrolysis of epoxide 7 was monitored by HPLC on a Chiralcell OD column using hexane–2-
propanol as the eluent (ratio 9:1, 1.0 ml/min, t (R)=17.2 min and t (S)=35.9 min).
r
r
3.4. Synthesis of epoxides
The epoxides 1 and 7 were bought from Aldrich and epoxide 5 was a gift from DSM. The epoxides
, 3, 4 and 6 were synthesized from their corresponding aldehydes and ketones with trimethylsulfonium
methylsulfate. The synthesis of para-chlorostyrene oxide is a typical procedure for the epoxidation
2
22
of aldehydes and ketones. To a solution of para-chlorobenzaldehyde (2.8 gram, 19.9 mmol) in CH Cl2
2
(20 ml) trimethylsulfonium methylsulphate was added (4.4 gram, 23.3 mmol). Aqueous NaOH (50%, 10
ml) was added and the reaction mixture was stirred overnight. Water was added and the organic phase
was separated. The water phase was extracted twice with CH Cl . The combined organic phase was
2
2
washed twice with 20 ml portions of water, subsequently shaken for 20 minutes with a saturated solution
of sodium metabisulphite and finally washed twice with 20 ml portions of water. The organic phase was
dried and the CH Cl was removed on a rotary evaporator. The para-chlorostyrene oxide was obtained by
2
2
bulb-to-bulb distillation or by flash chromatography on silica 60 H. In the case of epoxide 3 the reaction
was performed at the same conditions with the exception that the reaction mixture was heated to reflux.
1
13
2
: H NMR δ: 2.46 (s, 3H), 2.90 (dd, 1H), 3.25 (dd, 1H), 7.3 (m, 4H); C NMR δ: 20.9 (CH ), 50.9
3
(
C-2), 52.2 (C-1), 125.4, 128.9, 134.4, 137.8 (C-Ar).
1
13
3
: H NMR δ: 1.62 (s, 3H), 2.65 (d, 1H), 2.85 (d, 1H), 7.2 (m, 5H); C NMR δ: 21.6 (CH ), 56.5
3
(
C-1), 56.8 (C-2), 125.2, 127.4, 128.2, 141.1 (C-Ar).
1
13
4
: H NMR δ: 2.63 (dd, 1H), 3.19 (dd, 1H), 4.21 (dd, 1H), 7.20–7.38 (m, 4H); C NMR δ: 49.8 (C-2),
5
5
0.5 (C-1), 125.5, 126.9, 128.8, 135.4 (C-Ar).
1
13
6
: H NMR δ: 2.88 (dd, 1H), 3.27 (dd, 1H), 3.96 (dd, 1H), 7.32–7.46 (m, 4H); C NMR δ: 51.0 (C-2),
1.5 (C-1), 126.7, 128.6, 133.8, 136.1 (C-Ar).
The reference compounds (S)-1 and (R)-5 were bought from Aldrich and (R)-7 was synthesized from
(
S)-glycidol and phenol, according to a literature procedure.23

466
J. H. Lutje Spelberg et al. / Tetrahedron: Asymmetry 9 (1998) 459–466
Acknowledgements
This research was financially supported by the Innovation Oriented Research Program (IOP) on
Catalysis (no. 94007a) of the Dutch Ministry of Economic Affairs.
References
1
2
3
4
5
6
7
8
9
. de Bont, J. A. M. Tetrahedron: Asymmetry 1993, 4, 1331.
. Archer, I. V. J. Tetrahedron 1997, 53, 15617.
. Oesch, F. Xenobiotica 1973, 3, 305.
. Faber, K.; Mischitz, M.; Kroutil, W. Acta Chem. Scand. 1996, 50, 249.
. Pedragosa-Moreau, S.; Morisseau, C.; Zylber, J.; Archelas, A.; Baratti, J.; Furstoss, R. J. Org. Chem. 1996, 61, 7402.
. Weijers, C. A. G. M. Tetrahedron: Asymmetry 1997, 8, 639.
. Mischitz, M.; Kroutil, W.; Wandel, U.; Faber, K. Tetrahedron: Asymmetry 1995, 6, 1261.
. Kroutil, W.; Mischitz, M.; Plachota, P.; Faber, K. Tetrahedron Lett. 1996, 37, 8379.
. Archer, I. V. J.; Leak, D.; Widdowson, D. A. Tetrahedron Lett. 1996, 37, 8819.
1
1
1
1
1
0. Osprian, I.; Kroutil, W.; Mischitz, M.; Faber, K. Tetrahedron: Asymmetry 1997, 8, 65.
1. Jacobs, M. H. J.; van den Wijngaard, A. J.; Pentenga, M.; Janssen, D. B. Eur. J. Biochem. 1991, 202, 1217.
2. Rink, R.; Fennema, M.; Smids, M.; Dehmel, U.; Janssen, D. B. J. Biol. Chem. 1997, 272, 14650.
3. Lacourciere, G. M.; Armstrong, R. N. J. Am. Chem. Soc. 1993, 115, 10466.
4. Johnson, C. R.; Kirchhoff, R. A.; Reischer, R. J., Katekar, G. F. J. Am. Chem. Soc. 1973, 95, 4287 and Pedragosa-Moreau,
S.; Archelas, R.; Furstoss, R. Tetrahedron 1996, 52, 4593.
1
1
1
1
1
2
2
2
2
2
5. Watabe, T.; Ozawa, N.; Hiratsuka, A. Biochem. Pharmacol. 1981, 30, 1695.
6. Westkaemper, R. B.; Hanzlik, R. P. Arch. Biochem. Biophys. 1981, 208, 195.
7. Wistuba, D.; Schurig, V. Chirality 1992, 4, 178.
8. Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. Am. Chem. Soc. 1982, 104, 7294.
9. Fersht, A. Enzyme Structure and Mechanism, 2nd edn, W. H. Freeman & Co., New York, 1985.
0. Rakels, J. L. L.; Romein, B.; Straathof, A. J. J.; Heijnen, J. J. Biotech. Bioeng. 1994, 43, 411.
1. Hudlicky, T.; Boros, E. E.; Boros, C. H. Tetrahedron: Asymmetry 1993, 4, 1365.
2. Mosset, P.; Gree, R. Syn. Comm. 1985, 15, 749.
3. Waagen, V.; Hollingsæter, I.; Partali, V.; Thorstad, O.; Anthonsen, T. Tetrahedron: Asymmetry 1993, 4, 2265.
4. For compound 2: Pedragosa-Moreau, S.; Morisseau, C.; Zylber, J.; Archelas, A.; Baratti, J.; Furstoss, R. J. Org. Chem.
1996, 61, 7402. For compound 3: Johnson, C. R.; Kirchhoff, R. A.; Reischer, R. J.; Katekar, G. F. J. Am. Chem. Soc. 1973,
95, 4287.
25. Fruetel, J. A.; Mackman, R. L.; Peterson, J. A.; Ortiz de Montellano, P. R. J. Biol. Chem. 1994, 269, 28815.