Microbiological transformations. 33. Fungal epoxide hydrolases applied to the synthesis of enantiopure para-substituted styrene oxides. A mechanistic approach

Article abstract of DOI:10.1021/jo960558i

The biohydrolysis of differently para-substituted styrene oxide derivatives was studied, using whole cells of the fungi Aspergillus niger or Beauveria sulfurescens. These microorganisms proved to be equipped with epoxide hydrolases which are able to achieve these hydrolyses with high enantioselectivity. This allowed the preparation of the optically active epoxides and of the corresponding vicinal diols which were obtained with good to excellent enantiomeric purity. These two microorganisms proved to be enantiocomplementary. A mechanistic study, carried out using a crude lyophilized enzymatic extract from A. niger, indicated via Hammet coefficient plotting that this hydrolysis is very probably due to a general base-catalyzed process.

Full text of DOI:10.1021/jo960558i

7402
J . Org. Chem. 1996, 61, 7402-7407
Micr obiologica l Tr a n sfor m a tion s. 33. F u n ga l Ep oxid e Hyd r ola ses
Ap p lied to th e Syn th esis of En a n tiop u r e P a r a -Su bstitu ted Styr en e
Oxid es. A Mech a n istic Ap p r oa ch
S. Pedragosa-Moreau, C. Morisseau, J . Zylber, A. Archelas, J . Baratti, and R. Furstoss*
Groupe Biocatalyse et Chimie Fine, ERS 157 associe´e au CNRS, Faculte´ des Sciences de Luminy, Case
901, 163 avenue de Luminy, 13288 Marseille Cedex 9, France
Received March 26, 1996X
The biohydrolysis of differently para-substituted styrene oxide derivatives was studied, using whole
cells of the fungi Aspergillus niger or Beauveria sulfurescens. These microorganisms proved to be
equipped with epoxide hydrolases which are able to achieve these hydrolyses with high enantio-
selectivity. This allowed the preparation of the optically active epoxides and of the corresponding
vicinal diols which were obtained with good to excellent enantiomeric purity. These two
microorganisms proved to be enantiocomplementary. A mechanistic study, carried out using a
crude lyophilized enzymatic extract from A. niger, indicated via Hammet coefficient plotting that
this hydrolysis is very probably due to a general base-catalyzed process.
Sch em e 1
Epoxide hydrolases (EH) are very interesting enzymes
which appear more and more to be ubiquitous in nature.
They have been detected for example in organisms as
diverse as mammals,1 plants,2 and microorganisms.3,4
Because of their involvement in the metabolism of
various xenobiotics, many of which are suspected to be
carcinogenic, mammalian enzymes have been extensively
studied during the past two decades. For example, it has
been described very recently5 that epoxide hydrolases
were identified in 96% of prostate malignant tumors, and
it has been suggested that the expression of these
enzymes is one possible mechanism of anticancer drug
resistance. Besides these important biological studies,
at least two other aspects of these enzymes are of high
interest, i.e., (a) their use as asymmetric biocatalysts for
organic chemistry applications and (b) the determination
of their intimate mechanism of action.
As far as chemistry is concerned, the value of epoxides
and/or of their corresponding vicinal diols as synthetic
intermediates, and the now largely admitted fact that
most biologically active molecules ought to be studiedsor
even manufacturedsin enantiomerically pure form, em-
phasizes the need and desirability of obtaining these
intermediates in a high state of enantiomeric purity.6,7
We have ourselves described recently a new preparative
scale methodology based on the use of newly discovered
fungal epoxide hydrolases from Aspergillus niger or
Beauveria sulfurescens.3
Another very interesting question about these enzymes
is the detailed understanding of their mechanism of
action. It is now well established that these enzymes act
via a trans opening of the oxirane ring and that this
attack occurs very preferentially at the less substituted
carbon atom. Furthermore, the very elegant single
turnover experiments recently conducted by Lacourciere
and Armstrong8 using rat liver microsomal epoxide
hydrolase led to the conclusion that a two-step mecha-
nism does in fact take place. A similar result has been
very recently described by Hammock et al.9 for soluble
(cytosolic) recombinant murine epoxide hydrolase. How-
ever, it seems that these two types of enzymes may be
somewhat different as far as the oxirane ring activation
process is concerned. Indeed, whereas it is now well
established that, for microsomal epoxide hydrolases
(mEH) this mechanism corresponds to a general base
X Abstract published in Advance ACS Abstracts, August 1, 1996.
(1) (a) See ref 9 and references cited therein. (b) Knehr, M.; Thomas,
H.; Arand, M.; Gebel, T.; Zeller, H.-D.; Oesch, F. J . Biol. Chem. 1993,
268, 17623. (c) Belluci, G.; Chiappe, C.; Cordoni, A.; Marioni, F.
Chirality 1994, 6, 207. (d) Belluci, G.; Chiappe, C.; Ingrosso, G.
Chirality 1994, 6, 577. (e) Hammock, B. D.; Grant, D. F.; Stroms, D.
H. Epoxide Hydrolases. In Comprehensive Toxicology; Sipes, I., Mc-
Queen, C., Gandolfi, A., Eds.; Pergamon: Oxford, 1996; Vol. 3, Chapter
18.
(2) (a) Ble´e, E.; Schuber, F. Eur. J . Biochem. 1995, 230, 229. (b)
Stapleton, A.; Beetham, J . K.; Pinot, F.; Garbarino, J . E.; Rockhold,
D.R.; Friedman, M.; Hammock, B. D.; Belknap, W. R. Plant J . 1994,
6, 251.
(3) (a) Chen, X.-J .; Archelas, A.; Furstoss, R. J . Org. Chem. 1993,
58, 5528. (b) Pedragosa-Moreau, S.; Archelas, A.; Furstoss, R. J . Org.
Chem. 1993, 58, 5533.
(4) (a) Mischitz, M.; Kroutil, W.; Wandel, U.; Faber, K. Tetrahe-
dron: Asymmetry 1995, 6, 1261. (b) Zhang, J .; Reddy, J .; Roberge, C.;
Senanayake, C.; Greasham, R.; Chartrain, M. J . Ferment. Bioeng. 1995,
80, 244.
(5) Murray, G. I.; Taylor, V. E.; McKay, J . A.; Weaver, R. J .; Ewen,
S. W. B.; Melvin, W.T.; Burke, D. J . Pathol. 1995, 177, 147.
(6) Mahmoudian, M.; Michael, A. Appl. Microbiol. Biotechnol. 1992,
37, 28.
(7) (a) Besse, P.; Veschambre, H. Tetrahedron 1994, 50, 8885. (b)
Pedragosa-Moreau, S.; Archelas, A.; Furstoss, R. Bull. Chim. Fr. 1995,
132, 769.
(8) Lacourciere, G. M.; Armstrong, R. N. J . Am. Chem. Soc. 1993,
115, 10466.
(9) Borhan, B.; J ones, A. D.; Pinot, F.; Grant, D. F.; Kurth, M. J .;
Hammock, B. D. J . Biol. Chem. 1995, 270, 26923.
S0022-3263(96)00558-0 CCC: $12.00 © 1996 American Chemical Society

Microbiological Transformations
J . Org. Chem., Vol. 61, No. 21, 1996 7403
Ta ble 1. Bioh yd r olysis of P a r a -Su bstitu ted Styr en e
Oxid e Der iva tives Usin g A. n iger
Ta ble 2. Bioh yd r olysis of P a r a -Su bstitu ted Styr en e
Oxid e Der iva tives Usin g B. su lfu r escen s
yield of
yield of
epoxide, %
(reaction
ee of
epoxide,
epoxide, %
(reaction
time, h)
ee of
ee of diol,
yield of % (abs
yield of ee of diol,
epoxide,
substr
time, h) % (abs conf) product diol, % % (abs conf)
substr
% (abs conf)a product diol, %
conf)a
1
2
3
4
5
6
7
28 (2)
34 (1)
35 (0.75)
33 (1)
34 (1)
99 (S)3b
96 (S)a
8
9
50
40
47
49
51
42
54
65 (R)3b
66 (R)d
81 (R)d
79 (R)b
79 (R)g
76 (R)g
70 (R)21
1
2
3
4
5
6
7
34 (2)
30 (0.5)
25 (2)
20 (24)
33 (2)
59 (24)
50 (24)
98 (R)
>98 (R)
96 (R)
54 (R)
96 (R)
15 (R)
20 (S)
8
9
45
45
50
66
52
25
36
83 (R)
76 (R)
78 (R)
72 (R)
72 (R)
50 (R)
49 (R)
98 (S)16
>98 (S)e
>98 (S)f
>98 (S)f
>98 (S)c
10
11
12
13
14
10
11
12
13
14
38 (0.75)
37 (1)
a
a
Groves, J . T.; Myers, R. S. J . Am. Chem. Soc. 1983, 105, 5791.
Hudlicky, T.; Boros, E. E.; Boros, C. H. Tetrahedron: Asymmetry
See Table 1 for the determination of the absolute configura-
b
tions.
1993, 4, 1365. c Pluim, H.; Wynberg, H. J . Org. Chem. 1980, 45,
d
2498. The absolute configuration was deduced by using a chemi-
configuration, can be in most cases easily and efficiently
converted back to the corresponding (R) epoxides via
chemical cyclization, without loss of stereochemical in-
tegrity. Interestingly, some of these products have not
been described previously in optically pure form. As far
as the regioselectivity of the oxirane opening is concerned,
our previous experiments11 conducted on styrene oxide
(1) using 18O-labeled water, have shown a high regiose-
lectivity of the isotope incorporationsand therefore of the
nucleophilic attack, which occurred at the less hindered
carbon atom. The absolute configuration of the diols
obtained from epoxides 2-7 indicates that, in all cases,
the regioselectivity of the oxirane ring opening is similar,
i.e., that the nucleophilic attack again takes place very
preferentially at the methylene carbon atom.
Bioh yd r olysis Usin g B. su lfu r escen s. To the con-
trary of the results observed with A. niger, the data
obtained by studying the biohydrolysis of the para-
substituted styrene oxide derivatives (2-7) using whole
cells of B. sulfurescens show a remarkable effect of the
phenyl ring para-substitution (as compared to the non-
substituted styrene oxide (1)) (Table 2). They indicate
that substitution of the aromatic ring affects the reaction
rate as well as the substrate enantioselectivity to an
extent that seems to depend upon the electronic proper-
ties of the substituents. Thus, the p-CH3 group (epoxide
2), which is an electron-donating group (by hyperconju-
gation), caused a 4-fold increase of the reaction rate (as
compared to that of 1) without modification of the
enantioselectivity. On the other hand, the weakly elec-
tron-withdrawing p-F (3) and p-Br (5) substituents did
not cause any noticeable modification of either reaction
rate or enantioselectivity. Surprisingly enough, substitu-
tion by a p-Cl group (4) led to a dramatic lowering of both
reaction rate and enantioselectivity. At this stage of our
study, we do not have any good explanation for this
surprising fact: one could consider either the presence
of twosor severalsdistinct enzymes or isoenzymes op-
erating in this particular fungus and/or the influence of
some permeation parameters which could be quite dif-
ferent for this particular substrate. Finally, the presence
of the powerful cyano (6) or nitro (7) electron-withdraw-
ing groups decreased significantly both reaction rate and
enantioselectivity. Even more surprisingly, the p-NO2
moiety caused a dramatic switch in the substrate enan-
tioselectivity. Thus, predominant hydrolysis of the (R)
enantiomer was observed in this case.
cal correlation with the corresponding epoxide. e The absolute
configuration was established by using a chemical correlation with
the corresponding diol. f The absolute configuration were deduced
from the elution order on the chiral GLC, the (R) enantiomer being
g
eluted before the (S) antipode for all the epoxides studied. The
absolute configuration was deduced by using a chemical correlation
with the corresponding epoxide and confirmed from the elution
order on chiral GLC of the corresponding acetonide derivative, the
(R) enantiomer being eluted before the (S) enantiomer for all the
diols studied.
catalysis, a push-pull mechanism implying activation of
the epoxide by protonation or hydrogen bonding has been
postulated for cytosolic epoxide hydrolase. Thus, a
general acido-basic-catalyzed process implying activation
of the oxirane ring via either protonation or hydrogen
bonding could be operative in this case. This hypothesis
was essentially supported by the difference of regiose-
lectivity observed for the opening of simple monosubsti-
tuted epoxides when catalyzed by either cEH or mEH.10
In this study, we now report our results aimed, on one
hand, at exploring the possibility of using the two fungi
A. niger and B. sulfurescens as asymmetric catalysts for
organic synthesis and, on the other hand, getting some
more insight into the mechanism of the two epoxide
hydrolases contained in these fungal strains (Scheme 1).
Resu lts a n d Discu ssion
The various para-substituted styrene oxide derivatives
2-7 were submitted to biotransformation by the fungi
A. niger and B. sulfurescens.
Bioh yd r olysis Usin g A. n iger . The results of the
biohydrolysis of styrene oxide (1) and of its derivatives
2-7, using whole cells of the fungus A. niger, are reported
in Table 1. It clearly appears that, for all the epoxides
studied, a rapid hydrolysis showing a high substrate
enantioselectivity occurred. Indeed, after only 1 h of
bioconversion (or even less for some of these substrates),
all the residual epoxides were almost optically pure (ee
> 98%) and all the formed diols also exhibited high
enantiomeric excesses. Furthermore, it could be observed
that the nature of the substituent born by the phenyl ring
did not affect the absolute configuration of the products
obtained. As in the case of styrene oxide (1), all the
unreacted epoxides were of (S) absolute configuration,
whereas the formed diols were all shown to be of (R)-
configuration. We emphasize that all these epoxides and
diols can thus be prepared easily, on a preparative
multigram scale, by using this methodology. Further-
more the formed diols, which are always of (R) absolute
Mech a n istic Stu d ies. As emphasized, another aim
of this work was to gain more information about the
mechanism involved during these oxirane ring biohy-
drolyses. Interestingly, whereas a huge amount of work
has been devoted to the study of mEH, almost no
(10) Bellucci, G.; Chiappe, C.; Cordoni, A;; Marioni, F. Tetrahedron
Lett. 1994, 35, 4219.

7404 J . Org. Chem., Vol. 61, No. 21, 1996
Pedragosa-Moreau et al.
Ta ble 3. Su bstitu en t Effects on th e Kin etic P a r a m eter s
of Hyd r olysis of P a r a -Su bstitu ted Styr en e Oxid e by th e
Ep oxid e Hyd r ola se of A. n iger
substrate
Kma
Vma
103(Vm/Km)
σp
(()-1
(R)-1
(S)-1
(()-2
(()-4
(()-6
(()-7
(R)-7
(S)-7
4.7 ( 0.8
4.3 ( 0.6
9.9 ( 1.6
2.4 ( 0.3
1.1 ( 0.1
1.8 ( 0.2
1.0 ( 0.2
0.8 ( 0.1
4.7 ( 1.9
116 ( 10
221 ( 17
28 ( 2
104 ( 5
128 ( 3
191 ( 10
195 ( 10
238 ( 17
38 ( 10
34
51
3
0
43
-0.13
0.24
0.67
117
106
195
297
8
0.78
a
Km are expressed in mM, Vm in µm min-1 g-1 of protein.
F igu r e 1. Hammet plot of log(Vmx/VmH) for biohydrolysis
of (()-para-substituted styrene oxide vs σp, using soluble
enzyme extract of A. niger as biocatalyst.
information about the mechanism involved in microbial
EH is presently available. Our previous studies,11 con-
ducted on 1, led us to suggest the occurrence of a general
base-catalyzed process (similar to the one proposed for
mEH) in the case of A. niger, whereas for B. sulfurescens
we considered that activation of the oxirane ring via
protonation or hydrogen bonding had to be implied, as
in the case of soluble EH. These hypotheses were
essentially based on the regio- and stereoselectivities
observed. In order to confirm these hypotheses, another
appealing way was to determine the Hammet correlation
parameters of these biohydrolyses, using our para-
substituted aromatic epoxides. Unfortunately, because
of the physical state of fungal whole-cell cultures, the
substrate distribution into the culture medium was
heterogeneous, some substrate being preferentially ad-
sorbed on the cell surface. Thus, it was not possible,
using this methodology, to obtain accurate and reproduc-
ible data to conclude as to the influence of the electron-
donating or -withdrawing power of the para substituents
on the reaction rates (Table 1 or Table 2). However, we
could achieve this study without any experimental
problem using a soluble protein preparation from A.
niger.12
Mech a n ism Im p lied in th e F u n gu s A. n iger . Using
such an enzymatic extract, the biohydrolysis of the
racemic derivatives 1, 2, 4, 6, and 7 was carried out and
their initial hydrolysis rates were measured as a function
of the substrate concentration. The apparent kinetic
parameters (Km and Vm) were calculated via a nonlinear
regression of the Michaelis equation and are reported in
Table 3. Plotting of the relative rates (VmX/VmH)
logarithm versus the Hammet constant σp for each
substituent is given in Figure 1. This shows that a rather
linear relation (correlation coefficient r ) 0.985) for the
five substrates studied exists indeed, and the observed
linear relation suggests that the same mechanism oper-
ates throughout the series. Furthermore, the positive
slope observed can be associated with a developing
negative charge in the transition state during biohy-
drolysis. In the case of an oxirane ring opening mecha-
nism, an electron-withdrawing group located on the
phenyl moiety would be expected to stabilize, in the
transition state, the development of a negative charge
on the incipient oxygen anion at the benzylic carbon. The
small magnitude of F (+0.3), which reflects the impor-
tance of the charge development at the reactive center,
indicates a weak polar character of the “early” transition
state structure. Therefore, this result strongly suggests
that a nucleophilic mechanism was implied as the rate-
determining step for the biohydrolysis of the epoxides
studied. It should be noted that this conclusion is in
agreement with the results previously reported by Blu-
menstein et al.13 for the chemical base-catalyzed hydra-
tion (for the attack occurring at the methylene carbon)
of the same aromatic epoxide derivatives, where a value
of F ) +0.32 was estimated. These results are also in
accordance with a similar study previously carried out
using rat liver mEH where a comparable correlation was
obtained (F ) +0.2) for substrates bearing an electron-
withdrawing group.14
Some further interesting informations can be obtained
from these kinetic measurements. Comparison of the
various Km measurements shows a noticeable drop of
these values when the electron-withdrawing power of the
phenyl substituent increases, indicating that substrates
with high σp show a better affinity for the enzyme.
Furthermore, the specificity constant (Vm/Km) increases
sharply with the σp value, showing a higher selectivity
for these substrates (Table 3). Thus, this indicates that
the observed substrate selectivity is the result of both
higher affinity and higher catalytic rate.
Another valuable information about the mechanism
leading to the observed enantioselectivity can be deduced
from the kinetic parameters. Indeed, the Km and Vm
values were determined separately for each enantiomer
of 1 and 7. For both these substrates, the (R) enantiomer
showed a slightly lower Km (by a factor of 2.3 and 5.9,
respectively) and a significantly higher Vm (7.9 and 6.2
times higher) than their (S) enantiomer. These results
clearly indicate that the enantioselectivity is due to both
a better affinity and a higher reaction rate for the (R)
enantiomer. This may be due to a better positioning of
the (R) enantiomer at the active site and/or to a higher
stabilization of the transition state. The enantiomeric
ratios E, estimated for 1 and 7 from the ratio of the Vm/
Km of their enantiomers, were respectively 17 and 36.
As reported in Table 1, these values are in good agree-
ment with those obtained by the direct determination
method using the conversion ratio and the ee of the
remaining epoxide.12
Mech a n ism Im p lied in th e F u n gu s B. su lfu r e-
scen s. A similar study, aimed to confirm the fact that
the intimate process involved with the fungus B. sul-
furescens is different indeed, would have been obviously
(11) Pedragosa-Moreau, S.; Archelas, A.; Furstoss, R. Bioorg. Med.
Chem. 1994, 2, 609.
(12) Nellaiah, H.; Morisseau, C.; Archelas, A.; Furstoss, R.; Baratti,
J . Biotechnol. Bioeng. 1996, 49, 70.
(13) Blumenstein, J . J .; Ukachukwu, V. C.; Mohan, R. S.; Whalen,
D. L. J . Org. Chem. 1993, 58, 924.
(14) Dansette, P. M.; Makedonska, V. B.; J erina, D. M. Arch.
Biochem. Biophys. 1978, 187, 290.

Microbiological Transformations
J . Org. Chem., Vol. 61, No. 21, 1996 7405
Sch em e 2. P r op osed Mod els for th e Active Sites
of Ep oxid e Hyd r ola ses fr om (a ) A. n iger a n d (b)
B. su lfu r escen s
not only was the fast-reacting enantiomer shown to have
the (R) configuration, but the nucleophilic attack of the
oxirane ring also occurred preferentially at the less
substituted C(2) carbon atom. This may be explained
either (a) by a different positioning of this particular
compound into the active site, due for instance to some
specific bonding of the highly polar nitro group with some
ionized peptide residues, or (b) by considering the exist-
ence of a second (or even several) EH in this fungus which
would exhibit a higher catalytic power for this particular
compound together with a reversed enantioselectivity.
Con clu sion
In conclusion, these results show clearly that the two
fungi A. niger and B. sulfurescens are able to achieve an
enantioselective hydrolysis of several para-substituted
aromatic epoxides. This allows us to obtain both the (S)
and the (R) enantiomers of epoxides 1-6 in a high state
of enantiomeric purity. Interestingly, these microbiologi-
cal transformations can easily be carried out on a
preparative scale, thus allowing production of several
grams of products. Several of these epoxides (3,16 4,17
and 718 ) are known to be valuable chiral building blocks
involved in the synthesis of compounds showing biological
activities. Our results also confirm that the interesting
enantiocomplementarity which we previously did observe
during the biohydrolysis of styrene oxide (1) with these
two fungi was not affected by the presence of a para
substituent. Indeed, the residual epoxide antipodes were
always of (S) absolute configuration with A. niger,
whereas the ones of (R)-configuration were obtained
using B. sulfurescens, except for epoxide 7 where the (S)
enantiomer was less reactive. As far as the enzymatic
mechanism is concerned, we have shown, using a Ham-
met correlation approach, that a general base-catalyzed
process is very probably implied in the case of A. niger.
In contrast, the reaction rates and regioselectivities
observed using whole cells of B. sulfurescens strongly
suggest the implication of a general acido-basic catalyzed
mechanism. Work is in progress in our laboratory to
order to explore the scope and limitations of these
interesting biotransformations as well as to study the
possibilities of building up biotechnological processes
using such fungal enzyme extracts.
very interesting. Unfortunately, we have not yet man-
aged to prepare an active soluble enzyme extract from
this fungus. Nevertheless, the results we have obtained
using whole cells allow us to get some preliminary
information about the mechanism implied. Thus, the
results reported in Table 2 clearly show that the decrease
of reaction rate is concomitant with the increase of the
electron-withdrawing power of the para substituents.
This observation is in accordance with an acid-catalyzed
process in which some carbonium ion character would
be created during the transition site. Another important
observation, which confirms this hypothesis, is based on
the regioselectivity of the oxirane opening. Indeed, one
must notice that all the diols formed are, without
exception, of (R) absolute configuration. Since the fast-
reacting enantiomer is in most cases of (S) configuration,
this implies an attack by the enzyme at the C(1) benzylic
carbon atom, thus leading to inversion of configuration
at this center. This regioselectivity is an accordance with
the one previously observed during chemical acid-
catalyzed opening of the same para-substituted styrene
oxide derivatives.13
Active Site Top ology. In a previous study,15 the
results we obtained using several aromatic epoxides
bearing methyl substitution at the oxirane ring led us to
propose an active site model for both EH involved in the
two fungi employed in this study. For A. niger, we
considered the existence of a lipophilic pocket located at
the right back side of the active site (taking as a
convention that the oxirane ring is oriented with the
oxygen atom toward the top). The present results are
in perfect agreement with this model, the phenyl ring of
all the fast reacting (R)-enantiomers being located in the
appropriate part of the space (Scheme 2). Obviously,
positioning of the (S)-enantiomers into this model will
be less favored. These results also lead to the conclusion
that this pocket is able to accommodate bulky para
substituents like cyano or nitro groups, since all these
epoxides proved to be good substrates showing high
reactivity and enantioselectivity.
In the case of B. sulfurescens, we proposed that the
lipophilic pocket would be located at the right front side,
the nucleophilic entity of the enzyme (presumably an
aspartic residue) being situated beneath the benzylic
carbon atom C(1). This model is supported by the fact
that the fast-reacting enantiomers of the epoxides 1-6
are those of (S) absolute configuration. Again it appears
that this pocket is able to accommodate the bulky para
substituents born by the aromatic ring. As an exception,
epoxide 7 (p-NO2 group) does not fit into this model, since
Exp er im en ta l Section
Gen er a l. The strain of A. niger used in this work is
registered at the Museum d’Histoire Naturelle (Paris) under
no. LCP 521 (Lab. de Cryptogamie, 12 rue Buffon, 75005 Paris,
France) and B. sulfurescens was purchased at the ATCC
collection under no. 7159. Corn steep liquor (CSL) is from
Roquette SA. NMR spectra (1H and l3C) were recorded in
CDCl3 solution on a Bruker AC 250. Chemical shifts are
reported in δ from TMS as internal standard. Optical rotation
values were measured on a Perkin-Elmer 241C polarimeter
at 589 nm. Vapor phase chromatography analyses were
performed using either a 25 m capillary column (OV 1701) or
a chiral column [heptakis(6-O-methyl-2,3-di-O-pentyl)-â-cy-
clodextrin]. Determination of the enantiomeric excesses of
diols was performed after derivatization into their acetonides.
HPLC analyses were carried out using an UV detector and a
(16) Nieduzak, T. R.; Margolin, A. L. Tetrahedron: Asymmetry 1991,
2, 113.
(17) Nankai, M.; Nicholas, C.; Fage, D.; Carter, C. Br. J . Pharmacol.
1994, 112, PU142.
(15) Pedragosa-Moreau, S.; Archelas, A.; Furstoss, R. Tetrahedron
1996, 52, 4593.
(18) Marmo, E.; Caputi, A. P.; Rossi, F.; Lampa, E.; Vacca, C.; Brita,
G.; Saini, R. K.; Di Nola, R. Agressologie 1976, 17, 279.

7406 J . Org. Chem., Vol. 61, No. 21, 1996
Pedragosa-Moreau et al.
column (25.5/0.4 cm) filled with 5 µm silica gel using hexane-
ethanol as eluent. Separation and purification of the products
were achieved by flash chromatography (silica gel 60 H from
Merck and solvent mixtures consisting of pentane and ether
in the range of 100% pentane to 100% ether) or by bulb-to-
bulb distillation.
Syn th esis of Su bstr a tes 2-7. (()-p-Meth ylstyr en e
Oxid e (2). This epoxide was obtained by epoxidation of the
corresponding olefin using m-chloroperoxybenzoic acid under
biphasic conditions (CH2C12/phosphate buffer) for 5 h (86%
yield) as described previously.19 1H-NMR δ: 2.34 (s, 3H, CH3);
2.78 (dd, 1H, H2cis, J gem ) 5.4 Hz; J 1-2cis ) 2.6 Hz); 3.12 (dd,
lH, H2trans, J gem ) 5.4 Hz, J l-2trans ) 4.l Hz); 3.82 (dd, lH, Hl,
J l-2cis ) 2.6 Hz, J l-2trans ) 4.1 Hz); 7.2 (m, 4Harom). l3C-NMR
δ: 21.4 (p-CH3); 51.4 (C-2); 52.6 (C-1); 125.7; 129.6; 134.8; 138.2
(C-Ar).
(()-p-F lu or ostyr en e Oxid e (3). This epoxide was ob-
tained using the same procedure as described for 2 (72% yield).
1H-NMR δ: 2.76 (dd, 1H, H2cis, J gem ) 5.3 Hz, J l-2cis ) 2.5 Hz);
3.14 (dd, lH, H2trans, J gem ) 5.4 Hz, J l-2trans ) 4.1 Hz); 3.14 (dd,
1H, Hl, J l-2cis ) 2.6 Hz, J l-2trans ) 4.0 Hz); 6.9-7.1 (m, 2Harom);
7.2-7.4 (m, 2Harom). 13C-NMR δ: 51.2 (C-2); 51.9 (C-l); 115.3-
115.7 (d, C-Ar, J ) 21.8 Hz); 127.1-127.2 (d, C-Ar, J ) 8 Hz);
133.3 (C-Ar); 160.7-164.7 (d, C-Ar, J ) 246 Hz).
ether (three times), washed twice with brine, and dried over
magnesium sulfate. Evaporation of the solvent yielded 5.7 g
of crude product. Purification by flash chromatography (pen-
tane/ether) yielded 3 g (91% yield) of 7 as a yellow solid, mp
84-85 °C. 1H-NMR δ: 2.79 (dd, 1H, H2cis, J gem ) 5.5 Hz, J l-2cis
) 2.4 Hz); 3.24 (dd, lH, H2trans, J gem ) 5.5 Hz, J l-2trans ) 4.2
Hz); 3.97 (dd, lH, Hl, J l-2cis ) 2.4 Hz, J 1-2trans ) 4.0 Hz); 7.45
(d, 2Harom, J ) 8.7 Hz); 8.22 (d, 2Harom, J ) 8.8 Hz). 13C-NMR
δ: 51.5 (C-l); 51.7 (C-2); 123.8; 126.3; 145.3; 147.9 (C-Ar).
Syn th esis of Ra cem ic Diols 9-14. These diols were
prepared as follows by hydrolysis of the corresponding ep-
oxides. To 1 g of epoxide dissolved in a mixture of THF (50
mL) and water (10 mL) were added two drops of concentrated
sulfuric acid. After 24 h of stirring, the hydrolysis was
complete (checked by TLC) and the solution was neutralized
by addition of saturated NaHCO3 solution and extraction with
ether (three times). Evaporation of the washed (saturated salt
solution) and dried (MgSO4) extracts yielded a crude residue
which was further purified by flash chromatography.
(()-9: yield 65%, mp 70 °C. lH-NMR δ: 2.32 (s, 3H, CH3);
3.49 (m, 4H, (2OH+H2)); 4.73 (s large, lH, Hl); 7.11-7.25 (m,
4Harom). 13C-NMR δ: 21.0 (C-3); 67.9 (C-2); 74.5 (C-l); 126.0;
129.1; 137.4; 137.4 (C-Ar). Anal. Calcd for C9H12O2 (152.19):
C, 71.03; H, 7.95. Found: C, 70.65; H, 7.81.
(()-10: yield 73%, mp 56-57 °C, lit.22 mp 53 °C. lH-NMR
δ: 3.20 (s large, 2H, OH); 3.63 (m, 2H, H2); 4.77 (dd, 1H, Hl,
J l-2 ) 3.2 Hz and 8.11 Hz); 7.0 (m, 2Harom); 7.3 (m, 2Harom).
13C NMR δ: 68.1 (C-2); 74.2 (C-1); 115.3 and 115.7 (d, C-Ar, J
) 21.4 Hz); 127.8 and 127.9 (d, C-Ar, J ) 8.2 Hz); 136.2; 160.6
and 164.5 (d, C-Ar, J ) 246 Hz).
(()-p-Ch lor ostyr en e Oxid e (4). This epoxide20 was ob-
tained using the same procedure as described for 2 (70% yield).
lH-NMR δ: 2.75 (dd, lH, H2cis, J gem ) 5.5 Hz, J l-2cis ) 2.6 Hz);
3.15 (dd, 1H, H2trans, J gem ) 5.4 Hz, J 1-2trans ) 4.1 Hz); 3.84
(dd, 1H, Hl, J l-2cis ) 2.6 Hz, J l-2trans ) 4.0 Hz); 7.18-7.26 (m,
2Harom); 7.30-7.34 (m, 2Harom). 13C-NMR δ: 51.3 (C-2); 51.8
(C-l); 126.8; 128.7; 133.9; 136.2; 136.2 (C-Ar).
(()-p-Br om ostyr en e Oxid e (5). This epoxide23 was ob-
tained using the same procedure as described for 2 (70% yield).
lH-NMR δ: 2.74 (dd, 1H, H2cis, J gem ) 5.5 Hz, J l-2cis ) 2.5 Hz);
3.14 (dd, 1H, H2trans, J gem ) 5.4 Hz, J 1-2trans ) 4.1 Hz); 3.83
(dd, 1H, Hl, J l-2cis ) 2.6 Hz, J l-2trans ) 4.1 Hz); 7.13-7.17 (d,
2Harom); 7.45-7.48 (d, 2Harom). 13C-NMR δ: 51.2 (C-2); 51.8
(C-l); 122.0; 127.1; 131.6; 136.7 (C-Ar).
(()-Cya n ostyr en e Oxid e (6). To a stirred solution of 4 g
(18 mmol) of trimethyloxosulfonium iodide in 50 mL of dry
DMSO, placed under argon, was added 0.72 g (18 mmol) of
sodium hydride (60% mineral oil dispersion). After the solu-
tion was stirred for 30 min (formation of the dimethyloxosul-
fonium methylide), a solution (2 g, 15 mmol) of p-cyanoben-
zaldehyde in 10 mL of DMSO was added and the reaction
mixture stirred at room temperature for 2 h. After cooling
and addition of water, the mixture was extracted with ether
(three times). The combined extracts were washed with brine,
dried over sodium sulfate, and evaporated to yield 2.5 g of
crude product. Purification by flash chromatography (pentane/
ether) yielded 1.66 g (76% yield) of 6 as a white solid, mp 34
°C, lit.21 mp 38 °C. lH-NMR δ: 2.76 (dd, lH, H2cis, J gem ) 5.5
Hz, J l-2cis ) 2.4 Hz); 3.2 (dd, lH, H2trans, J gem ) 5.4 Hz, J l-2trans
) 4.2 Hz); 3.9 (dd, lH, Hl, J l-2cis ) 2.5 Hz, J l-2trans ) 4.0 Hz).
13C-NMR δ: 51.5 (C-l and C-2); 111.7 (CN); 118.6; 126.1; 132.2;
143.3 (C-Ar).
(()-p-Nitr ostyr en e Oxid e (7). Racemic 7 was prepared
by cyclization of the corresponding bromohydrin formed by
reduction of ω-bromopropiophenone. To a stirred solution of
5 g (20 mmol) of ω-bromo-4-nitroacetophenone in 50 mL of
MeOH, cooled in an ice bath, was added 0.83 g (22 mmol) of
sodium borohydride. After the ice bath was removed stirring
was continued for 3 h, 2.76 g (20 mmol) of potassium carbonate
was added in the same flask. After 20 h of stirring, 30 mL of
water was added and the mixture was extracted with
(()-11: yield 75%, mp 83-84 °C. lH-NMR δ: 2.89 (s large,
1 H, OH); 3.37 (s large, 1H, OH); 3.57 (dd, 1H, H2, J 1-2 ) 8.2
Hz, J gem ) 11.3 Hz); 3.70 (dd, lH, H2, J l-2 ) 3.2 Hz, J gem
)
11.4 Hz); 4.75 (dd, lH, Hl, J l-2 ) 3.3 Hz, J l-2 ) 8.1 Hz); 7.21-
7.40 (m, 4Harom). 13C-NMR δ: 67.9 (C-2); 74.0 (C-l); 127.4;
128.7; 133.7; 138.9 (C-Ar). Anal. Calcd for C8H9O2Cl
(217.06): C, 55.80; H, 5.27. Found: C, 55.65; H, 5.25.
(()-12: yield 78%, mp 103-104 °C. lH-NMR δ: 3.58 (m,
2H, H2); 3.92 (t, 1H, OH, J OH-H2 ) 5.7 Hz); 4.48 (d, lH, OH,
J OH-Hl ) 3.9 Hz); 4.71 (m, lH, Hl); 7.36 (d, 2Harom, J ) 8.4 Hz);
7.48 (d, 2Harom, J ) 8.5 Hz). 13C-NMR δ: 68.7 (C-2); 74.6 (C-
1); 121.2; 129.3; 131.8; 143.2 (C-Ar). Anal. Calcd for C8H9O2-
Br (217.06): C, 44.3; H, 4.2. Found: C, 44.38; H, 4.22.
(()-13: yield 71%, viscous oil. 1H-NMR δ: 2.25 (t, 1H, OH,
J ) 5.5 Hz); 2.91 (d, 1H, OH, J ) 3.4 Hz); 3.63 (m, 1H, H2);
3.80 (m, 1H, H2); 4.89 (m, 1H, H1); 7.50 (d, 2Harom, J ) 8.4
Hz); 7.66 (d, 2Harom, J ) 8.3 Hz). 13C-NMR δ: 67.7 (C-2); 73.9
(C-1); 111.6 (CN); 118.7; 126.8; 132.3; 145.9 (C-Ar). Anal.
Calcd for C9H9O2N (163.17): C, 66.25; H, 5.56; N, 8.58.
Found: C, 66.21; H, 5.52; N, 8.45.
(()-14: yield 60%, mp 78 °C. 1H-NMR δ: 3.55 (m, 2H, H2);
3.93 (t, 1H, OH, J ) 6.0 Hz); 4.65 (d, lH, OH, J ) 4.0 Hz);
4.79 (m, 1H, Hl); 7.62 (d, 2Harom, J ) 8.5 Hz); 8.11 (d, 2Harom
,
J ) 8.8 Hz). l3C-NMR δ: 68.4 (C-2);74.5 (C-l); 123.8; 128.2;
146.3; l5l.6 (C-Ar).
Gen er a l P r oced u r e for Der iva tiza tion of th e Diols in to
th e Cor r esp on d in g Aceton id es. The corresponding diol (5-
10 mg) and 2,2-dimethoxypropane (100 µL) were stirred in the
presence of a catalytic amount of TsOH for 30 min. After
neutralization with a saturated NaHCO3 solution, the reaction
mixture was extracted with ether, dried over MgSO4, and
directly analyzed by chiral GLC.
Bioh yd r olysis of Styr en e Oxid e Der iva tives (1-7) w ith
A. n iger a n d B. su lfu r escen s. Gen er a l P r oced u r e. The
fungal strains were cultured in a 2 L fermentor as previously
described.3 After incubation (40 h for A. niger, 48 h for B.
sulfurescens), the mycelium was filtered off, washed with
water, and then placed back in the same fermentor filled with
1 L of a pH 8 phosphate buffer (0.1 M) solution. The medium
was stirred at 700 rpm and maintained at 27 °C. The
appropriate racemic styrene oxide derivative (1 g) was added
to the medium as a solution in ethanol (10 mL). The course
of the bioconversion was followed by withdrawing two aliquots
(2 mL) at regular time intervals. One of these samples was
extracted with pentane (2 mL) and subjected to direct chiral
GC analysis, which allowed for determination of the ee of the
(19) Groves, J . T.; Myers, R. S. J . Am. Chem. Soc. 1983, 105, 5791.
(20) Hudlicky, T.; Boros, E. E.; Boros, C. H. Tetrahedron: Asymmetry
1993, 4, 1365.
(21) Pluim, H; Wynberg, H. J . Org. Chem. 1980, 45, 2498.
(22) Imuta, M.; Ziffer, H. J . Org. Chem. 1979, 44, 1351.
(23) Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P.
Tetrahedron: Asymmetry 1993, 4, 1325.
(24) Eistert, B.; J urazyk, H.; Arackal, T. J . Chem. Ber. 1976, 109,
640.
(25) Broquet, C.; Pasero, J . J . C. R. Hebd. Seances Acad. Sci., Ser.
C 1967, 265(16), 873.

Microbiological Transformations
J . Org. Chem., Vol. 61, No. 21, 1996 7407
residual epoxide. After saturation with NaCl, the second
sample was extracted with ether (2 mL), and after purification
through silica gel (using a prepacked silica gel Extrasep
column) and derivatization to the acetonide, the ees of the diols
were determined by chiral GC analysis. The bioconversion was
stopped by addition of ether (500 mL). The medium was
filtered off, and the fungal cake was separately extracted two
times with ether (2 × 200 mL). After decantation, the aqueous
phase was saturated with NaCl and then continuously ex-
tracted with dichloromethane (48 h). The combined organic
layers were dried (MgSO4), and purification of the products
was achieved by flash chromatography or/and by bulb-to-bulb
distillation. Preparative yields and ees of residual epoxides
and formed diol are given in Tables 1 and 2. All these
bioconversions were performed several times, in particular in
the case of the chloro derivative 4 using B. sulfurescens as
biocatalyst.
(1R)-10, [R]20 -49 (c 1.07, CHCl3) using A. niger; [R]20
D
D
-49.2 (c 1.01, CHCl3) using B. sulfurescens [100 °C, tR(1R) 8.2
min and tR(1S) 8.5 min].
(1R)-11, [R]20 -31.4 (c 0.97, EtOH) using A. niger; lit.20
D
[R]20 -27.60 (c 0.96, EtOH) for (1R) [110 °C tR(1R) 16.6 min
D
and tR(1S) 17.0 min].
(1R)-12, [R]20 -37.2 (c 1.03, CHCl3) using A. niger; [R]20
D
D
-36.5 (c 1.00, CHCl3) using B. sulfurescens [110 °C, tR(1R) 29.5
min and tR(1S) 30.4 min].
(1R)-13, [R]20 -17.1 (c 0.53, EtOH) using A. niger [120 °C
D
tR(1R) 32.0 min and tR(1S) 33.0 min].
(1R)-14, [R]20 -15.0 (c 0.71, MeOH) using A. niger; lit.21
D
[R]20 -20.0 (c 1.15, MeOH); for (1R)-14 [130 °C tR(1R) 36.8
D
min and tR(1S) 37.8 min].
Bioh yd r olyses b y Solu b le E n zym e E xt r a ct fr om A.
n iger . All enzymatic reactions were carried out at 25 °C in a
minivial (3 mL) containing phosphate buffer (1.84 mL, 0.1 M,
pH 7) and 8-53 mg of protein/mL.12 Reactions were initiated
by addition of 0.16 mL of the epoxide in ethanol (range of
0.125-24 mM). The course of the biohydrolysis was followed
by withdrawing samples (0.25 mL) at different time intervals.
After saturating with NaCl, samples were extracted with CH2-
Cl2 (0.5 mL). The produced diols were quantified by HPLC
analysis using a specific calibration curve for each diol. Initial
rates of hydration were calculated at each initial substrate
concentration from the amounts of diol formed. The apparent
kinetic parameters (Km and Vm) were calculated by nonlinear
regression of the Michaelis’ equation (see Table 3).
Optical rotations and chiral GLC data of the isolated
products after biohydrolysis are as follows:
(1S)-2, [R]20 +23.0 (c 0.61, CHCl3); (1R)-2, [R]20 +25.5 (c
D
D
1.3, PhH); lit.19 [R]20 (+) (PhH) for (1R)-2 [70 °C, tR(1R) 16.0
D
min and tR(1S) 17.5 min].
(1S)-3, [R]20 +15.6 (c 0.97, CHCl3); (1R)-3, [R]20 -15.1 (c
D
D
0.45, CHCl3); lit.16 [R]20 +16.7 (c 1.03, CHCl3) for (1S)-3 [80
D
°C, tR(1R) 6.2 min and tR(1S) 7.3 min].
(1S)-4, [R]20 +19.3 (c 1.16, CHCl3) [100 °C, tR(1R) 8.6 min
D
and tR(1S) 9.2 min].
(1S)-5, [R]20 +13.6 (c 1.46 CHCl3); (1R)-5, [R]20 -12.9 (c
D
D
1.04, CHCl3) [110 °C, tR(1R) 9.9 min and tR(1S) 10.4 min].
(1S)-6, [R]20 +7.3 (c 1.46, CHCl3) [120 °C tR(1R) 11.2 min
Ack n ow led gm en t. One of us (S.P.-M.) is indebted
to the Socie´te´ Roussel-Uclaf for financial support.
Financial support to this work and a fellowship from
the Socie´te´ Rhoˆne-Poulenc Rorer to C.M. are greatly
acknowledged.
D
and tR(1S) 11.8 min].
(1S)-7, [R]20D +37.6 (c 1.99, CHCl3); lit.21 [R]20D +37.4 (c 1.99,
CHCl3) for (1S)-7 [130 °C, tR(1R) 14.2 min and tR(1S) 15.1 min].
(1R)-9, [R]20D -44.7 (c 0.75, CHCl3) using A. niger; [R]20D -50
(c 0.87, CHCl3) using B. sulfurescens [100 °C, tR(1R) 14.1 min
and tR(1S) 14.6 min].
J O960558I