2
400
N. Monfort et al. / Tetrahedron: Asymmetry 13 (2002) 2399–2401
5
the fungal Aspergillus niger (AnEH) epoxide hydrolase,
which we have recently cloned and overexpressed, gave
triphenylphosphine in carbon tetrachloride (80°C, 7 h).
9
Indeed, this essentially afforded a single enantiomer of
the corresponding chloro-epoxide 2, which proved to be
the same (identical retention time on chiral GC) as the
one recovered after biohydrolysis. These experiments
serve to confirm that the products obtained from this
biohydrolysis were, without doubt, the unreacted
chloro-epoxide (S)-2 and the diol (R)-3, respectively.
the best results. Interestingly, this enzyme exhibited an
excellent enantioselectivity as well as a very high activ-
ity towards 2. Thus, a preparative scale experiment
carried out using (±)-2 (400 mg, 2 mM) and AnEH
crude enzyme (13 mg, specific activity 8 U/mg
1
0
protein) in phosphate buffer solution (100 mM con-
taining 10% DMSO, pH 7, 1 L) led, after a reaction
time of 4 h 45 min, to a conversion ratio of nearly 50%.
Stopping the reaction at this point allowed recovery of
the unreacted epoxide (S)-2, which showed an excellent
ee (98.3%) in good yield (41%) as well as of the formed
chloro-diol 3 in nearly enantiopure form (98.3% ee,
Since determination of the regioselectivity is also an
important factor in this type of reaction, we have also
determined the a(S)/b(S) and a(R)/b(R) ratios for each
enantiomer of 2. These were calculated on the basis of
the ee of the diol 3 obtained upon biohydrolysis of
enantiopure (S)-2 and the ee of the diol 3 formed after
38% yield) (Scheme 2).
1
1
total conversion of the racemic epoxide. The results
indicate that attack by water occurred almost exclu-
sively at the b-carbon atom for each enantiomer [a(S)/
b(S)=5/95 and a(R)/b(R)=1/99]. As a result, this
biohydrolysis essentially led to retention of configura-
tion at the stereogenic carbon atom of 2, an observa-
tion consistent with the fact that the ee of the formed
diol 3 decreased throughout conversion of (±)-2, down
to around 4% at total conversion.
The conversion ratio and the ee of the recovered epox-
ide (S)-2 were determined by chiral GC analysis
(
Lipodex G, 110°C, Macherey–Nagel). The ee of the
formed diol 3 could be determined (using the same
chiral column, 120°C) after intramolecular cyclisation
into the corresponding epoxy-alcohol 4. It was shown
to be also higher than 98%. The E value, calculated on
the basis of the ee of the residual epoxide and either the
ee of the formed diol or the conversion ratio, appeared
to be higher than 200.
The aim of this work was to explore the possibility of
developing a new strategy allowing further preparative
scale syntheses of enantiopure building blocks usable
for elaboration of the potentially important antifungal
drug D0870. Our results show that this goal could, in
principle, be reached by performing the resolution of
The absolute configuration of the obtained chloro-diol
3
could be established by chemical correlation with the
5
previously described epoxy-alcohol (S)-4. Thus,
intramolecular cyclisation of 3 (dry THF/NaH, 0°C, 1.5
2
6
H) led to 4 ([h] =+44.6 (c 1; THF); lit. for (S)-4:
D
(
±)-2 using our recombinant A. niger epoxide hydro-
[
h] =42 (c 1; THF)). This allowed us to conclude that
D
2
6
lase. Indeed, this enzyme shows good activity against 2
and the observed E value was shown to be excellent
the chloro-diol product 3 ([h] =+3.6 (c 1; THF)) had
D
R absolute configuration. As a consequence it could be
deduced that, most probably, this (R)-chloro-diol 3 was
formed by hydrolysis of the R enantiomer of 2, assum-
ing that nucleophilic attack by water occurred at the
less substituted carbon atom of the epoxide moiety.
(
>200). This therefore allows a very effective resolution
of substrate 2. Both the recovered chloro-epoxide (S)-2
and the formed chloro-diol (R)-3 were thus obtained in
very good yields and in nearly enantiopure form at a
conversion ratio of about 50%. Since both these prod-
ucts can be used in the synthesis of our target, this
approach interestingly allows the industrially very
important ‘100% yield/100% ee’ criteria to be fulfilled.
Further work is going on in our laboratory in order to
improve this resolution for preparative scale applica-
tion, and this will be described in due course.
(
Theoretically, formation of (R)-3 could also result
from attack of (S)-2 at the more substituted carbon
atom.) In this case, the absolute configuration of the
2
6
unreacted chloro-epoxide 2 ([h] =+45 (c 1.2; THF))
D
should be of S absolute configuration. This assignment
was confirmed by treatment of the epoxy-alcohol (R)-4
(
obtained from (R)-3 as described above) with
Scheme 2.
Monfort, Nicolas
