Microbial Baeyer-Villiger oxidation applied to the synthesis of the N-protected (1R,5R)-Geisman-Waiss lactone

Article abstract of DOI:10.1016/j.tetasy.2005.06.038

Using whole cell Baeyer-Villiger biooxidation as a key step and depending on the choice of the biocatalyst {E. coli TOP10[pQR239] or Acinetobacter TD63}, the synthesis of nearly enantiopure N-protected (1R,5R)-(-)-Geisman-Waiss lactone (92% ee) or its (1R

Full text of DOI:10.1016/j.tetasy.2005.06.038

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 2521–2524
Microbial Baeyer–Villiger oxidation applied to the synthesis
of the N-protected (1R,5R)-Geisman–Waiss lactone
*
´
´
Amparo Luna, Maria-Concepcion Gutierrez, Roland Furstoss and Veronique Alphand
Groupe Biocatalyse et Chimie Fine, FRE CNRS 2712, Universite´ de la Me´diterrane´e, Case 901, 163 Avenue de Luminy,
13288 Marseille Cedex 9, France
Received 17 May 2005; accepted 30 June 2005
Abstract—Using whole cell Baeyer–Villiger biooxidation as a key step and depending on the choice of the biocatalyst {E. coli
TOP10[pQR239] or Acinetobacter TD63}, the synthesis of nearly enantiopure N-protected (1R,5R)-(ꢀ)-Geisman–Waiss lactone
(92% ee) or its (1R,5S)-(ꢀ) regioisomer (98% ee) be performed.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
We therefore focused our attention on the biooxidation
of N-protected 2-aza-bicyclo[3.2.0]heptan-6-one 1, a
precursor of the so-called Geisman–Waiss lactone (2-
oxa-6-azabicyclo[3.3.0]octan-3-one). This lactone and
its N-protected derivatives 2 are versatile intermediates
for the synthesis of necine base pyrrolizidine alkaloids,7
a large class of natural products widely found in plants
(Scheme 1).8
Whole-cell or enzyme mediated Baeyer–Villiger (BV)
biooxidations allow the efficient transformation of
ketones into lactones with high regio-, stereo- and
enantioselectivity. They favourably compare with BV
oxidations using transition metal-based catalysts.1,2
The enzymes responsible for this type of reaction, that
is, Baeyer–Villiger monooxygenases (BVMOs), belong
to the flavin monooxygenase family and, because they
are NAD(P)H dependent, need cofactor recycling.1,3
However, due to several practical bottlenecks such as
substrate and/or product inhibition, such biotransfor-
mation processes have not been amenable to efficient
upscaling until now.4 As an interesting breakthrough,
we have recently shown that the combination of a whole
cell process (using a recombinant strain overexpressing a
BVMO) with a resin-based in situ substrate feeding
product removal methodology5 allowed us to overcome
these drawbacks and potentially pave the way to indus-
trial development, as exemplified by a very recent kilo-
gram scale application.6 This encouraged us to
broaden the scope of substrates transformable in high
regio- and/or enantioselectivity.
R1O
CH2OR2
H
N
R1, R2 : necic acid moieties
R1 = R2 = H (+)-retronecine
Scheme 1. General scheme of necine base pyrrolizidine alkaloids.
Pyrrolizidine alkaloids are responsible for livestock poi-
soning through contaminated crops and are involved in
the chemical defence of insects or amphibians. However,
in spite of their various biological activities (ranging
from anti-feedant to insecticidal, fungicidal, bactericidal
and virucidal effects8), studies focusing on the biological
effect of synthetic pyrrolizidine alkaloids analogues or of
unnatural stereoisomers of these compounds are scarce,
a feature that may be partly due to the lack of appropri-
ate synthetic methods.9
To date, only a few examples of BV biooxidation of
ketones bearing a nitrogen atom have been reported.1
Herein, we report the preparation of nearly enantiopure
N-protected (1R,5R)-Geisman–Waiss lactone 2 and of
its (1R,5S) regioisomer 3 via the whole cell mediated
Baeyer–Villiger oxidation of rac-1 (Scheme 2).
*
Corresponding author. Tel.: +33 (0)4 91 82 91 58; fax: +33 (0)4 91 82
91 45; e-mail: alphand@luminy.univ-mrs.fr
0957-4166/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.06.038

2522
A. Luna et al. / Tetrahedron: Asymmetry 16 (2005) 2521–2524
2. Results and discussion
OH
1)ClCbz, iPr2NEt, THF, 0°C
EtO
NH2
2) aq HCl
OEt
N
2.1. Synthesis of rac-1
4
5
Cbz
Over the course of studies aimed at the [2+2] cycloaddi-
tion of endocyclic enecarbamates,10 Correia et al.
described a synthesis of ketone 1 via the unstable
1-pyrroline intermediate.10c However, because of the
low yield we obtained following this method (<10%),
we decided to follow a slightly different route involving
cyclization of an aminobutyraldehyde11 (Scheme 2). The
commercially available aminobutyraldehyde diethyl ace-
tal 4 was protected by a Cbz (PhCH2OC(O)–) moiety at
the nitrogen atom and was cyclized under acid catalysis,
to afford the corresponding hydroxy-carbamate 5. Sub-
sequent dehydration of 5 in toluene (using a Dean–Stark
device) led to the corresponding endocyclic enecarba-
mate. Because of its easy rehydration, the [2+2] cyclo-
addition was carried out directly in the same pot and
solvent. The dichloroketene was formed in situ by reac-
tion of dichloroacetyl chloride with triethylamine.10 The
classical treatment of dichloroketone 6 with AcOEt/Zn
led to the undesired compound 7 (82% yield). However,
a reductive dechlorination could be successfully carried
out using a Zn/Cu mixture in methanol (prior saturated
with ammonium chloride).10b The desired N-Cbz-2-aza-
bicycloheptanone 1 was finally obtained in 50% overall
yield (Scheme 3).
O
1) toluene,
reflux
Zn /
(Dean Stark)
AcOH
N
Cbz
7
2) Cl2CHCOCl,
Et3, toluene,
N2, rt,
O
MeOH (sat. NH4Cl)
rac-1
50% overall yield
Cl
Zn / Cu alloy
N2, rt
N
Cl
Cbz
6
Scheme 3. Synthesis of racemic azabicycloheptanone 1.
(CHMO B)] and one recombinant strain overexpressing
CHMO A, E. coli TOP10[pQR239], which was con-
structed by Ward et al.14 Wild type cells were grown
in a mineral media supplemented with 1,2-cyclohexane-
diol as previously described.15 Recombinant cells were
grown in a glycerol-based medium while the enzyme
expression was induced by L-arabinose as previously
described.5
2.3. Biotransformations
O
O
(R)
(S)
(R)
(R)
O
biocatalyst
+
O
O
Analytical scale biotransformations were carried out in
250 mL shake flasks at 30 ꢁC using 2 mM of rac-1 and
50 mL of cell broth. These were monitored by GC anal-
ysis of samples taken at regular intervals. Ketone and
lactone ees were determined by HPLC (Chiralcel AS
column-Daicel) as well as by chiral GC (Chirasil
Dex-Chrompak- and CycloSilb-JW Scientific-columns)
analyses. The results reported in Table 1 were obtained
at ca. 50–60% conversion ratio.
R=Cbz
N
R
N
R
N
R
rac-1
(1R,5S)-(-)-3
(1R,5R)-(-)-2
N-protected
Geisman-Waiss lactone
Scheme 2. Whole-cell biotransformation of rac-1.
2.2. Biocatalysts
Whatever be the biocatalyst, the whole cell mediated
biotransformation afforded two lactones in high ees:
the ÔnormalÕ one, the N-protected Geisman–Waiss lac-
tone 2 (corresponding to the usual oxygen insertion
between the more substituted carbon atom and the
carbonyl group) and the ÔabnormalÕ lactone 3 (arising
from oxygen insertion on the other side of the carbonyl
group).
In order to explore the BV oxidation of rac-1, we
selected two Baeyer–Villiger monooxygenases among
the most studied. These enzymes are cyclohexanone
monooxygenases, named here CHMO A and B. The
employed biocatalysts were the two wild type bacterial
strains [respectively, Acinetobacter calcoaceticus
NCIMB 9871 (CHMO A)12 and Acinetobacter TD6313
Table 1. Biotranformations of rac-1
Enzyme
Microorganism
Reaction
time (min)
Ketone 1
yielda % (ee %)
Normal lactone 2
yielda % (ee %)
Abnormal lactone 3
yielda % (ee %)
Ratio 2:3
CHMO A
A. calcoaceticus NCIMB 9871
E. coli TOP10[pQR239]
120
41
(86)
ndb
36
(96)
37
10
(93)
11
3.6:1
3.4:1
60
(>95)
(94)
(87)
CHMO B
Acinetobacter TD63
300
45
11
29
1:2.6
(74)
(87)
(98)
a Ketone and lactone yields were measured by GC using internal standard method.
b An impurity from the growth medium prevented us from a correct calculation of the ketone yield.

A. Luna et al. / Tetrahedron: Asymmetry 16 (2005) 2521–2524
2523
2.4. Regioselectivity
100
80
In the case of both CHMO A expressing strains (wild
type or recombinant strain), the desired Geisman–Waiss
lactone 2 was the predominant product in a regioiso-
meric 2/3 ratio of about 3.5:1. This ratio was similar
to that obtained upon mCPBA-mediated BV oxidation
(3:1). However, an inversion of this ratio was observed
with Acinetobacter TD63 (CHMO B), which led to the
preferential formation of the ÔabnormalÕ lactone 3 (in a
regioisomeric ratio 2/3 of about 1:2.6). This was rather
unexpected since, until now, we had never observed
any noticeable qualitative differences between both
CHMOs, A and B, the latter exhibiting only a slightly
lower enantioselectivity.15,16 As observed previously
with other substrates, 1 also underwent a reduction to
afford alcohols in <5% yield.
60
40
20
0
0
5
10
15
20
25
30
35
40
45
time (min)
Figure 1. Time course of the semi-preparative biotransformation with
E. coli [pQR239] cells. Ketone: h ee; normal lactone: d yield, · ee;
abnormal lactone: m yield, + ee.
endo-alcohol was also isolated. The deprotection of 2
to give the (ꢀ)-Geisman–Waiss lactone could be easily
carried out by hydrogenation on Pd as previously
reported.7c
2.5. Enantioselectivity
In all the biotransformations, ketone ee regularly
increased whereas the lactone ee was quite high during
the first part of the reaction (87% < ee < 98%) and
decreased later on. Apparent E value (enantiomeric
ratio calculated using substrate and product ees), could
be estimated to superior to 80 for both CHMOs.
2.6. Absolute configuration assignment
Comparison of the specific rotation sign of purified 2
with the literature data7c led us to assign the (1R,5R)
absolute configuration to this lactone. Interestingly, it
was the desired enantiomer for the synthesis of natural
necine bases as (+)-retronecine. Chemical (mCPBA) oxi-
dation of the recovered substrate 1 (79% ee) afforded a
mixture of 2 and 3 exhibiting the same ee. Since all enan-
tiomers were separable by chiral GC, the retention time
comparison of the thus synthesized lactone 2 with that
of (1R,5R)-enantiomer formed by bioconversion entitled
us to assign the (1S,5S)-configuration to the recovered 1.
Finally, the retention time comparison of chemically
and microbiologically obtained 3 allowed us to assign
the (1R,5S)-configuration to the ÔmicrobialÕ abnormal
lactone.
In order to determine the absolute configuration of the
lactones and to validate this biotransformation as a
potential synthetic method, a semi-preparative scale
experiment was carried out in a 2 L fermenter (750 mL
broth, 30 ꢁC, pH 7.2) using E. coli [pQR239] cells. As
the recombinant strain mediated oxidation was very
fast, grown cells were diluted twofold before use (with
a fresh culture medium) to provide a cell suspension of
1.4 g Lꢀ1 dry weight. Thus, the observed drop of lactone
ee was slowed down enough to allow, thanks to a timely
monitoring, the quenching of the reaction at high
lactone ee. A solution of ketone rac-1 (in ethanol) was
added to reach
a 2.7 mM initial concentration
(0.67 g Lꢀ1). The biotransformation was stopped once
the unreacted ketone 1 reached about 80% ee (ca.
40 min). The time course of the biotransformation is
reported in Figure 1.
It is noteworthy that, using CHMOs, both regioisomeric
lactones were obtained from the same single ketone
enantiomer, allowing consequently the recovery of opti-
cally active 1. This feature is quite different from that
observed using bicyclo[3.2.0]hept-2-en-6-one (a structur-
ally very close compound) as substrate.16 In this experi-
ment, an almost perfect regiodivergent parallel kinetic
resolution was described:17 both lactones were formed
simultaneously, each of them arising from a different
ketone enantiomer. Obviously, the presence of a nitro-
gen atom and/or of an increased steric hindrance due
to the nitrogen protecting group strongly influences on
the overall enzyme behaviour.
After continuous extraction of the broth with dichloro-
methane and purification by flash chromatography, lac-
tone 2 was isolated in 32% yield and 92% ee
32
{½aꢁD ¼ ꢀ107 (c 0.7, CHCl3)}. The minor lactone 3à
32
was obtained in 10% yield and 94% ee (½aꢁD ¼ ꢀ170 (c
0.3, CHCl3)), whereas the unreacted ketone 1 was recov-
ered in 40% yield and 79% ee. A small amount (4%) of
(1R,5R)-N-Cbz-2-oxa-6-azabicyclo[3.3.0]octan-3-one 2 was identified
by comparison of its spectroscopic data with those already described
in the literature.7c
3. Conclusion
à For (1R,5S)-N-Cbz-3-oxa-6-azabicyclo[3.3.0]octan-2-one 3: 1H NMR
(CHCl3-d, 250 MHz): d 2.13 (m, 1H), 2.29 (m, 1H), 3.20 (m, 2H), 3.73
(m, 1H), 4.42 (m, 3H), 5.08 (m, 2H), 7.29 (s, 5H); 13C NMR (CHCl3-
d, 250 MHz), mixture of rotamers: d 27.0 (CH2), 27.6 (CH2), 43.9
(CH), 44.7 (CH), 45.2 (CH2), d 45.6 (CH2), 57.7 (CH2), 58.4 (CH),
67.3 (CH2), 72.5 (CH2), 73.4 (CH2), 128.0 (CH), 128.1 (CH), 128.6
(CH), 136.2 (C), 154.0 (C@O), 177.8 (C@O).
Nearly enantiopure (1R,5R)-N-Cbz-2-oxa-6-azabicyclo-
[3.3.0]octan-3-one 2 (N-protected Geisman–Waiss lac-
tone) and its regioisomer (1R,5S)-3 were obtained in
high ee by whole-cell Baeyer–Villiger biooxidations of
ketone rac-1. CHMO A expressing strains favoured
the formation of the desired lactone 2 whereas

2524
A. Luna et al. / Tetrahedron: Asymmetry 16 (2005) 2521–2524
Gutierrez, M. C.; Alphand, V.; Wohlgemuth, R.; Furstoss,
R. Org. Lett. 2004, 6, 1955–1958; (c) Simpson, H.;
Alphand, V.; Furstoss, R. J. Mol. Catal. B: Enzym.
2001, 16, 101–108; (d) Zambianchi, F.; Raimondi, S.;
Pasta, P.; Carrea, G.; Gaggero, N.; Woodley, J. M. J.
Mol. Catal. B: Enzym. 2004, 31, 165–171.
CHMO B expressing strain predominantly led to the
ÔabnormalÕ lactone 3.
Work is currently in progress in our laboratory to study
the influence of the size and nature of the nitrogen atom
protecting group towards the enzyme selectivity and to
determine the requirements for high enantioselectivity.
6. Hilker, I.; Alphand, V.; Wohlgemuth, R.; Furstoss, R.
Biotechnol. Bioeng., in press.
7. For syntheses of 2: (a) Ambrosio, J. C. L.; Santos, R. H.;
Correia, C. R. D. J. Braz. Chem. Soc. 2003, 1, 27–38; (b)
Wee, A. G. H. J. Org. Chem. 2001, 66, 8513–8517, and
references cited therein; (c) Takahata, H.; Banba, Y.;
Momose, T. Tetrahedron: Asymmetry 1991, 2, 445–448.
8. (a) Michael, J. P. Nat. Prod. Rep. 2001, 18, 520; (b)
Harborne, J. B. Nat. Prod. Rep. 2001, 18, 361, and
references cited therein.
Acknowledgements
This work was funded by the European Community 5th
Framework Program (QLK3-CT2001-00403). We are
very grateful to Prof. J. Ward (University College of
London) for providing us with recombinant strain and
also Dr. Correia (University of Campinas, Brazil) for
helpful discussions.
9. Montes de Oca, A. C. B.; Correia, C. R. D. Arkivoc 2003,
390–403.
10. (a) De Faria, A. R.; Salvador, E. L.; Correia, C. R. D.
J. Org. Chem. 2002, 67, 3651–3661; (b) Miranda, P. C. M.
L.; Correia, C. R. D. Tetrahedron Lett. 1999, 40, 7735–
7738; (c) De Faria, A. R.; Matos, C. R. R.; Correia, C. R.
D. Tetrahedron Lett. 1993, 34, 27–30.
References
1. For recent reviews, see: (a) ten Brink, G.-J.; Arends, R. A.
I. W. C. E.; Sheldon, R. A. Chem. Rev. 2004, 104, 4105–
4123; (b) Mihovilovic, M. D.; Rudroff, F.; Gro¨tzl, B. Curr.
11. Grotjahn, D. B.; Vollhardt, K. P. C. Synthesis 1993, 579–
605.
12. Stewart, J. D. Curr. Org. Chem. 1998, 2, 211–232.
13. Davey, J. F.; Trudgill, P. W. Eur. J. Biochem. 1977, 74,
115–127.
14. (a) Doig, S. D.; OÕSullivan, L. M.; Patel, S.; Ward, J.;
Woodley, J. M. Enzyme Microb. Technol. 2001, 28, 265;
(b) Doig, S. D.; Simpson, H.; Alphand, V.; Furstoss, R.;
Woodley, J. M. Enzyme Microb. Technol. 2003, 32, 347.
15. Alphand, V.; Furstoss, R.; Pedragosa-Moreau, S.; Rob-
erts, S. M.; Willetts, A. J. J. Chem. Soc., Perkin Trans. 1
1996, 1867–1872.
Org. Chem. 2004, 8, 1057; (c) Mihovilovic, M. D.; Muller,
¨
B.; Stanetty, P. Eur. J. Org. Chem. 2002, 22, 3711–3730.
2. Watanabe, A.; Uchida, T.; Irie, R.; Katsuki, T. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 5737–5742.
3. (a) Kamerbeek, N. M.; Janssen, D. B.; van Berkel, W. J.
H.; Fraaije, M. W. Adv. Synth. Catal. 2003, 345, 667–678;
(b) Fraaije, M. W.; Kamerbeek, N. M.; van Berkel, W. J.
H.; Janssen, D. B. FEBS Lett. 2002, 43–47.
4. Alphand, V.; Carrea, G.; Wohlgemuth, R.; Furstoss, R.;
Woodley, J. M. Trends Biotechnol. 2003, 21, 318–323.
5. (a) Hilker, I.; Alphand, V.; Wohlgemuth, R.; Furstoss, R.
Adv. Synth. Catal. 2003, 346, 203–214; (b) Hilker, I.;
16. Alphand, V.; Furstoss, R. J. Org. Chem. 1992, 67, 1306–
1309.
17. Dehli, J. R.; Gotor, V. Chem. Soc. Rev. 2002, 31, 365–370.