Biocatalytic synthesis towards both antipodes of 3-hydroxy-3-phenylpropanitrile a precursor to fluoxetine, atomoxetine and nisoxetine

Article abstract of DOI:10.1016/j.tetlet.2006.12.057

The bakers' yeast reduction of 3-oxo-3-phenylpropanenitrile (1) has been difficult to achieve due to a dominant alkylating mechanism. A library of 20 bakers' yeast reductases, that are overexpressed in Escherichia coli, were screened against (1). Four enzymes were found to reduce this substrate and by varying the enzyme both enantiomers of 3-hydroxy-3-phenylpropanitrile (2) could be prepared with a high enantiomeric excess. In addition, the Escherichia coli whole-cell system can be optimized to nearly eliminate the competing alkylating mechanism. By using this system, a formal biocatalytic synthesis of both antipodes of fluoxetine, atomoxetine and nisoxetine has been demonstrated.

Full text of DOI:10.1016/j.tetlet.2006.12.057

Tetrahedron Letters 48 (2007) 1217–1219
Biocatalytic synthesis towards both antipodes of
-hydroxy-3-phenylpropanitrile a precursor to fluoxetine,
atomoxetine and nisoxetine
3
a
a
b
a,
*
Richard J. Hammond, Benjamin W. Poston, Ion Ghiviriga and Brent D. Feske
a
Department of Chemistry and Physics, Armstrong Atlantic State University, 11935 Abercorn St. Savannah, GA 31419, USA
b
Department of Chemistry, University of Florida, PO Box 117200, Gainesville, FL 32611, USA
Received 10 October 2006; revised 1 December 2006; accepted 12 December 2006
Available online 8 January 2007
Abstract—The bakers’ yeast reduction of 3-oxo-3-phenylpropanenitrile (1) has been difficult to achieve due to a dominant alkylating
mechanism. A library of 20 bakers’ yeast reductases, that are overexpressed in Escherichia coli, were screened against (1). Four
enzymes were found to reduce this substrate and by varying the enzyme both enantiomers of 3-hydroxy-3-phenylpropanitrile (2)
could be prepared with a high enantiomeric excess. In addition, the Escherichia coli whole-cell system can be optimized to nearly
eliminate the competing alkylating mechanism. By using this system, a formal biocatalytic synthesis of both antipodes of fluoxetine,
atomoxetine and nisoxetine has been demonstrated.
Ó 2006 Elsevier Ltd. All rights reserved.
1. Introduction
ity to reduce an assortment of substrates can be
explained by the many reductases it contains. Unfortu-
5
The treatment of a major depressive disorder using sero-
tonin/norepinephrine reuptake inhibitors has become a
billion dollar industry since their inception. These inhib-
itors have also been used to treat other disorders such as
anxiety, alcoholism, chronic pain, migraine headaches,
urinary incontinence, sleep disorders, memory disorders,
obesity and bulimia. Specific interest has been placed on
nately, this large number of reductases often leads to a
mixture of products formed by competing enzymes. To
circumvent this problem, GST-reductase chimeras were
engineered and placed into Escherichia coli creating a
5
–7
bakers’ yeast reductase library. We have used this sys-
tem to screen the stereospecificity of a single reductase
for a given substrate by the use of the pure fusion pro-
TM
TM
7–9
Prozac
(fluoxetine) (3), Straterra
(atomoxetine) (4)
tein or used directly in whole cells.
and nisoxetine (5) because these inhibitors show little
1
,2
effect on dopaminergic and noradrenergic receptors.
This initiated our interest in the reduction of 3-oxo-3-
phenylpropanenitrile (1), because 3-hydroxy-3-phenyl-
propanitrile (2) is a simple precursor to these popular
2. Results and discussion
The asymmetric synthesis of (2) has been accomplished
1
0
serotonin/norepinephrine reuptake inhibitors (Scheme
by the oxidation of styrene, ruthenium catalyzed
3
11–13
14
1
).
reductions of (1),
enzyme reduction of (1), aldol-
1
5–17
type addition of alkylnitriles to benzaldehyde,
and
The biocatalytic
reduction of (1) has also been achieved by both bakers’
1
8–21
Bakers’ yeast (Saccharomyces cerevisiae) is a popular
biocatalytic tool for organic chemists. It has been shown
to catalyze the forming and breaking of C–C bonds, oxi-
kinetic resolutions using lipase.
2
2–30
yeast and the fungus Curvularia lunata (Scheme 2).
4
dations, hydrolysis and a variety of reductions. Its abil-
Surprisingly, these whole-cell systems did not afford
the expected alcohol (2) as the dominant product, but
instead these organisms gave a mixture of a-alkylated
product 2-cyano-1-phenylbutanone (6) and alcohol S-
(2). This led to the screening of (1) against the library
of bakers’ yeast reductases, using our E. Coli whole-cell
overexpression system. Out of the 20 reductases
TM
Keywords: Reductase; Bakers’ yeast; Prozac ; Fluoxetine; Atomoxe-
TM
tine; Nisoxetine; Straterra
.
*
Corresponding author. Tel.: +912 921 5658; fax: +912 961 3226;
e-mail: feskebre@mail.armstrong.edu
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.12.057

1218
R. J. Hammond et al. / Tetrahedron Letters 48 (2007) 1217–1219
CF3
ref. 3
O
NHMe
Fluxetine (3)
O
asymmetric
reduction
OH
*
O
ref. 3
CN
CN
NHMe
Atomoxetine (4)
(1)
(2)
H CO
3
ref. 3
O
NHMe
Nisoxetine (5)
(atomoxetine) (4) and nisoxetine (5).
TM
TM
Scheme 1. Overall strategy towards Prozac
(fluoxetine) (3), Straterra
O
O
OH
CN Bakers' Yeast
CN
CN
or
Curvularia lunata
24 h
1
6
S
-2
major
minor
Scheme 2. Whole-cell reduction of (1) with bakers’ yeast or C. lunata over 24 h.
screened, four were found to reduce (1) (Table 1).
Unlike the results from bakers’ yeast and C. lunata,
the heterologous E. coli system afforded both R-(2) and
S-(2) with a high enantiomeric excess. The asymmetric
synthesis of R-(2) is of particular importance because
it is a stereochemical mimic and building block for the
(Luria–Bertani) and a nitrogen deficient non-growing
media (phosphate buffer pH = 6, 4 mM glucose).
The whole-cell reduction of (1) by YOL151w and
YGL039w in a nutrient rich growth media afforded
ethylated product (6) at about 39% of the overall yield
as shown by GC analysis (Table 2). The two whole-cell
reactions were then performed in a nitrogen deficient
media (Scheme 3), which is a system that has proven
synthesis of (R)-atomoxetine (4) and (R)-nisoxetine (5).
TM
Additionally, synthesis of optically pure Prozac
(3)
may be advantageous since it is currently delivered
as the racemate, even though the stereochemistry
of the alcohol has shown to effect its inhibitory
8
,9
successful for previous substrates. Both non-growing
systems yielded ꢀ1 g/L final concentration of (2) and
the formation of (6) was decreased to a 5% overall yield.
Compared to the whole-cell bakers’ yeast reduction,
these results show a major improvement in the yield
and it also allows for the synthesis of optically pure R-
(2). In addition, these data further suggest that the het-
erologous E. coli system is able to focus more of the host
bacteria’s metabolism towards reducing substrates while
1
,29,30
properties.
The organisms overexpressing YOL151w and
YGL039w were chosen for the E. coli whole-cell scale-
up because of their products high enantiomeric excess
(
Table 1). These reactions were investigated in two dif-
ferent growth conditions: A nutrient rich growth media
Table 1. Enantiomeric excess and configuration of the reduction
Table 2. Triplicate average of the whole-cell reaction in Scheme 3 for
products from the four bakers’ yeast reductases compared to bakers’
YOL151w and YGL039w
yeast
a
b
b
Overexpressed gene Final concn of (2) (g/L) % (2)
% (6)
a
Overexpressed gene
ee
Configuration
c
YOL151w
YGL039w
0.60
0.51
1.01
0.96
0.11
60
62
96
94
10
40
38
4
6
90
c
c
YOL151w
YGL039w
YGL157w
YNL331c
99%
97
92
33
98
31
S
d
d
d
R
YOL151w
YGL039w
R
R
S
e
Bakers’ yeast
b
Bakers’ yeast
a
Final concentration is the amount of isolated product per reaction
volume (g/L).
a
b
c
Configuration confirmed by NMR.
Isolated in a 10% yield.
b
c
Percent calculated by GC analysis.
3
[
a]
D
À60.0 (c 1.6, CHCl
3
); lit. [a]
D
À60.5 (c 1.0, CHCl
3
).
LB (1 L, pH 6) shaken for 24 h at 20 °C.
Phosphate buffer (1L, pH 6) shaken for 24 h at 20 °C.
Tap water (100 mL) shaken for 24 h at 20 °C.
d
d
e
[
a]
D
+57.5 (c 1.2, CHCl ).
3

R. J. Hammond et al. / Tetrahedron Letters 48 (2007) 1217–1219
1219
overexpresseda
B.Y. reductase
OH
*
4. Csuk, R.; Glanzer, B. I. Chem. Rev. 1991, 91, 49–96.
O
CN
CN
5. Kaluzna, I. A.; Matsuda, T.; Sewell, A. K.; Stewart, J. D.
J. Am. Chem. Soc. 2004, 126, 12827–12832, and references
cited therein.
1
2b
6
. Rodriguez, S.; Kayser, M. M.; Stewart, J. D. J. Am. Chem.
Soc. 2001, 66, 733–738.
Scheme 3. Whole-cell reduction of (1) using engineered E. coli.
LB medium with 30 ug/mL kanamycin was inoculated with a single
colony of E. coli (containing either YOL151w or YGL039w overex-
pressed gene) and shaken 15 h at 37 °C. This preculture was diluted
a
7. Kaluzna, I. A.; Feske, B. D.; Wittayanan, W.; Ghiviriga,
I.; Stewart, J. D. J. Org. Chem. 2005, 70, 342–345.
8
. Feske, B. D.; Kaluzna, I. A.; Stewart, J. D. J. Org. Chem.
005, 70, 9654–9657.
. Feske, B. D.; Stewart, J. D. Tetrahedron: Asymmetry 2005,
6, 3124–3127.
2
1
:100 into 2 L of the same medium and shaken 2.5 h at 37 °C with a
9
stir rate of 400 rpm. Upon reaching an O.D.600 = 0.6, the cell culture
was cooled to 20 °C, and isopropylthio-b-D-galactoside was added to a
final concentration of 100 lM and shaken for 14 h. The cells were
collected by centrifugation (5000 g for 10 min at 4 °C) and then
resuspended in 1 L of 100 mM phosphate buffer (pH = 7) containing
g/L glucose. The bioconversion was carried out at 20 °C with a stir
rate of 400 rpm. The portions of neat (1) were added in small
increments (50 mg) and monitored by GCMS for conversion.
IR (Neat) 3421, 3032, 2915, 2256, 1454, 1054 cm
1
1
1
1
0. Pandey, R. J.; Fernandes, R. A.; Kumar, P. Tetrahedron
Lett. 2002, 43, 4425–4426.
1. Watanabe, M.; Murata, K.; Ikariya, T. J. Org. Chem.
2
002, 67, 1712–1715.
2. Li, Y.; Li, Z.; Wang, Q.; Tao, F. Org. Biomol. Chem. 2005,
, 2513–2518.
13. Liu, P. N.; Gu, P. M.; Wang, F.; Tu, Y. Q. Org. Lett.
004, 6, 169–172.
4
3
b
À1
1
;
H NMR
2
(
1
1
300 MHz, CDCl ) d: 7.38 (s, 5H), 5.01 (t, 1H, J = 6.0 Hz), 3.23 (br s,
3
13
14. Noriyuki, K.; Makoto, U.; Daisuke, M.; Naoaki, T.;
Yoshihiko, Y. European Patent WO2006046455, 2006.
5. Suto, Y.; Kumagai, N.; Matsunaga, S.; Kanai, M.;
Shibisaki, M. Org. Lett. 2003, 5, 3147–3150.
H), 2.73 (d, 2H, J = 6.0 Hz); C NMR (300 MHz, CDCl
3
) d: 141.1,
29.0, 128.9, 125.6, 117.4, 70.1, 28.0; mass (EI) 147, 129, 107, 79, 77.
1
1
6. Granander, J.; Eriksson, J.; Hilmersson, G. Tetrahedron:
Asymmetry 2006, 17, 2021–2027.
under non-growing conditions versus conditions that
allow bacterial reproduction and growth.
17. Suto, Y.; Tsuji, R.; Kanai, M.; Shibisaki, M. Org. Lett.
005, 7, 3757–3760.
2
1
8. Itoh, T.; Mitsukura, K.; Kanphai, W.; Takagi, Y.;
Kihara, H.; Tsukube, H. J. Org. Chem. 1997, 62, 9165–
3. Conclusions
9
172.
1
2
9. Kamal, A.; Ramesh Khanna, G. B.; Krishnaji, T.; Ramu,
R. Bioorg. Med. Chem. Lett. 2005, 15, 613–615.
0. Pamies, O.; Backvall, J. Adv. Synth. Catal. 2001, 343, 726–
The reduction of b-keto nitriles using whole-cell cataly-
sis has been a difficult task due to the dominating mech-
anism affording alkylated products. By using the
heterologous overexpression system the alkylated prod-
uct (6) is nearly eliminated. In addition, this method
allows for the synthesis of both antipodes of 3-hydro-
xy-3-phenylpropanitrile (2), which demonstrates the
advantage of using a single overexpressed bakers’ yeast
reductase versus bakers’ yeast alone. By using this sys-
tem, a formal biocatalytic synthesis of both antipodes
of fluoxetine (3), atomoxetine (4) and nisoxetine (5)
has been demonstrated.
7
31.
21. Itoh, T.; Takagi, Y. Chem. Lett. 1989, 1505–1506.
22. Itoh, T.; Takagi, Y.; Fujisawa, T. Tetrahedron Lett. 1989,
30, 3811–3812.
2
2
2
2
2
2
3. Itoh, T.; Fukuda, T.; Fujisawa, T. Chem. Soc. Jpn. 1989,
6
2, 3851–3855.
4. Smallridge, A. J.; Ten, A.; Trewhella, M. Tetrahedron
Lett. 1998, 39, 5121–5124.
5. Florey, P.; Smallridge, A. J.; Ten, A.; Trewhella, M. A.
Org. Lett. 1999, 1, 1879–1880.
6. Delhi, J. R.; Gotor, V. Tetrahedron: Asymmetry 2001, 12,
1
485–1492.
7. Gotor, V.; Delhi, J. R.; Rebolledo, F. J. Chem. Soc.,
Perkin Trans. 1 2000, 307–309.
8. Delhi, J. R.; Gotor, V. Tetrahedron: Asymmetry 2000, 11,
3693–3700.
References and notes
1
. Robertson, D. W.; Jones, N. D.; Swartzendruber, J. K.;
Yang, K. S.; Wong, D. T. J. Med. Chem. 1988, 31, 185–
29. Robertson, D. W.; Krushinski, J. H.; Fuller, R. W.;
Leander, J. D. J. Med. Chem. 1988, 31, 1412–1417.
30. Wong, D. T.; Fuler, R. W.; Robertson, D. W. Acta
Pharm. Nord. 1990, 2, 171–181.
31. Seco, J. M.; Quinoa, E.; Ruguera, R. Chem. Rev. 2004,
104, 17–117.
1
89.
. Ives, J. L.; Heym, J. Ann. Rep. Med. Chem. 1989, 24, 21–
9.
2
3
2
. Kamal, A.; Ramesh Khanna, G. B.; Ramu, R. Tetrahe-
dron: Asymmetry 2002, 13, 2039–2051.