A highly enantioselective conjugate reduction of 3-arylinden-1-ones using bakers' yeast for the preparation of (S)-3-arylindan-1-ones.

Article abstract of DOI:10.1021/ol991111+

[formula: see text] The bakers' yeast reduction of 3-(1,3-benzodioxol-5-yl)-6-propoxy-1H-inden-1-one 4 has been shown to give (S)-3-(1,3-benzodioxol-5-yl)-2,3-dihydro-6-propoxy-1H-indan-1-one 6 in 65% yield with high enantioselectivity (> 99.0% ee), a key

Full text of DOI:10.1021/ol991111+

ORGANIC
LETTERS
1
999
Vol. 1, No. 11
839-1842
A Highly Enantioselective Conjugate
Reduction of 3-Arylinden-1-ones Using
Bakers’ Yeast for the Preparation of
1
(
S)-3-Arylindan-1-ones†
William M. Clark,* Andrew J. Kassick, Michael A. Plotkin, Ann M. Eldridge, and
Ivan Lantos
Synthetic Chemistry Department, SmithKline Beecham Pharmaceuticals,
709 Swedeland Road, P.O. Box 1539, King of Prussia, PennsylVania 19406
william_m_clark@sbphrd.com
Received October 1, 1999
ABSTRACT
The bakers’ yeast reduction of 3-(1,3-benzodioxol-5-yl)-6-propoxy-1H-inden-1-one 4 has been shown to give (S)-3-(1,3-benzodioxol-5-yl)-2,3-
dihydro-6-propoxy-1H-indan-1-one 6 in 65% yield with high enantioselectivity (>99.0% ee), a key intermediate for the synthesis of the endothelin
receptor antagonist SB 217242. In addition, the substituted 3-arylinden-1-ones 10a−e gave equally high enantioselectivity for the 3-arylindan-
1-one products 13a−e. Mechanistic studies of the reaction indicate the operative pathway to be an asymmetric conjugate reduction, wherein
the hydride transfer from NAD(P)H occurs from the Re-face of the indenone substrate.
The reduction of ketones or aldehydes employing microor-
ganisms has been an area of active interest, and in this regard,
bakers’ yeast (Saccharomyces cereVisiea) has been the
favored microbe on the basis of its ready availability, safety,
Scheme 1
1
and low cost. Few investigations have been reported on the
extension of these studies to R,â-unsaturated carbonyl
compounds, and in this area some of the results have been
2
controversial. For example, Fuganti carried out a pioneering
study on â-phenyl-substituted enones using bakers’ yeast
3
which resulted in the isolation of the (S)-allylic alcohol and
only minor amounts of the saturated ketone. In subsequent
work, they studied the reduction of 3-methyl-4-phenyl-buten-
observed in 15% yield (Scheme 1). Upon reexamination of
4
5
the same reaction, first by Sakai and later by Kawai, each
found that the reduction takes place at the R,â-unsaturated
linkage, affording mostly the saturated ketone (S)-3. In
2-one 1, and again the formation of allylic alcohol (S)-2 was
†
This paper is dedicated to the memory of Prof. Jack T. Edward of
(3) Fuganti, C.; Grasselli, P.; Servi, S.; Spreafico, F.; Zirotti, C.; Casati,
P. J. Chem. Res. Synop. 1985, 22.
(4) Sakai, T.; Matsumoto, S.; Hidaka, S.; Imajo, N.; Tsuboi, S.; Utaka,
M. Bull. Chem. Soc. Jpn. 1991, 64, 3473.
(5) Kawai, Y.; Saitou, K.; Hida, K.; Hai-Dao, H.; Ohno, A. Bull. Chem.
Soc. Jpn. 1996, 69, 2633.
McGill University.
1) (a) Servi, S. Synthesis 1990, 1. (b) Csuk, R.; Glanzer, B. I. Chem.
ReV. 1991, 91, 49.
2) Fuganti, C.; Grasselli, P.; Servi, S.; Spreafico, F.; Zirotti, C.; Casati,
P. J. Chem. Res. Synop. 1984, 112.
(
(
1
0.1021/ol991111+ CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/02/1999

particular, Kawai has shown that a reductase isolated from
(25%) and further evaluation of the reaction mixture showed
the remaining mass balance to be that of the starting enone.
Noting a significant drop in the pH (pH 7 f pH 4) during
the course of the reaction, the reduction was repeated using
a pH probe and automated titrant apparatus for the addition
of 1 M NaOH. With the yeast mixture maintained at pH
7.2, the reduction afforded complete conversion of 4 after
24 h at 40 °C (Scheme 3). After purification by flash
bakers’ yeast gives the saturated ketone (S)-3 with good
6
enantioselectivity and yield (80% ee, 71%). In addition,
Kawai showed through labeling studies that delivery of
hydride at the â-position was derived from NAD(P)H with
the hydrogen at the R-position presumably coming from
water or a proton source within the reductase enzyme. Kawai
also observed that the overall addition of hydrogen to the
enone olefin occurred in a trans stereochemical fashion.
In our work, we required the chiral indanone (S)-6 in
connection with the synthesis of endothelin receptor antago-
nist SB 217242 currently in clinical development at Smith-
Scheme 3
7
Kline Beecham. While the compound could be readily
prepared by a catalytic hydride reduction using the Itsuno-
Corey oxazoborolidine method to (S)-5 followed by a formal
1
,3-hydride rearrangement to yield (S)-6,8 we became
interested in the possibility of applying bakers’ yeast as the
reductant for this transformation (Scheme 2). It is well
chromatography, the indanone (S)-6 was delivered in 65%
yield with >99.0% ee! In comparison, the best enantiose-
Scheme 2
lectivity we observed for the oxazoborolidine approach was
8
9
4% ee for (S)-6.
Due to the excellent results obtained for the bakers’ yeast
reduction of 4, we decided to investigate the bakers’ yeast
reduction of other indenone substrates. Subsequently, two
approaches were developed for the preparation of variously
substituted 3-arylindenones (Scheme 4). In our initial ap-
Scheme 4
documented that bakers’ yeast reductions of prochiral ketones
typically deliver the (S)-alcohol often with high enantiose-
lectivity, the absolute stereochemistry we desired for indenol
5. In addition, we felt substrate 4 would offer an opportunity
to investigate the bakers’ yeast reduction of 3-arylinden-1-
ones, which to date have not been studied in the literature
in this manner, and also potentially address the issue of a
9
1,2- versus 1,4-reduction in this class of compounds.
Initially we attempted the reduction of indenone 4 by
dissolving the substrate (1.25 g) in EtOH (12.5 mL) and
adding the solution to a mixture of bakers’ yeast (6.0 g) and
glucose (5.0 g) in deionized water (250 mL). The suspension
was then stirred vigorously for 2 days at 40 °C with no
control of pH. Interestingly, the yeast reduction afforded the
indanone (S)-6 with excellent enantioselectivity (>99.0% ee)
and the desired absolute stereochemistry but not the expected
alcohol product (S)-5. In addition, the yield was fairly low
(
6) Kawai, Y.; Hayashi, M.; Inaba, Y.; Saitou, K.; Ohno, A. Tetrahedron
Lett. 1998, 39, 5225.
7) Ohlstein, E. H.; Nambi, P.; Amparo, H.; Douglass, W. P.; Beck, G.;
(
Fong, K.-L.; Eddy, E. P.; Smith, P.; Ellens, H.; Elliot, J. D. J. Pharmacol.
Exp. Ther. 1996, 37(2), 609.
proach, compounds 10a-e were prepared by a Suzuki cross-
coupling reaction of the respective pinacol boronate esters
(
8) Clark, W. M.; Tickner-Eldridge, A. M.; Huang, G. K.; Pridgen, L.
N.; Olsen, M. A.; Mills, R. J.; Lantos, I.; Baine, N. H. J. Am. Chem. Soc.
998, 120, 4550.
9) This work was disclosed in part at Chiral USA 99, San Francisco,
CA, May 2, 1999.
1
0
9a-c with the 3-bromoinden-1-ones 7 and 8 which were
1
11
prepared using conditions developed by Joullie. Initially,
(
the Suzuki coupling proved to be difficult due to the base
1840
Org. Lett., Vol. 1, No. 11, 1999

sensitivity of the bromide at elevated temperatures. Eventu-
ally we determined that the cross-coupling reaction pro-
did not adversely affect the yield or enantioselectivity of the
reduction, i.e., 13a and 13d were isolated in 81% and 84%
1
6
gressed well when using aqueous NaHCO
PPh in DME at 50 °C to give yields of the 3-arylindenones
in the 40-70% range.
Attempts to prepare 3-arylindenones with electron-
withdrawing groups (Cl, Br, NO ) at the C(5)- or C(6)-
3
with Pd
2
(dba)
3
/
yields, respectively.
3
Realizing that an asymmetric conjugate reduction of
indenone 4 was a viable reaction pathway, we wanted to
determine the potential for rearrangement of (S)-5 when
subjected to the yeast reduction conditions. We reasoned that
if a fast 1,3-hydrogen rearrangement occurred under these
conditions, then only very low levels of (S)-5 would be
observed during the course of the reaction. On the other hand,
if indenol (S)-5 was stable to the yeast reaction conditions,
then a conjugate hydride addition of the indenone system
was most likely the operative mechanism.
To this end, the indenol (S)-5 (94% ee) was subjected to
the yeast reduction conditions at 40 °C. After 24 h, the
indanone (S)-6 was not observed by HPLC or TLC. In
addition, (S)-5 was completely recovered and the enantio-
meric purity of (S)-5 remained unchanged. In our subsequent
experiment, racemic 5 was added to the yeast reduction of
indenone 4 (5% added relative to 4). After 20 h, complete
consumption of 4 was achieved and the indanone (S)-6
12,13
2
position of the 3-bromoindenone using the Suzuki method-
ology failed to give appreciable amounts of the desired
product. In the examples investigated, the base and thermal
instability of the bromide and indenone product appeared to
be responsible for the poor results. Similar results were
observed when attempting to couple 2-furanboronic acid and
2-thiopheneboronic acid with bromides 7 and 8. Conse-
quently, the remaining indenone substrates 12a,b were
prepared using a previously reported procedure developed
in our laboratories.8
The enantioselective conjugate reduction of 3-aryl-inden-
1
-ones using bakers’ yeast appears to be a general transfor-
14
mation for this class of compounds (Scheme 5). Using our
Scheme 5
(12) General Procedure for the Preparation of 3-Arylinden-1-ones
10a-e): To a thoroughly degassed round-bottomed flask charged with
(
DME (8.5 mL) were added triphenylphosphine (0.05 g, 0.189 mmol) and
tris(dibenzylideneacetone)dipalladium (0.04 g, 0.042 mmol), and the
resulting yellow solution was stirred under nitrogen for 1.5 h at room
temperature. To this solution were added the â-bromoindenone (1.0 g, 4.2
mmol), aryl pinacol borate ester (4.6 mmol, 1.1 equiv), and a saturated
NaHCO3 solution (2.6 mL), and the mixture was heated to 50 °C for 1 h.
The reaction mixture was then cooled to 25 °C, diluted with H2O (25 mL),
and extracted with CH2Cl2 (50 mL). The organic layer was washed with a
saturated NaCl solution (25 mL), dried (MgSO4), filtered, and concentrated
under reduced pressure to furnish 10a-e after flash chromatography (SiO2,
1
:9 EtOAc/hexane).
1
(13) Data for 10d: mp 77-79 °C; H NMR (CDCl3) δ 7.46-7.25 (5
H, m), 6.88 (1H, d, J ) 2.2 Hz), 6.65 (1H, dd, J ) 2.2, 8.0 Hz), 5.96 (1H,
s), 3.82 (3H, s), 2.41 (3H, s); 13C NMR (300 MHz, CDCl3) δ 195.72, 163.88,
1
1
60.78, 146.64, 138.70, 132.99, 131.10, 128.80, 127.93, 125.13, 124.54,
24.47, 124.39, 111.22, 110.23, 55.72, 21.43. Anal. Calcd for C17H14O2:
C, 81.58; H, 5.64. Found: C, 81.57; H, 5.63.
14) General Procedure for the Preparation of 3-Arylindan-1-ones
(
Using Bakers’ Yeast (13a-g): To a 250 mL round-bottomed flask charged
with an aqueous phosphate buffer solution (70 mL, pH 7.2) warmed to 40
°C were added bakers’ yeast (6.43 g) and R-D-glucose (5.14 g, 28.5 mmol).
The 3-arylinden-1-one (1.0 g) was then dissolved in a minimum amount of
hot EtOH (4-6 mL) and added to the reaction mixture. The reaction was
stirred for 24 h, and a neutral pH was maintained through the addition of
1 mL aliquots of 1 M NaOH (10-15 mL). After 24 h, the reaction mixture
was cooled to room temperature, EtOH (100 mL) added, the mixture filtered
through Celite to remove the yeast, and the filtrate extracted with 2:1 EtOAc/
acetone (4 × 150 mL). The organic layer was washed with H2O (200 mL)
and a saturated NaCl solution (200 mL), dried (MgSO4), filtered, and
concentrated under reduced pressure to furnish 13a-g after flash chroma-
tography (SiO2, 1:9 EtOAc/hexane).
standard conditions, all of our substrates gave high enanti-
oselectivities (>99.0% ee) and good to excellent yields (50-
8
4%) for the indanone products (13a-g).15 Interestingly,
changing the placement of the electron-donating groups on
either aryl ring had little effect on the reaction outcome. For
example, a significant electronic influence on enantioselec-
tivity has been observed in numerous bakers’ yeast reductions
of prochiral substrates;1 however, placement of a methoxy
group at the C(5)- or C(6)-position did not change the
stereoselectivity (13f, 13g), nor did changing the substituents
on the C(3)-phenyl ring.
(
15) The enantiomeric purity of 13a-g was determined on a Chiralpak
OD column, 9:1 hexane/IPA, 1 mL/min, 230 nm.
16) Data for 13c: mp 115-116 °C; [R]D) -14.4°(c ) 0.500, CHCl3);
H NMR (CDCl3) δ 7.73 (1H, d, J ) 7.8 Hz), 7.17 (1H, d, J ) 7.6 Hz),
.05 (1H, d, J ) 7.6 Hz), 6.92 (3H, m), 6.65 (1H, d, J ) 1.8 Hz), 4.45
1H, dd, J ) 3.7, 8.0 Hz), 3.77 (3H, s), 3.17 (1H, dd, J ) 8.0, 19.0 Hz),
(
,6
1
7
(
2.64 (1H, dd, J ) 3.8, 19.0 Hz), 2.30 (3H, s); 13C NMR (CDCl ) δ 204.06,
3
1
1
65.59, 160.97, 143.64, 138.57, 130.23, 128.77, 128.29, 127.71, 125.04,
24.69, 115.93, 109.84, 55.63, 47.10, 44.42, 21.39. Anal. Cald for
A steric influence at the C(3)-position of the indenone
substrate was also not observed. For example, a methoxy
group in the ortho or meta positions of the C(3)-phenyl ring
C17H16O2: C, 80.93; H, 6.39. Found: C, 80.66; H, 6.42. Data for 13g:
1
mp 91-93 °C; [R]D ) -26.7°(c ) 0.375, CHCl3); H NMR (CDCl3) δ
7
.72 (1 H, d, J ) 8.5 Hz), 7.03 (2 H, d, J ) 8.7 Hz), 6.92 (1 H, d, J ) 8.5
Hz), 6.83 (2 H, d, J ) 8.7 Hz), 6.63 (1 H, s), 4.44 (1 H, dd, J ) 3.7, 8.0
Hz), 3.78 (3 H, s), 3.77 (3 H, s), 3.17 (1 H, dd, J ) 8.1, 19.0 Hz), 2.61 (1
H, dd, J ) 3.7, 19.0 Hz); C NMR (CDCl3) δ 204.06, 165.58, 161.18,
158.58, 135.74, 130.17, 128.57, 124.99, 115.91, 114.30, 109.72, 55.60,
55.26, 47.23, 43.69. Anal. Cald for C17H16O3: C, 76.1; H, 6.01. Found:
C, 75.83; H, 6.19.
13
(
10) The pinacol boronate esters were prepared according to the procedure
described by Tour: Lamba, J. J. S.; Tour, J. M. J. Am. Chem. Soc. 1994,
16, 11723.
11) Heffner, R. J.; Joullie, M. M. Synth. Commun. 1991, 21, 2231.
1
(
Org. Lett., Vol. 1, No. 11, 1999
1841

obtained with >99.0% ee along with complete recovery of
racemic 5. Both of these experiments strongly suggest that
indenol (S)-5 is not an intermediate in the yeast reduction
of 4. Further confirmation was obtained when racemic 5 was
treated with a phosphate buffer (pH 8.0) in EtOH and heated
to 40 °C for 24 h. The indanone 6 was not observed by
HPLC, confirming that neutral pH (or slightly higher) was
not basic enough to promote the rearrangement of (S)-5 to
(
S)-6.
Figure 1.
From these experiments, we conclude the yeast reduction
of 4 is a highly enantioselective conjugate reduction of the
R,â-unsaturated indenone system. Reductases or dehydro-
genases isolated from bakers’ yeast have been shown to
a bakers’ yeast reduction. The method is an attractive
alternative to the previously published procedure developed
for the synthesis of the endothelin receptor antagonist SB
217242. In addition, the biotransformation is an effective
method for the preparation of enantiomerically pure electron-
rich (S)-3-arylindan-1-ones. Our current efforts are now
focused on preparing 3-alkylinden-1-ones and studying their
behavior under similar reaction conditions.
5,6,17
reduce prochiral ketones with high enantioselectivity.
The
reductase enzyme requires a cofactor, NAD(P)H, which is
the actual hydride source. In our simplified model, we
propose the reductase enzyme complexes with the Si face
of the indenone substrate allowing for Re face attack of
NAD(P)H (Figure 1). We assume the overall hydrogen
addition occurs with trans stereochemistry as observed by
Kawai.6
Acknowledgment. We thank the Analytical Sciences
department of SmithKline Beecham for the analytical support
of this research and SB for the financial support of M.P.
and A.K.
In summary, a highly enantioselective transformation for
the synthesis of indanone (S)-6 has been accomplished using
(17) Faber, K. Biotransformations in Organic Chemistry; Springer-
Verlag: Berlin, 1997; pp 166-175.
OL991111+
1842
Org. Lett., Vol. 1, No. 11, 1999