Enzymatic reductions for the regio-and stereoselective synthesis of hydroxy-keto esters and dihydroxy esters

Article abstract of DOI:10.1021/ol3003833

Ketoreductases were utilized for the stereoselective synthesis of δ-hydroxy-β-keto esters, β-hydroxy-δ-keto esters, and β,δ-dihydroxy esters. Seven out of eight possible stereoisomers were obtained from the enzymatic reduction of the corresponding β,δ-diketo ester in high enantio-and diastereomeric excess.

Full text of DOI:10.1021/ol3003833

ORGANIC
LETTERS
2012
Vol. 14, No. 7
1792–1795
Enzymatic Reductions for the Regio- and
Stereoselective Synthesis of Hydroxy-
keto Esters and Dihydroxy Esters
Anna Bariotaki, Dimitris Kalaitzakis, and Ioulia Smonou*
Department of Chemistry, University of Crete, Heraklion 71003, Crete, Greece
smonou@chemistry.uoc.gr
Received February 16, 2012
ABSTRACT
Ketoreductases were utilized for the stereoselective synthesis of δ-hydroxy-β-keto esters, β-hydroxy-δ-keto esters, and β,δ-dihydroxy esters.
Seven out of eight possible stereoisomers were obtained from the enzymatic reduction of the corresponding β,δ-diketo ester in high enantio- and
diastereomeric excess.
Optically active δ-hydroxy-β-keto esters, β-hydroxy-δ-
keto esters and β,δ-dihydroxy esters are very useful inter-
mediates in asymmetric organic synthesis and have been used
for the synthesis of many natural products, pharmaceuticals,
and other high value chemical compounds.1 Although sev-
eral methods have been developed for their synthesis such as
chemical,2 chemoenzymatic,3 and enzymatic,4 there is not a
simple and straightforward method for the stereoselective
synthesis of all possible stereoisomers.
In recent years, the challenge to simplify a multistep
synthetic procedure to one step and if possible to combine
consecutive catalytic transformations in the same reaction
vessel has attracted interest in academic research and in the
chemical industry.5,6 In general, enzymes can facilitate an
otherwise complicated conventional chemical route, since
they offer great opportunities to control the enantioselec-
tivity of the products and to accomplish domino reactions
in the same vessel.7
Our continuing interest in enzymatic reductions for the
stereoselective synthesis of various chiral synthons8 led us
to present herein a straightforward enzymatic reduction
of tert-butyl 3,5-dioxohexanoate 1. This simple one-step
(1) (a) Patel, R. N. Adv. Synth. Catal. 2001, 343, 527. (b) Kumar,
R. N.; Meshram, H. M. Tetrahedron Lett. 2011, 52, 1003. (c)
Pfefferkorn, J. A.; Bowles, D. M.; Kissel, W.; Boyles, D. C.; Choi, C.;
Larsen, S. D.; Song, Y.; Sun, K.-L.; Miller, S. R.; Trivedi, B. K.
Tetrahedron 2007, 63, 8124. (d) Yeung, K.-S.; Paterson, I. Chem. Rev.
2005, 105, 4237. (e) Nakamura, R.; Tanino, K.; Miyashita, M. Org. Lett.
2005, 7, 2929. (f) Enders, D.; Vicario, J. L.; Job, A.; Wolberg, M.;
€
Muller, M. Chem.;Eur. J. 2002, 8, 4272. (g) Kiyooka, S.; Hena, M. A.
J. Org. Chem. 1999, 64, 5511. (h) Dvorak, C. A.; Schmitz, W. D.; Poon,
D. J.; Pryde, D. C.; Lawson, J. P.; Amos, R. A.; Meyers, A. I. Angew.
Chem., Int. Ed. 2000, 39, 1664. (i) Baker, D. D.; Chu, M.; Oza, U.;
Rajgarhia, V. Nat. Prod. Rep. 2007, 24, 1225. (j) Hamada, Y.; Shioiri, T.
Chem. Rev. 2005, 105, 4441. (k) Kobayashi, J.; Tsuda, M. Phytochem.
Rev. 2004, 3, 267. (l) Grkovic, T.; Blees, J. S.; Colburn, N. H.; Schmid,
T.; Thomas, C. L.; Henrich, C. J.; McMahon, J. B.; Gustafson, K. R.
J. Nat. Prod. 2011, 74, 1015.
(2) (a) Zacharia, J. T.; Tanaka, T.; Hayashi, M. J. Org. Chem. 2010,
75, 7514. (b) Xu, Q.; Yu, J.; Han, F.; Hua, J.; Chen, W.; Yang, L.
Tetrahedron: Asymmetry 2010, 21, 156. (c) Kumar, P.; Pandey, M.;
Gupta, P.; Dhavale, D. D. Tetrahedron Lett. 2010, 51, 5838. (d) Blandin,
V.; Carpentier, J.-F.; Mortreux, A. Eur. J. Org. Chem. 1999, 3421. (e)
Solladie, G.; Bauder, C.; Rossi, L. J. Org. Chem. 1995, 60, 7774.
(3) (a) Wolberg, M.; Filho, M. V.; Bode, S.; Geilenkirchen, P.;
(5) Andrews, I.; Dunn, P.; Hayler, J.; Hinkley, B.; Hughes, D.;
Kaptein, B.; Lorenz, K.; Mathew, S.; Rammelloo, T.; Wang, L.; Wells,
A.; White, T. D. Org. Process Res. Dev. 2011, 15, 22.
€
Feldmann, R.; Liese, A.; Hummel, W.; Muller, M. Bioprocess Biosyst.
(6) Tietze, L. F.; Brasche, C.; Gericke, K. M. Domino Reactions in
Organic Synthesis; Wiley-VCH: Weinheim, 2006; Chapter 8, p 529.
(7) (a) Hollmann, F.; Arends, I. W. C. E.; Holtmann, D. Green Chem.
2011, 13, 2285. (b) Schrittwieser, J. H.; Sattler, J.; Resch, V.; Mutti, F. G.;
Kroutil, W. Curr. Opin. Chem. Biol. 2011, 15, 249. (c) Rajagopalan, A.;
Kroutil, W. Mater. Today 2011, 14, 144. (d) Hall, M.; Bommarius, A. S.
Chem. Rev. 2011, 111, 4088. (e) Garcıa-Urdiales, E.; Alfonso, I.; Gotor, V.
Chem. Rev. 2011, 111, PR110.
Eng. 2008, 31, 183. (b) Sun, F.; Xu, G.; Wu, J.; Yang, L. Tetrahedron:
Asymmetry 2007, 18, 2454. (c) Ohrlein, R.; Baisch, G. Adv. Synth. Catal.
€
2003, 345, 713.
(4) (a) Wu, X.; Wang, L.; Wang, S.; Chen, Y. Amino Acids 2010, 39,
305. (b) Liljeblad, A.; Kallinen, A.; Kanerva, L. T. Curr. Org. Synth.
€
€
2009, 6, 362. (c) Ludeke, S.; Richter, M.; Muller, M. Adv. Synth. Catal.
2009, 351, 253.
r
10.1021/ol3003833
Published on Web 03/12/2012
2012 American Chemical Society

reaction proceeded in a totally regio- and stereoselective
manner, affording in excellent yields optically and chemi-
cally pure δ-hydroxy-β-keto esters 2, β-hydroxy-δ-keto
esters 3, or β,δ-dihydroxy esters 4, Figure 1.
product. The enantiomers of choice 2a and 2b were formed
in a relatively short time, 6 and 3 h respectively, free of any
byproducts. The absolute configuration of these two en-
antiomers was determined utilizing chiral derivatizing
agents.9 It should be emphasized here that the reduction
catalyzed by Kred-118 took place mostly at the β-carbonyl
group of 1 affording 3a as the major product (92%). This
important pathway led to the optically pure β-hydroxy-δ-
keto ester 3a, a useful chiral synthon and building block,
difficult to be synthesized otherwise.10 Along with major
product 3a, a small amount of dihydroxyester 4b(8%) was
also identified in high enantiomeric and diastereomeric
excess (Figure 3).
Figure 1. Enzymatic reduction of tert-butyl 3,5-dioxohexanoate 1.
At this point, to investigate the selectivity of Kred-118,
we tested the enzymatic reduction of optically pure δ-
hydroxy-β-keto ester 2a, synthesized earlier (Figure 3).
To our delight, this reduction afforded dihydroxy ester
4b in optically pure form (Figure 4). Furthermore, in a
control experiment we also found that attempts to reduce
3a (separated by flash column chromatography from the
3a and 4b mixture) utilizing the same enzyme Kred-118
failed (Figure 4). This result indicates that the reduction of
diketo ester 1 by Kred-118 initially leads to the formation
of 3a and 2a. A further reduction of 2a leads to the
observed minor product 4b.
The eight possible products of this reduction are the two
enantiomers of δ-hydroxy-β-keto ester 2a,b, two enantiomers
of β-hydroxy-δ-keto ester 3a,b, and four steroisomers of β,δ-
dihydroxy ester 4aꢀd, as shown in Figure 2. Our challenging
goal was the one-step preparation of every single stereo-
isomer out of these eight products, in optically pure form.
Figure 4. Pathway for the formation of syn-dihydroxy ester 4b,
catalyzed by Kred-118.
Figure 2. All possible products from the keto reduction of tert-
butyl 3,5-dioxohexanoate (1).
After this unexpected and significant regio- and enan-
tioselectivity of Kred-118, attempts to improve the yield of
the β-hydroxy-δ-keto ester3aby minimizing the formation
of product 4b were successful. Specifically, we combined
the enzymatic reduction of diketo ester 1 by Kred-118 with
an enzymatic oxidation utilizing Kred-104 in the same
reaction vessel. As shown before (Figure 3), Kred-104
catalyzes the reduction only at the δ-carbonyl group of 1.
Consequently, the reverse oxidation procedure is expected
to convert dihydroxy ester 4b to 3a. Indeed, after comple-
tion of the reduction of 1 by Kred-118, an oxidation step
was followed in the same vessel utilizing Kred-104 and
NADPþ (oxidant) converting dihydroxy ester 4b into the
desired product 3a (Figure 5). An excess of 2-pentanone
To this purpose, we explored the enzymatic reduction of
only one carbonyl group in order to produce either the δ-
hydroxy-β-keto ester 2a and 2b or the β-hydroxy-δ-keto ester
3a and 3b. A series of 23 enzymes were screened for this
transformation, and the most representative results are
shown in Figure 3.
(8) (a) Kalaitzakis, D.; Smonou, I. J. Org. Chem. 2010, 75, 8658. (b)
Kalaitzakis, D.; Smonou, I. Tetrahedron 2010, 66, 9431. (c) Kalaitzakis,
D.; Rozzell, D. J.; Kambourakis, S.; Smonou, I. Adv. Synth. Catal. 2006,
348, 1958. (d) Kalaitzakis, D.; Rozzell, D. J.; Kambourakis, S.; Smonou,
I. Org. Lett. 2005, 7, 4799.
Figure 3. Enzymatic reduction of diketo ester 1 catalyzed by the
enzymes Kred-104, Kred-107, and Kred-118.
(9) Seco, J. M.; Quinoa, E.; Riguera, R. Chem. Rev. 2004, 104, 17.
€
€
(10) (a) Mulzer, J.; Altmann, K.-H.; Hofle, G.; Muller, R.; Prantz,
K. C. R. Chimie 2008, 11, 1336. (b) Altmann, K.-H.; Bold, G.; Caravatti,
€
G.; Florsheimer, A.; Guagnano, V.; Wartmann, M. Bioorg. Med. Chem.
Lett. 2000, 10, 2765. (c) Nicolaou, K. C.; Ninkovic, S.; Sarabia, F.;
Vourloumis, O.; He, Y.; Vallberg, H.; Finlay, M. R. V.; Yang, Z. J. Am.
Chem. Soc. 1997, 119, 7974. (d) Zheng, Y.; Avery, M. A. Tetrahedron
2004, 60, 2091.
As shown in Figure 3, Kred-104 and Kred-107 catalyzed
the reduction of substrate 1, affording the S or R enantio-
mer of hydroxy ketoesters 2a,b respectively as the only
Org. Lett., Vol. 14, No. 7, 2012
1793

was used for NADPþ recycling. Specifically, after the first
reduction step, the reaction was left for 24 h. This period of
time is needed for NADPH and Kred-118 to be inactive.
Next, we used 2-pentanone for regeneration of NADPþ
since we have already demonstrated that this compound is
a good substrate for Kred-104.11 Thus, this novel cascade
enzymatic transformation led to the preparation of β-
hydroxy-δ-keto ester 3a as the only product, with high
chemical and optical purity (Figure 5).
products were isolated without any chromatographic puri-
fication. The relative and absolute configuration of 4aꢀ4d
were assigned by comparison with literature data.1e
To this end, having accomplished the preparation of
4aꢀ4d, startingfrom the optically pure hydroxyketo esters
2a and 2b, we investigated the proper combination of
specific biocatalysts for the one-pot synthesis of all stereo-
isomers of dihydroxy esters 4aꢀ4d starting from diketo
ester 1. Indeed, we managed to accomplish the one-pot
synthesis of all these compounds, utilizing two sequential
reduction steps, in the same vessel, without the isolation of
the intermediates. The results are summarized in Figure 7.
This figure shows that the reduction of tert-butyl 3,5-
dioxohexanoate 1 by Kred-107 followed by Kred-B1F
led to stereoisomer (3R,5R)-4a. The enantiomer (3R,5S)-
4c was produced by the reduction of substrate 1 by Kred-
104, followed by Kred-B1F. The sequential reduction of
substrate 1, by Kred-107 and Kred-118, afforded (3S,5R)-
4d. Finally, the stereoisomer (3S,5S)-4b was prepared
utilizing Kred-104 and Kred-118 simultaneously in the
same vessel. It should be emphasized here that the simul-
taneous reduction by both enzymes led to 4b because both
intermediate products 2a and 3a were reduced by these
enzymes leading to the same product 4b. In all these cases
the dihydroxy esters 4aꢀ4d were synthesized in high
optical and chemical purity without any chromatographic
purification process.
Figure 5. Preparation of β-hydroxy-δ-keto ester 3a in >98%
chemical purity and 82% isolated yield.
Furthermore, the preparation of optically pure dihy-
droxy esters was accomplished by the enzymatic reduction
of the optically pure hydroxy keto esters 2a, 2b, and 3a,
which were previously synthesized enzymatically
(Figure 3). These reactions are shown in Figure 6.
Figure 6. Enzymatic synthesis of optically pure dihydroxy esters
4aꢀ4d.
As shown in Figure 6, ketoreductases were active toward
the reduction of 2a, 2b, and 3a producing all possible
stereoisomers of the product 4aꢀ4d. In particular, Kred-
118 catalyzed the reduction of both substrates 2a and 2b,
producing in high optical purity the (3S,5S)-4b and
(3S,5R)-4d enantiomer respectively. Furthermore, we were
able to synthesize stereoselectively the enantiomers
(3R,5S)-4c and (3R,5R)-4a utilizing Kred-B1F, which
catalyzed the reduction of both substrates 2a and 2b.
Finally, β-hydroxy-δ-keto ester 3a was reduced stereose-
lectively to compound (3S,5S)-4b by Kred-104. All the
above enzymatic reactions were completed at 24 h, and the
Figure 7. Enzymatic one-pot synthesis of all the stereoisomers of
tert-butyl 3,5-dihydroxyhexanoate (dihydroxy esters 4aꢀ4d).
Furthermore, an interesting result was obtained from
the reduction of diketo ester 1 catalyzed by Kred-B1F
(Figure 8). In this case, the stereoisomer (3R,5S)-4c was
formed directly, indicating that only one enzyme, Kred-
B1F stereoselectively catalyzed the reduction of both β and
(11) Kallergi, M.; Kalaitzakis, D.; Smonou, I. Eur. J. Org. Chem.
2011, 3946.
1794
Org. Lett., Vol. 14, No. 7, 2012

δ carbonyl groups. This is a very impressive result because of
the direct formation of the desired stereoisomer in high
optical purity and chemical yield, utilizing only one enzyme.
Figure 8. Direct formation of anti dihydroxy ester 4c by enzy-
matic reduction of 1 with Kred-B1F.
Figure 10. Kred-B1F mediated reduction of compounds 2a, 2b,
and 3a.
It is also interesting to note that this reduction afforded
the anti diastereomer indicating that the enzyme shows an
opposite selectivity to the β-carbonyl group compared to
that of the δ-carbonyl group. In order to determine the
course of this reaction, we managed to identify the inter-
mediate products by terminating the reduction at an early
stage (1 and 2 h respectively). As shown in Figure 9, there
were two intermediate mixtures consisting of a δ-hydroxy-
β-keto ester 2, a β-hydroxy-δ-keto ester 3, and the corre-
sponding dihydroxy ester 4. These results indicate that, in
the initial step, the enzyme did not exhibit any regioselec-
tivity toward the reduction of β- or δ-carbonyl groups of
substrate 1. However, this enzyme showed high stereose-
lectivity toward the formation of only one final product 4c
after 24 h of reaction.
intermediates 2 and 3 as 5S and 3R respectively, we were
able to assign with confidence the absolute configuration
of dihydroxy ester 4c(Figure11) asthe antidihydroxyester
(3R,5S).
Figure 11. Route for the enzymatic formation of 4c by Kred-B1F.
In summary, NADPH dependent ketoreductases were
used as catalysts for the stereoselective synthesis of δ-
hydroxy-β-keto esters, β-hydroxy-δ-keto ester, and β,δ-
dihydroxy esters. This method proved very efficient for the
preparation of seven out of the eight stereoisomers of these
highly valuable chiral molecules depending on the choice
of the enzyme. Cascade enzymatic reactions were per-
formed for the one-pot synthesis of optically pure β,δ-
dihydroxy esters by avoiding time-, effort-, and solvent-
intensive steps starting from the same diketo ester 1. This
green procedure afforded these stereoisomers in high
stereoselectivity, with a low reaction time and without
any chromatographic purification of the isolated product.
Figure 9. Investigation of the reduction of diketo ester 1 cata-
lyzed by Kred-B1F.
Acknowledgment. This work was supported by the proj-
ect “IRAKLITOS II - University of Crete” of the Opera-
tional Programme for Education and Lifelong Learning
2007ꢀ2013 (E.P.E.D.V.M.) of the NSRF (2007ꢀ2013),
which is cofunded by the European Union (European Social
Fund) and National Resources. We are also thankful to the
PROFI (Proteomics Facility at IMBB-FORTH) for per-
forming all the HRMS analyses.
To determine the absolute configuration of intermedi-
ates 2ꢀ4 and the final product 4c, we accomplished enzy-
matic reductions of the optically pure compounds (5S)-2a,
(5R)-2b, and (3S)-3a with Kred-B1F (Figure 10).
These experiments showed that Kred-B1F was active
only toward the reduction of substrates 2a and 2b. More-
over, only 2a afforded the anti-dihydroxy ester (3R,5S)-4c.
This crucial observation led us to conclude that the inter-
mediate δ-hydroxy-β-keto ester (2, Figure 9) was the (5S)-
2a stereoisomer. Furthermore, Kred-B1F was inactive
toward substrate (3S)-3a indicating that the intermediate
β-hydroxy-δ-keto ester observed during the reduction
of 1 (3, Figure 9) was the (3R)-3b enantiomer shown in
Figure 11. Thus, by determining the stereochemistry of
Supporting Information Available. Detailed experimen-
1
tal procedures, HRMS, H and 13C NMR spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 7, 2012
1795