This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Tetrahedron Letters 54 (2013) 1921–1923

Contents lists available at SciVerse ScienceDirect

Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet

Thermally driven asymmetric domino reaction catalyzed by a thermostable

esterase and its variants
Reina Wada a, Takashi Kumon a, Robert Kourist b, Hiromichi Ohta a, Daisuke Uemura a, Shosuke Yoshida a,
Kenji Miyamoto a,⇑
a
Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa 223-8522, Japan
b
Junior Professorship for Microbial Biotechnology, Ruhr Universität Bochum, Universitätstr. 150, 44780 Bochum, Germany

a r t i c l e i n f o a b s t r a c t

Article history: We have developed a thermally driven domino reaction for the synthesis of (S)-a-arylpropionates
Received 14 December 2012 (profens) using a thermostable esterase from Sulfolobus tokodaii strain 7. Stereoselectivity was improved
Revised 15 January 2013 considerably by engineering of the active site. Stereoselective decarboxylation at the active site of an
Accepted 18 January 2013
esterase is a new reaction for the synthesis of optically active carboxylic acids.
Available online 29 January 2013
Crown Copyright ! 2013 Published by Elsevier Ltd. All rights reserved.

Keywords:
Domino reaction
Thermostable esterase
a-Arylpropionates
Thermal spontaneous decarboxylation
Active site

Climate change and the future limitation of energy and re-
sources create a high demand for efficient, sustainable, synthetic
strategies under the green chemistry philosophy.1 One approach
to this is the minimization of resources dedicated to the work-up
and purification of compounds. Domino reactions, where several
bond-forming and bond-breaking steps take place in one pot under
the same reaction conditions, provide a highly efficient strategy.
These reactions improve the synthetic efficiency by decreasing
the number of operations and thus the amount of chemicals and
solvents required. The application of enzymes in multistep reac-
tions permits the use of excellent stereoselectivity and mild reac-
tion conditions usually associated with enzymes.2 Akai et al.
reported the first example of lipase-catalyzed enantioselective
esterification followed by an intramolecular Diels–Alder reaction,
a sequence which gave products with up to 99% enantiomeric ex-
cess (ee).3
We have recently identified a thermostable esterase gene
(ST0071) from the thermophilic archeon Sulfolobus tokodaii by an
in silico screening of the total genome.4 Esterase (Est0071) ex-
pressed from the putative gene is highly similar to esterases with
a so-called GGG(A)X motif in their active sites. This motif has been
associated with the ability of GGG(A)X hydrolases to convert steri-
cally demanding tertiary alcohols.5 Indeed, Est0071 shows activity
toward several arylaliphatic tertiary alcohols, albeit with low
⇑ Corresponding author. Tel.: +81 45 566 1786; fax: +81 45 566 1783.
E-mail address: kmiyamoto@bio.keio.ac.jp (K. Miyamoto).
enantioselectivity (not published). Moreover, it is known that sev-
eral of these esterases have excellent enantioselectivity toward
bulky malonate diesters (1).6
The hydrolysis reaction can be used to synthesize malonic acid
monoesters (2) in asymmetrization with 100% theoretical yield. In
contrast to mesophilic enzymes, Est0071 can catalyze this reaction
at temperatures above 70 "C. Under these reaction conditions, the
monoester is spontaneously cleaved to the a-arylpropionic acid es-
ter (4). If decarboxylation occurs at the active site it may be stere-
oselective. This prompted us to attempt an enantioselective
domino reaction to produce the pharmacologically important com-
pound naproxen (5b). If the chiral environment of the enzyme
causes the product to be optically pure, the stereospecific outcome
should depend to some degree on the choice of amino acids at the
active site, thus creating the opportunity to optimize enantioselec-
tivity by protein engineering (see Table 2).
Initially, we examined whether this esterase could catalyze the
hydrolysis of prochiral malonate diesters (1a) at room tempera-
ture. As anticipated, cell-free extract of Est0071 catalyzed the
hydrolysis of prochiral diesters to the corresponding monoesters
((S)-2a) with excellent stereoselectivity (99% ee) and 96% yield.
Interestingly, the enantiopreference of Est0071 is the opposite to
that of pig liver esterase, which forms the (R)-enantiomer with
96% ee.6 Under high temperatures of 50–70 "C, the monoester
(2a) was cleaved to the racemic ester (4a).4 As the esterase shows
excellent enantioselectivity, the decrease in optical purity was
attributed to racemization during decarboxylation. Variations in

0040-4039/$ - see front matter Crown Copyright ! 2013 Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.01.080
 Author's personal copy

1922 R. Wada et al. / Tetrahedron Letters 54 (2013) 1921–1923

pH between 6 and 9 and temperature between 50 and 70 "C had no CH3 2) Asymmetric
influence on the ee of the product. Nevertheless, the optical purity Path A 1) Hydrolysis CO2H Decarboxylation
of 5a was increased to 48% ee when His-tag purified enzyme was Ph
CO2H
employed in the reaction. In addition, no decarboxylation of 2a oc-
CH3 6a CH3
curred without Est0071 (data not shown). These observations led
CO2Et CO2H
us to assume that cleavage of the carboxyl group might take place Ph Ph
CO2H H
at the active site of Est0071. Accordingly, the chiral environment of
2a (S)-5a
the active site induces stereoselective decarboxylation. As both CH3
steps occur in an enantioselective manner, the reaction is a one-en- CO2Et
Ph
zyme three-step domino reaction (see Fig. 1). 1) Asymmetric H 2) Non-selective
Path B
To obtain a better understanding of the reaction mechanisms, Decarboxylation (S)-4a Hydrolysis
several of the compounds considered to be intermediates were
subjected to reaction with Est0071. We considered two possible Figure 2. Two possible reaction courses.

reaction courses (Fig. 2). In the first pathway shown as Path A, a

half-ester is first hydrolyzed to prochiral phenylmalonic acid
decarboxylation of malonate (6a) and proceeds via a prochiral
(6a). Diacid (6a) is then decarboxylated to the corresponding
intermediate.7–9 In this two-step reaction, the formation of the
monoacid (5a). In this pathway, decarboxylation of malonic acid
enolate is the rate determining step, and the direction of the addi-
is the selectivity-determining step. To confirm this assumption,
tion of the proton decides the absolute configuration of the prod-
we incubated phenylmalonate with Est0071 at 70 "C. We found
uct. It is also important to note that the optically active ester
that decarboxylation proceeded with or without esterase at almost
(4a) is hydrolyzed to the corresponding acid in a non-selective
the same rate, and that both products (5a) were racemic. More-
manner. Interestingly, the esterase-catalyzed reaction forms
over, it is known that the esterase-catalyzed hydrolysis of the sec-
(S)-enantiomers, while arylmalonate decarboxylase is (R)-specific.
ond ester of malonic acid esters is slow. Based on these
These findings encouraged us to increase the stereoselectivity of
observations, a reaction via a prochiral dicarboxylic acid interme-
the domino reaction by protein engineering. Directed evolution
diate is unlikely. In the second proposed pathway (Path B), decar-
efficiently improves the enantioselectivity of esterases.10 Unfortu-
boxylation of the half-ester (2a) is followed by hydrolysis. In this
nately, no high-throughput assay for determining the optical pur-
pathway, a monoester (2a) is decarboxylated to an optically active
ity of 5 was available. It is rather difficult to predict the amino
ester 4a. The esterase then hydrolyzes 4a to the corresponding acid
acid substitutions for rational engineering of stereoselectivity be-
(5a). In order to test this possibility, the reaction intermediate was
cause the exact role of active-site amino acids in the decarboxyl-
analyzed. In the beginning of the reaction, the diester disappeared
ation mechanism remains unknown. Therefore, we chose a semi-
immediately, and the half ester (2a) generated. After that, the car-
rational approach. A combination of structural considerations for
boxylic acid (5a) was formed gradually. Next, both (R)-4a and
generating small, high-quality libraries has been shown to be effec-
(S)-4a were subjected to the reaction. We found that the hydrolysis
tive for improving esterases.10 A homology model of Est0071 based
rates of both enantiomers were nearly identical. If Path B is correct,
the decarboxylation product must be an optically active ester.
However, the ester (4a) could not be isolated because the ester
hydrolysis rate was faster than that of decarboxylation. Therefore,
the decarboxylation is a rate-determining step. Adding the racemic
monoester (2a) as a substrate to the reaction led to the formation
of optically enriched (S)-5a (45% ee). Selectivity was almost the
same as the reaction using the optically active half-ester (48% ee).
These results indicated that the intermediate of the domino
reaction should be planar (3). After hydrolysis forms the (S)-enan-
tiomer of the monoester (2a), decarboxylation is stereoselective or
at least preserves the absolute configuration to some extent.
Accordingly, protonation occurs mainly from one side of the planar
enolate intermediate. This mechanism strongly resembles that of
arylmalonate decarboxylase, which catalyzes the enantioselective

Thermal spontaneous
CH3 Esterase-catalyzed CH3 decarboxylation in the
hydrolysis active site of esterase
CO2Et CO2Et
Ar Ar
CO2Et CO2H
1a, b (S)-2a, b

CH3 CH3 Esterase-catalyzed CH3

O hydrolysis
Ar CO2Et CO2H
Ar Ar
OEt H H
(S)-4a, b (S)-5a, b
Planar enolate
intermediate (3)

a; Ar = b; Ar =
MeO Figure 3. (a) Residues of the active site pocket of Est0071 with potential impact on
the conversion of phenylmalonate diethyl ester (1a); (b) catalytic triad (D243, H273
Figure 1. Esterase-catalyzed asymmetric domino reaction. and S150) and oxyanion hole (G78, G78, and G80) of Est0071.
 Author's personal copy

R. Wada et al. / Tetrahedron Letters 54 (2013) 1921–1923 1923

Table 1 putative prochiral intermediate, based on better delocalization of

Enantioselectivity of Est0071 and its variantsa the charge on the naphthyl ring system. The effects of the amino
Variant Conversion (%) eeb (%) Configuration acid substitutions differed remarkably from the findings using 1a
A179M/L198Y 37 80 S as the substrate. Variant A179 M/L198Y, which produced 5a with
A179M/L198E 43 77 S 80% ee, catalyzed the formation of 5b with 31% ee. In contrast,
I203R 5 74 S A179 K gave rise to (S)-5b with good optical purity (64% ee).
A179K 56 70 S We developed the first enzymatic domino reaction for the syn-
I203E 15 67 S
A179M/L198S 35 66 S
thesis of optically active (S)-profens. A mechanistic analysis of the
Wild-type 85 48 S reaction showed that it proceeds via enantioselective hydrolysis of
a
the malonic acid ester to the monoester. Cleavage of the carboxyl
Reaction conditions: pH 7, 70 "C, 48 h.
b group at the active site of the enzyme leads to the formation of
The ee values were determined by chiral-phase HPLC.
the prochiral intermediate, which is then protonated to give the
enantioenriched monoacid. Engineering of the active site gave rise
Table 2 to a variant with good stereoselectivity. As the esterase-catalyzed
Asymmetric synthesis of naproxen (5b)a domino reaction does not need external cofactors and allows a the-
Variant Conversion (%) ee (%) Configuration
oretical yield of 100%, it offers clear advantages over alternative
enzymatic routes.11 The esterase-catalyzed domino reaction is thus
A179K >99 64 S
an important step on the road toward the sustainable synthesis of
I203R >99 63 S
I203E >99 52 S optically active profens.
A179M/198Y >99 31 S
A179M/198E >99 18 S References and notes
Wild-type >99 21 S
a 1. Poliakoff, M.; Licence, P. Nature 2007, 450, 810–812.
Reaction conditions: pH 7, 70 "C, 48 h.
2. Burda, E.; Hummel, W.; Gröger, H. Angew. Chem., Int. Ed. 2008, 47, 9551–9554.
3. Akai, S.; Tanimoto, K.; Kita, Y. Angew. Chem., Int. Ed. 2004, 43, 1407–1410.
4. (a) Suzuki, Y.; Miyamoto, K.; Ohta, H. FEMS Microbiol. Lett. 2004, 236, 97–102;
on the X-ray structure of an esterase from Archaeoglobus fulgidus Esterase (Est0071) is commercially available from Enzymicals AG (Germany),
(PDB 1JJI) was the basis for the identification of three amino acid http://www.enzymicals.com/.
residues around the substrate binding site (Fig. 3). 5. Kourist, R.; Bornscheuer, U. T. Appl. Microbiol. Biotechnol. 2011, 191, 505–517.
6. Domínguez de María, P.; García-Burgos, C. A.; Bargeman, G.; van Gemert, R. W.
Residues 179, 198, and 203 are situated in close proximity to Synthesis 2007, 10, 1439–1452.
the catalytic triad. These positions were simultaneously altered 7. Miyamoto, K.; Ohta, H. J. Am. Chem. Soc. 1990, 112, 4077–4078.
by mutagenic polymerase chain reaction with primers bearing 8. Miyamoto, K.; Tsuchiya, S.; Ohta, H. J. Am. Chem. Soc. 1992, 114, 6256–6257.
9. Ijima, Y.; Matoishi, K.; Terao, Y.; Doi, N.; Yanagawa, H.; Ohta, H. Chem. Commun.
the NNK codon. A library of 200 clones was constructed. The num-
2005, 877–879.
ber of clones was not sufficient for full coverage, but helped to en- 10. Kourist, R.; Jochens, H.; Bartsch, S.; Kuipers, R.; Padhi, S. K.; Gall, M.; Bottcher,
sure the potential impact of the residues on the outcome of D.; Joosten, H. J.; Bornscheuer, U. T. ChemBioChem 2010, 11, 1635–1643.
decarboxylation. All clones were expressed in Escherichia coli and 11. Kourist, R.; Domínguez de Maria, P.; Miyamoto, K. Green Chem. 2011, 13,
2607–2618.
purified by His-tag affinity chromatography. After purification, all 12. Typical procedure: Two hundred microliters of 50 mM substrate (1a or 1b)
variants were active in the hydrolysis of p-nitrophenyl butyrate solution in CH3CN was added to 4 ml of 400 mM HEPES–NaOH buffer (pH 7).
at room temperature. In the domino reaction at 70 "C, 69 variants To the resulting solution, 1 ml of Est007 solution (25 lg/ml) was added and the
mixture was stirred for 48 h at 70 "C. The reaction was stopped by addition of
showed activity. Using these positive mutants, we next confirmed 2 ml of 2 M HCl. The products were extracted with 5 ml diisopropyl ether. The
the stereoselectivity of the domino reaction by HPLC 12. organic layer (4 ml) was concentrated in vacuo, and to this was added 400 ll
All variants showed (S)-selectivity. Interestingly, 6 variants diisopropyl ether, 100 ll methanol, and an excess amount of Me3SiMeN2 in
hexane. The solution was vortexed and allowed to stand at room temperature
showed increased stereoselectivity compared to the wild-type en- for 30 min. Enatiomeric excess of methyl ester of 5a and 5b was determined by
zyme (Table 1). The best variants (A179 M/L198Y) showed 80% ee, HPLC analysis using CHRALCEL OJ (Daicel Corporation, Japan). To determine
at conversions around 37%. The variants with improved selectivity the conversion rate, 500 ll of the reaction mixture was transferred to the tube,
and the reaction was stopped by addition of 200 ll of CH3CN containing 2%
had hydrophobic residues such as leucine replaced with polar argi- CF3CO2H. The products were extracted with EtOAc which contains diethyl
nine, serine, glutamate, or lysine. phthalate as an internal standard. The conversion rate of the products was
The enantioselective domino reaction was applied to the syn- determined by reverse phase HPLC using COSMOSIL 5C18-PAQ (Nakarai
Tesque, Inc., Japan).
thesis of (S)-naproxen (5b). The conversion of 1b was much faster
than that of 1a. This can be explained by the higher stability of the