Improvements of enzyme activity and enantioselectivity in lipase-catalyzed alcoholysis of (R,S)-azolides

Article abstract of DOI:10.1016/j.molcatb.2009.11.001

With Candida antarctica lipase B (CALB)-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide at the optimal conditions (i.e. anhydrous MTBE as the solvent, and methanol as the acyl acceptor at 45°C) as the model system, the enzyme enantioselectivity

Full text of DOI:10.1016/j.molcatb.2009.11.001

Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Improvements of enzyme activity and enantioselectivity in lipase-catalyzed
alcoholysis of (R,S)-azolides
An-Chi Wu, Pei-Yun Wang, Yi-Sheng Lin, Min-Fang Kao,
Jin-Ru Chen, Jyun-Fen Ciou, Shau-Wei Tsai∗
Institute of Biochemical and Biomedical Engineering, Chang Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan, Tao-Yuan 33302, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 August 2009
Received in revised form 8 October 2009
Accepted 1 November 2009
Available online 10 November 2009
With Candida antarctica lipase B (CALB)-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide at the
optimal conditions (i.e. anhydrous MTBE as the solvent, and methanol as the acyl acceptor at 45 ◦C) as the
model system, the enzyme enantioselectivity in terms of VR/VS = 105.8 and specific activity for the fast-
reacting (R)-azolide VR/(Et) = 0.979 mmol/(h g) were greatly improved in comparison with VR/VS = 8.0 and
VR/(Et) = 0.113 mmol/(h g) of using (R,S)-naproxenyl 2,2,2-trifluoroethyl ester as the substrate. The resolu-
tion strategy was successfully extended to other (R,S)-profenyl 1,2,4-triazolides and lipases from Candida
rugosa (Lipase MY) and Carica papaya (CPL) having opposite enantioselectivity to CALB. Moreover, the
kinetic constants were estimated, compared with those obtained via hydrolysis, and employed for mod-
eling time-course conversions of (R,S)-naproxenyl 1,2,4-triazolide in anhydrous MTBE. The advantages
of easy substrate preparation, high enzyme reactivity and enantioselectivity, as well as easy product sep-
aration from the remaining substrate via reactive extraction demonstrate merits of using (R,S)-azolides
but not the corresponding esters for the alcoholytic resolution.
Keywords:
(R,S)-Azolide
(R,S)-Profen
Lipase
Alcoholysis
Kinetic analysis
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
active-site structure for increasing the enantiomer discrimina-
tion and reactivity, acyl donors containing an achiral activated
or irreversible leaving group have been utilized [24,25]. Recently,
(R,S)-azolides, i.e. N-acylazoles, as novel substrates for lipase-
catalyzed hydrolysis in water-saturated organic solvents were
reported [26]. The good enzyme performance was attributed to
the unique azolide structure that contains an unshared electron
pairs on the acyl-substituted nitrogen N(1), making the carbonyl
carbon more electrophilic and susceptible to nucleophilic attack
[27]. It is therefore aimed to extend the previous resolution pro-
cess to the alcoholysis of (R,S)-azolides in anhydrous organic
solvents for preparing optically pure carboxylic acids or their
derivatives.
Lipases as versatile biocatalysts are widely applied for prepar-
ing a variety of pharmaceuticals and fine chemicals containing
a chiral center. Their active site consisting of a Ser-His-Asp/Glu
catalytic triad catalyzes the hydrolysis or synthesis by following
an acylation–deacylation displacement mechanism. In comparison
with the resolution of secondary alcohols and amines [1–4], the
enantioselectivity toward carboxylic acids is usually low to modest
[5]. Therefore, various approaches of using substrate engineering
[6–8], medium engineering [9–11], enzyme engineering [12–17], or
their combinations have been proposed for improving the enzyme
performance.
Candida antarctica lipase B (CALB), showing low to modest
enantioselectivity toward (R,S)-profens via esterification and (R,S)-
profenyl esters via hydrolysis and transesterification [7,20,22],
was employed as the model lipase for resolving (R,S)-profenyl
azolides via alcoholysis in anhydrous organic solvents (Scheme 1).
Effects of solvent, temperature, alcohol, and substrate struc-
ture on varying the enzyme activity and enantioselectivity were
firstly studied. The resolution was then extended to Lipase
MY and CPL having opposite enantioselectivity to CALB. More-
over, the kinetic constants estimated and compared with those
obtained via hydrolysis, as well as separation of ester prod-
ucts from the remaining substrate via reactive extraction were
addressed.
In comparison with esterification of (R,S)-carboxylic acids or
hydrolysis of the corresponding (R,S)-esters, fewer studies relevant
to transesterification, aminolysis and ammonolysis of (R,S)-esters
are reported in the literature [7,18–23]. Apparently, this is due
to at least an additional synthetic step for preparing ester sub-
strates or hydrolytic step for obtaining acid products, as well as
the difficulty of separating resultant ester or amide products from
remaining ester substrates. In order to manipulate the enzyme
∗
Corresponding author. Tel.: +886 3 2118800x3415; fax: +886 3 2118668.
E-mail address: tsai@mail.cgu.edu.tw (S.-W. Tsai).
1381-1177/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2009.11.001

236
A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
2-fluoro-␣-methyl-4-biphenylacetic acid), (R,S)-ibuprofen (i.e.
Nomenclature
(R,S)-2-(4-isobutylphenyl)-propionic acid) and (R,S)-fenoprofen
(i.e. (R,S)-2-(3-phenoxyphenyl)-propionic acid) from Sigma (St.
Louis, MO); dichloromethane, diisopropyl ether (IPE), ethanol,
hexane, isooctane (ISO), isopropanol, methanol, methyl tert-butyl
ether (MTBE), propanol, butanol, pentanol, hexanol, tert-butyl
alcohol (TBA) and triethylamine from Tedia (Fairfield, OH).
Anhydrous solvent was prepared by adding calcium hydride
from Riedel-de Haen (Seelze, Germany) to the organic solvent
for 24 h.
eep, ees enantiomeric excesses for product and substrate,
respectively
E
(Et)
G
enantiomeric ratio, defined as k2SKmR/k2RKmS
enzyme concentration (mg/ml)
parameter defined in Eq. (3)
k2i, k4i
kinetic constants, i = R or S for (R)- or (S)-enantiomer
(mmol/(g h))
KI
inhibition constant (mM)
Kmi, Km3i kinetic constants, i = R or S for (R)- or (S)-enantiomer
2.2. Substrate preparations
(mM)
(Si)
(Si)0
Vi
substrate concentration, i = R or S for (R)- or (S)-
enantiomer (mM)
initial concentration, i = R or S for (R)- or (S)-
enantiomer (mM)
initial rate, i = R or S for (R)- or (S)-enantiomer
(mM/h)
The general procedures for preparing (S)- or (R,S)-profenyl
azolides from the acylation of azoles with acid chloride or via
direct activation of carboxylic acid by CDI or CDT were previously
described [26]. For example, to 5 ml benzene was added 1 mmol
(R,S)-naproxen and 1.5 mmol CDT and stirred at 55 ◦C for 2 h. The
resultant mixture was filtered and evaporated under reduced pres-
sure, giving the desired (R,S)-naproxenyl 1,2,4-triazolide. Thionyl
chloride of 13 mmol in 5 ml benzene was added dropwise to a
mixture consisting of 10 mmol 4-bromopyrazole, 13 mmol (R,S)-
naproxen, and 40 mmol triethylamine. The resultant solution was
stirred for 2 h at room temperature. After being quenched in suc-
cession with 0.1 M HCl solution, 0.1 M NaOH solution and 0.1 M
NaCl solution each for three times (3× 10 ml), the organic phase
was separated, dried over anhydrous MgSO4, filtered and concen-
trated under reduced pressure, giving the desired (R,S)-naproxenyl
4-bromopyrazolide.
To 20 ml benzene containing 27 mmol thionyl chloride and
15 mmol (R,S)-naproxen was refluxed for 1.5 h at 80 ◦C. After evap-
orating the solvent and remaining thionyl chloride, 30 ml benzene
containing 15 mmol pyridine and 27 mmol 2,2,2-trifluoroethanol
was added and refluxed for 4 h at 80 ◦C. The resultant solution
was quenched in succession with NaOH solution (pH = 12.5, 4×
50 ml) and deionized water (2× 100 ml). After purification in sil-
ica gel chromatography with the mobile phase of hexane/ethyl
acetate (2/1, v/v) and concentrated under reduced pressure, the
desired (R,S)-naproxen 2,2,2-trifluoroethyl ester in white powder
was obtained. The synthesized substrates were confirmed from the
retention time in HPLC analysis and 1H NMR spectra [8,26].
Xi, Xt
enantiomer conversion defined as [1 − (Si)/(Si)0],
i = R or S; racemate conversion defined as [XR + XS]/2
2. Materials and methods
2.1. Materials
Novozym 435 (CALB) from Candida antartica lipase B and
Lipase MY from Candida rugosa were provided by Novo Nordisk
(Bagsvaerd, Denmark) and Meito Sangyo Industries Ltd. (Tokyo,
Japan), respectively. A Carica papaya lipase (CPL) partially purified
from the aqueous-insoluble aggregate of crude papain was kindly
donated from Challenge Bioproducts (Yun-Lin Hsien, Taiwan).
Other chemicals of analytical grade were commercially available:
pyrazole from Acros (Geel, Belgium); 4-bromopyrazole, N,N-
carbonyldiimidazole (CDI), N,N-carbonyldi-1,2,4-triazole (CDT)
and 2,2,2-trifluoroethanol from Aldrich (Milwaukee, WI); (R,S)-
2-phenylpropionic acid from Fluck (Buchs, Switzerland); (R,S)-
ketoprofen (i.e. (R,S)-2-(3-benzoylphenyl)-propionic acid), (R,S)-
naproxen (i.e. (R,S)-2-(6-methoxy-2-naphthyl)-propionic acid) and
(S)-naproxen from TCI (Tokyo, Japan); (R,S)-flurbiprofen (i.e. (R,S)-
Scheme 1. Lipase-catalyzed alcoholysis of (R,S)-profenyl azolides and (R,S)-naproxenyl 2,2,2-trifluoroethyl ester in anhydrous organic solvents.

A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
237
Table 1
Effects of solvent on specific initial rates VR/(Et) and VS/(Et), VR/VS, Xt and ees for CALB-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide.
Solvent (log P)
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VR/VS
Time (h)
Xt (%)
ees (%)
ISO (4.50)
IPE (1.90)
MTBE (1.43)
TBA (0.79)
4.95E−1
6.03E−1
9.79E−1
1.33E−1
2.51E−2
1.25E−2
9.25E−3
3.50E−3
19.7
48.3
106
1.0
1.5
2.2
7.0
66.6
53.0
53.3
53.9
>99.8
>99.8
>99.8
>99.8
38.1
Conditions: 10 ml anhydrous solvent containing 3 mM racemate (but 1 mM racemate in ISO), 100 mM methanol, and 6 mg/ml lipase at 45 ◦C and 400 rpm. Symbols E−1 and
P for solvent as 10−1 and partition coefficient of the solvent between water and octanol, respectively.
2.3. Analysis
2.5. Kinetic analysis
Alcoholysis of (R,S)-naproxenyl 4-bromopyrazolide was mon-
itored by HPLC using a chiral (S,S)-whelk-01 column from Regis
(Morton Grove, IL) that was capable of separating the internal
standard of 2-nitrotoluene, (S)- and (R)-azolides with the reten-
tion time of 2.2, 3.0, 4.1 min, respectively. The mobile phase was
a mixture of n-hexane/isopropanol (90/10, v/v) at a flow rate of
2.0 ml/min. UV detection at 270 nm was employed for quantifica-
tion at the room temperature. For the analysis of (R,S)-naproxen
2,2,2-trifluoroethyl ester, the composition of mobile phase was
changed to n-hexane/isopropanol/acetic acid glacial (80/20/0.5,
v/v/v), giving the retention time for the internal standard of ace-
topheone, (R)- and (S)-esters as 2.4, 3.2, 4.0 min, respectively.
Detailed analytical conditions for other (R)- and (S)-profenyl
azolides were reported elsewhere [26]. The products of (R)- and (S)-
naproxen methyl esters were monitored at 270 nm by using a Daicel
chiral OD-H column (Tokyo, Japan) that was capable of separating
the internal standard of 2-nitrotoluene, (R)- and (S)-ester with the
retention time of 2.7, 3.4, 3.8 min, respectively. The mobile phase
was a mixture of n-hexane/isopropanol (96.5/3, v/v) at a flow rate of
2.0 ml/min.
The alcoholysis was performed in 10 ml anhydrous MTBE
consisting of 6 mg/ml CALB, methanol and (R,S)-naproxenyl 1,2,4-
triazolide of different concentrations at 45 ◦C, from which the
time-course conversions and initial rates for each enantiomer
were estimated. Similar experiments of only using (S)-naproxenyl
1,2,4-triazolide as the acyl donor were carried out. An irreversible
ping-pong Bi–Bi mechanism by considering alcohol inhibition (see
Supplementary Information) was employed for deriving the rate
equations as follows:
−d(SR)
k2R(SR)(Et)/KmR
VR
VS
=
=
=
=
(1)
(2)
dt
G
−d(SR)
k2S(SS)(Et)/KmS
dt
G
(SR)[1 + k2R/k4R
]
(SS)[1 + k2S/k4S
]
(M)
KI
G = 1 +
+
+
KmR
KmS
ꢀ
ꢁꢀ
ꢁ
k2RKm3R(SR)
k4RKmR
k2SKm3S(SS)
k4SKmS
1
(M)
+
+
(3)
Notations (Et), (M), (SR) and (SS) denoted the concentrations of
enzyme, alcohol, (R)- and (S)-azolides in the solvent, respectively.
All kinetic constants defined in the Nomenclature were estimated
and employed for modeling the time-course conversions of the
substrates.
2.4. Effects of solvent, temperature, alcohol, substrate structure,
and lipase source
To 10 ml anhydrous ISO, IPE, MTBE or TBA containing 3 mM (R,S)-
naproxenyl 1,2,4-triazolide and 100 mM methanol at 45 ◦C was
added 6 mg/ml CALB. The resultant solution was stirred with a mag-
netic stirrer, and samples were removed at different time intervals
for the HPLC analysis, from which the time-course conversions XR
and XS with the analytical error less than 1.3%, initial rates for both
enantiomers VR and VS based on several conversion determinations,
racemate conversion Xt, and enantiomeric excess for the substrate
ees were determined. Similar experiments were performed except
that anhydrous MTBE was employed as the solvent at 35 and 55 ◦C.
In order to investigate effects of acyl acceptor on the lipase perfor-
mance, more experiments in anhydrous MTBE containing 100 mM
of ethanol, propanol, butanol, pentanol or hexanol were carried out
at 45 ◦C.
Effects of substrate structure on the enzyme activity and
enantioselectivity were studied, in which the alcoholysis of
methanol with (R,S)-naproxen 2,2,2-trifluoroethyl ester or (R,S)-
naproxenyl azolides containing a leaving imidazole, pyrazole, or
4-bromopyrazole was performed in anhydrous MTBE at 45 ◦C.
Similar experiments of using other (R,S)-profenyl 1,2,4-triazolides
as the acyl donor were carried out, and the results were
compared with those via hydrolysis in anhydrous and water-
saturated MTBE. In order to test if the resolution process could
be applied to other lipases having opposite enantioselectivity to
CALB, the alcoholysis of (R,S)-naproxenyl azolides by methanol
via Lipase MY or CPL in anhydrous isooctane at 45 ◦C was also
performed.
2.6. Reactive extraction for separating the product
In order to simplify the synthesis procedures, 62.0 mM (R,S)-
naproxenyl 1,2,4-triazolide was prepared in 10 ml anhydrous MTBE
but not benzene via direct activation of (R,S)-naproxen by CDT at
55 ◦C for 12 h. After centrifuging the precipitate, to 8 ml of the solu-
tion was added 6 mg/ml CALB and 50 mM methanol at 45 ◦C for
2.5 h, giving Xt = 51.5%, eep = 94.1% and ees = 100% from the HPLC
analysis. As an example for the easy product separation, to 5 ml
of the resultant solution after removing CALB via centrifugation
was added 5 ml sodium carbonate buffer (pH = 9, 300 mM) at 45 ◦C
with stirring for 7 h. The methyl ester products of 38.1 mM and
eep = 91.5% without the traces of (S)-azolide in the organic phase
was found from HPLC analysis. Apparently, the hydrolysis of (S)-
naproxenyl 1,2,4-triazolide in the aqueous phase was more rapidly
than that of (R)- or (S)-naproxenyl methyl ester. Yet in order to
enhance the optical purity of (S)-naproxen in the aqueous phase,
other extraction conditions for decreasing the ester hydrolysis
should be found.
3. Results and discussion
3.1. Effects of solvent, temperature and alcohol
Table 1 demonstrates the effects of anhydrous organic solvent
on the initial specific activities VR/(Et) and VS/(Et), and enan-

238
A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
Table 2
Effects of temperature on specific initial rates VR/(Et) and VS/(Et), VR/VS, Xt and ees for CALB-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide.
Temperature (◦C)
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VR/VS
Time (h)
Xt (%)
ees (%)
35
55
4.82E−1
4.00E−3
1.63E−2
120
80.0
3.0
2.0
52.4
55.2
>99.8
>99.8
1.30
Conditions: 10 ml anhydrous MTBE containing 3 mM racemate, 100 mM methanol, and 6 mg/ml lipase at 400 rpm. Symbol of E−1 as 10−1
.
Table 3
Effects of alcohol on specific initial rates VR/(Et) and VS/(Et), VR/VS, Xt and ees for CALB-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide.
Alcohol
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VR/VS
Time (h)
Xt (%)
ees (%)
Ethanol
Propanol
Butanol
Pentanol
Hexanol
6.73E−1
6.15E−1
5.75E−1
7.86E−1
7.81E−1
7.13E−3
5.86E−3
6.50E−3
7.65E−3
7.33E−3
94.4
105
88.4
102
2.2
4.1
4.1
2.2
2.2
54.0
52.9
54.8
53.5
53.0
>99.8
>99.8
>99.8
>99.8
>99.8
106
Conditions: 10 ml anhydrous MTBE containing 3 mM racemate, 100 mM alcohol, and 6 mg/ml lipase at 45 ◦C and 400 rpm. Symbol of E−1 as 10−1
.
tioselectivity in terms of VR/VS for CALB-catalyzed alcoholysis of
(R,S)-naproxenyl 1,2,4-triazolide by methanol at 45 ◦C. In gen-
eral, the specific activity for VS/(Et) but not VR/(Et) increased
with the solvent hydrophilicity in terms of the log P value, yet
it reversed if the value further increased from 1.43 of MTBE to
1.90 of IPE. An optimal VR/(Et) for the fast-reacting enantiomer
and VR/VS = 105.8 in anhydrous MTBE was thus obtained. In com-
parison with VR/(Et) = 0.113 mmol/(h g) and VR/VS = 81.8 for the
lipase-catalyzed hydrolysis in water-saturated MTBE [26], the bet-
ter enzyme activity and enantioselectivity via alcoholysis was
apparent.
Table 2 shows effects of temperature on varying the enzyme
activity and enantioselectivity in anhydrous MTBE. A liner rela-
tionship of ln(VR/(Et)), ln(VS/(Et)) or ln(VR/VS) with inverse of the
absolute temperature was found (data not shown), implying that
the lipase was thermally stable at 55 ◦C. Yet, 45 ◦C was selected as
the best temperature by considering a compromise between the
enzyme activity and enantioselectivity.
of leaving imidazole or pyrazole to 3 of 1,2,4-triazole, the specific
activity for each enantiomer increased and resulted in maximum
enantioselectivity of VR/VS = 179.3 for (R,S)-naproxenyl pyrazolide.
One may attribute the high reactivity of azolides to the unshared
electron pairs on the acyl-substituted nitrogen N(1) that are part
of the cyclic ␲-system of the azole units. This unique structure can
lead to a partial positive charge on N(1) that interfaces with the nor-
mal carboxamide resonance and exerts an electron-withdrawing
effect on the carbonyl groups. Therefore, the carbonyl carbon is
more electrophilic and susceptible to nucleophilic attack, especially
when the number of nitrogens of leaving azoles increases [26,27].
The enantioselectivity further increased to VR/VS = 205.5 when
employing (R,S)-naproxenyl 4-bromopyrazolide as the substrate.
A
comparison of pKa = 0.64 for the conjugated acid of 4-
bromopyrazole (i.e. 4-bromopyrazolium) with 7.0 of imidazolium,
2.52 of pyrazolium, and 2.19 of 1,2,4-triazolium [28] indicated that
the leaving 4-bromopyrazole moiety might exert the strongest
electron-withdrawing effect on the carbonyl carbon atom, mak-
ing it the most electrophilic and susceptible to nucleophilic attack.
Yet, the specific activity VR/(Et) of the 4-bromopyrazolide was only
slightly higher than that of (R,S)-naproxenyl pyrazolide and even
lower than that of (R,S)-naproxenyl 1,2,4-triazolide, indicating that
extra interactions between 4-bromo substituent and the amino acid
residues of enzyme active site might exist to weaken the electron-
withdrawing ability of the substituent. This can be verified from
comparing the extent of interactions by docking different azolides
into the active site via molecular modeling techniques.
Table
3 further demonstrates the effects of alcohol on
VR/(Et), VS/(Et), and VR/VS for CALB-catalyzed alcoholysis of (R,S)-
naproxenyl 1,2,4-triazolide in anhydrous MTBE at 45 ◦C. In general,
the carbon-chain length of the acyl acceptor had minute influences
on the enzyme activity and enantioselectivity, implying that the
rate-limiting step for the alcoholysis was the acylation step. By con-
sidering the lowest cost, methanol was selected as the best acyl
acceptor for the following experiments.
The results of using (R,S)-naproxenyl 2,2,2-trifluoroethyl ester
as the substrate was also tabulated (Table 4). About 8.6-fold and
13.2-fold lower of VR/(Et) and VR/VS, respectively, were perceived
when comparing with those for (R,S)-naproxenyl 1,2,4-triazolide
(Table I). By further considering the difficulty of separating methyl
ester from remaining 2,2,2-trifluoroethyl ester, the benefit of using
(R,S)-azolides but not (R,S)-esters as the substrate was evident.
The analysis was then extended to other (R,S)-profenyl 1,2,4-
triazolides (Table 5), showing very good enzyme performances
similar to those for (R,S)-naproxenyl 1,2,4-triazolide. In general
the more bulky was the acyl group; the lower enzyme activity for
3.2. Effects of substrate structure and lipase source
In order to demonstrate that (R,S)-azolides acted as versatile
substrates for the kinetic resolution in anhydrous organic sol-
vents, Table 4 indicates the alcoholysis of (R,S)-naproxenyl azolides
and (R,S)-naproxenyl 2,2,2,-trifluoroetheyl ester by methanol in
anhydrous MTBE. The 172-fold enhancement of VR/(Et) for (R,S)-
1,2,4-triazolide in comparison with that for (R,S)-imidazolide led
to the rate-limiting acylation step for all fast-reacting enantiomers
as they had the same acyl-enzyme intermediate for performing
the deacylation step. By increasing the number of nitrogens from 2
Table 4
Effects of leaving moiety on specific initial rates VR/(Et) and VS/(Et), VR/VS, Xt and ees for CALB-catalyzed alcoholysis of (R,S)-naproxenyl azolides.
Leaving moiety
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VR/VS
Time (h)
Xt (%)
ees (%)
Imidazole
Pyrazole
4-Bromopyrazole
1,2,4-Triazole
Trifluoroethanol
5.69E−3
3.51E−1
4.26E−1
9.79E−1
1.13E−1
1.81E−3
1.96E−3
2.08E−3
9.25E−3
1.41E−2
3.1
179
205
105
8.0
7.1
8.2
3.0
2.2
7.0
29.2
57.0
51.2
53.3
64.2
18.6
>99.8
>99.8
>99.8
86.8
Conditions: 10 ml anhydrous MTBE containing 3 mM racemate, 100 mM methanol, and 6 mg/ml lipase at 45 ◦C and 400 rpm. Symbol of E−1 as 10−1
.

A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
239
Table 5
Effects of alcohol on specific initial rates VR/(Et) and VS/(Et), VR/VS, Xt and ees for CALB-catalyzed alcoholysis of (R,S)-profenyl 1,2,4-triazolides.
Acyl moiety and alcohol
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VR/VS
(Et) (mg/ml)
Time (h)
Xt (%)
ees (%)
2-Phenylpropionyl
Methanol
Ethanol
Propanol
Butanol
9.46
7.32
6.33
6.52
8.66
6.16
2.15E−1
1.28E−1
1.41E−1
1.50E−1
1.50E−1
1.43E−1
43.9
57.1
44.8
43.4
57.6
43.1
2.0
2.0
2.0
2.0
2.0
2.0
0.6
0.3
0.5
0.3
0.5
1.0
60.0
53.5
55.0
56.2
57.4
60.0
>99.8
>99.8
>99.8
>99.8
>99.8
>99.8
Pentanol
Hexanol
Fenoprofenyl
Methanol
Hexanol
4.94
6.62
4.94E−2
6.98E−2
100
94.9
2.0
2.0
2.0
0.7
55.0
53.4
>99.8
>99.8
Flurbiprofenyl
Methanol
Hexanol
1.08
7.53E−1
1.31E−2
7.38E−3
82.5
102
6.0
6.0
3.0
1.0
56.1
51.0
>99.8
>99.8
Ibuprofenyl
Methanol
Hexanol
1.29
4.77E−1
1.13E−2
4.80E−3
114
99.4
2.0
2.0
8.0
9.0
52.1
60.0
>99.8
>99.8
Ketoprofenyl
Methanol
Hexanol
3.96
4.61
6.25E−2
5.24E−2
63.4
88.0
2.0
2.0
1.0
1.0
51.4
50.9
>99.8
>99.8
Conditions: 10 ml anhydrous MTBE containing 3 mM racemate and 100 mM alcohol at 45 ◦C and 400 rpm. Symbol of E−1 as 10−1
.
Table 6
Effects of lipase-catalyzed alcoholysis of (R,S)-naproxenyl azolides on specific initial rates VR/(Et) and VS/(Et), VS/VR, Xt and ees.
Lipase and leaving azole
VR/(Et) (mmol/(h g))
VS/(Et) (mmol/(h g))
VS/VR
(Et) (mg/ml)
Time (h)
Xt (%)
ees (%)
Lipase MY
Pyrazole
4-Bromopyrazole
1,2,4-Triazole
2.40E−4
4.95E−4
2.04E−2
9.16E−3
6.21E−2
1.67E−1
38.2
125
8.2
50.0
10.0
6.0
23.5
10.0
4.0
58.8
51.5
60.5
>99.8
>99.8
>99.8
CPL
4-Bromopyrazole
1.35E−4
3.44E−2
6.71E−3
3.82E−2
49.7
1.1
10.0
6.0
22.9
5.0
32.7
50.1
45.6
2.7
1,2,4-Triazole
Conditions: 10 ml anhydrous isooctane containing 100 mM methanol and 3 mM racemate (but 1 mM (R,S)-naproxenyl 1,2,4-triazole) at 45 ◦C and 400 rpm. Symbol of E−1 as
10−1
.
k2S/KmS = 1.29 × 10−2 L/(h g), and E = k2RKmS/k2SKmR = 161, were
estimated, showing that the acylation step was rate-limiting in the
whole reaction steps. By substituting theses kinetic constants into
Eqs. (1)–(3), the time-course (SR) and (SS), and hence XR and XS,
were solved by using a fourth-order Runge–Kutta method. Some
experimental data in agreement with the theoretical predictions
were represented in Fig. S1.
the fast-reacting enantiomer but with better enantioselectivity was
obtainable.
In order to test if the resolution strategy could be applied to other
lipases having opposite enantioselectivity to CALB, the alcoholysis
of (R,S)-naproxenyl azolides by methanol in anhydrous isooctane at
45 ◦C via C. rugosa and C. papaya lipases was carried out. As shown in
Table 6, good to excellent enantioselectivity for CPL and Lipase MY
was obtained when selecting (R,S)-naproxenyl 4-bromopyrazolide
as the substrate. All the results represented in Tables 4–6 really
imply that for a custom-make (R)- or (S)-carboxylic acid, one may
find an efficient azole to prepare the corresponding (R,S)-azolide
for carrying out the alcoholysis.
3.3. Kinetic analysis
As azolides might act as activated acyl donors for carrying out
the lipase-catalyzed alcoholysis, an irreversible ping-pong Bi–Bi
mechanism by considering alcohol as a competitive inhibitor
was employed for deriving the rate equations. Variations of
initial VR/(Et) and 100 VS/(Et) with enantiomer and methanol
concentrations, respectively, for CALB-catalyzed alcoholysis of
(R,S)-naproxen 1,2,4-triazolide in anhydrous MTBE were illus-
trated in Figs. 1 and 2. As described in Supplementary Information,
the kinetic constants k2R = 181 mmol/(h g), k2S = 0.43 mmol/(h g),
Fig. 1. Variations of specific initial rate VR/(Et) (ꢁ) with (R)-naproxenyl 1,2,4-
triazolide and methanol concentrations in anhydrous MTBE consisting of 6 mg/ml
CALB at 45 ◦C. (—) Best-fit results.
KmR = 86.8 mM, KmS = 33.3 mM, KI = 34.7 mM, k4RKmR ꢀ k2RKm3R
,
and and then k2R/KmR = 2.08 L/(h g),
k4SKmS ꢀ k2SKm3S
,

240
A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
ety from 1,2,4-triazole to pyrazole and 4-bromopyrazole but
not imidazole at the optimal conditions yielded enhancements
of enzyme enantioselectivity but not specific activity for the
fast-reacting (R)-enantiomer. In comparison with the enzyme
performance of using (R,S)-naproxenyl 1,2,4-triazolide and (R,S)-
2,2,2-trifluoroethyl ester as the substrate, about 8.6-fold and
13.2-fold higher of VR/(Et) and VR/VS, respectively, for the for-
mer were obtained. The resolution strategy was then applied
to other (R,S)-profenyl 1,2,4-triazolides, showing good to excel-
lent enzyme enantioselectivity. Moreover, with a careful selection
of reaction conditions (i.e. anhydrous isooctane as the solvent
and 4-bromopyrazole as the leaving moiety), the strategy was
successfully extended to Lipase MY and CPL having opposite enan-
tioselectivity to CALB.
A
thorough kinetic analysis for the alcoholysis of (R,S)-
naproxenyl 1,2,4-triazolide at the optimal conditions led to
the rate-limiting acylation step with the alcohol as a competi-
tive inhibitor. The kinetic constants estimated were successfully
employed for modeling the time-course conversions for the sub-
strate. A comparison of k2R/KmR and k2S/KmS for the alcoholysis
in anhydrous MTBE with those via hydrolysis in anhydrous and
water-saturated MTBE implied that water also acted as an enzyme
inhibitor. An alkaline buffer solution was further employed for per-
forming the reactive extraction, in which the ester products in the
organic phase can be easily separately from the remaining azolide
substrate.
Fig. 2. Variations of specific initial rate 100 VS/(Et) (᭹) with (S)-naproxenyl 1,2,4-
triazolide and methanol concentrations in anhydrous MTBE consisting of 6 mg/ml
CALB at 45 ◦C. (—) Best-fit results.
It has been reported that the water content in MTBE had
profound influence on CALB-catalyzed hydrolysis of (R,S)-2-
phenylpropionyl 1,2,4-triazolide, in which an order-of-magnitude
higher KmR and KmS, as well as k2R but not k2S, were found when
anhydrous MTBE as the reaction medium was replaced with water-
saturated MTBE [26]. This implied that water molecules adsorbed in
the enzyme active site not only impeded the substrate affinity to the
binding pockets, but also played as lubricants for the easy proton
transfer from catalytic serine to histidine to perform the nucle-
ophilic reaction. However in comparison k2R/KmR = 7.91 l/(h g),
k2S/KmS = 0.605 l/(h g) and hence E = 13.1 in anhydrous MTBE
with k2R/KmR = 2.90 l/(h g), k2S/KmS = 2.92 × 10−2 l/(h g) and hence
E = 99.0 in water-saturated MTBE, the higher enantioselectivity
for the latter was offset by the lower enzyme reactivity for the
fast-reacting (R)-enantiomer. Similarly by regarding KmR ꢀ (SR)
and KmS ꢀ (SS) for the hydrolysis of (R,S)-naproxenyl 1,2,4-
triazolide, k2R/KmR = 1.47 l/(h g) and k2S/KmS = 1.52 × 10−2 l/(h g) in
anhydrous MTBE (Table S1), as well as k2R/KmR = 0.173 l/(h g) and
k2S/KmS = 2.12 × 10−3 l/(h g) in water-saturated MTBE [26], were
estimated from VR/(Et)/(SR) and VS/(Et)/(SS), respectively. Therefore
regardless of the acyl part of (R,S)-profenyl azolides, VR/(Et) and
VS/(Et) as well as k2R/KmR and k2S/KmS in anhydrous MTBE should
be greater than those in water-saturated MTBE, implying that one
might regard water as a competitive enzyme inhibitor in deriving
the rate equations [29].
Acknowledgement
Financial supports from National Science Council (Grant No. NSC
97-2221-E-182-018-MY3) are appreciated.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2009.11.001.
References
[1] M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Kesseler, R. Sturmer, T. Zelinski,
Angew. Chem. Int. Ed. 43 (2004) 788–824.
[2] R. Chenevert, N. Pelchat, F. Jacques, Curr. Org. Chem. 10 (2006) 1067–1094.
[3] V. Gotor-Fernández, V. Gotor, Curr. Org. Chem. 10 (2006) 1125–1143.
[4] E. Santaniello, S. Casati, P. Ciuffreda, Curr. Org. Chem. 10 (2006) 1095–1123.
[5] U.T. Bornscheuer, R.J. Kazlauskas, Hydrolases in Organic Synthesis: Regio- and
Stereoselective Biotransformations, 2nd ed., Wiley-VCH, Weinheim, 2006, pp.
37–275.
[6] T. Miyazawa, M. Shimaoka, T. Yamada, Biotechnol. Lett. 21 (1999) 309–312.
[7] H. Yang, E. Henke, U.T. Bornscheuer, J. Org. Chem. 64 (1999) 1709–1712.
[8] S.W. Tsai, C.C. Chen, H.S. Yang, I.S. Ng, T.L. Chen, BBA Protein Proteomics 1764
(2006) 1424–1428.
[9] U.T. Bornscheuer, C. Bessler, R. Srinivas, S.H. Krishna, Trends Biotechnol. 20
(2002) 433–437.
[10] E.P. Hudson, R.K. Eppler, D.S. Clark, Curr. Opin. Biotechnol. 16 (2005) 637–643.
[11] F. Theil, Tetrahedron 56 (2000) 2905–2919.
[12] P. Berglund, S. Park, Curr. Org. Chem. 9 (2005) 325–336.
[13] B.G. Davis, Curr. Opin. Biotechnol. 14 (2003) 379–386.
[14] T. Penning, Chem. Rev. 101 (2001) 3027–3046.
[15] M.T. Reetz, Adv. Catal. 49 (2006) 1–69.
[16] M.D. Toscano, K.J. Woycechowsky, D. Hilvert, Angew. Chem. Int. Ed. 46 (2007)
3212–3236.
[17] P.Y. Wang, S.W. Tsai, T.L. Chen, Biotechnol. Bioeng. 101 (2008) 460–469.
[18] C.S. Chang, S.W. Tsai, C.N. Lin, Tetrahedron: Asymmetry 9 (1998) 2799–2807.
[19] R. Chenevert, P. Morin, N. Pelchat, in: V. Gotor, I. Alfonso, E. Garcia-Urdiales
(Eds.), Asymmetric Organic Synthesis with Enzymes, Wiley-VCH, Weinheim,
2008, pp. 133–169.
A comparison of k2R/KmR (or k2S/KmS) values for the hydrolysis
and alcoholysis of (R,S)-naproxenyl 1,2,4-triazolide in anhydrous
MTBE indicated that the presence of methanol had minute influ-
ences on varying the intrinsic kinetic behavior of CALB. As the
water content at 45 ◦C in water-saturated MTBE was estimated
as 472.7 mM [30], one might evaluate k2R/KmR/[1 + (M)/KI] =
0.142 l/(h g)
and
k2S/KmS/[1 + (M)/KI] = 8.82 × 10−4 l/(h g)
if
(M) = 472.7 mM was employed for the alcoholysis. By consid-
ering effects of water content on the enzyme performance
via hydrolysis [26] as well as k2R/KmR = 0.173 l/(h g) and
k2S/KmS = 2.12 × 10−3 l/(h g) in water-saturated MTBE that were
comparable to those via alcoholysis in anhydrous MTBE, water
playing the same role as methanol on inhibiting the lipase activity
was deducible.
4. Conclusions
[20] E. Henke, S. Schuster, H. Yang, U.T. Bornscheuer, Monatsch. Chemie 131 (2000)
633–638.
[21] T. Miyazawa, K. Onishi, T. Murashima, T. Yamada, S.W. Tsai, Tetrahedron: Asym-
metry 16 (2005) 2569–2573.
[22] R. Morrone, G. Nicolosi, A. Patti, M. Piattelli, Tetrahedron: Asymmetry 6 (1995)
1773–1778.
With CALB-catalyzed alcoholysis of (R,S)-naproxenyl 1,2,4-
triazolide in anhydrous solvents as the model system, the optimal
conditions of 45 ◦C, MTBE as the solvent, and methanol as the
acyl acceptor were screened. A change of the leaving azole moi-
[23] S.W. Tsai, I.C. Cheng, C.M. Huang, Chem. Eng. Sci. 55 (2000) 4571–4582.

A.-C. Wu et al. / Journal of Molecular Catalysis B: Enzymatic 62 (2010) 235–241
241
[24] K. Faber, Biotransformation in Organic Chemistry, 5th ed., Springer-Verlag,
Berlin, 2004.
[25] U. Hanefeld, Org. Biomol. Chem. 1 (2003) 2405–2415.
[26] P.Y. Wang, Y.J. Chen, A.C. Wu, Y.S. Lin, M.F. Kao, J.R. Chen, J.F. Ciou, S.W. Tsai,
Adv. Synth. Catal. 351 (2009) 2333–2341.
[28] A.R. Katritzky, C.W. Rees, Comprehensive Heterocyclic Chemistry. Part 4A. Five-
membered Rings with Two or More Nitrogen Atoms, vol. 5, 1st ed, Pergamon
Press, Oxford, 1984, pp. 223, 384, 742.
[29] V. Leonard-Nevers, Z. Marton, S. Lamare, K. Hult, M. Graber, J. Mol. Catal. B:
Enzym. 59 (2009) 90–95.
[27] H.A. Staab, H. Bauer, K.M. Schneider, Azolides in Organic Synthesis and Bio-
chemistry, Weinheim, Wiley-VCH, 1998.
[30] J.A. Alkandary, A.S. Aljimaz, N.S. Fandary, M.A. Fahim, Fluid Phase Equilibr.
187–188 (2001) 131–138.