Enzymatic resolution of (RS)-2-arylpropionic acid thioesters by Candida rugosa lipase-catalyzed thiotransesterification or hydrolysis in organic solvents

Article abstract of DOI:10.1016/S0957-4166(98)00296-1

An enzymatic resolution process was developed to produce (S)-naproxen ester, (S)-naproxen or (S)-ibuprofen from the corresponding racemic thioesters by using lipase-catalyzed thiotransesterification or hydrolysis in organic solvents. Enzyme activity is greatly enhanced when activated naproxen thioesters containing an electron-withdrawing group are the substrates. Unlike other lipases, Candida rugosa lipase may discern the sulfur moiety of the thioesters, and yields lower enzyme activity when compared to the corresponding oxygen-containing analogues. Enzyme performances were further compared under various conditions, i.e. different combinations of reaction type (thiotransesterification or hydrolysis), solvent (isooctane or cyclohexane), substrate (naproxen or ibuprofen thioesters) and lipase sources.

Full text of DOI:10.1016/S0957-4166(98)00296-1

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 2799–2807
Enzymatic resolution of (RS)-2-arylpropionic acid thioesters by
Candida rugosa lipase-catalyzed thiotransesterification or
hydrolysis in organic solvents
∗
Chun-Sheng Chang, Shau-Wei Tsai and Chun-Nan Lin
Department of Chemical Engineering, National Cheng Kung University, Tainan, Taiwan 70101, ROC Taiwan
Received 19 May 1998; accepted 21 July 1998
Abstract
An enzymatic resolution process was developed to produce (S)-naproxen ester, (S)-naproxen or (S)-ibuprofen
from the corresponding racemic thioesters by using lipase-catalyzed thiotransesterification or hydrolysis in or-
ganic solvents. Enzyme activity is greatly enhanced when activated naproxen thioesters containing an electron-
withdrawing group are the substrates. Unlike other lipases, Candida rugosa lipase may discern the sulfur moiety of
the thioesters, and yields lower enzyme activity when compared to the corresponding oxygen-containinganalogues.
Enzyme performances were further compared under various conditions, i.e. different combinations of reaction
type (thiotransesterification or hydrolysis), solvent (isooctane or cyclohexane), substrate (naproxen or ibuprofen
thioesters) and lipase sources. © 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction
Hydrolytic enzymes such as esterases, proteases and, in particular, lipases have received much
attention because of their effectiveness in regio- and enantioselective hydrolysis, esterification and
transesterification of organic acids, alcohols, amines, amides and esters in the presence of organic
1,2
solvents. However, in the literature, very few studies deal with enzymatic transformations involving
thioacids and thioesters. Zaks and Klibanov, for instance, employed a porcine pancreatic lipase in the
thiotransesterification of methyl butyrate and butanethiol in various organic solvents and found rates
3
similar in magnitude to transesterification involving alcohols with the same enzyme. Caussette et al.
have successfully carried out the esterification of oleic acid and butanethiol catalyzed by an immobilized
Mucor miehei lipase in n-hexane, in which a low affinity of butanethiol to the enzyme was observed.4
Sproull et al. also considered the esterification of butanol and various thioacids with the same enzyme in
∗
Corresponding author. Fax: 886-6-2344496; e-mail: t62647@mail.ncku.edu.tw
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00296-1

2800
C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
a chloroform:hexane (25:75, v/v) mixture and found rates much faster than those for the corresponding
5
carboxylic acids.
In another study, Bianchi and Cesti considered the lipase-catalyzed thiotransesterification of mercapto
6
esters such as (RS)-3-(acetylthio)-2-methylpropionic acid methyl ester with propanol. They discovered
that porcine pancreatic lipase and Lipase P (Amano) have good enantioselectivity for the (R)-enantiomer
and obtained higher reaction rates by using the thio substrates instead of the corresponding oxa-
containing analogues. In the resolution of (RS)-2-octanol using (S)-ethyl thiooctanoate as an acyl donor,
7
Candida antarctica lipase was found to have high enantioselectivity for the (R)-enantiomer. Recently,
Tan et al. have employed Lipase P-30 (Amano) and trioctylamine as the catalysts for hydrolysis
and racemization, respectively, in the dynamic resolution of (RS)-2-(phenylthio)propionic acid ethyl
8
thioester. In their report, both a higher yield and higher enantiomeric excess for the desired enantiomer
than a normal resolution without racemization were obtained.
2
-Arylpropionic acids (i.e. profens), an important class of non-steroidal anti-inflammatory drugs, have
9
their pharmacological activity mainly on the (S)-enantiomer. Considerable efforts have been made on the
10
lipase-catalyzed enantioselective resolution of (S)-profens or their ester prodrugs. However, as far as we
know, there are still no reports of using profen thioesters as the starting material in the previous resolution
processes. Therefore, as shown in Scheme 1, with naproxen thioester 1a and ibuprofen thioester 1b as two
representative members of this class, an enzymatic resolution process was developed by using Candida
rugosa lipase-catalyzed thiotransesterification or hydrolysis in organic solvents.
Scheme 1.
2. Results and discussion
2.1. Naproxen thioester selection in thiotransesterification
With 4-morpholine ethanol as an acyl acceptor and Lipase MY (Meito Sangyo) as the biocatalyst in
isooctane, Table 1 shows the time–course variations of the enantiomeric excesses for the substrate (ees),
the main product of ester (eePM) and the byproduct of acid (ee ), as well as the conversion for the
PN
substrate (X ), the yields for the main product of ester (X ) and the byproduct of acid (X ) for various
t
M
N
naproxen thioesters. Compared with the enzyme performance in the enantioselective esterification of
racemic naproxen with ethanol, n-propanol or butanol,1 very low conversions of X and hence the
0e
t
enzyme activity were obtained when ethyl, n-propyl and n-butyl thioester were employed as the substrate
in thiotransesterification. When activated thioesters containing an electron-withdrawing moiety such as

C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
2801
Table 1
phenyl or trifluoroethyl group were used, great improvements in the substrate conversion were found
Table 1). These imply that, unlike other lipases,3,5,6 the oxygen hole for acyl binding in the active site
(
of Candida rugosa lipase may have difficulty in accommodating the large sulfur atom due to steric
and/or electrostatic effects and hence reduces the enzyme activity. This unfavorable factor is relaxed by
introducing an electron-withdrawing moiety in the thioesters. A detailed analysis of the yields (X and
M
X ) and enantiomeric excesses for the product (ee and ee ) for both activated thioesters indicates
N
PM
PN
that this enzyme has higher enantioselectivity in thiotransesterification and is more favorable when using
phenyl thioester as the substrate. However, this benefit is offset by its lower saturated solubility of 20.4
mM (compared with 290 mM for trifluoroethyl thioester) in isooctane at 37°C and the higher boiling
point of 169°C for thiophenol (35°C for trifluoroethanethiol). In order to decrease the hydrolysis side-
reaction, a further study on controlling a suitable water content of the crude lipase is needed.
The time–course variations of ee , ee , ee , X , X and X when racemic naproxen trifluoroethyl
s
PM
PN
t
M
N
ester was employed as the substrate are shown in Table 2. Comparing the results in both tables by
considering the applied enzyme concentration on the reaction rate, we conclude that Lipase MY has
higher enantioselectivity and activity when using the ester substrate instead of the corresponding sulfur-
containing analogue. This result indeed provides further strong evidence to support the above elucidation
of the lower enzyme activity of Candida rugosa lipase for thioesters.
By further inspecting the time–course yields in Table 2, a maximum of X =12.4% at 49 h for
N
the hydrolysis side-reaction was found. According to the Ping Pong kinetic mechanism as shown in
Scheme 2, the acyl donor (RCOOR ) at first liberates trifluoroethanol (HOR ) in the acylation step. Then,
1
1
the nucleophilic alcohol (HOR ) competes with water to form ester (RCOOR ) and acid (RCOOH),
2
2
respectively, in the deacylation step. Since trifluoroethanol is a good leaving group, one may regard
transesterification and hydrolysis to be irreversible. However, with the addition of an excess of 4-
morpholine ethanol and the formation of naproxen, a reversible lipase-catalyzed esterification between
the acid and the alcohol occurs to decrease X and enhances X . Finally, an equilibrium may exist among
N
M
water, 4-morpholine ethanol, naproxen and naproxen ester product. Therefore, further studies on finding

2802
C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
Table 2
optimal conditions such as best reaction time, substrate concentrations and water content are needed in
order to obtain the highest X and eePM.
M
Scheme 2.
When another Candida rugosa lipase purchased from Sigma was employed as the biocatalyst,
very similar enzyme performances for both lipases were obtained (Table 2). Therefore, we employed
this lipase preparation in the enantioselective thiotransesterification for both activated thioesters and
presented the time–course variations of eePM and X in Fig. 1(A) and 1(B). As expected, both enzymes
t
have very similar performances at a given thioester and organic solvent. However, increasing the solvent
hydrophilicity by replacing isooctane with cyclohexane results in the decrease of X for a given thioester
t
and enzyme preparation. This effect is especially evident for naproxen phenyl thioester [Fig. 1(A)],
although no explanations were given.

C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
2803
Fig. 1. Effects of solvent, lipase preparation, phenyl (A) and trifluoroethyl (B) thioesters on time–course variations of eePM and
−
1
X
X
t
for 3 mM thioester, 10 mM 4-morpholine ethanol and 5 mg ml lipase at 37°C. In isooctane: for Lipase MY, eePM (ꢀ) and
t
(ꢁ); for C. rugosa lipase, eePM (ꢂ) and X
(◦)
t
(•). In cyclohexane: for Lipase MY, eePM (5) and X
t
(ꢃ); for C. rugosa lipase,
eePM (4) and X
t
Table 3
2.2. Enantioselective hydrolysis for activated profen thio esters
By using Lipase MY as the biocatalyst, water contained in the crude enzyme (6.4%, w/w) and the
activated thioesters as substrates in isooctane, the time–course variations of the enantiomeric excess for
naproxen (ee ) and the conversion for the thioester (X ) are demonstrated in Table 3. After comparing the
P
t
results with those of thiotransesterification in Table 1, we conclude that Lipase MY has lower enzyme
enantioselectivity and activity for hydrolysis. Moreover, very little difference on the enzyme performance
was found when either phenyl thioester or trifluoroethyl thioester was the substrate. This result is also
valid when ibuprofen phenyl thioester is the substrate if enzyme performances for both phenyl thioesters
are compared by considering the effect of lipase concentration on enzyme activity (Table 3).
3. Conclusions
An enzymatic resolution process was developed to produce (S)-naproxen esters from racemic naproxen
thioesters by using Candida rugosa lipase-catalyzed thiotransesterification in organic solvents. Enzyme
activity is greatly enhanced when activated thioesters containing an electron-withdrawing group are the
substrate. Unlike other lipases, Candida rugosa lipase may discern the sulfur moiety of thioesters, and

2804
C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
Table 4
yields lower enzyme activity when compared with the result for the corresponding oxygen-containing
analogues. Lipase MY has been found to have higher enzyme enantioselectivity and activity when
thiotransesterification but not hydrolysis in isooctane was carried out. The enzymatic process was also
extended to obtain (S)-naproxen or (S)-ibuprofen by hydrolyzing the corresponding racemic activated
thioesters in isooctane, in which similar enzyme performances were found regardless of the source of
thioesters.
4. Experimental
4.1. General remarks on analytical procedure
The thiotransesterification and hydrolysis for naproxen thioesters or transesterification for naproxen
ester was monitored by HPLC using a chiral column (Chiralcel OD, Daicel Chemical Industries) capable
of separating the internal standard of 2-nitrotoluene, (R)- and (S)-naproxen as well as their derivatives
with retention time and capacity factor shown in Table 4. The mobile phase was a mixture of n-
−1
hexane:isopropanol:acetic acid (97:3:1, v/v) at a flow rate of 1.0 ml min . UV detection at 270 nm was
for quantification at the column temperature of 25°C. For monitoring the hydrolysis of ibuprofen phenyl
thioester, the same HPLC was employed except that a chiral column (Chiralcel OJ, Daicel Chemical
Industries), an internal standard of racemic flurbiprofen and the composition of the mobile phase (n-
hexane:isopropanol:acetic acid=100:0.4:0.4) were changed. The retention time and capacity factor for
(R)- and (S)-phenyl thioesters were also tabulated in Table 4.
4
.2. General procedure for the synthesis of racemic naproxen thioesters11
To 25 ml of ice-cooled 1,2-dimethoxyethane were added 200 mM of racemic naproxen, 600 mM of
anhydrous pyridine, 300 mM of phenyl dichlorophosphate and 320 mM of thiol, which were reacted
at room temperature for 16 h with stirring. The resulting solution was added to 50 ml of ice-cooled
NaOH solution (1 M), and then 70 ml of chloroform for extraction. The organic layer was separated

C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
2805
Table 5
and washed twice with 50 ml of 1 M NaOH solution and twice with 50 ml of saturated NaCl solution,
dried over MgSO , filtered and concentrated under vacuum. The resulting oil was purified by silica gel
4
liquid chromatography with the mobile phase of n-hexane:ethyl acetate (5:1, v/v), and concentrated in
1
vacuum. The desired racemic naproxen thioesters were confirmed from the retention time in HPLC. H
NMR spectra were also recorded at 200 MHz on a Bruker AC-200 spectrometer in deuteriochloroform
solutions with tetramethylsilane as an internal standard. Chemical shifts are reported in ppm (from TMS)
in Table 5. When peak multiplicities are reported, the following abbreviations are used: s, singlet; d,
doublet; t, triplet; q, quartet; m, multiplet.
4
.3. General procedure for the synthesis of racemic naproxen trifluoroethyl ester12
To 20 ml of benzene were added 15 mmol of racemic naproxen and 27 mmol of thionyl chloride, which
were heated to 80–90°C with reflux for 1.5 h. The resulting solution was concentrated under vacuum,
and 30 ml of benzene, 15 mmol of anhydrous pyridine and 27 mmol of trifluoroethanol were added.
The mixture was refluxed at 80–90°C for 4 h. The reaction mixture was cooled to room temperature and
extracted four times with 50 ml of NaOH solution (1 M) and twice with 100 ml of deionized water. The
organic layer was separated and dried over MgSO , centrifuged and concentrated under vacuum. The
4
resulting oil was purified by silica gel liquid chromatography with the mobile phase of n-hexane:ethyl
acetate (5:1), and concentrated under vacuum. The desired racemic naproxen trifluoroethyl esters, with
1
H NMR (CDCl ) data shown in Table 5, were confirmed from the retention time in HPLC.
3

2806
C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
4.4. General procedure for the thiotransesterification of racemic naproxen thioesters
To 15 ml of isooctane were added 3 mM of racemic naproxen thioester, 10 mM of 4-morpholine
−1
ethanol and 20 mg ml of Lipase MY (Meito Sangyo), which were reacted with stirring at 37°C.
The sample was injected onto the above HPLC system at different time intervals. The active naproxen
thioester was determined from the yield and the enantiomeric excess for the ester product. Similar
experiments were performed with activated thioesters as the substrate and Candida rugosa lipase (Sigma)
in cyclohexane (or isooctane) to study effects of lipase sources, solvent hydrophilicity and type of
activated thioester on enzyme performances.
4.5. General procedure for the transesterification of racemic naproxen trifluoroethyl ester
To 15 ml of isooctane were added 2 mM of racemic naproxen trifluoroethyl ester, 15 mM of 4-
−1
morpholine ethanol and 5 mg ml of Lipase MY or C. rugosa lipase, which were reacted with stirring
at 37°C. The sample was injected onto the above HPLC system at different time intervals, from which
yields and the enantiomeric excesses for the ester and acid products were calculated.
4.6. General procedure for the hydrolysis of racemic naproxen thioesters
To 15 ml of isooctane were added 20 mg ml 1 of Lipase MY and 3 mM of activated naproxen
−
thioesters, which were reacted with stirring at 37°C. The sample was injected onto the above HPLC
system at different time intervals, from which the yield and the enantiomeric excess for the acid product
were calculated.
4
.7. General procedure for the synthesis of racemic ibuprofen phenyl thioester11
To 10 ml of ice-cooled 1,2-dimethoxyethane were added 200 mM of racemic ibuprofen, 600 mM
of anhydrous pyridine, 300 mM of phenyl dichlorophosphate and 400 mM thiol, which were heated
with stirring to room temperature over 12 h. To the resulting solution was added 20 ml of ice-cooled
NaOH solution (1 M), and then 25 ml of chloroform for extraction. The organic layer was separated and
washed twice with 20 ml of 1 M NaOH solution and twice with 25 ml of saturated NaCl solution, dried
over MgSO , filtered and concentrated under vacuum. The resulting oil was purified by silica gel liquid
4
chromatography with the mobile phase of n-hexane:ethyl acetate (50:3, v/v), and concentrated under
1
vacuum. The desired racemic ibuprofen phenyl thioester, with H NMR (CDCl ) data shown in Table 5,
3
was confirmed from the retention time in HPLC.
4.8. General procedure for the hydrolysis of ibuprofen phenyl thioester
To 20 ml of isooctane were added 50 mg ml 1 of Lipase MY and 1.2 mM of ibuprofen phenyl thioester,
−
which were reacted with stirring at 37°C. The sample was injected onto the above HPLC system at
different time intervals, from which the yield and the enantiomeric excess for the acid product were
calculated.

C.-S. Chang et al. / Tetrahedron: Asymmetry 9 (1998) 2799–2807
2807
Acknowledgements
The financial support from the Chinese National Science Council with Grant NSC87-2214-E-006-006
is appreciated.
References
1
2
3
4
5
6
7
8
9
. Wong, C. H.; Whitesides, G. M. Enzymes in Synthetic Organic Chemistry; Elsevier Science: New York, 1994; pp. 41–111.
. Faber, K. Biotransformations in Organic Chemistry; Springer-Verlag: New York, 1992; pp. 23–125.
. Zaks, A.; Klibanov, A. M. Proc. Natl. Acad. Sci. USA 1985, 82, 3192–3196.
. Sproull, K. C.; Bowman, G. T.; Carta, G.; Gainer, J. L. 1997, 13, 71–76.
. Caussette, M.; Marty, A.; Combes, D. J. Chem. Tech. Biotechnol. 1997, 68, 257–262.
. Bianchi, D.; Cesti, P. J. Org. Chem. 1990, 55, 5657–5659.
. Frykman, H.; Ohrner, N.; Norin, T.; Hult, K. Tetrahedron Lett. 1993, 8, 1367–1370.
. Tan, D. S.; Gunter, D. G.; Drueckhammer, D. G. J. Am. Chem. Soc. 1995, 117, 9093–9094.
. Hutt, A. J.; Caldwell, J. Clin. Pharmacokin. 1984, 9, 371–373.
10. (a) Sih, C. J.; Gu, Q. M.; Fuelling, G.; Wu, S. H. Dev. Ind. Microbiol. 1988, 29, 221–229. (b) Mustranta, A. Appl.
Microbiol. Biotechnol. 1992, 38, 61–66. (c) Palomer, A.; Cabre, M.; Ginesta, J.; Mauleon, D.; Carganico, G. Chirality
1
993, 5, 320–328. (d) Arroyo, A.; Sinisterra, J. V. J. Org. Chem. 1994, 59, 4410–4417. (e) Tsai, S. W.; Wei, H. J. Enzyme
Microb. Technol. 1994, 16, 328–333. (f) Morrone, R.; Nicolosi, G.; Patti, A.; Piattelli, M. Tetrahedron: Asymmetry 1995,
, 1773–1778. (g) Chang, C. S.; Tsai, S. W. Enzyme Microb. Technol. 1997, 20, 635–639. (h) Tsai, S. W.; Lin, J. J.; Chang,
6
C. S.; Chen, J. P. Biotechnol. Prog. 1997, 13, 82–88.
1. Liu, H. J.; Sabesan, S. I. Can. J. Chem. 1980, 2645–2648.
2. Allenmark, S.; Ohlsson, A. Chirality 1992, 4, 98–102.
1
1