Dynamic Enzymatic Resolution of Thioesters
J. Am. Chem. Soc., Vol. 120, No. 23, 1998 5609
enantioselectivity made this a further challenging test for
dynamic resolution, as enrichment of substrate in the slower
reacting enantiomer over the course of the reaction resulting
from a slow rate of substrate racemization relative to hydrolysis
could in turn result in substantially diminished enantiomeric
purity of the product. Hydrolysis of the propargyl thioester 1l
under racemizing conditions taken to 95% completion gave
product in 80% ee. This is only slightly lower than the
theoretical value of 84% expected on the basis of the E value
for a reaction in which substrate remains racemic throughout
the reaction. The trifluoroethyl thioester, which has a greater
rate of racemization, gave product in 83% ee upon hydrolysis
to 97% completion under racemizing conditions. This is
equivalent to the theoretical value and demonstrates the effect
of racemization rate on enantiomeric purity of product.
In addition to the low rate of racemization, the dichlorophe-
noxypropionate thioesters posed further challenges due to their
susceptibility to nonenzymatic hydrolysis. This enhanced rate
of nonenzymatic hydrolysis relative to other thioesters studied
is attributed to the inductive effect of the dichlorophenoxy group,
which is much greater than the inductive effect of the phenyl
group. In proton exchange studies, the resonance effect of the
phenyl group on stabilization of the enolate makes up for the
smaller inductive effect so that the rate of R-proton exchange
is similar to that of the dichlorophenoxypropionate thioester.
The nonenzymatic hydrolysis precluded efficient enzymatic
dynamic resolution by hydrolysis. Instead, a transesterification
reaction was performed using n-butyl alcohol as the acyl
acceptor. Nonenzymatic transesterification was apparently less
prevalent than nonenzymatic hydrolysis, as the product was
obtained in moderate enantiomeric purity. A second advantage
of the transesterification procedure is that it provides a product
which can be used in a second enzymatic enantiomeric purity
enhancing step by hydrolysis using the same enzyme. This was
clearly demonstrated, as hydrolysis to 81% completion under
nonracemizing conditions increased the product ee from 75 to
93%.
transesterification reaction. These examples have demonstrated
that product can be obtained in good ee, even with substrates
for which racemization is slow and enzyme enantioselectivity
is modest.
This study demonstrates that substantial work is involved in
developing optimal enzymatic dynamic resolution procedures.
However, the potential to vary the choice of enzyme, the
substrate structure, and the reaction conditions should make this
a widely applicable and powerful technology.
Experimental Section
General Experimental. CH2Cl2 was distilled from calcium hydride
prior to use. 1H NMR were taken at 200 MHz and 13C NMR at 50
MHz using TMS as an internal standard. Lipase PS-30 (Pseudomonas
cepacia) was obtained from Amano. C. rugosa lipase as cross-linked
crystals was obtained from Altus. Other enzymes were from Sigma.
C. rugosa lipase immobilized on silica gel was prepared as previously
described.13
Synthesis of Thioesters of 2-Substituted Propionic Acids (1a-
m). To a solution of a 2-substituted propionic acid (10 mmol) in CH2-
Cl2 (50 mL) was added triethylamine (1.4 mL, 10 mmol), the mixture
was cooled to 4 °C, and oxalyl chloride (0.87 mL, 10 mmol) was added
dropwise over 15 min. After additional stirring for 2 h at 4 °C, a
solution of triethylamine (1.5 mL, 11 mmol) and the appropriate thiol
(11 mmol) in CH2Cl2 (5 mL) was added. After stirring overnight at
room temperature the solution was filtered, the filtrate was washed with
5% aqueous NaHCO3 (2 × 20 mL) and saturated aqueous NaCl (10
mL), dried over MgSO4, filtered, and concentrated. The crude product
was purified by column chromatography on silica gel using ethyl actate/
hexanes (1:9) as eluent. Products were obtained in 81-94% yield.
Propargyl Mercaptan (Prop-2-yne-1-thiol). To a solution of
propargyl chloride (2.07 mL, 30 mmol) in DMF (30 mL) was added a
solution of sodium thiophosphate26 (23.76 g, 60 mmol) in water (150
mL). The mixture was stirred for 5 h at room temperature, and the
pH was lowered to 4.0 with 1 N HCl. After the mixture was stirred
overnight at room temperature, the product, which separated off as a
yellow oil, was collected and further purified by distillation under
1
reduced pressure (1.4 g, 65%): bp 35-37 °C (100 mmHg); H NMR
(CDCl3) δ 3.28 (dd, 2H, J ) 2.7, 7.5 Hz), 2.32 (t, 1H, J ) 2.7 Hz),
2.08 (t, 1H, J ) 7.5 Hz); 13C NMR (CDCl3) δ 81.9, 70.7, 11.9.
(R,S)-Butyl 2-(2,4-Dichlorophenoxy)propanoate (5). To a solution
of a 2-(2,4-dichlorophenoxy)propionic acid (2.35 g, 10 mmol) in CH2-
Cl2 (50 mL) was added triethylamine (1.4 mL, 10 mmol). The mixture
was cooled to 4 °C, and oxalyl chloride (0.87 mL, 10 mmol) was added
dropwise over 15 min. After additional stirring for 2 h at 4 °C, a
solution of triethylamine (1.5 mL, 11 mmol) and 1-butanol (1.0 mL,
11 mmol) in CH2Cl2 (5 mL) was added. After the mixture was stirred
overnight at room temperature, the solution was filtered, the filtrate
was washed with 5% aqueous NaHCO3 (2 × 20 mL) and saturated
aqueous NaCl (10 mL), dried over MgSO4, filtered, and concentrated.
The crude product was purified by column chromatography on silica
gel using ethyl actate/hexanes (1:9) as eluent to give (R,S)-5 (2.4 g,
82%): 1H NMR (CDCl3) δ 7.40 (d, 1H, J ) 2.6 Hz), 7.17 (dd, 1H, J
) 2.6, 8.8 Hz), 6.80 (d, 1H, J ) 8.8 Hz), 4.75 (q, 1H, J ) 6.8 Hz),
4.18 (m, 2H), 1.70 (d, 3H, 6.8 Hz), 1.62 (m, 2H), 1.33 (m, 2H), 0.92
(t, 3H, J ) 7.4 Hz); 13C NMR (CDCl3) δ 171.4, 152.2, 130.3, 127.4,
126.9, 124.7, 115.9, 74.3, 65.3, 30.5, 18.9, 18.4, 13.6.
Conclusion
Studies reported here demonstrate that the choice of thiol
moiety can have a large effect on the R-proton acidity of
thioesters. The ethyl thioesters of some of the R-substituted
propionates studied in this work appear sufficiently acidic for
dynamic resolution while the acidity of thioesters of other acids
can be sufficiently enhanced by modification of the thiol moiety.
Thioesters of these thiols should be practical substrates for
enzymatic dynamic resolution.
This work has also shown that most hydrolytic enzymes will
readily accept thioesters as substrates. However, studies with
the C. rugosa lipase have shown that high activity and
enantioselectivity in oxoester hydrolysis does not guarantee high
activity and enantioselectivity in hydrolysis of a thioester of
the same acid. Enzymatic dynamic resolution may often require
the rediscovery of a suitable catalyst, even if one or more
enzymes have been previously identified for oxoester resolution.
Enzymatic dynamic resolution has been further demonstrated
in hydrolysis of a pair of R-phenylpropionate thioesters using
the protease subtilisin Carlsberg. Enzymatic dynamic resolution
has also been demonstrated in a transesterification reaction using
an alcohol rather than water as the acyl acceptor. This
minimizes nonenzymatic reactions, for substrates which are
prone to nonenzymatic hydrolysis under the basic racemizing
conditions, and facilitates further enhancement of enantiomeric
purity by enzymatic hydrolysis of the oxoester product of the
Deuterium Exchange Experiments. To a solution of the thioester
(0.25 mmol) and CD3OD (1.25 mmol) in toluene-d8 (0.5 mL) was added
trioctylamine (0.125 mmol). The decrease in intensity of the R-proton
1
signal was monitored over time by H NMR integration.
Measurement of Enzyme Activity for Oxoethyl and Thioethyl
Butyrate. To a biphasic mixture of aqueous PIPES buffer (20 mL,
0.01 M, pH 7.0) and 2 mL of toluene (or 2 mL of acetonitrile for
subtilisin Carlsberg) were added substrate (1 mmol) and enzyme. The
pH was maintained at 7.0 by addition of 0.02 M KOH solution using
a pH stat. Enzyme activity was determined by monitoring the volume
of base added vs time.
(26) Akerfeldt, S. Acta Chem. Scand. 1960, 14, 1980-1984.
Um, Pil-Je
