Evaluation of Pseudoenantiomeric Mixed Carbonates as Efficient Fluorogenic Probes for Enantioselectivity Screening

Article abstract of DOI:10.1002/cbic.201500509

We report mixed carbonates as enzyme substrates and demonstrate their application as fluorogenic probes for lipase and esterase enantiopreference screening. By the application of pseudoenantiomers with distinct fluorescence behaviors, it is possible to evaluate the activity and enantiopreference of hydrolytic enzymes. In order to validate our method, enantioselectivities calculated from fluorometric measurements were compared with the results obtained from larger-scale kinetic resolution. Pseudoenantiomeric mixed carbonates with distinct fluorogenic properties are shown to be very useful probes for screening lipase and esterase enantioselectivity. The simultaneous use of pseudoenantiomeric substrates overcomes conventional assays with single enantiomer substrates, as the effects of competitive binding are taken into account.

Full text of DOI:10.1002/cbic.201500509

DOI: 10.1002/cbic.201500509
Full Papers
Evaluation of Pseudoenantiomeric Mixed Carbonates as
Efficient Fluorogenic Probes for Enantioselectivity
Screening
[
a]
Anna Zadlo, Dominik Koszelewski, Filip Borys, and Ryszard Ostaszewski*
We report mixed carbonates as enzyme substrates and demon-
strate their application as fluorogenic probes for lipase and es-
terase enantiopreference screening. By the application of pseu-
doenantiomers with distinct fluorescence behaviors, it is possi-
ble to evaluate the activity and enantiopreference of hydrolytic
enzymes. In order to validate our method, enantioselectivities
calculated from fluorometric measurements were compared
with the results obtained from larger-scale kinetic resolution.
Introduction
Biocatalysis is becoming a key synthesis method for complex
compounds with high optical purity, with a place alongside
conventional chemocatalysis and chromatographic separation.
Enantiomerically pure secondary alcohols are undoubtedly one
of the most important broad classes of building blocks for the
total synthesis of medicaments, agricultural chemicals, liquid
crystals, organic nonlinear materials, and other fine chemicals.
The resolution of two enantiomers from racemic mixture is of
great importance, particularly in the pharmaceutical industry
because of the different pharmacological activities and phar-
macokinetic characteristics of each enantiomer. (For many
pharmaceuticals, only one enantiomer is active; the other can
ic selectivity is as the ratio of specificity constants [Equa-
[
3]
tion (1)]:
R
S
ðkcat=K
ðkcat=K
Þ
m
m
E ¼
ð1Þ
Þ
Unfortunately, measurement of the kinetic parameters for
each enantiomer separately ignores competitive binding to
the active site of the enzyme. This results in some differences
between values calculated from high-throughput measure-
ments and from larger-scale experiments for racemates. Al-
though many reports are available, most are based on single
enantiomer measurements. The “Quick E” method, reported by
[1]
cause adverse side effects.)
[
4]
Lipase- or esterase-catalyzed reactions have been applied to
solve a number of synthetic problems, one of which is the ki-
Janes and Kazlauskas, takes into consideration the competi-
tion between enantiomers. The addition of an achiral reference
compound was proposed as a solution to the problem
[Eq. (2)]:
[
2]
netic resolution of racemic mixtures of secondary alcohols.
However, the properties of enzymes do not always comply
with the requirements of an industrial application. Therefore,
the discovery of new, more efficient, and stereoselective var-
iants of enzymes by genome mining and protein engineering
is desirable. Recombinant DNA technology has enabled the in-
crease in gene sequences and the generation of enzyme libra-
vS=v
ref
Quick E ¼
ð2Þ
vR=v
ref
However, the reference compound is structurally different
from the substrates, so it does not mimic the real racemic
sample.
6
ries containing many random mutations. Therefore, up to 10
mutants are needed to be characterized. To avoid time-con-
suming work-up followed by determining conversions and
enantiomeric excesses of a huge number of enzymatic trans-
formations, approximate enantioselectivity (E) values can be
determined by high-throughput screening. Such an approach
affords selection of the most stereoselective enzymes and con-
siderably reduces the number of samples to be investigated by
chromatographic methods. A common expression of enzymat-
High-throughput process monitoring is possible with instru-
[
5]
ment assays, such as by chromatographic methods, IR spec-
[
6]
[7]
[8]
troscopy, mass spectrometry, capillary electrophoresis, chi-
[
9]
[10]
roptical methods, pH-stat assay, or (most popular) UV–visi-
[
11,12]
ble and fluorescence spectroscopies.
Most high-through-
put screening methods use chromogenic and fluorogenic sub-
strates, because of the real-time measurement capability.
Fluorescence assays have many advantages, for example, they
are very sensitive, specific, simple, and fast. Additionally, enzy-
matic activity can be measured in biological samples, because
[
a] A. Zadlo, Dr. D. Koszelewski, F. Borys, Prof. R. Ostaszewski
Institute of Organic Chemistry, Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
[13]
this method is less susceptible to interference.
Fluorogenic and chromogenic probes typically use umbelli-
feryl and nitrophenyl esters, as well as sulfates, phosphates,
and glycosides. Unfortunately, these substrates are not stable
E-mail: r.ostaszewski@icho.edu.pl
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201500509.
ChemBioChem 2016, 17, 71 – 76
71
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
(
tend to high spontaneous and nonspecific hydrolysis in aque-
dure. Butan-2-ol was reacted with triphosgene, thereby leading
to chlorocarbonate, followed by reaction with 4-methylumbel-
liferone or 4-trifluorometylumbelliferone. The desired umbelli-
feryl carbonates (1, 2) were obtained in up to 47% yield. Car-
bonates of 4-methylumbelliferone and 4-trifluorometylumbelli-
ferone are nonfluorescent compounds, whereas enzymatic
transformation of these probes to free 4-methylumbelliferone
or 4-trifluorometylumbelliferone causes strong fluorescent
emission. Compared to assays based on coumarin ethers,
which lead to harmful aldehydes or ketones (which might also
be enzymes inhibitors), our procedure is environmentally
friendly as it creates nontoxic coumarin, alcohol, and carbon
dioxide (Scheme 1).
ous buffer). Reymond and co-workers reported a procedure
based on a chiral secondary alcohol linked to an umbelliferone
[
14,15]
with an ether bond.
The nonfluorescent substrate was oxi-
dized to the corresponding ketone by an alcohol dehydrogen-
ase or by a sodium periodate, then treatment with bovine
serum albumin released the fluorescent umbelliferone by b-
elimination. This approach was adapted for a variety of en-
[
16–21]
zymes.
Available compounds did not allow screening the
activity of enzymes toward secondary alcohols, because the
alkoxy group function was provided by the fluorophore mole-
cule.
Recently, we reported a technique, based on replacement of
an ester bond with an isosteric carbonate bond, that allowed
spectrofluorometric monitoring of the enzymatic reaction
[22]
course with achiral and chiral alcohol derivatives. A set of
mixed carbonates of 4-methylumbelliferone and alcohols was
synthesized. We demonstrated their application as fluorescent
probes for lipase and esterase activity and enantiopreference
screening. These compounds represent a new class of fluoro-
genic probes that are particularly suited for screening enzymes
toward alcohols. The probes were easy to synthesize, not sus-
ceptible to nonspecific degradation, and are highly recom-
mended for high-throughput screening and for in vivo imag-
ing of enzyme activities in cells.
Here, we demonstrate the application of pseudoenantiomer-
ic pairs of substrates for determining the enantiopreference of
hydrolytic enzymes. Fluorogenic probes that differ in fluores-
cent behavior but are structurally similar can be mixed and fol-
lowed at two different emission wavelengths. The application
of two pseudoenantiomers in one reaction mixture mimics the
real sample conditions. The results obtained from fluorometric
measurements with pseudoracemic mixtures correlated well
with the results obtained from kinetic resolution experiments
for racemic mixtures.
3
Scheme 1. Synthetic pathway. a) triphosgene, pyridine; b) 3 or 4, Et N,
DMAP, THF; c) lipase or esterase, PBS, pH 7.4.
In general, the effective control and modeling of bioprocess-
es is possible only with systems exhibiting high sensitivity,
high specificity and providing real-time measurement. These
requirements are fulfilled by fluorescence spectroscopy. A hy-
drolytic process that takes place over minutes can be followed
by measuring changes in fluorescence intensity and emission
wavelength over time. 4-Substituted umbelliferones released
during the enzymatic reactions vary in their fluorescence spec-
tra. Dual-wavelength spectrofluorimetric tracking of two fluo-
rophores enabled the rapid determination of enantioselectivity
of the process. It was noticed that signals overlapped in some
ranges, therefore conditions were selected so as to permit
measurement of both pseudoenantiomers simultaneously:
data were recorded at l =360 nm, l =430 nm for the
Results and Discussion
We report the evaluation of mixed carbonates as pseudoenan-
tiomeric substrates for the determination of the enantioprefer-
ences of lipases and esterases. A pair of compounds that have
the opposite configuration at the stereogenic center and differ
[23–25]
by one substituent are called pseudoenantiomers.
If the
products of hydrolysis have different spectroscopic properties,
it is possible to determine reaction rates in real time for both
pseudoenantiomers in one reaction vessel. In order to provide
different wavelengths of maximum fluorescence emission for
pseudoenantiomeric substrates, coumarins carry different sub-
stituents (X=CH or X=CF ); these do not differ much in their
ex
em
methyl derivative and l =360 nm, l =550 nm for the tri-
ex
em
fluoromethyl derivative. By taking into account the difference
between fluorescence quantum yields (F =0.63 for 4-methyl-
F
[
26]
umbelliferone; F =0.34 for 4-trifluoromethylumbelliferone),
3
3
F
molecule size, but have different emission properties. The ap-
we were able to estimate the enantiopreference in experi-
ments with pseudoenantiomers.
plication of probes that release 4-methylumbelliferone (l =
ex
3
60 nm, lem =445 nm (blue emission)), and 4-trifluoromethy-
Compounds 1 and 2 vary in both absolute configuration
lumbelliferone (l =385 nm, lem =500 nm (green)) enables
and labeling at the functional group CX . The derivatives were
ex
3
monitoring reaction rates for both compounds simultaneously
at two different wavelengths.
prepared in enantiomerically pure form and assayed with a
series of commercially available lipases and esterases. Initially,
the fluorescence measurements were conducted for each
enantiomer separately (initial reaction rates in Table 1): 1 and 2
Various carbonates of 4-methylumbelliferone and 4-trifluoro-
metylumbelliferone were obtained by a straightforward proce-
ChemBioChem 2016, 17, 71 – 76
www.chembiochem.org
72
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
The hydrolysis rates for simple enantiomers (Table 1) were
compared with the rates measured for pseudoracemic sub-
strates (Table 2). In most cases, the pseudoenantiomeric mix-
ture resulted in a decrease in both reaction rates. This was as
expected, because the presence of a second enantiomer
caused competitive binding to the active site of the enzyme.
Where the rates between compounds with the same configu-
rations were similar, the results correlated well. Large differen-
ces in hydrolysis rates between 4-methyl and 4-trifluoromethyl
derivatives resulted in weaker correlation, including a shift in
enantiopreference. For example, with lipoprotein lipase from
Burkholderia sp. and esterase from Pseudomonas fluorescens,
the 4-trifluorometyl derivative was hydrolyzed several times
faster than the 4-methyl derivative, because of the chemo-
selectivity of the enzymes.
Table 1. Initial rates of reaction of fluorogenic substrates 1 and 2.
À1
Lipase (L) or esterase (E)
V
0
[nms
]
(R)-1
(S)-1
(R)-2
(S)-2
0.5
121
49
24
10
8
8
131
1
10
67
11
–
0.3
92
0.3
0.5
97
49
35
13
10
14
99
1
Candida rugosa L.
porcine liver E.
porcine pancreas L.
Candida antarctica L.
horse liver E.
126
142
20
12
11
29
139
3
154
36
10
13
20
95
2
1
10
3
hog pancreas L.
Candida cylindracea L.
Pseudomonas cepacia L.
Pseudomonas fluorescens E.
Burkholderia sp. lipoprotein L.
Pseudomonas sp. lipoprotein L.
3
7
7
6
64
3
À1
Conditions: 50 mm substrate, 50 mgmL enzyme in PBS (100 mm, pH 7.4),
3
08C.
In order to omit the chemoselectivity aspects but retain the
competitive action, we adapted the Janes equation by replac-
ing the rate for reference compound with the rate of the op-
posite-configured pseudoenantiomer [Eq. (3)]:
exhibited high specific reactivity with enzymes and acceptable
stability against autohydrolysis. The rates for compounds with
the same configuration ((R)-1 and (R)-2) were similar, with
three exceptions: reactions catalyzed by porcine liver esterase,
lipoprotein lipase from Burkholderia sp., and esterase from
Pseudomonas fluorescence. Although the CH and CF groups
vS=v
pseudoR
Quick E ¼
ð3Þ
vR=v
pseudoS
Larger-scale experiments were devised to assess the useful-
ness and applicability of this method. Equimolar amounts of
enantiomers (R)-1 and (S)-1 were mixed to form a racemate.
Then, the kinetic resolution processes were started by the ad-
dition of hydrolytic enzymes. The reactions lasted from 2 h to
four days. As analytical HPLC was used to establish the un-
reacted substrates, reactions was carried out until conversion
exceeded 50%. The fluorometric and preparative scale meas-
urements can be compared in Table 3. Enantioselectivity (E)
was calculated with the equation of Chen et al., based on de-
3
3
are far from the reaction center, they affected enzyme activity.
The results imply that these enzymes are very specific, and
[
27]
that a change in substrate structure affects the reaction rate.
A commonly used method to avoid the problem of competi-
tive binding at the active site is the addition of a reference
compound. But a structurally different achiral substrate does
not really mimic the conditions of the kinetic resolution pro-
cess. We employed the pseudoenantiomeric 4-trifluoromethyl-
umbelliferone derivative as the reference, as it is chirally and
structurally similar to the substrate. Mixing pairs of pseudo-
enantiomers (1:1, simulating a racemate) enabled monitoring
the reaction rates for both substrates simultaneously (Table 2).
The time-dependent increase in fluorescence intensity at two
different emission wavelengths allowed the enantiopreference
of the selected enzymes to be determined.
[
28]
termined enantiomeric excess (ee) and conversion (c).
Only a few reports have described the effective kinetic reso-
[
29]
lution of butan-2-ol. Determining the enantiomeric excess of
butan-2-ol was also problematic. With our method, derivatiza-
tion of butan-2-ol with 4-methylumbelliferone resulted in very
good chiral HPLC separation. The retention times between
enantiomers differed by 78 s. The use of fluorophores as deri-
vatizing agents is common to achieve effective separation of
[
30]
compounds on chromatographic columns.
Table 2. Initial rates of reactions of 1 and 2 with various enzymes.
A comparison of the larger-scale E values with Quick E
values from fluorometric measurements revealed good correla-
tion. The pseudoracemic mixture simulated a racemic mixture
À1
Lipase (L) or esterase (E)
V
(S)-2
0
[nms
]
(
R)-1
(R)-2
(S)-1
0.3
137
50
11
6
10
6
103
3
–
0.3
0.5
125
53
11
9
7
4
125
1
26
87
0.5
successfully. This contrasts with the approximate E values (Eapp
)
Candida rugosa L.
Porcine liver E.
Porcine pancreas L.
Candida antarctica L.
Horse liver E.
62
49
31
9
78
64
18
8
15
9
50
2
12
104
calculated from data in Table 1 (shown as the ratio of relative
values) which indicated high variation in the results. The Quick
E values were closer to those calculated from kinetic resolu-
tion, thus proving the superiority of this method over the Eapp
method. Variations in the E values can arise from small differ-
ences in initial reaction rates, and ratios between relative rates
cannot be used to estimate enzymatic enantioselectivity in our
case.
8
8
Hog pancreas L.
Candida cylindracea L.
Pseudomonas cepacia L.
Pseudomonas fluorescens E.
Burkholderia sp. lipoprotein L.
Pseudomonas sp. lipoprotein L.
40
3
4
4
9
0.3
12
0.6
0.7
0.5
The results indicate that 4-substituted umbelliferon units not
only are good derivatizing agents for chiral HPLC resolution,
but also positively affect the kinetic resolution process. Our ap-
À1
Conditions: 50 mm substrate, 50 mgmL enzyme in PBS (100 mm, pH 7.4),
08C.
3
ChemBioChem 2016, 17, 71 – 76
www.chembiochem.org
73
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
ture turned green, so the enzyme hydrolyzed substrate (S)-2 at
a higher rate. Similar results were obtained for a mixture of (R)-
Table 3. Enantioselectivity (E) calculated from larger-scale kinetic resolu-
tion and Quick E calculated from fluorometric measurements.
2
and (S)-1.
[
a]
[b]
[c]
Lipase (L) or esterase (E)
t
c
ee
s
E
Quick E
E
app
This methodology is a useful for real-time monitoring of the
[
h] [%] [%]
enzymatic processes and takes into account competitive bind-
ing between enantiomers to the active site of the enzyme. We
have demonstrated its use with pseudoenantiomeric, chiral,
fluorogenic carbonates as probes for a rapid determination of
the enantiopreference of hydrolytic enzymes. This class of
probes is easy to synthesize, and might be applicable for high-
throughput formats and for in vivo imaging of enzyme activity
in cells.
Candida rugosa L.
porcine liver E.
porcine pancreas L.
Candida antarctica L.
horse liver E.
hog pancreas L.
Candida cylindracea L.
Pseudomonas cepacia L.
4
2
80
69
37 1.6 (S)
1.1 (R)
44 3.2 (R)
62 1.6 (R)
3.5 (S)
1.2 (R)
4.6 (R)
1.3 (R)
1.7 (R)
3.0 (R)
6.4 (S)
2.0 (R)
2.2 (S)
1.6 (R)
1.4 (S)
1.1 (R)
1.8 (R)
1.2 (S)
1.2 (R)
1.5 (S)
1.5 (S)
1.6 (S)
3.9 (S)
1.4 (R)
6
72 55
6
4
48 74
72 84
4
94
59
3
1.1 (R)
83 4.2 (R)
>99 5.8 (S)
52
4
5
1.1 (R)
1.1 (R)
Pseudomonas fluorescens E. 96 57
Burkholderia sp.
lipoprotein L.
4
97
47 1.3 (R)
Pseudomonas sp.
lipoprotein L.
2
76
18 1.3 (R)
1.4 (R)
2.5 (S)
Conclusion
À1
Conditions: 50 mm substrate, 50 mgmL enzyme in PBS (100 mm, pH 7.4),
08C. [a] Calculated from HPLC analysis from ee and c. [b] Calculated
Mixed carbonates of chiral alcohols and 4-methylumbelliferone
and 4-trifluoromethylumbelliferone fluorophores have been
shown to be good enzyme substrates and fluorogenic probes
for screening lipase and esterase enantiopreference. Measuring
reaction rates for each enantiomer separately does not account
for competition to the active site of enzyme; this is achievable
with our pseudoracemic mixture. The presence of the oppo-
sitely configured pseudoenantiomer in the reaction vessel si-
mulated a racemic mixture. With the application of two differ-
ently labeled pseudoenantiomers, we have demonstrated
a useful tool for real-time monitoring of enzyme activity at two
emission wavelengths and estimating preference. The results
from the fluorometric measurements correlated well with
those from the larger-scale assay. The preparative-scale ap-
proach can be used in the kinetic resolution of a broad range
of chiral sec-alcohols.
3
from fluorometric measurements from data in Table 2 (e.g., Candia rugosa
L.: Quick E=((S)-1/(R)-2)/((R)-1/(S)-2)=(125/62)/(78/137)=3.5). [c] Calculat-
ed from data in Table 1 Eapp =Vrelfast/Vrelslow; Vrel =(V
Candia rugosa L.: Eapp =(126–0.3/0.3)/(92–0.3/0.3)=1.4).
0
ÀVnoE/VnoE) (e.g.,
proach can be used in the kinetic resolution of wide spectrum
of chiral sec-alcohols. Kinetic resolution catalyzed by Candida
cylindracea lipase gave the best result. The enantiomeric
excess of recovered substrate was better than 99%, which
gives a promising perspective for further investigations.
Our method is easy to apply in any laboratory, and does not
require an expensive (and not always available) spectropho-
tometer. The photograph of cuvettes (Figure 1) taken 5 min
after addition of C. cylindracea lipase to a solution of 1 and 2
demonstrates that enzyme activity that can be observed under
a UV-vis lamp. A comparison of light intensity for probes (R)-
1
and (S)-1 (similarly for (R)-2 and (S)-2) confirmed that C. cylin- Experimental Section
dracea lipase prefers the (S)-configured probes. When (R)-1 (re-
leasing blue 4-methylumbelliferone) and (S)-2 (releasing green
Materials and methods: Commercial enzyme preparations were
purchased from Sigma–Aldrich. All chemicals were commercial
products and of analytical grade. Melting points were determined
with Büchi SMP-20, specific optical rotation with Jasco P-2000.
4
-trifluoromethylumbelliferone) were mixed, the reaction mix-
1
13
H NMR and C NMR spectra were recorded on 200 and 400 MHz
spectrometers. UV/Vis spectra were recorded at ambient tempera-
ture on a U-1900 spectrophotometer (Hitachi) in 1 cm plastic cells.
Fluorescence spectra, fluorescence time profiles, and 2D measure-
ments were recorded at 308C in quartz cuvettes in an F7000 spec-
trofluorometer (Shimadzu). Enantiomeric separations of (R)-1 and
(S)-1 were performed by HPLC on a Varian ProStar instrument
equipped with Chiralpak AD-H column (Ø 4.6”250 mm, l=
2
1
70 nm; Daicel Chemical Industries, Osaka, Japan) at 258C with a
mLmin flow rate, and eluted in hexane/isopropanol (70:30,
À1
v/v).
Synthetic procedure
CAUTION: Phosgene (b.p. 88C) is highly toxic and corrosive, and
reacts violently with many nucleophiles. Excess phosgene (formed
by the amine-mediated decomposition of triphosgene) was caught
in the primary liquid nitrogen trap and destroyed when still cold in
CH Cl by dropwise addition to a vigorously stirred, cold mixture of
Figure 1. Enzymatic assay for substrates 1 and 2 with C. cylindracea lipase
À1
(
l
ex =366 nm). Candida rugosa lipase (50 mgmL ) in PBS (pH 7.4) was in-
2
2
cubated with A) 50 mm (R)-1, B) 50 mm (S)-1, C) 50 mm (R)-2, D) 50 mm (S)-2,
E) 50 mm (R)-1 and 50 mm (S)-2, or F) 50 mm (S)-1 and 50 mm (R)-2.
2-aminoethanol or piperidine (5 mL) and ethanol (30 mL) in di-
chloromethane (100 mL) in a well-ventilated hood.
ChemBioChem 2016, 17, 71 – 76
www.chembiochem.org
74
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Pyridine (40 mg, 0.5 mmol) was added dropwise to alcohol
0.5 mmol) and triphosgene (75 mg, 0.25 mmol) in CH Cl (10 mL),
4.90–4.80 (m, 1H), 1.82–1.63 (m, 2H), 1.41–1.36 (m, 3H), 1.00 (t, J=
13
(
7.5 Hz, 3H); C NMR (CDCl , 100 MHz): d=158.4, 155.1, 154.4,
2
2
3
and the mixture was stirred at room temperature for 2 h to afford
the chlorocarbonate in CH Cl . A mixture of DMAP (25 mg,
152.1, 126.3, 122.8, 120.0, 118.4, 115.3, 111.2, 110.5, 78.8, 28.7, 19.2,
9.5; UV/Vis (lmax, nm, phosphate buffer pH 7.4, loge): 215 (3.8), 285
(3.7), 315 (3.7); elemental analysis calcd (%) for C H F O : C 54.55,
2
2
0
.2 mmol), Et N (0.42 mL, 3 mmol), and 7-hydroxy-4-methyl-2H-
3
15 13
3
5
chromen-2-one (176 mg, 1 mmol) or 7-hydroxy-4-trifluoromethyl-
H-chromen-2-one (230 mg, 1 mmol) in anhydrous CH Cl2 (2 mL)
H 3.97; found: C 54.75, H 4.02.
2
2
Enzyme assay: Substrates were prepared as stock solutions
was slowly added to the CH Cl solution of chlorocarbonate cooled
2
2
(
5 mm) in acetonitrile, and tested with commercially available lipas-
to 08C. The mixture was stirred for 2 h. The resulting suspension
was diluted with CH Cl (10 mL) and then HCl (10 mL, 1m). The or-
es and esterases. Substrate (50 mm) was added to enzyme (100 mg)
in a phosphate buffer (2 mL, 100 mm, pH 7.4). The fluorescence
time profile was recorded in a cuvette in an F7000 spectrofluorom-
eter (Shimazu). The fluorescence data were acquired over one
hour, which was sufficient for completion of the reaction with the
more reactive enzymes. Fluorescence data were converted into 4-
metylumbelliferone and 4-trifluoromethylumbelliferone concentra-
tions by means of a calibration curve. The linear portion of the
curve was used to obtain the reaction rate.
2
2
ganic phase was separated, dried over MgSO , and evaporated
4
under reduced pressure. The crude product was purified by
column chromatography on silica gel (CH Cl /MeOH, 99:1, v/v).
2
2
(R)-2-Butyl-(2-oxo-4-methyl-2H-chromen-7-yl) carbonate (R)-1:
The procedure was applied starting with (R)-butan-2-ol (74 mg,
1
2
.0 mmol) and 7-hydroxy-4-methyl-2H-chromen-2-one (352 mg,
mmol) to yield carbonate (R)-5 as a white solid; yield 130.5 mg
1
(
47%); m.p. 51–548C; [a] =À11.9 (c=1.0, CHCl ); H NMR (CDCl ,
À1
D
3
3
Larger-scale procedure: Equimolar (R)-1 and (S)-1 (4 mgmL )
4
00 MHz): d=7.61 (d, J=8.8 Hz, 1H), 7.22 (d, J=2.2 Hz, 1H), 7.17
were diluted in PBS (4 mL, 100 mm, pH 7.4) with acetonitrile (80:20,
(
dd, J =8.7 Hz, J =2.3 Hz, 1H), 6.27 (d, J=1.2 Hz, 1H), 4.90–4.78
À1
1
2
v/v). Enzyme (4 mgmL ) was added, and the mixture was placed
(
m, 1H), 2.44 (d, J=1.2 Hz, 3H), 1.81–1.62 (m, 2H), 1.37 (d, J=
on a shaker; the reactions lasted from one to four days. The mix-
ture was extracted with ethyl acetate, the organic phase was sepa-
rated, and solvents were evaporated. The mixture was analyzed by
1
3
6
1
7
.4 Hz, 3H), 1.00 (t, J=7.5 Hz, 3H); C NMR (CDCl , 100 MHz): d=
3
60.4, 154.1, 153.4, 152.4, 151.8, 125.4, 117.8, 117.4, 114.5, 109.9,
8.4, 28.7, 19.2, 18.7, 9.5; UV/Vis (lmax, nm, phosphate buffer
HPLC with a Chiralpak AD-H column. Retention times: (R)-1 t =
R
pH 7.4, loge): 215 (4.1), 275 (3.9) 310 (3.9); elemental analysis calcd
%) for C H O : C 65.21, H 5.84; found: C 65.30, H 5.85; HPLC
1
0.8 min, (S)-1 t =12.1 min. E values were calculated from E=
R
(
1
5
16
5
ln[(1Àc)(1Àee )]/ln[(1Àc)(1+ee )], where c is conversion, and ee is
s
s
s
analysis t =10.8 min (AD-H Chiralpak).
R
the enantiomeric excess of unreacted substrate. Quick E values
were calculated from E=(V /V )/(V /V ), where the hydrolysis rate
S
ref
R
ref
(S)-2-Butyl-(2-oxo-4-methyl-2H-chromen-7-yl) carbonate (S)-1:
for the reference was replaced by a second pseudoenantiomer rate
The procedure was applied starting with (S)-butan-2-ol (74 mg,
(
e.g., E=((S)-1/(R)-2)/((R)-1/(S)-2), if V >V ).
S R
1
2
.0 mmol) and 7-hydroxy-4-methyl-2H-chromen-2-one (352 mg,
mmol) to yield carbonate (S)-5 as a white solid; yield 108.9 mg
1
(
39%); m.p. 51–548C; [a] =+13.0 (c=1.0, CHCl ); H NMR (CDCl ,
D
3
3
Acknowledgements
4
00 MHz): d=7.61 (d, J=8.8 Hz, 1H), 7.23 (d, J=2.2 Hz, 1H), 7.17
(
dd, J =2.3 Hz, J =8.7 Hz, 1H), 6.27 (q, J=1.1 Hz, 1H), 4.90–4.78
1 2
(
m, 1H), 2.47–2.40 (m, 3H), 1.82–1.61 (m, 2H), 1.41–1.35 (m, 3H),
This work was supported by project “Biotransformations for phar-
maceutical and cosmetics industry” No. POIG.01.03.01–00–158/
09–01 part-financed by the European Union within the European
Regional Development Fund and by Polish National Science
Center project No. 2013/11/B/ST5/02199.
1
3
1
1
9
(
5
.04–0.96 (t, 3H); C NMR (CDCl , 100 MHz): d=160.4, 154.1, 153.4,
3
52.4, 151.8, 125.4, 117.8, 117.4, 114.5, 109.9, 78.4, 28.7, 19.2, 18.7,
.5; UV/Vis (lmax, nm, phosphate buffer pH 7.4, loge): 215 (4.1), 275
3.9) 310 (3.9); elemental analysis calcd (%) for C H O : C 65.21, H
15 16 5
.84; found: C 65.38, H 5.87; HPLC analysis t =12.1 min (AD-H
R
Chiralpak).
Keywords: enantioselectivity · fluorescent probes · fluorogenic
assay · high-throughput screening · pseudoenantiomers
(
R)-2-Butyl-(2-oxo-4-trifluoromethyl-2H-chromen-7-yl) carbonate
(
(
R)-2: The procedure was applied starting with (R)-butan-2-ol
111 mg, 1.5 mmol) and 7-hydroxy-4-trifluoromethyl-2H-chromen-2-
[
[
1] B. Testa, W. F. Trager, Chirality 1990, 2, 129–133.
2] B. Martín-Matute, M. Edin, K. Bogµr, J.-E. Bäckvall, Angew. Chem. Int. Ed.
one (690 mg, 2 mmol) to yield carbonate (R)-6 as a white solid;
yield 214.1 mg (43%); m.p. 101–1028C; [a] =À9.3 (c=1.0, CHCl );
D
3
2004, 43, 6535–6539; Angew. Chem.2004, 116, 6697–6701.
1
H NMR (CDCl , 400 MHz): d=7.75 (dq, J =1.8 Hz, J =8.8 Hz, 1H),
3
1
2
[3] A. J. J. Straathof, J. A. Jongejan, Enzyme Microb. Technol. 1997, 21, 559–
7
.34 (d, J =2.2 Hz, 1H), 7.28–7.20 (m, 1H), 6.78 (s, 1H), 4.91–4.79
1
571.
(
m, 1H), 1.83–1.63 (m, 2H), 1.38 (d, J=6.4 Hz, 3H), 1.01 (t, J=
[
4] L. E. Janes, R. J. Kazlauskas, J. Org. Chem. 1997, 62, 4560–4561.
1
3
7
1
9
.5 Hz, 3H); C NMR (CDCl , 100 MHz): d=158.4, 155.1, 154.4,
[5] M. T. Reetz, K. M. Kühling, S. Wilensek, H. Husmann, U. W. Häusig, M.
Hermes, Catal. Today 2001, 67, 389–396.
[6] M. T. Reetz, M. H. Becker, K. M. Kühling, A. Holzwarth, Angew. Chem. Int.
Ed. 1998, 37, 2647–2650; Angew. Chem. 1998, 110, 2792–2795.
3
52.1, 126.3, 122.8, 120.0, 118.3, 115.4, 111.2, 110.5, 78.8, 28.6, 19.2,
.5; UV/Vis (lmax, nm, phosphate buffer pH 7.4, loge): 215 (3.8), 285
3.7), 315 (3.7); elemental analysis calcd (%) for C H F O : C 54.55,
15 13 3 5
(
[
[
[
7] M. T. Reetz, M. H. Becker, H.-W. Klein, D. Stçckigt, Angew. Chem. Int. Ed.
999, 38, 1758–1761; Angew. Chem. 1999, 111, 1872–1875.
8] M. T. Reetz, K. M. Kühling, A. Deege, H. Hinrichs, D. Belder, Angew. Chem.
Int. Ed. 2000, 39, 3891–3893; Angew. Chem. 2000, 112, 4049–4052.
9] M. T. Reetz, K. M. Kühling, H. Hinrichs, A. Deege, Chirality 2000, 12, 479–
H 3.97; found: C 54.78, H 3.96.
1
(
S)-2-Butyl-(2-oxo-4-trifluoromethyl-2H-chromen-7-yl) carbonate
(
(
S)-2: The procedure was applied starting with (S)-butan-2-ol
111 mg, 1.5 mmol) and 7-hydroxy-4-trifluoromethyl-2H-chromen-2-
482.
one (690 mg, 2 mmol) to yield carbonate (S)-6 as a white solid;
[
10] F. Morís-Varas, A. Shah, J. Aikens, N. P. Nadkarni, J. D. Rozzell, D. C. Demi-
yield 177.8 mg (36%); m.p. 101–1028C; [a] =+9.8 (c=1.0, CHCl );
D
3
rjian, Bioorg. Med. Chem. 1999, 7, 2183–2188.
1
H NMR (CDCl , 400 MHz): d=7.75 (dq, J =1.8 Hz, J =8.9 Hz, 1H),
[11] H. Liu, X. Yang, L. Liu, J. Dang, Y. Xie, Y. Zhang, J. Pu, G. Long, Y. Li, Y.
3
1
2
7
.34 (d, J=2.2 Hz, 1H), 7.28–7.21 (m, 1H), 6.78 (d, J=0.7 Hz, 1H),
Yuan, J. Liao, F. Liao, Anal. Chem. 2013, 85, 2143–2154.
ChemBioChem 2016, 17, 71 – 76
www.chembiochem.org
75
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
[
[
[
12] P. Babiak, J.-L. Reymond, Anal. Chem. 2005, 77, 373–377.
[24] T. Knüttel, H. Meyer, T. Scheper, Enzyme Microb. Technol. 2005, 37, 673–
13] M. Schmidt, U. T. Bornscheuer, Biomol. Eng. 2005, 22, 51–56.
14] G. Klein, J.-L. Reymond, Helv. Chim. Acta 1999, 82, 400–407.
15] F. Badalassi, D. Wahler, G. Klein, P. Crotti, J.-L. Reymond, Angew. Chem.
Int. Ed. 2000, 39, 4067–4070; Angew. Chem. 2000, 112, 4233–4236.
16] E. Nyfeler, J. Grognux, D. Wahler, J.-L. Reymond, Helv. Chim. Acta 2003,
686.
[25] T. Knüttel, T. Hartmann, H. Meyer, T. Scheper, Enzyme Microb. Technol.
2001, 29, 150–159.
[26] W.-C. Sun, K. R. Gee, R. P. Haugland, Bioorg. Med. Chem. Lett. 1998, 8,
[
[
3107–3110.
[
[
[
27] S.-i. Ueji, K. Watanabe, T. Koshiba, M. Nakamura, K. Oh-ishi, Y. Yasufuku,
T. Miyazawa, Biotechnol. Lett. 1999, 21, 865–868.
28] C. S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc. 1982,
8
6, 2919–2926.
17] R. Pꢁrez Carlón, N. Jourdain, J.-L. Reymond, Chem. Eur. J. 2000, 6, 4154–
162.
4
1
04, 7294–7299.
[
[
18] E. Henke, U. T. Bornscheuer, Anal. Chem. 2003, 75, 255–260.
19] E. M. Gonzµlez-García, J. Grognux, D. Wahler, J.-L. Reymond, Helv. Chim.
Acta 2003, 86, 2458–2467.
20] P. D. Jones, N. M. Wolf, C. Morisseau, P. Wheatstone, B. Hock, B. D. Ham-
mock, Anal. Biochem. 2005, 343, 66–75.
29] Z. Prokop, Y. Sato, J. Brezovsky, T. Mozga, R. Chaloupkova, T. Koudelako-
va, P. Jerabek, V. Stepankova, R. Natsume, J. G. E. van Leeuwen, D. B.
Janssen, J. Florian, Y. Nagata, T. Senda, J. Damborsky, Angew. Chem. Int.
Ed. 2010, 49, 6111–6115; Angew. Chem. 2010, 122, 6247–6251.
30] M. D. Mertens, M. Gütschow, Chirality 2013, 25, 957–964.
[
[
[
[
[
21] R. Sicard, L. S. Chen, A. J. Marsaioli, J.-L. Reymond, Adv. Synth. Catal.
2
005, 347, 1041–1050.
22] A. Zadlo, D. Koszelewski, F. Borys, R. Ostaszewski, ChemBioChem 2015,
6, 677–682.
23] S. Marose, C. Lindemann, T. Scheper, Biotechnol. Prog. 1998, 14, 63–74.
Manuscript received: October 5, 2015
1
Accepted article published: October 28, 2015
Final article published: November 23, 2015
ChemBioChem 2016, 17, 71 – 76
www.chembiochem.org
76
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim