High-throughput screening for the asymmetric transformation reaction of L-histidine to D-histidine by capillary array electrophoresis

Article abstract of DOI:10.1021/ac051377w

Asymmetric transformation reaction of L-histidine to D-histidine was studied by homemade capillary array electrophoresis for the first time. The enantiomeric excess value of asymmetric histidine products can be directly determined from the electrophoretogram of capillary array electrophoresis. The experiment results showed that the optimized asymmetric transformation reaction condition was in the presence of salicylaldehyde as catalyst and acetic acid as solvent.

Full text of DOI:10.1021/ac051377w

Anal. Chem. 2006, 78, 901-904
High-Throughput Screening for the Asymmetric
Transformation Reaction of L-Histidine to
D-Histidine by Capillary Array Electrophoresis
Jun Wang, Kaiying Liu, Guangming Sun, Jiling Bai, and Li Wang*
State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences,
Dalian, 116023, P. R. China
Scheme 1
Asymmetric transformation reaction of L-histidine to D-
histidine was studied by homemade capillary array elec-
trophoresis for the first time. The enantiomeric excess
value of asymmetric histidine products can be directly
determined from the electrophoretogram of capillary array
electrophoresis. The experiment results showed that the
optimized asymmetric transformation reaction condition
was in the presence of salicylaldehyde as catalyst and
acetic acid as solvent.
mass spectrometric analysis,8 the use of capillary array electro-
phoresis (CAE)9 and HPLC/circular dichroism (CD) spectros-
copy.10
There are also some disadvantages for the application of HTS;
for example, the bottleneck in such an application was the data
analysis and processing.11 In the case of slow analytical methods
such as GC and HPLC, it became crucial to perform measure-
ments at the correct sampling time.12 The possible way to solve
such a problem would be the application of time-resolved HTS.12
To broaden this scope and avoid the problem of such incorrect
sampling time, we developed a multichannel CAE system, which
could serve as a real-time measurement, to optimize the reaction
conditions of crystallization-induced asymmetric transformations
Combinatorial asymmetric catalysis involves time-saving paral-
lel synthesis and the testing of large numbers of chiral catalysts.1
The challenges in this interesting new area of asymmetric catalysis
are twofold, which center around strategies for the modular
synthesis of chiral ligands and on developing high-throughput
assays for determining the enantiomeric excess (ee).2 Several
catalysts have already been identified by parallel synthesis and
high-throughput screening (HTS) techniques.3 However, the factor
limiting an efficient extension of this research to asymmetric
catalysis remains the lack of efficient methods for rapid screening
of the enantioselective reaction.4 To overcome this major obstacle,
the first approach consists of screening the catalyst library for
activity using an HTS procedure and then testing each lead for
enantioselectivity by conventional methods,5 but this is only
amenable if few catalysts from the broad library are active. The
first high-throughput ee assay was a rather crude UV/visible-based
screening system for the lipase-catalyzed kinetic resolution of
chiral p-nitrophenol esters.6 Nevertheless, it forms the basis of
other tests as well, for example, the recent development using
coupled enzymatic transformations.7 Several more general screen-
ing systems have been devised, such as methods based on the
of
L
-histidine to
D-histidine.
CAE, which was initially designed for DNA analysis and
sequencing, had been applied in screening of enantioselective
catalysts of chiral amines first by Reetz et al.9 Their initial elegant
investigation indicated that CAE could be used for high-
throughput determination of enantiomeric purity. Evidently, a
high-throughput technique allowing quantification of both activity
and enantioselectivity would accelerate the rate at which asym-
metric catalysts are discovered. Continued efforts in this fascinat-
ing area are necessary, since no single assay is universal.
D-Histidine (His) can be prepared from inexpensive L-His by
asymmetric transformation,13 described by Scheme 1. Despite
great advances in biocatalysis and asymmetric synthesis, the
resolution of racemates is an important approach in the industrial
synthesis of enantiomeric pure compounds. It is often the most
economical and convenient way to prepare enantiomeric pure
* To whom correspondence should be addressed. Telephone: +0086-411-
84379243. Fax: +0086-411-84675584. E-mail: liwangye@dicp.ac.cn.
(1) Burgess, K.; Lim, H. J.; Porte, A. M.; Sulikowski, G. A. Angew. Chem., Int.
Ed. Engl. 1996, 35, 220-222.
(8) Reetz, M. T.; Becker, M. H.; Klein, H. W.; Sto¨ckigt, D. Angew. Chem., Int.
Ed. Engl. 1999, 38, 1758-1761.
(2) Edkins, T. J.; Bobbitt, D. R. Anal. Chem., 2001, 73, 488A - 496A.
(3) Lavastre, O.; Morken, J. P. Angew. Chem., Int. Ed. Engl. 1999, 38, 3163-
3165.
(4) Janes, L. E.; Lowendahl, A. C.; Kazlauskas, R. J. Chem. Eur. J. 1998, 4,
2324-2331.
(9) Reetz, M. T.; Kuhling, K. M.; Deege, A.; Hinrichs, H.; Belder, D. Angew.
Chem., Int. Ed. 2000, 39, 3891-3893.
(10) Ding, K.; Ishii, A.; Mikami, K. Angew. Chem., Int. Ed. Engl. 1999, 38, 497-
501.
(11) Tibiletti, D.; de Graaf, E. A. B.; Teh, S. P.; Rothenberg, G.; Farrusseng, D.;
(5) Copeland, G. T.; Miller, S. J. J. Am. Chem. Soc. 1999, 121, 4306-4307.
(6) Reetz, M. T.; Zonta, A.; Schimossek, K.; Liebeton, K.; Jaeger, K. E. Angew.
Chem., Int. Ed. Engl. 1997, 36, 2830-2832.
Mirodatos, C. J. Catal. 2004, 225, 489-497.
(12) Boelens, H. F. M.; Iron, D.; Westerhuis, J. A.; Rothenberg, G. Chem. Eur.
J. 2003, 9, 3876-3881.
(7) Baumann, M.; Sturmer, R.; Bornscheuer, U. T. Angew. Chem., Int. Ed. 2001,
40, 4201-4204.
(13) Shiraiwa, T.; Shinjo, K.; Masui, Y.; Ohta, A.; Natuyama, H.; Miyazaki, H.;
Kurokawa, H. Bull. Chem. Soc. Jpn. 1991, 64, 3741-3742.
10.1021/ac051377w CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/15/2005
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006 901

compounds. Traditional resolutions through selective crystalliza-
tion of diastereomeric salts imply the introduction of in situ
racemization of the unwanted isomer resulting in crystallization-
induced asymmetric transformations.14 For amino acids, this is a
well-known process whereby in situ racemization is achieved
under mild conditions through catalytic amounts of Schiff bases.
When
the presence of 0.1 equiv of aldehyde in organic acid, the salt of
-TA and -His crystallized, from which -His can be obtained in
the form of optical purity. The crystallization-induced asymmetric
transformation reaction can be divided into three steps. First, DL
His can be prepared by racemization method with aldehyde
catalysts and organic acid solvents; then, -His and optically active
TA continuously formed precipitable salt along with the continuous
racemization of -His in solution; finally, the salt was alkalified
and -His was separated from the mixture.
In this paper, the asymmetric transformation reactions,
-His,13 were successfully investigated for the first time by our
L-His was heated with 1 equiv of L-tartaric acid (TA) in
L
D
D
-
D
L
Figure 1. Electrophoretogram of racemic products of L-His by eight
aldehydes as catalyst in the presence of acetic acid.
D
L
-His
to
D
the given speed, each capillary was addressed at the interval of
60 ms. A homemade data acquiring card was employed to transfer
data to a computer. The data were analyzed and converted into
electrophoretogram by in-house C Program. By using this setup,
16 racemate products of the racemization reaction could be studied
in the same run. Before experiment, the capillary array was rinsed
with the corresponding running Borax buffer (20 mmol/L, pH
10) containing 2.5 mmol/L â-CD. The running buffer was the
optimal separation condition of DL-His, which was optimized with
different buffer concentrations or different additives by this CAE.
Samples were loading by 20 s of gravity flow. Separations were
performed at 6000 V (230 V/cm).
homemade 16-channel CAE setup, which solved the problem of
capillary orientation of previous CAE systems and recorded the
D-His signal in real-time scale.
EXPERIMENTAL SECTION
Reagents and Chemicals. Fluorescein isothiocyanate,
L-
histidine (L-His), and â-cyclodextrin (â-CD) were purchased from
Sigma Chemical Co. (St. Louis, MO). All other chemical com-
pounds are analytical pure grade.
Capillary Array Electrophoresis. Details about our rotary
CAE setup can be found elsewhere.15 Herein, we just gave a brief
description. Racemate separations of optically active amino acids
were completed on a CAE using a confocal fluorescent rotary
scanner. A total of 16 capillaries (50-µm inner diameter) of 26-cm
total length and 23-cm effective length were symmetrically
distributed on a round capillary holder. The inlet ends of 16
capillaries were spatially dipped into 16-well microtiter plates, in
which 16 platina electrodes were placed to make independent
injections possible into each capillary. The excitation laser, 488-
nm output of a 20-mW argon ion laser, passed through a dichroic
mirror (T <500 nm, R >500 nm), was focused, and reflected to
the capillaries by a rotary reflection mirror, which was seated in
the center of the round capillary holder. Fluorescent signal from
the irradiated capillary was confocally reflected by the same rotary
reflection mirror and the dichroic mirror to the photomultiplier
tube (PMT). The rotary reflection mirror and a rotary encoder
(Omron, model E6CP, absolute position readout) were coaxally
combined to a high-speed direct current motor. By reading the
output data of the encoder, the absolute position information of
each capillary could be oriented directly by the control computer.
The main advantage of this design was that both the accuracy
and the repeatability of capillary position were independent of the
speed and stability of the dc motor rotation. While rotating the
motor, the surrounding capillaries were scanned by the focused
excitation laser beam one by one, and the corresponding fluores-
cent signal output, detected by PMT, was recorded together with
the capillary position readout. Currently, the motor was set at
∼1000 rpm due to the speed limitation of the rotary encoder. For
Racemization Reaction of L-His. L-His was heated in the
presence of 0.1 equiv of aldehyde as catalyst in 5 mL of acetic
acid for 7 h. A 100-µL aliquot of the solution was pipetted out,
rapidly cooled, and diluted to 10-3mol/L, fluorescently labeled,
and finally diluted to 10-5mol/L for sample injection.
Asymmetric Transformation Reaction of
L-His. L-His was
heated and stirred for 10 h with 1 equiv of -TA in the presence
L
of salicylaldehyde or n-butyraldehyde in 3 mL of organic acid.
The enantiomeric His was fluorescently labeled, and the ee value
could be measured after the synthetic D-His salt was filtered, dried,
and dissolved with Borax buffer (10 mM, pH 10). The additional
purification step as reported in previous methods,13 the third step
in Scheme 1, was not necessary in our experiments.
RESULTS AND DISCUSSION
CAE for the Racemization of L-His. The racemic products
of L-His with eight kinds of aldehydes were respectively hydro-
dynamically injected into 16 capillaries. To estimate the reliability
of the separation and detection, two adjacent capillaries were used
for same conditions; for example, both the first channel and the
second channel were for samples in the presence of acetaldehyde
as catalyst. Figure 1 showed the racemization results of L-His in
the presence of acetic acid by eight aldehydes: acetaldehyde,
propionaldehyde, n-butyraldehyde, salicylaldehyde, benzaldehyde,
capric aldehyde, glutaraldehyde, and lauryl aldehyde. The inset
figure in Figure 1 illustrated the details of the eighth capillary,
containing salicylaldehyde as catalyst in the presence of acetic
acid.
D
-His and
L
-His were baseline separated. The sampling
-His or -His,
(14) Grigg, R. Chem. Soc. Rev. 1987, 16, 89-121.
(15) Wang, J.; Sun, G. M.; Bai, J. L.; Wang, L. Analyst 2003, 128, 1434-1438.
interval of 60 ms, compared with the full width of
D
L
902 Analytical Chemistry, Vol. 78, No. 3, February 1, 2006

both the first to second channels were for the reaction of L-His in
the presence of formic acid (FA) as solvent. Three kinds of organic
acids were investigated, i.e., acetic aid (AA), propionic acid (PA)
and butanoic acid (BA). To clearly demonstrate the racemization
results, eight channels were selected in Figure 2.
Figure 2 indicated that AA was the optimal solvent for the
racemization reaction of L-His in the presence of n-butyraldehyde
or salicylaldehyde. PA and BA were better than FA for the
racemization reaction. For BA, the racemization result in salicyl-
aldehyde was better than that of n-butyraldehyde.
From Figure 1 and Figure 2, we obtained the optimal condi-
tions for the racemization reaction of L-His in the presence of
salicylaldehyde as catalyst and AA as solvent. Since the racemi-
zation principle was also adapted for the asymmetric transforma-
tion of
as solvent for reaction conditions of the asymmetric transformation
from -His to -His.
L-His, we used salicylaldehyde as catalyst and acetic acid
Figure 2. Organic acid solvent effects on racemization of L-His in
the presence of n-butyraldehyde and salicylaldehyde.
L
D
CAE for Asymmetric Transformation of
L
-Histidine. We
investigated the asymmetric transformation of
L-His in the pres-
∼0.1 min, could provide almost real-time measurement without
peak profile distortion. In fact, during illumination of each capillary
by the excitation laser beam, the fluorescent signal from the
capillary was sampled at the rate of 100 kHz. Each data point in
Figure 1 was the average result of ∼10 real sampling points.
Although this sampling rate resulted in large numbers of data,
these data were automatically divided into different arrays ac-
cording to the encoder readout and corresponded to each capillary
by software.
ence of salicylaldehyde and
solvents. We also investigated the effects of n-butylaldehyde.
The asymmetric transformation products of -His in the
L-TA with various organic acids as
L
presence of salicylaldehyde or n-butylaldehyde with four kinds
of organic acid as solvents were respectively hydrodynamically
injected into different capillaries. To verify the reproducibility of
the setup, four adjacent channels were used for the same
asymmetric transformation products, i.e., 1st-4th channels for
FA, 5th-8th for AA, 9th-12th for PA, and 13th-16th for BA. To
clearly demonstrate the asymmetric transformation results, we
just selected four channels to show in Figure 3. Figure 3a
illustrated the electrophoretogram of asymmetric transformation
As illustrated in Figure 1, it was obvious that salicylaldehyde
was the best choice for the racemization of
aldehydes. Glutaraldehyde and lauryl aldehyde were the second
choices as catalysts for the racemization reaction of -His.
The racemates of -His in the presence of salicylaldehyde or
L-His among the eight
L
products of L-His in the presence of butyraldehyde with the four
L
organic acids. Figure 3b showed the results in the presence of
n-butyraldehyde.
The ee values were calculated from the integrated height of
peaks in the electrophoretogram, as illustrated in Figure 3,
n-butyraldehyde with four organic acids as solvents were respec-
tively hydrodynamically injected into different capillaries. To verify
the reproducibility of the setup, two adjacent channels were used
for the same asymmetric transformation products; for example,
Figure 3. Organic acid solvent effects on asymmetric products of L-His in the presence of (a) salicylaldehyde and (b) n-butyraldehyde.
Analytical Chemistry, Vol. 78, No. 3, February 1, 2006 903

the highest enantiomeric excess, 82 ( 4%, for the asymmetric
transformation reaction of -His. This optimal condition for
asymmetric transformation of -His was also supported by Figure
L
L
1 and Figure 2, in which salicylaldehyde showed the best
racemization results in different organic acids. The maximum
standard deviation of these 16 capillaries, (4%, was acceptable,
considering the differences among each channel.
CONCLUSIONS
We applied a homemade high-throughput CAE system to
optimize the reaction conditions of crystallization-induced asym-
Figure 4. Organic acid effects on ee of asymmetric transformation
of L-His with (a) salicylaldehyde and (b) n-butyraldehyde as catalyst.
metric transformation of L-histidine to D-histidine and rapidly
Table 1. ee Valuesa (%) in Organic Acids with
Salicylaldehyde and n-Butyraldehyde as Catalyst
obtained the optimal reaction conditions, i.e., salicylaldehyde as
the catalyst and acetic acid as solvent for the asymmetric
transformation reaction. By using this setup, the operational step
for determining enantiomeric excess was also reduced, compared
with previous methods.13,14 As shown in the present study, the
sampling rate and sensitivity were all acceptable in achieving high
throughput and high specificity with adequate sensitivity and
without sampling distortion.11,12 Our study indicated that the
development of CAE for optimization of the asymmetric transfor-
mation reaction could be used in a truly high-throughput manner
(16 ee determinations in less than 10 min). This high-throughput
acids
salicylaldehyde
n-butyraldehyde
FA (formic acid)
AA (acetic acid)
PA (propionic acid)
BA (butanoic acid)
-60 ( 3
82 ( 4
57 ( 3
65 ( 1
-54 ( 2
54 ( 3
34 ( 3
-21 ( 3
a Mean ( std dev (N ) 4).
according to the following definition:
analysis of asymmetric transformation of L-histidine by CAE also
demonstrated the capabilities of this method for quality assurance.
Further development of our CAE system, up to 128 capillaries
with higher electric field strength leading to the possibility of chiral
separation within seconds, is in progress.
ee% ) (ID - IL)/(ID + IL) × 100%
where ID and IL were the integrated peak heights of D-His and
L
-His, respectively.
Figure 4a showed the ee values of asymmetric transformation
-His in different organic acid solvents in the presence of
ACKNOWLEDGMENT
Funding support for the present studies was provided by the
of
L
Chinese Academy of Sciences (KGCX2-201-3) and the National
High Technology Research and Development Program of China
(2002AA321020).
salicylaldehyde. Figure 4b showed the results in the presence of
n-butyraldehyde.
To estimate the reproducibility and systemic error of our CAE
in the application of asymmetric transformation of L-His, we
Received for review August 2, 2005. Accepted November
19, 2005.
calculated the average ee and its standard deviation of each of
the four capillaries with same conditions, as listed in Table 1. As
one could find in the Table 1, AA with salicylaldehyde provided
AC051377W
904 Analytical Chemistry, Vol. 78, No. 3, February 1, 2006