A protein-based mixed selector chiral monolithic stationary phase in capillary electrochromatography

Article abstract of DOI:10.1039/c8nj02309c

A new mixed selector chiral stationary phase (CSP) was prepared with co-immobilized human serum albumin and cellulase on a poly(glycidylmethacrylate-co-ethylene glycol dimethacrylate) (poly(GMA-co-EDMA)) monolith and the evaluation of its usefulness in chiral separation research was presented. For comparison, two single selector chiral stationary phases (CSPs) were also fabricated with the corresponding proteins. The enantioseparation ability of these CSPs was investigated by capillary electrochromatography (CEC) with various racemates. The mixed selector CSP exhibited a broader range of enantioselectivities than the single selectors and it could separate 10 chiral analytes while the two single selector CSPs resolved 3 and 8 respectively. Moreover, for (±)-warfarin, the enantioresolution was improved on the mixed selector CSP. Meanwhile, compared with the single selector CSPs, no additional preparation stage or reagent consumption was required in the simultaneous immobilization of different proteins, which is more favorable from economical and practical points of view. Consequently, by mixing HSA and cellulase together, the composite column combines the enantioselectivities of both individual proteins, thus expanding their application range practically.

Full text of DOI:10.1039/c8nj02309c

NJC
View Article Online
View Journal
PAPER
A protein-based mixed selector chiral monolithic
stationary phase in capillary electrochromatography†
Cite this:DOI: 10.1039/c8nj02309c
ab
Shujuan Xu,ab Yuying Wang,ab Yixia Tangab and Yibing Ji
*
A new mixed selector chiral stationary phase (CSP) was prepared with co-immobilized human serum
albumin and cellulase on a poly(glycidylmethacrylate-co-ethylene glycol dimethacrylate) (poly(GMA-co-
EDMA)) monolith and the evaluation of its usefulness in chiral separation research was presented. For
comparison, two single selector chiral stationary phases (CSPs) were also fabricated with the
corresponding proteins. The enantioseparation ability of these CSPs was investigated by capillary
electrochromatography (CEC) with various racemates. The mixed selector CSP exhibited a broader
range of enantioselectivities than the single selectors and it could separate 10 chiral analytes while the
two single selector CSPs resolved 3 and 8 respectively. Moreover, for (ꢀ)-warfarin, the enantioresolution
was improved on the mixed selector CSP. Meanwhile, compared with the single selector CSPs, no
additional preparation stage or reagent consumption was required in the simultaneous immobilization of
different proteins, which is more favorable from economical and practical points of view. Consequently,
by mixing HSA and cellulase together, the composite column combines the enantioselectivities of both
individual proteins, thus expanding their application range practically.
Received 10th May 2018,
Accepted 9th July 2018
DOI: 10.1039/c8nj02309c
rsc.li/njc
multiple binding interactions between the protein and analyte.
To date, various proteins, including bovine serum albumin
1. Introduction
The enantioseparation of chiral drugs continues to be a hot topic (BSA),11–14 human serum albumin (HSA),15–18 penicillin G
in pharmaceutical analysis.1,2 Among the various chromato- acylase,19,20 a1-acid glycoprotein (AGP),21 ovomucoid (OVM),22
graphic separation techniques, capillary electrochromato- pepsin,23–25 lipase26 and avidin27 have been successfully immo-
graphy (CEC) has been widely utilized for chiral separation bilized on monoliths as a chiral stationary phase in high-
due to its important advantages over classical techniques, such performance liquid chromatography (HPLC) or CEC. Though
as the minimization of solvent consumption, the reduced a wide range of racemic compounds have been resolved on
environmental impact, and the improved suitability for coupling these CSPs, the chiral recognition abilities of different proteins
with mass spectrometry detection.3
are markedly complementary. Therefore, the combination of
In CEC, one of the most significant topics of research in the two (or more) different proteins in one stationary phase may
field of chiral separation is the development of CSPs, along lead to a broader range of enantioselectivities than those
with high selectivity for a variety of compounds. Monoliths, provided by either of the single protein based stationary
bearing the advantages of easy preparation, low resistance to mass phases. However, a mixed chiral monolithic stationary phase
transfer and high permeability, have become ideal stationary containing different proteins as a selector has not been pre-
phases for various separation formats. There are several chiral viously described.
selectors that have been utilized in CEC with monoliths such as
HSA, a non-glycosylated protein, having an isoelectric point
ligand exchangers,4 Pirkle-type selectors,5 cyclodextrin (CD) and its (pI) of 4.7 and a molecular mass of 66.5 kDa, is known to bind a
derivatives,6 polysaccharide derivatives,7 antibiotics,8 chiral crown variety of drugs and biological compounds.28 When immobi-
ethers9 and proteins.10 Among these, protein-based monoliths lized on a monolithic column, this can be used to study drug–
present high efficiency for enantiomeric separation due to protein interactions and separate some chiral solutes.29
multiple binding sites on the surface of the protein and HSA-based monoliths, as obtained by anchoring the protein
to a polymer or silica monolithic matrix, were first introduced
by Hage and coworkers,15 and these columns have been success-
fully applied to the enantiomeric separation of neutral and acidic
drugs including tryptophan, phenylalanine, ibuprofen and
warfarin by HPLC, while they exhibit much less efficiency for
a Department of Analytical Chemistry, China Pharmaceutical University,
Nanjing 210009, China. E-mail: jiyibing@msn.com
b Key Laboratory of Drug Quality Control and Pharmacovigilance,
Ministry of Education, Nanjing 210009, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c8nj02309c basic drugs.15–17,30
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
Cellulase from T. reesei is a mixture of many hydrolytically in all of the experiments, including synthetic reaction and
active enzymes containing three main components: cello- mobile phase preparation. (ꢀ)-Azelastine was obtained from
biohydrolase I (CBH I), cellobiohydrolase II (CBH II) and Jiangsu Hengrui Medicine Co., Ltd (Lianyungang, China).
endoglucanase II (EG II), whose molecular weights are 64 kDa, (ꢀ)-Warfarin was obtained from Nanjing Lide Biotechnology
53 kDa and 48 kDa, respectively and pIs are 3.9, 5.9 and Co., Ltd (Nanjing, China). Other racemic drugs were from
5.5, respectively.31–33 Recently, Matsunaga et al. immobilized Dalian Meilun Biotec Co., Ltd (Dalian, China). Bare fused-
cellulase on aminopropyl-silica gels via its amino and carboxy silica capillaries (50, 75, 100 mm i.d. ꢁ 365 mm o.d.) were
groups respectively, by using N,N0-disuccinimidyl carbonate supplied by Yongnian Rui-feng Chromatographic Devices Co.,
(DSC), and 1-ethyl-3-(30-dimethylaminopropyl)carbodimide (EDC) Ltd (Handan, China).
and N-hydroxysulfosuccinimide (HSSI).34 The new CSPs showed
good selectivity and resolution in the enantioselective analysis
2.2. Instruments
of b-blockers, such as propranolol, alprenolol, oxprenolol and
pindolol. In addition, the effect of silica particle diameters (5, 3
and 2.1 mm) on the chromatographic performance was also
investigated and it was shown that the 2.1 mm aminopropyl-
silica gels give the most efficient performance. Although smaller
silica particles can improve the separation efficiency, the back-
pressure increases rapidly with decreasing particle size.
Compared with packed columns, monoliths, due to their
porous properties, can avoid this effectively. Nevertheless, to
the best of our knowledge, there has been little work reported
on the use of monolithic columns for cellulase immobilization.
Considering the similar molecular mass and pI, as well as
complementary enantioseparation scope of HSA and cellulase,
in this work, we prepared a protein based mixed selector
polymer monolithic CSP (HSA–cellulase@poly(GMA–EDMA)
monolith) in order to extend the range of applications of single
selector CSPs. GMA was used as a functional monomer that can
be formed by the aminolysis of the epoxy groups and then
activation with glutaraldehyde to anchor proteins. The enantio-
separation ability of the newly prepared mixed selector CSP and
corresponding single selector CSPs was evaluated by using
different classes of racemic pharmaceuticals, namely, a- and
b-blockers, serotonin-reuptake inhibitors, antihistamines,
anticoagulants, and amino acids under the same separation
conditions by CEC. Moreover, the factors that affect the separa-
tion efficiency including protein concentration, capillary inner
diameter and CEC conditions were also investigated.
An Agilent 7100 CE system (Waldbronn, Germany) equipped
with a diode-array UV detector, an auto-sampler, a ꢀ30 kV
power supply, a temperature controlled column compartment,
an external nitrogen pressure, and a chromatographic work-
station (Chemistry Station, USA) was used. SEM was carried out
on a S-3400N II (Hitachi, Japan).
2.3. Preparation of the monolithic support
The scheme for preparation of the poly(GMA–EDMA) monolith
is shown in Fig. 1A. The fused-silica capillary was washed with
1.0 mol Lꢂ1 NaOH (1 h), 0.1 mol Lꢂ1 HCl (30 min), and
methanol (10 min) successively to activate the silanol groups
on the capillary wall. The capillary was then dried by nitrogen
gas for 2 h at 120 1C. In order to provide covalent attachment of
the polymer, the pretreated capillary was then silanized by
g-MAPS (50%, v/v) in methanol and reacted at 50 1C overnight
followed by a rinsing step with methanol and a drying step with
nitrogen. To obtain a monolithic bed suitable for further
experiments, a mixture of 21.96% (v/v) GMA, 7.47% (v/v) EDMA,
62.38% (v/v) n-propanol, 8.19% (v/v) 1,4-butanediol and 0.32%
(m/v) AIBN was introduced into the preconditioned fused-silica
capillary to the appropriate length by siphon action and reacted
at 46 1C for 12 h. Finally, the resulting column was immensely
flushed with methanol and water in order to remove the
porogenic solvents and the unreacted monomers. To obtain
sufficient quantities of the monolithic material for surface area
determination and ninhydrin reaction, polymerization of the
2. Experimental
2.1. Chemicals and materials
2,20-Azobisisobutyronitrile (AIBN), n-propanol, 1,4-butanediol,
sodium cyanoborohydride (NaCNBH3), g-methacryloxypropyl-
trimethoxylsilane (g-MAPS), GMA, EDMA, glutaraldehyde,
cellulase and tryptophan were purchased from Aladdin Chem-
istry (Shanghai, China). HPLC-grade methanol and acetonitrile
(ACN) were from CINC High Purity Solvents (Shanghai) Co., Ltd
(Shanghai, China) and Tedia Co. Inc. (OH, USA), respectively.
Thiourea, ammonium hydroxide, phosphoric acid, hydrochloric
acid and 2-propanol were commercially available from Nanjing
Chemical Reagent Co., Ltd (Nanjing, China). Sodium hydroxide
and disodium hydrogen phosphate dodecahydrate were from
Xilong Chemical Co., Ltd (Shantou, China). HSA was from
Sigma-Aldrich (Shanghai, China). Deionized water was used
Fig. 1 General scheme for the preparation of the HSA–cellulase@poly(GMA–
EDMA) monolith.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
same reaction mixture was carried out in quartz tubes (180 mm ꢁ protonated amino groups reduced. When the glutaraldehyde
3 mm i.d.) instead of a capillary tube under the same experi- (GA) was reacted with amino to form imine, the EOF of the
mental conditions.
GA@poly(GMA–EDMA) monolith reduced. After HSA and cellu-
lase were covalently immobilized to the GA@poly(GMA–EDMA)
monolith, a positive EOF from the anode to the cathode was
generated in the pH range of 6.0–8.0, which was attributed to
the fact that HSA and cellulase were negatively charged. When
the pH value was 5.0, the cellulase became positively charged
and HSA had almost no charge, so the direction of EOF
changed from cathode to anode. The above results should be
interpreted as convincing proof of successful immobilization of
HSA and cellulase onto the poly(GMA–EDMA) monolith.
Ninhydrin reaction was also carried out by treating the
monolithic materials obtained from each step of the reaction
with 0.2% ninhydrin (m/v) in ethanol. As seen in Fig. S2 (ESI†),
the pristine poly(GMA–EDMA) monolith was still colorless
when it was mixed with ninhydrin solution in a 100 1C water
bath for a few minutes. For the NH2@poly(GMA–EDMA) mono-
lith, it was purple after being treated with ninhydrin, proving
the successful modification with ammonium hydroxide. For
the GA@poly(GMA–EDMA) monolith, the purple disappeared,
suggesting the reaction between GA and amino-groups. For the
HSA–cellulase@poly(GMA–EDMA) monolith, the purple appeared
again, indicating that the protein has been modified on the
monolith.
2.4. Immobilization of the proteins
The aminolysis of epoxy rings was carried out using 25% (m/v)
ammonium hydroxide at 40 1C for 5 h. The activation with
carbonyl groups was performed using a 50% glutaraldehyde
solution (m/v) for 5 h at room temperature. Afterwards, the
immobilization was carried out with a 4 mg mLꢂ1 protein
mixture solution dissolved in 100 mmol Lꢂ1 phosphate buffer
(pH 7.0) containing 2 mg mLꢂ1 NaCNBH3 for 12 h. The protein
mixture consisted of HSA and cellulase at a concentration ratio
of 1 : 1. Finally, the prepared monolithic column was washed
with 100 mmol Lꢂ1 phosphate buffer (pH 7.0) for 2 h to remove
the nonspecifically absorbed proteins and then conditioned
with running buffer for 1 h before installation to the CEC
instrument. For comparison, the corresponding single selector
chiral stationary phases (HSA@poly(GMA–EDMA) monolith
and cellulase@poly(GMA–EDMA) monolith) were also fabri-
cated with the two individual proteins at the concentration
of 2 mg mLꢂ1 using the same procedure, respectively. The
procedure of protein immobilization is shown in Fig. 1B.
2.5. Electrochromatography conditions
The morphology of the HSA–cellulase@poly(GMA–EDMA)
monolith was investigated by SEM. As illustrated in Fig. 2,
the prepared monolith was homogeneous and tightly anchored
to the capillary wall without any disconnection. The large
through pores presented in this stationary phase could reduce
the flow resistance. This would guarantee effective mass
transfer and high stability of the fabricated column. The porous
properties of the prepared monoliths were also investigated.
The specific surface area was measured by nitrogen adsorption
on a Micromeritics ASAP 2020 V4.00 (Micromeritics, USA).
As described in the previous studies,35,36 total porosity was
estimated by using the gravimetric method and the perme-
ability was determined based on Darcy’s Law with water as the
The mobile phase was 10 mmol Lꢂ1 phosphate buffer solution
with or without organic modifier. The phosphate buffer was
freshly prepared by dissolving an exact amount of disodium
hydrogen phosphate in deionized water and adjusted to a
desired pH by phosphoric acid. The racemic drugs were dissolved
in methanol or deionized water and diluted to an appropriate
concentration. All above solutions were filtered through 0.22 mm
nylon membrane filters and degassed before use. Unless stated
otherwise, the monolithic columns (effective length 21 cm and
total length 34 cm) were equilibrated prior to CEC runs with
mobile phases until a stable current was observed. The column
was thermostated at 20 1C and the applied voltage was in the
range of 5–25 kV. Electrokinetic injections were adopted at 10 kV
for 1 s or 2 s unless otherwise stated.
3. Results and discussion
3.1. Characterization of the prepared monolithic column
In CEC mode, the direction and magnitude of electroosmotic
flow (EOF) reveal the property of the capillary columns. In order
to confirm that the protein-based mixed selector chiral mono-
lithic stationary phase has been successfully synthesized,
thiourea was used as the EOF marker to measure the EOF after
each step of the reaction. As seen in Fig. S1 (ESI†), in the pH
range of 5.0–8.0, the EOF of the pristine poly(GMA–EDMA)
monolith was too low to elute thiourea in 50 min due to the
lack of groups that can generate EOF. After reaction with
ammonium hydroxide, the NH2@poly(GMA–EDMA) monolith
generated a high reversed EOF from cathode to anode. The EOF
decreased with the pH increasing from 5.0 to 8.0, because the
Fig. 2 Scanning electron micrograph for the HSA–cellulase@poly(GMA–
EDMA) monolith.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
eluent at room temperature. The effective radius of the equivalent and then decreased with the increase in the concentration of
pores was calculated using the Kozeny–Carman equation. The protein from 2 to 8 mg mLꢂ1. When the protein concentration
numerical values for the specific surface area, porosity, pore was 4 mg mLꢂ1, the Rs and a reached the maximum. However, for
diameter and permeability were 3.33 m2 gꢂ1, 0.57, 3.6 mm and (ꢀ)-tryptophan the Rs and a were increased by increasing the
3.7 ꢁ 10ꢂ13 m2, respectively.
content of protein. The Rs (1.43, 2.12, 2.28) and a (1.06, 1.12, 1.15)
were observed for 2, 4, and 8 mg mLꢂ1 protein respectively. From
section 3.4, we know the enantiorecognition of (ꢀ)-metoprolol
comes from cellulase and only the HSA can resolve (ꢀ)-tryptophan,
so the above phenomenon may be due to the fact that HSA could
occupy more aldehyde binding sites at higher protein concen-
tration thus leading to the reduction of immobilized cellulase.
Meanwhile, a further increase of protein concentration resulted in
lower column efficiency. From the viewpoints of higher resolution
and column efficiency, the protein concentration of 4 mg mLꢂ1 was
chosen in this study. In addition, the concentration ratios of HSA to
cellulase at 1 : 2, 1 : 1 and 2 : 1 were also investigated. As seen in
Fig. S3 (ESI†), the resolution of (ꢀ)-tryptophan increased with the
HSA ratio increasing while the resolution of (ꢀ)-metoprolol
increased with the increase in cellulase ratio. Taking into account
both, the optimum concentration ratio of HSA to cellulase was
determined to be 1 : 1.
3.2. Enantioseparation on the HSA–cellulase@poly(GMA–
EDMA) monolith
3.2.1. Effect of protein concentration on enantioseparation.
The concentration of protein could affect the amounts of protein
immobilized on the monolith. As a result, the separation perfor-
mance of the HSA–cellulase@poly(GMA–EDMA) monolith may
also be affected by the protein concentration. Thus, the effect of
protein concentration on the enantioseparation effectiveness was
investigated. As seen in Fig. 3, the resolution (Rs) and selectivity
factor (a) of (ꢀ)-metoprolol, the model analyte, increased firstly
3.2.2. Effect of capillary inner diameter on enantioseparation.
The influence of the capillary inner diameter (50, 75 and 100 mm)
on column performance was investigated in detail. As shown
in Fig. 4, the current and electroosmotic flow (EOF) of the HSA–
cellulase@poly(GMA–EDMA) monolith increased with the capillary
inner diameter increasing while the column back pressure
decreased (the column back pressure was observed when the flow
rate was 3 mL minꢂ1 using water as the mobile phase). In
comparison with the 50 mm i.d. column, the 75 mm i.d. column
could provide more immobilization sites to load proteins (3.35,
5.55 and 6.81 mg of proteins were immobilized on 1 cm length
HSA–cellulase@poly(GMA–EDMA) monoliths prepared with 50 mm,
75 mm and 100 mm inner diameter capillaries, respectively, see
Fig. S4, ESI†), which strengthened the interaction between the
proteins and (ꢀ)-metoprolol, consequently leading to higher
Fig. 3 Effect of protein concentration on enantioseparation of
(ꢀ)-metoprolol on the HSA–cellulase@poly(GMA–EDMA) monolith. Experi-
mental conditions: Rs, resolution; a, selectivity factor; N1, plate count for the
first-eluting enantiomer; samples, 50 mg mLꢂ1 (ꢀ)-metoprolol; running buffer,
10 mmol Lꢂ1 phosphate buffer (pH 7.0, containing 10% 2-propanol); applied
voltage, 10 kV; injection, 10 kV ꢁ 2 s; detection wavelength, 225 nm; resolution. Whereas, when the inner diameter was increased to
temperature, 20 1C.
100 mm, baseline separation of (ꢀ)-metoprolol could not be
Fig. 4 Effect of capillary inner diameter on column performance of the HSA–cellulase@poly(GMA–EDMA) monolith. Thiourea was used as the EOF
marker. Experimental conditions: samples, 50 mg mLꢂ1 (ꢀ)-metoprolol; running buffer, 10 mmol Lꢂ1 phosphate buffer (pH 7.0, containing 10% 2-
propanol); applied voltage, 10 kV for (ꢀ)-metoprolol and 20 kV for thiourea; injection, 10 kV ꢁ 2 s for (ꢀ)-metoprolol and 10 kV ꢁ 1 s for thiourea;
detection wavelength, 225 nm for (ꢀ)-metoprolol and 210 nm for thiourea; temperature, 20 1C. The back pressures were observed when the flow rate
was 3 mL minꢂ1 using water as the mobile phase.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
obtained and the peaks obviously broadened, which was attri- and column efficiency, the Rs reached the maximum at pH 7.0.
buted to the large temperature gradients that existed in wider Further increasing the pH to 8.5, obviously poor enantio-
capillaries.37,38 Hence, the 75 mm inner diameter HSA–cellulase@ selectivity occurred, which may be due to the fact that the
poly(GMA–EDMA) capillary monolithic column was chosen.
stability of the protein would be adversely affected at pH 8.5.
3.2.3. Effect of organic modifiers on enantioseparation. Therefore, the optimum pH of the running buffer was determined
Since organic modifiers can alter the enantioseparation by to be 7.0.
influencing interactions involved in the chiral recognition
3.2.5. Effect of buffer concentration on enantioseparation.
mechanism and EOF,8 an organic modifier was considered to Since the concentration of buffer solution can affect the buffering
be added into the buffer in order to achieve baseline separa- capacity, the resolution and column efficiency, the effect of the
tion. The chiral separation of (ꢀ)-metoprolol was achieved in phosphate buffer concentration (from 5 to 20 mmol Lꢂ1) on
10 mmol Lꢂ1 phosphate buffer with 0 to 14% 2-propanol and the enantioseparation of metoprolol enantiomers was evaluated.
effect of 2-propanol on EOF, retention time (tR), Rs and a is As illustrated in Fig. S5 (ESI†), the retention time of
illustrated in Table S1 (ESI†). With the increase of 2-propanol (ꢀ)-metoprolol was gradually prolonged with increasing buffer
percentage, EOF exhibited a decreasing tendency, which can be concentration, due to the reduced zeta potential as usually
explained by the change of dielectric constant/viscosity ratio and observed in CEC. Resolution was increased with increasing
the influence of mobile phase polarity on the zeta potential.39 For buffer concentration up to 10 mmol Lꢂ1. A further increase
(ꢀ)-metoprolol, an increase in the retention time and resolu- in buffer concentration gave declined resolution, which is
tion was observed when 2-propanol increased from 0% to likely due to the weakened electrostatic interaction between
10%. It is obvious to note that the longer retention time was protein and (ꢀ)-metoprolol. Hence, a buffer concentration of
ascribed to the decrease in EOF. The addition of 2-propanol 10 mmol Lꢂ1 was chosen for further analysis.
increased the affinity of the enantiomers for the chiral selector,
3.2.6. Effect of applied voltage on enantioseparation.
hence strengthening the interaction between the enantiomers The effect of applied voltage in the range of 5–25 kV on the
and the chiral stationary phase. However, by further increasing enantioseparation of (ꢀ)-metoprolol was evaluated and the
the 2-propanol concentration to 14%, the Rs lowered due to the results are presented in Fig. S6 (ESI†). As expected, when
second peak broadening. Therefore, 10% 2-propanol was chosen the applied voltage increased, the EOF was increased so that
for further investigation.
the retention time of (ꢀ)-metoprolol reduced. The resolution
3.2.4. Effect of pH on enantioseparation. The pH of the values were decreased upon the increase of voltage probably
running buffer could directly influence the ionization state of because of extra Joule heating and peak broadening. In addition,
the analyte and the EOF within the column; therefore, a change as the voltage increased, especially when the voltage exceeded
in pH could have a great influence on chiral separation by CEC. 20 kV, the voltage and current deviated linearly, indicating an
The effect of pH from 6.5 to 8.5 on the enantioseparation of increase in Joule heat. To reduce the effect of Joule heating, a
(ꢀ)-metoprolol was evaluated. As shown in Fig. 5, when the lower voltage should be used. In consideration of good resolution,
buffer pH increased, the retention time of metoprolol was shorter retention time and higher column efficiency, 10 kV was
gradually shortened, owing to the increasing EOF. On the other selected as the appropriate voltage for the separation of metoprolol
hand, the enantioselectivity increased when the buffer pH enantiomers.
changed from 6.5 to 8.0, indicating that the interacting differ-
ences of two metoprolol enantiomers with the protein would be
3.3. Stability and repeatability
enhanced as the ionization degree of the protein increased. Based on the relative standard deviations (RSD) of retention
Conversely, the column efficiency decreased with increasing time and resolution of metoprolol enantiomers, the repeatability
pH. This is probably because the enhanced interaction inten- of the HSA–cellulase@poly(GMA–EDMA) monolith was evaluated.
sifies the tailing of the peaks. Under the double influence of a For the intra-day repeatability test, the racemic metoprolol
Fig. 5 Effect of pH on the chiral separation of (ꢀ)-metoprolol on the HSA–cellulase@poly(GMA–EDMA) monolith. Experimental conditions: samples,
50 mg mLꢂ1 (ꢀ)-metoprolol; running buffer, 10 mmol Lꢂ1 phosphate buffer (pH 6.5–8.5, containing 10% 2-propanol), other conditions are the same as
in Fig. 4.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
standard solution was analyzed using five replicates within one day. of the HSA–cellulase@poly(GMA–EDMA) monolith. Moreover, after
The assessment of inter-day repeatability was carried out with a 60 runs over 30 days on the column, little changes in resolution
single capillary and it was performed by conducting three runs (only from 1.89 to 1.73) were observed and the intra-day repeatability
for three consecutive days. As for the column-to-column repeat- was still good (the RSD values of retention time and resolution were
ability, the same test analyte was analyzed on three columns 1.4% and 3.9%), suggesting good stability of the prepared polymer
made from one polymerization mixture. Batch-to-batch repeatability monolith.
was evaluated by another three columns that were fabricated from
three polymerization mixtures prepared using the same method.
The intra-day, inter-day, column-to-column and batch-to-batch RSD
3.4. Comparison with the corresponding single selector chiral
stationary phases
of the analyte’s retention time and resolution shown in Table 1 were
The chiral recognition ability of the HSA–cellulase@poly(GMA–
all less than 2.1% and 5.4%, which confirmed the good repeatability
EDMA) monolith was evaluated by different classes of enantio-
mers including amino acid, basic and acid chiral drugs. In the
experiment, there were ten enantiomers that could be enantio-
separated on the HSA–cellulase@poly(GMA–EDMA) monolith
and six of them could be completely resolved. For comparison,
the corresponding single selector CSPs were also fabricated
with the individual HSA and cellulase, respectively. As illustrated
in Fig. 6, the mixed selector CSP could separate 10 chiral analytes,
while the two single selector CSPs resolved 3 and 8 respectively.
It is demonstrated that the mixed selector CSP exhibited a
broader range of enantioselectivities than the single selectors.
Table
monolith
1
Repeatability data for the HSA–cellulase@poly(GMA–EDMA)
Rs(avg),
t1(avg),
t2(avg),
(% RSD) min (% RSD) min (% RSD)
Intraday (n = 5)
Interday (n = 9)
Column-to-column (n = 9) 4.2
Batch-to-batch (n = 9) 5.4
3.7
4.4
1.1
1.5
1.9
1.1
0.82
1.6
2.1
1.9
Experimental conditions are the same as in Fig. 3.
Fig. 6 Resolution of ten pairs of enantiomers on the HSA–cellulase@poly(GMA–EDMA) monolith, HSA@poly(GMA–EDMA) monolith and cellulase@
poly(GMA–EDMA) monolith. Experimental conditions: samples, 0.7 mg mLꢂ1 (ꢀ)-tryptophan, 50 mg mLꢂ1 (ꢀ)-metoprolol, 0.1 mg mLꢂ1 (ꢀ)-bisoprolol,
0.2 mg mLꢂ1 (ꢀ)-azelastine, 0.1 mg mLꢂ1 (ꢀ)-esmolol, 1 mg mLꢂ1 (ꢀ)-warfarin, 0.1 mg mLꢂ1 (ꢀ)-atenolol, 0.1 mg mLꢂ1 (ꢀ)-labetalol, 30 mg mLꢂ1
(ꢀ)-citalopram and 50 mg mLꢂ1 (ꢀ)-terazosin; running buffer, 10 mmol Lꢂ1 phosphate buffer (pH 7.0, without organic modifier for tryptophan, with
20% ACN for warfarin and with 10% 2-propanol for others); applied voltage, 20 kV for azelastine, terazosin, 15 kV for tryptophan, labetalol, warfarin and
10 kV for others; injection, 10 kV ꢁ 2 s for metoprolol, bisoprolol, labetalol, esmolol, warfarin, 1 kV ꢁ 2 s for atenolol, and 10 kV ꢁ 1 s for others;
detection wavelength, 215 nm for azelastine, tryptophan, labetalol, warfarin, 254 nm for terazosin, 240 nm for citalopram and 225 nm for others;
temperature, 20 1C.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
Except for (ꢀ)-warfarin, the resolution for racemates that could Moreover, for (ꢀ)-warfarin, the chiral recognition capacity was
be enantiomerically separated by just one of the two selectors, improved on the mixed selector CSP, which might be brought
was slightly lower on such mixed CSP, because the content about from the same elution order of enantiomers resolved by
of this selector in the mixed CSP is actually lower than the the two individual selectors. In such a case, the two selectors on
phase containing such a selector only. For (ꢀ)-tryptophan, a mixed CSP worked in a coordinative way that promoted the
(ꢀ)-metoprolol, (ꢀ)-bisoprolol, (ꢀ)-esmolol and (ꢀ)-azelastine, resolution. In general, by mixing HSA and cellulase together,
the resolution values on the mixed selector CSP were all greater the composite column combines the enantioselectivities of
than 1.5, completely meeting the general separation requirements. both individual proteins, thus expanding their application
Fig. 7 Electrochromatograms for enantioseparation of ten chiral analytes on the HSA–cellulase@poly(GMA–EDMA) monolith. The conditions are the
same as in Fig. 6.
Table 2 Comparison of serum albumin or cellulase based CSPs
Types of
protein
Backbone/methods
Analyte
Ref.
Poly(GMA–EDMA) monolith; HPLC
Poly(GMA–EDMA) monolith; HPLC
Poly(GMA–EDMA) monolith/
HSA
HSA
HSA
Warfarin, tryptophan
Tryptophan, phenylalanine
Warfarin, tryptophan
15
16
17
poly(GMA-TRIM) monolith; HPLC
GO modified silica monolith; CEC
HSA
Warfarin, tryptophan, salbutamol, chlortrimeton, propranolol, nefopam,
phenylalanine, azelastine, ibuprofen (with GO); tryptophan, nefopam,
azelastine (without GO)
18
Silica monolith; HPLC
Silica monolith; CEC
Silica monolith; CEC
Hybrid monolith; CEC
Gold nanoparticle (GNPs) modified
silica monolith; CEC
HSA
BSA
BSA
BSA
BSA
Warfarin, tryptophan, ibuprofen
Tryptophan, benzoin
Tryptophan
Tryptophan
Phenylthiocarbamyl amino acids
30
40
11
12
13
Silica monolith; CEC
Polystyrene nanoparticles; open-tubular
CEC (OT-CEC)
BSA
BSA
Tryptophan, pantoprazole, atenolol
Tyrosine, tryptophan, warfarin
14
41
GNPs; OT-CEC
Silica gels; HPLC
Poly(GMA–EDMA) monolith; CEC
BSA
FITC-labeled ephedrine and norephedrine
42
34
Cellulase
HSA and
cellulase
Propranolol, alprenolol, oxprenolol, pindolol
Warfarin, tryptophan, terazosin, azelastine, atenolol, metoprolol,
bisoprolol, esmolol, citalopram, labetalol
This work
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.

View Article Online
Paper
NJC
range practically. Fig. 7 shows the resultant enantioseparation Acknowledgements
electrochromatograms of ten chiral analytes on the HSA–
cellulase@poly(GMA–EDMA) monolith.
The authors are thankful for financial support from the Post-
graduate Research & Practice Innovation Program of Jiangsu
Province (KYCX17_0707).
3.5. Comparison of the work with previous reports
In previous reports, various serum albumin (HSA and BSA) or
cellulase based CSPs have been successfully applied for
enantioseparation research. It would be meaningful to com-
pare the analytical feature of the present method with other
previous reports. As shown in Table 2, an obvious wider
range of chiral recognition ability was obtained on the HSA–
cellulase@poly(GMA–EDMA) monolith with high column effi-
ciency and low reagent consumption except for research.18
In research,18 the introduction of graphene oxide (GO)
improved the chiral selectivity and the scope of the chiral
drugs that could be separated by HSA was extended. However,
without GO, there were only three enantiomers that could be
resolved on the HSA based silica monolith, much less than on
the HSA–cellulase@poly(GMA–EDMA) monolith. Furthermore,
it is worth mentioning that the results acquired from the
enantiomeric separation of tryptophan and warfarin in this
work were better than the related research18 considering the
resolution and separation efficiency. This implied that the
HSA–cellulase@poly(GMA–EDMA) monolith could be expected
to be a promising microscale enantioseparation device.
References
1 Q. Fu, K. Zhang, D. Gao, L. Wang, F. Yang, Y. Liu and Z. Xia,
Anal. Chim. Acta, 2017, 969, 63–71.
2 J. Zhou, B. Yang, J. Tang and W. Tang, New J. Chem., 2018,
42, 3526–3533.
3 E. J. Carrasco-Correa, E. F. Simo-Alfonso, G. Ramis-Ramos
and J. M. Herero-Martinez, Curr. Med. Chem., 2017, 24,
781–795.
4 Z. Chen, T. Nishiyama, K. Uchiyama and T. Hobo, Anal.
Chim. Acta, 2004, 501, 17–23.
¨
´
5 M. Lammerhofer, F. Svec and J. M. J. Frechet, Anal. Chem.,
2000, 72, 4623–4628.
6 M. Guerrouache, M. C. Millot and B. Carbonnier, Macromol.
Rapid Commun., 2009, 30, 109–113.
7 R. Noel Echevarria, E. J. Carrasco-Correa, S. Keunchkarian,
M. Reta and J. M. Herrero-Martinez, J. Sep. Sci., 2018, 41,
1424–1432.
8 S. Dixit, I. S. Lee and J. H. Park, J. Chromatogr. A, 2017, 1507,
132–140.
9 T. Koide and K. Ueno, J. Chromatogr. A, 2001, 909, 305–315.
10 Y. Zheng, X. Wang and Y. Ji, Talanta, 2012, 91, 7–17.
11 M. Kato, N. Matsumoto, K. Sakai-Kato and T. Toyo’oka,
J. Pharm. Biomed. Anal., 2003, 30, 1845–1850.
12 M. Kato, H. Saruwatari, K. Sakai-Kato and T. Toyo’oka,
J. Chromatogr. A, 2004, 1044, 267–270.
13 J. Lu, F. Ye, A. Zhang, Z. Wei, Y. Peng and S. Zhao, J. Sep.
Sci., 2011, 34, 2329–2336.
14 T. Hong, Y. Zheng, W. Hu and Y. Ji, Anal. Biochem., 2014,
464, 43–50.
4. Conclusions
In this work, a novel HSA–cellulase@poly(GMA–EDMA) mono-
lith was successfully prepared and used as a stationary phase
for enantioseparation by CEC. The prepared column exhibited
powerful enantioseparation capability towards different
chemical classes and there were ten chiral compounds that
could be separated on this column. Compared with the
individual HSA or cellulase based CSPs, the HSA–cellulase@
poly(GMA–EDMA) monolith has an even wider range of
applications. These results indicated that mixed selector CSPs
combining the enantioselectivities of individual CSPs could
be a useful tool for chiral separation, particularly in the
simultaneous analysis of racemic mixtures that can only be
resolved on different stationary phases. Further study will
include covalently bonding other proteins onto the monolith
to create other CSPs for the separation of different kinds
of chiral compounds in the same capillary. Additionally,
the strategy used here could be extended to immobilize
different enzymes in one microreactor for proteomic analysis
as well as fabricating dual-enzymatic systems used as a plat-
form to discover multitargeted enzyme-inhibitor drug lead
compounds.
15 R. Mallik, T. Jiang and D. S. Hage, Anal. Chem., 2004, 76,
7013–7022.
16 C. Yao, L. Qi, J. Qiao, H. Zhang, F. Wang, Y. Chen and
G. Yang, Talanta, 2010, 82, 1332–1337.
17 E. L. Pfaunmiller, M. Hartmann, C. M. Dupper, S. Soman
and D. S. Hage, J. Chromatogr. A, 2012, 1269, 198–207.
18 T. Hong, X. Chen, Y. Xu, X. Cui, R. Bai, C. Jin, R. Li and
Y. Ji, J. Chromatogr. A, 2016, 1456, 249–256.
19 E. Calleri, G. Massolini, D. Lubda, C. Temporini, F. Loiodice
and G. Caccialanza, J. Chromatogr. A, 2004, 1031, 93–100.
20 C. Temporini, E. Calleri, G. Fracchiolla, G. Carbonara,
F. Loiodice, A. Lavecchia, P. Tortorella, G. Brusotti and
G. Massolini, J. Pharm. Biomed. Anal., 2007, 45, 211–218.
21 R. Mallik, H. Xuan and D. S. Hage, J. Chromatogr. A, 2007,
1149, 294–304.
22 L. Zhao, L. Yang and Q. Wang, J. Chromatogr. A, 2016, 1446,
125–133.
23 T. Hong, C. Chi and Y. Ji, J. Sep. Sci., 2014, 37, 3377–3383.
24 S. Xu, R. Mo, C. Jin, X. Cui, R. Bai and Y. Ji, J. Pharm. Biomed.
Anal., 2017, 140, 190–198.
Conflicts of interest
There are no conflicts of interest to declare.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018

View Article Online
NJC
Paper
25 C. Miao, R. Bai, S. Xu, T. Hong and Y. Ji, J. Chromatogr. A, 34 H. Matsunaga and J. Haginaka, J. Chromatogr. A, 2016, 1467,
2017, 1487, 227–234.
155–162.
26 M. Ahmed and A. Ghanem, Chirality, 2014, 26, 754–763.
27 Z. Liu, K. Otsuka, S. Terabe, M. Motokawa and N. Tanaka,
Electrophoresis, 2002, 23, 2973–2981.
35 Z. A. Alothman, A. Aqel, H. A. Al Abdelmoneim, A. Yacine
Badjah-Hadj-Ahmed and A. A. Al-Warthan, Chromatographia,
2011, 74, 1–8.
28 C. Bi, X. Zheng, S. Azaria, S. Beeram, Z. Li and D. S. Hage, 36 I. Gusev, X. Huang and C. Horvath, J. Chromatogr. A, 1999,
Separations, 2016, 3, 27. 855, 273–290.
29 C. Bertucci and D. Tedesco, Curr. Med. Chem., 2017, 24, 37 E. Grushka, R. M. McCormick and J. J. Kirkland, Anal.
743–757. Chem., 1989, 81, 241–246.
30 R. Mallik and D. S. Hage, J. Pharm. Biomed. Anal., 2008, 46, 38 C. Pan, W. Lv, G. Wang, X. Niu, H. Guo and X. Chen,
820–830. J. Chromatogr. A, 2017, 1484, 98–106.
31 R. Isaksson, C. Pettersson, G. Pettersson, J. Ståhlberg, 39 Y. Tian, C. Zhong, E. Fu and Z. Zeng, J. Chromatogr. A, 2009,
J. Hermansson and I. Marle, Trends Anal. Chem., 1994, 13,
431–439.
32 T. B. Vinzant, W. S. Adney, S. R. Decker, J. O. Baker,
1216, 1000–1007.
40 M. Kato, K. Sakai-Kato, N. Matsumoto and T. Toyo’oka,
Anal. Chem., 2002, 74, 1915–1921.
M. T. Kinter, N. E. Sherman, J. W. Fox and M. E. Himmel, 41 D. Wang, X. Song, Y. Duan, L. Xu, J. Zhou and H. Duan,
Appl. Biochem. Biotechnol., 2001, 91–93, 99–107. Electrophoresis, 2013, 34, 1339–1342.
33 Y.-H. P. Zhang and L. R. Lynd, Biotechnol. Bioeng., 2004, 88, 42 H. F. Li, H. Zeng, Z. Chen and J. M. Lin, Electrophoresis,
797–824. 2009, 30, 1022–1029.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2018
New J. Chem.