Experimental and computational study on the enantioseparation of four chiral fluoroquinolones by capillary electrophoresis with sulfated-β-cyclodextrin as chiral selector

Article abstract of DOI:10.1002/chir.23340

In this work, enantioseparation of four chiral fluoroquinolones (FQs), namely, ofloxacin, gemifloxacin, lomefloxacin, and gatifloxacin, was achieved by capillary electrophoresis with sulfated-β-cyclodextrin (S-β-CD) as chiral selector. Factors affecting the enantiomeric resolution, such as the concentrations of S-β-CD, BGE pH conditions, and the buffer types and concentrations, were optimized and discussed. A BGE consisting of 30 g/L S-β-CD and 30-mM phosphate at pH?4.0 was found fit for enantiomeric resolution of ofloxacin and gemifloxacin, while the same BGE at pH?3.0 was suitable for enantioseparation of lomefloxacin and gatifloxacin. The pH-dependent experiments showed that separation resolutions of four FQs enantiomers were significantly affected by BGE pH, which was thought to be related with the varying electrostatic attraction between the enantiomers and chiral selector. To verify this speculation, molecular docking studies were used for further investigation of the enantiomeric recognition mechanism of S-β-CD. Molecular model indicated that hydrophobic effect and hydrogen bond were involved in host-guest inclusion, but the electrostatic attraction enhanced the chiral discrimination by increasing the difference in binding energy between individual enantiomers and S-β-CD. This work provided a further insight into the chiral recognition mechanisms of CD derivatives.

Full text of DOI:10.1002/chir.23340

Received: 3 April 2021
DOI: 10.1002/chir.23340
Revised: 9 June 2021
Accepted: 6 July 2021
R E S E A R C H A R T I C L E
Experimental and computational study on the
enantioseparation of four chiral fluoroquinolones by
capillary electrophoresis with sulfated-β-cyclodextrin as
chiral selector
Qianyun Ma | Wei Cong | Ye Liu | Zikai Geng | Ying Lin | Zhaokun Wang
School of Pharmacy, Binzhou Medical
Abstract
University, Yantai, China
In this work, enantioseparation of four chiral fluoroquinolones (FQs), namely,
Correspondence
Zhaokun Wang, School of Pharmacy,
ofloxacin, gemifloxacin, lomefloxacin, and gatifloxacin, was achieved by capil-
lary electrophoresis with sulfated-β-cyclodextrin (S-β-CD) as chiral selector.
Binzhou Medical University, 346 Guanhai
Road, Laishan District, 264033, Yantai,
Factors affecting the enantiomeric resolution, such as the concentrations of
Shandong Province, China.
S-β-CD, BGE pH conditions, and the buffer types and concentrations, were
Email: wangzhaokun@bzmc.edu.cn
optimized and discussed. A BGE consisting of 30 g/L S-β-CD and 30-mM phos-
phate at pH 4.0 was found fit for enantiomeric resolution of ofloxacin and
gemifloxacin, while the same BGE at pH 3.0 was suitable for enantioseparation
of lomefloxacin and gatifloxacin. The pH-dependent experiments showed that
separation resolutions of four FQs enantiomers were significantly affected by
BGE pH, which was thought to be related with the varying electrostatic attrac-
tion between the enantiomers and chiral selector. To verify this speculation,
molecular docking studies were used for further investigation of the enantio-
meric recognition mechanism of S-β-CD. Molecular model indicated that
hydrophobic effect and hydrogen bond were involved in host-guest inclusion,
but the electrostatic attraction enhanced the chiral discrimination by increas-
ing the difference in binding energy between individual enantiomers and
S-β-CD. This work provided a further insight into the chiral recognition mech-
anisms of CD derivatives.
K E Y W O R D S
chiral fluoroquinolones, chiral recognition mechanism, enantioseparation, molecular
docking, S-β-CD
1 | INTRODUCTION
norfloxacin, lomefloxacin, gatifloxacin, pazufloxacin,
enrofloxacin, gemifloxacin, and moxifloxacin. Due to
their excellent efficacy in the treatment of both human
and animal bacterial infections, they are widely used and
the sales rank among the top of all antibiotics.2 Lots of
FQs are chiral pharmaceuticals, possessing one or more
chiral centers in their chemical structures, such as
Fluoroquinolones (FQs) are a class of synthetic broad-
spectrum antimicrobials which are piperazinyl deriva-
tives of quinolones. FQs interfere with bacterial DNA
replication by inhibiting DNA gyrase and topoisomerase
IV.1 Common FQs include ofloxacin, levofloxacin,
Chirality. 2021;1–9.
wileyonlinelibrary.com/journal/chir
© 2021 Wiley Periodicals LLC.
1

2
MA ET AL.
ofloxacin, lomefloxacin, gatifloxacin, pazufloxacin, gemi-
floxacin, and moxifloxacin. It has been reported that the
antimicrobial activity of these chiral FQs showed signifi-
cant enantioselectivity. For example, the S-enantiomer of
ofloxacin was 8–128 times higher than its R-enantiomer
in vivo activity,3,4 and therefore, the single isomer of
S-ofloxacin was developed as a single enantiomeric drug.
For pazufloxacin, the S-enantiomer is more effective, and
the R-enantiomer has no antibacterial effect in vitro, so
the commercialized pazufloxacin was only supplied in
the form of its S-enantiomer.5
gemifloxacin, lomefloxacin, and gatifloxacin, and then
the mechanism of enantiomeric recognition was further
investigated by combining the experimental results and
molecular docking research, which provide an insight for
the application and mechanism investigation of S-β-CD
as chiral selector for enantioseparation of chiral FQs.
2 | MATERIALS AND METHODS
2.1 | Chemicals and reagents
Enantioseparation of chiral FQs is necessary for the
determination of the enantiomeric purity of chiral
compounds and to study the different biological and
pharmacological activities of enantiomers. However, only
a few methods are available now. The use of chiral sta-
tionary phase (CSP) and ligand exchange methods for
enantioseparation of FQs in high performance liquid
chromatography (HPLC) have been reported in the previ-
ous literature.6–9 Recently, capillary electrophoresis
(CE) continues to be used extensively in the field of chiral
separations due to its advantages of high efficiency
and resolution, short analysis time, and low cost and
producing small volume of waste solution.10,11 The
enantioseparation in CE requires the addition of
chiral selectors into a background electrolyte (BGE).
β-Cyclodextrin (CD) is one of the most commonly used
chiral selectors in CE, which is naturally occurring cyclic
oligosaccharides consisting of seven glucose units, owing
the ability to form host-guest inclusion with a wide range
of chiral molecules.12 However, the low solubility of
native β-CD limited its further usefulness in chiral sepa-
ration.13 By chemical modification of native CDs, the
obtained CD derivatives displayed significant improve-
ment in water solubility and chiral recognition potency.
During the past few decades, more and more CD deriva-
tives with different functionality have been designed
for diverse purposes, such as methylated β-CDs,14,15
hydroxypropylated β-CDs,2,16 and carboxymethylated
and sulfated β-CDs.17–20 Among them, sulfated β-CD
(S-β-CD) was most widely used, which possessed the sul-
fonic groups that could provide the negative charges over
the entire pH range and thereby formed the electrostatic
interaction with cationic compounds in addition to the
hydrophobic inclusion and hydrogen-bond interaction.13
It has been reported that S-β-CD presented good
enantioselectivity for diverse types of chiral analytes.19–21
Our previous study also found that S-β-CD could be used
for enantioseparation of β-blockers, phenethylamines,
and anticholinergic agents.22
(R,S)-Ofloxacin (OFLX), (R,S)-gemifloxacin (GMLX),
(R,S)-lomefloxacin (LMLX), and (R,S)-gatifloxacin
(GTLX) were purchased from Chinese Food and Drug
Inspection Institute (Beijing, China) as racemic mix-
tures. A single enantiomer, S-ofloxacin was supplied by
Aladdin Biochemical Technology Co., Ltd. (Shanghai,
China). The corresponding chemical structures and pKa
values were provided in Table 1. S-β-CD with the
degree of substitution of 7–11 was supplied by Sigma-
Aldrich (St. Louis, MO, USA). Sodium hydroxide,
sodium dihydrogen phosphate, and phosphoric acid
were of analytical grade and purchased from Tianjin
Bodi Chemical Plant (Tianjin, China). Methanol of
HPLC grade was from Tianjin Concord Technology Co.,
LTD (Tianjin, China). All other used reagents were of
analytical grade, and the ultrapure water was used
throughout our study.
2.2 | Apparatus
Enantioseparation of four chiral FQs was performed with
a Beckman P/ACE™ MDQ Capillary Electrophoresis
System (Beckman, Fullerton, CA, USA), equipped with a
diode-array detection; 32 Karat 8.0 Software (Beckman,
Fullerton, CA, USA) was used for instrument control,
data collection, and data analysis. An uncoated, fused
silica capillary (Ruifeng, Yongnian, Hebei, China),
50 cm ꢀ 50 μm (effective length 40 cm), was used for chi-
ral separation.
2.3 | Methods
The background electrolyte solutions (BGEs) were pre-
pared as follows: First, sodium phosphate was dissolved
in water, and then the chiral selectors were added into
the solution. After that, the solution was adjusted to
the desired pH with phosphoric acid or sodium hydrox-
ide at the concentration of 1.0 mol/L. All analytical
samples were prepared at a concentration of 0.1 mg/ml
In the present work, S-β-CD was used as chiral
selector for enantioseparation of four typical chiral
FQs by capillary electrophoresis, namely, ofloxacin,

MA ET AL.
3
TABLE 1 The chemical structures and pKa values of four FQs
2 ðt2 ꢂt1Þ
W2 þW1
Rs ¼
ð1Þ
FQs
Structures
pKa
OFLX
5.20
7.37a
t2
α ¼
ð2Þ
t1
where t1 and t2 are the migration times. W1 and W2 are
the baseline peak widths of the two enantiomers,
respectively. The magnitude of electroosmotic mobility
(μeof) was calculated based on the migration time of
neutral marker (acetone), by the use of the following
equation:
GMLX
6.02
9.15a
LMLX
GTLX
5.64
5.69
8.70b
8.73b
Lꢃl
μeof
¼
ð3Þ
teof ꢃV
where L is the total capillary length, l is the effective
capillary length, teof is the migration time of acetone, and
V is the applied voltage. The magnitude of apparent
mobility (μap), electrophoretic mobility (μep), and electro-
osmotic mobility (μeof) can be correlated by the following
equation:
Note: * indicates chiral center.
ahttps://scifinder.cas.org
bhttps://go.drugbank.com/
μap ¼ μepþμeof
ð4Þ
2.5 | Molecular docking
by dissolving the standards into water:methanol (50:50,
v/v) and then passed through the 0.45-μm syringe filter
before injection. Before a new capillary was used for
separation, it was first rinsed with sodium hydroxide
(1.0 mol/L) for 30 min and then water for 10 min at
20 psi. For each day, the capillary was rinsed with
0.1 mol/L sodium hydroxide solution, followed by
water for 10 min, and then conditioned with BGE for
10 min before analysis. Between consecutive runs, the
capillary was rinsed with BGE for 2 min to eliminate
the interference from the previous analysis. Injections
were hydrodynamically made at 0.5 psi for 5 s. Separa-
tions were performed at the voltage of 20 kV (applied
to the cathode or anode), and all samples were
detected at 260 nm. The capillary was thermostated at
25ꢁC by a capillary cartridge.
Molecular construction was carried out on the
ChemBioDraw Ultra14.0 software package (CambridgeSoft,
Massachusetts, USA), and Discovery Studio 2019
Client software (Accelrys Inc., San Diego, CA, USA)
was applied for structure optimization. The three-
dimensional structure of S-β-CD was constructed by
referencing the crystal structures of β-CD extracted from
β-CD_α-hemolysin complex crystal structure (PDB entry
code: 3M3R23). The obtained structure of S-β-CD was
then optimized by Discovery Studio 2019 before docking
process.
Discovery Studio 2019 software was used for molecu-
lar docking. Consecutive attempt was made by auto-
docking the ligand to the host via different interaction
models. During the docking process, S-β-CD and each
enantiomer were set as the host and guest (the ligand),
respectively. For each enantiomer, the obtained best
model with the lowest binding energy was selected as
the optimal interaction model. The obtained binding
energy (BE) reflected the interaction intensity between
the host and guest, which was calculated by following
equation:
2.4 | Calculations
The chiral separation parameters in CE were calculated
in the following manner. The resolution (Rs) and the
effective electrophoretic selectivity (α) were calculated by
using Equations 1 and 2, respectively.
BE ¼ Ecomplex
ꢂ
Eguest
ꢂ
Ehost
ð5Þ

4
MA ET AL.
where Ecomplex, Eguest, and Ehost represented the energy
of the complex, free guest molecule, and free host
molecule, respectively. The difference in the binding
energy of R,S-enantiomers interacted with S-β-CD led to
the distinction existing in their migratory behavior,
which was the basics of chiral separation in capillary
electrophoresis.
3 | RESULTS AND DISCUSSION
In the preliminary study, three types of neutral β-CDs,
namely, native β-CD, methylated β-CD (Me-β-CD), and
hydroxypropylated β-CD (HP-β-CD), were first tested for
chiral separation. It was indicated that Me-β-CD and
HP-β-CD showed slight chiral recognition ability to the
enantiomers of OFLX, but no sign was observed for
enantioseparation of GMLX, LMLX, and GTLX. How-
ever, S-β-CD displayed good enantioselectivity for all the
four chiral FQs and was subsequently used as the chiral
selector in further studies. The effects of experimental
parameters on the enantioseparation were thoroughly
discussed.
3.1 | Effect of the concentration of
S-β-CD
The concentration of chiral selector is an essential
parameter for enantioseparation.2 To find out an opti-
mum concentration, S-β-CD at different concentrations
was investigated in the range of 5–40 g/L. The detection
was set at the anodic side (reversed polarity) with the
BGE conditions of 30-mM phosphate buffer solution at
pH of 3.0.
FIGURE 1 Variation of migration times (A), enantiomeric
resolution (B) of four chiral FQs with addition of S-β-CD from 10 to
40 g/L. Buffer condition: 30-mM phosphate, pH 3.0, voltage:
ꢂ20 kV (applied to the anode)
The migration times of four FQs enantiomers tended
to decrease with the increase of S-β-CD concentrations.
In general, a high concentration of chiral selector would
enhance the interaction between analytes and chiral
selectors, resulting in a prolonged migration time.22 How-
ever, the increasing viscosity and ionic strength due to
the dissociation of S-β-CD also suppressed the electroos-
motic flow and caused a lower electroosmotic mobility
(μeof) and a higher apparent mobility (μap), leading to a
shorter migration time in our study. As to the enantio-
meric resolutions, no obvious variation was observed
with addition of different concentrations of S-β-CD.
Usually, the decreased migration time meant the reduced
interaction intensity and poor resolutions. However, the
increase of S-β-CD also strengthened the interaction
between the enantiomers and chiral selector, which
was benefit for the improvement of enantioseparation.
Therefore, there was a balance between the migration
time and enantiomeric resolutions. According to the
results showed in Figure 1, when S-β-CD at 30 g/L was
used as chiral selector, both proper migration times and
enantiomeric resolutions were obtained for four FQs
enantiomers.
3.2 | Effect of buffer pH
Buffer conditions would affect the electrophoresis
mobility (μep) and electroosmotic mobility (μeof) by
changing the dissociation rate of analytes and Si–OH
groups on the inner capillary wall. It was reported that
enantioseparation of ofloxacin and its metabolites in cap-
illary electrophoresis by using sulfobutyl β-cyclodextrin

MA ET AL.
5
as chiral selector could be possible at all pH values
except at the isoelectric points.24 In our study, the
enantioseparation of four chiral FQs was investigated
over the pH range of 2.5–8.0 with addition of S-β-CD at
30 g/L. Table 2 showed the results of enantiomeric
resolutions (Rs), migration times (t), and effective
electrophoretic selectivity (α) at different BGE pH
conditions.
When BGE pH varied from 2.5 to 5.0, the voltage was
applied the anode (ꢂ20 kV) because the electrophoretic
mobility (μep, negative value) was larger than electroos-
motic mobility (μeof, positive value), and the migratory
direction of the enantiomers was in keep with the
electrophoretic mobility. With the rise of BGE pH, both
migration times and resolutions of the enantiomers
tended to increase, mainly due to the increasing EOF
reduced the magnitude of apparent mobility (μap).
Once BGE pH reached over 6.0, the increased EOF
caused a change of apparent mobility toward the
cathode. Therefore, over the pH range of 7.0–8.0,
enantioseparation of four chiral FQs was carried out at
normal polarity mode, and the migration times were
shortened with the rise of BGE pH.25 In some cases, the
magnitudes of μep and μeof were very close, which caused
that the μap was almost zero and the enantiomers could
not be detected on neither cathode or anode within
40 min, for example, at pH 6.0. As to the enantiomeric
resolutions, there were no significant variation tendency
when pH in the range of 2.5–5.0. For OFLX and GMLX,
the best enantiomeric resolutions were obtained at
pH 3.0. For LMLX and GTLX, the best BGE pH for
enantioseparation was 4.0. However, when pH was
higher than 7.0, the enantiomeric resolutions of four FQs
rapidly decrease. For GMLX, LMLX, and GTLX, the
enantiomeric recognition and separation potency even
completely disappeared.
Effects of BGE pH conditions on the mobility
parameters of the enantiomers were also evaluated.
Figure 2 showed the variation of μap, μeof, and μep with
the BGE pH changing in range of 2.5–8.0. As shown in
Figure 2, μap was obviously varied with μeof, which was
affected by BGE pH. However, the values of μep hardly
changed from pH 2.5 to 8.0, which may be the reason
why enantiomeric resolutions showed no obvious change
even though the migration times varied in the range of
pH 2.5–5.0. When pH got higher than 7.0, the charge
state of the enantiomer began to change. Four chiral FQs
are amphoteric compounds, possessing both carboxyl and
amino groups in their chemical structures, and their
isoelectric points are situated between 6.5 and 7.5.26 As
BGE pH in the range 2.5–5.0, the enantiomers mainly
existed as cations. Meanwhile, as a salt of strong acid,
TABLE 2 Results of migration
pH
times, enantiomeric resolutions, and
Analyte
2.5
3.0
4.0
5.0
6.0
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
7.0
8.0
5.83
5.83
1.00
0
selectivity factors of four FQs at
different BGE pH
OFLX
t1/min
t2/min
α
11.4
12.8
13.0
14.9
13.4
15.4
15.0
17.1
7.75
7.83
1.01
0.42
7.73
7.73
1.00
0
1.12
4.74
10.0
10.6
1.06
2.00
14.6
14.9
1.02
1.05
15.3
15.7
1.03
1.25
1.14
5.32
10.8
11.4
1.06
2.23
16.5
16.8
1.02
1.22
17.2
17.7
1.04
1.30
1.15
5.35
11.3
12.1
1.07
2.35
17.5
17.8
1.02
0.98
18.5
19.0
1.03
1.19
1.14
5.32
12.8
13.6
1.06
2.33
18.1
18.4
1.02
0.85
19.5
19.9
1.02
1.03
RS
GMLX
LMLX
GTLX
t1/min
t2/min
α
5.6
5.62
1.00
0
RS
t1/min
t2/min
α
7.13
7.13
1.00
0
5.56
5.56
1.00
0
RS
t1/min
t2/min
α
6.83
6.83
1.00
0
5.42
5.42
1.00
0
RS
Abbreviation: nd, no signal was detected.
Note: Buffer condition: 30 g/L S-β-CD, 30-mM phosphate, voltage: 20 kV (pH 2.5–5.0 applied to the anode;
pH 7.0–8.0 applied to the cathode).

6
MA ET AL.
3.4 | Effect of chemical structures of FQs
Typical electropherograms of four FQs enantiomers were
shown in Figure 3. Regrettably, even though the
separation conditions were systematically optimized in
our study, only the enantiomers of OFLX and GMLX
were baseline resolved. For LMLX and GTLX, the best
resolutions of the enantiomers were 1.22 and 1.30,
respectively. As reported in the previous literature,
chemical structures of the molecules were important in
the enantiomeric recognition, especially the position and
distribution of the chiral groups.2,27 It was speculated
that the presence of aromatic rings in the vicinity of
chiral center might be favorable to enantiomeric
recognition due to the steric repulsion.28 Four chiral FQs
are similar in their basic molecular skeleton, but only
OFLX has a chiral center close to its aromatic system,
which could explain why OFLX enantiomers got the best
resolution in the four FQs. For LMLX and GTLX, they
shared the same chiral center and the enantiomeric
resolutions of their enantiomers were also similar.
FIGURE 2 Variation of μap, μeof and μep of OFLX enantiomers
with increasing BGE pH from 2.5 to 8.0. Buffer condition:
30 g/L S-β-CD, 30-mM phosphate, voltage: 20 kV (pH 2.5–5.0
applied to the anode; pH 7.0–8.0 applied to the cathode)
the charge state of S-β-CD showed pH independence. The
electrostatic attraction between the host (S-β-CD) and the
guest (enantiomer) was favorable for enantioseparation.
Once pH got higher than 7.0, the enantiomers mainly
existed as anions, which had a negative effects on
enantiomeric recognition, causing the resolutions of FQs
enantiomers nearly disappeared except that OFLX still
can be partly separated at pH 7.0.
3.5 | Interactions responsible for
enantiomeric recognition
The molecular docking technique was further applied
to present the host–guest interaction models between
S-β-CD and FQs enantiomers, which would help us better
understand the interactions affecting enantioseparation.
It should be mentioned that the used S-β-CD in our
research is a sulfated mixture, and the degree of the
substitution (DS) is in the range of 7–11. In the docking
process, DS of the used S-β-CD was counted as 9, which
was the average DS of the multicomponent mixture.
Figure 4 displayed the interaction models of R,S-
enantiomers of OFLX (positive charge state) interacted
with S-β-CD, top view (A) and side view (B), respectively.
It was obvious that each enantiomer inserted into the
hydrophobic cavity of S-β-CD and the groups next to the
chiral center pointed to the rim of cyclodextrin. Litera-
ture reported that hydrophobic interaction (not shown in
Figure 4) between the phenyl rings and the cavity of
cyclodextrin was responsible for host–guest inclusion,26
and the difference in hydrogen bonding (green dotted
line in Figure 4) was helpful to increase the chiral dis-
crimination.27,29 However, in our study, the electrostatic
attraction (brown dotted line in Figure 4) seemed to play
an important role for enantiomeric recognition. When
R,S-enantiomers existed as the neutral molecules, the
difference in their biding energy with S-β-CD
(ΔBE = 0.08 kcal/mol) was insufficient to cause any
enantiomeric resolution. When as anions, the difference
3.3 | Effect of buffer type and
concentration
The chemical composition and concentration of buffer
solution also affected the enantioselectivity, baseline
stability, and peak shapes. In our study, three types of
buffer solutions, including phosphate, Tris–HCl, and
acetate solutions, were tested as BGEs. It was indicated
that when phosphate buffer was used, the best enantio-
meric resolutions and proper peak shapes were obtained.
Thus, the phosphate buffer was selected as BGE solution.
In addition, the buffer concentration was also optimized
according to the migration times and resolutions. With
rising phosphate buffer concentration from 10 to 30 mM
at pH 3.0, the migration times were prolonged, and
enantiomeric resolutions were also improved. However,
upon rising concentrations higher than 30 mM, no
obvious improvement on the resolutions was observed
even though the migration times continued increasing.
Considering both the separation efficiency and migra-
tion times, 30-mM phosphate buffer was selected for
enantioseparation.

MA ET AL.
7
FIGURE 3 Typical electropherograms of
the enantiomers of OFLX, GMLC, LMLX and
GTLX. BGE conditions: 30 g/L S-β-CD, 30-mM
phosphate, pH 3.0, voltage: ꢂ20 kV (applied to
the anode)
FIGURE 4 Molecular interaction models
of OFLX R,S-enantiomers interacted with
S-β-CD, top view (A) and side view (B)
TABLE 3 The calculated binding
energy (BE) of OFLX R,S-enantiomers
interacted with S-β-CD at different
charge states
BE (kcal/mol)
Charge states
Neutrality
Anion
R-enantiomer
ꢂ19.72
ꢂ20.28
S-enantiomer
ꢂ19.80
ꢂ19.10
ΔBE (kcal/mol)
0.08
1.18
2.12
Cation
ꢂ17.68
ꢂ19.80
in the biding energy (ΔBE = 1.18 kcal/mol) caused part
resolution of OFLX enantiomers, which was
corresponding to the enantioseparation of OFLX at
pH 7.0. When as cations, the difference in their binding
energy was further enlarged (ΔBE = 2.12 kcal/mol),
which gave rise to a complete enantiomeric resolution of

8
MA ET AL.
lomefloxacin in human plasma. J Pharm Biomed Anal. 1995;
13(10):1243-1248.
8. Lehr KH, Damm P. Quantification of the enantiomers of
ofloxacin in biological fluids by high-performance liquid
chromatography. J Chromatogr. 1988;425:153-161.
OFLX, corresponding to the enantioseparation of OFLX
at pH 2.5–5.0. Table 3 listed the data of the calculated
binding energy of OFLX R,S-enantiomer interacted with
S-β-CD.
9. Zeng S, Zhong J, Pan L, Li Y. High-performance liquid chroma-
tography separation and quantification of ofloxacin enantio-
mers in rat microsomes. J Chromatogr B. 1999;728(1):151-155.
10. Suntornsuk L. Recent advances of capillary electrophoresis in
pharmaceutical analysis. Anal Bioanal Chem. 2010;398(1):
29-52.
11. Pan C, Wang W, Chen X. Recent developments of chiral sepa-
ration by capillary electrophoresis. Chin J Chromatogr. 2016;
34(1):16-20.
12. Xiao Y, Ng SC, Tan TT, Wang Y. Recent development of
cyclodextrin chiral stationary phases and their applications in
chromatography. J Chromatogr A. 2012;1269:52-68.
13. Potiera IL, Tamisier-Karolaka SL, Morinb P, Megelc F,
Tavernaa M. Comparison of native, alkylated and charged
cyclodextrins for the chiral separation of labetalol stereoiso-
mers by capillary electrophoresis. J Chromatogr A. 1998;
829(1-2):341-349.
14. Yue J, Cheng C, Meng L, Chen J, Li M, Zhu Z. Simultaneous
and sensitive capillary electrophoretic enantioseparation of
three β-blockers with the combination of achiral ionic liquid
and dual CD derivatives. Talanta. 2012;89:149-154.
15. Li W, Zhao L, Zhang H, et al. Enantioseparation of new
triadimenol antifungal active compounds by electrokinetic
chromatography and molecular modeling study of chiral recog-
nition mechanisms. Electrophoresis. 2014;35(19):2855-2862.
16. Zhao M, Cui Y, Yu J, Xu S, Guo X. Combined use
of hydroxypropyl-cyclodextrin and ionic liquids for the
simultaneous enantioseparation of four azole antifungals by
CE and a study of the synergistic effect. J Sep Sci. 2014;37(1-2):
151-157.
4 | CONCLUSION
In this work, S-β-CD with the degree of substitution
of 7–11 was applied for efficient enantioseparation
of four chiral FQs, namely, ofloxacin, gemifloxacin,
lomefloxacin, and gatifloxacin. The important parameters
affected the enantioseparation, such as the concentra-
tions of chiral selector, BGE pH and buffer types and con-
centrations were optimized. Further, the enantiomeric
recognition mechanism of S-β-CD was also investigated
by combining the experimental results and molecular
docking study. It was indicated that the electrostatic
attraction between the enantiomer and S-β-CD played an
crucial role in chiral discrimination. This work provided
not only a applicable method for enantioseparation of
chiral FQs, but also a new insight into the chiral recogni-
tion mechanism of S-β-CD.
DATA AVAILABILITY STATEMENT
Research data are not shared.
ORCID
Zhaokun Wang https://orcid.org/0000-0003-2033-9880
REFERENCES
17. Yu J, Zhao Y, Song J, Guo X. Enantioseparation of meptazinol
and its three intermediate enantiomers by capillary electropho-
resis using a new cationic β-cyclodextrin derivative in single
and dual cyclodextrin systems. Biomed Chromatogr. 2014;28(6):
868-874.
1. Drusano G, Labro-T M, Cars O, et al. Pharmacokinetics and
pharmacodynamics of fluoroquinolones. Clin Microbiol Infect.
1998;4:2S27-2S41.
2. Zhou S, Ouyang J, Baeyens WRG, Zhao H, Yang Y. Chiral
separation of four fluoroquinolone compounds using capillary
electrophoresis with hydroxypropyl-cyclodextrin as chiral
selector. J Chromatogr A. 2006;1130(2):296-301.
3. Morrissey I, Hoshino K, Sato K, et al. Mechanism of
differential activities of ofloxacin enantiomers. Antimicrob
Agents Chemother. 1996;40(8):1775-1784.
4. Hassan RM, Yehia AM, Saleh OA, El-Azzouny AA,
Aboul-Enein HY. Structure-retention relationship for enantio-
separation of selected fluoroquinolones. Chirality. 2018;30(6):
828-836.
5. Zhang Z, Yang G, Wang D, Liang G, Chen Y. Chiral Separation
and enantiomeric purity determination of pazufloxacin
mesilate by HPLC using chiral mobile phase additives. J Liq
Chromatogr R T. 2004;27(5):813-827.
ꢀ
18. Gomara B, García-Ruiz C, Marina ML. Enantioselective
separation of the sunscreen agent 3-(4-methylbenzylidene)-
camphor by electrokinetic chromatography: quantitative
analysis in cosmetic formulations. Electrophoresis. 2005;26(20):
3952-3959.
19. Fonseca MC, Silva RCSD, Nascimento CS, Borges KB.
Computational contribution to the electrophoretic enantiomer
separation mechanism and migration order using modified
β-cyclodextrins. Electrophoresis. 2017;38(15):1860-1868.
20. Zhou L, Lun J, Liu Y, Jiang Z, Di X, Guo X. In situ immobiliza-
tion of sulfated-β-cyclodextrin as stationary phase for capillary
electrochromatography enantioseparation. Talanta. 2019;
200:1-8.
21. Lin CE, Ko TC, Kuo CM, et al. Determination of
enantiomerization barrier of thioridazine by dynamic capillary
electrophoresis using sulfated cyclodextrins as chiral selectors.
Electrophoresis. 2009;30(17):3071-3078.
6. Grellet J, Ba B, Saux MC. High-performance liquid
chromatographic separation of fluoroquinolone enantiomers: a
review. J Biochem Biophys Methods. 2002;54(1-3):221-233.
7. Foster RT, Carr RA, Pasutto FM, Longstreth JA. Stereospecific
22. Wang Z, Zhang Q, Luo L, Sun T, Guo X. Comparison of three
S-β-CDs with different degrees of substitution for the chiral
high-performance
liquid
chromatographic
assay
of

MA ET AL.
9
separation of 12 drugs in capillary electrophoresis. Chirality.
2017;29(9):558-565.
28. Bao Y, Lantz AW, Yao T, Huang Q, Larock RC,
Armstrong DW. Enantiomeric separation of furan deriva-
tives and fused polycycles by cyclodextrin-modified micel-
lar capillary electrophoresis. Electrophoresis. 2005;26(21):
4164-4171.
29. Zhou J, Ai F, Zhou B, Tang J, Ng SC, Tang W.
Hydroxyethylammonium monosubstituted cyclodextrin as
chiral selector for capillary electrophoresis. Anal Chim Acta.
2013;800:95-102.
23. Banerjee A, Mikhailova E, Cheley S, et al. Molecular bases of
cyclodextrin adapter interactions with engineered protein
nanopores. Proc Natl Acad Sci U S A. 2010;107(18):8165-8170.
24. Horstkötter C, Blaschke G. Stereoselective determination of
ofloxacin and its metabolites in human urine by capillary
electrophoresis using laser-induced fluorescence detection.
J Chromatogr B. 2001;754(1):169-178.
25. Rudaz S, Calleri E, Geiser L, Cherkaoui S, Prat J, Veuthey JL.
Infinite enantiomeric resolution of basic compounds using
highly sulfated cyclodextrin as chiral selector in capillary elec-
trophoresis. J Chromatogr A. 1997;766:215-224.
26. Sun SW, Chen LY. Optimization of capillary electrophoretic
separation of quinolone antibacterials using the overlapping
resolution mapping scheme. J Chromatogr A. 1997;766(1-2):
215-224.
27. Li W, Zhao L, Zhang H, et al. Molecular modeling study of
chiral separation and recognition mechanism of β-adrenergic
antagonists by capillary electrophoresis. Electrophoresis. 2014;
35(19):2855-2862.
How to cite this article: Ma Q, Cong W, Liu Y,
Geng Z, Lin Y, Wang Z. Experimental and
computational study on the enantioseparation of
four chiral fluoroquinolones by capillary
electrophoresis with sulfated-β-cyclodextrin as
chiral selector. Chirality. 2021;1-9. https://doi.org/
10.1002/chir.23340