Separation, Quantification and Structural Study of (+)-Catechin and (–)-Epicatechin by Ion Mobility Mass Spectrometry Combined with Theoretical Algorithms

Article abstract of DOI:10.1002/cjoc.201900057

Two catechin epimers and their non-covalent complexes with γ-cyclodextrin were studied by using ion mobility coupled with mass spectrometry (IM-MS). Rapid separation of complexes was achieved with the peak-to-peak resolution reaching 0.86 after optimizati

Full text of DOI:10.1002/cjoc.201900057

Accepted Article
Title: Separation，quantification and structural study of (+)-catechin and (-)-
epicatechin by ion mobility mass spectrometry combined with theoretical
algorithms
Authors: Xinyu Bian, Bing Zhao, Bo Pang, Zhong Zheng, Shu Liu,* Zhiqiang Liu and
Fengrui Song*
This manuscript has been accepted and appears as an Accepted Article online.
This work may now be cited as: Chin. J. Chem. 2019, 37, 10.1002/cjoc.201900057.
The final Version of Record (VoR) of it with formal page numbers will soon be
published online in Early View: http://dx.doi.org/10.1002/cjoc.201900057.
ISSN 1001-604X • CN 31-1547/O6
mc.manuscriptcentral.com/cjoc
www.cjc.wiley-vch.de

Separation，quantification and structural study of (+)-catechin and
(-)-epicatechin by ion mobility mass spectrometry combined with
theoretical algorithms
Xinyu Bian,a,b Bing Zhao,a Bo Pang,a Zhong Zheng,a Shu Liu,*,a Zhiqiang Liua and Fengrui Song*,a,b
aNational Center for Mass Spectrometry in Changchun & Jilin Province Key Laboratory of Chinese Medicine Chemistry and Mass Spectrometry, Changchun
Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China
bUniversity of Science and Technology of China, Hefei, 230029, China
Cite this paper:Chin. J. Chem.2019,37, XXX—XXX.DOI: 10.1002/cjoc.201900XXX
Two catechin epimers and their non-covalent complexes with γ-cyclodextrin were studied by using ion mobility coupled with mass spectrometry
(IM-MS). Rapid separation of complexes was achieved with the peak-to-peak resolution reaching 0.86 after optimization of IM condition. Collision cross
section (CCS) was measured to explore the structural difference of complexes. A gap of 11.75 Å2 between two complexes was found. Molecular modeling
and theoretical CCS calculation were adopted to explain the measurement results. Two binding ways of both complexes were found and the calculated
CCS corresponds accurately to the measured CCS. Quantification of catechins in mixtures were performed and the relative error was less than 15%,
indicating the effectiveness of quantification by IM-MS.
to LC or GC but spends much less time (miliseconds) than the two
methods (minutes). Combining IM with MS (IM-MS) provides an
additional analytic dimension, which makes the conformational
Background and Originality Content
Catechins are part of the chemical family of flavan-3-Ol and
widely found in plants such as tea,[1] cocoa,[2] peaches[3] and so on.
Catechins as well as their gallic acid conjugates are seen as a
family which is a kind of important functional components in
plants. Bioactivity studies showed that catechins have various
information of molecules as an important segment in explaining
structural differences or intermolecular interaction. Recently,
IM-MS has been used in studying biomolecules such as
oligonucleotides,[15] carbohydrates,[16] lipids,[17] proteins[18] and
structural description of complexes like protein-inhibitor.[19]
Collision cross section (CCS) is usually adopted to describe the
structural information by calculating rotationally averaged cross
section of a molecule based on the overall size and molecular
architecture.[20] CCS can be experimentally measured by IM to
provide a quantified result of structure information. In addition,
CCS can also be calculated from theoretical models derived by
molecular dynamics (MD) simulations or other modelling
techniques.[21] Comparing CCS from these two methods makes the
cognition of molecule structures more reliable.
In this study, we reported our attempt of separating CA and
EC by IM-MS method. In consideration of potential problem of
inadequate resolution of IM,[22] several kinds of cyclodextrins
were introduced to explore the possibility of separating the two
catechins at a higher resolution. Cyclodextrins (CDs) are chiral,
cyclic oligosaccharides comprised of D-glucopyaranoside units
linked together by α→1, 4 glycosidic linkage. Structures of three
typical CDs were shown in Figure 1. These amphiphilic molecules
were widely used in the chiral recognition method usually called
as host-guest method because of their multiple chiral centers.[23]
Our present work carried out a further study on the interaction
between CDs and catechins. It was the first attempt to study the
conformations of catechin-CD complexes by combining IM-MS
measurement with theoretical molecular model construction and
CCS calculation. In addition, the quantitative capability of IM-MS
method was evaluated by quantifying catechin-CD mixtures.
physiological
effects
like antibacterial,[4]
antioxidant,[5]
anticarcinogenic[6] and antiallergic[7] properties.
The non-gallated catechins consist of two phenyl rings A and B
with four phenolic hydroxyl groups and a pyran ring C with a
hydroxyl group on carbon 3 (Figure 1). Obviously, there are
two chiral centers on carbons
diastereoisomers which are
2
and 3, bringing four
(+)-catechin, (-)-catechin,
(+)-epicatechin and (-)-epicatechin. The most common isomers
found in plants and food are (+)-catechin (CA) and (-)-epicatechin
(EC). Studies indicated differences of the two catechins in
bioavailability,[8] antioxidant[9] and other properties. Therefore, it
is necessary to distinguish and quantify different catechin isomers
in complex samples.
Traditionally, high-performance liquid chromatography mass
spectrometry (HPLC-MS) and gas chromatography (GC) were the
most frequently used methods for separating and analyzing
complicated components including isomers.[10, 11] As for CA and EC,
thin-layer chromatography (TLC)[12] and capillary electrophoresis
(CE)[13] were also adopted as identification methods. Although
these techniques provide high sensitivity or separation resolution
for isomer analysis, long period of analysis and tedious
pretreatment may reduce the analysis efficiency.
Ion mobility spectrometry (IMS or IM) is another effective way
for separating molecules in gas phase based on their size and
shape.[14] Unlike differentiating ions analyzed by their mass to
charge (m/z) in MS or tandem MS, IM separates ions according to
their flight time in the buffer gas. This method is somehow similar
Results and Discussion
IM-MS analysis of CA/EC-CDs
Some isomers are highly different in structure, causing distinct
CCS and easy separation in IM. But some isomers, especially
Enough conformation difference is necessary for IM separation.
*E-mail: mslab20@ciac.ac.cn; songfr@ciac.ac.cn; Tel.:0086-0431-85262044; Fax: 0086-0431-85262044.
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which
may lead to differences between this version and the Version of Record. Please cite this
article as doi: 10.1002/cjoc.201900057
This article is protected by copyright. All rights reserved.

epimers with small group at isomerism sites, may have nearly the
same CCS. One strategy for chiral separation by IM is using chiral
gas as buffer gas.[24] However, as chiral gas molecules have more
complex structures than helium or nitrogen, theoretical model of
collisions and interactions in gas phase may be much more
unpredictable, which makes caculating CCS extremely difficult.
Another way, as described above, is making covalent or
non-covalent complexes of target analytes with other molecules
and measuring the complexes. Amino acids and cyclodextrins are
usually used as chiral reagent.[22,25] The selectivity from their chiral
center makes it easy to form different conformation of complexes.
In this work, five kinds of CDs were chosen to bind with CA/EC.
Individual epimeric catechin-cyclodextrin mixture was firstly
analyzed. MS spectra of 10 samples were shown in Supporting
Information (Fig. S1). It was found that all cyclodextrins could bind
with two catechins by 1:1, but the charge states of the complexes
were different. To determine the possibility of CA/EC separation by
IM, the IM spectra of CA/EC and their complexes with CDs were
extracted from MS peaks. The drift time shown in IM spectra in
Figure 2 indicated that there was nearly no resolving possibility for
negtive charged CA/EC. For complexes of catechins and β-CD (m/z
711.2285), HP-β-CD (m/z distributed from 800 to 950), DM-β-CD
(m/z distributed from 1500 to 1600), HP-γ-CD (m/z distributed from
900 to 1050), the results were similar. However, the IM spectra of
CA/EC-γ-CD complexes (m/z 792.3511) complexes showed an
inspiring result. The drift time of the two base peaks had a
difference of 0.61ms, which indicated the possibility of seperation.
and wave velocity (WV). The resolution of IM can be determined by
peak-to-peak resolution (R) as Equation (2) shown, where t1 and t2
represent the drift time of two peaks and w1 and w2 represent their
baseline width.
R=2(t2-t1)/(w1+w2)
(2)
Figure 3 shows a comparison of IM spectra of CA-γ-CD,
EC-γ-CD and their mixtures at the same IM condition. Obviously,
the two peaks of mixture corresponded to the two base peaks of
individual complex. Although the two weak peaks were overlapped
by the two strong ones because of similar drift time, qualitatively
judging that the two catechins were successfully separated was not
influenced.
The mixture was tested at six IM conditions to find a best one
of separation effective. Resolution of each spectrum was calculated
and listed in Table 1. The six IM spectra were shown in Supporting
Information, Figure S2. The two analytes could be separated at all
six conditions. Although the resolution differed in a small range, the
best condition was determined as WH 28V and WV 450m/s.
CCS measurements
Polyalanine was measured at eight IM conditions to build
calibration curves. As the charge state of components in mixture
were different (one negative charge for catechins and two for γ-CD
and CA/EC-γ-CD complexes), two calibration curve clusters of these
charge states were individually built. The calibration curves were
shown in Supporting Information Figure S3, S4 and Table S1, S2.
CA-γ-CD and EC-γ-CD samples were also measured at the same
IM conditions. In all cases, two peaks of complex ions were found.
Thus two CCS values were calcuted for both complexes. Table 2
shows the results of CCS measurement. Relative standard deviation
(RSD) of each group was <1%, illustrating a high precision of
measurement. There was only a disparity of 0.32Å2 between CA
and EC. It was an apparent result because of the structural
similarity and symmetry of buffer gas molecule. However, the
disparity came to 11.75 Å2 when comparing the main peaks of two
complexes. Each complex itself was divided into two groups by CCS
disparity of 12.33Å2 for CA-γ-CD and 13.62Å2 for EC-γ-CD. Table 3
shows the relative difference (RD) between two complexes as well
as two catechins. Such distribution of CCS indicated that
catechin-γ-CD complexes may adopt a variety of conformations.
Interestingly, the CCS only increased about 15~30 Å2 from γ-CD to
the complexes, which was extremely different from the CCS of
catechins. As reported,[27] CDs with sufficiently large hydrophobic
cavity mostly bind with small molecules taking the form of
“inclusion”, which means that the hydrophobic part of small
molecule enters into the cavity while the polar groups bind with the
hydroxyl groups at CDs’ cavity rims. Thus, the CCS measurement in
this work suggested that catechins may be highly included into the
cavity of γ-CD while forming complexes. The interaction between
the polar groups may cause shrinkage of the hydrophobic cavity.
Thus,
further
experiments
were
all
based
on
catechin-γ-cyclodextrin complexes.
Interestingly, there appeared two additional weak peaks in both
spectra of CA/EC-γ-CD complexes, and the drift time of one’s peak
was just nearly the same as the other’s base peak. This
phenomenon was not found in any other complexes or CA and EC.
A study about the structure of CA/EC-β-CD pointed out that
inclusion ways of CA/EC and β-CD were completely different.[26]
Considering that different cyclodextrins contains different
hydrophobic cavity area and polar groups, we speculated that
CA/EC-γ-CD complexes may also take different binding ways which
could cause CCS diversity. In addition, a larger hydrophobic cavity in
γ-CD may provide more freedom and flexibility than β-CD while
including catechins, leading to CCS difference of one single complex.
Therefore, a computing simulation was done to search the possible
conformations of catechin-γ-CD complexes. The results are shown
and discussed in later section.
Catechin separation in mixture and optimizing IM condition
A mixture of CA, EC and γ-CD was prepared at a concentration
ratio of 10μM:10μM:50μM to optimize the IM condition and
evaluate the resolution. As ralated work reported,[22] the separation
efficiency of T-wave IM was mostly decided by wave height (WH)
Conformational simulations and theoretical CCS calculations
Computational modeling was performed to generate the
theoretical structures of catechins-γ-CD complexes to explain the
separation of two complexes and heterogeneity of their
conformation. Docking and MC/MD simulation provided about 500
conformations totally. 31 of them were preliminary selected
according to energy, including 14 for CA-γ-CD and 17 for EC-γ-CD.
Two general kinds of binding ways were found from docking results
for both complexes. One was that the A-ring of catechins was
included into the cavity and the other one was that the B-ring was
included. Therefore, the selected conformations were further
divided into four groups for CCS/energy calculation and comparison.
Calculated CCS and energy of selected structures were listed in
Supporting Information Table S3. An energy-CCS map of the
conformations was ploted in Figure 4 to show their distribution.
The conformations at minimum energy from four groups were
This article is protected by copyright. All rights reserved.

shown in Figure 5.
similar energy.[21] In combination with the relative intensity of
measured peaks, it is valid to believe that the stronger peaks
corresponded to the “A-ring included” comformation and the other
peak corresponded to the “B-ring included” comformation for both
complexes.
Table 4 showed comparison of theoretical CCS (Ωcal) and
experimental CCS (Ωmes). A difference of about 5~10 Å2 between
theoretical and experimental CCS values of complexes was found
while the gap decreased a lot when it came to catechins and γ-CD.
Considering the calculating theory,[28] this slightly difference was
acceptable. The distribution width of theoretical CCS was nearly the
same as measured (about 20 Å2) and the theoretical CCS gap
between two complexes was 9.3 Å2, which verified the reliability of
conformation searching results.
According to the energy results, the most stable conformation
was the “A-ring included” conformation for both complexes (Figure
5 A and C). In this situation, the A-ring was included while B-ring
“lies” on the rim of γ-CD because of the possible H-bond interaction.
However, when it came to the situation of “B-ring included” (Figure
5 B and D), the B-ring was included deeper while the A-ring “stands”
outside the cavity. The energy of “B-ring included” conformations
was just slightly higher than the other one. Specifically, the energy
gap between lowest-energy “A-ring included” conformation and
“B-ring included” conformation was 20.715kJ/mol for CA-γ-CD and
23.905kJ/mol for EC-γ-CD. Such a gap allows the four lowest-energy
conformation to be formed and detected by IM. Other studies also
provided envidences for discovery of multiple conformations with
Quantitative analysis
calibration curves. Table 5 shows measured concentrations of
catechins and the relative error (RE) between calculated
concentrations and actual concentrations. All the measured
concentrations of seven validation groups were at a believable level
(RE<15%). Such a result proved that IM-MS was comparable to
HPLC-MS in the ability of quantifying concentrations of compounds
in complex mixture.
IM-MS have been proved as effective methods for quantitative
analysis.[22, 29] The way using IM to quantify compounds in complex
samples is similar as HPLC. In the present work, catechins-γ-CD
complexes were considered as substance to build standard curves
to indirectly quantify catechins. However, there remained
a
problem that the two complexes were not completely separated.
The overlapped peaks may influence the quantitative accuracy.
Thus the peak intensity provided by the overlapped weak peaks
should be removed while quantifying the complexes in mixture. In
particular, mixtures of individual catechin at different concentration
and γ-CD at fixed concentration were firstly measured to build
calibration curves as Equation set (3) shows.
Quantification of catechins in catechu
Components are firstly ionized and then separated in IM-MS
measurement. Such a characteristic brings a challenge quantifying
target compounds in complex system due to the limited ionization
efficiency. In this study, we made a preliminary attempt to quantify
catechins in extracts of a kind of herb, catechu, using the method
mentioned in previous section.
yCA=a1xCA+b1
yEC=a2xEC+b2
(3)
A mixture of diluted extractive solution with γ-CD was firstly
tested to make sure whether catechins had been extracted and
binded with γ-CD. The results showed that catechins in catechu
extracts formed complexes with γ-CD and the complexes could also
be detected by IM-MS dispite possible effects of matrix. Then,
quantification was performed according to the procedure
introduced before. The calibration curves and calculation results of
IM peak ratio were shown in Supporting Information Figure S6 and
Table S5. Measurement and calculation results of extracts were
shown in Table 6. It was found that (+)-catechin was the main
component in the extracts, which was about 150mg/g or 15%,
while (-)-epicatechin was only about 4%. Although as introduced
before, γ-CD could form complexes with variety kinds of
compounds and the detection of complexes may be hard because
of the matrix effects, catechins in extracts could still be mesured
and quantified. Considering the binding effect, γ-CD was at a high
concentration in all samples. But the singal intensity of γ-CD was
not at high level due to the strong ionization tendency of catechins.
This made the quantification less likely interpreted by γ-CD or other
compounds in the complex system。
Then the proportion (k) of the stronger peak to total peak area
(y’/y) was calculated as Equation set (4) shows.
yCA’=kyCA
yEC’= kyEC
(4)
The peak intensity of mixture of two complexes (A) was seen as
a superposition of stronger peak (y’) of one complex and weak
peak of the other complex (y-y’) as Equation set (5) shows.
A1=yEC’+yCA-yCA
A2=yEC-yEC’+yCA
The quantitative process is shown in Figure
’
’
(5)
and the
6
concentration of catechins (x) can be calculated by solving the
equation set (3), (4) and (5). The solution of was shown in equation
set (6).
(
)
A1−kECA2
xCA = ( 1−k
− b1)/a1
EC
1−kCA−kEC
(
)
A2−kCAA1
xEC = ( 1−k
− b2)/a2
(6)
CA
1−kCA−kEC
The measurements were performed under optimized
conditions mentioned above. Calibration curves were shown in
Supporting Information Figure S5. Both curves possessed good
linearity (R2>0.99). The peak area of all stronger peaks was
extracted to calculate the proportion (k). Results were listed in
Supporting Information Table S4. Mixtures of complexes at
different CA/EC ratios were then measured to verify the accuracy of
The same extracts were also measured by HPLC. Experimental
details were shown in Supporting Information. The results shown in
Table 6 gave similar weight percentages as the method in this work
measured, indicating a credible accuracy and higher time efficiency
(1 minute for one sample while more than 10 minutes using HPLC)
of current method compared with HPLC.
This article is protected by copyright. All rights reserved.

Sample preparation
To prepare stock solutions, standard (+)-catechin and
(-)-epicatechin were dissolved in methonal, while each kind of
cyclodextrin was dissolved in water. The concentration of all stock
solutions was 1mg/ml. For pre-seperation of two catechins,
eatablishing quantitative standard curve and CCS measurment,
(+)-catechin and (-)-epicatechin were individually mixed with each
cyclodextrin at a concentration ratio of 10μM:20μM. All samples
were prepared in ammonium acetate buffer (500μM) in
methonal/water (50/50, v/v).
For quantification of catechins in complex system, 1g of
catechu powder was mixed with 10mL of ethanol/water (70/30,
v/v) in a 50mL centrifuge tube. Then, the mixture was treated by
ultrasound for 1h at room temperature. The extractive solution
was obtained by centrifugation and stocked in -20℃for further use.
While preparing samples for quantification, the solution was
diluted to a series of concentration and mixed with γ-CD (50μM)
and ammonium acetate buffer (500μM) in methonal/water
(50/50, v/v).
Conclusions
In this study, we attempted to separate (+)-catechin and
(-)-epicatechin by using IM-MS. The negative charged epimers as
well as complexes of catechins and β-CD, HP-β-CD, DM-β-CD or
HP-γ-CD showed almost no separation in IM spectra. But when it
came to the complexes of catechins and γ-CD, a remarkable
improvement of separation was found. The catechins-γ-CD
complexes could also be separated from mixtures with drift time
difference >0.5ms and the peak-to-peak resolution up to 0.86
after condition optimization. Two individual peaks of both
complexes were detected, indicating a diversity of conformations
of complexes. Further efforts provided the experimental CCS of
the compounds in complex mixtures. The CCS difference between
CA and EC was only 0.32Å2 while it increased to 11.75 Å2 between
the two complexes. The slight increase of CCS from γ-CD to
catechin-γ-CD complexes sugggested that the epimers may highly
included into the hydrophobic cavity of γ-CD while forming the
non-covalent complexes.
Instrument parameters
Computational modeling and calculation was performed to
provide theoretical support to the experimental results. 31
models were selected from a large number of modeling results
according to energy and CCS of these models was calculated. Two
binding ways of both complexes were found and the
lowest-energy conformations of each way shared small energy
gaps, indicating the coexistence possibility of these conformations.
CCS calculation revealed that the two separated peaks in IM may
correspond to the “A-ring” included conformations of two
complexes because of their lower energy and theoretical CCS
difference. The two additional peaks of both complexes may
correspond to the “B-ring” included conformations with higher
energy.
Quantitative analysis by IM-MS proved it an effective method
to quantify catechins according to the contents of separated
complexes indirectly. Although IM peaks may overlap at a certain
degree because of the conformational diversity, mathematical
optimization for concentration calculation could reduce the
influence from overlapped peaks to ensure the accuracy of
quantitative analysis. The method was also intoduced to quantify
catechins in complex system and showed similar accuracy and
higher analysis speed comparaed with traditional method such as
HPLC.
The results obtained from the present study promised a
hopeful future of rapid separation and quantification of small
molecule isomers. The combination of IM-MS measurement and
theoretical calculation provided a clear sight of conformation
difference of isomers and their complexes with chiral selectors.
This way can not only be applied in isomer separation, but also in
other field such as protein-ligand interaction and inhibitor
screening. Our further experiment will aim at mechanisms of
interaction between flavonoids and proteins based on these
separation and theoretical methods.
All IM-MS measurements were conducted on a quardupole
ion-mobility time-of-flight (Q-IM-Tof) mass spectrometer (Synapt
G2-S HDMS, Waters) equipped with an electrospray ionization (ESI)
source. The IM instrument had an intrinsic resolution of >40
(Ω/ΔΩ). All the samples were directly infused by a syringe pump at
a flow rate of 10μl/min and detected on negative ion mode with a
capillary voltage at 2.1kV. The cone voltage was set at 40V. The
flow of cone gas was 50L/h and desolvation gas was 450L/h,
respectively. The source temperature was 80℃ and the
desolvation temperature was 350℃. As for IM gas parameters, the
helium cell gas flow was set at 200mL/min and IM gas flow was
set at 120mL/min.
CCS mesurement
As described everywhere,[30] since a T-wave IMS applies
constantly changing electric field in its mobility cell to transport
ions through the buffer gas, the CCS of analytes can’t be caculated
directly according to drift time. It must be determined by
calibration of standard materials with appropriate molecular
weight range. In this work, polyalanine was employed to build CCS
calibration. The form of calibration should be Equation (1). CCS (Ω)
of polyalanine at different polymerization degrees and charge
statements was known.[31] The calibration could be built by
measuring the drift time (tD) of polyalanine.
B
Ω=AtD
(1)
Stock solution of polyalanine (1mg/ml) was diluted to
0.1mg/ml in water/acetonitrile/acetic acid (50/50/1) and acquired
at eight different conditions of IM. The conditons of ESI source
were the same as those of acquiring analytes.
Computational methods
Autodock (version 4.2) was applied to build initial models of
catechin-cyclodextin complex for further simulation.[32]
Semiflexible docking method was used with genetic algorithm to
search complex models. The grid box was set to 0.3 and contained
the whole cyclodextrin. In every simulation, 10 results were
output and scored by a built-in function of the software.
According to the scores and structures, appropriate models were
Experimental
Materials
chosen to peform
simulation.
a second dock or molecular dynamics
β-cyclodextrin
(β-CD),
γ-cyclodextrin
(γ-CD),
(HP-β-CD),
(DM-β-CD),
(2-Hydroxypropyl)-β-cyclodextrin
2,6-Di-O-methyl-β-cyclodextrin
The molecular dynamics simulation and geometry
optimizations were performed by Ascalaph Designer with
MdynaMix software package.[33] The output structures from
Autodock were treated by a short optimization to remove atoms
clashes and then simulated by Monte Carlo (MC) method mixed
with molecular dynamics (MD). The simulation temperature was
300K, and the MD time was set to be 500ps with 2.0fs of time
step. Structures of last 100ps were extracted from every 5ps to be
optimized by Hybrid LS and CD method. Molecular mechanics
method was used to calculate energy. The OPLS force field was
(2-Hydroxypropyl)-γ-cyclodextrin (HP-γ-CD) were purchased from
Aladdin (China). (+)-Catechin, (-)-epicatechin, ammonium acetate
were obtained from Sigma-Aldrich (USA). Catechu was obtained
from Tongrentang Changchun Pharmacy (China). Methonal,
acetonitrile and acetic acid were supplied by TEDIA Company
(USA). The ultrapure water for all experiments was purified using
a Milli-Q Water System (USA).
This article is protected by copyright. All rights reserved.

chosen for all simulation and energy calculation.
[16] K. Pagel; D. J. Harvey. Ion mobility–mass spectrometry of complex
carbohydrates: collision cross sections of sodiated N-linked glycans. Anal.
Chem. 2013, 85, 5138–5145.
[17] M. Kliman; J. C. May; J. A. McLean. Lipid analysis and lipidomics by
structurally selective ion mobility-mass spectrometry. BBA Mol. Cell Biol. L.
2011, 1811, 935–945.
[18] Zhuang, X.; Zhao, B.; Liu, S.; Song, F.; Cui, F.; Liu, Z.; Li, Y. oncovalent
Interactions between Superoxide Dismutase and Flavonoids Studied by
Native Mass Spectrometry Combined with Molecular Simulations. Anal.
Chem. 2016, 88(23), 11720-11726.
[19] R. Beveridge; L. Migas; R. Kriwacki; Perdita, E. Barran. Ion mobility mass
spectrometry measures the conformational landscape of p27 and its
domains and how this is modulated upon interaction with Cdk2/cyclin A.
Angew. Chem. 2018, Doi:10.1002/ange.201812697.
Reasonable simulation results were then chosen to calculate
theoretical CCS by Collidoscope, an open source program
developed by Prell et al.[34] This program uses trajectory method
to calculate CCS. For all CCS computation, the net charge was set
to 0 and spherical nitrogen model was chosen as simulation gas
phase.
Supporting Information
The supporting information for this article is available on the
WWW under https://doi.org/10.1002/cjoc.2018xxxxx.
Acknowledgement
[20] Mason E. A.; Schamp H. W. Ann. Phys. 1958, 4, 233-270.
[21] Christopher, D. Chouinard; Vinícius, Wilian, D. Cruzeiro; Christopher, R.
Beekman; Adrian, E. Roitberg; Richard A. Yost. Investigating Differences in
Gas-Phase Conformations of 25-Hydroxyvitamin D3 Sodiated Epimers
using Ion Mobility-Mass Spectrometry and Theoretical Modeling. J. Am.
Soc. Mass Spectrom. 2017, 28, 1497-1505.
[22] Anna Trod; Magdalena Zimnicka; Witold Danikiewicz. Separation of
catechin epimers by complexation using ion mobility mass spectrometry. J.
Mass Spectrom. 2015, 50, 542–548.
This work was funded by the National Natural Science
Foundation of China (No.21673219 and 81873193)
References
[1] Yanagida, A.; Shoji, A.; Shibhbusawa, Y.; Shindo, H,; Tagashira, M.; Ikeda,
M.; Ito, Y. Analytical separation of tea catechins and food-related
polypeenols by high-speed counter-current chromatography. J.
Chromatogr. A. 2006, 1112, 195-201.
[2] Kwik-Uribe, C.; Bektash, RM. Cocoa flavanols-measurement,
bioavailability and bioactivity. Asia Pac. J. Clin. Nutr. 2008, 17, 280–3.
[3] Cheng Guiwen, W.; Crisosto, Carlos, H. Browning Potential, Phenolic
Composition, and Polyphenoloxidase Activity of Buffer Extracts of Peach
and Nectarine Skin Tissue. J. Amer. Soc. Hort. Sci. 1995, 120, 835–838.
[4] P.M. Aron; J.A. Kennedy. Flavan-3-ols: nature, occurrence and biological
activity. Mol. Nutr. Food Res. 2008, 52, 79-104.
[5] Katiyar, S.; Elmets, C. A.; Katiyar, S. K. Green tea and skin cancer:
Photoimmunology, angiogenesis and DNA repair. J. Nutr. Biochem. 2007,
18, 287.
[23] József Szejtli. Introduction and General Overview of Cyclodextrin
Chemistry. Chem. Rev. 1998, 98 (5), 1743–1754.
[24] P. Dwivedi; C. Wu; L. M. Matz; B. H. Clowers; W. F. Siems; H. H. Hill.
Gasphase chiral separations by ion mobility spectrometry. Anal. Chem.
2006, 78, 8200–8206.
[25] Fiori J.; Pasquini B.; Caprini C.; Orlandini S.; Furlanetto S.; Gotti R. Chiral
analysis of theanine and catechin in characterization of green tea by
cyclodextrin-modified micellar electrokinetic chromatography and high
performance liquid chromatography. J. Chromatogr. A. 2018, 1562,
115-122.
[6] Yanez J. A.; Remsberg C. M.; Miranda N. D.; Andrews P. K.; Vega-Villa K. R.;
Davies N. M. Pharmacokinetics of selected chiral flavonoids: hesperetin,
naringenin and eriodictyol in rats and their content in fruit juices.
Biopharm. Drug Dispos. 2008, 29, 63–82.
[7] Sano M.; Suzuki M.; Miyase T.; Yoshino K.; Maeda-Yamamoto M. Novel
antiallergic catechin derivatives isolated from oolong tea. J. Agric. Food
Chem. 1999, 47, 1906-10.
[8] S. Baba, N.; Osakabe, M. Natsume, Y. Muto. In vivo comparison of the
bioavailability of (+)-catechin, (−)-epicatechin and their mixture in orally
administered rats. The Journal of Nutrition. 2001, 131, 2885–2891.
[9] Florian, Labrouche; Andrew, C. Clark; Paul, D. Prenzler. Isomeric Influence
on the Oxidative Coloration of Phenolic Compounds in a Model White
Wine:ꢀ Comparison of (+)-Catechin and (−)-Epicatechin. J. Agric. Food
Chem. 2005, 53, 9993–9998.
[10] H. M. Merken; G. R. Beecher. Measurement of food flavonoids by
highperformance liquid chromatography: a review. J. Agric. Food Chem.
2000, 48, 577–599.
[11] D. Tsimogiannis; M. Samiotaki; G. Panayotou; V. Oreopoulou.
Characterization of flavonoid subgroups and hydroxy substitution by
HPLC-MS/MS. Molecules. 2007, 12, 593–606.
[12] Kunbo, W.; Zhonghua, L.; Jian-an, H.; Donghe, F.; Fang, L.; Yushun, G.;
Xiaosong, W. TLC separation of catechins and theaflavins on polyamide
plates. Journal of Planar Chromatography. 2009, 22(2), 97-100.
[13] EG, Yanes; SR, Gratz; AM, Stalcup. Tetraethylammonium
tetrafluoroborate: a novel electrolyte with a unique role in the capillary
electrophoretic separation of polyphenols found in grape seed extracts.
Analyst. 2000, 125(11), 1919-1923.
[26] Khedkar, J. K.; Gobre, V. V.; Pinjari, R. V.; Gejji, S. P. Electronic Structure
and Normal Vibrations in (+)-Catechin and (-)-Epicatechin encapsulated
β-cyclodextrin. J. Phys. Chem. A. 2010, 114, 7725.
[27] Dupuya, N.; Barbryb, D.; Briac, M.; Marquisd, S.; Vrielyncka, L.; Kistera, J.
1H NMR study of inclusion compounds of phenylurea derivatives in
β-cyclodextrin. J. Spectrochimica Acta Part A. 2005, 61, 1051.
[28] Iain Campuzano; Matthew, F. Bush; Carol, V. Robinson; Claire Beaumont;
Keith Richardson; Hyungjun Kim; Hugh I. Kim. Structural Characterization
of Drug-like Compounds by Ion Mobility Mass Spectrometry: Comparison
of Theoretical and Experimentally Derived Nitrogen Collision Cross
Sections. Anal. Chem. 2011, 84, 1026-1033.
[29] Rui Wang; Wen Gao; Bin Li; Chang-Jiang-Sheng Lai; Shan Lin; Ping Li; Hua
Yang.
A strategy for absolute quantitation of isomers using high
performance liquid chromatography-ion mobility mass spectrometry and
material balance principle. J. Chromatogr. A. 2018, 1571, 140-146.
[30] Xueqin, P.; Chenxi, J.; Zhengwei, C.; Lingjun, L. Structural Characterization
of Monomers and Oligomers of D-Amino Acid-Containing Peptides Using
T-Wave Ion Mobility Mass Spectrometry. J. Am. Soc. Mass Spectrom. 2016,
28(1), 110-118.
[31] Bush, M. F.; Campuzano, I. D. G.; Robinson, C. V. Ion Mobility Mass
Spectrometry of Peptide Ions: Effects of Drift Gas and Calibration
Strategies. Anal. Chem. 2012, 84, 7124-7130.
[32] G. M. Morris; R. Huey; W. Lindstrom; M. F. Sanner; R. K. Belew; D. S.
Goodsell; A. J. Olson. AutoDock4 and AutoDockTools4: Automated
docking with selective receptor flexibility. J. Comput. Chem. 2009, 30 (16),
2785-91.
[33] A.P. Lyubartsev; A. Laaksonen. MDynaMix - A scalable portable parallel
MD simulation package for arbitrary molecular mixtures. Computer
Physics Communications. 2000, 128, 565-589.
[14] A. B. Kanu; P. Dwivedi; M. Tam; L. Matz; H. H. Hill. Ion mobility–mass
spectrometry. J. Mass Spectrom. 2008, 43, 1–22.
[15] D. Balbeur, J.; Widart, B.; Leyh, L.; Cravello, E.; De Pauw. Detection of
oligonucleotide gas-phase conformers: H/D exchange and ion mobility as
complementary techniques. J. Am. Soc. Mass Spectrom. 2008, 19,
938–946.
[34] Ewing, S.A., Donor, M.T., Wilson, J.W., Prell J.S. Collidoscope: An
Improved Tool for Computing Collisional Cross-Sections with the
Trajectory Method J. Am. Soc. Mass Spectrom. 2017, 28, 587-596.
This article is protected by copyright. All rights reserved.

Accepted manuscript online: XXXX, 2019
Version of record online: XXXX, 2019
(The following will be filled in by the editorial staff)
Manuscript received: XXXX, 2019
Manuscript revised: XXXX, 2019
Manuscript accepted: XXXX, 2019
Figure legends
Figure 1 Suctures of catechins and three typical cyclodextrins.
Figure 2 IM spectra of CA/EC and their 1:1 noncovalent complexes with five kinds of CDs.
Figure 3 Overlapped IM spectra of CA-γ-CD, EC-γ-CD and their mixture. The intensity was normalized by the strongest peak of all spectra.
Figure 4 Energy-CCS distribution of the selected conformations.
Figure 5 Side view (upper) and top view (below) of theoretical conformations of catechins-γ-CDcomplexes at lowest energy states including: A)
A-ring of CA included (-111.686kJ/mol, Ωcal=405.43Å2), B) B-ring of CA included (-90.971kJ/mol, Ωcal=392.54Å2), C) A-ring of EC included
(-112.212kJ/mol, Ωcal=396.13Å2), D) B-ring of EC included (-88.738kJ/mol, Ωcal=405.10Å2). Possible H-bonds were also labeled.
Figure 6 Quantification process of CA and EC by catechins-γ-CD complexes in mixture.
Table 1 Resolution of IM spectra of CA/EC-γ-CD mixture at different wave
height (WH) and wave velocity (WV)
WH(V)/WV(m/s)
t1(ms)
t2(ms)
w1+w2(ms)
R
28/400
28/450
28/500
20/270
25/400
27/400
6.06
7.27
8.05
7.94
7.83
6.62
6.61
7.81
8.71
8.71
8.60
7.28
1.44
1.25
1.98
2.09
2.09
1.54
0.76
0.86
0.67
0.74
0.74
0.85
Table 2 Measured CCS of catechins, γ-CD and their non-covalent complexes under eight conditions. Ωstr and Ωweak mean CCS calculated from the drift time of
stronger peak and weaker peak of complex ions. Ωmes represents average CCS of each group.
ΩCA-γ-CD(Å2)
ΩEC-γ-CD(Å2)
WH(V)/WV(m/s)
ΩCA(Å2)
ΩEC(Å2)
Ωγ-CD(Å2)
Ωstr
Ωweak
Ωstr
Ωweak
30/400
35/400
28/450
30/450
35/450
40/450
35/500
399.77
400.28
400.03
401.78
394.78
392.06
399.52
388.16
387.66
386.47
389.24
381.97
380.82
386.71
386.60
390.41
388.20
387.78
386.30
383.93
384.64
401.26
402.48
401.68
401.78
398.93
397.52
399.52
153.47
153.00
154.80
154.49
152.99
152.18
153.88
153.47
154.31
154.80
154.49
153.84
153.75
152.66
372.48
372.55
373.61
371.64
366.51
369.65
371.20
This article is protected by copyright. All rights reserved.

40/500
Ωmes(Å2)
RSD
397.97
398.27
0.82%
386.52
385.94
0.77%
384.34
386.52
0.58%
397.97
400.14
0.47%
154.09
153.61
0.57%
154.09
153.93
0.43%
372.30
371.24
0.60%
Table 3 Comparison of measured CCS between CA and EC as well as their complexes. For CA-γ-CD and EC-γ-CD, the relative difference of CCS calculated from
strong and weak peak was given out. The relative difference of two complexes were calculated from their strong peaks.
CA-γ-CD
EC-γ-CD
CA-γ-CD/EC-γ-CD
CA/EC
0.21%
RD
3.19%
3.52%
3.04%
Table 4 Theoretical CCS (Ωcal) and experimental CCS (Ωmes). Difference between two values were also shown (ΔΩ=Ωcal-Ωmes).
CA-γ-CD
EC-γ-CD
CA
EC
γ-CD
A-ring
B-ring
A-ring
B-ring
Ωmes(Å2)
Ωcal(Å2)
ΔΩ(Å2)
398.27
405.43
7.16
385.94
392.54
6.60
386.52
396.13
9.61
400.14
405.10
4.96
153.61
152.81
-0.80
153.93
150.50
-3.43
371.24
373.47
1.23
Table 5 Results of complex mixtures measurements. A1 and A2 are area of two peaks. Measured concentrations (xCA and xEC) were calculated from Equation set
(6). Relative errors (RE) were calculated by xCA/cCA and xEC/cEC.
cCA(μM)
cEC(μM)
A1
A2
xCA(μM)
xEC(μM)
RECA
REEC
10
10
10
10
20
30
40
10
20
30
40
10
10
10
1.93E+05
3.79E+05
5.66E+05
7.81E+05
2.43E+05
2.71E+05
2.92E+05
1.82E+05
1.99E+05
2.25E+05
2.41E+05
3.65E+05
5.54E+05
7.66E+05
9.22
9.34
8.84
17.62
26.39
36.55
10.05
10.16
9.79
-7.85%
-6.56%
-0.78%
-1.12%
-8.83%
-7.86%
-4.41%
-11.55%
-11.90%
-12.05%
-8.62%
0.46%
9.92
9.89
18.23
27.64
38.23
1.58%
-2.14%
Table 6 Results of extracts measurements (Crude drug concentration is 10 μg/mL)
acquired by MS and HPLC methods.
Methods
ωCA
ωEC
MS
15.50%
15.71%
4.64%
4.62%
HPLC
This article is protected by copyright. All rights reserved.

Entry for the Table of Contents
Page No.
Separation，quantification and structural study
of (+)-catechin and (-)-epicatechin by ion
mobility mass spectrometry combined with
theoretical algorithms
Xinyu Bian, Bing Zhao, Bo Pang, Zhong Zheng,
Shu Liu,* Zhiqiang Liu and Fengrui Song*
Quantitative process of CA and EC by catechins-γ-CD complexes in mixture.
This article is protected by copyright. All rights reserved.

Fig 1
Fig 2
Fig 3
This article is protected by copyright. All rights reserved.

Fig 4
This article is protected by copyright. All rights reserved.

Fig 5
Fig 6
This article is protected by copyright. All rights reserved.

This article is protected by copyright. All rights reserved.