The use of diffusion-ordered spectroscopy and complexation agents to analyze mixtures of catechins

Article abstract of DOI:10.1039/b900164f

Mixtures of catechins, (+)-catechin (C), (-)-epicatechin (EC), (-)-epigallocatechin (EGC) and (-)-epigallocatechin gallate (EGCG) have been analyzed by diffusion-ordered spectroscopy (DOSY) using liquid and high-resolution magic angle spinning (HR-MAS) NM

Full text of DOI:10.1039/b900164f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
The use of diffusion-ordered spectroscopy and complexation agents
to analyze mixtures of catechinsw
a
b
b
Jun Xu,ab Tianwei Tan, Lennart Kenne and Corine Sandstrom*
¨
Received (in Montpellier, France) 7th January 2009, Accepted 19th January 2009
First published as an Advance Article on the web 20th February 2009
DOI: 10.1039/b900164f
Mixtures of catechins, (+)-catechin (C), (ꢀ)-epicatechin (EC), (ꢀ)-epigallocatechin (EGC)
and (ꢀ)-epigallocatechin gallate (EGCG) have been analyzed by diffusion-ordered spectroscopy
(DOSY) using liquid and high-resolution magic angle spinning (HR-MAS) NMR probes.
Beta-cyclodextrin (b-CD) and bovin serum albumin (BSA), often used as ligands in affinity
chromatography, were added to the mixture of catechins to mimic chromatographic conditions
and modify the average mobility. The influence of the solvent, water, dimethyl sulfoxide,
methanol, acetone and acetonitrile, was also investigated. The best separation of the components
was achieved with b-CD in the liquid probe using a 15% CD3CN–85% D2O solution, and this
was applied to the analysis of catechins from green tea extract. The parts of the catechin
molecules having the closest contact to the BSA protein were also determined by saturation
transfer difference (STD) NMR experiments.
high-speed counter-current chromatography (HSCCC)14,15
Introduction
has also been reported.
The preventive activity of tea against cancer is a topic of
intensive investigation. The active components and the mole-
cular mechanisms responsible are not clearly understood but
most mechanistic studies have focused on green tea catechins.1–5
The six major catechins displaying biological activity, are
(–)-epigallocatechin-3-gallate (EGCG), (ꢀ)-gallocatechingallate
(GCG), (–)-epigallocatechin (EGC), (–)-epicatechin-3-gallate (ECG),
(–)-epicatechin (EC) and (+)-catechin (C). Of these, EGCG is
the most abundant and most active catechin. It is a strong
antioxidant and the polyphenolic structure also allows binding
to many proteins such as plasma proteins, fibronectin and the
CD4 receptor at the attachment site of the glycoprotein gp120
present at the surface of the HIV virus.6,7 Thus, because of the
possible use of catechins as chemotherapeutic agents, there is a
strong need of developing highly sensitive and selective
methods for their isolation and characterization.
Diffusion-ordered spectroscopy (DOSY) is an NMR tech-
nique used to separate signals from different species according
to their diffusion rates and has sometimes been described as
‘‘in tube chromatography’’ or ‘‘NMR chromatography’’. It
has been used for identification of metabolites, characteriza-
tion of aggregates and hydrogen bonding and for affinity
NMR.16–23 The latter technique is based on the fact that
ligands bound to a receptor will have reduced translational
diffusion rates allowing their signals to separate from those of
non-binding molecules. It has also been shown that using high-
resolution magic-angle spinning (HR-MAS) NMR spectro-
scopy, the separation properties of DOSY can be enhanced
upon addition of a solid chromatographic phase such as silica
or C18 gels used in HPLC.24–27 The method mimics a part of
the chromatographic process in that the molecular average
mobility in the mixture is modified according to the individual
affinities for the stationary phase. This technique has been
applied to the separation of benzene, naphthalene and
anthracene26 as well as to the separation of acetone, butanone
and pentanone.25
The identification of molecules in a mixture usually requires
a separation step and a structural characterization step often
achieved by chromatographic techniques and by NMR or
mass spectrometry. The chromatographic procedures
frequently used for the separation of catechins are based on
derivatized silica adsorbents and good separation is achieved
with reversed-phase liquid chromatography using C18
columns.8–11 The use of soft chromatographic media, such as
Sephadex LH-2012,13 and partition chromatography such as
We have recently investigated the complexation between
EGCG and different cyclodextrins (CDs),28 as well as the
adsorption mechanism of EGCG on CD-substituted agarose29
and on the silica gel medium CYCLOBOND.28 The inter-
action between EGCG and b- or g-CD bonded to silica beads
1
was studied by H HR-MAS NMR spectroscopy and it was
a College of Life Science and Technology, Beijing University of
Chemical Technology, Beijing, 100029, China
demonstrated that the chromatographic retardation of EGCG
was due to interaction of EGCG with the CDs. Non-specific
interactions with the silica gel were not observed.28 The NMR
data obtained from hydroxy protons suggested that for
the complex, intermolecular hydrogen bonding in addition
to hydrophobic interaction stabilized the b-CD–EGCG
complex.28
b Department of Chemistry, Swedish University of Agricultural
Sciences, P.O. Box 7015, Uppsala, SE-750 07, Sweden
w Electronic supplementary information (ESI) available: 1H NMR
chemical shifts and complexation induced shifts for CH protons of
catechins and b-CD. Inter-residual NOEs between b-CD and
catechins. DOSY spectra of green tea extracts in water and methanol
and at different concentrations. See DOI: 10.1039/b900164f
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 1057–1063 | 1057

View Article Online
Saturation transfer difference (STD) NMR
The NMR samples were prepared in 50 mM sodium phos-
phate buffer (pH 6.8) and 99.96% D2O. The concentration of
catechin and BSA was 5 mM and 0.05 mM, respectively,
unless otherwise stated. The chemical shifts for 1H NMR
signals were referenced by setting the residual HDO signal to
dH = 4.70 ppm at 25 1C. Water suppression was achieved by
the WATERGATE pulse sequence.42
The STD NMR spectra were acquired using a series of 40
equally spaced 50 ms Gaussian-shaped pulses for selective
saturation, with 1 ms delay between pulses. The on-resonance
saturation frequency of the protein was set at d ꢀ3 ppm. The
off-resonance saturation frequency was set at d 30 ppm.
Subtraction of FIDs with on- and off-resonance saturation
was achieved by phase cycling. Investigation of the time
dependence of the saturation transfer with saturation times
from 0.5 to 5 s showed that 2 s was sufficient for efficient
transfer of saturation from the protein to the catechins.
2048 scans were collected. Relative STD values were calcu-
lated by dividing the STD signal intensities by the intensities of
the corresponding signals in the one-dimensional 1H NMR
reference spectrum of the same sample similarly processed.
Optimization of the experimental conditions was achieved
using samples without BSA. Several on-resonance irradiation
frequencies were tested (0, ꢀ1, ꢀ2, ꢀ3 and ꢀ4 ppm) and with
the irradiation frequency set at d ꢀ3 ppm, no signals were
present in the STD NMR spectrum, indicating that the effects
observed in the presence of the protein were due to true
saturation transfer.
Chart 1 The structure of (+)-catechin (1), (ꢀ)-epicatechin (2),
(ꢀ)-epigallocatechin (3) and (ꢀ)-epigallocatechin-3-gallate (4).
The absence of non-specific interactions between catechins and
the silica gel, together with the fact that different catechins bind to
b-CD with different affinities30 open the possibility of achieving
mixture analysis of catechins using DOSY but without solid-
phase chromatographic support. Indeed, DOSY has been used to
investigate the complexation of cyclodextrins with diverse host
molecules31–33 and recently with (+)-catechin34 but to our know-
ledge DOSY and b-CD have not been used to separate mixture of
catechins. Thus, in the present work, four of the major catechins
in green tea, C, EC, EGC and EGCG (Chart 1), were chosen to
study how HR-DOSY or HR-MAS-DOSY affinity experiments
can be used to analyze a mixture of similar compounds. This was
further applied on the catechin components from green tea
extract. Since the diffusion properties also can be manipulated
by changes in solvent just as changes in mobile phase are used to
control retention time in chromatography, different solvents were
investigated. The diffusion behavior of catechins in the presence
of bovine serum albumin (BSA), widely used as a chiral
recognition ligand35 and in affinity chromatography36 was also
studied. Fluorescence quenching,37,38 Fourier transform infrared,
circular dichroism,39 gel electrophoresis40 and isothermal titration
calorimetry41 experiments have demonstrated binding of some
catechins to BSA. The aim of this work was thus two-fold: first to
study how the diffusion properties of mixtures of catechins can be
modified upon addition of b-CD or BSA as complexation agents
in different solvents; secondly, to investigate if and how catechins
from green tea extract can be analyzed using this technique.
Catechin/b-CD interaction
A 1 : 1 molar ratio of catechins and b-CD in three different
solvents (D2O, DMSO-d6 and CD3CN) was used to study the
complex formation between catechins and b-CD, unless other-
wise stated. The concentration of each catechin was kept
below 7 mM to avoid self-association.43,44
DOSY experiments
Data acquisition and analysis were performed using the
Bruker TOPSPIN software (version 1.3). The DOSY experi-
ments were performed using the ledbpgp2s pulse sequence
from the Bruker library, with stimulated echo, longitudinal
eddy current compensation, bipolar gradient pulses, and two
spoil gradients using 16 different gradient values varying from
2 to 95% of the maximum gradient strength. Diffusion time
ranging from 50 to 250 ms and gradient length from 1 to 5 ms
were tested. 100 ms diffusion time was chosen for samples in
D2O and 200 ms for samples in DMSO-d6 and CD3CN. The
gradient length was set to 2.2 ms for all solvent systems.
Processing was achieved using 4096 points in the F2 dimension
and 256 points in F1. An exponential window function with
1 Hz line broadening was applied in the F2 dimension prior to
Fourier transformation. After baseline correction, the
diffusion dimension was processed with the DOSY processing
program (Bruker TopSpin software 2.0). A logarithmic scaling
was applied in the diffusion axis, and a noise sensitivity factor
of 4 and line width factor of 2 were used. The fitting of the
diffusion dimension in the 2D-DOSY spectra was obtained
Experimental
b-CD, BSA (purity 99%, fatty acid free and essentially
globulin free) and the catechins, C, EC, EGC and EGCG
were purchased from Sigma-Aldrich. Crude extracts of green
tea (GTE) were supplied by Shenglong Bioproduct Co.,
Guangxi, China.
The HR-NMR experiments were performed on a Bruker
1
400 MHz spectrometer using a 5-mm H/13C/15N/31P inverse
detection probe equipped with z-gradient, while the HR-MAS
NMR experiments were performed on a Bruker 600 MHz
1
spectrometer using a 4-mm HR-MAS SB BL4 H/13C inverse
detection probe equipped with z-gradient.
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
1058 | New J. Chem., 2009, 33, 1057–1063

View Article Online
using a single exponential fit (Nexp = 1). Each DOSY
experiment was carried out two or three times and for at least
two different sample preparations and the reported values for
the diffusion coefficients are the mean and mean deviation.
The error ranges given indicate the range of values obtained
from the experiments.
showing that the catechins 1–4 form complexes with b-CD.
The B ring of 1, 250 and 330 and the A and C rings of 428 are
inserted into the cavity of b-CD from the wide secondary
hydroxyl group side. The different mode of inclusion of 4 is
attributed to the additional gallate part. Both hydrogen
bonding and hydrophobic interactions were shown to be
involved in the complex formation.28,51
Prior to the DOSY, for experiments with CD3CN as a
co-solvent, the content of CD3CN in D2O ranging from 0%
to 20% (v%) was measured by liquid chromatography using a
column (300 mm ꢂ 5 mm i.d.) packed with CYCLOBOND
I 2000.
Binding to BSA. The STD NMR spectra of catechins 1–4 in
the presence of BSA show large effects (Fig. 1) indicating
binding to the protein. STD effects are observed for all protons
suggesting that all parts of the molecules are involved in the
interaction with the protein. Stronger STDs are seen for H20,
H50, H60 of 1 and 2, and H20, H60 of 3, showing that the B ring
in 1, 2 and 3 has closest contact to the protein. In 4, H90 and
H130 from the D ring in addition to H20 and H60 from the B
ring yield the most intense signals. This is to our knowledge
the first study providing direct experimental evidence on the
part of the catechins that have stronger interaction with BSA.
From docking studies with 4,39 it has been suggested that the
galloyl ring is very important for the interaction with BSA.
From the present work, it appears that the galloyl ring has no
closer contact to the protein than the B ring has, and the
increase in size of the catechin is probably responsible for the
higher binding affinity.
HR-MAS DOSY
b-CD bonded to silica gel (CYCLOBOND I 2000) was used
and obtained from ASTEC. The amount of b-CD bonded to
the silica beads has been quantified previously.28 Prior to
analysis, the gels were submerged in D2O for 1 h, and carefully
inserted into zirconia rotors (4 mm outer diameter, spherical
sample volume of 20 mL, Bruker, Karlsruhe, Germany). The
rotor was subsequently sealed with the rotor spacer, sealing
screw and finally the rotor cap. During the experiments
recorded at 25 1C, the samples were spun at 3 kHz at the
magic angle (54.71). The optimized diffusion time and gradient
length were 150 ms and 2.2 ms, respectively.
Preparation of green tea samples
Green tea extract (100 mg) was dissolved in deuterated solvent
(2 mL), the solution obtained was then clarified by micro-
filtration (0.45 mm) and 0.6 mL of a clear solution was used for
NMR experiments. The concentrated samples were prepared
as follows: 10 mL of 50 mg mLꢀ1 GTE solutions in different
solvents were filtered and dried using freeze drying for the
water solution and N2 for the organic solvents methanol,
acetone and acetonitrile. The dried sample was re-dissolved
in 2 mL deuterated solvents for deuterium exchange and then
dried, the procedure repeated, and finally dissolved in 1.5 mL
solvent for further NMR experiments.
Results and discussion
Binding of catechins to b-CD and BSA
Before analyzing mixtures of catechins by DOSY NMR experi-
ments, their binding to b-CD and BSA were studied. The NMR
techniques used to study complexation or binding between two
molecules are dependent on the size of the molecules involved in
the interaction. For the small molecules b-CD and catechins,
the formation of inclusion complexes was studied by measuring
1H NMR chemical shift changes and by 2D ROESY experi-
ments, whilst the binding of the catechins to the large BSA
receptor was investigated by saturation transfer difference
(STD) NMR.45–49 In this experiment, magnetization is trans-
ferred from the protein to the ligand thereby identifying the
binding of the ligand. The degree of saturation of individual
ligand protons reflects their proximity to the protein surface
and precise binding epitope on the ligand can be obtained.
Binding to b-CD. The chemical shift changes observed upon
addition of equimolar amount of b-CD and the intermolecular
ROEs (see ESIw) were consistent with previous reports30,34,50
Fig. 1 1H reference and STD NMR spectra of catechins (5 mM) and
BSA (0.05 mM): a, a0 (C), b, b0 (EC), c, c0 (EGC) and d, d0 (EGCG).
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 1057–1063 | 1059

View Article Online
DOSY spectra of catechins
Table 1 Diffusion coefficients (log D) in D2O (catechins alone 5 mM,
diffusion/m2 ꢀ1), log D+CD (catechins 5 mM with b-CD 5 mM),
s
DOSY spectra for each of the catechins 1–4 and for a mixture
of them were recorded for D2O, DMSO-d6 and CD3CN
solutions. The diffusion coefficients (D) measured for samples
in D2O are reported in Table 1. Fig. 2(a) shows that the signals
from 4 were separated from those from the other three
catechins due to slower diffusion caused by the higher mole-
cular mass of EGCG. 1, 2 and 3 had similar diffusion
coefficients in the three solvent systems. The complexation
agents b-CD and BSA were then added to the mixture of
catechins in order to see if and how the separation properties
of the DOSY experiment were changed.
log D+BSA (catechins 5.0 mM with BSA 0.05 mM) and log Dfree-DMSO
log Dfree-ACN in DMSO and acetonitrile, respectively
,
C (1)
EC (2)
EGC (3)
EGCG (4)
log Dfree
log D+CD
log D+BSA
log Dfree-DMSO
log Dfree-ACN
ꢀ9.38
ꢀ9.57
ꢀ9.53
ꢀ9.76
ꢀ8.89
ꢀ9.37
ꢀ9.51
ꢀ9.52
ꢀ9.76
ꢀ8.89
ꢀ9.39
ꢀ9.50
ꢀ9.44
ꢀ9.78
ꢀ8.89
ꢀ9.47
ꢀ9.63
ꢀ9.85
ꢀ9.85
ꢀ8.98
Here Dobs is the observed diffusion coefficient, and Dfree and
Dbound are diffusion coefficient of free and bound guest, respec-
tively. [H]0 and [G]0 are the total concentrations of the host and
guest, respectively. The association constants were derived using
single-point procedure53,54 in which it is assumed that the diffu-
sion coefficient of the host–guest complex is the same as that of
the host molecule. The binding strength of the four catechins 1–4
to b-CD is in the order C 4 EGCG 4 EC 4 EGC with K
values for the b-CD/C, b-CD/EC, b-CD/EGC and b-CD/EGCG
inclusion complexes at 25 1C around 21 000, 1000, 600 and
18 000 Mꢀ1, respectively. These values should only be taken as
an indication of the relative affinity of the different catechins for
b-CD due to the large uncertainty introduced by the single-point
approximation method. The values for the b-CD/EC,
b-CD/EGC complexes are similar to those obtained with other
DOSY spectra of catechins in the presence of b-CD. The
observed diffusion coefficients are the weighted average of
those of the free and bound molecules, due to the fast
exchange of free and bound species on the NMR timescale
(eqn (1)). The association constant K is determined by
eqn (2)52 on the premise of known mole fraction w of the
bound guest wb.
Dobs = wDbound + (1 ꢀ w)Dfree
(1)
wb
K ¼
ð2Þ
ð1 ꢀ wbÞð½Hꢃ0 ꢀ wb½Gꢃ0Þ
Fig. 2 DOSY spectra of catechins 1–4 (a: 2.0 mM for each); in the presence of BSA (b: BSA 0.05 mM, catechins 1, 2 and 4 2.0 mM for each;
c: BSA 0.05 mM, catechins 1–3 2.0 mM for each); in the presence of b-CD (d: b-CD 2.0 mM, catechins 2 mM for each in D2O; e: b-CD 8 mM,
catechins 2 mM for each in D2O; f: b-CD 8 mM, catechins 2 mM for each in 15% CD3CN–85% D2O). Only some protons are marked but the
NMR signals for all protons in each catechin were identified.
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
1060 | New J. Chem., 2009, 33, 1057–1063

View Article Online
agreement with results obtained from other types of studies37,38,40
and confirming the importance of the gallate moiety in the
interaction. Addition of BSA resulted in the separation of 3
from 2 and 1 (Fig. 2(c), Table 1). Despite being a larger molecule,
3 has a higher diffusion coefficient than that of 1 or 2,
demonstrating its weaker binding to BSA. The only difference
between 2 and 3 is the presence of a third hydroxyl group in ring
B, this ring being in close proximity to BSA as shown by STD
NMR experiments (Fig. 1). It is thus possible that the additional
OH is not favorable for the interaction. Since the B ring is
identical in 3 and 4, the latter showing the strongest binding to
BSA, it is thus most likely that the two compounds bind to BSA
in different ways.
Although the concentration of BSA was much less than that
of catechins, the aggregation of BSA due to the presence of
catechins resulted in poor spectral resolution. Better resolution
was obtained in the DOSY spectra using b-CD which was
therefore chosen as the complexation agent for the analysis of
green tea extract.
Fig. 3 Plot of the diffusion coefficients of catechins 1–4 in D2O at
25 1C as a function of the concentration of b-CD. The catechin
concentration was kept constant at 4 mM.
methods such as NMR chemical shifts or calorimetry, while
those for b-CD/C and b-CD/EGCG are significantly larger.30,50
The value for the b-CD/C complex is however very similar to the
K value obtained by Julian et al.34 using the same procedure. A
plot of the apparent diffusion coefficient of the catechins 1–4 as a
function of the concentration of b-CD (Fig. 3) shows a successive
decrease in the value of log D until an equimolar amount of
b-CD and catechin is reached. No changes are observed when
b-CD is in excess indicating a 1 : 1 stoichiometry for the
cyclodextrin–catechin complexes.
HR-MAS DOSY of b-CD and mixture of catechins. Fig. 4
shows the HR-MAS DOSY spectrum of a mixture of catechins
1, 2 and 4 and b-CD bonded to silica gel in D2O. Comparison
with the DOSY spectrum obtained with the liquid probe
(Fig. 2(e)) shows that the MAS probe produced a spectrum
with broader diffusion peaks. The lower quality of DOSY
spectra obtained with the HR-MAS probe has been previously
observed and attributed to larger statistical error due to the
vortexing sample.55 Extensive signal averaging was shown to
eliminate partially the observed erratic behavior.55
Upon addition of b-CD (2.0 mM) into the solution of
catechins (2.0 mM each), 1 exhibits a decreased motion because
of its favored complexation with b-CD, leading to the separa-
tion from 2 and 3 (Fig. 2(d)). The diffusion coefficients of 1 and
4 become similar when the amount of b-CD is increased and for
a 1 : 1 molar ratio of b-CD and catechins, the separation
between the two compounds is lost (Fig. 2(e)).
Since we have previously shown that there was no inter-
action between the silica gel column and the catechins and the
apparent diffusion coefficient leading to component identifica-
tion is entirely based on the affinity of a given component for
the b-CD ligand, all subsequent experiments were performed
with the liquid NMR probe.
Due to differences in binding affinity, organic solvents will
affect the equilibrium of the various catechin/b-CD complexes
to a different extent. Thus, the best separation of catechins was
obtained with acetonitrile as the co-solvent in a 15%
CD3CN–85% D2O (v/v) ratio (Fig. 2(f)).
DOSY of green tea extract (GTE)
Water, methanol, acetonitrile and acetone were used to investi-
gate the effect of the solvent on the extraction of compounds
from green tea. As shown in Fig. 5, the water and methanol
extractions on the one hand and the acetonitrile and acetone
extractions on the other hand produce relatively similar proton
NMR spectra. The low solubility of carbohydrates in aceto-
nitrile and acetone leads to relatively simpler NMR spectra
(e.g., the 2.5–4.5 ppm region) for extractions performed with
these solvents. Major metabolites such as caffeine and theanine
DOSY spectra of catechins in the presence of BSA. The
diffusion of catechins 1–4 in the presence of BSA was studied in
aqueous solution using a 40 molar ratio excess of catechins and
Fig. 2(b) shows that 4 still has the lower diffusion coefficient.
Also, the changes in the apparent diffusion coefficients (Table 1)
indicate that 4 is binding most strongly to BSA, in good
1
could be readily assigned from the one-dimensional H NMR
spectra by comparing the resonances to those of well-defined
standard compounds as well as to those reported for green
tea.56 The presence of other compounds such as quinic acid,
alanine, sucrose and 2-O-(b-L-arabinopyranosyl)myoinositol
was confirmed from 2D COSY, TOCSY and HSQC experi-
ments. Since the main goal of this study was to assess the use of
DOSY to identify and separate catechin components from
green tea, no further effort was made to assign the other
metabolites present in the NMR spectra of the green tea extract.
The GTE samples were concentrated in order to increase the
signal intensity. For concentrated samples in water or in
Fig. 4 DOSY spectra of catechins 1, 2 and 4 (2.0 mM for each) and
b-CD bonded to silica gel recorded with an HR-MAS probe in D2O.
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 1057–1063 | 1061

View Article Online
Fig. 6 DOSY spectra of the concentrated samples in (a) acetonitrile,
(b) acetone and (c) with b-CD in 15% CD3CN–85% D2O (sample 2).
Fig. 5 1H NMR spectra of GTE in different solvents (a) methanol-d4,
(b) acetone-d6, (c) acetonitrile-d3, (d) D2O. * Indicates solvent peak.
Fig. 2(f), confirming the identification of these catechins in
green tea extract. The signals of C and EC were weak if
compared to those from EGCG due to their lower amounts
in green tea (Fig. 6(c)). EGCG could also be assigned in the
DOSY spectra of sample 1 (data not shown) but signals from
EC and C were not observed. In addition to changing the
equilibrium of the cyclodextrin–catechin inclusion complexes,
the use of acetonitrile as a co-solvent to water (sample 2)
improved the resolution of the DOSY spectra.
methanol, greater shielding and lower diffusion coefficients
due to higher viscosity57 were observed (see ESIw). Thus, for
methanol and water extractions, concentrations above 15 and
50 mg mLꢀ1, respectively, lead to a decrease in diffusion rate
due to an increase of the medium viscosity. For the acetonitrile
and acetone extracts, the highly concentrated sample of
300 mg mLꢀ1 showed the same diffusion as the 10 mg mLꢀ1
GTE extraction because of the absence of sugars. Therefore,
the use of acetonitrile and acetone as solvents was preferred in
the study of catechins from green tea extract, while water and
methanol can be used to investigate compounds with higher
polarity such as carbohydrates. The catechins EGCG, EC and
C were clearly identified and separated in the DOSY spectra of
the concentrated sample (Fig. 6(a) and (b)). A better resolu-
tion in the diffusion dimension was obtained with acetonitrile
making the assignment of catechins in the mixture easier.
Thus, the sample obtained from acetonitrile extraction was
after solvent evaporation, dissolved in D2O (sample 1) and in
15% CD3CN–85% D2O (sample 2) and b-CD was added to
each sample. Due to the fact that b-CD forms inclusion
complexes with many of the small molecules present in green
tea, the amount of b-CD was increased until no chemical shift
changes were observed in the 1H NMR spectrum (Fig. 6(c)). It
was then assumed that full complexation between b-CD and
the binding molecules from the green tea extract were reached.
With sample 2, the diffusion coefficients of EGCG, EC and C
in complex with b-CD were consistent with those shown in
Conclusions
b-CD and BSA, coupled to solid chromatographic supports
such as silica gel or agarose, are widely used in chromato-
graphy to separate mixtures on the basis of molecular inter-
action. In the present work, it is shown that they can be used
to enhance or modify the spectral separation of the DOSY
spectra of catechins. Thus, while EGCG could be separated
from a mixture of four catechins due to its higher molecular
size, separation of EGC from C and EC on the basis of their
apparent diffusion coefficient was achieved with BSA while C,
EC and EGC could be differentiated using b-CD and different
solvent systems. A better knowledge of the binding interaction
between small metabolites and different types of receptors
should increase the ability to identify compounds in a mixture
by this type of experiment. A drawback is that chemical shift
changes are often induced upon binding, their amplitude being
dependant on the ligand and on the strength of binding.
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009
1062 | New J. Chem., 2009, 33, 1057–1063

View Article Online
26 G. Pages, C. Delaurent and S. Caldarelli, Angew. Chem., Int. Ed.,
2006, 45, 5950.
27 S. Caldarelli, Magn. Reson. Chem., 2007, 45, S48.
28 J. Xu, T. Tan, J. C. Janson, L. Kenne and C. Sandstrom,
¨
Carbohydr. Res., 2007, 342, 843.
Acknowledgements
The authors acknowledge financial support given by grants
from the Swedish Research Council, the National Science
Foundation of China (Grants 20636010, 20325622, 20576013
and 50373003) and the 973 Project (2007CB714304).
29 J. Xu, C. Sandstrom, J. C. Janson and T. Tan, Chromatographia,
¨
2006, 64, 7.
30 Y. Cai, S. H. Gaffney, T. H. Lilley, D. Magnolato, R. Martin,
C. M. Spencer and E. Haslarm, J. Chem. Soc., Perkin Trans. 2,
1990, 2197.
References
31 K. S. Cameron and L. Fielding, Magn. Reson. Chem., 2002, 40,
S106.
32 S. Simova and S. Berger, J. Inclusion Phenom. Macrocyclic Chem.,
2005, 53, 163.
´
33 A. Guerrero-Martınez, M. A. Palafox and G. Tardajos, Chem.
1 T. Yamamoto, J. Wataha, D. Dickinson, B. Singh, W. B. Bollag,
E. Ueta and T. Osaki, J. Pharmacol. Exp. Ther., 2004, 308, 317.
2 M. Kawase, R. Wang, T. Shiomi, R. Saijo and K. Yagi, Biosci.,
Biotechnol., Biochem., 2000, 64, 2218.
3 Y. K. Yee and M. W. L. Koo, Aliment. Pharmacol. Ther., 2000, 14,
635.
4 C. S. Yang, P. Maliakal and X. Meng, Annu. Rev. Pharmacol.
Toxicol., 2002, 42, 25.
5 P. M. Kris-Etherton and C. L. Keen, Curr. Opin. Lipidol., 2002, 13,
41.
6 C. F. Hung, T. F. Huang, H. S. Chiang and W. B. Wu, J. Cell.
Biochem., 2005, 96, 183.
7 M. P. Williamson, T. G. Mccormick, C. L. Nance and
W. T. Shearer, J. Allergy Clin. Immunol., 2006, 118, 1369.
8 E. Nishitani and Y. M. Sagesaka, J. Food Compos. Anal., 2004, 17,
675.
9 H. Wang, K. Helliwell and X. You, Food Chem., 2000, 68, 115.
10 J. Z. Daniel, C. N. Bryant, A. Klaus and J. D. Joseph, Anal. Chem.,
2000, 72, 5020.
11 T. Goto, Y. Yoshida, M. Kiso and H. Nagashiama,
J. Chromatogr., A, 1996, 749, 295.
Phys. Lett., 2006, 432, 486.
34 C. Jullian, S. Miranda, G. Zapate-Torres, F. Mendizabal and
C. Olea-Azar, Bioorg. Med. Chem., 2007, 15, 3217.
35 R. K. Gilpin, S. E. Ehtesham and R. B. Gregory, Anal. Chem.,
1991, 63, 2825.
36 M. Fiori, M. Farne, C. Civitareale, A. Nasi, L. Serpe and P. Gallo,
Chromatographia, 2004, 60, 253.
37 A. Papadppoulou, R. J. Green and R. A. Frazier, J. Agric. Food
Chem., 2005, 53, 158.
38 S. Soares, N. Mateus and V. D. Freitas, J. Agric. Food Chem.,
2007, 55, 6726.
39 T. K. Maiti, K. S. Ghosh and S. Dasgupta, Proteins, 2006, 64, 355.
40 M. Kusuda, T. Hatano and T. Yoshida, Biosci., Biotechnol.,
Biochem., 2006, 70, 152.
41 R. A. Frazier, A. Papadopoulou and R. J. Green, J. Pharm.
Biomed. Anal., 2006, 41, 1602.
42 M. Piotto, V. Saudek and V. Sklenar, J. Biomol. NMR, 1992, 2,
661.
43 A. J. Charlton, N. J. Baxter, M. L. Khan, A. J. G. Moir,
E. Haslam, A. P. Davies and M. P. Williamson, J. Agric. Food
Chem., 2002, 50, 1593.
12 R. Saijo, Agric. Biol. Chem., 1982, 46, 1969.
13 D. Baumann, S. Adler and M. Hamburger, J. Nat. Prod., 2001, 64,
353.
14 N. S. Kumar and M. Rajapaksha, J. Chromatogr., A, 2005, 1083,
223.
44 N. Hayashi, T. Ujihara and K. Kohata, Biosci., Biotechnol.,
Biochem., 2004, 68, 2512.
15 W. Si, J. Gonga, R. Tsao, M. Kalab, R. Yang and Y. Yin,
J. Chromatogr., A, 2006, 1125, 204.
45 M. Mayer and B. Meyer, Angew. Chem., Int. Ed., 1999, 38, 1784.
46 J. Klein, R. Meinecke, M. Mayer and B. Meyer, J. Am. Chem.
Soc., 1999, 121, 5336.
47 M. Vogtherr and T. Peters, J. Am. Chem. Soc., 2000, 122, 6093.
48 M. Mayer and B. Meyer, J. Am. Chem. Soc., 2001, 123, 6108.
49 B. Meyer and T. Peters, Angew. Chem., Int. Ed., 2003, 42, 864.
50 T. Ishizu, K. Kintsu and H. Yamamoto, J. Phys. Chem. B, 1999,
103, 8992.
16 M. Lin, M. J. Shapiro and J. R. Wareing, J. Am. Chem. Soc., 1997,
119, 5249.
17 K. Bleicher, M. Lin, M. J. Shapiro and J. R. Wareing, J. Org.
Chem., 1998, 63, 8486.
18 J. S. Gounarides, A. Chen and M. J. Shapiro, J. Chromatogr., B:
Biomed. Sci. Appl., 1999, 725, 79.
19 C. S. Johnson, Jr., Prog. Nucl. Magn. Reson. Spectrosc., 1999, 34,
203.
51 J. Xu, T. Tan and J. C. Janson, J. Chromatogr., A, 2006, 1137, 49.
52 R. Wimmer, F. L. Aachmann, K. L. Larsen and S. B. Peterson,
Carbohydr. Res., 2002, 337, 841.
20 M. D. Diaz and S. Berger, Carbohydr. Res., 2000, 329, 1.
21 M. D. Pelta, G. A. Morris, M. J. Stchedroff and S. J. Hammond,
Magn. Reson. Chem., 2002, 40, S147.
53 D. A. Stauffer, R. E. Barrans and D. A. Dougherty, J. Org. Chem.,
1990, 55, 2762.
54 K. S. Cameron and L. Fielding, J. Org. Chem., 2001, 66, 6891.
55 S. A. Bradley, J. Paschal and P. Kulanthaivel, Magn. Reson.
Chem., 2005, 43, 31.
22 H. Utsumi, H. Seki, K. Yamaguchi and M. Tashiro, Anal. Sci.,
2003, 19, 1441.
23 P. Croves, M. O. Rasmussen, M. D. Molero, E. Samain,
´
F. J. Canada, H. Driguez and J. Jimenez-Barbero, Glycobiology,
2004, 14, 451.
56 G. L. Gall, I. J. Colquhoun and M. Defernez, J. Agric. Food
Chem., 2004, 52, 692.
24 S. Viel, F. Ziarelli and S. Caldarelli, Proc. Natl. Acad. Sci. U. S. A.,
2003, 100, 9696.
57 G. Savary, E. Guichard, J. L. Doublier, N. Cayot and C. Moreau,
J. Agric. Food Chem., 2006, 54, 665.
25 G. Pages, C. Delaurent and S. Caldarelli, Anal. Chem., 2006, 78,
561.
ꢁc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2009 New J. Chem., 2009, 33, 1057–1063 | 1063