Preparation and characterization of Catechin sulfates, glucuronides, and methylethers with metabolic interest

Article abstract of DOI:10.1021/jf803140h

Catechins are major polyphenols in many plant foods that have been related to health promotion. In the human organism they are largely metabolized to different conjugates (sulfates, glucuronides, and methylethers), which are further found in plasma and would contribute to the biological effects associated with the intake of the parent compounds. Circulating metabolites are likely to possess biological properties different from those of the original compounds, and therefore, it is important to evaluate their activity, for which sufficient amounts of them are required that cannot be obtained by isolation from biological fluids. This paper describes the preparation of the methyl, sulfate, and glucuronide derivatives of catechins using different chemical syntheses and their characterization by HPLC-DAD-ESI/MS. MS² fragmentation of the compounds was also described that allowed the determination of the location of the different substituents on the catechin aglycones. The procedures optimized allowed the preparation of (epi)catechin sulfates, glucuronides, and methylethers conjugated at positions 3' and 4', as well as the sulfates at positions 5 and 7 with satisfactory yields for their further isolation by semipreparative-HPLC in view of their use in in vitro/ex vivo assays.

Full text of DOI:10.1021/jf803140h

J. Agric. Food Chem. 2009, 57, 1231–1238 1231
Preparation and Characterization of Catechin
Sulfates, Glucuronides, and Methylethers with
Metabolic Interest
´
´
´
SUSANA GONZALEZ-MANZANO, ANA GONZALEZ-PARAMAS,
˜
CELESTINO SANTOS-BUELGA,* AND MONTSERRAT DUENAS
Grupo de Investigacio´n en Polifenoles, Unidad de Nutricio´n y Bromatolog´ıa, Facultad de Farmacia,
Universidad de Salamanca, Campus Miguel de Unamuno, E-37007, Salamanca, Spain
Catechins are major polyphenols in many plant foods that have been related to health promotion. In
the human organism they are largely metabolized to different conjugates (sulfates, glucuronides,
and methylethers), which are further found in plasma and would contribute to the biological effects
associated with the intake of the parent compounds. Circulating metabolites are likely to possess
biological properties different from those of the original compounds, and therefore, it is important to
evaluate their activity, for which sufficient amounts of them are required that cannot be obtained by
isolation from biological fluids. This paper describes the preparation of the methyl, sulfate, and
glucuronide derivatives of catechins using different chemical syntheses and their characterization by
HPLC-DAD-ESI/MS. MS2 fragmentation of the compounds was also described that allowed the
determination of the location of the different substituents on the catechin aglycones. The procedures
optimized allowed the preparation of (epi)catechin sulfates, glucuronides, and methylethers conjugated
at positions 3′ and 4′, as well as the sulfates at positions 5 and 7 with satisfactory yields for their
further isolation by semipreparative-HPLC in view of their use in in vitro/ex vivo assays.
KEYWORDS: catechin metabolites; methylcatechin; sulfates; glucuronides
INTRODUCTION
ronide and sulfate residues; nonmethylated catechin was mainly
found as sulfate and sulfoglucuronide conjugates, whereas 3′-
O-methylcatechin was present primarily as a glucuronide.
Similar results were found in rats by Baba et al. (12), which
showed that absorbed catechin and epicatechin were mostly
present in plasma as glucuronides, sulfates, and sulfoglucu-
ronides, both in the form of nonmethylated or 3′-O-methylated
derivatives. Methylation and glucuronidation are already pro-
duced in the enterocyte, while additional methylation and
sulfation would further take place in the liver (13). Perfusion
of isolate rat jejunum with flavanols resulted in glucuronidation
(∼45%), O-methylation: 3′-O-methyl and 4′-O-methyl (30%)
and O-methyl glucuronidation (20%) during transfer across the
enterocytes to the serosal side (14). Nevertheless, differences
may exist in the profile of metabolites between rats and humans.
Natsume et al. (15) isolated epicatechin 3′-O-glucuronide, 4′-
O-methylepicatechin 3′-O-glucuronide, and 4′-O-methylepicat-
echin 5- or 7-O-glucuronide from human urine. These same
authors purified 3′-O-methylepicatechin, epicatechin-7-O-glu-
curonide, and 3′-O-methylepicatechin-7-O-glucuronide from rat
urine. Bhatathi et al. (16) indicated that sulfation constitutes
the major pathway in epicatechin metabolism in humans, in both
liver and intestine, with no glucuronidation occurring. This is
not in agreement with the observations made by Natsume et al.
(15) or Donovan et al. (11), who found glucuronide conjugates
of (methyl)catechins in either human urine or plasma. These
Flavan-3-ols (catechins and proanthocyanidins) are major polyphe-
nols in many plant foods that have been related to health promotion.
Indeed a large number of epidemiological studies demonstrate that
the intake of flavan-3-ol-rich products (i.e., tea, red wine, apples,
or chocolate) is inversely associated with the risk of coronary heart
disease, cancer, and inmunodysfunctions (1-4). Epidemiological
observations are supported by many in vitro and ex vivo studies
displaying a remarkable scope of biochemical and pharmacological
actions of these compounds, among them antiviral (5), anti-
inflammatory (6), antiallergic (7), antioxidant and free radical
scavenging properties (8, 9).
Because of the potential beneficial role of these compounds
in human health, it is essential to understand their bioavailability,
i.e., absorption, metabolism, and excretion. It has been shown
that catechins are absorbed from the human intestinal tract,
largely metabolized and distributed as conjugated derivatives
in blood, and that these forms are excreted in urine (10). Methyl,
sulfate, and glucuronide conjugates have been described as
circulating metabolites of catechins. In studies with human
volunteers, Donovan et al. (11) found that after consumption
of wine, catechin was present in plasma as both nonmethylated
and 3′-O-methylated (21%) derivatives conjugated with glucu-
* To whom correspondence should be addressed (telephone 00 34
923294537; fax 00 34 923294515; e-mail csb@usal.es).
10.1021/jf803140h CCC: $40.75  2009 American Chemical Society
Published on Web 01/30/2009

1232 J. Agric. Food Chem., Vol. 57, No. 4, 2009
Gonza´lez-Manzano et al.
carbonate (500 mg), the mixture was stirred for 2 h at room temperature.
The resulting mixture was added to 150 g of cooled water and then
adjusted to acidic condition by adding a few drops of formic acid. The
acetone was removed in a Bu¨chi R-124 rotary evaporator (Bu¨chi
Labortechnik AG, Switzerland) at 30 °C to dryness. The residue
obtained was redissolved in ultrapure water (Direct-Q, Millipore,
Molsheim, France) and submitted to semipreparative HPLC to remove
the remaining reagents and side products of the reaction. A fraction
containing intermediate compounds was collected, evaporated to dryness
under vacuum, and redissolved in 50 mL of methanol. This solution
was analyzed HPLC-DAD-MS, showing that it contained a mixture
of five compounds with a pseudomolecular ion [M - H]- at m/z 605
corresponding to the different catechin triacetylglucuronic methyl esters.
Then 750 µL of sodium methylate (28:72, v/v sodium methylate/
methanol) was added to the solution and kept at 4 °C for 30 min to
remove acetyl moieties bound to the glucuronide residue. Further
addition of 7.5 mL of ultrapure water was made and allowed to react
30 min at room temperature in order to hydrolyze the methyl esters.
The solution was neutralized (1.85 mL of 2 N hydrochloric acid),
adjusted to acidic condition by adding a few drops of formic acid, and
then evaporated to dryness and redissolved in water. The catechin
glucuronides obtained were characterized by HPLC-DAD-MS. For
their isolation a fractionation was first made in a Sephadex LH-20
column (350 mm × 30 mm) eluted with 10% aqueous ethanol (500
mL) and 20% aqueous ethanol (500 mL), and further purification of
the compounds was performed by semipreparative HPLC.
latter authors differ, however, in the preferential position of
substitution of the methyl residue (i.e., 3′ or 4′) on the catechin.
Circulating metabolites would reach the biological targets,
and therefore, they should be the main actors to explain the
health effects associated with the intake of catechins and
flavonoids in general. Conjugated metabolites are likely to
possess different biological properties than do parent com-
pounds, and therefore, in vitro studies should also consider
metabolites rather those only commercially available compounds
as found in foodstuffs (17). Differences should also exist
between the distinct type of conjugates (sulfates, glucuronides,
and methylated compounds) and the different position at which
the conjugation occurs. However, far less is known about the
activity and distribution patterns of the metabolites than about
their precursors, and relatively few analytical methods have been
developed in order to study these metabolites in biological fluids
and tissues. Previous work has shown that the positions of the
sugar moieties have a large effect on the bioactivity of flavonoids
(18) and a similar effect was noted for conjugated metabolites
(19). Thus, the ability to detect, identify, and elucidate the
structure of flavonoid metabolites would be extremely useful.
In this respect, liquid chromatography-mass spectrometry
(LC-MSn) is in many ways an ideal method for analyzing and
identifying flavonoid metabolites owing to its high sensitivity,
applicability to complex mixtures, and ability to provide
structural information.
Preparation of (Epi)Catechin Sulfates. These metabolites were
synthesized by a modification of the method described by Jones et al.
(24) for the preparation of quercetin sulfates. First, water associated
with (epi)catechin (500 mg) was removed by adding dry pyridine until
it dissolved. Pyridine was rotary-evaporated, and the dry compound
was dissolved in dioxane (50 mL) and allowed to react with a 10-fold
molar excess of sulfur trioxide-N-triethylamine complex under argon
to avoid contact with air. This reaction took place in a water bath (40
°C) for 90 min, after which products of sulfation precipitated out and
stuck to the glass. Then dioxane was decanted and the product (a
mixture of different mono- and disulfates of (epi)catechin, as well as
the aglycone) redissolved in ethanol and water (10:90). The compounds
were fractionated on a Sephadex LH-20 column (350 mm × 30 mm)
eluted with 10% aqueous ethanol (500 mL) and 20% aqueous ethanol
(500 mL) to separate (epi)catechin monosulfates, disulfates, and
aglycone. The fraction containing the monosulfates was collected,
concentrated to dryness under vacuum, redissolved in ultrapure water,
and analyzed by HPLC-DAD-MS. Different (epi)catechin sulfates
were then isolated from this fraction by semipreparative HPLC.
HPLC-DAD-ESI/MS Analyses. They were carried out in a
Hewlett-Packard 1100 chromatograph (Agilent Technologies, Wald-
bronn, Germany) with a quaternary pump and a diode array detector
(DAD) coupled to an HP Chem Station (revision A.05.04) data-
processing station. A Waters Spherisorb S3 ODS-2 C8, 3 µm (4.6 mm
× 150 mm) column thermostated at 30 °C was used. Solvents employed
for the analysis of catechin methylethers and glucuronides were (A)
2.5% acetic acid, (B) acetic acid/acetonitrile (10: 90, v:v), and (C)
acetonitrile. The elution gradient established was 100% A to 100% B
over 5 min, 0-15% C in B over 35 min, and 15-40% C in B over 10
min, using a flow rate of 0.5 mL ·min-1. For the analysis of catechin
monosulfates solvents were (A) 0.1% TFA in water and (B) acetonitrile;
the elution gradient was 0% A to 10% B over 5 min, 0-15% A in B
over 20 min, and 15-20% C in B over 5 min, using a flow rate of 0.5
Some studies have been published regarding the activity of
conjugated metabolites of quercetin (20, 21), but hardly any
data exist concerning catechin metabolites. In order to evaluate
the activity of catechin metabolites in in vitro and cell model
assays, as well as to optimize methods for their analysis,
sufficient amounts of pure compounds are required that cannot
be obtained by isolation from biological fluids. This work deals
with the synthesis and characterization of catechin metabolites,
i.e., sulfate, glucuronide, and methylated derivatives. The
knowledge of their chromatographic behavior, mass spectra, and
fragmentation patterns, as provided by HPLC-DAD-ESI/MS,
is expected contribute to their correct identification in biological
samples and to the understanding of their metabolism in vivo.
MATERIALS AND METHODS
Standards and Reagents. Catechin, acetobrom-R-D-glucuronic acid
methyl ester, sodium methylate, methyl iodide, and potassium carbonate
were purchased from Sigma-Aldrich (Milwaukee, WI). Pyridine,
dioxane, and sulfur trioxide-N-triethylamine were from Sigma (Poole,
U.K.). HPLC grade methanol and ethanol were purchased from Carlo
Erba (Milan, Italy). Acetone, glacial acetic acid, formic acid, and
hydrochloric acid were of analytical grade and obtained from Panreac
(Barcelona, Spain). Trifluoroacetic acid (TFA) was purchased from
Riedel-de Hae¨n (Seelze, Germany).
Preparation of (Epi)Catechin Methylethers. The methylethers of
(epi)catechin were synthesized on the basis of the protocol described
by Donovan et al. (22). A mixture of catechin or epicatechin (250 mg),
potassium carbonate (500 mg), and methyl iodide (1 mL) was prepared
in acetone (20 mL) and irradiated in an ultrasonic bath. The progress
of the reaction was monitored by HPLC. After a reaction time of 3.5 h
the solvent was filtered and concentrated in rotary evaporator to dryness.
The methylethers of (epi)catechin synthesized were characterized by
HPLC-DAD-MS. The major products of the reaction (3′- and 4′-
methylethers of catechin and epicatechin) were further purified by
semipreparative HPLC.
mL·min-1
.
MS detection was performed in a Finnigan LCQ detector (Thermo-
quest, San Jose, CA) equipped with an ESI source and an ion trap
mass analyzer, which were controlled by the LCQ Xcalibur software.
The mass spectrometer was connected to the HPLC system via the
DAD cell outlet. Both the auxiliary and the sheath gases were nitrogen
at flow rates of 20 and 80 L min-1, respectively. The source voltage
was 4.5 kV, the capillary voltage was 11 V, and the capillary
temperature was 220 °C. Spectra were recorded in positive (catechin
methylethers) and in negative (catechin sulfates and glucuronides) ion
modes between m/z 150 and 2000. The MS detector was programmed
to perform a series of three consecutive scans: a full mass scan, an
Preparation of Catechin Glucuronides. Synthesis of the catechin
glucuronides was based on the Koenigs-Knorr previously employed
by Tsushida et al. (23) for the preparation of quercetin glucosides.
Catechin (500 mg) and acetobrom-R-D-glucuronic acid methyl ester
(2.5 g) were dissolved in 25 mL of acetone. After addition of potassium

Catechin Sulfates, Glucuronides, and Methylethers
J. Agric. Food Chem., Vol. 57, No. 4, 2009 1233
Figure 1. Chromatograms recorded at 280 nm showing the result obtained in the hemisynthesis of catechin methylethers after (A) 2 h and (B) 3.5 h of
reaction. Peaks are as follows: (1) catechin; (2) 5-O-methylcatechin; (3) 3′-O-methylcatechin; (4) 7-O-methylcatechin; (5) 4′-O-methylcatechin.
MS/ MS scan of the most abundant ion the first scan, and an MS3 of
preferentially in the 3′ position, although in our case higher yield
the most abundant ion in the MS2 using normalized collision energy
in the formation of the 4′-methylether than of the 3′ derivative
of 45%.
was observed. Minor formation of the methylated derivatives
at 5- and 7-positions was also found (Figure 1), whereas no
Semipreparative HPLC. A Waters 600 chromatograph coupled to
a UV-vis model 486 detector and an Ultracarb C18ODS20 5 µm (10
production of the 3-methylether could be observed, indicating
mm × 250 mm, i.d.) column from Phenomenex (Supelco Ascentis,
that the B-ring is the preferential site for methylation in the
Bellefonte, PA) were used. The solvents were (A) 5% acetic acid and
(B) methanol. The flow rate was 3 mL ·min-1. The gradient used was
different for the different catechin derivatives. For the intermediate
products in the synthesis of catechin glucuronides (i.e., catechin
triacetylglucuronic mehylesters), the gradient was 10-20% B in 15
min, 20-40% B in 15 min, 40-55% B in 10 min. For the separation
of catechin glucuronides, the gradient was 0-10% B in 5 min and
10-30% B in 20 min. For the separation of the catechin sulfates, the
gradient was 0-10% B in 40 min and 10-12% B in 20 min, and for
the catechin methylethers, it was 0-20% B in 15 min, 20-30% B in
25 min, and 30-40% B in 5 min. Detection was carried out at 280
nm, and the peaks were collected in a fraction collector. Fractions
obtained were concentrated under vacuum at low temperature and
freeze-dried. The identity of the compounds was checked by
HPLC-DAD-MS.
reaction conditions used. No mention to the formation of none
of these latter products was made by Donovan et al. (22).
Application of the same procedure using epicatechin instead
catechin gave similar results, with formation of the correspond-
ing 3′- and 4′-methylethers.
Methods for the preparation of catechin glucuronides and
sulfates had to be optimized on the basis of methodologies
previously described for the hemisynthesis of similar quercetin
derivatives. It is noticed that relevant differences exist between
both types of flavonoids regarding the acidity of their various
hydroxyl groups, which provokes differences in the regiose-
lectivity of the syntheses. Thus, distinct types and/or distribution
of the reaction products should be expected in each case.
Flavonols have A- and B-rings conjugated to the 4-carbonyl,
and relevant differences exist between the distinct hydroxyls.
In the case of quercetin, the 5-hydroxyl is by far the least acidic
because of its hydrogen bonding with the 4-carbonyl. By
contrast, 7- and 4′-hydroxyls are the most acidic, which is
explained by their para positions regarding the conjugated chain.
Compared to them, 3- and 3′-hydroxyls show reduced acidity,
explained by the meta orientation of the 3′-hydroxyl in relation
to the conjugated chain, and the possibility of weak hydrogen
bonding with the 4-carbonyl in the case of the 3-hydroxyl; in
addition this position is more hindered than the others. The more
acidic hydroxyls of flavonols, made more nucleophilic by
deprotonation under the mildly basic conditions commonly
RESULTS AND DISCUSSION
Preparation of Catechin Conjugates. The method for the
preparation of catechin methylethers described by Donovan et
al. (22) was applied with small modifications regarding the time
of reaction. In the original method a time of 2 h was used;
however, in our conditions low amounts of conjugates were
accumulated at that time and the reaction was kept 3.5 h, at
which the peaks of the methylethers reached their maximum
area. The majority formation of two products was observed
(Figure 1), and they were assigned to the 3′- and 4′-methylethers
of catechin, as previously shown by Donovan et al. (22). These
same authors (11) indicated that the catechin was methylated

1234 J. Agric. Food Chem., Vol. 57, No. 4, 2009
Gonza´lez-Manzano et al.
Figure 2. Chromatogram recorded at 280 nm showing the profile of catechin glucuronides obtained by hemisynthesis. Peaks are as follows: (1) catechin-
5-O-glucuronide; (2) catechin-3-O-glucuronide; (3) catechin-3′-O-glucuronide; (4) catechin-7-O-glucuronide; (5) catechin-4′-O- glucuronide.
employed for conjugation, are often more reactive, and therefore,
preferential formation of those conjugates should occur. The
case of catechins is quite different. The lack of the 4-carbonyl
and the 2,3-double bond results in similar acidities for all the
phenolic hydroxyls, and therefore, no regioselective conjugation
is expected (25). Thus, modifications had to be introduced in the
protocols described for the preparation of the quercetin conju-
gates to adapt them to the synthesis of the catechin derivatives
and also particular attention had to be paid to the steps of
purification and characterization of the metabolites because of
their similar characteristics.
Catechin glucuronides were chemically synthesized using the
Koenigs-Knorr reaction following the protocol described in
the section of experimental. Major products obtained were
catechin-4′-O-glucuronide (40%) and catechin-3′-O-glucuronide
(30%), followed by catechin-5-O-glucuronide (11%), catechin-
7-O-glucuronide (9%), and catechin-3-O-glucuronide (3%)
Figure 3. Chromatogram recorded at 280 nm obtained in the same HPLC
(Figure 2), indicating that in the reaction conditions phenolic
conditions as the chromatograms of Figures 1 and 2 showing the result
hydroxyls in the B-ring are more reactive than those in the
of the hemisynthesis of catechin sulfates. Peaks are as follows: (1) mixture
A-ring and that the nonphenolic hydroxyl at position 3 is the
of catechin disulfates; (2) mixture of catechin monosulfates; (3) catechin.
less reactive. This is in agreement with the acidities calculated
for the four catechin hydroxyl groups by Cren-Olive´ et al. (26),
according to which hydroxyls at 5 and 7 positions possess
slightly higher pKa values (9.48 and 9.58) than those at 3′ and
4′ (9.02 and 9.12).
Figure 3 shows the HPLC chromatogram of the crude extract
resulting from the synthesis of catechin sulfates obtained in the
same conditions used for the separation of catechin methylethers
and glucuronides, in which mono- and disulfates appear as two
unresolved peaks. Both types of products could be easily
separated in different fractions in a Sephadex LH-20 column.
Monosulfates could be resolved by changing the analytical
HPLC gradient conditions using trifluoroacetic acid instead of
Figure 4. Chromatogram recorded at 280 nm showing the profile of
acetic acid as elution solvent (Figure 4). Four compounds were
catechin monosulfates obtained. HPLC conditions are different from those
detected corresponding to the monosulfates at positions 5, 3′,
employed for the chromatograms shown in the Figures 1-3. Peaks are
7, and 4′ of the catechin, respectively. Their elution order was
as follows: (1) catechin-5-O-sulfate; (2) catechin-3′-O-sulfate; (3) catechin-
established based on their chromatographic behavior compared
7-O-sulfate; (4) catechin-4′-O- sulfate.
with other catechin conjugates and MSn fragmentation pattern
(see below). No formation of the sulfate at position 3 was
observed. Higher yield was obtained in the formation of
catechin-3′-O-sulfate (38% of the monosulfates prepared) than
for the other three monosulfates (around 20% each). Similar
results were found when the hemisynthesis was carried out with
epicatechin. No good separation and characterization of the
complex mixture of catechin bisulfates obtained in the hemi-
synthesis procedure could be obtained.
HPLC of the following metabolites: 3′-methylether and 4′-
methylether of (epi)catechin, 3′-O-glucuronide and 4′-O-glu-
curonide of (epi)catechin-, and 3′-O-sulfate, 4′-O-sulfate, 5-O-
sulfate, and 7-O-sulfate of (epi)catechin. In the conditions
described these compounds were obtained at the level of tens
of milligrams, sufficient for their use in in vitro/ex vivo activity
assays, as well as to optimize their LC-MS characterization,
useful for their analysis in biological fluids.
The procedures above-described allowed the preparation with
satisfactory yields for their further isolation by semipreparative
CharacterizationofCatechinConjugatesbyHPLC-DAD-
ESI/MS. The same HPLC conditions were used for the

Catechin Sulfates, Glucuronides, and Methylethers
J. Agric. Food Chem., Vol. 57, No. 4, 2009 1235
Figure 5. MS2 spectra of different catechin metabolites: catechin methylethers (pseudomolecular ion [M + H]+ at m/z 305) conjugated on the A-ring (A)
and B-ring (B); catechin glucuronides (pseudomolecular ion [M - H]- at m/z 465) conjugated on the A-ring (C) and B-ring (D); catechin sulfates
(pseudomolecular ion [M - H]- at m/z 369) conjugated on the A-ring (E) and B-ring (F). Spectra of the catechin methylethers were obtained in positive
ion mode and spectra of catechin glucuronides and sulfates in negative ion mode.
separation of (epi)catechin methylethers and glucuronides,
whereas different conditions were optimized for the separation
of the monosulfates using TFA instead of acetic acid as HPLC
solvent. ESI/MS identification of the methylether derivatives
was performed using positive ionization mode so that compari-
son could be made with the results previously reported by
Donovan et al. (22) and Cren-Olive´ et al. (28). ESI/MS
characterization of catechin glucuronides and sulfates was made
using negative ionization mode. The fragmentation ion nomen-
clature recently proposed for flavones and flavonols (27) has
been adopted to denote the product ions. The labels ijA and ijB
refer to the fragment containing A and B ring, respectively, and
the superscripts i and j indicate the C-ring bonds that have been
broken.
The chromatogram in Figure 1 shows the result obtained in
the synthesis of the catechin methylethers. Four peaks with a
pseudomolecular [M - H]+ ion at m/z 305 were detected, which
were associated with different products of the methylation on
each of the phenolic hydroxyl groups of the catechin. Similar
synthesis carried out by Donovan et al. (22) revealed the
preferential formation of the 3′- and 4′-O-methyl derivatives of
catechin, as identified by NMR, the first one eluting earlier in
reversed-phase HPLC. Studies carried out by Cren-Olive´ et al.
(28) by LC-MS also showed that 3′-O-methylcatechin eluted
before the 4′-O-methyl derivative and that 5-O-methylcatechin
eluted before the corresponding 7-O-methyl analogue.
No relevant differences were found in the MS2 spectra of
compounds 2 and 4 (Figure 5A), indicating that they
possessed the same fragmentation pattern (Figure 6). The
ion at m/z 153 results from the retro Diels-Alder (RDA)
fission and gives rise to a 1,3A+ ion with the methyl group
(139 + 14)153). The ion at m/z 165 would correspond to
the 1,4B+ ion from the 1,4 fission of the C-ring. Further
dehydration of this product yields the ion at m/z 147. The
ion at m/z 123 results from the 1,2 fission of the B ring that
gives rise to a 1,2B+ ion. The ion at m/z 183 corresponds to
the 1,2 fission of the C ring that gives rise to a 1,2A+ ion
with the methyl group (169 + 14)183), and the ion at m/z
287 is attributed to the loss of the hydroxyl in C3 position,
which was suggested to be the most labile and easily lost
(28). These results clearly showed that the methyl residue in
peaks 2 and 4 was located on the A-ring, that is, at either
positions 5 or 7.
Peaks 3 and 5 also showed the same MS2 spectra (Figure
5B) and fragmentation pattern. This is in agreement with
Donovan et al. (22) who also found that the fragmentation
pattern did not differ significantly for methylation at the 3′ or
4′ position. In these compounds the observation of the product

1236 J. Agric. Food Chem., Vol. 57, No. 4, 2009
Gonza´lez-Manzano et al.
Figure 6. MS2 fragmentation patterns of catechin methylethers (pseudomolecular ion at m/z 305) in positive ion mode.
Figure 7. MS2 fragmentation patterns of catechin glucuronides (pseudomolecular ion at m/z 465) in negative ion mode.
ions 1,2B+ at m/z 137 (123 + 14 ) 137) and 1,3A+ at m/z 179
(165 + 14 ) 179) demonstrated the location of the methyl
residue on the B-ring (i.e., positions 3′ or 4′). On the basis of
these results and the previous observations made by Donovan
et al. (22) and Cren-Olive´ et al. (28) regarding elution of the
catechin methylethers in RP-HPLC, peaks 2-5 in the chro-
matogram of the Figure 1 were assigned to 5-O-methylcatechin,
3′-O-methylcatechin, 7-O-methylcatechin, and 4′-O-methylcat-
echin, respectively.
Chemical synthesis of catechin glucuronides gave rise to a
mixture of five intermediate compounds showing a pseudomo-
lecular ion [M - H]- at m/z 605, corresponding to the different
triacetylglucuronic methylesters at each of the five hydroxyl
groups of the catechin. Further removal of the acetyl and methyl
residues produced a mixture of five catechin monoglucuronides
(Figure 2). All of them possessed a pseudomolecular ion [M
- H]- at m/z 465 that released a main MS2 fragment at m/z
289 (catechin; loss of the glucuronide moiety [(M - H) -
176]-). Secondary product ions were useful to establish the
location of the glucuronide moiety on the catechin and thus to
identify the different metabolites. As it happened for the catechin
methylethers, similar MS2 spectra were observed for compounds
1 and 4 (Figure 5C) and compounds 3 and 5 (Figure 5D),
indicating that the glucuronide moiety was located in either the
A- or B-ring in each case.
The fragmentation patterns of these compounds are shown in
Figure 7. Four secondary ions were observed for peaks 1 and 4 at
m/z 313 due to the RDA fission that gives rise to a 1,3A- ion with
the glucuronide moiety (137 + 176 ) 313), at m/z 343 that
corresponds to the 1,2 fission of the C releasing a 1,2A- ion with
the glucuronide moiety (167 + 176 ) 343), at m/z 447 that results
from the loss of the OH at C3, and at m/z 245 from the loss of
CO2 [(M - H) - 44]- from the ion at m/z 289; this loss of CO2
seems to be characteristic of the fragmentation of flavonoids in
negative ion mode (29), and the resulting ion (m/z at 245) was
observed as the main product ion in the MS3 spectra. These results
demonstrated that in these compounds the glucuronide residue was
located on the catechin A-ring.
Product ions observed for peaks 3 and 5 were at m/z 327,
corresponding to the RDA fission yielding the 1,3B- fragment

Catechin Sulfates, Glucuronides, and Methylethers
J. Agric. Food Chem., Vol. 57, No. 4, 2009 1237
Figure 8. MS2 fragmentation patterns of catechin monosulfates (pseudomolecular ion at m/z 369) in negative ion mode.
with the glucuronide moiety (151 + 176 ) 327), at m/z 447
(loss of the OH in C3 position), at m/z 245 (loss of CO2 from
catechin), and at m/z 381 attributed to the loss of the A-ring
(205 + 176)381). This last product ion is unusual in catechins,
although it was reported by Cren-Olive et al. (28) for methyl-
catechins. This fragmentation pattern revealed that in peaks 3
and 5 the glucuronide moiety was attached to catechin B-ring.
The MS2 spectrum of peak 2 showed only a relevant signal at
m/z 289 (catechin).
with that of the methylcatechins allowed the identification of
peaks 1, 2, 3, and 4 in the chromatogram of Figure 3 as
catechin-5-O-sulfate, catechin-3′-O- sulfate, catechin-7-O- sul-
fate, and catechin-4′-O-sulfate, respectively.
Conclusions. Procedures for the hemisynthesis of (epi)cat-
echin sulfates, glucuronides, and methylethers of metabolic
interest have been optimized. In the conditions used, phenolic
hydroxyl groups at the 3′ and 4′ positions in the catechin B-ring
were shown as the most reactive for conjugation, followed by
phenolic hydroxyls at positions 5 and 7 on the A-ring, while
hardly any derivatives at position 3 (C-ring) were produced.
Five catechin monoglucuronides, four catechin monosulfates,
and four catechin methylethers could be obtained and character-
ized by HPLC-DAD-MS. Identification of the compounds was
made on the basis of their MS2 fragmentation and elution
characteristics in RP-HPLC, thus establishing patterns for their
further analysis in biological fluids. The methodologies used
allowed the preparation of (epi)catechin sulfates, glucuronides,
and methylethers conjugated at positions 3′ and 4′, as well as
the sulfates at positions 5 and 7 with satisfactory yields for their
further isolation by semipreparative-HPLC in view of their use
in in vitro/ex vivo assays for their biological activity.
Taking into account these observations and the comments
already made regarding the elution profile of the methylcat-
echins, peaks 1-5 obtained in the synthesis of catechin
glucuronides were assigned to catechin-5-O-glucuronide, cat-
echin-3-O-glucuronide, catechin-3′-O-glucuronide, catechin-7-
O-glucuronide, and catechin-4′-O-glucuronide, respectively.
The chromatogram of the Figure 3 shows the result obtained
in the synthesis of catechin sulfates. The three peaks observed
possessed negative ions at m/z 449, m/z 369, and m/z 289, which
correspond to catechin disulfates, monosulfates, and aglycone,
respectively. Further fractionation on Sephadex LH-20 allowed
the separation of the disulfates (first compounds to elute) from
the monosulfates and catechin (last one to elute). Monosulfates
were collected and analyzed by HPLC-DAD-MS using TFA
as acid modifier for the elution as above commented. Four
compounds were detected (Figure 4) that showed a pseudo-
molecular ion [M - H]- at m/z 369 that released a main MS2
fragment at m/z 289 (catechin). Various secondary product ions
were also produced that were useful to establish the location of
the sulfate residue on the catechin. Compounds 1 and 3 showed
similar MS2 spectra (Figure 5E) and fragmentation pattern
(Figure 8). The ion at m/z 217 corresponds to 1,3A- ion with a
sulfate moiety (137 + 80) produced from RDA fission of the
parent metabolite, and the small signal at m/z 247 is associated
with the 1,2A- ion with the sulfate moiety (167 + 80). These
fragments clearly demonstrated that the sulfate moiety in
compounds 1 and 3 was linked to the catechin A-ring.
ACKNOWLEDGMENT
Financial support for this work was obtained from the Spanish
Ministerio de Ciencia e Innovacio´n (Projects AGL2004-06685-
C04-03 and AGL2007-66108-C04-02), Junta de Castilla y Leo´n
˜
(Project SA003A07), and University of Salamanca (ref KANB).
The research group belongs to the Consortium FUN-C-FOOD
(ref CSD2007-00063) funded by the Consolider-Ingenio 2010
Programme.
LITERATURE CITED
(1) Nakachi, K.; Suemasu, K.; Suga, K.; Takeo, T.; Imai, K.; Higashi,
Y. Influence of drinking green tea on breast cancer malignancy
among Japanese patients. Jpn. J. Cancer Res. 1998, 89, 254–261.
(2) Cooper, K. A.; Donovan, J. L.; Waterhouse, A. L.; Williamson,
G. Cocoa and health: a decade of research. Br. J. Nutr. 2008, 99,
1–11.
(3) Cook, N. C.; Samman, S. Flavonoids. Chemistry, metabolism,
cardioprotective effects and dietary sources. J. Nutr. Biochem.
1996, 7, 66–76.
Peaks 2 and 4 also showed similar MS2 spectra (Figure 5F)
with product ions that revealed that their sulfate residues were
linked to the catechin B-ring. In particular, ions at m/z 231 (151
+ 80) and 201 (121 + 80) correspond respectively to the 1,3B-
and 1,2B- fragments with the sulfate moiety. The ion at m/z 245
produced in all peaks is attributed to the loss of CO2 [(M - H)
- 44]- from the aglycone; this fragment was observed as the
main product ion in the MS3 spectra. These fragmentation
patterns and the expected HPLC elution behavior by analogy
(4) Williamson, G.; Manach, C. Bioavailability and bioefficacy of
polyphenols in humans. II. Review of 93 intervention studies.
Am. J. Clin. Nutr. 2005, 81, 243S–255S.

1238 J. Agric. Food Chem., Vol. 57, No. 4, 2009
Gonza´lez-Manzano et al.
(5) Hanasaki, Y.; Ogawar, S.; Fukui, S. The correlation between active
oxygens savenging and antioxidative effects of flavonoids. Free
Radical Biol. Med. 1994, 16, 845–850.
(6) Saito, M.; Hosoyama, H.; Ariga, T.; Kataoka, S.; Yamaji, N.
Antiulcer activity of grape seed extract and procyanidins. J. Agric.
Food Chem. 1998, 46, 1460–1464.
(7) Hope, W. C.; Welton, A. F.; Fiedler-Nagy, C.; Batula-Bernardo,
C.; Coffey, J. W. In vitro inhibition of the biosynthesis of slow
reacting substance of anaphylaxis (SRS-A) and lipoxygenase
activity by quercetin. J. Biochem. Pharmacol. 1983, 32, 367–
371.
(8) Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Structure-
antioxidant activity relationships of flavonoids and phenolic acids.
Free Radical Biol. Med. 1996, 20, 933–956.
(9) Haenen, G. R. M. M.; Bast, A. Nitric oxide radical scavenging
of flavonoids. Methods Enzymol. 1999, 301, 490–503.
(10) Baba, S.; Osakabe, N.; Yasuda, A.; Natsume, M.; Takizawa, T.;
Nakamura, T.; Terao, J. Bioavailability of (-)-epicatechin upon
intake of chocolate and cocoa in human volunteers. Free Radical
Res. 2000, 33, 635–641.
(19) Day, A. J.; Bao, Y. P.; Morgan, M. R. A.; Williamson, G.
Conjugation position of quercetin glucuronides and effect on
biological activity. Free Radical Biol. Med. 2000, 29, 1234–1243.
(20) Yoshizumi, M.; Tsuchiya, K.; Suzaki, Y.; Kirima, K.; Kyaw, M.;
Moon, J. H.; Terao, J.; Tamaki, T. Quercetin glucuronide prevents
VSMC hypertrophy by angiotensin II via the inhibition of JNK
and AP-1 signaling pathway. Biochem. Biophys. Res. Commun.
2002, 293, 1458–1465.
(21) Lodi, F.; Jimenez, R.; Menendez, C.; Needs, P. W.; Duarte, J.;
Perez-Vizcaino, F. Glucuronidated metabolites of the flavonoid
quercetin do not auto-oxidise, do not generate free radicals and
do not decrease nitric oxide bioavailability. Planta Med. 2008,
74, 741–746.
(22) Donovan, J. L.; Luthria, D. L.; Stremple, P.; Waterhouse, A. L.
Analysis of (+)-catechin, (-)-epicatechin and their 3′ and 4′
O-methylated analogs: a comparison of sensitive methods. J. Chro-
matogr., B. 1999, 726, 277–283.
(23) Tsushida, T.; Suzuki, M. Isolation of flavonoid-glycosides in
onion and identification by chemical synthesis of the glycosides.
Jpn. Food Sci. Technol. 1995, 42, 100–108.
(24) Jones, D. J. L.; Jukes-Jones, R.; Verschoyle, R. D.; Farmer, P. B.;
Gescher, A. A. A synthetic approach to the generation of quercetin
sulfates and the detection of quercetin 3′-O-sulfate as a urinary
metabolite in the rat. Bioorg. Med. Chem. 2005, 13, 6727–6731.
(25) Barron, D.; Cren-Olive´, C.; Needs, P. W. Chemical Synthesis of
Flavonoid Conjugates. In Methods in Polyphenol Analysis; Santos-
Buelga, C., Williamson, G., Eds.; Royal Society of Chemistry:
Cambridge, U.K., 2003; Vol. 18, pp 7-228.
(26) Cren-Olive, C.; Lebrun, S.; Rolando, C. An efficient synthesis of
the four mono methylated isomers of (+)-catechin including the
major metabolites and of some dimethylated and trimethylated
analogues through selective protection of the catechol ring.
J. Chem. Soc., Perkin Trans. 2002, 6, 821–830.
(27) Ma, Y. L.; Li, Q. M.; Van den Heuvel, H.; Claeys, M.
Characterization of flavone and flavonol aglycones by collision-
induced dissociation tandem mass spectrometry. Rapid Commun.
Mass Spectrom. 1997, 11, 1357–1364.
(28) Cren-Olive, C.; Deprez, S.; Lebrun, S.; Coddeville, B.; Rolando,
C. Characterization of methylation site of monomethylflavan-3-
ols by liquid chromatography/electrospray ionization tandem mass
spectrometry. Rapid Commun. Mass Spectrom. 2000, 14, 2312–
2319.
(11) Donovan, J. L.; Bell, J. R.; Kasim-Karakas, S.; German, J. B.;
Walzem, R. L.; Hansen, R. J.; Waterhouse, A. L. Catechin is
present as metabolites in human plasma after consumption of red
wine. J. Nutr. 1999, 129, 1662–1668.
(12) Baba, S.; Osakabe, N.; Natsume, M. In vivo comparison of the
biioavailability of (+)-catechin, (-)-epicatechin and their mixture
in orally administered rats. J. Nutr. 2001, 131, 2885–2891.
(13) Donovan, J. L.; Crespy, V.; Manach, C.; Morand, C.; Besson,
C.; Scalbert, A.; Remesy, C. Catechin is metabolized by both the
small intestine and liver of rats. J. Nutr. 2001, 131, 1753–1757.
(14) Kuhnle, G.; Spencer, J. P. E.; Schroeter, H.; Shenoy, B.; Debnam,
E. S.; Srai, S. K. S.; Rice-Evans, C.; Hahn, U. Epicatechin and
catechin are O-methylated and glucuronidated in the small
intestine. Biochem. Biophys. Res. Commun. 2000, 277, 507–512.
(15) Natsume, M.; Osakabe, N.; Oyama, M.; Sasaki, M.; Baba, S.;
Nakamura, Y.; Osawa, T.; Terao, J. Structures of (-)-epicatechin
glucuronide identified from plasma and urine after oral ingestion
of (-)-epicatechin: differences between human and rat. Free
Radical Biol. Med. 2003, 34, 840–849.
(16) Bharathi, J.; Walle, T. Glucuronidation and sulfation of the tea
flavonoid (-)-epicatechin by the human and rat enzymes. Drug
Metab. Dispos. 2002, 30, 897–903.
(17) Kroon, P. A.; Clifford, M. N.; Crozier, A.; Day, A. J.; Donovan,
J. L.; Manach, C.; Williamson, G. How should we assess the
effects of exposure to dietary polyphenols in vitro? Am. J. Clin.
Nutr. 2004, 80, 15–21.
(29) Fabre, N.; Rustan, I.; de Hoffmann, E.; Quetin-Leclercq, J.
Determination of flavone, flavonol, and flavanone aglycones by
negative ion liquid chromatography electrospray ion trap mass
spectrometry. J. Am. Soc. Mass Spectrom. 2001, 12, 707–715.
(18) Spencer, J. P. E.; El Mohsen, M. M. A.; Rice-Evans, C. Cellular
uptake and metabolism of flavonoids and their metabolites:
implications for their bioactivity. Arch. Biochem. Biophys. 2004,
423, 148–161.
Received for review October 8, 2008. Revised manuscript received
December 14, 2008. Accepted December 17, 2008.
JF803140H