Nonenzymatic C-glycosylation of flavan-3-ols by oligo- and polysaccharides

Article abstract of DOI:10.1021/jf0719508

Model reactions between the polysaccharide amylose and the polyphenol (-)-epicatechin followed by partial enzymatic hydrolysis of the reaction products formed led to the detection of mono- and oligo-C-glucosylated flavan-3-ols by means of LC-MS/MS experim

Full text of DOI:10.1021/jf0719508

J. Agric. Food Chem. 2007, 55, 9685–9697 9685
Nonenzymatic C-Glycosylation of Flavan-3-ols by
Oligo- and Polysaccharides
TIMO STARK,† DANIELA KELLER,† KERSTIN WENKER,† HEDDA HILLMANN,† AND
THOMAS HOFMANN*,#
Institut für Lebensmittelchemie, Universität Münster, Corrensstrasse 45 D-48149 Münster, Germany,
and Lehrstuhl für Lebensmittelchemie und Molekulare Sensorik, Technische Universität München,
Lise-Meitner-Straße 34, Germany
Model reactions between the polysaccharide amylose and the polyphenol (-)-epicatechin followed by
partial enzymatic hydrolysis of the reaction products formed led to the detection of mono- and oligo-C-
glucosylated flavan-3-ols by means of LC-MS/MS experiments. To confirm the structure of these putative
flavan-3-ol/oligosaccharide conjugates, (-)-epicatechin was reacted with maltose and maltotriose,
respectively, giving rise to a series of previously unreported flavan-3-ol/maltose and flavan-3-ol/maltotriose
conjugates, namely, (-)-epicatechin-8-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-D-glucopyranoside, (-)-catechin-
8-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-D-glucopyranoside, (-)-catechin-6-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-
D-glucopyranoside, (-)-catechin-8-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-D-glucopyranosyl-(4f1)- O-R-D-
glucopyranoside, (-)-catechin-6-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-D-glucopyranosyl-(4f1)-O-R-D-gluc-
opyranoside, and (-)-epicatechin-6/8-C-ꢀ-D-glucopyranosyl-(4f1)-O-R-D-glucopyranosyl-(4f1)-O-R-D-
glucopyranoside. Furthermore, quantitative analysis of flavan-3-ol-C-glucosides in an enzymatic total
hydrolysate using a newly developed stable isotope dilution assay (SIDA) enabled a first insight into the
yield of the formation of polyphenol/polysaccharide cross-links, for example, an amount of 14.0, 9.0, and
0.15 µmol of flavan-3-ol-6-C-ꢀ-D-glucopyranoside, flavan-3-ol-8-C-ꢀ-D-glucopyranoside, and flavan-3-ol-
6-C,8-C-ꢀ-D-glucopyranoside were per mmol (-)-epicatechin when reacted with amylose.
KEYWORDS: Flavan-3-ol; epicatechin; catechin; C-glycosylation; starch; polysaccharides; cross-links;
(-)-catechin-8-C-ꢀ-D-maltopyranoside
INTRODUCTION
the detection of a velvety, smoothly astringent-tasting polyphe-
nol fraction that was predominantly present in the alkalized
sample (6). LC-MS/MS analysis, 1D/2D-NMR, and CD spec-
troscopy, as well as model reactions led to the discovery and
unequivocal identification of these sensory active molecules in
cocoa as a series of flavan-3-ol-C-ꢀ-D-glycopyranosides (Figure
1), namely, (-)-catechin-8-C-ꢀ-D-glucopyranoside (1), (-)-
catechin-6-C,8-C-ꢀ-D-diglucopyranoside (2), (-)-catechin-6-C-
ꢀ-D-glucopyranoside (3), (-)-epicatechin-8-C-ꢀ-D-glucopyra-
noside (4), (-)-epicatechin-6-C,8-C-ꢀ-D-diglucopyranoside (5),
and (-)-epicatechin-6-C-ꢀ-D-glucopyranoside (6) (6). Besides
exhibiting a velvety astringent taste, these flavan-3-ol-C-
glycosides were demonstrated to modify the bitter taste profile
and decrease the bitter taste intensity of cocoa beverages as well
as theobromine solutions. Most surprisingly, model experiments
revealed that these polyphenol C-glycoconjugates are formed
upon cocoa alkalization via a novel “non-enzymatic C-glyco-
sylation” of flavan-3-ols (6).
Over the last decade, scientists and food manufacturers
have become increasingly interested in polyphenols due to
their great abundance in our diet, their various biological
functions, and their probable role in the prevention of diseases
associated with oxidative stress, such as cardiovascular and
neurodegenerative diseases (1). However, polyphenols are
highly reactive compounds and suitable substrates for numer-
ous enzymatic and chemical reactions during postharvest food
processing and storage such as condensation reactions
between flavanols and anthocyanins (2), aldehydes (3), and
ellagitannins (4), respectively, as well as redox reactions (5).
Although the occurrence of such reactions and their contribu-
tion to food quality are well documented, the chemical
structures of the reaction products formed by postharvest
polyphenol transformation are still poorly understood.
Very recently, application of a molecular sensory science
approach on alkalized and nonalkalized cocoa powder revealed
Although flavan-3-ols were shown to react nonenzymatically
with glucose (6), it is unclear whether these polyphenols are
able to react covalently with the reducing terminus of polysac-
charides such as amylose, thus giving rise to polysaccharide/
polyphenol conjugates. Therefore, the purpose of the present
* To whom correspondence should be addressed. Phone: +49-8161/
71-2902. Fax: +49-8161/71-2949. E-mail: thomas.hofmann@
wzw.tum.de.
† Universität Münster.
# Technische Universität München.
10.1021/jf0719508 CCC: $37.00
2007 American Chemical Society
Published on Web 10/12/2007

9686 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
Figure 1. Chemical structures of flavan-3-ol-C-glycosides 1–17.
MATERIALS AND METHODS
Chemicals. The following compounds were obtained com-
mercially: (-)-epicatechin, (+)-catechin (Sigma, Steinheim,
study was to answer the question as to whether reducing oligo-
and polysaccharides such as amylose can also bind covalently
to plant flavan-3-ols as has been shown for monohexoses.

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9687
Germany); potassium carbonate, D-glucose, D-maltose, D-
maltotriose (Merck, Darmstadt, Germany); amylose from maize
(Roth, Karlsruhe, Germany). Amyloglucosidase (from Aspergil-
lus niger 119.7 U) (Fluka, Taufkirchen, Germany) showed an
activity of 120 U/mg, where 1 U is the amount of enzyme
necessary to break down 1 µmol of glucose per min at pH 4.8
and 60 °C. The thermostable exo-R-amylase Termamyl 120L
(from Bacillus licheniformis) (Novo Nordisk, Denmark) showed
an activity of 120 KNU/g, where 1 KNU is the amount of
enzyme necessary to break down 5.26 g of starch/h under
standard conditions (Termamyl Product Data Sheet, Novo
Nordisk, 1995). Solvents were HPLC grade (Merck, Darmstadt).
Deuterated solvents were obtained from Euriso-Top (Gif-Sur-
Yvette, France). Bottled water (Evian) adjusted to pH 6.0 with
aqueous formic acid (1%) was used for sensory evaluation.
Flavan-3-ol -C-ꢀ-D-glucopyranosides 1–6 (Figure 1) were
synthesized and purified following the procedure reported
recently (6). Throughout the paper, D-glucopyranoside is ab-
breviated as D-glc p.
plus or minus one dilution step; that is, a threshold value of 1.4
µmol/Lfor(-)-epicatecin-8-C-ꢀ-D-glcp-(4f1)-O-R-D-glcp(4f1)-
O-R-D-glc p represents a range from 0.7 to 2.8 µmol/L.
Preparation and Characterization of Flavan-3-ol C-Poly-
glucopyranosides from Flavan-3-ols and Amylose. Prepara-
tion of FlaVan-3-ol-C-Polyglucopyranosides. (-)-Epicatechin
(1 mmol) was mixed with amylose (5 g), K2CO3 (5 mmol),
and water (30 mL), and then thermally treated for 10 min at 80
°C while stirring. After the sample was cooled, the reaction was
stopped by the addition of acetic acid until a pH of 7.0 was
reached.
Analysis of FlaVan-3-ol-C-Polyglucopyranosides. After ad-
dition of the Termamyl preparation (1 mL), the reacted
epicatechin/amylose mixture (pH 7.0) was kept in a boiling
water bath for 15 min and then cooled to room temperature.
The pH of the reaction mixture was adjusted to 4.6 by the
addition of acetic acid, amyloglucosidase (2 mL, 10 mg/mL)
was added, and the solution was incubated for 30 min at 60
°C. After cooling, the mixture was centrifuged (3000 rpm),
the supernatant was collected and concentrated in a vacuum
(aqueous fraction), and the insoluble residue was extracted
four times with aqueous ethanol (40%, 30 mL each) for 30
min at room temperature upon stirring. After centrifugation,
the ethanolic extract was freed from solvent under reduced
pressure at 40 °C and combined with the water fraction to
give a crude extract of the reaction mixture (10 mL). Aliquots
(2 mL) of the crude extract were applied onto the top of a
Strata Gigatube C18 cartridge (5g) (Phenomenex, Germany)
preconditioned with methanol, followed by water. Fraction-
ation was performed by flushing the cartridge with water (50
mL, fraction S1), followed by methanol (50 mL, fraction S2).
Solvent was removed from fraction S2 in vacuum, the residue
was taken up in a methanol/water mixture (1/9, v/v; 1 mL)
adjusted to pH 2.5 with aqueous formic acid (0.1% in water),
and then aliquots (5 µL) were analyzed by means of HPLC-
MS/MS, which was equipped with a 150 × 2 mm i.d., 5 µm,
RP Phenylhexyl column (Phenomenex, Germany) operated with
a flow rate of 0.2 mL/min. Chromatography was performed
starting isocratically with aqueous formic acid (0.1%, pH 2.5)
for 10 min, then the methanol content was increased to 40%
within 40 min, increased to 100% within another 10 min, and,
finally, held at 100% for 10 min.
Sensory Analyses. Training of the Sensory Panel. Twelve
subjects (seven women and five men, age 25–38 years), who
gave their informed consent to participate in the sensory tests
of the present investigation and had no history of known taste
disorders, were trained to evaluate the taste of aqueous solutions
(3 mL each) of the following standard taste compounds by using
a triangle test as described in the literature (7): sucrose (12.5
mmol/L) for sweet taste; lactic acid (20 mmol/L) for sour taste;
NaCl (12 mmol/L) for salty taste, caffeine (1 mmol/L) for bitter
taste, and sodium glutamate (3 mmol/L) for umami taste. For
the puckering astringency and the velvety astringent, mouth-
drying oral sensation, the panel was trained with aqueous
solutions of gallotannic acid (0.05%) and quercetin-3-O-ꢀ-D-
glc p (0.002 mmol/L), respectively, using the half-tongue test
(8). Sensory analyses were performed in a sensory panel room
at 22–25 °C in three different sessions.
Half-Tongue Test. To overcome carry-over effects of astrin-
gent compounds, threshold concentrations of astringent com-
pounds were determined in bottled water by means of the
recently developed half-tongue test (8). Serial 1:1 dilutions of
the samples were presented in order of increasing concentrations
to a trained panel of 12 persons in three different sessions, using
the sip-and-spit method. At the start of the session and before
each trial, the subject rinsed with water and expectorated. An
aliquot (1 mL) of the aqueous solution containing the astringent
compound was randomly applied with a pipette on one side of
the tongue, whereas pure water was applied on the other side
of the tongue as the control. The sensory panelists were then
asked to move their tongue forward and backward towards the
palate for 15 s and to identify the place of astringent sensation
by comparison of both sides. After indicating which part of the
tongue showed the typical astringent sensation induced by the
tastant, the participant rinsed with water and, after 10 min,
received another set of one blank and one taste-active sample.
To prevent excessive fatigue, tasting began at a concentration
level two steps below the threshold concentration that had been
determined in a preliminary taste experiment. Whenever the
panelist selected incorrectly, the next trial took place at the next
higher concentration step. When the panelist selected correctly,
the same concentration was presented again besides one blank
as a proof for the correctness of the data. The geometric mean
of the last and the second last concentration was calculated and
taken as the individual recognition threshold. Values between
individuals and three separate sessions differed not more than
Preparation of Flavan-3-ol -C-Oligoglucopyranosides (7–12,
Figure 1). Analytical Scale. A mixture of (-)-epicatechin (0.04
mmol), maltose, or maltotriose (0.04, 0.08, 0.2, 0.4, 0.8, or 4
mmol, respectively), K2CO3 (0.21 mmol), and water (1 mL)
was heated at 80 °C while stirring. After 10, 20, 40, and 60
min, the reaction was stopped by adjusting the pH to 5.0 using
aqueous hydrochloric acid (1 mol/L) and, after membrane
filtration, aliquots (10–50 µL) of the mixture were analyzed by
means of RP–HPLC–DAD.
PreparatiVe Scale. (-)-Epicatechin (2 mmol) was mixed with
maltose or maltotriose (40 mmol, each), K2CO3 (10.4 mmol),
and water (40 mL). After homogenization, the mixture was
heated for 10 min at 80 °C while stirring. After cooling, the
reaction was stopped by the addition of aqueous hydrochloric
acid (1 mol/L) until a pH of 5.0 was reached. Thereafter, the
reaction mixture was concentrated under reduced pressure at
40 °C to about 10 mL and was then applied onto the top of a
water-cooled 140 × 40 mm RP-18 column, LiChroprep, 25–40
µm (Merck, Darmstadt, Germany) conditioned with aqueous
formic acid (0.1% in water; pH 2.5). Chromatography was
performed using aqueous formic acid (0.1% in water; pH 2.5)
as the effluent, followed by aqueous formic acid (0.1% in water;

9688 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
pH 2.5) containing increasing amounts of methanol. Monitoring
the effluent at 270 nm, the fractions containing the title
compounds were freed from solvent in vacuum to give 10
subfractions (f1–f10), freeze-dried, and finally purified by means
of HPLC (Jasco, Groꢀ-Umstadt, Germany) consisting of a
HPLC-pump system PU 2087, a high-pressure gradient unit,
and a PU-2075 UV-detector using a preparative 21.2 × 250
mm, 5 µm, RP-18 column, HyperClone micro ODS (C18)
(Phenomenex, Aschaffenburg, Germany) as the stationary phase.
Subfractions f3–f6 were dissolved in a mixture (30/70, v/v) of
methanol and aqueous formic acid (0.1% in water; pH 2.5),
and after membrane filtration aliquots (0.5–2.0 mL) were
fractionated by RP-HPLC. Monitoring the effluent at 280 nm,
chromatography was performed with a mixture (5/95, v/v) of
methanol and aqueous formic acid (0.1% in water, pH 2.5) for
10 min, increasing the methanol content to 40% over 30 min,
and then to 100% over 5 min, thereafter, eluting with methanol
for 10 min at a flow rate of 18.0 mL/min. After the solvent was
removed in vacuum, the title compounds were suspended in
water (10 mL), and freeze-dried two times to afford the
corresponding flavan-3-ol-C-oligoglucopyranosides 7–12 (Fig-
ure 1) as white, amorphous powders in high purities of more
than 99%.
4.89 [m, 2H, J ) 9.2 Hz, H-C(2, 1′′)], 5.19 [d, 1H, J ) 3.2 Hz,
H-C(1′′′)], 6.02 [s, 1H, H-C(6)], 6.78 [d, 1H, J ) 8.0 Hz,
H-C(5′)], 6.83 [d, 1H, J ) 8.0 Hz, H-C(6′)], 7.08 [s, 1H,
H-C(2′)]; 13C NMR (100 MHz, MeOD, HMQC, HMBC): δ
29.8 [C(4)], 62.2 [C(6′′/6′′′)], 62.7 [C(6′′′/6′′)], 67.4 [C(3)], 71.5
[C(4′′′)], 74.3 [C(2′′′)], 74.8 [C(2′′, 4′′)], 75.1 [C(5′′)], 76.5
[C(1′′)], 79.8 [C(2)], 79.9 [C(3′′)], 80.7 [C(5′′)], 81.8 [C(4′′)],
97.1 [C(6)], 100.0 [C(4a)], 103.3 [C(1′′′)], 104.2 [C(8)], 115.1
[C(2′)], 116.1 [C(5′)], 118.8 [C(6′)], 132.5 [C(1′)], 145.6 [C(3′/
4′)], 146.1 [C(4′/3′)], 157.1 [C(8a)], 158.2 [C(5, 7)].
(-)-Catechin-6-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p, 9 (Figure
1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v pH 2,5) λmax ) 218,
231, 280 nm; exact mass: m/z 637.1730 (calc. for [C27H34O16
+ Na+] ) 637.1739), m/z 613.1778 (calc. for [C27H33O16]-
)
613.1763); MS (ESI-): m/z 613 (100% [M – 1]-), 451 (41%
[M – 163]-), 433 (28% [M – 181]-), 289 (15% [M – 325]-),
331 (10% [M – 283]-); MS/MS (ESI-) (-50 V): m/z 331 (100),
123 (55), 109 (40), 373 (8), 613 (1); 1H NMR (400 MHz,
MeOD, COSY): δ 2.51 [dd, 1H, J ) 8.0, 16.0 Hz, H-C(4R)],
2.87 [dd, 1H, J ) 5.2, 16.4 Hz, H-C(4ꢀ)], 3.29 [pt, 1H, J )
9.2 Hz, H-C(4′′′)], 3.47 [dd, 1H, J ) 3.6, 9.6 Hz, H-C(2′′′)],
3.51 [m, 1H, H-C(5′′)], 3.63–3.75 [m, 6H, J ) 8.8, 10.0 Hz,
H-C(2′′, 3′′, 3′′′, 4′′, 5′′′, 6R′′)], 3.84–3.95 [m, 3H, J ) 3.2,
4.8, 10.4 Hz, H-C(6ꢀ′′, 6R′′′, 6ꢀ′′′)], 3.98 [ddd, 1H, J ) 5.6,
6.8, 8.0 Hz, H-C(3)], 4.57 [d, 1H, J ) 7.2 Hz, H-C(2)], 4.88
[d, 1H, J ) 8.6 Hz, H-C(1′′)], 5.24 [d, 1H, J ) 3.6 Hz,
H-C(1′′′)], 5.96 [s, 1H, H-C(8)], 6.71 [dd, 1H, J ) 1.2, 8.0 Hz,
H-C(6′)], 6.76 [d, 1H, J ) 8.0 Hz, H-C(5′)], 6.82 [d, 1H, J )
1.2 Hz, H-C(2′)]; 13C NMR (100 MHz, MeOD, HMQC,
HMBC): δ 28.5 [C(4)], 61.7 [C(6′′/6′′′)], 62.8 [C(6′′′/6′′)], 68.7
[C(3)], 71.6 [C(4′′′)], 74.1 [C(2′′)], 74.3 [C(2′′′)], 74.9 [C(3′′′/
5′′′)], 75.2 [C(5′′′/3′′′)], 77.3 [C(1′′)], 79.4 [C(3′′)], 81.0 [C(4′′,
5′′)], 82.8 [C(2)], 96.3 [C(8)], 102.2 [C(4a)], 103.0 [C(1′′′)],
105.3 [C(6)], 115.2 [C(2′)], 116.1 [C(5′)], 120.0 [C(6′)], 132.1
[C(1′)], 146.28 [C(3′/4′)], 146.30 [C(4′/3′)], 156.11 [C(5/7)],
156.13 [C(7/5)], 156.4 [C(8a)].
(-)-Catechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p, 7 (Figure
1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v, pH 2,5) λmax
)
218, 231, 280 nm; exact mass: m/z 637.1730 (calc. for
[C27H34O16 + Na]+ ) 637.1739), m/z 613.1778 (calc. for
[C27H33O16]- ) 613.1763); MS (ESI-): m/z 613 (100% [M –
1]-), 451 (40% [M – 163]-), 433 (30% [M – 181]-), 289 (20%
[M – 325]-), 331 (12% [M – 283]-); MS/MS (ESI-) (-50 V):
1
m/z 331 (100), 123 (55), 109 (40), 373 (8), 613 (1); H NMR
(400 MHz, MeOD, COSY): δ 2.55 [dd, 1H, J ) 7.0, 16.4 Hz,
H-C(4R)], 2.74 [dd, 1H, J ) 5.2, 16.4 Hz, H-C(4ꢀ)], 3.27 [pt,
1H, J ) 9.2 Hz, H-C(4′′′)], 3.42–3.46 [m, 2H, J ) 3.6, 4.0,
9.2, 9.6 Hz, H-C(5′′, 2′′)], 3.59–3.71 [m, 5H, J ) 8.0, 8.4, 10.0,
10.4 Hz, H-C(3′′, 3′′′, 4′′, 5′′′, 6R′′)], 3.81–3.89 [m, 3H, J )
2.0, 12.0 Hz, H-C(6ꢀ′′, 6R′′′, 6ꢀ′′′)], 3.99 [ddd, 1H, J ) 5.6,
6.4, 6.6 Hz, H-C(3)], 4.10 [dd, 1H, J ) 8.8, 9.2 Hz, H-C(2′′)],
4.76 [d, 1H, J ) 6.4 Hz, H-C(2)], 4.84 [d, 1H, J ) 9.2 Hz,
H-C(1′′)], 5.16 [d, 1H, J ) 3.6 Hz, H-C(1′′′)], 6.01 [s, 1H,
H-C(6)], 6.71 [dd, 1H, J ) 1.6, 8.0 Hz, H-C(6′)], 6.75 [d, 1H,
J ) 8.0 Hz, H-C(5′)], 6.83 [d, 1H, J ) 1.6 Hz, H-C(2′)]; 13C
NMR (100 MHz, MeOD, HMQC, HMBC): δ 27.0 [C(4)], 62.3
[C(6′′/6′′′)], 62.7 [C(6′′′/6′′)], 68.5 [C(3)], 71.5 [C(4′′′)], 73.0
[C(2′′)], 74.3 [C(2′′′)], 74.7 [C(5′′′/3′′′)], 75.1 [C(5′′′/3′′′)], 76.4
[C(1′′)], 79.8 [C(3′′)], 80.9 [C(5′′)], 81.6 [C(4′′)], 82.4 [C(2)],
97.1 [C(6)], 101.0 [C(4a)], 103.0 [C(1′′′)], 104.1 [C(8)], 114.7
[C(2′)], 116.2 [C(5′)], 119.5 [C(6′)], 132.4 [C(1′)], 146.1 [C(3′/
4′)], 146.2 [C(4′/3′)], 155.2 [C(8a)], 157.1 [C(5/7)], 157.6 [C(7/
5)].
(-)-Catechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)-O-R-
D-glc p, 10 (Figure 1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v,
pH 2.5) λmax ) 218, 231, 280 nm; exact mass: m/z 775.2314
(calc. for [C33H43O21]- ) 775.2291); MS (ESI-): m/z 775
(100% [M – 1]-), 331 (33% [M – 445]-), 289 (20% [M –
487]-), 451 (12% [M – 325]-); MS/MS (ESI-) (-50 V): m/z
1
331 (100), 109 (40), 123 (27), 203 (26); H NMR (400 MHz,
D2O, COSY): δ 2.61 [dd, 1H, J ) 8.4, 16.0 Hz, H-C(4R)], 2.96
[dd, 1H, J ) 5.2, 16.0 Hz, H-C(4ꢀ)], 3.47 [pt, 1H, J ) 9.2 Hz,
H-C(4′′′′)], 3.53–3.55 [m, 2H, H-C(5′′, 5′′′)], 3.60–3.96 [m, 14H,
J ) 2.8, 3.6, 4.0, 8.4, 8.8, 9.2, 10.8 Hz, H-C(2′′′, 2′′′′, 3′′, 3′′′,
3′′′′, 4′′, 4′′′, 5′′′′, 6R′′, 6R′′′, 6R′′′′, 6ꢀ′′, 6ꢀ′′′, 6ꢀ′′′′)], 4.10
[ddd, 1H, J ) 5.6, 6.8, 7.2 Hz, H-C(3)], 4.21 [pt, 1H, J ) 8.8
Hz, H-C(2′′)], 4.84 [d, 1H, J ) 7.2 Hz, H-C(2)], 4.95 [d, 1H,
J ) 8.8 Hz, H-C(1′′)], 5.31 [d, 1H, J ) 4.0 Hz, H-C(1′′′/1′′′′)],
5.41 [d, 1H, J ) 4.0 Hz, H-C(1′′′′/1′′′)], 6.18 [s, 1H, H-C(6)],
6.89 [s, 2H, H-C(6′, 5′)], 6.97 [s, 1H, H-C(2′)];13C NMR (100
MHz, D2O, HMQC, HMBC): δ 27.0 [C(4)], 60.3 [C(6′′/6′′′/
6′′′′)], 60.4 [C(6′′/6′′′/6′′′′)], 67.0 [C(3)], 69.3 [C(4′′′′)], 71.0
[C(2′′)], 71.4, 71.6, 72.9, 73.0, 73.4 [7C, C(3′′, 3′′′, 3′′′′, 4′′′,
5′′′′, 2′′′, 2′′′′)], 74.2 [C(1′′)], 76.8 [C(5′′′)], 78.0 [C(4′′)], 78.7
[C(5′′)], 80.8 [C(2)], 96.2 [C(6)], 99.9 [C(1′′′/1′′′′)], 100.2
[C(1′′′′/1′′′)], 100.7 [C(4a)], 102.4 [C(8)], 114.3 [C(2′)],
115.9 [C(5′)], 119.8 [C(6′)], 130.8 [C(1′)], 144.0 [C(3′, 4′)],
154.0 [C(8a)], 154.9 [C(5, 7)].
(-)-Epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p, 8 (Figure
1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v, pH 2,5) λmax
)
218, 231, 280 nm; exact mass: m/z 637.1730 (calc. for
[C27H34O16 + Na+] ) 637.1739), m/z 613.1778 (calc. for
[C27H33O16]- ) 613.1763); MS (ESI-): m/z 613 (100% [M –
1]-), 451 (37% [M – 163]-), 433 (31% [M – 181]-), 289 (18%
[M – 325]-), 331 (17% [M – 283]-); MS/MS (ESI-) (-50 V):
1
m/z 331 (100), 123 (55), 109 (40), 373 (8), 613 (1); H NMR
(400 MHz, MeOD, COSY): δ 2.83 [dd, 1H, J ) 1.6, 17.2 Hz,
H-C(4R)], 2.88 [dd, 2H, J ) 4.0, 16.8 Hz, H-C(4ꢀ)], 3.29 [pt,
1H, J ) 9.2 Hz, H-C(4′′′)], 3.45–3.48 [m, 2H, J ) 3.6, 9.6 Hz,
H-C(5′′, 2′′′)], 3.63–3.74 [m, 6H, J ) 8.4, 9.2, 9.6 Hz, H-C(2′′,
3′′, 3′′′, 4′′, 5′′′, 6R′′)], 3.83–3.95 [m, 3H, J ) 2.8, 3.6, 11.6,
12.0, 12.4 Hz, H-C(6ꢀ′′, 6R′′′, 6ꢀ′′′)], 4.16 [br s, 1H, H-C(3)],
(-)-Epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)- O-R-
D-glc p, 11 (Figure 1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v,
pH 2.5) λmax ) 218, 231, 280 nm; exact mass: m/z 775.2314

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9689
(calc. for [C33H43O21]- ) 775.2291); MS (ESI-): m/z 775
(100% [M – 1]-), 331 (35% [M – 445]-), 289 (23% [M –
487]-), 451 (13% [M – 325]-); MS/MS (ESI-) (-50 V): m/z
[C(6′′/ 6′′′)], 62.7 [C(6′′/ 6′′′)], 68.2 [C(3)], 70.8 [C(4′′/ 4′′′)],
71.4 [C(4′′/ 4′′′)], 74.6 [C(2′′, 2′′′)], 76.0 [C(1′′/1′′′)], 77.4 [C(1′′/
1′′′)], 79.2 [C(3′′/3′′′)], 82.1 [C(5′′/5′′′)], 82.3 [C(2)], 82.4 [C(5′′/
5′′′)], 100.8 [C(4a)], 103.5 [C(6/8)], 104.1 [C(6/8)], 114.9
[C(2′)], 115.9 [C(5′)], 119.4 [C(6′)], 130.7 [C(1′)], 144.6
[C(3′,4′)], 152.8 [C(8a)], 153.4 [C(5)], 154.0 [C(7)].
1
331 (100), 123 (80), 109 (51); H NMR (400 MHz, MeOD,
COSY): δ 2.82 [dd, 1H, J ) 2.0, 17.6 Hz, H-C(4R)], 2.88 [dd,
1H, J)4.0, 16.7 Hz, H-C(4ꢀ)], 3.27 [pt, 1H, J ) 9.6 Hz,
H-C(4′′′′)], 3.43–3.92 [m, 17H, J ) 1.2, 2.8, 3.6, 8.8, 9.2, 9.6,
12.4, 12.8 Hz, H-C(2′′, 2′′′, 2′′′′, 3′′, 3′′′, 3′′′′, 4′′, 4′′′, 5′′, 5′′′,5′′′′,
6R′′, 6R′′′, 6R′′′′, 6ꢀ′′, 6ꢀ′′′, 6ꢀ′′′′)], 4.16 [s, 1H, H-C(3)], 4.89
[d, 1H, J)9.6 Hz, H-C(1′′)], 4.90 [s, 1H, H-C(2)], 5.16 [d, 1H,
J ) 4.0 Hz, H-C(1′′′/1′′′′)], 5.20 [d, 1H, J ) 3.6 Hz, H-C(1′′′/
1′′′′)], 6.02 [s, 1H, H-C(6)], 6.78 [d, 1H, J ) 8.2 Hz, H-C(5′)],
6.84 [dd, 1H, J)2.0, 8.2 Hz, H-C(6′)], 7.09 [s, 1H, H-C(2′)].
(+)-Catechin-6-C-ꢀ-D-glcp, 14 (Figure 1): UV/vis (MeOH)
λmax ) 220, 241, 273 nm; MS (ESI-): m/z 331 (100% [M –
121]-), 451 (38% [M – 1]-); MS/MS (ESI-): 123 (100% [M
– 329]-), 331 (99% [M – 121]-), 109 (67% [M – 343]-), 149
(34% [M – 303]-); exact mass: m/z 451.1243 (calculated for
-
1
C21H23O11 451.1246); H NMR (400 MHz, MeOD, COSY):
δ 2.55 [dd, 1H, J ) 7.2, 7.6, 16.0 Hz, H-C(4R)], 2.79 [dd, 1H,
J ) 5.2, 16.0 Hz, H-C(4ꢀ)], 3.42 [m, 1H, J ) 2.0, 4.0 Hz,
H-C(5′′)], 3.48 [dd, 1H, J ) 8.4, 8.8 Hz, H-C(4′′)], 3.52 [dd,
1H, J ) 8.8, 9.2 Hz, H-C(3′′)], 3.66 [dd, 1H, J ) 8.8, 9.6 Hz,
H-C(2′′)], 3.79 [dd, 1H, J ) 4.4, 12.0 Hz, H-C(6R′′)], 3.86 [dd,
1H, J ) 2.4, 12.0 Hz, H-C(6ꢀ′′)], 4.00 [ddd, 1H, J ) 5.2, 5.6,
7.2, 7.6 Hz, H-C(3)], 4.62 [d, 1H, J ) 7.2 Hz, H-C(2)], 4.87
[d, 1H, J ) 9.6 Hz, H-C(1′′)], 5.96 [s, 1H, H-C(8)], 6.69 [dd,
1H, J ) 2.0, 8.0 Hz, H-C(6′)], 6.74 [d, 1H, J ) 8.0 Hz, H-C(5′)],
6.80 [d, 1H, J ) 2.0 Hz, H-C(2′)]; 13C NMR (100 MHz, MeOD,
HMQC, HMBC): δ 27.4 [C(4)], 61.7 [C(6′′)], 68.2 [C(3)], 70.8
[C(4′′)], 74.2 [C(2′′)], 77.1 [C(1′′)], 79.2 [C(3′′)], 82.1 [C(5′′)],
82.3 [C(2)], 96.0 [C(8)], 101.9 [C(4a)], 105.4 [C(6)], 114.8
[C(2′)], 115.6 [C(5′)], 119.5 [C(6′)], 132.0 [C(1′)], 145.0 [C(3′/
4′)], 146.0 [C(4′/3′)], 156.0 [C(8a)], 155.8 [C(5, 7)].
(-)-Catechin-6-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)-O-R-
D-glc p, 12 (Figure 1): UV/vis (MeOH/0.1% HCOOH, 3/7, v/v
pH 2.5) λmax ) 218, 231, 280 nm; exact mass: m/z 775.2314
(calc. for [C33H43O21]- ) 775.2291); MS (ESI-): m/z 775
(100% [M – 1]-), 331 (29% [M – 445]-), 289 (21% [M –
487]-), 451 (8% [M – 325]-); MS/MS (ESI-) (-50 V): m/z
1
331 (100), 109 (43), 123 (24), 203 (20); H NMR (400 MHz,
D2O, COSY): δ 2.59 [dd, 1H, J ) 8.0, 16.0 Hz, H-C(4R)], 2.95
[dd, 1H, J ) 5.6, 16.0 Hz, H-C(4ꢀ)], 3.46 [dd, 1H, J ) 9.2, 9.6
Hz, H-C(4′′′′)], 3.62 [dd, 1H, J ) 4.0, 10.0 Hz, H-C(2′′′′)],
3.66–3.71 [m, 2H, J ) 4.8, 5.2, 9.6 Hz, H-C(5′′, 5′′′)], 3.72
[dd, 1H, J ) 4.0, 9.6 Hz, H-C(2′′′)], 3.75–4.09 [m, 12H, J )
2.0, 4.8, 8.4, 8.8, 9.2, 10.4, 12.4 Hz, H-C(3′′, 3′′′, 3′′′′, 4′′, 4′′′,
5′′′′, 6R′′, 6R′′′, 6R′′′′, 6ꢀ′′, 6ꢀ′′′, 6ꢀ′′′′)], 4.03 [dd, 1H, J )
9.2, 9.6 Hz, H-C(2′′)], 4.20 [ddd, 1H, J ) 5.6, 6.8, 7.6 Hz,
H-C(3)], 4.77 [d, 1H, J ) 6.8 Hz, H-C(2)], 4.93 [d, 1H, J )
8.8 Hz, H-C(1′′)], 5.43 [d, 1H, J)4.0 Hz, H-C(1′′′′)], 5.46 [d,
1H, J ) 4.0 Hz, H-C(1′′′)], 6.12 [s, 1H, H-C(8)], 6.89 [d, 1H,
J ) 8.0 Hz, H-C(6′)], 6.97 [d, 1H, J ) 8.0 Hz, H-C(5′)], 6.98
[s, 1H, H-C(2′)]; 13C NMR (100 MHz, D2O, HMQC, HMBC):
δ 26.5 [C(4)], 60.4 [C(6′′, 6′′′, 6′′′′)], 66.5 [C(3)], 69.3 [C(4′′′′)],
71.1 [C(3′′′)], 71.4 [C(2′′, 2′′′)], 71.6 [C(2′′′′)], 72.6, 72.9, 73.2
[4C, C(3′′, 3′′′′, 4′′′, 5′′′′)], 74.8 [C(1′′)], 77.1 [C(5′′′)], 77.6
[C(4′′)], 78.7 [C(5′′)], 80.8 [C(2)], 95.8 [C(8)], 99.8 [C(1′′′,
1′′′′)], 101.8 [C(4a)], 104.2 [C(6)], 114.9 [C(2′)], 115.7 [C(5′)],
119.9 [C(6′)], 130.4 [C(1′)], 144.2 [C(3′/4′)], 144.4 [C(4′/3′)],
154.0 [C(5, 7)], 154.8 [C(8a)].
(+)-Catechin-8-C-ꢀ-D-glcp, 15 (Figure 1): UV/vis (MeOH)
λmax ) 220, 241, 273 nm; MS (ESI-): m/z 331 (100% [M –
121]-), 451 (72% [M – 1]-); MS/MS (ESI-): 331 (100% [M
– 121]-), 123 (71% [M – 329]-), 109 (70% [M – 343]-); exact
-
1
mass: m/z 451.1243 (calculated for C21H23O11 451.1246); H
NMR (400 MHz, MeOD, COSY): δ 2.55 [dd, 1H, J ) 7.6,
16.0 Hz, H-C(4R)], 2.79 [dd, 1H, J ) 5.2, 16.0 Hz, H-C(4ꢀ)],
3.35 [m, 1H, J ) 2.8, 3.6 Hz, H-C(5′′)], 3.39 [m, 2H, J ) 8.4,
9.2 Hz, H-C(3′′,4′′)], 3.72 [dd, 1H, J ) 5.2, 12.0 Hz, H-C(6R′′)],
3.86 [dd, 1H, J ) 2.0, 12.0 Hz, H-C(6ꢀ′′)], 4.00 [ddd, 1H, J )
5.6, 7.2 Hz, H-C(3)], 4.12 [dd, 1H, J ) 8.8, 9.0 Hz, H-C(2′′)],
4.72 [d, 1H, J ) 6.8 Hz, H-C(2)], 4.82 [d, 1H, J) 9.6 Hz,
H-C(1′′)], 6.01 [s, 1H, H-C(6)], 6.76 [d, 1H, J ) 8.0 Hz,
H-C(5′)], 6.80 [d, 1H, J ) 1.6, 8.0 Hz, H-C(6′)], 7.01 [d, 1H,
J ) 1.6 Hz, H-C(2′)]; 13C NMR (100 MHz, MeOD, HMQC,
HMBC): δ 27.4 [C(4)], 62.5 [C(6′′)], 68.2 [C(3)], 71.7 [C(4′′)],
72.9 [C(2′′)], 76.3 [C(1′′)], 79.7 [C(3′′)], 81.8 [C(5′′)], 81.9
[C(2)], 96.5 [C(6)], 99.4 [C(4a)], 103.8 [C(8)], 114.9 [C(2′)],
115.9 [C(5′)], 119.3 [C(6′)], 131.2 [C(1′)], 144.6 [C(3′,4′)], 154.8
[C(8a)], 155.9 [C(5, 7)].
Preparation of (+)-Flavan-3-ol-C-glucopyranosides 13–17
(Figure 4). Compounds 13–17 were prepared by reacting (+)-
catechin and D-glucose in aqueous K2CO3 solution for 10 min
at 80 °C following exactly the procedure reported recently for
the (-)-epicatechin (6).
(+)-Catechin-6-C,8-C-ꢀ-D-diglcp, 13 (Figure 1): UV/vis
(MeOH) λmax ) 220, 241, 273 nm; MS (ESI-): m/z 613 (100%
[M – 1]-), 373 (59% [M – 241]-), 493 (35% [M – 121]-), 331
(10% [M – 283]-); MS/MS (ESI-): 373 (100% [M – 124]-),
109 (96% [M-505]-), 123 (82% [M – 491]-), 129 (63% [M –
465]-), 331 (10% [M – 283]-); exact mass: m/z 613.1784
(+)-Epicatechin-6-C-ꢀ-D-glcp, 16 (Figure 1): UV/vis (MeOH)
λmax ) 220, 241, 273 nm; MS (ESI-): m/z 331 (100% [M –
121]-), 451 (75% [M – 1]-); MS/MS (ESI-): 331 (100% [M
– 121]-), 123 (75% [M – 329]-), 109 (64% [M – 343]-), 149
(30% [M – 303]-); exact mass: m/z 451.1243 (calculated for
-
1
(calculated for C27H33O16 613.1773); H NMR (400 MHz,
MeOD, COSY): δ 2.58 [dd, 1H, J ) 7.2, 16.4 Hz, H-C(4R)],
2.77 [dd, 1H, J ) 5.2, 16.4 Hz, H-C(4ꢀ)], 3.31 [m, 1H, H-C(5′′/
5′′′)], 3.43 [m, 3H, J ) 9.2 H-C(4′′/4′′′, 3′′/3′′′, 5′′/5′′′)], 3.52
[m, 2H, J ) 9.2 Hz, H-C(4′′/4′′′, 3′′/3′′′)], 3.63 [dd, 1H, J )
9.0, 9.6 Hz, H-C(2′′)], 3.74 [dd, 1H, J ) 4.8, 12.0 Hz,
H-C(6R′′)], 3.81 [m, 3H, J ) 2.0, 2.4, 4.4, Hz, H-C(6ꢀ′′/6R′′′/
6ꢀ′′′)], 3.91 [dd, 1H, J ) 9.6 Hz, H-C(2′′′)], 4.00 [ddd, 1H, J
) 5.2, 6.4, 6.8 Hz, H-C(3)], 4.72 [dd, 1H, J ) 6.8 Hz, H-C(2)],
4.87 [d, 1H, J ) 9.6 Hz, H-C(1′′′)], 4.90 [d, 1H, J ) 9.6 Hz,
H-C(1′′)], 6.74 [d, 1H, J ) 8.4 Hz, H-C(5′)], 6.80 [dd, 1H, J )
1.8, 8.4 Hz, H-C(6′)], 6.96 [d, 1H, J ) 1.8 Hz, H-C(2′)] 13C
NMR (100 MHz, MeOD, HMQC, HMBC): δ 27.6 [C(4)], 61.7
-
1
C21H23O11 451.1246); H NMR (400 MHz, MeOD, COSY):
δ 2.76 [dd, 1H, J ) 2.4, 16.4 Hz, H-C(4R)], 2.86 [dd, 1H, J )
4.4, 16.8 Hz, H-C(4ꢀ)], 3.39 [m, 3H, J ) 2.4, 4.0 Hz, H-C(5′′)],
3.51 [m, 2H, J ) 8.8, 9.2, 9.6 Hz, H-C(3′′,5′′)], 3.67 [dd, 1H,
J ) 8.8, 9.6 Hz, H-C(2′′)], 3.80 [dd, 1H, J ) 4.4, 12.0 Hz,
H-C(6R′′)], 3.85 [dd, 1H, J ) 2.4, 12.0 Hz, H-C(6ꢀ′′)], 4.18
[m, 1H, J ) 2.4, 4.4 Hz, H-C(3)], 4.80 [s, 1H, H-C(2)], 4.86
[d, 1H, J ) 10.2 Hz, H-C(1′′)], 6.00 [s, 1H, H-C(8)], 6.75 [d,
1H, J ) 8.4 Hz, H-C(5′)], 6.80 [dd, 1H, J ) 2.0, 8.4 Hz,
H-C(6′)], 6.97 [d, 1H, J ) 2.0 Hz, H-C(2′)]; 13C NMR (100
MHz, MeOD, HMQC): δ 29.4 [C(4)], 62.0 [C(6′′)], 67.3 [C(3)],
71.3 [C(4′′)], 74.3 [C(2′′)], 77.3 [C(1′′)], 79.6 [C(3′′)], 79.7

9690 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
[C(2)], 82.0 [C(5′′)], 96.6 [C(8)], 115.2 [C(2′)], 116.0 [C(5′)],
119.2 [C(6′)].
mixture was adjusted to pH 7.0 by the addition of aqueous acetic
acid. To measure the catechin- and epicatechin-C-glycosides
1–6 as the flavan-3-ol/carbohydrate linkages, the (-)-epicat-
echin/amylose reaction mixture was spiked with solutions (20
µL, each) of the labeled internal standards 18–20, dissolved in
methanol (1.0 mg/mL), then vortexed and equilibrated for 30
min at room temperature. After addition of Termamyl (100 µL),
the mixture was kept in a boiling water bath for 60 min and
was then cooled to room temperature. The pH of the solution
was adjusted to 4.6 by adding aqueous acetic acid, amyloglu-
cosidase (0.3 mL, 10 mg/mL) was added, and the solution was
maintained for 3 h at 60 °C while stirring. After centrifugation,
the liquid supernatant was collected (water fraction), and the
insoluble residue was extracted four times with aqueous ethanol
(40%, 10 mL each) for 30 min at room temperature with stirring.
The ethanolic extract was freed from solvent under reduced
pressure at 40 °C, combined with the water fraction, and freeze-
dried to give the ethanolic/water extract. Aliquots (∼50 mg) of
the ethanol/water extract were taken up in a methanol/water
mixture (1/9, v/v; 1 mL), which was acidified to pH 2.5 with
aqueous formic acid (0.1% in water). After membrane filtration,
aliquots (5 µL) were analyzed by means of LC-MS/MS using
the same conditions as described above. The labeled compounds
were used as the internal standards for the quantitative analysis
of the corresponding analytes. In addition, compound 4 was
quantified via 18, compound 5 via 19, and compound 6 via 20
as the internal standard. The amounts of the individual flavan-
3-ol-C-glycosides were calculated using the response factors
(given in parentheses) for the individual compounds 1 (0.97),
2 (0.96), 3 (0.98), 4 (0.91), 5 (0.92), and 6 (0.94), which had
been determined by analysis of solutions containing defined
amounts of the internal standards and the target compounds in
five mass ratios from 0.2 to 5.0.
(+)-Epicatechin-8-C-ꢀ-D-glcp, 17 (Figure 1): UV/vis (MeOH)
λmax ) 220, 241, 273 nm; MS (ESI-): m/z 451 (100% [M –
1]-), 331 (95% [M – 121]-); MS/MS (ESI-): 331 (100% [M
– 121]-), 123 (67% [M – 329]-), 109 (63% [M – 343]-), 149
(31% [M – 303]-); exact mass: m/z 451.1243 (calculated for
-
1
C21H23O11 451.1246); H NMR (400 MHz, MeOD, COSY):
δ 2.78 [dd, 1H, J ) 1.6, 16.8 Hz, H-C(4R)], 2.89 [dd, 1H, J )
4.4, 16.8 Hz, H-C(4ꢀ)], 3.40 [m, 3H, J ) 8.4, 9.2 Hz,
H-C(3′′,4′′,5′′)], 3.65 [dd, 1H, J ) 4.8, 12.0 Hz, H-C(6R′′)],
3.85 [dd, 1H, J ) 1.6, 12.0 Hz, H-C(6ꢀ′′)], 4.23 [m, 1H, J )
1.6, 4.4 Hz, H-C(3)], 4.27 [br m, 1H, H-C(2′′)], 4.85 [d, 1H, J
) 9.2 Hz, H-C(1′′)], 4.88 [s, 1H, H-C(2)], 6.02 [s, 1H, H-C(6)],
6.78 [d, 1H, J ) 8.0 Hz, H-C(5′)], 6.82 [dd, 1H, J ) 1.6, 8.0
Hz, H-C(6′], 7.04 [d, 1H, J ) 2.0 Hz, H-C(2′)]; 13C NMR (100
MHz, MeOD, HMQC, HMBC): δ 29.4 [C(4)], 62.7 [C(6′′)],
66.7 [C(3)], 71.8 [C(4′′)], 72.4 [C(2′′)], 76.0 [C(1′′)], 79.6 [C(2)],
79.6 [C(3′′)], 82.0 [C(5′′)], 96.4 [C(6)], 99.2 [C(4a)], 102.6
[C(8)], 115.0 [C(2′)], 115.9 [C(5′)], 119.4 [C(6′)], 130.8 [C(1′)],
144.2 [C(4′/3′)], 144.6 [C(3′/4′)],154.2 [C(8a)], 155.2 [C(5)],
156.9 [C(7)].
Preparation of [13C6]-Labeled Flavan-3-ol-C-glucopyra-
nosides 18–20. A mixture of (-)-epicatechin (0.55 mmol),
[13C6]-D-glucose (5.55 mmol), K2CO3 (3.0 mmol), and water
(7.1 g) were heated for 10 min at 80 °C while stirring. After
cooling, the title compounds (-)-catechin-8-C-[13C6]-ꢀ-D-glu-
copyranoside (18), (-)-catechin-6,8-C-di-[13C6]-ꢀ-D-glucopy-
ranoside (19), and (-)-catechin-6-C-[13C6]-ꢀ-D-glucopyranoside
(20) were isolated and purified as detailed above for the
corresponding nonlabelled analogues.
(-)-Catechin-8-C-[13C6]-ꢀ-D-glcp, 18. UV/vis (MeOH/0.1%
HCOOH, pH 2.5; 2/8, v/v) λmax ) 218, 231, 280 nm; MS
(ESI-): m/z 457 (100%, [M – 1]-), 333 (45%, [M – 125]-);
MS/MS (ESI-) (-50 V): m/z 333 (100), 123 (57), 109 (27),
457 (2); 13C NMR (100 MHz, MeOD) δ 62.8 [d, J ) 42.7 Hz,
C-6′′], 71.7 [dd, J ) 38.9, 40.5 Hz, C-4′′], 73.5 [dd, J ) 37.9,
39.4 Hz, C-2′′], 76.6 [d, J ) 39.6 Hz, C-1′′], 80.0 [dd, J )
38.4, 38.7 Hz, C-3′′], 82.3 [dd, J ) 41.4, 41.8 Hz, C-5′′].
High Performance Liquid Chromatography (HPLC). The
HPLC apparatus (Kontron, Eching, Germany) consisted of low
pressure gradient system 525 HPLC pump, a M800 gradient
mixer, a type 560 auto sampler, and a DAD type 540+ diode
array detector. Chromatography was performed on 250 × 4.6
mm stainless-steel columns packed with RP-Phenylhexyl mate-
rial, 5 µm (Phenomenex, Germany) operated with a flow rate
of 0.8 mL/min.
(-)-Catechin-6-C,8-C-di-[13C6]-ꢀ-D-glcp, 19. UV/vis (MeOH/
0.1% HCOOH, pH 2.5; 2/8, v/v) λmax ) 218, 231, 280 nm; MS
(ESI-): m/z 625 (100%, [M – 1]-), 501 (16%, [M – 125]-);
MS/MS (ESI-) (-50 V): m/z 377 (100), 123 (70), 109 (48),
408 (12), 333 (12), 625 (1); 13C NMR (100 MHz, MeOD) δ
62.1 [d, J ) 38.1 Hz, C-6′′/C-6′′′], 62.4 [d, J ) 37.9 Hz, C-6′′/
C-6′′′], 71.2 [dd, J ) 40.2, 40.7 Hz, C-4′′/C-4′′′], 71.5 [dd, J )
38.6, 40.3 Hz, C-4′′/C-4′′′], 73.4 [dd, J ) 38.3, 38.9 Hz, C-2′′/
C-2′′′], 74.8 [dd, J ) 37.7, 37.9 Hz, C-2′′/C-2′′′], 76.9 [d, J )
39.7 Hz, C-1′′/C-1′′′], 77.3 [d, J ) 39.9 Hz, C-1′′/C-1′′′], 79.3
[dd, J ) 37.0, 37.9 Hz, C-3′′/C-3′′′], 79.5 [dd, J ) 37.7, 37.9
Hz, C-3′′/C-3′′′], 82.5 [dd, J ) 41.6, 41.7 Hz, C-5′′/C-5′′′], 82.6
[dd, J ) 41.2, 42.1 Hz, C-5′′/C-5′′′].
Liquid Chromatography–Mass Spectrometry (LC-MS/
MS). LC-MS/MS analysis was performed using an Agilent 1100
HPLC-system connected to the API 3200 LC-MS/MS (Applied
Biosystems, Darmstadt, Germany) running in the negative
electrospray ionization (ESI-) mode. Zero grade air served as
nebulizer gas (35 psi), and as turbo gas (300 °C) for solvent
drying (45 psi). Nitrogen served as curtain (20 psi) and collision
5
gas (4.5 × 10- torr). Both quadrupoles were set at unit
resolution. By means of the multiple reaction monitoring (MRM)
mode, the individual flavan-3-ol-C-glycosides 1, 3, 4, 6, and
14–17 (m/z 451.2 f 331.0), 18 and 20 (m/z 457.1 f 333.0), 2,
5, and 13 (m/z 613.1 f 373.0), 19 (m/z 625.1 f 377.0), 7–9
(m/z 613.5 f 331.0), and 10–12 (m/z 775.5 f 331.0), as well
(-)-Catechin-6-C-[13C6]-ꢀ-D-glcp, 20. UV/vis (MeOH/0.1%
HCOOH, pH 2.5; 2/8, v/v) λmax ) 218, 231, 280 nm; MS
(ESI-): m/z 457 (100%, [M – 1]-), 333 (22%, [M – 125]-);
MS/MS (ESI-) (-50 V): m/z 333 (100), 123 (57), 109 (27),
457 (2). 13C NMR (100 MHz, MeOD/D2O) δ 61.9 [d, J ) 42.5
Hz, C-6″], 70.9 [dd, J ) 39.2, 40.0 Hz, C-4″], 74.2 [dd, J )
38.2, 38.9 Hz, C-2″], 77.1 [d, J ) 39.7 Hz, C-1″], 79.3 [dd, J
) 38.4, 38.2 Hz, C-3″], 82.1 [dd, J ) 41.6, 41.3 Hz, C-5″].
as flavan-3-ol-C-maltotetraoside (m/z 937.8
f 331.0),
flavan-3-ol-C-maltopentaoside (m/z 1099.9 f 331.0), flavan-
3-ol-C-maltohexaoside (m/z 1261.0 f 331.0), flavan-3-ol-C-
maltoheptaoside (m/z 1423.0 f 331.0), and flavan-3-ol-C-
maltooctaoside (m/z 1585.0 f 331.0) were analyzed using the
mass transitions (given in brackets) monitored for a duration
of 75 ms. In addition, the mass transitions of the corresponding
pseudomolecular ions of the C-glycosides to the daughter
fragment ions m/z 123.0 and 109.0 were recorded. ESI- mass
and product ion spectra were acquired with direct flow infusion.
Quantitative Analyses of the Degree of Flavan-3-ol Gly-
cosylation. A mixture of (-)-epicatechin (0.1 mmol), amylose
(0.5 g), K2CO3 (0.5 mmol), and water (5 mL) was treated for
10 min at 80 °C while stirring. After cooling, the reaction

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9691
For ESI-, the ion spray voltage was set at -4500 V in the
negative mode. The MS/MS parameters were tuned for each
individual compound, detecting the fragmentation of the [M –
H]- molecular ions into specific product ions after collision with
nitrogen (4.0 × 10- 5 torr).
Figure 3, C-glucosylated flavan-3-ols containing up to five
glucopyranose moieties were detected in the partial hydrolysate.
On the basis of these data, it can be suggested that the reaction
between (-)-epicatechin and amylose leads to the formation
of conjugates by covalently attaching the flavan-3-ol to the
anomeric carbon atom of the terminal carbohydrate moiety at
the reducing end of the polysaccharide.
LC/Time-of-Flight Mass Spectrometry (LC/TOF-MS).
Mass spectra of the target compounds were measured on a
Bruker Micro-TOF (Bruker Daltronics, Bremen, Germany) mass
spectrometer with flow injection referenced on sodium formiate
and polyethylene glycol (PEG) 600, respectively. The com-
pounds were dissolved in 1 mL of MeOH and 10 µL of a
saturated solution of NaBF4 in MeOH was added to measure
the exact mass of the sodium adducts.
To unequivocally confirm the presence of such flavan-3-ol-
C-oligoglucopyranosides in the partially hydrolyzed flavan-3-
ol/amylose reaction mixture, reference compounds needed to
be synthesized. Closely following the procedure reported
recently for the preparation of the flavan-3-ol-C-glucopyrano-
sides 1–6 (6), the corresponding flavan-3-ol-C-maltosides and
maltotriosides were prepared by reacting (-)-epicatechin either
with D-maltose, or with D-maltotriose, respectively, under
alkaline conditions. After prefractionation by means of RP-18
column chromatography and final purification by means of RP-
HPLC, the target compounds (-)-catechin-8-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p (7), (-)-epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-
glc p (8), (-)-catechin-6- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p(1
f 4) (9), (-)-catechin-8- C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)-
O-R-D-glc p (10), (-)-epicatechin-8-C-ꢀ-D-glc p-(4f1)- O-R-
D-glc p(4f1)-O-R-D-glc p (11), and (-)-catechin-6-C-ꢀ-D-glc
p-(4f1)-O-R-D-glc p(4f1)-O-R-D-glc p (12) (Figure 1) were
obtained as white amorphous powders in a purity of about 99%.
The structures of these previously unreported flavan-3-ol/
oligosaccharide conjugates were unequivocally determined by
means of UV/vis, LC-MS/MS, and 1D/2D-NMR spectroscopy.
1
Nuclear Magnetic Resonance Spectroscopy (NMR). H,
COSY, HMQC, HMBC, 13C and DEPT-135 NMR measure-
ments were performed on a DMX 400 spectrometer (Bruker,
Rheinstetten, Germany). Chemical shifts were referenced to the
solvent signal. Data processing was performed by using XWin-
NMR software (version 3.5; Bruker) as well as Mestre-C
software (Mestrelab Research, Santiago de Compostela, Spain).
RESULTS AND DISCUSSION
Very recent studies demonstrated that a series of velvety
astringent and bitterness supressing (-)-catechin- and (-)-
epicatechin-C-glucopyranosides (1–6, Figure 1) are formed
during the alkalization process of cocoa powder manufacturing
via a novel “nonenzymatic C-glycosylation” of flavan-3-ols with
D-glucose. On the basis of that finding, the question arose as to
whether this nonenzymatic C-glycosylation also takes place
between flavan-3-ols and polysaccharides such as amylose.
Therefore, binary mixtures of (-)-epicatechin and amylose were
thermally treated in an aqueous K2CO3 solution for 10 min at
80 °C to generate polyphenol/amylose conjugates such as
exemplified for the flavan-3-ol-6-C-polyglucopyranosides out-
lined in Figure 2. To confirm the formation of such flavan-3-
ol/amylose linkages, the reaction products formed were treated
with Termamyl, followed by amyloglucosidase to break-down
step-wise the O-glucosidic linkages in the polysacharide back-
bone of the conjugates and to release C-oligoglycosylated
flavan-3-ols. After RP18 cartridge clean-up, the partially
hydrolyzed reaction mixture was screened for the presence of
flavan-3-ol-C-mono- and oligoglycopyranosides by means of
LC-MS/MS(ESI-). To achieve this, the mass transitions for
various conjugates were calculated based on the fragmentation
pattern reported recently by us for the C-monoglucopyranosides
1–6 (6). Assuming that several glucose monomers in an
oligoglucopyranoside chain will give the same characteristic
main daughter ion fragment with m/z 331, which is typical for
the C-glycosides 1–6 (6), the mass transition of m/z 775.5 f
331.0 was selected for a flavan-3-ol C-oligoglucopyranoside
containing three glucose moieties. For each homologous C-
glucosylated flavan-3-ol with an additional glucose moiety, the
putative mass transitions were calculated by addition of another
162 mass units.
Exemplified for (-)-catechin-8-C-ꢀ-D-glc p-(4 f 1)-O-R-D-
glc p (7), this compound showed the typical UV/vis absorption
maxima expected for flavan-3-ols, and showed a pseudomo-
lecular ion [M – 1]- ion with m/z 613 as well as the fragment
ions m/z 331, 451, and 433 in the MS-ESI- spectrum (A in
Figure 4). High resolution LC-MS(ESI-) analysis confirmed
compound 7 to have the molecular formula C27H34O16. The 1H
NMR spectrum showed an aromatic singlet for H-C(6) at 6.01
ppm, three aromatic protons H-C(6′), H-C(5′), and H-C(2′)
resonating at 6.71, 6.75, and 6.83 ppm and showing an ABX
coupling system. In addition, the four aliphatic protons H-C(4R),
H-C(4ꢀ), H-C(3), as well as the proton H-C(2) were detectable
at 2.55, 2.74, 3.99, and 4.76 ppm coupling with each other and
confirming the flavan-3-ol aglycone. Besides the signals of the
1
polyphenol, the H NMR spectrum also exhibited 14 aliphatic
protons resonating at 3.27 ppm [H-C(4′′′)], 3.42–3.46 ppm
[H-C(5′′, 2′′′)], 3.59–3.71 ppm [H-C(3′′, 3′′′, 4′′, 5′′′, 6R′′)],
3.81–3.89 ppm [H-C(6ꢀ′′, 6R, ꢀ′′′)], 4.10 ppm [H-C(2′′)], 4.84
ppm [H-C(1′′)], and 5.16 ppm [H-C(1′′′)] as expected for two
hexose moieties. Considering all the coupling constants of the
sugar moieties in the molecule, and, in particular, the coupling
constants of 9.2 and 3.6 Hz measured for H-C(1′′) and H-C(1′′′)
with the corresponding protons H-C(2′′) and H-C(2′′′), respec-
tively, and comparing these values with the H-C(1)/H-C(2)
coupling constants reported for ꢀ-D-glucopyranosides and
R-D-glucopyranosides (9, 10), the two sugar moieties were
unequivocally identified as one ꢀ-D- and one R-D-glucopyranose.
As shown in Figure 3, recording of the mass transition m/z
451.2f 331.0 revealed that (-)-epicatechin-8-C-ꢀ-D-glc p (4),
(-)-catechin-8-C-ꢀ-D-glc p (1), (-)-catechin-6-C-ꢀ-D-glc p (3),
and (-)-epicatechin-6-C-ꢀ-D-glc p (6) were present in the partial
hydrolysate of the reaction products formed. In addition, we
screened for flavan-3-ol-C-oligoglucopyranosides by recording
the mass transitions expected for flavan-3-ol-C-maltosides (m/z
613.5 f 331.0), C-maltotriosides (m/z 775.5 f 331.0), C-
maltotetraosides (m/z 937.8 f 331.0), and C-maltopentaosides
(m/z 1099.9 f 331.0) as given in parentheses. As shown in
A comparison of the 13C NMR spectrum showing 27 signals
with the results of the DEPT-135 experiment exhibiting 19
signals, revealed eight signals corresponding to quarternary
carbon atoms. Unequivocal assignment of these quarternary
carbon atoms and the hydrogen-substituted carbon atoms,
respectively, could be successfully achieved by means of
heteronuclear multiple bond correlation spectroscopy (HMBC)
2
3
optimized for JC,H and JC,H coupling constants and hetero-
nuclear multiple-quantum correlation spectroscopy (HMQC)

9692 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
Figure 2. Formation of (-)-epicatechin/amylose conjugates and partial enzymatic hydrolysis.
optimized for 1JC,H coupling constants, respectively. Addition-
ally, the HMBC experiment revealed a correlation between the
sugar proton H-C(1′′) resonating at 4.84 ppm and neighbouring
carbon atoms C(7), C(8), and, in particular C(8a), thus dem-
onstrating clearly the intramolecular 8-C-linkage of the ꢀ-D-
As described recently (6), comparison of the chemical shifts
and the coupling constants of the protons H-C(4R), H-C(4ꢀ),
H-C(3), H-C(2) and the sequence of increasing chemical shifts
of the protons H-C(5′), H-C(6′), and H-C(2′) of (-)-epicatechin,
(+)-catechin, as well as their corresponding epimers, with the
isolated C-glycoside 7, allowed its configuration at carbon atom
C(2) and at C(3) to be deduced. The coupling constants and
the chemical shifts of the protons H-C(2), H-C(3), H-C(4Rꢀ),
H-C(2′), H-C(5′), and H-C(6′), as well as the chemical shifts
of their corresponding C-atoms, indicated the 2(S),3(R)-
configuration of the (-)-catechin aglycone of compound 7.
Taking all these spectroscopic data into consideration, the
structure of compound 7 prepared from (-)-epicatechin and
maltose could be unequivocally identified as (-)-catechin-
3
glucopyranose to the aglycone. In addition, a heteronuclear J
coupling between the sugar proton H-C(1′′′) resonating at 5.16
ppm and the carbon atom C(4′′) of the ꢀ-D-glucopyranose could
be detected, thus demonstrating the intramolecular O-linkage
of the ꢀ-D-glucopyranose with the R-D-glucopyranose. In
addition, the chemical shifts of the carbon atoms C(1′′) and
C(1′′′) resonating at 76.4 and 103.0 ppm confirmed the C-linkage
to the ꢀ-D-glucopyranose moiety to the aglycone and the
O-linkage of the R-D-glucopyranose to the other sugar moiety.

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9693
Figure 3. HPLC-MS/MS(ESI-)-MRM analysis of the (-)-epicatechin/amylose reaction mixture after partial enzymatic hydrolysis using Termamyl (15
min) and amyloglucosidase (30 min).
8- C-ꢀ-D-glc p-(4f1)-O-R-D-glc p (Figure 1). Using the
same analytical strategy, (-)-catechin-6-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p (9) and (-)-epicatechin-8-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p (8) were isolated as main reaction products
formed from (-)-epicatechin and D-maltose (Figure 1). To
the best of our knowledge, the compounds 7–9 have not
previously been reported.
Reaction of (-)-epicatechin and D-maltotriose revealed the
previously unreported, corresponding flavan-3-ol-C-maltotrio-
sides, identified as (-)-catechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-
glc p(4f1)-O-R-D-glc p (10), (-)-catechin-6-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p(4f1)-O-R-D-glc p (12), and (-)-epicatechin-6/
8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)- O-R-D-glc p (11)
(Figure 4) by means of LC-MS/MS and 1D/2D-NMR experi-
ments. Exemplified for compound 12, the mass spectrum (ESI-)
of these compounds showed a pseudomolecular ion [M – 1]-
ion with m/z 775 as well as the characteristic fragment ions
m/z 331, 451, and 289 (Figure 4B).
Identification of Flavan-3-ol-C-glycosides 1–12 in the
Partial Hydrolysate of the (-)-Epicatechin/Amylose Reaction
Mixture. To unequivocally verify the covalent binding of
flavan-3-ols to the anomeric carbon atom of the reducing
terminus of polysacharides and to confirm the presence of the
proposed flavan-3-ol-C-glycosides 1–12 in the partial hydroly-
sate, the mass spectrometer was tuned for the mass transitions
of the target compounds 1–12 and then the enzymatically treated
reaction mixture was analyzed by RP–HPLC–MS/MS(ESI-)
running in the multiple reaction monitoring mode. The mass
transition m/z 451.2 f 331.0 was recorded for the flavan-3-ol-
C-glucosides 1, 3, 4, and 6, the mass transition m/z 613.5 f
331.0 was monitored for the flavan-3-ol-C-maltopyranosides
7–9, and the mass transition m/z 775.5 f 331.0 was recorded
for the flavan-3-ol- C-glucosides 10–12, respectively. Besides
the previously reported (-)-catechin- and (-)-epicatechin-C-
glucopyranosides 1–6, the presence of the C-maltosides and
C-maltotriosides 7–12 was confirmed in the partial hydrolysate
of the (-)-epicatechin/amylose reaction mixture by comparing
the retention times and the mass transitions with those obtained
for the corresponding reference compounds and, finally, by co-
chromatography (data not shown).
The flavan-3-ol-6-C,8-C-ꢀ-D-diglc p 2 and 5 have very similar
retention times when compared to their corresponding flavan-
3-ol-8-C-ꢀ-D-maltosides 7 and 8, whereas (-)-epicatechin-6-
C,8-C-ꢀ-D-diglucopyranoside (5) and (-)-epicatechin-8-C-ꢀ-
D-maltoside (8) were found to coelute. By comparing the MS/
MS data obtained at a collision energy of - 50V and, in
particular, the intensities of the fragment ions m/z 331, 373,
and 403 measured for the C-diglucopyranosides 2 and 5 with
those observed for the corresponding C-maltosides 7 and 8
(Figure 5), it was possible to clearly distinguish between a
flavan-3-ol-6-C,8-C-digluco- or a -6- C- and -8-C-maltopyra-
noside. In the case of the C-diglucopyranosides 2 and 5 the ion
m/z 373 is observed as the main fragment, whereas the daughter
ion m/z 331 showed only an intensity of about 10% (Figure
5A). In comparison, the C-maltopyranosides 7 and 8 showed
the highest intensity for the fragment ion m/z 331 and only 8%
for the ion m/z 373 (Figure 5B). In addition, the C-diglucopy-
ranosides 2 and 5 exclusively showed a fragment ion with m/z
403. As shown in Figure 6, the differences in the mass
transitions can be used to clearly distinguish between C-
maltopyranosides (m/z 613.1 f 373/331) and C-diglucopyra-
nosides (m/z 613.1 f 403) showing the same pseudomolecular
ion of 613 [M – 1]-.
In the traces showing the mass transitions m/z 937.8 f 331.0
and m/z 1099.9 f 331.0 calculated for flavan-3-ol-C-maltotet-
raosides and maltopentaosides, respectively, either two or only
one peak was detectable (Figure 3). As shown for all the various
reference compounds 1–12 the flavan-3-ol-8-C-glycosides eluted
earlier than the corresponding 6-C-glycosides, and the catechin-
C-glycosides eluted earlier than the corresponding epicatechin-
C-glycosides (6), so that the compounds detected with the mass
transitions of m/z 937.8 f 331.0 (Figure 3) were tentatively

9694 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
Figure 4. MS (ESI-) spectrum of (A) (-)-catechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p (7) and (B) (-)-catechin-6-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p- (4f1)-
O-R-D-glc p (12).
Figure 5. LC-MS/MS spectrum (ESI-, collision energy -50V) of (A) (-)-catechin-6-C,8-C-ꢀ-D-diglc p (2) and (B) (-)-catechin-8-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p (7).
identified as (-)-catechin-8-C- and (-)-catechin-6-C-ꢀ-D-glc
p-(4f1)-O-R-D-glc p(4f1)-O-R-D-glc p(4f1)-O-R-D-glc p,
respectively. In addition, the compound detected with mass
transitions of m/z 1099.9 f 331.0 might be suggested to be
(-)-catechin-8-C-ꢀ-D-maltopentaoside.
To answer the question as to whether also (+)-catechin is
susceptible to nonenzymatic glycosylation, (+)-catechin and
D-glucose were heated in aqueous K2CO3 solution for 10 min
at 80 °C while stirring. Closely following the procedure
described recently for a (-)-epicatechin/D-glucose reaction
mixture (6), five reaction products (13–17), showing the
typical absorption maxima expected for flavan-3-ols, were
isolated and purified by means of semipreparative RP-HPLC
and their structures were determined by means of LC-MS/

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9695
Figure 6. HPLC-MS/MS(ESI-)-MRM analysis of the (-)-epicatechin/amylose reaction mixture after enzymatic hydrolysis using Termamyl (30 min) and
amyloglucosidase (60 min).
Figure 7. Excerpt of the 13C NMR spectra (100 MHz, MeOD-d3) of (A) nonlabelled (1) and (B) [13C6]-labelled (-)-catechin-8-C-ꢀ-D-glc p (18), respectively.
MS and NMR spectroscopy. High resolution LC-MS(ESI-)
analysis revealed compound 13 to have the molecular formula
C27H34O16, whereas the molecular formula C21H24O11 was
found for compounds 14–17. By means of LC-MS(ESI-),
the reaction products 14–17 showed an [M – 1]- ion with
m/z 451 as well as the fragment ions m/z 331, 123 and 109,
compound 13 showed an [M – 1]- ion with m/z 613 as well
as the fragment ions m/z 373, 123, 109, 331, and 403. 1D-
and 2D-NMR experiments led to the identification of the five
reaction products as (+)-catechin-6-C,8-C-ꢀ-D-diglc p (13),
(+)-catechin-6-C-ꢀ-D-glc p (14), (+)-catechin-8-C-ꢀ-D-glc
p (15), (+)-epicatechin-6-C-ꢀ-D-glc p (16), and (+)-epicat-
echin-8-C-ꢀ-D-glc p (17), the structures of which are given
in Figure 1. In addition, these compounds were identified in
a Termamyl/amyloglucosidase treated (+)-catechin/amylose
reaction mixture by means of HPLC-MS/MS, thus demon-
strating the formation of flavan-3-ol/amylose conjugates (data
not shown). To the best of our knowledge, compounds 13,
16, and 17 have not previously been reported in the literature.
Although compounds 14 and 15have been reported earlier
as phytochemicals in rhubarb (11), the formation of such (+)-
catechin- C-glycopyranosides by nonenzymatic C-glycosy-
lation of the corresponding flavan-3-ol has not been previ-
ously described.
To accurately determine the conversion rate of (-)-epicat-
echin with amylose, a stable isotope dilution analysis (SIDA)
was developed. To achieve this, first, corresponding labelled
internal standards needed to be synthesized.
Synthesis of [13C]-Labelled Flavan-3-ol-C-glucosides. As
the use of stable isotopologues of analytes is known to enable
the correction of compound discrimination during extraction,
cleanup, chromatographic separation, and MS detection, a
versatile and reliable stable isotope dilution assay (SIDA) was
developed to determine the rate of product formation in flavan-
3-ol/polysaccharide model reactions. Aimed at synthesizing
[13C]-isotopologues of the flavan-3-ol C-glucosides, an aqueous
solution of (-)-epicatechin and [13C6]-D-glucose were thermally
treated in the presence of K2CO3 for 10 min at 80 °C. The target

9696 J. Agric. Food Chem., Vol. 55, No. 23, 2007
Stark et al.
Figure 8. MS/MS chromatograms for the quantitative analysis of flavan-3-ol-C-glucopyranosides in a Termamyl/amyloglucosidase-treated (-)-epicatechin/
amylose reaction mixture via the internal standards 18–20 by using the multiple reaction monitoring (MRM) mode.
Table 1. Human Recognition Taste Thresholds for the Oral Sensation Induced by Flavan-3-ols and Flavan-3-ol-C-glycosides, Respectively
threshold conc [µmol/L]a for
compound
astringency
bitterness
(-)-epicatechin
800.0b
600.0b
63.6c
16.2c
10.0c
3.1c
800.0
1000.0
n.d.d
n.d.
(-)-catechin
(-)-catechin-6- C-ꢀ-D-glc p (3)
(-)-catechin-8- C-ꢀ-D-glc p (1)
(-)-epicatechin-6- C-ꢀ-D-glc p (6)
(-)-epicatechin-8- C-ꢀ-D-glc p (4)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
(-)-epicatechin-6- C,8- C-ꢀ-D-diglc p (5)
2.5c
(-)-catechin-6- C,8- C-ꢀ-D-diglc p (2)
1.1c
(-)-catechin-8- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p (7)
18.0c
14.0c
2.2c
(-)-catechin-6- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p (9)
(-)-epicatechin-8- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p (8)
(-)-catechin-6- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p(4 f 1)- O-R-D-glc p (12)
(-)-catechin-8- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p(4 f 1)- O-R-D-glc p (10)
(-)-epicatechin-8- C-ꢀ-D-glc p-(4 f 1)- O-R-D-glc p(4 f 1)- O-R-D-glc p (11)
3.0c
2.5c
n.d.
n.d.
1.4c
a Taste threshold concentrations were determined by means of the half-tongue test (4). b The astringent sensation was described a rough and extremely puckering. c The
astringent sensation was described as a very smooth and velvety mouthcoating. d n.d.: not detectable.
compounds were isolated by column chromatography on
RP-18 material, followed by a final HPLC purification yielding
(-)-catechin-8-C-[13C6]-ꢀ-D-glucopyranoside (18), (-)-catechin-
6-C,8-C-[13C12]-ꢀ-D-diglucopyranoside (19), and (-)-catechin-
6-C-[13C6]-ꢀ-D-glucopyranoside (20) each as a white, amorphous
powder in high purities of >98%.
When compared to the nonlabelled analogue 2, the mass
spectrum of (-)-catechin-6-C,8-C-ꢀ-D-[13C12]-diglucopyrano-
side (19), obtained in the ESI- mode, revealed an increase of
the pseudomolecular ion by 12 units, thus reflecting the
incorporation of two molecules of [13C6]-glucose into the
molecule. In addition, the incorporation of the 13C-atoms into
the target molecules was confirmed by means of 1H broad-band-
decoupled 13C NMR spectroscopy. For example, the 60–85 ppm
region of the 13C NMR spectrum obtained from (-)-catechin-
8-C-ꢀ-D-glucopyranoside (1) showed eight carbon signals
resonating as singlets (Figure 7A), whereas the same spectrum
segment obtained from the isotopologue 18 showed six coupled
resonance signals for the carbon atoms of the carbohydrate
moiety (Figure 7B) due to homonuclear 13C-13C connectivities.
As expected, the signals of the carbon atoms C(2) and C(3)
were not detectable in the NMR spectrum obtained from 18
due to their low 13C abundance.
Development of a Stable Isotope Dilution Assay (SIDA)
for the Quantitative Analysis of Flavan-3-ol-C-Glucosides.
To convert the measured ion intensities into the mass ratios of
labelled and nonlabelled flavan-3-ol-C-glycosates, a graph was
calculated from calibration mixtures of known mass ratios and the
corresponding peak area ratios obtained by means of LC-MS/MS.
Good linearity was found for mass ratios ranging from 0.2 to 5.0.
To quantify the degree of flavan-3-ol glycosylation in a
flavan-3-ol/polysaccharide reaction, the epicatechin/amylose
reaction mixture was spiked with defined amounts of the [13C]-
labelled internal standards 18, 19, 20, followed by homogeni-
sation, equilibration at room temperature, and enzymatic
depolymerization by means of Termamyl and amyloglucosidase
treatment. In order to ensure a total enzymatic hydrolysis of
the flavan-3-ol- C-oligoglucosides, aliquots of the reaction
mixture were treated with Termamyl for 2h, followed by an
amyloglucosidase treatment for an additional 4h. Thereafter, the

Structure of Flavan-3-ol-C-oligosaccharides
J. Agric. Food Chem., Vol. 55, No. 23, 2007 9697
hydrolysates obtained were analyzed by RP-HPLC-MS/MS
(Figure 8). Only trace amounts of flavan-3-ol-C-ꢀ-D-glc p-(4f1)-
O-R-D-glc p were detectable and neither flavan-3-ol-C-ꢀ-D-glc
p-(4f1)-O-R-D-glc p-(4f1)-O-R-D-glc p, nor higher C-oligo-
mers could be observed, thus demonstrating a nearly quantitative
hydrolytic break-down of the O-glucoside linkages in the flavan-
3-ol- C-oligoglucosides. On the basis of this finding, the yield
of flavan-3-ol-6-C-ꢀ-D-glc p (3), flavan-3-ol-8-C-ꢀ-D-glc p (1),
and flavan-3-ol-6-C,8-C-ꢀ-D-glc p termini (2) in the (-)-
epicatechin/amylose reaction was determined (Figure 8). The
SIDA revealed an average amount of 14.0 ((2.2), 9.0 ((1.4),
and 0.15 ((0.3) µmol per mmol (-)-epicatechin for compound
3, 1, and 2, respectively. These data clearly demonstrate that
the reducing terminus of the polysaccharide is reacting prefer-
entially with the 6- and 8-position of flavan-3-ols. To a much
smaller extent, flavan-3-ols seem also to be able to cross-link
the reducing termini of two polysaccharide moieties via the 6-
and 8-position.
In conclusion, individual flavan-3-ol-C-ꢀ-D-glycopyranosides
could be identified as the cross-linker molecules linking the
reducing terminus of a polysaccharide to the A-ring of flavan-
3-ols. These data allow first structural insights into the chemistry
of the formation of covalent polyphenol/polysaccharides con-
jugates generated upon food processing.
LITERATURE CITED
(1) Manach, C.; Scalbert, A.; Morand, C.; Rémésy, Ch.; Jiménez, L.
Polyphenols: food source and bioavailability. Am. J. Clin. Nutr.
2004, 79, 727–747.
(2) González-Paramás, A. M.; Lopes da Silva, F.; Martín-López, P.;
Macz-Pop, G.; González-Manzano, S.; Alcalde-Eon, Ch.; Joaquin
Pérez-Alonso, J.; Escribano-Bailón, M. T.; Rivas-Gonzalo, J. T.;
Santos-Buelga, C. Flavanol-anthocyanin condensed pigments in
plant extracts. Food Chem. 2006, 94, 428–436.
(3) Es-Safi, N. E.; Cheynier, V.; Moutounet, M. Study of the reactions
between (+)-catechin and furfural derivatives in the presence or
absence of anthocyanins and their implication in food color
change. J. Agric. Food Chem. 2000, 48, 5946–5954.
(4) Quideau, S.; Jourdes, M.; Saucier, C.; Glories, Y.; Pardon, P.;
Baudry, C. DNA topoisomerase inhibitor acutissimin A and other
flavano-ellagitannins in red wine. Angew. Chem. Int. Ed. 2003,
42, 6012–6014.
(5) Glabasnia, A.; Hofmann, T. Identification and sensory evaluation
of dehydro- and deoxy-ellagitannins formed upon toasting of oak
wood (Quercus alba L.). J. Agric. Food Chem. 2007, 55, 4109–
4118.
Sensory Activity of Flavan-3-ol-C-glycosides. As recent
studies decribed the velvetly astringency of the glucosides 1–6
(6), the taste thresholds of the flavan-3-ol-C-glycosides 7–12
was evaluated next. Prior to sensory analysis, the purity of all
compounds was checked by LC-MS as well as 1H NMR
spectroscopy. To study the sensory activity of 7–12, the human
sensory recognition thresholds were determined in bottled water
(pH 6.0) using the half-mouth test (5, 8) (Table 1).
Compared to the aglycones (-)-epicatechin and (-)-catechin,
exhibiting a puckering astringent sensation as well as bitter taste
at threshold concentrations between 600 and 1000 µmol/L, the
flavan-3-ol-C-glucosides 1, 3, 4, and 6 induced a smooth
astringent and velvety mouth-coating sensation at very low
threshold concentrations ranging from 3.1 to 63.6 µmol/L
without exhibiting any bitter taste (Table 1). In comparison,
the flavan-3-ol-C-diglucosides (2, 5) showed the same sensory
attributes at even lower thresholds of 1.1 and 2.5 µmol/L,
respectively. Sensory analysis of the flavan-3-ol-C-ꢀ-D-glc
p-(4f1)-O-R-D-glc p isomers 10 and 12 as well as the flavan-
3-ol-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4 f 1)-O-R-D-glc p (11)
revealed somewhat lower thresholds when compared to those
found for the flavan-3-ol-C-glucosides. In particular, the thresh-
old concentration of 1.4 and 2.2 µmol/L determined for (-)-
epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p(4f1)-O-R-D-glc
p and (-)-epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-D-glc p were
found to be extraordinarily low. These data clearly demonstrate
that the elongation of the carbohydrate chain up to three glucose
units favors the sensory impact of these flavan-3-ol glycoconju-
gates.
(6) Stark, T.; Hofmann, T. Application of a molecular sensory science
approach to alkalized cocoa (Theobroma cacao): structure deter-
mination and sensory activity of nonenzymatically C-glycosylated
flavan-3-ols. J. Agric. Food Chem. 2006, 54, 9510–9521.
(7) Wieser, H.; Belitz, H.-D. Studies on the structure/activity relation-
ship of bitter tasting amino acids and peptides (in German). Z.
Lebensm.-Unters. Forsch. 1975, 159, 65–72.
(8) Scharbert, S.; Holzmann, N.; Hofmann, T. Identification of the
astringent taste compounds in black tea infusions by combining
instrumental analysis and human bioresponse. J. Agric. Food
Chem. 2004, 52, 3498–3508.
(9) Altona, C.; Haasnoot, C. A. G. Prediction of anti and gauche
vicinal proton-proton coupling constants in carbohydrates: a simple
additivity rule for pyranose rings. Org. Magn. Reson. 1980, 13,
417–428.
(10) Markham, K. R.; Geiger, H. 1H nuclear magnetic resonance
spectroscopy of flavanoids and their glycosides in hexadeuteri-
odimethylsulfoxide. In The FlaVanoids; Harborne, J. B., Ed.;
Chapman & Hall: London, 1994; pp 441–497.
(11) Kashiwada, Y.; Nonaka, G.-I.; Nishioka, I. Tannins and related
compounds. XLV. Rhubarb. (5). Isolation and characterization
of flavan-3-ol and procyanidin glucosides. Chem. Pharm. Bull.
1986, 34, 3208–3222.
In line with our previous observations (6), the aglycone
moiety has a significant influence on the perception of astrin-
gency. Comparing (-)-epicatechin-8-C-ꢀ-D-glc p-(4f1)-O-R-
D-glc p (8) with the corresponding (-)-catechin-8-C-ꢀ-D-glc
p-(4f1)-O-R-D-glc p (7) showed that the aglycone strongly
influences the taste intensity of these compounds; for example,
compound 7 exhibited a threshold concentration of 18.0 µmol/
L, which is 8.2 times above the threshold concentration
determined for the corresponding (-)-epicatechin derivative.
Received for review June 30, 2007. Revised manuscript received August
28, 2007. Accepted August 30, 2007. This work was supported by the
International Graduate Research School Münster, Germany / Nagoya,
Japan (IRGT MS-NG 1143): “Complex Functional Systems in Chem-
istry: Design, Development and Applications” funded by the Deutsche
Forschungsgemeinschaft (DFG) and the Japan Society for the Promo-
tion of Science (JSPS).
JF0719508