Carbon-14 biolabeling of (+)-catechin and proanthocyanidin oligomers in willow tree cuttings

Article abstract of DOI:10.1021/jf981380z

Proanthocyanidin polymers, oligomers, and the structurally related monomer (+)-catechin were labeled by incorporation of radioactive precursors in shoots of willow tree (Salix caprea L.). [1-¹⁴C]Acetate and [U-¹⁴C]- phenylalanine pre

Full text of DOI:10.1021/jf981380z

J. Agric. Food Chem. 1999, 47, 4219−4230
4219
Ca r bon -14 Biola belin g of (+)-Ca tech in a n d P r oa n th ocya n id in
Oligom er s in Willow Tr ee Cu ttin gs
Ste´phanie De´prez,† Isabelle Mila,† and Augustin Scalbert*,‡
Laboratoire de Chimie Biologique (INRA), Centre de Biotechnologie Agro-Industrielle, Institut National
Agronomique ParissGrignon, 78850 Thiverval-Grignon, France, and Station de Technologie des Produits
ve´ge´taux, INRA, Domaine Saint-Paul, Site Agroparc, B.P. 91, 84914 Montfavet Cedex 9, France
Proanthocyanidin polymers, oligomers, and the structurally related monomer (+)-catechin were
labeled by incorporation of radioactive precursors in shoots of willow tree (Salix caprea L.). [1-14C]-
Acetate and [U-14C]-phenylalanine precursors were fed through the cut stems or petioles of leaves.
Optimization of several parameters such as the nature and origin of the plant material, leaf maturity,
nature, and quantity of radioactive precursor applied and the duration of metabolism led to
incorporation yields of 3.2% and to specific activities of 500 µCi/g. Detailed characterization of the
products (polymerization degree, procyanidin/prodelphinidin ratio, specific activities) and purification
by chromatography are reported. Some sugars bound to radiolabeled proanthocyanidin polymers
were removed by enzymic treatment with a mixture of glycosidases. A radioactive purity close to
100% and specific activities suitable for bioavailability studies were obtained.
Keyw or d s: Proanthocyanidins; Salix caprea; willow tree; biosynthesis; radiolabeling; bioavailability
INTRODUCTION
carbon. Their molecular weight most often varies be-
tween 600 (dimers) and 3000. They also differ by their
hydroxylation pattern and the stereochemistry of C-3.
The main PAs found in food are procyanidins, polymers
of catechin units (5, R ) H), and to a lesser extent
prodelphinidins, polymers of gallocatechin units (5, R
) OH) (Shahidi and Naczk, 1995). Two types of units
can be distinguished according to their stereochemis-
try: (gallo)catechin with a 2,3-trans configuration and
(gallo)epicatechin with a 2,3-cis configuration.
This high structural complexity makes PAs difficult
to analyze and estimate. Furthermore, pure PA mol-
ecules needed for biological studies are not commercially
available. For these reasons and in contrast to simpler
molecules such as quercetin, catechin or genistein, their
nutritional effects and their bioavailability have been
poorly studied and remain largely unknown.
Isotopic labeling has been an invaluable tool in
pharmakinetic studies of drugs or natural products to
trace the parent compounds or their metabolites. La-
beled polyphenols have been prepared either through
synthetic routes (full synthesis or aromatic proton
exchange) or by incorporation of a commercial radio-
labeled precursor in plant tissues known to accumulate
the molecule of interest (De´prez and Scalbert, 1999).
Few studies on the bioavailability of PAs have been
reported, and still fewer are carried out with labeled
substrates. PAs have been radiolabeled by incorporation
of 14CO2, 14C-acetate, or 14C-phenylalanine in Vitis
vinifera (Laparra et al., 1977; Harmand and Blanquet,
1978), Sorghum bicolor (Reddy and Butler, 1989; J ime-
nez-Ramsey et al., 1994), and Lotus pedunculatus
(Terrill et al., 1994). However, only crude mixtures of
oligomers or polymers have been prepared, and bio-
availability studies did not allow one to determine
precisely the effect of the polymerization degree on their
absorption (Laparra et al., 1977; Harmand and Blan-
Polyphenols are widespread in most fruits and veg-
etables regularly consumed, either raw or transformed
(Swain, 1962; Shahidi and Naczk, 1995). It is generally
assumed that we ingest 1 g of polyphenols per day with
our diet (Ku¨hnau, 1976). Their antioxidant properties
have raised considerable interest over the past years.
They may contribute to explain the general protective
effects of fruit and vegetable consumption against
cancers (Steinmetz and Potter, 1991) and of a moderate
consumption of wine against cardiovascular diseases
(Hertog, 1997).
The diversity of polyphenol structures is enormous.
They are classified into several classes, the most im-
portant in our food being phenolic acids and flavonoids
(Shahidi and Naczk, 1995). Proanthocyanidins (PAs,
syn. condensed tannins) differ from other flavonoids by
their polymeric nature and are probably the most
abundant polyphenols in our diet. They might well
represent over 50% of the 1 g of polyphenols consumed
every day (Ku¨hnau, 1976). PAs are responsible for the
astringency and to some extent the bitterness of many
fruits (grape, apple, persimmon, etc.) and derived bever-
ages (wine, cider). For example a glass of red wine (125
mL) contains 30-60 mg of PA dimers and trimers
(Ricardo da Silva et al., 1991, 1992) and probably as
much or more polymers.
PAs are polymers of flavan-3-ol units linked by a
carbon-carbon bond between the benzylic C-4 carbon
of the heterocyclic ring and principally the C-8 carbon
of the flavanol A ring of another unit but also the C-6
* To whom correspondence should be addressed: Labora-
toire des Maladies Me´taboliques et Micronutriments, INRA,
63122 Saint-Gene`s-Champanelle, France (phone +33 4 73 62
42 33; fax +334 73 62 46 38; email scalbert@clermont.inra.fr).
† Institut National Agronomique ParissGrignon.
‡ INRA.
10.1021/jf981380z CCC: $18.00 © 1999 American Chemical Society
Published on Web 09/24/1999

4220 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
An alysis of Flavan ols an d Th iolysis P r odu cts of P r oan -
th ocya n id in s by HP LC. Thiolysis was carried out as de-
scribed before and used to estimate polymers in leaf extracts
(Matthews et al., 1997; De´prez et al., 1999). Analytical RP-
HPLC was carried out on a Lichrospher 100 RP-18 column (5
µm, 250 × 4 mm i.d., Merck) with the following elution
conditions: solvent A, 0.1% H3PO4 in water; solvent B, MeOH;
linear gradient, 15-30% B in 30 min for flavanols and 30-
90% B in 30 min for thiolysis products; flow rate, 1 mL/min.
Detection was carried out at 280 nm or with an ESI-MS
Platform LCZ spectrometer (Micromass) operating in negative
mode. For ESI-MS detection, 1/5 of the eluent was admitted
in the source, masses were scanned from m/z 150 to 1000, and
a cone voltage of -30 V was applied. Retention times of
benzylthioethers were 3,4-cis-benzylthiocatechin (6), 18.7 min;
3,4-trans-benzylthiocatechin (7), 17.5 min; 3,4-trans-benzyl-
thioepicatechin (8), 20.8 min; 3,4-cis-benzylthiogallocatechin
(9), 14.6 min; 3,4-trans-benzylthiogallocatechin (10), 13.4 min;
3,4-trans-benzylthioepigallocatechin (11), 16.5 min (De´prez et
al., 1999).
quet, 1978; J imenez-Ramsey et al., 1994; Terrill et al.,
1994).
We report here the preparation of pure radiolabeled
PA oligomers and polymers by incorporation of 14C-
acetate or 14C-phenylalanine in willow tree cuttings.
Male catkins of willow tree were reported to contain
high quantities of dimer B3 2, trimer C2 4, PA polymers
5 and the related (+)-catechin monomer 1 (Thompson
et al., 1972). After checking that leaves contained
An a lysis of F r ee a n d Bou n d Su ga r s. Free sugars in
leaves (0.1 g leaf dry wt.) extracted with acetone/water 7:3 (20
mL) were redissolved in water (2 mL). A fucose internal
standard was added and the solution (0.1 mL) applied to a
Sep-Pak C18 cartridge (Waters) preliminary washed with
MeOH and conditioned in water. Bound sugars were hydro-
lyzed from PA polymers (1 mg) either by acid hydrolysis or
enzymatic treatment. Acid hydrolysis was performed by treat-
ment at 100 °C for 3 h in EtOH/2 N aq. HCl 1:1 (0.6 mL). After
cooling, acid was neutralized with NaHCO3 and filtered
(Harborne, 1965). After evaporation of EtOH, the aqueous
phase was applied to a Sep-Pak C18 cartridge as described
above. Enzymatic treatment was carried out at 37 °C with
â-glucosidase, cellulase Onozuka or AR2000 pectinase in water
(1 mL) with stirring and for 18 h (reaction time at which yields
reached a plateau). Water was then partially evaporated under
reduced pressure, and acetone (0.6 mL) was added to precipi-
tate enzymes. After centrifugation and evaporation of acetone,
compounds were redissolved in water and applied to a Sep-
Pak C18 cartridge as described above. Sugars were analyzed
by HPLC on a DIONEX 4500I series liquid chromatograph
equipped with a pulsed amperometric detector (PAD-2). A
CarboPac PA1 column (250 × 4 mm i.d.) was used with the
following elution conditions: solvent A, water; solvent B,
NaOH 50 mM; sequential isocratic elution of 92% A (27 min)
and 40% A (8 min); flow rate, 1 mL/min. PAD was employed
with a gold electrode and triple-pulse amperometry. Addition
of NaOH (300 mM, 1.7 mL/min) to the column effluent
increased the PAD sensitivity and minimized baseline drift.
Peak identification and quantification were based on retention
times, and response factors were determined with reference
compounds.
similar amounts of the same compounds, willow cuttings
were selected to prepare the radiolabeled molecules
otherwise difficult to obtain by synthetic routes. Various
parameters, such as selection of plant material, nature
of precursor, and its mode of administration, were
optimized and methods of purification of the radio-
labeled products developed.
Th in -La yer Ch r om a togr a p h y. Bidimensional chroma-
tography was carried out with high-performance cellulose
plates on aluminum foils (10 × 10 cm, 0.1 mm, Merck). Eluents
for PAs were t-BuOH/AcOH/water 3:1:1 (by vol., eluant A),
AcOH/water 6:94 (eluant B) and for sugars (Markham, 1982)
EtOAc/pyridine/AcOH/water 36:36:7:21 (eluant C) and n-
BuOH/toluene/pyridine/water 5:1:3:3 (eluant D). Plates were
sprayed with a mixture of vanillin (2% m/v) and toluene-p-
sulfonic acid (1% m/v) reagent in EtOH and heated at 60 °C
(PAs) or a mixture of equal volumes of aqueous K3Fe(CN)6 (2%
m/v) and FeCl3 (2% m/v) (all phenolic compounds). Sugars were
detected by spraying plates with aniline hydrogen phthalate
(aniline, 0.2 mL, phthalic acid, 0.4 g in 25 mL n-BuOH/Et2O/
H2O 24:24:1) and heating at 100 °C (brown spots) (Partridge,
1949). Phenylalanine was revealed by spraying ninhydrine (0.2
g in 100 mL EtOH) and appeared as a purple spot after heating
at 100 °C.
MATERIALS AND METHODS
Gen er a l. (+)-Catechin was purchased from Fluka, D-
glucose, D-arabinose, D-xylose, D-sucrose from Prolabo, D-
galactose from Serlabo, D-fucose and chlorogenic acid from
Sigma, and D-fructose from Merck. (+)-Gallocatechin was
kindly provided by Dr. Alan P. Davies (Unilever, Bedford,
U.K.). Radioactive precursors were purchased from Isotopchim
for labeling experiments ([1-14C]-acetic acid, sodium salt, 59
mCi/mmol; L-[U-Ring-14C]-phenylalanine, 450 mCi/mmol) and
from ICN for toxicity experiments ([1-14C]-acetic acid, sodium
salt, 10 mCi/mmol dry). Other radiochemicals were purchased
from Sigma ([UL-Ring-14C]-atrazine, 5-25 mCi/mmol dry) and
Amersham ([1,4-14C]-putrescine dihydrochloride, 114 mCi/
mmol in 2% EtOH). â-Glucosidase from almonds, cellulase
Onozuka, and AR2000 pectinase from Aspergillus niger were
obtained from Fluka, Yakult Pharmaceutical, and Littorale
Oenologie (Gist-Brocades), respectively. 1H NMR spectra were
recorded by Kate Herve´ du Penhoat (CERMAV, Grenoble) at
400 MHz on a Bruker DRX 400 spectrometer at 300 K;
chemical shifts (δ) were referenced to TMS. 13C NMR spectra
were recorded at 100 MHz on a Bruker AC300 instrument at
303 K; chemical shifts (δ) were referenced to solvent signal.
P u r ifica tion of Un la beled P r oa n th ocya n id in Oligo-
m er s a n d P olym er s. Fresh male catkins (230 g fresh wt.)
were collected on Salix caprea L. trees in February and
extracted by acetone/water 7:3 (2 × 100 mL) with a Waring
blender. The solution was filtered, and acetone was removed
under reduced pressure from the solution. The aqueous phase

Biolabeling of proanthocyanidins
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4221
was successively extracted with Et2O (3 × 100 mL) and EtOAc
(5 × 40 mL). The EtOAc extract was dried (4 g), redissolved
146.7-144.8 (C-3′, C-4′), 131.3 (C-1′), 120.0 (C-6′), 115.8 (C-
2′, C-5′), 106.2 (C-8), 103.9 (C-1′′, R or â anomer, sugar), 101.5
(C-4a), 100.3 (C-1′′, R or â anomer, sugar), 95.9 (C-6), 82.7 (C-
2, 2,3-trans linkage), 76.3-75.6 (C-2, 2,3-cis linkage; C-3′′,
C-5′′, sugar), 73.6-72.8 (C-3; C-2′′, sugar), 69.8-69.4 (C-4′′,
sugar), 65.6 (C-6′′, sugar), 37.8 (C-4). Mixed procyanidin/
prodelphinidin polymers (5, R ) H or OH) isolated from leaves
of cuttings grown in the greenhouse: 13C NMR (100 MHz,
acetone-d6/D2O 1:1) δ 156.5-152.3 (C-5, C-7, C-8a), 144.5-
143.0 (C-3′, C-4′, C-5′), 129.8 (C-1′), 114.0 (C-2′, C-6′), 105.2,
104.9 (C-8), 103.1 (C-1′′, sugar), 99.0 (C-4a), 95.4 (C-6), 81.5
(C-2, 2,3-trans linkage), 76.2-74.0 (C-2, 2,3-cis linkage; C-3′′,
C-5′′, sugar), 72.9-71.0 (C-3; C-2′′, sugar), 69.0 (C-4′′, sugar),
63.3 (C-6′′, sugar), 36.1 (C-4).
in EtOH/MeOH 2:1 (6 mL), and chromatographed on
a
Sephadex LH-20 column (55 × 3 cm i.d.; 100 µm; Pharmacia)
in absolute EtOH. Four fractions were successively eluted and
contained (+)-catechin 1 (470 mL, 444 mg), B3 dimer 2 (540
mL, 250 mg), B6 dimer 3 (670 mL, 55 mg), and C2 trimer 4
(950 mL, 187 mg). The fraction containing C2 trimer was
further purified by preparative RP-HPLC (Scalbert et al., 1990)
with H3PO4/H2O/MeOH 1:859:140 as eluant. MeOH was
removed from the fractions containing C2 trimer which were
applied on a Sephadex LH-20 column (140 × 14 mm i.d.) in
water to eliminate H3PO4. The adsorbed trimer was eluted
with MeOH and freeze-dried (42 mg of pure compound).
Ad m in istr a tion of 14C-La beled P r ecu r sor s to Willow
Sh oots. The incorporation chamber (120 (height) × 180
(length) × 40 (width) cm) was equipped with an air extractor.
The lighting produced by 18 neon tubes (Mazdafluor, TF 75,
58 W) reached 0.5 mmol photons‚m-2‚s-1 PAR. Light periods
lasted 16 h. The air temperature was most often around 25
°C but varied occasionally between 18 and 33 °C. Plates
containing 1 M NaOH were placed in the chamber to eventu-
ally trap respired 14CO2.
Leafy shoots were harvested on an adult willow tree (S.
caprea L.) at the Arboretum of the Museum National d'Histoire
Naturelle of Che`vreloup (Yvelines; shoots harvested from April
to September) or on young plants purchased from a nursery
and planted in November (the year before the incorporation
experiments) in a greenhouse (1 m height, 3 cm stem circum-
ference, shoots harvested from February to October). Four
sodium lamps (16 h light period) provided additional illumina-
tion. Overall illumination varied between 0.5 and 2 mmol
photons‚m-2‚s-1 PAR (400-700 nm).
The stem of the leafy shoot (15-20 cm long; 10-20 leaves)
or the leaf petiole was recut under water before administration
of the precursor. The cut end was immersed in water contain-
ing sodium [1-14C]-acetate (0.4 mCi in 1 mL, pH made to 7
with 1 N HCl). When the solution was almost totally absorbed,
more water was added in small aliquots (1 mL) until the
labeled precursor was fully absorbed at an average rate of 1
mL/h. The cutting was then fed with larger volumes of water.
Shoots were either kept in the growth chamber until full
desiccation of leaves (survival before wilting varied between
1 and 6 days) or dried at room temperature for time-course
labeling experiments. The same procedure was applied for
incorporation of L-[U-Ring]-phenylalanine.
Procyanidin and mixed procyanidin/prodelphinidin polymers
were purified from leaves of adult trees grown outdoors and
young trees grown in the greenhouse, respectively. Leaves (1
g dry wt.) were similarly extracted as catkins. PAs in the
residual aqueous phase were purified on the same Sephadex
LH-20 column as above. The following eluants were succes-
sively applied: water (100 mL), MeOH/water 1:1 (50 mL), and
acetone/water 7:3 (150 mL). PAs in the acetone/water eluant
were dried by removing the solvants under reduced pressure
and freeze-drying.
B3 Dim er 2. The purity was assessed by chromatography
and NMR spectroscopy. B3 dimer (15 mg) was dissolved in
anhydrous pyridine (500 µL) and dry acetic anhydride (500
µL) and the solution stirred at room temperature for 15 h in
the dark. Some toluene was added, and the solvents were
removed under reduced pressure. Residual pyridine was
eliminated by dissolving the acetylated dimer in CH2Cl2 (5 mL)
and washing with 10% HCl. The organic phase was neutralized
by addition of NaHCO3, filtered, and freeze-dried. Acetylated
dimer was recovered as a white powder (80% yield on a molar
basis). B3 dimer peracetate: 1H NMR (400 MHz, CDCl3) δ 7.18
(d, J ) 8.4 Hz, H-5′B or E), 7.17 (d, J ) 8.3 Hz, H-5′E or B),
7.06 (d, J ) 1.4 Hz, H-2′B), 7.02 (dd, J ) 8.4 Hz, J ) 1.4 Hz,
H-6′B), 6.97 (d, J ) 1.4 Hz, H-2′E), 6.76 (dd, J ) 8.4 Hz, J )
1.4 Hz, H-6′E), 6.68 (s, H-6D), 6.54 (d, J ) 2.1 Hz, H-8A), 6.52
(d, J ) 2.1 Hz, H-6A), 5.66 (t, J ) 9.7 Hz, H-3C), 5.05 (m,
H-3F), 4.99 (d, J ) 8.0 Hz, H-2F), 4.80 (d, J ) 10.0 Hz, H-2C),
4.52 (d, J ) 9.3 Hz, H-4C), 2.97 (dd, J ) 5.6 Hz, J ) 16.7 Hz,
H-4RF), 2.69 (dd, J ) 7.8 Hz, J ) 16.8 Hz, H-4âF), 1.9-2.4
(m, 30 H, 10CH3CO).
1
C2 Tr im er 4. It was peracetylated as above. H NMR (400
MHz, CDCl3) δ: 7.31 (d, J ) 8.4 Hz, H-5′), 7.10 (m, H-6′), 7.13
(d, J ) 9.1 Hz, H-5′), 6.90 (d, J ) 2.0 Hz, H-2′), 6.84 (d, J )
2.0 Hz, H-2′), 6.70 (s, H-6G), 6.66 (s, H-6D), 6.62 (dd, J ) 2.1
Hz, J ) 8.5 Hz, H-6′), 6.58 (d, J ) 2.4 Hz, H-8A), 6.57 (dd, J
) 1.9 Hz, J ) 8.7 Hz, H-6′), 6.26 (d, J ) 2.3 Hz, H-6A), 5.61
(dd, J ) 9.0 Hz, J ) 10.1 Hz, H-3C), 5.53 (m, H-3F), 5.29 (sl,
H-2I), 5.27 (m, H-3I), 4.79 (d, J ) 10.2 Hz, H-2F), 4.62 (d, J )
8.3 Hz, H-4F), 4.67 (d, J ) 10.1 Hz, H-2C), 4.18 (d, J ) 9.0
Hz, H-4C), 2.60 (m, H-4âI), 2.10 (m, H4RI masked by acetate
signals), 1.6-2.4 (m, 45 H, 15CH3CO). 13C NMR (100 MHz,
CDCl3) δ: 171.4-167.0 (CO of CH3CO), 156.3 (C8aA), 154.8
(C8aD), 151(C8aD),0.151.1 (C8aG), 150-147 (C7, C5, A, D and
G), 149.2 (C5A), 142.5-141.1 (C4′, C3′, B, E, and H), 136.6
(C1′H), 135.1 (C1′E), 125.6 (C6′B or E), 124.7-122.3 (C2′, C5′,
C6′, B, E and H), 119.5 (C2′B, E, and H), 118.8 (C8D), 117.2
(C4aD), 116.7 (C4aA), 116.6 (C8G), 110.4 (C6G), 109.8 (C4aG),
108.4 (C6D), 108.3 (C6A), 79.7 (C2F), 78.6 (C2C), 76.4 (C2I),
72.1 (C3F), 70.9 (C3C), 66.7 (C3I), 36.7 (C4C), 36.7 (C4F), 29.7
(C4I), 21.2-20.2 (CH3 of CH3CO).
B6 Dim er 3. It was characterized as a procyanidin dimer
by HPLC-ESI-MS (m/z 578) and identified by comparison of
its chromatographic characteristics with those described in the
literature. Elution order on Sephadex LH-20 column (Thomp-
son et al., 1972), retention time by RP-HPLC (Treutter et al.,
1994), and Rf values on bidimensional cellulose HPTLC
(Thompson et al., 1972) were all consistent.
P r oa n t h ocya n id in P olym er s. Procyanidins (5, R ) H)
isolated from leaves of an adult willow tree: 13C NMR (100
MHz, acetone-d6/D2O 1:1) δ 157.8-154.4 (C-5, C-7, C-8a),
Isola tion a n d P u r ifica tion of La beled F la va n ols. Dry
leaves (1 g, ∼20 leaves) were ground to a fine powder with a
porcelain pestle and liquid nitrogen in a porcelain mortar. The
powder was extracted with acetone/water 7:3 (3 × 20 mL), and
plant debris was filtered off on Whatman No. 4 paper. Extracts
were stored at -20 °C until further purification. A similar
experiment carried out on similar nonlabeled leaf extracts
showed that freezing (1 month) and thawing do not affect PA
solubility, as seen by the stable absorbance at 280 nm. Extracts
from about 3 g of dry leaves were combined, and water was
added to obtain a final 5:95 acetone/water ratio. This solution
was applied first to a Sephadex LH-20 column (300 × 20 mm
i.d.) to eliminate the remaining [1-14C]-acetate precursor which
would otherwise be lost in the rotary evaporator during
subsequent sample concentrations. [1-14C]-Acetate was eluted
with water (150 mL) together with other polar contaminants
(free carbohydrates, phenolic acids). PAs were then eluted with
acetone/water 7:3 (150 mL), and acetone was removed from
the solution under reduced pressure. This first Sephadex
chromatography step was omitted when [U-14C]-phenylalanine
was used as a precursor.
The aqueous phase (40 mL) was successively extracted with
n-hexane (3 × 30 mL) and EtOAc (3 × 30 mL). The EtOAc
phase contained (+)-catechin 1, PA dimers 2, 3, and trimer 4,
and the aqueous-phase PA polymers 5. Several EtOAc extracts
corresponding to 11 g of initial dry leaves were pooled, washed
with n-hexane to remove chlorophylls, and applied to a
Sephadex LH-20 column (250 × 20 mm i.d.) in absolute EtOH.
Each fraction collected manually (5 mL) was analyzed by

4222 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
HPTLC using eluant A and sprayed with the vanillin/toluene-
p-sulfonic acid reagent. Similar fractions were pooled. After
evaporation of EtOH and redissolution in water/MeOH 1:1,
fractions were analyzed by HPLC and 2D-HPTLC. The still
contaminated (+)-catechin and dimer B3 fractions were further
purified by cellulose HPTLC (12 × 8 cm, 0.1 mm thickness,
Merck); 0.5 µCi in about 40 µL were spotted on a 10 cm width
and eluted with eluant B. Labeled flavanols were localized by
autoradiography, the cellulose scraped, and the product finally
recovered by stirring in acetone/water 7:3. Labeled flavanols
were finally dried on phosphoric anhydride under vacuum.
This procedure yielded 9.4 mg of (+)-catechin 1 (368 µCi/g),
3.2 mg of B3 dimer 2 (150 µCi/g), and 3.2 mg of C2 trimer 4
(144 µCi/g).
PA polymers 5 in the residual aqueous phase were treated
with AR2000, a pectinase preparation used in oenology to free
aroma precursors in wine and containing rhamnosidase,
arabinosidase, apiosidase, galactosidase, xylosidase, and â-glu-
cosidase activities (Bayonove et al., 1991), to eliminate co-
valently bound sugars before purification on Sephadex. The
enzyme preparation (5 mg) was added to the aqueous phase
(30 mL), and the mixture was kept at 37 °C for 18 h under
stirring. Water was then evaporated under reduced pressure,
and acetone (1 mL) was added to precipitate enzymes. After
centrifugation, the acetone-soluble PAs were dried and redis-
solved in water (5 mL). This solution was centrifuged again
to eliminate residual insoluble materials and applied to a
Sephadex LH-20 column (250 × 20 mm i.d.) first eluted with
water (100 mL) to remove released carbohydrates. Phenolic
acids (such as chlorogenic acid) were eluted with MeOH/water
1:1 (50 mL) (Lu and Foo, 1997) and PAs with acetone/H2O 7:3
(150 mL). The PA solution was concentrated and washed again
with n-hexane and EtOAc (when residual PA monomers were
still detected). Labeled PA polymers (90 mg, 177 µCi/g) were
finally dried on phosphoric anhydride under vacuum.
Mea su r em en t of Ra d ioa ctivity. The radioactivity of PAs
and their thiolysis products was determined with a Berthold
liquid scintillation spectrometer (LB 506 C) on-line with RP-
HPLC. The liquid scintillation cocktail, Optiflow Safe 1 (EG&G
Berthold) was delivered at the same flow rate as the HPLC
eluent (1 mL/min). The counting efficiency in the Z cell (vol. 2
mL) was on average 50%. Counts were corrected for back-
ground (estimated at 20 cpm). The detection threshold was
50 dpm. Specific activites of PAs were determined by reference
to the UV absorbance (280 nm) also measured on-line and
compared to the molar extinction of purified nonlabeled
reference compounds. The specific activity of the labeled
precursors was controlled with a Packard Tri-Carb 1500 liquid
scintillation analyzer (liquid scintillation cocktail: Ecolite,
ICN); the counting efficiency was 95%.
F igu r e 1. HPLC-ESI-MS chromatograms (negative ioniza-
tion) of an acetone/water extract of leaves harvested on an
adult willow tree: (A) total ion current; (B) flavanol monomers
(m/z ) 289); (C) flavanol dimers (m/z ) 577); (D) flavanol
trimers (m/z ) 865). (1) (+)-Catechin; (2) dimer B3; (3) dimer
B6; (4) trimer C2.
of (+)-catechin 1 (Figure 1B), two main dimers (Figure
1C), and three trimers (Figure 1D). The main dimer and
trimer eluted first on HPLC chromatograms and not
separated from each other were purified by chromatog-
raphy on Sephadex LH 20 and RP-HPLC and identified
as dimer B3 2 and trimer C2 4 by comparison of the
NMR spectra of their acetylated derivatives with those
of the literature (Balas and Vercauteren, 1994; Balas
et al., 1995). The second dimer was identified as dimer
B6 3, according to its HPLC retention time and previous
identification in the same extract (Thompson et al.,
1972).
Electronic autoradiographies of thin-layer chromatograms
were obtained with an Instant Imager (Packard) and a Storm
860 (Molecular Dynamics) electronic system. Dried cellulose
plates were imaged for 3-24 h to obtain a maximal error
coefficient of 2%. To test the sensitivity and linearity for
carbon-14 of the Instant Imager, 5 µL aliquots of a series of
[1,4-14C]-putrescine dilutions were applied to a cellulose TLC
plate and imaged for about 2 h. The counting efficiency was
1.2% and detection threshold 0.015 cpm/mm2. Calibration of
the Storm system was systematically performed with [U-14C]-
phenylalanine.
The radioactivity in solid tissues was determined by com-
bustion. Dried tissues stored in a desiccator containing silica
gel were placed in combusto-cones containing cellulose and
combusted for 4 s in an oxidizer (Packard, model 307). The
14CO2 produced was trapped in Carbosorb (Packard Instru-
ment Co.), and Permafluor (Packard) was added. The combus-
tion and scintillation counting efficiency was estimated at 85-
90% with solutions of 14C-atrazine.
Ra d iola beled F la va n -3-ols P r od u ced by In cor -
p or a tion of 14C-La beled P r ecu r sor s in Willow Tr ee
Lea ves. 14C-Labeled flavan-3-ols were analyzed in leaf
extracts after incorporation of [1-14C]-acetate or [U-14C]-
phenylalanine in willow shoots by cellulose bidimen-
sional TLC and RP-HPLC. (+)-Catechin 1, PA dimers
2 and 3, trimer 4 (further named “oligomers”), and
polymers 5 appearing as a streak from the origin spot
are clearly seen on the TLC chromatogram of the crude
extract and separated from labeled contaminants such
as phenylalanine precursor (14C-acetic acid is too volatile
and is removed by chromatography before further
analysis, see Experimental Section), sugars, and chlo-
rogenic acid isomers (Figure 2A). Radiolabeled com-
pounds were identified by comparing their retention
times and Rf values to those of nonlabeled molecules
purified from willow tree catkins. Ethyl acetate/water
partition of the crude extract allows one to separate
polymers 5 (aqueous phase) from (+)-catechin and PA
oligomers analyzed by RP-HPLC (Figure 2B).
RESULTS
P r oa n th ocya n id in s in Willow Tr ee Lea ves. A
crude acetone/water extract of leaves harvested on an
adult willow tree was analyzed by HPLC-ESI-MS.
HPLC traces on selected ions clearly show the presence
To more selectively assess the incorporation of carbon-
14 into PA polymers, they were depolymerized by
thiolysis (Scheme 1). Radiolabeled benzylthioether prod-

Biolabeling of proanthocyanidins
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4223
F igu r e 2. Chromatograms of willow leaf flavan-3-ols radio-
labeled with [1-14C]-acetate. (A) Autoradiogram of a bidimen-
sional cellulose thin-layer chromatogram of a crude acetone/
water extract. (B) HPLC chromatogram with UV and radio-
activity detection of an ethyl acetate extract. (1) (+)-Catechin;
(2) dimer B3; (3) dimer B6; (4) trimer C2; 5, proanthocyanidin
polymers; (6) sugars; (7) apolar contaminant.
F igu r e 3. Chromatograms of thiolysis products formed by
degradation of willow leaf procyanidins radiolabeled with
[1-14C]-acetate. (A) Autoradiogram of a bidimensionnal cel-
lulose thin layer chromatogram; (B) HPLC chromatogram with
UV and radioactivity detection. (1, 2) 3,4-cis and 3,4-trans
isomers of benzylthiocatechin; (3) 3,4-trans-benzylthioepicat-
echin; (4) (+)-catechin; (T) benzylmercaptan.
ucts were analyzed by cellulose bidimensional TLC or
RP-HPLC (Figure 3). More details on the analysis of
prodelphinidin thiolysis products have already been
published (De´prez et al., 1999).
Selection of P la n t Ma ter ia ls for Op tim a l P r oa n -
th ocya n id in Ra d iola belin g. To make plant material
available throughout the year for the labeling experi-
ments, young willow plants were grown in a greenhouse.
However, flavanols present in leaves were found to differ
in several respects from those of leaves harvested on
adult trees grown outdoors (Table 1). Adult tree leaves
were found to contain high amounts of (+)-catechin,
dimer B3, trimer C2, and procyanidin polymers and low
amounts of prodelphinidin polymers. In contrast, leaves
from plants grown in the greenhouse contained no dimer
and trimer but bigger amounts of polymers. Further-
more these polymers differ in their constitutive units
and are predominantly prodelphinidins. This observa-
tion, first deduced from thiolysis analysis, was con-
Sch em e 1. P r ocya n id in (R ) H) a n d P r od elp h in id in (R ) OH) Dep olym er iza tion by Th iolysis in to Ster eoisom er s
of Ben zyl Th io(ep i)Ca tech in (R ) H) a n d Ben zyl Th io(ep i)Ga lloca tech in (R ) OH) a n d F la va n -3-ols

4224 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
Ta ble 1. F la va n -3-ols in Lea ves of Willow P la n ts Gr ow n
Ou td oor s or in th e Gr een h ou se a n d Collected in Ma y (%
d r y w t)
Ta ble 2. Ra d ioa ctivity Tr a n sloca ted to Stem s a n d Lea ves
a fter Ad m in istr a tion of [1-14C]-Aceta te or
[U-14C]-P h en yla la n in e to Willow Sh oots
willows
precursor
adult tree
grown outdoors
young plant grown
in the greenhouse
phenylalaninea
acetateb
48d
labeling duration (h)
48d
72d
144d
72c
15
1.7
ndf
0.2
11.8
0.2
2
(+)-catechin
(+)-gallocatechin
B3 dimer
B6 dimer
C2 trimer
polymers
procyanidin units
prodelphinidin units
0.8
nda
0.5
0.1
0.2
0.1
0.1
nda
nda
nda
radioactivity (%)g
leaf extract
*precursor
*flavanol monomers
*flavanol polymers
*sugars
*apolar compounds
leaf residue
12
6.2
0.0
0.9
2.8
0.1
2
17
16
41
0.3
0.1
1.4
10.1
0.1
5
ndf
0.2
3.4
8.7
0.2
5
NDe
0.3
1.1
2.7
0.4
21
1.0
0.1
0.6
2.3
a
stem
86
78
79
83
38
nd: not detected.
a
b
25 µCi. 90 µCi. c Shoot collected on an adult tree grown
firmed by 13C NMR spectroscopy (Czochanska et al.,
1980): polymers isolated from adult trees showed major
signals at 116 (C-2′, C-5′), 120 (C-6′), and 145 ppm (C-
3′, C-4′) with a 2:1:2 intensity ratio characteristic of
procyanidins, whereas polymers isolated from plants
grown in the greenhouse showed signals at 116 (C-2′,
C-6′) and 145 ppm (C-3′, C-4′, C-5′) in a 2:3 ratio
characteristic of prodelphinidins.
outdoors. Shoot grown in a greenhouse. e Not determined due to
d
the volatility of the precursor. f Not detected. The total radioac-
g
tivity recovered was close to that administered to the shoot, and
for the sake of comparison, results are expressed as percent of
the total radioactivity analyzed.
Some variations in the proportion of procyanidins and
prodelphinidins were observed according to the sample
analyzed: procyanidins represented 70-90% of the total
PAs in adult tree leaves and 10-20% in the leaves of
young trees grown in the greenhouse. No nonextractible
PAs were detected by thiolysis in the residue of extrac-
tion recovered after filtration of the crude acetone/water
leaf extracts (Matthews et al., 1997).
The content and composition of leaf flavanols also
varied with the season of harvest. The PA content in
leaves collected on adult trees grown outdoors and
estimated by thiolysis coupled with RP-HPLC increased
from 0.55% to 1.14% dry wt. from April to J uly (fully
matured leaves). It remained constant between Febru-
ary and August in leaves collected on greenhouse trees.
Shoots harvested from adult trees were thus selected
for labeling (+)-catechin 1, dimer B3 2, trimer C2 4, and
procyanidin polymers 5 (R ) H). Young plants grown
in the greenhouse were preferred to label PA polymers
5 (largely prodelphinidins) in high yield.
Tr a n sloca t ion of R a d iola b eled P r ecu r sor s
th r ou gh th e P la n t. The radiolabeled precursor solu-
tions were administered by translocation through the
cut stem to shoots placed in a culture chamber equipped
with neon tubes and an air extractor. Sodium [1-14C]-
acetate and [U-14C]-phenylalanine were tested and the
amount of radioactivity recovered in the different tissues
of the shoots determined (Table 2).
The extent of precursor translocated through the
plant varied with the nature of the precursor. With
phenylalanine, about 80% of the radioactivity was
retained in the stem whereas with acetate the major
part of the radioactivity was found in leaves. The
amount of precursor reaching the leaves was higher in
the lower part of the shoot with both [1-14C]-acetate
(Figure 4) and [U-14C]-phenylalanine precursors. In an
experiment using the latter precursor, the stem of the
shoot and its eight leaves were separated into three
segments and analyzed separately (Table 3). The ra-
dioactivity measured by combustion of the whole samples
was 2-3 times higher in the basal segment (both leaves
and stem) as compared to the upper segment. The
radioactivity was also higher at the base of the leaves
F igu r e 4. Autoradiogram of a 20-cm-long shoot collected on
a young willow plant fed [1-14C]-acetate (50 µCi) through the
cut stem. Full desiccation was reached after 72 h.
Ta ble 3. Ra d ioa ctivity (µCi/g d r y w t. p la n t m a ter ia l)
F ou n d in Differ en t Segm en ts of a Willow Sh oot Gr ow n in
a Gr een h ou se, 144 h a fter Ad m in istr a tion of 25 µCi
[U-14C]-P h en yla la n in e. Low er , In ter m ed ia te, a n d Ap ica l
Segm en t Bor e Tw o, Tw o, a n d F ou r Lea ves, Resp ectively
segment
basal
intermediate
apical
stem
leaves
75.9
9.4
50.5
7.7
25.5
4.2
as compared to their apex. When leaves of shoot fed
[U-14C]-phenylalanine were cut into three partss
proximal, intermediate, and distalsradioactivity in the
proximal part was on average from 1.6 (apical leaves)
to 4.1 (basal leaves) times higher as compared to the
distal one. Incorporation of [1-14C]-acetate through the
cut petiole to excised isolated leaves showed a similar
heterogeneity in translocation (Figure 5).

Biolabeling of proanthocyanidins
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4225
The same shoots were then placed outdoors, and incor-
poration yields measured again in J uly reached 3.2%
and dropped to 0.05% in August as the metabolic
activity decreased. When the shoots were brought back
to the greenhouse in early September under sodium
lamp illumination, new leaves were formed 2 weeks
later and the incorporation yield at that time reached
1.0%. However, in the middle of October, yields dropped
again to values as low as 0.01%.
The length of labeling largely affected the incorpora-
tion yield which regularly increased in the 6 days
following administration of [U-14C]-phenylalanine (Table
2). In further labeling experiments, the leafy stems were
kept in the incorporation chamber until the leaves fully
wilt. However, large variations in the resistance of the
shoot to desiccation were observed. In practice, the
incorporation length varied from 2 to 8 days.
F igu r e 5. Autoradiogram of excised leaves collected on a
young willow plant and fed [1-14C]-acetate (50 µCi) through
cut petioles. Full desiccation was reached after 54, 20, and 20
h, respectively, for leaves from the left to the right.
A final factor affecting the yield of incorporation is
the nature of the plant sample, either shoots or excised
leaves. The incorporation of [1-14C]-acetate into PAs was
1.5-fold higher when the precursor was fed to excised
leaves rather than to leafy shoots, despite the reduction
by one-half of the labeling length explained by a more
rapid wilting (Table 6).
Scalin g-Up of th e In cor por ation P r ocedu r e. Once
the incorporation yields have been optimized, an easy
way to increase the specific activity of labeled flavanols
is to increase the amount of labeled precursor applied
(Table 7). No toxicity was observed when up to 200 µCi
(0.4 µmol) [U-14C]-phenylalanine was administered.
However, some toxic effects on shoots (blackening of the
leaf periphery, premature wilting) were observed when
too high amounts of sodium [1-14C]-acetate precursor
were applied. Administration of 1 mCi acetate (100 µmol
in 1 mL) reduced the incorporation yield into PAs and
sugars to 0.05% and 0.8%, respectively, both values
lower than those reached when only 400 µCi of the same
precursor were fed (0.4% and 4.8% respectively). No
apparent damage to leaves was observed at this last
concentration.
Ta ble 4. In cor p or a tion of [1-14C]-Aceta te (445 µCi) for 48
h in to Lea f F la va n ols of a Sh oot Ha r vested on a n Ad u lt
Willow Tr ee
concentration specific activity incorporation yield
(% dry leaves)
(µCi/g)
(% radioactivity fed)
(+)-catechin
B3 dimer
C2 trimer
procyanidin
polymers
0.53
0.31
0.19
0.55
79.6
65.9
58.9
23.0
0.11
0.05
0.03
0.03
The major part of the activity in leaves (60-83%)
was extracted by aqueous acetone, and most of the
radioactivity in the extract was shared between some
unmetabolized precursor, flavanols, sugars, and some
apolar compounds (Table 2). Radioactivity in the pre-
cursor was virtually nil 3 days after administration of
[U-14C]-phenylalanine, whereas the radioactivity in
flavanols, sugars, and apolar compounds increased in
the meantime. The total radioactivity recovered in these
experiments shows the absence of significative loss
through exhaust of carbon dioxide or volatile organic
compounds.
In cor p or a tion of Ra d iola beled P r ecu r sor s in to
F la va n -3-ols. The yield of incorporation of the precur-
sor differed according to the flavanol considered. With
shoots collected on adult trees, the precursor [1-14C]-
acetate was preferentially incorporated into (+)-cat-
echin, followed by dimer, trimer, and polymers (Table
4). The same results were obtained in all leaves of a
same shoot whatever their age (Figure 6). Specific
activities of (+)-catechin were also higher than those of
dimer B3 and trimer C2. Specific activities of the
flavanols were also affected by the position of the leaves
on the shoot. With the exception of the younger apical
leaves, these activities were higher in the basal leaves
than in the upper ones (Figure 6).
Incorporation of the precursors into PA polymers was
further examined by thiolysis degradation and the
specific activity of the polymer units determined (Table
5). Procyanidins and prodelphinidins were both labeled,
and specific activities varied by a factor of 2-3 according
to the polymer unit considered.
To further examine the toxicity of acetate and its
effect on its incorporation into PAs, a low amount of
[1-14C]-acetate (5 µmol) was fed to a shoot in the
presence or absence of larger amounts of nonlabeled
acetate (500 µmol). Addition of this amount of non-
labeled acetate did not induce any leaf blackening but
reduced the incorporation of 14C-acetate into PAs from
(1.62 ( 0.34)% to (0.91 ( 0.46)%. Neutralization with
10% HCl of the precursor solution before administration
only partially restored its incorporation to a (1.13 (
0.17)% value. The absorption rate of the precursor
solution was also reduced from 200 to 75 µL/min and
rose to 125 µL/min when acetate was neutralized.
Some typical incorporation results are presented in
Table 8. The yields of incorporation into (+)-catechin
and PAs reached 3% of the radioactivity fed, and the
specific activities varied from 200 to 500 µCi/g depend-
ing on the flavanol considered, the radioactivity admin-
istered, and the length of labeling.
Con ta m in a n ts a n d P u r ifica tion of Ra d iola beled
F la va n ols. The main radiolabeled contaminants are
sugars (free sugars account for 7% of the leaf dry weight;
the main sugars were sucrose (3.9%), glucose (3.1%),
fructose (0.3%), and galactose (traces)), chlorogenic acid,
and nonidentified apolar compounds. Large variations
in incorporation yields into these contaminants were
The season also had a major impact on flavanol
labeling. Shoots collected on the adult tree showed
maximal yield of incorporation into flavanols (3.2%) in
early J uly. This yield dropped to 0.05% in August.
Incorporation yields in shoots grown in the greenhouse
were maximal in May (3.4%) and decreased to 2.7% in
J une as the temperature in the greenhouse increased.

4226 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
F igu r e 6. Specific activities (A) and contents (B) of (+)-catechin, B3 dimer and C2 trimer in leaves of a willow shoot harvested
on an adult tree and fed [1-14C]-acetate (45 µCi) for 48 h: (white bars) (+)-catechin; (gray bars) B3 dimer and C2 trimer (not
separated in the HPLC system used).
Ta ble 5. In cor p or a tion of [1-14C]-Aceta te (90 µCi) for 48 h in to Lea f P r oa n th ocya n id in s of a Sh oot Gr ow n in th e
Gr een h ou se. P r oa n th ocya n id in s Wer e Su bm itted to Th iolysis a n d Ra d ioa ctivity Deter m in ed in Ea ch P r od u ct by
RP -HP LC
concentration
(% dry leaves)
specific activity
incorporation yield
(% radioactivity fed)
thiolysis products
(µCi/g)
Prodelphinidin Units
3,4-cis-benzylthiogallocatechin 9
3,4-trans-benzylthiogallocatechin 10
3,4-trans-benzylthioepigallocatechin 11
0.69
0.21
0.59
41.3
23.1
18.8
0.21
0.03
0.08
Procyanidin Units
3,4-cis-benzylthiocatechin 6
3,4-trans-benzylthiocatechin 7
3,4-trans-benzylthioepicatechin 8
0.17
29.6
nda
23.8
0.04
nda
0.04
nda
0.23
Terminal Units
0.09
(+)-catechin 1
13.9
0.01
a
nd = not detected.
Ta ble 6. P a r a m eter s of th e Biola belin g in Excised
Lea ves a n d Lea fy Sh oots Collected on You n g Willow
P la n ts a n d Ad m in ister ed [1-14C]-Aceta te (50 µCi),
Resp ectively, th r ou gh Cu t P etioles a n d Cu t Stem s.
Resu lts Ar e th e Mea n of Tr ip lica tes
Ta ble 7. Effect of th e Am ou n t of Ra d iola beled P r ecu r sor
Ad m in ister ed to a Willow Sh oot on th e Sp ecific Activity
of P r oa n th ocya n id in P olym er s (la belin g d u r a tion 48-96
h )
radioactivity
administered
(µCi)
proanthocyanidin
specific activity
(µCi/g)
excised leaf
leafy shoot
65 ( 35
200
precursor
labeling duration (h)
average absorption rate of the
precursor (µL/h)
31 ( 20
50
[U-14C]-phenylalanine
25
200
90
30
301
48
[1-14C]-acetate
Incorporation Yields (% radioactivity fed)
300
400
221
261
catechin
PA oligomers and polymers
non-PA contaminants
0.17 ( 0.12
2.31 ( 1.39
7.75 ( 4.07
0.11 ( 0.02
1.62 ( 0.34
7.19 ( 3.25
contained impurities, which were finally eliminated by
cellulose thin-layer chromatography.
Specific Activities (µCi/g)
498 ( 391
catechin
PA oligomers and polymers
63 ( 30
108 ( 41
Purification of polymers required an additional step.
Analysis of nonlabeled PA polymers purified by the
same procedure and analyzed by 13C NMR spectroscopy
showed prominent signals in the 60-105 ppm region
attributed to sugars. These sugars, likely covalently
linked to PAs (Achmadi et al., 1994), represent about
3-4% of the PA fraction as seen by hydrolysis with HCl
(Table 9). The main sugars are glucose, xylose, galac-
tose, and arabinose. To remove these bound sugars from
the labeled PAs, the PA fraction was treated with
enzymes. The highest yields of hydrolysis were obtained
with the AR2000 pectinase preparation. They were close
to those obtained by hydrolysis with HCl.
841 ( 566
observed (Table 2). With [U-14C]-phenylalanine, the
radioactivity incorporated into sugars varied from 3.8%
to 74.1% of the total activity in the leaf extract;
incorporation into chlorogenic acid and other noniden-
tified compounds reached up to 9.0% and 36.7%, respec-
tively. With [1-14C]-acetate, the major labeled contami-
nants were sugars (from 20.6% to 79.2% of the total
activity in the leaf extract); no labeled chlorogenic acid
could be detected.
Labeled dimers and trimer were separated from the
polymers by ethyl acetate/water partitioning. Flavan-
3-ols in both EtOAc and aqueous extracts were easily
separated from labeled sugars by column chromatog-
raphy on Sephadex LH-20. Some oligomer fractions still
Purification yields are shown in Table 10. Each
Sephadex chromatography resulted in a loss of 20-40%
of flavanols. The overall yield of purification varied
between 30% and 60%.

Biolabeling of proanthocyanidins
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4227
Ta ble 8. Recover y of 14C-la beled (+)-ca tech in a n d P r oa n th ocya n id in Dim er , Tr im er a n d P olym er fr om Lea ves of Willow
Cu ttin gs a fter Th r ee In d ep en d en t In cor p or a tion s of La belled P h en yla la n in e or Aceta te. Len gth of La bellin g Va r ied
betw een 72 a n d 80 h
14C-precursor
(amount administered, µCi)
amount in
leaf extract (mg)
total activity
specific activity
incorporation yield
(% radio fed)
14C-Producta
(µCi)
(µCi/g)
[U-14C]-phenylalanine (200)
[1-14C]-acetate (400)
polymersb (2/8)
(+)-catechin
B3 dimer
18
9
11
11
70
19
5.3
4.8
2.5
2.2
12.7
9.5
301
543
234
198
180
506
2.7
1.2
0.6
0.5
3.2
2.7
C2 trimer
polymers (9/1)
[1-14C]-acetate (350)
polymersb (2/8)
a
Values in parentheses represent the ratio of procyanidin to prodelphinidin units in proanthocyanidin polymers. Experiments showing
a high procyanidin/prodelphinidin ratio were carried out on shoots collected on an adult tree; others with a low ratio were carried out on
b
shoots grown in the greenhouse. Free of monomer flavanol contaminant but may contain some dimers and trimers.
13C-phenylalanine were easily translocated through the
Ta ble 9. Su ga r s Rem oved by En zym a tic a n d Acid
Hyd r olyses fr om th e P r oa n th ocya n id in s Isola ted fr om
Willow Lea ves (m g/g p r oa n th ocya n id in )
roots of Fagus grandifoliia, whereas 13C-ferulic acid fed
in similar conditions did not reach the aerial parts of
the plant (Lewis et al., 1990). Administration of 14C-
phenylalanine to a sorghum panicle showed that 50%
of the radioactivity remained in the stem and did not
reach the seeds (Reddy and Butler, 1989). Translocation
of [U-14C]-phenylalanine and [3-14C]-cinnamic acid fed
to the detached leaves of Polygonum or wheat was
largely limited to their basal part (Barnes and Friend,
1975). In the present study, 78-86% of the radioactivity
applied as 14C-phenylalanine fed to willow shoots was
found in the stem and did not reach the leaves (Table
2), possibly because of some incorporation into insoluble
polymeric lignins. Better translocation was observed
with [1-14C]-acetate, but no increase of incorporation in
both flavanols and nonphenolic compounds could be
observed (compare incorporations at 48 h for both
precursors, Table 2). This tends to indicate that a major
part of acetate has not been metabolized. There is thus
no clear evidence for preferring one or the other precur-
sor on the basis of their translocation.
sugar
â-glucosidase
cellulase
AR2000
HCl
glucose
xylose
galactose
arabinose
total
5.4
0.0
0.8
0.0
6.2
12.3
2.8
1.0
0.0
16.1
14.3
17.6
2.1
3.0
37.0
21.9
7.7
1.3
4.8
35.7
DISCUSSION
Various parameters, such as the nature of the plant
material, the nature and quantity of the precursor, and
the method of incorporation, affect the incorporation
yield of the labeled precursor into both polyphenols and
nonphenolic compounds. It is essential to optimize these
parameters in order to obtain high specific activities and
high purity of the labeled polyphenols.
Selection of P la n t Ma ter ia l for P r oa n th ocya n i-
d in Ra d iola belin g. Haslam et al. found high amounts
of (+)-catechin 1, dimer B3 2, and trimer C2 4 in willow
tree catkins (Thompson et al., 1972). However, catkins
are available for only a short period of the year and
could not be used for labeling experiments. Leaves on
adult trees showed similar contents of the same mol-
ecules, but the season for labeling was still limited to
about 4 months (April-J uly). To further extend this
period, young willow trees were grown in a greenhouse,
and labeling could thus be carried out from February
to October, but the composition of the leaves differed
from that of leaves harvested on adult trees by the
absence of monomer, dimer, and trimer and the pres-
ence of large amounts of PA polymers, largely prodel-
phinidins. Leaves of these young willow shoots were
thus used to label PA polymers 5 and adult tree to label
(+)-catechin and PA dimer and trimer.
This difference in flavanol composition was not ex-
plained by the change in the environmental conditions.
Indeed, the PA pattern in leaves was not changed when
cuttings from the adult tree were grown in the green-
house or when the plants from the nursery initially
grown in the greenhouse were moved outdoors. The
difference in PA composition is thus explained by
genetic polymorphism (Harborne and Turner, 1984).
Similarly, two chemomorphs with a prodelphinidin
content of 55% and 90% of total PAs in needles have
been decribed in Scotch pines (Lebreton et al., 1990).
Tr a n sloca t ion of R a d iola b eled P r ecu r sor s
th r ou gh th e P la n t. Once the plant material has been
selected, the incorporation of the precursor into the
compounds of interest will depend on its translocation
to the metabolically active tissues. Translocation de-
pends on the nature of the precursor. Solutes such as
In cor p or a tion of Ra d iola beled P r ecu r sor s in to
F la va n -3-ols. Incorporation yields into flavanols de-
pend on the nature of the precursor, on the more or less
active biosynthesis of PAs in the plant tissues, and on
the competing incorporation of the precursor into non-
phenolic products. Specific radioactivities of PAs will
depend on the yield of incorporation and on the dilution
with preexisting nonlabeled molecules. To increase both
the incorporation yield and specific activities, it is
important to extend as much as possible the duration
of metabolism and to feed high amounts of radiolabeled
precursor.
Different studies have shown large variations in
incorporation yields according to precursors. 14CO2 was
3-5 times less efficient than 14C-acetate to label (+)-
catechin in Uncaria gambir (Das and Griffiths, 1967).
14CO2 was also less effective than 14C-phenylalanine
when fed to sorghum to label condensed tannins (J ime-
nez-Ramsey et al., 1994). On the other hand, shikimic
acid was 20 times more effective than L-phenylalanine
in the labeling of gallic acid (Zaprometov and Bukhlae-
va, 1968). In case of willow shoots, we could not observe
any significant difference between phenylalanine and
acetate. Since [U-14C]-phenylalanine was about 10 times
more expensive than [1-14C]-acetate, the latter was
preferred.
Large variations of PA specific activities were ob-
served according to the position of the leaves on the
shoot. They were 2-3 times higher in the lower leaves
as compared to the upper leaves (Figure 6A). This may
be explained by the limited translocation of the precur-

4228 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
Ta ble 10. P u r ifica tion Yield s of Ra d iola beled F la va n -3-ol Mon om er a n d P olym er s. F la va n ol Mon om er s Wer e Isola ted
fr om Lea ves of a n Ad u lt Tr ee a n d P olym er s fr om Th ose of You n g Willow Tr ees Gr ow n in a Gr een h ou se
(+)-catechin
PA polymers
PA polymers
flavanol precursor
[1-14C]-acetate (n ) 2)a
[1-14C]-acetate (n ) 6)a
[U-14C]-phenylalanine (n ) 2)a
Purification step (% of total labeled flavanol)
sephadex LH-20b + solvent fractionation
solvent fractionation
59
71
76
81
76
enzyme hydrolysisc + Sephadex LH-20
sephadex LH-20
60
83
29
0.1
cellulose TLC
All steps (% of total labeled flavanol)
Yield (% leaf dry wt.)
45
3.2
62
2.6
b
an: Number of repetitions. Residual acetate precursor is removed by a first Sephadex LH-20 chromatography. c Polymers are treated
by enzymes to remove bound sugars.
sor to the upper leaves (Figure 4, Table 3), the loss of
some precursor in the stem (Table 2), or a higher
biosynthetic activity in the lower parts of the shoot. The
higher concentration of flavanols in the lower leaves
(Figure 6B) shows that PAs continue to accumulate once
the leaves are fully developed. An attempt to use these
lower leaves to directly feed the precursor through the
petiole was carried out. The total incorporation yield
into flavanols (2.5%, Table 6) was of the same order than
the best ones obtained with the experiments on shoots
(Table 8). Specific activities were also significantly
increased in detached leaves despite the larger dilution
by the large content of nonlabeled PAs (Figure 6B).
Yields of incorporation also varied according to the
state of maturity of the shoot at the time of harvest.
Variations in the nature and content of polyphenols in
leaves according to their state of maturity have com-
monly been reported (Feeny and Bostock, 1968; Tissut,
1968; Scalbert and Haslam, 1987; Koupai-Abyazani et
al., 1993; Bohm et al., 1994; Lister et al., 1994).
Variations of the biosynthetic activity of the flavanols
with time was thus expected. The PA content in leaves
of the willow tree grown outdoors increased from April
to J uly (fully matured leaves). The incorporation of
[1-14C]-acetate into PAs was highest (3%) in early J uly.
In general, total incorporation yields into PAs com-
monly reached 3-5% and specific activities 200-500
µCi/g (Table 8). These values are consistent with the
best results reported in the literature. Specific activities
of 0.8 (J imenez-Ramsey et al., 1994) and 290 mCi/mmol
(Harmand and Blanquet, 1978) (calculated on a catechin
unit basis) were reported for condensed tannins. Our
own values of 147 µCi/mmol for PA polymers (calculated
on a polymer unit basis) compare well with those values.
Previous works on PA labeling show that less than 0.1%
of [1-14C]-acetate was converted into procyanidins of the
A type (Eastmond and Gardner, 1974) and 0.5% of
[U-14C]-phenylalanine into condensed tannins (Reddy
and Butler, 1989; see De´prez and Scalbert (1999) for a
review).
that were unlabeled and would then add progressively
more radiolabeled units.
Specific activities in procyanidin or prodelphinidin
units of PA polymers were not significantly different,
but they did vary with the unit stereochemistry. Thi-
olysis of the radiolabeled PAs showed that benzylthio-
ethers with a 2,3-trans stereochemistry (catechin 6, 7
and gallocatechin 9, 10) were more highly labeled than
those with a 2,3-cis stereochemistry (epicatechin 8 and
epigallocatechin 11). These differences reflect a late
accumulation of 2,3-trans units as compared to 2,3-cis
ones. Similar conclusions were reached by comparing
the composition of PAs in leaves of Onobrychis at
different stages of maturity (Koupai-Abyazani et al.,
1993). These differences of specific activities are then
the result of the PA biosynthetic pathway and of its
varying expression over time which leads to the ac-
cumulation of PAs with definite structures.
Con ta m in a n ts a n d P u r ifica tion of Ra d iola beled
F la va n ols. When using radiolabeled molecules as trac-
ers, it is essential to have molecules of high purity,
otherwise one might trace the impurities. The use of
plants for labeling polyphenols usually results in the
formation of large amounts of nonphenolic labeled
contaminants which need to be carefully eliminated. No
cheap precursor allows, so far, one to channel the
radioactivity specifically to PAs or other polyphenols of
interest.
The nature of plant materials used for labeling
polyphenols affects the relative importance of incorpora-
tion of the precursor into polyphenols or contaminants.
Incorporation of labeled phenylalanine into sugar con-
taminants was much lower when the precursor was fed
to whole leaves rather than to leaf disks (Hillis and Isoi,
1965). In the present willow tree materials, labeled
sugars were the main impurities. They accounted for
4-80% of the radioactivity in the leaf extract depending
on the choice of the precursor and the labeling duration
and correspond to 2-7 times more radioactivity than
labeled flavanols (Table 2). Not only acetate but also
phenylalanine was incorporated into sugars. A similar
conversion of phenylalanine to sugar was shown in
Eucalyptus species (Hillis and Isoi, 1965). A more
specific labeling of polyphenols with phenylalanine could
possibly be acheived with shorter incorporations, but
their specific activities would be reduced consequently.
Few authors have paid attention to the purification
of radiolabeled PA polymers. In the present study,
purification by chromatography on a Sephadex LH-20
column and cellulose TLC led to the isolation of (+)-
catechin, PA dimer B3, and trimer C2 with a radio
purity close to 100%, as shown by HPLC and bidimen-
sional TLC checking.
Heter ogen eou s In cor por ation of P r ecu r sor s in to
P r oa n th ocya n id in s. Incorporation yields and specific
activities vary with the flavanol considered or the unit
in the polymer. Specific activities decreased from mono-
mer to polymers (Table 4). In polymers, degradation by
thiolysis showed that less radioactivity was incorporated
into PA terminal units than into internal units (Table
5). Similar observations were previously made on Dou-
glas fir PAs and led Stafford to propose a biosynthetic
scheme for PAs (Stafford and Lester, 1982). This scheme
implies the existence of a mode of polymerization
consisting of the incorporation of labeled catechin
molecules into preexisting PA polymers or oligomers

Biolabeling of proanthocyanidins
J. Agric. Food Chem., Vol. 47, No. 10, 1999 4229
Purification of radiolabeled PA polymers required an
additional treatment with enzymes to remove some
bound sugars which, in contrast to free sugars, could
not be removed by simple chromatography on Sephadex
LH-20. Glycosylated PAs have been described on several
occasions (Porter et al., 1985; Gujer et al., 1986; Zhang
et al., 1988; Achmadi et al., 1994). Bound sugars must
be removed in order to study the intestinal absorption
of these radiolabeled PA polymers which would other-
wise be removed by glycosidases in the gut (Griffiths
and Smith, 1972; Day et al., 1998) and their absorption
confused with that of PAs.
Various commercial glycosidase preparations were
tested, and AR2000 was found to release the highest
amount of sugars (Table 9). The total amount (3.7%) and
the proportion of the four sugars are close to those
observed by hydrochloric acid hydrolysis. The higher
amount of xylose, liberated by AR2000 enzymes as
compared to the hydrochloric acid treatment could be
explained by the relative instability of xylose in the
conditions of acid hydrolysis (Barton et al., 1982).
Yields of purified PA polymers recovered at the end
of the purification process (3.2% and 2.6% leaf dry wt.,
depending on the precursor, Table 10) can be compared
to the initial content of PA polymers as determined by
thiolysis (Table 1). These purification yields and the
yield of purification at each step of the protocole (Table
10) allow one to calculate the actual content of PA
polymers in the leaves, i.e., 7.1% and 4.2% depending
on the precursor. These PA contents are higher than
those directly determined by thiolysis on leaves (2.9%
leaf dry wt., Table 1). This apparent discrepancy is
explained by an underestimation of the PA content of
leaves when assayed by thiolysis. Yields of depolymer-
ization of PAs by thiolysis varied between 30% and 70%
depending on the PA samples due to the presence of
thiolysis-resistant interunit bonds (Matthews et al.,
1997). It can be concluded that the actual PA content
of leaves is closer to 4-7%.
to their greenhouse facilities, to Mr. Hachette (Arbore-
tum de Che`vreloup, Rocquencourt) for allowing us to
collect catkins and shoots on an adult willow tree in
their collection, and to Pamela Scalbert-Menanteau for
corrections on the manuscript.
LITERATURE CITED
Achmadi, S.; Syahbirin, G.; Choong, E. T.; Hemingway, R. W.
Catechin-3-O-rhamnoside chain extender units in polymeric
procyanidins from mangrove bark. Phytochemistry 1994, 35,
217-219.
Balas, L.; Vercauteren, J . Extensive high-resolution reverse
2D NMR Analysis for the structural elucidation of procya-
nidin oligomers. Magn. Reson. Chem. 1994, 32, 386-393.
Balas, L.; Vercauteren, J .; Laguerre, M. 2D NMR structure
elucidation of proanthocyanidins: The special case of the
catechin-(4 alpha-8)-catechin-(4 alpha-8)-catechin trimer.
Magn. Reson. Chem. 1995, 33, 85-94.
Barnes, C.; Friend, J . The lack of movement of [14C]-phenyl-
alanine and [14C]-cinnamate after administration to leaves
of Polygonum and wheat. Phytochemistry 1975, 14, 139-
142.
Barton, F. E. I.; Windham, W. R.; Himmelsbach, D. S. Analysis
of sugar hydrolysates of forage cell walls by high-pressure
liquid chromatography. J . Agric. Food Chem. 1982, 30,
1119-1123.
Bayonove, C.; Gunata, Y. Z.; Sapis, J . C.; Baumes, R. L.;
Dugelay, I.; Grassin, C. Augmentation des aroˆmes dans le
vin et utilisation d′enzymes. Rev. Oenol. 1991, 4.
Bohm, B. A.; Singh, S.; Koupai-Abyazani, M. R. Biosynthesis
and turnover of tannins in sainfoin (Onobrychis viciifolia).
Polyphe´nols Actualite´s 1994, 11, 18.
Czochanska, Z.; Foo, L. Y.; Newman, R. H.; Porter, L. J .
Polymeric proanthocyanidins. Stereochemistry, structural
units and molecular weight. J . Chem. Soc., Perkin Trans. 1
1980, 2278-2286.
Das, N. P.; Griffiths, L. A. Studies on flavonoid metabolism.
Biosynthesis of (+)-[14C]-catechin by the plant Uncaria
gambir Roxb. Biochem. J . 1967, 105, 73-77.
Day, A. J .; DuPont, M. S.; Ridley, S.; Rhodes, M.; Rhodes, M.
J . C.; Morgan, M. R. A.; Williamson, G. Deglycosylation of
flavonoid and isoflavonoid glycosides by human small
intestine and liver â-glucosidase activity. FEBS Lett. 1998,
436, 71-75.
CONCLUSION
Administration of radiolabeled precursors, either
acetate or phenylalanine, to willow tree shoots led to
the isolation of PA dimer, trimer, and polymer of high
radiopurity. Such a method provided labeled molecules
which cannot yet be easily obtained through synthetic
routes due to their high structural complexity.
De´prez, S.; Mila, I.; Scalbert, A. Complete qualitative and
quantitative analysis of procyanidins and prodelphinidins
in plant extracts by thiolysis. J . Agric. Food Chem. 1999,
submitted for publication.
De´prez, S.; Scalbert, A. Isotopic labeling of dietary polyphenols
for bioavailability studies. In Plant Polyphenols. 2. Chem-
istry and biology; Gross, G. G., Hemingway, R., Yoshida,
T., Eds.; Plenum Press: New York, 1999; in press.
Eastmond, R.; Gardner, R. J . [14C]-Epicatechin and [14C]-
procyanidins from seed shells of Aesculus hippocastaneum.
Phytochemistry 1974, 13, 1477-1478.
Feeny, P. P.; Bostock, H. Seasonal changes in the tannin
content of oak leaves. Phytochemistry 1968, 7, 871-880.
Griffiths, L. A.; Smith, G. E. Metabolism of myricetin and
related compounds in the rat. Metabolite formation in vivo
and by the intestinal microflora in vitro. Biochem. J . 1972,
130, 141-151.
Gujer, R.; Magnolato, D.; Self, R. Glucosylated flavonoids and
other phenolic compounds from Sorghum. Phytochemistry
1986, 25, 1431-1436.
Biolabeling methods with excised tissues have, how-
ever, their limits, the major one being the limited
incorporation of the precursor into the molecule of
interest when plant survival does not exceed a few days.
An alternative would be to grow different tissues with
the appropriate labeled precursor, either plant cells with
the same precursors or entire plants grown in pots
under a 14CO2 atmosphere. These methods require a
much higher investment both in time and money to
establish and grow the cell cultures or to feed the
radiolabeled gas in a closed growth chamber. The
method developed here allowed us to obtain in a
relatively short delay and at a relatively low cost several
pure labeled PAs suitable for bioavailability studies.
Current work on their bioavailability in rodents and on
their metabolism and absorption by human liver and
intestinal cell lines is now being carried out.
Harborne, J . B. Plant Polyphenols. XIV. Characterization of
flavonoid glycosides by acidic and enzymatic hydrolysis.
Phytochemistry 1965, 4, 107-120.
Harborne, J . B.; Turner, B. L. Plant Chemosystematics;
Academic Press: London, 1984.
We are grateful to Catherine L. M. Herve´ du Penhoat
(CNRS Grenoble) for NMR analyses, to J . M. Chartier,
M. Lauransot and G. Raynal (INRA Grignon) for access
Harmand, M. F.; Blanquet, P. The fate of total flavanolic
oligomers (OFT) extracted from Vitis vinifera L. in the rat.
Eur. J . Drug Metabol. Pharmacokinet. 1978, 3, 15-30.

4230 J. Agric. Food Chem., Vol. 47, No. 10, 1999
De´prez et al.
Hertog, M. G. L. Flavonols in wine and tea and prevention of
coronary heart disease. In Polyphenols 96; Vercauteren, J .,
Che`ze, C., Triaud, J ., Eds.; INRA Editions: Paris, 1997; pp
119-131.
Ricardo da Silva, J . M.; Cheynier, V.; Souquet, J . M.; Mout-
ounet, M.; Cabanis, J . C.; Bourzeix, M. Interaction of grape
seed procyanidins with various proteins in relation to wine
fining. J . Sci. Food Agric. 1991, 56, 111-125.
Ricardo da Silva, J . M.; Rosec, J . P.; Bourzeix, M.; Mourgues,
J .; Moutounet, M. Dimer and trimer procyanidins in Cari-
gnan and Mourvedre grapes and red wines. Vitis 1992, 31,
55-63.
Scalbert, A.; Duval, L.; Peng, S.; Monties, B. Polyphenols of
Quercus robur L. II. Preparative isolation by low-pressure
and high-pressure liquid chromatography of heartwood
ellagitannins. J . Chromatogr. 1990, 502, 107-119.
Scalbert, A.; Haslam, E. Polyphenols and chemical defence of
the leaves of Quercus robur. Phytochemistry 1987, 26, 3191-
3195.
Shahidi, F.; Naczk, M. Food phenolics, sources, chemistry,
effects, applications; Technomic Publishing Co. Inc.: Lan-
caster, 1995.
Hillis, W. E.; Isoi, K. The biosynthesis of polyphenols in
Eucalyptus species. Phytochemistry 1965, 4, 905-918.
J imenez-Ramsey, L. M.; Rogler, J . C.; Housley, T. L.; Butler,
L. G.; Elkin, R. G. Absorption and distribution of C-14-
labeled condensed tannins and related sorghum phenolics
in chickens. J . Agric. Food Chem. 1994, 42, 963-967.
Koupai-Abyazani, M. R.; McCallum, J .; Muir, A. D.; Bo¨hm, B.
A.; Towers, G. H. N.; Gruber, M. Y. Develpmental changes
in the composition of proanthocyanidins from leaves of
sainfoin (Onobrychis viciifolia Scop.) as determined by
HPLC analysis. J . Agric. Food Chem. 1993, 41, 1066-1077.
Ku¨hnau, J . The flavonoids: a class of semi-essential food
components: their role in human nutrition. World Rev.
Nutr. Diet. 1976, 24, 117-191.
Laparra, J .; Michaud, J .; Masquelier, J . Etude pharmacocine´-
tique des oligome`res flavanoliques. Plant Med. Phytother.
1977, 11, 133-142.
Stafford, H. A.; Lester, H. H. Enzymic and nonenzymic
reduction of (+)-dihydroquercetin to its 3,4-diol. Plant
Physiol. 1982, 70, 695-698.
Steinmetz, K. A.; Potter, J . D. Vegetables, fruits and cancer.
I. Epidemiology. Cancer Causes Control 1991, 2, 325-357.
Swain, T. Economic importance of flavonoid compounds:
foodstuffs. In The chemistry of flavonoid compounds; Geis-
sman, T. A., Ed.; Pergamon Press: Oxford, 1962; pp 513-
552.
Terrill, T. H.; Waghorn, G. C.; Woolley, D. J .; McNabb, W. C.;
Barry, T. N. Assay and digestion of 14C-labeled condensed
tannins in the gastrointestinal tract of sheep. Brit. J . Nutr.
1994, 72, 467-477.
Thompson, R. S.; J acques, D.; Haslam, E.; Tanner, R. J . N.
Plant proanthocyanidins-Part 1-Introduction: the isola-
tion, structure and distribution in nature of plant procya-
nidins. J . Chem. Soc., Perkin Trans. 1 1972, 1387-1399.
Tissut, M. Etude du cycle annuel des de´rive´s phe´noliques de
la feuille de heˆtre, Fagus sylvatica L. Physiol. Veg. 1968, 6,
351-365.
Treutter, D.; Santos-Buelga, C.; Gutmann, M.; Kolodziej, H.
Identification of flavan-3-ols and procyanidins by high-
performance liquid chromatography and chemical reaction
detection. J . Chromatogr. A 1994, 667, 290-297.
Zaprometov, M. N.; Bukhlaeva, V. Y. Two pathways of bio-
synthesis of gallic acid. Biochemistry 1968, 33, 317-319.
Zhang, B.; Nonaka, G.-I.; Nishioka, I. Potentillanin, a bifla-
vanoid and procyanidin glycoside from Potentilla viscosa.
Phytochemistry 1988, 27, 3277-3280.
Lebreton, P.; Laracine-Pittet, C.; Bayet, C.; Lauranson, J .
Variabilite´ polyphe´nolique et syste´matique du pin sylvestre
Pinus sylvestris L. Ann. Sci. For. 1990, 47, 117-130.
Lewis, N. G.; Inciong, M. E.; Razal, R. A.; Yamamoto, E.;
Davin, L. B. Monolignol biogenesis, lignin structure and
biodegradation; 15 e`mes J ourne´es Internationales d'Etude
des Polyphe´nols: Strasbourg, 1990; pp 365-372.
Lister, C. E.; Lancaster, J . E.; Sutton, K. H. Developmental
changes in the concentration and composition of flavonoids
in skin of a red and a green apple cultivar. J . Sci. Food
Agric. 1994, 64, 155-161.
Lu, Y.; Foo, L. Y. Identification and quantification of major
polyphenols in apple pomace. Food Chem. 1997, 59, 187-
194.
Markham, K. R. Techniques of flavonoid identification; Aca-
demic Press: London, 1982.
Matthews, S.; Mila, I.; Scalbert, A.; Donnelly, D. M. X.
Extractible and nonextractible proanthocyanidins in barks.
Phytochemistry 1997, 45, 405-410.
Matthews, S.; Mila, I.; Scalbert, A.; Pollet, B.; Lapierre, C.;
Herve´ du Penhoat, C. L. M.; Rolando, C.; Donnelly, D. M.
X. Method for estimation of proanthocyanidins based on
their acid depolymerization in the presence of nucleophiles.
J . Agric. Food Chem. 1997, 45, 1195-1201.
Partridge, S. M. Aniline hydrogen phthalate as a spraying
reagent for chromatography of sugars. Nature 1949, 164,
443.
Porter, L. J .; Foo, L. Y.; Furneaux, R. H. Isolation of three
naturally occurring O-â-D-glucopyranosides of procyanidin
polymers. Phytochemistry 1985, 24, 567-569.
Received for review December 21, 1998. Revised manuscript
received May 26, 1999. Accepted May 26, 1999. This work was
supported by the European Community (FAIR CT Contract
95-0653).
Reddy, V.; Butler, L. G. Incorporation of 14C from [14C]-
phenylalanine into condensed tannin of sorghum grain. J .
Agric. Food Chem. 1989, 37, 383-384.
J F981380Z