Analysis of ethylidene-bridged flavan-3-ols in wine

Article abstract of DOI:10.1021/jf0626258

A method was developed to determine the amount of ethylidene-bridged flavan-3-ols in wine. The method was based upon the analysis of 2,2 -ethylidenediphloroglucinol (EDP), a product formed after acid-catalyzed cleavage of wine flavan-3-ols in the presence of excess phloroglucinol. In the developed analytical method, the wine was purified and concentrated using C ₁₈ solid-phase extraction before the phloroglucinolysis reaction was carried out. This procedure was used to quantify ethylidene-bridged flavan-3-ols in wine and the molar ratio between ethylidene-bridged linkages and native interflavan linkages. The method validation showed 9.2% repeatability. The recovery of the ethylidenebridged flavanols in wine was 90% for concentrations up to 4.5 mg-L^-1 of ethylidene-bridged linkages, and it decreases to 83% above and until the concentration reached 7.6 mg-L^-1. Initial results showed that the concentration of ethylidene-bridged flavan-3-ols measured in wines was very low (less than 1.3 mg-L^-1) and that they represented less than 1.3% of the total interflavonoid linkages on a molar ratio.

Full text of DOI:10.1021/jf0626258

J. Agric. Food Chem. 2007, 55, 1109−1116 1109
Analysis of Ethylidene-Bridged Flavan-3-ols in Wine
JESSICA DRINKINE,† PAULO LOPES,† JAMES A. KENNEDY,‡
PIERRE-LOUIS TEISSEDRE,† AND CEDRIC SAUCIER*,†
Faculte´ d’Oenologie de Bordeaux, Universite´ Victor Segalen Bordeaux 2, UMR 1219 INRA ISVV,
351 Cours de la Libe´ration, 33405 Talence Cedex, France, and Department of Food Science and
Technology, Oregon State University, 220A Weigand Hall, Corvallis, Oregon 97331-6602
A method was developed to determine the amount of ethylidene-bridged flavan-3-ols in wine. The
method was based upon the analysis of 2,2′-ethylidenediphloroglucinol (EDP), a product formed after
acid-catalyzed cleavage of wine flavan-3-ols in the presence of excess phloroglucinol. In the developed
analytical method, the wine was purified and concentrated using C18 solid-phase extraction before
the phloroglucinolysis reaction was carried out. This procedure was used to quantify ethylidene-
bridged flavan-3-ols in wine and the molar ratio between ethylidene-bridged linkages and native
interflavan linkages. The method validation showed 9.2% repeatability. The recovery of the ethylidene-
bridged flavanols in wine was 90% for concentrations up to 4.5 mg‚L-1 of ethylidene-bridged linkages,
and it decreases to 83% above and until the concentration reached 7.6 mg‚L-1. Initial results showed
that the concentration of ethylidene-bridged flavan-3-ols measured in wines was very low (less than
1.3 mg‚L-1) and that they represented less than 1.3% of the total interflavonoid linkages on a molar
ratio.
KEYWORDS: Tannins; flavan-3-ol; ethylidene bridge; red wine; aging; 2,2′-ethylidenediphloroglucinol;
EDP; phloroglucinolysis
INTRODUCTION
production (e.g., Saccharomyces cereVisiae) (13) and from the
oxidation of ethanol (14). In order to determine the importance
of acetaldehyde in wine aging, it would be useful to have
information on the quantity of ethylidene-bridged flavan-3-ols
in wine.
Phenolic compounds play an important role in wine quality
because of their color and taste properties. Condensed tannins
influence bitterness and astringency (1, 2) and are involved in
wine colloidal and color stability (3-6). In the grape, they exist
as polymers of flavan-3-ol units [(+)-catechin, (-)-epicatechin,
(-)-epigallocatechin, and (-)-epicatechin-3-O-gallate] with
C4-C6 or C4-C8 linkages (7).
During winemaking and aging, condensed tannins undergo
enzymatic or chemical modifications, and many of these
modifications result in the formation of new interflavonoid
linkages. One such modification is the acid-catalyzed cleavage
of the original plant-derived interflavanoid bond and a subse-
quent condensation reaction (4, 8, 9). Another type of modifica-
tion is nucleophilic substitution reactions involving the C6 or
C8 of the A-ring of multiple flavan-3-ol units (10). Acetaldehyde
has been identified as one of the most important electrophiles
in this type of bridging reaction (3, 11). The ethylidene-bridged
flavan-3-ol products of this reaction have also been referred to
as ethyl-bridged and methyl-methine-bridged polymers. To
date, only dimers and trimers of ethylidene-bridged flavan-3-
ols have been identified in wine (12). The acetaldehyde involved
in this reaction may be produced in two ways: by yeast
Condensed tannins are complex flavan-3-ol polymers, and
in general, only methods based upon their cleavage provide
compositional information. Several methods such as thiolysis
(15) or phloroglucinolysis (16, 17) have historically been used
for compositional analysis. Phloroglucinolysis is based upon the
acid-catalyzed cleavage of flavan-3-ol interflavanoid bonds in
the presence of phloroglucinol. The reaction releases the terminal
unit as a flavan-3-ol monomer and forms C4 phloroglucinol
adducts of the extension units (Figure 1). The products are then
analyzed by reversed-phase HPLC. It is thus possible insofar
as the cleavage is complete, to determine flavan-3-ol polymer
composition and mean average degree of polymerization (mDP).
The purpose of the present work was to develop a method
for the detection of ethylidene-bridged flavan-3-ols by conduct-
ing phloroglucinolysis on the flavan-3-ols to form an ethylidene-
bridged phloroglucinol adduct.
MATERIALS AND METHODS
Materials. Deionized water was purified with a Milli-Q water system
(Millipore, Bedford, MA). Acetonitrile (HPLC grade) was obtained
from Fisher Chemicals (Elancourt, France); ethyl alcohol (HPLC grade)
was from Carlo Erba (Val de Reuil, France); acetaldehyde (RP) was
from Riedel-De Hae¨n (Val de Reuil, France); and methanol (HPLC
* Corresponding author. Fax: 33-5-40-00-6468. E-mail: saucier@
oenologie.u-bordeaux2.fr.
† Universite´ Victor Segalen Bordeaux 2.
‡ Oregon State University.
10.1021/jf0626258 CCC: $37.00
© 2007 American Chemical Society
Published on Web 01/19/2007

1110 J. Agric. Food Chem., Vol. 55, No. 4, 2007
Drinkine et al.
B in 12.5 min, from 50 to 100% B in 0.5 min, 100% B for 3 min, from
100 to 0% B in 0.5 min, 0% B for 5 min. Peak identities were
determined by mass spectrometry. (+)-Catechin ethylidene-bridged
oligomers were quantified from the amount of (+)-catechin monomer
that disappeared during the condensation.
Synthesis and Characterization of 2,2′-Ethylidenediphloroglu-
cinol (EDP). EDP was prepared as follows: Phloroglucinol (2 g) and
ascorbic acid (200 mg) were dissolved in 10 mL of 0.1 N HCl methanol,
and 15 µL of acetaldehyde was combined with 2 mL of the phloro-
glucinol mixture and then the mixture stirred for 30 min at 50 °C. The
reaction was stopped by adding 10 mL of 40 mM aqueous sodium
acetate. EDP was purified on a C18 cartridge as follows: the column
(5 g) was conditioned and the sample applied (10 mL). The column
was washed with 50 mL of water followed by 50 mL of methanol/
water (25:75, v/v). EDP was then eluted with 50 mL of methanol:
water (1:1 by volume). The fraction was dried under reduced pressure
and lyophilized to a dry powder.
EDP purity was determined by LC-MS. It was quantified at 280 nm
and corresponded to the percentage of the EDP peak area compared to
the total peak area. The column used was a reversed-phase C18 Waters
Xterra protected with a guard column of the same material (3.5 µm
packing, 4.6 × 100 mm i.d.) (Agilent, Saint Quentin-en-Yvelines,
France). Solvent A was water/acetic acid (95:5, v/v) and solvent B
was acetonitrile. Elution was conducted at room temperature, and the
sample loop was 20 µL. The elution gradient was as follows: from 5
to 15% B in 6 min, from 15 to 40% B in 26 min, from 40 to 100% B
in 1 min, 100% B for 3 min, from 100 to 5% B in 1 min, 5% B for
4 min with a 1.5 mL‚min-1 flow for the column and 0.15 mL‚min-1
Figure 1. Reaction pathway of phloroglucinolysis. Procyanidins, R′ )
and R′′ ) OH; prodelphinidins, R′ ) OH and R′′ ) OH.
H
grade), acetic acid (RP), L-ascorbic acid, L-tartaric acid, hydrochloric
acid, and sodium acetate were from Prolabo-VWR (Fontenay s/Bois,
France). (+)-Catechin, (-)-epicatechin, (-)-epigallocatechin, and (-)-
epicatechin-3-O-gallate were purchased from Sigma-Aldrich (Saint
Quentin Fallavier, France). Phloroglucinol and 4-methylcatechol were
purchased from Fluka (Saint Quentin Fallavier, France).
MS Apparatus and LC-MS Analysis. LC-MS analyses were
performed on a Micromass Platform II simple quadruple mass
spectrometer (Micromass-Beckman, Roissy Charles-de-Gaulle, France)
equipped with an electrospray ion source. The mass spectrometer was
operated in negative-ion mode. Source temperature was 120 °C,
capillary voltage was set at ( 3.5 kV, and cone voltage was -30 V.
HPLC separations were performed on a Hewlett-Packard 1100 series
(Agilent, Massy, France) including a pump module and a UV detector.
Both systems were operated using Masslynx 3.4 software. The
absorbance was recorded at 280 nm and mass spectra were recorded
from 50 to 1500 amu.
1
for the MS source. EDP characterization was also determined by H
NMR analysis. The compound was dissolved in deuterated methanol.
Phloroglucinolysis of (+)-Catechin Ethylidene-Bridged Oligo-
mers. A solution of 0.1 N HCl in methanol containing 50 g‚L-1
phloroglucinol and 10 g‚L-1 ascorbic acid was prepared. A 5 g‚L-1
solution of (+)-catechin ethylidene-bridged oligomers (8 h) was
prepared in methanol, and 200 µL of the previous solution reacted in
200 µL of this solution at 50 °C for 10, 20, 40, 60, and 90 min and
then combined with 1 mL of 40 mM aqueous sodium acetate to stop
the reaction.
(+)-Catechin ethylidene-bridged oligomers cleavage products were
monitored by LC-MS using the same elution conditions used for
monitoring the synthesis and characterization EDP.
Wine EDP Phloroglucinolysis Method. The phloroglucinolysis
protocol was divided into two steps. The first step consisted of
purification and concentration of the wine and was carried out using
C18 solid-phase extraction (SPE). Water (15 mL) was added to 5 mL
of wine. The mixture was purified on a C18 cartridge as follows: the
column was conditioned and the sample applied (10 mL). The column
was washed with 50 mL of water and eluted with 50 mL of methanol.
The fraction was dried under reduced pressure and then dissolved in
2 mL of methanol. The second step was the phloroglucinolysis reaction.
A solution of 0.2 N HCl in methanol, containing 100 g‚L-1 phloro-
glucinol and 20 g‚L-1 ascorbic acid, was prepared, and 100 µL of wine
sample was reacted with 100 µL of the phloroglucinol reagent at
50 °C for 20 min and then combined with 200 µL 400 mM aqueous
sodium acetate to stop the reaction. Aqueous 4-methylcatechol (4MC;
20 µL of 500 mg‚L-1) was then added as internal standard.
Specific ionization masses, m/z 277 (EDP) and m/z 123 (4MC), were
recorded. EDP quantification was accomplished using the EDP/4MC
peak areas after calibration and considering the sample dilution (1.68).
The same elution conditions were used than for the synthesis and
characterization of EDP but using the following elution gradient: 10%
B for 2 min, from 10 to 50% B in 8 min, from 50 to 100% B in 1 min,
100% B for 4 min, from 100 to 10% B in 1 min, 10% B for 4 min
with a 1 mL‚min-1 flow for the column and 0.1 mL‚min-1 for the MS
source.
1H NMR Analysis. A Bruker Avance-300 NMR spectrometer
(Bruker, Wissembourg, France), operating at 300.13 MHz for 1H, using
the UXNMR software package, was used for NMR experiments;
chemical shifts are expressed in δ (parts per million) referred to the
solvent peaks δH 3.31 for CD3OD; coupling constants, J, are in hertz.
Compounds were dissolved in deuterated methanol.
Synthesis and Characterization of (+)-Catechin Ethylidene-
Bridged Oligomers. (+)-Catechin ethylidene-bridged oligomers were
synthesized in model wine (12% v/v ethanol, 5 g‚L-1 tartaric acid and
pH 3.2) as follows: (+)-Catechin (100 mg) was dissolved in model
wine (100 mL), and the solution (100 mL) was mixed with acetaldehyde
(5 mL) and model wine (95 mL). The mixture was incubated at 40 °C
for 8 and 24 h. To stop the reaction, a solid-phase extraction (SPE)
step was used. Each sample was purified on a C18 cartridge (Supelco,
St Quentin Fallavier, France) as follows: the column (5 g) was
conditioned (50 mL MeOH followed by 50 mL H20) and the sample
(5 mL) applied. The column was washed with 15 mL of water to remove
tartaric acid and excess acetaldehyde. The polymers were then eluted
with 5 mL of methanol. The methanol fraction was dried under reduced
pressure and lyophilized to a dry powder.
Oligomers were analyzed by LC-MS. The column used was a
reversed-phase C18 Interchrom UP3 ODB-10QS (3 µm packing, 100
× 4.6 mm i.d.) protected with a guard column of the same material
(Interchim, Montluc¸on, France). Solvent A was water:acetic acid (99:1
by volume), and solvent B was acetonitrile:water (4:1 by volume).
Separation was conducted at room temperature, flow rate was 1
mL‚min-1 for the column and 0.1 mL‚min-1 for the MS source, and
the sample loop was 20 µL. The elution gradient was as follows: from
0 to 20% B in 1 min, from 20 to 30% B in 5.5 min, from 30 to 50%
Wine EDP Phloroglucinolysis Calibration. Calibration was done
using an EDP extract (60% w/w purity, accounted for in the calibration).
Solutions containing increasing EDP concentrations and fixed 4MC
concentration (internal standard) were prepared. EDP concentrations
were 100, 50, 20, 10, 5, and 2 mg‚L-1. The 4MC concentration in all
samples was fixed at 23.8 mg‚L-1. Samples were separated and

Analysis of Ethylidene-Bridged Flavan-3-ols in Wine
J. Agric. Food Chem., Vol. 55, No. 4, 2007 1111
monitored by LC-MS with molecular ion masses of m/z 277 (EDP)
and m/z 123 (4MC). The calibration curve was obtained by fitting EDP/
4MC peak area as a function of EDP concentration.
Wine EDP Phloroglucinolysis Repeatability. Method repeatability
was carried out using one wine. Wine EDP phloroglucinolysis was
determined as described above. SPE purification was conducted in
triplicate. Phloroglucinolysis was then conducted in triplicate for each
replicate. Each replicate was injected into the LC-MS. For one
replicate, the injection was replicated three times. Repeatability was
defined as the percent variation of the EDP concentration measured.
Wine EDP Phloroglucinolysis Recovery Study. (+)-Catechin
ethylidene-bridged oligomers (24 h reaction) at three concentrations
were prepared in methanol (100, 50, and 25 mg‚L-1). These solutions
corresponded to 1.5, 3.0, and 5.9 mg‚L-1 ethylidene units within the
oligomers. Considering that 1 mol of combined acetaldehyde yields 1
mol of EDP after phloroglucinolysis, the expected EDP concentrations
were calculated. The phloroglucinolysis reaction was carried out, and
the EDP concentrations were measured by LC-MS. Measured EDP
concentrations were then compared to expected EDP concentrations,
and the method recovery was determined as follows:
Figure 2. Proposed compounds released by the phloroglucinolysis of
ethylidene-bridged flavan-3-ols.
[EDP]measured
recovery (%) )
× 100
[EDP]calculated
Recovery was also determined in model wine (12% ethanol, 5 g‚L-1
tartaric acid, pH 3.2) and wine. Solutions of (+)-catechin ethylidene-
bridged oligomers were then isolated. The EDP phloroglucinolysis
method was applied (SPE + phloroglucinolysis), and EDP concentra-
tions were measured by LC-MS. Measured EDP concentrations were
compared to expected EDP concentrations, and the method recovery
was determined.
Wine Phloroglucinolysis. The proanthocyanidin mDP and the
flavan-3-ols analyses were measured by phloroglucinolysis, which was
adapted from a previous study (17). The wine was purified and
concentrated similar to that for wine EDP phloroglucinolysis. A solution
of 0.2 N HCl in methanol, containing 100 g‚L-1 phloroglucinol and
20 g‚L-1 ascorbic acid, was prepared, and 100 µL of wine sample was
reacted with 100 µL of the phloroglucinol reagent at 50 °C for 20 min
and then combined with 1000 µL of 40 mM aqueous sodium acetate
to stop the reaction.
The proanthocyanidin composition was determined by phloroglu-
cinolysis and LC-MS analysis at 280 nm after peak identification and
calibration. The same elution conditions were used as for the synthesis
and characterization of EDP but using water:acetic acid (99:1, v/v) as
solvent A, methanol as solvent B, and the following elution gradient:
5% B for 25 min, from 5 to 20% B in 20 min, from 20 to 40% B in
25 min, from 40 to 100% B in 1 min, 100% B for 9 min, from 100 to
5% B in 1 min, 10% B for 4 min with a 1 mL‚min-1 flow for the
column and 0.1 mL‚min-1 for the MS source.
Figure 3. HPLC chromatogram of synthesized (
bridged oligomers. ( )-cat )-catechin, dimE
bridged dimers, trimE ) (+)-catechin ethylidene-bridged trimers.
+
)-catechin ethylidene-
+
)
(+
) (+)-catechin ethylidene-
the polymerization reaction. Previous work has shown that
during the reaction only ethylidene-bridged oligomers are
formed (18). Upon analysis, 61% and 97% of the (+)-catechin
existed as ethylidene-bridged oligomers, for the 8 and 24 h
reaction, respectively. Considering the ethylidene-bridged (+)-
catechin unit molecular weight (461 g‚mol-1) and the molecular
weight of the ethylidene bridge (28 g‚mol-1), ethylidene-bridged
(+)-catechin oligomers contained 6.07% ethylidene by weight.
By extension, the reaction mixtures after 8 and 24 h contained,
respectively, 3.7 and 5.9 mg of ethylidene per 100 mg of
material.
RESULTS AND DISCUSSION
This study was conducted in order to develop a method to
quantify ethylidene-bridged flavan-3-ols in wine by phloroglu-
cinolysis. As the interflavanoid bonds of native flavan-3-ols are
broken in acid with heat, we hypothesized that ethylidene
bridges would undergo cleavage. Considering this hypothesis,
phloroglucinolysis of ethylidene-bridged flavan-3-ols would
release flavan-3-ol monomers from terminal units and their
corresponding ethylideneflavan-3-ol-phloroglucinol adducts
from extension units (Figure 2). Ethylidene-bridged flavan-3-
ol phloroglucinolysis was first investigated using synthesized
(+)-catechin ethylidene-bridged oligomers.
Synthesized (+)-Catechin Ethylidene-Bridged Oligomers.
The synthesized extracts of (+)-catechin ethylidene-bridged
oligomers consisted primarily of (+)-catechin monomer and
dimer and trimer of ethylidene-bridged oligomers (Figure 3).
The quantity of ethylidene-bridged oligomers was estimated by
determining the (+)-catechin monomer disappearance during
Phloroglucinolysis of Ethylidene-Bridged (+)-Catechin
Oligomers. Ethylidene-bridged (+)-catechin oligomers (8 h)
and predicted cleavage products were observed by their mo-
lecular ion masses in the negative-ion mode: (+)-catechin (m/z
289), (+)-catechin ethylidene-bridged dimer (m/z 605), (+)-
catechin ethylidene-bridged trimer (m/z 921), and ethylidene-
flavan-3-ol-phloroglucinol adduct (m/z 441). After a phloro-
glucinolysis reaction time of 10 min (Figure 4), no (+)-catechin
ethylidene-bridged trimers were detected and very little (+)-
catechin ethylidene-bridged dimers remained. Ethylideneflavan-
3-ol-phloroglucinol adducts (two stereoisomers) were found
in small quantities.
The two major compounds, 1 and 2, formed during phloro-
glucinolysis were unknown (Figure 5). MS analysis showed
molecular ion masses of m/z 277 and 415, respectively. The
proposed structure for 1 is the phloroglucinol ethylidene-bridged
dimer, 2,2′-ethylidenediphloroglucinol (EDP), and 2 was pro-

1112 J. Agric. Food Chem., Vol. 55, No. 4, 2007
Drinkine et al.
Figure 4. HPLC chromatogram of cleavage products from (
of phloroglucinol. AA ascorbic acid, P phloroglucinol, (
)-catechin phloroglucinol adduct, 1 and 2 unknown compounds.
+)-catechin ethylidene-bridged oligomers (8 h) following 10 min acid-catalysis in the presence
)
)
+)-cat ) (+)-catechin, dimE ) (+)-catechin ethylidene-bridged dimer, ECP ) ethylidene-
(
+
−
)
Figure 5. Mass spectra in negative-ion mode and proposed structures for unknown compounds following phloroglucinolysis of (+)-catechin ethylidene-
bridged oligomers.
posed to be the 1-(3,4-dihydroxyphenyl)-1,3-bis(2,4,6-trihydroxy-
phenyl)propan-2-ol. This product has been previously described
(16) and results from the reaction of phloroglucinol with the
(+)-catechin B-ring quinone-methide following acid-catalyzed
C-ring cleavage.
(+)-catechin ethylidene-bridged oligomers was monitored as a
function of time (data not shown). The optimum time found was
20 min. The reaction was not complete and low amounts of ethyl-
idene-(+)-catechin-phloroglucinol adducts were found. The EDP
peak area represented 77% of the sum of EDP and ethylidene-
(+)-catechin-phloroglucinol adduct peak areas at 280 nm.
2,2′-Ethylidenediphloroglucinol Compound (EDP). As
EDP was the major compound formed by phloroglucinolysis,
it was synthesized to calibrate the method. The EDP extract,
analyzed at 280 nm and by LC-MS, was 60% EDP purity
(Figure 7). On the chromatogram, an additional peak other than
EDP appeared with m/z 429 molecular ion mass corresponding
to the phloroglucinol ethylidene-bridged dimer, consistent with
stepwise polymerization kinetics.
(+)-catechin and EDP were the major products of (+)-
catechin ethylidene-bridged oligomer phloroglucinolysis. This
led to the following hypothesis: ethylidene-bridged flavan-3-
ol polymers should release EDP upon phloroglucinolysis. The
pathway proposed for the ethylidene-bridged flavan-3-ol phlo-
roglucinolysis is an acid-catalyzed nucleophilic substitution
(Figure 6). The first step would consist of the acidic cleavage
of one part of the ethylidene bridge, releasing one flavan-3-ol
unit, followed by nucleophilic attack by phloroglucinol, leading
to an ethylideneflavan-3-ol-phloroglucinol adduct. The second
step would consist of the acidic cleavage of the second part of
the ethylidene bridge, releasing the second flavan-3-ol unit,
followed by phloroglucinol addition, leading to EDP formation.
EDP would be the only product of ethylidene-bridged flavan-
3-ol phloroglucinolysis.
The structure of EDP was confirmed by 1H NMR. The
spectrum (Figure 8) is readily interpretable: The 4.72 ppm
quartet is consistent with the H7 proton, and the 1.69 ppm
doublet is consistent with the ethylidene methyl group (CH3).
H4, H6, H4′, and H6′ are equivalent and resonated as a singlet at
5.89 ppm (Figure 8).
In order to optimize EDP formation and reduce ethylidene-
flavan-3-ol-phloroglucinol adducts, the phloroglucinolysis of
Quantification of Ethylidene-Bridged Flavan-3-ols in
Wine. The phloroglucinolysis method was adapted from previ-

Analysis of Ethylidene-Bridged Flavan-3-ols in Wine
J. Agric. Food Chem., Vol. 55, No. 4, 2007 1113
Figure 6. Hypothetical phloroglucinolysis reaction pathway for the formation of EDP from ethylidene-bridged flavan-3-ols. Procyanidins, R′ ) H and R′′
OH; prodelphinidins, R′ ) OH and R′′ ) OH.
)
Figure 7. HPLC chromatogram of purified EDP extract. P
) phloroglucinol, EDP ) phloroglucinol ethylidene-bridged dimer, ETP ) phloroglucinol
ethylidene-bridged trimer.
ous work (17) and was conducted as follows: wine dilution in
water, isolation, and concentration using SPE, followed by
phloroglucinolysis. Wine samples were diluted 1.68 times before
injection as a result of the SPE and phloroglucinolysis steps.
EDP quantification was carried out by mass spectrometry in
the negative-ion mode with an internal standard, 4-methylcat-
echol (4MC), added at the end of the phloroglucinolysis. EDP
quantification was calculated from the peak area ratio EDP/
4MC after calibration. The calibration followed a linear regres-
sion (R2 ) 0.997):
where the 1.68 coefficient corresponds with the wine dilution.
Method repeatability was determined to be 9.2%, which is in
agreement with criteria for intralaboratory analysis (19) (Table
1). The EDP concentration was then converted into ethylidene
bridge concentration considering the molecular weight of EDP
and ethylidene.
Known amounts of ethylidene-bridged oligomer (24 h) were
dissolved in methanol and model wine to determine the method
recovery. The recovery of ethylidene was determined in
methanol samples, and the recovery of the entire method was
determined in model wine samples. Expected EDP concentra-
tions were then calculated and compared to measured EDP
concentrations. The results showed that the measured concentra-
EDP/4MC
EDP (mg‚L-1) )
× 1.68
0.037

1114 J. Agric. Food Chem., Vol. 55, No. 4, 2007
Drinkine et al.
Figure 8. 1H NMR spectrum and spectroscopic data (300 MHz) of EDP in deuterated methanol.
Table 1. EDP Mean Concentrations of Repeatability Assay and
Corresponding Percentage Errors (
±
SD, N
) 3)
[EDP]
coefficient of
variation (%)
-
1
(mg
‚
L
)
injection
phloroglucinolysis and SPE
6.1
6.2
5.1
5.7
5.8
±
±
±
±
±
0.3
0.2
0.3
0.3
0.5
5.4
4.8
3.2
4.9
9.2
mean
tions are close to expected concentrations (Figure 9A). The
ethylidene recovery in methanol was 97% and the recovery of
the entire method was 90%. The same known amounts were
added to a wine, and the recovery was calculated again. The
results showed that the measured concentrations are close to
expected concentrations until an EDP concentration of 18
mg‚L-1(e.g., 4.5 mg‚L-1 ethylidene) is reached. The recovery
was then 90%. Above an EDP concentration of 30 mg‚L-1 (e.g.,
7.6 mg‚L-1 ethylidene), the recovery decreased to 83%.
Application to Wine. The EDP phloroglucinolysis and
flavan-3-ol determination were applied to three Bordeaux red
wines, produced by the same winery but from different
vintages: 2004, 1999, and 1991. Results showed that ethylidene
bridges were measurable and comparable to flavan-3-ol amounts
(Table 2). Nevertheless, their amounts were very low (0.7-
1.3 mg‚L-1) compared to flavan-3-ol amounts. Ethylidene-
bridged linkages represented less than 1.3% of the interflavanoid
linkages by mol (Figure 10). The analyses of more wines are
necessary to draw meaningful conclusions.
Figure 9. Recovery and linearity studies of the EDP phloroglucinolysis:
(A) in methanol and in model wine and (B) in wine (±SD, N ) 3).
Specific questions remain about the quantity of ethylidene-
bridged flavan-3-ols in wines. The surprisingly low amounts
of EDP suggest that ethylidene-bridged flavan-3-ols are unstable
and undergo further reaction. Given the relative ease with which
EDP is formed under the phloroglucinolysis conditions, it is
likely that ethylidene-bridged flavan-3-ols are easily hydrolyzed
under wine conditions. Given that acetaldehyde reacts with other
nucleophiles in wine such as anthocyanins (20-28) and the

Analysis of Ethylidene-Bridged Flavan-3-ols in Wine
J. Agric. Food Chem., Vol. 55, No. 4, 2007 1115
(5) Boulton, R. The copigmentation of anthocyanins and its role in
the color of red wine: A critical review. Am. J. Enol. Vitic. 2001,
52, 67-82.
(6) Liao, H.; Cai, Y.; Haslam, E. Polyphenol interactions. Antho-
cyanins: Co-pigmentation and colour changes in red wines. J.
Sci. Food Agric. 1992, 59, 299-305.
(7) Prieur, C.; Rigaud, J.; Cheynier, V.; Moutounet, M. Oligomeric
and polymeric procyanidins from grape seeds. Phytochemistry
1994, 36, 781-784.
(8) Haslam, E. In vino veritas: Oligomeric procyanidins and the
ageing of red wines. Phytochemistry 1980, 19, 2577-2582.
(9) Jurd, L. Review of polyphenols condensation reactions and their
possible occurrence in the aging of wines. Tetrahedron 1969,
25, 2367-2380.
Table 2. Ethylidene Bridge Amounts and Flavan-3-ol Composition in
-1
mg‚
L
for Three Bordeaux Wines (
±
SD, N
)
3)a
vintage
CHCH3 Flterm
Flext
Fl
mDP
2004
1999
1991
0.8
0.7
1.3
±
±
±
0.1
0.1
0.1
355
260
192
±
±
±
10
25
6
517
282
181
±
±
±
17
5
3
871
541
373
±
±
±
22
30
5
2.4
2.1
1.9
±
±
±
0.1
0.1
0.1
a CHCH3
)
ethylidene bridges, Fl
)
flavan-3-ols, Flterm
) flavan-3-ol terminal
units, Flext
)
flavan-3-ol extension units, mDP )
mean degree of polymerization.
(10) Timberlake, C. F.; Bridle, P. Anthocyanins: Color augmentation
with catechin and acetaldehyde. J. Sci. Food Agric. 1977, 26,
539-544.
(11) Es-Safi, N.; Fulcrand, H.; Cheynier, V.; Moutounet, M. Competi-
tion between (+)-catechin and (-)-epicatechin in acetaldehyde-
induced polymerization of flavanols. J. Agric. Food Chem. 1999,
47, 2088-2095.
(12) Saucier, C.; Guerra, C.; Laguerre, M.; Glories, Y. (+)-Catechin-
acetaldehyde condensation products in relation with wine-ageing.
Phytochemistry 1997, 46, 229-234.
(13) Romano, P.; Suzzi, G.; Turbanti, L.; Polsinelli, M. Acetaldehyde
production in Saccharomyces cerevisiae wine yeasts. Microbiol.
Lett. 1994, 118, 213-218.
(14) Singleton, V. L.; Kramling, T. E. Browning of white wines and
an accelerated test for browning capacity. Am. J. Enol. Vitic.
1976, 27, 157-160.
(15) Rigaud, J.; Perez-Ilzarbe, J.; Ricardo Da Silva, J. M. Micro
method for the identification of proanthocyanidin using thiolysis
monitored by high-performance liquid chromatography. J. Chro-
matogr. 1991, 540, 401-405.
(16) Kennedy, J. A.; Jones, G. P. Analysis of proanthocyanidin
cleavage products following acid-catalysis in the presence of
excess phloroglucinol. J. Agric. Food Chem. 2001, 49, 1740-
1746.
(17) Peyrot des Gachons, C.; Kennedy, J. A. Direct method for
determining seed and skin proanthocyanidin extraction into red
wine. J. Agric. Food Chem. 2003, 51, 5877-5881.
(18) Drinkine, J.; Saucier, C.; Glories, Y. (+)-Catechin-aldehyde
condensations: Competition between acetaldehyde and glyoxylic
acid. J. Agric. Food Chem. 2005, 53, 7552-7558.
(19) Official methods of analysis of the association of official
analytical chemists, 14th ed.; Horwitz, W., Ed.; Association of
Official Analytical Chemists: Arlington, VA, 1984; pp 1141.
(20) Timberlake, C. F.; Bridle, P. The effect of processing and other
factors on the colour characteristics of some red wines. Vitis
1976, 15, 37-49.
(21) Bakker, J.; Picinelli, A.; Bridle, P. Model wine solutions: Colour
and composition changes during ageing. Vitis 1993, 32, 111-
118.
(22) Dallas, C.; Ricardo-Da-Silva, M. J.; Laureano, O. Products
formed in model wine solutions involving anthocyanins, pro-
cyanidin B2, and acetaldehyde. J. Agric. Food Chem. 1996, 44,
2402-2407.
Figure 10. Molar ratio in percentage of ethylidene-bridged linkages
compared to interflavonoid linkages in three Bordeaux wines (
±SD, N )
3).
reactivity of these products to phloroglucinolysis has not been
determined, it is important to elucidate this chemistry in order
to fully understand the role of acetaldehyde in wine aging.
Determining this is the subject of ongoing studies. Nevertheless,
the results of our work shows that it should be possible to
measure ethylidene-bridged flavan-3-ols directly in wines by
quantifying the unique major cleavage product of phlorogluci-
nolysis, EDP.
ABBREVIATIONS USED
EDP, 2,2′-ethylidenediphloroglucinol; 4MC, 4-methylcat-
echol.
ACKNOWLEDGMENT
We thank Dr. Tristan Richard for the NMR analyses and their
interpretations.
LITERATURE CITED
(1) Santos-Buelga, C.; Scalbert, A. Proanthocyanidins and tannin-
like compoundssNature, occurrence, dietary intake and effects
on nutrition and health. J. Sci. Food Agric. 2000, 80, 1094-
1117.
(23) Bakker, J.; Timberlake, C. F. Isolation, identification, and
characterization of new color-stable anthocyanins occurring in
some red wines. J. Agric. Food Chem. 1997, 45, 35-43.
(24) Francia-Aricha, E. M.; Guerra, M. T.; Rivas-Gonzalo, J. C.;
Santos-Buelga, C. New anthocyanin pigments formed after
condensation with flavanols. J. Agric. Food Chem. 1997, 45,
2262-2266.
(25) Es-Safi, N.; Fulcrand, H.; Cheynier, V.; Moutounet, M. Studies
on the acetaldehyde-induced condensation of (-)-epicatechin and
malvidin 3-O-glucoside in a model solution system. J. Agric.
Food Chem. 1999, 47, 2096-2102.
(2) Llaudy, M.; Canals, R.; Canals, J. M.; Roze´z, N.; Arola, L.;
Zamora, F. New method for evaluating astringency in red wine.
J. Agric. Food Chem. 2004, 52, 742-746.
(3) Saucier, C.; Bourgeois, G.; Vitry, C.; Roux, D.; Glories, Y.
Characterization of (+)-catechin-acetaldehyde polymers:
A
model for colloidal state of wine polyphenols. J. Agric. Food
Chem. 1997, 45, 1045-1049.
(4) Vidal, S.; Cartalade, D.; Souquet, J.-M.; Fulcrand, H.; Cheynier,
V. Changes in procyanidin chain length in winelike model
solutions. J. Agric. Food Chem. 2002, 50, 2261-2266.

1116 J. Agric. Food Chem., Vol. 55, No. 4, 2007
Drinkine et al.
(26) Vivar-Quintana, A. M.; Santos-Buelga, C.; Rivas-Gonzalo, J.
C. Anthocyanin-derived pigments and colour of red wines. Anal.
Chim. Acta 2002, 458, 147-155.
vinylpyranoanthocyanin-catechin pigment (a portisin). Tetrahe-
dron Lett. 2004, 45, 3455-3457.
(27) Mateus, N.; Carvalho, E.; Carvalho, R. F. A.; Melo, A.; Gonzales-
Paramas, M. A.; Santos-Buelga, C.; De Freitas, A. P. V. Isolation
and structural characterization of new acylated anthocyanin-vinyl-
flavanol occuring in aging red wines. J. Agric. Food Chem. 2003,
51, 277-282.
Received for review September 13, 2006. Revised manuscript received
October 30, 2006. Accepted November 6, 2006. We thank Yvon Mau
Co. (Gironde sur Dropt, France) and ANRT (Association Nationale pour
la Recherche Technologique, Paris, Cifre Grant 855/2002) for their
financial support of this research.
(28) Mateus, N.; Oliveira, J.; Santos-Buelga, C.; Silva, A. M. S.; De
Freitas, A. P. V. NMR structure characterization of a new
JF0626258