Orosensory-directed identification of astringent mouthfeel and bitter-tasting compounds in red wine

Article abstract of DOI:10.1021/jf073031n

Application of sequential solvent extraction, followed by HPLC combined with the taste dilution analysis, enabled the localization of the most intense velvety astringent, drying, and puckering astringent, as well as bitter-tasting, compounds in red wine,

Full text of DOI:10.1021/jf073031n

1376 J. Agric. Food Chem. 2008, 56, 1376–1386
Orosensory-Directed Identification of Astringent
Mouthfeel and Bitter-Tasting Compounds in Red Wine
†
JAN CARLOS HUFNAGEL AND THOMAS HOFMANN*,#
Institut für Lebensmittelchemie, Universität Münster, Corrensstrasse 45, Universität Münster
Technische Universität München, D-48149 Münster, Germany, and Lehrstuhl für Lebensmittelchemie
and Molekulare Sensorik, Technische Universität München, Lise-Meitner-Strasse 34,
D-85350 Freising-Weihenstephan, Germany
Application of sequential solvent extraction, followed by HPLC combined with the taste dilution analysis,
enabled the localization of the most intense velvety astringent, drying, and puckering astringent, as
well as bitter-tasting, compounds in red wine, respectively. Isolation of the taste components involving
gel adsorption chromatography, ultrafiltration, and synthesis revealed the identification of 26 sensory-
active nonvolatiles, among which several hydroxybenzoic acids, hydroxycinnamic acids, flavon-3-ol
glycosides, and dihydroflavon-3-ol rhamnosides as well as a structurally undefined polymeric fraction
(>5 kDa) were identified as the key astringent components. In contradiction to literature suggestions,
flavan-3-ols were found to be not of major importance for astringency and bitter taste, respectively.
Surprisingly, a series of hydroxybenzoic acid ethyl esters and hydroxycinnamic acid ethyl esters were
identified as bitter compounds in wine. Taste qualities and taste threshold concentrations of the
individual wine components were determined by means of a three-alternative forced-choice test and
the half-mouth test, respectively.
KEYWORDS: Red wine; wine; astringency; bitterness; taste; taste dilution analysis; polyphenolic acid
ethyl esters; half-mouth test
INTRODUCTION
Although many attempts have been made to correlate
analytical data on distinct wine components with the sensory
data obtained from human subjects, the reports on the chemical
species imparting the typical bitter and astringent taste of red
wines are rather contradictory. It is believed that astringency is
due to the polyphenol-induced complexation and/or precipitation
of proline-rich salivary proteins in the oral cavity (2–4), thus
inducing a tactile sensation perceived by touch via mechanore-
ceptors (5). Consistent with this hypothesis, more than 40 years
ago water-soluble phenols with molecular masses from 500 to
3000 Da were found to be required for imparting astringency
(6), and bioassays based on protein complexation have been
developed for the measurement of polyphenols such as tannins
(7). However, first attempts trying to predict astringency by
analyzing the turbidity of solutions containing polyphenol-protein
complexes gave rather contradictory data (8). Furthermore,
simple polyphenols containing 1,2-dihydroxybenzene or 1,2,3-
trihydoxybenzene groups were reported to cross-link and
precipitate proteins (9). On the other hand, low molecular weight
compounds were found to elicit astringency by complexing
salivary proteins without precipitation (10).
As one of life’s finest pleasures, the alluring aroma, the
desirable taste, and the typical color of red wines have been
attracting consumers for more than 2000 years. Aroma-active
volatiles as well as nonvolatile chromophores of red wine were
thoroughly investigated in recent decades, but only a small
number of studies were targeted toward the nonvolatile taste
compounds in wine and, in particular, those eliciting an
astringent and/or bitter oral sensation. The typical astringent
mouthfeel, which is perceived as a long-lasting trigeminal
sensation in the oral cavity, can be classified into several
subqualities such as velvety, grainy, drying, or puckering (1)
and, together with bitterness, is of crucial importance for the
palatability of red wines. Whereas velvety astringency is
perceived as a silky and finely textured kind of astringent
sensation, puckering astringency is understood as a reflexive
action of cheek surfaces being brought together and released in
an attempt to lubricate mouth surfaces (1). Despite being aware
of the sensory importance of nonvolatiles in wines, the key
inducers of astringency and bitter taste in red wines are still
unclear on a molecular level.
Flavan-3-ol monomers and higher oligomers, so-called pro-
cyanidins, are well accepted to induce an astringent oral
sensation as well as bitter taste (11–13). Although the threshold
concentrations of these compounds were found to be rather high,
for example, 46.1 and 17.3 mg/L for (+)-catechin and procya-
nidin B3 (14), respectively, the procyanidins were intensively
* Author to whom correspondence should be addressed (telephone
+49-8161/71-2902; fax +49-8161/71-2949; e-mail thomas.hofmann@
wzw.tum.de).
† Universität Münster.
# Technische Universität München.
10.1021/jf073031n CCC: $40.75
2008 American Chemical Society
Published on Web 01/15/2008

Taste Compounds in Red Wine
J. Agric. Food Chem., Vol. 56, No. 4, 2008 1377
materials of procyanidin B3 and procyanidin C1 were isolated from
nonfermented cocoa beans following the procedure reported re-
cently (34).
investigated for the time/intensity course of the astringency
perceived as well as for interactions in mixtures (15–17). More
recent studies reported that highly polymeric, galloylated tannins
induce a puckering astringent mouthfeel (18). In contradiction,
studies performed on crude fractions isolated from red
wines (19–21) revealed that gallic acid and, in particular, flavan-
3-ols exhibiting molecular masses below 500 Da elicit astringent
and bitter taste qualities.
Besides the polyphenols, organic acids were reported to
impart astringency (22). On the contrary, other studies found
that organic acids are able to influence but, with the exception
of malic acid, do not evoke astringent taste sensations (23, 24).
Although more recent studies showed some correlations between
the concentrations of polyphenolic compounds and sensory
descriptive analysis, the key inducers of astringency could not
be unequivocally identified on a molecular level (25).
Similar to astringency, also the data available on bitter taste
compounds in wines are rather inconsistent. Multiple investigations
suggest procyanidins as bitter stimuli in wines (11–13, 16, 19, 20)
and, moreover, ethanol was reported to enhance the bitter
intensity perceived (11, 26). In addition, terpene glycosides were
reported to contribute to the bitter taste of Muscat wines (27),
but studies aimed at correlating chemical and sensory data of
red wines were not successful in generating a predictive model
for bitterness on wines (28).
To molecularize the sensometabolites driving the attractive
taste of foods, we have recently developed sensomics tools such
as the so-called taste dilution analysis (TDA) to screen for
sensory active nonvolatiles (29). Application of this approach
led to the identification of bitter compounds in thermally
processed sugar/amino acid mixtures (29), cooling compounds
in roasted malt (30), bitter off-tastants in carrots (31), a taste
enhancer in beef bouillon (32), the astringent and bitter key
compounds in tea infusions (33) as well as in roasted cocoa
(34), and, recently, the astringent ellagitannins migrating into
spirits and wines upon oak treatment (35).
The red wine used for the study was an Amarone della Valpolicella
DOC, vintage year 1997, 15% by volume ethanol; the grapes were
grown on a clay and limestone soil with a vine density of 2500 vines
per hectare; grape varieties used for manufacturing were 30% Corvina,
30% Corvinone, 30% Rondinella, and 10% a mixture of Molinara,
Rossignola, Oseleta, Negrara, and Dindarella. For its manufacturing,
grapes were harvested by hand at the end of September to the beginning
of October, placed in wooden trays, and then dried for about 4 months
at controlled temperature and humidity conditions. After the grapes
had lost about 40% of their weight, the grapes were pressed and the
must obtained was fermented at 15 °C in stainless steel tanks for about
40 days. The young wine was then matured for about 2 years in
barriques and another year in the bottle.
Sensory Analyses. Training of the Sensory Panel. Ten subjects (five
women and five men, ages 22–39 years), who gave informed consent
to participate in the sensory tests of the present investigation and had
no history of known taste disorders, were trained in sensory experiments
with purified reference compounds at regular intervals for at least 2
years as described earlier (34, 35) and were, therefore, familiar with
the techniques applied. Sensory analyses were performed in a sensory
panel room at 22 °C in three different sessions under red light.
Pretreatment of Fractions. Prior to sensory analysis, the fractions
or compounds isolated were suspended in water, and, after removal of
the volatiles under high vacuum (<5 mPa), were freeze-dried twice.
GC-MS and ion chromatographic analysis revealed that food fractions
treated by that procedure are essentially free of the solvents and buffer
compounds used.
Taste Profile Analysis. A freshly opened bottle of wine was kept at
room temperature for at least 2 h prior to sensory analysis. Freeze-
dried fractions isolated from red wine were taken up in water in
“natural” concentrations prior to analysis. The samples (5 mL) were
presented to the sensory panelists, who wore nose clips and who were
asked to briefly swirl the sample in the mouth and, then, to expectorate.
Using this sip-and-spit method, the panelists were asked to score the
taste qualities astringent, bitter, sour, sweet, salty, umami, and mouth-
fulness/body on a scale from 0 (not detectable) to 5.0 (strong taste
impression).
To bridge the gap between the sensory perception of red wine
and the chemical structure of the corresponding taste stimuli,
the objectives of the present investigation are to screen for the
key astringent and bitter compounds in a red wine by means of
the taste dilution analysis, to isolate and identify the chemical
structure, and to determine the sensory thresholds of the
compounds evaluated with the highest gustatory response.
Because Amarone della Valpollicella is considered to be one
of the most prestigious wines of excellent taste quality, this wine
was selected as the target for analysis.
Taste Dilution Analysis (TDA). A taste dilution (TD) factor (29) was
determined for each individual HPLC fraction isolated from red wine
using bottled water (pH 4.5) as the solvent. To achieve this, aliquots
of the individual fractions were dissolved in their “natural” concentration
ratio in exactly 10 mL of water and were then sequentially 1 + 1 diluted
with bottled water. The serial dilutions of each of these fractions were
then presented to the sensory panel in order of ascending concentrations,
and taste impressions in each dilution was evaluated by means of the
recently developed half-tongue test (33–35).
Sequential Solvent Extraction of Red Wine. An aliquot (100 mL)
of red wine was partially freed from EtOH under vacuum (10 kPa)
and then extracted with n-pentane (5 × 300 mL) at room temperature.
The combined organic layer was freed from solvent under vacuum,
thus achieving the pentane-soluble components (fraction A). Thereafter,
the aqueous residue was extracted five times with EtOAc (300 mL each)
at room temperature, the organic phases were combined, the solvent
was removed under vacuum, and after the addition of water (30 mL),
the EtOAc extractables (fraction B) were obtained after freeze-drying.
The remaining aqueous layer was lyophilized to give the water solubles
(fraction C). The individual fractions, stored at – 26 °C until use, were
used for sensory and chemical analysis.
HPLC/TDA of Fraction B. A sample (200 mg) of fraction B was
dissolved in a mixture (20:80, v/v; 2 mL) of MeOH and aqueous
HCOOH (0.1% in water; pH 2.5) using an ultrasonic bath. After
membrane filtration (0.45 µm, Satorius Hannover, Germany), portions
of 200 µL were analyzed by semipreparative RP-HPLC/UV–vis.
Monitoring the effluent at 272 nm, chromatography was performed
starting with a mixture (95:5, v/v) of aqueous HCOOH (0.1%) and
CH3CN, then increasing the CH3CN content to 17% within 35 min, to
20% within 15 min, and then to 50% within an additional 25 min, and
finally to 100% within 5 min. The eluent was separated into 30 fractions,
MATERIALS AND METHODS
Chemicals. The following compounds were obtained commercially:
(+)-catechin, (-)-epicatechin, gallic acid, caffeic acid, syringic acid,
ferulic acid ethyl ester, protocatechuic acid ethyl ester, formic acid,
acetic acid, lactic acid, (E)-aconitic acid, (Z)-aconitic acid, glutaric
acid, tartaric acid, succinic acid, malic acid, citric acid, isocitric acid,
galacturonic acid, and phosphoric acid (Sigma, Steinheim, Germany);
vanillic acid ethyl ester, caftaric acid (Apin Chemicals, Oxon, U.K.);
p-coumaric acid, vanillic acid, silica gel 60, gallic acid ethyl ester (Fluka
Chemika, Taufkirchen, Germany); isorhamnetin-3-O-ꢀ-D-glucopyra-
noside, quercetin-3-O-ꢀ-D-galactopyranoside, syringetin-3-O-ꢀ-D-glu-
copyranoside, procyanidin B1, procyanidin B2 (Extrasynthese, Genay
Cedex, France); and ethyl acetate (EtOAc), ethanol (EtOH), methanol
(MeOH), acetonitrile (CH3CN), and acetone (Me2CO) of HPLC grade
(Merck, Darmstadt, Germany). Deuterated solvents were from Euriso-
top (Saarbruecken, Germany). Deionized water used for chromatography
was purified by means of a Milli-Q Gradient A10 system (Millipore,
Billerica, MA). For sensory analyses, bottled water (Evian) was adjusted
to pH 4.5 with trace amounts of formic acid prior to use. Reference

1378 J. Agric. Food Chem., Vol. 56, No. 4, 2008
Hufnagel and Hofmann
namely, fractions B-1–B-30, which were individually collected into
ice-cooled, brown glass vials. The corresponding effluents obtained from
10 HPLC runs were combined and, after the solvent had been removed
under vacuum, freeze-dried and were directly used for the TDA.
Gel Adsorption Chromatography (GAC). A sample (600 mg) of
content to 100% within 5 min, monitoring the eluent at 272 nm with
a UV–vis detector. After the most active astringent compounds had
been located by means of HPLC degustation, individual peaks were
collected in several runs and the corresponding eluates were combined
and freeze-dried. Spectroscopic data (UV-vis, LC-MS, NMR) of the
taste compounds were identical with those measured for the corre-
sponding reference compounds. Finally, the identity of the taste
compounds as (+)-catechin (HPLC fraction B-10), (-)-epicatechin
(HPLC fraction B-13), procyanidin B1 (HPLC fraction B-10), procya-
nidin B2 (HPLC fraction B-12), procyanidin B3 (HPLC fraction B-1),
and procyanidin C1 (HPLC fraction B-14) was confirmed by cochro-
matography with the reference compounds.
fraction
B
was dissolved in
a
solution of MeOH/water
(20:80, v/v; 10 mL) and placed onto the top of a water-cooled 400 ×
50 mm XK 26/70 glass column (Amersham Pharmacia Biotech,
Uppsala, Sweden) filled with a slurry of Sephadex LH 20 (GE
Healthcare, Munich, Germany), which was conditioned with a solution
of MeOH/water (20:80, v/v) and then adjusted to pH 4 with a 1%
aqueous solution of HCOOH. For chromatography, the column was
eluted sequentially with solutions of MeOH/water containing 20, 40,
60, or 80% MeOH (400 mL each), with MeOH (1200 mL), and finally
with an Me2CO/water (70:30, v/v; pH 4; 800 mL), keeping a flow rate
of 1.3 mL/min by means of a P1-type pump (Pharmacia Biotech). With
the effluent monitored at 272 nm by means of a UV-2575-type UV-vis
detector (Jasco), the GAC fractions were collected every 10 min by
means of an LKB Bromma 7000 Ultrorac fraction collector and
combined to give 13 GAC fractions (I-XIII). The fractions were then
freed from solvent under vacuum, freeze-dried, and used for isolation
and identification of the taste-active compounds located by means of
HPLC/TDA.
Synthesis of Phenolic Acid Ethyl Esters. A solution of caffeic acid,
p-coumaric acid, or syringic acid (15 mmol) in EtOH (100 mL) and
H2SO4 (95%; 0.75 mL) was stirred for 12 h at 60 °C. After cooling,
water (300 mL) was added, the mixture was freed of ethanol under
vacuum, and the remaining aqueous layer was extracted four times with
EtOAc (200 mL each). The combined organic layer was concentrated
under vacuum to about 15 mL, and aliquots (5 mL) were added on top
of a water-cooled glass column (500 × 30 mm) filled with a slurry of
silica gel 60 (6% water) in toluene/EtOAc (60:40, v/v). Chromatography
was performed with toluene/EtOAc (60:40, v/v) with stepwise increas-
ing amounts of EtOAc. Fractions were collected every 150 mL, organic
solvents were removed under vacuum, and the isolates obtained were
analyzed for the target compounds by means of RP-HPLC/DAD, LC-
MS, and NMR spectroscopy.
(+)-Catechin: UV-vis (CH3CN/water, pH 2.5), λmax ) 237, 277;
LC-MS (ESI+), m/z 291 (100; [M + H]+), 598 (35; [M2 + H + H2O]+);
LC-MS (ESI-), m/z 289 (100; [M]-).
(-)-Epicatechin: UV-vis (CH
3CN/water, pH 2.5), λmax ) 237, 277;
LC-MS (ESI+), m/z 291 (100; [M + H]+), 598 (35; [M2 + H + H2O]+);
LC-MS (ESI-), m/z 289 (100; [M]-).
Procyanidin B1: UV-vis (CH3CN/water; pH 2.5), λmax ) 236, 278;
LC-MS (ESI+), m/z 579 (100, [M + 1]+), 601 (39, [2M + Na]+);
LC-MS (ESI-), m/z 577 (100; [M]-).
Procyanidin B2: UV-vis (MeOH/water; pH 2.5), λmax ) 236, 278;
LC-MS (ESI+), m/z 579 (100, [M + 1]+), 601 (39, [2M + Na]+);
LC-MS (ESI-), m/z 577 (100; [M]-).
Procyanidin B3. UV-vis (CH3CN/water; pH 2.5), λmax ) 236, 277;
LC-MS (ESI+), m/z 579 (100, [M + 1]+), 601 (40, [2M + Na]+);
LC-MS (ESI-), m/z 577 (100; [M]-).
Procyanidin C1: UV-vis (CH3CN/water; pH 2.5), λmax ) 236, 278;
LC-MS (ESI+), m/z 867 (100, [M + 1]+), 889 (45, [M + Na]+); LC-
MS (ESI-), m/z 865 (100; [M]-).
Identification of Flavon-3-ol- and Dihydroflavon-3-ol Glycosides.
Analytical HPLC analysis of the GAC fractions revealed that the taste
compounds detected in HPLC fractions B-19, B-21, B-23, and B-25
by means of HPLC/TDA were present in GAC fractions VII and VIII.
Preparative HPLC, followed by UV-vis, LC-MS/MS, and 1D/2D NMR
spectroscopy led to the identification of the key astringent compounds
as quercetin-3-O-ꢀ-D-galactopyranoside (B-19), 2R,3R-dihydroquerce-
tin-3-O-R-L-rhamnoside (B-21), 2R,3R-dihydrokaempferol-3-O-R-L-
rhamnoside (B-23), quercetin-3-O-ꢀ-D-glucuropyranoside (B-21), isor-
hamnetin-3-O-ꢀ-D-glucopyranoside (B-23), and syringetin-3-O-ꢀ-D-
glucopyranoside (B-25) in the HPLC fractions given in parentheses.
The latter three substances were confirmed by cochromatography with
the corresponding reference compound.
Quercetin-3-O-ꢀ-D-galactopyranoside: UV-vis (CH3CN/water; pH
2.5), λmax ) 215, 251, 355; LC-MS (ESI+), m/z 465 (100; [M + 1]+),
303 (53; [M - gal + 1]+); 1H and 13C NMR data were identical with
those measured for the reference compound.
Isorhamnetin-3-O-ꢀ-D-glucopyranoside: UV-vis (CH3CN/water; pH
2.5), λmax ) 254, 350; LC-MS (ESI+), m/z 479 (100; [M + 1]+), 317
(45; [M - glc + 1]+); 1H and 13C NMR data were identical with those
measured for the reference compound.
Caffeic acid ethyl ester: UV-vis (CH3CN/water; pH 2.5), λmax
)
241, 297, 323; LC-MS (ESI+), m/z 209 (100; [M + 1]+), 163 (53; [M
1
- C2H5 - H2O + 1]+), 181 (20; [M - C2H5 + 1]+); H NMR (400
MHz; CD3OD), δ 1.36 [t, 3H, H-C(2′)], 4.23 [dd, 2H, H-C(1′)], 6.26
[d, 1H, H-C(2)], 6.79 [d, 1H, H-C(8)], 6.96 [dd, 1H, H-C(9)], 7.05 [d,
1H, H-C(5)], 7.55 [d, 1H, H-C(3)]; 13C NMR (100 MHz; CD3COD),
δ 13.3 [C(2′)], 60.1 [C(1′)], 113.6 [C(5)], 113.9 [C(2)], 115.0 [C(8)],
121.5 [C(9)], 126,4 [C(4)], 145.3 [C(3)], 145.4 [C(7)], 148.1 [C(6)],
176.9 [C(1)].
p-Coumaric acid ethyl ester: UV-vis (CH3CN/water; pH 2.5), λmax
) 234, 323; LC-MS (ESI+), m/z 193 (100; [M + 1]+), 147 (53; [M -
C2H5 - H2) + 1]+), 165 (21; [M - C2H5 + 1]+); 1H NMR (400 MHz;
CD3OD), δ 1.35 [t, 3H, H-C(2′)], 4.23 [dd, 2H, H-C(1′)], 6.33 [d, 1H,
H-C(2)], 6.82 [d, 2H, H-C(6), H-C(8)], 7.47 [d, 2H, H-C(5), H-C(9)],
7.62 [d, 1H, H-C(3)]; 13C NMR (100 MHz; CD3COD), δ 13.2 [C(2′)],
60.0 [C(1′)], 113.9 [C(2)], 115.4 [C(6), C(8)], 125.8 [C)4)], 129.7 [C(5),
C(9)], 144.9 [C(3)], 159.9 [C(7)], 167.9 [C(1)].
Syringetin-3-O-ꢀ-D-glucopyranoside: UV-vis (CH3CN/water; pH
2.5), λmax ) 252, 355; LC-MS (ESI+), m/z 509 (100; [M + 1]+), 523
1
(60; [M + Na]+), 347 (20; [M - glc + 1]+); H and 13C NMR data
Syringic acid ethyl ester: UV-vis (CH3CN/water; pH 2.5), λmax
)
were identical with those measured for the reference compound.
Quercetin-3-O-ꢀ-D-glucuropyranoside: UV-vis (CH3CN/water; pH
2.5), λmax ) 255, 350; LC-MS (ESI-), m/z 477 (100; [M - 1]-), 301
(20; [M - glu - 1]-); 1H NMR (400 MHz; CD3COD), δ 3.46 [m,
1H, H-C(3′′)], 3.53 [m, 1H, H-C(2′′)], 3.58 [dd, 1H, H-C(4′′)], 3.74
[d, 1H, H-C(5′′)], 5.34 [d, J ) 7.6 Hz 1H, H-C(1′′)], 6.20 [d, 1H,
H-C(6)], 6.39 [d, 1H, H-C(8)], 6.84 [d, 1H, H-C(5′)], 7.62 [dd, 1H,
H-C(6′)], 7.64 [d, 1H, H-C(2′)]; 13C NMR (100 MHz; CD3COD), δ
71.7 [C(4′′)], 74.2 [C(2′′)], 75.9 [C(5′′)], 76.4 [C(3′′)], 93.5 [C(8)], 98.7
[C(6)], 103.1 [C(1′′)], 104.5 [C(10)], 114.5 [C(5′)], 116.0 [C(2′)], 121.7
[C(1′)], 122.3 [C(6′)], 134.2 [C(3)], 144.8 [C(3′)], 148.7 [C(4′)], 157.2
[C(9)], 157.9 [C(2)], 161.9 [C(5)], 164.8 [C(7)], 171.2 [C(6′′)], 178.1
[C(4)].
275; LC-MS (ESI+), m/z 227 (100; [M + 1]+), 199 (28; [M - C2H5
+ 1]+); 1H NMR (400 MHz; DMSO-d6), δ 1.31 [t, 3H, H-C(2′)], 3.81
[s, 6H, H-C(1′′), H-C(2′′)], 4.28 [dd, 2H, H-C(1′)], 7.21 [s, 2H, H-C(3),
H-C(7)]; 13C NMR (100 MHz; DMSO), δ 14.8 [C(2′)], 56.5 [C(1′′),
C(2′′)], 60.9 [C(1′)], 107.1 [C(3), C(7)], 119.9 [C(2)], 141.0 [C(5)],
147.9 [C(4), C(6)], 166.0 [C(1)].
Identification of Flavan-3-ols. Analytical HPLC analysis of the
GAC fractions revealed that the taste compounds detected in HPLC
fractions B-12 and B-14 by means of HPLC/TDA were present in GAC
fractions X and XI. Samples (200 mg) of the individual fractions were
dissolved in 1% aqueous HCOOH (20 mL), and after membrane
filtration, aliquots (2 mL) of the solution were used for preparative
RP-HPLC. The chromatographic column was eluted consecutively with
a solution of 1% aqueous HCOOH and CH3CN (95:5, v/v), then
increasing the CH3CN content to 17% within 35 min, keeping the
CH3CN content constant for 15 min, thereafter increasing the CH3CN
2R,3R-Dihydrokaempferol-3-O-R-L-rhamnoside, engelitin: UV-vis
(CH3CN), λmax ) 290; LC-MS (ESI+), m/z 435 (19; [M + 1]+), 457
(100; [M + Na]+), 311 (10; [M - rha + Na]+), 289 (5; [M - rha +
1
1]+); H NMR (400 MHz; CD3COD), δ 1.09 [d, 3H, H-C(6′′)], 3.26

Taste Compounds in Red Wine
J. Agric. Food Chem., Vol. 56, No. 4, 2008 1379
[dd, 1H, H-C(4′′)], 3.40 [dd, 1H, H-C(2′′)], 3.55 [dd, 1H, H-C(3′′)],
3.91 [d, J ) 1.6 Hz, 1H, H-C(1′′)], 4.16 [m, 1H, H-C(5′′)], 4.52 [d,
1H, H-C(3)], 5.04 [d, 1H, H-C(2)], 5.79 [d, 1H, H-C(6)], 5.82 [d, 1H,
H-C(8)], 6.75 [d, 2H, H-C(3′), H-C(5′)], 7.25 [d, 2H, H-C(2′), H-C(6′)];
13C NMR (100 MHz; CD3COD), δ 16.2 [C(6′′)], 68.9 [C(5′′)], 69.1
[C(2′′)], 70.8 [C(3′′)], 72.1 [C(4′′)], 76.8 [C(3)], 82.7 [C(2)], 94.4 [C(6)],
95.9 [C(8)], 100.4 [C(10)], 100.9 [C(1′′)], 127.1 [C(2′), C(6′)], 128.4
[C(3′), C(5′)], 127.7 [C(1′)], 157.8 [C(4′)], 162.9 [C(9)], 165.9 [C(7)],
167.4 [C(5)], 194.2 [C(4)].
2R,3R-Dihydroquercitin-3-O-R-L-rhamnoside,astilbin:UV-vis(CH3CN/
water; pH 2.5), λmax ) 290; LC-MS (ESI+), m/z 473 (100; [M + Na]+),
923 (49; [2M + Na]+), 451 (34; [M + 1]+), 305 (19; [M - rha +
1
1]+); H NMR (400 MHz; CD3COD), δ 1.19 [d, 3H, H-C(6′′)], 3.32
[dd, 1H, H-C(4′′)], 3.56 [m, 1H, H-C(2′′)], 3.67 [dd, 1H, H-C(3′′)],
4.08 [d, J ) 1.1 Hz 1H, H-C(1′′)], 4.27 [m, 1H, H-C(5′′)], 4.61 [d,
1H, H-C(3)], 5.10 [d, 1H, H-C(2)], 5.91 [d, 1H, H-C(6)], 5.94 [d, 1H,
H-C(8)], 6.82 [d, 1H, H-C(5′)], 6.86 [dd, 1H, H-C(6′)], 6.97 [d, 1H,
H-C(2′)]; 13C NMR (100 MHz; CD3COD), δ 16.3 [C(6′′)], 69.0 [C(5′′)],
70.2 [C(2′′)], 70.6 [C(3′′)], 72.3 [C(4′′)], 77.1 [C(3)], 82.4 [C(2)], 94.6
[C(6)], 96.3 [C(8)], 100.9 [C(10)], 105.4 [C(1′′)], 113.9 [C(2′)], 114.4
[C(5′)], 118.9 [C(6′)], 127.8 [C(1′)], 145.1 [C(3′)], 145.9 [C(4′)], 162.9
[C(9)],166.5 [C(7)], 167.2 [C(5)], 194.5 [C(4)].
Identification of Phenolic Acids and Their Esters. Analytical
HPLC analysis of the individual GAC fractions, followed by UV-vis,
LC-MS/MS, and 1D/2D NMR spectroscopy, led to the identification
of a series of astringent phenolic acids as well as their esters.
Spectroscopic data of the taste compounds caftaric acid, caffeic acid,
p-coumaric acid, syringic acid, protocatechuic acid, gallic acid, vanillic
acid, gallic acid ethyl ester, vanillic acid ethyl ester, protocatechuic
acid ethyl ester, ferulic acid ethyl ester, caffeic acid ethyl ester,
p-coumaric acid ethyl ester, and syringic acid ethyl ester were identical
with those measured for the corresponding reference compounds. The
structures of these compounds were confirmed by cochromatography
with commercially available reference compounds or with synthetic
compounds.
Figure 1. Taste profile of Amarone red wine.
mass fraction UF5 (<5 kDa) and the high molecular mass fraction UR5
(>5 kDa), respectively. For sensory analysis, these materials were
solubilized in “natural” concentrations in a 15% aqueous EtOH
solution.
High-Performance Liquid Chromatography (HPLC). The HPLC
apparatus (Jasco, Gross-Umstadt, Germany) consisted of a PU 1580
type pump with a DG-1580-53 degasser, an LG-1580-02 low-pressure
gradient unit, and an MD1515 diode array detector (DAD). Chromato-
graphic separations were performed on stainless steel columns packed
with ODS-Hypersil, 5 µm, RP-18 material (ThermoHypersil, Kleinos-
theim, Germany) either in an analytical (250 × 4.6 mm i.d., flow rate
) 1.0 mL/min), a semipreparative (250 × 10 mm i.d., flow rate ) 3.5
mL/min), or a preparative scale (250 × 21.2 mm i.d., flow rate ) 20
mL/min).
HPLC/TDA of Fraction C. An aliquot (500 mg) of fraction C was
dissolved in 0.1% aqueous HCOOH (10 mL; pH 2.5) by means of an
ultrasonic bath. After membrane filtration, aliquots (250 µL) were
analyzed by semipreparative RP-HPLC/DAD. With the effluent moni-
tored at 272 nm, chromatography was performed starting with 0.1%
aqueous HCOOH, then increasing the CH3CN content to 17% within
35 min, keeping the CH3CN content constant for additional 15 min,
then raising the CH3CN content to 100% within 15 min and maintaining
this solvent for 5 min. The effluent was separated into 17 fractions,
namely, C-1–C-17, which were individually collected into ice-cooled
brown glass vials. The corresponding fractions obtained from 25 HPLC
runs were combined and solvents evaporated under vacuum. After
lyophilization, fractions were used for TDA and qualitative studies.
Identification of Organic and Inorganic Acids. Formic acid, acetic
acid, lactic acid, (E)-aconitic acid, (Z)-aconitic acid, glutaric acid,
tartaric acid, succinic acid, malic acid, citric acid, isocitric acid,
galacturonic acid, hydrochloric acid, and phosphoric acid were identified
in HPLC fraction C-1 by means of ion exchange chromatography using
a Bio-LC system (Dionex) following the standard protocol (36). The
chromatographic system consisted of a GS 50 gradient pump with an
AS50 autosampler, an AS50 thermal compartment, and an ED50
detector and was operated with a Dionex IonPac AS11-HC column
(250 × 2 mm i.d.) and a Dionex IonPac AG11-HC (50 × 2 mm i.d.)
guard column. Analysis of data was performed with Chromeleon
software v 6.60 SP4.
Isolation of Polymeric Fraction by Means of Ultrafiltration. An
aliquot (250 mL) of the red wine was placed into a VIVACELL 250
static gas pressure filtration system (Vivascience) equipped with a 5
kDa molecular weight cutoff VIVACELL 250 5000 MWCO PES
membrane (Vivascience). After sealing, a nitrogen pressure of 4 bar
was applied using an air pressure controller. During filtration
VIVACELL 250 was moved on a type 3005 GFL laboratory shaker
(GFL) with 200 rpm at room temperature. After filtration, the retentate
was taken up with a 15% aqueous EtOH solution (3 × 100 mL). The
solubilized retentates as well as the filtrate were separately freed of
ethanol under vacuum and freeze-dried to afford the low molecular
Liquid Chromatography-Mass Spectrometry (LC-MS/MS). Spec-
tra were acquired in turbo spray electrospray ionization (ESI) mode
on an API 4000 Q-Trap LC-MS/MS system (AB Sciex Instruments,
Darmstadt, Germany) connected to an HPLC system (Agilent 1100
series, Karlsruhe, Germany). Samples were dissolved in a mixture
(50:50, v/v) of MeOH and 0.1% aqueous HCOOH and were analyzed
by direct loop injection (2–20 µL) at a flow rate of 200 µL/min. The
ion-spray voltage was set at -4500 V in the ESI- mode and at +5500
V in the ESI+ mode, and the temperature was set at 300 °C. Nitrogen
served as the curtain gas (20 psi), gas 1 (35 psi), and gas 2 (40 psi).
The declustering potential was set at -10 to -30 V in the ESI- mode
and at +60 V in the ESI+ mode. The mass spectrometer was operated
in the full-scan mode monitoring positive or negative ions. Fragmenta-
tion of [M - H]- and [M + H]+ molecular ions into specific product
ions was performed in enhanced product ion (EPI) mode induced by
collision with nitrogen (4 × 10–5 Torr) with a collision energy of +30
V in the positive and -30 V in the negative mode. Analysis of mass
spectroscopic data was performed by means of Analyst software v
1.4.1.
NMR Spectroscopy. The 1D/2D NMR experiments were performed
on a Bruker DPX 400 spectrometer (Rheinstetten, Germany) using
MeOH-d4 or DMSO-d6 as the solvent. Data processing was performed
by using the NMR software Mestre-C v 1.4.
RESULTS AND DISCUSSION
Preliminary wine tasting involving wine taste experts revealed
the Amarone della Valpolicella as an excellent red wine with a
velvety mouthcoating onset and a balanced astringent offset
accompanied by a mild bitter taste and pronounced mouthfulness
and body. To evaluate the taste profile of that red wine on a
scientific basis, a taste profile analysis using a five-point scale
was performed (Figure 1). A high score of 4.0 was given for
body/mouthfulness as well as for the puckering astringent offset,
followed by sourness (3.0). The velvety mouthcoating onset,

1380 J. Agric. Food Chem., Vol. 56, No. 4, 2008
Hufnagel and Hofmann
Table 1. Yields and Sensorial Evaluation of Fractions A-C Isolated from
Red Wine
intensitya perceived for
fractionb astringency bitterness sourness sweetness mouthfulness/ body
A
B
C
0.3
1.8
3.4
0.2
1.3
0.4
0.0
2.2
3.3
0.0
0.2
1.0
0.0
0.0
4.0
a The taste intensity of aqueous solutions of the individual fractions isolated
from 100 mL of Amarone red wine in bottled water is rated on a scale from 0 (not
detectable) to 5.0 (strongly detectable). b Individual fractions contain the n-pentane
solubles (A), ethyl acetate extractables (B), and water solubles (C) isolated from
Amarone red wine.
perceived by the sensory panel in the first part of the sensation,
was judged with an intensity of 2.5, closely followed by
sweetness (2.0). In comparison, bitterness was perceived with
a somewhat lower intensity of 1.5, whereas umami and salty
tastes were not detected at all.
To gain first insight into the hydrophobicity of the compounds
imparting the typical astringent as well as bitter taste sensation,
the red wine was extracted sequentially with solvents of
increasing polarity.
Figure 2. RP-HPLC chromatogram (left side) and taste dilution analysis
(right side) of fraction B isolated from red wine.
Table 2. Yields, Taste Qualities and Taste Dilution (TD) Factors of HPLC
Fractions Isolated from Amarone Red Wine Fraction B
Solvent Fractionation of Red Wine. A red wine sample was
extracted sequentially with n-pentane, followed by EtOAc, and
the extracts obtained were freed from solvent under vacuum to
give fractions A and B in yields of 0.1 and 4.6%, respectively.
The remaining aqueous layer was freeze-dried to give the water-
solubles (fraction C) accounting for >95% of the dry mass of
the red wine (Table 1). Fractions A-C were individually
dissolved in water (pH 4.5), each in natural concentration, and
the solutions obtained were then sensorially evaluated. Whereas
fraction A was found to be nearly tasteless, fractions B and C
reflected taste impressions already found for the red wine,
respectively (Table 1). Fraction C induced an puckering oral
astringent sensation as well as sweetness in high intensities of
3.4 and 1.0, respectively. Although fraction B was judged with
a somewhat lower score for total astringency (1.8), bitterness
was perceived with high intensity (1.3). In addition, sourness
was detected predominantly in fractions B and C, whereas body/
mouthfulness was exclusively perceived in fraction C. On the
basis of the results of these sensory data, the following
identification of astringent and bitter compounds was focused
on fractions B and C.
Taste-Active Compounds in Fraction B. To separate the
bulk of tasteless or less taste-active components from the
intensely tasting compounds, fraction B was fractionated by
means of RP-HPLC/DAD to give 30 fractions, namely, fractions
B-1-B-30 (Figure 2). To evaluate their taste impact, these
HPLC fractions were freeze-dried, taken up in water, and then
analyzed by means of the TDA using the half-tongue test (33–35).
The highest taste impact was located in fraction B-2 judged
with a TD factor of 256 for astringency, followed by fractions
B-6, B-12, and B-14 evaluated with a TD factor of 128 (Table
2). Fractions B-16 and B-29 still showed an astringent oral
sensation after a dilution of 1:32, and fractions B-1, B-11, B-21,
B-25, and B-27 were taste-active up to a dilution of 1:16.
Besides astringency, sourness was detected in fractions B-1–B-
8, and a bitter taste quality was found for fractions B-14, B-27,
and B-29 evaluated with a TD factor of 4 (Table 2).
fractiona
B-1
taste qualityb
sour
puckering astringent
puckering astringent
sour
TD factorb
compounds identifiedc
64
16
256
2
B-2
1
B-3
B-4
sour
sour
4
4
B-5
B-6
B-7
B-8
sour
4
128
2
puckering astringent
sour
sour
6
2
B-9
puckering astringent
puckering astringent
puckering astringent
puckering astringent
puckering astringent
puckering astringent
bitter
8
2
16
23, 24, 25
11
3, 4, 5
26
6, 7
B-10
B-11
B-12
B-13
B-14
16
128
4
128
4
6, 7
B-15
B-16
B-17
B-18
B-19
B-20
B-21
B-22
B-23
B-24
B-25
B-26
B-27
velvety astringent
puckering astringent
puckering astringent
puckering astringent
velvety astringent
velvety astringent
puckering astringent
puckering astringent
velvety astringent
puckering astringent
velvety astringent
puckering astringent
puckering astringent
bitter
2
32
4
4
8
4
16
4
8
2
8
22
12, 13, 14
17, 18
15
16
4
16
4
19, 20, 21
19, 20, 21
B-28
B-29
velvety astringent
puckering astringent
bitter
4
32
4
9, 10
9, 10
B-30
puckering astringent
2
a Number of HPLC fraction referring to Figure 2. b The taste quality and TD
factor were determined by using a half-tongue duo test. c The structures of the
compounds given as numbers are displayed in Figure 3.
To gain a first insight into the compounds inducing the taste
impression in the most active HPLC fractions, the fractions
evaluated with the highest TD factors for astringency and
bitterness were investigated by means of HPLC-MS operating
in the ESI+ mode. Fractions B-2 and B-6, both judged with
high TD factors for astringency, each contained only a single
component showing the pseudomolecular ions m/z 171 and 155,
respectively. Upon comparison of the UV-vis, LC-MS, and
NMR spectroscopic data with those found for the reference
compound, the key tastants in fractions B-2 and B-6 were

Taste Compounds in Red Wine
J. Agric. Food Chem., Vol. 56, No. 4, 2008 1381
Figure 3. Chemical structures of phenolic taste compounds identified in Amarone red wine.
identified as gallic acid (1) and protocatechuic acid (2),
respectively (Figure 3).
of spectroscopic (LC-MS/MS, UV-vis), chromatographic, and
sensory data with those obtained for the reference compound,
this compound was unequivocally identified as procyanidin B2
(5) (Figure 3), fitting well with data reported in the literature
(38).
Due to its high TD factor (Table 2), HPLC fraction B-14
was investigated next. LC-MS analysis of this fraction revealed
the presence of two compounds, one showed a pseudomolecular
ion of m/z 199 in the ESI+ mode, whereas the second compound
exhibited m/z 865 as the [M - H]- ion in the ESI- mode.
Fragmentation of the ion m/z 199 resulted in a daughter ion
with m/z 171 corresponding to the [M + 1]+ ion expected for
gallic acid. The loss of 28 amu as well as the galloyl-type
The astringent fraction B-12, judged with a TD factor of 128,
was more complex and contained three individual compounds.
Rechromatography by means of RP-HPLC revealed two com-
ponents showing the mass transitions m/z 181 f 163 and m/z
199 f 155, respectively, and were identified as caffeic acid
(3) and syringic acid (4) by comparison of the chromatographic,
spectroscopic, and sensory data with those obtained for the
reference compounds (Figure 3), thus confirming earlier
literature reports (37). The third component of fraction B-12
showed a pseudomolecular ion of m/z 579 in the ESI+ mode,
thus suggesting the presence of a procyanidin. By comparison

1382 J. Agric. Food Chem., Vol. 56, No. 4, 2008
Hufnagel and Hofmann
Figure 4. Gel adsorption chromatography of fraction B isolated from red
wine.
UV-vis spectrum of this compound suggested this astringent
and bitter compound to be gallic acid ethyl ester (6) (Figure
3). This was confirmed by comparison of spectroscopic and
chromatographic data, followed by cochromatography with the
reference compound. Although this compound was recently
identified in red wine (39), its bitter taste quality has not been
reported so far. In addition, the second taste compound eluting
in fraction B-14 was identified as the trimeric procyanidin C1
(7) by comparison with the reference compound (Figure 3),
thus confirming earlier reports in the literature (40).
Judged with a TD factor of 32, the astringent compound in
fraction B-16 was analyzed by means of LC-MS/MS, NMR,
and UV-vis spectroscopy, as well as by sensory analysis. On
the basis of comparison of the data obtained with those found
for the reference compounds, this astringent compound was
identified as p-coumaric acid (8) (Figure 3), which was
identified in red wine already 50 years ago (41).
Furthermore, fraction B-29, evaluated as astringent and bitter,
was analyzed by means of LC-MS. By using the ESI+ mode,
two compounds were detected showing pseudomolecular ions
of m/z 193 and 223, respectively. Fragmentation of both of these
pseudomolecular ions revealed a neutral loss of 28 amu, thus
implying the presence of one ethyl ester moiety. After isolation
and purification, the taste compound LC-MS/MS and NMR data
revealed the structure of the two taste compounds as ferulic
acid ethyl ester (9) and p-coumaric acid ethyl ester (10) (Figure
3). The identity of both compounds was confirmed by com-
parison of spectroscopic and chromatographic data with those
obtained for the corresponding reference compound. Although
the p-coumaric acid ethyl ester has already been isolated from
red and white wines (42, 43) and ferulic acid ethyl ester has
been identified in white wines (44), the bitter taste quality of
both esters has not been reported so far.
Evaluated with a TD factor of 16, fraction B-11 was
investigated next. LC-MS experiments revealed a pseudomo-
lecular ion with m/z 169 [M + 1]+ and a daughter ion with m/z
151, most likely due to the cleavage of one molecule of water.
To isolate suitable amounts for NMR spectroscopic structure
elucidation, fraction B was fractionated in a preparative scale
by means of GAC on Sephadex LH 20 to give the fractions
I–VIII (Figure 4). By comparison of spectroscopic data and
chromatographic data, this taste compound was identified to be
present in high amounts in GAC fraction V (Figure 5). After
isolation, NMR as well as LC-MS analysis identified the
Figure 5. RP-HPLC chromatograms of GAC fractions V-VIII isolated
from red wine fraction B.
astringent compound as vanillic acid (11) (Figure 3), well-
known as a red wine constituent (39).
The identification of the astringent compounds in fraction
B-21 seemed to be more challenging. LC-MS analysis of the
key astringent compound revealed several ions. In the ESI+
mode, the predominant ion detected showed m/z1 451, followed
by m/z2 183. In addition, the ion m/z3 477 was detected in the
ESI- mode. By HPLC degustation as well as comparison of
the LC-MS data obtained, two compounds, exhibiting pseudo-
molecular ions with m/z 451 [M + 1]+ and m/z 477 [M - 1]-,
respectively, were found to be enriched in GAC fraction VIII
(Figure 5). After purification by means of analytical RP-HPLC,
the structure of both compounds was elucidated by means of
NMR spectroscopy. Fragmentation of the minor compound,
showing m/z 477 [M - 1]- in the ESI- mode, revealed a
daughter ion of m/z 301 as expected for a quercetin aglycone.
This was supported by the adsorption maxima observed at 255
and 350 nm in the UV-vis spectrum. Also, NMR data
confirmed the presence of a quercetin aglycon as well as a
glycoside moiety, but the methylene proton signals expected
for H-C(6′′) of the hexose moiety were lacking. Due to the
unusually high 13C chemical shift of 171.2 ppm observed for
carbon C(6′′) as well as the neutral loss of 176 amu upon MS/
MS analysis, the taste compound was identified as quercetin-
3-O-ꢀ-D-glucuropyranoside (12) (Figure 3). Although this
glucuronic acid derivative was identified earlier (45), its
astringent activity was not reported so far. The major compound,
exhibiting the pseudomolecular ions m/z 451 [M + 1]+ and m/z
473 [M + Na+]+ in the LC-MS spectrum, showed a rather
different UV-vis spectrum with only one absorption maximum
centering around 290 nm. Fragmentation of the pseudomolecular
ion [M + 1]+ resulted in a daughter ion with m/z 305, indicating
the cleavage of one molecule of rhamnose. The difference of 2
amu between the aglycone of this taste compound and quercetin
suggested a dihydroquercetin as the aglycon. 1D/2D NMR
experiments and comparison of the spectroscopic data with those
reported earlier in the literature (46) enabled the determination
of the structure of that astringent compound as 2R,3R-dihy-
droquercitin-3-O-R-L-rhamnoside (13), known as astilbin (Fig-
ure 3) (46). The third compound present in fraction B-21 was

Taste Compounds in Red Wine
J. Agric. Food Chem., Vol. 56, No. 4, 2008 1383
detected in GAC fraction VI (Figure 5). After isolation, NMR,
LC-MS/MS, and UV-vis analyses resulted in the identification
of that taste compound as protocatechuic acid ethyl ester (14)
(Figure 3). Although this ester has already been identified in
wines (47), its bitter taste quality has not been reported so far.
Fraction B-25, evaluated with a TD factor of 16 for
astringency, showed UV-vis absorption maxima at 252 and
355 nm and two pseudomolecular ions with m/z 509 [M + 1]+
and 523 [M + Na+]+, respectively, indicating a molecular mass
of 508 Da. LC-MS/MS (ESI+) experiments revealed a neutral
loss of 162 amu between the parent ion m/z 509 and the daughter
ion m/z 347 as expected for the cleavage of one molecule of
hexose. HPLC-DAD and HPLC degustation identified the same
tastant to be present in GAC fraction VII (Figure 5), from which
the target compound could be isolated in amounts suitable for
1
NMR experiments. The H NMR data revealed two methoxy
groups located at a flavonol-like aglycone. By means of homo-
(gCOSY) and heteronuclear (HMBC) chemical shift correlation
experiments, the taste compound was identified as syringetin-
3-O-ꢀ-D-glucopyranoside (15; Figure 3). Although this com-
pound has already been detected in red wine (48), the astringent
activity of this glycoside has not been reported so far.
Furthermore, quercetin-3-O-ꢀ-D-galactopyranoside (22; Figure
3) was identified as the velvety astringent compound detected
in fraction B-19, thus confirming data published earlier for
Spanish wines (49).
Figure 6. COSY NMR spectrum (400 MHz, MeOD) of 2R,3R-dihydro-
kaempferol-3-O-R-L-rhamnoside (18).
209, respectively, and upon a neutral loss of 28 amu the daughter
ions m/z 199, 179, and 181, thus suggesting the presence of
ethyl esters. 1H NMR analysis of the main compound 19,
isolated from GAC fractions VII and VIII (Figure 5), showed
a triplet at 1.36 ppm integrating for three protons and coupling
with a double duplet resonating at 4.23 ppm, thus confirming
the ethyl ester function in the molecule. In addition, NMR
spectroscopy showed the typical signal pattern expected for
caffeic acid and allowed the identification of that taste compound
as caffeic acid ethyl ester. Furthermore, spectroscopic analysis
of the other two compounds eluting in fraction B-27 led to
syringic acid ethyl ester and vanillic acid ethyl ester as the
proposed structures. To verify the identity of these ethyl esters,
vanillic acid ethyl ester (20; Figure 3) and syringic acid ethyl
ester (21; Figure 3) were synthesized, and the spectroscopic,
chromatographic, and sensory data were compared to those
obtained from the isolates. On the basis of the identity of the
data obtained, the three bitter compounds detected in fraction
B-27 were unequivocally identified as the ethyl esters of caffeic
acid (19), vanillic acid (20), and syringic acid (21). Although
compounds 19 (53–55), 20 (56), and 21 (57) have already been
identified in wine as well as wine vinegar, the bitter taste activity
of these esters was not known.
As literature studies (19) reported on the flavan-3-ol mono-
mers (+)-catechin and (-)-epicatechin as well as the dimeric
procyanidins B1 and B3, respectively, as important taste
compounds in red wine, additionally the HPLC fractions
B-1-B-30 were screened for these compounds by means of
HPLC-MS/MS using the corresponding reference compounds
for MS tuning experiments. This MS-based screening led to
the identification of (+)-catechin (23), procyanidin B1 (24), and
procyanidin B3 (25) in fraction B-10 and (-)-epicatechin (26;
Figure 3) in fraction B-13. However, it is interesting to note
that the flavan-3-ol-containing fractions B-10 and B-13 were
not evaluated as tasting bitter by means of the TDA. In contrast,
the bitter-tasting fractions B-14, B-27, and B-29 were found to
contain the bitter-tasting ethyl esters of hydroxybenzoic and
hydroxycinnamic acids.
The astringent fraction B-9, evaluated with a TD factor of 8,
showed a pseudomolecular ion of m/z 311 in the LC-MS (ESI-)
spectrum. Detection of the same compound in GAC fraction V
(Figure 5), followed by isolation, NMR analysis, and compari-
son of the spectroscopic and chromatographic data with those
observed for the reference compound, led to the identification
of the taste-active substance in fraction B-9 as caftaric acid (16;
Figure 3), the taste contribution of which was declined to be
of interest in former studies (50).
HPLC-MS analysis of the astringent fraction B-23 gave two
compounds, 17 and 18, showing the pseudomolecular ions m/z
479 and m/z 435, respectively. MS/MS analysis of compound
17 indicated the cleavage of one molecule of hexose (162 amu)
from m/z 479 to give the daughter ion m/z 317, thus indicating
the presence of isorhamnetin as aglycone. Finally, cochromatog-
raphy and comparison of UV-vis spectra as well as MS data
identified this substance as isorhamnetin-3-O-ꢀ-D-glucopyra-
noside (17; Figure 3), thus confirming earlier literature reports
(51). The astringent compound detected with the pseudomo-
lecular ion m/z 435 showed a loss of 146 amu to give the
daughter ion m/z 289, thus demonstrating the presence of a
molecule of rhamnose. In addition, this compound exhibited a
UV-vis spectrum which was rather similar to that observed
for astilbin. Assuming this compound to be a dihydroflavonol
rhamnoside, 1H NMR analysis of the compound, isolated from
GAC fraction VIII (Figure 5), showed two strongly coupling
protons resonating at 5.04 and 4.52 ppm, respectively, similar
to those found in the NMR spectrum of astilbin (Figure 6).
Careful interpretation of all NMR data obtained led to the
identification of this taste compound as 2R,3R-dihydrokaempferol-
3-O-R-L-rhamnoside (18; Figure 3), known as engelitin.
Although engelitin has been identified earlier in grapes as well
as white wines (46, 52), this compound has not been previously
reported as a constituent of red wine, nor has its taste activity
been reported before.
Fraction B-27 was evaluated as one of three bitter-tasting
HPLC fractions. LC-MS/MS analysis revealed three compounds
exhibiting the pseudomolecular ions m/z 227, m/z 197, and m/z
Taste-Active Compounds in Fraction C. To locate the most
taste-active compounds in the water-soluble fraction isolated

1384 J. Agric. Food Chem., Vol. 56, No. 4, 2008
Hufnagel and Hofmann
Table 3. Sensorial Evaluation of Fractions Isolated by Means of
Ultrafiltration from Amarone Red Wine
intensityb perceived for
astringency bitterness sourness sweetness mouthfulness/body
samplea
red wine
LMW fractiona
HMW fractionc
4.0
3.1
3.7
1.5
1.5
0.2
3.0
2.7
0.3
2.0
2.0
0.0
4.0
3.7
0.3
a Low molecular weight fraction (<5 kDa) isolated from red wine by means of
ultrafiltration. b High molecular weight fraction (g5 kDa) isolated from red wine by
means of by ultrafiltration. c The taste intensity of solutions of the individual fractions
in 15% aqueous EtOH (50 mL, pH 4.5) was rated on a scale from 0 (not detectable)
to 5.0 (strongly detectable).
mouthfulness/body as well as the other taste qualities bitterness,
sourness, and sweetness either were evaluated with rather low
scores or were not detectable at all. These data clearly
demonstrate that besides the LMW compounds identified by
means of HPLC-TDA, also the HMW components contribute
to the astringent taste of red wine, thus fitting well with literature
data (11–13, 18).
Sensory Activity of Taste-Active Compounds. Prior to
sensory analysis, the purity of all compounds was checked by
HPLC-MS as well as 1H NMR spectroscopy. To determine the
human threshold concentrations for bitter taste and the astringent
oral sensation, aqueous solutions of the target compound were
evaluated by means of the three-alternative forced-choice test
(35) and half-tongue test (33–35), respectively. As given in
Table 4, the individual taste compounds have been grouped
into three classes: a group of velvety astringent compounds, a
group of puckering astringent compounds, and a group of bitter
and astringent compounds, respectively.
Figure 7. RP-HPLC chromatogram (left side) and taste dilution analysis
(right side) of red wine fraction C.
from red wine, fraction C was separated by means of RP-HPLC
and the TDA was applied onto the collected fractions C-1-C-
17 (Figure 7). Interestingly, only two fractions, namely,
fractions C-1 and C-16, were evaluated with TD factors above
16. Fraction C-16 was judged with a TD factor of 128 for
astringency, and fraction C-1 was evaluated with TD factors of
64, 32, and 16 for astringency, sourness, and sweetness,
respectively.
Analysis of fraction C-1 by means of HPLC-DAD as well
as ion chromatography revealed a broad spectrum of 14 or-
ganic acids and inorganic salts, namely, formic acid, acetic acid,
lactic acid, (E)-aconitic acid, (Z)-aconitic acid, glutaric acid,
tartaric acid, succinic acid, malic acid, citric acid, isocitric acid,
galacturonic acid, chloride, and phosphate. Sensory analysis
revealed that, among these organic acids, exclusively (E)/(Z)-
aconitic acid (27) induced a strongly puckering astringent
sensation.
Fraction C-16, exhibiting a rather shrinking and puckering
type of astringency, exhibited a UV-vis spectrum which was
similar to that of procyanidins. As LC-MS/MS experiments did
not reveal any reliable results, this fraction was suggested to
be composed of polymeric structures. To gain further insight
into the molecular weight of the components inducing the
puckering taste sensation, an aliquot of fraction C was separated
by means of ultrafiltration using a molecular weight cutoff of 5
kDa. The low molecular weight fraction (LMW) and the fraction
containing high molecular weight components (HMW, >5 kDa)
were freeze-dried and, after the residues had been dissolved in
15% aqueous EtOH in their “natural” concentration, were
evaluated by means of a taste profile analysis (Table 3). The
panelists judged the LMW fraction to exhibit the entire
sweetness of the wine judged with a score of 2.0. Furthermore,
sourness, bitterness, and mouthfulness/body induced by the
LMW fraction were evaluated with only slightly lower intensi-
ties when compared to the red wine. The astringency imparted
by the LMW fraction was judged with a rather high intensity
of 3.1, but the quality was described to be more velvety and
silky when compared to the HMW fraction, which was evaluated
as more puckering astringent with an intensity of 3.7. The
Among the velvety astringent compounds, syringetin-3-O-
ꢀ-D-glucopyranoside (15) and quercitin-3-O-ꢀ-D-galactopyra-
noside (22) were evaluated with the lowest recognition threshold
concentrations of 0.2 and 0.4 µmol/L, followed by the other flavon-
3-ol and dihydroflavon-3-ol glycosides 12, 17, 13, and 18 with
about 10-fold higher taste thresholds. The low threshold concentra-
tions of these velvety astringent compounds are well in line with
those reported for other flavon-3-ol glycosides (33, 34).
The group of puckering astringent compounds contained
various hydroxybenzoic acids, cinnamic acid derivatives, and
the polymeric fraction (UR5) isolated from red wine, as well
as (E)/(Z)-aconitic acid (Table 4). The lowest taste thresholds
were found for the aconitic acid inducing an astringent sensation
already at the low concentration of 0.5 µmol/L, whereas its
typical sour taste was perceived not below a concentration of
500 µmol/L. All of the other compounds in that group were
significantly less taste active, and vanillic acid was judged with
the highest taste threshold of 315 µmol/L. The threshold found
for the polymeric fraction was 22.0 mg/L, which is in the range
of the low molecular weight phenols 1-4 (Table 4).
The group of compounds inducing a bitter and astringent oral
sensation consisted of monomeric and dimeric flavan-3-ols and
various hydroxybenzoic acid and cinnamic acid ethyl esters
(Table 4). The human threshold concentrations for the astrin-
gency of these flavan-3-ols ranged from 200 to 930 µmol/L and,
with the exception of (-)-epicatechin, were always below the
recognition thresholds determined for their bitter taste. The
astringent taste threshold decreased from the flavan-3-ol mono-
mers (+)-catechin and (-)-epicatechin over the dimeric pro-
cyanidins B1, B2, and B3, to the trimeric procyanidin C1, thus
being well in line with data published on procyanidins from
cocoa (Theobroma cocoa L.) (34). The bitter taste thresholds

Taste Compounds in Red Wine
J. Agric. Food Chem., Vol. 56, No. 4, 2008 1385
Table 4. Human Taste Recognition Thresholds of Compounds Isolated
from Red Wine
for wine taste and, in particular, the different kinds of
astringency perceived.
taste thresholdb for
LITERATURE CITED
astringency
µmol/L mg/L
Velvety Astringent Compounds
bitterness
(1) Gawel, R.; Iland, G. G.; Francis, I. L. Characterizing the
astringency of red wine: a case study. Food Qual. Pref. 2001,
12, 83–94.
(2) Haslam, E.; Lilley, T. H. Natural astringency in foodstuffs - a
molecular interpretation. Crit. ReV. Food Sci. Nutr. 1988, 27, 1–
41.
(3) Luck, G.; Hua, L.; Murray, N. J.; Grimmer, H. R.; Warminski,
E. E.; Williamson, M. P.; Lilley, T. H.; Haslam, E. Polyphenols,
astringency and proline-rich proteins. Phytochemistry 1994, 37,
357–371.
(4) De Freitas, V.; Mateus, N. Structural features of procyanidin
interactions with salivary proteins. J. Agric. Food Chem. 2001,
49, 940–945.
(5) Breslin, P. A. S.; Gilmore, M. M.; Beachamp, G. K.; Green, B. G.
Psychophysical evidence that oral astringency is a tactile sensation.
Chem. Senses 1993, 18, 405–417.
(6) Bate-Smith, E. C.; Swain, T. In ComparatiVe Biochemistry;
Mason, H. S., Florkin, A. M., Eds.; Academic Press: New York,
1962; pp 755–809.
(7) Adams, D. O.; Habertson, J. Use of alkaline phosphatase for the
analysis of tannins in grapes and red wines. Am. J. Enol. Vitic.
1990, 50, 247–252.
(8) Monteleone, E.; Condelli, N.; Dinnella, C.; Bertuccioli, M.
Prediction of perceived astringency induced by phenolic com-
pounds. Food Qual. Pref. 2004, 15, 761–769.
(9) McManus, J.; Lilley, T. M.; Haslam, E. Plant polyphenols and
their association with proteins. In Plant Resistance to Insects;
Hedin, P. A., Ed.; ACS Symposium Series 208; American
Chemical Society: Washington, DC, 1983; pp 123–137.
(10) Yokotsuka, K.; Singleton, V. L. Interactive precipitation between
graded peptides from gelatin and specific grape tannin fractions
in wine-like model solutions. Am. J. Enol. Vitic. 1987, 38, 199–
206.
(11) Nobel, A. C. Why do wines taste bitter and feel astringent? In
Chemistry of Wine FlaVour; Waterhouse, H. L., Ebeler, S. E.,
Eds.; ACS Symposium Series 714; American Chemical Society:
Washington, DC, 1998; pp 156–165.
compounda (no.)
µmol/L
mg/L
dihydrokaempferol-3-O-R-L-rhap (18)
4.8
3.7
2.4
2.0
0.4
0.2
2.1
1.7
1.1
1.0
0.2
0.1
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
dihydroquercitin-3-O-R-L-rhap (13)
isorhamnetin-3-O-ꢀ-D-glcp (17)
quercetin-3-O-ꢀ-D-glcAp (12)
quercetin-3-O-ꢀ-D-galp (22)
syringetin-3-O-ꢀ-D-glcp (15)
Puckering Astringent Compounds
vanillic acid (11)
gallic acid (1)
syringic acid (4)
protocatechuic acid (2)
p-coumaric acid (8)
ferulic acid
315
292
263
206
139
67
53
50
52
32
23
13
13
5
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
caffeic acid (3)
72
16
(E)-caftaric acid (16)
(Z)/(E)-aconitic acid (27)c
polymeric fraction (>5 kDa)
0.5
0.1
22
Bitter and Astringent Compounds
(-)-epicatechin (26)
(+)-catechin (23)
procyanidin C1 (7)
930
410
300
277
240
200
190
185
143
125
67
270
119
260
58
139
116
110
37
27
25
15
9
930
1000
400
1100
400
500
485
2200
715
1500
710
1000
576
270
290
347
229
231
289
280
438
137
294
158
182
130
caffeic acid ethyl ester (19)
procyanidin B1 (24)
procyanidin B3 (25)
procyanidin B2 (5)
gallic acid ethyl ester (6)
p-coumaric acid ethyl ester (10)
vanillic acid ethyl ester (20)
ferulic acid ethyl ester (9)
protocatechuic acid ethyl ester (14)
syringic acid ethyl ester (21)
49
18
4
a The structures of the compounds are displayed in Figure 3. b Taste threshold
concentrations were determined in bottled water by means of a triangle test for
bitterness and by means of the half-tongue test for astringency. c Taste threshold
for sour is 500 µmol/L.
(12) Kielhorn, S.; Thorngate, J. H., III. Oral sensations associated with
the flavan-3-ols (+)-catechin and (-)-epicatechin. Food Qual.
Pref. 1999, 10, 109–116.
of these procyanidins were about 2-fold below the values found
for the flavan-3-ol monomers.
(13) Peleg, H.; Gacon, K.; Schlich, P.; Noble, A. C. Bitterness and
astringency of flavan-3-ol monomers, dimers and trimers. J. Sci.
Food Agric. 1999, 79, 1123–1128.
(14) Delcour, J. A.; Vandenberghe, M. M.; Corten, P. F.; Dondeyne,
P. Flavor thresholds of polyphenolics in Water. Am. J. Enol. Vitic.
1984, 35, 134–136.
(15) Noble, A. C. Bitterness and astringency in wine. In Bitterness in
Foods and BeVerage; Rousseff, R. L., Ed.; Elsevier Sciences: New
York, 1990; pp 145-158.
(16) Nobel, A. C. Astringency and bitterness of flavanoid phenols In
Chemistry of Taste; Given, P., Paredes, D., Eds.; ACS Symposium
Series 825; American Chemical Society: Washington, DC, 2002;
pp 193–201.
The sensory properties of phenolic acid ethyl esters were
described as primarily puckering astringent and were evaluated
with taste thresholds ranging from 18 to 185 µmol/L; for
example, syringic acid ethyl ester was found with the lowest
astringency threshold of 18 µmol/L in that group. It is interesting
to note that in contrast to the procyanidins the bitter taste
thresholds of all the phenolic acid ethyl esters were significantly
higher than their astringency threshold concentrations; for
example, the bitter threshold for protocatechuic acid ethyl ester
was about 20 times above the astringency threshold of that
compound (Table 4).
In conclusion, the data obtained by application of a sensomics
approach give strong evidence that velvety astringent and
puckering astringent as well as bitter and astringent compounds
contribute to the typical astringent and bitter taste of red wine.
In contrast to previous studies; most surprisingly, the procya-
nidin monomers, dimers, and trimers seem not to be the key
elicitors of the astringent taste of the red wine, but a variety of
different low molecular weight polyphenols as well as a
polymeric fraction were found with highest taste impacts.
Quantitative studies as well as taste reconstruction and omission
experiments in a wine-like matrix are currently ongoing to get
a lead on the importance of the individual substance classes
(17) Lesschaeve, I.; Noble, A. C. Polyphenols: factors influencing their
sensory properties and their effects on food and beverage
preferences. Am. J. Clin. Nutr. 2005, 81, 330S–335S.
(18) Vidal, S.; Francis, L.; Noble, A. C.; Kwiatkowski, M.; Cheynier,
V.; Waters, E. Taste and mouth-feel properties of different types
of tannin-like polyphenolic compounds and anthocyanins in wine.
Anal. Chim. Acta 2004, 57–65.
(19) Arnold, R. A.; Noble, A. C.; Singleton, V. L. Bitterness and
astringency of phenolic fractions in wine. J. Agric. Food Chem.
1980, 28, 675–678.
(20) Robichaud, J. L.; Noble, A. C. Astringency and bitterness of
selected phenolics in wine. J. Sci. Food Agric. 1990, 53, 343–
353.

1386 J. Agric. Food Chem., Vol. 56, No. 4, 2008
Hufnagel and Hofmann
(21) Cheynier, V.; Fulcrand, H.; Brossaud, F.; Asselin, C.; Moutounet,
M. Phenolic composition as related to red wine flavour. In
Chemistry of Wine FlaVour; Waterhouse, H. L., Ebeler, S. E.,
Eds.; ACS Symposium Series 714; American Chemical Society:
Washington, DC, 1998; pp 124–141.
electrode liquid chromatography-electrochemistry. Analyst 1988,
113, 99–102.
(41) Hennig, K.; Burkhardt, R. Detection of phenolic compounds and
of aromatic hydroxy acids in grape constituents, wine, and wine-
like beverages. I. Weinberg Keller 1958, 542–552.
(22) Lawless, H. T.; Corrigan, C. J.; Lee, C. B. Interactions of astringent
substances. Chem. Senses 1994, 19, 141–154.
(23) Kallithraka, S.; Bakker, J.; Clifford, M. N. Red wine and model
red wine astringency as affected by malic and lactic acid. J. Food
Sci. 1997, 2, 416–420.
(42) Vitrac, X.; Castagnino, C.; Waffo-Teguo, P.; Delaunay, J.-C.;
Vercauteren, J.; Monti, J.-P.; Deffieux, G.; Merillon, J.-M.
Polyphenols newly extracted in red wine from southwestern
France by centrifugal partition chromatography. J. Agric. Food
Chem. 2001, 49, 5934–5938.
(24) Kallithraka, S.; Bakker, J.; Clifford, M. N. Effect of pH on
astringency in model solutions and wines. J. Agric. Food Chem.
1997, 45, 2211–2216.
(25) Preys, S.; Mazerolles, G.; Courcoux, P.; Samson, A.; Fischer, U.;
Hanafi, M.; Bertrand, D.; Cheynier, V. Relationship between
polyphenolic composition and some sensory properties in red
wines using multiway analyses. Anal. Chim. Acta 2006, 563, 126–
136.
(26) Noble, A. C. Bitterness in wine. Physiol. BehaV. 1994, 56, 1251–
1255.
(27) Noble, A. C.; Strauss, C. R.; Williams, P. J.; Wilson, B.
Contribution of terpene glycosides to bitterness in Muscat wines.
Am. J. Enol. Vitic. 1988, 39, 129–131.
(28) Boselli, E.; Boulton, R. B.; Thorngate, J. H.; Frega, N. G.
Chemical and sensory characterization of red wines from Marche
(Italy) related to vintage and grape cultivars. J. Agric. Food Chem.
2004, 52, 3843–3854.
(29) Frank, O.; Ottinger, H.; Hofmann, T. Characterization of an intense
bitter-tasting 1H,4H-quinolizinium-7-olate by application of the
taste dilution analysis, a novel bioassay for the screening and
identification of taste-active compounds in foods. J. Agric. Food
Chem. 2001, 49, 231–238.
(30) Ottinger, H.; Bareth, A.; Hofmann, T. Characterization of natural
“cooling” compounds formed from glucose and L-proline in dark
malt by application of taste dilution analysis. J. Agric. Food Chem.
2001, 49, 1336–1344.
(31) Czepa, A.; Hofmann, T. Structural and sensory characterization
of compounds contributing to the bitter off-taste of carrots (Daucus
carota L.) and carrot puree. J. Agric. Food Chem. 2003, 51, 3865–
3873.
(32) Ottinger, H.; Hofmann, T. Identification of the taste enhancer
alapyridaine in beef broth and evaluation of its sensory impact
by taste reconstitution experiments. J. Agric. Food Chem., 2003,
51, 6791–6796.
(33) Scharbert, S.; Holzmann, N.; Hofmann, T. Identification of the
astringent taste compounds in black tea infusions by combining
instrumental analysis and human bioresponse. J. Agric. Food
Chem. 2004, 52, 3498–3508.
(34) Stark, T.; Bareuther, S.; Hofmann, T. Sensory-guided decomposi-
tion of roasted cocoa nibs (Theobroma cacao) and structure
determination of taste-active polyphenols. J. Agric. Food Chem.
2005, 53, 5407–5418.
(35) Glabansnia, A.; Hofmann, T. Sensory-directed identification of
taste-active ellagitannins in American (Quercus alba L.) and
European oak wood (Quercus robur L.) and quantitative analysis
in bourbon whiskey and oak-matured red wines. J. Agric. Food
Chem. 2006, 54, 3380–3390.
(36) Dionex Corp. Application Note 123, The determination of
inorganic acids and organic acids in fermentation broths; 1998.
(37) Ribereau-Gayon, P. Phenolic compounds of grapes and wine. I.
Phenolic acids. Ann. Physiol. Veg. 1964, 6, 119–139.
(38) Salagoity-Auguste, M. H.; Bertrand, A. Wine phenolics - analysis
of low molecular weight components by high performance liquid
chromatography. J. Sci. Food Agric. 1984, 35, 1241–7.
(39) Monagas, M.; Suarez, R.; Gomez-Cordoves, C.; Bartolome, B.
Simultaneous determination of nonanthocyanin phenolic com-
pounds in red wines by HPLC-DAD/ESI-MS. Am. J. Enol. Vitic.
2005, 56, 139–147.
(43) Somers, T. C.; Verette, E.; Pocock, K. F. Hydroxycinnamate ester
of Vitis Vinifera: changes during vinifications, and effect of
exogenous enzymic hydrolysis. J. Sci. Food Agric. 1987, 40, 67–
78.
(44) Nickenig, R.; Pfeilsticker, K. Investigation of the oxidative
browning of white wine. HPLC separation of preparatively
obtained wine phenols. Dtsch. Lebensm. Rundsch. 1980, 76, 115–
119.
(45) Romani, A.; Mancini, P.; Tatti, S.; Vincieri, F. F. Polyphenols
and polysaccharides in Tuscan grapes and wines. Ital. J. Food
Sci. 1996, 8, 13–24.
(46) Trousdale, E. K.; Singleton, V. L. Astilbin and engelitin in grapes
and wine. Phytochemistry 1983, 22, 619–620.
(47) Dugo, G.; Saitta, M.; Salvo, F.; Vilasi, F.; Ragusa, M.; Manzo,
G.; Bambara, G. Characterization of Sicilian wines. Note 1.
Determination of phenolic compounds in Sicilian red wine by a
HRGC-MS/MS system. VigneVini 2001, 28, 81–85.
(48) Wang, H.; Race, E. J.; Shrikhande, A. J. Anthocyanin transforma-
tion in Cabernet Sauvignon wine during aging. J. Agric. Food
Chem. 2003, 51, 7989–7994.
(49) Alonso, E.; Revilla, E.; Estrella, I.; Diez, C. Qualitative analysis
of flavonol of two wines, one red and one white, of “Valdepenas”
vintage. Sci. Aliments 1986, 6, 361–372.
(50) Singleton, V. L.; Noble, A. C. Wine flavour and phenolic
substances. In Phenolic, Sulfur and Nitrogen Compounds in Food
FlaVours; Charalambous, G., Katz, A., Eds.; ACS Symposium
Series 26; American Chemical Society: Washington, DC, 1976;
pp 47–70.
(51) Gutierrez, I. H.; Sanchez-Palomo, L.; Eva, E.; Almudena, V.
Phenolic composition and magnitude of copigmentation in young
and shortly aged red wines made from the cultivars, Cabernet
Sauvignon, Cencibel, and Syrah. Food Chem. 2005, 92, 269–283.
(52) Baderschneider, B.; Winterhalter, P. Isolation and characterization
of novel benzoates cinnamates, flavanoids, and lignans from
Riesling wine and screening for antioxidant activity. J. Agric. Food
Chem. 2001, 49, 2788–2798.
(53) Nickenig, R.; Pfeilsticker, K. Investigation of the oxidative
browning of white wine. HPLC separation of preparatively
obtained wine phenols (in German). Dtsch. Lebensm.-Rundsch.
1980, 76, 115–119.
(54) Garcia-Parrilla, M. C.; Gonzalez, G. A.; Heredia, F. J.; Troncoso,
A. M. Differentiation of wine vinegars based on phenolic
composition. J. Agric. Food Chem. 1997, 45, 3487–3492.
(55) Dugo, G.; Saitta, M.; Salvo, F.; Vilasi, F.; Ragusa, M.; Manzo,
G.; Bambara, G. Characterization of Sicilian wines. Note 1.
Determination of phenolic compounds in Sicilian red wines by a
HRGC-MS/MS system. VigneVini 2001, 28, 81–85.
(56) Ducruet, V.; Flanzy, C.; Bourzeix, M.; Chambroy, Y. Volatiles
of young wines obtained by carbonic maceration. Sci. Aliments
1983, 3, 413–426.
(57) Schneider, R.; Baumes, R.; Bayonove, C.; Razungles, A. Volatile
compounds involved in the aroma of sweet fortified wines (Vins
Doux Naturels) from Grenache Noir. J. Agric. Food Chem. 1998,
46, 3230–3237.
Received for review October 15, 2007. Revised manuscript received
December 14, 2007. Accepted December 17, 2007.
(40) Lunte, S.; Blankenship, K. D.; Read, S. A. Detection and
identification of procyanidins and flavanols in wine by dual-
JF073031N