Quantitative studies and sensory analyses on the influence of cultivar, spatial tissue distribution, and industrial processing on the bitter off-taste of carrots (Daucus carota L.) and carrot products

Article abstract of DOI:10.1021/jf0496393

Although various reports pointed to 6-methoxymellein (1) as a key player imparting the bitter taste in carrots, activity-guided fractionation experiments recently gave evidence that not this isocoumarin but bisacetylenic oxylipins contribute mainly to the off-taste. Among these, (Z)-heptadeca-1,9-dien-4,6- diyn-3-ol (2), (Z)-3-acetoxy-heptadeca-1,9-dien-4,6-diyn-8-ol (3), and (Z)-heptadeca-1,9-dien-4,6-diyn-3,8-diol (falcarindiol, 4) have been successfully identified. In the present study, an analytical procedure was developed enabling an accurate quantitation of 1-4 in carrots and carrot products. To achieve this, (E)-heptadeca-1,9-dien-4,6-diyn-3,8-diol was synthesized as a suitable internal standard for the quantitative analysis of the bisacetylenes. On the basis of taste activity values, calculated as the ratio of the concentration and the human sensory threshold of a compound, a close relationship between the concentration of 4 and the intensity of the bitter off-taste in carrots, carrot puree, and carrot juice was demonstrated, thus showing that compound 4 might offer a new analytical measure for an objective evaluation of the quality of carrot products. Quantitative analysis on the intermediate products in industrial carrot processing revealed that removing the peel as well as green parts successfully decreased the concentrations in the final carrot puree by more than 50%.

Full text of DOI:10.1021/jf0496393

4508 J. Agric. Food Chem. 2004, 52, 4508−4514
Quantitative Studies and Sensory Analyses on the Influence of
Cultivar, Spatial Tissue Distribution, and Industrial Processing
on the Bitter Off-Taste of Carrots (Daucus carota L.) and Carrot
Products
ANDREAS CZEPA† AND THOMAS HOFMANN*,‡
Deutsche Forschungsanstalt f u¨ r Lebensmittelchemie, Lichtenbergstrasse 4,
D-85748 Garching, Germany, and Institut f u¨ r Lebensmittelchemie, Universit a¨ t M u¨ nster,
Corrensstrasse 45, D-48149 M u¨ nster, Germany
Although various reports pointed to 6-methoxymellein (1) as a key player imparting the bitter taste in
carrots, activity-guided fractionation experiments recently gave evidence that not this isocoumarin
but bisacetylenic oxylipins contribute mainly to the off-taste. Among these, (Z)-heptadeca-1,9-dien-
4,6-diyn-3-ol (2), (Z)-3-acetoxy-heptadeca-1,9-dien-4,6-diyn-8-ol (3), and (Z)-heptadeca-1,9-dien-4,6-
diyn-3,8-diol (falcarindiol, 4) have been successfully identified. In the present study, an analytical
procedure was developed enabling an accurate quantitation of 1-4 in carrots and carrot products.
To achieve this, (E)-heptadeca-1,9-dien-4,6-diyn-3,8-diol was synthesized as a suitable internal
standard for the quantitative analysis of the bisacetylenes. On the basis of taste activity values,
calculated as the ratio of the concentration and the human sensory threshold of a compound, a close
relationship between the concentration of 4 and the intensity of the bitter off-taste in carrots, carrot
puree, and carrot juice was demonstrated, thus showing that compound 4 might offer a new analytical
measure for an objective evaluation of the quality of carrot products. Quantitative analysis on the
intermediate products in industrial carrot processing revealed that removing the peel as well as green
parts successfully decreased the concentrations in the final carrot puree by more than 50%.
KEYWORDS: Bitter taste; carrot; 6-methoxymellein; (Z)-heptadeca-1,9-dien-4,6-diyn-3,8-diol; falcarindiol
INTRODUCTION
bitter off-taste of stored carrots and commercial carrot puree,
respectively (8). Among these bitter compounds, (Z)-heptadeca-
It has been known for more than half a century that carrots
are able to produce a sporadic bitter off-taste (1). This bitter
taste is often the reason for consumers rejection of carrot
products such as carrot puree in the infant diet and is, therefore,
a major problem for vegetable processors.
1
,9-dien-4,6-diyn-3-ol (2), (Z)-3-acetoxy-heptadeca-1,9-dien-4,6-
diyn-8-ol (3), and (Z)-heptadeca-1,9-dien-4,6-diyn-3,8-diol (4)
Figure 1), called falcarindiol, could be identified for the first
time as bitter compounds on the basis of mass spectrometry
MS) as well as one- and two-dimensional nuclear magnetic
(
(
Several compounds such as 6-methoxymellein (1) (Figure
resonance (NMR) experiments. Preliminary quantitative studies
and relating the concentration of these compounds with their
individual bitter detection thresholds clearly excluded 1 as the
inducer of the bitter off-taste and pinpointed compound 4 as
one of the key players in the bitter taste of carrots. The
quantitative procedure used for determining the amounts of these
taste compounds in carrots was, however, not convenient.
1
) (2, 3), eugenin (4), terpenoids (5), and water soluble phenolics
(6) have been proposed in the last 50 years as chemical
principles of off-taste development. The data found so far are
very contradictory, and for none of the compounds detected in
carrots could a correlation be found between the sensory
evaluation of the bitter taste and the results obtained by
instrumental analysis (3, 5, 7).
Very recent studies, activity-guided fractionation involving
solvent extraction, gel permeation chromatography, and high-
performance liquid chromatography (HPLC) in combination
with human taste dilution analyses revealed that not a single
compound but a multiplicity of bitter tastants contribute to the
The objectives of the present investigation were, therefore,
(i) to develop a suitable analytical method enabling the
simultaneous quantification of bisacetylenic oxylipins 2-4 and
1
in carrots and carrot puree, (ii) to rate these compounds in
their bitter impact on the basis of a dose-activity relationship,
iii) to investigate the spatial distribution of the bitter compounds
(
*
To whom correspondence should be addressed. Tel: +49-251-83-33-
in the carrot root, (iv) to study the influence of different carrot
cultivars on bitter taste, and (v) to investigate the influence of
industrial puree manufacturing on bitter taste compounds 1-4.
3
91. Fax: +49-251-83-33-396. E-mail: thomas.hofmann@uni-muenster.de.
†
Deutsche Forschungsanstalt f u¨ r Lebensmittelchemie.
‡
Universit a¨ t M u¨ nster.
1
0.1021/jf0496393 CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/05/2004

Bitter Off-Taste of Carrots
J. Agric. Food Chem., Vol. 52, No. 14, 2004 4509
was cooled in an ice bath, water (100 mL) and an aqueous saturated
NH
washed with diethyl ether (100 mL). The organic phases were
combined, were washed with an aqueous saturated NaHCO solution
100 mL) as well as a saturated NaCl solution (100 mL), and were
then dried over Na SO . After the solution was filtered and the solvent
4
Cl solution (150 mL) were added, and the aqueous layer was
3
(
2
4
was removed in vacuo, the title compound 6 (Figure 2) was isolated
1
with a purity of more than 99% by distillation (20 mbar, 60 °C). H
3
NMR (400 MHz, CDCl
3
): δ 2.17 (1H, s), 2.58 (1H, d, J ) 2.3 Hz),
2
3
2
4
)
1
.89 (1H, m), 5.25 (1H, dt, J ) 1.4, J ) 10.2 Hz), 5.50 (1H, dt, J
1.4, J ) 17.0 Hz), 5.98 (1H, ddd, J ) 5.2 Hz, J ) 10.2 Hz, J )
): δ 63.3 (CH), 74.8 (C), 83.0
3
3
3
3
7.0 Hz). 13C NMR (75 MHz, CDCl
3
2
(CH), 117.2 (CH ), 136.9 (CH).
(E)-4-Dodecen-1-in-3-ol (8). A solution of 7 (8 g; Figure 2) in
anhydrous diethyl ether (30 mL) was added dropwise to a mixture of
an ethinylmagnesiumbromide solution (100 mL; 0.5 M in tetrahydro-
furan) and anhydrous diethyl ether (40 mL) while stirring and was then
refluxed at 60 °C for 2 h. After the reaction mixture was worked up
following the procedure detailed above, 8 (Figure 2) was obtained upon
high vacuum distillation (0.5 mbar, 80 °C). HRGC/MS (CI, NH
3
): m/z
80 (100), 181 (14), 163 (9), 172 (4). H NMR (400 MHz, CDCl ): δ
.88 (3H, t, 6.8 Hz), 1.35 (10H, m), 1.90 (1H, s), 2.07 (2H, m), 2.56
1
1
0
3
4
3
(1H, d, J ) 2.0 Hz), 4.84 (1H, d, J ) 5.7 Hz), 5.61 (1H, m), 5.92
1H, m). 13C NMR (75 MHz, CDCl
( ): δ 14.5 (CH), 22.8 (CH ), 29.2
(CH ), 29.5 (CH ), 29.5 (CH ), 32.2 (CH ), 32.3 (CH ), 63.2 (CH),
4.3 (C), 83.8 (CH), 128.8 (CH), 135.0 (CH).
E)-Heptadeca-1,9-dien-4,6-diyn-3,8-diol (9). A mixture of 6 (2.0
3
2
2
2
2
2
2
7
Figure 1. Structures of bitter-tasting 1−4. Recognition threshold concentra-
(
tions (mg/L) for bitterness are given in brackets.
g; 24 mmol), 8 (4.4 g; 24 mmol), copper(I) chloride (1.0 g; 10 mmol),
and ammonium chloride (0.75 g; 14 mmol) in methanol (50 mL) was
stirred at room temperature in a laboratory autoclave (Roth, Germany)
under an atmosphere of oxygen (3 bar) for 18 h. After the suspension
was centrifuged and filtered, the reaction mixture was extracted with
diethyl ether (100 mL), and the water phase was extracted three times
with diethyl ether (100 mL). The combined organic phases were dried
MATERIALS AND METHODS
Chemicals. The following compounds were obtained commer-
cially: (E)-2-decenal (7) (Acros Organics, Schwerte, Germany),
ethinylmagnesiumbromide, 7-methoxycoumarin (Aldrich, Steinheim,
Germany), acrolein (5) (Fluka, Neu-Ulm, Germany), ammonium
chloride, copper(I) chloride, sodium chloride, sodium sulfate, and
sodium hydrogen carbonate (Merck, Darmstadt, Germany). The solvents
were of HPLC grade (Merck). Fresh carrots were obtained commercially
from a local vegetable market and analyzed the same day. Different
carrot cultivars were supplied by the companies Bejo (Warmenhuizen,
Netherlands), Rijk Zwaan (Welver, Germany), and Hipp (Pfaffenhofen,
Germany). The carrot puree exhibiting a significant bitter off-taste was
supplied by the German food industry. The reference materials of
2 4
over Na SO and filtered, and the solvent was removed in vacuo.
Fractionation of the residue by gel permeation chromatography on a
400 mm × 55 mm Sephadex LH-20 column (Amersham Pharmacia
Biotech, Freiburg, Germany) with ethanol as the mobile phase was
followed by semipreparative RP-HPLC to yield 9 (Figure 2) with a
purity of more than 99%. HRGC/MS (EI): m/z 55 (100), 91 (94), 115
(87), 129 (74), 41 (71), 77 (70), 128 (69), 105 (58), 43 (56), 79 (43).
+
HRGC/MS (CI, NH
3
): m/z 260 (100; [M] ), 154 (68), 225 (58), 172
3
-methyl-6-methoxy-8-hydroxy-3,4-dihydroisocoumarin (1) and 2-4
(50), 242 (43), 243 (31), 204 (26), 229 (23), 261 (22), 205 (19), 245
1
3
were isolated from carrots and purified as reported recently (8).
Syntheses. 1-Pentene-4-in-3-ol (6). A solution of freshly distilled 5
(19). H NMR (360 MHz, CDCl
3
): δ 0.88 (3H, t, J ) 6.8 Hz), 1.27
3
(10H, m), 1.39 (2H, m), 1.89 (1H, s), 2.06 (2H, q, J ) 6.1 Hz), 4.89
3
3
2
(3 g; Figure 2) in anhydrous diethyl ether (15 mL) was added dropwise
(1H, d, J ) 5.2 Hz), 4.95 (1H, d, J ) 5.2 Hz), 5.27 (1H, m, J )
3 2 3
to a mixture of an ethinylmagnesiumbromide solution (100 mL; 0.5 M
in tetrahydrofuran) and anhydrous diethyl ether (40 mL) while stirring.
Thereafter, the mixture was refluxed at 60 °C for 2 h. After the mixture
0.91 Hz, J ) 10.0 Hz), 5.48 (1H, m, J ) 0.91 Hz, J ) 17.0 Hz),
3 3 3
5.57 (1H, m, J ) 6.1 Hz, J ) 15.2 Hz), 5.91 (1H, m, J ) 5.2 Hz,
3
3
3
3
J ) 15.2 Hz), 5.95 (1H, ddd, J ) 5.2 Hz, J ) 10.0 Hz, J ) 17.0
Figure 2. Synthesis of the internal standard 9.

4510 J. Agric. Food Chem., Vol. 52, No. 14, 2004
Czepa and Hofmann
Hz). 13C NMR (75 MHz, CDCl
): δ 14.1 (CH), 22.6 (CH
), 28.8 (CH
electron impact mode (MS/EI) was performed at 70 eV and in the
chemical ionization mode (MS/CI) at 115 eV with ammonia as the
reactant gas.
3
2
2
),
2
29.1 (CH ), 29.1 (CH
2
), 31.8 (CH ), 32.0 (CH ), 63.4 (CH), 63.5 (CH),
2
2
6
9.7 (C), 70.3 (C), 78.3 (C), 79.3 (C), 117.4 (CH ), 127.7 (CH), 135.3
2
NMR Spectroscopy. 1H, 13C, DEPT-135 NMR spectroscopy,
HMQC, and HMBC experiments were performed on a AM-360
spectrometer (Bruker, Rheinstetten, Germany). Deuterochloroform was
used as the solvent, and tetramethylsilane was used as the internal
standard.
(
CH), 135.8 (CH).
Quantification of Bitter Compounds 1-4 in Carrots and Carrot
Puree. Commercial carrot puree (20 g) or fresh puree (20 g) obtained
upon mincing fresh carrots using an Ultra-Turrax while cooling,
respectively, was spiked with the internal standards (E)-falcarindiol (250
µg in 2 mL of MeOH) and 7-methoxycoumarin (50 µg in 2 mL of
MeOH) and then intimately mixed with Na
2
SO
4
(100 g). The carrot
RESULTS AND DISCUSSION
material was then extracted three times with ethyl acetate (100 mL)
by stirring for 5 min at room temperature. After the material was filtered,
the organic layers were combined and then freed from the solvent in
vacuo. The ethyl acetate extractables were dissolved in n-pentane (2
mL) and centrifuged (3000 rpm), and the clear supernatant was applied
onto the top of a Sep-Pak Classic Silica cartridge (Waters, Ireland)
conditioned with n-pentane. After the cartridge was flushed with
n-pentane/diethyl ether (5 mL; 95/5, v/v), polyacetylenic oxylipins as
well as 1 were eluted with a mixture (5 mL; 40/60, v/v) of n-pentane
and diethyl ether. The effluent was collected, the solvent was removed
in vacuo, and the residue obtained was dissolved in methanol (2 mL)
and membrane filtered. An aliquot (20 µL) of that stock solution was
analyzed by HPLC for quantification of 1 using 7-methoxycoumarin
as the internal standard. For the quantification of the polyacetylenic
compounds, an aliquot (1 mL) of the methanolic stock solution was
diluted with methanol (5 mL) and then analyzed by gas chromatography
To evaluate the contribution of the bitter compounds 1-4
(
Figure 1) to the bitter off-taste of carrots and carrot products
on the basis of a dose-activity relationship, an analytical tool
needed to be developed first enabling a rapid and accurate
quantitation of the bitter tastants in foods. To achieve this,
7-methoxycoumarin was chosen as a suitable internal standard
for the quantification of 1, but no commercially available
compound proved suitable as the internal standard for the
analysis of the acetylenic alcohols 2-4. Because the double
bonds in these bitter compounds were confirmed in carrots to
exist only in the (Z)-configuration, the corresponding 9 was
synthesized as a suitable internal standard. Following the
synthetic sequence given in Figure 2, 5 was converted in a
Grignard reaction with ethinylmagnesiumbromide to give 6. In
parallel, 7 was ethinylated to 8. The acetylenic alcohols 6 and
(GC). The results are given as the means of triplicates, and the standard
8
9
were then linked by a Glaser coupling reaction giving rise to
, which was purified by gel permeation chromatography and
RP HPLC.
deviation was less than 10%.
Sensory Analyses. Training of the Sensory Panel. Twelve assessors
were trained to evaluate the taste of aqueous solutions (3 mL each) of
the following standard taste compounds by using a triangle test as
described in the literature (9): saccharose (50 mmol/L) for sweet taste,
lactic acid (20 mmol/L) for sour taste, NaCl (12 mmol/L) for salty
taste, caffeine (1 mmol/L) for bitter taste, sodium glutamate (8 mmol/
L, pH 5.7) for umami taste, and tannin (gallustannic acid; 0.05%) for
astringency. The sensory analyses were performed in a sensory panel
room at 22-25 °C in three different sessions.
Using 7-methoxycoumarin and 9 as the internal standards,
the following analytical procedure was developed for a straight-
forward quantification of the bitter tastants 1-4 in carrots and
carrot products. Either the carrot tissue minced with an Ultra-
Turrax, the carrot puree, or the carrot juice was spiked with the
defined amounts of 7-methoxycoumarin and 9 and was then
intimately mixed with sodium sulfate until a dry powder was
obtained. After the carrot powder was extracted with ethyl
acetate and the sample was cleaned up by means of a silica
cartridge, 1 was quantified by RP HPLC/diode array detection
using 7-methoxycoumarin as the internal standard (Figure 3),
and the bisacetylenic oxylipins 2-4 were analyzed by HRGC
using 9 as the standard (Figure 4). This straightforward
procedure enabled the quantification of the bitter compounds
Intensity Ranking Test. Following the procedure reported recently
(9), the bitter taste of the commercial carrot puree, carrot juice, and
chopped, fresh carrots was evaluated by the trained sensory panel using
a scale from 0 (no bitter taste detectable) to 3 (strong bitterness).
HPLC. The HPLC apparatus (BIO-TEK Kontron Instruments,
Eching, Germany) consisted of two pumps (type 522), a Rheodyne
Injector (250 µL loop), and a UV/vis detector (type 535). The HPLC
analysis of 1 was performed on a 250 mm × 4.8 mm i.d. analytical
scale RP-18, ODS-Hypersil, 5 µm column (ThermoHypersil, Kleinos-
theim, Germany), monitoring the effluent at 302 nm and using the
following methanol/water gradient (flow rate 1.0 mL/min): starting
with a mixture (40/60, v/v) of methanol and water, the methanol content
was increased to 60% within 10 min and then increased to 80% in 15
min.
1
-4 in several samples in less than 4 h.
Contribution of Compounds 1-4 to the Bitter Taste of
Carrots. To elucidate the taste contribution of the compounds
1
-4 in fresh carrot tissue, the concentrations of these bitter
compounds were quantitatively determined in seven carrot
samples. In parallel, these samples were presented to a trained
sensory panel who was asked to score the intensity of the bitter
perception on a scale from 0 (not detectable) to 3 (strong
detectable). The results in Table 1 revealed the highest
concentrations for 4 spanning from 21.7 to 84.3 mg/kg, thus
confirming the high concentrations of acetylenic alcohol in
carrots as reported earlier (10-12).
High-Resolution (HR) GC. For HRGC analysis, a Trace GC
(Thermo Quest CE Instruments) coupled with either a flame ionization
detector or a mass spectrometer was used. Helium was used as the
carrier gas with a column pressure of 240 kPa, and nitrogen was used
as the makeup gas (30 mL/min). The sample was injected cool on
column. For HRGC analysis, a 60 m × 0.25 mm Neutra Bond-1,
WCOT fused silica column (GL Sciences Inc., Tokyo, Japan) was used.
The injection (1 µL) was performed at an oven temperature of 50 °C.
After 1 min, the temperature was increased with a rate of 10 °C/min
up to 250 °C and held for 15 min.
The acetylenes 2 and 3 were present in somewhat lower
amounts of 8.1-27.5 and 7.7-40.8 mg/kg, respectively. By far,
the lowest concentrations were found for 1, which was present
in concentrations below 2 mg/kg, with the exception of sample
containing 6.8 mg/kg of compound 1 (Table 1).
Because quantitative data alone do not allow any insight into
the taste contribution of a compound, compounds 1-4 were
rated in their taste impact on the basis of a dose-activity
relationship. To achieve this, taste activity values (TAVs) were
determined as the ratio of the concentration of a compound and
its threshold concentration. Calculations of the ratio of the
HRGC/MS. HRGC was performed with a Type 5890 Series II gas
chromatograph (Fisons Instruments, Mainz, Germany) using a 30 m
5
×
0.32 mm DB-5 fused silica capillary, 0.25 mm (J&W Scientific,
Fisons) by on-column injection at 40 °C. After 2 min, the temperature
of the oven was raised at 10 °C/min to 260 °C and held for 15 min
isothermally. The flow of the carrier gas, helium, was 1.8 mL/min.
MS analysis was performed with a MAT 95 S (Finnigan, Bremen,
Germany) in tandem with the HRGC. Mass chromatography in the

Bitter Off-Taste of Carrots
J. Agric. Food Chem., Vol. 52, No. 14, 2004 4511
Figure 3. HPLC chromatogram of the quantification of 1 with the internal standard 7-methoxycoumarin.
carrots, whereas, in contrast, no correlation could be observed
for the other taste compounds. Quantitative as well as sensory
analysis of an additional 28 commercial carrot batches further
strengthened this relationship. The data displayed in Figure 6
clearly show that the analytical data run in parallel to the sensory
data, thus confirming compound 4 as a reasonable indicator
substance for the analytical monitoring of bitterness in carrot
tissue.
Influence of the Carrot Cultivar and Soil on Bitter
Compounds. To investigate the influence of the carrot cultivar
on the concentration of bitter compounds, compounds 1-4 were
quantified in freshly harvested carrots from 11 different cultivars
grown in the same soil, and in addition, the intensity of the
carrot bitter taste was rated by a trained sensory panel on a
scale from 0 (no bitterness detectable) to 3 (strong bitterness
detectable). As given in Table 2, these cultivars differed
significantly in their bitter taste intensity being evaluated with
scores spanning from 0.1 to 1.7. The cultivars Bersky, Florida,
Fayette, Infinity, Kingston, and Kazan were evaluated with bitter
intensities of 0.1-0.3, whereas the bitterness of the cultivars
Bangor, Nandrin, Kamaran, Kathmandu, and Sunset was gener-
ally rated with higher bitterness scores; for example, the cultivar
Sunset showed the most pronounced bitter off-taste rated with
an intensity of 1.7 (Table 2). Quantitative analysis of the bitter
tastants revealed that the concentrations of compound 4 in these
carrots increased with increasing bitter taste intensity. The
cultivars evaluated with bitter intensities between 0.1 and 0.3
contained compound 4 in concentrations of 16.2-38.5 mg/kg,
whereas the carrots scored with bitter intensities of 0.9-1.7
contained the tastant in amounts of 45.0-84.3 mg/kg (Table
Figure 4. HRGC chromatogram of the quantification of the bisacetylenic
oxylipins 2−4 with the internal standard 9.
Table 1. Concentrations of 1−4 and the Bitter Intensity of Fresh
Carrots
concentration (mg/kg) of
sample
1
2
3
4
bitter intensitya
1
2
3
4
5
6
7
0.1
<0.1
0.6
<0.1
6.8
8.1
11.1
27.5
24.1
20.0
18.5
10.0
8.0
17.1
9.2
29.7
7.7
21.7
31.7
42.3
38.5
42.7
46.8
84.3
0.1
0.2
0.9
1.0
1.0
1.5
1.7
2.0
0.1
8.6
40.8
a
The bitter intensity was evaluated by the trained sensory panel using a scale
from 0 (no bitter taste detectable) to 3 (strong bitterness).
2
). Relating these concentrations to the threshold concentration
of 10 mg/kg revealed TAVs ranging from 1.6 to 8.4, thus
demonstrating that compound 4 contributes to the bitter off-
taste of all of the carrot cultivars investigated. In contrast, neither
for the isocoumarin derivative 1 nor for the oxylipins 2 and 3
could any correlation between the concentration of these
compounds and the data of the sensory analysis be found. On
the basis of these findings, it can be concluded that bitter off-
taste is strongly dependent on the carrot cultivar and that
compound 4 can be used as an analytical measure for off-taste
development.
concentration (Table 1) and the threshold concentrations
reported recently for 1 (20 mg/kg), 2 (20 mg/kg), 3 (60 mg/
kg), and 4 (10 mg/kg) (8) revealed the highest TAVs for
compound 4 in all of the samples investigated, e.g., the
concentration of compound 4 in sample 7 was 8.4 times above
its threshold concentration (Figure 5). In contrast, the concen-
tration of compound 2 was above its detection threshold only
in the samples 3-5 (Figure 5). In contrast, the TAVs calculated
for compounds 1 and 3 were below 1, thus indicting that these
compounds do not contribute to the bitter off-taste of the carrots
under investigation. Taking these analytical as well as the
sensory data into consideration, it is obvious that the TAV of
compound 4 increased parallel to the intensity of the bitterness
perceived. These data indicate a close relationship between the
concentration of 4 and the intensity of the bitter off-taste in
Aimed at getting first insight into the influence of the soil
on bitter tastant formation, the concentrations of bitter com-
pounds 1-4 were quantified in the less bitter carrot cultivar
Kazan and the bitter carrot cultivar Kamaran, respectively, each
grown on three different fields, and in addition, the intensity of

4512 J. Agric. Food Chem., Vol. 52, No. 14, 2004
Czepa and Hofmann
Figure 5. Bitter activity values of 1 and bisacetylenic oxylipins (2−4) and the bitter intensity of fresh carrot tissue.
Table 3. Influence of the Soil on the Concentrations of 1−4 and the
Bitter Intensity of Fresh Carrots
d
concentration (mg/kg) of
cultivar
Kazana
1
2
3
4
TAV of 4
bitter intensitye
0.3
0.1
0.6
0.1
2.0
7.8
20.8
7.9
27.5
12.0
18.5
7.0
16.9
15.3
9.2
8.8
8.6
15.5
33.0
38.5
46.4
46.8
114.3
1.6
3.3
3.9
4.6
4.7
0.2
0.2
0.3
1.2
1.3
2.0
b
Kazan
Kazanc
Kamarana
b
Kamaran
Kamaranc
32.4
11.4
a−c
Carrots were grown in the same field. d Concentrations are given as the
mean of triplicates. e Bitter intensity was rated on a score from 0 (not detectable)
to 3 (strongly detectable).
Figure 6. Correlation between the concentration of oxylipin 1 and the
bitter taste of 28 carrot samples.
Table 4. Distribution of 1−4 in the Carrot Root
Table 2. Influence of the Carrot Cultivar on the Concentrations of 1−4
concentration (mg/kg) of
and the Bitter Intensity of Fresh Carrots
tissuea
bitter intensitya
1
2
3
4
TAV of 4
b
upper end
lower end
phloem
0.2
0.2
0.3
24.0
20.4
28.6
24.7
15.2
8.0
11.5
13.3
33.5
18.7
32.2
14.9
3.4
1.8
3.2
1.5
0.5
<0.1
0.6
concentration (mg/kg) of
cultivara
1
2
3
4
TAV of 4
bitter intensityc
xylem
<0.1
0.2
Bersky
Florida
Fayette
Infinity
Kingston
Kazan
Bangor
Nandrin
Kamaran
Kathmandu
Sunset
<0.1
0.3
6.1
<0.1
0.8
0.6
<0.1
0.4
0.1
<0.1
0.1
5.2
30.9
10.5
11.1
15.5
27.5
16.0
23.6
12.0
6.1
8.9
7.0
11.0
17.1
16.2
9.2
18.7
12.1
8.8
28.3
40.8
16.2
16.0
24.9
31.7
32.4
38.5
45.0
43.5
46.4
45.7
84.3
1.6
1.6
2.5
3.2
3.2
3.9
4.5
4.4
4.6
4.6
8.4
0.1
0.1
0.2
0.2
0.3
0.3
0.9
0.9
1.2
1.3
1.7
the maximum, whereas the more intensely bitter cultivar
Kamaran reached a maximum value of 11.4 (Table 3).
Confirming the data in Table 2, there was no correlation
between the sensorially perceived bitterness and the quantitative
data of compounds 1-3. Taking all of these data into account,
it can be concluded that the carrot cultivar rather than the soil
conditions is more the key to bitter off-taste development.
Spatial Distribution of Bitter Compounds in the Carrot
Root. To investigate the spatial distribution of bitter compounds
in the carrot root, compounds 1-4 were quantified in the outer
phloem, the inner xylem, and the lower as well as the upper
end of the root, and in addition, the intensity of the carrot bitter
taste was rated by a trained sensory panel on a scale from 0
(no bitterness detectable) to 3 (strong bitterness detectable). The
sensory panel detected an intense bitter taste in the phloem (0.6)
as well as the upper end (0.5) of the root, whereas the bitterness
of the xylem as well as the lower end was comparatively low
(Table 4). Relating the sensory data to the quantitative analytical
data again demonstrated that the more bitter the sample, the
higher the concentrations of compound 4. The more bitter upper
end and the phloem contained 33.5 and 32.3 mg/kg of compound
4, whereas in the less bitter lower end and the xylem lower
10.0
a
Carrots were grown on the same field. b Concentrations are given as the mean
of triplicates. c Bitter intensity was rated on a score from 0 (not detectable) to 3
(
strongly detectable).
the carrot bitter taste was rated by a trained sensory panel on a
scale from 0 (no bitterness detectable) to 3 (strong bitterness
detectable). Independent from the soil, the cultivar Kazan
imparted only a weak bitter off-taste evaluated with intensities
below 0.3, whereas the samples of the cultivar Kamaran were
evaluated with significantly higher bitter scores spanning from
1
.2 to 2.0 (Table 3). Quantitative analysis of the bitter
compounds and calculation of TAVs revealed that the concen-
trations of compound 4 were again above the threshold
concentrations in all of the samples, but in the less bitter-tasting
cultivar Kazan, the taste acitivity values (TAVs) reached 3.9 at

Bitter Off-Taste of Carrots
J. Agric. Food Chem., Vol. 52, No. 14, 2004 4513
Figure 7. TAVs of 1 and bisacetylenic oxylipins (2−4) in intermediate products of industrial carrot puree manufacturing.
Table 5. Concentrations of 1−4 and the Bitter Intensity of Commercial
Carrot Puree and Carrot Juice
concentrations of 1.8 and 1.5 mg/kg were found. As already
found for the entire carrot tissue (e.g., Tables 1-3), no
correlation was found between the concentration of compounds
concentration (mg/kg) of
1
-3 and the bitter off-taste intensity. However, it is interesting
sample
1
2
3
4
TAV of 4
bitter intensitya
to notice that among the group of bisacetylenic oxylipins,
compounds 2 and 3 have a different spatial distribution than
the key bitter tastant 4; for example, the concentrations of
compounds 2 and 3 in the phloem equal those found in the
xylem, whereas the content of compound 4 in the phloem is
double the amount determined in the xylem (Table 4).
Carrot Puree
1
2
3
4
5
6
7
8
0.3
0.3
0.8
0.3
0.5
1.2
0.2
2.3
7.1
7.9
9.8
5.2
8.1
10.9
13.2
13.8
14.0
16.9
21.0
21.6
0.8
1.1
1.3
1.4
1.4
1.7
2.1
2.2
0.5
1.0
1.0
1.0
1.0
1.5
2.0
2.0
4.7
5.8
6.9
6.3
8.9
5.0
7.2
6.9
10.3
11.4
7.4
Contribution of Compounds 1-4 to Bitter Off-Taste of
Carrot Products. To investigate the contribution of the
compounds 1-4 to the bitter off-taste of carrot products, the
compounds were quantified in eight commercial carrot puree
batches as well as in six batches of carrot juice, and the bitter
taste intensity of these samples was evaluated by a trained
sensory panel on a scale from 0 (no bitterness detectable) to 3
7.9
Carrot Juice
1
2
3
4
5
6
2.1
2.2
0.9
3.8
2.7
1.6
5.5
6.2
4.9
7.5
12.7
7.2
3.5
6.6
8.1
11.4
14.3
17.6
20.6
0.7
0.8
1.1
1.4
1.8
2.1
0.3
0.3
0.5
0.5
0.5
1.0
3.9
3.8
5.7
5.9
6.9
(strong bitterness detectable). The quantitative data showed that
compound 4 was present in the highest concentrations spanning
from 8.1 to 21.6 mg/kg in carrot puree and from 6.6 to 20.6
mg/kg in carrot juice (Table 5). The polyacetylenes 2 and 3
were present in lower concentrations, and also, the concentration
differences between the individual samples were much smaller,
e.g., compounds 2 and 3 were present in concentrations from
a
The bitter intensity was evaluated by the trained sensory panel using a scale
from 0 (no bitter taste detectable) to 3 (strong bitterness).
pounds do not contribute to the bitter taste of carrot puree and
juice, respectively.
6
.9 to 11.4 and from 4.7 to 8.9 mg/kg. The lowest concentra-
Influence of Industrial Carrot Processing on Compounds
1-4. One of the major products for the infant food industry is
carrot puree. For the industrial manufacturing of carrot puree,
whole carrot roots are steam-peeled, and the peeled carrots are
then cut into cubic pieces (10 mm i.d.). Using an optoelectronic
device, green and dark parts are detected and sorted out by
means of an air pressure jet; the high-quality carrot pieces are
then minced and cooked, filled into jars, and finally heat-
sterilized.
To investigate the influence of the carrot puree manufacturing
process on the taste activity of individual bitter compounds,
compounds 1-4 were quantified in samples drawn from each
step of the industrial process, and TAVs were calculated as the
ratio of their concentration to their individual bitter threshold
concentrations (Figure 7). As expected, the analysis of the whole
carrot root showed a high TAV of 3.7 for compound 4, whereas
the concentrations of compounds 1-3 did not reach the taste
threshold. After steam peeling, the TAV of compound 4 dropped
from 3.7 to 1.4, whereas the low taste impact of compounds
tions, however, found for 1, e.g., 0.3-2.3 and 0.9-3.8 mg/kg,
respectively, were quantified in carrot puree and carrot juice.
Relating the concentrations of these compounds to their
individual bitter taste thresholds demonstrated that the TAV of
compound 4 was close to or above 1.0, thus indicating this
compound as a contributor to the bitter off-taste of the carrot
products. Well in line with this suggestion, the overall bitter
taste intensity of the individual puree and juice samples ran in
parallel with the TAVs of compound 4, e.g., in puree sample
8
, which was evaluated with an overall bitter intensity of 2.0,
a TAV of 2.2 was determined for compound 4 (Table 5). As
the sensory panel described all of the carrot puree samples
evaluated with a bitter intensity of g1.5 with the term
“
intolerably bitter”, compound 4 might be used as a suitable
indicator substance for the analytical monitoring of bitter off-
taste development in carrot products. In contrast, the concentra-
tions of compounds 1-3 were significantly below their taste
thresholds (data not shown), demonstrating that these com-

4514 J. Agric. Food Chem., Vol. 52, No. 14, 2004
Czepa and Hofmann
1
-3 was not strongly influenced. These data indicate that during
(2) Sondheimer, E. The isolation and identification of 3-methyl-6-
methoxy-8-hydroxy-3,4-dihydroisocoumarin from carrots. J. Am.
Chem. Soc. 1957, 79, 5036-5039.
the peeling process a major amount of compound 4 is selectively
removed with the peel and clearly confirm the data found for
the spatial distribution of bitter compounds in the carrot root
(
(
(
3) M u¨ ller, H. Studies on the formation of 3-methyl-6-methoxy-
3
,4-dihydroisocoumarin in carrots during storage (in German).
(Table 4). After the peeled carrot was cut into cubic pieces
Phytopathol. Z. 1978, 93, 241-248.
and sorting out the green and dark pieces, compound 4 was
present in the premium pieces with a TAV of 1.6, whereas in
the colored particles the concentration of the bitter compound
exceeded its sensory threshold by a factor of 3.6. These data
indicate that not only the peel but also the green and dark parts
of the carrot root contained the highest concentrations of
compound 4. After the samples were minced and cooked as
well as after the heat sterilization process of the puree, the taste
activity of compound 4 equaled the value found in the carrot
pieces, thus indicating that heat processing is not an influence
on taste activity of compound 4.
Taking all of these data into consideration, compound 4 could
be used to analytically objectify the bitterness of carrot products
and might offer a new standard for an objective evaluation of
the quality of carrot products. Analysis of various cultivars
demonstrated that some cultivars seem to produce higher
amounts of that bitter compound than others, whereas the soil
seems not to be a key driver for bitter taste development. In
the course of industrial carrot processing, removing the peel as
well as green and dark parts was demonstrated to successfully
remove more than 50% of the bitter tastant from the carrot
material. Such information might be helpful for carrot breeders
and producers as well as the carrot-processing industry to find
possibilities as to how to reduce the bitter off-taste in carrots
and carrot products in the future.
4) Coxon, D. T.; Curtis, R. F.; Price, K. R.; Levett, G. Abnormal
metabolites produced by Daucus carota roots stored under
conditions of stress. Phytochemistry 1973, 12, 1881-1885.
5) Seljasen, R.; Hoftun, H.; Bengtsson, G. B. Sensory quality of
ethylene-exposed carrots (Daucus carota L., cv. “Yukon”) related
to the contents of 6-methoxymellein, terpenes and sugars. J. Sci.
Food Agric. 2000, 81, 54-61.
(6) Talcott, S. T.; Howard, L. R.; Brenes, C. H. Factors contributing
to taste and quality of commercially processed strained carrots.
Food Res. Int. 2001, 34, 31-38.
(
(
(
7) Yoshino, N.; Kawaguchi, T.; Tokuoka, K.; Ishitani, T.; Hirata,
T. 6-Methoxymellein levels in fresh carrots in relation to the
sensory quality. J. Jpn. Soc. Food Sci. Technol. 1993, 40, 17-
2
1.
8) 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,
5
1, 3865-3873.
9) Soldo, T.; Blank, I.; Hofmann, T. (+)-(S)-Alapyridainesa general
taste enhancer? Chem. Senses 2003, 28, 371-379.
(
10) Wulf, L. W.; Nagel, C. W.; Branen, A. L. Analysis of myristicin
and falcarinol in carrots by high-pressure liquid chromatography.
J. Agric. Food Chem. 1978, 26, 1390-1393.
(11) Yates, S. G.; England, R. E. Isolation and analysis of carrot
constituents: myristicin, falcarinol and falcarindiol. J. Agric.
Food Chem. 1982, 30, 317-320.
12) Lund, D. L.; White, J. M. Polyacetylenes in normal and water-
stressed “Orlando-Gold” carrots (Daucus Carota). J. Sci. Food
Agric. 1990, 51, 507-516.
(
ACKNOWLEDGMENT
We thank Kristin Kahlenberg for her skillful assistance.
Received for review March 4, 2004. Revised manuscript received May
2, 2004. Accepted May 3, 2004. This research project was supported
by the FEI (Forschungskreis der Ern a1 hrungsindustrie e.V., Bonn), the
AiF, and the Ministry of Economics and Labor; AiF Project 12226 N.
LITERATURE CITED
(1) Yamaguchi, H.; Howard, F. D.; McNelly, L. B. Observations
on bitterness of California grown carrots. Plant Dis. Rep. 1955,
3
9, 302.
JF0496393