J. Agric. Food Chem. 2007, 55, 10591–10598
10591
Molar Absorptivities and Reducing Capacity of
Pyranoanthocyanins and Other Anthocyanins
MONICA JORDHEIM,† KJERSTI AABY,‡ TORGILS FOSSEN,† GRETE SKREDE,‡
ØYVIND M. ANDERSEN*,†
AND
Department of Chemistry, University of Bergen, Allégt. 41, N-5007, Bergen, Norway, and Matforsk
AS, Norwegian Food Research Institute, Osloveien 1, N-1430 Aas, Norway
To improve accuracy in the determination of anthocyanin purity and succeeding antioxidant capacity,
1
H and 13C nuclear magnetic resonance spectroscopy have been combined with high-performance
liquid chromatography (HPLC) equipped with a diode array detector and UV–vis spectroscopy in the
analysis of anthocyanidin 3-glycosides and 5-carboxypyranoanthocyanidin 3-glycosides. The molar
absoptivity (ε) values were found to be relatively similar, in contrast to previously reported literature
values, and the average ε values for both anthocyanidin 3-monoglycosides and 5-carboxypyranoanthocyanidin 3-glycosides were proposed to be 22 000 and 23 000 in acidified aqueous and methanolic
solutions, respectively. To assess the influence of structure on the potential antioxidant capacity of
anthocyanins, the 3-glucosides of pelargonidin (1), cyanidin (2), peonidin (3), delphinidin (4), petunidin
(5), malvidin (6), 5-carboxypyranopelargonidin (8), 5-carboxypyranocyanidin (9), 5-carboxypyranodelphinidin (11), 5-carboxypyranopetunidin (12), and 5-carboxypyranomalvidin (13) and the 3-galactosides of cyanidin (7) and 5-carboxypyranocyanidin (14) were examined by a ferric ion reducing
antioxidant power (FRAP) assay. The reducing capacities of the individual anthocyanins were in the
range of 0.9-5.2 µmol of Trolox equivalents/µmol. The two 5-carboxypyranoanthocyanins 11 and 9
and the four common anthocyanins 2, 4, 7, and 14, all possessing pyrogallol or catechol type of B
rings, showed the highest antioxidant capacity measured by FRAP. However, the inclusion of the
5-hydroxyl in the D ring and just one oxygen substituent on the B ring in 8 diminished the reducing
capacity considerably. Correspondingly, electrochemical behavior of 5-carboxypyranoanthocyanidin
3-glucosides and anthocyanidin 3-glucosides was derived using HPLC coupled to a coulometric array
detector set from 100 to 800 mV in increments of 100 mV. The relative order of the reducing capacity
of the various 5-carboxypyranoanthocyanidin 3-glucosides and anthocyanidin 3-glucosides were nearly
alike, whether determined by coulometric array detection or FRAP.
KEYWORDS: Anthocyanins; pyranoanthocyanins; purity; molar absorptivity; reducing capacity; FRAP;
coulometry; antioxidant capacity
INTRODUCTION
Increasing evidence that reactive oxygen species and oxidative
damage are involved in degenerative diseases has resulted in
increased interests in the efficacy of antioxidant activity of
naturally occurring molecules including various vitamins and
polyphenolic compounds in food and biological systems.
However, some recent human intervention trials involving
vitamins have presented negative results in this context, while
other studies have shown that consumption of antioxidant-rich
foods decreases levels of oxidative damage in ViVo in humans
(1). The antioxidants serve to keep the levels of free radicals
low, permitting them to perform useful biological functions
* To whom correspondence should be addressed. Telephone: +475558-3460. Fax: +47-5558-9490. E-mail: oyvind.andersen@kj.uib.no.
†
University of Bergen.
‡
Matforsk AS, Norwegian Food Research Institute.
without too much damage. A variety of methods for the
determination of antioxidant capacity of anthocyanin samples have
been described (2–7), and one will find many terms used by
different researchers to describe antioxidant capacity (8, 9). To
obtain comparable results, the measurements of individual anthocyanins have usually been compared to similar measurements of
ascorbic acid or Trolox, often expressed as Trolox equivalents (10).
After surveying the literature, it is obvious that these results are
strongly dependent upon the antioxidant assays applied (2–4, 10–23),
and most comparative studies conclude that each methodology
gives different responses for the same compounds or samples
(24–27). The relative antioxidant capacity order of various anthocyanins has even been altered just by changing the concentration
of the examined compounds (3).
Our opinion is that the relevance of much antioxidant
anthocyanin literature is limited because of inadequate consid-
10.1021/jf071417s CCC: $37.00 2007 American Chemical Society
Published on Web 11/30/2007
10592
J. Agric. Food Chem., Vol. 55, No. 26, 2007
Jordheim et al.
Table 1. Variations of Molar Absorptivity (ε) Values in the Literature (ref
35 and References therein, 42 and 43) Showing the Lowest and Highest
Reported ε Values for Pg3glc, Cy3glc, Cy3gal, Pn3glc, Dp3glc, Pt3glc,
and Mv3glca
pigment
ε values in
the literature
Pg3glc
Cy3glc
Cy3gal
Pn3glc
Dp3glc
Pt3glc
Mv3glc
lowest
highest
14 300
36 600
16 520
34 300
30 200
46 230
11 300
15 100
13 000
29 000
12 900
21 300
13 900
36 400
a
Table 2. Molar Absorptivity (ε) Values of Selected Anthocyanidin
3-Glycosides and 5-Carboxypyranoanthocyanidin 3-Glycosides in Acidified
Aqueous and Methanolic Solvents (A and B)a
Structures of individual anthocyanins are shown in Figure 1.
pelargonidin 3-glc (1)
cyanidin 3-glc (2)
petunidin 3-glc (5)
malvidin 3-glc (6)
cyanidin 3-gal (7)
5-carboxypyranopelargonidin 3-glc (8)
5-carboxypyranocyanidin 3-gal (9)
molar
λvis-max
solvent absorptivity (ε) (nm)
A
B
A
A
B
A
B
A
B
B
B
21 000
24 000
20 000
21 000
23 000
23 000
25 000
23 000
22 000
22 000
21 000
497
502
510
515
527
517
529
508
519
495
506
Mrb
546.40
546.40
484.83c
514.86c
592.43
528.88c
606.45
562.40
562.40
614.43
630.43
a
A ) aqueous buffer (pH 1.0) consisting of 0.2 M KCl/0.2 M HCl (25:67, v/v).
B ) concentrated HCl/MeOH (0.01:99.99, v/v). b Molecular mass including the
mass of chloride counterion. See Figure 1 for structures. c Molecular mass including
the mass of trifluoroacetate counterion. See Figure 1 for structures.
MATERIALS AND METHODS
Figure 1. Structures and nomenclature of anthocyanins (1–7) and
carboxypyranoanthocyanins (8–14). 1 ) pelargonidin 3-O-β-glucopyranoside (Pg3glc); 2 ) cyanidin 3-O-β-glucopyranoside (Cy3glc); 3 )
peonidin 3-O-β-glucopyranoside (Pn3glc); 4 ) delphinidin 3-O-β-glucopyranoside (Dp3glc); 5 ) petunidin 3-O-β-glucopyranoside (Pt3glc); 6
) malvidin 3-O-β-glucopyranoside (Mv3glc); 7 ) cyanidin 3-O-βgalactopyranoside (Cy3gal); 8 ) 5-carboxypyranopelargonidin 3-O-βglucopyranoside (pPg3glc); 9 ) 5-carboxypyranocyanidin 3-Oβ-glucopyranoside (pCy3glc); 10 ) 5-carboxypyranopeonidin 3-O-βglucopyranoside (pPn3glc); 11 ) 5-carboxypyranodelphinidin 3-O-βglucopyranoside (pDp3glc); 12 ) 5-carboxypyranopetunidin 3-Oβ-glucopyranoside (pPt3glc); 13 ) 5-carboxypyranomalvidin 3-O-βglucopyranoside (pMv3glc); and 14 ) 5-carboxypyranocyanidin 3-O-βgalactopyranoside (pCy3gal).
erations concerning the purity state of the examined anthocyanin
samples. The majority of antioxidant tests subjected to single
anthocyanins have addressed purity by the use of highperformance liquid chromatography (HPLC) equipped with a
diode array detector (DAD), liquid chromatography-mass
spectrometry (LC-MS), or both methods (28–30). When molar
absorptivity values have been used, one has to bear in mind
that existing literature presents substantial variation between the
molar absorptivity values given for the same anthocyanin, even
in the same solvent, as well as inconsistent values for structurally
similar anthocyanins (Table 1).
The first part of the present paper illustrates limitation by
the use of HPLC-DAD alone in the determination of anthocyanin purity. Thereafter, a new set of molar absorptivity values
for anthocyanidin 3-monoglycosides and pyranoanthocyanidin
3-monoglycosides is presented. The second part of the paper
describes the electrochemical behavior of a series of pure
pyranoanthocyanidin 3-monoglycosides and corresponding anthocyanin 3-monoglycosides with the aim of linking variation
in anthocyanin structure to reducing potentials obtained by both
a ferric ion reducing antioxidant power (FRAP) assay and
hydrodynamic voltammograms.
Anthocyanins. Pg3glc (1) was isolated from strawberries (Fragaria
x ananassa), while Cy3glc (2) and Pn3glc (3) were isolated from black
rice (Oryza satiVa) (Figure 1). Dp3glc (4), Pt3glc (5), and Mv3glc (6)
were isolated from black beans (Phaseolus Vulgaris), while Cy3gal (7)
came from black chokeberry (Aronia melanocarpa). The acidified
aqueous methanolic extracts containing the various anthocyanins were
concentrated under reduced pressure, purified by partition with ethyl
acetate and various type of column chromatography support, Amberlite
XAD-7, Sephadex LH-20, and Toyopearl HW-40F (31, 32). In some
cases, the same type of chromatography technique was applied twice
at different stages in the isolation procedure. The 5-carboxypyranoanthocyanins (8-14) (Figure 1) were hemisynthesized from their
corresponding anthocyanidin 3-glycosides (1-7) according to the
published procedure (32, 33). The purity of individual anthocyanins
was checked by integration of their HPLC-DAD chromatograms
recorded at 520 ( 20 and 280 ( 10 nm, respectively, and confirmed
by 1H and 13C nuclear magnetic resonance (NMR) spectroscopy. Solid
samples were achieved by evaporation to dryness under nitrogen
followed by freeze-drying to stable mass.
HPLC with UV–Vis Detection. The HPLC system used with
UV–vis detection (Agilent 1100 Series, Waldbronn, Germany) was
equipped with a HP 1050 DAD, a 20 µL loop, and a 200 × 4.6 mm
i.d., 5 µm ODS Hypersil column (Supelco, Bellefonte, PA). Two
solvents, A, water [0.5% trifluoroacetic acid (TFA)] and B, acetonitrile
(0.5% TFA) were used for elution. The elution profile consisted of
initial conditions with 90% A and 10% B, followed by gradient elution
for 10 min (14% B), isocratic elution for 10–14 min, and the subsequent
gradient conditions: 18 min (16% B), 22 min (18% B), 26 min (23%
B), 31 min (28% B), and 32 min (40% B), isocratic elution for 32–40
min, gradient elution for 40–43 min (10% B), and final linear elution
for 43–46 min (10% B). The flow rate was 1.0 mL/min, and aliquots
of 15 µL were injected with a Micro Autosampler (Agilent 1100 Series).
The UV–vis absorption spectra were recorded online during HPLC
analysis over the wavelength range of 240–600 nm in steps of 2
nm.
HPLC with Coulometric Array Detection. The HPLC system (HP
1050 series, Waldbronn, Germany) used with coulometric array
detection was interfaced to a coulometric array detector (ESA, Inc.,
Chelmsford, MA) with eight porous graphite working electrodes with
associated palladium reference electrodes (5). The array detector
was set from 100 to 800 mV in increments of 100 mV. The ESA
CoulArray operating software was used to collect voltammetric data.
Raw data were processed using Microsoft Excel, and the results
were presented as peak areas at the electrodes expressed as
microcoulombs per nanomole of antioxidant (µC/nmol) and as
cumulative peak areas (µC/nmol). The cumulative peak area was the
response across several electrodes; for example, the cumulative
response at 300 mV was the sum of the peak areas at 100, 200, and
Molar Absorptivities and Reducing Capacity of Anthocyanins
J. Agric. Food Chem., Vol. 55, No. 26, 2007
10593
Figure 2. HPLC chromatograms [detected at 520 ( 20 nm (A) and 280 ( 10 nm (B), respectively] of Mv3glc (6) after purification by XAD-7, Sephadex
LH-20, and Toyopearl HW-40F column chromatography. (C) 1H NMR spectra (600.13 MHz) of pure 6 (concentrated 11 mM). (D) 1H NMR spectra
(600.13 MHz) of the same sample, 6, as shown in the HPLC chromatograms (A and B). NMR samples dissolved in 5:95 CF3CO2D/CD3OD (v/v) were
recorded at 25 °C.
300 mV. Hydrodynamic voltammograms (HDVs) were achieved by
plotting the cumulative response versus electrode potential. Chromatographic separation was performed on a Betasil C18 column
(250 × 4.6 mm i.d., 5 µm particle size) equipped with a 5 µm C18
guard column (4.0 × 4.6 mm i.d.) both from Thermo HypersilKeystone (Bellefonte, PA). The mobile phase consisted of acetonitrile (A) and 0.2% phosphoric acid and 2% acetic acid (v/v) in
water (B). The program followed a linear gradient from 5 to 20%
A in 25 min and from 20 to 40% A in 5 min. The column
temperature was 40 °C; the solvent flow rate was 1 mL/min; and
the column was allowed to equilibrate for 10 min before each sample
(15 µL) was injected. The samples were filtered through a Millex
HA 0.45 µm filter (Millipore, Molsheim, France) before
injection.
NMR Spectroscopy. One-dimensional 1H and 1D 13C compensated
attached proton test experiment (CAPT), two-dimensional heteronuclear
single-quantum coherence (1H-13C HSQC), heteronuclear multiple-
bond correlation (1H-13C HMBC), double-quantum-filtered correlation
(1H-1H DQF-COSY), total correlated (1H-1H TOCSY), and nuclear
Overhauser effect (1H-1H NOESY) spectroscopy were obtained at
600.13 and 150.90 MHz for 1H and 13C, respectively, on a Bruker
Avance 600 instrument (Fällanden, Switzerland), equipped with a 600
MHz Ultrashield Plus magnet (Bruker Biospin AG) and a tripleresonance cryogenic probe (5 mm CPTCI 1H-13C/15N/D Z-gradient
coil). Sample temperatures were stabilized at 298 K. The deuteriomethyl
13
C signal and the residual 1H signal of the solvent (5:95 CF3CO2D/
CD3OD, v/v) were used as secondary references [δ 49.0 and 3.40 from
tetramethylsilane (TMS), respectively].
Determination of Molar Absorptivity (ε) Values. Weighed portions
of purified anthocyanins were dissolved in aqueous buffer (pH 1.0)
consisting of 0.2 M KCl/0.2 M HCl (25:67, v/v) (solvent A in Table
2) or concentrated HCl/MeOH (0.01:99.99, v/v) (solvent B in Table
2) to give accurate pigment concentrations around 1.0 × 10-4 M. The
molar absorptivity values, δ (L cm-1 mol-1), were calculated according
10594
J. Agric. Food Chem., Vol. 55, No. 26, 2007
Jordheim et al.
Figure 3. Online UV–vis spectrum and HPLC chromatogram (detected at 280 ( 10 nm) of pure Pg3glc (1) and 1D
13
C NMR spectrum of pure 1.
Figure 4. Antioxidant activity of anthocyanidin and pyranoanthocyanidin 3-glucosides (1–9 and 11–14) in the FRAP assay expressed as Trolox equivalents
(TE). For pigment identification, see Figure 1.
to Lambert–Beer’s law using the molecular mass, including the mass
of the counterion of the individual anthocyanins. The absorbance
measurements were made in triplicate.
FRAP Assay. The FRAP assay was performed according to the
procedure described by Benzie and Strain (34), with modifications. The
measurements were carried out on a FLUOstar OPTIMA plate reader
Molar Absorptivities and Reducing Capacity of Anthocyanins
Figure 5. (A) HDVs showing cumulative peak areas (µC/nmol of
anthocyanin) of pPg3glc (8) (9), pCy3glc (9) ([), pPn3glc (10) (2),
pDp3glc (11) (b), and pPt3glc (11) (0). (B) HDVs showing cumulative
peak areas (µC/nmol of anthocyanin) of Pg3glc (1) (9), Cy3glc (2) ([),
Pn3glc (3) (2), Dp3glc (4) (b), and Pt3glc (5) (0). For pigment
identification, see Figure 1.
(BMG Labtech GmbH, Offenburg, Germany) using the 595 nm
absorbance filter. Anthocyanin solution (250 µM, 10 µL) was added
manually to the plate and mixed with freshly prepared FRAP reagent
(190 µL) added by the plate reader. The reaction was conducted at 27
°C, and absorbance was measured every 2 min for 60 min. Aqueous
solutions of FeII (FeSO4 · 6H2O) in the concentration range of 125–1000
µmol/L were used for the calibration of the FRAP assay. FRAP values,
derived as average values from triplicate analysis of three solutions of
the anthocyanins, were expressed as micromoles of Trolox equivalents
per micromole of anthocyanin (µmol of TE/µmol).
RESULTS AND DISCUSSION
Determination of Anthocyanin Purity. The merit of antioxidant capacity values for individual anthocyanins depends upon
the precision in the determination of pigment purity or sample
concentration. Whether anthocyanins are purchased from commercial sources or isolated by research groups, the purity of
individual compounds has mainly been determined by HPLC-DAD,
LC-MS, or both methods (28–30). However, as shown in Figure
2, anthocyanin purity values obtained by HPLC-DAD have their
limitations. After purification by successive use of various types
of column chromatography (Amberlite XAD-7, Sephadex LH-20,
and Toyopearl HW-40F), the absence of additional peaks in the
HPLC chromatograms recorded at 520 ( 20, 280 ( 10 (parts A
and B of Figure 2), 320 ( 10, and 360 ( 10 nm indicated high
sample purity. However, when the 1D 1H NMR spectrum of the
same sample (Figure 2D) was compared to a similar spectrum of
pure Mv3glc (6) (Figure 2C), it was clear that 6 in the former
sample was not even the major aromatic compound. Pure Mv3glc
was achieved by subjecting the sample shown in Figure 2D a
second time through a Toyopearl HW-40F column.
Purity analyses based on HPLC-DAD chromatograms have
to reflect the following considerations: Chromatograms recorded
in the visible area (typical between 500 and 550 nm) fail to
detect aromatic compounds absorbing at shorter wavelengths.
J. Agric. Food Chem., Vol. 55, No. 26, 2007
10595
When additional HPLC chromatograms recorded in the UV–
visible region of the spectrum are included (typically around
280 nm), other aromatic compounds may be detected. Impurities
lacking a UV-absorbing chromophore will still be invisible.
These impurities might be perceived by the use of MS. However,
eventual water and inorganic salt content will normally not be
determined by these methods. Furthermore, compounds with
different chromatographic properties to those of anthocyanins
might not show up in the HPLC chromatograms, independently
of the HPLC detector, because of the strong interaction with
the stationary phase of the column. This latter case is most likely
the reason for the discrepancy between the HPLC-DAD and
1
H NMR results obtained for the anthocyanin sample examined
in Figure 2. The reliability of HPLC-DAD purity analysis of
anthocyanin may thus be improved considerably by including
complementary determination of purity by NMR. Purity control
was, in this paper, performed by the integration of 1H NMR
signals. When the presence of the baseline separated signals
other than those belonging to the examined compound and the
solvent peaks were accumulated below 15% of those of the
examined anthocyanin, the sample was defined as pure. A
standardized method using NMR integration for the determination of impurity amounts has limitations caused by crowded
spectral regions with significant signal overlap and baseline
effects, which may artificially enhance the signal integrals of
trace compounds. The purity control was further checked by
observation of signals in the 13C NMR spectrum.
Another routinely used approach employed to define or
measure purity/concentration of anthocyanin samples includes
the use of molar absorptivity (ε) values. Major difficulties here
with respect to exact mass determinations are reflected by the
huge variations among the reported ε values (Table 1). For
instance, the ε value of Mv3glc (6) dissolved in 0.1% HCl in
methanol has been reported separately to be both 13 900 and
29 500 L cm-1 mol-1 (reviewed by Giusti et al., 35). In addition
to substantial variation between ε values given for the same
anthocyanin, even in the same solvent, there exist inconsistent
differences between structurally very similar anthocyanins
(Table 1). The presence of other impurities than anthocyanins
(and other pigments) implies in most cases the calculation of
too low of ε values, which according to Lambert–Beer’s law
(A ) εcl) give too high of anthocyanin concentrations.
Consequently, impurities or too low ε values will imply that
the measured antioxidant capacities are presented to be lower
than reality. Additionally, some reported ε values and purity
determinations are hampered by the lack of anthocyanin
counterions in the calculations.
To improve the control procedure in estimations of anthocyanin purity, we have combined 1H (Figure 2C) and 13C NMR
spectroscopy with HPLC-DAD and UV–vis spectroscopy in
the purity analysis of various anthocyanins. As shown in the
13
C NMR spectrum of Pg3glc (1) (Figure 3), there exist no
significant signals but those belonging to the examined pigment
(1). Samples obtained using this NMR standardization have been
applied for determination of ε values of various anthocyanidin
3-glycosides and 5-carboxypyranoanthocyanidin 3-glycosides
(Table 1). Contrary to what has been proposed in the literature
(reviewed by Giusti et al., 35), the ε values of various
anthocyanidin 3-monoglycosides have been found to be relatively similar for all components investigated (Table 1). Even
the two analogous pyranoanthocyanins showed ε values in a
similar range of those recorded for anthocyanidin 3-monoglycosides (Table 1). On the basis of these measurements, we
suggest the average ε values for both anthocyanidin 3-monogly-
10596
J. Agric. Food Chem., Vol. 55, No. 26, 2007
Jordheim et al.
Figure 6. Cumulative peak area (µC/nmol) recorded at 400 mV derived from the HDVs of anthocyanidin and pyranoanthocyanidin 3-glucosides (1–5 and
8–12). For pigment identification, see Figure 1.
cosides and 5-carboxypyranoanthocyanidin 3-monoglycosides
to be 22 000 and 23 000 L cm-1 mol-1 in acidified aqueous
and methanolic solutions, respectively.
Reducing Capacity of Anthocyanins Determined by the
FRAP Assay. To assess the influence of structure on the
potential antioxidant capacity of anthocyanins, the 3-glucosides
of Pg (1), Cy (2), Pn (3), Dp (4), Pt (5), Mv (6), pPg (8), pCy
(9), pDp (11), pPt (12), and pMv (13) and the 3-galactosides of
Cy (7) and pCy (14) were examined by the FRAP method
(Figure 4). The concentration of each anthocyanin dissolved
in acidified methanolic solutions was first determined by
absorption spectroscopy using the obtained molar absorptivity
(ε) value of 23 000 L cm-1 mol-1 at the visible absorption
maxima for the 13 anthocyanins. The reducing capacities of
the individual anthocyanins were in the range of 0.9 to 5.2 µmol
of TE/µmol. The two 5-carboxypyranoanthocyanins 11 and 9
showed the highest potential antioxidant capacity measured by
FRAP subjected to anthocyanins (5.2 and 4.9 µmol of TE/µmol,
respectively). However, nearly similar values were obtained for
2, 4, 7, and 14 (Figure 4). These six pigments possess vicinal
trihydroxyl (pyrogallol type) or o-dihydroxyl (catechol type)
groups on their B rings. Compounds 8 and 1 gave the lowest
reducing capacities (0.9 and 2.8 µmol of TE/µmol, respectively)
among the examined anthocyanins. These pigments have only
one hydroxyl group on their B rings. The large difference
between the latter two values shows that the inclusion of the
5-hydroxyl in the D ring seem to have a significant negative
effect on the reducing capacity, when there is just one oxygen
substituent on the B ring.
When the reducing capacity of ferric ions is monitored, the
FRAP method has previously been used to analyze the potential
antioxidant capacity of anthocyanin-containing extracts and
mixtures (36–39). In a few cases, this method has been used
for the determination of the potential antioxidant capacity of
pure anthocyanins (5, 10). García-Alonso et al. (10) have
reported the reducing capacity of 4–6 and 11–13 to be between
1.6 and 2.5 expressed as ascorbic acid equivalents. The FRAP
values for ascorbic acid and Trolox (water-soluble R-tocopherol
analogue) are considered to be identical (40), and thus, the
corresponding reducing capacities of 4–6 and 11–13 (2.9–5.2
TE) shown in Figure 4 are considerably higher than those
reported by Garcia-Alonso et al. (10). Their antioxidant capacities were measured after a relatively short reaction time (6 min)
(10), when the reactions may have been incomplete (5, 40).
Furthermore, their purity determinations seem to be based on
HPLC chromatograms detected at 520 nm, which is only
adequate for the detection of compounds absorbing in the visible
spectral region. The major difference concerning the reducing
capacities shown in Figure 4 compared to literature values is
that the 5-carboxypyranoanthocyanins in previous literature (10)
were reported to have lower antioxidant capacities than the
corresponding common anthocyanins.
Reducing Capacity of Anthocyanins Determined by Coulometric Array Detection. The reducing capacity of the
3-glucosides of Pg(1), Cy (2), Pn (3), Dp (4), Pt (5), pPg (8),
pCy (9), pPn (10), pDp (11), and pPt (12) were derived from
coulometric analyses using HPLC coupled to a coulometric array
detector set from 100 to 800 mV in increments of 100 mV.
HDVs for each of the 10 anthocyanins were achieved by plotting
the cumulative responses versus electrode potential (Figure 5).
Flavonoids present several waves of oxidation across the
coulometric array, corresponding to several moieties capable
of undergoing oxidation (5, 41). According to Aaby et al. (5),
the cumulative responses at low to medium oxidation potentials
(300–500 mV) were most relevant for addressing the potential
antioxidant capacity of various phenolics. Hence, the relative
cumulative peak area at 400 mV was used as a measure for the
reducing capacity of the individual anthocyanins (Figure 6).
When the reducing capacity of the individual anthocyanins
was examined, the most pronounced effect was observed for
pPg3glc (8) (Figure 5A). This pigment remained without any
significant cumulative responses even at electrode potentials as
high as 600 mV. In full agreement with FRAP measurements
(Figure 4), the reducing capacity of this compound with a D
ring and only one hydroxyl group on the B ring was very low
compared to the other examined anthocyanins. In fact, the
relative order of the reducing capacity of the 5-carboxypyranoanthocyanidin 3-glucosides were alike whether determined
by coulometric array detection (Figure 6) or FRAP (Figure
4). With the exemption of a slightly decreased value for Dp3glc
Molar Absorptivities and Reducing Capacity of Anthocyanins
(4) measured by coulometric array detection, there was similar
agreement between the relative reducing capacity of the
examined anthocyanidin 3-glucosides.
ACKNOWLEDGMENT
We are grateful to Dr. Saleh Rayyan, Ms. Heidi Blokhus,
Mr. Arve Fossen, and Ms. Unni Hauge for contributions during
the preparation of samples.
LITERATURE CITED
(1) Halliwell, B. Polyphenols; antioxidant treats for healthy living or
covert toxins. J. Sci. Food Agric. 2006, 86, 1992–1995.
(2) Satué-Gracia, M. T.; Heinonen, M.; Frankel, E. Anthocyanins
as antioxidants on human low-density lipoprotein and lecithin–
liposome systems. J. Agric. Food Chem. 1997, 45, 3362–3367.
(3) Kähkönen, M. P.; Heinonen, M. Antioxidant activity of anthocyanins and their aglycons. J. Agric. Food Chem. 2003, 51, 628–
633.
(4) Rahman, M. M.; Ichiyanagi, T.; Komiyama, T.; Hatano, Y.;
Konishi, T. Superoxide radical- and peroxynitrite-scavenging
activity of anthocyanins; structure-activity relationship and their
synergism. Free Radical Res. 2006, 40, 993–1002.
(5) Aaby, K.; Hvattum, E.; Skrede, G. Analysis of flavonoids and
other phenolic compounds using high-performance liquid chromatography with coulometric array detection: Relationship to
antioxidant activity. J. Agric. Food Chem. 2004, 52, 4595–4603.
(6) Tsuda, T.; Horio, F.; Osawa, T. Dietary cyanidin 3-O-β-Dglucoside increases ex vivo oxidation resistance of serum in rats.
Lipids 1998, 33, 583–588.
(7) García-Alonso, M.; Rimbach, G.; Sasai, M.; Ankara, M.; Matsugo,
S.; Uchida, Y.; Rivas-Gonzalo, J. C.; de Pascual-Teresa, S.
Electron spin resonance spectroscopy studies on the free radical
scavenging activity of wine anthocyanins and pyranoanthocyanins.
Mol. Nutr. Food Res. 2005, 49, 1112–1119.
(8) Huang, D.; Ou, B.; Prior, R. L. The chemistry behind antioxidant
capacity assays. J. Agric. Food Chem. 2005, 53, 1841–1856.
(9) Prior, R. L.; Wu, X.; Schaich, K. Standardized methods for the
determination of antioxidant capacity and phenolics in foods and
dietary supplements. J. Agric. Food Chem. 2005, 53, 4290–4302.
(10) García-Alonso, M.; Rimbach, G.; Rivas-Gonzalo, J. C.; de PascualTeresa, S. Antioxidant and cellular activities of anthocyanins and
their corresponding vitisins A—Studies in platelets, monocytes,
and human endothelial cells. J. Agric. Food Chem. 2004, 52,
3378–3384.
(11) Tsuda, T.; Watanabe, M.; Ohshima, K.; Norinobu, S.; Choi, S.W.; Kawakishi, S.; Osawa, T. Antioxidative activity of the
anthocyanin pigments cyanidin 3-O-β-D-glucoside and cyanidin.
J. Agric. Food Chem. 1994, 42, 2407–2410.
(12) Miller, N. J.; Rice-Evans, C. The relative contributions of ascorbic
acid and phenolic antioxidants to the total antioxidant activity of
orange and apple fruit juices and blackcurrant drink. Food Chem.
1997, 60, 331–337.
(13) Wang, H.; Cao, G.; Prior, R. L. Oxygen radical absorbing capacity
of anthocyanins. J. Agric. Food Chem. 1997, 45, 304–309.
(14) Pool-Zobel, B. L.; Bub, A.; Schröder, N.; Rechkemmer, G.
Anthocyanins are potent antioxidants in model systems but do
not reduce endogenous oxidative DNA damage in human colon
cells. Eur. J. Nutr. 1999, 38, 227–234.
(15) Degenhardt, A.; Knapp, H.; Winterhalter, P. Separation and
purification of anthocyanins by high-speed countercurrent chromatography and screening for antioxidant activity. J. Agric. Food
Chem. 2000, 48, 338–343.
(16) Stintzing, F. C.; Stintzing, A. S.; Carle, R.; Frei, B.; Wrolstad,
R. E. Color and antioxidant properties of cyanidin-based anthocyanin pigments. J. Agric. Food Chem. 2002, 50, 6172–6181.
(17) Seeram, N. P.; Schutzki, R.; Chandra, A.; Nair, M. G. Characterization, quantification, and bioactivities of anthocyanins in
Cornus species. J. Agric. Food Chem. 2002, 50, 2519–2523.
J. Agric. Food Chem., Vol. 55, No. 26, 2007
10597
(18) Seeram, N. P.; Nair, M. G. Inhibition of lipid peroxidation and
structure-activity-related studies of the dietary constituents anthocyanins, anthocyanidins, and catechins. J. Agric. Food Chem.
2002, 50, 5308–5312.
(19) Kim, M.-Y.; Iwai, K.; Onodera, A.; Matsue, H. Identification and
antiradical properties of anthocyanins in fruits of Viburnum
dilatatum Thunb. J. Agric. Food Chem. 2003, 51, 6173–6177.
(20) Chun, O. K.; Kim, D.-O.; Lee, C. Y. Superoxide radical
scavenging activity of the major polyphenols in fresh plums. J.
Agric. Food Chem. 2003, 51, 8067–8072.
(21) Kim, D.-O.; Lee, C. Y. Comprehensive study on vitamin C
equivalent antioxidant capacity (VEAC) of various polyphenolics
in scavenging a free radical and its structural relationship. Crit.
ReV. Food Sci. Nutr. 2004, 44, 253–273.
(22) Lapornik, B.; Wondra, A. G.; Prosek, M. Comparison of TLC
and spectrophotometric methods for evaluation of the antioxidant
activity of grape and berry anthocyanins. J. Planar Chromatogr.—
Mod. TLC 2004, 17, 207–212.
(23) Awika, J. M.; Rooney, L. W.; Waniska, R. D. Properties of
3-deoxyanthocyanins from Sorghum. J. Agric. Food Chem. 2004,
52, 4388–4394.
(24) Arnao, M. B.; Cano, A.; Acosta, M. Methods to measure the
antioxidant activity in plant material. A comparative discussion.
Free Radical Res. 1999, 31, 89–96.
(25) Baderschneider, B.; Luthria, D.; Waterhouse, A. L.; Winterhalter,
P. Antioxidants in white wine (cv. Riesling). Part 1. Comparison
of different testing methods for antioxidant activity. Vitis 1999,
38, 127–131.
(26) Perez, D. D.; Leighton, F.; Aspee, A.; Aliaga, C.; Lissi, E. A
comparison of methods employed to evaluate antioxidant capabilities. Biol. Res. 2000, 33, 71–77.
(27) Schwarz, K.; Bertelsen, G.; Nissen, L. R.; Gardner, P. T.;
Heinonen, M. I.; Hopia, A.; Huynh-Ba, T.; Lambelet, P.; McPhail,
D.; Skibsted, L. H.; Tijburg, L. Investigation of plant extracts for
the protection of processed foods against lipid oxidation. Comparison of antioxidant assays based on radical scavenging, lipid
oxidation and analysis of the principal antioxidant compounds.
Eur. Food Res. Technol. 2001, 212, 319–328.
(28) Santos-Buelga, C.; Williamson, G. Methods in Polyphenol Analysis; Royal Society of Chemistry: Cambridge, U.K., 2003; pp 383.
(29) Andersen, Ø. M.; Francis, G. W. Techniques of pigment identification. In Plant Pigments and Their Manipulation; Davies, K.,
Ed.; Blackwell Publishing: London, U.K., 2004; pp 293–341.
(30) Giusti, M. M.; Wrolstad, R. E. Characterization and measurement
of anthocyanins by UV-visible spectroscopy. In Handbook of
Food Analytical Chemistry: Pigments, Colorants, FlaVors, Texture, and BioactiVe Food Components; Hoboken, N. J., Ed.; John
Wiley and Sons: New York, 2005; pp 19–31.
(31) Andersen, Ø. M.; Fossen, T.; Torskangerpoll, K.; Fossen, A.;
Hauge, U. Anthocyanin from strawberry (Fragaria ananassa) with
the novel aglycone, 5-carboxypyranopelargonidin. Phytochemistry
2004, 65, 405–410.
(32) Jordheim, M.; Fossen, T.; Andersen, Ø. M. Preparative isolation
and NMR characterization of carboxypyranoanthocyanins. J.
Agric. Food Chem. 2006, 54, 3572–3577.
(33) Fulcrand, H.; Benabdeljalil, C.; Rigaud, J.; Cheynier, V.; Moutounet, M. A new class of wine pigments generated by reaction
between pyruvic acid and grape anthocyanins. Phytochemistry
1998, 47, 1401–1407.
(34) Benzie, I. F. F.; Strain, J. J. The ferric reducing ability of plasma
(FRAP) as a measure of “antioxidant power”: The FRAP assay.
Anal. Biochem. 1996, 239, 70–76.
(35) Giusti, M. M.; Rodríguez-Saona, L. E.; Wrolstad, R. E. Molar
absorptivity and color characteristics of acylated and non-acylated
pelargonidin-based anthocyanins. J. Agric. Food Chem. 1999, 47,
4631–4637.
(36) Moyer, R. A.; Hummer, K. E.; Finn, C. E.; Frei, B.; Wrolstad,
R. E. Anthocyanins, phenolics, and antioxidant capacity in diverse
small fruits: Vaccinium, Rubus, and Ribes. J. Agric. Food Chem.
2002, 50, 519–525.
10598
J. Agric. Food Chem., Vol. 55, No. 26, 2007
(37) Nilsen, I. L. F.; Haren, G. R.; Magnussen, E. L.; Dragsted, L. O.;
Rasmussen, S. E. Quantification of anthocyanins in commercial
black currant juices by simple high-performance liquid chromatography. Investigation of their pH stability and antioxidant
potency. J. Agric. Food Chem. 2003, 51, 5861–5866.
(38) Vangdal, E.; Slimestad, R. Methods to determine antioxidative
capacity in fruit. J. Fruit Ornamental Plant Res. 2006, 14, 123–
131.
(39) Walton, M. C.; Lentle, R. G.; Reynolds, G. W.; Kruger, M. C.;
McGhie, T. K. Anthocyanin absorption and antioxidant status in
pigs. J. Agric. Food Chem. 2006, 54, 7940–7946.
(40) Huang, D.; Ou, B.; Prior, R. L. The chemistry behind antioxidant
capacity assays. J. Agric. Food Chem. 2005, 53, 1841–1856.
(41) Manach, C. The use of HPLC with coulometric array detection
in the analysis of flavonoids in complex matrixes. In Methods
in Polyphenol Analysis; Santos-Buelga, C., Williamson, G., Eds.;
Jordheim et al.
Royal Society of Chemistry: Cambridge, U.K., 2003; pp
63–91.
(42) Cabrita, L.; Fossen, T.; Andersen, Ø. M. Colour and stability of
the six common anthocyanidin 3-glucosides in aqueous solutions.
Food Chem. 2000, 68, 101–107.
(43) Torskangerpoll, K.; Andersen, Ø. M. Colour stability of anthocyanins in aqueous solutions at various pH values. Food Chem.
2005, 89, 427–440.
Received for review May 14, 2007. Revised manuscript received October
23, 2007. Accepted October 24, 2007. Monica Jordheim gratefully
acknowledges the Norwegian Research Council, NFR, for her fellowship. This work is part of project 157347/I20, which receives financial
support from NFR.
JF071417S