Available online at
Antioxidant tannins from Syzygium cumini fruit
Liang Liang Zhang and Yi Ming Lin*
Department of Biology, School of Life Sciences, Xiamen University, Xiamen 361005, China.
Accepted 13 February, 2009
Tannins are significant plant secondary metabolites
in vascular plants. Condensed tannins are also known as
proanthocyanidins (PAs), the oligomeric and polymeric
flavan-3-ols, which are linked through C4 - C8 or C4 - C6
linkages. The diversity of condensed tannins is given by
the structural variability of the monomer units (different
hydroxylation patterns of the aromatic rings A and B, and
different configurations at the chiral centers C2 and C3)
(Figure 1). The size of PAs molecules can be described
by their degree of polymerization (DP). The molecules
are water-soluble and are able to form complexes with
polyphenolic compounds that are widely distributed in the
plant kingdom. They are esters of a polyol (most often β-
D-glucose) with either gallic acid (gallotannins) or
hexahydroxydiphenic acid (ellagitannins). These ester
forms vary from simple compounds such as β-D-
glucogallin to compounds with M
values in excess of
meric tannins is difficult because of their heterogeneous
character. Due to this complexity and diversity, the cha-
*Corresponding author. E-mail: firstname.lastname@example.org. Tel.: +86
racterization of highly polymerized condensed tannins
regarding structure-activity relationships. Various techni-
ques including NMR and mass spectroscopy (MS) have
been used to characterize hydrolysable and condensed
tannins (Behrens et al., 2003; Chen and Hagerman,
2004; Vivas et al., 2004).
The fruits of Syzygium cumini (L.) Skeels are edible
and are reported to contain gallic acid, tannins,
anthocyanins and other components (Benherlal and
Arumughan, 2007). The juice of unripe fruits is used for
preparing vinegar that is considered to be a stomachic,
carminative and diuretic. The ripe fruits are used for
making preserves, squashes and jellies. The fruits are
astringent. A wine is prepared from the ripe fruits in Goa
(Wealth of India, 1976). Extract of seed, which is
traditionally used in diabetes, has a hypoglycaemic action
and antioxidant property in alloxan diabetic rats (Prince et
al., 1998) possibly due to tannins (Bhatia et al., 1971).
Tannins are antioxidants often characterized by reduc-
ing power (Mi-Yea et al., 2003) and scavenging activities
(Minussi et al., 2003). The antioxidant capabilities of
tannins depend on: (1) the extent of their colloidal state;
(2) the ease of interﬂavonoid bond cleavage or its
stereochemical structure; (3) the ease of the pyran ring
(C-ring) opening; and (4) the relative number of –OH
groups on A and B rings (Noferi et al., 1997). Compounds
with a trihydroxyl structure in the B-ring have the greatest
= OH, afzelechin
= OH, epiafzelechin
antioxidant activity (Rice-Evans et al., 1996).
The fruit of S. cumini has been shown to contain a
large range of hydrolysable tannins (Benherlal and
Arumughan, 2007). However, neither hydrolysable nor
condensed tannins have been characterized in S. cumini.
In this study we undertook the structural characterization
of tannins in S. cumini fruit using a combination of NMR,
MALDI-TOF MS and HPLC analyses. This would
contribute to a better understanding of the chemical
composition of S. cumini and the applicability of NMR,
MALDI-TOF MS and HPLC in the analyses of food
tannins. We also report the evaluation of free-radical
scavenging properties and ferric reducing/antioxidant
power of tannins extracted from S. cumini fruit.
Plant materials and chemicals
Mature fruits of S. cumini were collected at the campus of Xiamen
ground. The resulting powder was extracted with acetone:water
(7:3, v/v) and the organic solvent was eliminated by evaporation
under vacuum. The remaining crude tannin fraction was chromato-
graphed on an LH-20 column (Pharmacia Biotech, Uppsala,
Sweden) which was first eluted with methanol:water (50:50, v/v)
and then with acetone : water (7:3, v/v). The last fraction, containing
the polymeric tannins was freezed-dried and stored at -20°C till
1,1-Diphenyl-2-picrylhydrazyl (DPPH), 2,4,6-tripyridyl-S-triazine
(TPTZ), potassium ferricyanide, butylated hydroxyanisole (BHA),
ascorbic acid, (+)-catechin, cesium chloride, gallic acid, ellagic acid
and tannic acid were purchased from Sigma Chemical Co. (St.
Louis, MO, USA). Reagents and solvents were of analytical or
HPLC grade. Deionised water was used throughout.
Varian Metcury-600 spectrometer (USA) at 150 MHz.
The tannin extract was mixed with a 1 M solution of dihydroxy-
mixture was spotted onto a ground stainless steel MALDI target for
MALDI analysis using the dry droplet method. Cesium chloride (1
mg/ml) was mixed with the analyte/matrix solution at the 1:3
volumetric ratio to promote the formation of a single type of ion
adduct (M + Cs
) (Xiang et al., 2006). A Bruker Reflex III MALDI-
laser (337 nm) was used
ion reflectron TOF mode.
Acid hydrolysis of the ellagitannins was performed as described by
Oszmianski et al. (2007). The ellagitannins from S. cumini fruit
stone (25 mg) were hydrolyzed with 2 ml of 2 mol/l hydrochloric acid
in a boiling water bath for 1 h. After cooling, 2 ml of 2 mol/l NaOH
and then 6 ml methanol were added to the vial. The slurry was
sonicated for 20 min with occasional shaking. Further, the slurry
was centrifuged at 10,000 g and the supernatant was used for
HPLC analysis. The HPLC apparatus consisted of an Agilent 1100
diode array detector and a quaternary pump.
The samples were previously dissolved in a mobile phase and
then filtrated through a membrane filter with an aperture size of
0.45 µm. 10 ml of the clear supernatant was injected. Separation
was performed on a Hypersil ODS column (4.6 × 250 mm, 5 µm)
thermostatted at 30°C. The mobile phase was composed of solvent
A (0.1% v/v) trifluoroacetic acid (TFA) in water) and solvent B (0.1%
v/v) TFA in acetonitrile). The gradient condition was: 0~2
100% A, 2
min 5% B, 10
min 10% B, 20
min 10~20% B, 30
min 20% B and 35
min 20~30% B. Other chromatographic
conditions were as follows: flow rate at 1 ml/min, detection at 280
and 254 nm, and scanning performed between 200 and 600 nm.
The identification of chromatographic peaks was made by
comparison of their relative retention times with those of external
standards, as well as by their UV–visible spectra. Ellagic acid
(detection 254 nm) and gallic acid (detection at 280 nm) was
quantified using the calibration curve established with ellagic and
gallic acid standards. Analysis was made in triplicate.
Free-radical scavenging activity
The free-radical scavenging activity was measured according to
Braca et al. (2001). A 100 µl of the sample at different concen-
trations (15 – 500 µg/ml) was added to 3 ml of DPPH solution (0.1
M methanolic solution). 30 min later, the absorbance was measured
at 517 nm. Lower absorbance of the reaction mixture indicates
higher free radical scavenging activity. The IC
the amount of antioxidant necessary to decrease the initial DPPH
concentration by 50%, was calculated from the results and used for
comparison. The capability to scavenge the DPPH radical was
calculated using the following equation:
DPPH scavenging effect (%) = [(A
] × 100
is the absorbance of the control reaction and A
ascorbic acid were used as controls.
Ferric reducing/antioxidant power (FRAP) assay
FRAP assay is a simple and reliable colorimetric method commonly
used for measuring total antioxidant capacity (Benzie and Strain,
1996). Briefly, 3 ml of FRAP reagent, prepared freshly, was mixed
with 100 µl of the test sample, or methanol (for the reagent blank).
The FRAP reagent was prepared from 300 mM, pH 3.6, acetate
buffer, 20 mM ferric chloride and 10 mM 2,4,6-tripyridyl-S-triazine
made up in 40 mM hydrochloric acid. All three solutions were mixed
together in the ratio of 10:1:1 (v/v/v). The absorbance of reaction
mixture at 593 nm was measured spectrophotometrically after
incubation at 25°C for 10 min. The FRAP values, expressed in
mmol ascorbic acid equivalents (AAE)/g dried tannins, were derived
from a standard curve.
All measurements were replicated three times and one-way analy-
considered to be significant at P < 0.05. All statistical analyses were
performed with SPSS 11.0.
The signal assignment was made based on the publica-
distinct signals at 157 ppm, which are assignable to C4’
Indeed, procyanidin units (catechin/epicatechin) and
145 - 146 ppm respectively (Czochanska et al., 1980;
Behrens et al., 2003). The absence of a clear signal with
such chemical shift in the spectra of the condensed
tannins from S. cumini fruit skin revealed that they are
only composed of propelargonidin units.
The region between 70 and 90 ppm is sensitive to the
stereochemistry of the C-ring. The determination of the
ratio of the 2,3-cis to 2,3-trans stereochemistries could
thus be achieved through distinct differences in their
respective C2 chemical shifts (Czochanska et al., 1980).
Whereas C3 of both cis and trans isomers occurs at 73
ppm, C2 gives a resonance at 76 ppm for the cis form
and at 84 ppm for the trans form. The absence of the
latter signal peak in the spectrum of the studied con-
Zhang and Lin 2303
epiafzelechin subunits. The presence of a signal at 35.9
ppm was consistent with a C4 being shifted upfield by the
presence of a 3-O-gallate unit. This was further confirmed
by the observation of signals for ester carbonyl carbons
at 175.5 ppm (Gal-C7) and galloyl ring carbons at 114.0
ppm (Gal-C2, Gal-C6), 130.8 ppm (Gal-C1) and 143 -
145 ppm (Gal-C4). These results thus showed that the
polymeric propelargonidin of the studied S. cumini fruit
skin is predominantly constituted of propelargonidin with
(-)-epiafzelechin as the main constitutive monomer, some
with galloyl groups attached (Spencer et al., 2007).
As indicated above,
C NMR spectroscopy
techniques were used to estimate the degree of polyme-
rization and the number-average molecular weight. The
C3 in terminal units generally have their chemical shift
around 67 ppm. Theoretically, its intensity relative to that
of the signal of the C3 in extension monomer units at 73
ppm could be used for elucidating the polymer chain
length. However, in the case of the spectra presented
here, the signal-to-noise ratio is too low to allow for such
To obtain more detailed information on the chemical
problems with determination of polymer chain lengths by
NMR spectroscopy, further characterization was conti-
nued by means of MALDI-TOF MS.
MALDI-TOF MS has
sensitivity and mass range. The single ionization event
produced by MALDI-TOF MS allows the simultaneous
determination of masses in complex mixtures of low and
high molecular weight compounds. Several factors must
be optimized to develop MALDI-TOF MS techniques.
These factors include the selection of an appropriate
matrix, optimal mixing and optimal selection of cationi-
zation reagent. In our study, the Cs
was used as the
for their MALDI-TOF analysis and resulted in a relatively
simple MALDI-TOF spectrum.
Figure 2 shows the MALDI-TOF mass spectrum of the
adducts in the
repeating propelargonidin polymers. The polymeric
character is reflected by the periodic of peak series
representing different chain lengths. The results indicated
that S. cumini fruit skin tannins are characterized by
mass spectra with a series of peaks with distances of 272
Da corresponding to a mass difference of one afzelechin/
epiafzelechin between each polymer. Therefore, pro-
longation of condensed tannins is due to the addition of
afzelechin/epiafzelechin monomers. The spectrum show-
ed a series polyflavan-3-ols extending from the dimer
(m/z 679) to the undecamer (m/z 3129) that did not
contain ions with ∆2 amu lower than predicted in the
Masses represent the epiafzelechin homopolymer of the polyflavan-3-ol series [M+Cs]
Table 1. Observed and calculated masses
of heteropolyflavan-3-ols by
Mass calculations were based on the equation 274 + 272a + 152b + 133,
the degree of polymerization (DP) contributed by the epiafzelechin
extending unit, b is the number of galloyl esters and 133 is the atomic
weight of cesium; N, no observed peaks corresponding to those
positive-ion reflectron mode (Table 1).
al. (2003), an equation
was formulated to predict
masses are the molecular ions minus 1 proton plus Cs
weight of the terminal epiafzelechin unit, a is the degree
of polymerization (DP) contributed by the epiafzelechin
extending unit, b is the number of galloyl esters, 133 is
the atomic weight of cesium. Application of this equation
to the experimentally obtained data revealed the
presence of a series of condensed tannins consisting of
well-resolved oligomers. The broad peaks in these
spectra indicate, however, that there is large structural
heterogeneity within each DP.
For the condensed tannins indicated, each peak was
corresponding to the addition of one galloyl group at the
heterocyclic C-ring. Thus, peak signals corresponding to
monogalloylated derivatives of various condensed tannin
oligomers were easily attributed. No propelargonidin
containing more than one galloyl group were detected.
Therefore, MALDI-TOF MS indicates the simultaneous
occurrence of a mixture of propelargonidin polymers,
monogalloylated derivatives of propelargonidin polymers.
This showed that there were a mixture of galloylated
propelargonidin and propelargonidin in S. cumini fruit skin
No series of compounds that are ∆2 amu multiples
heteropolyflavan-3-ols were detected. So there are no A-
type interflavan ether linkages occurring between
adjacent flavan-3-ol subunits. All compounds are linked
In the case of S. cumini fruit stone tannins (Figure 3), a
TOF mass spectrum and notably these tannins possess
very similar mass distributions but different average
masses. In the spectra, four sets of peaks that are
separated by 152 Da are evident and have been assign-
ed to a unit of galloyl group using modeling correlations.
Structural assignment of these tannins advocates that S.
fruit stone is composed of gallic acid units centred
upon a core of glucose unit that is different from S. cumini
fruit skin. So, our study confirms the general classification
of the S. cumini fruit stone tannins as ellagitannins, that is
consisting of a glucose core surrounded by gallic and
ellagic acid units (Table 2). Almost superimposable mass
spectra were obtained for all the S. cumini fruit stone
ellagitannins analysed. Masses between 1500 and 5000
correspond to structures of oligomeric ellagitannins in
which two or more core glucose units are cross-linked by
dehydrodigalloyl or valoneoyl units. This is in agreement
with previously reported data concerning the same genus
plant Syzygium aromaticum (Tanaka et al., 1996) for
which two new ellagitannins were reported. These
different chemical groups are frequently composed of the
same building blocks but in different combinations and
numbers. For example, gallic acid occurs naturally but
can dimerize to form ellagic acid. Ellagic acid can
dimerize to form gallagic acid. Ellagic acid can combine
with glucose to form the unique compounds punicalagin
and punicalin. The different combinations and polymers
of the aforementioned form the large, diverse group of
compounds known as polyphenols, which show potent
antioxidant capacity and possible protective effects on
human health (Santos-Buelga and Scalbert, 2000).
These oligomers have been detected in S. cumini for the
first time in this study, although they are known to occur
in other plants (Quideau and Feldman, 1996) and further
study is required to elucidate their structure.
Mass + Cs
Identification of hydrolytic products of ellagitannins
The ellagic acid in plants is present mainly in the form of
transforms glucosylated and esteriﬁed ellagic acid into
their aglycones, and liberates the parent compound
ellagic acid and gallic acid (Daniel et al., 1989). Ellagic
and gallic acids are major products, as shown by a typical
HPLC chromatogram of the hydrolyzed products of
ellagitannins from S. cumini fruit stone (Figure 4) with
35.46 ± 3.00 mg/g dry tannins for ellagic acid and 19.73 ±
0.81 mg/g dry tannins for gallic acid.
Most quantitative evaluation of ellagitannins in fruit has
been on the hydrolyzed ellagitannins as ellagic acid
equivalents (Wada and Ou, 2002; Siriwoharn and
Wrolstad, 2004). This has significant problems in terms of
relating data to possible health effects, because there is
significant evidence that larger molecular mass tannins
＞1000 Da), including ellagitannins and procyanidins,
state (Cerda et al., 2004). Knowledge of ellagitannin
molecular structure, composition and quantity is needed
to understand their role in determining potential health
Figure 5 shows the dose-response curve of DPPH radical
nm (a) and 254 nm (b). The peaks corresponding to gallic acid (1) and ellagic acid (2) are indicated on the chromatogram.
Other peaks are unidentified phenolics.
Antioxidant concentration (mg/ml)
Figure 5. Free radical-scavenging activities of tannins,
measured using ascorbic acid, BHA and (+)-catechin as DPPH
assay reference compounds.
fruits, compared with (+)-catechin, BHA and
ascorbic acid. The tannins of the stone had a higher
activity than that of the skin. At a concentration of 0.25
mg/ml, the scavenging activity of tannins of the skin
reached 65.85%, while at the same concentration that of
the stone was 93.31%. IC
values were compared with
the antioxidant property of S. cumini fruit tannins (Table
3). A lower value of IC
indicates greater antioxidant
values of tannins from stone were superior
(DPPH) free radical-scavenging assay and the (FRAP) ferric-
reducing antioxidant power assay.
The antioxidant activity was evaluated as the concentration of the
test sample required to decrease the absorbance at 517 nm by
50% in comparison to the control.
FRAP values are expressed in
Different letters on the same column show significant differences
from each other at P < 0.05.
to those of the reference ascorbic acid, (+)-catechin and
BHA. The effect of antioxidants on DPPH is thought to be
due to their hydrogen donating ability (Baumann et al.,
1979). Though the DPPH radical scavenging abilities of
tannins from S. cumini fruit skin were less than that of the
stone, the study showed that the tannins have proton-
donating ability and could serve as free radical inhibitors
or scavengers, acting possibly as primary antioxidants.
The reducing ability of the tannin fractions from S. cumini
skin (3.02 mmol AAE/g) (Table 3). The antioxidant poten-
tial of the tannins from S. cumini fruit were estimated from
FRAP (mmol AAE/g)
165.05 ± 3.90a
3.02 ± 0.06d
82.21 ± 0.77c
6.21 ± 0.19b
106.36 ± 4.28b
4.34 ± 0.07c
113.00 ± 4.28b
85.68 ± 0.46c
catechin and BHA standards
their ability to reduce TPTZ-Fe
complex to TPTZ-Fe
The FRAP value of the skin tannins was significantly
values for the stone tannins on the other hand were signi-
ficantly lower than that of BHA but higher than that of (+)-
catechin. Such potential reducing power activity might be
attributed due to the presence of hydrolysable tannins
present in the stone. Antioxidant activity increased pro-
portionally with tannins content, and all tannins showed
increased ferric reducing power with increasing concen-
tration (Figure 6). According to
Oktay et al. (2003), a
antioxidant activity appears to be the trend in many plant
The results obtained showed that the condensed tannins
stereochemistry. The mean degree of polymerization
determined through MALDI-TOF MS analysis was 5.0
and the number-average molecular weight was 1372.45
Da. Cationization by addition of Cs
allowed us to elimi-
with ∆16 mass differences that
results from pattern of hydroxylation. As a result of these
techniques, we have observed larger structural hetero-
geneity of oligomers than is generally appreciated in the
literature on plant tannins. The results from the appli-
cation of MALDI-TOF MS clearly demonstrate its power
as a tool to characterize the nature of tannins. Tannins
extracted from S. cumini fruit showed a very good DPPH
radical scavenging activity and ferric reducing/antioxidant
Project (No. 30671646) was supported by the National
New Century Excellent Talents in University (NCET-07-
0725) and the Program for Innovative Research Team in
Science and Technology in Fujian Province University.
Baumann J, Wurn G, Bruchlausen FV (1979). Prostaglandin synthetase
O radical scavenging properties of some ﬂavonoids and
Pharmacol. 307: 1-77.
mass spectrometry and PSD fragmentation as means for the analysis
of condensed tannins in plant leaves and needles. Phytochem. 62:
Benherlal PS, Arumughan C (2007). Chemical composition and in vitro
antioxidant studies on
fruit. J. Sci. Food Agric. 87:
Benzie IFF, Strain JJ (1996). The ferric reducing ability of plasma
(FRAP) as a measure of “antioxidant power”: The FRAP assay. Anal.
Biochem. 239: 70-76.
Bhatia IS, Bajaj KL, Ghangas GS (1971). Tannins in black plum seeds.
Phytochem. 10: 219-220.
Braca A, Tommasi ND, Bari LD, Pizza C, Politi M, Morelli I (2001).
Antioxidant principles from
. J. Nat. Prod. 64:
Cerda B, Espin JC, Parra S, Martinez P, Tomas-Barberan FA (2004).
The potent in vitro antioxidant ellagitannins from pomegranate juice
are metabolised into bioavailable but poor antioxidant hydroxy-
dibenzopyran-6-one derivatives by the colonic microﬂora of healthy
Chen Y, Hagerman AE (2004). Characterization of soluble non-covalent
complexes between bovine serum albumin and
-glucopyranose by MALDI-TOF MS. J. Agric. Food Chem.
Czochanska Z, Foo LY, Newman RH, Porter LJ (1980). Polymeric
proanthocyanidins: Stereochemistry, structural units and molecular
weight. J. Chem. Soc. Perkin Trans. 1: 2278-2286.
Daniel EM, Krupnick A, Heur YH, Blinzler JA, Nims RW, Stoner GD
(1989). Extraction, stability and quantiﬁcation of ellagic acid in various
fruits and nuts. J. Food Comp. Anal. 2: 338-349.
Haslam E (1992). In: Plant Polyphenols; Hemingway, R.W., Laks, P.E.,
Eds.; Plenum Press: New York, p. 172.
Haslam E (1998). Practical Polyphenolics: From Structure to Molecular
Recognition and Physiological Action. Cambridge University Press:
Krueger CG, Vestling MM, Reed JD (2003). Matrix-assisted laser
sorghum. J. Agric. Food Chem. 51: 538-543.
N (2003). Phenolic compounds and total antioxidant potential of
commercial wines. Food Chem. 82: 409-416.
Mi-Yea S, Tae-Hun K, Nak-ju S (2003). Antioxidants and free radical
). Food Chem. 82: 593-597.
Noferi M, Masson E, Merlin A, Pizzi A, Deglise X (1997). Antioxidant
characteristics of hydrolysable and polyﬂavonoid tannins: An ESR
kinetics study. J. Appl. Polym. Sci
Oktay M, Gulcin I, Kufrevioglu QI (2003). Determination of in vitro
antioxidant activity of fennel (
) seed extracts.
LWT-Food Sci. Tech. 36: 263-271.
Oszmianski J, Wojdylo A, Lamer-Zarawska E, Swiader K (2007).
Antioxidant tannins from Rosaceae plant roots. Food Chem. 100:
Prince PS, Menon VP, Pari L (1998). Hypoglycaemic activity of
seeds: Effect on lipid peroxidation in alloxan
diabetic rats. J. Ethnopharm. 61: 1-7.
Quideau S, Feldman K (1996). Ellagitannin chemistry. Chem. Rev. 96:
Rice-Evans CA, Miller NJ, Paganga G (1996). Structure-antioxidant
activity relationships of ﬂavanoids and phenolic acids. Free Rad. Biol.
Med. 20: 933-956.
Santos-Buelga C, Scalbert A (2000). Proantocyanidins and tannin-like
compounds-nature, occurrence dietary intake and effects on nutrition
and health. J. Sci. Food Agric. 80: 1094-1117.
Siriwoharn T, Wrolstad RE (2004). Polyphenolic composition of marion
and evergreen blackberries. J. Agric. Food Chem. 69: 233-240.
Spencer P, Sivakumaran A, Fraser K, Foo LY, Lane GA, Edwards PJB,
Meagher LP (2007). Isolation and characterization of procyanidins
. Phytochem. Anal. 18: 193-203.
Tanaka T, Orii Y, Nonaka GI, Nishioka I, Kouno I (1996).
. Phytochem. 43:
Zhang and Lin 2309
(2004). Differentiationof proanthocyanidin tannins from seeds, skins
and stems of grapes (
) and heartwood of Quebracho
) by matrix-assisted laser desorption/ionization
chromatography/electrospray ionization mass spectrometry. Anal.
Chim. Acta. 513: 247-256.
Wada L, Ou B (2002). Antioxidant activity and phenolic content of
Oregon caneberries. J. Agric. Food Chem. 50: 3495-3500.
Wealth of India (1976). Raw Materials. New Delhi: CSIR, Vol X: 100-
Xiang P, Lin YM, Lin P, Xiang C (2006). Effects of adduct ions on
trometry of condensed tannins: A prerequisite knowledge. Chinese J.
Anal. Chem. 34: 1019-1022.