Food Chemistry 145 (2014) 395–403

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Fingerprinting of anthocyanins from grapes produced in Brazil using

HPLC–DAD–MS and exploratory analysis by principal component
analysis
Karina Fraige a, Edenir R. Pereira-Filho b, Emanuel Carrilho a,c,⇑
a
Instituto de Química de São Carlos, Universidade de São Paulo, 13560-970 São Carlos, SP, Brazil
b
Departamento de Química, Universidade Federal de São Carlos, 13565-905 São Carlos, SP, Brazil
c
Instituto Nacional de Ciência e Tecnologia de Bioanalítica, 13083-970 Campinas, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: The anthocyanin profile was studied for 11 types of grapes of different varieties and geographical origins,
Received 20 February 2013 including 10 wine grape varieties (Vitis vinifera) and one hybrid variety. Twenty anthocyanins were iden-
Received in revised form 31 July 2013 tified by means of the absorbance spectrum and fragmentation pattern by tandem mass spectrometry.
Accepted 16 August 2013
The multivariate method of principal component analysis (PCA) was used to evaluate differences
Available online 26 August 2013
amongst the cultivars. The results show the hybrid grape separated from a cluster represented by vinif-
eras grapes, mainly due to anthocyanin diglucosides. Within this group, it was possible to discriminate
Keywords:
the different wine grapes. Clusters discriminating according to geographical origin were not observed.
Grapes
Anthocyanins
Ó 2013 Elsevier Ltd. All rights reserved.
Chromatographic fingerprint
Principal component analysis

1. Introduction traces and more resistance to pathogens. It is a well-adapted culti-

Grapevine (Vitis spp.) is one of the most cultivated fruit plants
throughout the world. Grapes and products derived from them
constitute an economically important factor worldwide. In Brazil,
viticulture practice is recent when compared to traditional Euro-
pean grape producing countries, however, there has been observed
a marked improvement in the quality of Brazilian wines due to the
improvement of fine cultivars and winemaking techniques. A vari-
ety of species are cultivated throughout the world, but Vitis vinifera
is known for the best quality wines. About 20% of the grapes cul-
tured in Brazil come from V. vinifera species, and 80% from Vitis
labrusca, Vitis bourquiana, and hybrid varieties (Mello, 2007).
Amongst the V. vinifera species cultured in Brazil are Merlot, Cab-
ernet Sauvignon, and Syrah. The first two are more adapted to
the southern states whilst the latter is found in the northeastern
and southeast states. The most widespread are non-V. vinifera spe-
cies consumed as table grapes, for example Isabel, Moscato, Niag-
ara, and Itália. The hybrid grape IAC 138-22, popularly known as
Maximo, is a cross of Syrah (V. vinifera) and Seibel 11342 (hybrid),
which was part of a genetic improvement program by Empresa
Brasileira de Agropecuária (EMBRAPA) to produce a grape with fine
⇑ Corresponding author at: Instituto de Química de São Carlos, Universidade de
São Paulo, 13560-970 São Carlos, SP, Brazil. Tel.: +55 16 3373 9441.
E-mail addresses: emanuel@iqsc.usp.br, ecarrilho@me.com (E. Carrilho).
var to São Paulo state and is used to make wines (Santos Neto,
Pereira, Martins, & Leitão Filho, 1968).
Phenolic compounds have been extensively studied due to their
potentially beneficial antioxidant, anti-inflammatory and anti-car-
cinogenic properties, which are spurring the interest of both indus-
try and consumers for flavonoid-rich foods (Dugas et al., 2000;
Moure et al., 2001; Ng, Liu, & Wang, 2000; Pérez-Gregorio, Regue-
iro, Simal-Gándara, Rodrigues, & Almeida, 2013). Many foods or
products that present functional activities are rich in flavonoids
content, such as anthocyanidins and flavonols, as quercetin, con-
tributing to the overall intake of these components that present
antioxidant activity (Pérez-Gregorio et al., 2013; Yang, Meyers,
Van Der Heide, & Rui, 2004). The effect of the degradation of flavo-
nols and anthocyanins on processing foods and beverages is largely
studied. For example, it has been reported losses of flavonoids
upon boiling of vegetables, as a result of its migration into the
cooking water. However, quercetin conjugates showed to be
remarkably resistant to degradation during normal processing
operations, pointing out that it is an important source of flavonols
even after cooking if the broth were consumed too (Rodrigues,
Pérez-Gregorio, García-Falcón, & Simal-Gándara, 2009).
Grapes are a rich source of phenolic compounds, which play an
important role in oenology due to their influence on some impor-
tant sensory properties of wines, such as colour, stability, bitter-
ness, and astringency. Anthocyanins constitute a large family of

0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodchem.2013.08.066
396 K. Fraige et al. / Food Chemistry 145 (2014) 395–403
polyphenols in plants and are widely distributed in nature. They
are responsible for some fruit and flower colours, including red
grapes, in which they are located in the skins and appear mainly
during ripening (Esteban, Villanueva, & Lissarrague, 2001; Mazza
& Miniati, 1993; Ryan & Revilla, 2003). They play a very important
role in plant metabolism and are important in the food industry.
The main anthocyanins found in grapes are derived from cyanidin,
peonidin, delphinidin, petunidin, and malvidin, and they generally
occur as glycosides and acylglycosides; malvidin-3-glucoside is the
most abundant in almost all grape varieties (Pomar, Novo, & Masa,
2005; Ryan & Revilla, 2003; Vian, Tomao, Coulomb, Lacombe, &
Dangles, 2006). Some studies suggest that anthocyanins present
higher antioxidant activity than vitamins C or E, being reported a
linear relationship between antioxidant capacity and anthocyanin
content in some fruits (Castañeda-Ovando, Pacheco-Hernandez,
Paez-Hernandez, Rodriguez, & Galan-Vidal, 2009; Heinonen,
Meyer, & Frankel, 1998). The antioxidant activity of the Vitis
coignetiae Pulliat anthocyanins was determined via 2,2-diphenyl-
2-picrylhydrazyl radical scavenging and 2,20-azinobis-(3-ethyl-
benzothiazoline-6-sulphonic acid) radical cation assays and the
results showed that anthocyanins are primarily responsible for
the antioxidant activity of this grape variety, as well as previously
reported in other grapes varieties. Moreover, the antioxidant activ-
ity of V. coignetiae Pulliat anthocyanins was higher than that of
ascorbic acid (Choi et al., 2010).
Quantitative analysis and identification of individual anthocya-
nins has been widely studied using high-performance liquid chro-
matography (HPLC), generally coupled to diode array detection as
the main analytical tool (Hong & Wrolstad, 1990; Ryan & Revilla,
2003; Vian et al., 2006; Wang, Tong, Chen, & Gangemi, 2010). Vian
et al. (2006) identified nine anthocyanins in the extracts of skin
from Syrah grapes by HPLC with diode array detection, including
3-monoglucosides of delphinidin, cyanidin, petunidin, peonidin,
and malvidin, and the acetylglucosides and p-coumarylglucosides
of malvidin and peonidin. These authors observed that the content
of anthocyanins in conventionally grown grapes was significantly
higher than in those from organic production. In the last few years,
HPLC coupled to mass spectrometry (MS) had become a very effi-
cient tool providing an increased sensitivity and structural infor-
mation compared to HPLC–DAD (Castañeda-Ovando et al., 2009;
de Rosso et al., 2012; de Villiers, Vanhoenacker, Majek, & Sandra,
2004; Kammerer, Claus, Carle, & Schieber, 2004; Mazzuca, Ferranti,
Picariello, Chianese, & Addeo, 2005; Wang, Race, & Shirkande,
2003; Wu & Prior, 2005), making it possible to elucidate the struc-
ture and identify the anthocyanins by the aglycone moiety, type
and number of sugars by tandem mass spectrometry. The MS
instruments used can vary from simple quadrupole to advanced
tandem and high-resolution systems, and when reference stan-
dards for anthocyanins are not available, compound identification
is often tentative and is usually based on elution order, UV–Vis
spectra and mass spectral information (molecular weight and
MS/MS fragmentation) (Abad-García, Berrueta, Garmón-Lobato,
Gallo, & Vicente, 2009). Mazzuca et al. (2005) used a mass spec-
trometry procedure by LC–ESI-MS/MS to differentiate V. vinifera
from hybrid grape cultivars from Italy based on the anthocyanin
profile. The results provided evidence of the incorrect classification
of a grape cultivar previously believed to belong to the V. vinifera
species and the correct classification of the Pallagrello cultivar as
V. vinifera. Application of MALDI-TOF/MS for analyses of anthocya-
nins in food samples has already been done (Carpentieri, Marino, &
Amoresano, 2007; Grant & Helleur, 2008; Ivanova et al., 2011). Re-
cently, four different MALDI matrices were tested for anthocyanin
analysis in wine and grape samples; it was found that 2,5-dihy-
droxybenzoic acid (2,5-DHB) provided the best results, allowing
the identification of the dominant anthocyanins present in these
samples (Ivanova et al., 2011). Alberts, Stander, and de Villiers
(2012) reported a high-efficiency ultra high pressure liquid chro-
matography (UPLC) procedure with tandem mass spectrometry
using a triple quadrupole instrument for screening of red wine
anthocyanins and derived pigments. They used MS/MS in neutral
loss scanning mode to observe the loss of dehydrated sugar moie-
ties to tentatively identify the compounds based on their molar
mass and elution order, and in a second experiment the character-
istic fragmentation patterns were used to identify the anthocyani-
din. Using this method they observed significant chemical
alteration of the anthocyanins profile during wine ageing identify-
ing 121 different compounds belonging to anthocyanins, pyrano-
anthocyanins, and flavanol–anthocyanin condensation products
in Pinotage wines (Alberts et al., 2012).
Anthocyanins have been postulated as chemical markers to dif-
ferentiate grape cultivars and the red wines made from, providing
valuable information about the adulteration of juices and wines
(Arozarena, Casp, Marin, & Navarro, 2000; Berente, De la Calle Gar-
cia, Reichenbächer, & Danzer, 2000; García-Beneytez, Revilla, &
Cabello, 2002). Statistical multivariate methods can be used to
facilitate data interpretation when a large number of variables
are analyzed. The multivariate method of principal component
analysis (PCA) has been applied to reduce the number of dimen-
sions of the original data system to visualize whether different
groups of 23 Spanish wines could be related to specified anthocy-
anins or not. The authors found that the anthocyanin with the
greatest discriminant power in the differentiation amongst varie-
ties was malvidin-3-acetylglucoside, in addition to the relative
contents of p-coumaric acid esters of malvidin 3-glucoside (Aroza-
rena et al., 2000). A study on the composition of Tempranillo,
Garnacha, and Cabernet Sauvignon grapes from high and low qual-
ity vineyards using different statistical multivariate methods
showed that the anthocyanin profile was primarily determined
by variety. The differences between vineyards, however, were very
small (Arozarena et al., 2002). The anthocyanin and flavonol pro-
files of three varieties of grapes from Galicia, a Spanish wine-pro-
ducing region, were investigated. The polyphenolic profile of
Brancellao, Gran Negro and Mouratón red grapes, together with
statistical analysis by cluster analysis and PCA allowed differentia-
tion of the varieties. Gran Negro could be mainly characterized by
peonidin and malvidin conjugates and syringetin-3-O-glucoside
and isorhamnetin-3-O-glucoside, Mouratón by their content on
petunidin, delphidin and myricetin derivatives and Brancellao by
the cyanidin content. Moreover the results showed that the
biosynthesis of flavonols and anthocyanins are closely related
(Figueiredo-González et al., 2012).
In the present study, the anthocyanin composition of Syrah,
Cabernet Sauvignon, and Merlot grapes, all V. vinifera species from
different locations of Brazil, and Maximo, a hybrid grape studied
for the first time in relation to individual anthocyanin composition
and used to produce wines in some regions of São Paulo state, was
identified by HPLC coupled with a diode array detector and ion trap
mass spectrometer (LC–DAD–MS). The relative amount of the iden-
tified anthocyanins was studied by PCA to visualize if the grape
varieties or producer regions could be related to specific
anthocyanins.
2. Materials and methods
2.1. Chemicals
HPLC grade acetonitrile, methanol, and formic acid were
purchased from J.T. Baker (Phillipsburg, USA). Hydrochloric acid
was purchased from Qhemis (São Paulo, Brazil), and sodium
metabisulfite from Sigma (Saint Louis, USA). Water was acquired
from a Milli-Q purification system (Millipore Corporation, Bedford,
USA).
 K. Fraige et al. / Food Chemistry 145 (2014) 395–403 397

2.2. Samples
1400
Bunches of Cabernet Sauvignon and Syrah grapes were provided
from different regions in Brazil and collected at harvest in the years 1200

Total Anthocyanins (mg/L)

2008, 2009, and 2010. Merlot and Maximo grapes were collected
during the harvests of 2008 and 2010, respectively (Table 1). The 1000
berries were randomly picked from both sides of each bunch to ob-
tain a representative sampling. To perform the analysis of total 800
anthocyanin content, the grapes were crushed and centrifuged at
3000 rpm for 10 min. The supernatants were used and analyses 600
were performed by the sodium bisulfite discolouration method
(Ribéreau-Gayon, Glories, Maujean, & Dubourdieu, 2006) that is 400
based on the difference in absorbance of the anthocyanins when
reacted with an excess of SO2. Measurements were carried out in 200
a UV–Vis Jasco V-630 Spectrophotometer (Tokyo, Japan) at 520 nm.
To perform the analysis of the anthocyanin fingerprint by HPLC, 0
the berries were ground in liquid nitrogen using a mortar and pes-

CS_SC_10

M_BG_08
CS_SC_08
S_CN_08
S_SC_10

CS_BG_08
S_SC_09

S_CN_10
S_SC_08

tle. Aliquots of 2 g of pulverized grapes were extracted with 50 mL
methanol/0.1% HCl (v/v) for 2 h under agitation. The extracts were
filtered, and the material was re-extracted with 50 mL of the or-
ganic solvent for 1 h under agitation. The combined supernatants
were evaporated to dryness in a vacuum at 35 °C, and the residues
were dissolved in 12 mL of acidified water (pH 3.0). Before analysis
by HPLC, the extracts were diluted in the mobile phase (1:1 v/v),
centrifuged at 8000 rpm for 1.5 min and injected into the chro-
matographic system.
2.3. HPLC analysis
HPLC analyses were carried out using a Finnigan Surveyor Plus
HPLC system (Thermo Finnigan, San Jose, USA) equipped with a
quaternary gradient pump, an autosampler, a column oven, and a
PDA detector. Separation was performed in a Zorbax Eclipse XBD
C18 column (Agilent Technologies, Palo Alto, USA) (250  4.6 mm
i.d.; 5 lm) at 25 °C. The diode array detector was set to an acquisi-
tion range from 200 to 600 nm and the monitoring was made in
520 nm. The mobile phase consisted of water/formic acid/acetoni-
trile (87:10:3, v/v/v; solvent A) and water/formic acid/acetonitrile
(40:10:50, v/v/v; solvent B) using a gradient program as follows:
from 10% to 25% B in 10 min, from 25% to 31% B in 5 min, from
31% to 40% B in 5 min, from 40% to 50% B in 10 min, from 50% to
100% B in 10 min, and from 100% to 10% B in 5 min at a flow rate
of 0.8 mL min 1. The total run time was 50 min. The injection vol-
ume for all samples was 10 lL.
As standards were not available, the percentage area of each
compound in the samples was carried out considering the total
and individual area of each peak obtained from the data acquired
by the PDA detector.
Mx_L_10
S_L_10
Fig. 1. Total anthocyanin concentration (mg L 1) in Syrah, Cabernet Sauvignon,
Merlot, and Maximo grapes from different geographical origins and harvest years.
2.4. LC–MS analysis
LC–MS analyses were performed on a Finnigan LCQ Deca XP
Max ion trap mass spectrometer equipped with an electrospray
ionization (ESI) interface (Thermo Finnigan, San Jose, USA), coupled
to liquid chromatography as described on Section 2.3. After passing
through the flow cell of the diode array detector, the column eluate
was split and 0.3 mL min 1 was directed to the MS. Data acquisi-
tion and processing were performed using Xcalibur 2.0 software.
Positive mode mass spectra of the eluate were recorded in the
range of m/z 150–1000. Nitrogen was used as the sheath gas and
air as the auxiliary gas at flow rates of 62 and 8 arbitrary units,
respectively. The capillary temperature was set at 275 °C and the
spray voltage was 5 kV. For MS/MS experiments, the molecular
ion was isolated in the ion trap, followed by collision-induced dis-
sociation (CID) at 1.5 V. The identification of individual anthocya-
nins was performed by correlating the absorbance and mass
spectra obtained with those previously reported in the literature
and by study of their fragmentation patterns.
2,3 9
250 mAbs
6 7

10 15 20
13 16
4,5 8 12
Table 1 1 11 18
Grapes studied during harvesting in 2008, 2009, and 2010, and their respective codes.
19 (D)
mAbs

9 20
Variety Origina Year Code 8 14
2 7 15 17 18 19
Shiraz São Carlos, SP 2008 S_SC_08 5 10 12
13 (C)
Shiraz São Carlos, SP 2009 S_SC_09
Shiraz São Carlos, SP 2010 S_SC_10 9
2 8
Shiraz Casa Nova, BA 2008 S_CN_08 15
5 7 10
20 (B)
Shiraz Casa Nova, BA 2010 S_CN_10 12 19
Shiraz Louveira, SP 2010 S_L_10
2 7 8 9 14
C. Sauvignon São Carlos, SP 2008 CS_SC_08 5 10 12 13 15 19 20 (A)
C. Sauvignon São Carlos, SP 2010 CS_SC_10
C. Sauvignon Bento Gonçalves, RS 2008 CS_BG_08
Merlot Bento Gonçalves, RS 2008 M_BG_08 0 10 20 30
Maximo Louveira, SP 2010 Mx_L_10 time (min)
a
SP, State of São Paulo (southeastern Brazil); BA, State of Bahia (northeastern Fig. 2. HPLC analysis of (A) Merlot, (B) Cabernet Sauvignon, (C) Syrah, and (D)
Brazil); RS, State of Rio Grande do Sul (southern Brazil). Maximo grapes. Peak identities are listed in Table 2.
398 K. Fraige et al. / Food Chemistry 145 (2014) 395–403

Table 2
Tentative identification of anthocyanins in the grapes studied. Presentation of peak labels, retention times (Tr), compound names, molecular ion, and fragmentation data found for
all varieties of grapes.

Peak Tr Compound Molecular ion (M+) Fragments (m/z) Grapes variety

1 4.96 Delphinidin 3,5-O-diglucoside 627 465/303 Mx
2 7.57 Delphinidin 3-O-glucoside 465 303 Mx, CS, S, M
3 8.00 Petunidin 3,5-O-diglucoside 641 479/317 Mx
4 9.60 Peonidin 3,5-O-diglucoside 625 463/301 Mx
5 9.75 Cyanidin 3-O-glucoside 449 287 Mx, CS, S, M
6 10.90 Malvidin 3,5-O-diglucoside 655 493/331 Mx
7 11.09 Petunidin 3-O-glucoside 479 317 Mx, CS, S, M
8 13.47 Peonidin 3-O-glucoside 463 301 Mx, CS, S, M
9 14.51 Malvidin 3-O-glucoside 493 331 Mx, CS, S, M
10 15.68 Delphinidin 3-O-(6-O-acetyl) glucoside 507 303 Mx, CS, S, M
11 16.47 Malvidin 3-O-(6-O-acetyl)-5-O-diglucoside 697 535/493/331 Mx
12 19.76 Petunidin 3-O-(6-O-acetyl) glucoside 521 317 Mx, CS, S, M
13 21.90 Delphinidin 3-O-(6-O-p-coumaryl) glucoside 611 303 Mx, S, M
14 22.70 Peonidin 3-O-(6-O-acetyl) glucoside 505 301 CS, S, M
15 23.52 Malvidin 3-O-(6-O-acetyl) glucoside 535 331 Mx, CS, S, M
16 23.80 Malvidin 3-O-(6-O-p-coumaryl)-5-O-diglucoside 801 639/493/331 Mx
17 24.50 Cyanidin 3-O-(6-O-p-coumaryl) glucoside 595 287 S
18 25.10 Petunidin 3-O-(6-O-p-coumaryl) glucoside 625 317 Mx, S, M
19 28.07 Peonidin 3-O-(6-O-p-coumaryl) glucoside 609 301 Mx, CS, S, M
20 28.67 Malvidin 3-O-(6-O-p-coumaryl) glucoside 639 331 Mx, CS, S, M

Compound R1 R2 R3 R4
Delphinidin 3,5-O-diglucoside OH OH Glu H
Delphinidin 3-O-glucoside OH OH H H
Petunidin 3,5-O-diglucoside OCH3 OH Glu H
Peonidin 3,5-O-diglucoside OCH3 H Glu H
Cyanidin 3-O-glucoside OH H H H
Malvidin 3,5-O-diglucoside OCH3 OCH3 Glu H
Petunidin 3-O-glucoside OCH3 OH H H
Peonidin 3-O-glucoside OCH3 H H H
Malvidin 3-O-glucoside OCH3 OCH3 H H
Delphinidin 3-O-(6-O-acetyl) glucoside OH OH H acetyl
Malvidin 3-O-(6-O-acetyl)-5-O-diglucoside OCH3 OCH3 Glu acetyl
Petunidin 3-O-(6-O-acetyl) glucoside OCH3 OH H acetyl
Delphinidin 3-O-(6-O-p-coumaryl) glucoside OH OH H p-coumaryl
Peonidin 3-O-(6-O-acetyl) glucoside OCH3 H H acetyl
Malvidin 3-O-(6-O-acetyl) glucoside OCH3 OCH3 H acetyl
Malvidin 3-O-(6-O-p-coumaryl)-5-O-diglucoside OCH3 OCH3 Glu p-coumaryl
Cyanidin 3-O-(6-O-p-coumaryl) glucoside OH H H p-coumaryl
Petunidin 3-O-(6-O-p-coumaryl) glucoside OCH3 OH H p-coumaryl
Peonidin 3-O-(6-O-p-coumaryl) glucoside OCH3 H H p-coumaryl
Malvidin 3-O-(6-O-p-coumaryl) glucoside OCH3 OCH3 H p-coumaryl

Fig. 3. Structures of the anthocyanins identified in the grapes studied.

2.5. Principal component analysis The data set was organized in a matrix with 11 lines corresponding
to samples and 20 columns corresponding to the identified anthocy-
PCA was carried out on the semi-quantitative data obtained by anins. The data set was autoscaled and the PCA calculation was per-
the relative percentage of the areas from the identified anthocyanins. formed using PirouetteÒ v. 4.0 (Infometrix Inc., Bothell, USA).
 K. Fraige et al. / Food Chemistry 145 (2014) 395–403 399

3. Results and discussion
3.1. Total anthocyanins
Fig. 1 presents the total anthocyanin content from the grapes
analyzed. The identity of each grape cultivar is presented in Sec-
tion 2, Table 1.
Except for Syrah produced in Casa Nova in the year of 2008 and
in Louveira in 2010, all grapes of this variety presented similar con-
centrations for total anthocyanins, around 150 mg L 1. For Caber-
net Sauvignon grapes, we found a slightly higher concentration
in grapes cultured in Bento Gonçalves than those cultured in São
Carlos; this was almost the same quantity for Merlot produced in
the same vineyard. It is remarkable that the concentration of
anthocyanins in Maximo grapes was about ten times higher than
the other grapes analyzed. This fact can be visualized by the colour
of the skins and the must produced by the maceration of these
grapes.
3.2. HPLC and LC–MS analysis
The chromatographic fingerprint of anthocyanins from grapes
was initially analyzed using HPLC with absorbance detection at
520 nm. Compounds were subsequently tentatively identified by
LC–ESI-MS/MS considering the generated fragments and data from
the literature (Burns et al., 2002; de Villiers et al., 2004; Kammerer
et al., 2004; Mazzuca et al., 2005; Wu & Prior, 2005). Fig. 2 presents
the chromatograms obtained for Merlot, Cabernet Sauvignon, Syr-
ah, and Maximo grapes. Twenty anthocyanins were identified and
baseline separation for 11 anthocyanins in Cabernet Sauvignon, 13
anthocyanins in Merlot, and 14 anthocyanins in Syrah grapes was
achieved in less than 30 min. The chromatographic profile found
for Maximo grapes was different from those of the V. vinifera
grapes, with a higher number of compounds (18 anthocyanins)
with peak intensities about three times higher, which confirmed
the results obtained by total anthocyanin analysis with spectro-
fragment ions at m/z 535, corresponding to a loss of glucose
(M+ 162) and m/z 331, corresponding to a loss of glucose and
acetylglucoside (M+ 366), the latter representing the aglycone of
malvidin. In this grape variety, the separation of the peaks corre-
sponding to delphinin 3-O-glucoside and petunidin 3,5-O-digluco-
side was very subtle, as well as cyanidin 3-O-glucoside and
peonidin 3,5-O-diglucoside. The baseline resolution was not
achieved for malvidin 3-O-(6-O-acetyl) glucoside and malvidin
3-O-(6-O-p-coumaryl)-5-O-diglucoside. However, all peaks could
be identified and fragmented in MS/MS experiments. The MS and
MS/MS spectra of malvidin 3-O-(6-O-p-coumaryl)-5-O-diglucoside,
the most complex molecule found in the hybrid grape, are pre-
sented in Fig. 4 as an example of the obtained results. The MS spec-
tra of peaks (2,3) and (4,5) in Fig. 2, indicating a slight separation of
delphinidin 3-O-glucoside/petunidin 3,5-O-diglucoside and peoni-
din 3,5-O-diglucoside/cyanidin 3-O-glucoside, respectively, are
presented in Figs. 5 and 6, as well as the MS/MS spectra of each
anthocyanins indicating the identification. The mass spectra of
the other compounds identified in the grapes studied are presented
in Supplementary Figs. 1–5.
The anthocyanins were tentatively identified by LC–MS/MS and
as standards were not available the percentage area of each com-
pound in the samples was carried out considering the total and
individual areas of each peak obtained in the data from the PDA
detector.
Although similar anthocyanins were found in the varieties, the
relative amounts of individual compounds were different. In all
varieties, malvidin-3-O-glucoside was the major anthocyanin, rep-
resenting more than 32% for the vinifera varieties and 22% for the
hybrid variety. For the hybrid variety, delphinidin 3-O-glucoside
and the diglucoside of malvidin were present in high concentra-
tions as well, with the latter detected only in this variety.
7000 (A) 801.06

photometry at 520 nm. The identities of the anthocyanins found 6000

in the grapes are presented in Table 2 together with the retention
time, molecular ion, and fragmentation information. Fig. 3 shows 5000
Intensity

the structures of the compounds detected. 4000

Five monoglucosides of the delphinidin, petunidin, cyanidin,
peonidin, and malvidin were identified in all analyzed grapes by 3000
UV spectral characteristics and by the presence of the characteris-
2000
tic protonated molecular ion (M+H)+ and the fragment (M+ 162),
corresponding to the loss of a glucose moiety. In the same way, 1000
four acetylglucosides and five (p-coumaryl) glucosides of these
0
anthocyanidins were identified in the grapes by the loss of 204 200 300 400 500 600 700 800 900
and 308 mass units, respectively, corresponding to the loss of an m/z
acetylglucoside and a p-coumarylglucoside group. Except for del- 638.98
phinidin 3-O-(6-O-p-coumaryl) glucoside, the elution order for 1400 (B)
the compounds was glucoside < acetylglucoside < (p-coumaryl) 1200
glucoside. In the V. vinifera grapes, we observed the presence of
1000
cyanidin 3-O-(6-O-p-coumaryl) glucoside only in Syrah grapes,
but in small quantities. The (p-coumaryl) glucosides of delphinidin 800
Intensity

and petunidin were not detected in Cabernet Sauvignon grapes, 600 331.15
regardless of the origin.
400
Maximo grapes were characterized by the presence of digluco-
492.99
sides of delphinidin, petunidin, and malvidin. The fragmentation of 200
the generated molecular ions corresponded to the monoglucoside 0
and the aglycone of each anthocyanin. Malvidin 3-O-(6-O-p-coum-
aryl)-5-O-diglucoside has a molecular ion at m/z 801, which yields -200
250 300 350 400 450 500 550 600 650
fragment ions at m/z 639, 493, and 331, corresponding to a loss of
m/z
glucose (M+ 162), p-coumarylglucose (M+ 308), and both
(M+ 470), respectively. Malvidin 3-O-(6-O-acetyl)-5-O-digluco- Fig. 4. Typical (A) MS and (B) MS/MS spectra of malvidin 3-O-(6-O-p-coumaryl)-5-
side has a protonated molecular ion at m/z 697, which yields O-diglucoside found in Maximo grapes after HPLC separation.
400 K. Fraige et al. / Food Chemistry 145 (2014) 395–403
Fig. 5. MS spectra of (A) the peak 2,3 of Maximo grapes indicating the co-elution of delphinidin 3-O-glucoside (m/z 465) and petunidin 3,5-O-diglucoside (m/z 641). (B) and
(C) indicate the respective MS/MS spectra of m/z 465 and m/z 641.
Almost the same content of each anthocyanin was observed for
all V. vinifera grapes, but considering the differences between spe-
cies, we found a higher concentration of cyanidin 3-O-glucoside
and delphinidin 3-O-glucoside in Cabernet Sauvignon than in
Syrah grapes, with the opposite pattern observed for peonidin 3-
O-(6-O-p-coumaryl) glucoside and malvidin 3-O-glucoside that
were found in higher concentrations in Syrah grapes (except for
those cultured in Louveira) than in Cabernet Sauvignon grapes.
Therefore, considering the same species from different places, we
could not observe significant differences.
3.3. Principal component analysis
To evaluate the differences in the anthocyanin composition of
Cabernet Sauvignon, Syrah, and Maximo grapes from different pro-
ducing regions of Brazil, PCA was carried out with the relative
amount of each identified anthocyanin. Fig. 7 shows the scores
and loading plots obtained by PCA.
The first two principal components (PC) accounted for about
73% of the total variance. Samples of each variety are grouped in
the plot of PC1 versus PC2. It is clear that the V. vinifera group is
separated from the hybrid grape, and considering only the vinif-
eras, it is possible to group Cabernet Sauvignon (CS) on the left
of the graph separated from Syrah (S) on the PC2 axis. Merlot
grapes are grouped together with the viniferas grapes and, as only
one sample of this variety was studied, it was not possible to attri-
bute the isolated position of this variety to the formation of an-
other group. The Syrah grapes produced in Casa Nova in 2008
(S_CN_08) and those produced in Louveira in 2010 (S_L_10) were
slightly distant from the other grapes of the same variety, but
the region of culture probably did not influence this because an-
other sample produced in the former location is grouped with
the other Syrah grapes. This separation is probably due to some cli-
matic factor that interfered with the biosynthesis of polyphenols,
including anthocyanins.
It has been reported that the difference between V. vinifera and
other species is the presence or absence of anthocyanidin digluco-
sides, since V. Vinifera contains only monoglucosides from the five
anthocyanidins (Arozarena et al., 2002; Burns et al., 2002; Hebrero,
García-Rodríguez, Santos-Buelga, & Rivas-Gonzalo, 1989; Mazzuca
et al., 2005). The observation of the score plots and loading plots
together indicate that the main anthocyanins responsible for the
 K. Fraige et al. / Food Chemistry 145 (2014) 395–403 401

Fig. 6. MS spectra of (A) the peak 4,5 of Maximo grapes indicating the co-elution of peonidin 3,5-O-diglucoside (m/z 625) and cyanidin 3-O-glucoside (m/z 449). (B) and (C)
indicate the respective MS/MS spectra of m/z 625 and m/z 449.

-2 0 2 4 6 8 10 -0,3 -0,2 -0,1 0,0 0,1 0,2 0,3 0,4

3 3
S_SC_08 0,4 20 0,4
S_CN_10
18
2 S_SC_10 2 17
0,3 19 13 0,3
S_SC_09
0,2 0,2
1 1 1, 3, 4, 6,
Mx_L_10
S_L_10 11, 16
0,1 0,1
PC2 (25.4 %)
PC2 (25.4 %)

S_CN_08
12
0 0 0,0 0,0

-0,1 14 7 -0,1
-1 -1 9
M_BG_08 8
-0,2 15 10 -0,2
-2 -2 -0,3 -0,3
CS_BG_08 2
CS_SC_08 -0,4 5 -0,4
-3 (A) -3 (B)
CS_SC_10
-0,5 -0,5
-2 0 2 4 6 8 10 -0,3 -0,2 -0,1 0,0 0,1 0,2 0,3 0,4
PC1 (47.2 %) PC1 (47.2 %)

Fig. 7. (A) Scores and (B) loading plots for PC1  PC2. j Syrah, d Cabernet Sauvignon, s Merlot, and e Maximo. Numbers in the loading plots represent the peak number
listed in Table 1.
402 K. Fraige et al. / Food Chemistry 145 (2014) 395–403
separation of Maximo from the V. vinifera grapes were all
diglucosides, including delphinidin 3,5-O-diglucoside (1), petuni-
din 3,5-O-diglucoside (3), peonidin 3,5-O-diglucoside (4), malvidin
3,5-O-diglucoside (6), malvidin 3-O-(6-O-acetyl)-5-O-diglucoside
(11), and malvidin 3-O-(6-O-p-coumaryl)-5-O-diglucoside (16).
Considering the V. vinifera grapes, the clustering of Cabernet
Sauvignon can be attributed mainly to the higher amount of cyani-
din 3-O-glucoside (5) in this variety than in Syrah and Maximo
grapes. In contrast to these results, the Syrah grouping may be pri-
marily due to the absence of cyanidin 3-O-(6-O-p-coumaryl) gluco-
side (17) and the lower amounts of peonidin 3-O-(6-O-p-coumaryl)
glucoside (19) and malvidin 3-O-(6-O-p-coumaryl) glucoside (20)
in Cabernet Sauvignon and Maximo compared to this group.
The results obtained by PCA demonstrate that the differences
between samples are due to varietal variability, but in terms of
geographical region, the samples could not be grouped considering
the composition of the anthocyanins alone.
4. Conclusions
Twenty anthocyanins were identified by HPLC–DAD–MS in
grapes from four varieties, including three V. vinifera (Cabernet
Sauvignon, Syrah, and Merlot) and a hybrid grape developed in
Brazil. By means of principal component analysis, it was possible
to see clearly clustering of data separating the viniferas from the
hybrid grape, and within the group of viniferas grapes, it was pos-
sible to separate two groups, one belonging to Syrah grapes and
another to Cabernet Sauvignon. In spite of only a few samples
being analyzed, these results are valid and can show that the
anthocyanin profile can classify grapes from vinifera and non-vinif-
era species cultivated in Brazil, which is important for the assess-
ment of adulteration in juices and wines that can contribute to a
decrease in the quality of these sub-products.
Acknowledgements
The authors would like to thank Mr. Antero Lisciotto, Mr. Daniel
Miqueleto and Miolo Winegroup for the donation of the samples,
Fundação de Amparo a Pesquisa do Estado de São Paulo (FAPESP
2007/07752-7) and Conselho Nacional de Desenvolvimento Cientí-
fico e Tecnológico (CNPq) for a scholarship (K.F.) and financial sup-
port, respectively.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foodchem.2013.
08.066.
References
Abad-García, B., Berrueta, L. A., Garmón-Lobato, S., Gallo, B., & Vicente, F. (2009). A
general analytical strategy for the characterization of phenolic compounds in
fruit juices by high-performance liquid chromatography with diode array
detection coupled to electrospray ionization and triple quadrupole mass
spectrometry. Journal of Chromatography A, 1216, 5398–5415.
Alberts, P., Stander, M. A., & de Villiers, A. (2012). Advanced ultra high pressure
liquid chromatography–tandem mass spectrometric methods for the screening
of red wine anthocyanins and derived pigments. Journal of Chromatography A,
1232, 92–102.
Arozarena, I., Ayestarán, B., Cantalejo, M. J., Navarro, M., Vera, M., Abril, I., et al.
(2002). Anthocyanin composition of Tempranillo, Garnacha and Cabernet
Sauvignon grapes from high- and low-quality vineyards over two years.
European Food Research and Technology, 214, 303–309.
Arozarena, I., Casp, A., Marin, R., & Navarro, M. (2000). Differentiation of some
Spanish wines according to variety and region based on their anthocyanin
composition. European Food Research and Technology, 212, 108–112.
Berente, B., De la Calle Garcia, D., Reichenbächer, M., & Danzer, K. (2000). Method
development for the determination of anthocyanins in red wines by
high-performance liquid chromatography and classification of German red
wines by means of multivariate statistical methods. Journal of Chromatography
A, 871, 95–103.
Burns, J., Mullen, W., Landrault, N., Teissedre, P.-L., Lean, M. E. J., & Crozier, A. (2002).
Variations in the profile and content of anthocyanins in wines made from
cabernet sauvignon and hybrid grapes. Journal of Agricultural and Food
Chemistry, 50, 4096–4102.
Carpentieri, A., Marino, G., & Amoresano, A. (2007). Rapid fingerprinting of red
wines by MALDI mass spectrometry. Analytical and Bioanalytical Chemistry, 389,
969–982.
Castañeda-Ovando, A., Pacheco-Hernandez, M. L., Paez-Hernandez, M. E., Rodriguez,
J. A., & Galan-Vidal, C. A. (2009). Chemical studies of anthocyanins: A review.
Food Chemistry, 113, 859–871.
Choi, J. Y., Lee, S. J., Lee, S. J., Park, S., Lee, J. H., Shim, J., et al. (2010). Analysis and
tentative structure elucidation of new anthocyanins in fruit peel of Vitis
coignetiae Pulliat (meoru) using LC–MS/MS: Contribution to the overall
antioxidant activity. Journal of Separation Science, 33, 1192–1197.
de Rosso, M., Tonidandel, L., Larcher, R., Nicolini, G., Ruggeri, V., Dalla Vedova, A.,
et al. (2012). Study of anthocyanic profiles of twenty-one hybrid grape varieties
by liquid chromatography and precursor-ion mass spectrometry. Analytica
Chimica Acta, 732, 120–129.
de Villiers, A., Vanhoenacker, G., Majek, P., & Sandra, P. (2004). Determination of
anthocyanins in wine by direct injection liquid chromatography–diode array
detection–mass spectrometry and classification of wines using discriminant
analysis. Journal of Chromatography A, 1054, 195–204.
Dugas, A. J. J., Castaneda-Acosta, J., Bonin, G. C., Price, K. L., Fischer, N. H., &
Winstone, G. W. (2000). Evaluation of the total peroxyl radical-scavenging
capacity of flavonoids: Structure–activity relationships. Journal of Natural
Products, 63, 327–331.
Esteban, M. A., Villanueva, M. J., & Lissarrague, J. R. (2001). Effect of irrigation on
changes in the anthocyanin composition of the skin of cv Tempranillo (Vitis
vinifera L) grape berries during ripening. Journal of the Science of Food and
Agriculture, 81, 409–420.
Figueiredo-González, M., Martínez-Carballo, E., Cancho-Grande, B., Santiago, J. L.,
Martínez, M. C., & Simal-Gándara, J. (2012). Pattern recognition of three Vitis
vinifera L. red grapes varieties based on anthocyanin and flavonol profiles, with
correlations between their biosynthesis pathways. Food Chemistry, 130, 9–19.
García-Beneytez, E., Revilla, E., & Cabello, F. (2002). Anthocyanin pattern of several
red grape cultivars and wines made from them. European Food Research and
Technology, 215, 32–37.
Grant, D. C., & Helleur, R. J. (2008). Rapid screening of anthocyanins in berry samples
by surfactant-mediated matrix-assisted laser desorption/ionization time-of-
flight mass spectrometry. Rapid Communications in Mass Spectrometry, 22,
156–164.
Hebrero, E., García-Rodríguez, C., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (1989).
Analysis of anthocyanins by high performance liquid chromatography-diode
array spectroscopy in a hybrid grape variety (Vitis vinifera  Vitis berlandieri
41B). American Journal of Enology and Viticulture, 40, 283–291.
Heinonen, I. M., Meyer, A. S., & Frankel, E. N. (1998). Antioxidant activity of berry
phenolics on human low-density lipoprotein and liposome oxidation. Journal of
Agricultural and Food Chemistry, 46, 4107–4112.
Hong, V., & Wrolstad, R. E. (1990). Use of HPLC separation/photodiode array
detection for characterization of anthocyanins. Journal of Agricultural and Food
Chemistry, 38, 708–715.
Ivanova, V., Dörnyei, A., Stefova, M., Stafilov, T., Vojnoski, B., Kilár, F., et al. (2011).
Rapid MALDI-TOF–MS detection of anthocyanins in wine and grape using
different matrices. Food Analytical Methods, 4, 108–115.
Kammerer, D., Claus, A., Carle, R., & Schieber, A. (2004). Polyphenol screening of
pomace from red and white grape varieties (Vitis vinifera L.) by HPLC–DAD–MS/
MS. Journal of Agricultural and Food Chemistry, 52, 4360–4367.
Mazza, G., & Miniati, E. (1993). Anthocyanins in fruits, vegetables, and grains. Boca
Raton: CRC Press.
Mazzuca, P., Ferranti, P., Picariello, G., Chianese, L., & Addeo, F. (2005). Mass
spectrometry in the study of anthocyanins and their derivatives: Differentiation
of Vitis vinifera and hybrid grapes by liquid chromatography/electrospray
ionization mass spectrometry and tandem mass spectrometry. Journal of Mass
Spectrometry, 40, 83–90.
Mello, L. M. R. (2007). Embrapa Uva e Vinho. Viticultura Brasileira: Panorama 2007.
Available from <http://www.cnpuv.embrapa.br/publica/artigos/
panorama2007_viticultura.pdf>. Accessed 15.09.08.
Moure, A., Cruz, J. M., Franco, D., Dominguez, J. M., Sinerio, J., Dominguez, H., et al.
(2001). Natural antioxidants from residual sources. Food Chemistry, 72,
145–171.
Ng, T. B., Liu, F., & Wang, Z. T. (2000). Antioxidative activity of natural products from
plants. Life Sciences, 66, 709–723.
Pérez-Gregorio, M. R., Regueiro, J., Simal-Gándara, J., Rodrigues, A. S., Almeida, D. P.
F. (2013). Increasing the added-value of onions as a source of antioxidant
flavonoids: A critical review. Critical Reviews in Food Science and Nutrition,
http://dx.doi.org/10.1080/10408398.2011.624283.
Pomar, F., Novo, M., & Masa, A. (2005). Varietal differences among the anthocyanin
profiles of 50 red table grape cultivars studied by high performance liquid
chromatography. Journal of Chromatography A, 1094, 34–41.
Ribéreau-Gayon, P., Glories, Y., Maujean, A., & Dubourdieu, D. (2006). Handbook of
enology: The chemistry of wine stabilization and treatments (Vol. 2, 2nd ed.).
Chichester: John Wiley and Sons.
 K. Fraige et al. / Food Chemistry 145 (2014) 395–403
Rodrigues, A. S., Pérez-Gregorio, M. R., García-Falcón, M. S., & Simal-Gándara, J.
(2009). Effect of curing and cooking on flavonols and anthocyanins in
traditional varieties of onion bulbs. Food Research International, 42, 1331–1336.
Ryan, J.-M., & Revilla, E. (2003). Anthocyanin composition of cabernet sauvignon
and tempranillo grapes at different stages of ripening. Journal of Agricultural and
Food Chemistry, 51, 3372–3378.
Santos Neto, J. R. A., Pereira, F. M., Martins, F. P., & Leitão Filho, H. F. (1968).
Características e possibilidades do cultivar de videira IAC 138-22. O Agronômico,
20, 1–8.
Vian, M. A., Tomao, V., Coulomb, P. O., Lacombe, J. M., & Dangles, O. (2006).
Comparison of the anthocyanin composition during ripening of Syrah grapes
grown using organic or conventional agricultural practices. Journal of Food
Chemistry, 54, 5230–5235.
403
Wang, H., Race, E. J., & Shirkande, A. J. (2003). Characterization of anthocyanins in
grape juices by ion trap liquid chromatography–mass spectrometry. Journal of
Agricultural and Food Chemistry, 51, 1839–1844.
Wang, X., Tong, H., Chen, F., & Gangemi, J. D. (2010). Chemical characterization and
antioxidant evaluation of muscadine grape pomace extract. Food Chemistry, 123,
1156–1162.
Wu, X., & Prior, R. L. (2005). Identification and characterization of anthocyanins by
high-performance liquid chromatography–electrospray ionization-tandem
mass spectrometry in common foods in the United States: vegetables, nuts,
and grains. Journal of Agricultural and Food Chemistry, 53, 3101–3113.
Yang, J., Meyers, K. J., Van Der Heide, J., & Rui, H. L. (2004). Varietal differences in
phenolic content and antioxidant and antiproliferative activities of onions.
Journal of Agricultural and Food Chemistry, 52, 6787–6793.