Food Chemistry 155 (2014) 132139

Contents lists available at ScienceDirect

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

The proteins of the grape (Vitis vinifera L.) seed endosperm: Fractionation
and identication of the major components
Diana Gazzola a, Simone Vincenzi a, Luca Gastaldon a, Serena Tolin b, Gabriella Pasini a, Andrea Curioni a,
a
Department of Agronomy, Food, Natural Resources, Animals and Environment (DAFNAE) and Centro Interdipartimentale per la Ricerca in Viticoltura ed Enologia (CIRVE),
Padova University, via dellUniversit 16, 35020 Legnaro (PD), Italy
b
Proteomics Center of Padova University, VIMM and Padova University Hospital, via G. Orus 2b, 35129 Padova, Italy

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

Article history: In the present study, grape (Vitis vinifera L.) seed endosperm proteins were characterized after sequential
Received 22 July 2013 fractionation, according to a modied Osborne procedure. The salt-soluble fraction (albumins and glob-
Received in revised form 9 December 2013 ulins) comprised the majority (58.4%) of the total extracted protein. The protein fractions analysed by
Accepted 14 January 2014
SDSPAGE showed similar bands, indicating different solubility of the same protein components. SDS
Available online 23 January 2014
PAGE in non-reducing and reducing conditions revealed the polypeptide composition of the protein
bands. The main polypeptides, which were similar in all the grape varieties analysed, were identied
Keywords:
by LCMS/MS as homologous to the 11S globulin-like seed storage proteins of other plant species, while
Electrophoresis
Globulins
a monomeric 43 kDa protein presented high homology with the 7S globulins of legume seeds.
Grape seed The results provide new insights about the identity, structure and polypeptide composition of the grape
Mass spectrometry seed storage proteins.
Seed proteins 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Grape is one of the worlds most widely grown fruit crops, with
a reported annual production higher than 69 million tons
(FAOSTAT, 2011), which is almost completely transformed in wine.
The main winemaking by-product is grape pomace, constituting
13% of the total grape weight (10 million tons) and containing
3852% (on a dry matter basis) of grape seeds (Maier, Schieber,
Kammerer, & Reinhold, 2009), which are then produced in several
tons per year. A small amount of this is used for the extraction of
the seed oil and natural antioxidants (Arvanitoyannis, Ladas, &
Mavromatis, 2006), and recently a grape seed protein extract was
proposed as a valuable ning agent for wine (Vincenzi et al.,
2013). However, most of the grape seeds are treated as a waste
material. The nutritional quality of the grape seed proteins have
been considered for a long time, with these proteins having nutri-
tional characteristics similar to those of other oilseeds and cereals
(Igartuburu et al., 1991). Moreover, grape seed proteins show some
functional properties such as good solubility and emulsifying activ-
ity (Zhou, Zhang, Liu, & Zhao, 2011).
Although several papers have been published on the identica-
tion and characterisation of the grape pulp and skin proteins (for a
review see Giribaldi & Giuffrida, 2010, and the citations therein),
Corresponding author. Tel.: +39 (0) 49 8272624; fax: +39 (0) 49 8272929.
E-mail address: andrea.curioni@unipd.it (A. Curioni).
grape seed proteins have received little attention. The protein
content of the grape seed was the object of early studies, which
reported varying values ranging from 8.44% (Igartuburu et al.,
1991) to 25.9% (Fazio, Gattuso, Cilluffo, & Arcoleo, 1983). In
contrast, only a few papers have been published on the character-
isation of the different protein components present in the grape
seed (Gianazza et al., 1989; Zhou, Li, Zhang, Bai, & Zhao, 2010).
In general, plant seeds contain different types of proteins, most
of which can be classied as storage proteins, a nitrogen reserve for
early seedling growth (Shewry, Napier, & Tatham, 1995). Early
studies by Osborne divided the seed proteins according to their
solubility in different solvents into albumins (water-soluble),
globulins (salt-soluble), prolamins (aqueous alcohol-soluble) and
glutelin (acidic or alkaline solution-soluble) (Osborne, 1924). This
approach is still considered a valid method to fractionate and
classify seed proteins. In several plant species, the main storage
proteins belong to the globulin fraction, commonly divided into
11S and 7S according to their sedimentation behaviour (Shewry
et al., 1995).
The major protein of the grape endosperm was initially shown
to be a 60 kDa globulin containing disulde-linked 1921 kDa and
3844 kDa peptides (Gianazza et al., 1989). More recently, a pro-
tein made of two polypeptides with molecular masses of 25.5
and 40.0 kDa was isolated and puried from grape seeds and iden-
tied by mass spectrometry as an 11S globulin-like protein (Zhou
et al., 2010). However, no other information was available on the

http://dx.doi.org/10.1016/j.foodchem.2014.01.032
0308-8146/ 2014 Elsevier Ltd. All rights reserved.
 D. Gazzola et al. / Food Chemistry 155 (2014) 132139
identity of the other protein components of the grape seed, nor on
their structural organisation. Moreover, during a study on the ef-
fects of water-decit stress on grape berry tissues, several grape
seed proteins were identied by the proteomic approach, most of
them belonging to the 11S globulins with storage function,
although in this case their structural organisation in the seed
was not investigated (Grimplet et al., 2009). Other studies focused
on grape varieties differentiation by their seed protein composi-
tions (Pesavento et al., 2008) although the proteins were not
characterised.
In this paper, the proteins of the grape seed endosperm were
fractionated according to their solubility and characterised by
electrophoretic analyses in terms of polypeptide compositions.
Moreover, the main electrophoretic bands, which were present in
several grape varieties as major components, were identied by
mass spectrometry.
2. Materials and methods
2.1. Plant material
Seeds were manually extracted from the berries of different
white (Glera, Trebbiano, Durella, Cortese, Moscato Colli Euganei,
Manzoni Bianco, Moscato Fiori dArancio, Garganega) and red
(Raboso Piave and Corvina) grape (Vitis vinifera L.) varieties,
harvested in the Conegliano area. For each variety, one hundred
randomly selected, fully ripe seeds were used. Seeds were washed
with distilled water and left to dry at room temperature. By means
of a razor blade, seeds were dissected and the integuments were
removed, while the endosperm was recovered and immediately
frozen for storage.
2.2. Lipid extraction
Approximately 2 g (cv. Glera) or 400 mg (other varieties) of
endosperm was ground to a ne our in a mortar in the presence
of an excess of n-hexane. The suspension was ltered at 0.45 lm
with lters for organic solvents (polyethersulfone, PESU, Sartorius,
Goettingen, Germany). The delipidated endosperm was recovered,
weighed and stored at 4 C.
2.3. Seed protein sequential fractionation based on solubility
Protein fractions were sequentially extracted at room tempera-
ture according to a modied Osborne procedure (Sogi, Arora, Garg,
& Bawa, 2002) as follows. The defatted our (1.4 g for cv. Glera and
100 mg for the other varieties) was extracted with 0.5 M sodium
chloride (1:10, w/v) with constant stirring for 30 min. The slurry
was centrifuged at 14,000g for 5 min. The supernatant was recov-
ered and two further extractions were performed on the pellet,
each stirring for 15 min. The three supernatants were pooled,
ltered at 0.45 lm and extensively dialyzed (3 kDa MW cut-off)
against distilled water. After dialysis, the suspension was centri-
fuged (14000g, 5 min). The precipitate (globulins) and the superna-
tant (albumins) were separated and freeze-dried.
The insoluble pellet resulting from the salt extraction was then
treated with 70% v/v aqueous ethanol (1:10 w/v) and centrifuged
(14,000g, 5 min). The procedure was repeated three times and
the supernatants were pooled, ltered at 0.45 lm and dialyzed
against distilled water. Finally, the content of the dialysis tube
(prolamins) was freeze-dried.
The insoluble pellet from the previous extraction was treated
with 0.05 M acetic acid (1:10 w/v). The suspension was stirred
for 30 min and centrifuged (14,000g, 5 min.). After dialysis against
distilled water, the fraction (glutelin) was freeze-dried.
133
The remaining insoluble material was nally extracted by stir-
ring (20 min.) at 100 C, using a solution containing 2% Sodium
Dodecyl Sulphate (SDS) (Bio-Rad Laboratories, Milan, Italy). The
sample was centrifuged and the residue extracted again with the
same buffer. Supernatants were combined, dialyzed against water
and freeze-dried (un-extractable proteins, UP). A scheme of the
whole extraction procedure is reported in the Supplementary data.
2.4. Protein content determination
Total nitrogen of the defatted grape seed endosperm (cv. Glera)
and of the extracted fractions was determined after sample miner-
alisation by a HACH Digesdahl apparatus (HACH Company,
Loveland, CO, USA). Ammonia was quantied with the Nessler
reagent (Vogel & Svehla, 1979), and protein content was computed
as ammonia  6.25.
2.5. Sodium Dodecyl SulfatePolyacrylamide Gel Electrophoresis (SDS
PAGE)
SDSPAGE analyses were performed according to Laemmli
(1970) in a Mini-Protean III apparatus (Bio-Rad). Aliquots of
freeze-dried albumin, globulin, prolamin and glutelin fractions
were solubilised in 10 ll of 0.5 M TrisHCl buffer, pH 6.8, contain-
ing 15% (w/v) glycerol, 1.5% (w/v) SDS (Bio-Rad) and 4% (v/v)
2-mercaptoethanol (2-ME) (SigmaAldrich, Milan, Italy) (loading
buffer). Samples were heated at 100 C for 5 min. before loading.
For SDSPAGE analyses under non-reducing conditions, the reduc-
ing agent 2-ME was omitted from the loading buffer. Electrophore-
sis was carried out at 25 mA constant current until the tracking dye
Bromophenol Blue ran off the gel. The molecular weight standard
proteins were the Broad Range Molecular Weight Markers (Bio-
Rad). 1.5 mm thick gels were prepared with T = 16% (acrylamide-
N, N0 methylenebisacrylamide 29:1; SigmaAldrich), and stained
with Colloidal Coomassie Brilliant Blue G-250 (SigmaAldrich).
2.6. Two-dimensional (transversal) SDSPAGE (non-
reducing  reducing)
After a standard SDSPAGE performed in non-reducing condi-
tions (rst dimension), the gel lane of interest was cut using a razor
blade, placed in a tube with 5 ml of loading buffer containing 4% (v/
v) 2-ME and heated for 5 min at 100 C. After cooling, the gel slice
was placed horizontally on the top of a second gel (second dimen-
sion) and xed seeping a 0.5% (w/v) agarose solution. Runs were
performed under the same conditions previously mentioned for
SDSPAGE.
2.7. Liquid chromatographytandem mass spectrometry (LCMS/MS)
analyses
After electrophoretic separation, the selected bands were
excised from the gel and subjected to in-gel protein digestion.
Briey, cysteines were reduced with 10 mM dithiotreitol (DTT)
(1 h, 56 C, in the dark) and alkylated with 55 mM iodoacetamide
(1 h, room temperature, in the dark). After washes with 50 mM
NH4HCO3 and acetonitrile, digestion was performed at 37 C over-
night using sequencing grade modied trypsin (Promega, Madison,
WI) (12.5 ng/lL). Peptides were extracted with 50% acetonitrile/1%
formic acid, dried under vacuum and dissolved in 10 ll of 0.1%
formic acid. Liquid chromatographytandem mass spectrometry
(LCMS/MS) analyses were performed with a 6520 Q-TOF mass
spectrometer (Agilent Technologies, Santa Clara, CA, USA) coupled
to a chip-based chromatographic interface. 4 ll of samples were
loaded into the enrichment column (C18, 4 mm, 40 nl volume) at
a ow rate of 4 ll/min. Peptides were separated in the C18
134 D. Gazzola et al. / Food Chemistry 155 (2014) 132139
nano-column (43 mm  75 lm) at a ow rate of 0.5 ll/min, using
water/formic acid 0.1% and acetonitrile/formic acid 0.1% eluents
A and B, respectively, with a gradient of eluent B from 3% to 50%
in 15 min. MS/MS spectra of the 3 most intense ions were acquired
for each MS scan in the range of 3502400 Da. Capillary voltage
was set to 1750 V. Raw data les were analyzed using Proteome
Discoverer Software (version 1.2, ThermoFisher Scientic),
connected to a Mascot Search Engine server version 2.2.4 (Matrix
Science, London, UK). Spectra were searched against UniRef100
database (version 2010, 10573053 sequences, 3710354253
residues) with enzyme specicity set to trypsin with 2 missed
cleavages, precursor and fragment ions tolerance set to 10 ppm
and 0.05 Da, respectively, with carbamidomethylcysteine as xed
modication and oxidation of methionine as variable modication.
Proteins were considered as positive hits if at least 2 peptides were
identied with medium condence (5% False Discovery Rate). In
order to obtain more information about the function of the
identied protein, a BLAST search was carried out.
3. Results and discussion
3.1. Protein fractionation based on solubility
The total protein content of the defatted seed endosperm was
found to be 40% as determined by nitrogen analysis. This percent-
age was much higher than that previous found for the grape seeds
(Castriotta & Canella, 1978; Fazio et al., 1983; Igartuburu et al.,
1991), probably due to the fact that the endosperm was separated
from the seed coat, which consists mainly of ber (Castriotta &
Canella, 1978). To study the grape seed proteins, a modied Os-
borne fractionation procedure (Sogi et al., 2002) with additional
modications was applied. The endosperm proteins were fraction-
ated into albumins (water-soluble), globulins (0.5 M NaCl-soluble),
prolamins (70% alcohol-soluble) and glutelins (0.05 M acetic
acid-soluble). Moreover, a further extraction step was performed
on the remaining material using the detergent SDS, thus obtaining
the unextractable protein fraction (UP). As conrmed here by
SDSPAGE analysis (Fig. 1), the Osborne method shows many inac-
curacies as not all the proteins strictly full the solubility criteria
(Ribeiro, Teixeira, & Ferreira, 2004). For this reason, the protein
extraction yields of the grape seeds were calculated considering
the pool of the salt-soluble proteins, in which both the so-called
albumin and globulin fractions, containing common compounds,
are present (Fig. 1). Two other works (Castriotta & Canella, 1978;
Fig. 1. SDSPAGE under non-reducing (a) and reducing (b) conditions of the protein
fractions extracted from the grape seed endosperm (cv. Glera). A: albumins, G:
globulins, P: prolamins, Gl: glutelins and UP: un-extractable proteins (fraction
extracted with SDS). MW: molecular weight standard proteins. MWs in kDa are
indicated on the left.
Fazio et al., 1983) proposed the Osbornes protocol to fractionate
the grape seed proteins, but starting from whole defatted grape
seeds. In this case the high amount of polyphenols, especially tan-
nins, of the seed coat (Fantozzi, 1981) can bind and precipitate
some protein components, thus removing them from the extract
(Fazio et al., 1983). In contrast, the extraction of the proteins from
the separated seed endosperm, as performed here, improves pro-
tein isolation and analysis by SDSPAGE (Gianazza et al., 1989).
Protein quantication by nitrogen analysis (N  6.25) (Table 1)
showed that the salt-soluble fraction comprised the majority
(58.4%) of the seed endosperm protein. These results are in con-
trast with those of Castriotta and Canella (1978), who found that
albumins accounted for only 5.7% of total seed protein content,
while the globulin amount was negligible. On the contrary, Fazio
et al. (1983) reported that the albumins accounted for 20.1
40.3% of the total protein, while globulins only for 2.84.2%. The
prolamin and the glutelin fractions had a negligible protein con-
tent, whereas the extraction with the SDS solution gave a fraction
(unextractable proteins) containing 17.3% of the total protein. In
addition, a quantity of nitrogen corresponding to 15.5% of the total
protein content could not be extracted with the solutions here
used. In previous work, the percentage of the proteins that could
not be solubilised by any solvent of the Osborne procedure ranged
from 40% (Fazio et al., 1983) to 80% (Castriotta & Canella, 1978),
whereas here, considering the protein content of the un-extract-
able fraction and residue, only 32.8% of total protein was not
extracted, conrming the convenience to use the puried endo-
sperm as the starting material.
3.2. SDSPAGE analysis of the protein fractions
The water-soluble proteins (albumins according to the Osborne
classication) present in the salt-soluble extract were separated
from those unsoluble in water (globulins) by dialysis against water
and centrifugation. The different protein fractions were then ana-
lysed by SDSPAGE under both non-reducing and reducing condi-
tions (Fig. 1).
A clear cross-contamination between the albumin and globulin
fractions could be detected (compare lane A with lane G in both
non-reducing and reducing conditions). Indeed, the bands of the
albumin fraction exactly matched with those present in the
globulin fraction although their relative amounts were different.
This result conrms that overlapping of albumins and globulins
is a major problem for the Osborne fractionation, as found for other
seeds (Ribeiro et al., 2004), with this occurring due to partial
protein solubilities in both water and salt solution.
In non-reducing conditions, the albumin and globulin fractions
(Fig. 1a, lanes A and G) showed many bands in the 2565 kDa
range. In particular, three major polypeptides of about 35, 40 and
60 kDa were detected, although in different relative amounts, in
both the fractions. In contrast a 8 kDa band appeared only in
the globulin fraction indicating this protein was completely insol-
uble in water.
When the same fractions were analysed under reducing condi-
tions (Fig. 1b), their patterns were very different from those
showed in non-reducing conditions, indicating that most of the
grape seed proteins are organized in structures stabilised by disul-
de bonds, which comprise different polypeptides, as already dem-
onstrated for other seed globulins (Shewry et al., 1995). Also in this
case differences between the albumin and globulin fractions were
noted only for the bands in the low MW region.
The fraction extracted with 70% ethanol showed only few faint
bands, which seem to derive from a cross-contamination with the
salt-soluble fraction. These bands were barely detectable in both
non-reducing and reducing conditions (Fig. 1, lane P), conrming
 D. Gazzola et al. / Food Chemistry 155 (2014) 132139 135

Table 1
Yield of extracted fraction (mg/g of defatted grape seed our), percentage of protein in each fraction and protein yield of each fraction. Mean values and standard deviations of
three determinations, calculated on a dry basis, are reported. N.d.: not determined.

Solvent Fraction Yield (mg/g) Protein content (%) Protein yield (mg/g)
0.5 M NaCl Salt-soluble (albumins + globulins) 202.11 25.72 75.14 0.37 151.81 18.57 (58.38 5.50%)
70% Ethanol Prolamins 4.31 1.55 N.d. N.d.
0.05 M Acetic acid Glutelins 10.96 5.00 N.d. N.d.
2% SDS Un-extractable protein 182.93 35.46 25.33 9.14 44.72 7.74 (17.27 3.47%)
Residue 301.29 3.23 13.37 0.39 40.29 0.73 (15.52 0.16%)
Total 701.60 64.49a 236.83 11.56 (91.17 1.87)
a
The total of the recovered material does not reach 1 g due to sample losses during the extraction procedure.

the presence of a very low quantity of proteins in the prolamin

fraction of the grape seed endosperm (Table 1).
The acetic acid-soluble fraction (Fig. 1, lane Gl) showed only one
band at 9 kDa, but only when the extract was analysed after
reduction. This band was contained exclusively in this fraction,
indicating solubility at low pH.
Finally, the protein pattern of the fraction extracted with SDS
(Fig. 1, lane UP) paralleled that of the water/salt-soluble protein,
suggesting the presence of similar proteins with different solubili-
ties. This conrms earlier results reported by Gianazza et al.
(1989). According to their unextractability in water/salt, the bands
detected in the UP fraction are likely to be present in the grape
seed endosperm as subunits of very high molecular weight aggre-
gates, which could be dissociated and rendered soluble only after
the action of the detergent SDS known to break non-covalent
interactions. As seen for the seed storage proteins of many plants
(Shewry et al., 1995), those of the grape are hetero-dispersed with
regards to their molecular size (Gianazza et al., 1989).
A joint consideration of the electrophoretic results of Fig. 1
leads to the conclusion that most of the grape seed protein frac- Fig. 2. Transversal (two-dimensional: non-reducing  reducing) SDSPAGE of the
tions obtained by the modied Osborne method comprise similar salt-soluble fraction of the grape seed endosperm (cv. Glera). MW: molecular
protein bands, although they can be extracted in different relative weight standard proteins. MWs in kDa are indicated on the left and on the top.
proportions by the different solvents. In particular, the polypeptide
composition of some fractions were almost identical (albumins,
transversal gel (Fig. 2), corresponding to the 43 kDa band detected
globulins and UP), conrming the previous results reported for
in SDSPAGE under reducing conditions (Fig. 1b). This behaviour
legume seeds (Freitas, Teixeira, & Ferreira, 2004; Ribeiro et al., 2004).
indicates that this protein is a monomer having intra-molecular
disulphide bonds which, if broken, makes its structure more
3.3. Transversal gel electrophoresis of the salt-soluble fraction
extended, thus increasing its hydrodynamic volume and apparent
MW in SDSPAGE, as shown for other proteins (Vincenzi & Curioni,
In order to precisely dene the subunit composition of the
2005). A similar protein has been found to be particularly reactive
protein bands detected under non-reducing conditions by one-
with tannins (Wu & Lu, 2004) and this reactivity has been
dimensional electrophoresis, transversal (two-dimensional:
conrmed here by SDSPAGE after incubation of the salt-soluble
non-reduced  reduced) SDSPAGE was used to analyse the whole
fraction with increasing amounts of grape seed tannin (data not
salt-soluble fraction, because albumins and globulins were not
shown). Therefore, probably due to its binding to the seed coat tan-
distinguishable in terms of protein bands (Fig. 1).
nins, this protein was missing in the extracts of the whole grape
With this technique, the 60 kDa band of the non-reduced SDS
seed (Zhou et al., 2010), but can be detected (as shown here) using
PAGE pattern was split into two polypeptides of 40 and 25 kDa
the dehulled seed.
which were detected after reduction (Fig. 2, see also Fig. 1b), indi-
cating that this 60 kDa band is a dimer stabilised by SS bonds.
This result conrms previous ndings obtained for the 11S globu- 3.4. SDSPAGE analysis of the seed proteins from different V. vinifera
lin-like protein from grape seeds, which showed the presence of varieties
components made of two polypeptides with molecular masses of
25.5 and 40 kDa (Zhou et al., 2010). Dimeric globulins of the 11S In order to conrm the presence of similar proteins in different
type comprising disulphide linked polypeptides with MW similar grape varieties, both the albumin and the globulin fractions of
to those found for the grape seed are widespread as storage some white (Trebbiano, Durella, Cortese, Moscato Colli Euganei,
proteins in several plant seeds (Marcone, 1999). Manzoni Bianco, Moscato Fiori dArancio, Garganega, Glera) and
The major protein at 35 kDa detectable in non-reducing condi- red (Raboso and Corvina) grape varieties were analysed by SDS
tions, when analysed by transversal SDSPAGE, gave rise to two PAGE. Beside differences attributable to the genetic variability of
heavy stained bands of 25 and 13 kDa (Fig. 2), indicating that the samples, the non-reduced (60, 40, 35 and 8 kDa) and reduced
this protein also comprised two disulde-linked polypeptides. To (43, 40, 25 and 13 kDa) bands described above for the cv. Glera
the best of our knowledge, this 35 kDa grape seed protein has were generally monomorphic by SDSPAGE (Fig. 3), suggesting
never been described. that these proteins, although showing minor differences, are
In contrast, the 40 kDa band detected in SDSPAGE in generally conserved among the grape varieties here considered.
non-reducing conditions moved slightly above the diagonal of However, the pattern of the cv. Glera shown in Fig. 3, which is
136 D. Gazzola et al. / Food Chemistry 155 (2014) 132139

Fig. 3. SDSPAGE under non-reducing (a and c) and reducing (b and d) conditions of the albumin (A) and globulin (G) fractions from different grape varieties: Trebbiano (T),
Durella (D), Cortese (C), Moscato Colli Euganei (E), Manzoni Bianco (M), Moscato Fiori dArancio (A), Raboso Piave (R), Garganega (G), Glera (L), Corvina (V).

similar to those of the other varieties tested, seems to differ from
that showed in Fig. 1, with the ratio of the intensities of the 60
and 35 kDa bands being much higher in the pattern shown in
Fig. 3. A possible explanation for this fact is that the 35 kDa band
is a degradation product of the 60 kDa protein occurring during
the process of fractionation and dialysis used in the early
experiments.
In the albumin fractions of the different varieties, the 60 kDa
band was the most relevant protein (Fig. 3a), its staining intensity
being apparently similar among the grape varieties examined. The
exception was cv. Manzoni Bianco, with a lower amount of this
band compared to the other varieties (Fig. 3a), which was likely
the reason for the disappearance of the 60 kDa band after reduc-
tion (Fig. 3b). In general, while the albumin bands in the range
3145 kDa were rather similar among the varieties, a certain poly-
morphism could be detected for the bands with apparent MWs
lower than 31 kDa (Fig. 3a).
When reduced, the albumin extracts gave rise to the 25 kDa
polypeptide of the 60 kDa dimer, which showed the same electro-
phoretic mobility in the different varieties (Fig. 3b). Also one of the
components of the doublet around 40 kDa (containing the other
polypeptide of the 60 kDa dimer) was similar among the varieties.
However differences in the number of bands, their electrophoretic
mobility and relative staining intensity could be detected around
this MW (Fig. 3b). These differences are not surprising, since minor
modications of the seed storage proteins are common for varie-
ties of the same species. Indeed, these modications do not result
in the loss of the protein function, which is related to the nitrogen
storage.
For the seeds of all the varieties analysed, the globulin fraction
was mainly represented by the band at 40 kDa (Fig. 3c), which
shifted to 43 kDa when reduced (Fig. 3d), conrming the results
of the transversal SDSPAGE (Fig. 2). Other minor bands were pres-
ent, some of which differed in electrophoretic mobility and stain-
ing intensity among the varieties in both non-reducing (Fig. 3c)
and reducing (Fig. 3d) conditions. In this case, the globulin fraction
of Manzoni Bianco differed from that of the other varieties, with a
lower quantity of the 40 kDa band.
3.5. Identication of the grape seed proteins by mass spectrometry
The identity of the main salt-soluble proteins (comprising both
albumins and globulins) of the grape seed endosperm was then
established by LCMS/MS, using the cv. Glera as the sample
material.
 D. Gazzola et al. / Food Chemistry 155 (2014) 132139
The following bands were selected: the dimeric 60 kDa (band
1), the monomeric 40 kDa (band 2), the dimeric 35 kDa (band 3)
and the monomeric 8 kDa (band 4) bands, which were excised
from a SDSPAGE gel run under non-reducing conditions (Fig. 4a).
While band 4 was directly submitted to LCMS/MS analysis
after digestion with trypsin, bands 1, 2 and 3 were reduced with
2-ME and loaded on a second SDSPAGE gel (Fig. 4b).
It was conrmed that the unreduced 60 kDa band (Fig. 4a, band
1), was split into two major polypeptides of 40 and 25 kDa after
reduction, each represented by a doublet (1A and 1B, 1C and 1D,
Fig. 4b, lane 1), which exactly reproduced the patterns recently
reported by Zhou et al. (2010, 2011). Also the 35 kDa protein
(Fig. 4a, band 3) gave rise, in addition to minor bands, to a doublet
around 25 kDa (3A and 3B, Fig. 4b, lane 3) and a band at 13 kDa
(3C, Fig. 4b, lane 3). In contrast, the unreduced 40 kDa band
(Fig. 4a, band 2) was a monomer changing its apparent MW from
40 to about 43 kDa when reduced (Fig. 4b, lane 2). All these results
thus conrmed those of the transversal SDSPAGE (Fig. 2).
The bands were excised from the gel, digested with trypsin and
analysed by LCMS/MS, which allowed the identication of the
proteins described above (Table 2). Both bands 1A and 1B could
be ascribed to the same protein identied in the Uniref100 data-
base with a high score as hypothetical protein isoform 2 of V.
vinifera. Therefore they might be two forms of the same polypep-
tide with a minimal difference in electrophoretic mobility, which
could be due, for example, to a post-translational modication of
the original protein. An identical identication resulted also for
the bands of the 25 kDa doublet (Table 2). When the sequence
found for these bands was introduced into the BLAST, bands 1A,
1B, 1C and 1D all showed high sequence homology with members
of the seed storage globulins of the 11S type (Table 2). All these re-
sults indicate that the polypeptides deriving from the reduction of
band 1 are similar to the acidic and basic polypeptides existing in
the subunits of the 11S seed globulins (Marcone, 1999; Shewry
et al., 1995) and conrms that this type of widespread proteins is
present also in V. vinifera seeds (Gianazza et al., 1989; Zhou
et al., 2010, 2011).
Zhou et al. (2010) puried a grape seed protein that they named
11S globulin-like protein because its sequence matched best
with that of the 11S globulin of a variety of Ficus pumila. In their
paper, this protein was shown to be composed of two polypeptides
with molecular masses of 25.5 and 40.0 kDa, (corresponding to
those here described as components of band 1) and these two poly-
peptides were supposed to be encoded by different genes. Here it is
Fig. 4. Globulin bands 1, 2 and 3 (in boxes) separated in non-reducing SDSPAGE (a) were cut and reloaded in reducing SDSPAGE. The resulting bands (1AD, 2 and 3AC)
(b) were identied by LCMS/MS analysis. Globulin band 4 (a) and glutelin band 5 (loaded in reducing SDSPAGE, (c) were directly submitted to LCMS/MS analysis.
137
demonstrated that these two polypeptides belong to a 60 kDa di-
mer (band 1), which is similar to the dimeric subunits of the 11S
legumin-type seed globulins. As has been reported for these pro-
teins, the polypeptides of the dimers are fragments of a precursor
protein that is proteolytically cleaved after the formation of an
intramolecular disulphide bond which maintains the two frag-
ments linked together (Shewry et al., 1995). Therefore, by analogy
to what has been observed for the 11S globulins of other plants, it
is likely that also in grape seeds these two polypeptides are
synthesised as a single precursor protein.
Bands 3B (25 kDa) and 3C (13 kDa) derived from the reduction
of the 35 kDa globulin (Fig. 4a, band 3), which has never been
described in V. vinifera seeds. They were identied by MS and,
when submitted to BLAST analysis, both bands showed sequence
similarity with the 11S seed storage globulin of Chenopodium
quinoa and Amaranthus hypochondriacus (Table 2). Therefore the
original unreduced 35 kDa band seems to belong to the 11S seed
globulins type, as occurred for the 60 kDa protein (band 1)
described above. The relationships between the proteins found
for the 60 and 35 kDa bands remain to be established, although
their similarity suggests a common origin. As mentioned above,
the 35 kDa band could be a proteolytic product of the 60 kDa
protein originating from the degradation of its 40 kDa polypeptide.
25 kDa polypeptides are found in both the 60 and 35 kDa proteins
as bands 1C, 1D and 3B (Fig. 4), which all show the same identity
by MS analysis (Table 2). Therefore, the low MW polypeptides
resulting from the reduction of the 35 kDa protein, including band
3C (Fig. 4b) would be fragments of the 40 kDa polypeptide.
Only two tryptic peptides were identied for band 3A, and thus
its identication remains uncertain. However, these peptides indi-
cated sequence homology with the basic 7S globulin of Medicago
truncatula and conglutin gamma of Lupinus albus (Table 2), which
are homologous proteins having a high content of sulphur amino
acids (Kagawa, Yamauchi, & Hirano, 1987). This protein contains
one main subunit of about 40 kDa, which comprises two polypep-
tide chains linked by disulde bonds, one of which having a MW of
about 26 kDa (Melo, Ferreira, & Teixeira, 1994). This situation may
be compatible with the results here reported for band 3 (Fig. 4a), if
the presence of two co-migrating protein is supposed around
35 kDa in unreducing conditions.
The 40 kDa protein of band 2 (Fig. 4a and b) was identied as
Putative uncharacterized protein of V. vinifera, which matched
best with the basic 7S globulin of M. truncatula and conglutin gam-
ma of Lupinus albus (Table 2). Contrary to the subunits of the latter
138 D. Gazzola et al. / Food Chemistry 155 (2014) 132139
Table 2
List of bands analyzed by LCMS/MS and identied using Mascot search engine.
Gel Protein Protein identication name
band accession
1A UPI00019839EA PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
UPI00019839EE PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
1B UPI00019839EE PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
UPI00019839EA PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
1C UPI00019839EE PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
D7U304 Whole genome shotgun sequence of line PN40024,
scaffold_5.assembly12x [Vitis vinifera]
1D UPI00019839EF PREDICTED: hypothetical protein isoform 3 [Vitis
vinifera]
UPI00019839EC PREDICTED: hypothetical protein isoform 4 [Vitis
vinifera]
2 A5C7L5 Putative uncharacterized protein [Vitis vinifera]
3A A5C7L5 Putative uncharacterized protein [Vitis vinifera]
3B UPI00019839EE PREDICTED: hypothetical protein isoform 2 [Vitis
vinifera]
UPI00019839EC PREDICTED: hypothetical protein isoform 4 [Vitis
vinifera]
3C D7U302 Whole genome shotgun sequence of line PN40024,
scaffold_5 assembly 12x [Vitis vinifera]
4 UPI00019839D3 Hypothetical protein isoform 1 [Vitis vinifera]
UPI00019839EA PREDICTED: hypothetical protein isoform 3 [Vitis
vinifera]
5 UPI000198433F PREDICTED: hypothetical protein [Vitis vinifera]
homologous legume protein, which is made of distinct polypep-
tides linked by disulde bonds (Melo et al., 1994; Restani, Duranti,
Cerletti, & Simonetti, 1981), the protein of band 2 did not split into
polypeptides after reduction (Figs. 2 and 4b) indicating that it is
monomeric in the grape seed. However, the presence in this mono-
mer of at least one intramolecular disulde bond can be inferred
from the shift of the apparent MW from 40 to 43 kDa as observed
in SDSPAGE after reduction (Fig. 2). Therefore it is possible that
the 40 kDa grape seed protein of band 2 is an analogue of the pre-
cursor protein that in legumes originates the two disulphide linked
polypeptides after being post-translationally cleaved by a specic
protease (Shewry et al., 1995). The presence of 7S globulin-like
proteins in V. vinifera seeds is described here for the rst time. In
another case, peptides matching a 7S globulin were detected by
MS with the bottom-up approach during a study on ancient grape
seeds (Cappellini et al., 2010) although the protein was not de-
scribed by SDSPAGE.
Band 4 (Fig. 4a), which did not change its mobility (8 kDa)
after reduction (Figs. 1 and 2), was similar to that found by
Gianazza et al., 1989, who also described this component as
monomeric. The protein of band 4 was identied as V. vinifera
hypothetical protein isoform which also showed high sequence
homology with the 11S globulin seed storage proteins (Table 2).
Similar low MW bands related to the 11S globulins were de-
scribed for other seeds as degradation products of a vicilin sub-
unit (Baumgartner & Chrispeels, 1977). Therefore it is possible
that band 4 is a proteolytic fragment related to the degradation
process of the 60 kDa protein which gave raise to the 35 kDa
component (see above).
Mascot # BLAST homology attribution
score Peptides
363 4 Legumin A precursor, putative [Ricinus communis]; 11S legumin
protein [Carya illinoinensis]
261 4 11S seed storage globulin B [Chenopodium quinoa]; 11S globulin
seed storage protein [Amaranthus hypochondriacus]
1251 6 11S seed storage globulin B [Chenopodium quinoa]; 11S globulin
seed storage protein [Amaranthus hypochondriacus]
974 4 Legumin A precursor, putative [Ricinus communis]; 11S legumin
protein [Carya illinoinensis]
608 4 11S seed storage globulin B [Chenopodium quinoa]; 11S globulin
seed storage protein [Amaranthus hypochondriacus]
440 4
471 3 11S globulin seed storage protein [Amaranthus hypochondriacus];
11S seed storage globulin [Chenopodium quinoa]
265 2 11S globulin seed storage protein [Amaranthus hypochondriacus];
11S seed storage globulin [Chenopodium quinoa]
648 7 Basic 7S globulin [Medicago truncatula]; conglutin gamma
[Lupinus albus]
375 2 Basic 7S globulin [Medicago truncatula]; conglutin gamma
[Lupinus albus]
322 4 11S seed storage globulin B [Chenopodium quinoa]; 11S globulin
seed storage protein [Amaranthus hypochondriacus]
251 2 11S globulin seed storage protein [Amaranthus hypochondriacus];
11S seed storage globulin [Chenopodium quinoa]
144 2 11S seed storage globulin A [Chenopodium quinoa]; 11S seed
storage globulin [Chenopodium quinoa]
140 4 11S globulin seed storage protein [Amaranthus hypochondriacus];
11S seed storage globulin B [Chenopodium quinoa]
140 4 Legumin A precursor, putative [Ricinus communis]; 11S legumin
protein [Carya illinoinensis]
134 2 Nonspecic lipid-transfer protein A, putative [Ricinus communis];
lipid transfer protein 2 [Euphorbia lagascae]
The globulin fraction was subjected to SDSPAGE analysis with
PAS staining, in order to detect the presence of sugars associated
with the different protein bands. No bands were observed to
have detectable carbohydrate moieties (data not shown). While
glycosylation is not commonly observed among 11S globulins,
the 7S are often glycosylated at different extents (Black, Bewley,
& Halmer, 2006), although cases of 7S globulins having no
carbohydrate have been reported (Garcia, Arocena, Laurena, &
Tecson-Mendoza, 2005).
Finally, two tryptic peptides from band 5 of the glutelin fraction
were indicated to have homology with a putative nonspecic
Lipid-Transfer Protein (nsLTP) A from Ricinus communis and Lipid
Transfer Protein 2 from Euphorbia lagascae (Table 2). The presence
of LTP peptides has been reported only in a study on ancient grape
seeds (Cappellini et al., 2010), but the corresponding protein was
not described by SDSPAGE.
4. Conclusions
The effectiveness of dehulling the grape seeds improves the
protein extraction and allows protein characterization and analy-
sis. However, the Osborne fractionation method adopted here does
not seem to be completely suitable to obtain clearly separated pro-
tein fractions. This might depend on the presence of different
aggregation states of the same proteins, showing different solubil-
ities. The action of the detergent SDS used for the electrophoretic
analyses was likely sufcient to abolish the interactions involved
in the stabilization of these native protein complexes, indicating
 D. Gazzola et al. / Food Chemistry 155 (2014) 132139
their non covalent nature. LCMS/MS analysis conrmed the pres-
ence of globulin-like proteins in V. vinifera seeds, where these com-
ponents, which are present in high amounts, should have a storage
function. Generally these types of proteins are present in the seeds
of several plant species as high molecular weight multimers
stabilised by non-covalent interactions occurring among different
subunits, which are in turn constituted by disulphide-linked
polypeptides originating from the cleavage of a precursor protein
(Shewry et al., 1995). Although the original multimers were not
characterised here, the polypeptide structure of their different sub-
units was assessed. These polypeptides have been shown to belong
to the globulin-like protein family, conrming that also in V. vinif-
era they are the main proteins of the seed (Gianazza et al., 1989;
Grimplet et al., 2009; Zhou et al., 2010). In particular, whereas
most of the proteins showed homology with the globulins of the
11S type, at least a component of about 40 kDa matching with a
7S globulin was also identied here for the rst time.
In conclusion, the grape seed contains storage proteins which
seem to be similar to those of several other plant species. Further
research is needed to elucidate the native structure of these seed
components as they are present in the endosperm in order to
improve the knowledge on their potential utilization for practical
applications (Vincenzi et al., 2013; Zhou et al., 2011).
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.2014.
01.032.
References
Arvanitoyannis, I., Ladas, D., & Mavromatis, A. (2006). Potential uses and
applications of treated wine waste: a review. International Journal of Food
Science and Technology, 41, 475487.
Baumgartner, B., & Chrispeels, M. J. (1977). Purication and characterization of
vicilin peptidohydrolase, the major endopeptidase in the cotyledons of mung-
bean seedlings. European Journal of Biochemistry, 77, 223233.
Black, M., Bewley, J. D., & Halmer, P. (2006). The encyclopedia of seeds: science,
technology and uses. Cambridge, USA: CABI publishing.
Cappellini, E., Gilbert, M. T. P., Geuna, F., Fiorentino, G., Hall, A., Thomas-Oates, J.,
et al. (2010). A multidisciplinary study of archaeological grape seeds.
Naturwissenschaften, 97, 205217.
Castriotta, G., & Canella, M. (1978). Protein classication and nitrogen extractability
of grape seed meal. Journal of Agricultural and Food Chemistry, 26, 763765.
Fantozzi, P. (1981). Grape seed: A potential source of protein. Journal of the American
Oil Chemists Society, 58, 10271031.
FAOSTAT (Food and Agriculture Organization Corporate Statistical Database).
(2011). URL <http://www.fao.org> Accessed 06.05.13.
Fazio, G., Gattuso, A., Cilluffo, V., & Arcoleo, G. (1983). Preparation and
characterization of protein materials from grapeseed meals. Rivista della
Societa Italiana di Scienza dellAlimentazione, 6, 469478.
139
Freitas, R. L., Teixeira, A. R., & Ferreira, R. B. (2004). Characterization of the proteins
from Vigna unguiculata seeds. Journal of Agricultural and Food Chemistry, 52,
16821687.
Garcia, R. N., Arocena, R. V., Laurena, A. C., & Tecson-Mendoza, E. M. (2005). 11S and
7S globulins of coconut (Cocos nucifera L.): Purication and characterization.
Journal of Agricultural and Food Chemistry, 53, 17341739.
Gianazza, E., Tedesco, G., Villa, P., Scienza, A., Cargnello, G., & Righetti, P. (1989).
Characterization of the major proteins from Vitis Vinifera seeds. Plant Science, 62,
7381.
Giribaldi, M., & Giuffrida, M. G. (2010). Heard it through the grapevine: Proteomic
perspective on grape and wine. Journal of Proteomics, 73, 16471655.
Grimplet, J., Wheatley, M. D., Jouira, H. B., Deluc, L. G., Cramer, G. R., & Cushman, J. C.
(2009). Proteomic and selected metabolite analysis of grape berry tissues under
well-watered and water-decit stress conditions. Proteomics, 9, 25032528.
Igartuburu, J. M., Del Ro, R. M., Massanet, G. M., Montiel, J. A., Pando, E., & Luis, F. R.
(1991). Study of agricultural by-products. Extractability and amino acid
composition of grapeseed (Vitis vinifera) proteins. Journal of the Science of Food
and Agriculture, 54, 489493.
Kagawa, H., Yamauchi, F., & Hirano, H. (1987). Soybean basic 7 S globulin represents
a protein widely distributed in legume species. FEBS Letters, 226, 145149.
Laemmli, U. (1970). Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature, 227, 680685.
Maier, T., Schieber, A., Kammerer, D. R., & Reinhold, C. (2009). Residues of grape
(Vitis vinifera L.) seed oil production as a valuable source of phenolic
antioxidants. Food Chemistry, 112, 551559.
Marcone, M. F. (1999). Biochemical and biophysical properties of plant storage
proteins: A current understanding with emphasis on 11S seed globulins. Food
Research International, 32, 7992.
Melo, T. S., Ferreira, R. B., & Teixeira, A. N. (1994). The seed storage proteins from
Lupinus albus. Phytochemistry, 37, 641648.
Osborne, T. (1924). The vegetable proteins (2nd ed.). London: Longmans Green and
Co.
Pesavento, I. C., Bertazzo, A., Flamini, R., Dalla Vedova, A., De Rosso, M., Seraglia, R.,
et al. (2008). Differentiation of Vitis vinifera varieties by MALDI-MS analysis of
the grape seed proteins. Journal of Mass Spectrometry, 43, 234241.
Restani, P., Duranti, M., Cerletti, P., & Simonetti, P. (1981). Subunit composition of
the seed globulins of Lupinus albus. Phytochemistry, 20, 20772083.
Ribeiro, A. C., Teixeira, A. R., & Ferreira, R. B. (2004). Characterization of globulins
from common vetch (Vicia sativa L.). Journal of Agricultural and Food Chemistry,
52, 49134920.
Shewry, P. R., Napier, J. A., & Tatham, A. S. (1995). Seed storage proteins: Structures
and biosynthesis. The Plant Cell, 7, 945956.
Sogi, D., Arora, M., Garg, S., & Bawa, A. (2002). Fractionation and electrophoresis of
tomato waste seed proteins. Food Chemistry, 76, 449454.
Vincenzi, S., & Curioni, A. (2005). Anomalous electrophoretic behavior of a chitinase
isoform from grape berries and wine in glycol chitin-containing sodium dodecyl
sulfatepolyacrylamide gel electrophoresis gels. Electrophoresis, 26, 6063.
Vincenzi, S., Dinnella, C., Recchia, A., Monteleone, E., Gazzola, D., Pasini, G., et al.
(2013). Grape seed proteins: A new ning agent for astringency reduction in red
wine. Australian Journal of Grape and Wine Research, 19, 153160.
Vogel, A. I., & Svehla, G. (1979). Vogels Textbook of macro and semimicro qualitative
inorganic analysis (5th ed.). London: Longman.
Wu, L., & Lu, Y. (2004). Electrophoretic method for the identication of a haze-active
protein in grape seeds. Journal of Agricultural and Food Chemistry, 52,
31303135.
Zhou, T., Li, Q., Zhang, J., Bai, Y., & Zhao, G. (2010). Purication and characterization
of a new 11S globulin-like protein from grape (Vitis vinifera L.) seeds. European
Food Research and Technology, 230, 693699.
Zhou, T., Zhang, T., Liu, W., & Zhao, G. (2011). Physicochemical characteristics and
functional properties of grape (Vitis vinifera L.) seeds protein. International
Journal of Food Science & Technology, 46, 635641.