Research article

Received: 27 December 2011 Revised: 27 February 2012 Accepted: 6 April 2012 Published online in Wiley Online Library

(wileyonlinelibrary.com) DOI 10.1002/jms.3010

Seed oil triglyceride profiling of thirty-twohybrid grape varieties†

Fabiola De Marchi,a,c Roberta Seraglia,b Laura Molin,b Pietro Traldi,b

Mirko De Rosso,a Annarita Panighel,a Antonio Dalla Vedova,a

Massimo Gardiman,a Mirella Giusta and Riccardo Flaminia*

Triglyceride profile of seed oil samples from 32 hybrid grape varieties not studied before was investigated. A new method forthe analysis of triacylglycerols (TAGs) has been developed based on the direct infusion in the electrospray ionization (ESI)source and employing tetrahydrofuran/methanol/water (85:10:5 v|v|v) as solvent; the formation of [M+Na]+ ions in high yieldhas been observed. TAGs were identified by ESI-tandemmass spectrometry analysis, and the matrix-assisted-laser-desorption-ionization and time-of-flight profile of samples was determined. Six were the principal TAGs identified in seed oil: trilinolein(LLL) was the most abundant (43%), followed by dilinoleoyl-oleoylglycerol (LOL, 23%), and dilinoleoyl-palmitoylglycerol(LPL, 15%). Compounds present in lower concentration were LSL and LOO (11%), LOP (6%), and LSP (2%). Compared with seedoils produced from V. Vinifera grapes, some significant differences in the relative abundances of TAGs were found, inparticular hybrid grape seed oils showed higher LOL and lower LPL content, respectively. Among the samples studied, aparticularly high content of LLL (rich in unsaturated fatty acids) was found in seed oils from two red varieties. Copyright © 2012John Wiley & Sons, Ltd.

Keywords: triglyceride; grape seeds; hybrid grapes; MALDI-TOF; ESI/MS

* Correspondence to: Riccardo Flamini, CRA-VIT, Viale XXVIII aprile 26, 31015Conegliano (TV), Italy. E-mail: riccardo.flamini@entecra.it

† This article is part of the Journal of Mass Spectrometry special issue entitled“2nd MS Food Day” edited by Gianluca Giorgi.

a CRA-VIT, Viale XXVIII aprile 26, 31015 Conegliano, TV, Italy

b CNR-ISTM, Sezione di Padova, Corso Stati Uniti 4, 35100 Padova, Italy

c Dip. Biotecnologie Agrarie, Università degli Studi di Padova, Italy

111

Introduction

In the last century, a number of hybrid grape varieties has been pro-duced by crossing between V. Vinifera and non-V. Vinifera varietieswith the aim of having plants resistant to diseases. The grapes pro-duced from these cultivars have been little studied because in severalcountries, such as in the European Community, these are notincluded in the list of varietieswhich canbe used for producingwine.[1]

Nevertheless, the study of chemical composition of thesegrapes is useful to assess their potential as source of foods, bio-products, nutraceuticals and renewable energies. Among thegrape products, the high level of unsaturated fatty acids (up to90%) of seed oil makes this product suitable for culinary andpharmaceutical purposes, and for other technical uses (e.g.cosmetics).[2] The content of linoleic acid is relevant, and this isan important compound involved in the prostaglandin synthesiswhich influences on the platelet aggregation and inflammatoryprocesses.[3] Due to its ability of increasing high density lipopro-tein ‘good’ cholesterol and reducing low density lipoprotein‘bad’ cholesterol, linoleic acid is active in the prevention of cardiocirculatory problems and heart diseases.[4] Other than for theirfatty acid content, the nutritional value and physicochemicalproperties of triglycerides are linked to the positional distributionof the acyl groups along the glycerol backbone. Characterized bylight flavour with fruity touches and high smoke point (216 �C),grape seed oil is commonly used for culinary and frying.[2,3]

Profiling of triacylglycerols (TAGs) allows a rapid characteriza-tion of oils based on the simultaneous detection of all TAGs inthe sample.[5] Characterization of triglyceride composition pro-vides information about processing conditions and the rawmaterial used,[2] and can be used for quality control in order toverify the purity or blending of the product.[6]

J. Mass Spectrom. 2012, 47, 1113–1119

Matrix-assisted-laser-desorption-ionization mass spectrometry(MALDI-MS) proved to be very effective technique in the profilingof TAGs and provides semiquantitative data with little sampleamount in a very short time.[2,7,8]

While TAG composition of seed oils produced from someV. Vinifera grape varieties has been studied,[2,9] oils producedfrom hybrid grapes have been little studied.[10] In the presentwork, seed TAG profiles of thirty-two hybrid grape varietiesselected among the potentially more interesting varieties presentin CRA-VIT (Viticulture Research Centre) grapevine collection,were studied. Structural identification of compounds was per-formed by electrospray ionization tandem MS (ESI-MS/MS)analysis of seed extract, while TAG profile of the samples wasdetermined by MALDI and time-of-flight (TOF) MS.

Materials and methods

Samples and reagents

Grapes from the 32 hybrid varieties reported in Table 1 (21 redand 11 white, respectively) were harvested in 2010 from the

Copyright © 2012 John Wiley & Sons, Ltd.

3

Table 1. TAG percentages in the seed oil MALDI profile of the 32 hybrid grape varieties studied. ESI-MS/MS fragments ions used for identification ofcompounds, are reported

MALDI (%) 877.63[LPL+Na]+

879.65[LOP+Na]+

881.64[LSP +Na]+

901.63[LLL +Na]+

903.64[LOL+Na]+

905.66[LSL/LOO+Na]+

ESI-MS/MS m/z with relative intensities 621.4(67%);597.4(100%)

623.4(100%);599.4(73%);597.6(49%)

624.5(100%);600.5(87%)

621.4 (100%) 623.4(100%)621.4(42%)

625.4(100%);623.4(75%);621.4(43%)

Red varieties

1 Bacò 1 13.1 4.3 1.5 48.2 24.7 8.2

2 Bacò 30-12 17.2 6.7 1.1 41.9 21.6 11.4

3 Bertille Seyve 1808 15.8 6.4 1.5 45.2 20.9 10.1

4 Bertille Seyve 4825 15.5 6.3 0.9 37.6 24.8 14.9

5 Burdin 4077 12.3 6.6 1.9 40.2 26.4 12.6

6 Clinton 15.2 5.9 1.4 44.2 22.4 10.9

7 Couderc 25 15.7 4.8 2.3 46.3 20.2 10.6

8 Galibert 238-35 16.6 6.8 1.6 37.2 22.8 15.0

9 Seibel 8357 11.6 7.1 1.9 38.1 25.4 15.9

10 Seibel 8745 13.0 5.9 1.8 43.6 25.4 10.3

11 Seibel 10878 14.5 5.9 1.2 46.1 22.2 10.1

12 Seyve Villard 12-347 15.9 6.4 1.3 42.3 23.2 10.9

13 Seyve Villard 12-390 13.5 5.7 1.2 42.8 23.7 13.1

14 Seyve Villard 23-369 14.1 5.4 1.4 44.2 25.1 9.7

15 Seyve Villard 23-399 16.8 7.5 1.3 39.2 24.1 11.2

16 Seyve Villard 29-522 15.5 6.7 1.4 38.9 25.0 12.5

17 Terzi 100-31 15.8 4.9 2.2 44.5 20.1 12.5

18 Terzi 108-6 13.1 6.0 2.2 45.1 21.5 12.0

19 Terzi 9746 16.6 5.6 1.7 41.5 23.1 11.6

20 Unknown red 1 15.1 9.1 3.4 33.5 25.5 13.4

21 Unknown red 2 16.3 4.6 0.9 51.6 19.0 7.7

White varieties

22 Bacò 2-16 12.8 6.3 1.6 39.8 26.3 13.1

23 Burdin 7419 15.3 6.5 1.5 41.7 23.9 11.2

24 GA 48-12 17.9 6.3 1.9 44.2 20.5 9.2

25 Galibert 152-26 14.7 5.5 0.9 45.7 24.5 8.8

26 Galibert 255-43 16.0 5.6 1.3 44.8 22.5 9.8

27 Galibert Treblanc 17.5 7.2 1.7 39.6 25.6 8.4

28 GF 84-21-9 15.6 5.8 1.0 44.2 22.0 11.4

29 Inc. Bruni 624 17.1 5.2 1.0 45.9 21.0 9.7

30 Seibel 9110 14.0 6.4 4.0 43.0 23.9 8.7

31 Seibel 10173 14.4 5.5 1.1 45.7 22.8 10.5

32 Seyve Villard 12-357 14.9 5.8 0.9 43.6 25.4 9.3

Mean 15.1 6.1 1.6 42.8 23.3 11.1

S.D. 1.6 0.9 0.7 3.6 2.0 2.0

F. De Marchi et al.

1114

CRA-VIT (Viticulture Research Centre) grapevine collection sitedin Susegana (Treviso, Italy) and immediately frozen.Hexane was purchased from Romil Ltd. (Cambridge, UK),

while tetrahydrofuran, methanol, chloroform, 2,5-dihydroxybenzoicacid were purchased from Sigma-Aldrich (Milan, Italy).

Grape seed extraction

Seeds from 20 berries were manually separated from skin andpulp, dried on a paper sheet, weighted and cold ground by usinga mortar and liquid nitrogen. Seed oil was extracted from thepowder with hexane (ratio 1:10 w/w) performing extractionunder stirring at room temperature, overnight in the dark. Theextract was filtered, and the solvent was removed under vacuum.The oil samples were stored in 2-mL vials under nitrogenatmosphere at �20 �C until analysis.

wileyonlinelibrary.com/journal/jms Copyright © 2012 Joh

To estimate the oil yield, 40 berries of each sample wereweighted, seeds were separated from the other parts of theberry, weighted, dried at room temperature for 1 day andpowdered by using an electric grinder. The oil was extracted inthe same conditions described above. Extracts were transferredin a small flask, the solvent was removed under vacuum andthe residue was weighted.

ESI/MS and ESI/MS/MS measurements

ESI mass spectra were obtained using an ion trap LCQ instrument(Thermo-Finnigan, San Jose, CA, USA), equipped with an ESI ionsource operating in positive ion mode. One milligram of oilsample was dissolved in 1mL of THF/CH3OH/H2O (85:10:5 v/v/v)solution prior to perform analysis by direct infusion using asyringe pump at a flow rate of 8 mL/min.

n Wiley & Sons, Ltd. J. Mass Spectrom. 2012, 47, 1113–1119

Seed triglyceride profiling of hybrid grapes

The instrumental conditions were: spray voltage: 4 kV; entrancecapillary temperature: 280 �C; sheath gas flow rate: 40 a.u.

MS/MS experiments were performed by resonant excitation ofthe ion of interest through a supplementary r.f. voltage in therange 35–40% of its maximum value (5 V peak to peak). Theisolation width was set at 1 mass unit.

MALDI-TOF MS analysis

MALDI/MSmeasurements were performed using a MALDI-TOF-TOFUltrafleXtreme (Bruker Daltonics, Bremen, Germany), equippedwith 1 kHz smartbeam II laser (l=355nm) and operating inreflectron positive ion mode.

The instrumental conditions employed to analyze molecularspecies in the m/z range 450–3500 were: ion source 1: 25.00 kV;ion source 2: 22.30 kV, lens: 7.70 kV, pulsed ion extraction: 80 ns,reflector: 26.45 kV, reflector 2: 13.45 kV. The matrix solution wasprepared by dissolving 10mg of crystalline DHB in 1mL of meth-anol. Oil samples were dissolved in chloroform at a concentrationof 1mg/mL. Five microlitres of matrix solution was mixed withthe same volume of oil solution, and 1 mL of the resulting mixturewas deposited directly on the stainless steel sample holder andallowed to dry before introduction into the mass spectrometer.External mass calibration (Peptide Calibration Standard) wasbased on monoisotopic values of [M+H]+ of Bradykinin,Angiotensin II, Angiotensin I, Substance P, Bombesin, ACTH clip(1–17), ACTH clip (18–39), Somatostatin 28 at m/z 757.39916,1046.5420, 1296.6853, 1347.7361, 1619.8230, 2093.0868, 2465.1990and 3147.4714, respectively.

MALDI-TOF-TOF analysis were carried out on the same oil solu-tions by LIFT device. The instrumental parameters were: IS1 =7.5 kV;IS2= 6.75 kV; LIFT1: 19 kV. LIFT2: 2.9 kV, delay time=70ns.

Statistical analyses

Factor analysis, one-way analysis of variance (ANOVA) and clusteranalysis were performed by using MINITAB Statistical SoftwareInc. (2003), Release 14 for Windows (State College, Pennsylvania).ANOVA applied Tukey test with a 95% significance level; com-parisons having p< 0.05 were considered significant. Clusteranalysis was performed to group together similar cases in thecontent of triglycerides and applied average linkage methodand Euclidean distance.

111

Results and discussion

ESI-MS/MS characterization of TAGs

TAG analysis is usually performed by LC/ESI/MS/MS methodsusing methanol, chloroform and isopropanol, added with ammo-nium acetate and/or sodium acetate and detecting [M+NH4]

+,[M+Na]+ and [M+H]+.[11,12] In the present paper, it was consid-ered of interest the use of direct injection ESI/MS/MS approachand the solvent used was a mixture of THF/CH3OH/H2O(85:10:5 v/v/v). In this condition, protonated molecules are notproduced while the adducts with Na+ lead to highly abundantsignals (106–107 DACS).

All the varieties studied were selected taking into account thegreat interest for their high production of grape per hectareand richness in chemical compounds (in particular antioxidant

J. Mass Spectrom. 2012, 47, 1113–1119 Copyright © 2012 John

compounds such as polyphenols and anthocyanins), and thatseed TAG composition of these grapes was not studied before.

The typical ESI-MS spectrum of grape seed oil is reported inFig. 1a. It is characterized by the presence of two clusters in them/z 877–881 range, corresponding to C55 triglyceride series,and in the m/z 901–905 range, due to C57 triglyceride series.

Structural identification of TAGs was achieved by performingESI-MS/MS analysis of seed oil. Triglycerides were named by peakintensities determined by ESI-MS/MS, assigning the loss with thelowest yield to the fatty acid in the sn-2 position of glycerol back-bone, while the loss of the highest yield was assigned to theisomeric sn-1 and sn-3 positions.

The ESI-MS/MS spectra of the ions at m/z 877, 879, 801, 901,903 and 905 indicate that they correspond to sodium adductsof six different TAGs and are characterized by the losses of theacid present in the structure and by the losses of the related acidsodium salt. As an example, the ESI-MS/MS spectrum of the ion atm/z 879 is reported in Fig. 1b. The fragment ions at m/z 623, 599and 597 corresponds to the loss of palmitic acid (P; 256Da),linoleic acid (L; 280Da) and oleic acid (O; 282Da), respectively,while the loss of their related sodium salts leads to the ionic spe-cies at m/z 601, 577 and 575. The ESI-MS/MS spectrum of the ionat m/z 901 showed two peaks: one at m/z 621, corresponding tothe loss of linoleic acid (280Da) and the other one at m/z 599,due to the loss of sodium linoleate. Consequently, the ion atm/z 901 corresponds to triglyceride LLL sodium adduct ([LLL+Na]+).By this approach, the other peaks were identified as follows:the ion at m/z 877 (losses of 280 and 256 Da) as [LPL+Na]+,the ion at m/z 903 as [LOL +Na]+, the ion at m/z 905 (lossesof 282 and 284 Da corresponding to O and S, respectively) as[LSL +Na]+ and [LOO +Na]+ and ion at m/z 881 as [LSP +Na]+

(see Table 1).From these data, TAG grape seed oil consist principally of four

fatty acids: linoleic acid (L, C18:2n-6,9), oleic acid (O, C18:1n-9),stearic acid (S, C18:0) and palmitic acid (P, C16:0) differentlylinked to the glycerol backbone to form six main TAGs.

MALDI-MS seed oil profiling

The MALDI/MS spectra of the grape seed oils samples showed,again the presence of the C55 and C57 triglyceride series (Fig. 2).The C57 triglycerides series, detected as sodium adducts, giverise to peak intensities higher than that of the C55 triglycerideone. The ions at m/z 901, 903 and 905 correspond to LLL, LOLand LPL/LOO triglycerides, respectively, while the ionic speciesat m/z 877, 879 and 881 are to LPL, LOP and LSP.

Study of the peak intensities (corrected for the isotopic patternof the most intense peak in the cluster, about 20% of the highestpeak in each series was subtracted from each of the peaks follow-ing) allowed to evaluate the relative percentage of each TAG inthe sample. No significant differences between red and whitegrape varieties were found. On the other hand, comparisonamong the different varieties showed peculiar characteristics forsome samples (data in Table 1). In particular, the mean percent-age of LLL (a TAG with high content of unsaturated fatty acids)was about 43%, but samples Bacò 1 and Unknown red 2 hadmean LLL content about 48% and 52%, respectively. On thecontrary, the variety Unknown red 1 showed the lowest LLLpercentage (33%). In the case of LSP, the highest saturated TAGof grape seed oil, the mean percentage was lower than 2%, butthe samples Unknown red 1 and Unknown red 2 had percen-tages of 3.4% and 0.9%, respectively. Mean total percentage of

Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

5

0

2

4

6

x104

Inte

ns. [

a.u.

]

860 870 880 890 900 910 920 930 940

m/z

[LPL + Na]+

[LOP + Na]+

[SLP + Na]+

[LLL + Na]+

[LOL + Na]+

[LSL/LOO + Na]+

+ 16

[LLL + K]+

Figure 2. MALDI-TOF spectrum of sample Bacò 1. The main Na+ and K+ TAGs adducts, are showed.

Burdin_40772_13 #203-245 RT: 4.55-5.50 AV: 31 NL: 7.04E3F: + c ESI Full ms2 879.80@cid40.00 [240.00-900.00]

250 300 350 400 450 500 550 600 650 700 750 800 850 90005

101520253035404550556065707580859095

100 623.37

879.63

599.41

577.39

813.82543.34 730.10693.09303.00 456.35

-256 Da: palmitic acid

-280 Da: linoleic acid

-256 Da: oleic acid

O OO

O

O H

O

Rel

ativ

e A

bund

ance

Burdin_40772_13 #7-573 RT: 0.15-11.61 AV: 104 NL: 4.18E6F: + c ESI Full ms [100.00-1500.00]

850 860 870 880 890 900 910 920 930 940 950

m/z

m/z

05

101520253035404550556065707580859095

100 901.80

903.81

877.73

905.79

879.84906.80

907.80880.87881.86 908.83

899.79 909.81882.87 917.67 933.71875.65861.42853.74 947.35923.72 937.61

Rel

ativ

e A

bund

ance

b

a

Figure 1. a. ESI-MS spectrum of Burdin 4077 grape seed oil sample. b. ESI-MS/MS spectrum of the signal at m/z 879 assigned to the triglyceride LOPsodium adduct ([LOP+Na]+).

F. De Marchi et al.

wileyonlinelibrary.com/journal/jms Copyright © 2012 John Wiley & Sons, Ltd. J. Mass Spectrom. 2012, 47, 1113–1119

1116

F1

F2

3210-1-2

3

2

1

0

-1

-2

0

032

3130

29

28

27

26

25

24

23

22

21

20

19

18

17

16

15

14 13

1211

109

87

6

5

4

3

2

1

Factor Analysis

E

A

B

CD

Figure 3. Factor analysis of 32 hybrid varieties with respect to the content ofthe six triglycerides determined in the grape seed oil samples. 70% of varianceis explained by the first two factors (45 and 25%, respectively). Factor 1 is mostlychargedby variables LLL, LSP, LOP and LSL/LOOandFactor 2 by LPL and LOL. The32 varieties are divided in five groups (A, B, C, D, E) according to cluster analysis.

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

LPL LOP LSP LLL

%

*

Figure 4. TAG percentages in the MALDI seed oil profile of four V. Vinifera grcentages of the 32 hybrid grape varieties studied (* = p< 0.05).[2]

0.0

0.5

1.0

1.5

2.0

x104

Inte

ns. [

a.u.

]

1010 1020 1030 1040 1050 1

[LOO + nonanal + Na]882 + 142 + 23

[LLL + undece878 + 154

Figure 5. Signals at m/z 1047.7, 1055.7 and m/z 1071.7 present in the MALD

Seed triglyceride profiling of hybrid grapes

J. Mass Spectrom. 2012, 47, 1113–1119 Copyright © 2012 John

unsaturated TAGs (calculated as the sum of LLL, LOL, LSL/LOOpercentages) was 77%, and also in this case Unknown red 1showed the lowest (72%) and Bacò 1 the highest (81%) values,respectively. These results are evidenced from the data of factoranalysis shown in Fig. 3: the samples Bacò 1 (sample 1 in Table 1),Unknown red 1 (20) and Unknown red 2 (21) are separatedfrom the others mainly by factor 1 (which includes the variablesLLL, LSP, LOP and LSL/LOO) while GA-4812 (24), Burdin 4077 (5)and Seibel 8357 (9) by factor 2 (which includes LPL and LOL).Cluster analysis put in evidence these results by collecting thesamples 1, 20 and 21 in different classes, the remaining 29varieties are collected into two groups the minor containing thesamples 5 and 9.

A study of TAG composition of cold-pressed seed oil samplesfrom four V. Vinifera grape varieties (Chardonnay, Merlot,Cabernet-Sauvignon, Zweigelt) performed by MALDI-TOF/MShas been recently reported.[2] The relative percentages normal-ized to the six compounds determined in our samples werebetween 39 and 42% for LLL, 15 and 17% for LOL, 11 and 12%for LSL or/and LOO, 22 and 24% for LPL, 4 and 6% for LOP and

LOL LSL-LOO

ChardonnayMerlotCabernet-SauvignonZweigelt32 hybrids

*

ape varieties reported in the literature compared with the mean TAG per-

060 1070 1080 1090 1100 1110

m/z

[LLL + undecanal + Na]878 + 170 + 23

ne + Na] + 23

I/MS profiles of the samples.

Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

1117

F. De Marchi et al.

1118

1 and 4% for LSP. By performing one-way ANOVA, significant dif-ferences - on a confidence level of 95% - between the 32 hybridsand the four V. Vinifera varieties, were found. Data of some com-pounds were similar between two sets of samples, but higherpercentages of LOL (23%, TAG with high content of unsaturatedfatty acids) and lower of LPL (15%) with significance <0.05 werefound in the hybrid grape samples (Fig. 4). Also, the mean percent-age of LSP was lower in oil from hybrid grapes, except for Merlot oilwhere very low content of this compound was found.Besides the peaks corresponding to the six TAGs, other

three intense signals at m/z 1047.7, 1055.7 and m/z 1071.7were found in the MALDI/MS profiles of all samples (Fig. 5).TOF-TOF mass spectrum of the ion at m/z 1071.7 showed lossof 280 Da (linoleic acid) and an intense peak at m/z 599,corresponding to a loss of 472 Da. In a previous paper, signalsat m/z 1055.7 and 1071.7 were found in sunflower seed oilsafter heating for 6 h at 180 �C, and were identified as ‘chain-branched addition products’.[13,14] These compounds areformed from reaction of alkyl and aldehydic fragments, deriv-ing from b-scission of fatty acids hydroperoxides, with intacttriglyceride. Hence, the peak at m/z 1047.7 would be theproduct of nonanal addition to LOO, the ion at m/z 1055.7of undecene addition to LLL, and the ionic species at m/z1071.7 would be due to LLL-undecanal addiction product.These species are completely undetectable in ESI conditions,and this suggests that their origin can be necessarily relatedto MALDI conditions. No acid was added to the starting sam-ple solutions for MALDI analysis, and consequently the originof these peaks cannot be ascribed to chemical phenomena.Looking at the data above described on the thermal behav-iour of TAGs, a possible activation of the same reactionsdue to laser irradiation could be invoked. Among the differentmechanisms proposed for MALDI, laser-induced shock-waves[15] (the laser irradiation produces local high pressureshocks, reflecting in the thermal excitation of a microvolumeof the matrix-analyte sample) and thermal ionization[16] (orig-inating by the photon–phonon transformation in the crystallattice) have been considered, on the basis of the high localtemperature increase due to interaction of the laser beamwith the solid state sample. However, experiments performedwith 1-palmitoyl-2-oleyl-3-stearyl glycerol, used as standardTAG, and showing the lack of formation of the adduct, seemto exclude this mechanism. Looking at the conjugation pres-ent in linoleic acid containing TAGs, it could be hypothesizedthat the production of the adducts is due to photochemicalreactivity of linoleic acid.

Seed oil production

Finally, the mean seed oil production of each variety was esti-mated on samples collected for two years (harvests 2010 and2011, data not shown). In general, seed oil contents between1.0 and 6.0 g/kg grape were found. The more productive sampleswere Bacò 30–12, Clinton, Seibel 10878, Terzi 9746, Inc. Bruni 624and Seibel 10173 with seed oil contents >4.5 g/kg grape. Lowerproductions were found for Bertille Seyve 4825 and Seibel 9110(about 1 g/kg grape). Among the high productive varieties, Seibel10878 and Terzi 9746 also showed high oil percentage content inthe seeds (>10%). It is interesting to report that Seibel 10878 alsoshowed a high production in terms of kg grape per plant, makingthis variety particularly suitable for cultures finalized to high seedoil production.

wileyonlinelibrary.com/journal/jms Copyright © 2012 Joh

Conclusions

The study of MALDI-TOF profile is effective in the TAG character-ization of grape seed oil. TAG profile can be useful to distinguishbetween seed oils produced from hybrid and V. Vinifera grapes inparticular on the basis of the LOL and LPL percentage content.On the contrary, no significant difference between white andred grape varieties has been found.

Due to its high percentage of unsaturated TAGs, the varietyBacò 1 is particularly interesting for production of edible seedoil and Seibel 10878 for high seed oil production.

Other uses of hybrid grape seeds can be linked to antioxidantproperties of their extracts (due to the relevant polyphenolic con-tent) and for biodiesel production.[17–19] Studies finalized to theseaims are in progress.

Acknowledgements

This study was developed within the VITENERGY1 Project, withfunding from the Italian Ministry of Agricultural, Food andForestry Policies, duration 2010–2013.

References[1] European Community (CE) Regulation n. 479/2008.[2] S. Bail, G. Stuebiger, S. Krist, H. Unterweger, G. Buchbauer. Character-

ization of various grape seed oils by volatile compounds, triacylgly-cerol composition, total phenols and antioxidant capacity. FoodChem. 2008, 108, 1122–1132.

[3] J. E. Pardo, E. Fernández, M. Rubio, A. Alvarruiz, G. L. Alonso. Charac-terization of grape seed oil from different grape varieties (VitisVinifera). Eur. J. Lipid Sci. Technol. 2009, 111, 188–193.

[4] V. Wijendran, K. C. Hayes. Dietary n-6 and n-3 fatty acid balance andcardiovasculary health. Ann. Rev. Nutr. 2004, 24, 597–615.

[5] A. Jakab, K. Nagy, K. Heberger, K. Vekey, E. Forgacs. Differentiation ofvegetable oils by mass spectrometry combined with statistical anal-ysis. Rapid Commun. Mass Spectrom. 2002, 16(24), 2291–2297.

[6] S. Krist, G. Stuebiger, S. Bail, H. Unterweger. Detection of adulterationof poppy seed oil with sunflower oil based on volatiles and triacyl-glycerol composition. J. Agr. Food Chem. 2006, 54(17), 6385–6389.

[7] O. Belgacem, A. Bowdler, I. Brookhouse, F. L. Brancia, E. Raptakis. Dis-sociation of biomolecules using a ultraviolet matrix-assisted laserdesorption/ionisation time-of-flight/curved field reflectron tandemmass spectrometer equipped with a differential-pumped collisioncell. Rapid Commun. Mass Spectrom. 2006, 20(11), 1653–1660.

[8] J. O. Lay, Jr., R. Liyanage, B. Durham, J. Brooks. Rapid characterizationof edible oils by direct matrix-assisted laser desorption/ ionizationtime-of-flight mass spectrometry analysis using triacylglycerols.Rapid Commun. Mass Spectrom. 2006, 20(6), 952–958.

[9] M. Lísa, M. Holčapek. Triacylglycerols profiling in plant oils importantin food industry, dietetics and cosmetics using high-performanceliquid chromatograph-atmospheric pressure chemical ionizationmass spectrometry. J. Chromatogr. A 2008, 1198–1199, 115–130.

[10] L. R. Mattick, A. C. Rice. Fatty Acid Composition of Grape Seed Oilfrom Native American and Hybrid Grape Varieties. Am. J. Enol. Vitic.1976, 27, 88–90.

[11] A. Zeb, M. Murkovic. Analysis of triacylglycerols in refined edible oilsby isocratic HPLC-ESI-MS. Eur. J. Lipid Technol. 2010, 112, 844–851.

[12] C. A. Dorschel. Characterization of the TAG of Peanut Oil by Electro-spray LC-MS-MS. JAOCS, 2002, 79, 749–753.

[13] G. Picariello, A. Paduano, R. Sacchi, F. Addeo. MALDI-TOF MassSpectrometry Profiling of Polar and Nonpolar Fractions in HeatedVegetable Oils. J. Agr. Food Chem. 2009, 57, 5391–5400.

[14] C. Byrdwell, W. Neff. Dual parallel electrospray ionization andatmospheric pressure chemical ionization mass spectrometry (MS),MS/MS and MS/MS/MS for the analysis of triacylglycerols and triacyl-glycerol oxidation products. Rapid Commun. Mass Spectrom. 2002,16, 300–319.

[15] D. D. Dlott, S. Hambir, J. Franken. The new wave in shock waves.J. Phys. Chem. B 1998, 102, 2121–2130.

n Wiley & Sons, Ltd. J. Mass Spectrom. 2012, 47, 1113–1119

Seed triglyceride profiling of hybrid grapes

[16] K. Dreisewerd, M. Schurenberg, M. Karas, F. Hillekamp. Influence oflaser intensity and spot size on the desorption of molecules and ionsin matrix-assisted laser-desorption/ionization with a uniform beamprofile. Int. J. Mass Spectrom. Ion Proc. 1995, 141, 127–148.

[17] T. Maier, A. Schieber, D. R. Kammerer, R. Carle. Residues of grape(Vitis vinifera L.) seed oil production as a valuable source of phenolicantioxidants. Food Chem. 2009, 112, 551–559.

J. Mass Spectrom. 2012, 47, 1113–1119 Copyright © 2012 John

[18] H. Lutterodt, M. Slavin, M. Whent, E. Turner, L. Yu. Fatty acid compo-sition, oxidative stability, antioxidant and antiproliferative propertiesof selected cold-pressed seed oils and flours. Food Chem. 2011, 128,391–399.

[19] C. M. Fernández, M. J. Ramos, A. Pérez, J. F. Rodríguez. Production ofbiodiesel from winery waste: Extraction, refining and transesterifica-tion of grape seed oil. Bior. Tech. 2010, 101, 7019–7024.

Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

1119