Antioxidant content and volatile composition of seedless table grape (Vitis vinifera L.) varieties

Grapes are important sources of phenolics, which are recognized to possess high biological value. In this work, we evaluated the total phenol content (TPC), total anthocyanin content (TAC), antioxidant activity (AA) and volatile composition of unstudied seedless table grape varieties (i.e., Autumn Crisp and Pristine as white cultivars, Scarlotta and Crimson as red cultivars and Adora and Melody as black cultivars). As a result, Adora and Melody exhibited higher TPCs, TACs and AA, measured by the 2,2-diphenyl-2-picrylhydrazil (DPPH) and photochemiluminescence (PCL) assays, than the rest of the varieties. A comparison between the two black cultivars proved that Adora possessed higher TPC (62.70 mg GAE 100 g−1), DPPH radical scavenging activity (IC50 of 3.69 mg ml−1) and PCL antiradical activity (72.14 µg Trolox ml−1) than Melody (47.30 mg GAE 100 g−1, IC50 of 5.0 mg ml−1 and 42.36 µg Trolox ml−1, respectively). The volatile composition, determined by solid phase microextraction (SPME)–gas chromatography/mass spectrometry (GC/MS), indicated a similar qualitative aroma profile for all varieties. However, quantitative differences were measured among them, which were reflected in distinct sensory perception by sensorial testing. From these results, Adora was selected for its antioxidant properties and flavor characteristics.


Introduction
Grape (Vitis vinifera L.) represents one of the major fruit crops around the world with many different cultivars varying in flavor and color. In addition to their sensory characteristics, grapes are also known for their antioxidant properties, which are associated with their high phenolic content. In this regard, grapes are one of the most important sources of phenolics as compared to other fruits and vegetables [1]. Phenolic compounds are an integral part of the human diet and regarded as non-nutrients with relevant biological activity [2], which is mainly attributed to their powerful antioxidant, metal chelating and antiradical activities [3]. Among the biological properties related to phenolics, the anti-inflammatory, antioxidant [4,5], antimicrobial [6] and antiaging [7] activities can be highlighted. Phenolics also play an important role in the sensorial characteristics of grapes, contributing to their color, astringency, bitterness and aroma [8]. The main aromatic compounds in grape berries are monoterpenes, benzene derivatives and C6 alcohols. Although all of them represent naturally occurring volatile compounds in grape cultivars, varietal features are frequent due to altitude, soil, climate and viticulture practices contributing to its intensity [9].
On the other hand, the basic characteristic of current table grape production is its adaptation to the requirements of consumers to improve grape quality. In the last few years, an important attribute of the grape berry quality is the seedlessness. Seedless grapes are grown from cuttings, which refer to amputated parts of a vine that is infected with the genetic defect that causes it to grow seedless grapes.
There are a number of bibliographic studies on both the phenolic content and the aroma profile in grape berries in the literature. However, most of them have been carried out on wine grape varieties [1,3,[8][9][10]. Information on seedless table grapes is in general scarce and, in particular, on the varieties here studied is almost inexistent.
The purpose of this research was to study the phenolic content, antioxidant properties and the volatile composition of unstudied seedless table grape varieties. To that end, we evaluated TPC, TAC and AA (i.e., DPPH • radical scavenging and PCL antiradical activities), of Autumn Crisp, Pristine, Scarlotta, Crimson, Adora and Melody grapes. All these cultivars are relatively recent since they have been developed to satisfy the current consumer demand for seedless table grapes. Although occasional reports can be found in the literature on Scarlotta and Crimson varieties [11][12][13], the antioxidant content and properties of the rest of the cultivars have not been, to our knowledge, assessed as yet. Also, a comparative study of volatile constituents by SPME-GC-MS together with sensory evaluation was performed to get an insight into the organoleptic characteristics of each cultivar. This latter aspect has never been studied in these varieties either.

Materials
For the experiments, seedless table grape berries of six varieties were used: Autumn Crisp and Pristine (as white varieties), Scarlotta and Crimson (as red varieties) and Adora and Melody (as black varieties). All of them were cultivated in Murcia (Spain) and supplied by SAT MOYCA (Murcia). Grape berries were collected at the optimum maturity stage. Immediately after reception, grapes were manually separated from the stems and kept at − 80 °C until analysis.

Extraction
Prior to the actual extraction, all samples were lyophilized. Then phenolics were extracted from each grape variety by following the same procedure. A 25 ml volume of methanol:water (70:30 v/v) was added to the sample (2 g). The mixture was then homogenized by using an Ultra-Turrax (T18 Digital, IKA) for 5 min and subsequently centrifuged at 2500 rpm for 15 min at 4 °C. The upper layer was removed and the remaining precipitate was re-extracted. Both extracts were finally combined and taken to dryness. The extract was kept at − 20 °C until analysis. Duplicates of each extraction for every cultivar were carried out. The extracts were used for the determination of TPC, TAC and AA, as explained below.

Determination of TPC
TPC measurements were performed using a Beckman Coulter DU-800 spectrophotometer (Barcelona, Spain). Based on the literature, we applied the Folin-Ciocalteu method [14]. In brief, we prepared the extract by dissolving 20 mg of dry extract in 1 mL −1 of methanol:water (70:30). Then, a 500 µL of Folin-Ciocalteu reagent and 10 mL of a sodium carbonate solution (75 g L −1 ) were added to a 100 µL of the extract. The mixture was made up to 25 mL with distilled water and the absorbance was measured at 750 nm after 1 h against a blank (i.e., mixture without Folin-Ciocalteu reagent). The results were expressed as milligrams of gallic acid equivalents (GAE) per 100 g of fresh weight (FW). All the analyses were carried out in triplicate.

Determination of TAC
The pH differential method was used to quantify anthocyanins [15]. The grape extracts were diluted with 0.025 M potassium chloride buffer solutions at pH 1 and with 0.4 M sodium acetate buffer at pH 4.5. A 400-700 nm sweep was carried out using a spectrophotometer (Beckman Coulter DU-800 spectrophotometer, Barcelona, Spain). The TAC values obtained were expressed as milligrams of cyanidin-3-O-glucoside equivalents (C3G) per 100 g of FW based on a molar extinction coefficient of 26,900 L cm −1 and a molecular weight of 449.4 g/L.
The following equation was applied to calculate the total absorbance:

DPPH • radical scavenging activity
The AA in terms of free radical scavenging activity was determined by the DPPH • assay [16] with slight modifications. A spectrophotometer (Beckman Coulter DU-800 spectrophotometer, Barcelona, Spain) was also used to carry out the measurements. First, the extracts were dissolved in methanol (i.e., 20 mg mL −1 ). This stock solution was further diluted to final concentrations of 15.6, 62.5, 125, 250 and 500 µg mL −1 . Each extraction solution, before adding DPPH • , was used as a blank. A 15 µL of DPPH • (400 µmol l −1 ) was added to a 50 µL of the sample. After that, the mixture was then incubated at 37 °C for 30 min and the absorbance nm was monitored at 517 nm. The value of absorbance obtained from the DPPH • reagent solution was used as a reference. Therefore, the measurement provided by the sample indicates percentage inhibition of the DPPH • by each dilution of samples. A plot of percentage inhibition versus concentration was represented and the IC 50 values, expressed as mg ml − 1, were calculated using linear regression analysis. The experiments were performed in triplicate.

PCL antiradical activity
The AA in terms of antiradical activity was also determined by using a PCL assay. This method was applied using Photochem ® device (Analytik Jena AG, Jena, Germany) and conducted by the ACL protocol [17]. A commercial reagent kit ACL (Analytik Jena AG, Jena, Germany) was acquired for this study [18]. For the assays, 20 μL of the sample (30 g L −1 of the extract dissolved in methanol:water (70:30)) was mixed with the ACL reagent. Subsequently, the mixture was placed in the Photochem device. Results were calculated on the basis of standard curves into ng Trolox equivalents per mL of sample (µg mL −1 ).

SPME
The extraction was carried out by using a fused-silica fiber coated with a 65 µm layer of polydimethylsiloxane/divinylbenzene (PDMS/DVB) (65 µm) installed in a SPME holder for manual use (Supelco, Madrid, Spain). The fiber was previously conditioned in the injector of the gas chromatograph at 250 ℃ for 30 min as recommended by the supplier. An approximate weight of 6 g of fresh grape berries was mushed and placed in a 10 ml vial. Different grapes were used to obtain a homogenous sample minimizing this way fruit-to-fruit variability. The vial was sealed with plastic film with characteristics suitable for the SPME extraction (i.e., insensitivity to usual reagents and low water permeability). Prior to the actual extraction, the equilibration of volatiles in the sample headspace was reached by heating the sample at the extraction temperature (i.e., 70 °C) for 5 min. After the equilibration time, the extraction was performed by exposing the fiber to the sample for 10 min. The extraction conditions were selected on the basis of the literature [19]. When the extraction was completed, the compounds retaining in the fiber were thermally desorbed by inserting the fiber into the injector port of the GC. Finally, the volatile compounds were analyzed by gas chromatography-mass spectrometry (GC-MS), as specified below.

GC analysis
A Hewlett-Packard model 6890 gas chromatograph fitted with a split/splitless injector and mass spectrometer (MS) model HP59 was used for the analyses. The SPME fiber was desorbed at 250 °C for 3 min into the GC injector. Splitless mode was used in all instances. GC separations were performed on 30 m × 0.25 mm i.d. fused-silica column coated with a 0.25 μm layer of poly (ethylene glycol) phase (i.e., carbowax, Quadrex, USA). The initial temperature was held at 40 °C for 5 min, then the column was first programmed at 5 °C/min to 240 °C and finally at 20 °C/min to 260 °C, which was held for 5 min. Helium was used as the carrier gas at a linear velocity rate of1 mL/min and a constant pressure mode was employed (10 psi). The source and the quadrupole temperatures were set at 230 °C and 280 °C, respectively. The SCAN mode was always used. Data acquisition from the MS was accomplished with HP-ChemStation system (Agilent Technologies, Palo Alto, CA, USA). The identification of the volatile compounds analyzed was made by matching the mass spectra with those provided by the Wiley library.  [20] and that some phenolics act as co-pigments to stabilize this color [21]. Therefore, the results here found support bibliographic reports on the relation between phenolic content and grape skin color [22]. Black varieties possessed in general, and Adora in particular, the highest TPC and TAC values as compared with red and especially with white cultivars. It is interesting to point out the particularly high TPC value obtained for Adora (i.e., 62.70 mg GAE 100 g −1 FW). It is also worth pointing out the relatively high TPC value measured in Pristine cultivar despite being a white variety (i.e., 32.28 mg GAE 100 g −1 FW). In fact, Pristine TPC value was even higher than those of red varieties (26.77 and 26.37 mg GAE 100 g −1 FW in Scarlotta and Crimson, respectively). This is probably owing to the occurrence of some relevant non-anthocyanin phenolic component contributing significantly to TPC in Pristine grapes. Figures 1 and 2 represent the AA in terms of DPPH • (expressed as IC 50 , mg ml −1 ) and PCL (expressed as µg Trolox ml −1 ), respectively, of seedless table grapes of six different varieties. On the basis of the results obtained from both assays, the black varieties, especially Adora, exhibited the highest AA, followed by the red varieties, Scarlotta and Crimson, and finally by the white varieties, Autumn Crisp and Pristine. These results agree with reports on the relationship between grape berry color and AA [23]. By comparing both red varieties, Scarlotta possessed higher AA than Crimson as measured by both DPPH (i.e., IC 50 5.30 vs 6.27 mg ml −1 ) and PCL (72.14 vs 42.30 µg Trolox ml −1 ) assays. This difference could be visually appreciated from the color, which was more intense in the case of Scarlotta. It is also noticeable that Melody showed an IC 50 value (i.e., 5.0 mg ml −1 ) close to those of the red varieties.

Results and discussion
A possible correlation between the results on AA (Figs. 1,  2) and TPC and TAC data (see Table 1) was studied by linear regression. All varieties exhibited direct correlation of both TPCs and TACs with DPPH radical scavenging activity (r 2 = 0.9563 and 0.9945, respectively) and PCL antiradical activity (r 2 = 0.8938 and 0.9912 respectively). These results indicate that, in general, phenolics and, particularly, anthocyanins are mainly responsible for AA measured in the grape samples included in this study. The only exception to this general trend was TPCs and TACs of the red cultivars (Scarlotta and Crimson) and DPPH data, which did not display a linear dependence (r 2 = 0.3640 and 0.4465, respectively).
As already mentioned, both DPPH and PCL assays showed, in general terms, similar results. However, it is interesting to highlight the slight differences found between them for the red cultivars, Scarlotta and Crimson. Whereas IC 50 values of both red varieties (i.e., 5.30 and 6.27 mg ml −1 ) were closer to those of the black cultivars (i.e., 3.69 mg ml −1 for Adora and 5.00 mg ml −1 for Melody), their PCL values (i.e., 72.14 and 42.40 µg Trolox ml −1 ) were comparable to those of the white varieties (i.e., 43.85 µg Trolox ml −1 for Autumn Crisp and 49.06 µg Trolox ml −1 for Pristine). This is explained by the different mechanism of action on which both assays are based to determine AA. Whereas the DPPH method determine the ability to act as free radical scavenger or hydrogen donor, the PCL assay measures the capacity to react in a single free radical reaction. Actually, differences in the results provided by both mechanisms of action have already been reported in the literature [24]. From these results, it is believed that the red varieties Scarlotta and Crimson possessed the ability to transfer electrons as high as those of the black cultivars, Adora and Melody, whereas they exhibited lower capacity to donate hydrogen. Table 2 represents the volatile compounds extracted by SPME and detected by GC-MS in six seedless table grape varieties. Data are given as relative areas (i.e., absolute peak areas with respect to the sum of the total area of all detected compounds), expressed in %. Data are expressed as mean values (n = 3). As seen, all varieties exhibited a similar qualitative volatile profile, and only semiquantitative differences were established. Approximately, 15 volatile compounds were detected and the most representative belonged to short   [25], they were not identified in the varieties here studied. It is likely that they were included within the nonidentified compounds shown in Table 2, although it is also necessary to bear in mind that the occurrence of monoterpenes is known to depend on a number of factors including not only the cultivar, but also climate, region, soil, agricultural practices, etc. [9]. Figure 3 shows the volatile distribution in terms of relative areas (%) of alcohols, aldehydes and non-identified compounds with respect to the sum of total areas of all detected compounds. As observed, aldehydes were the major group in all varieties, on average more than 70% of the volatile fraction, followed by non-identified compounds of 20% and at lower levels by alcohols, which represented less than 10%. It is worthy to mention that Adora, and particularly Scarlotta, did not follow this general pattern. Both varieties exhibited lower aldehyde proportions (i.e., 50.03% for Adora and 42.10% for Scarlotta) together with higher percentages of alcohols (i.e., 21.03% for Adora and 34.25% for Scarlotta) than the rest of the varieties.
From these results, fruity and pleasant flavor, associated with aldehydes, would be expected in Autumn Crisp, Pristine, Crimson and Melody varieties, whereas herbaceous and woody notes, attributed to alcohols, would predominant in Scarlotta. On the other hand, the particular combination of aldehydes (i.e., 50.03%), alcohols (i.e., 21.03%) and nonidentified compounds (i.e., 15.66%) measured in Adora would result in a more unpredictable flavor as compared with the other varieties. According to Table 2, hexanal and 2-hexenal were by far the major constituents in all varieties except Scarlotta. Hexanal scent has been described to resemble freshly herbal, and it is frequently used to produce fruity flavors [26]. On the other hand, 2-hexenal is perceived differently depending on its concentration. Its odor is usually described as pleasant green-leafy, oily aroma with fruitygreen banana and apple pips nuances. However, it becomes even fruitier with fresh green apple notes when very diluted.
A sensory testing performed in the laboratory confirmed the flavory perceptions mentioned above. The tasting attributes were alike in Autumn Crisp, Pristine, Melody and Crimson, whereas they were much more discriminant in Scarlotta and, especially, Adora. In particular, Pristine exhibited a marked fruity character, resembling a combination of apple

Fig. 3
Volatile distribution in terms of relative areas (%) of alcohols, aldehydes and non-identified compounds with respect to the sum of total absolute areas of all detected compounds in seedless table grape varieties and tropical and citric odors, together with flowery secondary notes (honey). Autumn Crisp showed a mixture of floral tasting notes and apple-like odor, whereas Crimson displayed a very pleasant fruity aroma reminding of strawberry flavor with some slight citric character. Melody exhibited similar flavor to Crimson, but with a more intense tropical note, in particular pineapple-like aroma. It is worth highlighting the distinctive and characteristic flavors of Scarlotta and Adora. In opposition to the rest of the varieties, Scarlotta possessed a pronounced green flavor (a combination between herb, green olive and tea), whereas Adora displayed a complex mixture of spicy (cinnamon), sweet (honey), fruity (apple) and undefined tropical flavors. This particularly pleasant combination of flavors resulted in a unique taste which brought about the selection of Adora in terms of sensory attributes. It is also interesting that no unpleasant odor or taste was reported in the sensory testing. This suggests that the molecules responsible for the off-flavors did not reach the perceived threshold, thus not affecting the organoleptic characteristics of the grape varieties studied.

Conclusions
Among the seedless table grape varieties examined in this study, black varieties (Adora and Melody) possessed higher TPC, TAC and AA in terms of DPPH free radical scavenging activity and PCL antiradical activity than red (Scarlotta and Crimson) and white (Autumn Crisp and Pristine) grapes. A linear regression study indicated that, in general, TPC and TAC values correlated directly with AA. This suggests that phenolics and, particularly anthocyanins, were the main contributors to the antioxidant characteristics of grapes. Interestingly, the red varieties Scarlotta and Crimson did not exhibit this general trend since their TPCs and TACs did not reveal a linear dependence with DPPH activity data. By comparing the two black varieties studied, Adora had higher phenolic content and AA than Melody. The study of the volatile profile by GC-MS and sensory panel showed a unique and pleasant flavor for Adora as compared to the rest of the varieties. All in all, Adora was regarded as the highest-quality variety because of its antioxidant properties and sensorial attributes.