Biochemical composition and antioxidant activity of three extra virgin olive oils from the Irpinia Province, Southern Italy

Abstract Extra virgin olive oil (EVOO), appraised for its healthy properties, represents an important element for the economy of several countries of the Mediterranean area, including Italy. Our study aimed to evaluate some biochemical characteristics (polyphenols and volatile compounds) as well as the antioxidant activity of three EVOOs obtained from the varieties Ravece, Ogliarola, and Ruvea antica, grown in the same field of an Irpinian village, Montella, in the Campania region, Southern Italy. Extra virgin olive oil Ruvea antica contained the greatest amount of total polyphenols and showed the highest antioxidant activity. Principal component analysis of the aromatic profiles indicated that the three EVOOs could be easily discriminated according to the cultivar. 1‐Hexanol, 2‐hexen‐1‐ol, 3‐pentanone, representing the most abundant volatiles of the EVOO Ruvea antica, and 2‐hexenal, which resulted the main component in EVOOs Ogliarola and Ravece, could be considered as markers to discriminate these three EVOOs, according to the ReliefF feature selection algorithm.

. However, olives of the same variety, cultivated under different environmental conditions or in diverse geographical areas, can produce EVOOs with different organoleptic characteristics and healthy properties (Angerosa, Basti, Vito, & Lanza, 1999).
Concurrently, fruits from different cultivars grown under the same environmental conditions could produce oils with different biochemical characteristics (Gorzynik-Debicka et al., 2018). In the composition of EVOOs, volatile organic compounds (VOCs) and polyphenols are of great importance. Volatile organic compounds are strongly related to oil aroma perceived during the assay of the product (Salas, Harwood, & Martinez-Force, 2013). They are produced at the beginning of the malaxation, during cell structure rupture, due to enzymatic reactions in the presence of oxygen. C6 aldehydes, C6 alcohols, and their corresponding esters, together with smaller amounts of C5 carbonyl compounds, are the main constituents of VOCs (60%-80%). Specifically, hexan-1-ol, hexanal, E-2-hexenal, and 3-methylbutan-1-ol generally dominate the VOCs pattern of the most common EVOOs from Mediterranean regions . volatile organic compounds profile can depend on cultivar and on degree of maturation (Angerosa et al., 1999).
The Mediterranean diet is the golden standard for healthy nutrition. It is characterized mainly by a high intake of fruit, vegetables, and cereals, which are rich in phytochemicals . Among these compounds, polyphenols stimulated particular attention, due to their versatility of action, being able to protect against oxidative stress and to inhibit the proliferation of cancer cells (Del Rio, Costa, Lean, & Crozier, 2010). The beneficial effects of the Mediterranean diet are also attributed to the EVOO (Visioli & Bernardini, 2011), which, even if more expensive than olive oil, is richer in polyphenols, vitamins, phytosterols, etc., concurring to reduce the risk of cardiovascular events (Estruch et al., 2013), so that US Food and Drug Administration compared it to a real drug.
Extra virgin olive oil is rich in polyphenols ranging between 50 and 1,000 mg gallic acid equivalents (GAE)/kg of product (Gorzynik-Debicka et al., 2018). Oleuropein, quercetin, and hydroxytyrosol, some of the main polyphenols present in EVOO, have antioxidant activity and ascertained effects in protecting against the coronary artery disease (Manna et al., 2002) or cancer (Owen et al., 2000).
The aim of our work was to determine the biochemical composition of three EVOOs obtained from traditional varieties of olives cultivated in the same field of Montella, a little village of the Irpinia region, Southern Italy, harvested in the same period and processed by cold pressure. Three varieties, Ogliarola, Ravece, and Ruvea antica, in particular, attracted our attention. These are typical varieties of the Mediterranean area, diffused in Campania. Tree of Ogliarola has a medium foliage, with elliptical-lanceolate leaves. It produces a low number of flowers. Its fruits are black and elliptical, with a weight of 2-4 g. The endocarp has a weight of 0.3-0.45 g. Ravece tree has a high foliage density. Leaves are elliptical-lanceolate. Fruits are elongated, purple, and have a weight of 4-6 g; the endocarp is heavy (weight >0.45 g). Ruvea antica tree has medium foliage.
Leaves are elliptical-lanceolate and longer more than 7 cm. Its fruits are purple, elliptical, and show a weight of 2-4 g. The endocarp has a weight of 0.3-0.45 g (Di Vaio & Nocerino, 2012). The biochemical characterization of resulting EVOOs involved the total antioxidant activity and the polyphenol content. The polyphenolic profile and VOCs were also evaluated. Statistical analysis allowed us to correlate some of the biochemical characteristics of the EVOOs; in particular, the antioxidant activity was correlated with total polyphenols and the singular components, identified in the oil by UPLC. Principal component analysis (PCA) of the aromatic profiles (obtained by Gas Chromatography/Mass Spectrometry) was carried out to discriminate oil samples according to cultivar. Moreover, a feature selection algorithm was used to identify and select putative volatile markers responsible for EVOO varieties discrimination.
Ultrapure water from a Milli-Q system (Millipore) with a resistivity at 25°C of 18 MΩ * cm was used throughout the analyses. Helium (Rivoira) at a purity of 99.999% was the GC carrier gas. The SPME glass vials and the fibers were from Supelco; the capillary GC-MS column HP-Innowax (30 m × 0.25 mm × 0.5 μm) was purchased from Agilent J&W (Agilent Technologies Inc.).

| Plant material
The EVOOs used in this study were produced in the same year by cold pressing of three different varieties (Ruvea antica, Ogliarola, and Ravece) grown in the same field located in the Montella village, in the Irpinia Province, Campania region, Southern Italy. Prof. Vincenzo De Feo identified the varieties. Voucher specimens of the three varieties were stored in the herbarium of the Department of Pharmacy, University of Salerno.

| Polyphenol analysis and free radical scavenging capacity
To isolate the phenolic fraction of the three EVOOs, 1.5 g of sample was mixed with 1.5 ml of hexane and charged onto cartridges SPE C 18 . Polyphenols were eluted through 3 ml of methanol 100% and recovered; this step was repeated other two times. The three residues were collected, grouped, dried, and re-suspended with 1 ml of methanol. The samples were filtered (mesh = 0.20 μm). The method of Singleton and Rossi (Singleton & Rossi, 1965) was used to evaluate the content of total polyphenols present in the three EVOO samples. Quantification was determined by using gallic acid as standard and reading the absorbance at 760 nm through a Cary UV/Vis spectrophotometer (Varian). Results were expressed as μg gallic acid equivalent (GAE)/g of EVOO ± standard deviation (SD).
The scavenging activity was expressed as effectiveness (%) of the sample to inhibit DPPH radical activity during a 60-min incubation.
Polyphenol profile was determined through UPLC (ultra highperformance liquid chromatography) by using an ACQUITY Ultra Performance system linked to a PDA 2996 photodiode array detector (Waters), setting the UV detection wavelength at 280 nm, following the method of Fratianni and coworkers .
Quantification of known components was performed by comparing the peak areas on the chromatograms of samples with those obtained from standard solutions.

| Analysis of VOCs profiles
The optimization of SPME parameters was achieved by examining samples of a commercial EVOO bought at a local supermarket. SPME GC-MS volatile analysis was accomplished according to Romero and coworkers (Romero, Garcıa-Gonzalez, Aparicio-Ruiz, & Morales, 2015), but using the DVB/CAR/PDMS (50/30 μm) fiber. For the sample preparation, 2 g of each sample was put into a 20-mL headspace vial with screw cap (Supelco) and 4-methyl-2-pentanol to a final concentration of 1.5 mg/g was added as an internal standard to guarantee the analytical reproducibility. Subsequently, vials, closed with a Teflon (PTFE) septum and an aluminum cap (Chromacol) and stirred, were put in the instrument dry block heater and held at 40°C for 10 min. After the equilibration time, the extraction and injection processes were automatically carried out using an autosampler MPS 2 (Gerstel).
Volatiles were analyzed by gas chromatography-quadrupole mass spectrometry (GC-qMS), introducing the SPME fiber into the injector port of the gas chromatographer, model GC 7890A, Agilent hyphenated with a mass spectrometer 5975C. Once desorbed, metabolites were directly transferred to the capillary column HP-Innowax for the analysis. The oven temperature program was initially set at 40°C for 3 min, increased to 200°C at 30°C/min, and then ramped to 240°C at 30°C/min, holding for 1 min. Volatiles were investigated according to the instrumental parameters as reported in the literature (Cozzolino, Martignetti, et al., 2016;Cozzolino, Pace, et al., 2016). Each sample was analyzed in duplicate in a randomized sequence where blanks were also run.
Volatile metabolites recorded in the headspace of the extra virgin olive oils under study were identified by three diverse methods, as previously reported (Cozzolino, Martignetti, et al., 2016;Cozzolino, Pace, et al., 2016). The areas of the identified volatiles were determined from the total ion current (TIC), and the semiquantitative data of each metabolite (Relative Peak Area, RPA%) were considered in relation to the area of the peak of 4-methyl-2pentanol, used as internal standard.

| Statistical analysis
Data were expressed as the mean ± standard deviation (SD) of triplicate measurements, and antioxidant activity was correlated with polyphenols. As concerns VOCs, analysis of variance (ANOVA) was used to compare results and significance was accepted at p < .05.
Principal component analysis (PCA) was then used to relate the obtained values and as an explorative tool for the preliminary visualization of the separation of the different EVOO samples, according to their VOCs profiles. Last, the ReliefF (Kononenko, Simec, & Robnik-Sikonja, 1997) feature selection algorithm was used to identify potential markers, among VOCs, responsible for EVOO discrimination.

| Total polyphenol content and antioxidant activity
The analysis of total polyphenols (TPF, Table  TA B L E 1 Total polyphenols (expressed as μg GAE/g of EVOO ± SD) and antioxidant activity (evaluated through the DPPH and expressed as percentage ± SD) of the three polyphenolic extracts from Ogliarola, Ravece, and Ruvea antica EVOOs

| Polyphenol profile
The amount (expressed as μg GAE/g of EVOO) of polyphenols identified through UPLC analysis is shown in representing an abundant polyphenol in Ogliarola and Ruvea antica, was found at concentrations much lower in EVOO Ravece (9.30 μg GAE/g), the 5.93% of the total polyphenols. This molecule is an ester of hydroxytyrosol; it gives rise from the mevalonic acid pathway (Omar, 2010 and Ruvea antica (9.77 and 5.09 μg GAE/g, respectively), but not in Ogliarola. The polyphenols identified in the three EVOOs are well known highly bioavailable molecules. The presence of high amounts of oleuropein, whose absorption in the body is about 55%-60% (Omar, 2010), is very significant, given the numerous and key effects of such metabolite including antioxidant, anti-inflammatory, anticancer, antiatherogenic activities, and cardioprotective, antihyschaemic and hypolipidemic properties (Visioli & Galli, 2002). Concomitantly, the high content of quercetin contributes to improve the biological value of the three EVOOs. The amount of quercetin and its derivative spiraeoside in the EVOOs Ogliarola and Ruvea antica represented the 51.74% and 35.04%, respectively, of the polyphenols.
A so high amount of these compounds is certainly essential: Like other flavonoids, they can affect the cellular function, by mediating gene expression and signal transduction rather than through a direct antioxidant effect (Nemeth et al., 2003). Dietary quercetin and other flavonoids are absorbed by a little percentage (5%-10%) in the small intestine; the residue of these molecules moves to the colon, where they are metabolized by the gut microbiota, influencing its composition. These molecules exert potential prebiotic effect, protecting from intestinal dysbiosis and all alterations interesting microbiota, and finally, they can concur to significantly influence host biochemistry and host susceptibility to diseases (Nazzaro, Fratianni, d'Acierno, & Coppola, 2013;Tamura et al., 2017).
Considering the almost complete linearity between the total polyphenol content and the antioxidant activity (corr = 99.9, Figure 1    . The analysis was performed taking into consideration the most abundant molecules present in the three extracts, which resulted, by the UPLC analysis, quercetin, oleuropein, spiraeoside, formononetin, naringenin, and luteolin. The results are shown in Figure 2. Naringenin, present in all three polyphenolic extracts, at amounts ranging between 7.90 and 21.05 μg GAE/g EVOO, did not seem to affect the antioxidant activity in marked way until 10.31 μg GAE/g. Its effect seemed stronger upper such threshold, so that, at twice amounts, a doubling of the antioxidant activity was observed (Figure 2a). Oleuropein appeared to exhibit a linear behavior, with an antioxidant activity growing concurrently to its amounts (corr = 88.75, Figure 2b). This molecule seemed to be the main responsible for the antioxidant activity exhibited by the three polyphenol extracts, although it did not represent the most abundant molecule. Therefore, the noticeable antioxidant activity of oleuropein is reported, mainly as a scavenger of chain-propagating lipid peroxyl radicals within the membranes (Saija et al., 1998 which decreased from 34% to 20.7% (Figure 2c). A similar behavior could be also attributed to luteolin (Figure 2d), which amounts in the three extracts ranged between zero and 9.77 μg GAE/g. Like quercetin, it apparently exerted an antioxidant activity (34%) until a specific threshold (that could be ascribable to 5.09 μg GAE/g); after which, increasing its content until 9.77 μg GAE/g, the antioxidant activity decreased from 34% to 16%. Formononetin (corr = −37.98) showed a variable trend so that increasing its amount until a certain percentage (7.6%) the antioxidant activity decreased, increasing again as the molecule's content increased (Figure 2e). Our results corroborated the hypothesis that a certain bioactive compound can modify its properties in the presence of other compounds. In the case of spiraeoside, for instance, it is possible that its influence on the antioxidant activity can be negligible (Figure 2f), although its presence in the extracts of EVOOs Ruvea antica and Ogliarola was practically the same.

| Volatile compounds analysis
The analysis of EVOO volatile compounds was performed through the SPME sampling followed by GC-MS (Torri, Sinelli, & Limbo, 2010). SPME, as an alternative technique for fractionation of volatiles from interfering non-volatile matrix compounds, is a pre-concentration technology, which integrates sample extraction, concentration, and sample introduction into a single solvent-free step, preventing the production of artifacts compared with conventional solvent extraction procedures (Pawliszyn, 2012). A total of 49 VOCs were identified, which belonged to hydrocarbons (3), aldehydes (11), alcohols (12), ketones (5), esters (5), carboxylic acids (6) conditions. SPME GC-MS semiquantitative data, calculated as the percent ratio of the respective peak area relative to the peak area of 4-methyl-2-pentanol, used as internal standard, were subject to a one-way ANOVA, in order to investigate the effect of cultivar on the identified VOCs. Table 3   . This compound provides the typical "green note" of olive oil and has been reported to be negatively correlated with the maturity and degree of oxidation of virgin olive oils (Pouliarekou et al., 2011). Alcohols were the most abundant volatiles present in the EVOO of Ruvea antica, representing the 58.2% of the total VOCs. The principal alcohols were (E) -hexen-1-ol (37.4%) and 1-hexanol (15%), both deriving from the LOX pathway and showing a characteristic odor described as green, grassy, leafy. These compounds, on the other hand, were present only at lower concentration in Ogliarola (1.2% and 1.8%, respectively) and in Ravece EVOOs (4.2% and 14.1%, respectively).
The variety could strongly affect the abundance of volatile compounds, which in turn have revealed to be extremely valuable as varietal markers (Kalua et al., 2005). For this reason, the volatile profiles of the three EVOOs were subjected to multivariate statistical analysis with the aim to build models able to explain the variations of the metabolic content dependently from genotype, and to identify putative volatile markers useful for cultivar discrimination.

| CON CLUS ION
Data obtained in this research clearly confirm the influence of genetic and environmental factors in determining the organoleptic properties of olive oil and permitted a distinction of the three EVOOs studied on the basis of their volatile constituents.

ACK N OWLED G M ENTS
This work was funded within the Project "SALVE," PSR 2007-2013, mis. 214, action f2, by the Campania Regional Council, Italy.

CO N FLI C T O F I NTE R E S T
Authors declare that they do not have any conflict of interests.

E TH I C A L A PPROVA L
Human and animal testing was unnecessary for this study.