**1. Introduction**

Extra virgin olive oil (EVOO), obtained from the fruit of *Olea europea* L. using only physical or mechanical methods, is a product of grea<sup>t</sup> importance because of its exclusive chemical-nutritional characteristics, and health properties. EVOO is one of the key ingredients in the Mediterranean diet [1].

The production of olive oils from monovarietal olives is carried out to produce oils with particular chemical composition and unique organoleptic properties, which depend on cultivar, geographic origin, and pedoclimatic conditions [2].

The check of the cultivars used to obtain an olive oil may contribute to highlight the oil origin. This aspect may have commercial interest in the case of monovarietal high-quality EVOO with typical marks (protected designation of origin-PDO, protected geographical indication-PGI, traditional specialty guaranteed-TSG), because these oils have high commercial value and may be adulterated by lower quality oils, using anonymous or less expensive cultivars [3].

There is an increasing need for developing appropriate methodologies in order to guarantee food traceability [4], as well as to identify geographical origin [5] or cultivar [6].

For EVOO traceability, several analytical approaches, from chromatographic to nondestructive spectroscopy [7,8] have been reported, together with DNA based methods or electrochemical devices [9,10]. Studies of authenticity have been reported for the classification of olive oils according to their botanical or geographical origin, based on determination of fatty acid (FA) profile or minor constituents, as phytosterols, phenols, or volatiles [11–13]. In addition, the classification was performed using a simultaneous combination of two or more components; for example nuclear magnetic resonance (NMR) [14], Fourier Transform Infra-Red spectroscopy [15], and stable isotopic techniques [16,17] have been used.

The differentiation of EVOO samples according to variety and geographical origin has been recently addressed by comprehensive two-dimensional gas chromatography [18] and by ultra-high-pressure liquid chromatography (UHPLC) coupled to an electrospray quadrupole–time-of-flight hybrid mass spectrometer (ESI/QTOF-MS) [19,20]. Omics-based massive molecular tools can help to circumvent limitations of traditional methodologies, and therefore genomics, proteomics and metabolomics-based methods are being developed for the authentication of a wide range of food commodities [21].

In this research, the characterization of olive oil varieties was performed initially by triacylglycerol (TAG) stereospecific analysis. Afterwards, volatile analysis by solid-phase microextraction gas chromatography-mass spectrometry (SPME-GC–MS) was carried out.

The objectives of this paper have been: (i) To study the TAG fraction of monocultivar EVOO, for total and positional FA compositions, resulting from the specificity of biosynthetic enzymes and correlated to the nutritional aspects; (ii) to obtain the qualitative and quantitative profile of EVOO volatile fraction, also depending on the enzymatic pool, directly related to genetic characteristics; and (iii) to investigate and compare the potential of stereospecific analysis of TAG and of volatile profile, combined with chemometric data analysis, to classify four monovarietal Italian EVOO (Dolce Agogia, Frantoio, Leccino, and Moraiolo) on the basis of varietal origin.

#### **2. Materials and Methods**

#### *2.1. Materials and Chemicals*

Acetone, diethyl ether, hydrochloric acid, formic acid, methanol, and petroleum ether were obtained from J.T. Baker B.V. (Deventer, the Netherlands). Anhydrous sodium sulfate, chloroform, ethanol, hexane, and potassium hydroxide were purchased from Carlo Erba Reagents (Milan, Italy). Deionized water was from a Milli-Q SP Reagent Water System (Bedford, MA, USA). Supelco ™ 37 component fatty acid methyl esters (FAME) mix (catalog n◦ 47885-U), containing the methyl esters of 37 fatty acids (the FA contents ranged between 2% and 4%, while the palmitic acid methyl ester was 6%), was bought from Supelco (Bellefonte, PA, USA). Divinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS) fiber, lipase from porcine pancreas (EC 3.1.1.3), and *sn*-1,2-diacylglycerol kinase from *Escherichia coli* (DAGK; EC 2.7.1.107) were acquired from Sigma-Aldrich (St. Louis, MO, USA).

## *2.2. EVOO Samples*

Sixteen EVOO samples from four different *O. European* cultivars (Dolce Agogia, Frantoio, Leccino, and Moraiolo), typical of Central Italy, were analyzed. Monovarietal bottled EVOO samples were purchased in 2016 from local producers, which guaranteed their origin and cultivar. The monovarietal EVOO samples were stored in the dark at 8 ◦C.

#### *2.3. Purification of TAG Fraction from EVOO Samples*

The TAG fraction was isolated from monovarietal EVOO samples by thin layer chromatography as reported in a previous paper [22].

#### *2.4. Stereospecific Analysis of TAG Fraction from EVOO Samples*

The stereospecific analysis procedure [23] was performed on purified TAG of EVOO samples, and the following steps were carried out:

(a) Hydrolysis by pancreatic lipase to obtain *sn*-2-monoacylglycerols and then the FA percent positional composition of TAG *sn*-2 position;

#### *Foods* **2019**, *8*, 58

(b) TAG deacylation by Grignard reagen<sup>t</sup> to obtain *sn*-1,3/*sn*-1,2(2,3)-diacylglycerols, followed by the DAGK enzymatic reaction in order to obtain the *sn*-1,2-phosphatidic acids and then the FA percent positional composition of TAG *sn*-1 and *sn*-2 positions.

#### *2.5. Preparation of Methyl Esters of Constituent Fatty Acids and Analysis*

Fatty acid methyl esters (FAME) were prepared by transesterification, as previously reported [24] and analyzed by high resolution gas chromatography (HRGC). A DANI 1000DPC gas-chromatograph (Norwalk, CT, USA) provided with a split–splitless injector and a flame ionization detector (FID) was used. A fused silica capillary column, named CP-Select CB for FAME (50 m × 0.25 mm i.d., 0.25 μm f.t.; Varian, Superchrom, Milan, Italy), was used for the chromatographic separation. The injector and detector temperature was 250 ◦C. The initial oven temperature, 180 ◦C, was held for 6 min, raised at 3 ◦C/min to 250 ◦C, and finally maintained for 10 min. Carrier gas was helium with flow rate of 1 mL/min; the injection volume was 1 μL with a split ratio of 1:70. To identify the FA, the standard mixture containing 37 FAME was used. The percentage of each FA was calculated using the peak area of the samples. The chromatograms were acquired and processed using Clarity integration software (DataApex Ltd., Prague, Czech Republic). The data were normalized considering only the main reported FA (% mol mean values ≥0.1).

#### *2.6. Analysis of Volatile Fraction*

Volatiles have been analyzed by SPME-GC–MS as reported in a previous paper [25]. The heated samples were placed in a 25 ◦C water bath and the DVB/CAR/PDMS fiber, previously cleaned for 15 min at 250 ◦C, was exposed to the sample headspace for 15 min.

Volatile compounds were analyzed with a Hewlett-Packard 5890 series II gas chromatograph (Palo Alto, CA, USA) equipped with a split–splitless injector, an Econo-Cap EC-5 capillary column (30 m × 0.25 mm i.d., 0.25μm f.t.; Alltech, Milan, Italy), and a 5971A quadrupole MS detector (Palo Alto, CA, USA).

Volatiles were desorbed from the SPME fiber at 250 ◦C for 5 min into the injector port, in splitless mode. The oven initial temperature, 35 ◦C, was maintained for 5 min, then the temperature was raised at 3 ◦C/min to 230 ◦C, and finally maintained for 5 min.

The mass spectrometer operated in electron impact mode with electron energy of 70 eV, and scanned, in full scan acquisition mode, in the mass range 35–500 m/z at 1.2 scans/s. The temperatures of interface and ion source were 280 and 180 ◦C, respectively. Data were collected by HP G1030 MS ChemStation (Hewlett-Packard). Compounds were identified by comparing their mass spectra with those reported in Wiley138 mass spectral library and using the linear retention indexes from literature [26,27]. A semi-quantitative analysis was performed.
