Effect of light exposure on the quality of extra virgin olive oils according to their chemical composition

Accepted Manuscript Effect of light exposure on the quality of extra virgin olive oils according to their chemical composition Esposto Sonia, Taticchi Agnese, Urbani Stefania, Selvaggini Roberto, Veneziani Gianluca, Di Maio Ilona, Sordini Beatrice, Servili Maurizio PII: DOI: Reference:

S0308-8146(17)30344-8 http://dx.doi.org/10.1016/j.foodchem.2017.02.151 FOCH 20706

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

18 August 2016 13 February 2017 27 February 2017

Please cite this article as: Sonia, E., Agnese, T., Stefania, U., Roberto, S., Gianluca, V., Ilona, D.M., Beatrice, S., Maurizio, S., Effect of light exposure on the quality of extra virgin olive oils according to their chemical composition, Food Chemistry (2017), doi: http://dx.doi.org/10.1016/j.foodchem.2017.02.151

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effect of light exposure on the quality of extra virgin olive oils according to their chemical composition Esposto Sonia, Taticchi Agnese, Urbani Stefania, Selvaggini Roberto, Veneziani Gianluca, Di Maio Ilona, Sordini Beatrice, and Servili Maurizio

Department of Agricultural, Food and Environmental Sciences, University of Perugia, Via San Costanzo s.n.c., 06126, Perugia, Italy

*Corresponding author. Phone: +39 075 5857952, [email protected] (S. Esposto)

Running title: Light-exposure effect on olive oil quality

Fax: +39 075 5857916, E-mail:

Abstract The influence of light exposure on the quality of commercially available extra-virgin olive oils (EVOOs) of different chemical composition was studied as a function of storage (11 weeks) under conditions simulating market storage. By mildly stripping the polyphenols from oil ‘A’, with high levels of polyphenols and oleic acid, and oil ‘B’, exhibiting a medium level of polyphenols and a low level of oleic acid, ‘C’ and ‘D’ EVOOs were obtained. Ten EVOOs were produced as mixtures of these four oils. The initial concentrations of oleic acid and polyphenols in the 14 oils ranged from 64.5–77.7% and 18.1–1476.7 mg/kg, respectively. The extinction coefficient K270 well reflected the EVOO product quality. EVOOs richer in oleuropein derivatives showed superior oxidative stability, which resulted in lower off-flavour volatile compound production and α-tocopherol and polyphenols losses, and thus, higher EVOO sensory and health benefits.

Keywords: Storage; Exposure to light; Tocopherols; Phenolic compounds; C7-C11 aldehydes; K270

1. Introduction Extra-virgin olive oil (EVOO) is one of the most important components of the Mediterranean diet (Estruch et al., 2006; Konstantinidou et al., 2010). Compared to other vegetable oils, it is characteristically richer in monounsaturated fatty acids (MUFAs) and antioxidants, mainly secoiridoid derivatives and volatile substances, which confer the exclusive sensory and health properties to EVOOs (Angerosa, Servili, Selvaggini, Taticchi, Esposto, & Montedoro, 2004; Terés et al., 2008; Servili et al., 2009; Bulotta, Celano, Lepore, Montalcini, Pujia, & Russo, 2014). Oxidation decreases the sensory and health-promoting qualities as well as the marketing value and consumer acceptability of an EVOO because it leads to the generation of low-molecularweight off-flavour substances, loss of antioxidants, and accumulation of toxic compounds such as free radicals (Choe & Min, 2005; 2006; 2009). The predisposition of an EVOO to these negative phenomena depends on its exposure to pro-oxidant factors such as oxygen, temperature, light, and other activators (chlorophylls and transition metals) (Khan, 1955; Choe & Min, 2006, Bendini, Cerretani, Salvador, Fregapane, & Lercker, 2010), which can affect its oxidative stability from production to consumption. However, the response of an EVOO to oxidation depends on its chemical composition, particularly, on those substances that can directly or indirectly influence these phenomena, such as natural antioxidants and oleic acid, respectively (Choe & Min, 2005; 2006; 2009, Bendini et al., 2010). Because the levels of these two chemical groups are highly influenced by genetic, agronomic, and technological factors, EVOOs typically widely vary in oleic acid and antioxidant contents. A recent study on 740 commercial EVOOs differing in terms of olive cultivar used, geographic origin, and mechanical extraction method used revealed ranges of 50.5– 80.5% for oleic acid, 91–665 mg/kg for α-tocopherol, and 50–900 mg/kg for secoiridoid derivatives and lignans (Servili et al., 2015). Generally, EVOO stability ranges roughly from 9 to 18 months, based on the mentioned internal (chemical composition) and external factors (presence of pro-oxidants). Furthermore, EVOO shelf life appears quite unsteady and is often compromised in marketplaces, where long

exposure to light can substantially reduce EVOO quality (Choe & Min, 2006; Bendini, Cerretani, Salvador, Fregapane & Lercker, 2012). Various studies have evaluated EVOO oxidative stability during shelf storage; however, in many cases, they involved laboratory tests, which do not consider several parameters that affect EVOO market storage (Rahmani & Saari Csallany, 1998; Psomiadou & Tsimidou, 2002a, 2002b; Luna, Morales, & Aparicio, 2006; Stefanoudaki, Williams, & Harwood, 2010; Nishida, Yamashita, Miki, 2007; Hachicha Hbaieb, Kotti, Gargouri, Msallem, & Vichi, 2016). Fast assays, such as the AOM, Rancimat, DPPH, and OSI tests, only allow determining the capability of EVOOs to resist drastic oxidative stresses (Gutfinger, 1981; Baldioli et al., 1996; Gòmez-Alonso, Salvador, & Fregapane, 2002; Carrasco-Pancorbo, Cerretani, Bendini, Segura-Carretero, Lercker, & Fernández-Gutiérrez, 2005; Servili et al., 2009; Mancebo-Campos, Desamparados, Salvador, & Fregapane, 2014; Condelli, Caruso, Galgano, Russo, Milella, & Favati 2015); hence, these studies cannot appropriately assess the stresses occurring in bottled EVOOs in the market. There are currently very few reports of in-depth studies on the evaluation of quality parameters and shelf life of commercial, bottled EVOOs, likely because such studies are timeconsuming and involve multiple analyses, and numerous EVOOs with highly variable composition are available (Pagliarini, Zanoni, & Giovanelli, 2000; Cinquanta, Esti, & Di Matteo, 2001; Morello, Motilva, Tovar, & Paz Romero, 2004; Caponio, Bilancia, Pasqualone, Sikorska, & Gomes, 2005; Kalua, Bedgood, Bishop, & Prenzler, 2006; Gutiérrez & Fernandez, 2002; Okogeri & TasioulaMargari, 2002, Del Nobile, Bove, La Notte, Sacchi, 2003; Coutelieris & Kanavouras, 2006; GõmezAlonso, Mancebo-Campos, Desamparados Salvador, & Fregapane 2007; Mendez & Falque, 2007; Pristouri, Badeka, & Kontominas, 2010). To the best of our knowledge, research evaluating the quality of EVOOs in terms of oleic acid percentage and polyphenol contents during storage simulating supermarket conditions has not been performed to date. Here, we report a time-course study of the effect of storage at room temperature for 6 months under 11 h of light exposure per day on parameters related to product, health-promoting, and sensory qualities of EVOOs with different chemical compositions.

2. Materials and methods 2.1. Materials Ethyl acetate, anhydrous sodium sulphate, methanol, ethanol, n-hexane, 2-propanol, and acetic acid were purchased from Sigma-Aldrich (Milan, Italy), and high-performance liquid chromatography (HPLC)-grade methanol and water for HPLC-MS were purchased from Fluka (Milan, Italy). Two different EVOOs with high and medium polyphenol levels and high and low oleic acid percentages (named ‘A’ and ‘B’, respectively) were obtained from bulk suppliers. The oils were analysed upon receipt to determine the initial quality and to provide a quality baseline for the experiments. The initial polyphenol contents of A and B were 1476.7 and 682.5 mg/kg, respectively. These concentrations represented the sum of the following compounds: hydroxytyrosol (3,4-DHPEA), tyrosol (p-HPEA), secoiridoid derivatives, such as the dyaldeidic form of the elenolic acid linked to 3,4-DHPEA (3,4-DHPEA-EDA), the oleuropein aglycon (3,4 DHPEA-EA), and the aldeydic form of elenolic acid linked to p-HPEA (p-HPEA-EDA), and the lignans (+)-1-acetoxypinoresinol and (+)-pinoresinol. The initial oleic acid percentages for sample A and sample B were 64.8 % and 77.7%, respectively. The α-tocopherol contents were 173.8 mg/kg for sample A and 220 mg/kg for sample B. The chlorophyll contents of EVOOs A and B were 20 and 19.8 mg/kg, respectively. Polyphenol stripping was conducted by repeating the following procedure thrice: oil and water (1:1, v/v) were immediately mixed by vortexing for 3 min. The mixture was centrifuged in a basket centrifuge at 352 G-force for 8 min. The oily phase of the supernatant was recovered and filtered with sodium sulphate to remove any trace water. By this procedure, two different EVOOs, C and D, were obtained from A and B, respectively. The entire process was performed with as little air contact as possible.

A statistical central composite design (CCD) was built with samples A, B, C, and D, which were placed at the vertices of the square as starting points. By mixing of the samples in different percentage ratios, 10 EVOOs characterized by intermediate contents of oleic acid and hydrophilic phenols were obtained. The 10 oils obtained by applying the CCD were named: ‘C+A’, ‘D+B’, ‘B+A’, ‘D+C’, ‘A+D’, ‘C+B’, ‘DC+DB’, ‘CA+BA’, ‘DB+BA’, and ‘DC+CA’, where the two samples that composed each new EVOO were each present at 50% of the volume.

2.2. Experimental set-up for market storage simulation Real marketplace storage conditions were simulated as follows. Eleven green-glass 750-mL bottles of each of the 14 EVOOs (oils A, B, C, D, C+A, D+B, B+A, D+C, A+D, C+B, DC+DB, CA+BA, DC+CA, and DB+BA) were placed in a climate chamber in 14 rows. The room temperature was set at 22 °C, and the bottles were exposed to 12 h of light (600 lx) per day by an automatic lighting system. The bottles were moved weekly from the first to the last position in the same row to ensure equal light exposure over the experimental period of 165 days. Every 15 days, the first bottle of each row was withdrawn and stored at 12 °C until analysis. Thus, the bottles were withdrawn from the climate chamber at times T0, T15, T30, T45, T60, T75, T90, T105, T120, T135, T150, and T165.

2.3. Analytical determinations All determinations were done for each of the 14 EVOOs at all 12 time points mentioned above, unless mentioned otherwise.

2.3.1. Merchandise parameters The free acidity content (g of oleic acid/100 g of oil), peroxide values (amount of hydroperoxides expressed as milli-equivalents of O2/kg), K232 and K270 extinction coefficients, and fatty acid

compositions of all the EVOO samples were determined according to the official methods of the European Commission (Commission Delegated Regulation (EU) 2015/1830).

2.3.2. Initial chlorophyll content determination The chlorophyll contents were analysed as described by Pokorny, Kalinova, & Dysseler (1998).

2.3.3. α-Tocopherol and phenolic compound determination The α-tocopherol contents were evaluated by HPLC with diode array and fluorescence detectors (HPLC-DAD-FLD; Agilent Technologies, Santa Clara, CA, USA) according to Esposto et al. (2015). Phenols were extracted according to the procedure reported by Esposto et al. (2013) and evaluated by HPLC-DAD (Agilent Technologies, Santa Clara, CA, USA) analysis.

2.3.4. Determination of the phenolic oxidation products Phenolic oxidation products were analysed in all EVOOs at T0, T30, T60, T90, T120, and T150 by HPLC/electrospray ionization tandem mass spectrometry (HPLC-ESI-MS) and MS/MS (Agilent Technologies, Santa Clara, CA, USA) according to Di Maio et al. (2013).

2.3.5. Determination of the volatile compounds The compositions of the volatile compounds of all EVOO head spaces were assessed using headspace solid-phase micro-extraction followed by gas chromatography/mass spectrometry (HSSPME-GC/MS; Agilent Technologies, Santa Clara, CA, USA) as described by Esposto et al. (2013).

2.4. Statistical analysis A CCD for two factors was created using the Modde 9.1 package, whereas principal component analysis (PCA) and partial least squares (PLS) regression were carried out using the SIMCA 13.0

chemometric package. Both packages were purchased from Umetrics AB (Umeå, Sweden). A priori one-way analysis of variance (ANOVA) followed by the Tukey test was conducted using the SigmaPlot software package, version 12.3 (Systat Software Inc., San Jose, California).

3. Results & discussion 3.1. Initial EVOO composition A first exploration of the qualitative parameters, including free fatty acid percentage and indicators of primary (PV and K232) and secondary (K270) oxidation products, of the 14 EVOOs showed that all samples were well below the upper legal limit values according to the current EU Regulation (2015/1830) for these parameters (data not shown). The chlorophyll contents of oils A and B were 6.1 and 7.3 mg/kg, respectively. Given the lipophilic character of this molecule, this value remained stable after polyphenol stripping, in both C and D. As expected, in the other 10 oils obtained by mixing A, B, C and D, the chlorophyll content ranged between these two values (data not shown). Regarding their initial acidic and phenolic compositions (Table 1), the EVOOs showed substantial differences, in accordance with the wide variability reported by Servili et al. (2015), who examined more than 740 commercial EVOOs and reported a range of 50.5–80.5% for oleic acid, with a mean of 70.4%, and a range of 187–977 mg/kg for hydrophilic phenols (oleuropein, ligstroside derivatives, and lignans), with a mean of 536.6 mg/kg.

3.2. EVOO quality evaluation during light exposure The effect of light exposure, the most important factor influencing EVOO shelf-life quality during supermarket storage, was evaluated by determining legal quality and health-promoting as well as sensorial characteristics at different moments during storage for 165 days. In particular, acidity, peroxide value (PV), extinction coefficients K232 and K270, and acidic composition were considered for product quality, whereas health-promoting property was assessed by measuring all phenolic compounds and oxidative products of secoiridoid derivatives. Furthermore, secoiridoid

compounds, which are responsible for the bitter and pungent sensations, as well as sensory properties were evaluated by measuring head-space volatile compounds. The results were assessed by different multivariate statistical methods. Initially, PCA was conducted to observe object dispositions in a two-dimensional space (data not shown). Briefly, that model, explaining 81 % of total variance, showed a distribution of EVOOs according to period of storage. The chemical compounds with the highest loading were secoiridoid derivatives. Next, partial least square (PLS) of latent variables analysis was performed to identify a potential correlation between the evolution of EVOO chemical composition and light-exposure duration. Three latent significant variables explained 92% of the total variance. The score plot of the first latent variable showed a narrow distribution of EVOOs containing oil D, which was located to the right of the plot, and EVOOs containing oil A were located to the left, in agreement with their peculiar chemical compositions, particularly, the lower concentrations of hydrophilic phenols (secoiridoid derivatives) for the former and the higher concentrations of secoiridoid derivatives for the latter (Fig. 1). The 14 oils were widely distributed according to the duration of light exposure, with the top of the score plot occupied by the EVOOs with the longest period of light exposure and the bottom by those exposed for 0 to 60 days. The relative loading plot (Fig. 1) showed that the latent dependent variable Y, representing light-exposure duration, displayed positive correlations, particularly with the extinction coefficient K270, C7-C11 aldehydes (including (E)-2-heptenal, (E, E)-2,4-heptadienal, (E, E)-2,4nonadienal, 2-undecenal, and 2,4-(E, E)-decadienal), and the oxidative products of the secoiridoid derivatives (the oxidized form of elenolic acid, ODFEA). In contrast, the latent variable Y showed negative correlations with α-tocopherol and all hydrophilic phenols, particularly the oleuropein derivatives 3,4-DHPEA-EDA and 3,4-DHPEA-EA. However, fatty acids and K232 did not correlate with the latent variable Y. Based on this first exploration, we concluded that among the several chemical compounds analysed during the entire shelf life of the 14 EVOOs, PV, K270, the antioxidant fractions α-tocopherol and secoiridoid derivatives as well as ODFEA, and C7-C11

aldehydes were highly influenced by storage duration; hence, each of these chemical parameters was investigated in more detail.

3.3. Evolution of legal product parameters The quality indices PV and K270 increased linearly during the 165-day storage (Table 2). The time ranges in which EVOOs exceeded the upper limits according to EU marketing standards are indicated in Table 2 (EU Reg. 2015/1830). Among the 14 EVOOs analysed, the PVs of oils C, D, D+B, and DC+DB exceeded the upper limit (20 meq. O2/kg) established by the European Regulation (EU Reg. 2015/1830) for the ‘extra virgin’ category at various time points during the 165 days of storage. This result suggested the importance of antioxidants in the oxidative stability of EVOOs; the lack of antioxidants in almost all oil mixtures containing D probably caused the rapid decrease in their relative capacity to inhibit oxidative phenomena. With respect to the K270 index, C, D, DC+DB, and D+B could strictly not be categorized as EVOOs according to the current UE limit of 0.22 (EU Reg. 2015/1830) at approximately 30 days before the time points determined by the PV analyses. It is well known that this extinction coefficient indicates the presence of volatile compounds originating from hydroperoxide decomposition during the last phase of oxidation, some of which absorb UV radiation at 270 nm (Gutierrez and Fernandes, 2002; Caponio, Bilancia, Pasqualone, Sikorska, Gomes, 2005; Kalua, Bedgood, Bishop, & Prenzler, 2006; GómezAlonso et al., 2007); these compounds are responsible for the rancid defect (Aparicio, Morales, & Alonso, 1996). The duration of light exposure and the relative K270 were strongly correlated as indicated by the high R2 value, suggesting that K270 should be useful for evaluating EVOO shelf life, based on our observations over a storage period of less than 165 days. In contrast, PV and light exposure duration were less well correlated (R2 < 0.80), indicating the lower reliability of PV as a potential predictor of EVOO freshness. This finding was in accordance with those of previous studies (Gutierrez and Fernandes, 2002; Caponio, Bilancia, Pasqualone, Sikorska, Gomes, 2005; Kalua, Bedgood, Bishop, & Prenzler, 2006; Gómez-Alonso et al., 2007) that confirmed the

importance of the K270 index for monitoring product quality during EVOO storage. It was also observed that, except for oil D+C, the four samples for which this parameter exceeded the limit at various time points all contained D, and thus, were characterized by low initial hydrophilic phenol content. However, EVOO A and its composite mixtures did not reach the maximum limit established for K270 at any time point of the experimental period. Furthermore, the rates of increase in both PV and K270 were consistently highest for D and its derivates (Table 2).

3.4. Evolution of phenolic compounds The evolution of all phenolic compounds was determined to explore the involvement of antioxidant substances in retarding or inhibiting oxidative phenomena in the EVOOs. Fig. 2a shows the evolution of the composition of polyphenols expressed as the sum of 3,4-DHPEA, p-HPEA, 3,4-DHPEA-EDA, p-HPEA-EDA, and lignans. All 14 samples showed gradual decreases in each of these substances; the loss percentages ranged from 72.1% to 90.8% (Table 3), with the highest losses observed in those EVOOs with the lowest initial contents. In particular, after 165 days of storage and light exposure, all EVOOs—except for D, for which polyphenol loss stabilized at 76%—with polyphenol contents lower than 200 mg/kg, such as C, D+C, and DC+DB, exhibited loss percentages that exceeded 86%. However, for EVOOs with initial concentrations between 400 and 700 mg/kg (D+B, C+B, and DC+CA), the decreases were 75.4% to 81.9%, with lower percentages corresponding to EVOOs with higher polyphenol concentrations. Even lower levels of polyphenol loss were observed for those EVOOs with initial polyphenol concentrations from 700 to more than 1400 mg/kg, as in oils DB+BA, A+D, C+A, B+A and A, with total decreases of 71.5%, 75.4%, 77.1%, 71.5%, and 72%, respectively, at the end of the experiment. Thus, EVOOs characterized by a higher initial polyphenol content showed smaller polyphenol losses during storage. It must be noticed that polyphenols are very important natural antioxidants that function in several reactions. For example, polyphenols can chelate transition metal ions, directly scavenge

molecular species of active oxygen, and inhibit lipid peroxidation by trapping the lipid alkoxyl radical (Roche, Dufour, Mora, & Dagles, 2005; Choe & Min, 2005; 2006; 2009; Sharma, Jha Ambuj, Dubey, & Pessarakli, 2012). During light exposure, their metal chelation property might engage the pro-oxidant activity of chlorophyll through an interaction with the ring of the magnesium atom, limiting its capability to transform stable 3O2 molecules into the very highly destructive reactive oxygen species (ROS) 1O2. Moreover, once 1O2 is formed, it can be quenched by polyphenols, although the contribution of polyphenols is marginalized by some authors who consider carotenoids as major 1O2 stabilizers (Nishida, Yamashita, & Miki, 2007; Kim & Choe, 2013). Furthermore, 1O2 is a short-lived (few nanoseconds) ROS that directly oxidizes unsaturated fatty acids, thereby forming lipid peroxy radicals and hydroperoxidases (Khan 1955; Choe & Min, 2005; 2006; 2009; Roche, Dufour, Mora, & Dagles, 2005) that can be stabilized by polyphenols, thus limiting or interrupting the oxidative process in the entire product. Thus, in this study, the initial quantity of these antioxidants was fundamental in inhibiting photo-oxidation in EVOOs through their combined antioxidant abilities, which strictly hinder the diffusion of free radicals and restrict peroxidative reactions. This feature also allowed for reduced polyphenol loss during oxidative processes. In fact, in all EVOOs characterized by higher initial polyphenol concentrations, reduced losses of polyphenols (Table 3) and lower PV and K270 (Table 2), which describe the primary and secondary phases of oxidation, respectively, were observed. Evolutions in oleuropein, ligstroside-derivative, lignan, and α-tocopherol contents were examined to verify the effects of these compounds on EVOO shelf life. Oleuropein derivatives 3,4DHPEA, 3,4-DHPEA-EA, and 3,4-DHPEA-EDA strictly followed the evolution seen for total polyphenols expressed as the sum of oleuropein, ligstroside derivatives, and lignans. As illustrated in Fig. 2b, the slope of each line representing the decrease in the amount of a given substance was highly similar to that of the slope for the total polyphenol content for each EVOO. EVOO polyphenols were mostly composed of oleuropein derivatives, except for oils C, D, and D+C, where this content was limited to 45–50% of the total polyphenol content. This finding was in agreement

with a previous study (Servili et al., 2015). In all other samples, oleuropein derivatives represented at least 70% of the entire polyphenol content (Table 3). However, even though the decrease was higher than 80% for all oils, losses were relatively larger in EVOOs characterized by lower phenolic and oleuropein derivative concentrations (Table 3). Furthermore, except for EVOO D+B, 3,4DHPEA, 3,4-DHPEA-EA, and 3,4-DHPEA-EDA were no longer observed at the end of the experiment in EVOOs with lower levels of polyphenols and oleuropein derivatives (D, C, D+C, and DC+DB). However, levels of approximately 100% were observed in those EVOOs with 400 mg/kg polyphenols, such as C+B and DC+CA (Table 3). For those EVOOs most abundant in polyphenols (700–1450 mg/kg) and thus, in oleuropein derivatives (530–1140 mg/kg), the mean decrease was reduced to 86.5% (Table 3). These findings, which corroborate previous hypotheses regarding the involvement of polyphenols in limiting photo-oxidation through various antioxidant mechanisms (metal chelation, radical quenching, and scavenging), prompted us to hypothesize that the antioxidant effects were mostly owing to oleuropein derivatives. Several other studies previously confirmed the higher antioxidant activities of o-diphenols, such as 3,4-DHPEA, 3,4-DHPEA-EA, and 3,4-DHPEA-EDA (Baldioli et al., 1996; Carrasco-Pancorbo, Cerretani, Bendini, SeguraCarretero, Lercker, & Fernández-Gutiérrez, 2005; Servili et al., 2009; Esposto et al., 2015). A different behaviour was observed for ligstroside derivatives (expression of p-HPEA and p-HPEA-EDA; Supplementary Fig. S1), which showed very little variation over the entire storage period. Except for oils C, D, and D+C, in which this fraction represented 43.6%, 50.3%, and 45% of total polyphenols, respectively, the concentrations of the ligstroside derivatives ranged between 18.6–23.1% (Table 3). At the end of storage, decreases in ligstroside derivatives of more than 65% were registered only for those samples having an initial polyphenol concentration lower than or equal to 200 mg/kg, such as the EVOOs D, C, D+C, and DC+DB (Table 3). For the remainder of samples, the losses in this fraction were approximately 25–57%, with lower decreases correlating with higher polyphenol concentrations. Thus, ligstroside derivatives showed an important involvement in inhibiting oxidation phenomena in EVOOs with very low concentrations of

oleuropein derivatives, confirming findings of previous studies on EVOOs that had been exposed to stresses (Baldioli et al., 1996; Carrasco-Pancorbo, Cerretani, Bendini, Segura-Carretero, Lercker, & Fernández-Gutiérrez, 2005; Servili et al., 2009; Esposto et al., 2015). Similarly, the lignans +-pinoresinol and 1(+)-acetoxypinoresinol showed low variability in each EVOO examined (Supplementary Fig. S1), indicating a very limited contribution to the entire antioxidant activities during storage. Their initial contents were 4.0–5.7% of total polyphenols (Table 3). At the end of the experiment, the reductions ranged between 4.6–26%, with most EVOOs displaying decreases below 15%. In particular, no correlations with the quantitative and qualitative compositions of the other polyphenols were observed.

3.5. Evolution of α-tocopherol α-Tocopherol is a lipophilic antioxidant involved in neutralizing lipid peroxy radicals and 1

O2 (Roche, Dufour, Mora, & Dagles, 2005; Choe & Min, 2005; 2006; 2009; Sharma, Jha Ambuj,

Dubey, & Pessarakli, 2012). Time-course analysis of α-tocopherol based on HPLC showed that its levels consistently decreased in all EVOOs upon light exposure (Fig. 2c), even though the decreases varied substantially among the samples (Table 3). The initial α-tocopherol content was similar among the 14 oils, with values between 172.6 and 218.6 mg/kg (Table 3). Moreover, higher initial α-tocopherol content did not correlate with higher polyphenol content (Table 3); however, its evolution during storage was largely influenced by the hydrophilic antioxidant fraction. Indeed, the oils with the lowest base levels of polyphenols, such as D, C, D+C, and DC+DB, showed substantial decreases of 96%, 100%, 80%, and 82%, respectively, in this lipophilic compound (Table 3). In particular, for these oils, the largest decreases were observed between the 45th and the 60th day of storage (Fig. 2c). For the other 10 EVOOs, the mean decreases were below 54.7%. This improved conservation of α-tocopherol can be attributed to the existence of phenols, which have been previously reported to have a higher antioxidant activity than α-tocopherol (Baldioli et al., 1996; Carrasco-Pancorbo, Cerretani, Bendini, Segura-Carretero, Lercker, & Fernández-Gutiérrez

2005; Servili et al., 2009; Esposto et al., 2015), and thus might be effective stabilizers of αtocopherol under the experimental conditions and during accelerated stresses (such as in AOM or RANCIMAT tests). This finding is important because it might contribute to finding ways to maintain high vitamin and antioxidant effects of α-tocopherol, and thus, the health-promoting value of EVOOs. Importantly, EVOOs with polyphenol contents exceeding 350 mg/kg showed higher oxidative stability, even at the early stages of storage, limiting the involvement of α-tocopherol in oxidation phenomena and thus potentially preserving its antioxidant effect over periods longer than 165 days. Furthermore, the results showed that this protective activity was proportional to the initial phenolic content (Table 3). In contrast, in oils with very low polyphenol base levels, α-tocopherol represented a useful antioxidant fraction for inhibiting negative oxidation processes. Thus, a larger reduction in α-tocopherol could be expected in such oils, as previously reported (Servili et al., 2009; Esposto et al., 2015).

3.6. Evolution of the oxidized phenolic products The relative content of the 3,4-DHPEA-EDA terpene degradation product, ODFEA, was monitored over time using LC-ESI-MS and LC/MS Q-TOF to confirm the involvement of polyphenols in oxidation processes during light exposure in the 14 EVOOs. As shown in Fig. 2d, initially, given the freshness of the samples, the HPLC-ESI-MS profile showed that ODFEA was not present; however, ODFEA gradually increased during storage. ODFEA appeared in each sample after 30 days of light exposure, even though at varying concentrations; samples characterized by higher oleuropein derivative concentrations, such as A, C+A, B+A, A+D, CA+BA, DC+CA, and DB+BA, were higher in ODFEA (Fig. 2d). In these samples, ODFEA accumulated quite quickly— within 30 days of light exposure—and, in particular for sample A, increased almost linearly during successive stages. However, for samples with lower polyphenol base levels, such as D, C, D+C, and DC+DB, ODFEA were 60-fold lower and remained relatively constant throughout storage. The accumulation of ODFEA in oils suggests the direct involvement of 3,4-DHPEA-EDA in oxidation

reactions (Fig. 2d), as previously proposed by Di Maio et al. (2013), who showed that oxidation phenomena occurring during EVOO stresses induced the oxidation of 3,4-DHPEA-EDA, which was hydrolysed to ODFEA. Furthermore, based on the time-course analysis results, it was reasonable to assume that the degradation of 3,4-DHPEA-EDA and subsequent oxidation product formation were proportional to the initial concentration of oleuropein derivatives in the oils, and that the rapid increases in ODFEA proved its prompt involvement in inhibiting oxidative phenomena by preventing or hampering ROS production. These findings also explain the minor losses of antioxidants in samples with high initial polyphenol concentrations.

3.7. Evaluation of volatile compounds The evaluation of volatile compounds was particularly focused on substances that originate from the decomposition of hydroperoxides during oxidation and that are responsible for rancid offflavours of EVOOs. To this end, we examined the evolution of the sum of headspace C7-C11 aldehydes (2-heptenal, (E)-2-heptenal, (E)-2,4-heptadienal, (E, E)-2,4-heptadienal, (E, E), 2,4nonadienal, (E, E)-2,4-decadienal, (E, E)-2,4-decadienal, and (E,E)-2-undecenal) which are reportedly the most abundant in rancid olive oils (Aparicio, Morales, & Alonso, 1996; Angerosa, Servili, Selvaggini, Taticchi, Esposto, & Montedoro, 2004; Luna, Morales, & Aparicio, 2006; Bendini, Cerretani, Salvador, Fregapane, & Lercker, 2010). At time point zero, none of these aldehydes were observed, proving the freshness of the EVOOs at the beginning of the experiment (data not shown). At 30 days, these compounds were increased in all 14 oils, with large amounts (50–110 µg/kg) observed for D, C, D+C, DC+DB, and D+B; however, in A, B, C+A, B+A, A+D, CA+BA, and DB+BA, the concentrations did not exceed 50 µg/kg and were less than 20 µg/kg in most samples (Fig. 3). The aldehydes gradually accumulated in the headspaces of all EVOOs until day 90. Between the 90 th and the 105th day of storage, the samples that showed substantial volatile compound accumulation, such as D, C, D+C, C+B, and DC+DB, reached concentrations up to 780– 1341 µg/kg. Furthermore, in these samples, the concentrations amounted to 2.5 mg/kg during the

late phases of light exposure (from the 150th to the 165 th day). The lowest levels were recorded in A and its mixtures until 105 days, with values ranging from 156.2 to 565.8 µg/kg. After that period, the sum of C7-C11 aldehydes was significantly increased for these EVOOs, even though the levels remained approximately 1000 µg/kg lower than those for EVOOs C and D (Fig. 3). These data corroborate that the higher the polyphenol content, which can inhibit hydroperoxidase homolytic cleavage, the lower the production of these substances promoted by hydroperoxidase activity. These findings confirmed previous reports on the capability of polyphenols to inhibit the evolution to the secondary phase of oxidation processes in several fat-containing food products exposed to oxidative stresses (Luna, Morales, & Aparicio, 2006; Esposto et al., 2015). Moreover, the values obtained validated the K270 coefficient as a qualitative parameter for monitoring the freshness of EVOOs during storage. This index is better than other product parameters, such as PV, because EVOOs that had concentrations higher than the legal value were also characterized by higher concentrations of C7-C11 volatile compounds (Table 2, Fig. 3), which, in fact, absorb at a wavelength of 270 nm (Gutierrez & Fernandes, 2002; Caponio, Bilancia, Pasqualone, Sikorska, & Gomes, 2005; Kalua, Bedgood, Bishop, & Prenzler, 2006; Gómez-Alonso et al., 2007). With the exception of C+B, all EVOOs characterized by low polyphenol concentrations, including D, D+B, DC+DB, and C, reached the legal K270 limit of 0.22 (EU Reg. 2015/1830) after 105 days of light exposure, the same period in which substantial C7-C11 aldehyde accumulation was registered in the headspaces of these oils.

Conclusions Simulations of real food shelf life are not often applied in scientific studies, mainly because of the time and budget required for such analyses. However, such studies are very useful for understanding the quality changes that EVOOs undergo during the often very long periods of light exposure in the market. With this in mind, we conducted a shelf-life simulation study of 14 EVOOs in 11 bottles per type over approximately 6 months, evaluating those substances that mainly

influence both the quality and stability during storage with light exposure, such as oleic acid content and quantitative as well as qualitative polyphenol composition. PCA and PLS modelling showed that there was a limited influence of oleic acid percentage and acidity on EVOO stability during light exposure. In contrast, factors affecting health-promoting and sensory qualities showed larger effects. The K270 extinction coefficient may be of capital importance for monitoring the quality of stored EVOOs by predicting the time at which they will lose their ‘extra’ status in a manner that is certainly easier and cheaper than a volatile compound instrumental analysis or an evaluation by a panel of assessors. The loss of those typical EVOO health-promoting and sensory properties during light exposure is highly dependent on their initial antioxidant capacity, particularly on the oleuropein derivative contents. The higher the oleuropein derivative contents, the higher the resistance to oxidation. However, in EVOOs with oleuropein derivative contents of at least 300 mg/kg, lower losses of secoiridoid derivatives and α-tocopherol and reduced accumulation of the volatile C7-C11 substances responsible for the rancid defect were observed. The fast increase in ODFEA content in EVOOs with higher polyphenol contents was a further indication of the immediate and primary involvement of oleuropein derivatives in inhibiting oxidation, according to their 1O2 quenching, radical scavenging, and metal chelation properties, which also presumably reduce the negative effect of chlorophyll on photo-oxidation. These results were corroborated by the observation that in EVOOs with lower oleuropein derivative concentrations, an important loss of polyphenols and tocopherols and a substantial increase in the negative volatile substances was found, even before the first stages of storage. Based on these results, it is reasonable to assume that the time beyond which an EVOO loses its sensory and health qualities can be predicted as a function of its initial oleuropein derivative content.

Acknowledgements The authors wish to thank the Ministero Italiano dell’Università della Ricerca MIUR (Project PROSIT, Cluster Agrifood Nazionale) for providing financial support.

References Andrewes, P., Busch, J. L. H. C., De Joode, T., Groenewegen, A., & Alexandre H. (2003). Sensory properties of virgin olive oil polyphenols: Identification of deacetoxy-ligstroside aglycon as a key contributor to pungency. Journal of Agricultural and Food Chemistry, 51, 1415. Angerosa, F., Servili, M., Selvaggini, R., Taticchi, A., Esposto S., & Montedoro, G. F. (2004). Volatile compounds in virgin olive oil: Occurrence and their relationship with the quality. Journal of Chromatography A, 1054, 17–31. Aparicio, R., Morales, M. T., & Alonso, M. V. (1996). Relationship between volatile compounds and sensory attributes by statistical sensory wheel. Journal of the American Oil Chemists' Society, 73, 1253–1264. Baldioli, M., Servili, M., Perretti, G., & Montedoro G. F. (1996). Antioxidant activity of tocopherols and phenolic compounds of virgin olive oil. Journal of the American Oil Chemists’ Society, 73 1589–1590. Bendini, A., Cerretani, L., Salvador, M. D., Fregapane, G., & Lercker, G. (2010). Stability of the sensory quality of virgin olive oil during storage. An overview. Italian Food & Beverage Technology, 60, 5–18. Bulotta, S., Celano, M., Lepore, S. M., Montalcini, T., Pujia, A., & Russo D. (2014). Beneficial effects of the olive oil phenolic components oleuropein and hydroxytyrosol: focus on protection against cardiovascular and metabolic diseases. Journal of Translational Medicine, 12, 1–9. Caponio, F., Bilancia, M. L., Pasqualone, A., Sikorska, E., & Gomes, T. (2005). Influence of the exposure to light on extra virgin olive oil quality during storage. European Journal of Food Research and Technology, 221, 92–98.

Carrasco-Pancorbo, A., Cerretani, L., Bendini, A., Segura-Carretero A., Lercker, G., & FernándezGutiérrez A. (2005). Evaluation of the antioxidant capacity of individual phenolic compounds in virgin olive oil. Journal of Agriculture and Food Chemistry, 53, 8918–8925. Choe E, & Min D. B. (2009). Mechanisms of antioxidants in the oxidation of foods. Comprehensive Reviews in Food Science and Food Safety, 8, 345–358. Choe, E, & Min, D. B. (2005). Chemistry and reactions of reactive oxygen species in foods. Journal of Food Science, 70, 142–159. Choe, E, & Min, D. B. (2006). Mechanisms and factors for edible oil oxidation. Comprehensive Reviews in Food Science and Food Safety, 5, 169–186. Cinquanta, L., Esti, M., & Di Matteo, M. (2001). Oxidative stability of virgin olive oils. Journal of the American Oil Chemists’ Society, 78, 1197–1202. Commission Delegated Regulation (EU) 2015/1830 of 8 July 2015 amending Regulation (EEC) No 2568/91 on the characteristics of olive oil and olive-residue oil and on the relevant methods of analysis. Condelli, N., Caruso, M. C., Galgano, F., Russo, D., Milella, L., & Favati, F. (2015). Prediction of the antioxidant activity of extra virgin olive oils produced in the Mediterranean area. Food Chemistry, 177, 233–239. Coutelieris, F. A., & Kanavouras, A. (2006). Experimental and theoretical investigation of packaged olive oil: Development of a quality indicator based on mathematical predictions. Journal of Food Engineering, 73, 85–92. Del Nobile, M. A., Bove, S., La Notte, E., & Sacchi, R. (2003). Influence of packaging geometry and material properties on the oxidation kinetic of bottled virgin olive oil. Journal of Food Engineering, 57, 189–197. Di Maio, I., Esposto, S., Taticchi, A., Selvaggini, R., Veneziani, G., Urbani, S., & Servili, M. (2013). Characterization of 3,4-DHPEA-EDA oxidation products in virgin olive oil by high

performance liquid chromatography coupled with mass spectrometry. Food Chemistry, 138, 1381–1391. Esposto, S., Taticchi, A., Di Maio, I., Urbani, S., Veneziani G., Selvaggini R., Sordini B., & Servili, M. (2015). Effect of an olive phenolic extract on the quality of vegetable oils during frying. Food Chemistry, 176, 184–192. Esposto, S., Veneziani, G., Taticchi, A., Selvaggini, R., Urbani, S., Di Maio, I., Sordini, B., Minnocci, A., Sebastiani, L., & Servili, M. (2013). Flash thermal conditioning of olive pastes during the olive oil mechanical extraction process: impact on the structural modifications of pastes and oil quality. Journal of Agricultural and Food Chemistry, 61, 4953–4960. Estruch, R., Martínez-González, M.A., Corella, D., Salas-Salvadó, J., Ruiz-Gutiérrez, V., Covas, M.I., Fiol, M., Gómez-Gracia, E., López-Sabater, M.C., Vinyoles, E., Arós, F., Conde, M., Lahoz, C., Lapetra, J., Sáez, G., & Ros, E. (2006). Effects of a Mediterranean-style diet on cardiovascular risk factors: a randomized trial. Annual Internal Medicine, 145, 1–11. Gõmez-Alonso, S., Mancebo-Campos, V., Desamparados Salvador, M., & Fregapane, G. (2007). Evolution of major and minor components and oxidation indices of virgin olive oil during 21 months storage at room temperature. Food Chemistry, 100, 36–42. Gutiérrez, F., & Fernandez, J. L. (2002). Determinant parameters and components in the storage of virgin olive oil. Prediction of storage time beyond which the oil is no longer of extra quality. Journal of Agricultural and Food Chemistry, 50, 571–577. Hachicha Hbaieb, R., Kotti, F., Gargouri, M., Msallem, M., & Vichi, S. (2016). Ripening and storage conditions of Chétoui and Arbequina olives: Part I. Effect on olive oils volatiles profile. Food Chemistry, 203, 548–558. Khan (1955). Infrared studies on initiation of the autoxidation of some fatty acid esters with and without light-sensitized chlorophyll, ultraviolet light and lipoxidase. Biochimica et Biophysica Acta, 16, 159-160.

Kalua, C. M., Bedgood, D. R., Bishop, JR., A. G,, & Prenzler P. D. (2006). Discrimination of storage conditions and freshness in virgin olive oil. Journal of Agricultural and Food Chemistry, 54, 7144-7151. Kim, N., & Choe, E. (2013). Contribution of minor compounds to the singlet oxygen-related photooxidation of olive and Perilla oil blend. Food Science Biotechnology, 22, 315–321. Konstantinidou, V., Covas, M. I., Munõz-Aguayo, D., Khymenets, O., de la Torre, R., Saez, G., del Carmen Tormos, M., Toledo, E., Marti, A., Ruiz-Gutiérrez V., Ruiz Mendez M. V., & Fito, M. (2010). In vivo nutrigenomic effects of virgin olive oil polyphenols within the frame of the Mediterranean diet: a randomized controlled trial. The FASEB Journal, 10, 2546–2557. Krichene, D., Allalout, A., Mancebo-Campos, V., Salvador, M. D., Zarrouk, M., & Fregapane, G. (2010). Stability of virgin olive oil and behaviour of its natural antioxidants under medium temperature accelerated storage conditions. Food Chemistry, 121, 171–177. Luna, G., Morales, M. T., & Aparicio, R. (2006). Changes induced by UV radiation during virgin olive oil storage. Journal of Agricultural and Food Chemistry, 54, 4790–4794. Mancebo-Campos, V., Desamparados, Salvador, M., & Fregapane, G. (2014). Antioxidant capacity of individual and combined virgin olive oil minor compounds evaluated at mild temperature (25 and 40 °C) as compared to accelerated and antiradical assays. Food Chemistry, 150, 374– 381. Mateos, R., Uceda, M., Aguilera, M. P., Escuderos, M. E., & Maza, G. B. (2006). Relationship of Rancimat method values at varying temperatures for virgin olive oils. European Food Research and Technology, 223, 246–252. Mendez, A., & Falque, E. (2007). Effect of storage time and container type on the quality of extra virgin olive oil. Food Control, 18, 521–529. Morello, J. R., Motilva, M. J., Tovar, M. J., & Paz Romero, M. (2004). Changes in commercial virgin olive oil (cv Arbequina) during storage, with special emphasis on the phenolic fraction. Food Chemistry, 85, 357–364.

Nishida, Y., Yamashita, E., & Miki, W. (2007). Quenching activities of common hydrophilic and lipophilic antioxidants against singlet oxygen using chemiluminescence detection system. Carotenoid Science, 11, 16–20. Okogeri, O., & Tasioula-Margari, M. (2002). Changes occurring in phenolic compounds and αtocopherol of virgin olive oil during storage. Journal of Agricultural and Food Chemistry, 50, 1077–1080. Pagliarini, E., Zanoni, B., & Giovanelli, G. (2000). Predictive study on Tuscan extra virgin olive oil stability under several commercial conditions. Journal of Agricultural and Food Chemistry, 48, 1345–1351. Pokorny J., Kalinova, L., & Dysseler, P. (1998). Determination of chlorophyll pigments in crude vegetable oils. Pure Applied Chemistry, 67, 1781–1787. Pristouri, G., Badeka, A., & Kontominas, M. G. (2010). Effect of packaging material headspace, oxygen and light transmission, temperature and storage time on quality characteristics of extra virgin olive oil. Food Control, 21, 412–418. Psomiadou, E., & Tsimidou, M. (1998). Simultaneous HPLC determination of tocopherols, carotenoids, and chlorophylls for monitoring their effect on virgin olive oil oxidation. Journal of Agriculture and Food Chemistry, 46, 5132–5138. Psomiadou, E., & Tsimidou, M. (2002a). Stability of virgin olive oil. Autoxidation studies. Journal of Agricultural and Food Chemistry, 50, 716–721. Psomiadou E., & Tsimidou M. (2002b). Stability of virgin olive oil. 2. Photo-oxidation studies. Journal of Agricultural and Food Chemistry, 50, 722–727. Rahmani, M., & Saari Csallany, A. (1998). Role of minor constituents in the photooxidation of virgin olive oil. Journal of the American Oil Chemists’ Society, 75, 837–843. Roche, M., Dufour, C., Mora, N., & Dagles, O. (2005). Antioxidant activity of olive phenols: mechanistic investigation and characterization of oxidation products by mass spectrometry. Organic & Biomolecular Chemistry, 3, 423–430.

Servili, M., Esposto, S., Taticchi, A., Urbani, S., Di Maio, I., Veneziani G., & Selvaggini, R. (2015). New approaches to virgin olive oil quality, technology, and by-products valorization. European Journal of Lipid Science and Technology, 117, 1882–1892. Servili, M., Esposto, S., Fabiani, R., Urbani, S., Taticchi, A., Mariucci, F., Selvaggini, R., & Montedoro, GF. (2009). Phenolic compounds in olive oil: Antioxidant, health and organoleptic activities according to their chemical structure. Inflammopharmacology, 17, 1–9. Sharma P., Jha A. B., Dubey R. S., & Pessarakli M. (2012). Reactive oxygen species, oxidative damage, and antioxidative defense mechanism in plants under stressful conditions. Journal of Botany, article ID 217037. Stefanoudaki, E., Williams, M., & Harwood, J. (2010). Changes in virgin olive oil characteristics during different storage conditions. European Journal of Lipid Science and Technology, 112, 906–914. Terés, S., Barceló-Coblijn, G., Benet, M., Alvarez, R., Bressani, R., Halver, J.E., & Escribá, P.V. (2008). Oleic acid content is responsible for the reduction in blood pressure induced by olive oil. Proceedings of the National Academy of Sciences of the United States of America, 105, 13811– 13816.

Figure captions Fig. 1. Score and loading plots of the PLS model built using all the analytical determinations (representing the independent variable X) for the 14 EVOOs exposed to light for different periods (representing the latent variable Y, defined as TIME). Total explained variance of Y: 93%, with three latent significant variables.

Fig. 2. Evolution of the concentrations of the following compounds in the 14 EVOOs during 165 days of light exposure: polyphenols (mg/kg), expressed as the sum of oleuropein and ligstroside derivatives and lignans (a); oleuropein derivatives (mg/kg, sum of 3,4-DHPEA, 3,4-DHPEA-EDA,

and 3,4-DHPEA-EA) (b); α-tocopherol (mg/kg) (c); and the oxidized form of elenolic acid (area counts) (d).

Fig. 3. Evolution of the concentrations of volatile compounds (µg/kg) expressed as the sum of 2heptenal, (E)-2-heptenal, (E)-2,4-heptadienal, (E, E)-2,4-heptadienal, (E, E)-2,4-nonadienal, (E, E)2,4-decadienal, (E, E)-2,4-decadienal, and (E, E)-2-undecenal, in the 14 EVOOs during 165 days of light exposure. One-way ANOVA was used to compare results among all EVOOs at the same time point; different letters above each bar indicate significant differences between columns of the same colour (P < 0.05).

Figures 10 8 6 4 2 u[1]

T0

C +CA +DB +B++BABB+AD BA DBCDAC+A D D+ C BC+AADC DC B BAA +BC+CA C +D ++B+A B B+ACDABAA CDC CD+D D+D CD A +CA +DB +BB+D +ADCB +B+ABDC D+C D C A+A BBA DCCD BA +BAA+BC+CA C+DB +C DB++A AC+DCD D DD AB+DA B C+CB B ++BBA+A+D B +C+AD C CDAD DCDCC+B D+ A B CB+A DCBA A B A A A C ++BB B+C BC+D D+C A D B+ADCBBAC+A A+DDD+CC+D +B+ABA B +CA+DB +C CD BDBC+ADC++DBCDC AB++DA A CA D B+AD+BCA +C +BA+B+D B+C DD DCDC+DBC A B CB++AACA DA A A B A B C +B+ D B BD+ C D+C D ++C A BC+AADCB+AA+ DD+CC D B A +BA+BD+B+A C+DC+B DC+D C DD+C +BA DB CCD CAB+A AA A B +CA+DB B++B C +B+A D ABB +DA C D D+ C+DA+DC DC BC AA C DBCA BBAA AB++CA+BDAC+++BDA DCD+C+ C D D+C B+A CBD

0 -2 -4 -6 -8

T15

Loading plot: w*c[1]/w*c[2]

T30 T45 T60

0,20

T75 T90

0,10

T105 T120 T135

0,0

T150 T165

-0,10

-0,20

p-HPEA-EDA (+)-Pinoresinol 3,4 -DHPEA-EA Sum of phenolic fractions alpha -tocopherol 3,4 -DHPEA-EDA Margaric acid acidity cis-11-Eicosenoic acid K232

-8

-6

-4

-2

0 t[1]

Figure 1.

2

4

6

8

10

12

ODFEA

-0,20

-0,10

0,00

w*c [1]

0,10

K270 Propanal

2-Undecenal TIME 2,4 -Nonadienal, ( E,E)2,4-Decadienal, ( E,E)-i 2,4 -Decadienal, ( E,E)2,4-Heptadienal, ( E,E)-i Pentanal 2-Heptenal, (E) Peroxide Value

Linoleic acid Palmitoleic acid Palmitic acid

-0,30

-10

Loadings : w*c[1]/w*c[2]

Oleic acid Stearic acid Linolenic acid Hexanal Arachidic acid cis-10 -Heptadecenoic acid 3,4-DHPEA p-HPEA (+)-1-Acetoxypinoresinol

0,30

w*c [2]

Score plot: t[1]/u[1]

0,20

0,30

1600,0

1200,0

mg/kg

mg/kg

b

a 1400,0 1000,0

1200,0 800,0

1000,0 600,0

800,0 600,0

400,0

400,0 200,0

200,0 0,0

0,0 T0

250,0

T15

T30

T45

T60

T75

T90

T105

T120

T135

mg/kg

T150

T165

T0

T15

14000000,0

c

T30

T45

T60

T75

T90

T105

T120

T135

Area counts

T150

d

12000000,0

200,0

10000000,0 150,0

8000000,0

6000000,0

100,0

4000000,0 50,0

2000000,0

0,0

0,0 T0

T15

T30

T45

T60

T75

T90

A Figure 2.

C+B

T105

T120

B+A D+B

T135

T150

T0

T165

CA+BA DC+CA

C+A DC+DB

T30

DB+BA C

T60

A+D D+C

T90

T120

B D

T150

T165

T15 2500,0

T45

T75

T105

T135

d

T165

µg/kg df i cf ce

c

2000,0 e c b

b

d

1500,0

b

b

b

b

b b

b

m

a

a

a

l f h

f

f

f

f

f

1000,0 c

e

d

a o i

in b

500,0

bn c

e

c a

b bh

a

0,0

a

a

A

Figure 3.

d d

b

en e

b

a

a

B

C

D

d

g

C+A

fi f c

b gmo

g d

eg

a

D+B

B+A

D+C

bf h

l

h i

a

A+D

f

C+B

e

gh m g

i fno a

a

ae p

DC+DB CA+BA DC+CA DB+BA

Table 1 Initial fatty acid (%) and antioxidant (mg/kg) compositions of the 14 EVOOs1. Antioxidants Oleuropein

Ligstroside

(α-tocopherol +

SFA

MUFA

Oleic acid

PUFA

polyphenols)

derivatives

derivatives

Lignans

%

%

%

mg/kg

mg/kg

mg/kg

mg/kg

mg/kg

A

13.2 ± 1.0 78.6 ± 4.9 77.7 ± 4.9 8.0 ± 0.6 1650.5 ± 4.4 1143.9 ± 3.0 274.3 ± 1.0 58.5 ± 0.2 1

B

19.9 ± 1.0 67.6 ± 5.5 64.6 ± 5.5 12.5 ± 1.0 903.0 ± 8.2

508.7 ± 1.7 135.4 ± 0.9 38.4 ± 1.5 2

C

13.4 ± 0.7 78.6 ± 4.6 77.6 ± 4.6 8.0 ± 0.6 301.0 ± 1.7

66.1

± 0.6 55.9 ± 0.0

6.2 ± 0.1 1

8.1

± 0.5

0.9 ± 0.2 2

D

20

± 1.1 67.4 ± 5.5 64.5 ± 5.5 12.5 ± 1.0 236.7 ± 0.8

9.1

± 0.5

C+A

13.5 ± 0.3 78.5 ± 4.6 77.7 ± 4.6 8.0 ± 0.2 974.8 ± 10.4 605.0 ± 0.4 164.7 ± 10.0 32.4 ± 0.0 1

D+B

19.9 ± 0.9 67.6 ± 4.2 64.6 ± 4.2 12.5 ± 0.6 570.8 ± 6.5

258.4 ± 0.3 72.3 ± 6.4 19.7 ± 0.0 2

B+A

16.8 ± 1.4 73

811.0 ± 1.1 201.8 ± 0.5 47.0 ± 0.2 1

D+C

16.5 ± 1.4 73.4 ± 2.5 71.7 ± 2.5 10.0 ± 0.9 266.1 ± 3.9

35.4

A+D

16.5 ± 0.6 73.1 ± 4.3 71.3 ± 4.3 10.4 ± 0.4 853.1 ± 5.5

578.0 ± 1.8 48.3 ± 1.3 29.4 ± 1.9 1

C+B

16.6 ± 0.5 73.2 ± 2.5 71.3 ± 2.5 10.1 ± 0.3 701.0 ± 6.5

298.3 ± 4.2 182.8 ± 0.6 22.7 ± 2.0 1

DC+DB 17.4 ± 0.4 70.8 ± 5.0 68.3 ± 5.0 11.7 ± 0.3 454.3 ± 2.9

143.9 ± 0.3 91.0 ± 0.4 11.6 ± 0.7 2

CA+BA 15.2 ± 1.2 75.7 ± 6.6 74.1 ± 6.6 9.1 ± 0.8 1058.0 ± 1.8

694.8 ± 0.9 136.7 ± 0.4 41.2 ± 1.4 1

DC+CA 15.5 ± 1.0 75.3 ± 3.4 74.0 ± 3.4 9.1 ± 0.7 666.7 ± 6.6

314.7 ± 5.8 140.7 ± 0.1 28.0 ± 0.0 1

DB+BA 18.1 ± 0.9 70.6 ± 3.5 68.0 ± 3.5 11.2 ± 0.6 870.2 ± 6.4

531.3 ± 4.2 96.6 ± 0.2 33.3 ± 0.1 2

± 4.7 71.0 ± 4.7 10.3 ± 0.9 1257.2 ± 2.6

1

± 0.3 31.9 ± 3.2

Data are the mean of two independent experimentations twice analysed. SFA, saturated fatty acids; MUFA,

monounsaturated fatty acids; PUFA, polyunsaturated fatty acids.

3.6 ± 0.3 1

Table 2 Correlation coefficients of the linear regressions of quality indices PV and K270, time required to exceed the legal limits, and increase rate during the exposure to light. PV K270 Time required to exceed legal

Time required Increase rate

2

to exceed legal 2

R

R limit

(meq O2/kg olio)

limit

(range of 15 days)

Increase rate

(range of 15 days)

A

0.84

-

0.71

0.94

-

0.005

B

0.69

-

0.70

0.88

C

0.67

150–165

1.19

0.81

105–120

0.011

D

0.60

135–150

1.35

0.94

105–120

0.010

C+A

0.80

-

0.75

0.78

-

0.011

D+B

0.66

150–165

1.34

0.84

105–120

0.014

B+A

0.47

-

1.14

0.90

-

0.009

D+C

0.79

-

1.16

0.83

-

0.007

A+D

0.90

-

1.10

0.80

-

0.015

C+B

0.70

-

1.06

0.91

-

0.008

DC+DB

0.69

150–165

1.35

0.74

135–150

0.014

CA+BA

0.74

-

0.83

0.83

-

0.009

DC+CA

0.71

-

1.11

0.97

-

0.007

DB+BA

0.81

-

1.07

0.76

-

0.013

0.005

Maximum values of EVOO quality parameter (EU Reg. 2015/1830): PV, 20 meq O2/kg; K270, 0.22.

Table 3 Initial (T0) and final (T165) compositions (mg/kg) and relative decrease percentage (D.P.) between these two times points in oleuropein and ligstroside derivatives, lignans and their sum (polyphenols), and α-tocopherol1. POLYPHENOLS

OLEUROPEIN DER.

D.P. T0

T165

T0-

D.P. T0

T165

T165 mg/kg mg/kg

1

%

LIGSTROSIDE DER.

T0-

D.P. T0

T165

T165 mg/kg mg/kg

%

LIGNANS

T0-

D.P. T0

T165

T165 mg/kg mg/kg

%

α-Tocopherol

T0-

D.P. T0

T165

T165

T0T165

mg/kg mg/kg

%

mg/kg mg/kg

%

A

1476.7

411.7

72.1

1143.9

150.7

86.8

274.3

205.1

25.2

58.54

55.87

4.6

173.8

79.0

54.6

B

682.5

160.3

76.5

508.7

65.4

87.2

135.4

66.4

51.0

38.41

28.48

25.9

220.5

76.7

65.2

C

128.3

11.8

90.8

66.1

0.0

100.0

55.9

7.1

87.3

6.25

4.74

24.0

172.6

0.0

100.0

D

18.1

3.43

76.2

8.1

0.0

100.0

9.1

2.9

67.9

0.93

0.53

43.0

218.6

8.6

96.0

C+A

802.1

183.9

77.1

605.0

65.8

89.1

164.7

91.5

44.4

32.39

26.56

18.0

172.7

89.2

48.3

D+B

350.3

74.8

78.6

258.4

27.6

89.3

72.3

30.6

57.6

19.67

16.56

15.8

220.5

124.1

43.7

B+A

1059.8

302.0

71.5

811.0

156.0

80.8

201.8

104.0

48.5

47.01

41.99

10.7

197.3

96.6

51.0

D+C

70.9

9.4

86.8

35.4

0.3

99.2

31.9

6.0

81.3

3.59

3.08

14.0

195.2

39.1

80.8

A+D

748.1

184.3

75.4

578.0

56.0

90.3

140.7

102.2

27.3

29.43

26.11

11.3

197.4

87.8

55.5

C+B

417.6

75.6

81.9

298.3

11.8

96.0

96.6

51.2

47.0

22.73

12.61

44.5

197.2

78.4

60.2

DC+DB

203.8

21.4

89.5

143.9

1.7

98.8

48.3

9.9

79.5

11.63

9.82

15.5

207.7

37.9

81.8

CA+BA

918.8

241.6

73.7

694.8

98.7

85.8

182.8

109.9

39.9

41.23

33.03

19.9

185.3

82.5

55.5

DC+CA

423.7

74.4

82.4

314.7

8.7

97.2

91.0

47.6

47.7

18.99

18.10

4.7

183.3

89.9

50.9

DB+BA

701.3

199.5

71.5

531.3

75.5

85.8

136.7

95.1

30.4

33.34

28.90

13.3

209.0

78.3

62.5

Data are the mean of two independent experimentations twice analysed.

31

THE EFFECTS OF MARKET STORAGE CONDITIONS ON EVOO QUALITY WERE SIMULATED. LONG LIGHT EXPOSURE TIMES AND CHEMICAL DIFFERENCES OF THE MARKET PRODUCTS WERE CONSIDERED. COMMERCIAL EVOO MERCHANDISE QUALITY CAN BE MONITORED BY THE K270 EXTINCTION COEFFICIENT. THE TREND OF QUALITY LOSS OF AN EVOO THAT IS EXPOSED TO A LIGHT SOURCE DEPENDS STRONGLY ON ITS ANTIOXIDANT HERITAGE. OEUROPEIN DERIVATIVES ARE THE MOST EFFICIENT ANTIOXIDANTS IN TERMS OF CONTRASTING PHOTOOXIDATION.

32