Characterization of apple seed oil with Denomination of Origin from Asturias, Spain

J.C. Bada^a, M. León-Camacho^b, P. Copovi^a and L. Alonso^a,*

^aInstituto de Productos Lácteos de Asturias (CSIC). Paseo Rio Linares s/n. 33300 Villaviciosa-Asturias, Spain

^bInstituto de la Grasa (CSIC), Avda Padre García Tejero 4, 41012 Sevilla, Spain

*Corresponding author: lalonso@ipla.csic.es

1. INTRODUCTIONTOP

In the last few years, there has been a considerable interest in finding new food sources that will meet the health and nutritional needs of the world´s population (Maazouzi et al., 2007; Ohama et al., 2006; Yukui et al., 2009). In Japan, many of the new foods with health claims have been wild vegetables (Cai et al., 2003), such as mustard leaf (Liang et al., 2006) and elm fruit (Hu et al., 2006). Such resources are limited, and more attention should be paid to existing waste resources such as apple seed oils. Denomination of origin (DO) is a special consideration for some foods and drinks to recognize that they have benefits and particular characteristics in many differentiated geographical areas, and the influence of human factor in their production (European Union, 2008). Spain is one of the most important countries of the European Union for the production of cider, producing about 585,000 tons of apples as fruit for human consumption and 75,000 tons of apples to make cider. Asturias is the region with the most important cultivation of apples for the production of cider and have 22 apple varieties included in the DO. Cider, in general, is produced with different varieties of apples, but the cider with DO from Asturias is produced only with apples recognized by the DO (BOE, 2003). In the process for producing cider, there are two important by-products, seeds and skins. The seeds have variable contents in oils. In the cider industry, this seed is not used for producing apple seed oil, and it is a waste product. Fruit seed oil has been widely applied in food, perfumes, toiletries and chemical additives (Stone and Kushner, 2000; Etherton and Etherton, 2003). Plant seed oil is one of the most interesting essential oils due to its properties (Anwar et al., 2006; Lei Tian et al., 2010). A few authors have published papers about apple seed oil and they describe some of its properties (Yu et al., 2007; Yukui et al., 2009; Lei Tian et al., 2010).

The aim of this research was to study the oils of seven different apple seed oils from Asturias (Spain), Blanquina, Raxao, Collaos, Durona, Riega, Solarina and Limón Montés species pertaining to the DO cider from Asturias. The composition in fatty acids, sterols, phospholipids, tocopherols and triglycerides were characterized, as well as the total contents in oil and moisture.

2. MATERIAL AND METHODSTOP

2.1. SamplesTOP

The apple seed species (Table 1) used in the present study came from different plantations in the region of Asturias (Spain), Serida (Principado de Asturias) and San Martín–Llanes (Asturias) collected during a period of suitable ripening. Once prepared at room temperature, the solid matter (skin and pulp) was separated from the seeds.

2.2. Solvent oil extraction and water contentTOP

Samples were blanched and ground in an electrical grinder. The oil was extracted in a Soxhlet glass apparatus using hexane as solvent (IUPAC, 1987a). Moisture was determined by weight loss after heating in an oven at 105 °C in accordance with (IUPAC, 1987b).

2.3. Fatty acid analysisTOP

Fatty acid methyl-esters (FAME) were analyzed by gas chromatography (GC). FAME were extracted with n-heptane after cold methylation with 2N KOH in methanol (IUPAC, 1987c). FAME analysis was performed on an HP-5890-II apparatus (Hewlett-Packard, Palo Alto, CA) using a fused silica capillary SP-2380 column (60m×0.25 mm, 0.2 μm film thickness) with a flame ionization detector (FID). The oven temperature was kept at 160 °C for 13 min and was then raised to 190 °C at a rate of 1.5 °C min⁻¹ and held isothermally for 20 min. The injector temperature was kept at 225 °C, while the detector temperature was 250 °C. Hydrogen (19 psi inlet pressure) was used as carrier gas, while the make-up gas was nitrogen. Standards of each fatty acid were used to identify the fatty acids. These were purchased from Sigma-Aldrich (St. Louis, MO): palmitic (P=C16:0), palmitoleic (Po=C16:1), stearic (S=C18:0), oleic (O=C18:1 ω9 cis), linoleic (L=C18:2) alpha and gamma, linolenic (Lo=C18:3). Fatty acids were identified in the samples by comparing the retention times for standards and those of the samples. The area was expressed as percentages of areas of the total fatty acids.

2.4. Sterol analysisTOP

The unsaponifiable fraction was extracted as described (European Communities 1991). 0.5 mL 5-α-cholestanol (Fluka, Buchs, Switzerland) solution in chloroform was added to 5 g of oil as an internal standard. The mixture was saponified for 0.5 hour with 50 mL of 2N ethanolic potassium hydroxide. The solution was then passed to a 500 mL decanting funnel, 100 mL distilled water were added and the mixture was extracted twice with three 80 mL portions of diethyl ether. The diethyl ether extracts were combined in another funnel and were washed several times with 100 mL portions of water, until the wash reached a neutral pH. The diethyl ether solution was dried over anhydrous sodium sulphate and evaporated to dryness in a rotary evaporator at 30 °C under reduced pressure. After purification by thin-layer chromatography (IUPAC, 1987d), the sterol fraction was analyzed and quantified by gas chromatography in an HP 5890-II apparatus equipped with a split-splitless injector and a flame ionization detector. An HP-5 fused silica capillary column (30 m×0.32 mm i.d., 0.25 mm film thickness) was used with hydrogen (7 psi inlet pressure) as the carrier gas and nitrogen the make-up gas. The oven temperature was maintained isothermally at 265 °C for 30 min. The injector temperature was 280 °C, while the detector was kept at 300 °C. Previously, the sterol fraction was derivatized as trimethylsilyl ethers (TMS) according to the method proposed in (European Communities, 1991). The sterols were identified by comparison of the mass spectral data with those of authentic reference compounds and by comparing their retention times with sterols from olive, sunflower and soybean oils (León-Camacho and Morales, 2000).

2.5. Phospholipid and triglyceride analysisTOP

Phospholipids were analyzed by reverse phase analysis of the lipids carried out on a Water Acquity Ultra Performance LC (UPLC) system using an Acquity UPLC-Bech C18 column (2.1×10 mm, 1.5 μm) at 40 °C. The solvent consisted of water/methanol (10/90) and the flow rate of 1 mL min⁻¹, which was coupled to an evaporating light scattering detector (ELSD 2424-Waters). As the nebulizing gas, N₂ was used at a flow rate of 1.4 L min⁻¹, and a nebulizing temperature of 80 °C. The injection volume was 10 μL (30 mg 10 mL⁻¹ oil sample). The assignment of chromatographic peaks was carried out by means of standards purchased from Sigma-Aldrich (St.Louis, MO). Tryglycerides were performed on a capillary gas chromatography column (25×0.25 mm) coated with TAB-CB in an HP 5890-II apparatus equipped with a split-splitless injector and a flame ionization detector (Chrompack) according to the procedure described by Alonso, 1993.

2.6. Tocopherol analysisTOP

Tocopherols were quantified by high performance liquid chromatography (HPLC). The HPLC system consisted of a low pressure quaternary pump HP-1050, a Rheodyne injection valve (20 mL loop), a thermostatic furnace and a fluorescence detector RF-235 (Shimadzu, Kyoto, Japan). Separation was performed in a 250×4 mm particle size 5 μm Lichrospher Si-60 (Merck, Darmstadt, Germany) column. The column and detector were kept at a temperature of 40 °C. The mobile phase was n-hexane/2-propanol 99/1 (v/v). The flow rate was supported at 1 mL min⁻¹ isocratic elution. Quantification was carried out by a calibration system based on standards (IUPAC, 1987d).

2.7. Statistical analysisTOP

In order to determine significant differences among species, the main effects analysis of variance (ANOVA) was performed according to the general lineal model procedure. The data were analyzed using the statistical package CSS: Statistica 8.0 software (StatSoft Inc., 1995). The compounds identified were considered as chemical descriptors. A data matrix was built where rows are the samples and columns are the variables. Each element of this matrix, expressed as xij, corresponds to the content of compound j for the sample i. In order to obtain a better understanding of the data trends, pattern recognition methods such as principal component analysis (PCA) and linear discrimination analysis (LDA) were applied.

3. RESULTS AND DISCUSIONTOP

3.1. Oil composition analysisTOP

Table 1 shows the content in oil, moisture and seed yield for Blanquina, Raxao, Collaos, Durona, Riega, Solarina and Limón Montés apple seeds. The highest oil content corresponded to Collao apples at 22.73±0.81 g 100 g⁻¹ seed, followed by Raxao and Riega at 20.19±0.74 g 100 g⁻¹ seed and 19.67±0.85 g 100 g⁻¹ seed, respectively. These values are slightly higher than those reported by Tian et al. (2010) in a study of antioxidant and antimicrobial activities of the oil in apple seeds. The moisture contents ranged between 7.13±0.12 g 100 g⁻¹ seed and 9.26±0.12 g 100 g⁻¹ seed in Limón Montés and Blanquina. The highest seed yields were for Limón Montés (7.49±0.21 g of seed 10 Kg⁻¹ of fruit) and Solarina (7.21±0.17 g of seed 10 Kg⁻¹ of fruit). Table 2 shows the fatty acid compositions of the apple seed oil. Linoleic acid was found to be the main component in the oil of the Limón Montés species (60.78±3.07 g 100 g⁻¹ of oil) followed by the Riega species (60.01±3.41 g 100 g⁻¹ of oil) these contents are higher than those found by other authors (Yukui et al., 2009; Tian et al., 2010). The highest concentrations of oleic (36.57±2.08 g 100 g⁻¹ of oil) and stearic acids (2.30±0.15 g 100 g⁻¹ of oil) were found for the Collaos species. These data are somewhat in agreement with those found by Yu et al. (2007) in a study of the proximate composition of the apple seed and the characterization of its oil. Almost all phospholipid (PL) classes were identified in the phospholipid fraction (Table 3). Phosphatidylcholine (70.58±3.85% of the total PL) was found to be the main constituent in the blanquina species followed by the Collaos species (55.55±2.96% of the total PL). Phophatidiletanolamine (30.80±1.54% of the total PL) and phophatidilinositol (31.80±0.25% of the total PL) showed a higher value for Limón Montés and Raxao species than the other species. These values were somewhat higher than those obtained by Zlatanov et al. (1997) in a study of the phospholipid composition of Rosaceae seed oil. Table 4 shows the content in sterols in Blanquina, Raxao, Collaos, Durona, Riega, Solarina and Limón Montés apple seed oils. β-sitosterol was the major sterol in all apple seed varieties. Solarina seed oil was the one with the highest content at 558.52±9.42 mg 100 g⁻¹ of oil while Blanquina seed oil was the one that presents the lowest values at 166.55±1.89 mg 100 g⁻¹ of oil. Likewise, Limón Montés seeds had the highest content in campesterol at 49.25±1.16 mg 100 g⁻¹ of oil and Blanquina the least at 17.70±0.22 mg 100 g⁻¹ of oil. The stigmasterol content in Solarina was worth noting at 4.05±0.37 mg 100 g⁻¹ of oil. With regard to Δ-5 avenasterol, Solarina seed contained 56.59±1.94 mg 100 g⁻¹ of oil, while Blanquina had only 14.37±0.43 mg 100 g⁻¹ of oil. Table 5 shows the content in tocopherols in the studied apple seed oils. Raxao presented the highest content in total β-tocopherol at 125.29±12.62 mg Kg⁻¹ of oil and Limón Montés seed oil the lowest content at 79.21±4.25 mg Kg⁻¹ of oil. α-tocopherol was the prominent tocopherol in Limón Montés, with a content of 84.68±5.61 mg Kg⁻¹ of oil followed by the Solarina species with 84.76±5.02 mg 100 Kg⁻¹ of oil. The main triglyceride composition of apple seeds (Table 6) was made up of LLP (41.17–39.32%) followed by LLL (27.12–17.80%). The triglycerides present in small amounts were SOL (0.81–0.50%) and PPO (0.31–0.17%). Triglycerides with three unsaturated acids presented the highest proportions, followed by those that have two unsaturated acids.

3.2. Principal component analysisTOP

A PCA was applied to the data set to obtain linear combinations of the variables called principal components (PCs). The first principal component (PC1) expresses the largest variability and each successive PC represents as much of the residual variability as possible. The results of this analysis are included in Table 7. As can be seen, the two first PCs obtained present eigenvalues greater than 1, the explanation of the original variance for each PC is shown in this table. For fatty acids PC1 is highly influenced by C16:1, C18:0, C18:1, C18:2, C20:0, C20:1 and C22:0. In the case of PC2, the most contributing variable is C18:3. In the case of triacylglycerols, PC1 is highly influenced by LLL, LLO, SOL, SOLn, OOO and OOP. The most contributing variables to PC2 are OOL and PPO. For phospholipids PC1 is highly influenced by the total content, PC, PE and PI. The most contributing variable to PC2 is LPC. In the case of unsaponifiable compounds, brasicasterol, campesterol, stigmasterol, sitosterol, sitostanol, avenasterol, stigmastenol, avenasterol, and tocopherol exert a great influence on PC1. The most contributing variables to PC2 are cholesterol and γ-tocopherol.

The two-dimensional plots of the PCs can be used to reveal the internal structure of the data and visualize data trends Jolliffe (2002). Figure 1 shows the distribution of the samples in the plane formed by the two first PCs, in this figure a good separation between specie samples can be observed, along with their remarkable distribution of positive or negative values according to the scores of PC1 and PC2. Taking into account these results, linear discriminant analysis was applied to obtain an adequate classification model.

3.3. Linear discriminant analysisTOP

The following step in our analysis was to apply the procedures of supervised pattern recognition in order to achieve a better separation. Accordingly, LDA (Coomans, et al., 1979) was applied to the data set to obtain suitable classification rules for the samples. The corresponding discriminant functions were calculated as linear combinations of the chemical descriptors. The criterion used for feature selection was the backward stepwise approach. In a first run, all the variables are present in the model; in each step the variable with least discriminant power, according to the Wilks’ l statistic test (Gardine, 1997), is rejected. Figure 2 shows the sample distribution in the space of the different obtained discriminant functions. The seven oil samples appear completely separated.

4. CONCLUSIONSTOP

On the basis of the present study, it may be concluded that the nutritional properties of the studied apple seed oils possess a high added value. The composition in lipids of these seeds presents high contents in polyunsaturated fatty acids C18:2, sterols (β-sitosterol), phospholipids (phosphatidilcoline), and tocopherols (alfa-tocopherol). The extraction of these apple seed oils is an option to obtain high added value oils and an additional channel for their use as rich in bioactive nutraceutical compounds to assess the potential use of the oils in foodstuffs. From the point of view of authenticity, it could be a useful tool to know the trazability of apple seed oil with DO compared to others in order to avoide adulteration through the study of the minor oil components.

ACKNOWLEDGMENTSTOP

This research was conducted with support from the regional Government of the Principado of Asturias (Spain). The authors would like to acknowledge Isabel Cuesta, Ana Hernández and Jorge Alvarez for their technical assistance and the Serida and San Martín de Llanes for the selection of DO apples. The authors also wish to thank Dra. M. Prieto from the University of Oviedo, a group associated with the Spanish Scientific Research Council (CSIC).

REFERENCESTOP