From By-Product to the Food Chain: Melon (Cucumis melo L.) Seeds as Potential Source for Oils

Fruit-processing industries annually discard large volumes of fruit by-products. Thousands of tons of melon seeds could be recovered through the year from melon production. These seeds are an excellent source of vegetable oil with significant health-promoting properties due to their unsaturated fatty acid profile and high content of specific bioactive compounds. However, little information exists about the influence of melon cultivars and oil-extraction methods on oil characteristics. In this study, oils from nine different melon cultivars were evaluated. Additionally, two oil-extraction methods (screw and hydraulic press) were studied. Results showed that melon seeds may be used as a novel source of healthy oils. Higher-quality oils were obtained with the hydraulic press; however, low yields reduced industrial interest in this method. Oils extracted from the different cultivars showed high variability in the content of linoleic (51–69%) and oleic (15–34%) acids. Regarding vitamin E, γ-tocopherol was the main isoform found in melon-seed oils (99.81–456.73 mg/kg), followed by α- and δ-tocopherols. Significant concentrations of tocotrienols (α, β, and γ) were also found. Although all cultivars showed positive attributes, principal-component analysis (PCA) showed that Honey Dew and Blanco de Ribatejo could be specifically considered as a potential source of polyunsaturated oils with high concentrations of vitamin E.


Introduction
The melon fruit (Cucumis melo L.) belongs to the Cucurbitaceae family, and it is grown in tropical and subtropical regions of the world. Global melon production has continuously risen in the last decade, reaching the current annual production of about 31.2 × 10 6 tons. Melon processing in the industry generates large quantities of by-products that are usually discarded. Within those by-products, melon seeds account for 10% of total melon weight [1]. However, melon seeds are not considered as waste in all regions of the world. In some Arabian countries, they are roasted and directly consumed [2], and in India they are dried and used to add flavor to traditional dishes and desserts [3]. This traditional use of seeds is not applied to melon production in Europe, where melon seeds are rarely used in the food chain.
Previous studies carried out mainly on melons grown in some developing countries confirmed the interest in melon seeds as a possible functional ingredient [3][4][5]. In this regard, the nutritional composition of melon-seed cultivars grown in different countries, including Egypt [6], Brazil [7][8][9], from cv. Piel de Sapo were obtained from three different conditions: traditional cultivation, organic cultivation, and seeds from Protected Geographical Indication (PGI) Melon de la Mancha, kindly supplied by the Regulatory Board of the PGI (Tomelloso, Spain). Commercial melon seeds were obtained from local suppliers in Albacete (Spain). Seeds were cleaned and washed to remove sugars and any adhered residues. Then, seeds were dried at room temperature for several days until the seeds from all cultivars reached a moisture of less than 10%.
To evaluate the proportion of the peel with respect to the total weight of the seeds, 100 seeds were selected from each cultivar and manually peeled. The peels were weighted, and the proportion of the peel was calculated as the peel weight divided by the total seed weight.

Oil Extraction
Oil extraction was carried out by using a Komet Oil Press CA59G screw press (IBG Monforts Oekotec GmbH & Co. KG, Monchengladbach, Germany). One kilogram of unpeeled seeds was introduced directly into the press once the barrel was heated to 100 • C to ensure the correct extraction of oil [14]. Medium rotational-speed conditions were selected (49 rpm). Additionally, seeds from cv. Piel de Sapo PGI were subjected to extraction by using a hydraulic press (MECAMAQ model DEVF 80, Vila-Sana, Lleida, Spain). For extraction with the hydraulic press, 1 kg of ground unpeeled seeds was placed on the press, and the seeds were subjected to a pressure of 150 bar for 10 min. After pressing, oil was centrifuged to remove remaining solids. Oil samples were stored in dark glass bottles at 5 • C to avoid degradation until analysis.

Regulated Quality Parameters
Regulated quality parameters consist of free acidity and peroxide values. To determine free acidity, expressed as % of oleic acid, a solution of melon-seed oil dissolved in ethanol/ether (1:1) was titrated with a 0.1 mol/L potassium hydroxide ethanolic solution [22]. On the other hand, the peroxide value, expressed in milliequivalents of active oxygen per kilogram of oil (meq O 2 /kg), was measured according to European Union (EU) regulations [22]. Briefly, chloroform and acetic acid were added to an oil sample, mixed vigorously, and left to react with a solution of potassium iodide in the darkness. Then, the free iodine was titrated with a sodium thiosulfate solution [22].

Fatty Acid Profile
Fatty acid profile was measured according to Santos et al. [23]. Briefly, 2 mL of n-hexane was added to 0.02 g of oil to obtain fatty acid methyl esters (FAME) by cold transmethylation with methanolic potassium hydroxide. Then, 200 µL of methanolic potassium hydroxide solution (2 N) was added and vigorously mixed. Then, the supernatant was carefully transferred to a glass vial and analyzed by gas chromatography in a Shimadzu GC-2010 Plus Gas Chromatograph (Shimadzu, Tokyo, Japan). This was performed using a CPSil 88 fused silica capillary column (50 m × 0.25 mm i.d.), 0.20 µm film thickness (Varian, Middelburg, The Netherlands), and helium was used as the carrier gas (120 kPa). The used temperature program was a first step of 5 min at 140 • C, followed by an increase of 5 • C/min from 140 to 220 • C, and then maintaining at 220 • C for 15 min. The temperature of the injector and detector was 250 and 270 • C, respectively, and the split ratio was 1:50 with an injection volume of 1 µL. Lastly, each FAME was identified by direct comparison with a standard mixture (FAME 37, Supelco, Bellefonte, PA, USA). All analyses were performed in duplicate, and results are expressed as the relative percentage of each FA on the basis of relative peak areas.

Vitamin E Determination
Analogously, vitamin E values were determined by HPLC analysis in oil samples according to Alves et al. [24]. Briefly, about 20 mg of oil was diluted in 1 mL of n-hexane (HPLC-grade, Merck, Darmstadt, Germany), where 20 µg/mL of tocol was added as internal standard. Then, 20 µL was injected to perform the separation on a normal-phase SupelcosilTM LC-SI column (3 µm; 75 × 3.0 mm; Supelco, Bellefonte, PA, USA). The used equipment was an HPLC system (Jasco, Tokyo, Japan) equipped with an AS-2057 automated injector, a PU-2089 pump, and an MD-2018 multiwavelength diode array detector (DAD) coupled with an FP-2020 fluorescence detector (Jasco, Japan). They were programmed for excitation at 290 nm and emission at 330 nm. Lastly, the identification of the compounds was accomplished by a comparison with commercial standards. Analyses were performed in duplicate, and results are expressed as mg/kg of oil.

Color Determination
Oil samples were filtered, and color was measured using a UV/Vis Jasco V-530 spectrophotometer (Jasco Analytical, Madrid, Spain). Basically, oil samples were placed in quartz cuvettes (1 cm path length) for analysis, using N-hexane as the blank reference. The obtained values were used to calculate CIELAB chromatic coordinates: L* (brightness), a* (red-green component), b* (yellow-blue component) as recommended by the Commission Internationale de l'Eclairage (CIE, Wien, Austria) [25].

Statistical Analysis
Data are expressed as mean ± standard deviation of the obtained results for the selected cultivars. Data were analyzed using the t-test and Duncan's test. Statistical significance was defined for p < 0.05 (95% confidence level). Pearson's correlations were examined. To perform principal-component analysis (PCA) for melon-seed cultivars, those variables that had previously shown significant differences were used. The Kaiser-Meyer-Olkin (KMO) test for sampling adequacy was used. All statistical analyses were carried out using the SPSS program v. 23.0 for Windows.

Oil-Extraction Yield
First, for a clearer idea about oil yield, it is important to consider seed parts that do not contain a significant composition in oil, which may contribute to reducing the yield value. In this sense, seed peel is mainly composed of carbohydrates, especially fiber [10]. The melon-seed peel constitutes about 30-40/100 g of seed weight. Significant differences were found between cultivars (Table 1). Cv. Piel de Sapo grown under conventional production (not organic) showed the lowest proportion of seed peel (30.65/100 g) followed by the organic Piel de Sapo. Cv. Piel de Sapo grown under PGI conditions showed one of the highest proportions of seed peel (39.00/100 g). Table 1. Proportion of seed peels and oil-extraction yields obtained with screw press in selected melon cultivars. PGI, Protected Geographical Indication.

Oil-Extraction Yield (g/100 g seeds)
Amarillo Mean ± standard deviation; a-g different letters in same column represent significant differences, p < 0.05 between samples. Melon seeds were identified as a good source of oil, with percentages of lipids about 30.7%-32.3% [8][9][10]. Oil yields from melon seeds were high enough to encourage their use for oil-production purposes when a screw press is used (Table 1). Obtained yields using the screw press were statistically different depending on the considered cultivar. Cvs. Piñonet and Piel de Sapo PGI showed the highest values, 29.90 and 26.23 g per 100 g, respectively. The lowest values were reported in the Arizo cultivar (16.95/100 g seeds). Negative but not statistically significant (r = −0.576, p = 0.64) correlation was found between peel proportion and oil yield obtained with the screw press.

Oil-Extraction Methods, Oil Quality, and Color
To evaluate the differences regarding extraction systems, seeds from cv. Piel de Sapo PGI were subjected to oil extraction with two presses, a hydraulic and a screw press. Extraction with the hydraulic press was performed under room temperature, while extraction with the screw press requires previous heating to obtain optimal performance. The data regarding oil yield, regulated quality (acidity and peroxide index) and color of the oils obtained are shown in Table 2. The pressure system selected for oil extraction had significant influence on oil quality and yield. Oil extraction in cv. Piel de Sapo PGI using the screw press resulted in an oil yield of 26.23/100 g, while extraction with the hydraulic press was almost four times smaller (6.80/100 g). This low yield makes the hydraulic press unsuitable for obtaining an economic benefit. Therefore, for analysis of the profile of fatty acids and vitamin E for the rest of the cultivars, only oils extracted with the screw press were used, since this could be the most appropriate method for obtaining commercial oils in the industry. Regarding oil quality, Codex Alimentarius [26] does not have specific regulation for melon-oil quality standards. Results showed that oils obtained with the hydraulic press were of slightly better quality than that of oils obtained with the screw press. In all cases, the values in oils obtained using pressure systems were significantly lower than the values reported for melon-seed oils obtained using solvent extraction [7,11].
Melon oils show a light yellow color. Nevertheless, CIELAB color parameters showed differences attending to the used extraction method. Oils obtained with the screw press showed more intense yellow colors, with higher values for the b* parameter ( Table 2). Oil extraction using the screw press requires high temperatures applied on the barrel to ensure proper oil extraction [27]. The processing temperature of screw press compared to room temperature used in hydraulic extraction may affect oil pigment content [28]. Previous studies on plant oils showed that lutein, which provides the yellow color to oils, is more resistant to high temperatures than other pigments are, such as chlorophylls [29]. Furthermore, some studies even reported an increase in the content of lutein after the thermal processing of food products due to the inactivation of enzymes responsible for oxidizing carotenoids [30]. Although the total content of carotenoids in melon oil was reported to be low [9], the balance of carotenoids in melon oil could be the reason for the observed change of color in melon oils depending on the extraction method.

Fatty Acids
As previously reported, melon-seed oils are mainly composed of linoleic (50.67%-69.22%) and oleic (15.23%-33.96%) acids. Saturated fatty acids, mainly palmitic and stearic, accounted for less than 15.62% in all studied cultivars (Table 3). Our data support previous results about the high variability of the fatty acid profile in melon-seed oils [5,7,8,10,11]. Some cultivars, such as Tendral Valenciano and Tendral Verde, showed a high content of linoleic acid, 69.22% and 69.15%, respectively, in comparison with cv. Piel de Sapo PGI, which showed the lowest values. As widely reported in other unsaturated plant oils, linoleic and oleic content were negatively correlated [31,32].
Extraction method had no effect on the fatty acid composition of the oils. When seeds from cv. Piel de Sapo PGI were used, oils extracted with the screw and hydraulic presses showed slight differences in fatty acid profile (Table 3). Similarly, small differences in the content of linoleic acid were found when the results for cv. Honey Dew were compared to those of the study of Bora, Narain, and de MeIio [5], who used solvent extraction for the same cultivar. In all cases, the differences reported from the cultivars were determinants compared to those small differences that could be attributed to the oil-extraction method.

Vitamin E
Regarding vitamin E content, the studied melon-seed oils showed significant differences (Table 4). Cvs. Honey Dew and Blanco de Ribatejo showed the highest contents of vitamin E, with 530.62 and 468.19 mg/kg, respectively. γand α-tocopherols were the main components of vitamin E. The amounts of α-tocopherol (37.42-74.71 mg/kg) was significantly higher than those in some previous studies for specific cultivars [9,10], but were in accordance with the data reported by Górnaś and Rudzińska [12]. α-tocopherol is crucial for oil quality, as it is the form preferentially absorbed and accumulated in humans. Cv. Tendral valenciano, which showed a small concentration of vitamin E, was, however, the one with the highest content in α-tocopherol. As reported with regard to oils from dicotyledonous plants, tocotrienol content was low compared to tocopherol content [33]; however, it was higher than that reported in some oils with similar fatty acid profiles, such as walnut oil [34,35]. γ-tocotrienol was the main tocotrienol found, in agreement with the study of Górnaś, Soliven, and Seglin , a [20]. The total concentration of vitamin E in melon oils and the content of the specific isoforms could be used for the authentication of products containing melon oils and flours as functional ingredients.
Oil-seed extraction methods significantly affect the concentration of tocopherols in oils, especially if solvent extraction is compared to pressing [36][37][38]. Our results showed that significant differences appear in the concentration of tocopherol and tocotrienol forms, and in the total content of vitamin E. The vitamin E content of the Piel de Sapo PGI cultivar, extracted by hydraulic pressing, was lower than the content obtained with screw pressing; however, as reported for fatty acids, cultivar had a larger effect on vitamin E content than extraction method did.

Principal-Component Analysis
The reported variability in the oil composition of cultivars is reflected in principal-component analysis (Figure 1). Principal Component 1 (PC1) was mainly composed of fatty acids (C18:2n6c; C18:1n9c; C18:0) and tocotrienols (β-T3), while Principal Component 2 (CP2) was composed of tocopherols (α-T; γ-T). The negative correlation between linoleic and oleic acid was clear. Similar negative correlation could be observed for the α-and γ-tocopherols. As γ-tocopherol is the main component of vitamin E in melon-seed oils, and major interest exists for oils with a high content in unsaturated fatty acids, the cultivars in the top right of Figure 1 are the most interesting for melonseed-oil production. These are cvs. Charentais and Honey Dew. However, the provided information also encourages the production of oils with more monounsaturated fatty acids and with a higher concentration of the most active homologous α-tocopherol. In this case, the use of cvs. Piñonet and Piel de Sapo PGI, in the bottom left of Figure 1, is preferable.

Conclusions
Melon-seed oil is proposed as a highly valuable product that can be obtained from the byproducts of the agroindustrial processing of melons. This oil showed a high content of polyunsaturated fatty acids, mainly linoleic, and a high concentration of vitamin E. However, the effect of cultivar and oil-extraction method must be considered, as they have crucial influence in melon-oil characteristics. The oil-extraction method influences oil-quality parameters and oil due to

Conclusions
Melon-seed oil is proposed as a highly valuable product that can be obtained from the byproducts of the agroindustrial processing of melons. This oil showed a high content of polyunsaturated fatty acids, mainly linoleic, and a high concentration of vitamin E. However, the effect of cultivar and oil-extraction method must be considered, as they have crucial influence in melon-oil characteristics. The oil-extraction method influences oil-quality parameters and oil due to the processing temperature in screw-press extraction. However, low yields obtained by hydraulic-press extraction could be inconvenient for industries to obtain an economic benefit. On the other hand, the use of different cultivars results in oils with different degrees of unsaturation and vitamin E content. Regarding vitamin E content, melon-seed oil may be considered as a rich source of tocopherols and tocotrienols. The industrial extraction of oil from melon seeds is a feasible option to obtain high-quality oil in order to meet the current demand of vegetable oils for human nutrition.