A fatty acid methyl ester mix C14−C22 (wt. % varied conc., ampule of 100 mg), DL-α-tocopherol acetate (≥ 96%), (+)-γ-tocopherol (≥ 96%), rac-β-tocopherol, δ-tocopherol and tocotrienol (α-, β-, γ- and δ-) analytical standards were purchased from Sigma-Aldrich (St. Louis, USA). All chemicals, reagents and solvents used were of analytical or HPLC grade and obtained from VWR International (Poole, UK), Sigma-Aldrich (St. Louis, MO, USA) or Merck (Darmstadt, Germany) companies. Deionized water (>18 MΩ.cm) was obtained from a Milli-Q system (Millipore, Brussels, Belgium). Standards for sterol (spinasterol, β-sitosterol, ∆7,22,25-stigmastatrienol, ∆7,25-stigmastadienol, ∆7-stigmasterol, ∆7-avenasterol), phenolic acid (arbutin, gallic acid, 5-hydroxymethyl furfural, protocatechuic acid, procyanidin, 4-hydroxybenzoic acid, catechin, chlorogenic acid, procyanidin B2, caffeic acid, syringic acid, vanillic acid, p-coumaric acid, ferulic acid, sinapic acid, quercetin-3-rutinoside, quercetin-3-galactoside, quercetin-3-glucoside, quercetin-3-xyloside, quercetin-3-arabinoside, quercetin-3-rhamnoside, isorhamnetin-3-glocoside, phloridzin, cinnamic acid, quercitin, phloretin) and carotenoid bioactives, and silylating agents (pyridine, hexamethyldisilazane and chlorotrimethylsilane) were obtained from Sigma-Aldrich (St. Louis, USA). Four different cold-pressed pumpkin seed (Cucurbita pepo L.) oil samples were studied to determine their bioactive compound contents. Commercially available pumpkin seeds (Cucurbita pepo L.) harvested in the regions of central Anatolia of Turkey (Celtik-Konya, Cumra-Konya, Iceri Cumra-Konya and Polatlı-Ankara), were purchased from local suppliers. Raw pumpkin seeds were of domestic origin and harvested in mid-autumn of 2015, and used for screw-pressing. The raw pumpkin seeds were stored in closed bags in the dark at 4 °C, until the extraction process.

Oils were extracted from sampled seeds according to the screw-pressing process. This pressing is performed by directly pressing raw/dried seeds onto a continuous screw press at low temperatures. A laboratory-scale screw press machine (15 kg seed.h^-1 capacity, single head, 2hp, 1.5 kw power) in our laboratory was used for the pressing of seed samples. 40 rpm screw rotation speeds and 40 °C ±0.05 temperatures were selected as process parameters. The temperature of the extracted oil was measured throughout the experiments with a digital thermometer. For screw-pressing, 10 kg of Celtik-Konya, 7 kg of Cumra-Konya, 10 kg of Iceri Cumra-Konya and 8 kg of Polatlı-Ankara region raw pumpkin seed samples were used. At the end of the cold-pressing process, the oil and meal with oily cake were collected, and then the oil was filtered to remove suspended materials. After one day sedimentation in closed dark bottles at temperatures below 15 °C, the extracted oil was separated from the sediment by decantation, and then it was centrifuged in a centrifuge system (Sigma 2-16K, Germany) for 20 min at 10 °C. The cold-pressed oils were stored in dark green glass bottles. The bottles were kept in the refrigerator at 4 °C, until the analysis.

The fatty acid composition analysis was performed in an Agilent 7890A gas chromatography system equipped with a flame ionization detector (FID) (Agilent Tech. Inc., USA). Fatty acid methyl esters (FAMEs) of the oil samples were prepared according to the method described by Arslan et al., (2017) with a 2N potassium hydroxide solution in methanol. Briefly, 0.1 g of the PSO was weighed into a centrifuge tube and dissolved in 10 mL of hexane. Then, 0.1 mL of 2N potassium hydroxide solution was added and the tube was vigorously shaken for 60 s. Then, the mixture was subjected to centrifugation for 5 min, at 2000 rpm. In the last stage, the upper layer was moved to a vial for analysis. The operating conditions for FAME analysis were carried out according to Gu et al., (2011) with a few modifications. Separations were performed on a HP-5MS (30m×0.25mm×0.25μm, Agilent) capillary column. The injection volume was 1 μL with a split ratio of 100:1. The inlet temperature was 230 °C. The oven temperature program was started at 60 °C, then linearly increased to 175 °C (15 °C/min) and then linearly increased to 240 °C (2 °C/min), and kept at this temperature for 10 min. Fatty acids were identified based on the retention times of a FAME mix C14:0–C22:0 standard injected under the proposed method conditions and expressed as g/100 g of totally fatty acids. The data were processed with the aid of Agilent ChemStation 2001–2010 data processor.

Phytosterol and squalene determination was performed according to the method described by Rabrenovic et al., (2014). The phytosterol derivatives were analyzed using an Agilent 7890N gas chromatography system equipped with FID and 5975C MSD detector systems (Agilent Tech. Inc., USA). 7.5 mL of a cholesterol solution as an internal standard, with a concentration of 1.33 mg/mL prepared with methylene chloride, were put into 4.5 g of sample. After the evaporation of the solvent, the sample was saponified with 20 mL of a 6 M KOH aqueous solution. The blend was diluted with 30 mL of ethanol (with up to 50 mL/L diethyl ether), and then it was refluxed for 90 min at 70 °C. After saponification, 30 mL of water were added, and the unsaponifiables were extracted three times with 45 mL of petroleum ether and 45 mL of diethyl ether, respectively. The extracts were merged, washed three times with 20 mL of 0.5 M KOH and 20 mL of 50 mg/L NaCl until the neutral pH of washing water. The organic fraction was dried with Na₂SO₄ and filtrated. The phytosterol derivatives were prepared by adding 1.5 mL of drypyridine, 0.2 mL hexamethyldisilazane, and 0.1 mL chlorotrimethylsilane in a vial and heated to 80 °C for 30 min. The derivatized parts were transferred into a vial and after derivatization they were analyzed by GC-MS within 6 h. A mixture of derivized phytosterol standard solutions was also prepared by the same procedure.

GC-MS separations were performed on a HP-5MS (30m×0.25mm×0.25μm, Agilent) capillary GC column. The flow rate of carrier gas (helium) was 1.5 mL/min and the injected sample volume was of 1 μL in the splitless mode. The temperature program was as follows: a gradient of 3 °C/min from 60 °C to 300 °C, and held at this temperature for 10 min. MS data were obtained in the EI mode with a scan range of 30e550 m/z; the source temperature was 230 °C, quadrupole temperature was 150 °C and hold-up time was 3 min.

The tocol (tocopherol and tocotrienol) profiles of the PSOs were identified using a normal-phase liquid chromatographic (NP-HPLC) method (Arslan et al., 2016). An Agilent 1200 series HPLC system (Agilent Technologies Inc., USA) was equipped with a G1321C model fluorescence detector (FLD) and G1316A model column compartment set at 20 °C. Separation was achieved on a LiChrospher 100Å diol column (250×4.0 mm id., 5μm, Teknokroma, Barcelona) equipped with a guard column (75×4.0 mm i.d.,5μm, Teknokroma, Barcelona). The data were recorded using an Agilent ChemStation 2001–2010 data processor. Before HPLC analysis, one gram of oil was weighed, then dissolved in 10 mL hexane, centrifuged and filtered through a PTFE syringe filter (0.45mm). 5 μL of the sample were injected and a mixture of hexane:2-propanol (99.4:0.6, v/v) was used as a mobile phase under isocratic conditions with a flow rate of 0.7 mL.min^-1. Detection of tocol isomers was performed at λ_exc= 295 nm and λ_em= 330 nm by FLD. Commercial standards of all of the tocols were used for calibration, and curves ranging from 0.1 to 3.0 mg.mL^-1 were linear for all tocols (R² > 0.99). Contents of each tocol were expressed in mg.100g^-1 oil.

The carotenoid composition analysis was performed according to a previously described method with some modifications (Parry et al., 2006). One mL of PSO was dissolved in 160 mL of methanol:tetrahydrofuran (THF) (50:50, v/v), and the carotenoid composition was determined using a HPLC-DAD-ESI/MS/MS (high performance liquid chromatography-diode array detector/ electrospray ionization/ tandem mass spectrometry) system. An Agilent 1100 HPLC system (Agilent Tech. Inc., USA) and TSQ Quantumtandem mass spectrometry (Thermo-Finnigan, San Jose, CA) was equipped with an ESI interface system and used for the analyses of β-carotene, lutein, cryptoxanthin, and zeaxanthin. The separations were performed using a Zorbax-SB C18 HPLC column (50×1.0 mm i.d., 3.5 μm) (Agilent Tech. Inc., USA). A mobile phase of solvent A (ultra-pure water) and solvent B [methanol: acetonitrile:2-propanol (54:44:2, v/v/v)] was used with a gradient program as follows; the gradient was linear from 50% to 99% of solvent B, and the flow rate was increased from 0.2 to 0.27 mL/min in the first 10 min, and 99% of solvent B and flow rate of 0.27 mL/min for 10 min. Prior to the next injection, the column was re-equilibrated for 10 min with 50% of solvent B. The quantification of carotenoids was determined using the total ion counts compared to external standards of the individual carotenoids. The wavelength of DAD detection was set at 440 nm. The identification of carotenoids was achieved by comparing the retention time and SRM analysis of the sample peaks with that of the commercial carotenoid standards. The TSQ Quantum was operated in the positive-ion mode under the following conditions: nitrogen (>99.7 %) was used for sheath gas and auxiliary gas at pressures of 30 psi and 5 units, respectively. The temperature of the heated capillary was kept at 300 °C, and the spray voltage of ESI was set at 4.5 kV. A collision-induced dissociation (CID) was performed using argon as the collision gas at the pressure regulated to more than 1.0 mTorr, and the applied collision offset energy was set to - 45 eV.

Total phenolic compounds (TPCs) were extracted from the cold-pressed PSOs according to the method proposed by Feng-Lin et al., (2010). Briefly, 200 μL of the extract obtained from the oil sample and gallic acid (GA) standards (concentration range between 0.05 to 3 mg GAE), were transferred into centrifuge tubes. Then, 1 mL of the Folin-Ciocalteau (FC) reagent and 2 mL of sodium carbonate (7.5%, w/v) were added, diluted with ultra-pure water to 7 mL and mixed thoroughly. After shaking, the mixtures were then allowed to stand for 2 hours in the dark before measuring the absorbance in Lambda-25 UV-Vis spectrophotometer (PerkinElmer Inc., Waltham, Massachusetts, U.S.A.) at 765 nm. The measurements were performed in triplicate and the results were expressed as milligrams of gallic acid equivalents (GAEs) per 100 grams of sample. The contents of TPCs in the studied oils is calculated and reported by using calibration graph.

The radical scavenging activity of PSO extract was determined according to the method described by (Molyneux, 2004). In brief, 500 μL of the extract obtained from the oil sample and GA standards (0.02−0.14 mg GAE) were transferred into a centrifuge tubes. Then, 3 mL of 6×10⁻⁵ M solution of 2,2-diphenyl-1-picrylhydrazyl (DPPH) were added and mixed thoroughly. The absorbance at 517 nm was measured after reaction at room temperature in the dark for 30 min. A control sample, without oil, was prepared and measured daily. The DPPH^’ concentration in the reaction medium was calculated from the following calibration curve, determined by the linear regression:

The radical scavenging activity toward DPPH’ was expressed as the % reduction in DPPH’ concentration by the constituents of the oils:

where [DPPH’]₀ and [DPPH’]₃₀ were the concentrations of DPPH’ in the control sample (t=0) and in the test mixture after 30 min reaction, respectively (Molyneux, 2004).

In last stage the antioxidant activities of the oil samples were determined according to the method of (Prieto et al., 1999). 300 μL of the extract obtained from the oil sample and GA standards (0.2−1.0 mg GAE), were transferred into centrifuge tubes. Three mL of reagent solution (mixture of 25 mL 0.6M H₂SO₄, 25 mL 28 mM Na₂HPO₄.12H₂O and 25 mL 4 mM ammonium molybdate) were added. The tubes were incubated at 95 °C for 90 min. After the mixtures had cooled to room temperature, the absorbance was measured at 695 nm. All analyses were performed in triplicate and the results were expressed as milligrams of GAEs per 100 grams of sample.

All analyses were carried out in triplicate and data were expressed as the mean ±standard deviation. All statistical analyses were performed by means of Excel (Microsoft, 2007) and OriginPro8 (OriginLab, USA) databases.

Table 1 presents an overview of the profile and content of fatty acids in the cold-pressed PSO samples. As can be seen in Table 1, the studied PSOs belong to the oleic-linoleic acid type of oil. A total of fourteen cis fatty acids and three trans fatty acids (TFA) in PSO was determined. Fourteen main components; seven saturated fatty acids (SFA), four monounsaturated fatty acids (MUFAs) and three polyunsaturated fatty acids (PUFAs) were identified. The main fatty acids in the studied oils were identified in the order of; linoleic (C18:2), oleic (C18:1), palmitic (C16:0), and stearic acid (C18:0). The content of polyunsaturated linoleic acid ranged from 53.19 ±0.04% to 53.27 ±0.04% and was followed by oleic acid which was also present in high contents varying from 27.52 ±0.02% up to 27.59 ±0.02%. As reported in previous studies, the high content of linoleic acid is an important nutritional aspect of oil because it is an essential fatty acid for humans and it is involved in the building of cellular membranes, vitamin D, and a variety of hormones (Jiao et al., 2014). The presence of high amounts of linoleic acid also suggested that the analyzed PSOs could be used as a good source of essential fatty acids. High oleic acid contents, detected in all the analyzed oil samples, is also favorable from a cardio-protective and nutritional viewpoint (Fruhwirth et al., 2007).

In all the pumpkin seed oils, palmitic and stearic acids were detected in relatively high amounts ranging from 11.90±0.01% to 11.99±0.01% and 5.26±0.01% to 5.29±0.01%, respectively. Trace amount of the fatty acids were also identified and these data are similar to those reported by Kostadinovic Velickovska et al., 2015. On the other hand, there are insignificant differences in fatty acid profiles between these results and those reported by Nederal et al., (2014). The differences in profile and amount of certain fatty acids depended on the origin of specific pumpkin (Cucurbita pepo L.) seed oil cultivars, climatic conditions, and post-harvest management (Rabrenovic et al., 2014).

The nutritional values of the studied PSOs were also calculated by their total MUFA (∑MUFA), total PUFA (∑PUFAs) and total SFA (∑SFA) contents (Jiao et al., 2014; Rezig et al., 2012). To the best of our knowledge, a higher intake of MUFAs may emerge as health-promoting food choices on the basis of both cardiovascular benefits and their capacity to decrease the risk of opportunistic diseases. For this reason, the vegetable oil industry has focused attention on vegetable oils with high oleic FAs or MUFAs (Jafari et al., 2012). Herein, the detected amounts of ∑MUFA ranged from 27.77±0.04% to 27.86±0.05% (Table 1). PUFAs are more susceptible to oxidation reactions, while SFAs are more stable, so the ratio of ∑PUFA/∑SFA is generally defined as the best indicator for verifying the degree of oil deterioration and the high ratios signified that the samples were nutritionally valuable. ∑PUFA/∑SFA values have also favorable nutritional implications and a positive physiological effect in the prevention of both heart and cancer diseases (Arslan et al., 2017; Omah et al., 2000). Herein, the detected amounts of ∑PUFA ranged from 53.60 ±0.06% to 53.73 ±0.05%, and ∑SFA ranged from 17.49 ±0.02% to 17.61 ±0.03% (Table 1). The detected ratios of ∑PUFA/∑SFA ranged from 3.05 ±0.05, 3.05 ±0.05, 3.07 ±0.04, and 3.04 ±0.05, respectively. Trans fatty acid contents were also detected and reported as 0.37 ±0.01%, 0.40 ±0.01%, 0.33 ±0.01% and 0.36 ±0.01% for the PSO samples of Celtik-Konya, Cumra-Konya, Iceri Cumra-Konya and Polatlı-Ankara, respectively. As expected, cold-pressed PSOs contain much less TFAs on average then refined seed oils (Jafari et al., 2012). Therefore, the fatty acid profiles of PSO can be an indicator of its stability and nutritional quality.

Phytosterols are of interest due to their cholesterol-lowering and physiological effects on human health. The quantities of phytosterols in vegetable oils have been used for the classification and assessment of their quality (Rabrenovic et al., 2014; Kostadinovic Velickovska et al., 2015). The cultivation and environmental conditions influence the concentration of phytosterols. Based on the position of the double bond in the ring, phytosterols could be classified as ∆5- and ∆7-sterols. Most species contain dominant ∆5-sterols, while ∆7-sterols are characteristic of only a few plant species and are specific bioactive compounds to pumpkin seed (Cucurbitaceae) oil (Rabrenovic et al., 2014). The phytosterol profile of the PSOs is given in Table 2 and regression analyses were also performed. There was a good correlation between the concentration of phytosterol bioactive compounds and detector intensities. Regression analysis between R²=0.9740 and 0.9990 values were indicated as satisfactory correlations for phytosterols. As can be seen in Table 2, ∆7,22,25-stigmastatrienol was the major phytosterol in the seed oils ranging from 28.3 ±0.1 to 31.6 ±0.3 mg/100 g of the total phytosterol contents. Spinasterol together with β-sitosterol were the second-most abundant phytosterol, showing 42.4 ±2.5-47.20 ±2.1 g/100 g of total phytosterols. They were followed by ∆7-stigmasterol with 15.1 ±0.1-17.1 ±0.6 g/100 g, ∆7,25-stigmastadienol with 4.9 ±0.2-5.9 ±0.4 g/100 g, and ∆7-avenasterol with 3.20 ±0.1-4.00 ±0.3 g/100 g of total phytosterols. Significantly higher contents of total phytosterols were determined and total phytosterol contents for the PSO samples which ranged from 782.1 ±9.7 to 805.2 ±11.3 mg/100 g oil. Total phytosterol content in the studied samples is significantly higher than the data reported in the literature (Rabrenovic et al., 2014; Kostadinovic Velickovska et al., 2015). Thus, the high total phytosterol contents in the PSO samples from different regions of Turkey may also be related to the species of pumpkin, although the techniques used for oil extraction from seeds and the isolation procedure of phytosterols from oil may also considerably influence data.

Squalene is a natural acyclic compound produced by plant sources (mainly vegetable oils) and shark liver oils. This bioactive compound has been recommended to be an essential part of the human diet as it may be an inhibitive component that protects humans from several diseases (Ryan et al., 2007). The obtained results for the content of squalene in cold-pressed PSO samples are given in Table 2. The squalene content in the samples ranged from 591.3 ±10.6 to 632.5±11.4 mg/100 g oil. These results are in good agreement with previously reported data for seed oils from different pumpkin cultivars (Rabrenovic et al., 2014). Differences among the contents of squalene could be attributed to the applied temperature, oil extraction procedure, and seed origin. Thus, the high content of squalene compounds contributes to the significant nutritional and therapeutic value of the cold-pressed PSOs from different regions of Turkey.

Tocols are very significant non-glyceride ingredients of vegetable oils. Tocols represent a family of tocopherols (α-, β-, γ-, and δ-T) and tocotrienols (α-, β-, γ-, and δ-TT) and natural lipophilic antioxidants mainly found in vegetable oils. They are believed to be primary chain-breaking antioxidants in free radical chain reactions or in converting lipid radicals to more stable products, therefore improving the antioxidant potentials of oils. Due to their role in the protection against the oxidative deterioration of PUFAs, the content and profile of tocol isomers in vegetable oil are exceptionally important. Tocols have also been associated with the reduction of heart disease, delay of Alzheimer and prevention of cancer diseases (Mariod et al., 2009). Table 3 presents the tocol profile and content of cold pressed PSOs from central Anatolia region of Turkey. As can be seen in Table 3, significantly high amount of tocols were determined and total contents for the studied oil samples ranged from 94.29±0.34 mg/100 g oil to 97.79±0.76 mg/100g oil. The amounts of tocols are higher than other reports and, in mentioned studies the maximum contents of tocols were detected as 70.90 mg/100 g oil, 35.10 mg/100 g oil, and 77.05 mg/100 g oil in pumpkin (Cucurbita pepo L.) seed oils, respectively (Stevenson et al., 2007; Nederal et al., 2012; Siger et al., 2008). The contents of these natural antioxidants in vegetable oils depends on a number of factors, such as variety characteristics, the kind of oil, extraction process, and they are possibly managed by the content of unsaturated FAs. The content of tocols in oils is also significantly affected by the processing conditions of the oil production (Rabrenovic et al., 2014).

In all the analyzed PSO samples, α-T, β-T, γ-T, β-TT and γ-TT isomers were identified, and α-TT isomers were detected only in Celtik-Konya and Polatlı-Ankara PSO samples (Figure 1). As is well known, the γ- and δ-tocol isomers possess a much higher antioxidant activity than the α- and β- isomers; however, α-tocol is considered to have a higher vitamin effectiveness than other tocol isomers (Nederal et al., 2012). As can be seen in Figure 1, β-T was identified as the major tocol isomer in all the varieties, and its contents ranged from 83.00 ±0.38 to 86.23 ±0.87 mg/100 g oil, while the other tocol isomers were present in relatively lower quantities. The amounts of α-tocopherol detected in the PSOs ranged from 3.66 ±0.03 to 4.39 ±0.04 mg/100 g oil.Thus, the obtained results demonstrated that the studied PSO samples were good sources of α-, β- and γ-tocols.

Phenolic acids are known as minor antioxidant constituents, and play a determinant role due to their characteristics such as flavor, quality and resistance against the oxidation process (Nyam et al., 2009). Common vegetable oils contain different kinds and contents of phenolic acids, and data about the phenolics in cold-pressed oils are very rare in the literature. In this research, six phenolic acids; protocatechuic acid, caffeic acid, syringic acid, vanillic acid, p-coumaric acid, and ferulic acid were identified. The chromatogram obtained from the analysis of a standard mixture of 27 phenolic acids at 280 nm, is illustrated in Figure 2. The concentration of the detected phenolic acids in the cold-pressed PSOs is given in Table 4 and regression analyses were also performed. There was a good correlation between the concentration of phenolic bioactive compounds and detector intensities. Values for the regression analysis between R²=0.9510 and 0.9950 are indicated as satisfactory correlations for phenolics.

As can be seen in Figure 2, the proposed method permits the baseline separation for all components including quercetin glycosides, hydroxybenzoic acids, hydroxycinnamic acids, flavanol monomers, dimers, and phloridzin. The obtained results showed that syringic acid (7.62 ±0.13-8.03 ±0.16 mg/100 g oil) was the predominant acid in all the samples, followed by ferulic acid (4.72 ±0.22-5.17 ±0.38 mg/100 g oil). Cumra-Konya cold-pressed PSO sample had the highest amount of phenolic acids (23.98 ± 0.46 mg/100 g oil) followed by Iceri Cumra-Konya cold-pressed PSO sample (23.66 ±0.34 mg/100 g oil). Minor amounts of phenolic acids such as protocatechuic acid, caffeic acid, vanillic acid and p-coumaric acids were also detected. The differences in the amounts of phenolic compounds may occur due to differences in the extraction methods, origin of seeds, environmental conditions and the treatments applied to the seeds before extraction (Alothman et al., 2009).

The carotenoid composition (β-carotene, lutein, zeaxanthin and cryptoxanthin) of the cold-pressed PSOs is alsogiven in Table 4. Zeaxanthin was the major carotenoid pigment present in all the studied PSOs with a range of 2.65 ±0.05-2.91 ±0.03 mg/100 g oil. The Cumra-Konya PSO sample contained the highest amount of total carotenoids (7.60 ±0.03 mg/100 g oil), followed by the Iceri cumra-Konya (7.21 ±0.05 mg/100 g oil) PSO samples. This PSOs also had the highest β-carotene content of 0.54 ±0.02 to 0.60 ±0.02 mg/ 100 g oil, among all the studied oil samples which was much higher than that observed in cold-pressed seed oils (0.01–0.24 mg/100 g), and also much higher than that of refined oils (0.01–0.12 mg/100 g) (Parry et al., 2006). The obtained results demonstrated that different contents and profiles of carotenoids are influenced by the genetics and phase of seed maturation, and also by oil extraction and storage conditions. Thus, these data suggest that the studied cold-pressed PSOs may serve as dietary sources of carotenoids, especially zeaxanthin (Parry et al., 2006).

Phenolic compounds which have unique chemical characteristics posing low-energy radicals through stable hybrids are thought to be good antioxidants for foodstuffs. They have also proven to possess important roles in their quality and nutritional values. Natural oils contain more polar compounds that have mainly phenolic characteristics. These phenolics are secondary components and represent only a minor fraction of the oil. Phenolic compounds may also exert high biological activity far beyond their antioxidant properties, so the determination of phenolic compounds in vegetable oil is of great interest (Latif et al., 2011). As can be seen in Table 5, the content of total phenolic compounds (TPCs) in cold-pressed PSOs is in the range of 3.96 ±0.13–5.82 ±0.15 mg GAE/100 g. The content of TPCs are in the following order: 4.32 ±0.09 mg GAE/100 g, 5.82 ±0.15 mg GAE/100 g, 3.96 ±0.13 mg GAE/100 g, 4.32 ±0.12 mg GAE/100 g for PSOs from Celtik-Konya, Cumra-Konya, Iceri Cumra-Konya and Polatlı-Ankara, respectively. Values from the regression analyses with R²=0.9944, 0.9933 and 0.9972 are indicated as satisfactory correlations for total phenolic compounds, free radical-scavenging and antioxidant activity analyses, respectively. There was also a directly relative correlation between TPCs and total tocol content. It is clear that Cumra-KonyaPSO is the richest cold-pressed oil in terms of both the TPCs and tocol content (Tables 3 and 5). According to the literature, the content of total phenolic compounds in oil is mainly related to the oil from roasted pumpkin seeds, and is reported to range from 0.28 to 2.46 mg GAE/100 g (Tuberoso et al., 2007; Siger et al., 2008). Andjelkovic et al., (2010) reported the contents of phenolics in PSO samples in a range from 2.47 to 5.09 mg GAE/100 g. Also, higher contents of phenolic compounds in a range from 7.33 to 128.84 mg GAE/100 g were reported by Jiao et al., (2014).

In order to identify the antioxidant activity of the studied pumpkin seed samples, 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical-scavenging activities of oil extracts were determined. The values of free radical-scavenging activity were detected in the following order; 5.70 ±0.13 mg GAE/100 g, 7.35 ±0.15 mg GAE/100 g, 6.32 ±0.11mg GAE/100 g, 6.67 ±0.14 mg GAE/100 g for the seed oils from Celtik-Konya, Cumra-Konya, Iceri Cumra-Konya and Polatlı-Ankara, respectively (Table 5). As shown in Table 5, the extract of Cumra-KonyaPSO exhibited superior antioxidant activity (7.35 ±0.15 mg GAE/100 g) compared with other oil extracts. As expected, Cumra-Konya PSO exhibited a higher antioxidant activity than others, which can be substantially explained by the higher tocol, carotenoidand phenolic contents of this oil than that in other studied oils. In fact, the antioxidant potential of vegetable oils is directly related to their tocol, carotenoid and phenolic compound contents (Fruhwirth et al., 2003; Latif, 2011).

The total antioxidant activities of the studied PSOs were determined at 695 nm. In the studied PSOs, the total antioxidant activity values were in the range of 26.67 ±0.97−38.89 ±1.41mg GAE/100 g. The lowest value for total antioxidant activity was determined for the Iceri Cumra-Konya PSO sample (26.67 ±0.97 mg GAE/100 g), whereas, the Cumra-Konya sample had the highest value (38.89 ± 1.41 mg GAE/100 g). In view of the oxidative stability of vegetable oils, the presence of natural bioactive compounds with anti-oxidative activity is a vital factor. Tuberoso et al., (2007) deduced that the changeability of the relationship between the antioxidant activity and profile of the seed oils could be characteristic of the differences in the contents of bioactive compound of the oils and their interactions. Frushwirth et al., (2003) also reported that PSO samples have 59% of their antioxidant activity due to the phenolic compounds, and 41% to the tocols. Similar to the other antioxidant activity data, Cumra-Konya seed oil exhibited higher antioxidant activity than others. As known from previous studies; the cultivar, genetic factors, extraction system, production and storage conditions of seed oils are quite critical factors for the contents of antioxidant compounds (Nawirska-Olszanska et al., 2013). Thus, the differences in antioxidant activity data of the PSOs in this study are predicted to be mainly derived from their cultivar variations and genetic factors.