Fatty acid methyl esters (FAMEs) were prepared in accordance with the method developed by AOCS Ce 2-66 (AOCS, 2000). The separation of FAMEs was performed by gas chromatography on a Shimadzu GC-2010 equipped with a flame ionization detector and the capillary column TR-CN 100 (60 m x 0.25 mm x 0.2 μm, Teknokroma, Barcelona, Spain). One μL of FAMEs was injected in split mode at a ratio of 50:1. Helium was used as carrier gas at a flow rate of 1 mL/min. The injector and detector temperatures were 250 °C. Separation was carried out at a temperature of 190 °C for 30 min. FAME identification was based on retention times as compared with those of a standard FAME mixture. The results were expressed as the percentage of total peak area. The analyses were carried out in triplicate.

Sterols were determined according to the AOCS Official Method of Ch 6-91 (AOCS, 2000). The sterols were analyzed on an Agilent 7890A gas chromatograph, equipped with a FID and a capillary column HP-5 (30 m × 0.32 mm id × 0.25 μm film thickness). The carrier gas was nitrogen with a flow rate of 1 mL/min and a split ratio of 5:1. The injector and detector temperatures were set at 300 °C. The oven temperature was held at 235 ^oC for 75 min. The identification of peaks was done by comparing the retention time of external standards (cholesterol, brassicasterol, campesterol, stigmasterol and β-sitosterol). The result of each sterol compound was expressed as percent concentration. LOD was 0.005% and LOQ was 0.05%. The analyses were done in triplicate.

Phenolic compounds were extracted from the oils as described by Yılmaz and Durmaz (2015). An oil sample (2 g) was dissolved in 5 mL of n-hexane and then the solution was extracted with 3 mL of aqueous methanol (80%). The mixture was blended in a vibration mixer for 3 min. After 5 min of centrifugation at 600 g the hydrophilic layer was filtered and the combined extract was brought to dryness in a rotary evaporator at 38 ºC. The TPC in the oil extract was determined colorimetrically at 760 nm, using the Folin-Ciocalteau reagent (Singleton et al., 1999; Ramadan et al., 2012) and expressed as gallic acid equivalents (GAE). Total carotenoids were determined according to the method of Silva et al. (2011). Carotenoids were measured at 455 nm and the results were presented as mg β-carotene per kg of oil. The analyses were done in triplicate.

The Tocopherols were determined following the AOCS Method Ce 8-89 (AOCS, 2000). External standards of α, β, γ and δ-tocopherols (Sigma-Aldrich, Inc., St. Louis, MO) were used to calculate the individual amounts of each tocopherol in the oil. An aliquot of sample (20 µL) was injected into a Shimadzu HPLC system equipped with a PDA detector and a silica gel column (Intertsil SIL 100 A, with a 5-mm particle diameter, 4.6 x 250 mm in size, Tokyo, Japan). The separation of tocopherols was achieved with an isocratic elution of hexane: isopropanol (99.5:0.5, v/v) at 1.0 mL/min flow rate. The tocopherols were measured at 292 nm and expressed as mg/kg. The analyses were done in triplicate.

The DPPH· method described previously (Espín et al., 2000; Yılmaz and Durmaz, 2015) was used for the determination of the radical scavenging capacity of the oil samples. In brief, 100 µL of ethyl acetate extract from the oil was added to 2.9 mL of DPPH· methanol solution (25 mg/L). The mixture was shaken vigorously and left in the dark for 30 min. The absorbance was measured at 520 nm against pure ethyl acetate (blank) using a spectrophotometer in a 1-cm quartz cell. The radical scavenging activity was expressed as the inhibition percentage and was calculated using the following formula:

The obtained data were used to calculate the α-tocopherol and trolox equivalents that correspond to the oil. The analyses were done in triplicate.

The antioxidant capacity of the oils was estimated using the ABTS⁺ method (Yılmaz and Durmaz, 2015). The ABTS⁺ radical solution was obtained by incubating 7 mM potassium peroxidisulfate solution in distilled water for 16 h in the dark at room temperature. The absorbance was maintained in the 0.700±0.010 range with a methanol:chloroform (1:1, v/v) mixture before the assay. A 100 μL oil sample, diluted 10-fold in methanol:chloroform (1:1, v/v) was allowed to react with 2.9 mL of the ABTS⁺ solution and the absorbance was recorded at 752 nm. The results of the ABTS⁺ tests were expressed as α-tocopherol and trolox equivalents. The analyses were done in triplicate.

The induction period (IP) of the oil samples was determined using a 679 Rancimat device (Metrohm Ltd., Herisau, Switzerland). Briefly, 3 g of oil sample were used at 110 °C with an air flow of 20 L/h. Each measuring vessel contained distilled water (60 mL) and the experiments were done in duplicate.

DSC was also used for the determination of the IP of rosehip seed oil. The IP of the oil sample was analyzed by DSC (Shimadzu Corporation, Tokyo, Japan) at different temperature ranges (100-150 °C). For this purpose, 3.0 mg of the oil were weighed in an aluminum pan while an empty pan was employed as the reference. The oven was heated from 50 °C to 150 °C at a rate of 10 °C/min under a nitrogen gas flow of 50 mL/min. After reaching 150 °C, oxygen gas was released into the sample instead of nitrogen at the same flow rate and the sample was held at 150 °C during analysis (Suja et al., 2004). The experiments were done in duplicate.

An oil sample (15 g) was weighed into glass Petri dishes (70 mm diameter) and placed in the dark in an oven at 30 °C and 60 °C. The oxidative stability of the oil was examined using the specific extinction (K₂₃₂ and K₂₇₀) values of the samples taken during the oxidation period (0-500 h) according to the AOCS Official Method of Ch 5-91 (AOCS, 2000). The experiments were done in duplicate. The tocopherol and carotenoid degradation in the oil were investigated during oxidation storage (0-160 h).

Accelerated photo oxidation was performed in a UV-light cabin (110 cm x 55 cm x 55 cm) fitted two 30 W UV lamps (254 nm wavelength). Petri dishes containing 15 g of oil were irradiated at a distance of 27 cm during 21 h. The temperature in the cabin was maintained at 30 °C. The oxidative stability of the oil was monitored by the determination of K₂₃₂ and K₂₇₀ values (AOCS, 2000). The experiments were done in duplicate. The tocopherol and carotenoid contents in the oil samples were also determined during photo oxidation.

The fatty acid composition of rosehip seed oil is shown in Table 1. The main fatty acid found in the oil sample was linoleic acid (54.8%), followed by linolenic acid (23.4%) and oleic acid (14.7%). These fatty acid profiles are similar to previous reports (Ilyasoğlu et al., 2014; Prescha et al., 2014). Grajzer et al., (2015) found fatty acid profiles in cold-pressed rose hip oils that were similar to those reported by Ozcan (2002) and Szentmihályi et al., (2002) for oils extracted from rose hip seeds. They concluded that rose hip oil is a valuable source of PUFA in the human diet.

The sterol composition of rosehip seed oil is also presented in Table 1. The main sterol component found in the oil was β-sitosterol (78%), followed by campesterol, Δ7-stigmastenol and Δ5-avenasterol, with 4.3%, 4.3% and 3.9%, respectively. Our results for β-sitosterol were similar to those reported by Grajzer et al., (2015) and Ilyasoğlu (2014). They stated that β-sitosterol was the most abundant sterol in rosehip oil. The percentages of campesterol, Δ7-stigmastenol and Δ5-avenasterol were similar to the data of Ilyasoğlu (2014). Grajzer et al., (2015) reported that the sterol levels of cold-pressed rose hip oils were high and ranged from 5891.6 mg/kg to 6485.4 mg/kg. The presence of brassicasterol reported by Zlatanov (1999) and Δ7-stigmastenol reported by Ilyasoğlu (2014) (5.4% and 6.2% of total sterols, respectively) were confirmed in our results. Phytosterols have health-promoting traits and aid in preventing cardiovascular diseases by inhibiting the intestinal absorption of cholesterol. Thus, sterols have been incorporated into functional foods (Ramadan, 2015).

The bioactive lipids in rosehip seed oil including tocopherols, carotenoids and phenolics are given in Table 2. The Total amount of tocopherol, carotenoid and phenolic contents were 786.3 mg/kg oil, 218.8 mg/kg oil and 37.97 mg gallic acid/kg oil, respectively. Regarding TPC, Grajzer et al., (2015) reported TPC in lower amounts (783.5 and 570.7 µg/kg) than those presented in our work. The level of total carotenoids in the oil was higher than those reported by Ilyasoğlu (2014) and Szentmihályi et al., (2002). Grajzer et al., (2015) reported that the carotenoid content of cold-pressed rose hip oils ranged from 36.4 mg/kg to 107.7 mg/kg mg/kg. Among tocopherol isomers, γ-tocopherol accounted for the highest value in rosehip oil (472.0 mg/kg oil). Besides, α-tocopherol (261.2 mg/kg oil) was found slightly higher than half of the γ-tocopherol, while δ-tocopherol was present only as a minor (53.1 mg/kg oil) isomer among the tocopherols. Grajzer et al., (2015) reported higher levels of total tocopherols (1124.7 and 1037.6 mg/kg) in rosehip oil in comparison with our results. However, the total tocopherol content of rosehip oil in our study was higher than the values (89.4 mg/kg) reported by Zlatanov (1999).

Several studies mentioned that oil composition is affected by the method of extraction. Also, studies reported that the qualitative and quantitative compositions of fatty acids, sterols, tocopherols and carotenoids differed significantly among different genera and also among cultivars of the same species. Grajzer et al., (2015) stated that γ- tocopherol was the major tocopherol isomer in rosehip oil. They found lower contents of α-tocopherol and higher contents of δ- tocopherol than those detected in our study (116.6, 147.3 mg/kg for α-tocopherol and 230.4, 259.9 mg/kg for δ- tocopherol). Andersson et al., (2011) studied the levels of tocols in four species of Rosa (R. dumalis, R. spinosissima, R. rubiginosa, and R. pimpinellifolia) and found α- and γ-tocopherol only in the fleshy parts of the rosehip. The amount of tocols and vitamin E activity varied with the date of harvesting and species, while fruit ripening had little influence on this activity.

The antiradical activities of the extracts from oil samples were assessed using DPPH· and ABTS⁺ radical scavenging tests. Both tests are commonly used to determine the antioxidant capacity of bioactive compounds. As exhibited in Table 2, the antioxidant activities of the extracts as assessed by the DPPH· test were 1.08 mg α-tocopherol/g oil and 4.18 µmol TEAC (Trolox equivalent antioxidant capacity)/g oil. The ABTS⁺ values for the oil extracts were determined as 1.00 mg α-tocopherol/g oil and 3.02 µmol TEAC/g oil. The antioxidant activity tested by ABTS⁺ for rosehip seed oil extracts was higher than the reported value (Ilyasoğlu, 2014) for the methanol extract of rosehip oil (1.77 μmol TEAC/g). The DPPH· results showed higher antiradical activity than those (2.32 mM TEAC/kg) reported by Prescha et al., (2014) and Grajzer et al., (2015), who reported that the antioxidant activity of rosehip oil extracts ranged between 2.32 and 3.00 mM TEAC/kg.

The differential scanning calorimetry (DSC) method is a rapid method to evaluate the oxidative stability of oils. In addition to this method, the Schaal oven test could be used as an accelerated oxidation method (Ramadan, 2012). Arain et al., (2009) determined the oxidative stability of Bauhinia purpurea oil and found a high correlation between DSC T0 values and the oxidative stability index (OSI) values. They concluded that the DSC method could be easily used as an alternative technique for the measurement of oil’s oxidative stability. To get comprehensive data about stability against oxidation, different thermal oxidation values should be compared. The values for the IP obtained by the Rancimat test and DSC are given in Table 3. The IP measured by Rancimat was 3.46 h at 110 °C. Regarding the DSC test, the IP of the oil samples decreased with the increase in the DSC experiment temperature. As expected, the highest IP (58.06 min) was recorded at 100 °C, while the lowest IP of the oil was 0.26 min at 150 °C. The IP obtained by DSC was lower than that obtained by the Rancimat method. These differences could be related to the higher surface-to-volume ratio, small amount of sample and purified oxygen used in DSC (Tan et al., 2002). In our study, the IP of rosehip oil determined by DSC was lower than the results reported by Grajzer et al., (2015) for cold-pressed rosehip oils.

Tocols have vitamin E activity and they are compounds with the ability to scavenge free radicals, and even tocols could be considered as the most important antioxidants. The high amount of tocopherols found in rosehip oil could be the main reason for the strong oxidative stability of the oil.

Lipid oxidation negatively influences the quality and nutritional value of foods during storage and this might also limit the utilization of oil in processed foods. Therefore, it is important to evaluate the stability of oils as affected by storage conditions (Mohdaly et al., 2010; Özkan et al., 2016). Several methods have been employed to measure the stability of oils and fats. The storage of oil samples under thermal and photo oxidation conditions was employed for monitoring the stability of oils (Özkan et al., 2016). On the other hand, the detection of conjugated dienes (CD) and conjugated trienes (CT) is a good parameter to determine the stability of oils. The formation of hydroperoxides is coincidental with the conjugation of double bonds in unsaturated fatty acids, as measured by absorption at the UV spectrum. Methylene-interrupted dienes or polyenes in lipids show a shift in their double bond position during oxidation. The resulting CD exhibit intense absorption at 232 nm, while CT exhibit intense absorption at 270 nm. The increase in CD and CT contents is proportional to the uptake of oxygen. The higher the levels of CD and CT in oil the lower the oil stability will be (Ramadan et al., 2006).

The changes in K₂₃₂ values during thermal and photo oxidation experiments are illustrated in Figure 1A. At all stages, the highest K₂₃₂ value of the oil was observed during UV light-induced oxidation. After 21 h, the K₂₃₂ value increased from 4.94 to 40.64 in the photo oxidation experiment, followed by the oil stored at 60 °C in the oven (15.38). There was a lower increase in the K₂₃₂ value in the oil during 21 h storage at 30 °C under dark conditions (8.47). With the increase in storage time, the K₂₃₂ value of the oil was decreased after a certain point under thermal storage conditions (60 °C). This behavior of the decrease in CD when the heating times increased could be explained by the induction of secondary oxidation products (hydroperoxides) which formed in the primary stages of oxidation (Shahidi et al., 1992; Ramadan and Moersel, 2004; Iqbal and Bhanger, 2007). Regarding the oil stored at 30 °C in the dark, the maximum value for K₂₃₂ was 17.69 after storage of up to 480 h.

The K₂₇₀ values for the oil samples were determined during storage under thermal and UV light-induced conditions (Figure 1B). After 21 h of storage at UV light, the K₂₇₀ value increased from 1.24 to 10.50. The K₂₇₀ values for samples over the same period stored at 60 °C and 30 °C were 2.83 and 1.14, respectively. Up to 144 h of storage at 60 °C, the K₂₇₀ value reached its maximum value (10.19) and after that, the value was decreased. However, the K₂₇₀ value of the oil increased slightly during storage at 30 °C. The results for K₂₃₂ and K₂₇₀ values were similar to those reported in a study of rosehip oil stored at 20 °C (Prescha et al., 2014), where an increase was observed for CD and CT values in rosehip oil during 6 months of storage.

The amounts of total tocopherols in the oils during different storage periods are presented in Figure 2A. The highest decrease in total tocopherols occurred during storage under UV light conditions. The amount of total tocopherols decreased from 784 mg/kg to 11.3 mg/kg after 11 h of storage under UV light. After that, the tocopherols were not detected in these samples. The other storage condition which caused a loss in tocopherols was the storage of oils at 60 °C. After 72 h of storage, total tocopherols decreased to 92.3 mg/kg; thereafter, tocopherols were not detected in the oil samples. The smallest loss in total tocopherols was determined at 30 °C of storage. Although there was a decline in the total tocopherols at the beginning of storage, after 21 h, the tocopherol content in the oil (132.6-139.8 mg/kg) exhibited stabile behavior for 120 h of storage.

In addition to total tocopherols, the levels of γ-tocopherol, the main tocopherol isomer in oil, was assessed during different storage conditions. The changes in the amounts of this isomer in the oil during storage is exhibited in Figure 2B. The highest decline in the level of this isomer was observed in the oils during storage under UV light. During 9 h of storage under UV light, the amount of γ-tocopherol decreased to 30.3 mg/kg from the starting value (472.0 mg/kg). The other significant decrease in γ-tocopherol was detected during storage at 60 °C. After 120 h under thermal storage, the concentration of γ-tocopherol declined to 15.3 mg/kg. The lowest decrease in γ-tocopherol between among the storage conditions was determined under dark storage at 30 °C. The amount of γ-tocopherol was 62.3 mg/kg for rosehip oil at the end of 144 h of storage.

The carotenoid degradation of rosehip oil under different storage conditions is summarized in Figure 2C. The degradation rate of carotenoids was more noticeable under UV light storage than under storage in the dark at 30 °C and 60 °C. After 15 h of storage under UV light, there was a tremendous decline in the carotenoid level in rosehip oil, where the total carotenoid content decreased to 0.6 mg/kg from the initial value (218.8 mg/kg). The other dramatic decrease in the content of carotenoids was observed in the oil samples stored under thermal conditions at 60 °C. Total carotenoid content decreased to 2.3 mg/kg after 48 h storage under thermal storage conditions. During storage at 30 °C, the minimum losses in total carotenoids was detected, where the total carotenoid content was 9.6 mg/kg even at the end of 144 h storage in the dark at 30 °C.