Seeds (moisture content 12%) were fine powdered using a high-speed mixer, and milled into a fine particle sized (ca. 20–50 µm) meal, then subjected to Soxhlet extraction using n-hexane for 14 h. The recovered TL were stored at 4 °C for further analysis.

The TL were separated into different classes by elution with solvents over a glass column (20 mm dia×30 cm) packed with a slurry of activated silicic acid (70–230 mesh; Merck, Darmstadt, Germany) in chloroform (1:5, w/v). NL were eluted with 3 times the column volume of chloroform. The major portion of GL was eluted with 5-times the column volume of acetone and PL with 4 times the column volume of methanol. By means of TLC on Silica gel F₂₅₄ plates (thickness=0.25 mm; Merck, Darmstadt, Germany) a further characterization of GL and PL subclasses was carried out with the following solvent system: chloroform/methanol/ammonia solution 25% (65:25:4, v/v/v). For the characterization of NL subclasses TLC plates were developed using n- hexane/diethyl ether/acetic acid (60:40:1, v/v/v). For the detection of the lipids, TLC plates were sprayed with the following agents: for the marking of all lipids with sulfuric acid (40%), for the marking of GL with α-naphthol/sulfuric acid and for the marking of PL with the molybdate-blue reagent (Kates, 1986; Ramadan et al., 2006). Each spot was identified with lipid standards as well as their reported retention factor (R^f) values. Individual bands were visualized under UV light, scraped from the plate and recovered by extraction with chloroform/methanol (2:1, v/v).

For the quantitative determination of NL subclasses individual bands were scraped from the plate and recovered by extraction with 10% methanol in diethyl ether, followed by diethyl ether. The data presented are the average of three gravimetrical determinations. For the quantitative estimation of GL subclasses, the acetone fraction obtained from CC was separated by TLC in the above given solvent system. The silica gel regions with the corresponding GL subclasses were scraped out followed by a hexose measurement at 485 nm using the phenol/sulfuric acid in acid-hydrolyzed lipids (Ramadan et al., 2006). The percent distribution of each component was obtained from the hexose values. From the extinction values the quantitative amount was determined and related to their portion of the GL fraction. The determined portion was compared with the amount of TL, which had been separated using CC in the main lipid fractions. For the determination of PL, the methanol fraction from CC was also separated using TLC in the above given solvent system. After the scraping out of the individual PL subclasses, classes were brought to react with the hydrazine sulfate/sodium molybdate reagent at 100 °C for 10 min and espectrophotometrically analyzed at 650 nm. From the obtained extinction values via a calibration chart for phosphorus the amount of PL was calculated. The individual values were compared to the PL fraction (methanol fraction from CC) and to the amount of TL.

The fatty acid compositions of TL, NL, GL and PL were determined using GC as described below. Fatty acids were transesterified into FAME using N-trimethylsulfoniumhydroxide (Macherey-Nagel, Düren, Germany) according to Arens et al. (1994). FAME were determined on a Shimadzu GC-14A equipped with flame ionization detector (FID) and C-R4AX chromatopac integrator (Kyoto, Japan). The flow rate of the carrier gas (helium) was 0.6 mL/min and the split value was a ratio of 1:40. A sample of 1 µL was injected onto a 30 m×0.25 mm×0.2 µm film thickness Supelco SP™-2380 (Bellefonte, PA, USA) capillary column. The injector and FID temperature was set at 250 °C. The initial column temperature was 100 °C programmed at 5 °C min⁻¹ until 175 °C and held for 10 min at 175 °C, then at 8 °C min⁻¹ until 220 °C and held for 10 min at 220 °C. A comparison was made between the retention times of the samples with those of authentic FAME standard mixture (ME 19, Sigma, St. Louis, MO, USA; 99% purity specific for GC) run on the same column under the same conditions to facilitate identification.

The separation of ST was performed after the saponification of oil according to Ramadan et al. (2008). TL (250 mg) were refluxed with 5 mL of ethanolic potassium hydroxide solution (6%, w/v) and a few anti-bumping granules for 60 min. The unsaponifiables were first extracted 3-times with 10 mL of petroleum ether. The extracts were combined and washed 3-times with 10 mL of neutral ethanol/water (1:1, v/v), then dried overnight with anhydrous sodium sulfate. The extract was evaporated in a rotary evaporator at 25 °C under reduced pressure, and then ether was completely evaporated under nitrogen. GC analyses of unsaponifiable residues were carried out using a Mega Series Chromatograph (HRGC 5160, Carlo Erba Strumentazione; Milan, Italy) equipped with an FID. The following GC conditions were used: DB 5 column (JandW scientific; Falsom, CA, USA) packed with 5% phenylmethylpolysiloxan, 30 m length, 0.25 mm i.d., 1.0 µm film thickness; carrier gas (helium) flow 0.38 mL min⁻¹. Detector and injector were set at 280 °C. The oven temperature was kept constant at 310 °C and the injected volume was 2 µL. The repeatability of the analytical procedure was tested and the relative standard deviation of three repeated analyses of a single sample was <5%. Quantitative analyses were performed with a Shimadzu (C-R6A Chromatopac; Kyoto, Japan) integrator.

NP-HPLC was selected to avoid extra sample treatment (e.g., saponification) according to Ramadan et al. (2010). The analysis was performed with a solvent delivery LC-9A HPLC (Shimadzu, Kyoto, Japan). The chromatographic system included a model 87.00 variable wavelength detector and a 250×4 mm i.d. LiChrospher-Si 60, 5 µm column (Knauer, Berlin, Germany). The separation of tocopherols was based on isocratic elution when the solvent flow rate was maintained at 1 mL min⁻¹ at a column back-pressure of about 65–70 bar. The solvent system selected for elution was isooctane/ethyl acetate (96:4, v/v) with detection at 295 nm. Twenty µL of the diluted solution of TL in the mobile phase were directly injected into the HPLC. Tocopherols were identified by comparing their retention times with those of authentic standards. Standard solutions were prepared by serial dilution to a concentration of approximately 5 mg·mL⁻¹ of each tocopherol. Standard solutions were prepared from a stock solution which was stored in the dark at −20 °C. Twenty µL were injected and peak areas were determined to generate standard curve data. All quantification was by peak area using a Shimadzu C-R6A chromatopac integrator (Kyoto, Japan). Standard curves (concentration versus peak area) were calculated from different concentration levels by linear regression. Based on the chromatographic conditions, repeated injections of different concentrations of the tocopherols were made 3 times in the HPLC system. Injections in triplicate were made at each concentration for both standards and samples. All work was carried out under subdued light conditions. All the experiments were repeated at least three times when the variation in any one was routinely less than 5%.

Aliquots of C. medium oil and extra virgin olive oil (2 g) were dissolved in n-hexane (5 mL) and mixed with 10 mL methanol-water (80:20, v/v) in a glass tube for two min in a vortex. After centrifugation at 3000 rpm for 10 min, the hydroalcoholic extracts were separated from the lipid phase using a Pasteur pipet and then combined and concentrated in vacuo at 30 °C until a syrupy consistency was reached. The lipid residue was re-dissolved in 10 mL methanol-water (80:20, v/v) and the extraction was repeated twice. Hydroalcoholic extracts were re- dissolved in acetonitrile (15 mL) and the mixture was washed three times with n-hexane (15 mL each). The purified phenols in acetonitrile were concentrated in vacuo at 30 °C and then dissolved in methanol for further analysis. Aliquots of phenolic extracts were evaporated to dryness under nitrogen. The residue was re-dissolved in 0.2 mL water and diluted (1:30) Folin-Ciocalteu's phenol reagent (1 mL) was added. After 3 min, 7.5% sodium carbonate (0.8 mL) was added. After a further 30 min, the absorbance was measured at 765 nm using UV-260 visible recording spectrophotometer (Shimadzu, Kyoto, Japan). Gallic acid was used for the calibration and the results of triplicate analyses are expressed as parts per million of gallic acid.

The RSA of C. medium oil and olive oil was assayed with DPPH radical previously dissolved in toluene according to Ramadan et al. (2009). A toluene solution of DPPH radicals was freshly prepared at a concentration of 10⁻⁴ M. The radical was stable for more than 2 h of kinetic assay in the absence of antioxidant compounds. For evaluation, 10 mg of C. medium oil or extra virgin olive oil (in 100 µL toluene) were mixed with 390 µL of the toluene solution of DPPH radicals and the mixture was vortexed for 20 s at ambient temperature. Against a blank of pure toluene without DPPH, the decrease in absorption at 515 nm was measured in 1 cm quartz cells after 1, 30 and 60 min of mixing using a UV-260 visible recording spectrophotometer (Shimadzu, Kyoto, Japan). RSA toward DPPH radicals was estimated from the differences in absorbance of the toluene DPPH solution with or without sample (control) and the inhibition percent was calculated from the following equation:

A Miniscope MS 100 Electron Spin Resonance (ESR) spectrometer (Magnettech GmbH; Berlin, Germany) was used (Ramadan, 2012). The experimental conditions were as follows: measurement at room temperature; microwave power, 6 db; centerfield, 3397 G, sweep width 83 G, receiver gain 10 and modulation amplitude 2000 mG. Ten mg of oils (in 100 µL toluene) were allowed to react with 100 µL of a toluene solution of galvinoxyl (0.125 mM). The mixture was stirred in a vortex stirrer for 20 s then transferred to a 50 µL micro pipette (Hirschmann Laborgeräte GmbH, Ederstadt, Germany) and the amount of galvinoxyl radical inhibited was measured exactly 60 s after the addition of the galvinoxyl radical solution. The galvinoxyl signal intensities were evaluated by the peak height of the signals against a control. Further ESR spectra have been recorded at intervals of 90 s for a total incubation time of 60 min. A quantitative estimation of the radical concentration was obtained by evaluating the decrease in the ESR signals in arbitrary units between 1 and 60 min incubation using a KinetikShow 1.06 Software program (Magnettech GmbH; Berlin, Germany). The reproducibility of the measurements was ± 5 as usual for kinetic parameters.

All experiments were carried out under subdued light conditions. The experiments were repeated at least three times when the variation in any one was routinely less than 5%. All experimental procedures were performed in triplicate and the mean values (± standard deviation) were given.