Mature seeds of A. graveolens were collected from the Experimental Research Station Derab, College of Agriculture, King Saud University, Riyadh, Saudi Arabia, during June, 2015. A voucher specimen (KSU no. 24-54-1) was deposited at the Herbarium of the Department of Botany and Microbiology, College of Science, King Saud University, Riyadh 1145, Saudi Arabia. After oven-drying the seeds at 40 °C for 24 h, the dried seeds were milled in a K/IKA-WERKE M20 grinder. A six -place Soxtec 8000 extraction unit (Foss, Suzhou, China) was used to extract the seed oil using hexane as solvent under the following conditions: Boiling temperature of 100 °C, boiling time of 5 min, rinsing time of 20 min, recovery time of 60 min, powder mass of 20 g, and hexane volume of 60 ml. The extract was stored in a refrigerator at −15 °C until analysis. Fatty acid methyl ester standards (FAMEs) were obtained from Sigma-Aldrich Corporation (Steinheim, Germany). Tocopherols and tocotrienols (tocols) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). The chemicals used in this investigation were of analytical grade or chromatographic purity. The sunflower oil used for the thermal analysis was obtained by the same extraction method using seeds purchased from a local market. Sunflower oil was used as standard for comparison purposes.

We used International Organization for Standardization (ISO) standards to determine the acidity (ISO 3960) and peroxide value (ISO 660). Carotenoid and chlorophyll contents were determined from the absorption spectra of the extract according to the method described by Nehdi et al., (2014a). The spectroscopic properties of A. graveolens extract diluted in n-hexane (1, 10%, v/v) were evaluated. The absorbance was measured at 290-400 and 400-800 nm using UV-1800 (Shimadzu, Kyoto, Japan).

We prepared FAMEs according to the laboratory protocol described by Nehdi et al., (2013) and analyzed using a GC-MS (QP2010 Ultra, Shimadzu, Kyoto, Japan). An RT-2560 column (100 m length, 0.25 mm internal diameter, 0.25-μm film thickness) was used for FAME separation. Helium was used as carrier gas at a flow rate of 1.50 mL/min. The oven temperature was increased from 115 °C to 240 °C at a rate of 2 °C/min and held for 15 min. A Shimadzu software (Cat. No. 225-21731-92) was used for the chromatogram analysis. In addition, the NIST analysis software and NIST11 library were used for the interpretation of the mass spectra and identification of each FAME.

The standard ISO 9936 procedure was used to determine the tocol content of A. graveolens seed extract. Briefly, a 0.5-g aliquot of the dill extract was dissolved in 25 mL of hexane, and 20 μL of the solution was injected into an LC-20AT high-performance liquid chromatography (HPLC) pump (Shimadzu, Kyoto, Japan). Tocols were separated on a Hypersil silica column (15 cm × 3 mm I.D., 3-μm particle size; Thermo Scientific). An isocratic elution with hexane/2-propanol (99.5:0.5; v/v) at a flow rate of 0.5 mL/min was used for separation. Each tocol was detected by a fluorescence detector set at 330 nm emission wavelength and 295 nm excitation wavelength. Authentic standards were used for the identification of tocols.

The induction time (IT) of each sample was analyzed with a 743 Rancimat analyzer (Metrohm AG, Herisau, Switzerland). Briefly, 3 g of the extract were incubated in the measuring vessel tube and heated up to 110 °C under an air flow rate of 20 L/h to obtain the IT of the extract.

Thermogravimetric (TGA) and first derivate thermogravimetric (DTGA) curves were obtained by a thermogravimetric analyzer TGA-50 (Shimadzu, Kyoto, Japan). A 5-mg sample mass placed in an alumina crucible was heated up to 600 °C at a heating rate of 10 °C/min under a synthetic zero air atmosphere (100 mL/min). The data of three independent measurements were analyzed by Shimadzu TA-60WS (2.20) software.

The results of GC/MS analysis of A. graveolens seed extract (Table 1) revealed that the extract comprised a mixture of fixed oil (74%) and essential oil (26%). The key fatty acids in the fixed oil component were oleic acid (OL) and its isomer cis-vaccenic acid (VA), which together represented more than 58% of the total compounds, followed by linoleic acid (LA) (5.5%) and palmitic acid (3.3%). Furthermore, linolenic, stearic, and erucic acids were present at a 4.5% concentration. Olive oil, which is rich in OL, is consumed to reduce the risk of coronary disease (Ruiz-Canela et al., 2011). Djoussé et al., (2014) observed a correlation between decreased risk of coronary heart disease and high plasma levels of cis-vaccenic acid.

The fatty acid LA has health-promoting effects. Topical application of oil containing LA may prevent skin diseases (acne, inflammation, dryness, and roughness) (Mokbli et al., 2018).

The principle volatile compounds were apiol (23%) and d-carvone (2.4%). Furthermore, the amount of trans-dihydrocarvone was only 0.44%. Apiol is a phenylpropanoid compound found in parsley (Petroselinum crispum) essential oil extracted from the seed or aerial part (Zhang et al., 2006; Farzaei et al., 2013). Apiol is the main constituent (90.7%) identified in the hexane extract of Piper aduncum leaves. This compound has two symmetrical methoxy groups. The radical-scavenging activity of the benzene ring is improved by the presence of these two methoxy groups, which are strong electron donors. Zhang et al., (2006) reported that apiol is the primary contributor to the antioxidant activity of parsley oil.

Elmastaş et al., (2006) found that d-carvone is an oxygenated monoterpene with stronger antioxidant activity than α-tocopherol.

It plays an important role as a free radical-scavenging agent against N(ω)-nitro-L-arginine methyl ester (L-NAME)-induced hypertension, a major factor involved in cardiovascular diseases (Rajeshwari and Raja, 2015).

A. graveolens seed extract is a healthy natural product rich in antioxidant compounds. It can be used in food as a natural preservative and antioxidant to prevent lipid oxidation as well as to retard rancidity (Galanakis et al., 2018). Furthermore, it may be used as a pharmaceutical ingredient (Aludatt et al., 2017).

Seven isomers of vitamin E were found in the A. graveolens extract at a 155 mg/100 g concentration. These included α-tocopherol, β-tocopherol, γ-tocopherol, δ-tocopherol α-tocotrienol, γ-tocotrienol, and δ-tocotrienol (Table 2). The principle isomers present were β-tocopherol and γ-tocotrienol at a concentration of 110 and 21 mg/100 g, respectively. β-isomer exists in low concentrations (0.3-2.5 mg/100 g) in most conventional vegetable oils such as soybean, sunflower, olive, and corn oils (Nehdi et al., 2014b). Hence, the A. graveolens lipid fraction may be considered as a source of this rare β-isomer, which inhibits lipid oxidation by suppressing free radicals such as hydroperoxides. The presence of other tocols maintains the flavor quality of the product. Thus, A. graveolens seed extract may be added into food products to prevent oxidation.

The extract content of A. graveolens seeds was 9.4%. The extract was green in color and had a pleasant, spicy aroma. The characteristic green color of the extract may be associated with the presence of chlorophyll (9.9 mg/kg, Table 2), as confirmed by the strong absorption band at 669 nm (Figure 1). Despite the presence of carotenoids in the extract (41 mg/kg), the green color of chlorophyll was dominant. Carotenoids are widely used in the pharmaceutical, medical, cosmetic, and food industries (Siger et al., 2017). The extract showed medium quality indices such as peroxide value (9.5 meq O₂/kg) and acidity (4.2%) (Table S2). The UV absorbances of the extract in UV-A and UV-B ranges (320-400 and 290-320 nm, respectively) were stronger than those of sunflower and Citrullus colocynthis seed oils (Nehdi et al., 2013). Thus, A. graveolens extract may have protective effects against UV-B-induced sun burn (Schwarz et al., 1995) and UV-A-induced premature skin aging (Tebbe et al., 1997). The A. graveolens extract may be used as an ingredient in the formulation of sunscreen lotions and creams (Nehdi et al., 2013). Furthermore, it may serve as a food additive owing to its pleasant aroma.

A. graveolens extract showed a high IT of 45.22 h (Figure 2). This observation revealed the high resistance of the extract against oxidation. The IT of this extract was higher than that of date seed oil (21 h difference) (Nehdi et al., 2018), soybean oil (6.5 h difference) (Taghvaei et al., 2014), safflower oil (2.87 h difference), and flax seed oil (1.57 h difference) (Bozan and Temelli, 2008). This observation may be attributed to the low content (6.4 %) of linoleic and linolenic acids and the presence of antioxidants such as tocols, apiol, and d-carvone. However, these antioxidants may together produce synergistic effects, thereby positively affecting the oxidative stability of the A. graveolens extract.

The comparison between the TGA and DTGA curves of A. graveolens extract and sunflower extract is shown in Figure 3. The TGA curve showed that the extract was thermally unstable compared to sunflower seed oil. A. graveolens extract lost 5% of its mass at 92 °C; however, sunflower oil showed a 5% mass loss at 247 °C. Heating resulted in the evaporation of the volatile fraction from A. graveolens extract, and maximum evaporation was observed at 353 °C. Further heating resulted in a similar effect to that observed with the sunflower oil. The extract showed mass loss at three steps with nearly the same maximum at 352-353 °C, 428 °C, and 516-541 °C. The mass loss was attributed to the evaporation of different groups of triacylglycerols.

Thermogravimetric analysis may be used as a successful method for the estimation of the purity of the seed oil. The DTGA curve of the mixture of seed oil and essential oil showed an additional evaporation peak of the essential oil.

Further studies are warranted to discover new therapeutics counteract the problems related to emerging resistance to existing cancer drugs. Plants are rich in bioactive compounds that play a crucial role in cancer chemotherapy, such as vincristine and taxol isolated from Vinca rosea and Taxus brevifolia, respectively (Prakash et al., 2013).

To evaluate the cytotoxicity of A. graveolens seed extract, breast cancer cells MDA-MB-231 and MCF-7 were treated with different concentrations of the extract and a MTT assay was performed to analyze cell viability. The MTT assay is an indicator of cellular metabolic activity based on the reduction of MTT by the mitochondrial enzymes to purple-colored formazan. It is used to study the cytotoxic effects of toxic compounds and plant extracts against cancer cell lines.

A. graveolens extract showed a dose-dependent inhibitory effect on MCF-7 and MDA-MB-231 cells with IC₅₀ values of 53 and > 100 μg/mL, respectively (Figure 4A). Maximum toxicity was observed at a concentration of 100 μg/mL, where the viabilities of MCF-7 and MDA-MB-231 cells reduced to approximately 31 and 63%, respectively. Figure 4B shows the effect of A. graveolens seed extract on LDH release from MCF-7 and MDA-MB-231 cells. LDH is a reliable marker of cytotoxicity because injured cells are fragmented following incubation with the extract. Therefore, LDH leakage from MCF-7 and MDA-MB-231 cells is attributed to the cytotoxic nature of A. graveolens seed extract, which confirms its antiproliferative activity.

This antiproliferative activity may be associated with the presence of apiol in the A. graveolens seed extract. di Stefano et al., (2011) observed that apiol exhibits cytotoxicity against K-562, NCI-H460, and MCF-7 cancer cells. The percentage of LDH released from MCF-7 and MDA-MB-231 cells significantly increased (p < 0.05) after 48 h of incubation (Figure 4B). The percentage increase in LDH release was significant for MCF-7 and MDA-MB-231, respectively, as compared to the control (Figure 4B) at the concentration of 100 μg/ml.

Figure 5 shows the morphology of MDA-MB-231 and MCF-7 cells exposed to dill extract. Cell morphology was assessed under a microscope after treatment with A. graveolens seed extract at 50 and 100μg/ml for 48 h. The cells from the treatment group were shrunken and showed condensed nuclei. In addition, these cells were detached. The higher the concentration of dill oil, the larger was the percentage of abnormal cells (Figure 5). These observations indicate that the effects observed in the treatment groups were associated with nuclear and cytosol shrinkage, which were the markers of apoptotic events (Kalinichenko and Matveeva, 2008). Our result is in agreement with Mohammed et al., (2018) who reported the anticancer potential of A. graveolens seed extract against HepG2 cells, though the concentration used was higher than 300 μg/ml. Interestingly, the extract showed a different response to the breast cancer cell lines tested and this could be attributed to the difference in the cell lines used. Thus, the seed oil of the A. graveolens plant could be a chemotherapeutic option against breast cancer.