Butylated hydroxytoluene (BHT) was purchased from Sigma. Nitroblue tetrazolium (NBT), phenazine methosulfate (PMS), nicotinamide adenine dinucleotide (NADH) and 1,1-Diphenyl-2-picrylhydrazyl (DPPH) were purchased from Sigma. Folin–Ciocalteu reagent and sodium azide were purchased from Aldrich. Authentic standards of phenolic compounds were purchased from Sigma and Fluka. Stock solutions of these compounds were prepared in HPLC-grade methanol. These solutions were wrapped in aluminium foil and stored at 4 °C. All other chemicals used were of analytical grade.

N. sativa shoots and roots were collected in May 2005 at the flowering stage from cultivated plants from the Menzel Temime region (northeastern Tunisia).

The air-dried and finely ground shoots and roots (2.5 g) were extracted by stirring with 25 ml of absolute methanol at room temperature for 30 min. Extracts were kept for 24 h at +4 °C, and then filtered through Whatman filter paper. Extracts were evaporated under vacuum to dryness to give yields of 29.3 and 19% for shoots and roots respectively, and stored at +4 °C until analysis.

Total phenolics of N. sativa extracts were determined using the Folin–Ciocalteu (F-C) reagent, according to the method described by Dewanto et al. [15]. An aliquot of 0.125 ml of diluted extract was added to 0.5 ml of deionized water and 0.125 ml of the (F-C) reagent. After shaking, the mixture was incubated for 3 min at room temperature. Then, 1.52 ml of 7% Na₂CO₃ solution was added. The volume obtained was adjusted to 3 ml using distilled water, mixed vigorously, and held for 90 min at ambient temperature. The absorbance of the solution was then measured at 760 nm against a blank. The total phenolic content was expressed as mg of gallic acid equivalents (GAE) per gram of dry weight through the calibration curve of gallic acid. The sample was analyzed in three replications.

Dried samples from shoots and roots were hydrolysed according to the method of Proestos et al. [16], slightly modified. Forty millilitres of methanol containing BHT (1 g l⁻¹) were added to 0.5 g of a dried N. sativa sample. Then 10 ml of 6 M HCl were added. The mixture was stirred carefully and then sonicated for 15 min and refluxed in a water bath at 90 °C for 2 h. The obtained mixture was injected to HPLC. The phenolic compounds' analysis was carried out using an Agilent Technologies 1100 series liquid chromatograph (RP–HPLC) coupled with an UV-Vis multiwavelength detector. The separation was carried out on a 250×4.6-mm, 4-μm Hypersil ODS C18 reversed phase column at ambient temperature. The mobile phase consisted of acetonitrile (solvent A) and water with 0.2% sulphuric acid (solvent B). The flow rate was kept at 0.5 ml min⁻¹. The gradient programme was as follows: 15% A/85% B 0–12 min, 40% A/60% B 12–14 min, 60% A/40% B 14–18 min, 80% A/20% B 18–20 min, 90% A/10% B 20–24 min, 100% A 24–28 min. The injection volume was 20 μl, and peaks were monitored at 280 nm. Samples were filtered through a 0.45-μm membrane filter before injection. Peaks were identified by congruent retention times compared with standards. Analyses were performed in triplicates.

The electron donation ability of the obtained methanol extracts was measured by bleaching of the purple-coloured solution of 1,1-diphenyl-2-picrylhydrazyl radical (DPPH) according to the method of Hanato et al. [17]. Extracts (2 ml, 10–1000 μg ml⁻¹) were added to 0.5 ml of 0.2 mM DPPH methanolic solution. After an incubation period of 30 min at room temperature, the absorbance was determined against a blank at 517 nm. Percentage inhibition of free radical DPPH (PI%) was calculated as follow: PI%=[(Ablank−Asample)/Ablank]×100, where Ablank is the absorbance of the control reaction and Asample is the absorbance in the presence of plant extract. Extract concentration providing 50% inhibition (IC₅₀) was calculated form the regression equation prepared from the concentration of the extracts and the inhibition percentage. BHT was used as a positive control. Samples were analyzed in triplicate.

The superoxide anion was generated in a PMS-NADH system by oxidation of NADH and assayed by the reduction of NBT. The method described by Duh et al. [18] was used to determine the superoxide anion radical scavenging activity. The reaction mixture, containing 0.2 ml of extract (10–1000 μg ml⁻¹) in distilled water, 0.2 ml PMS (60 μM), 0.2 ml of NADH (677 μM), and 0.2 ml NBT (144 μM), was incubated at room temperature for 5 min. Then the absorbance was read at 560 nm. All solutions were prepared in a phosphate buffer (0.1 M, pH 7.4). The scavenging activity was calculated as follows: PI%=[(Ablank−Asample)/Ablank]×100, where Ablank is the absorbance of the control reaction and Asample is the absorbance in the presence of plant extract. The IC₅₀ was calculated from the plot of the inhibition percentage against the extract concentration. BHT was used as a reference compound control. Samples were analyzed in triplicate.

The method of Oyaizu [19] was used to assess the reducing power of N. sativa shoot and root extracts. Methanol extracts (1 ml) were mixed with 2.5 ml of a 0.2 M sodium phosphate buffer (pH = 6.6) and 2.5 ml of 1% potassium ferricyanide (K₃Fe (CN)₆), and incubated in a water bath at 50 °C for 20 min. Then, 2.5 ml of 10% trichloroacetic acid were added to the mixture that was centrifuged at 650 g for 10 min. The supernatant (2.5 ml) was then mixed with 2.5 ml distilled water and 0.5 ml of 0.1% ferric chloride solution. The intensity of the blue–green colour was measured at 700 nm. The EC₅₀ value (mg ml⁻¹) is the extract concentration at which the absorbance was 0.5 for the reducing power and was obtained from the linear regression equation prepared from the concentrations of the extracts and the absorbance values. High absorbance indicates high reducing power. Ascorbic acid was used as a positive control.

Extracts (0.1 ml) were added to a solution of 2 mM FeCl₂ (0.05 ml) [20]. The reaction was initiated by the addition of 5 mM ferrozine (0.1 ml) and 2.75 ml of distilled water. The mixture was shaken vigorously and left at room temperature for 10 min. The absorbance of the solution was then measured at 562 nm. The scavenging activity was calculated as follows: PI%=[(Ablank−Asample)/Ablank]×100, where Ablank is the absorbance of the control reaction and Asample is the absorbance in the presence of plant extract. IC₅₀ was calculated from the plot of inhibition percentage against extract concentration. EDTA was used as a positive control. Samples were analyzed in triplicate.

The mutagenicity assay was performed with Salmonella typhimurium strain TA 1535 using the plate incorporation procedure described by Maron and Ames [21]. Fifteen microlitres of bacterial stock were incubated in 5 ml of oxoid nutrient for 16 h at 37 °C in an orbital shaker. Then 100 μl of the overnight culture of bacteria (approximate cell density: 10⁸ cells ml⁻¹) and 100 μl of extracts dissolved in DMSO (0.05–0.5 mg ml⁻¹) were added to tubes containing 2 ml of top Agar (supplemented with 0.5 mM l-histidine and 0.5 mM d-biotine). The mixture was then poured on minimal agar plates previously prepared as described by Maron and Ames [21]. The plates were incubated at 37 °C for 48 h and the revertant bacterial colonies of each plate were counted. Negative control cultures gave numbers of revertants per plate that were within the normal limits found in the laboratory. An extract was considered mutagenic if the number of revertants per plate was at least doubled. The spontaneous revertants are known to be responsive to sodium azide [21]. This mutagen was chosen to study the antimutagenic activity, and 15 μl of the mutagenic agent (dissolved in DMSO) were added per tube of top agar containing the bacterial strain and the extract. The resulting mixture was poured on the minimal agar plate. The plates were incubated at 37 °C for 48 h and the revertant bacterial colonies of each plate were counted. The inhibition percentage of mutagenicity (PI%) was calculated according to the following formula: PI = [1 – (number of revertants on test plates/number of revertants on control plates)] × 100.

All data are reported as the mean ± SD of three measurements. Means of total phenolic content, IC₅₀ on DPPH, superoxide and chelating powers and EC₅₀ on reducing power were statistically compared using the STATI-CF program with the Student t-test at the P<0.05 significance level. A one-way analysis of variance (ANOVA) and the Newman–Keuls multiple-range test were carried out to test any significant differences between shoot and root phenolic content used at P<0.05. A two-way analysis of variance (ANOVA), with the organ (O) and dose (D) as factors, was achieved for mutagenic and antimutagenic data, using the STATI-CF statistical program. Means were compared using the Newman–Keuls test at the P<0.05 level, when significant differences were found.

Based on the absorbance values of extract solutions reacted with the Folin–Ciocalteu reagent and compared with the standard solutions of gallic acid equivalents, total phenolics are given in Table 1. Phenolic contents varied significantly between organs. Shoots exhibited an important amount of 10.04 mg of GAE/g DW, which was about 2.5 times higher than the total polyphenols of roots (4.01 mg of GAE/g DW). The levels of total polyphenols were superior to those reported in N. sativa seeds [22].

The free forms of phenolic compounds are rarely present in plants. More often, they occur as esters, glycosides or bound to the cell wall. For this reason, acidic hydrolysis was used to release the aglycones in order to simplify the identification process [23]. Moreover, BHT, a powerful antioxidant, was added to prevent degradation of phenolics during hydrolysis [24]. RP-HPLC coupled with a UV-Vis multiwavelength detector was employed to separate and to quantify phenolic compounds. Fig. 1 shows the chromatograms of authentic standards and N. sativa shoots and roots. Fourteen phenolic compounds were successfully identified, including gallic acid, (–)-p-hydroxybenzoic acid, chlorogenic acid, vanillic acid, p-coumaric, ferulic acid, trans-2-hydroxycinnamic acid, trans-cinnamic acid, epicatechin, (+)-catechin, quercetin, apigenin, amentoflavone, and flavone. These compounds have been identified according to their retention time and the spectral characteristics of their peaks compared to those of standards, as well as by spiking the sample with standards. Results demonstrated that differences in N. sativa shoots and roots phenolic composition were significantly more quantitative than qualitative. With the exception of p-coumaric and ferulic acids detected only in roots, and amentoflavone present in shoots, N. sativa plant parts possess similar composition.

The amounts of the different identified phenolic acids and flavonoids are shown in Table 2. Vanillic acid was detected to be the major phenolic component in the two plant parts (shoots and roots), contributing about 66% to the total amount and showing the levels of 143.21 and 89.94 mg/100 g DW in shoots and roots, respectively. Gallic acid was also predominant, but slightly higher in roots (30.59 mg/100 g DW) than in shoots (27.86 mg/100 g DW). trans-Cinnamic acid, (+)-catechin and apigenin were also found with appreciable level in shoots. The amounts of the detected compounds are close to those reported in similar aromatic and medicinal herbs [25].

The levels of total phenolic compounds in N. sativa shoots and roots determined by HPLC were 2.15 and 1.35 mg/g DW for shoots and roots, respectively, and thus lesser than the ones obtained by the Folin–Ciocalteu method. This result is predictable due to the weak selectivity of the Folin–Ciocalteu reagent, as it reacts positively with different antioxidant compounds (phenolic and non-phenolic substances).

The stable DPPH radical is widely used to evaluate the free radical scavenging activity in many plant extracts [26]. The assessment of antioxidant activity showed that both N. sativa shoots and roots were able to scavenge this radical (Table 1). Shoots displayed a higher activity than roots (IC₅₀ = 280 and 450 μg ml⁻¹, respectively). Although this scavenging effect was lower than that of BHT (Table 1), it was stronger than the antiradical activity reported in the seeds of the same specie from India [22], which showed an IC₅₀ value of 1240 μg ml⁻¹.

Results demonstrated also that both N. sativa shoots and roots exhibited significant superoxide anion scavenging capacity (Table 1). As for DPPH, shoots were more effective O⁻₂ scavengers than roots. The IC₅₀ values of N. sativa shoots and roots extracts on superoxide radical scavenging activity were 12 and 18 μg ml⁻¹, respectively. These results revealed that methanolic extracts of N. sativa organs were free radical scavengers, acting possibly as primary antioxidants.

The reducing capacity of a compound may serve as indicator of its potential antioxidant activity [27]. The presence of reducers (i.e. antioxidants) causes the conversion of the Fe³⁺/ferricyanide complex to the ferrous form. N. sativa shoots exhibit a significant reducing power (Table 1); in fact EC₅₀ (57 μg ml⁻¹) was comparable to that of the standard ascorbic acid (38 μg ml⁻¹). Roots showed lower reducing ability than roots, presenting an EC₅₀ of 1.85 mg ml⁻¹. These results suggest that N. sativa shoots were electron donors reacting with free radicals to convert them into more stable products and to terminate radical chain reactions as described by [28].

Although iron is essential for oxygen transport, respiration, and enzymes activity, it is a reactive metal that catalyzes oxidative damage in living tissues and cells [29]. Ferrozine can quantitatively form complexes with Fe²⁺. In the presence of chelating agents, the complex formation is disrupted with the result that the red colour of the complex is decreased. Results indicate that methanol extracts of N. sativa shoots and roots interfered with the formation of ferrous and ferrozine complex (Table 1). However, this chelating power activity was weak, in fact, compared to the standard EDTA, IC₅₀ were high (3.8 and 7.5 mg ml⁻¹ for shoots and roots, respectively).

The strong antioxidant activity of N. sativa shoots and roots assessed by the different systems could be attributed to their high total polyphenolic contents; in fact, it has been found that polyphenols are one of the most effective antioxidative constituents in the plant [30]. Moreover, the high yield of the different phenolic compounds found in N. sativa plant parts (Table 2) might contribute to the potent antioxidant activity of the methanol extracts, since a positive correlation between phenolic composition and antioxidant activity was proved [31]. Thus, antioxidant property of shoots and roots could be attributed to the significant amount of benzoates, especially vanillic acid, present in our study with the high amounts of 143.21±0.23 and 89.94±0.92 μg/100 g DW, respectively in shoots and roots. Numerous studies indicated that this phenolic acid was a potent antioxidant agent, quenching radicals, singlet oxygen and hydrogen peroxide [32–34]. Moreover, the amount of vanillic acid was about 1.5 higher in shoots than in roots, which may explain the superiority of shoots extracts as antioxidants. Other minor phenolic compounds should not be neglected, since synergy of the different chemicals with each other should be taken into consideration for the biological activity. Likewise, the molecular structure and differences in number and position of the hydroxyl group on the aromatic ring influence the antioxidant activity [5]. The presence of the CHCHCOOH group in the hydroxycinnamic acids found in N. sativa shoots and roots (trans-cinnamic, trans-2-hydroxy-cinnamic, p-coumaric, ferulic and chlorogenic acids) is considered to be key element for the antioxidative efficiency [35]. As well, catechin was found to be the most abundant flavonoid in our study for both shoots (7.26±0.97 mg/100 g DW) and roots (3.4±0.74 mg/100 g DW). This flavonol is a well-known antioxidant, due to the presence of the o-dihydroxy and o-hydroxyketo groups [3].

The results of the Ames test are reported in Table 3. Spontaneous revertants were 18 in number and none of the tested extracts, even at high concentrations, induced significant increase of the revertants' number in Salmonella typhimurium (TA1535) strains. The absence of mutagenicity within the extracts of N. sativa in the Salmonella tested strain indicated that DNA does not seem to be a relevant target.

Sodium azide damages DNA and thus induces mutagenecity. A dose of 1.5 μg per plate of this mutagen was chosen for the antimutagenicity studies, since this dose was not toxic and induced 1320 revertants in S. thyphimurium TA1535. The inhibitory effect of the methanol N. sativa extracts on the mutagenicity of sodium azide using the plate incorporation assay is illustrated in Table 3.

The results indicated that shoots and roots demonstrated significant inhibitory effect. The highest antimutagenic activity for N. sativa shoots (45.84% inhibition) and roots (71.32% inhibition) was obtained at a low concentration (50 μg/plates), although higher doses failed to increase antimutagenicity (Table 3). This type of response has been already obtained in other antimutagenic studies [36].

The relevant antimutagenic activity of N. sativa extracts associated with the absence of genotoxicity suggests that these extracts contain interesting active compounds and our data on antioxidant activity of N. sativa shoots and roots imply that their antigenotoxic activity is mediated by their antioxidative property, as it is has already been reported [37,38]. In fact, many carcinogens and/or mutagens produce oxygen-free radicals for interaction with cellular macromolecules; thus the antioxidant and antiradical properties of antioxidants, especially polyphenols, give them the ability to inhibit DNA lesions generated by oxidative stress, and to prevent mutagenicity [37]. In our study, shoots and roots exhibited a high level of phenolic compounds (Table 2) and several of these phenolic acids and flavonoids, including gallic acid, catechin and apigenin, were reported as powerful antimutagenic and anticancer agents [9,39,40]. Moreover, Sawa [41], who studied the antiradical property of 17 authentic phenolics, demonstrated that vanillic acid, the major phenolic component in N. sativa shoots and roots, has a strong anti-tumour-promoter effect through its remarkable antiradical capacity.

Finally, the fact that roots of N. sativa have a stronger antimutagenic capacity than that of shoots, in spite of their lower total polyphenolic content, suggests that the activity may be due more to the combinatory effect of different compounds than to their quantity.