Fresh and mature pears fruits of two Opuntia species, two forms of O. ficus indica: spiny (green-yellow peel and yellow pulp) and thornless (green peel and red-purple pulp), and O. stricta (purple peel and pulp) were collected in summer 2011 (September). The spiny wild form was from the region of Al-Ala in the center of Tunisia, located at 35°36′N (North) latitude, 9°34′E (East) longitude, and 450 m (meter) altitude. The thornless cultivated form was from pilot cultivar of the Bou Argoub region in the northeast of Tunisia, located at 36°32′N latitude, 10°33′E longitude, and 62 m altitude. O. stricta was from a botanical garden in Tunis located at 36°49′N latitude, 10°11′E longitude, and 5 m altitude.

Methanol (CH₄O), acetonitrile (CH₃CN), formic acid (CH₂O₂) and acetic acid (CH₃COOH) requested for chromatographic analysis and Folin Ciocalteu reagent were purchased from Merck Company (Merck, Darmstadt, Germany) and DPPH from Sigma Aldrich (St. Louis, MO, USA). Water was purified on a MilliQ system (Millipore S.A., Molsheim, France). The used standards were provided by Extrasynthese S.A. (Lyon, France) and were the following: rutin (quercetin 3-O-rutinoside), myricitrin (myricetin 3-O-rhamnoside), hyperoside (quercetin 3-O-galactoside), kaempferol 3-O-rutinoside, isorhamnetin 3-O-rutinoside, kaempferide (4′-methylkaempferol), rhamnetin (7-methylquercetin), isorhamnetin (3-methylquercetin), quercitrin (quercetin 3-O-rhamnoside) and myricetin.

Opuntia fruits were washed with distilled water, air-dried, and hand-peeled. Both, peel and pulp were freeze-dried and reduced into powders. The ground dried sample (30 mg) was extracted three times with 1.2 mL MeOH:acetic acid (99:1), sonicated in a water bath at room temperature for 15 min and then centrifuged at 3900 rpm for 15 min (Fish Bioblock Scientific). The argon is used to degas the combined supernatants. The mixture was immediately filtered through polytetrafluoroethylene (PTFE) membrane (0.45 μm) and stored in refrigerator at −30 °C. The described procedure is recommended for the total phenolic assay, total flavonoid assay, and RP-HPLC and ESI-MS analysis [34].

The antioxidant capacity of the methanol extracts was tested by DPPH (1,1-diphenyl-2-picrylhyydrazyl) according to method adopted by Yen and Duth [35]. The DPPH method is the most used for the evaluation of the antioxidant properties and antiradical activity of natural products. DPPH is a stable free radical in a methanolic solution. An aliquot of the extracts above (10 μL, 20 μL, 40 μL, 80 μL, 180 μL, 380 μL, 580 μL, 780 μL, 980 μL, 1180 μL, 1580 μL, 1980 μL, 2380 μL, 2980 μL, 3780 μL) was mixed with 2 mL of DPPH, solution of varying concentrations were obtained. The mixtures were vortexed vigorously for 30 s and then immediately placed in an UV-visible lambda 25 model “Perkin Elmer” spectrophotometer. Scans of the solution were performed at wavelength ranging from 400 to 800 nm. The maximum absorbance was then read at 515 nm. When DPPH is placed in an antioxidant solution, its free radical is inhibited by the antioxidant before an absorbance measurement is performed. The inhibition of free radical DPPH (I%) was calculated as I% = [(A₀−A₁)/A₀] × 100 (1) where A₀ and A₁ are the absorbance values of the blank (all reagents except the test compounds) and of the tested samples, respectively. The I% were plotted against respective concentrations used. The slope of the linear portion of each graph was used to calculate IC₅₀% which is the concentration when 50% of the antioxidant is reduced. IC₅₀% values of the extracts were compared to the IC₅₀% value of a standard antioxidant, Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) used as positive control and obtained by the same procedure.

The amount of total polyphenolic compounds was determined according to Folin–Ciocalteu and adapted from Singleton and Rossi [36]. A 0.1 mL amount of methanol extracts was diluted to 0.5 mL with 2.5% (v/v) acetic acid. A 0.25 mL amount of Folin–Ciocalteu reagent (Merck) was added to 0.5 mL of the diluted extract and was allowed to stand for 3 min at room temperature. One mL of 200 g·L⁻¹ Na₂CO₃ solution was added, and the volume was adjusted to 5 mL with distilled water. The mixture was then heated at 70 °C for 10 min. After cooling and color development, the absorbance was measured at 700 nm with a spectrophotometer UV (Spectrometer Spectra Max 384 PLUS Molecular Devices) using blank samples composed of distilled water and reagents. The amount of polyphenolic compounds is determined by comparing absorbance values of the samples to the absorption values of rutin and gallic acid (GA) (mg rutin or GA·g⁻¹ fresh weight) standards.

Total flavonoid content in the methanol extract was determined by using the colorimetric method (trichloride aluminum method) adopted by Bahorun et al. [37] in Nigella sativa extracts with slight modifications. An aliquot of methanol extracts (0.3 mL) was added to a 5 mL volumetric flask, containing 0.45 mL of distilled deionized water. After 5 min, 0.75 mL of 2% aluminum chloride (AlCl₃·H₂O) solution was added. The mixture was shaken and allowed to rest for 10 min of reaction. The absorbance was measured at 415 nm versus prepared methanol blank with an UV-VIS spectrophotometer (Spectrometer Spectra Max 384 PLUS Molecular Devices). The concentration of flavonoids was determined by comparing absorbance values of the sample to the absorption values of rutin used as a standard. Results were expressed as equivalent rutin (mg rutin·g⁻¹of fresh weight) ± SD (standard deviation) with 3 replications.

RP-HPLC analyses were performed using a Surveyor AS Autosampler including a binary Agilent HP 1100 pumping system, a thermostated column oven and a thermostated automatic injection module. The system was coupled to a TSP UV6000 detector UV used in 240–600 nm range. The column (Merck) was a reversed phase column (150 × 2.1 mm i.d., 3.5 μm, Agilent Eclipse XDB-RP). The injected volume was 4 μL and the column was thermostated at 30 °C. The elution solvent was a mixture of solvent A, consisting of ultrapure water/formic acid (99.9:0.1, v/v) and solvent B, consisting of acetonitrile/formic acid (99.9:0.1, v/v), and the following gradient was applied: initial, 3% B; 0–5 min, 9% B linear; 5–15 min, 16% B linear; 16–45 min, 50% B linear; the gradient was followed by washing and reconditioning of the column. Helium was used for degassing solvents. Two particular wavelengths were used for quantification of polyphenols: 280 nm for phenolic acids, 350 nm for flavonols [30]. Quantitative determination was carried out using calibration curves of standards. Phloretic acid and hyperoside were chosen as external standards for quantification of phenolic acids, and flavonols. ESI-MS analyses were performed in negative mode for phenolic compounds (molecular fragmentation losses hydrogen). For better identification of the peaks, the positive mode was used for betalains (molecular fragmentation with hydrogen acquisition). These tests were performed on an LCQ DECA ion trap mass spectrometer (Thermo-Finnigan, San José, CA, USA) equipped with an ESI source and run by Xcalibur (version 1.2) software. The parameters were as follows: ion spry voltage, 3.69 kV; capillary voltage, −70.78 V; capillary temperature, 240.4 °C; sheath nitrogen gas flow rate, 66.65 (arbitrary units); auxiliary gas flow rate, 3.81 (arbitrary units); scan range of m/z 50–2000. Samples corresponding to collected HPLC peaks were directly introduced into the ESI source by a built-in syringe pump at 3 μL. For the generation of MSⁿ data, the precursor ions were fragmented by helium gas collision in the ion trap by optimizing the collision energy in order to obtain the intensity of the precursor ion close to 10% of the relative scale spectrum.

All experiments were the result of three runs that averaged together. The standard deviations were based on triplicate measurements (n = 3). The value for each sample was expressed as the mean (M) ± standard deviation (SD). The analyses of variance were performed by ANOVA with software SPSS version 11.5 for Windows [38]. Differences among the means were compared using the Fisher–Snedecor distribution with a level of significance p < 0.05.

The antioxidant activity of Tunisian O. stricta and O. ficus indica fruit methanol extracts of both peel and pulp are compared and shown in Table 1. Results show that there is little difference in the antioxidant activity between the spiny and the thornless O. ficus indica (for both peel and pulp). The antioxidant activity of the O. stricta and O. ficus indica are lower compared to trolox (for both peel and pulp). Results also show that the peel extract for the O. stricta has a higher antioxidant activity than the pulp (about 3% higher). However, for the O. ficus indica, the pulp has a higher antioxidant activity than the peel (about 2%). The average antioxidant activity of the peel and the pulp of the O. stricta is higher than the average antioxidant activity for O. ficus indica (about 13% higher). Statistically, the methanol extract from Tunisian Opuntia fruits have similar reductive ability (F_2,6 = 5.14, p < 0.05; Table 1).

Our results are in contradiction with the results of Moussa-Ayoub et al. [39], who used the electron paramagnetic resonance spectroscopy to determine the antioxidant activity. Moussa-Ayoub et al. [39] showed that the O. ficus indica (from Sicily/Italy) peel has an antioxidant activity higher than the pulp (about a 35% discrepancy). They explained their results by the presence of large amounts of flavonols, phenolics, as well as betacyanins in the fruit’s peel of O. ficus indica compared to its pulp. Maataoui et al. [40] showed that the purple juice of the O. ficus indica has a higher antioxidant activity than the yellow-orange juice. Ammar et al. [41] studied flower extracts at the post-flowering stage for both O. stricta and O. ficus indica. Their results show that O. stricta has a higher antioxidant activity than O. ficus indica. All these results indicate that O. stricta has higher antioxidant activity than O. ficus indica, regardless of the part of fruit studied. This is possibly related to the darker color of the flower and the fruit.

Polyphenols are an important group of natural compounds, recently considered to be of high scientific and therapeutic interest. The total polyphenols content (TPC) was estimated in different methanol extracts (Figure 1). Results show that the amounts of polyphenols were much greater in the peel than in the pulp. The TPC in the peel of the spiny O. ficus indica was three-fold higher than the pulp. The TPC in the peel of the thornless was 2.21-fold higher than the pulp. Similar results were observed for the O. stricta, for which the peel was 1.66-fold higher than the pulp. The comparison between the two cactus pear species shows that the average TPC in the O. ficus indica is higher than the average TPC in the O. sticta (1.94-fold higher for spiny form, and 1.76-fold higher for the thornless form). The total phenolic contents were not statistically different among the various forms of Tunisian Opuntia fruits (p < 0.05). In conclusion, results show that the TPC is higher in the peel than in the pulp and is also higher in the O. ficus indica than the O. stricta. These results are in agreement with the work of Moussa Ayoub et al. [39] and Díaz Medina et al. [42]. All of which showed that the highest concentrations of phenolic compounds in fruits occurred in the skin tissue.

Our work shows that there are much less flavonoids than phenolic compounds in all the cactus pear species studied. The total flavonoid content is expressed as a concentration in mg of equivalent rutin g⁻¹ FW. Results show that the flavonoid content in the peel of the O. ficus indica is higher than in the pulp (4-fold higher for the thornless and 2.30-fold higher for the spiny). For the O. stricta, the total flavonoid content in the peel is 1.42-fold higher than the pulp. The total flavonoid content of the thornless O. ficus indica is higher than the total flavonoid content of the O. stricta (1.47-fold higher). The total flavonoid is higher in the thornless O. ficus indica than the spiny (1.76-fold higher). These differences were statistically non-significant (p < 0.05). In conclusion, the peel contained more flavonoids than the pulp; the thornless contained more flavonoids than the spiny, and the O. ficus indica has more flavonoids than the O. stricta. These results are in agreement with the work of Ndhlala et al. [43].

It is known that large amounts of polyphenols and flavonoids increase antioxidant activity. Polyphenols are endowed with potent antioxidant activities in vitro, but in vivo, the scantiness of biomarkers was the major limitation by the lack of robust biomarkers [44]. The monofloral Cuban honey analysis showed high concentrations of phenolic acids and flavonoids, which are responsible, at least in part, for their antioxidant activity [45]. However, this is contrary to what was found in this work. In this study, we observed in the pulp of the O. ficus indica, a higher antioxidant activity and lower content of polyphenols and flavonoids. (Table 1, Figure 2, Figure 3). Recent studies on the antioxidant activity of O. ficus indica and O. stricta flowers reported by Ammar et al. [41] showed that, during the initial flowering stage, the O. stricta contains a low amount of total phenolics and a high antioxidant activity. The antioxidant capacity is determined not only by concentration, but also by several other factors such as the reactivity toward radicals, and the distribution, localization, and fate of antioxidant-derived radicals in interaction with other antioxidants [46]. In order to evaluate the antioxidant capacity, these factors should be separately assessed and considered. We can attribute the antioxidant activity of the pulp to other compounds and explain our results in comparison to Niki’s studies on antioxidant capacity [46]. According to the literature, the content of vitamin C and betalains in Opuntia pulp fruits provided higher antioxidant activity and offered a preventive potential against oxidative stress in the human body. Many investigations reported the non-negligible antioxidant activity of betalains due to their redox potentials. Several works have demonstrated the potent antiradical scavenging activity of betalains in vitro [47]. In conclusion, the antioxidant activity as shown in this work is related to more than the amount of polyphenols and flavonoids.

Very little work is available in the literature on the presence of flavonols in the O. stricta fruit. In this investigation, we identify and compare amounts of flavonols and other bioactive compounds in the peel and pulp of the Tunisian O. ficus indica and the O. stricta. The method used in this work is RP-HPLC coupled with UV and ESI-MS. The identification of the compounds (Table 2) was based on a comparison of their HPLC retention times (RT), UV-visible spectrum, mass spectrometry (MS and MS² fragments) with authentic standards compounds and published data. The betacyanins peaks were identified by order of elution, spectral characteristics, and molecular ions [M + H]⁺ with the red beet methanol extract injected in the same conditions as the samples. According to the available standards, we identified nine chromatographic peaks: A₃₅, A₃₆, B₃₈, B₄₃, C₃₅, C₃₆ and C₄₅ as isorhamnetin-3-O-rutinoside, A₃₆ as isorhamnetin-3-O-glucoside and B₃₇ as quercetin-3-O-rutinoside (Figure 3). All the identified flavonoids belong to the group of flavonol aglycones. Table 2 also shows the quantification of phenolic compounds.

Three types of compounds were detected by UV: two classes of phenolic compounds (flavonols detected at 350 nm and phenolic acids detected at 280 and 320 nm) and betalain pigments detected at 540 nm and 470 nm (Table 2).