Chemical Composition and Antioxidant Activity of Euterpe oleracea Roots and Leaflets

Euterpe oleracea (açaí) is a palm tree well known for the high antioxidant activity of its berries used as dietary supplements. Little is known about the biological activity and the composition of its vegetative organs. The objective of this study was to investigate the antioxidant activity of root and leaflet extracts of Euterpe oleracea (E. oleracea) and characterize their phytochemicals. E. oleracea roots and leaflets extracts were screened in different chemical antioxidant assays (DPPH—2,2-diphenyl-1-picrylhydrazyl, FRAP—ferric feducing antioxidant power, and ORAC—oxygen radical absorbance capacity), in a DNA nicking assay and in a cellular antioxidant activity assay. Their polyphenolic profiles were determined by UV and LC-MS/MS. E. oleracea leaflets had higher antioxidant activity than E. oleracea berries, and leaflets of Oenocarpus bacaba and Oenocarpus bataua, as well as similar antioxidant activity to green tea. E. oleracea leaflet extracts were more complex than root extracts, with fourteen compounds, including caffeoylquinic acids and C-glycosyl derivatives of apigenin and luteolin. In the roots, six caffeoylquinic and caffeoylshikimic acids were identified. Qualitative compositions of E. oleracea, Oenocarpus bacaba and Oenocarpus bataua leaflets were quite similar, whereas the quantitative compositions were quite different. These results provide new prospects for the valorization of roots and leaflets of E. oleracea in the pharmaceutical, food or cosmetic industry, as they are currently by-products of the açaí industry.


Introduction
Euterpe oleracea (E.oleracea), known as açaí worldwide and wassaye in French Guiana, is a palm tree native from the Amazonian rainforest, which gives purple berries with high potential currently used as dietary food supplements or in cosmetics thanks to their high antioxidant activity.Traditionally, açaí berries are consumed as a beverage called "açaí vino" in Brazil.Since the mid 2000s, many scientific studies focused on E. oleracea berries due to its biological activities including anti-proliferative, anti-inflammatory, antioxidant, and cardioprotective properties.E. oleracea was also found to be extremely rich in antioxidant compounds including phenolic acids, flavonoids and anthocyanins [1,2].
While açaí berries are mainly used as foodstuffs, E. oleracea vegetative organs (such as leaves and roots) are traditionally used in medicine, curing snake bites [3], diabetes, kidney and liver pains, fever, anemia, as well as arthritis, or used as hemostatic [4,5].Roots of Euterpe precatoria, a close species to E. oleracea, also showed anti-inflammatory [6] and antioxidant properties [7].The vegetative organs of E. oleracea are considered as waste from the açaí palm heart industry and could be potential sources of phytochemicals with cosmetic, pharmaceutical or nutritional applications.Currently, there is interest in valorizing polyphenols-containing food wastes such as grape skins, olive mill waste water, citrus peels, seed wastes or berry leaves [8,9] as there is a growing demand for polyphenols for pharmaceutical, cosmetic and food industries, especially to treat cardiovascular, and skin diseases, cancer, aging, and diet issues.Some studies have already been carried out on açaí seeds, another waste from the açaí industry.Antioxidant activity [10] and anti-nociceptive activity have been demonstrated [11], showing that this waste could find applications in the food or pharmaceutical industries.Therefore, there is also a real value in enhancing the knowledge of the phytochemical composition and pharmacological activities of açaí roots and leaves.A previous study showed that roots and leaflets of two other Amazonian palm trees (Oenocarpus bacaba and Oenocarpus bataua) had high antioxidant activity, underlining both Oenocarpus vegetative organs as good sources of antioxidant compounds [12].In this paper, we therefore investigated the antioxidant activity of root and leaflet crude extracts of E. oleracea using different chemical antioxidant assays (DPPH-2,2-diphenyl-1-picrylhydrazyl, FRAP-ferric reducing antioxidant power, and ORAC-oxygen radical absorbance capacity), a DNA nicking assay and a cellular antioxidant activity assay, while LC-MS/MS was performed to elucidate the structures of the bioactive compounds.

Antioxidant Activity
2.1.1.Antioxidant Activity in the Chemical Assays (DPPH, FRAP, and ORAC) Antioxidant properties of root and leaflet extracts from E. oleracea were investigated using different chemical assays (DPPH, FRAP and ORAC) having different modes of action: hydrogen transfer (ORAC), electron transfer (FRAP) or a mixed mode (DPPH) [13].In all antioxidant tests, best results were obtained for acetone extracts independently of the organ used.Leaflet extracts were more active than root extracts, independently of the solvent used.
The ORAC assay, considered as one of the most relevant amongst antioxidant chemical assays, gave ORAC values from 1600 to 2200 µmol TEq/g DM, for leaflet extracts.They were at least three-fold more active than palm berries in the ORAC assay [14], at least 1.5-fold more active than the best Oenocarpus leaflet extracts [12] and almost equivalent to green tea with 2400 µmol TEq/g DM (Table 1).Besides, even if E. oleracea roots were less active than E. oleracea leaflets in all the antioxidant tests, their activities were as high or higher than those of E. oleracea and O. bataua berries, as illustrated by ORAC results (300-1300 µmol TEq/g DM) (Table 1).
Root and leaflet extracts from E. oleracea were active across the board in the different chemical antioxidant assays (DPPH, FRAP, and ORAC) showing that they act according to different mechanisms of action such as electron and/or hydrogen transfer.Root and leaflet extracts were at least as active as palm berries, considered as super fruits.Leaflet extracts were the most antioxidant extracts, almost equivalent to green tea leaves, which have one of the strongest recorded antioxidant activity.Despite optimization of the DNA nicking assay conditions, the sensitivity using methanol extracts was low [16], therefore only aqueous and acetone extracts of E. oleracea were assessed in the DNA nicking assay.All root extracts of E. oleracea were found to be active in this assay.The aqueous and acetone root extracts had an antioxidant effect at all concentrations tested; they protected form I from degradation and reduced the formation of linear nicked form III in a similar way to Trolox at 1 mg/mL (Figure 1).Aqueous leaflet extracts had an antioxidant effect in the DNA nicking assay at all concentrations tested (1 mg/mL and 10 mg/mL), while acetone leaflet extracts had an antioxidant effect at 10 mg/mL and a prooxidant effect at a lower concentration of 1 mg/mL.The overall high potency of the root and leaflet extracts of E. oleracea in the DNA nicking assay encouraged the evaluation of their antioxidant activity in an assay physiologically more relevant, i.e., a cell-based antioxidant assay.Despite optimization of the DNA nicking assay conditions, the sensitivity using methanol extracts was low [16], therefore only aqueous and acetone extracts of E. oleracea were assessed in the DNA nicking assay.All root extracts of E. oleracea were found to be active in this assay.The aqueous and acetone root extracts had an antioxidant effect at all concentrations tested; they protected form I from degradation and reduced the formation of linear nicked form III in a similar way to Trolox at 1 mg/mL (Figure 1).Aqueous leaflet extracts had an antioxidant effect in the DNA nicking assay at all concentrations tested (1 mg/mL and 10 mg/mL), while acetone leaflet extracts had an antioxidant effect at 10 mg/mL and a prooxidant effect at a lower concentration of 1 mg/mL.The overall high potency of the root and leaflet extracts of E. oleracea in the DNA nicking assay encouraged the evaluation of their antioxidant activity in an assay physiologically more relevant, i.e., a cell-based antioxidant assay.
Root and leaflet extracts from E. oleracea were quite active in the cellular antioxidant assay with the median effective concentration (EC 50 ) ranging from 10 to 30 µg/mL (Table 1).Most of the palm extracts were more active than fruits including kiwis, blueberries, carrots and strawberries, which had a high antioxidant activity (EC 50 from 50 to 200 µg/mL) in a similar cell-based assay [18].
Leaflet extracts of E. oleracea were more active in the CAA assay (45-208 µmol Quercetin Eq/g DM) than root extracts (10-30 µmol Quercetin Eq/g DM) and acetone extracts were more active than methanol extracts.Activities for E. oleracea were quite similar to Oenocarpus roots (10-50 µmol Quercetin Eq/g DM), whereas E. oleracea leaflets were at least two-fold superior than Oenocarpus leaflets (7-100 µmol Quercetin Eq/g DM) [12].Interestingly, there was a very high correlation between the ORA activity and the cell-based activity (Pearson's r correlation coefficients of 0.976) (Figure S1), which are both based on the ROO• radical scavenging capacity of the extracts.These correlations were also found in the work of Girard-Lalancette et al., 2009 and Leba et al., 2016 [12,18].The high potency of extracts in the ORAC assay translated well in the CAA assay, meaning that compounds were able to act in a cellular environment, either by crossing membrane cells or by directly scavenging radicals outside the cells.These results encourage investigating the chemical composition of E. oleracea root and leaflet extracts.

Total Phenolic Content (TPC)
The Total Phenolic Content (TPC) of root and leaflet extracts from E. oleracea ranged from 14 to 84 µg Gallic acid equivalent/mg Dry Matter (µg GAEq/mg DM) (Table 1).
Best results were found with acetone independently of the organ used.The highest values were found for the leaflet extracts of E. oleracea, with 84 µg GAEq/mg DM, which was two-fold higher than the TPC of palm berries including E. oleracea, O. bataua [14] and O. bacaba [15], known to be antioxidants.The TPC of E. oleracea roots were globally similar to that of Oenocarpus roots extracts [12], whereas E. oleracea leaflets were richer in polyphenols than Oenocarpus leaflets [12].Moreover, the TPC of E. oleracea leaflets was almost similar to the TPC of green tea leaves (100-125 µg GAEq/mg), which has one of the strongest recorded antioxidant activities.There was a high correlation between the total polyphenolic content (TPC) and the antioxidant activity in the different assays (p < 0.05) with high Pearson's r correlation coefficients between TPC and DPPH (0.945), TPC and FRAP (0.983), TPC and ORAC (0.986), and TPC and CAA activity (0.968) (Figure S1).

Identification of Compounds in Root and Leaflet Extracts by LC-MS/MS
The UV chromatograms recorded at 320 nm showed that both E. oleracea root and leaflet extracts (Figure 2) contained polyphenols, whose structures (Figure 3) were then determined by LC-MS/MS.Leaflet extracts showed more complex polyphenol profiles (fourteen compounds) than root extracts (six compounds), while some peaks (1, 2, and 3) were common to both extracts.In positive mode, the ions at m/z 163 and 145, produced by the loss of the quinic acid moiety [19] were not conclusive of the structure of the CQA.The structure of CQA isomers was assigned according to their diagnostic ions at m/z 191, 179 and 173 in the negative mode [20].Compound 2   In positive mode, the ions at m/z 163 and 145, produced by the loss of the quinic acid moiety [19] were not conclusive of the structure of the CQA.The structure of CQA isomers was assigned according to their diagnostic ions at m/z 191, 179 and 173 in the negative mode [20].Compound 2 In positive mode, the ions at m/z 163 and 145, produced by the loss of the quinic acid moiety [19] were not conclusive of the structure of the CQA.The structure of CQA isomers was assigned according to their diagnostic ions at m/z 191, 179 and 173 in the negative mode [20].Compound 2 (t R 7.2 min) was assigned to 4-CQA, having a base peak at m/z 173 in the MS 2 spectrum.Compounds 1 (t R 4.2 min) and 3 (t R 9.1 min), both having m/z 191 as a base peak were distinguished by the intense ion at m/z 179 in the MS 2 spectrum, characteristic of 3-CQA for compound 1, whereas compound 3 was assigned to 5-CQA (Table 2).The retention order of the CQA isomers (3 < 4 < 5) was similar to the study of Regos and Treutter, 2010 [21] using a similar PFP column.

Characterization of Caffeoylshikimic Derivatives (Mr = 336) in Root Extracts
In the root extracts, three peaks (4, 5, 6) at t R 12.2, 14.2 and 16.2 min yielded [M − H] − ions at m/z 335, [M + Na] + at m/z 359, and showed UV spectra (λ = 298 nm and 320 nm) corresponding to caffeoylshikimic acids (CSA, Mr = 336).In positive mode, the ions at m/z 163 and 145 were characteristic of the caffeoyl moiety [19].The structure of CSA was tentatively assigned according to their diagnostic ions at m/z 317, 291, 179, 161 and 135 in negative mode [22][23][24] (Table 2).The fragmentation pattern of compound 4 with an intense ion at m/z 161 was conclusive of a 4-CSA.The absence of the ion at m/z 161 in the MS 2 spectrum of compound 5 and the presence of ions at m/z 317 and m/z 291 was indicative of a 5-CSA isomer.The fragmentation pattern of compound 6 was close to that of a 4-CSA isomer but was not conclusive of the structure.  spectrum showed a first loss of 80 Da, characteristic of sulfates, followed by successive losses of 90 Da, 120 Da and 150 Da, characteristics of hexose residues in both positive and negative modes.Compounds 8 and 9 were tentatively assigned to 6,8-di-C-hexosyl apigenin sulfates.The lack of a hypsochromic shift in the band I of 8 and 9 compared to compound 7 suggest the sulfation occurred on position 7 or on the sugar moieties [26].Six peaks (10, 12, 14, 15, 16) (Mr = 564) at t R 21.2, 22.3, 23.2, 24.9 and 25.6 min yielded [M − H] − ions at m/z 563 in the negative mode.Their MS 2 spectrum showed losses of 60 Da, 90 Da, 120 Da and 150 Da, characteristic of pentose and hexose residues.They were assigned as 6,8-di-C-pentosyl-hexosyl apigenin derivatives.The [M − H-60] − ion resulting from the fragmentation of the pentose moiety was more intense in 12 (100%) and 16 (25%) than in other compounds, indicating that the pentose moiety was on position 6, which fragments preferentially over sugars on position 8 [25].Therefore, compounds 12 and 16 were identified as 6-C-pentosyl-8-C-hexosyl apigenin isomers.The others compounds (10, 14, and 15) were assigned as 6-C-hexosyl-8-C-pentosyl apigenin isomers.[25].Based on retention order and injection of synthetic standards, 11 was identified as 8-C-glycosyl luteolin and 13 as 6-C-glycosyl luteolin.

Quantification of Chemical Compounds in Root and Leaflet Extracts
Root extracts of E. oleracea were mainly characterized by the presence of hydroxycinnamic acids HCA, namely three caffeoylquinic acids (CQA) and three caffeoylshikimic acids (CSA) (Figure 4).Hydroxycinnamic acids are known to display biological properties such as antiviral, anti-inflammatory and antioxidant activity [27].They also play important physiological role of defense against pathogens and disease resistance [27], which is essential for vulnerable vegetative organs such as roots.The CSA content was higher than the CQA content in the root extracts, (2.5 mg/g DM against 0.5 mg/g DM for the most concentrated extract), i.e., 70%-85% of all HCA (Figure 4).5-CQA was the major compound amongst CQA, while each of the three CSA was found in almost similar amounts.The qualitative composition of E. oleracea roots was similar to that of O. bataua and O. bacaba [12].However, the quantitative compositions of O. bacaba were different with a global two-fold lower content in compound compared to E. oleracea.Di-C-glycosyl derivatives of apigenin (compounds 7-9) were the major compounds in leaflet extracts of E. oleracea, accounting for almost 50% of all C-glycosylflavones, the sulfated flavones 8 and 9 being the major ones with about 2 mg/g DM.Apigenin C-glycosides are phytochemical markers specific to a few Arecaceae species, while C-glycosyl luteolin derivatives are more common amongst palm tree species [28].More recently, C-glycosyl derivatives of apigenin, close to those found in this work, were identified in another Amazonian palm tree species, Mauritia flexuosa [29].The presence of sulfated flavones in the leaflet extracts of E. oleracea is interesting as natural sulfated flavonoids inspire the design of smaller bioactive compounds with sulfate groups [30].E. oleracea leaflets composition was qualitatively similar to that of Oenocarpus leaflets [12] but the quantitative composition was conspicuously different (maximum 1500 and 5000 µg/g DM for O. bataua and O. bacaba, respectively, against 16000 µg/g DM for E. oleracea).It was interesting to observe that these three palm tree species (E.oleracea, O. bacaba and O. bataua) have the same qualitative composition of their roots or of their leaflets but have quite different antioxidant activities.Indeed, E. oleracea leaflets were two fold more active than O. bacaba and O. bataua leaflets.The difference of activity could come from the quantity of polyphenolic compounds in the extracts, rather than the qualitative composition.These results interrogate the fact that there might be a particular regulation of key enzymes in the biosynthetic pathway of polyphenols in E. oleracea leaflets.To validate this hypothesis, numerous molecular and biochemical studies will be required.2); Di-Glc-Api: 6,8-di-C-hexosyl apigenin; Di-Glc-Api Sulf: 6,8-di-C-hexosyl apigenin sulfate; n = 3 biological replicates.

Relationship between the Chemical Composition and the Antioxidant Activity
Principal Component Analysis (PCA) was carried out using the data from the chemical antioxidant assays (DPPH, FRAP, ORAC) and the quantification of compounds in root and leaflet extracts of E. oleracea.The antioxidant activity of the roots was linked to the CSA and CQA contents (negative side of PC1) (Figure 5).There was no clear differentiation of the methanol and acetone root extracts according to the PCA, as they had very similar antioxidant activity and chemical composition.
For the leaflet extracts, the most active extracts in all the chemical assays were on the positive side of PC2, which was correlated with high contents of flavones 8 and 9 (Figure 5).These results translated quite well to the CAA assay, where the leaflet extracts of E. oleracea were very active.The PCA highlighted a correlation between the antioxidant activity of E. oleracea leaflet extracts and the presence of di-C-glycosyl apigenin sulfates (8 and 9).Indeed, hydroxyl substituted flavones Leaflet extracts were characterized by two main compound families: three caffeoylquinic acids, as in the root extracts, and eleven flavones (Figure 4).5-CQA was the main CQA in the leaflet extracts (40%-70% of total CQA).Higher contents of CQA (up to 4.5 mg/g DM) were found in the leaflet extracts compared to the root extracts.Leaflet flavones were quantified mainly as C-monoglycosyl and C-diglycosyl derivatives of apigenin, and to a lesser extent as C-glycosyl derivatives of luteolin.Di-C-glycosyl derivatives of apigenin (compounds 7-9) were the major compounds in leaflet extracts of E. oleracea, accounting for almost 50% of all C-glycosylflavones, the sulfated flavones 8 and 9 being the major ones with about 2 mg/g DM.Apigenin C-glycosides are phytochemical markers specific to a few Arecaceae species, while C-glycosyl luteolin derivatives are more common amongst palm tree species [28].More recently, C-glycosyl derivatives of apigenin, close to those found in this work, were identified in another Amazonian palm tree species, Mauritia flexuosa [29].The presence of sulfated flavones in the leaflet extracts of E. oleracea is interesting as natural sulfated flavonoids inspire the design of smaller bioactive compounds with sulfate groups [30].E. oleracea leaflets composition was qualitatively similar to that of Oenocarpus leaflets [12] but the quantitative composition was conspicuously different (maximum 1500 and 5000 µg/g DM for O. bataua and O. bacaba, respectively, against 16000 µg/g DM for E. oleracea).It was interesting to observe that these three palm tree species (E.oleracea, O. bacaba and O. bataua) have the same qualitative composition of their roots or of their leaflets but have quite different antioxidant activities.Indeed, E. oleracea leaflets were two fold more active than O. bacaba and O. bataua leaflets.The difference of activity could come from the quantity of polyphenolic compounds in the extracts, rather than the qualitative composition.These results interrogate the fact that there might be a particular regulation of key enzymes in the biosynthetic pathway of polyphenols in E. oleracea leaflets.To validate this hypothesis, numerous molecular and biochemical studies will be required.

Relationship between the Chemical Composition and the Antioxidant Activity
Principal Component Analysis (PCA) was carried out using the data from the chemical antioxidant assays (DPPH, FRAP, ORAC) and the quantification of compounds in root and leaflet extracts of E. oleracea.The antioxidant activity of the roots was linked to the CSA and CQA contents (negative side of PC1) (Figure 5).There was no clear differentiation of the methanol and acetone root extracts according to the PCA, as they had very similar antioxidant activity and chemical composition.

Chemicals and Reagents
Solvents used for extraction and LC-MS analysis were of HPLC grade, obtained from Carlo Erba Reagents (Val de Reuil, France).Standards of 5-O-caffeoylquinic acid, orientin (>99%), isoorientin (>99%), and isovitexin (>99%) were obtained from Extrasynthèse (Genay, France).Gallic acid-1hydrate (99%) was purchased from Panreac Quimica (Barcelona, Spain).For the leaflet extracts, the most active extracts in all the chemical assays were on the positive side of PC2, which was correlated with high contents of flavones 8 and 9 (Figure 5).These results translated quite well to the CAA assay, where the leaflet extracts of E. oleracea were very active.The PCA highlighted a correlation between the antioxidant activity of E. oleracea leaflet extracts and the presence of di-C-glycosyl apigenin sulfates (8 and 9).Indeed, hydroxyl substituted flavones such as apigenin or luteolin derivatives are known to have high antioxidant activity due to their structures [31].

Plant Materials
Roots and leaflets of three specimens of E. oleracea were harvested on February 2013, in Macouria, French Guiana.Samples were washed, cut, freeze-dried and grinded immediately after collection.Dried matter was then stored at −20 • C to limit degradation until extraction.

Extraction
Extractions were performed following the method previously published [12].Three solvents were used for extraction of roots and leaflets from E. oleracea and green tea leaves: water (W), methanol/water 70/30 v/v (M) and acetone/water 70/30 v/v (A).The combination of acetone or methanol with water is commonly used to extract polyphenols [32].Ultrasounds were used to assist the extraction of polyphenols of leaflets and roots from E. oleracea, as they are known to be a green alternative to conventional extraction methods and give optimized extraction efficiency and reduced extraction time [33].Preliminary studies using ultrasound extraction were carried out to determine that four successive extractions gave optimized extraction efficiency.

Total Phenolics by Folin-Ciocalteu
Total Phenolic Content (TPC) was determined using the Folin-Ciocalteu method as previously described [12].

Chemical Antioxidant Assays (DPPH, FRAP, ORAC)
Chemical antioxidant properties were assessed using DPPH, FRAP and ORAC assays as described in our previous paper [12].These assays are based on different antioxidant mode of action like electron transfer (FRAP), hydrogen transfer (ORAC) or a mixed mechanism implying both electron and hydrogen transfer (DPPH) [13].

Cytotoxicity Assay
Cytotoxicity of the extracts in NHDF was measured using the MTT (3-(4,5-dimethylthiazol-2 -yl)-2,5-diphenyltetrazolium bromide) method [34].NHDF cells were incubated for 24 h at 37 • C, in presence or absence of E. oleracea extracts at 100, 200, 300, 400 and 500 µg/mL.The cells were washed twice with 500 µL of a phosphate buffer solution (PBS), then incubated at 37 • C for 3 h with MTT at 0.5 mg/mL.MTT was removed using 1 mL of DMSO and the plate was kept in the dark at room temperature during 30 min before recording the absorbance at 595 nm using a plate-reader (Dynex, Magellan Biosciences, Tampa, FL, USA) to detect viable cells.A decrease of the absorbance by more than 20% was set as the limit for cytotoxicity.

Cellular Antioxidant Activity (CAA) Assay
The cell-based assay was done according to the method of Wolfe and Liu [35], with some modifications.Briefly, NHDF cells were seeded for 24 h on a 96-well microplate at a density of 11 × 10 4 cells by well in 100 µL of RPMI growth medium.The growth medium was removed and the wells were washed with PBS.Wells were treated in triplicates with 100 µL of extracts at four concentrations (10, 25, 50 and 100 µg/mL), not cytotoxic to the NHDF cells, plus 50 µM 2 ,7 -dichlorodihydrofluorescein diacetate (DCFH-DA) dissolved in the treatment medium and kept in the dark for 1 h.Control wells (cells treated with 50 µM of DCFH-DA, without extract), and blank wells (cells treated only with growth medium) were included.Then, the wells were drained from the treatment medium and 100 µL of 2,2 -Azobis(2-amidinopropane) dihydrochloride (AAPH) at 250 µM was added in each well (except in blank wells).Immediately after the AAPH addition, the plate was placed into a plate-reader (Dynex, Magellan Biosciences) at 37 • C and the fluorescence was recorded for 30 min (λ excitation 485 nm, λ emission 538 nm).After blank subtraction from the fluorescence readings, the area under the curve of fluorescence versus time was integrated to calculate the CAA value at each extracts concentration tested as follows: CAA unit: 100 − ( SA/ CA) × 100 where SA is the integrated area under the sample fluorescence versus time curve and CA is the integrated area from the control curve.The median effective concentration EC 50 (µg/mL), was determined for E. oleracea extracts from the median effect plot of log (ƒa/ƒu) versus log (concentration), where ƒa is the fraction affected and ƒu is the fraction unaffected by the treatment.Quercetin was used as a standard in the CAA assay and the cellular antioxidant activity of E. oleracea extracts was then expressed in µmol Quercetin equivalents/g of dried matter (µmol QEq/g DM).

DNA Nicking Assay
The DNA nicking assay was performed using the pUC18 plasmid.All conditions required to analyze aqueous and organic extracts were previously optimized [16].Four microliters of E. oleracea extracts (max 7% acetone) at various concentrations were added to 4 µL of the Fenton reaction mixture, containing: plasmid DNA (150 µg/µL), phosphate buffer (50 mM, pH 7.4), H 2 O 2 (30 mM), FeSO 4 (2 mM/8 mM) and EDTA-Na 2 (3.75 mM/15 mM) for aqueous and acetone extracts respectively.The final volume was adjusted to 24 µL using distilled water and incubated for 20 min and 15 min at 37 • C for aqueous and acetone extracts respectively.Two microliters of loading dye was added to the incubated mixture, and 10 µL were loaded onto 1% (w/v) agarose gel.After electrophoresis in TAE buffer (0.04 M tris-acetate and 1 mM EDTA, pH 7.4) using a DNA subcell (Bio-Rad Laboratories, Hercules, CA, USA), the agarose gel was stained with ethidium bromide for 15 min and DNA bands were analyzed using a Bio-Rad Gel Doc™ XR (Bio-Rad Laboratories).The prooxidant and antioxidant activity of the extracts was assessed using band intensity (NIH Image J) of form I (supercoiled DNA) and form III (nicked linear DNA) compared to the appropriate controls.DNA incubated with and without Fenton's reagent was used as positive and negative control, respectively.Trolox was used as a control of DNA protection, at a concentration of 0.01 and 0.1 mg/mL for aqueous and acetone extracts, respectively.

Analysis by LC-MS/MS
LC-MS/MS analysis was performed on an ion trap (500-MS Varian, Palo Alto, CA, USA) equipped with an electrospray ionization source and coupled to a Pro Star LC system (Agilent Technologies, Santa Clara, CA, USA) and a Pro Star 335 PDA (Agilent).Extracts were filtered using 0.2 µm PTFE filters.Twenty microliters of samples were injected onto a Kinetex PFP column, 100 × 4.6 mm, 2.6 µm (Agilent) using a gradient of 1% formic acid (A) and acetonitrile (B) at 25 • C. The gradient was as follows: 5%-10% B in 10 min, 10%-20% in 10 min, 20% held for 10 min, 20%-100% in 5 min, 100% held for 5 min, and returned to initial conditions for 9 min to reequilibrate the column.The flow rate was 1 mL/min.The total run time per sample was 60 min.
Analyses were first performed using Full scan mode (m/z 120-750), both in negative mode and positive mode to identify the molecular ions and then in TurboDDS ™ mode (Data Dependent Scanning, Varian) to acquire MS 2 spectra.Helium was used as the damping and collision gas at 0.8 mL/min.The operation parameters were as follows: nebulizer gas pressure 50 psi, drying gas pressure 25 psi, drying gas temperature 350 • C, needle voltage −5000 V/5000 V, sprayshield voltage −600 V, spray chamber 50 • C, capillary voltage 100 V/−70 V (negative ionization mode/positive ionization mode).

Statistical Data Analysis
Results of antioxidant activity were presented as the mean of three biological replicates and standard error (SE) or standard error of the mean (SEM).Comparisons between extracts in the TPC, ORAC, FRAP, DPPH and DNA nicking assays were performed using an ANOVA followed by multiple comparisons using Fisher's Least Significant Difference test (p < 0.05).Principal component analysis (PCA) was performed on the phytochemical composition and chemical antioxidant activity (DPPH, FRAP, and ORAC) of roots and leaflet extracts from E. oleracea using the Statistica program (Statsoft, Paris, France).

Conclusions
Root and leaflet extracts from E. oleracea were active across the board in different chemical antioxidant assays (DPPH, FRAP, ORAC, DNA nicking) and biological assay (CAA).The high antioxidant activity of the root and leaflet extracts of E. oleracea was respectively correlated to the presence of hydroxycinnamic acids and apigenin C-glycosides.The high antioxidant activity of E. oleracea leaflets seemed to come from their highest content of polyphenols, suggesting a particular regulation of the polyphenols biosynthetic pathway.This study shows for the first time that Euterpe oleracea roots and leaflets, which are currently by-products of the palm heart industry, could, along with the berries, be valorized as a new non-cytotoxic source of antioxidants containing hydroxycinnamic acids and flavonoids for pharmaceutical, nutraceutical or cosmetic applications.

Figure 3 .
Figure 3. Structure of main components of root and leaflet extracts of Euterpe oleracea; Glc: glucose.

Figure 3 .
Figure 3. Structure of main components of root and leaflet extracts of Euterpe oleracea; Glc: glucose.2.3.1.Characterization of Caffeoylquinic Derivatives (Mr = 354) in Root and Leaflet Extracts In the root and leaflet extracts, three peaks (1, 2, and 3) at retention time (tR) 4.2, 7.2 and 9.1 min yielded [M − H] − ions at m/z 353 and [M + Na] + at m/z 377 and showed UV spectra (at λ = 298 nm and 320 nm) characteristic of caffeoylquinic acids (CQA, Mr = 354).In positive mode, the ions at m/z 163 and 145, produced by the loss of the quinic acid moiety[19] were not conclusive of the structure of the CQA.The structure of CQA isomers was assigned according to their diagnostic ions at m/z 191, 179 and 173 in the negative mode[20].Compound 2

Figure 3 .
Figure 3. Structure of main components of root and leaflet extracts of Euterpe oleracea; Glc: glucose.
Int. J. Mol.Sci.2017, 18, 61 10 of 15 such as apigenin or luteolin derivatives are known to have high antioxidant activity due to their structures[31].

Table 1 .
Antioxidant activity and total phenolic content (TPC) of root and leaflet extracts of Euterpe oleracea.

Table 1 .
Antioxidant activity and total phenolic content (TPC) of root and leaflet extracts of Euterpe oleracea.

Table 2 .
Identification of main components of root and leaflet extracts of Euterpe oleracea.Compounds 11 and 13 (Mr = 432) gave [M − H] − at m/z 447 and [M + H] + at m/z 449.Characteristics losses of 90 Da and 120 Da were diagnostics of hexose residues.The fragments ions at m/z 327 (aglycone + 42) and m/z 357 (aglycone + 72) in the negative mode clearly indicated C-glycosyl luteolin derivatives