Food Science and Technology 61 (2015) 346E351

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

LWT - Food Science and Technology 61 (2015) 346e351

Contents lists available at ScienceDirect

LWT - Food Science and Technology

journal homepage: www.elsevier.com/locate/lwt

Chemical characterization and use of artichoke parts for protection from oxidative stress in canola oil

* Thiago Claus , Swami A. Maruyama, Sylvio V. Palombini, Paula F. Montanher, Elton G. Bonafe, Oscar de Oliveira Santos Junior, Makoto Matsushita, Jesuí V. Visentainer

State University of Maringa, Department of Chemistry, Av. Colombo, 5790, 87020-900, Maringa, Parana State, Brazil

article info abstract

Article history: The residues of artichoke processing are commonly discarded by the food industry owing to lack of Received 4 August 2014 detailed knowledge of their chemical composition, which would facilitate research into practical uses for Received in revised form them. We therefore evaluated proximate and fatty acid compositions, total phenolic compound content 13 December 2014 and total antioxidant activity of artichoke parts (bracts, receptacle and spikes) cultivated in Brazil and Accepted 21 December 2014 performed principal components analysis. Analysis indicated higher concentrations of phenolic com- Available online 30 December 2014 pounds and antioxidant activity in the spikes than in the bracts and receptacle. In a follow-up experiment we assessed the lipid protective value of artichoke parts in tests on canola oil using a new instrument, Keywords: “ ” Antioxidants called Oxitest . The addition of 5 g/100 g of spikes almost doubled the canola oil induction point. These Fat acids results suggest that artichoke spikes may have industrial applications as natural antioxidant additives to Oxitest foods such as canola oil. QUENCHER procedure © 2014 Elsevier Ltd. All rights reserved. PCA

1. Introduction edible; cultivation produces a large volume of byproducts which are usually discarded. The artichoke plant (Cynara cardunculus var. scolymus (L.) Fiori) The edible parts of artichoke and its byproducts contain many originated in southern Europe, in the Mediterranean region, and it chemical compounds known for their beneficial effects in the hu- belongs to the Asteraceae family, which also includes daises and man body, such as phenolic acids, flavonoids and sesquiterpenes, sunflowers (Lombardo et al., 2010). It is the most important hor- these compounds help to protect the liver, reduce blood cholesterol ticultural crop in Italy, followed by potato and tomato (Fratianni, levels, have anticarcinogenic activity, and increase the volume of Tucci, De Palma, Pepe, & Nazzaro, 2007). Italy is also the world's bile secretion from liver as well as the rate of discharge from the biggest producer of artichokes with about 480,000 tons harvested digestive system (Jacociunas et al., 2013; Lattanzio, Kroon, Linsalata, in the year of 2010, followed by Spain (167,000 tons) and France & Cardinali, 2009; Noldin et al., 2003; Pandino, Lombardo, (42,000 tons). Together these three countries are responsible for Mauromicale, & Williamson, 2011; Shen, Dai, & Lu, 2010; 85% of the total global cultivation area (Lutz, Henríquez, & Escobar, Sonnante, De Paolis, Lattanzio, & Perrino, 2002). As well as these 2011; Saez, Fasoli, D'Amato, Simo-Alfonso, & Righetti, 2013). Arti- benefits the compounds listed above are also responsible for the choke was introduced to Brazil at the beginning of the 20th century antioxidant activity of artichoke parts; they stabilize free radicals in by Italian immigrants, and nowadays it is mainly grown in the the human body (Biglari, Alkarkhi, & Easa, 2008; Ferracane et al., states of Sao~ Paulo and Rio Grande do Sul. It is important to 2008; Rufino, Alves, Fernandes, & Brito, 2011; Vasco, Ruales, & remember that Brazilian consumption much lower than European Kamal-Eldin, 2008). consumption and sale and production costs are higher because of The ability of antioxidant compounds to delay or inhibit this lower demand, which reduces accessibility (Di Giulio, 2004). oxidation of substrates has other applications, including protecting Costs are also high because only a small part of the artichoke is edible fats and oils e and even engine lubricants (Focke et al., 2012; Impellizzeri et al., 2012) e against oxidation. In many cases the additive used is a synthetic antioxidant such as * Corresponding author. Tel.: þ55 4430113661. 2,6-bis(1,1-dimethylethyl)-4-methylphenol (BHT) due to the high E-mail address: [email protected] (T. Claus). activity per unit weight of these compounds, however there is

http://dx.doi.org/10.1016/j.lwt.2014.12.050 0023-6438/© 2014 Elsevier Ltd. All rights reserved. T. Claus et al. / LWT - Food Science and Technology 61 (2015) 346e351 347 evidence that they may have harmful effects on human health Official Method 942.05 and crude protein (CP) was measured ac- (Arabshahi-Delouee & Urooj, 2007; Botterweck, Verhagen, cording to AOAC Official Method 960.52 (1990), using a factor of Goldbohm, Kleinjans, & Van Den Brandt, 2000; Dolatabadi & 6.25 to convert percentage nitrogen to percentage protein (Asso- Kashanian, 2010; LeClercq, Arcella, & Turrini, 2000). Chemical and ciation of Official Analytical Chemists AOAC, 2000). Crude fiber pharmacological research is therefore underway to find natural content was determined according to the IAL method (Instituto antioxidants which are suitable for use in the food industry (Moure Adolfo Lutz, 2008, chap. 4) using digestion with sulfuric acid et al., 2001). Byproducts of artichoke represent a potential alterna- 0.23 mol/L and sodium hydroxide 0.66 mol/L. tive to synthetic antioxidants, because of their low cost and the re- ported antioxidant activity of artichoke cultivated in other 2.4. Fatty acid composition countries; however there is very limited data on the antioxidant content and activity of artichokes cultivated in Brazil. Fatty acid methyl esters were prepared by saponification of Samples of this type can also be tested for ability to protect a sample lipid contents with sodium hydroxide 0.5 mol/L (in meth- substrate against oxidation. Lipid matrices rich in unsaturated fatty anol) followed by methylation with triboronfluoride (12.0 ml BF3 in acids (UFAs) are usually more prone to oxidative degradation than 100.0 ml methanol) (Joseph & Ackman, 1992) and separated by gas saturated fatty acids (SFAs). The fatty acid composition of canola oil chromatography (GC) using a Thermo 3300 gas chromatograph (7.3 g/100 g saturated fatty acids; SFAs, 66.9 g/100 g mono- equipped with a flame ionization detector (FID) and a fused-silica unsaturated fatty acids; MUFAs and 25.8 g/100 g polyunsaturated CP-7420 (SELECT FAME) capillary column (100 m 0.25 mm fatty acids; PUFAs; Huang et al., 2008) makes it an excellent sub- i.d. 0.25 mm cyanopropyl). The operation parameters were as strate for lipid degradation tests. follows: detector temperature, 240 C; injection port temperature, The aim of this study was therefore to determine the antioxidant 230 C. Column temperature was maintained at 165 C for 7 min, capacity of artichoke using DPPH radical capture assay, alone and in then raised by 4 C/min to 185 C and maintained at this temper- combination with the QUENCHER procedure, and to evaluate the ature for 4.67 min. Following this the temperature was raised to proximate and fatty acid composition of artichoke parts in terms of 235 C at a rate of 6 C/min and maintained at this level for 5 min, total phenolic content using high performance liquid chromatog- giving a 30-min chromatographic run. Carrier gas flow (hydrogen) raphy (HPLC). The ability of samples to delay canola oil oxidation was 1.2 ml/min; nitrogen at 30 ml/min was used as the makeup gas. was tested using a new instrument, “Oxitest”. Data on antioxidant The sample split ratio was 1:80. activity were correlated with themselves and with Oxitest data and Fatty acids were identified by comparing retention times to assays of total phenolic content to extract more information about those of standard methyl esters (Sigma, USA). Quantification of the chemical composition of sampled artichoke parts. fatty acids (FAs) was performed using tricosanoic acid methyl ester (Sigma, USA) as an internal standard (Joseph & Ackman, 1992). Peak 2. Materials and methods areas were determined using the Chromquest 5.0 software.

2.1. Sampling 2.5. Preparation of extracts

About 36 kg of artichokes (Cynara cardunculus var. scolymus (L.) Methanolic, ethanolic and aqueous extracts were obtained from Fiori) were acquired commercially in three periods separated by all prepared sample in proportions of 1:10 (m/v). Methanolic and 30-day intervals, in the city of Maringa, Parana state, Brazil. Sam- ethanolic extracts were stirred for 4 h at room temperature (25 C) ples were washed with running water, dried with paper towels and then subjected to rotary evaporation to remove solvents. Aqueous manually separated into bracts, spikes and receptacle (heart). extracts were stirred for 8 min at 80 C. After cooling the solvent Samples from the three different sampling periods were combined was sublimated in a lyophilizer (model Alpha 1-2 LD plus, Martin and homogenized in order to minimize the effects of edaphocli- Christ, Germany). All extracts were stored in eppendorfs, protected matic variation, which might affect the properties of interest in from light, in a freezer at 18 C until the analysis. these samples. Artichoke parts were dried in a laboratory oven with forced ventilation at 45 C until constant weight was attained. 2.6. Analysis of antioxidant capacity using conventional DPPH assay Dried samples were packed under vacuum in polypropylene bags and kept in a freezer at 18 C. Prior to analysis samples were Initially, the antioxidant capacity of extracts was measured triturated in a knife mill and passed through a 0.177 mm sieve to through a conventional DPPH (2,2-Diphenyl-1-picrylhydrazyl) free ensure that particle size did not influence the analytical procedures radical capture method, following the procedure described by Brand- (Gokmen,€ Serpen, & Fogliano, 2009). Williams, Cuvelier, and Berset (1995), with some modifications (Ma et al., 2011). Extracts were prepared as 2.0 mg/ml solutions then 2.2. Reagents and standards 25.0 ml aliquots were mixed with 2.0 ml of a methanolic solution of 6.25 10 5 mol/L DPPH. The mixture was kept in the dark for 30 min Reagents and solvents in analytical and HPLC grades were ac- and absorbance at 517 nm was measured in a spectrophotometer quired from Fisher Scientific (Fair Lawn, NJ) and Sigma Chemicals (Genesys 10 uv, Thermo Scientific, Madison, USA). Methanolic solu- Co (St. Louis, MO). Chlorogenic acid, p-coumaric acid, ferulic acid, tions of Trolox ((±)-6-hydroxi-2,5,7,8-tetramethylcromane-2-car- apigenin and cynarin standards for HPLC were acquired from Sigma boxylic acid) with different concentrations were used to achieve the Chemicals Co (St. Louis, MO). Ultrapure water (Milli-Q system, calibration curve and the antioxidant capacity was expressed in Millipore Corp, Bedford, MA) was used in all analyses. mmol Trolox equivalent (TE)/100 g dried sample.

2.3. Chemical analysis 2.7. Antioxidant capacity analysis through DPPH assay paired with QUENCHER procedure All analyses were carried out in triplicate. Total lipid content (TL) of the samples was determined according to the procedure The QUENCHER (QUick, Easy, New, CHEap and Reproducible) described by Bligh and Dyer (1959). Moisture content was deter- procedure is a recently developed procedure for measuring the mined using AOAC Official Method 930.15, ash content using AOAC antioxidant capacity of foods directly (Gokmen€ et al., 2009). It has 348 T. Claus et al. / LWT - Food Science and Technology 61 (2015) 346e351

Table 1 Proximate composition of dried artichoke.

Parts Initial moisture Moisture after Ash (g/100 g) Crude protein Total lipids Fiber (g/100 g) (g/100 g) drying (g/100 g) (g/100 g) (g/100 g)

Bracts 80.65b ± 2.12 4.49a ± 0.01 5.37b ± 0.16 10.35c ± 0.05 2.04c ± 0.03 44.23a ± 1.29 Receptacle 88.44a ± 0.30 4.63a ± 0.06 12.32a ± 0.12 24.27a ± 0.12 1.34d ± 0.02 13.07d ± 0.56 Spikes 80.98b ± 0.75 2.94c ± 0.09 7.31b ± 0.12 20.61b ± 0.03 9.17a ± 0.28 27.79c ± 1.14

Results expressed as mean ± standard deviation for analysis in three replicates. Means followed by different superscript letters (a) in same column are significantly different by Tukey's test (p < 0.05). the advantages that it can be combined with DPPH radical assay 2.9. Preparation of standards for HPLC and avoids the use of solvent extraction procedures which affect the evaluation of antioxidants in ways which are explained below Every standard used was carefully weighed, dissolved in meth- (Section 3.3). Analyses were done according to the procedure anol then submitted to sonication for 10 min. Calibration curves described by Serpen, Gokmen,€ and Fogliano (2012a). A solution of were obtained from chlorogenic acid, p-coumaric acid, ferulic acid, DPPH was obtained by dissolving 40 mg DPPH in 200 ml ethanol/ apigenin and cynarin solutions at concentrations ranging from 0.01 water mixture (1:1, ml:ml). An absorbance value of 0.75e0.80 was to 0.5 mmol/L (Fratianni et al., 2007). achieved by diluting the solution in approximately 800 ml ethanol/ water (1:1, ml:ml) mixture. About 10 mg of sample was weighed; 2.10. Quantification of phenolic compounds using HPLC bract and receptacle samples were mixed with 10.0 ml DPPH, spike samples were mixed with 50.0 ml DPPH. Mixtures were protected All HPLC experiments were done at room temperature in a Varian from light and subjected to magnetic stirring for 5, 10, 15, 30, 45 or Pro Star equipment with a Star Chromatography Workstation 60 min reaction time. After this mixtures were centrifuged at running LC control software (Varian Analytical Instruments), a 9200 g for 5 min, then the absorbance of the supernatant was Rheodyne manual injection valve (50 ml loop), a Pro Star 350 measured in a spectrophotometer (Genesys 10 uv, Thermo Scien- photodiode array detection (DAD) connected to a Polyview 2000TM tific, Madison, USA), with a wavelength of 525 nm. Methanolic program and a Microsorb C18 (250 mm 4.6 mm, particle size: 5 mm) Trolox solutions of different concentrations were used to achieve column. The method was based that of Fratianni et al. (2007). The 10 ml and 50 ml calibration curves and the antioxidant capacity mobile phase (MP) consisted of two solutions: Milli-Q water with was expressed in mmol TE/100 g of dried sample. 1 ml/100 ml acetic acid (solution A) and acetonitrile solution con- taining 1 ml/100 ml acetic acid (solution B). The MP gradient started 2.8. Total phenolic compounds with an A:B ratio of 90:10, this was adjusted to 50:50 over 30 min at a linear rate of change. Shortlyafter this the ratio was changed to 0:100 The total phenolic compound content (TPC) was determined over 5 min. The ratio was held at 0:100 for 2 min before being according to the procedure described by Shahidi and Naczk (1995). returned to 90:10 over 2 min after which it remained unchanged Extracts were prepared as 2.5 mg/ml solutions; a 250 ml aliquot of until the end of the run. Total run time was 49 min; MP flow rate was the solution was added to a tube with 250 ml Folin-Ciocalteau re- 0.8 ml/min and the detector wavelength was set at 280 nm. agent (diluted 1:1 in distilled water), 500 ml saturated Na2CO3 so- lution and 4.0 ml distilled water. These mixtures were kept allowed 2.11. Oxidation of canola oil by artichoke parts to rest for 25 min at room temperature then centrifuged at 3800 g for 10 min. Supernatant absorbance at 725 nm was measured with Canola oil oxidation tests were conducted according to the a Genesys 10 uv (Thermo Scientific, Madison, USA) spectropho- method developed by Verardo et al. (2013) using a reactor called tometer. Gallic acid solutions of different concentrations were used “Oxitest” (Velp Scientifica, Usmate, Milan, Italy), equipped with two to achieve the calibration curves, and results were expressed as mg separated oxidation chambers. The sample of interest was placed gallic acid equivalent (GAE)/100 g dried sample. in a chamber, then this system was sealed, heated to a certain

Table 2 Fatty acid quantification (mg/100 g) for dried artichoke parts.

Fatty acids Artichoke parts

Bracts Receptacle Spikes

16:0 (Palmitic acid) 229.95c ± 20.94 327.74bc ± 31.61 2303.58a ± 146.02 18:0 (Stearic acid) 13.37b ± 2.27 16.80b ± 0.91 297.20ª ± 15.08 18:1n-9 (Oleic acid) 17.80b ± 2.75 19.44b ± 1.59 99.87a ± 8.60 18:1n-7 (cis-Vaccenic acid) 6.74b ± 1.06 12.99b ± 1.72 51.84a ± 6.31 18:2n-6 (Linoleic acid) 518.92b ± 39.93 784.66b ± 71.32 3462.87a ± 189.87 18:3n-3 (alpha-Linolenic acid) 242.60b ± 19.04 151.67c ± 14.49 987.74a ± 34.42 20:0 (Arachidic acid) 7.51b ± 0.69 10.50b ± 0.82 180.52a ± 21.45 22:0 (Behenic acid) 5.66b ± 0.67 4.36b ± 0.86 52.91ª ± 5.89 20:4n-6 (Arachodonic acid) 9.00b ± 0.94 11.10b ± 1.48 127.94ª ± 8.80 20:5n-3 (Eicosa-pentaenoic acid) 20.49b ± 1.75 16.07b ± 1.21 175.60ª ± 11.32

SFA 256.49c ± 21.08 359.40bc ± 31.64 2600.78a ± 148.47 MUFA 24.54b ± 2.95 32.43b ± 2.34 151.71ª ± 10.67 PUFA 796.41b ± 44.2 963.50b ± 72.80 4754.15a ± 193.50 n-6/n-3 2.00c ± 0.35 4.74a ± 0.96 3.09b ± 0.35

Results expressed as mean ± standard deviation for analysis in three replicates. Means followed by different superscript letters (a) in same column are significantly different by Tukey's test (p < 0.05). SFA: sum of saturated fatty acids. MUFA: sum of monounsaturated fatty acids. PUFA: sum of polyunsaturated fatty acids. T. Claus et al. / LWT - Food Science and Technology 61 (2015) 346e351 349

Table 3 Total content of phenolic compounds and total antioxidant capacity of methanolic (ME), ethanolic (EE), aqueous (AE) extracts and QUENCHER for the artichoke parts.

Artichoke part TPC (mg EAG/100 g) DPPH (mmol Trolox/100 g) QUENCHER (mmol Trolox/100 g) ME EE AE ME EE AE

Bracts 706.48c ± 11.88 307.89c ± 9.46 280.76c ± 9.21 155.73c ± 0.65 129.15c ± 0.93 138.85c ± 0.74 273.52b ± 15.19 Receptacle 1637.89b ± 37.00 1452.74b ± 18.22 1310.05b ± 23.66 294.17b ± 2.23 249.09b ± 0.86 198.85b ± 1.24 356.13b ± 11.72 Spikes 4290.23a ± 51.38 3429.86a ± 60.25 2538.18a ± 37.78 2333.12a ± 17.20 1438.88a ± 10.74 1067.52a ± 3.15 8573.36a ± 254.70

Results expressed as mean ± standard deviation for analysis in three replicates. Means followed by different superscript letters (a) in same column are significantly different by Tukey's test (p < 0.05). temperature and oxygen was injected into the chamber to achieve a 3.3. Total phenolic content and antioxidant capacity pre-defined oxygen pressure. When the oxygen has been added the chamber was electronically locked and the analysis begins. Any The results of assays of TPC and antioxidant capacity assays are oxidizable compound will react with the oxygen in the chamber listed in Table 3. The significant differences between the results of thus reducing the gas pressure inside the chamber. The pressure in conventional DPPH assay and the QUENCHER procedure notwith- the chamber is monitored throughout the procedure and the in- standing it was clear that TPC was higher in spikes than in any other duction point (IP) of the sample was obtained using the two- artichoke part sampled; spikes also had the highest antioxidant tangent method. If a compound which delays sample oxidation is activity. added to the system the latency to a measurable decrease in oxygen Combining the QUENCHER procedure with the DPPH assay led pressure will be increased. The Oxitest method thus offers an effi- to greater results of antioxidant activity from artichoke parts than cient method of assessing the ability of a given compound to delay the same method executed with sample extracts. or inhibit the oxidation of a given substrate. All tests for this study Conventional methods use solvents to extract the compounds were performed at a temperature of 90 C with an initial oxygen possessing antioxidant activity from samples. A fraction of every pressure of 588 kPa, 99.9999% purity. Samples of artichoke part sample will remain insoluble regardless of the solvent chosen so (0.25 g) were added to 5 g canola oil. Oxidation of the pure oil extraction cannot be complete; some of the compounds of interest (control) and of canola oil plus 0.001 g BHT (reference) were also may be retained in the insoluble fraction, leading to underestima- assessed. It is very important to note that the percentage BHT used tion of antioxidant activity (Serpen et al., 2012) by the QUENCHER in the reference test is the maximum allowed for use in foods by the procedure. Estimates of antioxidant activity based on extraction are Food and Drug Administration (FDA). Each experiment was usually less accurate when the substance contains amino acids, repeated three times. fibers or uronic acids (Perez-Jim enez & Saura-Calixto, 2006).

2.12. Statistical analysis 3.4. Capacity of artichoke parts to delay oxidation of canola oil

Data were subjected to ANOVA and means were compared using The Oxitest method proved a very effective method of the Tukey test conducted using Statistica, version 7.0. Correlation comparing oxidation of canola oil with and without addition of coefficients (R) were calculated to determine the relationship be- artichoke parts. Data from tests in the Oxitest reactor are given in tween the obtained results using Microsoft Office Excel, version Table 4. Canola oil without additives starts to oxidize in 2007. Principal components analysis (PCA) was carried out using 892 ± 7.0 min; adding 0.001 g BHT delays oxidation by 347 min MATLAB 2007. (38.90% increase in oxidation latency). Bracts and receptacle were less effective than BHT in delaying oxidation (oxidation latency was 3. Results and discussion increased by 15.58% and 24.10% respectively), but adding 0.25 g artichoke spikes delayed oxidation by 649 (a 72.75% increase in 3.1. Yield and chemical analysis oxidation latency compared to the control). This is a very impres- sive result; artichoke spikes show superior antioxidant activity to Processing the artichoke into parts yielded the following pro- BHT and they are from a natural source e unused byproducts of the portions by mass for the various parts, receptacle: 22.70 g/100 g; food industry e unlike BHT. bracts: 56.14 g/100 g and spikes: 21.16 g/100 g. Only the receptacle is edible, so all the other parts (about 77% by mass of the whole 3.5. Correlation analysis artichoke) are discarded by the food industry, generating a large amount of residue. The calculation of correlation coefficients (R) is useful for con- The results of the chemical analyses of moisture content, ash firming the strength of predicted relationships. In this work we used content, TL and crude fiber content are presented in Table 1. Initial moisture values of artichoke were high, ranging from 80.65 g/100 g (bracts) to 88.44 g/100 g (receptacle), however after the drying Table 4 Induction points of the oxidation experiments with canola oil. process moisture content of all parts was less than 4.7 g/100 g. Canola oil Induction point (minutes) 3.2. Fatty acid composition Without antioxidants 892.0e ± 7.0 b ± Table 2 shows the ten main FAs which were detected in dried With 0.02 g/100 g of BHT 1239.0 12.5 With 5 g/100 g of artichoke bracts 1031.0d ± 6.3 artichoke parts, four SFAs, two MUFAs and four PUFAs (two n-6; With 5 g/100 g of artichoke receptacle 1107.0c ± 9.6 two n-3). The main compounds in the SFA, MUFA and PUFA groups With 5 g/100 g of artichoke spikes 1541.0a ± 9.3 were palmitic (16:0), oleic (18:1 n-9) and linoleic (18:2 n-6) acids Results expressed as mean ± standard deviation for analysis in three replicates. fi a respectively. Alpha-linolenic acid (18:3 n-3) was also a signi cant Means followed by different superscript letters ( ) in same column are significantly proportion of PUFA content. different by Tukey's test (p < 0.05). 350 T. Claus et al. / LWT - Food Science and Technology 61 (2015) 346e351

Table 5 Quantification of phenolic compounds from methanolic extracts of dried artichoke parts.

Compound Retention Bracts (mg/100 g) Receptacle (mg/100 g) Spikes (mg/100 g) time (minutes)

Chlorogenic acid 10.50 706.40c ± 18.24 2494.41b ± 99.76 5011.10a ± 150.33 Cynarin 11.70 nd 31.35a ± 0.93 nd p-Coumaric acid 16.98 141.35b ± 4.23 388.70b ± 12.45 1447.59a ± 69.97 Ferulic acid 18.04 440.56c ± 14.08 1515.99a ± 75.40 1180.39b ± 55.77 Apigenin 27.61 nd 8.19b ± 0.24 1206.68a ± 48.24

Results expressed as mean ± standard deviation for analysis in three replicates. Means followed by different superscript letters (a) in same line are significantly different by Tukey's test (p < 0.05). the correlation coefficients to check the correlations between the standards for UVevisible light spectra. Quantification of com- TPC assays and oxidation experiments. Correlations between TPC pounds was based on comparison of sample calibration curves with and DPPH, QUENCHER and Oxitest estimates of antioxidant activity those of standards. Table 5 shows the results of the HPLC analyses. were R ¼ 0.798, R ¼ 0.642 and R ¼ 0.744 respectively. The corre- Ferulic and p-coumaric acids were detected in the samples, but lations between DPPH assay results and QUENCHER and Oxitest the main phenolic compound detected in every artichoke part was estimates of antioxidant activity were R ¼ 0.822 and R ¼ 0.830 chlorogenic acid. Spikes had the highest chlorogenic acid content, respectively. Finally, the correlation between QUENCHER and about 5000 mg/100 g dried sample, spikes also had an apigenin Oxitest was R ¼ 0.981. The results of the Oxitest procedure were content of 1200 mg/100 g. most highly correlated with those achieved by combining the DPPH The cynarin content of the receptacle was low (31.35 mg/100 g) and assay with the QUENCHER procedure (R ¼ 0.981), correlations with the cynarin content of bracts and spikes was below the quantification conventional DPPH and assay of TPC were lower (R ¼ 0.830 and threshold. These results were probably due to limited interaction R ¼ 0.744, respectively). We had expected this pattern of results as between the extraction solvent and cynarin and the instability of this both methods (QUENCHER and Oxitest) are directly applied to phenolic compound which degrades into more stable isomers samples, thereby increasing the proportion of molecules with (Carvalho, Gosmann, & Schenkel, 2004; Pandino et al., 2011). antioxidant capacity which are available for reaction with the analysis medium. The lowest correlation coefficient values were 3.7. PCA analysis those related the TPC assay to other estimates of antioxidant ac- tivity, probably because TPC assay only quantifies some antioxidant PCA was used to provide an overview of the relationships among activity; it is known that compounds other than phenols, including the artichoke parts and investigate the influence of the chemical certain vitamins, carotenoids, tocopherols and other compounds characteristics of the components on the discrimination of data. We can act as antioxidants (Serpen et al., 2012a, 2012b). used a 3 14 data set. The three artichoke parts investigated (bracts, receptacle and spikes) constituted the rows of the matrix; 3.6. HPLC analysis the columns consisted of the mean content for each variable investigated, i.e. TPC and DPPH of methanolic, ethanolic and TPC and antioxidant capacity were greatest in the methanolic aqueous extracts, chlorogenic acid, cynarin, p-coumaric acid, ferulic extracts, indicating that methanol was the best solvent for extrac- acid, apigenin, PUFA, QUENCHER and Oxitest analysis. tion of antioxidant compounds; we therefore used methanolic ex- PCA decomposes the data into separate sets of scores and loadings tracts in the HPLC analysis. The main phenolic compounds in each for the samples and variables, and the whole data variability is artichoke part were identified by chromatographic analysis and explained in order to provide a clear and more interpretable visu- comparison of sample retention times with those of analytical alization of data structure in a reduced dimension. The variables

Fig. 1. Biplot graph from results of scores and loadings. Triangles refer to result type (TPC and DPPH of methanolics, ethanolics and aqueous extracts, chlorogenic acid, cynarin, p-coumaric acid, ferulic acid, apigenin, AGPI, QUENCHER and Oxitest), while asterisks refer to sample type (bracts, receptacle and spikes). Data regarding TPC and DPPH results were circled in order to demonstrate a grouping of the different extracts, while QUENCHER and Oxitest were circled to emphasize a greater distance from central point (intersection between dotted lines). T. Claus et al. / LWT - Food Science and Technology 61 (2015) 346e351 351 investigated were measured on different scales so we chose to use an Dolatabadi, J. E. N., & Kashanian, S. (2010). A review on DNA interaction with e autoscaled data matrix which allowed the variance of all variables to synthetic phenolic food additives. Food Research International, 43, 1223 1230. Ferracane, R., Pellegrini, N., Visconti, A., Graziani, G., Chiavaro, E., Miglio, C., et al. be identical in the first instance. (2008). Effects of different cooking methods on antioxidant profile, antioxidant The two principal components explained 91.71% of all variance capacity, and physical characteristics of artichoke. Journal of Agricultural and e in the data. The entire data set was visualized using a biplot Food Chemistry, 56, 8601 8608. Focke, W. W., Westhuizen, I. V. D., Grobler, L., Nshoane, K. T., Reddy, J. K., & Luyt, A. S. (combined scores and loadings plot for two components) which is (2012). The effect of synthetic antioxidants on the oxidative stability of bio- shown in Fig. 1. diesel. Fuel, 94,227e233. The spikes (*17) differ from bracts (*15) and receptacle (*16) by Fratianni, F., Tucci, M., De Palma, M., Pepe, R., & Nazzaro, F. (2007). Polyphenolic composition in different parts of some cultivars of globe artichoke (Cynara presenting opposing positions on the axis of PC2. The results of the cardunculus L. var. scolymus (L.) Fiori). Food Chemistry, 104, 1282e1286. DPPH and TPC assays were grouped around the central Cartesian Gokmen,€ V., Serpen, A., & Fogliano, V. (2009). Direct measurement of the total point indicating that they were not influential components. How- antioxidant capacity of foods: the ‘QUENCHER’ approach. Trends in Food Science fi and Technology, 20,278e288. ever, QUENCHER procedure showed a signi cant correlation with Huang, S. S. Y., Fu, C. H. L., Higgs, D. A., Balfry, S. K., Schulte, P. M., & Brauner, C. J. spikes because point *17 is the closest to point 13. This pattern of (2008). Effects of dietary canola oil level on growth performance, fatty acid results indicates that the QUENCHER procedure is valuable for composition and ionoregulatory development of spring chinook salmon parr e evaluation of artichoke parts as the radical used (DPPH) reacts with (Oncorhynchus tshawytscha). Aquaculture, 274,109 117. Impellizzeri, D., Esposito, E., Mazzon, E., Paterniti, I., Di Paola, R., Bramanti, P., et al. more antioxidant compounds than the common DPPH assay which (2012). The effects of a polyphenol present in olive oil, oleuropein aglycone, in relies on extraction. an experimental model of spinal cord injury in mice. Biochemistry Pharma- cology, 83,1413e1426. Instituto Adolfo Lutz. (2008). Normas Analíticas do Instituto Adolfo Lutz. Metodos 4. Conclusion químicos e físicos para analise de alimentos (4th ed.). Sao Paulo: Digital. Jacociunas, L. V., Andrade, H. H. R., Lehmann, M., Abreu, B. R. R., Ferraz, A. B. F., Silva, J., et al. (2013). Artichoke induces genetic toxicity in the cytokinesis-block As in other countries the only part of artichoke cultivated in micronucleus (CBMN) cytome assay. Food and Chemical Toxicology, 55,56e59. Brazil (Cynara cardunculus var. scolymus (L.) Fiori) which is edible is Joseph, J. D., & Ackman, R. G. (1992). Capillary column gas-chromatographic method the receptacle. Knowledge about the artichoke receptacle increased for analysis of encapsulated fish oils and fish oil ethyl-esters e collaborative study. Journal of AOAC International, 75, 488e506. following evaluation of their proximate and FA composition. All Lattanzio, V., Kroon, P. A., Linsalata, V., & Cardinali, A. (2009). Globe artichoke: a parts of the artichoke have antioxidant activity, as the TPC and functional food and source of nutraceutical ingredients. Journal of Functional DPPH/QUENCHER assays in this study have demonstrated. These Foods, 1,131e144. assays indicated that TPC and antioxidant activity are concentrated LeClercq, C., Arcella, D., & Turrini, A. (2000). Estimates of the theoretical maximum daily intake of erythorbic acid, gallates, butylated hydroxyanisole (BHA) and in the spikes of artichokes. This finding was corroborated by HPLC, butylated hydroxytoluene (BHT) in Italy: a stepwise approach. Food and in which the greatest concentrations of chlorogenic, ferulic and p- Chemical and Toxicology, 38,1075e1084. € coumaric acids and apigenin were detected in the spikes. Lombardo, S., Pandino, G., Mauromicale, G., Knodler, M., Carle, R., & Schieber, A. (2010). Influence of genotype, harvest time and plant part on polyphenolic Unlike the other artichoke parts spikes showed a remarkable composition of globe artichoke [Cynara cardunculus L. var. scolymus (L.) Fiori]. capacity to delay oxidation of canola oil, producing results superior Food Chemistry, 119,1175e1181. to that of BHT added at 0.02 g/100 g. This effect can be attributed to Lutz, M., Henríquez, C., & Escobar, M. (2011). Chemical composition and antioxidant properties of mature and baby artichokes (Cynara scolymus L.), raw and cooked. the overall antioxidant activity of spikes, indicated by the strong Journal of Food Composition and Analysis, 24,49e54. correlation between the Oxitest results and the results of DPPH Ma, X., Wu, H., Liu, L., Yao, Q., Wang, S., Zhan, R., et al. (2011). Polyphenolic compounds radical capture assay combined with QUENCHER procedure and antioxidant properties in mango fruits. Scientia Horticulturae, 129,102e107. ¼ Moure, A., Cruz, J. M., Franco, D., Dominguez, J. M., Sineiro, J., Dominguez, H., et al. (R 0.9814). Artichoke spikes are therefore strong candidates for (2001). Natural antioxidants from residual sources. Food Chemistry, 72,145e171. use in the food industry as natural additives against oxidation of Noldin, V. F., Filho, V. C., Monache, F. D., Benassi, J. C., Christmann, I. L., Pedrosa, R. C., foods such as canola oil although assessments of taste modification et al. (2003). Composiçao~ química e atividades biologicas das folhas de Cynara scolymus L. (Alcachofra) cultivada no Brasil. Química Nova, 26,331e334. or product allergenicity are required to evaluate the food safety of Pandino, G., Lombardo, S., Mauromicale, G., & Williamson, G. (2011). Profile of such a product. polyphenols and phenolic acids in bracts and receptacles of globe artichoke (Cynara cardunculus var. scolymus) germplasm. Journal of Food Composition and Analysis, 24,148e153. Acknowledgments Perez-Jim enez, J., & Saura-Calixto, F. (2006). Effect of solvent and certain food constituents on different antioxidant capacity assays. Food Research Interna- The authors would like to thank CAPES for financial support and tional, 39,791e800. fi e Ru no, M. S. M., Alves, R. E., Fernandes, F. A. N., & Brito, E. S. (2011). Free radical Cocamar Cooperativa Agroindustrial de Maringa, for the donation scavenging behavior of ten exotic tropical fruits extracts. Food Research Inter- of canola oil. national, 44, 2072e2075. Saez, V., Fasoli, E., D'Amato, A., Simo-Alfonso, E., & Righetti, P. G. (2013). Artichoke and cynar liqueur: two (not quite) entangled proteomes. Biochimica et Bio- References physica Acta, 1834,119e126. Serpen, A., Gokmen,€ V., & Fogliano, V. (2012a). Solvent effects on total antioxidant Arabshahi-Delouee, S., & Urooj, A. (2007). Antioxidant properties of various solvent capacity of foods measured by direct QUENCHER procedure. Journal of Food extracts of mulberry (Morus indica L.) leaves. Food Chemistry, 102, 1233e1240. Composition and Analysis, 26,52e57. € Association of Official Analytical Chemists. (2000). Official methods of analysis of the Serpen, A., Gokmen, V., & Fogliano, V. (2012b). Total antioxidant capacities of raw association of official analytical chemists. Gaithersburg (17th ed.). and cooked meats. Meat Science, 90,60e65. Biglari, F., Alkarkhi, A. F. M., & Easa, A. M. (2008). Antioxidant activity and phenolic Shahidi, F., & Naczk, M. (1995). Food phenolics: Sources, chemistry, effects and content of various date palm (Phoenix dactylifera) fruits from Iran. Food Chem- applications. Lancaster: Technomic Publishing Company. istry, 107, 1636e1641. Shen, Q., Dai, Z., & Lu, Y. (2010). Rapid determination of caffeoylquinic acid Bligh, E. G., & Dyer, W. J. (1959). A rapid method of total lipid extraction and derivatives in Cynara scolymus L. by ultra-fast liquid chromatography/tandem purification. Canadian Biochemistry Physiology, 37,911e917. mass spectrometry based on a fused core C18 column. Journal of Separation Botterweck, A. A. M., Verhagen, H., Goldbohm, R. A., Kleinjans, J., & Van Den Science, 33,3152e3158. Brandt, P. A. (2000). Intake of butylated hydroxyanisole and butylated Sonnante, G., De Paolis, A., Lattanzio, V., & Perrino, P. (2002). Genetic variation in hydroxytoluene and stomach cancer risk: results from analyses in the wild and cultivated artichoke revealed by RAPD markers. Genetic Resources and Netherlands Cohort study. Food and Chemical Toxicology, 38, 599e605. Crop Evolution, 49,247e252. Brand-Williams, W., Cuvelier, M. E., & Berset, C. (1995). Use of a free radical method Vasco, C., Ruales, J., & Kamal-Eldin, A. (2008). Total phenolic compounds and anti- to evaluate antioxidant activity. LWT e Food Science and Technology, 28,25e30. oxidant capacities of major fruits from ecuador. Food Chemistry, 111,816e823. Carvalho, J. C. T., Gosmann, G., & Schenkel, E. P. (2004). Farmacognosia: da planta ao Verardo, V., Riciputi, Y., Sorrenti, G., Ornaghi, P., Marangoni, B., & Caboni, M. F. medicamento (5th ed.). Brazil: Porto Alegre/Florianopolis . (2013). Effect of nitrogen fertilisation rates on the content of fatty acids, sterols, Di Giulio, G. (2004). Falta pesquisa para aumentar produçao~ de alcachofra no Brasil. tocopherols and phenolic compounds, and on the oxidative stability of walnuts. Ci^encia e Cultura, 56,13e14. LWT e Food Science and Technology, 50,732e738.