Cent. Eur. J. Biol. • 7(1) • 2012 • 146-158 DOI: 10.2478/s11535-011-0095-8

Central European Journal of Biology

Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

Research Article

Roberto Lo Scalzo1,*, Massimo Morassut2, Paolo Rapisarda3

1National Council for Agricultural Research (C.R.A.), Research Unit of Food Technology (CRA-IAA), 20133 Milano, 2National Council for Agricultural Research (C.R.A.), Research Unit for Enology of Central Italy (CRA-ENC), 00049 Velletri, Italy 3National Council for Agricultural Research (C.R.A.), Research Center of Citriculture and Mediterranean Crops (CRA-ACM), 95024 Acireale, Italy Received 12 May 2011; Accepted 19 October 2011

Abstract: The aim of the present study was the evaluation of the antioxidant content in phenolic and non-phenolic extracts of ten wine samples, trying to elucidate the potential role of unusual antioxidant compounds. Samples of wines processed from red and white (

L.), deprived of the volatile fraction at low temperature and buffered at physiological pH, were fractionated by C18 into two fractions: FR1 and FR2. Non-phenolics, such as tartaric, malic, lactic, and succinic acids; glucose; fructose; and glycerin were mainly found in FR1, while polyphenols were present exclusively in FR2. Peroxyl radical quenching was assayed by the ORAC method, while superoxide and hydroxyl radical scavenging activity were assayed by electron paramagnetic resonance. In the ORAC and superoxide assays, most of the activity was found in FR2, while in hydroxyl radical assay, the activity was found in FR1. Model solutions were used to attribute a role to the single compounds in the evaluation of wine’s ROS scavenging capacity: the ORAC and superoxide anion scavenging effects were mainly attributed to the polyphenols, averaging 94.8%, with some contribution from glycerin, particularly in white wines. Unexpectedly, the main chemical responsible for hydroxyl radical scavenging activity was glycerin (56.1%), with the polyphenols scavenging at 18.1%.

Keywords: Wine • Phenolic compounds • Antioxidant • C18 fractionation • Peroxyl radical • Superoxide anion • Hydroxyl radical © Versita Sp. z o.o.

1. Introduction Wine is known for its healthy properties, if appropriately assumed. These properties are strictly Wine represents an application of a traditional related to polyphenols, known as antioxidants, able to biotechnology known for thousands of years. It is now a scavenge free radicals produced by oxygen catabolism very important worldwide aspect of food culture, having or fatty acid peroxidation [2,3]. spread from the Mediterranean basin countries to other Many studies have been made to identify the active emerging producers, such as the United States, Chile, compounds of wine antioxidant action, both on whole Australia and South Africa. extracts and separately on its polyphenol classes, such The chemical composition of the grapes as anthocyanins, flavones, stilbenes, cinnamates and is subjected to deep changes during the phenolic acids [4-11]. biotechnological processing into wine [1]. The wine The common opinion is that the free radical scavenging composition profile is obtained by the progressive capacity of and wine extracts is related to its alcoholic extraction of grape components, usually polyphenol content, mainly located in the grape skin [12-16]. with maceration for red-skinned and not for white- It has been found that phenols show different skinned grape varieties. rankings of reactivity with different antioxidant assays

* E-mail: [email protected] 146 R. Lo Scalzo et al.

[17-19]. In addition, some studies on the antioxidant 2. Experimental Procedures potential of grape extracts found that the use of different tests gave different responses, not directly correlated 2.1 Rationale of the experiment with the polyphenolic content of wine. For example, the Different red and white wines made from ten grape activity of grape components against hydroxyl radicals, varieties (Vitis vinifera L.) were evaporated to subtract one of the most harmful free radicals generated as an the volatile and ethanolic fraction and re-solubilized oxygen byproduct, showed that the activity of grape in phosphate buffer to give the unfractionated (UN) extracts was poorly related to polyphenol content. samples. An aliquot of UN was fractionated on C18, A previous work by Lo Scalzo et al. [20] demonstrated obtaining two fractions: FR1, the directly eluted fraction that the comparison of two different antioxidant assays, deprived of polyphenols, containing the non-phenolic such as DPPH and hydroxyl radical scavenging, resulted hydrophilic compounds, mainly hydroxy acids, glycerin in a strong difference between grape skin and flesh and residual simple sugars; and FR2, the retained extracts. The scavenging activity of DPPH was higher fraction eluted with MeOH, with practically all the in grape skin than flesh, as was expected in accordance polyphenols present in the original wine sample. with their significantly different phenol content. The UN, FR1 and FR2 samples were analyzed and activity against hydroxyl radicals did not give the same compared for their content in hydroxy acids such as result: the hydroxyl radical scavenging was about the tartaric, malic, lactic and succinic; for glycerin, glucose same between the grape skin and flesh extracts. The and fructose; and for total polyphenols index (Table 1). experiment was repeated on table grape samples [21], Three free radical scavenging tests against the by comparing three antioxidant assays: DPPH, Fremy’s main Reactive Oxygen Species (ROS) on the different salt, and hydroxyl radical scavenging on grape skin and fractions of red and white wines were also performed. flesh extracts. The results were confirmed as previously The scavenging of peroxyl radicals was evaluated with described, with the indices of DPPH and Fremy’s salt the fluorescein quenching by a peroxyl radical generated scavenging yielding higher results in skin than flesh, by thermal degradation of an azo-bis-compound, with no difference for hydroxyl radical scavenging. well known as the ORAC test (Oxygen Radical Furthermore, the grape extracts were fractionated by Absorbance Capacity). Superoxide anion (SASC) and solid phase extraction on C18, and the result was that hydroxyl radical scavenging (HRSC) were measured the activity against DPPH and Fremy’s salt was found by 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) spin- in the polyphenol-enriched fraction, while most of the trap reaction and electronic paramagnetic resonance activity against hydroxyl radical was in the fractions detection, respectively. poor in polyphenols. These results were confirmed unexpectedly in vivo, where experimental rats were 2.2 Winemaking and sampling divided into three groups, two separately fed with grape Wine samples were obtained from the experimental wine skin or flesh, and the third used as a control, fed without cellar of the CRA-ENC, Research Unit for the Enology grape supplementation. The interesting result was that of Central Italy, Velletri (Rome). 10 wine samples were the two groups separately fed with grape skin and flesh obtained from the winemaking of single certified grape were equally cardioprotected with respect to the control varieties (red: 1 Abbuoto, 2 Bombino nero, 3 Marzemino, group [22]. 4 , 5 Petit verdot, 6 Refosco dal peduncolo At this point, the identification of the compounds rosso, 7 Teroldego; white: 8 Bombino bianco, 9 Greco, responsible for the antioxidant activity in wine grapes 10 Malvasia del ), harvested in the experimental has become the most relevant question, starting with vineyard of the CRA-ENC during the 2007 season, and the most abundant grape compounds, such as simple processed according to established protocols. Red grapes sugars and organic acids. Previous experiments (about 100 kg) were destemmed and crushed. The must [23,24] demonstrated that •OH could be scavenged by was treated with 6 g/hL of potassium metabisulfite and simple sugars, with disaccharides more active than inoculated with 20 g/hL of Saccharomyces cerevisiae r.f. monosaccharides [25], and also that organic acids can uvarum strain S6u (LALLEMAND Inc 151 Skyway Avenue act as antioxidants [26,27]. Rexdale, Ontario M9W 4Z5). Fermentation temperature Many types of grape are destined to become wine was 25°C, the maceration time of skins was 10 days. and the logical continuation of these studies is to shift White grapes (about 100 kg) were destemmed, crushed, previous considerations of grapes to wine, so the aim and 10 g/hL of potassium metabisulfite was added. After of the present study was to investigate the role of clarification, musts were racked and inoculated with polyphenols in the antioxidant capacities of different 25 g/hL of of Saccharomyces cerevisiae r.f. uvarum, wine samples before and after fractionation. strain S6u. Fermentation took place in 100-L stainless

147 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines and eluted with 18 succinic acid (mg/100 ml) total polyphenols (mg/100 ml)

FR1 FR2 UN FR1 FR2 FR1 FR2 UN FR1 FR2 lactic acid (mg/100 ml) glycerin (g/l) FR1 FR2 UN FR1 FR2 UN fructose (mg/100 ml) malic acid (mg/100 ml) FR1 FR2 UN FR1 FR2 UN glucose (mg/100 ml) tartaric acid (mg/100 ml) UN UN 9.7 a 9.3 a 0.0 b 102.0 a 91.9 a 0.0 b 12.6 a 15.6 a 0.0 b 260.5 a 12.8 b 277.1 a 29.5 a 27.1 a 0.0 b 37.4 a 36.5 a 0.0 b 11.5 a 11.5 a 0.0 b 99.4 a 7.6 b 86.8 a 18.2 a 15.4 a 0.0 b 44.1 a 43.7 a 0.0 b 13.0 a 13.8 a 0.0 b 323.6 a 16.0 c 271.8 b 12.1 a 13.3 a 0.0 b 60.1 a 59.6 a 0.0 b 17.5 a 21.0 a 0.0 b 120.8 a 17.2 b 129.3 a 49.6 a 44.4 a 0.0 b 94.4 a 90.1 a 0.0 b 10.4 a 10.6 a 0.0 b 31.0 a 4.6 c 19.0 b 73.5 a 71.9 a 0.0 b 102.6 a 103.0 a 0.0 b 9.1 a 9.4 a 0.0 b 24.6 a 2.9 b 20.8 a 198.9 a 205.6 a 0.2 b 66.1 a 63.9 a 0.0 b 41.9 a 42.3 a 0.0 b 28.3 a 26.1 a 0.0 b 201.0 a 196.8 a 3.1 b 20.9 a 16.3 a 0.0 b 181.4 a 179.3 a 0.0 b 36.6 a 27.9 a 0.0 b 444.7 a 436.1 a 0.1 b 175.4 a 174.9 a 0.0 b 8.1 a 7.4 a 0.0 b 40.5 a 35.9 a 0.0 b 499.1 a 512.7 a 1.7 b 73.4 a 79.9 a 0.0 b 5.2 a 4.8 a 0.0 b 16.8 a 16.9 a 0.0 b

between each UN, FR1 and FR2 value. MeOH. Different letters are for significant differences (P<0.05) between each UN, FR1 Average values of the main chemicals ten samples wine (1-7 red, 8-10 white, n=30). UN, sample untreated by fractionation; FR1, directly eluted fraction; FR2, retained C Average Samples Samples average red 178.6 a 178.0 a 1.2 b 45.2 a 41.4 a 0.0 b 123.9 a 125.4 a 0.0 b 24.4 a 22.4 a 0.0 b average red 20.1 a 18.6 a 0.0 b 59.2 a 57.5 a 0.0 b 12.6 a 14.3 a 0.0 b 184.7 a 12.4 b 173.9 a average white 402.4a 406.4a 0.7 b 118.5 a 121.2 a 0.0 b 5.9 a 5.7 a 0.0 b 27.9 a 25.4 a 0.0 b average white 57.9 a 54.7 a 0.0 b 88.3 a 86.8 a 0.0 b 10.3 a 10.4 a 0.0 b 26.4 a 3.9 b 20.3 a 1 Abbuoto

2 Bomb nero 258.9 a 257.7 a 2.1 b 148.0 a 145.0 a 0.0 b 13.2 a 12.5 a 0.0 b 20.2 a 17.7 a 0.0 b 3 Marzemino 187.6 a 190.4 a 0.8 b 22.5 a 18.2 a 0.0 b 179.8 a 184.9 a 0.0 b 16.1 a 16.6 a 0.0 b 4 Montepulciano 84.5 a 81.7 a 0.3 b 24.8 a 21.6 a 0.0 b 207.5 a 207.1 a 0.0 b 22.2 a 21.7 a 0.0 b 5 verdot Pet 133.8 a 124.3 a 1.2 b 20.2 a 16.4 a 0.0 b 219.7 a 233.0 a 0.0 b 21.7 a 20.3 a 0.0 b 6 Refosco 7 Teroldego 185.8 a 189.1 a 0.9 b 13.9 a 8.4 a 0.0 b 23.9 a 18.9 a 0.0 b 25.9 a 26.5 a 0.0 b 8 Bomb bianco 263.4 a 270.4 a 0.2 b 106.8 a 108.8 a 0.0 b 4.5 a 4.8 a 0.0 b 26.3 a 23.3 a 0.0 b 9 Greco 1 Abbuoto

2 Bomb nero 21.9 a 20.4 a 0.0 b 56.4 a 58.1 a 0.0 b 11.5 a 11.9 a 0.0 b 120.6 a 17.7 c 101.5 b 3 Marzemino 39.6 a 35.1 a 0.0 b 52.5 a 49.6 a 0.0 b 11.6 a 14.3 a 0.0 b 168.5 a 8.1 b 168.0 a 4 Montepulciano 9.6 a 9.7 a 0.0 b 62.1 a 63.4 a 0.0 b 10.2 a 11.8 a 0.0 b 199.6 a 7.1 b 182.4 a 5 verdot Pet 6 Refosco 7 Teroldego 8 Bomb bianco 50.7 a 47.9 a 0.0 b 68.0 a 67.2 a 0.0 b 11.3 a 11.3 a 0.0 b 23.7 a 4.2 b 21.2 a 9 Greco 10 Malvasia 10 Malvasia Table 1. Table

148 R. Lo Scalzo et al.

steel tanks at 18°C. At the end of alcoholic fermentation 2.5 Hydroxy acids and glycerin (residual sugars <3.0 g/L), wines were racked with 6 g/hL Hydroxy acids, glycerin, glucose and fructose content of of potassium metabisulfite, cold stabilized for two months wine extracts were measured by HPLC with a JASCO at 10°C, bottled and kept in a wine cellar at 14-15°C for system equipped with a diode array and refractive 6 months. index detectors (MD-910 JASCO, RI-930). The pump Three bottles of wine were randomly selected for the (PU-980 JASCO) was coupled with a ternary gradient analysis: two aliquots of about 100 mL were taken from unit (LG-1580-02 JASCO). The analytical data were each bottle. 10 mL of each aliquot was evaporated to evaluated using a software-management system dryness in a centrifugal evaporator at 2-4°C temperature. of chromatographic data (ChromNAV, Jasco). The The residue from one aliquot was buffered to the original methods proposed and validated by Caccamo et al., volume with cold (2-4°C) phosphate buffer 0.1 M, pH 7.4 [30] and by Lòpez-Tamames et al., [31], were followed

(PBS) and successively 4-fold diluted with cold H2O to with slight modifications. obtain the unfractionated (UN) samples, which were then For hydroxy acids, the separation was carried out centrifuged in Eppendorff tubes (20000×g, 5 minutes) on an Inertsil ODS-3 column of 5 μm particle diameter and supernatants were stored at –80°C before analysis. and 0.46×25 cm dimension. The elution was carried out

Samples from the second aliquot remained undiluted, to at 30°C with H3PO4 0.02 M in water as mobile phase at be used for the fractionation. 0.6 mL/min. The detection was spectrophotometrically

The evaporation in the vacuum completely eliminated made at 214 nm. Samples (1 mL) were filtered on C18 the sulphur dioxide, both free and combined as checked by support (ICN Silica RP 18, 32-63 μm, 5 mm height, 9 mm titration with iodine in presence of a starch-water indicator, internal diameter) and 5-fold diluted with the mobile following the rapid method of Ripper-Schmitt [28]. phase before injection (20 μl). Tartaric, malic, lactic and succinic acid solutions were used as external standards 2.3 Fractionation (retention times 6.4, 7.7, 9.8 and 16.7 min, respectively), 1 mL of undiluted buffered wine sample was placed on a and the results were expressed as mg/100 mL. The

C18 column (ICN Silica RP 18, 32-63 μm, 25 mm height, analyses were repeated twice. 9 mm internal diameter), previously conditioned with For glucose, fructose and glycerin, separation was PBS 0.025 M, and eluted at 0.5 mL/min. The column was done with a CarboSep Coregel 87C carbohydrate three times rinsed with 1 mL of H2O, the rinsing solvent column with a 0.78×30 cm bed packed with a cation- was pooled to the eluate and constituted the samples exchange resin in the Ca2+ ionic form. The mobile phase named FR1. Successively, the column was eluted with was water at 0.7 mL/min, the elution was performed 1 mL of MeOH for 4 times, the eluates pooled and at 85°C, and the signals were revealed by a refractive evaporated to dryness, taken to a 4 mL volume with cold index detector. Samples were 20-fold diluted with PBS 0.025 M, pH 7.4 to constitute the samples named mobile phase before injection (10 μl). Glucose, fructose FR2. Fractions, exactly diluted as UN samples, were and glycerin solutions were used as external standards centrifuged as before and stored at –80°C until analysis. (retention times 10.1, 12.4 and 16.8 min) and the results were expressed as mg/100 mL for glucose and fructose 2.4 Total polyphenols index and as g/L for glycerin. The analyses were repeated The UN, FR1 and FR2 wine extracts were assayed for twice. polyphenols by the Folin-Ciocalteu reaction, according The analyte concentrations were calculated from the to the protocol proposed by Benherlal and Arumughan experimental peak areas by analytical interpolation in a [29], with some modifications: 0.1 mL of each sample standard calibration curve for each compound. was diluted with 2 mL water in a 5 mL flask, then was treated with 0.3 mL of Folin-Ciocalteu reagent and mixed 2.6 Peroxyl radical test (ORAC) well. After 5 min at room temperature, the mixture was The method was performed as described by Ou et al., treated with 0.6 mL of Na2CO3 20% solution in water, [32] with some modifications. The measurements were kept to a final volume of 5 mL, and placed for 1 hr in the carried out on a Wallac 1420 Victor II 96 well plate reader dark. After the complete reaction time, the absorbance (EG & Wallac, Turku, Finland) with a fluorescence filter values of the solutions were read at 730 nm in 1 cm (excitation 485 nm, emission 535 nm). Fluorescein cuvettes against a blank of the same composition as the (116 nM) was the target molecule for free radical attack wine’s extraction solution. from 2,2’-Azobis-2-Amidinopropane (AAPH) (153 mM) Results were expressed as mg/100 mL, after as the peroxyl radical generator. The reaction was calibration with solutions of gallic acid treated in the same conducted at 37°C with Trolox (1 µM) as the control way as the samples. The analyses were repeated twice. standard and phosphate buffer (pH 7.0) as the blank.

149 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

All solutions were freshly prepared prior to analysis. modulation amplitude, 1000 mG; microwave attenuation All samples were diluted with phosphate buffer (1:25- 5 dB; receiver gain, 8×102. The resultant 5.6 mM DMPO- 100, v/v) and results were reported as an amount of •OOH was registered after exactly 1 minute reaction at Trolox (6-Hydroxy-2,5,7,8-tetramethylchromane-2- 25°C, obtaining a spectrum consisting of a quadruplet, carboxylic acid), an analogue of vitamin E, after data with 1:1.1.1 relative intensities, whose spectral interpolation with a calibration made with model solutions constants were close to previous literature findings a b at known concentration. Hence, the scavenging (g=2.00616, an=1.38 mT, ah =1.11 mT, ah =0.08 mT). activity was expressed as mM Trolox equivalents. The For the measurement, the reference solution without correlation data (rxy) were calculated by simple linear the scavenging compound contained: 0.3 mL of 50 mM regression. DMPO in PBS 0.1 M; 0.2 mL of PBS 0.02M; 0.3 mL of

15 mM KO2 in DMSO. 2.7 EPR scavenging tests The solution with the test scavenging compound . Ferrous sulphate heptahydrate (FeSO4 7H2O) was contained: 0.3 mL of of 50 mM DMPO in PBS 0.1 M; 0.2 mL

purchased from Fluka Chemie AG (Switzerland). All of wine extract 20-fold diluted; 0.3 mL of 15 mM KO2 in other chemicals used were from Sigma-Aldrich (US). For DMSO. Each assay was repeated three times. each dissolution or dilution, phosphate buffer solutions (PBS) and all other solutions were previously deprived 2.7.2 •OH test (HRSC)

of O2 by bubbling with pure N2. Active oxygen species, such as hydroxyl radicals The methods for the scavenging evaluation of (•OH), are formed through a one–electron reduction of • - • • superoxide anion ( O2 , SASC) and hydroxyl ( OH, HRSC) hydrogen peroxide (H2O2). The OH is generated by a radicals are performed by the Electronic Paramagnetic process known as redox cycling or Fenton reaction, and Resonance Spectroscopy (EPR) by DMPO-spin trap of is catalysed by a transition metal such as Fe2+: the two free radicals. The EPR apparatus was a Miniscope 2+ 3+ - • MS200 equipped with a probe for a quartz capillary Fe + H2O2 → Fe + OH + OH tube of 100 mm length and 1.3 mm internal diameter to place the reaction solution. The EPR spectra identities The 2 mM hydroxyl radical produced in the Fenton were validated by referring the spectral data to a control reaction solution was then trapped by DMPO, following solution consisting of a 0.05 mM DPPH• in benzene an already validated methodology [25,35]. The resultant with spectra constants described in literature [33]. The adduct DMPO–•OH, consisting in a quadruplet with percent scavenger activity of each tested compound or 1:2:2:1 relative intensities had spectral constants close

mixed solution was expressed as the following formula: to the literature (g=2.0055, an=1.49 mT, ah=1.47 mT) and was registered after 1 minute reaction at 25°C.

I=100–[(hx / ho)×100] The instrumental constants were as follows: field set, 3350 G; scan range, 100 G; scan time, 1 min; modulation

where I was the scavenger activity, ho and hx were amplitude, 3000 mG; microwave attenuation 20 dB; the heights of the same EPR signals in a reaction receiver gain, 1×102. • mixture without (ho) and with (hx) the scavenger solution, For OH measurement, the solution without the respectively. scavenging compounds contained: 0.1 mL of PBS; The scavenging activity was expressed as mM 0.1 mL of PBS 5-fold diluted; 0.1 mL of 10mM Fe-EDTA equivalents of Trolox, as previously reported for the in PBS; 0.1 mL of 50 mM DMPO in PBS; 0.1 mL of

ORAC test. The correlation data (rxy) were calculated 10 mM H2O2 in PBS. by simple linear regression. The sample with the test scavenging compound contained: 0.1 mL of PBS; 0.1 mL of wine extract 5-fold • - 2.7.1 O2 test (SASC) diluted; 0.1 mL of 10 mM Fe-EDTA in PBS; 0.1 mL of

This test followed the recommendations described 50 mM DMPO in PBS; 0.1 mL of H2O2 in PBS. Each in previous studies [34,35], with some modifications. assay was repeated three times. • - The O2 was generated in a freshly prepared 15.0 mM + solution of KO2, complexing K with an equimolecular 2.8 Contribution of single compounds to amount of 18-crown-6-ether in anhydrous antioxidant data dimethylsulphoxide (DMSO) and trapped with 50 mM of Gallic acid solutions were used as the standard of wine 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) dissolved in polyphenols, and all other compounds were separately PBS 0.1M pH 7.4. The experimental conditions were: analyzed: tartaric, malic, lactic and succinic acids; field set, 3350 G; scan range, 150 G; scan time, 2 min; glucose, fructose and glycerin.

150 R. Lo Scalzo et al.

Stock solutions of gallic acid, glycerin, tartaric peaking at 17.5 g/L in “Teroldego” wine. The high acid, malic acid, lactic acid, succinic acid, glucose level of glycerine in this variety might be explained by and fructose (100 mM) were made in PBS 0.025 M, the late ripening stage of the grape, with the lowest subsequently diluted in the same solutions as the wine amount of malic acid (Table 1), yielding a high amount extracts, and assayed with the three ROS scavenging of simple sugars. Hence, it could be supposed that tests, ORAC, SASC and HRSC with exactly the same the glycerol-pyruvic fermentation, transforming the methodology used for wine samples. sugars into glycerin at the initial stage of the wine- Oxygen radicals scavenging data (% scavenging with making process, triggers the production of high respect to the control, fixed at 0% activity, see previous glycerin levels [36]. formula) were expressed, as for previous wine samples, After fractionation, glucose, fructose and glycerin as Trolox equivalents (mM) and have been plotted against were exclusively found in FR1, with no significant concentrations of single compounds and linearly forced difference in UN samples. to straight lines. The respective equations represented Total Polyphenols Index, measured by Folin- the reactivity of the assayed compounds: the higher the Ciocalteu reaction showed (Table 1) a higher amount in slope or angle coefficient, the higher its activity (Table 2). UN red (average 184.7 mg/100 mL) than UN white wines The minimum R2 value obtained was 0.85. (26.4 mg/100 mL). The highest value was in “Refosco” (323.6 mg/100 mL), while “Greco” showed the highest value for UN white wine samples (31.0 mg /100 mL). 3. Results and Discussion The lowest value was found in “Bombino bianco” with an amount of 23.7 mg/100 mL. The above mentioned 3.1 Chemical composition polyphenols amount are in good accordance with those The organic acid profile of wine (Table 1) was elucidated found in the literature for the same varieties [37-39]. by measuring the four main hydroxy acids of wine, two of The fractionation on C18-SPE resulted in an almost which are already present in grapes - tartaric and malic acid exclusive presence of polyphenols in FR2, with no - and two which result from fermentation - lactic and succinic statistical difference between UN and FR2, respectively. acid. A low presence of malic acid corresponded to a higher The only varieties with a significant difference between presence of lactic, indicating that malolactic fermentation UN and FR2 were “Bombino nero”, “Refosco” and had occurred. This was evident in four samples of red “Greco”. However, the average values resulted in a lack wines: 3, 4, 5 and 6. By contrast, sample 7, Teroldego, of significance between UN and FR2, while FR1 showed showed low values of both malic and lactic acids. The negligible values of polyphenols, significantly lower than tartaric acid amount was higher in UN white wines (average in UN and FR2, yielding an average recovery of 6.7% for 402.4 mg/100 mL) compared to the 178.6 mg/100 mL found red and 14.8% for white wines. in UN red wines. The quantity of succinic acid was lower than previously evaluated for the other acids: in red wines, 3.2 ROS scavenging it averaged 24.4 mg/100 mL, with the maximum in sample As expected, the antioxidant capacity (Figure 1) 6. In white wines it had an average of 27.9 mg/100 mL, with measured by the ORAC test was higher in UN red wine the highest value in sample 9 (Table 1). than UN white wine extracts, averaging 120.3 mM Trolox The fractionation process indicated the presence eq. and 41.3 mM Trolox eq., respectively. The highest of hydroxy acids only in FR1. The amount of tartaric value was in “Marzemino” (136.3 mM Trolox eq.), while acid in UN did not significantly vary from that of FR1 the lowest was in “Bombino bianco” (38.1 mM Trolox (average of red wines was 178.0 and of white was 406.4 eq.). The peroxyl scavenging indices found in previous mg/100 mL). Only traces of tartaric acid were shown in works resulted in a strong variability [40-42]. This FR2, due to the higher amount with respect to the other concern might be due to the different sample history - assayed acids. The amount of malic, lactic and succinic very significant for wines - where the technological acid did not significantly differ between UN and FR1, variables are important. As an example, in the present whereas in FR2 these acids were completely absent. work, the complete absence of sulphur dioxide due to The amounts of glucose and fructose were very the sample preparation provided a strong difference: low. The average glucose and fructose contents were this compound has demonstrated a significant role in 20.1 and 59.2 mg/100 mL in red wines; and 57.9 and the oxygen consumption capacity of wine samples [43]. 88.3 mg/100 mL in white wines, respectively. However, the difference between the activity of red and The most prevalent organic compound in wine white wines was in good accordance with previous except for ethanol is glycerin. It averaged (Table 1) results [40], also consistent with a high correlation with 12.6 g/L in UN red and 10.3 g/L in UN white wines, the polyphenol content (see after) [41].

151 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

After fractionation, the ORAC activity was found and white wine samples, respectively. The average almost exclusively in all FR2 extracts (phenolics), with percent recovery of ORAC activity in FR2 from red average values of 94.8 and 65.1 mM Trolox eq. for red wines was 85%. On the other hand, the FR2 values of

Figure 1. ROS scavenging capacities of ten wine samples (1-7 red, 8-10 white, n=30) measured with three different scavenging assays: peroxyl • - • radical (ROO., ORAC), superoxide anion ( O2 , SASC), and hydroxyl radical ( OH, HRSC). UN, unfractionated sample; FR1, directly

eluted fraction; FR2, retained by C18 and eluted with MeOH. All units are in mM Trolox equivalents. Bars represent standard deviations. Different letters are for significant differences (P<0.05) between each UN, FR1 and FR2 value.

152 R. Lo Scalzo et al.

white wines were significantly higher than UN samples (2.4 and 1.1 mM Trolox eq., respectively). It has to be (Figure 1), possibly due to the positive effect of the partial pointed out that literature data about HRSC by wine extract purification in FR2, resulting in a significantly often disagreed with those presented here, with high higher value in FR2 than UN in the average data. A very values and great sample dilutions with respect to those small fraction of the ORAC activity was present in FR1 used in the present study [44,45]. The possible cause (non-phenolic fraction), 8.8 and 3.6 mM Trolox eq. in red is that ethanol, demonstrated to be a hydroxyl radical and white wines, corresponding to 7.9 and 8.7% of the scavenger [46] has been included in the scavenging ORAC values in UN red and white wines, respectively. calculation: where the ethanol has not been considered For each grape variety, the differences between FR1 and [45,47], the data are in good accordance with those of other extracts was significant in all analyzed samples. the present work. · - In regard to O2 scavenging (SASC), a total The most important aspect of the HRSC assays was 80-fold dilution was achieved to significantly quench that the activity of FR1 samples was always significantly · - 5.6 mM O2 in 1 minute at 25°C. The SASC activity was higher than FR2 and close to UN samples, an inverse significantly higher in UN red (average 4.4 mM Trolox situation with respect to ORAC and SASC profiles. eq.) than in UN white wines (average 1.1 mM Trolox eq.) The samples “Abbuoto”, “Refosco”, “Bombino bianco” as showed in Figure 1. The highest value was found in and “Malvasia”, resulted in a significant difference in “Refosco” (5.38 mM Trolox eq.); the lowest one was in scavenging index between UN and FR1 extract: this fact “Bombino bianco” (0.71 mM Trolox eq.), confirming their could be due to the wide range of compounds active performances already found for total polyphenols. In red against •OH. The FR1 indices however, were never wines, FR2 values were close to UN ones (4.3 mM Trolox similar to FR2 values, which remained very low. eq.), without significant differences, while significantly The FR1 contribution to the scavenging value in red lower values were found in FR1 extracts (0.6 mM wines was 94.4%, while the average value of white wines Trolox eq.). White wines showed different results: FR1 was slightly higher than UN samples, with no statistical and FR2 average values (1.4 and 1.2 mM Trolox eq., difference. FR2 contribution was much lower: 16.7% in respectively) were similar with no significant differences red and 5.6% in white wines. In fact, no linearity was between the two fractions and the unfractionated found with polyphenols (rxy=-0.17, n=60), while high sample. In fact, the percent of SASC calculated on the regression values were found for glycerin (rxy=0.85, values of UN samples was 13.6% for FR1 and 97.7% for n=60), succinic acid (rxy=0.82, n=60) and fructose FR2 in red wines, while in white wines the FR1 and FR2 (rxy=0.76, n=60). Other hydroxy acids correlated with values were equally divided. HRSC to a lower extent, but with positive values, A high regression value (rxy=0.93, n=60) was found 0.73 and 0.56 for tartaric and malic acids, respectively. between total polyphenols amount and ORAC data. The scarce linearity of HRSC with polyphenol amount Moreover, the simple regression between the SASC has been also discussed by Arnous et al., [44] and assay and the measured chemical indices of the wine Makris et al. [48], reinforcing the present findings. were close to linearity for total polyphenols (rxy=0.90, As for the comparison between the different tests, a n=60). No other measured chemical parameter was high rxy value (0.90, n=60) was found between ORAC found to be significantly correlated either with ORAC or and SASC, while other combinations gave low values SASC activity. (ORAC vs HRSC rxy=-0.35, n=60; SASC vs HRSC The scavenging data against •OH (HRSC) returned rxy=-0.24, n=60). completely different results from those found using ORAC and SASC (Figure 1). First, a 20-fold dilution 3.3 Contribution of single compounds to was necessary to obtain a significant quenching of antioxidant data 2 mM •OH after 1 minute reaction at 25°C, resembling The rankings of reactivity for gallic acid and Trolox in the the different reactivity between SASC and HRSC found three ROS scavenging tests returned results similar to by Garcia-Alonso et al. [19]: the reactivity of superoxide those found in previous studies on similar compounds scavengers is considered higher than that of hydroxyl [17,29,49,50]. radicals, because a lower amount of scavengers are The reactivity profile for ORAC, SASC and HRSC tests needed to obtain the same scavenging index during the (Table 2) was measured by comparing each compound same time of reaction. The activity differences between with Trolox as calibrant. Gallic acid was 33.8-fold more red and white wines were not evident as shown before reactive than Trolox, being about 440 times more reactive for the previous assays; the average was 1.8 mM Trolox than the average of other compounds. The hydroxy acids eq. for both UN wine types. The highest value was found had very low values, and slightly higher reactivity values in “Marzemino”, while the lowest was in “Petit verdot”, were found for glycerin, glucose and fructose.

153 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

As for SASC, (Table 2), gallic acid stood out for its as gallic acid simulating that it was the only phenolic scavenging strength, with a similar ranking of ORAC— wine component, and it is evident that the situation is far 28.8-fold more reactive than Trolox, showing a strong more complex than is described here. accordance with the findings of Tabart et al., [51], who A percent contribution was evaluated for each reported a relative index of reactivity of 24.9. Other single class of measured compounds compared to the compounds gave low (glycerin, glucose and fructose) overall theoretical scavenging activity. The contribution or negative (hydroxy acids) values. These negative of the phytochemicals measured in wine to the ORAC values were due to the increased production of spin- response was almost completely due to the polyphenols, · trap adduct with hydroperoxyl radical (HO2 ) formation with a higher value for red samples (96.9%) than white given by the presence of acid sources in the reaction ones (83.0%). All other compounds gave very low solutions. In short, gallic acid was about 2700-fold more indices, except for glycerin, which contributed 13.4% to reactive than the average of the other compounds. the antioxidant capacity of white wines. The reactivity slopes of single compounds for the For SASC, the activity was largely due to the HRSC assay (Table 2) showed a high activity for Trolox, polyphenols (99.8 and 99.5% in red and white wines, with gallic acid 2.5-fold less reactive. The reactivity of respectively); there was only a low contribution from gallic acid versus other compounds was much less than glycerin (0.4 and 2.5%), yet it was still higher than other that found in ORAC and SASC tests. Gallic acid was components. Hydroxy acids generally gave negative only 6.5 times more reactive than the average of other responses. compounds; hydroxy acids being slightly more reactive As for HRSC, the situation completely changed: than glycerin and simple sugars. in red wines the response was mainly due to glycerin A theoretical scavenging index (Table 3) was (51.5%) and polyphenols (29.9%); tartaric, malic and calculated by the single contributions from the lactic acids having an average value of about 5%; concentrations of the measured chemicals in UN wines. succinic acid, glucose and fructose being around 1%. The theoretical scavenging values were close to the In white wines, glycerin showed the highest contribution real measured values only for HRSC data, with higher (60.8%); then tartaric acid (18.7%) and a low contribution average values for red wine samples with respect to of polyphenols (6.2%), followed by other components. white ones. The ratio between the theoretical value This fact could be explained by the significant reactivity and the actual measured one was 1.5 for red and 1.1 of hydroxy acids and glycerin if compared to gallic acid for white wines (Table 3), so demonstrating that the towards •OH. In the other two ROS scavenging tests, measured chemical profile of wine could explain its hydroxy acids and glycerin were much less reactive antioxidant potential against •OH. Instead, the average than gallic acid against the assayed ROS. SASC and ORAC data were much higher than the real The very different reactivity among various free measured values, with a higher ratio in red than in white radicals is in accordance with previous findings: some wines (Table 3). The ratio between theoretical and anthocyanins, selected for superoxide and hydroxyl calculated values was higher for red wine samples in scavenging by an EPR approach, resulted in a very both SASC and ORAC assays (average 6.4) compared different ranking of activity towards the two free radicals. to white wine extracts (average 3.1). This discrepancy Malvidines showed the highest reactivity towards was probably due to the use of a single compound such superoxide anions, while delphinidines had the best score against hydroxyl radicals [19]. On the other hand, a close relationship was found between the antioxidant ORAC SASC HRSC capacity of selected vegetables measured by fluorimetry gallic acid 33.820 28.770 0.405 on ROO• and •OH [52]. It is evident that the type of tartaric acid 0.040 -0.019 0.073 molecular probe used in the assay plays a crucial role in malic acid 0.070 -0.034 0.075 the antioxidant measurement. lactic acid 0.040 -0.013 0.058 succinic acid 0.040 -0.025 0.072 glucose 0.100 0.004 0.054 4. Conclusions fructose 0.110 0.002 0.058 glycerin 0.140 0.009 0.050 It is our opinion that peroxyl, superoxide anion and hydroxyl radical assays have to be considered together Table 2. Slopes or angle coefficient in a plot concentration (mM) vs . in the evaluation of the total antioxidant profile of a food activity (mM Trolox equivalents) of single compounds found product. They are relevant in the onset and etiology of in wine extracts for peroxyl radicals (ORAC) superoxide anions (SASC) and hydroxyl radicals (HRSC). many diseases (a check on PubMed gave 934, 2907

154 R. Lo Scalzo et al.

3.66 2.61 6.46 1.48 1.05

1.11 4.37 1.78 1.75

4.06 2.65 1.84 28.21

0.10 0.12 1.37 2.73 100.00 1.12 0.442.48 100.00 100.00 17.58 644.48 120.30 5.36 14.39 107.59 41.30 13.38 100.00 51.54 100.00 60.75 100.00

0.65 0.00 0.97 0.00 0.04 0.10 0.06 0.91 0.00 0.02 1.44 3.07

0.20 0.00 0.58 0.00 0.01 0.03 0.03 0.54 0.00 0.03 0.48 1.89 glucose fructose glycerin total measured theor./meas.

acid 0.10 0.11 0.03 0.02 0.03 0.10 1.12 1.85 -0.01 -0.01 -0.02 -0.15 succinic % contribution

acid 0.48 0.02 0.00 0.15 0.07 0.01 0.02 5.78 0.41 -0.02 -0.06 -0.02 lactic

acid 0.30 0.83 0.13 0.05 0.13 0.77 4.95 7.12 -0.03 -0.03 -0.11 -0.73 malic

acid 1.42 0.13 0.08 0.34 1.32 4.75 -0.04 18.73 tartaric tartaric acid malic acid lactic acid succinic acid glucose fructose glycerin total

total total 0.79 polyphenols polyphenols Average theoric scavenging activity (mM Trolox equivalents) and % contribution to the total antioxidant capacity (peroxyl radical, ORAC; superoxide anion, SASC; hydroxyl radical HRSC) calculated theoric scavenging activity (mM Trolox Average by the calibration of each measured substance, using gallic acid as standard for total polyphenols. The column “measured” represents the actual measured values, calculated from the averages of UN samples of red and white wines.

SASC HRSC ORAC average red 624.66 0.52 average red average red 28.15 -0.02 average red 96.92 average red 99.80 -0.06 average red 29.93 SASC theoric HRSC theoric ORAC theoric average white 89.27 average white 4.04 average white 0.11 average white 82.96 average white 99.46 -1.10 average white 6.18 Table 3. Table

155 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

and 81 results for “hydroxyl radical”, “superoxide anion”, to the diffusion rate, so the present results are in “peroxyl radical”, and “disease”, respectively). This accordance with a previous review by Halliwell et al., fact was already stated by Ou et al., [32] who affirmed: [53], dealing with the characterization of antioxidants, : “However, the ORAC assay cannot be considered a affirming that sugars - often present in relevant amount total antioxidant activity assay to elucidate a full profile in foods - can be effective scavengers of hydroxyl of antioxidant activity against various ROS/RNS, such radicals. • - • • as O2 , OH and NO , the development of different The present work has tried to quantify the reciprocal methods specific for each ROS/RNS is needed.” action of wine components against different ROS. Many Peroxyl and superoxide anion scavenging had a different chemical species other than the polyphenol similar trend and gave a high reciprocal correlation value. fraction are found in raw plant extracts; this study Moreover, the total polyphenol index of wine extracts emphasized the protective action of such compounds strongly correlated with peroxyl and superoxide scavenging. against oxygen radicals. In the case of wine, there is The present study, however, demonstrated that a very different reactivity of wine components towards hydroxyl radicals can be effectively scavenged by wine different types of ROS, and it should be underlined the compounds other than polyphenols. The antioxidant non-exclusive contribution of polyphenols as traditionally action of wine found for •OH fit with what was previously considered antioxidants. A role of the scavenging of •OH found in grape extracts [21]. and also, to a lesser extent, a significant scavenging The reason for this particular behaviour of hydroxyl of ROO• in white wines has been attributed to the radical can be found in its very strong reactivity, close contribution of glycerin.

References

[1] Betes-Saura C., Andrès-Lacueva C., Ramuela- [9] Pascual-Martì M.C., Salvador A., Chafer A., Berna Laventos R.M., P, J. Agr. Food Chem., 1996, 44, A., Supercritical fluid extraction of resveratrol from 3040-3046 grape skin of Vitis vinifera and determination by [2] Rice-Evans C., Packer L., Flavonoids in health and HPLC, Talanta, 2001, 54, 735-740 disease, Marcel-Dekker, New York, 1998 [10] Sànchez-Moreno C., Larrauri J.A., Saura-Calixto [3] Pratt D.E., Phenolic compounds in food and their F., Free radical scavenging capacity and inhibition effects on health, American Chemical Society, of lipid oxidation of wines, grape juices and related Washinton DC, 1992 polyphenolic constituents, Food Res. Int., 1999, [4] Cantos E., Espìn J.C., Tomàs-Barberàn F.A., 32, 407-412 differences among the polyphenols profiles [11] Mannino S., Brenna O., Buratti S., Cosio M.S. A new of seven table grape cultivars studied by LC-DAD- method for the evaluation of the ‘Antioxidant Power’ MS-MS, J. Agr. Food Chem., 2002, 50, 5691-5696 of wines, Electroanalysis, 1998, 10, 908-912 [5] Saint-Cricq de Gaulejac N., Glories Y., Vivas N., [12] Waterhouse A.L., The phenolic wine antioxidants, Free radical scavenging effect of anthocyanins in p. 401, In: Packer L., Cadenas E. (Eds.), Handbook red wines, Food Res. Int., 1999, 32, 327-333 of antioxidants, CRC Press, Cleveland (Ohio), [6] Gardner P.T., McPhail D.B., Crozier A., Duthie G., 2002 ESR spectroscopic assessment of the contribution [13] Cho Mi J., Howard L.R., Prior R.L., Clark J.R., of quercetin and other flavonols to the antioxidant Flavonoid glycosides and antioxidant capacity capacity of red wines, J. Sci. Food Agr., 1999, 79, of various blackberry, blueberry and red grape 1011-1014 genotypes determined by HPLC/MS, J. Sci. Food [7] Yilmaz Y., Toledo R.T., Major flavonoids in grape Agr., 2004, 84, 1721-1782 seeds and skins: antioxidant capacity of catechin, [14] Stockley C.B., Høj P.B., Better wine for better epicatechin and gallic acid, J. Agr. Food Chem., health: fact or fiction?, Austr. J. Grape and Wine 2004, 52, 255-260 Res., 2005, 11, 127-138 [8] Burns J., Gardner P.T., O’Neil J., Crawford S., [15] Duda-Chodak A, Tarko T., Antioxidant properties Morecroft I., McPhail D.B., et al., Relationship of different fruit seeds and peels, Acta Sci. Pol. among antioxidant activity, vasodilation capacity Technol. Alim., 2007, 6, 29-36 and phenolic content of red wines, J. Agr. Food [16] Plumb G.W., de Pascual-Teresa S., Santos- Chem., 2000, 48, 220-230 Buelga C., Cheynier V., Williamson G., Antioxidant

156 R. Lo Scalzo et al.

properties of catechins and proanthocyanidins: méthodes internationales d’analyse des vins et des effect of polymerisation, galloylation and moûts], Organisation International de la Vigne et glycosilation, Free Rad. Res., 1998, 29, 351-358 du Vin (O.I.V.), Paris, France, 1990, (in French) [17] Yeh C.T., Yen G.C., Effects of phenolic acids on [29] Saraswathy Benherlal P., Arumughan C., Chemical human phenolsulfotransferases in relation to their composition and in vitro antioxidant studies on antioxidant activity, J. Agr. Food Chem., 2003, 51, Syzygium cumini fruit, J. Sci. Food Agr., 2007, 87, 1474-1479 2560-2569 [18] Arnous A., Makris D.P., Kefalas P., Effect of [30] Caccamo F., Carfagnini G., Di Corcia A., principal polyphenolic components in relation to Samperi R., Improved high-performance liquid antioxidant characteristics of aged red wines, J. chromatographic assay for determining organic Agr. Food Chem., 2001, 49, 5736-5742 acids in wines, J. Chrom. A, 1986, 362, 47-53 [19] Garcia-Alonso M., Rimbach G., Sasai M., Nakahara [31] López-Tamames E., Buxaderas S., Puig-Deu M.A., M., Matsugo S., Uchida Y., et al., Electron spin Teixeira E., Organic Acids, Sugars, and Glycerol resonance spectroscopy studies on the free radical Content in white winemaking products determined scavenging activity of wine anthocyanins and by HPLC: relationship to climate and varietal pyranoanthocyanins, Mol. Nutr. Food Res., 2005, factors, Am. J. Enol. Vitic., 1996, 47, 193-198 49, 1112-1119 [32] Ou B., Hampsch-Woodill M., Prior R.L., [20] Lo Scalzo R., Morassut M., Bianchi G., Morelli R., Development and validation of an improved Free radical scavenging activity of grape’s (Vitis Oxygen Radical Absorbance Capacity assay using vinifera L.) skin and pulp, Vol. 17, Ch. 15, p. 321, fluorescein as the fluorescent probe, J. Agr. Food In: “Chronic and Common Disease I”, Govil J.N. Chem., 2001, 49, 4619-4626 (Ed.), Studium Press LLC, Houston, (Tx), 2006 [33] Weil J.A., Bolton J.R., Wertz J.E., Electron [21] Lo Scalzo R., Iannoccari T., Summa C., The Paramagnetic Resonance. Elementary theory and relationship between the composition of different practical applications, Appendix E, Experimental table grape (Vitis vinifera L.) extracts and three considerations, pp. 492-519, Wiley Interscience methods of measuring their free radical scavenging Inc., New York, 1994 properties, It. J. Food Sci., 2007, 19, 329-341 [34] Privat C., Telo J.P., Bernardes-Genisson V., [22] Falchi M., Bertelli A., Lo Scalzo R., Morassut M., Vieira A., Souchard J.P., Nepveu F., Antioxidant Morelli R., Das S., et al., Comparison between properties of trans-ε-viniferin as compared to cardioprotective abilities bertween the flesh and stilbene derivatives in aqueous and nonaqueous skin of grapes, J. Agr. Food Chem., 2006, 54, media, J. Agr. Food Chem., 2002, 50, 1213-1217 6613-6622 [35] Valavanidis A., Nisiotou C., Papageorgiou Y., [23] Gray J., Mower H.F., The role of simple KremLi I., Satravelas N., Zinieris N., et al., carbohydrates in the suppression of hydroxyl free Comparison of the radical scavenging potential radicals in gamma-irradiated papaya juice, Food of polar and lipidic fractions of olive oil and other Chem., 1991, 41, 293-301 vegetable oils under normal conditions and after [24] Wehmeier K.R., Mooradian A.D., Autoxidative and thermal treatment, J. Agr. Food Chem., 2004, 52, antioxidative potential of simple carbohydrates, 2358-2365 Free Rad. Biol. Med., 1994, 17, 83-86 [36] Moreno-Arribas M.V., Polo M.C., Wine Chemistry [25] Morelli, R., Russo-Volpe S., Bruno N., Lo Scalzo and Biochemistry, Springer Science Business R., Fenton-dependent damage to carbohydrates: Media, New York, 2009 free radicals scavenging activity of some simple [37] de Palma L., Lopriore G., Tarricone L., Novello V., sugars, J. Agr. Food Chem., 2003, 51, 7418-7425 Soleti F., Vegetative and production development, [26] Kayashima T, Katayama T., Oxalic acid is available water system and grape and wine quality in as a natural antioxidant in some systems, Biochim. “Montepulciano” brought up with different distances Biophys. Acta, 2002, 1573, 1-3 on the row [Sviluppo vegetativo e produttivo, stato [27] Gil M.I., Tomás-Barberán F.A., Hess-Pierce B., idrico e qualità dell’uva e del vino in “Montepulciano” Holcroft D.M., Kader A.A., Antioxidant activity allevato con diverse distanze sul filare], Quad. Vitic. of pomegranate juice and its relationship with Enol. Univ. Torino, 2005, 28, 223-234 (in Italian) phenolic composition and processing, J. Agr. Food [38] Rigo A., Vianello F., Clementi G., Contribution Chem., 2000, 48, 4581-4589 of Proanthocyanidins to the peroxyl radical [28] Compendium of the international analytical scavenging capacity of some Italian red wines, J. methods of wines and grape juices [Recueil des Agr. Food Chem., 2000, 48, 1996-2002

157 Oxygen radical scavenging capacity of phenolic and non-phenolic compounds in red and white wines

[39] Pellegrini N., Simonetti P., Gardana C., Brenna [47] Puškaš V., Jović S., Antov M., Tumbas V., O., Brighenti F., Pietta P., Polyphenol content and Antioxidative activity of red wine with the increased total antioxidant activity of Vini novelli (Young red share of phenolic compounds from solid parts of wines), J. Agr. Food Chem., 2000, 48, 732-735 grape, Chem. Ind. Chem. Eng. Quarterly, 2010, 16, [40] Dàvalos A., Gòmez-Cordovès C., Bartolomè 65-71 B., Extending applicability of the oxygen radical [48] Makris D.P., Boskou G., Andrikopoulos N.K., absorbance capacity (ORAC-fluorescein) assay, J. Polyphenolic content and in vitro antioxidant Agr. Food Chem., 2004, 52, 48-54 characteristics of wine industry and other agri-food [41] Yilmaz Y., Toledo R.T., Oxygen radical absorbance solid waste extracts, J. Food Comp. Anal., 2007, capacities of grape-wine industry byproducts and 20, 125-132 effect of solvent type on extraction of grape seed [49] Pérez-Jiménez J., Saura-Calixto F., Effect of polyphenols, J. Food Comp. Anal., 2006, 19, 41-48 solvent and certain food constituents on different [42] Rossetto M., Vanzani P., Zennaro L., Mattivi F., antioxidant capacity assays, Food Res. Int., 2006, Vrhovsek U., Scarpa M., et al., Stable free radicals 39, 791-800 and peroxyl radicals trapping capacity in red wines, [50] da Silva J.M.R., Darmon N., Fernandez, Y., Mitjavilat J. Agr. Food Chem., 2004, 52, 6151- 6155 S., Oxygen free radical scavenger capacity in [43] Manzocco L., Mastrocola D., Nicoli M., Chain- acqueous models of different procyanidins from breaking and oxygen scavenging properties of wine grape seeds, J. Agr. Food Chem., 1991, 39, 1549- as affected by some technological procedures, 1552 Food Res. Int., 1999, 31, 673-678 [51] Tabart J., Kevers C., Pincemail J., Defraigne [44] Arnous A., Makris D.P., Kefalas P., Correlation J-O., Dommes J., Comparative antioxidant of pigment and flavanol content with antioxidant capacities of phenolic compounds measured properties in selected aged regional wines from by various tests, Food Chem., 2009, 113, 1226- Greece, J. Food Comp. Anal., 2002, 15, 655-665 1233 [45] Ghiselli A., Nardini M., Baldi A., Scaccini C., [52] Cao G., Sofic E., Prior R.L., Antioxidant capacity of Antioxidant activity of different phenolic fractions tea and common vegetables, J. Agr. Food Chem., separated from an Italian red wine, J. Agr. Food 1996, 44, 3426-3431 Chem., 1998, 46, 361-367 [53] Halliwell B., Aeschbach R., Löliger J., Aruoma O.I., [46] Halliwell B., Gutteridge J.M.C., Aruoma O.I., The The characterization of antioxidants, Food Chem. deoxyribose method: a simple “test tube” assay for Toxicol., 1995, 33, 601-617 the determination of rate constants for reactions of hydroxyl radicals, Anal. Biochem., 1987, 165, 215-219

158