Detailed Phenolic Composition of White Grape by-products by RRLC/MS and Measurement of the Antioxidant Activity

Elsevier Editorial System(tm) for Talanta Manuscript Draft Manuscript Number: Title: Detailed phenolic composition of white grape by-products by RRLC/MS and measurement of the antioxidant activity Article Type: Research Paper Keywords: By-products; phenolic compounds; grape; rapid resolution liquid chromatography (RRLC); mass spectrometric (MS) Corresponding Author: Prof. Francisco J. Heredia, PhD Corresponding Author's Institution: Universidad de Sevilla First Author: M. José Jara-Palacios Order of Authors: M. José Jara-Palacios; Dolores Hernanz, PhD; Susana González-Manzano, PhD; Celestino Santos-Buelga, PhD; M. Luisa Escudero- Gilete, PhD; Francisco J. Heredia, PhD Abstract: The development and validation of a rapid method of RRLC has been carried out to determine the phenolic composition of winemaking byproducts (pomaces, seeds, skins and stems). Thirty-one phenolic compounds belonging to three groups (flavanols, flavonols and phenolic acids) have been identified by use of standards and mass spectrometric detection, and quantified by using the corresponding external standard calibration plot, in a 16-minute run. The validation was realized calculating the repeatability, the reproducibility and the limits of detection (LOD) and quantification (LOQ), from standards solutions. The limits of detection and quantification were in the range of 0.16-1.09 and 0.52-3.63 mg/L, respectively, and good repeatability (R.S.D. values < 1.5%) and reproducibility (R.S.D. values < 5.5%) were found. Results confirmed that the method is effective and suitable for determination of phenolic compounds in winemaking by-products. Seeds, skins, stems and pomaces exhibited a different qualitative and quantitative phenolic profile and different antioxidant activities. Highlights 1. A rapid RRLC method has been validated to determine the phenolic composion of winemaking by-products. 2. Pomaces, seeds, skins and stems have been analyzed by RRLC and MS. 3. Thirty-one phenolic compounds have been idenfied and quanfied in samples in a 16-minute run. 4. Samples exhibited different phenolic profile and anoxidant acvity. Seeds Stems Skins Grape Pomace RRLC method Validation 16 minutes LOD/LOQ 31 polyphenols Repeatability Reproducibility FRAP ABTS Analysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Detailed phenolic composition of white grape by-products by RRLC/MS and 1 measurement of the antioxidant activity. 2 3 M. José Jara-Palacios1, Dolores Hernanz2, Susana González-Manzano3, Celestino 4 Santos-Buelga3, M. Luisa Escudero-Gilete1, Francisco J. Heredia1 5 6 1Food Colour & Quality Lab., Dept. Nutrition & Food Science, Universidad de Sevilla. 7 Facultad de Farmacia, 41012 Sevilla, Spain. 8 2Department of Analytical Chemistry, Universidad de Sevilla. Facultad de Farmacia,9 41012 Sevilla, Spain.10 3Grupo de Investiga -11 Bromatología, Facultad de Farmacia, Universidad de Salamanca, Salamanca, Spain.12 13 * Corresponding author: 14 Francisco J. Heredia 15 Food Colour & Quality Lab., Dept. Nutrition & Food Science, Universidad de Sevilla.16 Facultad de Farmacia, 41012 Sevilla, Spain. 17 Tel.: +34 954556495 Fax: +34 95455611018 e-mail: [email protected]19 20 21 22 23 24 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Abstract 26 27 The development and validation of a rapid method of RRLC has been carried out to 28 determine the phenolic composition of winemaking by-products (pomaces, seeds, skins 29 and stems). Thirty-one phenolic compounds belonging to three groups (flavanols, 30 flavonols and phenolic acids) have been identified by use of standards and mass 31 spectrometric detection, and quantified by using the corresponding external standard 32 calibration plot, in a 16-minute run. The validation was realized calculating the 33 repeatability, the reproducibility and the limits of detection (LOD) and quantification 34 (LOQ), from standards solutions. The limits of detection and quantification were in the 35 range of 0.16-1.09 and 0.52-3.63 mg/L, respectively, and good repeatability (R.S.D. 36 values < 1.5%) and reproducibility (R.S.D. values < 5.5%) were found. Results 37 confirmed that the method is effective and suitable for determination of phenolic 38 compounds in winemaking by-products. Seeds, skins, stems and pomaces exhibited a 39 different qualitative and quantitative phenolic profile and different antioxidant activities.40 41 42 43 44 Keywords 45 By-products; phenolic compounds; grape; rapid resolution liquid chromatography 46 (RRLC); mass spectrometric (MS). 47 48 49 50 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 INTRODUCTION 52 Phenolic compounds have been widely studied for decades because of their beneficial 53 properties on the health and their influence on the organoleptic characteristics of the 54 food. Grape pomace, consisting of seeds, skins and stems, is a winemaking by-product 55 recognized as a rich source in phenolic compounds with interest by their potential 56 natural antioxidant [1], anti-inflammatory [2] and antimicrobial activities [3], which 57 have been related with the prevention of important chronic pathologies such as 58 cardiovascular disorders [4], neurodegenerative decline [5] or cancer [6]. 59 Winemaking generates a high amount of by-products that cause environmental and 60 economic problems, which could be minimized by the exploitation and valorisation of 61 those products, such as their use in pharmaceutical and food industries. 62 Seeds, skins and stems present different qualitative and quantitative composition in 63 phenolic compounds. Seeds and stems are rich in flavanols whereas skins also present 64 flavonols, and it is well known that different phenolic compounds may show different 65 biological and antioxidant properties [7]. Several in vitro methods had been employed 66 to measure the antioxidant activity, such as ABTS and DPPH assays, and ferric 67 reducing antioxidant power (FRAP), based on an electron transfer mechanism and 68 reduction of a coloured oxidant, and others based on a hydrogen atom transfer 69 mechanism, such as oxygen radical absorbance capacity (ORAC), in which antioxidants 70 and substrate compete for thermally generated peroxyl radicals [8,9]. 71 Different techniques have been used for the separation of phenolic compounds, such as 72 high speed counter current chromatography (HSCCC), supercritical fluid 73 chromatography (SFC), capillary electrophoresis (CE) and especially high performance 74 liquid chromatography (HPLC), the most commonly used for the separation and 75 analysis of these compounds in grape, wine and related products [10]. However, these 76 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 methods either require longer analysis time or consume relatively large amounts of 77 organic solvents used as mobile phase. Considering the complexity of the grape 78 pomace, with a diversity of phenolic compounds from different groups, it is very 79 difficult to achieve good separations with a single chromatographic run. Several authors 80 have analysed polyphenols in seeds, skins and stems using different chromatographic 81 conditions with good results but usually requiring long retention times [1,11-13]. Rapid 82 resolution liquid chromatography (RRLC) is a technique of liquid chromatography in 83 which small particles are packed into short columns run with small particle size and 84 diameter. The advantages of RRLC are higher resolution and sensitivity, and shorter 85 retention times than HPLC. However, as far as we know, this technique has not been 86 used for the analysis of phenolics in grape pomace. Liquid-liquid extraction has been 87 widely used in sample preparation for further analysis of phenolic compounds. In this 88 regard, it is important to stress that the choice of the extraction solvents must be made 89 as a function of the type of sample to be analysed and the information required [14]. 90 Thus, several extraction solvents (ethanol, methanol, ethyl acetate, and sulphured water)91 have been used for the analysis of phenolic compounds in grape seeds [14-16], grape 92 skin and seeds [17], or grape pomace [18]. 93 The aim of this work was to determine the phenolic composition of white grape 94 winemaking by-products (pomaces, seeds, skins and stems) by the development and 95 validation of a rapid and effective RRLC method using MS as a detection technique for 96 compound identification and/or confirmation. In addition, the antioxidant activity of the 97 by-product extracts was measured by ABTS and FRAPS assays, and correlations with 98 the phenolic composition were established. 99 100 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 MATERIALS AND METHODS 101 Samples and Reagents 102 103 harvest collected after winemaking wa104 (Bollullos Par del Condado, Spain). Seeds, skins and stems were manually separated 105 from the grape pomace samples and all samples were further freeze-dried. 106 Hydrochloric acid, formic acid, HPLC-grade acetonitrile, methanol, ethanol, glycine, 107 Folin reagent, and iron trichloride (FeCl3·6H2O) were obtained from Panreac 108 (Barcelona, Spain). ABTS (2,2-azino-bis-(3-ethylbenzothiazolne-6-sulfonic acid) 109 diammonium salt) and Trolox (6-hydroxy-2,5,7,8-tetramethyl-chroman-2-carboxylic 110 acid) were purchased from Fluka (Madrid, Spain). 111 Gallic acid, protocatechuic acid, (+)-catechin (C), (-)-epicatechin (EC), quercetin, 112 kaempferol, ferulic acid, caffeic acid, p-coumaric acid, sodium carbonate, potassium 113 persulphate, potassium metabisulphite, 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ) and 114 phosphate buffered saline (PBS) were purchased from Sigma-Aldrich (Madrid, Spain). 115 Quercetin 3-O-glucoside and kaempferol 3-O-glucoside were obtained from 116 Extrasynthese (Lyon, France). Procyanidin dimers B1, B2, B3 and B4 and trimer C1 117 were isolated in the laboratory by semi-preparative HPLC [19].118 Sample preparation 119 The ability of different solvents to extract the polyphenols in the by-products samples 120 (pomace, seeds, skins and stems) was investigated. For this, analyses were carried out 121 after extraction with four different solvents: 70% ethanol, 40% ethanol, 75% methanol and 122 1% potassium metabisulphite, in water. The by-product sample (5 g) was homogenized in 123 25 mL of the solvent, kept under shaking for 1 h in an incubating mini shaker (VWR 124 International, Barcelona. Spain), and further centrifuged at 4190g for 15 min; the 125 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 supernatant was collected and the residue submitted to the same process twice, and the 126 supernatants combined. The extracts thus obtained were used for determination of total 127 phenols content by spectrophotometry, and the average recoveries were selected as 128 responses of interest. 129 After selection of the solvent, the extracts obtained was used for determination of the 130 antioxidant activity by FRAP and ABTS assays. Furthermore, the extracts (2 mL) were 131 concentrated to dryness and further re-dissolved in 1 mL of 0.1% formic acid to be 132 analysed by RRLC after filtration through a hydrophilic PVDF Millex-HV 0.45 m133 syringe filter (Millipore, Bedford, MA, USA). 134 Chromatography 135 Analyses were carried out in an Agilent 1260 chromatograph (Agilent Technologies, Palo 136 Alto, CA, USA) equipped with a diode-array detector, which was set to scan from 200 to 137 x 4.6 mm) using an injection 138 volume of 15 L.139 The solvents were 0.1% formic acid in water (solvent A) and acetonitrile (solvent B) at 140 the following gradient: 0-5 min, 5% B linear; 5-20 min 50% B linear; 20-25 min, 141 washing and re-equilibration of the column. The flow-rate was 1.5 mL/min and the 142 temperature of the column was set at 25 ºC. Detection was also performed in an API 143 3200 Qtrap (Applied Biosystems, Darmstadt, Germany) equipped with an ESI source 144 and a triple quadrupole-ion trap mass analyser, which was connected to the HPLC 145 equipment via the DAD cell outlet, as described by Jara-Palacios et al. [9]. Phenolic 146 compounds were identified by their retention time, UV-vis spectra and mass spectra, as 147 well as by comparison with our data library and standards when available. 148 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 presented the lowest concentration (122 mg/100 g). In contrast, flavonols were most 297 abundant in skins and pomace (146 and 144 mg/100 g, respectively) with no significant 298 differences between them, whereas seeds were poor in these compounds skins (8 299 mg/100 g), as also reported by other authors [11]. Phenolic acid derivatives were 300 minority compounds in the four by-products (Fig. 3). 301 Considering individual compounds, procyanidins B1 and B2 3-O-gallate and catechin 302 were the predominant flavanols in all by-products. The highest amounts of procyanidin 303 B2 3-O-gallate and catechin were found in seeds (146 and 65 mg/100 g, respectively) 304 and procyanidin B1 was highest in stems and seeds (89 and 85 mg/100 g, respectively). 305 Procyanidin B2, B3, B4 and B7 were also found in all by-products, being B4 the 306 predominant non-galloyled dimer after B1. A not identified galloyled procyanidin was 307 also found in relatively high amount in seeds (60 mg/100 g) while it was in lower levels 308 in the other extracts (between 4 and 26 mg/100 g). Epicatechin, which was described as 309 an important flavanol in by-products from grape and wine [13], showed very different 310 concentrations among by-products ranging between 5 and 43 mg/100 g, being more 311 abundant in seeds than in skins. As for other flavanols, procyanidin trimers and 312 tetramers also showed relevant contribution to the levels of total flavanols in seeds, 313 stems, pomace and skins (114, 54, 48 and 20 mg/100 g). 314 The main flavonols in the by-products were quercetin glycosides, with quercetin 3-O-315 glucoside and quercetin 3-O-glucuronide, accounting for 32-42% and for 35-37% of 316 total flavonol content, respectively. As expected, both compounds were more abundant 317 in pomace and skins (51-58 mg/100 g) than in stem (25-27 mg/100 g) and seed extracts 318 (around 3 mg/100 g). These compounds have been also identified as the most abundant 319 flavonols in other grape varieties grown in warm climate Spanish regions, such as 320 Airén, Chardonnay, Listán Huelva, Pedro Ximénez or Verdejo [27]. Other flavonol 321 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 glycosides were also found in the by-products but in lower concentrations than 322 quercetin glycosides, being among them kaempferol 3-O-glucoside the most abundant 323 in skins and pomaces (19 and 15 mg/100 g, respectively). Isorhamnetin 3-O-glucoside 324 was also found in the extracts in low concentrations as reported by other authors 325 [27,28]. These results are in accordance with Castillo-Muñoz et al. [27] that reported 326 quercetin glycosides as the dominant flavanols in white grapes, followed by kaempferol 327 glycosides, considered as the second in importance, and isorhamnetin glycosides that 328 occurred as very minor flavonols. In this study, quercetin and kaempferol aglycones, 329 which were not reported by other authors [27,29], were also detected in low 330 concentrations in the extracts of skins, stems and pomaces. 331 Gallic acid was the most abundant non-flavonoid phenolic compound in the samples, 332 representing between 36% (skins) and 74% (pomace) of the total contents of phenolic 333 acids and derivatives. Caftaric acid was the most abundant cinnamoyl derivative in all 334 by-products, showing higher concentrations in pomaces and stems (10 and 7 mg/100 g, 335 respectively) than in seeds and skins (around 2 mg/100 g). Rodriguez-Montealegre et al. 336 [28] described trans-caftaric as the main acid in white grape skins although they did not 337 identify it in the seeds. 338 Antioxidant activity in the studied by-products 339 The antioxidant activity of seeds, skins, stems and pomaces was measured by ABTS 340 and FRAP assays (Table 4). Seed extracts presented the greatest antioxidant capacity 341 (97 mmol TE/100g) in the ABTS assay, followed by pomace, stem, and skin extracts 342 (70, 52 and 42 mmol TE/100 g, respectively), with significant differences among them 343 (p<0.05). Similar results were obtained in the FRAP assay for seed and pomace (39 and 344 35 mmol TE/100 g, respectively) and for stem and skin extracts (22 and 20 mmol 345 TE/100 g, respectively). These results were in agreement with the total phenolic 346 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 contents determined by RRLC, which was significantly correlated with ABTS (r2=347 0.89, p<0.05) and FRAP (r2= 0.70, p<0.05) values, as showed by regression analysis. 348 Different authors also reported significant correlation between antioxidant activity and 349 the total phenolic content in winemaking by-products from different grape varieties 350 [12,16,30].351 SLDA analysis 352 To ascertain whether it was possible to discriminate between pomace, seeds, skins and 353 stems as a function of the phenolic contents, one stepwise linear discriminant analysis 354 (SLDA) was carried out. Eight variables were found significant (p<0.05): trimer 2, 355 tetramer 1, cis-coutaric acid, quercetin 3-O-glucuronide, protocatechuic acid, caftaric 356 acid, kaempferol and procyanidin B2, indicated in descending order of discriminating 357 power. Two classification functions were obtained, which yielded a good separation 358 (100% correct classification) among samples (Fig. 4). The discriminant function 1 was 359 mainly related to trimer 2, procyanidin B2 and protocatechuic acid (with positive sign), 360 and kaempferol (negative sign), whereas the discriminant function 2 was mainly linked 361 to quercetin 3-O-glucuronide, caftaric acid and cis-coutaric acid (positive sign), and 362 tetramer 1 (negative sign). 363 364 CONCLUSIONS 365 A chromatographic method for the rapid analysis of phenolic compounds in extracts of 366 winemaking by-products has been described, whose applicability is demonstrated by 367 validation criteria considering the linearity, repeatability and reproducibility. Analysis 368 of real samples of by-products has been further carried out, which constitutes, in our 369 knowledge, the first investigation of the simultaneous identification and quantification 370 by RRLC/MS of phenolic compounds belonging to different phenolic groups in 371 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 different winery by-products (pomace, seeds, skins and stems). Thirty-one phenolic 372 compounds were identified in the different samples belonging to the groups of 373 flavanols, flavonols and phenolic acid derivatives, showing quantitative differences 374 among the distinct by-products from the white grape Zalema, a Vitis vinifera variety 375 south Spain. Eight 376 phenolic compounds (a procyanidin trimer, a procyanidin tetramer, cis-coutaric acid, 377 quercetin 3-O-glucuronide, protocatechuic acid, caftaric acid, kaempferol, and 378 procyanidin B2) allowed classifying correctly 100% of the by-product samples. The 379 antioxidant activity of the different by-product samples was also determined by the 380 FRAP and ABTS assays and a correlation between it and the phenolic composition was 381 established. 382 383 Acknowledgements 384 Consejo Regulador D.O. Condado de 385 Huelva386 of Education. The GIP-USAL is financially supported by the Spanish Government 387 through the project CSD2007-00063 (Fun-c-Food, Consolider-Ingenio 2010 388 Programme).389 390 391 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 References 392 [1] I.I. Rockencbach, L.V. Gonzaga, V.M. Rizelio, A.E. de Souza Schmidt Gonçalves, 393 M.I. Genovese, R. Fett, Food Res. Int. 44 (2011) 897-901.394 395 396 - , M. Blay, J. Agric. Food Chem. 55 (2007) 397 4357-4365.398 399 [3] N. G. Baydar, O. Sagdic, G. Ozkan, S. Cetin, Int. J. Food Sci. 41 (2006) 799-804.400 401 [4] M. Falchi, A. Bertelli, R. Lo Scalzo, M. Morassut, R. Morelli, S. Das, J. H. Cui, 402 D.K. Das, J. Agric. Food Chem. 54 (2006) 6613-6622. 403 404 [5] M. Balu, P. Sangeetha, G. Murali, C. Panneerselvam, Brain Res. Bull. 68 (2006) 405 469-473.406 407 [6] M.C. , J.M. 408 Antolín Rodriguez, N. Fabris, L. Bianchi, J. Med. Food 12 (2009) 561-568. 409 410 [7] E.Q. Xia, G.F. Deng, Y.J. Guo, H.B. Li, Int. J. Mol. Sci. 11 (2010) 622-646.411 412 [8] C. Xu, Y. Zhang, L. Cao, J. Lu. Food Chem. 119 (2010) 1557-1565. 413 414 - - -415 - , F.J. Heredia, C. Santos-Buelga, J. Agric. Food Chem. 416 61 (2013) 5114-5121. 417 418 [10] A.R. Fontana, A. Antoniolli, R.J. Bottini. Agric. Food Chem. 61 (2013) 8987-419 9003.420 421 [11] D. Kammerer, A. Claus, R. Carle, A.J. Schieber, Agric. Food Chem. 52 (2004) 422 4360-4367.423 424 [12] A.M. Alonso, D.A. Guillén, C. Barroso, B. Puertas, A. García, J. Agr. Food Chem. 425 50 (2002) 5832-5836. 426 427 [13] M. Anastasiadi, H. Pratsinis, D. Kletsas, A. Skaltsounis, S. Haroutounian, Food 428 Res. Int. 43 (2010) 805-813. 429 430 [14 - Srécko, L. Jakabek, M. Seruga. Int. J. Food Sci. 431 Technol. 44 (2009) 2394-2401.432 433 [15] E. Karvela, D. Makris, N. Kalogeropoulos, V. Karathanos, P. Kefalas, Eur. Food 434 Res. Technol. 229 (2009) 731-742.435 436 [16] R. Guendez, S. Kallithraka, D.P. Makris, P. Kefalas. Food Chem. 89 (2005) 1-9. 437 438 [17 - - - , C. 439 Santos-Buelga, Food Chem. 114 (2009) 649-656. 440 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 441 [18] B. Lapornik, M. Prosek, A.G. Wondra. J. Food Eng. 71 (2005) 214-222. 442 443 [19 - - - -444 - , J. Agric. Food Chem. 54 (2006) 4326-4332. 445 446 [20] UNE 82009-1:1998. Exactitud (veracidad y precision) de resultados y métodos de 447 medición. AENOR. 448 449 [21] V.L. Singlenton, J.A. Rossi, Am. J. Enol. Vitic. 16 (1965) 144-158. 450 451 [22] I.F. Benzie, J.J. Strain, Anal. Biochem. 239 (1996) 70-76. 452 453 [23] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang, C. Rice-Evans, Free 454 Radical Biol. Med. 26 (1999) 1231-1237. 455 456 [24] StatSoft Inc. STATISTICA (data analysis software system), v 8.0; StatSoft Inc.: 457 Tulsa, OK, 2007. 458 459 [25] R. Perestrelo, Y. Lu, S.A.O. Santos, A.J.D. Silvestre, C.P. Neto, J.S. Câmara, S.M. 460 Rocha, Food Chem. 135 (2012) 94-104. 461 462 [26] O. Sagdic, I. Ozturk, G. Ozkan, H. Yetim, L. Ekici, M.T. Yilmaz, Food Chem. 126 463 (2011) 1749-1758. 464 465 [27] N. Castillo-Muñoz, S. Gómez-Alonso, E. García-Romero, I. Hermosín-Gutiérrez, 466 J. Agric. Food Chem. 55 (2007) 992-1002. 467 468 [28] R. Rodríguez Montealegre, R. Romero Peces, J.469 , E. García Romero, Food Compos. Anal. 19 (2006) 687-693.470 471 [29] N. Castillo-Muñoz, S. Gómez-Alonso, E. García-Romero, I. Hermosín-Gutiérrez, 472 J. Food Compos. Anal. 23 (2010) 699-705. 473 474 [30] B. Bozan, G. Tosun, D. Ozcan. Food Chem. 109 (2008) 426-430. 475 476 477 478 479 480 481 482 483 484 485 486 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 FIGURE CAPTIONS 487 488 Figure 1. RRLC chromatograms recorded at 280, 320 and 370 nm of a mixture of 489 standards in the optimized chromatography conditions. Peaks: 1, gallic acid; 2, 490 protocatechuic acid; 3, catechin; 4, epicatechin; 5, caffeic acid; 6, p-coumaric acid; 7, 491 ferulic acid; 8, quercetin 3-O-glucoside; 9, kaempferol 3-O-glucoside; 10, quercetin; 11, 492 kaempferol. 493 Figure 2. Average recoveries corresponding to the total phenolic content after extraction 494 of seeds, skins, stems and pomace using four different extraction solvents. Different letters 495 in the same by-product indicate significant differences by ANOVA test (p<0.05).496 Figure 3. Concentration of flavanols, flavonols and phenolic acids in seeds, skins, 497 stems and pomace of Vitis vinifera cv. Zalema. Different letters in the same phenolic 498 group indicate significant differences by ANOVA test (p<0.05).499 Figure 4. Scatterplot of the by-products samples in the plane defined by the canonical 500 function when phenolic composition is considered for discrimination. 501 502 Table 1. Analytical parameters of calibration curves of standards solutions. Compound Wavelength (nm) Intercept ± SD Slope ± SD Correlation coefficient (r2) Linear range (mg/L) LOD (mg/L) LOQ (mg/L) Catechin 280 -0.368 ± 0.77 12.016 ± 0.00 1.0000 0.4-500 0.192 0.641 Epicatechin 280 -1.506 ± 0.61 11.838 ± 0.00 1.0000 0.4-500 0.155 0.518 Quercetin 370 9.299 ± 11.07 95.705 ± 0.25 0.9998 1.5-100 0.347 1.156 Kaempferol 370 3.841 ± 9.38 79.675 ± 0.22 0.9998 1.5-100 0.353 1.778 Q-3-O-glucoside 370 -4.395 ± 7.62 21.026 ± 0.08 0.9994 0.9-250 1.088 3.628 K-3-O-glucoside 370 -4.0479 ± 3.89 20.222 ± 0.20 0.9975 0.2-50 0.578 1.927 Gallic acid 280 -7.100 ± 2.88 35.214 ± 0.05 0.9987 1.5-150 0.245 0.818 Protocatechuic acid 280 -4.499 ± 1.22 23.118 ± 0.12 0.9998 0.2-25 0.158 0.527 Caffeic acid 320 -34.080 ± 9.83 66.639 ± 0.28 0.9995 0.2-100 0.442 1.475 Ferulic acid 320 -7.028 ± 4.66 84.166 ± 0.46 0.9998 0.2-25 0.166 0.554 p-Coumaric acid 320 -8.393 ± 4.86 90.350 ± 0.48 0.9998 0.2-25 0.161 0.537 Table 2. RSD (%) values of the retention time (RT) and peak area (PA) for each standard. Intra-day (n=6) Inter-day (n=6) Compound RT PA RT PA Catechin 0.04 1.29 0.02 5.52 Epicatechin 0.03 1.43 0.02 4.14 Quercetin 0.03 0.46 0.04 3.58 Kaempferol 0.02 0.36 0.04 3.38 Q-3-O-glucoside 0.17 0.62 0.15 1.19 K-3-O-glucoside 0.17 1.46 0.12 1.12 Gallic acid 0.02 0.13 0.26 4.18 Protocatechuic acid 0.14 0.28 0.16 1.38 Caffeic acid 0.04 0.28 0.02 0.78 Ferulic acid 0.01 0.26 0.02 3.42 p-Coumaric acid 0.03 0.20 0.02 1.52 Table 3. Mass spectrometry data and concentrations of phenolic compounds in winery by-products of Vitis vinifera cv. Zalema Compound MS (m/z) 1 [M-H] MS/MS (m/z)1Seeds2Skins2Stems2Pomaces2 Flavanols Catechin (C) 289 245 65.13 17.06 a 17.13 7.82 b 57.16 11.58 a 37.84 6.30 c Epicatechin (EC) 289 245 43.35 9.14 a 4.73 1.02 b 10.64 1.97 c 13.70 3.28 c Procyanidin B1 577 425, 405, 289 84.61 17.79 a,c 32.14 6.52 b 88.79 13.90 a 72.62 9.73 c Procyanidin B2 577 425, 405, 289 19.09 4.26 a 7.50 1.63 b 8.31 1.61 b 7.59 0.95 b Procyanidin B3 577 425, 405, 289 21.77 8.79 a 7.84 2.61 b 11.99 2.84 b 8.28 3.44 b Procyanidin B4 577 425, 405, 289 31.85 6.37ª 15.12 3.35 b 33.19 5.69ª 23.82 2.29 c Procyanidin B7 577 425, 405, 289 18.25 3.38ª 1.77 1.41 b 8.22 2.24 c 5.51 1.31 d Procyanidin trimer 1 865 577, 289 13.69 2.81 a 1.85 0.71 b 5.78 1.38 c 4.07 1.16 c Procyanidin trimer 2 865 577, 289 72.35 11.29ª 7.56 2.31 b 13.84 3.45 c 23.11 3.98 d Procyanidin tetramer 1 1153 863, 577, 287 26.46 4.94ª 9.34 2.19 b 27.52 4.78ª 18.04 2.04 c Procyanidin tetramer 2 1153 863, 577, 287 0.87 0.66 a 1.50 1.26 a,c 6.45 1.52 b 2.28 0.79 c Galloyled procyanidin 729 577, 425, 407, 289 60.12 9.83 a 3.63 1.47 b 26.32 4.10 c 19.18 3.12 d Procyanidin B2 3-O-gallate 729 577, 425, 407, 289 145.62 30.45 a 11.79 5.19 b 49.97 8.16 c 46.61 11.14 c Flavonols Quercetin 3-O-rutinoside 609 301 0.88 0.34 a 10.71 3.07 b 3.28 0.49 c 8.39 1.27 d Quercetin 3-O-glucuronide 477 301 2.51 1.28 a 51.01 12.84 b 24.74 3.84 c 53.70 5.17 b Quercetin 3-O-glucoside 463 301 2.72 1.38 a 55.26 13.91 b 26.81 4.16 c 58.18 5.60 b Quercetin pentoside 433 301 0.00 0.00 a 0.44 0.21 b 0.05 0.15 a 0.52 0.09 b Kaempferol hexoside 447 285 0.24 0.24 a 4.81 1.56 b 1.38 0.22 c 3.63 0.65 d Kaempferol 3-O-glucoside 447 285 0.81 0.37 a 18.75 4.64 b 4.97 0.56 c 15.15 2.39 d Kaempferol 3-O-glucuronide 461 285 n.d. a n.d. b 0.37 0.28 c 0.04 0.00 d Isorhamnetin 3-O-glucoside 477 315 0.20 0.07 a 2.87 0.86 b 1.01 0.15 c 2.36 0.35 d Isorhamnetin 3-O-glucuronide 491 315 0.63 0.24 a 1.80 0.71 b 0.82 0.23 a,c 1.14 0.37 c Quercetin 301 0.00 0.00 a 0.50 0.30 b 0.30 0.11 c 0.87 0.45 d Kaempferol 285 0.00 0.00 a 0.35 0.29 b 0.53 0.18 c 0.52 0.29 b,c Phenolic acids Gallic acid 169 125 16.85 9.01 a 3.39 1.44 b 12.32 5.81 a 12.53 5.97 a Protocatechuic acid 153 109 2.08 0.53 a 0.55 0.33 b 1.07 0.36 c 1.00 0.46 c Caffeic acid 179 135 0.45 0.34 a 1.28 0.17 b 1.66 0.27 c 1.74 0.35 c Caftaric acid 311 179 1.96 0.59 a 1.82 0.31 a 7.30 2.13 b 9.99 3.74 c Fertaric acid 325 193 0.35 0.30 a 0.99 0.11 b 0.84 0.07 b,c 0.80 0.13 c cis-Coutaric acid 295 163 0.64 0.05 a 0.73 0.08 b 0.68 0.05 a,b 0.32 0.03 c trans-Coutaric acid 295 163 0.77 0.09 a 0.66 0.08 a 1.10 0.18 b 1.35 0.69 b 1Fragment ion detected in negative ion MS/MS. 2mg phenolic compound/100 g dry matter. Each value represents mean (n=3) SD. Values in the same row followed by different letters are significantly different by ANOVA test (p<0.05).