Phenolic Compounds Extraction in Enzymatic Macerations of Grape Skins Identified as Low-level Extractable Total Anthocyanin Content

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is the peer reviewed version of the following article: Nogales-Bueno J, Baca-Bocanegra B, Heredia FJ, Hernández-Hierro JM. Phenolic compounds extraction in enzymatic macerations of grape skins identified as low-level extractable total anthocyanin content. J Food Sci. 2020 Feb;85(2):324-331. doi: 10.1111/1750-3841.15006. Epub 2020 Jan 22. PMID: 31968392., which has been published in final form at https://doi.org/10.1111/1750-3841.15006 This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. This article may not be enhanced, enriched or otherwise transformed into a derivative work, without express permission from Wiley or by statutory rights under applicable legislation. Copyright notices must not be removed, obscured or modified. The article must be linked to Wiley’s version of record on Wiley Online Library and any embedding, framing or otherwise making available the article or pages thereof by third parties from platforms, services and websites other than Wiley Online Library must be prohibited." Phenolic Compounds Extraction in Enzymatic Macerations of Grape Skins Identified as Low 1 Level Extractable Total Anthocyanin Content 2 3 Julio Nogales-Bueno1,2, Berta Baca-Bocanegra1, Francisco José Heredia1* and José Miguel 4 Hernández-Hierro1. 5 6 1 Food Colour and Quality Laboratory, Área de Nutrición y Bromatología, Facultad de Farmacia, 7 Universidad de Sevilla, 41012 Sevilla, Spain. 8 9 2 Instituto de Ciências Agrárias e Ambientais Mediterrânicas, Universidade de Évora, 10 Departamento de Fitotecnia, Apartado 94 7002 – 554 Évora, Portugal. 11 12 13 14 * Corresponding author: Francisco José Heredia 15 Phone: +34 9545 56495 16 E-mail: [email protected] 17 18 19 20 Word count: 5473 words 21 22 23 Short version of title: Improve grape skin phenolic extraction 24 25 26 Journal/section: Journal of Food Science / Food Chemistry 27 28 ABSTRACT: Anthocyanins in wine principally depends on grape skin extractable anthocyanin 29 content, i.e., the amount of anthocyanins present in grape skin that are released to wine during 30 the maceration stage. This amount of extractable anthocyanins is closely linked to the cell wall 31 degradation of skin cells. Indeed, among other methodologies, the maceration in presence of 32 different enzymes can be used to increase cell wall degradation, and therefore, the amount of 33 anthocyanins extracted from grape skins to wine. 34 Vitis vinifera L. cv. Tempranillo and Syrah red grapes have been identified as samples with low 35 anthocyanin extraction potential by near infrared hyperspectral imaging. Grape skins have been 36 macerated in presence of cellulase, glucosidase and pectinase. Then, colour of the supernatants 37 and phenolic compounds extracted from grape skins (total phenols, total flavanols and total 38 and individual anthocyanins) have been determined. 39 Cellulase and glucosidase have shown a positive effect in the extraction of phenolic compounds 40 from these grapes. Macerations carried out in the presence of cellulase have produced 41 supernatants with a more intense colour (lower lightness and higher chroma values), and a 42 higher extraction of flavanols and anthocyanins than the respective control essays. However, 43 pectinase treatments have produced the opposite effect, which could be partially explained by 44 an eventual interaction between the cell wall polysaccharides liberated by pectinase and the 45 phenolic compounds extracted. 46 Synergy effects do not appear between cellulase and glucosidase. Moreover, the negative 47 effect of the addition of pectinase might be due to the interactions between the cell wall 48 material liberated by pectinase and the phenolic compounds extracted. 49 50 Keywords: Wine; grape skins; phenolic compounds; enzymes; extraction. 51 52 Practical Application: In the present study, grape samples with a low anthocyanin extraction 53 potential have been identified and these samples have been macerated in presence of different 54 enzymes. The applied enzymes were three of the most common enzymes that are applied in 55 the wine industry. Individual enzymes and mixtures have been applied to Syrah and 56 Tempranillo grape skin samples and the results have been compared to control macerations. 57 Knowledge in this topic will help the production of quality wines. 58 59 Introduction 60 Enzymes play an important role in the winemaking process. They usually come from grape 61 itself, the indigenous microbiota on the grape and the microorganisms present during 62 winemaking. Since the action of these enzymes is difficult to control in some cases and is not 63 enough in others, manufacturers traditionally have developed better adapted enzymatic 64 preparations with the aims of improving the juice extraction, must clarification, aroma freeing 65 and colour extraction and stabilization (Ribéreau-Gayon et al., 2006). 66 Regarding to the colour extraction and stabilization, it mainly depends on the quantity and type 67 of phenolic compounds extracted from grape skins to wine. Anthocyanins are the compounds 68 directly responsible of the colour of red wines. However, other phenolic compounds, such as 69 flavanols or phenolic acids, can modulate and stabilize wine colour (Boulton, 2001). Similarly, 70 phenolic compounds are also responsible for other wine attributes related to wine quality 71 (astringency, sourness, bitterness, etc.) or to their health properties (antioxidant activity, free 72 radical scavenging capacity, etc.) (Jackson, 2000; Martín Bueno et al., 2012). In consequence, 73 the effect that different enzymes or enzymatic preparation have on the phenolic compounds 74 extraction from grape skin is being widely studied. 75 Romero-Cascales, Fernández-Fernández, Ros-García, López-Roca, and Gómez-Plaza (2008) 76 characterized and quantified the activities of six different enzymatic preparations. They studied 77 the phenolic composition and wine colour in post-fermentative and 12-month Monastrell 78 wines. They obtained a general increase in the total phenol content and higher colour intensity 79 after 12 month of storage. In the case of glucosidase activity, which can degrade anthocyanins 80 by breaking down the linkage of the glucose to the anthocyanidin (side effect in enzymatic 81 preparations), these authors found a positive significant difference in the extraction of total 82 phenols and tannins using the suppliers dosage of enzymatic preparations. However, they did 83 not obtain any difference in colour parameters even taking into account that the aforesaid side 84 effect may cause anthocyanin’s degradation. Parley (1997) and Río Segade et al. (2015) also 85 found differences in colour and phenolic content respectively using different enzymatic 86 preparations. Benucci et al. (2017) evaluated the effectiveness of enzymes applied, individually 87 and in enzyme mixes, on the extraction of anthocyanins, oligomeric flavanols and polymeric 88 flavanols from withered grape skins during simulated maceration. They found that the 89 effectiveness of individual enzymes and mixes of enzymes in improving the extractability of 90 phenolic compounds were variety dependent. Finally, Castro-López, Gómez-Plaza, Ortega- 91 Regules, Lozada, and Bautista-Ortín (2016), Samoticha, Wojdyło, Chmielewska, Politowicz, and 92 Szumny (2017) and A. Bautista-Ortín, Ben Abdallah, Castro-López, Jiménez-Martínez, and 93 Gómez-Plaza (2016) studied the proanthocyanidin extraction in presence of maceration 94 enzymes. 95 In all these studies, enzymatic macerations were carried out taking into account grape samples 96 collected at veraison or in the harvest, i.e. grape samples in the same stage of maturity. 97 However, it is well known that there is heterogeneity in the content of polyphenols and 98 extractable polyphenols within grape samples with the same maturity degree (Kontoudakis et 99 al., 2011; Martinez-Sandoval et al., 2016; Nogales-Bueno, Baca-Bocanegra, Rodríguez-Pulido, 100 Heredia, & Hernández-Hierro, 2015; Zouid, Siret, Jourjon, Mehinagic, & Rolle, 2013). In 101 consequence, it is not possible to know if the abovementioned enzymatic macerations would 102 affect similarly to grape samples with different contents of extractable polyphenols (e.g. high 103 extractable content or low extractable content). Actually, it would be of interest to know if the 104 presence of enzymes during the maceration stage also increases the amount of polyphenols 105 extracted from grape samples of low extractable polyphenol contents. 106 In the present study, grape samples have been classified according to their content of 107 extractable polyphenols by means of near infrared hyperspectral imaging. Then, samples with a 108 low anthocyanin extraction potential have been selected and these samples have been 109 macerated in presence of different enzymes. The applied enzymes were cellulase, glucosidase 110 and pectinase, three of the most common enzymes that make up the commercial enzymatic 111 preparations that are applied in the wine industry. Individual enzymes and mixtures have been 112 applied to Syrah and Tempranillo grape skin samples and the results have been compared to 113 control macerations. Total phenols, flavanols, total and individual anthocyanins and colour have 114 been evaluated in the different model wine extractions. To the best of our knowledge, the 115 effect that different enzymes have in the extraction of phenolic compounds has not been 116 evaluated yet in grapes with a low anthocyanin extraction potential. 117 118 Materials and Methods 119 In order to facilitate the understanding of the entire procedure carried out in this study, a 120 schematic representation of this process is shown in Fig. 1. 121 Sample selection by near infrared hyperspectral imaging 122 Sample collection. 123 Two hundred V. vinifera L. cv. Tempranillo and Syrah red grape samples, one hundred per 124 variety, were collected from two vineyards located in the Condado de Huelva Designation of 125 Origin D.O. (Andalusia, Spain). Samples were collected on two different dates, when Syrah and 126 Tempranillo vineyards were respectively harvested (August 11 and 07, 2014). Due to the 127 heterogeneity of the grape within a grape cluster (Ribéreau-Gayon et al., 2006), single grapes 128 were collected from the top, middle and bottom of the cluster and from the sunlight and shade 129 side in order to obtain a representative sample set. The samples were refrigerated and 130 immediately transported to the laboratory. 131 Near infrared data acquisition and predicted extractable anthocyanin content 132 determination. 133 Near infrared hyperspectral images were acquired for each individual grape. Hyperspectral 134 imaging device is described in detail elsewhere in Nogales-Bueno, Hernández-Hierro, 135 Rodríguez-Pulido, and Heredia (2014). 136 Afterward, hyperspectral images were calibrated (a two point calibration) and segmented and 137 the average absorbance spectrum was calculated for each grape sample. Then, the partial least 138 square (PLS) prediction model developed for Nogales-Bueno et al. (2015) was applied to grape 139 spectra in order to predict the extractable anthocyanin content for each sample. This 140 methodology has already been applied with good results for predicting the extractable 141 anthocyanin contents in different studies (Baca-Bocanegra, Nogales-Bueno, Heredia, & 142 Hernández-Hierro, 2018; Nogales-Bueno, Baca-Bocanegra, Jara-Palacios, Hernández-Hierro, & 143 Heredia, 2017). Previously to the extractable anthocyanin content prediction, it was evaluated 144 if these spectral samples were within the spectral space where the calibration model can be 145 applied. For this purpose, a Principal Component Analysis (PCA) was applied to the spectral 146 matrix and samples were ordered according to their distance from the centre of the spectral 147 space. This distance was measured following the advice of the software developer (Infrasoft 148 International LLC, 2000). According to this advice, Mahalanobis distance (H) was calculated for 149 each sample and samples with a H distance from the spectral space greater than 3 were 150 removed from the data set (Garrido-Varo, Garcia-Olmo, & Fearn, 2019; Infrasoft International 151 LLC, 2000). 152 Sample selection. 153 Grape samples were sorted according to their predicted extractable anthocyanin content. In 154 this way, samples with a low capacity for the extraction of anthocyanins were identified 155 following the procedure described elsewhere in Nogales-Bueno, Baca-Bocanegra, Jara-Palacios, 156 et al. (2017) and they were used in the subsequent analysis. 157 Enzymatic treatments 158 Grape skins, from grapes identified as samples with a low anthocyanin extraction potential, as 159 described in section 2.1., were removed from the whole grapes, divided into two parts and 160 weighed. For each sample, a half skin was macerated in presence of enzyme (enzymatic 161 maceration) while the other half was macerated without enzyme (control maceration). 162 Enzymatic macerations were carried out in presence of three of the most common enzymes 163 that make up the commercial enzymatic preparations: cellulase (C), glucosidase (G) and 164 pectinase (P) (Fluka Biochemika, references 22178, 49291 and 17389 respectively). Enzymes 165 were applied individually and in mixtures of enzymes (CG, CP and GP). The activities for C, G 166 and P are respectively 0.8, 0.8 and 1.0 U mg-1 at their respective optimal conditions of pH and 167 temperature. Although these conditions are different from the pH and temperature typical in 168 the vinification processes, the macerations were carried out under vinification conditions (pH 169 colorimetric parameters and phenolic contents as dependent variables. The different enzymatic 298 treatments and the type of maceration were evaluated as factors. 299 The scores of the first two principal components after PCA performed on supernatant colour 300 and phenolic parameters of the grape skins are shown in Fig. 2a. Samples are plotted in the 301 space defined by the first and second principal components, which described 67.61% (PC1) and 302 13.31% (PC2) of the data variability. Apparently, there are no trends in this graph. The samples 303 processed in control or enzymatic macerations or with different enzymatic treatments seem to 304 be overlaid. Nevertheless, if only samples treated with cellulase and glucosidase are shown (Fig. 305 2b), a trend can be inferred. Samples macerated in presence of their respective enzyme appear 306 on the left of samples macerated without the enzyme. This trend is surely linked to the 307 differences observed in the experimental data (Tables 1, 2 and 3). Moreover, this trend is 308 mainly observed in the PC1 direction, thus, this principal component surely collects the major 309 part of the variability of the data linked to the different extraction rate observed. 310 The influence that each variable have in PC1 and PC2 are shown in the Fig. 2c. It can be seen 311 that the colorimetric variable L* is which have an effect more intense (positive) on PC1, 312 whereas a number of variables have a negative effect on this principal component. If graph is 313 zoomed in on this region (Fig. 2d), more details can be observed. Chroma, total and 314 monoglucoside anthocyanins, and different individual anthocyanins show a negative influence 315 on the PC1. Notably, most of anthocyanin monoglucosides and their sum (MTotal) and the three 316 cyaniding derivatives appear in this region. Therefore, samples with a low value of L*, high 317 value of C*ab and a high extraction of these anthocyanins show a negative PC1 score, they 318 appear on the left region of the Fig. 2a graph, and vice versa. 319 Discussion 320 Among the different enzymes tested, glucosidase and, to a larger extent, cellulase have a 321 positive influence on the extraction of phenolic compounds from grape skins previously 322 classified as samples with a low capacity for the extraction of anthocyanins. In this study, an 323 increase of approximately 120% in the extraction of phenolic compounds has been achieved in 324 the macerations in presence of cellulase. A result larger than those obtained in wine (110%) 325 (Romero-Cascales et al., 2008) and in grape skin (Benucci et al., 2017) were the total 326 anthocyanins and proanthocyanins were increased in a 10% in presence of cellulase activity. 327 This slight increase can be due to the fact that grapes used in the present study were previously 328 identified as grapes with low extraction potential. Therefore, the enzyme had a larger room for 329 improvement. 330 Samples extracted in presence of these enzymes show lower L* and higher C*ab values than 331 control extractions. These enzymes should have weakened the skin cell wall. The relationship 332 between cell wall degradation and phenolic extractability has been previously studied 333 (Hernández-Hierro et al., 2014; Nogales-Bueno, Baca-Bocanegra, Rooney, et al., 2017; Quijada- 334 Morín, Hernández-Hierro, Rivas-Gonzalo, & Escribano-Bailón, 2015). A number of previous 335 studies also obtain a degradation of the cell wall and an increase of phenolic compounds 336 extraction with the application of cellulase (Apolinar-Valiente, Romero-Cascales, Gómez-Plaza, 337 & Ros-García, 2016; A. Bautista-Ortín et al., 2016; Benucci et al., 2017; Castro-López et al., 338 2016) and glucosidase (Romero-Cascales et al., 2008) enzymes to grape samples. However, 339 mixtures of cellulase and glucosidase enzymes have not shown a synergy effect. It has been 340 found that synergetic effects are not always obtained in the enzymatic macerations of grape 341 skins (Benucci et al., 2017). 342 Another interesting result of this study is that pectinase does not show a positive influence on 343 the phenolic extraction from these samples. This enzyme seems to produce the opposite effect. 344 It is known that the degradation of the cell wall produces the liberation of different 345 polysaccharides and these polysaccharides present a high affinity for the proanthocyanidins 346 and may reduce the proanthocyanidin content in the solution (A. B. Bautista-Ortín, Cano- 347 Lechuga, Ruiz-García, & Gómez-Plaza, 2014; A. Bautista-Ortín et al., 2016; Bindon, Smith, & 348 Kennedy, 2010; Zhu, 2017). 349 350 Conclusion 351 In this study, different enzymes have been tested in order to know if they can modify the 352 amount of phenolic compounds extracted from grape skins previously identified as samples 353 with a low anthocyanin extraction potential. The chosen enzymes have been cellulase, 354 glucosidase and pectinase, three enzymes typically present in commercial enzymatic 355 preparations. Their individual and combined effects have been tested. 356 Cellulase has shown a positive effect in the extraction of phenolic compounds from these low- 357 anthocyanic-extraction grapes. Macerations carried out in the presence of this enzyme have 358 produced supernatants with lower L* and higher C*ab values than control macerations. 359 Moreover this enzyme also produced a higher extraction of proanthocyanidins and 360 anthocyanins than the respective control essays. To a lesser extent, glucosidase produced a 361 similar effect than cellulase, whereas their enzymatic mixture produced more inconsistent 362 results. 363 On the contrary, supernatants obtained from pectinase treatments have less colour (more L* 364 and less C*ab values) and extractable phenols than those obtained from controls. This effect 365 could be explained by an interaction between the cell wall material liberated by pectinase and 366 the phenolic compounds extracted. 367 Further studies should be developed in order to clarify several aspects. Firstly, the non- 368 appearance of a synergy effect between cellulase and glucosidase. Secondly, how to reduce the 369 interactions between the cell wall material liberated by pectinase and the phenolic compounds 370 extracted in grapes with low potential for the phenolic extraction. And thirdly, understanding 371 why the pectinase, an enzyme widely apply in oenology, produced this negative effect in grapes 372 with low potential for the extraction anthocyanic compounds. 373 Acknowledgments 374 This work was supported by the Spanish Ministerio de Economía y Competitividad [grant 375 number AGL2017-84793-C2]; and by Universidad de Sevilla [grant numbers VPPI-II.2, VPPI-II.4, 376 VIPPIT-2019-II.4]. The authors thank the technical staff of Biology Service [Servicios Generales 377 de Investigación (SGI), Universidad de Sevilla]. 378 379 Author Contributions (required) 380 Julio Nogales-Bueno developed the statistical analyses and drafted the manuscript. Berta Baca- 381 Bocanegra designed and executed the experimental part. Francisco José Heredia supervised 382 the study, results and manuscript. José Miguel Hernández-Hierro planned the study and 383 interpreted the results. 384 385 Nomenclature or Appendix (optional) 386 a* and b*, CIELAB colorimetric coordinates; C, cellulase; C*ab, CIELAB chroma; CG, enzymatic 387 mixture of cellulase and glucosidase; CP, enzymatic mixture of cellulase and pectinase; DMACA, 388 p-dimethylaminocinnamaldehyde; G, glucosidase; GP, enzymatic mixture of glucosidase and 389 pectinase; H, Mahalanobis distance; hab, CIELAB hue; L*, CIELAB lightness; MANOVA, 390 multivariate analysis of variance; P, pectinase; PC, principal component; PCA, principal 391 component analysis. 392 References 393 Apolinar-Valiente, Rafael, Romero-Cascales, Inmaculada, Gómez-Plaza, Encarna, & Ros- 394 García, José María. (2016). Degradation of Syrah and Cabernet Sauvignon grapes skin: 395 application of different enzymatic activities: a preliminary study. European Food 396 Research and Technology, 242(12), 2041-2049. doi: 10.1007/s00217-016-2702-4 397 Baca-Bocanegra, Berta, Nogales-Bueno, Julio, Heredia, Francisco José, & Hernández- 398 Hierro, José Miguel. (2018). Influence of oak wood chips–grape mix maceration on the 399 extraction of anthocyanins from low-extractable anthocyanin content red grapes. 400 European Food Research and Technology, 244(4), 729-734. doi: 10.1007/s00217-017- 401 2999-7 402 Bautista-Ortín, Ana Belén, Cano-Lechuga, Mario, Ruiz-García, Yolanda, & Gómez-Plaza, 403 Encarna. (2014). Interactions between grape skin cell wall material and commercial 404 enological tannins. Practical implications. Food Chemistry, 152(Supplement C), 558-565. 405 doi: https://doi.org/10.1016/j.foodchem.2013.12.009 406 Bautista-Ortín, Ana, Ben Abdallah, Rim, Castro-López, Liliana, Jiménez-Martínez, María, 407 & Gómez-Plaza, Encarna. (2016). Technological Implications of Modifying the Extent of 408 Cell Wall-Proanthocyanidin Interactions Using Enzymes. International Journal of 409 Molecular Sciences, 17(1), 123. 410 Benucci, I., Río Segade, S., Cerreti, M., Giacosa, S., Paissoni, M. A., Liburdi, K., . . . Rolle, L. 411 (2017). Application of enzyme preparations for extraction of berry skin phenolics in 412 withered winegrapes. Food Chemistry, 237(Supplement C), 756-765. doi: 413 https://doi.org/10.1016/j.foodchem.2017.06.003 414 Bindon, Keren A., Smith, Paul A., & Kennedy, James A. (2010). Interaction between 415 Grape-Derived Proanthocyanidins and Cell Wall Material. 1. Effect on Proanthocyanidin 416 Composition and Molecular Mass. Journal of Agricultural and Food Chemistry, 58(4), 417 2520-2528. doi: 10.1021/jf9037453 418 Boulton, Roger. (2001). The Copigmentation of Anthocyanins and Its Role in the Color of 419 Red Wine: A Critical Review. American Journal of Enology and Viticulture, 52(2), 67-87. 420 Castro-López, Liliana del Rocío, Gómez-Plaza, Encarna, Ortega-Regules, Ana, Lozada, 421 Daniel, & Bautista-Ortín, Ana Belén. (2016). Role of cell wall deconstructing enzymes in 422 the proanthocyanidin–cell wall adsorption–desorption phenomena. Food Chemistry, 423 196, 526-532. doi: http://dx.doi.org/10.1016/j.foodchem.2015.09.080 424 CIE. (2004). Technical Report: Colorimetry (3rd ed.). Vienna, Austria: Commission 425 Internationale de l’Eclairage Central Bureau. 426 Cozzolino, D., Kwiatkowski, M.J., Parker, M., Cynkar, W.U., Dambergs, R.G., Gishen, M., 427 & Herderich, M.J. (2004). Prediction of phenolic compounds in red wine fermentations 428 by visible and near infrared spectroscopy. Analytica Chimica Acta, 513(1), 73-80. 429 García-Marino, Matilde, Hernández-Hierro, José Miguel, Rivas-Gonzalo, Julián C., & 430 Escribano-Bailón, M.Teresa. (2010). Colour and pigment composition of red wines 431 obtained from co-maceration of Tempranillo and Graciano varieties. Analytica Chimica 432 Acta, 660(1-2), 134-142. 433 Garrido-Varo, A, Garcia-Olmo, J, & Fearn, T. (2019). A note on Mahalanobis and related 434 distance measures in WinISI and The Unscrambler. Journal of Near Infrared 435 Spectroscopy, 27(4), 253-258. doi: 10.1177/0967033519848296 436 Heredia, F. J., Álvarez, C., González-Miret, M. L., & Ramírez, A. (2004). Spain Patent No. 437 Hernández-Hierro, José Miguel, Nogales-Bueno, J., Rodríguez-Pulido, F.J., & Heredia, F.J. 438 (2013). Feasibility study on the use of near-infrared hyperspectral imaging for the 439 screening of anthocyanins in intact grapes during ripening. Journal of Agricultural and 440 Food Chemistry, 61(41), 9804-9809. 441 Hernández-Hierro, José Miguel, Quijada-Morín, Natalia, Martínez-Lapuente, Leticia, 442 Guadalupe, Zenaida, Ayestarán, Belén, Rivas-Gonzalo, Julián C., & Escribano-Bailón, 443 M.Teresa. (2014). Relationship between skin cell wall composition and anthocyanin 444 extractability of Vitis vinifera L. cv. Tempranillo at different grape ripeness degree. Food 445 Chemistry, 146(0), 41-47. doi: 10.1016/j.foodchem.2013.09.037 446 Infrasoft International LLC. (2000). WinISI II (Version 1.50.). Port. Matilda, PA, USA. 447 Jackson, Ron S. (2000). Chemical Constituents of Grapes and Wine. In R. S. Jackson (Ed.), 448 Wine science: principles, prectice and perception (pp. 232-280). San Diego, California: 449 Academic Press. 450 Janik, L.J., Cozzolino, D., Dambergs, R., Cynkar, W., & Gishen, M. (2007). The prediction 451 of total anthocyanin concentration in red-grape homogenates using visible-near- 452 infrared spectroscopy and artificial neural networks. Analytica Chimica Acta, 594(1), 453 107-118. 454 Kim, Y. J., Kim, D. O., Chun, K., Shin, D. H., Jung, H., Lee, C. Y., & Wilson, D. B. (2005). 455 Phenolic extraction from apple peel by cellulases from Thermobifida fusca. Journal of 456 Agricultural and Food Chemistry, 53(24), 9560-9565. doi: 10.1021/jf052052j 457 Kontoudakis, Nikolaos, Esteruelas, Mireia, Fort, Francesca, Canals, Joan Miquel, De 458 Freitas, Victor, & Zamora, Fernando. (2011). Influence of the heterogeneity of grape 459 phenolic maturity on wine composition and quality. Food Chemistry, 124(3), 767-774. 460 Martín Bueno, Julia, Ramos-Escudero, Fernando, Sáez-Plaza, Purificación, Muñoz, Ana 461 María, Navas, María José, & García Asuero, Agustin. (2012). Analysis and Antioxidant 462 Capacity of Anthocyanin Pigments. Part I: General Considerations Concerning 463 Polyphenols and Flavonoids. Critical Reviews in Analytical Chemistry, 42(2), 102-125. 464 Martinez-Sandoval, J. R., Nogales-Bueno, J., Rodriguez-Pulido, F. J., Hernandez-Hierro, J. 465 M., Segovia-Quintero, M. A., Martinez-Rosas, M. E., & Heredia, F. J. (2016). Screening of 466 anthocyanins in single red grapes using a non-destructive method based on the near 467 infrared hyperspectral technology and chemometrics. J Sci Food Agric, 96(5), 1643-1647. 468 doi: 10.1002/jsfa.7266 469 Nicolaï, Bart M., Beullens, Katrien, Bobelyn, Els, Peirs, Ann, Saeys, Wouter, Theron, 470 Karen I., & Lammertyn, Jeroen. (2007). Nondestructive measurement of fruit and 471 vegetable quality by means of NIR spectroscopy: A review. Postharvest Biology and 472 Technology, 46(2), 99-118. 473 Nogales-Bueno, Julio, Baca-Bocanegra, Berta, Jara-Palacios, María José, Hernández- 474 Hierro, José Miguel, & Heredia, Francisco José. (2017). Evaluation of the influence of 475 white grape seed extracts as copigment sources on the anthocyanin extraction from 476 grape skins previously classified by near infrared hyperspectral tools. Food Chemistry, 477 221, 1685-1690. doi: http://dx.doi.org/10.1016/j.foodchem.2016.10.118 478 Nogales-Bueno, Julio, Baca-Bocanegra, Berta, Rodríguez-Pulido, Francisco José , Heredia, 479 Francisco José , & Hernández-Hierro, José Miguel (2015). Use of near infrared 480 hyperspectral tools for the screening of extractable polyphenols in red grape skins. Food 481 Chemistry, 172, 559-564. 482 Nogales-Bueno, Julio, Baca-Bocanegra, Berta, Rooney, Abigail, Miguel Hernández-Hierro, 483 José, José Heredia, Francisco, & Byrne, Hugh J. (2017). Linking ATR-FTIR and Raman 484 features to phenolic extractability and other attributes in grape skin. Talanta, 485 167(Supplement C), 44-50. doi: https://doi.org/10.1016/j.talanta.2017.02.008 486 Nogales-Bueno, Julio, Hernández-Hierro, J.M., Rodríguez-Pulido, F.J., & Heredia, F.J. 487 (2014). Determination of technological maturity of grapes and total phenolic 488 compounds of grape skins in red and white cultivars during ripening by near infrared 489 hyperspectral image: A preliminary approach. Food Chemistry, 152, 586-591. 490 Parley, A. (1997). The effect of pre-fermentation enzyme maceration on extraction and 491 colour stability in Pinot noir wine. (Master of Applied Science), Lincoln University, Lincoln 492 (New Zealand). 493 Qadir, R., Anwar, F., Gilani, M.A., Zahoor, S., Misbah ur Rehman, M., & Mustaqeem, M. 494 (2019). RSM/ANN based optimized recovery of phenolics from mulberry leaves by 495 enzyme-assisted extraction. Czech Journal of Food Science, 37, 99-105. doi: 496 https://doi.org/10.17221/147/2018-CJFS 497 Quijada-Morín, Natalia, Hernández-Hierro, José Miguel, Rivas-Gonzalo, Julián C., & 498 Escribano-Bailón, M. Teresa. (2015). Extractability of Low Molecular Mass Flavanols and 499 Flavonols from Red Grape Skins. Relationship to Cell Wall Composition at Different 500 Ripeness Stages. Journal of Agricultural and Food Chemistry, 63(35), 7654-7662. doi: 501 10.1021/acs.jafc.5b00261 502 Ribéreau-Gayon, P., Dubourdieu, D., Doneche, B., Lonvaud, A., Glories, Y., Maujean, A., 503 & Branco, J.M. (2006). Handbook of Enology, The Microbiology of Wine and 504 Vinifications. West Sussex, England: J. Wiley & Sons. 505 Río Segade, Susana, Pace, Caterina, Torchio, Fabrizio, Giacosa, Simone, Gerbi, Vincenzo, 506 & Rolle, Luca. (2015). Impact of maceration enzymes on skin softening and relationship 507 with anthocyanin extraction in wine grapes with different anthocyanin profiles. Food 508 Research International, 71(Supplement C), 50-57. doi: 509 https://doi.org/10.1016/j.foodres.2015.02.012 510 Romero-Cascales, Inmaculada, Fernández-Fernández, José Ignacio, Ros-García, José 511 María, López-Roca, José María, & Gómez-Plaza, Encarna. (2008). Characterisation of the 512 Figures (graphs, charts, line drawings, photographs) 559 560 Fig. 1: Schematic representation of the entire process. Sample collection, prediction of extractable anthocyanin content by near 561 infrared hyperspectral imaging, selection of samples with a low anthocyanin extraction potential, enzymatic and control maceration 562 of grape skins and supernatant and statistical analyses. 563 564 565 Fig. 2: (a) Score plot of the first two principal components after PCA performed on supernatant colour and phenolic parameters of 566 the grape skins for all the enzymatic treatments. Non-filled markers correspond to control macerations whereas filled markers 567 correspond to enzymatic macerations. (b) Score plot represented only for C, G and CG enzymatic treatments. Arrows indicate the 568 trends produced by the presence of the enzyme in the enzymatic macerations. (c) Projection of the colorimetric and phenolic 569 variables on the principal component plane. (d) Detail of this projection where variables with a higher influence on PC1 have been 570 identified. Abbreviations in this figure: please, see Tables 1 and 3 foot notes. 571