article crtique

Studying the evolution of anthocyanin-derived pigments in a typical red wine of Southern Italy to assess its resistance to aging

Tiziana Dipalmo a, Pasquale Crupi a, b, *, Sandra Pati c, Maria Lisa Clodoveo a, Aldo Di Luccia c a Department of Agricultural and Environmental Science, University of Bari, Via Amendola 165/A, 70126 BA, Italy b Consiglio per la ricerca in agricoltura e per l'economia agraria (CREA), Unit!a di ricerca per l'uva da tavola e la vitivinicoltura in ambiente mediterraneo, Via Casamassima 148, 70010 Turi, BA, Italy c Department of the Sciences of Agriculture, Food and Environment, University of Foggia, Via Napoli 25, 71121 Foggia, Italy

a r t i c l e i n f o

Article history: Received 23 September 2015 Received in revised form 14 February 2016 Accepted 8 March 2016 Available online 10 March 2016

Keywords: Primitivo Wine aging Wine pigments HPLC-UV-ESI-MSn

a b s t r a c t

In this study, the pigments of a young and 2 years old Primitivo red wines were investigated by using a column chromatography purification/fractionation step followed by HPLC-UV-ESI-MSn. Compounds such as malvidin 3O-acetyl-4-vinyl-procyanidin, malvidin 3O-glucoside-di(epi)-catechin, and A-type peonidin and malvidin 3O-glucoside-(epi)catechin were identified in Primitivo for the first time; furthermore, this is a further evidence that malvidin 3O-glucoside-malvidin 3O-(p-coumaroyl)-glucoside and (malvidin 3O-glucoside)3 can survive to the wine aging. Overall, the found anthocyanin-derived pigments were formed by malvidin, peonidin, and petunidin, which were mainly involved in the formation of pyr- anonathocyanins derivatives, accounting for 49% of the total content in the aged wine. The relative amounts of compounds, such as flavanol-anthocyanin adducts, which gave blue hues to the samples decreased whilst the reddish or violet ethyledene-bridge favanol-anthocyanin adducts increased in the aged wine. The predominance of pyranoanthocyanins, characterized by known higher stability compared to ethylidene-bridge favanol-anthocyanin and anthocyanin-flavanol compounds, and maybe their early formation (within two years) could be responsible for the reported rapid change of Primitivo color into orange hues, but also it would be able to slow down the natural loss of structure which wine is subject to.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Primitivo is a red grape variety cultivated in Apulia region (Southern Italy), often resulting in a tannic, ruby-purple wine with spicy and red-fruit flavor. Primitivo is genetically identical to Cali- fornia Zinfandel (Bowers, Bandman, & Meredith, 1993) and Croa- tian Crljenak Ka"stelanski and constitutes one parent of Croatian Plavac mali (Maleti#c et al. 2004).

Anthocyanins are the main compounds involved in the color of black and red grapes; in Vitis vinifera L. the monoglucosides of delphinidin, cyanidin, petunidin, peonidin, and malvidin but also their acetyl-, p-coumaroyl-, and caffeoyl derivatives are present in

quantities varying among varieties, climate, and viticulture condi- tions (Crupi et al., 2012; Coletta et al., 2013). At winemaking, they are partially extracted to the must becoming responsible for the color of young red wine (Rib#ereau-Gayon, Dubourdieu, Don!eche, & Lonvaud, 2006); then, during wine maturation and aging, their levels decrease with time, since they begin to react with other wine constituents leading to the formation of more stable pigments that are responsible for color changes as well as loss in astringency (Sun & Spranger, 2005).

Anthocyanins can condense with flavan-3-ols, as demonstrated in wines and model solutions, either directly (Salas et al., 2004; Hayasaka & Kennedy, 2003; Sun, Fernandes, & Spranger, 2010) or mediated by acetaldehyde (Atasanova, Fulckrand, Cheynier, & Moutounet, 2002; He et al., 2012) or other compounds (Pissarra et al., 2004). Moreover, they can give rise to cyclo-addition reac- tion at the carbon in position 4 and the hydroxyl group in position 5 with pyruvic acid (Fulcrand, Benabdeljalil, Rigaud, Cheynier, & Moutounet, 1998; Romero & Bakker, 1999), acetaldehyde (Bakker

* Corresponding author. Consiglio per la ricerca in agricoltura e per l'economia agraria (CREA), Unit!a di ricerca per l'uva da tavola e la vitivinicoltura in ambiente mediterraneo, Via Casamassima 148, 70010 Turi, BA, Italy.

E-mail addresses: [email protected], [email protected] (P. Crupi).

Contents lists available at ScienceDirect

LWT - Food Science and Technology

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

http://dx.doi.org/10.1016/j.lwt.2016.03.012 0023-6438/© 2016 Elsevier Ltd. All rights reserved.

LWT - Food Science and Technology 71 (2016) 1e9

& Timberlake, 1997), vinylphenol derivatives (Fulcrand, Dos Santos, Sarni-Manchado, Cheynier, & Favre-Bonvin, 1996; Hakånsson, Pardon, Hayasaka, De Sa, & Herderich, 2003; Hayasaka & Asen- storfer, 2002), and vinylflavanols (Mateus, Silva, Santos-Buelga, Rivas-Gonzalo, & De Freitas, 2002; Mateus et al., 2003), origi- nating pyranoanthocyanins, which have been detected in red wine (Pati et al., 2006; Alcade-Eon, Escribano-Bail#on, Santos-Buelga, & Rivas-Gonzalo, 2006).

Because of the large number of anthocyanin-derived pigments and the low levels of some of them, especially in old red wines (Alcade-Eon et al., 2006), their analysis generally requires pre- liminary fractionation/purification steps (i.e. solid-phase extraction or column chromatography) prior to off-line chromatographic separation and identification (Gergely et al., 2000; Alcade-Eon, Escribano-Bail#on, Santos-Buelga, & Rivas-Gonzalo, 2007; Sun et al., 2010). Moreover, the attempt of overcoming the possible drawbacks (mainly time-consuming, hydrolysis phenomena, and irreversible absorption) of the aforementioned fractionation/puri- fication methods by direct injection onto HPLC systems of raw red wine diluted samples without any pre-treatment has been limited by the difficulty of identifying and separating some minor pigments in aged wines from more concentrated compounds with which they coexist (Pati et al., 2006; De Villers, Vanhoenacker, Majek, & Sandra, 2004). This is particularly hard among the anthocyanins and their corresponding type A vitisins, which in the majority of the reversed phase LC methods elute very close together (Alcade-Eon, Escribano-Bail#on, Santos-Buelga, & Rivas-Gonzalo, 2004; Rentzsch, Schwartz, Winterhalter, Blanco-Vega, & Hermosin- Guti#errez, 2010).

High performance liquid chromatography (HPLC) coupled to UV and mass spectrometry (MS) detections have provided decisive information for the identification of a large number of pigment families based on their spectroscopic features and fragmentation patterns (Alcade-Eon et al., 2007; Pati et al., 2006; Sun et al., 2010; He et al., 2012). However, in spite of the numerous wine poly- phenols nowadays identified, the influence of a long aging period on the presence of oligomeric and polymeric pigments in Primitivo wine has been still few investigated.

Therefore, the goal of this study was to characterize the evolution of the anthocyanin-derived pigments present in a young and 2 years old Primitivo red wines by using a column chromatography purifi- cation/fractionation step followed by HPLC-UV-ESI-MSn. Grouping this information could get new insight about the low color stability of this wine, a consistent problem still facing the wine makers, in order to better define the effective potential to age of Primitivo.

2. Materials and methods

2.1. Wine samples

2005 vintage grapes from a Primitivo 25 year old vineyard in Manduria territory (Taranto, Southern Italy), were picked at the technological maturity and processed according to the following winemaking procedure. About 300 kg of grapes (sugar content 19.3 !Brix, pH 3.35, total acidity 7.3 g L"1 of tartaric acid equivalents) previously destemmed, were crushed and collected in stainless steel 100-Kg-capacity tanks (three replicates). Before starting the fermentation, the crushed grapes were supplied with potassium metabisulphite (15 g 100 kg"1), dry yeasts (15 g 100 kg"1, Saccha- romyces cerevisiae, Zymasil, AEB, Brescia, Italy), and yeast activator (25 g 100 kg"1, Bioact Plus, Oliver Ogar, Montebello Vicentino, Italy). During the alcoholic fermentation (10 days long), two daily gentle punching down of the pulp cap into the fermenting juice were made. After running-off and light pressure of the marcs, the malolactic fermentation occurred spontaneously; afterwards, the

wine was decanted and a second aliquot of potassium meta- bisulphite (15 g 100 kg"1) was added. Finally, after another decantation performed in March, the wine was bottled in 1 L glass bottles and stored in the dark under controlled conditions of temperature (15e18 !C), for 2 years.

Red wine was analyzed at the end of fermentations and after 2 years.

2.2. Chemicals

Trifluoroacetic acid, ethanol RPE, and HPLC grade water were purchased from J.T. Baker (Deventer - Holland). LC-MS grade sol- vent acetonitrile was purchased from Riedel-de Ha€en (Steinheim - Germany). Delphinidin, cyanidin, petunidin, peonidin, and malvi- din 3-O-glucoside chloride was purchased from Extrasynthese (Genay e France).

2.3. Wine fractionation

For fractionation of the aged wine, 10 mL of sample was placed onto a 230 # 15 mm i.d. Toyopearl HW-40(S) gel column (Tosoh, Japan) which was then passed with 150 mL distilled water to wash out phenolic acids. The elution solvent was ethanol/0.1% aqueous trifluoroacetic acid (60:40, v/v) by which all the pigments retained in the column were eluted and collected in 10 fractions of 5 mL. Flow rate was regulated at 0.2 mL min"1 using a peristaltic pump. Finally, since there was no interest in correlating the fractions composition to sensorial characteristic such as in other works (Weber, Greve, Durner, Fischer, & Winterhalter, 2013; Wollmann & Hofmann, 2013), for reducing the analysis time the fractions 1 and 2 (A), 3e6 (B), 7e10 (C) were grouped together, respectively, on the basis of their UVeVis maxima wavelengths (lmax) registered in the range 190e700 nm and of their chromatographic behavior.

The young wine was analyzed without fractionation as reported by Pati et al. (2006).

2.4. HPLC-UV-ESI-MS analysis

As previously described (Pati et al., 2006), separation and identification of anthocyanin-derived pigments were carried out by means of an HPLC-UV-ESI-MSn system consisting of a 600-MS multi solvent delivery pump (Waters, Milford, MA, USA), an UVeVis de- tector (model G1315B DAD system, 1100 series Agilent Technolo- gies, Palo Alto, CA, U.S.A.) and an LCQ ion trap mass spectrometer (Thermo Electron Co., San Jose, CA, USA).

Briefly, after filtration through 0.45 mm pore size regenerated cellulose filters (VWR International, USA) wine diluted samples (1:2, v/v with water) and/or the separated fractions were injected onto a reversed stationary phase column, Luna C18 (150 # 2 mm i.d., particle size 3 mm, Phenomenex, USA) protected by a C18 Guard Cartridge (4.0 # 2.0 mm i.d., Phenomenex). HPLC separation was accomplished using a binary mobile phase composed of (solvent A) water/acetonitrile (95:5, v/v) and (solvent B) water/acetonitrile (10:90, v/v), both containing 0.1% (v/v) trifluoroacetic acid. The following gradient was adopted: isocratic 2% B for 2 min; from 2 to 10% B for 6 min; from 10 to 13% B for 22 min; from 13 to 20% B for 20 min; from 20 to 30% B for 25 min; followed by washing and re- equilibrating the column. The column temperature was not controlled and the flow was maintained at 0.2 mL min"1. UVeVis detection wavelength was set at 520 nm. HPLC-ESI-MS data were acquired under positive ion mode using Xcalibur (Thermoquest) software. In the preliminary stage, an optimization of the ESI interface and ion optics parameters was accomplished in order to maximize the signal of malvidin 3-O-glucoside ion (m/z 493), chosen as the reference compound. To this aim, a 40-ppm solution

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e92

of malvidin 3-O-glucoside in solvent A was infused (at a flow rate of 20 mL/min) in the ESI interface using the syringe pump incorpo- rated into the mass spectrometer. The optimized electrospray/ion optics parameters were as follows: spray voltage, 4.5 kV; sheath gas (nitrogen), 0.9 L min"1; capillary voltage, 35.0 V; capillary tem- perature, 200 !C; tube lens offset voltage, 15.0 V.

Typically, three runs were performed during HPLC-ESI-MS analysis of each sample. First, an MS full-scan acquisition (m/z range 50e2000) was performed to obtain preliminary information on the predominant m/z ratios observed during the elution. Subse- quently, MS2 spectra were acquired by fragmenting some precursor ions selected in the MS spectra. Tentative compound identification was achieved by combining different information: UV absorption, capacity factors (k'), and mass spectra (MS and MS2) which were compared with those from pure standards, when available, and interpreted with the help of structural models already hypothesized in the literature (Alcade-Eon et al., 2004; Alcade-Eon et al., 2007; Pati et al., 2006). Semi-quantitation was performed using extracted ion chromatograms (EIC): for each compound, the EIC at the corre- sponding molecular ion was obtained and the relevant peak was integrated; subsequently, peak areas were summed with respect to the type of pigment to calculate the percentage content of the different classes determined in the wines.

3. Results and discussion

3.1. HPLC-UV-MS analysis of raw wine

Fig. 1 shows the chromatogram of a young and 2 years old

Primitivo raw wine samples diluted (1:2, v/v) with water. Unlike the young Primitivo wine (Fig. 1a), it was difficult to achieve a total identification of the compounds present in the older wine (Fig. 1b). This drawback is very frequent in ESI-MS analysis of minor con- stituents in complex matrix (like aged wine) and is probably caused by competitive ionization between the low concentrated analytes of interest and the matrix. Moreover, as it usually happens for aged wines (Wang, Race, & Shrikhande, 2003; Alcade-Eon et al., 2004; Blanco-Vega, G#omez-Alonso, & Hermosín-Guti#errez, 2014), the “hump” under the chromatogram indicated the presence of a lot of not-well resolved peaks corresponding to co-eluting species. Only the identity of 7 pigments could be tentatively assigned by means of their k' and molecular ions [M]þ in their mass spectra (Table 1). Therefore, a fractionation of this wine was needed to obtain a more complete analysis of the pigments present, by eliminating those compounds responsible of the “hump” in the chromatogram of the raw aged wine (Fig. 1b).

In the elution of the wine through the Toyopearl HW-S gel column with ethanol/0.1% aqueous trifluoroacetic acid (60:40, v/v), different colored bands were formed. As previously stated, the el- uates were collected in 10 fractions, which were further grouped in 3 fractions according to their UVeVis spectra and chromatographic behavior in order to simplify the analysis. Indeed, preliminary HPLC-UV-Vis analyses of all the eluates were carried out in order to determine the chemical nature of the major components of each; the eluates 1 and 2, 3e6, and 7e10, respectively, showed similar chromatograms, with peaks that had the same retention times and UVeVis spectra in all cases, only differing in the proportions in which they were present in the different eluates. Thus, “fraction A” was formed by the combination of eluates 1 and 2 and was mainly characterized by anthocyanidins monoglucosides and different types of pyranoanthocyanins, whilst “fraction B” and “fraction C” contained prevalently vinyl-linked anthocyanin-flavonol and ethyl- linked anthocyanin-flavonol, were obtained grouping together the eluates 3e6 and 7e10, respectively (Table 1, Fig. 2).

3.2. HPLC-UV-MS analysis of young wine and aged wine fractions

Fig. 3 shows the UVeVis chromatograms at 520 nm of fractions A-C of the 2 years old Primitivo wine, in which more than 50 compounds were identified. Their k' and ions from MS and MS/MS analyses were reported in Table 1. Peaks 8, 10, and 12 were iden- tified as the monoglucosides of petunidin, peonidin, and malvidin, respectively, as reported elsewhere (Pati et al., 2006); likewise, peaks 22, 24, and 27 were attributed to malvidin 3O-acetylgluco- side, malvidin 3O-caffeoylglucoside, and mavidin 3O-p-coumar- oylglucoside. As shown by the fragmentation pattern equivalent to compound 27, another malvidin 3O-p-coumaroylglucoside was also observed into the fraction B (peak 33, Fig. 3b), thus, on the basis of their elution order, it was assumed that those molecules (peaks 27 and 33) were cis- and trans-isomers (Crupi et al., 2012). It is worth pointing that the afore mentioned compounds were found in both wines, whilst other acyl anthocyanins, such as peonidin 3O-ace- tylglucoside (peak 20) and cyanidin, petunidin, and peonidin 3O- (p-coumaroyl)-glucoside (peaks 25, 26, and 32), as well as delphi- nidin and cyanidin glucosides (peaks 3 and 5) were only detected in the young wine (Fig. 1a, Table 1).

Five compounds corresponding to carboxy-pyranoanthocyanins derived from the reaction between glucosides anthocyanidins and pyruvic acid (A-type vitisins, Fig. 2b) were also identified in both wines. In particular, peaks 9, 11, and 13 were attributed to petuni- din, peonidin, and malvidin 3O-glucoside pyruvic derivatives because their molecular ions in the MS analysis (m/z 547, 531, and 561, respectively) were fragmented into aglycones (m/z 385, 369, and 399, respectively) by losing a dehydrated-glucose moiety

Fig. 1. HPLC-UV-Vis chromatogram of the raw a) young and b) 2 years old red wine (Primitivo cv.) recorded at 520 nm.

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e9 3

Table 1 Chromatographic and mass spectral data of the compounds identified into Primitivo young wine and 2 year old wine fractions.

Peak k' Compounda [M]þ (m/z) MS/MS Sample

Young Old

Fraction A Free anthocyanins 3 5.83 Delphinidin 3O-glucosideb 465 303 * n.d. 5 6.71 Cyanidin 3O-glucosideb 449 287 * n.d. 8 7.50 Petunidin 3O-glucosideb 479 317 * * 10 9.04 Peonidin 3O-glucosideb 463 301 * * 12 9.90 Malvidin 3O-glucosideb 493 331 * * 20 19.04 Peonidin 3O-acetylglucosidec 505 301 * n.d. 22 19.50 Malvidin 3O-acetylglucosidec 535 331 * * 24 21.78 Malvidin 3O-caffeoylglucosidec,i 655 331 * * 25 21.90 Cyanidin 3O-(p-coumaroyl)-glucosidec,i 595 287 * n.d. 26 22.76 Petunidin 3O-(p-coumaroyl)-glucosidec,i 625 317 * n.d. 27 22.83 Malvidin 3O-cis-(p-coumaroyl)-glucosided,i 639 331 * * 32 24.63 Peonidin 3O-(p-coumaroyl)-glucosidec,i 609 301 * n.d. A-type vitisins 9 7.67 Petunidin 3O-glucoside pyruvic derivativee 547 385 * * 11 9.78 Peonidin 3O-glucoside pyruvic derivativee 531 369 * * 13 10.57 Malvidin 3O-glucoside pyruvic derivativee 561 399 * * 19 18.53 Peonidin 3O-(p-coumaroyl)-glucoside pyruvic derivativee,i 677 369 * * 21 19.06 Malvidin 3O-(p-coumaroyl)-glucoside pyruvic derivativee,i 707 399 * * Hydroxyphenyl pyranoanthocyanins 41 24.03 Petunidin 3O-glucoside-4-vinyl-phenole 595 433 n.d. * 43 24.61 Peonidin 3O-glucoside-4-vinyl-guaiacole 609 447 n.d. * 44 24.65 Malvidin 3O-glucoside-4-vinyl-guaiacole 639 477 n.d. * 45 24.71 Peonidin 3O-glucoside-4-vinyl-catechole 595 433 n.d. * 36 25.88 Malvidin 3O-glucoside-4-vinyl-catechole 625 463 * * 37 26.11 Peonidin 3O-glucoside-4-vinyl-phenole 579 417 * * 39 26.71 Malvidin 3O-glucoside-4-vinyl-phenole 609 447 * * Flavanol pyranoanthocyanins 23 21.04 Malvidin 3O-acetylglucoside-4-vinyl-(epi)catechinc 847 643,491 * * 30 23.79 Malvidin 3O-glucoside-4-vinyl-(epi)catechinc 805 643,491 * * 34 25.19 Malvidin 3O-glucoside-4-vinyl-(epi)catechinc 805 643,491 * * 8,8-Ethylidene Flavanol-Anthocyanin adducts 14 15.63 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 * * 16 17.08 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 * * 18 17.94 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 * * 29 23.50 Peonidin 3O-(p-coumaroyl)-glucoside-8-ethyl-(epi)catechinc,e,i 925 635,617,327 * * 31 24.21 Malvidin 3O-(p-coumaroyl)-glucoside-8-ethyl-(epi)catechinc,e,i 955 665,647,357 * * 40 27.95 Malvidin 3O-(p-coumaroyl)-glucoside-8-ethyl-(epi)catechinc,e,i 955 665,647,357 * * Flavanol-anthocyanin adducts 1 5.29 (epi)catechin-peonidin 3O-glucosidec 751 589, 463, 437 * n.d. 2 5.50 (epi)catechin-malvidin 3O-glucosidec 781 619, 493, 467 * n.d.

5.88 di(epi)catechin-malvidin 3O-glucosidec 1069 907, 781, 619 * n.d. 6 7.21 (epi)catechin-malvidin 3O-glucosidec 781 619, 493, 467 * n.d. 7 7.25 di(epi)catechin-malvidin 3O-glucosidec 1069 907, 781, 619 * n.d. Anthocyanin dimers and trimers 62 7.5 (Malvidin glucoside)3

h 1477 1315, 1153, 822 * * 42 24.08 Malvidin 3O-glucoside malvidin 3O-(p-coumaroyl)-glucosidee 1131 969,823,661 * * Fraction B Anthocyanin-tannin derivative 46 8.09 Malvidin 3O-glucoside-(epi)cateching 783 631,621,469 n.d. * 47 9.39 Peonidin 3O-glucoside-(epi)cateching 753 601,591,439 n.d. * Flavanol pyranoanthocyanins 15 17.00 Malvidin 3O-glucoside-4-vinyl-di(epi)catechinc 1093 931,803,641 * * 17 17.75 Malvidin 3O-glucoside-4-vinyl-di(epi)catechinc 1093 931,803,641 * * 48 18.53 Malvidin 3O-acetylglucoside-4-vinyl-di(epi)catechinf 1135 931,845 n.d. * 49 23.50 Petunidin 3O-glucoside-4-vinyl-(epi)catechin 791 629,477 n.d. * 50 25.34 Peonidin 3O-glucoside-4-vinyl-(epi)catechin 775 613,461 n.d. * 35 25.82 Malvidin 3O-(p-coumaroyl)-glucoside-4-vinyl-catechinc,i 951 643,491 * * 38 26.50 Malvidin 3O-(p-coumaroyl)-glucoside-4-vinyl-(epi)catechinc,i 951 643,491 * * Free Anthocyanins 27 22.76 Malvidin 3O-cis-(p-coumaroyl)-glucosided,i 639 331 * * 33 25.00 Malvidin 3O-trans-(p-coumaroyl)-glucosided,i 639 331 * * Fraction C Anthocyanin-Flavanol adducts 51 10.80 Malvidin 3O-glucoside-di(epi)catechinf 1071 919,909,781 n.d. * 8,8-Ethylidene Flavanol-Anthocyanin adducts 53 14.96 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 n.d. * 54 16.41 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 n.d. * 55 17.27 Malvidin 3O-glucoside-8-ethyl-(epi)catechinc,e 809 647,519,357 n.d. * 59 24.49 Malvidin 3O-(p-coumaroyl)-glucoside 8-ethyl-(epi)catechinc,e,i 955 665,357 n.d. * 60 27.14 Malvidin 3O-(p-coumaroyl)-glucoside 8-ethyl-(epi)catechinc,e,i 955 665,357 n.d. * 61 27.45 Malvidin 3O-(p-coumaroyl)-glucoside 8-ethyl-(epi)catechinc,e,i 955 665,357 n.d. *

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e94

(Table 1). While compounds 19 and 21 can be proposed as peonidin and malvidin 3O-p-coumaroyl-glucoside pyruvic derivatives, since they had molecular ions (m/z 677 and 707) 146 u heavier respect to peaks 11 and 13 (Pati et al., 2006; Alcade-Eon et al., 2004; Heier, Blaas, Drob, & Wittkowski, 2002). Acetylated A-type vitisin was absent in according with what reported in model solutions (Romero & Bakker, 1999) and in wine (Alcade-Eon et al., 2006).

Similarly to other red wines (Hayasaka & Asenstorfer, 2002), seven hydroxyphenyl-pyranoanthocyanins (pinotins) originated from the cycloaddition between anthocyanidins glucosides and

vinylphenol, vinylcatechol, or vinylguaiacol were revealed (Fig. 2c). By matching the information from their MS intense signals (m/z 595, 579, and 609) and MS/MS product ions (m/z 433, 417, and 447), together with their elution order to those present in the literature (Alcade-Eon et al., 2004), peaks 41, 37, and 39 were assigned to petunidin, peonidin, and malvidin 3O-glucoside-4-vinyl-phenol, respectively. Then, because of the appearance in their mass spectra of molecular ions and aglycones of 16 and 30 u heavier than compounds 37 and 39 (maybe due to an additional hydroxyl or methoxyl group in the vinylphenol structure) and considering the

Table 1 (continued )

Peak k' Compounda [M]þ (m/z) MS/MS Sample

Young Old

Flavanol-pyranoanthocyanins 52 14.02 Malvidin 3O-glucoside-4-vinyl-tri(epi)catechinc 1381 1219,803 n.d. * 56 21.03 Malvidin 3O-(caffeoyl)-glucoside 4-vinyl-di(epi)catechinc,i 1255 965,931,641 n.d. * 28 23.08 Malvidin 3O-(p-coumaroyl)-glucoside 4-vinyl-di(epi)catechinc,i 1239 931,641 * * 57 23.51 Malvidin 3O-(p-coumaroyl)-glucoside 4-vinyl-di(epi)catechinc,i 1239 931,641 n.d. * 58 23.83 Malvidin 3O-(p-coumaroyl)-glucoside 4-vinyl-di(epi)catechinc,i 1239 931,641 n.d. *

k’: capacity factor values; Mþ: molecular ions; MS/MS: product ions; *: detected; n.d.: not detected. a Detailed structures are reported as Supplementary Material (Figure S1 e S5). b Identified by comparison with standard compounds. c Identified by comparison with literature data (Pati et al., 2006; Asenstorfer et al., 2001). d Identified by comparison with literature data (Crupi et al., 2012). e Identified by comparison with literature data (Alcade-Eon et al., 2004; Heier et al., 2002). f Identified by comparison with literature data (Blanco-Vega et al., 2014). g Identified by comparison with literature data (Duen~as et al., 2006). h Identified by comparison with literature data (Oliveira, da Silva, Parola, Mateus, Br#as, Jo~ao Ramos, & de Freitas, 2013). i The characteristic absorption maximum around 320 nm in the UVeVis spectra confirmed the presence of coumaroyl or caffeoyl moieties in the molecules.

Fig. 2. Structures of the main anthocyanins and anthocyanin-derived red wine pigments found in Primitivo: a) grape native anthocyanins; b) A-type vitisins; c) hydroxyphenyl- pyranoanthocyanins or pinotins (R3 ¼ H, phenol-pyranoanthocyanins; R3 ¼ OH, catechol-pyranoanthocyanins; R3 ¼ OCH3, guaiacol-pyranoanthocyanins); d) (4 / 8)-anthocyanin- flavanol adducts (R3 ¼ H or (epi)catechin); e) 8,8-ethylidene-flavanol-anthocyanin adducts (R3 ¼ H or (epi)catechin); f) 10-flavanol-pyranoanthocyanins (R3 ¼ H or (epi)catechin). For all structures: R ¼ glucose, p-coumaroyl, or caffeoyl glucose; R1 ¼ H and R2 ¼ OH, OCH3, R1 ¼ OH and R2 ¼ OH, OCH3, or R1 and R2 ¼ OCH3.

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e9 5

elution order which respect the decreasing polarity, the peaks 43, 44 and the peaks 45, 36 were identified as pyranoanthocyanins formed by reaction of peonidin and malvidin 3O-glucoside with vinylguaiacol and vinylcatechol, respectively (Table 1). It is worth noting that the compounds 36, 37, and 39 were found into both wines, while the other pinotins were only present into the older sample (Figs. 1a and 3a).

Some well resolved chromatographic peaks referring to isobaric ions with similar MS/MS spectra were achieved for species with [M]þ at m/z 805 and 809 (Figs. 1a and 3a, Table 1), which are vinyl- linked flavanol anthocyanin (Fig. 2f) and ethyliden-bridged flavanol anthocyanin (Fig. 2e) pigments, respectively. In particular, the MS/ MS fragmentation of the two molecular ions at m/z 805 gave rise to the base peak corresponding to the aglycon ion (loss of 162 u) which was consecutively fragmented to m/z 491 (loss of 152 u) arising from a RDA cleavage, indicative for the presence of catechin or epicatechin unit in the compound. Then, peaks 30 and 34 could be attributed to malvidin 3O-glucoside-4-vinyl-(epi)catechin iso- mers due to the stereoisomery of catechin/epicatechin attached to the C10 position. On the basis of its fragmentation pattern, in which the dominant fragment was originated by the loss of a dehydrated- acetylglucoside [M-204]þ from the [M]þ at m/z 847 (Table 1), also

compound 23 was imputed to a vinyl-linked anthocyanin flavanol pigment, namely malvidin 3O-acetylglucoside-4-vinyl-(epi)cate- chin. Fraction B revealed the presence of other vinyl-linked flavanol anthocyanins, generating from the reaction between malvidin coumaroyl-glucosides and vinylcatechin and vinylepicatechin (peaks 35 and 38), also observed into Primitivo young wine (Fig. 1a). Indeed, the molecular ion at m/z 951, and fragments at m/z 643 and 491, respectively, originating from the consecutive loss of a coumaroyl-glucose moiety [M-308]þ and 152 u by retro Diels-Alder (RDA) fission of an heterocyclic ring of a catechin or epicatechin unit, were detected. Similar observation about the MS fragmenta- tion patterns allowed to attribute the peaks 49 and 50, with mo- lecular ions at m/z 791 and 775, to petunidin 3O-glucoside-4- vinyl(epi)catechin and peonidin 3O-glucoside-4-vinyl(epi)cate- chin (Table 1); these compounds, only revealed into the older sample, have been never reported in Primitivo wine. Showing the same [M]þ at m/z 1093 and the same fragmentation with the ions at m/z 931 and 803 due to the loss of 162 u (dehydrated-glucose moiety) and 290 u by the elimination of an (epi)catechin molecule, peaks 15 and 17 can be imputed to two isomers of malvidin 3O- glucoside-4-vinyl-procyanidin; analogously, in the MS/MS spec- trum of peak 48, the elimination of 162 and 290 u from its mo- lecular ion at m/z 1135, which is 42 u higher than 1093, allowed to name it malvidin 3O-acetylglucoside-4-vinyl-procyanidin. It is very interesting that the latter compound was identified in Primitivo for the first time, unlike the other two procyanidin-anthocyanins iso- mers were previously found into a year old wine (Pati et al., 2006).

Also the highest peaks in the chromatogram of fraction C (peaks 28, 52, 56e58, Fig. 3c) can be attributed to vinyl-linked pigments, namely to three isomers of the trimer malvidin 3O-(p-coumaroyl)- glucoside-4-vinyl-di(epi)catechin and to the trimer malvidin 3O- (caffeoyl)-glucoside-4-vinyl-di(epi)catechin and the tetramer malvidin 3O-glucoside-4-vinyl-tri(epi)catechin, as confirmed by their molecular ion at m/z 1239,1255, and 1381, respectively, and by the presence of dominant fragments corresponding to the aglycon ions (loss of 308, 324, and 162 u, respectively) together with the common loss of terminal flavanol moieties (at 290 and 578 u) in their fragmentation patterns (Pati et al., 2006; Asenstorfer, Hayasaka, & Jones, 2001).

The MS/MS pattern related to the three molecular ions at m/z 809 (peaks 14, 16, 18, Figs. 1a and 3a) showed three main fragment ions at m/z 647 (loss of a dehydrated-glucose residue), 519 (loss of an (epi)catechin unit), and 357 (consecutive loss of dehydrated glucose and (epi)catechin) (Table 1); these data confirmed the identification of the pigments as three isomers of malvidin 3O- glucoside-8-ethyl-(epi)catechin (Pati et al., 2006). By matching their MS and MS/MS patterns with literature, peaks 29, 31, and 40 were identified as peonidin 3O-(p-coumaroyl)-glucoside-8-ethyl- (epi)catechin and two malvidin 3O-(p-coumaroyl)-glucoside-8- ethyl-(epi)catechin isomers, respectively (Figs. 1a and 3a, Table 1) (Pati et al., 2006; Alcade-Eon et al., 2004). Four others stereoiso- mers of malvidin 3O-glucoside-8-ethyl-(epi)catechin (peaks 53 and 54) and malvidin 3O-(p-coumaroyl)-glucoside-8-ethyl-(epi)cate- chin (peaks 60 and 61), having different k' respect to those into faction A and young wine, were only detected into the fraction C of the 2 years old Primitivo (Fig. 3c, Table 1). Furthermore, a direct adduct between malvidin 3O-glucoside and flavanol dimer was found (peak 51) for the first time into an aged Primitivo; indeed, the mass of its molecular ion (m/z 1071) together with the appearance of the heterocycle-ring-fission fragment ion at m/z 290 in the MS2

spectrum allow to hypothesize that adduct linkage involved the C4 position of the anthocyanin unit and the C8 position of the flavanol unit (Table 1, Fig. 3c) (Blanco-Vega et al., 2014).

As regards anthocyanin-tannin derivatives, two protonated molecules [MþH]þ at m/z 783 (peak 46) and 753 (peak 47) were

Fig. 3. HPLC-UV-Vis chromatogram of a) fraction A, b) fraction B, and c) fraction C recorded at 520 nm.

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e96

detected (Table 1), whose fragmentation patterns, characterized by the major ions at m/z 631 and 601, respectively, due to the neutral loss of 152 u by RDA decomposition of the flavanol unit and less abundant [MþH-162]þ product ions corresponding to the loss of a glucose, were similar to that obtained by Duen~as, Fulcrand, and Cheynier (2006) for a malvidin 3O-glucoside-catechin adduct (A- T) with an A-type linkage (Fig. 4). These data, together with the lack of fragments at m/z 343 and 313 or at m/z 331 and 301, respectively, typical of a flavene form (S#anchez-Il#arduya et al. 2014), as well as the absence of absorption in the visible region, corroborated the hypothesis that the peaks 46 and 47 were the colorless A-type peonidin 3O-glucoside-(epi)catechin and malvidin 3O-glucoside- (epi)catechin, respectively. These bycyclic compounds, which indirectly influence the wine color, were previously identified into model solutions but also into Roja aged red wines (Duen~as et al., 2006; S#anchez-Il#arduya et al. 2014). On the contrary, tannin- anthocyanin derivatives (peaks 1, 2, 4, 6, and 7) having molecular ions and fragmentation patterns typical of (epi)catechin-petunidin or malvidin 3O-glucoside adducts (Pati et al., 2006), were only revealed into young Primitivo (Fig. 1a, Table 1).

Finally, it is really interesting that the anthocyanin dimer mal- vidin 3O-glucoside- malvidin 3O-(p-coumaroyl)-glucoside (peak 42) and trimer (malvidin 3O-glucoside)3 (peak 62) showing a molecular ion at m/z 1131 and 1477, respectively, and fragmentation patterns consistent with what reported in literature (Vidal, Meudec, Cheynier, Skouroumounis, & Hayasaka, 2004; Alcade-

Eon et al., 2007; Oliveira, da Silva, Parola, Mateus, Br#as, Jo~ao Ramos, & de Freitas, 2013), were detected into both wines because this is the first evidence that such compounds survive to the wine aging. Indeed, their presence in red wine is maybe due to their extraction from grape skin during fermentation rather than ob- tained from the degradation of larger instable oligomers, which could be rapidly formed in the young wine as intermediates. This statement is confirmed by the fact that the synthesis of dimers and trimers was not observed in wine model solutions containing malvidin 3O-glucoside (Oliveira, da Silva, Parola, Mateus, Br#as, Jo~ao Ramos, & de Freitas, 2013).

3.3. Color contribution of each pigment class

The aging capacity of a wine is influenced by a number of factors, including the amount of tannin, acid, grape variety, and alcohol present in the wine, as well as the storage conditions of the wine; however, knowing the type and evolution of anthocyanin derived pigments can be of crucial importance for assessing the resistance to aging of a red wine (Rib#ereau-Gayon et al., 2006). Therefore the percentage content of the seven different classes of pigments considered in this work were compared between young and 2 years old Primitivo wines (Fig. 5), in order to establish which is the predominant pigment family at the two moments of wine aging. As expected, a strong decrease of monomeric anthocyanins was revealed in the aged wine respect to young Primitivo; in particular

Fig. 4. MS/MS spectra of the A-type (epi)catechin-malvidin 3O-glucoside (a), and (epi)-catechin-peonidin 3O-glucoside (b).

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e9 7

the absence of both free and derivative forms of delphinidin and cyanidin (Table 1), which were probably degraded during the maturation but not oxidized to form hydroxyphenyl or catechin adducts, as revealed in other aged wines (Alcade-Eon et al., 2006; Wang et al., 2003). Also a relative high percentage (39%) of the grape native pigments, especially malvidin compounds, was still found in the older wine (Fig. 5b, Table 1).

Interestingly, all the found anthocyanin-derived pigments were formed by malvidin, peonidin, and petunidin which are known to be the most stable anthocyanins (He et al., 2012) and are mainly involved in the formation of pyranonathocyanins derivatives, known to have their absorbance maxima within the range of 500e515 nm and accounting for 49% of the total content in the aged wine (Fig. 5b). Furthermore, it can be seen that the relative amounts of compounds, such as flavanol-anthocyanin adducts, which gave blue hues to the samples decreased (from 4.8 to 0.9%) as wine became older, and the reddish or violet ethyledene-bridge favanol- anthocyanin adducts were greatly affected by disappearance during aging, passing from a relative ratio of almost 1:2 up to 1:4 respect to vinyl-linked anthocyanins in the older wine (Fig. 5). Then, overall, it is the predominance of pyranoanthocyanins, characterized by higher stability compared to ethyledene-bridge favanol-anthocy- anin and anthocyanin-flavanol compounds (Alcade-Eon et al., 2006; Boido, Alcade Eon, Carrau, Dellacassa, & Rivas-Gonzalo, 2006), and maybe their early formation (within two years) that would be responsible for the more rapid change of Primitivo color into orange hues compared to other international wines (Blanco- Vega et al., 2014), but also it would be able to slow down the nat- ural loss of structure which wine is subject to (Rib#ereau-Gayon et al., 2006).

4. Conclusion

In this work, the percentage content of seven different classes of pigments were compared between young and 2 years old Primitivo wines. More than fifty anthocyanin-derived pigments were detec- ted whose structures have been elucidated through a comparison with the literature data and interpretation of MS and MS/MS spectra. Compounds such as malvidin 3O-acetyl-4-vinyl-procyani- din, malvidin 3O-glucoside-di(epi)-catechin, and A-type peonidin and malvidin 3O-glucoside-(epi)catechin were identified in Primi- tivo for the first time; then, malvidin-3O-glucoside-malvidin-3O-p- coumaroylglucoside dimer and (malvidin-3O-glucoside)3 trimer anthocyanins were detected into both wines showing a further evidence that such compounds survive to the wine aging. The pyranoanthocyanin pigments were the most abundant in the aged wine, accounting for 49% of the total pigment content. This, and maybe their early formation (within two years), can be responsible for the often reported rapid change of Primitivo color into orange hues compared to other international wines, but also it would be able to slow down the natural loss of wine structure as pyr- anoanthocyanin are the most stable known pigments.

Acknowledgment

This study was supported by grant from the Italian Ministry of University and Research-MIUR.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://

Fig. 5. Distribution (percentage) of anthocyanin-derived red pigments in Primitivo: a) young wine; b) two years old wine.

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e98

dx.doi.org/10.1016/j.lwt.2016.03.012.

References

Alcade-Eon, C., Escribano-Bail#on, M. T., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (2004). Separation of pyranoanthocyanins from red wine by column chroma- tography. Analytica Chimica Acta, 513, 305e318.

Alcade-Eon, C., Escribano-Bail#on, M. T., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (2006). Changes in the detailed pigment composition of red wine during maturity and ageing. A comprehensive study. Analytica Chimica Acta, 563, 238e254.

Alcade-Eon, C., Escribano-Bail#on, M. T., Santos-Buelga, C., & Rivas-Gonzalo, J. C. (2007). Identification of dimeric anthocyanins and new oligomeric pigments in red wine by means of HPLC-DAD-ESI/MSn. Journal of Mass Spectrometry, 42, 735e747.

Asenstorfer, R. E., Hayasaka, Y., & Jones, G. P. (2001). Isolation and structures of oligomeric wine pigments by bisulfite-mediated ion-exchange chromatog- raphy. Journal of Agricultural and Food Chemistry, 49, 5957e5963.

Atasanova, V., Fulckrand, H., Cheynier, V., & Moutounet, M. (2002). Effect of oxygenation on polyphenol changes occurring in the course of wine-making. Analytica Chimica Acta, 458, 15e27.

Bakker, J., & Timberlake, C. F. (1997). Isolation, identification, and characterization of new color-stable anthocyanins occurring in some red wines. Journal of Agri- cultural and Food Chemistry, 45, 35e43.

Blanco-Vega, D., G#omez-Alonso, S., & Hermosín-Guti#errez, I. (2014). Identification, content and distribution of anthocyanins and low molecular weight anthocyanin-derived pigments in Spanish commercial red wines. Food Chem- istry, 158, 449e458.

Boido, E., Alcade Eon, C., Carrau, F., Dellacassa, E., & Rivas-Gonzalo, J. C. (2006). Aging effect on the pigment composition and color of Vitis vinifea L. Cv. Tannat wines. Contribution of the main pigment families to wine color. Journal of Agricultural and Food Chemistry, 54, 6692e6704.

Bowers, J. E., Bandman, E. B., & Meredith, C. P. (1993). DNA fingerprint character- ization of some wine grape cultivars. American Journal of Enology and Viticulture, 44, 266e274.

Coletta, A., Trani, A., Faccia, M., Punzi, R., Dipalmo, T., Crupi, P., et al. (2013). Influ- ence of viticultural practices and winemaking technologies on phenolic composition and sensory characteristics of Negroamaro red wines. International Journal of Food Science and Technologies, 48, 2215e2227.

Crupi, P., Coletta, A., Milella, R. A., Perniola, R., Gasparro, M., Genghi, R., et al. (2012). HPLC-DAD-ESI-MS analysis of flavonoid compounds in five seedless table grapes grown in Apulian Region. Journal of Food Science, 77, C174eC181.

De Villers, A., Vanhoenacker, G., Majek, P., & Sandra, P. (2004). Determination of anthocyanins in wine by direct injection liquid chromatography-diode array detection-mass spectrometry and classification of wines using discriminant analysis. Journal of Chromatography A, 1054, 195e204.

Duen~as, M., Fulcrand, H., & Cheynier, V. (2006). Formation of anthocyanin-flavonol adducts in model solutions. Analytica Chimica Acta, 563, 15e25.

Fulcrand, H., Benabdeljalil, C., Rigaud, J., Cheynier, V., & Moutounet, M. (1998). A new class of wine pigments generated by reaction between pyruvic acid and grape anthocyanins. Phytochemistry, 47, 1401e1407.

Fulcrand, H., Dos Santos, P. J. C., Sarni-Manchado, P., Cheynier, V., & Favre-Bonvin, J. (1996). Structure of new anthocyanin-derived wine pigments. Journal of Chemical Society Perkin Transactions, 1(7), 735e739.

Gergely, A., Kiss, C., Forgacs, E., Cserhati, T., Candeias, M., Vilas-Boas, L., et al. (2000). Solid-phase extraction and high-performance liquid chromatographic separa- tion of pigments of red wines. Journal of Chromatography A, 889, 51e57.

Hakånsson, A. E., Pardon, K., Hayasaka, Y., De Sa, M., & Herderich, M. (2003). Structures and colour properties of new red wine pigments. Tetrahedron Letters, 44, 4887e4891.

Hayasaka, Y., & Asenstorfer, R. E. (2002). Screening for potential pigments derived from anthocyanins in red wine using nanoelectrospray tandem mass spec- trometry. Journal of Agricultural and Food Chemistry, 50, 756e761.

Hayasaka, Y., & Kennedy, J. A. (2003). Mass spectrometric evidence for the forma- tion of pigmented polymers in red wine. Australian Journal of Grape and Wine

Research, 9, 210e220. Heier, A., Blaas, W., Droß, A., & Wittkowski, R. (2002). Anthocyanin analysis by

HPLC/ESI-MS. American Journal of Enology and Viticulture, 53, 78e86. He, F., Liang, N. N., Mu, L., Pan, Q. H., Wang, J., Reeves, M. J., et al. (2012). Antho-

cyanins and their variation in red wines II. Anthocyanin derived pigments and their color evolution. Molecules, 17, 1483e1519.

Maleti#c, E., Peji#c, I., Karoglan Konti#c, J., Piljac, J., Dangl, G. S., Vokurta, A., et al. (2004). Zinfadel, Dobrici#c, and Plavac mali : the genetic relatioship among three cul- tivars of the Dalmatian coast of Croatia. American Journal of Enology and Viti- culture, 55, 174e180.

Mateus, N., Carvalho, E., Carvalho, A. R. F., Melo, A., Gonz#alez-Param#as, A. M., San- tos-Buelga, C., et al. (2003). Isolation and structural characterization of new acylated anthocyanin-vinyl-flavanol pigments occurring in aging red wines. Journal of Agricultural and Food Chemistry, 51, 277e282.

Mateus, N., Silva, A. M. S., Santos-Buelga, C., Rivas-Gonzalo, J. C., & De Freitas, V. (2002). Identification of anthocyanin-flavanol pigments in red wines by NMR and mass spectrometry. Journal of Agricultural and Food Chemistry, 50, 2110e2116.

Oliveira, J., da Silva, M. A., Parola, A. J., Mateus, N., Br#as, N. F., Ramos, M. J., et al. (2013). Structural characterization of a A-type linked trimeric anthocyanin derived pigment occurring in a young Port wine. Food Chemistry, 141, 1987e1996.

Pati, S., Losito, I., Gambacorta, G., La Notte, E., Palmisano, F., & Zambonin, P. G. (2006). Simultaneous separation and identification of oligomeric procyanidins and anthocyanin-derived pigments in raw red wine by HPLC-UV-ESI-MSn. Journal of Mass Spectrometry, 41, 861e871.

Pissarra, J., Lourenco, S., Gonzalez-Paramas, A. M., Mateus, N., Santos-Buelga, C., & de Freitas, V. (2004). Formation of new anthocyanin-alkyl/aryl-flavanol pig- ments in model solutions. Analytica Chimica Acta, 513, 215e221.

Rentzsch, M., Schwartz, M., Winterhalter, P., Blanco-Vega, D., & Hermosin- Guti#errez, I. (2010). Survey on the content of vitisin A and hydroxyphenyl- pyranoanthocyanins in tempranillo wines. Food Chemistry, 119, 1426e1434.

Rib#ereau-Gayon, P., Dubourdieu, D., Don!eche, B., & Lonvaud, A. (2006). Handbook of enology, the microbiology of wine and vinifications. InRed winemaking (Kindle Edition, Vol. 1, pp. 327e395). John Wiley & Sons Ltd.

Romero, C., & Bakker, J. (1999). Interactions between grape anthocyanins and py- ruvic acid, with effect of pH and acid concentration on anthocyanin composi- tion and color in model solutions. Journal of Agricultural and Food Chemistry, 47, 3130e3139.

Salas, E., Atasanova, V., Poncet-Legrand, C., Meidec, E., Mazauric, J. P., & Cheynier, V. (2004). Demonstration of the occurrence of flavonol-anthocyanin adducts in wine and in model solutions. Analytica Chimica Acta, 513, 325e332.

S#anchez-Il#arduya, M. B., S#anchez-Fern#andez, C., Garm#on-Lobato, S., Abad-García, B., Berrueta, L. A., Gallo, B., et al. (2014). Detection of non-coloured anthocyanin- flavonol derivatives in Rioja aged wines by liquid chromatography-mass spec- trometry. Talanta, 121, 81e88.

Sun, B., Fernandes, T. A., & Spranger, M. I. (2010). A new class of anthocyanin- procyanidin condensation products detected in red wine by electrospray ioni- zation multi-stage mass spectrometry analysis. Rapid Communication in Mass Spectrometry, 24, 254e260.

Sun, B. S., & Spranger, M. I. (2005). Changes in phenolic composition of Tinta Miúda red wines after 2 years of ageing in bottle: effect of winemaking technologies. European Food Research and Technology, 221, 305e312.

Vidal, S., Meudec, E., Cheynier, V., Skouroumounis, G., & Hayasaka, Y. (2004). Mass spectrometric evidence for the existence of oligomeric anthocyanins in grape skins. Journal of Agricultural and Food Chemistry, 52, 7144e7151.

Wang, H., Race, E. J., & Shrikhande, A. J. (2003). Anthocyanin transformation in Cabernet Sauvignon wine during aging. Journal of Agricultural and Food Chem- istry, 51, 7989e7994.

Weber, F., Greve, K., Durner, D., Fischer, U., & Winterhalter, P. (2013). Sensory and chemical characterization of phenolic polymers from red wine obtained by gel permeation chromatography. American Journal of Enology and Viticulture, 64, 15e25.

Wollmann, N., & Hofmann, T. (2013). Compositional and sensory characterization of red wine polymers. Journal of Agricultural and Food Chemistry, 61, 2045e2061.

T. Dipalmo et al. / LWT - Food Science and Technology 71 (2016) 1e9 9