Aims: The aim of this research was to determine and quantify the ability of enological tannins to reduce laccase activity and, consequently, to protect wine color against enzymatic browning and/or oxidasic haze.

Methods and results: Botrytized grape juice with laccase activity was obtained by inoculating Botrytis cinerea in healthy mature grapes. Laccase activity was determined in grape juice before and after supplementation with enological tannins using the syringaldazine method. White micro-fermentations were performed in the presence or not of laccase activity and supplemented or not with enological tannins in order to determine how the color was affected. Similarly, red micro-fermentations were performed using white grape juice supplemented with malvidin-3-O-glucoside. All enological tannins inhibited laccase activity and protected the wine color.

Conclusion: Supplementation with enological tannins is an interesting tool to inhibit laccase activity and protect the color of white wines from browning and the color of red wines from oxidasic haze.

Significance and impact of the study: This is the first scientific study evidencing the inhibitory effect of enological tannins on laccase activity in winemaking conditions.

Keywords: enological tannins, Botrytis cinerea, grey mould, laccase activity, contact time, dose effect

Enological tannins are usually classified in two families, namely hydrolysable and condensed tannins. Their use in winemaking is a common practice (Obradovic et al., 2005), with doses ranging from 5 to 100 g/hL depending on the addition strategy and the desired effect. Nevertheless, up to date, they are only authorized by the International Organization of Vine and Wine (OIV) in order to facilitate the fining of musts and wines (OIV, 2015).

On the one hand, the family of hydrolysable tannins comprises gallotannins and ellagitannins. Gallotannins are polymers formed by esterification between D-glucose and gallic acid. Tannic acid is the commercial name for gallotannin extract comprising mixtures of polygalloyl quinic acid ester or polygalloyl glucoses (Pascual et al., 2017). The main sources of commercial gallotannins are nut galls and tara. Ellagitannins are polymers of ellagic, gallic and/or hexahydroxydiphenic acids (Versari et al., 2013). More precisely, a nonahydroxyterphenoyl unit (NHTP) is esterified in positions 2, 3 and 5 with a C-glycosidic bond, while an open-chain glucose is esterified in positions 4 and 6 with a hexahydroxydiphenoyl unit (HHDP) forming the chemical structure of ellagitannins (Quideau et al., 2004). In addition to the major phenolic compounds, hydrolysable tannin preparations contain monosaccharides: oak, quebracho and hazel contain arabinose, xylose, glucose and fructose, while tannins from grapevine bunch and nut gall contain only fructose and glucose (Versari et al., 2013). The main sources of commercial ellagitannins are oak, chestnut and myrobalan.

On the other hand, the family of condensed tannins, also called proanthocyanidins, mainly includes procyanidins, prodelphinidins and profisetinidins. They differ mainly in respect to the type of monomer released after acidic cleavage, the degree of polymerization (mDP), and their levels of galloylation (Versari et al., 2013). Grape-skin tannins are composed of procyanidins and prodelphinidins since their acidic cleavage releases cyanidin and delphinidin. However, grape-seed tannins are only composed of procyanidins. Grape-skin tannins have a high mDP and a low level of galloylation, while grape-seed tannins have a lower mDP and a high level of galloylation (Souquet et al., 1996). Quebracho tannins are profisetinidins, because their acidic cleavage gives fisetinidin, and they are characterized by a high level of ramification, while mimosa tannins are prorobinetidins releasing robinetinidin (Celzard et al., 2015). The main sources of commercial condensed tannins are grapevine seeds and skins and other plant sources such as quebracho, mimosa and acacia (Versari et al., 2013).

Furthermore, each group of tannins has different composition, nature, and chemical structure associated with varied properties. For this reason, enological tannins are also used for other goals than fining of must and wine. These properties include the improvement and stabilization of red wine color by means of their effect as copigments (Neves et al., 2010) and their participation in the formation of new pigments (Versari et al., 2013). Enological tannins are also used as antioxidant (González-Centeno et al., 2012; Pascual et al., 2017) and as antioxidasic (Versari et al., 2013). Concerning the interaction with proteins, enological tannins are mainly used to help protein fining and prevent protein haze (Ribéreau-Gayon et al., 2006), while avoiding gelatin over-fining (Peynaud, 1984). Finally, they are also used to improve wine structure and mouthfeel (Preys et al., 2006) and to eliminate reduction odors (Vivas, 2001).

Although these properties of enological tannins have been widely described in the literature, the antioxidasic properties (anti-laccase) are not yet well documented. Botrytis bunch rot (BBR), caused by the necrotrophic pathogenic fungus Botrytis cinerea, is responsible for huge economic losses each year. The infection of the bunch with B. cinerea provokes serious biological and chemical changes that impact negatively on the organoleptic qualities of the wine (Ribéreau-Gayon et al., 1980). When climatic conditions are wet under mild weather conditions (Ciliberti et al., 2015), B. cinerea can infect directly grape berries from veraison onwards. A vintage contaminated by the pathogen at the rate of 5% in severity already shows irreversible consequences on the organoleptic features of a qualitative red wine (Ky et al., 2012). The infection of grape berries by B. cinerea is accompanied by the excretion of fungal metabolites (glycerol, gluconic acid and -glucans) and enzymes (pectinases, proteases, tyrosinases and laccases) in the host cells. Laccases (EC 1.10.3.2) are o-diphenol and p-diphenol: dioxygen oxidoreductases. Generally, in B. cinerea, two genes are found encoding laccases with molecular weights of about 60 kDa (Claus et al., 2014). These multi-copper glycoproteins catalyze the oxidation of mono- and/or di-phenolic substrates in the presence of oxygen. One of the important features of laccase is that it is very stable at wine pH (More et al., 2011). The oxidation of the polyphenols, and thus the alteration of the color, begins in the grape berry and continues in the grape juice. In skins from infected grape berries and in musts derived from botrytized grapes, the laccase produced by the fungus oxidizes the polyphenols and leads to the formation of quinones. These quinones will polymerize and form brown compounds, which is called oxidase case and leads to color degradation and instability (Pourcel et al., 2007). This phenomenon takes place in red and white grape juices, giving brick and brown tints, respectively. Furthermore, the organoleptic qualities of wines are also impacted by changes in their equilibrium, body and mouthfeel (Claus et al., 2014). Laccase activity in musts can be determined by a manual spectrophotometric method developed previously by Grassin and Dubourdieu (1986).

Until now, it is only possible to inactivate or inhibit the enzyme by thermovinification or addition of sulfur dioxide (Ribéreau-Gayon et al., 2006) or to protect the grape juice against oxidation using inert gas or ascorbic acid (Li et al., 2008). In our days, the consumer wants healthier and more eco-friendly products and for this reason, the wine industry is searching for alternative products for reducing or even eliminating the use of sulfur dioxide. Thus, the aim of this research was to determine the ability of enological tannins to reduce the laccase activity produced from the pathogenic infectious process by B. cinerea and, consequently, to protect the wine color against enzymatic browning and oxidasic haze.

All samples and standards were handled without any exposure to light. L-(+)-tartaric acid, sodium hydroxide, sodium acetate, polyvinylpolypyrrolidone (PVPP), Tween 80 and syringaldazine were purchased from Sigma-Aldrich (Madrid, Spain). Malvidin-3-O-glucoside was purchased from Extrasynthese (Genay, France). D-(+)-glucose, peptone, agar and yeast extract were purchased from Panreac (Barcelona, Spain) and potassium metabisulfite was purchased from Acros Organics (Madrid, Spain). Ethanol (96%) and hydrochloric acid were supplied by Fisher Scientific (Madrid, Spain). Yeast (Zymaflore® Spark) and nutrients (Nutristart®) were provided by Laffort (Floirac, France).

The equipment used was as follows: a spectrophotometer UV-Vis Helios Alpha™ (Thermo Fisher Scientific Inc., Waltham, MA, USA); an incubator IPP 260 (DD Biolab, Barcelona, Spain); a centrifuge Heraeus™ Primo™ (Thermo Fisher Scientific Inc., Waltham, MA, USA); and a CB Standard Balance (Cobos, Barcelona, Spain). All the materials for the micro-vinification were provided by the cellar “Mas dels Frares” of the Enology Faculty of the Rovira i Virgili University (Constanti, AOC Tarragona, Spain).

2. Commercial tannins

Five commercial tannins, representing the main botanical origins, were considered in this study. Specifically, three condensed tannins were used: one procyanidin from grape seeds, one procyanidin/prodelphinidin from grape skin and one profisetinidin from quebracho, as well as two hydrolysable tannins: one gallotannin from nut galls and one ellagitannin from oak. All these tannins were provided by Laffort (Floirac, France).

During the 2017 vintage, healthy grapes (Vitis vinifera cv. Muscat d’Alexandrie) were collected on September 18th (around 50 kg) from the experimental vineyard, planted in 1992, at the Enology Faculty of the Rovira i Virgili University in Constanti (AOC Tarragona; 41°8ʼ54.17” N and 1°11ʼ53.89” E). The vineyard is at 87 m above sea level, and groundwater is located at a depth of around 4 m. The vines were trained on a vertical trellis system and arranged in rows 2.80 m apart, with 1.20 m spacing between vines. They were pruned using a double “Cordon de Royat” system, with 16 buds, 8 on each cane. Half of the grapes were kept at 4°C to obtain healthy grape juice, whereas the other half was inoculated with B. cinerea.

The B. cinerea single-spore isolate 213, originally isolated from grapevine leaves in 1998, was selected from the collection of UMR SAVE, Bordeaux (Martinez et al., 2003). It was selected because of its virulence on grapevine leaves and berries and because it is a transposa type strain (Martinez et al., 2003; Martinez et al., 2005; Ky et al., 2012). The pathogen was inoculated on Yeast Peptone Dextrose (YPD) Petri plates (20 g/L of peptone and glucose, 10 g/L of yeast extract and 17 g/L of agar in distilled water) and grown about 1 week at 20°C in an incubator. Half of the grapes harvested (around 25 kg) were placed in five plastic boxes (600x400x200 mm) and inoculated by spraying a spore suspension (1^.10⁶ conidia/mL, 1 drop of Tween 80 and 50 g/L of glucose in sterilized water) until the complete fruit surface was covered by the spore suspension. The plastic boxes containing the grapes were then incubated for around 3 weeks, at 20°C, surrounded by two plastic boxes containing sterile water to maintain the humidity (90-100%).

The healthy grapes were crushed and pressed to obtain a healthy grape juice using a small pneumatic press (Venmhidrprei-040, Invia, Vilafranca del Penedès, Spain). The juice was also recovered under dry ice to keep the grape juice protected from oxidation and was slightly sulphited (60 mg of K₂S₅O₇/L). This healthy grape juice was settled with previous addition of 20 mg/L of pectolytic enzymes at 4°C for 18 hours. The botrytized grapes were previously sorted visually to remove undesirable rotten berries due to other fungal development, notably by Penicillium spp. (blue-green color) or other fungal species (Alternaria spp. or Clostridium spp.), or acetic bacteria (red-pink color). Then, these selected botrytized grapes were crushed and pressed to obtain the botrytized grape juice using a small pneumatic press. The juice was also recovered under dry ice in order to keep it protected from oxidation, but without any addition of sulfur dioxide in order not to inhibit laccase activity. This botrytized grape juice was centrifuged at 8,500 rpm for 5 minutes. Both grape juices, healthy and botrytized, were then immediately stocked in glass bottles at - 4°C.

Laccase activity was determined using an adaptation of the syringaldazine test method (Grassin and Dubourdieu, 1986). Five mL of the different samples were added with 0.8 g of PVPP (to remove phenolic compounds that can cause interference), stirred and centrifuged for 10 minutes at 8,500 rpm. One mL of the supernatant was introduced into a plastic spectrophotometer cuvette to which were also added: 1.4 mL of buffer solution (8.2 g/L of sodium acetate in deionized water, pH 5.5) and 0.6 mL of syringaldazine solution (60 mg/L of syringaldazine in ethanol 96%). The solution was then homogenized by inverting the cell and the absorbance was measured at 530 nm every minute for 5 minutes (including time measurement at 0 minute). All analyses were performed in triplicate. By definition, a laccase unit (UL) corresponds to the amount of enzyme catalyzing the oxidation of a micromole of syringaldazine per minute. The following equation was used for the calculation of laccase activity by using the slope of the line obtained by a calibrating linear regression (ΔA) expressed in absorbance units/minute:

L a c c a s e   a c t i v i t y = 46.15   x   ∆ A   µ m o l . m l   - 1 = 46.15   x   ∆ A   U L

A solution of each one of the different tannins was prepared at 2 g/L in a model wine solution (12% ethanol, 4 g/L of tartaric acid, pH 3.5). Then, the first tube, representing the control, was filled with 4 mL of botrytized grape juice supplemented with 1 mL of deionized water. The second tube, corresponding to a dose of 20 g/hL, was filled with 4 mL of botrytized grape juice and completed to 5 mL by adding 0.5 mL of tannin solution and 0.5 mL of deionized water. The third tube, corresponding to a dose of 40 g/hL, was filled with 4 mL of botrytized grape juice and completed to 5 mL by adding 1 mL of tannin solution. After 5 minutes, the tubes were supplemented with 0.8 g of PVPP, stirred, centrifuged for 10 minutes at 8,500 rpm and used for laccase activity measurement. All analyses were performed in triplicate. The following equation gives the residual laccase activity:

%   o f   r e s i d u a l   a c t i v i t y =   ( l a c c a s e   a c t i v i t y   s a m p l e / l a c c a s e   a c t i v i t y   c o n t r o l )   x   100    

The estimation of the inhibition kinetics of enological tannins on laccase activity was carried out with the highest dose (40 g/hL) at 0, 1, 2, 3, 4, 5 or 10 minutes of incubation before adding the PVPP and centrifuging. These data were used to determine the time needed to reach the maximal inhibition of laccase activity. All analyses were performed in triplicate.

Two types of winemaking trials were performed using the grape juices obtained from the healthy and botrytized grapes. The first trial was carried out directly with the white grape juices (white winemaking), whereas the second trial was performed with the white grape juices supplemented with malvidin-3-O-glucoside (pseudo-red winemaking) in order to approach what happens in red winemaking. This strategy was selected instead of using real red winemaking because we wanted to determine the protective effect of the different enological tannins on the color of anthocyanins without the presence of the natural proanthocyanidins from seeds and skins that can also exert an inhibitory effect on laccase activity. All winemaking trials were carried out in plastic tubes of 30 mL which were used as fermentation vessels. Working solutions of the five enological tannins were prepared at 10 g/L in a model wine solution (12% ethanol, 4 g/L of tartaric acid, pH 3.5) for supplementing the different winemaking trials. The healthy grape juice was used without any treatment. However, the botrytized juice was drastically treated with PVPP (160 g/L) and centrifuged for 10 minutes at 8,500 rpm to eliminate the interferences that its dark brown color could cause in the final wines. In the case of fermentations without laccase addition, 22 mL of healthy grape juice were added to each vessel. In the case of fermentations with laccase addition, 17 mL of healthy grape juice were supplemented with 5 mL of botrytized grape juice (corresponding to a 1.5 UL/mL of final volume). Immediately, 1 mL of deionized water was added for the controls or 0.5 mL of each tannin solution and 0.5 mL of deionized water (corresponding to a dose of 20 g/hL), or 1 mL of each tannin solution (corresponding to a dose of 40 g/hL). All the vessels were inoculated with 1 mL of a solution containing the yeast (10 g/L of Zymaflore Spark®, Laffort, Floirac, France) and 1 mL of nutrient solution (10 g/L of Nutristart®, Laffort, Floirac, France). The fermentations were carried out at room temperature (20 ± 2°C) and were monitored by weighing the vessels. Residual sugars were determined to ensure that alcoholic fermentations were finished. With the aim of approaching red winemaking, the same experiment was done with supplementation of 50 mg/L of malvidin-3-O-glucoside (Extrasynthese, Genay, France). All these fermentations were performed in triplicate.

Once the fermentations were completed, the impact of B. cinerea on the color of the wine was determined by measuring the absorbance at 420 nm (yellow) for white winemaking and at 520 nm (red) for “red winemaking” (Glories, 1984). In the case of “red winemaking”, the total anthocyanin concentration was also analyzed (Ribéreau-Gayon and Stonestreet, 1966). Finally, the total color difference (ΔEab*) between samples was obtained using the CIELAB coordinates (Ayala et al., 1997) and was calculated using the following equation:

Δ E a b *   = L 2 * - L 1 * 2 + a 2 * - a 1 * 2 + ( b 2 * - b 1 * ) ²

The ΔEab* represents a measure of the difference between two colors. ΔEab* is used to know whether the difference between two samples can be detected visually by the human eye. Generally, it is considered that the difference is visible to the human eye when ΔEab* > 3 (García-Marino et al., 2013).

All the chemical and physical data are expressed as mean values ± standard deviation. The statistical analyses were carried out using the XLSTAT 2017 statistical package. The two hypotheses of normality and homoscedasticity of the data were tested, for all parameters, by using the Shapiro-Wilk test and Levene’s test, respectively. When populations were distributed normally and presented homogeneity in variance, parametric tests (ANOVA and Tukey) were used to detect significant differences at p-value < 0.05. In contrast, when populations were not distributed normally and/or presented heterogeneity in variance, non-parametric tests (Kruskal-Wallis and pairwise-Wilcox) were used. Differences were considered to be statistically significant at p-value < 0.05.

Figure 1 shows the residual laccase activity in botrytized grape juice supplemented with the different enological tannins at 0, 20 and 40 g/hL. The initial laccase activity of the grape juice was 8.58 ± 1.34 units. Residual activity was expressed in all cases as a percentage of residual activity with respect to the control (% of residual activity). The supplementation with all enological tannins caused a significant decrease in the residual laccase activity compared to the control. The residual activities following supplementation with the different enological tannins ranged between 55.3 ± 4.5% (in the case of the grape-seed tannin at 40 g/hL) and 81.5 ± 7.5% (for ellagitannin at 20 g/hL). Regarding the doses of enological tannins used, the higher the dose, the lower the residual activity. In fact, the residual activity ranged between 75.5 ± 8.6% and 81.5 ± 7.5% for the dose of 20 g/hL and between 55.3 ± 4.5% and 68.9 ± 5.4% for the dose of 40 g/hL. The differences between doses were significant at P = 0.05, with the only exception of the grape-skin tannin. In contrast, ellagitannins presented the highest difference between the two doses, decreasing from 81.5 ± 7.5% to 58.5 ± 3.1%. All the tannins exerted a similar inhibitory effect at the low dose (20 g/hL). Specifically, the percentage of residual laccase activity was 81.5 ± 7.5% for ellagitannin, 79.9 ± 5.0% for quebracho tannin, 78.7 ± 1.6% for gallotannin, 76.7 ± 5.6% for grape-seed tannin and finally 75.5 ± 8.6% for grape-skin tannin. However, significant differences were observed by testing the highest dose (40 g/hL). Specifically, grape-seed tannins and ellagitannins were the most effective, reaching at the highest dose 55.3 ± 4.5% and 58.5 ± 3.1% of residual activities, respectively. Quebracho tannins, gallotannins and grape-skin tannins were the less effective, reaching levels of inhibition of 63.9 ± 1.1%, 66 ± 2.6% and 68.9 ± 5.4%, respectively. These data confirm that all enological tannins significantly inhibited laccase activity at the usual dose of use.

Table

All the data are the mean ± SD of three replicates. The capital letters represent the significant differences between the different doses for the same tannin. The Greek letters represent the significant differences between the different tannins for the same dose.

The authors are grateful to Laffort industries (Floirac, France) for providing the enological tannins, nutrients and yeast and for financial support.