Article pubs.acs.org/JAFC
Effect of Commercial Enzymes on Berry Cell Wall Deconstruction in the Context of Intravineyard Ripeness Variation under Winemaking Conditions Yu Gao,† Jonatan U. Fangel,§ William G. T. Willats,§ Melané A. Vivier,† and John P. Moore*,† †
Institute for Wine Biotechnology, Department of Viticulture and Oenology, Faculty of AgriSciences, Stellenbosch University, Matieland 7602, South Africa § Department of Plant and Environmental Sciences, Faculty of Science, University of Copenhagen, DK-1001 Copenhagen, Denmark S Supporting Information *
ABSTRACT: Significant intravineyard variation in grape berry ripening occurs within vines and between vines. However, no cell wall data are available on such variation. Here we used a checkerboard panel design to investigate ripening variation in pooled grape bunches for enzyme-assisted winemaking. The vineyard was dissected into defined panels, which were selected for winemaking with or without enzyme addition. Cell wall material was prepared and subjected to high-throughput profiling combined with multivariate data analysis. The study showed that significant ripening-related variation was present at the berry cell wall polymer level and occurred within the experimental vineyard block. Furthemore, all enzyme treatments reduced cell wall variation via depectination. Interestingly, cell wall esterification levels were unaffected by enzyme treatments. This study provides clear evidence that enzymes can positively influence the consistency of winemaking and provides a foundation for further research into the relationship between grape berry cell wall architecture and enzyme formulations. KEYWORDS: cell walls, intravineyard variation, ripening, fermentation, grape pomace, wine enzymes, multivariate data analysis
■
INTRODUCTION Winemakers are challenged each year to produce consistent wine quality (defined loosely as a recognized style and typicity associated with a specific cultivar). However, various factors (i.e., genotype and environmental) can positively or negatively affect the grape developmental and ripening processes in vineyards, which in turn results in a significant degree of intravineyard variation (e.g., lack of synchronicity of berry ripening), which affects the biochemical composition of the final harvested grapes.1 Cabernet Sauvignon is one of the most iconic wine grape cultivars and has been studied with respect to the role of environmental factors (e.g., UV exposure and water availability)2 and viticultural practices3 that influence berry composition (e.g., sugars, organic acids, flavonols, anthocyanins, and tannins). Variability in chemistry and biochemistry is a phenomenon that is observed at the vine-to-vine, bunch-tobunch, and even berry-to-berry level within vineyards; this has been validated from several studies on a number of cultivars.4−6 In a standard winemaking scenario, it is important to consistently release favorable compounds (e.g., tannins, anthocyanins, etc.) from harvested berries into the final wine via maceration and fermentation. Berry skin cells contain high concentrations and diversity of these beneficial compounds7,8 and need to be broken down to increase the final wine’s structure, body, potential for aging (commonly related to the quality of tannins obtained), as well as expressing the typicity (i.e., wine style as expressed with specific cultivars and winemaking processes, e.g., a Bordeaux blend) of the final wine. In the past couple of decades, various techniques have been tested to aid the maceration process; these include using different yeasts, manipulating the temperature (e.g., cold soak © 2016 American Chemical Society
maceration), and physical disruption (e.g., punch downs or electrical pulse treatment)9−11 to “open up” the skin cells. Enzymes are also major additives that have been applied to effectively degrade the skin cell wall polymers, thereby improving maceration.12 These enzymes are crude extracts from source microorganisms (e.g., fungi) that are considered generally recognized as safe (GRAS) for human health. A number of experimental studies in the past have worked on the relationship (i.e., correlation) between the release of favorable wine compounds and the use of maceration enzymes, with positive and negative impacts reported on wine quality.13−15 However, the direct effect of enzymes on the variation presumably present in harvested berry cell walls, in the context of intravineyard variation, and possible relationship to the fermentation process and final wine produced have not been investigated to any depth. The information generated from most previous studies on cell walls is still exclusively reliant on the indirect (i.e., inferred) approach based on monosaccharide composition and concentration obtained from analyzing isolated and degraded polymers.16,17 The application of glycan microarray technology (i.e., comprehensive microarray polymer profiling (CoMPP)) with well-characterized mAbs and CBMs targeting specific epitopes associated with specific cell wall polymers (e.g., mAb LM15, which targets the XXXG motif glucan backbone in unsubsituted xyloglucan), provides the opportunity to directly Received: Revised: Accepted: Published: 3862
February April 20, April 28, April 28,
25, 2016 2016 2016 2016 DOI: 10.1021/acs.jafc.6b00917 J. Agric. Food Chem. 2016, 64, 3862−3872
Article
Journal of Agricultural and Food Chemistry evaluate enzyme actions on polysaccharides by virtue of changes in their epitope abundance and/or exposure.18−21 This approach allows for a more direct evaluation of enzymemediated changes at the cell wall polymer level. By combining this high-throughput approach with classical cell wall profiling methods, conserved development profiles (but also some differences) were found for Cabernet Sauvignon (a wine grape cultivar) and Crimson Seedless (a table grape cultivar) in a comparative study of samples from (deseeded) whole berries collected at green (berry touch, E-L code L), veraison (E-L code M), and ripe (E-L code N) stages.22 To test the relationship between ripening levels and enzyme-mediated maceration on Vitis vinifera cv. Pinotage skin cell walls, a study was performed over two vintages (with °Brix (°B) at 22.7 and 26.5 °B).20 This study demonstrated the importance of grape ripening levels, which resulted in a greater degree of degradation (presumably due to endogenous enzymes and senescence) of more ripe berries, which lessened any impact of enzyme addition on berry cell walls under winemaking conditions.20 A further study on Pinotage also demonstrated under optimized buffered in vitro experiments using isolated skin cell walls that different enzyme combinations helped to (i) unravel the berry skin cell wall and probably enhanced (ii) degradation (depectination) significantly; the former appears to “loosen” up the pectin enhancing extraction.21 This fits well with the study on whole berries of Cabernet Sauvignon, which was processed to wine and pomace.23 The berry pomace samples from the winemaking were then fractionated chemically and enzymatically.23 A combination CoMPP analysis with classical fractionation, as well as gas chromatography−mass spectrometry (GC-MS) and Fourier transform infrared (FTIR) analyses, revealed evidence for a two-layered pomace structure (an esterified pectin-rich inner layer and a hemicellulose-rich pectin-coated outer layer).23 The less strongly bound pectin-rich inner layer being presumably “loosened-up” (unraveled) by enzymes.21 However, there is no study as far as we are aware, that has focused on investigating the intravineyard grape variation of berry cell walls at harvest and the influence of this on the maceration enzymes used commercially during winemaking. The aims of this study are to assess the relationship between intravineyard grape berry cell wall variation at harvest and the use of commercial enzyme preparations. The study employs a checkerboard design (see Figure 1) for dissecting the Cabernet Sauvignon vineyard into untreated and enzyme-treated panels (similar to a recent study)24 and then analyzes the resulting berry cell wall data sets generated using cell wall profiling approaches22,23 coupled with multivariate data analysis.21,25
■
Figure 1. (A) Harvest plan and (B) ripening level variation of Cabernet Sauvignon. Each block represents a panel, which consists of six vines. U refers to untreated fermentations. (B) Variation of berry soluble solids (°B) at different panels in the vineyard. The level of ripening was categorized into three stages depending on °B value (mean values from three biological repeats), including stage 1 (19− 21), stage 2 (21−23), and stage 3 (23+). The shades of green (dark green for 19−21 °B, light green for 21−23 °B, and pink for 23+ °B) indicate the ripeness stages. these rows represented only a small subsection of the vineyard, and we included six buffer vines from the end of the row to ensure edge-effects were minimized. Panels were harvested individually at a commercial harvest estimation of 24 °B (advised by viticulturists using classical random sampling design, even though we later evaluated for individual panel level °B variation in results) for three different maceration enzyme treatments with fermentation (see Figure 1 and Supplementary Table 1 in the Supporting Information), and grape berries (approximately 15 kg) from each panel (six vines) were pooled together and split into three biological replicates (5 kg each) for subsequent fermentations. Vinification of Wine. After harvest, berry bunches were pooled from each panel, destemmed and crushed, and then split into polypropylene buckets (5 kg each times 3 for each panel (as 3 biological repeats, 72 buckets in total), sodium metabisulfite was added (30 ppm of SO2) after crushing, and then the must was inoculated with prerehydrated yeast (Sacchromyces cereviseae VIN13, Anchor Yeast, South Africa) according to the manufacturer’s instructions and recommendation for fermentation. Three commercial enzyme preparations (Supplementary Table 1), named Lafase HE Grand Cru (for extended maceration), Lafase Fruit (for fruity wines and color extraction), and Lafase XL Extraction (for enhanced color and tannins extraction) (Laffort, France), were applied to the must, using the layout presented in Figure 1, according to the manufacturer’s recommendation (i.e., 5 g/100 kg grape for Grand Cru and Fruit, 4 mL/100 kg grape for XL). Fermentation was conducted at 25 °C, and
MATERIALS AND METHODS
Grape Berry Sampling Layout. In this study, the grape samples (V. vinifera cv. Cabernet Sauvignon) were harvested in March 11, 2014, from the Welgevallen experimental vineyard of the Department of Viticulture and Oenology, Stellenbosch University, South Africa. The vineyard is situated at 33°56′42″ S, 18°51′44″ E and composed of alluvial soils with light to medium texture arranged in a north−south row orientation. Vines are drip-irrigated and trained on a seven-wire vertical trellis system; each row was divided into panels (six vines) using wooden poles as markers. The harvested panels (i.e., 24) were arranged alternating with each other (see layout in Figure 1), and the panels were carefully selected for producing representative samples; diseased/damaged vines were not present in the experimental block, which received similar sunlight exposure and identical viticultural treatments during the 2013−2014 season. It is important to note that 3863
DOI: 10.1021/acs.jafc.6b00917 J. Agric. Food Chem. 2016, 64, 3862−3872
Article
Journal of Agricultural and Food Chemistry
Figure 2. Monosaccharide composition (μg/mg) of AIR sourced from fresh berry from different panels on the vineyard and the berry soluble solid (°B) values: (A) U1−U4 (untreated), Cru1−Cru4 (Lafase HE Grand Cru); (B) U5−U8 (untreated), fruit 1−4 (Lafase fruit); (C) U9−U12 (untreated), XL1−XL4 (Lafase XL extraction). The composition is expressed in μg/mg (dry weight in AIR sample). Ara, arabinose; Rha, rhamnose; Fuc, fucose; Xyl, xylose; GalA, galacturonic acid; Man, mannose; Gal, galactose; Glc, glucose; GlcA, glucuronic acid. Error bars represent the standard derivation of the mean value of three biological repeats. Different letters indicate a significant difference for a specific sugar between the different stages (95% confidence level, ANOVA, P = 0.05).
the pomace cap of each bucket was punched down twice a day to assist maceration. At the end of fermentation, the wines were pressed off-
skins on the same day using a basket press to generate the free run wine and pomace samples for subsequent analyses. 3864
DOI: 10.1021/acs.jafc.6b00917 J. Agric. Food Chem. 2016, 64, 3862−3872
Article
Journal of Agricultural and Food Chemistry General Enological Parameters of Wine. Before the inoculation with the yeast, the must from each panel (Figure 1) was measured for sugar level °B values (soluble solids) to assess the variability of ripening levels within the vineyard experimental block. After the fermentation, the enzyme-treated and untreated wines were sampled at ca. 50 mL units, which were analyzed using FT-IR spectroscopy with a WineScan FT120 Basic (Foss Analytical, Hillerød, Denmark) instrument to assess standard oenological parameters (see Supplementary Table 2). Cell Wall Sample Preparation from Pomace. After the wine was pressed, composite sampling was performed (10 positions on the linearized press-cake of each fermentation) to get approximately 5 g of fully representative pomace as outlined in the theory of sampling;26 the pomace samples were then deseeded and milled (30 Hz for 30s) using a Retsch MM400 mixer mill (Retsch, Haan, Germany) under liquid nitrogen. The resulting powder was processed to alcoholinsoluble residue (AIR).23 Briefly, the powder was heated in 80% (v/v) ethanol at 95 °C for 20 min and then centrifuged to discard the supernatant; the residue was washed using a series of organic solvents (i.e., methanol, chloroform, and acetone).25 At the end of washing steps, the AIR was air-dried and resuspended in Milli-Q water, frozen at −80 °C overnight, and then freeze-dried. Monosaccharide Composition Analysis of Pomace Cell Wall Material. The AIR samples sourced from fresh deseeded berries and fermented pomace were analyzed for their monosaccharide composition according to the method (see Supplementary File 1) described and performed previously.23 CoMPP Analysis of Pomace Cell Wall Material. To assess the variation of whole berry cell wall samples on the polymer level, AIR (10 mg) sourced from grape pomace (with and without enzyme treatment) was used for CoMPP analysis (see Supplementary File 2) of diaminocyclohexanetetraacetic acid (CDTA) and NaOH extracts.18 Multivariate Data Analysis. Multivariate data analysis tools (SIMCA, MKS, Sweden) were used to create models of the data and perform principal component analysis (PCA) plots to visualize the CoMPP data. These analyses were performed using the SIMCA 14 software package (Umetrics AB, Umea, Sweden) (see Supplementary File 3 for statistical analysis methods used).
the season. However, clearly some unknown factors presumably affected sugar accumulation of berry bunches within each panel; here it should be noted that berry bunches were manually harvested per panel and visually inspected, with those bunches showing evidence of shriveling and disease symptoms being discarded. It must be borne in mind that the must tested was sourced from the crushed berries from a single panel, so these °B values actually reflect the mean sugar levels for each panel. The variation study was not performed on a vine or singlebunch level because we also wished to remain at a semiindustrial level relevant for the testing of enzyme preparations (and we also have found volumes
