Grapes, the essential raw material determining wine volatile composition. It s not just about varietal characters. Paul Boss and Eric Dennis Food Futures Flagship and CSIR Plant Industry, P Box 350 Glen smond, SA 5064, Australia. Introduction Having succeeded in convincing a consumer to purchase a bottle of wine, it is important to satisfy their expectations to ensure subsequent purchases and guarantee a positive impression of the label and/or winery. Flavour and aroma will have an important influence on the repeat purchase of a wine. It is the mixture of volatile and semi-volatile compounds in the wine that a consumer will sense as aroma and flavour, but the origin of these wine components is varied. Some are produced by fermentation microbes or are sourced from any oak used during the vinification process. However, a vital input to the process is the grapes and these provide essential raw materials that determine the flavour and aroma characteristics of the wine. This is obvious when one considers the varietal impact on wine sensory attributes, but is also apparent when wines of the same variety are compared between vintages, regions or vineyards. The types of compounds grapes contribute to the pool of volatile and semi-volatile compounds in the wine are vary widely (Dunlevy et al. 2009). First, there are grape 1
components that do not get transformed in any way during the winemaking process (Figure 1A). These are rare, but can be extrememly important contributors to varietal wine characters. Methoxypyrazines, which influence vegetal wine attributes in varieties such as Cabernet Sauvignon and Sauvignon blanc (Dunlevy et al. 2010), and rotundone, which is responsible for the peppery character of Shiraz (Wood et al. 2008) are two such examples. Second, grapes contain non-volatile precursors of certain wine volatile components that can be released during winemaking or bottle aging of wines (Figure 1B). These generally fall into two classes; the glycosides and the cysteine conjugates. The glycosides consist of simple sugars linked to volatile compounds such as monoterpenes, C 13 -norisoprenoids and volatile phenols rendering them flavourless and non-volatile. A portion of these glycosides undergoes hydrolytic cleavage during winemaking to release the volatile components into the wine matrix and this can be enhanced through the use of exogenous fungal enzyme preparations. A fraction of the cysteine conjugates are also cleaved during winemaking to release extremely potent volatile thiols. Their importance to wine aroma has focussed mostly on Sauvignon blanc, although they have been reported in Colombard, Melon B., Chenin blanc and Semillon. Third, grapes contribute compounds that can be metabolised during fermentation to produce volatile components of wine (Figure 1C). In some cases, these grape compounds can also be produced by yeast themselves via the metabolism of sugars and nitrogenous precursors, for example the amino acid valine (Figure 1C). Therefore, the combination of the grape and yeast pools of these compounds will contribute to the volatile profiles of the wine. In several other viticultural studies conducted in our laboratory we found that, even with controlled winemaking procedures, the levels of many of these fermentation-derived volatile compounds varied, suggesting the different grape samples could alter their 2
production. Therefore, we set out to identify those wine components, produced by yeast, that could be influenced by grape composition (Keyzers and Boss 2010). Altering the levels of grape juice in a fermentation Many fermentation volatiles can be produced by yeast from primary metabolites such as sugars and amino acids. This can be modelled in artificial solutions or musts containing primary metabolites and various vitamins and micronutrients. Model musts have been used to examine of the effects of nutrients or fermentation conditions on the production of fermentation volatiles. However, these experiments often assume that the major role the grape plays in the production of fermentation-derived volatiles is as a source of sugars, amino acids and nitrogen. To test this assumption we conducted sets of small-scale fermentations in which the major variable was the amount of grape juice in the fermentation. In order to do this, ferments of equal volume but varying in the amount of grape juice (either Cabernet Sauvignon or Riesling) were prepared (Figure 2). Six different compositions were chosen to measure (0, 5, 10, 20, 50 and 100% v/v grape juice), three of which were at low juice concentrations (0-20%) to highlight significant changes that may occur with only small addition of juice and minimal changes in yeast assimilable nitrogen (YAN). Volatile components were analysed in these wines and those whose concentration increased as the proportion of grape juice in the musts increased were identified. These compounds were of interest as such a pattern suggests that their production is dependent or enhanced by the presence of grape components in the fermentation. Must composition can dramatically alter levels of fermentation-derived esters 3
Several of the compounds found to increase as the grape percentage increased are known to be grape-derived. Examples of these are the terpenes in Riesling and β- damasceneone in both Cabernet Sauvignon and Riesling (Figure 3). These observed changes in the model fermentations were due to a dilution effect. That is, less of these compounds were present as less grape juice was used. These findings confirmed that the experimental approach we have taken could identify compounds dependent on grape composition. The most important discovery from these experiments was the observation that the concentrations of many fermentation-derived esters increased as the amount of Cabernet Sauvignon grape juice present in the fermentations increased (Figure 4). These compounds can not be detected in grape juice and were made in fermentations with model must alone, but their production significantly increased as a greater percentage of juice was present in the fermentations. This was despite an excess of amino acids and other forms of yeast assimilable nitrogen being added to all the fermentations to minimise the effect of these variables, both of which can influence ester production. This increase in ester production was not observed to the same extent in the Riesling series of fermentations. This again suggests the production of these ester-derived compounds is influenced by grape composition. These experiments suggest that grape juice contributes significantly to the pool of substrates the yeast uses to produce fermentation-derived volatiles aside from sugars and amino acids. Alternatively, compounds in the grape juice may stimulate the production yeast-derived esters without being direct precursors of these compounds. Clearly the properties of different yeast strains, such as their ability to transport compounds into the cell and the substrate preferences of their enzymes will influence 4
the mixture of volatile compounds present in wine. However, the experiments described in this article have been used to identify fermentation esters that may be influenced by grape composition. It has been shown that even minor variations in the ester profiles of wines can have a major sensory impact (Pineau et al 2009). Therefore, investigation into the origins and fates of such esters and the role the grape plays in their production will provide valuable information that has the potential to benefit the industry in several ways. Industry implications of these findings There is an increasing body of knowledge regarding the compounds which contribute to aroma and flavour in the finished wine. However, there is a large gap in our knowledge about how compounds in the grape berries contribute to the final flavour and aroma characteristics of the wine. There is no technology for the objective measurement of grape flavour attributes that growers and wineries can easily use to assess their product. There are also no scientifically validated methods of flavour management in the vineyard that provide producers with the ability to better manage the flavour potential of their grapes. The major discovery of this work was that wine ester production is greatly influenced by the grape component of the must. Previously these compounds have been considered yeast-derived and few studies have investigated how grape composition may alter their production beyond altering the nitrogen status of the must. This extends the impact grapes have on wine flavour and aroma beyond varietal characteristics. Even with advanced modern winemaking technologies, the flavour and aroma of 5
wines depends on the composition of the starting material: the grapes used to produce the wine. bjective measures of fruit flavour potential and a means of predicting wine sensory attributes from grape composition could provide a step change in improving our ability to efficiently grow grapes to suit desired wine styles. It will help to promote improved decisions about harvest timing as well as batching and streaming of fruit to consistently produce desired wine styles. More importantly, it will provide tools to optimise grape flavour potential in the vineyard and deliver the means of producing grapes with a desired chemical profile that can be used to make wines of a specified flavour profile. Acknowledgements This work is, in part, funded by the Grape and Wine Research and Development Council and we would like to thank Yalumba for access to vineyards and grape samples. References Dunlevy JD, Kalua CM, Keyzers RA, Boss PK (2009) The production of flavour and aroma compounds in grape berries. In Molecular Biology and Biotechnology of Grapevine, 2nd Edition, K.A. Roubelakis-Angelakis ed. Springer, Berlin, Germany, pp 293-340. Dunlevy JD, Soole KL, Perkins MV, Keyzers RA, Kalua CM, Boss PK (2010) Two -methyltransferases involved in the biosynthesis of methoxypyrazines, key grape 6
derived aroma compounds important to wine flavour. Plant Molecular Biology, DI 10.1007/s11103-010-9655-y. Keyzers RA, Boss PK (2010) Changes in volatile production in fermentations made from musts with increasing grape content. Journal of Agricultural and Food Chemistry 58: 1153-1164. Pineau, B.; Barbe, J.-C.; van Leeuwen, C.; Dubourdieu, D. Examples of perceptive interactions involved in specific "Red-" and "Black-berry" aromas in red wines. J. Agric. Food Chem. 2009, 57, 3702-3708. Wood, C.; Siebert, T. E.; Parker, M.; Capone, D. L.; Elsey, G. M.; Pollnitz, A. P.; Eggers, M.; Meier, M.; Vossing, T.; Widder, S.; Krammer, G. Sefton, M. A.; Herderich, M. J. From wine to pepper: Rotundone, an obscure sesquiterpene, is a potent spicy aroma compound. Journal of Agricultural and Food Chemistry, 2008, 56; 3738-3744. 7
Figure 1 Grapes contribute to wine volatile composition via several mechanisms. A. Compounds that are not altered during fermentation. B. Volatile compounds released by cleavage of non-volatile conjugates. C. Volatile compounds produced by yeast metabolism of a grape precursor. Grape Wine 1A N N Me IBMP N N Me IBMP 1B Rotundone -Glycoside Rotundone Glycosylated precursor H 2 N CH S H S-Cysteine congugate -damascenone SH H 3-mercaptohexanol(3-MH) 1C H NH 2 H Valine isobutyric acid Et ethyl-isobutanoate H isobutyl alcohol isobutyl acetate 8
Figure 2 A series of model musts made with increasing amounts of Cabernet Sauvignon juice. 9
Figure 3 Changes in the concentration of some terpenoids in the Riesling fermentations (A) and b-damascenone in the Riesling and Cabernet Sauvignon fementations (B) in relation to the percentage of grape juice present in the model musts. 10
Figure 4 Changes in the concentration of some acetate esters (A) and isoamyl esters (B) in relation to the percentage of Cabernet Sauvignon grape juice present in the model musts fermented into wine. 11