The flavour or aroma of wine is one of the most important factors when determining quality and intrinsic value.

Even small variations in the occurrence and concentration of volatile flavour components may account for the difference between a premium, gold medal winner and an ordinary table wine. It is common knowledge that consumers tend to purchase wine “with the nose”.

Some of the most distinct flavour components encountered in wine are the volatile thiols, especially 4-mercapto-4-methylpentan-2-one (4MMP), 3-mercaptohexan-1-ol (3MH) and 3-mercaptohexyl acetate (3MHA). The term thiol or mercapto refers to the sulphur-containing functional SH-group (sulphur and hydrogen) present in these components. The perception threshold values are 0.8 ng/L for 4MMP, 60 ng/L for 3MH and 4 ng/L for 3MHA, which explains why these components are so distinctly perceptible. To illustrate this point, theoretically 1 mg of pure 4MMP is able to adjust the aroma of one million litres of wine!

Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations Wynboer - December 2007 - Enhancement of Sauvignon blanc wine aroma through yeast combinations

In Sauvignon blanc wine these components are extremely important with regard to the cultivar character, since they contribute to the characteristic cultivar flavour which is reminiscent of boxwood (4MMP), passion-fruit, grapefruit, gooseberry and guava (3MH and 3MHA) (Dubourdieu et al., 2006). The flavour components, 4MMP, 3MH and 3MHA have been identified in various concentrations in wines made from Riesling, Colombard, Smillon, Cabernet Sauvignon and Merlot grapes, and might well influence the flavour of these cultivars (Tominaga et al., 2000; Murat et al., 2001b).

Sauvignon blanc wines have characteristic flavour profiles which may be described as capsicum, asparagus, boxwood, tomato leaf, grapefruit, gooseberry and passion-fruit. Work by Augustyn (1982) and later Allen (1991) identified the methoxypyrazines, specifically isobutyl-methoxypyrazine (IBMP), as the predominant flavour component that determines capsicum and asparagus flavours. Methoxypyrazines derive exclusively from the grape and are largely influenced by the microclimate of the vineyard, where shaded grape bunches usually have higher IBMP concentrations compared to bunches that are exposed to the sun (Marais et al., 1999).

On the other hand, the volatile thiols are practically absent in grape juice and develop only during the fermentation process. This explains the commonly held notion that the wine yeast Saccharomyces cerevisiae is responsible for the formation of volatile thiols during fermentation. However, yeasts do no synthetise this kind of volatile thiol de novo. Darriet et al. (1995) found that 4MMP and 3MH occur in the grape in the form of aroma-free, non-volatile, cysteine-bound compounds and that yeast is only involved in splitting the aromatic thiols from the aroma-free grape precursor compounds.

An enzymatic mechanism for the release of thiols was suggested after it was found that a bacterial enzyme extract containing carbon sulphur lyase enzymes, was able to release 4MMP from the precursor S-4-(4-methylpentan-2-one)-L-cysteine (Cys-4MMP) (Tominaga et al., 1995). It was then suggested that the enhancement of Sauvignon blanc specific cultivar flavours during fermentation possibly occurred through the action of yeast carbon sulphur lyases. Researchers at the Australian Wine Research Institute (AWRI) in Adelaide investigated how the ability of yeasts to release 4MMP from Cys-4MMP, was influenced when their carbon sulphur lyase coding gene was removed. Four laboratory yeast genes suspected of coding for carbon sulphur lyase enzymes, were found to influence the release of the volatile thiol 4MMP. These research results indicated that multiple genes were possibly involved in the release mechanism. These findings were also confirmed in a commercial wine yeast, thereby indicating that the removal of the four suspected carbon sulphur lyase genes resulted in a reduction in the amount of 4MMP released (Howell et al., 2005). AWRI researchers showed furthermore that the volatile thiol 3MHA was formed by yeast from 3MH through the action of the ester forming enzyme, alcohol acetyl transferase (Swiegers et al., 2006b). This work first proved the relationship between the volatile thiol and ester metabolisms of yeast cells. This enzymatic release mechanism was recently confirmed when the same researchers showed that excessive expression of a bacterial carbon sulphur lyase gene in a wine yeast resulted in a dramatic increase in the concentration of volatile thiols (Swiegers et al., 2007). In this study Sauvignon blanc grapes from a warm Australian wine region (Riverland) were used to make experimental wine using the latter yeast, thereby causing the passion-fruit flavour in the wine to increase dramatically (Figure 1). It is important, however, to note that current Australian policy does not provide for any commercial wine to be made using genetically modified (GMO) yeast strains (Swiegers et al., 2007). The wine yeast mentioned above that contains the bacterial carbon sulphur lyase gene is used purely as an experimental model to investigate the underlying mechanisms for thiol release.

Professor Denis Dubourdieu’s research group in Bordeaux proved that when the concentration of a pre-added, chemically synthetised Cys-3MH-precursor compound in model fermentations decreased during the fermentation process, the 3MH-concentration increased. However, only a small part (1.6% on the sixth day of fermentation) of the precursor compound was released as 3MH. Which means that wine yeasts have a limited ability to release volatile thiols (Dubourdieu et al., 2006). In Cabernet Sauvignon and Merlot grape must it was proved that the amount of 3MH released, corresponds proportionally to the Cys-3MH-concentration in the unfermented grape juice. The higher the concentration of the thiol precursors in the grape juice, the higher the volatile thiol concentration in the wine (Murat et al., 2001b). The fact that only 3.2% of the Cys-3MH-precursor in the latter study was released during fermentation, confirmed that most wine yeasts had a very limited ability to release volatile thiols in wine. In other words, wine yeasts do not have the ability to develop the full aroma potential of grape must and therefore a large source of flavour in the grapes remains unused.

Research has shown that the amount of 4MMP released during fermentation depends on the type of yeast strain used to complete the fermentation (Dubourdieu et al., 2006). It seems therefore that the genetic and physiological characteristics of the particular yeast strain have a significant effect on its ability to release thiols. Research at the AWRI confirmed these findings by showing that different wine yeast strains have varying abilities to release 4MMP from the Cys-4MMP-precursor in model fermentations (Howell et al., 2004). The ability of different commercial wine yeast strains to convert 3MH to 3MHA was also further investigated. Large variations in the ability of commercial wine yeasts to convert 3MH were observed and in most instances this did not correlate with the ability to release 4MMP (Swiegers et al., 2005). In a more recent study AWRI researchers showed that different commercial wine yeast strains, as a result of the variation in thiol release, may adjust the aroma of Sauvignon blanc dramatically. In this study Anchor VIN7 resulted in the highest 4MMP concentration and Anchor VIN13 in the highest 3MH concentration (Swiegers et al., 2006a). Winemakers preferred the wines made with yeasts that release high concentrations of thiols. A correlation therefore seems to exist between wine preference and thiol concentrations in certain wine styles.

Co-inoculations refer to the simultaneous inoculation of two or more yeasts in a given grape juice. The yeasts conduct the fermentation jointly. Previous research has shown that co-inoculations of several yeast strains adjust the chemical and sensorial characteristics of wine, compared to single yeast strain fermentations or an equal blend of single yeast strain fermentations (Rojas et al., 2003; Howell et al., 2006). Our hypothesis is that yeast strains interact and exchange metabolites during a co-fermentation. This hypothesis was partly proved by comparing the redox potential of co-fermentations to single yeast strain fermentations (Cheraiti et al., 2005). The purpose of the study below was to determine whether co-inoculations of two or more commercial wine yeast strains enhance the aroma profile, especially the volatile thiols, of Sauvignon blanc wine. The studies were conducted on 2006/2007 Sauvignon blanc grapes from the Adelaide Hills (Adelaide, Australia).

Experimental procedure

In 2006 unfiltered and homogenised Sauvignon blanc juice was fermented using two commercial yeast strains, Anchor VIN7 and Lalvin QA23, as well as a 50:50 combination of Anchor VIN7 / Lalvin QA23. In 2007 the Sauvignon blanc juice was fermented using three experimental yeast combinations, Anchor Alchemy I, Anchor Alchemy II and yeast combination III, as well as four standard commercial yeast strains (some of which occur in the Alchemy yeast combinations). The yeast cells were inoculated in active, dry form at 30 g/HL to produce a cell population of 5 X 106 cells/ml in the grape must. Fermentations were executed at 15C in 20-L closed, stainless steel cylinders. Sugar concentrations were monitored during the fermentation process. The identity of the yeast strains in the lees was determined once the fermentations were completed. Chemical analyses of the wines, including pH, titratable acid, residual sugar, alcohol and volatile acid, were conducted after bottling by the analytical services of the AWRI. Two months after bottling a formal, descriptive sensorial analysis of the 2006 wines was conducted by 12 trained judges. Thiol analyses were conducted by the SARCO Laboratoire in France.


Fermentation and chemical analysis of the 2006 Adelaide Hills Sauvignon blanc wines

All the wines which were made with VIN7, QA23 and the VIN7/QA23 co-inoculation fermented dry. Residual sugar concentrations were below 3 g/L. The VIN7/QA23 co-inoculation had similar fermentation tempos to that of the QA23 fermentations, while the VIN7 fermentations were slightly slower. In the VIN7/QA23 co-inoculation both yeast strains were present at the end of the fermentation. However, the initial 50:50 VIN7/QA23 inoculation ratio was adjusted during the fermentation process to the extent that VIN7 made up only about 10% and QA23 90% of the yeast population at the end of the fermentation. The reason for this phenomenon could possibly be because VIN7 usually has a slow fermentation tempo. Nevertheless it did not have a dramatic influence on the overall fermentation.

Table 1 indicates the basic chemical analysis of the wine. All parameters between the different fermentations were relatively the same, except for the volatile acid which varied considerably. The VIN7 wines had the highest volatile acid concentrations (0.84 g/L). The VIN7/QA23 co-inoculation wine had a lower volatile acid concentration (0.45 g/L), closer to the volatile acid concentration of the QA23 wine (0.40 g/L). It was interesting to see that the VIN7/QA23 co-inoculation wine had lower volatile acid concentrations than the VIN7+QA23 wine blend (consisting of 50% VIN7 wine and 50% QA23 wine). This data, together with the aroma data below, proves that co-inoculations result in a unique wine aroma profile, which cannot be emulated by wine blends.

Chemical analyses of the wines’ aroma components showed that the co-inoculations resulted in a more complex composition of esters, higher alcohols and volatile acids (data not shown). In the case of volatile thiols, the analyses showed that the VIN7/QA23 co-inoculation wines had the highest concentration of 3MH and 3MHA (Figure 2). The 3MH and 3MHA flavour components produce grapefruit/passion-fruit and passion-fruit aromas respectively. The data indicates that the co-inoculations resulted in wines with enhanced aroma profiles.

The formal descriptive sensorial analysis showed that VIN7 wines had the highest acetic acid character which was caused by the high volatile acid concentration (Figure 3). The QA23 wines were high in floral/rose aromas and estery characters. As expected, the VIN7+QA23 wine blend had intermediary characters. The VIN7/QA23 yeast blend wines had the highest estery, floral/rose flavours and passion-fruit characters,which correlates with the high 3MH and 3MHA concentrations.

The three yeast combinations and the four commercial yeast strains fermented dry. Yeast strain C fermented the quickest, followed by Alchemy I, Alchemy II and yeast combination III, yeast strain B, yeast strain D and yeast strain A. Table 2 indicates the basic chemical analyses of the wines. All parameters, except for the volatile acid, were relatively constant among the wines.

Volatile thiol analyses indicated that Alchemy I and Alchemy II resulted in the highest 3MH and 3MHA concentrations in the wines (Figure 4). Informal trials in the Australian wine industry indicated that these yeast combinations resulted in highly aromatic white wines (especially Sauvignon blanc) with citrus and passion-fruit flavours, most likely as a result of the high thiol concentrations. These wines are currently subject to a formal sensorial analysis.


Co-inoculations, using both VIN7/QA23 and the Anchor Alchemy yeast combinations, have enhanced chemical and sensorial aroma profiles when compared to single yeast strain fermented wines and blends of the latter single yeast strain fermentation wines. This phenomenon is most likely due to metabolic interactions between wine yeasts (Howell et al., 2006). At the moment very little information is available about the metabolic interaction of various wine yeasts during fermentation. Researchers at the AWRI are currently involved in generating more information about possible yeast-yeast interactions during wine fermentation.

The low thiol concentrations of yeast combination III compared to Alchemy I and Alchemy II emphasise the importance of correct ratios of yeasts in a yeast combination, seeing that all three combinations contain the same yeasts in different ratios. It was interesting to note than no 4MMP could be measured in the 2007 wines. This may be due to the dramatic climatic conditions in Australia during the 2006/2007 period, when the country suffered a severe drought. Due to the composition of yeast combination III, it is expected to have high 4MMP concentrations provided sufficient precursors are available in the grapes.

Sauvignon blanc wines produced using VIN7/QA23 co-inoculations had increased passion-fruit and floral flavours. Preliminary data indicate that the Alchemy I and Alchemy II yeast combinations resulted in increased fruity flavour in white wines, especially Sauvignon blanc. Future research at the AWRI aims to determine the relationship between fruity white wines and consumer preferences. In today’s competitive wine market it is imperative for the consumer to be one of the most important focus points in our thinking around technological development for the wine industry.

From the above study it is clear that yeast combinations are potentially a powerful mechanism to increase fruitiness in white wines, thereby increasing the wine’s consumer appeal. The Anchor Alchemy yeast combination series, which was developed in collaboration with the AWRI, will be available commercially in certain countries in 2008.


Research on flavour active wine yeasts at the AWRI is funded by the Australian wine industry via their investment body, Grape and Wine Research Development Corporation. We would like to thank Anchor Yeast for their financial contribution to this project. Thanks are due to AWRI researchers Robyn Willmott and Belinda Bramley for their contribution. Our thanks also to Chris Day of Provisor for making the wine and Marie-Laure Murat of SARCO Laboratoire (France) for the thiol analyses.


Allen, M.S., Lacey, M.J., Harris, R.L.N., Brown, W.V. 1991. Contribution of methoxypyrazines to Sauvignon Blanc aroma. Am. J. Enol. Vitic., 42, 109 – 112.

Augustyn, O.P.H., Rapp, A., Van Wyk, C.J. 1982. Some volatile aroma compounds of Vitis vinifera L. cv. Sauvignon Blanc. S. Afr. Enol. Vitic., 3, 53 – 60.

Cheraiti, N., Guezenec, S., Salmon, J.M. 2005. Redox interactions between Saccharomyces cerevisiae and Saccharomyces uvarum in mixed culture under enological conditions. Appl. Environ. Microbiol., 71, 255 – 260.

Darriet, P., Tominga, T., Lavigne, V., Boidron, J., Dubourdieu, D. 1995. Identification of a powerful aromatic compound of Vitis vinifera L. var. Sauvignon wines: 4-Mercapto-4-methylpentan-2-one. Flavour Fragrance J., 10, 385 – 392.

Dubourdieu, D., Tominaga, T., Masneuf, I., Peyrot Des Gachons, C., Murat, M.L. 2006. The role of yeast in grape flavour development during fermentation: the example of Sauvignon Blanc. Am. J. Enol. Vitic., 57, 81 – 88.

Howell, K.S., Swiegers, J.H., Elsey, G.M., Siebert, T.E., Bartowksy, E.J., Fleet, G.H., Pretorius, I.S., de Barros Lopes, M.A. .2004. Variation in 4-mercapto-4-methylpentan-2-one release by Saccharomyces cerevisiae commercial wine strains. FEMS Micro. Lett., 240, 125 – 129.

Howell, K.S., Klein, M., Swiegers, J.H., Hayasaka, Y., Elsey, G.M., Fleet, G.H., Hj, P.B., Pretorius, I.S., de Barros Lopes, M.A. 2005. Genetic determinants of volatile thiol release by Saccharomyces cerevisiae during wine fermentation. Appl. Environ. Microbiol., 71, 5420 – 5426.

Howell, K.S., Cozzolino, D., Bartowsky, E., Fleet, G.H., Henschke, P.A. 2006. Metabolic profiling as a tool for revealing Saccharomyces interactions during wine fermentation. FEMS Yeast Res., 6, 91 – 101.

Marais, J., Hunter, J.J. and Haasbroek, P.D. 1999. Effect of canopy microclimate, season and region on Sauvignon Blanc grape composition and wine quality. S. Afr. J. Enol. Vitic., 20: 19 – 30.

Murat, M.L., Masneuf, I., Darriet, P., Lavigne, V., Tominga, T., Dubourdieu, D. 2001a. Effect of Saccharomyces cerevisiae yeast strains on the liberation of volatile thiols in Sauvignon blanc wine. Am. J. Enol. Vitic., 52, 136 – 139.

Murat, M.L., Tominaga, T., Dubourdieu, D. 2001b. Assessing the aromatic potential of Cabernet Sauvignon and Merlot musts used to produce rose wine by assaying the cysteinylated precursor of 3-mercaptohexan-1-ol. J. Agric. Food Chem., 49, 5412 – 5417.

Rojas, V., Gill, J.V., Piaga, F., Manzanares, P. 2003. Acetate ester formation in wine by mixed cultures in laboratory fermentations. Int. J. Food Microbiol., 86, 181 – 188.

Swiegers, J.H., Bartowksy, E.J., Henschke, P.A., Pretorius, I.S. 2005. Yeast and bacterial modulation of wine aroma and flavour. Austral. J. Grape Wine Res., 11, 127 – 138.

Swiegers, J.H., Capone, D.L., Elsey, G.M., Sefton, M.A., Bramley, B.R., Francis, I.L., Pretorius, I.S. 2007. Engineering volatile thiol release in Saccharomyces cerevisiae for improved wine aroma. Yeast, 24, 561 – 574.

Swiegers, J.H., Francis, I.L., Herderich, M.J., Pretorius, I.S. 2006a. Meeting consumer expectations through management in vineyard and winery: The choice of yeast for fermentation offers great potential to adjust the aroma of Sauvignon Blanc wine. Austral. NZ. Wine Ind. J. 21, 34 – 42.

Swiegers, J.H., Willmott, R., Hill-Ling, A., Capone, D.L., Pardon, K.H., Elsey, G.M., Howell, K.S., de Barros Lopes, M.A., Sefton, M.A., Lilly, M., Pretorius, I.S. 2006b. Modulation of volatile thiol and ester aromas in wine by modified wine yeast. In Developments in Food Science 43, Flavour Science Recent Advances and Trends, Bredie WLP, Petersen MA (eds). Elsevier: Amsterdam, The Netherlands.

Tominaga, T., Masneuf, I. and Dubourdieu, D. 1995. A S-cysteine conjugate, precursor of aroma of white Sauvignon. J. Int. Sci. Vigne Vin. 29, 227 – 232.

Tominaga, T., Baltenweck-Guyot, R., Peyrot des Gachons, C., Dubourdieu, D. 2000. Contribution of volatile thiols to the aromas of white wines made from several Vitis vinifera grape varieties. Am. J. Enol. Vitic., 51, 178 – 181.

Hentie Swiegers 1, Ellie King 2, Brooke Travis 1, Leigh Francis 1 & Sakkie Pretorius 1

1 The Australian Wine Research Institute, PO Box 197, Glen Osmond, Adelaide, SA 5064, Australia, email:
2 School of Agriculture, Food and Wine, University of Adelaide, Private Bag 1, Glen Osmond, Adelaide, SA 5064, Australia


You may like to read these:

Go Back