International Journal of Food Microbiology 108 (2006) 385 390 www.elsevier.

com/locate/ijfoodmicro

Influence of fermentation temperature on volatile thiols concentrations in Sauvignon blanc wines

Isabelle Masneuf-Pomarde a,, Chantal Mansour b , Marie-Laure Murat b , Takatoshi Tominaga c , Denis Dubourdieu c
b a ENITA de Bordeaux, 1 Cours du Gnral de Gaulle, BP 201, CS 40201, 33175 Gradignan cedex, France SARCO laboratory, B.P. 40, 33015 Bordeaux cedex, France, Socit d'Application de Recherche et de Conseil nologique, research subsidiary of Laffort nologie c Facult d'nologie, Universit Victor Segalen, Bordeaux 2, 351 cours de la Libration, 33400 Talence, France

Received 25 May 2005; received in revised form 6 December 2005; accepted 7 January 2006

Abstract The effect of Saccharomyces cerevisiae strains on the amount of 4-mercapto-4-methylpentan-2-one, a major varietal aroma of Sauvignon blanc wines, was demonstrated by previous research work. However, the influence of different alcoholic fermentation parameters on the levels of volatile thiols (4-mercapto-4-methylpentan-2-one, 3-mercaptohexan-1-ol and 3-mercaptohexyl acetate) in wines has not yet been investigated. The impact of fermentation temperature on the final amount of volatiles thiols and on some other analytical parameters (ethanol, total acidity, residual sugars, volatile acidity) was determined in a model medium and in grape juice. Interaction between fermentation temperature and yeast strain was also tested. The fermentation temperature influenced the amount of volatile thiols irrespective of the yeast strain used. The final levels of 4MMP and 3MH in model medium and in wines were higher when the alcoholic fermentation is conducted at 20 C than at 13 C. The 3MHA, which was correlated with the amount of 3MH determined in wines, was also higher when the alcoholic fermentation was conducted at 20 C. From a technological point of view, the choice of yeast strain and fermentation temperature has a decisive influence on the concentrations of the varietal aromas of Sauvignon blanc wines. 2006 Elsevier B.V. All rights reserved.
Keywords: Volatile thiols; Cysteinylated precursors; Fermentation temperature; Yeast strains

1. Introduction Many factors such as must composition and juice clarification, the temperature of fermentation, the yeast strain inoculated strongly affect alcoholic fermentation and the aromatic profile of white and ros wines (Ollivier et al., 1987; Aragon et al., 1998; Antonelli et al., 1999; Ribreau-Gayon et al., 2000). Low temperature enhanced the wine contents of some volatile compounds produced by the yeast during the alcoholic fermentation (esters, acetates, medium-chain fatty acids) (Killiam and Ough, 1979; Cotrell and MC Lellan, 1986; Torija et al., 2003). For this reason, low temperature fermentation (10

Corresponding author. Fax: +33 557. E-mail address: i-masneuf@enitab.fr (I. Masneuf-Pomarde). 0168-1605/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ijfoodmicro.2006.01.001

15 C) is of interest to winemakers to enhance the production of some volatile compounds and improve the wine aromatic profile. However, this low temperature can easily cause sluggish or stuck fermentations. In some simple or non-floral grape varieties, like Vitis vinifera L. var. Sauvignon blanc, the characteristic varietal aromas that are synthesized in wines during the alcoholic fermentation were shown to be due to volatile thiols ( Darriet et al., 1995; Tominaga et al., 1996, 1998a). Three major aromatic volatile thiols, 4-mercapto-4methylpentan-2-one (4MMP), 3-mercaptohexan-1-ol (3MH) and 3-mercaptohexyl acetate (3MHA), were identified as responsible for the box tree, grapefruit and passion fruit nuances of wines made from this variety. These compounds are also present in wines made from Gewrztraminer, Riesling, Colombard, Petit manseng, Cabernet Sauvignon and Merlot grapes (Tominaga et al., 2000; Murat et al., 2001a). Two sulphur

386

I. Masneuf-Pomarde et al. / International Journal of Food Microbiology 108 (2006) 385390

compounds (4MMP and 3MH) exist in the must in a non-volatile form as an S-cysteine conjugate (Tominaga et al., 1995, 1998a; Murat et al., 2001b). The volatile thiols are released during the alcoholic fermentation due to the transformation of the corresponding S-cysteine conjugate by the yeast (Tominaga et al., 1998a). Recent investigations using genetic screens led to the identification of four genes that influence the release of the volatile thiol 4MMP in laboratory strain (Howell et al., 2005). However, the mechanisms of transformation of cysteinylated precursors by yeasts into aroma remained unknown. Yeasts from Saccharomyces cerevisiae have a significant influence, variable according to the specific strain, on the concentrations of 4MMP in wines (Murat et al., 2001c; Howell et al., 2004). S. bayanus var. uvarum strains and hybrids S. cerevisiae S. bayanus var. uvarum were shown to present a high ability to release the volatile thiols from their natural precursors (Masneuf et al., 2002). The choice of yeast strain has a decisive impact on the varietal aroma of wines made from V. vinifera L. var. Sauvignon blanc. The aim of the present work was to study the impact of different fermentation temperatures on the amount of volatile thiols in Sauvignon blanc wines. The combined effect of fermentation temperature and starter cultures commonly used in white winemaking is also tested. 2. Materials and methods 2.1. Yeast strains Four industrial strains (A, B, C, D) and one hybrid strains S. cerevisiae S. bayanus var. uvarum, noted H9, (Masneuf et al., 2002) were chosen in this study because of their good ability to liberate the volatile thiols from their odorless precursor. All the strains were Active Dry Yeast. They were rehydrated in model fermentation medium or must diluted with warm water (1 : 1) at 37 C for 30 min. 2.2. Fermentation experiments The model fermentation medium used was described by Marullo et al. (2004). This medium was strongly buffered to pH 3.3 and contained (g l 1): glucose (85), fructose (85), L + tartaric acid (3), citric acid (0.3), and L-malic acid (0.3); nitrogen source: 190 mg l 1 available nitrogen provided by 300 mg l 1 (NH4)2SO4 and 600 mg l 1 asparagine; Mineral salts (mg l 1): KH2PO4 (2000), MgSO47H2O (200), MnSO4 H2O (4), ZnSO4 7H2O (4), CuSO45H2O (1), KI (1), CoCl2 6H2O (0.4), (NH4)6Mo7O244H2O (1), and H3BO3 (1); 5 vitamins (g): myo-inositol (300), biotin (0.04), thiaminHCl (1), pyridoxineHCl (1), nicotinic acid (1), calcium panthothenate (1), and para-amino benzoic acid (1). The low level of sugars was chosen to ensure that the alcoholic fermentation would run completely even at low temperature. The medium was sterilized by filtration through a 0.45 m nitrate-cellulose membrane and supplemented with sulfur dioxide (20 mg l 1) corresponding to enological treatment. A sample of 10 nmol l 1 of S-4-(4methylpentan-2-one)-L-cystein and of 2000 nmol l 1 of S-3-

(hexan-1-ol)-L-cystein (Tominaga et al., 1998b) was added in this medium. The concentration of S-4-(4-methylpentan-2one)-L-cystein was chosen according to the amount previously found in Sauvignon blanc must whereas the concentration of S-3-(hexan-1-ol)-L-cystein was 10 times higher (Peyrot des Gachons et al., 2005). Fermentation was carried out in 750 ml sterile bottles in triplicate in incubators at three temperatures (13, 20 and 24 C) usually used in dry white winemaking. During the harvest in 2003, Sauvignon blanc must and solids were collected from two wineries (Bordeaux) noted R and L. Sugar concentrations (g l 1) for must R and L were 198 and 185 respectively. This concentration was corrected in must L in order to obtain in wine 12% v/v of ethanol. In the laboratory, turbidity was adjusted to 200 NTU by adding the must solids (Ollivier et al., 1987). Fermentation was carried out in 350 ml sterile bottles in duplicates at two temperatures (13 and 24 C). After two days of alcoholic fermentation, 6 mg l 1 of oxygen was added by air bubbling according to white winemaking practical treatment in cellars. In both experiments, the fermentative medium was inoculated with 200 mg l 1 of rehydrated dry yeast. When the reducing sugar content was less than 2 g l 1, SO2 was added at 50 mg l 1 and model media or wines were kept at 4 C before analyzed. 2.3. CHEF (contour clamped homogeneous electric field) gel electrophoresis At mid-fermentation, model medium or must samples of 500 l were used to inoculate a 50 ml sterile flask containing YPD liquid medium (Glucose 20 g l 1, yeast extract 10 g l 1 and polypeptone g l 1). After 24 h in culture at 25 C, chromosomal DNA was prepared in agarose plugs from total biomass obtained after centrifugation of 30 ml of YPD liquid medium (Frzier and Dubourdieu, 1992). Comparative analysis of the karyotypes of the inoculated yeast strain and the total biomass was used to check the implantation of the strains (Frzier and Dubourdieu, 1992). A CHEF DRII apparatus (Bio-Rad, Richmond, California) was used to separate chromosomal DNAs. Electrophoresis buffer (0.5 TBE) was circulated around the gel and cooled to 14 C. Electrophoresis was carried out at 200 V for 15 h with a switching time of 60 s and then for 9 h with a switching time of 90 s. 2.4. Chemical analysis The quantity of volatile thiols released in the model medium (4MMP and 3MH) and wines (4MMP, 3MH and 3MHA) was measured using the method described by Tominaga et al. (2003). Ethanol concentration (vol.%) was measured by the infrared reflectance method (infra Analyser 450 Technicon, Bran+Luebbe, Plaisir, France). Volatile acidity and residual sugars were analyzed after distillation by colorimetry (A460 nm) in continuous flux (Sanimat, Montauban, France). Total acidity was analysed by potentiometry with a vinilog V2 (Sanimat, Montauban, France).

I. Masneuf-Pomarde et al. / International Journal of Food Microbiology 108 (2006) 385390

387

was previously demonstrated (Murat et al., 2001c). All the fermentations were completed (residual sugars were below 2 g l 1). The duration of alcoholic fermentation was 9 to 10, 19 to 21 and 24 to 28 days at 24, 20 and 13 C respectively. The results regarding the amount of 4MMP, 3MH and volatile acidity are presented in the Fig. 1a, b and c respectively. For the strains VL3c and VIN13, the amount of 4MMP and 3MH decreased with the fermentation temperature. This decrease in the level of 4MMP and 3MH is dramatically significant for the strain VL3c between 20 and 13 C. The amount of 3MH liberated by the strain NW3 decreased with fermentation temperature. However, for that strain, there was no relation between the levels of 4MMP and the temperature of the alcoholic fermentation. Strains VL3c and NW3 produced higher amount of volatile acidity at 13 C compared with 20 and 24 C. A different behavior is observed for the strain VIN13, which produced less of volatile acidity compared to the other strains. The lower amount of volatile acidity for the strain Vin13 is obtained at 13 and 20 C. Moreover, the results clearly showed the influence of the yeast strains on the liberation of the 4MMP (Fig. 1a) and on the acetic acid production (Fig. 1c). As for the 4MMP, there is a significant effect of the yeast strain studied on the amount of 3MH liberated from their cysteinylated precursors added to model medium. The higher amounts of 4MMP and 3MH were obtained with the strain VIN13. The values of 4MMP, 3MH and volatile acidity are submitted to a double-factor (strain temperature) analysis of variance with repetition ( = 0.05). The ANOVA test was statistically significant for both factors (results not shown).
Fig. 1. Effect of the fermentation temperature on the production of 4MMP (a), 3MH (b), and volatile acidity (c) by the three S. cerevisiae strains in the synthetic medium. Values are the mean of three independent fermentations.

3.2. Effect of the temperature and the yeast strains on analytical parameters and on the liberation of volatile thiols in grape juice Fermentation experiments were conducted at two temperatures (13 and 20 C) using Sauvignon blanc must originated from two wineries (noted R and L) and collected in 2003. Three S. cerevisiae yeast strains (VL3c, VIN13 and X5) and one hybrid strain S. cerevisiae S. bayanus var. uvarum (H9) were
Table 1 Effect of the fermentation temperature and the yeast strains on the residual sugars, total acidity, ethanol and volatile acidity concentrations in white wines (average of two replicates) Strains TA (g l 1 Residual 1 sugars (g l ) tartaric acid) Ethanol (% v/v) VA (g l 1 acetic acid)

2.5. Statistical analysis The analytical parameters were subjected to double-factor analysis of variance (yeast strain/temperature) with repetitions ( = 0.05, ANOVA, Statbox software). A Principal Components Analysis was performed by using Statbox Software. The PCA was carried out on the data obtained for must experiments. This analysis generated the components that can be used to represent the main differences between strains and temperature tested. 3. Results 3.1. Effect of the temperature and the yeast strains of S. cerevisiae on the liberation of volatile thiols and the production of volatile acidity in model medium The concentrations of volatile thiols obtained from their cysteinylated precursors added to the model medium at the end of the alcoholic fermentation were measured for three S. cerevisiae strains (VL3c, NW3 and VIN13) at three different temperatures (13, 20 and 24 C). The absence of volatile thiols in unfermented control under the same experimental conditions

13 C 20 C 13 C 20 C 13 C 20 C 13 C 20 C Wine R VL3c 3.4 VIN13 0.6 X5 2.0 H9 2.7 Wine L VL3c 15.9 VIN13 1.1 X5 2.3 H9 3.4 0.7 0.8 0.9 0.7 0.8 1.0 0.9 1.7 5.30bc 5.12d 5.16cd 4.97e 5.43e 5.58e 5.51e 5.99d 5.35b 5.30bc 5.58a 5.30bc 6.61a 6.23c 6.30bc 6.42b 12.25 12.22 12.27 12.12 11.90 11.90 11.90 11.70 12.27 12.17 12.35 12.15 11.95 11.80 11.95 11.67 0.39a 0.22d 0.34b 0.08f 0.29a 0.12d 0.21b 0.06f 0.27c 0.21d 0.34b 0.12e 0.19c 0.19c 0.21b 0.09e

Values followed by different letters are statistically different (double-factor analysis of variance with repetition, = 0.05, ANOVA).

388

I. Masneuf-Pomarde et al. / International Journal of Food Microbiology 108 (2006) 385390

Table 2 Effect of the fermentation temperature and the yeast strains on the liberation of 4MMP, 3MH, and 3MHA (average of two replicates) Strains 4MMP (ng l 1) 13 C Wine R VL3c VIN13 X5 H9 VL3c VIN13 X5 H9 nd nd nd 5 nd nd 3 18 20 C 5 7 14 16 9 17 17 23 3MH (ng l 1) 13 C 222c 1664b 392c 243c 168 205 162 164 20 C 2564a 3002a 3198a 837c 312 311 289 200 3MHA (ng l 1) 13 C 114c 479b 175c 112c 22c 26c 22c 24c 20 C 519b 866a 878a 360bc 56b 79a 46b 57b

Wine L

Values followed by different letters are statistically different (double-factor analysis of variance with repetition, = 0.05, ANOVA). nd: not detected.

tested. At mid-fermentation, the analysis of the karyotypes of the total biomass for each yeast tested confirmed the implantation of all strains in the two experiments (data not shown). The duration of the alcoholic fermentation was 8 to 13 and 21 to 25 days at 20 and 13 C respectively. Results regarding the amounts of the residual sugars, the total acidity, the ethanol and volatile acidity in the white wines are presented in Table 1. The levels of residual sugars were higher at 13 C than at 20 C, except for the strain VIN13. For the strains VL3c and H9, the alcoholic fermentation was not completed at low temperature. No significant differences were noticed regarding the level of ethanol whatever the yeast strains used or the fermentation temperature. Irrespective of the yeast strains used, the level of total acidity was higher at 20 C than at 13 C for both experiments. Concerning the production of volatile acidity, the behavior of strains X5 was the same at 13 and at 20 C. In contrast, the strain VL3c seems to be strongly affected by the fermentation temperature concerning production of acetic acid, which is higher at 13 C than at 20 C. The level of volatile

acidity produced by the hybrid strain H9 in both experiment is very low in contrast with S. cerevisiae strains. Three volatile thiols (4MMP, 3MH and 3MHA) were quantified at the end of the alcoholic fermentation in the wines R and L (Table 2). The 4MMP was not detected in wines fermented by strains VL3c and VIN13 in the two experiments at 13 C. The amount of that compound was systematically higher at 20 C than 13 C whatever the yeast strains used and the origin of the must. Strains X5 and H9 released major quantities of 4MMP in the two experiments. The influence of the yeast strain on the amounts of 3MH was statistically significant only for the wine R. However, significant differences between the yeast strains on the amount of 3MHA were obtained for both experiments. At 20 C all the strains tested released between 2 fold (wine R, strain VIN13) and up to 10 fold (wine R, strain VL3c) more 3MH than when fermented at 13 C. For the 3MHA levels, the same phenomenon was observed with between 2 fold (wines L) and 3 fold (wines R) more compounds in wines fermented at 20 C than in wines fermented at 13 C. The higher level of 3MH and 3MHA for the wine R was obtained with the strains VIN13 and X5. There is a linear correlation between the amount of 3MH and 3MHA measured in wines for the two experiments whatever the strains considered and the temperature of alcoholic fermentation tested (correlation coefficient = 0.95). The Principal Component Analysis (PCA) was applied to the results obtained for the seven parameters measured in wines R and L (residual sugars, total acidity, ethanol, volatile acidity, 4MMP, 3MH and 3MHA) and the four strains tested. The PCA give similar results for both experiments (results not shown). As an example, the PCA applied to the results obtained for the wine L is reported Fig. 2. The temperature is used as additional qualitative variable. The projection conserves 85% of the information with two components, explaining respectively 35% and 50% of the total inertia. Axis 1 clearly separates two

Fig. 2. Principal Component Analysis (PCA) of four strains for seven enological parameters. PCA was carried out on the values regarding the must L.

I. Masneuf-Pomarde et al. / International Journal of Food Microbiology 108 (2006) 385390

389

temperatures. The groups of strains corresponding to 20 C are located on the right of the axis 1 and corresponded to the high values of total acidity, 4MMP, 3MH and 3MHA. The position of sugars on the axis 1 is explained by the level of residual sugars obtained at 13 C. The axis 2 discriminates the strain H9 cluster from the other strains. Strain H9 cluster is principally separated from the other strains by the amount of residual sugars and 4MMP on one hand, and by volatile acidity and ethanol production on the other hand. 4. Discussion Several studies have described the influence of the temperature and the yeast strains on the production of secondary metabolites during the alcoholic fermentation and the final quality of wine (Aragon et al., 1998; Antonelli et al., 1999; Reynolds et al., 2001). We reported in previous work the effect of winemaking yeast species and strains on some grape aromas (Murat et al., 2001c; Masneuf et al., 2002). The impact of yeast strains on the liberation of 4MMP from its corresponding cysteinylated precursors was clearly demonstrated. The effect of yeast strains on the levels of 3MH at the end of the alcoholic fermentation was previously reported in ros wines made from Merlot, Cabernet franc and Cabernet sauvignon (Murat, 2001). In our experimental conditions, it is possible to report the influence of the yeast strain on the liberation of 3MH from its synthetic precursor in model medium and from natural precursor in Sauvignon blanc must. This result is, nevertheless, not systematic and seems to be strongly dependent on the must origin. Our data confirm the high liberating ability of the hybrid strain to liberate 4MMP. However, this was not confirmed for 3MH amounts in wines. Beside the yeast strain effect, the aim of the work was to evaluate the influence of fermentation temperature on some analytical parameters and on the amounts of volatile thiols in Sauvignon blanc wines. The PCA analysis showed that four parameters among the seven studied are correlated with high fermentation temperature. The values for total acidity are higher at 20 C than at 13 C. Increased total acidity with increasing fermentation temperature was previously reported by C.S. Ough et al. (1969). It may be partially due to the production of significant amounts of acetic and succinic acids with increasing temperature. In our conditions, more tartaric acid may have precipitated when alcoholic fermentation is conducted at 13 C thus explaining the higher level of total acidity found in wines obtained at 20 C. Irrespective of the yeast strain used and the fermentative medium, the results presented here clearly show that fermentation temperature influences the amount of volatile thiols in wines. The final levels of 4MMP and 3MH in wines were higher when the alcoholic fermentation was conducted at 20 C than at 13 C. The influence of fermentation temperature (18 and 28 C) on the amount of 4MMP released by different yeast strains was previously studied by Howell et al. (2004) in synthetic medium. The precursor and 4MMP concentrations used in their experimental conditions were

respectively 10,000 and 1000 times higher than concentrations tested in our work. The authors reported a positive effect of high temperature on 4MMP levels for two yeast strains tested. These results, obtained for a fermentation temperature of 28 C compared with 18 C, are in agreement with our results for fermentation temperatures of 20 and 13 C. The formation and retention of fermentation esters, such as acetate, were reported to be affected by fermentation temperature (Killiam and Ough, 1979). The authors found greater amounts of ethyl acetate, isoamyl acetate and n-hexyl acetate in wines fermented at 10 C than at 20 C. In our study, concentrations of 3MH acetate are higher in samples fermented at high temperature. The 3MHA levels are also well correlated with the amount of 3MH. Thus, in the case of 3MHA, the temperature does not seem to have an impact on yeast metabolism but an indirect effect on the level of 3MHA through the amount of 3MH obtained in wines. The mechanisms by which the odorless precursors are converted into aromas by yeast during alcoholic fermentation have not yet been fully elucidated. A -lyase type enzyme activity found in baker's and brewer's yeasts was shown to be able to hydrolyze S-cysteine conjugates and release the corresponding thiols (Huynh-Ba and Tuong, 1997; Tominaga et al., 1998a). This enzyme activity is probably present in winemaking yeast. Not knowing this mechanism, it is difficult to make hypothesis regarding the influence of fermentation temperature on yeast cysteinylated precursors metabolism. It was previously described that low fermentation temperature enhanced the aromatic characteristics of wines, possibly because of greater synthesis and a greater retention of volatile flavors (Ribreau-Gayon et al., 2000). Our results concerning levels of volatile thiols in wines show opposite conclusions. Fermentation temperature can influence the membrane fattyacyl composition (Hunter and Rose, 1972). The unsaturation degree of membrane fatty-acyl is higher at low temperature (Watson, 1987). Some authors reported that a higher saturation degree in the membrane speeds up the removal of ethanol and volatile compounds from yeast (Rotmann and Rehm, 1991; Rosi and Bertuccioli, 1992). The role of the yeast membrane on the transport and excretion of the volatile thiols and the relation with its fatty-acyl composition have not yet been studied. The change of membrane fatty acids depending on fermentation temperature can be one hypothesis to explore the negative impact of low temperature on the level of some varietal aromas in Sauvignon blanc wines. The response of yeast to the fermentation temperature is variable depending of the strains tested. Two strains tested (VIN13 and X5) have nearly similar fermentative properties at low and high temperature: complete alcoholic fermentation and same amount of volatile acidity produced. Fermentative properties of strain VL3c are altered by low fermentation temperature (stuck fermentation and increasing production of volatile acidity). As shown in previous studies (Kishimoto, 1994; Shinohara et al., 1994), the hybrid strain has distinct fermentative properties from S. cerevisiae strains in all wine parameters tested: low level of volatile acidity and ethanol, stuck fermentation.

390

I. Masneuf-Pomarde et al. / International Journal of Food Microbiology 108 (2006) 385390 segregation of industrial wine yeast strains. FEMS Yeast Res. 7 (4), 711719 (May). Masneuf, I., Murat, M.L., Naumov, G., Tominaga, T., Dubourdieu, D., 2002. Hybrids Saccharomyces cerevisiae Saccharomyces bayanus var. uvarum having a high liberating ability of some sulfur varietal aromas of Vitis vinifera Sauvignon blanc wines. J. Int. Sci. Vigne Vin 36 (4), 205212. Murat, M.L., 2001. Recherches sur la vinification des vins ross et clairets de BordeauxDiplme d'Etudes et de Recherches de l'Universit de Bordeaux II. Murat, M.L., Tominaga, T., Dubourdieu, D., 2001a. Mise en vidence de composs clefs dans l'arme des vins ross et clairets de Bordeaux. J. Int. Sci. Vigne Vin 35 (2), 99105. Murat, M.L., Tominaga, M.L., Dubourdieu, D., 2001b. Assessing the aromatic potential of cabernet sauvignon and Merlot musts used to produce ros wine by assaying the cysteinylated precursor of 3-mercaptohexan-1-ol. J. Agric. Food Chem. 49 (11), 54125417. Murat, M.L., Masneuf, I., Darriet, P., Lavigne, V., Tominaga, T., Dubourdieu, D., 2001c. Effect of the Saccharomyces cerevisiae yeast strains on the liberation of volatile thiols in Sauvignon blanc wines. Am. J. Enol. Vitic 52 (2), 136139. Ollivier, C., Stonestreet, T., Larue, F., Dubourdieu, D., 1987. Incidence de la composition colloidale des mots blancs sur leur fermentescibilit. Conn. Vigne Vin 21, 5970. Ough, C.S., Amerine, M.A., Sparks, T.C., 1969. Studies with controlled fermentations. XI Fermentation temperature effects on acidity and pH. Am. J. Enol. Vitic. 20, 127139. Peyrot des Gachons, C., Van Leeuwen, K., Tominaga, T., Soyer, J.P., Gaudillre, J.P., Dubourdieu, D., 2005. Influence of water and nitrogen deficit on fruit ripening and aroma potential of Vitis vinifera L cv Sauvignon blanc in field conditions. J. Sci. Food Agric. 85, 7385. Reynolds, A., Cliff, M., Girard, B., Kopp, T.G., 2001. Influence of fermentation temperature on composition and sensory properties of Semillon and Shiraz wines. Am. J. Enol. Vitic. 52 (3), 235240. Ribreau-Gayon, P., Dubourdieu, D., Donche, B., Lonvaud, A., 2000. Handbook of enology. The Microbiology of Wine and Vinifications, vol. I. Wiley, West Sussex, England. Rosi, I., Bertuccioli, M., 1992. Influences of lipid addition on fatty acid composition of Saccharomyces cerevisiae and aroma characteristics of experimental wines. J. Inst. Brew. 98, 305314. Rotmann, B.H., Rehm, H.J., 1991. Relationship between fermentation capability and fatty acid composition of free and immobilized Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol. 34, 502508. Shinohara, T., Saito, K., Yanagida, F., Goto, S., 1994. Selection and hybridisation of wine yeasts for improved winemaking properties: fermentation rate and aroma productivity. J. Ferment. Bioeng. 77, 428431. Tominaga, T., Masneuf, I., Dubourdieu, D., 1995. A S-cysteine conjugate, precursor of aroma of white sauvignon. J. Int. Sci. Vigne Vin 29, 227232. Tominaga, T., Darriet, P., Dubourdieu, D., 1996. Identification de l'actate de 3mercaptohexanol, compos forte odeur de buis, intervenant dans l'arme des vins de Sauvignon. Vitis 35 (4), 207210. Tominaga, T., Furrer, A., Henry, R., Dubourdieu, D., 1998a. Identification of new volatile thiols in the aroma of Vitis vinifera L. var. Sauvignon blanc wines. Flavour Fragr. J. 13, 159162. Tominaga, T., Peyrot des Gachon, C., Dubourdieu, D., 1998b. A new type of flavor precursors in Vitis vinifera L. cv. Sauvignon Blanc: S-cysteine conjugates. J. Agric. Food Chem. 46, 52155219. Tominaga, T., Baltenweck-Guyot, R., Peyrot des Gachon, 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 (2), 178181. Tominaga, T., Guimberteau, G., Dubourdieu, D., 2003. Contribution of benzenemethanethiol to smoky aroma of certain Vitis vinifera L. wines. J. Agric. Food Chem. 51, 13731376. Torija, M.J., Beltran, G., Novo, M., Poblet, M., Guillamon, J.M., Mas, A., Rozs, N., 2003. Effects of fermentation temperature and Saccharomyces species on the cell fatty acid composition and presence of volatile compounds in wine. Int. J. Food Microbiol. 85, 127136. Watson, K., 1987. Temperature relations. In: Rose, A.H., Harrison, J.S. (Eds.), The Yeast. Yeast and Environment, vol. 2. Academic Press, London, pp. 4171.

5. Conclusion The results of our study clearly demonstrate that a high fermentation temperature (1820 C) is better for the expression of the varietal aroma of Sauvignon blanc wines and probably also for other varieties containing S-cystein conjugate precursors such as Gros manseng (Tominaga et al., 1998a,b) and Gewrztraminer (personal communication). This fermentation temperature is inexpensive (temperature control) for winemakers and preferred to prevent sluggish or stuck fermentations. However, some yeast strains seem to be well adapted to low fermentation temperature. The choice of yeast strain and fermentation temperature is important to control the varietal aroma of wines made from Sauvignon blanc grapes. Acknowledgements The authors acknowledge Olivier Lavialle (Ecole national des Travaux Agricoles, Bordeaux, France) for assistance in the statistical study, and Christine Barbe for her contribution to the English correction on the manuscript. Sauvignon blanc must were provided by Chateau Reynon (Bordeaux, France) and Chateau La Louvire (Pessac-Lognan, France). References
Antonelli, A., Castellari, L., Zambonelli, C., Carnacini, A., 1999. Yeast influence on volatile composition of wines. J. Agric. Food Chem. 47 (3), 11391144. Aragon, P., Atienza, J., Climent, M.D., 1998. Influence of clarification, yeast type and fermentation temperature of the organic acid and higher alcohols of Malvasia and Muscatel wines. Am. J. Enol. Vitic. 49 (2), 211219. Cotrell, T.H.E., MC Lellan, M.R., 1986. The effect of fermentation temperature on chemical and sensory characteristics of wines from seven white grape cultivars grown in New York State. Am. J. Enol. Vitic. 37 (3), 190194. Darriet, P., Tominaga, T., Lavigne, V., Boidron, J.N., Dubourdieu, D., 1995. Identification of a powerful aromatic component of Vitis vinifera L. var. Sauvignon wines: 4-mercapto-4-methylpentan-2-one. Flavour Fragr. J. 10, 385392. Frzier, V., Dubourdieu, D., 1992. Ecology of yeast strains Saccharomyces cerevisiae during spontaneous fermentation in a Bordeaux winery. Am. J. Enol. Vitic. 43 (4), 375380. Howell, K.S., Swiegers, J.H., Elsey, G.M., Siebert, T.E., Bartowsky, E.J., Fleet, G.H., Pretorius, I.S., De barros Lopes, M.A., 2004. Variation in 4-mercapto4-methyl-pentan-2-one release by Saccharomyces cerevisiae commercial wine strains. FEMS Microbiol. Lett. 240, 125129. Howell, K.S., Klein, M., Swiegers, 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 (9), 54205426. Hunter, K., Rose, A.H., 1972. Lipid composition of Saccharomyces cerevisiae as influenced by growth temperature. Biochim. Biophys. Acta 260, 639653. Huynh-Ba, Tuong, 1997. Brevet europen EP 0 770 686 A1, Socit des produits Nestle S. A., Bulletin 1997, 18. Killiam, E., Ough, C.S., 1979. Fermentation esters formation and retention as affected by fermentation temperature. Am. J. Enol. Vitic. 30 (4), 301305. Kishimoto, M., 1994. Fermentation characteristics of hybrids between the cryophilic wine yeast Saccharomyces bayanus and the mesophilic wine yeast Saccharomyces cerevisiae. J. Ferment. Bioeng. 77, 432435. Marullo, P., Bely, M., Masneuf, I., Aigle, M., Dubourdieu, D., 2004. Inheritable nature of enological quantitative traits is demonstrated by meiotic