Interactions Between Industrial Yeasts and Chemical Contaminants in Grape Juice Affect Wine Composition Profile

Abstract

The interaction between four industrial wine yeast strains and grape juice chemical contaminants during alcoholic fermentation was studied. Industrial strains of Saccharomyces cerevisiae (AWRI 0838), S. cerevisiae mutant with low H^sub 2^S production phenotype (AWRI 1640), interspecies hybrid of S. cerevisiae and S. kudriavzevii (AWRI 1539) and a hybrid of AWRI 1640 and AWRI 1539 (AWRI 1810) were exposed separately to fungicides pyrimethanil (Pyr, 10 mg/L) and fenhexamid (Fhx, 10 mg/L), as well as to the most common toxin produced by moulds on grapes, ochratoxin A (OTA, 5 µg/L), during alcoholic fermentation of Vitis vinifera L. cv. Sauvignon blanc juice. Contaminants were found to strongly impair fermentation performance and metabolic activity of all yeast strains studied. The chemical profile of wine was analyzed by HPLC (volatile acidity, concentrations of ethanol, fructose, glucose, glycerol and organic acids) and the aromatic profile was analyzed using a stable isotope dilution technique using GC/MS (ethyl esters, acetates and aromatic alcohols) and Kitagawa tubes (H^sub 2^S). The chemical composition of wine with added contaminants was in all cases significantly different from the control. Of particular note is that the quantity of aromatic compounds produced by yeast was significantly lower. Yeast's capacity to remove contaminants from wine at the end of the alcoholic fermentation, and after extended contact (7 days) was determined. All the strains were able to remove contaminants from the media, moreover, after extended contact, the concentration of contaminants was in most cases lower.

Headnote

Summary

The interaction between four industrial wine yeast strains and grape juice chemical contaminants during alcoholic fermentation was studied. Industrial strains of Saccharomyces cerevisiae (AWRI 0838), S. cerevisiae mutant with low H^sub 2^S production phenotype (AWRI 1640), interspecies hybrid of S. cerevisiae and S. kudriavzevii (AWRI 1539) and a hybrid of AWRI 1640 and AWRI 1539 (AWRI 1810) were exposed separately to fungicides pyrimethanil (Pyr, 10 mg/L) and fenhexamid (Fhx, 10 mg/L), as well as to the most common toxin produced by moulds on grapes, ochratoxin A (OTA, 5 µg/L), during alcoholic fermentation of Vitis vinifera L. cv. Sauvignon blanc juice. Contaminants were found to strongly impair fermentation performance and metabolic activity of all yeast strains studied. The chemical profile of wine was analyzed by HPLC (volatile acidity, concentrations of ethanol, fructose, glucose, glycerol and organic acids) and the aromatic profile was analyzed using a stable isotope dilution technique using GC/MS (ethyl esters, acetates and aromatic alcohols) and Kitagawa tubes (H^sub 2^S). The chemical composition of wine with added contaminants was in all cases significantly different from the control. Of particular note is that the quantity of aromatic compounds produced by yeast was significantly lower. Yeast's capacity to remove contaminants from wine at the end of the alcoholic fermentation, and after extended contact (7 days) was determined. All the strains were able to remove contaminants from the media, moreover, after extended contact, the concentration of contaminants was in most cases lower.

Key words: wine fermentation, Saccharomyces spp., interspecies hybrids, pyrimethanil, fenhexamid, ochratoxin A, aromatic profile, fermentation kinetics, H^sub 2^S, Sauvignon blanc

Introduction

The chemical composition of grape juice is mainly a consequence of vine physiological processes (1). How- ever, grapes may undergo microbiological spoilage by moulds among which some Aspergillus and Penicillium species, especially A. carbonarius, are producers of ochra- toxin A (OTA) (2). Many studies have investigated its re- moval from grape juice, wine and other media, to re- duce its negative impact on human health (reviewed by Amézqueta et al. (3)). It was shown that yeasts are able to reduce OTA concentration by its adsorption on their cell wall during alcoholic fermentation; predominantly by mannoproteins, which are released from the yeast cell wall after the end of alcoholic fermentation (4-6). The ef- fect of yeasts on the concentration of OTA in the fermen- tation media was widely studied, which is not the case with the influence of OTA on yeast metabolism in fer- mentation media. In a previous study (4) we showed that OTA at higher concentrations (approx. 5 mg/L) impaired the yeast fermentative capacities and induced a higher volatile acidity during alcoholic fermentation in synth- etic media.

To prevent microbiological spoilage of grapes, many fungicides and other phytopharmaceuticals are used. However, in years when the climate conditions for Bo- trytis infection are favourable, such control measures may cause maximum permitted residue levels to be exceeded (7), even though the concentrations of fungicides are sig- nificantly reduced during processing of grapes into wine (8,9). It has been demonstrated that some fungicides are able to affect the ecology and kinetics of alcoholic fer- mentations (10,11). Additionally, it was found that the aromatic compound production of yeasts is negatively affected (12-14).

In recent years the trend has been to use selected yeasts for alcoholic fermentation, because this can guar- antee the smooth development of the process, avoid the production of off-flavours and generate positive aroma that improves the sensory properties of the wine (15). In order to improve wine aromatic composition, other al- ternative techniques of inoculation have been adopted, especially mixed/sequenced inocula of selected non-Sac- charomyces yeasts (16,17) and the use of interspecies hy- brids of Saccharomyces yeasts, i.e. hybrids of S. cerevisiae and S. kudriavzevii, which were found to have good aro- matic potential (18).

The aim of this study is to uncover the interaction of four genetically different industrial wine yeast strains: S. cerevisiae (AWRI 0838), S. cerevisiae mutant with low H2S production phenotype (AWRI 1640), an interspecies hybrid of S. cerevisiae and S. kudriavzevii (AWRI 1539) and the triple hybrid of the last two strains (AWRI 1810) with the mycotoxin OTA and fungicides pyrimethanil (Pyr) and fenhexamid (Fhx). We have tried to answer the question whether the abusive use of fungicides affects more nega- tively the final product than the presence of OTA during alcoholic fermentation by determining the removal po- tential of the strains with different genomic background and the potential of contaminants to affect yeast metabo- lism (fermentation kinetics and aromatic compound pro- duction).

Materials and Methods

Yeast strains

The yeast strains used in this study were all indus- trial wine yeasts: Saccharomyces cerevisiae AWRI 0838, the mutant Saccharomyces cerevisiae AWRI 1640 with low H2S phenotype (19), an interspecies hybrid of Saccharomyces cerevisiae and Saccharomyces kudriavzevii, AWRI 1539, and the hybrid of AWRI 1640 and AWRI 1539, AWRI 1810 (18). Three-day-old cultures grown on yeast extract, pep- tone, dextrose (YPD) agar plates (2 % D-glucose,1%yeast extract, 1 % peptone and 2 % agar) were inoculated into 8 mL of sterile YPD broth incubated for 24 h at 28 °C (in 12-mL sterile Falcon tubes), and later transferred into 16 mL of sterile chemically defined must (CDM) (20) and incubated for 24 h (in sterile 50-mL Falcon tubes). The concentrations of yeast cells for the inocula were count- ed by haemocytometer.

Fermentation assays

Alcoholic fermentations were carried out in sterile 250-mL fermentation flasks (Schott, Mainz, Germany) containing 200 mL of Sauvignon blanc 2005 (SB05) grape juice (128.5 g/L of reducible sugars, titratable acidity (pH=8.2) 5.1 g/L, pH=3.19, SO2 (free) 10 mg/L, SO2 (to- tal) 19 mg/L, yeast assimilable nitrogen (YAN) 235 mg/L) sterilized by filtering through 0.65- and 0.45-mm cartridge filters (Sartorius, Bohemia, NY, USA). The grape juice with low concentration of reducible sugars and high amount of YAN was chosen to make sure that H2S production will not be affected by high osmotic pressure and low amount of YAN (19). Each of the four strains was inocu- lated at the final concentration of 106 cells per mL (all in triplicate). Four types of media were prepared: (i) con- trol, which was composed of SB05 and 1 mL of 80 % ethanol; (ii) Pyr or (iii) Fhx, which were composed of SB05 and 10 mg of pyrimethanil or fenhexamid suspend- ed in 1 mL of 80 % ethanol, respectively, and (iv)OTA, composed of SB05 and 5.0 mg/L of OTA suspended in 1 mL of 80 % (by volume) Et-OH (4). All contaminants were obtained as analytical standards from Sigma-Aldrich (St. Louis, MO, USA). The selection of concentrations was done considering previous research on synthetic media (4,10). The fermentation flasks were equipped with pre- cision gas detector tubes (Kitagawa, Hiroshima, Japan) and a trap-based method for H2S quantification during fermentation was used (18).

The assays were performed at 17 °C, with rotary shaking at 150 rpm. The fermentation kinetics was fol- lowed by CO2 measuring the mass loss every 24 h. Fer- mentations were considered finished when CO2 release was lower than 0.1 g per 100 mL per day and the con- centration of reducible sugars was lower than 2 g/L (Clinitest®, Bayer, Leverkusen, Germany).

After the end of fermentation, samples of wine were taken for the determination of volatile and non-volatile chemical compounds as well as the concentration of contaminants. The samples were taken from homoge- nized media under aseptic conditions, centrifuged (for 5 min at 11 200×g), and the clean supernatant was frozen for analysis.

Extended contact between wine and yeast lees

In order to determine the potential of yeast lees to remove the contaminants, a 7-day extended contact time with daily mixing was performed (4,10). After the con- tact period, the samples were taken from the homoge- nized media under aseptic conditions, centrifuged (for 5 min at 11 200×g), and the clean supernatant was frozen for analysis.

Analysis of the principal chemical compounds in wine

Contents of glucose, fructose, ethanol, glycerol, and acetic, citric, malic and tartaric acids in fermented SB05 wines were analyzed. Their concentration in the media was analyzed by HPLC, using a Bio-Rad HPX-87H col- umn (Bio-Rad, Hercules, CA, USA) as described previ- ously (19).

Analysis of fermentation products

Samples of fermented SB05 wines were prepared as follows: from each treatment the same aliquots of the three replicates were taken and mixed together into one sample (from 48 fermentations, 16 final samples). Sam- ples were prepared in 2 dilutions (1/20 and 3/10) with model wine (13.8 % ethanol, 10 % potassium hydrogen tartrate, pH adjusted with tartaric acid to 3.5). Samples were prepared and the content of ethyl esters, acetates and aromatic alcohols was analyzed by Metabolomics Australia (AWRI, Adelaide, Australia) as detailed by Bizaj et al. (18).

Determination of fungicide residues

Extraction and determination of pyrimethanil and fenhexamid residues in fermented SB05 wines was done using a gas chromatography-mass spectrometry system (GC-MS) and liquid chromatography-tandem mass spec- trometry system (LC/MS/MS), respectively, according to the previously described methods (10,21).

Determination of ochratoxin A residues

The concentration of OTA in the fermented SB05 wines and wine samples of that were collected after the extended contact phase with yeast lees were analyzed by means of immunoaffinity column clean-up and HPLC, as described previously (22).

Statistical analysis

Statistical analysis was done using general linear model (SAS software 8.01, SAS Institute Inc., Cary, NC, USA) using the equation:

yijk=m+Ti+Sj+T·Sij /1/

where yijk is the controled value; µ is the average value; Ti is the effect of i-treatment, i=1-8; Sj is the effect of j-strain, j=1-4; and T·Sij is the effect of interaction between i-treatment and j-strain.

The data of fermentation kinetics were statistically analysed by intervals of standard deviation (Microsoft Office Excel 2003, Durham, NC, USA). The statistical level of significance was set at p£0.05.

Results and Discussion

Fermentation kinetics

The fermentation kinetics at 17 °C in the SB05 grape juices varied with different yeast strains; the fastest fer- mentation kinetics in control treatments was performed by AWRI 1539 in 7 days; followed by AWRI 0838 and 1810, which finished the fermentation in 9 days; while AWRI 1640 displayed the slowest fermentation rate, com- pleting it in 12 days (Fig. 1). Differences between the strains were expected in control media (10,23). More in- terestingly, all strains responded with significantly slow- er fermentation rates when challenged with contaminants OTA, Pyr and Fhx. However, in Fhx fermentations, the stationary phase was reached earlier, showing that Fhx was less inhibitory if compared to OTA and Pyr. This was also shown by Cabras et al. (24), who found no neg- ative effect of Fhx on wine yeast fermentation kinetics, even at concentrations greater than that permitted by legislation (3 mg/L). Recently, Bizaj et al. (10) demon- strated the negative effect of Fhx on the fermentation ki- netics at higher concentrations, i.e. 10 mg/L, during wine yeast fermentation in synthetic media.

The toxicity of Pyr on alcoholic fermentation was al- ready shown to be greater than the toxicity of Fhx in synthetic media (10), but it has now been confirmed for grape juice (Fig. 1). The effect of Pyr was most promi- nent when using AWRI 1539. Its fermentation rate was initially the fastest, but after the 3rd day of fermentation, a drastic decrease in the fermentation rate can be seen. Similar trend was also observed by Bizaj et al. (10), where strains exhibiting faster fermentation rates were more affected by the toxicity of contaminants. Moreover, this trend can be strengthened by less negative effect of Pyr on the more slowly fermenting strains, i.e. AWRI 1810, 0838 and 1640.

As shown in Fig. 1, OTA had negative effect on fer- mentation kinetics of all wine yeast strains in grape juice, as determined before for synthetic media (4). More- over, its toxicity seemed to be similar to that of Pyr.

In the assays with OTA and Pyr, all the fermenta- tions tended to be slower, producing less CO2 in com- parison with the control and Fhx fermentations, suggest- ing their more intense negative effect on all wine yeast strains studied. The only case when fermentation re- mained stuck was the assay with Pyr and the strain AWRI 1539.

Off-flavour production during alcoholic fermentation

All AWRI 1640 fermentations performed in this study were free from detectable H2S, as expected (19). The three other strains had different capacities for H2Spro- duction: AWRI 1539 was known to be a relatively high H2S producer (18), AWRI 0838, a yeast strain commonly used in commercial fermentations and AWRI 1810, the hybrid between AWRI 1539 and AWRI 1640, were known to be intermediate level producers (18). As shown in Ta- ble 1 all the strains in control assay produced H2S. How- ever, the contaminants seem to have different effects on the H2S production during fermentation. The strain AWRI 0838 produced the highest amount of H2S in control fer- mentations, suggesting that all contaminants impaired H2S production. On the other hand, when using AWRI 1810 the addition of Fhx and Pyr stimulated H2Spro- duction, while in the case of OTA and control assays, the amount of produced H2S was significantly lower. The high- est H2S producer, AWRI 1539 in the assays with OTA and Pyr produced lower amounts of H2S than in assays with Fhx or in the control one. In the more slowly fer- menting assays and especially in the case of the stuck fermentation when using the combination of Pyr and strain AWRI 1539, the quantities of H2S produced were significantly lower. These results suggest that there were interactions between the contaminants and the yeast strains with regard to H2S formation. Fenhexamid, pirymethanil and ochratoxin A had previously been found to have ef- fects on metabolic pathways during alcoholic fermenta- tions (4,10,25). Moreover, we have demonstrated here that all of them have the potential to affect H2S production pathway during alcoholic fermentation of grape juice.

Principal chemical compounds in wine

The principal chemical compounds analyzed at the end of fermentation were citric, tartaric, malic, succinic, acetic and lactic acids, glycerol, ethanol, glucose and fruc- tose (Table 2). All the strains in all four types of assays were able to consume all the sugars, since no residual sugar was found in the media after the end of alcoholic fermentation.

Except for the strain AWRI 1640, all strains degraded more malic acid if a contaminant was present in the me- dia. The highest concentrations of malic acid were found in all assays when AWRI 1640 was used, suggesting that, since this strain was produced by random chemical mu- tagenesis (19), malic degradation enzyme might have been affected in the way of reducing the malate degradation by malo-ethanolic fermentation (26). Interestingly, higher concentration of malic acid was characteristic also for a triple hybrid AWRI 1810, which showed intermediate phe- notype, suggesting that it has inherited part of AWRI 1640 genome background. Similar trend was observed also for citric acid in terms of contaminant effect; on the other hand, there were not any particular trends present be- tween different strains. Succinic acid is considered to be particularly important for sensory wine quality (1). The production of succinic acid was significantly lower in all assays where the contaminants were present, meaning that their presence in the media negatively affected its production. Strain dependency was evident again; the in- herent ability to degrade malic acid as well as to pro- duce succinic acid was present. The highest production by AWRI 0838 and 1640 was observed, the lowest by AWRI 1539 and intermediate by AWRI 1810. Strain de- pendency was also observed to be more important in the case of lactic and acetic acid production. AWRI 0838 was the highest producer of lactic acid, followed by AWRI 1810, 1539 and 1640 in control assays; on the other hand, a particular trend of the effect of contaminant on its pro- duction was not present, except when AWRI 1640 was used, in which case contaminants negatively affected lac- tic acid production. AWRI 1539 is known for its high po- tential for acetic acid production (18), which can also be seen in this study; but none of the contaminants were able to affect acetic acid production. Similarly to lactic acid, AWRI 1640 was strongly affected by the presence of contaminants, suggesting that metabolic pathways af- ter the synthesis of pyruvate are particularly susceptible to the contaminants in this strain.

Glycerol production during alcoholic fermentation was affected when the contaminant was present in the medium. The ability of AWRI 1640 and 1539 strains to produce glycerol was clearly impaired by all three con- taminants, especially the latter in the assay with Pyr when the fermentation remained stuck. On the other hand, AWRI 0838 and 1810 strains were overall less sen- sitive; moreover, Pyr seemed to stimulate their glycerol production. The production of ethanol seemed to be the most negatively affected. Even though all the reducible sugars are consumed by yeasts during fermentation, it seems that due to the effect of contamination on meta- bolic pathway and the increasing concentration of etha- nol in media, yeasts were not able to convert sugars to ethanol. Yeasts in assays with contaminants produced roughly 14 % less ethanol. Again, Pyr was found to be the most toxic contaminant during fermentation. Its ef- fect was the most negative on AWRI 1539, where a 35.8 % decrease in ethanol production was observed. How- ever, this was expected from the fermentation kinetics (Fig. 1c). Similar data were found for AWRI 1640, where pyrimethanil was found to be the most toxic. On the other hand, we could not observe any significant differ- ences between assays with contaminants when using AWRI 0838 or 1640 strain. The primary fermentation prod- uct concentrations were significantly lower in the case of slower fermentation kinetics and especially in the case when the fermentation remained stuck.

Volatile fermentation products in wine

In Table 3 the production of aromatic compounds by the yeast strains in fermentative assays with and with- out contaminants is presented.

Ethyl acetate was the highest produced ethyl ester by all strains. This compound imparts pleasant smell to wine when present in concentrations lower than 80 mg/L (1), which is the case in all our assays. Its production was impaired in all assays using contaminants, especial- ly when pyrimethanil was present (AWRI 1640 and 1539), when the fermentation kinetics was slow or remained stuck for AWRI 1539. Strain-related sensitivity can be seen as well. Interestingly, in the case of S. cerevisiae AWRI 0838, ethyl acetate production was negatively af- fected by all contaminants to a similar degree, suggest- ing high sensitivity of the strain to the three chemical compounds. This is in contrast to the studies of García et al. (13), where the non-hybrid S. cerevisiae strain was found to be the most resistant to pesticides, including pyrimethanil.

Ethyl hexanoate is well known for its important and positive effect on the aroma, especially of young wines (27). In all our assays, the presence of contaminants in the media negatively affected its production, which is in accordance with the results obtained by García et al. (13). Moreover, the final concentration in the media was shown to be strain dependent.

Ethyl propanoate and ethyl butanoate were all pro- duced below the perception threshold (28). Ethyl buta- noate production was in all assays negatively affected by the contaminants, and showed a strong strain depen- dency. Interestingly, strain dependency was also present in the production of ethyl propanoate, but except for AWRI 1810, in all other assays the contaminants stimu- lated its production.

Ethyl 2-methylpropanoate, ethyl 2-methylbutanoate and ethyl 3-methylbutanoate are derivates of acids con- sidered as indicators of lower quality of wine (29). Strains AWRI 0838 and 1810 were the highest producers of ethyl 2-methylpropanoate. Interestingly, contaminants posi- tively affected its production, especially Pyr (AWRI 1640 and 0838) and OTA (AWRI 1539 and 1810), which may have the effect of lowering the quality of wines. Ethyl 2-methylbutanoate was found in traces and ethyl 3-meth- ylbutanoate could not be detected at all in all assays.

2-Phenyl ethyl acetate, which confers the flavour of roses and violets to young wines, was found to be be- low the perception thresholds in all our assays. Strains AWRI 1539 and 1810 were high producers of 2-phenyl ethyl acetate. When contaminants were added, the pro- duction was significantly lower in most of the cases, es- pecially when Pyr was added in the assays with AWRI 0838, where 2-phenyl ethyl acetate was not detected at all. Interestingly, in the assays with AWRI 1640, no 2-phe- nyl ethyl acetate was detected. This data suggest that during its production by mutagenesis (19), a mutation occurred in genes involving enzymes for 2-phenyl ethyl acetate production. However, this negative trait was not conferred to AWRI 1810 (18).

The concentration of hexyl acetate, giving the fla- vour of cherries or pears, in all assays was below the perception threshold. Strain-dependent production can be observed; the order from the highest to the lowest producer was as follows: AWRI 1810>1539>0838>1640. In all assays where contaminants were added, they had a negative effect on the hexyl acetate production; the only exception was OTA used with strain AWRI 1539.

3-Methyl butyl acetate, 2-methyl butyl acetate and 2-methyl propyl acetate are very important in determin- ing young wine flavour, conferring banana or fruity fla- vours in white wines. In all assays where contaminants were present their negative effect can be observed, espe- cially that of pyrimethanil. This is opposite to what was observed by García et al. (13), where its production was stimulated. However, a different yeast strain was used in their study.

Higher alcohols are compounds that are produced by yeasts during alcoholic fermentation, and their con- centrations below 300 mg/L positively affect wine fla- vour. Two isoamyl alcohols were analyzed, 2-methyl bu- tanol and 3-methyl butanol, which are considered to be among the major volatiles that confer intensity of fruity flavour to wine (30). Their perception threshold was ex- ceeded in all assays and their concentration was highly dependent on yeast strain, with interspecies hybrids be- ing higher producers. The presence of contaminants in the media had a significant effect on the concentration of produced amyl alcohols; the production of 3-methyl bu- tanol was negatively affected in all cases, on the other hand the production of 2-methyl butanol in assays with yeasts AWRI 0838 and 1810 was positively affected by Pyr and Fhx, respectively. Productions of 2-methyl pro- panol, butanol and hexanol were negatively affected by contaminants in all our assays.

The production of secondary metabolites was signif- icantly affected by the presence of contaminants in the media; however, no direct link to the fermentation kinet- ics was found.

Removal of contaminants during alcoholic fermentation and after extended contact with wine yeasts

Yeast strains were found to remove contaminants from grape juice as well as synthetic media in fermentative and stationary assays (4,5,10,31). In previous studies, it was demonstrated that Pyr, Fhx and OTA can be re- moved only by the adsorption of the cell wall manno- proteins (4,6,10,32) and not by degradation in synthetic media and in grape juice, in contrast to some other pes- ticides (33).

The capacity of four yeast strains to remove Pyr, Fhx and OTA from Sauvignon blanc must is presented in Table 4. The adsorption potential was evaluated after alcoholic fermentation, and after the extended contact be- tween yeasts and media containing contaminants. Nunez et al. (6) and Bizaj et al. (10) demonstrated that the main release of mannoproteins from the yeast cell wall occurs within seven days after the end of fermentation, when they also adsorb a fraction of contaminants. Our results confirmed this, as the amount of removed OTA was sig- nificantly higher after the extended contact in all assays, except for strain AWRI 1640. The fractions of adsorbed contaminants were higher in natural wine than in syn- thetic media (4,10). Additionally, the work of Bejaoui et al. (32) showed that contaminants such as OTA can be released back into the synthetic media after being ad- sorbed onto yeast components. This did not happen in our work, where real grape must was used. In this way we confirmed the importance of environmental condi- tions for the adsorption capacity of yeast cell wall, espe- cially pH, which determines the charge of functional groups on mannoproteins and binding contaminants (34). The importance of the final amount of produced biomass in the media is crucial for the removal potential (32), and this was found to be dependent on the fermentation ki- netics.

From the results shown in Table 4, it is also evident that the removal potential of yeast strains in fermenta- tive assays is strain and species dependent, as probably different genetic background of the four strains and the induced mutations in AWRI 1640 define yeast cells mor- phologically, chemically and metabolically.

Conclusion

In this study we highlighted the complexity of inter- actions of genetically different industrial wine yeasts and their contaminants, originating on the one hand from natural spoilage mycobiota on grape berries (ochratoxin A), and on the other from fungicides (pyrimethanil and fenhexamid), working antagonistically against spoilage mycobiota. It was demonstrated for the first time that OTA, Fhx and Pyr negatively affect fermentation kinet- ics of industrial yeast cells in natural grape juice. How- ever, their intensiveness was dependent on the genetic background of the yeast strain. In all assays the contami- nants affected metabolic pathways that dictate the aro- matic and basic composition of wines. Moreover, meta- bolic pathways were found to be affected differently by the same contaminant. This suggests that these interac- tions define the composition of the final product.

Furthermore, the final composition of wines was af- fected by the ability of yeasts to remove contaminants. A significant part was removed already during alcoholic fermentation, and not only after the extended contact of yeast lees with wine.

None of the three contaminants was found to increase the concentration of any compounds known to confer desirable sensory characteristics, but on the other hand, they were found to increase the concentration of unde- sirable compounds.

Acknowledgements

The authors would like to thank Dr. Franc ^u{ from Agricultural Institute of Slovenia for the determination of pesticide residues. This research was supported by the Ministry of Higher Education, Science and Technol- ogy of Slovenia (project no. J4-0838), Vinska Klet Gori- {ka Brda z.o.o., Dobrovo, Slovenia, and the Australian Wine Research Institute, a member of the Wine Innova- tion Cluster in Adelaide, supported by Australian grape growers and winemakers through their investment body, the Grape and Wine Research Development Corpora- tion, with matching funds from the Australian Govern- ment.

References

References

1. P. Ribereau-Gayon, D. Dubordieu, B. Doneche, A. Lonvaud: Handbook of Enology: The Microbiology of Wine and Vinifica- tions, Vol. 1, John Wiley & Sons Ltd., London, UK (2006).

2. N. Belli, S. Marin, I. Coronas, V. Sanchis, A.J. Ramos, Skin damage, high temperature and relative humidity as detri- mental factors for Aspergillus carbonarius infection and ochratoxin A production in grapes, Food Control, 18 (2007) 1343-1349.

3. S. Amézqueta, E. González-Pen~ as, M. Murillo-Arbizu, A. López de Cerain, Ochratoxin A decontamination: A review, Food Control, 20 (2009) 326-333.

4. E. Bizaj, J. Mavri, F. ^u{, A. Raspor, Removal of ochratoxin Ain Saccharomyces cerevisiae liquid cultures, S. Afr. J. Enol. Vitic. 30 (2009) 151-155.

5. F. Cecchini, M. Morassut, E.G. Moruno, R. Di Stefano, In- fluence of yeast strain on ochratoxin A content during fer- mentation of white and red must, Food Microbiol. 23 (2006) 411-417.

6. Y.P. Nunez, E. Pueyo, A.V. Carrascosa, A.J. Martinez-Ro- driguez, Effects of aging and heat treatment on whole yeast cells and yeast cell walls and on adsorption of ochratoxin A in a wine model system, J. Food Prot. 71 (2008) 1496- 1499.

7. P. Cabras, A. Angioni, Pesticide residues in grapes, wine, and their processing products, J. Agric. Food Chem. 48 (2000) 967-973.

8. F. Cus, H. Ba{a-^esnik, S.V. Bolta, A. Gregor~i~, Pesticide residues in grapes and during vinification process, Food Con- trol, 21 (2010) 1512-1518.

9. F.K. Miller, U. Kiigemagi, P.A. Thomson, D.A. Heatherbell, M.L. Deinzer, Methiocarb residues in grapes and wine and their fate during vinification, J. Agric. Food Chem. 33 (1985) 538-545.

10. E. Bizaj, F. ^u{, P. Raspor, Removal of pyrimethanil and fenhexamid from Saccharomyces cerevisiae liquid cultures, Food Technol. Biotechnol. 49 (2011) 474-480.

11. F. Cus, P. Raspor, The effect of pyrimethanil on the growth of wine yeasts, Lett. Appl. Microbiol. 47 (2008) 54-59.

12. J. Oliva, S. Navarro, A. Barba, G. Navarro, M.R. Salinas, Effect of pesticide residues on the aromatic composition of red wines, J. Agric. Food Chem. 47 (1999) 2830-2836.

13. M.A. García, J. Oliva, A. Barba, M.A. Cámara, F. Pardo, E.M. Díaz-Plaza, Effect of fungicide residues on the aro- matic composition of white wine inoculated with three Sac- charomyces cerevisiae strains, J. Agric. Food Chem. 52 (2004) 1241-1247.

14. R.M. Gonzalez-Rodriguez, R. Noguerol-Pato, C. Gonzalez- -Barreiro, B. Cancho-Grande, J. Simal-Gandara, Application of new fungicides under good agricultural practices and their effects on the volatile profile of white wines, Food Res. Int. 44 (2011) 397-403.

15. I.S. Pretorius, F.F. Bauer, Meeting the consumer challenge through genetically customized wine-yeast strains, Trends Biotechnol. 20 (2002) 426-432.

16. E.S. King, R.L. Kievit, C. Curtin, J.H. Swiegers, I.S. Preto- rius, S.E.P. Bastian et al., The effect of multiple yeasts co- -inoculations on Sauvignon Blanc wine aroma composition, sensory properties and consumer preference, Food Chem. 122 (2010) 618-626.

17. N.P. Jolly, O.P.H. Augustyn, I.S. Pretorius, The role and use of non-Saccharomyces yeasts in wine production, S. Afr. J. Enol. Vitic. 27 (2006) 15-39.

18. E. Bizaj, A.G. Cordente, J.R. Bellon, P. Raspor, C.D. Curtin, I.S. Pretorius, A breeding strategy to harness flavor diver- sity of Saccharomyces interspecific hybrids and minimize hy- drogen sulfide production, FEMS Yeast Res. 12 (2012) 456- 465.

19. A.G. Cordente, A. Heinrich, I.S. Pretorius, J.H. Swiegers, Isolation of sulfite reductase variants of a commercial wine yeast with significantly reduced hydrogen sulfide produc- tion, FEMS Yeast Res. 9 (2009) 446-459.

20. P.A. Henschke, V. Jiranek: Metabolism of Nitrogen Com- pounds. In: Wine Microbiology and Biotechnology, G.H. Fleet (Ed.), Harwood Academic Publishers, Camberwell, UK (1992) pp. 77-165.

21. H. Ba{a-^esnik, A. Gregor~i~, F. ^u{, Pesticide residues in grapes from vineyards included in integrated pest man- agement in Slovenia, Food Addit. Contam. A, 25 (2008) 438- 443.

22. A. Visconti, M. Pascale, G. Centonze, Determination of ochratoxin A in wine by means of immunoaffinity column clean-up and high-performance liquid chromatography, J. Chromatogr. A 864 (1999) 89-101.

23. F.N. Arroyo-Lopez, S. Orlic, A. Querol, E. Barrio, Effects of temperature, pH and sugar concentration on the growth parameters of Saccharomyces cerevisiae, S. kudriavzevii and their interspecific hybrid, Int. J. Food Microbiol. 131 (2009) 120-127.

24. P. Cabras, A. Angioni, V.L. Garau, F.M. Pirisi, F. Cabitza, M. Pala et al. , Fenhexamid residues in grapes and wine, Food Addit. Contam. 18 (2001) 625-629.

25. P. Cabras, V.L. Garau, A. Angioni, G.A. Farris, M. Budroni, L. Spanedda, Interactions during fermentation between pesticides and enological yeasts producing H2S and SO2, Appl. Microbiol. Biotechnol. 43 (1995) 370-373.

26. F. Radler: Yeast Metabolism of Organic Acids. In: Wine Mi- crobiology and Biotechnology, G.H. Fleet (Ed.), Harwood Ac- ademic Publishers, Camberwell, UK (1992) pp. 165-182.

27. J.L. Robichaud, A.C. Noble, Astringency and bitterness of selected phenolics in wine, J. Sci. Food Agr. 53 (1990) 343- 353.

28. P. Dubois, Wine aromas and their defects, Rev. Fr. Oenol. 145 (1994) 27-40 (in French).

29. C.G. Edwards, R.B. Beelman, C.E. Bartley, A.L. McConnell, Production of decanoic acid and other volatile compounds and the growth of yeast and malolactic bacteria during vini- fication, Am. J. Enol. Vitic. 41 (1990) 48-56.

30. P.X. Etiévant: Wine. In: Volatile Compounds in Foods and Beverages, H. Maarse (Ed.), Marcel Dekker, New York, NY, USA (1991) pp. 241-252.

31. P. Cabras, A. Angioni, V.L. Garau, M. Melis, F.M. Pirisi, E.V. Minelli et al., Fate of some new fungicides (cyprodinil, fludioxonil, pyrimethanil, and tebuconazole) from vine to wine, J. Agric. Food Chem. 45 (1997) 2708-2710.

32. H. Bejaoui, F. Mathieu, P. Taillandier, A. Lebrihi, Ochra- toxin A removal in synthetic and natural grape juices by selected oenological Saccharomyces strains, J. Appl. Micro- biol. 97 (2004) 1038-1044.

33. A. Angioni, V.L. Garau, A.A. Del Real, M. Melis, E.V. Mi- nelli, C. Tuberoso et al., GC-ITMS determination and deg- radation of captan during winemaking, J. Agric. Food Chem. 51 (2003) 6761-6766.

34. D. Gonzalez-Ramos, E. Cebollero, R. Gonzalez, A recombi- nant Saccharomyces cerevisiae strain overproducing manno- proteins stabilizes wine against protein haze, Appl. Envi- ron. Microbiol. 74 (2008) 5533-5540.