Free access
Research Article
22 February 2012

Candida zemplinina Can Reduce Acetic Acid Produced by Saccharomyces cerevisiae in Sweet Wine Fermentations


In this study we investigated the possibility of using Candida zemplinina, as a partner of Saccharomyces cerevisiae, in mixed fermentations of must with a high sugar content, in order to reduce its acetic acid production. Thirty-five C. zemplinina strains, which were isolated from different geographic regions, were molecularly characterized, and their fermentation performances were determined. Five genetically different strains were selected for mixed fermentations with S. cerevisiae. Two types of inoculation were carried out: coinoculation and sequential inoculation. A balance between the two species was generally observed for the first 6 days, after which the levels of C. zemplinina started to decrease. Relevant differences were observed concerning the consumption of sugars, the ethanol and glycerol content, and acetic acid production, depending on which strain was used and which type of inoculation was performed. Sequential inoculation led to the reduction of about half of the acetic acid content compared to the pure S. cerevisiae fermentation, but the ethanol and glycerol amounts were also low. A coinoculation with selected combinations of S. cerevisiae and C. zemplinina resulted in a decrease of ∼0.3 g of acetic acid/liter, while maintaining high ethanol and glycerol levels. This study demonstrates that mixed S. cerevisiae and C. zemplinina fermentation could be applied in sweet wine fermentation to reduce the production of acetic acid, connected to the S. cerevisiae osmotic stress response.


Candida zemplinina is a psycrotolerant and osmotolerant yeast, properties that can be advantageously exploited in sweet wine production, which is characterized by a high sugar concentration and low fermentation temperatures (26). Sipiczki recognized it as a distinct new species and named it C. zemplinina in 2003 (25). However, already back in 2002, in the Napa valley, California, Mills et al. isolated a strain of Candida sp., named Candida species strain EJ1, from Botrytis cinerea-infected grapes, with interesting features, such as the ability to deplete fructose from a Chardonnay juice without affecting the glucose concentration (19).
Yeast ecology studies carried out over the last 5 years have highlighted the frequent presence of this species in wine fermentations (3, 16, 17, 20, 28, 29, 31, 33). Moreover, it has been demonstrated that strains of C. stellata, deposited in several culture collections and isolated from grapes, belong to the C. zemplinina species (10). Altogether, all this evidence points out the need for a better understanding of the role of this yeast during wine transformation.
Sweet wines, such as “Passito wines,” “Icewines,” “Sauternes,” and others, are produced from must obtained from dried grapes and are characterized by a very high sugar concentration. The Saccharomyces cerevisiae strains used for these fermentations should be able to promptly respond to osmotic stress and be able to increase their load immediately after inoculation. It has been shown that the response of S. cerevisiae to osmotic stress can result in increased acetic acid contents due to the upregulation of genes encoding for aldehyde dehydrogenases. In fact, yeast grown in 40% (wt/vol) sugar juice produced 1.35 g of acetic acid/liter compared to 0.3 g/liter at the lower sugar concentration (22% [wt/vol]) (12). The acetic taste in these wines is in part masked by the high residual sugars after fermentation. However, winemakers would like to reduce its content, which generally penalizes the final sensory quality of wines, becoming a limit to its commercialization, also in view of international legal limits for the acetic acid (11, 14).
Several studies have demonstrated that non-Saccharomyces yeasts are able to survive during alcoholic fermentations (9, 19). Such evidence opens the way toward new applications of non-Saccharomyces in wine fermentation, which would make the organoleptic profiles of the wines more complex due to enzymatic activities that such yeasts possess (27).
Mixed Saccharomyces and non-Saccharomyces fermentations have been tested since the 1990s (32), but it is only in recent years that the interest of researchers has been growing, as evidenced by recent reviews published on this subject (5, 13). Hanseniaspora uvarum, Torulaspora delbrueckii, Lachancea (Kluyveromyces) thermotolerans, and C. stellata have been the main species investigated thus far in mixed fermentations with the aim of adding complexity to the wine (4, 6, 9, 15, 32). The mixed fermentation strategy has also been used in high sugar musts in order to reduce the acetic acid content of the final wine. For this purpose, strains of T. delbrueckii, which have been described as low acetic acid producers (23), have been combined with S. cerevisiae, and a 53% reduction in volatile acidity has been obtained (2).
In this context, one possibility that could be exploited is to combine S. cerevisiae with C. zemplinina during fermentation. Since the latter yeast is osmotolerant and fructophylic and generally produces low amounts of acetic acid, together with relevant quantities of glycerol from sugar fermentation (18, 29), it might be able to consume sugars at the very beginning of fermentation, in this way alleviating the S. cerevisiae osmotic stress, thereby reducing production of acetic acid.
In the present study, we performed mixed fermentations, combining C. zemplinina obtained from grapes and wines of different origin with three strains of S. cerevisiae (one commercial and two wine isolates), in order to evaluate the potential effect on the reduction of acetic acid.


Yeast strains.

Thirty-five C. zemplinina isolates, mainly from Italy, but also including one isolate from Greece and one from the United States, were used in the present study (Table 1). Three strains of S. cerevisiae were also selected for mixed fermentations. Two of these, coded FB40 and ELCF3WC, came from the spontaneous fermentation of sweet wines in northern Italy, from Picolit and Erbaluce, respectively. The third S. cerevisiae strain, Lalvin EC1118 (Lallemand, Montreal, Quebec, Canada), a commercial active dried yeast (ADY), was selected because it is widely used by local wineries for the production of sweet wines.
Table 1
Table 1 Source of isolation of the C. zemplinina strains used in this study
Geographical region (country)WinerySourceStrain(s)
Abruzzo (Italy)GGrape juiceL37
  Cooked mustL191, L35, L491, L23, L34, L364, L344, L36, L365, L477
Friuli Venezia Giulia (Italy)BPicolit grapesBC16, BC20
 Picolit grape juiceBC55, BC60
  Picolit fermentation (3 days)BC115, BC116
  Picolit fermentation (14 days)BC224, BC226
 FPicolit grapesFC50, FC54
 RRamandolo grapesR1, R5
Trentino Alto Adige (Italy)ToNosiola dried grapesTOHA07
TNosiola dried grapesTOBLINO02
 PdNosiola dried grapesPEDRO10
 PiNosiola grapesPISO02, SANTA01
Veneto (Italy)MAmarone grape juiceC2CY2, C1AY2, T1Y3
  Amarone fermentation (7 days)C2AY9, C2BY10
California (USA)NDaBotritized grapesEJ1
Attika (Greece)NDGrape juiceD1
ND, not defined.
All of the isolates were identified through molecular methods, and they are deposited in the culture collections of the Universities of Turin, Teramo, and Verona (Italy). They were routinely cultivated in YPD broth (glucose [20 g/liter], peptone [20 g/liter], and yeast extract [10 g/liter], all from Oxoid, Milan, Italy) for 36 to 48 h at 30°C.

Molecular characterization of the C. zemplinina strains.

DNA was extracted from the C. zemplinina strains by using a mechanical treatment in a bead-beater machine (FastPrep-24; MP Biomedical, Solon, OH) as described elsewhere (7). The DNA was subsequently quantified with a NanoDrop instrument (Celbio, Milan, Italy) and standardized at 100 ng/μl. Molecular characterization of the isolates was carried out using RAPD [random(ly) amplified polymorphic DNA] and SAU-PCR methods, as described by Cocolin et al. (8). All of the C. zemplinina strains were subjected to both methods at least twice.

C. zemplinina fermentations.

Laboratory fermentations were carried out in a total volume of 100 ml of an Erbaluce dried grape must, containing 210 g of glucose, 193 g of fructose, and 3.5 g of glycerol/liter and 1.8% ethanol. The pH and titratable acidity, expressed as g of tartaric acid/liter, were 3.23 and 8.12, respectively. Before inoculation, the must was pasteurized at 90°C for 90 min, and the absence of live yeast populations was assessed by plating 100 μl on YPD agar, incubated at 30°C for 48 h. Despite the long time necessary for the pasteurization, no caramelization of the sugars occurred. The C. zemplinina isolates were preadapted in the same sweet must for 48 h and then inoculated to reach a final concentration of 105 to 106 cells/ml, which was determined through a microscopic count. Fermentations were carried out at 25°C for 14 days. Samples were collected in triplicate at 0, 1, 2, 6, 8, 12, and 14 days of fermentation and microbiological counts and high-pressure liquid chromatography (HPLC) analysis were carried out. The CFU/ml were determined by plating the serially diluted samples on the differential WLN medium (21) (Oxoid) and by incubating them for 3 to 5 days at 30°C. S. cerevisiae forms convex, creamy-white colonies in this medium, whereas C. zemplinina forms flat, light to intense green ones, this difference enabling their concurrent enumeration on the same medium (31). The glucose and fructose consumption, as well as the glycerol, ethanol, and acetic acid production, were quantified by means of HPLC (Thermo Electron Corp., Waltham, MA) equipped with a UV detector (UV100), set to 210 nm, and a refractive index detector (RI-150). The analyses were performed isocratically at 0.8 ml/min and 65°C with a cation-exchange column (300 by 7.8 mm [inner diameter]; Aminex HPX-87H) and a Cation H+ Microguard cartridge (Bio-Rad Laboratories, Hercules, CA), using 0.0026 N H2SO4 as the mobile phase (14, 24).
The three S. cerevisiae strains were also tested for their fermentation performance in the dried grape must, as described for the C. zemplinina strains. All of the fermentation trials were carried out twice.

Mixed fermentations.

Three S. cerevisiae strains and five strains of C. zemplinina were selected and used for mixed fermentations in a total volume of 100 ml of the same must described above. Two different approaches were used: (i) inoculation of both yeasts at the same time (coinoculation) and (ii) the addition of S. cerevisiae 48 h after C. zemplinina inoculation (sequential inoculation). All of the strains were preadapted in the same sweet must and always added to a final concentration of 105 to 106 cells/ml. The fermentations lasted 14 and 16 days for the coinoculation and sequential inoculation, respectively. Both the plate counts and the HPLC analysis of these experiments were performed as described above. Samplings were carried out in triplicate at 0, 1, 6, 8, 12, and 14 days of fermentation. In the case of sequential inoculation, samples were also collected 48 and 24 h prior to inoculation of S. cerevisiae (−2 and −1 days of fermentation). All of the fermentation trials were carried out twice.

Statistical analysis.

Gels containing the RAPD and SAU-PCR profiles of the C. zemplinina isolates were normalized, using the 1-kb molecular ladder (Sigma, Milan, Italy), loaded into each gel, and subjected to cluster analysis using BioNumerics software (Applied Maths, Kortrijk, Belgium). The Pearson product moment correlation coefficient was used to calculate the similarities in profile patterns, and dendrograms were obtained by means of the unweighted pair group method using arithmetic averages (UPGMA) clustering algorithm.
Statistical analyses of the chemical composition of the wines were performed using the SPSS statistical software package (version 17.0; SPSS, Inc., Chicago, IL). The Tukey test for P < 0.05 was used in order to establish any statistical differences by one-way analysis of variance (ANOVA). A multifactorial ANOVA test was used to explore the effect of the three tested factors (i.e., the strain of S. cerevisiae, the strain of C. zemplinina, and the type of inoculum) and to verify the existence of any interaction between them.


Molecular characterization of the C. zemplinina strains.

The dendrogram, which combines the results of the RAPD and SAU-PCR molecular characterization of the C. zemplinina strains included in the present study, is shown in Fig. 1. As can be observed, 26 of 35 strains clustered in two groups (1 and 2), while the remaining 9 strains formed three minor clusters, each including 2 to 4 strains, and one single-strain cluster. If the composition of the clusters is analyzed, it can be observed that no correlation could be found with the source of isolation; strains from different Italian regions, from Greece, and from the United States in fact clustered together. Only cluster 4 was composed of four strains from the same geographic region in the northeast of Italy, and the majority of the strains from winery G, in the Abruzzo region (Central Italy), were grouped in cluster 2.
Fig 1
Fig 1 Dendrogram of similarity constructed taking into consideration the RAPD and SAU-PCR profiles of the C. zemplinina strains used in the present study. Clusters are indicated by Latin numerals.

Dynamics and analytical results of the fermentations inoculated with pure cultures of C. zemplinina and S. cerevisiae.

The growth kinetics of the C. zemplinina strains, when inoculated in pure culture in must obtained from dried grapes, are summarized in Fig. 2, where the counts of all of the tested strains are taken into consideration, and the results are shown as the minimum, the maximum, and the median growth. As can be seen, C. zemplinina was able to grow, reaching a count of ∼107 CFU/ml in the first 2 days of fermentation, and then started to decrease from day 6 onwards. At the end of the monitored period, the strains still presented good vitality, with counts ranging from 105 to 106 CFU/ml.
Fig 2
Fig 2 Growth dynamics of C. zemplinina during fermentation of the must obtained from dried grapes. The results of the 35 strains are reported as minimums and maximums (indicated by the lines) and median values.
The results obtained from the HPLC analyses of the wines, after 14 days of fermentation, using either strains of C. zemplinina or S. cerevisiae in pure culture, are shown in Table 2. The C. zemplinina strains were characterized by a higher consumption of fructose than glucose, and some strains showed a total fructophylic character, having left the content of glucose originally found in the must before inoculation untouched (Table 2). The C. zemplinina strains generally produced relevant quantities of glycerol and low amounts of acetic acid and ethanol, although it should be underscored that some isolates were able to produce up to an 8.0% volume of ethanol. A completely different picture emerged from the chemical analysis of the wines obtained from the S. cerevisiae fermentation. Strains FB40 and ELCF3WC performed rather well, consuming more than half of the sugars present in the must and producing more than an 13.5% volume ethanol (Table 2). However, both strains showed a very high production of acetic acid: 1.54 and 1.29 g/liter for FB40 and ELCF3WC, respectively. A different behavior was observed for the EC1118 commercial strain, which showed a limited consumption of sugars (∼130 g/liter), which was correlated to a very low ethanol and acetic acid content. All of the S. cerevisiae strains were able to produce a relevant amount of glycerol (Table 2).
Table 2
Table 2 HPLC analyses of wines obtained from pure cultures of C. zemplinina and S. cerevisiaea
StrainResidual reducing sugars (g/liter)ConsumptionProductionEthanol/consumed sugar yield (g/g)
Glucose (g/liter)bFructose (g/liter)Glycerol (g/liter)Acetic acid (g/liter)Ethanol (% vol)
All C. zemplinina strains       
All S. cerevisiae strains       
Selected C. zemplinina strains       
Results are shown as minimum, maximum and average amounts for the 35 strains of C. zemplinina. Moreover, the analytical results for the three S. cerevisiae and the five C. zemplinina selected in this study are presented as averages of two fermentation trials. The glucose and fructose contents of the sweet must used in the fermentations were 210 and 193 g/liter, respectively.
ND, not detectable.
Five C. zemplinina strains were selected, on the basis of the fermentation performances and their genetic diversity, from the 35 tested. The details regarding sugar consumption, yield, and glycerol, acetic acid, and ethanol production are reported in Table 2. These strains were chosen because of the low acetic acid, low ethanol, and relevant glycerol content of the final wines and the preferred consumption of fructose with respect to glucose. The C. zemplinina EJ1strain did not comply with the above-mentioned criteria; however, it was selected due to its origin (from a different continent) and limited genetic similarity with the other C. zemplinina isolates, as shown in Fig. 1.

S. cerevisiae and C. zemplinina mixed fermentations.

The microbial dynamics of the mixed fermentations are shown in Fig. 3. Two different strategies were tested, a coinoculation of the two species and a sequential inoculation, with a delay of 48 h in the addition of S. cerevisiae with respect to C. zemplinina. The data are presented separately for the three S. cerevisiae strains, and the results of the counts (CFU/ml) are expressed as means ± the standard deviations. The C. zemplinina trends are the results of the five fermentations conducted in duplicate with the selected strains. The counts of S. cerevisiae FB40 when mixed separately with the 5 C. zemplinina strains are reported in Fig. 3A. In this case in either the coinoculated or the sequential inoculation, the C. zemplinina strains were not able to compete with S. cerevisiae, although a balance of the two species can be observed in the first 6 days. After this point, the C. zemplinina counts started to decrease and reached 103 to 104 CFU/ml after 14 days, compared to a population of S. cerevisiae that remained stable at 106 to 107 CFU/ml throughout the whole period. Very similar behavior was observed for S. cerevisiae ELCF3WC (Fig. 3B). Again, in this case, the C. zemplinina strains could not compete and started to decrease in numbers after 6 days. In order to exclude possible killer toxin activity of S. cerevisiae FB40 and ELCF3WC toward the C. zemplinina strains, tests were carried out as described by Pérez et al. (22) and resulted in no inhibition in any of the cases (data not shown). A completely different picture emerged for the EC1118 commercial strain, which was only able to dominate the fermentation in the case of the coinoculation (Fig. 3C). It is interesting that in the case of the sequential inoculation, the C. zemplinina strains were able to dominate EC1118 until day 12, and this was most probably due to the scarce capability of growth that this S. cerevisiae showed when inoculated after C. zemplinina.
Fig 3
Fig 3 Growth dynamics of the S. cerevisiae strains (FB40 [A], ELCF3WC [B], and EC1118 [C]) coinoculated (solid symbols) or sequentially inoculated (empty symbols) with C. zemplinina, as determined on the WLN medium. The mean CFU/ml values ± the standard deviations of five fermentations (each with a different C. zemplinina strain) are shown for each S. cerevisiae, whereas all of the data were combined and are presented as the mean CFU/ml ± the standard deviations for the five selected strains of C. zemplinina. Abbreviations: Cz_C, C. zemplinina in coinoculation; Sc_C, S. cerevisiae in coinoculation; Cz_S, C. zemplinina in sequential inoculations; Sc_S, S. cerevisiae in sequential inoculations.
The chemical composition of the wines obtained from the fermentations carried out by coinoculation and sequential inoculation are presented in Tables 3 and 4, respectively. In the coinoculated fermentations, the three different S. cerevisiae strain-C. zemplinina combinations resulted in a significantly different consumption of sugars. As can be seen, the EC1118 strain combination always performed poorly, leaving high quantities of sugars at day 14, regardless of what C. zemplinina was used. On the contrary, the ELCF3WC combination always showed good fermentation properties, and this resulted in high ethanol content and low residual sugars. The glucose/fructose (G/F) ratio was similar for strains FB40 and ELCF3WC, while it was significantly higher for all of the wines produced with EC1118. The acetic acid content was influenced to a great extent by the S. cerevisiae strain that was used. It is interesting that S. cerevisiae EC1118, which showed the worst performance, also produced the highest amount of acetic acid unless it was inoculated with C. zemplinina T1Y3. This observation could be considered in conflict with its behavior in pure culture; however, in such a condition, the strain consumed very little sugar and produced limited amounts of ethanol and acetic acid. On the contrary, wines fermented with the ELCF3WC strain always contained higher amounts of alcohol (>12.6%vol.) and glycerol (>15 g/liter), whereas strain FB40 always produced low quantities of glycerol. However, it did not always produce low quantities of alcohol.
Table 3
Table 3 Chemical composition of wines obtained from coinoculated fermentations of S. cerevisiae and C. zemplinina strains
Strain(s)Avg ± SDa
C. zemplininaS. cerevisiaeReducing sugar (g/liter)Glucose/fructose (−)Ethanol (% vol)Ethanol/consumed sugar yield (g/g)Glycerol (g/liter)Acetic acid (g/liter)
EJ1FB40240 ± 2b,γ0.69 ± 0.01a,α11.2 ± 0.1a,α0.454 ± 0.009a,α13.2 ± 0.1a,α1.21 ± 0.01b,β
 EC1118255 ± 5c,β0.87 ± 0.07b,α10.6 ± 0.4a,α0.467 ± 0.028a,α14.0 ± 0.2b,α1.12 ± 0.07b,β
 ELCF3WC223 ± 4a,δ0.70 ± 0.05a,α12.6 ± 0.3b,α0.471 ± 0.023a,α15.2 ± 0.1c,α0.92 ± 0.05a,α
    Sig1 ********NS******
T1Y3FB40228 ± 3b,β0.66 ± 0.04a,α12.3 ± 0.2b,β0.471 ± 0.018a,α14.2 ± 0.1a,δ1.50 ± 0.04b,γ
 EC1118235 ± 4c,α0.91 ± 0.05b,α11.6 ± 0.3a,β0.462 ± 0.025a,α14.7 ± 0.1b,β0.96 ± 0.05a,α
 ELCF3WC183 ± 3a,α0.60 ± 0.04a,α14.9 ± 0.2c,γ0.468 ± 0.015a,α15.2 ± 0.1c,α1,02 ± 0.04a,β
    Sig1 *********NS******
BC60FB40206 ± 3a,α0.61 ± 0.05a,α13.4 ± 0.3b,γ0.465 ± 0.019a,α13.8 ± 0.1a,γ1.12 ± 0.05a,αβ
 EC1118248 ± 5b,β0.83 ± 0.07b,α11.1 ± 0.4a,αβ0.462 ± 0.037a,α14.5 ± 0.2b,β1.55 ± 0.07b,γ
 ELCF3WC197 ± 4a,β0.65 ± 0.06a,α13.6 ± 0.4b,β0.454 ± 0.023a,α15.1 ± 0.2c,α1.22 ± 0.06a,γ
    Sig1 *******NS******
L37FB40222 ± 2b,β0.70 ± 0.03a,α12.3 ± 0.1b,β0.460 ± 0.011a,α13.5 ± 0.1a,β1.04 ± 0.03a,α
 EC1118249 ± 3c,β0.91 ± 0.05b,α10.8 ± 0.3a,αβ0.463 ± 0.025a,α14.5 ± 0.1b,β1.62 ± 0.05b,γ
 ELCF3WC189 ± 1a,α0.63 ± 0.02a,α14.7 ± 0.1c,γ0.474 ± 0.007a,α15.3 ± 0.2c,αβ1.02 ± 0.02a,β
    Sig1 *********NS******
PEDRO10FB40225 ± 3b,β0.68 ± 0.04a,α12.0 ± 0.2b,β0.453 ± 0.017a,α13.4 ± 0.1a,β1.13 ± 0.04a,αβ
 EC1118238 ± 3c,α0.82 ± 0.04b,α11.4 ± 0.2a,β0.459 ± 0.020a,α14.6 ± 0.1b,β1.22 ± 0.04ab,β
 ELCF3WC209 ± 2a,γ0.65 ± 0.05a,α13.8 ± 0.1c,β0.488 ± 0.011a,α15.7 ± 0.1c,β1.27 ± 0.05b,γ
    Sig1 ********NS****
Sig2 ***, **, ***NS, NS, NS***, *, ***NS, NS, NS***, **, ******, ***, ***
The ethanol content of the initial must was 1.8%. All data are expressed as averages (n = 2). Different superscript Roman letters within the same column indicate significant differences (Sig1) among the different S. cerevisiae strains inoculated with the same strain of C. zemplinina (Tukey's test; P < 0.05). Different superscript Greek letters within the same column indicate significant differences (Sig2) for different C. zemplinina strains inoculated with the same strain of S. cerevisiae (Tukey's test; P < 0.05). *, **, ***, and NS indicate significance at P < 0.05, P < 0.01, and P < 0.001 and no significant difference, respectively.
Table 4
Table 4 Chemical composition of wines obtained from sequential inoculated fermentations of S. cerevisiae and C. zemplinina strains
Strain(s)Avg ± SDa
C. zemplininaS. cerevisiaeReducing Sugar (g/liter)Glucose/fructose (−)Ethanol (% vol)Ethanol/consumed sugar yield (g/g)Glycerol (g/liter)Acetic acid (g/liter)
EJ1FB40242 ± 2b,α2.06 ± 0.03b,γ11.3 ± 0.2b,β,γ0.465 ± 0.016a,β14.6 ± 0.1c,δ1.03 ± 0.03b,γ
 EC1118288 ± 3c,γ2.04 ± 0.04b,β8.9 ± 0.2a,βγ0.466 ± 0.030a,β12.1 ± 0.1a,β0.70 ± 0.04a,β
 ELCF3WC228 ± 3a,βγ1.56 ± 0.04a,α11.4 ± 0.3b,β,γ0.432 ± 0.019a,α13.5 ± 0.1b,β0.75 ± 0.04a,γ
    Sig1 *********NS******
T1Y3FB40245 ± 5b,α1.54 ± 0.07a,α10.9 ± 0.4b,β0.453 ± 0.036b,β12.9 ± 0.2b,β0.58 ± 0.07a,α
 EC1118262 ± 2c,α2.13 ± 0.03b,β8.4 ± 0.1a,αβ0.373 ± 0.013a,α11.4 ± 0.1a,α0.50 ± 0.03a,α
 ELCF3WC236 ± 3a,γ1.56 ± 0.04a,α11.4 ± 0.2b,β0.454 ± 0.019b,α13.9 ± 0.1c,β,γ0.56 ± 0.04a,β
    Sig1 **************NS
BC60FB40247 ± 4a,α1.64 ± 0.05a,α9.1 ± 0.3b,α0.370 ± 0.023a,α12.1 ± 0.1b,α0.74 ± 0.05b,β
 EC1118271 ± 5b,αβ1.85 ± 0.07b,α7.7 ± 0.4a,α0.355 ± 0.035a,α11.2 ± 0.2a,α0.61 ± 0.07a,α,β
 ELCF3WC250 ± 4a,δ1.51 ± 0.05a,α10.4 ± 0.3c,α0.444 ± 0.026b,α14.0 ± 0.1c,γ1.01 ± 0.05c,δ
    Sig1 ***************
L37FB40247 ± 3b,α1.64 ± 0.04a,α9.6 ± 0.2b,α0.391 ± 0.018a,α12.2 ± 0.1b,α0.59 ± 0.04ab,α
 EC1118264 ± 4c,α2.04 ± 0.06b,β8.5 ± 0.3a,α,β0.378 ± 0.029a,α11.5 ± 0.2a,α0.50 ± 0.06a,α
 ELCF3WC214 ± 5a,α1.52 ± 0.07a,α11.9 ± 0.4c,β0.422 ± 0.029a,α13.7 ± 0.2c,β,γ0.67 ± 0.07b,β,γ
    Sig1 *********NS******
PEDRO10FB40242 ± 3b,α1.79 ± 0.05a,β11.9 ± 0.3b,γ0.475 ± 0.023b,β14.1 ± 0.1b,γ0.56 ± 0.05b,α
 EC1118278 ± 4c,β2.09 ± 0.06b,β9.4 ± 0.4a,γ0.462 ± 0.020b,β12.3 ± 0.2a,β0.51 ± 0.06b,α
 ELCF3WC226 ± 5a,β1.88 ± 0.05a,β11.4 ± 0.4b,β0.417 ± 0.031a,α12.0 ± 0.2a,α0.37 ± 0.05a,α
    Sig1 *************
Sig2 NS, ***, ******, **, ******, **, *****, **, NS***, ***, ******, **, ***
The ethanol content of the initial must was 1.8%. All data are expressed as averages (n = 2). Different superscript Roman letters within the same column indicate significant differences (Sig1) among the different S. cerevisiae strains inoculated with the same strain of C. zemplinina (Tukey's test; P < 0.05). Different superscript Greek letters within the same column indicate significant differences (Sig2) for different C. zemplinina strains inoculated with the same strain of S. cerevisiae (Tukey's test; P < 0.05). *, **, ***, and NS in indicate significance at P < 0.05, P < 0.01, P < 0.001, and no significant differences, respectively.
In the case of the sequential inoculation (Table 4), S. cerevisiae EC1118 was again the worst performer, producing wines with high residual sugars and with a G/F ratio of between 1.85 and 2.13. When this inoculation approach was used, it is important to note that the acetic acid content was not connected to the S. cerevisiae strain but was influenced by the S. cerevisiae-C. zemplinina strain combination used in the fermentation process. C. zemplinina EJ1 was associated with wines with a low ethanol content and a generally high acetic acid content, and similar results were also obtained for the BC60 strain. The combination C. zemplinina PEDRO10 and S. cerevisiae ELCF3WC produced wines with less acetic acid (<0.40 g/liter) without affecting the ethanol (11.4% [volume]) or glycerol (12 g/liter) contents.
If the chemical parameters determined for the wines obtained with the two different inoculation approaches are analyzed together, significant differences emerge in their compositions, especially for the G/F ratio, glycerol, and acetic acid (Table 5). The influence of the strains (both S. cerevisiae and C. zemplinina), their interaction, as well as the type of inoculation used, always resulted in significant differences (P < 0.001) (data not shown). The wines obtained with sequential inoculation resulted to have higher residual sugars, with an increased G/F ratio, thereby potentially influencing their final organoleptic properties. In these cases, the fermentation products, such as ethanol and glycerol, decreased. However, the quantity of acetic acid produced was between 0.60 and 0.75 g/liter and was half that of the fermentations conducted inoculating only the S. cerevisiae strains or both species at the same time.
Table 5
Table 5 Statistical differences for the chemical composition of wines obtained from coinoculated (data from Table 3) and sequentially inoculated (data from Table 4) fermentations of S. cerevisiae and C. zemplinina strains
C. zemplininaS. cerevisiaeReducing sugar (g/liter)Glucose/fructose (−)Ethanol (% vol)Ethanol/consumed sugar yield (g/g)Glycerol (g/liter)Acetic acid (g/liter)
*, **, ***, and NS indicate significance at P < 0.05, P < 0.01, P < 0.001, and no significant differences, respectively.


In the last couple of years, a number of studies that have focused on the potential application of C. zemplinina in wine fermentations have been published (1, 18, 29, 30), mainly due to its ethanol and low temperature tolerance, osmotic resistance and fructophylic character.
In the present study, we have specifically investigated the possibility of using C. zemplinina in sweet wine fermentations, sequentially or coinoculated with S. cerevisiae. All of the fermentations were carried out in natural must obtained from dried grapes in order to mimic the real conditions encountered during the production of sweet wines and avoid the use of laboratory media that can give a totally different picture in terms of yeast fermentative behavior.
A set of 35 isolates of C. zemplinina was first molecularly characterized, and the results obtained underlined a relative genetic homogeneity within the strains tested. There were no differences, in terms of clustering, on the basis of the geographic distribution, and most of the strains formed two large clusters. When these strains were tested in fermentation trials of must obtained from dried grapes, their fructophylic character was confirmed, as previously described (25, 26). Moreover, their ability to produce relevant quantities of glycerol and low amounts of acetic acid was also confirmed, in agreement with other studies (18, 29). Interestingly, the behavior of S. cerevisiae was different between the commercial strain and the wild isolates from sweet wine fermentations. As shown in Table 2, the EC1118 commercial strain performed worse than the two wine isolated strains, since a high level of residual sugar, associated with lower ethanol production, was detected at the end of the monitoring period. A notable production of acetic acid was concomitantly observed, most likely due to the osmotic stress provoked by the high concentration of sugars, responsible for the upregulation of the genes encoding for aldehyde dehydrogenases (12). It should be underlined that differences in the vitality counts were observed for the three S. cerevisiae strains when inoculated as pure culture in the must utilized here. Although the wild strains were able to promptly increase in cell counts, reaching 108 CFU/ml at day 1, EC1118 never reached a load of 107 CFU/ml and started to decrease after 6 days of fermentation (data not shown). This evidence allows us to speculate that the wild strains may adapt well to an environment, similar to the one from where they were isolated (i.e., musts with a high sugar concentration), and underlines the need for better performing strains than the commercial one used in the present study.
Five strains of C. zemplinina were selected, on the basis of their genetic characteristics and fermentation performances, for the mixed fermentation experiments. As shown in Fig. 1, the BC60 and PEDRO10 strains were located in cluster 1 (although in two different subclusters), while the T1Y3 and L37 strains grouped in cluster 2 (but again in two different subclusters). The last strain selected, EJ1, was genetically far from the first four described. Considering the chemical parameters of the wines obtained with these strains (Table 2), it can be observed that they all resulted to be homogeneous, apart from C. zemplinina EJ1, which again showed relevant differences with respect of the other strains selected.
Remarkably, all of the selected C. zemplinina showed comparable growth kinetics, regardless of which S. cerevisiae strain was used. This aspect is highlighted by the contained standard deviations reported in Fig. 3. In other words, the five selected C. zemplinina showed similar growth curves when coupled with S. cerevisiae, both in the coinoculation and in the sequential inoculation. The main differences were observed in the trends of S. cerevisiae EC1118, which was not able to dominate the fermentation in the case of sequential inoculation. It is interesting that the counts reached by this strain in mixed fermentations were higher than in pure culture, underscoring that C. zemplinina strains could facilitate the growth of S. cerevisiae EC1118.
Although no variations were observed, in terms of growth, in the mixed fermentation experiments for each strain of S. cerevisiae, relevant differences were detected for the chemical composition of the wines. Again, in this case, S. cerevisiae EC1118 showed poor fermentation power, whereas S. cerevisiae ELCF3WC was able to produce more ethanol and glycerol and less acetic acid. However, as described above, C. zemplinina influenced the production of acetic acid by S. cerevisiae. More specifically, a coinoculation of C. zemplinina T1Y3 or L37 with S. cerevisiae ELCF3WC resulted in wines with a high ethanol content (>14.5% [by volume]) and acetic acid of ∼1 g/liter (21% acetic acid reduction, compared to the pure ELCF3WC fermentation).
This study has demonstrated that the fermentation of musts, characterized by a high sugar content, with S. cerevisiae and C. zemplinina mixtures, may contribute to control the acetic acid production by S. cerevisiae. The data presented here support the use of this non-Saccharomyces species, but at the same time the specific S. cerevisiae-C. zemplinina combination is important. As shown, the possibility to reduce the acetic acid content is closely connected to the strain combination and the type of inoculation performed. Moreover, other chemical parameters, such as higher alcohols and acetaldehyde should be monitored, since it has been pointed out that C. zemplinina can contribute to a great extent to their increase or decrease, respectively (1, 18). Since all of the data presented here were obtained from pasteurized must, more investigations are necessary to assess the ability of C. zemplinina to compete with the natural microbiota of grape musts and to confirm that the mechanism responsible for the acetic acid reduction is due to S. cerevisiae osmotic stress relief.


We thank Kyria Boundy-Mills from the Phaff Collection, University of Davis, Davis, CA, and George-John Nychas from the Agricultural University of Athens, Athens, Greece, for the C. zemplinina EJ1 and D1, respectively.


Andorrà I et al. 2010. Effect of pure and mixed cultures of the main wine yeast species on grape must fermentations. Eur. Food Res. Technol. 231:215–224.
Bely M, Stoeckle P, Masneuf-Pomarède I, and Dubourdieu D. 2008. Impact of mixed Torulaspora delbrueckii-Saccharomyces cerevisiae culture on high-sugar fermentation. Int. J. Food Microbiol. 122:312–320.
Brezna B et al. 2010. Evaluation of fungal and yeast diversity in Slovakian wine-related microbial communities. Antonie Van Leeuwenhoek 98:519–529.
Ciani M, Beco L, and Comitini F. 2006. Fermentation behavior and metabolic interactions of multistarter wine yeast fermentations. Int. J. Food Microbiol. 108:239–245.
Ciani M, Comitini F, Mannazzu I, and Domizio P. 2010. Controlled mixed culture fermentation: a new prospective on the use of non-Saccharomyces yeasts in wine making. FEMS Yeast Res. 10:123–133.
Ciani M and Ferraro L. 1998. Combined use of immobilized Candida stellata cells and Saccharomyces cerevisiae to improve the quality of wines. J. Appl. Microbiol. 85:247–254.
Cocolin L, Bisson LF, and Mills DA. 2000. Direct profiling of the yeast dynamics in wine fermentations. FEMS Microbiol. Lett. 189:81–87.
Cocolin L, Pepe V, Comitini F, Comi G, and Ciani M. 2004. Enological and genetic traits of Saccharomyces cerevisiae isolated from former and modern wineries. FEMS Yeast Res. 5:237–245.
Comitini F et al. 2011. Selected non-Saccharomyces wine yeasts in controlled multistarter fermentations with Saccharomyces cerevisiae. Food Microbiol. 28:873–882.
Csoma H and Sipiczki M. 2008. Taxonomic reclassification of Candida stellata strains reveals frequent occurrence of Candida zemplinina in wine fermentations. FEMS Yeast Res. 8:328–336.
Erasmus DJ, Cliff M, and van Vuuren HJJ. 2004. Impact of yeast strain on the production of acetic acid, glycerol, and the sensory attributes of Icewines. Am. J. Enol. Vitic. 55:371–378.
Erasmus DJ, van der Merwe GK, and van Vuurem HJJ. 2003. Genome-wide expression analyses: metabolic adaptation of Saccharomyces cerevisiae to high sugar stress. FEMS Yeast Res. 3:375–399.
Fleet G. 2008. Wine yeast for the future. FEMS Yeast Res. 8:979–995.
Giordano M, Rolle L, Zeppa G, and Gerbi V. 2009. Chemical and volatile composition of three Italian sweet white Passito wines. J. Int. Sci. Vigne Vin. 43:159–170.
Herraiz T, Reglero G, Herraiz M, Martin-Alvarez PJ, and Cabezudo MD. 1990. The influence of the yeast and type of culture on the volatile composition of wines fermented without sulfur dioxide. Am. J. Enol. Vitic. 41:313–318.
Li SS et al. 2010. Yeast species associated with wine grapes in China. Int. J. Food Microbiol. 138:85–90.
Magyar I and Bene ZS. 2006. Morphological and taxonomic study of mycobiota of noble rotten grapes in the Tokay wine district. Acta Alimentaria 35:237–246.
Magyar I and Toth T. 2011. Comparative evaluation of some oenological properties in wine strains of Candida stellata, Candida zemplinina, Saccharomyces uvarum, and Saccharomyces cerevisiae. Food Microbiol. 28:94–100.
Mills DA, Johannsen EA, and Cocolin L. 2002. Yeast diversity and persistence in botrytis-affected wine fermentations. Appl. Environ. Microbiol. 68:4884–4893.
Nisiotou A and Nychas GJE. 2007. Yeast populations residing on healthy or Botrytis-infected grapes from a vineyard in Attica, Greece. Appl. Environ. Microbiol. 73:2765–2768.
Pallmann CL et al. 2001. Use of WL medium to profile native flora fermentations. Am. J. Enol. Vitic. 52:198–203.
Pérez F, Ramírez M, and Regodón JA. 2001. Influence of killer strains of Saccharomyces cerevisiae on wine fermentation. Antonie Van Leeuwenhoek 79:393–399.
Renault P et al. 2009. Genetic characterization and phenotypic variability in Torulaspora delbrueckii species: potential applications in the wine industry. Int. J. Food Microbiol. 134:201–210.
Schneider A, Gerbi V, and Redoglia M. 1987. A rapid HPLC method for separation and determination of major organic acids in grape musts and wines. Am. J. Enol. Vitic. 38:151–155.
Sipiczki M. 2003. Candida zemplinina sp. nov., an osmotolerant and psychrotolerant yeast that ferments sweet botrytized wines. Int. J. Syst. Evol. Microbiol. 53:2079–2083.
Sipiczki M. 2004. Species identification and comparative molecular and physiological analysis of Candida zemplinina and Candida stellata. J. Basic Microbiol. 44:471–479.
Strauss MLA, Jolly NP, Lambrechts MG, and van Rensburg P. 2001. Screening for the production of extracellular hydrolytic enzymes by non-Saccharomyces wine yeasts. J. Appl. Microbiol. 91:182–190.
Tofalo R et al. 2009. Molecular identification and osmotolerant profile of wine yeasts that ferment a high sugar grape must. Int. J. Food Microbiol. 130:179–187.
Tofalo R et al. 2012. Diversity of Candida zemplinina strains from grapes and Italian wines. Food Microbiol. 29:18–26.
Tofalo R et al. 2011. Influence of organic viticulture on non-Saccharomyces wine yeast populations. Ann. Microbiol. 61:57–66.
Urso R et al. 2008. Yeast biodiversity and dynamics during sweet wine production as determined by molecular methods. FEMS Yeast Res. 8:1053–1062.
Zironi R, Romano P, Suzzi G, Battistuta F, and Comi G. 1993. Volatile metabolites produced in wine by mixed and sequential cultures of Hanseniaspora guillermondii or Kloeckera apiculata and Saccharomyces cerevisiae. Biotechnol. Lett. 15:235–238.
Zott K, Miot-Sertier C, Claisse O, Lonvaud-Funel A, and Masneuf-Pomarede I. 2008. Dynamics and diversity of non-Saccharomyces yeasts during the early stages in winemaking. Int. J. Food Microbiol. 125:197–203.

Information & Contributors


Published In

cover image Applied and Environmental Microbiology
Applied and Environmental Microbiology
Volume 78Number 615 March 2012
Pages: 1987 - 1994
PubMed: 22247148


Received: 1 September 2011
Accepted: 1 January 2012
Published online: 22 February 2012


Request permissions for this article.



Kalliopi Rantsiou
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy
Paola Dolci
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy
Simone Giacosa
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy
Fabrizio Torchio
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy
Rosanna Tofalo
Dipartimento di Scienze degli Alimenti, University of Teramo, Teramo, Italy
Sandra Torriani
Dipartimento di Biotecnologie, University of Verona, Verona, Italy
Giovanna Suzzi
Dipartimento di Scienze degli Alimenti, University of Teramo, Teramo, Italy
Luca Rolle
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy
Luca Cocolin
DIVAPRA, Agricultural Microbiology and Food Technology Sector, University of Turin, Turin, Italy


Address correspondence to Luca Cocolin, [email protected].
K.R. and P.D. contributed equally to this article.

Metrics & Citations



  • For recently published articles, the TOTAL download count will appear as zero until a new month starts.
  • There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
  • Citation counts come from the Crossref Cited by service.


If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media






Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy