Mathier.ch

Hindawi Publishing CorporationBioMed Research InternationalVolume 2013, Article ID 540465, 13 pageshttp://dx.doi.org/10.1155/2013/540465 Research Article
Characterization and Dynamic Behavior of Wild Yeast during
Spontaneous Wine Fermentation in Steel Tanks and Amphorae
Cecilia Díaz,1 Ana María Molina,2 Jörg Nähring,1 and Rainer Fischer1
1 Molecular Biology Division, Fraunhofer Institute for Molecular Biology and Applied Ecology, 57392 Schmallenberg, Germany2 Facultad de Ingenier´ıa y Tecnolog´ıa, Universidad San Sebasti´an, 4030000 Concepcion, Chile Correspondence should be addressed to Cecilia D´ıaz; [email protected] Received 22 January 2013; Revised 28 March 2013; Accepted 9 April 2013 Academic Editor: George Tsiamis Copyright 2013 Cecilia D´ıaz et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We studied the dynamic behavior of wild yeasts during spontaneous wine fermentation at a winery in the Valais region ofSwitzerland. Wild yeasts in the winery environment were characterized using a PCR-RFLP method. Up to 11 different yeastspecies were isolated from the vineyard air, whereas only seven were recovered from the grapes surface. We initially investigateda cultureindependent method in pilot-scale steel fermentation tanks and found a greater diversity of yeasts in the musts from twored grape varieties compared to three white grape varieties. We found that the yeasts Metschnikowia pulcherrima, Rhodotorulamucilaginosa, Pichia kluyveri, P. membranifaciens and Saccharomyces cerevisiae remained active at the end of the fermentation. Wealso studied the dynamic behavior of yeasts in Qvevris for the first time using a novel, highlysensitive quantitative real-time PCRmethod. We found that non-Saccharomyces yeasts were present during the entire fermentation process, with R. mucilaginosa and P.
anomala the most prominent species. We studied the relationship between the predominance of different species and the output ofthe fermentation process. We identified so-called spoilage yeasts in all the fermentations, but high levels of acetic acid accumulatedonly in those fermentations with an extended lag phase.
because no additional wine yeasts are introduced into theprocess.
Low-intervention winemaking methods based on sponta- Wine flavor is influenced by the large number of neous fermentation are becoming more popular among yeast species present during spontaneous fermentation [3, wine producers and consumers [1, 2]. Some wine producers 8–11], including those from the genera Hanseniaspora, and viticulturists have readopted traditional winemaking Metschnikowia and Candida, and more occasionally Toru- methods to generate unique attributes that differentiate their laspora and Pichia. Most of these non-Saccharomyces yeasts products, improve wine quality, and increase the variety of grow during the early fermentation stages, whereas the complex flavors that characterize regional vineyards.
process is eventually completed by Saccharomyces cerevisiae Spontaneous fermentation is a complex process influ- because it can tolerate higher levels of alcohol and lower levels enced by many factors, including the endogenous microbial of oxygen [9–15].
flora, the grape variety, climatic conditions, and the winemak- Previous studies have shown that non-Saccharomyces ing process [3–7]. The outcome of the fermentation process yeasts can be detected throughout the fermentation process can therefore be difficult to predict and can differ from [15]. They influence the course of fermentation and the char- year to year. The natural yeast flora, found on grapes and acteristics of the resulting wine by producing extracellular in wineries, play a significant role during fermentation and enzymes and metabolites of oenological significance that are particularly important during spontaneous fermentation modify the sensory and organoleptic properties of wine, BioMed Research International introducing a broader spectrum of aromas and flavors [16– Table 1: Grape varieties used for the analysis of dynamic wild yeast populations during spontaneous fermentation in stainless steel Microbiology techniques are often used to isolate and tanks and Qvevris.
identify wild yeasts, but this requires different types of media and different culture protocols that influence specieswhich are recovered. The metabolic status of the cells also 2008-2009 (steel tank) 2010-2011 (Qvevris) results in the presence of viable but nonculturable (VBNC) Pinot Noir (R) Chardonnay (W) Resi (W) microbes whose influence on fermentation can be underesti- mated because the population dynamics cannot be evaluated Petit Arvine (W) Ermitage (W) accurately. Quantitative real-time PCR (qPCR) is a faster and more reliable alternative to identify and quantify yeasts (R): red variety; (W): white variety.
during fermentation [19] and is particularly advantageousfor VBNC yeasts because of its sensitivity [20]. Althoughthe technique cannot distinguish living cells from intact from the inner surface of clean fermentation tanks before dead cells, it remains the most widely used method for filling them with grape juice, and 1000 L of air inside the the evaluation of wild yeast dynamics during fermentation cellar was also filtered and the collected residues were plated because VBNC cells continue to influence wine flavor and as above. All the environmental samples were collected in palatability regardless of their actual status [21].
We compared microbiology methods (viable counts) Five different grape varieties from the 2008-2009 harvests and novel molecular biology techniques: polymerase chain (Table 1) were processed by spontaneous fermentation to reaction/restriction fragment length polymorphism (PCR- determine the predominant yeast species at the different RFLP) and qPCR for the identification of yeast species, and fermentation stages. Prefermentation steps such as harvesting we characterized their dynamic behavior during spontaneous and pressing were carried out according to routine winery wine fermentation in the Valais region of Switzerland in procedures. Pressed berries were fermented with the skin the 2008–2011 harvests. We used these new methods to to make red wine, or were clarified before fermentation to identify the predominant species present during spontaneous produce white wine. Duplicate fermentations were carried fermentation, establishing a standard for the semiquantitative out in the winery cellar using new 110-L stainless steel tanks, detection of yeasts with antibodies in a biochip format. Such a without starting yeast cultures. Liquid samples (50 mL, in device would allow winemakers to make early decisions about duplicate) from the fermenting musts were collected daily, the suitability of grapes and the likely success of spontaneous frozen immediately at −20∘C, and stored in the dark prior fermentation. Also for the first time, we studied the dynamic to analysis. Immediately after defrosting, liquid samples were behavior of wild yeasts during spontaneous fermentation in centrifuged at 4000 rpm for 5 min. The supernatant was Qvevris (amphora-like clay vessels), the use of which is an tested for chemical parameters, and the pellet was resus- emerging trend among European winemakers.
pended in 100 𝜇L distilled water, plated on RBCA mediumand incubated at 30∘C for 3–7 days, and then stored at 4∘C 2. Materials and Methods
prior to analysis.
2.1. Samples. Grapes and must samples were collected from 2.3. Fermentation Parameters. The fermentations were mon- the vineyards of the winery Albert Mathier et Fils S.A., in itored by measuring glucose/fructose consumption and Salgesch, Valais, Switzerland, during the 2008–2011 harvest ethanol formation during fermentation, and the acetic acid seasons. The grape varieties we studied are listed in Table 1.
content at the end of fermentation. The parameters were Samples were collected in situ and frozen at −20∘C during determined by spectrophotometry at 20 ± 1∘C using the D- transport prior to analysis.
Glucose/D-Fructose, Ethanol, and Acetic Acid enzymatic kitsprovided by R-Biopharm (Germany), according to the manu- 2.2. Isolation of Yeasts from the Winery Environment and facturer's instructions. Standards and controls were provided Fermentation Samples. During 2008 and 2009, we screened in the kits. All measurements (duplicate fermentations) were yeasts present in the winery environment (i.e., the vineyard, taken in triplicate.
winery facilities and cellar) and in the fermenting wine musts.
In the vineyard, grape berries were placed in direct contact 2.4. Identification of the Predominant Yeast Species. A ter- with plates containing Rose Bengal Chloramphenicol Agar minal restriction fragment length polymorphism (T-RFLP) (RBCA), a selective medium for yeasts and molds (15 g L−1 method was developed and optimized for yeast identification, agar, 10 g L−1 glucose, 5 g L−1 papain-digested soybean meal, based on restriction patterns generated from the genomic 1 g L−1 KH2PO4, 0.5 g L−1 MgSO4 × 7H2O, 0.05 g L−1 Rose region spanning the internal transcribed spacers (ITS1 and Bengal, and 10 g L−1 chloramphenicol). We pumped 1000 L of ITS2) and the 5.8S rRNA gene. These regions show low vineyard air surrounding the grapes through a Millipore M intraspecific polymorphism and high interspecific variability Air Tester T (Millipore, USA) and plated the collected residue and have previously been shown to distinguish 26 yeast on RBCA as above. We also sampled environmental yeast species found on grapes, in cellars and/or in wine musts flora from the winery facilities. Contact samples were taken BioMed Research International Total DNA from the isolated colonies was extracted agar (10 g L−1 yeast extract, 20 g L−1 peptone, 20 g L−1 glucose, using the First-Beer Magnetic DNA kit (GEN-IAL GmbH, and 0.1 mg L−1 chloramphenicol) at 27∘C for 24 h. The cells Troisdorf, Germany) and amplified using primers ITS-5 (5󸀠- were counted using a Neubauer chamber. DNA was extracted GGA AGT AAA AGT CGT AAC AAG G-3󸀠) and ITS-4 (5󸀠- using the First-Beer Magnetic DNA kit and serially diluted TCC TCC GCT TAT TGA TAT GC-3󸀠) followed by a second (1 : 10) from 107-108 down to 1 cell mL−1. Each point on the round of amplification with the nested primers ITS-1 (5󸀠-TCC calibration curve was measured in duplicate. Conventional GTA GGT GAA CCT GCG G-3󸀠) and ITS-2 (5󸀠-GCT GCG and real-time PCR was carried out using a range of yeast TTC TTC ATC GAT GC-3󸀠) [24]. The first reaction mixture species to verify the specificity of each primer set.
comprised 200 𝜇M of each dNTP, 10x PCR buffer, 0.5 𝜇M ofeach primer, 1 𝜇L of extracted yeast DNA and 1.25 U Hot Start 2.6. Wild Yeast Dynamics during Spontaneous Fermentation in Polymerase (5 Prime, Hamburg, Germany) in a total volume Qvevris. Spontaneous fermentation in Qvevris was studied of 25 𝜇L. The samples were amplified in a thermocycler during the 2010 and 2011 harvest seasons. The white grape (VRW, Pennsylvania, USA) by denaturing at 95∘C for 3 min varieties Resi and Ermitage (Table 1) were harvested, crushed, followed by 15 cycles of denaturing at 95∘C for 30 s, annealing and fermented in 1500-L Qvevris without clarification. We at 57∘C for 30 s and extension at 72∘C for 1 min, and a final took 50-mL samples in triplicate at 2-3-day intervals through- extension step at 72∘C for 5 min. The nested amplification out fermentation; that is, every time the Qvevris were opened mixture comprised 200 𝜇M of each dNTP, 10x PCR buffer, to stir the must. The samples were frozen immediately at 0.5 𝜇M of each primer (labeled if necessary for product size −20∘C and stored in the dark prior to analysis. DNA was determination, see below), 2.5 U Hot Start Polymerase (5 extracted from the must using a modified CTAB method Prime, Hamburg, Germany), and 0.5 𝜇L template DNA (from [25] in which 10 mL samples were centrifuged for 1 min at the first-round PCR) in a total volume of 50 𝜇L. The mixture 3000 rpm to sediment the skin and seeds before the standard was denatured at 95∘C for 3 min then amplified by 20 cycles protocol was applied. The extracted DNA was then tested of denaturing at 95∘C for 30 s, annealing at 62∘C for 30 s and by qPCR to identify the wild yeast species present during extension at 72∘C for 1 min, followed by a final extension spontaneous fermentation as discussed above.
at 72∘C for 5 min. The products were digested with BstYI(New England BioLabs, Ipswich) at 60∘C for 1 h. The length ofthe terminal fragment was determined using a 3130 Genetic 3. Results
Analyzer (Applied Biosystems, Darmstadt, Germany) priorto the purification of the samples using the Cycle Pure Kit 3.1. Establishing a PCR-RFLP Method for Yeast Identification. (Omega Bio-tek, USA).
Yeast genomic DNA was amplified using primers ITS4and ITS5 (first round), and the products were amplifiedwith the nested primers ITS1 and ITS2. The sizes of both 2.5. Primer Design and Real-Time PCR. Primers specific the digested and undigested PCR products are unique to for the 10 predominant yeasts found in the winery and particular yeast genera and also allow the differentiation of in fermentation samples during the 2008-2009 harvest certain species, resulting in the unambiguous identification season were designed to anneal within the 26S rDNA of up to 28 species (Table 3). There were only three cases in region and amplify products 150–200 bp in length (Table 2).
which we were unable to distinguish two different species: Each primer pair was designed by processing available (1) Hanseniaspora guilliermondii and H. uvarum; (2) Saccha- sequences using CLC Combined Workbench 3 Software romyces bayanus and S. pastorianus; (3) Dekkera bruxellensis (CLC-Bio, Denmark), and the properties of each primer and Cryptococcus flavus. The method was optimized using were verified using Primer Tool (Sigma-Aldrich, USA; http:// species obtained from the Deutsche Sammlung von Mikroor- ganismen und Zellkulturen GmbH (DSMZ), Braunschweig, of each primer pair was controlled by searching GenBank Germany. Even so, wild yeast species in wineries are often using BLAST (http://www.ncbi.nlm.nih.gov /BLAST/).
local subspecies that are subject to different environmental Real-time PCR was carried out using an ABI 7300 Real- selection conditions and their sequences and PCR product Time PCR System (Applied Biosystems, Hitachi, Japan).
sizes can differ slightly from purchased strains. Therefore, Each reaction comprised 7.5 𝜇L Platinum SYBR Green qPCR and in order to validate the method, we selected 4 isolated SuperMix-UDG (Bio-Rad, Hercules, CA, USA), 200 nM of yeasts and sequenced the first-round PCR products (NCBI each primer (Metabion, Germany), and 0.3 𝜇L template DNA accession numbers KC869927, KC869928, KC869929, and extracted from must, in a total volume of 15 𝜇L. The mixture KC869930) to compare these empirical sequences to those in was heated to 50∘C for 2 min and then 95∘C for 2 min, GenBank by using the empirical sequences as BLAST queries.
followed by 40 cycles of denaturation at 95∘C for 15 s, andannealing/extension at 60–63∘C (depending on the primers) 3.2. Natural Flora in Vineyard and Cellar Environments. for 45 s. The cycling temperature was then increased by 0.3∘C Yeasts naturally present in the vineyard environment were every 10 s from 63 to 95∘C to obtain the melting curve. The isolated from the grape surface and from the air around the DNA concentration in the samples was limited to 50 ng per grapes using culture-dependent methods (see Section 2.2).
analysis, except for standard curves prepared from samples During 2008 and 2009, up to 11 different yeast species could containing a known number of yeast cells. All yeast species be isolated from the vineyard air although Bulleromyces albus were cultivated in Yeast Extract Peptone Dextrose (YPD) and Sporidiobolus pararoseus were the only species recovered BioMed Research International Table 2: Specific primers used for qPCR analysis.
Annealing temperature ∘C Pichia kluyveri Pichia angusta Pichia anomala Candida glabrata Pichia fermentans in both years (Table 4). We recovered seven yeast species from varieties Pinot Noir and Cornalin, which contained 12 and the grape surface, and species appear to be dependent on 9 different yeast species, respectively, and the white varieties the variety and the year of harvest (Table 4). Aurebasidium Gutedel, Chardonnay, and Petite Arvine, which contained 12, pullulans, Cryptococcus magnus, Rhodotorula mucilaginosa, 12, and 7 yeast species, respectively (Table 5).
and Zygosaccharomyces florentinus were the only species Most of the yeast species we identified were present isolated from both the air and the grape surface. Most of in more than one of the musts (Table 5). M. pulcherrima, the yeasts isolated from the vineyard air were also present S. cerevisiae, S. bayanus, and T. delbrueckii were found in the grape juice at the beginning of fermentation. In the in all five musts at some point during fermentation. Six cellar environment, yeasts were isolated from the surface of yeast species were only found in one type of must, and clean and empty barrels (i.e., before filling the fermentation only at the beginning of fermentation. Four yeast species tanks with the grape must) and from the air inside the cellar found in Gutedel musts did not grow in any of the other room. In 2008, four different species were isolated from musts: Bulleromyces albus, Candida zeylanoides, Cryptococcus the cellar air and three from the clean fermentation tank flavus/Dekkera bruxellensis (the latter could not be distin- (Table 4), whereas in 2009 only one species (R. mucilaginosa) guished on the basis of their PCR-RFLP patterns), and Filoba- was isolated from the clean fermentation tank. All the species sidium floriforme. Similarly, Pichia burtonii and P. holstii only in the cellar environment were also found in the vineyard, and found in Cornalin musts (Table 5). There were no species all species identified in the cellar environment were also later associated exclusively with red or white grape varieties.
found in the fermenting must.
The composition of the yeast populations also changed significantly during fermentation. Initially, 7–12 differentspecies were found in the musts (depending on the variety), 3.3. Yeast Flora in Steel-Tank Fermentations. Changes in the but this declined to 1–5 species by the midfermentation, when composition of the yeast population during spontaneous nitrogen becomes limiting and the ethanol concentration fermentation in steel tanks were measured using culture- begins to increase rapidly (Table 5). By the end of fermenta- dependent methods. We investigated the musts of five grape tion, only six different yeast species could be recovered from varieties during the 2008 and 2009 harvest seasons: the red the musts. The ethanol-resistant strain S. bayanus made up BioMed Research International Table 3: Sizes of digested and undigested nested PCR products representing different yeast species derived using the T-RFLP Candida glabrata Cornalin Chardonnay Petit Arvine Gutedel Pichia klyveri Figure 1: Yeast population recovered at the end of spontaneous fermentation in steel tanks during the 2008 and 2009 harvest as well as acetic acid production (Table 6). The red varieties Pinot Noir and Cornalin reached dryness (less than 4 g L−1 of total sugar) 6–11 days after pressing (Figure 2). The lag phase Metschnikowia sp. of the Pinot Noir fermentations in 2008 (i.e., the period before Pichia angusta glucose consumption increases rapidly) was relatively longcompared to the Cornalin fermentations in the same year (2- Pichia anomala 3 days) (Figure 2). In contrast, the white grape varieties failed Pichia fermentans to reach dryness in fermentations during 2008 and 2009, and Pichia holstii the Petite Arvine and Chardonnay vessels contained high Pichia kluyveri levels of residual sugar at the end of fermentation (Figure 2).
In 2008, the fermentation of Gutedel grapes was delayed at the midexponential phase (days 6–13) whereas Petite Arvinewas characterized by sluggish fermentation from the late exponential phase (day 11) onwards (Figure 2).
3.4. Real-Time PCR. Thirteen pairs of specific primers were designed for the rapid identification and quantification of Torulaspora delbrueckii (wild yeast the yeast species we detected. The primers designed for Z. florentinus, C. glabrata, and P. fermentans showed evidence of nonspecific annealing and were therefore eliminated from the study. The sequences and annealing temperatures of the remaining primers are summarized in Table 2. The melt curveanalysis for each PCR showed a single peak (data not shown).
Standard curves were established for each pair of primers. The a substantial proportion of the yeasts in all musts (Figure 1) reaction efficiencies ranged between 72.54% (P. anomala) and and was the only strain detected in Gutedel musts during the 98.68% (S. cerevisiae) with high reproducibility. The lowest mid- and late fermentation stages. In contrast, S. cerevisiae detection limit was 102 cells L−1.
was found in the Chardonnay, Pinot Noir, and Petite Arvinemusts at the end of fermentation, and M. pulcherima was 3.5. Yeast Flora in Qvevri Fermentations. The dynamic behav- present in the Chardonnay, Pinot Noir, and Cornalin musts ior of the yeast populations in Qvevri spontaneous fermen- at the end of fermentation. The other species retrieved at tations was monitored by qPCR during the 2010 and 2011 the end of the fermentation were P. klyveri (Chardonnay and harvest seasons. There was a slight tendency towards higher Cornalin musts), P. membranifaciens (Chardonnay must), yeast diversity in the Resi variety compared to Ermitage, with and R. mucilaginosa (Pinot Noir must) (Figure 1).
10 and 8 different yeast species, respectively (Table 7). Most The progress of fermentation was monitored by measur- of the species were present in varieties, and M. pulcherima, ing sugar consumption and ethanol production (Figure 2), R. mucilaginosa, P. anomala, H. uvarum, S. cerevisiae, and BioMed Research International n (g tiotra 100ncenoC Figure 2: Spontaneous fermentation profile, expressed in g L−1, for the white grape varieties in stainless steel tanks during the 2008 and 2009harvest seasons. (a) Pinot Noir; (b) Cornalin; (c) Chardonnay; (d) Petit Arvine; (e) Gutedel.
BioMed Research International Table 4: Yeasts isolated from the winery environment during the 2008 and 2009 harvest seasons.
Pichia angustaPichia anomala Pichia kluyveri ∗Varieties of grapes and musts: Pinot Noir (Pn); Cornalin (Co); Chardonnay (Ch); Petite Arvine (Pa); Gutedel (Gu).
Table 5: Yeasts found during spontaneous fermentation in stainless steel tanks, during the 2008 and 2009 harvest seasons.
Candida Krusei or Issatchenkia orientalis Cryptococcus flavus or Dekkera bruxellensis Pichia anomala Pichia burtonii Pichia holstii Pichia kluyveri a: detected at the beginning of the fermentation; b: detected during log phase; c: detected during stationary phase.
T. delbrueckii were also found at every fermentation stage fermentations during 2010 and 2011 (Table 7), and in both (Figures 3 and 4). In contrast, C. zemplinina and P. angusta cases the dominant species in the must before fermentation were found only in the Resi variety, and although P. kluyveri were R. mucilaginosa, and P. anomala although they were was found in both varieties, it was present only at certain more abundant in 2011 (Figure 3). The less-abundant species fermentation stages during the 2010 harvest and was not were H. uvarum, S. cerevisiae, T. delbrueckii, and C. zem- detected in 2011 (Table 7).
plinina, although all of them were present throughout the R. mucilaginosa was the dominant species in the 2011 fermentation. These species were 10 times more abundant Ermitage fermentations whereas P. anomala was the domi- in 2010 than in 2011, except R. mucilaginosa, which was nant species in the Resi fermentations in both harvest years.
more abundant in 2011. S. cerevisiae was the most abundant Up to eight yeast species were detected in the Ermitage species at the beginning of the 2010 fermentations and it BioMed Research International M. pulcherrima S. cerevisiae M. pulcherrima S. cerevisiae P. anomala P. anomala R. mucilaginosa R. mucilaginosa Figure 3: Dynamic behavior of wild yeast populations during the spontaneous fermentation of Ermitage grapes in Qvevris, measured byqPCR: (a) 2010 harvest; (b) 2011 harvest.
M. pulcherrima S. cerevisiae M. pulcherrima S. cerevisiae P. anomala P. anomala R. mucilaginosa R. mucilaginosa Figure 4: Dynamic behavior of wild yeast populations during the spontaneous fermentation of Resi grapes in Qvevris, measured by qPCR:(a) 2010 harvest; (b) 2011 harvest.
proliferated rapidly, reaching its maximum concentration 2010 than 2011 although the onset of fermentation in 2011 (1 × 106 cells mL−1) by day 6. In contrast, R. mucilaginosa was more rapid, beginning after 1 day (Figure 4). In 2010, was the most abundant species in the 2011 fermentations the highest concentration of S. cerevisiae (1 × 106 cells mL−1) (1 × 106 cells mL−1) and S. cerevisiae proliferated more slowly, was achieved 2 days after fermentation began, whereas in reaching its maximum concentration after 14 days. In 2010, 2011 the concentration increased rapidly during the first day the yeast population declined slowly during fermentation and remained high until the end of the fermentation (1 × whereas in 2011 the P. anomala, S. cerevisiae, and R. mucilagi- 106 cells mL−1).
nosa populations remained high (Figure 3).
The progress of the Ermitage and Resi fermentations was The Resi fermentations during 2010 and 2011 began monitored and compared. The onset of the Ermitage fer- rapidly (before 5 days in both cases) with P. anomala mentation took longer in 2010 but was nevertheless complete dominating throughout fermentation and H. uvarum and T. after 14 days in both 2010 and 2011 (Figure 5). The Ermitage delbrueckii present at lower levels (Figure 4). The concentra- fermentations did not reach dryness by day 14 in 2011, and tion of yeast, including S. cerevisiae, was slightly higher in the ethanol content was lower than in the 2010 fermentation, BioMed Research International Figure 5: Spontaneous fermentation profile in Qvevris during the 2010 and 2011 harvests: (a) Ermitage; (b) Resi.
Table 6: Acetic acid production in the spontaneous fermentations.
Fermentation vessel Acetic acid (mg L−1) concomitant with the production of significant amounts of yeasts that were present in the winery environment and in acetic acid (Table 6). No measurements were taken beyond wine musts undergoing spontaneous fermentation in steel day 14 because the Ermitage and Resi wines were blended tanks, thus favoring the detection and proliferation of some at this stage. The Resi fermentations become more rapidly in yeast species over others. Rose Bengal Chloramphenicol Agar 2011 than in 2010, beginning on the same day (or shortly after) (RBCA) medium was chosen instead of Sabouraud medium the Qvevris were filled. However the fermentation process because the latter favored mold growth over yeasts (data not reached dryness in both years. The Ermitage must took longer shown). Freezing the samples prior to analysis may have to begin fermentation than Resi, starting 2 and 10 days later reduced the viability of the yeast although it is thought that in 2010 and 2011, respectively.
this is a minor effect [26, 27]. Therefore, we acknowledgethat yeast species present in low numbers are unlikely to be detected using this method, whereas abundant species aremore likely to be recognized. Thus, only seven yeast species 4.1. Isolation and Identification of Predominant Yeast Species. were isolated from the grape surface (A. pullulans, C. magnus, The isolation media we used enabled us to select different F. floriforme, R. mucilaginosa, W. saturnus, and Z. florentinus), BioMed Research International Table 7: Yeasts identified by qPCR during spontaneous fermentations in Qvevris during the 2010 and 2011 harvest seasons.
Pichia kluyveri Pichia angusta Pichia anomala a: detected at the beginning of the fermentation; b: detected during log phase; c: detected during stationary phase.
with A. pullulans and R. mucilaginosa previously reported the yeast populations through the different stages of fermen- as colonizers of the grape surface [28]. A further 11 yeast tation in steel tanks also differed among grape varieties. The species were isolated from vineyard air samples, all of detection of some yeast species only during the later stages which had previously been detected in winery environmental of fermentation probably reflects their proliferation to cell samples [28, 29]. The anamorphic yeast Kloeckera apiculata, numbers above the detection threshold of our assay rather previously reported as the predominant yeast species on the than their genuine absence at the beginning of fermentation.
grape surface and in air samples [28], was not found in our The relative greater diversity of yeast species in red compared investigation but instead the teleomorphic species H. uvarum to white wines is consistent with the higher pH of red wines, was found in our environmental samples.
providing favorable conditions for yeast growth [34]. In white We found that differences in yeast diversity were often wines, yeasts isolated from the grape skin were not found in dependent on the grape variety. This phenomenon can be the must, probably because they remained in the skin fraction attributed to several factors, including the different stages during clarification, and this may also have contributed to the of berry ripening at harvest, physical damage to the grape lower species diversity we observed.
surface, and pest management practices [29]. Although we The higher yeast diversity during the early stages of studied different grape varieties grown in the same area and fermentation predominantly reflects the low ethanol toler- processed at the same winery, microclimatic conditions and ance of non-Saccharomyces species [3, 9, 10, 17, 35, 36].
viticultural practices may have influenced the yeast diversity Nevertheless, we found that non-Saccharomyces yeasts such we detected.
as P. klyveri, P. membranifaciens, R. mucilaginosa, and M. Most of the yeasts isolated from the vineyard air were also pulcherima were active in the late fermentation stages in present in the grape juice at the beginning of fermentation.
some must varieties. This is consistent with previous reports All the yeasts identified in the cellar were also found later of ethanol tolerance in M. pulcherima [10, 35, 37], but R. in the fermenting must. R. mucilaginosa was found in air mucilaginosa is usually found during the early stages of samples from both the vineyard and the cellar, and on the fermentation, and its presence along with the Pichia species grape surface, but not on the tank surface. During 2008, Z. later in fermentation could add complexity but also reduce florentinus was the only species found in all environmental the wine quality [34, 38].
samples (air and contact samples, from both the vineyard and Considering the results from the 2008 and 2009 harvests the cellar).
together, we observed that the generally higher yeast diver- The viable counts of the environmental samples showed sity in the must at the beginning of the fermentation was the presence of only non-Saccharomyces species. Although coincident with the rapid onset of the exponential phase.
S. cerevisiae and related species such as S. bayanus are We evaluated the interrelation between the yeast species predominantly responsible for fermentation, they represent and the success of fermentation. We found that despite the only a small fraction of the diversity we identified, which diversity of yeasts in red and white varieties, white musts is consistent with other reports showing that S. cerevisiae is generally contained higher residual sugar levels than red rarely isolated from natural sources such as berry and leaf musts and that sluggish fermentation was more likely. Such surfaces when using viable count methods [30–33]. The small fermentations were characterized by the initial predominance number of species isolated from the cellar environment (air of C. zemplinina and S. bayanus, as well as lower levels of and tank surface) during 2009 compared to 2008 may have M. pulcherima and S. cerevisiae, contrasting with the red been caused by the sanitary conditions adopted by the winery wine musts. The impact of these properties on fermentation after the sampling results in 2008. The dynamic behavior of reflects the better performance of S. cerevisiae compared with BioMed Research International the lower fructose uptake capacity of S. bayanus [39], which further experiments to determine the influence of Qvevris on is consistent with our results.
spontaneous fermentation. Comparative studies with steel-tank fermentations, using the same raw materials (grape 4.2. Dynamic Behavior of Wild Yeasts during Spontaneous variety and harvest year), should be carried out to investigate Fermentation in Qvevris. We developed a novel qPCR the impact of Qvevris in more detail.
method for the rapid, sensitive, and culture-independentdetection of yeast species throughout fermentation, revealing that the non-Saccharomyces yeast R. mucilaginosa and P.
anomala dominated the final stages of spontaneous fermen- The predominant yeasts found in the winery (i.e., tation in Qvevris. These results are important because non- Saccharomyces yeasts can influence the flavor and quality of and some Pichia species) were used as a basis for the wine in both positive and negative ways [40–42] despite their development of an antibody chip for the identification and metabolic activity and abundance [19, 20, 43].
semiquantitative detection of wild yeast. The effect of the The diversity of the yeast species was variety dependent initial yeast concentration and the berry/must temperature and vintage dependent, with C. zemplinina and P. angusta on the length of the fermentation lag phase and thus the present only in the variety of Resi, and P. kluyveri present quality of spontaneous fermentation will be investigated in both wines but only during the 2010 harvest. H. uvarum in more detail to improve the performance of this device.
has previously been identified as the predominant species We have also provided the first quantitative evidence during the early stages of fermentation [9–11, 35] but we describing the dynamic behavior of yeast populations during found no evidence for this species on the grape surface spontaneous fermentation in amphora vessels.
(viable cell count method) and found it was less prevalentduring amphora fermentations (qPCR method). In con- Conflict of Interests
trast, R. mucilaginosa was found to be abundant in boththe amphora and steel-tank fermentations using qPCR and The authors declare that there is no conflict of interests.
culture-dependent methods, respectively.
The Resi fermentations commenced almost immediately in 2011, even though similar numbers of yeast cells werepresent at the beginning of fermentation in both years, The authors thank Albert Mathier et Fils winery, especially and S. cerevisiae was less abundant in 2011 than 2010. The the owner (Am´ed´ee Mathier) and oenologist (Fadri Kuonen) minimal lag phase and rapid fermentation (completed in 3 for kindly providing the samples for testing. This work was days) could be explained by the climatic conditions in the funded by the Fraunhofer Institute.
weeks prior to harvest, which increased the temperature ofthe berries and the must after crushing (data not shown),favoring the rapid proliferation of S. cerevisiae. This sug- gests that berry temperature before pressing could play a [1] M. R. Provenzano, H. El Bilali, V. Simeone, N. Baser, D.
key role in the success of spontaneous fermentation. We Mondelli, and G. Cesari, "Copper contents in grapes and wines tested this hypothesis by studying parallel fermentations.
from a Mediterranean organic vineyard," Food Chemistry, vol.
Ermitage fermentations underwent a longer lag phase in 122, no. 4, pp. 1338–1343, 2010.
2011 (11 days) than 2010 (5 days, reaching dryness by day [2] W. Kaltzin, "Natural wines, als Trend," 2012, http://www.der- 14). During 2011, R. mucilaginosa and P. anomala were the predominant species throughout fermentation, and these [3] G. H. Fleet, "Yeast interactions and wine flavour," International are considered spoilage yeasts [44]. Several previous studies Journal of Food Microbiology, vol. 86, no. 1-2, pp. 11–22, 2003.
have shown that longer lag phases provide an opportunityfor non-Saccharomyces yeasts and other microorganisms to [4] A. Cuadros-Inostroza, P. Giavalisco, J. Hummel, A. Eckardt, L. Willmitzer, and H. Pe˜na-Cort´es, "Discrimination of wine outcompete beneficial microbes and produce toxic and/or attributes by metabolome analysis," Analytical Chemistry, vol.
noxious compounds, causing spoilage [45–49]. Accordingly, 82, no. 9, pp. 3573–3580, 2010.
we found that 0.65 g L−1 acetic acid was produced in this [5] J. Marais, "Effect of different wine-making techniques on the fermentation, which is above the upper range in normal wines composition and quality of pinotage wine. I. Low-temperature and is considered undesirable [8, 50, 51].
skin contact prior to fermentation," South African Journal of Despite the presence of spoilage yeast, the success of Enology and Viticulture, vol. 24, no. 2, pp. 70–75, 2003.
spontaneous fermentation seems to correlate with the length [6] P. Hernandez-Orte, M. Cersosimo, N. Loscos, J. Cacho, E.
of the lag phase, since fermentations with a longer lag Garcia-Moruno, and V. Ferreira, "Aroma development from phase were more likely to fail. The onset of fermentation non-floral grape precursors by wine lactic acid bacteria," Food also depended on the temperature of the must, so this is Research International, vol. 42, no. 7, pp. 773–781, 2009.
a key factor to consider when predicting the outcome of a [7] P. Romano, C. Fiore, M. Paraggio, M. Caruso, and A. Capece, "Function of yeast species and strains in wine flavour," Interna- Our integration of novel analytical methods with tra- tional Journal of Food Microbiology, vol. 86, no. 1-2, pp. 169–180, ditional winemaking using Qvevris provides the basis for BioMed Research International [8] M. G. Lambrechts and I. S. Pretorius, "Yeast and its importance medically important yeasts," Journal of Clinical Microbiology, to wine aroma—a review," South African Journal of Enology and vol. 39, no. 11, pp. 4042–4051, 2001.
Viticulture, vol. 21, special issue, pp. 97–129, 2000.
[24] T. White, T. Bruns, S. Lee, and J. Taylor, "Amplification and [9] M. Combina, A. El´ıa, L. Mercado, C. Catania, A. Ganga, and C.
direct sequencing of fungal ribosomal RNA genes for phyloge- Martinez, "Dynamics of indigenous yeast populations during netics," in PCR Protocols: A Guide to Methods and Applications, spontaneous fermentation of wines from Mendoza, Argentina," M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White, Eds., International Journal of Food Microbiology, vol. 99, no. 3, pp.
pp. 315–322, Academic Press, 1990.
237–243, 2005.
[25] M. A. Lodhi, G. N. Ye, N. F. Weeden, and B. I. Reisch, "A [10] E. Di Maro, D. Ercolini, and S. Coppola, "Yeast dynamics during simple and efficient method for DNA extraction from grapevine spontaneous wine fermentation of the Catalanesca grape," cultivars and Vitis species," Plant Molecular Biology Reporter, International Journal of Food Microbiology, vol. 117, no. 2, pp.
vol. 12, no. 1, pp. 6–13, 1994.
201–210, 2007.
[26] S. Pardo, M. A. Galvagno, and P. Cerrutti, "Studies of via- [11] A. Rementeria, J. A. Rodriguez, A. Cadaval et al., "Yeast bility and vitality after freezing of the probiotic yeast Sac- associated with spontaneous fermentations of white wines from charomyces boulardii: physiological preconditioning effect," the "Txakoli de Bizkaia" region (Basque Country, North Spain)," Revista Iberoamericana de Micologia, vol. 26, no. 2, pp. 155–160, International Journal of Food Microbiology, vol. 86, no. 1-2, pp.
201–207, 2003.
[27] C. Alves-Ara´ujo, M. J. Almeida, M. J. Sousa, and C. Le˜ao, [12] G. H. Fleet, Wine Microbiology and Biotechnology, Taylor & "Freeze tolerance of the yeast Torulaspora delbrueckii: cellular Francis, 1993.
and biochemical basis," FEMS Microbiology Letters, vol. 240, no.
[13] E. H. Hansen, P. Nissen, P. Sommer, J. C. Nielsen, and N.
1, pp. 7–14, 2004.
Arneborg, "The effect of oxygen on the survival of non-Saccharomyces yeasts during mixed culture fermentations of [28] A. Martini, "Origin and domestication of the wine yeast grape juice with Saccharomyces cerevisiae," Journal of Applied Saccharomyces cerevisiae," Journal of Wine Research, vol. 4, no.
Microbiology, vol. 91, no. 3, pp. 541–547, 2001.
3, pp. 165–176, 1993.
[14] R. B. Boulton, V. L. Singleton, L. F. Bisson, and R. E. Kunkee, [29] P. Raspor, D. M. Milek, J. Polanc, S. Smole Moˇzina, and N.
Principles and Practices of Winemaking, Springer, 1998.
Cadeˇz, "Yeasts isolated from three varieties of grapes cultivated [15] N. P. Jolly, O. P. H. Augustyn, and I. S. Pretorius, "The role and in different locations of the Dolenjska vine-growing region, use of non-saccharomyces yeasts in wine production," South Slovenia," International Journal of Food Microbiology, vol. 109, African Journal of Enology and Viticulture, vol. 27, no. 1, pp. 15– no. 1-2, pp. 97–102, 2006.
[30] A. Martini, "Origin and domestication of the wine yeast [16] M. Ciani and F. Maccarelli, "Oenological properties of non- Saccharomyces cerevisiae," Journal of Wine Research, vol. 4, no.
Saccharomyces yeasts associated with wine-making," World 3, pp. 165–176, 1993.
Journal of Microbiology and Biotechnology, vol. 14, no. 2, pp. 199– [31] A. Vaughan-Martini and A. Martini, "Facts, myths and legends on the prime industrial microorganism," Journal of Industrial [17] C. M. Egli, W. D. Edinger, C. M. Mitrakul, and T. Henick-Kling, Microbiology, vol. 14, no. 6, pp. 514–522, 1995.
"Dynamics of indigenous and inoculated yeast populations and [32] M. J. De La Torre, M. C. Millan, P. Perez-Juan, J. Morales, and J.
their effect on the sensory character of Riesling and Chardonnay M. Ortega, "Indigenous yeasts associated with two Vitis vinifera wines," Journal of Applied Microbiology, vol. 85, no. 5, pp. 779– grape varieties cultured in southern Spain," Microbios, vol. 100, no. 395, pp. 27–40, 1999.
[18] A. Soden, I. L. Francis, H. Oakey, and P. A. Henschke, "Effects [33] I. S. Pretorius, "Tailoring wine yeast for the new millennium: of co-fermentation with Candida stellata and Saccharomyces novel approaches to the ancient art of winemaking," Yeast, vol.
cerevisiae on the aroma and composition of Chardonnay wine," 16, no. 8, pp. 675–729, 2000.
Australian Journal of Grape and Wine Research, vol. 6, no. 1, pp.
21–30, 2000.
[34] T. Deak and L. R. Beuchat, "Yeasts associated with fruit juice concentrates," Journal of Food Protection, vol. 56, no. 9, pp. 777– [19] N. Hierro, B. Esteve-Zarzoso, A. Gonz´alez, A. Mas, and J. M.
Guillam´on, "Real-time quantitative PCR (QPCR) and reversetranscription-QPCR for detection and enumeration of total [35] M. J. Torija, N. Roz es, M. Poblet, J. M. Guillam´on, and A. Mas, yeasts in wine," Applied and Environmental Microbiology, vol.
"Yeast population dynamics in spontaneous fermentations: 72, no. 11, pp. 7148–7155, 2006.
comparison between two different wine-producing areas over [20] K. Zott, O. Claisse, P. Lucas, J. Coulon, A. Lonvaud-Funel, and I.
a period of three years," International Journal of General and Masneuf-Pomarede, "Characterization of the yeast ecosystem in Molecular Microbiology, vol. 79, no. 3-4, pp. 345–352, 2001.
grape must and wine using real-time PCR," Food Microbiology, [36] P. Satora and T. Tuszynski, "Biodiversity of yeasts during plum vol. 27, no. 5, pp. 559–567, 2010.
Wegierka Zwykla spontaneous fermentation," Food Technology [21] L. Cocolin and D. Ercolini, Molecular Techniques in the Micro- and Biotechnology, vol. 43, no. 3, pp. 277–282, 2005.
bial Ecology of Fermented Foods, Springer, 2007.
[37] A. Querol, M. Jimenez, and T. Huerta, "Microbiological and [22] B. G. Baldwin, "Phylogenetic utility of the internal transcribed enological parameters during fermentation of musts from poor spacers of nuclear ribosomal DNA in plants: an example from and normal grape-harvests in the region of Alicante (Spain)," the compositae," Molecular Phylogenetics and Evolution, vol. 1, Journal of Food Science, vol. 55, no. 6, pp. 1603–1606, 1990.
no. 1, pp. 3–16, 1992.
[38] V. Loureiro and M. Malfeito-Ferreira, "Spoilage yeasts in the [23] Y. C. Chen, J. D. Eisner, M. M. Kattar et al., "Polymorphic wine industry," International Journal of Food Microbiology, vol.
internal transcribed spacer region 1 DNA sequences identify 86, no. 1-2, pp. 23–50, 2003.
BioMed Research International [39] I. Magyar and T. T´oth, "Comparative evaluation of some oenological properties in wine strains of Candida stellata, Can-dida zemplinina, Saccharomyces uvarum and Saccharomycescerevisiae," Food Microbiology, vol. 28, no. 1, pp. 94–100, 2011.
[40] J. Mora, J. I. Barbas, and A. Mulet, "Growth of yeast species dur- ing the fermentation of musts inoculated with kluyveromycesthermotolerans and saccharomyces cerevisiae," American Jour-nal of Enology and Viticulture, vol. 41, no. 2, pp. 156–159, 1990.
[41] E. Longo, J. B. Vel´azquez, C. Sieiro, J. Cansado, P. Calo, and T.
G. Villa, "Production of higher alcohols, ethyl acetate, acetalde-hyde and other compounds by 14 Saccharomyces cerevisiaewine strains isolated from the same region (Saln´es, N.W. Spain),"World Journal of Microbiology & Biotechnology, vol. 8, no. 5, pp.
539–541, 1992.
[42] C. Lema, C. Garcia-Jares, I. Orriols, and L. Angulo, "Contribu- tion of Saccharomyces and non-Saccharomyces populations tothe production of some components of Albari˜no wine aroma,"American Journal of Enology and Viticulture, vol. 47, no. 2, pp.
206–216, 1996.
[43] I. Andorr a, S. Landi, A. Mas, J. M. Guillam´on, and B. Esteve- Zarzoso, "Effect of oenological practices on microbial popula-tions using culture-independent techniques," Food Microbiol-ogy, vol. 25, no. 7, pp. 849–856, 2008.
[44] J. I. Pitt and A. D. Hocking, Fungi and Food Spoilage, Springer, 3rd edition, 2009.
[45] G. S. Drysdale and G. H. Fleet, "The growth and survival of acetic acid bacteria in wines at different concentrations ofoxygen," American Journal of Enology and Viticulture, vol. 40,no. 2, pp. 99–105, 1989.
[46] L. F. Bisson, "Stuck and sluggish fermentations," American Journal of Enology and Viticulture, vol. 50, no. 1, pp. 107–119,1999.
[47] C. G. Edwards, R. B. Beelman, C. E. Bartley, and A. L.
Mcconnell, "Production of decanoic acid and other volatilecompounds and the growth of yeast and malolactic bacteriaduring vinification," American Journal of Enology and Viticul-ture, vol. 41, no. 1, pp. 48–56, 1990.
[48] C. G. Edwards, K. M. Haag, and M. D. Collins, "Identification and characterization of two lactic acid bacteria associated withsluggish/stuck fermentations," American Journal of Enology andViticulture, vol. 49, no. 4, pp. 445–448, 1998.
[49] M. J. Torija, G. Beltran, M. Novo et al., "Effects of fermentation temperature and Saccharomyces species on the cell fatty acidcomposition and presence of volatile compounds in wine,"International Journal of Food Microbiology, vol. 85, no. 1-2, pp.
127–136, 2003.
[50] M. Vilanova, Z. Genisheva, L. Bescansa, A. Masa, and J. M.
Oliveira, "Volatile composition of wines from cvs. Blancolex´ıtimo, Agudelo and Serradelo (Vitis vinifera) grown inBetanzos (NW Spain)," Journal of the Institute of Brewing, vol.
115, no. 1, pp. 35–40, 2009.
[51] I. Mato, S. Su´arez-Luque, and J. F. Huidobro, "Simple determi- nation of main organic acids in grape juice and wine by usingcapillary zone electrophoresis with direct UV detection," FoodChemistry, vol. 102, no. 1, pp. 104–112, 2007.

Source: http://www.mathier.ch/uploads/media/540465.pdf