Despite its biotechnological interest, hybridization, which can result in hybrid vigor, has not commonly been studied or exploited in the yeast genus. From a diallel design including 55 intra- and interspecific hybrids between Saccharomyces cerevisiae and S. uvarum grown at two temperatures in enological conditions, we analyzed as many as 35 fermentation traits with original statistical and modeling tools. We first showed that, depending on the types of trait – kinetics parameters, life-history traits, enological parameters and aromas –, the sources of variation (strain, temperature and strain * temperature effects) differed in a large extent. Then we compared globally three groups of hybrids and their parents at two growth temperatures: intraspecific hybrids S. cerevisiae * S. cerevisiae, intraspecific hybrids S. uvarum * S. uvarum and interspecific hybrids S. cerevisiae * S. uvarum. We found that hybridization could generate multi-trait phenotypes with improved oenological performances and better homeostasis with respect to temperature. These results could explain why interspecific hybridization is so common in natural and domesticated yeast, and open the way to applications for wine-making.
Citation: da Silva T, Albertin W, Dillmann C, Bely M, la Guerche S, Giraud C, et al. (2015) Hybridization within Saccharomyces Genus Results in Homoeostasis and Phenotypic Novelty in Winemaking Conditions. PLoS ONE 10(5): e0123834. https://doi.org/10.1371/journal.pone.0123834
Academic Editor: Joseph Schacherer, University of Strasbourg, FRANCE
Received: November 24, 2014; Accepted: February 21, 2015; Published: May 6, 2015
Copyright: © 2015 da Silva et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: SARCO and BIOLAFFORT provided support in the form of salaries for authors (PM, SlG), but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.
Competing interests: The authors have the following interests. Staphane la Guerche and Philippe Marullo are employed by SARCO and BIOLAFFORT, respectively. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.
Plant hybrids commonly harbor non-additive inheritance for polygenic traits, with phenotypic values usually different from the mean parental values. These “monsters” produced by hybridization  have evolutionary implications [2, 3] and are extensively exploited for producing improved crops. For instance in maize, F1 hybrids between homozygous lines show heterosis of 100 to 400% for grain yield [4, 5], and many other complex traits, such as height, leaf area, grain size, germination rate, root growth and root nitrogen uptake, also display heterosis [6, 7].
Hybridization affects not only the phenotypic values, but also their stability over environmental changes. Homeostasis, canalization or robustness— the term depends on the biological field [8–11] allows the organisms to buffer the effects of external perturbations through metabolic, physiological or developmental adjustments, and thus to maintain fitness in diverse habitats. Homeostasis is usually higher in intra- or interspecific hybrids than in their parents, as shown for instance for yield, tolerance to soil acidity and to soil moisture stress in maize  or morphometric traits in mice .
Consequences of hybridization have been studied in a large range of wild as well as domesticated species, but have scarcely been studied and exploited in industrial eukaryotic micro-organisms such as yeast. Recently some authors investigated the heterosis phenomenon within natural and domesticated strains of Saccharomyces cerevisiae [14–17], but these studies were mainly focused on cell growth in laboratory conditions.
In the Saccharomyces sensu stricto clade, yeast species showed a severe reproductive isolation (less than 1% of viable spores). However, the prezygotic barrier can be easily bypassed leading to viable interspecific hybrids . Numerous interspecific hybrids between S. cerevisiae and psychrophilic species S. uvarum or S. kudriavzevii have been isolated in wine and natural environment [20–28]. These natural hybrids have technological properties differing from those of their respective “parental” species, with sometimes better robustness [29–32]. Moreover some wine starters empirically selected proved to be interspecific hybrids [25, 33, 34], promoting the idea that interspecific hybridization is a good way for obtaining valuable strains for wine fermentation. However in the previous works the parental strains of the hybrids were not known, so it was not possible to state definitely that interspecific hybridization conferred novel phenotypes and possibly better homeostasis. Moreover, since their genesis, these natural hybrids may have undergone genomic modifications that can drastically affect their phenotype such gross chromosomal rearrangement [35, 36], loss of heterozygosity [23, 37], particular mitotypes , aneuploidies  and introgressions .
In order to assess rigorously the phenotypic impact of intra- and interspecific hybridization, the hybrids must be compared to their parental strains. As previously reviewed [19, 40] various laboratories have produced such hybrids between Saccharomyces species [27, 41–43]. However, only a few interspecific hybrids were compared to their parents, and for quite a small number of traits [44–46]. As the Saccharomyces strains harbor huge genetic and phenotypic diversity [47, 48], the behavior of few hybrids is not sufficient to have an overall view on the effects of hybridization.
In this work we examined the extent to which hybridization within and between Saccharomyces species modified a large series of traits measured during and at the end of fermentation at two temperatures, with particular attention to homeostasis. We focused on S. cerevisiae and S. uvarum (formerly S. bayanus var. uvarum) [28, 49, 50], two related species naturally associated with wine fermentations [51–53]. S. cerevisiae is the main yeast able to achieve grape must fermentation, but S. uvarum can display similar fermentation performance, particularly at low temperature [40, 54–56]. Although these sister species share large synteny [57, 58], they differ for several technological traits. First, S. cerevisiae has a higher resistance to high temperature stress (up to 37°C)  while S. uvarum is more tolerant to low temperatures . Second, S. uvarum exhibits a specific aromatic profile by producing higher amounts of phenyl-2-ethanol and phenyl-2-ethanol acetate than S. cerevisiae strains [31, 41, 44]. Finally, although S. uvarum harbors a high ethanol resistance (up to 15% ), it is less resistant than S. cerevisiae . Several natural hybrids between these two species have been described [26, 30, 61], and the possibility to produce synthetic inter-specific hybrids  established S. cerevisiae and S. uvarum as model systems for hybridization studies.
Measuring an unprecedented number of traits, we investigated the physiological and technological properties of a collection of four S. uvarum and seven S. cerevisiae parental strains and their 55 possible hybrids, namely 27 intraspecific hybrids and 28 inter-specific hybrids, under winemaking conditions at two temperatures. We analyzed the sources of phenotypic variation— genetic and/or environmental —for various categories of traits (fermentation kinetics, life-history, wine composition and organoleptic quality), we compared the intra- and interspecific hybrids and assessed the extent to which hybridization increased homeostasis at a multi-trait level.
Materials and Methods
Parental strains and culture conditions
The starting genetic material of the experimental design were seven S. cerevisiae strains and four S. uvarum strains, associated to various food-processes (enology, brewery, cider fermentation and distillery) or isolated from natural environment (oak exudates) (Table 1). These strains could not be used as such as parents of a diallel design because they were suspected to be heterozygous at many loci. Monosporic clones were isolated by tetrad dissection using a micromanipulator (Singer MSM Manual; Singer Instrument, Somerset, United Kingdom). All original strains but Alcotec 24 were homothallic (HO/HO), therefore fully homozygous diploid strains were spontaneously obtained by fusion of opposite mating type cells. For A24 (ho/ho), one isolated haploid meiospore was diploidized via transient expression of the HO endonuclease . These strains, called W1, D1, D2, E2, E3, E4 and E5 for S. cerevisiae and U1, U2, U3 and U4 for S. uvarum, were used as the parental strains for the construction of a half diallel design (Fig 1).
Fully homozygous diploid strains were used as parental strains. W1, D1, D2, E2, E3, E4 and E5 are S. cerevisiae strains. U1, U2, U3 and U4 are S. uvarum strains. Fermentations were carried out in Sauvignon blanc grape juice and run at two temperatures 18°C and 26°C in triplicate for a total of 396 experiments. Thirty-five traits were collected and grouped into four classes (Fermentation Kinetics Traits, Life History Traits, Basic Enological Parameters and Aromatic Traits). Fruits, plant and sugar images were reprinted from CC BY licenses, the copyright holders being Steve Hopson, 2006 (bananas); Uwe Hermann, 2007 (sugar cubes); Sharon Mollerus, 2007 (strawberries); Sten Porse, 2005 (boxwood); Lebensmittelfotos, 2013 (grapes).
All strains were grown at 24°C in YPD medium containing 1% yeast extract (Difco Laboratories, Detroit, MI), 1% Bacto peptone (Difco), and 6% glucose, supplemented or not with 2% agar. When necessary, antibiotics were added at the following concentrations: 100 μg/mL for G418 (Sigma, l’Isle d’Albeau France) or nourseothricin (Werner Bioagent, Jena, Germany), and 300 μg/mL for hygromycin B (Sigma).
Construction of the diallel design
In order to produce interspecific hybrids, the eleven diploid parental strains were transformed with a cassette containing the HO allele disrupted by a gene of resistance to either G418 (ho::KanR), hygromycin B (ho::HygR) or nourseothricin (ho::NatR) as previously described . After transformation, monosporic clones were isolated, and the mating-type (MATa or MATα) of antibiotic-resistant clones was determined using testers of known mating-type. Strain transformation allowed (i) conversion to heterothallism for the homothallic strains (all but D2, see Table 1) and (ii) antibiotic resistance allowing easy hybrid selection.
For each hybrid construction, parental strains of opposite mating type were put in contact for 2 to 6 hours in YPD medium at room temperature, and then plated on YPD-agar containing the appropriate antibiotics. The 55 possible hybrids from the 11 parental strains, namely 21 S. cerevisiae intraspecific hybrids, 6 S. uvarum intraspecific hybrids and 28 interspecific hybrids, were obtained. For each cross, a few independent colonies were collected. After recurrent cultures on YPD-agar corresponding to ~ 80 generations, the nuclear chromosomal stability of the hybrids was controlled by pulsed field electrophoresis (CHEF-DRIII, Biorad, CA) as well as homoplasmy (only one parental mitochondrial genome) as detailed in Albertin et al. .
Yeast strain characterization
In order to discriminate rapidly the hybrids and parental strains, we used two polymorphic microsatellites specific to S. cerevisiae (Sc-YFR038, Sc-YML091)  and two specific to S. uvarum (locus 4 and 9) . These four markers were amplified in a multiplex PCR reaction with the labeled primers (S1 Table). The PCR was carried out in a final volume of 8 μL using the following program: 95°C for 5 min for initial denaturation step; 95°C for 30 s, 55°C for 90 s and 72°C for 60 s repeated 35 times; a final elongation step of 30 min at 60°C. The PCR products were analyzed on an ABI3730 apparatus (Applied Biosystem, Villebon-sur-Yvette, France) by the genotyping facilities of Bordeaux University. Microsatellite lengths were analyzed using the Peak Scanner tool (Applied Biosystem, Villebon-sur-Yvette, France)
Alcoholic fermentation experiments
The 66 strains (11 parental and 55 hybrids) were grown in three replicates at two temperatures, 26°C and 18°C. The 396 fermentations (66 strains x 2 temperatures x 3 replicates) were performed following a randomized experimental design. The design was implemented considering a block as two sets of 27 fermentations (26 fermentations and a control without yeast to check for contamination), one carried out at 26°C and the other at 18°C. The distribution of the strains within the blocks was randomized to minimize the residual variance of the estimators of the Strain and Temperature effects.
Grape must and fermentation conditions.
White grape must was obtained from Sauvignon blanc grapes, harvested in vineyards in Bordeaux area (2009 vintage). This grape juice was obtained by Ducourt Vignoble 18 Le Hourc, 33760 Ladaux that gives the autorization to use this material for the study. The harvest was realized by the owner of the field and did not involve endangered or protected species.Tartaric acid precipitation was stabilized and turbidity was adjusted to 100 NTU (Nephelometric Turbidity Unit) before storage at – 20°C. Grape juice had a sugar concentration of 189 g.L–1, a pH of 3.3 and an assimilable nitrogen content of 242 mg N.L–1. The indigenous yeast population was estimated by YPD-plate counting after must thawing and was consistently lower than 20 CFU (Colony-Forming Unit) per mL.
Yeast pre-cultures (20 mL) were run in half-diluted must filtered through a 0.45 μm nitrate-cellulose membrane, during 24 h, at 24°C with orbital agitation (150 rpm). Cell concentration was quantified using a flow cytometer (see below) and grape must was inoculated at 106 viable cells per mL. Fermentations were run in 125 mL glass-reactors, locked to maintain anaerobiosis, with permanent stirring (300 rpm) at 18°C or 26°C. Yeast strain implantation in grape must was checked when the stationary phase was reached (40% of alcoholic fermentation). The DNA of fermenting yeast was extracted using FTA clone saver cards (Whatman, France), and strain identity was controlled by microsatellite analysis.
Flow cytometry analysis.
The monitoring of population growth, cell size and viability was performed using a cytometer FC500 MPL (Beckman Coulter, Villepinte, France). Collected samples were filtered before flow cytometry analysis with 10 μm disposable filters, CellTric (Partec, Münster, Germany). Samples were diluted with McIlvaine buffer pH 4 (0.1 M citric acid, 0.2 M sodium phosphate dibasic) added with propidium iodide (0.3% v/v) in order to stain dead cells, and dilution was adapted to reach a flow rate lower than 2500 particules/sec. Fluorescent beads (Cell Counter, Beckman Coulter) were used to normalize the quantification of cellular concentration.
Multi-trait phenotyping in winemaking conditions
For each alcoholic fermentation, four sets of experimental data were obtained: fermentation kinetics parameters (FK), life-history traits (LHT), basic enological parameters (BEP) and aromatic traits (AT).
Fermentation kinetics parameters (FK).
The amount of CO2 released was monitored daily by the weight loss of the bioreactors.
The amount of CO2 released (Yit) for the fermentation i at time t was modeled by a Weibull function f as described in S1 Supporting Information. Seven kinetics parameters were computed from the model (S1 Fig):
- - t-lag (h) = , the fermentation lag-phase that is the time between inoculation and the beginning of CO2 release (when the CO2 production rate was higher than 0.05 g.L–1.h–1);
- - t-Vmax (h), the time to reach the inflexion point, out of the fermentation lag phase;
- - t-45 (h), the fermentation time at which 45 g.L–1 of CO2 was released, out of the fermentation lag phase;
- - t-75 (h), the fermentation time at which 75 g.L–1 of CO2 was released, out of the fermentation lag phase;
- - AFtime (Alcoholic Fermentation time, h), the time between t-lag and the time at which the CO2 emission rate became less than, or equal to, 0.05 g.L–1.h–1;
- - Vmax (g.L–1.h–1), the value of the first derivative of the Weibull function f, at t-Vmax (h), (f’ = αdb * (t-Vmax —t0)b–1 * exp [– α (t-Vmax—t0)b]), and corresponded to the maximum CO2 released rate;
- - CO2max (g.L–1) = the total amount of CO2 released at the end of the fermentation.
Life-history traits (LHT).
During the alcoholic fermentation, cell samples were taken and analyzed as described in the “flow cytometry analysis” session. The experimental measurement of the logarithm of cell concentration was modeled by a discontinuous function of time as described in S2 Supporting Information. The cell size and viability were modeled using a linear model (Wit) as described in S3 Supporting Information.
These models allowed computing eight life-history traits (S1 Fig).
- - t-N0 (h) = , the growth lag-phase (time between inoculation and the beginning of population growth);
- - t-Nmax (h) = the time at which the carrying capacity K was reached;
- - r (logarithm of the number of cell divisions per mL per hour), the intrinsic growth rate;
- - K (log[cells/mL]), was the carrying capacity computed as: K = I + r (t-Nmax – t-lag) + C.
- - Jmax (g.h–1.10–8 cell–1) was the maximum value of the estimated CO2 production rate divided by the estimated cell concentration.
- - Size-t-Nmax (μm), the average cell size at t-Nmax
- - Viability.t-Nmax (%), the percentage of living cells at t-Nmax
- - Viability.t-75 (%), the percentage of living cells at t-75.
Basic Enological Parameters (BEP).
At the end of the fermentation, six Basic Enological Parameters (BEP) were quantified: Residual Sugar (g.L–1), Ethanol (%vol), Sugar/Ethanol Yield (g.L–1.%vol–1) (ratio between the amount of metabolized sugar and the amount of released ethanol), Acetic acid (g.L–1 H2SO4), Total SO2 and Free SO2 (mg.L–1). Residual Sugar and Ethanol were measured by infrared reflectance using an Infra-Analyzer 450 (Technicon, Plaisir, France). For some strains, Residual Sugar was below the threshold of detection. In these cases, instead of inferring the value “0”, which is not biologically realistic, we used the value: (x/1.05) + y, where x is the lowest value measured in the whole data and y is a value drawn in a uniform distribution ~ U(0, 0.001). Acetic acid was quantified by colorimetry (A460) in continuous flux (Sanimat, Montauban, France). Total SO2 and Free SO2 were assayed by Pararosaniline titration .
Aromatic Traits (AT).
The aromatic profile of fermenting yeast was estimated by quantifying 14 aromatic traits (AT). The main volatile compounds were measured at the end of the alcoholic fermentation by GC-MS. For esters, higher alcohols and volatile acids, HSSE HeadSpace Sorptive Extraction followed by GC-MS analysis was used according to Weldegergis et al. . For volatile thiols, a specific extraction was performed according to Tominaga et al.. These analytical methods allowed us to detect up to 22 compounds in the analyzed samples (S2 Table). However only 13 of them were quantified in a sufficient number of samples and were retained after statistical analysis. These compounds were: two higher alcohols (Phenyl-2-ethanol and Hexanol, mg.L–1), seven esters (Phenyl-2-ethanol acetate, Isoamyl acetate, Ethyl-propanoate, Ethyl-butanoate, Ethyl-hexanoate, Ethyl-octanoate and Ethyl-decanoate, mg.L–1), three medium chain fatty acids (Hexanoic acid, Octanoic acid and Decanoic acid, mg.L–1) and one volatile thiol (4-methyl-4-mercaptopentan-2-one, or 4MMP, ng.L–1). For Ethyl-decanoate and Ethyl-octanoate, which were sometimes below the threshold of detection, we proceeded as described above for Residual Sugar. Finally the Acetate ratio, ratio between Phenyl-2-ethanol acetate and Phenyl-2-ethanol, was computed. This parameter represents the acetylation ratio of higher alcohols.
For each of the 35 traits collected, the effects of the strain, of the temperature and of the strain-by-temperature interaction, as well as the random block effect, were estimated through the following mixed model of analysis of variance (R program, lme4 package): where Yijk was the value of the trait for strain i (i = 1, …, 66) at temperature j (j = 1, 2) obtained the week k (k = 1, …, 22), m was the overall mean, si was the fixed strain effect, tempj was the fixed temperature effect, (s*temp)ij was the interaction effect between temperature and strain, Wk was the random block effect and Eijk the residual error. For each trait, normality of residual distributions and the homogeneity of the variances were checked. Some of them displayed heteroscedasticity, which decreased the power of the ANOVA. This was due to strains with weak fermentation abilities (t-lag > 40 h, t-Vmax > 20 h, CO2max < 88 g.L-1 and/or t-75 > 110 h). The predicted means were computed from the model's parameters, as well as their standard deviations. For many traits a significant block effect was found (α = 0.05). Therefore, the decomposition of the total phenotypic variance of each trait into its genetic and environmental components was computed after correction for the random block effects. Multiple non parametric comparisons (Campbell and Skillings analysis) were carried out using nparcomp package of the R program with adjusted p-values .
Large-scale phenotyping of a half yeast diallel under winemaking conditions
Eleven parental strains (seven strains of S. cerevisiae and four strains of S. uvarum, Table 1) and their 55 intra- and inter-specific hybrids were phenotyped under enological conditions, at two temperatures (26°C, favorable for S. cerevisiae, and 18°C, favorable for S. uvarum), in three replicates (Fig 1 and S1 Fig). A total of 396 alcoholic fermentations were performed (66 strains * 2 temperatures * 3 replicates). A few fermentations (26) were discarded due to incomplete or absent implantation of the expected strain. The alcoholic fermentations were characterized in depth through 35 phenotypic traits, leading to almost 13 000 numerical data (35 * 370) for all the fermentations. The traits were classified into four categories: (i) Seven fermentation kinetics (FK) parameters: t-lag, t-Vmax, t-45t-45, t-75, AFtime, Vmax, CO2max; (ii) Eight life-history traits (LHT): t-N0, t-Nmax, r, K, Jmax (growth traits), Size t-Nmax, Viability t-Nmax, Viability t-75 (size and viability traits); (iii) Six basic enological parameters (BEP): Residual Sugar, Ethanol, Sugar/Ethanol Yield, Acetic acid, Total SO2 and Free SO2; (iv) Fourteen aromatic traits (AT): Phenyl-2-ethanol, Hexanol, Phenyl-2-ethanol acetate, Isoamyl acetate, Ethyl-propanoate, Ethyl-butanoate, Ethyl-hexanoate, Ethyl-octanoate, Ethyl-decanoate, Hexanoic acid, Octanoic acid, Decanoic acid, 4MMP (4-methyl-4-mercaptopentan-2-one) and Acetate ratio, the acetylation rate of higher alcohols.
The sources of phenotypic variation differ according to trait categories
The sources of variation of each phenotypic trait were studied by analyses of variance (ANOVA) to estimate the Strain, Temperature, and Strain*Temperature interaction effects (Table 2). The part of phenotypic variation explained by the model (block effect removed) depended on the trait category, with Fermentation Kinetics parameters (FK) showing the highest R2 values (0.60 to 0.92) and Aromatic Traits (AT) the smallest (0.09 to 0.66). All the traits but three (Isoamyl acetate, Ethyl-butanoate and Ethyl-octanoate) displayed a significant Strain effect, accounting for 11 to 67% of the variance explained (p-value < 0.05). The temperature had contrasted effects according to the trait category: the ten traits for which temperature explained at least 10% of the model variance were mainly found in the Fermentation Kinetics (FK) and Life-history Traits (LHT) categories, with R2 values up to 79%: t-45, AFtime, Vmax, t-75 and t-lag (FK), r, Jmax and t-Nmax (LHT), Acetic acid (BEP), and Hexanol (AT). Finally highly significant Strain*Temperature interactions were found for CO2max, t-lag and t-Vmax (FK), t-N0, K, Size.t-Nmax and Viability.t-75 (LHT), Ethanol, Residual Sugar and Sugar/Ethanol Yield (BEP) and Acetate ratio (AT).
Overall, FK and LHT traits displayed Strain effects and large Temperature effects, and in a lesser extent Strain*temperature interactions (except for CO2max with R2 = 0.32), BEP traits had both Strain and Strain*temperature effects with almost no effect of temperature, and finally AT traits had almost exclusively Strain effects.
Strain types differ for many traits of biotechnological interest
For each trait*temperature combination, we compared the means of the S. cerevisiae strains (parents and intraspecific hybrids), of the S. uvarum strains (parents and intraspecific hybrids) and of the interspecific hybrids, using non-parametric comparison tests (α = 0.05). In 42 cases out of 70 (2 temperatures x 35 traits), at least one mean was significantly different from the others (S2 and S3 Figs). For 12 traits, a difference was observed at both temperatures, for 5 traits at 18°C only and for 12 traits at 26°C only. For some traits of biotechnological interest the three strain types were well separated. As shown Fig 2 (panels A and B), interspecific hybrids had a production of Phenyl-2-ethanol and Phenyl-2-ethanol acetate that was roughly intermediate between the ones of the parental species. It is well documented that these compounds discriminate S. cerevisiae and S. uvarum during wine fermentation. Moreover there was a significant Species*Temperature interaction for these compounds (2% and 6% of variance explained for Phenyl-2-ethanol and Phenyl-2-ethanol acetate, respectively). Their concentration was significantly lower at 26°C than at 18°C in the S. uvarum group but this was not the case neither in the S. cerevisiae group nor in the interspecific hybrids. As a consequence the interspecific hybrids are intermediate between parental species at 18°C and close to the S. uvarum group at 26°C. Another striking difference between groups was the yield of alcoholic fermentation, a key parameter in winemaking industry because strains with high Sugar/Ethanol Yield are required to reduce ethanol content in wine. At 18°C, the S. uvarum group and the interspecific hybrids required respectively 0.56 and 0.35 g/L more sugar than S. cerevisiae group for producing 1% vol. of ethanol (Fig 2C). This species discrepancy was highly significant and showed a slight Species*Temperature interaction with a reduced difference between species at 26°C as compared to 18°C. The production levels of several ethyl-esters, was higher in interspecific hybrids than in either parental species at both temperatures. This global heterosis effect was illustrated by summing the concentrations of all ethyl-esters (Ethyl-propanoate, Ethyl-butanoate, Ethyl-hexanoate, Ethyl-octanoate and Ethyl-decanoate) (Fig 2D). For other traits differences between strain types were found (S2 Fig and S3 Fig). At 26°C the interspecific hybrids produced less acetic acid than the parental species. Finally the production of 4MMP was significantly lower in the S. cerevisiae group than in the two other groups. This analysis revealed the existence of large differences between the strain groups analyzed. In addition, the numerous trait*temperature interactions (Table 2) also generated a large phenotypic diversity that may be of interest from a biotechnological viewpoint.
Phenyl-2-ethanol acetate (A1), Phenyl-2-ethanol (A2), Sugar/Ethanol Yield (B) and sum of ethyl esters (C) concentrations in S. cerevisiae (SC), S. uvarum (SU) and interspecific hybrid (H) strains at 18°C and 26°C. Statistical differences between the species groups were tested for each temperature using a multiple non-parametric test with corrected p values (Holm test). Different letters indicate groups showing significant differences (p < 0.01). Capital and lower cases were used for 18° and 26°C, respectively.
Intra and Interspecific hybridizations reshape multi-trait phenotypes and improve homeostasis with respect to temperature
In order to capture the consequence of hybridization at the multi-trait phenotypic level, a Principal Component Analysis (PCA) from the whole data set was carried out. The first PCA axis (PCA1) accounted for 20% of the total variance and clearly separated the strains according to the fermentation temperature (Fig 3A). As expected, the first axis was mainly explained by traits showing a large temperature effect in the ANOVA (p-value < 0.0001) (Fig 3C). All the FK time traits (t-lag, t-75, t-45, t-Vmax, t-Nmax) had low values at 26°C, which reduced the alcoholic fermentation time (AFtime). These traits were strongly correlated with each other (S4 Fig), which explained the major temperature effect seen on PCA1. The first axis was also clearly explained by Vmax and r, two traits with highly significant temperature effects, and in a lesser extent by K, Jmax, Acetic acid, Hexanol, t-N0, Total SO2 and three ethyl esters. The second and the third axes of the PCA (Fig 3B) accounted for 13% and 12% of the total inertia, respectively, and clearly separated the data according to the strain type (S. cerevisiae, S. uvarum and interspecific hybrids). Volatile compounds such as Phenyl-2-ethanol and Phenyl-2-ethanol acetate, as well as most of ethyl esters (AT traits), largely contributed to discriminate the S. uvarum and S. cerevisiae groups, which is consistent with their large R2 for Strain effects. Life-history traits such as cell size and viability also contributed to separate the species. S. uvarum strains had smaller cell size and a lower viability than S. cerevisiae strains. Note the negative correlation between carrying capacity K and cell size (S4 Fig), previously reported in various studies [65, 76], which were here confirmed in a multi-species context. Finally, three basic enological parameters, Ethanol, Residual Sugar and Sugar/Ethanol Yield, and one FK trait, CO2max, were also highly correlated to PCA2 or PCA3. Thus both ANOVA and PCA showed that genetic and environmental variations did not affect in the same way the different trait categories. Temperature strongly influenced fermentation kinetics and life-history traits, while fermentation byproducts (AT and BEP) were mainly influenced by strain origin.
Each point represents one of the 55 hybrid strains: S. cerevisiae intraspecific hybrids S. uvarum intraspecific hybrids and interspecific hybrid at 18°C and 26°C. A: axes 1 and 2 (33% of the total inertia). B: axes 2 and 3 (25% of the total inertia). C: correlation of the variables to discriminant axes PCA1, PCA2 and PCA3. Only variables showing a significant correlation (p-value < 0.0001) are shown. The four-color palette corresponds to the four variable categories (FK: Fermentation Kinetics, LHT: Life-history Trait, BEP: Basic Enological Parameters, AT: Aromatic Traits).
To assess the relative position of the hybrids and their parents we performed another PCA including in silico hybrids. The phenotypic values of in silico intraspecific and interspecific hybrids were computed assuming additivity (i.e. mid-parent value) for all 35 traits. The first discriminating axes accounted for 33% of variance (Fig 4A). As expected, whatever the temperature, the intraspecific in silico hybrids perfectly overlapped the groups of their respective parental strains, and the interspecific in silico hybrids were intermediary between S. cerevisiae and S. uvarum groups. Interestingly, in vivo and in silico hybrids usually did no overlap very well, meaning that both intra- and interspecific hybridizations created original multi-trait phenotypes that were not mid-way between their parents. The distance between in silico and in vivo hybrids depended on the hybrid type and on temperature. For example, in vivo S. uvarum intraspecific hybrids were rather far from their mid-value expectation at both temperatures, as it was the case for interspecific hybrids at 26°C but not at 18°C.
A. First two dimensions of the PCA, explaining 33% of total variation. The phenotypic values of in silico intraspecific and interspecific hybrids were computed assuming additivity (i.e. mid-parent value) for all 35 traits. B. Multi trait phenotype distance between the two temperatures was measured for each hybrid type. Bar plot represents the mean and the standard error for the six groups of hybrids. Differences between in silico and in vivo hybrids were tested using a non-parametric (Wilcoxon test).
For each hybrid type, we computed the multi-trait distance between 18°C and 26°C (Fig 4B). The smaller the distance, the higher the homeostasis with respect to temperature. The S. cerevisiae hybrids globally showed a higher homeostasis than the others two groups. Moreover homeostasis is significantly higher for in vivo hybrids than for in silico hybrids for both S. uvarum and interspecific hybrids. The homeostasis of interspecific hybrids came mainly from strain x temperature interactions for various traits of biotechnological interest, which display contrasted average values between S. uvarum hybrids and S. cerevisiae hybrids (Fig 5). As a result the average values of the interspecific hybrids at 18°C and 26°C were intermediary and close to each other (e.g. Phenyl-2-ethanol, Ethanol, Sugar/Ethanol Yield). For other traits, such as Residual sugar, CO2max and sum of ethyl esters, the strict homeostasis observed is not, or not only, due to strain x temperature interactions (Fig 5).
A. Fermentation kinetics traits displaying large Temperature effects and moderate Strain effects. B and C. Traits displaying Strain effects and Strain x Temperature interactions. Traits in A, B and C are the mainly correlated to axes 1, 2 and 3 of the PCA, respectively. The colors and symbol are the same than previous figures.
All these findings indicate that both intra and interspecific hybridization result in multi-trait phenotypes that are hardly predictable from the parental strains and that display, at least for some traits, more homeostasis than expected under the additivity hypothesis.
In this study, a diallel design of 55 newly synthetized hybrids was obtained from 11 parental strains belonging to the two main species involved in grape juice fermentation, S. cerevisiae and S. uvarum. This kind of genetic design has been widely used in plant and animal breeding to analyze the genetic bases of complex traits and identifying heterotic groups [4, 77]. In yeast, a diallel design has been recently developed by different authors using a collection of S. cerevisiae yeast strains [14–17]. Our design included for the first time interspecific hybrids, allowing us to investigate possible synergies between the genomes of S. cerevisiae and S. uvarum. From 370 controlled fermentation experiments at two temperatures in a natural grape juice (Sauvignon blanc), we measured or estimated through mathematical models various fermentation kinetics parameters, life-history traits and a series of metabolites including wine aromatic compounds such as esters and volatile thiols, resulting in about 13 000 data points for 35 phenotypic traits. Global characteristics of intra- and interspecific hybrids were described, focusing in particular on their possible biotechnological interest.
Interspecific hybridization between S. cerevisiae and S. uvarum strains provides yeasts with new and suitable traits for winemaking
Multivariate analysis clearly showed that interspecific hybrids can be separated from S. cerevisiae and S. uvarum strains mostly by aromatic traits and other parameters crucial for enology (Fig 3B and 3C). Hybridization between S. cerevisiae and S. uvarum strongly reshapes the production of several secondary metabolites in interspecific hybrids (S2 Fig and S3 Fig). This finding was previously reported for glycerol , acetic acid [41, 45], volatile thiols (4MMP)  and higher alcohols such as Phenyl-2-ethanol [31, 41, 44]. Except for glycerol that was not assayed here, these observations were confirmed for a large set of hybrids. At 26°C, the interspecific hybrids produced less acetic acid than the parental species, which can be useful for wine yeast selection . Interestingly, at the same temperature, the 4MMP production was three fold higher in interspecific hybrids and S. uvarum group than in S. cerevisiae group. This can be explained by the inheritance of the dominant Irc7p allele of S. uvarum encoding for a fully active a cystathionin β-lyase able to cleave efficiently the cysteinylated precursor of this compound [79, 80]. The production of Phenyl-2-ethanol and its acetate in the interspecific hybrids confirmed to be intermediate between the parental species . The high level of these molecules is a major feature of S. uvarum species and could be due to the more active shikimate and phenylalanine pathways found in this species [54, 81]. Interestingly interspecific hybrids produced lower amount of these compounds than S. uvarum. In wine these compounds can mask more subtle fragrances , so their moderate production during alcoholic fermentation is desired.
Beside these already described features, our data provide new interesting findings. First, interspecific hybrids display a much higher production of ethyl esters (2.45 folds) than parental species at both temperatures. These compounds positively impact wine quality by conferring fruity notes . The production of ethyl esters can be related to two factors: (i) the availability of short and medium fatty acid chains which depends on the must composition, the fermentation temperature [83, 84] and the species . S. uvarum and S. cerevisiae show large differences in lipid metabolism  with a higher level of Medium Chain Fatty Acids (MCFA) production in S. uvarum , likely due to a more active fatty acid pathway ; (ii) the esterification of these fatty acids with ethanol that is mediated by specific ethanol-O-Acyl transferases (EOAT) (Eeb1p, Eht1p, Ymr210p) [85, 86]. Recent gene expression surveys demonstrated that allele Eeb1p of the major EOAT was much more expressed in S. cerevisiae than in S. eubayanus , or in S uvarum and S kudriavzevii . This could indirectly suggest that S. cerevisiae might have a higher EOAT than S. uvarum, a species closely related to S. eubayanus . The high production of ethyl esters in interspecific hybrids could therefore result from the combination of these two factors. A second interesting result was the higher Sugar/Ethanol Yield found in both interspecific hybrids and S. uvarum strains as compared to S. cerevisiae strains. To date the natural intraspecific variation among S. cerevisiae strains was very low for this trait [47, 65, 66]. Due to the continuous increasing level of ethanol in wines, the Sugar/Ethanol Yield is becoming an important trait for wine yeast selection [90–92]. Recent works demonstrated that S. uvarum and S. kudriavzevii species have a Sugar/Ethanol Yield higher than S. cerevisiae, especially at low temperature [29, 41]. However these species are susceptible to high ethanol content and elevated temperature and are not adapted to harsh fermentation conditions. Additional investigations with higher sugar concentrations confirmed that some of these hybrids can reduce the ethanol content in wine up to 0.4% without excessive production of acetic acid .
The collection of 28 interspecific hybrids obtained also allows the investigation of mitochondrial inheritance effect for the 35 traits. Using the multi locus (ATP6, COX2, COX3) molecular typing described by Albertin et al.  we determined the mitochondrial inheritance of these hybrids (10 mt-Sc, 17 mt-su, 1 not determined). However we failed to establish a statistical link between this mitochondrial inheritance and trait variation (data not shown). This result suggested that the mitochondrial inheritance has a small impact during the alcoholic fermentation. This conclusion has been previously reported using isogenic interspecific hydrids harboring different mitochondrial inheritance in two previous studies [62, 94].
Hybridization results in homeostasis for some traits and creates phenotypic novelties
Temperature had a major effect on many variables, particularly on the fermentation kinetics traits, and numerous strain x temperature interactions could be detected (Fig 5 and Table 2). From a multivariate analysis including both in vivo and in silico hybrids, we showed that the in vivo S. cerevisiae hybrids are more homeostatic than interspecific and S. uvarum hybrids, but it is worth noting that hybridization has the largest effect on homeostasis in S. uvarum. In the interspecific hybrids, homeostasis was observed for traits of biotechnological interest such sugar/ethanol yield and aroma production (Figs 2 and 5).
Beside homeostasis, both intra- and interspecific hybridization were shown to create novel multi-trait phenotypes. The occurrence of such an effect of hybridization was previously occasionally described, essentially for plant inter-specific hybrids of Brassica sp., Gossypium sp., etc. . In addition, the extent of phenotypic novelty depends on temperature, suggesting that environmental conditions may modulate the phenotypic innovation associated with intra- and inter-specific hybridization. Our results show that both intra- and inter-specific hybridization can generate hybrids departing from their parents. For example, interspecific hybrids performed better than their parents for ethyl esters production. Such phenotypic transgression, associated with homeostasis, is particularly interesting from an evolutionary viewpoint. Interspecific hybrids with robust fitness are more likely to colonize winemaking environments that are basically changing. Alternatively, homeostasis for Basic Enological Parameters, Fermentation Kinetics and Aromatic Traits may have been selected by human for winemaking, allowing the dissemination of strains having quite stable phenotypes over temperature changes. Conscious or unconscious anthropic selection may explain why intra- and inter-specific hybridization is so frequent in yeast. Indeed, numerous natural hybrids were described associated with enology [20, 30, 33], but also with other bioprocesses producing alcoholic beverages (beer, cider, etc.) [41, 89, 96]. Altogether, homeostasis, phenotypic novelties and transgressive phenotypes may explain the evolutionary role of hybridization in natural or domesticated yeasts.
S1 Fig. Fermentation kinetics and population dynamics parameters during alcoholic fermentation.
Panel A. Fermentation kinetics: CO2 released was expressed in g.L–1; t-lag (h) corresponded to the time between inoculation and the beginning of CO2 release; t-45 (h) and t-75 (h) were respectively the fermentation time at which 45 g.L-1 and 75 g.L-1 of CO2 were released, excluding t-lag; AFtime (h) was the time necessary to ferment all the sugars in the medium excluding t-lag, and CO2max (g.L–1) corresponded to the total amount of CO2 released at the end of the fermentation. Panel B. CO2 production rate was expressed in g.L–1.h–1; Vmax (g.L–1.h-1) corresponded to the maximum CO2 production rate; t-Vmax (h) was the fermentation time at which Vmax was reached. Panel C. Cell growth: the carrying capacity K was expressed in cell.mL–1; t-N0 (h) and t-Nmax (h) were respectively the time to reach the initial growth point and the carrying capacity K. Panel D. CO2 flux, J, computed by dividing the CO2 production rate by the estimated cell concentration (g.h–1.108 cell–1). Jmax is the maximum flux. Panel E. Evolution of cell Size (diameter, μm) over time. Size-t-Nmax (μm) was the average cell size at t-Nmax. Panel F. Evolution of Viability over time. Viability.t-Nmax and Viability.t-75 (%) were the percentages of living cells at t-Nmax and t-75, respectively.
S2 Fig. Traits with a significant species effect at 18°C.
S3 Fig. Traits with a significant species effect at 26°C.
S4 Fig. Correlation between the 35 fermentation traits analyzed at 18°C (A) and 26°C (B).
Only parameters showing a significant correlation (p-value < 0.05 after Benjamini-Hochberg adjustment) were represented by a dot. Green and red tones correspond to positive and negative correlation, respectively.
S1 Supporting Information. CO2 production measurement.
S2 Supporting Information. Cell growth measurement.
S3 Supporting Information. Cell size and viability traits.
S1 Table. List of primers used for microsatellite analysis.
Conceived and designed the experiments: WA CD MB CG SH DS DdV PM. Performed the experiments: TdS WA MB SlG PM. Analyzed the data: TdS WA CD MB CG SH DS IMP DdV PM. Contributed reagents/materials/analysis tools: WA SlG IMP CD SH CG DS PM. Wrote the paper: TdS WA CD DS DdV PM.
- 1. . Newman HH. On the Production of Monsters by Hybridization. Biological Bulletin. 1917;32(5):306–21.
- 2. Baack EJ, Whitney KD, Rieseberg LH. Hybridization and genome size evolution: timing and magnitude of nuclear DNA content increases in Helianthus homoploid hybrid species. New Phytologist. 2005;167(2):623–30. pmid:15998412
- 3. Mallet J. Hybridization as an invasion of the genome. Trends in Ecology & Evolution. 2005;20(5):229–37.
- 4. Hallauer AR, Miranda J. Quantitative Genetics in Maize Breeding. Iowa St Univ Press, Ames, Ia. 1981.
- 5. Sprague GF. Heterosis in Maize: Theory and Practice. In: Frankel PDR, editor. Heterosis. Monographs on Theoretical and Applied Genetics: Springer Berlin Heidelberg; 1983. p. 47–70.
- 6. Coque M, Gallais A. Genetic variation for N-remobilization and postsilking N-up-take in a set of maize recombinant inbred lines. 2. Line per se performance and comparison with testcross performance. Maydica. 2008;53(1).
- 7. Sarkissian IV, Kessinger MA, Harris W. Differential rates of development of heterotic and nonheterotic young maize seedlings, i. Correlation of differential morphological development with physiological differences in germinating seeds. Proceedings Of The National Academy Of Sciences Of The United States Of America. 1964;51(2):212–8. pmid:16591144
- 8. Jarosz DF, Taipale M, Lindquist S. Protein Homeostasis and the Phenotypic Manifestation of Genetic Diversity: Principles and Mechanisms. Annual Review of Genetics. 2010;44(1):189–216.
- 9. Lerner IM. Genetic homeostasis. 1954: Wileey, New York, 134 pp.
- 10. Meyers L, Bull J. Fighting change with change: adaptive variation in an uncertain world. Trends in Ecology & Evolution. 2002;17(12):551–7.
- 11. Waddington C. Canalization of development and the inheritance of acquired characters. Nature. 1942;150(3811):563–5.
- 12. Adams MW, Shank DB. The Relationship of Heterozygosity to Homeostasis in Maize Hybrids. Genetics. 1959;44(5):777–86. pmid:17247856
- 13. Leamy L, Thorpe RS. Morphometric studies in inbred and hybrid house mice. Heterosis, homeostasis and heritability of size and shape. Biological Journal of the Linnean Society. 1984;22(3):233–41.
- 14. Plech M, de Visser JAGM, Korona R. Heterosis Is Prevalent Among Domesticated but not Wild Strains of Saccharomyces cerevisiae. G3: Genes|Genomes|Genetics. 2014;4(2):315–23. pmid:24347627
- 15. Shapira R, Levy T, Shaked S, Fridman E, David L. Extensive heterosis in growth of yeast hybrids is explained by a combination of genetic models. Heredity (Edinb). 2014.
- 16. Timberlake WE, Frizzell MA, Richards KD, Gardner RC. A new yeast genetic resource for analysis and breeding. Yeast. 2011;28(1):63–80. pmid:20853274
- 17. Zorgo E, Gjuvsland A, Cubillos FA, Louis EJ, Liti G, Blomberg A, et al. Life History Shapes Trait Heredity by Accumulation of Loss-of-Function Alleles in Yeast. Mol Biol Evol. 2012;29(7):1781–9. pmid:22319169
- 18. Greig D. Reproductive isolation in Saccharomyces. Heredity. 2008;102(1):39–44. pmid:18648383
- 19. Morales L, Dujon B. Evolutionary Role of Interspecies Hybridization and Genetic Exchanges in Yeasts. Microbiology and Molecular Biology Reviews. 2012;76(4):721–39. pmid:23204364
- 20. Lopandic K, Tiefenbrunner W, Gangl H, Mandl K, Berger S, Leitner G, et al. Molecular profiling of yeasts isolated during spontaneous fermentations of Austrian wines. FEMS Yeast Res. 2008: 1063–1075. pmid:18435743
- 21. Lopes CA, Barrio E, Querol A. Natural hybrids of S. cerevisiae x S. kudriavzevii share alleles with European wild populations of Saccharomyces kudriavzevii. FEMS Yeast Res. 2010;10(4):412–21. pmid:20337723
- 22. Peris D, Belloch C, Lopandic K, Alvarez-Perez JM, Querol A, Barrio E. The molecular characterization of new types of Saccharomyces cerevisiae x S. kudriavzevii hybrid yeasts unveils a high genetic diversity. Yeast. 2011;29(2):81–91 pmid:21287575
- 23. Gonzalez SS, Barrio E, Querol A. Molecular characterization of new natural hybrids of Saccharomyces cerevisiae and S. kudriavzevii in brewing. Appl Environ Microbiol. 2008;74(8):2314–20. pmid:18296532
- 24. Le Jeune C, Lolier M, Demuyter C, Enry C, Legras J, Aigle M, et al. Characterization of natural hybrids of Saccharomyces cerevisiae and Saccharomyces bayanus var. uvarum. FEMS Yeast Res. 2007;7(4):540–9. pmid:17302940
- 25. Erny C, Raoult P, Alais A, Butterlin G, Delobel P, Matei-Radoi F, et al. Ecological success of a group of Saccharomyces cerevisiae/Saccharomyces kudriavzevii hybrids in the northern european wine-making environment. Appl Environ Microbiol. 2012;78(9):3256–65. pmid:22344648
- 26. Masneuf I, Hansen J, Groth C, Piskur J, Dubourdieu D. New hybrids between Saccharomyces sensu stricto yeast species found among wine and cider production strains. Appl Environ Microbiol. 1998;64(10):3887–92. pmid:9758815
- 27. Antunovics Z, Irinyi L, Sipiczki M. Combined application of methods to taxonomic identification of Saccharomyces strains in fermenting botrytized grape must. J Appl Microbiol. 2005;98(4):971–9. pmid:15752344
- 28. Nguyen HV, Lepingle A, Gaillardin CA. Molecular typing demonstrates homogeneity of Saccharomyces uvarum strains and reveals the existence of hybrids between S. uvarum and S. cerevisiae, including the S. bayanus type strain CBS 380. Syst Appl Microbiol. 2000;23(1):71–85. pmid:10879981
- 29. Arroyo-Lopez FN, Orlic S, Querol A, Barrio E. Effects of temperature, pH and sugar concentration on the growth parameters of Saccharomyces cerevisiae, S. kudriavzevii and their interspecific hybrid. Int J Food Microbiol. 2009;131(2–3):120–7. pmid:19349089
- 30. Belloch C, Orlic S, Barrio E, Querol A. Fermentative stress adaptation of hybrids within the Saccharomyces sensu stricto complex. Int J Food Microbiol. 2008;122(1–2):188–95.
- 31. Gamero A, Tronchoni J, Querol A, Belloch C. Production of aroma compounds by cryotolerant Saccharomyces species and hybrids at low and moderate fermentation temperatures. J Appl Microbiol. 2013;114(5):1405–14. pmid:23294204
- 32. Tronchoni J, Gamero A, Arroyo-Lopez FN, Barrio E, Querol A. Differences in the glucose and fructose consumption profiles in diverse Saccharomyces wine species and their hybrids during grape juice fermentation. Int J Food Microbiol. 2009;134(3):237–43. pmid:19632733
- 33. Borneman AR, Desany BA, Riches D, Affourtit JP, Forgan AH, Pretorius IS, et al. The genome sequence of the wine yeast VIN7 reveals an allotriploid hybrid genome with Saccharomyces cerevisiae and Saccharomyces kudriavzevii origins. FEMS Yeast Res. 2012;12(1):88–96. pmid:22136070
- 34. Fukuda K, Yamamoto N, Kiyokawa Y, Yanagiuchi T, Wakai Y, Kitamoto K, et al. Balance of activities of alcohol acetyltransferase and esterase in Saccharomyces cerevisiae is important for production of isoamyl acetate. Appl Environ Microbiol. 1998;64(10):4076–8. 35. pmid:9758847
- 35. Dunn B, Paulish T, Stanbery A, Piotrowski J, Koniges G, Kroll E, et al. Recurrent Rearrangement during Adaptive Evolution in an Interspecific Yeast Hybrid Suggests a Model for Rapid Introgression. PLoS Genet. 2013;9(3):e1003366. pmid:23555283
- 36. Piotrowski JS, Nagarajan S, Kroll E, Stanbery A, Chiotti KE, Kruckeberg AL, et al. Different selective pressures lead to different genomic outcomes as newly-formed hybrid yeasts evolve. BMC evolutionary biology. 2012;12:46. pmid:22471618
- 37. Peris D, Lopes CA, Belloch C, Querol A, Barrio E. Comparative genomics among Saccharomyces cerevisiae x Saccharomyces kudriavzevii natural hybrid strains isolated from wine and beer reveals different origins. BMC Genomics. 2012;13:407. pmid:22906207
- 38. de Barros Lopes M, Bellon JR, Shirley NJ, Ganter PF. Evidence for multiple interspecific hybridization in Saccharomyces sensu stricto species. FEMS Yeast Research. 2002;1(4):323–31. pmid:12702336
- 39. Almeida P, Gonçalves C, Teixeira S, Libkind D, Bontrager M, Masneuf-Pomarède I, et al. A Gondwanan imprint on global diversity and domestication of wine and cider yeast Saccharomyces uvarum. Nat Commun. 2014;5.
- 40. Sipiczki M. Interspecies hybridization and recombination in Saccharomyces wine yeasts. FEMS Yeast Res. 2008.
- 41. Masneuf-Pomarede I, Bely M, Marullo P, Lonvaud-Funel A, Dubourdieu D. Reassessment of phenotypic traits for Saccharomyces bayanus var. uvarum wine yeast strains. Int J Food Microbiol. 2010;139(1–2):79–86. pmid:20388574
- 42. Sebastiani F, Barberio C, Casalone E, Cavalieri D, Polsinelli M. Crosses between Saccharomyces cerevisiae and Saccharomyces bayanus generate fertile hybrids. Res Microbiol. 2002;153(1):53–8. pmid:11881899
- 43. Solieri L, Antunez O, Perez-Ortin JE, Barrio E, Giudici P. Mitochondrial inheritance and fermentative: oxidative balance in hybrids between Saccharomyces cerevisiae and Saccharomyces uvarum. Yeast. 2008;25(7):485–500 pmid:18615860
- 44. Antonelli A, Castellari L, Zambonelli C, Carnacini A. Yeast influence on volatile composition of wines. J Agric Food Chem. 1999;47:1139–44. pmid:10552428
- 45. Caridi A, Cufari A, Ramondino D. Winemaking from gaglioppo grapes with hybrid strains ofSaccharomyces. Folia microbiologica. 2002;47(4):407–8. pmid:12422518
- 46. Masneuf I, Murat M, Naumov G, Tominaga T, Dubourdieu D. Hybrids Saccharomyces cerevisiae x Saccharomyces bayanus var. uvarum having a high liberating ability of some sulfur varietal aromas of Vitis vinifera Sauvignon blanc wines. Journal international des sciences de la vigne et du vin. 2002;36(4):205–12.
- 47. Camarasa C, Sanchez I, Brial P, Bigey F, Dequin S. Phenotypic Landscape of Saccharomyces cerevisiae during Wine Fermentation: Evidence for Origin-Dependent Metabolic Traits. PLoS ONE. 2011;6(9):e25147. Epub 2011/09/29. pmid:21949874
- 48. Liti G, Carter D, Moses A, Warringer J, Parts L, James S, et al. Population genomics of domestic and wild yeasts. Nature. 2009;458:337–41. pmid:19212322
- 49. Naumov GI. [Saccharomyces bayanus var. Uvarum comb.nov., a new variety established by genetic analysis]. Mikrobiologiia. 2000;69(3):410–4. pmid:10920813
- 50. Rainieri S, Zambonelli C, Hallsworth JE, Pulvirenti A, Giudici P. Saccharomyces uvarum, a distinct group within Saccharomyces sensu stricto. FEMS Microbiol Lett. 1999;177(1):177–85. pmid:10436934
- 51. Naumov GI, Masneuf I, Naumova ES, Aigle M, Dubourdieu D. Association of Saccharomyces bayanus var. uvarum with some French wines: genetic analysis of yeast populations. Res Microbiol. 2000;151(8):683–91. pmid:11081582
- 52. Naumov GI, Naumova ES, Antunovics Z, Sipiczki M. Saccharomyces bayanus var. uvarum in Tokaj wine-making of Slovakia and Hungary. Appl Microbiol Biotechnol. 2002;59(6):727–30. pmid:12226732
- 53. Torriani S, Zapparoli G, Suzzi G. Genetic and phenotypic diversity of Saccharomyces sensu stricto strains isolated from Amarone wine. Antonie Van Leeuwenhoek. 1999;75(3):207–15. pmid:10427409
- 54. Blein-Nicolas M, Albertin W, Valot B, Marullo P, Sicard D, Giraud C, et al. Yeast proteome variations reveal different adaptive responses to grape must fermentation. Mol Biol Evol. 2013;30(6):1368–83. pmid:23493259
- 55. Demuyter C, Lollier M, Legras JL, Le Jeune C. Predominance of Saccharomyces uvarum during spontaneous alcoholic fermentation, for three consecutive years, in an Alsatian winery. J Appl Microbiol. 2004;97(6):1140–8. pmid:15546404
- 56. Tosi E, Azzolini M, Guzzo F, Zapparoli G. Evidence of different fermentation behaviours of two indigenous strains of Saccharomyces cerevisiae and Saccharomyces uvarum isolated from Amarone wine. Journal Of Applied Microbiology. 2009;107(1):210–8. pmid:19245401
- 57. Bon E, Neuveglise C, Casaregola S, Artiguenave F, Wincker P, Aigle M, et al. Genomic exploration of the hemiascomycetous yeasts: 5. Saccharomyces bayanus var. uvarum. FEBS Lett. 2000;487(1):37–41. pmid:11152880
- 58. Fischer G, Neuveglise C, Durrens P, Gaillardin C, Dujon B. Evolution of gene order in the genomes of two related yeast species. Genome Res. 2001;11(12):2009–19 pmid:11731490
- 59. Rainieri S, Zambonelli C, Tini V, Castellari L, Giudici P. The enological traits of thermotolerant Saccharomyces strains. Am J Enol Vitic. 1998;49(3):319–24.
- 60. Castellari L, Ferruzzi M, Magrini A, Giudici P, Passarelli P, Zambonelli C. Unbalanced wine fermentation by cryotolerant vs. non-cryotolerant Saccharomyces strains. Vitis. 1994;33(1):49–52.
- 61. Belloch C, Pérez-Torrado R, González SS, Pérez-Ortín JE, García-Martínez J, Querol A, et al. Chimeric genomes of natural hybrids of Saccharomyces cerevisiae and Saccharomyces kudriavzevii. Applied And Environmental Microbiology. 2009;75(8):2534–44. pmid:19251887
- 62. Albertin W, da Silva T, Rigoulet M, Salin B, Masneuf-Pomarede I, de Vienne D, et al. The mitochondrial genome impacts respiration but not fermentation in interspecific Saccharomyces hybrids. PLoS One. 2013;8(9):e75121. pmid:24086452
- 63. Albertin W, Marullo P, Aigle M, Bourgais A, Bely M, Dillmann C, et al. Evidence for autotetraploidy associated with reproductive isolation in Saccharomyces cerevisiae: towards a new domesticated species. J Evol Biol. 2009;22(11):2157–70. pmid:19765175
- 64. Liti G, Carter DM, Moses AM, Warringer J, Parts L, James SA, et al. Population genomics of domestic and wild yeasts. Nature. 2009;458(7236):337–41. pmid:19212322
- 65. Albertin W, Marullo P, Aigle M, Dillmann C, de Vienne D, Bely M, et al. Population size drives industrial Saccharomyces cerevisiae alcoholic fermentation and is under genetic control. Appl Environ Microbiol. 2011;77(8):2772–84 pmid:21357433
- 66. Marullo P, Bely M, Masneuf-Pomarede I, Pons M, Aigle M, Dubourdieu D. Breeding strategies for combining fermentative qualities and reducing off-flavor production in a wine yeast model. FEMS Yeast Res. 2006;6(2):268–79. pmid:16487348
- 67. Marullo P, Mansour C, Dufour M, Albertin W, Sicard D, Bely M, et al. Genetic improvement of thermo-tolerance in wine Saccharomyces cerevisiae strains by a backcross approach. FEMS Yeast Res. 2009;9(8):1148–60. Epub 2009/09/18. pmid:19758333.
- 68. Marullo P, Bely M, Masneuf-Pomarede I, Aigle M, Dubourdieu D. Inheritable nature of enological quantitative traits is demonstrated by meiotic segregation of industrial wine yeast strains. FEMS Yeast Res. 2004;4(7):711–9. pmid:15093774
- 69. Masneuf-Pomarede I, Le Jeune C, Durrens P, Lollier M, Aigle M, Dubourdieu D. Molecular typing of wine yeast strains Saccharomyces bayanus var. uvarum using microsatellite markers. Syst Appl Microbiol. 2007;30(1):75–82. pmid:16621402
- 70. Richards KD, Goddard MR, Gardner RC. A database of microsatellite genotypes for Saccharomyces cerevisiae. Antonie Van Leeuwenhoek. 2009;96(3):355–9. pmid:19396625
- 71. Pate JB, Lodge JP, Wartburg AF. Effect of Pararosaniline in the Trace Determination of Sulfur Dioxide. Analytical Chemistry. 1962;34(12):1660–2.
- 72. Weldegergis BT, Tredoux AG, Crouch AM. Application of a headspace sorptive extraction method for the analysis of volatile components in South African wines. journal of agricultural and food chemistry. 2007;55(21):8696–702. pmid:17927151
- 73. Tominaga , Furrer A, Henry R, Dubourdieu D. Identification of new volatile thiols in aroma of Vitis vinifera L. var. Sauvignon blanc wines. Flavour Fragr J. 1998;13:159–62.
- 74. Gao X, Alvo M, Chen J, Li G. Nonparametric multiple comparison procedures for unbalanced one-way factorial designs. Journal of Statistical Planning and Inference. 2008;138(8):2574–91. pmid:18709183
- 75. Dray S, Dufour A-B. The ade4 package: implementing the duality diagram for ecologists. Journal of Statistical Software. 2007;22(4):1–20.
- 76. Spor A, Wang S, Dillmann C, de Vienne D, Sicard D. "Ant" and "Grasshopper" Life-History Strategies in Saccharomyces cerevisiae. PLoS ONE. 2008;3(2):e1579. pmid:18270570
- 77. Crusio WE, Kerbusch JML, Abeelen JHF. The replicated diallel cross: A generalized method of analysis. Behav Genet. 1984;14(1):81–104. pmid:6712552
- 78. Vilela-Moura A, Schuller D, Mendes-Faia A, Corte-Real M. Reduction of volatile acidity of wines by selected yeast strains. Appl Microbiol Biotechnol. 2008;80(5):881–90. pmid:18677471
- 79. Dufour M, Zimmer A, Thibon C, Marullo P. Enhancement of volatile thiol release of Saccharomyces cerevisiae strains using molecular breeding. Appl Microbiol Biotechnol. 2013;97(13):5893–905. pmid:23423325
- 80. Roncoroni M, Santiago M, Hooks DO, Moroney S, Harsch MJ, Lee SA, et al. The yeast IRC7 gene encodes a beta-lyase responsible for production of the varietal thiol 4-mercapto-4-methylpentan-2-one in wine. Food Microbiol. 2011;28(5):926–35. pmid:21569935
- 81. Lopez-Malo M, Querol A, Guillamon JM. Metabolomic comparison of Saccharomyces cerevisiae and the cryotolerant species S. bayanus var. uvarum and S. kudriavzevii during wine fermentation at low temperature. PLoS One. 2013;8(3):e60135. pmid:23527304
- 82. Sumby KM, Grbin PR, Jiranek V. Microbial modulation of aromatic esters in wine: current knowledge and future prospects. Food Chemistry. 2010;121(1):1–16.
- 83. Beltran G, Novo M, Guillamon JM, Mas A, Rozes N. Effect of fermentation temperature and culture media on the yeast lipid composition and wine volatile compounds. International Journal of Food Microbiology. 2008;121(2):169–77. pmid:18068842
- 84. Torija MJ, Beltran G, Novo M, Poblet M, Guillamon JM, Mas A, et al. Effects of fermentation temperature and Saccharomyces species on the cell fatty acid composition and presence of volatile compounds in wine. International Journal of Food Microbiology. 2003;85(1–2):127–36. pmid:12878388
- 85. Mason AB, Dufour JP. Alcohol acetyltransferases and the significance of ester synthesis in yeast. Yeast. 2000;16(14):1287–98. pmid:11015726
- 86. Saerens SM, Verstrepen KJ, Van Laere SD, Voet AR, Van Dijck P, Delvaux FR, et al. The Saccharomyces cerevisiae EHT1 and EEB1 genes encode novel enzymes with medium-chain fatty acid ethyl ester synthesis and hydrolysis capacity. J Biol Chem. 2006;281(7):4446–56. pmid:16361250
- 87. He Y, Dong J, Yin H, Chen P, Lin H, Chen L. Monitoring of the production of flavour compounds by analysis of the gene transcription involved in higher alcohol and ester formation by the brewer's yeast Saccharomyces pastorianus using a multiplex RT-qPCR assay. Journal of the Institute of Brewing. 2014; 120(2):119–126.
- 88. Gamero A, Belloch C, Ibanez C, Querol A. Molecular analysis of the genes involved in aroma synthesis in the species S. cerevisiae, S. kudriavzevii and S. bayanus var. uvarum in winemaking conditions. PLoS One. 2014;9:e97626. pmid:24854353
- 89. Libkind D, Hittinger CT, Valerio E, Goncalves C, Dover J, Johnston M, et al. Microbe domestication and the identification of the wild genetic stock of lager-brewing yeast. Proc Natl Acad Sci U S A. 2011;108(35):14539–44. pmid:21873232
- 90. Tilloy V, Ortiz-Julien A, Dequin S. Reducing ethanol and improving glycerol yield by adaptive evolution of Saccharomyces cerevisiae wine yeast under hyperosmotic conditions. Applied And Environmental Microbiology. 2014;10.1128/aem.03710-13. doi: 10.1128/aem.03710-13.
- 91. Schmidtke LM, Blackman JW, Agboola SO. Production Technologies for Reduced Alcoholic Wines. Journal of Food Science. 2012;77(1):R25–R41. pmid:22260123
- 92. Aguera E, Bes M, Roy A, Camarasa C, Sablayrolles J-M. Partial removal of ethanol during fermentation to obtain reduced-alcohol wines. American Journal of Enology and Viticulture. 2010;61(1):53–60.
- 93. Bely M, Renault P, Da Silva T, Masneuf-Pomarède I, W A, Moine V, et al., editors. Non conventional yeasts and alcohol levels reduction. 1st International Symposium; 2013 6 september 2013; Bordeaux.
- 94. Picazo C, Gamero-Sandemetrio E, Orozco H, Albertin W, Marullo P, Matallana E, et al. Mitochondria inheritance is a key factor for tolerance to dehydration in wine yeast production. Letters in Applied Microbiology. 2014;10.1111/lam.12369:n/a-n/a. doi: 10.1111/lam.12369.
- 95. Albertin W, Balliau T, Brabant P, Chevre AM, Eber F, Malosse C, et al. Numerous and rapid nonstochastic modifications of gene products in newly synthesized Brassica napus allotetraploids. Genetics. 2006;173:1101–13. pmid:16624896
- 96. Nguyen HV, Legras JL, Neuveglise C, Gaillardin C. Deciphering the hybridisation history leading to the Lager lineage based on the mosaic genomes of Saccharomyces bayanus strains NBRC1948 and CBS380. PLoS One. 2011;6(10):e25821. pmid:21998701