OENO One Genomic insights into the glutathione metabolism of the wine yeast Starmerella bacillaris 1 Department of Agronomy Food Natural Resources Animals and Environment, University of Padova, Legnaro, Italy 2 Science and Technology Faculty, Free University of Bozen, 39100, Bolzano, Italy 3 Department of Environmental Engineering, Technical University of Denmark, Denmark 4 Department of Biology, University of Padova, Padova, Italy 5 Interdepartmental Centre for Research in Viticulture and Enology, University of Padova, Conegliano, Italy 6 Department of Microbiology, Universidade Federal de Viçosa, Viçosa, Brazil 7 CRIBI Biotechnology Center, University of Padova, Padova, Italy 2† WilsonJoséFernandesLemosJunior1 4† LauraTreu3 Nadai1 Chiara 5 Duarte1 ViníciusdaSilva 6 Campanaro4 Stefano 7 Fabrega-Prats Marta 1 Giacomini1 Alessio * 5 Corich1 Viviana 5 * corresponding author: alessio.giacomini@unipd.it Abstract

Glutathione (GSH) is an antioxidant molecule of great technological interest due to its wide range of applications in the food and beverage industry. In winemaking, although glutathione is produced during fermentation, its addition is possible for the control of oxidative spoilage of wine. Recently, to improve wine quality, mixed fermentation has been proposed by introducing a selection of non-Saccharomyces yeasts as complementary starters to the oenological species S. cerevisiae. Among them, Starmerella bacillaris, an osmophilic and high glycerol producer yeast, has been extensively studied.

In the present study, the genomes of two S. bacillaris strains were compared with S. cerevisiae to identify the GSH metabolic pathway. The results showed that GSH biosynthesis includes the GSH1 and GSH2 genes in both species. The identification of a new transcription factor which binds sites in the promoter region of these genes underlined differences in the transcriptional regulation of both species. Additionally, between S. bacillaris strains, a high number of polymorphisms was found in genes involved in GSH redox balance. Preliminary laboratory scale fermentations revealed marked differences in the cell glutathione content of the two S. bacillaris strains. By comparing genomes, it was possible to gain a better understanding of the genes involved in the GSH metabolism pathway in S. bacillaris.

Keywords: Saccharomyces cerevisiae, Candida zemplinina, glutathione, GSH1, GSH2

Introduction

During alcoholic fermentation, sugars present in grape must are transformed into ethanol. This process is generally carried out by a single Saccharomyces cerevisiae strain added to the grape must as a starter culture. Many non-Saccharomyces yeasts are present on the grape surface and, therefore, in the grape must. In the past, they were considered undesirable spoilage microorganisms, because they were often isolated from stuck or sluggish fermentations, or from wines with anomalous analytical compositions or negative sensorial profiles (Jolly et al., 2013; Ciani and Comitini, 2015). Throughout the past decade, they have become popular in the wine industry and their role has been re-considered, as many species contribute to wine fermentation and may positively affect wine quality (Ciani and Comitini, 2015; Ivit and Kemp, 2018). Indeed, there has been an increasing number of claims that wines made with Saccharomyces starter cultures are more standardised. The yeast Starmerella bacillaris (synonym Candida zemplinina)is an osmophilic, fructophilic early-colonising wine yeast, which is preferentially isolated from high sugar grape musts (Sipiczki, 2003; Wang et al., 2014). It is part of microbiota in the grape pomace, a main winemaking by-product (Bovo et al., 2011). When used as a starter culture in sequential fermentation with S. cerevisiae strains, this yeast contributes to decreasing ethanol levels and increasing glycerol content while maintaining moderate volatile acid production (Jolly et al., 2014; Bely et al., 2013). Recently, this novel non-conventional wine yeast showedbiocontrol activity against the fungal pathogens Botrytis cinerea and Penicillium expansum (Lemos Junior et al., 2016; Nadai et al., 2018).

Wine is sensitive to oxygen exposure, which can lead to a loss of aroma, the development of atypical aging characters, and undesirable colour changes. In winemaking, glutathione addition has been proposed for the control of oxidative spoilage of wine. Glutathione (L-g-glutamyl-L-cysteinyl-glycine, GSH) is a hydrosoluble tripeptide, an important molecule containing thiol residues that are responsible for its antioxidant property (Kiriyama et al., 2013). GSH present in wine exerts a protective effect on various aromatic compounds (Ugliano et al., 2011). Its concentration can change due to its release from yeast cells via autolysis (Lavigne et al., 2007). Therefore, after fermentation, wine aging on yeast lees can potentially play a role in protecting wine from oxidation.

Moreover, as a result of the winemaking process in wineries, one ton of grapes generates approximately 0.06 t of lees (Oliveira and Duarte, 2014). Therefore, it is worth exploring how GSH recovered from lees can improve sustainability in winemaking.

In S. cerevisiae, genes involved in GSH production and regulation are well-known, as is the biological role of GSH. Indeed, GSH is involved in protection against oxidative stress and the elimination of heavy metals and toxic endogenous metabolites, and it is also a source of cysteine. Additionally, it can take part in nitrogen metabolism and other pathways generating sulphur compounds (Mezzetti et al., 2014).

In S. cerevisiae, GSH is synthesised via two ATP-dependent steps (Figure 1).

Figure 1. Gluthatione metabolism.

GSH1 Gamma glutamylcysteine synthetase, GSH2 Glutathione synthetase, GLR1 Glutathione oxidoreductase, GRX1-8 Glutaredoxins, GTO1-3 Glutathione transferases Omega-like, ROS Reactive oxygen species, RX Xenobiotics, GTT1-2 Glutathione transferases, GRX1-2 Glutaredoxins, YCF1 Vacuolar glutathione S-conjugate transporter, ECM38 Gamma-glutamyltranspeptidase, DUG1-3 GSH degradosomal complex

Figure 1. Gluthatione metabolism.

GSH1 Gamma glutamylcysteine synthetase, GSH2 Glutathione synthetase, GLR1 Glutathione oxidoreductase, GRX1-8 Glutaredoxins, GTO1-3 Glutathione transferases Omega-like, ROS Reactive oxygen species, RX Xenobiotics, GTT1-2 Glutathione transferases, GRX1-2 Glutaredoxins, YCF1 Vacuolar glutathione S-conjugate transporter, ECM38 Gamma-glutamyltranspeptidase, DUG1-3 GSH degradosomal complex

γ-Glutamylcysteine synthetase (GSS1; coded by GSH1) catalyses the first and rate-limiting step, during which dipeptide γ-Glu-Cys is formed from glutamate and cysteine (Lisowsky, 1993). The second step is catalysed by glutathione synthetase (GSS2) coded by GSH2, which ligates γ-Glu-Cys with glycine (Grant et al., 1997). GSH is degraded to form cysteinyl-glycine. Therefore, the cell content of γ-Glu-Cys precursor and cysteinyl-glycine catabolite can give information about GSH turnover in the cell (Penninckx, 2002). GSH activity is important during oxidative stress conditions as a cofactor for stress defence enzymes. Oxidative stress converts GSH to its oxidised disulphide form (GSSG) by ROS (reactive oxygen species) or in reactions catalysed by glutaredoxines (GRX). Eight related glutaredoxins have been identified in S. cerevisiae (coded by GRX1-8 genes). GRX1-2 are regulated via stress-responsive STRE elements. The same reaction is catalysed by Gto1–3 proteins as well. These enzymes are induced in response to oxidants under the control of Yap1p and STRE-responsive elements (Morano et al., 2011). Reduced GSH from GSSG is regenerated in an NADPH-dependent reaction catalysed by a glutathione reductase (GRS) coded by GLR1. GSH can be conjugated to xenobiotics (RX) by glutathione S-transferases (GST), which comprise a major family of proteins involved in the detoxification of many xenobiotic compounds. GSH conjugates are transported to the vacuole by the Ycf1 GS-X pump. The GST family includes Gtt1–2 and Grx1–2 proteins, as well as Gto1-3, although the latter proteins are not active as GSTs with a xenobiotic compound, but, as already mentioned, as thioltransferases (glutaredoxins, GRX) (Lu, 2009). Gto1-3 proteins also function as glutathione peroxidases (GPX) and provide major enzymatic defence against oxidative stress caused by hydroperoxides (Barreto et al., 2006). Moreover, specific Gpx proteins (coded by GPX1-3) are present and protect membrane lipids from peroxidation (Avery and Avery, 2001). GSH is a source of cysteine whose cellular concentration can be modulated by GSH degradation, which involves the gamma-glutamyl transpeptidase encoded by ECM38 (Kumar et al., 2003) or, with an alternative pathway, the Cys-Gly metallo-di-peptidase encoded by DUG1 (Kaur et al., 2012).

Knowledge about S. cerevisiae's GSH metabolism has been used to investigate the possibility of increasing glutathione content using new S. cerevisiae wine starters during alcoholic fermentation in order to prevent wine oxidative spoilage (Mezzetti et al., 2014). However, little information is available regarding genes involved in non-Saccharomyces yeast glutathione production and the contribution of these non-conventional starters to glutathione content in wines.

In the present study, two S. bacillaris strains whose genome were recently sequenced (Lemos Junior et al., 2017a; Lemos Junior et al., 2017b; Lemos Junior et al., 2018) were considered. By comparing the genomes of S. bacillaris and S. cerevisiae, the genes involved in the GSH metabolism pathway were identified and genomic variations of both S. bacillaris strains were investigated. Moreover, for both strains, the GSH cell content was determined in synthetic must, mimicking grape must composition.

 Materials and methods 1. Bioinformatic analysis

The identification of single nucleotide polymorphisms (SNPs) was performed using SAMtools mpileup (Li et al., 2009) and the output file was converted to a VCF format using BCFtools (https://github.com/samtools/BCFtools). SNPs annotation in the genes of interest was conducted using SNPeff software (version 4.3g) (Cingolani et al., 2012). The Neural Network Promoter Prediction (Reese, 2001) was chosen as the tool for the prediction of GSH1 and GSH2 promoter regions, whereas putative binding sites for transcriptional factors near promoter sequences were identified with Yeastract (Teixeira et al., 2013). The automated protein-structure homology-modeling server SWISS-MODEL (Biasini et al., 2014) was used to carry out GSH1 and GSH2 modelling to compare with those obtained from S. cerevisiae EC1118 (reference templates). Ortholog analysis was carried out with SPOCS (Species Paralogy and Orthology Clique Solver) (Curtis et al., 2013) to compare protein sequences related to the GSH pathway from S. bacillaris FRI751, S. bacillaris PAS13 and S. cerevisiae S288c. Protein multiple sequence alignment, visualisation and analysis were performed with Jalview v2.11 (Waterhouse et al., 2009). Finally, a phylogenetic tree was constructed with crucial proteins involved in GSH biosynthesis using MEGA 7.0 (Kumar et al., 2016). The Maximum Likelihood method based on the JTT matrix-based model with 1,000 bootstrap replicates was chosen. Due to its remarkable difference in terms of GSH content, S. cerevisiae EC1118 was included in all in silico analyses.

 2. Yeast strains and growth conditions

Two S. bacillaris strains (FRI751 and PAS13) were isolated from dried grapes of the Raboso Piave variety, as described by Lemos Junior et al. (2016). S. cerevisiae EC1118 (Lallemand Inc., Montreal, Canada) was used as the control. The starter cultures were prepared from YPD agar plate (yeast extract 10 g/L, peptone 10 g/L, dextrose 20 g/L) to inoculate 5 mL of YPD broth in 15 mL tubes. A stationary phase culture with approximately 108 cells/mL was obtained after 24 hours of incubation at 30 °C; it was determined by OD measurements and confirmed by plate counts (CFU/mL).

3. Fermentations trials

Pre-cultures of each strain used in this work were prepared as described by Bovo et al. (2016). A suitable aliquot of each yeast culture, corresponding to a final cell concentration of 1.5 x 106 cells/mL, was used to inoculate 120 mL bottles, fitted with closures that enabled the carbon dioxide to escape, containing 100 mL of MS300 synthetic must according to Bely et al. (1990) with the following modifications: 100 g/L of glucose, 100 g/L of fructose and 6 g/L of malic acid, pH 3.3.

After yeast inoculation, the bottles were incubated at 20 °C. All experiments were performed in triplicate. CO2 production was monitored by weighing the bottles twice a day and calculating the weight loss of each culture. The fermentations were terminated when the weight loss rate had dropped to under 0.05 g/day.

HPLC analysis was performed to determine the concentrations of residual glucose and fructose, acetic acid, glycerol and ethanol when the fermentations had been terminated, as described by Lemos Junior et al. (2019).

4. Thiols measurement

Yeast lees from synthetic wine were collected at the end of fermentation. The supernatant was discarded, and the pellet was washed twice (5 mL of 0.9 % NaCl). The pellet was weighed and re-suspended in 10 % w/v of 0.1 N HCl, 1 mmol/L Na2-EDTA. To prompt cellular lysis, glass beads were added and the suspension was vortexed for 3 min. Once cellular lysis had occured, 50 µL of supernatant was used for derivatisation with the fluorescent dye SBD-F, as described by Masi et al. (2002). 20 μl of filtered samples was analysed by reverse-phase HPLC separation. The HPLC instrument (Shimadzu, Tokio, Japan) was equipped with a refractive index detector fluorescent (excitation wavelength: 386 nm; emission wavelength: 516 nm) for the determination of cysteinyl-glycine, gGluCys, GSH. The chromatographic conditions were realised with the LC C18 100Å column (150 mm × 4.6 mm I.D., 5 µm particle size; Luna, Phenomenex, USA), running at a flow rate of 1 mL/min and at room temperature. The mobile phase was NH4+ formate 50 mM pH 2.9 containing 3 % methanol.

 5. Statistical Analysis

The statistical data analysis was performed using the XLSTAT software, version 2016.02 (Addinsoft, Paris, France). Data were subjected to a Student’s t-test. Differences were considered statistically significant for a p-value of less than 0.05.

 Results and discussion 1. Comparison of the genes involved in the GSH metabolism pathway in S. bacillaris strains and S. cerevisiae

To better understand the genes involved in the GSH metabolism pathway in S. bacillaris, the genomes of the two previously sequenced (Lemos Junior et al., 2017a; Lemos Junior et al., 2017b; Lemos Junior et al., 2018) strains were analysed and compared to S. cerevisiae EC1118, a well-known commercial wine strain.

The conventional GSH biosynthetic pathway (Figure 1) was evaluated by comparing S. bacillaris strains and S. cerevisiae EC1118/S288c in terms of enzymatic classes, localisation of genes across the genome (synteny), presence of transcriptional factors binding sites in promoter regions, and the putative effect of SNPs on GSH-related genes.

KEGG pathway analysis was performed using a data-set of 24 protein-coding genes (Supplementary Datasheet 1) obtained after ortholog prediction between S. bacillaris FRI751, S. bacillaris PAS13, and S. cerevisiae S288c. Proteins that belong to the enzymatic classes involved in GSH metabolism were found, such as GSS1 (γ-Glutamylcysteinesynthetase), GSS2 (glutathione synthetase), GRX (glutaredoxin), GST (glutathione S-transferase), GPX (glutathione peroxidase) and GSR (glutathione reductase). Among the three GST-coding genes, two out of three are in a tandem array separated by an intergenic region of 1,698 bp on the chromosome sequence in both S. bacillaris strains. Additionally, three genes encoding for components of the GSH degradosomal complex (DUG1, DUG2, and DUG3) were present, as well as two hydroxyacylglutathione hydrolases coding-sequences(GLO1 and GLO4). The two hydrolases are involved the detoxification of methylglyoxal (a by-product of glycolysis; Inoue et al., 2011). No genes related to GSH plasma-membrane transport or vacuolar GS–X pump were found, although a mitochondrial outer membrane protein (POR1) was identified.

To better understand whether genomic variants were present and their potential effects on protein primary structure, sequence comparison of the GSH-related proteins between the two S. bacillaris strains and their orthologs in S. cerevisiae was performed. Although the same enzymatic activities participate in the GSH-pathway in both fungal species, the comparison between S. bacillaris strains revealed a high number of SNPs able to impact the functioning of twenty genes (Table 1). According to Schacherer et al. (2007), the interpretation of biological assays can be facilitated by understanding the number and the position of genomic variants between strains that show different phenotypes. In a previous study conducted by Junior Lemos et al. (2018), the authors reported the existence of 33,771 high-quality variants in 1,146 genes in S. bacillaris FRI751 and PAS13. A similar content of genomic variants (37,424 SNPs) was identified in S. cerevisiae S288c and SK1 (Schacherer et al., 2007), although SNP content can vary significantly (from 13,787 to 57,463) due to the reference strain of choice and the phylogenetic distance between the genomes of interest, as described in several studies (Borneman et al., 2008; Otero et al., 2010; Meijnen et al., 2016).

Table 1. Annotation of twenty-four* proteins involved in GSH metabolism and number of variants for each impact category in S. bacillaris strains.
titre du tableau
Reference pathway (KO) Gene ID Ortholog in S. cerevisiae Enzyme Commission number (E.C.M) SNPs
High Low Moderate Modifier
K00432 S05_1418 HYR1 1.11.1.9 0 0 0 1
K00432 S03_775 GPX2 1.11.1.9 0 1 0 49
K15040 S07_1803 POR1 # 0 1 0 87
K00383 S10_2233 GLR1 1.8.1.7 0 0 1 0
K18802 S05_1468 DUG3 # 0 8 0 91
K15428 S09_2037 DUG1 3.4.13.- 0 1 0 33
K01070 S31_3701 YJL068C 3.1.2.12 0 3 2 122
K11204 S05_1424 GSH1 6.3.2.2 0 0 0 1
K00799 S32_3749 URE2 2.5.1.18 0 1 0 33
K01013 S09_2034 RDL1 2.8.1.- 0 0 0 31
K01069 S25_3474 GLO4 3.1.2.6 0 2 0 13
K01759 S17_3012 GLO1 4.4.1.5 0 0 0 2
K13566 S05_1296 NIT2 3.5.1.3 0 0 0 72
K01920 S25_3478 GSH2 6.3.2.3 0 0 0 11
K07393 S03_730 ECM4 # 0 5 3 127
K07232 S02_518 GCG1 4.3.2.7 0 1 0 5
K14262 S09_2045 DUG2 3.4.-.- 0 18 2 109
K03233 S01_244 TEF4 # 0 0 0 3
K00799 S06_1485 GTT1 2.5.1.18 0 6 1 78
K00799 S06_1486 GTT1 2.5.1.18 0 6 0 137
K00121 S11_2303 SFA1 1.1.1.284 0 0 0 0
K07390 S10_2243 GRX5 # 0 0 0 0
K03676 S15_2809 GRX2 # 0 0 0 0
K03386 S15_2788 PRX1 1.11.1.15 0 0 0 0

High (variant is assumed to have disruptive impact in the protein: stop codon gained and loss of function), low (harmless or unlikely to change protein behavior: synonymous codon), moderate (non-disruptive variant that might alter protein effectiveness: missense codon and inframe codon loss) and modifier (non-coding variants or variants affecting non-coding genes: exon variant and downstream gene variant) are parameters used by SnpEff to correlate SNPs and their putative impact on protein functioning. * TEF4 and URE2: GST-like proteins. HYR – HYdroperoxide Resistance; GPX – Glutathione PeroXidase; POR – PORin; GLR – GLutathione Reductase; DUG – Deficient in Utilization of Glutathione; GSH – Glutathione; URE – UREidosuccinate transport; RDL1 – RhoDanese-Like protein; GLO – GLyOxalase; NIT2 – NITrilase superfamily; ECM4 - ExtraCellular Mutant; GCG – Gamma-glutamyl Cyclotransferase acting on Glutathione; TEF – Translation Elongation Factor; GTT – GlutaThione Transferase; SFA – Sensitive to FormAldehyde; GRX – GlutaRedoXin; PRX – PeroxiRedoXin.

At the amino acid level, substitutions were observed in two glutathione S-transferases (ECM4 and GTT1), one glutathione peroxidase, and one glutathione reductase (Figure 2).

Figure 2. Different sequence similarities displayed by four GSH-related proteins as a result of amino acid substitution and/or deletion in S. bacillaris strains associated with genomic variants.

A-GTT1 (K00799); B-ECM4 (K07393); C-GSR (K00383); D-GPX (K00432).

Figure 2. Different sequence similarities displayed by four GSH-related proteins as a result of amino acid substitution and/or deletion in S. bacillaris strains associated with genomic variants.

A-GTT1 (K00799); B-ECM4 (K07393); C-GSR (K00383); D-GPX (K00432).

The enzymes annotated as glutathione S-transferases displayed the highest number of substitutions (K00799: S06_1486p.Thr80Ile, p.Ile93Lys, p.Leu127Ile, p. Arg131Lys, p.Gly135Glu, p. Gly137Ser, p. Asp153His, p. Lys169Ser, p. Glu191Asp, p. Lys216Arg, p. Val226Ile and p. His233Gln; K07393: p. Asp222Asn and p. Gly354Arg), and almost all of them were localised in domain regions (Figures 2a and 2b). Only one amino acid substitution was identified in the glutathione reductase enzymes (K00383: p. Thr461Ile) (Figure 2c), whereas one of the glutathione peroxidase enzymes (K00432; gene ID: S03_775) encoded by S. bacillaris FRI751 displayed a shorter protein length than S. bacillaris PAS13 and S. cerevisiae due to the deletion of the first eight amino acids (Figure 2d). Together, these results suggest that the described amino acid substitutions are not all neutral and they have a functional effect, evidencing a marked inter- and intra-strain variability that can be relevant to determine the differences in GSH production.

Since the three GST-coding genes showed a remarkable number of variants between S. bacillaris strains, reflecting substitutions at the amino acid level in two of them, further investigations have focused on these genes. According to Ma et al. (2009), seven proteins (Gtt1p, Gtt2p, Gto1p, Gto2p, Gtop3, Grx1p and Grx2p) display GST activity in S. cerevisiae. Phylogenetic analysis of GST amino-acid sequences for S. cerevisiae and S. bacillaris strains revealed that the two GST proteins encoded by genes in a tandem array are closely related to the paralogs encoded by GTT1, while the third is grouped with URE2 (Figure 3a). URE2 encodes a bifunctional protein that is not generally included in the glutathione pathway as it is involved in both nitrogen catabolite repression (Coschigano and Magasanik, 1991) and oxidative stress response (Rai and Cooper, 2005). The protein has been shown to exhibit glutathione peroxidase activity and can mutate to acquire GST activity (Coschigano and Magasanik, 1991; Bai et al., 2004). This finding confirms the presence of relevant differences in proteins involved in the GSH system between S. cerevisiae and S. bacillaris.

Figure 3. A-Molecular Phylogenetic analysis using the Maximum Likelihood method.

The evolutionary history was inferred by using the Maximum Likelihood method based on the JTT (Jones Taylor-Thornton) matrix-based model (Jones et al., 1992). The tree with the highest log likelihood (-2135.54) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search was/were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 16 amino acid sequences of GSTs from S. cerevisiae and 6 from S. bacillaris FRI751 and PAS13. All positions containing gaps and missing data were eliminated. There was a total of 70 positions in the final dataset. Evolutionary analyses were conducted in MEGA7 (Kumar et al., 2016); B and C-Transcriptional factors prediction using upstream regions from GSH1 and GSH2 respectively. In each panel, sequences obtained from S. cerevisiae S288c, S. bacillaris FRI751 and S. bacillaris PAS13 are displayed from top to bottom respectively.

Figure 3. A-Molecular Phylogenetic analysis using the Maximum Likelihood method.

The evolutionary history was inferred by using the Maximum Likelihood method based on the JTT (Jones Taylor-Thornton) matrix-based model (Jones et al., 1992). The tree with the highest log likelihood (-2135.54) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search was/were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The analysis involved 16 amino acid sequences of GSTs from S. cerevisiae and 6 from S. bacillaris FRI751 and PAS13. All positions containing gaps and missing data were eliminated. There was a total of 70 positions in the final dataset. Evolutionary analyses were conducted in MEGA7 (Kumar et al., 2016); B and C-Transcriptional factors prediction using upstream regions from GSH1 and GSH2 respectively. In each panel, sequences obtained from S. cerevisiae S288c, S. bacillaris FRI751 and S. bacillaris PAS13 are displayed from top to bottom respectively.

The regulation of the two genes involved in GSH synthesis, GSH1 and GSH2, is crucial for determining the GSH level in the yeast cell. In S. cerevisiae, according to Murata and Kimura (1990), Gsh1p and Gsh2p enzymes are regulated through a feedback-loop by GSH and GSSG respectively, indicating that the step regulated by Gsh1p is rate limiting in GSH biosynthesis.

GSH1 expression is co-regulated by Met4p, a transcription activator that also controls the expression of genes involved in sulfur assimilation (Wheeler et al., 2002). In addition, GSH1 expression is induced by the levels of oxidants, such as hydrogen peroxide, and by heat shock. Both types of regulation involve the transcriptional factor Yap1p (Sugiyama et al., 2000).

Bioinformatic analysis of transcription factor binding sites (TFBS) localised in the promoter region of GSH1 of S. bacillaris revealed two proteins that can potentially be involved in its transcriptional regulation (Azf1p and Tec1p), while five other proteins (Aft2p, Aft1p, Gis1p, Pho4p and Rlm1p) are putatively involved in the regulation of GSH2. These TF binding sites were not identified in the corresponding regions of S. cerevisiae S288c (Figures 3b and 3c). Moreover, no Met4p or Yap1p analogs were found to regulate GSH1 or GSH2 in S. bacillaris strains. The TFBS predicted in the promoter region of GSH1 is associated with diauxic shift under nutrient depletion and transformation of non-fermentable carbon source through oxidative metabolism (Azf1p, Gis1p, Pho4p). In these conditions, cells require protection against ROS and oxidative stress in general (Cherry et al., 2011). Interestingly, Aft2p and Aft1p, putatively involved in the control of S. bacillaris GSH2,are involved in iron utilisation and homeostasis in S. cerevisiae. Although in this yeast no co-regulation has been found between GSH biosynthesis genes and iron metabolism, the involvement of glutathione in maturation Fe/S proteins has been demonstrated (Kumar et al., 2011).

GSH1 and GSH2 genes from S. bacillaris strains displayed 26 and 20 insertions or deletions (INDELs) respectively when compared to S. cerevisiae EC1118, but none were located in the GSH2 regions of substrate binding (Table 2). Three amino acid substitutions were localised in three out of four domains in GSH2. Both GSH genes encode proteins with a reduced length in comparison to the S. cerevisiae ones. Additionally, a high isoelectric point was predicted for Gsh2p, evidencing great differences in the protein structure.

Table 2. GSH1 and GSH2 comparisons between S. bacillaris strains and S. cerevisiae.
titre du tableau
Parameters GSH1 GSH2
Strain EC1118 FRI751 PAS13 EC1118 FRI751 PAS13
Length (a.a) 678 601 601 491 478 478
Mol. Weight (Da) 78253.60 68767.87 68767.87 55815.23 53662.36 53662.36
Isoelectric Point 5.87 5.30 5.30 5.50 6.85 6.85
Identity with S. cerevisiae (%) - 47.25 47.25 - 42.55 42.55
*GMQE with S. cerevisiae - 0.72 0.73 - 0.74 0.74
Numbers of INDELs by Swiss-Model - 26 26 - 20 20
Substrate binding GSH1 Amino acid sequence GSH2 Amino acid sequence
Region 1 (150 – 153) - - - VSVS VSVS VSVS
Region 2 (228 – 230) - - - ERN ETN ETN
Region 3 (285 – 479) - - - RTGY RAGY RAGY
Region 4 (478 – 479) - - - VA IA IA

Reduced length and molecular weight were found for both enzymes, however GSH2 displayed high isoelectric point and substitutions on its active site. *GMQE is the Swiss-Model global quality estimation and has a range of 0 to 1.

GSH production by S. bacillaris and S. cerevisiae in synthetic must

Glutathione is relevant in winemaking, due to its antioxidant property. With the aim of evaluating S. bacillaris potential contribution to glutathione production, a lab-scale fermentation trial was performed. GSH content was determined in yeast cells produced during a standard fermentation process that involved S. bacillaris as a starter. The industrial wine strain S. cerevisiae EC1118 was used as a control. The fermentations were run in MS300 synthetic must at 20 °C, mimicking winemaking conditions (Supplementary Figure 1). Due to the presence of GSH in natural musts (Lavigne et al., 2007), the synthetic must was used to evaluate GSH content produced exclusively by yeasts. All the fermentations were stopped when S. cerevisiae had completed the sugar transformation. As reported in the literature, S. bacillaris strains were not able to consume all the sugars (Magyar and Tóth, 2011; Bely et al., 2013; Lemos Junior et al., 2016; Lemos Junior et al., 2019). PAS13 and FRI751 showed 109.3 and 102.7 g/l of residual sugar respectively, and were confirmed to be fructophilic yeasts (Table 3a). Conversely, EC1118 fermentation showed no residual sugar.

Table 3. a) Concentrations of residual glucose and fructose, of the main fermentation products and thiol content at 618 hours of fermentation in MS300 synthetic must; b) Statistical analysis (Student’s t-test) results.
titre du tableau
a) Strain Glucose (g/L) Fructose (g/L) Glycerol (g/L) Acetic acid (g/L) Ethanol (%v/v) Dry cell weight (g) GSH (nmol/L) GSH content (µg/g) γGluCys content (µg/g) CysGly content (µg/g)
EC1118 0.00±0.00 0.00±0.00 5.77±0.14 0.38±0.07 13.16±0.02 0.30±0.01 709.42±60.46 73±6 84±16 10±2
FRI751 78.22±2.41 24.50±1.84 5.72±0.20 0.38±0.05 4.53±0.11 0.05±0.00 119.44±13.11 80±4 17±6 1±0
PAS13 84.56±2.11 23.88±1.31 7.81±0.37 0.40±0.02 4.94±0.14 0.05±0.00 207.60±5.87 122±3 30±84 2±0
b) Comparison Glucose Fructose Glycerol Acetic acid Ethanol Dry cell weight GSH GSH content γGluCys content CysGly content
EC1118 / PAS13 * * * ns * * * * * *
EC1118 / FRI751 * * ns ns * * * ns * *
FRI751 / PAS13 * ns * ns * ns * * * *

Levels of glutathione (GSH) and GSH, γ-glutamylcysteine (γGluCys) and cysteinyl-glycine (Cys-Gly) content (µg/g dry cell weight) from total cell extract derivatised and quantified by reverse-phase HPLC. Data are expressed as the average of three replicates ± standard deviations. * indicate significant differences between values (Student’s t-test, p < 0.05).

The biomass produced (dry cell weight) by S. bacillaris during fermentation was much lower than S. cerevisiae. This could be due to the remarkable difference in the cell size of the two species (2.2-3.0 x 3.0-5.2 µm S. bacillaris (Sipiczki, 2003), 3-8x5-10 µm S. cerevisiae (Vincenzini et al., 2005).

GSH is synthesised starting from glutamate and cysteine to form γ-glutamylcysteine (γ-GluCys), which is subsequently conjugated to glycine. Moreover, GSH is degraded to form cysteinyl-glycine. Therefore, the cell content of the two GSH-related molecules can give information about GSH turnover in the cell. At the end of fermentation, the concentration of the precursor, γGluCys, the catabolite cysteinyl-glycine and the GSH present in the yeast cells were all measured (Table 3a). Both S. bacillaris strains produced GSH at lower concentrations than S. cerevisiae. When the strain biomass was considered, the GSH cell content (mg/g dry cell weight) of S. bacillaris PAS13 was 67 % higher than EC1118, while the corresponding values for FRI751 were not significantly different from EC1118 (Table 3b). 61. At industrial-scale production, the cellular glutathione concentration of Candida utilis and S. cerevisiae strains, the most commonly used microorganisms for commercial glutathione production, is 0.1-1 % dry cell weight (Li et al., 2004). These concentrations are higher than those found in S. bacillaris strains PAS13 (0.012 %) and FRI751 (0.008 %), and S. cerevisiae EC1118 (0,007 %) in this study. In industrial production, however, growth conditions are optimised and they are very different from those of the oenological environment, mainly in terms of glucose and cysteine concentration in the medium. Interestingly, in the tested conditions, S. bacillaris PAS13 produces higher GSH concentrations than S. cerevisiae EC1118. These results are consistent with those of Gamero-Sandemetrio et al. (2018), who tested the GSH production of different non-Saccharomyces strains during the active dry yeast manufacturing process; among them, C. stellata (the previous taxonomic classification of S. bacillaris) was the only non-Saccharomyces species that produced higher GSH levels than S. cerevisiae. S. cerevisiae GSH precursor and catabolite contents were significantly higher than for S. bacillaris strains, suggesting a different GSH turnover. When the two S. bacillaris strains were compared, GSH and GSH-related molecules contents were higher in FRI751 compared to PAS13. Differences between strains were evidenced at a genomic level, involving genes related to GSH/GSSG balance and responsible for glutathione peroxidase, reductase and S-transferases activity.

Conclusions

In conclusion, the comparison of genomes makes it possible to gain a better understanding of the genes involved in the GSH metabolism pathway in S. bacillaris. Several orthologous genes were found to be involved in the synthesis, degradation and oxidation/reduction of GSH. No genes related to GSH plasma-membrane transport or vacuolar GS–X pump were identified. The gene comparison between both S. bacillaris strains revealed a significant number of SNPs present in four GSH-related genes, which directly impacted protein primary structure causing amino acid substitutions or deletions, which in turn can potentially impact protein function. These genes are mainly involved in a GSH/GSSG balance that contributes to modulating cell GSH content. The different GSH cell contents of the two S. bacillaris strains, determined at the end of fermentation, indicate a possible involvement of the genomic variations investigated in this study. Further studies are required to understand the role of these genomic variations in a large number of S. bacillaris strains. Moreover, in order to increase wine quality, S. bacillaris PAS13, which showed high GSH cell content, could be tested in mixed or sequential fermentations to evaluate the influence of high-producing strains in the GSH content of wines. Finally, our results can be considered as a starting point for strain selections focused on GSH production, in order to properly assess the efficacy of S. bacillaris as a source of glutathione.

 Acknowledgements

The authors wish to thank Anna Rita Trentin for her skillful assistance in the HPLC analysis.

The research financial support was provided by CAPES-Coordenação de Aperfeiçoamento de Pessoal de Nível Superior and MIUR ex 60 % (Ministero dell’Istruzione, dell’Università e della Ricerca) (Grant No. 60A08/3022/15).

 Avery A. M. , Avery S. V. , 2001. Saccharomyces cerevisiae expresses three phospholipid hydroperoxide glutathione peroxidases. Journal of Biological Chemistry, 276(36), 33730–33735. https://doi.org/10.1074/jbc.m105672200 . Bai M. , Zhou J. M. , Perrett S. , 2004. The yeast prion protein Ure2 shows glutathione peroxidase activity in both native and fibrillar forms. Journal of Biological Chemistry, 279(48), 50025–50030. https://doi.org/10.1074/jbc.m406612200 . Barreto L. , Garcerá A. , Jansson K. , Sunnerhagen P. , Herrero E. , 2006. A peroxisomal glutathione transferase of Saccharomyces cerevisiae is functionally related to sulfur amino acid metabolism. Eukaryotic Cell, 5(10), 1748–1759. https://doi.org/10.1128/ec.00216-06. Bely M. , Renault P. , Da Silva T. , Masneuf-Pomarède I. , Albertin.,W. Moine , V. Coulon , J. Sicard , D. De Vienne , D. & Marullo , P. , 2013. Non-conventional yeasts and alcohol levels reduction. In: P. L. Teissedre (Ed.), Alcohol Level Reduction in Wine (pp. 33-37). Villenave d'Ornon: Vigne et Vin Publications Internationales.. Bely M. , Sablayrolles J. M. , Barre P. , 1990. Description of Alcoholic Fermentation Kinetics: Its Variability and Significance. American Journal of Enology and Viticulture, 41, 319–324.. Biasini M. , Bienert S. , Waterhouse A. , Arnold K. , Studer G. , Schmidt T. , Kiefer F. , Cassarino T. G. , Bertoni M. , Bordoli L. , Schwede T. , 2014. SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information. Nucleic Acids Research, 42(W1), W252–W258. https://doi.org/10.1093/nar/gku340. Borneman A. R. , Forgan A. H. , Pretorius I. S. , Chambers P. J. , 2008. Comparative genome analysis of a Saccharomyces cerevisiae wine strain. FEMS Yeast Research, 8(7), 1185–1195. https://doi.org/10.1111/j.1567-1364.2008.00434.x . Bovo B. , Giacomini A. , Corich V. , 2011. Effects of grape marcs acidification treatment on the evolution of indigenous yeast populations during the production of grappa. Journal of Applied Microbiology, 111(2), 382–388. https://doi.org/10.1111/j.1365-2672.2011.05060.x. Bovo B. , Nadai C. , Vendramini C. , Fernandes Lemos Junior W. J. , Carlot M. , Skelin A. , Giacomini A. , Corich V. , 2016. Aptitude of Saccharomyces yeasts to ferment unripe grapes harvested during cluster thinning for reducing alcohol content of wine. International Journal of Food Microbiology, 236, 56–64. https://doi.org/10.1016/j.ijfoodmicro.2016.07.022. Cherry J. M. , Hong E. L. , Amundsen C. , Balakrishnan R. , Binkley G. , Chan E. T. , Christie K. R. , Costanzo M. C. , Dwight S. S. , Engel S. R. , Fisk D. G. , Hirschman J. E. , Hitz B. C. , Karra K. , Krieger C. J. , Miyasato S. R. , Nash R. S. , Park J. , Skrzypek M. S. , Simison M. , Weng S. , Wong E. D. , 2011. Saccharomyces Genome Database: the genomics resource of budding yeast. Nucleic Acids Research, 40(D1), D700–D705. https://doi.org/10.1093/nar/gkr1029. Ciani M. , Comitini F. , 2015. Yeast interactions in multi-starter wine fermentation. Current Opinion in Food Science, 1, 1–6. https://doi.org/10.1016/j.cofs.2014.07.001. Cingolani P. , Platts A. , Wang L. L. , Coon M. , Nguyen T. , Wang L. , Land S. J. , Lu X. , Ruden D. M. , 2012. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff. Fly, 6(2), 80–92. https://doi.org/10.4161/fly.19695. Coschigano P. W. , Magasanik B. , 1991. The URE2 gene product of Saccharomyces cerevisiae plays an important role in the cellular response to the nitrogen source and has homology to glutathione s-transferases. Molecular and Cellular Biology, 11(2), 822–832. https://doi.org/10.1128/mcb.11.2.822. Curtis D. S. , Phillips A. R. , Callister S. J. , Conlan S. , McCue L. A. , 2013. SPOCS: software for predicting and visualizing orthology/paralogy relationships among genomes. Bioinformatics, 29(20), 2641–2642. https://doi.org/10.1093/bioinformatics/btt454. Gamero-Sandemetrio E. , Payá-Tormo L. , Gómez-Pastor R. , Aranda A. , Matallana E. , 2018. Non-canonical regulation of glutathione and trehalose biosynthesis characterizes non-Saccharomyces wine yeasts with poor performance in active dry yeast production. Microbial Cell, 5(4), 184–197. https://doi.org/10.15698/mic2018.04.624. Grant C. M. , MacIver F. H. , Dawes I. W. , 1997. Glutathione synthetase is dispensable for growth under both normal and oxidative stress conditions in the yeast Saccharomyces cerevisiae due to an accumulation of the dipeptide gamma-glutamylcysteine. Molecular Biology of the Cell, 8(9), 1699–1707. https://doi.org/10.1091/mbc.8.9.1699. Inoue Y. , Maeta K. , Nomura W. , 2011. Glyoxalase system in yeasts: Structure, function, and physiology. Seminars in Cell & Developmental Biology, 22(3), 278–284. https://doi.org/10.1016/j.semcdb.2011.02.002. Ivit N. N. , Kemp B. , 2018. The impact of non-Saccharomyces yeast on traditional method sparkling Wine. Fermentation, 4(3), 73. https://doi.org/10.3390/fermentation4030073. Jolly N. P. , Varela C. , Pretorius I. S. , 2013. Not your ordinary yeast: non-Saccharomyces yeasts in wine production uncovered. FEMS Yeast Research, 14(2), 215–237. https://doi.org/10.1111/1567-1364.12111. Jones D. T. , Taylor W. R. , Thornton J. M. , 1992. The rapid generation of mutation data matrices from protein sequences. Bioinformatics, 8(3), 275-282. https://doi.org/10.1093/bioinformatics/8.3.275. Kaur H. , Ganguli D. , Bachhawat A. K. , 2012. Glutathione degradation by the alternative pathway (DUG pathway) in Saccharomyces cerevisiae is initiated by (Dug2p-Dug3p) 2 complex a novel glutamine amidotransferase (GATase) enzyme acting on glutathione. Journal of Biological Chemistry, 287(12), 8920–8931. https://doi.org/10.1074/jbc.m111.327411. Kiriyama K. , Hara K. Y. , Kondo A. , 2013. Oxidized glutathione fermentation using Saccharomyces cerevisiae engineered for glutathione metabolism. Applied Microbiology and Biotechnology, 97(16), 7399–7404. https://doi.org/10.1007/s00253-013-5074-8. Kumar C. , Igbaria A. , D'Autreaux B. , Planson A.-G. , Junot C. , Godat E. , Bachhawat A. K. , Delaunay-Moisan A. , Toledano M. B. , 2011. Glutathione revisited: a vital function in iron metabolism and ancillary role in thiol-redox control. The EMBO Journal, 30(10), 2044–2056. https://doi.org/10.1038/emboj.2011.105. Kumar C. , Sharma R. , Bachhawat A. K. , 2003. Investigations into the polymorphisms at the ECM38 locus of two widely used Saccharomyces cerevisiae S288C strains, YPH499 and BY4742. Yeast, 20(10), 857–863. https://doi.org/10.1002/yea.1012. Kumar S. , Stecher G. , Tamura K. , 2016. MEGA7: Molecular Evolutionary Genetics Analysis Version 7.0 for Bigger Datasets. Molecular Biology and Evolution, 33(7), 1870–1874. https://doi.org/10.1093/molbev/msw054. Lavigne V. , Pons A. , Dubourdieu D. , 2007. Assay of glutathione in must and wines using capillary electrophoresis and laser-induced fluorescence detection. Journal of Chromatography A, 1139(1), 130–135. https://doi.org/10.1016/j.chroma.2006.10.083. Lemos Junior W. J. F. , Nadai C. , Crepalde L. T. , de Oliveira V. S. , de Matos A. D. , Giacomini A. , Corich V. , 2019. Potential use of Starmerella bacillaris as fermentation starter for the production of low-alcohol beverages obtained from unripe grapes. International Journal of Food Microbiology, 303, 1–8. https://doi.org/10.1016/j.ijfoodmicro.2019.05.006. Lemos Junior W. J. F. , Bovo B. , Nadai C. , Crosato G. , Carlot M. , Favaron F. , Giacomini A. , Corich V. , 2016. Biocontrol ability and action mechanism of Starmerella bacillaris (synonym Candida zemplinina) isolated from wine musts against gray mold disease agent Botrytis cinerea on grape and their effects on alcoholic fermentation. Frontiers in Microbiology, 7. https://doi.org/10.3389/fmicb.2016.01249. Lemos Junior W. J. F. , da Silva Duarte V. , Treu L. , Campanaro S. , Nadai C. , Giacomini A. , Corich V. , 2018. Whole genome comparison of two Starmerella bacillaris strains with other wine yeasts uncovers genes involved in modulating important winemaking traits. FEMS Yeast Research, 18(7). https://doi.org/10.1093/femsyr/foy069. Lemos Junior W. J. F. , Treu L. , Duarte V. da S. , Campanaro S. , Nadai C. , Giacomini A. , Corich V. , 2017a. Draft genome sequence of the yeast Starmerella bacillaris (syn. Candida zemplinina) FRI751 isolated from fermenting must of dried Raboso grapes. Genome Announcements, 5(17). https://doi.org/10.1128/genomea.00224-17. Lemos Junior W. J. F. , Treu L. , da Silva Duarte V. , Carlot M. , Nadai C. , Campanaro S. , Giacomini A. , Corich V. , 2017b. Whole-Genome Sequence of Starmerella bacillaris PAS13, a Nonconventional Enological Yeast with Antifungal Activity. Genome Announcements, 5(32). https://doi.org/10.1128/genomea.00788-17. Li Y. , Wei G. , Chen J. , 2004. Glutathione: a review on biotechnological production. Applied Microbiology and Biotechnology, 66(3), 233–242. https://doi.org/10.1007/s00253-004-1751-y. Li H. , Handsaker B. , Wysoker A. , Fennell T. , Ruan J. , Homer N. , Marth G. , Abecasis G. , Durbin R. , 2009. The Sequence Alignment/Map format and SAMtools. Bioinformatics, 25(16), 2078–2079. https://doi.org/10.1093/bioinformatics/btp352. Lisowsky T. , 1993. A high copy number of yeast  y -glutamylcysteine synthetase suppresses a nuclear mutation affecting mitochondrial translation. Current Genetics, 23(5–6), 408–413. https://doi.org/10.1007/bf00312627. Lu S. C. , 2009. Regulation of glutathione synthesis. Molecular Aspects of Medicine, 30(1–2), 42–59. https://doi.org/10.1016/j.mam.2008.05.005. Ma X. , Jiang Y. , He Y. , Bao R. , Chen Y. , Zhou C. , 2009. Structures of yeast glutathione ‐ S ‐ transferase Gtt2 reveal a new catalytic type of GST family. EMBO Reports, 10(12), 1320–1326. https://doi.org/10.1038/embor.2009.216. Magyar I. , Tóth T. , 2011. Comparative evaluation of some oenological properties in wine strains of Candida stellata, Candida zemplinina, Saccharomyces uvarum and Saccharomyces cerevisiae. Food Microbiology, 28(1), 94-100. https://doi.org/10.1016/j.fm.2010.08.011. Masi A. , Ghisi R. , Ferretti M. , 2002. Measuring low-molecular-weight thiols by detecting the fluorescence of their SBD-derivatives: application to studies of diurnal and UV-B induced changes in Zea mays L. Journal of Plant Physiology, 159(5), 499–507. https://doi.org/10.1078/0176-1617-00655. Meijnen J.-P. , Randazzo P. , Foulquié-Moreno M. R. , van den Brink J. , Vandecruys P. , Stojiljkovic M. , Dumortier F. , Zalar P. , Boekhout T. , Gunde-Cimerman N. , Kokošar J. , Štajdohar M. , Curk T. , Petrovič U. , Thevelein J. M. , 2016. Polygenic analysis and targeted improvement of the complex trait of high acetic acid tolerance in the yeast Saccharomyces cerevisiae. Biotechnology for Biofuels, 9(1). https://doi.org/10.1186/s13068-015-0421-x. Mezzetti F. , De Vero L. , Giudici P. , 2014. Evolved Saccharomyces cerevisiae wine strains with enhanced glutathione production obtained by an evolution-based strategy. FEMS Yeast Research, 14(6), 977–987. https://doi.org/10.1111/1567-1364.12186. Morano K. A. , Grant C. M. , Moye-Rowley W. S. , 2011. The response to heat shock and oxidative stress in Saccharomyces cerevisiae. Genetics, 190(4), 1157–1195. https://doi.org/10.1534/genetics.111.128033. Murata K. , Kimura A. , 1990. Overproduction of glutathione and its derivatives by genetically engineered microbial cells. Biotechnology Advances, 8(1), 59–96. https://doi.org/10.1016/0734-9750(90)90005-v. Nadai C. , Fernandes Lemos W. J. , Favaron F. , Giacomini A. , Corich V. , 2018. Biocontrol activity of Starmerella bacillaris yeast against blue mold disease on apple fruit and its effect on cider fermentation. PLOS ONE, 13(9), e0204350. https://doi.org/10.1371/journal.pone.0204350. Oliveira M. , Duarte E. , 2014. Integrated approach to winery waste: waste generation and data consolidation. Frontiers of Environmental Science & Engineering, 10(1), 168–176. https://doi.org/10.1007/s11783-014-0693-6. Otero J. M. , Vongsangnak W. , Asadollahi M. A. , Olivares-Hernandes R. , Maury J. , Farinelli L. , Barlocher L. , Østerås M. , Schalk M. , Clark A. , Nielsen J. , 2010. Whole genome sequencing of Saccharomyces cerevisiae: from genotype to phenotype for improved metabolic engineering applications. BMC Genomics, 11(1). https://doi.org/10.1186/1471-2164-11-723. Penninckx M. , 2002. An overview on glutathione in versus non-conventional yeasts. FEMS Yeast Research, 2(3), 295–305. https://doi.org/10.1016/s1567-1356(02)00081-8. Rai R. , Cooper T. G. , 2005. In vivo specificity of Ure2 protection from heavy metal ion and oxidative cellular damage in Saccharomyces cerevisiae. Yeast, 22(5), 343–358. https://doi.org/10.1002/yea.1215. Reese M. G. , 2001. Application of a time-delay neural network to promoter annotation in the Drosophila melanogaster genome. Computers & Chemistry, 26(1), 51–56. https://doi.org/10.1016/s0097-8485(01)00099-7. Schacherer J. , Ruderfer D. M. , Gresham D. , Dolinski K. , Botstein D. , Kruglyak L. , 2007. Genome-wide analysis of nucleotide-level variation in commonly used Saccharomyces cerevisiae strains. PLoS ONE, 2(3), e322. https://doi.org/10.1371/journal.pone.0000322. Sipiczki M. , 2003. Candida zemplinina sp. nov., an osmotolerant and psychrotolerant yeast that ferments sweet botrytized wines. International Journal of Systematic and Evolutionary Microbiology, 53(6), 2079–2083. https://doi.org/10.1099/ijs.0.02649-0. Sugiyama K. , Izawa S. , Inoue Y. , 2000. The Yap1p-dependent induction of glutathione synthesis in heat shock response of Saccharomyces cerevisiae. Journal of Biological Chemistry, 275(20), 15535–15540. https://doi.org/10.1074/jbc.275.20.15535. Teixeira M. C. , Monteiro P. T. , Guerreiro J. F. , Gonçalves J. P. , Mira N. P. , dos Santos S. C. , Cabrito T. R. , Palma M. , Costa C. , Francisco A. P. , Madeira S. C. , Oliveira A. L. , Freitas A. T. , Sá-Correia I. , 2013. The YEASTRACT database: an upgraded information system for the analysis of gene and genomic transcription regulation in Saccharomyces cerevisiae. Nucleic Acids Research, 42(D1), D161–D166. https://doi.org/10.1093/nar/gkt1015. Ugliano M. , Kwiatkowski M. , Vidal S. , Capone D. , Siebert T. , Dieval J.-B. , Aagaard O. , Waters E. J. , 2011. Evolution of 3-mercaptohexanol hydrogen sulfide and methyl mercaptan during bottle storage of Sauvignon blanc wines. Effect of glutathione copper oxygen exposure and closure-derived oxygen. Journal of Agricultural and Food Chemistry, 59(6), 2564–2572. https://doi.org/10.1021/jf1043585. Vincenzini M. , Romano P. , Farris G. A. , 2005. Microbiologia del vino. Casa Editrice Ambrosiana, Milano, 487.. . Wang C. , Esteve-Zarzoso B. , Mas A. , 2014. Monitoring of Saccharomyces cerevisiae, Hanseniaspora uvarum, and Starmerella bacillaris (synonym Candida zemplinina) populations during alcoholic fermentation by fluorescence in situ hybridization. International Journal of Food Microbiology, 191, 1–9. https://doi.org/10.1016/j.ijfoodmicro.2014.08.014. Waterhouse A. M. , Procter J. B. , Martin D. M. A. , Clamp M. , Barton G. J. , 2009. Jalview Version 2--a multiple sequence alignment editor and analysis workbench. Bioinformatics, 25(9), 1189–1191. https://doi.org/10.1093/bioinformatics/btp033. Wheeler G. L. , Quinn K. A. , Perrone G. , Dawes I. W. , Grant C. M. , 2002. Glutathione regulates the expression of γ-glutamylcysteine synthetase via the Met4 transcription factor. Molecular Microbiology, 46(2), 545–556. https://doi.org/10.1046/j.1365-2958.2002.03174.x.