APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 1998, p Vol. 64, No. 10

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Oct. 1998, p Vol. 64, No /98/$ Copyright 1998, American Society for Microbiology. All Rights Reserved. Improvement of Nitrogen Assimilation and Fermentation Kinetics under Enological Conditions by Derepression of Alternative Nitrogen-Assimilatory Pathways in an Industrial Saccharomyces cerevisiae Strain JEAN-MICHEL SALMON* AND PIERRE BARRE Laboratoire de Microbiologie et de Technologie des Fermentations, Institut des Produits de la Vigne, Institut National de la Recherche Agronomique, Montpellier Cedex 1, France Received 1 December 1997/Accepted 17 July 1998 Metabolism of nitrogen compounds by yeasts affects the efficiency of wine fermentation. Ammonium ions, normally present in grape musts, reduce catabolic enzyme levels and transport activities for nonpreferred nitrogen sources. This nitrogen catabolite repression severely impairs the utilization of proline and arginine, both common nitrogen sources in grape juice that require the proline utilization pathway for their assimilation. We attempted to improve fermentation performance by genetic alteration of the regulation of nitrogen-assimilatory pathways in Saccharomyces cerevisiae. One mutant carrying a recessive allele of ure2 was isolated from an industrial S. cerevisiae strain. This mutation strongly deregulated the proline utilization pathway. Fermentation kinetics of this mutant were studied under enological conditions on simulated standard grape juices with various nitrogen levels. Mutant strains produced more biomass and exhibited a higher maximum CO 2 production rate than the wild type. These differences were primarily due to the derepression of amino acid utilization pathways. When low amounts of dissolved oxygen were added, the mutants could assimilate proline. Biomass yield and fermentation rate were consequently increased, and the duration of the fermentation was substantially shortened. S. cerevisiae strains lacking URE2 function could improve alcoholic fermentation of natural media where proline and other poorly assimilated amino acids are the major potential nitrogen source, as is the case for most fruit juices and grape musts. A wide variety of nitrogen-containing compounds are present in grape juice, depending upon the grape variety and time of harvest. During fermentation, these compounds are taken up during the first part of the Saccharomyces cerevisiae growth phase. Biosynthetic pools of amino acids are filled and the remaining nitrogenous compounds are utilized as nitrogen sources (17). Once pools are filled and growth begins, nitrogenous compounds are taken up and degraded in a specific order depending on environmental, physiological, and strain-specific factors (30, 32). Ammonium ions, which may constitute up to 10% of the total assimilable nitrogen in the must (26), reduce catabolic enzyme levels and transport activity for nonpreferred nitrogen sources through a phenomenon known as nitrogen catabolite repression (18). Nitrogen catabolite repression is attributed to the action of three proteins, GLN3, URE2, and GAP1 (36). The GLN3 and URE2 gene products are required for the transcription of many genes involved in alternative nitrogen-assimilatory pathways (22). GLN3 activates their transcription when preferred nitrogen sources are not available (38, 39), and URE2 represses their transcription when alternative nitrogen sources are not needed (20). GAP1, the general amino acid permease that transports all biological amino acids across the plasma membrane (28), is regulated at the transcriptional level by GLN3 and URE2 and is inactivated by dephosphorylation in the presence of glutamate and glutamine (48). Alternative nitrogen-assimilatory pathways are not expressed * Corresponding author. Mailing address: Laboratoire de Microbiologie et de Technologie des Fermentations, Institut des Produits de la Vigne, Institut National de la Recherche Agronomique, 2 place Viala, Montpellier Cedex 1, France. Phone: (33) Fax: (33) jmsalmon@ensam.inra.fr when ammonium is present. In grape juice, ammonium is the preferred nitrogen source. As ammonium is consumed, amino acids are taken up in a pattern determined by their concentration relative to cell needs for biosynthesis and to total nitrogen availability (40 42). Two exceptions are known: (i) proline is not taken up from grape juice under anaerobic fermentative conditions (27) and proline metabolism requires oxygen and a functioning electron transport chain to cleave the proline ring (51) and (ii) arginine and -aminobutyrate are usually taken up during the latter stages of fermentation under enological conditions and are always detectable in the final wine (9). Proline and arginine are the most common nitrogenous compounds in grape juice and represent 30 to 65% of the total amino acid content of grape juices (26). Both amino acids require the proline utilization pathway for conversion to glutamate and ammonia (12). Proline is transported into S. cerevisiae by the general amino acid permease and a proline-specific permease (product of PUT4 [36]). Proline is converted to glutamate in the mitochondria by proline oxidase (product of the PUT1 gene [51]) and 1 -pyrroline-5-carboxylate dehydrogenase (product of PUT2 [33]). The expression of the PUT genes is regulated by the PUT3 activator protein. This protein responds to the presence of proline in the and increases transcription of PUT1 and PUT2 genes (10, 13). URE2 represses transcription of the PUT genes and proline transporters under nitrogen-repressing conditions; the GLN3 protein has no effect on these genes (13, 53). The objective of our work was to isolate mutants of an industrial strain of S. cerevisiae that were no longer subject to nitrogen catabolite repression, while studying the fermentation kinetics of these mutants on simulated standard grape juice under enological conditions. The ultimate goal of this research

2 3832 SALMON AND BARRE APPL. ENVIRON. MICROBIOL. is to enhance the degradation of proline and other poorly assimilated amino acids during the growth phase and evaluate the potential impact of these physiological changes on yeast metabolism and fermentation kinetics. MATERIALS AND METHODS Strains, vectors, and culture conditions. Yeast strains. S. cerevisiae strains used in this study were V5 (MATa ura3) and A45 (MAT ). These two strains were derived from the same diploid industrial wine strain. Both strains exhibited identical fermentation kinetics under enological conditions. The ura3 genotype was introduced in the V5 strain at the haploid stage. This strain is preserved in the Collection Nationale de Cultures de Microorganisms (CNCM, Institut Pasteur, Paris, France) under reference no. I The A45 strain is preserved in our laboratory collection (Institut National de la Recherche Agronomique, Montpellier, France). Isogenic laboratory S. cerevisiae strains MYC1 (MAT ade2-1) and MYC2 (MATa ade2-1) were used as mating-type tester strains (J. Conde, Sevilla, Spain). Culture media. All media were heat sterilized (110 C, 30 min). The standard nutrient used for the general cultivation of yeast strains contained 1% yeast extract (Difco), 2% Bacto Peptone (Difco), and 2% glucose (YPD). Glucose-glutamine, glucose-proline, and glucose-ammonia liquid media contained 0.17% yeast nitrogen base (YNB) without amino acids and ammonium sulfate (Difco), 2% glucose, 0.002% uracil, and 0.1% glutamine, 0.1% proline, or 0.2% NH 4 H 2 PO 4, respectively. The synthetic fermentation media used in this study (symbolized as MSx in the text, where x represents the concentration of assimilable nitrogen [milligrams of N liter 1 ]) were simulated standard grape juices strongly buffered to ph 3.3 (5). These media contained the following ingredients (per liter): glucose, 200 g; citric acid, 6 g; DL-malic acid, 6 g; uracil, 20 mg; mineral salts (KH 2 PO 4, 750 mg; K 2 SO 4, 500 mg; MgSO 4 7H 2 O, 250 mg; CaCl 2 2H 2 O, 155 mg; NaCl, 200 mg; MnSO 4 H 2 O, 4 mg; ZnSO 4, 4 mg; CuSO 4 5H 2 O, 1 mg; KI, 1 mg; CoCl 2 6H 2 O, 0.4 mg; H 3 BO 3, 1 mg; NaMoO 4 2H 2 O, 1 mg); vitamins (myoinositol, 20 mg; nicotinic acid, 2 mg; calcium panthothenate, 1.5 mg; thiamin-hcl, 0.25 mg; pyridoxine-hcl, 0.25 mg; biotin, mg); anaerobic growth factors (ergosterol, 15 mg; sodium oleate, 5 mg; Tween 80, 0.5 ml); nitrogen source, 80 to 300 mg of N as ammoniacal nitrogen (18.6% NH 4 Cl); and amino acids (L-proline, 20.5%; L-glutamine, 16.9%; L-arginine, 12.5%; L-tryptophan, 6%; L-alanine, 4.9%; L-glutamic acid, 4%; L-serine, 2.6%; L-threonine, 2.6%; L-leucine, 1.6%; L-aspartic acid, 1.5%; L-valine, 1.5%; L-phenylalanine, 1.3%; L-isoleucine, 1.1%; L-histidine, 1.1%; L-methionine, 1.1%; L-tyrosine, 0.6%; L- glycine, 0.6%; L-lysine, 0.6%; and L-cysteine, 0.4%). The ammonium salts and -amino acids (all amino acids except proline) in the were considered assimilable nitrogen. For the proline degradation assay, the following was used: 0.17% YNB without amino acids and ammonium sulfate, 20% glucose, 0.002% uracil, 0.25% proline (0.3 g of N liter 1 ), and 0.009% (NH 4 ) 2 SO 4 (20 mg of N liter 1 ). Ammonium was provided at a low initial level to initiate cell growth. Growth conditions. For YPD and glucose-ammonia and glucoseproline liquid media, yeasts were inoculated at 10 6 cells ml 1 in 25-ml Erlenmeyer flasks containing 5 ml of liquid and incubated at 28 C on a rotary shaker. For MSx fermentation media, yeasts were precultured at 28 C in small fermentors (250 ml) with fermentation locks under discontinuous magnetic stirring (30 s every 5 min). Inoculation was standardized at 10 6 cells ml 1. Cells were harvested by centrifugation (500 g, 5 min), rinsed twice with sterile 0.9% (wt/vol) NaCl, and inoculated in culture. Yeast cultures were grown in fermentors (1.2 liters) with fermentation locks (CO 2 bubbling outlets filled with water). Fermentation media were normally deaerated by bubbling argon prior to inoculation (initial oxygen concentration, 1 mg liter 1 ). Filling conditions were controlled, and fermentations were carried out during anaerobiosis with continuous stirring under isothermal conditions (28 C). When oxygenation of the was required at the beginning of fermentation, the was oxygenated by adding different amounts of the same saturated with pure O 2 at 4 C. During fermentation, the fermentation was oxygenated by adding a synthetic solution saturated with pure O 2 at 4 C. This solution contained malic acid (6 g liter 1 ), citric acid (6 g liter 1 ), sugar, and ethanol at the same concentrations as in the fermenting. The addition of 6 mg of dissolved O 2 liter 1 by this method caused 10% dilution of the fermentation. Control fermentations involved 10% dilution with deaerated synthetic solution. For proline degradation assays, yeasts were inoculated at 10 6 cells ml 1 in 30-ml Erlenmeyer flasks containing 29 ml of argon-deaerated liquid containing 0.17% YNB without amino acids and ammonium sulfate), 20% glucose, 0.002% uracil, 0.25% proline (0.3 g of N liter 1 ) and 0.009% (NH 4 ) 2 SO 4 (20 mg of N liter 1 ). Ammonium was provided at a low initial level to initiate cell growth. Cultures were incubated for at least 48 h at 28 C without agitation. Oxygen diffusion in the was prevented by using bubbling CO 2 outlets. Genetic methods. (i) Mutagenesis and mutant selection. V5 cells were spread on plates containing 0.17% YNB without amino acids and ammonium sulfate, 20% glucose, 2% agar, 0.002% uracil, 0.25% proline (0.3 g of N liter 1 ), and 0.48% methylamine (1 g of N liter 1 ) at a cell density of 10 5 cells per plate. UV mutants were obtained by irradiating plates with UV (Philips UV-C 15 W; G15T8) for 40 s with a UV dose of 1,000 ergs mm 2. This dose killed 90 to 98% of the cells. V5 cells are impaired at methylamine concentrations above 0.24% (0.5 g of N liter 1 ). All media were adjusted to ph 6.5 with concentrated KOH prior to sterilization. (ii) Mating type. Mating type was determined by observing zygote formation after mixed inoculation of cells with tester strains MYC1 (MAT ) and MYC2 (MATa) on 2% agar-ypd plates. (iii) Sporulation. Approximately 10 7 cells were grown for 24 h on a plate of presporulation (1% yeast extract, 0.5% Bacto Peptone, 2% agar, and 10% glucose) and then spread on a plate of sporulation (1% yeast extract, 2% Bacto Peptone, 2% agar, and 1% potassium acetate) and incubated at 28 C. Sporulation efficiency was expressed as the ratio of asci to vegetative cells in a total population of at least 10 3 cells. (iv) Tetrad dissection. The ascus sac was digested with Helix pomatia gut juice (SHP; IBF-Sepracor) at 28 C for 20 min according to the method described by Johnston and Mortimer (31), and spores were separated with a micromanipulator. (v) Plasmid. Centromeric plasmid p1c-cs contained the URE2 gene inserted into the ClaI/SalI site of the Ycp50 plasmid (20). For this plasmid, V5 strain transformation was carried out on yeast spheroplasts (14). (vi) URE gene disruption. URE2 disruption was obtained by internal deletion of the open reading frame (ORF) by the method described by Wach et al. (50). A 1.4-kb PCR fragment containing a dominant resistance module, kanmx, was amplified by using plasmid pfa6-kanmx4 as template and two oligonucleotides, TTGTTTTAAGCTGCAAATTAAGTTGTACACCAAATGATGACGTACG CTGCAGGTCGAC and AAGCAGCCTTCATTCACCACGCAATGCCTTG ATGACCGCGGATGAATTCGAGCTCG, containing 18 and 16 nucleotides, respectively, homologous to the pfa6-kanmx4 multicloning site. In addition, these primers have 40-nucleotide extensions homologous to regions surrounding the start codon (nucleotides 33 to 6) or the stop codon (nucleotides 1034 to 1073) of the URE2 ORF. The PCR product was used directly to transform strain V5 by the lithium acetate method (47). Cells were incubated at 28 C in YPD for 14 h and plated on YPD containing 150 mg of G418 (Geneticin) liter 1. Correct replacement of the URE2 ORF by the kanmx4 module was checked by PCR with total genomic DNA and two oligonucleotides homologous to a region upstream of the start codon (nucleotides 88 to 72) or to that downstream of the stop codon (nucleotides 1114 to 1132) of the URE2 ORF and having the following sequences, respectively: ATCCCCCGTACGAA CTT and GCCTATATACATACCCTTA. PCR with transformants carrying a correctly integrated kanmx4 module gave a 1.5-kb fragment instead of the 1.2-kb fragment corresponding to the wild-type fragment. Analytical methods. (i) Cell counting. Cells were counted after sonication (30 s, 10 W) with an electronic Coulter Counter (model ZBI; Coulter Coultronics, Margency, France) fitted with a 100- m probe. (ii) Cellular dry weight. Cellular dry weight was obtained by filtering 10 ml of culture through membrane filters (pore size, 1.2 m). Filters were rinsed with the same amount of distilled water, and cells were dessicated at 108 C until a constant weight was obtained (24 h). (iii) Total cell protein. Total cell proteins were extracted as described by Jayamaran et al. (29). (iv) Protein determination. The protein concentration was determined with the bicinchoninic acid protein assay reagent (Pierce Chemicals, Rockford, Ill.), with crystalline bovine serum albumin as standard. (v) Determination of assimilable nitrogen in fermentation media. Ammonium and -amino acid concentrations were measured by enzymatic assay (8) and the TNBS (2,4,6-trinitrobenzenesulfonic acid) method (23), respectively. Proline concentrations in fermentation media were determined by the method of Yemm and Cocking (54). (vi) Determination of amino acid profiles in fermentation media. An aliquot of each fermentation (10 ml) was mixed with 50 ml of 96% (vol/vol) ethanol and allowed to stand for 48 h at 20 C to precipitate proteins and polysaccharides. After centrifugation (20,000 g, 20 min), the supernatant was dried under vacuum and resuspended in 0.2 N lithium citrate buffer (ph 2.2). Amino acids were separated by ion-exchange chromatography on an anionic Ultropac-8 lithium form resin (Pharmacia) with a Chromakon 400 analyzer (Kontron) and detected after reaction with ninhydrin (6, 7). (vii) Proline uptake experiments. We estimated high-affinity proline uptake by using the proline-specific permease (PUT4 gene product) and low-affinity proline uptake by using both the general amino acid and the proline-specific permeases (GAP1 and PUT4 gene products, respectively). The methodology described by Brandriss and Magasanik (11) was used. Since the affinities of these permeases for proline are very different (2.5 mm and 31 M, respectively), uptake was studied with L-[U- 14 C]proline (ICN Biomedicals, Oxfordshire, United Kingdom) at a final proline concentration of 10 mm (10 Ci mmol 1 )or100 M (500 Ci mmol 1 ) for studying low-affinity or high-affinity proline uptake activity, respectively. (viii) NAD-linked glutamate dehydrogenase assay. Yeast cell crude extracts were prepared by vortexing exponentially growing cells with glass beads as previously described (15). NAD-dependent glutamate dehydrogenase assays were performed as described by Miller and Magasanik (37). (ix) Fermentation kinetics. The amount of CO 2 released was determined by automatic measurement of fermentor weight loss every 20 min (45). Loss of ethanol and water by CO 2 stripping represented less than 2% of the total

3 VOL. 64, 1998 NITROGEN REPRESSION AND WINE FERMENTATION 3833 TABLE 1. Proline assimilation capacities of mutant strains under oxygen-limiting conditions Strain(s) a Final cell number b (10 7 ml 1 ) Residual proline concentration c (g liter 1 ) Nitrogen assimilation efficiency d (mg of N/10 12 cells) V UV Other mutant strains a V5, wild type; UV9, mutant strain. A total of five mutant strains were tested for data presented for other mutant strains. b Cells were counted after a 48-h incubation at 28 C without agitation on argon-deaerated liquid containing 0.17% YNB without amino acids and ammonium sulfate, 20% glucose, 0.002% uracil, 0.25% proline (0.3 g of N liter 1 ), and 0.009% (NH 4 ) 2 SO 4 (20 mg of N liter 1 ). c Residual proline concentration was determined by spectrophotometry at 460 nm after reaction with ninhydrin (residual ammonium was not detectable by enzymatic analysis). d Nitrogen assimilation efficiency was calculated by dividing the total amount of degraded nitrogen by the final cell number. fermentor weight loss. The CO 2 production rate was calculated by polynomial smoothing of the last 10 measurements of fermentor weight loss. The numerous acquisitions (one datum point every 20 min) and the precision of the fermentor weighing (0.01 g) allowed calculation of the CO 2 production rate with good precision (5). (x) Dissolved oxygen measurements. Dissolved oxygen measurement were routinely performed with a dissolved oxygen probe (OXI90 model; Wissenschaftlich Technische Werkstätten, Weilheim, Germany). RESULTS Isolation and characterization of an S. cerevisiae mutant strain V5 derepressed for proline utilization. Derepressed mutants were obtained after UV mutagenesis by selecting for resistance to the ammonium analog methylamine. Mutant strains were screened for their ability to grow on plates containing proline as the sole nitrogen source and methylamine at a repressive concentration. Except for one, all 40 isolated mutants exhibited clear derepression of amino acid utilization under nitrogen repression (data not shown). Most of these strains could not grow on plates with proline as the sole nitrogen source under oxygen-limiting conditions (Table 1). Only one mutant (UV9) could degrade proline under such conditions and increase biomass. UV9 cells had better nitrogen assimilation efficiency and therefore higher nitrogen contents than the other tested strains. This result suggests that UV9 was altered in its ability to regulate proline utilization. Proline uptake requires either the general amino acid permease or a specific proline permease. Since both systems are subject to nitrogen repression, we estimated the capacities of wild-type and UV9 mutant strains for proline uptake by both permeases TABLE 2. Proline uptake capacity of wild type and UV9 mutant a Strain High-affinity proline uptake (nmol/10 8 cells min 1 ) on: MS300 Glucose-proline Low-affinity proline uptake (nmol/10 8 cells min 1 ) on: MS300 Glucose-proline V UV a Yeast cells were harvested during the initial growth phase on MS300 or glucose-proline synthetic. Ammonium ions were still detectable in the MS300 fermentation at the time of harvest. High-affinity proline uptake (PUT4 permease activity) was determined at a proline concentration of 100 M (500 Ci mmol 1 ), and low-affinity proline uptake (both GAP1 and PUT4 permease activities) was determined at a proline concentration of 10 mm (10 Ci mmol 1 ). Mean values and standard errors of two different experiments are shown. FIG. 1. Amino acid composition of MS80 synthetic before (A) and after 28-h fermentation by wild-type V5 (B) and mutant UV9 (C) strains. At this harvest time, fermentation progress was for both strains. Mean values and standard errors of three different experiments are shown. Abbreviations: Gaba, -aminobutyrate; Ala, alanine; Asp, aspartate; Cys, cysteine; Etn, ethanolamine; Glu, glutamate; Gln, glutamine; Gly, glycine; His, histidine; Ile, isoleucine; Leu, leucine; Lys, lysine; Met, methionine; Orn, ornithine; Phe, phenylalanine; Ser, serine; Thr, threonine; Tyr, tyrosine; Val, valine; NH4, ammonium ions; Pro, proline; and Arg, arginine. under different culture conditions (Table 2). The UV9 strain had a higher capacity for high-affinity proline uptake (PUT4 function) than the wild type under nitrogen-repressing conditions (MS300 ). Control experiments performed under derepressing conditions (glucose-proline ) revealed a significant increase in proline-specific permease activity in the UV9 mutant. We also studied amino acid utilization by UV9 cells on simulated standard grape juice. Mutant and wild-type strains were inoculated at the same cell density on MS80 and MS300 synthetic media and harvested at the end of the growth phase. The amino acid composition was determined in both fermentation media (Fig. 1 and 2). In both cases, UV9 utilized more amino acids than the V5 strain, except for histidine, leucine, lysine, methionine, and threonine. Glutamate, which is one of the two end products of nitrogen catabolic pathways of yeasts (34), is excreted at low concentration into the fermentation by wild-type cells during the growth phase (52) but is not excreted by UV9. This result may indicate the presence of a strongly derepressed catabolic NAD-linked glutamate dehydrogenase (NAD-GDH) favoring the interconversion of glutamate into ammonia within UV9 cells (21, 37). We tested this hypothesis by measuring the level of NAD-GDH in both strains under strongly repressing conditions (glucose-glutamine ). We observed that the mutant strain exhibited higher NAD-GDH specific activity than the wild type under such conditions (490 8 nmol min 1 mg of protein 1 versus 59 6 nmol min 1 mg of protein 1, three determinations). In the absence of oxygen on these two fermentation media, the UV9 mutant always reached a higher final biomass than V5 cells (2.5 versus 2.3 g [dry weight] liter 1 and 5.2 versus 3.4 g [dry weight] liter 1 on MS80 and MS300 media, respectively). Characterization of the mutated URE2 allele. We crossed UV9 with the wild-type A45 strain. The diploids could not

4 3834 SALMON AND BARRE APPL. ENVIRON. MICROBIOL. FIG. 2. Amino acid composition of MS300 synthetic before (A) and after 28-h fermentation by wild-type V5 (B) and mutant UV9 (C) strains. At this harvest time, fermentation progress was and for V5 and UV9 strains, respectively. Mean values and standard errors of three different experiments are shown. Abbreviations are the same as those in the legend for Fig. 1. grow on plates containing proline as the sole nitrogen source and methylamine at a repressive concentration. We characterized these cells for their ability to use proline as the nitrogen source for growth under oxygen-limiting conditions (Table 3). Heterozygous diploids (initial cross A45 UV9: Z6 and Z14) reached the same biomass as the wild-type strain, indicating Strain Mating type Phenotype Sporulation TABLE 3. Genetic analysis of the mutation carried by the UV9 strain that the mutation is recessive. All 25 tetrads from this cross segregated 2 /2 for this character, indicating that a single mutation is responsible for this phenotype. Feedback crosses with the parental UV9 strain or direct crossing of haploids allowed us to construct homozygous diploids for the corresponding mutation (Table 3). Since ure2 mutant alleles (also known as usu and gdhcr) have been isolated in a number of screens designed to isolate mutants with increased amino acid permease activity (22) or genetic derepression of NAD-linked glutamate dehydrogenase (24), we also checked the identity of the isolated mutation as a recessive mutation in the URE2 gene. Strains with recessive URE2 gene mutations were previously characterized as possessing nitrogen catabolic enzymes insensitive to nitrogen catabolite repression (16), mainly in the pathways involved in glutamate, glutamine, arginine, allantoin, urea, -aminobutyrate, and proline assimilation (13). As previously described for ure2 mutants (53), the UV9 strain grew aerobically more slowly than the wild-type strain V5 on a glucose-ammonia (doubling time of 3.5 versus 2.6 h) or on a glucoseproline (doubling time of 8.9 versus 5.5 h). Similarly, exposure of the mutant strain to heat shock (45 C, 3 h) resulted in reduced recovery at 30 C compared with the V5 wild-type strain on YPD plates: survival rates were 13 and 49%, respectively. All diploids homozygous for the mutation failed to sporulate (Table 3). Very low sporulation efficiency is a characteristic of homozygous ure2/ure2 diploids (53). To confirm that the isolated mutation had occurred in ure2, the UV9 mutant was transformed with the centromeric plasmid p1c-cs containing URE2 (20). The resulting strain UV9/p1C-CS exhibited a phenotype analogous to that of the wild-type strain V5. Finally, in the wild-type strain V5, we disrupted URE2, leading to a URE2 null allele (mutant strain V5/ure2::kan). This mutant behaved like UV9 for most of its growth phenotypes (Table 3). Final cell population on containing proline a (10 7 cells ml 1 ) V5 a Ura (5) Wild type A45 Ura (3) Wild type UV9 a Ura (5) Mutant of V5 Z6 a/ Ura 2.6 (1) Initial cross of A45 and UV9 Z14 a/ Ura (2) Initial cross of A45 and UV9 Z6-5A Ura 1.8 (1) Segregant of Z6 Z6-5B Ura 3.6 (1) Segregant of Z6 Z6-5C a Ura 4.1 (1) Segregant of Z6 Z6-5D a Ura 1.1 (1) Segregant of Z6 Z6-11A Ura 4.4 (1) Segregant of Z6 Z6-11B a Ura 1.0 (1) Segregant of Z6 Z6-11C a Ura 1.3 (1) Segregant of Z6 Z6-11D Ura 4.1 (1) Segregant of Z6 Z14-9A Ura 1.2 (1) Segregant of Z14 Z14-9B a Ura 1.9 (1) Segregant of Z14 Z14-9C a Ura 4.1 (1) Segregant of Z14 Z14-9D Ura 4.4 (1) Segregant of Z14 UV9/Z14-9D a/ Ura 3.8 (1) Cross of UV9 and Z14-9D Z6-5D/Z14-9A a/ Ura 2.4 (1) Cross of Z6-5D and Z14-9A Z14-9C/Z6-11A a/ Ura 3.3 (1) Cross of Z14-9C and Z6-11A Z14-9D/Z6-11B a/ Ura 2.4 (1) Cross of Z14-9D and Z6-11B UV9/p1C-CS a Ura (5) UV9 strain carrying plasmid p1c-cs V5/ure2::kan a Ura (2) V5 strain carrying disruption of URE2 a Cells were counted after a 48-h incubation at 28 C without agitation on argon-deaerated liquid containing 0.17% YNB without amino acids and ammonium sulfate, 20% glucose, 0.002% uracil, 0.25% proline (0.3 g of N liter 1 ), and 0.009% (NH 4 ) 2 SO 4 (20 mg of N liter 1 ). Numbers in parentheses indicate the number of determinations for each strain. Source

5 VOL. 64, 1998 NITROGEN REPRESSION AND WINE FERMENTATION 3835 FIG. 3. Variations in the CO 2 production rate by wild-type V5 (F) and mutant UV9 (E) and V5/ure2::kan ( ) strains in MS300 culture at 28 C in the absence (A and B) or in the presence (C and D) of 6 mg of dissolved oxygen liter 1. The arrows indicate the time of dissolved oxygen addition for each strain. The CO 2 production rate patterns are represented as a function of fermentation progress (panels A and C) or of fermentation time (panels B and D). Final cell populations were , , and cells ml 1 in the absence of oxygen and , , and cells ml 1 in the presence of oxygen for V5, UV9, and V5/ure2::kan strains, respectively. Potential technological application of ure2 mutant strains in wine fermentations. Typical concentrations of total available nitrogen in real grape juices ranged from 50 to 800 mg of N liter 1, although assimilable nitrogen represents only 30 to 500 mg of N liter 1 (4, 5). An assimilable nitrogen concentration of 80 mg of N liter 1 is considered limiting for both growth and fermentation of industrial S. cerevisiae strains under enological conditions (5). In the absence of oxygen, both ure2 strains had a higher maximum CO 2 production rate and final biomass than V5 on two simulated standard grape juices, leading to quicker fermentation (about 100 instead of 115 h) (Fig. 3A and B and 4A and B). Under such conditions, regardless of initial must nitrogen content, assimilation of all nitrogen substrates was better in ure2 strains than in the wild type (Table 4). The increased level of amino acid permeases and derepression of amino acid utilization under ammoniacal nitrogen repression in ure2 strains may be solely responsible for these effects. We also tested the ability of ure2 mutants to degrade proline in simulated standard grape juices in the presence of oxygen. Air or oxygen diffusion during wine fermentation is a legal practice (23a); some authors have shown that oxygen addition improves the synthesis of anaerobic growth factors (ergosterol and unsaturated fatty acids [2, 3]) at the end of the cell growth Strain phase (43, 44). Nevertheless, in enological conditions, this oxygen requirement is low and is estimated at 5 to 10 mg liter 1 (43). We tested growth with increased initial dissolved oxygen concentrations in a nitrogen-limited simulated grape juice (MS80) containing anaerobic growth factors. Under these conditions, at oxygen concentrations equal to or above 6 mg liter 1, both ure2 strains utilized proline more efficiently than wild type, produced more biomass, and exhibited a higher maximum CO 2 production rate (Table 4). We also checked the effect of dissolved oxygen addition at the end of the cell growth phase on MS80 and MS300 media (Fig. 3 and 4). On MS80, oxygen addition had a slight effect on the maximum fermentation rate of UV9 and no effect on V5; this effect was not sufficient to reduce the fermentation duration (Fig. 4). This effect was amplified on MS300, where there were higher levels of assimilable nitrogen compounds and proline. V5 was weakly affected by oxygen addition (Fig. 3C and D), but ure2 strains produced more biomass and maintained a higher CO 2 production rate than V5 throughout the fermentation. Consequently, the fermentation duration decreased from 100 to only about 85 h. This effect of oxygen addition could be attributed primarily to deregulation of the proline utilization pathway in ure2 mutants. TABLE 4. Effect of initial oxygen addition on fermentation characteristics of wild-type V5 and UV9 and UV9/ure2::kan mutant strains in MS80 synthetic Initial oxygen concentration (mg liter 1 ) Maximum CO 2 production rate (g of CO 2 liter 1 h 1 ) Final cell population (10 8 cells ml 1 ) Assimilated nitrogen a (mg of N liter 1 ) Assimilated proline b (mg of N liter 1 ) Nitrogen assimilation efficiency c (mg of N/10 12 cells) V UV V5/ure2::kan a Ammonium salts and -amino acids (all amino acids except proline) in the were considered residual assimilable nitrogen (initial values: ammonium salts, 32 mg of N liter 1 ; -amino acids, 48 mg of N liter 1 ). b The residual proline concentration was determined by spectrophotometry at 460 nm after reaction with ninhydrin (initial value, 19.9 mg of N liter 1 ). c Nitrogen assimilation efficiency was calculated by dividing the total amount of degraded nitrogen by the final cell population.

6 3836 SALMON AND BARRE APPL. ENVIRON. MICROBIOL. FIG. 4. Variations in the CO 2 production rate by wild-type V5 (F) and mutant UV9 (E) strains in MS80 culture at 28 C in the absence (A and B) or presence (C and D) of 6 mg of dissolved oxygen liter 1. The arrows indicate the times of dissolved oxygen addition. The CO 2 production rate patterns are represented as a function of fermentation progress (panels A and C) or of fermentation time (panels B and D). Final cell populations were cells ml 1 for both strains in the absence of oxygen, and cells ml 1 and cells ml 1 for V5 and UV9 strains in the presence of oxygen, respectively. DISCUSSION Metabolism of nitrogen compounds by S. cerevisiae may govern the efficiency of alcoholic fermentation and affect the final product quality. Proline and arginine are the most abundant amino acids in fruit juices, but S. cerevisiae is not able to completely utilize these two amino acids during alcoholic fermentation. Derepression for the assimilation of amino acids in ure2 mutant strains of S. cerevisiae leads to better amino acid assimilation during alcoholic fermentation. Moreover, these strains can assimilate a significant amount of proline, especially following incorporation of small amounts of oxygen in the fermentation (as low as 6 mg of dissolved oxygen liter 1 ) at the end of the yeast growth phase. Cleavage of the proline ring requires oxygen and a functioning electron transport chain (51). As high hexose concentrations inhibit respiration by first closing mitochondrial voltage-dependent anionselective channels (1) and then repressing key enzymes in the respiratory chain (19), this observed proline degradation in ure2 strains under strong glucose-repressive conditions indicates that mitochondria retain their full potential for this degradation. Further research is needed to clarify the specific role of mitochondria under such conditions. The more efficient use of amino acids allowed ure2 strains to reach a higher final biomass and consequently to ferment natural media faster than wild-type cells. Thus, natural and industrial yeasts might be expected to lose URE2 repressor function during evolution. The much longer generation time of ure2 mutants on glucose-containing media could explain why selection has not favored spontaneous ure2 mutants. The ure2 mutants also were more sensitive to thermal stress than the corresponding wild type. This sensitivity might carry over to other stress situations, such as ethanol stress, but remains to be examined in detail. From a technological point of view, S. cerevisiae strains lacking URE2 function could improve alcoholic fermentation of natural media where proline and other poorly assimilated amino acids represent the major potential nitrogen source. The metabolism of nitrogen-containing compounds yields some end products of sensory importance for wine quality. For example, amino acids are deaminated catabolically to release their nitrogenous components and leave their carbon skeletons. This deamination step can result in the formation of -keto acids or of higher (fusel) alcohols. Further research is needed to identify the specific impact of ure2 strain fermentation on the overall flavor and aroma profile of wines. ACKNOWLEDGMENTS We thank M. J. Biron for technical assistance, especially with the genetic methods, E. Baptista for the construction of ure2 disruptants, and M. Pradal for assistance with the amino acid analyses. The plasmid p1c-cs was kindly provided by D. Rowen (Massachusetts Institute of Technology, Cambridge, Mass.). REFERENCES 1. Ahmadzadeh, M., A. Horng, and M. Colombini The control of mitochondrial respiration in yeasts: a possible role of the outer mitochondrial membrane. Cell Biochem. Funct. 14: Andreasen, A. A., and T. J. B. Stier Anaerobic nutrition of Saccharomyces cerevisiae. I. Ergosterol requirement for growth in a defined. J. Cell. Comp. Physiol. 41: Andreasen, A. A., and T. J. B. Stier Anaerobic nutrition of Saccharomyces cerevisiae. II. Unsaturated fatty acid requirement for growth in a defined. J. Cell. Comp. Physiol. 43: Bely, M., J. M. Sablayrolles, and P. Barre Automatic detection of assimilable nitrogen deficiencies during alcoholic fermentations in enological conditions. J. Ferment. Bioeng. 70: Bely, M., J. M. Sablayrolles, and P. Barre Description of alcoholic fermentation kinetics: its variability and significance. Am. J. Enol. Vitic. 40: Benson, J. V Multipurpose resins for analysis of amino acids and ninhydrin-positive compounds in hydrolysates and physiological fluids. Anal. Biochem. 50: Benson, J. V., M. J. Gordon, and J. A. Patterson Accelerated chromatographic analysis of amino acids in physiological fluids containing glutamine and asparagine. Anal. Biochem. 18: Bergmeyer, H. U., and H. O. Beutler Ammonia, p In H. U. Bergmeyer (ed.), Methods of enzymatic analysis. Metabolites. 3. Lipids, amino acids and related compounds, vol. 8, 3rd ed. VCH, Weinheim, Federal Republic of Germany. 9. Boulton, R. B., V. L. Singleton, L. F. Bisson, and R. E. Kunkee Nitrogen metabolism during fermentation, p In Principles and practices of winemaking. International Thomson Publishing, New York, N.Y. 10. Brandriss, M. C Proline utilization in Saccharomyces cerevisiae: analysis of the cloned PUT2 gene. Mol. Cell. Biol. 3: Brandriss, M. C., and B. Magasanik Genetics and physiology of proline utilization in Saccharomyces cerevisiae: enzyme induction by proline. J. Bacteriol. 140: Brandriss, M. C., and B. Magasanik Proline: an essential intermediate in arginine degradation in Saccharomyces cerevisiae. J. Bacteriol. 143: Brandriss, M. C., D. A. Falvey, S. A. G. des Etages, and S. Xu The roles of PUT3, URE2, and GLN3 regulatory proteins in the proline utilization pathway of Saccharomyces cerevisiae. Can. J. Bot. 73(Suppl. 1):S153 S Burgers, P. J. M., and K. J. Percival Transformation of yeast spheroplasts without cell fusion. Anal. Biochem. 163: Camarasa, C., S. Prieto, R. Ros, J. M. Salmon, and P. Barre Evidence for a selective and electroneutral K /H -exchange in Saccharomyces cerevisiae using plasma membrane vesicles. Yeast 12: Coffman, J. A., H. M. el Berry, and T. G. Cooper The URE2 protein regulates nitrogen catabolic gene expression through the GATAA-containing UAS NTR element in Saccharomyces cerevisiae. J. Bacteriol. 176: Cooper, T. G Nitrogen metabolism in Saccharomyces cerevisiae, p In J. N. Strathern, E. W. Jones, and J. B. Broach (ed.), The molecular biology of the yeast Saccharomyces. Metabolism and gene expression. Cold Spring Harbor Laboratory, New York, N.Y. 18. Cooper, T. G., and R. A. Sumrada What is the function of nitrogen catabolite repression in Saccharomyces cerevisiae? J. Bacteriol. 155: Correa Garcia, S., M. Bermudez Moretti, M. Cardalda, M. V. Rossetti, and A. M. del C. Batlle The role of ALA-S and ALA-D in regulating porphyrin biosynthesis in a normal and a HEM R mutant strain of Sac-

7 VOL. 64, 1998 NITROGEN REPRESSION AND WINE FERMENTATION 3837 charomyces cerevisiae. Yeast 9: Coschigano, P., and B. Magasanik 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. Mol. Cell. Biol. 11: Coschigano, P. W., S. M. Miller, and B. Magasanik Physiological and genetic analysis of the carbon regulation of the NAD-dependent glutamate dehydrogenase of Saccharomyces cerevisiae. Mol. Cell. Biol. 11: Courchesne, W. E., and B. Magasanik Regulation of nitrogen assimilation in Saccharomyces cerevisiae: roles of the URE2 and GLN3 genes. J. Bacteriol. 170: Crowell, E. A., C. S. Ough, and A. Bakalinsky Determination of alpha amino nitrogen in musts and wines by TNBS method. Am. J. Enol. Vitic. 36: a.European Economic Community Council regulation 822/87 of 16 March 1987 on the common organization of the market of wine. Official journal no. L084, annexe VI, point 3b, p Office for Official Publications of European Communities, Luxembourg, Luxembourg. 24. Grenson, M., E. Dubois, M. Piotrwska, R. Drillen, and M. Aigle Ammonia assimilation in Saccharomyces cerevisiae as mediated by the two glutamate dehydrogenases. Mol. Gen. Genet. 128: Grenson, M., C. Hou, and M. Crabeel Multiplicity of the amino acid permeases in Saccharomyces cerevisiae. IV. Evidence for a general amino acid permease. J. Bacteriol. 103: Henschke, P. A., and V. Jiranek Hydrogen sulfide formation during fermentation: effects of nitrogen composition in model grape musts, p In J. M. Rantz (ed.), International nitrogen symposium on grapes and wine. American Society for Enology and Viticulture, Davis, Calif. 27. Ingledew, W. M., C. A. Magnus, and F. W. Sosulski Influence of oxygen on proline utilization during the wine fermentation. Am. J. Enol. Vitic. 38: Jauniaux, J. C., and M. Grenson GAP1, the general amino acid permease gene of Saccharomyces cerevisiae. Eur. J. Biochem. 190: Jayamaran, J., J. C. Cotman, H. M. Mahler, and C. V. Sharp Biochemical correlation of respiration deficiency. VII. Glucose repression. Arch. Biochem. Biophys. 116: Jiranek, V., P. Langridge, and P. A. Henschke Nitrogen requirements of yeast during wine fermentation, p In P. J. Williams, D. M. Davidson, and T. H. Lee (ed.), Proceedings of the 7th Australian wine industry technical conference. Australian Industrial Publishers S.A., Adelaide, Australia. 31. Johnston, J. R., and R. K. Mortimer Use of snail digestive juice in isolation of yeast spore tetrads. J. Bacteriol. 78: Jones, M., and J. S. Pierce Adsorption of amino acids from wort by yeasts. J. Inst. Brew. 70: Krzywicki, K. A., and M. C. Brandriss Primary structure of the nuclear PUT2 gene involved in the mitochondrial pathway for proline utilization in Saccharomyces cerevisiae. Mol. Cell. Biol. 4: Large, P. J Degradation of organic nitrogen compounds by yeasts. Yeast 2: Lasko, P., and M. C. Brandriss Proline transport in Saccharomyces cerevisiae. J. Bacteriol. 148: Magasanik, B Regulation of nitrogen utilization, p In E. W. Jones, J. R. Pringle, and J. B. Broach (ed.), The molecular and cellular biology of the yeast Saccharomyces, vol. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 37. Miller, S. M., and B. Magasanik Role of NAD-linked glutamate dehydrogenase in nitrogen metabolism in Saccharomyces cerevisiae. J. Bacteriol. 172: Minehart, P. L., and B. Magasanik Sequence and expression of GLN3, a positive nitrogen regulatory gene of Saccharomyces cerevisiae encoding a protein with a putative zinc finger DNA-binding domain. Mol. Cell. Biol. 11: Mitchell, A. P., and B. Magasanik Regulation of glutamine-repressible gene products by the GLN3 function in Saccharomyces cerevisiae. Mol. Cell. Biol. 4: Monteiro, F. F., and L. F. Bisson Amino acids utilization and urea formation during vinification. Am. J. Enol. Vitic. 42: Monteiro, F. F., and L. F. Bisson Nitrogen supplementation of grape juice. I. Effect on amino acid utilization during fermentation. Am. J. Enol. Vitic. 43: Monteiro, F. F., and L. F. Bisson Nitrogen supplementation of grape juice. II. Effect on amino acid and urea release following fermentation. Am. J. Enol. Vitic. 43: Sablayrolles, J. M., and P. Barre Evaluation des besoins en oxygène de fermentations alcooliques en conditions oenologiques simulées. Sci. Alim. 6: Sablayrolles, J. M., P. Barre, and P. Grenier Design of a laboratory automatic system for studying alcoholic fermentations in anisothermal enological conditions. Biotechnol. Tech. 1: Sablayrolles, J. M., J. M. Salmon, and P. Barre Carences nutritionnelles des moûts. Efficacité des ajouts combinés d oxygène et d azote ammoniacal. Rev. Fr. Oenol. 159: Salmon, J. M., O. Vincent, J. C. Mauricio, M. Bely, and P. Barre Sugar transport inhibition and apparent loss of activity in Saccharomyces cerevisiae as a major limiting factor of enological fermentations. Am. J. Enol. Vitic. 44: Schiestl, R. H., and R. D. Gietz High efficiency transformation of intact cells using single stranded nucleic acid as carrier. Curr. Genet. 16: Stanbrough, M., and B. Magasanik Transcriptional and posttranslational regulation of the general amino acid permease of Saccharomyces cerevisiae. J. Bacteriol. 177: Stucka, R., S. Dequin, J. M. Salmon, and C. Gancedo DNA sequences in chromosomes II and VII code for pyruvate carboxylase isoenzymes in Saccharomyces cerevisiae: analysis of pyruvate carboxylase-deficient strains. Mol. Gen. Genet. 229: Wach, A., A. Brachat, R. Pöhlmann, and P. Philippsen New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 10: Wang, S. S., and M. C. Brandriss Proline utilization in Saccharomyces cerevisiae: sequence, regulation, and mitochondrial localization of the PUT1 gene product. Mol. Cell. Biol. 7: Watson, T. G Amino-acid pool composition of Saccharomyces cerevisiae as a function of growth rate and amino acid nitrogen source. J. Gen. Microbiol. 96: Xu, S., D. A. Falvey, and M. C. Brandriss Roles of URE2 and GLN3 in the proline utilization pathway in Saccharomyces cerevisiae. Mol. Cell. Biol. 15: Yemm, E. W., and E. C. Cocking The determination of amino acids with ninhydrin. Analyst 80: