OUP user menu

Two-dimensional protein map of an “ale”-brewing yeast strain: proteome dynamics during fermentation

Dominique Kobi, Sandra Zugmeyer, Serge Potier, Laurence Jaquet-Gutfreund
DOI: http://dx.doi.org/10.1016/j.femsyr.2004.07.004 213-230 First published online: 1 December 2004


The first protein map of an ale-fermenting yeast is presented in this paper: 205 spots corresponding to 133 different proteins were identified. Comparison of the proteome of this ale strain with a lager brewing yeast and the Saccharomyces cerevisiae strain S288c confirmed that this ale strain is much closer to S288c than the lager strain at the proteome level. The dynamics of the ale-brewing yeast proteome during production-scale fermentation was analysed at the beginning and end of the first and the third usage of the yeast (called generation in the brewing industry). During the first generation, most changes were related to the switch from aerobic propagation to anaerobic fermentation. Fewer changes were observed during the third generation but certain stress-response proteins such as Hsp26p, Ssa4p and Pnc1p exhibited constitutive expression in subsequent generations. The ale brewing yeast strain appears to be quite well adapted to fermentation conditions and stresses.

  • Brewing yeast
  • Fermentation
  • Proteome comparisons
  • Two-dimensional electrophoresis
  • Taxonomy

1 Introduction

During fermentation, brewing yeasts are exposed to severe environmental stresses such as variable temperature, high ethanol concentration, high sugar concentration (high-gravity brewing), nutrient limitation, oxygen shortage, and low pH. In general, any environmental factor which could have an adverse effect on cell growth is considered as a stress. A good deal of work on protein and/or mRNA expression has been carried out with laboratory or industrial strains [18], but only a few reports have been devoted to lager-brewing yeast transcription under fermentation conditions [9,10] and, to our knowledge, nothing has been published about ale brewing yeast protein expression. Moreover, “top-fermenting” ale and “bottom-fermenting” lager-brewing strains exhibit very different genomic, physiological and fermentation properties. Stress tolerance [40], aroma and flavour formation, fermentation time, reserve carbohydrate accumulation [34] and flocculation capacity [41] vary greatly. Most lager-brewing yeasts are allopolyploid, containing parts of at least two diverged genomes. The first is Saccharamyces cerevisiae, but the second has not yet been clarified despite many investigations focused on this second ancestor. Some homology was reported with S. monacensis[11,12], S. bayanus[13] or S. pastorianus[14,15]. Conversely, ale yeasts are closely related to the S. cerevisiae laboratory strain S288c [16]. Random Amplified PCR (RAPD) analysis has revealed similar patterns between S. cerevisiae strains and the ale yeasts but ale-brewing strains exhibit greater intraspecific heterogeneity than the lager yeasts [16,17]. Amplified fragment length polymorphism (AFLP) analysis of the laboratory and industrial strains also confirmed this pattern of relatedness: 93.7% shared fragments were found between the ale and laboratory strains, but only 74.6% between the lager strain and the same laboratory strains [18]. Two dimensional electrophoresis, a large-scale protein analysis technique, offers great potential for investigating protein expression in a given cell type or cell state and can also provide information on taxonomic relationships between Saccharomyces species [14]. Boucherie and coworkers [19] have pioneered proteomic investigations by 2-D electrophoresis on the budding yeast laboratory strain (S288c). They established a protein reference map with more than 350 identifications. Another protein reference map of a lager brewing strain was recently published by Joubert and coworkers [20] who identified more than 230 proteins.

In this work, we established the first 2-D map of an ale-brewing strain with 205 identified spots corresponding to 133 different proteins. This map will be very helpful for proteomic studies on ale-brewing yeasts. Two-dimensional electrophoresis enabled us to confirm the relatedness between the reference S. cerevisiae S288c and the ale-brewing strains and the discrepancies between these strains and the lager-brewing strain. The aim of this work was:

  • (i) first, to characterise physiological changes occurring during fermentation in the ale yeast, i.e. the adaptation to stressful conditions and the influence of propagation on the first fermentation in order to improve the brewing process,

  • (ii) second, to identify biological markers for the prediction of optimal fermentation progress.

Protein expression was studied at the beginning and the end of the first and the third generation of ale-brewing yeast production-scale fermentation. This study demonstrated great variations related to the switch from respiratory to fermentative metabolism occurring when yeast cells are transferred from propagation to fermentation conditions.

2 Materials and methods

2.1 Strains and culture conditions

Yeast strains used in this study were ale-brewing yeasts A12 and A38 and [14] lager-brewing strains, K3, K6 and K11 as well as S. cerevisiae S288c (Yeast genetics Stock Center, University of California, Berkeley, CA, USA: R.K. Mortimer) used as the reference strain for proteome map comparisons. All comparisons of 2-D maps between ale- and lager-brewing yeasts and the reference S. cerevisiae S288c strain were performed with cells grown in rich YPD medium at 28 °C with agitation (200 rpm). Samples were collected during the exponential growth phase (OD660 nm= 0.6). The A38 ale strain proteome was monitored during fermentation. The industrial fermentation process (Fig. 1) was performed in a 10-hl Tank OutDoor (TOD) pilot device. The A38 ale strain was grown in aerobic conditions with saccharose as the sole carbon source before pitching in wort for the first fermentation (called first generation in brewing terminology). As in the industrial process, the yeasts were harvested at the end of the fermentation and re-inoculated in fresh wort for a second generation. Up to five or six generations can be performed for lager yeasts, many more for ale strains. Proteomic investigations were carried out during the first and third generations. Samples collected at the start, one-quarter and end of both generations were noted F1, F2, F3 (first generation) and T1, T2, T3 (third generation), respectively. Degradation of wort sugars (maltose, glucose, fructose, saccharose and maltotriose) was determined by HPLC (Fig. 1B and C). The A38 ale strain reference map was established on the gel corresponding to the end of the first generation (F3). This gel was chosen as reference since its quality and resolution were the best among all gels performed.

Figure 1

(A) Diagram of the industrial fermentation process. (B) Sugar utilisation profile during the first generation. Sampling times for 2-D electrophoresis are indicated by F1, F2 and F3. (C) Sugar utilisation profile during the third generation. Sampling times for 2-D electrophoresis are indicated by T1, T2 and T3. (B) and (D) Strains were cultured as described in Section 2. (◻) – total sugars, (△) – maltose, (▪) – glucose, (▪) – fructose, (x) – saccharose, (+) – maltotriose.

2.2 Two-dimensional gel electrophoresis

Protein extractions were carried out as previously described by Boucherie and coworkers [21] from 5 × 107 cells pelleted after centrifugation. After separation, proteins were fixed and visualized using Silver, Coomassie Blue or Sypro Ruby staining. The image acquisition was scanned with UMAX Power LookII for the Silver and Coomassie Blue coloration and with Molecular Imager FX (BioRad) for the Sypro Ruby coloration.

2.3 Protein identification

Identifications were performed by MALDI-TOF analysis. Protein spots (randomly selected) were cut out from the gel, and trypsin-digested before mass spectrometry analysis as described by Joubert and coworkers [20]. Monoisotopic peptide masses obtained by MALDI-TOF analysis were used for database searches with the MS-Fit ProteinProspector package (http://prospector.ucsf.edu/ucsfhtml4.0/msfit.htm).

2.4 Quantitative gel analysis

Quantitative gel comparisons were performed with BioImage 2-D Analyser Software running on a SUN station. The comparisons were carried out with four Sypro Ruby-stained gels for each stage. In a first step, the software went through a detection and quantification process for each spot of each gel. In a second step, the four Sypro Ruby-stained gels of each stage were matched in a single image gel (called composite) by selecting 20 landmarks, allowing the software to match automatically additional spots. Each individual spot on the 2-D gels was normalized with respect to a normalization factor calculated from total valid spots found on the respective gel. Finally, two composites were also manually landmarked, automatically matched and normalized. The integrated intensity ratios were estimated to search for statistically significant differences in spot intensities between two composites. Ratios greater than 2 (+ or −) were considered significant.

2.5 Quantitative sugar analysis

Glucose, fructose, saccharose, maltose and maltotriose in fermentation samples were monitored by HPLC with a La Chrom system (Merck). Samples were injected onto a Shodex NH2P-504E column at 40 °C with acrylonitrile (70%) eluant (flow rate 1 ml min−1).

Resolved sugars were detected by a refractive index detector and quantified with a standard concentration scale.

3 Results and discussion

3.1 Two-dimensional protein map of an industrial ale-brewing yeast

The ale strain A38 was chosen as the reference for establishing a 2-D electrophoresis map of an ale-brewing strain (Fig. 2). The pattern was composed of 1200 polypeptide spots. The first dimension ranged from pH 3.8 to 6.8 and the second dimension from 15 to 180 kDa. Two hundred and fifty spots were selected among the most abundant proteins and excised from Coomassie-blue-stained gel to be identified by MALDI-TOF: 133 different proteins were identified (Table 1). One-third of the identified proteins are involved in amino acid and protein biosynthesis, whereas 23%, 12% and 7.5%, respectively, are involved in carbon metabolism, stress pathways and nucleic acid biosynthesis.

Figure 2

Two-dimensional protein reference map of the A38 ale-brewing yeast. The gel presenting the highest quality and resolution was chosen to report protein identifications. The pattern is composed of 1200 polypeptide spots. Spot names followed by an asterisk correspond to protein fragments; 205 spots corresponding to 133 different proteins were identified.

View this table:
Table 1

List of the 133 proteins identified by MALDI-TOF from the 2-D reference map of an ale-brewing yeast (Fig. 2)

No.pIMw (kDa)nameORF nameProtein namefunction
16.358.7ACH1YBL015WAcetyl-CoA hydrolaseLipid metabolism
26.275.5ACS2YLR153CAcetyl-CoA synthetase 2To Krebs cycle
45.186.1ADE5.7YGL234WPhosphoribosylamine-glycine ligase and phosphoribosylformylglycinamidine cyclo-ligaseAcid nucleic metabolism
56.165.3ADE16YLR028CAICAR transformylase/IMP cyclohydrolaseAcid nucleic metabolism
66.165.3ADE17YMR120CAICAR transformylase/IMP cyclohydrolaseAcid nucleic metabolism
76.736.9ADH1YOL086cAlcohol dehydrogenase 1Fermentation
8536.4ADO1YJR105WAdenosine kinaseAcid nucleic metabolism
95.4107ALA1YOR335CAlanyl-tRNA synthetaseProtein metabolism
106.356.7ALD4YOR374WAldehyde dehydrogenaseCarbon metabolism
115.354.4ALD6YPL061WAldehyde dehydrogenaseCarbon metabolism
125.738.9ARA1YBR149Wd-arabinose dehydrogenaseGlycolyse/Gluconeogenese
135.546.9ARG1YOL058WArgininosuccinate synthetaseAmino acid metabolism
145.756.2ARO8YGL202WAromatic amino acid aminotransferase IAmino acid metabolism
155.764.5ASN1YPR145WAsparagine synthetase 1Amino acid metabolism
165.554.8ATP2YJR121WF1-β ATP synthaseRespiration
175.834.8BEL1YMR116C40S small subunit ribosomal proteinProtein metabolism
184.830.1BMH1YER177WHomology with 14-3-3 proteinCytoskeleton
194.831.1BMH2YDR099WHomology with 14-3-3 proteinCytoskeleton
206.546.1CAR2YLR438WOrnithine aminotransferaseAmino acid metabolism
214.958.4CDC48YDR168WATPase of AAA familyCellular process
225.951.4CIT2YCR005CCitrate synthaseTo Krebs cycle
236.850.2COR1YBL045CUbiquinol cytochrome-c reductase core protein 1Respiration
246.142.5CYS3YAL012WCystathionine γ-LyaseAmino acid metabolism
256.356CYS4YGR155WCystathionine β-SynthaseAmino acid metabolism
266.465.3DLD1YDL174Cd-lactate dehydrogenaseTo Krebs cycle
276.455.2DLD3YEL071WHomology with d-lactate dehydrogenaseEnergy
284.322.7EFB1YAL003WTranslation elongation factor eEF1betaProtein metabolism
295.993.3EFT1YOR133WTranslation elongation factor eEF2Protein metabolism
306.246.8ENO1YGR254WEnolase 1Glycolyse/Gluconeogenese
315.746.9ENO2YHR174WEnolase 2Glycolyse/Gluconeogenese
325.543.4ERG6YML008CS-adenosyl-methionine delta-24-sterol-C-methyltransferaseLipid metabolism
335.539.6FBA1YKL060CFructose-bisphosphate aldolase IIGlycolyse/Gluconeogenese
345.567.4FRS1YLR060WPhenylalanyl-tRNA synthetase, alpha subunitProtein metabolism
355.657.5FRS2YFL022Cphenylalanine–tRNA ligase β chainProtein metabolism
368.553.1FUM1YPL262WFumarate hydrataseKrebs cycle
375.649.6GDH1YOR375CGlutamate dehydrogenase 1Amino acid metabolism
395.941.7GLN1YPR035WGlutamine synthaseAmino acid metabolism
406.253.5GND1YHR183W6-Phosphogluconate dehydrogenasePentose phosphate pathways
415.342.9GPD1YDL022WGlycerol-3-phosphate dehydrogenaseCellular process
426.130.4GPP1YIL053Wdl-glycerol phosphataseCarbon metabolism
435.827.8GPP2YER062Cdl-glycerol phosphataseCarbon metabolism
445.838.2GRE2YOL151WSimilarity to plant dihydroflavonol-4-reductasesStress
455.775.4GRS1YBR121CGlycyl-tRNA synthetaseProtein metabolism
466.124.8GSP1YLR293CGTP-binding protein of the ras superfamilyAcid nucleic metabolism
476.158.5GUA1YMR217WGMP synthetaseAcid nucleic metabolism
486.337.7HEM13YDR044WCoproporphyrinogen oxidaseProtein metabolism
494.880.9HSC82YMR186WHeat shock proteinStress
505.323.9HSP26YBR072WHeat shock proteinStress
514.881.4HSP82YPL240CHeat shock proteinStress
525.3102HSP104YLL026WHeat shock proteinStress
535.260.7HSP60YLR259CHeat shock proteinStress
545.253.7HXK1YFR053CHexokinase IGlycolyse/Gluconeogenese
555.253.9HXK2YGL253WHexokinase IIGlycolyse/Gluconeogenese
568.674.9ILV2YMR108WAcetolactate synthaseAmino acid metabolism
571113.6ILV5YLR255CCetol-acide reductoisomeraseAmino acid metabolism
585.432.3IPP1YBR011CInorganic pyrophosphatasePhosphate metabolism
594.874.5KAR2YJL034WHeat shock proteinProtein metabolism
606.7114KDG1YIL125W2-Oxoglutarate dehydrogenaseKrebs cycle
618.950.4KGD2YDR148C2-Oxoglutarate dehudrogenaseKrebs cycle
625.867.9KRS1YDR037WLysyl-tRNA synthetaseProtein metabolism
637.651.8LAT1YNL071WDihydrolipoamide S-acetyltransferaseTo Krebs cycle
645.538.9LEU2YCL018W3-Isopropylmalate dehydrogenaseAmino acid metabolism
65775.1LYS4YDR234WHomoaconitaseAmino acid metabolism
666.847.1LYS20YDL182WHomocitrate synthaseAmino acid metabolism
675.148.9LYS9YNR050CSaccharopine dehydrogenaseAmino acid metabolism
698.474.4MAE1YKL029CMalate dehydrogenaseCarbon metabolism
705.557.7MET3YJR010WATP-sulfurylaseAmino acid metabolism
716.185.9MET6YER091CHomocysteine methyltransferaseAmino acid metabolism
72648.7MET25YLR303WO-acetylhomoserine lyaseAmino acid metabolism
735.444.9OYE3YPL171CNADPH dehydrogenaseEnergy
745.764.3PAB1YER165WmRNA polyadenylate-binding proteinAcid nucleic metabolism
755.240.1PDB1YBR221CPyruvate dehydrogenaseTo Krebs cycle
765.861.5PDC1YLR044CPyruvate decarboxylase 1Fermentation
779.857.7PDH1YPR002WSimilarity to B. subtilis mmgE proteinUnknown
787.844.7PGK1YCR012WPhosphoglycerate kinaseGlycolyse/Gluconeogenese
805.969.6PRB1YEL060CProtease B, vacuolarProtein metabolism
81639.5PSA1YDL055CMannose-1-phosphate guanyltransferaseCarbon metabolism
826.455.2PYK2YOR347CPyruvate kinase 2Glycolyse/Gluconeogenese
838.223.3RIP1YEL024WUbiquinol-cytochrome C reductase iron–sulfur subunitRespiration
845.140.1RNR4YGR180CRibonucleotide reductase small subunitAcid nucleic metabolism
856.433.7RPL1YPL131W60S large subunit ribosomal proteinProtein metabolism
868.625RPS5YJR123WRibosomal protein S5.eProtein metabolism
875.849.1SAH1YER043CS-adenosyl-l-homocysteine hydrolaseAmino acid metabolism
88541.8SAM1YLR180WS-adenosylmethionine synthetase 1Amino acid metabolism
895.242.3SAM2YDR502CS-adenosylmethionine synthetase 2Amino acid metabolism
905.533SBP1YHL034CSingle-strand nucleic acid binding proteinAcid nucleic metabolism
915.928SCL1YGL011C20S proteasome subunit YC7ALPHA/Y8 (alpha1)Protein metabolism
925.129.1SEC53YFL045CPhosphomannomutaseProtein metabolism
936.143.4SER1YOR184WPhosphoserine transaminaseAmino acid metabolism
945.853.3SES1YDR023WSeryl-tRNA synthetaseProtein metabolism
955.615.9SOD1YJR104CCopper–zinc superoxide dismutaseCellular process
968.525.8SOD2YHR008CSuperoxide dismutaseCellular process
975.333.3SPE3YPR069CSpermidine synthaseAmino acid metabolism
98569.8SSA1YAL005CHeat shock protein of HSP70 familyStress
99569.5SSA2YLL024CHeat shock protein of HSP70 familyStress
100569.7SSA4YER103WHeat shock protein of HSP70 familyStress
1015.366.6SSB1YDL229WHeat shock protein of HSP70 familyStress
1025.466.6SSB2YNL209WHeat shock protein of HSP70 familyStress
1035.570.6SSC1YJR045CMitochondrial heat shock protein 70-related proteinStress
1045.177.4SSE1YPL106CHeat shock protein of HSP70 familyStress
1055.566.3STI1YOR027WStress-induced proteinStress
1066.535.8TDH2YJR009CGlyceraldehyde-3-phosphate dehydrogenaseGlycolyse/Gluconeogenese
1076.535.8TDH3YGR192CGlyceraldehyde-3-phosphate dehydrogenaseGlycolyse/Gluconeogenese
1086.124.3TFS1YLR178CCDC25-dependent nutrient- and ammonia-response cell-cycle regulatorCellular process
1095.557.5THR4YCR053WThreonine synthaseAmino acid metabolism
110544.7TIF1YKR059WTranslation initiation factor 4AProtein metabolism
1115.324.3TIF45YOL139CTranslation initiation factor eIF4EProtein metabolism
1125.726.8TPI1YDR050CTriose-phosphate isomeraseGlycolyse/Gluconeogenese
1135.756.1TPS1YBR126CTrehalose-6-phosphate synthaseGlycolyse/Gluconeogenese
1145.734.2TRR1YDR353WThioredoxine-dependante peroxide reductaseCellular process
115521.6TSA1YML028WThiol-specifique antioxidantStress
116519.1TSA2YLR109WAlkyl hydroperoxide reductaseStress
1186.248TUF1YOR187WTranslation elongation factor TUProtein metabolism
1196.352.9UGA1YGR019WButyrate transaminaseAmino acid metabolism
1205.834.8URA1YKL216WDihydroorotate dehydrogenaseAcid nucleic metabolism
1215.8118VMA1YDL185WH+-ATPase V1 domain 69 KD catalytic subunit, vacuolarCellular process
122557.7VMA2YBR127CH+-ATPase V1 domain 60 KD subunit, vacuolarCellular process
1235.935.6YDL124WYDL124WSimilarity to aldose reductasesUnknown
1245.325.7YDR533CYDR533CStrong similarity to hypothetical proteins YPL280w. YOR391c and YMR322cUnknown
1255.7116YEF3YLR249WTranslation elongation factor eEF3Protein metabolism
1265.719YER067WYER067WStrong similarity to hypothetical protein YIL057cUnknown
1275.452.9YFR044CYFR044CSimilarity to hypothetical protein YBR281cUnknown
1285.368.6YGR287CYGR287CStrong similarity to maltaseGlycolyse/Gluconeogenese
1295.944.6YHB1YGR234WFlavohemoglobineCellular process
1305.927.3YHR049WYHR049WSimilarity to S. pombe dihydrofolate reductase and YOR280cUnknown
1314.936.7YIL041WYIL041WSimilarity to S. pombe hypothetical proteinUnknown
1324.727.9YST2YLR048W40S ribosomal protein p40 homolog BProtein metabolism
1335.738.6YPR127WYPR127WSimilarity to C-term, of N. tabacum auxin-induced proteinUnknown
  • The Mw and pI of each protein are those given in the MIPS catalogue.

Two or more spots were identified for several proteins. For some, the only difference was a change in pI, generally due to post-translational modifications. The others displayed changes in both pI and molecular weight, which would probably correspond to protein fragments generated by vacuolar or proteasome degradation [22]. Moreover, about 20% of proteins present on the 2-D gels have been shown to be N-acetylated by an N-terminal acetyltransferase which leads to a shift in their pI position [19,23]. Other post-translational modifications, such as phosphorylation, glycosylation, lipidation and sulfation [2427], could be identified on the 2-D gel. As also observed by Larsen and coworkers [28], Eno2p was represented by several spots, probably corresponding to different C- and/or N-terminal processed forms. Finally, 21 identified spots appeared to be protein fragments since their molecular weight was smaller than that of the corresponding protein, whereas the covering percentage ranged between 14% and 50%. Most of them belong to the glycolytic pathway. But in general the shift in spot position, for most of the proteins, remains unexplained.

3.2 Qualitative comparison between proteomes of the ale- and lager-brewing yeasts, and the reference strain S. cerevisiae S288c

It has previously been reported that 2-D electrophoresis can be used to define the relatedness between yeast strains [14,29]. Therefore, 2-D protein patterns of three yeast strains, the ale A38 ‘top-fermenting’ brewing strain, the K11 ‘bottom-fermenting’ lager-brewing strain, and S. cerevisiae S288c were compared (Fig. 3). All were grown on rich medium, in respiro-fermentative conditions and harvested during the exponential phase. Whereas protein patterns of A38 and S288c appeared similar, the pattern of K11 differed from those of A38 and S288c. This is consistent with the fact that A38 has been described as a tetraploide of S. cerevisiae whereas K11 is a hybrid made up of at least two different genomes: one derived from S. cerevisiae and the other still not clearly identified. Fig. 4 presents detail of patterns of three different lager strains (A–C), S. cerevisiae S288c (D), and two different ale strains (E and F). Previous co-migration experiments between S. cerevisiae S288c and the K11 strains performed by Joubert and coworkers [14] have shown that spots a′, b′ and c′, identified by micro-sequencing as Pdc1p, Eno2p and Fba1, respectively, correspond to S. cerevisiae proteins. Spots a, b and c were, however, identified on K11 strains as isoforms of respectively a′, b′ and c′ and as belonging to S. pastorianus NRRL Y-1551. The pattern of the “second parental strain” appeared clearly on the 2-D gel of K11, but not on the A38 ale strain gel (Fig. 4F). Consequently, the A38 ale strain exhibited the same pattern as the S. cerevisiae strain. These observations can be extended to other ale and lager strains. For example, another ale strain called A12 (Fig. 4E) displayed a protein pattern similar to that of the A38 ale strain (Fig. 4F). Protein patterns of the lager yeasts K3 (Fig. 4B), K6 (Fig. 4C) and K11 (Fig. 4A) also displayed common patterns.

Figure 3

Comparison of the 2-D gel electrophoresis pattern between an ale brewing strain, a lager-brewing strain and a laboratory strain. Sypro Ruby protein staining. (A) A38 ale-brewing yeast, (B) S. cerevisiae S288c reference strain and (C) K11 lager-brewing yeast. a′, b′, c′ and their isoforms a, b, c were identified as Pdc1p, Eno2p and Fba1p, respectively.

Figure 4

Two-dimensional gel electrophoresis patterns of ale- and lager-brewing yeasts and of a laboratory strain. Proteins were revealed by silver staining. (A) K11 lager- brewing yeast, (B) K3 lager-brewing yeast, (C) K6 lager-brewing yeast, (D) S. cerevisiae S288c reference strain, (E) A12 ale-brewing yeast and (F) A38 ale-brewing yeast. a′, b′, c′ and their isoforms a, b, c were identified as Pdc1p, Eno2p and Fba1p, respectively.

Protein identifications performed for S288c and A38 reference maps confirmed that their genomes are closely related. However, this does not mean that these strains are identical, as the comparison was performed only on a detail part of the map. When looking at the entire gel for both strains, some discrepancies appear. For example, Adh4p is present in the brewing strain but not in S288c (Fig. 3). Adh4 expression seems to be specific for the brewing strain in such conditions. To summarize, 2-D electrophoresis is a powerful technology for confirming genetic background and distinguishing between ale- and lager-brewing strains, and for monitoring protein expression during a fermentation process.

3.3 Protein expression in ale strain A38 during fermentation

After aerobic propagation, yeasts were inoculated in wort for the first generation. During that generation, cells replicated two to four times. Yeasts were then harvested and re-used for inoculating the next generation (Fig. 1). Two-dimensional gels were obtained with protein extracts collected at the beginning, one quarter and the end of the first generation and at the end of the third re-used cell generation. Fig. 1B and C indicate the sampling times on sugar consumption curves (F1, F2, F3 and T1, T2, T3). Propagation for the first generation was carried out under aerobic conditions and on synthetic medium containing saccharose as the sole carbon source. Qualitative 2-D gel comparisons showed few variations between the samples harvested at the beginning (F1, T1) and at the first quarter (F2, T2) of the generations. Whereas samples collected at the end of generations (T2/T3), compared to the first quarter samples (F2/F3), present similar protein variations as those observed in the comparison between the beginning and the end of the generations (F1/F3, T1/T3). As no supplementary information was presented on the F2 and T2 step gels, quantitative analysis was only performed on gels corresponding to the beginning (F1,T1) and the end (F3,T3) stages of the first and third fermentations. Ratios greater than 2 (+ or −) were considered significant. All spots that exhibited significant intensity differences were subjected to MALDI-TOF identification. Some spots with weak intensity gave few peptide mass peaks, which were insufficient for unambiguous identification of the protein. However, the identification rate reached 70% and revealed that the first generation displayed the greatest number of changes, with 85 significant changes in comparison with 27 during the third generation. Results are reported in Table 2 and Fig. 5A for the first generation, and Table 3 and Fig. 5B for the third generation.

View this table:
Table 2

List of decreased or increased quantities of proteins from the comparison of the beginning (F1) and the end (F3) of the first generation

No.pIMwORF nameGene or ORF name% CoverageaFactorName or functionGroup functions
1569.7YER103wSSA429PresentHeat shock protein of HSP70 family, cytosolicStress
24.874.5YJL034wKAR244PresentNuclear fusion proteinStress
46.337.7YDR044wHEM1329PresentCoproporphyrinogen III oxidaseProtein biosynthesis
95.825YGL037cPNC137PresentSimilarity to PIR:B70386 pyrazinamidase/nicotinamidase - Aquifex aeolicusUnknown
115.323.9YBR072wHSP26∗394.4Heat shock proteinStress
135.738.6YPR127wYPR127w363.79Similarity to C-term, of N.tabacum auxin-induced proteinUnknown
146.337.7YDR044wHEM13293.75Coproporphyrinogen III oxidaseProtein biosynthesis
155.325.7YDR533cYDR533c293.47Hypothetical proteins YPL280w, YOR391c and YMR322cUnknown
165.861.5YLR044cPDC1∗233.44Pyruvate decarboxylase, isozyme 1Carbon metabolism
176.124.4YLR178cTFS1303.37Nutrient- and ammonia-response cell-cycle regulatorUnknown
215.323.9YBR072wHSP26392.78Heat shock proteinStress
22YGR254w/YHR174wENO∗18/212.73Enolase 1 or enolase 2Carbon metabolism
247.144.7YCR012wPGK1∗282.6Phosphoglycerate kinaseCarbon metabolism
275.539.6YKL060cFBA1∗172.41Fructose-bisphosphate aldolaseCarbon metabolism
286.455.2YOR347cPYK2∗142.24Pyruvate kinase, glucose-repressed isoformCarbon metabolism
306.546.1YLR438wCAR2352.17Ornithine aminotransferaseAmino acid biosynthesis
31YGR254w /YHR174wENO50/502.168Enolase 1 or enolase 2Carbon metabolism
345.861.5YLR044cPDC1∗62.09Pyruvate decarboxylase, isozyme 1Carbon metabolism
365.570.6YJR045cSSC130−2.01Mitochondrial heat shock protein 70-related proteinStress
384.936.7YIL041wYIL041w26−2.09Similarity to S. pombe hypothetical proteinUnknown
396.7114YIL125wKGD122−2.12-oxoglutarate dehydrogenase complex E1 componentUnknown
405.242.3YDR502cSAM231−2.1S-adenosylmethionine synthetase 2Amino acid biosynthesis
418.553.2YPL262wFUM110−2.11Fumarate hydrataseCarbon metabolism
435.738.9YBR149wARA150−2.17d-arabinose dehydrogenase, large subunitCarbon metabolism
446.248YOR187wTUF143−2.2Translation elongation factor TU, mitochondrialProtein biosynthesis
456.158.5YMR217wGUA122−2.22GMP synthase (glutamine-hydrolyzing)Nucleotides biosynthesis
465.543.4YML008cERG616−2.22S-adenosyl-methionine delta-24-sterol-c-methyltransferaseLipid biosynthesis
485.855.4YCL040wGLK128−2.26Aldohexose specific glucokinaseCarbon metabolism
495.333.3YPR069cSPE38−2.3Putrescine aminopropyltransferase (spermidine synthase)Amino acid biosynthesis
505.554.8YJR121wATP236−2.32F1F0-ATPase complex, F1 β subunitRespiration
526.535.7YGR192cTDH356−2.4Glyceraldehyde-3-phosphate dehydrogenase 3Carbon metabolism
536.236.8YOL086cADH140−2.41Alcohol dehydrogenase ICarbon metabolism
54648.7YLR303wMET1741−2.42O-acetylhomoserine sulfhydrylaseAmino acid biosynthesis
55536.4YJR105wADO139−2.48Strong similarity to human adenosine kinaseNucleotides biosynthesis
585.260.8YLR259cHSP6034−2.53Heat shock protein – chaperone, mitochondrialStress
596.236.8YOL086cADH140−2.56Alcohol dehydrogenase ICarbon metabolism
606.236.8YOL086cADH140−2.6Alcohol dehydrogenase ICarbon metabolism
615.570.6YJR045cSSC130−2.62Mitochondrial heat shock protein 70-related proteinStress
626.356.7YOR374wALD436−2.62Aldehyde dehydrogenase, mitochondrialCarbon metabolism
639.144.4YLR355cILV547−2.63Ketol-acid reducto-isomeraseAmino acid biosynthesis
645.546.9YOL058wARG114−2.74Argininosuccinate synthetaseAmino acid biosynthesis
655.543.4YML008cERG616−2.87S-adenosyl-methionine delta-24-sterol-c-methyltransferaseLipid biosynthesis
665.570.6YJR045cSSC130−2.88Mitochondrial heat shock protein 70-related proteinStress
678.950.4YDR148cKGD222−3.112-oxoglutarate dehydrogenase complex E2 componentCarbon metabolism
685.368.6YGR287cYGR287c32−3.22Strong similarity to maltaseUnknown
695.353.7YFR053cHXK112−3.36Hexokinase ICarbon metabolism
706.185.9YER091cMET638−3.49Homocysteine methyltransferaseAmino acid biosynthesis
716.185.9YER091cMET6∗38−3.5Homocysteine methyltransferaseAmino acid biosynthesis
725.253.9YGL253wHXK233−3.57Hexokinase IICarbon metabolism
738.223.4YEL024wRIP126−3.59Ubiquinol–cytochrome-c reductase iron-sulfur protein precursorRespiration
745.353.7YFR053cHXK112−3.64Hexokinase ICarbon metabolism
755.554.8YJR121wATP236−3.67F1F0-ATPase complex, F1 β subunitRespiration
765.242.3YDR502cSAM231−4.09S-adenosylmethionine synthetase 2Amino acid biosynthesis
775.554.8YJR121wATP236−4.1F1F0-ATPase complex, F1 β subunitRespiration
78541.8YLR180wSAM130−4.15S-adenosylmethionine synthetase 1Amino acid biosynthesis
795.354.4YPL061wALD617−4.73Aldehyde dehydrogenase, cytosolicCarbon metabolism
806.850.2YBL045cCOR139−5.09Ubiquinol—cytochrome-c reductase 44 K core proteinRespiration
815.951.4YCR005cCIT218−5.37Citrate (si)-synthase, peroxisomal/Ubiquinol–cytochrome-c reductase 44K core proteinRespiration
836.358.7YBL015wACH146−6.26Acetyl-CoA hydrolaseCarbon metabolism
845.868YDR037wKRS116−10.56Lysyl-tRNA synthetase, cytosolicProtein biosynthesis
855.242.3YDR502cSAM231AbsentS-adenosylmethionine synthetase 2Amino acid biosynthesis
  • “Present” as factor indication indicates that the spot was observed only on the end of fermentation. Inversely “absent” as factor indication indicates that the spot was missing at the end of fermentation.

  • a Percentage of matched peptide covering the protein obtained after spot identification by mass spectrometry and MS-Fit questioning. Spot identifications with percentage coverage below 10% were confirmed by Q-TOF analysis. The Mw and pI of each protein are those given in the MIPS catalogue. NI indicates that a spot was not identified by MALDI-TOF. These proteins are reported by their numbers in Fig. 5A.

Figure 5

Comparison between the beginning and the end of the first and the third generations. Proteins were revealed by Sypro Ruby staining and quantitatively compared by BioImage software. Only proteins of variable expression are annotated. Spot intensities increasing more than 2-fold are underlined whereas intensities decreasing less than 2-fold are not. Protein names followed by an asterisk designate fragments. Spots indicated by a number were also variable but not identified by MALDI-TOF. (A) First fermentation: 2-D gel electrophoresis comparison between (a) the beginning (F1) and (b) the end (F3). (B) Third fermentation: 2-D gel electrophoresis comparison between (a) the beginning (T1) and (b) the end (T3).

View this table:
Table 3

List of decreased or increased quantities of proteins from the comparison of the beginning (T1) and the end (T3) of the third generation

No.pIMw (kDa)ORF nameGene or ORF name% CoverageaFactorName or functionGroup functions
15.849.1YER043cSAH1245.5S-adenosyl-l-homocysteine hydrolaseAmino acid biosynthesis
2569.8YAL005cSSA1434.96Heat shock protein of HSP70 family, cytosolicStress
39.257.7YPR002wPDH1∗433.77Similarity to B. subtilis mmgE proteinCarbon metabolism
45.554.8YJR121wATP2∗233.53F1F0-ATPase complex, F1 β subunitRespiration
54.874.5YJL034wKAR2443.35Nuclear fusion proteinStress
65.539.6YKL060cFBA1∗172.78Fructose-bisphosphate aldolaseCarbon metabolism
75.861.5YLR044cPDC1262.65Pyruvate decarboxylase, isozyme 1Carbon metabolism
106.236.8YOL086cADH1262.53Alcohol dehydrogenase ICarbon metabolism
11569.8YAL005cSSA1282.48Heat shock protein of HSP70 family, cytosolicStress
126.253.5YHR183wGND1342.456-phosphogluconate dehydrogenaseCarbon metabolism
165.444.9YPL171cOYE3112.35NAPDH dehydrogenase (old yellow enzyme). isoform 3Energy
176.185.9YER091cMET6382.32Homocysteine methyltransferaseAmino acid biosynthesis
185.738.6YPR127wYPR127w362.29Similarity to C-term, of N. tabacum auxin-induced proteinUnknown
205.557.7YJR010wMET3142.17Sulfate adenylyltransferaseAmino acid biosynthesis
225.7115.9YLR249wYEF311−2.3Translation elongation factor eEF3Protein biosynthesis
235.354.4YPL061wALD617−2.32Aldehyde dehydrogenase, cytosolicCarbon metabolism
255.366.6YDL229wSSB159−2.83Heat shock protein of HSP70 familyStress
  • a Percentage of matched peptide covering the protein obtained after spot identification by mass spectrometry and MS-Fit questioning. The Mw and pI of each protein are those given in the MIPS catalogue. NI indicates that a spot was not identified by MALDI-TOF. These proteins are reported by their numbers in Fig. 5B.

3.3.1 First generation (Fig. 5A)

A comparison between the beginning (a) and the end (b) of the first generation shows that the most significant changes concerned proteins involved in carbohydrate metabolism, respiration, and amino acid and protein biosynthesis.

On one hand, 50 of the 85 differentially expressed proteins during the first fermentation were repressed (Fig. 5A; a and b; Table 2). Most glycolytic enzymes, proteins involved in acetyl-CoA formation, proteins of the tricarboxylic acid cycle, but also proteins involved in respiration were down regulated. The abundance of other proteins involved in amino acid, nucleotide, or lipid synthesis decreased. On the other hand, most of the proteins induced were fragments belonging to either carbon metabolism, or protein or amino acid biosynthesis pathways.

Most of these drastic changes in protein expression reveal an adaptive response to anaerobic conditions. This is not surprising as yeasts had been propagated in aerobic conditions. Indeed, the down-regulation of carbohydrate metabolism, and in particular the decreased expression of the acetyl-CoA pathway and some mitochondrial proteins, directly correlates with a switch from respiratory to fermentative metabolism. The main metabolic pathways affected during this switch are represented in Fig. 6.

Figure 6

Metabolic pathways affected by aerobic-to-anaerobic switch during the first generation. Protein names followed by an asterisk designate fragments. Impact of aerobic-to-anaerobic switch on glycolysis and acetyl-CoA pathways

Glycolysis was only moderately affected, since the only significant change was the repression of Glk1p, Hxk1p, Hxk2p and Tdh3p. However, the switch from respiratory to fermentative metabolism was accompanied by an increase of glycolytic protein fragments. Trabalzini and coworkers [22] have reported that most of the protein fragments which appear after glucose exhaustion, in a wine strain grown on YPD, are specific isoforms of glycolytic enzymes. This suggests that intracellular proteolysis has an impact on the regulation of the abundance of these isoenzymes. Consequently, the proteolysis observed here could be related to the stress caused by the oxidative-to-fermentative switch. There was, however, a clear inhibition of acetyl-CoA metabolism, especially of the pyruvate dehydrogenase bypass (PDH bypass). Pyruvate is processed into acetyl-CoA via three pathways: the pyruvate dehydrogenase complex (PDH complex), and the cytosolic and the mitochondrial pyruvate dehydrogenase bypasses. The PDH complex was represented by Pdb1p (pyruvate dehydrogenase) on our gel but its decrease was not significant (−1.4-fold). In contrast, expression of Ald6p and Ald4p, involved respectively in cytoplasmic and mitochondrial PDH bypasses, was strongly inhibited.

Ald6p is thought to be involved in the production of acetate leading to acetyl-CoA synthesis. Acetyl-CoA is then supplied principally to biosynthetic pathways, in particular for lipid synthesis and in smaller amounts for the Krebs cycle [30]. Ald4p has recently been identified as a mitochondrial PDH bypass [31] leading to acetate production in mitochondria. It has already been shown that both cytosolic and mitochondrial aldehyde dehydrogenases are required for growth on ethanol [32], but only the former has a role in acetate formation during sugar fermentation [33]. Consequently, down-regulation of Ald6p and Ald4p during the first generation suggests that they play a role during aerobic propagation on saccharose, and probably are involved in adaptation to fermentation conditions. Indeed, a strong (about 6-fold) decline in Ald6p expression probably leads to a consistent reduction in acetyl-CoA supply for the Krebs cycle but also partially for lipid synthesis. The protein encoded by the ACH1 gene, which catalyses hydrolysis of acetyl CoA in the mitochondrial compartment [35], also displayed significant repression (about 6-fold) during the first generation. Interestingly, Ach1p and Ald6p are both subject to repression by the cAMP-signalling pathways via the stress-responsive cis element (STRE) [36]. The simultaneous repression of both proteins thus would suggest that Ach1p is also involved in oxidative metabolism during propagation. Summarising, slowdown and adaptation to fermentation conditions appear to involve a down-regulation of glycolytic flux and less active PDH bypasses. Fine regulation of the latter will have to be studied further to better understand their contribution to biosynthesis and energy metabolism. Other pathways

Concerning stress-response proteins, variable expression was noted for six of the fifteen proteins compared quantitatively. Ssc1p and Hsp60p, both involved in mitochondrial protein folding, decreased whereas expression of other heat-shock proteins such as Ssa4p and Hsp26p, as well as Kar2p and Pnc1p, two proteins also related to stress-response, were strongly enhanced.

Other pathways were also modified during the first generation. For example, two proteins, Yhb1p and Hem13p, were affected by the change from oxidative to fermentative metabolism. YHB1 is a gene encoding a flavohemoprotein whose expression is higher in aerobic conditions [38]. In fact, the decrease in the level of Yhb1p during the first generation goes in line with the switch from aerobic to anaerobic metabolism. Hem13p, known to be involved in heme biosynthesis, presented enhanced expression, which correlates with previous observations indicating that oxygen inhibits the encoding gene [39]. Since there is a major slowdown of the respiratory system, decreased Krebs cycle activity (Fum1p, Cit2p, Kgd1p and Kgd2p) is not surprising, as confirmed by the down-regulation of Atp2p, Cor1p and Rip1p, involved in the mitochondrial respiratory chain. Consequently, the slowdown of the PDH bypass seems to have an impact on lipid synthesis and the Krebs cycle which corroborates with the general slowdown of biosynthetic pathways such as the synthesis of sulphur (Sam2p, Sam1p, Met6p, and Met17p) and branched (Ilv5p) amino acids as well as arginine (Arg1p).

3.3.2 Third generation (Fig. 5B)

The yeast was re-pitched in a second and then in a third subsequent generation. Two-dimensional proteome analysis was performed at the same steps as for the first generation (beginning T1/end T3). Results are reported in Table 3 and illustrated in Fig. 5B (a and b). The overall pattern shows much less variation than during the first generation. Quantitatively, the degree of variation in expression was much lower than in the first fermentation; maximum variability was 5-fold, compared with 10-fold for the first generation. Spots exhibiting variability corresponded essentially to proteins involved in carbon metabolism, methionine biosynthesis and stress-response pathways. Only 27 spots with at least 2-fold changes were found and 74% corresponded to enhanced expression. Proteins involved in methionine biosynthesis such as Sah1p, Met6p and Met3p, as well as those involved in carbon metabolism, are found in these 74%. But the increase of the Adh1p and Pdc1p cannot be considered as really enhanced since variable spots of these proteins were considered as minor spots. Concerning stress-response proteins, Ssb1p was down-regulated, whereas Ssa1p and Kar2p clearly presented enhanced expression.

Unsurprisingly, in contrast to the first generation, no protein variations related to the aerobic–anaerobic switch were observed during the third generation.

3.3.3 Comparisons between first and third generation: stress-response proteins

Changes in stress-responsive proteins between the end of the first generation (F3) and the beginning (T1) of the third generation, especially the evolution of stress proteins, were studied (Table 4).

View this table:
Table 4

Evolution of some stress proteins during the fermentation process

ProteinFirst generationComparison between F3/T1 stagesThird generation
SSA4IncreasedNo change*No change
HSP26IncreasedNo changeNo change
PNC1IncreasedNo changeNo change
HSP60DecreasedNo changeNo change
SSC1DecreasedNo changeNo change
SSANo changeNo changeIncreased
SSB1No changeIncreasedNo change
  • “Increased” and “Decreased” indicate at least 2-fold change. All variations were established by quantitative analysis, except for proteins designated by an asterisk (∗) indicating visual comparison.

The adaptation to anaerobic conditions was characterized by enhancement of Hsp26p, Ssa4p, Pnc1p and Kar2p during the first generation. In the third generation, the enhanced level was maintained for all of these proteins except for Kar2p which decreased. Hsp26p is a general stress-responsive protein known to be induced upon exposure to a range of stress factors, such as entry into stationary phase or diauxic shift [42], sporulation [43,44], ethanol stress [45], osmostress [46], weak-acid exposure [37], heat shock [45] and H2O2 exposure [47]. Ssa4p, a member of the Hsp70 protein family, is heat- and ethanol-inducible [45,48] and is involved in chaperone functions preventing aggregation and allowing refolding of stress-damaged proteins. Pnc1p is also classified in the stress-response protein family since it is induced by sorbic acid stress [37,49] and ethanol treatment [50]. Thus, Hsp26p, Ssap4 and Pnc1p are stress proteins induced by a variety of treatments. Their induction during the first generation suggests that the oxidative-to-fermentative switch can be also considered as an environmental stress. Moreover, these proteins are also constitutively expressed in subsequent generations, and so are probably important for maintaining the viability of cells encountering stressful fermentative conditions.

KAR2 also encodes a chaperone protein of the HSP70 family and is required for protein folding in the endoplasmic reticulum (ER). In our study, Kar2p seems to be related to the process of adapting to fermentation-related stresses, since its expression was highly induced in both the first and third generations, but to a lesser extent at the beginning of the third fermentation (T1) than at the end of the first fermentation (F3).

On the other hand, Ssc1p and Hsp60p, involved in subcellular mitochondrial chaperone-assisted folding [51], both exhibited decreased expression, but only during the first fermentation. The down-regulation of these proteins is directly related to the general slowdown of mitochondrial activity especially for protein biogenesis.

Two other proteins, Ssb1p and Ssa1p, which both belong to the HSP family, show variable expression. Ssb1p and Ssb2p are ribosome-associated proteins which play a role in the folding of nascent polypeptides emanating from ribosomes [52]. Ssa1 to Ssa4 proteins form an essential chaperone group: at least one of them must be present at high levels to perform their chaperoning activity and assure cell viability. Ssb1 expression increased between the end of the first generation and the beginning of the third, whereas Ssa1p expression increased during the third generation. The role of these two proteins could not be clearly identified in these conditions.

4 Conclusions

This study confirms the relatedness between S. cerevisae S288c and the ale-brewing strain. The first protein reference map of an ale-brewing strain established here will be very helpful for future investigations concerning these kinds of brewing yeasts, but it also allowed us to study variations in protein expression during production-scale fermentation.

The most pronounced changes in protein expression occur in the first generation, during the switch from oxidative to fermentative metabolism. Besides the oxidative effect, the yeast has to cope with considerable change in nutrient supply due to the switch from synthetic media to wort. Yeast propagation is performed in synthetic saccharose medium whereas the wort used in the fermentation process principally contains maltose as well as glucose, fructose and small amounts of maltotriose and saccharose as carbon sources. Nevertheless, no drastic protein changes directly related to these important changes in sugar source were observed. The most significant changes observed were probably indirectly linked to glucose repression mechanisms involved in the adaptation of cells entering fermentation metabolism. In any case, the main effect of this adverse environmental change appears to be a general slowdown of acetyl-CoA metabolism via the PDH bypass and mitochondrial proteins. This result also highlights the huge impact of yeast preculture on metabolic changes occurring in the first generation of fermentations. Moreover, the great adaptability of brewing strains to fermentation conditions is revealed in the third generation where very few stress proteins are affected.

The complexity of the production-scale fermentation with many adverse effects does not allow to clearly identify biomarkers for the prediction of optimal fermentation progress. Although this study allows us to identify that Hsp26p, Ssa4p and Pnc1p as essential during fermentation, they cannot be used to predict a bad progression of fermentation.

Yeast cells must cope with many different kinds of stress during the brewing fermentation process. As highlighted in this study, stresses on brewing yeast must be tested on an individual strain basis to obtain a better understanding of the impact of each kind of stress on cell physiology and behaviour.


The authors would like to thank Pr. Alain Van Dorsselaer and Dr. Jean-Marc Strub for MS protein identification. Dominique Kobi is supported by a CIFRE fellowship from ANRT (Association nationale de la recherche technique).


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
View Abstract