SEARCH

SEARCH BY CITATION

Keywords:

  • iron;
  • lager beer;
  • oxidative stability;
  • Saccharomyces pastorianus;
  • thioredoxin;
  • yeast

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Aims:  In this study, we investigated the relationship between the ability of lager brewing yeast strains to tolerate oxidative stress and their ability to produce oxidative stable model beer.

Methods and Results:  Screening of 21 lager brewing yeast strains against diamide and paraquat showed that the oxidative stress resistance was strain dependent. Fermentation of model wort in European Brewing Convention tubes using three yeast strains with varying oxidative stress resistances resulted in three model beers with different rates of radical formation as measured by electron spin resonance in forced ageing experiments. Interestingly, the strain with the lowest oxidative stress resistance and lowest secretion of thioredoxin, as measured by Western blotting, resulted in the highest uptake of iron, as measured by inductively coupled plasma-mass spectrometry, and the slowest formation of radicals in the model beers.

Conclusions:  A more oxidative stable beer is not obtained by a more-oxidative-stress-tolerant lager brewing yeast strain, exhibiting a higher secretion of thioredoxin, but rather by a less-oxidative-stress-tolerant strain, exhibiting a higher iron uptake.

Significance and Impact of the Study:  To obtain lager beers with enhanced oxidative stability, yeast strains should be screened for their low oxidative stress tolerance and/or high ability to take up iron rather than for their high oxidative stress tolerance and/or high ability to secrete thioredoxin.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

During industrial beer fermentation, the brewers yeast will upregulate genes encoding antioxidants, thus indicating that it has been subjected to oxidative stress during fermentation (Higgins et al. 2003; James et al. 2003). Oxidative stress typically leads to the formation of reactive oxygen species (ROS), which cause damage to the lipids, proteins and DNA of yeast cells (Halliwell and Aruoma 1991; Temple et al. 2005). The exact cause of this oxidative stress during the fermentation process has, as yet, not been found, but it seems to be related more to starvation than to the presence of oxygen during the initial stages of the fermentation (Gibson et al. 2008). To prevent intracellular oxidative damage, the yeast cell possesses an oxidative stress response, consisting of both enzymatic and nonenzymatic defence mechanisms to remove or detoxify ROS. Enzymatic defence mechanisms include peroxidases, catalases and superoxide dismutases, and the nonenzymatic defence mechanisms include antioxidants such as glutaredoxin, glutathion and thioredoxin (Grant et al. 1998; Jamieson 1998). The expression of thioredoxin may be induced by chemicals that cause oxidative stress, such as H2O2, paraquat and diamide (Delaunay et al. 2000; Garrido and Grant 2002; Braconi et al. 2010). It is well known that the tolerance of yeast to various other stress conditions, such as ethanol, is strain dependent (Kubota et al. 2004). So far, however, the oxidative stress tolerance of different strains of brewers yeast has not been studied.

In the past decade, the role of oxidation processes in the formation of beer off-flavour compounds has received considerable interest, because these compounds most likely are oxidation products of components present in the wort and/or the beer (Bamforth and Lentini 2009). The oxidation of flavour compounds in wort and/or beer is initiated by the reaction of oxygen with transition metals, such as iron and copper, thus generating the superoxide anion (inline image), which may be further reduced and protonated to hydrogen peroxide (H2O2). The superoxide anion and hydrogen peroxide can then undergo the Haber–Weiss and Fenton reactions, respectively, with iron and/or copper ions to produce the very reactive hydroxyl radical (OH˙) (Kaneda et al. 1988). The generated ROS species may react with various wort and beer compounds, resulting in flavour changes in the beer (Vanderhaegen et al. 2006).

One of the strategies to prevent oxidation in beer is to capture ROS and free radicals by antioxidants (Vanderhaegen et al. 2006). Sulphite is the most abundant antioxidant in beer, produced and secreted by the brewers yeast during fermentation (Kaneda et al. 1994, 1996; Andersen et al. 2000). Thioredoxin, a cytosol-localized antioxidant, has been reported to be secreted out of the cell during sake and beer production (Swan et al. 2003; Inoue et al. 2007). Furthermore, a proteomic study of 11 lager beers identified thioredoxin as one out of only four proteins originating from yeast (Iimure et al. 2010). Thus, besides sulphite, thioredoxin may also be a potential candidate for protecting the beer against oxidation of flavour compounds. Moreover, it may be postulated that more-oxidative-stress-tolerant yeast strains may produce and secrete more thioredoxin to the beer during fermentation, thereby resulting in a beer with enhanced oxidative stability.

In this study, we identify lager beer yeast strains with different tolerances towards thiol-depleting and superoxide-generating oxidative stress. We find that model beers fermented with the most-oxidative-stress-tolerant strain contain the highest amount of thioredoxin as compared with beers fermented with two less-oxidative-stress-tolerant strains. However, the oxidative stability of the beer, as measured by electron spin resonance (ESR), is not governed by the ability of the yeast to secrete thioredoxin. Rather, it seems to depend on the ability of the yeast to take up iron.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Yeast strains and media

The yeast strains (n = 21) used in this study were all lager brewing strains, belonging to the species Saccharomyces pastorianus, obtained from White Labs (WL, San Diego, CA, USA) and our own collection (KVL) at the Department of Food Science, Food Microbiology, University of Copenhagen (see Table 1). Yeast strains were grown in synthetic beer media (SBM), composed of 4% maltose; 1·5% sucrose; 1% dextrin; 1% glucose; 0·5% fructose; 2·94% balanced peptone; and 1·7% yeast nitrogen base w/o aa adjusted to pH 5·5.

Table 1.   Screening for pro-oxidant effect on lager yeast*
Yeast stainSynthetic beer media
Paraquat (mmol l−1)DIAMIDE (mmol l−1)
0.50.751.51.752
  1. (+++), no effect of additive; (++), minor effect – smaller colonies; (+), large effect – smal colonies; (−), no growth.

  2. *Yeast strains were spotted in 10-fold dilutions on SBM. Colony sizes were determined after 6 days at 11°C.

KVL001++++++++++
KVL002+++++++++++++
KVL003+++++
KVL004
KVL005+++++
KVL006++++++
KVL007+++
KVL008++++
KVL009++++
KVL010++++++
KVL016+++++
KVL017+++++
KVL018
KVL019+++++++
KVL020++++++
WLP800++
WLP810+++++++++++
WLP830++++++
WLP833+++++++++++
WLP838+++
WLP840+++++++++++

Screening for oxidative stress tolerance

Yeast was grown to exponential phase in 10 ml of SBM at 11°C without shaking and diluted to OD600 0·1, 0·01 and 0·001. Cells were spotted on SBM plates containing either diamide (1·5; 1·75 or 2 mmol l−1) or paraquat (0·5 or 0·75 mmol l−1). Plates were incubated at 11°C for 6 days. For each strain, growth was scored against colony size without any additives annotated as (+++) no effect of additive, (++) minor effect – smaller colonies, (+) large effect – tiny colonies and (−) no growth.

Model beer fermentation

Aerobic propagation of yeast was started from a single colony in 10 ml of SBM – in duplicates. After incubation at 20°C for 24 h, the suspensions were transferred to 100 ml SBM in 250-ml Erlenmeyer flasks with a magnetic stirrer at 200 rev min−1. Yeast suspensions were transferred after 2 days at 20°C to 400 ml SBM and incubated for 24 h at 20°C. Yeast cells were harvested (3000 g, 10 min, 20°C) and pitched at 7 × 106 cells ml−1 in 2 l of SBM saturated with air. Fermentations were carried out in 2·5-litres European Brewing Convention (EBC) tubes at 11°C for 12 days. Samples of culture broth were collected aseptically with a syringe from the top of the EBC-tubes at days 0, 1, 2, 3, 6, 9 and 12. Cell density was determined by measuring the optical density at 600 nm (UV-1800; Shimadzu Scientific Instruments, Columbia, MD), and pH was determined using a pH-meter (pHM220; Radiometer Analytical SAS, Villewbanne Cedex, France) before further treatment of samples.

Protein analysis

Immediately after sampling, proteins were precipitated from a 15-ml sterile filtered sample with trichloroacetic acid (TCA; finale concentration 12·5%) at −20°C O/N. Proteins were pelleted (15 000 g, 4°C, 30 min), washed twice in 1/3 volume ice-cold acetone (15 000 g for 15 min at 4°C) and re-suspended in buffer (400 μl of 10 mmol l−1 MES pH 6·0, 1 mmol l−1 EDTA).

2D Quant (80-6483-56; GE Healthcare Bio-Sciences, Hillerød, Denmark) was used to determine the protein concentration according to the manufacturer’s protocol, with BSA as standard. 60 μg of protein was separated on a 20% RunBlue minigel using an Xcell II electrophoresis system (Life Technologies Europe BV, Naerum, Denmark).

Western blotting

Separated proteins were blotted on polyvinylidene diflouride membranes. Bound anti-yeast thioredoxin (a kind gift from Professor Chris M. Grant) was visualized by alkaline phosphatase after incubation with anti-rabbit immunoglobulin-alkaline phosphatase conjugate. ImageJ 64 (Rasband, W.S., ImageJ; US National Institutes of Health, Bethesda, MD, USA, http://rsb.info.nih.gov/ij/, 1997–2009) was used to quantify band intensity of the developed blots by estimating the peak intensity. Thioredoxin concentrations were determined from a dilution row of yeast-purified thioredoxin (Genway, San Diego, CA, USA) on the same blot.

Sulphite analysis

Enzymatic determination of total sulphite was performed using a Sulphite UV determination kit according to the manufacturer’s protocol with some modifications (10725854035; Roche Diagonostics A/S, Hvidovre, Denmark). In brief, reaction size was scaled down 10 times for measurement in microtitre plates. Samples were sterile filtered through a 0·22-μm filter to separate yeast cells from sample and diluted threefold in modified reaction buffer (700 mmol l−1 triethanolamine, 0·4 g l−1 NADH, pH 8·0). 0·01 U NADH peroxidase was added, and the start absorbance was recorded at 340 nm followed by the addition of 0·1225 U sulphite oxidase. The reduction in NADH was measured at 340 nm, and sodium sulphite was used as standard. Values were determined as a minimum in triplicates.

Sugar and ethanol determination

Samples were filtrated through a 0·22-μm sterile filter and kept at −20°C until analysis. Sugar and ethanol concentrations were determined using a HPLC (HP series 1100; Hewlett-Packard ApS, Allerød, Denmark) with a MicroGuard cation H cartridge followed by an Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA, USA) connected to a RI detector (HP1047A; Hewlett-Packard). The column was eluted with a degassed mobile phase containing 2·5 mmol l−1 H2SO4, pH 2·75, at 50°C and at a flow rate of 0·6 ml min−1.

ESR experiments

The lag phase experiments were performed by heating approximately 1 ml of sterile filtered samples containing 30 mmol l−1 of the spin trap phenyl N-tert-butylnitrone (PBN) in 3·6-ml screw-cap tubes in a 60°C water bath. The first sample (50 μl) was withdrawn after 10 min followed by intervals of 20 min. ESR spectra of the samples were recorded with a Miniscope MS 200 X-band spectrometer (Magnetteech Gmbh, Berlin, Germany) using 50-μl micropipettes as sample cells. The settings used were as follows: microwave power, 10 mW; sweep width, 60 G; modulation frequency, 2000 mG; receiver gain, 900; and sweep time 30 s. All spectra consisted of single scans and were recorded at room temperature. The amplitudes of the spectra were measured and are reported as the height of the central doublet. The lag phases of the ESR spectra, that is, the antioxidant capacity, were determined according to Uchida and Ono (2000), and the relative radical formation rates, that is, the oxidative stability, were determined by linear regression of the straight line and related to the slowest formation. All samples were measured in duplicates.

Metal ion determination

To determine the concentrations of iron and copper, samples were filtered through a 20-μm sterile filter, condensed and acid digested in 45% HNO3 and 10% H2O2 (Wyrzykowska et al. 2001). Ion analysis of samples were analysed using inductively coupled plasma-mass spectrometry (ICP-MS) (Agilent 7500c; Agilent Technologies Denmark ApS, Hørsholm, Denmark). Settings for the ICP-MS were as described in Hansen et al. (2009). All samples were measured in duplicates.

Statistical analysis

All results represent the mean values ± standard error of the mean (SEM) from two independent fermentations with at least duplicate measurements. Statistical analysis was performed by one-way analysis of variance (anova) and Tukey’s post hoc using StatPlus software (AnalystSoft, Inc., Vancover, British Columbia, Canada). Probabilities <0·05 were considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Screening of brewers yeast against pro-oxidants

Initially, a collection of 21 lager brewing yeast strains was screened for their ability to withstand oxidative stress. Two pro-oxidants, that is, diamide – a thiol-depleting agent and paraquat – a superoxide-generating agent, were used to induce oxidative stress by different pathways (Bus and Gibson 1984; Kosower and Kosower 1995). Yeast was grown on agar plates in aerobic conditions in tenfold dilutions against near lethal concentrations of the selected pro-oxidants. The screening results show that some yeast strains were clearly more tolerant to both pro-oxidants than others (Table 1). Moreover, paraquat seemed to be more reactive than diamide (Table 1).

Model beer fermentations in EBC-tubes

Based on the screening results on SBM, a highly-oxidative-stress-resistant strain, that is WLP833; and a low-resistant strain, that is KVL018, were selected for model beer fermentations. Besides, KVL001 was included as a medium-resistant strain. Growth and metabolic profiles for fermentations in 2 l SBM in EBC-tubes using yeast strains KVL001, KVL018 and WLP833 are shown in Fig. 1. The three fermentations were all similar and resembled typical beer fermentation (Fig. 1).

image

Figure 1.  Fermentation profiles for yeast strains KVL001 (a), KVL018 (b) and WLP833 (c), respectively, grown in 2 l synthetic beer media in European Brewing Convention tubes, showing (○) glucose; (□) fructose; (bsl00066) maltose; (△) ethanol; (▪) OD and (•) pH. Values are means of two biologically independent fermentations, and error bars indicate standard deviations.

Download figure to PowerPoint

Oxidative stability of model beers

Forced ageing of the fermentation samples followed by ESR were performed to determine the oxidative stability of the model beers after 12 days. ESR measures the level of radicals in samples indirectly by detecting the formation of radicals trapped by the spin trap, PBN. The formation of stable spin trap radical follows a two-phase course; first a lag phase is observed and then the rate of radical formation increases and signals from the spin trap increase linearly with time (Fig. 2). The lag phase is proposed to be the time where antioxidants quench the radicals before reacting with a spin trap and may thus reflect the antioxidant capacity (Uchida and Ono 1996, 2000). The model beers showed similar lag phase (Table 2), indicating that the amount of antioxidants produced by the different yeast strains were not significantly different. However, model beers fermented with KVL018 generated radicals at a low rate, whereas those fermented with KVL001 and WLP833 produced radicals at 3- and 1·8-fold higher rates, respectively (Table 2).

image

Figure 2.  Formation of spin trap radicals measured by electron spin resonance during forced ageing at 60°C of model beers fermented in duplicate with (△) KVL001; (○) KVL018 and (□) WLP833.

Download figure to PowerPoint

Table 2.   Sulphite concentration, antioxidante capacity and radical formation rate in model beers after 12 days of fermentation*
Yeast strainSulphite (mg l−1)ESR lag phase (min)Relative radical formation rate compared to KVL018 (fold)
  1. ESR, electron spin resonance.

  2. *Values represent the mean ± SEM from duplicate measurements from two independent fermentations. Values with different capital letters in superscripts, within a column, are significantly different (P < 0.05).

KVLO013.75 ± 0.0855 ± 8.13.2 ± 0.33A
KVL0182.50 ± 0.0653 ± 1.91 ± 0.22B
WLP8333.48 ± 0.0460.5 ± 2.31.8 ± 0.03B

Sulphite content of model beers

The concentration of sulphite were similar in all three model beers after 12 days of fermentation, ranging from 3·75 mg l−1 in model beer fermented with KVL001 to 3·48 and 2·50 mg l−1 in model beers fermented with WLP833 and KVL018, respectively (Table 2).

Thioredoxin content of model beers

Swan et al. (2003) have earlier reported that brewers yeast are able to secrete thioredoxin to the beer during fermentation. Our data support these findings. Western blotting of protein samples during fermentation showed that the level of thioredoxin in the model beers increased from being nondetectable in the beginning of the fermentation (data not shown) to 40–80 pg μg−1 total protein after 12 days of fermentation (Fig. 3). Here, KVL001 and WLP833 had produced twofold more thioredoxin, that is 80 pg μg−1 total protein, than KVL018, that is, 40 pg μg−1 total protein (Fig. 3).

image

Figure 3.  Western blot detection of extracellular thioredoxin after 12 days of fermentation with KVL001, KVL018 and WLP833 in 2 l synthetic beer media in European Brewing Convention tubes. 60 μg of TCA precipitated proteins from two independent fermentations (A/B) were separated on a 20% SDS-PAGE gel blotted on to a PVDF membrane, detected with an anti-yeast-thioredoxin antibody, and quantified from the dilution row of 1·3–10·4 ng yeast thioredoxin (TRX2) standard.

Download figure to PowerPoint

Metal contents of model beers

According to Zufall and Tyrell (2008), the iron and copper concentrations in wort are in the range of 100–270 and 20–400 μg l−1, respectively. In the model wort used in this study, the iron concentration is higher (600–815 μg l−1), yet the copper concentration is in the interval reported by Zufall and Tyrell (25–30 μg l−1) (Table 3). Both the copper and iron ion concentrations detected in the model beer were comparable to those reported earlier (Wyrzykowska et al. 2001; Pohl and Prusisz 2010; Sancho et al. 2011).

Table 3.   Transition metal concentration in wort*
Yeast strainFe (ppb)Cu (ppb)
Days of fermentationDays of fermentation
012012
  1. *Metal ion concentrations were determinated by ICPMS. Values represent the mean ± SEM from duplicate measurements from two independent fermentations. Values with different capital letters in superscripts, within a column, are significantly different (P < 0.05).

KVL001831 ± 35396 ± 27A28 ± 3.012 ± 1.4
KVL018665 ± 3257 ± 9.6B27 ± 0.89 ± 1.2
WLP833624 ± 46160 ± 14C26 ± 1.99 ± 0.9

The copper ion concentrations in the model beers were virtually similar (9–12 μg l−1) after 12 days of fermentation (Table 3). The iron ion concentrations were, however, very different ranging from 57 ± 9 μg l−1 in the model beer fermented with KLV018 to 396 ± 27 μg l−1 in the one fermented with KVL001 and 160 ± 14 in the WLP833 fermented model beer (Table 3).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

To the best of our knowledge, this is the first report comparing the ability of brewing yeast strains to endure oxidative stress. We show here that the ability of lager brewing yeast to tolerate the two different oxidative-stress-inducing compounds, diamide and paraquat is strain dependent. Because brewers yeast somehow seem to be subject to oxidative stress during beer fermentation (Higgins et al. 2003; James et al. 2003), and because intracellularly produced antioxidants, such as thioredoxin, have been reported to be secreted to the beer during fermentation (Swan et al. 2003), we expected that strains with increased resistance to pro-oxidants would be able to circumvent this stress condition better by producing and secreting more antioxidants to the beer. Interestingly, this seems to be the case, that is, the two more-oxidative-stress-tolerant strains (KVL001 and WLP833) produced beer with higher amounts of thioredoxin than the low-oxidative-stress-tolerant strain (KVL018).

In this study, the antioxidant capacity of the three model lager beers, as represented by the lag phase in the ESR experiments, are virtually similar. This may be explained by the fact that the sulphite concentrations in all three model beers are the same, and it is in accordance with previous studies reporting that the ESR lag phase is correlated with the amount of sulphite in beer (Uchida et al. 1996; Andersen et al. 2000).

Our results, however, also demonstrate that the rates of radical formation are different in the three model lager beers. The rate of radical formation is expected to be slow when the sample has either a high oxidant defence, which hinders oxidative reactions, and/or a low level of pro-oxidants favouring radical formation (Uchida et al. 1996). As the most-oxidative-stress-tolerant strains secrete the most thioredoxin, we expected that these strains would also result in beers with an increased oxidative stability. Surprisingly, our results show more or less the opposite, that is, the yeast strain most prone to oxidative stress in the screening experiment (KVL018) gives beer with the slowest formation of radicals. Together, our data indicate that the oxidative stability of lager model beers is not related to yeast strains having a high oxidative stress tolerance, and thus a good ability to secrete thioredoxin.

Rather, it seems as if KVL018 is superior to take up iron, which, in turn, may explain its low oxidative stress tolerance. Perhaps its higher iron uptake may lead to increased intracellular iron levels and thus higher ROS levels as a result of Haber–Weiss and Fenton reactions (Vanderhaegen et al. 2006). The good ability of KVL018 to take up iron, however, seems to decrease the rate of radical formation outside the cells and may thus be beneficial for the oxidative stability of the lager model beer. Interestingly, the three strains do not differ in their copper uptake. Further experiments are required to identify the mechanisms underlying this phenomenon. Many strategies have so far been proposed to inhibit metal-induced, oxidative changes in beer (Vanderhaegen et al. 2006). Based on our results presented here, we suggest a new strategy for lager beers: to use lager yeast strains that are susceptible to oxidative stress and/or are highly capable of capturing iron ions.

It remains to be investigated if other yeast produced antioxidants; for example, glutaredoxin and glutathione have an influence on the oxidative stability of lager beer. It should also be noted that although high levels of antioxidants may be desired to obtain oxidative stable beer, they may have a negative influence on foam stability of beer, as reported for thioredoxin (Iimure et al. 2008).

In conclusion, we find that lager beer yeast resistant to oxidative stress does not produce more oxidative stable model beer. On the other hand, we see increased thioredoxin production and secretion to the beer correlating with the tolerance towards pro-oxidants. We have identified a lager beer yeast strain that has a low oxidative stress tolerance and is able to assimilate almost all the iron present in model wort, thereby decreasing the rate of radical formation in model beer. Thus, our results suggest that, to obtain lager beers that are oxidatives, yeast strains should be screened for their low oxidative stress tolerance and/or their good ability to take up iron rather than for their high oxidative stress tolerance and/or high secretion of thioredoxin.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This project was financed by the Danish Ministry of Food, Agriculture and Fisheries, project no. 3304-FVFP-07. We thank Christopher White from White Labs, San Diego, USA, for kindly providing yeast strains and are grateful to Professor Chris Grant, University of Manchester, UK, for donating the yeast thioredoxin antibody.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References