Trehalose promotes the survival of Saccharomyces cerevisiae during lethal ethanol stress, but does not influence growth under sublethal ethanol stress

Authors


  • Editor: Ian Dawes

Correspondence: Grant A. Stanley, School of Engineering and Science, Victoria University, PO Box 14428, Melbourne, Vic. 8001, Australia. Tel.: +61 3 9919 8104; fax: +61 3 9919 8284; e-mail: grant.stanley@vu.edu.au

Abstract

Trehalose is known to protect cells from various environmental assaults; however, its role in the ethanol tolerance of Saccharomyces cerevisiae remains controversial. Many previous studies report correlations between trehalose levels and ethanol tolerance across a variety of strains, yet variations in genetic background make it difficult to separate the impact of trehalose from other stress response factors. In the current study, investigations were conducted on the ethanol tolerance of S. cerevisiae BY4742 and BY4742 deletion strains, tsl1Δ and nth1Δ, across a range of ethanol concentrations. It was found that trehalose does play a role in ethanol tolerance at lethal ethanol concentrations, but not at sublethal ethanol concentrations; differences of 20–40% in the intracellular trehalose concentration did not provide any growth advantage for cells incubated in the presence of sublethal ethanol concentrations. It was speculated that the ethanol concentration-dependent nature of the trehalose effect supports a mechanism for trehalose in protecting cellular proteins from the damaging effects of ethanol.

Introduction

Microbial-based ethanol production serves large and diverse industries, from alcoholic beverages to biofuel production. Although ethanol is the desired end product of yeast fermentation, it becomes a significant stress factor as it accumulates in the culture broth. In general, Saccharomyces spp. are highly ethanol tolerant relative to most microbial species, with some strains being able to produce up to 18% (v/v) ethanol in a single batch. High ethanol concentrations, however, inhibit yeast growth and viability, and affect fermentation performance, resulting in decreased fermentation productivity and ethanol yield (Norton et al., 1995; Galeote et al., 2001; Aguilera et al., 2006). A better understanding of the cellular consequences of microbial ethanol stress and of the underlying cell-based ethanol stress defence mechanisms is important for improving the performance of yeast strains during industrial fermentations.

Trehalose is known for its role as a reserve carbohydrate in yeast, but it is also associated with the protection of cells against many environmental stressors, including ethanol stress. Trehalose is commonly found in organisms, as diverse as yeast and other fungi, bacteria, a variety of plants and invertebrates, in which it accumulates significantly during adverse environmental conditions (Gaff, 1971; Thevelein, 1984; Singer & Lindquist, 1998; Zentella et al., 1999; Chen et al., 2002; Schluepmann et al., 2004).

In yeast, trehalose is synthesized by a large enzyme complex comprising two enzymes: trehalose-6-phosphate synthase (Tps1) (Bell et al., 1992) and trehalose-6-phosphate phosphatase (Tps2) (de Virgilio et al., 1993). Two other enzymes, Tsl1 and Tps3, are believed to be alternative, regulatory or stabilizing subunits of the complex (Bell et al., 1998). Trehalose is synthesized from UDP-glucose and glucose-6-phosphate, which are converted to trehalose-6-phosphate by Tps1 and then to trehalose by Tps2 (Thevelein & Hohmann, 1995; Ferreira et al., 1996).

Trehalose is hydrolysed into two glucose molecules by trehalase. Yeast has two trehalase enzymes: neutral trehalase (Nth1) and acid trehalase (Ath1). Nth1 is a cytosolic enzyme encoded by NTH1, the expression of which is regulated by stress (Thevelein, 1984; Hohmann, 2002). Nth1, with optimum activity at pH 6.8–7.0, is responsible for the breakdown of accumulated trehalose. Ath1, with optimum activity at pH 4.5–5.0, is reportedly involved in growth using trehalose as a carbon source (Nwaka et al., 1995). All genes involved in trehalose metabolism contain one or more copies of stress response elements in their promoter regions and are induced by general transcription factors Msn2/4 in response to a range of stresses (Parrou et al., 1997).

Depending on the environmental conditions, trehalose can represent from <1% to >25% of cell dry weight (Hohmann, 2002), and over the past 20 years, a number of studies have demonstrated a correlation between intracellular trehalose levels and the ability of yeast to survive various environmental stresses, such as starvation, desiccation, dehydration, osmotic and oxidative stress and extremes in temperature (Thevelein & Hohmann, 1995; Hounsa et al., 1998; Elbein et al., 2003; Herdeiro et al., 2006). The dramatic and rapid variations in the level of trehalose in response to a variety of environmental changes may reflect a complex regulatory process governing the metabolism of this carbohydrate. Soto et al. (1999) showed that mutant Schizosaccharomyces pombe strains, which were unable to synthesize trehalose, were sensitive to temperature, freeze/thawing, dehydration, sodium chloride and ethanol stresses. These authors speculated that trehalose is a key determinant in general stress tolerance.

A role for trehalose in the ethanol tolerance of Saccharomyces cerevisiae, however, remains controversial, with some reports of a positive effect on yeast tolerance to ethanol (Kim et al., 1996; Mansure et al., 1997; Soto et al., 1999; Pataro et al., 2002; Jung & Park, 2005) and others reporting that trehalose has no impact (Lewis et al., 1997; Alexandre et al., 1998; Ribeiro et al., 1999; Gomes et al., 2002). It has also been shown that trehalose synthesis increases in yeast during ethanol stress (Mansure et al., 1997) and that several genes associated with this (TPS1, TPS2 and TSL1) are highly expressed in S. cerevisiae when exposed to sublethal ethanol concentrations (Alexandre et al., 2001; Chandler et al., 2004). These results suggest that trehalose, or its biosynthetic pathway, plays a role in the ethanol tolerance of S. cerevisiae, and yet investigations into ethanol tolerance using genome-wide screens have either not identified a phenotype associated with ethanol tolerance for null mutants of genes involved in the trehalose metabolism (Fujita et al., 2006), identified a phenotype for TPS1 and TPS2 null mutants at 11% (v/v) ethanol, but not at 8% (v/v) (Kubota et al., 2004), or identified a phenotype only for a TPS1 null mutant in 6% (v/v) ethanol (van Voorst et al., 2006). These various outcomes from the relatively small number of studies in this area have led to confusion over the importance of the trehalose molecule and/or trehalose metabolism in the ethanol tolerance of S. cerevisiae.

The approach used in the current project was to conduct a series of physiological investigations using wild-type and knockout strains, tsl1Δ and nth1Δ. Viability profiles of these knockout strains and the wild type were compared in the absence and presence of ethanol stress across a wide range of ethanol concentrations to determine whether trehalose plays a role in the ethanol tolerance of S. cerevisiae.

Materials and methods

Organism and culture conditions

Saccharomyces cerevisiae strains in this study comprised the genetic background of wild-type strain BY4742 (MATαhis3Δ1 leu2Δ0 lys2Δ0 ura3Δ0). Single-deletion mutants were from the Euroscarf deletion collection (http://www-sequence.stanford.edu) and had one of the genes TSL1 or NTH1 replaced by the kanMX4 module to generate two knockout strains, BY4742 (MATαhis3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) tsl1Δ∷kanMX4 and nth1Δ∷kanMX4. Yeast cultures were grown in a defined medium containing per litre: 20 g d-glucose, 5 g ammonium sulphate and 1.7 g yeast nitrogen base (without amino acids and ammonium sulphate), or on solid nutrient-rich YEPD medium (yeast extract 1% w/v, Bacto peptone 2% w/v, d-glucose 2% w/v and Bacto agar 1.5% w/v). YEPD-Geneticin plates, for the selection of deletion strains, comprised YEPD medium with the addition of 200 mg L−1 G418 Geneticin (Sigma). Yeast cultures were grown under aerobic conditions in defined medium at 30 °C/130 r.p.m. in an orbital-shaker incubator, unless otherwise stated. The culture vessels were Erlenmeyer or sidearm flasks with cotton wool plug stoppers.

Inoculum preparation

Inocula were prepared by aseptically transferring a loopful of cells from stock cultures into a 250-mL Erlenmeyer flask containing 50 mL of fresh defined medium. The culture was incubated overnight at 30 °C/130 r.p.m. Two subsequent serial subcultures were prepared before each experiment. Parent cultures containing 100 mL (or 200 mL depending on the experiment) of fresh defined medium in 500-mL Erlenmeyer flasks were inoculated to an initial OD620 nm reading of 0.1 and grown for approximately 8 h to late exponential phase (OD620 nm of 0.8) at 30 °C/130 r.p.m. Late exponential phase parent culture cells were collected by centrifugation at 3313 g in a swinging rotor centrifuge. The supernatant was discarded and the cells washed in prewarmed fresh defined medium before addition to the experimental cultures.

Ethanol stress experiments

Aliquots of a cell suspension from an exponential phase parent culture were inoculated to an initial OD620 nm of 0.1 (approximately 3 × 106 cells mL−1), into control and experimental flasks (500-mL Erlenmeyer flasks) containing prewarmed defined medium (100 mL) without and with added ethanol. Once inoculated, the cultures were quickly transferred to the shaker incubator and grown under aerobic conditions at 30 °C/130 r.p.m. Samples for OD and viable plate counts were taken at regular intervals during incubation.

Competition experiments were performed in 250-mL Erlenmeyer flasks containing 48 mL of medium either with or without added ethanol. To create a mixed population, equal cell numbers of wild type and a single-deletion strain (approximately 350 μL of each) from late exponential phase parent cultures were inoculated into defined medium in the presence or absence of ethanol, grown for six generations and transferred while in the late exponential phase. Aliquots of mixed populations (750 μL) were transferred into fresh medium with similar conditions. Altogether, five serial transfers were performed (around 36 generations). At the end of each incubation, samples from the mixed cultures were plated on YEPD medium with or without Geneticin to score the population of each strain. All of the cultures were grown under aerobic conditions at 30 °C/130 r.p.m. OD620 nm was used to determine biomass levels and plate counts were used to determine cell viability in all experiments.

Trehalose determination

Trehalose concentrations were determined using trehalose assay (K-TREH 01/05) (Megazyme) as recommended by the supplier. Using this assay, trehalose was hydrolysed to glucose-6-phosphate, with stoichiometric production of NADPH, in a two-step reaction involving trehalase and hexokinase; the increase in NADPH was measured spectrophotometrically at 340 nm and a standard curve was used to determine relative trehalose concentrations. To determine trehalose concentrations, 15 mL of culture was harvested and the cells were immediately centrifuged for 5 min at 4 °C and 3313 g. Pelleted cells were washed with 0.1 M phosphate buffer (pH 5.9) to remove external glucose, followed by further centrifugation. Cell pellets were then frozen in liquid nitrogen and stored at −80 °C. The frozen cell pellets were thawed on ice and resuspended in a 0.25 M Na2CO3 solution at a cell density of 1 × 108 cells mL−1. Samples (1 mL) were boiled for 20 min to extract intracellular trehalose. After cooling, the samples were centrifuged at 12 000 g using a bench-top microfuge (Eppendorf) for 3 min to remove cell debris. Aliquots (300 μl) of supernatant were neutralized by adding 1 M acetic acid (150 μL) and then mixed with 650 μl distilled water, 100 μL imidazole buffer (2 M imidazole, 100 mM magnesium chloride and 0.02% w/v sodium azide; pH 7.0), 50 μL NADP+/ATP (12.5 mg mL−1 NADP+ and 36.7 mg mL−1 ATP) and 10 μL suspension of HK/G-6-PDH (425 U mL−1 hexokinase and 212 U mL−1 glucose-6-phosphate dehydrogenase). Samples were incubated at 28 °C for 10 min. Then, the sample was mixed with 10 μL trehalase (490 U mL−1) and incubated at 28 °C for 20 min. A340 nm was recorded against the reagent blank to determine the trehalose content. Pre-existing glucose was determined in a control reaction without added trehalase.

Results

Genetic confirmation and stability of the knockout strains

Saccharomyces cerevisiae BY4742 (wild type) and two knockout strains in this background, tsl1Δ∷kanMX4 and nth1Δ∷kanMX4, were used throughout. Replacement of the TSL1 and NTH1 ORFs with the kanMX4 module was verified by PCR analysis. The presence and the position of the kanMX4 module was confirmed using a flanking primer upstream of each ORF in combination with a downstream primer complementary to an internal region of the kanMX4 cassette. The stability of wild-type and knockout strains was confirmed by serially subculturing each strain three times in nonselective YEPD medium. The strains were then plated onto both YEPD and YEPD-Geneticin plates. The YEPD-Geneticin plates did not support the growth of the wild type, yet YEPD and YEPD-Geneticin plates for both knockout strains had equivalent colony numbers, indicating that the kanMX4 module had been conserved after three serial transfers.

Notable absences in the knockout strains used in this project are tps1Δ and tps2Δ, for the following reasons. Previous studies with strain tps1Δ have shown that it does not grow on glucose due to deregulation of the glycolytic pathway (Bell et al., 1992; González et al., 1992; Neves et al., 1995). The TPS1 gene encodes the small subunit of trehalose-6-phosphate synthase in the trehalose synthesis complex (Bell et al., 1992) and disruption of this gene blocks the initial step of trehalose synthesis involved in trehalose-6-phosphate production. Trehalose-6-phosphate plays an important role in the regulation of glycolysis, mainly through the inhibition of hexokinase II (Blazquez et al., 1993; Hohmann et al., 1996). Therefore, a tps1Δ strain lacking TPS1 gene has a growth defect on fermentable carbon sources including glucose (Bell et al., 1992; González et al., 1992; Neves et al., 1995). Saccharomyces cerevisiae tps2Δ strains are known to have viability issues due to intracellular accumulation of trehalose-6-phosphate, especially during exposure to stressors such as heat shock, osmotic stress and nutrition limitation (de Virgilio et al., 1993; Hounsa et al., 1998). The intracellular accumulation of trehalose-6-phosphate is known to be toxic to yeast (Thevelein & Hohmann, 1995; Gancedo & Flores, 2004).

Intracellular trehalose levels

Trehalose levels in all strains were determined at various stages of growth and in the presence or absence of ethanol (7% v/v). In the absence of ethanol stress, the wild-type and nth1Δ strains accumulated trehalose from late exponential phase onward. The nth1Δ strain, having a disruption in trehalose mobilization, had a trehalose content that was more than two times higher than that of the wild type in the late exponential phase (Fig. 1). Strain tsl1Δ, containing disrupted trehalose biosynthesis, did not have detectable amounts of trehalose during the late exponential phase; however, this strain accumulated trehalose during the stationary phase such that its trehalose concentration at the late stationary phase was close to 78% and 53% of that measured in the wild type and nth1Δ strains, respectively. In the presence of 7% (v/v) ethanol, late exponential phase cells of all three strains had detectable amounts of trehalose, which were considerably higher than that of the exponential phase cells in the absence of ethanol. Late exponential phase cells of strains tsl1Δ and nth1Δ grown in the presence of 7% (v/v) ethanol had 19% less or 20% more trehalose, respectively, compared with the wild type grown under the same conditions. In both the presence or the absence of ethanol, the intracellular trehalose concentration increased in wild type, tsl1Δ and nth1Δ cells when they entered the stationary phase and continued to increase with increasing incubation time, although the magnitude of this change in trehalose concentration from the early to the late stationary phase was considerably less for cells grown in the presence of 7% (v/v) ethanol. Compared with the wild type, the lowest and highest trehalose concentrations were always measured in strains tsl1Δ and nth1Δ, respectively, at all stages of incubation.

Figure 1.

 Intracellular trehalose concentrations in Saccharomyces cerevisiae wild type, tsl1Δ and nth1Δ strains incubated to late exponential phase (□), early stationary phase (inline image) and late stationary phase (▪) of growth. Cultures were incubated in defined medium at 30oC/130 r.p.m. in the presence or the absence of added ethanol (7% v/v). Error bars indicate SD from the mean of triplicate experiments.

Batch growth in sublethal ethanol stress

Batch incubations were performed to determine the influence of trehalose and its metabolism on the growth of S. cerevisiae in medium containing various ethanol concentrations in the range of 0–9% (v/v). Preliminary experiments identified this concentration range as being inhibitory, but not lethal, for the wild type (data not shown).

Wild type, tsl1Δ and nth1Δ strains all had increasingly long lag periods and lower growth rates as the ethanol concentration increased from 6% to 9% (v/v) (Fig. 2). There was no difference in the growth profiles of these three strains under each of the conditions used. In the absence of ethanol stress, none of these cultures had a detectable lag period and all three strains had similar specific growth rates. In the presence of 6–8% (v/v) ethanol, all three strains had similar lag periods and specific growth rates at each ethanol concentration (data not shown). There was no clear exit from the lag period by any of these cultures after a 12-h incubation in 9% (v/v) ethanol, with the viable cell population remaining constant at around 5 × 106 cells mL−1. The difference in intracellular trehalose concentration had no apparent impact on the ability of the cells to adapt to sublethal ethanol concentrations or on their subsequent growth rate in the presence of ethanol.

Figure 2.

 The effect of sublethal ethanol concentrations on the viable cell population of Saccharomyces cerevisiae wild-type, tsl1Δ (a) and nthl1Δ (b) cultures. Late exponential phase cells were inoculated into defined medium in the absence (•, ○) or presence of added ethanol at (v/v) 6% (♦, ⋄), 7% (▴, ▵), 8% (▪, □) or 9% (▾, ▿). Wild-type and deletion cultures are represented by closed and open symbols, respectively. The cultures were incubated at 30°C/130 r.p.m. Error bars indicating SD from the mean of triplicate batch cultures are shown for one time point.

Competitive growth at sublethal ethanol concentrations

The ability of conventional batch growth experiments to differentiate phenotypes is limited because small differences in cell vitality may not be detected. Knowledge of whether trehalose influences the yeast growth rate during ethanol stress, no matter how small that influence may be, is important if the role of trehalose and its mode of action in the ethanol tolerance of yeast are to be established. Competition cultures are capable of detecting small differences in the growth rate of strains that would not normally be observed using conventional batch growth experiments (Baganz et al., 1998; Thatcher et al., 1998). With this in mind, competitive growth experiments were used in this work to determine whether there are marginal differences in the growth rates (competitive fitness) of the wild type and tsl1Δ or nthl1Δ strains during sublethal ethanol stress.

The competitive growth experiments comprised equal cell populations of wild type and knockout strains incubated together in a flask containing defined medium with 7% (v/v) added ethanol, and with the population being subcultured after every six generations; a control experiment comprised the strains grown in competition, but with no added ethanol. The relative population of each strain in the coculture was determined by viable cell counts using YEPD and YEPD-Geneticin plates. The wild type, tsl1Δ and nthl1Δ strains were indistinguishable in their competitive fitness over 36 generations in the presence or absence of 7% (v/v) ethanol (data not shown). This suggests that the relative amount of intracellular trehalose affords no competitive advantage for S. cerevisiae growing at sublethal ethanol concentrations.

Survival at lethal ethanol concentrations

Given that many studies on the role of trehalose in the ethanol tolerance of yeast use lethal ethanol concentrations, it was decided to investigate the ethanol tolerance of the wild type and null mutants in the presence of lethal ethanol concentrations, the premise being that different biochemical and physical mechanisms may be involved in survival when exposed to lethal ethanol concentrations, compared with acclimation and growth at sublethal ethanol concentrations. Late exponential phase cells were inoculated into a defined medium in the absence or presence of 10%, 12%, 14%, 15%, 16% or 18% (v/v) added ethanol. Samples were taken at regular time intervals and the viable cell population was monitored by duplicate plate counts.

All null deletion cultures in the survival experiments grew similarly to the wild type in the absence of added ethanol (data not shown). There was no change in the viable cell population of wild type, tsl1Δ and nth1Δ over a 12-h period at an ethanol concentration of 10% (v/v) (Fig. 3); when these strains were exposed to ethanol concentrations >10% (v/v), the viable cell population began to decrease over time. The tsl1Δ strain was more sensitive than the wild type to lethal ethanol conditions, i.e. the viable cell population of tsl1Δ decreased at a substantially higher rate (Fig. 3). When the nth1Δ strain was exposed to lethal ethanol concentrations, it had a higher survival rate compared with the wild type and the difference in survival rate was even higher when compared with that of strain tsl1Δ (Fig. 3). It appears that higher intracellular trehalose concentrations improve the resilience of the cell when incubated in the presence of lethal ethanol concentrations.

Figure 3.

 The effect of lethal ethanol concentrations on the viable cell population of Saccharomyces cerevisiae wild type, tsl1Δ (a) and nth1Δ (b) strains. Late exponential phase cells were inoculated into defined medium containing added ethanol at (v/v) 10% (•, ○), 12% (♦, ⋄), 14% (▴, ▵), 15% (▪, □), 16% (▾, ▿) or 18% (inline image). Wild type and deletion cultures are represented by closed and open symbols, respectively. The cultures were incubated at 30°C/130 r.p.m. Error bars showing the SD from the mean of triplicate batch cultures are shown for one time point.

Effect of growth phase and trehalose concentration on survival rate

It is likely that S. cerevisiae metabolism is compromised in the presence of lethal ethanol concentrations to the extent that the cell is not capable of producing significant amounts of trehalose. Differences between strain survival rates when exposed to lethal ethanol concentrations are probably due to the differences in the amount of trehalose in the strains at the time of inoculation. To investigate this further, tsl1Δ and nth1Δ inocula were harvested at different growth stages and inoculated into medium containing 16% (v/v) ethanol; the viable cell populations were compared over time. The inoculum cultures of both strains were grown in defined medium and inocula were harvested at late exponential (8 h), early stationary (10 h) or late stationary (13 h) phases.

The results show that the inoculum age (at least up to the late stationary phase) influences the survival of cells exposed to lethal ethanol concentrations. Early and late stationary phase inocula had improved survivability (low death rate) compared with late exponential phase inocula (higher death rates) (Fig. 4). The nth1Δ strain always had a higher survival rate than tsl1Δ regardless of inoculum age.

Figure 4.

 The effect of inoculum from different stages of growth on the viable cell population of tsl1Δ and nth1Δ cultures in the presence of a lethal ethanol concentration. Cells from late exponential (▪, □), early stationary (▴, ▵) and late stationary (•, ○) phase parent cultures were inoculated into a defined medium containing 16% (v/v) ethanol. Closed and open symbols represent tsl1Δ and nth1Δ cultures, respectively. The cultures were incubated at 30°C/130 r.p.m. Error bars showing the SD from the mean of triplicate batch cultures are shown for one time point.

Discussion

The role of trehalose in the ethanol tolerance of S. cerevisiae has been a subject of debate, with some reports of positive effects (Kim et al., 1996; Mansure et al., 1997; Pataro et al., 2002; Jung & Park, 2005; Vianna et al., 2008) and others finding that trehalose has no influence (Lewis et al., 1997; Ribeiro et al., 1999; Gomes et al., 2002). A number of these studies examined the relationship between intracellular trehalose levels and ethanol tolerance by comparing natural and industrial yeast strains, with some reporting a correlation between trehalose levels and ethanol tolerance (Mansure et al., 1997; Pataro et al., 2002; Vianna et al., 2008) while others did not (Lewis et al., 1997; Ribeiro et al., 1999; Gomes et al., 2002). The lack of consensus across these investigations may be due to the diversity in genetic background of the strains used within each experiment. Experiments that compare trehalose levels and corresponding ethanol tolerance across different strains are unable to distinguish the effects of trehalose from those of other factors that assist the cell to cope with ethanol stress, such as heat shock proteins or differences in plasma membrane composition; the influence of such factors on ethanol tolerance are likely to vary from strain to strain.

Using S. cerevisiae BY4742 and knockout strains in this background, tsl1Δ and nth1Δ, trehalose was found to play a role in ethanol tolerance, with this role confined to yeast exposed to lethal ethanol concentrations. Differences in intracellular trehalose content did not influence the accumulation rate of S. cerevisiae to sublethal ethanol concentrations or the growth rate of such cultures. Competitive growth experiments are extremely sensitive at detecting slight differences in growth rates of competing strains, yet such experiments could not separate S. cerevisiae strains with differences in intracellular trehalose content of around 19–20% between the wild type and knockout strains, and around 40% between the two knockout strains, in the presence of 7% (v/v) ethanol.

Trehalose does, however, influence S. cerevisiae survival in the presence of lethal ethanol concentrations. Inocula containing cells with higher intracellular trehalose levels had higher survival rates compared with cells carrying less trehalose. The amount of trehalose produced during incubation at lethal ethanol concentrations was not determined, but it is expected to be relatively small, especially at higher ethanol concentrations where metabolic function is severely compromised. Supporting this outcome was the effect of inoculum age on yeast survival rate. Late stationary phase cells had higher trehalose levels and better survival rates in 16% (v/v) ethanol than late exponential phase cells, noting that other factors may play a role in the survival of stationary phase inocula, such as the switching on of heat shock protein synthesis. One important outcome is that regardless of inoculum age, strain nth1Δ always had the highest survival rate compared with strain tsl1Δ, with the former having higher intracellular trehalose concentrations.

Few studies have used closely related S. cerevisiae strains to investigate the relationship between trehalose content and ethanol tolerance (Kim et al., 1996; Soto et al., 1999; Pereira et al., 2001; Jung & Park, 2005). All of these investigations reported a positive correlation between trehalose accumulation and ethanol tolerance, with three studies measuring strain survival in lethal ethanol concentrations of 18% (v/v) (Kim et al., 1996), 15% (v/v) (Soto et al., 1999) and 10% (v/v) (Pereira et al., 2001), which agrees with our findings. Jung & Park (2005) found that a S. cerevisiae strain with decreased ATH1 expression had a higher trehalose level and growth rate in 8% (v/v) ethanol than the wild type, suggesting that a sublethal ethanol concentration was used. However, the high initial cell populations used (around 1 × 108 CFU mL−1), the extremely low doubling times of the nonstressed control cultures (c. 6 h) and very small population differences between the control and experimental cultures suggest that stressors other than ethanol were impacting on cell vitality and growth.

Investigations on ethanol tolerance using genome-wide screens mostly agree with our findings with trehalose metabolism-associated genes being important for ethanol tolerance at high ethanol concentrations (Kubota et al., 2004), but not during sublethal ethanol stress (Kubota et al., 2004; Fujita et al., 2006). van Voorst et al. (2006) identified a reduced growth phenotype for a TPS1 null mutant during sublethal ethanol stress and, although we cannot account for this result, it may be related to the tps1Δ strain having a growth defect on fermentable carbon sources such as glucose, as described earlier (González et al., 1992; Neves et al., 1995; Bell et al., 1998).

The finding that trehalose influences the ethanol tolerance of S. cerevisiae only when exposed to lethal ethanol concentrations can shed light on previously suggested roles for trehalose in stress tolerance; the two main proposed roles being stabilizing the plasma membrane and protecting cytoplasmic proteins by assisting in their proper folding and repair, and preventing protein aggregation (Mansure et al., 1994; Singer & Lindquist, 1998; Sola-Penna & Meyer-Fernandes, 1998; Simola et al., 2000). The disruptive effects of ethanol on the plasma membrane occur at sublethal ethanol concentrations as low as 4% (v/v), making it unlikely that the ethanol tolerance role of trehalose is associated with interference of membrane function (Juroszek et al., 1987; Pascual et al., 1988; Petrov & Okorokov, 1990; Marza et al., 2002; Aguilera et al., 2006). Ethanol is known to inhibit cytosolic protein function in S. cerevisiae only at ethanol concentrations of 10% (v/v) or higher (Pascual et al., 1988; Jones, 1990; Hallsworth et al., 1998; Walker, 1998; Sebollela et al., 2004), with some reports demonstrating that trehalose protects enzyme integrity at high ethanol concentrations (Sebollela et al., 2004). A role for trehalose in protecting cytosolic proteins from ethanol damage, facilitating cell survival at high ethanol concentrations, is the most likely model to account for our results.

This study demonstrated a role for trehalose in protecting S. cerevisiae against the damaging effects of high ethanol concentrations, but it does not have a protective role in the presence of sublethal ethanol concentrations. The ethanol concentration-dependent nature of the trehalose effect on S. cerevisiae ethanol tolerance supports a mechanism for trehalose in protecting cytosolic proteins from ethanol-related damage. Although a role for trehalose in protecting the membrane stability cannot be discounted, it does not appear to influence cell fitness during growth at sublethal ethanol concentrations.

Acknowledgement

The authors are grateful to the Australian Government for providing an Australian Postgraduate Award to support A.B.

Ancillary