Plasma membrane integrity, ability to transport substrates and maintenance of homeostasis represent obligatory requirements for efficient ethanol production by Saccharomyces cerevisiae. The effect of ethanol on water diffusion through the bilayer and on mediated water movements was evaluated by stopped flow spectroscopy. Ethanol stimulated water diffusion and inhibited mediated water transport. In a strain overexpressing AQY1, the activation energy for water transport increased progressively (from 5.9 to 12.7 kcal mol−1) for increasing ethanol concentrations (up to 12% v/v), indicating that mediated water transport lost importance as compared with water diffusion through the bilayer. The effect of ethanol on proton movements (inward by passive diffusion and outward through the PMA1 H+-ATPase) was evaluated by measuring the rate of extracellular alcalinization and acidification of unbuffered cell suspensions at different temperatures. Above 10% ethanol, H+ diffusion was strongly increased at 30 °C, but no effect was observed at 20 °C up to 12%, indicating the existence of a threshold above which ethanol has a marked effect. On H+ extrusion, ethanol had no effect at 20 °C, but induced a monotonous decrease at higher temperatures. Our results support the view that above a threshold of ethanol concentration, the membrane structure is disrupted, becoming very leaky to H+.
Yeasts conduct alcoholic fermentation and are of strong economic interest, and besides all the advantages for food and beverages, we are now dealing with fermenting yeasts as a potential tool with good prospects in the energy field (Wald, 2007). It is obvious that the construction of ethanol-resistant strains or the manipulation of environmental factors that increase resistance to this alcohol may contribute to the overall optimization of the process (Zhao & Bai, 2009). Any small improvement may have a decisive impact on the delicate balance of the economical efficiency of ethanol production.
Around 20 years ago, attention was paid to the effect of ethanol on membrane processes (van Uden, 1985). Experiments were performed by measuring the initial transport rates of substrates and it was shown that ethanol had an exponential inhibitory effect on the uptake of hexoses (Leão & van Uden, 1982), maltose (Loureiro-Dias & Peinado, 1982), amino acids (Leão & van Uden, 1984a) and ammonium (Leão & van Uden, 1983). In these studies, assays were also performed with other alkanols and good correlations were established between the partition coefficient of these compounds and their effect on transport, supporting the view that their action was mediated by their presence in the lipid bilayer (van Uden, 1985). This effect of inhibition on mediated transport has been confirmed by several subsequent studies (Sousa, 1996; Santos, 2008).
Ethanol had the opposite effect when transport occurred through the lipid bilayer: nonmediated diffusion processes were stimulated by ethanol and other alkanols. This was first shown for H+ passive diffusion (Leão & van Uden, 1984b), and it was also shown for the uptake of hydrophobic weak acids (Henriques, 1997). Stimulation of H+ passive diffusion was also demonstrated in the bacterium Oenococcus oeni during malolactic fermentation (Da Silveira, 2002).
In most organisms, water crosses the plasma membrane by two parallel pathways: (1) the lipid bilayer with a high activation energy (Ea) for transport and lower osmotic permeability coefficients (Pf), and (2) the channel pathway (through aquaporins), with a low Ea and higher Pf values. The activation energy for osmotic water transport through the membrane is the result of the contribution of these two pathways and depends mainly on the number and activity of the channels present in the membrane (Verkman, 2000). In the Saccharomyces cerevisiae genome, two aquaporins were found (AQY1 and AQY2), and the relevance of their activity in yeast physiology has been under debate (Bonhivers, 1998; Tanghe, 2006). A role for aquaporins in industrial fermentations and the effect of ethanol on their performance were never considered in the literature.
As for H+, the movements outwards in yeast are driven by an H+-ATPase (PMA1), the most abundant protein in the plasma membrane of S. cerevisiae. This ATPase is the main mechanism responsible for cytosolic pH control and for creating proton motive force across the plasma membrane (Goffeau & Slayman, 1981). The movements inwards are not so well characterized. H+ are taken up by symport together with substrates, but it has been assumed that H+ also cross the membrane passively through the bilayer (Leão & van Uden, 1984b).
In the present work, we applied the fluorescence stopped-flow technique to evaluate the effect of ethanol and temperature on S. cerevisiae mediated and passive water transport. The osmotic water permeability coefficient (Pf) and the activation energy (Ea) were calculated for aqy1aqy2 double deletion and AQY1 overexpressing S. cerevisiae strains. We also monitored the combined effect of ethanol and temperature on S. cerevisiae H+ movements, following the external medium acidification or alkalinization, in cells subjected to different experimental conditions. Results are discussed taking into account recent reports on different mechanisms proposed for diffusion of water and protons through lipid bilayers.
Materials and methods
The strains used in this work were S. cerevisiae 10560-6B MATαleu2∷hisG trp1∷hisG his3∷hisG ura3-52; S. cerevisiae 10560-6B/pYX012 KanMX (further indicated as the parental strain); S. cerevisiae 10560-6B/pYX012 KanMX AQY1-1 (further indicated as the AQY1 overexpressing strain); and S. cerevisiae 10560-6B aqy1∷KanMX4 aqy2∷HIS3 (further indicated as the aqy1aqy2 double-deletion strain). These strains were kindly provided by Prof. Patrick Van Dijck, Institute of Botany and Microbiology, Katholieke Universiteit Leuven and Flanders Interuniversity Institute for Biotechnology (VIB), Kasteelpark Arenberg, Belgium.
Water fluxes assays
Growth conditions and fluorophore loading
Cells were grown in YPD medium (1% w/v peptone, 0.5% w/v yeast extract, 2% w/v glucose), with orbital shaking, at 28 °C. Exponential-phase cells (OD640 nm≈1) were harvested by centrifugation (10 000 g; 3 min; 4 °C), washed and resuspended in ice-cold 1.4 M sorbitol (3 mL g−1 wet weight). Cells were preloaded for 10 min at 30 °C with the nonfluorescent precursor 5-(and-6)-carboxyfluorescein diacetate (CFDA, 1 mM in an isosmotic solution) which is cleaved by nonspecific esterases generating the fluorescent form, expected to remain mainly in the cytoplasm. Although some of the probe may be either accumulated in the vacuole or exported to the medium (Breeuwer & Abee, 2000), this effect should not affect our results because the duration of each experiment was very short (3 s). Before the osmotic challenges, a 1 : 10 dilution was incubated for 5 min in an isosmotic resuspension buffer with ethanol solution in the desired concentrations. To avoid pH interference in fluorescence, cell suspensions and osmotic solutions were prepared in potassium citrate buffer (50 mM, pH 5).
Cell volume determination
Equilibrium cell volumes were obtained by loading cells with CFDA under an epifluorescent microscope (Olympus BX51) equipped with a digital camera. Cells were assumed to have a spherical shape with a diameter calculated as the average of the maximum and minimum dimensions of each cell. Sorbitol osmotic shocks of increasing tonicity were imposed on a microscope slide and an average of six pictures with four to six cells each was taken before (V0) and within 10–40 s after the osmotic challenge (V∞). The tonicity of the osmotic shock is defined as the ratio of the final to the initial osmolarity of the outside medium [Λ=(osmout)∞/(osmout)0].
The stopped-flow technique was used to monitor cell volume changes, in cells loaded with a concentration-dependent self-quenching fluorophore (Soveral, 2007). Experiments were performed on a HI-TECH Scientific PQ/SF-53 stopped-flow apparatus, which has a 2-ms dead time, controlled temperature, interfaced with an IBM PC/AT compatible 80386 microcomputer. Experiments were performed at temperatures ranging from 10 to 35 °C. Four runs were usually stored and analyzed in each experimental condition. In each run, 0.1 mL of cell suspension [initial osmolarity (osmout)0=1.4 M] was mixed with an equal amount of hypo- or hyperosmotic sorbitol solutions to reach different out- or inwardly directed solute gradients. Fluorescence was excited using a 470-nm interference filter and detected using a 530-nm cut-off filter. The time course of volume change was followed by fluorescence quenching of the entrapped fluorophore (CF). The recorded fluorescence signals were fitted to a single exponential from which the rate constant (k) was calculated.
Pf and Ea evaluation
The osmotic water permeability coefficient, Pf, was estimated from the linear relationship between Pf and k (van Heeswijk & van Os, 1986), Pf=k(V0/A)(1/Vw(osmout)∞), where Vw is the molar volume of water, V0/A is the initial volume to area ratio of the cell population, and (osmout)∞ is the final medium osmolarity after the osmotic shock. The activation energy (Ea) of water transport was evaluated from the slope of the Arrhenius plot (ln Pf as a function of 1/T).
Proton movements assays
Cells were grown in mineral medium K with 2% w/v glucose, at 28 °C, with orbital shaking (van Uden, 1967). Briefly, the medium contained (per L) (NH4)2SO4, 5 g; KH2PO4, 5 g; MgSO4·7H2O, 0.5 g; and CaCl2·2H2O, 0.132 g (dissolved separately). The medium was supplemented with 0.5 mL of each of three solutions: vitamins (biotin, 0.01 g; calcium pantothenate, 0.8 g; inositol, 40 g; niacin, 1.6 g; pyridoxin·HCl, 1.6 g; thiamine·HCl, 1.6 g in H2O, 1 L), solution A (H3BO3, 1.0 g; KI, 0.2 g; Na2MoO4·2H2O, 0.4 g in H2O, 1 L) and solution B (CuSO4·5H2O, 0.08 g; FeCl3·6H2O, 0.4 g; MnSO4·4H2O, 0.8 g; ZnSO4·7H2O, 0.8 g in 1 mM HCl, 1 L). The salt solutions were sterilized in the autoclave. Sugars and vitamins were filter sterilized separately and added to the medium to obtain the desired concentrations. Exponential-phase cells (OD640 nm≈1) were harvested (10 000 g; 3 min; 4 °C) and washed twice with ice-cold water. Cells were resuspended in water at a concentration of 50 mg wet weight mL−1 and kept on ice.
Proton movements were measured by recording the pH of unbuffered cell suspensions with a standard pH meter (PHM 82; Radiometer) connected to a potentiometer recorder (BBC-GOERZ METRAWATT, SE460). The pH electrode was immersed in a 2-mL capacity water-jacketed cell, kept at different temperatures and with magnetic stirring. Distilled water, ethanol and cell suspension were mixed in the chamber to a final volume of 1 mL.
Passive proton influx
The initial pH (between 5 and 6) was brought to four by the addition of HCl (100 mM). The time course of proton uptake was immediately monitored through the rate of alkalinization and calibrated with 10 mM HCl. The prior addition of 2-deoxy-d-glucose (DOG) (1 mM) and antimycin (2 μg mL−1) prevented the H+ movements created by the plasma membrane ATPase activity (Leão & van Uden, 1984b).
The initial pH was adjusted to 5 using HCl (100 mM) and a baseline was established. The addition of 2% w/v glucose triggered H+ efflux, leading to an acidification of the extracellular environment. The rate of acidification, calibrated with 10 mM HCl, was taken as a measure of proton extrusion activity.
Effect of ethanol on water fluxes
In order to study the effect of ethanol on yeast water fluxes, we measured the osmotic water permeability and evaluated the activation energy values (Ea) for an aqy1aqy2 double-deletion strain and an AQY1 overexpressing S. cerevisiae strain. Different phenotypes for water transport associated with the different levels of aquaporin expression were reported previously for these strains (Soveral, 2006).
Figure 1 shows the stopped-flow records obtained for the aqy1aqy2 double-deletion strain at 35 °C. Intact cells were loaded with the fluorescent dye CFDA (Soveral, 2007) and subjected to isosmotic and hyperosmotic shocks, either in the presence or in the absence of ethanol. The curves are identified by the tonicity of the osmotic shock (Λ), defined as the ratio of the final to initial osmolarity of the outside medium [Λ=(osmout)∞/(osmout)0]. These traces show that the presence of ethanol accelerated the efflux of water, the final equilibrium volume being reached earlier.
Stopped flow fluorescence signals obtained for the aqy1aqy2-deleted strain. Cells were loaded with CFDA (1 mM in an isosmotic solution) and subjected to hyperosmotic shocks (Λ=1.25) at 35°C (308 K), in the presence or absence of ethanol 15% v/v. As a negative control, cells were loaded with CFDA (1 mM in an isosmotic solution), under the same ethanol conditions, and were subjected to an isosmotic solution (Λ=1.0). (Representative signals of experiments, repeated at least three times).
Equivalent traces were obtained for the aqy1aqy2 double deletion and the overexpressing AQY1 strains for a broad temperature range (10–30 °C) (data not shown). Rate constants (k) derived from these traces were used to estimate the osmotic water permeability coefficients (Pf) based on the linear relationship between Pf and k (van Heeswijk & van Os, 1986) (see Materials and methods). Equilibrium volumes of cells loaded with CFDA were also measured under an epifluorescence microscope to determine the Pf values (see Materials and methods). Figure 2a shows that ethanol enhances the Pf values in the aqy1aqy2 double-deletion strain, this effect being more pronounced for higher temperatures. Because the aqy1aqy2 double-deletion strain only displays the contribution of the lipid bilayer, these results point to the stimulation of the passive water transport by ethanol, especially at higher temperatures. As for the AQY1 overexpressing strain, it was observed that low temperatures together with ethanol resulted in a significant reduction of the Pf values (Fig. 2b). However, the higher the temperature, the narrower the difference between Pf values obtained in the presence or in the absence of ethanol. In fact, at 35 °C, a stimulation of the water fluxes by ethanol occurred (the results for intermediate ethanol concentrations are not shown).
Effect of ethanol on the osmotic water permeability coefficients (Pf) of aqy1aqy2 double deletion (a) and AQY1 overexpressing (b) strains, at different temperatures. The symbols represent the data regarding the presence (x– 15% v/v and 12% v/v for the aqy1aqy2 and AQY1 overexpressing strains, respectively) and the absence of ethanol (◊).
The Ea values for water transport for the different experimental conditions were estimated from Arrhenius plots (Table 1). The Ea value for the aqy1aqy2 double-deletion strain was approximately 15 kcal mol−1, regardless of the concentration of ethanol, whereas for the AQY1 overexpressing strain, the higher the ethanol concentration, the higher the Ea value.
Activation energy (Ea) values obtained from the Arrhenius plots for the aqy1aqy2 double deletion and AQY1 overexpressing yeast strains incubated in different ethanol concentrations
ethanol (% v/v)
Ea (kcal mol−1)
ethanol (% v/v)
Ea (kcal mol−1)
At least six pairs of values of temperature and permeability were considered for the determination of each slope.
Effect of ethanol on proton fluxes
Figure 3 illustrates the typical changes in the pH values of S. cerevisiae cell suspensions upon fast acidification (a) or upon the addition of a glucose pulse (b). Proton inward movements were registered after the application of an acid pulse to a cell suspension. The sudden decrease of the external pH (from 5–6 to 4) was followed by a steady increase, corresponding to the H+ influx into the cells (Leão & van Uden, 1984b). To minimize H+ efflux through the ATPase, cells were ATP-depleted using antimycin and DOG. Antimycin is an inhibitor of the respiratory chain and DOG is phosphorylated after entering the cell, but not further metabolized (Loureiro-Dias & Santos, 1989). Each proton influx record was calibrated by the addition of a known amount of HCl.
Schematic representation of the pH changes observed during proton fluxes measurements for (a) proton influx and (b) proton efflux. The rate of passive influx was calculated from the slope of trace (a), after acidification to pH 4. The rate of proton efflux was calculated from the maximum slope of curve (b) a few seconds after the addition of glucose.
Figure 4 shows the effect of ethanol together with temperature on the passive proton influx. Up to 6% v/v of ethanol, only the effect of temperature on the increase of proton influx was observed. For higher ethanol concentrations, an increasingly higher proton influx was observed and this was more pronounced at the maximum temperatures assayed.
Effect of ethanol and temperature on the rates of passive proton influx in the Saccharomyces cerevisiae parental strain. The symbols represent the experimental temperatures: 25°C (◻), 30°C (△) and 35°C (○). (Points represent the mean values of at least two experiments).
Protons are usually present at higher concentrations in the extracellular environment and yeast cells rely on the proper functioning of the proton pump ATPase to maintain the intracellular environment close to neutral. When glucose is added to S. cerevisiae cells, protons are extruded after a delay of 10–30 s and the external pH declines (Fig. 3b). In this work, proton outward movements were monitored when a glucose pulse was added to a cell suspension and the rate constants were calculated from the maximum instant slope of the records. The first relevant observation is that in the absence of ethanol, temperature affected the efflux remarkably (Fig. 5). This stimulating effect of temperature was suppressed by the increase of the ethanol concentration. At 42 °C, for concentrations above 10% v/v, proton efflux was not observed.
Effect of ethanol and temperature on the rates of the net proton efflux for the Saccharomyces cerevisiae parental strain. The symbols represent the experimental temperatures: 20°C (◊), 25°C (◻), 30°C (△), 35°C (○) and 42°C (▲).
The results described in this work on the effect of ethanol on water fluxes through the yeast plasma membrane agree very well with our previous knowledge on water transport. In the aqy1aqy2 deletion strain, water crosses the membrane exclusively through the lipid bilayer. In this case, the activation energy of the process was high and was not affected by the presence of ethanol (the mechanism does not change). It is remarkable that, as Fig. 2 shows, in the presence of ethanol, the values for permeability increased. Ethanol will interact with components of the membrane, which becomes leakier. This effect is potentiated by temperature, as expected. In the strain overexpressing AQY1, the activation energy for water transport was low in the absence of ethanol, but this value progressively increased with the concentration of ethanol. This means that the mediated transport was inhibited, while the diffusion component was stimulated. It is interesting to notice that a low concentration of ethanol (4% v/v) had a remarkable effect on the activation energy, suggesting that AQY1 is strongly inhibited. At a low temperature, the inhibitory effect dominated and a decrease in water flux was observed, but at the highest temperature tested (35 °C), the inhibitory effect was masked by the stimulatory effect on the diffusion process. These results support the idea that aquaporins will certainly play a poor role in yeast in the presence of high concentrations of ethanol.
Proton homeostasis is a key mechanism for good yeast performance for both appropriate intracellular pH and proton motive force across membranes. Our results indicate that ethanol enhanced the diffusion process for rather high concentrations and this phenomenon was much more evident at higher temperatures. Cells were grown at 28 °C and these cells were quite impermeable to protons in assays performed at 25 °C. However, in assays performed at 30 °C, H+ permeability increased and ethanol became more effective. The shape of the curve suggests that the integrity of the membrane was not affected up to rather high concentrations (10% v/v), but above a certain threshold, a strong increase in H+ permeability occurred. The strain tested by Leão & van Uden (1984b) seemed to be much more sensitive to ethanol, but the results presented in our work are similar to those previously described for another strain of S. cerevisiae and for Zygosaccharomyces bailii (Quintas, 2000). We are now aware that temperature plays a very important role. Above a threshold of temperature and of ethanol concentration, the cells would be invaded by protons. These results may be interpreted in the light of the phenomenon of phospholipid interdigitation. Above the threshold of ethanol concentration, instead of a bilayer, phospholipids would form nonbilayer structures (Gurtovenko & Anwar, 2009), and the homeostasis would be severely compromised. Zeng (1993) attributes the abrupt changes observed in H+ permeability upon increase of ethanol concentration to interdigitation. According to Dickey (2009), ergosterol plays a fundamental role in the prevention of interdigitation in yeast membranes.
The ability of the yeast to acidify the environment (acidification power) has long been considered as an indication of good fermentative activity (Opekarova & Sigler, 1982; Jimenez & van Uden, 1985; Sigler, 2006). When we measure the effect of ethanol on proton extrusion through the extracellular acidification rate, a complex set of events is being considered: glucose transport and metabolism and ATPase activity and regulation events. As a whole, proton extrusion was stimulated by temperature and the inhibitory effect of ethanol was also enhanced by temperature. In this aspect, our work just illustrates the previous idea that the deleterious effect of ethanol is potentiated by high temperature.
In recent years, several studies have been performed on yeast ethanol tolerance in global (omics) perspectives. Authors compared strains with different tolerances (Zuzuarregui, 2006; Dinh, 2009) or compared cells under different ethanol stress conditions (Albers & Larsson, 2009; Zhao & Bai, 2009). These studies represent a wealth of information to be taken into consideration when more focused studies are performed. Shobayashi (2005) reported that the genes responsible for ergosterol synthesis are overexpressed under brewing conditions in the ethanol-tolerant yeast. Certainly, these results fit with an important role for membrane fluidity and permeability that may play a definitive role in ion homeostasis. Also, the PMA1-ATPase has been reported in some studies (Aguilera, 2006; Lei, 2007). Both H+ extrusion and H+ permeability are certainly involved in H+ homeostasis.
This work was supported by Fundação para a Ciência e a Tecnologia (Project POCTI/AGR/47891/2002 and Project POCTI/AGR/57403/2004 and grant SFRH/BPD/32511/2006 to L.L.).