Correspondence: Olga Kinclova-Zimmermannova, Institute of Physiology, Academy of Sciences of the Czech Republic, Videnska 1083, 142 20 Prague 4, Czech Republic. Tel.: +420 241 062 673; fax: +420 241 062 488; e-mail: email@example.com
There are three different sodium transport systems (Ena1-4p, Nha1p, Nhx1p) in Saccharomyces cerevisiae. The effect of their absence on the tolerance to alkali-metal cations and on the membrane potential was studied. All three sodium transporters were found to participate in the maintenance of Na+, Li+, K+ and Cs+ homeostasis. Measurements of the distribution of a fluorescent potentiometric probe (diS-C3(3) assay) in cell suspensions showed that the lack of all three transporters depolarizes the plasma membrane. The overexpression of the Na+,K+/H+ antiporter Nha1 resulted in the hyperpolarization of the plasma membrane and consequently increased the sensitivity to Cs+, Tl+ and hygromycin B. This is the first evidence that the activity of a Na+,K+/H+ antiporter could play a role in the homeostatic regulation of the plasma membrane potential in yeast cells.
Generation and maintenance of the membrane potential is one of the crucial tasks of any living cell. In Saccharomyces cerevisiae, the plasma membrane potential is mainly created by the electrochemical gradient of protons generated by Pma1 H+-ATPase (Serrano et al., 1986). This potential is indispensable for nutrient uptake and for the regulation of intracellular pH (Serrano, 1988). On the other hand, the potassium uptake systems Trk1p and Trk2p are major expenders of membrane potential and prevent excessive hyperpolarization of the yeast plasma membrane (Madrid et al., 1998; Kuroda et al., 2004). The steady-state value of the electrochemical membrane potential set by Pma1p, Trk1p and Trk2p is also necessary for the maintenance of cation homeostasis, as hyperpolarization of the plasma membrane promotes the non-specific influx of monovalent cations into yeast cells (Madrid et al., 1998; Mulet et al., 1999; Goossens et al., 2000; Navarre & Goffeau, 2000).
The Nha1 antiporter is a typical secondary active transporter of the yeast plasma membrane and, as well as in the elimination of toxic cations, it is involved in other functions, such as regulation of the cell cycle (Simon et al., 2001, 2003) and intracellular pH (Sychrova et al., 1999; Brett et al., 2005) or in the immediate cell response to an osmotic shock (Kinclova et al., 2001; Proft & Struhl, 2004). The Nha1 protein (985 amino acids long) has 12 transmembrane domains and a long hydrophilic C-terminal domain (56.2% of the protein). Most of the Nha1p C-terminus (amino acids 473-985) is neither important for the localization of the antiporter in the plasma membrane nor for its substrate specificity (Kinclova et al., 2001), but is important for Na+ and Li+ transport velocity and participates in the regulation of the intracellular potassium content and the cell cycle (Kinclova et al., 2001; Simon et al., 2001, 2003).
In this work, we prepared a set of isogenic strains carrying combinations of ena1-4, nha1 and/or nhx1 null mutations. Mutant strains were tested for their tolerance to different alkali metal cations or hygromycin B, and for the phenotype of Nha1p overexpression. The overexpression of the plasma membrane Na+, K+/H+ antiporter rendered the cells sensitive to caesium and hygromycin B. By measuring the distribution of a fluorescent potentiometric probe in cell suspensions, we (1) compared the membrane potential of yeast strains lacking sodium transport systems, and (2) tested whether the sensitivity of cells overexpressing Nha1p to toxic cations is due to hyperpolarization of the plasma membrane, and (3) whether the long Nha1p C-terminus is involved in this process.
Materials and methods
Yeast strains, plasmids and media
The yeast strains used are listed in Table 1. The AXT3 triple mutant (ena1-4Δnha1Δnhx1Δ) (Quintero et al., 2000) was white and grew better on standard media than its parental strain W303-1B, though both strains harbour the ade2 mutation, causing the accumulation of a toxic red pigment (Woods, 1969). Therefore, the AXT3 was crossed with W303-1A (Wallis et al., 1989). The resulting diploids were sporulated and subjected to tetrad analysis to select a series of pink haploid strains with different combinations of deletions of genes encoding sodium transporters (Table 1). The plasmids pNHA1-985 and pNHA1-472 are derivatives of YEp352 (Hill et al., 1986) and contain either the complete Saccharomyces cerevisiae NHA1 gene or its 3′-truncated version encoding 472-amino acid-long Nha1p (lacking 513 amino acids of the hydrophilic C-terminus; Kinclova et al., 2001). Yeast cells were grown at 30°C in standard YNB medium without amino acids (Difco, Detroit, MI) containing 2% glucose and appropriate auxotrophic supplements. The growth of yeast cells in the presence of Na+, Li+, K+, Cs+ chlorides, Tl+ nitrate or hygromycin B was tested on a series of YNB plates supplemented with increasing concentrations of salts or hygromycin B (200–1000 mM NaCl, 10–50 mM LiCl, 800–1600 mM KCl, 30–50 mM CsCl, 1–4 mM TlNO3, 50–250 μg mL−1 HygB). Salts of alkali metal cations and TlNO3 were added prior to, and hygromycin B after autoclaving. Growth was recorded over a period of 4 days.
Table 1. Saccharomyces cerevisiae strains used in this study
MATa leu2-3/112 ura3-1 trp1-1 his3-11/15 ade 2-1 can1-100GAL SUC2 mal10
Fluorescence measurement of membrane potential, diS-C3(3) assay
The differences in membrane potential between isogenic yeast strains were estimated by fluorescence assay based on the redistribution of the fluorescent dye 3,3′-dipropylthiacarbocyanine iodide (diS-C3(3)) in cells as described earlier (Gaskova et al., 1998, 1999). Yeast cells were cultured at 30°C in liquid YNB media to the early exponential phase (OD578≈0.2), harvested, washed with double-distilled water and resuspended in a 10-mM citrate-phosphate buffer of pH 6.0 to OD578=0.1±5%. The fluorescent dye diS-C3(3) (10-μM stock solution in ethanol) was added to 3 mL of yeast cells to produce a final concentration of 10 nM. Fluorescence emission spectra were measured on a FluoroMax 3 spectrofluorimeter (Jobin-Yvon, SPEX, Edison, NJ) equipped with a xenon lamp. The excitation wavelength was 531 nm, emission range 560–590 nm, duration of one spectral scan 20 s. Scattered light was eliminated by an orange glass filter with a cut-off wavelength at 540 nm. The staining curves, i.e. the dependence of the maximum emission wavelength λmax(t) on the duration of staining was fitted to the equation as in Malac et al. (2005), where λmax(t) reflects the actual intracellular probe concentration, λmaxeq represents the position of the fluorescence maximum at equilibrium, C2 reflects the rate of change of the fluorescence maximum and C1 reflects the difference between the starting point of the staining curve and λmaxeq, which depends on the time of the first sampling point. For every staining, the fluorescence parameter λmaxeq was determined and used for the expression of differences in membrane potential between isogenic yeast strains. For each experiment, data are reported as an average of at least three independent experiments±standard deviation (SD).
Glucose-induced medium acidification
Yeast cells were cultured at room temperature in liquid YNB media (100 mL) to the exponential phase of growth (OD600≈0.3), pelleted by centrifugation (3050 g, 4 min) and resuspended in YNB without glucose. The changes in extracellular pH were monitored at room temperature in a stirred cell suspension by a pH electrode (pH/mV/Temperature Meter, Model 3505, Jenway, Dunmow, Essex, UK). At time zero, a 2-mL aliquot of cells (OD600=10) was added to 18 mL of the YNB media without glucose, adjusted to pH 5.4 with Ca(OH)2. After 5 min, when a stable pH base line was established, 2% glucose (from 40% stock solution) was added to initiate acidification, and extracellular pH was recorded at 30-s intervals. The maximal rate of acidification (ΔpH) per minute was calculated and reported as an average of the four independent experiments ±SD.
Results and Discussion
Growth and plasma membrane potential characteristics of yeast strains lacking sodium transporters
We have constructed a series of haploid yeast strains lacking one, two or three sodium transporters (ena1-4Δ, nha1Δ, nhx1Δ), all isogenic to W303-1A (MATa leu2-3/112 ura3-1 trp1-1 his3-11/15 ade 2-1 can1-100 GAL SUC2 mal10). The effect of the deletion of particular sodium transport systems on the cell tolerance to alkali metal cations and on their membrane potential was characterized. All strains grew at identical rates on medium without salts, indicating that none of the deletions changed the growth properties of the strain (Fig. 1a). Further, we compared the growth of these mutants on a series of plates (non-buffered YNB medium, pH 4.5) supplemented with increasing amounts of NaCl, LiCl, KCl or CsCl, respectively, or with hygromycin B, which is an amino glycoside antibiotic used as an indicator of the hyperpolarized/depolarized state of the yeast plasma membrane (McCusker et al., 1987; Perlin et al., 1988).
Plasma membrane Ena ATPases and the Nha1 antiporter participate in tolerance to Na+, Li+ and K+ (Bañuelos et al., 1998), whereas the intracellular Nhx1p is required mainly for Na+ tolerance and detoxification from hygromycin B (Gaxiola et al., 1999). In agreement with published results, our GW19 strain lacking genes encoding Ena ATPases was sensitive to sodium, lithium and potassium (Fig. 1a, compare W303-1A and GW19). The growth of nha1Δ or nhx1Δ single mutants was unaffected in the presence of 200 mM NaCl or 10 mM LiCl, and these deletions had only a slight negative effect in the presence of 800 mM KCl (Fig. 1a, CW25 and AW11, respectively). Similarly, a negative effect of high KCl concentrations on cell growth was observed for nhx1 mutants in liquid media (Fukuda et al., 2004; Brett et al., 2005). On plates, the participation of the two antiporters in the tolerance to Na+, Li+ and K+ could be clearly demonstrated when genes encoding Nha1p and Nhx1p were deleted together with genes encoding Na+-ATPases. The ena1-4Δnha1Δ mutant was more sensitive to Na+, Li+ and K+ than the strain carrying ena1-4Δ alone (Fig. 1a, BW31 vs. GW19), and the nhx1 deletion in the ena1-4Δnha1Δ background further enhanced the sensitivity to Na+, Li+ and K+ (Fig. 1a, AB11c vs. BW31).
A single nha1 deletion had no visible effect on the tolerance to caesium cations (Fig. 1a, CW25). On the other hand, the absence of Ena ATPases (ena1-4Δ) or the intracellular antiporter (nhx1Δ) reduced tolerance to caesium significantly, and this effect was enhanced in the triple-deletion strain (Fig. 1a). This indicates that Ena1-4 and Nhx1 systems are involved in Cs+ detoxification.
Yeast mutants with impaired K+ uptake (trk1Δ) are hypersensitive to hygromycin B, as the reduced K+ uptake hyperpolarizes the plasma membrane potential (negative inside), which drives the influx of hygromycin B (Madrid et al., 1998; Mulet et al., 1999). On the other hand, mutations that reduce the H+-pumping activity of the plasma membrane Pma1 H+-ATPase result in the depolarization of the plasma membrane potential and confer resistance to hygromycin B (McCusker et al., 1987; Perlin et al., 1988). Besides being used as an indicator of the level of yeast plasma membrane potential, hygromycin B is an important tool for the assessment of the Nhx1 antiporter function (Gaxiola et al., 1999; Brett et al., 2005). Whereas our mutant strains lacking the two plasma membrane transporters did not show any difference in tolerance to hygromycin B compared to the wild type (Fig. 1a), the absence of the intracellular Nhx1 antiporter resulted, as expected, in a high sensitivity. Absence of all three transporters rendered cells extremely sensitive to hygromycin B (Fig. 1a).
To find out whether the sensitivity of the nhx1Δ mutant to hygromycin B and Cs+ resulted from plasma membrane hyperpolarization leading to an increased uptake of these cations, we compared the plasma membrane potential of the wild-type strain and isogenic mutants lacking alkali-metal-cation transporters (Fig. 1b). Although the exact value of electrical membrane potential of yeast cells is not known, comparative assessments of cell membrane potentials may be obtained, for instance with fluorescent cyanine dyes (Madrid et al., 1998). To assess differences in the plasma membrane potential of our strains, we used a well-established diS-C3(3) assay based on the fact that the cyanine dye diS-C3(3) enters yeast cells according to their membrane potential (Gaskova et al., 1999). Inside the cells, the probe binds to cytosolic compounds, which results in an increase in fluorescence intensity and in a red shift of the dye emission maximum (λmax). A larger red shift of λmax indicates a higher intracellular concentration of the probe (Gaskova et al., 1999). Differences in λmax attained in the equilibrium state (λmaxeq) correspond to differences in membrane potential (Cadek et al., 2004; Malac et al., 2005). As shown in Fig. 1b, there was no significant difference in λmaxeq (i.e. membrane potential status) between the control wild-type cells (W303-1A) and cells lacking either of the two plasma membrane transporters (GW19, CW25, BW31) or the intracellular Nhx1 antiporter (AW11). Only the λmaxeq of the triple mutant AB11c (ena1-4Δnha1Δnhx1Δ) was always lower by about 1 nm (Fig. 1b), indicating that the lack of all three transporters simultaneously caused a mild depolarization of cell plasma membranes. It is worth noting that under non-stressed conditions (cells exponentially growing in YNB, pH 4.5), the expression of all three systems is low (Garciadeblas et al., 1993; Bañuelos et al., 1998; Nass & Rao, 1998) and their activity is maintained at a basic level. Thus their contribution to the generation of plasma membrane potential should also be low.
Results summarized in Fig. 1 show that (1) all three transport systems, originally described as sodium transporters, participate in the cell homeostasis of Na+, Li+ and K+, (2) Ena ATPases and Nhx1 antiporter are involved also in Cs+ detoxification, and (3) the single deletion of a plasma membrane transporter or of the intracellular alkali-metal-cation transporter does not cause significant hyperpolarization of the plasma membrane in cells growing in minimal medium. Nevertheless, the lack of all three transporters together causes mild depolarization of the plasma membrane. As there was no difference in the plasma membrane potential between the wild-type strain (W303-1A) and the strain lacking intracellular Nhx1p (AW11), the high sensitivity of nhx1Δ cells to Cs+ and hygromycin B is more due to a defective sequestration of toxic cations in intracellular compartments than to their increased uptake. The mechanism by which Nhx1p confers tolerance to hygromycin B is believed to be coupled to its role in intracellular pH regulation (Brett et al., 2005).
Overexpression of Nha1p increases the membrane potential of yeast cells
The plasma membrane Nha1 antiporter was thought to mediate an electroneutral (1 : 1) exchange of alkali metal cations for protons. However, recent biochemical analysis of the molecular function of Nha1p in secretory vesicles isolated from a temperature-sensitive secretory mutant has shown that Nha1p activity is electrogenic, i.e. associated with a net charge movement across the membrane, transporting more protons per single sodium cation (Ohgaki et al., 2005). As shown in Fig. 1, the single nha1 deletion affected neither the cell tolerance to caesium or hygromycin B nor the steady-state cell plasma membrane potential, suggesting that the activity of Nha1p is not a major contributor to the generation or maintenance of membrane potential in exponentially growing non-stressed cells. Nevertheless, the overexpression of NHA1 from a multi-copy vector rendered cells significantly more sensitive to Cs+ and hygromycin B. Figure 2 shows the growth characteristics of wild-type and isogenic mutant strains carrying the nha1Δ null allele either alone (CW25), or in combination with ena1-4Δ(BW31) and nhx1Δ(AB11c) transformed with an empty vector (Y; control), or overexpressing Nha1p (N9). On plates containing the Nha1p's natural substrates Na+ and K+, the overexpression of Nha1p significantly improved the growth of cells lacking the NHA1 chromosomal copy, but the growth of all four strains overexpressing Nha1p (including the wild type) was inhibited in the presence of both 30 mM CsCl and 250 μg mL−1 hygromycin B (Fig. 2). Transformants of the hygromycin B-sensitive AB11c strain did not grow on plates containing hygromycin B at concentrations higher than 120 μg mL−1, but the effect of increased hygromycin B sensitivity due to Nha1p overexpression was observed on plates containing 100 μg mL−1 hygromycin B (not shown).
According to these results, monovalent Cs+ does not seem to be extruded by the Nha1 antiporter. Similarly, Cs+/H+ antiportation was not detected in the in vitro Nha1p study (Ohgaki et al., 2005). As mentioned above, the increased Cs+ and hygromycin B sensitivity of cells overexpressing Nha1p could reflect an enhanced influx caused by plasma membrane hyperpolarization. To verify this hypothesis, we tried to estimate the changes in plasma membrane potential caused by Nha1p overexpression.
The diS-C3(3) assay was performed with exponentially growing cells transformed either with the empty vector (control) or with pNHA1-985. A representative membrane potential-dependent diS-C3(3) staining of BW31 (ena1-4Δnha1Δ) cells is shown in Fig. 3. Throughout staining, the value of λmax (or corresponding calculated λmaxeq) of control cells was significantly lower than that of cells overexpressing the Nha1p (Fig. 3). The membrane potential could be abolished by the addition of carbonylcyanide-3-chlorophenyl-hydrazone (CCCP), upon which the λmax in both cases rapidly decreased to the level nearly identical to the λmax of the free probe (570 nm; Fig. 3). This control assay confirmed that the differences in λmax (or λmaxeq) really corresponded to differences in membrane potential status between isogenic yeast strains, and that the higher λmax (or λmaxeq) can be taken as hyperpolarization of the plasma membrane. Fig. 4 shows comparison of λmaxeq for each pair of transformed strains. Similarly as in the case of BW31 cells, in the other three strains tested, the presence of pNHA1-985 (i.e. overexpression of Nha1p) resulted in a significant increase in membrane potential; the λmaxeq of control cells was always lower than in cells overexpressing Nha1p (Fig. 4). The highest relative enhancement of the membrane potential was observed in the triple AB11c (ena1-4Δnha1Δnhx1Δ) mutant. These results confirmed our hypothesis that an increased amount of Nha1p in cells could cause hyperpolarization of the cell plasma membrane, which results in a higher sensitivity of these cells to hygromycin B and caesium.
The electrogenic activity of Nha1p determined by Ohgaki et al. (2005) should decrease the proton gradient across the plasma membrane. Moreover, the presence of Nha1p was found to enhance the kinetic parameters of Trk1p (Bañuelos et al., 2002). The resulting higher influx of H+via Nha1p and K+via Trk1p should cause a mild depolarization of the membrane. Thus, the results presented here pose a question: how could Nha1p contribute to plasma membrane hyperpolarization? Our explanation is based on the hypothesis that Nha1p's involvement in membrane potential generation is rather indirect and could be coupled to intracellular pH changes. The deletion of nha1 results in an increase in intracellular pH (Sychrova et al., 1999; Brett et al., 2005), whereas Nha1p overexpression is connected with a decrease in internal pH (Sychrova et al., 1999). It seems probable that the drop in internal pH caused by the enhanced influx of protons in cells overexpressing Nha1p can activate and/or block other efflux and influx systems that contribute to plasma membrane potential generation. The increased/decreased electrogenic activity of these systems can lead to a net loss of positive charge from inside the cell, i.e. to the observed hyperpolarization of the plasma membrane. The most probable candidate is the main generator of plasma membrane potential, Pma1 H+-ATPase, as the decrease of cytosolic pH is one of its most important activating stimuli (Benito et al., 1992). Pma1p activity estimated in secretory membrane vesicles is about five times higher at pH 6.25 than at pH 7.5 (Ohgaki et al., 2005). Hence, the increased influx of protons via Nha1p results in a drop of intracellular pH that could launch excessive outward proton pumping, hyperpolarizing the plasma membrane. The activity of Pma1 ATPase in vivo can be observed as the acidification of an aqueous suspension of cells after addition of glucose (Serrano, 1980). To strengthen our hypothesis, we compared the changes of extracellular pH after addition of glucose in a suspension of BW31 (ena1-4Δnha1Δ) cells containing the empty vector or pNHA1-985 (overexpressing the Nha1p). A higher acidification rate (ΔpH min−1) after addition of glucose was observed for cells overexpressing the Nha1p than for cells with the empty vector (0.058±0.002 vs. 0.049±0.002 pH units min−1, respectively), which indicates that the proton efflux is higher in cells overexpressing the Nha1p.
The C-terminal region of Nha1p is important for its contribution to plasma membrane hyperpolarization
In our previous work, we have found that the long Nha1p C-terminus is important for antiporter transport activity and for the regulation of the intracellular K+ content (Kinclova et al., 2001). To know whether the Nha1p C-terminus is also implicated in the hyperpolarization of the plasma membrane in cells overexpressing Nha1p, we first assessed the growth characteristics of BW31 (ena1-4Δnha1Δ) cells overexpressing the complete Nha1p or its truncated version with a short C-terminus (Nha1p 472 aa long) on YNB plates containing different monovalent cations or hygromycin B (Fig. 5). The presence of both Nha1p versions improved the tolerance to sodium and potassium, but the shorter Nha1p provided cells with a lower Na+ tolerance than the complete antiporter (Fig. 5). This corresponds to a decreased transport activity of the truncated version for Na+ (Kinclova et al., 2001). On the other hand, in the presence of Cs+ or also Tl+ (a monovalent toxic cation used usually as an analogue of K+ (Lapathitis & Kotyk, 1998)), cells with the complete Nha1p did not grow as well as cells with the truncated version of Nha1p (Fig. 5). Cells expressing the complete version were also more sensitive to hygromycin B than cells harbouring the short Nha1p (Fig. 5). Compared to control cells without any antiporter, cells with the truncated Nha1p tolerated a slightly lower amount of Cs+ and hygromycin B (Fig. 5). Similar results were observed also when the caesium and hygromycin B tolerance of the AB11c triple mutant (ena1-4Δnha1Δnhx1Δ) overexpressing both Nha1p versions was tested (data not shown).
Results obtained in drop tests indicated that cells with the short Nha1p version could be less hyperpolarized than cells with the complete Nha1p, and hence a smaller amount of toxic cations should enter these cells. To confirm this presumption, we carried out the diS-C3(3) assay with exponentially growing BW31 and AB11c cells transformed with the two versions of Nha1p or with the empty vector as a control. Though the plasma membrane of BW31 and AB11c cells expressing the truncated Nha1p is hyperpolarized compared to control cells without any antiporter, it is clearly less hyperpolarized than in cells with the complete Nha1p (Fig. 6). The λmaxeq of cells transformed with pNHA1-472 was significantly lower than of cells overexpressing the full-length Nha1p (Fig. 6). This corresponds well with the observed lower sensitivity of cells expressing Nha1p-472 to toxic cations (Fig. 5).
We have previously demonstrated that the Nha1p C-terminus is not important for the initial rate of K+ efflux or for Nha1p-mediated cell tolerance to high external concentrations of KCl (Kinclova et al., 2001). However, the activity of Nha1p with a short C-terminus retained a higher potassium content in the cells (Kinclova et al., 2001). The lower overall K+/H+ exchange activity of the truncated Nha1p could mean fewer protons entering cells, i.e. less intracellular acidification. Consequently, the response of cells (apparently activation of Pma1p) with Nha1p-472 is modest compared to that of cells with the complete Nha1p and, hence, cells overexpressing Nha1p-472 are less hyperpolarized.
In conclusion, the diS-C3(3) fluorescent dye proved to be usable for the assessment of plasma membrane potential differences in isogenic yeast strains growing under the same conditions. We have shown that neither the plasma membrane Ena-ATPase nor the Nha1 alkali metal cation/H+ antiporter are major contributors to the generation of plasma membrane potential in cells exponentially growing in standard YNB medium (the deletion of the corresponding genes does not influence the potential level). The role of these two transport systems in normally growing, non-stressed cells is more probably a fine tuning of the plasma membrane potential. Significant potential changes could be observed only upon high overexpression of the transporter. Measurements of differences in membrane potential showed that the higher sensitivity of cells overexpressing Nha1p to toxic caesium and thallium cations and to hygromycin B is due to hyperpolarization of the plasma membrane. This is the first evidence that the activity of the Nha1 alkali metal cation/H+ antiporter could influence the generation and maintenance of the plasma membrane potential. In our future work we will focus on a detailed elucidation of the mechanism by which the presence of Nha1p could participate in the homeostatic regulation of the yeast cell membrane potential.
We thank Dr F. J. Quintero for the AXT3 strain. We are grateful to Eva Urbankova for help with λmaxeq determination. This work was supported by Czech Grant agencies (GA AS CR 5011407, GA CR 204/02/D092, AV0Z 50110509) and by Research Concepts MSM 0021620835.