OUP user menu

Cnh1 Na+/H+ antiporter and Ena1 Na+-ATPase play different roles in cation homeostasis and cell physiology of Candida glabrata

Yannick Krauke, Hana Sychrova
DOI: http://dx.doi.org/10.1111/j.1567-1364.2010.00686.x 29-41 First published online: 1 February 2011


Yeasts tightly regulate their intracellular concentrations of alkali metal cations. In Saccharomyces cerevisiae, the Nha1 Na+/H+-antiporter and Ena1 Na+-ATPase, mediate the efflux of toxic sodium and surplus potassium. We report the characterization of Candida glabrata CgCnh1 and CgEna1 homologues. Their substrate specificity and transport properties were compared upon expression in S. cerevisiae, and their function characterized directly in C. glabrata. The CgCnh1 antiporter and the CgEna1 ATPase transport both potassium and sodium when expressed in S. cerevisiae. CgEna1p fully complements the lack of S. cerevisiae own Na+-ATPases but the activity of the CgCnh1 antiporter is lower than that of ScNha1p. Candida glabrata deletion mutants and analyses of their phenotypes revealed that though both transporters have a broad substrate specificity, their function in C. glabrata cells is not the same. Their differing physiological roles are also reflected in their regulation of expression, CgENA1 is highly upregulated by an increased osmotic pressure or sodium concentration, whereas CgCNH1 is expressed constitutively. The Cnh1 antiporter is involved in the regulation of potassium content and the Ena1 ATPase in sodium detoxification of C. glabrata cells. This situation differs from S. cerevisiae, where the Nha1 antiporter and Ena ATPases both participate together in Na+ detoxification and in the regulation of K+ homeostasis.

  • Candida glabrata
  • salt tolerance
  • potassium homeostasis
  • sodium transporter
  • potassium transporter


The maintenance of intracellular ion homeostasis is crucial for any organism, including yeast cells. It means to provide the cytosol with a high amount of potassium, which is required for many physiological functions, and maintain low intracellular level of toxic sodium. To maintain an optimal cytoplasmic concentration of potassium and a stable high intracellular K+/Na+ ratio, yeast cells possess a series of transporters with varying transport mechanisms that provide cells with the required amount of potassium or eliminate the surplus of alkali metal cations from the cytoplasm, either by exporting them from the cell or sequestrating them in organelles.

Candida glabrata is a human fungal pathogen and is, in spite of its name, phylogenetically more related to Saccharomyces cerevisiae than to the group of other Candida species (Dujon, 2004; Kaur, 2005; Fitzpatrick, 2006; Li, 2007). Though found mainly in humans and animals, it has also been isolated from hypersaline water (Butinar, 2005). In general, pathogen Candida species, including C. glabrata, are more tolerant to alkali metal cations and grow better in the presence of high external concentrations of salts than S. cerevisiae (Krauke & Sychrova, 2008, 2010). However, the molecular basis of Candida's high salt tolerance, including the involved transport systems, and its relation to virulence/pathogenicity, have not been completely elucidated.

Two plasma-membrane transport systems are involved in the halotolerance of the most-studied yeast S. cerevisiae: P-type Ena Na+-ATPases mainly export alkali metal cations when the cells are in media with neutral or higher pH levels, and a Na+/H+-antiporter (Nha1p) takes part in the alkali-metal-cation efflux at lower external pH levels (Banuelos, 1998; Platara, 2006).

The ScNha1 antiporter is a member of the large and ubiquitous family of Na+/H+ exchangers (Brett, 2005a), and it has a broad substrate specificity for various alkali metal cations (K+, Na+, Li+, Rb+). Because of its ability to export potassium, it is involved in many other cellular processes besides detoxification of sodium, such as regulation of the cell cycle (Simon, 2001), the response to osmotic shock (Kinclova-Zimmermannova & Sychrova, 2006), or regulation of plasma-membrane potential (Kinclova-Zimmermannova, 2006) and intracellular pH (Sychrova, 1999; Brett, 2005b).

In most other yeast species, including Candida albicans, Candida dubliniensis, Candida parapsilosis, Candida tropicalis and the closely related nonpathogenic Debaryomyces hansenii, similar antiporters capable of exporting both toxic cations (Na+, Li+) and surplus K+ exist, however their physiological role is not known in detail as their properties have been characterized mainly by heterologous expression in transporter-less mutants of S. cerevisiae (Kinclova, 2001a; Kamauchi, 2002; Velkova & Sychrova, 2006; Krauke & Sychrova, 2008). The only exception so far is Schizosaccharomyces pombe, Zygosaccharomyces rouxii and C. albicans for which the role and properties of alkali-metal-cation/proton antiporters have been studied both upon heterologous expression in S. cerevisiae and in the original yeast species. Schizosaccharomyces pombe and Z. rouxii differ (together with Yarrowia lipolytica) from other yeast species as they possess two types of antiporters in their plasma membranes, one involved in sodium and lithium detoxification, and the other one exporting mainly potassium and thus participating in the regulation of homeostasis, internal pH and cell volume (Jia, 1992; Papouskova & Sychrova, 2007; Pribylova, 2008). The C. albicans Cnh1 antiporter expressed in S. cerevisiae has many features (e.g. substrate specificity, transport capacity) that are highly similar to S. cerevisiae's own Nha1p (Kinclova, 2001a) but its physiological role in C. albicans cells is different. Its activity does not contribute directly to C. albicans tolerance to sodium cations, as its deletion results only in sensitivity to high potassium and rubidium concentrations and a diminished potassium efflux, suggesting that Cnh1p plays a role in potassium homeostasis and not in sodium detoxification (Soong, 2000; Kinclova-Zimmermannova & Sychrova, 2007).

As the yeast Na+/H+ antiporter's function is dependent on the proton gradient created by the H+-ATPase across the plasma membrane, and thus is fully active at lower external pH, yeast cells usually possess a second Na+ extrusion system, the Ena ATPase. Saccharomyces cerevisiae harbours, depending on the strain, up to five ENA genes that are organized in a tandem array (Wieland, 1995). Expression of ScENA1 can be highly upregulated via several signalling pathways, mainly upon exposure to different stresses (Ruiz & Arino, 2007). This is in contrast to S. cerevisiae NHA1, which is constitutively expressed at low transcription levels and independently of exposure to stress (Banuelos, 1998). Nevertheless, in S. cerevisiae, the Nha1 antiporter and Ena1 ATPase mediate the efflux of at least four alkali metal cations, and participate together in Na+ detoxification and the regulation of K+ homeostasis.

A search in the yeast genome sequences of other yeast species revealed that most of them contain no more than two copies of ENA genes (Benito, 2002; Gorjan & Plemenitas, 2006). As far as pathogenic Candida species are concerned, the role of Ena ATPases has been studied in detail only in C. albicans and C. dubliniensis (Enjalbert, 2009). The expression of ENA genes upon osmotic stress differs significantly in both species, and the expression of C. albicans Ena21 ATPase in C. dubliniensis is able to substantially increase its NaCl tolerance (Enjalbert, 2009). Nothing is known about the Ena ATPase's ability to export potassium from Candida cells.

Also, the activity of the C. albicans Na+/H+-antiporter (encoded by the CNH1 gene) is much higher than the activity of C. dubliniensis Cnh1p, if both antiporters are expressed under the same conditions in S. cerevisiae (Krauke & Sychrova, 2008). As C. albicans has more powerful plasma-membrane systems to extrude surplus sodium from cells than C. dubliniensis, and as these two closely related species differ significantly in their halotolerance, C. albicans being able to grow in much higher salt concentrations than C. dubliniensis, it seems evident that the alkali-metal-cation efflux systems play a crucial role in the salt tolerance of Candida species.

In this study, we report the first characterization of two C. glabrata transport systems involved in the efflux of alkali metal cations. The activity and substrate specificity of both transporters, the Cnh1 antiporter and the Ena1 ATPase, were characterized upon heterologous expression in a salt-sensitive S. cerevisiae strain, and in C. glabrata cells. Taken together, the results lead to the proposition that the CgCnh1 antiporter plays an important housekeeping role in potassium homeostasis but not in sodium export, whereas CgEna1 ATPase is mainly involved in sodium and lithium detoxification.

Materials and methods

Strains and media

The S. cerevisiae and C. glabrata strains used in this study are listed in Table 1. All strains were routinely grown in rich YPD (1% yeast extract, 2% peptone, 2% glucose) or minimal YNB [0.17% YNB w/o amino acids and ammonium sulphate, 0.5% (NH4)2SO4, 2% glucose and auxotrophic supplements if necessary] media at 30 °C.

View this table:

Saccharomyces cerevisiae and Candida glabrata strains used in this study

S. cerevisiae
W303-1AMATa leu2-3/112 ura3-1 trp1-1 his3-11/15 ade2-1 can1-100 GAL SUC2 mal10Wallis (1989)
CW25W303-1A nha1LEU2Kinclova-Zimmermannova (2006)
BW31aW303-1A ena1-4HIS3 nha1LEU2Kinclova-Zimmermannova (2006)
C. glabrata
ATCC2001Wild typeATCC collection
cnh1ΔATTC2001 cnh1Δ∷FRTThis study
ena1ΔATTC2001 ena1Δ∷FRTThis study
CNH1reATTC2001 cnh1Δ∷FRTCNH1-SAT1This study
ENA1reATTC2001 ena1Δ∷FRTENA1-SAT1This study


The plasmids for the expression of C. glabrata transporters in S. cerevisiae (listed in Table 2) were constructed as described previously (Krauke & Sychrova, 2008). Briefly, the coding sequences of CgCNH1 and CgENA1 were cloned into the two multicopy plasmids, YEp352 and pGRU1, behind the ScNHA1 promoter by homologous recombination in S. cerevisiae. In pGRU1 the green fluorescent protein (GFP) sequence was attached to the 3′ end of the coding sequence.

View this table:

Plasmids constructed in this study

PlasmidsGene of interest
S. cerevisiaeExpression
YEp352-SAT1Complete SAT1 deletion cassette
C. glabrataDeletion or reintegration
  • UR, DR – 400 bp upstream and downstream noncoding sequences, respectively.

For the construction of CgCNH1 and CgENA1 deletion cassettes, first the whole SAT1 cassette (Reuss, 2004) was cloned by homologous recombination into S. cerevisiae YEp352. Then the upstream and downstream flanking regions of the CgCNH1 and CgENA1 genes (each about 400 bp long) were amplified using the appropriate primers (Table 3) and cloned by homologous recombination upstream and downstream of the SAT1 cassette resulting in the Y-CgCNH1Δ and Y-CgENA1Δ plasmids, respectively. For reintegrating CgCNH1 and CgENA1 into the C. glabrata genome, CgCNH1 and CgENA1 genes were cloned by homologous recombination in S. cerevisiae between the corresponding upstream flanking region and SAT1 cassette in Y-CgCNH1Δ and Y-CgENA1Δ, respectively, resulting in the Y-CgCNH1Re and Y-CgENA1Re plasmids (Tables 2 and 3).

View this table:

Oligonucleotides used in this study

OligonucleotidesSequences (5′–3′)
Gene amplification
Gene deletion
Gene reintegration
  • * Sequences in italics, region homologues to amplified DNA.

Gene deletion/reintegration in C. glabrata

The deletion/reintegration cassettes were cut out of the plasmids using unique restriction sites and transformed by electroporation into C. glabrata ATCC 2001. Clones were selected by growth on YPD+200 μg mL−1 nourseothricin (Werner Bioagents, Jena, Germany). In deletion mutants, the integrated SAT1 cassette was eliminated from the locus by growth in YPM (1% yeast extract, 2% peptone, 2% maltose) and subsequent selection of nourseothricin-sensitive clones on YPD plates with 25 μg mL−1 nourseothricin. Deletion mutants were further transformed with the reintegration cassettes to place the genes back at their loci. The integration and reintegration of cassettes were verified by PCR and Southern blot.

Salt tolerance assay

To determine the salt tolerance of S. cerevisiae or C. glabrata strains, 3 μL of 10-fold serial dilutions of yeast cell suspensions were spotted on YNB or YPD plates containing alkali metal cations at various concentrations. Plates with their pH adjusted to 3.5, 5.5 and 7.0, respectively, were prepared as described previously (Kinclova, 2001b). Plates were incubated at 30 °C and cell growth recorded for 7 days.

Cation efflux assay

The efflux of K+ and Na+ from yeast cells was measured as described previously (Papouskova & Sychrova, 2006). Saccharomyces cerevisiae and C. glabrata cells were grown in YNB-NH4 to OD600 nm=0.17–0.25, harvested and washed. For Na+ efflux experiments, cells were preloaded with Na+ in YNB supplemented with NaCl as indicated. Freshly harvested (for potassium efflux measurements) or preloaded (for sodium efflux measurements) cells were transferred to either pH 5.5 [20 mM MES, pH adjusted to pH 5.5 with Ca(OH)2, 0.1 mM MgCl2 and 2% glucose] or pH 4.5 [10 mM Tris, pH adjusted to 4.4 with citric acid and then brought up to 4.5 by Ca(OH)2, 0.1 mM MgCl2 and 2% glucose] incubation buffer, as indicated in the text. For efflux measurements in S. cerevisiae, 10 mM RbCl or KCl were added to the incubation buffers to prevent reuptake of lost K+ or Na+, respectively, and for C. glabrata experiments the concentration of external cations was increased to 50 mM. Cell samples were withdrawn at distinct time points for 60 min, collected on Millipore membrane filters, washed, acid extracted and the intracellular concentration of K+ and Na+ estimated by atomic absorption spectroscopy (Kinclova, 2001b). All experiments were repeated two to four times; mean values are given and representative experiments shown.

Gene expression analysis by quantitative reverse transcriptase (qRT)-PCR

For qRT-PCR experiments, C. glabrata cells were grown to OD600 nm≈1 in YPD. Immediately, 20 mL of cell culture was pelleted and frozen (for time point t=0) and the remainder of the cells were collected, resuspended in YPD with 1000 mM NaCl, 1000 mM KCl or 1000 mM sorbitol and incubated at 30 °C. After 30, 60 and 90 min of incubation, 20-mL aliquots were taken, centrifuged, cell pellets frozen with liquid nitrogen and stored at −80 °C.

RNA isolation from cell pellets was performed using the Trizol agent (Invitrogen) following the manufacture's instructions with slight modification. Ten to one hundred times more cells than suggested were broken. Candida glabrata cells were broken by vortexing with glass beads 8 × 30 s. The quality of obtained RNA was verified by gel electrophoresis and concentration measured with Nanodrop apparatus. The remaining DNA was removed from samples by DNase treatment and 1 μg of RNA was used for cDNA synthesis by reverse transcription using the First Strand cDNA Synthesis Kit (Fermentas).

Real-time PCR was performed using 96-well plates, a 480 Lightcycler instrument (Roche) and SYBR Green I Master Stain (Roche). In each well, a 20 μL reaction mix was added containing 10 μL of SYBR Green I Master Mix (Roche), 2 μL of the specific primer pair (1 μmol final concentration) (Table 3), 2 μL of cDNA (25 ng in well) and 6 μL of water. Each sample was prepared in triplicate. The PCR program was as follows: 10 min 95 °C initial denaturation, 45 cycles of 10 s 95 °C, 30 s 60 °C and 30 s 72 °C. To verify the specificity of the generated products a melting curve was measured by heating the products. The Ct values were calculated using the lightcycler 480 software version 1.5. Transcript levels were normalized against ACT1 expression, and gene expression changes (relative to expression levels at t=0) calculated using the ΔΔCt-method. Each experiment was repeated at least three times, each time with a triplicate of analysed samples, a representative result is shown.


Our first insight into the salt tolerance of four pathogenic Candida species (Krauke & Sychrova, 2008) has shown that all of them are in general more tolerant to high concentrations of alkali metal cations than the model yeast S. cerevisiae. However, C. glabrata seemed to be not as salt tolerant as C. albicans and C. parapsilosis, mainly its tolerance of toxic lithium cations was not very high (Krauke & Sychrova, 2008). We decided to clone C. glabrata genes encoding putative Na+-ATPases and Na+/H+-antiporters, i.e. systems that are involved in Li+ detoxification in other yeast species, and characterize the properties of their products in order to elucidate whether the transport capacity of these two systems reflects the observed level of C. glabrata salt tolerance and their role in C. glabrata.

Candida glabrata Na+/H+-antiporter

Heterologous expression of CgCnh1p in S. cerevisiae

There is one ORF in the C. glabrata genome encoding a putative plasma-membrane Na+/H+-antiporter, and the sequence of its product shares a much higher level of similarity with ScNha1p than with the antiporters from other Candida species (Krauke & Sychrova, 2008). To characterize the CgCnh1 antiporter's substrate specificity and activity, the coding sequence was expressed in a salt-sensitive S. cerevisiae BW31a strain lacking its own alkali-metal-cation exporters (ena1-4Δnha1Δ). To test if the CgCnh1 antiporter is functional, which cations are its substrate and how active it is in S. cerevisiae cells, the salt tolerance of BW31a cells expressing CgCnh1p was compared with the tolerance of BW31a cells either expressing ScNha1p and CaCnh1p (under the same conditions, i.e. from the same multicopy vector and ScNHA1 promoter) or lacking any efflux systems.

The overexpression of CgCnh1p was not toxic for S. cerevisiae cells (Fig. 1, left panel), and the antiporter was functional. Its presence significantly increased the ability of cells to grow in the presence of a high external concentration of three alkali-metal-cation salts, suggesting a broad substrate specificity (Fig. 1). NaCl was tolerated at lower concentrations than KCl due to the toxicity of sodium cations. However, an increased tolerance of CgCnh1-containing cells to the smallest cation tested, Li+, was not observed, which could mean that Li+ is not one of the preferred substrates of CgCnh1. Comparison of the salt-tolerance levels of cells expressing the three antiporters revealed that the presence of ScNha1 or CaCnh1 antiporters brings about a significantly higher tolerance to NaCl and slightly higher tolerance to RbCl compared with the expression of CgCnh1p. On the other hand, the three antiporters increased the tolerance to KCl similarly (Fig. 1). The expression of CgCNH1 with the GFP sequence attached to its 3′ (from pGRU1 and the ScNHA1 promoter) also resulted in a similarly increased salt tolerance of BW31a cells, which suggested that, as in the case of other yeast Na+/H+-antiporters, the tagging of the C-terminus does not influence the CgCnh1 antiporter's activity. The CgCnh1–GFP fusion protein was exclusively localized to the plasma membrane of BW31a cells; again similarly to other yeast Na+/H+-antiporters (not shown).

Figure 1

Growth of Saccharomyces cerevisiae BW31a (ena1-4Δnha1Δ) cells expressing CgCNH1 in YNB supplemented with salts of various alkali metal cations. Cells containing the empty YEp352 vector (–) and expressing ScNHA1 or CaCNH1 genes were used as controls.

To test if the activity of the CgCnh1 antiporter is proton-gradient dependent, BW31a cells expressing the three antiporters were grown on YNB plates with adjusted/buffered pH and different concentrations of various salts (Table 4). The sodium and potassium tolerance of BW31a cells expressing CgCnh1p (as well as the other two antiporters) was the highest at acidic external pH, confirming the inward gradient of protons as the driving force to pump surplus alkali metal cations out of cells. Nevertheless, even at external pH 7.0, cells expressing all three antiporters tolerated higher salt concentrations than cells without any antiporter, which suggests that even when the proton gradient across the plasma membrane is small, the antiporters are partially functional (Table 4). The observed tolerances confirmed (at all tested pH values) the results shown in Fig. 1, i.e. the lower capacity of CgCnh1p for Na+, if compared with ScNha1 or CaCnh1 antiporters, and its inability to significantly diminish cell sensitivity to Li+.

View this table:

Maximum salt concentrations allowing the growth of BW31a cells expressing yeast Na+/H+ antiporters in YNB-NH4 media with different pHout levels

NaCl (mM)KCl (mM)LiCl (mM)

To verify if the observed increased salt tolerance of BW31a cells expressing CgCnh1p was due to the efflux of cations via this antiporter, sodium and potassium efflux was measured. As the internal sodium concentration of S. cerevisiae cells grown in YNB is negligibly low, cells were preloaded with sodium (100 mM NaCl, 60 min) under conditions where the antiporter's activity is rather low (i.e. at pH 7.0). After preloading, cells were transferred to a low pH buffer (without NaCl but with KCl to prevent sodium reuptake), aliquots of cells were withdrawn over 60 min, and the intracellular concentration of sodium in samples estimated. Cells expressing CgCnh1p exported significantly more Na+ than cells with the empty plasmid but less than cells expressing ScNha1p (Fig. 2). Within 1 h, cells expressing CgCnh1p exported 40.5 ± 3.8%, cells with ScNha1p 80.6 ± 1.9% and cells with the empty plasmid only 8.6 ± 1.8% of their initial Na+. The initial concentration of Na+ after preloading was 78.3 nmol mg−1 dry wt for cells with ScNha1p, i.e. about 20% lower than for the cells expressing CgCnh1p (105.3 ± 18.6 nmol mg−1 dry wt) or cells with the empty plasmid (100.8 ± 3.0 nmol mg−1 dry wt). The initial intracellular sodium concentrations after preloading suggests that at pH 7.0, the ScNha1 antiporter is more active than CgCnh1p though this difference was not observed in the drop tests summarized in Table 2.

Figure 2

Efflux of intracellular Na+ from Saccharomyces cerevisiae BW31a cells containing the empty vector ( Embedded Image), expressing CgCNH1 (▲) or ScNHA1 ( Embedded Image), respectively.

Preloading is not necessary for potassium efflux measurements, as the internal potassium concentration in cells grown in standard YNB is high enough to measure K+ efflux. The initial K+ concentration in cells with ScNha1p (410.9 ± 18.1 nmol mg−1 dry wt) was significantly lower than in cells with CgCnh1p (473.2 ± 25.2 nmol mg−1 dry wt) or in cells with the empty plasmid (478.4 ± 45.8 nmol mg−1 dry wt). Saccharomyces cerevisiae's own Nha1 seemed to be more active during the exponential phase of growth than the heterologous CgCnh1 and this was confirmed by efflux measurements. Within 60 min in K+-free buffer (supplemented with RbCl to prevent K+ reuptake), cells expressing CgCnh1p lost significantly less K+ (11.2 ± 1.6% of the initial intracellular concentration), than cells with ScNha1p (42.1 ± 4.1%), but this was still three times more than cells without an antiporter (3.7 ± 0.5%). The observed low level of potassium efflux via CgCnh1p is surprising, considering the antiporter's ability to confer as high a potassium tolerance as the ScNha1p (Fig. 1).

Role of Cnh1p in C. glabrata

Initial results from the functional characterization of CgCnh1p in S. cerevisiae showed that the protein transports sodium and potassium and thus may play a role in both the salt tolerance and potassium homeostasis of C. glabrata cells. To confirm this, C. glabrata cnh1Δ deletion mutants were constructed using the SAT1 deletion cassette and subsequent excision of the cassette, resulting in the replacement of the whole CNH1-coding sequence with the short FRT sequence (Reuss, 2004). Two independent cnh1Δ strains were further characterized; in all experiments they behaved the same. In the first phenotypic characterization, wild-type and mutant cells were grown on a series of YPD plates containing alkali metal cations in increasing concentrations. Surprisingly, the growth of cnh1Δ mutants was only inhibited in the presence of high KCl (>2200 mM) and RbCl (>2000 mM) concentrations (Fig. 3a). Deletion of CNH1 in C. glabrata cells had no effect on tolerance to NaCl and LiCl (Fig. 3a), in spite of the fact that CgCnh1p was an efficient Na+ exporter when expressed in S. cerevisiae (Figs 1 and 2, Table 4). These results suggested that the role of CgCnh1p in C. glabrata might be different from the role of S. cerevisiae Nha1p which is important for cell survival in the presence of both high sodium and potassium (Banuelos, 1998). To confirm the observed mutant phenotypes, (1) the CNH1 gene was reintegrated to its locus (Fig. 3a), and (2) Na+ and K+ efflux from the C. glabrata wild type and cnh1Δ mutant was measured. Cells from the exponential phase of growth in YNB medium were either directly transferred to the incubation buffer (pH 4.5, supplemented with RbCl) for K+ efflux experiments or first preloaded in YNB with 400 mM NaCl for 30 min before transfer to the incubation buffer (pH 5.5, supplemented with KCl). The deletion of CNH1 strongly affected the ability of C. glabrata cells to export potassium. Whereas wild-type cells lost approximately 62% of their intracellular potassium within 60 min in potassium-free incubation buffer, cnh1Δ mutants exported about 9% of their internal potassium content (Fig. 3b). The experiment was repeated three times, and in each of the experiments, three parallel samples were analysed. It is worth noting that in all nine samples analysed for each strain, the initial intracellular K+ concentration was always higher in the cnh1Δ mutants. The average K+ concentration in exponentially growing cells was 545.0 ± 29.2 nmol mg−1 dry wt for the wild type, and 589.7 ± 56.5 nmol mg−1 dry wt for cnh1Δ. Preloading wild-type and mutant cells with 400 mM NaCl resulted in similar intracellular sodium concentrations, 131.1 ± 29.2 nmol mg−1 dry wt for the wild type and 126.9 ± 33.2 nmol mg−1 dry wt for the cnh1Δ mutant, suggesting that both strains do not significantly differ in their sodium-exporting capacity during preloading. Also the amount of sodium exported from preloaded cells during 60 min of incubation in Na+-free buffer was almost the same. The wild-type cells lost 91.4 ± 1.8% and mutant cells 86.0 ± 1.9% of their sodium content (Fig. 3b).

Figure 3

Growth of Candida glabrata wild-type (wt) (ATCC 2001), cnh1Δ and cnh1Δ∷CNH1 strains in YPD supplemented with salts of various alkali metal cations (a); and the amount of intracellular K+ and Na+ cations lost from wt and cnh1Δ cells in 60 min (b).

The reduced tolerance of cells lacking the CNH1 gene to high potassium (and rubidium) concentrations, together with the observed strongly diminished efflux capacity of and higher intracellular potassium content in cnh1Δ cells, suggest that the CgCnh1 antiporter is the major potassium export system in C. glabrata cells growing in standard conditions, and that its physiological role might be connected to the maintenance of cell homeostasis and membrane potential. On the other hand, the same tolerance to sodium and a similar loss of sodium from cnh1Δ and wild-type cells indicate that the Cnh1 antiporter does not play an important role in the Na+ detoxification of C. glabrata cells growing under standard conditions. This result is in contrast to results with S. cerevisiae nha1Δ cells, where both the tolerance to Na+ and K+ is diminished (Banuelos, 1998) but similar to C. albicans in which the cnh1/cnh1 deletion increases the cell sensitivity to high potassium and has no effect on sodium tolerance and transport (Kinclova-Zimmermannova & Sychrova, 2007).

Candida glabrata Na+-ATPase

In silico analysis

Our results described above revealed the existence of another, highly active transport system for sodium extrusion in C. glabrata cells. Saccharomyces cerevisiae harbours, beside the Nha1 antiporter, a second alkali-metal-cation exporting system, the Ena1 Na+-ATPase (Haro, 1991). A search in the genome sequence of C. glabrata (http://cbi.labri.fr/Genolevures/) found three ORFs that had a high similarity score to the ScEna1p sequence, CAGL0K12034g, CAGL0J01870g and CAGL0I04312g. All three C. glabrata protein sequences were compared in more detail with the sequences of already characterized Ena ATPases (from S. cerevisiae, C. albicans, C. dubliniensis), putative Ena ATPases (C. parapsilosis) and S. cerevisiae Pmr1 ATPase, which transports Ca2+ in the membranes of the Golgi apparatus (Antebi & Fink, 1992). A sequence comparison (Table 5) revealed that the C. glabrata CAGL0K12034g sequence has a high degree of identity to all proved or potential Candida Ena ATPases, on average approximately 60%. Nevertheless, CAGL0K12034g shares the highest identity with all three S. cerevisiae Ena proteins (almost 72%) which probably reflects the closer relationship of C. glabrata to S. cerevisiae than to other Candida species (Fitzpatrick, 2006). The analysis also suggested that the other two C. glabrata ORFs encoding putative cation ATPases are highly identical with ScPmr1p, indicating that CAGL0I04312g and CAGL0J01870g might both encode a Ca2+ ATPase (Table 5).

View this table:

Sequence comparison of three putative Candida glabrata Ena ATPases with Ena and Pmr1 homologues

% of identity
C. glabrata CAGL0K12034pC. glabrata CAGL0J01870pC. glabrata CAGL0I04312p
C. albicans Ena2p58.529.129.9
C. albicans Ena21p59.329.529.5
C. dubliniensis 070111p58.729.430.5
C. dubliniensis 070112p61.430.630.6
C. parapsilosis 005504p63.531.031.7
C. parapsilosis 005569p60.929.730.4
S. cerevisiae Ena1p71.729.529.5
S. cerevisiae Ena2p71.729.729.5
S. cerevisiae Ena5p71.729.729.5
S. cerevisiae Pmr1p30.564.480.0

The sequence alignment of Ena ATPases revealed their high degree of identity in known motifs as well. A typical yeast Ena ATPase possesses five regions important for their function: a phosphatase motif (DEXLLTGESL), a phosphorylation motif (DKTGTLT) and three ATP-binding motifs (DPPR), (MLTGD) and (GDGVNDSPSLK) (Gorjan & Plemenitas, 2006). With the exception of the phosphatase motif, all other motifs were 100% identical in our alignment. Significant sequence differences among the Ena ATPases only occurred in the first approximately 50–60 amino acids of the N-terminal region, and we have identified three regions (amino-acid residues in CgEna1p 436–446, 498–523, 617–626) existing only in S. cerevisiae and C. glabrata Ena ATPases (though not with highly conserved sequence) and absent in the Ena ATPases of other Candida species. According to the results of our sequence comparison, we named C. glabrata CAGL0K12034g CgEna1 and characterized its properties using the same two approaches as described above for CgCnh1p.

Heterologous expression of CgEna1p in S. cerevisiae

For characterizing the substrate specificity and transport capacity of the putative Na+-ATPase, the CgENA1 gene was expressed in the S. cerevisiae BW31a strain (ena1-4Δnha1Δ) similarly as the CgCNH1. The CgEna1 tagged C-terminally with the GFP sequence was exclusively localized to the plasma membrane. The phenotypes of BW31a cells expressing CgEna1p were compared with those of the S. cerevisiae W303-1A wild type (which harbours four ENA genes and NHA1), CW25 (ENA1-4 nha1Δ) and BW31a (ena1-4Δnha1Δ). A growth comparison of S. cerevisiae cells on YNB plates with increasing salt concentrations showed that CgEna1p restores the tolerance of BW31a to all four tested alkali metal cations (Fig. 4). Cells expressing CgEna1p grew on various salts similarly to the CW25 strain, indicating similar broad substrate specificity for CgEna1p and ScEna1p and indicating the role of both Na+-ATPases in S. cerevisiae potassium homeostasis (Fig. 4, second panel).

Figure 4

Growth of Saccharomyces cerevisiae BW31a (ena1-4Δnha1Δ) cells expressing CgENA1 in YNB supplemented with salts of various alkali metal cations (a); and the intracellular cation concentration in S. cerevisiae BW31a cells containing the empty vector (control) or expressing CgENA1 at time 0 (black bars) and after 60 min of incubation (grey bars) (b).

To confirm the broad substrate specificity and estimate the transport activity of CgEna1p in S. cerevisiae, potassium and sodium efflux was measured. As has been already mentioned, preloading with sodium is required to measure the Na+ efflux from S. cerevisiae cells. Our first attempts to preload cells in a standard manner (i.e. incubate them in YNB with 100 mM NaCl at pH 7.0 for 60 min) resulted in a very low Na+ concentrations (<30 nmol mg−1 dry wt) in cells expressing CgEna1p. This indicated a high activity of the ATPase during the preloading. The NaCl concentration for successful preloading was increased to 300 mM NaCl, but even then, the initial sodium concentration of cells expressing CgEna1p was only 55.2 ± 6.8 nmol mg−1 dry wt, i.e. more than three times lower than cells with the empty plasmid (183.3 ± 17.8 nmol mg−1dry wt). When the preloaded cells were incubated in Na+-free buffer for 60 min, cells with CgEna1p lost 46 ± 5.5% of their initial sodium concentration, and cells with the empty plasmid only 8.2 ± 1.2% (Fig. 4b). CgEna1p clearly transports Na+ when expressed in S. cerevisiae.

The expression of CgEna1p did not diminish the cell potassium content of growing cells and similarly neither did the expression of CgCnh1p under the same conditions (cf. above). The intracellular K+ concentrations in cells from the exponential phase of growth in YNB were 492.9 ± 30.2 nmol mg−1 dry wt for cells expressing CgEna1p, and 477.0 ± 30.0 nmol mg−1 dry wt in control cells lacking export systems. To estimate CgEna1p's K+ efflux capacity, changes in intracellular potassium concentration in cells were followed in K+-free buffer for 60 min. Cells expressing CgEna1p exported significantly more potassium than control cells. During 1 h they lost 17.5 ± 1.2% of the initial internal amount, i.e. approximately five times more than control cells (3.6 ± 1.5%; Fig. 4b). The drop tests and cation efflux measurements showed that CgEna1p has broad substrate specificity for alkali metal cations and that it is able to transport both sodium and potassium when expressed in S. cerevisiae.

Role of Ena1p in C. glabrata

The functional characterization of CgEna1p in S. cerevisiae cells provided some insights into the protein function in sodium and potassium efflux in C. glabrata. To confirm this theory, a C. glabrata ena1Δ strain was constructed using a similar protocol to that described above for the C. glabrata cnh1Δ mutant strain.

To characterize the phenotype of ENA1 deletion in C. glabrata cells, the growth of the ena1Δ strain was tested on YPD plates supplemented with alkali metal cations. As shown in Fig. 5a, deletion of CgENA1 had no effect on the tolerance to potassium and rubidium; mutant cells grew as well as the wild type at very high concentrations of KCl and RbCl. In contrast, C. glabrata ena1Δ cells were unable to grow on media containing higher concentrations of sodium and lithium cations. These initial results suggested that the Na+-ATPase is involved in C. glabrata tolerance to high concentrations of Na+ and Li+ but probably not of K+ or Rb+. The reintegration of ENA1 into its locus restored the cell tolerance to sodium and lithium salts (Fig. 5a).

Figure 5

Growth of Candida glabrata wild-type (wt) (ATCC 2001), ena1Δ and ena1Δ∷ENA1 strains in YPD supplemented with salts of various alkali metal cations (a); and the amount of intracellular K+ and Na+ cations lost from wt and ena1Δ cells in 60 min (b).

To confirm the role of Ena1p, sodium and potassium loss from cells was measured (Fig. 5b). For testing their ability to export sodium, cells were preloaded with 400 mM NaCl in YNB for 30 min and then transferred to the incubation buffer. The initial concentration of sodium after preloading was much higher in the ena1Δ mutant (172.3 ± 34.7 nmol mg−1 dry wt) than in wild-type cells (131. ± 29.2 nmol mg−1 dry wt) in all experiments. Within 1 h of incubation in Na+-free buffer wild-type cells exported 91.4 ± 1.8% of the initial sodium content, and ena1Δ cells 67.5 ± 6.3%. The deletion of ENA1 in C. glabrata clearly affected its ability to export sodium, even though nearly 70% of the initial sodium content was still exported from the ena1Δ strain. The deletion of ENA1 had no effect on potassium efflux; within 1 h, wild-type cells lost 62.0 ± 4.6% of their internal potassium content, and ena1Δ cells 68.5 ± 1.3%. The cation efflux experiments in C. glabrata cells showed that the deletion of ENA1 only affects the sodium export and the rate of potassium efflux remains unchanged.

Expression of CNH1 and ENA1 genes in C. glabrata

In S. cerevisiae, NHA1 is expressed constitutively at a low level, even under salt stress, whereas ENA1 expression is highly upregulated upon various stresses (Prior, 1996; Ruiz & Arino, 2007). Also in C. albicans, the salt stress increases the expression of ENA21-encoding Na+-ATPase (Enjalbert, 2009) and has no effect on the expression of CNH1 (Soong, 2000). To elucidate whether the expression of CgCNH1 and CgENA1 genes is also regulated in response to stress, the amount of both transcripts was estimated (using quantitative real-time PCR) after the exposure of C. glabrata cells to salt stress. First, relative transcription levels (normalized to ACT1 expression and to the time point t=0) were estimated in samples withdrawn at three time points (30, 60 and 90 min) after cell transfer to media with 1000 mM NaCl. As shown in Fig. 6a, the expression of CgENA1 increased maximally (approximately 25-fold) during the first 60 min of exposure to 1000 mM NaCl and then partly dropped (12-fold higher expression after 90 min, compared with the level at t=0). To distinguish if the induction of CgENA1 expression was sodium specific or if it was a general response to hyperosmotic stress, the changes in CgENA1 transcription levels in the presence of nontoxic KCl or a nonionic solute (sorbitol) were followed. The presence of high KCl or sorbitol also led to the induction of CgENA1, however the amount of detected transcripts at all three time points was significantly lower compared with that observed upon exposure to NaCl (Fig. 6a). The expression of CgCNH1 hardly changed at all under the same conditions (Fig. 6b). Neither exposure to NaCl nor to KCl influenced the level of CgCNH1 expression significantly, only the sorbitol stress (longer than 60 min) resulted to a slight decrease in CgCNH1 transcription (Fig. 6b). Compared with the expression of CgENA1, the expression of CgCNH1 is low and unchanged during salt stress.

Figure 6

Relative normalized expression of CgENA1 (a) and CgCNH1 (b) genes in the ATCC 2001 wild type upon osmotic stress during 90 min. Transcript levels were normalized to ACT1 level of expression and to time point t=0. Black bars, expression in the presence of 1000 mM NaCl; dark grey bars, expression in the presence of 1000 mM KCl; light grey bars, expression in the presence of 1000 mM sorbitol.


Genes encoding putative plasma-membrane Na+/H+-antiporters and Na+-ATPases have been found in all yeast genomes, including pathogenic Candida species. Their products are thought to play an important role in the survival of salt stress, regulation of cell volume, intracellular pH and cation homeostasis (Arino, 2010), however, experimental proof is missing for most of these species. The most intensively studied are by far the alkali-metal-cation efflux systems of S. cerevisiae (Arino, 2010) and their collaboration in the maintenance of sodium and potassium homeostasis in this model yeast has been taken to be a general rule for all yeast species. This assumption was mainly due to the fact that most of the other yeast plasma-membrane Na+/H+ antiporters have only been characterized upon heterologous expression in S. cerevisiae and there they have shown the ability to transport both sodium and potassium. Only in S. pombe, Z. rouxii and Y. lipolytica, are there two plasma-membrane antiporters with different physiological roles, one specialized for potassium efflux and the other with a higher affinity for sodium cations, as shown by their heterologous expression in S. cerevisiae and demonstrated for Z. rouxii and S. pombe by characterizing of strains lacking or overexpressing the two corresponding genes (Arino, 2010). As far as Ena ATPases are concerned, only those from Z. rouxii and D. hansenii have been studied in more detail (genes expressed in S. cerevisiae and Z. rouxii deletion mutants prepared) and their role in sodium detoxification was confirmed. Unfortunately, their ability to mediate the efflux of potassium has never been tested.

In pathogenic C. albicans, external alkali metal cations are thought to affect several virulence factors (Hermann, 2003) and intracellular potassium content probably plays a role in the morphology switch from budding to hyphae form (Watanabe, 2006). Our recent work compared the salt tolerance of four Candida species (Krauke & Sychrova, 2008, 2010) and characterized the Nha1p homologues from the most salt-tolerant C. parapsilosis and the least salt-tolerant C. dubliniensis (Krauke & Sychrova, 2008). Upon heterologous expression in S. cerevisiae, the CpCnh1p protein had higher transport activities for Na+ and K+ than the CdCnh1p. Also the second sodium-extruding system, the Ena ATPase, was shown to not be very active in C. dubliniensis (compared with the salt-tolerant C. albicans) (Enjalbert, 2009). These results lead to a proposal that the low level of salt tolerance observed for C. dubliniensis reflects the low activities of its two transport systems for sodium extrusion. In this work, we have characterized the two transporters in C. glabrata, another pathogenic yeast, which is haploid, phylogenetically related to S. cerevisiae, and whose salt tolerance lies somewhere in between those of the tolerant C. albicans and C. parapsilosis on one hand, and the sensitive C. dubliniensis on the other (Krauke & Sychrova, 2008, 2010). For the first time, the mutants of C. glabrata lacking one or the other of the two putative alkali-metal-cation exporters have been been constructed and their phenotypes compared.

The C. glabrata genome contains only one copy of the CNH1 and ENA1 genes. CNH1 gene is present in one copy in other Candida genomes as well, however, the ENA genes are usually present in two copies (Table 5). The obtained results clearly showed that heterologous expression in another yeast species does not always provide reliable evidence of the physiological role of the transporter in its native yeast species. We have shown that although both transporters are able to recognize and transport various alkali metal cations (i.e. have broad substrate specificity) upon heterologous expression in S. cerevisiae, their physiological role in C. glabrata cells is specialized, the Cnh1 antiporter being involved in the regulation of intracellular potassium concentration and the Ena1 ATPase in sodium detoxification. This is in contrast with the situation in S. cerevisiae, where the two systems with different transport mechanisms (primary active ATPase and secondary active cation/proton antiporter) complement each other and, under standard growth conditions, contribute both to sodium detoxification and potassium homeostasis (Banuelos, 1998).

As far as the C. glabrata Cnh1 antiporter's activity in S. cerevisiae cells is concerned, CgCnh1p seems to be slightly more active than C. dubliniensis Cnh1p, but less active than C. parapsilosis and C. albicans Cnh1 proteins expressed from the same plasmid and promoter in the same S. cerevisiae strain (Krauke & Sychrova, 2008). Similarly, C. glabrata cells are more salt tolerant than C. dubliniensis and less salt tolerant than C. albicans and C. parapsilosis, so as in the abovedescribed case of C. dubliniensis, the level of C. glabrata salt tolerance may reflect the low contribution of the Cnh1 antiporter. On the other hand, C. glabrata Ena1p has a high activity in S. cerevisiae cells and is fully able to complement the absence of S. cerevisiae's own ATPases, both in sodium detoxification and the elimination of surplus potassium cations (Fig. 4).

Construction of the first C. glabrata strains lacking either Cnh1 or Ena1 transporters, and the characterization of the deletion and reintegration mutants showed that though both proteins are able to bind and transport sodium and potassium (Figs 1, 2 and 4), they are not involved in eliminating the surplus of both cations in C. glabrata cells (Figs 3, 5 and 6). The ena1Δ mutant was significantly less tolerant to sodium and lithium but had similar growth to wild-type cells in the presence of potassium and rubidium, and vice versa, cnh1Δ cells did not grow well in the presence of high K+ and Rb+ concentrations but withstood Na+ and Li+ stress well (Figs 3 and 5). Transport measurements strengthened these findings, as the deletion of ENA1 had no effect on potassium export but led to the reduction of sodium efflux, and in contrast, the deletion of CNH1 left Na+ extrusion unchanged but almost completely abolished K+ efflux. Surprisingly, deletion of the ENA1 gene, though it led to a high sodium sensitivity (Figs 5 and 6) did not abolish the Na+ efflux as much as CNH1 deletion for K+ export (ena1Δ cells were still able to eliminate about 70% of internal sodium within an hour), which suggests the presence of another, as yet unknown, efficient sodium exporter that is probably active in the short period of efflux measurement under nongrowing conditions and not under the conditions when the cells adapt to the presence of salt and begin to grow (Fig. 5). This is in contrast to S. cerevisiae, where the Nha1 antiporter and Ena ATPases are the only two exporters of sodium from cells (Arino, 2010).

The different physiological role of the two C. glabrata transporters (detoxification of sodium and lithium cations by Ena1 and maintenance of potassium homeostasis by Cnh1) is also reflected in the regulation of their expression, as CgENA1 is highly upregulated when cells react to increased osmotic pressure or sodium concentration, whereas CgCNH1 is expressed constitutively and in rather low levels under all tested conditions. The only significant change in CgCNH1 expression observed was a slight decrease in the amount of transcripts after 60 and 90 min in sorbitol. This might suggest that upon nonionic osmotic shock, cells try to diminish the efflux of potassium, as intracellular K+ is needed to compensate for the external osmotic pressure. The effect of sorbitol on the Na+/H+ antiporter's activity has been already observed at the post-translational level [transient inactivation of ScNha1p upon sorbitol shock (Kinclova-Zimmermannova & Sychrova, 2006)] and our results suggest that the regulation of intracellular potassium content may also proceed via the repression of a potassium-exporting system.

CgENA1 was most upregulated in the presence of NaCl, but a significant induction of expression was also observed in the presence of KCl and the nonionic osmolyte sorbitol. This might mean that at least two signalling pathways contribute to the induction, as is the case with S. cerevisiae, where ENA1 expression is regulated by several signalling pathways. High sodium stress acts via the calcineurin (Mendoza, 1994; Marquez & Serrano, 1996) and TOR pathways (Crespo, 2001); osmotic stress regulates ENA1 expression via the HOG pathway (Proft & Serrano, 1999) and alkaline stress via the calcineurin, Rim101 and Snf1 pathways (Ruiz & Arino, 2007). In our experiments, the level of CgENA1 transcription increased for about 60 min and then dropped slightly. A similar time-dependent pattern of ScENA1 and CaENA21 expression was observed under similar stress conditions in S. cerevisiae and C. albicans cells (Hirasawa, 2006; Enjalbert, 2009). On the other hand, in contrast to S. cerevisiae, where the expression of ENA genes under standard conditions (i.e. without stress) is thought to be very low, the difference in sodium content between the C. glabrata wild-type and ena1Δ cells observed after only 30 min of incubation with 400 mM NaCl (131. ± 29.2 and 172.3 ± 34.7 nmol mg−1 dry wt, respectively) suggests that CgEna1 is significantly active in growing nonstressed cells.

In conclusion, our results show that though C. glabrata is phylogenetically more related to S. cerevisiae than to C. albicans, the involvement of two types of alkali-metal-cation transporters in C. glabrata cell physiology might be more similar to the situation in C. albicans than in S. cerevisiae. We propose that the CgCnh1 antiporter plays an important housekeeping role in potassium homeostasis as the major K+ exporter but not in sodium export, whereas CgEna1 ATPase is mainly involved in the sodium and lithium detoxification of C. glabrata cells via its efflux capacity.


This work was supported by the EU grant MRTN-CT-2004-512481 CanTrain and Czech grants MSMT LC531 and AV0Z50110509. We thank Pavla Herynkova for construction of YEp352-SAT1 and technical assistance, Karl Kuchler and Joachim Morschauser for providing the C. glabrata strain and SAT1 deletion cassette.


  • Editor: André Goffeau


View Abstract