OUP user menu

Genetic interactions among the Arl1 GTPase and intracellular Na+/H+ antiporters in pH homeostasis and cation detoxification

Lydie Marešová, Hana Sychrová
DOI: http://dx.doi.org/10.1111/j.1567-1364.2010.00661.x 802-811 First published online: 1 November 2010


The roles of intracellular GTPase Arl1 and organellar cation/H+ antiporters (Kha1 and Nhx1) in Saccharomyces cerevisiae tolerance to various stress factors were investigated and interesting new phenotypes of strains devoid of these proteins were found. The role of Arl1 GTPase in their tolerance to various cations is not caused by an altered plasma-membrane potential. Besides the known sensitivity of arl1 mutants to high temperature, we discovered their sensitivity to low temperature. We found for the first time that in the absence of Arl1p, Kha1p increases potassium, sodium and lithium tolerance, and can thus be categorized as an antiporter with broad substrate specificity. Kha1p also participates in the detoxification of undesired chemical compounds, pH regulation and growth at nonoptimal temperatures. Cells with the combined deletions of all three genes have considerable difficulty growing under nonoptimal conditions. We conclude that Arl1p, Kha1p and Nhx1p collaborate in survival strategies at nonoptimal pH, temperatures and cation concentrations, but work independent of each other.

  • Saccharomyces cerevisiae
  • cation detoxification
  • Kha1p
  • Nhx1p
  • Arl1p


In all eukaryotic cells, vesicle-mediated transport is a powerful tool for intracellular traffic. A large number of proteins are involved in vesicle assembly, mobility and disassembly, including primarily small G proteins, such as members of the ADP-ribosylation factor/ADP-ribosylation factor-like (ARF/ARL) family (Li, 2004). Arl proteins are highly conserved through phylogenetic groups from yeast to mammals (Bourne, 1991). Saccharomyces cerevisiae expresses two Arl proteins (Arl1p and Arl3p), both involved in the regulation of vesicular trafficking at the trans-Golgi network (Graham, 2004).

There has been strong evidence that Arl1p also plays a significant role in the cation homeostasis of yeast cells (Love, 2004; Munson, 2004a), but the specific molecular basis for this function has not been fully elucidated. The easiest explanation would be that Arl proteins, as regulators of vesicle trafficking, control the delivery of cation transporters to their proper cellular localization. In some cases, it was actually shown that the Sys1p-Arl3p-Arl1p-Imh1p signaling cascade is necessary for a plasma-membrane protein to reach its destination (Liu, 2006). However, other experimental data also suggest other mechanisms – for example that potassium influx is influenced by Arl1p through the positive regulation of Hal4 and Hal5 kinases, rather than influencing the delivery of the Trk1 potassium transporter to the plasma membrane (Munson, 2004b). Another work by the same group shows that the lithium sensitivity of arl1 mutants can be suppressed by overexpression of the plasma-membrane ATPase Ena1 (Munson, 2004a). We decided to perform a more detailed study of the interconnections between Arl1p and the cation transporters in the tolerance of cells to extracellular cations and other stresses, focusing mainly on the intracellular part of this process (using strains without alkali-metal-cation exporters). In the absence of the main players in alkali-metal-cation detoxification (the plasma-membrane Ena1-4 ATPases and the Nha1 antiporter), the role of intracellular transport systems becomes more evident – surplus alkali metal cations cannot be actively extruded from cells and internal defensive mechanisms (e.g. organellar sequestration) must prevent their adverse effects. In particular, we investigated the relationships between Arl1p and intracellular Nhx1 and Kha1 transporters. The first of them, the Nhx1 antiporter, was shown to be localized to the membrane of late endosomes (LE, called also prevacuolar compartments), regulating their trafficking and intracellular pH (Nass & Rao, 1998; Brett, 2005). This antiporter is involved in the cells' tolerance to salts, presumably by intracellular sequestration of alkali metal cations into the vacuole (Nass, 1997). The second, Kha1p, is localized in the Golgi apparatus (GA) (Flis, 2005), and although it belongs to the family of alkali-metal-cation antiporters, until now, it has failed to show any phenotype connected to alkali-metal-cation tolerance. Similar to Nhx1p, Kha1p plays a role in intracellular pH regulation (Marešová & Sychrová, 2005), and is believed to mediate K+/H+ antiport, although its substrate specificity has not been confirmed by direct transport measurements. Both Kha1p and Arl1p are important for growth in an alkali environment; deletion of the relevant genes results in an inability to grow at a high external pH (Giaever, 2002; Marešová & Sychrová, 2005). Deletion of the NHX1 and ARL1 genes, on the other hand, should result in reduced growth in acidic environments (Munson, 2004b; Brett, 2005).

Proper vesicular budding, routing and fusion are known to be of critical importance for the functionality of the secretory pathway, for vacuole biogenesis, organellar inheritance, etc. Large groups of mutants defective in some of these processes have been identified in yeast cells (e.g. vps mutants with affected vacuolar protein sorting or sec mutants with an affected secretory pathway). Nhx1p has already been identified as being responsible for one of the vps phenotypes [identical to Vps44p and the first ion transporter known in this group of proteins (Bowers, 2000)]. Several GTPases were matched with sec phenotypes (Schekman & Novick, 2004). Arl1p, although not listed among Sec or Vps proteins, was shown to be important for proper vacuole formation (Lee, 1997) and vesicle fusion with the trans-Golgi (Panic, 2003). As the role of the Nhx1 protein in vesicle transport and its interactions with small GTPases have been described quite well (Ali, 2004; Brett, 2005), we decided to focus more on the Kha1 antiporter. Current knowledge of its role in cell physiology is rather limited and no genetic interactions have been described.

Here, we show that Arl1 GTPase and both the intracellular antiporters are among the proteins important for the intracellular detoxification of various chemical compounds. Our results place Kha1p among the proteins regulating intracellular and intraorganellar cation homeostasis by vesicle trafficking; for the first time, we show its role in alkali-metal-cation tolerance, which is not restricted to potassium. We also demonstrate the role of Arl1p in alkali-metal-cation tolerance, independent of the plasma-membrane transporters Ena1-4 and Nha1. arl1 deletion, in many cases, enhances the kha1 and nhx1 deletion phenotypes.

Materials and methods

Yeast strains

Most S. cerevisiae strains used in this study were derivatives of the W303-1A strain (MATaade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1 ura3-1 mal10) (Wallis, 1989). Additional mutations are listed in Table 1. The AWA0, LMA01, ABA2, AWK0, WAK0, AWAK0, ABK0 and ABAK0 strains were prepared by replacing a gene in the corresponding parent strain with the loxP-KanMX-loxP sequence according to Güldener (1996). Where necessary, for further work, the deletion cassette was removed with Cre recombinase. Insertion of the deletion cassette into the correct chromosome locus was verified by PCR for all constructed strains. The KKB40-3 strain was constructed by backcrossing W303-1A and a W303-1B-derived kha1KanMX mutant kindly provided by Dr M.A. Bañuelos. One reference experiment was also performed with BY4742 (MATαhis3 leu2 lys2 ura3; Invitrogen) and an arl1 mutant derived from this strain (kindly provided by Dr Rajini Rao).

View this table:

Saccharomyces cerevisiae derivatives of W303-1A

WAR0arl1Δ∷loxP-KanMX-loxPMarešová & Sychrová (2007)
KKB 40-3kha1Δ∷KanMXThis work
AW11nhx1Δ∷TRP1Kinclova-Zimmermannova (2006)
AWA0nhx1Δ∷TRP1 arl1Δ∷loxP-KanMX-loxPThis work
AWK0nhx1Δ∷TRP1 kha1Δ∷loxP-KanMX-loxPThis work
WAK0arl1Δ∷loxP kha1Δ∷loxP-KanMX-loxPThis work
AWAK0nhx1Δ∷TRP1 arl1Δ∷loxP kha1Δ∷loxP-KanMX-loxPThis work
BW31ena1-4Δ∷HIS3 nha1Δ∷LEU2Kinclova-Zimmermannova (2005)
BWA0ena1-4Δ∷HIS3 nha1Δ∷LEU2 arl1Δ∷loxP-KanMX-loxPMarešová & Sychrová (2007)
LMB01ena1-4Δ∷HIS3 nha1Δ∷LEU2 kha1Δ∷loxP-KanMX-loxPMarešová & Sychrová (2005)
LMB11ena1-4Δ∷HIS3 nha1Δ∷LEU2 kha1Δ∷loxPMarešová & Sychrová (2006)
AB11cena1-4Δ∷HIS3 nha1Δ∷LEU2 nhx1Δ∷TRP1Marešová & Sychrová (2005)
LMA01ena1-4Δ∷HIS3 nha1Δ∷LEU2 kha1Δ∷loxP arl1Δ∷loxP-KanMX-loxPThis work
ABA2ena1-4Δ∷HIS3 nha1Δ∷LEU2 nhx1Δ∷TRP1 arl1Δ∷loxP-KanMX-loxPThis work
ABK0ena1-4Δ∷HIS3 nha1Δ∷LEU2 nhx1Δ∷TRP1 kha1Δ∷loxP-KanMX-loxPThis work
ABAK0ena1-4Δ∷HIS3 nha1Δ∷LEU2 nhx1Δ∷TRP1 arl1Δ∷loxP kha1Δ∷loxP-KanMX-loxPThis work


The plasmids are listed in Table 2. pNHX1 was prepared as follows: the NHX1 gene was amplified including its own promoter region (823 nt) from S. cerevisiae genomic DNA and inserted into YEp352 by homologous recombination. The resulting plasmid was verified by restriction analysis and sequencing of the NHX1 coding region.

View this table:


YEp352Multicopy vector, selection markers: URA3, AmpRHill (1986)
pARL1ARL1 with its own promoter in YEp352Rosenwald (2002)
pKHA1KHA1 with its own promoter in YEp352Marešová & Sychrová (2005)
pNHX1NHX1 with its own promoter in YEp352This work


Cells were grown in YPD medium (1% yeast extract, 2% Bacto peptone, 2% glucose, adenine 15 μg mL−1 and 2% agar for solid media). For salt and drug sensitivity tests and for the cultivation of strains transformed with a plasmid, YNB medium was used (0.67% yeast nitrogen base without amino acids, 2% glucose and 2% agar for solid media), with auxotrophic supplements added after autoclaving (adenine 20 μg mL−1, histidine 20 μg mL−1, leucine 30 μg mL−1, tryptophan 20 μg mL−1 and uracil 20 μg mL−1; uracil was excluded when working with strains harboring plasmids). Media were adjusted to the desired pH with KOH or HCl, pH 5.5–6.5 media were buffered with 20 mM MES, pH 6.6–7.0 media with 20 mM MOPS and media <pH 5.5 were unbuffered.

Growth assays

Growth was monitored either on solid (drop tests) or on liquid (growth curves) media supplemented as indicated in the text. For drop tests, fresh cells of each tested strain were resuspended in water and adjusted to the same initial OD600 nm=1.0 (corresponding to approximately 5 × 106 cells mL−1). Tenfold serial dilutions were prepared and 3-μL aliquots of each dilution were spotted on appropriate YPD or YNB agar plates supplemented as indicated in the text. Plates were incubated at 30 °C (usually for 2 days) and digital grayscale images were obtained using a Nikon Coolpix4500 digital camera.

Growth curves were measured at 595 nm in an ELx808 Absorbance Microplate Reader (Biotek) as described in Marešová & Sychrová (2007). Differences in the OD reached were compared for individual strains. The comparison is expressed as a percentage of the OD reached by the particular strain under standard conditions (media without added salts/drugs, at optimal pH, etc.) after 18 or 24 h, as indicated in Results.

Relative membrane potential

A comparison of the relative membrane potential was performed as described earlier (Marešová, 2006, 2009). Briefly, exponential cells were resuspended in pH 6, 10 mM citrate–phosphate buffer (with or without 2% glucose, as indicated in Results) and 3,3′-dipropylthiacarbocyanine iodide [diS-C3(3)] was added to a final concentration of 0.2 μM. Fluorescence emission spectra were measured on an ISS PC1 spectrofluorometer. The excitation wavelength was 531 nm, emission intensities were measured at 560 and 580 nm and the intensity ratio at equilibrium was compared for the studied strains.

Lithium uptake

Li+ influx was estimated using a method derived from Kinclova (2001). Cells were grown in YNB medium to the early exponential phase, harvested by centrifugation and resuspended in fresh YNB containing 100 mM LiCl. After 60 min, cells were harvested, washed in deionized water and resuspended in Li+-free buffer consisting of 0.1 mM MgCl2, 20 mM RbCl and 20 mM MES (Sigma-Aldrich) adjusted to pH 5.5 with Ca(OH)2. Samples of cells were filtered through Millipore membrane filters (0.22 μm pore diameter), washed rapidly with 20 mM MgCl2 and acid extracted. The Li+ concentration in the extraction solution was analyzed by atomic absorption spectrophotometry. Samples were measured in triplicate; data shown are the average from at least three independent experiments.


Cation and drug sensitivity

First of all, we wanted to know whether deletion of the ARL1 or KHA1 gene influences the growth rate of strains under standard conditions or their osmotolerance. We tested the growth of the W303 wild type and the corresponding arl1 and kha1 mutants. The tested strains differed significantly neither in the duplication time nor in the tolerance to sorbitol in the growth medium (data not shown).

On the other hand, we were able to observe an increased sensitivity of the mutant strains to three cationic drugs believed to enter cells proportional to the plasma-membrane potential (hygromycin B, tetramethylammonium, spermine, Fig. 2). This suggests that the tested proteins Arl1 and Kha1 influence the uptake and/or the intracellular processing of various drugs. The arl1 mutant was sometimes more sensitive than the kha1 mutant (hygromycin B, tetramethylammonium); sometimes, their sensitivities were the same.


Relative plasma-membrane potential comparison of W303 and WAR0 (arl1) strains. The equilibrium fluorescence intensity ratio (I580/560 ratio) reflects the relative electrochemical membrane potential (a higher ratio corresponds to a higher potential). Data are average of three independent experiments; error bars correspond to SD.

It was suggested earlier that the increased sensitivity of arl1 mutants to cationic compounds such as hygromycin B or tetramethylammonium might be due to the increased electrochemical potential at the plasma membrane (Munson, 2004b). However, we did not observe any difference in diS-C3(3) staining (an assay used for monitoring changes in the membrane potential) for the arl1 mutant (Fig. 2). Both strains (W303 and WAR0) had a naturally higher membrane potential in the presence of glucose, which activates Pma1 H+-ATPase (Serrano, 1983), but there was no significant difference between them. We showed previously that neither the kha1 nor the nhx1 mutant, both sensitive to drugs such as hygromycin B, had an altered membrane potential (Kinclova-Zimmermannova, 2006; Marešová, 2006). The sensitivity of mutants to various drugs is thus most likely caused by something other than an increased uptake due to membrane potential driving force. This assumption is also supported by the fact that both mutants were also sensitive to the anionic compound deoxycholate (Fig. 1) – a negatively charged drug does not enter cells in direct proportion to their membrane potential.


Drug sensitivity of W303, WAR0 (arl1) and KKB40-3 (kha1) strains. Cells were incubated for 24 h at 30°C and the OD reached was compared. Hygromycin B, tetramethylammonium (TMA) and spermine were tested in YNB medium; deoxycholate was tested in YPD, because in YNB, deoxycholate precipitated at the concentration required for the experiment. Data are representative of three independent experiments. Each sample was measured in quadruplicate; the average SD was 5%.

To determine how the intracellular antiporter and the Arl1 GTPase are involved in coping with the increased cytosolic concentrations of cations, we decided to characterize the phenotype of their absence in cells lacking plasma-membrane efflux systems (strain BW31 ena1-4 nha1), i.e. in cells with a higher salt sensitivity and a higher alkali-metal-cation content upon salt stress. For comparison, we also included the nhx1 mutant (known to be more sensitive to salts in this genetic background).

In the ena1-4 nha1 genetic background, the arl1 and nhx1 mutant strains were more sensitive to sodium and lithium (Fig. 3), while the kha1 mutant strain grew as well as the control strain (BW31). We selected lithium for a more detailed investigation and performed a comparison of all possible mutation combinations (arl1, kha1, nhx1) (Fig. 4a). The lithium sensitivity of yeast cells reflects the toxicity of the cation, while high potassium sensitivity often mainly reflects the osmotic effect. NaCl sensitivity combines both, as the sodium cation is more toxic than potassium, but far less than lithium (Gaxiola, 1992; Glaser, 1993). Figure 4a shows that the effect of mutations on lithium sensitivity is cumulative – the most sensitive strain of all was the triple mutant (arl1 kha1 nhx1). Unexpectedly, it was also found that Kha1p plays a role in lithium tolerance, although this was not observed in a previous study [comparing just the strains BW31 and LMB01 (Marešová & Sychrová, 2005)]. Here, in the absence of the ARL1 gene, the kha1 deletion significantly increased lithium sensitivity (compare the arl1 mutant with arl1 kha1 or arl1 nhx1 with the triple mutant, Fig. 4a). Surprised by this finding, we also tested the sodium and potassium sensitivity of the double and triple mutants, and also in the presence of NaCl and KCl, the arl1 nhx1 double mutant grew better than the arl1 kha1 nhx1 triple mutant (Fig. 4b). For the first time, the kha1 deletion phenotype showed that the Kha1 antiporter could probably transport all alkali metal cations, and similar to Nha1p and Nhx1p, thus belongs to the category of antiporters with broad substrate specificity.


Salt sensitivity in the absence of Ena1-4p and Nha1p. BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1) and AB11 (ena1-4 nha1 nhx1) strains were grown in YNB medium supplemented with salts as indicated for 18 h at 30 °C and the OD reached was compared. Data are representative of three independent experiments. Each sample was measured in quadruplicate; the average SD was 5%.


Salt sensitivity of strains with combined mutations. (a) Lithium sensitivity. BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1), AB11 (ena1-4 nha1 nhx1), LMA01 (ena1-4 nha1 arl1 kha1), ABA2 (ena1-4 nha1 arl1 nhx1), ABK0 (ena1-4 nha1 nhx1 kha1) and ABAK0 (ena1-4 nha1 arl1 kha1 nhx1) strains were grown in YNB medium supplemented with LiCl for 18 h at 30°C and the OD reached was compared. Each sample was measured in quadruplicate; the average SD was 5%. (b) Sodium and potassium sensitivity of strains differing in the presence of the KHA1 gene. ABA2 (ena1-4 nha1 arl1 nhx1) and ABAK0 (ena1-4 nha1 arl1 kha1 nhx1) were grown in YNB medium supplemented with salts for 24 h at 30°C and the OD reached was compared. Each sample was measured in quadruplicate; the average SD was 5%.

To verify the influence of particular proteins in cation tolerance, we overexpressed the ARL1, KHA1 or NHX1 genes in the triple mutant from a multicopy vector (Fig. 5). All the tested proteins increased the LiCl tolerance of cells compared with cells harboring the empty plasmid, although the effect of Kha1p was much weaker than that of Arl1p and Nhx1p. Complementation of sodium and potassium sensitivity could only be observed for the Arl1 and Nhx1 proteins.


Complementation of lithium sensitivity. The strain ABAK0 (ena1-4 nha1 arl1 kha1 nhx1) transformed with quoted plasmids was grown on YNB supplemented with salts as indicated. Four transformants of each plasmid were tested; a representative picture is shown.

We inferred from these results that the differences in sensitivity to various chemicals caused by the mutations arl1 and/or kha1 reflect neither differing osmotolerance nor differing membrane potential, but perhaps something like an intracellular trafficking and detoxification process, which can include for example vacuole sequestration. The final experiment performed to support our hypothesis was cation influx measurements in an Li+-efflux less background (strains derived from BW31, which lacks Ena1-4 ATPases and the Nha1 antiporter). Cells were incubated with 100 mM extracellular LiCl for 60 min and the total amount of intracellular Li+ was estimated. Mutant cells (arl1 or kha1 or nhx1 or double mutants in any combination) did not exhibit a significantly increased lithium accumulation compared with the control strain BW31 (Table 3). This was perfectly consistent with our presumption that the lithium sensitivity of the mutants was not caused by increased lithium influx. Only the triple mutant (arl1 kha1 nhx1) accumulated slightly more Li+ than the other strains. The reason could possibly be more drastically affected secretory and vesicle recycling pathways and hampered biogenesis and/or degradation of proteins responsible for membrane integrity maintenance. The slower growth of the triple mutant (duplication time in YNB is 124±8 min for ABAK0 strain, 98±6 min for BW31) provides another hint that the combination of all three deletions arl1, kha1 and nhx1 causes considerably more problems for the cells than any one of the single deletions.

View this table:

Li+ uptake during incubation with LiCl

StrainsDeletions in addition to ena1-4 nha1Li+ content ± SD (nmol mg−1 dry weight) P
Compared with BW31 (%)Compared with ABAK0 (%)
BW3159 ± 60.04
BWA0arl158 ± 6800.08
LMB01kha160 ± 6920.16
AB11cnhx161 ± 4740.06
LMA0arl1 kha166 ± 380.28
ABA2arl1 nhx160 ± 1770.10
ABK0kha1 nhx164 ± 3190.16
ABAK0arl1 kha1 nhx184 ± 80.04
  • * Student's two-sided t-test with unequal variation; P≤1% is considered to indicate a significant difference.

pH homeostasis

We next tested the ability of strains to grow at various extracellular pH levels.

We could observe previously published sensitivity of the arl1 mutant to both high and low pH (Giaever, 2002; Munson, 2004b). The previously observed high pH sensitivity of the kha1 mutant (Marešová & Sychrová, 2005) was also obvious in our experiments and we have shown the growth defect to be cumulative at a high pH (the arl1 kha1 double mutant grew worse than both single mutants, Fig. 6).


Growth of strains with combined mutations at various extracellular pH levels. BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1) and LMA01 (ena1-4 nha1 arl1 kha1) were grown in YNB medium with pH adjusted to 2.7–7.8 (>pH 5 media were buffered) for 24 h and the OD reached was compared (separately for buffered and unbuffered media). Data are average of four independent experiments; error bars correspond to SD.

It is known that the high pHout sensitivity of kha1 mutants can be suppressed by the addition of extracellular potassium (Marešová & Sychrová, 2005), and other mutations causing sensitivity to high external pH could be suppressed by an increased concentration of glucose or divalent cations, for example Cu2+ (Serrano, 2004). We tested whether this was also true for the arl1 and/or the kha1 mutations. Figure 7 shows that 6% glucose slightly increases the growth ability of both the arl1 and the kha1 mutants at pH 8, and hardly increases that of the double mutant at all. KCl (50 mM) restores the growth of both kha1 mutants (the single kha1 and the double arl1 kha1) to the level of the corresponding control strain (BW31 and the arl1 mutant, respectively). For the first time, we show that 10 mM NaCl has an effect similar on the kha1 mutant to potassium, although to a lower extent, which provides further evidence of broader substrate specificity for Kha1p. CuSO4 (20 μM) almost eliminates the differences between the mutations and restores the growth of the mutant strains to almost the level of the control strain.


Suppression of pH sensitivity. Cell suspensions of strains BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1) and LMA01 (ena1-4 nha1 arl1 kha1) were spotted on YNB plates with 2% or 6% glucose (glc), pH 5.6 or 8.0 and supplemented or not with salts as indicated. A representative picture from two independent experiments.

This suggests that Kha1p and Arl1p collaborate in the cell tolerance to increased pH on different molecular bases. Copper seems to reduce the pH sensitivity on a more general basis, although not all known high-pH-sensitive mutants can restore their growth after the addition of copper (Serrano, 2004).


Cells defective in vesicle transport from the late Golgi to the LE, in direct transport from the late Golgi to the vacuole or in a homotypic vacuolar fusion, were previously shown to grow slowly at the restrictive temperature of 37 °C (Stein, 2009). Also, the arl1 mutation caused high temperature sensitivity, although not in all genetic backgrounds (Rosenwald, 2002). We found that in the W303 background, the arl1 mutant is not only sensitive to supraoptimal temperature but also to suboptimal (Fig. 8a). These phenotypes are clearly strain specific. For comparison, we show another widely used strain, BY4742, in which the arl1 mutation did not influence its growth at various temperatures.


Temperature sensitivity. (a) arl1 mutants in various genetic backgrounds. Cell suspensions were spotted on YPD medium and incubated at determinate temperatures for 2 days. (b) Strains with the arl1 and/or the kha1 mutation. Cell suspensions of BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1) and LMA01 (ena1-4 nha1 arl1 kha1) strains were spotted on YPD plates and incubated at various temperatures. Images are representative from three independent experiments.

The bidirectional temperature sensitivity of the arl1 mutant observed in the W303 background was also confirmed in the BW31 (ena1-4 nha1) strain (Fig. 8b). Interestingly, the additional deletion of kha1 slightly exacerbates arl1 sensitivity to 37 °C, but not to 21 °C.

Mutual suppression by overexpression

Previous results have shown that although single deletions and their combinations do not have the same phenotypes, the studied proteins (Arl1p, Kha1p and Nhx1p as well) take part in the detoxification of alkali metal cations and various chemical compounds, and their presence is important for the growth of cells at nonoptimal temperatures and pH levels. These similar roles of Arl1p, Kha1p or Nhx1p led us to investigate their mutual substitutability. Mutants derived from BW31 strain were transformed with multicopy plasmids bearing the ARL1, KHA1 or NHX1 gene and the suppression of their phenotypes was tested. Figure 9a shows that the overexpression of ARL1, KHA1 or NHX1 cannot replace the function of another gene in tolerance to hygromycin B, although each gene can complement its own deletion phenotype or increase the hygromycin B tolerance of a triple mutant (Fig. 9b). Similar results were observed for the same strains grown at increased pH, supraoptimal temperature or in the presence of salts (data not shown). This was the final proof that Arl1p, Kha1p and Nhx1p do not redundantly play the same role and additional copies of one cannot replace any of the others.


Mutual irreplaceability of Arl1, Kha1 and Nhx1 proteins. BW31 (ena1-4 nha1), BWA0 (ena1-4 nha1 arl1), LMB01 (ena1-4 nha1 kha1) and ABAK0 (ena1-4 nha1 arl1 kha1 nhx1) strains transformed with the YEp352 plasmid (empty or bearing the ARL1, KHA1 or NHX1 gene) were tested for hygromycin B sensitivity on YNB plates. (a) BW31, BWA0 and LMB01 strains were tested at 300 μg mL−1 hygromycin B and (b) the ABAK0 strain was tested at 50 μg mL−1 hygromycin B. Images are representative from two independent experiments.


Our work, which aimed to investigate the cooperation between Arl1 GTPase and intracellular alkali-metal-cation transporters, yielded several interesting findings. First, we were finally able to demonstrate that the Kha1 transporter plays a role in tolerance to alkali metal cations, which is only revealed in the absence of Ena1-4p, Nha1p and Arl1p. We also traced its substrate specificity and found that it was not limited to potassium, which is manifested by the salt sensitivity of the mutants with multiple deletions including kha1 (KCl, NaCl, LiCl) and by the complementation of its high pH sensitivity by both KCl and NaCl.

Second, we have confirmed the results of Anne Rosenwald and others that Arl1 GTPase plays a crucial role in cation tolerance; however, the theory of an increased membrane potential in arl1 mutants was disproved, as the sensitivity of the mutants does not depend on the charge of the toxic compound and the diS-C3(3) assay did not show any difference for the deletion mutant.

Mislocalization of plasma-membrane alkali-metal-cation transporters (Ena1-4, Nha1) is not the reason for the alteration in salt sensitivity either as the arl1 deletion also increases salt sensitivity in strains without those proteins.

The late endosomal/prevacuolar compartment itself has been suggested to be a target of salt toxicity (Hernandez, 2009), hinting at an important role for the LE resident Nhx1p. Our results strongly indicate that Golgi proteins also take part in the complex process of intracellular detoxification, not only for inorganic cations but also for various organic drugs. As we did not find a significant difference in the membrane potential or the toxic cation content between mutants with altered sensitivity to the tested cations, we speculate that the phenotype differences are not primarily caused by an altered uptake and/or an efflux of the chemicals. The tested compounds had diverse chemical structures and mechanisms of action; thus, the involvement of the studied proteins in their specific detoxification is rather improbable. Affected vesicular transport in the mutants (which can hamper, e.g., vacuolar sequestration of undesired compounds or the correct trafficking of the enzymes necessary for detoxification) is an attractive hypothesis for future investigation.

The phenotypes of the arl1 and kha1 mutations usually exhibited cumulative effects when combined. This finding disproves the possibility of the three corresponding proteins following on from each other in one physiological process. There are several clues indicating that Arl1p, Kha1p and Nhx1p, although mediating similar physiological processes, work in an independent manner, each contributing to the desired process in a specific way. As an example, we can stress the cell tolerance to increased pH, in which both Arl1p and Kha1p play an important role, and the deletion of both exhibits a cumulative effect. Nhx1p does not participate in this process, perhaps related to the GA. We believe that Kha1p, which is a cation/H+ antiporter, might play a similar role at the Golgi membrane as Nhx1p in endosomes. By transporting K+ to the Golgi lumen and H+ out, it might change the pH of the inner milieu and help to maintain the balance between anterograde/retrograde vesicle transports. This hypothesis is supported by the fact that additional K+ in the cytosol (from extracellular KCl, Fig. 7) can complement the function of Kha1p at a high pH; thus, we speculate that the K+ concentration within GA might be the important factor in the balance of Golgi pH and/or intraorganellar turgor necessary for vesicle formation and trafficking. We believe that the Arl1 GTPase is another protein important for this trafficking, although not in direct interaction with Kha1p. The deletion of both genes (ARL1, KHA1) is not lethal, but significantly hampers the growth of cells at alkaline pH. We suggest that there are at least two molecular mechanisms by which the desired pH is maintained inside the GA: (1) proton/K+ transport across the membrane – mediated by the Kha1 antiporter and (2) balancing the vesicle budding and fusions, promoting either anterograde or retrograde transport, in which the small GTPases play an important regulatory role. Anyway, we and others have shown that the proper function of the GA is indispensable for yeast cells to grow at alkaline pH and any disturbance of its function would affect the cell's ability to survive environment alkalization.

The triple mutant (arl1 kha1 nhx1), besides cumulating the phenotypes of the single mutations (e.g. salt sensitivity), also showed some specific phenotypes not present in the single mutants (increased lithium accumulation, reduced growth rate). It seems that cells devoid of all three proteins, partially overlapping in their functions (Arl1, Kha1, Nhx1), are more severely affected in their secretory and vesicle recycling pathways, protein sorting, etc. However, the viability of this mutant shows that there is still another escape route, and under optimal conditions, cells are able to manage essential life processes without these proteins.

Although deletions of the genes ARL1, KHA1 or NHX1 have partially overlapping phenotypes, additional copies of one of them cannot replace the function of any other. These findings indicate again the existence of several physiological processes collaborating in intracellular detoxification processes and stress responses, one of which can usually only partially compensate for the absence of the others.

We conclude that all three proteins – the two intracellular antiporters and the Arl1 GTPase – participate in the detoxification and/or in the cell response to various extracellular stresses (high concentration of salts, drugs, nonoptimal temperatures and pH), although at different organelles and with diverse molecular processes (Fig. 10). The deletion phenotypes of mutant strains are similar, but the proteins do not participate in one physiological route; they contribute to the processes independent of each other. A more detailed study of their influence on intracellular vesicle transport, collaboration in restoring cation homeostasis, in maintaining optimal pH and the cation content in organelles (LE, GA) might yield interesting results in the future.


Proteins investigated in this study, involved in alkali-metal cation transport, drug resistance and vesicle trafficking. Plasma-membrane transporters (Ena1-4 and Nha1) mediate resistance by extruding cations from cells. Intracellular antiporters (Kha1, Nhx1), although also involved in cation resistance, mainly influence vesicle trafficking, intracellular sequestration of undesired chemicals and last, but not the least organellar pH. Arl1 GTPase participates in the same process, but does not seem to share the same pathway or interact directly with any of the cation transporters studied. Intraorganellar pH of GA can be influenced directly (proton transport mediated by Kha1p) or indirectly (vesicle fusion facilitated by Arl1p). GA, Golgi apparatus; LE, late endosome; V, vacuole.


We thank Dr Anne Rosenwald for the pARL1 plasmid and Dr Rajini Rao for the strains derived from BY4742. This work was supported by Czech grants MSMT LC531, AV0Z 50110509 and GA AS CR IAA 500110801 and KJB500110701.


  • Editor: André Goffeau


View Abstract