Two genes in Saccharomyces cerevisiae, ALR1 and ALR2, encode proteins putatively involved in Mg2+ uptake. The present study supports this role for ALR1 and provides the first electrophysiological characterisation of this protein. The patch-clamp technique was used to measure whole-cell ion currents in protoplasts prepared from the wild-type strain, the alr1 alr2 double mutant (CM66), and the double mutant over-expressing the ALR1 gene (CM66+ALR1). With 50 mM Mg2+ in the bathing solution, the inward current in protoplasts of CM66+ALR1 averaged −264±48 pA at −150 mV. Inward currents measured in the wild-type and CM66 protoplasts were more than five-fold smaller. When Mg2+ was the major cation in the pipette solution, time-dependent outward currents were also detected in CM66+ALR1 protoplasts suggesting ALR1 can facilitate Mg2+ efflux as well as uptake. We conclude that the ALR1 gene encodes a transport protein. The large magnitude of the Mg2+-dependent currents suggests that ALR1 could function as a cation channel.
Magnesium is the most abundant divalent cation in living cells yet, by comparison to most other macro- and micronutrients, relatively little is known about its role in biology. While Mg2+ has recognised roles in maintaining membrane stability and as a cofactor with ATP, it is increasingly clear that changes in free cellular Mg2+ concentrations can regulate different metabolic reactions in plants, animals and microorganisms [1,,,,,6]. The poor understanding of Mg2+ transport processes has hampered progress in this area, especially in eukaryotic cells. In bacteria, Mg2+ uptake primarily occurs via the constitutively expressed CorA transport system. Homologues of CorA have been identified in most fully sequenced microbial genomes as well as in the model plant species Arabidopsis thaliana where a 10-member family was described recently. Saccharomyces cerevisiae has two genes, designated ALR1 and ALR2, that encode proteins with a low degree of similarity to CorA. These genes were first identified from their ability to increase Al3+ tolerance in yeast cells. MacDiarmid and Gardner suggested that ALR1 and ALR2 encode Mg2+ transporters because the alr1 alr2 mutant grew poorly on standard medium unless high concentrations of Mg2+ were supplied in the medium. Additional evidence for involvement of ALR1 in Mg2+ uptake was presented by Graschopf et al.. They showed that ALR1 is localised in the plasma membrane of yeast cells and that its expression and turnover are controlled by Mg2+ concentration. Furthermore, mutants lacking ALR1 contained less Mg2+ and were unable to replenish their intracellular pools when supplied with Mg2+ following a period of starvation. The details of how the CorA or ALR proteins function to transport Mg2+ into cells remain unclear. Sensitivity of Mg2+ uptake to inhibitors such as 2,4-dinitrophenol could indicate an active transport mechanism coupled to proton movement. However this result could equally indicate that ALR1 is a passive transporter, such as a channel, whose function depends upon the maintenance of the electrical potential difference across the plasma membrane (ΔEio). Indeed the driving force generated by ΔEio in yeast and bacteria is sufficient to energise Mg2+ uptake for normal growth [11,12]. In this study, we provide electrophysiological evidence that the ALR1 gene encodes a transport protein involved in Mg2+ uptake by yeast. We conclude that ALR1 is likely to function as a cation channel.
Materials and methods
Yeast strains and growth conditions
Yeast strains were grown on standard synthetic medium (SC with the required auxotrophic supplements and 2% glucose) or on complete medium (YPD). Two yeast strains were derived in this study as isogenic derivatives of FY833. CM52 (MATαhis3-Δ200, ura3-52, leu2-Δ1, lys2-Δ202 trp1-Δ63) is wild-type for the two ALR genes while CM66 has both ALR genes deleted (MATa alr1::HIS3 alr2::TRP1 his3-Δ200, ura3-52, leu2-Δ1, lys2-Δ202 trp1-Δ63). The ALR1 ALR2 deletion mutants were generated by one-step gene disruption using an alr1::HIS3 polymerase chain reaction product or a palr2::TRP1 plasmid (analogous to palr2::URA3). Wild-type (CM52) and CM66 were derived from segregation of the same heterozygous diploid. Growth of CM66 strains was achieved by supplementing SC or YPD with 0.25–0.4 M MgCl2. Methods for yeast transformation were as described by Kaiser et al. and Gietz et al.. The plasmid pFLN2 was derived from the pFL61 high copy number plasmid by the addition of Xho I and Sal I sites to the multiple cloning region (MacDiarmid, unpublished). pFLN2 was used to express ALR1 in CM66 using the constitutive yeast phosphoglycerate kinase promoter.
Preparation of yeast protoplasts
Yeast were grown on a rotary shaker (100 rpm at 25°C) and protoplasts were prepared using a procedure described previously. Cells were harvested by centrifugation at 500×g for 5 min, and resuspended in 3 ml buffer A (50 mM KH2PO4 pH 7.2, titrated with KOH, plus 0.2%β-mercaptoethanol). The cells were incubated in a water bath at 30°C for 30 min with slow shaking (∼30 rpm). Buffer B (3 ml, buffer A supplemented with 2.4 M sorbitol, 2 mg ml−1 zymolyase, 2 mg ml−1 glucuronidase, and 50 mg ml−1 bovine serum albumin) was added to the yeast cells and mixed gently at 30°C for 30 min. The cells were collected by centrifugation (200×g, 5 min) and the protoplasts were released after resuspension in buffer C (150 mM KCl, 10 mM CaCl2, 5 mM MgCl2, 5 mM MES, pH 7.5 adjusted with Trizma). The protoplasts were collected and resuspended in buffer C plus 1% glucose and typically had diameters of 2–7 μm. Experiments were performed on protoplasts prepared on the same day.
Whole-cell patch clamping
Patch-clamping procedures were similar to those described by Bertl et al. with some differences. Uncoated, borosilicate glass pipettes (Clark Electromedical Instruments, UK, 1.5 mm O.D. and 0.86 mm I.D.) were prepared with a P-97 pipette puller (Sutter Instruments, USA). The protoplasts were transferred to a recording chamber containing a sealing solution to assist seal formation (150 mM KCl, 10 mM CaCl2, 5 mM MgCl2, 1 mM MES, pH 7 adjusted with Tris base). The cells were allowed to settle for 10–20 min and then the debris and unattached cells were washed away by steady perfusion of sealing solution for 5–10 min. Unless stated otherwise the standard pipette solution was: 150 mM K-acetate, 5 mM Mg-acetate, 5 mM KCl, 4 Tris–ATP, 1 mM EGTA, pH 7 with Tris base. Resistance of the pipettes was 5–10 MΩ with the sealing solution in the bath. The chloride concentration in the pipette solution was kept low to avoid the possibility that inward currents generated from chloride efflux could be confused with Mg2+-dependent inward currents. A GΩ seal was successfully achieved in a minority of cells after 10–20 min of constant and gentle suction. The following three criteria were used to help identify when the whole-cell patch configuration was obtained: (i) a sudden increase of membrane conductance, (ii) increase of background noise, or (iii) the appearance of a small time-dependent outward current characteristic of the DUK1 K+ channel. Small outward currents were often observed in protoplasts from all the strains but their magnitude and the time dependence were significantly smaller than reported previously for the K+ channel DUK1. It is likely that the low Ca2+ concentrations used in the pipette here inhibited the DUK1-dependent currents (A. Bertl, personal communication). Once the whole-cell configuration was obtained, the sealing solution was replaced with a bath solution that contained a high Mg2+ concentration (80 mM Tris acetate, 50 mM Mg-acetate, 5 mM KCl, 1 mM CaCl2, 1 mM MES, titrated to pH 7 with Tris base). In some experiments the high Mg2+ bath solution was replaced with a Mg2+-free solution (150 mM Tris–acetate, 5 mM KCl, 2 mM Ca-acetate, 1 mM MES, pH 7). The effect of aluminium on whole-cell currents was tested by using a high Mg2+ bath solution adjusted to pH 4.1 and then replacing it with a similar solution containing 0.5 mM AlCl3. The concentration of free Al3+ in this solution was calculated with the GEOCHEM program to be approximately 200 μM which corresponds to an activity of 20 μM. Current–voltage data were collected by holding the membrane potential at 0 mV and then stepping the voltage between −180 and +80 mV for 2 s intervals in 20 mV increments. The range of command voltages was reduced when the currents were very large or when they appeared unstable during the pulses. Current–voltage curves were generated from currents measured near the end of 2 s voltage pulses. Currents were amplified and filtered at 2 kHz using an Axopatch 200B amplifier (Axon Instruments) and digitised online by an IBM computer at 10 kHz (pClamp8.01/Digidata 1200B, Axon Instruments). Junction potentials for the high Mg2+ and Mg2+-free solutions were 10 mV and 12 mV respectively. The current–voltage curves are corrected for the liquid junction potentials. Leak currents were not subtracted from the data presented.
Whole-cell currents were initially compared in three yeast strains with 50 mM Mg2+ in the bathing solution. Measurements were made on the CM66 (alr1 alr2 knockout strain), CM66+ALR1 (CM66 transformed with the ALR1 gene) and the wild-type strain (CM52). Inward and outward currents in CM66 were small and neither exceeded 200 pA in any individual cell. The average current at −150 mV was −51±26 pA (n=12). Currents measured in the wild-type strain (CM52) were very similar to CM66 with inward currents averaging −50±27 pA (n=12) at −150 mV. By contrast, large and variable inward currents were observed in 78% of the CM66+ALR1 protoplasts examined. The magnitude of these currents ranged from −200 to −1500 pA (Fig. 1b) with a mean inward current of −264±48 pA (n=41) at −150 mV (Fig. 1c). The kinetics of current activation varied with voltage and between protoplasts but instantaneous and time-dependent components were observed at the more negative voltages. When Mg-acetate was removed from the bath solution and replaced with Tris–acetate, the large inward currents were reduced in 90% of protoplasts tested (n=9, Fig. 2). The outward currents also became smaller when Mg2+ was removed from the bathing solution. In protoplasts where the inward currents were very large (>1000 pA) the tail currents were often slower to relax to the pre-pulse levels. This has been reported previously in other yeast strains exhibiting large inward currents (see Fig. 3b in)
A new set of solutions were used to determine whether ALR1 can facilitate Mg2+ efflux as well as influx. In these experiments the pipette solution contained 50 mM Mg2+, 5 mM K+ and 10 mM TEA-Cl to inhibit any outward K+ currents. In these conditions, large inward and outward currents were observed in seven of 13 protoplasts examined. The remaining six protoplasts had small currents which possibly indicated that the expression of the ALR1 was poor in these cells. The average currents measured over all 13 protoplasts were −412±107 pA at −150 mV and +212±62 pA at 90 mV respectively (Fig. 3b). The currents averaged from the seven protoplasts showing the large currents only were −668±136 pA at −150 mV and +367±74 pA at 90 mV. In four experiments the 50 mM Mg2+-acetate in the bathing solution was replaced with 75 mM Tris–acetate. With this treatment both the inward and outward currents were significantly inhibited (Fig. 3b).
The effect of aluminium on the Mg2+-dependent inward currents was measured in the CM66+ALR1 strain. High Mg2+ concentrations were included in the pipette and bathing solutions in these experiments and the bathing solution was adjusted to pH 4.1 to ensure that the Al3+ cation was the predominant species present in the bathing solution. Large inward and outward currents were still observed at this pH (Fig. 4a). The addition of 500 μM AlCl3 to the bath solution (equivalent to an activity of free Al3+ cations of 20 μM) significantly decreased the inward and outward currents in three separate cells. Removal of Al3+ led to a recovery of both currents after a 5–20-min washout period. The CM66 control was omitted from this experiment because no Mg2+-dependent currents were detected in this strain.
Our results provide evidence that ALR1 encodes a protein involved in Mg2+ transport in yeast. We used the whole-cell patch-clamp technique to compare the Mg2+-dependent currents in the alr1 alr2 mutant (CM66) with the same strain over-expressing ALR1 (CM66+ALR1). Large Mg2+-dependent inward currents were measured in a majority of CM66+ALR1 protoplasts while no comparable currents were detected in the CM66. High external Mg2+ concentrations were used in this study to exaggerate the effects on the whole-cell currents. While 50 mM Mg2+ is a higher concentration than yeasts normally experience in their environment, it is not a toxic concentration and wild-type S. cerevisiae can grow well in 400 mM Mg2+.
It is unlikely that the large currents measured in the CM66+ALR1 strain are caused by other endogenous transporters, because inward currents of this magnitude are uncommon in wild-type S. cerevisiae. We are aware of only one report showing endogenous currents of an equivalent magnitude. In that study, Bihler et al. described large inward currents in wild-type yeast which are activated by low external divalent cation concentrations (<10 μM) and the authors attributed these currents to a non-specific cation channel designated NSC1. However, the inward currents measured in the present study cannot be attributed to NSC1 because activation of NSC1 requires a low external cation concentration while the ALR1-dependent currents were observed in the presence of 50 mM Mg2+. NSC1-dependent currents are also inhibited by low external pH but, in the present work, the inward currents were observed at pH 4.1. Finally, unpublished results show that the NSC1 currents are present in the CM66 strain which demonstrates conclusively that NSC1 cannot be attributed to ALR1 (H. Bihler and A. Bertl, personal communication). It is also unlikely that the large inward currents we measured can be attributed to Trk2p, another transporter found to generate inward currents in yeast. The magnitude of Trk2p-dependent currents at pH 7.0 are very small (<−50 pA at −200 mV) and independent of external Mg2+. By contrast, the inward currents measured in the present study at pH 7.0 are up to 40-fold greater and dependent upon external Mg2+. However, Trk2p-dependent currents do increase as pH declines because Trk2p appears to facilitate proton uptake. Therefore, it remains possible that a proportion of the inward current presented in Fig. 4 is due to Trk2p activity since pH 4.1 was used in those experiments.
Sodium-coupled Mg2+ transport has been demonstrated in mammalian systems [22,23] and Mg2+/H+ exchange has been measured across the tonoplast of plant cells but other transporters may also exist. Sodium was not included in the pipette solutions in these experiments, which precludes the involvement of ALR1 in a Mg2+/Na+ exchange reaction. In addition, the Mg2+-dependent inward currents measured here were observed when the external pH was 7.1 and 4.1 (Figs. 1 and 4), making it unlikely that ALR1-dependent Mg2+ uptake relies on symport or antiport with H+. Instead, the present evidence supports the hypothesis that ALR1 functions as an ion channel, rather than another type of carrier. Furthermore, the magnitude of the inward current is too large to be attributed to a non-channel transporter which typically functions 1000-fold more slowly than channels. For instance, the currents measured in the CM66+ALR1 strain exposed to high external Mg2+ often exceeded 500 pA per cell and sometimes reached 1500 pA per cell. A current of 500 pA per cell is equivalent to a current density of approximately 1000 μA cm−2 or 3×107 divalent ions μm−2 s−1 (assuming a cell diameter of 4 μm). A carrier-type transporter can facilitate ion movement at up to approximately 1000 ions s−1 which means that 30 000 transport proteins per μm−2 membrane area would be required to account for the currents measured. If we assume the ALR1 protein has a diameter of 60 Å and occupies an area of 3×10−5μm−2, then the calculated protein density exceeds the maximum protein distribution possible in the membrane. Such a density of proteins would prohibit normal growth and development. Yet the CM66+ALR1 cells have a normal phenotype with no special nutritional requirements suggesting that the other membrane proteins function normally.
The ALR1 and ALR2 genes were first identified from their ability to confer tolerance to aluminium toxicity when over-expressed in S. cerevisiae. The same study showed that aluminium reduced Mg2+(57Co2+) uptake by S. cerevisiae and it was concluded that aluminium toxicity was related to the inhibition of Mg2+ uptake via the ALR1 transporter. The present study supports this conclusion by demonstrating that Al3+ reversibly inhibits the Mg2+-dependent inward currents present in CM66+ALR1 cells.
The yeast ALR proteins show structural similarities to the bacterial CorA family of magnesium transporters. In particular, both the ALR and CorA proteins have a large, highly charged N-terminal domain with two hydrophobic regions at the C-terminal end. Gene fusion experiments with the Salmonella typhimurium CorA gene indicate that there are likely to be three membrane-spanning domains at the C-terminus of this protein. The electrophysiological evidence presented here suggests that ALR1 may function as a Mg2+ channel and cation channels with two or three transmembrane domains have been characterised in bacteria. For instance, KcsA, a K+-selective channel, and MscL, a bacterial mechanosensitive non-selective cation channel, form homo-tetramers and homo-pentamers respectively (reviewed by). Szegedy and Maguire recently reported ‘preliminary evidence’ that CorA also forms pentamers.
The presence of Mg2+-dependent inward and outward currents in CM66+ALR1 suggests that ALR transports Mg2+ into and out of yeast protoplasts. The reduction in both outward and inward currents when Mg2+ was removed from the bathing solution was initially surprising but both these characters have been observed for the CorA system in bacteria. For instance, Mg2+ efflux via the CorA transporter in Salmonella typhimurium is dependent on external Mg2+ such that when Mg2+ is absent from the external solution efflux from the cells ceases. Snavely et al. concluded that the effect of external Mg2+ on efflux could not be explained by a simple exchange reaction. They suggested instead that external Mg2+ somehow regulates the CorA transporter. Indeed, Mg2+ has been proposed to regulate ion channel activity in a number of animal and plant cells [6,29,,31]. In conclusion we suggest that the large Mg-dependent currents measured in CM66+ALR1 provide evidence that ALR1 functions as an ion channel.
The authors are grateful to Adam Bertl for his comments on an early version of the manuscript.
- pulmonary arteriogram
- pulmonary artery
- antimicrobial chemoprophylaxis
- aldehyde reductase
- carrier proteins
- patch-clamp techniques
- physician assistant
- plasminogen activator
- saccharomyces cerevisiae
- pipette (physical object)