Ability of Sit4p To Promote K+ Efflux via Nha1p Is Modulated by Sap155p and Sap185p

We demonstrate here that SAP155 encodes a negative modulatorof K⁺ efflux in the yeast Saccharomyces cerevisiae. Overexpressionof SAP155 decreases efflux, whereas deletion increases efflux.In contrast, a homolog of SAP155, called SAP185, encodes a positivemodulator of K⁺ efflux: overexpression of SAP185 increases efflux,whereas deletion decreases efflux. Two other homologs, SAP4and SAP190, are without effect on K⁺ homeostasis. Both SAP155and SAP185 require the presence of SIT4 for function, whichencodes a PP2A-like phosphatase important for the G₁-S transitionthrough the cell cycle. Overexpression of either the outwardlyrectifying K⁺ channel, Tok1p, or the putative plasma membraneK⁺/H⁺ antiporter, Kha1p, increases efflux in both wild-typeand sit4 ${Delta}$ strains. However, overexpression of the Na⁺-K⁺/H⁺ antiporter,Nha1p, is without effect in a sit4 ${Delta}$ strain, suggesting that Sit4psignals to Nha1p. In summary, the combined activities of Sap155pand Sap185p appear to control the function of Nha1p in K⁺ homeostasisvia Sit4p.

INTRODUCTION

Top

Introduction

K⁺, implicated in translation and thus replication (22, 25-27),is the most abundant cation found in cells. K⁺ has been implicatedin numerous other important cellular processes. For example,depletion of K⁺ causes inhibition of endocytosis (13, 19-21),presumably because clathrin cages cannot be disassembled byHsc70, an enzyme requiring K⁺ for function (35, 48). K⁺ is crucialfor establishing membrane potential, for import and export ofmolecules into and out of cells, for pH control, and for osmolaritycontrol in plants and fungi (11, 42). Finally, recent work hasshown that ion homeostasis is linked to cell cycle regulation(7, 9, 10, 31, 50) and is mediated in part by a G₁-S regulatoryprotein, the phosphatase, Sit4p (31). Despite the importanceof K⁺ in this broad spectrum of cell functions, little is understoodabout the molecular mechanisms controlling intracellular K⁺levels.

Cytosolic K⁺ homeostasis is controlled by the competing activitiesof influx, efflux, and storage. In the yeast, Saccharomycescerevisiae, Trk1p mediates high-affinity uptake (12). Trk2p,which is 55% identical to Trk1, mediates medium-affinity uptake(17, 18, 38). Low-affinity uptake is mediated by a variety ofother transporters, including NSC1, an activity described electrochemically,but for which no protein has yet been identified (5, 6, 39).

Efflux is thought to be mediated by one of three different transporters.The K⁺ channel, Tok1p, shows similarities to outward rectifyingchannels of higher eukaryotes (4, 23, 24, 51), although evidencesuggests that this protein can also mediate influx under certainconditions (4). Nha1p, a Na⁺-K⁺/H⁺ antiporter, localizes tothe plasma membrane and is involved in regulation of both intracellularion concentration and regulation of the cell cycle (2, 3, 47).Another antiporter, Kha1p, which exchanges K⁺ for H⁺, was initiallythought to localize to the plasma membrane (37), but recentevidence suggests that it is instead localized to internal membranes,perhaps the Golgi apparatus (30). The vacuole is responsiblefor K⁺ storage (36), carried out by proteins whose identitiesare currently unknown.

The regulation of K⁺ influx and efflux activities across theplasma membrane so as to maintain the cytosolic level of K⁺is not well understood. Recently, we found that loss of theguanine-nucleotide-binding protein gene, ARL1, causes a defectin K⁺ influx (33), suggesting a novel means for regulation ofthis process. Loss of ARL1 is accompanied by sensitivity toa number of different toxic cations, including the translationinhibitor, hygromycin B (33, 34). A high-copy suppressor screenfor genes that permit growth of arl1 ${Delta}$ strains in the presenceof hygromycin B identified SAP155 (33).

SAP155 and three homologs—SAP4, SAP185, and SAP190—encodeproteins that interact with Sit4p (28). Sit4p, a PP2A-like phosphatase,was first identified as a regulator of G₁-S transition (46)but also plays a role in K⁺ homeostasis, since overexpressionof SIT4 causes increased K⁺ efflux from cells (31). The Sapproteins may be substrates for Sit4p's phosphatase activitysince they are hyperphosphorylated in a sit4 ${Delta}$ strain (28). Recentwork has demonstrated that the Sap proteins, in conjunctionwith Sit4, function to control Tor-mediated transcription andtranslation (40). Interestingly, only SAP155 overexpressionprotects cells from the deleterious effects of the cell cycleinhibitor, zymocin, from Kluyveromyces lactis (16). Similarly,only SAP155 suppresses the hygromycin B-sensitive phenotypeof the arl1 ${Delta}$ mutant (33).

Based on this information, we hypothesized that Sap155p inhibitsSit4p's ability to promote K⁺ efflux. Through a variety of experiments,we establish the validity of this hypothesis and further demonstratethat a homolog of Sap155p, Sap185p, promotes K⁺ efflux. Finally,we identify Nha1p as a downstream target of Sit4p. Thus, thecombined activities of Sap155p and Sap185p control Sit4p's abilityto regulate K⁺ efflux through Nha1p.

MATERIALS AND METHODS

Top

Materials and Methods

Yeast strains. For strains used in the present study, see Table 1. All haploidstrains were derived from BY4742 from the Deletion Collection(Open Biosystems, Huntsville, AL) (49), and all mutations wereconfirmed by PCR analysis with an oligonucleotide specific tothe gene used to replace the open reading frame (KanMX or HIS3)and one specific to the gene deletion of interest in the reversedirection from the 3' downstream region. Strains were routinelygrown in either rich medium (YPAD) or minimal medium (SD) withappropriate supplements to cover auxotrophies (1, 43). In oneexperiment, dextrose was replaced with galactose for SGal medium.Genotypes of the diploid strains from the deletion collectionwere also confirmed by PCR analysis. Antibiotics (Geneticin[0.2 mg/ml] and hygromycin B [up to 0.1 mg/ml]) were added toYPAD medium after autoclaving.

View this table:
[in this window]
[in a new window]
TABLE 1. Strains used in this study

Plasmids. For plasmids used in the present study, see Table 2. Plasmidswere propagated in Escherichia coli (DH10B or DH5 ${alpha}$ ) grown onLB medium containing ampicillin (100 mg/liter), which was addedafter autoclaving. Plasmids were isolated from E. coli by usingcommercially available kits from either Promega (Madison, WI)or Qiagen (Valencia, CA). Transformation of yeast was done byusing the method of Ito et al. (15), but the 42°C incubationwas extended from 15 min to a maximum of 2 h.

View this table:
[in this window]
[in a new window]
TABLE 2. Plasmids used in this study^a

⁸⁶Rb⁺ studies. ⁸⁶Rb⁺ influx was performed as previously described (33) by themethod of Mulet et al. (32). Efflux was measured by using thefollowing protocol. Cells were grown overnight in SD medium(with appropriate supplements to cover auxotrophies) to an opticaldensity at 600 nm of 0.8, washed twice with distilled H₂O, resuspendedin K⁺ starvation medium (50 mM succinic acid [pH 5.5], 2% glucose,1% casein digest), and incubated for 4 h at 30°C. Cellswere then washed three times with distilled H₂O and resuspendedin reaction buffer (100 mM succinic acid [pH 5.5], 8% glucose)to a concentration of 10 optical density units (OD)/ml. ⁸⁶RbCl₂(2.2 µCi/ml, 0.2 mM RbCl₂ [final concentration]) was addedto each 10 OD of cells. One 100-µl aliquot (correspondingto 1 OD of cells) was taken immediately (zero time) and a secondwas taken after incubation at room temperature for 60 min (theloading/influx period). Cells were then collected by centrifugation,resuspended in room temperature reaction buffer containing 0.2mM cold RbCl₂, and then incubated for up to 3 h. Aliquots wereremoved over time and then added to 10 ml of ice-cold 20 mMMgCl₂. Cells were collected on nitrocellulose and washed extensivelywith ice-cold 20 mM MgCl₂, and the filters counted by scintillationspectroscopy.

Atomic absorption spectroscopy. Total internal K⁺ concentrations were determined by atomic absorptionspectroscopy using a method similar to that described previously(36). Briefly, cells were grown in YPAD medium overnight andthen grown for 12 h in SD medium with the required growth supplements.Cells (15 OD units) were then incubated for 8 h in 5 mM PIPES[piperazine-N,N'-bis(2-ethanesulfonic acid); pH 6.5] containing100 mM KCl and 30 mM glucose. Cells were then washed three timesin 5 mM PIPES with 30 mM glucose. Cell pellets were dissolvedin 500 µl of 6 M HNO₃ and incubated at room temperaturefor 4 h. Samples were then diluted to 5 ml total with distilledH₂O. K⁺ content was determined by using a Buck 200A spectrophotometerwith a K⁺ lamp tuned to 763.5 nm compared to standards.

RESULTS AND DISCUSSION

Top

Results and Discussion

SAP155 and SAP185 modulate hygromycin B sensitivity. We previously demonstrated that an arl1 ${Delta}$ mutant strain was sensitiveto hygromycin B and other toxic cations and that SAP155, butnot SAP4, SAP185, or SAP190, suppressed the hygromycin B-sensitivephenotype (33). We initially assumed that SAP155 was downstreamof ARL1 and expected that an arl1 ${Delta}$ sap155 ${Delta}$ double mutant wouldhave the same phenotype as an (ARL1) sap155 ${Delta}$ mutant. However,the double mutant was actually more sensitive to hygromycinB than either single mutant (Fig. 1), suggesting that SAP155functions in a separate pathway, one that does not contain ARL1.As observed in our previous studies with the arl1 ${Delta}$ mutant, increasedK⁺ in the medium suppressed the hygromycin B-sensitive phenotypeof the sap155 ${Delta}$ and arl1 ${Delta}$ sap155 ${Delta}$ mutants (data not shown).

View larger version (42K):
[in this window]
[in a new window]
FIG. 1. An arl1 ${Delta}$ sap155 ${Delta}$ mutant is more sensitive to hygromycin B than either single mutant. Strains AM150 (ARL1 SAP155), AM151 (arl1 ${Delta}$ SAP155), SP101 (ARL1 sap155 ${Delta}$ ), and DH101 (arl1 ${Delta}$ sap155 ${Delta}$ ) were grown overnight in YPAD medium and then pelleted by centrifugation, resuspended to an optical density at 600 nm of 5 OD/ml, and subjected to 10-fold serial dilutions in water. Cells were spotted onto solid YPAD medium with or without 0.05 mg of hygromycin B/ml by using a replicator tool. Plates were photographed after overnight incubation at 30°C.

To examine the effects of ARL1 and SAP155 on steady-state levelsof intracellular K⁺, cells were prepared for atomic absorptionspectroscopy. Wild-type (ARL1 SAP155) cells contained 83.4 ±3.0 µmol/OD, arl1 ${Delta}$ SAP155 cells contained 39.5 ±1.2 µmol/OD (47.1% of wild type), ARL1 sap155 ${Delta}$ cells contained63.1 ± 2.4 µmol/OD (75.7% of wild type), and thedouble mutant (arl1 ${Delta}$ sap155 ${Delta}$ ) cells contained 34.7 ± 3.0µmol/OD (41.6% of wild type). Thus, loss of either genelowers the steady-state levels of K⁺ in cells, whereas lossof both genes lowers the steady state more. This result, likethe hygromycin B sensitivity phenotypes of these strains, isconsistent with the hypothesis that ARL1 and SAP155 act in separatepathways.

The presence of SAP155 on a high-copy vector conferred resistanceto hygromcyin B in wild-type cells. Surprisingly, SAP185 conferredsensitivity. The presence of either SAP4 or SAP190 on a high-copyvector had no effect compared to empty vector (Fig. 2, top panel).Similar results were observed in the arl1 ${Delta}$ strain, although suchstrains were more sensitive to hygromycin B overall (not shown).Further, loss of SAP155 conferred hygromycin B sensitivity,whereas loss of SAP185 conferred resistance. Loss of eitherSAP4 or SAP190 was without effect (Fig. 2, bottom panel).

View larger version (86K):
[in this window]
[in a new window]
FIG. 2. SAP155 and SAP185 modulate responses of cells to hygromycin B. (Top) Overexpression of the SAP genes. Wild-type cells (strain BY4743) were transformed with either YEp24 (empty vector), CB2643 (SAP155 in YEp24), CB2606 (SAP4 in YEp24), CB2819 (SAP185 in YEp24), or CB2925 (SAP190 in YEp24) and selected for growth on SD medium lacking uracil. Transformants were then grown overnight in selective medium, diluted to an optical density at 600 nm of 5 OD/ml, serially diluted as described above, and spotted onto solid YPAD with or without 0.075 mg of hygromcyin B/ml. Plates were photographed after overnight incubation at 30°C. (Bottom) Deletion of the SAP genes. Homozygous diploid strains lacking both copies of each SAP gene were grown overnight in YPAD and then treated as in the top panel.

SAP155 and SAP185 modulate K⁺efflux. Since intracellular K⁺ levels are controlled by both influxand efflux, we hypothesized that SAP155 encodes a negative modulatorof K⁺ efflux, reasoning that preventing the loss of K⁺ fromcells should protect against taking up too much hygromycin B.Indirect support for this hypothesis came from the genetic experimentshown in Fig. 1, since the loss of SAP155 in an arl1 ${Delta}$ mutantexacerbated the arl1 ${Delta}$ phenotype, suggesting that SAP155 was notinvolved in K⁺ influx.

To test the efflux hypothesis directly, we loaded cells with⁸⁶Rb⁺, a congener of K⁺ that has been used extensively for studiesof K⁺ flux in yeast (29, 31-33, 50) and then monitored the lossof radioactivity over time. As predicted, overexpression ofSAP155 inhibited efflux, whereas overexpression of SAP185 increasedefflux (Fig. 3, top panel). In contrast, a sap155 ${Delta}$ strain exhibitedincreased efflux and a sap185 ${Delta}$ strain exhibited decreased effluxrelative to wild type (Fig. 3, bottom panel). The presence orabsence of SAP4 or SAP190 was without effect (not shown). Theseresults are completely consistent with the hygromycin B sensitivitydata shown in Fig. 2.

View larger version (14K):
[in this window]
[in a new window]
FIG. 3. SAP155 and SAP185 modulate K⁺ efflux. Strains as listed in both panels were grown and prepared for efflux as described in Materials and Methods. After a 60-min influx period, cells were washed and resuspended in buffer containing cold RbCl. Aliquots were removed over time, cells were collected by filtration, and filters were counted by scintillation spectroscopy. The data shown are average of triplicate determinations (± the standard deviation). Each of these experiments was repeated twice with similar results. (Top) Overexpression of the SAP genes. ⁸⁶Rb⁺ efflux from wild-type cells (strain BY4743) transformed with either YEp24 (empty vector), CB2643 (SAP155 in YEp24), or CB2819 (SAP185 in YEp24) was determined. (Bottom) Deletion of the SAP genes. ⁸⁶Rb⁺ efflux from wild-type, sap155 ${Delta}$ , and sap185 ${Delta}$ strains was determined.

We examined the steady-state levels of K⁺ in strains with eitherSAP155 or SAP185 deleted or overexpressing one of these twoSAP genes. Wild-type cells contained 86.1 ± 4.0 µmol/OD,the sap155 ${Delta}$ strain contained 40.4 ± 2.05 µmol/OD(46.9% of wild type), and the sap185 ${Delta}$ strain contained 129.9± 2.0 µmol/OD (151% of wild type). Wild-type cellswith empty vector contained 87.2 ± 8.0 µmol/OD,with the SAP155 vector contained 130.0 ± 5.4 µmol/OD(149% of wild type with empty vector), and with the SAP185 vectorcontained 43.8 ± 4.3 µmol/OD (50% of wild typewith empty vector). Thus, the results from steady-state analysisof intracellular K⁺ levels are consistent with the ⁸⁶Rb⁺ effluxand hygromycin B sensitivity data.

Further, overexpression of SAP155 in a sap185 ${Delta}$ mutant tippedthe balance more toward retention, whereas overexpression ofSAP185 in a sap155 ${Delta}$ mutant tipped the balance more toward efflux(Table 3). In summary, these results suggest that the combinedactivities of Sap155p and Sap185p modulate K⁺ efflux. In contrast,the loss of ARL1 had no effect on efflux (data not shown).

View this table:
[in this window]
[in a new window]
TABLE 3. Sap155p inhibits efflux and Sap185 promotes efflux^a

Previous work by Arndt's laboratory (28) suggested that SAP185and SAP190 were functional homologs of one another, since thesetwo proteins are 42% identical overall, the highest level ofhomology between any two of the Sap proteins; nevertheless,our study clearly distinguishes between the functions of thesetwo: SAP185 encodes a regulator of K⁺ efflux, whereas SAP190does not. This result may suggest that instead of two pairsof functional homologs (SAP185 and SAP190 compared to SAP4 andSAP155), all four Sap proteins have independent, unique functions.

Sap155p and Sap185p require Sit4p for function. The Sap proteins bind to Sit4p and Sit4p requires the Saps forfunction, since a strain missing all four SAP genes has thesame phenotype as a sit4 ${Delta}$ strain (28). The Sap proteins are hyperphosphorylatedin a sit4 ${Delta}$ strain, thus potential substrates for the phosphataseactivity of Sit4p (28). Further, overexpression of SIT4 causesincreased K⁺ efflux and expression of SIT4 is induced by K⁺(31). We therefore investigated the role of SIT4 with respectto the two SAP genes of interest. A sit4 ${Delta}$ strain was transformedwith empty vector or vector containing SAP155, SAP185, or SIT4,and then transformants were loaded with ⁸⁶Rb⁺. Neither overexpressionof SAP155 or SAP185 affects K⁺ efflux in a sit4 ${Delta}$ strain, demonstratingthat both SAP155 and SAP185 require the presence of SIT4 forthe modulation of K⁺ efflux (Fig. 4 and Table 3).

View larger version (13K):
[in this window]
[in a new window]
FIG. 4. SIT4 is required for the function of the SAP genes. ⁸⁶Rb⁺ efflux from the sit4 ${Delta}$ strain transformed with YEp24 (empty vector), CB2643 (SAP155 in YEp24), CB2819 (SAP185 in YEp24), and pSIT4 was determined as in Fig. 3 above. This experiment was repeated twice with similar results.

The ability of Nha1p to promote K⁺ efflux is modulated by Sit4p. Three proteins have been suggested to play roles in efflux ofK⁺ from cells. These are Tok1p, an inward rectifying K⁺ channel(4, 23, 24, 51); Nha1p, a Na⁺-K⁺/H⁺ antiporter (2, 3, 47), andKha1p, a K⁺/H⁺ antiporter (37). However, recent evidence suggeststhis last protein is not localized to the plasma membrane andmay therefore not play roles in efflux from the cytosol acrossthe plasma membrane (30).

To distinguish among these three possibilities, we examinedthe effect of overexpression of each of the three in wild-typeand sit4 ${Delta}$ strains, reasoning that the absence of SIT4 would renderoverexpression of one or more of these proteins without effect.Overexpression of any one of the three increased efflux in wild-typecells (Fig. 5). Overexpression of TOK1 (Fig. 5B) or KHA1 (Fig.5C) in the sit4 ${Delta}$ strain also led to increased efflux. In contrast,overexpression of NHA1 (Fig. 5A) was without effect in the sit4 ${Delta}$ strain, supporting the hypothesis that Sit4p signals to Nha1pto direct K⁺ efflux but not to Kha1p or Tok1p.

View larger version (35K):
[in this window]
[in a new window]
FIG. 5. Overexpression of NHA1, but not TOK1 or KHA1, is without effect in a sit4 ${Delta}$ strain. ⁸⁶Rb⁺ efflux was measured in cells transformed with either empty vector (EV) or vector containing NHA1, TOK1, or KHA1 to determine which gene required the presence of SIT4 for efflux activity. These experiments were repeated once with similar results. (A) Overexpression of NHA1. Wild-type and sit4 ${Delta}$ strains transformed with empty vector (EV) or vector containing NHA1 (pNHA1). Transformants were grown in SD medium lacking uracil and then loaded with ⁸⁶Rb⁺ as described in Materials and Methods. Efflux was measured as in Fig. 3. (B) Overexpression of TOK1. Wild-type and sit4 ${Delta}$ strains were transformed with empty vector or vector containing TOK1 under the control of the GAL1/10 promoter (pYGW1). The experiment was performed as described in panel A, except transformants were grown in SGal medium lacking uracil to induce expression of TOK1. Efflux was then measured as in Fig. 3. (C) Overexpression of KHA1. Wild-type and sit4 ${Delta}$ strains transformed with empty vector (EV) or vector containing KHA1 (pKHA1). Transformants were grown in SD medium lacking uracil and then loaded with ⁸⁶Rb⁺ as described in Materials and Methods. Efflux was measured as described for Fig. 3.

Our results suggest that the competing functions of Sap155pand Sap185p act as a rheostat to control Sit4p-mediated K⁺ effluxfrom cells. This is the first description of the involvementofany of the Sap proteins in control of intracellular K⁺ levels.Of the four Sap proteins, these two are the least homologousto one another, showing only 14% identity overall. The mosthomologous region among the four proteins is the central regionwhich may represent the Sit4p-binding domain, whereas the N-and C-terminal regions are more divergent. Previous resultsdemonstrated that binding of Sap155p and Sap190p to Sit4p wasmutually exclusive (28). Although the same experiment was notperformed with Sap155p and Sap185p in that study, mutually exclusivebinding is a plausible mechanism for the results observed here.Sap155p and Sap185p may therefore control Sit4p's phosphataseactivity, further suggesting that Nha1p is regulated by phosphorylation.If Nha1p is indeed phosphorylated, not only is there regulationat the level of the kinase and phosphatase, but from the presentstudy the activity of the phosphatase, Sit4p, may be regulatedby Sap155p and Sap185p. Future experiments will be geared towardidentifying the specific regions of Sap155p and Sap185p thatinhibit and promote K⁺ efflux, respectively.

Further, our results demonstrate that Sit4p signals to the Na⁺-K⁺/H⁺antiporter, Nha1p, a finding consistent with results from Ariño'slaboratory, which have shown that overexpression of NHA1 suppressesthe synthetic lethal phenotype caused by loss of both SIT4 andHAL3. In addition, the C-terminal domain of Nha1p contains regionsthat are important for regulation of the cell cycle (44, 45).Future experiments will be aimed at determining whether Nha1pis in fact phosphorylated and whether it is directly dephosphorylatedby Sit4p.

ACKNOWLEDGMENTS

We thank the members of the Rosenwald laboratory for commentsabout this study. Judith F. Rubinson (Chemistry Department,Georgetown University) assisted with the atomic absorption spectroscopystudies. Finally, we thank Joaquin Ariño (UniversitatAutòma de Barcelona) for the KHA1 and NHA1 plasmids,Kim Arndt (Wyeth-Ayherst) for the SAP plasmids, and CliffordSlayman (Yale University) for the TOK1 plasmid.

C.M.A.M. and D.H.H. were supported by Zukowski Summer ResearchAwards. C.M.A.M. is supported by a grant to Joseph Neale fromthe Howard Hughes Medical Institute and a scholarship from theClare Booth Luce Foundation. Work in A.G.R.'s laboratory issupported by a CAREER grant (MCB-9875762) from the NationalScience Foundation.

FOOTNOTES

* Corresponding author. Mailing address: Department of Biology, Georgetown University, 406 Reiss Science Center, Box 571229, Washington, DC 20057-1229. Phone: (202) 687-5997. Fax: (202) 687-5662. E-mail: rosenwaa{at}georgetown.edu.

Present address: University of Florida College of Medicine,Gainesville, Florida.

REFERENCES

Top