An Osmotically Induced Cytosolic Ca2+ Transient Activates Calcineurin Signaling to Mediate Ion Homeostasis and Salt Tolerance of Saccharomyces cerevisiae *

Hyperosmotic stress caused by NaCl, LiCl, or sorbitol induces an immediate and short duration (∼1 min) transient cytosolic Ca2+([Ca2+] cyt ) increase (Ca2+-dependent aequorin luminescence) inSaccharomyces cerevisiae cells. The amplitude of the osmotically induced [Ca2+] cyt transient was attenuated by the addition of chelating agents EGTA or BAPTA, cation channel pore blockers, competitive inhibitors of Ca2+ transport, or mutations (cch1Δ ormid1Δ) that reduce Ca2+ influx, indicating that Ca ext 2 + is a source for the transient. An osmotic pretreatment (30 min) administered by inoculating cells into media supplemented with either NaCl (0.4 or 0.5 m) or sorbitol (0.8 or 1.0 m) enhanced the subsequent growth of these cells in media containing 1 m NaCl or 2 msorbitol. Inclusion of EGTA in the osmotic pretreatment media or the cch1Δ mutation reduced cellular capacity for NaCl but not hyperosmotic adaptation. The stress-adaptive effect of hyperosmotic pretreatment was mimicked by exposing cells briefly to 20 mm CaCl2. Thus, NaCl- or sorbitol-induced hyperosmotic shock causes a [Ca2+] cyt transient that is facilitated by Ca2+ influx, which enhances ionic but not osmotic stress adaptation. NaCl-induced ENA1expression was inhibited by EGTA, cch1Δ mutation, and FK506, indicating that the [Ca2+] cyt transient activates calcineurin signaling to mediate ion homeostasis and salt tolerance.

Exposure of cells to high salt causes both ionic and hyperosmotic stresses. In yeast, at least two signal transduction pathways are activated to regulate processes necessary for ion homeostasis and osmotic adjustment. The high osmolarity glycerol (HOG) 1 pathway is primarily responsible for the con-trol of osmotic adaptation while the calcineurin pathway regulates ion homeostasis. The HOG pathway, which includes as a key intermediate the mitogen-activated protein (MAP) kinase Hog1p, is initiated as a consequence of hypertonic stress perception by either of two osmosensing proteins Sln1p or Sho1p (1)(2)(3). Signaling transduced through the HOG cascade activates transcription of GPD1 and of other stress-responsive genes necessary for osmotic adaptation (1, 4 -6). GPD1 encodes glycerol-3-phosphate dehydrogenase, which catalyzes a critical step in the biosynthesis of glycerol, the principal osmolyte of yeast cells.
The Ca 2ϩ /calmodulin-dependent type 2B phosphatase calcineurin is the pivotal intermediate in this signal pathway that mediates cellular Na ϩ , K ϩ , and Ca 2ϩ homeostasis (7,8). Hypersaline stress causes Ca 2ϩ activation of calmodulin, which in turn activates calcineurin (7,8). Signaling through calcineurin regulates the expression of ENA1/PMR2A, which encodes the P-type ATPase that is primarily responsible for Na ϩ efflux across the plasma membrane, and of PMC1 and PMR1, which encode endomembrane-localized Ca 2ϩ -ATPase pumps. Calcineurin dephosphorylates the zinc finger transcription factor Crz1p/Tcn1p/Hal8p, a process that facilitates its mobilization to the nucleus (9 -12). Crz1p/Tcn1p/Hal8p interacts with the calcineurin-dependent response element (CDRE) that is located in the promoters of ENA1, PMC1, PMR1, and other calcineurin-activated genes (9 -12). Hog1p also activates ENA1 expression by rendering the bZIP transcriptional repressor Sko1p and the Ssn6-Tup1 co-repressor complex inactive (13,14). Calcineurin post-transcriptionally activates the Trk1-Trk2 K ϩ transport system to the high affinity state resulting in higher K ϩ /Na ϩ selective uptake and reduced Na ext ϩ influx (8,15). The Hal4p/Hal5p signal pathway also activates the Trk1-Trk2 transport system high affinity transition, although the post-translational process has not been defined nor is it known whether Ca 2ϩ is required for the function of this pathway (16). Calcineurin post-transcriptionally regulates Vcx1p, which together with Pmc1p and Pmr1p function to evacuate Ca 2ϩ from the cytosol (17,18).
Hypotonic shock or a combination of hypotonic shock and cold stress induces a rapid, short duration [Ca 2ϩ ] cyt pulse in yeast cells that is hypothesized to activate the protein kinase C cell integrity pathway (19). Recently, it was determined that hypertonic shock triggers a transient increase in [Ca 2ϩ ] cyt that was attributed to the function of a transient receptor potentiallike channel (Yvc1p), which mediates the release of Ca 2ϩ from the vacuole (20). Although Ca 2ϩ is implicated as a primary effector of Na ϩ and Ca 2ϩ homeostasis necessary for salt tolerance (21), the molecular entities and processes involved in salt stress-dependent Ca 2ϩ activation of calmodulin and calcineurin have yet to be elucidated.
Herein, direct evidence is presented that Ca ext 2ϩ influx facilitates salt adaptation of yeast. Hyperosmotic shock induces a short duration transient increase in [Ca 2ϩ ] cyt through Cch1p-Mid1p that mediates ionic but not hyperosmotic stress tolerance. The osmotically induced [Ca 2ϩ ] cyt transient activates signaling through calcineurin leading to ENA1 induction.

EXPERIMENTAL PROCEDURES
Yeast Strains and Culture Conditions-The strains were derived from Saccharomyces cerevisiae W303-1A (MAT a: his3 leu2 trp1 ade2 ura3) (30). The cch1::kanMX4 deletion-disruption mutation was constructed using the PCR-based gene-targeting system (31). The forward primer 5Ј-TGA TGA AAG GCT ACG CAG GCG TAG CTT CAG TAG TTA TAG CCG ATC atc gat gaa ttc gag ctc g-3Ј (S1 primer for pFA6a-MX in lowercase) and the reverse primer 5Ј-TCT CTG ATC CAC AGT CCG TAT AGG TTG AAT TAT CAT CAG AGG AGC atc gat gaa ttc gag ctc g-3Ј (S2 primer for pFA6a-MX in lowercase) were used to amplify kanMX4 from plasmid pFA6a-kanMX4 (provided by Dr. P. Philippsen, Ref. 31). Homologous recombination of the PCR product introduced a kanMX4 disruption in the CCH1 (YGR217W) open reading frame beginning at ϩ719 bp from the translation start site to Ϫ941 bp from the translation stop site. The disrupted allele was detected using primers corresponding to sequence beginning at position Ϫ260 bp (5Ј-CGG GAA AAT GTA ATT TGG CAT GTC A-3Ј) from the translation start site to ϩ262 bp (5Ј-TTT TTG GTC TTG TGA AGA CTA CG-3Ј) from the translation stop site.
The mid1::kanMX4 mutation was constructed using the forward primer 5Ј-GGC AAG CAC TAT TCG TGG TTT ACT GCC TAT TTA CCA CTT CTA TTC atc gat gaa ttc gag ctc g-3Ј and reverse primer 5Ј-CTA CGT ATC GTC CAA TGG ATG AAT TAC CAT CAA GGA TGA GTT ACC atc gat gaa ttc gag ctc g-3Ј to amplify kanMX4. Homologous recombination of the PCR product introduced a disruption in the MID1 (YNL291C) open reading frame position ϩ55 bp from the translation start site to Ϫ45 bp from the translation stop site. The disrupted allele was detected using primers corresponding to sequence beginning at position Ϫ391 bp (5Ј-GGG CCT GTT CCT ACA CCT TTT ATA A-3Ј) from the translation start site to ϩ125 bp (5Ј-GCG AAC TTA TTC AGA TCT CTT CAA CCA-3Ј) from the translation stop site. The cch1⌬ mid1⌬ double mutation was generated by using the PCR-based gene-targeting system and HIS3MX6 (31) in place of the kanMX4 for the mid1⌬ mutation and was introduced into cch1::kanMX4 mutant.
Aequorin Luminescence Measurements and [Ca 2ϩ ] cyt Quantification-Aequorin luminescence was determined with a luminometer (TD-20/20 Turner Designs, Sunnyvale, CA) following procedures used to quantify [Ca 2ϩ ] cyt in yeast and tobacco cells (19,33). Cells were transformed with the 2-plasmid pEVP11/AEQ (provided by Dr. P. Masson, 19) containing the APOAEQUORIN gene. Leucine autotrophic cells at stationary phase were harvested by centrifugation and then resus-pended in YPD medium to an OD 600 of ϳ0.1-0.2. Cells were grown to an OD 600 of ϳ0.4 -0.6. The cells were concentrated about 5-fold by centrifugation and 1 l of 590 M coelenterazine (Biotum Inc., Fremont, CA) in absolute methanol was added to 100 l of culture. After incubation for 30 min on a rotating drum in the dark, a cellular luminescence baseline was determined by 1 min of recordings at 5/s intervals.
Hypertonic shock was administered to yeast cells by diluting the suspension 1:1 with YPD medium containing NaCl, LiCl, or sorbitol at 2ϫ final concentration. All inhibitory or chelating compounds were diluted in 50 mM MES (pH 6.5) and added to the cultures as indicated. EGTA and BAPTA solutions (0.5 M) were made in 50 mM MES buffer (pH 6.5). NiCl 2 , MgCl 2 , or GdCl 3 was dissolved in water to a concentration of 0.3 M and diluted with MES buffer (50 mM final concentration, pH 6.5) to 10-to 20-fold the concentration in the treatment solution. These chemical agents or an equal volume of 50 mM MES were added to cells immediately before the acquisition of the basal luminescence reading and in the osmotic pretreatment media. Luminescence from aequorin that remained in cells at the end of an experiment was determined after treating cells with Triton X-100 (5-10%) and CaCl 2 (2-5 M). These data were used to calculate the [Ca 2ϩ ] cyt using Equation 1, where L is the luminescence intensity at any time point and L max is the integrated luminescence intensity (34).
Measurement of 45 Ca 2ϩ Accumulation-Procedures for determining Ca 2ϩ uptake into cells were as described, with modifications (17,28). Cells in the exponential growth phase (YPD medium) were inoculated into fresh medium (OD 600 ϳ0.1) that contained 1 Ci/ml of 45 Ca 2ϩ ( 45 CaCl 2 , ICN Biomedicals Inc., Costa Mesa, CA). Cells were grown for 5 h and harvested for determination of intracellular radioactivity.
Osmotic Pretreatment of Cells-Cells in the log-growth phase (OD 600 of ϳ0.4 -0.6) were subjected to osmotic pretreatment (30 min) by diluting the suspension 1:1 with YPD medium containing NaCl or sorbitol at 2ϫ final concentration. Cells were collected by centrifugation (Eppendorf Centrifuge 5810) at 3000 rpm for 3 min. Cells were resuspended in growth medium to an OD 600 of ϳ0.2-0.3, and growth was monitored (OD 600 ).
ENA1-lacZ ␤-Galactosidase Assay-Wild type and cch1⌬ cells were transformed with pKC201 containing an ENA1-lacZ fusion (provided by Dr. K. Cunningham,Ref. 9). Cells from three independent transformation events were grown to late log phase in S.D. medium. Cells were collected, resuspended in YPD medium to an OD 600 of ϳ0.1, and grown to log phase (OD 600 of ϳ0.4 -0.6). These cells were exposed to hypertonic shock by diluting the culture 1:1 with medium containing NaCl, or 1:1 with CaCl 2 at 2ϫ final concentration. A stock solution of 2 mg/ml FK506 (provided by Fujisawa Healthcare Inc., Deerfield, IL) that was dissolved in 90% ethanol and 10% Tween-20 was added to medium as indicated. Cells were collected after 4 h, frozen in liquid nitrogen, and processed for determination of ␤-galactosidase activity (35). The same period of 4 h was used previously to determine calcineurin-dependent induction of FKS2, PMC1, and ENA1 expression, after stimulation by NaCl, Ca 2ϩ , or mating pheromone (9).

Hyperosmotic Stress Induces a Transient Increase in [Ca 2ϩ ] cyt That Is Derived from an Extracellular Pool-NaCl
shock initiates an immediate and transient elevation in [Ca 2ϩ ] cyt of about 1 min in duration (Fig. 1A). Because similar Ca 2ϩ transient profiles were obtained after treatment of cells with equiosmolar concentrations of NaCl, sorbitol, LiCl (Fig. 1, A-C), or Na 2 SO 4 or KCl (not shown), the increase in [Ca 2ϩ ] cyt is caused, in a concentration-dependent manner, by hypertonic shock and not ionic stress. A new and higher [Ca 2ϩ ] cyt steady state is established after the initial transient (Fig. 1B). This higher [Ca 2ϩ ] cyt steady state may be due to continued influx from the extracellular pool, release from an internal cellular store, or limited vacuolar sequestration (36).
The hyperosmotic stress-induced increase in [Ca 2ϩ ] cyt was attenuated by the inclusion of 2.5 or 5 mM NiCl 2 , or by 5 mM MgCl 2 in the medium (Fig. 1D). Ni 2ϩ effectively inhibits L-type Ca 2ϩ channel activity of guinea pig ventricular myocytes (37), and Mg 2ϩ competes with Ca 2ϩ for channel conductance in tobacco cells (33). Reduction in the amplitude of the transient rise in [Ca 2ϩ ] cyt by Ni 2ϩ and Mg 2ϩ implicates the extracellular pool as the cation source.
The transient elevation in [Ca 2ϩ ] cyt that is triggered by 1 M NaCl was inhibited also by inclusion in the medium of the membrane impermeable, divalent cation chelating agent EGTA (Fig. 1E). Similar attenuation of the transient increase in [Ca 2ϩ ] cyt occurred if BAPTA, a chelating agent with 10 5 -fold higher affinity for Ca 2ϩ than for Mg 2ϩ , was substituted for EGTA in the medium (Fig. 1E). These results also confirm that the hyperosmotically induced transient rise in [Ca 2ϩ ] cyt is dependent on influx across the plasma membrane of cells that are inoculated into YPD medium, which contains about ϳ75 M Ca 2ϩ (38).
Gd 3ϩ (GdCl 3 ) reduced the amplitude of the hypertonic shockinduced [Ca 2ϩ ] cyt transient in cells, albeit incompletely (Fig.   1F). Because incubation of cells with GdCl 3 was initiated 1 min prior to addition of NaCl, it is unlikely that Gd 3ϩ permeated into the cell and inhibited endomembrane localized Ca 2ϩ -permeable channel activity to cause the attenuation of the transient elevation in [Ca 2ϩ ] cyt . Gd 3ϩ is an inhibitor of stretchactivated Ca 2ϩ channels of yeast, plant, and animal cells (39,40), and effectively reduces Ca cyt 2ϩ transients in yeast and tobacco cells (19,33). These results implicate the involvement of a Gd 3ϩ -sensitive, plasma membrane Ca ϩ channel protein, perhaps Mid1p, in the Ca ext 2ϩ influx that results in increased [Ca 2ϩ ] cyt (29). cch1⌬, mid1⌬, or cch1⌬ mid1⌬ disruption/deletion mutations were generated in W303-1A cells. These cells were then transformed to express apoaequorin from pEVP11/AEQ (ADH::AEQ). Net 45 Ca 2ϩ accumulation in W303-1A cch1⌬ cells was reduced to about 60% of the level that accumulated in wild type cells (not shown). cch1⌬ caused a similar reduction in intracellular Ca 2ϩ accumulation in another yeast strain (28). The NaCl-induced [Ca 2ϩ ] cyt transient in cch1⌬ or mid1⌬ cells was substantially, and to a similar degree, lower than in wild type cells (Fig. 1, G and H). The cch1⌬ mid1⌬ double mutation did not cause an additive effect. The NaCl-induced transient elevation in [Ca 2ϩ ] cyt in cch1⌬, mid1⌬, and cch1⌬ mid1⌬ cells is comparable to that in wild type cells after treatment with 5 mM GdCl 3 . It can be concluded that the hyperosmotically induced [Ca 2ϩ ] cyt transient is mediated predominantly by influx through Cch1p and Mid1p that are genetically functioning as a single Ca 2ϩ uptake system. However, there may be contribution to the [Ca 2ϩ ] cyt transient through an alternative uptake system or from internal stores (20) because the elevation in [Ca 2ϩ ] cyt is not completely abrogated by pharmacological agents or by null mutations in the plasma membrane Ca 2ϩ channel.
Ca ext 2ϩ Influx through Cch1-Mid1p Is Required for Ionic but Not Hyperosmotic Stress Adaptation-cch1⌬ and mid1⌬ cells are Li ϩ sensitive (28). Furthermore, cch1⌬ and mid1⌬ cells are Na ϩ but not hyperosmotic sensitive (mid1⌬ data not shown), and Na ϩ and Li ϩ sensitivities of these cells are exacerbated by EGTA or attenuated by Ca 2ϩ supplementation to the media (Fig. 2). Because cells exhibit Ca 2ϩ -dependent growth enhancement on media supplemented with Na ϩ or Li ϩ but not with sorbitol, it is concluded that the divalent cation facilitates ionic adaptation (Fig. 2). Neither Na ϩ nor Li ϩ sensitivity, nor the modulation of sensitivity to these ions by EGTA or Ca ext 2ϩ , is additive in cch1⌬ mid1⌬ cells compared with cch1⌬ or to mid1⌬ cells (not shown). These results indicate that the hyperosmotic shock-induced transient increase in [Ca 2ϩ ] cyt that is mediated in part by Cch1p and Mid1p is necessary for ionic stress adaptation and that the proteins function together as an individual influx system.

Osmotic Pretreatment Facilitates Ionic Stress Adaptation That Is Ca ext
2ϩ Transient-dependent-Cell growth in ionic or osmotic stress media was enhanced by a 30-min hyperosmotic pretreatment (Fig. 3). Interestingly, pretreatment with 20 mM Ca 2ϩ enhanced cell growth only in medium that was supplemented with NaCl (Fig. 3, B-D). Cells exposed to high [Ca 2ϩ ] ext exhibit an immediate and transient increase in [Ca 2ϩ ] cyt (not shown; Refs. 20 and 36). Hypertonic shock-induced NaCl adaptation was enhanced by Ca ext 2ϩ or attenuated by EGTA (Fig. 3C). EGTA did not affect salt adaptation when the chelating agent was added to hyperosmotic pretreatment medium 25 min (30 min total) after cell inoculation, indicating that sufficient Ca 2ϩ influx had occurred by this time to mediate ionic stress adaptation (not shown). Osmotic pretreatment was substantially less effective at enhancing growth of cch1⌬ cells compared with wild type cells in NaCl-containing medium (Fig. 3C). The re-

FIG. 1. NaCl-induced [Ca 2؉ ] cyt transient in yeast cells is caused by hypertonic shock that is derived from Ca ext 2؉ influx.
Yeast cells expressing reconstituted aequorin (see "Experimental Procedures") were diluted with an equal volume of fresh medium supplemented with NaCl, LiCl, or sorbitol (at 2ϫ final concentration) to initiate the osmotic stress treatment (arrow). A, NaCl; B, NaCl or sorbitol; or C, NaCl or LiCl were added and [Ca 2ϩ ] cyt were determined from aequorin luminescence measurements. The NaCl-induced [Ca 2ϩ ] cyt transient is attenuated by D, divalent cations Ni 2ϩ (NiCl 2 ) or Mg 2ϩ (MgCl 2 ); E, by membrane impermeable Ca 2ϩ chelating agents EGTA or BAPTA; or F, by the Ca 2ϩ channel blocker Gd 3ϩ (GdCl 3 ). The NaCl-induced [Ca 2ϩ ] cyt transient was detected in cells of wild type (W303-1A) (A-F); wild type (wt) or cch1⌬ (G); wt, cchl⌬, mid1⌬, or cch1⌬ mid1⌬ (H). Data illustrated are from one representative experiment (minimum of three independent experiments). duced salt-adaptive capacity of cch1⌬ cells was similar to that determined when wild type cells were exposed to EGTA. In contrast, growth of cch1⌬ cells was equivalent to wild type after osmotic pretreatment and inoculation into medium with 2 M sorbitol (Fig. 3D). Together, these findings link the hyperosmotically generated [Ca 2ϩ ] cyt transient to ionic stress adaptation and implicate Cch1p and Mid1p as the major Ca ext 2ϩ influx determinants.
The Osmotically Induced Ca cyt 2ϩ Transient Activates ENA1 Expression through Calcineurin Signaling-Monitoring of ENA1 expression using the ENA1-lacZ promoter-reporter fusion system (9) confirmed that ␤-galactosidase activity is constitutively lower in cch1⌬ compared with wild type cells (Fig. 4 and Ref. 28). ENA1 expression is induced by inclusion of NaCl or CaCl 2 into the medium and the induction is greater in wild type than in cch1⌬ cells (Fig. 4). EGTA abrogates the difference in NaCl-induced ENA1 expression between wild type and cch1⌬ cells, which indicates that Ca ext 2ϩ influx through Cch1p is required for transcriptional activation of this gene (Fig. 4). The calcineurin-specific inhibitor FK506 (9, 10) reduced ENA1-lacZ expression to near basal levels in cells that were in medium with CaCl 2 but not with NaCl. The difference may reflect the contribution of the HOG pathway to ENA1 induction that is elicited by NaCl treatment. Almost total abrogation of NaClinduced ENA1-lacZ expression by a combination of EGTA and FK506 is evidence that an alternative Ca 2ϩ -dependent pathway may regulate the expression of the gene during a salt stress episode. However, ENA1 expression is potentiated substantially by calcineurin. Calcineurin-dependent nuclear translocation of the Crz1p/Tcn1p/Ha18p transcription factor, which activates ENA1 expression, occurred within 10 min in response to 200 mM Ca 2ϩ or 0.8 M NaCl (12). Nuclear translocation did not occur if the Ca 2ϩ -chelating agent EGTA is included in the medium. Furthermore, Ca 2ϩ -induced CDRE-lacZ reporter expression is attenuated by FK520, another inhibitor of calcineurin (12). Together, these findings establish that the hyperosmotically induced [Ca 2ϩ ] cyt transient is dependent on influx through Cch1p, presumably the Cch1p-Mid1p system, and activates calcineurin, which induces ENA1 expression.

Transient Increase in Ca cyt
2ϩ Mediates Ionic Adaptation-Hyperosmotic shock induces [Ca 2ϩ ] cyt transients in organisms as diverse as Chara (41), Dunaliella salina (42), Anabaena (43), and Arabidopsis thaliana (44) (reviewed in Ref. 45). In plants, salt stress increases [Ca 2ϩ ] cyt that is presumed to facilitate adaptation (46,47). Furthermore, plants or cultured cells that are exposed to sublethal levels of salt acquire greater stress adaptive capacity (48,49), and tomato seedlings subjected to hyperosmotic pretreatment exhibit enhanced salt tolerance (50). The results of this study provide direct evidence that hyperosmotic stress, generated by salt or a non-ionic osmotic solute, induces a transient [Ca 2ϩ ] cyt increase in yeast cells, and this [Ca 2ϩ ] cyt transient mediates ionic but not osmotic stress adaptation. Because most salts are toxic at concentrations that are also osmotically stressful, it is possible that yeast cells detect and respond to salt stress by an osmosensing and not an ion-specific sensing mechanism. Perhaps the Sln1p or Sho1p osmosensor perceives salt stress to activate the HOG pathway for osmotic adjustment and Ca 2ϩ -dependent ion homeostasis pathways, similar to the calcineurin signal cascade (1).

Hyperosmotic Stress-Induced [Ca 2ϩ ] cyt Is Mediated by Ca ext 2ϩ
Influx through Mid1p-Cch1p-Results presented here indicate that hypertonic shock induces Ca ext 2ϩ influx through Cch1p-Mid1p. Electrophysiological studies of yeast plasma membrane patches indicate that the activity and selectivity of mechanosensitive ion channels may function in osmotic sensing and turgor regulation (40). It is not known if the activity of the L-type channel-like protein Cchlp is modulated by mechanical stress but Ca 2ϩ permeability of Mid1p is stretch-activated (29). Ni 2ϩ , an inhibitor of L-type Ca 2ϩ channel activity (37), or Gd 3ϩ , an inhibitor of stretch-activated Ca 2ϩ channel activity (39), substantially diminishes hyperosmotically induced Ca ext 2ϩ influx into rat osteoblast-like cells (51).
Hypertonic shock also induces the release of Ca 2ϩ from internal stores through the vacuolar membrane localized TRP channel-like protein Yvc1p and this contributes to elevated [Ca 2ϩ ] cyt (20). Release of Ca vac 2ϩ through Yvc1p requires activation by Ca 2ϩ on the cytosolic side of the vacuolar membrane (52). Perhaps Ca 2ϩ influx through Cch1p-Mid1p, as indicated by the results presented herein, causes the activation of Yvc1p.
Our data indicate that influx through Cch1p-Mid1p mediates a hypertonic shock-induced transient increase [Ca 2ϩ ] cyt but do not exclude a contribution from an internal source. However, it is not possible to resolve why yvc1⌬ cells that have a functional Cch1p-Mid1p transport system are unable to generate a [Ca 2ϩ ] cyt transient in response to hypertonic shock (20). Interestingly, yvc1⌬ cells have no apparent growth defects, and overexpression of YVC1 causes Ca 2ϩ but not Na ϩ sensitivity (20). Herein, we show that Ca 2ϩ influx through Cch1p-Mid1p facilitates ionic adaptation (Fig. 3).
Activation of Calcineurin by a [Ca 2ϩ ] cyt Transient Facilitates Ion Homeostasis-The [Ca 2ϩ ] cyt transient signature is integral to signal diversity that is transduced through interactions with Ca 2ϩ -binding proteins resulting in outputs that control specific physiological processes (45,(53)(54)(55)(56). Because Ca 2ϩ is relatively immobile in the cytosol and quickly becomes associated with Ca 2ϩ -binding proteins, the degree of Ca 2ϩ saturation within the cell even after a substantial Ca 2ϩ influx is most pronounced in microdomains that are localized near the pore of the channel through which cation influx occurs (57). Recent evidence indicates that Ca 2ϩ -binding proteins (calmodulin, Ca 2ϩ -dependent protein kinases (CDPKs), phosphatases, heterotrimeric G-proteins, etc.) and their interacting proteins are associated physically with the cytosolic-localized domains of Ca 2ϩ channels (55, 58 -60). Some of these proteins are implicated in channel autoregulation, and others function as pivotal components of signal transduction. Thus, both Ca 2ϩ signal differentiation and propagation may result from the properties of the channel that control its gating and the Ca 2ϩ -binding proteins that are associated with the channel cytosolic domain(s), and are activated by the transient.
The results presented here provide direct evidence that hyperosmotic stress induces a transient increase in [Ca 2ϩ ] cyt that activates ENA1-lacZ expression through calcineurin signaling. Calcineurin regulates the Crz1p/Tcn1p/Hal8p transcription factor to activate ENA1 expression. The nuclear localization of this transcription factor is maintained for an extended period after a stimulus-induced transient elevation in [Ca 2ϩ ] cyt and perhaps this is the reason why calcineurin activation of gene expression is sustained for some period after the transient has dissipated (12). Furthermore, the new [Ca 2ϩ ] cyt steady state that is established may also continue signaling through calcineurin and Crz1p/Tcn1p/Hal8p that results in ENA1 transcription. NFAT activation by calcineurin is potentiated by a FIG. 4. Ca 2؉ influx through Cch1p activates calcineurin, which induces ENA1-lacZ expression. Wild type (W303-1A, gray bars) and cch1⌬ (white bars) cells were transformed with the plasmid pKC201 to express ENA1-lacZ (9). Cells were treated with 1 M NaCl or 20 mM CaCl 2 in media without or with EGTA (10 mM) and/or FK506 (0.2 g/ml). Values are the average for three determinations of ␤-galactosidase activity in cells from three independent colonies.
FIG. 5. A model depicting how hypertonic stress induces Ca 2؉ influx through Cch1p-Mid1p to activate calcineurin signaling that mediates ion stress tolerance. Hyperosmotic shock imposes a substantial mechanical stress/strain on the plasma membrane that potentiates Ca ext 2ϩ influx through Cch1p-Mid1p (28) and subsequent Ca 2ϩ release from the vacuole through Yvc1p (20). The [Ca 2ϩ ] cyt increase in the Cch1p-Mid1p microdomain activates calmodulin (Cmd1p) that is tethered either to the channel or its associated proteins. Cmdlp activates calcineurin that dephosphorylates Crz1p/Tcn1p/Hal8 resulting in its translocation to the nucleus where it activates ENA1 expression that facilitates ion homeostasis and salt tolerance. Ca 2ϩ -activated Cmd1p also post-transcriptionally upregulates Ena1p activity (71), and calcineurin also mediates the transition of Trk1p-Trk2p system to high affinity for K ϩ . The vacuolar membrane Ca 2ϩ /H ϩ exchanger Vcx1p is a major determinant that reduces [Ca 2ϩ ] cyt to the steady-state level after influx from the extracellular pool or release from internal stores (20,36). low but sustained level of [Ca 2ϩ ] cyt (reviewed in Refs. 61 and 62).
A model that describes how hyperosmotic stress induces a transient increase in [Ca 2ϩ ] cyt that mediates ion homeostasis and salt tolerance in yeast is illustrated in Fig. 5. This yeast paradigm may be a reasonable representation of the entities and the events that function in plants to initiate Ca 2ϩ -dependent signaling involved in salt adaptation (63,64). Salt induces an increase in [Ca 2ϩ ] cyt (44), and Ca ext 2ϩ enhances salt tolerance of plants (46). Current understanding indicates that in plants, as in yeast, separate pathways are involved in osmotic adjustment and ion homeostasis (65,66). Indeed, a putative twocomponent osmosensor that may be orthologous to Sln1p has been identified in Arabidopsis, together with entities that may comprise a MAP kinase cascade responsible for osmotic signaling (67)(68)(69). The Ca 2ϩ -dependent SOS (salt oversensitive) pathway of Arabidopsis may be the functional equivalent of the yeast calcineurin pathway, at least for ion homeostasis that is necessary to mediate salt tolerance (65,66). It is reasonable to expect that activation of the SOS pathway might be initiated by a hyperosmotically induced [Ca 2ϩ ] cyt transient that is dependent on influx through a plasma membrane localized transport system (70).