The budding yeast Saccharomyces cerevisiae generates calcium signals during the response to mating pheromones that promote survival of unmated cells. A Ca2+ channel composed of Cch1p and Mid1p was previously shown to be necessary for the production of these calcium signals. However, we find that the Cch1p-Mid1p high-affinity Ca2+ influx system (HACS) contributes very little to signaling or survival after treatment with α-factor in rich media. HACS activity was much greater after calcineurin inactivation or inhibition, suggesting the Cch1p-Mid1p Ca2+ channel is subject to direct or indirect regulation by calcineurin. Instead a distinct low-affinity Ca2+ influx system (LACS) was stimulated by pheromone signaling in rich medium. LACS activity was insensitive to calcineurin activity, independent of Cch1p and Mid1p, and sufficient to elevate cytosolic free Ca2+ concentrations ([Ca2+]c) in spite of its 16-fold lower affinity for Ca2+. Overexpression of Ste12p or constitutive activation of this transcription factor in dig1 dig2 double mutants had no effect on LACS activity but stimulated HACS activity when calcineurin was also inactivated. Ste12p activation had no effect on Cch1p or Mid1p abundance, suggesting the involvement of another target of Ste12p in HACS stimulation. LACS activation required treatment with mating pheromone even in dig1 dig2 double mutants and also required FAR1, SPA2, and BNI1, which are necessary for proper cell cycle arrest and polarized morphogenesis. These results show that distinct branches of the pheromone-signaling pathway independently regulate HACS and LACS activities, either of which can promote survival during long-term responses.
INTRACELLULAR calcium signals can be defined as transient increases in cytosolic free calcium concentrations ([Ca2+]c) that are potentially translated into cellular responses. All eukaryotic cells employ calcium signaling to regulate a wide variety of functions, including gene expression, exocytosis, cytoskeletal rearrangement, and cell physiology. For example, the process of fertilization in mammalian cells is regulated by several different calcium signals. One such signal is initiated by the binding of sperm cell receptors to glycoproteins on the surface of eggs, which triggers a calcium signal that is necessary and sufficient for initiating the acrosome reaction (Darszonet al. 1999; Wassarman 1999). Binding and fusion of sperm also induce calcium signals within the egg that are important for release from cell cycle arrest, exocytosis of cortical granules, and metabolic activation (Swann and Parrington 1999). The Ca2+ channels, their regulators, and the downstream signaling factors involved in these complex reactions are not fully understood.
The yeast Saccharomyces cerevisiae also undergoes a mating cycle where two haploid cells of opposite mating type (a or α) fuse to form a diploid zygote. This process is initiated by the binding of secreted peptide pheromones (a-factor or α-factor) to a seven-transmembrane receptor linked to a heterotrimeric G-protein and subsequent activation of the pheromone-signaling cascade (Sprague and Thorner 1992; Posaset al. 1998). Approximately 45 min after initiation of pheromone signaling, the rate of Ca2+ influx increases and [Ca2+]c becomes elevated, leading to activation of calcium signaling pathways (Ohsumi and Anraku 1985; Iidaet al. 1990). The precise role of calcium signaling during the response to mating pheromones is not known. Depletion of extracellular Ca2+ or mutations that inactivate the Ca2+ influx proteins (Mid1p and Cch1p), calmodulin (Cmd1p), calmodulin-dependent protein kinases (Cmk1p and Cmk2p), or the calmodulin-dependent protein phosphatase calcineurin (Cna1p-Cnb1p and Cna2p-Cnb1p) have little or no effect on mating efficiency. However, influx of extracellular Ca2+ and activity of all the calcium signaling factors are essential for maximum survival during long-term treatment with pheromone in the absence of a mating partner (Iida et al. 1990, 1994; Cyertet al. 1991; Cyert and Thorner 1992; Moseret al. 1996; Fischeret al. 1997; Paidhungat and Garrett 1997; Witheeet al. 1997). How the pheromone-signaling pathway promotes death and how the Ca2+ signaling pathway prevents it have not yet been determined.
The mechanism by which pheromone signaling stimulates Ca2+ influx in yeast is also not fully understood. Downstream of the cell type-specific pheromones and receptors, the pheromone-signaling cascade in both a and α cells is identical. The Gβγ subunits of the heterotrimeric G-protein stimulate a Rho-type small GTPase, a p21-activated protein kinase (PAK), and a mitogen-activated protein (MAP) kinase signaling module (Sprague and Thorner 1992; Posaset al. 1998). Activation of the MAP kinase Fus3p induces the expression of many genes by directly phosphorylating and inactivating the related proteins Dig1p and Dig2p, which bind and inhibit the transcription factor Ste12p in nonsignaling conditions (Cooket al. 1996; Tedfordet al. 1997). Fus3p also phosphorylates and activates Far1p, which then binds and inhibits the G1-specific cyclin-dependent protein kinase Cdc28p, resulting in cell cycle arrest (Chang and Herskowitz 1990; Elionet al. 1993; Peteret al. 1993; Tyers and Futcher 1993; Cherkasovaet al. 1999). Far1p together with Spa2p, Bni1p, and other factors promotes polarized cell growth (shmooing) in the direction of the pheromone gradient (Valtzet al. 1995; Blondelet al. 1999; Milleret al. 1999; Shimadaet al. 2000). Ca2+ influx and calcium signaling factors were not required for induction of mating-specific genes, cell cycle arrest, or the initial stages of polarized morphogenesis. Whether any of these processes specifically trigger Ca2+ influx is the subject of this study.
Mutants lacking both Dig1p and Dig2p exhibit constitutive Ste12p activity and high expression of mating-specific genes even in the absence of mating pheromone, cell cycle arrest, or shmooing (Cooket al. 1996; Tedfordet al. 1997). Here we show that activation of the Ste12p transcription factor is sufficient to stimulate the activity of the Cch1p-Mid1p high-affinity Ca2+ influx system (HACS) but has no obvious effect on the cellular levels of these proteins. However, HACS is seldom utilized due to regulation by calcineurin and the stimulation of a distinct low-affinity Ca2+ influx system (LACS) by a separate branch of the pheromone-signaling cascade, which leads to polarized morphogenesis. The data suggest two Ca2+ influx pathways can be independently and differentially regulated during the mating response of yeast.
MATERIALS AND METHODS
Media, yeast strains, and plasmids: Synthetic complete (SC), synthetic minimal (SD), and complex (YPD) media were prepared and supplemented with 2% glucose as described previously (Shermanet al. 1986), using reagents from Difco (Detroit) and Sigma Chemical Company (St. Louis). Synthetic media lacking calcium were prepared similarly using reagents obtained from BIO 101 (Vista, CA). Where indicated YPD was buffered to pH 5.5 with 5 mm succinic acid and supplemented with CaCl2. The synthetic mating pheromone α-factor was obtained from Star Biochemicals (Torrance, CA). FK506 was provided by Fujisawa USA (Tokyo, Japan).
Table 1 lists all yeast strains employed in this study, which were derived from either strains BY4741 (MATa his3-1 leu2-2 met15-0 ura3-0) or from W303-1a (MATa ade2-1 can1-100 his3-1 leu2-3,112 trp1-1 ura3-1; Walliset al. 1989), using standard methods of transformation and/or crossing. The ste12::URA3 disruption was created by transformation of strain MT1153 with a 3-kbp ClaI fragment of pNC163 (Companyet al. 1988) to create strain EMY149. The sst1::URA3 disruption was created by transformation of W303-1A with the 5.1-kbp fragment of pJGsst1 (Elionet al. 1993) produced by digestion with EcoRI plus SalI. Other sst1 strains were derived by isogenic crosses with K410. The far1::URA3 disruption was constructed by transformation of W303-1A with pFC13 (Chang and Herskowitz 1990) after digestion with NotI. All knockouts were confirmed by examining phenotype and/or molecular genotype.
The plasmid pKC147 [2μ URA3 PMA1-aequorin] was constructed by subcloning a 1.1-kbp fragment of the PMA1 promoter region (from nucleotides −936 to +171 relative to codon 1) plus a 0.7-kbp BclI-KpnI fragment of pRSV-AQ (Tanahashiet al. 1990) into the HindIII plus SalI and BamHI plus KpnI sites of YEp356R (Myerset al. 1986), respectively. In some aequorin experiments, plasmid pEVP11/AEQ89 [2μ LEU2 ADH-aequorin] (Batizaet al. 1996) was utilized. The 4X-CDRE-LacZ reporter construct pAMS366 was described previously (Stathopoulos and Cyert 1997). Plasmid pNC252 [2μ URA3 GAL-STE12] was provided by B. Errede.
Methylene blue viability assays: Cells grown to logarithmic phase in YPD or SD-100 media at 30° were diluted to a concentration of 106 cells/ml and exposed to 20 μm α-factor (Star Biochemicals) in the appropriate medium. At each time point indicated an aliquot of cells was harvested by centrifugation and resuspended in fresh media containing 100 μg/ml methylene blue (Sigma). The number of methylene blue-positive and -negative cells was determined immediately by bright field microscopy. A minimum of 200 cells was counted for each strain. The number of viable cells at time 0, for each respective strain, was set to 100%.
45Ca2+ accumulation assays: Accumulation of 45Ca2+ into yeast cells growing in various culture media was measured as described previously (Cunningham and Fink 1996). In a standard experiment, log-phase cells were harvested by centrifugation, resuspended in fresh medium, treated with 20 μm α-factor and/or 2.0 μg/ml FK506, labeled with tracer amounts of 45CaCl2 (Amersham Life Sciences, Arlington Heights, IL) for 4 hr at 30°, collected by filtration onto GFF filters (Whatman), washed three times with 5 ml ice-cold buffer A (5 mm Na-HEPES at pH 6.5, 10 mm CaCl2), dried, and processed for liquid scintillation counting. The specific activity of the culture medium was determined in each experiment and used to convert counts per minute into nanomoles of Ca2+. Cell number was determined by measurements of optical density at 600 nm.
β-Galactosidase assays: Yeast strains carrying reporter genes were grown at 30° in SC minus uracil medium to log phase, harvested, and resuspended in fresh YPD (pH 5.5) supplemented with 40 μm α-factor, 100 mm CaCl2, or 2 μg/ml FK506; and shaken for 4 hr at 30° in 24-well flat-bottomed dishes. β-Galactosidase assays were performed at room temperature using chloroform/SDS-permeabilized cells and colorimetric substrate as described previously (Guarente and Mason 1983).
Aequorin luminescence assays: Cells were grown to log phase in SC media lacking either uracil or leucine, harvested by centrifugation, resuspended in fresh medium to ~10 OD600 units, and loaded with 25 μg/ml coelenterazine (Molecular Probes, Eugene, OR) for 20 min at room temperature. Loaded cells were raised in the indicated media to an OD600 of 0.250 and treated with 20 μm α-factor and/or 2.0 μm FK506. Luminescence was monitored in a LB9507 luminometer (EG&G Wallac) and expressed as relative luminescence units (RLU). This procedure resulted in equivalent loading of different strains as judged by measuring total RLU after cell lysis with digitonin.
Western blots: Cells were grown to log phase in SC media lacking uracil or leucine, harvested, and extracted for membrane proteins as described (Katzmannet al. 1999). Briefly, 2 OD600 units of log phase cells pretreated for 4 hr at 30° with 1 μm α-factor and/or 2.0 μg/ml FK506 were lysed in breaking buffer [0.3 m sorbitol, 0.1 m NaCl2, 5 mm MgCl2, 10 mm Tris-Cl (pH 7.6) plus protease inhibitor cocktail with BAME, TAME, TLCK, AEBSF, leupeptin, and pepstatin (Sigma)] with glass beads by vortexing at high speed for 30 sec followed by incubation on ice for 30 sec a total of four times. Extracted proteins were solubilized in sample buffer [40 mm Tris-Cl (pH 6.8), 8 m urea, 15% SDS, 0.1 mm EDTA, 1% β-mercaptoethanol, 0.01% bromophenol blue, and protease inhibitor cocktail], heated at 37° for 1 min, centrifuged, fractionated by 8% SDS-PAGE, transferred to polyvinylidene difluoride (Millipore, Bedford, MA), and probed with either 12CA5 (Boehringer Mannheim, Indianapolis) or 9E10 (Santa Cruz Biotechnologies) monoclonal antibodies specific for Mid1p-HA (Iidaet al. 1994) and Cch1p-MYC (Lockeet al. 2000) tagged proteins, respectively. Cross-reacting proteins were then detected, using an ECL kit (Amersham).
Two independent Ca2+ influx pathways are stimulated by pheromone signaling: Ca2+/calmodulin-dependent activation of protein kinases (Cmk1p, Cmk2p) and phosphatases (calcineurin) has been shown to be important for long-term survival of a cells treated with α-factor (Moseret al. 1996; Witheeet al. 1997). Cch1p and Mid1p, putative components of a HACS (Lockeet al. 2000), have also been shown to be important for survival in SD-100 medium, a synthetic minimal growth medium containing 100 μm Ca2+ (Iidaet al. 1994; Fischeret al. 1997; Paidhungat and Garrett 1997). To confirm that HACS and Ca2+/calmodulin function in the same pathway, we compared the survival rates of cch1 null mutants, cmd1-3 mutants, and cch1 cmd1-3 double mutants at various times after treatment with α-factor. After treatment with α-factor in SD-100 medium, survival of cch1 mutants was diminished relative to an isogenic wild-type strain (Figure 1A). A cmd1-3 mutant, which carries six point mutations in calmodulin that abolish all high-affinity binding sites for Ca2+ (Geiseret al. 1991), also exhibited decreased survival after α-factor treatment (Figure 1A). Survival of a cch1 cmd1-3 double mutant was not significantly different from that of the single mutants, suggesting HACS and calmodulin do indeed function nonadditively within a common pathway.
Surprisingly, different results were obtained using YPD growth medium, a complex rich medium containing 140 μm Ca2+. In YPD medium, cch1 mutants showed no decrease in survival relative to wild type whereas cmd1-3 mutants and cch1 cmd1-3 double mutants continued to exhibit decreased survival (Figure 1B) much like the cmd1-6 mutant used in previous studies (Moseret al. 1996). Therefore, in rich YPD medium functional Ca2+/calmodulin promotes survival independent of the HACS protein Cch1p. These results suggest that a mechanism distinct from HACS may supply Ca2+ to calmodulin under these conditions.
To test whether cytosolic Ca2+ signals were generated independent of HACS in rich YPD medium, expression of the calcineurin-dependent reporter gene CDRE-lacZ was measured after treatment of a cells with α-factor. In wild-type cells, treatment with α-factor strongly induced CDRE-lacZ expression up to ~50% maximal levels obtained by adding 100 mm CaCl2 extracellularly (Figure 2A) and this induction was totally blocked by addition of FK506, a potent inhibitor of calcineurin (Schreiber and Crabtree 1992). Calcineurin-dependent induction of the reporter gene in cch1, mid1, and cch1 mid1 double mutants was diminished only slightly under these conditions (Figure 2A). Thus, in rich YPD medium, Ca2+ signals were produced independent of HACS in response to treatment with α-factor.
The HACS-independent Ca2+ signals described above might be generated through a variety of mechanisms, such as decreased Ca2+ efflux from the cytoplasm or increased Ca2+ influx from either intracellular stores or extracellular spaces. To help discriminate between these possibilities, we determined the initial rates of Ca2+ influx into cells from the culture medium using 45Ca2+ as a tracer (see materials and methods). A cch1 mid1 sst1 triple mutant, lacking the secreted protease Sst1p that degrades α-factor (Chan and Otte 1982), was grown in fresh YPD medium for 90 min with or without excess α-factor before addition of 45Ca2+ tracer. In duplicate experiments, 45Ca2+ accumulation was found to be linear with time over 60 sec (Figure 2B) and therefore useful to calculate the initial rates of Ca2+ influx. Cells pretreated with α-factor exhibited a sevenfold higher initial rate of Ca2+ influx than the untreated cells, suggesting that a novel HACS-independent Ca2+ influx system can be stimulated by pheromone signaling in cells grown in rich YPD medium. In contrast, cells grown in minimal SD-100 medium exhibited little or no stimulation of a HACS-independent Ca2+ influx system (Iidaet al. 1994; Fischeret al. 1997; Paidhungat and Garrett 1997).
Long-term 45Ca2+ accumulation experiments represent a composite of Ca2+ influx and efflux rates. When 45Ca2+ accumulation was measured over the entire 4-hr period of α-factor treatment (Figure 3), the results correlated well with the short-term measurements of Ca2+ influx (Figure 2B). In YPD medium, for example, α-factor treatment stimulated Ca2+ accumulation to similar degrees in wild-type and cch1 mid1 double mutant strains (Figure 3A). In minimal SD-100 medium, however, α-factor treatment stimulated Ca2+ accumulation in wild type but had no significant effect on the cch1 mid1 double mutant (Figure 3B). Thus, a HACS-independent Ca2+ accumulation system was stimulated by α-factor treatment in YPD medium but not SD-100 medium. Minimal SD-100 medium lacks amino acids and many of the other nutrients present in complex YPD medium. To determine the physiological inducer of the HACS-independent system, we began by adding back nutrients to SD-100 media. Supplementing SD-100 medium with all 20 amino acids plus adenine and uracil (termed synthetic complete SC-100 medium) restored the HACS-independent 45Ca2+ accumulation activity in cch1 mid1 double mutants to near wild-type levels (Figure 3C). These results indicate that culture media rich in amino acids are necessary for stimulation of the HACS-independent Ca2+ influx activity and that this activity, like HACS, can be detected in long-term Ca2+ accumulation experiments.
Interestingly, a DNA microarray experiment revealed that expression of many calcineurin-dependent genes was induced in wild-type cells after treatment with α-factor but not in bni1 mutants (Robertset al. 2000). These results suggest that Bni1p, a protein required for polarized morphogenesis but not cell cycle arrest or induction of mating-specific genes, might be required for the generation of Ca2+ signals after α-factor treatment. To test this possibility directly, induction of the calcineurin-dependent CDRE-lacZ reporter gene was measured in bni1 null mutants growing in YPD medium. After 4 hr of treatment with α-factor, induction of CDRE-lacZ expression was almost completely absent in the bni1 mutant although treatment with 100 mm CaCl2 resulted in wild-type levels of induction (Figure 4A). Moreover, α-factor treatment completely failed to stimulate 45Ca2+ accumulation in bni1 mutants relative to wild type (Figure 4B). Similar results were obtained with another mutant (spa2) that fails to undergo polarized morphogenesis in response to α-factor (Figure 4). Thus, factors specifically involved in polarized morphogenesis after α-factor treatment were essential for the stimulation of the HACS-independent system even in YPD medium.
To estimate the affinity of the novel influx system for Ca2+ we varied extracellular Ca2+ concentrations in SC media and monitored 45Ca2+ accumulation. To do this, cch1 mid1 sst1 triple mutant cells growing in SC-100 medium were washed in the same medium but lacking Ca2+ (SC-0 medium), aliquoted, and labeled with increasing concentrations of 45CaCl2 for 3 hr with or without α-factor treatment. In both cases, 45Ca2+ accumulation was saturable and fit to the Michaelis-Menten equation by nonlinear regression. In mutants lacking a HACS (Figure 5A), α-factor treatment stimulated the appearance of a LACS with apparent Km for Ca2+ estimated at ~3 mm. Data in earlier studies have indicated that calcineurin may negatively regulate Ca2+ accumulation stimulated by treatment with α-factor (Matheoset al. 1997; Witheeet al. 1997). However, addition of FK506 during treatment with α-factor increased the Vmax of 45Ca2+ accumulation by only 1.7-fold with no significant effect on the apparent Km (Figure 5A).
To summarize these results, a novel LACS was strongly stimulated in cells undergoing polarized morphogenesis in response to α-factor treatment and perhaps weakly modulated by calcineurin. This response was evident in cells growing in either rich YPD medium or synthetic complete SC-100 medium but not in synthetic minimal SD-100 medium. LACS activation can account for the HACS-independent cell survival and CDRE-lacZ expression we observed under these conditions.
Regulation of HACS by calcineurin and Ste12p activation independent of polarized morphogenesis: The above experiments revealed little or no contribution of HACS in YPD medium after treatment with α-factor. In complete SC medium containing a wide range of Ca2+ concentrations, treatment of sst1 mutants with α-factor stimulated Ca2+ accumulation to a level that was only slightly higher than that of cch1 mid1 sst1 triple mutants (Figure 5B, compare solid curves to dashed curves reproduced from Figure 5A). The difference between the two strains, representing HACS, was fit to the Michaelis-Menten equation but large statistical error prevented accurate estimation of Km and Vmax. Addition of FK506 during the α-factor treatment caused a striking difference between the two strains. After subtracting LACS activity and fitting the data to the Michaelis-Menten equation, the residual HACS activity exhibited a Km for Ca2+ that was estimated at 0.19 mm, ~16-fold lower than that of LACS. Thus, α-factor treatment strongly stimulates HACS activity only in the absence of calcineurin activity.
HACS activity requires at least Cch1p and Mid1p (Fischeret al. 1997; Paidhungat and Garrett 1997; Lockeet al. 2000) although it has been suggested that Mid1p may function as an ion channel in the absence of Cch1p (Kanzakiet al. 1999). To test whether these proteins might function independently after calcineurin-dependent regulation is removed, we performed long-term 45Ca2+ accumulation experiments on cch1 mutants and mid1 mutants in addition to the double mutant. In SD-100 medium (Figure 6A) where LACS function is undetectable, addition of α-factor alone, FK506 alone, or FK506 plus α-factor strongly stimulated 45Ca2+ accumulation into wild-type cells but had no effect on cells lacking Cch1p, Mid1p, or both. In YPD medium (Figure 6B), α-factor treatment stimulated 45Ca2+ accumulation in all strains (LACS activity); but simultaneous addition of FK506 strongly increased 45Ca2+ accumulation in the wild-type cells (~3.0-fold) but did so very weakly in cch1 mutants, mid1 mutants, and cch1 mid1 double mutants (1.1- to 1.3-fold). Thus, there was no evidence that Cch1p and Mid1p independently function in Ca2+ accumulation under any conditions tested in yeast. In YPD medium, LACS appears to be the primary Ca2+ influx system due to (direct or indirect) inhibition of HACS activity by calcineurin.
We next investigated which branches of the pheromone-signaling pathway stimulate HACS activity. In the presence of FK506, α-factor treatment strongly stimulated 45Ca2+ accumulation in bni1 mutants and was abrogated only by further deletion of cch1 (Figure 6C), indicating that, in contrast to LACS, HACS stimulation still occurred in the absence of polarized morphogenesis. A far1 mutant, which fails to arrest in the cell cycle or undergo normal polarized morphogenesis (Elionet al. 1993; Peteret al. 1993; Tyers and Futcher 1993; Cherkasovaet al. 1999), closely resembled the 45Ca2+ accumulation profile of the bni1 mutants (Figure 6C). However, a ste12 mutant, which fails to induce any genes in response to α-factor, exhibited no response to α-factor with or without FK506. HACS stimulation therefore required induction of Ste12p-dependent genes but not cell cycle arrest or polarized morphogenesis.
If activation of the mating transcription factor Ste12p is sufficient for stimulation of HACS or LACS, cells bearing constitutively activated Ste12p would be expected to exhibit high levels of 45Ca2+ accumulation independent of added α-factor. To test this hypothesis, we first examined the properties of a dig1 dig2 double mutant, which constitutively expresses Ste12p-dependent genes (Cooket al. 1996; Tedfordet al. 1997). 45Ca2+ accumulation into dig1 dig2 double mutants was similar to wild type in nonsignaling conditions, but, unlike wild-type cells, was also strongly stimulated after addition of FK506 (Figure 7A). FK506 addition failed to stimulate 45Ca2+ accumulation in dig1 dig2 strains that also lacked Ste12p, Mid1p, or Cch1p (Figure 7A). Alternatively, overexpression of Ste12p from the galactose promoter was capable of stimulating 45Ca2+ accumulation in the presence of FK506 (Figure 7A). These results suggest constitutively active Ste12p is sufficient to activate HACS but unable to overcome the inhibitory effects of calcineurin. Introduction of the cnb1 mutation to a dig1 dig2 strain stimulated 45Ca2+ accumulation to high levels with or without FK506 addition (Figure 7B). This finding confirms that the Ca2+ influx activity of Cch1p and Mid1p is regulated by calcineurin activity as opposed to a side effect of the drug.
As calcineurin also modulates the activities of vacuolar Ca2+ transporters Vcx1p and Pmc1p (the latter mediated by the transcription factor Tcn1p), we tested whether any of these factors were required for the effect of FK506 on dig1 dig2 strains. 45Ca2+ accumulation into dig1 dig2 vcx1 and dig1 dig2 tcn1 triple mutants was stimulated by FK506 to the same degree as in dig1 dig2 double mutants (Figure 7B), demonstrating that Vcx1p and Tcn1p were not required for the effect. The dig1 dig2 pmc1 triple mutant also responded to FK506; however, the levels of 45Ca2+ accumulation were lower due to the decreased Ca2+ sequestration into the vacuoles of pmc1 mutants (Cunningham and Fink 1994). FK506 addition also stimulated luminescence of dig1 dig2 pmc1 vcx1 quadruple mutants expressing cytoplasmic aequorin but not of pmc1 vcx1 double mutants (Figure 7C). Thus, Ste12p activation alone in the absence of signaling by upstream factors was sufficient to stimulate the Ca2+ influx activity of Cch1p and Mid1p through a process that was sensitive to calcineurin but independent of several known calcineurin targets (Pmc1p, Tcn1p, and Vcx1p).
To determine whether the expression of Cch1p and/or Mid1p themselves was stimulated by Ste12p activation, the abundance of these proteins was measured by Western blot analysis after treatments with α-factor and/or FK506 for 4 hr. Expression of an epitope-tagged Mid1p-HA fusion protein from a low-dosage plasmid in an sst1 mutant was not significantly affected by addition of α-factor, FK506, or both (Figure 8A). Similarly, expression of an epitope-tagged Cch1p-MYC fusion protein from the chromosomal locus was also unaffected by treatments with α-factor and/or FK506 (Figure 8B). The findings suggest Ste12p induces expression of an unknown activator of HACS rather than stimulating expression of Cch1p and Mid1p themselves.
The results presented here demonstrate the existence of two Ca2+ influx systems in yeast that can be independently activated in response to the pheromone-signaling cascade. The previously described system (HACS) exhibited relatively high affinity for Ca2+, required Cch1p and Mid1p function and Ste12p activation, and was activated strongly only when calcineurin was inactivated. The previously undescribed system (LACS) was characterized by a relatively low affinity for Ca2+, little or no sensitivity to calcineurin, independence of Cch1p or Mid1p, and dependence on Bni1p and Spa2p. LACS was detected in rich YPD medium and synthetic complete SC-100 medium but not in the minimal SD-100 medium lacking amino acid supplements. Either LACS or HACS was sufficient to activate Ca2+/calmodulin-dependent signaling pathways. These findings extend earlier studies that indicate survival in SD-100 medium requires Cch1p and Mid1p (Iidaet al. 1994; Fischeret al. 1997; Paidhungat and Garrett 1997) and afford a more careful characterization of the mechanisms that coordinate HACS and LACS activities.
Regulation of LACS: A surprising finding was that cch1 mid1 double mutants lacking HACS were still capable of producing robust calcium signals in response to pheromone. In these cells, α-factor stimulated 45Ca2+ influx and accumulation, elevated [Ca2+]c, and induced calcineurin-dependent genes to nearly the same levels as those seen in wild-type cells. This activity has not been described in previous reports and is characterized by a relatively low affinity for Ca2+, insensitivity to calcineurin, and dependence on the functions of Far1p, Bni1p, and Spa2p, in addition to treatment with α-factor. Far1p, Bni1p, Spa2p, and numerous interacting proteins become localized to the morphological projections in cells responding to α-factor (Doreret al. 1995; Valtzet al. 1995; Buttyet al. 1998; Blondelet al. 1999; Milleret al. 1999; Nern and Arkowitz 1999; Shimadaet al. 2000). Mutants lacking Bni1p or Spa2p form abnormal projections and also fail to induce a set of genes governed by the cell integrity signaling pathway, which includes the MAP-kinase Mpk1p (Robertset al. 2000). LACS activation does not require Mpk1p because pheromone signaling stimulated LACS activity to indistinguishable levels in mpk1 mutants and wild type (data not shown). Therefore, it is possible that normal polarized morphogenesis and/or membrane stretch, which occurs in wild-type cells but not in bni1 mutants, stimulate LACS. Finally, LACS function appears only in rich media such as YPD or synthetic complete SC media. Because the only difference between our SC-100 and SD-100 was the presence of additional amino acids and nutrients, it seems likely that nitrogen deficiency of SD-100 prevents LACS activation or expression. Further dissection of the regulatory mechanism that activates LACS would be facilitated by identification of the membrane proteins that compose this ion transporter or channel.
Regulation of HACS activity: The studies of Iida and others identified Mid1p and Cch1p as factors required for Ca2+ influx and survival of cells responding to α-factor (Iidaet al. 1994; Fischeret al. 1997; Paidhungat and Garrett 1997). We have confirmed these results and demonstrated that both proteins function in the same pathway as Ca2+/calmodulin in the minimal SD-100 medium (Figure 1). Cch1p/Mid1p (HACS) activity was strongly stimulated by the response to α-factor plus FK506 after a lag time of ~45 min (Figures 5 and 6 and data not shown). The abundance of Cch1p and Mid1p was unchanged in these conditions (Figure 8), consistent with prior analysis of the transcripts from the CCH1 and MID1 genes (Iidaet al. 1994; Robertset al. 2000). Nevertheless, we determined that activation of the Ste12p transcription factor was necessary and sufficient to stimulate HACS activity as long as calcineurin was inhibited or inactivated by mutation. For example, dig1 dig2 double mutants responded rapidly to FK506 with increased 45Ca2+ accumulation through a process dependent on Ste12p, Cch1p, and Mid1p (Figure 7).
The targets of Ste12p involved in HACS activation have not yet been identified. On the basis of the hypothesis that calcineurin inhibits some step in HACS activation, we speculated that a protein kinase may be the critical target of Ste12p. Of the 124 genes in yeast predicted to encode protein kinases, only FUS3 and KSS1 were significantly induced in dig1 dig2 double mutants (Robertset al. 2000). However, we found that FK506 stimulated 45Ca2+ accumulation in dig1 dig2 fus3 kss1 quadruple mutants normally (data not shown), so these kinases are not likely to oppose calcineurin in the regulation of HACS. This experiment also reaffirms other data shown here that factors upstream of Ste12p in the pheromone-signaling cascade are not required for HACS stimulation as long as Ste12p is active. Additional studies will be required to determine the mechanisms by which Ste12p and calcineurin regulate HACS.
Expression of Mid1p alone in human embryonic kidney cells resulted in the appearance of Ca2+-permeable stretch-activated channel activity (Kanzakiet al. 1999), suggesting Mid1p might function alone as an ion channel. Mid1p and Cch1p colocalize to the plasma membrane in yeast cells and can be co-immunoprecipitated from solubilized cell extracts (Kanzakiet al. 1999; Lockeet al. 2000). Additionally, overexpression of Mid1p in yeast did not alter the rates of Ca2+ accumulation with or without α-factor (Iidaet al. 1994). Because cch1 and mid1 single mutants exhibit phenotypes quantitatively identical to cch1 mid1 double mutants in a variety of contexts (Fischeret al. 1997; Paidhungat and Garrett 1997; Lockeet al. 2000), we suggest that the two proteins comprise two essential subunits of HACS that independently have no significant Ca2+ transport activity in yeast.
The role of Ca2+ in mating and the pheromone response: Our findings support a model where yeast cells growing in complete or rich media primarily employ LACS to generate calcium signals during the pheromone response. LACS activation was sufficient to stimulate calcineurin-dependent gene expression and presumably calcineurin-dependent inhibition of HACS. If LACS activity were insufficient to activate calcineurin, HACS activity might become more obvious and more significant physiologically. HACS activity alone in dig1 dig2 double mutants treated with FK506 was sufficient to elevate [Ca2+]c (Figure 7) and presumably to stimulate calcium signaling pathways. Therefore, LACS and HACS may simply represent a means of activating calcium signaling pathways over a broader range of environmental conditions. Alternatively, HACS and LACS activation may be spatially or temporally separated in the responding cell leading to distinct types of responses. Cch1p and Mid1p localize uniformly to the plasma membrane of vegetative and shmooing cells (Kanzakiet al. 1999; Lockeet al. 2000) and therefore may contribute to calcium signaling processes in both the cell body as well as the morphological projections of shmooing cells. The differential regulation of LACS and HACS by distinct branches of the pheromone-signaling pathway may lead to subtle differences in their physiological contributions. However, LACS and HACS activation after α-factor treatment exhibited identical kinetics and sensitivity to α-factor doses (data not shown). Finally, the distinct properties of the two Ca2+ influx activities may cause different effects, as observed previously in neurons (Badinget al. 1993; Hardinghamet al. 1998) and sperm cells (Darszonet al. 1999; Wassarman 1999). That yeast cells employ two separate Ca2+ influx activities during the pheromone response suggests an important role for calcium signaling in controlling downstream events.
What are the roles of calcium signaling during the pheromone response? Yeast cells responding to α-factor tend to survive for longer periods of time when sufficient Ca2+ is present in the environment and when Cch1p, Mid1p, calmodulin, calmodulin-dependent protein kinases, and calcineurin are all functional (Iida et al. 1990, 1994; Moseret al. 1996; Witheeet al. 1997). However, the critical targets of the calmodulin-regulated enzymes responsible for promoting cell survival have not been completely defined. Calcineurin activates the transcription factor Tcn1p (also known as Crz1p and Hal8p), which then induces expression of a number of genes involved in ion homeostasis and cell wall biogenesis (Matheoset al. 1997; Stathopoulos and Cyert 1997; Mendizabalet al. 1998). Tcn1p overexpression partially suppresses the survival defect of calcineurin-deficient mutants and tcn1 mutants exhibit a slight survival defect (Stathopoulos and Cyert 1997), suggesting that targets of Tcn1p at least partially promote survival. Many of these targets of Tcn1p were also induced following α-factor treatment, particularly in a later wave of transcription that was absent in bni1 mutants (Robertset al. 2000). It is not yet clear which of these targets are important for survival during the pheromone response.
A number of new questions are raised by this study. What are the targets of Ste12p and how do they activate HACS? How does calcineurin regulate HACS activity? What are the components of LACS and how are they regulated by cell morphogenesis factors? The complete sequence of the yeast genome does not reveal any gene products with obvious similarity to the presently known families of Ca2+ channels, except for Cch1p, the yeast homolog of voltage-gated Ca2+ channels. However, it seems likely that not all families of Ca2+ influx channels have been identified. Answers to these questions may define a new family of Ca2+ channels conserved in other species and produce new insight into mechanisms of ion channel regulation.
We are grateful to Beverly Errede, Elaine Elion, Patrick Masson, and Hidetoshi Iida for plasmids, and to Fujisawa USA for the generous gift of FK506. We also thank all members of our department and laboratory, especially Elizabeth P. O'Sullivan, for advice and technical support. This study was supported by grants from the DuPont Young Professors Award and the National Institutes of Health (GM53082).
Communicating editor: M. D. Rose
- Received August 1, 2001.
- Accepted September 25, 2001.
- Copyright © 2001 by the Genetics Society of America