Essential Role for Induced Ca2+ Influx Followed by [Ca2’]i Rise in Maintaining Viability of Yeast Cells Late in the Mating Pheromone Response Pathway A STUDY OF [Ca2+li IN SINGLE SACCHAROMYCES CEREVISIAE CELLS WITH IMAGING OF FURA-2*

We established an ex~rimental system for measur- ing the cytosolic-free Ca2+ concentration ([Ca”]if in individual ~u~~rom yces cereuisiae cells using fura- 2 as a Ca’+-specific probe in conjunction with digital image processing and examined changes in [Ca2+lj in response to a-factor in single cells of a mating type. The addition of a-factor to a cells raised [Ca2+]i to several hundred nanomolar in the cells from a basal level of approximately 100 nM, simultaneous with the induction of Ca2+ influx. When the cells were incubated with a-factor in a Ca2+-deficient medium, Ca2’ influx was greatly reduced, and the rise in {Ca2+]i was not detected. This indicates that the a-factor-induced rise in [Ca2+]i is generated by Ca2+ influx through the plasma membrane and not by release from internal stores. In the Ca2+-deficient

This indicates that the a-factor-induced rise in [Ca2+]i is generated by Ca2+ influx through the plasma membrane and not by release from internal stores. In the Ca2+-deficient medium, a cells died specifically after they had changed into cells with one projection on the cell surface. This indicates that the rise in [Ca2+]i is essential for the late response to afactor. The duration of Ca2+ requirement for maintaining viability was limited to this stage, and the earlier and later stages were not affected by Ca2+ deprivation. Mating between a and a mating type cells was impaired in this medium due to cell death at and before the stage of conjugation.
These findings are the first evidence for an essentia1 role for mobilized Ca2+ in the yeast life cycle.
In higher eukaryotic cells, Ca*+ plays essential roles in the regulation of cellular functions, including hormone secretion, neurotransmitter release, muscle contraction, fertilization, and lymphocyte activation by acting as a second messenger in response to extracellular stimuli (for reviews, see Refs. 1, 2). These stimuli induce a rapid and transient rise in [Ca*+]i in target cells either by opening Ca*+ channels in the plasma membrane to allow extracellular Ca2+ to enter the cytoplasm or by activating cell surface receptors that trigger hydrolysis * This work was supported in part by Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan and a fund from the Uehara Memorial foundation, Japan. This work contributed to the NIBB Program for Biomembrane Research. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "uduert~e~e~t" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Present address: Division of Biology, Jichi Medical School, Yakushiji, Minamikawachi, Tochiti 329-04, Janan.
In S~c~u~myces cerevisiue, the mating process of haploid cells is controlled by the mating pheromones, a-and a-factors, that are synthesized and secreted by a and cy mating type cells, respectively (for reviews, see Refs. 3,4). The mating pheromones induce several responses in cells of the opposite mating type: in the early stage, they rapidly alter the pattern of gene expression (5,6), induce cell surface agglutinins that facilitate (but are not essential for) mating (7,8), and arrest mitotic cell division in the G, phase (9,10). In the late stage, they induce morphological changes into cells (so-called shmoos) with one or more projections on the cell surface that act as the points of contact between mating cells (11,12), and lead to cell and nuclear fusions (13)(14)(15).
The mating pheromone response pathway has been shown to involve mechanisms similar to those found in mammalian cells, such as the pheromone-receptor interaction (16-M) and the function of the pheromone receptor-coupled guanine nucleotide-binding regulatory (G) protein (19)(20)(21), which are essential for mating pheromone signal transduction. However, nothing is known about the role of extracellular and intracellular Ca2+ in this pathway except that Ca2+ influx in a cells is stimulated by a-factor after a lag of 30-40 min in a dosedependent manner (22).
The study presented here was designed to investigate whether Ca" influx induced by a-factor correlates with changes in [Ca""]i that may have a regulatory role for signal transduction of the mating pheromone and to elucidate the role of induced Ca2+ influx and changes in [Ca2+li in the mating pheromone response pathway. To measure [Ca*+]; in individual yeast cells, we employed fura-as a Ca*+-specific probe, which has been widely used to measure [Ca*+]i in mammalian cells, in conjunction with digital image processing (23,24). To elucidate the role of induced Ca*+ influx and changes in [Ca'+]:, we investigated the effect of Ca'+ deprivation on Ca2+ influx, [Ca2*]i changes and progression of the mating process.
The results show that [Ca"]; rises following an influx of Ca" which is induced by a-factor. When this influx and consequent rise are prevented by incubating a cells with (Yfactor in a Ca*+-deficient medium, the cells die specifically after they have changed into shmoos. The duration of the Ca2+ requirement for maint~ning viability is relatively short and the requirement is specific to this limited stage. Mating between a and cy cells is thereby impaired due to cell death at or before the stage of conjugation. These results indicate that induced Ca2+ influx followed by a rise in [Ca'+& is essential for the late stage of the mating pheromone response. showed that the efficiency of mating was essentially the same between cells with sufficient fura-and those with insufficient fura-and that zygotes with sufficient fura-budded at normal frequency (Table  II and  Cells of strain RC629 growing in SD medium at 30 "C were divided into two equal parts. One portion (0) was subjected to electroporation with fura-2, washed three times with distilled water, and resuspended in the same medium, as described under "Materials and Methods." The other portion (0) was kept in the same growth conditions. Cell density was measured in a hemocytometer and viability was examined by the methylene blue method after resuspending cells in the medium. Note that a decrease in the cell density for treated cells is due to the loss of cells during recovering them from the filtration apparatus. ; into shmoos and the formacells.
Experimental conditions were described in the legends for Tables  I and II. a-d, cells of a mating type subjected to electroporation with fura-were exposed to 6 pM <u-factor for 2 h. Two shmoos are shown. One shmoo contains sufficient fura-in the vacuole and the other does not. e-h, cells of a mating type subjected to electroporation with fura-were mixed with cells of (\' mating type and incubated for 4 h as described in the legend for Table II for SD and 2 PM (74 KBq)/ml for SD-Ca) for 10 min. The cells were washed as described above and resuspended in distilled water. Portions of the suspensions were taken and radioactivity of the cells were measured after filtration on GF/C filters as described above. The remainders were divided into two equal parts. One received 80 PM fura-and were subjected to electroporation, washed, and resuspended in the same media used for incubation before electroporation as described above. The others were treated as above except that electroporation was omitted. The samples were then incubated for 5 and 10 min in the media and filtrated on GF/C filters. The radioactivity retained on the filter was measured.
The results showed that the rates of Ca" efflux were essentially the same between cells subjected to electroporation and those not subjected to it. We have also confirmed that Ca" content of cells that had been labeled with ""CaCl? as described above was not perturbed by washing with distilled water during filtration which was used before and after electroporation. Fura-acetoxymethyl ester, a compound more permeable to the mammalian plasma membrane than fura-2, appeared not to be permeable to the yeast envelope.
Loaded  ; in a cells (strain RC629) growing exponentially in the complete synthetic medium SD was low (Fig. 3, a and e). The average [Ca2+li in the cells was found to be 116 nM (SD = 90 nM), although [Ca"]; appeared to vary from cell to cell (Fig. 3e). We found no significant difference in [Ca2+]i between budded cells and unbudded cells. When m-factor was added to a concentration of 6 FM, a cells underwent a rise in [Ca"], with the onset of Ca"+ influx. When Ca'+ influx was induced 40 min after the addition of a-factor and 30% of the cells had changed into shmoos ( Fig. 4a and b), [Ca'+li in some shmoos (5 out of 11) rose above 200 nM (281-720 nM) (Fig. 3e). [Ca'+]i in cells that had not changed into shmoos remained low (data not shown). In addition, no significant rise in [Ca*+]; was observed 5 min (data not shown) and 20 min (Fig. 3e) after the addition of CXfactor. One h after the addition, when the level of Cay+ accumulation was about half its plateau level and 56% of cells had changed into shmoos with one projection (Fig. 4, a and  b), [Ca'+]i rose above 200 nM (226-1091 nM) in 9 out of 11 shmoos (Fig. 3e). A similar result was obtained with cells incubated with a-factor for 2 h in which Ca'+ accumulation had reached plateau level and 91% of the cells had changed into shmoos (Fig. 3e and Fig. 4, a and b). It should be noted that [Ca'+]i in all shmoos was not necessarily high. The reason for this apparent deviation in [Ca'+li will be given under "Discussion." The [Ca"'], rise was not detected when (Y cells were treated with cu-factor (data not shown) and when a cells were incubated with a-factor in a Ca'+-deficient medium, SD-Ca (Fig. 3, c, d,  this medium was 0.24 ELM, compared with 681 PM in SD. Vegetative growth of cells of strain RC629 in SD-Ca was very similar to that in SD (7'1, = 1.8 + 0.1 h in SD and 2.0 f 0.1 h in SD-Ca), even when serial reinoculations into SD-Ca (seven times in succession) were performed. Similar results were reported earlier (32). We then tested whether the responses of a cells to a-factor were affected by incubating cells in SD-Ca. In most experiments we used the sstl (or burl) mutant that produces a defective extracellular pepsin-like protease that degrades a-factor and is lo-30 times more sensitive to the factor than wild-type strains (25,33,34). Fig. 4a shows that the amount of Ca*+ accumulated 2 h after the addition of a-factor was loo-fold lower in cells in SD-Ca than those in SD (a positive control) and that the apparent initial rates of Ca*+ uptake of cells in SD-Ca and those in SD were 0.029 and 7.5 pmol/106 cells/min, respectively. The data obtained with cells grown in SD are comparable to those obtained with cells grown in YPD (22) . Fig. 3, c, d, and f, shows that [Ca'+]i in a cells incubated with LYfactor did not rise during a-factor treatment. Under these conditions, the calcium content of the medium had no effect on the kinetics of the responses to a-factor, including growth arrest (data not shown), disappearance of budded cells (Fig.  4b), appearance of shmoos (Fig. 4b), and increase in the population of cells with G1 DNA content (data not shown). These results indicates that the [Ca2+]i rise generated by Ca'+ influx are not responsible for progression of these early events during the mating pheromone response pathway.
However, we found differences in size and shape between shmoos incubated in SD-Ca and those in SD when the cells were incubated for 4 h or more. In SD-Ca, the shmoos were smaller in size than in SD, and most of the shmoos in SD-Ca had only one projection, while those in SD had two or three projections. This observation suggests that cellular growth of shmoos formed in SD-Ca may be restricted or the shmoos may die due to a lack of Ca" accumulation. To test these possibilities, viability was determined by the methylene blue method (28) and by measuring colony-forming ability. Microscopic observation showed that small shmoos with one projection after 10 h of incubation in SD-Ca were all methylene blue-positive, indicating that they were inviable (Fig. 5, u-c). The time course of the appearance of these inviable cells showed that 85% died within 5 h of the addition of a-factor, while 80% of the cells remained viable when incubated in complete medium (Fig. 4~). Similar results were obtained when colony-forming units were determined (Fig. 4~). On the basis of those microscopic observations and kinetic data for viability and shmoo formation, it is obvious that this cu-factorinduced cell death occurred after a cells had changed into shmoos. We have noticed that this phenomenon is more severe and occurs more rapidly when a low pH medium (pH 3.5) is used, in which degradation of a-factor is diminished (30). In addition, we have found that o( cells also die when incubated with u-factor in SD-Ca, indicating that the same phenomenon applies to both factors. The results shown in this paper were obtained with a cells incubated in media with the usual pH of 5.7. Fig. 6 shows that a-factor-induced cell death is specifically caused by the lack of Ca'+ in the medium. Addition of CaCl, above 1 FM resulted in rescue of the cells from cu-factorinduced death, the most effective concentrations being 681 PM and 1 mM both of which gave 71% viability. MgC& had no effect showing this response is specific to calcium. Fig. 7 shows the concentration dependence of a-factor on the induction of cell death. The dose resulting in 50% lethality (MATa) grown in SD-Ca were mixed, spun down, and incubated without shaking in 2 ml of SD-Ca or SD for 5 h at 30 'C (see legend to Table I  (LD5J was 3 X 10e6 M in the wild-type cells. This doseresponse curve is similar to that for a-factor-induced Ca*+ influx observed in a complete medium YPD in which cells are alive (22). The mutant sstl which is defective in a-factor degradation and the mutant sst2 which is defective in recovery from pheromone-induced growth arrest and is supersensitive to the pheromone (25,35) gave LDhO values of 1 X 10T7 M and 2 x lo-' M, respectively (Fig. 7). We have found that a-factor does not induce cell death in o( cells, a/a diploid cells or a Aste2 mutant cells which lack the a-factor receptor (36, 37) (data not shown). Furthermore, we used a mutant conditionally defective in expression of the GPAl (or SCGI) gene that encodes the LY subunit of the mating factor receptor-coupled G protein (19,20). In this mutant, the GPAl promoter is replaced by the GAL1 promoter, and GPAl expression is thereby induced in galactose-based media and repressed in glucose-based media. This mutant behaves like mating pheromone-treated cells when incubated in glucose-based media in the absence of mating pheromones (19) . Fig. 5, d-f and Fig.  8 show that this mutant died in a Ca'+-deficient, glucosebased medium as GPAl expression stopped and the popula- tion of unbudded cells rose. These results indicate that cell death is a consequence of signal transduction through the mating pheromone signaling pathway including specific pheromone-receptor interaction and the function of the pheromone receptor-coupled G protein under low Ca2+ conditions.
Ca"-requiring Period-To determine the Ca*+-requiring period during the a-factor response pathway, we performed the following two experiments. First, 6 PM a-factor was added to cells growing exponentially in SD-Ca, and 681 pM CaClz was then supplemented to the medium at the times indicated in Fig. 9a after the addition of the factor. Viability of each culture was measured 10 h after the addition. The result showed that cells could escape death if CaC12 was supplemented to the medium within 4 h of the addition of the factor. Second, cells growing exponentially in SD received 6 pM afactor and were shifted to SD-Ca at the times indicated in Fig. 9b and viability of each sample was measured. The result showed that cells could escape death only if high Ca*+ was present for at least 3 h after the addition of a-factor; halfmaximal viability was observed when Ca" was present for 4.2 h after the addition. After this time, cells could survive in SD-Ca (Fig. 95). Since shmoos can be formed in the absence of extracellular Ca*+, these results suggest that the period of Ca*+ requirement is between the time at which cells have changed into shmoos and 4.2 h after the addition of the factor.

Mating between a and a Cells in Ca'+-deficient Conditions-
We tested whether the lack of Ca*' in the medium affects mating between a and (Y cells by mixing and incubating them in SD-Ca. Table III shows that mating efficiency, scored by counting diploids formed, was 3-fold lower in SD-Ca than in SD, Microscopic observation of the cells 5 h after mixing shows that in SD-Ca, viability of unbudded zygotes was mostly diminished (Table III and Fig. 5, g and h) and viability of unbudded cells was also slightly diminished, probably due to exposure to the mating pheromone from cells of the opposite mating type. We could not confidently distinguish between growing unbudded cells and some shmoos in the mixed projection of the cell surface due to the action of a-factor. However, this prevention affects neither the kinetics of early responses to the factor, such as G1 arrest and morphological changes into shmoos with one projection on the cell surface nor viability of cells that have been exposed to the factor for 4.2 h or more ( Fig. 9), most of which (>98%) have changed into shmoos. Thus, the Ca'+-requiring stage appears to be limited between the time at which cells have changed into shmoos and 4.2 h after the addition of a-factor, and specific to shmoos with one projection, Time of CaCI, additlon (h) The [Ca'+]i rise appeared not to necessarily take place in a synchronous manner after the addition of a-factor (Fig. 3). We think that this was not due to technical error but rather due to a mixed cell population used in this study. Exponential growing cells respond to a-factor randomly and thus the time of [Ca'+]i rise should be random. Another compatible explanation is that the duration that cells have high [Ca*+]i is considerably short. a-Factor-induced cell death requires a high dose of a-factor (LDEo = 3 x 10e6 M) in wild-type cells (Fig. 7). This result applies to a model that high affinity a-factor receptor mediates early responses such as G1 arrest and agglutinin induction, and low affinity a-factor receptor mediates projection formation and Ca2+ accumulation (22,38). It is still possible to speculate that death may be an unphysiological response to a high dose of a-factor. However, this possibility is unlikely because mating between a and ry cells is significantly impaired in the Ca*+-deficient medium (Table III). In this condition, cells die mainly at the stage of conjugation or of unbudded cells that have been probably arrested in G1 due to the action of a mating pheromone produced by opposite mating type cells. Thus, we conclude that mobilized Ca*+ observed in this study is required for the late stage of the mating pheromone response pathway. A schematic model for Ca*+-requiring stages during mating pheromone response pathway is presented in Fig. 10. at the indicated times after the start -of incubation. Viability was determined bv the methvlene blue method 10 h after the addition of a-factor. b, the effect of the timing of CaC12 removal on o-factortreated cells incubated in SD. Cells of strain RC629 were incubated with 6 pM a-factor and at the indicated times, aliquots were filtered, washed twice with SD-Ca containing 6 pM a-factor and resuspended in the same medium. Viability of the cells was determined as in a.
population, and therefore both were counted as unbudded cells. Table III also shows that inviable unbudded zygotes accumulate in SD-Ca, and the formation of viable budded zygotes is thereby reduced. These results account for the low mating efficiency in SD-Ca and indicate that Ca2+ influx followed by [Ca*+]; rise is physiologically important in the mating process.

DISCUSSION
Mating Pheromone-induced Cell Death in Ca"-lacking Conditions-In this report, we have presented the first evidence that Ca*+ influx followed by a rise in [Ca'+]i induced by (Yfactor is essential for maintaining viability of S. cerevisiae cells late in the mating pheromone response pathway. Prevention of the Ca*+ influx and the [C!a*+]; rise by incubating a cells in a Ca*+-deficient medium correlates with induction of death of the cells that have changed into cells with one Some of the possible targets of mobilized Ca*+ may be Ca'+binding proteins. S. cereuisiae cells have calmodulin (39,40), a major Ca*+-binding protein, and other, putative, Ca*+-binding proteins such as the CDC31 gene product (41) and the CLS4 (or CDC24) gene product (42). Among them, the CLS4 (CDC.24) gene product is of particular interest. This gene product has two putative Ca*+-binding regions (42) and is essential for the morphogenetic processes of the cell division cycle which involve ordered polar cell surface growth (43). At the nonpermissive temperature, the temperature-sensitive cdc24 mutant of the a mating type does not form projections in response to a-factor (44), is defective in localized secretion of acid phosphatase (44), and has low mating efficiency (45). Thus, it is possible to speculate that disordered cell surface growth may occur due to the aberrant function of Ca'+dependent processes, such as that of the CLS4 (CDC24) gene product, in Ca*+-lacking conditions and this may cause serious damage leading to cell death. Although other yeast mutants defective in the function of calmodulin or the CDC31 gene product, in Ca2+ transport (46) or in Ca2+ metabolism (47)(48)(49)(50) have been isolated and characterized, none have been intensively tested in terms of the regulation of the mating process. The study of these mutants may uncover the molecular mechanisms underlying the mating process from the viewpoint of Ca*+ function.
Changes in [CV+]i and Ca*+ Homeostasis-By employing fura-as a Ca'+-specific probe, we have shown that [Ca*+], in individual, exponentially growing yeast cells is about 100 nM (Fig. 3). This concentration is comparable to that in many higher eukaryotic cells (51). In addition, we have determined the total Ca2+ content of yeast cells to be 49 pmol/106 cells with an atomic absorption spectrophotometer and the mean cell volume to be 31 pm3 with a Coulter counter. Thus, the total Ca2+ content of yeast cells was calculated to be 1.6 mM. We conclude that the cytosolic-free Ca2+ in yeast cells is 0.006% of the total. When a cells are exposed to a-factor in a complete medium, the average [Ca*+]i, the amount of Ca" accumulated, and the mean cell volume reach approximately 500 nM, 150 pmol/106 cells (Figs. 3 and 4a), and 42 pm3, respectively, 1 h after the exposure. These values would correspond to a concentration of 3.6 mM if all of this Ca'+ was in solution. These observations indicate that in absolute amounts, the rise in [Ca*+]i is small compared with the total Ca*+ accumulation and suggest that most of the Ca*+ incorporated into a cells is sequestered into internal stores. The major Ca*+-sequestering organelle in yeast cells is thought to be the vacuole whose membrane vesicles have a capacity to accumulate Ca*+ by a Ca'+/H+ antiport (52,53). Eilam and Chernichovsky (54) have shown that most cellular 45Ca2f is concentrated in the vacuole as a bound form after only 3 min. In contrast, unlike in mammalian cells, the mitochondria in yeast cells are incapable of Ca2+ uptake (55). It is, therefore, likely that most of Ca2+ incorporated into a cells by the action of a-factor is sequestered in the vacuole as a bound form. Several lines of evidence indicate that the rise in [Ca'+]i observed in this study is not a signal for the early response to a-factor. In mammalian cells, the rise in [Ca*+]i elicited by extracellular stimuli is rapid and transient, while the [Ca'+]i rise observed in this study is slow. In addition, like ol-factorinduced cell death, the [(?a*+]; rise in wild-type cells requires a high dose of a-factor, whereas early responses require a low dose of a-factor as discussed above. The rise in yeast could be a signal for the late response leading to conjugation. Alternatively, Ca2+ thus mobilized could act as a catalyst for reactions required for morphological changes which need the breakdown and synthesis of cell membrane and cell wall.
Our study strongly suggests that the a-factor signaling pathway including interaction between the factor and its receptor and the function of mating pheromone-coupled G protein are not dependent on extracellular Ca*+. Prevention of the Ca*+ influx followed by the [Ca*+], rise does not affect the early response to a-factor. Loss of GPAl function results in G1 arrest in a Ca'+-deficient medium just like in a complete medium. However, the possibility that a rapid and transient Ca2+ release from internal stores may participate in this signaling pathway remains to be examined.
Very rapid Ca*+ influx in response to a-factor at an unphysiological temperature (13 "C) was recently reported (56). However, we were not able to detect a significant, rapid increase in [Ca*+]i in our experimental conditions using fura-2. One explanation for this discrepancy may be a difference in experimental approaches. We are currently developing another method suitable for a rapid rise in [Ca*+]i.
The fact that only 0.006% of the total Ca2+ inside yeast cells is present as the cytosolic free Ca2+ (see above) gives an answer to the question of why yeast cells can grow infinitely in "Ca2+-free" media containing no mating pheromones (32). Even when incubated in a medium completely free from Ca*+, yeast cells are calculated to grow for I4 generations until Ca*+ content reaches a basal level of [Ca*']i, about 100 nM. The bound form of Ca*+ (1.6 mM) may well be released to generate the cytosolic-free Ca*+. Practically, since Ca*+-free media contain a considerable amount of contaminant Ca*+ salts, yeast cells can grow for further generations by using both extracellular and intracellular Ca*+ salts unless an appropriate mating,pheromone are added to the media.
Finally, our method established in this study for measuring [Ca'+], in individual yeast cells may be applied to the study on mitosis, secretion, and bud formation, all of which are thought to be dependent on the function of the cytosolic free Ca'+.