Implication of Ca²⁺ in the Regulation of Replicative Life Span of Budding Yeast

Abstract

In eukaryotic cells, Ca²⁺-triggered signaling pathways are used to regulate a wide variety of cellular processes. Calcineurin, a highly conserved Ca²⁺/calmodulin-dependent protein phosphatase, plays key roles in the regulation of diverse biological processes in organisms ranging from yeast to humans. We isolated a mutant of the SIR3 gene, implicated in the regulation of life span, as a suppressor of the Ca²⁺ sensitivity of zds1Δ cells in the budding yeast Saccharomyces cerevisiae. Therefore, we investigated a relationship between Ca²⁺ signaling and life span in yeast. Here we show that Ca²⁺ affected the replicative life span (RLS) of yeast. Increased external and intracellular Ca²⁺ levels caused a reduction in their RLS. Consistently, the increase in calcineurin activity by either the zds1 deletion or the constitutively activated calcineurin reduced RLS. Indeed, the shortened RLS of zds1Δ cells was suppressed by the calcineurin deletion. Further, the calcineurin deletion per se promoted aging without impairing the gene silencing typically observed in short-lived sir mutants, indicating that calcineurin plays an important role in a regulation of RLS even under normal growth condition. Thus, our results indicate that Ca²⁺ homeostasis/Ca²⁺ signaling are required to regulate longevity in budding yeast.

Introduction

The calcium ion (Ca²⁺) is a universal second messenger important in the regulation of diverse biological processes such as cell proliferation, muscle contraction, fertilization, development, motility, memory, and apoptosis (1). In addition, it was reported that α-Klotho, which was originally identified in short-lived mutant mice, is a major player in the regulatory system of Ca²⁺ homeostasis, suggesting a link between Ca²⁺ and aging (2).

In budding yeast, the Ca²⁺/calmodulin-dependent protein phosphatase calcineurin has been implicated in stress-induced gene expression, ion homeostasis, cell cycle regulation, and maintenance of viability after exposure to the mating pheromone (1, 3, 4). A recent report demonstrated that a calcineurin-deficient mutant of Caenorhabditis elegans displays an extended life span (5, 6). However, the relationship between Ca²⁺ and aging in yeast is not known.

Most of the identified regulatory factors of aging are evolutionally conserved, suggesting the implication of common processes/pathways in regulation of life span and aging in diverse organisms. For example, calorie restriction is a dietary regimen that is known to extend the life span and to increase stress resistance in organisms from yeast to mammals (7, 8). In yeast, a decrease in the glucose concentration of the culture medium extends the life span (9). This extension requires the activation of the silent information regulator (Sir) protein Sir2, a central determinant of yeast life span (9, 10). Sir2 is an NAD-dependent histone deacetylase (11–13) required for the chromatin silencing in mating-type loci HML and HMR (14), telomeres (15, 16), and the ribosomal DNA (rDNA)² locus RDN1 (17–19). Because deletion of either SIR2, SIR3, or SIR4 results in a shortened life span (20), the Sir complex functions to promote longevity in the wild-type cells.

Mutations in ZDS1 (zillion different screens) gene have been isolated in numerous genetic screenings for its ability to suppress the phenotypes caused by defects in various genes when it is overexpressed on a high-copy vector (21–23). The Zds1 functions reported include chromatin silencing, the establishment of cell polarity, cell cycle progression, and numerous other multiple processes (21–29). Moreover, Zds1 is also implicated in the regulation of aging (24). Although Zds1 (and its homolog Zds2) appears to be important in a wide range of cellular events, its biochemical function is not completely known.

We showed previously that ZDS1 null mutant budding yeast cells cultured in medium containing a high concentration of CaCl₂ were delayed in the G₂ phase and displayed polarized bud growth because of the activation of cellular Ca²⁺ signaling pathways (25). Two Ca²⁺-activated pathways, namely, the Mpk1 MAP kinase cascade and the calcineurin pathway, coordinately regulate the G₂/M cell cycle transition. To search for novel signaling components that mediate or affect the Ca²⁺-dependent regulation of cell growth and morphogenesis, we isolated mutant strains and classified them into 14 genetic complementation groups (designated scz1 to scz14 for suppressor of Ca²⁺-induced abnormalities of the zds1Δ strain). One of the identified mutants, scz14, was a mutant allele of sir3. In this study, we investigated a relationship between Ca²⁺ signaling and replicative life span (RLS) in yeast. Our results indicate that Ca²⁺ homeostasis/Ca²⁺ signaling are required for regulation of RLS.

EXPERIMENTAL PROCEDURES

Strains

Yeast strains DHT22-1b (MATa trp1 leu2 ade2 ura3 his3 can1–100), YAT1 (MATa zds1Δ::TRP1), YJY10 (MATa zds1Δ::TRP1 sir3Δ::kanMX4), YJY9 (MATa sir3Δ::kanMX4), YGA13 (MATa pmc1Δ::kanMX4), YGA14 (MATa pmr1Δ::kanMX4), DHT14 (MATa cnb1Δ::HIS3), YJY12 (MATa cnb1Δ::HIS3 sir3Δ::kanMX4), DMY2798 (MATa leu2::mURA3-LEU2), YSS3 (zds1Δ::KanMX4 in DMY2798), YGA4 (cnb1Δ::KanMX4 in DMY2798), YSS8 (sir3Δ::KanMX4 in DMY2798), YGA9 (cnb1Δ::HIS3 sir3Δ::KanMX4 in DMY2798), DMY2827 (sir2Δ::KanMX4 in DMY2798), DMY2800 (MATa NTS2::mURA3-LEU2), YSS5 (zds1Δ::KanMX4 in DMY2800), YGA6 (cnb1Δ::KanMX4 in DMY2800), YSS10 (sir3Δ::KanMX4 in DMY2800), YGA11 (cnb1Δ::HIS3 sir3Δ::KanMX4 in DMY2800), DMY2831 (sir2Δ::KanMX4 in DMY2800), DMY2804 (MATa NTS1::mURA3-LEU2), YSS4 (zds1Δ::KanMX4 in DMY2804), YGA5 (cnb1Δ::KanMX4 in DMY2804), YSS9 (sir3Δ::KanMX4 in DMY2804), YGA10 (cnb1Δ::HIS3 sir3Δ::KanMX4 in DMY2804), DMY2835 (sir2Δ::KanMX4 in DMY2804), DMY2895 (MATa adh4::URA3), YSS1 (zds1Δ::KanMX4 in DMY2895), YGA2 (cnb1Δ::KanMX4 in DMY2895), YSS6 (sir3Δ::KanMX4 in DMY2895), YGA7 (cnb1Δ::HIS3 sir3Δ::KanMX4 in DMY2895), DMY2841 (sir2Δ::KanMX4 in DMY2895), DMY2896 (MATa TEL VIIL::URA3), YSS2 (zds1Δ::KanMX4 in DMY2896), YGA3 (cnb1Δ::KanMX4 in DMY2896), YSS7 (sir3Δ::KanMX4 in DMY2896), YGA8 (cnb1Δ::HIS3 sir3Δ::KanMX4 in DMY2896), and DMY2839 (sir2Δ::KanMX4 in DMY2896) were the derivatives of strain W303-1A. Also used were YG880 (hmrΔE::TRP1 TEL VR::URA3 RDN1::ADE2-CAN1 his3 leu2), YG882A (zds1Δ::KanMX4 in YG880), YG884A (cnb1Δ::KanMX4 in YG880), YG883A (sir3Δ::KanMX4 in YG880), YG885A (cnb1Δ::HIS3 sir3Δ::KanMX4 in YG880), and YG881 (sir2Δ::KanMX4 in YG880).

Gene Disruption and Strain Construction

The zds1Δ, cnb1Δ, sir2Δ, and sir3Δ strains were constructed by gene replacement. Genomic DNA was isolated from the zds1::kanMX4, cnb1::kanMX4, sir2::kanMX4, and sir3::kanMX4 strains on a BY4741 background (Invitrogen). The amplified fragment was used to transform a strain. The oligonucleotide primer sequences used in this study are available on request.

Yeast RLS

RLS analyses were determined as described previously (30). Mother cells were discarded and buds were used as starting virgin cells. The RLSs of these cells were determined by noting and removing all subsequent daughters that were generated. All RLS analyses in this study were carried out on YPD plates with or without CaCl₂ at least three times independently. Statistical significance was determined by performing a log rank test. Average RLSs were taken to be significantly different at p < 0.05. Results from a single experiment are shown.

β-Galactosidase Assays

Cells carrying pKC201 (31) were grown to mid-log phase at 28 °C in defined minimal medium (synthetic complete (SC)) lacking uracil. The cells were harvested and resuspended in fresh YPD medium (pH 5.5). After incubation for 4 h at 28 °C with shaking, the cells were provided for assay of β-galactosidase activity as described previously (32). Error bars represent mean ± S.D. Significance levels for comparisons between the wild-type and other strains were determined with t tests.

Silencing Assays

Silencing assays were performed as described previously (33, 34). Cells were grown in nonselective liquid medium, and serial dilutions were spotted onto plates. The plates were incubated for 2–3 days at 25 °C and then photographed.

RESULTS

Deleting SIR3 Confers Ca²⁺ Resistance to zds1Δ Cells

The growth of ZDS1 null mutant (zds1Δ) yeast cells in medium containing 300 mm CaCl₂ is severely inhibited, with the cells exhibiting G₂ arrest and highly polarized bud growth. As shown in supplemental Fig. 1, the growth inhibition and polarized bud growth but not the G₂ delay of zds1Δ cells because of exogenous CaCl₂ were suppressed by an additional scz14 mutation. The plasmid that complemented the phenotype of the scz14 mutants was cloned, and the mutant gene was identified. Genetic linkage analysis showed that the scz14 mutation resided in the SIR3 gene (data not shown). The deletion of the SIR3 gene, similar to the scz14 mutation, suppressed the growth inhibition and the polarized bud growth of zds1Δ cells, suggesting that the scz14 mutation is a loss of function allele of the SIR3 gene (Fig. 1, A–C). Therefore, we used the sir3 deletion (sir3Δ) in further experiments.

FIGURE 1.

The deletion of the SIR3 gene suppresses various phenotypes of the zds1Δ strain. A, effect of the deletion of the SIR3 gene on the growth of the zds1Δ mutant strain on solid medium. WT, sir3Δ, zds1Δ, and zds1Δ sir3Δ cells were spotted on YPD solid medium containing 300 mm CaCl₂, after which the plates were incubated at 25 °C for 2 days. B and C, cell morphology after 6 h of incubation with 100 mm CaCl₂ at 25 °C (B). Flow cytometry analysis (FACS) of propidium iodide-stained cells of various strains after 6 h of incubation with 100 mm CaCl₂ at 25 °C (C). DIC, differential interference contrast; 1C, one DNA copy; 2C, two DNA copies. D, Cln2-HA and Cdc28 were detected by Western blotting using early log phase growing cells of those strains indicated in A, suspended in YPD containing 100 mm CaCl₂, incubated for 6 h, and then used for Western blotting. Cln2-HA or Cdc28 was detected by immunoblotting with anti-HA or anti-PSTAIRE antibody, respectively.

Why did the sir3Δ mutation suppress only the Ca²⁺-induced hyperpolarized bud growth but not the G₂-delay? We showed previously that the Ca²⁺ signal induces polarized bud growth by elevating the level of G₁-cyclin Cln2 (35). To investigate whether the sir3 deletion affected the amount of Cln2, we examined the protein level of Cln2 by performing a Western blot analysis. For this purpose, we used cells containing the chromosomally integrated constructs for HA-tagged Cln2. The Cln2 levels in the presence and absence of external CaCl₂ were decreased by the sir3Δ (compare zds1Δ and zds1Δ sir3Δ in Fig. 1D). The suppression of the Ca²⁺ sensitivity of the zds1Δ strain by the sir3 deletion was thus caused by a defect in the elevation of the Cln2 level in response to Ca²⁺. Basically, a similar effect of the sir3Δ disruption on the Cln2 level was seen on the WT background (compare WT and sir3Δ in Fig. 1D). These results indicate that the sir3Δ caused a defect in the elevation of the Cln2 level. The Sir3 protein in budding yeast is required for telomere silencing (15) and affects the life span (20, 24, 36). Therefore, in this study, we investigated a relationship between Ca²⁺ signaling and life span.

Increase in External and Cellular Ca²⁺ Levels Causes a Shortened RLS

We first investigated whether the level of external Ca²⁺ would affect the RLS. Yeast cells undergo asymmetric divisions, producing a smaller daughter cell from a larger mother cell, and yeast aging is measured as a RLS, i.e. the number of times a mother cell can divide before it dies (37–39). Cells grown in the YPD medium supplemented with 100 mm CaCl₂ had a shortened RLS relative to those grown on YPD (Fig. 2A). To check the specificity of Ca²⁺ on the RLS, we examined the effect of MgCl₂ and KCl on the RLS because some cations (such as Na⁺, Li⁺, Mn²⁺, Co²⁺, and Ni²⁺) but not Mg²⁺ and K⁺ have been shown to affect the Ca²⁺ homeostasis (40, 41). As the result, neither MgCl₂ (100 mm) nor KCl (150 mm) altered the RLS of WT cells significantly (Fig. 2A). These results established that external Ca²⁺ reduced RLS.

FIGURE 2.

An increase in Ca²⁺ levels led to a reduced RLS. A, RLS analysis of WT cells grown on YPD (control), YPD plus 100 mm CaCl_2, YPD plus 100 mm MgCl₂, and YPD plus 150 mm KCl at 25 °C. The mean RLS in terms of numbers of cell divisions are control, 27.4 (n = 52); 100 mm CaCl₂, 22.5 (n = 36); 100 mm MgCl₂, 29.8 (n = 52); and 150 mm KCl, 28.3 (n = 52). B, RLS analysis of WT and zds1Δ mutant cells grown on YPD at 25 °C. The mean RLSs are WT, 24.3 (n = 43) and zds1Δ, 19.6 (n = 44). C, expression of the calcineurin-dependent response element-dependent reporter gene in WT, zds1Δ, zds1Δ cnb1Δ, pmc1Δ, pmr1Δ, and cnb1Δ mutant cells. *, p < 0.001. D, RLS analysis of WT, pmc1Δ, and pmr1Δ mutant cells grown on YPD at 25 °C. The mean RLSs are WT, 27.5 (n = 47); pmc1Δ, 26.3 (n = 44); and pmr1Δ, 9.4 (n = 49).

It is possible that external Ca²⁺ causes the shortened RLS by activating intracellular Ca²⁺ signaling. To investigate this possibility, we measured the mean RLS of zds1Δ cells, in which Ca²⁺ signaling pathways might be activated by an increase in internal Ca²⁺ level (25). As expected, the mean RLS of zds1Δ cells was shorter than that of WT cells under the experimental conditions (Fig. 2B), although it was reported previously that the mean RLS of zds1Δ cells was longer than that of WT cells (24).

To investigate the activation of Ca²⁺ signaling pathways in zds1Δ cells under the experimental conditions, we estimated the calcineurin activity in vivo by using a reporter assay system (31). In response to the elevation of the cellular Ca²⁺ level, the transcription factor Crz1 binds to the calcineurin-dependent response element in the promoter and activates several stress-responsive target genes. We used the calcineurin-dependent response element-driven LacZ reporter gene (31) and measured the calcineurin activity in the WT and zds1Δ cells under normal growth condition (YPD) by β-galactosidase assay. The β-galactosidase activity in WT cells indicated the basal level of this reporter system because the activity in WT cells was comparable with that in the calcineurin-deleted cells (cnb1Δ) (Fig. 2C). As expected, β-galactosidase activity in the zds1Δ cells increased significantly by 3-fold compared with that in WT cells (Fig. 2C), and this increase was indeed dependent on calcineurin (Fig. 2C, compare zds1Δ and zds1Δ cnb1Δ). These results indicate that the increase in cellular Ca²⁺ level in the zds1Δ cells causes the shortened RLS by activating Ca²⁺ signaling pathway(s).

Increase in Cellular Ca²⁺ Level by PMR1 Deletion Shortened RLS

We previously isolated two genes, PMR1 (42, 43) and PMC1 (44, 45), encoding Golgi and vacuolar Ca²⁺-ATPases, respectively, as multicopy suppressors of CaCl₂ sensitivity of the zds1Δ cells (46). Both Pmr1 and Pmc1 transport Ca²⁺ from the cytosol into internal compartments. In addition, it has been reported that a lack of Pmr1 results in alterations in cellular Ca²⁺ homeostasis, including an increased rate of cellular Ca²⁺ uptake from the extracellular environment and an enhanced sensitivity to high extracellular Ca²⁺ levels (47, 48).

To confirm that an increase in cellular Ca²⁺ shortens the RLS in yeast, we examined RLS of the pmr1- and pmc1-deleted cells. The pmr1Δ cells but not the pmc1Δ showed a severe reduction in mean RLS of ∼ 64% (Fig. 2D), indicating that Pmr1-mediated Golgi Ca²⁺ sequestration but not a Pmc1-mediated vacuolar one is important for longevity under this experimental condition. To further to investigate the contribution of these Ca²⁺-ATPases in cellular Ca²⁺ homeostasis, we measured the calcineurin activity in these cells with the reporter assay. Consistent with the result of RLS, β-galactosidase activity in the pmr1Δ cells but not in the pmc1Δ significantly increased by 4-fold compared with the basal level in the WT cells (Fig. 2C). These results indicate that Pmr1-mediated Ca²⁺ sequestration to the Golgi plays an important role in regulation of RLS through cellular Ca²⁺ homeostasis.

Hyperactivation of Calcineurin Reduces the RLS

To further confirm the above suggestion that activation of Ca²⁺ signaling pathways shortens the RLS, we investigated whether overexpression of the constitutively activated calcineurin (CMP2ΔC, C-terminal autoinhibitory domain-truncated catalytic subunit) promotes aging and shortens the RLS. This was indeed the case. The WT cells expressing the activated calcineurin (YEp24-CMP2ΔC) showed a reduced RLS compared with the WT cells carrying the empty plasmid (YEp24) (Fig. 3A). This result established that activation of calcineurin signaling shortens the RLS.

FIGURE 3.

Effect of overexpression and deletion of calcineurin on the RLS. A, RLS analysis of the cells overexpressing the constitutively active form of calcineurin. WT cells transformed with empty plasmid (YEp24) or YEp24-CMP2ΔC plasmid grown on YPD at 25 °C were examined. Mean RLSs are YEp24, 25.5 (n = 40) and YEp24-CMP2ΔC, 21.9 (n = 39). B, RLS analysis of WT, zds1Δ, cnb1Δ, and zds1Δ cnb1Δ mutant cells grown on YPD at 25 °C. The mean RLSs are WT, 26.9 (n = 44); zds1Δ, 21.8 (n = 44); cnb1Δ, 23.5 (n = 44); and zds1Δ cnb1Δ, 27.6 (n = 43). C, RLS analysis of WT, cnb1Δ, sir3Δ, and cnb1Δ sir3Δ mutant cells grown on YPD at 25 °C. The mean RLSs are WT, 25.2 (n = 18); cnb1Δ, 19.7 (n = 33); sir3Δ, 19.6 (n = 20); and cnb1Δ sir3Δ, 15.0 (n = 36).

Calcineurin Deletion Overcomes the Shortened RLS of zds1Δ Cells

We reported previously that calcineurin and Zds1 play an antagonistic role in the regulation of cell growth and morphogenesis in the presence of high Ca²⁺ (25). As shown in Fig. 2B, deletion of ZDS1 resulted in about 20% decrease in the mean RLS with hyperactivation of calcineurin. We investigated whether calcineurin deletion overcomes the shortened RLS in zds1Δ cells. As expected, the mean RLS of zds1Δ cnb1Δ double deletion cells was longer than that of zds1Δ single deletion cells (Fig. 3B). Interestingly, the mean RLS of zds1Δ cnb1Δ cells was comparable with that of the WT cells. These results indicate that calcineurin and Zds1 play an antagonistic role in the regulation of RLS and that the activated calcineurin promotes aging in the zds1Δ cells.

Calcineurin and Sir3 Act Redundantly in the Regulation of RLS

We next examined the effect of the calcineurin deletion on the RLS of WT cells. Surprisingly, the calcineurin deletion per se on the WT background reduced the RLS under normal growth conditions (Fig. 3, B and C), suggesting that the basal activity of calcineurin is required for the maintenance of the RLS under this condition.

In the Sir3-mediated regulation of RLS, the Mpk1 MAPK cascade was identified as the pathway that phosphorylates Sir3, which leads to the shortening of the RLS (49). We and others showed earlier that calcineurin and Mpk1 act redundantly in cellular events (25, 50, 51). To investigate whether calcineurin is related to the Sir3-mediated regulation of RLS, we examined the RLS of the double deletion mutant cnb1Δ sir3Δ. As reported previously (20), deletion of SIR3 resulted in a 22% decrease in the mean RLS (Fig. 3C). The mean RLS of cnb1Δ sir3Δ double deletion cells was shortened more compared with that of each single deletion mutant (Fig. 3C). Thus, it appears that calcineurin and Sir3 act redundantly in the regulation of the RLS as parallel pathways.

Calcineurin Does Not Affect Silencing at the rDNA, Telomeres, or Mating-type Locus

To further seek the calcineurin-related regulation of RLS, we examined whether calcineurin is involved in gene silencing. Silencing at NTS1 has been proposed to counteract rDNA recombination, a phenomenon that has been implicated in yeast aging (33). In budding yeast, each 9.1-kb repeat yields a 35 S precursor rRNA (transcribed by RNA polymerase (Pol) I) and a 5 S rRNA (transcribed by RNA Pol III), separated by two nontranscribed spacers, NTS1 and NTS2 (52). To investigate whether calcineurin affects silencing at the rDNA, we constructed the CNB1-deleted cells carrying an mURA3 reporter gene integrated into one of the following sites: outside of the rDNA array at the non-silencing LEU2 locus (euchromatic locus) or within the rDNA unit at the strong silencing region (NTS1 or NTS2 locus) (34). As reported (33, 34), the reporter gene was strongly silenced at either the NTS1 or NTS2 sites in WT cells, as indicated by poor growth on SC minus uracil (-Ura) medium compared with the same reporter inserted at the euchromatic locus (Fig. 4A). As reported previously, defective silencing was observed in the sir2-deleted cells (Fig. 4A). No change was observed in the growth of cnb1Δ cells on -URA medium plates as in WT cells (Fig. 4A), indicating that calcineurin did not affect the silencing at a rDNA region.

FIGURE 4.

Effect of the deletion of calcineurin (cnb1Δ) on silencing at rDNA, telomeres, and the mating-type locus. A, silencing at rDNA. 5-fold serial dilutions of cultures of the indicated strains were spotted onto -URA medium. The synthetic complete medium (Complete) was using strains containing an mURA3 insertion at outside rDNA, NTS1, or NTS2. SIR2 is required for silencing at NTS1 and NTS2. B, silencing at telomeres. 5-fold serial dilutions of cultures of the indicated strains were spotted onto synthetic complete medium containing 5-fluoroorotic acid (+5-FOA) to assess telomeric silencing. C, silencing at mating-type locus. 5-fold serial dilutions of cultures of the indicated strains were spotted onto synthetic complete medium lacking tryptophan (-TRP).

Next, to investigate whether calcineurin was required for silencing at other heterochromatic regions, we examined silencing at telomeres and the mating-type locus in the cnb1-deleted cells. We constructed the CNB1-deleted cells, in which a URA3 reporter gene was integrated into the telomeric repeats of Chromosome VIIL or the non-silencing ADH4 locus as the negative control. Cells were spotted onto complete medium as a plating and growth control or onto medium supplemented with 5-fluoroorotic acid, a drug that is toxic to URA3-expressing cells (53). Both the WT and the cnb1Δ cells were able to grow on the 5-fluoroorotic acid medium, whereas the sir2Δ and sir3Δ strain, in which silencing fails and URA3 is expressed, did not grow at all (Fig. 4B). Further, to monitor silencing of the mating-type locus, we used a strain with a TRP1 marker integrated into the HMR locus. Similar to the effect at other regions (rDNA and telomeres), the silencing at the mating-type locus was normal in the Δcnb1 cells (Fig. 4C). These results indicate that calcineurin is not important for these genes silencing and plays an important role in a novel regulation of the RLS.

DISCUSSION

In conclusion, we reported that calcineurin plays an important role in regulating the RLS/aging in budding yeast. Hyperactivation of calcineurin signaling reduced the RLS. Indeed, the shortened RLS of zds1Δ cells, in which calcineurin is activated by the increase in cellular Ca²⁺ level, was suppressed by the calcineurin deletion. These results indicate that the constitutive activation of calcineurin would be harmful to yeast in the long term, although it could enhance resistance to several stresses, including ionic stress (Na⁺, Li⁺, Mn²⁺), high pH, and the presence of cell wall-disrupting compounds (calcofluor-white and Congo red). Surprisingly, loss of function of the calcineurin gene also promoted the aging in the WT cells, indicating that basal activity of calcineurin is required for maintenance of RLS under normal growth conditions. However, in the multicellular organism Caenorhabditis elegans, a calcineurin-deficient mutant displayed an extended lifespan (5, 6). It is possible that an optimal level of calcineurin activity/expression would be important for longevity in the unicellular organism yeast.

How do calcineurin and Zds1 act in the regulation of the RLS? Our results suggested that calcineurin and Zds1 play an antagonistic role in the regulation of RLS. We reported previously that the polarized bud growth, the defect of cell proliferation, and the G₂ delay caused by high calcium in the zds1Δ strain were abolished by the calcineurin deletion mutations (25). In addition, calcineurin activity was enhanced in the zds1Δ cells (Fig. 2D). Therefore, it is possible that Zds1 may be required for the longevity in the negative regulation of calcineurin. This possibility needs to be studied further.

In this study, we showed that the zds1Δ deletion reduced the RLS. However, previously, the another group reported that the zds1Δ cells led to an extended RLS (24). Although both studies used the W303 strain, the mean RLS of their WT cells (∼20 times) was much shorter than that of our WT cells (25∼27 times), indicating that both experimental conditions are different from each other. The contradiction of RLS in the zds1Δ cells would be caused by the differences in the experimental conditions. Further, it was reported that the zds1Δ strain on the BY4742 background had no effect on the RLS (54). We noted that the zds1Δ cells on the W303 background were more sensitive to Ca²⁺ than those on the BY4742 background.³ It is possible that the short-lived zds1Δ cells on W303 were due to the elevated level of intracellular Ca²⁺.

Why did the pmr1Δ cells exhibit a severe reduction in mean RLS compared with other cells (zds1Δ, cnb1Δ, and sir3Δ)? The Pmr1 protein plays an important role in both secretion and proton transport (55, 56). Although the gene that is related to the secretary pathway has not been identified yet in the systematic genetic screening for the RLS regulator, it has been reported that some proton ATPase mutants, such as the VMA genes encoding vacuolar H₁-ATPases, reduced the RLS (57). These results suggest that the severe RLS reduction of pmr1Δ cells was derived not only from alternation of the Ca²⁺ homeostasis but also from defect of proton transport. As the double deletion between calcineurin and the VMA gene shows synthetic lethality (51, 58), the shortened RLS of the calcineurin deletion might be caused at least in part by the impairment in control of the intracellular pH. Although a functional interaction between secretion and RLS has not been shown, we have not ruled out this possibility.

We showed previously that calcineurin leads to the Cln2 up-regulation by a mechanism that is mediated by the degradation of Yap1 (46). We showed that Sir3 was required for the elevation of the Cln2 level. In the regulation of the RLS, a functional link between Sir3 and Cln2 has not yet been identified. However, our results suggest the possibility that the optimum levels of Cln2 might be important for longevity. Our findings also suggest that Ca²⁺ homeostasis is necessary for longevity in yeast.

It will be interesting to determine whether altered Ca²⁺ homeostasis influences the life span of higher eukaryotes and whether the corresponding pathways are a part of a conserved network involved in the control of biological aging.