A novel gene, ecl1⁺, extends the chronological lifespan in fission yeast

Abstract

We have identified a novel gene from Schizosaccharomyces pombe that we have named ecl1⁺ (extender of the chronological lifespan). When ecl1⁺ is provided on a high-copy number plasmid, it extends the viability of both the Δsty1 MAP kinase mutant and the wild-type cells after entry into the stationary phase. ecl1⁺ encodes an 80-amino acid polypeptide that had not been annotated in the current database. The ecl1⁺-mRNA increases transiently when the growth phase is changed from the log phase to the stationary phase. The Ecl1 protein is localized in the nucleus. Calorie restriction extends the chronological lifespan of wild-type and Δecl1 cells but not ecl1⁺-overproducing cells. The Δpka1 mutant shows little, if any, additional extension of viability when Ecl1 is overproduced. The ste11⁺ gene that is negatively controlled by Pka1 is up regulated when Ecl1 is overproduced. From these results we propose that the effect of Ecl1 overproduction may be mainly linked to and negatively affects the Pka1-dependent pathway.

chronological lifespan, fission yeast, aging

Introduction

The chronological lifespan of yeast cells is measured by viability after entry into the stationary phase. We determined previously that the mutant of lcf1⁺, which encodes a long-chain fatty acyl-CoA synthetase, showed rapid loss of viability after entry into the stationary phase, and suggested that fatty acid utilization and/or metabolism is important to determine viability in the stationary phase in fission yeast (Oshiro et al., 2003; Fujita et al., 2007). Pka1 and Sck2 are regulators of chronological lifespan in fission yeast; each mutant showed a long-lived phenotype, and the pka1 and sck2 double mutant displayed an additive effect on chronological lifespan extension, suggesting that these two factors regulate related but independent pathways (Roux et al., 2006).

In Saccharomyces cerevisiae, two of the major pathways that control the chronological lifespan have been identified: the Ras/PKA/Msn2/4 pathway and the Sch9 pathway (Fabrizio et al., 2001; Fabrizio & Longo et al., 2003). The down-regulation of either pathway promotes lifespan extension. Importantly, similar pathways (insulin/IGF-I-like) regulate longevity in higher eukaryotes, suggesting a common evolutionary origin for the life span-regulatory mechanisms (Longo & Fabrizio et al., 2002; Longo & Kennedy et al., 2006). However, the molecular mechanism responsible for the regulation of lifespan has not yet been elucidated.

In the present study, we used Schizosaccharomyces pombe as a model system to study the chronological lifespan, and we screened for a gene that, when cloned on a high-copy number plasmid, could extend cell viability after entry into the stationary phase.

Materials and methods

Strains and media

Schizosaccharomyces pombe strain JY333 (h⁻ leu1-32 ade6-M216), SP14000 (h⁻ leu1-32 ade6-M210 ura4-D18), PR 109 (h⁻ leu1-32 ura4-D18) and TK107 (h⁻ leu1-32 ura4-D18 phh1::ura4⁺) were used. The strains were grown in SD medium [0.67% yeast nitrogen base without amino acids (Difco), 2% glucose] supplemented with necessary growth requirements in standard amounts. For caloric restriction, SD low-glucose medium [0.67% yeast nitrogen base without amino acids (Difco), 0.5% glucose] was used. EMM medium (Moreno et al., 1991) was also used.

Analysis of viability

To determine cell viability, cells were grown in SD liquid medium or EMM liquid medium, sampled in each growth phase, and then plated on yeast extract peptone dextrose agar plates after suitable dilution. After 7 days of incubation at 30 °C, we counted the number of colonies derived from 1 mL of culture on a plate and then divided the number by the cell turbidity at the sampling time. Cell growth was monitored by turbidity using a Bactomonitor (BACT-550) equipped with a 600-nm filter (Nissho Denki Co., Tokyo, Japan). For the calorie-restriction experiment, overnight cultures grown in SD liquid medium were transferred into an SD low-glucose medium, and the viability was determined as described above. To determine the viability in water, cells were grown in SD liquid medium for 24 h (OD_{600 nm}=2.5), collected, and then suspended in H₂O at a cell turbidity around OD_{600 nm}=2.5. Cell viability was determined as described above. All data are represented as mean of at least two independent experiments.

Construction of Δecl1, Δpka1, and Δsck2 mutants

For the ecl1⁺ disruption, the ORF region of the Ecl1 protein was replaced with a kan^R cassette using the methods described by Krawchuk & Wahls (1999). Both the upstream and the downstream regions of ecl1⁺ were PCR-amplified using F1 and F2 primers and R1 and R2 primers, respectively, and both fragments were purified. After mixing both DNA fragments with pFA6a-kanMX6, a PCR reaction was performed with the F1 and R1 primers. JY333 was transformed with the amplified DNA fragment, and stable G418-resisitant transformants were selected. Then, the ecl1::kan^R construct on the chromosome was confirmed by PCR using appropriate primers. The primers used were: F1, GATCTATCACACTATATATA; F2, TTAATTAACCCGGGGATCCGCGAGTGAGTAAACTAACAAG; R1, CCGCGGATTAAATTGCTGGT; and R2, GTTTAAACGAGCTCGAATTCTCTACTTAGAATCGGTTTAT.

Δpka1 and Δsck2 mutants were constructed as described above. The primers used for the construction of Δpka1 were: F1, CTTTGAAGGACTCAGAGTCG; F2, TTAATTAACCCGGGGATCCGGCAACGGCAGTCGTATCCAT; R1, TAGGACCGCGATGATTAAGG; and R2, GTTTAAACGAGCTCGAATTCTGGAACATCTGAAGATCCTG. The primers used for the construction of Δsck2 were: F1, TAGGTACTGCAACCCACTTC; F2, TTAATTAACCCGGGGATCCGGACAGCCCGAACCAATTTTTTG; R1, CTACTAATGTCAGCCAATTCAACG; and R2, GTTTAAACGAGCTCGAATTCCGTGCGAACGATGAGGTATT.

Plasmid construction

The plasmid, pSht3, that was used for overexpression of ecl1⁺ was described in Nakamichi (2000). In brief, the 1.34-kb DNA fragment shown in Fig. 2 was cloned in the vector plasmid pLDblet (Yamada et al., 1997). To introduce a stop codon into Ecl1 protein, site-directed mutagenesis was performed on the pSht3 plasmid using the site-directed mutagenesis primer: CTGCCATCCTAAGTGGTGGCT. To make the Ecl1-EGFP fusion construct, a DNA fragment encompassing the ecl1⁺ gene and its upstream region was amplified from pSht3 using two primers, GTAATACGACTCACTATAGGGC and TTCTAAGTCGACATCGCATGTAGTTTTGTGTAAGCAC. The amplified DNA fragment was digested with PstI and SalI and then cloned into the same sites of plasmid pREP41-EGFP-C to construct pREP41-Ecl1-EGFP (Craven et al., 1998). In the plasmid pREP41-Ecl1-EGFP, the nmt1 promoter was deleted and the ecl1-EGFP fusion gene was transcribed by the natural promoter of the ecl1⁺ gene.

2

Nucleotide sequence of a cloned DNA fragment in pSht3 and a possible translated protein, Ecl1. The 1.34-kb nucleotide sequence cloned in pSht3, which corresponds to nucleotide positions 9960 to 11296 of the sequence deposited in GenBank (Accession no. AL023794), is shown. A possible translated product, Ecl1 protein, is shown below the nucleotide sequence. The amino acid position that was mutated to make a stop codon (Gln 14) is double underlined. The canonical promoter (TATA) sequence was underlined.

To make plasmid pREP1-Ecl1, the ORF region of Ecl1 was amplified using two primers: CTCACCATATGGATTTGGATTTTTGCACAGTG and CTAAGGATCCTACGCATGTAGTTTTGTGTAAGC. After digestion with NdeI and BamHI, the fragment was cloned into the same sites of plasmid pREP1 (Maundrell et al., 1993).

Fluorescence microscopy of EGFP-tagged Ecl1 protein

Wild-type cells (JY333) carrying the pREP41-Ecl1-EGFP plasmid were grown in SD or EMM medium. Green fluorescent protein (GFP) fluorescence in living cells was monitored using an Eclipse E600 microscope (Nikon) with an appropriate set of filters.

Northern hybridization of ecl1⁺ mRNA

Northern hybridization analysis was carried out as described previously (Aiba et al., 1995). Cells were grown in SD or EMM medium. In the logarithmic growth phase, cells were collected and a total RNA fraction was prepared. After denaturation with formamide-formaldehyde, the RNA (10 μg) was analyzed on a 1.4% agarose gel containing formaldehyde, followed by alkali-blotting onto Hybond-N+(Amersham Int). Hybridization was carried out with a³²P-labelled probe, which specifically encompasses the ecl1⁺ coding sequence or the ste11⁺ coding sequence at 65 °C for two hours in Rapid-hyb buffer, as recommended by the supplier (Amersham Int.).

Results

Screening of genes suppressing rapid loss of viability in the stationary phase of a Sty1–MAP kinase–defective mutant

Sty1-MAP kinase is part of a stress-activated MAP kinase pathway consisting of Wis1(MAPKK)-Sty1(MAPK). The Δsty1 mutant has a pleiotropic phenotype that includes stress sensitivity to osmotic, oxidative, heat, and UV stresses, G2/M cell cycle delay, and decreased viability after entry into the stationary phase (Wilkinson & Millar et al., 1998). Because little is known about the mechanism of the rapid loss of viability in the stationary phase in the Δsty1 mutant, we screened for a plasmid clone that could suppress this phenotype. First, we analyzed the suppressing ability of some plasmid clones in our laboratory stock to determine whether they could improve the reduced viability of the Δsty1 mutant. Previously we screened an S. pombe genomic library for a plasmid that could suppress the heat-stress sensitivity of the Δsty1 mutant (Yamada et al., 1997). Among the suppressing plasmids, pSDS23, which encodes protein that functions as a multicopy suppressor for mutations of the PP1 protein phosphatase and the 20S cyclosome/anaphase-promoting complex (APC), and pHXK2, which encodes hexokinase 2, were revealed not to suppress the short-lived phenotype of the Δsty1 mutant (data not shown). By chance, we found that when plasmid pSht3 was introduced into the Δsty1 mutant, it suppressed its short-lived phenotype (Fig. 1a). pSht3 had been identified as a multicopy suppressor plasmid for the high-osmolarity-sensitive phenotype of the hos2 mutant (Nakamichi et al., 2000). Mutations in the hos2⁺ and the hos3⁺ genes cause a high-osmolarity-sensitive phenotype (Aiba et al., 1998; Aoyama et al., 2000). Both the hos2⁺ and hos3⁺ genes encode small proteins of 94 amino acids, and they are members of the DASH complex, which is involved in chromosome attachment (Aoyama et al., 2000; Nakamichi et al., 2000; Liu et al., 2005; Sanchez-Perez et al., 2005).

1

Identification of plasmid that increases cell viability after entry into the stationary phase. Cell viability (upper panels) and cell growth (lower panels) of Δsty1 mutant cells in SD medium (column a), wild-type cells in SD medium (column b), wild-type cells in EMM medium (column c), or wild-type cells in water (column d) carrying the vector plasmid, pLB-Dblet (○) or pSht3 (●), were monitored.

The Δsty1 cells display several phenotypes including temperature sensitivity at 37 °C, high-osmolarity sensitivity (0.9 M KCl or 2 M Sorbitol), elongated cell morphology indicative of delayed G2/M cell cycle progression, and 2 mM H₂O₂ sensitivity. Transformation of the plasmid pSht3 did not suppress these Δsty1 phenotypes (data not shown).

Next, we introduced pSht3 into wild-type cells (JY333), and the viability after entry into the stationary phase was monitored. We found that pSht3 extends viability after entry into the stationary phase for not only the Δsty1 mutant (TK107) but also for the wild-type cells (Fig. 1b). We introduced pSht3 into other wild-type strains carrying different genetic backgrounds (PR109 and SP14000) and confirmed that pSht3 also extends the viability of these cells (data not shown).

In our assay, we used synthetic dextrose (SD) medium because it had been used for the chronological lifespan assay of S. cerevisiae (Fabrizio et al., 2001) and we had used it for the screening of mutants (Oshiro et al., 2003). But the survival of wild-type cells at the stationary phase in the SD medium seemed to be short. Hence, to confirm that pSht3 increases the chronological survival in other growth conditions, we analyzed the survival of the cells in Edinburgh minimal medium (EMM) medium and in water. As shown in Fig. 1c, we found that, in the EMM medium, wild-type cells retained their viability longer than in the SD medium. Under this growth condition, pSht3 extended the viability of the cells at the stationary phase (Fig. 1c). We also analyzed the viability of cells in water. To carry out this analysis, cells grown in SD medium were collected, and then suspended in water. As shown in Fig. 1d, the viability of wild-type cells carrying pSht3 was higher than that carrying the vector plasmid. From these results, we concluded that pSht3 extends the chronological lifespan of S. pombe.

Identification of the ORF responsible for extending the chronological lifespan of S. pombe

pSht3 contains a 1.34-kb segment derived from chromosome III (GenBank Accession no. AL023794; nucleotide position 9960–11296). In the database, this particular region was noted to be ‘a low-complexity gene-free region’, and in fact contains only a few, very short ORFs. We performed a series of experiments to identify the gene responsible for the extension of viability after entry into the stationary phase (i.e., chronological lifespan).

First, we identified an ORF in this region that could encode an 80-amino acid protein (Fig. 2). To determine whether this ORF is responsible for the phenotype, we cloned this ORF into the expression vector pREP1. The resultant plasmid (pREP1-Ecl1) was designed to express the 80-amino acid protein under the control of the nmt1 promoter. Wild-type cells transformed with pREP1-Ecl1 retained their viability longer than cells transformed with vector plasmid, pREP1 (Fig. 3a). This result indicates that the ORF that encodes the 80-amino acid protein is sufficient for extending the chronological lifespan of S. pombe.

3

The ecl1⁺ gene is responsible for increased cell viability after entry into the stationary phase. (a) Cell viability (upper panel) and cell growth (lower panel) of wild-type cells (JY333) carrying the vector plasmid, pREP1 (○) or pREP1-Ecl1 (●), are shown. (b) Cell viability (upper panel) and cell growth (lower panel) of wild-type cells (JY333) carrying vector plasmid, pLB-Dblet (○), pSht3 (●), or pSht3-Q14stop (Δ), are shown.

To verify that the 80-amino acid protein encoded by the ORF is indeed responsible for the phenotype, we introduced a nonsense mutation (Gln-14 to stop) in the ORF of pSht3. As shown in Fig. 3b, the plasmid carrying the nonsense mutation (pSht3-Q14stop) did not extend the chronological lifespan. We also confirmed that pSht3-Q14stop, but not pSht3, did not suppress the high-osmolarity-sensitive phenotype of the hos2 mutant (Nakamichi et al., 2000 and data not shown).

To detect the transcript for the ORF, Northern hybridization analysis was carried out. As shown in Fig. 4a, c. a 1.6 kb of transcript that was not detected in the deletion mutant (Δecl1 mutant) was detected. Canonical TATA sequence could be assigned upstream of the ORF (Fig. 2). Next, to determine whether the protein was translated in vivo, we fused the EGFP protein to the C terminus of the canonical 80-amino acid protein, and the fusion protein was detected by Western blotting analysis using an anti-GFP antibody. We detected a 35-kDa fusion protein (Fig. 4b). This size is consistent with the calculated molecular weight of the fusion protein (Ecl1: 9 kDa+EGFP: 27 kDa). These results suggest that the canonical protein was expressed in vivo.

4

Expression of the ecl1⁺gene. (a) Total RNA were isolated from the indicated strains, and then subjected to Northern hybridization analysis using a probe for ecl1⁺. In the lower panel, the ethidium bromide stained agarose gel is shown as a control for the amounts of RNA loaded. (b) Wild-type cells carrying vector plasmid (pREP1) (lane 1), pREP41-Ecl1-EGFP (lane 2), or pREP41-EGFP-C (lane 3) were grown in SD medium, and the Ecl1-EGFP fusion protein was detected by Western blotting analysis using an anti-GFP antibody. (c) Wild-type cells (JY333) carrying pREP1 (marked as vector), pREP41-Ecl1-EGFP (marked as Ecl1-GFP), or pREP41-EGFP-C (marked as GFP) were grown in SD medium until the mid-log phase, and cells were observed under fluorescence microscopy.

Based on these results, we concluded that we identified a new gene that encodes an 80-amino acid protein. We named this gene ecl1⁺ (extender of the chronological lifespan).

Ecl1 protein is localized to the nucleus

To analyze the cellular localization of the Ecl1 protein, the localization of Ecl1-EGFP fusion protein was determined by fluorescent microscopy. For this analysis, we used the Ecl1-EGFP fusion construct used in Fig. 4b. We confirmed that the plasmid pREP41-Ecl1-EGFP, which carries the Ecl1-EGFP fusion construct, extends the chronological lifespan of wild-type cells (data not shown). This observation indicates that the addition of EGFP to the C terminus of Ecl1 protein does not interfere with its function. As shown in Fig. 4c, the Ecl1-EGFP fusion protein was localized mainly in the nucleus (also see Fig. 6).

6

Expression of Ecl1-EGFP protein decreased in the stationary phase. (a) Wild-type cells (JY333) carrying pREP41-Ecl1-EGFP were grown in EMM medium and the localization of the Ecl1-EGFP protein was observed under fluorescence microscopy in the log phase (OD_{600 nm}=0.95) or the stationary phase (OD_{600 nm}=3.5). DAPI-stained images of the cells are also shown. (b) Cell lysate was prepared from equivalent number of cells in the log phase (OD_{600 nm}=0.95, lane 1) or the stationary phase (OD_{600 nm}=3.5, lane 2) and subjected to Western blotting analysis using an anti-GFP antibody to detect the Ecl1-EGFP protein.

Regulation of ecl1⁺ expression

Because the effect of overexpression of the ecl1⁺ gene conferred a striking phenotype, particularly in the stationary growth phase, we examined whether or not this gene is predominantly expressed in the stationary phase. The wild-type cells were grown in EMM medium, and then the expression profile of the ecl1⁺ mRNA was followed along with growth by means of Northern hybridization. As shown in Fig. 5, the expression of ecl1⁺ mRNA up-regulated transiently when the growth phase was changed from the log phase to the stationary phase. After the up-regulation, the amount of ecl1⁺ mRNA decreased gradually.

5

Northern blot analysis of the ecl1⁺ transcript with growth. (a) Growth curve of wild-type cells (JY333) in EMM medium at 30°C. (b) Total RNAs were isolated from cells at each time point indicated in panel (a) and subjected to Northern blotting analysis with a radiolabelled ecl1⁺ probe. In the lower panel, the ethidium bromide-stained agarose gel is shown as control for the amounts of RNA loaded.

Next, we analyzed whether the expression and localization of Ecl1-EGFP protein is affected by growth phase. The cell carrying pREP41-Ecl1-EGFP was grown in EMM medium and the localization of Ecl1-EGFP protein was analyzed in the log and stationary phases. As shown in Fig. 6a, Ecl1-EGFP protein was mainly localized in nucleus at the log phase (also see Fig. 4c), but the expression and localization of the protein was not evident at stationary phase. Then we next analyzed the amount of Ecl1-EGFP protein at the log and stationary phases by means of Western blotting analysis. As shown in Fig. 6b, the amount of Ecl1-EGFP protein, that is present in the log phase decreased in the stationary phase.

These results suggested that the expression of ecl1⁺-mRNA (and maybe the stability of Ecl1 protein) is regulated by growth phase.

Calorie restriction extends the chronological lifespan of wild-type cells but does not extend the long-lived phenotype conferred by Ecl1 overproduction

The Ecl1 protein has no sequence homology to any other proteins with a known function, and so we could not speculate on its function. We constructed a Δecl1 mutant and analyzed its phenotype. However, we did not detect any prominent difference in the chronological lifespan of the wild-type cells and the Δecl1 mutant (Fig. 7a and c). We analyzed the stress sensitivity of the Δecl1 mutant, such as osmotic sensitivity (growth on a plate containing 0.9 M KCl, 2 M glucose, or 1.3 M sorbitol), oxidative sensitivity (2 mM H₂O₂), high- or low-temperature sensitivity (37 °C or 20 °C), and UV sensitivity (100 J); however, we did not detect any prominent phenotype of the Δecl1 mutant (data not shown).

7

Effect of calorie restriction on cell viability after entry into the stationary phase. (a) Wild-type cells (JY333) carrying vector plasmid pLB-Dblet were grown in SD medium (2% glucose; ○) or SD low-glucose medium (0.5% glucose; ●). Cell viability (upper panel) and cell growth (lower panel) were monitored. (b) Wild-type cells (JY333) carrying plasmid pSht3 were grown in SD medium (○) or SD low-glucose medium (●), and cell viability (upper panel) and cell growth (lower panel) were monitored. (c) Δecl1 cells (JY333 Δecl1) carrying vector plasmid pLB-Dblet were grown in SD medium (○) or SD low-glucose medium (●), and cell viability (upper panel) and cell growth (lower panel) were monitored.

In S. pombe, chronological aging is accelerated by glucose signalling; cells bearing mutations in genes controlling this pathway, such as pka1 and sck2, live longer (Roux et al., 2006). In S. cerevisiae, deletion of the genes SCH9, CYR1, and RAS2, which mediate glucose signaling, extends the chronological lifespan (Fabrizio et al., 2001, 2003). To investigate the relationship between ecl1⁺ and glucose signalling in more detail, we analyzed cell viability in a glucose-limiting medium. The analysis of viability in a glucose-limiting medium is known as a calorie-restriction experiment, and calorie restriction is a condition that extends lifespan in a variety of species (Guarente & Picard et al., 2005; North & Sinclair et al., 2007). As shown in Fig. 7a and c, calorie restriction extended the viability of wild-type and Δecl1 cells. However, there was no extension due to caloric restriction of viability in ecl1⁺-overexpressing cells (Fig. 7b).

The Pka1 pathway is involved in the extension of viability by Ecl1 overproduction

As mentioned above, the chronological lifespan of S. pombe is regulated by two genes: pka1⁺ and sck2⁺ (Roux et al., 2006). The deletion mutant of each gene was reported to increase viability in the stationary phase. Because the Δpka1, Δsck2 double-deletion mutant showed an increase in lifespan compared with the single mutants Δpka1 and Δsck2, the lifespan was thought to be regulated by these factors independently. To determine how Ecl1 overproduction acts to increase viability, first we made and confirmed that both Δpka1 and Δsck2 mutants increase viability in the stationary phase under our assay condition (Fig. 8). Then we introduced plasmid pSht3 into Δpka1 and Δsck2 mutants and viability was analyzed. As shown in Fig. 8b, the Δpka1 mutant showed little, if any, extension of viability by introduction of pSht3. On the other hand, in the Δsck2 mutant, the presence of the plasmid extended viability to the extent of the wild type (Fig. 8a and c).

8

Effect of Δpka1 or Δsck2 mutation on the extension of viability in the stationary phase by Ecl1 overproduction. Cell viability (upper panels) and cell growth (lower panels) in SD liquid medium of wild-type cells (column a), Δpka1 mutant (column b), or Δsck2 mutant (column c) carrying vector plasmid, pLB-Dblet (○) or pSht3 (●), were monitored.

These results suggested that Ecl1 overproduction might affect the Pka1-dependent pathway negatively. To support this possibility, we analyzed the expression of the ste11⁺ gene because the ste11⁺ gene is known to be negatively controlled by Pka1 through phosphorylation of the Zn finger transcriptional activator Rst2 (Higuchi et al., 2002). As shown in Fig. 9, the expression of ste11⁺-mRNA increased in wild-type cells when Ecl1 was overexpressed. The expression of ste11⁺-mRNA also increased in the Δpka1 mutant as shown previously (Higuchi et al., 2002). The level of the pka1⁺-mRNA was not affected when Ecl1 was overexpressed (data not shown). These results suggested that the activity of the Pka1-dependent pathway might be negatively affected by Ecl1 overproduction.

9

Overproduction of Ecl1 up-regulates ste11⁺ expression. Total RNAs were isolated from the following strains, wild-type cells (JY333) carrying pLB-Dblet (lane 1), wild-type cells (JY333) carrying pSht3 (lane 2), and pka1-deletion mutant (JY333 Δpka1) carrying pLB-Dblet (lane 3) grown in SD medium in the log phase and subjected to Northern blotting analysis with a radiolabelled ste11⁺ probe.

Overproduction of Ecl1 renders cells resistant to oxidative stress

In S. cerevisiae, it was shown that most chronologically long-lived mutants show high oxidative stress resistance (Fabrizio & Longo et al., 2003). Hence, we analyzed the oxidative stress sensitivity by spotting cells on a plate containing H₂O₂. As shown in Fig. 10, wild-type cells carrying the vector plasmid was sensitive to 5 mM H₂O₂. However, the cells carrying pSht3 was found to be resistant to this oxidative stress.

10

Effect of Ecl1 overproduction on oxidative sensitivity. Wild-type cells (JY333) carrying vector plasmid, pLB-Dblet or pSht3, were spotted on an SD plate (panel a) or an SD plate containing 5 mM H₂O₂ (panel b) with serial dilution. The plates were incubated for 4 days at 30°C and photographed.

Discussion

We have identified an unannotated gene of S. pombe that extends the chronological lifespan when provided from a high-copy number plasmid. The gene, which we named ecl1⁺, encodes an 80-amino acid protein. Ecl1 was originally identified as a multicopy suppressor of an hos2 mutant displaying a high-osmolarity-sensitive phenotype (Nakamichi et al., 2000). Later, Hos2 was identified as a component of the DASH complex. The DASH complex of S. pombe is transiently associated with kinetochores only during mitosis and is required for precise chromosome segregation (Liu et al., 2005). Thus, Ecl1, as it is localized in the nucleus, could be involved in chromosome maintenance. A relationship between replicative lifespan and genetic stability has been reported; the accumulation of extrachromosomal rDNA circles, which are generated by homologous recombination within the rDNA repeat, was a major factor in the replicative lifespan of S. cerevisiae (Piper et al., 2006). However, there was no evidence indicating linkage between chronological lifespan and genetic stability or chromosome maintenance in S. pombe. Further characterization of Ecl1 might shed light on this point.

We have not detected a prominent phenotype, especially for the chronological lifespan, for the ecl1 deletion mutant. It is possible that there might be other factors that functionally overlap with Ecl1. Concerning this possibility, we carried out a blast search at DDBJ (DNA Data Bank of Japan) using tblastn program and the nr-nt database, and we found two other short ORFs (Ecl2 and Ecl3) that are homologous to Ecl1 in the genome sequence of S. pombe. Ecl2 and Ecl3 are encoded at nucleotide positions 2361–2612 (complementary) of chromosome II (GenBank Accession no. AL163702) and 32191–32457 (complementary) of chromosome II (GenBank Accession no. AL021815), respectively. Neither has been shown to encode proteins. We have cloned DNA fragments encompassing Ecl2-ORF and Ecl3-ORF on the vector plasmid (pLB-Dblet) and confirmed that both clones extend viability of wild-type cells (H. Ohtsuka, Y. Ogawa and H. Aiba, unpublished data). These results support the idea that Ecl2 and Ecl3 might have overlapping functions with Ecl1. To clarify this possibility, we are now characterizing these two ORFs in more detail. We could not identify the orthologous protein of Ecl1 in other organisms by the blast search. Hence, the family consisting of Ecl1, Ecl2, and Ecl3 may be unique in S. pombe.

The expression of the ecl1⁺ gene showed an intriguing pattern. The amount of ecl1⁺-mRNA increased transiently when growth phase was changed from the log phase to the stationary phase. The limitation of some nutrient may be a signal that induces the ecl1⁺-mRNA. The expression of Ecl1-EGFP protein in the nucleus was evident in log phase, but not in the stationary phase. Although we could not directly compare the amount of ecl1⁺-mRNA and Ecl1-EGFP protein because Ecl1-EGFP protein was expressed from plasmid, these results may suggest that not only the transcription but also the stability of Ecl1 protein is regulated by the growth phase. Increased expression of Ecl1 protein at the log phase might be sufficient for the extension of viability after entry into stationary phase. If so, the possible mechanism (program) for the extension of chronological lifespan should be operating even in cells growing in the log phase.

We found that calorie restriction extends the chronological lifespan of S. pombe. Calorie restriction extends the lifespan in a variety of species; in the present report, S. pombe has also been established as a model system to analyze the mechanism of lifespan extension by calorie restriction. We demonstrated that calorie restriction extended the chronological lifespan of the Δecl1 mutant. This result indicates that the extension of chronological lifespan by calorie restriction does not depend on Ecl1. As mentioned above, there might be other factors compensating for Ecl1. On the other hand, calorie restriction did not extend the lifespan of Ecl1-overproducing cells; thus, there was no additive effect on extension of lifespan between Ecl1 overproduction and calorie restriction. One of the explanations for this phenotype is that the overexpression of Ecl1 might cause physiological changes that are equivalent to the changes caused by calorie restriction. It should be noted that the lowering of the glucose level may simply be acting by decreasing the level of glucose repression, thereby increasing the respiration competence of the cells, their oxidative stress resistance, and their ability to survive in the stationary phase. Clarification of these and other possibilities is awaited in the future.

At present, molecular mechanisms of lifespan regulation have not been elucidated clearly. To date, the factors known to be involved in lifespan regulation in S. pombe are Sck2 and Pka1 (Roux et al., 2006). Sck2 and Pka1 were known to function as negative factors for the extension of chronological lifespan in separate (in redundant or parallel signalling) ways. In the present study, we showed that overproduction of Ecl1 does not extend viability in the Δpka1 mutant. We also presented evidence that the ste11⁺ gene was up-regulated in wild-type cells when Ecl1 was overexpressed. Moreover, we showed that overproduction of Ecl1 renders the cells resistant to oxidative stress. Roux et al. reported that the Δpka1 mutant was resistant to oxidative stress due to exposure to H₂O₂. Taken together with these results, we propose that the overproduced Ecl1 negatively act on (or regulate) the Pka1 pathway, directly or indirectly. Because Ecl1 was originally identified as a multicopy suppressor of hos2 mutation (Hos2 is a component of the DASH complex), Ecl1 may be a key player linking the state of chromosome maintenance and/or genetic stability and chronological lifespan.

In this study, we provide a clue to regulate the chronological lifespan by a novel factor. At present, we have not been able to identify the orthorogous protein of Ecl1 in other organisms, and so we do not know how widely conserved the mechanism of life-span extension by Ecl1 is. Further characterization of Ecl1 in S. pombe should shed light on the complex mechanisms that regulate the lifespan.

Acknowledgements

We thank I. Kawagishi and S. Banno (Nagoya University) for technical assistance with fluorescence microscopy; A. Roux and L. Rokeach (University of Montreal) for strain; and H. Yamada, E. Yamamoto, T. Mizuno, and T. Yamashino (Nagoya University) for helpful discussions. This work was supported by a Grant-in-Aid for Scientific Research (to H.A.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Author notes

Editor: Ian Dawes