OUP user menu

A novel assay for replicative lifespan in Saccharomyces cerevisiae

Stefanie Jarolim, Jonathan Millen, Gino Heeren, Peter Laun, David S. Goldfarb, Michael Breitenbach
DOI: http://dx.doi.org/10.1016/j.femsyr.2004.06.015 169-177 First published online: 1 November 2004


The replicative lifespan of Saccharomyces cerevisiae is determined by both genetic and environmental factors. Many of the same factors determine the lifespan of metazoan animals. The lack of fast and reliable lifespan assays has limited the pace of yeast aging research. In this study we describe a novel strategy for assaying replicative lifespan in yeast, and apply it in a screening of mutants that are resistant to pro-oxidants. The assay reproduces the lifespan-shortening effects of deleting SIR2 and of growth in the presence of paraquat, a pro-oxidant. The lifespan-increasing activity of resveratrol is also reproduced. Compared to current assays, this new strategy promises to significantly increase the possible number of replicative-lifespan determinations.

  • Saccharomyces cerevisiae
  • Yeast
  • Longevity
  • Aging
  • Replicative lifespan

1 Introduction

Vegetatively dividing Saccharomyces cerevisiae cells are mortal. Mother cells produce a limited number of daughter cells before they begin to exhibit signs of functional senescence, cease replicating and, ultimately, die [1]. Mortimer and Johnston [2] were the first to count the maximum number of daughters produced by individual mother cells, using micromanipulation to remove the daughters before they themselves replicated. In this fashion it was determined, and later verified, that the number of daughter cells a population of mother cells produce is empirically described by the Gompertz survival function, which is commonly used to describe the increased tendency of older animals to die (reviewed in [3]). Interest in the replicative lifespan of S. cerevisiae increased when it became apparent that the aging of fungi, worms, and flies held some promise for revealing the molecular mechanisms of cellular aging of humans [4,5]. Another approach to the study of aging in yeast, called chronological aging, measures the length of time stationary-phase cells can survive in depleted medium. In contrast, replicative aging does not depend on calendar time. This property is a boon to experimentalists who can refrigerate the cells overnight without affecting their replicative capacity [6]. The relative merits of replicative and chronological aging assays, and their respective relevance to human aging, have been discussed elsewhere [79]. Chronological and replicative aging in yeast are related because events in stationary phase affect subsequent replicative lifespan in rich medium and many of the same genes affect both modes of aging [10,11].

Mother cell-specific replicative aging of yeast cells is based on the asymmetric division of mother and daughter cells. The clock for daughters is generally reset to zero, although daughters of older mothers, which replicate more slowly, have reduced lifespans [11]. Presumably, mother cells age and die because they become damaged. The chemical agents of the damage, the macromolecular targets, and the cell processes that are directly and indirectly affected are incompletely understood. Oxidative stress has been proposed as having a significant effect on lifespan [12,13]. Down-regulation of glucose signalling pathways also increases oxidative stress resistance by mechanisms dependent on proteins such as superoxide dismutases and catalases. These enzymes normally promote growth and increase protection against oxidative stress and the mutation of their genes in yeast decreases replicative lifespan [1417]. Tied into these pathways are genes controlling the mitochondrial retrograde response pathway [18,19]. The disruption of the signalling pathways that control caloric restriction in yeast increases lifespan by a mechanism which according to some authors depends on the histone deacetylase SIR2[20], but according to others does not [21]. Sir2p is a conserved NAD+-dependent histone deacetylase that plays an important role in gene silencing, and promotes long life in S. cerevisiae and Caenorhabditis elegans[22]. Sir2p activity appears to be controlled by enzymes that regulate the concentration of nicotinamide and/or NADH, both of which act as Sir2p inhibitors. Pnc1p positively regulates Sir2p-mediated silencing and longevity by preventing the accumulation of nicotinamide during stress [20,23,24]. Others argue that it is the concentration of the inhibitor NADH that regulates Sir2p activity via the NAD+/NADH ratio [25]. Extrachromosomal DNA minicircles (ERCs) as well as damaged mitochondrial macromolecules like protein carbonyls, which are products of oxidative stress, are candidate “death factors” since they accumulate in older mothers and are not inherited by their daughters, except in sir2 mutants [12,26].

The replicative lifespan of a yeast strain is typically described by the median lifespan (“n”), which can vary widely among laboratory strains, but is typically around 20 generations. In any population of logarithmically replicating cells, only a small fraction (1/2n) is senescent. Research in the field of yeast mother cell-specific aging has been hampered by the fact that old cells are rare in any population and there is no direct and easy way to efficiently separate old (terminally senescent) mother cells from young cells [26]. The purification methods that were tried include gradient purification, immobilization on magnetic beads [2729] and elutriation centrifugation [13]. However, in most cases the preparation of old cells still contains young cells which adhere to their mothers due to the peculiarities of yeast cytokinesis [13]. An even more efficient purification would be necessary for the construction of a genetic selection system for long-lived mutants. The development of faster, higher-throughput, and less labor-intensive replicative lifespan assays would greatly facilitate the study of aging in yeast. In this study we describe such an assay and apply it in a screen for long-lived mutants that are resistant to oxidative stress.

2 Materials and methods

2.1 Yeast strains

Mutant strain K6001 in W303 background is marked MATa, ade2-1, trp1-1, can1-100, leu2-3,112, his3-11,15, GAL, psi+, ho::HO::CDC6 (at HO), cdc6::hisG, ura3::URA3 GAL-ubiR-CDC6 (at URA3) [30]. K6001 sir2Δ was produced by replacing the SIR2 gene with a kanamycin resistance marker (kanR) cassette.

SIR2 plasmid: SIR2 was amplified via PCR from genomic DNA isolated from K6001 using primers with EcoRI and HindIII restriction sites. The PCR product was integrated between the HindIII and EcoRI restriction site in plasmid pRK2 [31]. All purification steps were done with Qiagen PCR purification kit.

2.2 Media

Yeast strains were grown on YPGlucose containing 2% (w/v) glucose, 1% (w/v) yeast extract, 2% (w/v) peptone or on YPGalactose with same components as YPGlucose but containing 3% (w/v) galactose instead of glucose. For resistance and sensitivity tests on different oxidants galactose minimal medium (SC) containing 3% (w/v) galactose, 0.5% ammonium sulphate, 0.017% yeast nitrogen base with amino acids (0.002% Arg, 0.001% His, 0.006% Ile, 0.006% Leu, 0.004% Lys, 0.001% Met, 0.006% Phe, 0.005% Thr, 0.004% Trp, 0.001% Ade, 0.004% Ura, 0.005% Tyr (Sigma)) was used. Agar plates were made by adding 2% (w/v) agar to the media.

2.3 Lifespan assays

Standard lifespan assays were performed as described previously [13]. All lifespans were determined on defined SCGlucose or SCGalactose media for a cohort of at least 40 cells. In a previous review we have discussed that the median of a lifespan distribution, and its standard deviation at a 95% level, are in practice the best parameters to describe lifespan (Figs. 1(c), 2 and 3) [32].

Figure 1

Replicative lifespan assays of K6001 cells. (a) Growth of K6001 SIR2+ (♦) and sir2Δ (●) cells in YPGlucose. Cells were cultured at 30 °C. (b) Growth of K6001 SIR2+ (♦) and sir2Δ (●) cells in YPGalactose. (c) Standard replicative lifespan survival curves for K6001 cells on SCGlucose (Embedded Image) and SCGalactose (Embedded Image) plates. Lifespan determination was done by micromanipulation. Percent viable cells was plotted against cell generations. Perpendicular dotted lines indicate median lifespans. Standard deviations are indicated by horizontal error bars. Assays were performed as described in Section 2.

Figure 2

Genetic and environmental factors that affect the lifespan of K6001 cells. (a) Survival plot of SIR2+ (Embedded Image), sir2Δ, (○) and sir2Δ cells plus SIR2+ integrated on a plasmid (△). In this and in the remaining parts of Fig. 2, lifespans were determined by the microcolony method on YPGlucose plates. (b) Survival curve of K6001 on plates containing 0 μM (Embedded Image), 0.5 μM (○), and 1.0 μM paraquat (△). (c) Survival curve of K6001 on plates containing 0 μM (Embedded Image), 10 μM (○), and 100 μM resveratrol (△). (d) Survival curves of K6001 sir2Δ cells on plates containing 0 μM (Embedded Image), 10 μM (○), and 100 μM resveratrol (⋄). Median lifespan is indicated by perpendicular dotted lines. Standard deviations are indicated by horizontal error bars. See Section 2 for details.

Figure 3

Lifespans of WT K6001 and long-lived mutants grown on SCGlucose media. Lifespan determination in (a) was done by lifespan assay on solid agar plates, by counting the cells of a microcolony. Percents viable cells are plotted against cell generations, perpendicular dotted lines indicate median lifespans and (b) shows the same strains as in (a) but lifespan determination was done by classical micromanipulation. Standard deviations are indicated by horizontal error bars.

Lifespan assay with spectrophotometer: cells were grown to logarithmic phase in YPGalactose media, washed twice with water and transferred to liquid YPGlucose or SCGlucose media. Optical density at 600 nm at the beginning was between 0.1 and 0.3. Cells were shaken at 28 °C for two days and optical density determined. The ratio between optical density after two days and at the beginning of the experiment was calculated and used in initial screening tests of mutants as a measure of lifespan. Comparison of these ratios with either standard lifespans done by micromanipulation or with lifespans determined by the microcolony method (see below) showed that the spectrophotometrically determined lifespans were proportional to the true lifespans determined by micromanipulation.

Lifespan assay on solid agar plates by counting cells in a microcolony: cells were grown to logarithmic phase in YPGalactose media, washed twice with water, and using a micromanipulator a single cell with a small bud was transferred to a defined spot on a SCGlucose or YPGlucose plate. A minimum of 40 cells were put on the plate, grown for two days at 28 °C and then the cell number in each microcolony was counted under the microscope. As the results show, cell numbers are proportional, but not numerically equal to the number of generations as determined by the standard micromanipulation lifespan assay.

2.4 Nutrient determination in media

The presence of glucose in the media was estimated with Diabur-Test® 5000 dipsticks (Roche Diagnostics, Germany).

2.5 Mutagenesis

Cells were mutagenized to 10% survivors by treating cells with UV or ethyl methane sulfonate (EMS). For UV mutagenesis cells were grown in YPGalactose for two days to stationary phase, washed twice, diluted to 6.7 × 106 cells ml−1 in water. Of this suspension 15 ml was poured into a Petri dish and exposed to UV light (UV-source: Phillips TUV 15W/615T8, UV-C) at 40 cm distance from the light source for 20 s. Survival rate was determined by plating out proper dilutions.

For EMS mutagenesis logarithmically growing cells in YPGalactose were washed twice with 50 mM potassium phosphate buffer at pH 6.8, resuspended in 10 ml of this buffer with 300 μl EMS (Sigma) added. After incubation for 1 h at 28 °C, 10% of the cells survived. Mutagenesis was stopped by adding 10 ml sodium thiosulfate (10% (w/v)) to the solution.

2.6 Determination of respiration

Strain K6001 is not able to grow on glycerol (the essential CDC6 gene is under control of the GAL1-10 promoter which is not active on glycerol). Therefore, in order to determine mitochondrial respiration, oxygen consumption was determined with a Clark type electrode (Oxygraph 2k, Oroboros, Innsbruck, Austria) following a standard protocol.

2.7 Oxidative stress sensitivity and resistance

Oxidants were added to SC plates with galactose in concentrations at which the most resistant mutant strain was slightly decreased in viability down to the concentration where the most sensitive strain was still able to grow. Final concentrations were 0.1–0.2 mM cumene hydroperoxide (CHP, Sigma), 0.4–1 mM diamide (Sigma), 1.5–3.0 mM hydrogen peroxide (H2O2, Sigma), 0.05–0.6 mM menadione (Sigma) and 1–3 mM tert-butyl hydroperoxide (tBHP, Sigma). Cultures were grown overnight, serially diluted to OD600 values of 3.0, 1.0, 0.3 and 0.1. Of each dilution 10 μl aliquots were spotted. Sensitivity or resistance was determined by comparing the wild-type strain with mutant strain growth after two days at 28 °C.

3 Results

3.1 A novel strategy for assaying replicative lifespan in yeast

Replicative lifespan in S. cerevisiae is defined as the number of daughter cells produced by individual mothers before they senesce and die. Replicative lifespan of a strain is usually expressed as the median lifespan of a cohort of cells. The replicative lifespan currently in wide use typically involves counting the number of daughters produced by 30–60 individual mother cells positioned carefully on an agar plate. The replicative cycles of the mothers are monitored by light microscopy throughout the day. Each new daughter cell is counted and removed by micromanipulation, leaving only the mothers to replicate, or die. Depending on the maximum lifespan of the strain, it takes between one to two weeks for a skilled person to complete a set of up to four assays, assuming the cells are refrigerated overnight. By example, the replicative lifespan of a strain with an average doubling time of 120 min and a maximum lifespan of 50 generations can be completed in ten 10-h days.

Here, we describe a novel replicative lifespan assay that requires neither the constant monitoring of cell divisions nor the manual removal of daughter cells. This assay exploits a W303-based strain, K6001, which was originally constructed to study the mother cell specifity of mating type switching [30,33]. K6001 cells contain two chromosomally integrated, independently segregating copies of CDC6 whose expression is regulated by the repressible GAL1-10 and mother-cell specific HO promoter, respectively [30]. Cdc6p is an essential protein required for DNA replication. Transcription from the HO promoter in daughter cells is prevented by Ash1p, which is expressed exclusively in daughter cells [34]. When K6001 cells are grown permissively on galactose, GAL1-10::CDC6 is expressed both in mother and daughter cells and the culture grows logarithmically (Fig. 1(b)). However, when expression of the GAL1-10::CDC6 gene is repressed by glucose, only the mother cell-specific expression of HO::CDC6 remains to support growth. cdc6 Mutants undergo mitosis without replicating their DNA, but subsequently arrest in G1 [33]. As a result, K6001 cultures grow more or less linearly in glucose until the mothers exhaust their capacity to replicate (Fig. 1(a)). Since Cdc6p is naturally unstable, and the GAL1-10 promoter is efficiently shut-off in glucose, the transition from log growth to linear growth is relatively tight and rapid. Both the HO and CDC6 genes are normally expressed during late G1 phase [33,35].

The replicative lifespan of strain K6001 in glucose can be assessed using the standard micromanipulation assay, which involves removing the dead or dying daughter cells after every cell cycle. More simply, the number of daughters produced by K6001 mother cells in glucose medium can be counted, either directly by micromanipulation after two days growth on glucose plates, or indirectly by monitoring their accumulation in liquid glucose medium by optical density (OD600). Standard micromanipulation lifespan assays in SC medium revealed a significant difference between the replicative capacity of K6001 cells in galactose and glucose, although theoretically they should be the same. As shown in Fig. 1(c), the median lifespans in this experiment were ∼4 generations in glucose and ∼17 generations in galactose. This discrepancy is likely due to properties of Cdc6p and its expression, or lack of proper expression, from the HO promoter in K6001 cells on glucose. Normally, the expression of CDC6 occurs in a transient burst at the end of G1 [33]. The HO promoter also drives expression during G1 [35], but it may do so at a lower level and over a more restricted period than the native CDC6 promoter. If the HO promoter is insufficient to maintain adequate or properly timed levels of Cdc6p, then the K6001 cells grown in glucose may die, in part, from a lack of Cdc6p, and consequently have a shorter lifespan. Expression of CDC6 from the GAL1-10 promoter is sufficient for survival, and thus we see the true lifespan of the strain, which is relatively close to the lifespan of the CDC6+ parent strain, W303 [16]. To test if the shortened lifespan of K6001 cells in glucose medium precludes its use in the study of lifespan we investigated the effects of mutations and environmental agents that are known to increase or decrease replicative lifespan.

3.2 Growth of SIR2+ and sir2Δ K6001 cells in galactose and glucose-containing medium

The application of K6001 cells to the study of replicative lifespan was tested by comparing the growth of SIR2 and sir2Δ K6001 cells in either galactose (YPGalactose) or glucose-containing (YPGlucose) liquid media. Here, the optical density at 600 nm (OD600) is used as a relative measure of the linear growth and senescence of K6001 cells. The deletion of SIR2 is well known to result in a shortening of replicative lifespan [36]. Since 88% of cells in a log phase culture are one, two, or three generations old (1/2 + 1/4 + 1/8…), K6001 cells growing in galactose are effectively age-synchronized. Therefore, the replicative capacity of K6001 cells can be estimated by the time-dependent cessation of growth of cells incubated in glucose. As shown in Fig. 1(b), both SIR2 and sir2Δ K6001 cells grew logarithmically with identical doubling times in permissive YPGalactose. In contrast, in YPGlucose SIR2 cells grew to a much higher OD600 than sir2Δ cells (Fig. 1(a)). The cessation of replication of both strains in YPGlucose was not due to nutrient limitation, since normal laboratory strains grow to significantly higher cell densities and glucose was still present in the media as shown by glucose measurements with test dipsticks. The initial growth of both SIR2 and sir2Δ K6001 cells in YPGlucose was approximately linear. After extended periods of incubation in YPGlucose the block preventing daughter cell division in glucose is occasionally lost, and these cultures revert to logarithmic growth (data not shown).

Qualitatively similar results were obtained by counting the numbers of daughter cells produced by a cohort of individual mother cells forming microcolonies on glucose-containing agar plates (see Section 2). The mother cell in each microcolony can usually be distinguished from the daughters by her large and sometimes irregular shape. By determining cell numbers of these microcolonies, survival curves can be obtained similar to those obtained by standard micromanipulation assay (see Figs. 2 and 3(a)). However, because the arrest point of a cell lacking CDC6 expression is late G2, and some cells complete mitosis without replicating their DNA, the number of cells in microcolonies is higher than the number of generations counted by the standard micromanipulation assay in which daughter cells are removed after each generation (Fig. 3(b)). Therefore, lifespans determined by the two methods are proportional but not numerically equal. As shown in Fig. 2(a), the lifespan of SIR2 K6001 cells determined by cell counts are significantly greater than sir2Δ cells, and the defect was rescued by a plasmid expressing SIR2 (Fig. 2(a)).

3.3 Paraquat and resveratrol affect the replicative lifespan of K6001 cells

We sought to test if the usefulness of K6001 cells extended to the study of chemical agents that either increased or decreased replicative lifespan. Pro-oxidants such as paraquat are known to reduce the replicative lifespan of yeast [16]. We tested the effect of paraquat on the growth of K6001 cells in liquid YPGalactose. Twenty micromoles of paraquat had no effect on the permissive growth of K6001 cells in liquid YPGalactose (not shown). Fig. 2(b) shows survival curves for colonies of K6001 cells in 0, 0.5, and 1 μM paraquat on SCGlucose plates. Lifespans were not reduced further by paraquat concentrations in excess of 1 μM (not shown). Higher concentrations of paraquat were needed to reduce the lifespan of K6001 cells in rich YPGlucose versus minimal SCGlucose (not shown), presumably because some component(s) present in YP medium counteracted the pro-oxidant activity of paraquat.

Resveratrol was recently reported to increase replicative lifespan in a SIR2-dependent fashion [37]. As shown in Fig. 2(c) the lifespan of K6001 cells was significantly increased by 10 μM resveratrol, and to a lesser extent by 100 μM, presumably because resveratrol is toxic to yeast at higher concentrations [37]. Resveratrol is thought to increase lifespan by activating the histone deacetylase activity of Sir2p, based mainly on the fact that resveratrol had no influence on the lifespan of sir2Δ cells [37]. Consistent with this result, resveratrol did not affect the lifespan of sir2Δ cells derived from K6001 (Fig. 2(d)).

3.4 A novel selection system for long-lived mutants

We investigated the utility of K6001 cells in the selection of mutants with prolonged lifespans. Large numbers of preselected mutants can be screened by monitoring their growth in microtiter wells. As suggested by the results shown in Fig. 1(a), the relative lifespans of mutant K6001 cells is proportional to the number of times they divide in glucose medium, which can be assessed by measuring plateau values of OD600. Mutants with substantially longer lifespans should grow to higher optical densities than the parental K6001 strain.

We mutagenized strain K6001, grew the strain for two generations under non-selective conditions on galactose to enable proper mutant fixation and segregation, and then applied a number of different oxidative-stress selective conditions (see Section 2). Menadione, CHP, hydrogen peroxide, tBHP, and diamide were used to induce oxidative stress. Under these conditions, an average of three colonies were formed on the selection plates for every million cells plated. Several hundred survivors of oxidative stress were tested for an increase in OD600 after linear growth in microtiter plates. Lifespans of mutants with a substantially increased OD600 were confirmed by microcolony counting and standard lifespan determination (see Section 2).

For a first selection, lifespan in microtiter plates was measured by inoculating a certain amount of cells in glucose media. OD600 at the beginning and at the end was measured with an ELISA reader and the ratio between starting and final OD600 was calculated (data not shown). Twenty mutants, which had an increased ratio, were taken and lifespan was determined on agar plates containing glucose (microcolony method). Four strains that showed an increased number of cells per microcolony were further investigated by standard lifespan determination. The lifespans of these four mutant strains were significantly increased on SCGlucose relative to the parental K6001 strain as determined by two methods (Fig. 3(a) and (b)). For example, mutant strains U9 and Y10 exhibited 80% and 60% increase in lifespan, respectively. Surprisingly, the lifespans of these mutants, when performed by standard micromanipulation assay on galactose medium, were not significantly greater than those of the parental strain (data not shown).

Spot tests were performed to characterize the sensitivity of the four mutants to oxidative stress. The sensitivity to several oxidants of the mutants compared to parental K6001 cells are shown in Table 1 and Fig. 4. Interestingly, each mutant exhibited a unique sensitivity profile to the oxidants. For example, O7 cells exhibited increased resistance to all four oxidants, whereas Y10 cells exhibited increased resistance only to H2O2 and were more sensitive than parental cells to diamide.

View this table:
Table 1

Resistance and sensitivity of K6001 and long-lived mutants to different oxidants (compare Fig. 4)

  • wt indicates no sensitivity or resistance, compared to wt K6001; +++ very strong resistance, ++ resistance, + slight resistance. Strong sensitivity is indicated by −−−, sensitivity by −− and weak sensitivity by −.

Figure 4

Sensitivity of K6001 and K6001-O7 mutant to different oxidants. Sensitivity and resistance to hydrogen peroxide (H2O2), diamide, CHP and tBHP was determined as described in Section 2. Concentration of oxidants used is indicated in the left column. All strains were grown to stationary phase in SCGalactose media, adjusted to an OD600 of 3.0, 1.0, 0.3 and 0.1 before spotting 10 μl aliquots onto appropriate plates. Plates were incubated for three days at 28 °C and photographed.

The respiration capacity of all four mutants was tested with a Clark-type electrode. Parental K6001, O7, U9 and K11 cells showed similar respiration levels. Y10 cells are respiratory-deficient, as they apparently consumed no oxygen.

4 Discussion

In this study we have tested a novel strategy for assaying the replicative lifespan of S. cerevisiae. This approach has the potential to replace the current method, which requires the laborious removal of newly budded daughter cells before they have a chance to divide. The new strategy exploits the fact that K6001 cells grown on glucose produce daughters that do not need to be removed because they die. The daughters die because they do not express the mother cell-specific HO::CDC6 gene. K6001 cells grown on galactose produce daughters that replicate permissively because they are supported by a GAL1-10::CDC6 gene. In principle, the replicative capacity of an individual K6001 cell on glucose can be determined by counting the number of dead daughters that surround their senescent mothers. Alternatively, the optical densities of liquid K6001 cultures in glucose medium should, in principle, be limited by the replicative capacities of the mother cells in the inoculum. In practice, this strategy produced mixed but promising results.

On the positive side we were able to demonstrate that lifespans of K6001 cells obtained with this strategy reproduced certain established characteristics of yeast aging. K6001 cells containing a sir2Δ mutation exhibited a reduced median lifespan that was rescued by wild-type SIR2 on a plasmid. The assay also reproduced the lifespan-shortening effect of paraquat and the SIR2-dependent lifespan-extending effect of resveratrol. We also employed the assay in a screen of oxidative stress-resistant mutants, and successfully isolated a number of long-lived mutant strains.

A shortcoming in the current configuration of this assay was exposed when the lifespans of K6001 cells were compared on glucose and galactose using the standard micromanipulation assay. In principle, the lifespan of K6001 cells should be the same in glucose and galactose, since the mother cells should express CDC6 on both carbon sources. In practice, the median lifespan of K6001 cells, when calculated by removing daughters as they bud, was considerably less on glucose than galactose, suggesting that, perhaps, the HO::CDC6 gene is not sufficient to maintain healthy replication of mother cells. Other explanations are possible, but inadequate level of Cdc6p in mother cells grown on glucose is the most likely explanation. This raises the caveat that the lifespan of K6001 cells in glucose is affected by a combination of factors, most notably a deficiency of Cdc6p. How a putative Cdc6p deficiency might interact with other genetic and environmental factors to affect lifespan is not known. Happily, the effects on lifespan of a sir2Δ mutation, and of paraquat and resveratrol, were not overridden by the putative Cdc6p deficiency. However, we were not able to reproduce experimental results obtained in other genetic backgrounds [21] pointing to a lifespan-extending effect of caloric restriction in yeast. Specifically, decreasing concentrations of glucose from 2% to 0.1% actually decreased rather than increased lifespan (data not shown). This problem may well be related to the notion that K6001 cells have short lifespans even when grown on high glucose concentrations.

The number of daughter cells produced by K6001 cells on glucose is somewhat higher than the intrinsic replicative capacity of the strain, because the daughter cells can complete mitosis without replicating their DNA after they have separated from the mother cells. In some cases the daughter cells can replicate a few times, possibly because they are sustained by residual Cdc6p expressed at high levels from the GAL1-10::CDC6 gene prior to their transfer into glucose medium. We do not think this explanation very likely, since the ability of daughter cells to continue to divide was unaffected by culturing the cells on 2% raffinose and 0.1% galactose prior to the shift to glucose. These conditions should have significantly reduced cellular Cdc6p levels compared to levels produced on 2% galactose. For certain applications, the replication of some daughter cells proportionally amplifies the median lifespan value of the strain, and explains why the OD600 values of K6001 cells in liquid glucose media never completely attain plateau (see Fig. 1(a)). Finally, a small fraction of K6001 cells revert to logarithmic growth in glucose, although this occurs infrequently enough not to interfere with either the liquid or solid medium-based assays. The reason for this is likely loss of repression of either the GAL1-10::CDC6 or HO::CDC6 constructs, in either mother or nascent daughter cells. Bobola et al. [30] have isolated a number of mutations in ASH1 that suppressed the inability of K6001 cells to grow on glucose, and these could be reverted by expression of wild-type ASH1 on a plasmid. Ash1p is expressed only in daughter cells and is required for the repression of HO expression [30]. In our hands, the reversion to K6001 cells to growth on glucose media was not prevented by expression of ASH1 from a CEN plasmid (data not shown).

The effect of the pro-oxidant paraquat on the lifespan of K6001 cells in glucose are consistent with published reports, and indicates that the strain is suitable for studying the effect of oxidative stress on lifespan.

The four long-lived mutants shown in Figs. 3 and 4 and in Table 1 were isolated in a two-step process. In the primary selection colonies were isolated which grew on plates containing the oxidants shown in Table 1. Six hundred resistant mutants were screened on microtiter plates for the OD600 reached after inoculation with a defined number of cells, resulting in 20 “high-OD” mutants. As we are showing here, high OD is correlated with a longer lifespan than WT. The 20 mutants were now tested for lifespan in two ways, the microcolony assay and the conventional lifespan assay by micromanipulation. Four mutants remained that showed a significantly longer lifespan than WT. Interestingly, we see strong resistance to one oxidant together with strong sensitivity to another oxidant in some mutants.

Fig. 3(a) presents the lifespans of the four strains as determined by the microcolony assay and Fig. 3(b) shows the same strains analyzed by micromanipulation. A comparison clearly demonstrates that both methods yield similar results. However, for instance mutant strain O7 is extremely long-lived in the microcolony assay, but only moderately long-lived in the micromanipulation assay. This could be explained by the fact that the mutation(s) in strain O7 not only influence the number of daughters generated by a mother, but also the terminal phenotype of the daughter, resulting in a higher number of descendants derived from each daughter cell in comparison to the starting strain, K6001. This is indeed the case (data not shown). We crossed each one of the long-lived mutants to the isogenic WT (data not shown) and observed the segregation pattern in the haploid progeny of these crosses. In each case, an irregular pattern resulted pointing to multigenic inheritance of the phenotype seen. As a control, some randomly selected resistant mutant strains (with unchanged lifespan) were also tested for segregation and were shown to segregate in a strictly Mendelian fashion. We deduce that the relationship between oxidative stress resistance and longevity is a very complex one, not allowing predictions in a simple one-to-one manner.

We conclude that measurements of optical density can reveal relative differences in the replicative lifespan of wild-type and mutant K6001 cells and may, with further work, facilitate the high-throughput analysis of lifespan required for large-scale genetic and chemical screens.


We are grateful to Lars Olsen for technical assistance. This work has been supported by grant no. P14574-MOB and P16402-MOB (FWF Austria; to M.B), and funds from the National Institutes of Health to D.S.G. (RO1 GM67838). A special thank to K. Nasmyth and N. Bobola for strain K6001.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
View Abstract