SEARCH

SEARCH BY CITATION

Keywords:

  • ageing;
  • stress resistance;
  • replicative senescence;
  • chronological lifespan;
  • Saccharomyces cerevisiae;
  • yeast;
  • general stress response

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Budding yeast can be considered to have two distinct lifespans: (a) a replicative (budding, non-chronological) lifespan, measured as the number of daughters produced by each actively dividing mother cell; and (ii) a chronological lifespan, measured as the ability of stationary cultures to maintain viability over time. In non-dividing cells, essential components that become damaged cannot be diluted out through cell division but must, of necessity, be turned over and renewed. By elevating stress resistances, many of the activities needed for such renewal should be elevated with commensurate reduction in the steady-state levels of damaged cell components. Therefore, chronological lifespan in particular might be expected to relate to stress resistance. For yeast to attain a full chronological lifespan requires the expression of the general stress response. It is more important, though, that the cells should be efficiently adapted to respiratory maintenance, since it is cultures grown to stationary phase on respiratory media that usually display the longest chronological lifespans. For this reason, respiration-adapted cells potentially provide a better model of chronological ageing than cultures pre-grown on glucose. Copyright © 2001 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Several recent studies have shown that the lifespans of model organisms can be increased by increasing their stress resistance and especially their levels of antioxidant enzymes. The initial reports of the extension of Drosophila lifespan by the insertion of transgenes coding for Cu/Zn superoxide dismutase (Cu/Zn–SOD) plus catalase (Orr and Sohal, 1994) did not survive close scrutiny (Kaiser et al., 1996; Tower, 1996). Despite this, it has since been demonstrated that increasing Cu/Zn–SOD levels in adults (Sun and Tower, 1999) or expressing human Cu/Zn–SOD in motor neurons (Parkes et al., 1998) can extend Drosophila lifespan.

Long-lived Drosophila can also be obtained by the selective breeding of individuals with long lifespan. Remarkably, these lines have a greater resistance to environmental stresses (Service et al., 1985) in the nematode Caenorhabditis elegans long-lived mutants can be selected directly, since this is a species that is self-fertilizing (Lithgow, 1996). Of the many mutations lengthening life in C. elegans, those affecting general metabolic rates (e.g. clk1) and those in the dauer/insulin receptor signalling pathway have been the most revealing (Ewbank et al., 1997; Lithgow, 1996; Martin et al., 1996; Wood, 1998). Mutations in daf-2 (Kenyon et al., 1993) and daf-23 (Larson et al., 1995) result in up to two-fold life prolongation and affect the starvation signalling that regulates dauer formation. daf-23 Encodes a PI3 kinase (Morris et al., 1996), signalling to the gene for a forkhead transcription factor (daf-16), whose action possibly serves to downregulate stress genes (Murakami and Johnson, 1996). In Drosophila (Orr and Sohal, 1994) and in C. elegans (Taub et al., 1999) there is evidence that increasing the scavenging of reactive oxygen species (ROS) increases longevity. These findings are all consistent with the notion that, for most organisms, stress protection and repair systems are not optimal for maximization of lifespan. They raise the important issue of why natural selection acts to prevent the higher constitutive expression of antioxidant defences and perhaps heat shock proteins, whereby lifespans would be maximal (Lithgow, 1996).

Yeast can potentially provide one of the simplest models of ageing. Traditionally, yeast ageing has been viewed as the events that limit the number of daughters an individual cell can produce. This replicative (budding) lifespan of mother cells is usually limited by proliferation of extrachromosomal circular forms of the ribosomal DNA (rDNA) repeat sequence on chromosome XII (ERCs). These ERCs mostly remain within the mother at cell division, causing this mother cell to senesce (Defossez et al., 1999; Sinclair and Guarente, 1997). In certain filamentous fungi (Neurospora crassa and Podospora anserina), linear mitochondrial plasmids regulate senescence (Osiewacz and Borghouts, 2000). It has yet to be shown though that similar genetic instabilities are a major cause of ageing in non-fungal systems. As a result, it is not very clear at present how the work on yeast replicative senescence relates to ageing studies in most other organisms. The precise relationship of stress resistance to the replicative lifespan of S. cerevisiae is also unclear. The deletion of RAS1 (Sun et al., 1994) or the altered transcriptional silencing associated with the sir4-42 mutation (Kennedy et al., 1995) both cause an increased replicative lifespan and increased stress resistance. However, the basis of these effects is unknown. They might reflect some of the genes that influence ERC formation being regulated by oxidative stress (Bandara et al., 1998). Alternatively, the increases in NAD+ as cells become progressively more prooxidant may promote histone deacetylation by the Sir2p NAD-dependent histone deacetylase (Imai et al., 2000). This in turn may suppress ERC formation by increasing transcriptional silencing (Imai et al., 2000), leading to the influences of Sir2p on replicative senescence (Lin et al., 2000).

Yeast replicative lifespan may not generally increase with increase in stress resistance. Cells that are growing on a respiratory carbon source do not consistently produce more daughters than cells growing mainly by the fermentation of glucose (Barker et al., 1999), even though stress resistances are generally higher in respiratory cells (Martinez-Pastor et al., 1996). Also, the constitutive overactivation of the yeast heat shock response, as in the cpr7,hsc82 mutant (Duina et al., 1998), does not increase the mean or maximal budding lifespan (Harris et al., 2000). An impaired oxidative defence shortens yeast replicative lifespan (Barker et al., 1999), although this may reflect a lowered ‘fitness’ of cells rather than any effect on ageing per se. In higher organisms also there is relatively little evidence to suggest that the replicative potential of somatic cells is influenced strongly by increases in stress resistance. In mammals, replicative senescence can be due to the shortening of the telomeres at the ends of chromosomes, an oxidative damage-induced accumulation of single-strand breaks in the telomeric DNA being one influence over this telomere shortening (von Zglnicki et al., 2000). However, telomere shortening is not normally a cause of senescence in S. cerevisiae (Austriaco, 1996; Austriaco and Guarente, 1997), making it improbable that oxidative damage to telomeric DNA is an influence over ageing in this species.

Yeast can be considered to have two discrete lifespans: (a) a replicative (budding, non-chronological) lifespan, measured as the number of daughters produced by each actively dividing mother cell; and (b) a chronological lifespan, measured as the ability of non-dividing cells to maintain viability over time. While it is convenient to consider these lifespans as totally distinct, they may not in fact be totally independent, since a period in stationary phase appears to be correlated with a reduction in replicative lifespan (Ashrafi et al., 1999). The increases in the lifespans of adultDrosophila and C. elegans, with increases in stress resistance, discussed above, relate to chronological lifespans that primarily reflect the ageing of postmitotic tissues. Dietary restriction, the only protocol that consistently extends lifespan in mammals, increases stress tolerance in rodents and leads to an increased capacity for the expression of molecular chaperone genes in response to acute stress (Heydari et al., 1996). Also, Ames dwarf mice, which have a pituitary deficiency, exhibit an extended lifespan and increased stress resistance (Brown-Borg et al., 1996). Again, though, these extensions to lifespan may involve primarily the chronological ageing of postmitotic tissues in the adult. In these non-dividing cells and tissues, essential components that become damaged cannot be diluted out through cell division but must, of necessity, be turned over and renewed. By elevating stress resistances, the gene products needed for such renewal will be elevated, with commensurate reduction in the steady-state levels of damaged cell components. A slower chronological ageing may be one manifestation of these events.

We reasoned that for yeast, as with Drosophila and C. elegans, stress resistance might relate more to the chronological lifespan of non-dividing cells than to the replicative senescence of budding cells. Nematodes in the dauer-larval state and yeast in stationary phase share certain similarities. Both are physiological states induced by nutrient limitation and characterized by high-stress resistances (thermotolerance, oxidative stress tolerance and tolerance of starvation; Lithgow, 1996; Werner-Washburne, et al., 1993; Werner-Washburne, et al., 1996). The long-term survival of S. cerevisiae, as with survival of the dormant (dauer-larval) state of C. elegans (Taub et al., 1999), is severely compromised by the loss of key antioxidant activities (Longo et al., 1996). These survivals are therefore influenced strongly by the long-term effects of ROS-mediated damage. The yeast cytosol also appears to become steadily more prooxidant during stationary maintenance (Jakubowski et al., 2000), a feature characteristic of ageing in a wide variety of species (Sohal and Weindruch, 1996).

In this study we have tried to identify some of the determinants of maximal chronological lifespan for laboratory S. cerevisiae strains, since these appear not to have been examined in detail from the standpoint of using yeast as a model of the ageing of postmitotic cells. Any model of chronological ageing is most likely to provide data relevant to the ageing process when physiological conditions are being employed that allow the attainment of long lifespans, lifespans close to the maximum attainable for the organism under study. At the outset we therefore considered it imperative to understand how aerobic survival of stationary yeast could be optimized. Only then, perhaps, could yeast survival provide a model of ageing processes that might affect the postmitotic tissues of metazoans. Throughout we have sought to establish simple protocols for optimizing and measuring stationary phase survival, protocols that would allow genetic studies of the determinants of chronological lifespan in non-dividing cells. We show that S. cerevisiae needs to be pre-adapted to efficient respiratory maintenance in order to attain maximal chronological lifespan. Such conditions probably maximize stress resistances, including those conferred by the general stress response activated when cells are deprived of glucose (Martinez-Pastor et al., 1996; Smith et al., 1998). A similar high expression of stress genes may produce the longevity of C. elegans mutants such as age1 (Lithgow, 1996), mutants that result from the inactivation of genes with a pro-senescence function. It might also contribute to the extended lifespans of several species subject to caloric restriction.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Strains, media and growth conditions

The S. cerevisiae strains used are listed in Table 1. Unless stated otherwise, they were grown aerobically at 30°C in liquid YP medium (1% Difco Yeast Extract, 2% Bacto peptone, 20 mg/l adenine), containing as carbon source either 2% glucose (YPD), 3% glycerol (YPGlycerol) or 2% ethanol (YPEthanol)(all concentrations % w/v). Flask cultures (flask volume:medium volume ratio of 5:1) were first inoculated from an overnight stationary YPD culture, at an initial density of 5×105 cells/ml. These cultures were then grown to stationary phase, by shaking in a rotary shaker at 200 rpm 30°C. They were confirmed as being in stationary phase by phase-contrast microscopy (large cells, <10% budded, with a single vacuole). They were then harvested, washed twice in sterile water, resuspended in water at time zero at 1×108 cells/ml and then shaken aerobically (flask volume:medium volume ratio of 5:1) at either 30°C or 37°C. At the indicated time intervals, three serial dilutions of these cells were prepared, these being plated on plate count agar plates (0.25% w/v Yeast Extract, 0.5% Bacto tryptone, 0.1% glucose, 0.9% agar). After growth for 5 days at 30°C, viable cells were determined by the counting of >300 colonies for each dilution, the plotted data being the mean and SD of these determinations. Survival is expressed as semilogarithmic plots, since over the linear part of the plot, when cell inactivation approximates to first-order kinetics, this survival may be expressed according to the equation:

  • equation image

where Nt is the number of viable cells after time t, N0 is the initial number of cells analysed (assumed as 100% viability at the time corresponding to the straight line extrapolated to 100% cells viable) and kD is the specific death rate.

Table 1. Yeast strains used in this study
StrainGenotypeReference
  • a

    A HO–gene bearing vector (pCY204; Adams et al., 1997) was used to prepare an isogenic diploid from FF18733.

W303-1aMATaura3-Δ1, trp1-Δ2, leu2-3,112, his3-11, ade2-1, can1-100 
msn2,msn4W303-1a msn2::HIS3 msn4::TRP1Martinez-Pastor et al., 1996
FY1679-28cMATaura3-52, his3-Δ200, leu2-Δ1, trp1-Δ63Piper et al., 1998
FF18733MATahis7-2, leu2-113, trp1-289, ura3-52, lys1-1,112Alseth et al., 1999
FF18733 diploidMATahis7-2, leu2-113, trp1-289, ura3-52, lys1-1,112This studya
MATα his7-2, leu2-113, trp1-289, ura3-52, lys1-1,112

Following the counting of colonies, plates were subsequently overlayed with 1% (w/v) triphenyltetrazolium chloride (Sigma) in 1% agar, then incubated for 1 h at 30°C in the dark so as to determine the fraction of the colonies that were respiratory-deficient. Respiring cells rapidly turn red in this test, whereas non-respiring colonies remain white. For all the strains analysed in this work, there was no appreciable accumulation of respiration-deficient cells amongst long-term survivors in stationary phase.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Yeast cells adapted to efficient respiratory maintenance have a longer chronological lifespan than cells grown by batch culture on glucose

As nutrients become limiting, S. cerevisiae enters stationary phase. Metabolic rates and protein synthesis decline, storage carbohydrates are synthesized and the cells become considerably more tolerant of potentially lethal stresses, such as exposure to oxidants and heat (Drebot et al., 1990; Lillie and Pringle, 1980; Werner-Washburne et al., 1993; Werner-Washburne et al., 1996). Survival in stationary phase is known to involve protein synthesis, as it is reduced considerably by the presence of cycloheximide (Granot and Snyder, 1993). It also requires efficient antioxidant defences, since it is severely compromised in sod mutant strains but assisted by conditions of low aeration (Longo et al., 1996).

Granot and Snyder have shown that incubating stationary cells in water, not spent medium, prolongs their survival (Granot and Snyder, 1991, 1993). For this reason we adopted the simple strategy of transferring cells to water at time zero in our measurements of chronological ageing. In our initial experiments we consistently observed that cells that had been initially cultured on a respiratory carbon source (YPGlycerol; see Methods) survived better in stationary phase than cells pre-grown on fermentative media (YPD) (Figure 1a). As a consequence, we conducted most of our subsequent experiments on YPGlycerol-grown cells.

thumbnail image

Figure 1. Chronological ageing of stationary S. cerevisiae FF18733 cells at 30°C (a) and at 37°C (b). Haploid (▪, •) and diploid (□) cells pregrown to early stationary phase on YPGlycerol (▪,□) or YPD (•), were transferred to water at time zero and their viability determined at the indicated times during subsequent aerobic maintenance (see Materials and methods ). The YPD-grown cells displayed much poorer survival in stationary phase compared to the YPGlycerol-grown cells (▪,•)

In higher organisms there appears to be an association between the losses of cell function during ageing and accumulations of ROS-induced damage to DNA, proteins and lipids (Migliaccio et al., 1999; Sohal and Weindruch, 1996). ROS production and therefore ROS-induced damages will be significantly reduced in yeast not fully adapted to respiratory maintenance, since it is the mitochondrial electron transport chain that provides the major endogenous source of ROS (Kowaltowski and Vercesi, 1999). It can be argued, therefore, that glucose-grown yeast, with its depressed respiratory activity, lowered endogenous ROS production and non-maximal stationary survival (Figure 1a), may not be a good model for processes that affect chronological ageing in more complex systems. Despite this, most of the published work of how yeast ages in stationary phase relates to cells initially grown by batch culture on glucose (Longo et al., 1996; Ashrafi et al., 1999; Jakubowski et al., 2000).

The improved stationary survival of YPGlycerol-, as compared to YDP-grown, cells (Figure 1a) may be attributed to the higher stress resistances of the former (see below) or their better adaptation to respiratory maintenance. The survival of stationary yeast cultures is known to be mainly through low-level respiratory metabolism (Werner-Washburne et al., 1993). Use of glucose as a carbon source substantially represses respiratory functions during the initial respirofermentative phase of rapid growth, when the sugar is being fermented (Lagunas, 1986). During aerobic batch growth on YPD the cells adapt to respiratory maintenance after the glucose becomes exhausted. However, even at this diauxic shift, when cells adapt to enable further respiratory growth on the ethanol generated in the earlier glucose fermentation, full respiratory function is generally not attained. Cells in the post-diauxic phase of growth on YPD, if transferred to fresh YPEthanol or YPGlycerol medium, will only adopt maximal YPEthanol or YPGlycerol growth rates after a lag period of several hours (data not shown). At no phase of batch glucose (YPD) culture, therefore, do they achieve optimal adaptation to respiratory maintenance. This may be due, in part, to the sensitivity of respiration to inhibition by ethanol (Piper, 1995).

Even for glycerol-grown cells, the phase of growth at time of starvation influences survival

It is known that yeast cells need to exit the cell cycle and enter a starvation-resistant state in order to maintain viability over long periods. Mutants that cannot attain this state, such as those with a constitutively high activity of protein kinase A, die extremely rapidly when starved (Thevelein, 1994). Even though lifespans in stationary phase are greater when cells are pre-grown on respiratory media (Figure 1a), there appears to be no published data on whether the phase of respiratory growth at the time of starvation influences subsequent survival. We therefore studied the 30°C stationary phase survival of cells transferred to water at the mid-exponential, late-exponential and early stationary phases of YPGlycerol growth. As shown in Figure 2, cells grown to early stationary phase displayed the longest survival. A sudden transfer to water in the mid-exponential phase of YPGlycerol growth evidently hinders attainment of the starvation-resistant physiological state needed for maximal long-term survival. Such cells, it seems, need to execute a slower entry into stationary phase in order to subsequently exhibit their maximal chronological lifespan. As a result of this observation, we now routinely grow YPGlycerol cultures to early stationary phase, transferring the cells to water at time zero in our measurements of chronological ageing (see Materials and methods ).

thumbnail image

Figure 2. Haploid FF18733 cells were transferred to water during the mid-exponential, late exponential and early stationary phases of 30°C growth on YPGlycerol (at Δ, 1.09×107 cells/ml; ▴, 1.23×107 cells/ml; □, 1.525×107 cells/ml and ▪ 4.02×107 cells/ml, respectively). Survival during aerobic 37°C maintenance was then determined at the indicated time points

Raising the temperature accelerates loss of viability in stationary yeast

As described above, it is YPGlycerol cultures grown to stationary phase that, in our hands, have provided the longest chronological lifespans. These cells show very little viability loss in their initial ‘robust’ phase of stationary maintenance. This ‘robust’ phase is followed, in turn, by a phase of much more rapid cell death (the ‘frail’ phase; see Figures 1a and 4) when inactivation adopts approximately first-order kinetics.

For cultures grown to stationary phase on YPD, the initial ‘robust’ phase of very little viability loss can be extremely short (Figures 1a, 3a). This contrasts with YPGlycerol-grown cells, where this ‘robust phase’ can last for up to 15–35 days at 30°C, depending on the strain (Figures 1a, 3b, 4). It can, as a result, take months rather than several days to conduct survival experiments on respiration-adapted (YPGlycerol-grown) cells maintained at 30°C. This rather negates one of the potentially attractive aspects of yeast as an ageing model, namely that data acquisition should not be a long, laborious process. The inactivation of respiration-adapted cultures is so slow at 30°C that we adopted 37°C for many of our subsequent experiments. At 37°C inactivation is considerably faster than at 30°C (cf. Figure 1a with 1b; also 3 with 4). We reasoned that maintaining cells at 37°C would allow measurements of chronological ageing over a realistic time-scale, but at a temperature still 1–2°C below the maximum for respiratory growth of the strains used in this work. The incubation of stationary cells at higher temperatures will almost certainly entail higher levels of endogenous ROS production, such that ageing processes measured at higher temperatures might be subject to a greater ROS influence. This is potentially not a serious problem, since attempts can always be made, if deemed appropriate, to reproduce interesting 37°C lifespan differences at a lower temperature.

thumbnail image

Figure 3. Stationary survival of the msn2,msn4 mutant (□) and its W303-1a parent (▪). Cells were grown to early stationary phase on either YPD (a) or YPGlycerol (b), transferred to water and then maintained aerobically at 37°C

thumbnail image

Figure 4. The strong influence of genetic background on stationary phase survival. Cells of the haploid strains FF18733 (▪), FY1679-28c (•) and W303-1a (▴) were grown to early stationary phase on YPGlycerol, transferred to water and their viability determined during aerobic maintenance at 30°C

The general stress response influences long-term aerobic survival of respiratory yeast

Much of the stress resistance of stationary phase yeast cells is due to the large number of genes induced by the general stress response. These are genes controlled by the Msn2p/Msn4p transcription factors, the trans-activators that bind to STRE promoter element sequences (Gorner et al., 1998; Martinez-Pastor et al., 1996; Ruis and Schüller, 1995; Schuller et al., 1994). The general stress response is switched-on in respiratory cultures and activated by the decrease in protein kinase A activity that accompanies the exhaustion of glucose in cultures growing by fermentation. STRE-regulated genes include several that encode antioxidant activities (Flattery-O'Brien et al., 1997; Ruis and Schüller, 1995). STRE-mediated gene expression is therefore important for maximizing resistance to oxidative stress, a stress resistance thought to be particularly important in ageing (see Discussion ). We investigated whether Msn2p/Msn4p-mediated gene expression is important for the improved long-term survival of stationary cells efficiently adapted to respiration. Respiratory adaptation is known not to be severely impaired in msn2,msn4 cells (Martinez-Pastor et al., 1996; Ruis and Schüller, 1995).

In full agreement with earlier investigations (Martinez-Pastor et al., 1996; Ruis and Schüller, 1995), we found that glucose (YPD)-grown msn2,msn4 mutant cells had a much-reduced starvation survival (Figure 3a). In contrast, msn2,msn4 mutant cells that had been grown to stationary phase on YPGlycerol exhibited better survival and died only slightly faster than the wild-type (Figure 3b). The effects of Msn2p/Msn4p loss on starvation survival are therefore less marked for glycerol-grown than for glucose-grown cells. While Msn2p/Msn4p-directed gene expression contributes to the attainment of a full chronological lifespan, it is more important for long-term survival of YPD-grown cells (Figure 3). We noted that the glucose-grown cells of the W303-1a genetic background used for this aspect of the work survived starvation slightly better than glucose-grown FF18733 cells (compare Figures 1a and 3a). In contrast glycerol-grown W303-1a exhibited a lower starvation survival than glycerol-grown FF18733 or FY1679-28c (Figure 4). It is clear therefore that there is a strong effect of genetic background on stationary survival. This may in part reflect different auxotrophies in the strains under investigation. In our hands, it is YPGlycerol-grown cultures of strain FF18733 that have displayed the longest stationary phase survival (Figures 1a, 4).

Influence of ploidy on the long-term survival of respiratory yeast

If single or double strand breaks in DNA, possibly due to ROS-induced damage, are a significant factor in losses of viability with time, these losses of viability should be influenced by ploidy. Pioneering work on the sensitivity of S. cerevisiae to both γ-irradiation and the hydroxyl radical-generating (Renzing et al., 1996) ultraviolet irradiation revealed that cell inactivation is markedly influenced by ploidy (see Piper et al., 1987, for literature). Cells of higher ploidies are more resistant than haploids to both ultraviolet and γ-irradiation, due in part to the decreased lethality of double-strand DNA breaks in diploids.

Using a HO gene-bearing vector, we prepared an isogenic diploid from our longest-lived haploid strain, FF18733 (Table 1). We then investigated if this increase in ploidy influenced chronological lifespan (Figure 1a). It transpired that the increase in ploidy exerted only a very small effect on survival. In four separate experiments (one of which is shown in Figure 1a) we consistently observed a very slightly reduced survival of diploid, as compared to isogenic haploid YPGlycerol-grown FF18733 cells. If potentially lethal DNA damage contributes to the critical losses of cell function that limit stationary survival, a higher ploidy might be expected to improve survival. This was not observed. Instead the slightly poorer survival of diploid as compared to haploid cells might possibly reflect the slightly larger size of the former, therefore their larger surface area (Piper et al., 1987). A larger surface area might prove to be a negative factor in the long-term maintenance of homeostasis, especially for frail cells.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The oxidative damage theory has steadily become one of the most widely-accepted views as to why organisms age (Halliwell and Gutteridge, 1998). It suggests that ageing, the progressive losses of function in tissues and cells over time, is due to accumulations of unrepaired oxidative damages to DNA, proteins and membrane lipids. There is significant evidence to suggest that, as organisms age, their cells progressively lose homeostasis and the cytosol becomes more prooxidant (Jakubowski et al., 2000; Sohal and Weindruch, 1996). It is still an open question, though, whether such increases in oxidative damage are the main cause of ageing, or merely biomarkers associated with what is primarily a genetically programmed, yet non-adaptive functional decline.

Cells that are maintaining themselves through the reduction of molecular oxygen inevitably suffer endogenous ROS production, the major source of these ROS being the mitochondrial electron transport chain. Most of these ROS are neutralized by antioxidant activities before they can cause cellular damage. However, ROS neutralization is never 100% efficient, such that all aerobic cells have a burden of oxidatively damaged DNA, lipids and proteins. It has been estimated that the DNA of the average rat cell attains 100 000 ROS ‘hits’ per day (Ames et al., 1993). Most of this DNA damage is efficiently repaired, yet oxidatively-damaged DNA, protein and lipid is detectable and its levels increase as organisms age. It is therefore reasonable to suppose: (a) that DNA repair, also the activities that catalyse turnover of aberrant proteins and lipids, might counteract the ageing process; and (b) that these repair processes might be less efficient in older individuals. Nevertheless, it is still very difficult to design direct experimental tests of such important precepts.

So important is the requirement to counteract ROS-induced damages in aerobic cells, that evolution will probably have provided strong selection for the most efficient repair systems. There are therefore strong reasons for suspecting that most of these repair activities will be widespread and highly conserved amongst eukaryotes. The ubiquitination system for intracellular protein turnover, the main system for turnover of aberrant or defective protein, is extensively conserved from yeast to man (Hershko and Ciechanover, 1998). Yeast molecular genetics has proved invaluable in unravelling this system. It has also led to the identification of activities that repair oxidative damage to DNA (Alseth et al., 1999), or which confer resistance to the toxic products of lipid peroxidation (Evans et al., 1997; Turton et al., 1997). The list of activities that might contribute to repair of oxidative damage in yeast is therefore extensive and a large number of S. cerevisiae mutants are now available that lack, or have conditional defects in, these repair activities. We reasoned that studying chronological ageing of these strains might provide indications as to the critical losses of cell function associated with the ageing of postmitotic cells and tissues (i.e. whether it is damages to DNA, to protein, or to membrane lipids that are paramount in the critical losses of function associated with ageing). Another potential advantage of using stationary yeast as a model of chronological ageing is that very large numbers of cells can be analysed rapidly, so that even at<0.1% viability survival can still be determined by standard microbiological dilution and plating techniques.

In this study we sought to:

  • Establish simple procedures for measuring chronological ageing in yeast, procedures that would facilitate the direct reproduction and comparison of ageing data obtained in different laboratories.

  • Identify factors needed to maximize the lifespan of stationary yeast.

  • Obtain chronological ageing data on W303-1a (the genetic background of many of the best-characterized stress gene mutants (Martinez-Pastor et al., 1996; Morgan et al., 1997); the S288C-derived strain FY1679-28c (isogenic to many of the gene deletant strains in the Euroscarf collection); also strain FF18733 (the parental strain of many of the mutants lacking activities of DNA repair; Alseth et al., 1999).

We found that it was cells grown to stationary phase on respiratory medium that display the longest chronological lifespans, possibly reflecting the fact that they are more stress-resistant. Yeast adapted to efficient respiratory maintenance is therefore potentially a better model of chronological ageing than cultures pre-grown on glucose. However, even with YPGlycerol-grown cultures, survival is highly strain-dependent. This may reflect longevity being a polygenic trait. Of the three strains that we investigated, YPGlycerol cultures of W303a were the shortest- and FF18733 the longest-lived (Figure 4).

Evolutionary pressure is exerted mainly on the young individuals of a species. As a result the force of natural selection declines dramatically with chronological age. This can provide a strong selection for prosenescence functions, genes that are beneficial in young individuals but detrimental in older individuals. The direct selection of mutants in such prosenescence functions, mutations that generate major increases in adult chronological lifespan, has hitherto been confined to C. elegans (Lithgow, 1996). This is because C. elegans is a habitual self-fertilizer, enabling screens for recessive mutations to be conducted much more readily than in outbreeders such as Drosophila and rodents. Yeast, with its stable haploid state, also provides a system that may allow direct selection of mutations in prosenescence functions. It is possible that such screening would solely yield mutants with elevated stress resistance (such as sir4-42, a mutation that increases both replicative lifespan and stationary survival; Kennedy et al., 1995). Alternatively, it might identify mutations that are detrimental for the propagation of the species yet beneficial for long-term stationary maintenance. One candidate for the latter class of mutation in S. cerevisiae is the hsp82 allele, encoding the E381K mutant form of the Hsp90 chaperone protein. This leads to slowed growth and a reduced replicative lifespan, yet increases in both stress resistance and stationary survival (Harris et al., 2000).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank H. Ruis, M. Bjoras and P. Seeburg for gifts of strains and plasmids. This project was supported by a BBSRC Project Grant 31/SAG09922 (to P.W.P.) and a postgraduate studentship (to N.H.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References