Saccharomyces cerevisiae cells lacking Isc1p, an orthologue of mammalian neutral sphingomyelinase 2, display a shortened lifespan and an increased sensitivity to oxidative stress. A lipidomic analysis revealed specific changes in sphingolipids that accompanied the premature ageing of Isc1p-deficient cells under severe calorie restriction conditions, including a decrease of dihydrosphingosine levels and an increase of dihydro-C26-ceramide and phyto-C26-ceramide levels, the latter raising the possibility of activation of ceramide-dependent protein phosphatases. Consequently, deletion of the SIT4 gene, which encodes for the catalytic subunit of type 2A ceramide-activated protein phosphatase in yeast, abolished the premature ageing and hydrogen peroxide sensitivity of isc1Δ cells. SIT4 deletion also abolished the respiratory defects and catalase A deficiency exhibited by isc1Δ mutants. These results are consistent with catabolic derepression associated with the loss of Sit4p. The overall results show that Isc1p is an upstream regulator of Sit4p and implicate Sit4p activation in mitochondrial dysfunction leading to the shortened chronological lifespan and oxidative stress sensitivity of isc1Δ mutants.
Sphingolipids are ubiquitous structural components of cell membranes, and their metabolites such as sphingosine, sphingosine-1-phosphate and ceramide play important roles in the regulation of several cellular processes. Sphingosine, a hydrophobic long-chain base that provides the backbone of sphingolipids, can be phosphorylated into sphingosine-1-phosphate by sphingosine kinases or acylated to ceramide by ceramide synthase. The fatty acids added to sphingosine can possess different chain lengths. Therefore, ceramides constitute a large and complex group of molecules that share a common structure. Ceramides with different fatty chain lengths and localizations may play distinct biological functions. Ceramide may also be generated through hydrolysis of complex sphingolipids; for example, in mammalian cells, sphingomyelin is hydrolysed by sphingomyelinases. Sphingosine and ceramide modulate cellular processes such as apoptosis, cell senescence and cell cycle arrest whereas sphingosine-1-phosphate plays a key role in proliferation, mitogenesis, cell migration and protection from apoptosis. Changes in the relative amounts of sphingosine-1-phosphate and sphingosine/ceramide are, therefore, major factors that determine cell fate (Hannun and Obeid, 2008).
In yeast, as described for mammalian cells (Won and Singh, 2006), changes in sphingolipid metabolism affect oxidative stress resistance and life span (Jiang et al., 2004; Aerts et al., 2006; 2008; Almeida et al., 2008). Furthermore, five genes (LAG1, YPC1, YSR3, IPT1 and LCB5) involved in sphingolipid metabolism are differentially expressed in both senescent and apoptotic yeast cells (Laun et al., 2005). We have recently shown that Isc1p, the yeast orthologue of mammalian neutral sphingomyelinase-2 (nSMase2) which acts on yeast inositol phosphosphingolipids, plays a key role in oxidative stress resistance and chronological lifespan, modulating redox homeostasis, iron levels and cell death by caspase-dependent apoptosis (Almeida et al., 2008). Mitochondrial deficits leading to an increased production of reactive oxygen species and accumulation of oxidative damages have been implicated in the ageing process (Murphy, 2009). Consistent with the shortened lifespan of isc1Δ mutants, several lines of evidence suggest that Isc1p is important for mitochondrial function. Isc1p is post-translationally regulated by translocation from the endoplasmic reticulum into mitochondria when aerobic respiration is induced (e.g. when glucose-grown cells reach the post-diauxic phase; Vaena de Avalos et al., 2004). Yeast mitochondria are enriched in α-hydroxylated phytoceramides generated by Isc1p and these lipids contribute to the normal function of mitochondria (Kitagaki et al., 2007). Cells lacking Isc1p preserve normal intrinsic mitochondrial functions, including intact mitochondrial DNA and oxygen consumption, but display a defective aerobic respiration in the post-diauxic phase associated with the incapacity to upregulate genes required for non-fermentable carbon source metabolism (Vaena de Avalos et al., 2005; Kitagaki et al., 2009).
Changes in the level of sphingolipid metabolites during cell ageing and the modulation of its downstream targets and signal transduction pathways remain poorly characterized. Several protein targets, including ceramide-activated protein kinases and phosphatases, protein kinase C, cathepsin D and JNK (Hannun and Obeid, 2008), have been suggested to mediate the effects of bioactive sphingolipids. In yeast, Sit4p is the catalytic subunit of the ceramide-activated protein phosphatase, a heterotrimeric complex that also includes Tpd3p and Cdc55p as regulatory subunits (Nickels and Broach, 1996). Sit4p is a serine-threonine protein phosphatase related to type 2A family of protein phosphatases (PP2A), and it has a high homology to other protein phosphatases, including the fission yeast PP2A and human protein phosphatase 6 that are involved in cell cycle regulation (Shimanuki et al., 1993; Bastians and Ponstingl, 1996). Sit4p has a role in the control of the Swi4p factor that regulates the transcription of G1 cyclin genes (Fernandez-Sarabia et al., 1992; Di Como et al., 1995) and is required for downregulation of cell functions that depend on Pkc1p, such as cell wall integrity pathway, cytoskeleton organization and ribosomal gene expression (Angeles de la Torre-Ruiz et al., 2002). Sit4p has also been functionally linked to the regulation of the ubiquitin-proteasome system (Singer et al., 2003), silencing in the subtelomeric region (Hayashi et al., 2005), monovalent ion and pH homeostasis (Masuda et al., 2000), nutrient signalling (Di Como and Arndt, 1996; Tate et al., 2006) and to reinitiate endocytosis under stress in cells deficient in amphiphysin function (McCourt et al., 2009). Interestingly, respiration is derepressed in sit4Δ cells grown in glucose medium, and these mutants are unable to grow under anaerobic conditions, indicating that energy produced by mitochondrial respiration is essential for their viability (Jablonka et al., 2006).
Here we show that the shortened chronological lifespan of isc1Δ cells submitted to a severe calorie restriction regime is accompanied by altered levels of bioactive sphingolipids, including an increase in dh-C26-Cer and phyto-C26-Cer. Consistent with ceramide activating protein phosphatases, loss of the ceramide-activated protein phosphatase Sit4p suppresses the shortened chronological lifespan, oxidative stress sensitivity and mitochondrial dysfunctions of isc1Δ cells.
Levels of sphingolipids during chronological ageing
Sphingolipids have been shown to play essential roles in both cell death and survival, and the balance between sphingosine-1-phosphate and sphingosine/ceramide is a key determinant of cell fate. To identify sphingolipid species that may be involved in the shortened chronological lifespan (CLS) of isc1Δ mutant cells, we analysed changes in the levels of 27 sphingolipid species, including long-chain sphingoid bases [dihydrosphingosine (DHS), phytosphingosine (PHS) and its 1-phosphate forms], and ceramides [dihydroceramides (dh-Cer), phytoceramides (phyto-Cer) and α-hydroxylated phytoceramides (αOH-phyto-Cer)] in Saccharomyces cerevisiae BY4741 and isc1Δ cells during chronological ageing. The main ceramide species found were αOH-C26-phyto-Cer and C26-phyto-Cer (C26 indicates an acyl chain of 26 carbon atoms). A map of sphingolipid metabolism in yeast is shown in Fig. 1.
These studies were performed using cells cultured to post-diauxic phase, washed and resuspended in water. These experimental conditions model a severe calorie restriction (CR) regime that induces a low metabolism stationary phase, resulting in the longest survival for wild-type yeast strains. Yeast optimized for chronological survival allows establishing a better correlation between ageing and cellular responses, including the modulation of bioactive molecules that regulate signalling pathways, stress resistance and mitochondria function (Fabrizio and Longo, 2003). The parental strain showed increasing levels of DHS (9.7-fold), DHS-1-P (18-fold) and PHS (1.8-fold) in cells aged during 10 days, but PHS-1-P decreased fivefold (Fig. 2). Total dihydroceramide levels of parental cells increased 1.5-fold by day 10 (Fig. 3A), with C24- (4.5-fold) and C26-dh-Cer (2.8-fold) increasing the most (Table S1; Fig. 3B). Regarding total phytoceramides (Fig. 3C), no significant changes were observed in parental cells by day 10, but individual changes in specific species were observed; thus, phyto-C16-Cer levels increased twofold, and phyto-C26-Cer levels decreased 2.2-fold (Table S2; Fig. 3D). The level of total α-hydroxylated phytoceramides increased 5.9-fold during ageing of parental cells (Fig. 3E). The greatest increases were seen in αOH-phyto-C18-Cer (4.4-fold), αOH-phyto-C24-Cer (4.4-fold) and αOH-phyto-C26-Cer (9.6-fold) (Table S3). These results demonstrate significant changes in several distinct species of yeast sphingolipids during chronologic ageing.
In contrast with parental cells, DHS and DHS-1-P levels did not change from day 0 to day 10 in isc1Δ cells. The final DHS levels were fivefold lower in isc1Δ mutants compared with parental cells (Fig. 2). The constitutive levels of PHS were elevated in isc1Δ cells, but then did not change over time, a dynamic similar to that observed in the parental strain. As observed in parental cells, PHS-1-P levels decreased from day 0 to day 10 in isc1Δ cells (Fig. 2).
Isc1p deficiency also affected changes in ceramide levels associated with chronological ageing (Fig. 3). Compared with aged parental cells, total dihydroceramides and phytoceramides were 1.5-fold and 2.8-fold higher in day 10 isc1Δ mutants. The analysis of the different ceramide species showed a two- to ninefold increase in the levels of dh-C16-Cer, dh-C20-Cer, dh-C26-Cer, phyto-C26-Cer and phyto-C26:1-Cer in aged isc1Δ cells, compared with parental cells, whereas the levels of dh-C24-Cer and phyto-C16-Cer were threefold lower (see Tables S1 and S2; Fig. 3A–D). Despite a 3.8-fold increase from day 0, the levels of total α-hydroxylated phytoceramides in day 10 isc1Δ mutants were twofold lower compared with parental cells (Fig. 3E). The most significantly decreased species were αOH-phyto-C16-Cer (twofold) and αOH-phyto-C26-Cer (fourfold) (see Table S3). Consistent with published data, the constitutive levels of several α-hydroxylated phytoceramides were significantly reduced in cells lacking Isc1p (Kitagaki et al., 2007). Thus taken together, there were several changes noted in senescent cells whose levels where differentially regulated in the isc1Δ strain.
One of these prominent lipids that showed differential changes was DHS-1-P. The low levels of DHS-1-P observed in isc1Δ cells suggest that it might be associated with the premature ageing of this mutant strain. To test this hypothesis, chronological lifespan and a lipidomic analysis were performed using cells deficient in Lcb4p, the major long-chain base kinase that phosphorylates DHS into DHS-1-P. The results showed that chronological lifespan was not affected in lcb4Δ cells (Fig. S1). As expected, in the lcb4Δ strain, the baseline levels of DHS-1-P (3.8-fold) and PHS-1-P (fivefold) were significantly reduced compared with the parental, whereas DHS (2.8-fold) and PHS (6.1-fold) were increased. Furthermore, in 10-day-old cells, DHS levels in the lcb4Δ mutants were similar to those of parental cells but DHS-1-P were 18-fold lower (Fig. 2). These results implicate Lcb4p in DHS phosphorylation during cell ageing but suggest that the shortened lifespan of isc1Δ cells is not associated with a decrease in DHS-1-P. The lipidomic analysis also showed higher levels of phytoceramides in lcb4Δ cells, compared with the levels of parental cells, both at day 0 and in aged cells (Fig. 3). All phytoceramide species contributed to this increase with the exception of phyto-C26-Cer and phyto-C26:1-Cer (Table S2; Fig. 3D).
The overall results indicate that there were significant changes specifically associated with the premature ageing of isc1Δ cells: (i) decreases in the levels of DHS and most αOH-phyto-ceramides, and (ii) increases in the levels of dh-C26-Cer and phyto-C26-Cer (Fig. 3B and D).
SIT4 disruption suppresses the shortened CLS and oxidative stress sensitivity of isc1Δ cells
Ceramide-activated protein phosphatase (CAPP) has been implicated in cell signalling elicited by ceramide. In yeast, Sit4p is the primary catalytic subunit of CAPP, being activated by a dose-dependent increase in ceramide levels (Nickels and Broach, 1996). Our results show that several ceramide species were increased in aged isc1Δ cells. These changes are probably implicated as a cause, rather than a consequence, of death/senescence, since cells remain metabolically active as suggested by the progressive increase of DNA fragmentation and caspase-like activity in isc1Δ cells from day 0 to day 14 of ageing (Almeida et al., 2008). This led us to wonder if Sit4p was involved downstream of the changes in ceramide. To begin to probe this, we first assessed whether loss of Sit4p could suppress the phenotypes of these mutant cells. Chronological lifespan and hydrogen peroxide resistance were measured in parental (BY4741), isc1Δ, sit4Δ and isc1Δsit4Δ cells. As previously shown (Almeida et al., 2008), isc1Δ cells displayed a shortened chronological lifespan associated with increased levels of protein carbonyls, a biomarker of oxidative damage (Fig. 4A and B). SIT4 disruption abolished the reduced lifespan of isc1Δ cells: at day 14, cell viability was 62% in parental cells, 1% in isc1Δ cells, 55% in sit4Δ cells and 46% in sit4Δisc1Δ cells. This protective effect was accompanied by a suppression of protein carbonylation in cells aged for 14 days (Fig. 4A and B, Fig. S2). Notably, loss of Sit4p extended the chronological lifespan of yeast cells: at day 35, cell viability in parental and sit4Δ mutants was 3% and 32% respectively.
The lifespan assay used in this study (cell grown to post-diauxic phase and then transferred into water) models a severe CR condition known to increase chronological survival (Fabrizio and Longo, 2003). To test if SIT4 deletion also increased lifespan under non-CR conditions, similar studies were performed using cells grown to post-diauxic phase on SC-2% glucose and survival of cells kept in the medium was followed over time. As expected, under these conditions, parental cells displayed a shorter lifespan (Fig. S3), compared with the observed under CR (Fig. 4A). The lifespan of isc1Δ cells also decreased under non-CR. Loss of Sit4p significantly increased chronological survival of both parental and isc1Δ cells, as observed for CR cultures. These results indicate that CR is not required for the effect of SIT4 deletion on lifespan extension.
The sit4Δ mutants were also hyper-resistant to H2O2, and SIT4 disruption suppressed the H2O2 sensitivity of isc1Δ cells (Fig. 5A): cellular viability was 20% in parental cells, 5% in isc1Δ cells, 80% in sit4Δ cells and 82% in isc1Δsit4Δ double mutants. The increased resistance to oxidative stress observed in sit4Δ and isc1Δsit4Δ cells was accompanied by a decrease in oxidative stress markers, namely intracellular oxidation [assessed using a molecular probe sensitive to reactive oxygen species, 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA)] and protein carbonylation (Fig. 5B and C). The increase in H2O2-induced intracellular oxidation and protein carbonylation was higher in isc1Δ cells (5.4- and 1.7-fold respectively), compared with the observed in parental cells (4.5- and 1.5-fold respectively). Remarkably, the protein carbonyl content in sit4Δ cells was always lower to that of parental cells. Moreover, SIT4 deletion significantly decreased oxidative stress markers in Isc1p-deficient cells. These results suggest that Sit4p is a key downstream mediator of the effects of Isc1p on lifespan and oxidative stress.
Based on the above, it became important to investigate if Sit4p is activated in response to deletion of ISC1. Sit4p is negatively controlled by the target of rapamycin complex 1 (TORC1), which is inhibited by rapamycin, and Sit4p in turn regulates the dephosphorylation of the Gln3p transcription factor and its translocation into the nucleus (Tate et al., 2006). To assess if Sit4p/CAPP activity was increased in isc1Δ cells, we measured the Gln3p-dependent MEP2 expression in parental (BY4741), isc1Δ, sit4Δ and isc1Δsit4Δ cells transformed with a MEP2-lacZ reporter. Consistent with an induction of the Gln3p reporter in isc1Δ cells, β-galactosidase activity increased 55% in this mutant strain, an effect that was suppressed by SIT4 disruption (Fig. S4). In addition, the analysis of isc1Δ microarrays data (Almeida et al., 2008) revealed that 2.8%, 33.3% and 53.2% of the genes differentially expressed in isc1Δ mutants are regulated by Gln3p based on direct, indirect or potential evidence respectively (Table S4). These results are consistent with an increase of Sit4p/CAPP activity in Isc1p-deficient cells.
SIT4 disruption relieves mitochondrial dysfunction and catalase A deficiency of isc1Δ cells
Cells lacking Isc1p display a mitochondrial dysfunction (Vaena de Avalos et al., 2005; Kitagaki et al., 2007; 2009), and respiration is derepressed in a sit4Δ mutant grown in glucose (Jablonka et al., 2006). Therefore, we raised the hypothesis that the effect of SIT4 deletion on the phenotypes of isc1Δ cells could be due to the modulation of mitochondrial function. To test this hypothesis, we measured oxygen consumption, cytochrome c oxidase (COX) activity and capacity to grow on glycerol as non-fermentable carbon source (Fig. 6). In exponential-phase cells, loss of Isc1p had no significant effect on oxygen consumption and COX activity, probably due to the fact that metabolism in parental and isc1Δ cells relies on fermentation for growth. In agreement with the catabolite derepression described for sit4Δ cells, oxygen consumption increased 60% and COX activity increased 9.2-fold in this strain. Moreover, COX activity increased 5.3-fold in sit4Δisc1Δ double mutants. In the post-diauxic phase, as expected, oxygen consumption and COX activity increased in parental cells. However, both oxygen consumption and COX activity were almost undetected in isc1Δ cells. The incapacity of isc1Δ cells to upregulate genes required for non-fermentable carbon source metabolism (Vaena de Avalos et al., 2005; Kitagaki et al., 2009) likely contributes to the decrease of oxygen consumption and the suppression of COX activation observed in isc1Δ mutants grown from exponential to post-diauxic phase. Notably, SIT4 disruption abolished the respiratory defect of isc1Δ cells: oxygen consumption in sit4Δisc1Δ double mutant was 57% of that in parental cells and COX activity was similar in both strains. In sit4Δ single mutants, oxygen consumption was not affected but COX activity increased 50% compared with that in post-diauxic phase parental cells. Consistent with the increase in mitochondrial function, sit4Δisc1Δ cells, in contrast with isc1Δ cells, were able to grow on a non-fermentable carbon source (Fig. 6C). It should be noted that catabolite derepression per se does not seem to account for the increased H2O2 resistance of sit4Δ cells since oxidative stress resistance was not affected in mig1Δ and hxk2Δ mutants that are also known to be derepressed (Fig. S5).
Antioxidant defences play a key role in cellular defence against oxidative stress and ageing by preventing oxidative damages (Moradas-Ferreira and Costa, 2000). We have previously shown that glutathione levels and the activity of catalase T (Ctt1p; cytosolic form) and superoxide dismutases (Sod) is not affected in isc1Δ cells (Almeida et al., 2008). However, Kitagaki et al. (2009) showed that isc1Δ cells fail to induce CTA1 gene expression in the post-diauxic phase. In agreement, we observed a decrease of 66% in total catalase activity in isc1Δ cells that was associated with a decrease in the activity of Cta1p, the catalase A form present in mitochondria and peroxisomes (Petrova et al., 2004). In addition, SIT4 disruption in Isc1p-deficient cells restored catalase activity by increasing Cta1p levels (Fig. 7A and B) and, thus, the capacity to detoxify H2O2. The overexpression of CTA1 gene in isc1Δ cells partially suppressed the shortened lifespan of this mutant (Fig. 7C and D). This result indicates that catalase A deficiency contributes to the premature ageing of cells lacking Isc1p, although it is not the only factor leading to this phenotype. A recent study showed that CTA1 overexpression shortens chronological lifespan whereas inactivation of catalases promotes longevity by increasing hydrogen peroxide levels (Mesquita et al., 2010). Under the severe CR conditions used in the current study, the viability of parental cells overexpressing CTA1 or transformed with the empty vector was 100% up to 14 days of ageing (Fig. 7C and D). The viability of these strains was not evaluated at longer time points in this study.
Isc1p, the yeast orthologue of mammalian neutral sphingomyelinase-2, is important for oxidative stress resistance and chronological lifespan (Almeida et al., 2008). The hydrolysis of inositolphosphosphingolipids by Isc1p generates ceramide, a signalling molecule conserved during evolution. However, the mechanisms and downstream mediators of these actions of Isc1p are not known. Elucidation of these mechanisms is further complicated by the interconnectivity of bioactive sphingolipids. As such, enzymes of sphingolipid metabolism function as a network that regulates the levels of individual bioactive lipids and their metabolic interconversion (Hannun and Obeid, 2008). The interaction between de novo sphingolipid biosynthesis versus Isc1p-mediated sphingolipid production is shown by the finding that ISC1 deletion is synthetically lethal with the lcb1-100 mutation (Cowart et al., 2006). Moreover, the LAC1 gene that encodes for a ceramide synthase component is upregulated in isc1Δ cells (Almeida et al., 2008). Given this metabolic interconnectivity, we performed a lipidomic analysis to identify specific changes in sphingolipids that might mediate the shortened lifespan of isc1Δ cells. To extend the monitoring of these changes, we used a severe CR regime (incubation in water) that results in the longest survival for wild-type yeast strains (Fabrizio and Longo, 2003). This treatment seems to activate an anti-ageing response similar to the observed under CR conditions induced by growing the cells in medium containing low glucose concentrations such as 0.05% or 0.5%, instead of the standard 2% (Smith et al., 2007).
In day 10 cells of the parental strain, the sphingoid bases had increases in DHS, DHS-1-P and PHS along with a decrease in PHS-1-P. Ceramides increased in total dh-Cer and αOH-phyto-Cer. Under these conditions, cell viability remained high; therefore, these changes may reflect an adaptive response to maintain stationary-phase viability. In isc1Δ mutants, lower levels of DHS and higher levels of dh-C26-Cer and phyto-C26-Cer are the changes that best correlate with loss of viability during ageing. The low levels of DHS observed in aged isc1Δ cells are probably a limiting factor for DHS-1-P formation, but the decrease in the levels of this long-chain base does not seem to be associated with cell ageing. Indeed, the formation of DHS-1-P during ageing was also suppressed in Lcb4p-deficient cells, which displayed a lifespan similar to that of parental cells. Other changes associated with ageing of isc1Δ mutants included an increased accumulation of dh-C16-Cer and dh-C20-Cer, and lower levels of phyto-C16-Cer, αOH-phyto-C16-Cer and αOH-phyto-C26-Cer. The decrease in αOH-phyto-Cer is in agreement with recent data showing that yeast mitochondria are enriched in these lipids, which are generated by Isc1p and contribute to mitochondrial function (Kitagaki et al., 2007). Mitochondrial dysfunction has been implicated in the ageing process (Murphy, 2009). Thus, alterations in the levels of α-hydroxylated phytoceramides may also function in the modulation of lifespan.
Ceramides containing different length fatty acids seem to have distinct effects in cell physiology, either by affecting the biophysical properties of the membrane lipid bilayer or by interacting with specific downstream components in signalling pathways (Hannun and Obeid, 2008). Here we present evidence that Sit4p, a serine-threonine phosphatase that is also a component of ceramide-activated PP2A, functions downstream of Isc1p. SIT4 disruption suppressed the shortened lifespan and H2O2 sensitivity of isc1Δ cells and decreased oxidative stress markers in these mutants. SIT4 disruption also increased chronological lifespan and H2O2 resistance of parental cells. Sit4p seems to be specifically associated with oxidative stress since its mutation or deficiency confers resistance to thiol-specific oxidants (López-Mirabal et al., 2008), t-butyl-hydroperoxide and menadione (a generator of superoxide radicals), but not to heat shock or cadmium (Fig. S6).
Sit4p functions downstream of Tor1p, a conserved nutrient-responsive protein kinase associated with lifespan regulation in organisms ranging from yeast to mice (Di Como and Arndt, 1996). Inhibition of TOR contributes to the beneficial effect of CR on cell longevity (Wei et al., 2008). In agreement with published data, our results show that chronological survival increased in cells submitted to CR. However, SIT4 deletion further delayed ageing of calorie restricted cells, indicating that Sit4p functions, at least in part, through a mechanism that is not modulated by CR.
Several lines of evidence were consistent with CAPP activation in isc1Δ cells. The Gln3p transcription factor is controlled by Sit4p and the expression of a Gln3p-regulated LacZ reporter increased in ISC1-deleted cells. Moreover, an in silico analysis suggested that many of the genes differentially expressed in isc1Δ cells (Almeida et al., 2008) are probably regulated by Gln3p. Since TOR1 or GLN3 deletion significantly increases chronological lifespan (Powers et al., 2006; Wei et al., 2009), it is likely that the activation of Gln3p associated with Isc1p deficiency contributes to the premature ageing phenotype of this mutant strain. Interestingly, TORC2 complex controls ceramide synthase activity as well as the turnover of complex sphingolipids through the activation of Slm1p and Slm2p that inhibit Isc1p (Tabuchi et al., 2006; Dickson, 2008). The characterization of changes in TOR signalling in isc1Δ cells will contribute to our understanding of the role of sphingolipids in the regulation of oxidative stress responses and chronological lifespan.
The identification of a key role of Sit4p activation downstream of Isc1p suggests that the most relevant changes in sphingolipids are the elevations in the levels of ceramides which would be capable of activating of Sit4p. However, significant more studies are needed to conclusively establish the specific downstream effects mediated by specific sphingolipids. Notably, the reciprocal changes in two subgroups of ceramides (increased phytoceramides and decreased αOH-phytoceramides) illustrate the complexity and interconnection of sphingolipid metabolism as one would have expected primarily a decrease in ceramides in the ISC1 deletion.
During exponential growth on glucose, yeast cells are under catabolite repression: oxygen consumption is low and respiratory metabolism is repressed (Santangelo, 2006). Sit4p is involved in catabolite repression and, therefore, respiration is derepressed in sit4Δ cells (Jablonka et al., 2006; Jin et al., 2007; Fig. 6A and B). Notably, SIT4 disruption alleviated mitochondrial dysfunction and catalase A (Cta1p) deficiency of isc1Δ cells. The expression of CTA1 gene is repressed by glucose but isc1Δ cells are unable to derepress the CTA1 gene after the diauxic shift when glucose decreases to low levels (Kitagaki et al., 2009). The induction of Cta1p activity probably promotes oxidative stress resistance and lifespan in isc1Δsit4Δ double mutants by decreasing oxidative damages. Indeed, CTA1 overexpression partially suppressed the shortened lifespan of isc1Δ cells.
The identification of proteins related to mitochondria function that are Sit4p targets will contribute to our understanding of how Sit4p controls chronological lifespan. Published data show that Sit4p interacts with two mitochondrial proteins, namely the Atp3p subunit of the ATP synthase and Nde1p, the external NADH dehydrogenase. Moreover, a mutation in Atp3p (ATP3-1) or the deletion of Nde1p are both capable of increasing chronological lifespan of yeast cells (Li et al., 2006; Francis et al., 2007). Thus, Sit4p may affect mitochondria function through modulation of mitochondrial proteins.
Mitochondria have been also linked to replicative lifespan, which is measured by counting the number of daughter cells generated by a mother cell before cell division stops. Mitochondria function is required for the maintenance of an age asymmetry between the mother and daughter yeast that prevents the inheritance of damaged molecules by the daughter cell. A retrograde response signals mitochondria dysfunction and induces cellular adaptations that compensate for age changes. This response links metabolism with genome stability during ageing (Jazwinski, 2005). It should be noticed that Sit4p also regulates replicative lifespan. Specifically, Sit4p dephosphorylates Sir3p, which plays a key role in maintaining a repressed chromatin structure near telomeres, and loss of Sit4p decreases replicative lifespan (Hayashi et al., 2005). These results point to distinct mechanisms by which Sit4p controls replicative and chronological lifespan of yeast cells.
In summary, our data suggest that Isc1p functions upstream of Sit4p through the modulation of ceramide levels and implicate Sit4p in mitochondrial dysfunction, premature ageing and oxidative stress sensitivity. SIT4 deletion restores mitochondrial function, increasing the longevity of isc1Δ cells (Fig. 8). These results offer new insights on the regulation of redox homeostasis by sphingolipid signalling, a link that has been suggested in several studies (Won and Singh, 2006).
Yeast strains, plasmids and growth conditions
The S. cerevisiae strains used in this study are listed in Table 1. Yeast cells were grown aerobically at 26°C in a gyratory shaker (at 140 r.p.m.), with a ratio of flask volume/medium volume of 5:1, to early exponential phase (OD600 = 0.6) or to post-diauxic phase (OD600 = 7–8). The growth media used were YPD [1% (w/v) yeast extract, 2% (w/v) bactopeptone, 2% (w/v) glucose], YPGlycerol [1% (w/v) yeast extract, 2% (w/v) bactopeptone, 4% (v/v) glycerol] or synthetic complete (SC) drop-out medium containing 2% (w/v) glucose 0.67% yeast nitrogen base without amino acids and supplemented with appropriate amino acids (80 mg histidine l−1, 400 mg leucine l−1, 80 mg tryptophan l−1). For ISC1 disruption in the sit4Δ strain, a deletion fragment containing URA3 and the flanking regions of ISC1 was amplified by PCR using pYES2 (Invitrogen) and the following primers: F (ATTTGCGCTTTCCGCGTAAAAAGGGAAAAAAAGCAGATATCTAGCTTTTCAATTCAATTC) and R (TCAGTAATTTTTTTACATATGCTAAAGAAAATCGATAATATTAGTTTTGCTGGCCGCATC). Cells were transformed by electroporation, and isc1Δsit4Δ double mutants were selected in minimal medium-GLC [0.67% (w/v) yeast nitrogen base without amino acids, 2% (w/v) glucose supplemented with appropriate amino acids (40 mg histidine l−1, 80 mg leucine l−1, 40 mg methionine l−1)]. For CTA1 gene overexpression, BY4741 and isc1Δ cells were transformed by electroporation with pPGK-M28-I (empty vector) or pCTA1-GFP (Petrova et al., 2004) and selected in minimal medium lacking uracil. For Gln3p activity, yeast cells were transformed by electroporation with YCpMEP2-lacZ plasmid (Marini et al., 1997), and selected in minimal medium lacking uracil.
Table 1. Saccharomyces cerevisiae strains used in this study.
Mata his3Δ1, leu2Δ0, met15Δ0, ura3Δ0
BY4741 sit4Δ::KanMX4 isc1Δ::URA3
BY4741 carrying YCpMEP2-lacZ
isc1Δ carrying YCpMEP2-lacZ
sit4Δ carrying YCpMEP2-lacZ
isc1Δsit4Δ carrying YCpMEP2-lacZ
BY4741 carrying pPGK-M28-I
BY4741 carrying pPGK-M28-I CTA1-GFP
isc1Δ carrying pPGK-M28-I
isc1Δ carrying pPGK-M28-I CTA1-GFP
ESI/MS/MS analyses of endogenous phyto- and dihydrosphingosine bases, their phosphates, α-hydroxylated phytoceramide, and non-hydroxylated phyto- and dihydroceramide species were performed on a Thermo Finnigan TSQ 7000 triple quadrupole mass spectrometer, operating in a Multiple Reaction Monitoring positive ionization mode, as described (Kitagaki et al., 2007). The mass of each species was normalized to total lipid phosphate.
Stress resistance and chronological lifespan
Stress resistance was determined in cells treated with H2O2, menadione, tert-butyl hydroperoxide, cadmium or submitted to heat shock (45°C), as indicated in figure legends, during 30 min. Chronological lifespan was assayed as described previously (Almeida et al., 2008). Briefly, overnight cultures were diluted to OD600 = 0.5 and grown for 24 h (to post-diauxic phase). Cells were then centrifuged at 4 000 r.p.m. for 5 min, washed twice with water, resuspended in water and incubated at 26°C for the indicated times. Cell viability was determined by standard dilution plate counts on YPD medium containing 1.5% agar. Colonies were counted after growth at 26°C for 3 days. Viability was expressed as the percentage of the colony-forming units.
Enzymatic activities and oxygen consumption
All the procedures were carried out at 0–4°C. Yeast cells were harvested by centrifugation. For enzyme activities, yeast extracts were prepared in 50 mM potassium phosphate buffer (pH 7.0) containing protease inhibitors (Complete, Mini, EDTA-free Protease Cocktail Inhibitor Tablets; Boehringer Mannhein), by vigorous shaking of the cell suspension in the presence of glass beads for 5 min. Short pulses of 1 min were used, with 1 min intervals on ice. Cell debris was removed by centrifugation at 13 000 r.p.m. for 15 min and protein content was determined by the method of Lowry, using bovine serum albumin as a standard. Catalase activity was analysed in situ, in the presence of 3,3′-diaminobenzidine tetrahydrochloride, using the H2O2/peroxidase system (Conyers and Kidwell, 1991) or spectrophotometrically as described previously (Aebi, 1984). COX activity was determined as previously described (Poyton et al., 1995), by measuring cytochrome c oxidation. Oxygen consumption rate was measured for 7.5 × 108 cells in the culture media using an oxygen electrode (Oxygraph, Hansatech). Data were analysed using the Oxyg32 V2.25 software. For the β-galactosidase assay, yeast cells containing YCpMEP2-lacZ were grown in SC-glucose medium lacking uracil to exponential phase, and the activity was measured as previously described (Ausubel et al., 1998), with the following modifications: a cellular extract was prepared, as described above, in 100 mM Tris-HCl, 1 mM DTT, 10% (v/v) glycerol, and 40 µg of total protein was used in the assay.
Oxidative stress markers
Protein oxidation was determined by immunodetection of protein carbonyls, as previous described (Costa et al., 2002). Protein content of cellular extracts was estimated as described above. Protein carbonylation assays were performed by slot blot analysis or SDS-PAGE using 12.5% gels followed by protein electrotransfer onto nitrocellulose membranes using rabbit IgG anti-DNP (Dako, Glostrup, Denmark) at a 1:5000 dilution, as the primary antibody, and goat anti-rabbit IgG-peroxidase (Sigma, St. Louis, MO, USA) at a 1:5000 dilution, as the secondary antibody. Immunodetection was performed by chemiluminescence, using a kit from GE Healthcare (RPN 2109). Quantification of carbonyls was performed by densitometry (GS-800, Bio-Rad). Intracellular oxidation was measured by using the oxidant-sensitive probe 2′,7′-dichlorofluorescein diacetate (H2DCF-DA), as described before (Almeida et al., 2008), and values were expressed as fluorescence units mg−1 protein.
Data are expressed as mean values ± SD of at least three independent experiments. Values were compared by Student's t-test. The 0.05 probability level was chosen as the point of statistical significance throughout.
We are grateful to the Lipidomic Core Facility at Medical University of South Carolina for sphingolipid analysis and to Dr Bruno André (Université Libre de Bruxelles; Belgium) and Dr. Manfred J. Schmitt (Universität des Saarlandes, Germany) for generously providing plasmids used in this study. This project was financially supported by FCT, POCTI/BCI/45066/2002, FSE-FEDER, Grants GM63265 and GM43825, and MUSC Lipidomics Core NIH C06 RR018823. A.D.B. (SFRH/BD/37129/2007), T.A. (SFRH/BD/9358/2002) and H.O. (SFRH/BPD/20280/2004) were supported by FCT fellowships.