|Home | About | Journals | Submit | Contact Us | Français|
Loss of the protein kinase Sch9p increases both the chronological life span (CLS) and the replicative life span (RLS) of Saccharomyces cerevisiae by mimicking calorie restriction, but the physiological consequences of SCH9 deletion are poorly understood. By transcriptional profiling of an sch9Δ mutant, we show that mitochondrial electron transport chain genes are upregulated. Accordingly, protein levels of electron transport chain subunits are increased and the oxygen consumption rate is enhanced in the sch9Δ mutant. Deletion of HAP4 and CYT1, both of which are essential for respiration, revert the sch9Δ mutant respiratory rate back to a lower-than-wild-type level. These alterations of the electron transport chain almost completely blocked CLS extension by the sch9Δ mutation but had a minor impact on the RLS. SCH9 thus negatively regulates the CLS and RLS through inhibition of respiratory genes, but a large part of its action on life span seems to be respiration independent and might involve increased resistance to stress. Considering that TOR1 deletion also increases respiration and that Sch9p is a direct target of TOR signaling, we propose that SCH9 is one of the major effectors of TOR repression of respiratory activity in glucose grown cells.
The SCH9 gene of the yeast Saccharomyces cerevisiae encodes a serine-threonine protein kinase with a catalytic domain very similar to that of human AKT1 (12). This kinase participates in a conserved signaling network involving upstream acting kinases: Pkh1/2p, which phosphorylate its activation loop, and Tor1/2p, which target its hydrophobic motif (34, 36, 38).
There are several lines of evidence that Sch9p plays an important role in glucose signaling in the budding yeast. Early work showed parallelism and complementarity of SCH9 signaling with the cyclic AMP-dependent protein kinase (PKA) pathway, which signals hexose abundance (25, 29, 37, 40). Overexpression of SCH9 suppresses the growth defect of many PKA pathway component mutations (37), and the sch9Δ null mutant is synthetically lethal with each of gpr1Δ, gpa2Δ, and ras2Δ mutations, which act upstream of PKA (25, 29, 37). It was also recently shown that Sch9p integrates nutrient signals with cell size regulation. In fact, the sch9Δ mutation was one of the most potent modifiers of cell size identified in a genome-wide screen for pathways coupling cell growth and division in yeast (18). A subsequent study showed that Sch9p is an activator of ribosomal protein and ribosomal biogenesis regulons and is required for carbon source modulation of cell size (19).
SCH9 is also a negative regulator of both the chronological life span (CLS) and the replicative life span (RLS), both of which are influenced by glucose availability (12, 21, 22, 24). The CLS is the relative survival over time of yeast cells grown to stationary phase in liquid culture, while the RLS is defined as the number of daughter cells produced by a given yeast mother cell before senescence (11, 16, 21). In addition to the sch9Δ deletion, targeting of either the TOR or cyclic AMP/PKA pathway also extends the CLS and RLS (12, 27, 33). As well as these genetic manipulations, calorie restriction (CR) lengthens the S. cerevisiae CLS and RLS (9, 17, 24, 27). Although this is currently controversial, respiration is suggested to be critical for the physiology of life span because (i) the effect of CR on the RLS is mimicked by overexpressing HAP4, which activates expression of electron transport chain genes and is blocked by deletion of CYT1 (28), and (ii) affecting respiration by blocking mitochondrial transcription, ATP synthesis, or the electron transport chain (atp2Δ, coq3Δ, or ndi1Δ deletions or antimycin A treatment) causes a reduction of the yeast CLS (5, 10). HAP4 overexpression also causes an increased CLS, suggesting that increasing respiration may promote RLS and CLS extension (32). Recently, it was convincingly shown that both the sch9Δ and the tor1Δ mutations mimic CR and that reduction in TOR signaling increases yeast respiratory activity (4, 24). However, the physiological effects of perturbing Sch9p signaling remain largely uncharacterized. Here we show that the sch9Δ mutation is required for a derepression of respiration in glucose-grown cells and that this causes part of the extended CLS and RLS phenotype of the sch9Δ mutant.
All Saccharomyces cerevisiae strains used in this study are BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0) and its isogenic derivatives (39). The sch9Δ heterozygous mutant was provided by Mike Tyers (18). The hap4Δ and cyt1Δ mutants were obtained from the Research Genetics yeast knockout library (Research Genetics). Double mutants were generated by standard yeast genetic manipulations. All strains were cultivated in SD-complete containing 2% dextrose, following standard protocols (1). The plasmid pMT3569, encoding hemagglutinin-Sch9p was a gift of Mike Tyers (19). This construct was subjected to PCR site-directed mutagenesis to introduce a kinase-disabling mutation (K441A). Transformations were performed with the lithium acetate method (13).
All cultures used for RNA extractions were grown to an optical density at 600 nm OD600 of 1, and the glucose concentration in the medium was quantitated with a glucose assay kit (Sigma) to ensure that culture conditions were the same for all analyzed strains. Total RNA was extracted by the hot-phenol extraction method (6) with the difference that glass beads (Sigma) were added to samples. Poly(A)+ RNAs were purified using the Invitrogen MicroFast Track kit (Invitrogen). For labeling, 5 μg of poly(A)+ RNA was reverse transcribed using oligo(dT)21 in the presence of Cy3 or Cy5-dCTP (Perkin-Elmer-Cetus/NEN) and Superscript II reverse transcriptase (Invitrogen). Reverse transcription reaction mixtures were treated with a mix of RNase H (Sigma) and RNase A (1 mg/ml) for 15 min at 37°C. The labeled cDNAs were purified with the QIAquick PCR purification kit (Qiagen).
Saccharomyces cerevisiae Y6.4k cDNA arrays containing 6,218 genes printed in duplicate were obtained from University Health Networks (Toronto, Canada) and used for all hybridizations. The microarray slides were prehybridized for 2 h at 42°C with DIGeasy hybridization buffer containing yeast tRNA and salmon sperm DNA and subsequently washed with 0.1× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and air dried. The slides were then hybridized with labeled cDNAs overnight at 42°C in DIGeasy hybridization buffer with yeast tRNA and salmon sperm DNA. The slides were washed twice with SSC-0.2% sodium dodecyl sulfate (SDS), once with 0.1× SSC-0.2% SDS, and three times with 0.1× SSC. A final wash with isopropanol was performed, and slides were air dried and scanned with a ScanArray 5000 scanner (Perkin-Elmer-Cetus) at 10-μm resolution. Signal intensity was quantified with QuantArray software (Perkin-Elmer-Cetus), and final normalization and inspection of the data were done with the GeneSpring package.
DNA extracts used for Southern blotting were obtained by the rapid glass beads method (15). Genomic DNA was digested with HaeIII and run on a 1% agarose gel. DNA was then transferred by capillarity action to a Zeta-Probe nylon membrane (Bio-Rad). To obtain Northern blots, total RNA (80 μg) was electrophoresed on a 7.5% formaldehyde-1% agarose gel and transferred by blotting to a Zeta-Probe nylon membrane (Bio-Rad).
Southern probes were a XbaI-PstI restriction fragment of the ACT1 gene from plasmid pGEM-ACT1 (41) and a COX2 probe generated by PCR. Northern probes were all generated by PCR. All probes were labeled by random priming with the RediPrime kit (Amersham Biosciences). Southern and Northern blot hybridizations were carried out as described previously (30).
Whole-cell extracts were obtained by bead beating in a buffer containing 0.1% NP-40, 250 mM NaCl, 50 mM NaF, 5 mM EDTA, and 50 mM Tris-HCl, pH 7.5. Complete protease inhibitor cocktail (Roche) was added just before use. Protein concentrations were determined by the standard Bio-Rad protein assay (Bio-Rad). Proteins were separated on a 10% SDS-polyacrylamide gel and transferred to a polyvinylidene difluoride membrane (Millipore). Antibodies were prepared in phosphate-buffered saline-0.05% Tween 20-5% skim milk powder. Rabbit polyclonal antibodies directed against Atp1p, Atp2p, and Atp7p were kind gifts of Jean Velours. They were used at dilutions of 1:100,000, 1:100,000 and 1:10,000, respectively. Antibodies directed against Cox4p, Cox5p, and Cyt1p were obtained from Alexander Tzagoloff and were all used at a dilution of 1:1000 (2). Horseradish peroxidase (HRP)-conjugated anti-rabbit and anti-mouse secondary antibodies (Santa Cruz) were used at 1:10,000. The HRP signal was revealed with Immobilon HRP substrate (Millipore).
Cells were grown to an OD600 of 1, spun down, and kept on ice. One hour before the measurements, cells were resuspended in SD-2% glucose and put on a shaker at 30°C for 1 hour. OD600 measurements and cell counts were performed immediately before putting cells in the oxygen sensing setup. The oxygen depletion rate was monitored by using a lab-built respirometer comprising a polarographic dissolved oxygen probe in a glass syringe with constant agitation (26). Oxygen consumption rates were expressed as percent O2 per minute per 1 × 106 cells.
Wild-type or sch9Δ mutant cells expressing Cox4p-red fluorescent protein (RFP) (3) were stained with DAPI (4′,6′-diamidino-2-phenylindole) (10 μg/ml in phosphate-buffered saline plus 2% glucose) for 10 min at room temperature, washed twice, and visualized with a Leica DM-IRE2 inverted microscope with a 63× objective and a 10× projection lens. Pictures were acquired with a Sensys charge-coupled-device camera. Images were manipulated with the Openlab software (Improvision).
We profiled the sch9Δ mutant by microarray analysis in the haploid and diploid states (Fig. (Fig.1A).1A). We also compared sch9Δ/sch9Δ diploid mutants transformed with a wild-type SCH9 or a kinase-dead SCH9 version (kd-SCH9; K441A). The profiles obtained are highly reproducible (Fig. (Fig.1A)1A) and reveal that numerous genes are upregulated in the sch9Δ mutant (643 genes; P < 0.05). The list of upregulated genes was systematically tested for enrichment in gene ontology (GO) terms from the SGD database (http://www.yeastgenome.org/). A randomized set of GO annotations was used to set the threshold P value (P < 1 × 10−4). Most of the nonredundant GO terms significantly associated with the upregulated gene list are related to energy derivation and ATP synthesis (Table (Table1).1). The other GO terms are associated with chromatin organization and RNA polymerase II-dependent transcription. Genes involved in mitochondrial functions that are upregulated in the sch9Δ mutant are listed in Table Table22 (P < 0.05). Almost all components of the five respiratory complexes are upregulated with fold changes that range from 1.2 to 2.5. In addition to canonical members of the respiratory complexes, cytochrome c oxidase chaperones (PET100, PET191, COX14, and COX17), heme biosynthesis enzymes (HEM4, HEM12, and HEM15), holocytochrome c synthases (CYC3 and CYT2), and mitochondrial ribosomal subunit genes are upregulated (Table (Table2;2; Fig. Fig.1B).1B). These observations suggest that the sch9Δ mutation affects respiratory physiology. We therefore decided to study the mitochondrial compartment of the sch9Δ mutant in more detail.
The sch9Δ mutant has a slow-growth phenotype that is a hallmark of petite mutants. However, our sch9Δ strain could grow on ethanol, glycerol, pyruvate, and succinate as sole carbon sources, suggesting that it is not a petite (data not shown). Since ABF2, MGM101, KGD2, and ACO1 are upregulated in the sch9Δ mutant (Table (Table2)2) and these genes are involved in mitochondrial DNA (mtDNA) maintenance, we also tested whether the sch9Δ mutant displays modified mtDNA abundance (7, 31, 35, 41). For this, Southern blot analysis with probes for a mitochondrial gene (COX2) and a single-copy nuclear gene (ACT1) was performed. As expected, the mtDNA/genomic DNA ratio increased as the culture density increased, but we observed no differences in mtDNA amounts between the sch9Δ mutant and the wild-type strain (Fig. (Fig.2A).2A). We also microscopically examined the mitochondrial compartment in sch9Δ and observed that it is more densely stained and speckled than in wild-type cells, where DAPI staining and Cox4p-RFP distribution are diffuse and reticulated (Fig. (Fig.2B2B).
We validated that the increased expression of the respiratory regulon in sch9Δ cells corresponds to enhanced respiratory activity. First, Western blotting of cultures at different cell densities established that protein levels of Atp1p, Atp2p, Atp7p, Cox4p, Cox5p, and Cyt1p are higher in sch9Δ cells throughout culture growth (Fig. (Fig.3A).3A). At all time points evaluated, glucose availability was the same in the wild-type and sch9Δ cultures (data not shown). Since the expression of electron transport chain mRNAs and proteins is increased in the sch9Δ mutant, we measured oxygen consumption rates in the sch9Δ/sch9Δ diploid and found that it consumes oxygen faster than the isogenic wild-type BY4743 cells, as predicted (Fig. (Fig.3B3B).
We intersected the list of genes of the sch9Δ mutant that were highly significantly upregulated (P < 0.001) with ChIP/CHIP and with conserved cis-regulatory element data for all transcription factors available (http://fraenkel.mit.edu/yeast_map_2004/) (8, 14). Hap4p-bound intergenic regions and conserved CCAAT box elements were highly enriched upstream of genes whose expression was perturbed in sch9Δ cells (Table (Table3).3). Accordingly, we observed a modest but significant increase in expression of all CCAAT-binding factor subunits in the sch9Δ mutant (Tables (Tables11 and and2;2; Fig. Fig.1B).1B). It thus seems likely that the CCAAT box-binding complex is influenced by SCH9. Since HAP4 overexpression has been associated with an extended RLS, we investigated the role of HAP4 in the sch9Δ mutant. We first confirmed that the hap4Δ and sch9Δ/hap4Δ strains are unable to grow on nonfermentable carbon sources (data not shown). We then tested the effect of the hap4Δ mutation on the sch9Δ upregulation of respiratory genes and observed that it is reverted (Fig. (Fig.4A).4A). Protein levels of many components of the electron transport chain and of the ATP synthase were decreased in the sch9Δ/hap4Δ mutant compared to the wild type (Fig. (Fig.4B).4B). Increased oxygen consumption by sch9Δ cells was reverted by deletion of HAP4 or of the cytochrome c1 gene (CYT1) to levels close to that of a rho0 strain (Fig. (Fig.4C4C).
As previously observed for CR cells, the sch9Δ mutation is associated with increased respiratory activity. We thus tested whether the CLS and RLS of the sch9Δ mutant are dependent on intact electron transport chain regulation or function. For this, we assessed the RLS of the sch9Δ mutant with or without HAP4 and CYT1. A hap4Δ single mutant has essentially the same replicative longevity as the wild-type congenic strain, while the sch9Δ mutant and the sch9Δ/hap4Δ double mutant have significantly increased RLSs (P = 1.8 × 10−4 and P = 8.2 × 10−3, respectively) (Table (Table4;4; Fig. Fig.5A).5A). Similarly, the sch9Δ and sch9Δ/cyt1Δ mutants both show RLS extension compared to wild-type cells (P = 4.4 × 10−4 and P = 8 × 10−3, respectively) (Table (Table4),4), while we confirmed the previous finding that the cyt1Δ mutation causes a significant decrease in the RLS (Fig. (Fig.5B)5B) (28). Since the effect of shutting down respiration in sch9Δ cells was small in terms of RLS, we tested whether the hap4Δ or cyt1Δ mutation could perturb the inability of the sch9Δ mutation to extend the CLS. As previously reported, we observed a dramatic extension of the CLS in the sch9Δ mutant (Fig. 5C and D). In contrast, the CLS of hap4Δ and cyt1Δ mutants was dramatically reduced compared to that of wild-type cultures, while the sch9Δ/hap4Δ and sch9Δ/cyt1Δ double mutants had slightly elongated CLSs compared to the hap4Δ and cyt1Δ single mutants (Fig. 5C and D). Interestingly, the sch9Δ/hap4Δ and sch9Δ/cyt1Δ double mutants fail to phenocopy hap4Δ and cyt1Δ mutants, respectively, in both CLS and RLS assays. Our data show that sch9Δ has respiration-independent effects on both the CLS and RLS. However, it seems that respiration or its proper regulation is required for full CLS extension in the sch9Δ mutant, while RLS extension by this mutation seems mostly respiration independent.
The sch9Δ mutation causes significant changes in expression of 643 genes, including a systematic increase in mitochondrial respiratory chain gene expression. Here, we have shown that mitochondrial activity of the sch9Δ mutant is increased. This finding was previously reported for CR yeast and in the tor1Δ strain (4).
Increasing respiration has drawbacks for cellular physiology by producing detrimental by-products such as reactive oxygen species (ROS) that limit the life span of cells in stationary-phase cultures, likely by oxidizing proteins, lipids, and nucleic acids (16). Besides ROS, ethanol also limits the CLS of yeast cells (9). Previous work has shown that the sch9Δ deletion causes an increase in heat and oxidative stress resistance, and SOD2 was reported to be a downstream effector of the sch9Δ deletion increase in oxidative stress resistance and CLS (10, 12). Our expression profiling experiments show that many stress response genes are upregulated in sch9Δ cells (Table (Table5):5): we confirm increased expression of the mitochondrial superoxide dismutase gene SOD2 and of several heat shock protein genes, and we show increased transcription of ethanol-degrading enzyme genes ADH1, -2, and -3. This provides additional molecular evidence for the sch9Δ mutant's resistance to stress and accelerated ethanol depletion (9).
Our study shows that not only is the sch9Δ mutant a genetic mimic of CR, but it also recapitulates many molecular hallmarks of CR in yeast. In fact, inactivation of SCH9 provokes shunting of the fermentative metabolism of yeast to a more respirative mode and promotes the heat shock response and the turnover of ROS and alcohol. We find that a large part of the SCH9 effect on the CLS is mediated by respiration, since deletion of HAP4 or CYT1 extensively abrogated sch9Δ-dependent extension of the CLS while RLS extension of sch9Δ was only slightly reduced by blocking respiration. Therefore, our data suggest that the sch9Δ mutation influences the CLS and RLS by both respiration-dependent and -independent mechanisms. Our data also indicate that the conditions required for extending the CLS and RLS are probably encountered by different means in the sch9Δ mutant, since disrupting respiration has a more dramatic effect on the CLS than on the RLS.
Our data support a recent finding that has challenged the relationship between CR-induced RLS extension and respiration (20). Those authors demonstrated that CR causes RLS extension in respiration-deficient yeast strains. Since sch9Δ is thought to be a CR genetic mimic and since it does not fully rely on respiration to promote CLS and RLS, our work further weakens the evidence linking CR-induced RLS and increased respiration.
The respiration-independent effect of the sch9Δ mutation might be explained by it being resistant to oxidative and heat stress through increased ROS and ethanol turnover, which we detect in expression analysis of exponentially growing cultures. Thus, reduced production of detrimental by-products during the growth phase caused by a respirative metabolism and faster turnover of ethanol and ROS might lead to increased survival of the sch9Δ mutant during CLS analysis.
It is believed that Sch9p, Tor1p, and PKA kinases belong to highly integrated signaling pathways that are all involved in nutrient sensing (19, 24). SCH9 and TOR inhibition increase CLS and RLS and are thought to mimic CR (23, 24). Recent findings show that Sch9p is the direct target of the TORC1 complex and that the tor1Δ strains display increased CLS and respiration (4, 38). This, in addition to our findings, supports the view that the TOR-SCH9 pathway feeds into the regulation of respiration and that Sch9p might be one of the major effectors of TOR repression of respiratory activity, as it is the major effector of translational activation by the TOR pathway. The CCAAT box-binding complex (Hap2/3/4/5p) is a likely downstream effector of this nutrient-dependent signal transduction pathway. Hap4p was originally proposed to be the regulatory moiety of the CCAAT box-binding complex, and the TOR-SCH9 pathway could thus reduce respiratory chain expression in high-glucose conditions by repressing Hap4p by a direct or indirect mechanism; in contrast, when glucose is depleted, decreased Sch9 signaling would induce the respiratory regulon as observed in our sch9Δ mutant (19).
It has been proposed that one of the major mitochondrial targets of TOR is translation (4). Here, we suggest that the TOR-SCH9 pathway not only impinges on mitochondrial translation but also affects transcriptional activity of the nuclear respiratory regulon through Sch9p.
We thank Mario Jolicoeur and Steve Hisiger for technical help with the oxygen consumption experiments. We thank Hervé Hogues for help with setting up a database for statistical analysis of gene ontology and transcription factor binding enrichment. We are grateful to Mike Tyers for providing the sch9Δ heterozygous mutant and the hemagglutinin-tagged Sch9 plasmid. Thanks go to Alexander Tzagolof for providing anti-Cyt1p, anti-Cox4p, and anti-Cox5p antibodies and to Jean Velours for anti-Atp1p, anti-Atp2p, and anti-Atp7p. The Cox4-RFP-expressing plasmid was provided by Benjamin Glick.
This work was supported by grants from the Canadian Institute for Health Research (CIHR). H.L. was supported by a CIHR fellowship. H.L also acknowledges partial support from an NCIC and CNRC fellowship. This is NRCC publication 49523.
Published ahead of print on 9 May 2008.