Epigenetic Silencing Mediates Mitochondria Stress-Induced Longevity

Cell Metabolism

Article Epigenetic Silencing Mediates Mitochondria Stress-Induced Longevity Elizabeth A. Schroeder,1,2 Nuno Raimundo,1 and Gerald S. Shadel1,2,* 1Department

of Pathology of Genetics Yale University School of Medicine, New Haven, CT 06520, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cmet.2013.04.003 2Department

SUMMARY

Reactive oxygen species (ROS) play complex roles in aging, having both damaging effects and signaling functions. Transiently elevating mitochondrial stress, including mitochondrial ROS (mtROS), elicits beneficial responses that extend lifespan. However, these adaptive, longevity-signaling pathways remain poorly understood. We show here that Tel1p and Rad53p, homologs of the mammalian DNA damage response kinases ATM and Chk2, mediate a hormetic mtROS longevity signal that extends yeast chronological lifespan. This pathway senses mtROS in a manner distinct from the nuclear DNA damage response and ultimately imparts longevity by inactivating the histone demethylase Rph1p specifically at subtelomeric heterochromatin, enhancing binding of the silencing protein Sir3p, and repressing subtelomeric transcription. These results demonstrate the existence of conserved mitochondria-to-nucleus stress-signaling pathways that regulate aging through epigenetic modulation of nuclear gene expression.

INTRODUCTION As multifunctional organelles, mitochondria influence aging, lifespan, and healthspan through a variety of mechanisms. Perhaps the most recognized is the production of reactive oxygen species (ROS), byproducts of mitochondrial respiration that damage macromolecules to promote cell dysfunction, apoptosis, and aging (Balaban et al., 2005). However, ROS also function as signaling molecules that regulate physiological processes (Finkel, 2011; Sena and Chandel, 2012) and participate in conserved signal transduction pathways, including the target of rapamycin (TOR) and insulin-sensing pathways that regulate longevity (Longo et al., 2012; Pan et al., 2011; Raimundo et al., 2012; Zarse et al., 2012). Additionally, studies in model organisms show that increasing ROS levels can positively influence health- and lifespan (Lee et al., 2010; Pan et al., 2011; Schulz et al., 2007; Yang and Hekimi, 2010; Zarse et al., 2012). Chronological lifespan (CLS) of the budding yeast, Saccharomyces cerevisiae, is measured by assaying the amount of 954 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.

time that cells survive in the late post-diauxic and stationary phases of culturing that follow exponential growth (Longo et al., 2012). Yeast CLS studies, which model postmitotic cellular aging in higher eukaryotes, have led to the identification of conserved pathways and processes involved in lifespan regulation (Fabrizio and Longo, 2007; Kaeberlein, 2010; Longo et al., 2012). We recently demonstrated that elevating mitochondrial ROS (mtROS) during yeast exponential growth elicits an adaptive response that reduces ROS in post-diauxic and stationary phases and extends CLS (Pan et al., 2011). Interventions that elevate mtROS in C. elegans also extend lifespan (Lee et al., 2010; Schulz et al., 2007; Yang and Hekimi, 2010; Zarse et al., 2012), and exercise in mammals is believed to improve health via increasing mtROS (Ristow et al., 2009). These examples indicate that hormetic adaptation to ROS is an emerging paradigm in lifespan and healthspan regulation (Gems and Partridge, 2008; Ristow and Schmeisser, 2011; Tapia, 2006). However, the signaling pathways that sense ROS and coordinate hormetic changes in cellular function to curtail aging and promote longevity remain largely unknown. Hormetic ROS might extend lifespan by increasing stress resistance or decreasing ROS levels late in life (Gems and Partridge, 2008). However, other cellular processes may be targets of mtROS signaling. In particular, aspects of telomere function determine replicative capacity and impact lifespan (Blasco, 2005). In yeast, telomeres consist of a repetitive TG1–3 sequence bound by several highly conserved proteins that regulate telomere length, transcription, and packaging (Grunstein, 1997). The silent information regulator (Sir) proteins maintain telomeric heterochromatin to repress transcription (Rusche et al., 2003). Histone modifications adjacent to telomeres regulate the spread of silencing, termed telomere position effect (TPE), by enhancing or preventing Sir complex binding (Altaf et al., 2007; Dang et al., 2009; Kozak et al., 2010; Park and Lustig, 2000). Yeast impaired in specific histone modifications that support TPE have a shorter replicative lifespan (Dang et al., 2009), and decreased heterochromatin formation at telomeres and subtelomeres in mammals induces telomere dysfunction to trigger apoptosis or senescence (Blasco, 2007). However, the signaling pathways that regulate histone modifications to support subtelomeric silencing are not fully understood. Additionally, although substantial progress has been made toward understanding the relationships between telomere dysfunction and replicative lifespan, how telomere function might influence postmitotic aging remains unknown.

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 1. The Histone Demethylase Rph1p Mediates Hormetic mtROS Longevity Signaling (A) Strategy to identify factors that regulate the global transcriptional response to hormetic mtROS signaling. (i) Identification of genes induced under conditions of elevated mtROS. (ii) Promoter sequences of induced genes were analyzed to identify enriched motifs. (iii) Known transcription factor binding sequences were scanned for the presence of motifs similar to those enriched in mtROS-induced gene promoters. (B) Results of promoter analysis. The AG5 motif is significantly enriched in promoters of genes induced by mtROS and is similar to the DNA sequences recognized by the proteins listed. The percentages of mtROS-induced genes containing the exact motifs recognized by the candidate proteins are indicated to the right of the motif. See also Table S1. (C) CLS of wild-type (DBY2006) and rph1D strains treated with 50 mM menadione (MD) or vehicle (not treated, nt) during logarithmic growth. In all CLS curves, data points represent the mean of three biological replicates inoculated from single colonies ± SEM. (D) ROS levels in wild-type and rph1D strains treated with menadione (MD+) or vehicle ( ) during exponential growth. Samples were analyzed at day 1 of CLS. In all graphs, data points represent the mean of three biological replicates ± SEM unless otherwise indicated. ‘‘ns’’ designates p > 0.05, *p % 0.05, and **p % 0.01. (E) RT-PCR of transcripts induced by mtROS. Wildtype and rph1D were treated with 50 mM menadione (MD) or vehicle (nt) for 2 hr in exponential phase, and the expression of three genes selected from the list of 45 mtROS-responsive transcripts (Figure 1A) was analyzed. See also Figure S1.

RESULTS

Herein, we focused on the process of hormetic mtROS signaling and adaptation in yeast, reasoning that our studies would reveal conserved mechanisms of ROS sensing and longevity regulation. We show that mtROS signal via Tel1p and Rad53p, homologs of mammalian ATM and Chk2, uncovering a function for the DNA damage response in communicating mitochondria functional status to the nucleus. This signaling pathway inhibits the histone demethylase Rph1p specifically at subtelomeric regions, which enhances transcriptional silencing to extend CLS. In yeast and mammals, telomere dysfunction signals changes in mitochondrial biogenesis and function (Nautiyal et al., 2002; Sahin et al., 2011). Thus, our results demonstrate that two-way communication pathways exist between mitochondria and telomeres that regulate aging and longevity.

The H3K36 Demethylase Rph1p Mediates Hormetic Responses to mtROS Reduced TOR complex 1 (TORC1) signaling extends lifespan in many model organisms (Kenyon, 2005), and we showed previously that mtROS hormesis contributes to this mode of CLS extension in yeast (Pan et al., 2011). Furthermore, transient treatment with a sublethal dose of the redoxcycling compound menadione during exponential growth generates mtROS (Castro et al., 2008) and extends yeast CLS (Pan et al., 2011) (Figure S1A). Menadione treatment and reduced TORC1 signaling decrease cellular ROS levels in post-diauxic and stationary phases, supporting that hormetic adaptation to ROS occurs under these conditions (Bonawitz et al., 2007; Pan et al., 2011). To identify factors that mediate mtROS longevity signaling, we crossed the results of two published microarrays from conditions that increase mtROS in yeast (1 mM menadione and 100 nm rapamycin, a TORC1 inhibitor) and identified the commonly induced genes (Figure 1Ai and Table S1) (Gasch et al., 2000; Hardwick et al., 1999). Rather than Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc. 955

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 2. A Mitochondrial Signal Extends CLS

Superoxide

(A) Western blot analysis of Sod2p (MnSOD) in PTDH::SOD2 and wild-type (DBY2006). Samples were collected at an OD600 of 1.0. MnSOD band intensity was normalized to actin, and the DBY2006 protein level was set to 1. Values under the blot represent the mean ± SEM for three biological replicates. (B) CLS in wild-type and PTDH::SOD2 strains following exponential phase treatment with menadione (MD) or ethanol (nt). (C) CLS in wild-type and sod2D strains as in (B). (D) CLS in wild-type and rho as in (B).

systematically analyzing all identified genes, which may not show individual robust contributions to mtROS adaptation, we searched first for transcriptional regulators that globally coordinate these gene expression responses and second for signaling events that determine their activity (Figures 1Aii, 1Aiii, and S1B). We identified a promoter motif, AG5, enriched in these genes (Figure 1B) (Romer et al., 2007) and transcriptional regulators that bind motifs similar to AG5 (Abdulrehman et al., 2011). Candidate transcriptional regulators included Gis1p and Msn2p/Msn4p, which have known roles in oxidative stress resistance, nutrient sensing, and CLS (Wei et al., 2008). Strains lacking Msn2p or Msn4p demonstrated a moderately reduced adaptive CLS response relative to the wild-type strain (72% and 81% of the menadione-dependent CLS extension observed in wild-type; Figures S1C and S1D), indicating nonredundant contributions of each transcription factor. Combined deletion of MSN2 and MSN4 abrogated menadione-induced CLS extension to 53% of that observed in the wild-type strain (Figure S1E). This additive effect suggests that Msn2p and Msn4p have partially redundant functions in mtROS adaptation, consistent with their overlapping regulation of gene expression in response to stress and downstream of TORC1 signaling (Gasch et al., 2000; Wei et al., 2008, 2009). Deletion of GIS1 reduced menadione-dependent CLS extension to 47% of the response seen in the wild-type (Figure S1F), indicating that it also participates in mtROS adaptation. Gis1p and Msn2p/Msn4p cooperatively contribute to yeast CLS extension under conditions of reduced TORC1 signaling and caloric restriction (Wei et al., 2008). However, a gis1D/msn2D/msn4D mutant strain retained an adaptive CLS response that was 47% of wild-type (Figure S1G), identical to that of the gis1D single mutant. This epistatic relationship indicates that Gis1p, Msn2p, and Msn4p partially contribute to the mtROS adaptive response in the same genetic pathway. Only deletion of the histone 3 lysine 36 956 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.

(H3K36) demethylase Rph1p completely abrogated CLS extension (Figure 1C), eliminated the hormetic decrease in ROS at day 1 of CLS (Figure 1D), and impaired expression of mtROS-responsive genes (Figure 1E). Rph1p also contributed to the CLS extension observed in a tor1D strain (Figure S1H), consistent with Rph1p participating in TORC1-regulated mtROS hormesis (Pan et al., 2011). Intrigued by the notion that mtROS adaptation might involve epigenetic changes in nuclear gene expression, Rph1p became our primary focus. Hormetic Menadione Treatment Induces Mitochondria Superoxide to Extend CLS Overexpression of SOD2, which encodes the mitochondria matrix superoxide dismutase, curtails CLS extension in tor1D yeast, supporting that elevated mtROS initiates an adaptive signal (Pan et al., 2011). To determine if hormetic menadione treatment elicits adaptive responses via a similar mechanism, we measured CLS in strains with different mtROS detoxification capacities. We utilized a strain in which the strong TDH3 promoter (Ko¨tter et al., 2009) drives SOD2 expression (PTDHSOD2) (Figure 2A) and a sod2D strain, with increased or decreased mtROS detoxification capabilities, respectively. The PTDH-SOD2 strain demonstrated reduced CLS extension relative to the wild-type strain (Figure 2B). In contrast, sod2D amplified CLS extension (Figure 2C). Menadione treatment also failed to extend CLS in a rho strain, which lacks mitochondrial DNA (mtDNA), and hence is unable to respire (Figure 2D). We therefore conclude that menadione is affecting CLS via alteration of mtROS levels (i.e., superoxide). Tel1p and Rad53p, Components of the Nuclear DNA Damage Response, Contribute to mtROS Hormesis The DNA damage checkpoint kinase Rad53p (Chk2 in mammals) phosphorylates Rph1p in response to DNA damage (Kim et al., 2002) (Figure 3A). Subsequent Rph1p-mediated chromatin modifications permit transcription of the DNA repair enzyme PHR1 (Liang et al., 2011). Mec1p (ATR in mammals) activates and phosphorylates Rad53p (Pellicioli et al., 1999), and phosphorylation of Rph1p in response to DNA damage is Mec1p dependent (Kim et al., 2002); therefore, we investigated

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 3. Hormetic mtROS Longevity Signaling Requires the Tel1p and Rad53p DNA Damage Response Kinases (A) Schematic of DNA damage signaling in Saccharomyces cerevisiae. (B) CLS of sml1D and rad53D/sml1D strains treated with 50 mM menadione (MD) or vehicle (nt) during exponential growth. (C) ROS levels in sml1D and rad53D/sml1D strains treated with menadione (+) or vehicle ( ) during exponential growth as in Figure 1D. Samples were collected at day 1 of CLS. (D) CLS of wild-type (DBY2006) and tel1D strains as in (B). See also Figure S2.

the requirement for the Rad53p and Mec1p kinases in mtROS signaling. Deletion of SML1 suppressed the lethality of rad53D and mec1D and allowed us to examine their role in mtROS signaling and adaptation (Zhao et al., 1998). Although Rad53p was required for mtROS-mediated CLS extension (Figure 3B) and hormetic reduction of ROS at day 1 of CLS (Figure 3C), Mec1p was not (Figure S2A). Therefore, we investigated the involvement of the related kinase Tel1p, which has partially overlapping functions with Mec1p (Figure 3A) but unique roles in telomere maintenance (Harrison and Haber, 2006). Additionally, the mammalian homolog of Tel1p, ATM, senses ROS (Guo et al., 2010), making Tel1p an attractive candidate to mediate mtROS signaling in yeast. Menadione treatment did not extend CLS in a tel1D strain (Figure 3D), implicating Tel1p as an upstream kinase in the adaptive mtROS pathway. The lack of adaptation to mtROS in rph1D, rad53D, and tel1D strains is not due to endogenous differences in ROS levels or an inability of these strains to elevate mtROS upon menadione treatment during exponential growth (Figure S2B). Thus far, these results indicate that Tel1p, Rad53p, and Rph1p together mediate mtROS signaling and adaptation to extend yeast CLS. Hormetic Menadione Treatment Activates DNA Damage-Responsive Kinases in the Absence of Nuclear DNA Damage Tel1p phosphorylates Rad53p upon telomere dysfunction or DNA damage (Sanchez et al., 1996). Hormetic menadione treatment caused an upward mobility shift in Rad53p, but this shift was intermediate to that observed following treatment with a comparable dose of methylmethane sulfate (MMS), a direct DNA damaging agent (Figures 4A, left, and S3A–S3C). The menadione-induced mobility shift was abolished by phospha-

tase treatment and did not occur in a tel1D strain, yet MMS-induced Rad53p phosphorylation was not affected by tel1D (Figure 4A, right). Rad53p contains multiple Tel1p/Mec1p target phosphorylation sites, characterized by an (S/T)Q motif (Kim et al., 1999; Lee et al., 2003). Analysis of (S/T)Q phosphoprotein profiles revealed that hormetic menadione treatment induced phosphorylation of a specific subset of (S/T)Q-containing targets compared to an MMS treatment that had similar effects on cell growth and toxicity (Figure 4B). This distinct profile, combined with partial phosphorylation of Rad53p in response to menadione, indicated to us that hormetic mtROS signaling via Tel1p and Rad53p is distinct from canonical Mec1p-mediated DNA damage signaling. We additionally demonstrated that the transcriptional outcomes of hormetic mtROS signaling are distinct from a DNA damage transcriptional signature (DDTS), which corresponds to genes similarly altered by 0.02% MMS and 170 Gy of ionizing radiation (Gasch et al., 2001) (Figure S3D and Table S2). RT-PCR analysis of three well-characterized DNA damageinduced transcripts confirmed that they are not induced by hormetic menadione treatment (Figure 4C). Instead, further characterization of menadione-induced genes identified several involved in oxidative stress responses that were elevated in wild-type, but not rph1D, strains (AHP1, CTA1, CUP1-1, LOT6, MXR1, TRX1, TRX2, TRX3, TSA1, and YPR1). Collectively, these likely explain the hormetic decrease in ROS observed at day 1 of CLS (Figure S3E and Table S2). Furthermore, although menadione treatment extends CLS via modulating mtROS levels (Figure 2), hormetic menadione treatment did not induce oxidative DNA damage in a wildtype strain or an ogg1D strain that lacks the oxidized purine repair enzyme and is therefore hypersensitive to oxidative DNA damage (Karahalil et al., 1998) (Figure 4D). Finally, mtROS adaptation does not require DNA double-strand break detection or signaling mediated by Mre11p, a component of the MRN/MRX complex (Nakada et al., 2004) (Figure S3F). Together, these data support that mtROS produced during hormetic menadione treatment activates specific DNA damage response kinases without inducing oxidative nuclear DNA damage or a canonical nuclear DNA damage response. Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc. 957

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 4. mtROS Activate a Noncanonical, Tel1p-Dependent DNA Damage Response (A) Left: western blot analysis of Rad53p phosphorylation in the wild-type treated with 50 mM menadione (MD+), 0.01% methylmethane sulfate (MMS+), or ethanol ( ) for 30 min. Right: wild-type and tel1D strains treated with menadione, MMS, ethanol, and/or calf intestinal phosphatase (CIP). (B) Protein phosphorylation at (S/T)Q sites in response to 30 min treatment with 50 mM menadione or 0.01% MMS. Molecular weights (kDa) are indicated on the left. Starred bands indicate proteins that are phosphorylated in response to MMS, but not menadione, treatment. (C) RT-PCR of transcripts induced by DNA damage. Wild-type and rph1D were treated with 50 mM menadione (MD), 0.01% MMS, or vehicle (nt) for 30 min in exponential phase. See also Table S2. (D) Quantification of 8-hydroxyguanosine (8-oxoG) formation in wild-type and ogg1D. Cells were treated for 30 min prior to DNA extraction. Values were normalized to untreated wild-type, which was set to 1. Data points represent the mean of three biological replicates inoculated from single colonies ± SEM. ‘‘ns’’ designates p > 0.05, *p % 0.05, **p % 0.01. See also Figure S3.

958 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.

Hormetic mtROS Signaling Inactivates Rph1p to Alter Subtelomeric Chromatin Rad53p-dependent phosphorylation of Rph1p causes it to dissociate from chromatin, elevating H3K36me3 levels to regulate transcription (Figure 5A) (Liang et al., 2011). We examined Rph1p phosphorylation in response to mtROS and found that hormetic menadione treatment induces a partial Rph1p upward mobility shift relative to treatment with MMS (Figures 5B, left, and S4A), mirroring the partial activation of Rad53p under these conditions (Figure 4A). Rph1p phosphorylation in response to menadione, but not MMS, also required Tel1p and Rad53p (Figure 5B). These data show that mtROS signaling via Tel1p and Rad53p induces phosphorylation of Rph1p, which may subsequently alter the chromatin binding capacity of Rph1p and H3K36me3 levels. In yeast, histone methylation and acetylation contribute to heterochromatin formation at the ribosomal DNA (rDNA) gene cluster, mating loci, and subtelomeric regions (defined as within 25 kb of telomeric repeat DNA) (Barton and Kaback, 2006; Dang et al., 2009; Millar and Grunstein, 2006; Santos-Rosa et al., 2004). Yeast deficient in heterochromatin formation at these silenced loci exhibit decreased replicative lifespan (Dang et al., 2009; Kaeberlein et al., 1999), yet it is unknown how silencing at these regions influences chronological aging. H3K36me3 is implicated in both transcriptional repression and activation within euchromatic regions (Wagner and Carpenter, 2012), but the role of this modification at silenced loci has not been described. Since Rph1p phosphorylation regulates its chromatin binding capacity (Liang et al., 2011) and Rph1p is phosphorylated in response to mtROS (Figure 5B), we investigated whether mtROS signaling altered the association of Rph1p with silenced regions of the yeast genome. Chromatin immunoprecipitation (ChIP) of Rph1-HA at day 1 of CLS (when adaptive response to menadione would be evident; Figure S4B) revealed that, in the absence of hormetic menadione treatment, Rph1p binding at all examined silenced genes was enhanced relative to ACT1 (Figures 5C and S4C). However, Rph1p binding at subtelomeric regions was specifically altered in response to hormetic mtROS signaling in a manner dependent on Rad53p (Figures 5C and S4C). Consistent with loss of Rph1p, H3K36me3 levels increased at subtelomeric genes in an mtROS- and RPH1dependent manner, suggesting that mtROS signaling regulates the H3K36me3 demethylase activity of Rph1p (Figures 5D and S4D). The requirement for H3K36 demethylation in adaptive CLS extension is specific to Rph1p, as deletion of Jhd2p, which also demethylates di- and trimethyl H3K36 (Tu et al., 2007), did not affect mtROS adaptation (Figure S4E). These data suggest that, although Rph1p may function at several silenced regions in chronologically aged yeast, hormetic mtROS signaling via Rad53p specifically targets Rph1p activity at subtelomeric regions, eliciting epigenetic changes that promote lifespan extension. Signaling by mtROS Represses Transcription of Subtelomeric Genes We next investigated the adaptive transcriptional consequences of Rph1p-mediated epigenetic changes in response to mtROS signaling. We performed gene expression microarrays of wildtype and rph1D yeast, that had been treated with menadione or a vehicle control during exponential growth, at day 1 of CLS. To

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 5. mtROS Regulate Rph1p to Modify Subtelomeric Heterochromatin (A) Schematic of Rad53p-dependent regulation of Rph1p chromatin binding and histone demethylase activity. (B) Western blot analysis of Rph1-HA phosphorylation in wild-type, tel1D, and sml1D/rad53D strains treated with 50 mM menadione or 0.01% MMS. Samples for western blots were collected after 30 min of treatment during exponential growth. (C) ChIP of Rph1-HA at a control region (ACT1) and two subtelomeric genes (YPS6 and YEL077C) in a wild-type or sml1D/rad53D background at day 1 of CLS. Data points represent the mean of two independent ChIP experiments; error bars represent the range. Values were normalized to the wildtype untreated ACT1 sample, which was set to 1. (D) ChIP using antibodies against trimethyl H3K36 or total H3 in wild-type or rph1D as in (C). See also Figure S4.

focus the analysis on Rph1p-dependent transcriptional changes, we excluded transcripts that were altered in an rph1D strain treated with menadione. Additionally, we compared the remaining transcripts with those altered in an untreated rph1D strain. Since mtROS signaling to Rad53p essentially inactivates Rph1p demethylase activity, we reasoned that although Rph1p has mtROSindependent functions (Orzechowski Westholm et al., 2012), a null strain might reveal a subset of the transcriptional outcomes relevant to mtROS-mediated CLS extension. Gene list comparisons (Table S3) are summarized in Figure 6A. We observed 3-fold enrichment for the Rph1p binding motif (WAG4) among transcripts induced by both mtROS signaling and Rph1p deletion (sector C; Figure S5A), consistent with the involvement of Rph1p in regulating gene induction following hormetic mtROS signaling. mtROS signaling specifically affects Rph1p binding at subtelomeric regions (Figure 5C), and subtelomeric transcripts were enriched only among genes repressed by mtROS signaling and RPH1 deletion (sectors A and C; Figures 6B, S5B, and S5C and Table S4). Using a subtelomeric URA3 reporter and RT-PCR, we confirmed that adaptive mtROS signaling represses transcription from subtelomeric regions (Figures S5D and S5E), but not at other silenced regions (Figure S5F and Table S5). Together, these results indicate that hormetic mtROS signaling to Rph1p selectively silences subtelomeric transcription. Hormetic mtROS Signaling Modifies Higher-Order Chromatin Structure to Extend CLS Histone modifications at subtelomeric regions can influence higher-order chromatin formation, which determines gene

expression (Schoeftner and Blasco, 2009). The Sir complex is initially recruited to silenced regions via Sir2pmediated histone deacetylation (Rusche et al., 2003). This enhances the binding of Sir3p, which recruits additional Sir2p to nucleate heterochromatin spreading (Norris and Boeke, 2010). Histone methylation can enhance Sir3p binding, thereby influencing transcriptional silencing (Ng et al., 2002). ChIP experiments revealed that treatment with menadione during exponential growth strongly enhanced the association of Sir3p with subtelomeric regions in a wild-type strain, but not in an rph1D strain, at day 1 of CLS (Figure 7A). Sir3p binding was not similarly affected at other silenced loci (Figure S6A). Additionally, Sir3p was required for subtelomeric silencing following mtROS signaling (Figure 7B); however, at other silenced loci, menadione-induced transcriptional changes were mainly Sir3p independent (Figure S6B). Finally, hormetic mtROS signaling failed to extend CLS of a sir3D strain (Figures 7C and S6C). These findings demonstrate that Rph1p-regulated subtelomeric silencing, mediated by Sir3p, is a key outcome of hormetic mtROS signaling that extends lifespan. DISCUSSION Mitochondria, epigenetics, and telomere function have all been independently implicated in lifespan regulation. Our investigation of mitochondrial stress signaling in yeast demonstrates coordination among these processes to promote lifespan extension. Specifically, we have elucidated a mitochondria-to-nucleus signaling pathway (Figure 7D) in which the DNA damage kinase Tel1p transduces an mtROS signal to activate Rad53p, which in turn regulates the ability of the histone demethylase Rph1p to bind subtelomeric chromatin. The loss of Rph1p from subtelomeric chromatin elevates H3K36me3 and enhances binding of the silencing protein Sir3p to repress subtelomeric transcription. Sir3p-mediated subtelomeric silencing is an essential outcome Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc. 959

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 6. Hormetic mtROS Signaling to Rph1p Promotes Subtelomeric Silencing (A) Gene expression changes induced by exponential phase menadione treatment in a wild-type strain (mtROS dependent, sector A), rph1D strain (Rph1p dependent, sector B), or mtROS-dependent inactivation of Rph1p (sector C) at day 1 of CLS. Gray font indicates the number of induced genes, and black font indicates the number of repressed genes. See Table S3 for full gene lists. (B) Repressed genes located the indicated distance (kb) from telomeres. Subtelomeric genes are within 25 kb (gray region) of telomeres and are listed in Table S4. The significance of enrichment of repressed subtelomeric genes was determined using hypergeometric probability distribution. See also Figure S5.

of mtROS signaling and adaptation that ultimately extends CLS. Although heterochromatin maintenance at several loci influences yeast lifespan (Dang et al., 2009; Kaeberlein et al., 1999), mtROS signaling specifically targets subtelomeric heterochromatin. We propose that this specificity is a key component of the mtROS hormetic response and corroborates previous studies in which Sir proteins target distinct genomic regions under different conditions, with unique consequences for chronological and replicative aging (Ruault et al., 2011; Salvi et al., 2013). For example, sir2D enhances CLS extension under caloric restriction or in sch9D strains by inducing stress response and metabolic genes that promote longevity (Fabrizio et al., 2005) but shortens RLS due to a failure to maintain heterochromatin at the rDNA cluster (Kaeberlein et al., 1999). Subtelomeric and telomeric heterochromatin maintenance preserves telomere length and genomic stability in yeast and mammals (Schoeftner and Blasco, 2009). Telomere function is intimately connected with health- and lifespan but is most often associated with replicative senescence. Our finding that altered subtelomeric silencing supports CLS extension implies that telomere function regulates postmitotic cellular health and aging. At this point, it is unclear how repressing subtelomeric transcription extends CLS. Transcription of subtelomeric coding or noncoding genes may limit lifespan by altering metabolism, telomere length, 960 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.

or general transcription (Pryde and Louis, 1997; Schoeftner and Blasco, 2009; Smith et al., 2011; Yamada et al., 1998). Alternatively, increased telomere packaging would both repress transcription and protect telomeres from DNA damage-induced attrition, which can lead to gross chromosomal rearrangements that destabilize the nuclear genome (Davoli and de Lange, 2011). We propose that hormetic mtROS stimulate pre-emptive telomere and subtelomere heterochromatin formation early in life, which minimizes telomeric DNA damage and maintains genomic stability during aging. Although epigenetic outcomes are essential to the adaptive response, our data also indicate that transcriptional changes mediated by Msn2p/Msn4p, Gis1p, and Rph1p support mtROS adaptation. Menadione treatment induces a number of oxidative stress response genes in a manner dependent on Rph1p, which may account for the hormetic decrease in ROS during post-diauxic and stationary phases (Figure S3E and Table S2D). All of the candidate mtROS-responsive proteins we investigated bind a similar DNA motif (Figure 1B), and since all proteins are at least partially required for mtROS adaptation (Figure S1), we speculate that Rph1p-dependent chromatin remodeling might facilitate Msn2p/Msn4p- and/or Gis1p-mediated transcription at mtROS-responsive genes. Future studies to explore the relationship between Rph1p, Msn2p/Msn4p, and Gis1p in transcriptional regulation and mtROS hormesis may uncover additional cooperation among these proteins in lifespan regulation. Our results also indicate that Tel1p propagates an mtROS signal in the absence of nuclear DNA damage (Figures 4C, 4D, S3E, and S3F), which suggests that certain DNA damage response kinases function as both ROS and DNA damage sensors, and can regulate distinct cellular processes in a stimulusspecific manner. Recent studies show that ATM (the mammalian homolog of Tel1p) acts as a redox sensor (Guo et al., 2010). Our finding that Tel1p mediates the hormetic mtROS response in yeast suggests that this function is evolutionarily conserved. Hormetic mtROS signaling ultimately inhibits the chromatin binding capacity of the histone demethylase Rph1p at subtelomeres, thereby increasing H3K36me3 at these regions (Figures 5C and 5D). Although the role of histone acetylation in transcriptional silencing at subtelomeres is well established (Dang et al., 2009; Kozak et al., 2010; Santos-Rosa et al., 2004), it is increasingly evident that histone methylation can also contribute to this process in both yeast and mammals (Blasco, 2007; Caslini et al., 2009; Dang et al., 2009; Kozak et al., 2010; Santos-Rosa et al., 2004). Thus, it will be important to determine which methyltransferases and demethylases function at subtelomeres during mammalian aging. Our study would support members of the Jumonji-domain demethylases related to yeast Rph1p as prime candidates. In conclusion, our results provide insight into multiple aspects of aging by integrating mtROS, DNA damage signaling, telomeres, and epigenetics into a prolongevity pathway. The mtROS signaling pathway identified herein connects mitochondria and telomere functional status and supports recent findings that these processes likely do not regulate aging in isolation (Sahin et al., 2011). The hormetic mtROS signaling pathway we describe should provide a platform for more integrated network theories of aging and lifespan regulation.

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Figure 7. mtROS Signaling via Rph1p Promotes Sir3p Binding at Subtelomeres to Repress Gene Expression and Extend CLS (A) ChIP of Sir3p in wild-type (DBY2006) and rph1D at a control region (ACT1) and two subtelomeric genes (YPS6 and YEL077C). Cells were analyzed at day 1 of CLS following treatment with 50 mM menadione or vehicle during exponential growth. (B) Quantitative RT-PCR of subtelomeric transcripts relative to ACT1 at day 1 of CLS in wildtype or sir3D yeast following 50 mM menadione or vehicle during exponential growth. (C) CLS of wild-type and sir3D treated with 50 mM menadione (MD) or vehicle (nt) during logarithmic growth. (D) Model for mtROS signaling via DNA damage response kinases and Rph1p to regulate subtelomeric heterochromatin formation and CLS.

trol). Cell growth was monitored for 24 hr of treatment, and viability was assessed via trypan blue staining. Flow Cytometry Measurement of cellular superoxide using dihydroethidium (DHE) was performed as described (Bonawitz et al., 2007; Pan et al., 2011).

EXPERIMENTAL PROCEDURES Yeast Strains All strains used in this study are derivatives of DBY2006 (MATa his3-D200 leu23,-112 ura3-52 trp1-D1 ade2-1) and are listed in Table S7. Deletion strains were generated via gene replacement with URA3, KanMX6, or hphMX3 cassettes and transformed using the lithium acetate method (Gietz and Schiestl, 2007). Endogenous RPH1 was tagged at the C terminus with a HA-KanMX6 cassette (Longtine et al., 1998). To generate the SOD2 overexpression strain (Figure S3D), the endogenous SOD2 promoter ( 600 to 1) was replaced with the TDH3 promoter (Ko¨tter et al., 2009). All strains were verified by PCR. Chronological Lifespan Measurement Yeast were grown in minimal dextrose medium supplemented with essential amino acids (Sherman, 1991). For lifespan assays, saturated cultures from single colonies were diluted to an optical density 600 (OD600) of 0.01 in 50 ml fresh media and grown at 30 C, 200 rpm until the OD reached 0.5. Either menadione (50 mM final concentration; Sigma-Aldrich) or an equivalent volume of ethanol (no treatment control) was added, and cultures were grown until the OD reached 2.0. At this point, cells were pelleted and resuspended in media from an untreated parallel culture. CLS was determined by trypan blue staining as described (Bonawitz et al., 2006). Measurements were initiated at day 2 of CLS. All strains and treatments were analyzed in triplicate. To quantitatively compare CLS extension between different strain backgrounds, the area under the viability curve was calculated in GraphPad Prism. The difference between the untreated and menadione-treated wild-type areas was set to 1, and differences within strain backgrounds are expressed as percent of wild-type CLS extension. Chemical Toxicity Wild-type yeast were grown to an OD600 of 0.5 and treated with 50 mM menadione, 200 mM menadione, 0.01% MMS, 0.02% MMS, or ethanol (vehicle con-

Immunoblotting Yeast whole-cell extracts were prepared from cells in exponential growth phase treated with menadione, MMS (0.01% final concentration), or ethanol for 30 min (Keogh et al., 2006). Proteins were separated in 6% SDS-PAGE gels (to resolve Rad53p and Rph1-HA phosphorylation) or 12% SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). Blocking and antibody incubations were carried out in Tris-buffered saline and Tween (TBST; 10 mM Tris [pH 8.0], 150 mM NaCl, 0.05% Tween 20) containing 5% nonfat dry milk. Membranes were incubated with anti-Rad53 (1:300; Santa Cruz Biotechnology), anti-phospho-(S/T)Q (1:1000; Cell Signaling), antiMnSOD (1:500; Assay Designs), anti-HA (1:1000; Covance), anti-actin (1:1000; Chemicon), or anti-GAPDH-HRP (horseradish peroxidase) (1:1000; Abcam). Immunoblotting was performed on protein isolated from at least two independent experiments. Band density traces and quantification were determined using ImageJ (National Institutes of Health). 8-Hydroxyguanosine Quantification Yeast cultures in exponential growth phase were treated with ethanol, 50 mM menadione, or 200 mM menadione for 30 min prior to DNA extraction (Hoffman, 2001). Total DNA was treated with ribonuclease (RNase) A and proteinase K (Sigma-Aldrich) and then precipitated with phenol:chloroform and ethanol. DNA (5 mg per sample) was blotted onto a nitrocellulose membrane. The membrane was dried for 1 hr at 85 C, blocked in TBST with 5% BSA (SigmaAldrich), and incubated overnight in anti-8-hydroxyguanosine (Chemicon) diluted 1:500 in BSA-TBST. After washing in TBST, the membrane was incubated in HRP-conjugated anti-goat secondary at a 1:2000 dilution for 1 hr and visualized with SuperSignal West Pico Chemiluminescent Substrate. To normalize dot intensity to total DNA, the membrane was incubated in 2.5 mg DAPI and visualized on a VersaDoc imager (Bio-Rad). Antibody and DAPI dot intensity from triplicate independent treatments were quantified using ImageJ. RNA Preparation and Microarray Analysis RNA extraction and purification were performed as described (Bonawitz et al., 2008) from wild-type DBY2006 and rph1D cultures grown for 2 or 24 hr after addition of 50 mM menadione or ethanol. RNA quality bioanalysis,

Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc. 961

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

complementary DNA (cDNA) synthesis, and hybridization to a Roche NimbleGen 12 3 135 K array were performed in collaboration with the Yale University W.M. Keck Foundation Biotechnology Resource Laboratory according to the manufacturer’s protocol. Two independent biological replicates were prepared for each sample, and hybridization was performed in duplicate to generate technical replicates. Raw data were normalized accordinig to the Robust Multi-array Average (RMA) algorithm using NimbleScan software (Irizarry et al., 2003). Probe expression values were used to determine fold change and filtered to significantly induced and repressed probe lists (p % 0.01 as determined by Student’s t test). Gene lists were generated based on the occurrence of at least two probes for a particular gene present exclusively in induced or repressed probe lists. To map the chromosomal location of genes in sectors A, B, and C, the distance from the transcription start site to the nearest telomere was determined. Gene lists are described in the Tables S2 and S3. Normalized expression data were downloaded from the Gene Expression Omnibus at NCBI (GSM1102, record GDS108 [Gasch et al., 2000]) and from the supplemental online material associated with Hardwick et al. (1999). Gene lists were ranked according to fold change, and the top 5% induced genes were selected for further analysis to identify genes induced in both data sets. The coinduced genes were further analyzed using WebMOTIFS (Romer et al., 2007) and YEASTRACT (yeast search for transcriptional regulators and consensus tracking; Abdulrehman et al., 2011) to generate the data shown in Figure 1C. Quantitative RT-PCR RNA was isolated as described above and converted to cDNA using the High Capacity cDNA Reverse Transcription Kit (Agilent) according to the manufacturer’s instructions. cDNA was diluted 1:10, and each 20 ml reaction contained 10 ml diluted cDNA, 0.4 ml dilutions of each primer (from a 25 mM stock), and 9.2 ml Bio-Rad iQ SYBR Mix. PCR conditions were as described (Cotney et al., 2007).

ACKNOWLEDGMENTS This work was supported by grant W911NF-11-1-0376 from the Army Research Office to G.S.S. E.A.S. was supported by NIH Genetics Training Grant T32 GM07499 and is currently supported by NIH predoctoral fellowship F31 AG043242. N.R. was supported by a postdoctoral fellowship from the United Mitochondrial Disease Foundation. We thank Drs. Megan King, Patrick Lusk, and David Stern for reagents used in the study and insightful discussions. Received: December 1, 2012 Revised: February 27, 2013 Accepted: April 1, 2013 Published: June 4, 2013 REFERENCES Abdulrehman, D., Monteiro, P.T., Teixeira, M.C., Mira, N.P., Lourenc¸o, A.B., dos Santos, S.C., Cabrito, T.R., Francisco, A.P., Madeira, S.C., Aires, R.S., et al. (2011). YEASTRACT: providing a programmatic access to curated transcriptional regulatory associations in Saccharomyces cerevisiae through a web services interface. Nucleic Acids Res. 39(Database issue), D136–D140. Altaf, M., Utley, R.T., Lacoste, N., Tan, S., Briggs, S.D., and Coˆte´, J. (2007). Interplay of chromatin modifiers on a short basic patch of histone H4 tail defines the boundary of telomeric heterochromatin. Mol. Cell 28, 1002–1014. Aparicio, O., Geisberg, J.V., and Struhl, K. (2004). Chromatin immunoprecipitation for determining the association of proteins with specific genomic sequences in vivo. Curr. Protoc. Cell Biol. Chapter 17, 17, 7. Balaban, R.S., Nemoto, S., and Finkel, T. (2005). Mitochondria, oxidants, and aging. Cell 120, 483–495. Barton, A.B., and Kaback, D.B. (2006). Telomeric silencing of an open reading frame in Saccharomyces cerevisiae. Genetics 173, 1169–1173.

Chromatin Immunoprecipitation Yeast cultures were treated with 50 mM menadione or ethanol from OD 0.5 to OD 2.0. At 24 hr after the initiation of menadione treatment at OD 0.5, cells were processed for ChIP analysis, which was performed essentially as described (Aparicio et al., 2004), with the following modifications. DNA was sheared in a Bioruptor UCD-200 Sonicator at 4 C five times for 20 s each. Chromatin was precleared by incubating with protein A agarose beads (Millipore) for 30 min at room temperature. Antibodies used were anti-HA (Covance), anti-trimethyl H3K36 (Epigentek), anti-H3 (Abcam), and anti-Sir3 (Santa Cruz Biotechnology). Quantitative PCR of three technical replicates per sample was performed using the same conditions described for RT-PCR, except the reaction proceeded for 40 cycles. Primers are listed in Table S8.

Blasco, M.A. (2005). Telomeres and human disease: ageing, cancer and beyond. Nat. Rev. Genet. 6, 611–622.

Statistical Analysis For CLS, fluorescence-activated cell sorting (FACS), 8-hydroxyguanosine, and RT-PCR analysis, all data points represent the mean of biological triplicate samples inoculated from isolated single colonies. Error bars represent SEM. p values were determined using Student’s unpaired t test, and values for relevant comparisons are indicated within figures. Statistical analysis for enrichment of subtelomeric transcripts was determined using a hypergeometric probability distribution (Martin et al., 2004). For ChIP analysis, data points represent the mean of two biological replicate samples inoculated from isolated single colonies. Error bars represent the range.

Bonawitz, N.D., Chatenay-Lapointe, M., Wearn, C.M., and Shadel, G.S. (2008). Expression of the rDNA-encoded mitochondrial protein Tar1p is stringently controlled and responds differentially to mitochondrial respiratory demand and dysfunction. Curr. Genet. 54, 83–94.

ACCESSION NUMBERS Raw microarray data were deposited in the Gene Expression Omnibus repository under the number GSE45383.

Blasco, M.A. (2007). The epigenetic regulation of mammalian telomeres. Nat. Rev. Genet. 8, 299–309. Bonawitz, N.D., Rodeheffer, M.S., and Shadel, G.S. (2006). Defective mitochondrial gene expression results in reactive oxygen species-mediated inhibition of respiration and reduction of yeast life span. Mol. Cell. Biol. 26, 4818– 4829. Bonawitz, N.D., Chatenay-Lapointe, M., Pan, Y., and Shadel, G.S. (2007). Reduced TOR signaling extends chronological life span via increased respiration and upregulation of mitochondrial gene expression. Cell Metab. 5, 265–277.

Caslini, C., Connelly, J.A., Serna, A., Broccoli, D., and Hess, J.L. (2009). MLL associates with telomeres and regulates telomeric repeat-containing RNA transcription. Mol. Cell. Biol. 29, 4519–4526. Castro, F.A., Mariani, D., Panek, A.D., Eleutherio, E.C., and Pereira, M.D. (2008). Cytotoxicity mechanism of two naphthoquinones (menadione and plumbagin) in Saccharomyces cerevisiae. PLoS ONE 3, e3999. Cotney, J., Wang, Z., and Shadel, G.S. (2007). Relative abundance of the human mitochondrial transcription system and distinct roles for h-mtTFB1 and h-mtTFB2 in mitochondrial biogenesis and gene expression. Nucleic Acids Res. 35, 4042–4054.

SUPPLEMENTAL INFORMATION

Dang, W., Steffen, K.K., Perry, R., Dorsey, J.A., Johnson, F.B., Shilatifard, A., Kaeberlein, M., Kennedy, B.K., and Berger, S.L. (2009). Histone H4 lysine 16 acetylation regulates cellular lifespan. Nature 459, 802–807.

Supplemental Information includes six figures and eight tables and can be found with this article online at http://dx.doi.org/10.1016/j.cmet.2013.04.003.

Davoli, T., and de Lange, T. (2011). The causes and consequences of polyploidy in normal development and cancer. Annu. Rev. Cell Dev. Biol. 27, 585–610.

962 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Fabrizio, P., and Longo, V.D. (2007). The chronological life span of Saccharomyces cerevisiae. Methods Mol. Biol. 371, 89–95. Fabrizio, P., Gattazzo, C., Battistella, L., Wei, M., Cheng, C., McGrew, K., and Longo, V.D. (2005). Sir2 blocks extreme life-span extension. Cell 123, 655–667. Finkel, T. (2011). Signal transduction by reactive oxygen species. J. Cell Biol. 194, 7–15. Gasch, A.P., Spellman, P.T., Kao, C.M., Carmel-Harel, O., Eisen, M.B., Storz, G., Botstein, D., and Brown, P.O. (2000). Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11, 4241– 4257. Gasch, A.P., Huang, M., Metzner, S., Botstein, D., Elledge, S.J., and Brown, P.O. (2001). Genomic expression responses to DNA-damaging agents and the regulatory role of the yeast ATR homolog Mec1p. Mol. Biol. Cell 12, 2987–3003. Gems, D., and Partridge, L. (2008). Stress-response hormesis and aging: ‘‘that which does not kill us makes us stronger’’. Cell Metab. 7, 200–203. Gietz, R.D., and Schiestl, R.H. (2007). High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method. Nat. Protoc. 2, 31–34. Grunstein, M. (1997). Histone acetylation in chromatin structure and transcription. Nature 389, 349–352.

Lee, S.J., Schwartz, M.F., Duong, J.K., and Stern, D.F. (2003). Rad53 phosphorylation site clusters are important for Rad53 regulation and signaling. Mol. Cell. Biol. 23, 6300–6314. Lee, S.J., Hwang, A.B., and Kenyon, C. (2010). Inhibition of respiration extends C. elegans life span via reactive oxygen species that increase HIF-1 activity. Curr. Biol. 20, 2131–2136. Liang, C.Y., Hsu, P.H., Chou, D.F., Pan, C.Y., Wang, L.C., Huang, W.C., Tsai, M.D., and Lo, W.S. (2011). The histone H3K36 demethylase Rph1/KDM4 regulates the expression of the photoreactivation gene PHR1. Nucleic Acids Res. 39, 4151–4165. Longo, V.D., Shadel, G.S., Kaeberlein, M., and Kennedy, B. (2012). Replicative and chronological aging in Saccharomyces cerevisiae. Cell Metab. 16, 18–31. Longtine, M.S., McKenzie, A., 3rd, Demarini, D.J., Shah, N.G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J.R. (1998). Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14, 953–961. Martin, A.M., Pouchnik, D.J., Walker, J.L., and Wyrick, J.J. (2004). Redundant roles for histone H3 N-terminal lysine residues in subtelomeric gene repression in Saccharomyces cerevisiae. Genetics 167, 1123–1132. Millar, C.B., and Grunstein, M. (2006). Genome-wide patterns of histone modifications in yeast. Nat. Rev. Mol. Cell Biol. 7, 657–666.

Guo, Z., Kozlov, S., Lavin, M.F., Person, M.D., and Paull, T.T. (2010). ATM activation by oxidative stress. Science 330, 517–521.

Nakada, D., Hirano, Y., and Sugimoto, K. (2004). Requirement of the Mre11 complex and exonuclease 1 for activation of the Mec1 signaling pathway. Mol. Cell. Biol. 24, 10016–10025.

Hardwick, J.S., Kuruvilla, F.G., Tong, J.K., Shamji, A.F., and Schreiber, S.L. (1999). Rapamycin-modulated transcription defines the subset of nutrientsensitive signaling pathways directly controlled by the Tor proteins. Proc. Natl. Acad. Sci. USA 96, 14866–14870.

Nautiyal, S., DeRisi, J.L., and Blackburn, E.H. (2002). The genome-wide expression response to telomerase deletion in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 99, 9316–9321.

Harrison, J.C., and Haber, J.E. (2006). Surviving the breakup: the DNA damage checkpoint. Annu. Rev. Genet. 40, 209–235. Hoffman, C.S. (2001). Preparation of yeast DNA. Curr. Protoc. Mol. Biol. Chapter 13, t13, 11. Irizarry, R.A., Hobbs, B., Collin, F., Beazer-Barclay, Y.D., Antonellis, K.J., Scherf, U., and Speed, T.P. (2003). Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4, 249–264. Kaeberlein, M. (2010). Lessons on longevity from budding yeast. Nature 464, 513–519. Kaeberlein, M., McVey, M., and Guarente, L. (1999). The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two different mechanisms. Genes Dev. 13, 2570–2580. Karahalil, B., Girard, P.M., Boiteux, S., and Dizdaroglu, M. (1998). Substrate specificity of the Ogg1 protein of Saccharomyces cerevisiae: excision of guanine lesions produced in DNA by ionizing radiation- or hydrogen peroxide/ metal ion-generated free radicals. Nucleic Acids Res. 26, 1228–1233. Kenyon, C. (2005). The plasticity of aging: insights from long-lived mutants. Cell 120, 449–460. Keogh, M.C., Kim, J.A., Downey, M., Fillingham, J., Chowdhury, D., Harrison, J.C., Onishi, M., Datta, N., Galicia, S., Emili, A., et al. (2006). A phosphatase complex that dephosphorylates gammaH2AX regulates DNA damage checkpoint recovery. Nature 439, 497–501. Kim, S.T., Lim, D.S., Canman, C.E., and Kastan, M.B. (1999). Substrate specificities and identification of putative substrates of ATM kinase family members. J. Biol. Chem. 274, 37538–37543. Kim, E.M., Jang, Y.K., and Park, S.D. (2002). Phosphorylation of Rph1, a damage-responsive repressor of PHR1 in Saccharomyces cerevisiae, is dependent upon Rad53 kinase. Nucleic Acids Res. 30, 643–648. Ko¨tter, P., Weigand, J.E., Meyer, B., Entian, K.D., and Suess, B. (2009). A fast and efficient translational control system for conditional expression of yeast genes. Nucleic Acids Res. 37, e120. Kozak, M.L., Chavez, A., Dang, W., Berger, S.L., Ashok, A., Guo, X., and Johnson, F.B. (2010). Inactivation of the Sas2 histone acetyltransferase delays senescence driven by telomere dysfunction. EMBO J. 29, 158–170.

Ng, H.H., Feng, Q., Wang, H., Erdjument-Bromage, H., Tempst, P., Zhang, Y., and Struhl, K. (2002). Lysine methylation within the globular domain of histone H3 by Dot1 is important for telomeric silencing and Sir protein association. Genes Dev. 16, 1518–1527. Norris, A., and Boeke, J.D. (2010). Silent information regulator 3: the Goldilocks of the silencing complex. Genes Dev. 24, 115–122. Orzechowski Westholm, J., Tronnersjo¨, S., Nordberg, N., Olsson, I., Komorowski, J., and Ronne, H. (2012). Gis1 and Rph1 regulate glycerol and acetate metabolism in glucose depleted yeast cells. PLoS ONE 7, e31577. Pan, Y., Schroeder, E.A., Ocampo, A., Barrientos, A., and Shadel, G.S. (2011). Regulation of yeast chronological life span by TORC1 via adaptive mitochondrial ROS signaling. Cell Metab. 13, 668–678. Park, Y., and Lustig, A.J. (2000). Telomere structure regulates the heritability of repressed subtelomeric chromatin in Saccharomyces cerevisiae. Genetics 154, 587–598. Pellicioli, A., Lucca, C., Liberi, G., Marini, F., Lopes, M., Plevani, P., Romano, A., Di Fiore, P.P., and Foiani, M. (1999). Activation of Rad53 kinase in response to DNA damage and its effect in modulating phosphorylation of the lagging strand DNA polymerase. EMBO J. 18, 6561–6572. Pryde, F.E., and Louis, E.J. (1997). Saccharomyces cerevisiae telomeres. A review. Biochemistry (Mosc.) 62, 1232–1241. Raimundo, N., Song, L., Shutt, T.E., McKay, S.E., Cotney, J., Guan, M.X., Gilliland, T.C., Hohuan, D., Santos-Sacchi, J., and Shadel, G.S. (2012). Mitochondrial stress engages E2F1 apoptotic signaling to cause deafness. Cell 148, 716–726. Ristow, M., and Schmeisser, S. (2011). Extending life span by increasing oxidative stress. Free Radic. Biol. Med. 51, 327–336. Ristow, M., Zarse, K., Oberbach, A., Klo¨ting, N., Birringer, M., Kiehntopf, M., Stumvoll, M., Kahn, C.R., and Blu¨her, M. (2009). Antioxidants prevent health-promoting effects of physical exercise in humans. Proc. Natl. Acad. Sci. USA 106, 8665–8670. Romer, K.A., Kayombya, G.R., and Fraenkel, E. (2007). WebMOTIFS: automated discovery, filtering and scoring of DNA sequence motifs using multiple programs and Bayesian approaches. Nucleic Acids Res. 35(Web Server issue), W217–W220.

Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc. 963

Cell Metabolism Mitochondrial ROS Signal Epigenetic Adaptation

Ruault, M., De Meyer, A., Loı¨odice, I., and Taddei, A. (2011). Clustering heterochromatin: Sir3 promotes telomere clustering independently of silencing in yeast. J. Cell Biol. 192, 417–431. Rusche, L.N., Kirchmaier, A.L., and Rine, J. (2003). The establishment, inheritance, and function of silenced chromatin in Saccharomyces cerevisiae. Annu. Rev. Biochem. 72, 481–516. Sahin, E., Colla, S., Liesa, M., Moslehi, J., Mu¨ller, F.L., Guo, M., Cooper, M., Kotton, D., Fabian, A.J., Walkey, C., et al. (2011). Telomere dysfunction induces metabolic and mitochondrial compromise. Nature 470, 359–365. Salvi, J.S., Chan, J.N., Pettigrew, C., Liu, T.T., Wu, J.D., and Mekhail, K. (2013). Enforcement of a lifespan-sustaining distribution of Sir2 between telomeres, mating-type loci, and rDNA repeats by Rif1. Aging Cell 12, 67–75. Sanchez, Y., Desany, B.A., Jones, W.J., Liu, Q., Wang, B., and Elledge, S.J. (1996). Regulation of RAD53 by the ATM-like kinases MEC1 and TEL1 in yeast cell cycle checkpoint pathways. Science 271, 357–360. Santos-Rosa, H., Bannister, A.J., Dehe, P.M., Ge´li, V., and Kouzarides, T. (2004). Methylation of H3 lysine 4 at euchromatin promotes Sir3p association with heterochromatin. J. Biol. Chem. 279, 47506–47512. Schoeftner, S., and Blasco, M.A. (2009). A ‘higher order’ of telomere regulation: telomere heterochromatin and telomeric RNAs. EMBO J. 28, 2323–2336. Schulz, T.J., Zarse, K., Voigt, A., Urban, N., Birringer, M., and Ristow, M. (2007). Glucose restriction extends Caenorhabditis elegans life span by inducing mitochondrial respiration and increasing oxidative stress. Cell Metab. 6, 280–293.

Tapia, P.C. (2006). Sublethal mitochondrial stress with an attendant stoichiometric augmentation of reactive oxygen species may precipitate many of the beneficial alterations in cellular physiology produced by caloric restriction, intermittent fasting, exercise and dietary phytonutrients: ‘‘Mitohormesis’’ for health and vitality. Med. Hypotheses 66, 832–843. Tu, S., Bulloch, E.M., Yang, L., Ren, C., Huang, W.C., Hsu, P.H., Chen, C.H., Liao, C.L., Yu, H.M., Lo, W.S., et al. (2007). Identification of histone demethylases in Saccharomyces cerevisiae. J. Biol. Chem. 282, 14262–14271. Wagner, E.J., and Carpenter, P.B. (2012). Understanding the language of Lys36 methylation at histone H3. Nat. Rev. Mol. Cell Biol. 13, 115–126. Wei, M., Fabrizio, P., Hu, J., Ge, H., Cheng, C., Li, L., and Longo, V.D. (2008). Life span extension by calorie restriction depends on Rim15 and transcription factors downstream of Ras/PKA, Tor, and Sch9. PLoS Genet. 4, e13. Wei, M., Fabrizio, P., Madia, F., Hu, J., Ge, H., Li, L.M., and Longo, V.D. (2009). Tor1/Sch9-regulated carbon source substitution is as effective as calorie restriction in life span extension. PLoS Genet. 5, e1000467. Yamada, M., Hayatsu, N., Matsuura, A., and Ishikawa, F. (1998). Y’-Help1, a DNA helicase encoded by the yeast subtelomeric Y’ element, is induced in survivors defective for telomerase. J. Biol. Chem. 273, 33360–33366. Yang, W., and Hekimi, S. (2010). A mitochondrial superoxide signal triggers increased longevity in Caenorhabditis elegans. PLoS Biol. 8, e1000556.

Sherman, F. (1991). Getting started with yeast. Methods Enzymol. 194, 3–21.

Zarse, K., Schmeisser, S., Groth, M., Priebe, S., Beuster, G., Kuhlow, D., Guthke, R., Platzer, M., Kahn, C.R., and Ristow, M. (2012). Impaired insulin/ IGF1 signaling extends life span by promoting mitochondrial L-proline catabolism to induce a transient ROS signal. Cell Metab. 15, 451–465.

Smith, J.J., Miller, L.R., Kreisberg, R., Vazquez, L., Wan, Y., and Aitchison, J.D. (2011). Environment-responsive transcription factors bind subtelomeric elements and regulate gene silencing. Mol. Syst. Biol. 7, 455.

Zhao, X., Muller, E.G., and Rothstein, R. (1998). A suppressor of two essential checkpoint genes identifies a novel protein that negatively affects dNTP pools. Mol. Cell 2, 329–340.

Sena, L.A., and Chandel, N.S. (2012). Physiological roles of mitochondrial reactive oxygen species. Mol. Cell 48, 158–167.

964 Cell Metabolism 17, 954–964, June 4, 2013 ª2013 Elsevier Inc.