Robert C. Dickson, Department of Molecular and Cellular Biochemistry and the Lucille Markey Cancer Center, 741 S. Limestone, University of Kentucky College of Medicine, Lexington, KY 40536, USA. Tel. 859 323 6052; fax 859 323 1037; e-mail: firstname.lastname@example.org
Disease incidence rises rapidly with age and increases both human suffering and economic hardship while shortening life. Advances in understanding the signaling pathways and cellular processes that influence aging support the possibility of reducing the incidence of age-related diseases and increasing lifespan by pharmacological intervention. Here, we demonstrate a novel pharmacological strategy that both reduces signs of aging in the budding yeast Saccharomyces cerevisiae and generates a synergistic increase in lifespan. By combining a low dose of rapamycin, to reduce activity of the target of rapamycin complex 1 (TORC1) protein kinase, and myriocin, to reduce sphingolipid synthesis, we show enhancement of autophagy, genomic stability, mitochondrial function, and AMP kinase pathway activity. These processes are controlled by evolutionarily conserved signal transduction pathways that are vital for maintaining a healthy state and promoting a long life. Thus, our data show that it ought to be possible to find pharmacological approaches to generate a synergistic reduction in the incidence of human age-related diseases to improve health quality in the elderly and enhance lifespan.
Most people in western societies die from diabetes, cancer, cardiovascular or immune dysfunction, or neurodegeneration, the so-called age-related diseases. There is a growing interest in finding strategies to reduce the incidence of these diseases and improve human healthspan and perhaps lifespan (Niccoli & Partridge, 2012). One promising strategy uses rapamycin to reduce activity of the target of rapamycin protein kinase complex 1 (TORC1). TORC1 is nearly ubiquitous in eukaryotes and plays evolutionarily conserved roles in aging and lifespan (Blagosklonny, 2006; Fontana et al., 2010; Kapahi et al., 2010; Niccoli & Partridge, 2012). One TORC1 substrate, the mammalian S6 protein kinase 1 (S6K1) and its budding yeast ortholog Sch9 (Fig. 1A), plays key roles in longevity, as deleting the S6K1 or SCH9 gene enhances lifespan in yeasts and mice, respectively (Fabrizio et al., 2001; Selman et al., 2009; Fontana et al., 2010).
Recently we showed that Sch9 activity can be reduced, and yeast chronological lifespan (CLS) increased by treating cells with the natural product myriocin (Huang et al., 2012). Myriocin inhibits the first enzyme in sphingolipid synthesis, and moderate drug doses effectively reduce the steady-state level of several, but not all, sphingolipids. As a consequence, the Pkh1 and Pkh2 protein kinases, functional orthologs of mammalian phosphoinositide-dependent protein kinase 1 (PDK1) (Casamayor et al., 1999), are less active as is Sch9, one of their substrates (Fig. 1A) (Urban et al., 2007; Loewith & Hall, 2011; Huang et al., 2012). In addition, we found that myriocin treatment extends CLS in sch9 mutant cells, indicating that myriocin influences lifespan by both Sch9-dependent and independent mechanisms.
Dual control of Sch9 by TORC1 and Pkh1/2 suggested the possibility of producing a synergistic increase in CLS by treating cells with a low-dose combination of myriocin and rapamycin. CLS is a measure of how well cells survive in stationary phase while in a nondividing or G0 state. The other measure of yeast longevity, replicative lifespan, measures how many times a cell can bud to yield a new cell. TORC1 and Sch9 regulate both forms of yeast lifespan, and it is thought that they modulate metabolism in ways that enhance survival, similar to what calorie restriction does in many organisms to extend lifespan (Fontana et al., 2010; Anderson & Weindruch, 2012; Longo et al., 2012).
We show that combination drug treatment creates a synergistic increase in CLS by modulating a wide array of processes and signaling pathways including the Snf1/AMP kinase, TORC1-Sch9, and protein kinase A (PKA), which play evolutionarily conserved roles in regulating aging and longevity. Part of the mechanism for generating synergy appears to lie upstream of TORC1, and we discuss ways in which a low dose of the two drugs may reduce TORC1 activity.
Lifespan and stress resistance
We identified low concentrations of myriocin or rapamycin having little effect on the CLS of haploid DBY746 cells when used individually, but produced a large and potentially synergistic increase (Table 1, bold numbers) in CLS when used together (Fig. 1B). As aging and longevity are complex traits and are influenced by many factors, we examined other strains backgrounds (Longo et al., 2012) to show the general feasibility of lifespan enhancement by drug synergy. Diploid BY4743 cells show a smaller, but statistically significant increase in CLS when treated with the drug combination but not with the single drugs (Fig. 1D), indicating that the drug combination also enhances lifespan in BY4743 cells, although the increase is less than in DBY746 cells.
Table 1. Combination drug treatment produces a synergistic increase in lifespan
Calorie restriction (CR) enhances lifespan in many eukaryotes ranging from yeasts to mammals (Fontana et al., 2010; Anderson & Weindruch, 2012). We determined whether CR improves the ability of the drug combination to enhance CLS by transferring drug-treated or untreated cells from culture medium to water after entry into stationary phase (72 h of incubation) (Longo et al., 2012). This strategy allows the drugs to promote remodeling of metabolism during the growth phase and then examines survival after growth stops and the drugs are removed. With this strategy, the combination drug treatment produces a large enhancement of lifespan in BY4743 cells (Fig. 1E), similar to the enhancement found in DBY746 cells treated in the same manner (Fig. 1C). Thus, severe calorie restriction imposed at CLS day 1 changes some element of BY4743 physiology that enables the drug combination to generate a very large and potentially synergistic increase in lifespan, similar to the increase seen in DBY746 cells either under CR or non-CR conditions.
Increased resistance to nutritional and other stresses is a nearly universal feature of strategies that increase lifespan and is governed by several signaling pathways in yeast (Fig. 1A) (Kourtis & Tavernarakis, 2011; De Virgilio, 2012; Longo et al., 2012). We find that dual drug treatment increases resistance to both heat and oxidative stress in DBY746 and BY4743 cells even without CR treatment (Fig. 1F,G), consistent with the idea that dual drug treatment is modulating multiple signaling pathways with roles in stress resistance.
To assess whether the increase in CLS produced by the combination of myriocin and rapamycin is synergistic, not additive, we performed survival experiments similar in design to those used to develop drug combinations that produce a synergistic decrease in cancer cell survival, only we measured an increase in survival. We chose to measure DBY746 survival on CLS day 5, because at this time, cells treated with the drug combination survive far better than cells treated with a single drug or no drug (Fig. 1B).
The effect of myriocin and rapamycin alone on survival was assessed to obtain the median effective concentrations (EC50) for each drug. The EC50 of rapamycin was 1.1 nm and that of myriocin was 314.7 nm. Nonlinear regression was used to fit the survival verses concentration data (data not shown), and the resulting curves were used to determine the concentrations that would be required for each drug alone to elicit the same survival effect as that observed with combinations of rapamycin and myriocin in different molar ratios (Table 1).
Using these data, we calculated the combination index (CI) according to Chou and Talalay (Chou & Talalay, 1984). To test the robustness of this result, we also calculated the alpha and beta4 parameters that are used to determine deviation from additivity in the more restrictive models of Greco and Plummer, respectively (Greco et al., 1990; Plummer & Short, 1990). The prediction of synergy was consistent across all mathematical models (Table 1). The molar ratio of the two drugs has no effect on synergy, but the drug concentration does. The effect approaches additivity as the concentrations approach 1 nm rapamycin and 180 nm myriocin. Below those concentrations, for both drugs, there is strong synergy, and above those concentrations, for both drugs, there is antagonism (Table 1, shaded values).
Growth rate and cell density were measured also using a wide range of drug concentrations. Growth rate slowed with myriocin concentrations > 200 ng mL−1 (498 nm) (Fig. S1A) but was not affected even by the highest dose of rapamycin (Fig. S1B). Concentrations of myriocin above 30 ng mL−1 (75 nm) increase the final density of cells as do concentrations of rapamycin above 500 pg mL−1 (0.55 nm) (Fig. S1A,B). This mass increase may be due to changes in metabolism that promote better use of available carbon sources and other nutrients.
Growth curves and cell density values for a broad array of drug combinations gave a range of responses with some combinations not slowing growth and others completely impairing it (Fig. S1C). All combinations that produce a synergistic increase in survival (Table 1) also increase the final cell density. However, the converse is not true because some combinations that increase cell density produce additive or antagonistic effects on survival, indicating that above a certain level of one or the other drug there is no correlation between increased cell density and a synergistic increase in survival.
Autophagy is an evolutionarily conserved process for promoting intracellular reutilization of resources and maintaining organelle homeostasis in ways that slow aging and disease progression and enhance lifespan (Alvers et al., 2009; Green et al., 2011; Rubinsztein et al., 2011; Niccoli & Partridge, 2012). The effect of the drug combination on autophagic flux was evaluated by measuring the cleavage of GFP from GFP-Atg8 (Cheong & Klionsky, 2008). Appearance of free GFP requires formation of autophagosomes, their fusion to vacuoles (liposome) and protease activation. Only the combination drug treatment increases the level of free GFP above background, indicating an increase in autophagic flux (Fig. 2A, lane 4). This increase in free GFP is due to autophagy, as it does not appear in atg1 mutant cells where autophagy is blocked (Fig. 2A, lane 6). The slight increase in the level of Atg8-GFP protein in wild-type cells treated with both drugs may be due to increased gene expression.
To determine whether autophagy is necessary for the drug combination to produce an increase in lifespan, we measured CLS in atg1∆ cells. Dual drug treatment promotes an increase in the CLS of atg1 cells (Fig. 2B), although the increase is less than in wild-type cells (Fig. 1C). Thus, autophagy is required for the drug combination to produce a maximal increase in lifespan but is not needed to produce submaximal increases.
DNA damage and genomic instability are hallmarks of cancer and aging (Salk et al., 2010; Pfau & Amon, 2012). To determine whether drug synergy reduces genomic instability, we used sgs1Δ cells because they show an age-dependent increase in genome instability that is regulated by Sch9 (Madia et al., 2008). Additionally, the low rate of gross chromosomal rearrangements in wild-type yeast cells hampers frequency measurements, but the sgs1Δ allele increases the frequency and improves quantification of drug effects. Sgs1, a member of the RecQ helicase family of DNA unwinding proteins which maintain genome stability, is defective in the premature aging disorder Werner's syndrome and in the cancer-prone disorder Blooms’ syndrome. We find that the combination drug treatment reduces the frequency of small chromosomal changes involving point mutations, deletions, insertions, and frameshifts along with gross chromosomal rearrangements in sgs1Δ cells down to the level found in wild-type cells at days 7 and 13 of a CLS assay (Fig. 2C,D). The individual drugs produce no such reductions. Thus, the drug combination generates a greater than additive increase in genomic stability, which is at least in part due to down-regulation of the TORC1-Sch9 pathway (see below).
We also measured the effect of drug treatment on the frequency of small chromosomal mutations in wild-type DBY746 cells and found that dual drug treatment produces a statistically significant reduction in mutation frequency, whereas the individual drugs do not reduce mutation frequency (Fig. S2).
Mitochondria are the primary organelle for eukaryotic energy generation whose by-products include reactive oxygen species (ROS) that seem to protect cells at relatively low levels, but damage cells at high levels and increase the rate of aging while reducing lifespan (Pan, 2011; Ristow & Schmeisser, 2011; Rubinsztein et al., 2011; Finkel, 2012). Mitochondrial function declines with age and is thought to hasten the onset of age-associated diseases (Green et al., 2011; Niccoli & Partridge, 2012). Superior mitochondrial function, as measured by respiration rate, is vital for yeast survival and is observed in long-lived mutant cells or wild-type cells treated with a high concentration of rapamycin or myriocin (Pan, 2011; Pan et al., 2011; Huang et al., 2012; Longo et al., 2012). With dual drug treatment, there is an increase in oxygen consumption in both log- and stationary-phase cells compared with untreated cells or cells treated with a single drug (Fig. 3A), consistent with increased lifespan (Fig. 1B–E). As mitochondria generate most cellular ROS (Pan et al., 2011), we examined the level of superoxide using dihydroethidium (DHE). There is an equal amount of superoxide in untreated or drug-treated log-phase cells, but by stationary phase, the level of superoxide is lower in cells treated with both drugs (Fig. 3B). While DHE primarily detects superoxide, it also detects other ROS such as the abundant hydrogen peroxide. To specifically measure hydrogen peroxide, we used 2′,7′-dichlorodihydrofluorescein (H2DCF-DA), and the results show that hydrogen peroxide is more abundant in log phase, but lower in stationary cells treated with both drugs (Fig. 3C). These results support the view that combination drug treatment modulates superoxide and ROS levels to enhance CLS in ways that are similar to what has been observed in long-lived cells with reduced TORC1 or Sch9 activity (Mesquita et al., 2010; Pan et al., 2011).
Effect of the drug combination on TORC1
Activation of Sch9 requires phosphorylation by both TORC1 and Pkh1/2 (Urban et al., 2007). To determine whether synergy is created upstream of one or both of these kinases, we examined phosphorylation of Sch9. Moderately high doses of myriocin are known to lower Pkh1/2 activity, causing a reduction in Sch9 phosphorylation on residue T570 (Huang et al., 2012). Neither the low dose of myriocin or rapamycin used here alters T570 phosphorylation (Fig. 4A, top panel). In contrast, the combination treatment causes about a 40% reduction in phosphorylation and, in addition, lowers the level of Sch9 protein by about 30%. Furthermore, the mobility of Sch9 shifts downward in the sample treated with the drug combination (Fig. 4A, lane 4), indicative of reduced C-terminal phosphorylation by TORC1 (Urban et al., 2007).
The effect of the combination drug treatment on C-terminal phosphorylation was quantified using chemically fragmented Sch9 with a C-terminal 5-HA tag and then calculating the ratio of phosphorylated to nonphosphorylated C-terminal fragments (Fig. 4B, +P/−P). The combination drug treatment reduces phosphorylation by about 80% compared with no drug or single drug treatment (Fig. 4B, compare lane 4 with lanes 1–3; see complete blots in Fig. S3). For comparison, inhibiting TORC1 with a high dose of rapamycin nearly eliminates all Sch9 C-terminal phosphorylation (Fig. 4B, lane 5). The drug combination does not reduce the Sch9 protein level in this assay (Fig. S3). This difference may be due to the C-terminal epitope tag or overproduction of Sch9 protein, as in this assay, cells express the SCH9-5HA allele from a CEN-vector and the wild-type SCH9 allele from the chromosome. The reduction in TORC1 phosphorylation of Sch9 caused by the combination drug treatment is larger than an additive effect of the two drugs, suggesting that at least some degree of synergy is generated upstream of TORC1. The drug combination is having a similar but smaller effect on Pkh1/2-mediated phosphorylation of T570. Determining how the drug combination influences the level of Sch9 and the Pkh1/2 pathway will require further study.
Another branch of the TORC1 pathway controls activity of the Gln3 transcription activator which regulates expression of genes involved in nitrogen metabolism, including the MEP2 gene encoding an ammonium transporter (Marini et al., 1997). Active TORC1 represses MEP2 expression, but the combination drug treatment induces a 13-fold increase in expression of a MEP2-lacZ reporter gene compared with no drug or a single drug treatment, indicating reduction in TORC1 activity by dual drug treatment (Fig. 4C). For comparison, strongly reducing TORC1 activity by high-dose rapamycin treatment induces MEP2 expression by 32-fold (Fig. 4C). These data verify that the combination drug treatment reduces TORC1 activity to a greater degree than just an additive effect of the single drug treatments.
Activation of PKA
Target of rapamycin protein kinase complex 1 and Sch9 function in networks to regulate cellular processes, and one network link is with (PKA) which, like TORC1 and Sch9, plays roles in nutrient and stress sensing and CLS (Fig. 1A) (Soulard et al., 2010; De Virgilio, 2012; Longo et al., 2012). To determine whether the combination drug treatment lowers PKA activity, we examined Atg13 phosphorylation, as the phosphorylated form of this protein maintains autophagic flux at a low basal level in log-phase cells (Stephan et al., 2009). Combination drug treatment causes a decrease in PKA-mediated phosphorylation of Atg13 compared with single or no drug treatment (Fig. 4D, top panel). Quantification of the ratio of PKA-phosphorylated Atg13 to total Atg13 protein shows that the effect of dual drug treatment is greater than additive (Fig. 4D, Atg13-P/Atg13). These results along with the GFP-Atg8 cleavage results and atg1Δ CLS data (Fig. 2A,B) show that the combination drug treatment increases autophagic flux, and this contributes to enhancement of lifespan, all of which support the idea that combination drug treatment reduces PKA pathway activity.
The AMP-activated protein kinase (AMPK) is an evolutionarily conserved cellular energy sensor with vital roles in aging and lifespan (Hardie et al., 2012). Snf1, the yeast paralog of human AMPK, also performs numerous functions in yeast including ones essential for long life (De Virgilio, 2012). Snf1 activity is low during log-phase growth due to glucose repression, but rises as glucose is consumed. This decrease in glucose triggers the switch in energy generation from fermentation to respiration or the diauxic shift. Log-phase cells treated with both drugs have an increased rate of respiration, suggesting that they have higher Snf1 activity compared with cells treated with one drug or no drug (Fig. 3A).
As another way to examine Snf1 activity, we measured expression of the ADH2 gene, encoding alcohol dehydrogenase, using an ADH2-lacZ reporter assay (Young et al., 2003). Combination drug treatment produces a small, but statistically significant increase in ADH2-lacZ expression in log-phase cells, a fourfold increase at the postdiauxic shift and a threefold increase at stationary phase relative to no drug or single drug treatment (Fig. 4E). Thus, dual drug treatment remodels intracellular signaling to promote activation of Snf1 and the processes it controls including respiration (Fig. 1A).
Our data for DBY746 cells demonstrate that it is possible to produce a synergistic increase in lifespan by using low concentrations of rapamycin and myriocin (Table 1). As far as we know, this is the first example of synergistic lifespan enhancement produced by a drug combination. There seems to be no simple relationship between synergistic enhancement of survival and the concentration of drugs used to treat cells, indicating that synergy and survival result from multiple effects of the drugs, and these in turn depend upon the concentration of each drug rather than their ratio.
Our data show that rather than simply reducing Sch9 activity, dual drug treatment produces more widespread effects and increases Snf1/AMPK activity, while lowering activity of the PKA and TORC1 pathways. Many stresses are known to trigger a decrease in TORC1 activity in yeast, and one mechanism is through binding with the Rho1 GTPase (Yan et al., 2012), an essential element of the mechanisms for sensing cell wall and membrane stress. We suggest that the combination drug treatment may, at least in part, produce a large reduction in TORC1 activity by inducing membrane and cell wall stress, which activates Rho1 causing it to bind and inhibit TORC1. Even if true, however, this explanation is likely to be only part of the overall mechanism for producing the synergy effects we observe. A thorough understanding of how the combination of myriocin and rapamycin produces a synergistic increase in yeast lifespan will likely require systems biology techniques, as the number and complexity of the signaling pathways and cellular processes uncovered by our data will be a challenge to fully understand and integrate into a coherent system by using purely intuitive approaches.
Drug synergy is an established protocol for treating infectious diseases and cancers, but is not widely touted as a strategy to prevent disease or slow aging. The idea of using rapamycin to lower TOR activity and reduce the incidence of age-related diseases and promote a healthier, longer life has been well publicized [e.g., (Blagosklonny, 2006, 2009; Bjedov & Partridge, 2011)], particularly, because it was shown to enhance lifespan in mice (Harrison et al., 2009). The therapeutic efficacy of rapamycin for treatment of cancers and other diseases is currently being evaluated in clinical trials using rapamycin alone or in combination with other drugs (http://clinicaltrials.gov). However, low-dose rapamycin treatment over a long time frame has the undesirable effect of inducing insulin resistance (Lamming et al., 2012). Greater efficacy and fewer side effects may be achieved by using rapamycin in combination with myriocin or another drug to generate synergy. Moreover, low-dose rapamycin therapy in combination with myriocin may be an effective way to lower the incidence of age-related diseases, as studies in rodents indicate the value of myriocin to treat diabetes, cancer, cardiomyopathy, and atherosclerosis (Summers, 2010; Jiang et al., 2011; Lee et al., 2012).
The cellular processes modulated by combination drug treatment in yeasts to enhance lifespan are evolutionarily conserved and this, along with data establishing the therapeutic value of rapamycin and myriocin in mammals, argues that reducing the incidence of age-related diseases in humans and improving the quality of life in the elderly by drug synergy is an achievable goal.
Yeast strains and plasmids used in this study are listed in Supplemental Information. The composition of synthetic dextrose complete (SDC) was as previously defined (Huang et al., 2012). For most experiments, cells were cultured in SDC medium buffered to pH 4.5 with 200 mm succinate (Fig. 1B,C,F, Figs. 2-4, and Figs. S1, S2, and S3). Experiments with BY4743 cells use SDC medium buffered to pH 6.0 with citrate phosphate (64.2 mm Na2HPO4, 17.9 mm citric acid) (Fig. 1D,E,G).
Chronological lifespan was measured as previously described (Huang et al., 2012). For CLS assays involving a switch from SDC medium to water, the cells were centrifuged after 72 h of incubation (CLS day 1), washed twice with water, and then cultured in the same volume of water as the volume of the initial culture medium. Every third day, the cells were washed twice and suspended in fresh water. Stress resistance, age-dependent genomic stability, and immunoblotting assays were performed as previously described (Huang et al., 2012). Antibodies for immunoblotting included anti-GFP (1:4000) (Cat#: MMS-118P; Covance, Princeton, NJ, USA), anti-Sch9 (Batch 2872, 1:1000), phospho-specific anti-Sch9T570-P (1:10 000) (Urban et al., 2007), phospho (Ser/Thr) anti-PKA substrate (1:1000) (Cat#: 9621; Cell signaling, Danvers, MA, USA), anti-HA (1:4000) (Sigma-Aldrich, St. Louis, MO, USA), anti-Vma2 (1:4000) (Cat#: A6427; Invitrogen, Grand Island, NY, USA). Other antibodies included alkaline phosphatase-linked anti-rabbit or anti-mouse IgG (Sigma-Aldrich). Fluorescent signals from membranes exposed to an ECF substrate (Amersham Biosciences, Pittsburgh, PA, USA) were analyzed using a PhosphorImager (ChemiDoc MP Imaging System; Bio-Rad, Hercules, CA, USA) and quantified using Image Lab 4.0.1 Software (Belgrade, MT, USA). C-terminal phosphorylation of Sch9 was measured as described previously (Urban et al., 2007). GFP-Atg8 processing was measured as described previously (Cheong & Klionsky, 2008).
To immunoprecipitate Atg13-3HA, DBY746 cells carrying pRS426-ATG13-3HA were cultured in SDC-ura supplemented with 100 μm CuSO4. Cells were grown from 0.005 to 1.0 A600 nm. Then 25 A600 nm units of cells were collected, and total protein was extracted as described above. Anti-HA-coupled Sepharose beads, 30 μL (Cat#: AFC-101P; Covance), were washed three times with 800 μL of wash buffer [50 mm Tris-HCl, pH 7.5, 1 mm EDTA, 1% NP-40, 1 × phosphatase inhibitors, 10 mm NaF, 10 mm NaN3, 10 mm p-nitrophenylphosphate, 10 mm Na2P2O7, and 10 mm beta-glycerophosphate] in a 1.5 mL microfuge tube followed by centrifugation for 2 min at 800 g. The supernatant was carefully removed, and the beads were resuspended using 1 volume of protein extract (3 mg total protein) and 11 volumes of IP buffer [50 mm Tris-HCl, pH 7.5, 150 mm NaCl, 1 mm EDTA, 1% NP-40, 1 mm PMSF (add freshly), 1 × phosphatase inhibitors and protease inhibitor cocktail (Cat#: 05892970001; Roche, Indianapolis, IN, USA)]. Samples were incubated overnight at 4 °C with gentle rotation. After incubation, the beads were washed three times with 1 mL wash buffer (this buffer contains 1 × phosphatase inhibitors, but no PMSF or protease inhibitors) and centrifuged 2 min at 2500 rpm. All liquid was removed from beads and then, 25 μL 2 × Laemmli sample buffer was added followed by heating at 95 °C for 2 min. The sample was gently mixed and heated again for 4 min, followed by 1 min on ice, and loading on a 7.5% SDS-PAGE.
The rate of oxygen consumption was measured as previously described (Huang et al., 2012) using cells grown from an A600 nm of 0.005 to 1. Oxygen consumption was measured using a Hansatech Oxytherm monitor, and the data are expressed as nanomoles of oxygen consumed per min per 106 cells. For the ROS measurements, DHE staining was used to measure superoxide, and 2′,7′-dichlorodihydrofluorescein staining was used to measure hydrogen peroxide (Mesquita et al., 2010). Log (A600 nm = 1.0)- or stationary-phase cells were collected, washed once with PBS, and resuspended in PBS containing a 10 μm DHE or 10 μm H2DCF-DA. For DHE staining, cells were incubated for 10 min, and for H2DCF-DA staining, they were incubated for 90 min at 30 °C and then washed twice with PBS and resuspend in 1 mL PBS. Fluorescence was measured in a fluorimeter (55 Luminescence Spectrometer; Perkin-Elmer, Waltham, MA, USA) using 488 nm wavelength for excitation and 670 nm for emission with DHE-stained cells, and 488 nm for excitation, and 530 nm emission with H2DCF-DA-stained cells. Data are presented as DHE or H2DCF-DA fluorescence per 108 cells corrected for the background fluorescence of unstained cells.
β-galactosidase activity was measured in cells grown to log (A600 nm = 1.0), postdiauxic shift or stationary phase. Cells were collected by centrifugation and resuspended in 1 mL of Z buffer (0.06 m Na2HPO4, 0.04 m NaH2PO4, 0.01 m KCl, 0.001 m MgSO4, 0.05 m β-mercaptoethanol, pH 7.0). Cells were made permeable by adding 25 μL chloroform and 40 μL 0.1% SDS followed by vortexing. β-galactosidase activity was expressed as nmol per min per 108 cells.
Statistical analysis was performed using a two-tailed Student's t-test.
We thank Drs. B. Andre, M. Carlson, M. Hall, P. Herman, Y. Jiang, D. Klionsky, R. Loewith, and T. Young for reagents and suggestions. This work was supported by grant AG024377 from the National Institutes of Health (RCD) and in part by NIH Grant Number P20GM103486 from the National Institute of General Medical Sciences; its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH or the NIGMS. Jun Liu was sponsored by the China Scholarship Council.
Drs. Dickson, Leggas, and Huang designed the study. Dr. Huang, Jun Liu, Aaron Samide, and Brad Withers performed experiments. Dr. Leggas performed data analysis to assess deviation from additivity in the survival experiments using drug combinations. All authors contributed reagents and suggestions to the studies. Dr. Dickson wrote the text, and the other authors contributed to the final text presentation.