Biochimica et Biophysica Acta 1813 (2011) 315–321

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bbamcr

Evidence for a second messenger function of dUTP during Bax mediated apoptosis ofyeast and mammalian cells

Drew Williams a, Grant Norman b, Chamel Khoury c, Naomi Metcalfe b, Jennie Briard b, Aimee Laporte b,Sara Sheibani b, Liam Portt b, Craig A. Mandato a, Michael T. Greenwood b,⁎a Department of Anatomy and Cell Biology, McGill University. Montreal, Quebec, Canadab Department of Chemistry and Chemical Engineering, Royal Military College (RMC), PO Box 17000, Station Forces, Kingston, Ontario, Canada K7K 7B4c University of Graz, Institute for Molecular Biosciences, Austria

⁎ Corresponding author. Tel.: +1 613 641 6000x3575E-mail address: michael.greenwood@rmc.ca (M.T. G

0167-4889/$ – see front matter © 2010 Elsevier B.V. Aldoi:10.1016/j.bbamcr.2010.11.021

a b s t r a c t

a r t i c l e i n f o

Article history:Received 7 September 2010Received in revised form 2 November 2010Accepted 29 November 2010Available online 8 December 2010

Keywords:ApoptosisCell deathAnti-deathAnti-apoptosisAnti-apoptoticCell survivalYeastC2C12 cellsdUTP

The identification of novel anti-apoptotic sequences has lead to new insights into the mechanisms involved inregulating different forms of programmed cell death. For example, the anti-apoptotic function of free radicalscavenging proteins supports the pro-apoptotic function of Reactive Oxygen Species (ROS). Using yeast as amodel of eukaryotic mitochondrial apoptosis, we show that a cDNA corresponding to the mitochondrialvariant of the human DUT gene (DUT-M) encoding the deoxyuridine triphosphatase (dUTPase) enzyme canprevent apoptosis in yeast in response to internal (Bax expression) and to exogenous (H2O2 and cadmium)stresses. Of interest, cell death was not prevented under culture conditions modeling chronological aging,suggesting that DUT-M only protects dividing cells. The anti-apoptotic function of DUT-M was confirmed bydemonstrating that an increase in dUTPase protein levels is sufficient to confer increased resistance to H2O2 incultured C2C12mouse skeletal myoblasts. Given that the function of dUTPase is to decrease the levels of dUTP,our results strongly support an emerging role for dUTP as a pro-apoptotic second messenger in the same veinas ROS and ceramide.

; fax: +1 613 542 9489.reenwood).

l rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Apoptotic cell death (type I) is a genetically programmedmechanism that allows the cell to commit suicide when conditionsare judged to be unfavorable [1]. Although the term apoptosis is oftenused to denote all forms of programmed cell death (PCD), it should benoted that genetically encoded cell death can occur via a number ofdifferent but likely interrelated mechanisms such as autophagy (typeII PCD) and necrosis (type III PCD) [2]. The decision on whether a cellcommits suicide in response to a given stress is dependent on theinterplay of a number of regulatory pro- and anti-apoptotic proteins[3]. For example, Bax, the most well studied pro-apoptotic member ofthe Bcl-2 family of proteins, functions, at least in part, by interactingwith the mitochondrial membrane and assisting the release of pro-apoptotic molecules such as cytochrome c. Other proteins likecaspases become activated by stress responsive cascades and theyserve to promote cell death by cleaving a variety of cellular substratesthat lead to death [4]. In contrast, many of the anti-apoptotic proteinsfunction by directly antagonizing apoptotic proteins. For example, the

anti-apoptotic Bcl-2 is probably the most studied inhibitor ofapoptosis. It functions in part, by opposing the effects of pro-apoptoticBcl-2 members such as Bax [5]. Other anti-apoptotic proteins includethe structurally diverse group of heat shock proteins (HSPs), a varietyof antioxidant proteins as well as a number of other proteins ofunknown function such as the GTPase GIMAP8 [6,7]. In spite of recentadvances, more is known about the processes involved in initiatingdeath than in the processes involved in preventing cell death [8,9].One factor that likely serves to limit our knowledge of anti-apoptosisis the fact that the complete repertoire of cell survival proteins is notknown.

A number of different approaches have been successful inidentifying novel anti-apoptotic genes. One successful strategy is tocharacterize genes that are up-regulated in a variety of differentconditions where cells have increased resistance to apoptosis. Anumber of anti-apoptotic genes have thus been identified fromcommonly described apoptotic resistant cells including cancer cells,cells treated with sublethal levels of stresses (i.e. pre-conditioning)and neuronal cells subjected to increased stimulation [10–14]. Thedisadvantage here is the large number of genes that are up-regulatedin such cells and the lack of knowledge regarding their potentialas anti-apoptotic sequences. Another widely used strategy involvesthe screening of heterologous cDNA expression libraries in yeast

316 D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

ectopically expressing a mammalian cDNA corresponding to the pro-apoptotic Bax [15]. Such screens have the advantage of yieldingsequences that have pro-survival functions. Such an approach hasbeen validated by the fact that the characterization of numeroussequences identified as Bax suppressors in yeast have been shown tobe anti-apoptotic when overexpressed in mammalian cells [16–18].This is not surprising given that many different stresses including ROSdonors and ectopically expressed Bax leads to Programmed Cell Deathin yeast that shares a great deal of similarities with mammalianmitochondrial centered apoptosis [15]. The similarity is such thatyeast is now a widely used model for apoptosis and other aspects ofPCD [19–21]. In addition, this functional interchangeability of yeastand mammalian genes, has for many years lead to the use of so-called“humanized yeast” as a useful model to study the structure andfunction of heterologously expressed mammalian genes [18]. Wehave previously carried out Bax suppressor screens and havedescribed the isolation of several potential mammalian anti-apoptoticsequences [22]. Here, we report that one of these uncharacterizedclones encodes a cDNA corresponding to mitochondrially targeteddUTPase protein encoded by the human DUT gene. We further showthat overexpression of the dUTPase cDNA renders yeast andmammalian cells resistant to stress induced cell death. Our resultsprovide compelling evidence that serves to support the notion thatdUTP is a secondmessenger that mediates the pro-apoptotic effects ofcellular stresses in growing cells.

2. Materials and methods

2.1. Yeast strains and plasmids

The Saccharomyces cerevisiae BY4741 (MATa his3Δ1 leu2Δ0met15Δ0 ura3Δ0) strain was used throughout this study (EURO-SCARF). The DUT-M and HSP90β clones, expressed under the controlof the galactose inducible GAL1 promoter in pYES-DEST52, wasisolated in our previous screen of a human cardiac cDNA library forBax suppressors [22]. Of the multitude of different clones isolated inthe screen, DUT-M was the only cDNA that was isolated threeindependent times.

2.2. Yeast transformations, growth and viability assays

Yeast were grown in synthetic minimal media containing yeastnitrogen base (YNB), 2% glucose and supplemented with the requiredamino acids (aa) or bases.When expression of the GAL1 promoter wasrequired, the glucose was substituted with 2% galactose and 2%raffinose. Yeast cells were transformed with the appropriate plasmidsusing the lithium acetate method and transformants were selected forby the omission of the appropriate nutrient (uracil for pYES-DEST52plasmids). The richmedia used consisted of 2% bactopeptone, 1% yeastextract and 2% glucose (YEPD). The clonogenecity assay was used todetermine viability. Briefly, freshly saturated overnight cultures of thedifferent yeast transformants were diluted in fresh, pre-warmedgalactose-containing media, incubated for 4 h to induce geneexpression, and subsequently treated with the indicated concentra-tion of H2O2 or cadmium. Aliquots were harvested, serially diluted andtriplicate samples of 300 cells were then plated on YEPDmedia, grownat 30 °C and the number of colonies that formed after 2 days werecounted. The same procedure was used to generate chronologicallyaged cultures except that the cells were continuously incubated for14 days and samples were harvested daily to determine viability.Alternatively, viability was also determined by microscopic examina-tion of cells stained with the vital dye trypan blue. Cells were stainedwith 0.1% trypan blue for 5 min and at least 300 cells were scored foreach time point.

2.3. C2C12 cell culture

Mouse C2C12 myoblasts were cultured at 37 °C in a controlledincubator of 5% CO2, maintained in growthmedium containing DMEM(Dulbecco's modified Eagle's medium) and supplemented with 10%fetal bovine serum and antibiotics (streptomycin and penicillin) [23].Cells were infected in 100-mm tissue culture plates at ≈80%confluency with HSV (herpes simplex virus) infectious stock toallow for 1:150 dilution of virus:growth medium. Cells were thenincubated in 37 °C CO2 incubator for 4 h, after which cell culture plateswere washed 3 times with PBS and replaced with growth medium.Cells were then incubated for 24 h in a 37 °C CO2 incubator. Oxidativestress was stimulated by exposing cells to different concentrations ofH202 in DMEM without serum or antibiotics for 24 h in a 37 °C CO2

incubator. Following H2O2 exposure, cells were pelleted by centrifu-gation at 3000 rpm for 3 min and analyzed by microscopy followingstaining with DAPI or via APO-BrdU TUNEL assay (Invitrogen).

2.4. Expression of Recombinant HSV constructs

The cDNAs corresponding to the entire open reading frame ofDUT-M and HSP90β were individually inserted into an ampliconplasmid pHSV-PrpUC, containing the immediate early promoter 4/5 ofHSV, and an HSV-packaging site [24]. The plasmid (2 μg) wassubsequently transfected into the HSV packaging cell line 2-2(3×105) with lipofectamine (12 μl) and then infected with 5 dl5HSV helper virus after 24 h after incubation. The recombinant viruswas amplified through three sequential rounds of infection and thenstored at -80 °C.

2.5. APO-BrdU TUNEL and flow cytometry analysis

Cells were grown in 100-mm tissue culture plates until ≈80%confluent, infected with HSV-constructs, and exposed to H2O2 asdescribed above. Apoptotic cells were detected by the terminaldeoxynucleotidyl transferase biotin-dUTP nick end labeling (TUNEL)staining using the fluorescence method with the APO-BrdU TUNELassay kit (Invitrogen) according to manufacturer's instructions.Following TUNEL protocol, cells were immediately analyzed in aFACScan (Becton Dickinson). The percentage of apoptotic positivecells was performed with Cell Fit software (Becton Dickinson).

2.6. Protein extraction and western blot analysis

Soluble protein was extracted from cultured cells using 1 ml of ice-cold lysis buffer per dish (50 mM Tris–HCl (pH 8.0), 150 mM NaCl,0.1% SDS, 100 μg/ml PMSF, and 1% NP-40). Twenty μg of solubleproteins were separated by SDS-PAGE, transferred to nitrocellulosemembrane and incubated with different primary anti-serum [25].Commercially available rabbit anti-dUTPase, HSP90 and of α-tubulinfrom Santa-Cruz Biotechnology (Santa Cruz, California) were used asdescribed by the manufacturer. HRP-conjugated secondary anti-serum was used and signals were subsequently detected with ECLplus (Amersham Bioscience) followed by exposure to X-ray film(Kodak X-Omat).

3. Results

Our lab previously isolated multiple mammalian cDNAs capable ofsuppressing the negative effects of Bax in yeast [22]. We have sincecharacterized four of these Bax suppressors and have determined thatthey are anti-apoptotic sequences since they are capable of preventingor delaying cell death in response to a variety of other pro-apoptoticstimuli in yeast [15,22,26,27]. Here, we report the characterization ofanother Bax suppressor corresponding to a human cDNA encoding apredicted 252 residue protein. Analysis of the sequence by comparing

317D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

it to the sequences in GenBank revealed that our cDNA was identicalto the sequence of the DUT gene encoding deoxyuridine tripho-sphatase (dUTPase). The human DUT gene is alternatively spliced toproduce 2 different proteins a 252 mitochondrial variant (DUT-M)and an N-terminally truncated 164 residue nuclear localized variant(DUT-N) [28]. Our sequence corresponds to the mitochondrial variantand we will refer to it as DUT-M from here on in.

3.1. DUT-M suppresses the lethal effects of 5-fluouracil

The anti-apoptotic function of DUT-M may be related to the factthat it is a mitochondrially localized protein. Overexpression of such aproteinmay somehow interfere with the ability of stress to trigger thenecessary mitochondrial changes required to trigger apoptosis. Morelikely, the anti-apoptotic property of DUT-M is related to the dUTPaseactivity of its protein product. It is known that dUTP is cytotoxic andwe thus reasoned that dUTPase, which serves to hydrolyze dUTP intodUMP and pyrophosphate (PPi), should be able to protect cells fromelevated dUTP levels [29]. To test this possibility the uracil analog,5-fluouracil (5-FU) which is a thymidylate synthase inhibitor, wasused. Blocking thymidylate synthase (TS) leads to an accumulation ofits substrate, dUMP, which subsequently leads to an elevation in thelevels of dUTP [30,31]. The physiology of TS inhibition is very wellstudied since it a commonly used chemotherapeutic approach [32,33].The biochemical pathways involved in thymidylate metabolism arelargely conserved between yeast and man, thus it not surprising thatthe inhibition of TS using 5-FU similarly leads to death and increaseddUTP levels [34]. Thus both control and DUT-M expressing yeast weretreated with 5-FU and viability was determined using the clonogenicassay 0, 2 and 4 h after treatment. The viability of control cells wasreduced to 75±1.8% and to 51±1.8% following a respective 2 and 4 htreatment with 5-FU (Fig. 1). In contrast, viability was remainedelevated in DUT-M expressing cells to a respective to 97.2±1.8% and70.1±1.8% after 2 and 4 h of 5-FU exposure (Fig. 1). These resultsdemonstrate that elevated levels of dUTPase confer resistance to 5-FUtoxicity in yeast as it has previously been observed in mammaliancells [32].

3.2. DUT-M does not protect non cycling cells

The inhibition of TS by agents such as 5-FU leads to thyminelessdeath that occurs in response, in part, to the perturbations in pools ofDNA synthesis precursors and the lack of the DNA synthesis precursordTTP [35]. The accompanying increase in dUTP levels further

Fig. 1. Expression of human DUT-M protects yeast cells from the toxic effects of thethymidylate synthase inhibitor 5-fluorouracil (5-FU). Yeast cells harbouring controlempty plasmid (VECTOR) or the DUT-M expressing vector (DUT-M) were grown ingalactose nutrient media and treated with 5-FU. Viability was determined using theclonegicity assay in samples harvested prior to (black bars), 2 h (grey bars) and 4 h(white bars) after the addition of 5-FU. Viability is shown as the percentage (%) of cellsthat remain viable after treatment. The data represent the mean of 3 independentexperiments (±SEM). *Student t test pb0.001.

exasperates the situation, since it can replace dTTP and getincorporated into DNA and thus lead to damaged DNA. The fact thatTS inhibition impinges on cells actively undergoing DNA synthesishelps to explain the increased sensitivity of tumors to 5-FU and theincreased resistance exhibited by non cycling cells. In order to confirmthat DUT-M does not protect G0 yeast cells from cell death we usedthe fact that in yeast cultures, it has been shown that chronologicalageing leads to physiologically induced apoptosis. Thus, to furtherexamine the potential anti-apoptotic role of DUT-M, we examinedwhether DUT-M overexpression allows for increased survival duringchronological ageing of yeast. A period of 14 days in cultivation canlead to apoptotic cell death for over 90% of the yeast culture [36]. Wethus examined the viability of control and DUT-M expressing yeastcells in cultures that were allowed to enter stationary phase. Viabilitywas found to decrease to less than 1% over a period of 13 days with noobservable differences between control and DUT-M expressing cells(Fig. 2). These results demonstrate that DUT-M does not protect dyingG0 yeast cells and is consistent with the notion that elevated dUTP isunlikely to play a role in mediating all forms of cell death.

3.3. Overexpression of DUT-M protects yeast against apoptotic inducingstresses in yeast

A number of Bax suppressors from a variety of different specieshave been identified in yeast [18]. Many of these have beensubsequently shown to represent anti-apoptotic sequences. Never-theless it remains that proteins may be able to suppress the effects ofBax in yeast without actually being anti-apoptotic. For example, giventhat Bax expression has pleiotropic effects including the inhibition ofcell growth, it remains that a sequence may serve to promote cellgrowth and thus act as a Bax suppressor. Alternatively a sequencemayact as a Bax binding protein and thus sequester Bax and reduce itsability to induce cell death. Therefore, in order to determine if DUT-Mis capable of functioning as an anti-apoptotic sequence, we observedwhether DUT-M overexpression in yeast could prevent cell death inresponse to H2O2 and cadmium treatment, each of which are knownto induce apoptotic cell death in yeast [37,38].

In the absence of H2O2 or cadmium (Cd++) viability of yeasttransformed with empty vector or with DUT-M was close to 100%(Fig. 3). Following treatment with 4 mM H2O2 for 4 h the viability ofyeast cells expressing a control empty vector was decreased to 52.7±2.6% while the viability of yeast cells overexpressing DUT-M treatedwith H2O2 was significantly higher at 71.0±2.7% (pb0.001; n=3)(Fig. 3A). Treating cells with Cd++ reduced the viability of yeast cellsharbouring a control empty vector to 12.4±1.9% (Fig. 3C). A pro-tective effect was observed when yeast cells overexpressing DUT-M

Fig. 2. The effect of DUT-M expression on the viability of chronologically aging yeastcultures. Yeast cultures were inoculated into galactose containing nutrient media at lowdensity and the cells were allowed to incubate at 30 °C for 13 days. Samples were takendaily and viability was determined using the clonogenic assay. The data represent themean of three samples and similar results were obtained in three independentexperiments.

Fig. 3. Expression of human DUT-M protects yeast cells from apoptotic inducingstresses. Yeast cells harbouring control empty plasmid (VECTOR) or the DUT-Mexpressing vector (DUT-M) were grown in galactose nutrient media for 4 h prior to theaddition of the apoptotic inducing agents. Yeast cells were treated with A, 4 mM H2O2

for 4 h and B and C, 1.5 mM cadmium sulfate for 24 h and viability was determinedusing two different methodologies. In A and C, viability was determined using theclonegicity assay while in B, viability was determined by microscopic examination ofcells stained with the vital dye trypan blue. The data are shown as the percentage ofcells that remain viable (Viability (%)) for A and C and as the percentage of cells that donot stain with trypan blue (Trypan blue negative (%)) for B. The data represent themean of at least 5 independent experiments (±SEM). *Student t test pb0.001.

Fig. 4. Dose dependant effect of H2O2 on the viability of C2C12 cells. A, C2C12 cells weretreated with 1.2 mM H2O2 (+H2O2) or without H2O2 (−H2O2) for 24 h, fixed, stainedwith DAPI, visualized by fluorescence microscopy and photographed. The scale barrepresents 10 μm. B. Identical cultures of C2C12 cells were treated without (0) orincreasing concentrations of H2O2 (0.3 to 1.2 mM) for 24 h. The cells were processed forTUNEL and the percentage of apoptotic cells (% TUNEL positive) was determined byflow cytometry. Viability is shown as the percentage of cells that are undergoing DNAdamage as a prelude to apoptosis (% TUNEL positive). The data represent the mean of 3independent experiments (±SEM). The percentage of TUNEL positive cells issignificantly increased from untreated cells for all concentrations of hydrogen peroxide(H2O2) used (Student t test pb0.001).

318 D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

were exposed to cadmium, viability increased to 48.4±1.4%,respectively (pb0.001; n=3) (Fig. 3C). As an independent assay forcell death, cell viability was also monitored using the vital dye trypanblue. Using trypan blue and following the same treatment with Cd++,

viability of control cells was found to be decreased to 49.9±1.8%(Fig. 3B). Because trypan blue is used to examine cell viability directlyafter treatment it overestimates the viability of cultures since anumber of cells exclude vital dyes even though these cells are unableto form colonies. Such differences in the viability have been notedbefore [22]. Nevertheless in spite of the overestimation of viability,cells overexpressing DUT-M show increased viability at 73.4±3.5%

using trypan blue (pb0.0001; n=3) (Fig. 3B). These data suggest thatdUTPase is another one of a multitude of mammalian anti-apoptoticsequences, such as the ceramide utilizing sphingomyelin synthase andthe small G-protein Ran, that were first identified as Bax suppressorsin yeast viability screens [17,22].

3.4. Optimization of H2O2 as an apoptotic inducer in C2C12 cells

Next, we wanted to test if DUT-M displayed anti-apoptotic activityin mammalian cells. Although dUTPase has been shown to protectcells mammalian cells from 5-FU, its ability to protect from apoptoticinducing stresses has not been examined. We decided to use themouse skeletal muscle cell line C2C12 since we and others have usedthese cells to study a variety of different aspects of apoptosis [25,39].In addition, H2O2, a commonly used and effective inducer ofmitochondrial apoptosis has been shown to be effective in thesecells [39]. We first had to establish conditions that would allow thedetection of sufficient cell death in order to be able and detect anyincreases in viability due to DUT-M overexpression. Control C2C12myoblasts were treated with increasing concentration of H2O2 (up to1.2 mM) for 24 h before being fixed and stained with DAPI.Fluorescent microscopy of cells revealed that a significant proportionof cells (ca. 50%, not shown) were undergoing apoptosis aftertreatment with 1.2 mM H2O2 for 24 h as observed by enlarged nuclei(Fig. 4A). To quantitatively assess H2O2-induced apoptosis in C2C12cells, DNA fragmentationwas determined via TUNEL labeling and flowcytometric analysis. As seen in Fig. 4B, the appearance of apoptoticcells was increased by H2O2 in a dose-dependent manner: 3.5±0.9%(control cells), 9.6±1.2% (0.3 mM), 25.9±2.1% (0.6 mM), 35.6±2.4%(0.9 mM), and 58.4±2.9% (1.2 mM).We thus decided to pursue theseexperiments using 1.0 mM H2O2.

319D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

3.5. Overexpression of DUT-M inhibits H2O2-induced apoptosis in C2C12cells

To determine the effect of overexpressing DUT-M on apoptotic celldeath in C2C12 cells, we first carried out western blot analysis usingextracts prepared from control C2C12 cells or from cells over-expressing HSP90β and DUT-M using protein specific antibodies.HSP90β was used as a positive control it is a known anti-apoptoticsequence [40]. The results obtained demonstrate that cells transducedwith the HSP90β and DUT-M viral expression constructs showincreased levels of these proteins (Fig. 5A). A viral system was usedto overexpress our cDNAs because transduction with this systemresulted in negligible increases in basal levels of apoptosis (from 3% to5%) and efficiency was close to 100% (as determined by transducing aGFP expressing viral construct). Traditional methodologies used tointroduce plasmid based cDNA overexpression constructs into C2C12lead to lower levels of transfection efficiencies (30–40%, in our hands)and to a significant increase in the basal level of apoptosis (from 3% to15%). Control cells with an empty virus and cells transduced withvirus overexpressing HSP90β or DUT-Mwere then exposed to 1.0 mMH2O2 for 24 h, labeled with TUNEL and analyzed by flow cytometry(Fig. 5B). Under these conditions, we found that 38.8±1.1% of controlH2O2-treated C2C12 cells were apoptotic (Fig. 5C). In contrast,apoptosis was reduced to 20.7±1.0% (pb0.0001; n=3) in H2O2-treated cells overexpressing HSP90β. Similarly, H2O2-treated cellsoverexpressing DUT-M were also protected as observed by thereduced the proportion of apoptotic cells to 14.6±0.8% (pb0.0001;n=3). Thus dUTPase has the ability to prevent stress mediated

Fig. 5. The expression and characterization of the effects of DUT-M and HSP90β overexprerecombinant HSV constructs expressing DUT-M or HSP90β or empty HSV which served as thvirus or with the DUT-M or HSP90β viruses and analyzed by western blot using protein speccells overexpressing (O/E) their respective viral expression vector as compared to cells transAb was also determined and served as a loading control. The experiment was reproduced t(Vector), HSP90β or DUT-M were treated with 1.0 mM H2O2 for 24 h and the TUNEL assay wcount of TUNEL cells having different distribution of fluorescence corresponding to cells havin(104). The results obtained in C were compiled and the percentage of cells corresponding to tmean of 3 independent experiments (±SEM). *Student t test pb0.0001.

apoptosis and thus it has one of the major characteristic of an anti-apoptotic gene [14].

4. Discussion

The identification of genes that are anti-apoptotic when over-expressed has been important in furthering our understanding of theprocesses involved in regulating apoptosis [3,41,42]. In effect, theearly realization that increased expression of Bcl-2 confers the abilityto evade apoptosis in many tumors was such a landmark event [43].Others include the fact that proteins acting as scavengers of ReactiveOxygen Species (ROS) are powerful negative regulators of stressmediated apoptosis [44]. Similarly, the identification of the ceramideutilizing sphingomyelin synthase 1 (SMS1) as anti-apoptotic se-quence strongly supports the long held view that ceramide is a stressinduced pro-apoptotic second messenger [45]. The identification of acDNA encoding the dUTP hydrolyzing enzyme dUTP pyrophosphatase(dUTPase) as a Bax suppressor, provides compelling evidence thatdUTP is involved in triggering apoptosis. Consistent with previousreports in mammalian cells, we have shown that the heterologousexpression of DUT-M in yeast serves to increase dUTPase activity sinceit prevented cell death in cells having increased levels of dUTP due to a5-FUmediated inhibition of thymidylate synthase (Fig. 1) [33,35]. Thefact that we were able to show that DUT-M suppresses the pro-apoptotic effects of ROS stress in both yeast and mammalian cells(Figs. 3 and 5) further supports a central role for dUTP as amediator ofcell death. Our demonstration that DUT-M does not prevent cell deathin stationary phase chronologically aged yeast cultures (Fig. 2) is

ssion on the viability of H2O2 treated C2C12 cells. C2C12 cells were transduced withe control cells. (A) Protein extracts were prepared from the cells transduced with emptyific Ab for DUT-M or HSP90β. The levels of DUT-M or HSP90β proteins were elevated induced with the empty HSV virus (Basal). The levels of α-tubulin using a protein specificwice with similar results. (B, C) Identical C2C12 cultures transduced with empty virusas subsequently carried out and analyzed by FACS. (C) Graphs showing the relative cellg intact DNA (102) and cells in which the DNA is cleaved and labeled by TUNEL staining

he percentage of apoptotic cells (%) is shown as a histogram in B. The data shown are the

320 D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

consistent with the notion that dUTP serves to kill cells by interferingwith DNA synthesis [29,46].

Thus given that the primary role of dUTPase is to rid the cell ofdUTP together with our results showing that dUTPase is anti-apoptotic, we hypothesize that dUTP increases in response to certainstresses and it has a regulatory role as an intracellular pro-apoptoticinducer. A number of previous observations serve to strongly supportthis hypothesis: 1—DUT is an essential gene [47], 2—dUTP is toxicto the cell since it can replace dTTP and get incorporated into DNA[48], 3—blocking thymidylate synthase leads to cell death in partbecause of an increase in dUTP levels while overexpression ofdUTPase is protective [30–32], 4—dUTP induced cell death has beenreported to kill unwanted cells during fruit fly development and inresponse to specific genotoxic stress [49,50], and 5—dUTP incorpora-tion into the genome of viruses may be a common mechanism that isused by infected cells to kill viruses or to undergo suicide [46]. Notsurprisingly, many viruses encode their own dUTPase as a safeguardagainst this cellular defense [46].

The biochemical pathways involved in the de novo generation ofthe dNTPs required for DNA synthesis have received a great deal ofattention [35]. The synthesis of dTTP is unique since Thymine is theonly nucleotide that is present exclusively in DNA while uracil is theequivalent nucleotide that is present exclusively in RNA. The synthesisof dTMP is thus unique and convoluted by the fact that TMP is notused as a precursor. Instead dUTPmust first be synthesized, convertedto dUMP which is then methylated by thymidylate synthase to makedTMP. Subsequent phosphorylation then leads to the production ofthe DNA synthesis precursor dTTP. Thus the hydrolysis of dUTP bydUTPase has two functions, one is to produce the dUMP substrate forthymidylate synthase and the other is to rid the cell of excess dUTP[29,46]. With the exception of the few cases described above, thetraditional view held that excess dUTP is simply a necessary aberrantgenotoxic intermediate that is required for dTTP synthesis. A role for

Fig. 6. Amodel for the role of dUTP and dUTPase in the regulation of cell death. A varietyof stresses (here, activated Bax, chemicals that serve as Reactive Oxygen Species (ROS)donors such as hydrogen peroxide (H2O2) as well as other environmental stresses suchas the heavy metal cadmium (Cd2+)) are able to induce apoptosis, at least in part, byeliciting an increase in the production of ROS. The mechanism by which ROS inducesapoptosis is not completely clear but likely involves ROS mediated damage to a varietyof cellular components including proteins, lipids and DNA. The importance of ROS in theprocess is made clear by the observation that proteins as well as chemicals that canscavenge ROS have potent anti-apoptotic properties that can prevent cell death inresponse to a large variety of cellular stresses [14]. In analogous fashion, we proposethat stress increases the levels of dUTP, which presumably causes cell death only individing cells due to the fact that it must get incorporated into DNA, also serves topromote apoptosis. The ability of overexpressed dUTPase to protect from the samemultiple stresses that are known to kill cells by increased ROS levels suggests that thereis a cross-talk between the processes involved in regulating the levels of ROS and dUTP.A more detailed discussion regarding the possible cross-talk between the different pro-apoptotic second messengers has been recently presented [14].

dUTP as a pro-apoptotic second messenger suggests that the cell haschosen to make dUTP not as an aberration, but instead dUTP has theimportant function of serving as a nexus point that can be used totrigger apoptosis. The importance of anti-apoptotic genes in control-ling cell fate is highlighted by the fact that many viruses encode cellsurvival genes within their genomes. The expression of these survivalgenes allows the virus to prevent cellular suicide and thus facilitatestheir highjacking of cells [51]. As mentioned above, the presence ofgenes encoding dUTPase within the genomes of many virusessuggests that this gene can play a pro-survival function [46].

The use of several pro-apoptotic second messengers includingROS, ceramide and calcium suggests that cells have many differentweapons that can serve to trigger cell death. The myriad of processesinvolved in triggering cell death may be a mechanism of ensuring thatapoptosis can be triggered when required. This is in keeping with theobservation that MEF cells which are unable to carry out apoptosisbecause of the knock out of Bax and Bak genes, will utilize alternativepathways to undergo cell death in response to apoptotic inducingstresses [52]. The interrelationship between the known pro-apoptoticsecondmessengerswith themselves andwith Bax is complex [53–56].For example, the heterologous expression of activated Bax in yeastleads to cell death that can be inhibited by the co-expression of Baxinhibitory protein Bcl-2 [57], the ceramide utilizing enzyme SMS1[22], free radical scavengers [58] as well as the expression of DUT-M(this study) (Fig. 6). This suggests that stresses can activate many pro-apoptotic pathways, yet the inhibition of a single such pathway issufficient to prevent cell death. Although many questions regardingdUTP and other pro-apoptotic messengers remain to be answered,these later studies clearly indicate that the use of anti-apoptoticsequences can serve as tools to help decipher the mechanismsregulating different forms of PCD.

Acknowledgements

This work was supported by a grant from the Heart and StrokeFoundation of Canada (Quebec) to CM and MTG and from the NaturalSciences and Engineering Research Council of Canada (NSERC) toMTGand from the Canadian Institutes of Health Research (CIHR) to CM.DW was supported by a post-graduate fellowship from NSERC whileNM, JB and AL were the recipients of USRA fellowships from NSERC.

References

[1] R.S. Hotchkiss, A. Strasser, J.E. McDunn, P.E. Swanson, Cell death, N. Engl. J. Med.361 (2009) 1570–1583.

[2] L. Duprez, E. Wirawan, T. Vanden Berghe, P. Vandenabeele, Major cell deathpathways at a glance, Microbes Infect. 11 (2009) 1050–1062.

[3] S. Fulda, A.M. Gorman, O. Hori, A. Samali, Cellular stress responses: cell survivaland cell death, Int. J. Cell. Biol. 2010 (2010) 214074.

[4] C. Wang, R.J. Youle, The role of mitochondria in apoptosis*, Annu. Rev. Genet. 43(2009) 95–118.

[5] J.E. Chipuk, D.R. Green, How do BCL-2 proteins induce mitochondrial outermembrane permeabilization? Trends Cell Biol. 18 (2008) 157–164.

[6] J. Krucken, M. Epe, W.P. Benten, N. Falkenroth, F. Wunderlich, Malaria-suppressible expression of the anti-apoptotic triple GTPase mGIMAP8, J. Cell.Biochem. 96 (2005) 339–345.

[7] L. Tutar, Y. Tutar, Heat shock proteins; an overview, Curr. Pharm. Biotechnol. 11(2010) 216–222.

[8] K. Zheng, J.W. Pan, L. Ye, Y. Fu, H.Z. Peng, B.Y. Wan, Q. Gu, H.W. Bian, N. Han, J.H.Wang, B. Kang, J.H. Pan, H.H. Shao, W.Z. Wang, M.Y. Zhu, Programmed cell death-involved aluminum toxicity in yeast alleviated by antiapoptotic members withdecreased calcium signals, Plant Physiol. 143 (2007) 38–49.

[9] L. Galluzzi, N. Joza, E. Tasdemir, M.C. Maiuri, M. Hengartner, J.M. Abrams, N.Tavernarakis, J. Penninger, F. Madeo, G. Kroemer, No death without life: vitalfunctions of apoptotic effectors, Cell Death Differ. 15 (2008) 1113–1123.

[10] P. Balakumar, A. Rohilla, M. Singh, Pre-conditioning and postconditioning to limitischemia-reperfusion-induced myocardial injury: what could be the next footstep?Pharmacol. Res. 57 (2008) 403–412.

[11] S. Fulda, Tumor resistance to apoptosis, Int. J. Cancer 124 (2009) 511–515.[12] S. Fulda, Evasion of apoptosis as a cellular stress response in cancer, Int. J. Cell. Biol.

2010 (2010) 370835.

321D. Williams et al. / Biochimica et Biophysica Acta 1813 (2011) 315–321

[13] S.J. Zhang, M.N. Steijaert, D. Lau, G. Schutz, C. Delucinge-Vivier, P. Descombes, H.Bading, Decoding NMDA receptor signaling: identification of genomic programsspecifying neuronal survival and death, Neuron 53 (2007) 549–562.

[14] L. Portt, G. Norman, C. Clapp, M. Greenwood, M.T. Greenwood, Anti-Apoptosis andCell Survival: a review, Biochim. Biophys. Acta 1813 (2011) 238–259.

[15] C.M. Khoury, M.T. Greenwood, The pleiotropic effects of heterologous Baxexpression in yeast, Biochim. Biophys. Acta 1783 (2008) 1449–1465.

[16] S.Y. Eun, I.S. Woo, H.S. Jang, H. Jin, M.Y. Kim, H.J. Kim, J.H. Lee, K.C. Chang, J.H. Kim,H.G. Seo, Identification of cytochrome c oxidase subunit 6A1 as a suppressor ofBax-induced cell death by yeast-based functional screening, Biochem. Biophys.Res. Commun. 373 (2008) 58–63.

[17] I.S. Woo, H.S. Jang, S.Y. Eun, H.J. Kim, S.A. Ham, J.H. Lee, K.C. Chang, J.H. Kim, C.W.Han, H.G. Seo, Ran suppresses paclitaxel-induced apoptosis in human glioblas-toma cells, Apoptosis 13 (2008) 1223–1231.

[18] M.T. Greenwood, P. Ludovico, Expressing and functional analysis of mammalianapoptotic regulators in yeast, Cell Death Differ. 17 (2010) 737–745.

[19] T. Eisenberg, D. Carmona-Gutierrez, S. Buttner, N. Tavernarakis, F. Madeo, Necrosisin yeast, Apoptosis 15 (2010) 257–268.

[20] V. Khurana, S. Lindquist, Modelling neurodegeneration in Saccharomycescerevisiae: why cook with baker's yeast? Nat. Rev. Neurosci. 11 (2010) 436–449.

[21] D. Carmona-Gutierrez, C. Ruckenstuhl, M.A. Bauer, T. Eisenberg, S. Büttner, F.Madeo, Cell death in yeast: growing applications of a dying buddy, Cell DeathDiffer. 17 (2010) 733–734.

[22] Z. Yang, C. Khoury, G. Jean-Baptiste, M.T. Greenwood, Identification of mousesphingomyelin synthase 1 as a suppressor of Bax-mediated cell death in yeast,FEMS Yeast Res. 6 (2006) 751–762.

[23] G. Jean-Baptiste, Z. Yang, C. Khoury, M.T. Greenwood, Lysophosphatidic acidmediates pleiotropic responses in skeletal muscle cells, Biochem. Biophys. Res.Commun. 335 (2005) 1155–1162.

[24] M.G. Kaplitt, H. Makimura, Defective viral vectors as agents for gene transfer in thenervous system, J. Neurosci. Methods 71 (1997) 125–132.

[25] G. Jean-Baptiste, X. Li, Z. Yang, J. Heubach, S. Gaudio, C. Khoury, U. Ravens, M.T.Greenwood, Beta adrenergic receptor-mediated atrial specific up-regulation ofRGS5, Life Sci. 76 (2005) 1533–1545.

[26] G. Ring, C.M. Khoury, A.J. Solar, Z. Yang, C.A. Mandato, M.T. Greenwood, Trans-membrane protein 85 from both human (TMEM85) and yeast (YGL231c) inhibithydrogen peroxide mediated cell death in yeast, FEBS Lett. 582 (2008) 2637–2642.

[27] C.M.Khoury, Z. Yang,X.Y. Li,M.Vignali, S. Fields,M.T. Greenwood, A TSC22-likemotifdefines a novel antiapoptotic protein family, FEMS Yeast Res. 8 (2008) 540–563.

[28] R.D. Ladner, S.J. Caradonna, The human dUTPase gene encodes both nuclear andmitochondrial isoforms. Differential expression of the isoforms and characteriza-tion of a cDNA encoding the mitochondrial species, J. Biol. Chem. 272 (1997)19072–19080.

[29] B.G. Vertessy, J. Toth, Keeping uracil out of DNA: physiological role, structure andcatalytic mechanism of dUTPases, Acc. Chem. Res. 42 (2009) 97–106.

[30] K. Dornfeld, M. Johnson, AP endonuclease deficiency results in extreme sensitivityto thymidine deprivation, Nucleic Acids Res. 33 (2005) 6644–6653.

[31] B.A. Tinkelenberg, M.J. Hansbury, R.D. Ladner, dUTPase and uracil-DNA glycosy-lase are central modulators of antifolate toxicity in Saccharomyces cerevisiae,Cancer Res. 62 (2002) 4909–4915.

[32] P.M. Wilson, W. Fazzone, M.J. LaBonte, J. Deng, N. Neamati, R.D. Ladner, Novelopportunities for thymidylate metabolism as a therapeutic target, Mol. CancerTher. 7 (2008) 3029–3037.

[33] G. Lurje, P.C. Manegold, Y. Ning, A. Pohl, W. Zhang, H.J. Lenz, Thymidylate synthasegene variations: predictive and prognostic markers, Mol. Cancer Ther. 8 (2009)1000–1007.

[34] L. Seiple, P. Jaruga, M. Dizdaroglu, J.T. Stivers, Linking uracil base excision repairand 5-fluorouracil toxicity in yeast, Nucleic Acids Res. 34 (2006) 140–151.

[35] C.K. Mathews, DNA precursor pools and genomic stability, FASEB J. 20 (2006)1300–1314.

[36] P. Fabrizio, V.D. Longo, Chronological aging-induced apoptosis in yeast, Biochim.Biophys. Acta 1783 (2008) 1280–1285.

[37] F. Madeo, E. Frohlich, M. Ligr, M. Grey, S.J. Sigrist, D.H. Wolf, K.U. Frohlich, Oxygenstress: a regulator of apoptosis in yeast, J. Cell Biol. 145 (1999) 757–767.

[38] A.M. Nargund, S.V. Avery, J.E. Houghton, Cadmium induces a heterogeneous andcaspase-dependent apoptotic response in Saccharomyces cerevisiae, Apoptosis 13(2008) 811–821.

[39] B. Jiang, W. Xiao, Y. Shi, M. Liu, X. Xiao, Role of Smac/DIABLO in hydrogenperoxide-induced apoptosis in C2C12 myogenic cells, Free Radic. Biol. Med. 39(2005) 658–667.

[40] C. Zhao, E. Wang, Heat shock protein 90 suppresses tumor necrosis factor alphainduced apoptosis by preventing the cleavage of Bid in NIH3T3 fibroblasts, Cell.Signal. 16 (2004) 313–321.

[41] A. Busca, M. Saxena, M. Kryworuchko, A. Kumar, Anti-apoptotic genes in thesurvival of monocytic cells during infection, Curr. Genomics 10 (2009) 306–317.

[42] S. Fulda, L. Galluzzi, G. Kroemer, Targeting mitochondria for cancer therapy, Nat.Rev. Drug Discov. 9 (2010) 447–464.

[43] D.L. Vaux, S. Cory, J.M. Adams, Bcl-2 gene promotes haemopoietic cell survival andcooperates with c-myc to immortalize pre-B cells, Nature 335 (1988) 440–442.

[44] M.L. Circu, T.Y. Aw, Reactive oxygen species, cellular redox systems, and apoptosis,Free Radic. Biol. Med. 48 (2010) 749–762.

[45] J.C.M. Holthuis, C. Luberto, Tales and mysteries of the enigmatic sphingomyelinsynthase family, in: C.D.P. Chalfant, Maurizio (Eds.), Advances in ExperimentalMedicine and Biology, vol. 668, 2010.

[46] J. Sire, G. Querat, C. Esnault, S. Priet, Uracil within DNA: an actor of antiviralimmunity, Retrovirology 5 (2008) 45.

[47] M.H. Gadsden, E.M. McIntosh, J.C. Game, P.J. Wilson, R.H. Haynes, dUTPpyrophosphatase is an essential enzyme in Saccharomyces cerevisiae, EMBO J. 12(1993) 4425–4431.

[48] M. Guillet, P.A. Van Der Kemp, S. Boiteux, dUTPase activity is critical to maintaingenetic stability in Saccharomyces cerevisiae, Nucleic Acids Res. 34 (2006)2056–2066.

[49] B. Dudley, A. Hammond, W.A. Deutsch, The presence of uracil-DNA glycosylase ininsects is dependent upon developmental complexity, J. Biol. Chem. 267 (1992)11964–11967.

[50] S. Kajimoto, M. Horie, H. Manabe, Y. Masuda, T. Shibayama-Imazu, S. Nakajo, X.F.Gong, T. Obama, H. Itabe, K. Nakaya, A tyrosine kinase inhibitor, beta-hydroxyisovalerylshikonin, induced apoptosis in human lung cancer DMS114cells through reduction of dUTP nucleotidohydrolase activity, Biochim. Biophys.Acta 1782 (2008) 41–50.

[51] S.M. Best, Viral subversion of apoptotic enzymes: escape from death row, Annu.Rev. Microbiol. 62 (2008) 171–192.

[52] S. Shimizu, T. Kanaseki, N. Mizushima, T. Mizuta, S. Arakawa-Kobayashi, C.B.Thompson, Y. Tsujimoto, Role of Bcl-2 family proteins in a non-apoptoticprogrammed cell death dependent on autophagy genes, Nat. Cell Biol. 6 (2004)1221–1228.

[53] V. Ganesan, M. Colombini, Regulation of ceramide channels by Bcl-2 familyproteins, FEBS Lett. 584 (2010) 2128–2134.

[54] L.J. Siskind, T.D. Mullen, K. Romero Rosales, C.J. Clarke, M.J. Hernandez-Corbacho, A.L. Edinger, L.M. Obeid, The BCL-2 protein BAK is required forlong-chain ceramide generation during apoptosis, J. Biol. Chem. 285 (2010)11818–11826.

[55] D.A. Kubli, M.N. Quinsay, C. Huang, Y. Lee, A.B. Gustafsson, Bnip3 functions as amitochondrial sensor of oxidative stress during myocardial ischemia andreperfusion, Am. J. Physiol. Heart Circ. Physiol. 295 (2008) H2025–2031.

[56] T. Goldkorn, N. Balaban, M. Shannon, V. Chea, K. Matsukuma, D. Gilchrist, H.Wang,C. Chan, H2O2 acts on cellular membranes to generate ceramide signaling andinitiate apoptosis in tracheobronchial epithelial cells, J. Cell Sci. 111 (Pt 21) (1998)3209–3220.

[57] W. Tao, C. Kurschner, J.I. Morgan, Modulation of cell death in yeast by the Bcl-2family of proteins, J. Biol. Chem. 272 (1997) 15547–15552.

[58] S. Chen, Z. Vaghchhipawala, W. Li, H. Asard, M.B. Dickman, Tomato phospholipidhydroperoxide glutathione peroxidase inhibits cell death induced by Bax andoxidative stresses in yeast and plants, Plant Physiol. 135 (2004) 1630–1641.