3171Research Article

IntroductionApoptosis is a cell-suicide process with an important role in awide variety of mammalian physiological processes, but causesdisease when inappropriately controlled. When its crucial rolein the prevention and promotion of different diseases wasrecognised, apoptosis gained much attention as a potentialtherapeutic target in the treatment of pathologies, such ascancer and neurodegenerative disorders (Meier et al., 2000).Recently, it has emerged that the apoptosis modulation-associated proteins are regulated through phosphorylation bydifferent protein kinases, such as protein kinase C (PKC)(Gutcher et al., 2003). PKC is a family of serine/threoninekinases with at least ten isoforms grouped into threesubfamilies based on their structure and cofactors required foractivation: the conventional or classical (�, �I, �II and �), thenovel (�, �, � and �) and the atypical ( and /�) isoforms(Spitaler and Cantrell, 2004). Regulated by distinct means andpresenting different tissue distribution patterns, PKC isoformshave deserved particular attention owing to their involvementin the regulation of prominent apoptotic members, such asproteins of the Bcl-2 family (Deng et al., 2000; Kornblau et al.,2000; Musashi et al., 2000). The Bcl-2 family, consisting ofanti-apoptotic (such as Bcl-2 and Bcl-xL) and pro-apoptotic(such as Bax and Bad) members, function at a point in the celldeath pathway that dictates whether or not cells are committedto die (Reed, 1997; Antonsson and Martinou, 2000). Therefore,owing to both inhibitory and stimulatory effects on Bcl-2proteins, and consequently on apoptosis (Gutcher et al., 2003),PKC isoforms have been recognised as potential targets for

therapeutic modulation (Deng et al., 2000; Kornblau et al.,2000; Gutcher et al., 2003).

Among the several isoforms, PKC-�, -�, -� and -, expressedin the majority of tissues, have been considered as key playersin apoptosis (Musashi et al., 2000; Gutcher et al., 2003). Ingeneral, classical and atypical PKCs appear to be associatedwith cell survival, whereas novel PKCs are associated withapoptosis stimulation (Gutcher et al., 2003). However, thefunction carried out by individual PKC isoforms in apoptosisremains to be clarified (Gutcher et al., 2003; Hofmann, 2004).Such information could lead to the identification of PKCisoforms that might be targeted to obtain a selective modulationof apoptosis (Gutcher et al., 2003). Yet, for that purposeindependent analysis of the role of each isoform must becarried out (Musashi et al., 2000; Hofmann, 2001; Gutcher etal., 2003), a goal that is difficult to achieve with mammaliancells, as a result of the co-existence of several PKC isoformsin the same cell.

In the present work, we propose a new methodologicalapproach to address the issues raised above. We havepreviously shown that expression of individual PKC isoformsin Saccharomyces cerevisiae allows the study of themodulatory activity of several compounds on individualisoforms (Saraiva et al., 2003; Saraiva et al., 2004).Furthermore, several authors have demonstrated that yeastundergoes apoptotic cell death similar to that seen inmammalian cells (reviewed by Madeo et al., 2002a; Madeo etal., 2004; Ludovico et al., 2005).

The present study reveals that PKC-�, -�, -� or - stimulates

Mammalian protein kinase C (PKC) isoforms have beensubject of particular attention because of their ability tomodulate apoptotic proteins. However, the roles played byeach PKC isoform in apoptosis are still unclear. Here,expression of individual mammalian PKC isoforms inSaccharomyces cerevisiae is used as a new approach tostudy the role of each isoform in apoptosis. The fourisoforms tested, excepting PKC-��, stimulate S. cerevisiaeacetic-acid-induced apoptosis essentially through amitochondrial ROS-dependent pathway. However, their co-expression with Bcl-xL reveals a PKC-isoform-dependent

modulation of Bcl-xL anti-apoptotic activity. A yeastpathway homologue to the mammalian SAPK/JNK isresponsible for acetic-acid-induced Bcl-xL phosphorylationthat is differently modulated by PKC isoforms. The dataobtained suggest conservation of an ancient mechanism ofapoptosis regulation in yeast and mammals and offer newinsights into mammalian apoptosis modulation by PKCisoforms.

Key words: Apoptosis regulation, Bcl-xL, Mammalian PKCisoforms, Yeast

Summary

Specific modulation of apoptosis and Bcl-xLphosphorylation in yeast by distinct mammalianprotein kinase C isoformsLucília Saraiva1, Rui D. Silva2, Gil Pereira1, Jorge Gonçalves3 and Manuela Côrte-Real2,*1Laboratório de Microbiologia and 3Laboratório de Farmacologia, Centro de Estudos de Química Orgânica, Fitoquímica e Farmacologia daUniversidade do Porto (CEQOFFUP), Faculdade de Farmácia, Universidade do Porto, Rua Aníbal Cunha 164, 4050-047 Porto, Portugal2Centro de Biologia, Universidade do Minho, Campus de Gualtar, 4710-057 Braga, Portugal*Author for correspondence (e-mail: mcortereal@bio.uminho.pt)

Accepted 3 May 2006Journal of Cell Science 119, 3171-3181 Published by The Company of Biologists 2006doi:10.1242/jcs.03033

Jour

nal o

f Cel

l Sci

ence

3172

S. cerevisiae acetic-acid-induced apoptosis. Moreover, it wasshown that PKC isoforms modulate the Bcl-xL anti-apoptoticactivity differently through interference with thephosphorylated or dephosphorylated forms of Bcl-xL. Theseresults support the interpretation that control of mammalianapoptosis by PKC-�, -�, -� and - can be attributed to theirdistinct modulation of Bcl-xL and possibly of other apoptoticproteins.

ResultsEnhancement of S. cerevisiae acetic-acid-inducedapoptosis by expression of mammalian PKC-�, -�, -�or -Previous studies showed that expression of mammalian PKC-�, -�, -� or - in yeast does not induce cell death (Saraiva etal., 2003; Saraiva et al., 2004). Therefore, to study whetherthese isoforms interfere with the apoptotic pathway in S.cerevisiae, acetic acid was selected as an inducer of apoptosis(Ludovico et al., 2001; Ludovico et al., 2002).

We found that expression of PKC-�, -�, -� or - isoformscaused a marked increase in the percentage of dead cellscompared with yeast transformed with the empty vector (Fig.1A). This increase in cell death was accompanied by anincrease in the number of TUNEL-positive cells (Fig. 1B),but not by an increase of the number of PI-positive cells (Fig.1C). This indicates that the additional cell death resultingfrom expression of PKC isoforms is also apoptotic. Indeed,

Journal of Cell Science 119 (15)

DNA degradation with preservation of plasma membraneintegrity is a characteristic marker of apoptotic death inmammalian (Vermes et al., 2000) and in yeast cells (Madeoet al., 1997).

The effect of PKC-�, -�, -� or - expression onmitochondrial reactive oxygen species (ROS) production inacetic-acid-induced apoptosis was also studied using themitochondria-selective probe MitoTracker Red CM-H2XRos.Fig. 2 shows the flow cytometric analysis of a representativeexperiment with cells treated in the absence (0 mM) and with300 mM acetic acid. A very low ROS production was detectedin yeast transformed with the empty vector (YEplac181 orYEp51 for PKC-�) after treatment with 300 mM acetic acid(Fig. 2, compare dark-grey fill at 0 and 300 mM). Expressionof each of the four PKC isoforms tested had very little effecton ROS production in cells not exposed to acetic acid,but resulted in the appearance of a second subpopulationdisplaying higher mean fluorescence intensity, indicative ofincreased ROS production in cells treated with 300 mM aceticacid (Fig. 2, compare grey line at 0 mM with 300 mM). Thiseffect was particularly marked for PKC-� and -� and verysmall for PKC-�. We confirmed, by epifluorescence analysis,that the subpopulation with increased mean fluorescenceintensity displayed a bright red fluorescence localised inmitochondria (not shown).

The possibility that PKC-�, -�, -� or - expression interfereswith the activity of the yeast caspase-related protease (Yca1p)

Fig. 1. Expression of mammalian PKC-�, -�, -� or - in S. cerevisiae stimulates acetic-acid-induced apoptosis. Yeast transformed with theempty vector, YEplac181 or YEp51 (black) or yeast expressing PKC-�, -�, -� or - (grey) were incubated with acetic acid (300, 350 or 400mM) for 1 hour at 30°C. (A) Cell death experiments. Percentage of dead cells was determined by c.f.u. counts, considering as 0% dead cells(100% survival) the number of c.f.u. obtained after 1 hour incubation with no acetic acid. Data are the mean ± s.e.m. of three independentexperiments with six replicates. (B) TUNEL staining and (C) Propidium iodide (PI) staining. Cells treated in the absence of acetic acid (0 mM)were used as negative control. Data are the mean ± s.e.m. of two independent experiments; means correspond to counts of at least 600 cells persample. Significant differences from those obtained with the empty vector are indicated by *P<0.05; **P<0.001 (unpaired Student’s t-test).

Jour

nal o

f Cel

l Sci

ence

3173Apoptosis regulation by mammalian PKC isoforms

during acetic-acid-induced apoptosis was also tested. Since invivo yeast labelling with the caspase inhibitor FITC-VAD-FMK has been attributed to caspase activation (Madeo et al.,2002b; Herker et al., 2004), this activation was initiallymeasured by flow cytometry of FITC-VAD-FMK stained cells.

For yeast transformed with the empty vector (YEplac181 orYEp51 for PKC-�), only about 20-30% of FITC-positive cellswere found for the maximum acetic acid concentration tested(400 mM; see Fig. 3A). For this acetic acid concentration,expression of PKC isoforms did not increase the percentage ofFITC-positive cells found with yeast transformed with theempty vector. The only exception occurred with cellsexpressing PKC-� where a small increase in this percentagewas observed (see Fig. 3A). Since caspase activation was onlydetected for the highest acetic acid concentration tested, andbecause a recent work reports that FITC-VAD-FMK binds non-specifically to PI-positive cells, claiming that staining with thisfluorogenic caspase substrate is subject to artefacts (Wysockiand Kron, 2004), additional tests to evaluate the possibility ofnonspecific staining, were carried out. Thus, we investigatedthe effect of the nonfluorescent caspase inhibitor Z-VAD-FMKon the viability of acetic-acid-treated yeast cells expressingPKC-�, for which the highest percentage of FITC-positive cellshad been obtained. Although significant, Z-VAD-FMK onlycaused a decrease of ~16% in the percentage of dead cells (Fig.3B). A similar value could be obtained by subtracting thepercentage of PI-positive cells from the percentage of FITC-positive cells expressing PKC-� after treatment with 400 mMacetic acid. This suggests that the real percentage of cells withcaspase activation only corresponds to ~16%. We also carriedout a cell-free fluorimetric assay to assess caspase activity inyeast expressing PKC-�, after treatment with 400 mM aceticacid. No caspase activation was detected for the threefluorogenic caspase substrates used (Fig. 3C). Non-detectionof caspase activation by the in vitro fluorimetric assay isconsistent with the low percentage of caspase activationdetected by the other assays performed. Together these resultssuggest that stimulation of acetic-acid-induced apoptosis byPKC isoform is not associated with an enhancement of Yca1pactivity.

Mammalian PKC-�, -�, -� or - differently modulate Bcl-xL anti-apoptotic effects in S. cerevisiae acetic-acid-induced apoptosisFirst, acetic-acid-induced cell death in yeast expressing Bcl-xLand in yeast transformed with the two empty vectors (pOW4and YEplac 181 or YEp51) was compared. For all acetic acidconcentrations tested, Bcl-xL expression caused a significantreduction in the acid-induced yeast cell death compared withyeast transformed with the empty vectors (Fig. 4A). Thisreduction was accompanied by a decrease in the percentage ofTUNEL-positive cells, particularly for 350 mM acetic acid(Fig. 4B) and by a marked decrease in ROS production (Fig.5, compare grey lines with dark-grey fill at 300 mM acetic-acid-treated cells).

In order to investigate a putative influence of PKC-�, -�, -�or - on the observed Bcl-xL anti-apoptotic effect, Bcl-xL andindividual mammalian PKC isoforms were co-expressed in S.cerevisiae. Afterwards, the effect of co-expression on cellsurvival, evaluated by c.f.u. counts (Fig. 4A), DNA degradationby TUNEL staining (Fig. 4B) and mitochondrial ROSproduction by MitoTracker Red CM-H2XRros staining (Fig. 5)was assessed. The results showed that PKC-�, -�, -� and -interfered with the Bcl-xL anti-apoptotic effect, but in differentways. The protective effect of Bcl-xL in acetic-acid-inducedapoptotic death was completely abolished when PKC-� was

Fig. 2. Stimulation of S. cerevisiae acetic-acid-induced apoptosis byexpression of mammalian PKC-�, -� or - but not PKC-� isoform isassociated with significant ROS production. ROS production bymitochondria was detected by flow cytometry, using MitoTrackerRed CM-H2XRos. Overlays of red fluorescence histograms wereobtained with yeast treated with 0 and 300 mM acetic acid: black fill,yeast transformed with the empty vector (YEplac181 or YEp51);light-grey line, yeast expressing PKC-�, -�, -� or -. Cells treated inthe absence of acetic acid (0 mM) were used as a negative control.Data represent one of two independent experiments.

Jour

nal o

f Cel

l Sci

ence

3174

co-expressed with Bcl-xL (Figs 4, 5). On the other hand, theeffects obtained with co-expression of PKC-� and Bcl-xL onacetic-acid-induced yeast apoptotic death were close to thoseobtained with yeast co-transformed with the two emptyvectors, suggesting that co-expression of PKC-� and Bcl-xLneutralises the effect caused by each protein when individuallyexpressed. On the contrary, co-expression of PKC-� or - withBcl-xL caused a marked enhancement of the Bcl-xL anti-apoptotic effects (Figs 4, 5). In the cell death assays, apronounced reduction in acetic-acid-induced yeast cell deathwas obtained for co-expression of both isoforms. Values closeto those obtained with 0 mM acetic acid were found with 250and 300 mM acetic acid for PKC-�, or with 300 and 350 mMacetic acid for PKC- (Fig. 4A). Surprisingly, a better survival

than that obtained with untreated control, was observed with250 mM acetic acid for PKC- (Fig. 4A). The percentages ofapoptotic cells assessed by TUNEL assay, obtained with 300and 350 mM of acetic acid were also close to that obtainedwith 0 mM acetic acid, especially for PKC-� (Fig. 4B). Co-expression of PKC-� and Bcl-xL practically abolished ROSproduction induced by 300 mM acetic acid (Fig. 5, third row;compare black lines at 0 and 300 mM). With PKC- and Bcl-xL co-expression, ROS production was also lower than thatobtained with cells only expressing Bcl-xL (Fig. 5, fourth row;compare black line with grey line at 300 mM acetic-acid-treated cells). Together these data show a PKC-isoform-dependent modulation of Bcl-xL anti-apoptotic effects in S.cerevisiae acetic-acid-induced apoptosis.

Journal of Cell Science 119 (15)

Fig. 3. Stimulation of S. cerevisiae acetic-acid-induced apoptosis by expression of mammalian PKC-�, -�, -� or - is not mediated by yeastcaspase activation. (A) Flow cytometric analysis. Percentage of cells with active caspases obtained with yeast transformed with the emptyvector (YEplac181 or YEp51) and with yeast expressing PKC-�, -�, -� or -, after 1 hour treatment with 400 mM acetic acid. Treated cells werelabelled with 50 �M FITC-VAD-FMK for 25 minutes at 30°C and analysed by flow cytometry. Cells treated in the absence of acetic acid (0mM) were used as negative control. Data represent one of two independent experiments. (B) Effect of the caspase inhibitor Z-VAD-FMK on thesurvival of yeast expressing PKC-� treated with 400 mM acetic acid. Before treatment with acetic acid, cells were pre-treated without (control)or with 100 �M Z-VAD-FMK for 2 hours at 30°C. The percentage of dead cells was determined by c.f.u. counts, as in Fig. 1. Data are the mean± s.e.m. of two independent experiments with six replicates. Significant differences from those obtained with the empty vector are indicated by*P<0.05 (unpaired Student’s t-test). (C) Fluorimetric analysis. Cell extracts obtained from yeast expressing PKC-� and treated with 0 mM(negative control) or 400 mM acetic acid, were incubated with 50 �M of the fluorogenic caspase substrates Ac-DEVD-AMC, Ac-IETD-AMCor Ac-VEID-AMC. Recombinant caspase 3 (for Ac-DEVD-AMC), caspase 6 (for Ac-VEID-AMC) and caspase 8 (for Ac-IETD-AMC) wereused as positive control. Aminomethylcoumarin (AMC) release was monitored for 90 minutes at 30°C in a spectrofluorimeter. Caspase activityis expressed in arbitrary fluorescence units (FU)/minute. Data are the mean ± s.e.m. of two independent experiments.

Jour

nal o

f Cel

l Sci

ence

3175Apoptosis regulation by mammalian PKC isoforms

Mammalian PKC-�, -�, -� or - co-expression in S.cerevisiae differently interferes with acetic-acid-inducedBcl-xL phosphorylationTreatment of yeast expressing Bcl-xL with 300 mM acetic acidcaused the appearance of a second, slow-migrating, Bcl-xLband at ~32 kDa. Furthermore, the intensity of the two Bcl-xLbands was dependent on the co-expressed PKC isoform (Fig.6A). Co-expression with PKC-� strongly reduced the intensityof the fast-migrating Bcl-xL band, whereas co-expression withPKC-� or - strongly reduced the intensity of the slow-migrating band. Co-expression with PKC-� only caused aslight decrease in the intensity of the fast-migrating Bcl-xLband.

To confirm whether the slow-migrating band correspondedto a phosphorylated form of Bcl-xL (Basu and Haldar, 2003;Du et al., 2005), a dephosphorylation reaction using the -protein phosphatase (-PPase) was carried out. As shown inFig. 6B, the -PPase treatment of protein extracts obtainedfrom yeast expressing Bcl-xL and exposed to 300 mM aceticacid completely abolished the 32 kDa band but did not affectthe 29 kDa band. This confirms that the slow-migrating Bcl-xL band corresponds to a phosphorylated form of Bcl-xL (P-Bcl-xL).

JNK inhibitor II reduces acetic-acid-induced cell deathand abolishes Bcl-xL phosphorylation in yeastexpressing Bcl-xL, but not in yeast co-expressing PKC-�and Bcl-xLRecently, the role of c-Jun N-terminal kinase (SAPK/JNK)signalling pathway on taxol- or 2-methoxyestradiol-triggered

Bcl-xL phosphorylation was demonstrated using the potentcell-permeable JNK inhibitor II, which eliminated Bcl-xLphosphorylation and reduced concomitant cell apoptosis (Basuand Haldar, 2003).

In order to assess whether a yeast pathway homologue to themammalian SAPK/JNK pathway was involved in Bcl-xLphosphorylation induced by acetic acid, the JNK inhibitor IIwas used before exposure to acetic acid. Pre-treatment of yeastcells expressing Bcl-xL with 20 �M of JNK inhibitor II for 8hours rendered the cells more resistant to acetic-acid-inducedcell death and led to the elimination of the Bcl-xLphosphorylated form (Fig. 7, YEplac181 + Bcl-xL and YEp51+ Bcl-xL). By contrast, for yeast co-expressing PKC-� andBcl-xL, pre-treatment of cells with 20 �M of JNK inhibitor IIfor 8 hours did not affect either acid-induced cell death nor theBcl-xL phosphorylated form (Fig. 7, PKC-� + Bcl-xL). Theresults support the existence of a yeast pathway homologue tothe mammalian JNK pathway responsible for acetic-acid-induced Bcl-xL phosphorylation. Furthermore, they suggestthat PKC-� is also involved in Bcl-xL phosphorylation by apathway independent of the yeast homologue to themammalian JNK pathway.

DiscussionThe aim of the present work was to validate a cell modelallowing the study of the role of individual mammalian PKCisoforms in apoptosis regulation. Though yeast has beenrecognised recently as an ‘unclean system’ to study severalfeatures of the complex mammalian apoptotic process, namelyproteins of the Bcl-2 family, they are still accepted as a

Fig. 4. Expression of Bcl-xL abrogates S. cerevisiae acetic-acid-induced apoptosis. Co-expression of mammalian PKC-�, -�, -� or - differentlymodulate Bcl-xL effects in S. cerevisiae acetic-acid-induced apoptosis. Yeast co-transformed with the empty vectors, pOW4 and YEp51 orpOW4 and YEplac181 (black), expressing Bcl-xL (pale grey) and co-expressing Bcl-xL and PKC-�, -�, -� or - (dark grey) were incubatedwith acetic acid for 1 hour at 30°C. (A) Cell death experiments. Percentage of dead cells was determined by c.f.u. counts as in Fig. 1. Data arethe mean ± s.e.m. of three independent experiments with six replicates. (B) TUNEL staining. Cells treated in the absence of acetic acid (0 mM)were used as negative control. Data are the mean ± s.e.m. of two independent experiments; means correspond to counts of at least 600 cells persample. Significantly differences from those obtained with yeast co-transformed with the empty vectors and with yeast only expressing Bcl-xLare indicated by *P<0.05 and #P<0.05 (unpaired Student’s t-test), respectively.

Jour

nal o

f Cel

l Sci

ence

3176

powerful model for apoptosis understanding (Ludovico et al.,2005). Furthermore, S. cerevisiae has also been used as anadequate model to study the modulatory activity of several

compounds on individual PKC isoforms (Saraiva et al., 2003;Saraiva et al., 2004). Although S. cerevisiae harbours a singlePKC enzyme structurally similar to mammalian PKC isoforms,Pkc1p, with a key role in signalling pathways, none of themammalian isoforms are functionally complementary to theyeast Pkc1p (Perez and Calonge, 2002). Additionally, it wasdemonstrated that Pkc1p does not respond to typicalmammalian PKC modulators (Shieh et al., 1996; Keenan et al.,1997; Saraiva et al., 2003; Saraiva et al., 2004). The resultspresented here support the view that S. cerevisiae is a valuablecell model to study the role of each mammalian PKC isoformin apoptosis.

Expression of mammalian PKC-�, -�, -� or - stimulatesS. cerevisiae acetic-acid-induced apoptosisHere it is shown that expression of PKC-�, -�, -� and - causesa pronounced stimulation of the acetic-acid-induced apoptoticcell death and of DNA fragmentation, with preservation ofplasma membrane integrity. Therefore, this additional celldeath observed when a PKC isoform is expressed is alsoapoptotic. The observed stimulation of mitochondrial ROSproduction in cells expressing individual PKC isoforms, aftertreatment with acetic acid, is consistent with the involvementof a mitochondrial-dependent pathway in S. cerevisiae acetic-acid-induced apoptosis (Ludovico et al., 2002). It should bestressed that, in the present study, a higher resistance to acetic-acid-induced apoptosis is detected in transformed yeast leadingto the use of acetic acid at concentrations higher than thoseused by Ludovico et al. (Ludovico et al., 2001; Ludovico et al.,2002). This greater resistance to apoptosis is particularlyevident in the ROS assay where 300 mM acetic acid does notcause ROS production in yeast transformed with the emptyvector. Madeo et al. (Madeo et al., 2002b) have alreadyreported a higher resistance to apoptosis for S. cerevisisaetransformed with YEp52.

Based on genetic evidence, Madeo et al. (Madeo et al.,2002b) reported that the yeast caspase-related protease(Yca1p) is an executor of acetic-acid-induced apoptosis in S.cerevisiae. However, only 20-30% of caspase-positive cellswere now detected by flow cytometry in the populationundergoing apoptosis induced by 400 mM acetic acid. Inaddition, the interference of expression of PKC isoforms in thisreduced Yca1p activation is rather restricted, an observationthat was confirmed by the small increase in survival with theinhibitor z-VAD-FMK. These results are consistent with anapoptotic pathway induced by acetic acid, in which the role ofyeast caspase seems less relevant compared with otherapoptotic stimuli like oxidative (Madeo et al., 2002b) orhyperosmotic (Silva et al., 2005) stresses. Therefore,stimulation of S. cerevisiae acetic-acid-induced apoptosis byPKC isoform expression seems to involve an Yca1p-independent pathway. Nevertheless, since it can be expectedthat the yeast genome contains other metacaspases besidesYca1p, which might not cleave the fluorogenic caspasesubstrates used in this study, we cannot preclude the hypothesisthat the observed effect of PKC isoform on S. cerevisiae acetic-acid-induced apoptosis is mediated through a caspase-dependent process.

Because expression of PKC-�, in contrast to the other PKCisoforms tested, does not cause a prominent stimulation ofmitochondrial ROS production, we proposed that different

Journal of Cell Science 119 (15)

Fig. 5. Expression of Bcl-xL reduces S. cerevisiae acetic-acid-induced ROS production. Co-expression of PKC-�, -�, -� or -differently modulates Bcl-xL effects on ROS production. ROSproduction by mitochondria was detected by flow cytometry, usingMitoTracker Red CM-H2XRos. Overlay of red fluorescencehistograms was obtained with yeast treated with 0 and 300 mMacetic acid. Dark grey fill, cells co-transformed with the emptyvectors (pOW4 and YEp51 or pOW4 and YEplac181); light greyline, cells expressing Bcl-xL; black line, cells co-expressing Bcl-xLand PKC-�, -�, -� or -. Cells treated in the absence of acetic acid (0mM) were used as negative control. Data represent one of twoindependent experiments.

Jour

nal o

f Cel

l Sci

ence

3177Apoptosis regulation by mammalian PKC isoforms

mechanisms might lie behind the observed stimulation ofacetic-acid-induced apoptosis. It is conceivable that under theconditions tested, PKC-�, -� or - expression triggers anoxidative DNA-damage-dependent pathway, whereas PKC-�

expression induces a preferential DNA-damage apoptoticpathway, which seems to be ROS independent. Actually, inmammalian cells there is increasing evidence that the keytargets for PKC-� are located in the nucleus, such as DNA-dependent protein kinase (DNA-PK) and nuclear lamin B(Bharti et al., 1998; Gutcher et al., 2003; Martelli et al., 2004).Ectopic expression of the catalytically active form of PKC-�failed to induce apoptosis in cells deficient in DNA-PK (Bhartiet al., 1998). In yeast, the DNA-PK components, Ku70 andKu80, are required for DNA-damage response and havefunctions similar to their mammalian counterparts (Burhans etal., 2003). Whether stimulation of yeast apoptosis by PKC-�is mediated by phosphorylation of the yeast orthologues ofDNA-PK and consequent enhancement of DNA damage is aquestion that deserves further study.

Co-expression of mammalian PKC-�, -�, -� or -modulates the anti-apototic activity of Bcl-xL inS. cerevisiaeTogether, the aforementioned results show that all the PKCisoforms tested cause a similar pro-apoptotic effect when

Fig. 6. Mammalian PKC-�, -�, -� or - co-expression in S. cerevisiae differentially interferes with acetic-acid-induced Bcl-xL phosphorylation.(A) Western blot analysis reveals two different Bcl-xL migrating bands in protein extracts obtained from yeast expressing Bcl-xL and treatedwith 300 mM acetic acid for 1 hour. The intensity of these two Bcl-xL bands is dependent on the PKC co-expressed isoform. Without aceticacid, only a single band of Bcl-xL at ~ 9 kDa can be observed (see Fig. 8B); the lane corresponding to protein extracts obtained from yeast co-expressing PKC- and Bcl-xL was located in a different part of the same gel. (B) In vitro phosphatase treatment of protein extracts obtainedfrom yeast expressing Bcl-xL and treated with 300 mM acetic acid for 1 hour. Yeast extract with 25 �g protein was incubated with 400 U -protein phosphatase (+-PPase), at 30°C for 30 minutes. Yeast extract without -PPase was used as a control. Western blot analysis showed thatthe slow-migrating Bcl-xL band (~32 kDa) was completely abolished by -PPase, indicating that this band corresponds to a phosphorylatedform of Bcl-xL (P-Bcl-xL). For western blot analysis, protein extracts were separated by 15% SDS-PAGE. For Bcl-xL detection an anti-Bcl-xLgoat polyclonal antibody, followed by horseradish-peroxidase-conjugated rabbit anti-goat IgG, were used. For �-actin detection, used as aloading control, membranes were stripped and then reprobed with an anti-actin rabbit antibody. Immunoblots were developed by enhancedchemiluminescence.

Fig. 7. JNK inhibitor II reduces acetic-acid-induced cell death andabolishes Bcl-xL phosphorylation in yeast expressing Bcl-xL, but notin yeast co-expressing PKC-� and Bcl-xL. Yeast cells expressingBcl-xL (YEplac181 + Bcl-xL; YEp51 + Bcl-xL) and co-expressingPKC-� and Bcl-xL (PKC-� + Bcl-xL) were pre-treated in thepresence of solvent (DMSO) or in the presence of 20 �M JNKinhibitor II for 8 hours followed by treatment with 0 and 300 mMacetic acid for 1 hour at 30°C. Percentage of dead cells wasdetermined by c.f.u. counts as in Fig. 1. Data are the mean ± s.e.m.of five to eight independent experiments with six replicates.Significant differences from those obtained with DMSO are indicatedby *P<0.05 (unpaired Student’s t-test). For western blot analysis,protein extracts were separated by 15% SDS-PAGE. For Bcl-xLdetection an anti-Bcl-xL goat polyclonal antibody, followed byhorseradish peroxidase-conjugated rabbit anti-goat IgG, were used.For �-actin detection, used as a loading control, membranes werestripped and then reprobed with an anti-actin rabbit antibody.Immunoblots were developed by enhanced chemiluminescence.

Jour

nal o

f Cel

l Sci

ence

3178

expressed in yeast, in contradiction with that described formammalian cells. In mammalian cells, PKC-�, -�, -� and -are considered major isoforms in apoptosis and have distinctinfluences on apoptosis regulation (Musashi et al., 2000;Hofmann, 2001; Gutcher et al., 2003). Many proteins areknown to be involved in the regulation of apoptosis, and themembers of the Bcl-2 family are recognised central regulators,dictating the fate of the cell either to live or die (Reed, 1997;Antonsson and Martinou, 2000). Several authors havesuggested that the inhibitory and stimulatory influences ofPKC isoforms on apoptosis can be mainly due to regulation,through phosphorylation, of proteins of the Bcl-2 family (Denget al., 2000; Kornblau et al., 2000; Musashi et al., 2000). Theabsence in yeast of endogenous proteins of the Bcl-2 familymay be a possible explanation for the similar pro-apoptoticeffects caused by PKC-�, -�, -� or - expression in S.cerevisiae.

To test this hypothesis we co-expressed in yeast the PKCisoform with Bcl-xL, an anti-apoptotic protein of the Bcl-2family, and evaluated the effect of this co-expression in S.cerevisiae acetic-acid-induced apoptosis. Expression of Bcl-xL in S. cerevisiae results in an increase in cell survival aftertreatment with acetic acid, accompanied by a reduction inDNA degradation and mitochondrial ROS production. Theseresults demonstrate an anti-apoptotic cytoprotective effect ofBcl-xL in acetic-acid-induced yeast cell death and add newevidence pointing to the involvement of this mode of cell deathin an apoptotic pathway. It was previously shown that Bcl-xLinhibits yeast apoptotic death induced by other apoptoticstimuli, such as hydrogen peroxide, menadione and heat that,like acetic acid, induce ROS production (Chen et al., 2003).The anti-apoptotic effect exhibited by Bcl-xL expression inyeast cells parallels that described for mammalian cells(Antonsson and Martinou, 2000). Together, these data furtherreinforce the idea that a bona fide cell death program occursin yeast and that pathways downstream of Bcl-xL areconserved between yeast and higher eukaryotes, as suggestedby Burhans et al. (Burhans et al., 2003). In addition, theoccurrence of an anti-apoptotic effect of Bcl-xL in thisunicellular eukaryote corroborates the use of yeast as a tool tounravel the complexities of the function of the Bcl-2 membersin apoptosis.

We then evaluated if PKC-�, -�, -� and - interfered with theBcl-xL anti-apoptotic activity. The results obtained reveal anisoform-dependent influence of PKC on the Bcl-xL anti-apoptotic activity. Co-expression of PKC-� or -� with Bcl-xLcompletely abolishes the Bcl-xL anti-apoptotic activity,whereas co-expression of PKC-� or - with Bcl-xL causes amarked enhancement of the Bcl-xL anti-apoptotic activity.

These results reinforced the hypothesis of a distinctmechanism for acetic-acid-induced apoptosis stimulation byPKC-�. In fact, PKC-�, -� or -, which elicited a pronouncedstimulation of mitochondrial ROS production, also interferedmarkedly with the mitochondrial anti-apoptotic proteinBcl-xL. Consistently, PKC-�, which only caused a slightstimulation of mitochondrial ROS production, displayed areduced effect on this mitochondrial protein. Furthermore, theycorroborate the interpretation that the different influences ofmammalian PKC isoforms on apoptosis can be due to distinctmodulation of apoptotic members of the Bcl-2 family by eachisoform (Gutcher et al., 2003).

A yeast pathway homologue to the mammalianSAPK/JNK is responsible for acetic-acid-induced Bcl-xLphosphorylation that is differently modulated by PKCisoformsUsing a cell-free phosphatase assay we detect, for the first time,the existence of a phosphorylated form of Bcl-xL aftertreatment with acetic acid in yeast expressing this protein.These results suggest that phosphorylation of Bcl-xL by astress-response kinase-signalling pathway hinders the anti-apoptotic function of Bcl-xL and permits yeast cells to die byacid-induced apoptosis, as previously reported for mammaliancells (Fan et al., 2000; Basu and Haldar, 2003).

Moreover, we find that the abolishment of the Bcl-xL anti-apoptotic effect by PKC-� co-expression is accompanied by apronounced decrease of the Bcl-xL dephosphorylated form,and that the remarkable increase in the Bcl-xL anti-apoptoticeffect by PKC-� or - co-expression is accompanied by apronounced decrease of the Bcl-xL phosphorylated form.Therefore, a possible explanation for the distinct modulationof Bcl-xL anti-apoptotic activity by each PKC isoform can beascribed to a PKC-isoform-dependent modulation of Bcl-xLphosphorylation. These data further reinforce the hypothesis ofa distinct mechanism for acetic-acid-induced apoptosisstimulation by expression of PKC-�, as discussed above. Infact, PKC-�, -� or -, which elicited a pronounced stimulationof mitochondrial ROS production also interfered markedlywith Bcl-xL, an anti-apoptotic protein primarily targeted to themitochondria (Polcic and Forte, 2003). By contrast, PKC-�caused only a slight stimulation of mitochondrial ROS andconsistently displayed a reduced effect on that protein.

Although the involvement of protein kinases in Bcl-xLphosphorylation, in response to distinct apoptotic stimuli, isreported in several studies with mammalian cells, thekinase(s) implicated still remain unclear (Du et al., 2005).Nevertheless, Jun N-terminal kinase (JNK) is a kinaseresponsible for phosphorylation of Bcl-xL, as well as of otheranti-apoptotic members of the Bcl-2 family, in response tooxidative stress (Fan et al., 2000; Basu and Haldar, 2003).Consistently with the results obtained in prostate cancercells treated with taxol or 2-methoxyestradiol (Basu andHaldar, 2003), pre-treatment of yeast cells expressing Bcl-xLwith JNK inhibitor II renders the cells more resistant toacetic-acid-induced apoptosis and eliminates Bcl-xLphosphorylated form. These results support the idea that Bcl-xL phosphorylation disables its anti-apoptotic function. Inaddition, they suggest the existence of a pathway in yeasthomologous to the mammalian SAPK/JNK pathwayresponsible for acetic-acid-induced Bcl-xL phosphorylation.The mammalian protein kinases p38 and SAPK/JNK arestructurally and functionally homologous to Hog1p in yeast(Kültz et al., 1997). Therefore, it is expected that they sharesimilar regulatory mechanisms (Takekawa et al., 1998) and itshould be considered that Hog1p mediates the observed Bcl-xL phosphorylation in response to acetic acid treatment inyeast. Supporting this possibility, Hog1p activation hasalready been observed in response to weak carboxylic acidstress (Lawrence et al., 2004). By contrast, in yeast co-expressing PKC-� and Bcl-xL, JNK inhibitor II does notinterfere with either cell death or Bcl-xL phosphorylation.These results suggest that Bcl-xL phosphorylation can occurby two independent pathways and that PKC-� can be

Journal of Cell Science 119 (15)

Jour

nal o

f Cel

l Sci

ence

3179Apoptosis regulation by mammalian PKC isoforms

involved in phosphorylation of Bcl-xL through a pathwayindependent of SAPK/JNK.

Du et al. (Du et al., 2005) proposed the existence inmammalian cells of a kinase and phosphatase system that maybe operating in tandem, leading to a coordinatedphosphorylation-dephosphorylation cycle that modulates Bcl-xL activity. However, the precise mechanisms for thismodulation remain to be determined. Our study offers newinsights into the role of phosphorylation on the modulation ofthe Bcl-xL function, identifying individual PKC isoforms asmodulators of the phosphorylation-dephosphorylation of Bcl-xL when expressed in yeast. Indeed, we demonstrate that PKC-� can phosphorylate Bcl-xL directly, or indirectly through akinase other than the yeast homologue to the mammalian JNK,and the possibility of modulation of Bcl-xL dephosphorylationthrough a yeast phosphatase activation by PKC-� and - iscurrently under investigation.

In conclusion, the results presented here point toconservation of targeted Bcl-xL phosphorylation in yeast andmammals and hence to the existence of a common ancientmechanism of apoptosis regulation. In the context of ourfindings and the recognised involvement of PKC onmammalian apoptosis regulation there is the interestingpossibility that the isoforms studied preserve their functionalrole regarding Bcl-xL phosphorylation. In addition, theelucidation of the enzymology of Bcl-xL phosphorylation maycontribute to understand the molecular mechanism of action ofanti-neoplasic chemotherapeutic drugs (Du et al., 2005) and todesign new strategies for apoptosis modulation. Finally, thegenetically tractable yeast promises advances in theunravelling of apoptosis regulation, namely of the Bcl-2members by individual PKC isoforms, towards a betterexploitation of mammalian PKC as a therapeutic target.

Materials and MethodsPlasmidsFor PKC-� expression, the bovine PKC-� cDNA was cloned into the YEp51 yeastexpression plasmid. For PKC-�, PKC-� and PKC- expression, the rat PKC-�, themouse PKC-� or PKC- cDNA, respectively, was cloned into the YEplac181 yeastexpression plasmid. These plasmids contain the promoter GAL (GAL1-10; GAL10in YEp51 and GAL1 in YEplac181) that is galactose inducible, and the LEU2 geneas a selectable marker.

For Bcl-xL expression, the human Bcl-xL cDNA was cloned into the pOW4 yeastexpression plasmid. This plasmid contains an ADH1 promoter for Bcl-xL expressionand a URA3 gene as a selectable marker.

All the plasmids used were amplified in Escherichia coli DH5� and confirmedby restriction analysis.

Yeast strain, transformation and growth conditionsSaccharomyces cerevisiae CG379 [� ade5 his7-2 leu2-112 trp1-289 ura3-52 (Kil-O)] strain was used for the yeast expression studies. Yeast cells were transformedusing the lithium acetate method (Ito et al., 1983). To ensure selection oftransformed yeast, cells were routinely grown in a minimal selective medium,consisting of 2% (w/v) glucose, 0.67% (w/v) yeast nitrogen base with ammoniumsulphate (Difco) and all the amino acids required for yeast growth (50 �g/ml),except leucine (for yeast only transformed with YEp51 or YEplac181 expressionplasmids) or leucine and uracil (for yeast co-transformed with pOW4 andYEplac181 or YEp51 expression plasmids). Transformed yeast cells werepreviously grown in this medium at 30°C with continuous shaking (150 r.p.m.), toapproximately 1 OD600 (Cary 1E Varian Spectrophotometer). Cells were afterwardscollected by centrifugation, diluted to 0.05 OD600 in a 2% (w/v) galactose-selectivemedium (PKC transcription inducer) with 3% (v/v) glycerol (alternative carbonsource) and grown to 0.5 OD600 (mid-log phase).

Western blotCell extracts of about 109 cells were prepared in 50 �l lysis buffer containing 0.5%NP-40, 20 mM HEPES (pH 7.4), 84 mM KCl, 10 mM MgCl2, 0.2 mM EDTA, 0.2mM EGTA, 1 mM DTT, 5 �g/ml aprotinin, 1 �g/ml leupeptin, 1 �g/ml pepstatin

and 1 mM phenylmethylsulfonyl fluoride. Cell suspensions were mixed with thesame volume of glass beads (0.45-0.5 mm diameter; Sigma) and lysed mechanicallyby six 30-second vortexing steps interrupted by cooling on ice for 1 minute. Celllysates were then clarified by centrifugation at 20,000 g for 15 minutes at 4°C. Thesupernatant was collected and protein concentration was determined using a kit forprotein quantification (Coomassie® Protein Assay Reagent Kit, Pierce). Proteinlysates (10-15 �g/lane) were separated on 15% SDS-polyacrylamide gels (SDS-PAGE) and transferred to nitrocellulose membranes (Bio-Rad). Preblocking andimmunostaining of blots were performed in Tris-buffered saline (TBS) containing0.05% Tween 20 and 5% non-fat milk at room temperature. After preblocking for1 hour, membranes were incubated with the primary antibody for 2 hours, followedby incubation with the secondary antibody for 1 hour. For PKC detection,membranes were incubated with a specific rabbit antibody to the individualmammalian PKC isoform: anti-PKC-� (1:20,000 dilution); anti-PKC-� (1:11,000dilution); anti-PKC-� (1:16,000 dilution) or anti-PKC- (1:20,000 dilution) (Sigma),followed by the horseradish-peroxidase-conjugated goat anti-rabbit IgG (1:5000dilution) (Sigma). For Bcl-xL detection, membranes were incubated with a Bcl-xLgoat polyclonal antibody (1:300 dilution) (Santa Cruz Biotechnology), followed bythe horseradish-peroxidase-conjugated rabbit anti-goat IgG (1:5000 dilution)(Zymed Laboratories). Membranes were stripped of bound antibody by incubatingthe strip in a buffer comprising 62.5 mM Tris-HCl (pH 6.8), 2% SDS, and 100 mM�-mercaptoethanol at 50°C for 30 minutes, and then reprobing with an actin rabbitantibody (1:200 dilution) (Sigma) followed by the horseradish peroxidase-conjugated goat anti-rabbit IgG (1:5000 dilution). Immunoblots were developed byenhanced chemiluminescence (ECL Western Blotting Analysis System; AmershamPharmacia Biotech).

Western blot analysis showed that expression of the four PKC isoforms in S.cerevisiae, after incubation in galactose-selective medium, occurred at comparablelevels (Fig. 8A, third lane). Expression of Bcl-xL was confirmed by the occurrenceof a single band at ~29 kDa (Fig. 8B). Expression of individual PKC isoforms orof Bcl-xL was not affected by co-expression of both proteins (Fig. 8A,B).

Cell death assaysExponential phase transformed yeast cells, previously grown in 2% (w/v) galactose-selective medium, were harvested and suspended in Sabouraud Dextrose BrothMedium (Difco) containing 0, 250, 300, 350 or 400 mM acetic acid and with a finalpH 3.0 (set with HCl). The treatments were carried out for 1 hour at 30°C withmechanical shaking (200 r.p.m.). In all the experiments, the extracellular pH did notchange during incubations.

Yeast cells expressing Bcl-xL (YEplac181 + Bcl-xL; YEp51 + Bcl-xL) and co-expressing PKC-� and Bcl-xL (PKC-� + Bcl-xL) were pre-treated only in thepresence of solvent (DMSO) or in the presence of 20 �M JNK inhibitor II(Calbiochem) for 8 hours followed by incubation with 300 mM acetic acid for 1hour at 30°C.

Cell death was determined by colony forming unit (c.f.u.) counts after 2 daysincubation at 30°C on Sabouraud Dextrose Agar (Difco) plates. No further coloniesappeared after that incubation period. The percentage of dead cells was estimatedconsidering 0% dead cells (100% survival) the number of c.f.u. obtained after 1hour incubation in the absence of acetic acid (0 mM).

Propidium iodide and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labelling (TUNEL) stainingPropidium iodide (PI) staining was used to monitor cell membrane integrity, aspreviously described (Ludovico et al., 2001). After treatment with acetic acid, 300�l of yeast samples were taken and incubated with 1.5 �l of 1 mg/ml PI (MolecularProbes), for 10 minutes at room temperature. TUNEL staining, to detect DNAstrand breaks, was performed using the In Situ Cell Death Detection Kit,Fluorescein (Roche Applied Science) and basically as described (Ludovico et al.,2001).

Cells incubated in the absence of acetic acid (0 mM) were used as negativecontrol. The samples were observed under an Eclipse E400 fluorescence microscope(Nikon, Japan) equipped with a 100 W mercury lamp and appropriate filter setting.

Determination of ROS productionReactive oxygen species (ROS) production by mitochondria was monitored by flowcytometry using a mitochondria-selective probe, MitoTracker Red CM-H2XRos(Molecular Probes), mainly as described (Ludovico et al., 2002). After treatmentwith acetic acid, about 107 cells/ml were harvested, washed once and resuspendedin 1 ml phosphate-buffered saline solution (PBS, pH 7.4) and preloaded with 0.4�g/ml dye for 20 minutes at 37°C. Cells incubated in the absence of acetic acid (0mM) were used as a negative control.

Determination of caspase activityFlow cytometric analysis: Detection of caspase activation was performed using theCaspACE, FITC-VAD-FMK In Situ Marker Kit (Promega). Briefly, after treatmentwith acetic acid, 106 cells were washed with PBS, resuspended in 100 �l stainingsolution containing 50 �M of FITC-VAD-FMK and incubated for 25 minutes at30°C. After incubation, cells were washed once with PBS and resuspended in 1 ml

Jour

nal o

f Cel

l Sci

ence

3180

PBS. Cells incubated in the absence of acetic acid (0 mM) were used as a negativecontrol.

Cell death assay in the presence of a caspase inhibitorIn experiments with the nonfluorescent caspase inhibitor Z-VAD-FMK (Promega),cells were treated for 2 hours at 30°C with 100 �M of Z-VAD-FMK, beforetreatment with 300 mM acetic acid. Cell death was assessed as described above.

Fluorimetric analysisCell extracts were obtained as described for western blot. Caspase activity wasdetermined by incubation of 25 �g protein lysate with 50 �M of the fluorogeniccaspase substrate N-acetyl-Asp-Glu-Val-Asp-aminomethylcoumarin (Ac-DEVD-AMC), N-acetyl-Ile-Glu-Thr-Asp-aminomethylcoumarin (Ac-IETD-AMC) or N-acetyl-Val-Glu-Ile-Asp-aminomethylcoumarin (Ac-VEID-AMC) (Biomol) in 200�l buffer containing 50 mM HEPES (pH 7.3), 100 mM NaCl, 10% sucrose, 0.1%CHAPS, and 10 mM DTT. Cells incubated in the absence of acetic acid (0 mM)were used as a negative control. The recombinant caspase 3 (for Ac-DEVD-AMC),caspase 6 (for Ac-VEID-AMC) and caspase 8 (for Ac-IETD-AMC) were used fortesting the fluorogenic substrates and assay conditions (positive control).Aminomethylcoumarin (AMC) release was monitored for 90 minutes at 30°C in aSpectraMAX Gemini Fluorescence Plate Reader (Molecular Devices), using anexcitation wavelength of 360 nm and an emission wavelength of 475 nm. Caspaseactivity was determined as the slope of the resulting linear regressions and expressedin arbitrary fluorescence units per minute.

Flow cytometric data acquisition and analysisFlow cytometric analysis was performed in an Epics® XL-MCLTM (BeckmanCoulter), flow cytometer equipped with an argon-ion laser emitting a 488 nm beamat 15 mW. Green fluorescence was collected through a 488 nm blocking filter, a 550nm long-pass dichroic and a 525 nm band-pass filter. Red fluorescence was collectedthrough a 560 nm short-pass dichroic, a 640 nm long-pass, and another 670 nmlong-pass filter. 20,000 cells were analysed per sample at low flow rate. Data wereanalysed by WinMDI 2.8 software.

In vitro phosphatase treatmentCell extracts of co-transformed yeast treated with 300 mM acetic acid, wereprepared as described for western blot. Aliquots of 25 �g protein lysate wereincubated with and without 400 units of -phosphatase (-PPase; New England

BioLabs) in 25 �l phosphatase buffer containing 50 mM Tris-HCl (pH 7.5), 0.1 mMEDTA, 5 mM DTT, 0.01% Brij 35 and 2 mM MnCl2, for 30 minutes at 30°C. Thereaction was stopped by boiling for 5 minutes in Laemmli sample buffer. Cellextracts were afterward analysed by western blot as described above.

Statistical analysisData were analysed statistically using the SigmaStat 3.1 programme (SYSTATSoftware, Inc.). Differences between means were tested for significance using theunpaired Student’s t-test. P values of 0.05 or lower were considered statisticallysignificant. Results are expressed as the mean ± s.e.m. of the indicated number ofexperiments.

We thank Manuel T. Silva and Frank Madeo for critical reading ofthe manuscript and all the helpful comments and suggestions. We aregrateful to Dr Heimo Riedel for YEp51 and YEp51-PKC-�, to DrNigel Goode for YEplac181, YEplac181-PKC-�, YEplac181-PKC-�and YEplac181-PKC- and to Dr Charles Rudin for pOW4 andpOW4-Bcl-xL. We are also grateful to IZASA for the use of theEpics® XL-MCLTM (Beckman Coulter); to Joana Tavares for her helpin some experiments; to Vitor Costa and Natércia Teixeira forproviding some reagents; and to Elsa Bronze for technical advice inthe phosphatase assay. We thank Fundação para a Ciência e aTecnologia, (FCT, Lisbon, Portugal) (I&D No. 226/94 and 655)POCTI and FEDER for financial support.

ReferencesAntonsson, B. and Martinou, J. C. (2000). The Bcl-2 protein family. Exp. Cell Res. 256,

50-57.Basu, A. and Haldar, S. (2003). Identification of a novel Bcl-xL phosphorylation site

regulating the sensitivity of taxol- or 2-methoxyestradiol-induced apoptosis. FEBS Lett.538, 41-47.

Bharti, A., Kraeft, S. K., Gounder, M., Pandey, P., Jin, S., Yuan, Z. M., Lees-Miller,L. P., Weichselbaum, R., Weaver, D., Chen, L. B. et al. (1998). Inactivation of DNA-dependent protein kinase by protein kinase C delta: implications for apoptosis. Mol.Cell. Biol. 18, 6719-6728.

Journal of Cell Science 119 (15)

Fig. 8. Expression of individualmammalian PKC isoforms or of Bcl-xL inS. cerevisiae is not affected by co-expression of both proteins.(A) Comparable levels of PKC isoformexpression were detected in proteinextracts obtained from yeast expressingPKC-�, -�, -� or - (PKC-pOW4 lanes)and from yeast co-expressing the PKCisoform and Bcl-xL (PKC-Bcl-xL lanes),after incubation in 2% (w/v) galactoseselective medium. Protein extractsobtained from yeast co-transformed withthe empty vectors were used as negativecontrols (YEp51-pOW4 or YEplac181-pOW4 lanes). (B) Comparable levels ofBcl-xL expression were detected inprotein extracts obtained from yeastexpressing Bcl-xL (YEp51-Bcl-xL orYEplac181-Bcl-xL lanes) and from yeastco-expressing the PKC isoform and Bcl-xL (PKC-�, -�, -� or --Bcl-xL lanes),after incubation in 2% (w/v) galactoseselective medium. Protein extractsobtained from yeast co-transformed withthe empty vectors were used as negativecontrols (YEp51/pOW4 or YEplac181-pOW4 lanes). Protein extracts (10-15�g/lane) were separated on 15% SDS-PAGE followed by western blot analysis. For PKC isoform detection, a specific rabbit antibody to each PKC isoform, followed by horseradishperoxidase-conjugated goat anti-rabbit IgG, was used. For Bcl-xL detection, anti-Bcl-xL goat polyclonal antibody, followed by horseradishperoxidase-conjugated rabbit anti-goat IgG, was used. For �-actin detection, used as a loading control, membranes were stripped and thenreprobed with an anti-actin rabbit antibody. Immunoblots were developed by enhanced chemiluminescence

Jour

nal o

f Cel

l Sci

ence

3181Apoptosis regulation by mammalian PKC isoforms

Burhans, W. C., Weinberger, M., Marchetti, M. A., Ramachandran, L., D’Urso, G.and Huberman, J. A. (2003). Apoptosis-like yeast cell death in response to DNAdamage and replication defects. Mutat. Res. 532, 227-243.

Chen, S.-R., Dunigan, D. D. and Dickman, M. B. (2003). Bcl-2 family members inhibitoxidative stress-induced programme cell death in Saccharomyces cerevisiae. FreeRadic. Biol. Med. 15, 1315-1325.

Deng, X., Kornblau, S. M., Ruvolo, P. P. and May, W. S. (2000). Regulation of Bcl2phosphorylation and potential significance for leukemic cell chemoresistance. J. Natl.Cancer Inst. Monographs 28, 30-37.

Du, L., Lyle, C. S. and Chambers, T. C. (2005). Characterization of vinblastine-inducedBcl-xL and Bcl-2 phosphorylation: evidence for a novel protein kinase and acoordinated phosphorylation/dephosphorylation cycle associated with apoptosisinduction. Oncogene 24, 107-117.

Fan, M., Goodwin, M., Vu, T., Brantley-Finley, C., Gaarde, W. A. and Chambers, T.C. (2000). Vinblastine-induced phosphorylation of bcl-2 and bcl-xL is mediated byJNK and occurs in parallel with inactivation of the Raf-1/MEK/ERK cascade. J. Biol.Chem. 275, 29980-29985.

Gutcher, I., Webb, P. R. and Anderson, N. G. (2003). The isoform-specific regulationof apoptosis by protein kinase C. Cell Mol. Life Sci. 60, 1061-1070.

Herker, E., Jungwirth, H., Lehmann, K. A., Maldener, C., Frohlich, K. U., Wissing,S., Büttner, S., Fehr, M., Sigrist, S. and Madeo, F. (2004). Chronological aging leadsto apoptosis in yeast. J. Cell Biol. 164, 501-507.

Hofmann, J. (2001). Modulation of protein kinase C in antitumor treatment. Rev. Physiol.Biochem. Pharmacol. 142, 1-96.

Hofmann, J. (2004). Protein kinase C isozymes as potential targets for anticancer therapy.Curr. Cancer Drug Targets 4, 125-146.

Ito, H., Fukuda, Y., Murata, K. and Kimura, A. (1983). Transformation of intact yeastcells treated with alkali cations. J. Bacteriol. 153, 163-168.

Keenan, C., Goode, N. and Pears, C. (1997). Isoform specificity of activators andinhibitors of protein kinase C � and �. FEBS Lett. 415, 101-108.

Kornblau, S. M., Vu, H. T., Ruvolo, P., Estrov, Z., O’Brien, S., Cortes, J., Kantarjian,H., Andreeff, M. and May, W. S. (2000). Bax and PKC-� modulates the prognosticimpact of Bcl-2 expression in acute myelogenous leukaemia. Clin. Cancer Res. 6,1401-1409.

Kültz, D., Garcia-Perez, A., Ferraris, J. D. and Burg, M. B. (1997). Distinctregulation of osmoprotective genes in yeast and mammals. J. Biol. Chem. 272,13165-13170.

Lawrence, C. L., Botting, C. H., Antrobus, R. and Coote, P. J. (2004). Evidence of anew role for the high-osmolarity glycerol mitogen-activated protein kinase pathway inyeast: regulating adaptation to citric acid stress. Mol. Cell. Biol. 24, 3307-3323.

Ludovico, P., Sousa, M. J., Silva, M. T., Leao, C. and Corte-Real, M. (2001).Saccharomyces cerevisiae commits to a programmed cell death process in response toacetic acid. Microbiology 147, 2409-2415.

Ludovico, P., Rodrigues, F., Almeida, A., Silva, M. T., Barrientos, A. and Corte-Real,M. (2002). Cytochrome c release and mitochondria involvement in programmed celldeath induced by acetic acid in Saccharomyces cerevisiae. Mol. Biol. Cell 13, 2598-2606.

Ludovico, P., Madeo, F. and Silva, M. T. (2005). Yeast programmed cell death: anintricate puzzle. IUBMB Life 57, 129-135.

Madeo, F., Frohlich, E. and Frohlich, K. U. (1997). A yeast mutant showing diagnosticmarkers of early and late apoptosis. J. Cell Biol. 139, 729-734.

Madeo, F., Engelhardt, S., Herker, E., Lehmann, N., Maldener, C., Proksch, A.,Wissing, S. and Frohlich, K. U. (2002a). Apoptosis in yeast: a new model systemwith applications in cell biology and medicine. Curr. Genet. 41, 208-216.

Madeo, F., Herker, E., Maldener, C., Wissing, S., Lächelt, S., Herlan, M., Fehr, M.,Lauber, K., Sigrist, S. J., Wesselborg, S. et al. (2002b). A caspase-related proteaseregulates apoptosis in yeast. Mol. Cell 9, 911-917.

Madeo, F., Herker, E., Wissing, S., Jungwirth, H., Eisenberg, T. and Frohlich, K. U.(2004). Apoptosis in yeast. Curr. Opin. Microbiol. 7, 655-660.

Martelli, A. M., Mazzotti, G. and Capitani, S. (2004). Nuclear protein kinase Cisoforms and apoptosis. Eur. J. Histochem. 48, 89-94.

Meier, P., Finch, A. and Evan, G. (2000). Apoptosis in development. Nature 407, 796-801.

Musashi, M., Ota, S. and Shiroshita, N. (2000). The role of protein kinase C isoformsin cell proliferation and apoptosis. Int. J. Hematol. 72, 12-19.

Perez, P. and Calonge, T. M. (2002). Yeast protein kinase C. J. Biochem. 132, 513-517.

Polcic, P. and Forte, M. (2003). Response of yeast to the regulated expression of proteinsin the Bcl-2 family. Biochem. J. 374, 393-402.

Reed, J. C. (1997). Double identity for proteins of the Bcl-2 family. Nature 387, 773-776.

Saraiva, L., Fresco, P., Pinto, E., Sousa, E., Pinto, M. and Gonçalves, J. (2003).Inhibition of �, �I, �, � and protein kinase C isoforms by xanthonolignoids. J. EnzymeInhib. Med. Chem. 18, 357-370.

Saraiva, L., Fresco, P., Pinto, E. and Gonçalves, J. (2004). Characterization of phorbolesters activity on individual mammalian protein kinase C isoforms, using the yeastphenotypic assay. Eur. J. Pharmacol. 491, 101-110.

Shieh, H.-L., Hansen, H., Zhu, J. and Riedel, H. (1996). Activation of conventionalmammalian protein kinase C isoforms expressed in budding yeast modulates the celldoubling time-a potential in vivo screen for protein kinase C activators. Cancer Detect.Prev. 20, 576-589.

Silva, R. D., Sotoca, R., Johansson, B., Ludovico, P., Sansonetty, F., Silva, M. T.,Peinado, J. M. and Côrte-Real, M. (2005). Hyperosmotic stress induces metacaspase-and mitochondria-dependent apoptosis in Saccharomyces cerevisiae. Mol. Microbiol.58, 824-834.

Spitaler, M. and Cantrell, D. A. (2004). Protein kinase C and beyond. Nat. Immunol. 5,785-790.

Takekawa, M., Maeda, T. and Saito, H. (1998). Protein phosphatase 2C� inhibitsthe human stress-responsive p38 and JNK MAPK pathways. EMBO J. 17, 4744-4752.

Vermes, I., Haanen, C. and Reutelingsperger, C. (2000). Flow cytometry of apoptoticcell death. J. Immunol. Methods 243, 167-190.

Wysocki, R. and Kron, S. J. (2004). Yeast cell death during DNA damage arrest isindependent of caspase or reactive oxygen species. J. Cell Biol. 166, 311-316.Jo

urna

l of C

ell S

cien

ce