Oncogene (2000) 19, 2354 ± 2362 ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc Caspases and mitochondria in c-Myc-induced : identi®cation of ATM as a new target of caspases

Anneli Hotti1,2, Kristiina JaÈ rvinen1,2, Pirjo Siivola1 and Erkki HoÈ lttaÈ *,1

1Haartman Institute, Department of Pathology, University of Helsinki, P.O. Box 21 (Haartmaninkatu 3), FIN- 00014, Finland

The mechanism(s) of c-Myc transcription factor-induced transcription of the target (Galaktionov et al., apoptosis is still obscure. The activation of c-Myc has 1996; Packham and Cleveland, 1994; Shim et al., 1997). been found to lead into the processing/activation of On the other hand, c-Myc has been reported to induce caspases (caspase-3), but the signi®cance of this for the apoptosis without transcriptional activation (Evan et al., demise is debatable. Here we report that several 1992; Xiao et al., 1998). Also, it is not clear whether the targets of caspases (PKCd, MDM2, PARP, replication tumour suppressor , a transcriptional target for c- factor C, 70 kDa U1snRNP, fodrin and lamins) are Myc, is needed for apoptosis induced by c-Myc (Lotem cleaved during c-Myc-induced apoptosis in Rat-1 and Sachs, 1995; Shim et al., 1998; Wagner et al., 1994; MycERTM cells, indicating an important role for caspases Zindy et al., 1998; reviewed in Prendergast, 1999; in the apoptotic process. We further found that the ATM Thompson, 1998). Recently, c-Myc-induced apoptosis (ataxia telangiectasia mutated) ± is a novel key has been shown either indirectly (Kagaya et al., 1997; substrate of caspases. In in vitro assays, puri®ed McCarthy et al., 1997) or directly (Kangas et al., 1998) recombinant ATM protein was found to be cleaved by to be associated with the activation of caspases, the e€ector caspases 3 and 7. The functional signi®cance members of the CED3/ICE-family of cysteine proteases. of the ATM cleavage is supported by the ®nding that However, it has also been suggested that caspases would ectopic expression of ATM protected in part against not be critically involved in the execution of the apoptosis. We also show that c-Myc-induced apoptosis apoptotic cell death induced by c-Myc (McCarthy et involves loss of mitochondrial transmembrane potential, al., 1997; Tsunoda et al., 1999, reviewed in Green and release of cytochrome c from mitochondria into the Kroemer, 1998). Further, data demonstrating a require- cytosol and subsequent processing of caspase-9. The ment for the CD95 receptor-ligand pathway (known to cleavage of caspase-9 is, however, minimal and a much involve the caspase and/or Jun N-terminal kinase later event than the processing/activation of caspase-3, pathways) in c-Myc-induced apoptosis has been ob- suggesting that it is not the apical caspase. Evidence is tained (Hueber et al., 1997). provided that there is, nevertheless, an upstream In this study, our aim was to assess the role of caspase(s) regulating the functions of caspase-3 and caspases in c-Myc-induced apoptosis by analysing a mitochondria. Additionally, it was found that p53 variety of essential documented to become becomes upregulated, together with its transcriptional cleaved by caspases in other apoptotic systems. We also targets MDM2 and p21, upon c-Myc induction, but this searched for novel key substrates for caspases by occurs also at a later time than the activation of caspase- studying known pivotal proteins with potential caspase 3. Oncogene (2000) 19, 2354 ± 2362. cleavage sites. One such putative target protein identi®ed is the ATM (ataxia telangiectasia mutated) protein, Keywords: apoptosis; ATM (ataxia telangiectasia mu- which is homologous to DNA-dependent protein kinase tated); caspases; c-Myc; mitochondria; p53 (DNA ± PK), a known substrate of caspases (Casciola- Rosen et al., 1996; Song et al., 1996). ATM is responsible for the ataxia telangiectasia (AT) -disease, Introduction the clinical features of which are numerous including dilated blood vessels, cerebellar degeneration, defects of c-Myc is a helix ± loop ± helix transcription factor immune system, chromosomal aberrations and predis- associated with cell growth and di€erentiation, but also position to cancer (for a review see Morgan and Kastan, with programmed cell death, apoptosis. The molecular 1997). The AT-cells are extremely sensitive to radiation. mechanisms of apoptosis induced by c-Myc are still They fail to arrest in the cell cycle (in G1, S, and G2) poorly understood if compared, for example, to the after ionising radiation-induced DNA damage (reviewed various death receptor-mediated forms of apoptosis. in Meyn, 1995) and proceed in the cycle but eventually One major enigma is how c-Myc, a nuclear protein, die at G2, presumably by apoptosis. The AT-cells initiates apoptosis involving cytoplasmic events. The display an altered expression of p53 protein (Canman transactivation domain of c-Myc is essential for all et al., 1994; Kastan et al., 1992), and also the expression identi®ed functions of c-Myc, which implies that also of the p53 target genes (p21, gadd45, mdm2) is altered apoptosis induced by c-Myc may be achieved by (Canman et al., 1994; Kastan et al., 1992). The ATM protein has a region homologous to the protein kinase domain of the phosphatidylinositol-3-kinase protein (Savitsky et al., 1995) and possesses kinase activity. *Correspondence: E HoÈ lttaÈ ATM can associate with and phosphorylate the product 2A Hotti and K JaÈ rvinen contributed equally to the work Received 24 August 1999; revised 2 March 2000; accepted 7 March of c-abl proto-oncogene (Baskaran et al., 1997; Shafman 2000 et al., 1997) suggested to play a role in the regulation of Mechanisms of c-Myc-induced apoptosis A Hotti et al 2355 G1/S transition. Moreover, ATM has been shown ®broblasts. The target proteins can be divided roughly recently to phosphorylate p53 on Ser15 after DNA into three subgroups by their function in cells. First, damage (Banin et al., 1998; Canman et al., 1998), which PKCd and MDM2 (shown in Figure 8) are regulatory may be of importance for the p53-mediated cell cycle proteins involved in signal transduction and cell cycle arrest and apoptosis. progression. Second, poly(ADP-ribose) polymerase Further, possible mitochondrial alterations (loss of (PARP), replication factor C (RFC), and the 70 kDa mitochondrial transmembrane potential, release of subunit of U1 snRNP are involved in the maintenance cytochrome c), and accumulation of p53 protein, and of the integrity of the genome and replication of DNA, their kinetic relationships with the activation of and in the processing of the mRNA transcripts, caspases were analysed. respectively. Third, fodrin/BIIspectrin together with lamins A, B1 and B2 are structural (cytoskeletal and nuclear) proteins, which get proteolytically processed. Most of the proteins studied above are beginning to be Results cleaved relatively early (by 24 h) after the induction of c-Myc expression in relation, for example, to the Cleavage of previously known substrates of caspases degradation of DNA occurring later in these cells during c-Myc-induced apoptosis (Kangas et al., 1998). That c-Myc is responsible for To assess the functional role of caspases in c-Myc- induction of these caspase-mediated protein cleavages, induced apoptosis, we studied the fate of potential is evidenced by the fact that treatment of parental, caspase targets in the ®broblastic Rat-1 MycERTM cell serum-starved Rat-1 cells (without MycER) with OHT line expressing a chimeric c-Myc/oestrogen receptor did not lead into the activation of caspases and protein, activated by 4-hydroxytamoxifen (OHT) (Evan proteolysis (data not shown). These data lend support et al., 1992; Littlewood et al., 1995). The Rat-1 to the idea that the function of caspases is important MycERTM cells were starved for 24 h in 0.5% FCS, for c-Myc-induced apoptosis. whereafter 100 nM OHT was added to induce Of the proteins found to become degraded during apoptosis. As depicted in Figure 1a,b, many proteins the cell death induction by c-Myc, many are thought to identi®ed previously as targets of caspases are the be substrates for caspase-3 or caspase-7. So far, PKCd victims also in c-Myc-induced apoptosis in rat is documented to be cleaved only by caspase-3 and not

Figure 1 (a,b) Proteolytical processing of known substrates of caspases during c-Myc-induced apoptosis. Rat-1 MycERTM cells, starved in 0.5% FCS for 24 h, were treated with or without OHT (+/7) for the indicated times. Total cell lysates were analysed by Western blotting for the indicated proteins. In (a) panels, the samples are derived from the same lysates and the loading control shows reblotting of the particular ®lter with an anti-actin antibody, and in (b) the loading control represents a non-speci®c, low molecular weight band reacting with the anti-fodrin antibody and stands for all the panels. Asterisk (*) denotes the intact protein and the cleavage products are indicated by arrows. Sizes of molecular weight markers are shown on the left

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2356 by caspases 1, 2, 4, 5, 6, or 7 in vitro (Ghayur et al., Figure 2a, the full-length ATM protein of 350 kDa is 1996), and thus it is most likely that caspase-3 is the degraded relatively early during c-Myc-induced apop- enzyme responsible for the processing of PKCd in vivo tosis to yield at least four cleavage products in rat in rat ®broblasts. The lamins again are known to be ®broblasts. Their sizes were estimated to be about 50, processed only by caspase-6 (Nicholson and Thornber- 60, 150 and 230 kDa. (A better visualization of the ry, 1997). Based on these data, it seems that at least 230 kDa band can be seen in Figure 2e where a 6% two e€ector caspases, caspase-3 and caspase-6, become SDS ± PAGE gel better separates this fragment from a activated in c-Myc-induced apoptosis. Additionally, we non-speci®c band nearby). It is noteworthy that the have also found that the third e€ector, caspase-7, cleavage of ATM is not limited to c-Myc-induced becomes processed/activated (Hotti et al. manuscript in apoptosis, as ATM becomes proteolytically cleaved preparation). also in staurosporine treated Rat-1 and HL-60 -cells It may also be noteworthy that actin (Figure 1a), (Figure 2b,c) concomitant with the cleavage of PARP gelsolin, Rb, cPLA2 and SREBP (data not shown) were upon caspase activation (Figure 2d). not found to be cleaved after the induction of c-Myc To assess the dependency of the proteolytic proces- expression. Further, we were unable to detect expres- sing of ATM on caspase activation, the expression of sion of Gas2 or D4-GDP dissociation inhibitor for the c-Myc and apoptosis was induced in the presence of Z- Rho family with the antibodies used. VAD-fmk, a pan-inhibitor of caspases. As shown in Figure 2e, the inhibition of caspase activity by Z-VAD- fmk prevented the cleavage of the full-length ATM ATM is proteolytically cleaved during c-Myc-induced protein and the appearance of the proteolytic frag- apoptosis ments, demonstrating that a caspase(s) is directly or In searching for novel important targets of caspases, indirectly responsible for the proteolytic cleavage of the we investigated whether the ATM protein, which has rat ATM protein. In the human and mouse ATM many potential caspase cleavage sites, could be a target protein, there are several potential recognition sites for of caspases in c-Myc-induced apoptosis. As shown in caspase-3, which is activated in Rat-1 MycERTM-cells

Figure 2 ATM is proteolysed in apoptosis. (a) Rat-1 MycERTM cells starved in 0.5% FCS for 24 h were treated with or without OHT (+/7) for the indicated times. Total cell lysates were subjected to Western blot analysis of ATM protein. Asterisk (*) denotes the intact protein and the cleavage products are indicated by arrows. Sizes of molecular weight markers are shown on the left. (b) Rat-1 ®broblasts were incubated without (7) or with (+) 0.5 mM staurosporine for 4 h, whereafter total cell lysates were analysed for the ATM cleavage by immunoblotting. (c) HL-60 cells were treated with (+) or without (7)1mM staurosporine for 2 or 6 h and total cell lysates were subjected to Western blot analysis of ATM. (d) The cleavage of PARP in staurosporine treated HL-60 cells is shown to verify the apoptotic status of the cells. (e) Rat-1 MycERTM cells were grown as described in (a) in the presence or absence of the caspase inhibitor Z-VAD (100 mM, 24 h) and total cell lysates were subjected to Western blot analysis of ATM protein

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2357 following c-Myc induction and apoptosis (Kangas et al., 1998), as well as for the other caspases. The antibody used to detect the rat ATM protein was generated against the amino acid region 1980 ± 2338 of the human ATM protein. A di€erential usage of the possible caspase cleavage sites along the protein could yield several protein fragments that could represent the detected proteolytic products. The exact cleavage sites on the rat ATM protein remain to be determined ± once the rat ATM has been cloned ± by using proteins with mutations in the speci®c Asp-sites and by in vitro degradation studies with di€erent caspases.

Recombinant human ATM protein is cleaved by caspase-3 and caspase-7 To identify the potential caspase(s) responsible for the cleavage of human ATM, recombinant ATM protein was isolated from 293-EBNA cells transfected with an expression vector carrying a FLAG-tagged, full-length human ATM cDNA (see Materials and methods) and incubated in the presence of equal amounts of various active recombinant caspases. As shown in Figure 3a, the 350 kDa intact ATM protein was cleaved to a Figure 3 Recombinant human ATM is cleaved by caspases 3 fragment of about 230 kDa by caspases 3 and 7 (the and 7. (a) Puri®ed recombinant human ATM was incubated in latter producing a doublet), while caspases 6 and 8 the absence or presence of the indicated recombinant, active (that were con®rmed to be active; see Figure 3b) did caspases (0.5 mg of each) for 20 min at 378C as described in Materials and methods. The reaction products were resolved on not seem to be capable of cleaving ATM at the SDS ± PAGE, and ATM and its cleavage products were detected concentration and incubation time used. As shown by immunoblotting with the anti-ATM AHP392 antibody. above, a fragment of this size is also generated in the Asterisk (*) denotes the intact ATM protein and the cleavage apoptotic cells in vivo, suggesting that the caspases 3 products are indicated by the arrow. (b) Testing the activities of the recombinant caspases 6 and 8 towards their known substrates. and 7, both of which we have found to become The activity of caspase-6 was checked by a speci®c cleavage of activated by Myc in the Rat-1 MycERTM cells, may lamin A in isolated HeLa cell nuclei incubated in the presence of also contribute to the degradation of ATM in vivo. the recombinant caspases (Orth et al., 1996). The reaction products were solubilized in Laemmli sample bu€er and subjected to Western blot analysis with antibodies to lamin A+B2. To Ectopic expression of human ATM protein inhibits verify the caspase-8 activity, the recombinant enzyme was c-Myc-induced apoptosis incubated with HeLa cell cytosolic extracts (or recombinant caspase-3), and the samples were then analysed for the cleaved To study the functional signi®cance of the proteolysis proteins by immunoblotting with antibodies to caspase-3 (or Bid, of ATM in c-Myc-induced apoptosis, the Rat-1 data not shown) MycERTM cells were transfected with the FLAG-tagged wild-type ATM expression vector (pEBS7-YZ5). The empty vector (pEBS7) was used as a control. The dihexyloxacarbocyanine iodide [DioC6(3)]. At 12 h transfectants were selected for hygromycin resistance there was no detectable change in the mitochondria for 3 ± 4 weeks, and the surviving population was tested yet, but, as shown in Figure 5, at about 24 h there was for the expression of the recombinant ATM and for its a reduction in the level of DioC6(3)-incorporation in e€ects on c-Myc-induced apoptosis (Figure 4). Nota- cells with induced c-Myc expression, indicating a loss bly, ectopic expression of the ATM protein protected of the transmembrane potential. in part from c-Myc-induced apoptosis, indicating that Treatment of the cells with the broad-spectrum the cleavage of ATM contributes to the cell death caspase inhibitor Z-VAD-fmk (150 mM) inhibited the induced by c-Myc in these cells. loss of the transmembrane potential during the 24 h period (Figure 5). Z-DEVD-fmk (50 mM, see Kangas et al., 1998), an inhibitor of caspase-3-like proteases, was c-Myc-induced apoptosis is associated with loss of much less e€ective (data not shown). This demon- mitochondrial transmembrane potential in a strates that the inhibition of the activity of caspases or caspase-dependent manner other Z-VAD sensitive protease(s) prevents or delays Given that mitochondrial changes may activate the loss of mitochondrial transmembrane potential caspases or independently induce apoptosis (Green suggesting that there is an upstream caspase regulating and Reed, 1998 and references therein), the possible the mitochondrial changes. Though caspase-3 starts to involvement of mitochondria in the apoptotic process be cleaved 8 h after the initiation of apoptosis and is induced by c-Myc was studied. Rat-1 MycERTM cells further activated at 16 ± 24 h (Kangas et al., 1998), the were starved for 24 h in 0.5% FCS and then treated inability of Z-DEVD to inhibit the loss of the with OHT as above to induce apoptosis. The cells were transmembrane potential at 24 h suggests that a harvested at 12, 24 or 48 h and analysed for the discrete caspase, acting upstream of or parallel with mitochondrial transmembrane potential by exploiting caspase-3, is activated during the early phase of the the ability of the mitochondria to incorporate 3,3'- apoptotic process.

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2358

Figure 4 Expression of recombinant human ATM in Rat-1 MycERTM cells protects in part against c-Myc-induced apoptosis. (a) Rat-1 MycERTM cells were transfected with an episomal expression vector carrying a FLAG-tagged, full-length human ATM cDNA and hygromycin resistance marker. As a control, cells were transfected with the empty vector. After selection of Figure 5 Apoptosis induced by c-Myc is accompanied by a loss stable transfectants, the populations of surviving cells were of mitochondrial transmembrane potential. To measure mito- cultured with or without OHT (+/7) for the indicated times, chondrial transmembrane potential the Rat-1 MycERTM cells and apoptosis was scored by counting the number (percentage) of were exposed to the mitochondria-speci®c DioC6(3)-dye and its dead, detached cells in a Coulter counter. The results are shown incorporation was analysed by FACS. (a) Rat-1 MycERTM cells as means (+) standard deviation of three or four independent were starved in 0.5% FCS (24 h) and cultured without 4- experiments at both the indicated times. It is notable that, the hydroxytamoxifen (OHT) (dark grey) or with OHT (light grey), protective e€ect of ATM against the c-Myc-induced apoptosis or with both OHT and the caspase-inhibitor Z-VAD for 24 h was also found at the other points analysed, i.e. at the earlier (white), whereafter the uptake of DioC6(3) was determined. (b) (24 ± 30 h) and later (72 h) phases of apoptosis. (b) The stable Incorporation of DioC6(3) into cells grown in 5% FCS (control, populations of the transfected cells were subjected to Western with normal membrane potential) (light grey) or into cells treated blotting analysis of the recombinant ATM protein. Total protein with the uncoupling agent mCICCP (membrane potential lost) were separated by SDS ± PAGE and blotted with monoclonal (white) anti-FLAG M2 antibody whereafter it steadily increases and reaches its max- imum at 48 h (Figure 6 and data not shown). Release of cytochrome c during c-Myc-induced apoptosis As the cytoplasmic cytochrome c is known to form a Subsequently, the possible release of mitochondrial complex (called apoptosome) with Apaf-1, ATP, and proteins was studied. After inducing the c-Myc activity, caspase-9 and thereby promote the autocatalytic the Rat-1 MycERTM cells were fractionated into the proteolytic processing and activation of caspase-9 crude mitochondrial and cytoplasmic fractions, and (Zou et al., 1997), the e€ect of the cytochrome c immunoblotted for the presence of cytochrome c. As release on caspase-9 was studied in the Rat-1 shown in Figure 6, a release of cytochrome c from the MycERTM cells. Figure 7a depicts the immunoblotting mitochondria into the cytoplasm takes place. The of caspase-9 during apoptosis induced by c-Myc. translocation can be observed relatively early at about Surprisingly, no proteolytic processing of caspase-9 8 ± 16 h after the induction of c-Myc expression, was observed until 48 h, when a minor proteolytic

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2359

Figure 6 c-Myc induces a release of cytochrome c from mitochondria to cytosol. Rat-1 MycERTM cells were cultured and treated with OHT for the indicated times as in Figure 1. After harvesting, the cells were gently disrupted in a glass homogenizer and subjected to subcellular fractionation as described in Materials and methods. Proteins (40 mg) from the cytosolic fraction and mitochondrial preparation were separated by SDS ± PAGE (12%), and immunoblotted with anti-cytochrome c antibody. Cytochrome c is indicated by the arrow. The 46 kDa protein cross-reacting with the cytochrome c antibody (Ziv et al., 1997) is shown for the loading control

caspase-9 antibody with the rat protein was further con®rmed by side by side blotting of a human B-cell lymphoma lysate, in which also the typical generation of the 35 kDa intermediate fragment of caspase-9 after the addition of 1 mM dATP and 0.2 mg cytochrome c in vitro (Li et al., 1997) was observed (data not shown).

p53 in c-Myc-induced apoptosis Having discovered that MDM2, a negative regulator of p53, and ATM, a p53 kinase (Banin et al., 1998; Canman et al., 1998), are cleaved, we investigated the behaviour of p53 during c-Myc-induced apoptosis. In our previous studies with Rat-1 MycERTM cells which were not pre-starved for serum, we did not observe any signi®cant changes in the level of p53 protein within 16 h after the induction of c-Myc expression (Kangas et al., 1998). However, in cells starved ®rst in 0.5% serum for 24 h, the expression of p53 showed an increase at 24 h after the c-Myc activation (Figure 8). This was accompanied by an increase in MDM2 and p21, the transcriptional targets of p53 (Figure 8). In contrast, the expression of Bax, which is also a Figure 7 Caspase-9 processing during the c-Myc- and stauros- transcriptional target of p53, remained unchanged porine-induced apoptosis in Rat-1 MycERTM cells. (a) Aliquots of (data not shown). the same cytosolic fractions (50 mg of protein) from the Rat-1 MycERTM cells as shown in Figure 6, were resolved on SDS ± PAGE (12%) and subjected to immunoblotting analysis with polyclonal anti-caspase-9 antibody (Pharmingen). Of note, similar Discussion results were also obtained in independent experiments using total cellular proteins for the Western blot analysis of caspase-9 (not Caspases in c-Myc-induced apoptosis shown). (b) Rat-1 MycERTM cells were exposed or not to 0.2 mM staurosporine for 5 h, whereafter the cytosolic fractions were The mechanisms of c-Myc-induced apoptosis are still prepared and subjected to Western blot analysis of caspase-9 as above. The typical release of cytochrome c following the unclear. Although it has been suggested that caspases staurosporine treatment is demonstrated by immunoblotting with would not be responsible for the apoptotic cell death in the anti-cytochrome c antibody in the bottom panel. The 46 kDa apoptosis induced by c-Myc (Green and Reed, 1998; protein unspeci®cally reacting with the anti-cytochrome c anti- McCarthy et al., 1997; Tsunoda et al., 1999), we body is shown for the loading control. In a and b the intact propose the opposite. This conclusion is supported by 46 kDa caspase-9 protein (doublet) is shown by the asterisk (*) and the about 17 kDa cleavage product representing the large the following observations. First, the caspase inhibitor subunit of caspase-9 is indicated by the arrow Z-VAD is able to inhibit or delay the progression of apoptosis induced by c-Myc (Kagaya et al., 1997; Kangas et al., 1998; McCarthy et al., 1997). Here, it is fragment of caspase-9 was detected. As a reference, a important to take into account that a 100% inhibition clear, rapid processing of caspase-9 in the Rat-1 of the activity of the caspases may be impossible to MycERTM cells is seen after induction of apoptosis achieve whereby the caspase cascade may eventually and release of cytochrome c by treatment with become activated despite the presence of the inhibitor, staurosporine (Figure 7b), demonstrating that the which complicates the interpretations. Second, caspase- antibody used (Pharmingen; raised against human 3 has been shown to become cleaved to active subunits caspase-9) recognizes well the proteolytic fragments of early during the c-Myc-induced apoptosis (Kangas et rat caspase-9 protein. The cross-reactivity of the al., 1998). Third, several of the proteins previously

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2360 implies that the mitochondria may function in mediating the apoptotic signal in c-Myc-induced cell death or be direct e€ectors of apoptosis as such. In various other forms of apoptosis, the loss of the mitochondrial transmembrane potential and the for- mation of permeability transition pores have been suggested to result in a release of mitochondrial proteins into the cytoplasm (e.g. AIF and cytochrome c) (reviewed in Green and Reed, 1998; Kroemer et al., 1997), although the release of cytochrome c may also Figure 8 The expression patterns of p53, MDM2 and p21 occur independently of the loss of the DCm (Bossy- during c-Myc-induced apoptosis. Rat-1 MycERTM cells were Wetzel et al., 1998). Cytochrome c is then thought to cultured as in Figure 2. Total cell lysates were analysed by trigger the activation/autocleavage of caspase-9, which Western blotting for the indicated proteins. The intact proteins in turn activates caspase-3 (Zou et al., 1997). We found are shown by asterisks and the cleavage product by an arrow. TM Note that there is an unknown band above the p53 protein signal that also in Rat-1 MycER ®broblasts cytochrome c (representing nonspeci®c or p53-related protein) becomes translocated from the mitochondria into the cytoplasm relatively early (at 8 ± 16 h) during c-Myc- induced apoptosis. However, the processing of caspase- demonstrated to be speci®c targets of caspases (and 9 occurred only after 48 h, i.e. considerably later than thought to be important for various cellular functions) the activation of caspase-3 (at 8 ± 16 h) (Kangas et al., get proteolytically cleaved during the apoptotic cell 1998). This could be taken as an indication that death initiated by c-Myc (this study). Fourth, mouse caspase-9 is not responsible for the processing of embryo ®broblasts de®cient in caspase-9 are resistant caspase-3. However, it is notable that caspase-9 may to c-Myc-induced apoptosis (Soengas et al., 1999). also be activated, at least in vitro, without proteolytic processing by the cytosolic factors released from Proteolytic cleavage of ATM in apoptosis mitochondria (Stennicke et al., 1999). Anyway, our ®nding that the proteolytic processing of caspase-3 is There is previous evidence that ATM, the protein inhibited by the caspase inhibitor Z-VAD in the Rat-1 product of the gene mutated in ataxia telangiectasia, MycERTM cells (Kangas et al., 1998) indicates that could be associated with the regulation of apoptosis. there is some apical caspase to be identi®ed. Whether First, autopsies of AT-homozygote patients reveal the same upstream caspase could also be responsible extensive apoptosis in di€erent tissues, and second, for the Z-VAD sensitive loss of the mitochondrial there is increased spontaneous apoptosis in AT-cells in membrane potential following c-Myc induction re- cell culture as compared to their normal counterparts mains interesting to be seen. We are currently studying (Meyn, 1995 and references therein). Also, it has been the possible involvement of caspase-8 in the regulation reported recently that if the expression of ATM is of these processes. prevented by an anti-sense expression of ATM, the viability of cells decreases following ionizing radiation (Zhang et al., 1998). Of speci®c interest, in our search p53 and caspases for new caspase substrates we identi®ed ATM as a The role of p53 in c-Myc-induced apoptosis is a candidate. ATM is a large, multifunctional protein that controversial issue (for reviews see Prendergast, 1999; could regulate apoptosis in multiple ways. However, Thompson, 1998). In the Rat-1 MycERTM cells, we the carboxyterminal PI-3-kinase domain likely has the found that the stabilization of p53 occurs signi®cantly most signi®cant activity (Morgan and Kastan, 1997). later than the activation of caspases (caspase-3). Thus, Finding that ATM is cleaved when apoptosis occurs, the accumulation of p53 is unlikely to be an initiating and that ectopic overexpression of full-length ATM signal for the activation of caspases and the induction protects in part from apoptosis, it seems likely that the of apoptosis by c-Myc. However, p53 might well proteolytic cleavage of ATM results in loss of the trigger or potentiate the caspase-9 activity (Soengas et activity of the protein. In support of this, treatment of al., 1999) or induce a novel caspase (Pochampally et recombinant ATM with caspases 3 and 7 abrogated its al., 1999), and thereby promote the apoptotic cell death kinase activity toward p53 (data not shown). The in these cells. ®nding that ATM is cleaved also during the staur- osporine-induced apoptosis in Rat-1 and HL-60 cells suggests that ATM may be an important apoptotic target in general. However, a recent contradictory Materials and methods study shows that a subpopulation of neuronal cells Cell culture lacking functional ATM is more resistant to apoptosis following ionizing irradiation (Herzog et al., 1998). The Rat-1 MycERTM ®broblast cell line expressing a Presumably, the results may depend on the apoptotic conditionally active c-Myc-oestrogen receptor chimera has stimulus (signalling pathways) or cell type. been described previously (Littlewood et al., 1995). The cells were routinely cultured in Dulbecco's modi®ed Eagle's medium (DMEM) supplemented with 5% foetal calf serum Mitochondria and caspases (FCS) and antibiotics. For experiments, the cells were allowed to adhere to culture dishes in normal growth media Our ®nding that the mitochondrial transmembrane overnight, washed in serum-free medium, replenished with potential is lost during c-Myc-induced apoptosis and DMEM with antibiotics and 0.5% FCS and cultured for that the disruption of DCm is inhibited by Z-VAD-fmk 24 h. To activate c-Myc and induce apoptosis, 100 nM 4-

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2361 hydroxytamoxifen (Research Biochemicals International, extraction bu€er diluted at a 1 : 3 ratio with 20 mM HEPES, USA) was added to the media and the cells were harvested pH 7.5, 0.3% NP-40). at the indicated times. To assess the role of caspases in the various apoptotic processes the cells were treated with cell In vitro ATM cleavage by caspases penetrative inhibitors N-benzyloxycarbonyl-Val-Ala-Asp- ¯uoromethyl-ketone (Z-VAD-fmk) and N-benzyloxycarbo- Puri®ed recombinant human ATM protein (isolated from the nyl-Asp-Glu-Val-Asp-¯uoromethyl-ketone (Z-DEVD-fmk) 293-EBNA cells as described above) was incubated with (Enzyme System Products, USA) 2 h prior to and during 0.5 mg of the di€erent recombinant active caspases (from the induction of apoptosis. Pharmingen) in a 20 ml reaction mixture containing 25 mM For comparison, the Rat-1 MycERTM cells were also HEPES, pH 7.5, 100 mM NaCl, 1 mM EDTA, 5 mM DTT subjected to apoptosis by treatment with staurosporine and 0.1% CHAPS at 378C for 20 min. The reactions were (Calbiochem; 0.2 mM) in the presence of normal growth stopped by adding Laemmli sample bu€er. The samples were medium (DMEM supplemented with 5% FCS) for the resolved on SDS ± PAGE followed by immunoblotting with indicated times. HL-60 cells were grown in RPMI supple- the anti-ATM AHP392 antibody. mented with 5% FCS and treated with 1 mM staurosporine to induce apoptosis. Measurement of mitochondrial transmembrane potential Cells were harvested by centrifugation and washed twice with Generation of Rat-1 MycERTM fibroblasts stably expressing PBS, incubated at 378C for 15 min in the presence of 40 n human ATM M 3,3'-dihexyloxacarbocyanine iodide [DioC6(3)] (Molecular Rat-1 MycERTM ®broblasts were transfected with 1 mg of the Probes, USA) and washed (Petit et al., 1990; Zamzami et episomal pEBS7-YZ5 expression vector carrying a FLAG al., 1995). As a control, the cells were pre-treated with epitope-tagged, full-length human ATM cDNA and the 100 mM carbonyl cyanide m-chlorophenylhydrazone hygromycin resistance gene (Ziv et al., 1997); kindly provided (mCICCP, Sigma Chemical Co.), an uncoupling agent, for by Dr Y Shiloh, University of Tel Aviv, Israel) and with the 45 min. The FACS analysis was performed with 488 nm empty vector (pEBS7) by using the FUGENETM 6 transfec- excitation and 525 nm emission. tion reagent (Boehringer Mannheim). Stable transfectants were selected in the presence of 200 ± 400 mg/ml of hygro- Subcellular fractionation for analysis of cytochrome c release mycin B for 3 ± 4 weeks. A total pool of the transfectants was from mitochondria used in the experiments. Rat-1 MycERTM cells cultured with or without OHT were collected by centrifugation and suspended in 0.25 ml of the Western blotting and antibodies extraction bu€er (250 mM sucrose, 10 mM Tris-HCl, pH 7.4, Protein analyses by SDS ± PAGE and immunoblotting were 10 mM NaCl, 3 mM MgCl2, 0.5 mM EGTA, 1 mM dithio- performed as described previously (Kangas et al., 1998). threitol, 1 mM AEBSF and 10 mg/ml each of the protease The following antibodies were used for immunoblotting: inhibitors aprotinin, leupeptin and pepstatin). The cells were Anti-Rb G3-245, anti-p53 pAb240, and polyclonal anti- disrupted by homogenization with 20 strokes in a Kontes caspase-9 (Pharmingen, USA); monoclonal anti-gelsolin glass homogenizer (pestle number 21) on ice. The cell lysate (Transduction Laboratories, USA); anti-ATM AHP392 was centrifuged at 700 g for 10 min at 48C to remove nuclei (Serotech, UK); anti-SREBP K-10, anti-MDM2 SMP14, and cell debris, and the resulting supernatant was further anti-p53 N-19, and anti-cPLA2 N-216 (Santa Cruz Bio- centrifuged at 10 000 g for 15 min at 48C. The supernatant technologies, USA); anti-lamin B1 and anti-lamin A+B2 (crude cytosolic fraction) was collected, mixed 5 : 1 with (Zymed Laboratories, USA); polyclonal anti-PKCd (Gibco ± ®vefold Laemmli bu€er and the pellet containing mitochon- BRL); anti-actin A-2668 and monoclonal anti-FLAG M2 dria was resuspended in 50 ml of Laemmli sample bu€er. The antibody (Sigma Chemicals Co.); polyclonal anti-D4-GD1 concentration of proteins was determined by the BCA protein (Na et al., 1996); monoclonal anti-fodrin (NaÈ rvaÈ nen et al., assay reagent (Pierce), the proteins separated by SDS ± 1987); polyclonal anti-Gas2 (Brancolini et al., 1995); anti- PAGE, and the content of cytochrome c analysed by PARP C2-10 (Lamarre et al., 1988); polyclonal anti-RFC immunoblotting. (Luckow et al., 1994); human anti-U1-70 kDa (Casciola- Rosen et al., 1995).

Expression and purification of the recombinant ATM protein from 293 cells Human embryonic kidney 293-EBNA cells were grown in DMEM and 5% FCS with antibiotics. The cells were Note added in proof transfected overnight with 1 mg of the FLAG-tagged, full- After submission of this work, Smith et al. (1999) have also length human ATM expression vector (Ziv et al., 1997) by reported that ATM is a substrate for caspase-3. using the FUGENETM 6 reagent. After culturing for 2 days the cells were harvested, lysed in 250 ml of freshly prepared extraction bu€er (20 mM HEPES, pH 7.5, 50 mM NaF, Acknowledgements 450 mM NaCl, 25% glycerol, 0.2 mM EDTA, 0.5 mM DTT, Drs Dennis Danley, Guy G Poirier, Anthony Rosen, 0.5 mM AEBSF and 10 mg/ml of aprotinin and leupeptin) by Wolfgang Schmid, Claudio Schneider, Donald Nicholson, freezing and thawing twice, and the lysate was centrifuged at Razquallah Hakem, Tak Mak and Ismo Virtanen are 14 000 g for 7 min at 48C to pellet cell debris (for refs see Ziv warmly thanked for providing us with antibodies and Dr et al., 1997). The supernatant was diluted at a 1 : 3 ratio with Yosef Shiloh for the bacterial strains of the pEBS7 and 20 mM HEPES, pH 7.5, 0.3% NP-40 and the recombinant pEBS7-YZ5 vectors. Ms Monica Schoulz is thanked for the ATM protein was anity-puri®ed by incubation of the cell FACS analysis. This work was supported by the University extract with 15 mg of anti-FLAG M2 antibody at 48C of Helsinki, the Finnish Cancer Organizations, the Ida overnight (Ziv et al., 1997). The immunocomplexes were Montin Foundation, the Research and Science Foundation harvested by using agarose-beads with covalently conjugated of Farmos, the Finnish Life and Pension Insurance goat anti-mouse IgG (40 ml) and washed three times (with the Companies and the Finnish Academy of Sciences.

Oncogene Mechanisms of c-Myc-induced apoptosis A Hotti et al 2362 References

Banin S, Moyal L, Shieh S-Y, Taya Y, Anderson CW, Na S, Chuang TH, Cunningham A, Turi TG, Hanke JH, Chessa L, Smorodinsky NI, Prives C, Reiss Y, Shiloh Y Bokoch GM and Danley DE. (1996). J. Biol. Chem., 271, and Ziv Y. (1998). Science, 281, 1674 ± 1677. 11209 ± 11213. Baskaran R, Wood LD, Whitaker LL, Canman CE, Morgan NaÈ rvaÈ nen O, NaÈ rvaÈ nen A, Wasenius VM, Partanen P and SE,XuY,BarlowC,BaltimoreD,Wynshaw-BorisA, Virtanen I. (1987). FEBS Lett., 224, 156 ± 160. Kastan MB and Wang JY. (1997). Nature, 387, 516 ± 519. Nicholson DW and Thornberry NA. (1997). Trends Biol. Bossy-Wetzel E, Newmeyer DD and Green DR. (1998). Sci., 22, 299 ± 306. EMBO J., 17, 37 ± 49. Orth K, Chinnaiyan AM, Garg M, Froelich CJ and Dixit Brancolini C, Benedetti M and Schneider C. (1995). EMBO VM. (1996). J. Biol. Chem., 271, 16443 ± 16446. J., 14, 5179 ± 5190. Packham G and Cleveland JL. (1994). Mol. Cell. Biol., 14, Canman CE, Lim D-S, Cimprich KA, Taya Y, Sakaguchi K, 5741 ± 5747. Appella E, Kastan MB and Siliciano JD. (1998). Science, Petit PX, O'Connor JE, Grunwald D and Brown SC. (1990). 281, 1677 ± 1679. Eur. J. Biochem., 194, 389 ± 397. Canman CE, Wol€ AC, Chen C-Y, Fornace Jr AJ and Pochampally R, Fodera B, Chen L, Lu W and Chen J. (1999). Kastan MB. (1994). Cancer Res., 54, 5054 ± 5058. J. Biol. Chem., 274, 15271 ± 15277. Casciola-Rosen L, Nicholson DW, Chong T, Rowan KR, Prendergast GC. (1999). Oncogene, 18, 2967 ± 2987. Thornberry NA, Miller DK and Rosen A. (1996). J. Exp. Savitsky K, Bar-Shira A, Gilad S, Rotman G, Ziv Y, Med., 183, 1957 ± 1964. Vanagaite L, Tagle DA, Smith S, Uziel T, Sfez S, Casciola-Rosen LA, Anhalt GJ and Rosen A. (1995). J. Exp. Ashkenazi M, Pecker I, Frydman M, Harnik R, Patanjali Med., 182, 1625 ± 1634. SR, Simmons A, Clines GA, Sartiel A, Gatti RA, Chessa Evan GI, Wyllie AH, Gilbert CS, Littlewood TD, Land H, L,SanalO,LavinMF,JaspersNGJ,TaylorAMR,Arlett Brooks M, Waters CM, Penn LZ and Hancock DC. CF, Miki T, Weissman SM, Lovett M, Collins FS and (1992). Cell, 69, 119 ± 128. Shiloh Y. (1995). Science, 268, 1749 ± 1753. Galaktionov K, Chen X and Beach D. (1996). Nature, 382, Shafman T, Khanna KK, Kedar P, Spring K, Kozlov S, Yen 511 ± 517. T, Hobson K, Gatei M, Zhang N, Watters D, Egerton M, Ghayur T, Hugunin M, Talanian RV, Ratnofsky S, Quinlan Shiloh Y, Kharbanda S, Kufe D and Lavin MF. (1997). C,EmotoY,PandeyP,DattaR,HuangY,KharbandaS, Nature, 387, 520 ± 523. Allen H, Kamen R, Wong W and Kufe D. (1996). J. Exp. Shim H, Chun YS, Lewis BC and Dang CV. (1998). Proc. Med., 184, 2399 ± 2404. Natl. Acad. Sci. USA, 95, 1511 ± 1516. Green D and Kroemer G. (1998). Trends Cell Biol., 8, 267 ± Shim H, Dolde C, Lewis BC, Wu CS, Dang G, Jungmann 271. RA, Dalla-Favera R and Dang CV. (1997). Proc. Natl. Green DR and Reed JC. (1998). Science, 281, 1309 ± 1312. Acad.Sci.USA,94, 6658 ± 6663. Herzog KH, Chong MJ, Kapsetaki M, Morgan JI and Smith GC, di Fagagna F, Lakin ND and Jackson SP. (1999). McKinnon PJ. (1998). Science, 280, 1089 ± 1091. Mol. Cell. Biol., 19, 6076 ± 6084. Hueber AO, ZoÈ rnig M, Lyon D, Suda T, Nagata S and Evan Soengas MS, Alarcon RM, Yoshida H, Giaccia AJ, Hakem GI. (1997). Science, 278, 1305 ± 1309. R, Mak TW and Lowe SW. (1999). Science, 284, 156 ± 159. Kagaya S, Kitanaka C, Noguchi K, Mochizuki T, Sugiyama Song Q, Lees-Miller SP, Kumar S, Zhang Z, Chan DW, A, Asai A, Yasuhara N, Eguchi Y, Tsujimoto Y and Smith GC, Jackson SP, Alnemri ES, Litwack G, Khanna Kuchino Y. (1997). Mol. Cell. Biol., 17, 6736 ± 6745. KK and Lavin MF. (1996). EMBO J., 15, 3238 ± 3246. Kangas A, Nicholson DW and HoÈ lttaÈ E. (1998). Oncogene, Stennicke HR, Deveraux QL, Humke EW, Reed JC, Dixit 16, 387 ± 398. VM and Salvesen GS. (1999). J. Biol. Chem., 274, 8359 ± Kastan MB, Zhan Q, El-Deiry WS, Carrier F, Jacks T, 8362. Walsh WV, Plunkett BS, Vogelstein B and Fornace Jr AJ. Thompson EB. (1998). Annu. Rev. Physiol., 60, 575 ± 600. (1992). Cell, 71, 587 ± 597. Tsunoda H, Terasawa T, Yageta M, Nakajima T, Tomooka Kroemer G, Zamzami N and Susin SA. (1997). Immunol. Y, Tsuchida N and Oda K. (1999). Biochem. Biophys. Res. Today, 18, 44 ± 51. Comm., 255, 722 ± 730. Lamarre D, Talbot B, de Murcia G, Laplante C, Leduc Y, Wagner AJ, Kokontis JM and Hay N. (1994). Genes Dev., 8, Mazen A and Poirier GG. (1988). Biochim. Biophys. Acta, 2817 ± 2830. 950, 147 ± 160. Xiao Q, Claassen G, Shi J, Adachi S, Sedivy J and Hann SR. Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, (1998). Genes Dev., 12, 3803 ± 3808. Alnemri ES and Wang X. (1997). Cell, 91, 479 ± 489. Zamzami N, Marchetti P, Castedo M, Zanin C, Vayssiere JL, Littlewood TD, Hancock DC, Danielan PS, Parker MG and Petit PX and Kroemer G. (1995). J. Exp. Med., 181, Evan GI. (1995). Nucl. Acids. Res., 23, 1686 ± 1690. 1661 ± 1672. Lotem J and Sachs L. (1995). Proc. Natl. Acad. Sci. USA, 92, Zhang N, Chen P, Gatei M, Khanna KK and Lavin MF. 9672 ± 9676. (1998). Oncogene, 17, 811 ± 818. Luckow B, Bunz F, Stillman B, Lichter P and Schutz G. Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, (1994). Mol. Cell. Biol., 14, 1626 ± 1634. Sherr CJ and Roussel MF. (1998). Genes Dev., 12, 2424 ± McCarthy NJ, Whyte MK, Gilbert CS and Evan GI. (1997). 2433. J. Cell Biol., 136, 215 ± 227. Ziv Y, Bar-Shira A, Pecker I, Russell P, Jorgensen TJ, Meyn MS. (1995). Cancer Res., 55, 5991 ± 6001. Tsarfati I and Shiloh Y. (1997). Oncogene, 15, 159 ± 167. Morgan SE and Kastan MB. (1997). Adv. Cancer Res., 71, Zou H, Henzel WJ, Liu X, Lutschg A and Wang X. (1997). 1±25. Cell, 90, 405 ± 413.

Oncogene