Ionizing radiation but not anticancer drugs causes cell cycle arrest andfailure to activate the mitochondrial death pathway in MCF-7 breastcarcinoma cells

Reiner U JaÈ nicke*,1, Ingo H Engels1, Torsten Dunkern2, Bernd Kaina2, Klaus Schulze-Ostho�1,4

and Alan G Porter3,4

1Department of Immunology and Cell Biology, University of MuÈnster, RoÈntgenstrasse 21, 48149 MuÈnster, Germany; 2Institute ofToxicology, University of Mainz, Obere Zahlbacher Str. 67, 55131 Mainz, Germany; 3Institute of Molecular and Cell Biology, 30Medical Drive, Singapore 117609, Republic of Singapore

There is considerable evidence that ionizing radiation(IR) and chemotherapeutic drugs mediate apoptosisthrough the intrinsic death pathway via the release ofmitochondrial cytochrome c and activation of caspases -9and -3. Here we show that MCF-7 cells that lackcaspase-3 undergo a caspase-dependent apoptotic celldeath in the absence of DNA fragmentation and a-fodrincleavage following treatment with etoposide or doxor-ubicin, but not after exposure to IR. Re-expression ofcaspase-3 restored DNA fragmentation and a-fodrincleavage following drug treatment, but it did not alterthe radiation-resistant phenotype of these cells. Incontrast to the anticancer drugs, IR failed to inducethe intrinsic death pathway in MCF-7/casp-3 cells, anevent readily observed in IR-induced apoptosis of HeLacells. Although IR-induced DNA double-strand breakswere repaired with similar e�ciencies in all cell lines, cellcycle analyses revealed a persistent G2/M arrest in thetwo MCF-7 cell lines, but not in HeLa cells. Together,our data demonstrate that caspase-3 is required for DNAfragmentation and a-fodrin cleavage in drug-inducedapoptosis and that the intrinsic death pathway is fullyfunctional in MCF-7 cells. Furthermore, they show thatthe radiation-resistant phenotype of MCF-7 cells is notdue to the lack of caspase-3, but is caused by the failureof IR to activate the intrinsic death pathway. Wepropose (1) di�erent signaling pathways are induced byanticancer drugs and IR, and (2) IR-induced G2/Marrest prevents the generation of an apoptotic signalrequired for the activation of the intrinsic death pathway.Oncogene (2001) 20, 5043 ± 5053.

Keywords: apoptosis; intrinsic death pathway; DNAdamage; cell cycle arrest

Introduction

Programmed cell death (apoptosis) can be triggered bya variety of stimuli via two principal signalingpathways, both of which depend on the formationof multimeric protein complexes and subsequentactivation of death proteases, called caspases (Crynsand Yuan, 1998; Thornberry and Lazebnik, 1998).The extrinsic death pathway involves the ligation ofdeath receptors (Fas/CD95; tumor necrosis factorreceptor I/CD 120a) that leads to the recruitment ofthe adaptor molecule Fas-associated death domainand pro-caspase-8 into a death-inducing signalingcomplex (DISC) (Kischkel et al., 1995; Muzio et al.,1996). The intrinsic death pathway is initiated at themitochondrion by the release of cytochrome c thattogether with dATP and Apaf-1 binds to pro-caspase-9 to form the apoptosome (Cryns and Yuan, 1998; Liet al., 1997b). Upon formation of the DISC or theapoptosome, pro-caspases-8 or -9, respectively, areautoproteolytically processed resulting in the activa-tion of downstream caspases such as caspases-3, -6and -7 (Srinivasula et al., 1996; Slee et al., 1999).Although both pathways are linked through thecaspase-8-mediated cleavage of the pro-apoptotic Bcl-2-related protein Bid that triggers cytochrome crelease, DNA-damaging agents are believed to mediateapoptosis via mitochondrial cytochrome c releaseindependent of the death receptor pathway (Ferrariet al., 1998; Wesselborg et al., 1999; Belka et al.,1999). The signals, however, leading to cytochrome crelease are not understood.

Fourteen distinct mammalian caspases have beenidenti®ed that make a limited number of strategic cutsin a variety of key cellular proteins, resulting either ininhibition or deregulation of their function (Nicholsonand Thornberry, 1997; Porter et al., 1997; Van deCraen et al., 1998). Although these proteases can beplaced into three groups on the basis of their in vitrosubstrate preferences (Nicholson and Thornberry,1997; Talanian et al., 1997), little is known of their invivo substrate speci®cities. The majority of knowncaspase substrates are cleaved by caspase-3 or a relatedprotease, including important proteins like gelsolin,

Oncogene (2001) 20, 5043 ± 5053ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00

www.nature.com/onc

*Correspondence: RU JaÈ nicke, E-mail: r.janicke@uni-muenster.de4Both authors share equal senior authorshipReceived 19 March 2001; revised 9 May 2001; accepted 23 May2001

protein kinase C-d, p21-activated kinase-2, a-fodrin,the retinoblastoma protein, and DNA fragmentationfactor, whose proteolysis most likely contributes toapoptosis in various ways (Nicholson and Thornberry,1997).

Some tumor cells readily undergo apoptosis whenexposed to ligands of the TNF1 family (such as TNF-a and CD95L), anticancer drugs or ionizing radiation(IR), while others do not; but the molecular basis ofthese di�erences is poorly understood (Wertz andHanley, 1996; Hannun, 1997; Ashkenazi and Dixit,1998). Many tumor cell lines harboring the casp-3gene are resistant to TNF-induced apoptosis. Incontrast, MCF-7 breast carcinoma cells, which havelost caspase-3 owing to a genomic deletion in thecasp-3 gene, are still sensitive to TNF or staurospor-ine, demonstrating the existence of caspase-3-indepen-dent death pathway(s) induced by these agents(JaÈ nicke et al., 1998a,b). However, caspase-3 isrequired for cell death resulting from microinjectedcytochrome c (Li et al., 1997a) and for TNF-inducedapoptotic DNA fragmentation and membrane bleb-bing (JaÈ nicke et al., 1998b), which are classic hall-marks of apoptosis.

The cellular response to DNA damage includes theinduction of apoptosis (Hannun, 1997; Fisher, 1994;Wang, 1998), but the failure to respond properly toDNA damage can lead to genetic alterations thatpromote tumor progression. The tumor suppressorp53 plays a major role in modulating the apoptoticresponse following DNA damage that is eitherdependent or independent of its transcriptionalactivity (Ko and Prives, 1996; Levine, 1997). Loss ofp53 function is observed in over half of all humancancers (Hollstein et al., 1991), and in many cases thesusceptibility of tumor cells to anticancer drugs or IRdepends on their p53 status (Clarke et al., 1993; Fanet al., 1994). However, some tumor cells withfunctional p53 are relatively resistant to anticancerdrugs or IR (Zhan et al., 1994; Fan et al., 1995).MCF-7 belongs to such a class of tumor cell(Levenson and Jordan, 1997), which harbors a wild-type p53 gene and is relatively resistant to p53-dependent apoptosis induced by DNA-damagingagents (Zhan et al., 1994; Fan et al., 1995). Weinvestigated more closely the role of caspase-3 inapoptosis induced by DNA-damaging agents, becausecaspases can act downstream of p53 (Sabbatini et al.,1997; Schuler et al., 2000), and caspase-3-likeproteases may play a role in DNA damage-inducedapoptosis (Dubrez et al., 1996; Faleiro et al., 1997;Fuchs et al., 1997; Martins et al., 1997; Simizu et al.,1998; Yu and Little, 1998).

In this study, we address the related questions of (i)whether caspase-3 is required for apoptosis induced byDNA-damaging agents; (ii) whether caspase-3 isessential for DNA fragmentation and speci®c cleavageof a-fodrin induced by DNA-damaging agents; and (iii)whether di�erent DNA-damaging agents utilize thesame apoptotic pathway leading to the activation ofthe intrinsic death pathway.

Results

Etoposide and doxorubicin, but not IR induce apoptosis inMCF-7 cells

Using the human MCF-7 breast carcinoma cell linethat is devoid of caspase-3, we and others havepreviously demonstrated that caspase-3 is not essentialfor TNF-, Fas- or staurosporine-induced apoptotic celldeath (JaÈ nicke et al., 1998a,b; Li et al., 1997a). Toinvestigate whether these caspase-3-de®cient cells mayhave acquired a selective growth advantage in thepresence of DNA-damaging agents, MCF-7 cells wereexposed to the anticancer drugs etoposide anddoxorubicin or IR. As a control, caspase-3-expressingHeLa H21 cells were treated in a similar fashion.Etoposide (100 mM) or doxorubicin (1 mg/ml) treatmentof HeLa and MCF-7 cells resulted in a time-dependentdecrease of cell viability as measured with the standardcrystal violet assay (JaÈ nicke et al., 1994; Figure 1a,b).Similar results were obtained when cell death wasassessed by the trypan blue exclusion test (data notshown). The rate of this decrease was slower in MCF-7cells, which were slightly more resistant than HeLacells to cell death induced by these agents 3 days after

Figure 1 Etoposide or doxorubicin, but not IR induce apoptosisin caspase-3-de®cient MCF-7 cells. Cytotoxicity assays of HeLaH21 cells (circles) and MCF-7 cells (squares) treated for theindicated times with etoposide (100 mM) (a) or doxorubicin (1 mg/ml) (b), or exposed to IR (20 Gy) (c) followed by incubation withgrowth medium. Cell death was assessed using the crystal violetassay as described (JaÈ nicke et al., 1994). Similar results wereobtained when cell death was assessed by trypan blue uptake(data not shown). The values are derived from one representativeexperiment of four performed in triplicates. (d) HeLa H21 cells(circles) and MCF-7 cells (squares) were exposed to the indicatedradiation doses in the absence (open symbols) or presence (®lledsymbols) of zVAD-fmk (50 mM). After 3 days, cell death wasassessed by propidium iodide uptake and subsequent FACSanalyses

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5044

Oncogene

treatment (Figure 1). Anticancer drug-induced killingof both cell lines was inhibitable by the broad spectrumcaspase inhibitor zVAD-fmk (data not shown) demon-strating that even MCF-7 cells lacking caspase-3 dievia a caspase-dependent apoptotic pathway. This resultis in agreement with our previous studies showing thatTNF or staurosporine induce apoptosis in caspase-3-de®cient MCF-7 cells (JaÈ nicke et al., 1998a,b).

In contrast to the anticancer drugs, we observed adramatic di�erence in apoptosis sensitivity when HeLaand MCF-7 cells were exposed to IR. A single dose of20 Gy caused the death of 75% HeLa cells after 3days, whereas the same treatment resulted in onlymarginal killing of MCF-7 cells (Figure 1c). Cell deathmeasurement by the dye exclusion test con®rmed theseresults (data not shown). Various radiation dosesranging from 1 to 10 Gy resulted in the dose-dependentdeath of HeLa cells which was partially inhibitable byzVAD-fmk, but did not alter the resistance of MCF-7cells to IR (Figure 1d). Together, these data suggestthat the radiation-resistant phenotype of MCF-7 cellsmight be caused, at least in part, by the lack ofcaspase-3.

Anticancer drugs and IR induce the caspase-3-dependentevents of DNA fragmentation and a-fodrin cleavage inHeLa, but not in caspase-3-deficient MCF-7 cells

Caspase-3 is activated by various death stimuliincluding DNA-damaging agents (Dubrez et al., 1996;Fuchs et al., 1997; Simizu et al., 1998). To investigatewhether caspase-3 is activated in HeLa cells under-going DNA damage-induced apoptosis, Western blotanalyses were performed. A consistent decrease in pro-caspase-3 (indicative of the cleavage activation ofcaspase-3) was not detectable in etoposide- or doxor-ubicin-treated or irradiated HeLa cells (Figure 2a,upper panel), which may be due to the upregulation ofthe casp-3 gene induced by DNA-damaging agents andthe de novo synthesis of pro-caspase-3 (Droin et al.,1998). Nevertheless, all these DNA-damaging agentsgave rise to active caspase-3 fragments as judged byWestern blotting (Figure 2b) and by the DEVD-speci®c¯uorogenic substrate assay (Figure 2c). As caspase-3 isrequired for DNA fragmentation and cleavage of a-fodrin in TNF-or staurosporine-induced apoptosis ofMCF-7 cells (JaÈ nicke et al., 1998a,b), we examinedwhether caspase-3 is also required for the induction ofthese events in apoptosis induced by DNA-damagingagents. FACS analysis con®rmed that apoptosis ofHeLa cells induced by etoposide, doxorubicin, or IRcorrelated well with extensive DNA fragmentation asshown by the time-dependent increase in the apoptoticsub-G1 peak in treated cells (Figure 2d). Usingconventional agarose gel electrophoresis, the typicalapoptotic DNA laddering was observed when HeLacells were exposed to etoposide, doxorubicin or IR(data not shown). In addition, all these treatmentsresulted in the e�cient caspase-3-dependent cleavage ofa-fodrin into the typical 120 kDa fragment in thesecells (Figure 2a, lower panel). Together, these results

demonstrate that caspase-3 is processed and fullyfunctional in HeLa cells treated with DNA-damagingagents.

Unlike in HeLa cells, a-fodrin was not cleaved intothe 120 kDa fragment when caspase-3-de®cient MCF-7cells were exposed to anti-cancer drugs or IR (Figure3a). Moreover, neither DEVD activity (Figure 3b) norDNA fragmentation (Figure 3c and data not shown)were detectable in response to these treatments. Hence,our results indicate that, like apoptosis induced byTNF or staurosporine (JaÈ nicke et al., 1998a,b), theseevents are also not required for the anticancer drug-induced apoptotic death of MCF-7 cells. However, thecaspase substrate poly(ADP-ribose)polymerase (PARP)was e�ciently cleaved following treatment of MCF-7cells with etoposide or doxorubicin, but not when thecells were exposed to IR (Figure 3d). Drug-induced celldeath (data not shown) and PARP cleavage (Figure3d) were blocked by zVAD-fmk, demonstrating thatcaspase-3-de®cient MCF-7 cells die by apoptosis,although some of the classic apoptotic hallmarks aremissing.

Caspase-3 expression in MCF-7 cells restores a-fodrincleavage and DNA fragmentation in response to theanticancer drugs, but not to IR

As HeLa cells, unlike MCF-7 cells, are susceptible toIR-induced killing and undergo classical alterations ofapoptosis, we next examined whether the lack ofcaspase-3 is responsible for the radiation-resistantphenotype of MCF-7 cells. For this purpose, MCF-7cells stably expressing caspase-3 (clone MCF-7.3.28;JaÈ nicke et al., 1998b) were analysed for their sensitivityto DNA-damaging agents. As judged by crystal violetstaining (Figure 4a), trypan blue or propidium iodideuptake (data not shown), re-expression of caspase-3 didnot signi®cantly enhance the sensitivity of MCF-7 cellsto drug treatment. Surprisingly, expression of caspase-3had also no e�ect on the susceptibility to IR treatmentas MCF-7/casp-3 cells remained as radiation-resistantas parental MCF-7 cells. In agreement with theseresults, microscopic examination revealed that onlyHeLa cells exposed to IR, but not similarly treatedMCF-7 or MCF-7/casp-3 cells displayed the typicalapoptotic morphology (data not shown).

Similar to the results obtained with HeLa cells(Figure 2), etoposide- or doxorubicin-induced apopto-sis of MCF-7/casp-3 cells resulted in the generation ofactive caspase-3 fragments as shown by Westernblotting (Figure 4b, lower panel) and by the DEVD-speci®c ¯uorogenic substrate assay (Figure 4c). Inaddition, cancer drug-induced caspase-3 activationresulted in cleavage of a-fodrin into the typical120 kDa fragment (Figure 4b, upper panel), and inDNA fragmentation as evidenced by the appearance ofthe apoptotic sub G1 peak in FACS analyses (Figure4d) and by agarose gel electrophoresis (data notshown). These results clearly demonstrate that cas-pase-3 is also required for these events in anticancerdrug-induced apoptosis. In contrast, exposure of the

Oncogene

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5045

cells to IR did not result in caspase-3 activation(Figure 4b, lower panel; Figure 4c), a-fodrin cleavage(Figure 4b, upper panel) or DNA fragmentation(Figure 4d and data not shown). These results weresimilar to those obtained with irradiated parentalMCF-7 cells (Figure 3). Together, our data demon-strate that the pathway to caspase-3 activation is fullyfunctional in MCF-7/casp-3 cells treated with etoposideor doxorubicin, but is not triggered when the cells areexposed to IR.

Anticancer drugs, but not IR induce cytochrome c releaseand caspase-9 activation in casp-3-transformed MCF-7cells

Although the exact mechanism of apoptosis induced byDNA-damaging agents is presently unknown, therelease of cytochrome c from mitochondria andsubsequent activation of caspase-9 and caspase-3appear to be essential (Hakem et al., 1998; Kuida etal., 1998; Chen et al., 2000). To determine in moredetail the reason for the inability of IR to activatecaspase-3 in MCF-7/casp-3 cells, we monitored therelease of cytochrome c from mitochondria. In HeLacells cytosolic cytochrome c levels were greatlyenhanced after exposure to etoposide, doxorubicin or

IR (Figure 5a). In addition, pro-caspase-9 processingwas detected in HeLa cells exposed to all these DNA-damaging agents (data not shown) which is inagreement with the e�cient activation of caspase-3(Figure 2). In MCF-7/casp-3 cells, however, only theanticancer drugs etoposide or doxorubicin, but not IRinduced cytochrome c release (Figure 5b) and pro-caspase-9 processing (Figure 5c). Taken together, ourdata demonstrate that the intrinsic death pathwayinduced by anticancer drugs is fully functional inMCF-7/casp-3 cells, in contrast to the death pathwayinduced by IR, which appears to be defective in thesecells presumably at a point upstream of mitochondrialcytochrome c release.

IR-induced DNA double-strand breaks result in G2/Marrest of the MCF-7 cell lines

The initial event following exposure to IR is thegeneration of DNA double-strand breaks (Wang,1998). To analyse whether DNA double-strand breaksare generated in MCF-7 cells following exposure to IR,the neutral single-cell gel electrophoresis system (cometassay) was employed. As shown in Figure 6a, IRinduced signi®cant DNA double-strand breaks in allcell lines including the two radiation-resistant MCF-7

Figure 2 DNA-damaging agents induce caspase-3 activation, cleavage of a-fodrin and DNA fragmentation in HeLa H21 cells. (a)Western blot analysis of the status of pro-caspase-3 (upper panel) and a-fodrin (lower panel) in HeLa cells. Cells were either leftuntreated (lane 1), exposed to 20 Gy followed by incubation with medium for 2 days (lane 2) or 3 days (lane 3), or treated for 2days with 100 mM etoposide (lane 4) or 1 mg/ml doxorubicin (lane 5). One representative experiment of three is shown. (b) Westernblot analysis of active caspase-3 in HeLa cells. Cells were either left untreated (lane 1), or were exposed for the indicated times toetoposide (lanes 2 and 3), doxorubicin (lanes 4 and 5) or IR (lanes 6 and 7). (c) E�ect on caspase-3-like activity. Cell lysates ofHeLa cells treated for the indicated times with etoposide (squares), doxorubicin (triangles) or IR (circles) were incubated with the¯uorogenic substrate DEVD-Amc. The catalytic activities are given in arbitrary units (AU). DEVDase activity of untreated HeLacells was 537 AU. (d) FACS analysis of HeLa H21 cells either left untreated (control) or treated for the indicated number of days(d) with etoposide (100 mM) or doxorubicin (1 mg/ml) or exposed to 20 Gy followed by incubation with growth medium. The variouscell cycle phases and the apoptotic sub-G1 peak are indicated. One representative experiment of three is shown

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5046

Oncogene

lines. In addition, these DNA double-strand breaksdecreased to the same extent and with similar kineticsin all three cell lines indicating no di�erences in theavailability and functionality of the DNA repairmachinery between these cells.

Although the IR-induced DNA double-strand breaksdid not induce the intrinsic death pathway in MCF-7or MCF-7/casp-3 cells, the possibility remained thatthese cells die via the necrotic pathway. Therefore weanalysed the release of cytosolic lactate dehydrogenase(LDH), another measurement for cell death, into thesupernatant from irradiated cells. As shown in Figure6b, LDH activity was only detected in supernatants ofIR-treated HeLa cells, but not in supernatants ofsimilar treated MCF-7 or MCF-7/casp-3 cells. Theseresults are in agreement with the data obtained usingseveral other methods of cell death detection includingcrystal violet (Figures 1 and 4a), DNA fragmentation(Figures 3c and 4d), trypan blue exclusion, orpropidium iodide uptake (Figure 1 and data notshown), thus con®rming the radiation-resistant pheno-type of MCF-7 and MCF-7/casp-3 cells.

Cells exposed to IR arrest in various phases of thecell cycle in order to repair damaged DNA. Dependingon the extent of the damage, cells either enter mitosisor die via apoptosis. Our previous experimentsindicated that, in contrast to radiation-sensitive HeLacells (Figure 2d), radiation-resistant MCF-7 cells(Figure 3c) and MCF-7/casp-3 cells (Figure 4d)remained in G2/M following exposure to IR. Toexamine these di�erences in more detail, we determinedthe percentages of cells in each phase of the cell cyclefor a period of up to 4 days following IR treatment. Inagreement with our previous data, we found that onlyHeLa cells succumb to apoptosis 2 days post IR with aprior accumulation in G2/M at day 1 (Figure 6c). Incontrast, MCF-7/casp-3 cells accumulated and re-mained in G2/M over the 4-day period, and noapoptotic cells were detected (Figure 6d). Similarresults were obtained with the parental MCF-7 cells(Figure 3c and data not shown) indicating that thisevent might be the reason for the failure of IR toactivate the intrinsic death pathway in MCF-7 breastcarcinoma cells.

Discussion

Chemotherapy and radiation are important treatmentmodalities for many cancers, but the frequent occur-rence of drug- and radiation-resistant tumors is acommon clinical problem. Many di�erent mechanismscan account for poor patient prognosis and treatmentfailure including the loss or mutation of pro-apoptoticgenes that regulate the intrinsic death pathway such asp53 or Apaf-1 (Lowe et al., 1993; Schmitt and Lowe,1999; Soengas et al., 1999, 2001). Several investigatorshave also proposed a crucial role for caspase-3 in DNAdamage-induced apoptosis, as in various tumor cellsthis protease is frequently activated during apoptosisinduced by anticancer drugs and ionizing radiation

Figure 3 DNA-damaging agents do not induce DEVD activity,a-fodrin cleavage or DNA fragmentation in caspase-3-de®cientMCF-7 cells. (a) Western blot analysis of the status of a-fodrin inuntreated cells (lane 1), or in cells treated for the indicatednumber of days with etoposide (lanes 2 ± 4), or doxorubicin (lanes5 ± 7), or in cells exposed to IR, followed by incubation withmedium for the indicated number of days (lanes 8 ± 10). Onerepresentative experiment of three is shown. (b) E�ect on caspase-3-like activity. Cell lysates of MCF-7 cells treated for theindicated times with etoposide (squares), doxorubicin (triangles)or IR (circles) were incubated with the ¯uorogenic susbstrateDEVD-Amc. The catalytic activities are given in arbitrary units(AU). DEVDase activity of untreated MCF-7 cells was 320 AU.(c) FACS analysis of MCF-7 cells either left untreated (control)or treated for the indicated number of days (d) with etoposide(100 mM) or doxorubicin (1 mg/ml) or exposed to 20 Gy followedby incubation with growth medium. The various cell cycle phasesare indicated. One representative experiment of three is shown. (d)Western blot analysis of PARP cleavage in MCF-7 cells. Cellswere either left untreated (lane 1), or were exposed for theindicated times to etoposide (lanes 2 ± 4), doxorubicin (lanes 5 ± 7)or IR (lanes 8 ± 10) in the absence (lanes 1 ± 3, 5 ± 6, 8 ± 9) orpresence (lanes 4, 7 and 10) of zVAD-fmk (50 mM)

Oncogene

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5047

(Dubrez et al., 1996; Fuchs et al., 1997; Martins et al.,1997; Yu and Little, 1998; Datta et al., 1997). As theinitiation of caspase-3-dependent DNA fragmentationand cleavage of several death substrates will surely leadto cell death, it was reasonable to speculate that thefunctional deletion of the casp-3 gene may be oneimportant event leading to apoptosis resistance and to

tumorigenesis (JaÈ nicke et al., 1998b). However, basedon our present and previous studies, MCF-7 cells canbe killed via apoptosis by a variety of apoptotic stimuliincluding anticancer drugs in the absence of caspase-3(JaÈ nicke et al., 1998a,b; Li et al., 1997a). Together withthe fact that a de®ciency of caspase-3 has not beenreported for primary tumors, our results suggest that

Figure 4 Re-expression of caspase-3 does not alter susceptibility of MCF-7/casp-3 cells to DNA-damaging agents. (a) Cytotoxicityassays of MCF-7 cells (open circles) and MCF-7/casp-3 cells (®lled circles) treated for the indicated times with etoposide ordoxorubicin or exposed to IR followed by incubation with growth medium. Cell death was assessed using the crystal violet assay asdescribed (JaÈ nicke et al., 1994). Similar results were obtained when cell death was assessed by trypan blue uptake (data not shown).The values are derived from one representative experiment of three performed in triplicates. (b) Western blot analysis of the statusof a-fodrin (upper panel), pro-caspase-3 (middle panel) and active caspase-3 (lower panel) in MCF-7/casp-3 cells is shown. Cellswere either left untreated (lane 1), treated with etoposide (lanes 2 ± 4), or doxorubicin (lanes 5 ± 7), or were exposed to IR followedby incubation with growth medium for the indicated number of days (lanes 8 ± 10). One representative experiment of four is shown.(c) E�ect on caspase-3-like activity. Cell lysates of MCF-7/casp-3 cells treated for 1 ± 3 days with etoposide (squares), doxorubicin(triangles), or exposed to IR (circles) followed by incubation with growth medium for 1 ± 3 days were incubated with the ¯uorogenicsubstrate DEVD-Amc. The catalytic activities are given in arbitrary units (AU). DEVDase activity of untreated MCF-7/casp-3 cellswas 430 AU. (d) FACS analysis of MCF-7/casp-3 cells treated as in b. The various cell cycle phases and the apoptotic sub-G1 peakare indicated. One representative experiment of ®ve is shown

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5048

Oncogene

the lack of caspase-3 by itself may not be the majorcause for the frequent occurrence of chemoresistanttumor cells.

Here we made the observation that in contrast to theanticancer drugs etoposide and doxorubicin, IR fails toinduce the apoptotic death of MCF-7 cells. It has beenreported that formation of free radicals might manip-ulate cell death pathways, diverting the cell's responseto a noxious stimulus from apoptosis toward necrosis(Lee and Shacter, 1999). However, using severaldi�erent techniques for the detection of apoptotic ornecrotic cell death including DNA fragmentation,caspase activation, a dye exclusion assay, crystal violetassay, propidium iodide uptake, MTT assay and LDHrelease, we could not detect any signi®cant death ofIR-treated MCF-7 or MCF-7/casp-3 cells, clearlydemonstrating the IR resistance of these cells. Accord-ing to our present data, the radiation-resistantphenotype of MCF-7 cells can be readily explained

by the failure of IR to trigger the intrinsic deathpathway in these cells upstream of cytochrome crelease and the activation of caspases-9 and -3. Ascytochrome c is released and caspases-9 and -3 areactivated in IR-treated HeLa cells (this study) andseveral other cell types (Fuchs et al., 1997; Yu andLittle, 1998; Datta et al., 1997), our results suggest thatthe apoptotic death pathway induced by IR is defectivein MCF-7 cells upstream of mitochondrial cytochromec release.

Agents known to induce apoptosis via the intrinsicdeath pathway such as the anticancer drugs etoposideand doxorubicin or the protein kinase inhibitorstaurosporine e�ciently kill MCF-7/casp-3 cells viathe intrinsic death pathway resulting in cytochrome crelease and activation of caspases -9 and -3 (JaÈ nicke etal., 1998b; Engels et al., 2000; this study). Evencaspase-3-de®cient MCF-7 cells undergo caspase-de-pendent apoptosis following treatment with etoposideor doxorubicin as both, cell death and PARP cleavagecould be blocked by zVAD-fmk. Thus, MCF-7 cells donot exhibit a general defect in the intrinsic apoptoticpathway. Our results rather suggest that anticancerdrugs and IR utilize di�erent signaling pathwaysupstream of mitochondrial cytochrome c release. Itseems surprising that two classes of DNA-damagingagents that are known to induce apoptosis via DNAdouble-strand breaks should not utilize the same deathpathway upstream of mitochondria. In support of ourconclusion, however, are several reports demonstratingthat, in contrast to the frequent occurrence of cross-resistance between various anti-cancer drugs, cross-resistance between cytotoxic drugs and ionizing radia-tion is relatively rare (Lehnert et al., 1989; Oshita et al.,1992; Heenan et al., 1996).

What are the components involved in the radiation-resistant phenotype of MCF-7 cells? The release ofcytochrome c is regulated by pro- and anti-apoptoticmembers of the Bcl-2 family such as Bax and Bcl-XL,respectively, and the deregulated expression of theseproteins is known to result in aberrant apoptoticresponses (Vander Heiden and Thompson, 1999).However, no signi®cant di�erences in the constitutiveand DNA damage-inducible expression levels of theseproteins were observed in MCF-7 cells compared toHeLa cells (data not shown). Alternatively, IR mightinduce an extremely e�cient DNA repair system inMCF-7 cells, thus preventing the generation of anapoptotic signal. Our data obtained with the cometassay, however, argue against this possibility, as theDNA double-strand breaks that are induced by IR tothe same extent in HeLa and in MCF-7 cells are alsorepaired with similar e�ciencies in both cell lines.

Might the G2/M cell cycle arrest observed only inMCF-7 cells but not in HeLa cells provide a clue to themechanism of IR resistance in MCF-7 cells? Incontrast to HeLa cells, MCF-7 cells express afunctional p53 gene. p53 plays a pivotal role inregulating a checkpoint in the G1 phase of the cellcycle, and it is also required for IR-induced G2/Marrest (Bunz et al., 1998; Sionov and Haupt, 1999).

Figure 5 Etoposide or doxorubicin, but not IR, inducemitochondrial cytochrome c release and pro-caspase-9 processingin MCF-7/casp-3 cells. Western blot analyses of cytosoliccytochrome c in HeLa cells (a), and in MCF-7/casp-3 cells (b),treated for the indicated times with etoposide or doxorubicin, orexposed to IR followed by incubation with growth medium. In (b)the upper part of the blot is shown as a loading control. (c)Western blot analysis of caspase-9 in MCF-7/casp-3 cells. Cellswere either left untreated or treated for the indicated number ofdays with etoposide or doxorubicin, or exposed to IR followed byincubation with growth medium. Pro-caspase-9 and cleavedcaspase-9 are indicated

Oncogene

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5049

Whereas the cyclin-dependent kinase inhibitor p21/waf1/cip1 contributes to both G1 and G2/M arrestfollowing radiation treatment, the p53-dependentinduction of 14-3-3 sigma and GADD45 selectivelymediate the G2/M arrest via interference with apathway that is controlled by the ATM and chkkinases (Sionov and Haupt, 1999). Our ®nding thatirradiated MCF-7 cells (but not HeLa cells) arrestpreferentially in G2/M is remarkable, because itindicates that there is a selective defect in the p53-dependent G2/M but not G1 checkpoint in these cells.We propose that the preferential G2/M arrest inirradiated MCF-7 cells prevents the generation of anas yet unknown apoptotic signal required for theactivation of the intrinsic death pathway.

In conclusion, our results demonstrate that caspase-3, although not essential for the process of apoptosisitself, is required for cancer drug-induced DNAfragmentation and a-fodrin cleavage in MCF-7 cells.

Caspase-3 has now been shown to be indispensable forthese hallmarks of apoptosis induced by numerousdeath stimuli and in every cell type examined (JaÈ nickeet al., 1998a,b; Woo et al., 1998; Zheng et al., 1998;Porter and JaÈ nicke, 1999; Kuida et al., 1996). As IRfails to induce the mitochondrial death pathway,including activation of caspase-3 in MCF-7/casp-3cells, the question of whether caspase-3 contributes toIR-induced apoptosis cannot be properly answered yet.However, our results suggest that the radiation-resistant phenotype of MCF-7 breast carcinoma cellsis not caused by the functional deletion of the casp-3gene or any other gene of the intrinsic death pathway,but may be due to the IR-induced G2/M arrest thatprevents the generation of an apoptotic signal requiredfor the activation of this pathway. Our results alsoindicate the requirement of di�erent initiating eventsfor apoptosis pathways induced by either drugtreatment or radiation.

Figure 6 IR-induced DNA double-strand breaks do not result in cell death, but in G2/M arrest in MCF-7 cells. (a) The DNAdouble-strand breaks were determined following exposure to IR (0 h) at the time points indicated by the comet assay. Data are themeans of three independent experiments. Bars, s.d. (b) Supernatants of cells were collected at the indicated time points post IR andassayed for LDH activity. One representative experiment of two is shown. (c,d) Cell cycle analyses of HeLa (c) and MCF-7/casp-3cells (d) exposed to IR followed by incubation with growth medium for the indicated times. Cell viability is expressed as the sum ofcells in the three cell cycle phases (G1, S and G2/M) from which apoptotic cells in the sub G1 phase are substracted. Onerepresentative experiment of three is shown

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5050

Oncogene

Materials and methods

Cell lines, reagents and antibodies

HeLa H21 cervical carcinoma cells (JaÈ nicke et al., 1994),MCF-7 breast carcinoma cells and MCF-7/casp-3 cells stablyexpressing caspase-3 (JaÈ nicke et al., 1998a,b) were maintainedin RPMI 1640, supplemented with 10% FCS, 10 mM

glutamine, and 50 mg (each) of streptomycin and penicillin/ml. The protease inhibitors aprotinin, bacitracin, antipain,leupeptin and phenylmethylsulfonyl ¯uoride as well asetoposide and doxorubicin were purchased from Sigma. Thebroad range caspase inhibitor zVAD-fmk was purchasedfrom Enzyme Systems (Dublin, CA, USA). The monoclonalantibodies to pro-caspase-3 and active caspase-3 were fromTransduction Laboratories and from R&D Systems, respec-tively. A polyclonal antibody to caspase-9 was purchasedfrom New England BioLabs, Inc., and the monoclonalcytochrome c and PARP antibodies were from Pharmingen,Inc. The monoclonal anti-a-fodrin antibody (mAb1622) wasfrom Chemicon International.

Preparation of cell extracts and Western blotting

Cell extracts were prepared as described (JaÈ nicke et al., 1996).To con®rm equal loadings, protein concentrations weredetermined with the Bio-Rad protein assay. Proteins wereseparated in SDS-polyacrylamide gels, and subjected toWestern blotting. The proteins were visualized with theAmersham ECL kit.

Measurements of cell death

Cells were treated with etoposide (100 mM), doxorubicin(1 mg/ml) or were exposed to IR (usually 20 Gy) using agamma chamber 4000 A (Bhabha Atomic Research Centre,Trombay, Bombay, India). Cell death was assessed byvarious methods including microscopic examination, trypanblue uptake, or with the standard cytotoxicity assay (crystalviolet assay) which is a measurement of cell viability (JaÈ nickeet al., 1994). The release of lactate dehydrogenase (LDH) intothe supernatant of irradiated cells (Vassault, 1983) wasemployed as another measurement of cell death. LDHactivity was assessed according to the protocol of themanufacturer (Roche Molecular Biochemicals). Cell deathwas also assessed by the uptake of propidium iodide (2 mg/ml, Sigma) into non®xed cells and subsequent ¯ow cytometricanalyses with the FSC/FL2 pro®le (Wesselborg et al., 1999).

DNA fragmentation and cell cycle analyses

Due to the requirement of caspase-3 for DNA fragmentation(JaÈ nicke et al., 1998b; this study), the method of Nicoletti(Nicoletti et al., 1991) to measure the leakage of fragmentedDNA from apoptotic nuclei could only be assessed incaspase-3-expressing cells (HeLa H21 and MCF-7/casp-3),but not in caspase-3-de®cient MCF-7 cells. Brie¯y, apoptoticnuclei were prepared by lysing cells in a hypotonic lysis bu�er(0.1% sodium citrate, 0.1% Triton X-100 and 50 mg/mlpropidium iodide) and subsequently analysed by ¯owcytometry. Nuclei to the left of the G1 peak containinghypodiploid DNA were considered as apoptotic. For cellcycle analyses, cells were ®xed in ice-cold 80% ethanol,washed with PBS and stained with propidium iodide (50 mg/ml, Sigma) at 378C for 60 min in the presence of ribonuclease(20 mg/ml, Sigma) and 0.1% Triton X-100. All ¯ow cytometryanalyses were performed on a FACScalibur (Becton Dick-

inson) by using CellQuest analysis software. For eachdetermination, a minimum of 20 000 cells was analysed.

Fluorimetric determination of caspase-3 activity

Caspase-3 activity was determined by incubation of cell lysateswith 50 mM of the ¯uorogenic substrate DEVD-AMC (N-acetyl-Asp-Glu-Val-Asp-aminomethylcoumarin; Bachem,Heidelberg, Germany) in 200 ml bu�er containing 50 mM

HEPES pH 7.4, 100 mM NaCl, 10% sucrose, 0.1% CHAPSand 10 mM dithiothreitol. The release of aminomethylcou-marin was measured by ¯uorometry using an excitationwavelength of 360 nm and an emission wavelength of 475 nm.

Neutral single cell gel electrophoresis (comet assay)

The procedure originally described for the detection of DNAdouble-strand breaks on an individual cell level (Klaude etal., 1996) was modi®ed as follows. Cells were trypsinized andwashed twice with PBS. Ten ml of the cell suspension (16106/ml) was resuspended in 120 ml 0.5% low melting pointagarose, spotted onto a microscope slide and covered with acoverslide. After removal of the coverslide, cells wereincubated for 1 h at 48C in neutral lysis bu�er (2.5 M NaCl,100 mM EDTA, 10 mM Tris, 1% sodium laurylsarcosinate(pH 7.5). Electrophoresis was carried out at 48C for 15 minin 90 mM Tris, 90 mM boric acid, and 2 mM EDTA. Theethanol-®xed and ethidium bromide-stained slides wereanalysed with a ¯uorescence microscope. Analysis of DNAmigration was performed using an image analysis system(Kinetic Imaging Ltd., Komet 4.0.2; Optilas) determining themedian tail moment (percentage of DNA in the tail6taillength) of 50 cells per sample.

Measurement of cytochrome c release

For analysis of cytochrome c release, approximately 46106

cells were resuspended in 200 ml of bu�er A containing250 mM sucrose, 20 mM HEPES pH 7.4, 1.5 mM Mg Cl2,10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 1 mM

phenylmethylsulfonyl ¯uoride, and 10 mg/ml of each of theprotease inhibitors aprotinin, bacitracin, antipain, leupeptin.Cells were homogenized, and the homogenates werecentrifuged at 1000 g for 10 min at 48C to remove cellnuclei. The supernatants were transferred to a fresh tube andcentrifuged at 10 000 g for 10 min at 48C to depletemitochondria. The resulting supernatants were loaded on a0.1% SDS and 15% polyacrylamide gel. Cytochrome crelease was analysed by immunoblotting with the mousemonoclonal antibody 7H8.2Cl2 (Pharmingen, Inc.).

AbbreviationsTNF, tumor necrosis factor; IR, ionizing radiation; zVAD-fmk, benzyloxycarbonyl-Val-Ala-Asp-¯uoromethylketone;PARP, poly(ADP-ribose)polymerase; DEVD-amc, N-acet-yl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin

AcknowledgmentsWe thank Alwin Teo for his help with the cell cycleanalysis and Mr Hwang (Physics Department, NationalUniversity of Singapore) for the use of the gammachamber. This work was funded by the Institute ofMolecular and Cell Biology, the Deutsche Krebshilfe andthe Deutsch-Israelische Projektkoordination (DIP).

Oncogene

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5051

References

Ashkenazi A and Dixit VM. (1998). Science, 281, 1305 ±1308.

Belka C, Marini P, Lepple Wienhues A, Budach W, Jekle A,Los M, Lang F, Schulze-Ostho� K, Gulbins E andBamberg M. (1999). Oncogene, 18, 4983 ± 4992.

Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S,Brown JP, Sedivy JM, Kinzler KW and Vogelstein B.(1998). Science, 282, 1497 ± 1501.

Chen Q, Gong B and Almasan A. (2000). Cell Death Di�er.,7, 227 ± 233.

Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC,Hooper ML and Wyllie AH. (1993). Nature, 362, 849 ±852.

Cryns V and Yuan J. (1998). Genes Dev., 12, 1551 ± 1570.Datta R, Kojima H, Banach D, Bump NJ, Talanian RV,

Alnemri ES, Weichselbaum RR, Wong WW and KufeDW. (1997). J. Biol. Chem., 272, 1965 ± 1969.

Droin N, Dubrez L, Eymin B, Renvoize C, Bre ard J,Dimanche-Boitrel MT and Solary E. (1998). Oncogene,16, 2885 ± 2894.

Dubrez L, Savoy I, Hamman A and Solary E. (1996). EMBOJ., 15, 5504 ± 5512.

Engels IH, Stepczynska A, Stroh C, Lauber K, Berg C,Schwenzer R, Wajant H, JaÈ nicke RU, Porter AG, Belka C,Gregor M, Schulze-Ostho� K and Wesselborg S. (2000).Oncogene, 19, 4563 ± 4573.

Faleiro L, Kobayashi R, Fearnhead H and Lazebnik Y.(1997). EMBO J., 16, 2271 ± 2281.

Fan S, El-Deiry WS, Bae I, Freeman J, Jondle D, Bhatia K,Fornace Jr AJ, Magrath I, Kohn KW and O'Connor PM.(1994). Cancer Res., 54, 5824 ± 5830.

Fan S, Smith ML, Rivet II DJ, Duba D, Zhan Q, Kohn KW,Fornace Jr AJ and O'Connor PM. (1995). Cancer Res., 55,1649 ± 1654.

Ferrari D, Stepczynska A, Los M, Wesselborg S andSchulze-Ostho� K. (1998). J. Exp. Med., 188, 979 ± 984.

Fisher DE. (1994). Cell, 78, 539 ± 542.Fuchs EJ, McKenna KA and Bedi A. (1997). Cancer Res., 57,

2550 ± 2554.Hakem R, Hakem A, Duncan GS, Henderson JT, Woo M,

Soengas MS, Elia A, de la Pompa JL, Kagi D, Khoo W,Potter J, Yoshida R, Kaufman SA, Lowe SW, PenningerJM and Mak TW. (1998). Cell, 94, 339 ± 352.

Hannun YA. (1997). Blood, 89, 1845 ± 1853.Heenan M, Kavanagh K, Redmond A, Maher M, Dolan E,

O'Neill P, Moriarty M and Clynes M. (1996). Cytotech-nology, 19, 237 ± 242.

Hollstein M, Sidransky D, Vogelstein B and Harris CC.(1991). Science, 253, 49 ± 53.

JaÈ nicke RU, Lee FHH and Porter AG. (1994). Mol. Cell.Biol., 14, 5661 ± 5670.

JaÈ nicke RU, Lin XY, Lee FHH and Porter AG. (1996). Mol.Cell. Biol., 16, 5245 ± 5253.

JaÈ nicke RU, Ng P, Sprengart ML and Porter AG. (1998a). J.Biol. Chem., 273, 15540 ± 15545.

JaÈ nicke RU, Sprengart ML, Wati MR and Porter AG.(1998b). J. Biol. Chem., 273, 9357 ± 9360.

Kischkel FC, Hellbardt S, Behrmann I, Germer M, PawlitaM, Krammer PH and Peter ME. (1995). EMBO J., 14,5579 ± 5588.

Klaude M, Eriksson S, Nygren J and AhnstroÈ m G. (1996).Mutat. Res., 363, 89 ± 96.

Ko LJ and Prives C. (1996). Genes Dev., 10, 1054 ± 1072.

Kuida K, Haydar TF, Kuan C-Y, Gu Y, Taya C,Karasuyama H, Su MS-S, Rakic P and Flavell RA.(1998). Cell, 94, 325 ± 337.

Kuida K, Zheng TS, Na S, Kuan C-Y, Yang D, KarusuyamaH, Rakic P and Flavell RA. (1996). Nature, 384, 368 ± 372.

Lee Y-J and Shacter E. (1999). J. Biol. Chem., 274, 19792 ±19798.

Lehnert S, Greene D and Batist G. (1989). Radiat. Res., 118,568 ± 580.

Levenson AS and Jordan VC. (1997). Cancer Res., 57, 3071 ±3078.

Levine AJ. (1997). Cell, 88, 323 ± 331.Li F, Srinivasan A, Wang Y, Armstrong RC, Tomaselli KJ

and Fritz LC. (1997a). J. Biol. Chem., 272, 30299 ± 30305.Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M,

Alnemri ES and Wang X. (1997b). Cell, 91, 479 ± 489.Lowe SW, Ruley HE, Jacks T and Housman DE. (1993).

Cell, 74, 957 ± 967.Martins LM, Kottke T, Mesner PW, Basi GS, Sinha S,

Frigon Jr N, Tatar E, Tung JS, Bryant K, Takahashi A,Svingen PA, Madden BJ, McCormick DJ, Earnshaw WCand Kaufmann SH. (1997). J. Biol. Chem., 272, 7421 ±7430.

Muzio M, Chinnaiyan AM, Kischkel FC, O'Rourke K,Shevchenko A, Ni J, Sca�di C, Bretz JD, Zhang M, GentzR, Mann M, Krammer PH, Peter ME and Dixit VM.(1996). Cell, 85, 817 ± 827.

Nicholson DW and Thornberry NA. (1997). TrendsBiochem. Sci., 8, 299 ± 306.

Nicoletti I, Migliorati G, Pagliacci MC, Grignani F andRiccardi C. (1991). J. Immunol. Methods, 139, 271 ± 279.

Oshita F, Fujiwara Y and Saijo NJ. (1992). Cancer Res. Clin.Oncol., 119, 28 ± 34.

Porter AG and JaÈ nicke RU. (1999). Cell Death Di�er., 6,99 ± 104.

Porter AG, Ng P and JaÈ nicke RU. (1997). BioEssays, 19,501 ± 507.

Sabbatini P, Han J, Chiou SK, Nicholson DW and White E.(1997). Cell Growth Di�., 8, 643 ± 653.

Schmitt CA and Lowe SW. (1999). J. Pathol., 187, 127 ± 137.Schuler M, Bossy-Wetzel E, Goldstein JC, Fitzgerald P and

Green DR. (2000). J. Biol. Chem., 275, 7337 ± 7342.Simizu S, Takada M, Umezawa K and Imoto M. (1998). J.

Biol. Chem., 273, 26900 ± 26907.Sionov RV and Haupt Y. (1999). Oncogene, 18, 6145 ± 6157.Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA,

Newmeyer DD, Wang HG, Reed JC, Nicholson DW,Alnemri ES, Green DR and Martin SJ. (1999). J. Cell.Biol., 144, 281 ± 292.

Soengas MS, Alarco n RM, Yoshida H, Giaccia AJ, HakemR, Mak TW and Lowe SW. (1999). Science, 284, 156 ± 159.

Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M,Opitz-Araya X, McCombie R, Herman JG, Gerald WL,Lazebnik YA, Cordon-Cardo C and Lowe SW. (2001).Nature, 409, 207 ± 211.

Srinivasula SM, Ahmad M, Fernandes-Alnemri T, LitwackG and Alnemri ES. (1996). Proc. Natl. Acad. Sci. USA, 93,14486 ± 14491.

Talanian RV, Quinlan C, Trautz S, Hackett MC, MankovichJA, Banach D, Ghayur T, Brady KD and Wong WW.(1997). J. Biol. Chem., 272, 9677 ± 9682.

Thornberry NA and Lazebnik Y. (1998). Science, 281,1312 ± 1316.

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5052

Oncogene

Van de Craen M, Loo GV, Pype S, Criekinge WV, Van denBrande I, Molemans F, Fiers W, Declercq W andVandenabeele P. (1998). Cell Death Di�., 5, 838 ± 846.

Vander Heiden MG and Thompson CB. (1999). Nature CellBiol., 1, E209 ± E216.

Vassault A. (1983). In Methods of Enzymatic Analysis.Bergmeyer, HU (ed) pp. 118 ± 126, Verlag Chemie, Basel,Switzerland.

Wang YJ. (1998). Curr. Opin. Cell Biol., 10, 240 ± 247.Wertz IE and Hanley MR. (1996). Trends Biochem. Sci., 21,

359 ± 364.Wesselborg S, Engels IH, Rossmann E, Los M and Schulze-

Ostho� K. (1999). Blood, 93, 3053 ± 3063.

Woo M, Hakem R, Soengas MS, Duncan GS, Shahinian A,KaÈ gi D, Hakem A, McCurrach M, Khoo W, KaufmanSA, Senaldi G, Howard T, Lowe SW andMak TW. (1998).Genes Dev., 12, 806 ± 819.

Yu Y and Little JB. (1998). Cancer Res., 58, 4277 ± 4281.Zhan Q, Fan S, Bae I, Guillouf C, Liebermann DA,

O'Connor PM and Fornace Jr AJ. (1994). Oncogene, 9,3743 ± 3751.

Zheng TS, Schlosser SF, Dao T, Hingorani R, Crispe N,Boyer JL and Flavell RA. (1998). Proc. Natl. Acad. Sci.,USA, 95, 13618 ± 13623.

Oncogene

Radiation-induced G2/M arrest but not apoptosis in MCF-7 cellsRU JaÈnicke et al

5053