* Department Clinical and Experimental Medicine (IKE), UNCORRECTED PROOF

Transcription

1 Volume xx Number xx Month 214 pp Wiem Chaabane *,, rtur Cieślar-Pobuda *,, Q1 7 Mohamed El-Gazzah, Mayur V. Jain *, 8 Joanna Rzeszowska-Wolny, Mehrdad Rafat *, 9 Joerg Stetefeld, Saeid Ghavami and Marek J. Łos *,# 2 Human-Gyrovirus-poptin 3 Triggers Mitochondrial Death 4 Pathway is Required for 1 5 poptosis Triggering * Department Clinical and Experimental Medicine (IKE), 11 Division of Cell iology, and Integrative Regenerative Medicine 12 Center (IGEN), Linköping University, Linköping Sweden; 13 Department of iology, Faculty of Sciences, Tunis University, 14 Tunis, Tunisia; iosystems Group, Inst. of utomatic Control, 15 Silesian Univ. of Technology, Gliwice, Poland; Department of 16 Chemistry, University of Manitoba, Winnipeg, Canada; 17 Department of Human natomy and Cell Science, University 18 of Manitoba, Winnipeg, Canada; # Department of Pathology, 19 Pomeranian Medical University, Szczecin, Poland 2 21 bstract 22 The human gyrovirus derived protein poptin (HGV-poptin) a homologue of the chicken anemia virus poptin 23 (CV-poptin), a protein with high cancer cells selective toxicity, trigger apoptosis selectively in cancer cells. In this 24 paper, we show that HGV-poptin acts independently from the death receptor pathway as it induces apoptosis in similar rates in Jurkat cells deficient in either FDD (fas-associated death domain) function or caspase-8 (key players of 26 the extrinsic pathway) and their parental clones. HGV-poptin induces apoptosis via the activation of the mitochondrial 27 intrinsic pathway. It induces both mitochondrial inner and outer membrane permebilization, characterized by the loss 28 of the mitochondrial potential and the release into cytoplasm of the pro-apoptotic molecules including apoptosis 29 inducing factor and cytochrome c. HGV-poptin acts via the apoptosome, as lack of expression of apoptotic protease- 3 activating factor 1 in murine embryonic fibroblast strongly protected the cells from HGV-poptin induced apoptosis. 31 Moreover, QVD-oph a broad-spectrum caspase inhibitor delayed HGV-poptin induced death. On the other hand, 32 overexpression of the anti-apoptotic CL-XL confers resistance to HGV-poptin induced cell death. In contrast, cells 33 that lack the expression of the pro-apoptotic X and K are protected from HGV-poptin induced apoptosis. 34 Furthermore, HGV-poptin acts independently from p53 signal but triggers the cytoplasmic translocation of. 35 Taking together these data indicate that HGV-poptin acts through the mitochondrial pathway, in a caspase- 36 dependent manner but independently from the death receptor pathway Neoplasia (214) xx, Introduction 4 poptosis is the process whereby individual cells of multicellular 41 organisms undergo systematic self-destruction in response to a wide 42 variety of stimuli [1]. poptosis is a genetically encoded program that 43 is involved in normal development and homeostasis and in diverse 44 patho-physiological processes [2]. poptosis functions to eliminate 45 cells during development when they become redundant or as an 46 emergency response after radiation damage, viral infection, or 47 aberrant growth induced by the activation of oncogenes [1]. The 48 morphology of apoptosis is orchestrated by the proteolytic activity of 49 the caspase proteases [3 5] which through the cleavage of many bbreviations: 7D, 7-amino-actinomycin D; IF, poptosis inducing factor; CL-XL, -cell lymphoma extra-large; CV-poptin, Chicken anemia virus apoptin; cyt c, cytochrome c; DISC, Death-inducing signal complex; FDD, Fas-associated death domain; HGV-poptin, Human gyrovirus apoptin; MEF, Mouse embryonic fibroblast; MOMP, Mitochondrial outer membrane permeabilization; TMRM, Tetramethylrhodamine methyl ester perchlorate. ddress all correspondence to: Marek Łos, MD/PhD, Dept. Clinical and Experimental Medicine (IKE), Integrative Regenerative Medicine Center (IGEN), Linköping University, Cell iology uilding, Level 1, Linköping, Sweden. marek.los@liu.se Received 4 July 214; Revised 31 July 214; ccepted 5 ugust Published by Elsevier Inc. on behalf of Neoplasia Press, Inc. This is an open access article under the CC Y-NC-ND license ( by-nc-nd/3./)

2 2 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, 214 proteins largely orchestrate apoptotic process [6]. In vertebrate cells, 51 apoptosis typically occurs through one of two major signaling 52 pathways termed the extrinsic/cell death receptor pathway and the 53 intrinsic/mitochondrial-initiator pathway [7]. In the extrinsic 54 pathway, the ligation of the death receptors leads to the recruitment 55 of the adaptor molecule FDD (fas-associated death domain) that 56 bind, trimerize, and activate an initiator caspase (caspase-8), that in 57 turn directly cleaves and activates the apoptosis executioner caspases 58 (caspase-3 and -7) [2,7,8]. In the intrinsic pathway, the mitochondria 59 respond to apoptotic stimuli through mitochondrial outer membrane 6 permeabilization (MOMP). MOMP leads to the release of 61 pro-apoptotic proteins from the mitochondrial intermembrane 62 space. Following its release, cytochrome c (cyt c) binds apoptotic 63 protease-activating factor 1 (PF1), inducing its conformational 64 change and oligomerization and leading to the formation of a caspase 65 activation platform termed apoptosome. The apoptosome recruits, 66 dimerizes, and activates an initiator caspase, caspase-9, which, in turn, 67 cleaves and activates caspase-3 and -7 [2,7,8]. Thus the caspase 68 cascade activation result from the remarkable MOMP and its 69 subsequent intermembrane space mitochondrial proteins release. 7 MOMP is highly regulated by the cell lymphoma 2 (CL-2) family 71 members [2] which have been classified into 3 classes [9,1]. One 72 class inhibits apoptosis (CL-2, CL-XL, MCL-1, etc), the second 73 class promotes apoptosis (X, K), and a third class termed the 74 H3-only proteins (D, IK, ID, IM, OK, etc.) binds and regulates the anti-apoptotic CL-2 proteins to promote apoptosis [4]. 76 While the pro-apoptotic family members X and K are crucial 77 for the induction of MOMP and the release of the pro-apoptotic 78 molecules, the anti-apoptotic family members CL-2 and CL-XL 79 inhibit X and K [4,11]. Following MOMP, the mitochondrial 8 transmembrane potential is dissipated through caspase-dependent 81 and caspase-independent means [2,12,13]. The intrinsic death 82 pathway is induced by many different stress signals including 83 DN-damaging agents, viral and cellular oncogenes, and transcrip- 84 tional blockade [12,14]. The stimuli are transmitted from the nucleus 85 to the mitochondria by two main molecules: the tumor suppressor 86 gene p53 and the orphan steroid receptor [15]. 87 poptosis plays an important role in the treatment of cancer as it is 88 induced by many treatments [16]. While the most used strategies aims at 89 targeting the apoptotic defects [16], some of the emerging strategies aim 9 at the development of cancer selective therapies by molecules that target 91 and kill preferentially cancer cells. One of the potential tools for cancer 92 selective therapy is CV-poptin as it induces apoptosis selectively in 93 cancer cells [17,18]. CV-poptin is a viral protein of 14 kda derived 94 from the chicken anemia virus [19,2]. The selective toxicity of 95 CV-poptin is associated at least in part to its tumor specific nuclear 96 localization and its tumor specific phosphorylation at Theorine-18, 97 which are essential for its nuclear accumulation and its induction of 98 apoptosis [21,22]. Recently, the human homolog of the CV named the 99 human gyrovirus (HGV) has been identified [23]. Its genome presents 1 an overall organization similar to that of CV [23,24], itconsistsofa 11 single negative-strand circular DN of 2315 nucleotides. HGV has a 12 similar organization of the promoter region and the encoded proteins as 13 the CV as revealed by both virus sequence alignment. It encodes a 1 14 amino-acid homologue of the CV-poptin VP3 protein that despite a 15 low overall identity has conserved important sites including nuclear 16 localization and export signals and phosphorylation sites [23,]. 17 HGV-poptin has the same subcellular distribution as the CV- 18 poptin, it localizes in the nuclei of cancer cells where it shows a granular distribution that later clusters to form aggregates while it remains in the cytoplasm of normal human cells []. Like CV-poptin, HGV- poptin induces apoptosis selectively in cancer cells but not in normal cells [] and is therefore a potential biologics anti-tumor candidate. In this paper we focus on the molecular mechanisms of HGV- poptin selective toxicity. Using cells with defective FDD or caspase-8 (key players in death receptor signaling), PF1 deficient cells, K/X-deficient cells, and other molecular tools, we demonstrate that HGV-poptin induces apoptosis independently of the death receptor pathway. Hence, it triggers the activation of the mitochondrial death pathway via MOMP and the release of cyt c, and apoptosis-inducing factor (IF) from mitochondria, in a caspase dependent manner. HGV-poptin induced apoptosis is modulated by the CL-2 family members, is independent from p53 signal and causes the cytoplasmic translocation of. Material and Methods Chemotherapeutics Inhibitors Staurosporine (2.5 μm) was purchased from Roche Diagnostics, Mannheim, Germany. The broad-range caspase inhibitor QVD-OPh was from Enzyme Systems (Dublin, C, US) and was added to the cells at a concentration of μm immediately after transfection. ntibodies The following primary antibodies were used: mouse anti-active caspase-8 (Cell Signaling Technology, everly, US), rabbit anti FS, and anti FDD antibodies (Santa Cruz iotechnology), murine activating anti-cd95/fs ( ng/ml) from upstate signaling, rabbit anti-if IgG (Sigma-ldrich), mouse anti-p53 antibody (Millipore), Rabbit anti-flag (Thermo/Fisher Scientific), rabbit anti-nur77 IgG (Santa Cruz), mouse anti β-actin (bcam). The following secondary antibodies were used: Infrared dye 8cw goat, anti-rabbit antibody, Infrared dye 68cw goat, anti-mouse antibody (Licor), Rhodamine redx anti-rabbit antibody (life technologies), cy5 anti-murine antibody (bcam). Cell Culture and Reagents Jurkat (T-cell leukaemia) cells, Jurkat clones stably transfected with FDD-DN (a dominant-negative FDD mutant lacking the N-terminal death-effector domain), caspase-8 deficient Jurkat cells, Jurkat cells overexpressing CL-XL and MCF7 (breast adenocarcinoma), MCF7 expressing caspase-3, MCF7 overexpressing CL-XL cells, and MCF7 expressing GFP cyt c were grown in RPMI-164 medium supplemented with 1% fetal calf serum (Hyclone), 1 μg/ml penicillin and.1 μg/ml streptomycin (Gibco RL). HCT116 (colon carcinoma), MEF (mouse embryonic fibroblasts) immortalized by retroviral transduction with a temperature-sensitive simian virus 4 large T antigen as described in [26], MEF-PF1 /,andmef-x-k / were grown in DMEM medium supplemented with 1% fetal calf serum (Hyclone), 1 μg/ml penicillin and.1 μg/ml streptomycin (Gibco RL). Human primary fibroblasts weregrowninfibrogromediafor culture of human fibroblast (Millipore). Cells were grown at 37 C with 5% CO 2 in a humidified incubator. The broad-range caspase inhibitor QVD-oph was from Enzyme Systems (Dublin, C, US). Plasmids and Transient Transfections The expression vectors of HGV-poptin GFP-HGV-PT and FLG- HGV-PT were provided by Dr M. Tavassoli []. The empty vector pegfpc1 was used as negative control. Cells were

3 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al transfected using XtremeGENE HP DN Transfection Reagent 165 according to the manufacturer's instructions (Roche), Jurkat cells 166 were transfected by electroporation using a IO-RD electroporator 167 at a density of 1 7 cells per electroporation with 6 μg of DN. The 168 expression of GFP and GFP-HGV-poptin was confirmed by 169 fluorescence microscopy. The broad-range caspase inhibitor 17 QVD-oph was added to the cells at a concentration of μm 171 immediately after transfection. 172 Measurement of poptosis by Flow Cytometry 173 poptosis was quantified at the indicated time points after 174 transfection using PO-PRO and 7-amino-actinomycin D (7D) 1 staining and according to the manufacture's instructions (Invitrogen). 176 riefly, at the indicated time points, cells were harvested and washed 177 twice with PS, then stained in PS with PO-PRO and 7D for min on ice. Cells were analysed using a Gallios flow-cytometer. 179 The population of GFP positive cells that corresponds to the cells 18 transfected with either the control vector pegfpc1 or with 181 GFP-HGV-PT were gated and analyzed among the whole 182 population of cells, by staining cells with PO-PRO -1 and 7-D 183 apoptotic cells show violet/blue fluorescence, dead cells show violet/ 184 blue and red fluorescence, and live cells show little or no fluorescence. 185 Measurement of Mitochondrial Membrane Potential 186 Mitochondrial permeability transition was determined by staining 187 the cells with tetramethylrhodamine methyl ester (TMRM) (Life 188 Technologies). The cells were harvested and washed once with PS 189 then stained with TMRM 15 min at 37 C and mitochondrial 19 membrane potential was quantified by flow cytometric determination 191 of the FL2 fluorescence of the cells. Data were collected and analyzed 192 using a flow analysis eckman-coulter software (Kaluza). 193 Immunocytochemistry and Confocal Imaging 194 Cells were grown overnight on coverslips and then transfected with 195 pegfpc1, FLG, GFP-HGV-PT, and FLG-HGV-PT. fter the 196 indicated time points, cells were washed with PS and then fixed in 4% 197 paraformaldehyde for 2 min then permeabilized with.1% Triton 198 X-1 and blocked in.1% S. To detect IF release, cells were 199 incubated with anti-if rabbit IgG (Sigma-ldrich); diluted 1:. 2 Followed by three wash steps with PS, the anti-if-antibody 21 complexes were stained with the corresponding Rhodamine redx 22 secondary antibody (diluted 1:2) then washed 3 times with PS. To 23 visualize nuclei, cells were stained with DPI (1 μg/ml). The same was 24 done for using anti- rabbit IgG (Santa-Cruz). The 25 fluorescent images were then observed and analyzed with a Zeiss LSM inverted laser-scanning confocal fluorescence microscope using (N 5 1.4) oil planochromat objective (Carl Zeiss, Thornwood, NY). 28 Cell Extracts and Immunoblotting 29 The activation of caspase8 and the level of protein expression of Fas 21 and FDD, were detected by immunoblotting. riefly, cells 211 were transfected with pegfpc1 vector alone or with 212 GFP-HGV-PT. The expression of GFP and GFP-HGV-PT was 213 confirmed by fluorescence microscopy. t the indicated time periods 214 green cells were sorted using D FCSria III cell sorter (ecton 215 Dickinson, Palo lto C, US), whole cell lysates were prepared 216 from green fluorescent cells using RIP uffer and according to the 217 manufacture's instructions (Thermo-scientific). Proteins (3 μg) 218 were separated by denaturing SDS-PGE and then transferred onto 219 a PVDF membrane. The membranes were blocked in 5% non-fat dry milk in TS.1% Tween and then incubated overnight with the following primary antibodies anti-active caspase8 (Cell Signaling Technology, everly, US), anti FS, and anti FDD antibodies (Santa Cruz iotechnology, Inc.) at 4 C. Then, the blots were incubated with the corresponding InfraRed dye secondary antibodies (Licor). The visualization of the membrane was carried using Licor membrane scanner. Detection and Quantification of poptotic Cells Using Laser Scanning Cytometry Cells grown on cover slips and transfected with pegfpc1 control vector or GFP- HGV-PT were stained at the indicated time points post-transfection with DPI and then analysed by Laser scanning cytometry, using the one primary (DPI fluorescence) and one sub (GFP fluorescence) protocol. For the detection of the chromatin condensation, three parameters were used, the nuclear area, the green maxpixel, and the blue maxpixel values. In this case, the nuclear area was set on the blue channel, cells with high blue maxpixel values (condensed chromatin) [27] are considered apoptotic. Results HGV-poptin Induce poptosis Independently from the Death Receptor Pathway oth CD95 ligand and its receptor are rapidly expressed following T cell activation. In both cases the soluble CD95L can act either in an autocrine manner and trigger apoptosis at a single cell level or can also mediate at high concentrations apoptosis of short distanced neighboring cells in a paracrine manner. CD95L can also act at the membrane bound form and also trigger apoptosis in an autocrine/paracrine manner. To examine the role of the CD95 system in HGV-poptin induced apoptosis we have tested the role of components of the death-inducing signal complex (DISC) which is critical for initiation of death-receptor-mediated apoptosis [28]. DISC is a complex of proteins formed upon the recruitment of the initiator caspase, pro-caspase-8, through its pro-domain together with FDD, a bipartite molecule with a death effector domain, to the death receptor [28 3]. This leads to the activation of the initiator caspase-8 that will directly mediate the down-stream caspase cascade and the execution of apoptosis through the death receptor pathway [3,31]. First we have tested the role of caspase-8 using Jurkat T cells that lacked the expression of caspase-8 (Jurkat-caspase-8 / ). Transiently expressed HGV-poptin in these cells accumulates in the nucleus similar to that of their wild type counterparts, and induces apoptotic morphological changes including cell contraction and nuclear condensation (Figure 1). Thus, suggesting that the lack of caspase-8 does not affect the ability of HGV-poptin to induce apoptosis in Jurkat cells. To further confirm the above findings, caspase-8 activation was monitored by Western blot of the cell lysates from GFP-positive MCF7 and HCT116 at the indicated time points post-transfection with GFP-HGV-poptin and its corresponding pegfpc1 control vector. Up to 72 hours post-transfection, caspase-8 activation was not detected in both cell lines in response to HGV-poptin transient expression unlike the cells that were treated with anti-fs antibody or daunorubucin or both activators of the death receptor pathway and caspase-8 (Figure 1). Taking together this data show that caspase-8 is neither required nor activated during HGV-poptin induced apoptosis. Next, we tested the role of the FDD adaptor molecule in HGV-poptin triggered apoptosis in Jurkat T cells that are overexpressing a truncated, dominant-negative form of FDD

4 4 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, 214 DPI GFP MERGE Jurkat-WT pegfpc1 HCT116 MCF7 Jurkat-WT GFP-HGV-PT Jurkat-cas-8 -/- GFP-HGV-PT. 24h 48h 72h 72h Caspase-8 β actin Caspase β actin Figure 1. HGV-poptin mediated cell death is independent of the death receptor pathway. ) Cellular localization and morphological changes induced by HGV-poptin in Jurakt WT or Jurkat caspase-8 -/- cells. Cells were transfected by electroporation with GFP-HGV-PT and the corresponding pegfpc1 control plasmid, cells were fixed 48 h post transfection and stained with DPI for the detection of nuclear morphology. ) HGV-poptin does not activate caspase-8. Western blot analysis of cell lysates from HCT116 (colon carcinoma) and MCF7 cells transfected either with pegfpc1 or with GFP-HGV-PT for caspase-8 activation. t the indicated time points post-transfection, GFP positive cells were sorted by flow cytometry; cells treated with anti-fs activating antibodies, or, Daunorubicin or both represent a positive control. C) Detection of apoptosis induced by HGV-poptin in Jurkat WT cells or Jurkat FDD-DN. Cells were transfected by electroporation with GFP-HGV-PT or the corresponding pegfpc1 control plasmid. 48 h later, transfected Jurkat cells were fixed and stained with DPI for the detection of nuclear morphology. D) Quantification of cell survival in Jurkat WT and Jurkat FDD-DN as percentage of PO-PRO/7D double negative cells (cell survival) in the GFP-positive population from 24 to 12 h post-transfection. E) Control experiment to assess cell death mediated by CD95/FS, both Jurkat WT and FDD-DN cells were treated with anti-fs (IgM) antibody or with Staurosporine (STUR), and cell death was measured by flow cytometry after 12 h. F) Western blot analysis of FS and FDD level of expression in Jurkat cells 48 h post transfection with GFP-HGV-PT or the corresponding pegfpc1 control vector. In panels and F β actin was used as a protein loading control. G) Quantification of the level of expression of FS and FDD 48 h post-transfection (Western blot shown in G ).

5 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al. 5 C DPI GFP MERGE Jurkat-WT pegfpc1 D 1 F Fas FDD β actin Jurkat-WT GFP-HGV-PT Jurkat-FDD-DN GFP-HGV-PT JWT-Ctrl JWT-GFP-HGV-PT J-FDD-DN-Ctrl J-FDD-DN-GFP-HGV-PT 24h 48h 72h 96h 12h E 1 G Protein expression level Untreated CD95 STUR p-egfpc1 J-WT J-FDD-DN fas FDD GFP-HGV-PT Figure 1 (continued) 277 (FDD-DN). The expression of a dominant negative form of FDD 278 prevents the formation of a functional DISC and will block the 279 signaling via the extrinsic pathway. Jurkat-FDD-DN cells show 28 apoptotic morphology similar to the Jurkat WT cells (Jurkat 281 expressing the intact DDD) after transient expression of HGV- 282 poptin (Figure 1C). Moreover the monitoring of apoptosis induced 283 by HGV-poptin in both Jurkat-WT and Jurkat-FDD-DN cells from 24 to 12 hours post-transfection by flow cytometry shows that both WT and FDD-DN cells have similar sensitivity to HGV-poptin (Figure 1D). Thus, the state of FDD has no effect in HGV-poptin induced apoptosis. In a control experiment, Jurkat WT and FDD-DN were treated for 12 hours with the anti-fs human activating antibody, which induce cell death via the recruitment of FDD and the activation of

6 6 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, caspase-8. fter 12 hours of treatment while almost all the Jurkat WT 292 cells were dead the Jurkat-FDD-DN were completely resistant to 293 treatment with FS (Figure 1E). 294 The above observations were further confirmed by monitoring the 295 expression of FS and FDD in GFP-HGV-poptin transiently 296 expressing Jurkat cells by Western blot. The protein expression 297 profiles of these apoptotic regulators were monitored in cell lysates of 298 the GFP-positive cells at 48 h post-transfection. Expression of FS 299 and FDD were detected in all the samples with no significant 3 increase in response to HGV-poptin expression. (Figure 1, F and G) 31 Taking together this data indicates that HGV-poptin similar to its 32 homolog CV-poptin induce apoptosis independently from the 33 death receptor pathway HGV-poptin Triggers the Mitochondrial Pathway of Cell Death and Induces Mitochondrial lterations 36 To gain more insight into the mechanism of apoptotic signaling 37 triggered by HGV-poptin, we investigated the role of the 38 mitochondrial intrinsic pathway in HGV-poptin induced apoptosis. 39 The intrinsic signaling pathways that initiate apoptosis involve diverse 31 non-receptor mediated stimuli that produce intracellular signals. These 311 signals act directly on targets within the cell and are mitochondrial- 312 initiated events [32] that lead to mitochondrial alterations and loss of 313 the mitochondrial membrane potential [33]. Thus in order to assess 314 changes in the mitochondrial potential induced by HGV-poptin we 315 have used the TMRM dye that is rapidly sequestrated by healthy GFP-HGV-PT pegfpc1 Loss of mitochondrial membrane potential (%) 1 mitochondria while the depolarized mitochondria does not accumulate the dye [34]. Indeed, cells transfected with HGV-apoptin show severe decrease in TMRM fluorescence (Figure 2) compared to control treated cells. Moreover monitoring of the mitochondrial potential by flow cytometry in HCT116 shows a huge decrease in the mitochondrial potential among HGV-poptin transient expressing cells (Figure 2). Similar results were obtained in Jurkat WT cells along with the mutant Jurkat FDD-DN cells after transfection with HGV-poptin showing that FDD have no effect on the loss of the mitochondrial potential caused by HGV-poptin (Figure 2C). This data indicates that HGV-poptin triggered apoptosis is associated with the loss of the mitochondrial potential and further confirms that this occurs independently from the death receptor pathway. HGV-poptin Induces the Release of the Mitochondrial Components Cytochrome c and IF and Triggers the poptosome Pathway Most pro-apoptotic stimuli that involve mitochondrial outer membrane permeabilization leading to intra-cytosolic release of mitochondria intermembrane space proteins such as cytochrome c (cyt c) will trigger caspase activation and/or IF mediated caspase independent death pathway [35]. ctivation of both pathways often coexists. Thus to further examine the effect of HGV-poptin on mitochondria, we have first checked cyt c release. For that purpose we have used MCF7 cells stably transfected with GFP cyt c fusion protein (MCF7-cytc-GFP). s shown in Figure 3, the transient DPI GFP TMRM MERGE pegfpc1 12h 24h 48h 72h 96h Loss of mitochondrial membrane potential (%) C 1 J-WT-GFP-HGV-PT J-WT-pEGFPC1 J-FDD-DN-pEGFPC1 J-FDD-DN-GFP-HGV-PT 12h 24h 48h 72h 96h Time post transfection Figure 2. Mitochondrial membrane potential is lost in response to transient expression of HGV-poptin. ) Mitochondrial staining of HCT116 cells transfected with GFP-HGV-PT or the corresponding pegfpc1 control plasmid. Cells were stained with TMRM dye for 15 min, then fixed and counterstained with DPI. ) Monitoring of mitochondrial membrane potential (Ψm) in HCT116, either transfected with the HGV-poptin tagged to GFP: GFP-HGV-PT or the control plasmid: pegfpc1, for the indicated time periods. Ψm was determined using TMRM dye, which shows a decrease in the red fluorescence upon loss of the mitochondrial potential. C) Monitoring of the mitochondrial potential in Jurkat WT and Jurkat-FDD-DN from 12 to 96 h post-transfection with GFP-HGV-PT and pegfpc1.

7 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al. 7 DPI CYTC FLG MERGE FLG-HGV-PT FLG DPI GFP IF DPI/IF/MERGE MERGE pegfpc1 HGV-GFP-PT Non-transfected Figure 3. HGV-poptin triggers the release of pro-apoptotic mitochondrial proteins. ) HGV-poptin induces cyt c release from transformed cells. MCF7-cyt c-gfp cell line was transiently transfected with FLG-HGV-PT and the corresponding FLG control vector, at 48 h post-transfection cells were fixed, stained with a primary rabbit anti-flg and secondary Red-X anti-rabbit antibody and counterstained with DPI. ) The cellular localization of IF in HCT116 cells as determined by confocal microscopy. HCT116 cells were transiently transfected with GFP-HGV-PT and the corresponding pegfpc1 control plasmid, after 48 h cells were fixed, stained with a primary rabbit anti-if and secondary Red-X anti-rabbit antibody and counterstained with DPI for nuclear visualization (the non-transfected cells along with p-egfpc1 transfected cells were used as negative controls).

8 8 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, expression of FLG-HGV-PT in MCF7-cytc-GFP leads to the 342 release of cyt c into the cytoplasm but not the transient expression of 343 the FLG empty vector (please note the selective loss of the green 344 granular pattern in cells transfected with FLG-HGV-PT). 345 Next, we checked for the release of IF. Similar to cyt c release, the 346 release of IF from mitochondria was also detected by confocal 347 microscopy in HCT116 in response to GFP-HGV-poptin transient 348 expression (Figure 3). This data show that HGV-poptin triggers 349 the release of the mitochondrial components cyt c and IF. 3 While IF can cause apoptosis through high molecular weight 351 DN fragmentation and chromation condensation independently 352 from caspase activity [35,36], cyt c, along with dtp binds PF1 353 and form the death signaling platform: apoptosome. This event is a 354 key step for caspase-9 activation and is crucial for the initiation of 355 apoptosis via the intrinsic pathway [37]. Therefore, in order to give 356 more insight into the role of the apoptosome in HGV-poptin 357 induced cell death we have checked the role of PF1 using 358 transformed mouse embryonic fibroblast genetically modified to 359 lack the expression of PF1 in transformed mouse embryonic 36 fibroblast (MEF-PF1 / ). Expression of GFP-HGV-poptin 361 in MEF-PF1 / is non-toxic as compared to the WT cells 362 (Figure 4, and ). These results show that HGV-poptin induce 363 apoptosis via the apoptosome and further confirm the implication of the 364 mitochondrial intrinsic pathway in HGV-poptin induced apoptosis. 365 HGV-poptin Mediated Cell Death is Caspase Dependent 366 To further elucidate the HGV-poptin triggered death pathway 367 we have checked the role of caspases, the key mediator of the 368 apoptotic program [5]. Upon apoptosis triggering, the upstream 369 caspases (caspase-8, caspase-9) cleave and activate downstream 37 caspases (caspase-3, caspase-7) leading to the execution of apoptosis. 371 For this purpose, Jurkat T cells were transfected with 372 GFP-HGV-PT in the presence or the absence of μm ofthe 373 broad-range caspase inhibitor (QVD-oph). t 24 hours post- 374 electroporation cells were examined microscopically for protein 3 expression and the effects of QVD-oph. In the presence of QVD-oph, 376 cells that express HGV-poptin were partially protected from cell 377 death. In contrast, QVD-oph non-treated cells were highly sensitivity 378 to HGV-poptin. To investigate the effect of caspase inhibition in 379 detail, apoptosis was monitored in both Jurkat non-treated cells and 38 Jurkat cells treated with caspase inhibitor QVD-oph from 24 to h post-transfection. s shown in Figure 5 the inhibition of 382 caspase activity by QVD-oph delayed HGV-poptin induced 383 apoptosis in Jurkat cells. 384 To further investigate whether the inhibition of caspase could 385 block HGV-poptin induced apoptosis cell death was monitored 386 among caspase-3-deficient MCF7 cells and MCF7 cells stably 387 overexpressing caspase-3 (MCF7/C3) either left untreated or treated 388 with μm of the caspase inhibitor QVD. s shown in Figure the ectopic expression of caspase-3 renders MCF7 more sensitivity to 39 HGV-poptin. Moreover, treatment with the caspase inhibitor 391 Q-VD-oph strongly inhibited HGV-poptin induced apoptosis in 392 both MCF7 and MCF7/C3 cells, these results further confirm that 393 caspase activity is required for HGV-poptin induced apoptosis HGV-poptin Mediated Cell Death is Modulated by the CL-2 Family Members 396 Mitochondrial outer membrane permeabilization, which plays a 397 crucial role in apoptosis induction, is controlled by pro- and MEF-PF1KD pegfpc1 MEF-PF1KD GFP-HGV-PT MEF-WT GFP-HGV-PT DPI GFP MERGE 24h 48h 72h MEF WT MEF-PF1 -/- Figure 4. Requirement of PF1 for HGV-poptin induced apoptosis. ) Immortalized PF1 knockout mouse fibroblasts were transfected with GFP-HGV-poptin or the corresponding pegfpc1 control plasmid and the relative confocal images of MEF WT and MEF-PF1 / 48 h post-transfection, cells were fixed and stained with DPI for the detection of nuclear morphology. ) Quantification of cell survival as percentage of PO-PRO/7D double negative cells in the GFP-positive population from 24 to 72 h post-transfection of MEF-WT and MEF-PF1 -/- cells. anti-apoptotic members of the CL-2 family [35]. Therefore we examined the role of both pro- and anti-apoptotic CL-2 family in HGV-poptin induced apoptosis. First we checked the role of the anti-apoptotic CL-XL in HGV-poptin induced cell death. We compared the sensitivity towards HGV-poptin induced apoptosis between the parental MCF7 cell line and its derivates stably over-expressing CL-XL. The transient expression of HGV-poptin leads to the apoptosis of MCF7 cells within 48 h whereas MCF7-CL-XL cells remained viable and appeared morphologically healthy. These observations were further confirmed by time-lapse studies. The monitoring of apoptosis induced by HGV-poptin by flow cytometry in both MCF7 and MCF7-CL-XL cells from 24 to 96 h shows that over expression of the anti-apoptotic CL-XL render the cells resistance to HGV-poptin as they remain viable compared to the parental clone (Figure 6). Since the protective effect of CL-XL against HGV-poptin could be cell type specific, we further tested the inhibitory activity of

9 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al Ctrl GFP-HGV-PT-untreated GFP-HGV-PT+QVD 24h 48h 72h 96h 12h Time post-electroporation (h) MCF7 MCF7+ QVD MCF7/C3 MCF7/C3 + QVD 24h 48h 72h 96h Time post transfection GFP-HGV-PT (h) Figure 5. HGV-poptin mediated cell death is caspase-dependent. ) Quantification of cell survival as percentage of PO-PRO/7D double negative cells in the GFP-positive population from 24 to 12 h post-transfection. Jurkat cells were transfected either with GFP-HGV-PT or with pegfpc1 control plasmid and then either left untreated or incubated with μm of QVD-oph and then stained with PO-PRO/7D. ) Quantification of cell survival as percentage of PO-PRO/7D double negative cells in the GFP-positive population from 24 to 96 h post-transfection. Wild-type caspase-3 deficient MCF7 cells and MCF7 cells expressing caspase-3 (MCF7/C3) were transiently transfected either with GFP-HGV-PT or with pegfpc1 control plasmid and then either left untreated or incubated with μm of QVD-oph. Cells were then stained with PO-PRO/7D and analyzed by flow cytometry. 415 CL-XL in the Jurkat T cells, and compared the sensitivity toward 416 HGV-poptin via apoptosis monitoring by flow cytometry in both 417 the parental clone and Jurkat cells overexpressing CL-XL. s shown 418 in Figure 6C over-expression of CL-XL also protected Jurkat cells 419 from HGV-poptin induced apoptosis. This data show that 42 HGV-poptin induced apoptosis is regulated by the anti-apoptotic 421 CL-XL independent of cell type. 422 Next we have tested the role of the pro-apoptotic X and K in 423 HGV-poptin induced apoptosis using MEF cells that lack the 424 expression of X and K (MEF-X-K -/- ). While MEF WT 4 cells were sensitive to the transient expression of HGV-poptin 426 MEF-X-K -/- cells remained viable and appeared morphologi- 427 cally healthy (Figure 6E). The monitoring of apoptosis induced by 428 HGV-poptin by flow cytometry in both MEF-WT and 429 MEF-X-K -/- cells from 24 to 72 h post-transfection shows 43 that the lack of expression of the pro-apoptotic family members X 431 and K protected MEF cells from HGV-poptin induced cell death. To further expand the above observations we have measured the mitochondrial membrane potential of MCF7-WT, MCF7-CL-XL (Figure 6), Jurkat-WT, Jurkat-CL-XL (Figure 6D) MEF-WT, and MEF-X-K -/- (Figure 6F) at 48 and 72 hours post transfection with HGV-poptin expressing vector. The parental clones showed a decrease in the mitochondrial potential while the CL-XL-expressing cells have maintained their mitochondrial membrane potential. lso, MEF-X-K -/- cells failed to respond with the decrease of mitochondrial membrane potential upon the expression of HGV- poptin. Taking together, this data show that HGV-poptin induced apoptosis is regulated by the CL-2 family members, both the pro- and the anti-apoptotic ones. HGV-poptin Mediated Cell Death is Independent from p53 Signal and Induces the Cytoplasmic Translocation of HGV-poptin activates the mitochondrial cell death pathway. p53 and are the main molecules capable of transmitting the apoptotic stimuli from nucleus to the cytoplasm. Thus we tested their respective role in HGV-PT induced cell death. First, we investigated the role of p53, the tumor suppressor protein. p53 functions as a transcriptional regulator of multiple apoptotic genes, also has a direct pro-apoptotic function in the mitochondria by directly interacting with both the anti- and pro-apoptotic CL-2 family members [15,34]. s shown in Figure 7, lack of expression of p53 in HCT116 cells does not prevent HGV-poptin from induction of apoptosis as both HCT116 WT and their counterparts lacking p53 both were equally sensitive to HGV-poptin expression. Hence, both cell-types show apoptotic morphology in response to HGV-poptin expression (48 h time-point). To get further insight in the role of p53 in HGV-poptin induced apoptosis, we have performed a time course studies. We have monitored apoptosis in HCT116-WT and their p53-deficient progeny between 24 and 12 h post-transfection with HGV-poptin by laser-scanning cytometry. Cells with a condensed chromatin were considered apoptotic and quantified (Figure 7). These experiments show that HGV-poptin does not require p53 for induction of apoptosis. Next, we tested the role of NUR77, an orphan nuclear receptor and member of the steroid/thyroid receptor family, in HGV-poptin triggered apoptosis. NUR77 is capable not only of transmitting the apoptotic stimuli from the nucleus to the mitochondria [38], but also can modulate apoptosis via activating the transcription of pro- and anti-apoptotic genes [39]. Expression of HGV-poptin leads to the translocation of NUR77 from nucleus to the cytoplasm in both HCT116 (Figure 8) and MCF7 cells (Figure 8), indicating that NUR77 is a signaling molecule transmitting apoptotic signal in HGV-poptin pathway. Discussion The selective targeting of cancer cells has been a holy grail of cancer therapy. Since cancer cells are derived from our own healthy tissues, the task seems very difficult. However, in the past two decades, growing number of proteins have been discovered, that display selective anti-cancer activity [4]. The protoplast of the group is CV-poptin, discovered in the mid-9s by Noteborn and colleagues [41]. CV-poptin causes apoptosis of cultured tumorigenic and transformed human cell lines but is harmless to normal cells [19]. This cancer cell selective toxicity made CV-poptin among the most potential candidate for tumor cells selective therapy. Recently, in 211, the human homolog of the CV-poptin has been identified. The human gyrovirus-poptin (HGV-poptin) [23]

10 1 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, displays the same differential behavior between normal and tumor 491 cells as its homolog. Recently published studies show that 492 HGV-poptin translocates to the nucleus of transformed cells 493 while it remains in the cytoplasm of normal cells, similarly to 494 CV-poptin. Moreover, HGV-poptin also induces apoptosis in 495 cancer cells while leaving normal cells intact []. This prompted us 496 to deeper study the cytotoxycity mechanism of HGV-poptin. etter understanding of the molecular mechanism of HGV-poptin may allow for the development apoptin-protein derivates better suited for the selective killing of cancer cells. number of anti-cancer drugs up-regulate CD95L, which in turn may contribute to their anti-cancer activity in an autocrine and/or paracrine manner [42 44]. lso, activation of T-cells leads to the expression of CD95L that kills T-cells in autocrine or paracrine fashion C E MCF7-WT-Ctrl MCF7-WT-GFP-HGV-PT 24h 48h 72h 96h 24h 48h 72h 96h J-WT-Ctrl J-WT-GFP-HGV-PT J-CLxl-Ctrl J-CLxl-GFP-HGV-PT MEF-WT-Ctrl MEF-WT-GFP-HGV-PT MEF-X-K-KO-Ctrl MEF-X-K-KO-GFP-HGV-PT Loss of mitochondrial membrane potential (%) D Loss of mitochondrial membrane potential (%) F Loss of the mitochondrial potential (%) MCF7-WT-pEGFPC1 MCF7-WT-GFP-HGV-PT MCF7-CLxl-pEGFPC1 MCF7-CLxl-GFP-HGV-PT 48h 72h J-WT-pEGFPC1 J-WT-GFP-HGV-PT J-CLxl-pEGFPC1 J-CLxl-GFP-HGV-PT 48h 72h MEF-WT-pEGFPC1 MEF-WT-GFP-HGV-PT MEF-X-K-KOl-pEGFPC1 MEF-X-K-KO-GFP-HGV-PT 24h 48h 72h 96h 48h 72h

11 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al. 11 DPI GFP MERGE HCT116- P53KD-GFP- HGV-PT HCT116-WT-GFP- HGV-PT HCT116-WT- PEGFPC [45]. Thus, we checked the role of the CD95 signaling pathway in 54 HGV-poptin induced apoptosis. FDD and caspase-8 are the most 541 important upstream components of the CD95/Fas signaling pathway, 542 so using cells that overexpress a dominant negative form of FDD and 543 other molecular tools we show that neither FDD nor caspase-8 are 1 HCT116-WT-Ctrl HCT116-WT-GFP-HGV-PT HCT116-P53 -/- Ctrl HCT116-P53 -/- GFP-HGV-PT h 24h 48h 72h 96h 12h Time post transfection ctin required for the apoptotic activity of HGV-poptin. Moreover, since some previous reports indicate that Fas can still signal atypical apoptotic cell death in the absence of caspase-8 [46], we examined by Western blot if HGV-poptin upregulates Fas expression. Our results show that in response to HGV-poptin expression, no significant increase in Fas P53 C HCT116 P53 +/+ HCT116 P53-/- Figure 7. HGV-poptin induced apoptosis is independent from p53 signal. ) p53 deficient HCT116 (HCT116-p53 / ) and the wild type (HCT116-WT) colon carcinoma cells were transfected with pegfpc1 or GFP-HGV-PT. The cells were fixed and stained with DPI 48 h post-transfection for the detection of the nuclear morphology; arrows indicate apoptotic cells expressing GFP-HGV-PT. ) Quantification of apoptosis in HCT116-WT and HCT116-p53 / cells using laser-scanning cytometry. Cells were fixed and stained with DPI then analysed by laser scanning cytometry, cells with a high blue max-pixel values (condensed chromatin) are considered apoptotic. C) Western blot data confirming that HCT116-p53 / lack p53. Figure 6. HGV-poptin induced apoptosis is regulated by the anti- and pro-apoptotic CL-2 family members. ) Quantification of cell survival as percentage of PO-PRO/7D double negative cells in the GFP-positive population from 24 to 96 h post-transfection. MCF7 cells and MCF7-CL-XL cells were transfected with GFP-HGV-PT or the corresponding control plasmid pefpc1 and stained with PO-PRO/7D. ) Monitoring of the mitochondrial membrane potential in MCF7 WT and MCF7 CL-XL cells after transfection with either GFP-HGV-PT or the control plasmid using the TMRM dye. C) Quantification of cell survival in Jurkat wild type and Jurkat-CL-XL as percentage of PO-PRO/7D double negative cells in the GFP-positive population from 24 to 96 h post-transfection. Cells were transfected by electroporation with GFP-HGV-PT or the corresponding control plasmid. D) Monitoring of the mitochondrial membrane potential in Jurkat WT and Jurkat CL-XL after transfection with either GFP-HGV-PT or the control plasmid using the TMRM dye. E) Quantification of apoptosis in the wild-type murine embryonic fibroblasts (MEF-WT) and MEF X and K deficient cells (MEF-X-K-KO). Cells were transfected with GFP-HGV-poptin or the corresponding control plasmid pefpc1. Cell survival was quantified by flow cytometry after PO-PRO/7D staining as percentage of PO-PRO/7D double negative cells in the GFP-positive population of cells. F) Monitoring of the mitochondrial membrane potential in MEF-WT and MEF-X-K-KO using the TMRM dye staining. Due to space constrain CL-XL was labeled in some figures as CLxl

12 12 Human-Gyrovirus-apoptin action Chaabane et al. Neoplasia Vol. xx, No. xx, 214 DPI GFP NUR77 MERGE EGFPC1 HGV-GFP-PT EGFPC1 HGV-GFP-PT 549 level is detectable. Taken together, our data indicate that HGV- 5 poptin, like its homolog CV-poptin, cancer cells independently 551 from the CD95 system. 552 The mitochondrial intrinsic death pathway is ultimately activated 553 by a number of anti-cancer drugs and stress stimuli that cause the loss 554 of the mitochondrial membrane potential [47] and the release of a 555 number of pro-apoptotic molecules from the mitochondria. mong 556 others, released are cyt c and IF [48,49]. Hence, we have checked 557 the mitochondrial alterations induced by HGV-poptin, also the 558 cellular localization of cyt c and IF. We have observed that 559 HGV-poptin induces a decrease in the mitochondrial potential, and 56 release of both cyt c and IF from the mitochondria. 561 lthough caspase-independent forms of apoptosis have been 562 described [], among them the nuclear translocation of IF where 563 it induces apoptosis via DN fragmentation, caspases play a crucial 564 role in apoptosis and are activated in both the extrinsic and intrinsic 565 pathways. Caspases are activated through the formation of large 566 protein complexes namely death inducing signal complex (DISC) 567 within the extrinsic pathway or the apoptosome within the intrinsic 568 pathway [7]. Lack of expression of PF1 (a major component of the 569 apoptosome), strongly delayed HGV-poptin induced apoptosis (up 57 to 96 h post-transfection). 571 poptosome is the molecular platform for caspase-9 activation. 572 Caspase-9 then activates down-stream caspases namely caspase-3 and 573 caspase-7 [51,52] leading to the execution of apoptosis (Figure 9). So DPI GFP NUR77 MERGE Figure 8. HGV-poptin induced cell death involves the nucleo-cytoplasmic translocation of. Confocal microscopy visualization of in ) MCF7 cells, and ) HCT116. Cells were fixed 48 h post-transfection and stained with a primary rabbit anti NUR77 antibody and a secondary anti-rabbit RED-X antibody and then counterstained with DPI for the visualization of the nuclei. in order to investigate the requirement of caspase activity in HGV-poptin induced apoptosis we have used the broad-spectru caspase inhibitor QVD-oph that inhibits both up-stream and down-stream caspases. QVD-oph delayed HGV-poptin induced cell death, indicating that HGV-poptin induced apoptosis occurs in a caspase-dependent manner. The inhibitory effect of QVD on HGV-poptin induced cell death was not only observed in Jurkat T cells but also in caspase-3 deficient MCF7 cells and its retransfected derivative MCF7/C3, which were only protected from HGV-poptin induced cell death in the presence of the caspase inhibitor. It should be noted that HGV-poptin kills MCF7 cells that lack the expression of caspase-3, which indicates that caspase-3 may be dispensable for HGV-poptin induced cell death, and substituted by other down-stream caspases such as caspase-6 and caspase-7. Moreover the ectopic expression of caspase-3 in MCF7 cells renders them slightly more sensitive to HGV-poptin. Thus, HGV-poptin requires caspase activity to kill its targets. The mitochondrial pathway is modulated by the pro- and antiapoptotic CL-2 family members [53]. In order to check the role of the CL-2 members in HGV-poptin induced cell death we have used MCF7 cells that overexpress the anti-apoptotic CL-XL. Over-expression of CL-XL inhibits the pro-apoptotic molecules X and K that in turn, inhibit MOMP and the release of pro-apoptotic molecules from the mitochondria [54]. Over-expression of CL-XL protected MCF7 cells from HGV-poptin induced

13 Neoplasia Vol. xx, No. xx, 214 Human-Gyrovirus-apoptin action Chaabane et al. 13 ΔΨ ΔΨ m m MOMP CL-2 subfamily ax/ak channels X k Cyt. c PF1 poptosome 599 apoptosis. Moreover, murine embryonic fibroblasts that lack the 6 expression of the pro-apoptotic X and K are protected from 61 HGV-poptin induced apoptosis compared to their parental clones. 62 Thus, both pro- and anti-apoptotic CL-2 family members regulate 63 HGV-poptin induced cell death. 64 In conclusion, in this paper we give an insight into the molecular 65 mechanism of HGV-poptin induced cell death (Figure 9). We show 66 that HGV-poptin acts via the intrinsic mitochondrial pathway in a 67 caspase-dependent manner, and that its activity is modulated by 68 CL-2 family members. Elucidating the molecular mechanism of 69 HGV-poptin's signaling pathway is of great importance, and it may 61 lead into the development of novel anticancer strategy allowing for 611 selective targeting of cancer cells. Comparative study of HGV- 612 poptin and CV-poptin signaling pathways, may lead to the 613 discovery of new molecular targets uniquely available only in cancer 614 cells. HGV-poptin does not require p53 tumor suppressor signaling IF Pro-caspase-9 Caspase-9 Caspase-3 poptosis whose inactivation is a frequent event in tumorigenesis [55], and which leads to the resistance to available cancer therapies [56]. References Nucleus? cknowledgments MJL kindly acknowledges the core/startup support from Linkoping University, from Integrative Regenerative Medicine Center (IGEN), from Cancerfonden (213/391), and from VR-NanoVision (K212-99X ). HGV- poptin HGV- HGV- poptin poptin HGV- poptin Figure 9. Schematic representation of HGV-poptin mechanism of action in cancer cells. HGV-poptin activation and nuclear accumulation induces the nuclear export of nur77, which might induce the activation of the CL-2 family members of proteins, leading to MOMP and the subsequent cytoplasmic release of IF and cyt c leading to apoptosis through PF1 activation, the apoptosome formation, and the subsequent caspase activation [1] Teodoro JG and ranton PE (1997). Regulation of apoptosis by viral gene 6 products. J Virol 71, [2] Tait SW and Green DR (21). Mitochondria and cell death: outer membrane 627 permeabilization and beyond. Nat Rev Mol Cell iol 11, [3] Kerr JF, Wyllie H, and Currie R (1972). poptosis: a basic biological phenomenon 629 with wide-ranging implications in tissue kinetics. r J Cancer 26,