Doxorubicin-Induced Death in Neuroblastoma Does Not Involve Death Receptors in S-Type Cells and Is Caspase-Independent in N-Type Cells

Home , Wild type

Doxorubicin-induced death in neuroblastoma does not involve death receptors in S-type cells and is caspase-independent in N-type cells

Sally Hopkins-Donaldson1, Pu Yan1, Katia Balmas Bourloud1, Annick Muhlethaler1, Jean-Luc Bodmer2 and Nicole Gross*,1

1Department of Pediatric Onco-Hematology, Centre Hospitalier Universitaire Vaudois, CH1011 Lausanne, Switzerland; 2Institute of Biochemistry, University of Lausanne, CH-1066, Epalinges, Switzerland

Death induced by doxorubicin (dox) in neuroblastoma tumors in infants frequently undergo spontaneous (NB) cells was originally thought to occur via the Fas remission. Failure to undergo programmed cell death pathway, however since studies suggest that caspase-8 is a putative mechanism for cancer cell resistance to expression is silenced in most high stage NB tumors, it is chemotherapy, while in contrast spontaneous remission more probable that dox-induced death occurs via a is thought to be a result of increased apoptosis (Oue et different mechanism. Caspase-8 silenced N-type invasive al., 1996). NB cell lines LAN-1 and IMR-32 were investigated for Apoptosis signaling occurs via two different path- their sensitivity to dox, and compared to S-type ways which converge on a common machinery of cell noninvasive SH-EP NB cells expressing caspase-8. All demise that is activated by caspases (Strasser et al., cell lines had similar sensitivities to dox, independently of 2000). The death receptor pathway is activated by caspase-8 expression. Dox induced caspase-3, -7, -8 and ligands of the tumor necrosis factor (TNF) family -9 and Bid cleavage in S-type cells and death was which induce the trimerization of their cognate blocked by caspase inhibitors but not by oxygen radical receptors (Daniel et al., 2001). This aggregation scavenger BHA. In contrast, dox-induced death in N- triggers the recruitment of adaptor protein FADD, type cells was caspase-independent and was inhibited by which in turn binds procaspase-8, inducing its activa- BHA. Dox induced a drop in mitochondrial membrane tion by limited proteolytic cleavage. Effector caspases-3 permeability in all cell lines. Dox-induced death in S- and -7 are subsequently processed and activated, which type cells gave rise to apoptotic nuclei, whereas in N- ultimately results in DNA fragmentation and death. type cells nuclei were non-apoptotic in morphology. The mitochondrial pathway is activated by endoge- Transfection of SH-EP cells with a dominant negative neous stress such as DNA damage and is characterized FADD mutant inhibited TRAIL-induced death, but had by the loss of mitochondrial transmembrane potential no effect on dox-induced apoptosis. These results suggest and the release of cytochrome c from mitochondria that S-type cells undergo apoptosis after dox treatment (Green and Reed, 1998). Cytosolic cytochrome c independently of death receptors, whereas N-type cells triggers the assembly of the apoptosome which recruits are killed by a caspase-independent mechanism. and activates caspase-9 and in turn caspases-3 and -7. Oncogene (2002) 21, 6132 – 6137. doi:10.1038/sj.onc. Anti-apoptotic Bcl-2 family members such as Bcl-2 1205879 inhibit cytochrome c release by modulating the ability of Bax and Bak to facilitate opening of pores in the Keywords: apoptosis; caspases; doxorubicin; caspase-8; outer mitochondrial membrane (Wei et al., 2001). neuroblastoma Crosstalk between the two apoptosis signaling pathways occurs in some cells when caspase-8 converts an inactive form of Bid into a pro-apoptotic form that Introduction interacts with Bax and induces cytochrome c release (Scaffidi et al., 1999). Neuroblastoma (NB) is a heterogeneous pediatric Recent studies have shown that in addition to cancer derived from the sympathetic nervous system. apoptosis, death receptor and mitochondrial signaling Children over 1 year of age with stage IV NB tumors may activate alternative mechanisms of programmed respond poorly to treatment, with a 5 year survival rate cell death which are caspase-independent (Holler et al., of only 30% (Spix et al., 2001), whereas low stage 2000; Leist and Jaattela, 2001). These alternative mechanisms include necrosis and caspase-independent apoptosis which although are rare events in normal cells may be the prominent mode of death in tumor *Correspondence: N Gross, Department of Pediatric Onco-Hematol- cells which have defects in the two apoptosis signaling ogy, University Hospital Lausanne (CHUV), CH-1011 Lausanne, Switzerland; E-mail: [email protected] pathways. Received 10 April 2002; revised 15 July 2002; accepted 16 July Several years ago, Fulda et al. (1997) reported that 2002 death in NB cells induced by chemotherapeutic drugs Doxorubicin-induced death in neuroblastoma cells S Hopkins-Donaldson et al 6133 including doxorubicin (dox) was triggered via the Fas demonstrating that the involvement of the Fas path- signaling pathway. However, conflicting evidence from way in dox-induced apoptosis is cell type specific. In other laboratories demonstrated that lymphocytes type I cells dox activates both the death receptor and derived from Fas-deficient lpr mice were equally as mitochondrial pathways, whereas in type II cells only sensitive to dox as those from wild type mice (Newton the mitochondrial pathway is activated, explaining why and Strasser, 2000) and transfection of cells with in some cells inhibitors of the death receptor pathway specific inhibitors of the Fas pathway such as cFLIP have no effect on dox-induced death. failed to block dox-induced death (Kataoka et al., If the death receptor pathway is important in dox- 1998). Recently, Fulda et al. (2001) attempted to induced death in NB cells, i.e. they are type I cells, it provide an explanation for these discrepancies by follows that resistance to dox should be conferred by

Figure 1 Dox-induced death in NB cells. (a) SH-EP, IMR-32, LAN-1 and Jurkat cells (105 cells/well in 96-well plates) were incubated with dox for 48 h. Tetrazolium dye solution (Promega) was then added and the production of blue formazan product pro- duced by viable cells measured at an absorbance of 570nm. Percentage cell viability compared to untreated controls was calculated. Assays were performed at least three times and a representative experiment shown. (b) Cells were treated with 5 mg/ml dox for 24 or 48 h. Thirty mg of cellular NP-40 (1%) extracts were analysed by SDS – PAGE (12%) and Western blotting. Blots were incubated with anti-caspases-3 (Transduction Laboratories), -7 (PharMingen), -8 (MBL), -9 (New England Biolabs) or anti-Bid (R&D) antibodies followed by peroxidase-conjugated second antibodies (Nordic). Bound antibodies were detected using chemiluminescence. (c) Cells (105 cells/well in 96-well plates) were incubated with different concentrations of dox for 48 h alone or in the presence of 5 100 mM zVADfmk and cell viability assays were performed as above. (d) Cells (10 cells/well in 96-well plates) were incubated with 5 mg/ml dox for 48 h alone or in the present of 50 mM DEVD-FMK, 50 mM zIETD-FMK or 50 mM LEHD-FMK. Cell viability assays were performed as above. Each condition was performed in quadruplicate and means+s.d.s are shown

Oncogene Doxorubicin-induced death in neuroblastoma cells S Hopkins-Donaldson et al 6134 the downregulation of the essential initiator caspase-8 to TNF-related apoptosis inducing-ligand (TRAIL)- or by the expression of a FADD dominant negative induced death (Hopkins-Donaldson et al., 2000). We mutant. The majority of high stage NB tumors and cell also assessed the role of the death receptor pathway in lines of neuronal (N-type) phenotype are in fact dox-induced death of the S-type NB cell line SH-EP silenced for caspase-8 expression by promoter hyper- used previously by Fulda et al. (1998) which are highly methylation, a phenomenon that correlates with sensitive to TRAIL-induced death (Hopkins-Donald- amplification of the N-myc oncogene (Teitz et al., son et al., 2000). Cell viability tests were performed in 2000). N-myc amplification is strongly linked to order to compare the sensitivity to dox-induced death malignancy and chemoresistance in high stage tumors, of N-type and S-type NB cell lines. No differences were and N-type cells are invasive in soft agar assays and observed between N-type and S-type cells in their form tumors when injected subcutaneously into nude sensitivity to dox after 48 h (Figure 1a) or 72 h (data mice (La Quaglia et al., 1996). In contrast, similar to not shown). While the IC50 for dox was 0.5 mg/ml in low stage NB tumors, the substrate adherent (S-type) Jurkat T cells, the three NB cell lines were more cell lines such as SH-EP used by Fulda et al. (1998) are resistant, with IC50 values of between 1 and 5 mg/ml. non-invasive, are not amplified for N-myc, and express Others have observed that dox concentrations as low caspase-8. as 0.5 mg/ml induced death in NB cell lines (Fulda et In the present study we investigated the mechanism al., 1997). This discrepancy with our findings might be of dox-induced death in N-type NB cell lines LAN-1 due to different techniques used to measure cell death, and IMR-32, which are silenced for caspase-8 expres- or to variations in behavior of the same cell lines in sion, amplified for N-myc, and are completely resistant different laboratories. In summary, whereas caspase-8

Figure 2 NB cell death induced by dox involves the mitochondria. (a) NB cells (105 cells/well in 96-well plates) were incubated with 5 mg/ml dox for 48 h alone or in the presence of 50 mM BHA. Cell viability assays were performed and percentage of cell viability calculated as described in Figure 1. Each condition was performed in quadruplicate and means+s.d.s are shown. (b) NB cells were treated with medium (untreated) or 5 mg/ml dox for 48 h. DCm was determined by ﬂow cytometry after staining mitochondria with 40 nM DiOC6(3). The percentages of cells with high or low DCm was calculated. Each staining was performed at least twice and representative data are shown

Oncogene Doxorubicin-induced death in neuroblastoma cells S Hopkins-Donaldson et al 6135 expression is crucial for death receptor-induced The majority of nuclei in these cells appeared intact, apoptosis in NB cells, sensitivity to dox is independent with no visible blebbing although chromatin was more of caspase-8 expression. condensed in comparison to untreated cells. Co- Dox-induced death in NB cells was investigated in incubation of cells with dox and zVADfmk resulted more detail by measuring the extent of caspase and in a decrease in apoptotic SH-EP nuclei (Figure 3c), Bid cleavage by Western blotting (Figure 1b). Cells but had no effect on the morphology of LAN-1 and were incubated with relatively high concentrations of IMR-32 cells (Figure 3f,i). Therefore it would seem dox (5 mg/ml) to ensure that any cleavage was that N-type cells die by a predominantly caspase- detected. Cleavage of caspases-3, -7, -8 and -9 independent mechanism distinct from apoptosis. precursors was detected in SH-EP cells after 24 h Dox-induced death of SH-EP cells has been dox treatment, and Bid cleavage occurred after 48 h. previously reported to occur via the death receptor As expected, N-type cells were deficient for caspase-8 pathway, as incubation of cells with F(ab’)2 anti-Fas expression, and no caspases-3, -7 or Bid cleavage was antibody inhibited dox-induced death (Fulda et al., observed after dox treatment. To further determine the 1998). We investigated this further by transfecting SH- importance of caspases in dox-induced death, NB cells EP cells with dominant negative FADD prior to dox were incubated with dox in the presence of the pan treatment. SH-EP FADD-DN clones expressed large caspase inhibitor zVADfmk (Figure 1c), or with amounts of recombinant protein (Figure 4a), and were specific inhibitors DEVD-FMK, IETD-FMK and completely resistant to TRAIL-induced death, suggest- LEHD-FMK which inhibit caspases-3 and -7, ing that the death receptor pathway had been caspase-8 or caspases-9 activation respectively (Figure effectively blocked (Figure 4b). In contrast, no increase 1d). All caspase inhibitors blocked dox-induced death in resistance to dox was observed in SH-EP FADD- in S-type cells, whereas they had a much smaller effect DN cells, and dox-induced cleavage of caspases-3, -7, on dox-induced death in N-type cells. Although -8 and -9 could be detected (Figure 4b,c). These results cleavage of caspase-9 was detected in IMR-32 cells suggest that caspase-8 can be activated by chemother- after 48 h treatment by Western blotting (Figure 1b), apeutic drugs independently of death receptors, which the caspase-9 inhibitor LEHD-FMK had no effect on has also been described for a non small cell lung cancer dox-induced death (Figure 1d). Together these data cell line and type II Jurkat T cells (Ferreira et al., 2000; suggest that whereas in S-type cells dox-induced death Fulda et al., 2001). It is possible that the mitochondrial is caspase-dependent, caspases are not involved in the pathway may be responsible for this activation as a death of N-type cells. result of an amplification loop whereby caspase-8 is Reactive oxygen species (ROS) are generated by activated by caspase-3, an observation that has been mitochondria in conditions of stress and can result in made in cell-free experiments (Slee et al., 1999). If this the induction of either necrosis or apoptosis (Green is the case, it is likely that SH-EP cells are type II and Reed, 1998). Treatment of cells with dox in the presence of the oxygen scavenger 2(3)-t-butyl-4-hydro- xyanisole (BHA) inhibited death in N-type cells but had no effect on S-type cell death (Figure 2a). This suggests that ROS production is involved in the induction of caspase-independent N-type cell death induced by dox, whereas they are less important in the caspase-mediated death of S-type cells. Similarly, Vercammen et al. (1998) demonstrated that caspase- independent necrosis induced by anti-Fas antibodies in the presence of zVADfmk was inhibited by BHA, whereas apoptosis induced by anti-Fas antibodies alone was not effected. Disruption of the mitochondrial membrane potential (DCm) after 48 h dox treatment was observed in all three cell lines (Figure 2b) with SH- EP having the greatest drop in DCm and IMR-32 the least. These results suggest that the mitochondria plays a role in the death of both S and N-type cells regardless of whether caspases are involved. In order to determine the nature of death induced by dox in NB cell lines, the morphology of nuclei after dox treatment was investigated by 4,6-diamidino-2- Figure 3 Morphology of NB nuclei after treatment with dox. phenylindole (DAPI) staining (Figure 3). SH-EP nuclei SH-EP (a,b,c), IMR-32 (d,e,f) and LAN-1 (g,h,i) cells were incu- contained condensed chromatin after 48 h incubation bated with medium (a,d,g), 1 mg/ml dox for 48 h alone (b,e,h) with 1 mg/ml dox with classic nuclear fragmentation or in the presence of 100 mM zVADfmk (c,f,i). Cells (36105 cells/ml) were cytospun onto glass slides and fixed in acid that denotes apoptosis (Figure 3b). In contrast, LAN-1 acetic/ethanol (1 : 4), then stained for 20 min at 378C with 1 mg/ml and IMR-32 cells treated with the same concentration DAPI in PBS. After washing, slides were mounted and examined of dox contained few apoptotic nuclei (Figure 3e,h). under a fluorescent microscope 663 magnification

Oncogene Doxorubicin-induced death in neuroblastoma cells S Hopkins-Donaldson et al 6136 rather than type I cells, as was originally suggested in N-type cells was recently suggested by Bian et al. (Fulda et al., 1998). (2001) who showed that transfection of N-type cells Teitz et al. (2000) demonstrated that a NB cell line with the inhibitor Ik-BaM blocked dox-induced death. deleted for the caspase-8 gene underwent a 75% Further investigations into the roles of AIF and NF- increase in dox induced apoptosis when transfected kB in dox-induced death in N-type cells are therefore with caspase-8, which suggests that the presence of warranted. caspase-8 is crucial for dox-induced apoptosis in NB. The tumor suppressor gene p53 is involved in the Drug-induced caspase-independent death has been cellular response to cytotoxic agents, and exerts anti- described previously in cancer cells deficient for proliferative effects such as the induction of cell cycle caspases or Apaf-1 (Sartorius and Krammer, 2002; arrest and apoptosis. Inactivation of p53 in cancer cells Miyazaki et al., 2001). The mechanisms involved in this results in genomic instability and diminished apoptosis process are still unclear, but it is generally agreed that and has been proposed as a mechanism of drug mitochondria are involved. Damage to mitochondria resistance (McCurrach et al., 1997). In contrast to can result in the production of oxygen free radicals, many human cancers, p53 mutations are rare in and in a drop in mitochondrial transition permeability neuroblastoma tumors and cell lines and while several which coincides with the release of apoptosis-inducing studies have reported that p53 is functionally intact in factor (AIF). AIF relocates from the mitochondria to neuroblastoma cells, (Goldman et al., 1996; McKenzie the nucleus and is responsible for caspase-independent et al., 1999), cytoplasmic sequestration and defective large scale DNA fragmentation (Susin et al., 2000). We translocation of p53 protein have also been suggested found that the oxygen radical scavenger BHA blocks as mechanisms of non-mutational inactivation (Oster- dox-induced death in N-type cells, and that dox meyer et al., 1996). induces a drop in mitochondrial membrane perme- The cell lines used in this study have variable p53 ability, confirming a role for the mitochondria in dox- statuses, from wild type (SH-EP, IMR-32) to loss of induced death. A role for NF-kB in dox induced death expression (LAN-1) (data not shown). p53 mutations

Figure 4 Transfection of SH-EP cells with a FADD dominant negative mutant. (a) SH-EP cells were stably transfected with empty vector (SH-EP pcDNA) or vector containing FADD-DN cDNA (SH-EP FADD-DN) and Western blots to detect endogenous and recombinant FADD were performed as described in Figure 2. Blots were incubated with mouse-anti-FADD antibody (BD Trans- duction Laboratories). (b) SH-EP pcDNA and SH-EP FADD-DN transfectants were incubated with 100 ng/ml crosslinked TRAIL for 24 h or with 5 mg/ml dox for 48 h and cell viability tests were performed and percentage of cell viability calculated as described in Figure 1. Each condition was performed in quadruplicate and means+s.d.s are shown. (c) SH-EP pcDNA and SH-EP FADD- DN transfectants were incubated with 5 mg/ml dox for 24 or 48 h. Cell lysates were prepared and Western blots were performed as described in Figure 1

Oncogene Doxorubicin-induced death in neuroblastoma cells S Hopkins-Donaldson et al 6137 do not seem to play a major role in the sensitivity to isms is probably insuﬃcient to prevent cancer spread. dox as all three cell lines were similarly killed by dox. Combinations of dox with agents that induce apoptosis However, it is currently not known whether the p53 via the mitochondrial apoptosis signaling pathway may expressed by IMR-32 cells is fully functional, and therefore be more successful than current treatments. therefore the role of p53 in the inability of N-type cells to undergo caspase-dependent death in response to DNA damaging agents is not clear and is currently Acknowledgments being investigated in further detail. We would like to thank Professor Ju¨ rg Tschopp, Institute The results of this study suggest that resistance of of Biochemistry, Lausanne, Switzerland for his kind gift of high stage NB tumors to chemotherapeutic drugs such the vector encoding FADD-DN. This work was funded by as dox may be a result of an absence of caspase Oncosuisse SLF605-1-1998, Foundation FORCE and the activation, and death via caspase-independent mechan- Emma Muschamp Foundation.

References

Bian X, McAllister-Lucas LM, Shao F, Schumacher KR, Miyazaki K, Yoshida H, Sasaki M, Hara H, Kimura G, Mak Feng Z, Porter AG, Castle VP and Opipari Jr AW. (2001). TW and Nomoto K. (2001). J. Biochem. (Tokyo), 129, J. Biol. Chem., 276, 48921 – 48929. 963 – 969. Daniel PT, Wieder T, Sturm I and Schulze-Osthoff K. Newton K and Strasser A. (2000). J. Exp. Med., 191, 195 – (2001). Leukemia, 15, 1022 – 1032. 200. Ferreira CG, Span SW, Peters GJ, Kruyt FA and Giaccone Ostermeyer AG, Runko E, Winkfield B, Ahn B and Moll G. (2000). Cancer Res., 60, 7133 – 7141. UM. (1996). Proc. Natl. Acad. Sci. USA, 93, 15190 – 15194. Fulda S, Meyer E, Friesen C, Susin SA, Kroemer G and Oue T, Fukuzawa M, Kusafuka T, Kohmoto Y, Imura K, Debatin KM. (2001). Oncogene, 20, 1063 – 1075. Nagahara S and Okada A. (1996). J. Pediatr. Surg., 31, Fulda S, Sieverts H, Friesen C, Herr I and Debatin KM. 251 – 257. (1997). Cancer Res., 57, 3823 – 3829. Sartorius UA and Krammer PH. (2002). Int. J. Cancer, 97, Fulda S, Susin SA, Kroemer G and Debatin KM. (1998). 584 – 592. Cancer Res., 58, 4453 – 4460. ScaffidiC,SchmitzI,ZhaJ,KorsmeyerSJ,KrammerPH Goldman SC, Chen CY, Lansing TJ, Gilmer TM and Kastan and Peter ME. (1999). J. Biol. Chem., 274, 22532 – 22538. MB. (1996). Am. J. Pathol., 148, 1381 – 1385. Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Green DR and Reed JC. (1998). Science, 281, 1309 – 1312. Newmeyer DD, Wang H-G, Reed JC, Nicholson DW, Holler N, Zaru R, Micheau O, Thome M, Attinger A, Alnemri ES, Green DR and Martin SJ. (1999). J. Cell Biol., Valitutti S, Bodmer JL, Schneider P, Seed B and Tschopp 144, 281 – 292. J. (2000). Nat. Immunol., 1, 489 – 495. Spix C, Aareleid T, Stiller C, Magnani C, Kaatsch P and Hopkins-Donaldson S, Bodmer JL, Bourloud KB, Brognara Michaelis J. (2001). Eur. J. Cancer, 37, 722 – 729. CB, Tschopp J and Gross N. (2000). Cancer Res., 60, Strasser A, O’Connor L and Dixit VM. (2000). Annu. Rev. 4315 – 4319. Biochem., 69, 217 – 245. Kataoka T, Schroter M, Hahne M, Schneider P, Irmler M, Susin SA, Daugas E, Ravagnan L, Samejima K, Zamzami N, Thome M, Froelich CJ and Tschopp J. (1998). J. Immunol., Loeffler M, Costantini P, Ferri KF, Irinopoulou T, 161, 3936 – 3942. Prevost MC, Brothers G, Mak TW, Penninger J, Earnshaw La Quaglia MP and Manchester KM. (1996). J. Pediatr. WC and Kroemer G. (2000). J. Exp. Med., 192, 571 – 580. Surg., 31, 315 – 318. Teitz T, Wei T, Valentine MB, Vanin EF, Grenet J, Valentine Leist M and Jaattela M. (2001). Nat. Rev. Mol. Cell Biol., 2, VA, Behm FG, Look AT, Lahti JM and Kidd VJ. (2000). 589 – 598. Nat. Med., 6, 529 – 535. McCurrach ME, Connor TM, Knudson CM, Korsmeyer J Vercammen D, Brouckaert G, Denecker G, Van de CM, and Lowe SW. (1997). Proc. Natl. Acad. Sci. USA, 94, Declercq W, Fiers W and Vandenabeele P. (1998). J. Exp. 2345 – 2349. Med., 188, 919 – 930. McKenzie PP, Guichard SM, Middlemas DS, Ashmun RA, Wei MC, Zong WX, Cheng EH, Lindsten T, Panoutsako- Danks MK and Harris LC. (1999). Clin. Cancer Res., 5, poulou V, Ross AJ, Roth KA, MacGregor GR, Thompson 4199 – 4207. CB and Korsmeyer SJ. (2001). Science, 292, 727 – 730.

Oncogene