FLIP overexpression inhibits death receptor-induced ... - Nature

Oncogene (2004) 23, 7753–7760

& 2004 Nature Publishing Group All rights reserved 0950-9232/04 $30.00 www.nature.com/onc

ORIGINAL PAPERS

FLIP overexpression inhibits death receptor-induced apoptosis in malignant mesothelial cells Maria Rita Rippo*,1, Simona Moretti1, Silvia Vescovi1, Marco Tomasetti1, Sara Orecchia3, Giuseppe Amici4, Alfonso Catalano1,2 and Antonio Procopio1,2 1 Department of Molecular Pathology and Innovative Therapies, Polytechnic University of Marche, 60100 Ancona, Italy; 2Laboratory of Cytology, Italian National Research Center on Aging, 8 60124 Ancona, Italy; 3Department of Anatomical Pathology, S Antonio and Biagio Hospital, 15100 Alessandria, Italy; 4Pediatric Hospital G Salesi, 60100 Ancona, Italy

Tumors have developed several forms of resistance to receptor-induced cell death. Here, we show that malignant mesothelial (MM) cell lines as well as primary MM cells and normal mesothelial (NM) cells express Fas and TNFrelated apoptosis-inducing ligand (TRAIL) receptors DR4 and DR5. We found that, although Fas expression levels are comparable, only MM cells are resistant to cell death. Furthermore, MM cells show resistance to TRAILinduced apoptosis. Caspase-8 (FLICE) is not activated by death receptors triggering in malignant cells whereas it is well activated by nonreceptor stimuli, such as UV radiation. We found that FLIP (FLICE-Inhibitory Protein) is constitutively expressed in all MM cell lines and is more expressed in primary MM cells than in NM cells. Knockdown of FLIP expression in MM cell lines, by a FLIPsiRNA, re-established the normal response to apoptosis induced by Fas or DR4/DR5, which was blocked by pretreatment with the caspase-8 inhibitor z-IETD-fmk. These results indicate that MM cells develop an intrinsic resistance to apoptosis induced by death receptors upregulating the expression of the antiapoptotic protein c-FLIP. Oncogene (2004) 23, 7753–7760. doi:10.1038/sj.onc.1208051 Published online 30 August 2004 Keywords: malignant mesothelioma; FLIP; caspase-8; apoptosis

Fas;

TRAIL;

Introduction Pleural malignant mesothelioma (MPM) is a fatal tumor, which is linked to asbestos exposure (Attanoos and Gibbs, 1997) and associated to SV40 infection (Carbone et al., 1994). It is expected that the incidence of malignant mesothelioma (MM) will increase in the next decades. The survival of patients with MPM is not more than 6% with a median survival time after diagnosis of 6–12 months. Mesothelioma is a chemoresistant malignancy. Only multimodality regimens of *Correspondence: MR Rippo; E-mail: [email protected] Received 22 April 2004; revised 22 July 2004; accepted 23 July 2004; published online 30 August 2004

treatment, based on surgery, radiotherapy and chemotherapy, improve the long-term survival for a very small subgroup of patients (Jaklitsch et al., 2001; Tomek et al., 2003). Furthermore, it has been shown that MM patients have impaired immune responsiveness (Manning et al., 1991). In normal cells, death is induced by several immunological, physiological stimuli, among them the triggering of the death receptor Fas (CD95/Apo-1) by its ligand (FasL). TRAIL receptors, DR4 and DR5, are expressed and selectively induce apoptosis in a wide variety of tumors without exhibiting toxicity to normal cells or tissues (French and Tschopp, 1999). Fas and DR4 and DR5 have an intracytoplasmic death domain (DD), which is important for the transduction of the proapoptotic signaling cascade. In fact, death receptor trimerization, through the DD, induces the recruitment of the adaptor protein FADD that, in turn, binds to and activates the caspase-8 proenzyme. The protein complex formed by all these proteins has been called DISC (Death-Inducing Signaling Complex) (Peter and Krammer, 2003). Activated caspase-8, in turn, mediates the dissipation of the mitochondrial transmembrane potential and the activation of the postmitochondrial caspase-9. Mitochondria, however, take part also in nonreceptor proapoptotic processes. Expression of Fas and TRAIL death receptors on normal mesothelial (NM) cells has been recently demonstrated. FasL induces apoptosis in cultured NM cells (Catalan et al., 2003). The sensitivity to Fas-induced cell death of NM cells may play an important role in controlling homeostasis in normal mesothelium, which is not only a protective layer but it is also a dynamic structure regulating serosal responses to injury, disease and infection (Mutsaers, 2002). The alteration of the Fasinduced apoptotic program could therefore play a role in the tumorigenesis of mesothelial cells. Many tumors are resistant to Fas- and TRAIL-induced cell death through different molecular mechanisms (Wright et al., 1994). Some produce decoy receptors (DcRs) (Roth et al., 2001) some, such as colon cancer cells, express FasL to perform the so-called ‘tumor counterattack’ (Hahne et al., 1996; O’Connel et al., 1996), others upregulate the expression of apoptosis inhibitors (French and Tschopp, 2002; Mitsiades et al., 2002;

FLIP inhibits Fas and TRAIL-induced apoptosis in MM cells MR Rippo et al

7754

Thomas et al., 2002). The caspase-8 inhibitor FLIP (FLICE-Inhibitory Protein) has a caspase-like domain, but it lacks amino-acid residues that are critical for caspase activity. Two forms of FLIP, long (FLIPL) and short (FLIPS), exist both with inhibiting functions. The mechanism of cell death inhibition by FLIP may reside in its ability to preclude recruitment of caspase-8 to the DISC and thereby its proteolytic cleavage (Krueger et al., 2001). Much information is available about the mechanisms mediating the proliferation and progression of MM, but little is known about MM cells sensitivity to death receptors. We and others have previously demonstrated that MM cell lines show low susceptibility to TRAIL (Liu et al., 2001; Tomasetti et al., 2004) (Vivo et al., 2003) and that this is not due to Bcl-2 (Narasimhan et al., 1998). Here, we show that, although MM primary cells and cell lines express Fas and DR4/DR5, they are not killed by their crosslinking. Furthermore, we demonstrate that the mechanism of resistance to Fas and TRAIL in MM cells systems is due to FLIPL, which is overexpressed in malignant, but not in normal cells.

Results MM cells are resistant to Fas- and TRAIL-induced cell death Little information exists on the sensitivity of MM cells to death receptor-induced apoptosis, thus, we have first analysed Fas expression in normal and malignant mesothelial (MM) cells. Seven MM cell lines (Ist-Mes3, MM-B1, MPP-89, Meso-2, H-Meso, Ist-Mes1, Ist-Mes2), five MM primary cultures (MesRF, Mes-SS, Mes-ST, Mes-PR, Mes-DP), and four NM cell cultures (NM-1, NM-8, NM-12, NM-13) were subjected to cytofluorimetric analysis, which revealed that Fas is expressed in all MM primary cell cultures, cell lines and NM cells and its levels appear apparently comparable, as shown in Figure 1a, where a representative cytofluorimetric analysis of MM-B1, Mes-DP and NM-1 of the extended one (Table 1) is indicated. Moreover, normal and malignant cells express DR4 and DR5, with highest DR5 expression in all cells analysed (Figure 1b and Table 2). Furthermore, Fas- and TRAIL-induced cell death in mesothelial cells was investigated by propidium iodide staining. All the primary cultures and cell lines,

Figure 1 Fas and TRAIL receptors/decoy receptors expression in MM and NM cells. Receptors expression was investigated by flow cytometry, using anti-Fas DX2 (a), anti-DR4 (gray), -DR5 (black) (b); -DcR1 (gray), -DcR2 (black) (c) monoclonal antibodies and anti-CD14 as a negative control. MM-B1 MM cell line, Mes-DP MM primary cells and NM-1 cells are representative of analysed samples (Tables 1 and 2) Table 1 Evaluation of Fas cell surface expression in MM cell lines and primary cells and NM cells MM cell lines

Fas

MM primary cells

Fas

NM cells

Fas

Ist-Mes3 MM-B1 MPP-89 Meso-2 H-Meso Ist-Mes1 Ist-Mes2

2.770.3 7.770.8 4.570.4 5.170.5 2.670.2 5.670.2 6.370.6

Mes-RF Mes-SS Mes-ST Mes-PR Mes-DP

3.970.3 2.170.2 2.570.4 3.170.5 5.670.3

NM-1 NM-8 NM-12 NM-13

5.370.3 2.870.5 3.370.2 4.670.3

Expression of Fas was evaluated by flow cytometry. The values represent the fluorescence intensity of the receptors normalized for the negative control. Data are expressed as mean7s.d. from three independent experiments

Table 2 Evaluation of TRAIL receptors cell surface expression in MM cell lines, MM primary cells and NM cells MM cell lines

DR4 DR5 DcR1 DcR2

MM primary cells

NM cells

MM-B1

Meso-2

Ist-Mes2

Ist-Mes3

Mes-SA

Mes-DP

NM-1

NM-12

4.270.4 7.570.7 2.570.2 570.4

1.670.2 3.970.3 1.770.1 270.2

370.4 5.370.3 2.570.1 3.370.2

2704 3.570.1 1.870.3 2.370.2

2.370.3 5.170.2 2.770.2 270.1

1670.8 1970.6 370.3 4.370.6

2.470.2 570.5 570.7 6172

1.970.5 3.570.4 770.8 3171.3

Expression of DR4, DR5, DcR1 and DcR2 was evaluated by flow cytometry. The values represent the fluorescence intensity of the receptors normalized for the negative control. Data are expressed as mean7s.d. from three independent experiments Oncogene


7755

Figure 3 Caspase-8 is not activated by Fas in MM cells. (a) Three MM cell lines (H-Meso, Meso-2 and MM-B1) were treated at the indicated time points with anti-Fas CH11 (500 ng/ml). Lysates were subjected to Western blot analysis with anti-caspase-8-specific antibody. Decrement of the 55 kDa procaspase-8 and formation of the active fragment (20 kDa) are not detectable. (b) NM-1 cells were pretreated (where indicated) with z-IETD-fmk and treated for 24 h with anti-Fas or TRAIL and processed as in (a). Representative experiments of n ¼ 3 are shown. (c) Procaspase-8 and -9 activation was assessed in UV-treated MM-B1 cell line. (d) The same samples were also processed for the analysis of hypodiploid nuclei. Results (mean7s.d.) are from two separate experiments with duplicates. P-values are indicated Figure 2 Resistance to Fas- and TRAIL-induced cell death of MM cells. (a) MM cell lines (Meso-2, MM-B1, Ist-Mes2), MM primary cells (Mes-DP and Mes-RF) and NM cells (NM-1) were treated with different doses of anti-Fas (CH11) for 24 h or (b) at different times with 500 ng/ml anti-Fas and then subjected to cytofluorimetric analysis of hypodiploid nuclei. (c) Apoptosis induced by anti-Fas (500 ng/ml) and TRAIL (50 ng/ml) in MM cell lines, MM primary cells and NM cells was evaluated after 24 h by propidium iodide and the percentage of apoptotic cells was compared. Results represent the mean7s.d. of n ¼ 3. *Pp0.05 compared to control

previously tested for Fas expression, were treated in a dose (250–1000 ng/ml) (Figure 2a) and time-dependent manner (24–48 h) (Figure 2b) with an anti-Fas agonistic antibody (CH11 clone), which induces Fas trimerization. No significant proapoptotic effect was detected on both MM primary cultures and cell lines compared to NM cells (Figure 2a, b). We have previously described that MM cell lines are resistant to TRAIL-induced apoptosis, in a dose- and time-dependent manner (Tomasetti et al., 2004). Like MM cell lines, primary MM and NM cells show a strong resistance to TRAIL-induced cell death (Figure 2c).

Similar results were also obtained by annexin-V staining (data not shown). Since the resistance to TRAIL may depend on its DcRs expression (Roth et al., 2001), we investigated whether mesothelial cells have DcR1 and DcR2 proteins (Figure 1c and Table 2) and mRNA (data not shown). NM cells had higher amounts of DcRs than MM cells, suggesting that NM cells may be intrinsically resistant to TRAIL-induced apoptosis by expressing its DcRs and that additional mechanisms could be involved in the resistance of MM cells. Caspase-8 is not activated in MM cells Fas-induced caspase-8 activation was investigated by Western blot analysis in MM and NM cells. Fas crosslinking did not induce procaspase-8 cleavage in MM cells, as the proenzyme band disappearance and the cleavage partial product (40 kDa) and the active fragment (20 kDa) production did not occur. A representative Western blot of three MM cell lines (H-Meso, Meso-2 and MM-B1), out of seven, is indicated (Figure 3a). Similarly, TRAIL did not activate Oncogene


7756

caspase-8 up to 20 h of treatment (Tomasetti et al., 2004). On the contrary, caspase-8 was well activated after 24 h of anti-Fas treatment in NM-1 cells. Notably, no caspase-8 activation was observed in NM cells treated with TRAIL (Figure 3b), which may be associated to DcRs expression. Similar results were also obtained with NM-12 and NM-13 cells (data not shown). Procaspase-8 and -9 were specifically cleaved in UV-treated MM cells that underwent cell death (Figure 3c, d), indicating that the postmitochondrial apoptotic pathway is not inhibited in MM and that caspase-8 can be activated by nonreceptor stimuli. These data suggest that MM cells have a mechanism of resistance to programmed cell death upstream of caspase-8 activation, at the receptor level. Flip is expressed in MM cells FLIP overexpression has been demonstrated in different kinds of tumors (Djerbi et al., 1999; Medema et al., 1999; Tepper and Seldin, 1999). Therefore, we analysed mRNA and protein levels of FLIPL/S in primary MM and NM cells as well as MM cell lines. Figure 4a shows that FLIPL and FLIPS mRNA are present in all primary MM cells and cell lines, whereas, in NM cells, FLIPL mRNA is poorly represented (NM-12) or not detectable (NM-13) as well as FLIPS mRNA. Similarly, FLIPL protein is expressed in all malignant cells, both in primary cultures and cell lines but is not detectable in NM cells (Figure 4b). FLIPS is well expressed in only three (MM-B1, H-Meso and Meso-2) of the MM cell lines (data not shown). These data indicate that FLIPL more than FLIPS may block MM cell death.

Figure 4 FLIPS/L expression in MM and NM cells. FLIPS/L expression was studied by RT–PCR and Western blot analysis. (a) RNA from human primary MM cells (Mes-RF, SS, ST, PR and SC), from MM cell lines (Ist-Mes3, MM-B1, Meso-2, H-Meso, MPP-89, Ist-Mes2 and Ist-Mes1,) and from NM cells (NM-12 and NM-13) were subjected to RT–PCR using specific primers for human FLIPL/S. b-Actin amplification was used as an internal control. (b) Western blot of untreated primary MM cells, MM cell lines and NM cell lysates with monoclonal anti-FLIP antibody. Data are from one experiment representative of three that gave comparable results

siRNA suppression of FLIP promotes death receptors caspase-8 activation in MM cells To correlate caspase-8 inhibition and FLIPL expression in MM cells, we selectively downregulated FLIPS/L with double-stranded small interfering RNA [FLIPsiRNA] (Siegmund et al., 2002) and analysed caspase-8 activation upon Fas and DR4/DR5 triggering by Western blot. In FLIPsiRNA-transfected Ist-Mes2 cells, FLIP expression was downregulated after 24 h, compared to control nonspecific siRNA (CsiRNA)-transfected cells (Figure 5a). FLIP suppression restored the ability of Fas and TRAIL to transduce a signal, since procaspase-8 partial cleavage product (40 kDa) and, more importantly, the 20 kDa active fragment appeared within 3 h of anti-Fas and TRAIL treatment of FLIPsiRNA MMtransfected cells but not in CsiRNA-transfected cells (Figure 5b). Furthermore, caspase-8 is specifically activated by death receptors in transfected cells and its activation is due to its autoproteolytic properties since the caspase-8-specific inhibitor z-IETD-fmk completely inhibits the cleavage product accumulation upon the triggering of death receptors (Figure 5b). This pattern is similar to that observed for the additional MM-B1 cell line (Figure 5c) Oncogene

Figure 5 FLIPsiRNA suppresses FLIPL expression and sensitizes MM cells to Fas-induced caspase-8 activation. The MM cell line Ist-Mes2 and MM-B1 were transfected with FLIPsiRNA or CsiRNA. Suppression of FLIPL protein expression was controlled by Western blot (a). Ist-Mes2 (b) and MM-B1 (c) transfected cells were treated for 3 h with the anti-Fas antibody (500 ng/ml) or with TRAIL (50 ng/ml) (c). Cell lysates were analysed by Western blotting for caspase-8 processing


7757

FLIP suppression sensitizes MM cells to Fas and TRAILinduced cell death Since NM cells do not express FLIP and are susceptible to Fas-induced cell death, we further investigated whether FLIP is functionally sufficient to protect MM cells from Fas and DR4/DR5 triggering. MM Ist-Mes2 cells were transfected for 24 h with FLIPsiRNA or CsiRNA, treated for only 10 h with the anti-Fas agonistic antibody or TRAIL and the percentage of apoptotic cells assessed by annexin-V and propidium iodide staining. As shown in Figure 6a, 50 and 40% of FLIPsiRNA-treated Ist-Mes2 cells underwent DNA fragmentation under stimulation of Fas and TRAILRs, compared with 10% of control cells. Furthermore, annexin-V staining showed, in a similar manner, that pretreatment of FLIPsiRNA-transfected cells with the inhibitor z-IETD-fmk blocked death receptor-induced apoptosis (Figure 6b), supporting data that cell death resistance of MM cells is due to FLIP-mediated caspase8 inhibition. Similar results were obtained with MM-B1 cell line (data not shown).

Discussion The Fas/FasL system plays a crucial role in regulating cellular and tissue homeostasis and represents one of the mechanisms by which immune cells kill target tumor cells. It has been published that the rejection of tumors by NK cells depends on a mechanism of death receptormediated apoptosis and that the introduction of FLIPL in tumor cells blocks the rejection (Screpanti et al., 2001). Furthermore, the expression of Fas on tumor targets makes them more immunogenic and susceptible to IL-2-induced LAK activity and CTL (Bradley et al., 1998). Fas plays an important role also in chemotherapy since drug-resistant, Fas þ -expressing prostate tumor cells, can be sensitized by cisplatin and etoposide to killing by FasL-bearing CTL, TIL and LAK cells (Frost et al., 1997). However, there is some evidence that asbestos-exposed people and malignant mesothelioma patients have impaired immune responsiveness. In particular, it has been observed that LAK and NK cells activity against mesothelioma target cells is reduced. The CD4 subset of lymphocytes is reduced in number, while the CD8 subset is unchanged (Lew et al., 1986; Manning et al., 1991). Furthermore, we have observed a strong presence of lymphocytes in pleural exudates (unpublished observation). The molecular mechanism responsible for this immune unresponsiveness is mainly unknown. To elucidate the mechanisms of MM cells immunoresistance, we have examined their response to the triggering of death receptors. We have observed that, although the expression levels of membrane Fas in NM and MM cells are comparable (Table 1 and Figure 1a), MM cells are much more resistant to Fasinduced cell death (Figure 2). Another group of researchers have recently published that in a small number of MM cell lines the Fas/FasL pathway is intrinsically intact (Stewart et al., 2002). This study does

Figure 6 FLIPsiRNA sensitizes MM cells to Fas and TRAIL killing. Ist-Mes2 cells were CsiRNA or FLIPsiRNA transfected and the proapoptotic effect of anti-Fas (500 ng/ml) and TRAIL (50 ng/ml) revealed, after 10 h, by (a) PI and (b) annexin-V staining, followed by cytofluorimetric analysis. z-IETD-fmk inhibited apoptosis (b). Results represent the mean7s.d. of n ¼ 3. P-values are indicated

not elucidate the sensitivity of primary MM cells to death receptor-induced cell death and does not unravel the molecular mechanism of resistance of insensitive cell lines. Here, we have shown that MM primary cells express TRAIL receptors (Table 2), but they are not killed by TRAIL (Figure 2c). These data are supported by previous observations that MM cell lines are also resistant to TRAIL-induced apoptosis (Liu et al., 2001; Tomasetti et al., 2004). Furthermore, NM cells are not killed by TRAIL and it could be due to decoy receptors expression (Figure 1 and Table 2). Taken together these data indicate that MM cells can develop a mechanism of resistance to death receptors, which, in normal conditions, may play a still unknown important role in the maintenance of mesothelium function. Many tumors acquire the ability to resist to death receptor-induced cell death upregulating antiapoptotic proteins that are distributed at different levels of the proapoptotic pathway. To explain the molecular mechanisms mediating the resistance of MM cells to Fas and TRAIL-triggered apoptosis, we investigated the ability of Fas to transduce an early signal, such as caspase-8 activation. Our data Oncogene


7758

show that in MM cells, the procaspase-8 is not processed, even 24 h after treatment with the anti-Fas agonistic antibody, in contrast to NM cells (Figure 3a, b). However, caspase-8 and the postmitochondrial caspase-9 can be activated in MM cells induced to die by nonreceptor stimuli, such as UV radiation. These data suggest that the antiapoptotic protein FLIP inhibits procaspase-8 recruitment and activation at the receptor level. Since Fas and DR4/DR5 signaling could be blocked by FLIPL and FLIPS isoforms, we studied FLIPL/S expression. Results obtained show that all the MM cell lines and primary cells tested express high levels of FLIPL in contrast to NM cells (Figure 4a, b). FLIPS mRNA is present in all cells analysed, but the protein is not detectable in primary MM cells and in around 50% MM cell lines (data not shown). Therefore, FLIPL, more than FLIPS seems to be important to inhibit death receptor-mediated apoptosis in MM cells. The importance of FLIP was confirmed when its expression was downregulated in MM cells. FLIPL downregulation (Figure 5a) allowed Fas and TRAIL to activate caspase-8 (Figure 5b, c) and more importantly, through the activation of the initiator caspase-8, to kill MM cells (Figure 6). Caspase-8 activation is important to propagate the proapoptotic signal in our cell system since pretreatment of FLIPsiRNA MM-transfected cells with the specific caspase-8 inhibitor z-IETD-fmk blocks the apoptotic process (Figure 6b). These data show clearly that overexpression of FLIP in MM cells can block the apoptotic pathway induced by Fas and TRAIL. We had previously demonstrated that MM cells express the death receptors DR4/DR5, but have a low susceptibility to their ligand (Tomasetti et al., 2004), through mechanisms still unknown; we have now given a good explanation for this resistance. This is an important point of view because TRAIL is selective for cancer cells and knocking down FLIP could be a good future strategy for treatment of malignant mesothelioma. The use of RNAi for genetic-based therapies is actually studied in cancers (Wall and Shi, 2003). Furthermore, these data may have another relevant clinical impact since FLIP, which is selectively expressed in MM but not in NM cells, can be considered in the future as a diagnostic and prognostic factor for MM. Finally, this work opens up two important questions: Firstly, which are the factors that induce FLIP expression in MM cells? Since MM cell produce angiogenic cytokines, promoting their survival and proliferation (Kumar-Singh et al., 1999; Romano et al., 2001; Strizzi et al., 2001), we suppose that these cytokines could modulate FLIP expression as a mechanism of selfprotection from proapoptotic cytokines. Secondly, our data suggest how MM cells can resist immune system attack but they do not explain how, in turn, they suppress the activity of immune system cells. Do express MM cells FasL to kill CTL and NK cells? These hypotheses are under investigation in our laboratory. In conclusion, these data clearly demonstrate that FLIP suppression in MM cells correlates with a strong susceptibility to Fas and TRAIL-induced cell death and Oncogene

that, therefore, FLIP play an important role in MM cell survival in response to the triggering of death receptors.

Materials and methods Cell cultures and treatments Human MM cell lines and primary MM and NM cells were established from patients and identified morphologically by phenotypic analysis. After 2 weeks in culture 100% of NM cells stained positive for calretinin. Cells were cultured in HAM’S/F-10 supplemented with 10% FBS, 1% L-glutamine and 1% penicillin–streptomycin (Hy-Clone, Rome, Italy) at 371C and 5% CO2. Fas and DR4/DR5 activation was induced by the IgM anti-Fas antibody CH-11 (UBI, Lake Placid, NY, USA) and by hrTRAIL (prepared as described elsewhere (Plasilova et al., 2002)), respectively. To assay procaspase-9 cleavage, MM cells were exposed for 2 min with a UV transilluminator (Euroclone) with a peak intensity of 9000 mW/cm2 at the filter surface and a peak emission of 313 nm, and left 18 h in culture before processing for Western blot analysis. The pancaspase inhibitor zVAD-fmk and the caspase-8-specific inhibitor z-IETD-fmk (Alexis, Lausen, Switzerland) were used at a concentration of 20 mM. AnnexinV-FITC was from Sigma (St Louis, MO, USA). Immunofluorescence and flow cytometry analysis Purified anti-CD14 (negative control) was provided by Becton Dickinson (San Jose, CA, USA), purified anti-DR4, DR5, DcR1 and DcR2 was from Alexis (Lausen, Switzerland) and purified anti-Fas (DX2) was kindly donated by Dr R Testi. FITC labeled goat anti-mouse (GaM-FITC) was from PharMingen (BD). For immunostaining, cells (2 105/ml) were incubated for 30 min at 41C with the appropriate amounts of the primary antibodies, washed and incubated again for 30 min with the GaM-FITC antibody. Samples were analysed by flow cytometry on a FACScaliburs (Becton Dickinson). Evaluation of apoptotic cells Nuclear hypodiploidy and condensation was assessed by propidium iodide staining. Briefly, cells at 2 105/well were plated in six-well plates, treated with anti-Fas (CH-11 clone) or TRAIL (50 ng/ml), trypsinized, washed in PBS and resuspended in hypotonic fluorochrome solution (50 mg/ml propidium iodide in 0.1% sodium citrate plus 0.1% Triton X-100), kept for 8 h at 41C in the dark and analysed by FACScaliburs cytofluorimeter for the evaluation of the percentage of hypodiploid nuclei. Apoptosis was quantified using the annexin V-FITC method, which detects phosphatidyl serine externalized in the early phases of apoptosis. Briefly, cells were washed twice in PBS, resuspended in 0.1 ml binding buffer (10 mM HEPES, 140 mM NaCl, 5 mM CaCl2, pH 7.4), incubated for 20 min at RT with annexin V-FITC, supplemented with 10 ml of propidium iodide (10 mg/ml) and analysed by flow cytometry. RT–PCR analysis Total RNA was isolated from MM and NM cells using OMNIzol (EuroClone Ltd, UK) according to the manufacturer’s instructions. The cDNA was synthesized from 1 mg total RNA using ProtoScript First-Strand cDNA Synthesis kit (New England BioLabs Inc., USA) and 1 ml of cDNA was used


7759 for each amplification reaction. PCR analyses were performed with EuroTaq DNA polymerase (EuroClone Ltd, UK). b-Actin primers were used for PCR internal control. The primers used for amplification of Fas, FLIPL/S, DR4, DR5, DcR1, DcR2 and b-actin have been described (Nagaraju et al., 2000; Bernard et al., 2001; Zang et al., 2001). FLIP and b-actin primers have the following sequences: FLIPL: 50 -AATTCA AGGCTCAGAAGCGA-30 (sense), 50 -GGCAGAAACTCTG CTGTTCC-30 (antisense) generating a 226 bp product (annealing temperature 581C for 35 cycles), FLIPS: 50 -ACCTTGT GGTTGAGTTGGAGAAAC-30 (sense), 50 -ACAATTTCCA AGAATTTTCAGATCAG-30 (antisense) generating a 268 bp products (annealing temperature 581C for 35 cycles); b-actin: 50 -TCATGTTTGAGACCTTCAA-30 (sense), 50 -GTCTTT GCGGATGTCCACG-30 (antisense) amplifying a 500 bp fragment (annealing temperature 581C for 35 cycles).

chemiluminescence (SuperSignal West Pico Chemiluminescent substrate, PIERCE, Rockford, USA) and autoradiography. Preparation of small interfering RNA (siRNA) and transfection The FLIP-specific siRNA was previously described (Siegmund et al., 2002). A nonspecific, double-stranded siRNA (QiagenXeragon, Germantown, MD, USA) with identical length was also generated based on the sequence of an unrelated protein (lamin-C) and used as control (CsiRNA). In addition, a nonsilencing siRNA showing no known homology to mammalian genes, labeled with fluorescein, was used to monitor the transfection efficiency. Cells (2 105/well) were grown in sixwell plates and transfected with annealed siRNA using the RNAi Starter Kit (Qiagen, MI, Italy) following the manufacturer’s instructions. Cells were then harvested at different times for further assays. Gene silencing was monitored at the protein level by Western blot analysis.

Western blot analysis After treatments, cells were lysed in lysis buffer (150 mM NaCl, 10 mM Tris-HCl pH 7.4, 1 mM EDTA, 1% Triton X-100, 0.5% NP40 and proteases inhibitors). Proteins (80 mg) were separated on a 12.5% SDS–PAGE and transferred onto nitrocellulose membrane (Schleicher and Schuell Bioscience, Germany). Membranes were blocked with 5% nonfat dry milk and blotted with the following antibodies: anti-human caspase-8 monoclonal antibody (3–1–9) (BD Biosciences Pharmingen, San Diego, CA, USA); mouse anti-human FLIP (NF-6) (Alexis Biochemical, Switzerland); rabbit anti-human caspase-9 (H-83) and rabbit anti-actin (H-300) (Santa Cruz Biotechnology Inc., CA, USA). For detection, anti-mouse IgG and anti-rabbit IgG peroxidase conjugate (SIGMA, St Louis, MO, USA) were used, followed by enhanced

Statistical analysis Results are expressed as mean7s.d. The two-sided Student’s t-test was used for statistical comparisons. A P-value p0.05 was considered statistically significant. Acknowledgements This work was supported by grants from the Associazione Italiana per la Ricerca contro il Cancro (AIRC), from the Polytechnic University of Marche Research Fund and from the Italian Department of Scientific Research to AP and MRR and by special Grant DLgs 502/92 from Italian Ministery of Health to MRR. AC is a FIRC fellowship holder. The authors thank Dr Roberto Testi for providing DX2 hybridoma and Dr Michael Freeman for technical assistance.

References Attanoos R and Gibbs A. (1997). Histopathology, 30, 403–418. Bernard D, Quatannens B, Vanderbunder B and Abbadie C. (2001). J. Biol. Chem., 276, 27322–27328. Bradley M, Zeytun A, Rafi-Janaireh A, Nagarkatti P and Nagarkatti M. (1998). Blood, 11, 4248–4255. Carbone M, Pass HI, Rizzo P, Marinetti M, Muzio MD, Mew DJY, Levine A and Procopio A. (1994). Oncogene, 9, 1781–1790. Catalan MP, Subira D, Reyero A, Selgas R, Ortiz-Gonzalez A, Egido J and Ortiz A. (2003). Kidney Int., 64, 321–330. Djerbi M, Screpanti V, Catrina I, Bogen B, Biberfeld P and Grandien A. (1999). J. Exp. Med., 190, 1025–1032. French L and Tschopp J. (1999). Nat. Med., 5, 146–147. French LE and Tschopp J. (2002). Semin. Cancer Biol., 12, 51–55. Frost P, Ng CP, Belldegrun A and Bonavida B. (1997). Cell. Immunol., 180, 70–83. Hahne M, Rimoldi D, Schroter M, Romero P, Schreier M, French LE, Schneider S, Bornard T, Fontana A, Lienard D, Cerottini J-C and Tschopp J. (1996). Science, 274, 1363–1366. Jaklitsch MT, Grondin SC and Sugarbaker DJ. (2001). World J. Surgery, 25, 210–217. Krueger A, Baumann S, Krammer PH and Kirchoff S. (2001). Mol. Cell. Biol., 21, 8247–8254. Kumar-Singh S, Weyler J, Martin M, Vermeulen P and Van Marck E. (1999). J. Pathol., 189, 72–78. Lew F, Tsang P, Holland J, N W, IJ S and JG B. (1986). J. Clin. Immunol., 6, 225–233.

Liu W, Bodle E, Chen JY, Gao M, Rosen GD and Broaddus VC. (2001). Am. J. Respir. Cell Mol. Biol., 25, 111–118. Manning L, Davis M and Robinson B. (1991). Clin. Exp. Immunol., 83, 85–91. Medema JP, Tong TJD, Hall Cjv and Melief Offringa R. (1999). J. Exp. Med., 190, 1033–1038. Mitsiades N, Mitsiades CS, Poulaki V, Anderson KC and Treon SP. (2002). Blood, 99, 2162–2171. Mutsaers SE. (2002). Respirology, 7, 171–191. Nagaraju K, Casciola-Rosen L, Rosen A, Thompson C, Loeffler L, Parker T, Danning C, Rochon P, Gillespie J and Plotz P. (2000). J. Immunol., 164, 5459–5465. Narasimhan SR, Yang L, Gerwin BI and Broaddus VC. (1998). Am. J. Physiol., 275 (Lung Cell. Mol. Physiol.), L165–L171. O’Connel J, O’Sullivan GC, Collins JK and Shanahan F. (1996). J. Exp. Med., 184, 1075–1082. Peter ME and Krammer PH. (2003). Cell Death Differ., 10, 26–35. Plasilova M, Zivny J, Jelinek J, Neuwirtova R, Cermak J, Necas E, Andera L and T S. (2002). Leukemia, 16, 67–73. Romano M, Catalano A, Nutini M, D’urbano E, Crescenzi C, Claria J, Libner R, Davi G and Procopio A. (2001). FASEB J., 15, 2326–2336. Roth W, Isenmann S, Nakamura M, Platten M, Wick W, Kleihues P, Bahr M, Ohgaki H, Ashkenazi A and Weller M. (2001). Cancer Res., 61, 2759–2765. Screpanti V, Wallin R, Ljunggren H-G and Grandien A. (2001). J. Immunol., 167, 2068–2073. Oncogene


7760 Siegmund D, Hadwiger P, Pfizenmaier K and Vornlocher H-P. (2002). Mol. Med., 8, 725–732. Stewart JH, Nguyen DM, Chen GA and Schrump DS. (2002). J. Thoracic Cardiovasc. Surg., 123, 295–302. Strizzi L, Catalano A, Vianale G, Orecchia S, Casalini A, Tassi G, Puntoni R, Mutti L and Procopio A. (2001). J. Pathol., 193, 468–475. Tepper CG and Seldin MF. (1999). Blood, 94, 1727–1737. Thomas RK, Kallenborn A, Wickenhauser C, Schultze JL, Draube A, Vockerodt M, Re D, Diehl V and Wolf J. (2002). Am. J. Pathol., 160, 1521–1528.

Oncogene

Tomasetti M, Rippo MR, Moretti S, Alleva R, Andera L, Neuzil J and Procopio A. (2004). Br. J. Cancer, 90, 1644–1653. Tomek S, Emrl S, Krejcy K and Manegold C. (2003). Br. J. Cancer, 88, 167–174. Vivo C, Liu W and Broaddus VC. (2003). J. Biol. Chem., 278, 25461–25467. Wall NR and Shi Y. (2003). The Lancet, 362, 1401–1403. Wright SC, Zhong J and Larrick JW. (1994). FASEB J., 8, 654–660. Zang D, Goodwin R, Loken M, Bryant E and Deeg J. (2001). Blood, 98, 3058–3065.