Oncogene (2003) 22, 1653–1662 & 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Interferon-a sensitizes human hepatoma cells to TRAIL-induced through DR5 upregulation and NF-jB inactivation

Masaya Shigeno1, Kazuhiko Nakao2, Tatsuki Ichikawa1, Kasumi Suzuki3, Atsushi Kawakami1, Seigou Abiru1, Seiji Miyazoe1, Yuichi Nakagawa1, Hiroki Ishikawa1, Keisuke Hamasaki1, Keisuke Nakata1, Nobuko Ishii2 and Katsumi Eguchi*,1

1First Department of Internal Medicine, Nagasaki University School of Medicine, 1-7-1, Sakamoto, Nagasaki 852-8501, Japan; 2Health Research Center, Nagasaki University, 1-7-1, Sakamoto, Nagasaki 852-8501, Japan; 3Department of Clinical Pharmacology, Nagasaki University, 1-7-1, Sakamoto, Nagasaki 852-8501, Japan

Tumor necrosis factor (TNF)-related apoptosis-inducing superfamily, is a promising candidate for cancer therapy (TRAIL), a member of the TNF superfamily, since it preferentially induces apoptosis in a variety of induces apoptosis in a variety of cancer cells with little or cancer cells with little or no effect on normal cells no effect on normal cells. Human hepatoma cells, (Sheridan et al., 1997; Ashkenazi et al., 1999; Walczak however, are resistant to TRAIL-induced apoptosis. Since et al., 1999). TRAIL triggers apoptosis through inter- interferon-a (IFN-a)is capable of enhancing TNF- a- action with death receptors DR4 (TRAIL-R1) and DR5 induced apoptosis in certain cancer cells, we evaluated the (TRAIL-R2) (Pan et al., 1997; Schneider et al., 1997), effect of IFN-a on TRAIL-induced apoptosis of human whereas TRAIL-R3, TRAIL-R4 and osteoprotegerin hepatoma cells. IFN-a pretreatment enhanced TRAIL- serve as decoy receptors to block TRAIL-mediated induced apoptosis of HuH-7 and Hep3B cells, in which apoptosis (Degli-Esposti et al., 1997a, b; Emery et al., IFN-a upregulated the expression of DR5, a death 1998). DR4 and DR5 do not only signal apoptosis receptor of TRAIL, and downregulated the expression through Fas-associated death domain (FADD) and of survivin, which has an antiapoptotic function. In -8 (Kischkel et al., 2000), but also activate contrast, IFN-a did not enhance TRAIL-induced apopto- nuclear factor-kappaB (NF-kB), which regulates the sis of HepG2 cells, in which expression of DR5 and expression of survival factors such as members of the survivin was not affected by IFN-a. On the other hand, inhibitor of apoptosis (IAP) family (c-IAP1, c-IAP2, TRAIL activated NF-jB composed of RelA-p50 hetero- XIAP) and Bcl-xL (Wang et al., 1998; Chen et al., 2000). dimer, a key regulating survival, These findings suggest the existence of several physio- in HuH-7 and HepG2 cells. However, IFN-a pretreatment logical mechanisms that protect normal cells against repressed the TRAIL-mediated activation of NF-jB and TRAIL-mediated apoptosis. Indeed, administration of decreased its transcriptional activity in HuH-7 but not in recombinant TRAIL to experimental animals, including HepG2 cells. Moreover, IFN-a pretreatment clearly mice and primates, induced significant tumor regression augmented TRAIL-mediated caspase-8 activation in without systemic toxicity (Ashkenazi et al., 1999; HuH-7 cells. Our results suggest that IFN-a could Walczak et al., 1999; Chinnaiyan et al., 2000). In sensitize certain human hepatoma cells to TRAIL-induced addition, a recent study revealed that recombinant apoptosis by stimulating its death signaling and by TRAIL, trimerized through the endogenous cysteine repressing the survival function in these cells. residue at position 230, does not induce apoptosis of Oncogene (2003) 22, 1653–1662. doi:10.1038/sj.onc.1206139 normal human hepatocytes (Lawrence et al., 2001), which have been reported to be susceptible to the other Keywords: TRAIL; interferon-a, hepatoma; DR5 type of recombinant TRAIL (Jo et al., 2000). On the survivin; NF-kB other hand, high concentrations of TRAIL are neces- sary to induce apoptosis of certain cancer cells, including human hepatoma cells (Yamanaka et al., 2000). Thus, combinations of TRAIL with several Introduction chemotherapeutic agents or radiation have been eval- uated for rendering such cells susceptible to TRAIL, and (TNF)-related apoptosis-indu- some combinations have successfully enhanced TRAIL- cing ligand (TRAIL), a novel member of the TNF mediated cancer cell death (Chinnaiyan et al., 2000; Gibson et al., 2000; Nagane et al., 2000; Sun et al., 2000; *Correspondence: Prof. Katsumi Eguchi, Chairman, First Depart- Yamanaka et al., 2000; Zhou et al., 2000; Belka et al., ment of Internal Medicine, Nagasaki University School of Medicine, 2001; Di Pietro et al., 2001; Lacour et al., 2001; 1-7-1 Sakamoto, Nagasaki 852-8501, Japan; E-mail: [email protected] Nimmanapalli et al., 2001). Received 13 May 2002; revised 7 October 2002; accepted 16 Interferon (IFN)-a/b plays an essential role in both October 2002 antiviral and antitumor defenses and has been used Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1654 clinically for the treatment of viral infections and NF-kB, and also analysed whether IFN-a regulated the malignancies (Gutterman, 1994). Several investigators expression of TRAIL, DR4, DR5 and apoptosis-related have recently reported that IFN-a/b stimulates the in these cells. expression of TRAIL in T cells or dendritic cells, resulting in upregulation of their cytotoxic activities (Fanger et al., 1999; Kayagaki et al., 1999), and that the Results apoptosis observed in several cancer cells susceptible to IFN-a/b is mediated by induction of endogenous IFN-a pretreatment sensitizes human hepatoma cells to TRAIL in these cells (Chawla-Sarkar et al., 2001; Chen TRAIL-induced apoptosis et al., 2001; Toomey et al., 2001). Taken together, these findings suggest that IFN-a/b may, at least in part, To elucidate the effect of IFN-a on TRAIL-induced exhibit its antitumor activity through induction of apoptosis of human hepatoma cells, HuH-7, HepG2, TRAIL. Conversely, IFN-a/b has been reported to Hep3B and PLC/PRF/5 were preincubated with or sensitize human cervical cancer cells or Jurkat lympho- without 1000 IU/ml of IFN-a for 48 h, followed by ma cells to TNF-a-induced apoptosis through inhibition treatment with varying concentrations of TRAIL for of NF-kB activation (Manna et al., 2000; Suk et al., 24 h. As shown in Figure 1, TRAIL dose-dependently 2001). Since both TRAIL and TNF-a trigger death reduced cell viability of HuH-7 and HepG2, but higher signaling and activate NF-kB (Yamanaka et al., 2000; concentrations of TRAIL were required to elicit Ravi et al., 2001), we hypothesized that IFN-a/b could significant cytotoxicity. Hep3B and PLC/PRF/5 were sensitize cancer cells to TRAIL-induced apoptosis. In much more resistant to TRAIL-induced cytotoxicity. In the present study, we investigated whether IFN-a contrast, preincubation with IFN-a enhanced the pretreatment could augment TRAIL-induced apoptosis TRAIL-induced cytotoxicity in HuH-7, Hep3B and of human hepatoma cells that exhibit resistance to PLC/PRF/5, but not in HepG2, although IFN-a alone TRAIL. In addition, we evaluated the effect of IFN-a could reduce cell viability in PLC/PRF/5. Next, to on TRAIL-mediated activation of both caspase-8 and examine whether IFN-a pretreatment actually enhanced TRAIL-induced apoptosis of HuH-7 and Hep3B, we examined the content of hypodiploid DNA in those cells. IFN-a pretreatment enhanced TRAIL-induced apoptosis of HuH-7 and Hep3B (Figure 2). However, such enhancement was not observed in HepG2 cells. These results indicated that IFN-a sensitized HuH-7 and Hep3B, but not HepG2 cells, to TRAIL-induced apoptosis.

IFN-a upregulates DR5 expression and downregulates survivin expression in HuH-7 and Hep3B, but not in HepG2 Recent studies have shown that expression of death receptors, DR4 or DR5, is upregulated in cancer cells by chemotherapeutic agents or by radiation, resulting in enhancement of TRAIL-induced apoptosis of these cells (Gibson et al., 2000; Nagane et al., 2000; Sun et al., 2000; Zhou et al., 2000; Di Pietro et al., 2001; Lacour et al., 2001; Nimmanapalli et al., 2001). We therefore examined the effect of IFN-a on expression of DR4 and DR5, as well as the expression of several apoptosis- related proteins, including procaspase 8, FLICE-inhibi- tory (FLIP), CIAP2, XIAP, survivin, Bcl-xL and Bax, by Western blotting. As shown in Figure 3, these proteins were constitutively expressed in HuH-7, HepG2 and Hep3B cells. DR5 expression was upregu- lated by IFN-a in HuH-7 and Hep3B but not in HepG2 cells. On the contrary, the expression of survivin, a Figure 1 Effects of IFN-a on TRAIL-induced cytotoxicity in member of the IAP family that has an antiapoptotic human hepatoma cells. (a) HuH-7, (b) HepG2, (c) Hep3B and (d) function, was repressed by IFN-a in HuH-7 and Hep3B PLC/PRF/5 cells were preincubated with or without 1000 IU/ml of but not in HepG2 cells. These results appeared to be IFN-a for 48 h, followed by treatment with varying concentrations parallel to the effect of IFN-a on TRAIL-induced of TRAIL for 24 h. Cell viability was then determined by the colorimetric method. Results are expressed as a percentage of the apoptosis in these cells. Conversely, the expression of control. Data represent the mean7s.d. values of the four procaspase 8, FLIP and XIAP was almost unchanged, experiments irrespective of the presence of IFN-a. DR4 expression

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1655

Figure 2 Effect of IFN-a on TRAIL-induced apoptosis of human hepatoma cells. HuH-7, HepG2 and Hep3B cells were preincubated with or without 1000 IU/ml of IFN-a for 48 h, followed by treatment with or without 100 ng/ml of TRAIL for 24 h. Cells were then stained with propidium iodide and subjected to DNA content analysis by flow cytometry. The percentages of cells with hypodiploid DNA are indicated. Results shown are from one representative experiment from a total of four performed was only slightly stimulated by IFN-a in HuH-7 but not to apoptosis induced by DR5-specific agonistic antibody in other cells. CIAP2 expression was clearly repressed in (Figure 6b) as well as to apoptosis induced by TRAIL Hep3B and slightly repressed in HuH-7 but not in (Figure 1). These results suggested that up-regulation of HepG2 cells. The effects of IFN-a on expression of Bcl- DR5 by IFN-a could contribute to enhanced TRAIL xL and Bax differed among these cells, and even sensitivity in HuH-7 cells. Next, we also examined the between HuH-7 and Hep3B, in which IFN-a equally involvement of survivin in resistance against TRAIL- enhanced TRAIL-induced apoptosis (Figure 2). DR5 mediated apoptosis of HuH-7 cells. In comparison with and DR4 expression on the cell membrane was also the vehicle vector transfection, transfection of the analysed by flow cytometry. As shown in Figure 4, IFN- survivin expression vector which produced the flag/ a upregulated the surface expression of DR5 in HuH-7 survivin fusion protein (Figure 7a) partially but dose- and Hep3B but not in HepG2 cells. On the other hand, dependently inhibited TRAIL/IFN-a-induced apoptosis the surface expression of DR4 was slightly stimulated by of HuH-7 cells (Figure 7b). These results suggested that IFN-a in HuH-7 but not in other cells. These results survivin might play a role in resistance against TRAIL- were almost similar to those of Western blotting. We mediated apoptosis of HuH-7 cells. next examined the effect of IFN-a on mRNA levels of TRAIL, DR4, DR5, survivin and several apoptosis- IFN-a inhibits TRAIL-mediated NF-kB activation in related in HuH-7 cells by RNase protection assay HuH-7 but not in HepG2 (Figure 5). IFN-a clearly upregulated the transcripts of DR5, and weakly stimulated the transcripts of DR4 and Since TRAIL can both trigger death signaling and TRAIL, which appeared as faint bands (Figure 5a). activate NF-kB, thus regulating the expression of Procaspase 8 and CIAP2 mRNAs were also upregulated several factors involved in cell survival, we examined by IFN-a (Figure 5a and b) although this was not the effect of IFN-a on TRAIL-mediated transcriptional detectable by Western blotting (Figure 3). In contrast, activity of NF-kB in HuH-7 and HepG2 cells by survivin mRNA expression was repressed by IFN-a reporter transfection assay. TRAIL dose-depen- (Figure 5b). dently stimulated luciferase activity from the plasmid To clarify the functional role of DR5 in enhancement pNFkB-luc (containing four repeats of NF-kB binding of TRAIL-mediated apoptosis by IFN-a, we examined sequences) in both cells. However, IFN-a pretreatment the effect of DR5-specific blocking chimera antibody on suppressed this induction in HuH-7 but not in HepG2 TRAIL/ IFN-a-induced apoptosis and also checked the cells (Figure 8). We then used EMSA to determine impact of IFN-a on apoptosis triggered by DR5-specific whether IFN-a could inhibit the TRAIL-mediated NF- agonistic antibody in HuH-7 cells. Addition of DR5- kB binding activity to oligonucleotides containing a NF- specific blocking chimera antibody dose-dependently kB binding (kB) sequence. Nuclear extracts from inhibited TRAIL/ IFN-a-induced apoptosis of HuH-7 unstimulated HuH-7 cells formed a complex band with cells (Figure 6a). In addition, IFN-a sensitized the cells the oligonucleotide probe (Figure 9a, lane 1). Addition

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1656

Figure 4 Flow cytometric analysis of DR4 and DR5 expression in human hepatoma cells. HuH-7, HepG2 and Hep3B cells were incubated with or without 1000 IU/ml of IFN-a for 48 h, and membrane expression of DR4 and DR5 was analysed by FACScan. Figure 3 Effect of IFN-a on the expression of DR4, DR5 and The horizontal and vertical bars indicate the fluorescence intensity several apoptosis-related proteins in human hepatoma cells. HuH- and the relative number of cells, respectively. Shadowed histograms 7, HepG2 and Hep3B cells were incubated with or without represent IFN-a-treated cells, and clear histograms represent 1000 IU/ml of IFN-a for 48 h, and the expression of DR4, DR5, untreated cells. Histograms with gray lines represent negative pro-caspase 8, FLIP, CIAP2, XIAP, survivin, Bcl-xL, Bax and b- controls. The percentages of positive cells are indicated. Results actin in the cells was analysed by Western blotting using the shown are from one representative experiment from a total of three appropriate antibodies. Results shown are from one representative performed experiment from a total of four performed. The density of each band was quantified with NIHimage analysis software using b- actin as reference. Data are expressed relative to the density of control; without IFN-a in each cell (lanes 1, 3 and 5) treatment for 2 h did not form such a slowly migrating band (Figure 9b, lane 4). These results indicated that TRAIL activated the NF-kB composed of RelA-p50 of a 100-times molar excess of the nonlabeled NF-kB heterodimer in HuH-7, which was inhibited by IFN-a oligonucleotide reduced the density of the complex band pretreatment. Such a slowly migrating band was also (Figure 9a, lane 5), but addition of nonlabeled AP-1 formed in HepG2 cells treated by TRAIL (which was oligonucleotide did not (Figure 9a, lane 6). Moreover, not inhibited by IFN-a pretreatment) (Figure 9b, lanes 7 addition of anti-p50 but not anti-c-Rel or anti-RelA led and 8), although similar supershifts were detected by to a supershift of the band (Figure 9a, lanes 2–4), addition of anti-p50 or anti-RelA (data not shown). suggesting that NF-kB composed of p50 homodimer was interacting with the kB sequence under nonstimu- IFN-a enhances TRAIL-induced caspase-8 activity in lated conditions. Nuclear extract from HuH-7 cells HuH-7 treated with TRAIL for 2 h formed an additional slowly migrating band (Figure 9a, lane 7 and Figure 6b, lane 3), Finally, we examined the effect of IFN-a on the TRAIL- which was reduced by addition of nonlabeled NF-kB mediated caspase-8 activity of cleaving IETD-pNA- oligonucleotide (Figure 9a, lane 11) and led to a conjugated substrate. TRAIL time-dependently induced supershift by addition of both anti-p50 and anti-RelA caspase-8 activity in HuH-7 cells. On the other hand, but not by addition of anti-c-Rel (Figure 9a, lanes 8–10). IFN-a pretreatment clearly enhanced its activity, which In contrast, nuclear extract from HuH-7 cells preincu- reached a peak level 4 hour after addition of TRAIL bated with IFN-a for 48 h, followed by TRAIL (Figure 10).

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1657

Figure 5 Effect of IFN-a on mRNA levels of DR4, DR5, TRAIL, pro-caspase 8 and IAP family genes in HuH-7 cells. HuH-7 cells were incubated with or without 1000 IU/ml of IFN-a for 48 h. Total RNA (10 mg) from the cells was subjected to analysis for mRNA levels of DR4, DR5, TRAIL, procaspase 8 and IAP family genes by RNase protection assay using human apoptosis multiprobe template sets, hAPO3-c (A) and hAPO-5c (B), respectively. Results shown are from one representative experiment from a total of three performed

Discussion of survivin by IFN-a may, at least in part, account for the enhancement of TRAIL-mediated apoptosis by In this study, we have demonstrated that IFN-a IFN-a. In contrast, several investigators have demon- pretreatment sensitized HuH-7 and Hep3B, but not strated that IFN-a/b stimulated TRAIL expression in HepG2, cells to TRAIL-induced apoptosis, and that certain cancer cells (Toomey et al., 2001; Chen et al., IFN-a upregulated the expression of DR5 and down- 2001), which could be involved in the antitumor regulated the expression of survivin in HuH7 and activities of IFN-a. Chawla-Sarkar et al. (2001) reported Hep3B, but not in HepG2 cells. Recent studies have that IFN-b, rather than IFN-a, induced TRAIL indicated that upregulation of DR5 by chemotherapeu- expression in cells. We also showed that tic agents or radiation sensitizes many cancer cells to IFN-a weakly stimulated TRAIL gene expression in TRAIL-induced apoptosis through activation of death HuH-7 cells, which may not, however, be the major signaling (Gibson et al., 2000; Nagane et al., 2000; Sun cause of the enhancement of TRAIL-induced apoptosis et al., 2000; Zhou et al., 2000; Lacour et al., 2001; by IFN-a because IFN-a alone could not induce Nimmanapalli et al., 2001). In addition, we have shown significant apoptosis of HuH-7 cells (Figures 1 and 2). that DR5-specific blocking chimera antibody inhibited The NF-kB transcription factor family consists of TRAIL/ IFN-a-induced apoptosis in HuH-7 cells, and several structurally related proteins such as c-Rel, RelA, that IFN-a could sensitize the cells to apoptosis induced RelB, p50/p105, and p52/p100, which form homo- or by DR5 specific agonistic antibody. Taken together, it heterodimers with each other and regulate the expression seems that upregulation of DR5 by IFN-a enhances of a number of genes (Barkett and Gilmore, 1999). In the TRAIL sensitivity. Survivin, which is expressed in present study, nuclear extracts from HuH-7 and HepG2 cancer tissues including but cells exhibited binding activity of NF-kBcomposedofa not in normal tissues, represses the activities of p50 homodimer without any stimulation, as reported in (Tamm et al., 1998; Ito et al., 2000), resulting in other types of cells (Inan et al., 2000). TRAIL induced resistance of cancer cells to Fas- or chemotherapeutic additional binding activity of NF-kB comprising a RelA- agent-mediated apoptosis (Li et al., 1998; Tamm et al., p50 heterodimer in these cells. Ravi et al. (2001) have 1998). In addition, inhibition of survivin expression, by recently reported that RelAÀ/À mouse fibroblasts are transduction with antisense oligonucleotides or a highly sensitive to TRAIL-induced apoptosis, and that dominant-negative expression vehicle, sensitized cancer anti-CD40-mediated activation of NF-kB, including cells to chemotherapeutic agent-mediated apoptosis RelA, effectively blocked TRAIL-induced apoptosis. (Olie et al., 2000; Grossman et al., 2001). Recently, Therefore, NF-kB composed of RelA-p50 appears to Griffith et al. (2002) reported that the cellular level of play a key role in the resistance of HuH-7 and HepG2 survivin is closely relevant to the resistance against cells to TRAIL-induced apoptosis. Indeed, IFN-a TRAIL-mediated apoptosis of pretreatment inhibited TRAIL-mediated activation of cells. We also demonstrated that ectopic expression of RelA-p50 NF-kB in HuH-7 cells that were sensitized to survivin repressed TRAIL/IFN-a-induced apoptosis of TRAIL-induced apoptosis by IFN-a. In contrast, IFN-a HuH-7 cells. These findings suggest that downregulation pretreatment could not inhibit activated HepG2 cells that

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1658

Figure 7 Effects of ectopic expression of survivin on TRAIL/IFN- a-induced apoptosis. (a) HuH-7 cells in six-well multiplates were transfected with 2 mg of flag/survivin expression vector (lanes 3 and Figure 6 Effect of DR5-specific blocking chimera antibody on 4) or vehicle vector alone (lanes 1 and 2) and incubated with or TRAIL/ IFN-a-induced apoptosis and effect of IFN-a on apoptosis without 1000 IU/ml of IFN-a for 48 h. Then, the expression of triggered by DR5-specific agonistic antibody in HuH-7 cells. (a) survivin or flag/survivin fusion protein was detected by anti- HuH-7 cells were preincubated with or without 1000 IU/ml of IFN- survivin antibody (upper column) or anti-flag antibody (lower a for 48 h, followed by treatment with or without 50 ng/ml of column), respectively. (b) HuH-7 cells in 24-well multiplates were TRAIL for 24 h in the presence of indicated concentrations of DR5- transfected with indicated amount of survivin expression vector specific blocking chimera antibody. (b) HuH-7 cells were preincu- (lanes 4 and 5) or vehicle alone (lanes 1–3) and incubated with or bated with or without 1000 IU/ml of IFN-a for 48 h, followed by without 1000 IU/ml of IFN-a for 48 h, followed by treatment with treatment with varying concentrations of DR5-specific agonistic or without 50 ng/ml of TRAIL for 24 h. Cell viability was then antibody. Cell viability was determined by the colorimetric method. determined by the colorimetric method. Results are expressed as Results are expressed as a percentage of the control. Data represent percentages of the control. Data represent the mean7s.d. values of 7 the mean s.d. values of four experiments (a and b) four experiments

remained resistant to TRAIL-induced apoptosis even in activity by IFN-a may be also involved in the stimulation the presence of IFN-a. Similar observations have been of caspase-8 activity, since the NF-kB-induced products, reported where IFN-a suppressed activation of RelA-p50 c-IAP, c-IAP2 and TNFR-associated factor-1 (TRAF1), NF-kB and potentiated TNF-a-induced apoptosis of are known to cooperatively block caspase-8 activity human cervical cancer cells and Jurkat lymphoma cells (Wang et al., 1998). However, why IFN-a did not inhibit (Manna et al., 2000; Suk et al., 2001). Furthermore, Wen NF-kB activity in HepG2 cells, even though IFN-a can et al. (2000) have recently reported that overexpression of stimulate the expression of IFN-a-inducible genes in p202, an IFN-a-inducible protein, is capable of sensitiz- HepG2 and HuH-7 cells (Ichikawa et al., 2001), remains ing breast cancer cells to TNF-a-induced apoptosis to be clarified. through inactivation of NF-kB by its interaction with In conclusion, we have demonstrated in the present p202. Taken together, these results indicate that IFN-a study that IFN-a potentiated the apoptotic effect of could possibly sensitize HuH-7 cells to TRAIL-induced TRAIL in human hepatoma cells by regulating the apoptosis by inhibiting the activation of NF-kBincluding expression of DR5 or survivin and by repressing RelA. We also demonstrated that IFN-a stimulated TRAIL-mediated NF-kB activation. These results sug- TRAIL-mediated caspase-8 activity in HuH-7 cells, gest that IFN-a may exhibit antitumor activity not only which could be explained by IFN-a-induced upregulation by endogenous TRAIL induction, but also by sensitizing of DR5 in these cells. In addition, suppression of NF-kB cancer cells to TRAIL-mediated apoptosis, and that

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1659

Figure 8 Effects of IFN-a on TRAIL-mediated transcriptional activity of NF-kB in HuH-7 and HepG2 cells. PNFkB-luc was cotransfected with pRL-CMV-luc into HuH-7 (a) and HepG2 (b) cells. After 6 h hours later, the cells were incubated with or without 1000 IU/ml of IFN-a for 36 h, followed by treatment with the indicated concentrations of TRAIL for 7 h. Luciferase activity in the cells was then analysed by dual-luciferase assay. Data represent the ratios of firefly-luc activity derived from pNFkB-luc over renilla-luc activity derived from pRL-CMV-luc relative to the control (no treatment), and are expressed as mean7s.d. of three separate experiments

TRAIL, in combination with IFN-a, may have ther- azide, 0.1% SDS, 100 mg/ml PMSF, 1 mg/ml of aprotinin, 1% apeutic potential in the treatment of human hepatocel- NP40, and 0.5% sodium deoxycholate) for 10 min at 41C and lular carcinoma. passed several times through a 25-gauge needle, and the protein concentration was determined using a Bio-Rad protein assay (Melville, NY, USA). The same amount of protein Materials and methods from each lysate (20 mg/well) was subjected to 8–16% SDS– PAGE. Proteins were transferred onto nitrocellulose mem- Cell culture and detection of apoptosis branes, that were then blocked for 1.5 h using 5% nonfat dried The human hepatoma cell lines, HepG2, Hep3B and PLC/ milk in PBS containing 0.1% Tween 20 (PBS-T), washed with PRF/5 were maintained in a chemically defined medium, IS- PBS-T and incubated at room temperature for 1 h in the RPMI (Nakabayashi et al., 1984) containing 10% fetal bovine presence of each antibody (rabbit polyclonal anti-human DR4, serum (FBS). HuH-7 was maintained in IS-RPMI containing Imgenex, San Diego, CA, USA; rabbit polyclonal anti-human 5% FBS. Recombinant human TRAIL (leucine zipper DR5, R&D systems; mouse monoclonal anti-human caspase- construct) was supplied by Immunex Co. (Seattle, WA, 8, MBL, Nagoya, Japan; rabbit polyclonal anti-human FLIP, USA). Recombinant human IFN-a 2a was provided by PharMingen, San Diego, CA, USA; rabbit polyclonal anti- Nippon Roche Co. (Tokyo, Japan). To analyse cell viability, human Bax, Santa Cruz Biotechnology, Santa Cruz, CA, 5 Â 103 cells were placed into 96-well multiplates. A day later, USA; mouse monoclonal anti-human Bcl-xL, Trevigen, the medium was replaced with fresh medium in the presence or Gaithersburg, MD, USA; mouse monoclonal anti-human absence of 1000 IU/ml of IFN-a, and the cells were incubated XIAP, MBL; rabbit polyclonal anti-human survivin, Alpha for 48 h, followed by treatment with varying concentrations of Diagnostic International Inc., San Antonio, TX, USA; rabbit TRAIL for 24 h. In some experiments, varying concentrations polyclonal anti-human c-IAP2, Santa Cruz Biotechnology; of recombinant human TRAIL R2/Fc chimera, DR5-specific mouse monoclonal anti-flag antibody, Sigma; mouse mono- blocking chimera antibody (R&D systems, Minneapolis, MN, clonal anti-human b-actin (as an internal control for Western USA), or goat anti-human TRAIL R2/DR5 antibody, and a blot analysis), Sigma). The membranes were washed with PBS- DR5-specific agonistic antibody (R&D systems), were added T and incubated with horseradish peroxidase-conjugated anti- to the cell cultures. Cell viability was determined by the rabbit IgG or anti-mouse IgG for 1 h. Following washing with colorimetric method using a Cell Counting kit (Wako Life PBS-T, immunoreactive bands were visualized by the ECL Science, Osaka, Japan). The absorbance of each well was chemiluminescence system (Amersham Life Science, Buckin- measured at 405 nm with a microtiter plate reader (Multiskan ghamshire, UK). JX, Thermo BioAnalysis Co., Japan). DNA fragmentation was quantified by the percentage of cells with hypodiploid Flow cytometry for DR4 and DR5 DNA. In brief, cells were fixed with 70% ethanol and treated Cells were analysed for the surface expression of DR4 and with RNase (100 mg/ml, Sigma Chemical Co., St Louis, MO, DR5 by indirect staining with primary goat anti-human DR4 USA) and then stained with propidium iodide (100 mg/ml, and DR5 (R&D systems), followed by phycoerythrin (PE)- Sigma) for 30 min on ice. The stained cells were analysed by a conjugated rabbit anti-goat IgG (Sigma). Briefly, cells (5 105 flow cytometer (Epics XL, Beckman Coulter, Hialeah, FL, Â ) USA) to detect the presence of cells with hypodiploid DNA. were stained with 200 ml PBS containing saturating amounts of anti-DR4 or anti-DR5 on ice for 30 min. After incubation, cells were washed twice, and reacted with PE-conjugated Western blotting rabbit anti-goat IgG on ice for 30 min. After double washing Cells were washed three times with PBS, lysed by addition of with PBS, the expression of DR4 and DR5 was analysed by lysis buffer (50 mm Tris (pH8.0), 150 m m NaCl, 0.02% sodium FACScan (Epics XL, Beckman Coulter).

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1660

Figure 9 Effect of IFN-a on DNA-binding activity of NF-kB induced by TRAIL in HuH-7 and HepG2 cells. (a) Nuclear extracts from HuH-7 cells incubated in the absence (lanes 1–6) or presence (lanes 7–12) of 50 ng/ml TRAIL for 2 h were subjected to EMSA using a 32P-labeled NF-kB oligonucleotide probe. To identify the subunits of NF-kB, the nuclear extract was preincubated for 30 min with the rabbit polyclonal anti-p50 (lanes 2 and 8), anti-c-Rel (lanes 3 and 9) and anti-RelA (lanes 4 and 10). For analysis of the specific binding to the kB sequence, a 100-times molar excess of the nonradiolabeled NF-kB oligonucleotides (lanes 5 and 11) or AP-1 oligonucleotides (lanes 6 and 12) was added to the sample as a competitor. (b) Nuclear extracts from HuH-7 and HepG2 cells preincubated with (lanes 2, 4, 6 and 8) or without (lanes 1, 3, 5 and 7) 1000 IU/ml of IFN-a, followed by treatment with (lanes 3, 4, 7 and 8) or without (lanes 1, 2, 5 and 6) 50 ng/ml of TRAIL for 2 h were subjected to EMSA using a 32P-labeled NF-kB oligonucleotide probe. Results shown are from one representative experiment from a total of three performed

RNase protection assay Madison, WI, USA) containing the CMV immediate early enhancer/ and renilla luciferase gene were used in the The RNase protection assay was performed using a Ribo- assay. Cells were grown in 24-well multiplates in triplicate the Quant Multi-Probe RNase Protection Assay System (Phar- day before transfection. In the next step, 200 ng of pNFkB-luc Mingen). According to the instructions provided by the and 10 ng of pRL-CMV-luc were transfected into the cells by manufacturer, hAPO3c (death receptors and death ligands) the lipofection method in serum-free medium. After a 6-h and hAPO5c (inhibitors of apoptosis) template sets, including incubation, the medium was replaced with fresh medium a ribosomal protein L32 and a glyceraldehyde-3-phosphate containing 10% FBS with or without 1000 IU/ml of IFN-a, dehydrogenase (GAPDH) template as internal controls, were 32 and the cells were incubated for 36 h, followed by treatment labeled with [a P] UTP using T7 RNA polymerase. The with varying concentrations of TRAIL for 7 h. Luciferase labeled RNA probes were hybridized with 10 mg of total RNA activity in the cells was then determined by a dual-luciferase from HuH-7 cells treated with or without 1000 IU/ml of IFN-a reporter assay system and a TD-20/20 luminometer (Prome- for 48 h. Samples were digested with RNase to remove single- ga). In some experiments, pCR2FL; a flag/survivin fusion stranded (nonhybridized) RNA. The remaining probes were protein expression vector (Kobayashi et al., 1999), kindly resolved on 6% urea-polyacrylamide-bis-acrylamide gels. Gels provided by Prof. T Tokuhisa (Department of Developmental were dried and analysed using an image analyser (BAS, Fuji Genetics, Chiba University) or pCR3.1, a vehicle vector Film Co., Tokyo). (Invitrogen, Carlsbad, CA, USA), was transfected into HuH- 7 cells cultured in 6- or 24-well multiplates. Cell transfection and luciferase assay Electrophoretic mobility shift assay (EMSA) The plasmids pNFkB-luc (Stratagene, La Jolla, CA, USA) containing four copies of the binding sequence of NF-kB and Cells were incubated in the presence or absence of 1000 IU/ml the firefly luciferase gene, and pRL-CMV-luc (Promega, of IFN-a for 48 h, followed by treatment with or without

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1661 [g-32P]ATP using T4 polynucleotide kinase. Nuclear extracts (4 mg) were incubated with 10 fmol of the labeled probe for 30 min at room temperature in the presence of 40 mm KCl, 20 mm HEPES (pH 7.9), 1 mm MgCl2, 0.1 mm EDTA, 1 mm dithiothreitol, 2 mg of poly (dI-dC), and 8% glycerol. The reaction mixture was electrophoresed on a 4% polyacrylamide gel containing 25 mm Tris-borate and 0.25 mm EDTA. Gels were dried and analysed using an image analyzer.

Caspase-8 activity Caspase-8 activity was determined using the ApoAlert Caspase-8 Colorimetric Assay kit (Clontech Laboratories, Palo Alto, CA, USA). In brief, cell pellets obtained from 2 Â 106 cells were resuspended in 50 ml of chilled lysis buffer and incubated on ice for 10 min. The cell lysates were centrifuged at 12 000 r.p.m. at 41C for 3 min, and the supernatants were collected. Reaction buffer (50 ml) containing 10 mm dithio- threitol and 5 mlof1mm IETD-pNA (N-acetyl-Ile-Glu-Thr- Asp-p-nitroaniline) substrate was added to the supernatants and incubated at 371C for 2 h in the dark. The absorbance of Figure 10 Effects of IFN-a on the TRAIL-mediated caspase-8 activity in HuH-7 cells. HuH-7 was preincubated with or without each sample was measured at 405 nm with a microtiter plate 1000 IU/ml of IFN-a for 48 h, followed by treatment with 100 ng/ reader. ml of TRAIL for the indicated periods, and caspase-8 activity cleaving IETD-pNA substrate was determined by caspase-8 colorimetric assay. Data represent the mean7s.d. values of four Abbreviations experiments EMSA, electrophoretic mobility shift assay; FADD, Fas- associated death domain; FLIP, FLICE-inhibitory protein; IAP, inhibitor of apoptosis; IETD-pNA, N-acetyl-Ile-Glu- 50 ng/ml of TRAIL for 2 h, and the nuclear extract was Thr-Asp-p-nitroaniline; IFN-a, interferon-a; TNF, tumor prepared as described previously (Nakao et al., 1999). Rabbit necrosis factor; TRAIL, tumor necrosis factor-related apop- polyclonal anti-human p50, anti-human c-Rel and anti-human tosis-inducing ligand; XIAP, X-chromose–linked IAP. RelA were purchased from Santa Cruz Biotechnology. The following double-stranded oligonucleotides were used as probes or competitors in the assays (only the sense strand is Acknowledgements shown): NF-kB oligonucleotide, AGTTGAGGG- We thank Prof. T Tokuhisa (Department of Developmental GACTTTCCCAGG; AP-1 oligonucleotide, CGCTTGAT- Genetics, Chiba University, Japan) for generously providing GAGTCAGCCGGAA. The probe was end-labeled with pCR2FL.

References Ashkenazi A, Pai RC, Fong S, Leung S, Lawrence DA, Di Pietro R, Secchiero P, Rana R, Gibellini D, Visani G, Marsters SA, Blackie C, Chang L, McMurtrey AE, Hebert Bemis K, Zamai L, Miscia S and Zauli G. (2001). Blood, 97, A, DeForge L, Koumenis IL, Lewis D, Harris L, Bussiere J, 2596–2603. Koeppen H, Shahrokh Z and Schwall RH. (1999). J. Clin. Emery JG, McDonnell P, Burke MB, Deen KC, Lyn S, Invest., 104, 155–162. Silverman C, Dul E, Appelbaum ER, Eichman C, DiPrinzio Barkett M and Gilmore TD. (1999). Oncogene, 18, 6910–6924. R, Dodds RA, James IE, Rosenberg M, Lee JC and Young Belka C, Schmid B, Marini P, Durand E, Rudner J, Faltin H, PR. (1998). J. Biol. Chem., 273, 14363–14367. Bamberg M, Schulze-Osthoff K and Budach W. (2001). Fanger NA, Maliszewski CR, Schooley K and Griffith TS. Oncogene, 20, 2190–2196. (1999). J. Exp. Med., 190, 1155–1164. Chawla-Sarkar M, Leaman DW and Borden EC. (2001). Clin. Gibson SB, Oyer R, Spalding AC, Anderson SM and Johnson Cancer Res., 7, 1821–1831. GL. (2000). Mol. Cell Biol., 20, 205–212. Chen C, Edelstein LC and Gelinas C. (2000). Mol. Cell Biol., Griffith TS, Fialkov JM, Scott DL, Azuhata T, Williams RD, 20, 2687–2695. Wall NR, Altieri DC and Sandler AD. (2002). Cancer Res., Chen Q, Gong B, Mahmoud-Ahmed AS, Zhou A, Hsi ED, 62, 3093–3099. Hussein M and Almasan A. (2001). Blood, 98, Grossman D, Kim PJ, Schechner JS and Altieri DC. (2001). 2183–2192. Proc. Natl. Acad. Sci. USA, 98, 635–640. Chinnaiyan AM, Prasad U, Shankar S, Hamstra DA, Gutterman JU. (1994). Proc. Natl. Acad. Sci. USA, 91, 1198– Shanaiah M, Chenevert TL, Ross BD and Rehemtulla A. 1205. (2000). Proc. Natl. Acad. Sci. USA, 97, 1754–1759. Ichikawa T, Nakao K, Nakata K, Hamasaki K, Takeda Y, Degli-Esposti MA, Dougall WC, Smolak PJ, Waugh JY, Kajiya Y, Higashi S, Ohkubo K, Kato Y, Ishii N and Eguchi Smith CA and Goodwin RG. (1997a). Immunity, 7, 813–820. K. (2001). Biochem. Biophys. Res. Commun., 280, 933–939. Degli-Esposti MA, Smolak PJ, Walczak H, Waugh J, Huang Inan MS, Rasoulpour RJ, Yin L, Hubbard AK, Rosenberg CP, DuBose RF, Goodwin RG and Smith CA. (1997b). J. DW and Giardina C. (2000). Gastroenterology, 118, 724– Exp. Med., 186, 1165–1170. 734.

Oncogene Enhancement of TRAIL-induced apoptosis by IFN-a M Shigeno et al 1662 Ito T, Shiraki K, Sugimoto K, Yamanaka T, Fujikawa K, Ito Pan G, O’Rourke K, Chinnaiyan AM, Gentz R, Ebner R, Ni J M, Takase K, Moriyama M, Kawano H, Hayashida M, and Dixit VM. (1997). Science, 276, 111–113. Nakano T and Suzuki A. (2000). Hepatology, 31, Ravi R, Bedi GC, Engstrom LW, Zeng Q, Mookerjee B, 1080–1085. Gelinas C, Fuchs EJ and Bedi A. (2001). Nat. Cell Biol., 3, Jo M, Kim TH, Seol DW, Esplen JE, Dorko K, Billiar TR and 409–416. Strom SC. (2000). Nat. Med., 6, 564–567. Schneider P, Thome M, Burns K, Bodmer JL, Hofmann K, Kayagaki N, Yamaguchi N, Nakayama M, Eto H, Okumura Kataoka T, Holler N and Tschopp J. (1997). Immunity, 7, K and Yagita H. (1999). J. Exp. Med., 189, 1451–1460. 831–836. Kischkel FC, Lawrence DA, Chuntharapai A, Schow P, Kim Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, KJ and Ashkenazi A. (2000). Immunity, 12, 611–620. Baldwin D, Ramakrishnan L, Gray CL, Baker K, Wood Kobayashi K, Hatano M, Otaki M, Ogasawara T and WI, Goddard AD, Godowski P and Ashkenazi A. (1997). Tokuhisa T. (1999). Proc. Natl. Acad. Sci. USA, 96, 1457– Science, 277, 818–821. 1462. Suk K, Kim YH, Chang I, Kim JY, Choi YH, Lee KY and Lacour S, Hammann A, Wotawa A, Corcos L, Solary E and Lee MS. (2001). FEBS Lett., 495, 66–70. Dimanche-Boitrel MT. (2001). Cancer Res., 61, 1645–1651. Sun SY, Yue P, Hong WK and Lotan R. (2000). Cancer Res., Lawrence D, Shahrokh Z, Marsters S, Achilles K, Shih D, 60, 7149–7155. Mounho B, Hillan K, Totpal K, DeForge L, Schow P, Tamm I, Wang Y, Sausville E, Scudiero DA, Vigna N, Hooley J, Sherwood S, Pai R, Leung S, Khan L, Gliniak B, Oltersdorf T and Reed JC. (1998). Cancer Res., 58, 5315– Bussiere J, Smith CA, Strom SS, Kelley S, Fox JA, Thomas 5320. D and Ashkenazi A. (2001). Nat. Med., 7, 383–385. Toomey NL, Deyev VV, Wood C, Boise LH, Scott D, Liu LH, Li F, Ambrosini G, Chu EY, Plescia J, Tognin S, Marchisio Cabral L, Podack ER, Barber GN and Harrington Jr WJ. PC and Altieri DC. (1998). Nature, 396, 580–584. (2001). Oncogene, 20, 7029–7040. Manna SK, Mukhopadhyay A and Aggarwal BB. (2000). J. Walczak H, Miller RE, Ariail K, Gliniak B, Griffith TS, Immunol., 165, 4927–4934. Kubin M, Chin W, Jones J, Woodward A, Le T, Smith C, Nagane M, Pan G, Weddle JJ, Dixit VM, Cavenee WK and Smolak P, Goodwin RG, Rauch CT, Schuh JC and Lynch Huang HJ. (2000). Cancer Res., 60, 847–853. DH. (1999). Nat. Med., 5, 157–163. Nakabayashi H, Taketa K, Yamane T, Miyazaki M, Miyano Wang CY, Mayo MW, Korneluk RG, Goeddel DV and K and Sato J. (1984). Gann, 75, 151–158. Baldwin Jr AS. (1998). Science, 281, 1680–1683. Nakao K, Nakata K, Yamashita M, Tamada Y, Hamasaki K, Wen Y, Yan DH, Spohn B, Deng J, Lin SY and Hung MC. Ishikawa H, Kato Y, Eguchi K and Ishii N. (1999). J. Biol. (2000). Cancer Res., 60, 42–46. Chem., 274, 28075–28078 Yamanaka T, Shiraki K, Sugimoto K, Ito T, Fujikawa K, Ito Nimmanapalli R, Perkins CL, Orlando M, O’Bryan E, M, Takase K, Moriyama M, Nakano T and Suzuki A. Nguyen D and Bhalla KN. (2001). Cancer Res., 61, 759–763. (2000). Hepatology, 32, 482–490. Olie RA, Simoes-Wust AP, Baumann B, Leech SH, Fabbro D, Zhou Q, Fukushima P, DeGraff W, Mitchell JB, Stetler Stahel RA and Zangemeister-Wittke U. (2000). Cancer Res., Stevenson M, Ashkenazi A and Steeg PS. (2000). Cancer 60, 2805–2809. Res., 60, 2611–2615.

Oncogene