Cancer-specific enhancement of cisplatin-induced cytotoxicity ... - Nature

Oncogene (2008) 27, 4603–4614

& 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $30.00 www.nature.com/onc

ORIGINAL ARTICLE

Cancer-specific enhancement of cisplatin-induced cytotoxicity with triptolide through an interaction of inactivated glycogen synthase kinase-3b with p53 Y Matsui1, J Watanabe1, M Ikegawa2, T Kamoto1, O Ogawa1 and H Nishiyama1 1

Department of Urology, Graduate School of Medicine, Kyoto University, Kyoto, Japan and 2Department of Genomic Medical Sciences, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan

To improve conventional chemotherapeutic efficacy, a combination use of traditional medicines is effective but detailed mechanisms have been rarely elucidated. In the this study, we attempted to clarify how triptolide (PG490), an oxygenated diterpene derived from a Chinese herb, enhances the cisplatin (CDDP)-induced cytotoxicity in urothelial cancer cells. Our results showed that a combined CDDP/triptolide therapy induced apoptosis in urothelial cancer cell lines with wild-type p53, but not in those with mutant-type p53 or normal human urothelium. As the mechanism, triptolide suppressed CDDP-induced p53 transcriptional activity, leading to p21 attenuation, which promoted apoptosis via the activation of c-Jun N-terminal kinase (JNK) and Bax. We further demonstrated that the functional regulation of p53 by triptolide was mediated by an intranuclear association of p53 with glycogen synthase kinase-3b (GSK3b), which was inactivated by protein kinase C (PKC). This modulation of the PKC-GSK3b axis by triptolide was observed in a cancer-specific manner. A mouse xenograft model also showed that a combined CDDP/triptolide therapy completely suppressed tumor growth without any side effects. We expect that cancer-specific enhancement of CDDP-induced cytotoxicity with triptolide may effectively overcome the resistance to a CDDP-based conventional chemotherapy as a treatment for urothelial cancer. Oncogene (2008) 27, 4603–4614; doi:10.1038/onc.2008.89; published online 7 April 2008 Keywords: urothelial cancer; cisplatin; triptolide; p53; p21; GSK3b

Correspondence: Dr O Ogawa, Department of Urology, Graduate School of Medicine, Kyoto University, 54 Shogoin Kawaharacho, Sakyo-ku, Kyoto 606-8507, Japan. E-mail: [email protected] Received 4 October 2007; revised 24 January 2008; accepted 29 February 2008; published online 7 April 2008

Introduction Urothelial cancers are common malignancies and significant causes of morbidity and mortality. Invasive bladder cancers are very aggressive and have poor prognosis because of metastatic recurrence after radical surgery (Nishiyama et al., 2004). Cisplatin (CDDP)based chemotherapy has been adopted for these advanced urothelial cancers, although the use of chemotherapy is associated with significant toxicity and provides long-term survival in only 15–20% of patients due to innate or acquired resistance to chemotherapy. To improve chemotherapeutic efficacy, it is important to identify possible targets enhancing specific cell-killing mechanisms in urothelial cancer cells. The p53 tumor suppressor gene is one of the most investigated genes in the chemosensitivity of urothelial cancer, but previous findings have been conflicting on whether it is responsible for the resistance to the effects of CDDP-based systemic chemotherapy (Sarkis et al., 1995; Cote et al., 1997). One of the main reasons can be explained by two distinct functions of p53 in response to DNA damage, one of which is cell-cycle arrest, mainly via p21, and the other is the induction of apoptosis via such genes as PUMA and PIG3. Previously, we demonstrated that dicoumarol, an NADPH quinone oxidoreductase 1 inhibitor, enhanced the cytotoxicity of CDDP through c-Jun N-terminal kinase (JNK) activation by the suppression of the p53–p21 pathway in urogenital cancer cells (Watanabe et al., 2006). These findings indicated that the induction of p53–p21 in malignant cells may possibly suppress the apoptotic process, and inversely, attenuation of p21 expression may make genotoxic chemotherapeutic agents more effective by subverting the normal repair process. Triptolide (PG490) is a diterpenoid triepoxide derived from the herb tripterygium wilfordii, which has been used as a traditional medicine in China (Kupchan et al., 1972). Although it was considered effective in the treatment of inflammatory or autoimmune disorders by the inhibition of cytokine production (Qiu et al., 1999; Chen et al., 2000; Cibere et al., 2003), triptolide also has an antiproliferative effect for tumor cells in vitro and in vivo (Chang et al., 2001; Jiang et al., 2001; Yang et al., 2003; Fidler et al., 2003; Miyata et al., 2005;

Triptolide-mediated enhancement of CDDP cytotoxicity Y Matsui et al

4604

Carter et al., 2006; Wan et al., 2006). Among several intracellular pathways suggested to be responsible for the antiproliferative effect of triptolide, the suppression of p21 is one of the most convincing mechanisms (Chang et al., 2001; Fidler et al., 2003) but the details have never been elucidated. In this study, we demonstrated the effectiveness of triptolide to enhance CDDPinduced cytotoxicity via the modulation of p53 transcriptional activity in a cancer-specific manner, and clarified that p53 transcriptional activity was modulated through an intranuclear association of functional p53 with glycogen synthase kinase-3b (GSK3b) phosphorylated at serine 9 by protein kinase C (PKC), which suppressed p21 expression and evoked the JNKmitochondria apoptotic pathway. Furthermore, we attempted combination use of triptolide with CDDP in a mouse xenograft model of urothelial cancer cells, and demonstrated the possibility of p21 attenuation therapy with triptolide in clinical application. Results Triptolide enhances susceptibility to CDDP in urothelial cancer cells with wild-type p53 through the suppression of p53 transcriptional activity The effects of triptolide were analysed using human normal urothelial (HNU) cells and five urothelial cancer cell lines (RT112, RT4, TCCsup, EJ and J82). Triptolide inhibited cell proliferation in all five cancer cell lines dose-dependently within 100 nM, but in the range over 100 nM, the effect plateaued. The cell viability rates 24 h after 40 nM triptolide exposure were 85.7±14.0, 76.1±4.5, 80.5±18.6, 63.1±5.9 and 77.7±23.4, 79.1±4.9% in RT112, RT4, TCCsup, EJ, J82 and HNU, respectively. In this study, we used the concentration at which triptolide did not induce apoptosis on its own. Susceptibility to CDDP differed among cells; IC50 monitored by MTT assay were 42.1±11.3, 46.9±0.8, 53.6±0.4, 109.7±13.4, 556.2±55.7 and 969.0±106.2 mM in HNU cell, RT4, RT112, TCCsup, EJ and J82, respectively. Pretreatment with triptolide enhanced the susceptibility to CDDP dose-dependently in RT112 and RT4, which retained wild-type p53, but not in three other cell lines with mutant p53 or HNU cells at 40 nM of triptolide (Figure 1a). Synergic effects were not observed even when the concentration of triptolide increased to 80 nM in bladder cancer cell lines with mutant p53 (Figure 1a). Nuclear staining with Hoechst33342 in RT112 demonstrated that pretreatment with triptolide (30 nM) induced significant apoptosis in combination with CDDP (30 mM) (60.0±1.0%), compared with CDDP alone (30 mM) and triptolide alone (30 nM) (8.7±3.4 and 1.9±0.4%, respectively; Figure 1a). The necessity of functional p53 in the effect of triptolide was elucidated in two stable clones transfected with p53 dominant-negative mutant (p53DN) (RT112 p53DN-2 and -11). Although the antiproliferative effect of triptolide itself was not significantly changed by p53DN transfection, MTT assay demonstrated that Oncogene

triptolide failed to enhance the susceptibility to CDDP in both p53DN transfectants (Figure 1a). Western blot analysis also demonstrated that the combination of CDDP and triptolide activated caspases in RT112 pCMVneo but not in RT112 p53DN-11, in which normal transcriptional function of p53 was attenuated by the dominant-negative effect (Figure 1b). On the other hand, no significant difference was shown in the intracellular distribution of nuclear factor-kB and mRNA levels of the inhibitor of apoptosisfamily regardless of the addition of triptolide in RT112 cells (Supplementary Figure 1). All together, these data suggested that p53 function is necessary for triptolide to enhance CDDP-induced apoptosis in urothelial cancer cells. To investigate the effect of triptolide to p53 function, change of p53 expressions and its downstream molecules were assessed. In RT112 cells, treatment with CDDP induced p53 expression in a time-dependent manner and pretreatment with triptolide slightly increased its induction (Figure 1c). However, whereas CDDP induced sequentially the expressions of p21, MDM2, PUMA, pretreatment with triptolide suppressed their induction after 7 h, although there was transient induction within 7 h. Bax expression was influenced neither by CDDP nor triptolide. Semiquantitative reverse transcription–PCR demonstrated that pretreatment with triptolide suppressed the induction of mRNAs for p21/MDM2/ PUMA by CDDP in RT112 cells (Figure 1c). The effect of triptolide on p53 transcriptional activity was further examined in KK47 cells transiently transfected by wildtype p53 expressing vector (pcDNA flag-p53). CDDP failed to induce p53 in KK47 cells, although KK47 cells have no mutation in p53 genomic sequence (Supplementary Figure 2). In KK47 cells, transient transfection of pcDNA flag-p53 induced p21 and triptolide attenuated the expression of p21 without the suppression of p53 protein level (Figure 1d). Taken together, these results indicated that triptolide suppressed p53 transcriptional activity without suppression of p53 protein level. The attenuation of p21 expression by triptolide is critical to evoke CDDP-induced apoptosis via the activation of the JNK–Bax pathway As CDDP-induced apoptosis is also related to mitogenactivated protein kinase cascades (Mansouri et al., 2003), the effects of triptolide on mitogen-activated protein kinase cascades were assessed (Figure 2a). Although CDDP alone did not induce the phosphorylation of JNK and p38, the combination treatment of triptolide and CDDP induced the phosphorylation of JNK 7 h after treatment and it was sustained over the following 14 h period in RT112, according to the induction of apoptosis. p38 was also phosphorylated by the combination treatment, but its timing was delayed. Although treatment with triptolide alone did not induce the phosphorylation of JNK at the same concentration, the combination treatment induced Bax transactivation and released cytochrome c from mito-


4605

Figure 1 Triptolide enhances susceptibility to cisplatin (CDDP) in urothelial cancer cells with wild-type p53 through the suppression of p53 transcriptional activity. (a) (Upper panel) The effects of triptolide on CDDP-induced cytotoxicity assessed by the status of p53, using human normal urothelial (HNU) cells, five urothelial cancer cell lines and two stable clones of RT112 transfected with p53 dominant-negative mutant (RT112 p53DN-2 and RT112 p53DN-11). The results of MTT assays for 24 h with CDDP at approximate concentrations of IC25 with or without triptolide at indicated concentrations are shown as average values±s.d. from three independent experiments. Cell viability is indicated by the ratio to cell viability with CDDP at the concentration of IC25 of each cell line. The concentration of IC25 was the concentration at which cell growth was inhibited by 25%, calculated from the regression line based on 24 h MTT assay. (Lower panel) Representative figures of Hoechst 33342 staining of RT112 treated with indicated drugs for 24 h. (b) Comparison of the response to CDDP and triptolide between RT112 control transfectants (RT112 mock) and p53 dominant-negative mutant (p53DN) transfectant (RT112 p53DN-11), demonstrated by western blot analysis. Cells were treated with CDDP at the concentration corresponding to IC25 (30 mM) with or without 1 h pretreatment of 30 nM triptolide for 12 h, and then lysed in RIPA buffer and subjected to western blot analysis. (c) Western blot analysis (left panel) and semiquantitative reverse transcription (RT)–PCR (Right panel) demonstrating the changes of p53-downstream molecules in RT112 treated by CDDP with or without triptolide. Cells were treated with 30 mM CDDP with or without 1 h pretreatment of 30 nM triptolide for the indicated times. ‘0 h’ refers to the time of CDDP addition. (d) Effect of triptolide on the expression of p21 under the existence of wild-type p53. Wild-type p53 expression vector (pcDNA flag-p53) or control vector (pcDNA3) was transiently transfected to KK47 cells. Thirty-six hours later, 35 nM triptolide was added, and cells were incubated for 12 h. Protein or mRNA were then extracted from cells and subjected to western blot analysis or RT–PCR.

chondria to cytosol following JNK phosphorylation (Figure 2b). The significance of the activation of the JNK–Bax pathway in the induction of apoptosis was confirmed with SP600125 (40 mM), a specific inhibitor of JNK1/2, or Bax Channel Blocker (10 mM), both of which apparently attenuated the apoptosis (Figure 2b). Similar effects of SP600125 and Bax Channel Blocker were also confirmed in RT4 cells (data not shown). These results indicated that activation of the JNK–Bax pathway was necessary for the synergic effect of triptolide on CDDP. Among several p53 downstream molecules, p21 has been reported to suppress the activation of JNK (Huang

et al., 2003; Watanabe et al., 2006). We established two stable clones of RT112 transfected with p21siRNA vector (RT112 p21siRNA-8 and -12) and investigated JNK status and susceptibility to CDDP and triptolide. These clones showed higher susceptibility to CDDP than RT112 mock, and triptolide had no significant enhancement on the apoptotic ratio (Figure 2c). Moreover, although CDDP exposure induced p21 but not the phosphorylation of JNK in RT112 mock, JNK activation was observed in RT112 p21siRNA at the same concentration of CDDP alone (Figure 2d). These results indicated that p21 attenuation by pretreatment with Oncogene


4606

Figure 2 Triptolide promotes cisplatin (CDDP)-induced apoptosis via the activation of the JNK–Bax–mitochondria pathway induced by p21 attenuation. (a) Western blot analysis demonstrating changes of the mitogen-activated protein kinase cascade in RT112 treated by CDDP with or without triptolide. Cells were treated with 30 mM CDDP with or without 1 h pretreatment of 30 nM triptolide for the indicated times. ‘0 h’ refers to the time of CDDP addition. (b) The significance of the JNK–Bax–mitochondria apoptotic pathway in the effect of triptolide; (Left panel) Western blot analysis showing the activation of JNK–Bax–mitochondria apoptotic pathway by the combination treatment of CDDP and triptolide in RT112. RT112 cells were treated with the indicated drugs for 12 h, after which cells were lysed in lysis buffer; conformationally changed Bax protein was immunoprecipitated using anti-Bax 6A7 antibody and subjected to western blot analysis using polyclonal Bax antibody. Simultaneously JNK phosphorylation status and cytosolic release of cytochrome c were examined by western blot analyses of whole cell lysate or cytosolic cell fraction. (Right panel) Western blot analysis showing the effect of inhibition of the JNK–Bax pathway on triptolide-mediated enhancement of CDDP-induced apoptosis in RT112. The status of molecules associated with apoptosis in RT112 treated with CDDP, triptolide and specific inhibitors of JNK or Bax were shown. Cells were treated by the indicated drugs for 12 h, and then subjected to western blot analysis. (c) MTT assay showing the effect of p21 suppression on the susceptibility to CDDP and triptolide in RT112. Two stable clones of RT112 transfected with psiRNA-h7SKhp21 vector containing shRNA targeting p21 were established (RT112 p21siRNA-8 and -12), and the suppression of p21 expression was confirmed by RT–PCR. Cells were then pretreated with (black bar) or without (white bar) 30 nM triptolide for 1 h followed by treatment of 30 mM CDDP for an additional 24 h. Cell viability was measured by MTT assay as average percent values±s.d. to cell viability of RT112 mock treated by 30 mM CDDP from three independent experiments. (d) Western blot analysis showing the induction of apoptosis via JNK activation by the attenuation of p21 expression. RT112 cells stably transfected with control vector (RT112 mock) or p21siRNA vector (RT112 p21siRNA-8) were treated with 30 mM CDDP for 12 h, and then p53 and p21 expression, JNK phosphorylation and poly (ADP-ribose) polymerase (PARP) cleavage were examined.

triptolide was essential for CDDP-induced apoptosis via JNK activation. Intranuclear association of p53 with GSK3b phosphorylated at serine 9 by PKC suppresses p53 transcriptional activity p53 transcriptional activity is influenced by various factors including post-translational modifications, the interaction of cofactors and subcellular distribution (Bode and Dong, 2004; Erster et al., 2004; Coutts and La Thangue, 2005). The phosphorylation and acetylation of p53 were induced in our condition, but there were no remarkable changes between pretreatment with Oncogene

triptolide and that without pretreatment (Figure 3a). We identified that triptolide induced a gradual increase of the total amount of intracellular GSK3b and evoked the phosphorylation at serine 9 of GSK3b (p-Ser9 GSK3b) in combination with CDDP treatment in RT112 cells after 10.5 h. Interestingly, in combination treatment, the expression of p21 was transiently induced by 7 h and attenuated at 10.5 h following the induction of p-Ser9 GSK3b. Immunocytochemical analysis of RT112 cells revealed that whereas p53 accumulation in the nucleus was not changed by the presence or absense of triptolide treatment, GSK3b tended to accumulate in the nucleus by combination treatment, showing the same subcellular distribution to p53 (Figure 3b).


4607

To clarify the interaction between p53 and GSK3b, we used two-dimensional (2D) BN/SDS–polyacrylamide gel electrophoresis (PAGE). The position of two protein spots in one vertical line in a 2D BN/SDS membrane could be indicative of their presence in the same multiprotein complex (MPC) (Camacho-Carvajal et al., 2004). The spots of p53 and GSK3b were not on the same vertical line by the treatment of CDDP alone, whereas the spot of GSK3b was shifted onto the vertical line of p53 by CDDP/triptolide combination treatment (Figure 3c). To confirm the formation of MPC, KK47 cells transiently transfected with GSK3b and p53-expressing vector were treated with triptolide. Immunoprecipitation analysis revealed that p-Ser9 GSK3b was coimmunoprecipitated with p53 after treatment by triptolide (Figure 3d). This association of GSK3b with p53, triptolide using treatment, was further confirmed within PC3 cells, p53-null prostate cancer cell line containing one base-deletion at codon 138, in which triptolide suppressed the transcriptional activity of transfected wild-type p53 (Figure 3e). Taken together, these data indicated that triptolide affects GSK3b and p53 to form MPC in the nucleus. To elucidate the importance of p-Ser9 GSK3b in synergic effects of triptolide and CDDP, RT112 cells were stably transfected with pSUPER neo-gfp GSK3b vector containing shRNA-targeting GSK3b. The effects of triptolide on CDDP-induced cytotoxicity were attenuated in the stable transfectant of shRNA (clone-6) (Figure 4a). The same result was also obtained in other clones (data not shown). Suppression of the expression of GSK3b attenuated JNK activation, and also restored the expression of p21 (Figure 4a). On the other hand, transient transfection of GSK3b(S9A), constitutively active mutant GSK3b, with wild-type p53 into KK47 cells also restored p21 expression partially after exposure to triptolide both in RNA and protein levels, compared with control transfectant (Figure 4b). Induction of p-Ser9 GSK3b was similarly observed in RT4 cells treated with triptolide and CDDP, which showed synergism, but not in HNU, where no synergic effect was observed (Figure 4c). These results indicated that the interaction between p-Ser9 GSK3b and p53 was a key mechanism for triptolide to enhance CDDP-induced apoptosis via modulation of p53 transcriptional activity. Furthermore, we examined the synergic effect of triptolide with other chemotherapeutic agents in RT112 cells. Our results showed that triptolide had a synergic effect with doxorubicin, which induced functional p53, but not with paclitaxel, which did not induce p53 (Supplementary Figure 3), indicating that triptolide can modulate p53 transcriptional activity under the existence of functional p53 and GSK3b, and it is not CDDP-specific. To explore the upstream mechanisms that phosphorylated GSK3b, we evaluated the change of the phosphorylated level of PKC, one of the upstream molecules of GSK3b, and 3-phosphoinositide-dependent protein kinase 1. Interestingly, PKC was activated by combination treatment in RT4 cells although there was no remarkable change in HNU (Figure 4c).

Moreover, there was no remarkable change in the phosphorylated level of 3-phosphoinositide-dependent protein kinase 1 both in RT4 and HNU cells. To confirm the importance of PKC as the upstream of GSK3b influenced by triptolide, the effects of GF109203x, PKC inhibitor, or LY294002, phosphatidylinositol-3 kinase inhibitor, were assessed in RT112 cells. The results showed that, although inhibition of phosphatidylinositol-3 kinase by LY294002 did not influence the phosphorylation of GSK3b, inhibition of PKC by GF109203x diminished GSK3b phosphorylation and restored p21 expression, resulting in suppression of apoptosis (Figure 4d). The effect of PKC inhibition was also shown using other PKC inhibitors, Go¨6983 and Go¨6976 (Figure 4d). Furthermore, we confirmed that triptolide enhanced CDDP-induced cytotoxicity in cancer cells of other organs (MCF7, breast cancer cell line, and U2OS, osteosarcoma cell line) as same as urothelial cancer cells but did not show a chemosensitizing effect in human normal fibroblast (Supplementary Figure 4). These data indicated that the intracellular target of triptolide was the PKC-GSK3b axis and that the difference in the regulation of PKC activity between human normal cells and cancer cells caused the difference in response to triptolide. Triptolide effectively enhances CDDP-induced cytotoxicity in mouse in vivo model For the clinical use of triptolide for urothelial carcinomas, we examined the effects of triptolide alone or in combination with CDDP to the in vivo growth of RT112 cells, using a subcutaneous xenograft model. CDDP was given weekly but triptolide was four times per week due to the short half-life of triptolide (Figure 5a). To investigate the safety of this combination, body weight was assessed. During the treatment duration, combination treatment of CDDP and triptolide did not have an adverse effect on body weight in comparison with CDDP or triptolide alone. In our conditions, triptolide or CDDP only partially delayed tumor cell growth on its own and the tumor grew steadily, but their combination suppressed tumor growth almost completely (Figure 5a). Hematoxylin and eosin staining of tumors demonstrated that combination treatment of triptolide and CDDP remarkably induced apoptosis, compared with CDDP treatment alone. Furthermore, immunohistochemistry revealed that JNK phosphorylation was highly promoted by the combination of CDDP and triptolide (Figure 5b). These results indicated that triptolide could also enhance CDDP-induced cytotoxicity via JNK activation in vivo as well as in vitro.

Discussion Triptolide, a diterpenoid triepoxide derived from the herb tripterygium wilfordii, has been reported to have an antiproliferative effect against a broad spectrum of tumors (Lee et al., 1999; Chang et al., 2001; Jiang et al., Oncogene


4608

2001; Fidler et al., 2003; Yang et al., 2003; Miyata et al., 2005; Yinjun et al., 2005; Carter et al., 2006; Wan et al., 2006). Its high antiproliferative efficiency is attractive for clinical applications, and phase E´ clinical trials using PG490-88, a water-soluble prodrug of triptolide, have been already attempted in Europe. However, as Shamon et al. (1997) reported using a xenograft model, its narrow therapeutic window would be a problem when used as a single

Oncogene

agent, and therefore the benefit of the combination use of triptolide at low concentrations with conventional chemotherapeutic agents such as doxorubicin and CPT-11 has been intended (Fidler et al., 2003; Carter et al., 2006). In this study, we demonstrated the usefulness of triptolide at subtoxic concentrations to enhance CDDP-induced cytotoxicity in urothelial cancer cells with wild-type p53, without synergic toxicity to normal urothelium.


4609

Although several intracellular pathways such as the modification of nuclear factor-kB, Bcl-2, phosphatidylinositol-3 kinase, cyclins and c-myc have been reported to be responsible for the mechanisms underlying the action of triptolide (Lee et al., 1999; Chang et al., 2001; Jiang et al., 2001; Fidler et al., 2003; Yang et al., 2003; Miyata et al., 2005; Yinjun et al., 2005; Carter et al., 2006; Wan et al., 2006), our results showed that the attenuation of p53-dependent p21 induction was a key mechanism to enhance CDDP-induced cytotoxicity. Several recent studies have demonstrated that p53-dependent p21 induction inhibits the apoptotic response, and p21 attenuation may make genotoxic chemotherapeutic agents more effective by subverting the normal repair process or, more directly, by promoting the apoptotic process through inhibition of p21 interaction with apoptosis signal-regulating kinase 1, which is upstream of JNK (Asada et al., 1999; Gartel and Tyner, 2002; Weiss, 2003). In our condition, CDDP/triptolide combination attenuated p53 downstream molecules such as p21, MDM2 and Puma which have short half-lives for mRNA and protein, but had no remarkable influence on the expression of Bax, which has relatively longer half-life, and activated JNK evoked mitochondrial apoptosis using Bax transactivation as a proapoptotic mediator. Previously, flavopiridol, a transcriptional inhibitor, showed a similar change of p53–p21 by downregulating thousands of mRNAs that have a short half-life (Demidenko and Blagosklonny, 2004). However, as triptolide had no suppressive effect on transcription of XIAP and CIAP2 (Supplementary Figure 1) and slightly increased p53 protein level, we consider that triptolide functioned by modulating p53 transcriptional activity but not as a pantranscriptional inhibitor such as flavopiridol. The increase of p53 protein level after combination treatment may be due to the suppression of MDM2 protein level or upregulation of intracellular reactive oxygen species by CDDP/triptolide. 30 mM CDDP treatment did not evoke apoptosis in RT112 p53DN cells, although they already lost p21 inducibility. We propose that this is because we used stable transfectants in our experiments, and some

undetermined change during clonal selection may bias cell viability. Concerning the detailed mechanism of triptolide to modulate p53 transcriptional activity, Chang et al. (2001) previously reported that the phosphorylation of p53 induced by triptolide may be responsible. Our results, however, showed that intranuclear association of p53 with inactive p-Ser9 GSK3b was responsible for suppressing p53 transcriptional activity. Recent studies revealed that nuclear GSK3b directly associated with p53 after DNA damage, and that its activity was necessary for full transcriptional activity of p53, whereas inhibition of GSK3b reduces the expression of p53regulated proteins (Watcharasit et al., 2002, 2003). Although it is still controversial whether the reduction of GSK3b activity inhibits p53-induced apoptosis or contributes to promoting apoptosis via its association with p53 in nuclei or mitochondria (Watcharasit et al., 2003; Tan et al., 2005), our results showed that inhibition of p53 transcriptional activity by inactive p-Ser9 GSK3b in nuclei contributed to promote apoptosis under the existence of genotoxic stress of chemotherapeutic agents. Concerning the regulation of binding of GSK3b to p53, GSK3b activity was reported not to be required; on the contrary, inhibition of GSK3b activity stabilized its association with p53 (Watcharasit et al., 2003). As Ser-9 phosphorylation of GSK3b inhibits its activity, this phosphorylation helps GSK3b associate with p53, according to these authors’ theory, although many other factors regulating the formation of protein complexes should be further elucidated. In this study, we also demonstrated that PKC may be an upstream target of triptolide to affect GSK3b function. As Go¨6976, an inhibitor of classical PKCa/b, attenuated the effect of triptolide, classical PKCs were considered upstream targets. Generally, changes in the balance of PKC activity have a close relation to various cellular processes such as proliferation, differentiation, apoptosis, migration or tumorigenesis (Koivunen et al., 2004). Among PKCs, classical PKCa/bI have been linked to increased cell growth (Mandil et al., 2001; Jiang et al., 2004). In urothelial cancers, not only the

Figure 3 Intranuclear association of p53 with GSK3b was induced by combination use of cisplatin (CDDP) and triptolide. (a) Western blot analysis showing post-translational modification of p53 and the change of GSK3b examined in a time-dependent manner. RT112 cells were treated with 30 mM CDDP with or without 1 h pretreatment of 30 nM triptolide for the indicated times and then subjected to western blot analysis. (b) Immunocytochemistry showing the intracellular localization of p53 and GSK3b after treatment of CDDP with or without triptolide. RT112 cells were treated with 30 mM CDDP with or without 30 nM triptolide. After 12 h exposure, intracellular localization of p53 and GSK3b was detected as described in ‘Materials and methods.’ (c) 2D BN/SDS– polyacrylamide gel electrophoresis (PAGE) analysis showing multiprotein complex formation involving p53 and GSK3b in RT112 cells cotreated with triptolide and CDDP. Cells were treated by the indicated drugs as described in (a) and lysed in cell lysis buffer. They were then separated by 2D BN/SDS–PAGE and immunoblotted using antibodies against p53 and GSK3b. The spots of p53 and GSK3b were indicated by black arrow and black arrowhead, respectively. (d) Immunoprecipitation analysis showing the association of p53 with p-Ser9 GSK3b in KK47 treated by triptolide. KK47 cells were transiently co-transfected with pcDNA flag-p53 (6 mg) and pcDNA GSK3b (18 mg) per 10 cm dish. Thirty-six hours later, 35 nM triptolide was added, and cells were incubated for an additional 12 h. The levels of coimmunoprecipitated p53 and phosphorylated or total GSK3b were measured by western blot analysis. (e) Effect of triptolide on the expression of p21 under the existence of wild-type p53 examined in PC3 cells. (Upper panel) Wild-type p53 expression vector (pcDNA flag-p53) or control vector (pcDNA3) was transiently transfected to PC3 cells. Thirty-six hours later, 35 nM triptolide was added, and cells were incubated for 12 h. Protein was then extracted from cells and subjected to western blot analysis. (Lower panel) Immunoprecipitation analysis showing the association of p53 with GSK3b in PC3 cells treated by triptolide. PC3 cells were transiently co-transfected with pcDNA flag-p53 and pcDNA GSK3b. Thirty-six hours later, 35 nM triptolide was added, and cells were incubated for an additional 12 h. The levels of coimmunoprecipitated p53 and GSK3b were measured by western blot analysis. Oncogene


4610

expression but also the activity of PKCa/bI has been reported to increase, particularly in the most proliferative areas of the tumor (Aaltonen et al., 2006). Our results showed that, compared with normal urothelium, cancer cells showed hyperphosphorylation of PKCs by CDDP/triptolide combined therapy, leading to the phosphorylation at Ser9 of GSK3b. We consider that the difference in the response of PKC to triptolide

Oncogene

between normal cells and cancer cells would be advantageous to promote CDDP-induced apoptosis in a cancer-specific manner. From our results in which p21 did not suppress immediately after drug exposure (Figure 1c), triptolide may require multiple steps of signal activation or intracellular translocation of important molecules to modulate the PKC–GSK3b signaling pathway, which finally attenuated p53


4611

Figure 5 Combination treatment of cisplatin (CDDP) and triptolide significantly suppressed the growth of RT112 tumor in vivo. (a) The change of subcutaneous RT112 tumor volume measured by photon emission in the in vivo imaging system. Nude mice bearing established subcutaneous tumors were grouped together to constitute similar mean tumor photon emissions in each group 6 days after the inoculation of RT112 cells expressing luciferase gene (day 0). Mice in each group were treated as described in ‘Materials and methods,’ and the effect of indicated drugs on tumor growth was measured by photon emission using the in vivo imaging system. (b) Histological examination of the effect of CDDP and triptolide on RT112 tumor in vivo. At day 16, the mice were killed, and the tumors were dissected. Morphological change was observed by hematoxilin-eosin (HE) stain, and the phosphorylation of JNK was examined by immunohistochemistry.

Figure 4 The interaction of GSK3b phosphorylated at serine 9 by PKC independently of 3-phosphoinositide-dependent protein kinase 1 activity is essential for the modulation of p53 transcriptional activity by triptolide. (a) Effect of the suppression of GSK3b expression on the synergic effect of triptolide and CDDP in RT112 cells shown by MTT assay (upper panel) and western blot analysis (lower panel). MTT assay shows the effect of GSK3b suppression on the susceptibility to CDDP and triptolide in RT112. A stable clone of RT112 transfected with pSUPER neo-gfp GSK3b vector containing shRNA targeting GSK3b were established (RT112 siGSK3b-6). RT112 siGSK3b-6 or RT112 mock (transfected by pSUPER vector) cells were then pretreated with or without 30 nM triptolide for 1 h followed by treatment with 30 mM CDDP for an additional 24 h. Cell viability was measured by MTT assay as average percent values±s.d. to cell viability of RT112 mock treated only by dimetyl sulfoxide (DMSO) used as vehicle from three independent experiments. Western blot analysis shows the influence of GSK3b suppression on the p53–p21–JNK proapoptotic pathway in RT112 cells. RT112 siGSK3b-6 or RT112 mock were treated by the indicated drugs for 12 h, and subjected to western blot analysis. (b) Quantitative reverse transcription–PCR (upper panel) and western blot analysis (lower panel) showing the effect of the expression of a constitutively active GSK3b mutant (GSK3b(S9A)) on the modulation of p53 transcriptional activity by triptolide. KK47 cells were transiently co-transfected with pcDNA flag-p53 (2 mg) and pcDNA flag-GSK3b(S9A) (6 mg) or pcDNA3 (6 mg) per 6 cm dish. Thirty-six hours later, 35 nM triptolide was added, and cells were incubated for an additional 12 h. Quantitative RT–PCR was shown as average values±s.d. of relative expression ratios of p21 (p21/GAPDH 1000) from three independent experiments. (c) Western blot analysis showing the change of signal transduction assessed in RT4 bladder cancer cell lines with wild-type p53 and human normal urothelium (HNU) after treatment with CDDP and triptolide. Cells were treated by CDDP with or without triptolide at the indicated concentrations. After 12 h, cells were collected and subjected to western blot analysis (in RT4 cells, ‘*’ indicates nonspecific bands detected by GSK3b antibody). (d) (Upper and middle panels) The change of signal transduction in RT112 treated with CDDP, triptolide and GF109203x, a specific inhibitor of PKC, or LY294002, a specific inhibitor of phosphatidylinositol-3 kinase. Cells were treated by the indicated drugs for 12 h, and then subjected to western blot analysis. The relative intensity of cleaved poly (ADP-ribose) polymerase (PARP)/b-actin indicates the relative intensity of cleaved PARP expression in each lane normalized to b-actin expression calculated from the results of western blot analysis. (Lower panel) The change of signal transduction in RT112 treated with CDDP, triptolide and Go¨6983 (4 mM) or Go¨6976 (1 mM). Cells were treated as described in the upper and middle panels and subjected to western blot analysis. Oncogene


4612

transcriptional activity. As cancer cells with mutant p53 did not exhibit the change of PKC–GSK3b signals (Supplementary Figure 5) and p-Ser9 GSK3b was induced by triptolide in KK47 cells transfected with wild-type p53 (Supplementary Figure 6), we also speculate that transcriptional activity-independent undetermined function of p53 may be required for triptolide to modulate the activity of the PKC–GSK3b signaling pathway. The detailed regulation mechanism of PKC activity other than 3-phosphoinositide-dependent protein kinase 1 stimulation should be further elucidated in the future. To identify and characterize protein interaction under native conditions, 2D BN/SDS–PAGE, the combination of BN–PAGE and SDS–PAGE in a 2D approach, was a powerful tool (Camacho-Carvajal et al., 2004). It was very helpful to detect the association of p53 with GSK3b in this study. It should be combined with other methods such as immunoprecipitation and immunocytochemistry to eliminate the possibility that each protein is part of separate complexes migrating at the same position in BN–PAGE. Furthermore, because the ratio of the complexes detected by 2D approach is different depending on the cells, treatments and detergent used, validation using other methods may be recommended in some cases. However, this method is suitable for highthroughput screening to identify MPCs in combination with liquid chromatography-tandem mass spectrometry (LC-MS/MS). In the future, this method will help us clarify the relationship between p53 and the PKCGSK3b axis, and resolve the dynamic change within cells falling into apoptosis. To assess the clinical usefulness and safety, we used a subcutaneous xenograft model. In this model, triptolide was administered at 0.15 mg/kg four times per week. This dose was lower than those reported previously (Shamon et al., 1997; Fidler et al., 2003; Yang et al., 2003) and we could use safely without any remarkable adverse effects, but this low concentration could still enhance cancer-specific CDDP-induced cytotoxicity. When we consider clinical application of triptolide at a low concentration with chemotherapeutic agents to overcome its disadvantage, which is the narrow therapeutic window, a combination with CDDP, which is one of the most frequently used agents for urothelial cancer, is a very attractive choice. In conclusion, we demonstrated that triptolide could promote CDDP-induced apoptosis in urothelial cancer cells with wild-type p53 by the suppression of p53dependent p21 induction and that the modulation of p53 transcriptional activity with triptolide was mediated through the intranuclear association of p53 with GSK3b inactivated by PKC in cancer-specific manner. Compared with the use of peptides, small molecule inhibitors or antisense techniques (Lam et al., 2003; Weiss et al., 2003), we consider that triptoide may be a more convenient, realistic and attractive p21 attenuation method, because triptolide is a common drug and exhibits cancer-specific p21 attenuation. Furthermore, this synergic effect of triptolide was not limited within urothelium or CDDP treatment. We are convinced that Oncogene

p21 target therapy using triptolide will develop as a promising clinical strategy applied to a broad spectrum of cancer treatments in the near future.

Materials and methods Antibodies and reagents Antibodies were obtained as follows: anticaspase-3, anticaspase-8 and anticytochrome c from BD PharMingen (San Diego, CA, USA), anti-PUMA from Sigma (St Louis, MO, USA) and anti-p53 (Ab-2) from Calbiochem (San Diego, CA, USA), p21 and MDM2 from Santa Cruz (Santa Cruz, CA, USA), anti-acetyl p53 (Lys373 and 382) from Upstate (Lake Placid, NY, USA) and anti-actin b from Abcam (Cambridge MA, USA). Other antibodies from Cell Signaling Technology (Beverly, MA, USA). Triptolide, Bax Channel Blocker ((±)-1-(3,6-Dibromocarbazol-9-yl)-3-piperazin-1-ylpropan-2-ol, bis TFA), LY294002, GF109203x, Go¨6983 and Go¨6976 were obtained from Calbiochem. SP600125 was obtained from Tocris Cookson Ltd. (Bristol, UK), and Cis-diamminedichloroplatinum (cisplatin, CDDP) was from WAKO (Osaka, Japan). Cell culture Human normal urothelial cells, six urothelial cancer cell lines (RT112, RT4, KK47, TCCsup, EJ and J82) and PC3 prostate cancer cell line were used for this study. Sequencing analysis and genotyping of p53 confirmed that RT112, RT4 and KK47 had wild-type p53 whereas four other cell lines harbored mutant p53. Five bladder cancer cell lines (RT112, KK47, TCCsup, EJ and J82) and PC3 were maintained in an RPMI1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. RT4 cells were in McCoy 5A. HNU cells were isolated from surgical specimens of patients with benign conditions (Southgate et al., 1994). HNU cells were maintained in a keratinocyte serum-free medium and used for experiments in the second passage. Cells were incubated at 37 1C in a humidified atmosphere containing 5% CO2. Plasmid and transfection pCMV-neo was a gift from Dr K Cho (University of Michigan Medical School). P53DN mutants containing one missense mutation at codon 135 (TGC-TAC), which were purchased from BD Bioscience, were ligated into pCMVneo. psiRNAh7SKhp21 vector containing shRNA targeting p21 was purchased from Invivogen (San Diego, CA, USA). pcDNA3 flag-p53 was a gift from Dr. Shimotohno (Institute for Virus Research, Kyoto University). pcDNA3 GSK3b, expressing wild-type GSK3b, pcDNA3 flag-GSK3b(S9A), expressing constitutively active mutant GSK3b (serine 9 was replaced with an alanine), and pSUPER neo-gfp GSK3b vector containing shRNA targeting GSK3b, were gifts from Dr Blattner (Institute for Toxicology and Genetics, Forschungszentrum Karlsruhe, Germany) (Kulikov et al., 2005). Cells were transfected with the indicated vectors by using Lipofectamine 2000 (Life Technologies Inc., Grand Irand, NY, USA) following the manufacturer’s instructions. Stable transfectants were selected by appropriate selection antibiotics, and were confirmed by sequencing, RT–PCR or western blot analysis. Cell viability assay and detection of apoptosis Cell viability was assessed by 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay as described previously (Watanabe et al., 2006). To detect apoptosis,


4613 nuclear staining was performed. For nuclear staining, cells were stained with 1 mM Hoechst 33342 solution (Wako, Osaka, Japan), and analysed with a fluorescence microscope. Apoptotic cells were identified by morphology and by condensation and fragmentation of their nuclei. The percentage of apoptotic cells was calculated as the ratio of apoptotic cells to total cells counted. RNA isolation, RT–PCR and Real-time quantitative RT–PCR Total RNA preparation and first-strand cDNA synthesis was performed as described previously (Matsui et al., 2007). Primer sequences used for cDNA amplification were as follows: p21 sense 50 -ATTAGCAGCGGAACAAGGAGTCAGACAT-30 and antisense 50 -CTGTGAAAGACAGAACAGTACAG GGT-30 ; MDM2 sense 50 -AATCATCGGACTCAGGTAC A-30 and antisense 50 -GTCAGCTAAGGAAATTTCAGG-30 ; PUMA sense 50 -CCAAACGTGACCACTAGCCT-30 and antisense 50 -ACAGGATTCACAGTCTGGGC-30 . Primers for amplification of the housekeeping gene, human GAPDH, were sense 50 -CGAGCCACATCGCTCAGACA-30 , and antisense 50 -TGAGGCTGTTGTCATACTTCTC-3. The RT–PCR exponential phase was determined to allow semiquantitative comparisons. PCR products were separated by electrophoresis on 2% agarose gels. Quantitative real-time PCR with SYBRgreen PCR Master Mix (Applied Biosystems, Foster City, CA, USA) was performed using the Gene Amp 5700 Sequence Detection System (Applied Biosystems) with the relative standard curve method. Immunoblotting and immunoprecipitation After drug treatments, cells were collected and lysed as described previously (Matsui et al., 2007). Thirty to 50 micrograms of protein samples were separated by SDS–PAGE and transferred onto a polyvinylidene difluoride membrane following standard methods. Membranes were probed with appropriate dilutions of primary antibodies followed by incubation with horseradish peroxidase-conjugated secondary antibodies. After extensive washing, proteins were visualized by a chemiluminescent detection system (GE Healthcare, Buckinghamshire, UK). For immunoprecipitation, cells were washed with phosphate-buffered saline and lysed in an appropriate volume of ice-cold lysis buffer (50 mM Tris–HCl pH 7.4, 150 mM NaCl, 0.5% Nonidet P-40, containing 1 mM Na3VO4, 1 mM NaF, 1 mM phenylmethylsulfonyl fluoride and protease inhibitor cocktail tablets). The lysates were incubated with the indicated antibodies for 1 h to overnight after preclearance with protein G-Sepharose beads for 30 min. Subsequently, they were incubated with protein G-Sepharose beads for 1 h. Beadbound proteins were eluted by boiling for 5 min and subjected to immunoblot analysis.

Immunocytochemical studies RT112 cells were grown and treated with 30 mM CDDP with or without 30 nM triptolide for 12 h in chamber slides. For permeabilization and fixation, cells were treated with methanol at 20 1C for 10 min. Cells were washed and blocked with 10% serum containing phosphate-buffered saline, and then the primary antibody was added. Finally, fluorescein isothiocyanate-conjugated or tetramethylrhodamine isothiocyanate-conjugated secondary antibody was added, and cells were visualized using a Nikon Eclipse E1000 (Nikkon, Tokyo, Japan) and photographed with a DC300FX digital camera and IM1000 software (Leica Microsystems Digital Imaging, Cambridge, UK). 2D blue-native (BN)/SDS–PAGE 2D BN/SDS–PAGE was performed generally according to the protocol described previously (Camacho-Carvajal et al., 2004). Briefly, RT112 cells preserved in 80 1C after drug treatment were lysed in appropriate lysis buffer and centrifuged at 15 000 g at 4 1C for 20 min, and the supernatant was used for BN–PAGE. BN gels were prepared as described (Camacho-Carvajal et al., 2004) and were run overnight at 4 1C. Next, for further separation in 2D SDS–PAGE, the lanes from first-dimension BN–PAGE were placed in 2D SDS–PAGE of the same thickness and overlaid with 1xSDS Laemmli loading buffer. The 2D run and immunoblotting were performed according to the standard protocols. Mouse xenograft model RT112 stably transfected with the pCAGGS-luc vector containing the luciferase gene was established (Niwa et al., 1991). To produce a subcutaneous xenograft model, tumor cells (5 106) were injected into the subcutaneous tissue of pathogen-free 6-week-old BALB/cAJcl-nu/nu mice (Nihon Clea, Japan). After growing for 6 days, mice with tumor xenografts were divided into four groups (4 mice per group) and treated as follows: (a) vehicle alone (phosphate-buffered saline); (b) CDDP at 6 mg/kg weekly; (c) Triptolide at 0.15 mg/ kg four times per week; (d) combination of (b) and (c). The doses and injection regimens for these drugs were based on reports published previously (Fidler et al., 2003; Yang et al., 2003). Sixteen days later, the mice were killed, and the tumors were dissected. Tumor growth patterns were measured twice a week on the basis of the photons emitted from luciferaseexpressing cells counted using the light-light chamber of a CCD camera system (Xenogen, Alameda, CA, USA) (Nogawa et al., 2005). Approval for these studies was obtained from the Animal Research Committee of Graduate School of Medicine, Kyoto University.

References Aaltonen V, Koivunen J, Laato M, Peltonen J. (2006). Heterogeneity of cellular proliferation within transitional cell carcinoma: correlation of protein kinase C alpha/betaI expression and activity. J Histochem Cytochem 54: 795–806. Asada M, Yamada T, Ichijo H, Delia D, Miyazono K, Fukumuro K et al. (1999). Apoptosis inhibitory activity of cytoplasmic p21 (Cipl/WAFl) in monocytic differentiation. EMBO J 18: 1223–1234. Bode AM, Dong Z. (2004). Post-translational modification of p53 in tumorigenesis. Nat Rev Cancer 4: 793–805. Camacho-Carvajal MM, Wollscheid B, Aebersold R, Steimle V, Schamel WW. (2004). Two-dimensional Blue native/SDS gel electro-

phoresis of multi-protein complexes from whole cellular lysates: a proteomics approach. Mol Cell Proteomics 3: 176–182. Carter BZ, Mak DH, Schober WD, McQueen T, Harris D, Estrov Z et al. (2006). Triptolide induces caspase-dependent cell death mediated via the mitochondrial pathway in leukemic cells. Blood 108: 630–637. Chang WT, Kang JJ, Lee KY, Wei K, Anderson E, Gotmare S et al. (2001). Triptolide and chemotherapy cooperate in tumor cell apoptosis. A role for the p53 pathway. J Biol Chem 276: 2221–2227. Chen BJ, Liu C, Cui X, Fidler JM, Chao NJ. (2000). Prevention of graft-versus-host disease by a novel immunosuppressant, PG490-88, Oncogene


4614 through inhibition of alloreactive T cell expansion. Transplantation 70: 1442–1447. Cibere J, Deng Z, Lin Y, Ou R, He Y, Wang Z et al. (2003). A randomized double blind, placebo controlled trial of topical Tripterygium wilfordii in rheumatoid arthritis: reanalysis using logistic regression analysis. J Rheumatol 30: 465–467. Cote RJ, Esrig D, Groshen S, Jones PA, Skinner DG. (1997). p53 and treatment of bladder cancer. Nature 385: 123–125. Coutts AS, La Thangue NB. (2005). The p53 response: emerging levels of co-factor complexity. Biochem Biophys Res Commun 331: 778–785. Demidenko ZN, Blagosklonny MV. (2004). Flavopiridol induces p53 via initial inhibition of Mdm2 and p21 and, independently of p53, sensitizes apoptosis-reluctant cells to tumor necrosis factor. Cancer Res 64: 3653–3660. Erster S, Mihara M, Kim RH, Petrenko O, Moll UM. (2004). In vivo mitochondrial p53 translocation triggers a rapid first wave of cell death in response to DNA damage that can precede p53 target gene activation. Mol Cell Biol 24: 6728–6741. Fidler JM, Li K, Chung C, Wei K, Ross JA, Gao M et al. (2003). PG490-88, a derivative of triptolide, causes tumor regression and sensitizes tumors to chemotherapy. Mol Cancer Ther 2: 855–862. Gartel AL, Tyner AL. (2002). The role of the cyclin-dependent kinase inhibitor p21 in apoptosis. Mol Cancer Ther 1: 639–649. Huang S, Shu L, Dilling MB, Easton J, Harwood FC, Ichijo H et al. (2003). Sustained activation of the JNK cascade and rapamycininduced apoptosis are suppressed by p53/p21(Cipl). Mol Cell 11: 1491–1501. Jiang XH, Tu SP, Cui JT, Lin MC, Xia HH, Wong WM et al. (2004). Antisense targeting protein kinase C alpha and betal inhibits gastric carcinogenesis. Cancer Res 64: 5787–5794. Jiang XH, Wong BC, Lin MC, Zhu GH, Kung HF, Jiang SH et al. (2001). Functional p53 is required for triptolide-induced apoptosis and AP-1 and nuclear factor-kappaB activation in gastric cancer cells. Oncogene 20: 8009–8018. Koivunen J, Aaltonen V, Koskela S, Lehenkari P, Laato M, Peltonen J. (2004). Protein kinase C alpha/beta inhibitor Go6976 promotes formation of cell junctions and inhibits invasion of urinary bladder carcinoma cells. Cancer Res 64: 5693–5701. Kulikov R, Boehme KA, Blattner C. (2005). Glycogen synthase kinase 3-dependent phosphorylation of Mdm2 regulates p53 abundance. Mol Cell Biol 25: 7170–7180. Kupchan SM, Court WA, Dailey Jr RG, Gilmore CJ, Bryan RF. (1972). Triptolide and tripdiolide, novel antileukemic diterpenoid triepoxides from Tripterygium wilfordii. J Am Chem Soc 94: 7194–7195. Lam KS, Liu R, Miyamoto S, Lehman AL, Tuscano JM. (2003). Applications of one-bead one-compound combinatorial libraries and chemical microarrays in signal transduction research. Acc Chem Res 36: 370–377. Lee KY, Chang W, Qiu D, Kao PN, Rosen GD. (1999). PG490 (triptolide) cooperates with tumor necrosis factor-alpha to induce apoptosis in tumor cells. J Biol Chem 274: 13451–13455. Mandil R, Ashkenazi E, Blass M, Kronfeld I, Kazimirsky G, Rosenthal G et al. (2001). Protein kinase Calpha and protein kinase Cdelta play opposite roles in the proliferation and apoptosis of glioma cells. Cancer Res 61: 4612–4619. Mansouri A, Ridgway LD, Korapati AL, Zhang Q, Tian L, Wang Y et al. (2003). Sustained activation of JNK/p38 MAPK pathways in response to cisplatin leads to Fas ligand induction and cell death in ovarian carcinoma cells. J Biol Chem 278: 19245–19256. Matsui Y, Ueda S, Watanabe J, Kuwabara I, Ogawa O, Nishiyama H. (2007). Sensitizing effect of galectin-7 in urothelial cancer to cisplatin through the accumulation of intracellular reactive oxygen species. Cancer Res 67: 1212–1220.

Miyata Y, Sato T, Ito A. (2005). Triptolide, a diterpenoid triepoxide, induces antitumor proliferation via activation of c-Jun NH2-terminal kinase 1 by decreasing phosphatidylinositol 3-kinase activity in human tumor cells. Biochem Biophys Res Commun 336: 1081–1086. Nishiyama H, Habuchi T, Watanabe J, Teramukai S, Tada H, Ono Y et al. (2004). Clinical outcome of a large-scale multi-institutional retrospective study for locally advanced bladder cancer: a survey including 1131 patients treated during 1990–2000 in Japan. Eur Urol 45: 176–181. Niwa H, Yamamura K, Miyazaki J. (1991). Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108: 193–199. Nogawa M, Yuasa T, Kimura S, Kuroda J, Sato K, Segawa H et al. (2005). Monitoring luciferase-labeled cancer cell growth and metastasis in different in vivo models. Cancer Lett 217: 243–253. Qiu D, Zhao G, Aoki Y, Shi L, Uyei A, Nazarian S et al. (1999). Immunosuppressant PG490 (triptolide) inhibits T-cell interleukin-2 expression at the level of purine-box/nuclear factor of activated T-cells and NF-kappaB transcriptional activation. J Biol Chem 274: 13443–13450. Sarkis AS, Bajorin DF, Reuter VE, Herr HW, Netto G, Zhang ZF et al. (1995). Prognostic value of p53 nuclear overexpression in patients with invasive bladder cancer treated with neoadjuvant MVAC. J Clin Oncol 13: 1384–1390. Shamon LA, Pezzuto JM, Graves JM, Mehta RR, Wangcharoentrakul S, Sangsuwan R et al. (1997). Evaluation of the mutagenic, cytotoxic, and antitumor potential of triptolide, a highly oxygenated diterpene isolated from Tripterygium wilfordii. Cancer Lett 112: 113–117. Southgate J, Hutton KA, Thomas DF, Trejdosiewicz LK. (1994). Normal human urothelial cells in vitro: proliferation and induction of stratification. Lab Invest 71: 583–594. Tan J, Zhuang L, Leong HS, Iyer NG, Liu ET, Yu Q. (2005). Pharmacologic modulation of glycogen synthase kinase-3beta promotes p53-dependent apoptosis through a direct Bax-mediated mitochondrial pathway in colorectal cancer cells. Cancer Res 65: 9012–9020. Wan CK, Wang C, Cheung HY, Yang M, Fong WF. (2006). Triptolide induces Bcl-2 cleavage and mitochondria dependent apoptosis in p53-deficient HL-60 cells. Cancer Lett 241: 31–41. Watanabe J, Nishiyama H, Matsui Y, Ito M, Kawanishi H, Kamoto T et al. (2006). Dicoumarol potentiates cisplatin-induced apoptosis mediated by c-Jun N-terminal kinase in p53 wild-type urogenital cancer cell lines. Oncogene 25: 2500–2508. Watcharasit P, Bijur GN, Song L, Zhu J, Chen X, Jope RS. (2003). Glycogen synthase kinase-3beta (GSK3beta) binds to and promotes the actions of p53. J Biol Chem 278: 48872–48879. Watcharasit P, Bijur GN, Zmijewski JW, Song L, Zmijewska A, Chen X et al. (2002). Direct, activating interaction between glycogen synthase kinase-3beta and p53 after DNA damage. Proc Natl Acad Sci USA 99: 7951–7955. Weiss RH, Marshall D, Howard L, Corbacho AM, Cheung AT, Sawai ET. (2003). Suppression of breast cancer growth and angiogenesis by an antisense oligodeoxynucleotide to p21(Wafl/Cipl). Cancer Lett 189: 39–48. Weiss RH. (2003). p21Wafl/Cipl as a therapeutic target in breast and other cancers. Cancer Cell 4: 425–429. Yang S, Chen J, Guo Z, Xu XM, Wang L, Pei XF et al. (2003). Triptolide inhibits the growth and metastasis of solid tumors. Mol Cancer Ther 2: 65–72. Yinjun L, Jie J, Yungui W. (2005). Triptolide inhibits transcription factor NF-kappaB and induces apoptosis of multiple myeloma cells. Leuk Res 29: 99–105.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc).

Oncogene