5-aza-2 -deoxycytidine upregulates caspase-9 expression ... - Nature

Oncogene (2004) 23, 6779–6787

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

5-aza-20 -deoxycytidine upregulates caspase-9 expression cooperating with p53-induced apoptosis in human lung cancer cells Yoshihito Gomyo1,3, Ji-ichiro Sasaki*,1,3, Cynthia Branch1, Jack A Roth1,2 and Tapas Mukhopadhyay1,4 1 Department of Thoracic and Cardiovascular Surgery and Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA; 2Department of Molecular and Cellular Oncology, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA

Treating lung cancer cell lines using low-dose 5-aza-20 deoxycytidine (DAC) caused an accumulation of procaspase-9 through mRNA upregulation, but the cells did not undergo apoptosis. However, when cells were treated with DAC and infected with a low dose of a recombinant wildtype p53 adenovirus vector (Ad-p53), a synergistic growth inhibitory effect was observed. Combination treatment induced Apaf-1 and procaspase-9 expression in which cytochrome c releases by Ad-p53 triggered the mitochondrial pathway of apoptosis. Selective blockage of caspase9 activities by Z-LEHD-FMK completely attenuated DAC-induced enhancement of apoptosis mediated by Adp53 infection, and ectopic overexpression of procaspase-9 sensitized cells to Ad-p53-induced apoptosis in p53-null cells. In addition, DAC sensitized lung cancer cells to cisplatin and paclitaxel. Induction of the mitochondrial pathway of apoptosis using a slightly toxic dose of DAC may therefore be a strategy for treating lung cancer, and DAC treatment may have clinical implications when combined with chemotherapy or apoptosis-inducing gene therapy. Oncogene (2004) 23, 6779–6787. doi:10.1038/sj.onc.1207381 Published online 26 July 2004 Keywords: 5-aza-2-deoxycytidine; caspase-9; p53; lung tumor; apoptosis

Introduction Induction of apoptotic pathways in cancer therapy is often complicated by molecular defects in apoptosis *Correspondence: J Sasaki, Department of Thoracic and Cardiovascular Surgery, Unit 023, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030, USA; E-mail: [email protected] 3 These authors contributed equally to this work 4 Current address: National Center for Human Genome Studies and Research, Panjab University Pharmaceutical Extension, Block Sector 14, Chandigarh 160014, India Received 29 July 2002; revised 28 August 2003; accepted 17 November 2003; published online 26 July 2004

signaling pathways. One way tumor cells bypass apoptotic pathways is by silencing several genes through DNA methylation. Aberrant hypermethylation of CpG islands in cancer cells has been implicated in the transcriptional inactivation of many genes, including several tumor suppressor genes: Rb (Ohtani-fujita et al., 1993), p16 (Herman et al., 1995), estrogen receptor (Lapidus et al., 1996), E-cadherin (Yoshiura et al., 1995), glutathione S-transferase (Lee et al., 1994), and some genes involved in cell cycle control (Myeroff et al., 1995). Methylation of CpG sites near promoter regions also has long been associated with inactivation of tissue-specific and developmentally regulated genes (Doerfler, 1983). The DNA methyltransferase inhibitor 5-aza-20 -deoxycytidine (DAC) has been used to reverse methylation and reconstitute the expression of silenced genes (Jones, 1996), which has sparked interest in the use of DAC in the treatment against cancer. DAC has been shown to have some clinical utility in the treatment of human hemoglobinopathies and malignancies (Charache et al., 1983; Pinto et al., 1984). In the present study, we sought to determine whether DAC could be used at a nontoxic dose to sensitize lung cancer cells to therapy. We found that agents that induce cytochrome c release from the mitochondria could initiate apoptosis in DAC-treated cells. Our experiments were performed to elucidate the molecular mechanisms that regulate the sensitivity of DAC-treated cells to apoptosis. We showed that an adenovirus vector carrying the human wild-type p53 gene regulated by the cytomegalovirus promoter (Ad-p53) induced cytochrome c release from the mitochondria, while DAC treatment resulted in upregulation of procaspase-9 and Apaf-1. Also, synergy in tumor-cell killing was achieved using a combination treatment strategy. We hypothesize that DAC induces recruitment of the Apaf-1 apoptosome with procaspase-9 and that p53 triggers release of cytochrome c from mitochondria, which in turn activates the executioner caspases, caspase-3 and caspase-7. These active executioners kill cells via proteolysis of key cellular substrates, thus activating the p53 apoptotic and mitochondrial pathways of apoptosis.

Enhancement of p53 by a methyltransferase inhibitor Y Gomyo et al

6780

Results Effect of DAC treatment on the growth of lung cancer cell lines To investigate the effect of DAC on the growth of H460 and H1299, cells from each line were treated using DAC at various concentrations, and a cell-growth assay was performed. Following DAC treatment for 72 h, both cell lines showed dose-dependent growth inhibition (Figure 1a). The dose–response curve indicated that dose-dependent growth inhibition occurred at a low DAC dose range; however, high-dose DAC treatment did not sensitize cells to further growth inhibition. DAC was not highly toxic in these cell lines at concentrations as high as 2.5 mM. However, at much higher concentrations (10–20 mM), DAC induced p53 and p21 protein expression, activated the p53 damage-response pathway, and arrested cells at G2/M (data not shown). Based on these observations, we further investigated the effect of DAC treatment mainly at 2.5 mM on molecular mechanisms regulating growth inhibition and programmed cell death. DAC dose- and time-dependent accumulation of procaspase-9 Because activation of mitochondrial pathways has been observed in most cases of drug-induced apoptosis (Debatin, 2000), we first sought to determine whether DAC treatment was associated with the activation of caspase-9, an active component of the apoptosome. H460 cells were treated using DAC at various doses for 72 h, and total cell lysates were subjected to sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis and immunoblot analysis using a specific anticaspase antibody. As shown in Figure 1b, DAC treatment induced procaspase-9 protein expression in a dose-dependent manner. Furthermore, in a time-dependent study, 2.5 mM DAC treatment slightly induced procaspase-9 accumulation after 24 h; however, incubation for a longer period (24–72 h) significantly increased procaspase-9 accumulation and produced minor accumulation of its cleavage product, a supposedly active form of caspase-9. The kinetics of procaspase-9 induction after DAC treatment was consistent with the drug’s dependence on cellular replication. Also, expression of procaspase-9 protein was 2.4-fold higher in H460 cells treated with 2.5 mM DAC for 72 h than in untreated cells. Similar results were observed with DAC treatment of H1299 cells (data not shown). These results suggested that DAC caused the dose-dependent, time-responsive accumulation of procaspase-9 in H460 cells, although the kinetics of active proteolysis was slow.

Figure 1 Induction of caspase-9 and Apaf-1 expression following DAC treatment. (a) H460 and H1299 cells were exposed to various doses of DAC for 72 h, and cell viability was measured using the XTT assay. A dose–response curve was established. (b) Dose- (left panel) and time-dependent (right panel) induction of procaspase-9 expression in H460 cells after DAC treatment using Western blot analysis. Actin was used as a control for equal protein loading. Expression of procaspase-9 protein was 2.4-fold higher in cells treated using 2.5 mM DAC for 72 h than in untreated control cells. (c) H460 cells were treated using 2.5 mM DAC for 24, 48, or 72 h. Caspase-9 mRNA expression was assessed using an RNase protection assay. Caspase-9 mRNA expression was induced in H460 cells following treatment using 2.5 mM DAC (upper panel). Expression of caspase-9 mRNA was 5.6-fold higher in cells treated using DAC for 72 h than in untreated control cells. L32 and GAPDH mRNA expressions were shown as internal controls for RNA loading (lower panel). C, control cells. (d) Induction of Apaf1 expression in DAC-treated H460 cells. Immunoblotting showed induction of Apaf-1 expression after 48 h of treatment using various doses of DAC (left panel). H460 cells were infected with Ad-p53 or Ad-LacZ and harvested at the indicated time points (right panel). Actin was used as a control for equal loading

Upregulation of caspase-9 mRNA by DAC Because DAC can reactivate several methylated genes in tumors and tumor-derived cell lines, we sought to determine whether procaspase-9 accumulation in lung cancer cells after DAC treatment was due to mRNA upregulation or post-translational modification of proOncogene

caspase-9. To that end, H460 cells were treated using 2.5 mM DAC for different durations, and total RNA was subjected to an RNase protection assay using caspase-9 as a template. DAC enhanced procaspase-9 mRNA expression in a time-dependent manner (Figure 1c). In


6781

addition, scanning densitometry values indicated 5.6fold higher levels of caspase-9 mRNA expression in cells treated with DAC for 72 h compared with that in untreated control cells. Induction of Apaf-1 protein expression by DAC or Ad-p53 Apaf-1, which has been reported as a molecule reexpressed by DAC as well as a transcriptional target of p53, has been identified as a proximal activator of caspase-9 during cell death (Zou et al., 1997; Moroni et al., 2001; Soengas et al., 2001). As reported previously, Western blotting demonstrated that H460 cells induced Apaf-1 protein expression in DAC-treated cells in a dose-dependent manner, reaching more than twofold that in untreated control cells (Figure 1d). Interestingly, the induction of Apaf-1 protein expression alongside a high level of procaspase-9 expression after DAC treatment did not cause caspase activation. Furthermore, the level of Apaf-1 protein expression increased in a time-dependent manner in Ad-p53-treated cells but not in Ad-LacZ-treated cells. DAC cooperation with Ad-p53-induced growth inhibition Results shown above gave us a hypothesis that DAC may sensitize cells to Ad-p53-induced apoptosis via induction of both Apaf-1 and caspase-9 expression. To determine whether DAC enhanced p53-mediated cellgrowth inhibition, we treated H1299 and H460 cells with DAC alone, Ad-p53 alone, or a combination of the two. Infection with Ad-p53 plus DAC extensively inhibited cell growth in these cells (Figure 2a). However, infection with the control virus Ad-LacZ in combination with DAC treatment did not significantly enhance growth inhibition in either cell line (data not shown). To further analyse the combined effects of DAC treatment and Ad-p53-induced growth suppression, we performed a two-dimensional inhibitory concentration (IC50) isobologram analysis. H460 and H1299 cells were treated with Ad-p53 and DAC at various doses, and cell viability was determined 72 h after treatment using the XTT assay. Isobolograms were then generated from the models to determine the presence of synergy, additivity, or antagonism between Ad-p53 and DAC. Figure 2b shows the isobolograms for the combination of Ad-p53 and DAC in each cell line. The area surrounded by the three lines representing modes 1, 2a, and 2b predicted the additive zone of DAC and Ad-p53. The combined data points fell within the envelope of additivity or were smaller than the predicted minimum data points, indicating that exposure to Ad-p53 and DAC treatment produced synergistic growth inhibition in both H460 and H1299 cells. Effect of DAC on p53-induced apoptosis To determine whether DAC treatment mediated the p53-induced apoptotic response, tumor cells from each treatment group were examined for apoptotic activity. Most of the morphological hallmarks associated with

apoptosis, including cell shrinkage, DNA fragmentation, and chromatin condensation, were detected in H460 cells infected with both Ad-p53 and DAC for 72 h (Figure 2c). DNA-strand breaks typical of apoptosis were also demonstrated with DNA ladder analysis (Figure 2d). Next, we sought to determine whether the combination treatment with DAC with Ad-p53 activated the apoptosis pathway. Therefore, we infected H460 and H1299 cells with Ad-p53 or control Ad-LacZ alone or in combination with DAC and performed Western blot analysis for apoptosis-related protein markers. In 48 h after treatment, treatment with DAC alone increased the level of Apaf-1 protein expression twofold; this expression also increased in Ad-p53infected cells as observed in Figure 1d. The combination of Ad-p53 and DAC had an additive effect on Apaf-1 protein expression when compared with Ad-p53 or DAC alone (Figure 3a). A clearly cleaved, active caspase-9 product was detected in Ad-p53- and DACtreated cells. In parallel, we found that a significant poly(ADP-ribose) polymerase (PARP) cleavage product, which indicated caspase-3 activation, was associated with caspase-9 activation in the Ad-p53- and DAC-treated cells but not in the untreated control cells. Expression of p53 was marginally induced after treatment with Ad-p53 in both cell lines and increased with treatment with DAC. However, DAC treatment with Ad-p53 did not enhance p21 expression compared with treatment with Ad-p53 alone. No changes in the level of Bcl-2 or Bcl-xL expression were observed in either cell lines after treatment with DAC and Ad-p53 (data not shown). Additionally, we detected expression of p16 protein, which has been reported as a methylated gene in cancer cells, after DAC treatment in H1299 cells having a normal allele of the p16 gene. This observation suggested that DAC treatment was enough to re-express proteins from methylated genes such as p16. Trigger for apoptosis by combination treatment with DAC and ad-p53 Overexpression of p53 has been shown to lead to the release of cytochrome c from mitochondria via both direct localization on mitochondrial membrane and transcriptional induction of releasing proteins such as Bax and p53AIP1 (Johnstone et al., 2002; Mihara et al., 2003). Thus, we treated H460 cells with Ad-p53 alone, DAC alone, or Ad-p53 plus DAC and determined the amount of cytochrome c released from the mitochondria using Western blotting. As shown in Figure 3b, cytochrome c was detected in the cytosol after Ad-p53 infection. The level of cytochrome c in the cytosol of DAC-treated, Ad-p53-infected cells was similar to that in the cytosol of Ad-p53-infected cells after 48 h of treatment. No significant differences in the level of cytosolic cytochrome c between DAC-treated and untreated control cells were detected. This finding suggested that DAC treatment did not cause cytochrome c release in these cells but that Ad-p53 induced cytochrome c release from the mitochondria. These results also suggested that the liberated cytochrome c Oncogene


6782

Figure 2 Sensitization of lung cancer cells to p53-induced growth inhibition and apoptosis by DAC. (a) H460 cells were treated with 1 mM DAC and/or 2500 vp/cell Ad-p53, and H1299 cells were treated with 0.5 mM DAC and/or 40 vp/cell Ad-p53. The XTT assay was performed 3 days after treatment. The absorbance ratio in the XTT assay of the nontreatment (con) to the treatment group is shown. (b) Two-dimensional isobologram analysis (at IC50) of H460 and H1299 cell growth after DAC and Ad-p53 treatment. Data from the combined treatment group () show the synergistic action of DAC and Ad-p53 (Po 0.05). (c) H460 cells were treated with DAC, Adp53, Ad-LacZ, or a combination of the three, and morphological changes associated with apoptosis were identified using a phasecontrast light microscope. Chromatin condensation was noted using a fluorescence microscope via staining of the nuclei with Hoechst 3342 fluorescent dye. (d) DNA-strand breaks in different treatment groups were detected after 48 h using DNA ladder analysis. Only cells that received DAC plus Ad-p53 showed DNA fragmentation

Oncogene


6783

Figure 4 Sensitization of cells to Ad-p53 by caspase-9 overexpression. (a) Western blot analysis of cell lysates derived from a mock transfectant (H1299/PCDNA) and caspase-9 transfectants (H1299/CAS9-5 and H1299/CAS9-6) 48 h after adenoviral treatment. Although caspase-9 proteins were expressed moderately and highly in H1299/CAS9-5 and H1299/CAS9-6, respectively, the transduction efficiency of GFP and p53 in these cells was equivalent. (b) Induction of apoptosis was further assessed using FACS analysis. A 48 h Ad-GFP treatment did not significantly affect production of sub-G0/G1 populations, but Ad-p53 treatment enhanced cell death in caspase-9 transfectants in a proteinexpression-dependent manner

Figure 3 DAC enhancement of p53-induced apoptosis via the mitochondrial caspase-9 pathway. (a) Western blot analysis of DAC-sensitized H460 and H1299 cells infected with Ad-LacZ or Ad-p53. P53 protein expression (upper panels), p21 protein expression (second panels), PARP protein expression and cleavage (third left panel), p16 protein expression (third right panel), Apaf-1 protein expression (fourth panels), and accumulation of procaspase-9 and cleavage products (fifth panels) are shown. Treatment using DAC alone induced the accumulation of procaspase-9 with no cleavage products, while treatment using DAC plus Ad-p53 reduced the accumulation of procaspase-9 and induced cleavage products. Actin was used as a loading control (bottom panels). Con, control cells. (b) Cytochrome c release was detected using Western blot analysis in the cytosolic fractions of H460 cells treated using Ad-p53 alone or with DAC but not in cells treated using DAC alone. (c) Induction of apoptosis was further assessed using FACS analysis. A 48 h DAC treatment alone had no effect on the production of sub-G0/G1 populations, but DAC plus Ad-p53 treatment enhanced cell death. Inclusion of 100 mM Z-LEHDFMK, a selective inhibitor of caspase-9, completely abolished the synergistic increase of apoptotic fraction with DAC and Ad-p53

initiated the formation of an apoptosome consisting of Apaf-1 oligomers as a trigger for apoptosis. Requirement of procaspase-9 overexpression for synergistic apoptosis by DAC and Ad-p53 Propidium iodide (PI) staining and flow cytometric analysis were performed to quantitate the percentage of cells undergoing apoptosis in response to different

treatments. We used the sub-G1 DNA content as a measure of apoptotic populations. Treatment with DAC alone did not induce significant apoptotic activity in these cells, even 72 h after treatment (Figure 3c). However, DAC sensitized the cells to Ad-p53 and induced the apoptotic population to 65% when compared with treatment with Ad-p53 alone, which induced 40% apoptosis in H460 cells. To address the significance of procaspase-9 overexpression, we determined the effect of Z-LEHD-FMK, a specific inhibitor of caspase-9, on the induction of apoptosis in Ad-p53transduced cells treated with DAC. Cells in different treatment groups were incubated for 72 h in the presence or absence of 100 mM Z-LEHD-FMK before both floating and attached cells were collected, fixed, and stained with PI. Treatment with Z-LEHD-FMK reduced the apoptotic population in Ad-p53-infected cells from 40 to 30%. The same treatment reduced the apoptotic population in Ad-p53-transduced cells with DAC from 65 to 30%. These results suggest that blockage of caspase-9 activity can inhibit the synergistic enhancement of apoptosis mediated by DAC when combined with Ad-p53. To confirm the role of caspase-9 in sensitization to p53-mediated apoptosis, we established stable cell lines that expressed caspase-9 at various protein levels from H1299 cells and examined apoptotic induction after Adp53 infection. As shown in Figure 4a, H1299/CAS 9-5 and H1299/CAS 9-6 cells expressed procaspase-9 Oncogene


6784

protein at different levels. The efficiency of protein expression induced by adenoviral gene transfer was equivalent in these cells, because the GFP and p53 proteins were equally expressed. The results from PI staining and flow cytometric analysis using these cell lines clearly indicated that procaspase-9 protein overexpression enhanced induction of apoptosis mediated by Ad-p53 (Figure 4b). Sensitization to chemotherapeutic agents by DAC Chemotherapeutic agents such as cisplatin and paclitaxel have been reported to induce apoptosis via mitochondrial pathways (Debatin, 2000). To determine whether DAC sensitizes cells to chemotherapeutic drugs, we performed an XTT assay to determine the IC50 of cisplatin and paclitaxel in the presence or absence of DAC in non-small cell lung cancer cells. The mean IC50 of cisplatin under normal culture condition was 3.39, 2.03, 10.27, and 8.77 mM in A549, H460, H1299, and H322 cells, respectively. In all of the cell lines analysed, DAC administered at the lowest concentration (0.25 mM) reduced the IC50 of cisplatin significantly (Figure 5a). The impact on reduction of ICs by DAC was greater in A549 and H460 cells (both of which had a wild-type p53 genotype) than those in H1299 and H322 cells (H1299, p53-null genotypes; H322, mutant p53 genotype). After paclitaxel treatment, the mean IC50 without DAC was 1.77, 3.23, 5.45, and 2.33 nM in A549, H460, H1299, and H322 cells, respectively. Although the 0.25 mM DAC treatment did not reduce the IC50, the 2.5 mM DAC

Figure 5 Sensitization of cells to cisplatin and paclitaxel by DAC. Cells were treated with various doses of cisplatin and paclitaxel with and without DAC. The XTT assay was performed 3 days after treatment. The relative IC50 values of cisplatin and paclitaxel in each cell line were calculated and plotted. (a) DAC significantly reduced the IC50 of cisplatin at a low concentration (0.25 mM) in all of the cell lines. (b) DAC significantly sensitized H460 cells to paclitaxel at a low concentration and the other cells at a high concentration (2.5 mM) (Po0.05) Oncogene

treatment reduced those significantly in all of the cell lines (Figure 5b).

Discussion Herein we report for the first time that DAC treatment induces caspase-9 expression in lung cancer cell lines. Caspase-9, the apex caspase of the mitochondrial pathway of apoptosis, plays a critical role in apoptotic initiation and progression (Kuida et al., 1998). However, regulation of the caspase-9 gene is not completely understood, because the gene promoter is not well characterized. Determining whether the induction of procaspase-9 mRNA expression after DAC treatment is due to direct dimethylation or mRNA stabilization will require further experimentation. However, it is possible that re-expression of methylated transcription factors following DAC treatment enhances caspase-9 transcription. A recent report on the rat caspase-9 promoter indicated that the 50 -flanking sequence of rat caspase-9 contained putative Sp-1 binding elements (Nishiyama et al., 2001). It was also reported that DAC treatment of MCF-7L cells induces the expression of transforming growth factor-b receptors II and I through increased Sp1 protein levels or activities. Transcriptional control of transforming growth factor-b receptors is dependent on Sp1/Sp3 protein levels or activities; DAC has been shown to have a combination of effects on Sp1 and Sp3 (Ammanamanchi et al., 1998). Also, modulation of Sp1 has been shown to be post-transcriptional, whereas Sp3 repression occurs through decreased transcription of the Sp3 gene (Ammanamanchi and Brattain, 2001). Furthermore, we have observed that DAC treatment increases Sp1 protein expression twofold in H460 cells (data not shown). However, the role of increased Sp1 expression on procaspase-9 transcriptional activation remains elusive. While DAC treatment had no effect on caspase-3, caspase-8, or proapoptotic proteins of the Bcl-2 family (data not shown), single treatment with DAC or Ad-p53 induced expression of Apaf-1, an upstream transactivator of apoptosis. Apoptosis via the mitochondrial pathway requires release of cytochrome c into the cytosol to initiate the formation of an oligomeric Apaf-1 apoptosome (Zou et al., 1997; Yoshida et al., 1998). This apoptosome recruits and activates caspase-9, which in turn activates caspase-3 and caspase-7, which then kill the cell through proteolysis. For example, Apaf-1-negative melanomas are invariably chemoresistant and unable to execute typical apoptosis in response to p53 activation. Restoring physiological levels of Apaf-1 through gene transfer or DAC treatment markedly rescues the apoptotic defects associated with Apaf-1 loss in melanoma cells (Soengas et al., 2001). In our study, DAC treatment induced Apaf-1 protein expression, suggesting that Apaf-1 is a member of a gene family that is highly methylated in tumors and tumor cell lines. Accumulation of the Apaf-1 protein following DAC treatment was documented in the study


6785

by Soengas et al. (2001), who showed that, although the Apaf-1 promoter had a CpG island, there was no change in methylation in this region after DAC treatment. Therefore, the mechanism of Apaf-1 accumulation following DAC treatment remains elusive. It is possible that methylation-related alterations of an enhancer element of Apaf-1 or of a methylated transcription factor lead to Apaf-1 expression. The results of the present study clearly show that DAC and Ad-p53 acted synergistically to inhibit cell growth and that DAC-induced sensitization of Ad-p53induced apoptosis was crucially mediated through increased caspase-9 expression, as inhibition of caspase-9 by Z-LEHD-FMK inhibited p53-induced apoptosis. Recently, it was reported that caspase-9 is required for p53-dependent apoptosis and chemosensitivity in a human ovarian cancer cell line (Wu and Ding, 2002). Also, it was suggested that genetic knockout of methyltransferase or DAC treatment activates the p53 DNA-damage response pathway (Jackson-Grusby et al., 2001; Karpf et al., 2001). Of course, the effect of DAC treatment on apoptosis-modulating genes in addition to caspase-9 or on other cellular genes may also contribute to its ability to sensitize cells to apoptosis, because several gene promoters are frequently methylated in tumors, including caspase-8 and tumor suppressor genes such as Rb1 and p16 (Jones and Laird, 1999; Kataoka et al., 2000; Fulda et al., 2001). However re-expression of these methylated genes by DAC may depend on the cell type, as the methylation status of genes is different in cell lines. Indeed, we detect re-expression of p16 protein after DAC treatment in H1299 cells in vitro as previously reported in human bladder tumor cells in vivo (Bender et al., 1998), but we did not observe caspase-8 mRNA and protein re-expression in the non-small cell lung cancer cells analysed in this study. In addition, combination treatment with DAC and Ad-p53 had synergistic growth-inhibitory effect and induced apoptosis in H460 cells, which have a homozygous deletion of p16 gene. These results support our hypothesis that sensitization by DAC may be related to overexpression of the apex caspase-9 pathway of apoptosis rather than re-expression of a tumor suppressor gene. In addition to Ad-p53 treatment, the use of chemotherapeutic agents produced synergistic growth inhibition with DAC in non-small cell lung cancer cell lines. One reported clinical trial evaluated the combination of DAC and cisplatin, which demonstrated synergy in vitro; however, the mechanism was not understood (Lenzi et al., 1995). Despite differences in the anticancer action and dependency on p53 between cisplatin and paclitaxel, the type of cell death with these drugs appears to be the same: apoptosis. Cisplatin and paclitaxel have been found to induce apoptosis via the mitochondria caspace-9 pathway (Jordan and CarmoFonseca, 2000; Wang et al., 2000). Our finding that DAC sensitized cells to chemotherapeutic agents was consistent with the hypothesis that DAC sensitization to apoptosis may be based on activation of mitochondria pathway with upregulation of caspase-9 and apaf-1 expression.

As summarized in Figure 6, we hypothesize that DAC treatment induces procaspase-9 and Apaf-1 expression at high levels but cannot induce enough apoptosis. We also postulate that p53 overexpression causes cytochrome c release, initiates apoptosis of DAC-sensitized cells, and has a synergistic effect with DAC on cell death. Inducing the mitochondrial pathway of gene expression in tumor cells may considerably affect the efficacy of p53 gene therapy and chemotherapeutic drugs. Furthermore, DAC may play a role in increasing the efficacy of various forms of chemotherapy. Our results suggest the potential use of DAC treatment as an adjuvant modality for cancer if administered along with apoptosis-inducing gene therapy.

Materials and methods Reagents and antibodies DAC (Sigma-Aldrich, St Louis, MO, USA) was dissolved in dimethyl sulfoxide (DMSO) at a concentration of 10 mM and stored at 201C as a master stock solution. Cisplatin and paclitaxel were purchased from Bristol-Myers Squibb Company (Princeton, NJ, USA). Mouse monoclonal and rabbit polyclonal antibodies used as primary antibodies for immunoblotting included anti-p53 (Bp53-12; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-p21 (Ab-1; Oncogene Research Science, Cambridge, MA, USA), anti-p16 (C-20; Santa Cruz Biotechnology), anti-b-actin (AC-15; SigmaAldrich), anti-PARP (4C10-5; BD Biosciences, San Jose, CA, USA), anti-Apaf-1 (24; BD Biosciences), anti-cytochrome-c (7H8.2C12; BD Biosciences), and anti-caspase 9 (catalog #9502; Cell Signaling Technology, Beverly, MA, USA). Cell lines and plasmids The cell lines H460, A549, H1299, and H322 (American Type Culture Collections, Manassas, VA, USA) were grown in RPMI medium containing 10% heat-inactivated fetal bovine

Figure 6 Schematic representation of the speculated mechanism of DAC function in enhancement of p53-induced apoptosis Oncogene


6786 serum, glutamine, and antibiotics (Invitrogen, Carlsbad, CA, USA) at 371C in a humidified atmosphere containing 5% CO2. For establishment of stable cell lines expressing caspase-9, we obtained a plasmid containing human caspase-9 cDNA (MGC-3054) from the American Type Culture Collection. After double digestion with EcoR1 and HindIII, the fragment (1305 bp) was ligated into a pcDNA3.1() expression vector (Invitrogen). The pcDNA/caspase-9 construction was then transfected into H1299 cells by FuGENE 6 according to the manufacturer’s protocol. Drug selection in 400 mg/ml G418 (Invitrogen) was begun 48 h after transfection. At 2 weeks after drug selection, colonies were harvested with cloning cylinders and expanded to cell lines. Adenoviral vectors The recombinant adenoviral vector Ad-p53, which contains the wild-type p53 gene, was used as a gene therapy agent. AdLacZ and Ad-GFP, which contain a b-galactosidase and GFP gene, respectively, were used as nonspecific negative control. Construction of these recombinant adenoviral vectors was described previously (Zhang et al., 1993). Each viral stock was prepared for our experiments by the Vector Core Facility at The University of Texas MD Anderson Cancer Center. Based on the transduction efficiency of H460 and H1299 (Gomyo et al., unpublished data), the cell lines were infected with 2500 and 40 vp/cell, respectively. Cytotoxicity/growth-inhibition assay Inhibition of tumor-cell growth was analysed by quantitatively determining cell viability using an improved XTT assay (Roche Molecular Biochemicals, Indianapolis, IN, USA) according to the manufacturer’s instructions. Cells (2 103) were plated and precultured for 24 h in 96-well plates and then incubated in the presence of an adenovirus, cisplatin, paclitaxel, or DAC for 24–72 h. The XTT assay was used to determine percent cell growth 3 days after the initial addition of DAC. The experiments were repeated at least two times. Isobologram The effect of combined DAC and Ad-p53 treatment was analysed using two-dimensional isobolograms originally described by Steel and Peckham (1979) and improved by our group (Nishizaki et al., 2001). The envelope of additivity in the isobologram was defined by three isoeffect lines: mode 1, mode 2a, and mode 2b. The effects of the combined therapeutic agents on tumor-cell growth were analysed quantitatively and statistically by plotting the observed experimental data onto the corresponding isobolograms according to methods described previously (Kano et al., 1998). DNA fragmentation To identify DNA fragmentation, cells were harvested and lysed using lysis buffer (10 mM Tris–HCl buffer, 10 mM ethylenediaminetetraacetic acid, 0.5% Triton X-100) after indicated treatments. After centrifuging, supernatants containing DNA fragments were collected and treated with RNase and proteinase K, and the DNA fragments were precipitated. DNA was then dissolved in TE buffer, electrophoresed on a 2% agarose gel, and stained using ethidium bromide. RNase protection assay H460 cells were treated using 2.5 mM DAC for 24, 48, or 72 h. Total RNA was isolated from these cells using the TRIzol Oncogene

reagent (Invitrogen) according to the manufacturer’s protocol. Next, an RNA protection assay was performed using the RiboQuant kit (BD Biosciences). Hybridized RNA was resolved on a 5% denaturing polyacrylamide gel; the gel was then dried and exposed to an X-ray film (Eastman Kodak, Rochester, NY, USA). Glyceraldehyde-3-phosphate dehydrogenase and L32 served as controls. Flow cytometric analysis At 48 h after treatment using DAC at various doses and for various durations, both floating and adherent cells were harvested, fixed in 70% ethanol, and stained with PI. Apoptotic cells were quantified via flow cytometric analysis. A caspase inhibition assay was also performed, in which H460 cells were treated with DAC alone, Ad-p53 alone, or DAC plus Ad-p53 as described above. Appropriate control and treated groups were then incubated for 48 h with 100 mM Z-LEHDFMK (CALBIOCHEM, La Jolla, CA, USA), a selective inhibitor of caspase-9. Cells were then harvested and fixed in 70% ethanol. Finally, cells were stained with PI, and the percentages of sub-G0/G1 populations were determined using fluorescence-activated cell sorting analysis. Cell lysate, cellular fractionation for cytochrome c and immunoblotting For general immunoblotting, cells were lysed in 1 SDS sample buffer (62.5 mM Tris–HCl, pH 6.8, 2% SDS, 10% glycerol, 1 proteinase inhibitor cocktail (Complete; Roche), and 6 M urea). The protein concentrations in the extracts were measured using the bicinchoninic acid (BCA) protein assay kit (Pierce Chemical Co., Rockford, IL, USA). A measure of 50 mg of the protein was then used for immunoblotting. For cytochrome c analysis, cell fractionation was performed using the ApoAlert cell fractionation kit (Clontech, Palo Alto, CA, USA) according to the manufacturer’s instructions. The protein concentration in each fraction was determined using the Bradford method (Bio-Rad, Hercules, CA, USA), and immunoblotting was performed. Blots were then probed with antibodies, and immunocomplexes were detected using the enhanced chemiluminescence kit according to the manufacturer’s instructions (Amersham Biosciences, Arlington Heights, IL, USA). Afterwards, blots were reprobed with an anti-actin antibody. Statistical analysis Descriptive statistics, such as means and standard deviations, were reported to summarize the study results. In addition, the two-sample t-test was performed to compare apoptotic populations and relative IC50 values in cell lines treated under various conditions. To determine the significance of synergism in combinations in isobolograms, Wilcoxon’s signed rank tests were performed to compare the mean values of the observed data with the predicted minimum or maximum values for the additivity. All of the statistical analyses were performed using the Statistica software program (StatSoft, Inc., Tulsa, OK, USA). Acknowledgements We thank Dr Masahiko Nishizaki for technical advice, and Carmelita Concepcion and Peggy James for the preparation of this manuscript. This work was partially supported by a grant from the National Cancer Institute, National Institutes of Health (PO1 CA78778-01A1; to JAR), the SPORE grant (2P50-CA70970-04), a developmental grant from the National Cancer Institute to MD Anderson Cancer Center SPORE in


6787 lung cancer; P50-CA70907 to TM, the Cancer Center Support Core Grant CA 16672 to the Division of Surgery at MD Anderson, a grant from the Tobacco Settlement Funds as

appropriated by the Texas Legislature (Project 8), the WM Keck Foundation, and a sponsored research agreement with Introgen Therapeutics, Inc. (SR93-004-1).

References Ammanamanchi S, Kim SJ, Sun LZ and Brattain MG. (1998). J. Biol. Chem., 273, 16527–16534. Ammanamanchi S and Brattain MG. (2001). J. Biol. Chem., 276, 32854–32859. Bender CM, Pao MM and Jones PA. (1998). Cancer Res., 58, 95–101. Charache S, Dover G, Smith K, Talbot Jr CC, Moyer M and Boyer S. (1983). Proc. Natl. Acad. Sci. USA, 80, 4842–4846. Debatin KM. (2000). Toxicol. Lett., 112–113, 41–48. Doerfler W. (1983). Annu. Rev. Biochem., 52, 93–124. Fulda S, Kufer MU, Meyer E, Valen F, Dockhorn-Dworniczak B and Debatin KM. (2001). Oncogene, 20, 5865–5877. Herman JG, Merlo A, Mao L, Lapidus RG, Issa JP, Davidson NE, Sidransky D and Baylin SB. (1995). Cancer Res., 55, 4525–4530. Jackson-Grusby L, Beard C, Possemoto R, Tudor M, Fambrough D, Csankovszki G, Dausman J, Lee P, Wilson C, Lander E and Jaenisch R. (2001). Nat. Genet., 27, 31–39. Johnstone RW, Ruefli AA and Lowe SW. (2002). Cell, 108, 153–164. Jones PA and Laird PW. (1999). Nat. Genet., 21, 163–167. Jones PA. (1996). Cancer Res., 56, 2463–2467. Jordan P and Carmo-Fonseca M. (2000). Cell Mol. Life Sci., 57, 1229–1235. Kano Y, Akutsu M, Tsunoda S, Mori K, Suzuki K and Adachi KI. (1998). Cancer Chemother. Pharmacol., 42, 91–98. Karpf AR, Moore BE, Ririe TO and Jones DA. (2001). Mol. Pharmacol., 59, 751–757. Kataoka M, Wiehle S, Spitz F, Schumacher G, Roth JA and Cristiano RJ. (2000). Oncogene, 19, 1589–1595. Kuida K, Haydar TF, Kuan CY, Gu Y, Taya C, Karasuyama H, Su MS, Rakic P and Flavell RA. (1998). Cell, 94, 325–337. Lapidus RG, Ferguson AT, Ottaviano YL, Parl FF, Smith HS, Weitzman SA, Baylin SB, Issa JJ and Davidson NE. (1996). Clin. Cancer Res., 2, 805–810. Lee WH, Morton RA, Epstein JI, Brooks JD, Campbell PA, Bova GS, Hsieh WS, Isaacs WB and Nelson WG. (1994). Proc. Natl. Acad. Sci. USA, 91, 11733–11737.

Lenzi R, Raber MN, Gravel D, Frost P and Abbruzzese JL. (1995). Int. J. Oncol., 6, 447–450. Mihara M, Erster S, Zaika A, Petrenko O, Chittenden T, Pancoska P and Moll UT. (2003). Mol. Cell, 11, 577–590. Moroni MC, Hickman ES, Denchi EL, Caprara G, Colli E, Cecconi F, Muller H and Helin K. (2001). Nat. Cell Biol., 3, 552–558. Myeroff LL, Parsons R, Kim SJ, Hedrick L, Cho KR, Orth K, Mathis M, Kinzler KW, Lutterbaugh J and Park K. (1995). Cancer Res., 55, 5545–5547. Nishiyama J, Yi X, Venkatachalam MA and Dong Z. (2001). Biochem. J., 360, 49–56. Nishizaki M, Meyn RE, Levy LB, Atkinson EN, White RA, Roth JA and Ji L. (2001). Clin. Cancer Res., 7, 2887–2897. Ohtani-Fujita N, Fujita T, Aoike A, Osifchin NE, Robbins PD and Sakai T. (1993). Oncogene, 8, 1063–1067. Pinto A, Attadia V, Fusco A, Ferrara F, Spada OA and Di Fiore PP. (1984). Blood, 64, 922–929. Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, Opitz-Araya X, McCombie R, Herman JG, Gerald WL, Lazebnik YA, Cordon-Cardo C and Lowe SW. (2001). Nature, 409, 207–211. Steel GG and Peckham MJ. (1979). Int. J. Radiat. Oncol. Biol. Phys., 5, 85–91. Wang TH, Wang HS and Soong YK. (2000). Cancer, 88, 2619–2627. Wu GS and Ding Z. (2002). Oncogene, 21, 1–8. Yoshida H, Kongr YY, Yoshida R, Elia AJ, Hakem A, Hakem R Penninger JM and Mak TW. (1998). Cell, 94, 325–337. Yoshiura K, Kanai Y, Ochiai A, Shimoyama Y, Sugimura T and Hirohashi S. (1995). Proc. Natl. Acad. Sci. USA, 92, 7416–7419. Zhang WW, Fang X, Branch CD, Mazur W, French BA and Roth JA. (1993). Biotechniques, 15, 868–872. Zou H, Henzel WJ, Liu X, Lutschg A and Wang X. (1997). Cell, 90, 405–413.

Oncogene