Apoptosis 2004; 9: 797–805
C 2004 Kluwer Academic Publishers Bcl-2 and CCND1/CDK4 expression levels predict the cellular effects of mTOR inhibitors in human ovarian carcinoma D. Aguirre∗ , P. Boya∗ , D. Bellet, S. Faivre, F. Troalen, J. Benard, P. Saulnier, S. Hopkins-Donaldson, U. Zangemeister-Wittke, G. Kroemer∗ and E. Raymond∗ Department of Clinical Biology, Department of Medicine (D. Aguirre, D. Bellet, S. Faivre, F. Troalen, J. Benard, P. Saulnier, E. Raymond), and CNRS-UMR 8125 (P. Boya, G. Kroemer), Institute Gustave-Roussy, Villejuif, Tumor Immunology Laboratory FRE2443 CNRS, University René Descartes Paris 5; France; Division of Medical Oncology, University Hospital, Zurich, Switzerland (S. Hopkins-Donaldson, U. Zangemeister-Wittke) Molecular markers enabling the prediction of sensitivity/resistance to rapamycin may facilitate further clinical development of rapamycin and its derivatives as anticancer agents. In this study, several human ovarian cancer cell lines (IGROV1, OVCAR-3, A2780, SK-OV-3) were evaluated for susceptibility to rapamycin-mediated growth inhibition. The differential expression profiles of genes coding for proteins known to be involved in the mTOR signaling pathway, cell cycle control and apoptosis were studied before and after drug exposure by RT-PCR. In cells exposed to rapamycin, we observed a dose-dependent downregulation of CCND1 (cyclin D1) and CDK4 gene expression and late G1 cell cycle arrest. Among these cell lines, SK-OV-3 cells resistant to both rapamycin and RAD001 were the sole to show the expression of the anti-apoptotic gene Bcl-2. Bcl-2/bclxLspecific antisense oligonucleotides restored the sensitivity of SK-OV-3 cells to apoptosis induction by rapamycin and RAD001. These results indicate that baseline Bcl2 expression and therapy-induced downexpression of CCND1 and CDK4 may be regarded as molecular markers enabling the prediction and follow-up of the cellular effects on cell cycle and apoptosis induction of rapamycin in ovarian cancer. Furthermore, strategies to down regulate Bcl-2 in ovarian cancer may prove useful in combination with rapamycin or RAD001 for ovarian cancer. Keywords: apoptosis; cell cycle; gene expression; RAD001; rapamycin; surrogate marker. Abbreviations: 4EBP, eukaryotic initiation factor 4E binding protein; CDK, cyclin-dependent kinase; EGFR, epithelial growth factor receptor; IGF: insulin growth factor; MDR-1, multidrug resistance; PI, propidium iodide; PI3K, phosphoinositol- ∗ They equally contributed to this work and should be considered as joint first and senior authors, respectively. Correspondence to: Guido Kroemer, M.D., Ph.D. Tumor Immunology Laboratory, Institut Gustave Roussy, 39, Rue Camille Desmoulins, 94805 Villejuif, Cedex, France. Tel.: 33 1 4211 6046; Fax: 33 1 4211 5244; e-mail: Kroemer@igr.fr 3 kinases, PTEN, phosphatase and tensin homologue; S6K or p70S6K, ribosomal p70 S6 kinase; TBBP, tata box binding protein; mTOR, mammalian target of rapamycin; VEGF, vascular endothelial growth factor. Introduction Genetic alterations that disrupt the regulation of proliferation, cell death, and senescence are hallmarks of malignant transformation of ovarian epithelial cells.1 Ovarian cancer cells produce and/or respond to various growth factors such as EGF, TGFβ, IGF, PDGF, and stem cell growth factor.1 Interactions between growth factors and their specific receptors can trigger downstream signaling pathways involving PI3K, MAPK, AKT and the mammalian target of rapamycin (mTOR).1,2 In cancer cells, AKT and mTOR act as “master switch” proteins which simultaneously modulate metabolism, the cell cycle, and apoptosis.3 In endometrioid ovarian cancer cells, AKT-dependent cellular proliferation is repressed by PTEN, a phosphatase that antagonizes the kinase activity of PI3K.1,2 Activation of the AKT/mTOR pathway in ovarian cancers modulates the response to apoptotic stress induced by cytotoxic agents such as cisplatin and paclitaxel.4,5 Rapamycin (Sirolimus, RapamuneTM ), a macrolide produced by Streptomyces hygroscopicus, binds FKBP-12 (FK506 binding protein), thereby creating a molecular complex that specifically inhibits the function of mTOR.6 The anticancer effects of rapamycin have stimulated the development of derivatives such as RAD001 (SDZ RAD, Everolimus, CerticanTM ), which are being included in clinical trials and show early evidence of antitumor activity either as a single agent or in combination with chemotherapy.6 Inhibition of mTOR by rapamycin or RAD001 leads to the downregulation of G1 cyclin/Cdk complexes, thereby blocking the cell cycle in the late G1/S phase.7 Recent studies have revealed that rapamycin inhibits the Apoptosis · Vol 9 · No 6 · 2004 797 D. Aguirre et al. oncogenic transformation of human cells induced by either PI3K or AKT and the loss of PTEN function.8 Furthermore, rapamycin antagonizes tumor growth induced by loss of the PI3K antagonist PTEN. PTEN+/− mice spontaneously develop neoplasia associated with loss of the normal PTEN allele and increased activation of AKT/PKB and the ribosomal p70 S6 kinase (S6K), a known downstream target of mTOR.8 In vivo treatment of such mice with the rapamycin analog CCI-779 normalizes S6K activity and reduces neoplastic proliferation.9 Similarly, PTEN-deficient human tumors are more sensitive to CCI-779-mediated growth inhibition than PTENexpressing cells. The mechanisms through which mTOR inhibits apoptosis remain to be elucidated.10–12 As a possibility, its downstream target S6K, which can bind to mitochondrial membranes, may phosphorylate the proapoptotic molecule BAD on serine 136, a reaction that disrupts BAD’s binding to the mitochondrial death inhibitors bcl-xL and bcl-2 and thus inactivates BAD.13,14 Rapamycin and its derivatives may thus suppress cancer cell proliferation due to cell cycle inhibition and/or due to an increase in apoptosis, depending on the type of cancer cell. The identification of molecular determinants of cell cycle and apoptotic effects of rapamycin should ultimately facilitate the clinical development of rapamycin derivatives. This study was designed to identify molecular markers associated with cell cycle and apoptotic effects by quantifying the expression of genes modulated by rapamycin. The selection of putative marker genes was based on previous studies identifying essential genes involved in the mTOR signaling pathway, cell cycle control, and apoptosis in Saccharomyces cerevisiae as well as in mammalian cells.15,16 In this study, we show that expression of Cyclin D1/CDK4 can be used to monitor the cell cycle effects of rapamycin in ovarian cancer. Moreover, we show that Bcl-2 expression is a marker of resistance to apoptosis induced by rapamycin. Materials and methods Cell culture and cytotoxic assays The ovarian cancer cell lines OVCAR-3, SK-OV-3 (American Type Culture Collection), IGROV1 (kindly provided by Jean Benard, Institute Gustave-Roussy, Villejuif, France), and A2780 (kindly provided by Dr. Larsen, Institute Gustave Roussy, Villejuif, France) were maintained in RPMI 1640 medium (GIBCO/BRL) supplemented with 10% FCS (BioWhittaker) and 2 mM L-glutamine (GIBCO/BRL) at 37◦ C in a humidified atmosphere of 5% CO2 . Cells were determined to be free of mycoplasma infection, and cell viability was assessed by means of a tetrazolium salt-based colorimetric assay (WST from Roche Diagnostics GmbH, Mannheim, Germany). 798 Apoptosis · Vol 9 · No 6 · 2004 Cells (200 µl per microtiter well) were treated with increasing concentrations of rapamycin (Sigma, France) or RAD001 (Novartis Institutes for BioMedical Research, Basel, Switzerland) 6, 24, 48 or 72 h. The assay was terminated by the addition of the WST reagent (20 µl/well) and incubation of the cells at 37◦ C for 30′ for SK-OV3 and 1 h for the other cell lines. Percent cell survival was defined as the relative absorbance of control versus treated cells. All assays were performed in quadruplicate and repeated at least twice. Real-time PCR (TaqMan) Total RNA was extracted from cell lines by RNA-Plus according to the practical recommendations of the manufacturer (Q-Biogene, Illkirch, France). Total RNA concentration was determined at 260 nm (Gene-Quant, Pharmacia) and its quality was assessed by conventional gel electrophoresis. Standard curves were constructed with 5 serial dilutions of total RNA. We used the universal human reference RNA (20, 4, 0.8, 0.16 and 0.032 ng per tube) from Stratagene (La Jolla, CA, USA) as an internal standard. Real-time PCR was carried out with the Applied Biosystems PCR system. 1 µg of RNA was reverse-transcribed under standard conditions (1× RTPCR buffer, 2 mM dNTP, 5 mM MgCl2 , 20 U RNase inhibitor, 50 U reverse transcriptase, 5 µM random hexamers). The samples were incubated at 42◦ C for 45 min, and reverse transcriptase was inactivated by heating at 100◦ C for 5 min and cooling at 4◦ C for 10 min. All PCR reactions were performed using an ABI Prism 7700 Sequence Detection system (Perkin-Elmer Applied Biosystems Inc, Foster City, CA, USA). For each PCR run, we used final concentrations of the master PCR mix (TaqMan UMM buffer, 20 pmol of each primer, 10 pmol of probe). 10 µl of each appropriately diluted cDNA were added to 40 µl of the PCR master-mix. The thermal cycling conditions comprised an initial denaturation step at 95◦ C for 10 min, 40 cycles at 95◦ C for 15 s, and 60◦ C for 1 min. Each PCR run included the five points of the standard curve, a no-template control, the calibrator cDNA, and three samples of cDNAs from a specific cell line exposed to different drug concentrations (10− 7 , 10− 6 and 10− 5 M). The target gene mRNA copy value was obtained in 2 h with this assay format. For each experimental sample, the amounts of the targets and endogenous reference were determined from the standard curve. Then the target amount was divided by the endogenous reference amount to obtain a normalized target value. The relative gene target expression level was also normalized to a control without treatment for each cell line (calibrator), or 1× sample. Each of the normalized target values was divided by the calibrator normalized target value to generate the final relative expression levels. Primers and probes for the reference gene TBBP Biomarkers for mTOR inhibitors in ovarian carcinoma (TATA-box binding protein, a component of protein complex TFIID) and target genes were chosen with the assistance of the computer program Primer Express (Perkin-Elmer Applied Biosystems, Foster City, CA, USA). Primers and probes were purchased from MWGBiotech AG. The nucleotide sequences of the oligonucleotide hybridization probes and primers are available under request. Experiments were performed with duplicates for each data point. The following genes were analyzed: AKT2, BAD, BCL2, CASP9, CCNB1, CCND1, CDK1, CDK4, CDKN1A, CDKN2A, EGFR, ERBB2, FAS, IGF1, KRAS2, MDR1, mTOR, MYC, PIK3CA, PTEN, RPLPO, TGFB1, TP53, VEGFA, and TBBP. Antisense oligonucleotides and transfection The chemoresistant line SK-OV-3 was transfected with the Bcl-2/Bcl-xL antisense oligonucleotide 4625, a second-generation phosphothioate oligonucleotide with modifications at the 5′ and 3′ ends consisting of methoxy-ethoxy residues at the 2′ -alpha-position of the deoxyribose.17,18 This oligonucleotide has 100% complementarity with Bcl-2 and 3 mismatches with Bcl-xL. As the scrambled control, we used 4626 (the same nucleotide composition, but in a random order). SK-OV-3 cells were plated at 75,000 cells/ml. On the next day, oligonucleotides (200 nM) were delivered to cells in the form of complexes with the transfection reagent Lipofectin (InVitrogen-Life Technologies) as described previously. Lipofectin was allowed to complex with oligonucleotides in serum- and antibiotic-free medium before dilution and addition to cells, which were exposed to rapamycin or Rad 001 (10− 5 M or 10− 7 M) 18 h later. To confirm the downregulation of Bcl-2, cells were harvested after 18 h transfection evaluated by RT-PCR. Western blot Cells were lysed for 15 min in 50 mM HEPES, 150 mM NaCl, 5 mM EDTA, 0.1% NP-40, supplemented with protease inhibitor cocktail (Roche), 1 mM dithiothreitol and 1 mM PMSF, and then centrifuged at 13,000 g for 10 min to remove debris. 40 µg of protein were loaded onto a 12% SDS-PAGE, transferred on nitrocellulose and blotted with mouse anti-human Bcl-2 (Ab 124, DAKO, Denmark) and anti-GAPDH (Chemicon). V-FITC and propidium iodide for 15 min at room temperature, followed by evaluation with a FACS Vantage cytofluorometer (Becton Dickinson). Results Gene expression profile in ovarian cancer cells Quantitative comparison of mRNA species revealed that a panel of four different ovarian cancer cell lines (IGROV1, SK-OV-3, OVCAR-3, and A2780) all expressed AKT2 and mTOR (Figure 1A), as well as mRNA species coding for cell cycle regulatory protein required for G1/S transition (cyclin D1 and Cdk4), G2/M transition (CCNB1 and CDK1) and inhibition of cell cycle progression (CDKN1A and CDKN2A) (Figure 1B). A2780 expressed the lowest level of several mRNA species (e.g. PI3K, PTEN, EGFR, CDKN2A), consistent with reports by the National Cancer Institute using cDNA microarrays (genome-www.edu/nci60/). The expression of mRNA coding BAD, c-myc and p53 was detected in all ovarian cancer cell lines (Figure 1C). Expression of mRNA species coding for Bcl-2 and Bcl-xL proteins was found only in SK-OV-3 cells, a result that was confirmed by immunoblot detection at the protein level (Figure 1C and D). Antiproliferative effects of mTOR inhibitors in ovarian cancer cells Inhibition of cellular proliferation was evaluated using several concentrations of rapamycin (chemical structures shown in Figure 2A). In our study, rapamycin more readily inhibited the growth of IGROV1 than that of the other three ovarian cancer cell lines, with SK-OV-3 cells being the most resistant (Figure 2B). Attempts to determine the GI50 of RAD001 was not satisfactory due to the poor solubility of this compound in media. Based on these data, we focused on IGROV1 and SK-OV-3 and selected concentrations ranging 10− 7 –10− 5 at which subsequent experiments were performed. In our study, steady state (baseline) expression levels of human homologues of yeast mRNA species modulated by rapamycin were found not to be predictive of the antiproliferative effects of rapamycin in human ovarian cancer cells. Exposure to rapamycin was shown to induce a downregulation of several genes involved in the the PI3 kinase and AKT pathway. Up regulation of caspase 9 and FAS was observed in IGROV-1 which was not found in SK-OV-3 (Figure 2C). Quantitation of apoptosis For the assessment of phosphatidyl serine exposure and viability, the Annexin V-FITC kit was used (Bender Medsystems). After the indicated treatments, cells were trypsinized and incubated in binding buffer with Annexin Effects of mTOR inhibitors on cell cycle mRNA species and cellular death Recent studies have shown that the Cyclin D1 is essential for mediating the proliferation signaled via the mTOR Apoptosis · Vol 9 · No 6 · 2004 799 D. Aguirre et al. Figure 1. Expression of mRNA species coding for key signaling molecules in ovarian cancer cells. RT-PCR analysis of genes coding for the mTOR pathway (A), cell cycle modulators (B) and apoptosis (C) were analyzed in a panel of 4 human cancer cell lines. Quantification was normalized using the housekeeping gene TBPP as control. Western blot using anti-Bcl-2 antibody showed that SK-OV-3 cells express higher Bcl-2 levels than IGROV1 cells (D): 50 µg pf protein from IGROV1 (1), SK-OV-3 (2), and HeLa cells stably transfected with a plasmid coding for human Bcl-2 (3) were analyzed using anti-GAPDH as loading control. pathway, and that rapamycin can modulate cellular proliferation by inhibiting cyclin D1 expression. In our study, IGROV1 and SK-OV-3 cells were incubated for 60 h in the presence of 10− 6 and 10− 7 M rapamycin, fixed and DNA content was analyzed by flow cytometry after DAPI staining. In ovarian cancer cells sensitive to rapamycin such as IGROV1, rapamycin and its derivatives blocked cell cycle progression in G1/S (not shown). Furthermore, exposure to rapamycin induced a decrease in mRNA encoding Cyclin D1, CDK4, CDKN1A, CDKN2A, and CCNB1 (Figure 2D), all genes involved in cell cycle control. Most notably, exposure of IGROV1 to rapamycin (and RAD001 data not shown) was associated with concentration-dependent inhibition of the expression of Cyclin D1 and CDK4 mRNA and cyclin D1 protein (data not shown). In contrast, levels of CDK1 mRNA increased after rapamycin treatment (Figure 2C). Cyclin D1 and CDK4 mRNA downregulation in IGROV1 was associated with increased expression of caspase 3 (data not shown) and increased cell death, as indicated by apoptotic shrinkage (Figure 3A and B). In rapamycin-resistant SK-OV-3 cells, exposure to mTOR inhibitors was also associated with a decrease 800 Apoptosis · Vol 9 · No 6 · 2004 in CCND1 and CDK4 expression, cell cycle blockage in G1/S (Figure 3B and not shown), but no apoptosis induction (Figure 4B). In contrast to that observed in IGROV1, exposure to rapamycin was associated with a decreased expression of CDK1 in SK-OV-3, OVCAR3 and A2780 preventing cells to undergo mitosis (Figure 2C). To identify factors that mediated resistance to rapamycin-induced apoptosis in SK-OV-3 cells, we investigated the expression of mRNA species coding for apoptosis regulators and compared it to that of rapamycinsensitive IGROV1 cells. At baseline, we observed that SK-OV-3 cells expressed higher levels of mRNA coding for BAD, CD95 and TGFβ than did IGROV1 cells. Conversely, expression of mRNA coding for caspase 9 was lower in SK-OV-3 than in IGROV1 (Figure 1C and data not shown). In our panel, SK-OV-3 was the sole ovarian cancer cell line in our panel that expressed mRNA coding for Bcl-2 (Figure 1C and D). However, since the level of Bcl-2 mRNA species was low in SK-OV3, we confirmed by western blot that, unlike IGROV1, SK-OV-3 expressed the bcl-2 protein. Expression of bcl-xL, another bcl-2 family member, was similar in SK-OV-3 and IGROV1. Based on these data, we speculated that Bcl-2 could prevent apoptosis induced by rapamycin. Biomarkers for mTOR inhibitors in ovarian carcinoma Figure 3. Death of ovarian cancer cells after exposure to mTOR inhibitors and/or Bcl-2/Bcl-xL antisense oligonucleotides. (A) Flow cytometry analysis of IGROV1 cells treated with 10− 5 M rapamycin and RAD001 for 15 h showed increased cell death compared to control (Co). Conversely, no significantly increased cell death was shown in SK-OV-3 cells exposed to mTOR inhibitors. Thus, SK-OV-3 cells were pretreated with the oligonucleotides antisense oligo4625 and oligo4626 for 16 h, incubated with rapamycin for 16 h and analyzed by flow cytometry. (B) Quantification of cell death showed that pretreatment of SK-OV-3 with anti bcl-2 oligo4625 (but not scramble oligo4626) increased cellular death induced by rapamycin and RAD001. Downregulation of Bcl-2 restores sensitivity to mTOR inhibitors To examine the role of bcl-2 in cellular death induced by rapamycin and RAD001, we exposed SK-OV-3 cells to the Bcl-2 anti-sense phosphorothioate olignonucleotide oligo4625 or a scrambled anti-sense oligonucleotide control (oligo4626). Down modulation of bcl2 with oligo4625 consistently increased the apoptotic demise of SK-OV-3 cells in response to rapamycin and RAD001, indicating a synergistic death-inducing effect between mTOR inhibition and Bcl-2 down-regulation. In contrast, treatment with mTOR inhibitors or oligo4625 alone had no apoptotic effects on SK-OV-3 cells at the concentrations used. These results were obtained using two different readouts, namely by measuring the apoptotic shrinkage of cells by cytofluorometry (Figure 3B) or by determining apoptotic phosphatidylserine exposure (Annexin V staining) on the plasma membrane surface (Figure 4A and B), detectable either by cytofluorometry (Figure 4A) or fluorescence microscopy (Figure 4B). Cell death triggered by the combination of oligo4625 with mTOR inhibitors was clearly apoptotic, as indicated by pronounced chromatin condensation (Figure 4B). The Bcl-2/Bcl-xL anti-sense oligonucleotide synergized with rapamycin and RAD001 in terms of induction of apoptosis (Annexin V+ /PI− ) and secondary necrosis (PI+ ) (Figure 4C). Discussion In the nineteen-eighties, screening performed on the 60human-tumor-cell-line panel of the NCI revealed that raApoptosis · Vol 9 · No 6 · 2004 801 D. Aguirre et al. Figure 2. Effects of rapamycin on cellular proliferation and gene expression ovarian cancer cells. Chemical formula of rapamycin (A). Antiproliferative effects corresponding to 50% growth inhibition (GI50 ) of rapamycin in our panel of ovarian cancer cell lines were determined using the MTT assay (B). Semi-quantitative analysis of gene expression in ovarian cancer cells after 48 h exposure to 10− 6 M rapamycin (C). Quantitative RT-PCR analysis of cell cycle genes in IGROV-1 and SK-OV-3 showed concentration-dependent downexpression of several genes involved in cell cycle progression, such as CCND1 and CDK4 (D). Figure 4. Apoptosis induction in SK-OV-3 cells exposed to bcl-2 antisense oligonucleotides and mTOR inhibitors. (A). SK-OV-3 cells were transfected with the oligonucleotides 4625 and 4626 for 16 h and treated with RAPA 10− 7 M for a further 16 h, stained with Annexin V-FITC conjugate and PI and analyzed by flow cytometry. (B) Cells treated as in A were observed under a fluorescence microscope. (C) Cells were treated with the antisense oligonucleotides as in A and incubated for 16 h with 10− 7 and 10− 5 with RAPA and stained with Annexin PI; apoptotic cells (Annexinv+ /PI− ) and dead cells (PI+ ) were determined by flow cytometry. 802 Apoptosis · Vol 9 · No 6 · 2004 Biomarkers for mTOR inhibitors in ovarian carcinoma pamycin and its derivatives exerted an original spectrum of antiproliferative activity, presumably via mechanisms that differed from those of other anticancer drugs.6 Recently, several rapamycin derivatives including CCI-779 and RAD001 have entered clinical trials, either as single agents or in combination with additional chemotherapeutic drugs.6 Surprisingly, while rapamycin derivatives were expected only to slow down tumor progression by interfering with the mTOR-controlled metabolic pathway and protein synthesis, some evidence, including early tumor regression, suggested that mTOR-inhibitors were also capable of triggering cancer cell death. The precise molecular mechanisms by which mTOR inhibitors are capable of inducing cancer cell death are elusive.11,13,14,17–19 Recent reports have indicated that phosphorylation of mTOR by AKT can be obtained through the activation of upstream molecules including PI3K (resulting from mutated or deleted PTEN) and/or through the MAPK pathway. Indeed, rapamycin was found to exert more potent antiproliferative effects in cells characterized by intrinsically phosphorylated (activated) AKT and loss of expression of PTEN (see 6 for review). More recently, mTOR was also discovered to serve as an important apoptosis modulator in human cells.13,14 However, the downstream targets connecting mTOR inhibition to cellular demise remain poorly understood. Two recent studies have suggested that members of the Bcl-2 family could be downstream mediators of IGF-I-stimulated PI-3 kinase-dependent survival of cells.14,17,19 Furthermore, several growth factors that activate the PI-3 kinase and S6K pathways were recently shown to increase expression of Bcl-2, thereby promoting cell survival in myeloid progenitor cells.19 The role of Bcl-2 in resistance to anticancer agents has been investigated extensively in in vitro systems, animal models and clinical studies. In studies with patients suffering from ovarian carcinoma, high levels and/or aberrant patterns of Bcl-2 expression have been correlated with resistance to commonly used anticancer agents.20 In this study, we found, within a panel of 4 ovarian cancer cell lines, that the IGROV1 cell line derived from a polymorphous endometrioid tumor was most sensitive to rapamycin. Attempts to identify molecular markers that predict sensitivity to rapamycin were based on previous studies performed in Saccharomyces cereviciae.15,16 We first evaluated the human homologues of yeast genes were associated with rapamycin sensitivity in yeast. However, baseline expressions of mRNA species coding for those selected signaling molecules involved in the mTOR pathway and cell cycle control did not determine the rapamycin response in ovarian carcinoma cells. We observed that under exposure to rapamycin, IGROV1 displayed downregulation of mRNA coding for Cyclin D1 and CDK4, an alteration that could lead to cell cycle arrest in the late G1 phase. Therefore, downregulation of Cyclin D1 was considered as a potential biomarker to evaluate the cell cycle effects of rapamycin and rapamycin derivatives at the cellular level in clinical trials. However, downregulation of Cyclin D1 was observed in both sensitive and resistant cells and was not predictive of rapamycin-induced antiproliferative effects. Therefore, other mechanisms downstream from Cyclin D1/CDK4 downregulation were suspected to play a role in cytotoxicity induced by rapamycin. CDK1/cyclin B kinase acts as a universal M-phase promoting factor and might also play a role in cellular apoptosis.13 Interestingly, exposure to rapamycin was associated to an increased CDK1 expression in IGROV1 contrasting with CDK1 downexpression in other less sensitive ovarian cancer cell lines. This suggested that ovarian cancer cells enable to maintain or increase CDK1 expression under exposure to rapamycin would have reduced cell proliferation. Intriguingly, CDK1 overexpression was associated with FAS overexpression in IGROV1. Although specific molecular interactions linking mTOR, Cyclin B1/cdk1, and Fas ligand in cancer cell are not fully determined, authors have recently found that activation-induced Fas ligand expression appears to be mediated by the Cyclin B1/cdk1 complex.21 Taken together; our data suggest that Cyclin D1/CDK4 downregulation together with CDK1 overexpression may represent a molecular pattern associated with sensitivity to rapamycin. Subsequently, we focused on two cell lines representative of the highest and lowest resistance to rapamycin cell lines in our panel. We found that exposure of IGROV1 cells to rapamycin and RAD001 was associated with substantial cellular apoptosis. Conversely, SK-OV-3 cells were only marginally sensitive to rapamycin and appeared incapable of undergoing apoptosis. In both SK-OV-3 and IGROV1, exposure to rapamycin and RAD001 was associated with a decrease in the expression of Cyclin D1 and CDK4, as well as G1 arrest. This confirmed that decreased expression of Cyclin D1 resulting from pharmacological mTOR inhibition did not necessarily indicate apoptotic effects in our cancer cells. Thus, we sought parameters other than cell cycle gene modulation that might play a role in sensitivity/resistance to rapamycin. In this study, we found that mRNA species of genes involved in apoptosis were expressed at different levels in our ovarian cancer cell lines. However, Bcl-2 was the sole gene to be expressed (although at a relatively low level) in SK-OV-3 cells and not in other cancer cell lines. Conversely, expression of bcl-xL was found in both SK-OV-3 and IGROV1. To determine whether Bcl-2 could play a role in preventing SK-OV-3 cells from undergoing apoptosis, we used anti-Bcl-2 oligonucleotides designed to interact with expressed Bcl-2 mRNA species. We found that sensitivity to rapamycin was restored by prior exposure to oligo-4625, which targets bcl-2/bcl-xL mRNA. This was associated with an increase in SK-OV-3 cells undergoing apoptosis. As expected, the use of anti-Bcl-2 oligonucleotides had Apoptosis · Vol 9 · No 6 · 2004 803 D. Aguirre et al. no effect in IGROV1. Therefore, our data suggest that bcl-2 plays a pivotal role in preventing apoptosis induction by rapamycin and derivatives in some ovarian cancers. Recent data also showed that in mice bearing transgenes encoding both AKT and Bcl-2, prostate intraepithelial neoplastic cells remain sensitive to RAD001-induced inhibition of proliferation but are resistant to apoptosis.22 In those transgenic mice as well as in our human ovarian cancer cells, apoptosis induction is mediated through caspase 3 suggesting a mechanism that requires an intact mitochondrial apoptotic pathway. Taken together, those data suggest that mTOR inhibition may be less effective in human cancers characterized by Bcl-2 overexpression, a situation eventually associated with the loss of PTEN in cancer cells. In that case, apoptosis induction by rapamycin derivatives could be restored using combinations with Bcl-2 inhibitors. Those results will be used in future clinical trials to better characterize tumor types likely to respond to mTOR inhibitors.23 Conclusion In summary, our data indicate that Cyclin D1/CDK4 and Bcl-2 may serve as surrogate markers to monitor sensitivity and to detect resistance to rapamycin and its derivative RAD001. Attempts will be made to use these surrogate molecular markers to monitor the biological activity doses of rapamycin derivatives in clinical trials, for instance by performing RT-PCRs on cancer cells isolated from blood, from ascites, or from solid tumors. Moreover, our data provide a rationale for exploring the combination of mTORand Bcl-2-targeted drugs in the treatment of ovarian carcinoma. Acknowledgments We thank Isabelle Pouillon, Nelly Motté and Jean Louis Bobot for their excellent technical assistance with cell cultures and quality control of mRNA analysis, and Heidi Lane, Novartis Institutes for BioMedical Research, for her critical reading of the manuscript. This work has been supported by a special grant from the Ligue contre le Cancer (to G.K.), as well as by grants from the European Commission (QLK3-CT-2002-01956 to G.K. and MCFI2000-00943 to P.B.). References 1. Mills GB, Lu Y, Fang X, et al. The role of genetic abnormalities of PTEN and the phospatidylinositol 3-kinase pathway in breast and ovarian tumorigenesis, prognosis, and therapy. Seminars in Oncol 2001; 28(suppl 16): 125–141. 804 Apoptosis · Vol 9 · No 6 · 2004 2. Hu L, Zaloudek C, Mills G, Gray J, Jaffe RB. In vivo and in vitro ovarian carcinoma growth inhibition by a phosphatidylinositol 3-kinase inhibitor (LY294002). Clin Cancer Res 2000; 6(3): 880–886. 3. Castedo M, Ferri KF, Kroemer G. Mammalian target of rapamycin (mTOR). Pro-and anti-apoptotic. Cell Death Differ 2002; 9: 99–100. 4. Shi Y, Frankel A, Radvanyi LG, Penn LZ, Miller RG, Mills GB. Rapamycin enhances apoptosis and increase sensitivity to cisplatin in vitro. Cancer Res 1995; 55(9): 1982– 1988. 5. Hu L, Hofmann J, Lu Y, Mills GB, Jaffe RB. Inhibition of phosphatidylinositol 3′ -kinase increases efficacy of paclitaxel in in vitro and in vivo ovarian cancer models. Cancer Res 2002; 62: 1087–1092. 6. Hidalgo M, Rowinsky EK. The rapamycin-sensitive signal transduction pathway as a target for cancer therapy. Oncogene 2000; 19(56): 6680–6686. 7. Majewski M, Korecka M, Kossev P, et al. The immunosuppressive macrolide RAD inhibits growth of human Epstein-barr virus-transformed B lymphocytes in vitro and in vivo: Potential approach to prevention and treatment of posttransplant lymphoproliferative disorders. Proc Natl Acad Sci USA 2000; 97(8): 4285–4290. 8. Neshat MS, Mellinghoff IK, Tran C, et al. Enhanced sensitivity of PTEN-deficient tumors to inhibition of FRAP/mTOR. Proc Natl Acad Sci USA 2001; 98(18): 10314–10319. 9. Podsypanina K, Lee RT, Politis C, et al. An inhibitor of mTOR reduces neoplasia and normalizes p70/S6 kinase activity in Pten+/− mice. Proc Natl Acad Sci USA 2001; 98(18): 10320– 10325. 10. Dudkin L, Dilling MB, Cheshire PJ, et al. Biochemical correlates of mTOR inhibition by the rapamycin ester CCI-779 and tumor growth inhibition. Clin Cancer Res 2001; 7(6): 1758– 1764. 11. Huang S, Liu LN, Hosoi H, Dilling MB, Shikata T, Houghton PJ. p53/p21(CIP1) cooperate in enforcing rapamycin-induced G(1) arrest and determine the cellular response to rapamycin. Cancer Res 2001; 61(8): 3373–3381. 12. Hosoi H, Dilling MB, Liu LN, et al. Studies on the mechanism of resistance to rapamycin in human cancer cells. Mol Pharmacology 1998; 54: 815–824. 13. Castedo T, Roumier J, Blanco KF, et al. Sequential involvement of Cdk1, mTOR and p53 in apoptosis induced by the HIV-1 envelope. EMBO J 2002; 21: 4070–4080. 14. Decaudin D, Hirsch T, Geley M, et al. Bcl-2 and Bcl-XL antagonize the mitochondrial dysfunction preceding nuclear apoptosis induced by chemotherapeutic agents. Cancer Res 1997; 57: 62–67. 15. Chan T-F, Carvalho J, Riles L, Zheng XSF. A chemical genomics approach toward understanding the global functions of the target of rapamycin protein (TOR). Proc Natl Acad Sci USA 2000; 97(24): 13227–13232. 16. Hardwick JS, Kuruvilla FG, Tong JK, Shamji AF, Schreiber SL. Rapamycin-modulated transcription defines the subset of nutrient-sensitive signaling pathways directly controlled by the Tor proteins. Proc Natl Acad Sci USA 1999; 96(26): 14866–14870 (http://www.pnas.org/cgi/content/full/96/26/ 14866/DC1). 17. Zangemeister-Wittke U, Leech SH, Olie RA, et al. A novel bispecific antisense oligonucleotide inhibiting both bcl-2 and bcl-xL expression efficiently induces apoptosis in tumor cells. Clin Cancer Res 2000; 6: 2547–2555. 18. Gautschi O, Tschopp S, Olie RA, et al. Activity of a novel bcl2/bcl-xL-bispecific antisense oligonucleotide against tumors of Biomarkers for mTOR inhibitors in ovarian carcinoma diverse histologic origins. J Natl Cancer Inst 2001; 93: 463– 471. 19. Shinjyo T, Kuribara R, Inukai T, et al. Downregulation of bim, a proapoptotic relative of Bcl-2, is a pivotal step in cytokineinitiated survival signaling in murine hematopoietic progenitors. Mol Cell Biol 2001; 21(3): 854–864. 20. Shinoura N, Yoshida Y, Nishimura M, et al. Expression level of bcl-2 determines anti- or proapoptotic function. Cancer Res 1999; 59: 4119–4128. 21. Torgler R, Jakob S, Ontsouka E, et al. Regulation of activationinduced Fas (CD95/Apo-1) ligand expression in T cells by the Cyclin B1/cdk1 complex. J Biol Chem 2004; Epub ahead of print June 23. 22. Majumder PK, Febbo PG, Bikoff R, et al. mTOR inhibition reverses Akt-dependent prostate intraepithelial neoplasia through regulation of apoptotic and HIF-1dependent pathways. Nat Medicine 2004; 10(6): 594– 601. 23. Raymond E, Alexandre J, Faivre S, et al. Safety and Pharmacokinetics of Escalated Doses of Weekly Intravenous Infusion of CCI-779, a Novel mTOR Inhibitor, in Patients With Cancer. J Clin Oncol 2004; 22(12): 2336–2347. Apoptosis · Vol 9 · No 6 · 2004 805