Synergistic Activity of Fenretinide and the Bcl-2 Family Protein Inhibitor ABT-737 against Human Neuroblastoma

Neuroblastoma is an aggressive childhood tumor of the sympathetic nervous system accounting for 8% to 10% of all childhood cancers and approximately 15% of cancer deaths in children (1). Treatment of high-risk neuroblastoma (stage 4 patients above 1 year at diagnosis and stage 3 disease with MYCN amplification or unfavorable histopathology) with multiagent chemotherapy, radiotherapy, and myeloablative chemotherapy supported stem cell transplant followed by treatment of minimal residual disease by 13-cis-retinoic acid (13-cis-RA; refs. 2, 3), and more recently, anti-GD2 antibody, cytokines, and 13-cis-RA have significantly improved the outcome for these patients (4, 5). However, as many high-risk patients still ultimately die from tumor that is refractory to initial therapy or recurrent, resistant disease, novel therapies effective against MDR neuroblastoma are needed.

4-HPR is a synthetic derivative of retinoic acid that has a broad spectrum of cytotoxic activity against primary tumor cells, cell lines, and/or xenografts of various cancers (including neuroblastoma; refs. 6–9) and has been tested in early phase clinical trials in recurrent neuroblastoma (10–12). In contrast to retinoids, such as all-trans-retinoic acid (ATRA) and 13-cis-RA, that cause arrest of cell growth and morphologic differentiation of human neuroblastoma cell lines, 4-HPR induces cell death in neuroblastoma cells via both apoptotic and nonapoptotic mechanisms (6–8). The observation that 4-HPR remains cytotoxic in neuroblastoma cell lines that are resistant to ATRA, 13-cis-RA, alkylating agents, and etoposide suggests that it may be active in high-risk neuroblastoma patients who are resistant to standard therapy (7, 9, 13). Indeed, a recent phase I trial in patients with refractory/recurrent neuroblastoma of a novel oral powder 4-HPR formulation (4-HPR LXS oral powder) that improves 4-HPR exposures documented 4 complete responses (11).

Overexpression of antiapoptotic Bcl-2 family proteins is a common mechanism by which cells become resistant to conventional chemotherapy, providing attractive therapeutic targets (14, 15). Neuroblastoma cell lines obtained from patients with recurrent disease after treatment exhibit increased levels of Bcl-2 expression, which may be responsible for their drug resistance (16). Strong Bcl-2 immunoreactivity was also detected in islets of residual neuroblastoma cells in MYCN nonamplified primary tumors of treated patients (17). These observations suggest that an increased expression of antiapoptotic Bcl-2 family of proteins may be one mechanism for the resistance of neuroblastoma cells to cytotoxic agents including 4-HPR.

ABT-737 is a small molecule that mimics the direct binding of the BH3-only protein Bad to antiapoptotic Bcl-2 family proteins such as Bcl-2, Bcl-X_L, and Bcl-w and therefore displaces BH3-only death-activating proteins (Bim, Bid; ref. 14). By binding to proteins of the Bcl-2 family, ABT-737 prevents proteins, such as Bid and Bim (direct activators of Bax and Bak), from forming heterodimers with antiapoptotic Bcl-2 family proteins, thereby promoting Bax and Bak activation via oligomerization (18). ABT-737 has been reported to be cytotoxic as a single agent and to sensitize a wide variety of cancer cells to several chemotherapeutic agents in vitro and in vivo (14, 19–23). We have previously shown that ABT-737 synergistically enhances 4-HPR cytotoxicity in acute lymphoblastic leukemia cells but that 4-HPR + ABT-737 was not toxic for normal nonproliferating lymphocytes (24).

We therefore hypothesized that the combination of an inhibitor of proteins of the Bcl-2 family (such as ABT-737) with 4-HPR would be synergistic against neuroblastoma with minimal systemic toxicities. To test this hypothesis, we evaluated the activity of ABT-737 and 4-HPR (alone and in combination) on a panel of human neuroblastoma cell lines and in a xenograft model of recurrent MDR human neuroblastoma.

We used a panel of 11 human neuroblastoma cell lines obtained from patients at various stages of disease: 2 cell lines established at diagnosis prior to any therapy (CHLA-15 and SMS-KAN), 7 cell lines obtained at the time of progressive disease during induction therapy [SK-N-BE(2), SK-N-RA, CHLA-119, LA-N-6, CHLA-20, SMS-KCNR, and CHLA-140], and 2 cell lines established at relapse after myeloablative therapy and bone marrow transplantation (CHLA-79 and CHLA-136). All cell lines were established in the senior author's laboratory, except SK-N-RA and SK-N-BE(2) which were a gift of Dr. L. Helson; characterization of these neuroblastoma cell lines has been previously reported (25, 26). Cell line identity was confirmed at time of the experiments using a 15-loci short tandem repeat (STR) assay + amelin for sex determination (27), with the genetic signature compared with the Children's Oncology Group STR database (http://www.COGcell.org). We also tested the human normal fibroblast cell line CRL-2076 obtained from the American Type Culture Collection.

SMS-KAN, SK-N-BE(2), SK-N-RA, LA-N-6, SMS-KCNR, and CRL-2076 fibroblasts were cultured in RPMI-1640 medium (Irvine Scientific) supplemented with 10% heat-inactivated FBS (Gemini Bio-Products, Inc.). CHLA-15, CHLA-119, CHLA-20, CHLA-140, CHLA-79, and CHLA-136 were cultured in Iscove's Modified Dulbecco's Medium (BioWhittaker) containing 20% heat-inactivated FBS and supplemented with 3 mmol/L l-glutamine (Gemini Bio-Products, Inc.), insulin, and transferrin (5 μg/mL each) and selenium (5 ng/mL; ITS Culture Supplement, Collaborative Biomedical Products). All cell lines were continuously cultured at 37°C in a humidified incubator containing 95% air + 5% CO₂ without antibiotics. Experiments were carried out using neuroblastoma cell lines at passage 15 to 35. Cells were detached from culture plates or flasks with the use of a modified Puck's Solution A plus EDTA (Puck's EDTA), containing 140 mmol/L NaCl, 5 mmol/L KCl, 5.5 mmol/L glucose, 4 mmol/L NaHCO₃, 0.8 mmol/L EDTA, 13 μmol/L phenol red, and 9 mmol/L HEPES buffer (pH = 7.3; ref. 28).

ABT-737 was kindly provided by Abbott Laboratories. 4-HPR was obtained from the Developmental Therapeutics Program of the National Cancer Institute (Bethesda, MD). Fenretinide was formulated as LYM-X-SORB oral powder (3% 4-HPR by weight, 4-HPR LXS) by Avanti Polar Lipids, Inc. (29) and was kindly provided to the investigators by Barry J. Maurer, MD, PhD. Eosin Y was purchased from Sigma Chemical Co., and fluorescein diacetate (FDA) was obtained from Eastman Kodak Co. Mitochondrial membrane potential probe JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolyl-carbocyanine iodide) was obtained from Molecular Probes; the Terminal Deoxynucleotidyl Transferase–Mediated dUTP Nick End Labeling (TUNEL) Kit and the caspase-8 enzyme inhibitor Z-IETD-FMK were obtained from BD Biosciences (APO-DIRECT). The pan-caspase enzyme inhibitor Boc-d-fmk was purchased from MP Biomedicals, LLC. Stock solutions of ABT-737 (5 mmol/L), 4-HPR (10 mmol/L), FDA (1 mg/mL), JC-1 (2 mg/mL), Z-IETD-FMK (10 mmol/L), and Boc-d-fmk (20 mmol/L) were dissolved in dimethyl sulfoxide, except 4-HPR which was dissolved in 95% ethanol. All reagents were stored at −20°C.

The cytotoxicity of ABT-737, 4-HPR, and their combination (at a 1:1 molar ratio) was determined in 96-well plates using the semiautomatic fluorescence-based Digital Imaging Microscopy System (DIMSCAN; refs. 30, 31). DIMSCAN uses digital imaging microscopy to quantify viable cells, which selectively accumulate FDA. DIMSCAN is capable of measuring cytotoxicity over a 4 log dynamic range by quantifying total fluorescence per well (which is proportional to the number of viable cells) after elimination of the background fluorescence by digital thresholding and eosin Y quenching. Cells were seeded into 96-well plates in 100 μL of complete medium at 5,000 to 10,000 cells per well, depending on cell line growth rate. After overnight incubation, ABT-737, 4-HPR, or their combination was added to each well at various concentrations in 50 μL of culture medium. We used the following drug concentrations: 0, 1.25, 2.5, 5, 10 μmol/L or 0, 2.5, 5, 7.5, 10 μmol/L in replicates of 12 wells for each experimental condition. After incubation with the drugs for 96 hours at 37°C, FDA (final concentration: 10 μg/mL) and eosin Y [final concentration: 0.1% (w/v)] were added to each well and the cells were incubated for an additional 20 minutes at 37°C. Total fluorescence per well was then measured using DIMSCAN, and the results were expressed as the ratio of the fluorescence in treated wells to the fluorescence in untreated wells (survival fraction).

Apoptosis in cells was examined by flow cytometry using a commercial TUNEL kit (APO-DIRECT; BD Biosciences) according to the manufacturer's instructions. Cells treated with ABT-737, 4-HPR, or their combination in the presence or absence of the pan-caspase inhibitor Boc-d-fmk (40 μmol/L for 1 hour) were collected, washed, centrifuged, and resuspended in 1% (w/v) paraformaldehyde in PBS (pH 7.4). They were then kept on ice for 60 minutes, washed in PBS, centrifuged, and fixed in 70% (v/v) ice-cold ethanol at −20°C for 12 to 18 hours, before being stained with terminal deoxynucleotidyltransferase (TdT) and fluorescein isothiocyanate (FITC)-labeled dUTP (FITC-dUTP) for 2 hours at 37°C. After washing with PBS, cells were resuspended in 0.5 mL of propidium iodide (PI) and RNase containing buffer (5 μg/mL PI, 200 μg/mL RNase). Cells were then incubated in the dark for 30 minutes at room temperature prior to analysis by flow cytometry. The percentage of TdT-mediated fluorescent cells was measured by flow cytometry using band pass filters of 525 ± 25 nm for FITC and 610 ± 25 nm for PI (7, 32) in a BD LSR II system (BD Biosciences) equipped with the DiVA software (version 4.1.2; BD Biosciences).

Cells were treated with ABT-737, 4-HPR, or the combination, collected in 5 mL polystyrene tubes, centrifuged at 300 × g for 5 minutes, and resuspended in 1 mL of medium containing 10 μg/mL of JC-1, incubated at 37°C for 10 minutes, and analyzed by flow cytometry. The detection of a fluorescence emission shift from red (610 ± 10 nm) to green (525 ± 10 nm) was an indication of mitochondrial membrane depolarization (33).

Cells were lysed in radioimmunoprecipitation (RIPA) lysis buffer (Upstate Biotechnology), containing 15 μL/mL of phenylmethylsulfonylfluoride and 40 μL/mL of Protease Inhibitor Cocktail (Sigma). The lysates were left on ice for 15 minutes, briefly sonicated, and centrifuged at 12,000 × g for 15 minutes. Protein concentration in the supernatants was determined using the BCA Protein Assay Kit (Pierce Biotechnology), and 20 μg of protein in each sample was resolved by electrophoresis in a 10% to 20% gradient acrylamide gel containing 0.1% SDS (Invitrogen). After electrophoresis, the gels were transferred to a protein nitrocellulose transfer membrane (Whatman GmbH). The membrane was hybridized with primary antibodies followed by horseradish peroxidase (HRP)-conjugated secondary antibodies and immunocomplexes detected by chemiluminescence (Pierce Biotechnology) and visualized on autoradiography film (Denville Scientific, Inc.). Quantification was obtained by scanning the immunoblots using an Epson Expression 1680 system (Epson). Antibodies used were as follows: anti-Bax rabbit polyclonal antibody (#554104) from BD Biosciences; anti-β-actin goat polyclonal antibody, anti-GAPDH mouse monoclonal antibody, and HRP-conjugated secondary anti-mouse, anti-goat, and anti-rabbit antibodies from Santa Cruz Biotechnology; anti-caspase-9 rabbit polyclonal antibody (#9502), anti-caspase-3 rabbit polyclonal antibody (#9662), anti-Bid rabbit polyclonal antibody (#2002), anti-Bcl-2 rabbit polyclonal antibody (#2870), anti-Bcl-XL rabbit polyclonal antibody (#2764), anti-Bcl-w rabbit polyclonal antibody (#2724), anti-Mcl-1 rabbit polyclonal antibody (#5453), and anti-Bak rabbit polyclonal antibody (#3814) from Cell Signaling Technology; anti-Bak mouse monoclonal antibody (Ab-1) which recognizes only conformationally active Bak was from Calbiochem; and anti-cytochrome c rabbit polyclonal antibody and anti-OxPhos Complex IV (COX IV) mouse monoclonal antibody were from Clontech. Densitometric analysis was conducted using ImageJ digital imaging software (NIH, Bethesda, MD).

After treatment with ABT-737, 4-HPR, or the combination for 6 or 24 hours, cells were subjected to a digitonin-based subcellular fractionation (34) to separate the cytosol (supernatant) from intact mitochondria (pellet). The pellets were then lysed in RIPA lysis buffer and Western blot analysis conducted as described above.

A caspase-8/FLICE colorimetic assay (Invitrogen) was used according to the manufacturer's instructions to detect the activation of caspase-8.

Validated siRNAs specific for Bax (5′-GCUCUGAGCAGAUCAUGAATT-3′) and Bak (5′-GCGAAGUCUUUGCCUUCUCTT-3′) were purchased from Qiagen. A nonspecific nonsilencing siRNA (AllStars Negative Control siRNA; QIAGEN) was used as negative control. Neuroblastoma cells were transfected with 100 nmol/L of Bax or Bak siRNA or control siRNA using Lipofectamine iMax transfection reagent (Invitrogen) according to the manufacturer's instructions.

The CHLA-119 cell line, established from a patient with neuroblastoma at time of disease progression during chemotherapy (26), was injected at 17 × 10⁶ cells subcutaneously between the shoulder blades of 4- to 6-week-old female athymic (nu/nu) mice. Once palpable and progressing tumors of 100 to 200 mm³ had developed, mice were treated with a maximum tolerated dose of 4-HPR LXS oral powder (240 mg/kg/d) and ABT-737 (50 mg/kg/d) for 5 days per week, alone or in combination or with the vehicle as control. When given alone, the dose of ABT-737 was increased to 100 mg/kg/d. 4-HPR LXS was prepared as slurry in sterile water and given orally by gavage, whereas ABT-737 was administered by intraperitoneal injection. Tumor volume was determined from measurements taken twice weekly using the formula 0.5 × height × width × length (35). When tumor volumes reached 1,500 mm³, the mice were sacrificed

For in vitro experiments, synergistic drug interactions were determined by fixed ratio dose–response assays of the drugs alone and in combination (4-HPR:ABT-737 = 1:1). The IC₉₀ (the drug concentration that is cytotoxic or growth inhibitory for 90% of a cell population) and the combination index (CI) values were obtained by DIMSCAN analysis and calculated using Calcusyn software (Biosoft). Calculation of a CI is a method to numerically quantify drug synergism based on the multiple drug–effect equation of Chou–Talalay derived from enzyme kinetic models (36, 37). With this method, a CI lower than 0.9 indicates synergism; a CI of 0.9 to 1.10 indicates additive activity; and a CI greater than 1.10 indicates antagonism.

In vivo data were analyzed using the software GraphPad Prism (GraphPad Software, Inc., version 4.03). Event-free survival (EFS) curves were compared by Kaplan–Meier log-rank test and quantified as the time taken from the initiation of treatment until an event, with events defined as tumor volume reaching 1,500 mm³ or when mice died or were sacrificed because of treatment-related toxicity. An EFS T/C value was defined as the ratio of the median time to event of the treatment group and the median time to event of the respective control group (38). For the EFS T/C measure, agents are considered highly active if they met 3 criteria: (a) an EFS T/C greater than 2; (b) a significant difference (P ≤ 0.05) in EFS distribution between treated and control (or single agent); and (c) a net reduction in median tumor volume for animals in the treated group at the end of treatment as compared with treatment initiation. Agents meeting the first 2 criteria, but not having a net reduction in median tumor volume for treated animals at the end of the study, are considered to have intermediate activity. Agents with an EFS T/C less than 2 are considered to have low levels of activity.

Relative tumor volumes (RTV) for control (C) and treatment (T) mice (tumor volume T/C values) were calculated on day 21 (38). The mean RTVs for control and treatment mice for each study were then calculated, and the T/C value was the mean RTV for the treatment group divided by the mean RTV for the control group. For the tumor volume T/C response measure, agents producing a T/C of ≤15% are considered highly active, those with a mean tumor volume T/C of ≤45% but >15% are considered to have intermediate activity, and those with mean T/C values >45% are considered to have low levels of activity.

Results with a value of P < 0.05 were regarded as significant (log-rank for EFS by GraphPad Prism or unpaired 2-sided Student t test by Microsoft Excel 2000 for other data).

We first determined the cytotoxicity of ABT-737, 4-HPR, and ABT-737 plus 4-HPR (0–10 μmol/L for each drug) at 1:1 molar ratio in 11 neuroblastoma cell lines using the DIMSCAN cytotoxicity assay. Representative drug cytotoxicity dose–response curves are shown in Fig. 1 and CI values calculated at the fixed ratio drug concentrations tested for the cytotoxicity assays and IC₉₀ values of both agents are shown in Table 1. 4-HPR + ABT-737 showed synergistic activity in all 11 cell lines (CI < 0.9). In 8 of the 11 cell lines, including those resistant to 4-HPR alone (CHLA-136), ABT-737 alone [SK-N-BE(2) and CHLA-119], or both (SK-N-RA and CHLA-20), 4-HPR + ABT-737 showed very strong to good synergistic activity. Only in 3 cell lines (LA-N-6, CHLA-79, and CHLA-140) did 4-HPR + ABT-737 show only moderate or slight synergistic activity. There was no apparent association between sensitivity to single agents and synergy of the drugs in combination. ABT-737 and 4-HPR as single agents or in combination were minimally toxic for CRL-2076 fibroblasts at concentrations up to 5 μmol/L and the combination showed only modest toxicity at 10 μmol/L for each drug (Fig. 1D).

The basal levels of proapoptotic and antiapoptotic proteins in 10 of these cell lines was examined by Western blot analysis (Fig. 2). There was no apparent association between protein basal levels of Bcl-2, Bcl-XL, Bcl-w, or Mcl-1 and response to 4-HPR, ABT-737, or the 2 drugs in combination (Fig. 1).

Because apoptosis is a major mechanism of action for both 4-HPR and ABT-737, we examined the effect of single agents and their combination on apoptosis. In all 4 cell lines tested (CHLA-119, CHLA-15, SMS-KAN, and CHLA-136), we observed a higher level of apoptosis when the 2 drugs were combined in comparison with each drug alone (Fig. 3A). For example, in CHLA-119 cells, the percentage of apoptotic cells at 24 hours was 0.6% ± 0.4% (control), 19.2% ± 2.9% in the presence of 2.5 μmol/L ABT-737, 26.9% ± 0.9% in the presence of 2.5 μmol/L 4-HPR, and 82.1% ± 6.0% in the presence of both drugs (P < 0.005 and P < 0.001 relative to ABT-737 and 4-HPR as single agents, respectively). In contrast, 4-HPR did not cause a significant increase of apoptosis in the other cell lines (CHLA-15, SMS-KAN, and CHLA-136), but the combination of 4-HPR and ABT-737 significantly increased apoptosis compared with ABT-737 alone. ABT-737 and 4-HPR (either alone or in combination) did not induce apoptosis in CRL-2076 normal fibroblasts (P > 0.05; Supplementary Fig. S1).

We next investigated the effect of 4-HPR and ABT-737 on the loss of mitochondrial membrane potential (Δψ_m) in CHLA-119, CHLA-15, and SMS-KAN cells. Treatment with ABT-737 and 4-HPR resulted in a greater than additive loss of Δψ_m when compared with the sum of the losses observed with each single agent (Fig. 3B). ABT-737 or 4-HPR alone, and especially the combination, caused release of cytochrome c from the mitochondria into the cytosol, and a decrease of cytochrome c in the mitochondrial fraction in CHLA-119 cells (Fig. 3C).

Caspase-9 is a key intermediate in the mitochondrial or intrinsic apoptotic pathway, and caspase-3 is the main final “effector caspase” inducing cell death. We therefore examined the effects of ABT-737, 4-HPR, and their combination on the expression and activation of these 2 caspases (Fig. 3D). The data indicated that both caspase-9 and caspase-3 were cleaved and activated upon treatment with ABT-737 alone or in combination with 4-HPR. To determine whether the effect of 4-HPR and ABT-737 solely involved activation of the intrinsic pathway, we also examine their effect on the activation of caspase-8, the major caspase involved in the extrinsic apoptosis (Fig. 3E). Caspase-8 activity was significantly higher in CHLA-119 and CHLA-15 cells treated with the 4-HPR + ABT-737 than in untreated cells or in cells treated with either single agent. These data suggest that ABT-737 plus 4-HPR induced both extrinsic and intrinsic apoptotic pathways.

To confirm that apoptosis induced by 4-HPR, ABT-737, or the combination was caspase dependent, we pretreated the cells with a pan-caspase inhibitor Boc-d-fmk (40 μmol/L) 1 hour before drug exposure and examined cells for apoptosis by TUNEL assay (Fig. 3F). In all cases (single drug and combination), Boc-d-fmk returned the levels of apoptosis to levels observed in the absence of any drug (control), indicating that that the apoptotic effects of ABT-737, 4-HPR, and the combination were caspase dependent.

To further explore the mechanisms of apoptosis induced by ABT-737 and 4-HPR, we determined the effects of ABT-737, 4-HPR, and the combination on the expression of proapoptotic proteins Bid and Bax by Western blot analysis (Fig. 4A). The activation of Bid was shown by the presence of the 15 kDa truncated form of Bid (t-Bid), and the activation of Bax by the presence of Bax-α (21 kDa), the most active form of Bax that plays a key role in cytochrome c release (14). In CHLA-136 cells (sensitive to ABT-737 and resistant to 4-HPR), ABT-737 activated Bid as indicated by the presence of t-Bid and Bax-α, and a greater effect was seen with the combination whereas 4-HPR alone had little effect. In CHLA-119 cells, which are relatively sensitive to 4-HPR but resistant to ABT-737, treatment with 4-HPR induced the truncation of Bid in a time-dependent manner, and a maximum activation was observed in the presence of 4-HPR + ABT-737. Bax activation occurred earlier in CHLA-136 than CHLA-119, perhaps due to differences in their relative sensitivity to ABT-737 as a single agent (Fig. 1).

The cleavage of Bid into t-Bid further supported the contribution of the extrinsic apoptotic pathway to the apoptosis induced by 4-HPR and ABT-737 because Bid is a substrate for caspase-8 (39). To confirm this possibility, we tested whether the activation of Bid into t-Bid in cells treated with ABT-737 and 4-HPR could be prevented by the caspase-8 inhibitor Z-VAD-FMK (20 μmol/L, 1 hour preexposure). As shown in Fig. 4B, Z-VAD-FMK markedly decreased Bid cleavage into t-Bid in CHLA-119 cells treated with 4-HPR and ABT-737.

In addition to Bax and Bid activation upon treatment with 4-HPR and ABT-737 (Fig. 4A), activation of another proapoptotic Bcl-2 family protein, Bak, was detected by Western blot using an antibody which specifically recognizes activated Bak (Fig. 4C). To determine whether the activation of Bax and Bak is involved in the apoptosis induced by ABT-737 + 4-HPR, we knocked down Bax or Bak by siRNA and examined apoptosis by TUNEL assay induced by ABT-37 + 4-HPR compared with control siRNA. Knockdown of Bax (Fig. 4D) or Bak (Fig. 4E) significantly decreased apoptosis induced in CHLA-119 cells by ABT-737 + 4-HPR.

To determine whether ABT-737 and 4-HPR would have a synergistic activity in vivo, we selected CHLA-119, a cell line derived from a patient whose disease progressed during intensive multiagent chemotherapy (26) and that is sensitive to 4-HPR but resistant to ABT-737 and for which strong synergism was observed in vitro (Table 1).

The tumor growth rate over time was recorded (Fig. 5A), and tumor volume T/C on day 21 was calculated to evaluate the capability of inhibiting tumor growth (Supplementary Table S1). We observed that ABT-737 only showed a low activity (tumor volume T/C = 66.8%), but both 4-HPR as a single agent and 4-HPR + ABT-737 showed intermediate activity with tumor volume T/C = 45.9% and 30.9% respectively, which was consistent with CHLA-119 cells being resistant to ABT-737 but sensitive to 4-HPR alone and the combination in vitro. Median EFS and EFS T/C values were also calculated to evaluate prolonged survival by the log-rank analysis (Fig. 5B and Supplementary Table S1), and the results were consistent with tumor growth rate analysis. Only 4-HPR (EFS T/C = 1.7) and 4-HPR + ABT-737 (EFS T/C = 3.3) statistically increased EFS (P = 0.0002) relative to controls, whereas ABT-737 had no significant effect (EFS T/C = 1.1, P > 0.05). ABT-737 + 4-HPR also increased EFS relative to either single drug alone (P < 0.001).

The Bcl-2 family of proteins provides one mechanism by which malignant cells can survive various cytotoxic drugs and inhibition of Bcl-2 family antiapoptotic proteins by drugs, such as ABT-737, is a promising approach for the treatment of cancer. ABT-737 is a small-molecule inhibitor of Bcl-2, Bcl-X_L, and Bcl-w that has been reported to be active against lymphoid cancers (19–23) and to have minimal systemic toxicity in animal preclinical models (18). A related orally bioavailable compound, ABT-263, is currently undergoing early phase adult clinical trials (40, 41). We have recently shown that ABT-737 synergizes 4-HPR in lymphoid malignancies in vitro (24). 4-HPR is a well-tolerated drug that has shown activity against recurrent neuroblastoma in the laboratory (6, 7) and in patients (10, 11). Here, we showed that ABT-737 is active against most neuroblastoma cell lines in vitro and has a synergistic antitumor effect when combined with 4-HPR at a concentration range that is active and tolerable in mouse xenograft models (18). Concentrations of 4-HPR used in this study were within the range achieved in children in clinical trials (10) with little hematopoietic toxicity observed.

In most of the neuroblastoma cell lines and drug concentrations tested in this study, we found that 4-HPR alone had a modest effect on apoptosis, whereas ABT-737 alone readily induced apoptosis in sensitive cell lines. In some cell lines, such as CHLA-119, the cells were found relatively resistant to ABT-737. However, in spite of this resistance, in all cell lines, the combination of the 2 drugs was synergistic in vitro. We also found that ABT-737 enhanced the activity of 4-HPR for CHLA-119 in vivo. Interestingly, we observed that the combination of these 2 agents had minimal toxic effect on normal human fibroblasts, which is consistent with similar observations made in normal resting lymphocytes (24), and suggests that the combination should have minimal toxicity for nonneoplastic cells in vivo.

The combination of 4-HPR and ABT-737 induced the loss of Δψm and the release of cytochrome c to the cytoplasm (Fig. 3B and C), indicating that this drug combination acts via the mitochondrial death pathway. Activation of caspase-9 and caspase-3 by ABT-737 in the presence or absence of 4-HPR further confirmed that ABT-737 acts via the mitochondria-dependent apoptotic pathway (Fig. 3D). In addition, the observation that caspase-8 was activated by these 2 agents (Fig. 3E), and that ABT-737 + 4-HPR–induced cleavage of Bid into t-Bid was inhibited by a caspase-8 inhibitor, indicate the involvement of the extrinsic apoptotic pathway as well (Fig. 4B). Furthermore, the majority of apoptosis induced by ABT-737 alone and by the combination of ABT-737 + 4-HPR occurred largely via caspase-dependent pathways as indicated by the fact that the pan-caspase inhibitor Boc-d-fmk inhibited the effect of ABT-737 and 4-HPR on apoptosis (Fig. 3B).

The mitochondrial apoptotic pathway is controlled by a balance between the proapoptotic protein members (i.e., the multidomain proapoptotic Bax, Bak, and BH3-only proapoptotic Bid, Bim, Bad, Bik, Noxa, Puma, Bmf, Hrk) and antiapoptotic protein members (i.e., the multidomain antiapoptotic Bcl-2, Bcl-X_L, Bcl-w, Mcl-1, Bfl/A1) of the Bcl-2 family. Those multidomain Bcl-2 proteins are functionally regulated by the BH3-only proteins (42, 43). Bid is localized in the cytosolic fraction of cells as an inactive precursor (44) and truncated Bid (t-Bid), the active form of Bid, is generated upon proteolytic cleavage by caspase-8 (45, 46).

We showed that ABT-737 together with 4-HPR increased the release of sequestered t-Bid from Bcl-2 and Bcl-X_L and the induction of Bax and Bak oligomerization (Fig. 4). The observation that Bid was cleaved into t-Bid in cells treated with ABT-737 alone or in combination with 4-HPR is similar to the previously reported activity of ABT-737 on leukemic and human pancreatic cancer cells (20, 21). The fact that Bax or Bak knockdown significantly impaired the apoptosis induced by ABT-737 + 4-HPR further confirms the importance of Bax and/or Bak activation in the synergistic effect observed with ABT-737 and 4-HPR. Another antiapoptotic Bcl-1 family protein Mcl-1 has been shown to be overexpressed in cells resistant to ABT-737 (22, 47–50), and a recent study in ALL showed that the combination of ABT-737 and 4-HPR was associated with Mcl-1 inactivation by 4-HPR (24). However, ABT-737 + 4-HPR did not alter Mcl-1 levels in neuroblastoma cells (Supplementary Fig. 2).

We extended our in vitro cytotoxicity studies by showing that 4-HPR and ABT-737 enhanced EFS when compared with either drug alone in CHLA-119 tumor–bearing mice, a mouse xenograft model of recurrent, TP53-mutated, p53-nonfunctional, MDR neuroblastoma (Fig. 5). Thus, we have shown a positive synergistic interaction between 4-HPR and the BH3-mimetic agent ABT-737 in neuroblastoma cell lines, and consistent with the in vitro activity, we showed antineuroblastoma activity of this novel combination in a neuroblastoma mouse xenograft model. These data support clinical trials combining 4-HPR with BH3-mimetic drugs in children with recurrent neuroblastoma.

C.P. Reynolds is a coinventor on a patent covering LXS fenretinide owned by Childrens Hospital Los Angeles. No potential conflicts of interest were disclosed by other authors.

H. Fang, T.M. Harned, Y.A. DeClerck, and C.P. Reynolds designed the research and analyzed the data; H. Fang, T.M. Harned, O. Kalous, and V. Maldonado carried out the research and analyzed the data; H. Fang wrote the manuscript; and Y.A. DeClerck and C.P. Reynolds edited the manuscript.

The authors thank Abbott Laboratories for providing ABT-737. 4-HPR LXS was kindly provided by Dr. B.J. Maurer.

The study was supported by National Cancer Institute grants CA81403 (C.P. Reynolds), CA82830 (C.P. Reynolds), and CA129377 (Y.A. DeClerck), and also by Cancer Prevention & Research Institute of Texas grant RP10072 (C.P. Reynolds). H. Fang was supported by The Saban Research Institute of CHLA as a recipient of pre-doctoral award. 4-HPR LXS was produced via a National Cancer Institute Rapid Access to Intervention Discovery (RAID) grant (B.J. Maurer).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.