There is a crucial need for better therapeutic approaches for the treatment of Ewing sarcoma (EWS). Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) induces apoptosis in EWS cells in vitro. However, in vivo, acquired resistance to TRAIL is a major limiting factor. Platelet-derived growth factor receptor-β (PDGFR-β) is highly expressed on EWS cells. Thus, the authors evaluated whether PDGFR-β blockade could sensitize EWS cells to TRAIL-induced apoptosis in vitro and in vivo.
The effect of combined imatinib mesylate (Gleevec) and TRAIL on EWS cell growth and apoptosis was tested in vitro. Stable PDGFR-β knockdown in EWS cells was achieved by short-hairpin RNA transduction, and TRAIL sensitivity of these cells was measured. Expression of death receptors was measured by fluorescence-activated cell-sorting (FACS) analysis, and caspase 8 activity was evaluated by Western blot analysis. An orthotopic human xenograft model of EWS growth and spontaneous metastasis in nude mice was used to assess the in vivo affect of combined imatinib mesylate and TRAIL.
Imatinib mesylate induced a significant TRAIL proapoptotic effect in EWS cells in vitro. Specific PDGFR-β silencing in EWS cells enhanced the effects of TRAIL, possibly through an increase in the expression of death receptors 4 and 5. The combination of imatinib mesylate and TRAIL significantly inhibited the growth of primary tumors and decreased the incidence of spontaneous EWS pulmonary metastasis compared with either drug alone.
Ewing sarcoma (EWS) is a disease of children and adolescents, and most cases occur in the second decade of life.1, 2 Patients who present with localized disease have better survival (5-year survival rate, 60%-70%) than patients who present with metastasis, who have a 5-year disease-free survival rate of 30% to 40%, depending on the secondary disease site.2, 3 Currently, there is no effective treatment available for patients with EWS who harbor metastatic disease, and even recently used approaches with high-dose chemotherapy plus stem cell rescue have not been successful in this context.4 Therefore, novel therapies for metastatic EWS are urgently needed.
Induction of apoptosis has been suggested as a potentially useful strategy for cancer therapy. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is a tumor necrosis superfamily member that has demonstrated the ability to induce programmed cell death in tumor cells without causing apoptosis in normal cells.5 TRAIL acts by binding to death receptors 4 (DR-4) and 5 (DR-5), which, in turn, activate a complex apoptotic downstream cascade.3, 6-10 It has been demonstrated that TRAIL induces apoptosis in vitro in several pediatric sarcoma histologies, such as rhabdomyosarcoma.11, 12 EWS cell lines also have been identified as highly sensitive to TRAIL in vitro, especially in cells that express TRAIL receptors.13-16 Sensitization of death receptor-induced apoptosis and abrogation of TRAIL resistance in vitro leads to increased tumor cell sensitivity to TRAIL and the concomitant inhibition of EWS growth.17
TRAIL as a single agent in EWS induces apoptosis in 7 of 9 EWS cell lines. Preincubation with interferon gamma (IFN gamma) rendered the 2 resistant cell lines sensitive.13-15 When TRAIL was administered in an orthotopic model, tumor growth was slowed in 60% of animals with a durable remission in only 11% to 19%, but metastatic disease was not effected. The addition of doxorubicin chemotherapy did not improve response rates.18 Thus, if TRAIL is to be used as part of the therapeutic armamentarium for EWS, there is a need to identify means of sensitizing EWS to the effects of TRAIL to address local and metastatic disease.
Small-molecule tyrosine kinase inhibitors, such as the epidermal growth factor receptor (EGFR) inhibitor gefitinib, have been identified recently that interact with TRAIL, leading to the reversal of TRAIL resistance and a significant increase in TRAIL-induced apoptosis in bladder carcinoma cell lines.19 Other small-molecule tyrosine kinase inhibitors have been evaluated in EWS but have not been proven effective as single agents. In EWS tumors, platelet-derived growth factor receptor β (PDGFR-β), and not EGFR, is highly expressed.20 It is believed that PDGFR-β expression in EWS cells enhances their motility and growth,20 although an inhibitor of PDGFR-β, imatinib mesylate (Gleevec; Novartis International AG, Basel, Switzerland), as a single agent was not successful in inducing response in a phase 2 study. In another phase 2 study of imatinib mesylate as a single agent in pediatric patients with solid tumors, only 1 of 24 patients with EWS had a response.21 However, to our knowledge, the role of PDGFR-β in combination with TRAIL has not been studied previously.
The objective of the current study was to evaluate the effects of combining PDGFR-β blockade with TRAIL on the growth of EWS in vitro and in vivo. Our results demonstrate that PDGFR-β inhibition in EWS enhances the proapoptotic effects of TRAIL, possibly through the increased expression of the DR-4 and DR-5 receptors and caspase 8. Furthermore, using a preclinical human EWS orthotopic murine model, we demonstrate that combining PDGFR-β inhibition (using imatinib mesylate) with TRAIL results in superior EWS growth inhibition and apoptosis in vivo.
MATERIALS AND METHODS
Cell Culture and Reagents
Human TC71 and A673 EWS cell lines obtained from the American Type Culture Collection (ATCC, Manassas, Va) were cultured in Eagle modified essential medium with 10% fetal bovine serum, 2 mmol/L L-glutamine, 1 mmol/L sodium pyruvate, 1 × nonessential amino acids, and 2 × minimal essential medium vitamin solution (Life Technologies, Carlsbad, Calif). TC71 cells stably transfected with PDGFR-β short-hairpin RNA (shRNA) (TCpdsh), vector scrambled controls (TCpd), and control cells were maintained in the TC71 medium with 200 μg/mL hygromycin B (Life Technologies). All cells were negative for Mycoplasma, as determined by a Mycoplasma Plus polymerase chain reaction (PCR) primer set (Stratagene, La Jolla, Calif).
The small-molecule tyrosine kinase inhibitor imatinib mesylate was obtained from the University of Texas M. D. Anderson Cancer Center pharmacy. TRAIL was produced as follows: combinational DNA (cDNA) from the extracellular domain of TRAIL corresponding to amino acids 114 through 281 was subcloned into the pET17/b bacterial expression vector and expressed in the BL21(DE3)pLysE (both from Novagen, Calgary, Alberta) bacterial host. After induction of TRAIL expression using isopropyl-β-thio-galactosidase (1 mM), bacterial pellets were harvested, and TRAIL was purified after passage through a nickel column (Ni-NTA) followed by a size exclusion column (Amersham Biosciences, Little Chalfont, United Kingdom) according to previously published procedures.14
Measurement of Cell Proliferation
Cell growth assays were processed using CellTiter96 Aqueous Non-Radioactive Cell Proliferation Assay kit (Promega, Madison, Wis) according to the manufacturer's instructions. Growth rates were analyzed using increasing TRAIL and imatinib mesylate doses either alone or in combination. Absorbance was measured at 490-nm wavelengths, and treated cell absorbance values are presented as the percentages of untreated cell absorbance.
Fluorescence-Activated Cell-Sorting Analysis of Propidium Iodide Stained Nuclei
The appearance of a subdiploid G1 peak characteristic of apoptosis was detected by fluorescence-activated cell-sorting (FACS) analysis of propidium iodide (PI)-stained nuclei with FACS scanning. Briefly, cells were collected at the end of incubation periods and were fixed in 70% ethanol and stored at 44°C until use. After washing with phosphate-buffered saline (PBS), the fixed cells were treated with RNase in PI buffer for 30 minutes. At the end of treatment, PI (5 μg/mL) was added, and the cells were stained for 30 minutes in the dark. Ten thousand events were acquired (excluding the debris) and were analyzed using CellQuest software (Becton-Dickinson and Company, Franklin Lakes, NJ) for the detection of apoptosis. Modfit software (Verity Software House, Topsham, Me) was used for cell cycle analysis.
Western Immunoblot Analysis
Western blot analysis was performed as described previously.22 Briefly, 50 μg of proteins extracted from cultured cells were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (Millipore Company, Bedford, Mass); then, membranes were blocked and blotted with relevant antibodies. Horseradish peroxidase-conjugated secondary antibodies were detected by enhanced chemiluminescence (ECL; Amersham Biosciences, Little Chalfont, United Kingdom). IRDye 680-conjugated and IRDye 800-conjugated secondary antibodies (Molecular Probes, Eugene, Ore) were detected using Odyssey Imaging (LICOR Biosciences, Lincoln, Neb)
Stable Anti-PDGFR-β shRNA Transfection
shRNA expression vector pSilencer2.1-U6 Hygro was purchased from Ambion (Austin, Tex). shRNA-expressing plasmids targeting human PDGFR-β were constructed according to the manufacturer's instructions. Briefly, 4 pairs of cDNA oligonucleotides targeting human PDGFR-β messenger RNA at 4 different locations were synthesized by Integrated DNA Technologies (Coralville, Iowa). Each pair of oligonucleotides was annealed at 90°C for 3 minutes, cooled to 37°C, and incubated for 1 hour. The annealed double-stranded DNA oligonucleotides were ligated between the BamHI and HindIII sites on the pSilencer2.1-U6 Hygro vector. The control vector (nontargeting shRNA) was constructed by inserting a sequence that expresses limited homology to sequences in the human and mouse genomes. The targeted PDGFR-β sequences were 1) AACTATTCATCTTTCTCACGG, 2) AATGAGGTGGTCAACTTCGAG, 3) AAGGTGATTGAGTCTGTGAGC, and 4) AATGAAGTCAACACCTCCTCA; and the control (cells transfected with PDGFR-β small-interfering RNA [pdsh]) was CTACCGTTGTTATAGGTGTCTCTTGAACACCTATAACAACGGTAGT. All inserted sequences were verified by DNA sequencing. Transfections were done with Superfect (Qiagen, Valencia, Calif) as directed by the manufacturer and were selected in hygromycin B (Invitrogen Life Technologies, Carlsbad, Calif) that contained medium at 400 μg/mL for TC71 cells. Stable transfected cell clones were tested for PDGFR-β expression by reverse transcriptase (RT)-PCR and Western blot analyses. PDGFR-β shRNA used in these experiments was derived from a single clone named TCpdsh.
Total RNA was extracted from specified cell lines. cDNA was synthesized using an RT system (Invitrogen Life Technologies) amplified by PCR with specific primers for PDGFR-β (sense, 5′gACACCAgCTCCgTCCTCTA; antisense, ggCTgTCACAggAgATggTT-3′). The initial denaturation was done at 94°C for 10 minutes. Then, the products were subjected to denaturation at 94°C for 1 minute, and 58°C for 1 minute, extension at 72°C for 1 minute (35 cycles), and a final elongation at 72°C for 10 minutes. The PCR products were subjected to electrophoresis on a 2% agarose gel with ethidium bromide and were observed under ultraviolet light. The PDGFR-β was 232 base pairs, and 18S primers (Ambion) were used as internal controls.
Detection of TRAIL Receptors by Flow Cytometry
Briefly, 5 × 105 cells were plated in 24-well plates and treated as needed every 24 hours. Wells were washed with PBS, and cells were collected in 1.5-mL centrifuge tubes. Cells were washed twice with 1 mL FACS buffer, then resuspended in 100 μL FACS buffer, and the primary antibody (or isotope immunoglobulin G control) was added. Then, the cells were incubated on ice for 45 minutes, washed twice with PBS, the secondary antibody labeled with fluorescein 5′-isothiocyanate was added for 45 minutes, the cells were washed 2 more times, and then they were and resuspended cell in 1 mL FACS buffer before being measured by flow cytometry (BD FACScalibur; Becton-Dickinson and Company).
In Vivo Ewing Animal Model
Specific pathogen-free athymic (T-cell deficient) nude mice (6-8 weeks old) were purchased from the National Cancer Institute (Bethesda, Md). TC71, TCpdsh, and TCpd control cells in mid-log growth phase were harvested by trypsinization. Single-cell suspensions (0.5 × 106 cells/20 μL PBS) were injected into the rib after the skin was incised. Three phenotypes consistently emerged as experiments were done in triplicate: 1) Mice developed large chest wall tumor (∼60%); or 2) no minimal local tumor was apparent, but mice developed bilateral lung metastasis (∼30%); or 3). both chest wall tumors and bilateral pulmonary metastasis occurred (∼10%). Mice with chest wall tumors were followed for tumor size, which was measured 3 times weekly using a caliper, and the greatest dimensions were recorded. Tumor volume was calculated according to the formula a2b/2, where a and b were the 2 greatest dimensions. Based on previous experiments, mice without chest wall tumors had pulmonary metastasis. Once chest wall tumors reached 0.5 cm in greatest dimension, the mice were allocated to the following treatment arms (n = 10 mice per group): 1) control (PBS), 2) oral imatinib mesylate (50 mg/kg daily), 3) intraperitoneal (ip) TRAIL (25 mg/kg once weekly), or 4) imatinib mesylate plus TRAIL. Treatment continued for 3 weeks. Mice were followed for tumor size and body weight and were killed when control group tumors reached 1.5 cm in greatest dimension.
Mice with lung metastases (identified by the <2-mm chest wall tumors) were allocated to the same 4 treatment arms described above; however, TRAIL was delivered intranasally at 20 mg/kg weekly. Treatment for these mice began 2 weeks after the injection of tumor cells. At the end of each experiment, animal lungs were weighed and subjected to gross examination for the presence of visible metastases and to histologic analyses for the presence of micrometastases.
Tumors and lungs were frozen or fixed in formalin and paraffin embedded for further immunohistochemical studies. Experiments were conducted with 10 mice per group, and each experiment was repeated in triplicate. All animal experiments were approved by The University of Texas M. D. Anderson Cancer Center Institutional Animal Care and Use Committee.
Terminal deoxynucleotidyl transferase deoxyuridine triphosphate (dUTP) nick-end labeling (TUNEL) was performed by using a commercially available apoptosis detection kit (Promega) with the following modifications: Samples were fixed with 4% paraformaldehyde (methanol free) for 10 minutes at room temperature, washed twice with PBS for 5 minutes. and then incubated with 0.2% Triton X-100 for 15 minutes at room temperature. After 2 washes for 5 minutes each in PBS, the samples were preincubated with equilibrium buffer (from the kit) for 20 minutes at room temperature. New buffer was added with nucleotide mix and terminal deoxynucleotidyl transferase enzyme, and the samples were incubated in a humid atmosphere at 37°C for 1 hour in the dark. The reaction was terminated by immersing the samples in 2 × standard saline citrate for 20 minutes. Samples were washed 3 times for 5 minutes each with PBS to remove unincorporated fluorescein-dUTP. TUNEL staining was quantified by observing the percentage of cells stained on each slide per high-power field.
Cell culture-based assays were repeated at least 3 times, and the mean ± standard deviation was calculated. Cell lines were examined separately. For outcomes that were measured at a single time point, 2-sample t tests were used to assess the differences. Differences in xenograft growth and metastasis in vivo were assessed using a 2-tailed Student t test. Significance was set at P ≤ .05.
Combined TRAIL and Imatinib Mesylate Results in Superior Anti-EWS Effects Compared With Each Therapy Alone
TRAIL is emerging as a novel anticancer therapeutic because of its unique, selective, antitumor, proapoptotic effects, which spare normal cells.14-16 EWS cells are sensitive to TRAIL in culture; however, only a limited effect can be observed in vivo.20, 21 Thus, there is a need to identify means to sensitize these tumors to the effect of TRAIL. Recent data suggest that inhibition of tyrosine kinase receptors may result in TRAIL sensitization.23-25 Our initial experiment demonstrated that EWS cells express significantly higher PDGFR-β levels compared with other sarcoma cell lines (Fig. 1A). On the basis of this finding, next, we evaluated whether PDGFR-β blockade might enhance TRAIL sensitivity in vitro. Human EWS cells were treated with imatinib mesylate (a multitargeted small-molecule inhibitor known to block the PDGFR-β), TRAIL, or both agents combined. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay demonstrated a synergistic growth-inhibitory effect when the combination was used (Fig. 1B). FACS analysis with PI staining also demonstrated that the addition of imatinib mesylate to TRAIL induced significantly higher rates of apoptosis (Fig. 1C).
PDGFR-β Inhibition Sensitizes EWS Cells to TRAIL-Induced Apoptosis
Combining imatinib mesylate with TRAIL, as indicated above, resulted in superior anti-EWS effects. However, imatinib mesylate is not a specific PDGFR-β inhibitor, because it has been to shown to inhibit the activity of additional tyrosine kinase receptors. To further confirm our hypothesis that the observed increase in EWS cell apoptosis was caused at least in part by PDGFR-β inhibition, next, we stably silenced PDGFR-β in TC71 cells using a PDGFR-β–specific shRNA (Fig. 2A,B). TC71 cells, TC71-nontargeting controls, and TC71 PDGFR-β shRNA cells were treated further with increasing TRAIL doses for 48 hours. A dose-dependent decrease in cell viability (Fig. 2C) and an increase in apoptosis (Fig. 2D) were demonstrated in all cells that we tested. However, a more pronounced, statistically significant (P < .05) effect was noticed in the PDGFR-β–silenced cells (TC71pdsh) compared with parental TC71 cells and nontargeting shRNA-transfected cells (Fig. 2C,D). These data demonstrated that PDGFR-β knockdown sensitizes EWS cells to TRAIL-induced cell death.
The effect of TRAIL on cancer cells is conveyed through the death receptors DR-4 and DR-5 and the consequential activation of the extrinsic apoptotic pathway. Previous studies suggested that decreases in death receptor and/or caspase 8 levels in tumor cells in vivo are potential mechanisms for acquired TRAIL resistance.13 Thus, we attempted to further evaluate the effect of PDGFR-β on the expression of these proteins. PDGFR-β knockdown in TC71 cells resulted in a significant increase in DR-4 and DR-5 receptor expression, as illustrated in Figure 3A. These results offer 1 potential mechanism for the observed PDGFR-β inhibition-induced TRAIL sensitization. PDGFR-β expression in EWS potentially inversely regulates the expression of DR-4 and DR-5 on these cells.
Next, we evaluated the potential effect of PDGFR-β on caspase 8 expression. Figure 3B shows that no change in caspase 8 levels could be observed in PDGFR-β–silenced cells compared with controls. However, activated caspase 8 levels were increased significantly in TC71pdsh cells compared with either nontargeting or control cells (Fig. 3B), further supporting the observed increased TRAIL-induced apoptosis in PDGFR-β knocked down EWS cells.
Taken together, the data presented demonstrate that PDGFR-β inhibition enhances TRAIL-induced apoptosis in EWS cells. This effect, at least in part, is the result of PDGFR-β blockade-induced DR-4/DR-5 expression.
Combined Imatinib Mesylate and TRAIL Therapy Inhibits EWS Growth In Vivo
To further investigate the effect of the imatinib mesylate/TRAIL combination on tumor growth, we used the novel orthotopic human EWS xenograft model described above (see Materials and Methods). After localized tumors grew to 5 mm in the chest wall, mice received either imatinib mesylate alone by oral gavage, ip TRAIL alone, or the combination of both agents. At termination of the study, the average size of the tumors was 499 mm3 without treatment (PBS), 443 mm3 with imatinib mesylate treatment, 310 mm3 with TRAIL ip (P = .028), and 224 mm3 with imatinib mesylate plus TRAIL (P = .0048). Primary chest wall tumor growth was inhibited significantly using combination therapy compared with either treatment alone (Fig. 4A). These data suggest that the superior anti-EWS effect of combined imatinib mesylate and TRAIL identified in vitro also can be recapitulated in vivo.
Combined Imatinib Mesylate and TRAIL Therapy Inhibits EWS Metastases Growth In Vivo
Next, we evaluated the effect of combination therapy on the growth of EWS lung metastases. When human EWS cells were injected into the ribs of nude mice, a subset of mice did not develop local tumors but reproducibly developed spontaneous lung metastases. This subset of mice was used for a therapeutic experiment, as described above, and the mice were allocated to 1 of 4 treatment groups: 1) control (vehicle only), 2) imatinib mesylate (by gavage), 3) TRAIL (intranasally), or 4) combined imatinib mesylate and TRAIL. At the completion of 21 days of treatment, total lung weights were compared. The average lung weight was 1.02 g in mice without treatment (PBS), 0.85 g in mice that received imatinib mesylate alone, 0.58 g in mice that received intranasal TRAIL alone, 0.38 g in mice that received intranasal TRAIL plus imatinib mesylate (the average normal lung weight in mice without metastasis was 0.20 g). Treatment with intranasal TRAIL plus imatinib mesylate provided superior inhibition of pulmonary metastasis compared with either therapy alone (P < .01) (Fig. 4B). Furthermore, TUNEL staining demonstrated increased apoptosis in combination-treated pulmonary metastasis compared with metastases that were treated with either agent alone (Fig. 4C). Similarly, increased apoptosis was observed in chest wall tumors when combination therapy was used (Fig. 4D). There was no difference in apoptosis observed in chest wall tumors or pulmonary metastases when comparing single-agent treatment with either control (PBS), or TRAIL, or imatinib mesylate (Fig. 4C,D). Taken together, the data presented here suggest that the imatinib mesylate/TRAIL combination is a potential treatment strategy that may have superior inhibitory effects on EWS growth and metastasis.
TRAIL is emerging as a novel anticancer therapy. It has been demonstrated that TRAIL induces apoptosis in many tumors types, including several sarcomas, such as EWS, fibrosarcoma, liposarcoma, rhabdomyosarcoma, Wilms tumor, clear cell sarcoma of the kidney, and neuroblastoma.26-33 However, in vivo TRAIL therapy has not proven efficacious for many tumors,26-33 possibly because of acquired TRAIL resistance. One possible mechanism that may underlie the observed in vivo EWS resistance to TRAIL is defects in DR-4 and DR-5 signaling.34 It has been observed that histone deacetylase (HDAC) inhibitors in vivo enhance TRAIL-induced apoptosis by caspase 8 but fail to increase the expression of DR-4 and DR-5 on EWS cells.23 Others have reported that EWS growing in vivo can acquire TRAIL resistance with the associated down-regulation of TRAIL receptors.35 In contrast, combination therapy (eg, IFN gamma, HDAC, and proteasome inhibitors) has been associated with the inhibition of EWS growth in vivo.22, 24, 25, 27 It also has been observed that IFN gamma treatment increases EWS expression of TRAIL receptors; the combination of TRAIL agonists and IFN gamma significantly decreased primary EWS growth and the incidence of metastatic disease in a xenograft model.27 In the current report, we describe an approach to overcoming TRAIL resistance in the treatment of EWS by inhibiting PDGFR-β that, in turn, potentiates the expression of DR-4 and DR-5 expression in human EWS cells, possibly through caspase 8.
We observed an increase in cleaved caspase 8 with TRAIL treatment of EWS cells in which PDGFR-β has been silenced compared with wild-type cells, suggesting caspase 8 dependence and the intrinsic pathway as a mechanism of increased TRAIL-induced apoptosis in the absence of PDGFR-β. Other authors have reported that the transfection of wild-type (but not mutant) caspase 8 into caspase 8-deficient EWS cells restored their sensitivity to TRAIL, indicating that up-regulation of caspase 8 may be sufficient to restore TRAIL sensitivity.25 However, the caspase 8 status of EWS tumors in humans has not been correlated with survival after anti-EWS chemotherapy.25
Imatinib mesylate as single-agent therapy has been proven effective in gastrointestinal stromal tumors in adults.35 Disappointingly, in a phase 2 study of imatinib mesylate as a single agent in pediatric patients with solid tumors, only 1 of 24 patients with EWS had a response.21 However, Hamai and colleagues36 reported that imatinib potentiated TRAIL-induced apoptosis in T1 melanoma cells by a mechanism that involved cleaved caspase 3, caspase 8, and caspase 9. Those authors also observed that resistance to TRAIL-induced apoptosis in metastatic melanoma cells was independent of c-kit activation.
Because there are no currently approved, specific inhibitors of PDGFR-β available, we used imatinib mesylate to evaluate the effect of PDGFR-β inhibition on tumor growth in an EWS orthotopic xenograft model. We observed growth inhibition of both EWS primary tumors and spontaneous pulmonary metastasis when imatinib mesylate was given in combination with TRAIL in vivo. To our knowledge, our study is the first in vivo report of this combination therapy. The efficacy we have demonstrated in the current study using combined imatinib mesylate and TRAIL has not been reported previously in either primary or metastatic EWS.
CONFLICT OF INTEREST DISCLOSURES
Supported in part by a grant from the Robert Wood Johnson Foundation, Harold Amos Award, and the Amschwand Sarcoma Foundation.