Background
Atypical Teratoid Rhabdoid Tumor of the central nervous system (CNS AT/RT) is a highly malignant neoplasm of infants and young children. A biallelic inactivation of the
hSnf5/Ini1 gene located in 22q11.2 is a characteristic molecular defect in these tumors [
1‐
4]. Murine knock-out models have confirmed that hSnf5
/Ini1 is a tumor-suppressor gene [
5], but the details of its exact role in the initiation and growth of the AT/RT are still being investigated. To date, studies showing INI1 interaction with key signaling molecules suggest its potential to modify the response to factors that mediate cell growth and differentiation programs [
6‐
8]. There is emerging evidence for the existence of autocrine and/or paracrine growth factor signaling pathways in these cells. Previously, we were able to maintain disseminated AT/RT cells in culture by the addition of autologous CSF to culture medium [
9]. Agents that inhibit IGF-IR activity have been shown to diminish tumor cell growth and targeting of IGF-IR expression with antisense oligonucleotides resulted in increased apoptosis and sensitivity to a number of chemotherapeutic agents [
10]. Furthermore, Arcaro and colleagues have shown evidence for autocrine signaling by insulin and its receptor (IR) in AT/RT cells, which involves the PI3K/Akt pathway [
11]. These findings suggest that abnormally regulated cytokine pathways and their downstream signaling molecules can be effective targets for therapeutics in AT/RT.
Ultra structurally, AT/RT often presents as a polymorphous tumor with overlapping morphologic features consisting of primitive neuroectodermal tumor (PNET), mesenchymal, rhabdoid and epithelial components. This phenotypic heterogeneity is likely to be aided by multi-level cross stimulation of growth and survival pathways and signaling molecules. As such, a single-targeted agent may not be the optimal choice, as these agents may permit the development of salvage or escape mechanisms. However, by virtue of their ability to interfere with a diverse array of signaling molecules, including cytokine receptor kinases, multi-targeted inhibitors may provide a therapeutic advantage in the treatment of AT/RT. In the recent past, tyrosine kinase inhibitors with multiple targets have been found to have clinically achievable activity and acceptable tolerability in studies against heterogeneous malignancies [
12]. In this study, we have evaluated two such agents, sunitinib and sorafenib, for
in vitro activity and drug combinability against three AT/RT cell lines.
Discussion
Currently, the prognosis for children with AT/RT is very poor. Occasional anecdotal reports of successful treatment are noted; but optimal therapy or even effective therapy has not been achieved in most cases. The chemotherapeutic agents classically used are cyclophosphamide, cisplatin, etoposide, vincristine, carboplatin and ifosfamide [
1]. The setback is that tumors seem to be responsive initially but develop resistance [
20]. However, recent evidence suggests that improved survival can be achieved with the use of more aggressive treatment approaches, including dose-intense chemotherapy and adjuvant radiation therapy [
21‐
23]. It has also been shown that radiotherapy is crucial to improve the survival rate of children with AT/RT [
24,
25]. Chi and colleagues have described an innovative treatment approach consisting of an aggressive multimodality approach [
26]. This protocol is the first prospective investigation consisting of surgery, radiation therapy combined with multi-agent systemic and IT chemotherapy and has resulted in a significant improvement in time to progression and overall survival of AT/RT patients. In general, the striking potential for long-term consequences of treatments that include radiation in these very young children necessitates trials with new therapeutics and treatment regimens.
The role of cytokine receptor mediated growth and survival signals in rhabdoid tumors has been investigated by a number of laboratories. In addition to the effects of IGF-I described previously [
27], our studies have shown the expression of significant quantities of VEGF and PDGF by all three cell lines (Table
1). Based on this, we have explored the effects of two multi-kinase inhibitors that have been shown to inhibit growth stimulatory pathways mediated by the receptors of these cytokines. Sorafenib and sunitinib are two oral multi-targeted receptor-tyrosine kinase inhibitors that are currently in clinical trials for various malignancies. Sorafenib is a multi-kinase inhibitor that inhibits the activity of c-Raf, b-Raf, vascular endothelial growth factor receptor family (VEGFR-2 and VEGFR-3), platelet-derived growth factor receptor family (PDGFR-β ) and stem cell factor receptor (c-Kit). Sunitinib is a multitargeted inhibitor of VEGFR, PDGFR-α and -β, c-Kit and Flt-3. These two agents offer broad anti-tumor efficacy through their ability to directly and indirectly inhibit these targets in concert to ultimately interfere with tumor growth, survival, and angiogenesis [
28]. It has been shown in that the antiproliferative effect of sorafenib is mediated through its effect on the MAP kinase pathway (i.e., reduced Erk1/2 phosphorylation) [
29]. Our studies have shown a decrease in activated Erk1/2 in two of the three cell lines (Figure
4A). In addition, we have found a decrease in the anti-apoptotic protein Mcl-1 in all three cell lines. Interestingly, the down regulation of Mcl-1 by sorafenib has been shown previously in other tumor models [
30]. Mcl-1 has also been implicated in the generation of resistance to chemotherapeutic agents [
31]. Though we have shown significant alterations in the activity of key signaling molecules in AT/RT cells, the contribution of off-target effects by sorafenib cannot be ruled out and awaits further analysis in biological correlative studies in xenografts and in future clinical trials of this agent.
Recently, sorafenib has been shown to inhibit proliferation and induce apoptosis in two medulloblastoma cell lines (Daoy and D283) and a primary culture (VC312) of human medulloblastoma at inhibitory concentrations very similar to that we have observed against AT/RT cells (IC
50 approximately 2.5 µM) [
32].
In vivo activity of sorafenib against medulloblastoma cells has also been demonstrated in a mouse xenograft model [
32]. Sunitinib has been shown to induce apoptosis and growth arrest in medulloblastoma cells by inhibiting Stat3 and Akt signaling pathways [
33]. In pre-clinical testing studies, Maris and co-workers have observed activity of sunitinib against rhabdoid tumor xenografts [
34]. These findings support the potential of sorafenib and sunitinib as effective treatments in AT/RT. However, as a way to increase treatment efficacy and to reduce potential adverse effects of these agents, we explored additional drug combination studies.
Irinotecan has been shown to have the ability to cross the blood-brain barrier and, in preclinical investigations, has demonstrated cytotoxic activity against central nervous system tumor xenografts [
35]. Recently, a Phase I trial of irinotecan by Pediatric Oncology Group (POG) was performed in children with refractory solid tumors where stable disease (4-20 cycles) was observed in seven patients with a variety of malignancies, including a patient with CNS AT/RT [
36]. In recurrent malignant gliomas, combination therapy with bevacizumab and irinotecan has been shown to prolong progression-free survival in comparison with historical controls [
37,
38]. Our studies have also shown the ability of irinotecan to inhibit the growth of AT/RT cells and significant synergy in drug combinations involving irinotecan with either sorafenib or sunitinib (Table
3). In previous trials, despite the initial response to therapy, most patients treated with irinotecan developed resistance and showed tumor progression [
39]. In the colorectal cancer model, treatment with irinotecan has been shown to lead to the activation of NF-κB. [
40]. As such, the activation of the NF-κB pathway constitutes a potential mechanism of inducible resistance by malignant cells exposed to irinotecan [
41]. NF-
κ B interferes with the effect of most anti-cancer drugs through induction of anti-apoptotic genes. Targeting NF-
κ B is therefore expected to potentiate conventional treatments in adjuvant strategies. In addition, recent studies have shown that the administration of siRNA directed against the p65 subunit of NF-κB can effectively enhance
in vitro and
in vivo sensitivity to chemotherapeutic agents [
42]. Thus, reducing NF-κB-mediated activation may help prevent resistance potentially generated upon exposure to irinotecan. This has been confirmed in studies where a pharmacological inhibitor of the IKK2 kinase (AS602868, Serono International SA, Geneva, Switzerland) which blocks NF-
κ B activation has been found to enhance the action of irinotecan metabolite [
43]. We have explored the possibility of decreased NF-
κ B activation as a potential mechanism in the enhanced cytotoxicity of irinotecan in the presence of sorafenib. Our studies have provided evidence for irinotecan mediated loss of cytoplasmic NF-
κ B in AT/RT cells. However, the presence of sorafenib appears to retain NF-
κ B in the cytoplasm as shown by Western blot analysis and indirect immunofluoresence studies. Interestingly, in Alzheimer's disease research, a similar observation was noted where the chronic treatment with sorafenib inhibited c-Raf and NF-
κ B in the brains of the aged APPswe mice [
44].
Methods
Cell lines and cell culture
BT12 and BT16 cell lines were a gift from Drs. Peter Houghton and Jaclyn Biegel (St. Jude Children's Research Hospital, Memphis, TN, USA). These cell lines have been established from infants with CNS AT/RT (BT12 from a six-week-old female; BT16 from a two-year-old male). KCCF1 was established in our laboratory with cells obtained from the Cerebral Spinal Fluid (CSF) of a two-month-old male infant with AT/RT. Characterization of this cell line has been described previously [
9].
Cells were cultured in Opti-MEM medium (Gibco, Invitrogen Corporation, Burlington, Ontario, Canada) supplemented with 5% fetal bovine serum (Gibco), 100 units/ml each of penicillin and streptomycin (Gibco). Confluent cells were trypsinized with 0.25% Trypsin-EDTA in Ca2+ and Mg2+ free balanced salt solution (Gibco) every three to five days. All cell cultures were maintained at 37 °C in a humidified atmosphere with 5% CO2.
Multiplex cytokine assay
Cells (2 × 10
6) in 2 ml of medium were grown for 48 hours and culture supernatants were analyzed for cytokine levels by multiplex technology as described previously [
45]. Briefly, culture supernatants were diluted 1:4 in sample diluent buffer and mixed with beads containing capture antibodies. After incubation and washing steps, beads were mixed and incubated with biotin conjugated detection antibodies. Following the detection antibody incubation, each well was filter-washed and incubated in the dark with streptavidin-phycoerythrin (PE) conjugate. The plates were washed and the contents of each well were re-suspended in assay buffer. The plates were read in a Luminex100™ multiplex assay detection system (Luminex Corp., Austin, Texas, USA) and quantitatively analyzed using STarStation 2.0 software (Applied Cytometry, Sheffield, United Kingdom). In this study, a panel containing the following 65 human cytokines and chemokines was used: EGF, Eotaxin, FGF(basic), Flt-3, Ligand, Fractalkine, G-CSF, GM-CSF, GRO, IFN alpha2, IFN gamma, IL-1alpha, IL-1beta, IL-1Ra, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12(p40), IL-12(p70), IL-13, IL-15, IL-17, IP-10, MCP-1, MCP-3, MDC, MIP-1alpha, MIP-1beta, PDGF-AA, PDGF-AB/BB, RANTES, CD40L, SDF-1, sIL-2R alpha, TGF alpha, TNFbeta, MCP-2, MCP-4, ENA-78, BCA-1, I-309, IL-16, MIP-1delta, TARC, 6Ckine, Eotaxin-2, Eotaxin-3, CTACK, IL-23, LIF, TPO, TRAIL, SCF, TSLIP, VEGF, IL-20, IL-21, IL-28A and IL-33.
Antineoplasic agents
Sorafenib, sunitinib, irinotecan and SN-38 were obtained from ChemieTek (Indianapolis, Indiana, USA) and the Oncology pharmacy at the Alberta Children's Hospital. These agents were dissolved in DMSO (Sigma-Aldrich, Oakville, Ontario, Canada) to a final concentration of 10 mM and stored in aliquots at -20°C. At the time of study, agents were then appropriately diluted in culture medium.
Cell growth inhibition assay
AT/RT cells were detached from the flask by trypsinization and plated in 96 well plates at a concentration of 1 × 103 to 5 × 103 cells per well. Increasing concentrations of study agents were added to these wells to a final volume of 200 µl per well. Corresponding dilutions of the vehicle DMSO was used as control. After four days in culture, cell survival was quantified by Alamar Blue Assay (Medicorp, Montreal, Quebec, Canada), according to manufacturer's protocol. Briefly, cells were incubated with 2.5% Alamar blue (which incorporates a proprietary redox indicator that changes color in response to metabolic activity) for 2 to 24 hours, and the absorbency at 570-620 nm was measured (Opsys MR Plate Reader, Dynex Technologies, Chantilly, Virginia, USA). Percent cell survival was calculated by:% Survival = (Abs 570-620 test well/Abs 570-620 Control) × 100.
From these values, inhibitory concentrations inducing 50% cell death (IC
50) compared to DMSO wells were calculated. For drug combination studies irinotecan at IC
25 concentration was added to cultures containing increasing concentrations of sorafenib or sunitinib. IC
50 values were then calculated for these agents alone or in combination with irinotecan and used to derive Combination Index (CI) values as described previously [
16]. A CI of less than 1 indicates synergy between the two agents under the experimental conditions used.
Western Blot Analysis
In order to determine the expression of cellular targets of sorafenib and sunitinib in AT/RT cell lines, cells were grown to confluence in six well culture plates (Nunc, Rochester, New York, USA) over a 24 hour period. The media was removed and cells were washed with ice cold PBS and lysed in buffer containing 50 mM Tris, 5 mM EDTA, 0.1% SDS, 1% Triton X-100, 0.5% sodium deoxycholate with phosphatase and protease inhibitors (Sigma). Protein content of the lysates was measured by BCA Protein Assay Kit (Pierce, Rockford, Illinois, USA). Proteins (30µg/sample) were separated on an 8% polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (Bio-Rad, Mississauga, Ontario, Canada). The membranes were blocked for two hours at 4°C with 5% skim milk powder in PBS containing 0.1% Tween-20 (Sigma). The blots were incubated with primary antibodies to c-Kit (1:1000, Santa Cruz Biotechnology, Santa Cruz, California, USA), PDGFR-β (1:1000, Santa Cruz), VEGFR2 (1:1000, Millipore, Billerica, Massachusetts, USA), Flt-3 (1:1000, Santa Cruz), c-Raf (1:1000, Cell Signaling Technology, Danvers, Massachusetts, USA), p-38α (1:1000, Santa Cruz) and β-actin (1: 10000, Sigma). After incubation overnight at 4°C, membranes were washed and probed with appropriate secondary antibodies conjugated to horseradish peroxidase (HRPO) (Sigma), followed by a luminal based substrate (Mandel, Guelph, Ontario, Canada) and developed by exposure to x-ray film (Fisher Scientific, Ottawa, Ontario, Canada). For intracellular signaling studies, cells were grown to confluence in six well culture plates and culture supernatant (spent medium) was removed, filtered and stored at 4°C and fresh serum free medium containing 10 μM sorafenib or vehicle (DMSO) control was added to the cells. After an additional two hours in culture the spent medium was added (10% v/v). Following further 30 min in the incubator, cells were lysed as described above and analyzed by Western blot using primary antibodies to Erk1/2 (1:1000, Cell Signaling), phospho-Erk1/2 (1:2000, R & D Systems, Minneapolis, Minnesota, USA), Akt1/2 (1:1000, Santa Cruz), phospho-Akt1/2 (1:1000, Santa Cruz), c-Raf (1:1000, Cell Signaling Technology), phospho-c-Raf (1:1000, Cell Signaling), Stat3 (1:1000, Cell Signaling), phospho-Stat3 (1:1000, Cell Signaling), Mcl-1 (1:1000, Santa Cruz) and β-actin. For the analysis of cytoplasmic NF-κ B, phospho-NF-κ B and Iκ Bα, cells were grown in culture for 2 days, after which the media was removed and replaced with serum free media. The cells were then treated with sorafenib (10 µM) or DMSO control for 30 minutes, then irinotecan (10 µM) for an additional 2 hours. Cell lysates were then analyzed by using primary antibodies to NF-κ Bp65 (1:1000, Santa Cruz), phospho-NFκ Bp65 (1:1000, Cell Signaling), NF-κ Bp50 (1:2000, Millipore), Iκ Bα (1:1000, Cell Signaling) and β-actin. For analysis of p27Kip1 expression, cells were treated with either sorafenib (1 µM) or irinotecan (1 µM) or both for 48 hours. Expression of p27Kip1 was determined using anti-p27Kip1 (1:1000, Cell Signaling).
Immunofluorescence analysis of cytoplasmic NF-κB
BT12 cells were cultured in six well plates (2 × 10
6/well) overnight and treated with (1) vehicle alone, (2) sorafenib (10 μM), (3) irinotecan (10 μM) and (4) sorafenib for 30 minutes followed by treatment with irinotecan for an additional 2 hours. Indirect immunofluoresence studies were carried out as described previously [
46]. Briefly, after various treatments, cells were washed with cold PBS, fixed and incubated with antibodies to NF-κB (Santa Cruz) for 1 hour, followed by fluorescent labeled secondary antibodies (30 min). Concurrent DAPI staining was performed to locate nuclei in each slide. Slides were visualized under a fluorescent microscope and random fields were photographed.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AJ planned and performed drug combination studies and Western blot analysis. DB planned and performed cytotoxicity studies. PB and KR facilitated the indirect immunofluorescence analysis and helped to write the manuscript. AN planned and supervised the studies and wrote the manuscript. All authors have read and approved the final manuscript.