Monotherapies have proven largely ineffective for the treatment of glioblastomas, suggesting that increased patient benefit may be achieved by combining therapies. Two protumorigenic pathways known to be active in glioblastoma include RAS/RAF/mitogen-activated protein kinase and phosphatidylinositol 3-kinase/AKT/target of rapamycin (TOR). We investigated the efficacy of a combination of novel low molecular weight inhibitors LBT613 and RAD001 (everolimus), which were designed to target RAF and TOR, respectively. LBT613 decreased phosphorylation of extracellular signal-regulated kinase 1 and 2, downstream effectors of RAF, in a human glioma cell line. RAD001 resulted in decreased phosphorylation of the TOR effector S6. To determine if targeting RAF and TOR activities could result in decreased protumorigenic glioma cellular behaviors, we evaluated the abilities of LBT613 and RAD001 to affect the proliferation, migration, and invasion of human glioma cells. Treatment with either LBT613 or RAD001 alone significantly decreased the proliferation of multiple human glioma cell lines. Furthermore, LBT613 and RAD001 in combination synergized to decrease glioma cell proliferation in association with G1 cell cycle arrest. Glioma invasion is a critical contributor to tumor malignancy. The combination of LBT613 and RAD001 inhibited the invasion of human glioma cells through Matrigel to a greater degree than treatment with either drug alone. These data suggest that the combination of LBT613 and RAD001 reduces glioma cell proliferation and invasion and support examination of the combination of RAF and TOR inhibitors for the treatment of human glioblastoma patients. [Mol Cancer Ther 2007;6(9):2449–57]
Despite advances in the ability to diagnose and treat many cancers, successful treatment of primary brain tumors remains extremely challenging. Even with the most advanced therapies, the median survival of patients with glioblastoma multiforme (WHO grade IV glioma) has reached only 14.6 months (1). This limited survival for glioblastoma patients indicates a need for innovative therapies targeting glioblastomas. Novel therapies are likely to arise from the targeting of multiple molecular pathways altered during glioma formation and progression.
Two signal transduction cascades known to contribute to the formation of glioblastomas are the RAS/RAF/mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K)/AKT/target of rapamycin (TOR) pathways. Both RAS (2) and AKT (3, 4) activities are elevated in glioblastoma patient specimens. Activation of AKT and RAS promotes tumor progression by transducing signals from growth factor receptors to regulate multiple protumorigenic cellular behaviors, including proliferation, survival, and invasion (2, 5–12). Indeed, AKT activation is associated with the conversion of human glioma cells from anaplastic astrocytoma to glioblastoma in a cell culture model (13), and mutations of the AKT inhibitor phosphatase and tensin homologue are observed in glioblastomas but not in low-grade gliomas (14).
Cooperation of the RAS/RAF/MAPK and PI3K/AKT/TOR signal transduction pathways exists during brain tumor formation. Simultaneous enhancement of phosphorylation of MAPK, S6 kinase, and TOR—shared components of these signaling pathways—is associated with poor patient outcome (15). Additionally, a murine brain tumor model shows that the combination of RAS and AKT activation induces tumors in mice, but neither RAS nor AKT alone is sufficient to cause glioblastoma formation (4). This promotion of tumor formation by RAS and AKT is likely due to the combined enhancement of translation of distinct mRNAs (16) to affect multiple protumorigenic cellular behaviors.
The cooperation of RAS/RAF/MAPK and PI3K/AKT/TOR signaling pathways during brain tumor formation suggests that the dual targeting of these signal transduction cascades may be beneficial for patient therapies. Previous results from our laboratory indicate that targeting RAF or TOR activities alone, with the small-molecule inhibitors AAL881 (17) and RAD001 (18), respectively, decreased glioma growth. Other researchers have determined that inhibiting AKT activity with the small-molecule inhibitors KP-372-1 and KP-372-2 decreased glioma cell proliferation due to increased apoptosis (6). A PI3K and TOR inhibitor, PI-103, was also shown to decrease glioma proliferation (19), showing that targeting the PI3K/AKT/TOR pathway at multiple steps inhibits glioma growth. Targeting integrin-linked kinase, a known activator of AKT, in combination with the RAF inhibitor GW5074 also showed a synergistic reduction in the colony-forming ability of glioma cells, which was associated with an increase in apoptosis (20). Based on these data and the known interaction between RAS/RAF/MAPK and PI3K/AKT/TOR signaling, we targeted both RAS and PI3K signaling using a unique combination of small-molecule inhibitors. We combined LBT613, an ATP-mimetic inhibitor of the kinase activity of RAF and vascular endothelial growth factor receptor 2, with the mammalian TOR inhibitor RAD001. We find that the combination of LBT613 and RAD001 significantly inhibits glioma cell proliferation more than either monotherapy. LBT613 and RAD001 synergize to decrease glioma growth through a mechanism associated with arrest of cells in the G1 phase of the cell cycle. LBT613 and RAD001 also cooperate to reduce glioma cell migration and invasion, indicating that the dual targeting of RAS and PI3K signaling has multiple antitumorigenic effects on glioma cell behaviors.
Materials and Methods
Cell Lines and Culture
The human malignant glioma cell lines U87MG and U373MG were purchased from the American Type Culture Collection. U251MG was provided by Dr. Jan Poten (University of Uppsala, Uppsala, Sweden). DNA fingerprinting was done to eliminate the possibility of cross-contamination of U373MG and U251MG cells. D54MG is the Duke University subline of A-172. D54MG, U251MG, U87MG, and U373MG human glioma cells were maintained in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (Sigma-Aldrich).
LBT613 and RAD001 were generously provided by Novartis Institutes for Biomedical Research (Cambridge, MA). Each agent was dissolved in DMSO (Sigma-Aldrich) at a concentration of 10 mmol/L and stored at −80°C. For all experiments, LBT613 and/or RAD001 1000× stocks were added directly to the medium of treated cells to produce the indicated final concentrations of inhibitors.
Western Blotting and Antibodies
Cells were serum starved overnight, pretreated with the indicated concentrations of drug or DMSO control for 2 h, and treated with serum (final 10%) or medium without serum for 15 min. Cells were then lysed in buffer [62.5 mmol/L Tris-HCl, 2% (w/v) SDS, 10% glycerol, 40 mmol/L DTT, protease inhibitors]. Protein content was determined using Bio-Rad Protein Assay Reagent. Equal protein amounts were run on SDS-polyacrylamide gels (Invitrogen), transferred to polyvinylidene difluoride membranes (Millipore), and probed with phosphorylated AKT, total AKT, phosphorylated S6, total S6, cyclin D1 (Cell Signaling Technology), phosphorylated extracellular signal-regulated kinase (ERK) 1/2 (Promega), total ERK1, total ERK2, phosphorylated epidermal growth factor receptor, total epidermal growth factor receptor (Santa Cruz Biotechnology), and α-tubulin (Sigma-Aldrich). Proteins were detected using an enhanced chemiluminescence system (Pierce Biotechnology).
Thymidine Incorporation Assay
Cells were plated into 12-well plates at a density of 2 × 104 per well, serum starved overnight, and treated with DMSO or increasing concentrations of RAD001 and/or LBT613 for 44 h in 10% serum. Cells were labeled for 4 h with 4 μCi [3H]thymidine, fixed in 10% trichloroacetic acid, and lysed in 0.2 N NaOH. [3H]thymidine incorporation into DNA was measured with a scintillation counter.
Combinatorial Index Calculation
Thymidine incorporation data were analyzed using CalcuSyn software (Biosoft) to generate a combination index (CI). A CI greater than one indicates antagonism, a CI of one indicates an additive effect, and a CI less than one indicates synergism.
Flow Cytometric and Annexin V Analysis
For cell cycle analysis, cells were plated into 10-cm dishes at a density of 3 × 105 per plate, serum starved overnight, and then treated for 48 h with DMSO control, 1 μmol/L LBT613, 0.1 nmol/L RAD001, or 1 μmol/L LBT613 and 0.1 nmol/L RAD001. Cells undergoing cell death become detached so conditioned media were collected. Treated cells were trypsinized and added to the conditioned medium. Cells were collected by centrifugation, washed with Dulbecco's PBS (Invitrogen), fixed in ethanol, and stained with propidium iodide. For Annexin V staining, cells were plated into six-well plates at a density of 1.5 × 105 per well and then treated for 48 h with DMSO control, 1 μmol/L LBT613, 0.1 nmol/L RAD001, or 1 μmol/L LBT613 and 0.1 nmol/L RAD001. Condition medium and trypsinized cells were collected, washed with PBS, and incubated with Annexin V stain and propidium iodide as per the manufacturer's instructions (EMD Chemicals). For both cell cycle and apoptosis, analysis was done on FACScan gated to exclude cellular debris and collecting 104 events. Calculations were done using BD software.
Kits were purchased from BD Biosciences and used according to the manufacturer's instructions. Cells were serum starved for 48 h and treated with DMSO control or 1 μL/mL of 1000× RAD001 and/or LBT613 dissolved in DMSO to produce indicated final concentrations of 1 μmol/L LBT613, 1 nmol/L RAD001, or 1 μmol/L LBT613 and 1 nmol/L RAD001. Treated cells were trypsinized, centrifuged, counted, and resuspended in serum-free medium to a concentration of 100,000 cells/mL. Cells were treated with DMSO control or indicated concentrations of RAD001 and/or LBT613, and 500 μL (50,000 cells) were placed in the upper Transwell chambers of inserts uncoated (migration) or coated with Matrigel (invasion). Bottom Transwell chambers were filled with 500 μL medium containing 10% fetal bovine serum as a chemoattractant. After 48 h, inserts were fixed and stained with Diff-Quick Fixative Solutions (Dade Behring, Inc.). Attached cells were imaged with a digital camera mounted to a light microscope (Olympus CK40) and quantified using ImageJ software.8
Significance was determined using GraphPad InStat 3 Software (GraphPad Software, Inc.) with ANOVA analysis except for CI statistics where the one sample t test was used to compare the mean to a hypothetical mean of 1.
LBT613 and RAD001 Effectively Target Different Signal Transduction Pathways in Human Glioma Cells
To validate the abilities of RAD001 and LBT613 to effectively target the RAF and TOR signal transduction pathways, respectively, we examined the phosphorylation of downstream components of each pathway, ERK for RAF signaling and S6 ribosomal protein for TOR signaling, as well as AKT due to the potential for loss of upstream inhibitory feedback loops in the TOR pathway (21). The efficacy of both drugs was evaluated in D54MG, a human glioma cell line previously shown in our laboratory to express B-RAF and C-RAF proteins without mutations (17). AKT signaling is active in D54MG cultures because, like many human gliomas, these cells express a nonfunctional mutant form of phosphatase and tensin homologue, which does not inhibit the PI3K/AKT pathway (data not shown). As expected, serum induced the phosphorylation of ERK and AKT in serum-starved D54MG cells (Fig. 1). Treatment with LBT613 decreased phosphorylation of ERK1 and ERK2 in a concentration-dependent manner with 5 μmol/L LBT613 sufficient to decrease ERK phosphorylation to baseline (Fig. 1A and C). This decrease in ERK phosphorylation was not due to a decrease in total ERK levels (Fig. 1A and C). LBT613 did not affect the phosphorylation or total levels of S6 ribosomal protein but decreased AKT phosphorylation without affecting total AKT levels (Fig. 1A and C). These effects of LBT613 on downstream signal transduction were consistent with the effects of other RAF kinase small-molecule inhibitors, such as AAL881 (17).
In contrast to the effects of LBT613, RAD001 treatment did not alter the phosphorylation or total levels of ERK1, ERK2, or AKT (Fig. 1B and C). RAD001 decreased the activating phosphorylation of the downstream S6 ribosomal protein in a concentration-dependent manner with 0.1 nmol/L sufficient to achieve complete inhibition (Fig. 1B and C). This decrease in phosphorylation was not due to a decrease in total protein levels as both total S6 ribosomal protein and tubulin control were not affected by RAD001 treatment (Fig. 1B and C).
The combination of LBT613 and RAD001 attenuated the activating phosphorylation of ERK1, ERK2, AKT, and S6 ribosomal protein without modulating total protein levels (Fig. 1C). The decrease in target protein phosphorylation was not enhanced by the combination of LBT613 and RAD001, but the use of both small-molecule inhibitors does simultaneously targeted RAF and TOR signal transduction with 1-h pretreatment (Fig. 1C) and may be beneficial long term (Supplementary Fig. S1).9
Supplementary material for this article is available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/)
The Combination of LBT613 and RAD001 Synergistically Decreases Glioma Cell Proliferation
To determine if the simultaneous targeting of RAF and TOR signal transduction could decrease protumorigenic glioma cell behaviors, we first determined the effects of LBT613, RAD001, or the combination of these inhibitors on human glioma cell proliferation. The proliferation of D54MG, U251MG, U87MG, and U373MG human glioma cells was decreased in a concentration-dependent manner by treatment with either LBT613 or RAD001 (Fig. 2). LBT613 (0.5 μmol/L) or RAD001 (0.05 nmol/L) was sufficient to significantly decrease glioma cell proliferation, as measured by thymidine incorporation, in all of the cell lines studied (Fig. 2; *, P < 0.001 by ANOVA compared with control). Every glioma cell line tested also proliferated less when LBT613 and RAD001 were used in combination compared with either monotherapy alone (Fig. 2; *, P < 0.05 by ANOVA compared with either LBT613 or RAD001). However, the concentrations necessary to achieve a combinatorial benefit differed slightly between cell lines. The minimal concentration of the combined inhibitors necessary to inhibit proliferation more than either drug alone was 0.2 μmol/L LBT613 and 0.01 nmol/L RAD001 for D54MG (Fig. 2A) and U87MG cells (Fig. 2C), 0.2 μmol/L LBT613 and 0.05 nmol/L RAD001 for U251MG cells (Fig. 2B), and 0.2 μmol/L LBT613 and 0.1 nmol/L RAD001 for U373MG cells (Fig. 2D). These data show that, although there are differences in the sensitivity of glioma cells to the combination of LBT613 and RAD001, LBT613 and RAD001 cooperate to reduce the growth of multiple human glioma cell lines.
To determine if the cooperation between LBT613 and RAD001 was additive or synergistic, the CI for each combination of drug concentrations was calculated using CalcuSyn software. We found that, in D54MG (Fig. 3A), U251MG (Fig. 3B), and U373MG (Fig. 3C), LBT613 and RAD001 synergized to reduce cell proliferation as indicated by the ability of multiple drug concentration combinations to produce a CI of less than one. The lowest CI, or greatest synergy, was produced with the combination of 0.2 μmol/L LBT613 and 0.1 nmol/L RAD001 in D54MG (CI = 0.22; Fig. 3A) and with 0.5 μmol/L LBT613 and 0.1 nmol/L RAD001 in U251MG (CI = 0.41; Fig. 3B) and U373MG (CI = 0.37; Fig. 3C). Therefore, LBT613 and RAD001 synergize to reduce the proliferation of human glioma cells.
The Combination of LBT613 and RAD001 Significantly Increases the Proportion of Cells in G1 Arrest More than Either Monotherapy
To determine the mechanism by which glioma cell proliferation was inhibited by LBT613 and RAD001, we compared the cell cycle of control and treated cells. LBT613 (1 μmol/L) or RAD001 (1 nmol/L) significantly increased the percentage of D54MG (Fig. 4A), U251MG (Fig. 4B), and U87MG (Fig. 4C) cells in the G1 phase of the cell cycle (*, P < 0.001 by ANOVA compared with control). The combination of 1 μmol/L LBT613 and 1 nmol/L RAD001 further increased the percentage of cells in the G1 phase of the cell cycle (‡, P < 0.05; #, P < 0.001 by ANOVA compared with LBT613 or RAD001 alone). The increase in G1 cell cycle arrest on LBT613 and RAD001 resulted in a significant decrease in the percentage of cells in the S and G2 (Fig. 4A–C) phases of the cell cycle and may be due, in part, to a decrease in expression of cyclin D1 (Fig. 1C). Thus, LBT613 and RAD001 reduced glioma cell proliferation by inducing a G1 cell cycle arrest that prevents continuation through the cell cycle.
In the cell cycle analysis, we also observed no significant change in the sub-G0 fraction of cells (Fig. 4) with LBT613 and RAD001 treatment, indicating there was no increase in the percentage of dead cells as detected with propidium iodide alone. To more fully determine if LBT613 and RAD001 could increase apoptosis, we examined the percentage of Annexin V–positive cells with monotherapy and the drug combination. Apoptosis increased with the combination of LBT613 and RAD001 compared with control in D54MG and U87MG (Fig. 5A and C; *, P < 0.01), whereas there was no effect in U251MG (Fig. 5B; P = 0.42). These data indicate that, although there is an increase in G1 arrest and a decrease in proliferation with the combination of LBT613 and RAD001 across all glioma cell lines tested, there was not a similar broad increase in apoptosis at an identical concentration and time course of inhibitor treatment.
The Combination of LBT613 and RAD001 Significantly Decreases Glioma Migration and Invasion
To further investigate the effect of targeting RAF and TOR signals on protumorigenic glioma cell behaviors, we measured the effect of LBT613 and RAD001 on glioma cellular migration and invasion. In Boyden chamber assays, the combination of 1 μmol/L LBT613 and 1 nmol/L RAD001 inhibited migration in D54MG (Fig. 6A), U251MG (Fig. 6B), and U373MG (Fig. 6C) cells (**, P < 0.01; ***, P < 0.001 by ANOVA compared with control). The combination of LBT613 and RAD001 inhibited invasion in all of the cell lines tested (Fig. 7; *, P < 0.05; ***, P < 0.001 by ANOVA compared with control). The decrease in glioma cell migration with LBT613 and RAD001 was significantly different from either monotherapy alone in U251MG, one of the three glioma cell lines tested (Fig. 6B; #, P < 0.05 by ANOVA compared with LBT613 or RAD001). However, the combination decreased invasion more than either monotherapy in D54MG and U373MG, two of the three glioma cell lines tested (Fig. 7A and C; #, P < 0.05 by ANOVA compared with LBT613 or RAD001). These data showed cell line–specific differences in the extent of the response to LBT613 and RAD001 treatment compared with either monotherapy. However, all of the human glioma cell lines displayed decreased migration and invasion on treatment with the combination of LBT613 and RAD001, showing that the combinatorial therapy is likely to be beneficial for reducing glioma cell movement.
As our understanding of the molecular mechanisms involved in brain tumor formation increases, so will our potential avenues for therapeutic intervention. For example, elucidation of the cooperation between RAS and PI3K signaling in the formation of glioblastoma (4, 16) suggests that dual targeting of these signal transduction cascades could be beneficial for patient therapy. Therefore, we determined the efficacy of combining inhibitors of RAF and TOR, proteins downstream of RAS and PI3K, respectively, on glioma growth and invasion.
The proliferation of each glioma cell lines tested synergistically decreased with the combination of the RAF inhibitor LBT613 and the TOR inhibitor RAD001, and this decrease in growth was associated with a G1 cell cycle arrest. However, identical treatments with LBT613 and RAD001 did not produce a consistent increase in the percentage of apoptotic cells. Although the combination of LBT613 and RAD001 induced apoptosis in D54MG and U87MG cells, no apoptotic effect was observed in U251MG. When compared with the status of major tumor suppressor genes commonly altered in glioma, these data may indicate that loss of p53 mediates resistance to apoptosis induced by the combination of LBT613 and RAD001. D54MG and U87MG are p53 wild-type, whereas U251MG expresses a mutant form of p53. All of the cell lines have mutated or deleted phosphatase and tensin homologue and the cyclin-dependent kinase inhibitor p16INK4A. Thus, variations in p53 status, but not phosphatase and tensin homologue or p16INK4A, could help to explain the cell type–specific differences in apoptosis in response to LBT613 and RAD001 treatment. This variability in cell survival also likely indicates that tumor responses may differ depending on the genetic mutations, which have accumulated during glioma development and progression.
In addition to its effects on cell growth and survival, the combination of LBT613 and RAD001 decreases glioma cell migration and invasion, suggesting an added benefit of the combination of RAF and TOR inhibitors for the treatment of gliomas. Indeed, the ability to prevent glioma cell invasion and target infiltrative glioma cell growth would represent a potential advance in our ability to treat brain tumors. Unlike most other types of cancer, the morbidity and mortality from glioblastomas comes from local invasion rather than metastasis. This local invasion prevents complete surgical resection as shown by the recurrence of the majority of glioblastomas (80–90%) within 2 cm of the original tumor site (22, 23). As glioma cells frequently extend through much of the neural axis before diagnosis, even local control of invasion will unfortunately not be curative. Therefore, novel therapies targeting the infiltrative glioma cells must be able to act some distance from the surgical site on cells that are disseminated into normal brain. Although it remains to be determined if the combination of RAF and TOR inhibitors could be an anti-invasive therapy, our in vitro data do show that this combination could have benefits for patient therapies beyond a simple reduction of cancer cell proliferation.
Our data strongly suggest that the combination of RAF and TOR inhibitors could reduce glioma cell growth and invasion and that further development of this novel therapeutic paradigm is warranted, including examination of the effects of LBT613 and RAD001 in vivo. Although quantities of these novel compounds are limited, we were able to do a preclinical animal trial of LBT613 and RAD001 in a D54MG xenograft model. Although there were two deaths in a total of 10 animals attributed to toxicity, treatment of immunocompromised mice with established tumors with 100 mg/kg LBT613 and 15 mg/kg RAD001 significantly delayed the growth of the tumors by 6.7 days (P < 0.001 by the Wilcoxon rank-sum test) as assessed by the average difference in days required for tumors to reach a five times the initial tumor volume. Although these in vivo data further support further examination of the use of LBT613 and RAD001 in combination for glioma therapy, it is too early to determine if the in vitro antitumorigenic effects we have observed across multiple cell lines will translate into a reduction in the growth of multiple xenografts in vivo. However, the use of LBT613 and RAD001 in combination to reduce glioma growth and invasion seems promising, and our data suggest that glioma cell growth could be decreased by other combinations of available novel small-molecule inhibitors that target different components of the RAS/RAF/MAPK and PI3K/AKT/TOR pathways.
Grant support: Pediatric Brain Tumor Foundation of the United States (J.N. Rich); Accelerate Brain Cancer Cure (J.N. Rich); Childhood Brain Tumor Foundation (J.N. Rich); Southeastern Brain Tumor Foundation (A.B. Hjelmeland); and NIH grants NS047409, NS054276, CA108786, and CA116659 (J.N. Rich). A.B. Hjelmeland is a Paul Brazen/American Brain Tumor Association Fellow. J.N. Rich is a Damon Runyon-Lilly Clinical Investigator supported by the Damon Runyon Cancer Research Foundation and a Sidney Kimmel Foundation for Cancer Research Scholar.
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.