Purpose: Glioblastoma multiforme is the most frequent primary brain tumor in adults and represents one of the most lethal malignancies with a median survival of 12-16 months. We have previously shown that an oncolytic measles virus derivative expressing soluble human carcinoembryonic antigen (MV-CEA) has significant antitumor activity against glioblastoma multiforme cell lines and xenografts. Radiation therapy (RT) represents one of the mainstays of glioma treatment. Here we tested the hypothesis that the combination of RT with MV-CEA would have synergistic activity against gliomas.
Experimental Design: 3-(4,5-Dimethyl-thiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) and clonogenic assays were used to test cytoxicity of the combination treatment in vivo. To examine the mechanism of synergy, one-step viral growth curves, terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assays, and Western blot analyses were performed. In vivo assessment of synergistic antitumor activity was conducted in a U87 glioma model.
Results: MTS and clonogenic assays showed a strong synergistic interaction between MV-CEA and RT in glioblastoma multiforme cells including both primary and established glioma lines. Furthermore, significant antitumor efficacy was observed in vivo in a subcuteneous U87 xenograph model. There was significant prolongation of survival (P = 0.001) in the combination treatment group as compared with single modality– or control-treated animals. One-step viral growth curves showed increased viral burst size by up to 2 log in MV/RT combination–treated cells, as compared with single agent MV-CEA–treated glioma cells. Changes in CEA levels and expression of viral N and H protein were also consistent with increased viral production. Furthermore, TUNEL assays and Western blot analysis showed increase in apoptosis in MV/RT combination–treated cells. The pan-caspase inhibitor Z-VAD-FMK and the caspase-8 inhibitor Z-IETD-FMK, but not the caspase-9 inhibitor Z-IEHD-FMK, protected glioma cells from MV-CEA/RT–induced cleavage of poly(ADP-ribose) polymerase (PARP), indicating that the apoptotic death in combination-treated cells is mostly mediated via the extrinsic caspase pathway. The Fas/Fas ligand interaction blocking antibody NOK-1 blocked MV/RT–induced PARP cleavage whereas the Fas agonistic antibody CH11 increased PARP cleavage in MV/RT combination–treated cells. Reverse transcription-PCR, fluorescence-activated cell sorting analysis and immunohistochemistry showed up-regulation of Fas in combination-treated tumor in vitro and in vivo cells.
Conclusions: There is synergy between MV-CEA and RT in vitro and in vivo. The synergistic effect of the combination seems to be due to increase in viral burst size and increase in apoptotic cell death. This latter effect is mostly mediated via the extrinsic caspase-8 pathway, activated via increased signaling through the Fas death receptor pathway. These results could have translational implications in glioma therapy.
Glioblastoma multiforme is the most common primary brain tumor in adults and remains lethal with a median survival of 12 to 16 months despite the use of multimodality treatment including surgery, chemotherapy, and radiation therapy (RT; ref. 1). RT is one of the main therapeutic modalities in the treatment of glioblastoma multiforme. Nevertheless, the majority of glioblastoma multiforme recur within the radiation field (2–4). As recently shown for the combination of RT and temozolomide in newly diagnosed glioblastoma multiforme patients (1), combinations of RT with active novel therapeutic agents could hold significant promise for improving the treatment outcome in glioblastoma multiforme patients.
Given their limited ability to metastasize, gliomas represent a promising target for gene transfer and virotherapy approaches (5, 6). Measles virus is an RNA virus that belongs to the family of Paramyxoviridae (7). Tumor cells including glioblastoma cells overexpress the measles virus receptor CD46, which allows for preferential tumor targeting (8, 9). We have previously shown that oncolytic measles virus derivatives have significant antitumor activity in glioma lines and xenografts (8). Using a measles virus derivative engineered to overexpress the human carcinoembryonic antigen (MV-CEA), we have shown that MV-CEA infection of glioma cells in vitro or in xenograft models resulted in the characteristic cytopathic effect [i.e., formation of multinuclear cell aggregates (syncytia) followed by apoptotic cell death and prolongation of survival in subcutaneous and orthotopic tumor models; ref. 8]. Moreover, CEA expression levels during viral replication represent a good correlate of viral gene expression (8, 10). On the basis of these promising preclinical data, MV-CEA is currently in clinical testing in recurrent glioblastoma multiforme patients.
Combining measles virus with external beam RT is appealing because both treatment modalities can be used in a locoregional manner to improve local control, the main mode of glioblastoma multiforme relapse, and possibly therefore outcome. Given the significant preclinical activity of measles virus derivatives in glioma models and the major role of apoptosis in measles virus– and RT-induced glioma cell death, we hypothesized that ionizing RT and MV-CEA virotherapy combination treatment would significantly augment the antitumor potential of either modality. We showed that oncolytic measles virus derivatives in combination with radiotherapy have synergistic activity against glioma cell lines and xenografts, which is due to an increase in viral bust size and apoptotic death. The latter is predominantly mediated via the extrinsic caspase-8 pathway through Fas signaling. These findings could have immediate translational implications in the treatment of gliomas.
Materials and Methods
Cell lines. Human malignant glioma cell lines U87 and U251 were obtained from American Type Culture Collection and were cultured in DMEM supplemented with 10% fetal bovine serum. The primary glioblastoma multiforme line GBM12 used in this study is derived from a glioblastoma patient, and it is radiation resistant (11, 12). GBM12 cells were kept and passaged s.c. in the flank of BALB/c nude mice. Two to three weeks before the in vitro or in vivo experiments, freshly isolated glioblastoma multiforme cells from the host mice were expanded in MEM with 2.5% fetal bovine serum for 2 weeks followed by culture in DMEM with 10% fetal bovine serum for 1 week before being used for in vivo or in vitro experiment (11).
3-(4,5-Dimethyl-thiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium assays. 3-(4,5-Dimethyl-thiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega) was done following manufacturer's instructions. Glioma cells were seeded into 96-well plates (14) at a density of 4,000 per well for U87 and U251 and at 12,000 per well for GBM12 1 day before treatment initiation. Dose-response curves were generated using a range of viral multiplicities of infection (MOI) from 0 to 1.0 and different RT doses from 0 to 15 Gy, delivered with a 137Cs irradiator. Each treatment condition was tested in five replicate wells and the cell proliferation and viability were measured at days 0 to 7 after administration of RT. Doses resulting in 60% to 80% cell viability were chosen for subsequent combination experiments, and the following sequences were tested: (a) RT treatment followed 24 h later by MV-CEA infection; (b) MV-CEA infection followed 24 h later by RT treatment; and (c) RT and MV-CEA administered concomitantly.
Results were analyzed using the CalcuSyn software program (Biosoft) based on the median effect method of Chou and Talalay (15). Combination indices (CI) were used to assess the interaction between the two treatment modalities (15).
Clonogenic assays. Clonogenic assays were done using the optimal MV-CEA/RT sequence as determined by MTS assays. Glioma cells were seeded into 96-well plates at a density of 4,000 per well 1 day before treatment. RT doses and MV-CEA MOIs were similar to those used in MTS assays. Cells were infected with MV-CEA 24 h before RT. Six hours after the radiation treatment, the cells were transferred into six-well plates at densities varying from 500 to 50,000 per well (U251 cells) to 1,000 to 30,000 per well (U87 cells) and incubated at 37°C with 5% CO2 incubator for 10 to 14 days. Thereafter, the cells were fixed with 0.5% glutaraldehyde and stained with 0.025% crystal violet. Colonies consisting of >50 cells were counted and the colony formation (16) for each condition was presented in relation to values obtained from untreated control cells. Each treatment condition was tested in four replicate wells, and each experiment was repeated twice.
Determination of CD46 expression levels. One million U87, U251, or GBM12 cells were harvested before and at 24, 48, and 72 h after RT, washed, and then incubated with FITC-labeled mouse anti-human CD46 (PharMingen). Washed cells were fixed in PBS containing 0.5% paraformaldehyde and analyzed on a Becton Dickinson FACScan Plus cytometer. Analysis was done using the CellQuest software (BD Biosciences).
One-step viral growth curves. Virus titers of MV-CEA in infected tumor cells with or without the addition of RT treatment were determined as described previously (8). Tumor cells were plated in six-well plates at a density of 1 × 105 per well and, 24 h later, were infected with MV-CEA at an MOI of 0.1. The cells were then exposed to a single dose of 137Cs radiation at 7.5 or 10 Gy 24 h after infection and harvested at 24, 72, and 120 h after radiation. The medium was collected for CEA detection. The viruses were released by two cycles of freeze/thawing in liquid nitrogen and 37°C water bath. Viral titers were determined on Vero cells by limited dilution assay as previously described (8).
Western immunoblotting. Glioma cells (2 × 105) were plated into each well of six-well plates 24 h before MV infection. Cells were infected with MV-CEA at MOI 0.1 for 2 h at 37°C, and the medium was changed to fresh growth medium. Cells were irradiated at 7.5 Gy 24 h later, and culture was continued at 37°C for another 72 h. In experiments with caspase inhibitors, fresh medium containing the caspase-8 inhibitor Z-IETD-FMK (Calbiochem), the caspase-9 inhibitor 2-IEHD-FMK (Calbiochem), or the pan-caspase inhibitor Z-VAD-FMK (Promega) at a concentration of 100 μmol/L was added following infection with MV-CEA. In experiments using a Fas agonist or a Fas/Fas ligand interaction antagonist, fresh growth medium with 1 to 5 μg/mL of the Fas/Fas ligand antagonistic antibody NOK-1 (BD PharMingen) or 100 ng/mL of the Fas agonistic antibody CH11 (Upstate) was added following infection with MV-CEA.
The treated cells were lysed with modified radioimmunoprecipitation assay lysis buffer (17). Whole-cell lysates from different treatment groups (20-50 μg) were electrophoresed in 4% to 15% gradient SDS-PAGE gel and were transferred onto the nitrocellulose membrane blot. The membrane blot was then blocked in a casein blocking solution (Vector Laboratories) for at least 45 min at room temperature and followed by incubation with the primary antibodies against measles virus N and H protein (kindly provided by Dr. Roberto Catanneo, Mayo Clinic, Rochester, MN), caspase-3 (Imgenex), caspase-8 (Cell Signaling), caspase-9 (Cell Signaling; ref. 18), or poly(ADP-ribose) polymerase (PARP; Promega; ref. 19) at 4°C overnight. Horseradish peroxidase–conjugated secondary antibodies, mouse anti-rabbit immunoglobulin G (IgG), or goat anti-mouse IgG + immunoglobulin M (Pierce) were applied and incubated for 45 min at room temperature and, subsequently, the membrane was visualized with a chemiluminescent detection kit (Pierce). β-Actin was used as loading control.
Fluorescence-activated cell sorting assessment of Fas expression. U251, U87, and GBM12 cells were plated in six-well plates at a density of 2 × 105 per well and incubated for 24 h in growth medium. The cells were infected with MV-CEA in opti-MEM at MOIs 0.01 to 1.0, as previously described, and then the medium was changed to growth medium with 40 μg/mL of fusion inhibitory peptide (Bachem) to prevent measles virus–induced syncytia. Twenty-four hours after MV infection, the cells were irradiated at 2.5, 7.5, or 10 Gy. Cells (106) collected before and at 24 and 48 h after RT were harvested, washed, and then incubated with FITC-labeled mouse anti-human CD95/Fas (DX2) or FITC-labeled mouse anti-human CD46 (PharMingen). FITC-labeled mouse IgG1κ and FITC-labeled mouse IgG2aκ (PharMingen) served as the isotype controls. Washed cells were fixed in PBS containing 0.5% paraformaldehyde and analyzed on a Becton Dickinson FACScan Plus cytometer. Analysis was done using the CellQuest software (BD Biosciences).
Reverse transcription-PCR. Total RNA was isolated from untreated and RT-, MV-CEA–, and MV/RT–treated glioma cells with RNeasy mini kit (Qiagen) following the manufacturer's instructions. cDNA was synthesized from 0.5 μg of total RNA using the First-Strand synthesis system (Invitrogen). A 426-bp Fas receptor fragment was amplified in 40 cycles using PCR Supermix (Invitrogen) primers 5′-ACTTGGGGTGGCTTTGTCTT-3′ (forward) and 5′-GGATGATAGTCTGAATTTTCTCTG-3′ (reverse; ref. 20) and the following cycling conditions: 4 min at 94°C, then 40 cycles of 45 s at 94°C (denaturation), 30 s at 58.6°C (annealing), and 60 s at 72°C (extension), followed by 5 min at 72°C. A 251-bp fragment of the human glyceraldehyde-3-phosphatase dehydrogenase gene (GAPDH) was amplified with forward primer 5′-CCATGTTCGTCATGGGTGTGAACCA-3′ and reverse primer 5′-GCCAGTAGAGGCAGGGATGATGTTC-3′ (21) and used as loading control. PCR products were separated on a 2% agarose gel.
Assessment of apoptosis by terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) assay. Glioma cells were plated into two-well culture slides (Becton Dickinson) at a density of 1 × 105 per well. The cells were then treated with MV-CEA alone, RT alone, or MV-CEA/RT combination at different MV-CEA MOIs and RT doses, respectively. Cells were analyzed for the presence of apoptosis using the In Situ Cell Death Detection kit, TMR red (Roche) at 48, 96, and 144 h postinfection. Air-dried cells were fixed with 4% paraformaldehyde, washed with PBS, permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate, and incubated with TUNEL reaction mixture (consisting of calf thymus terminal deoxynucleotidyltransferase and nucleotide mixture in reaction buffer) for 60 min at 37°C. Apoptotic cells were visualized by fluorescence microscopy. Infected cells treated with reaction mixture, from which terminal deoxynucleotidyltransferase was omitted, were used as negative control. Noninfected cells, exposed to DNase I for 10 min before the addition of the TUNEL reaction mixture, were used as positive control.
In vivo antitumor efficacy. All animal protocols were approved by the Mayo Clinic Institutional Animal Care and Use Committee. U87 cells (6 × 106) in 100-μL PBS were mixed with 100 μL of Matrigel (BD Biosciences) and s.c. injected into the right flanks of 5-week-old BALB/c nude mice with a 27-gauge needle. The mice were randomized into four groups of 9 to 10 mice each: RT, MV-CEA, MV-CEA/RT, and UV-inactivated MV-CEA (as control). Seven days after tumor implantation, 6 × 106 TCID50 of MV-CEA in 200-μL opti-MEM medium were administered i.v. (via the orbital vein) to the animals in MV-CEA and MV-CEA/RT groups every 2nd day for a total of six doses (total MV-CEA dose, 3.6 × 107 TCID50). Control group mice received UV-inactivated MV-CEA in a similar time course. In the RT alone and MV-CEA/RT combination groups, mice received 2 Gy every other day for a total of six doses (12 Gy), starting 24 h after the MV-CEA administration in the MV-CEA/RT combination group. The radiation was restricted to the tumor site by using a lead block to shield the remainder of the animal body. Mice were observed thrice a week following tumor implantation and the tumor size was measured with a caliper. Tumor volumes were calculated as length × width2 / 2 (8). As per the Mayo Institutional Animal Care and Use Committee guidelines, mice were euthanized when the tumor size exceeded 1.2 cm in diameter or when tumor ulceration occurred. Tumor samples were frozen or paraffin embedded for further histologic analysis. In addition, one animal per group was euthanized 24 h after treatment completion. Each tumor sample was split into two halves. One part was embedded in paraffin for histopathologic analysis, whereas the other half was minced in smaller pieces and used for Vero cell overlay. for the overlays, Vero cells were plated into six-well plates at 5 × 105 per well 24 h before the overlay. Minced tiny pieces of the tumor in 0.5-mL growth medium were added to the monolayer of Vero cells. The overlaid cells were incubated at 37°C and observed for up to 2 weeks for syncytia formation.
Analysis of tumor tissue for Fas expression (immunohistochemistry). Paraformaldehyde-fixed, paraffin-embedded tumor tissue samples of U87 xenografts were deparaffinized, rehydrated, antigen retrieval citra (BioGenex) treated, and blocked with 10% normal goat serum (Vector Laboratories) in PBS, and subsequently incubated with the respective primary antibody (IHC-Mouse anti-Fas IgG, B-10, Santa Cruz Biotechnology) in PBS containing 1.5% to 3% normal goat serum for 45 min at room temperature. The secondary antibody, biotinylated anti-rabbit-mouse-IgG (Vector Laboratories), was applied at 1:100 dilution and incubated for 45 min at room temperature, followed by incubation with the ABC reagent (Vector Laboratories) for 30 min (22). Antibody binding was visualized with a 3,3′-diaminobenzidine solution (Vector Laboratories) and tissues were counterstained with hematoxylin.
Statistical analysis. Cumulative survival probabilities for each group (UV-inactivated MV-CEA, MV-CEA, RT, and MV-CEA/RT) were estimated with the Kaplan-Meier method. The log-rank test was used to compare survival of groups of animals. Repeated-measures ANOVA with an autoregressive correlation structure was used to compare tumor volume over time across the four groups. For this model, tumor volume was the dependent variable and group was the independent variable. In all case, P < 0.05 was considered statistically significant.
The combination treatment of MV-CEA/RT has synergistic antitumor effect in glioma cells. To assess the effect of different treatment modalities on cell proliferation and survival, MTS and clonogenic assays were done on U87 and U251 glioma cell lines as well as the primary cell line GBM12. Several viral MOIs and radiation doses were tested in three different glioma lines to determine dose-response curves. Doses of virus and RT resulting in small cytopathic effect (cell viability of 60-80%) were used in subsequent combination experiments. To determine the optimal sequence of the two therapeutic modalities, we tested different sequences of MV-CEA and RT treatment. MV-CEA infection followed by RT 24 h later resulted in the highest cytopathic effect, as compared with the other two combination sequences (i.e., RT followed by MV-CEA infection or concomitant administration of the two modalities; data not shown). Infection of glioma cells with MV-CEA, followed 24 h later by RT, was associated with significant increase of cytopathic effect suggestive of synergy (Fig. 1A). This effect was observed both in radiation-sensitive and radiation-resistant lines. For example, in the radiation-resistant glioma line U87, treatment with single agent MV-CEA alone at an MOI of 0.01 resulted in 70% cell viability at 7 days after treatment (Fig. 1B). Treatment with RT at a dose of 10 Gy resulted in 80% cell viability. After combination treatment, however, at the same MOIs and RT dose, cell survival dropped to only 20%. Similar results were obtained with the radiation-sensitive line U251 and the radiation-resistant primary glioma line GBM12 (Fig. 1B). To evaluate the statistical significance of these results, CIs were calculated using the CalcuSyn software program based on the median effect method of Chou and Tallalay (15). CI <1 indicates synergistic interaction; CI = 1 indicates additive interaction; and CI >1 indicates antagonism between the two modalities. In the MV-CEA/RT combination treatment group, a synergistic cytotoxic interaction between the two treatments was observed in all cell lines and for most combination conditions, as indicated by CIs <1 (Table 1).
|MV/RT (MOI/Gy) .||CIs .||MV/RT (MOI/Gy) .||CIs .||MV/RT (MOI/Gy) .||CIs .|
|MV/RT (MOI/Gy) .||CIs .||MV/RT (MOI/Gy) .||CIs .||MV/RT (MOI/Gy) .||CIs .|
NOTE: CIs <1 were obtained for all three cell lines for most combination conditions, pointing to synergistic interaction of the two modalities.
Clonogenic assays further confirmed the results of the MTS assay. Representative results for U251 and U87 cells are shown in Fig. 1C. The sensitizer enhancement ratio at 10% survival (SER10) was calculated for each combination as the ratio of RT doses required for 90% cell killing following treatment with RT alone versus RT + virus (Fig. 1C; ref. 23). SER10 values >1.5 were observed for MOIs of 0.01 and 0.05 in U87 cells and for MOIs of 0.1 and 0.5 in U251 cells, confirming synergistic activity (Supplementary Table S1).
MV-CEA/RT combination treatment has synergistic antitumor effect in U87 xenograft mouse model. To assess if the in vitro synergy between MV-CEA and RT would persist in vivo, we tested the combination in a U87 xenograft BALB/c nude mouse model. The MV-CEA/RT combination treatment resulted in significant tumor regression and inhibition of tumor growth as compared with mice treated with single modality RT, single agent MV-CEA, or the control group treated with UV-inactivated virus (P = 0.001; Fig. 2A, pairwise comparisons: P = 0.009, MV/RT versus UV-inactivated virus; P = 0.005, MV/RT versus RT; P = 0.006, MV/RT versus MV). Complete tumor regression was seen in five of eight mice in the combination treatment group that persisted at 200+ days, as compared with none of eight mice in each of the other three groups (RT, MV-CEA, and UV-MV-CEA). Kaplan-Meier survival curves were generated for each group and compared using the log-rank test. As shown in Fig. 2B, there is significant prolongation of survival in MV-CEA/RT combination–treated mice as compared with MV-CEA alone–, RT alone–, or UV-inactivated MV-CEA–treated mice (Fig. 2B; P < 0.0001; pairwise comparisons: P < 0.0001, MV/RT versus UV-inactivated virus; P < 0.0003, MV/RT versus RT; P = 0.0012, MV/RT versus MV). Of equal importance, five of eight mice in MV-CEA/RT combination–treated group survived more than 200 days without evidence of tumor recurrence. All of the mice in the other three groups had to be euthanized due to oversized tumors in the first 80 days of the experiment. The median survival of MV-CEA/RT–treated mice was 200+ days, as compared with 56 days in MV-CEA–treated mice, 42 days in RT-treated mice, and 35 days in UV-inactivated MV/CEA–treated mice.
Measles virus was recovered from tumors of animals treated with MV-CEA or MV-CEA/RT on Vero cell overlay, thus indicating that the antitumor effect in MV-CEA/RT combination–treated animals is, at least in part, mediated via viral propagation in the tumor (Fig. 2C).
Radiation increases MV-CEA propagation in glioma cells. To examine the mechanism by which the MV-CEA/RT synergistic effect is mediated, we examined the effect of RT on virus replication by carrying out one-step viral growth curves (8). MV-CEA treatment in combination with RT enhanced the MV-CEA propagation in glioma cells by up to 2 log as compared with MV-CEA treatment alone without RT. Representative results in U87 cells are shown in Fig. 3A. Changes in CEA levels over time in MV-CEA/RT–treated groups paralleled the changes in viral titers (Fig. 3B). Furthermore, Western immunoblotting for viral N and H proteins showed increased viral protein expression in combination-treated cells (Fig. 3C), further supporting increased viral replication in RT-treated cells. This increase in viral burst size was not due to increased viral entry resulting from measles virus receptor up-regulation. Indeed, fluorescence-activated cell sorting analysis of measles virus CD46 receptor expression in glioma lines following RT showed no significant change in the CD46 expression levels (Supplementary Fig. S1).
The MV-CEA virotherapy/RT combination enhances apoptotic cell death via the extrinsic caspase-8 pathway. We have previously shown that measles virus derivatives kill glioma cells by inducing apoptosis (8). To test the effect of RT on MV-CEA–induced apoptosis in MV-CEA/RT combination–treated cells, TUNEL assays were performed at days 2, 4, and 6 post infection. Increase in TUNEL positivity was observed at days 4 and 6 in the MV-CEA/RT group as compared with either MV-CEA or RT treatment alone (data not shown). To further elucidate the mechanism underlying glioma cell apoptosis, we carried out Western immunoblotting to evaluate caspase-8, caspase-9, and caspase-3 activation and PARP cleavage at days 2, 4, and 6 postinfection. There was increase in cleaved caspase-8 and PARP following MV-CEA/RT combination treatment (Fig. 4A and B) on days 4 and 6 after infection, as compared with MV-CEA–treated cells or RT-treated cells, consistent with the TUNEL results. In inhibition experiments, we used the pan-caspase inhibitor Z-VAD-FMK, the caspase-8 inhibitor Z-IETD-FMK, and the caspase-9 inhibitor Z-IEHD-FMK to block specific caspase signaling pathways. As shown in Fig. 4C, complete inhibition of PARP cleavage in MV alone–treated groups and almost complete inhibition of PARP cleavage in MV/RT–treated cells were observed after the addition of pan-caspase inhibitor or caspase-8 inhibitor, but not caspase-9 inhibitor, indicating a predominantly caspase-8–dependent activation of apoptosis in combination-treated cells.
To further investigate the mechanism of activation of the extrinsic caspase signaling pathway in combination-treated cells, we investigated the Fas/Fas ligand interaction, which represents a key death signaling pathway implicated in the apoptotic death of hematopoietic cells after measles virus infection (24–26). Fas up-regulation has also been observed after sublethal doses of ionizing radiation (27, 28). By Western immunoblotting, we showed that the anti-CD95 antibody (CH11), a Fas agonist, significantly increased PARP cleavage in MV/RT–treated glioma cells (Fig. 5A). In contrast, the anti-CD95L antibody (NOK-1), which blocks the interaction between Fas and Fas ligand, significantly decreased the MV/RT–induced PARP activation (Fig. 5B). Furthermore, reverse transcription-PCR and fluorescence-activated cell sorting for Fas showed increased Fas expression in combination-treated glioma cells (Fig. 5C and D). Similarly, increase in FAS expression by immunohistochemistry was observed in MV-CEA/RT combination–treated U87 xenografts, as compared with tumors treated with single agent MV-CEA, single modality RT, or UV-inactivated virus. These results support the role of Fas up-regulation in inducing apoptosis in MV-CEA/RT combination–treated tumor cells and xenografts.
Radiotherapy has been used in the treatment of gliomas for more than 30 years (29, 30) and represents a key therapeutic modality in the management of these tumors. Nevertheless, more than 90% of gliomas recur within the radiation field (2, 3).
Gliomas represent a good target for gene transfer approaches, given their limited ability to metastasize; however, despite promising preclinical data, significant clinical benefit has not been shown to date. Development of novel, more potent viral agents may represent a possible solution to this problem. Oncolytic measles virus derivatives have such a potential; we have previously shown significant antitumor activity in several glioma lines and xenografts (8, 31, 32) and a phase I trial of MV-CEA in patients with recurrent glioblastoma multiforme is ongoing.
Combining external beam ionizing radiation and measles virotherapy is appealing because both modalities can be used in a locoregional manner. Given the fact that most glioblastoma multiforme relapses occur within the radiation field, synergistic combinations of measles virotherapy with RT could lead to more complete and durable tumor reductions and reduce the possibility of emergence of resistant clones.
Our data support the synergistic interaction of the two modalities. In vitro experiments using MTS and clonogenic assays showed synergistic cytotoxicity in different glioma lines, both established, such as U87 and U251 cells, and the primary line GBM12 deriving from a glioblastoma patient and propagated as s.c. xenografts in nude mice. It is of note that two of these lines, U87 and GBM12, are radioresistant in vivo (33), thus emphasizing the potential of the combination of measles virus RT as a means of enhancing the eradication of radiation resistant-glioma clones, an observation with significant practical implications. Furthermore, in vivo experiments in radioresistant U87 xenografts confirmed the synergistic activity with the combination treatment leading to statistically significant tumor regression (P = 0.001) and prolongation of survival (P < 0.0001). Five of the eight animals in the combination-treated group survived more than 200 days without evidence of tumor recurrence, indicating that the combination treatment led to complete tumor eradication. The median survival of MV-CEA/RT–treated mice was 200+ days as compared with 56 days in virus only–treated mice and 42 days in RT-treated mice.
Our data represent the first demonstration of synergy between an RNA oncolytic virus and external beam RT and potentially indicate a therapeutic direction worth exploring with other RNA oncolytic viruses. They also emphasize the potential utility of measles virus derivatives expressing the human thyroidal sodium iodine symporter gene (MV-NIS; ref. 34) in the treatment of gliomas. Cells infected with the MV-NIS measles virus strain express NIS, a membrane channel protein normally expressed on thyroid follicular cells (35). This allows uptake of γ-ray–emitting or β-particle–emitting iodine radioisotopes and could represent a means of intracavitary delivery of ionizing radiation. We are currently in the process of testing the therapeutic efficacy of MV-NIS in glioma models.
Synergistic or superadditive activity when combined with RT has previously been shown for DNA viruses such as replication-selective herpes simplex virus-1 strains (R7020 and G207; refs. 36–39) and the conditionally replicating adenoviruses ONYX-015, CV706, and AD5Δ24 RGD (40–43) in a variety of tumor models including prostate cancer, hepatocellular carcinoma, cervical cancer, thyroid cancer, and gliomas.
Our initial assessment of the synergistic activity of the measles virus derivative MV-CEA in combination with radiation treatment reveals a few different mechanistic possibilities. First, radiation at the synergistic dose significantly increases viral replication. One-step viral growth curves show that radiation increased the burst size of MV-CEA in glioma cells by up to 2 log without altering the kinetics of viral replication. Detection of the CEA transgene in the supernatant, which represents a good correlative of the viral gene expression, as well as assessment of the expression of the viral proteins H and N, supports the above conclusion. RT induces double-strand DNA breaks, which are rapidly repaired by constitutively expressed DNA repair mechanisms. DNA repair becomes more active in irradiated cells, potentially allowing for greater replication of the episomal viral RNA. Given its very small target size, the measles virus genome is far less likely to sustain radiation-induced damage. Therefore, the more active cellular DNA synthesis machinery in irradiated cells may facilitate viral RNA synthesis and virus replication, as it has previously been shown for DNA viruses (44).
In addition, combination treatment significantly increased apoptotic glioma cell death, as shown by TUNEL assays and Western immunoblotting assessing the caspase activation pathways. Changes in levels of cleaved PARP after treatment with caspase-8, caspase-9, and pan-caspase inhibitors showed that increase in apoptosis is mostly mediated via the extrinsic caspase-8 pathway.
To further investigate the mechanism of this activation, we examined the Fas ligand signaling pathway, given the association of the latter both with measles and RT-induced apoptosis (25–28). Fas (CD95, APO-1) is a 317-amino-acid type I transmembrane glycoprotein with three extracellular cysteine-rich domains that are characteristic of the tumor necrosis factor receptor superfamily. The Fas cytoplasmic portion contains the death domain that rapidly recruits the adapter molecule Fas-associated death domain protein and the caspase-8 enzyme after binding of Fas ligand or agonistic antibodies, leading to caspase-8 activation and apoptosis (45–54). Fas loss of function or expression has been reported to accompany malignant phenotypes (55). Measles virus infection has previously been shown to result in Fas up-regulation in peripheral blood mononuclear cells (25) and dendritic cells (24), both in cell culture and in measles patients (26). Fas up-regulation is also a significant determinant of lymphopenia observed in the context of natural measles virus infection (26). Furthermore, it has been shown that sublethal doses of ionizing radiation result in Fas up-regulation (27, 28). Our data support a significant role of the Fas signaling pathway in inducing apoptotic cell death via activation of the extrinsic caspase pathway in measles/RT–treated glioma cells. Incubation with a Fas agonistic antibody significantly increased PARP cleavage in combination-treated cells. In contrast, the NOK-1 antibody that blocks the Fas/Fas ligand interaction significantly decreased PARP activation in combination-treated cells. Reverse transcription-PCR and fluorescence-activated cell sorting analysis for Fas expression showed up-regulation of Fas in MV/RT–treated tumor cells. Furthermore, increased FAS expression was shown in MV-CEA/RT–treated U87 xenografts.
In this work we demonstrated that oncolytic measles virus derivatives in combination with radiation have synergistic activity against glioma cell lines and xenografts, a finding with potential translational implications. Although low levels of the measles virus CD46 receptor are also expressed in normal brain, this does not limit the potential of the virus as a therapeutic agent for central nervous system malignancies. Indeed, a certain receptor level threshold is required for fusion to occur (9, 56). Furthermore, toxicology studies involving central nervous system administration of MV-CEA in Rhesus macaques, the prototype animal model for measles neurotoxicity, have shown the safety of MV-CEA administration in the central nervous system, with no clinical neurotoxicity or encephalitis being observed (57).
We are currently in the process of testing the safety and efficacy of MV-CEA in a phase I trial for patients with recurrent glioblastoma multiforme. The results presented here, in combination with the safety and efficacy data resulting from the ongoing phase I trial, could serve as a rational basis for translation of this novel potent combination approach in the treatment of glioblastoma multiforme patients.
Grant support: National Brain Tumor Foundation (E. Galanis), NIH Specialized Program of Research Excellence grant P50 CA 108961 (E. Galanis, J.N. Sarkaria, and W. Wu), and R21 CA123839 (E. Galanis).
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).