Mouse 005 GSCs, a gift from Dr I Verma (Salk Institute, San Diego), were established from GBM generated with lentiviral transduction of H-Ras and activated Akt in Cre-GFAP/p53^+/− mice of a somewhat mixed (C57BL/6 and some FVB/N) background.32 They were cultured as spheres in serum-free stem cell medium composed of advanced DMEM/F12 (ThermoFisher Scientific), supplemented with 2 mM L-glutamine (Corning), 1% N2 supplement (ThermoFisher Scientific), 2 µg/mL heparin (Sigma-Aldrich), 0.5% penicillin G-streptomycin sulfate-amphotericin B complex (Corning), recombinant human EGF (20 ng/mL; R&D Systems), and recombinant human FGF-basic (20 ng/mL; Peprotech), as described,13 14 and dissociated with Accutase (Innovative Cell Technologies) for passaging. Low-passage cells were used, which were free from mycoplasma (LookOut mycoplasma kit; Sigma).

G47Δ-mIL12 was constructed from G47Δ (deletions in γ34.5 and α47 genes and inactivating insertion of LacZ into ICP633) by insertion of CMV IE promoter-driven mouse IL-12 cDNA,13 and grown in Vero cells after low multiplicity of infection (MOI), and purified as described.34

O6-BG (Sigma-Aldrich) and TMZ (Sigma-Aldrich) were dissolved in dimethyl sulfoxide (DMSO) for in vitro studies, and diluted in 005 GSC culture medium before adding to cells.

Dissociated mouse 005 GSCs were seeded into 96-well cell culture plates (2000 cells/well). For virus studies, cells were plated using 005 medium with ‘no heparin’, and TMZ and/or virus, diluted in 005 medium without heparin, added to cells at indicated doses immediately after seeding. O6-BG (5 µM) was added to cells 1 hour prior to TMZ treatment. Two hours post-treatment, medium ‘with heparin’ was added and incubated for 4 days at 37°C before MTS assays were performed following manufacturer’s instruction (Promega). Each experiment was repeated at least two independent times and performed in triplicate.

shRNA-mediated knockdown of MSH6 in 005 GSCs was performed as previously described.24 Briefly, plasmid construct containing shRNA sequences against MSH6 mRNA (TRCN0000071163, designated shRNA 1; TRCN0000071164, designated shRNA 2; TRCN0000071165, designated shRNA 3; TRCN0000071166, designated shRNA 4; TRCN0000071167, designated shRNA 5) were purchased from Dharmacon or non-targeting shRNA (SHC002) from Sigma. Generation of lentiviral constructs, lentiviral transduction of 005 GSCs for MSH6 knockdown, and selection with puromycin were performed as previously described.24 MSH6 knockdown was performed at least two times and the level of target gene was assessed by western blot. MTS cytotoxicity assays in MSH6 knockdown cells were performed as described above.

C57BL/6 mice (7–8 weeks old) were obtained from the National Cancer Institute (Frederick, Maryland). All mouse procedures were approved by the Institutional Animal Care and Use Committee at the Massachusetts General Hospital. Dissociated 005 GSCs (2×10⁴) were implanted stereotaxically into the striatum (2.2 mm lateral from Bregma and 2.5 mm deep) on day 0 to generate orthotopic intracranial tumors. Mice were randomly divided into groups at day 7 and treated with G47Δ-mIL12 (5×10⁵ pfu in 2 µL/mouse) or phosphate buffered saline (PBS) injected intratumorally at the same stereotaxic coordinates on day 12, and TMZ (7.5 mg/kg dissolved in 0.93% DMSO or 25 or 50 mg/kg dissolved in 3.1% DMSO) or vehicle solution injected intraperitoneally from days 10 to 14. For O6-BG+TMZ+Virus combination studies, O6-BG (0.3 mg/mouse dissolved in 40% polyethylene glycol-400 (Sigma) in PBS) or vehicle solution was injected intraperitoneally 1 hour prior to TMZ administration (as in Kanai et al 24). Mice were followed for neurological symptoms and euthanized before becoming moribund. Animal caretakers were blinded to the treatment knowledge.

C57BL/6 mice implanted with 005 GSCs were treated with TMZ from days 19 to 23. On day 26 mice were sacrificed, brains removed and fixed in 10% formalin, embedded in paraffin, and 5 µm sections subjected to immunohistochemical staining with primary antibodies against CD4 (anti-mouse CD4; eBioscience, Cat. # 14-9766-80), CD8 (anti-mouse CD8a; eBioscience, Cat. # 14-0808-80), or CD68 (anti-CD68; Abcam, Cat. # ab125212), followed by incubation with appropriate secondary antibodies (HRP anti-rat or anti-rabbit IgG; Vector Laboratories), as described previously.16 35 The number of positive cells were counted from three random fields/tumor section (one section/mouse). Counter was blinded to the treatments.

For 10-color flow cytometric analysis, single cell suspensions from harvested spleens were prepared (as in Cheema et al 13) and stained with fluorochrome-conjugated anti-mouse antibodies (PerCP-Cy-5.5 anti-mouse CD4, PE-Cy7 anti-mouse CD69, Alexa Fluor 647 anti-mouse FoxP3, Brilliant Violet 510 anti-mouse CD8a, Brilliant Violet 421 anti-mouse NK 1.1, Brilliant Violet 605 anti-mouse/human CD11b, APC-Cy7 anti-mouse CD11c, Alexa Fluor 700 anti-mouse Ly-6G/Ly-6C Gr-1, and FITC anti-mouse CD19), as well as appropriate isotype control antibodies, as described.14 36 All antibodies were obtained from Biolegend. Zombie UV live/dead fixable viability kit was used to stain dead cells. We followed a ‘no-wash’ sequential staining protocol (Biolegend) to stain dead cells and for surface staining. Intracellular FoxP3 staining was performed following the FoxP3 intracellular staining protocol (Biolegend). Fluorescent minus one and single-color compensation controls were included for each color, as we described.14 36 All samples were run in a LSRII flow cytometer (BD Biosciences) and flow cytometric data were analyzed by FlowJo software V.10.5.3 (Tree Star). Technician acquiring and gating the data were blinded to the treatments.

005 GSCs were pelleted and lysed in radioimmunoprecipitation buffer (Boston Bioproducts) with a cocktail of protease and phosphatase inhibitors (Roche). Protein (20 µg) was separated by 4%–15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to polyvinylidene difluoride (PVDF) membranes by electroblotting. Western blotting was performed with primary antibodies to MGMT (Biovision, Cat. # 3820–100) and MSH6 (Cell Signaling Technology, Cat. # 3995). Primary antibodies to β-actin (Cell Signaling Technology, Cat. # 4970) and Vinculin (Cell Signaling Technology, Cat. # 13901) were used as loading controls.

Dose–response curves and IC₅₀ values were calculated using Prism 7 GraphPad software V.7.0e. Cell viabilities at each time point, positive cell counts for each group (immunohistochemistry), and flow cytometry were compared by unpaired two-tailed Student's t-test. One-way analysis of variance followed by Tukey’s multiple comparison test was used to compare bodyweights between indicated treatments. Survival data were analyzed by Kaplan-Meier survival curves, and comparisons were performed by log-rank test. A p value of <0.05 was considered significant. All statistical analyses were performed using Prism 7 GraphPad software V.7.0e.

We first investigated the cytotoxic effects of TMZ and/or G47Δ-IL12 treatments in vitro. 005 GSCs are somewhat sensitive to TMZ (IC₅₀ ~100 µM; figure 1A) and G47Δ-IL12 (IC₅₀=MOI 0.7; figure 1B), as determined by MTS assay. In patients, TMZ treatment results in peak tumor concentrations of <40 µM.37 To evaluate combination effects, 005 GSCs were treated with ~IC₂₀ of TMZ (40 µM), ~IC₁₀ of G47Δ-IL12 (MOI 0.1), or both. Viability of 005 GSCs was significantly reduced after each single treatment compared with mock (p<0.0001), and the combination (TMZ+G47Δ-IL12) was significantly better than either agent alone (p<0.0001, figure 1C). In mouse 005 GSCs, knockdown of the MSH6 gene, which is associated with TMZ resistance in GBM,24 did not alter their sensitivity to oHSV as compared with control shRNA treated 005 cells (figure 1D–E).

Because the combination significantly improved GSC killing in vitro, we hypothesized that the combination would induce superior antitumor efficacy in vivo, as seen with human GSCs in immunodeficient mice.24 TMZ dosing regimens can vary from 75 to 200 mg/m²/day in patients with GBM.37 According to the Medscape body surface area-based dosing conversion, 75 mg/m²/day is equal to ~125 mg/day for 60 kg bodyweight, which is equivalent to 25 mg/kg for a 20 g mouse, according to species to species dosing conversion described in Freireich et al.38 To test the combination in vivo, mice bearing orthotopic allografts of 005 GSC-derived tumors were treated with intraperitoneal TMZ (25 mg/kg once daily from days 10 to 14 postimplantation) and/or intratumoral G47Δ-IL12 (5×10⁵ pfu at day 12, figure 2A, lower), and animals followed until they became moribund. No extension of survival was seen with TMZ treatment alone (median survival=36 days; p=0.81), while G47Δ-IL12 treatment alone significantly extended survival (median survival=53.5 days) compared with mock (median survival=38 days; p=0.0002). Surprisingly, combination therapy abrogated the efficacy of G47Δ-IL12, with the median survival of the combination group (39.5 days) similar to the mock group (38 days, figure 2A). We followed bodyweight to evaluate treatment toxicity. While there was a dip in median weight at days 13 and 14, after virus treatment, it was not significantly different from Mock (figure 2B).

Low-dose TMZ (7.5 mg/kg) treatment has been shown to enhance oncolytic adenovirus therapy in a relatively immunogenic GL261 orthotopic brain tumor model.25 Therefore, we tested whether low-dose TMZ would improve the therapeutic outcome of G47Δ-IL12 therapy, using the same schema (figure 2C, lower) as in figure 2A. Low-dose TMZ did not improve survival alone or when used in combination with G47Δ-IL12 (figure 2C), but did not abrogate the therapeutic efficacy of G47Δ-IL12, as with 25 mg/kg (figure 2A).

To try to understand how TMZ is altering the immune response, we examined the effects of TMZ locally on brain tumor-infiltrating immune cells and systemically in the spleen. Mice bearing 005 GSC-derived tumors were treated with TMZ (7.5 mg or 25 mg/kg) from days 19 to 23 postimplantation, and brains and spleens collected 3 days later (day 26). Formalin-fixed paraffin-embedded brain tumor sections were subjected to immunohistochemistry for tumor-infiltrating immune cells, and spleens mechanically dissociated to single cells for multicolor flow cytometry (as in Cheema et al 13). In the tumor, low-dose TMZ significantly reduced the number of CD4⁺ and CD8⁺ tumor infiltrating lymphocytes (TILs) by 1.7-fold and 2.7-fold, respectively, and CD68⁺ tumor-associated macrophages (TAMs) by 1.5-fold compared with the mock group (figure 3A–B). Higher dose TMZ, while not significant, also reduced CD4⁺ TILs (1.7-fold) and TAMs (1.3-fold), and CD8⁺ TILs (4-fold) compared with mock (figure 3A–B). In the spleen, only higher dose TMZ significantly reduced total splenocytes (figure 4A) and CD4⁺ T cells (figure 4B). Other immune cells, such as regulatory (CD4⁺FoxP3⁺) and activated T cells (CD4⁺CD69⁺), CD8⁺ T cells, natural killer cells (NK1.1⁺), myeloid cells (CD11b⁺), dendritic cells (CD11b⁺CD11c⁺), monocytes (CD11b⁺Gr1⁺), and B cells (CD19⁺) were not significantly altered by either TMZ dose (figure 4C–J).

005 GSCs express MGMT (figure 5A), which is associated with TMZ resistance.39 40 If MGMT expression in 005 GSCs contributes to TMZ resistance, its inhibition should improve efficacy.24 O6-BG, an inactivating pseudosubstrate of MGMT, significantly reduced MGMT expression (figure 5A), as seen previously in oral cancer cells,41 and sensitized 005 GSCs to TMZ, shifting the IC₅₀ by ~5-fold (figure 5B). O6-BG also sensitized 005 GSCs to the combination treatment (TMZ+G47Δ-IL12) by ~2 fold compared with no O6-BG (figure 5C). Because O6-BG significantly reduced the IC₅₀ dose of TMZ in vitro (figure 5B) to a dose within the peak tumor concentrations in patients,37 we hypothesized that O6-BG treatment would improve the efficacy of TMZ treatment alone or in combination with G47Δ-IL12 in vivo. The rationale for the combination with MGMT inhibitor was that TMZ did not inhibit 005 tumor growth alone, and therefore it was possible that TMZ-induced cell death was necessary for immune-mediated effects that might enhance TMZ+G47Δ-IL12 combination effects. Mice bearing 005 GSC-derived tumors were treated with TMZ (50 mg/kg) and O6-BG or vehicle, and/or G47Δ-IL12 as illustrated in the schema (figure 5D). The increased sensitivity of 005 GSCs in vitro to TMZ following O6-BG treatment was reproduced in vivo, with the median survival of the O6-BG+TMZ treatment group marginally, but significantly longer than TMZ alone (40 days vs 35 days; p=0.02, figure 5D). As in figure 2A, TMZ combined with G47Δ-IL12 (median survival=42 days) significantly (p=0.0004) abrogated the beneficial effects of oncolytic G47Δ-IL12 treatment (median survival=70 days). Unfortunately, even with the chemotherapeutic effects on tumor growth, the addition of O6-BG to the combination of TMZ+G47Δ-IL12 (median survival=42 days) did not overcome the antagonistic effects of TMZ treatment on oHSV therapy (figure 5D). There was no significant difference in bodyweights between any of the treatment groups and mock (figure 5E).