Combination of adenoviral virotherapy and temozolomide chemotherapy eradicates malignant glioma through autophagic and apoptotic cell death in vivo

Abstract

Conditionally replicative adenoviruses (CRAds) represent a novel treatment strategy for malignant glioma. Recent studies suggest that the cytopathic effect elicited by these vectors is mediated through autophagy, a form of programmed cell death. Likewise, temozolomide (TMZ), a chemotherapeutic agent used for the treatment of malignant gliomas, also triggers autophagic cell death. In this study, we examined the potential to combine the two treatments in the setting of experimental glioma. In vitro, pretreatment with TMZ followed by CRAd-Surivin-pk7 enhanced cytotoxicity against a panel of glioma cell lines. Western blot analysis showed increased expression of BAX and p53, decreased expression of BCL2 and elevated level of APG5. Treatment with TMZ followed by CRAd-Survivin-pk7 (CRAd-S-pk7) led to a significant over-expression of autophagy markers, acidic vesicular organelles and light-chain 3 (LC3). These results were further evaluated in vivo, in which 90% of the mice with intracranial tumours were long-term survivors (>100 days) after treatment with TMZ and CRAd-S-pk7 (P<0.01). Analysis of tumours ex vivo showed expression of both LC3 and cleaved Caspase-3, proving that both autophagy and apoptosis are responsible for cell death in vivo. These results suggest that combination of chemovirotherapy offers a powerful tool against malignant glioma and should be further explored in the clinical setting.

Figures

Figure 1

Temozolomide alone or in combination with CRAd-S-pk7 induces cell death that is additive in effect. (A) Growth-inhibitory effect of TMZ on human glioma cells. Kings, No.10, U87MG and U373MG cells were seeded at 104 cells per well in 96-well plate and incubated overnight at 37°C. After exposure to TMZ (1, 5, 10, 50, 100, 250 and 500 μM) for 120 h, the cells were subjected to cytotoxicity assay by detecting LDH release. Percentage of dead cells from two independent experiments was plotted for a dose-response curve. The four curves represent different sensitivity of human gliomas to TMZ treatments. (B) Comparison of the toxicity mediated by TMZ in the presence or absence of oncolytic virus. Cells were treated only with TMZ or CRAd-S-pk7; pretreated with TMZ and then infected with CRAd-S-pk7; or infected with CRAd-S-pk7 first and then treated with TMZ at doses indicated earlier. *P<0.05 vs CRAd-S-pk7 or TMZ.

Figure 2

Induction of autophagy in U87MG cells treated with TMZ followed by CRAd-S-pk7 infection. Effect of combined treatment of U87MG cells was detected by (A) light microscopy, (B) western blot, (C) membrane potential, (D) flow cytometry and (E) LDH toxicity. (A) Decrease in cell density and morphological changes associated with treatment with TMZ followed by CRAd-S-pk7. (B) Modulation of pro-apoptotic, anti-apoptotic and autophagic proteins in response to treatment with TMZ (100 μM), CRAd-S-pk7 (100 vp per cell) or combination (TMZ and CRAd-S-pk7) as determined by Western blot analysis. Pretreatment of U87MG cells with TMZ followed by CRAd-S-pk7 infection showed over-expression of p53, Bax and APG5 proteins and downregulation of Puma, Noxa, BNIP3 and BCL-2 proteins. There was no evidence of cleaved Caspase-3. (C) To show that mitochondrial pathway is not activated, we measured mitochondrial potential changes. TMZ, CRAd-S-pk7 and combination group did not induce significant changes in Δψ. (D) Autophagy was determined by staining with acridine orange (AO) and α-LC3B antibody followed by flow cytometry analysis. Experiment was performed in triplicates and the mean of two independent experiments is shown here. (E) Effect of Bafilomycin A1 (BAF-A1) and 3-MA treatments on co-treatment induced toxicity. Cells were pretreated with 3-MA, BAF-A1 or vehicle control for 12 h before exposure to TMZ, CRAd-S-pk7 or TMZ followed by CRAd-S-pk7. Figure summarises data from two independent experiments each having six replicates per condition. (*), (**) and (***) P<0.05. (*), (**) and (***) P-value determined by comparing mock vs TMZ alone, CRAd-S-pk7 alone or combination TMZ then CRAd-S-pk7, respectively.

Figure 3

Mitotic catastrophe does not contribute to drug-induced toxicity and cell cycle arrest. U87MG cells were treated with either TMZ (100 μM), CRAd-S-pk7 (100 vp per cell) or combination (TMZ and CRAd-S-pk7). Treated cells were harvested either on day 1, 3, 5 and 10 and subjected to (A) cell cycle distribution analysis, (B) Annexin V/7AAD assay, or (C) western blotting with antibodies recognising Cdc20 protein.

Figure 4

In vivo anti-tumour activity of CRAd-S-pk7 in combination with TMZ. (A) Five consecutive injections of TMZ at 70 mg kg−1 per day (▾) achieve 100% survival for 80 days after U87MG implantation, whereas five TMZ injections at 10 mg kg−1 per day (○) had significant increase in survival compared with mock treatment (•) (P<0.05). (B) Single intracranial (i.c.) injection of CRAd-S-pk7 at 5 × 109 vp per mouse (▵) achieves significant increase in survival compared with mice that received two (○) or one (▾) injection of CRAd-S-pk7 at a dose of 3 × 109 vp per mouse or mock control (•) (P<0.05) (C) Five consecutive injections of TMZ at 10 mg kg−1 per day followed by two CRAd-S-pk7 treatments each at 3 × 109 vp per mouse (Δ) showed significant additive effect on mice survival compared with mock (▪) (P<0.02), double injections of CRAd-S-pk7 each at 3 × 109 vp per mouse (▵) (P<0.02) or five consecutive injections of TMZ at 10 mg kg−1 per day (▾) (P<0.02).

Figure 5

Histopathological analysis of U87MG glioma xenografts. Mice with i.c. xenografts were treated with different regimens and the brain tissue was harvested at specific time points. (A) Sections (at day 41 after tumour implantation) were stained with H&E (a) or with antibodies against Ki-67 (b), LC3 (c), or cleaved Caspase-3 (d). Percent averages of Ki-67 (B) or TUNEL (C) positive cells at different time points (days) after treatment with indicated course. ‘*’ represents P-value <0.05 compared with mock treatment. (D) The polynomial model was used to determine the relationship between the level of LC3 or cleaved Caspase-3 expressions and days after treatment. There was no difference between any of the treatment groups and only a marginal difference in the group shown additive effect (P=0.11). Solid lines represent a model fit for LC3, whereas the dashed line represents a model fit for cleaved Caspase-3.