Oncolytic adenoviral mutants induce a novel mode of programmed cell death in ovarian cancer

Abstract

Oncolytic adenoviral mutants have considerable activity in ovarian cancer. However, the mechanisms by which they induce cell death remain uncertain. dl922-947, which contains a 24 bp deletion in E1A CR2, is more potent than both E1A wild-type adenoviruses and the E1B-55K deletion mutant dl1520 (Onyx-015). We investigated the mode of death induced by three E1A CR2-deleted replicating adenoviruses in models of ovarian cancer and also the importance of E3 11.6 (adenovirus death protein) in determining this mode of death. Ovarian cancer cells were infected with dl922-947 (E3 11.6+) and dlCR2 (E3 11.6-). We also generated dlCR2 tSmac, which also encodes the gene for processed Smac/DIABLO. Classical apoptosis does not occur in adenoviral cell death and there is no role for mitochondria. Expression of Smac/DIABLO does not enhance cytotoxicity nor increase apoptotic features. A role for cathepsins and lysosomal membrane permeability was excluded. Autophagy is induced, but is not the mode of death and may act as a cell survival mechanism. There is no evidence of pure necrosis, while the presence of E3 11.6 does not modulate the mode or extent of cell death. Thus, E1A CR2-deleted oncolytic adenoviral cytotoxicity in ovarian cancer may define a novel mode of programmed cell death.

Figure 1

Activity of E1A CR2-deleted adenoviral vectors in ovarian cancer. (a) IGROV1 and OVCAR4 cells were infected with dl922-947, dlCR2 and dlCR2 tSmac (MOI 0.001-10000 plaque-forming unit (PFU) per cell). Cell viability was assessed 120 h later. Results are presented as percentage cell survival compared to mock-infected cells (mean±s.d. n=3). IC₅₀ results are presented in tabular form. (b) IGROV1 cells were infected with dl922-947, dlCR2, dlCR2 tSmac and Ad LM-X (all MOI 10), mock-infected or incubated with cisplatin (10 μm) in triplicate. After 72 h, caspase activation was measured using PhiPhi Lux-G₁D₂. (c) Following infection with dl922-947 (MOI 0.001-1000), IGROV1 cells were treated with zVAD-fmk (0-25 μm). Cell viability was assessed 144 h later. (d) IGROV1 cells were treated with cisplatin (1 and 10 μm) with and without 25 μm zVAD-fmk. Cell survival was assessed after 72 h by MTT assay. ***P<0.001. (e) IGROV1 cells were harvested up to 120 h following infection with dlCR2 (MOI 10). About 20 μg of protein was separated on sodium dodecyl sulphate (SDS)-PAGE gels and analysed for expression of caspase-3, E1A and PARP by immunoblot.

Figure 2

Markers of classical apoptosis. (a) IGROV1 cells were infected with dlCR2 (MOI 10). Cell surface exposure of PS and loss of mitochondrial membrane permeability were analysed using Annexin V-FITC and TMRE over 120 h. DAPI-positive cells were gated out. (b) The presence of hypodiploid DNA in IGROV1 cells infected with dlCR2 (MOI 10) was assessed following incubation with PI. (c) Markers of classical apoptosis in IGROV1 cells treated with cisplatin (10 μm) for 72 h are shown. PS, phosphatidylserine exposure, MMP, mitochondrial membrane potential loss and sub-G₁, sub-G₁ DNA levels. (d) IGROV1 cells were treated with dl922-947 (MOI 10), cisplatin (3 μm) alone and in combination. After 72 h, cells were harvested and analysed for markers of apoptosis as above. (e) OVCAR4 and OVCAR4-Bcl-2 cells were infected with viruses (all MOI 10) or treated with cisplatin (10 μm). Cell viability was assessed 72 h later. All results represent mean±s.d. n=3.

Figure 3

Assessment of non-classical apoptosis. (a) Following infection with dlCR2 (MOI 10), IGROV1 cells were treated with E64 cysteine protease inhibitor (0-100 μm). Cell viability was assessed up to 144 h later by MTT assay (72 h results shown). Each bar represents normalized survival compared to mock-transfected cells at the same dose of E64; mean±s.d. n=3. (b) IGROV1 cells were infected with dlCR2 (MOI 10) or treated with cisplatin (10 μm). After 72 h, lysosomal stability was assessed following AO staining.

Figure 4

Autophagy. (a) IGROV1 cells were prepared for TEM 72 h following infection with dlCR2 and Ad LM-X (both MOIs 3 and 10). Features of autophagy include increased numbers of lysosomes (closed arrow), the presence of autophagosomes with double membranes (open arrow). (b) IGROV1 cells were infected with dl922-947 (MOI 10) and imaged 48 h later following staining with monodansylcadaverine. (c) IGROV1 cells were co-infected with dl922-947 (MOI 10) and Ad GFP-LC3 (MOI 0.2) and assayed 72 h later for the formation of autophagosomes by confocal microscopy. As positive control, cells infected with Ad GFP-LC3 only were incubated in HBSS for 16 h prior to confocal imaging. (d) (i) IGROV1 cells were infected with dlCR2 (MOI 3 and 10) in the presence of 3-methyladenine (3MA; 0-100 μm). Cell viability was assessed 72 h later. (ii) IGROV1 cells were infected with dl922-947 (MOI 0.03-300) and re-fed with 0-10 μm chloroquine 3 h later. Cell survival was assayed 120 h later and results represent survival normalized for mock-transfected cells at each dose of chloroquine.

Figure 5

Cell architecture and necrosis. (a) Cells were infected with dlCR2 and imaged up to 72 h later under phase contrast microscopy. After 48 h, individual cells show membrane blebbing (solid arrow) and appear rounded (open arrow). By 72 h, widespread cell rounding and detachment is seen, but without disruption to cell membranes. (b) IGROV1 cells were infected with Ad LM-X (MOI 10) or dlCR2 (MOI 3 and 10) and prepared for TEM 72 h later. (i and ii) Mock-infected control cells; (iii) Ad LM-X (MOI 10); (iv and v) dlCR2 (MOI 10); (vi) dlCR2 (MOI 3). C, condensed chromatin; G, disrupted Golgi; M, mitochondria; N, nucleus; V, intra-nuclear virus assembly. (c) IGROV1 cells were infected with Ad LM-X, dlCR2 and dlCR2 tSmac (all MOI 10). Intracellular adenosine triphosphate (ATP) levels were measured 72 h later using a luciferase-based assay as detailed in the ‘Materials and methods’ section. Results represent mean±s.d., n=3.