A hypoxia- and telomerase-responsive oncolytic adenovirus expressing secretable trimeric TRAIL triggers tumour-specific apoptosis and promotes viral dispersion in TRAIL-resistant glioblastoma

Abstract

Glioblastoma is a highly aggressive and malignant type of cancer that is apoptosis resistant and difficult to cure by conventional cancer therapies. In this regard, an oncolytic adenovirus that selectively targets the tumour tissue and induces tumour cell lysis is a promising treatment option. We designed and constructed a hypoxia-responsive and cancer-specific modified human telomerase reverse transcriptase (H5CmTERT) promoter to drive replication of an oncolytic adenovirus (H5CmTERT-Ad). To enhance the anti-tumour efficacy of H5CmTERT-Ad against malignant glioblastoma, we also generated an H5CmTERT-Ad expressing secretable trimeric tumour necrosis factor-related apoptosis-inducing ligand (H5CmTERT-Ad/TRAIL). H5CmTERT promoter-regulated oncolytic adenoviruses showed cancer-specific and superior cell-killing effect in contrast to a cognate control oncolytic adenovirus replicating under the control of the endogenous adenovirus promoter. The cancer cell-killing effects of H5CmTERT-Ad and H5CmTERT-Ad/TRAIL were markedly higher during hypoxia than normoxia owing to hypoxia responsiveness of the promoter. H5CmTERT-Ad/TRAIL showed more potent anti-tumour efficacy than H5CmTERT-Ad did in a xenograft model of TRAIL-resistant subcutaneous and orthotopic glioblastoma through superior induction of apoptosis and more extensive virus distribution in the tumour tissue. Altogether, our findings show that H5CmTERT-Ad/TRAIL can promote dispersion of an oncolytic adenovirus through robust induction of apoptosis in a highly TRAIL-resistant glioblastoma.

Figure 1

Construction and generation of H5CmTERT promoter-regulated oncolytic adenoviruses. (a) A schematic representation of the genomic structures of oncolytic adenoviruses used in this study. Rb7Δ19 contains retinoblastoma (Rb)-binding site-mutated E1A gene and E1B 55 kDa gene, but lacks E1B 19 kDa gene, and E1A expression is controlled by endogenous adenovirus promoter. H5CmTERT-Ad has a backbone similar to that of Rb7Δ19 but replicates under the control of the H5CmTERT promoter rather than the endogenous adenovirus promoter. H5CmTERT-Ad/TRAIL expresses stTRAIL from the adenovirus E3 region of the H5CmTERT-Ad backbone (★ indicates a mutation of the Rb-binding site of E1A). (b) U87MG human glioblastoma cells were infected with Rb7Δ19, H5CmTERT-Ad, or H5CmTERT-Ad/TRAIL at multiplicity of infection (MOI) of 0.5. The concentration of stTRAIL in the culture supernatant was measured by an ELISA. Data are presented as mean ± SD of triplicate experiments. **P < 0.01.

Figure 2

Cancer cell-specific killing effect of the oncolytic adenoviruses. (a) Glioblastoma cells (U87MG and U251N) and (b) normal cells (BJ and SVG-P12) were infected with Rb7Δ19, H5CmTERT-Ad, or H5CmTERT-Ad/TRAIL at the indicated  MOI and incubated under normoxic or hypoxic conditions. At 48 h after the infection, cell viability was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Each cell line was tested at least three times, and data are shown as mean ± SD of triplicate experiments; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3

stTRAIL-mediated potent induction of apoptosis. (a) U87MG or (b) U251N cells were treated with 1 μM of camptothecin (CPT), Rb7Δ19, H5CmTERT-Rd19-Ad, or H5CmTERT-Ad/TRAIL. At 48 h post-treatment, a terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay was conducted. The cells were counted (TUNEL-positive vs total cell count) in five independent fields within microscopic images. The data are representative of three independent experiments conducted in triplicate; *P < 0.05, ***P < 0.001. CPT served as a positive control of induction of apoptosis. A representative field from three independent experiments is shown. Original magnification: ×100 and ×400.

Figure 4

Transmission electron microscopy images of U87MG cells after infection with one of the oncolytic adenoviruses. U87MG cells were treated with PBS, Rb7Δ19, H5CmTERT-Ad, or H5CmTERT-Ad/TRAIL. At 36 h post-infection, the cells were harvested and analysed by transmission electron microscopy. Original magnification: ×5,000 and ×10,000.

Figure 5

Assessment of the anti-tumour effect and survival rate in a subcutaneous glioblastoma xenograft model. U87MG glioblastoma xenograft tumours received three doses of PBS, H5CmTERT-Ad, or H5CmTERT-Ad/TRAIL (5 × 10⁹ VPs) via intratumoural injection on days 1, 3, and 5 (vertical arrows). Tumour growth was monitored at 2-day intervals by measuring the width (W) and length (L) of the tumour. Tumour volume was estimated with the following formula: volume = 0.523 × L × W². *P < 0.05, **P < 0.01.

Figure 6

Histological and immunohistochemical analyses of tumour tissues treated with oncolytic adenoviruses. U87MG tumours established in nude mice were treated with PBS, H5CmTERT-Ad or H5CmTERT-Ad/TRAIL on days 1, 3, and 5, and the tumours were harvested on the 3^rd day after administration of the final treatment. (a) Representative sections were stained with hematoxylin and eosin (H & E). The expression levels of TRAIL and E1A were assessed by immunohistochemical analysis. A TUNEL assay was performed to detect apoptosis. Data are representative of three independent experiments. Original magnification: ×200 and ×400. (b) Semi-quantitative analysis of TRAIL-, TUNEL-, or E1A-stained sections in the MetaMorph image analysis software or ImageJ software. All the results are shown as means ± SD; **P < 0.01, ***P < 0.001.

Figure 7

Anti-tumour effects of oncolytic adenoviruses in an orthotopic glioblastoma tumour model. Orthotopic brain tumour model was established by stereotaxically implanting U87MG cells into the left forebrain of nude mice. Seven days after tumour cell injection, PBS, H5CmTERT-Rd19-Ad, or H5CmTERT-Ad/TRAIL (5 × 10⁹ VPs) were administered via intracranial injection. (a) At 17 days after the cell injection, the mice were euthanised and the brains were collected. Representative tissue slices were stained with H & E to assess the tumour burden in each treatment group in the ImageJ software. Original magnification: ×100. *P < 0.05, **P < 0.01, ***P < 0.001. (b) The expression levels of TRAIL and E1A were assessed by immunohistochemistry. A TUNEL assay was carried out to detect apoptosis. Original magnification: ×400. (c) Semi-quantitative analysis of TRAIL-, E1A-, or TUNEL-stained sections using MetaMorph image analysis software. Images are representative of three independent experiments. All results are shown as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 8

Anti-tumour effects of oncolytic adenoviruses in an U87MG/Fluc orthotopic glioblastoma tumour model. Orthotopic glioblastoma tumour model was established by stereotaxically implanting firefly luciferase-expressing U87MG (U87MG/Fluc) cells into the left forebrain of nude mice. At 7 days after the tumour cell injection, PBS, H5CmTERT-Rd19-Ad, or H5CmTERT-Ad/TRAIL (5 × 10⁹ VPs) were administered via intracranial injection. (a) Bioluminescence whole-body imaging was performed every 7 days following the treatments. (b) Bioluminescence signals were calculated after background subtraction in total flux photons/s from a body region of interest. Data presented as mean ± SD; **P < 0.01. (c) Histological and immunohistochemical analyses of tumour tissues from mice treated with PBS, H5CmTERT-Rd19-Ad, or H5CmTERT-Ad/TRAIL. At 10 days after the cell injection, the mice were euthanised and the brains were collected. Serially sectioned tumour tissues were stained with H & E. The expression levels of TRAIL and E1A were assessed by immunohistochemical analysis. A TUNEL assay was performed to detect apoptosis. Data are representative of three independent experiments. Original magnification: ×50, ×100, and ×400 magnification of the boxed area. T, tumour tissues; N, normal tissues in the H5CmTERT-Ad/TRAIL-treated group.