Ixovex-1, a novel oncolytic E1B-mutated adenovirus

Abstract

There is a great demand for improved oncolytic viruses that selectively replicate within cancer cells while sparing normal cells. Here, we describe a novel oncolytic adenovirus, Ixovex-1, that obtains a cancer-selective replication phenotype by modulating the level of expression of the different, alternatively spliced E1B mRNA isoforms. Ixovex-1 is a recombinant adenovirus that carries a single point mutation in the E1B-93R 3' splice acceptor site that results in overexpression of the E1B-156R splice isoform. In this paper, we studied the characteristics of this novel oncolytic adenovirus by validating its in vitro behaviour in a panel of normal cells and cancer cells. We additionally studied its anti-tumour efficacy in vivo. Ixovex-1 significantly inhibited tumour growth and prolonged survival of mice in an immune-deficient lung carcinoma tumour implantation model. In complementation experiments, overexpression of E1B-156R was shown to increase the oncolytic index of both Ad5wt and ONYX-015. In contrast to prior viruses of similar type, Ixovex-1 includes a functional E3B region for better in vivo efficacy. Throughout this study, the Ixovex-1 virus has been proven to be superior in competency compared to a virus with multiple deletions.

Fig. 1. The adenovirus genome and E1B cassette.

a The full-length genome with the E1 and E3 gene cassettes is indicated. b E1B splice map showing alternative E1B gene products. The full-length E1B transcript carries the E1B-176R, E1B-496R and pIX ORF. Through alternative splicing within the E1B-496R ORF, another three additional mRNAs are produced encoding for proteins, E1B-93R, E1B-156R and E1B-84R. The spliced E1B products have the 79 N-terminal aa in common with E1B-496R but differ in the C-terminal, except for E1B-156R, which splices in frame with E1B-496R.

Fig. 2. An amino acid change in the E1B-496R protein results in a loss of protein accumulation.

A549 cells were infected with the respective virus at 2.5 pfu/cell and total cell lysate was collected at 48 hpi. Shown is a western blot stained with polyclonal α-capsid (late) protein antibody (top panel), a monoclonal α-E1B-496R antibody (2A6, middle panel) and a monoclonal α-actin antibody as a loading control (lowest panel). The Control lane is lysate from non-infected cells.

Fig. 3. The point mutation in the E1B-496R ORF in Ixovex-1 blocks splicing to the E1B-93R splice acceptor (SA1).

A549 cells were infected with the respective virus at 2.5 pfu/cell and a total RNA was collected at 48 hpi. cDNA was made using an oligo-dT oligonucleotide. PCR was performed using a common sense oligonucleotide upstream of the splice donor 1 and a specific antisense oligonucleotide downstream of the respective splice acceptor. The PCR reactions were run on a 2% agarose gel. b In parallel, total protein was harvested at 48 hpi and western blot was performed using the 2A6 antibody that binds the common N-terminus shared by E1B-496R and its smaller spliced versions, an anti-E1A and a polyclonal anti-capsid (late) protein antibody.

Fig. 4. Ixovex-1 fails to induce p53 degradation, but does not induce p53 accumulation.

A549 cells were infected with the respective virus at 2.5 pfu/cell and total cell lysate was collected 48 hpi. Shown is a western blot using a monoclonal α-p53 antibody (top panel) and the monoclonal α-actin antibody as a loading control (lower panel). Mock lane is lysate from non-infected cells.

Fig. 5. Replication-efficiencies in H460 and H1299 cancer cells.

Each cell line was infected with 2.5 pfu/cell of the respective virus. Cells and media were harvested at 24, 48 and 72 hpi and analysed by a limited dilution assay. CPE was scored after 10 days and TCID₅₀ (pfu/cell) results were calculated.

Fig. 6. Replication-competencies in NHBE.

Cells were infected with 1 pfu/cell of the respective virus. Cells and media were harvested at 24, 48 and 72 hpi and analysed by a limited dilution assay. CPE was scored after 10 days and TCID₅₀ (pfu/cell) results were calculated. The result for Ixovex-1 was adjusted upwards because the level of replication was slightly below the detection limit.

Fig. 7. Ixovex-1 significantly inhibits tumour growth in nude mice and prolongs survival.

Nude NMRI mice were injected subcutaneously with H1299 cells in PBS mixed 1:1 with matrigel. After 2 weeks, mice with tumour growth (size ~70–100 mm³) were randomised into four groups with seven mice in each. Tumours were injected on days -2, -1, and 0 with (10⁸ viral particles) Ad5wt, ONYX-015, Ixovex-1 or PBS only. a All tumours were measured using a digital caliper two times per week. Tumour size was calculated as: length × width × depth × 0.66. Mice were sacrificed when tumour size exceeded 1000 mm³. Data plotted are the volume means ± SEM. One-way ANOVA analysis of variance was performed comparing the growth curves. Prism 6 software (GraphPad Software Inc.) was used throughout the analysis. b The proportions of live mice are plotted against time. The median time of survival was determined by Kaplan–Meier survival analysis (*p < 0.05, log-rank test for statistical significance).