Efficient antitumor effects of a novel oncolytic adenovirus fully composed of species B adenovirus serotype 35

Abstract

Oncolytic adenoviruses (OAds) are among the most promising oncolytic viruses. Almost all oncolytic adenoviruses are composed of human adenovirus serotype 5 (Ad5) (OAd5). However, expression of the primary infection receptor for Ad5, coxsackievirus-adenovirus receptor (CAR), often declines on malignant tumor cells, resulting in inefficient infection in CAR-negative tumor cells. In addition, at least 80% of adults have neutralizing antibodies against Ad5. In this study, we developed a novel OAd fully composed of OAd35. OAd35 recognizes CD46, which is ubiquitously expressed on almost all human cells and is often upregulated on malignant tumor cells, as an infection receptor. Moreover, 20% or fewer adults have neutralizing antibodies against Ad35. OAd35 mediated efficient cell lysis activities at levels similar to OAd5 in CAR-positive tumor cells, while OAd35 showed higher levels of cell lysis activities than OAd5 in CAR-negative tumor cells. Anti-Ad5 serum significantly inhibited in vitro tumor cell lysis activities of OAd5, whereas OAd35 exhibited comparable levels of in vitro tumor cell lysis activities in the presence of anti-Ad5 and naive serum. OAd35 significantly suppressed growth of the subcutaneous CAR-positive and CAR-negative tumors following intratumoral administration. These results indicated that OAd35 is a promising alternative oncolytic virus for OAd5.

Keywords: CAR; CD46; adenovirus serotype 35; neutralizing antibody; oncolytic virus.

Graphical abstract

Figure 1

Construction strategy for an OAd35 plasmid, pAdMS2-hTERT-E1 pAd35-hTERT-E1 and pAdMS2 were digested by SalI and SwaI, respectively. pAdMS2-hTERT-E1 was produced via homologous recombination in Escherichia coli BJ5183 using these fragments. mE1, mutated E1 gene; ITR, inverted terminal repeat.

Figure 2

Flow cytometric analysis of CAR and CD46 expression on human tumor cells Cells were labeled with mouse anti-CAR monoclonal antibody and PE-conjugated goat anti-mouse IgG secondary antibody or PE-conjugated mouse anti-CD46 monoclonal antibody. The data were analyzed by using FlowJo flow cytometry data analysis software. Light gray histogram, isotype control; gray histogram, anti-CAR or anti-CD46 antibody.

Figure 3

Tumor cell lysis activities and safety profiles of OAd35 (A and B) Viabilities of (A) human tumor cell lines and (B) normal cells were assessed by crystal violet staining assay. Cells were infected with OAds and wild-type Ads at the indicated VP/cell. Cells were stained with crystal violet following a 5-day incubation after infection. The representative images from at least 2 independent experiments were shown. (C and D) Viabilities of (C) human tumor cells and (D) normal cells were also evaluated by WST-8 assay. Cell lines were infected with OAds and wild-type Ads at 300 VP/cell. At the indicated time points, cell viabilities were determined by WST-8 assay. The viability in the mock-infected group was normalized to 100%. These data are expressed as the means ± SD (n = 4).

Figure 4

The E1A gene expression following infection with OAds and wild-type Ads (A and B) Human tumor (A) and normal (B) cells were infected with OAds and wild-type Ads at 100 VP/cell. Total RNA was recovered at 72 h after infection, followed by real-time RT-PCR analysis of the E1A genes. The values were normalized by the mRNA levels of a housekeeping gene, human GAPDH. The E1A gene expression levels of wild-type Ads were normalized to 1. These data are expressed as the means ± SD (n = 4).

Figure 5

OAd35 genome copy numbers in tumor cells (A) Human tumor cell lines were infected with OAds at 100 VP/cell. Total DNA was recovered at the indicated time points, followed by real-time PCR analysis of viral genome copy numbers. These data are expressed as the means ± SD (n = 3). (B) HepG2 cells were infected with OAds at 0.001 VP/cell. Following an 8-day incubation after infection, phase-contrast photomicrographs were obtained. Scale bar, 100 μm.

Figure 6

Tumor cell lysis activities of OAd35 in the presence of anti-Ad5 serum (A and B) HepG2 (A) and T24 (B) cells were infected with OAds at 300 VP/cell in the presence or absence of mouse anti-Ad5 serum. As a control, serum collected from naive mice was used. Cell viabilities were determined by WST-8 assay following a 5-day incubation. The viability in the mock-infected group was normalized to 100%. These data are expressed as the means ± SD (n = 4).

Figure 7

Tumor growth following intratumoral administration of OAd35 (A and B) OAds were intratumorally injected into (A) H1299 and (B) T24 tumor-bearing mice at a dose of 2.4 × 10⁹ VP/mouse. Arrows indicate the number of days after virus injection (days 0 and 3). Tumor volume is expressed as the mean tumor volume ± SE. ∗p < 0.05 (versus PBS; H1299 tumor; n = 7, T24 tumor; n = 6).