Enhanced Oncolytic Activities of the Telomerase-Specific Replication-Competent Adenovirus Expressing Short-Hairpin RNA against Dicer Free

Oncolytic viruses are highly promising for cancer treatment by virtue of their potential oncolytic activity (1, 2). Various types of oncolytic viruses that can efficiently replicate in tumor cells and induce tumor lysis in vitro and in vivo have been developed. Several clinical trials using oncolytic viruses have been carried out, and promising results have been reported (1, 2). For example, an engineered oncolytic herpesvirus (T-VEC, also known as OncoVEX^GM-CSF) exhibited potential antitumor effects in clinical trials (3) and was recently approved by an FDA committee.

Adenoviruses (Ad) also function as frameworks for oncolytic viruses. Various types of oncolytic Ads have been developed by genetic recombination (4). In order to enhance the efficiency and specificity of oncolytic virotherapy, a variety of improvements have been made to oncolytic viruses. For example, expression cassettes of inflammatory cytokine genes and tumor-suppressive genes were incorporated into the oncolytic virus genome (5, 6). Although these approaches significantly enhanced tumor cell death, there have been few studies of techniques that can enhance the replication efficiencies of oncolytic viruses. Such improvement should be crucial for the further enhancement of their oncolytic activities. The enhancement of an oncolytic virus by regulating viral or cellular factors is a promising strategy for obtaining more efficient antitumor effects.

Viral-associated (VA)-RNAs, which are approximately 160-nucleotide-long noncoding RNAs, are rapidly transcribed by RNA polymerase III after infection and accumulate to very high levels (VA-RNA I: 10⁸ molecules per cell; VA-RNA II: 5 × 10⁶ molecules per cell; refs. 7, 8). VA-RNAs support the efficient amplification of Ads by antagonizing the antiviral action associated with the Ad-induced activation of double-stranded RNA-dependent protein kinase (PKR; refs. 9, 10). After transcription from the Ad genome in Ad-infected cells, VA-RNA I binds to PKR with high affinity and then inhibits PKR activation, leading to efficient Ad protein synthesis and the promotion of Ad replication (9, 11).

VA-RNAs are processed in a manner similar to miRNAs (12–17), resulting in the production of VA-RNA-derived miRNAs (mivaRNA), which are approximately 22 nucleotides long. It remains to be clarified whether mivaRNAs are crucial for Ad replication. We demonstrated recently that the processing of VA-RNAs mediated by the endoribonuclease Dicer results in a decrease in VA-RNA copy numbers, leading to the suppression of Ad replication (18). A knockdown of Dicer significantly promoted Ad replication (18), suggesting that oncolytic Ad replication would also be upregulated by a Dicer knockdown.

In this study, we examined whether a knockdown of Dicer promotes the replication of the telomerase-specific replication-competent Ad (TRAD), which carries the human telomerase reverse transcriptase (hTERT) promoter-driven E1 gene expression cassette (19, 20). TRAD has shown promising antitumor effects, and a phase I clinical trial using TRAD has already been completed (21). In the present study, we developed a TRAD carrying an expression cassette of short-hairpin RNA (shRNA) against Dicer (shDicer; TRAD-shDicer). TRAD-shDicer mediated a significant reduction in Dicer expression following infection in a tumor cell–specific manner. Compared with the conventional TRAD, TRAD-shDicer exhibited more efficient replication and oncolytic activity in vitro and in vivo.

HeLa (a human epithelial carcinoma cell line: RCB0007), HepG2 (a human hepatocellular carcinoma cell line: RCB1648), and Huh7 (a human hepatocellular carcinoma cell line: RCB1366) cells were obtained from the JCRB Cell Bank. The other cell lines were obtained from ATCC. All cell lines were obtained within the period from 2007 to 2013. All experiments were done using cells passaged less than 20 times. The authentication was performed by the manufacturer, not by the authors.

HeLa, HepG2, HuH7, HEK293 (a transformed embryonic kidney cell line), SK HEP-1 (a human hepatoma cell line), and SK-OV-3 (a human ovarian carcinoma cell line) cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), streptomycin (100 μg/mL), and penicillin (100 U/mL). H1299 (a non-small cell lung carcinoma cell line) cells were cultured in RPMI1640 supplemented with 10% FBS, streptomycin (100 μg/mL), and penicillin (100 U/mL). HeLa-shDicer cells were previously constructed (22) and cultured in DMEM supplemented with 10% FBS, streptomycin (100 μg/mL), and penicillin (100 U/mL).

Normal human umbilical vein endothelial cells (HUVEC), normal human lung fibroblasts (NHLF), and normal human prostate stromal cells (PrSC) were cultured in the medium recommended by the supplier (Lonza).

Control siRNA (Allstars Negative Control siRNA) was purchased from Qiagen. siRNA against Dicer (siDicer) was obtained from Gene Design. The target sequence of siDicer was 5′-gaatcagcctcgcaacaaa-3′. Efficient knockdown of Dicer at both the mRNA and protein levels after transfection with siDicer was previously demonstrated (22).

pAdHM19-hAIB-shDicer, a plasmid for a telomerase-specific replication-competent Ad expressing shDicer, was constructed as follows. First, pENTR1A-H1T2-shDicer (18) was digested with BamHI/XbaI and then ligated with BamHI/XbaI-digested pHM13 (23), resulting in pHM13-H1T2-shDicer. Next, pHM5-hAIB (20), in which the E1A and E1B genes linked with an internal ribosomal entry site (IRES) are located downstream of an hTERT promoter, was digested with I-CeuI/PI-SceI and then ligated with I-CeuI/PI-SceI-digested pAdHM19 (23). The resulting plasmid, pAdHM19-hAIB, was digested with Csp45I and then ligated with a ClaI-digested fragment of pHM13-H1T2-shDicer, resulting in pAdHM19-hAIB-shDicer. pAdHM19-hAIB-shLuc was similarly constructed using pENTR1A-H1T2-shLuc, which was generated using oligonucleotides shLuc-S and shLuc-AS (24; Supplementary Table S1). Further details on the construction methods are available upon request.

Recombinant Ads were prepared as follows. PacI-digested pAdHM19-hAIB, pAdHM19-hAIB-shLuc, or pAdHM19-hAIB-shDicer were each transfected into HEK293 cells using Lipofectamine 2000 (Life Technologies), resulting in the conventional TRAD, TRAD expressing shRNA against luciferase (shLuc; TRAD-shLuc), and TRAD-shDicer, respectively. These recombinant Ads were amplified and purified by two rounds of cesium chloride gradient ultracentrifugation, dialyzed, and stored at −80°C (25). The determination of infectious titer units (IFU) was accomplished using an Adeno-X Rapid Titer Kit (Clontech).

The Western blotting assay was performed as described (26). Briefly, whole-cell extracts were prepared and electrophoresed on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels under reducing conditions, followed by electrotransfer to PVDF membranes (Millipore). After blocking with 5% skim milk prepared in TBS-T (Tween-20, 0.1%), the membrane was incubated with a mouse anti-Dicer antibody (13D6; Abcam; dilution 1:1,000) or mouse anti-β-actin (AC-15; Sigma-Aldrich; dilution 1:5,000), followed by incubation in the presence of horseradish peroxidase (HRP)-labeled anti-mouse or anti-rabbit IgG antibody (Cell Signaling Technology).

Real-time reverse transcription polymerase chain reaction (RT-PCR) analysis was performed as described (27). Complementary DNA was synthesized using 500 ng of total RNA with a Superscript VILO cDNA synthesis kit (Life Technologies). The real-time RT-PCR analysis was performed with Thunderbird SYBR qPCR Mix using the StepOnePlus real-time PCR system (Life Technologies) as described (28). The primer sequences used in this study are described in Supplementary Table S1.

We isolated total DNA, including Ad genomic DNA, from the cells infected with Ads by using a DNeasy Blood & Tissue Kit (Qiagen). After isolation, the Ad genome copy numbers were quantified by using StepOnePlus as described (29). The primer and probe sequences used are described in Supplementary Table S1.

Following infection with Ads, the cells were recovered and subjected to three cycles of freezing and thawing. After centrifugation, the supernatants were added to HEK293 cells. After 48 hours of incubation, the cells infected with Ads were counted using the Adeno-X Rapid Titer Kit (Clontech).

Cells were infected with oncolytic Ads at a multiplicity of infection (MOI) of 2. Cell viabilities were determined at the indicated time points by staining with AlamarBlue (Life Technologies) according to the manufacturer's instructions.

H1299 cells or SK-OV-3 cells (2 × 10⁶ cells per mouse) were injected subcutaneously into the flank of 5-week-old female BALB/c nu/nu mice and grown to approximately 5 to 6 mm in diameter. At that time, the mice were randomly assigned into four groups and intratumorally injected with PBS, the conventional TRAD, TRAD-shLuc, or TRAD-shDicer at a dose of 1 × 10⁷ IFU per mouse, followed by reinjection after 3 days. The tumors were measured every 3 days, and tumor volume was calculated by the following formula: tumor volume (mm³) = a × b² × 3.14 × 6⁻¹, where a is the longest diameter and b is the shortest.

The H1299 xenograft tumors were resected 14 days after administration, followed by homogenization. Total DNA, including the viral genome, was extracted from the whole-tumor homogenates by using a DNeasy Blood and Tissue Kit (Qiagen). After isolation, the Ad genome copy numbers were quantified as described above.

These experiments were approved by the Animal Experiment Committee of Osaka University.

Statistical significance was determined using the Student t test. Data are presented as the means ± SD or SE.

First, in order to determine whether the knockdown of Dicer promotes TRAD replication, we infected Dicer-knockdown cells with the conventional TRAD. Higher amounts of VA-RNAs were detected in the cells transfected with siDicer than in the cells transfected with a control siRNA (siControl) following infection with the conventional TRAD (Fig. 1A). Approximately 4- and 6-fold higher copy numbers of the viral genome of TRAD were found in the HeLa and H1299 cells, respectively, when Dicer was knocked down (Fig. 1B). The IFU titers of progeny TRAD were also increased by approximately 3- and 7-fold in Dicer-knockdown HeLa and H1299 cells, respectively (Fig. 1C). In order to further demonstrate that the knockdown of Dicer enhances TRAD replication, HeLa transformants inducibly expressing shDicer (HeLa-shDicer cells; ref. 22) were infected with TRAD. We previously demonstrated that the knockdown of Dicer in HeLa-shDicer cells was induced by doxycycline (Dox) in a dose-dependent manner (22). When HeLa-shDicer cells were infected with TRAD in the presence of various concentrations of Dox, copy numbers of the viral genome of TRAD in the cells were inversely correlated with the concentrations of Dox (Fig. 1D). These results indicate that Dicer expression negatively regulates TRAD replication and that the knockdown of Dicer leads to the promotion of TRAD replication.

Next, in order to apply the findings described above to TRAD-mediated virotherapy, we developed a TRAD expressing shDicer (TRAD-shDicer), which carries the hTERT promoter-driven E1 gene expression cassette and an H1 promoter-driven shDicer in the E1- and E3-deletion regions, respectively (Fig. 2A). The real-time RT-PCR and Western blotting analyses showed a significant knockdown of Dicer following infection with TRAD-shDicer in HeLa cells (Fig. 2B–D). A slight reduction in Dicer expression was also found after infection with the conventional TRAD and TRAD-shLuc (Fig. 2C).

To determine the copy numbers of VA-RNA following infection with TRAD-shDicer and the replication efficiencies of TRAD-shDicer, we infected several tumor cells with the TRADs at an MOI of 2. Approximately 8-fold higher VA-RNA copy numbers were observed in the HeLa cells following infection with TRAD-shDicer compared with the conventional TRAD and a TRAD expressing shRNA against luciferase (TRAD-shLuc; Fig. 3A). Approximately 8- and 12-fold higher amounts of the viral genome and IFU titers of progeny TRADs, respectively, were observed for TRAD-shDicer compared with the conventional TRAD and TRAD-shLuc in HeLa cells 48 hours after infection (Fig. 3B and C), probably because infection with TRAD-shDicer resulted in higher copy numbers of VA-RNA (Fig. 3A). In the H1299, SK HEP-1, and SK-OV-3 cells, TRAD-shDicer also showed significantly higher replication efficiencies than the conventional TRAD and TRAD-shLuc (Fig. 3B and C; Supplementary Fig. S1A). We next examined the cell lysis activity of TRAD-shDicer. TRAD-shDicer showed significantly higher cell lysis activities against all tumor cells examined compared with the conventional TRAD and TRAD-shLuc (Fig. 4A and B). Almost all of the tumor cells were lysed by TRAD-shDicer at 5 days after infection, whereas approximately 60% of the tumor cells were still viable at 5 days after infection with the conventional TRAD and TRAD-shLuc (Fig. 4A). TRAD-shDicer also exhibited significantly higher cell lysis activities in SK-OV-3 cells, compared with the conventional TRAD and TRAD-shLuc (Supplementary Fig. S1B).

These results indicate that TRAD-shDicer shows higher replication efficiencies and tumor lysis activities in vitro compared with the conventional TRAD.

To assess the safety profiles of TRAD-shDicer, we infected several normal cells with TRAD-shDicer at an MOI of 2. We previously demonstrated that genome copy numbers of the conventional TRAD were slightly but significantly increased in several normal cells (20). TRAD-shDicer did not mediate a significant knockdown of Dicer following infection in NHLFs (Fig. 5A). TRAD-shDicer exhibited slightly higher viral genome copy numbers in NHLFs and PrSCs compared with the conventional TRAD and TRAD-shLuc, 48 hours after addition to the cells (Fig. 5B). However, the IFU titers of the progeny Ad were below the detectable levels in the normal cells (<10 IFU per well). In addition, no apparent reduction in the cell viabilities was observed in the normal cells after infection with TRAD-shDicer, or with the conventional TRADs and TRAD-shLuc (Fig. 5C and D). These results indicate that TRAD-shDicer as well as the conventional TRADs do not cause significant cellular toxicity in normal cells.

Finally, to determine the antitumor effects of TRAD-shDicer on xenograft tumors, we intratumorally administered TRAD-shDicer into mouse xenograft H1299 and SK OV-3 tumors. The growth of subcutaneous H1299 and SK OV-3 tumors in the mice was efficiently suppressed following an intratumoral injection of the conventional TRAD or TRAD-shLuc (Fig. 6). Compared with the conventional TRAD or TRAD-shLuc, TRAD-shDicer exhibited stronger antitumor effects following the intratumoral injection (Fig. 6A). Approximately 7-fold higher amounts of the viral genome were found for TRAD-shDicer, compared with the conventional TRAD in subcutaneous H1299 tumors in the mice 14 days after administration (Fig. 6B). These results indicate that TRAD-shDicer also shows higher replication efficiencies and tumor lysis activities in vivo compared with the conventional TRAD.

VA-RNAs, which are Ad-encoded small RNAs, promote Ad replication via several mechanisms, including the inhibition of PKR (7, 8). We previously demonstrated that VA-RNAs are cleaved by Dicer, leading to a loss of promotion activity for Ad replication. Dicer is a negative cellular factor for Ad replication (18). Our goal in the present study was to improve the efficiencies of oncolytic Ad replication and the cell lysis activities of oncolytic Ads in tumor cells by inserting an shDicer-expressing cassette. Our results demonstrated that TRAD-shDicer efficiently induced the knockdown of Dicer in a tumor-specific manner (Fig. 2) and showed significantly higher replication efficiency and tumor cell lysis activity compared with the conventional TRAD (Figs. 3, 4, and 6). The viabilities of the normal cells were not altered by infection with TRAD-shDicer (Fig. 5).

To further enhance the antitumor effects of oncolytic viruses, it would be necessary to improve their replication efficiencies. For example, several compounds that can enhance oncolytic virus growth were identified using high-throughput screening (30). Diallo and colleagues demonstrated that treatment with some chemical compounds tumor-specifically enhanced the replication of oncolytic vesicular stomatitis virus (VSV) and its oncolytic activities via the inhibition of interferon (IFN)-induced antiviral responses (30). Those results suggest that the knockdown of antiviral genes would enhance the replication efficiencies of oncolytic viruses.

We recently observed that Dicer functions as an antiviral factor against Ad infection (18). In the present study, the suppression of Dicer expression significantly upregulated the replication of TRAD via several mechanisms, including the elevation of the VA-RNA copy numbers. This is a novel approach to more efficient oncolytic virotherapy. In addition, upregulation of the replication efficiencies of TRAD can reduce the injected doses of TRAD needed to achieve tumor regression, resulting in a reduction in side effects due to TRAD.

Rauschhuber and colleagues demonstrated that Ad replication was promoted by inhibition of miRNA-mediated posttranscriptional gene silencing using the tomato bushy stunt virus P19 protein (31). As shown in our present and previous studies (22), knockdown of Dicer also resulted in the enhancement of Ad replication. These results suggest that the arming with RNAi suppressors, such as shDicer and P19 protein, is highly beneficial for the development of novel oncolytic Ads showing high replication efficiencies.

As shown in Figs. 2B and 5A, TRAD-shDicer mediated a tumor cell-specific reduction in Dicer expression, whereas Dicer expression was not significantly reduced in the normal cells following treatment with TRAD-shDicer. This is probably because the expression of shDicer can be amplified only in tumor cells by telomerase-specific viral replication. Kojima and colleagues reported a similar finding: TRAD containing a GFP expression cassette mediated GFP expression in a tumor cell–specific manner (32). These results suggest that a transgene incorporated in the TRAD genome is significantly expressed in tumor cells but not normal cells.

We observed that Dicer expression was slightly suppressed after infection with the conventional TRAD and TRAD-shLuc (Fig. 2B). This is probably attributable to VA-RNA-mediated inhibition of Dicer protein expression. Bennasser and colleagues demonstrated that VA-RNA inhibited the export of Dicer mRNA from the nucleus to the cytoplasm by competitive binding for Exportin5 between Dicer mRNA and VA-RNA, leading to the reduction of Dicer protein levels after Ad infection (33). However, the suppression levels of Dicer expression in the conventional TRAD- and TRAD-shLuc-infected cells in the present study were much lower than those in the TRAD-shDicer-infected cells (Fig. 2B). TRAD-shDicer would efficiently mediate the inhibition of Dicer expression through both VA-RNA and shDicer expressions, leading to the efficient replication of TRAD-shDicer.

The additional insertion of an shDicer expression cassette into other types of oncolytic virus constructs would be beneficial for the enhancement of the viral replication efficiencies and tumor cell lysis activities. The replication of several types of viruses is promoted in cells with diminished Dicer expression (34, 35). For example, Otsuka and colleagues reported that the replication of VSV was enhanced in cells with a mutant Dicer allele due to the decreased expression of miR-24 and miR-93, which directly target the VSV genome (35). VSV was reported to exhibit potent oncolytic activities, and clinical trials using VSV are in progress (1, 2). The insertion of an shDicer expression cassette might be highly promising for VSV-based oncotherapy.

Infection with an oncolytic virus directly mediates the destruction of tumor cells. In addition to the direct tumor cell killing activity, oncolytic viruses induce antitumor effects via the release of intracellular materials, including tumor antigens and viral antigens, in association with tumor lysis events. The released antigens are recognized by immune cells, leading to the activation of antitumor immunity (1, 2). We and another group reported that VA-RNAs were recognized by RIG-I and/or MDA5, leading to the induction of type I IFN (36, 37). In the present study, the use of TRAD-shDicer resulted in increased expression levels of VA-RNAs after infection (Fig. 3A). After tumor cell lysis by TRAD-shDicer, higher amounts of VA-RNAs might be released and activate antitumor immunity via an activation of innate immunity as a type of adjuvant.

In summary, we developed a novel oncolytic Ad expressing shDicer, which shows significantly higher replication efficiency and antitumor activity compared with the conventional oncolytic Ad. The viability of several types of normal cells was not apparently altered following infection with TRAD-shDicer, as was the case with the conventional TRAD. TRAD-shDicer offers great potential for safer and more effective cancer virotherapy.

T. Fujiwara and H. Mizuguchi are consultants of Oncolys BioPharm Inc. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M. Machitani, F. Sakurai

Development of methodology: M. Machitani

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Machitani, K. Wakabayashi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Machitani, F. Sakurai, M. Tachibana, H. Mizuguchi

Writing, review, and/or revision of the manuscript: M. Machitani, F. Sakurai, H. Mizuguchi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T. Fujiwara

Study supervision: F. Sakurai, H. Mizuguchi

We thank Sayuri Okamoto and Eri Hosoyamada (Graduate School of Pharmaceutical Sciences, Osaka University, Osaka, Japan) for their help.

This work was supported by grants-in-aid for Scientific Research (A) to H. Mizuguchi, (B) to F. Sakurai, and for Young Scientists (A) to F. Sakurai from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan and by grants from the Mochida Memorial Foundation for Medical and Pharmaceutical Research to H. Mizuguchi. M. Machitani and K. Wakabayashi are Research Fellows of the Japan Society for the Promotion of Science.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.