Search

Background Oncolytic adenoviruses provide a promising alternative for cancer treatment. Recently, adjuvant interferon (IFN)-alfa has shown significant survival benefits for pancreatic cancer, yet was impeded by systemic toxicity. To circumvent these problems adenovirus with high-level targeted IFN-alfa expression can be generated. Methods Conditionally replicative adenoviruses (CRAds) with improved virulence and selectivity for pancreatic cancer were generated. The vectors were tested in vitro, in vivo, and in human pancreatic cancer and normal tissue specimens. Results Adenoviral death protein and fiber modifications significantly improved oncolysis. CRAds selectively replicated in vitro, in vivo and showed persistent spread in cancer xenografts. They showed high-level replication in human pancreatic cancer specimens, but not in normal tissues. Improved IFN-CRAd oncolytic efficiency was shown. Conclusions Optimized cyclooxygenase-2 CRAds show highly favorable effects in vitro and in vivo. We report a pancreatic cancer–specific, highly virulent, IFN-expressing CRAd, and we believe that adenovirus-based IFN therapy offers a new treatment opportunity for pancreatic cancer patients.

Top of pageAbstract Although anti-androgen therapy is initially effective for advanced prostate cancer, all patients eventually progress to hormone-refractory (HR) prostate cancer (PC) due to biological selection. In this stage, a median survival is only 6-12 months. Therefore, new therapeutic approaches for hormone-refractory prostate carcinoma (HRPC) are needed. Conditionally replicative adenoviruses (CRAds) are attractive anti-cancer agents. These agents are designed to replicate specifically in tumor cells followed by spread of the viral progeny to neighboring cancer cells. We used the cyclooxygenase-2 (COX-2) promoter for transcriptional targeting of tumor, which is active in many cancers but minimally active in normal tissues, including the liver. The COX-2 promoter (-1432/+59: COX-2L) was tested with luciferase (Luc) expression vectors in PC-3, Du-145, and LnCap prostate cancer cell lines. Very strong COX-2 promoter activity comparable to the CMV promoter was observed in PC-3 and Du-145 HRPC cell lines. We next examined infectivity enhancement in the PC cells by using fiber modification of Ad5 vectors, including incorporation of an RGD4C motif into the HI loop (RGD) and replacement of the Ad5 knob region with the Ad3 knob (Ad5/3 chimera),. Three CMV promoter driven Luciferase expression vectors with the wild type fiber (Ad5Luc1), RGD insertion (AdRGDLuc1), and Ad5/3 chimera fiber (Ad5/3Luc1) were used. The RGD-modified vector did not show enhanced infectivity in either of these PC cell lines, while the Ad5/3 chimera demonstrated significantly higher levels of infection in PC-3 and Du-145 HRPC cells. To analyze the effect of fiber modification on the cytocidal effect of CRAds we constructed COX-2 CRAds with wild type, RGD-modified, and Ad5/3 chimeric fibers, respectively. All COX-2 CRAds showed replication and subsequent oncolytic killing in both PC-3 and Du-145 cell lines while the COX-2 negative LnCap cell line was not affected. Ad5/Ad3 chimera CRAds demonstrated the strongest cytocidal effect among COX-2 CRAds in PC3 and Du-145 cell lines. Analysis of in vivo anti-tumor efficacy was performed using Du-145 prostate carcinoma subcutaneous xenografts. COX-2 CRAd agents with unmodified, RGD-modified, or Ad5/3 chimera fibers as well as wild type Ad5 and a non-replicative E1 deleted vector were used for intratumoral administration (5×109 vp/tumor). Our data indicate that the anti-tumor effect of COX-2 CRAd with Ad5/3 chimera fiber was better than unmodified or RGD-modified CRAds. The present results demonstrate that selective replication of CRAd under the control of the COX-2 promoter and infectivity enhancement with Ad5/3 chimeric fiber may be combined to generate promising viral therapeutic agents for the treatment of hormone refractory prostate carcinomas.

The employment of conditionally replicative adenoviruses (CRAds) provides a novel way to treat lung cancer. To maintain safety and efficacy in clinical application, a noninvasive detection system of viral replication is required. However, there are currently no available methods to dynamically monitor adenovirus replication and spreading ability in clinical settings. Recently we have shown that the EGFP reporter gene placed in the deleted E3 region under the adenoviral major late promoter could be exploited to noninvasively monitor adenovirus replication. In this study, we developed a system which allows monitoring of CRAd replication via the detection of bioluminescence reporter and applied it to derive novel Cox-2 promoter-based CRAds as therapeutic agents for lung cancer. In our reporter vectors most E3 genes except adenoviral death protein (ADP) were deleted or mutated and replaced with the firefly luciferase gene. We combined the Cox-2 promoter-controlled E1 expression cassette with the E3 vector to generate imaging capable Cox-2 CRAds. We designed six different structures to optimize monitoring including configurations with and without ADP and vectors equipped with different fibers accommodating the coxsackie-adenovirus receptor deficiency of cancer cells. The presence or absence of a poly-adenylation signal following the reporter gene has been examined for effect on stability of transgene expression. In vitro, all viruses showed transgene expression augmentation and killing effect over time and in a dose-dependent manner during infection of A549 human lung adenocarcinoma cells. The detected gene expression closely correlated with the viral DNA quantity resulting from viral replication. No expression or cytocidal effect was observed in mouse hepatoma BNL-1NG-A.2 cells in which human adenoviruses do not productively replicate, indicating the dependence of reporter expression on productive replication. All vectors with ADP showed a stronger killing ability without hampering luciferase expression compared to ADP lacking viruses. In vivo optical studies performed in nude mice with A549 subcutaneous xenografts visualized viral replication after both intratumoral and intravenous administration of Cox2-based CRAds. The bioluminescent signal was detected as early as the next day after adenoviral injection, increased with time, persisted over four weeks, and was comparable to that of a matched wild type E1 vector. Surprisingly, in one mouse, bioluminescence signal from a Cox2Crad was detected not only in the A549 subcutaneous nodule but also in spontaneously developed tumor metastases. Of note, in both in vitro and in vivo studies, Cox2-based CRAds demonstrated efficient replication and subsequent oncolysis in Cox2-positive A549 cells but not in Cox-2-negative A431 cells, thus confirming the dependence of their replication on the Cox-2 promoter activity. These data validate the applicability of our promoter-controlled E3 reporter system for noninvasive monitoring of CRAd replication and further development of oncolytic viruses for lung cancer.

The employment of conditionally replicative adenoviruses (CRAds) offers a promising alternative for cancer treatment. To maintain safety in clinical trials, a noninvasive detection system of viral replication is required. Our previous studies revealed the effective and transcriptionally-restricted cytocidal effect of infectivity-enhanced cyclooxygenase-2 (Cox-2) promoter-driven CRAds. However, there are no currently available methods to noninvasively monitor dynamic adenovirus replication, spreading ability, and extent of persistence in vivo. In this study, we have developed a system which allows monitoring of CRAds replication via the detection of fluorescence or luminescence reporters expressed from the E3 region. We deleted or mutated most E3 genes except adenovirus death protein (ADP), and incorporated the enhanced green fluorescent protein (EGFP) or the firefly luciferase (Luc) into the E3 region under the control of the adenovirus major late promoter (MLP). Since MLP turns on only when the virus is replicating, we hypothesized that the transgene expression would closely represent the viral replication in this configuration. Finally, we combined E1 shuttle vector with E1 expression cassette under control of Cox-2 promoter with E3ADPLuc cassette and generated imaging capable Cox-2 CRAds with different fiber modifications in order to overcome coxsackie-adenovirus receptor deficiency of cancer cells. In addition, we considered the structure of the E3 region with and without ADP since a strong cytocidal effect of ADP may not be optimal for imaging. To assess the replication dependent expression of the reporter gene from the E3 region, a replication permissive human lung adenocarcinoma A549 cells and mouse hepatoma BNL-1NG-A.2 cells, in which human adenoviruses do not productively replicate, were applied in our experiments. In vitro, all replication-competent reporter viruses showed transgene expression augmentation over time and in a dose-dependent manner during infection of A549 cells. The detected signal closely correlated with the viral DNA quantity resulting from viral replication. On the other hand, no EGFP/Luc expression was observed in BNL-1NG-A.2 cells, thus indicating the dependence of fluorescence/luminescent signal on productive replication. In vivo optical studies performed with a custom made in vivo imaging system revealed that after intratumoral administration of adenoviruses with the E3 imaging cassette the EGFP expression could be detected two days post-injection. Cox-2 CRAds with the luciferase expression showed the signal on day one and maintained it over four weeks. Importantly, in vivo reporter detection strongly correlated with the underlying level of replication. These data validate the applicability of our E3 reporter system for noninvasive in vivo monitoring of CRAd replication.