Development of a melanoma-specific adenovirus

Concerns regarding the hepatotoxicity of adenovirus for cancer gene therapy have led to attempts to engineer viruses for tissue-specific gene expression and tissue-specific replication. The Tyrex2 (a tandem murine melanocyte-specific enhancer) system was used to express luciferase, purine nucleoside phosphorylase (PNP), and the essential adenoviral gene E1A. In nonmelanoma cell lines, the CMV promoter/enhancer (CMV p/e) was 969 times stronger than the Tyrex2 construct, whereas in melanoma cells it was only 2.6 times stronger. An adenovirus with Tyrex2 regulating PNP (Ad2Tyr2-PNP) was tested for cytotoxicity. In melanoma cells, treatment with Ad2Tyr2-PNP plus the prodrug 6-methylpurine deoxyriboside (6-MPDR) resulted in 90% cytotoxicity by day 4. In non-melanoma cell lines, only the CMV p/e resulted in significant cytotoxicity. We compared the intrinsic E1A promoter/enhancer (E1A p/e) system with the melanoma-specific constructs and found that the Tyrex2 system achieved higher levels of luciferase than the E1A p/e in all melanoma lines tested. In non-melanoma cell lines, the E1A p/e is 12.4 times stronger than the Tyrex2 construct. Tyrex2 was then used to regulate adenoviral E1A expression to construct a melanoma-specific replicating adenovirus. We were unable to achieve selective replication. As E1A is a known transactivator of the adenovirus major late promoter (MLP), we studied the ability of the MLP to express a transgene in the context of tissue-selective E1A expression. We were able to demonstrate high levels of luciferase activity with this construct; however, selectivity was lost. Melanoma-specific adenovirus expression was achieved with the Tyrex2 construct, and this led to melanoma-specific cytotoxicity by the potent PNP suicide gene. Selective melanoma-specific replication was not successful. The MLP may be a useful promoter in the context of a tissue-specific replicating adenovirus.