Human adenoviruses, which are widespread in the population, are non-enveloped viruses with double-stranded DNA genomes. The study of these viruses has yielded fundamental insights into the mechanisms by which genes are expressed in human cells, as well as the intricate circuits that control cell proliferation but become deranged during malignant transformation. There is considerable interest in developing derivatives of adenoviruses for therapeutic applications, for example, as gene transfer vectors and oncolytic viruses for cancer therapy. However, relatively little attention has been paid to the interactions among viral and cellular gene products required for optimal virus reproduction in normal human cells (as opposed to highly abnormally transformed cells derived from human tumors). Furthermore, the molecular functions of the products of genes routinely deleted from “therapeutic” adenoviruses are not fully understood. Our research seeks to address such issues, focusing on the contributions of the viral E1B 55 kDa protein to the infectious cycle in normal human cells: the gene for this protein is deleted from all adenovirus gene transfer vectors, and such deletion (in the absence of other alterations) confers tumor cell-selective reproduction and host cell lysis.

One function of this E1B protein implicated in the cell-selective reproduction is induction of selective export of viral mRNAs during the late phase of infection. This process requires assembly in infected cells of a virus-specific E3 ubiquitin ligase that contains the viral E1B 55 kDa and E4 Orf6 proteins, as well as several cellular proteins. We have established that this virus-specific enzyme targets one or more components of the major mRNA export pathway in mammalian cells, the Nxf1 pathway, and that efficient export of viral late mRNAs is more dependent on the E1B 55 kDa protein in normal than in tumor cells. As the relevant substrates of the virus specific E3 ubiquitin ligase have not been identified, we are applying proteomic methods to compare the populations of ubiquitinylated proteins in cells infected by mutant viruses defective for production of the E1B 55 kDa protein and their wild-type parent.

One of the first activities ascribed to the E1B 55 kDa protein was repression of transcription in simplified experimental systems, such as transient expression assays. However, the contribution of this function to viral reproduction was not known, until the results of our genome-wide analysis of cellular gene expression in infected normal human cells established that the E1B protein inhibits expression of a substantial number of genes. This set of genes proved to highly enriched for those associated with anti-viral defenses, particularly that mediated by the cytokine type I interferon. In fact, the E1B 55 kDa protein is required to block inhibition of viral reproduction by interferon, and when the E1B protein is not made, viral genome replication is inhibited severely in infected cells exposed to the interferon. The results of molecular and genetic analysis have established that this viral protein represses transcription of interferon-inducible genes and that his function correlates with preventing inhibition of viral DNA synthesis by interferon. This cytokine does not induce inhibition of synthesis of viral replication proteins, but rather leads to a block in formation of the specialized intranuclear structures in which viral DNA synthesis takes place, so-called viral replication centers. The mechanism by which the E1B 55 kDa protein represses transcription of specific human genes is currently under investigation, as is that by which the products of one or more of the some 130 human genes that are both interferon-inducible and E1B protein repressed disrupt formation of viral replication centers.