Vaccinia virus is a large (2 um), double-stranded DNA virus. Unlike many other DNA viruses, vaccinia completes its entire life cycle in the host cytoplasm and therefore, encodes for its own machinery for gene expression including a multisubunit DNA-dependent RNA polymerase. Vaccinia transcription has proven to be a good model for investigating eukaryotic transcription, particularly transcription termination.

The genes of vaccinia are expressed in a temporal fashion, early, intermediate, and late, named according to the time of expression after infection. Early gene transcription occurs entirely within the virus core immediately upon infection. Termination of early genes requires the virion form of the RNA polymerase (vRNAP), Nucleoside triphosphate phosphohydrolase I (NPH I), ATP, the vaccinia termination factor (VTF), and the U5NU termination signal located in the nascent transcript 25-50 nts upstream from the site of termination (reviewed in Moss 2007). My research focused on elucidating the roles and mechanisms of the various components of early gene transcription termination machinery.

Previous work by other labs showed that early transcription termination occurs in at least two steps: recognition of the U5NU termination signal, apparently by VTF, and ATP hydrolysis by NPH I to release the transcripts. VTF, also known as the viral capping enzyme, binds the nascent RNA specifically to the termination signal during transcription elongation. Previous studies also demonstrated that the NPH I releases transcripts in the presence of VTF and U₅NU.

One focus of my dissertation research was understanding the mechanism by which NPH I releases transcripts. NPH I appears to cause transcript release by a mechanism known as forward translocation. According to the forward translocation model, NPH I propels itself downstream along the DNA and subsequently moves vRNAP downstream as well. When NPH I moves vRNAP forward in the absence of RNA chain elongation, the DNA/RNA hybrid shortens and becomes unstable, resulting in transcript release. Therefore, in the absence of transcription elongation, NPH I can release transcripts independent of VTF and U₅NU. However, when NPH I moves vRNAP downstream during active transcription elongation, the DNA/RNA hybrid is maintained such that transcript release does not occur. In this circumstance, NPH I requires VTF and U₅NU to release transcripts.

The second focus of my project was discerning the role of VTF in transcription termination. Based on the results of the NPH I studies above, VTF was hypothesized to inhibit transcription elongation thereby allowing NPH I to release the transcripts. Indeed, VTF was shown to inhibit transcription elongation by enhancing a pause near the site of termination. This pause appears to be induced by structure in the nascent RNA immediately downstream of the termination signal. This RNA 2° structure-induced pause is insufficient to stimulate termination without pause-enhancement of VTF. Conversely, VTF is able to cause termination in the absence of RNA structure, but termination is neither efficient nor site-specific. We also provide evidence to suggest that VTF causes transcription termination by a mechanism beyond binding the U₅NU termination signal.