the long term aims are to understand the mechanisms of transcription initiation in vivo at a model eukaryotic promoter, the adenovirus major late promoter (MLP), to gain insights into the evolution of its genetic structure, and to understand the regulatory circuits of gene expression control at late times in a virus infection. Previous genetic analyses showed that the TATA box is important, but not essential, to the functioning of the MLP, and suggested that there is some mechanistic interaction, between the TATA box, and the initiator element (INR) encompassing the start site of transcription. The genetic analysis will be expanded to gain insights both into the precise sequence requirements for the accurate and quantitative functioning of each element, and into the mechanism of interaction between them. Mutated TATA box sequences will also be examined by in vitro binding assays with purified, cloned, human TATA-binding protein (TBP), to determine if there is a correlation between the transcriptional activity measured in vivo, and the ability to bind TBP in vitro. Similar binding experiments will be performed with the putative initiator-binding protein TFII-I, and mutated INR sequences. The MLP contains several upstream activating elements, and sequence comparisons of different adenovirus serotypes show that the elements are conserved. This suggests either that each element plays an important specific role in the natural host, or that the ancestral progenitor acquired particular activating sequences by chance, and the activating functions have been conserved. The requirements for specific elements will be examined by replacing the wild type elements with others widely used in cellular and viral genes. Successful replacement would suggest that there is no specificity to the upstream elements, but that conservative selection has maintained the activating functions. If specificity is observed, it may imply that different elements are used in different cell types. Mutant viruses will be used to screen a variety of cell types derived from tissues likely to be infected with adenovirus in the natural host. Cells derived from the respiratory and gastro- intestinal tracts, and from lymphoid tissue, will be infected, and replicative and transcriptional abilities will be examined. If host range phenotypes are observed, the restrictive cells will be examined for the absence of one of the expected transcription factors, or for the presence of inhibitors, using both immunological techniques and gel retardation assays. Viruses with lowered transcription rates from the MLP, have a variety of contingent transcriptional and translational phenotypes. These presumably reflect alterations to the circuits of gene expression control normally acting at late times in infection, an area that is still poorly understood. These phenomena will be examined in detail by standard molecular virological techniques to determine if late gene expression down-regulates early gene expression, and whether this occurs in cis or in trans. Attempts will also be made to identify a non-viral protein prominently over-produced in cells infected with MLP-deficient viral mutants.
