The papovavirus BK is widely distributed in the human population. Like other papovaviruses, it is tumorigenic in rodents. The overall genome structure of BK is similar to the well- characterized monkey papovavirus SV40, with early and late RNAs transcribed in opposite directions from a common transcriptional control region. Different BK isolates show rearrangements, deletions, and insertions in the control region, leading to alterations in the virus's tumorigenic potential. This control region is the only part of the BK genome with little sequence homology to SV40, and it seems likely that BK interacts with a different set of host cell transcription factors. We propose a number of experiments to try a understand how the BK control region works. Using DNase I footprinting and other assays of DNA binding, we will map the precise sites of interaction of host cell proteins, and then will determine whether the proteins that bind these sites are the same or different from transcription factors known to bind other viral and cellular promoter/enhancer regions. Using information from the binding assays, we will proceed to correlate sequence-specific binding with biological function. We will develop an in vitro transcription system dependent on some or all of the proteins identified as binding to the control region, and will in addition develop in vivo assays capable of separately measuring early and late promoter function, and transcriptional enhancer activity. We will contrast sequence-specific protein binding in BK (Dunlop strain) with that in other BK variants, with special attention to sites that may be formed by the hyper-variable junctions within the region of direct sequence repeats. Experiments along the lines proposed here should thoroughly illuminate the workings of the BK control region, and thus will allow rational interpretation of present and future observations relating to tumorigenicity or other pathogenic effects of BK virus and its variants.
