Sequences at the capped 5'-ends of polyoma virus late region mRNAs: an example of extreme terminal heterogeneity.

We have localized with respect to the genomic DNA sequence the capped 5'-termini of polyoma virus late region mRNAs. A minimum of fifteen different purine termini were found within a 94 base pair region (66.36 to 68.12 map units, nt 5075-5168) immediately preceding the sequence determining the late region mRNA leader repeat (1-3). The most common termini occur at nearly every possible purine within a 25 bp sequence proximal to the leader repeat unit. These do not bear the usual positional relationship to a sequence resembling the 'TATA' box consensus. Deletion mutants lacking minor cap sites and sequences upstream from the principal cap sites were viable. A deletion mutant lacking one of the principal cap sites formed small plaques, while a slightly larger deletion further impinging on the principal cap site region was non-viable. The principal cap sites, which we assume to be transcriptional initiation points, are included in a DNaseI hypersensitive region of polyoma virus chromatin (4).