Sequence elements upstream of the 3' cleavage site confer substrate strength to the adenovirus L1 and L3 polyadenylation sites.

The adenovirus major late transcription unit is a well-characterized transcription unit which relies heavily on alternative pre-mRNA processing to generate distinct populations of mRNA during the early and late stages of viral infection. In the early stage of infection, two major late transcription unit mRNA transcripts are generated through use of the first (L1) of five available poly(A) sites (L1 through L5). This contrasts with the late stage of infection when as many as 45 distinct mRNAs are generated, with each of the five poly(A) sites being used. In previous work characterizing elements involved in alternative poly(A) site use, we showed that the L1 poly(A) site is processed less efficiently than the L3 poly(A) site both in vitro and in vivo. Because of the dramatic difference in processing efficiency and the role processing efficiency plays in production of steady-state levels of mRNA, we have identified the sequence elements that account for the differences in L1 and L3 poly(A) site processing efficiency. We have found that the element most likely to be responsible for poly(A) site strength, the GU/U-rich downstream element, plays a minor role in the different processing efficiencies observed for the L1 and L3 poly(A) sites. The sequence element most responsible for inefficient processing of the L1 poly(A) site includes the L1 AAUAAA consensus sequence and those sequences which immediately surround the consensus hexanucleotide. This region of the L1 poly(A) site contributes to an inability to form a stable processing complex with the downstream GU/U-rich element. In contrast to the L1 element, the L3 poly(A) site has a consensus hexanucleotide and surrounding sequences which can form a stable processing complex in cooperation with the downstream GU/U-rich element. The L3 poly(A) site is also aided by the presence of sequences upstream of the hexanucleotide which facilitate processing efficiency. The sequence UUCUUUUU, present in the L3 upstream region, is shown to enhance processing efficiency as well as stable complex formation (shown by increased binding of the 64-kDa cleavage stimulatory factor subunit) and acts as a binding site for heterogeneous nuclear ribonucleoprotein C proteins.