Transcription termination and polyadenylation in retroviruses.

The provirus structure of retroviruses is bracketed by long terminal repeats (LTRs). The two LTRs (5' and 3') are identical in nucleotide sequence and organization. They contain signals for transcription initiation as well as termination and cleavage polyadenylation. As in eukaryotic pre-mRNAs, the two common signals, the polyadenylation signal, AAUAAA, or a variant AGUAAA, and the G+U-rich sequence are present in all retroviruses. However, the AAUAAA sequence is present in the U3 region in some retroviruses and in the R region in other retroviruses. As in animal cell RNAs, both AAUAAA and G+U-rich sequences apparently contribute to the 3'-end processing of retroviral RNAs. In addition, at least in a few cases examined, the sequences in the U3 region determine the efficiency of 3'-end processing. In retroviruses in which the AAUAAA is localized in the R region, the poly(A) signal in the 3' LTR but not the 5' LTR must be selectively used for the production of genomic RNA. It appears that the short distance between the 5' cap site and polyadenylation signal in the 5' LTR precludes premature termination and polyadenylation. Since 5' and 3' LTRs are identical in sequence and structural organization yet function differently, it is speculated that flanking cellular DNA sequences, chromatin structure, and binding of transcription factors may be involved in the functional divergence of 5' and 3' LTRs of retroviruses.