UCSF

Many genomes contain overlapping genes in which nucleotides are shared in alternative open reading frames. Despite the fact that this phenomena occurs quite frequently in viruses, the manner in which shared nucleotide sequence impacts protein evolution has not yet been fully elucidated. We propose two simple models: 1) A "selfish" arrangement in which one protein dictates the coding sequence of the other and 2) A "compromising" model in which a group of nucleotides encodes critical residues in both frames, thereby forming a viral Achilles' heel that must satisfy multiple selection pressures.

To test these models, we examine human immunodeficiency virus type 1 (HIV-1) which contains substantial coding overlap including the essential, regulatory genes tat< and rev. Tat and Rev play known and critical roles in the production and processing of the viral RNA, making them ideal candidates for study. Here we combine statistical analyses of existing patient data, a residue-resolution, functional dissection of Tat and Rev, and directed evolution experiments that decouple the overlapped regions to demonstrate that the selfish organization of the tat/rev overlap minimizes the constraining effects of overlapped reading frames. This arrangement of overlapped genes may extend to other overlapped genes and also represent an important mechanism for the creation of genomic novelty and post-transcriptional regulation in small genomes.