The discovery of retroviral reverse transcriptase by Howard Temin and David Baltimore in 1970 revolutionized the field of molecular biology. Retroviruses have since become an invaluable research tool. At the same time, retroviruses continue to pose significant public health risk in terms of being causative agents for particularly devastating diseases like AIDS. Understanding the molecular mechanisms underlying essential functions in retroviruses is of immense interest. Rev-like proteins are essential regulatory proteins that function to mediate nuclear export of incompletely spliced mRNAs in retroviruses such as HIV-1. Following translation, HIV-1 Rev localizes to the nucleus and binds a unique sequence in the viral RNA, termed the Rev-responsive element (RRE). HIV-1 Rev subsequently multimerizes along the RNA and interacts with cellular Crm1 to export the RNA-protein complex to the cytoplasm. The Rev protein of equine infectious anemia virus (EIAV) is atypical in terms of organization of functional domains and the presence of a bipartite RNA binding domain (RBD). It was previously unknown how the bipartite RBD, made up of two short arginine-rich motifs (ARM1 and ARM2), bound the RRE.

To gain insight into the topology of the bipartite RNA binding domain, a computational approach was used to model the tertiary structure of EIAV Rev. Computational models suggested that ARM1 and ARM2 do not form a single RNA binding interface on the Rev monomer. A coiled-coil region was identified in the Rev sequence and computationally characterized. Critical residues in the coiled-coil were shown to be essential for Rev dimerization, and dimerization mutants were deficient for RNA binding. Together, these results suggest that EIAV Rev RNA binding requires dimerization to juxtapose ARM-1 and ARM-2; dimerization is mediated by a coiled-coil motif not previously reported in other Rev-like proteins, including HIV-1 Rev.

Since Rev-like proteins are functionally homologous, it was important to determine whether coiled-coil motifs are predicted for other members and the extent to which structural features are shared. Therefore, a comparative analysis of predicted secondary structural elements was performed in a phylogenetics framework. A common domain architecture was found in Rev-like proteins, with the single exception of EIAV Rev. In addition, the target RNA binding site of all Rev-like proteins was located in the 3' half of the viral genome. A common pattern of two alpha helices was observed among Rev proteins of the primate lentiviruses. The Rev proteins of non-primate lentiviruses and the Rev-like proteins of betaretroviruses contained more alpha helical segments, which also showed common patterns of distribution in the protein. Deltaretrovirus Rev-like proteins lacked predicted alpha helices or beta sheets. Coiled-coils were found in all lentivirus Rev groups but not in all members of each group. Coiled-coils were also found in two betaretrovirus Rev-like proteins; in contrast, coiled-coils were not found in any of the Rev-like proteins of deltaretroviruses. Phylogenetic reconstruction of the ancestral state of coiled-coils for all Rev-like proteins suggests a single origin followed by two major losses, leading to the absence of coiled-coils in some lentivirus groups and in all deltaretroviruses. These results reveal similarities among Rev-like proteins despite significant sequence divergence, a possible result of common ancestry. The fact that coiled-coils have been maintained across divergent Rev-like proteins suggests they play an important role in function, presumably oligomerization and other key protein-protein interactions. Some retroviruses, including HIV-1, may have evolved alternate sequences to replace coiled-coils motifs in their Rev-like proteins.