Analysis of the Relationship between the Structure and Function of the HIV-1 Rev Response Element (RRE)

HIV-1 is a retrovirus of global health interest as the causative agent of Acquired Immunodeficiency Syndrome (AIDS). It is also used as a molecular tool to study various eukaryotic cellular processes. For instance, major discoveries on the Crm1 cellular nuclear export were made using the HIV Rev protein and its RNA binding partner, the Rev Response Element (RRE). The RRE is a cis-acting RNA element with multiple stem-loops present in all intron-retaining HIV mRNAs. Binding of the Rev protein to the primary binding site, and Rev multimerization on other regions of the RRE, is required for the nucleo-cytoplasmic export of these mRNAs. This is an essential step in HIV replication. However, the precise secondary structure of the HIV-1 RRE remains controversial. Studies have reported that the RRE has either 4 or 5 stem-loops, which differ only in the rearrangement of regions that lie outside of the primary Rev binding site. To understand the role played by these regions, we have examined the relationship between these two structures and Rev-RRE activity.

In vitro transcribed NL4-3 RRE was found to migrate as a “doublet” band on a native polyacrylamide gel. Using in-gel Selective 2’ Hydroxyl Acylation analyzed by Primer Extension (SHAPE), we found that one of these bands consisted of an RRE with a 5 stem-loop structure, whereas the other was a 4 stem-loop structure. Thus, our data demonstrate, for the first time, that the NL4-3 RRE exists in two alternative structures.

To study the significance of these alternative structures, we made RRE mutants predicted to allow only one or the other of the structures to form. The predictions were confirmed using SHAPE. We then compared the activity of the two forms to each other and to the wt RRE. Analysis of the complexes that each RRE formed with purified Rev protein in vitro showed no significant difference in Rev binding affinity between the two structures. However, we observed differences in the migration rates of these complexes on native gels, suggesting structural differences. The RREs were also tested for their abilities to promote viral replication, by inserting each RRE into the Nef region of an RRE-defective provirus which contained RRE mutations that do not change the Env protein. Growth kinetics and competition assays showed that the virus with the 5 stem-loop RRE had a higher replicative fitness than the virus with either the wt or the 4 stem-loop RRE. Between the wt and the 4 stem-loop RRE-containing viruses, the virus with the wt RRE appeared more fit. These results suggest that HIV may use two alternative RRE secondary structures to modulate replication, potentially allowing adaptation to environmental demands in time and/or space, analogous to the use of riboswitches in bacteria.