Abstract Body

Background: HIV replication is a highly regulated process, and the binding of the viral Rev protein to the Rev Response Element (RRE) is essential for the nucleocytoplasmic export of mRNAs that retain introns and thus for packaging of genomic RNA. Previous studies have shown that there is variation in Rev and RRE sequence both within and between patients. We previously assessed the activity of the Rev-RRE system in viruses isolated from patients. Variation of Rev/RRE function was observed between patients and within a patient as the infection progressed. In the present study, we sought to examine whether Rev/RRE functional variation could help to explain changes in the relative prevalence of HIV subtypes in populations.

Methods: A functional assay was developed, which utilizes an HIV vector system that packages HIV vector RNA into particles in a Rev-dependent fashion. The virus particles formed in this system transduce hygromycin resistance to target cells, allowing Rev/RRE activity to be read-out by counting the number of resistant colonies. Rev and RRE sequences can be tested in this system to measure the function of naturally occurring or artificial Rev-RRE pairs. Eight sequences were selected from the Los Alamos database representing two subtype A, four AG, and two G examples. Four subtype B patient isolates were also included.

Results: There was a 24-fold difference in functional activity between the most and least active naturally occurring Rev/RRE combinations (cognate pairs). There was no clustering within withsubtypes. When selected Rev and RRE sequences were tested with NL4-3 RRE and Rev, the absolute activity could not be predicted from the activity of the cognate pair. In a linear regression analysis, there was low correlation of RRE activity (R=0.32) but high correlation of Rev activity (R=0.95) with the functional activity of the original cognate pair.

Conclusions: Rev/RRE functional activity varied dramatically between naturally occurring cognate pairs. This functional variation appeared to be primarily driven by differences in Rev. The clinical significance of this finding is still unclear, but our results demonstrate that there are clear differences in the Rev activity of circulating viruses. This may play a role in viral fitness and replication dynamics and be reflective of immune pressure. Identification of the most efficient Rev/RRE combination would also likely be of use in optimizing lentiviral vector systems for gene therapy.