Analytical study of rat retrotransposon VL30 RNA dimerization in vitro and packaging in murine leukemia virus.

A sequence of the rat retrotransposon virus-like 30 S RNA (VL30) located next to the 5' end of the Harvey murine sarcoma virus (HaMSV) genome was recently found to form stable dimeric RNA in vitro and to direct the efficient packaging of VL30-derived recombinant RNAs into MuLV virions. To study the structure-function relationships of the rat VL30 dimerization-encapsidation signal (E/DLS), we have performed biochemical and genetic studies of rat VL30 RNA dimerization in vitro. The results show that temperature and specific cation/RNA interactions are important for VL30 dimerization in vitro. VL30 RNA dimerization is optimal at 55 degrees C and Li+ dramatically enhances the stability of VL30 dimeric RNA. In addition, a genetic analysis of VL30 RNA dimerization reveals that a 5' G-rich sequence is critical for dimer formation and that a UGUCUUGUC repeat contributes to VL30 dimer stability. Interestingly enough, substitution of an A for a G in the 5' G-rich sequence is sufficient to abolish VL30 RNA dimerization in vitro. Taken together, these biochemical and genetic data indicate that dimerization of VL30 RNA involves non-canonical base-pairings and possible purine-purine interactions. Nucleocapsid protein NCp10 of murine leukemia virus (MuLV), a gag-encoded protein that is tightly associated with genomic RNA in the virion core, has been shown to have nucleic acid binding and annealing activities. Here we report that the viral NCp10 protein is able to bind tightly to annealing activities. Here we report that the viral NCp10 protein is able to bind tightly to the retrotransposon VL30 RNA and to activate its dimerization. Moreover, mutations in the 5' G-rich sequence of the VL30 dimerization sequence impaired NCp10 binding to RNA. Recombinant MLV-VL30 vectors with mutations in the VL30 dimerization sequence were constructed. Results obtained in vivo clearly show that the mutations that had a deleterious effect on the packaging of MLV-VL30 retroviral vector in vivo were those that impaired VL30 RNA dimerization and interactions with NCp10 in vitro, even the single mutation in the 5' G-rich region. Therefore, these findings suggest that packaging of VL30 RNA into MuLV virions requires specific interactions between RNA dimerization sequences and viral NC protein molecules.