A DOUBLE HAIRPIN STRUCTURE IS NECESSARY FOR THE EFFICIENT ENCAPSIDATION OF SLEEN NECROSIS VIRUS RETROVIRAL RNA

We conducted a mutational analysis within the previously defined encap sidation sequence (E) for spleen necrosis virus (SNV), an avian retrov irus. We found that two regions are necessary for efficient SNV replic ation. The first region is a double hairpin structure as proposed by K onings et al. (1992, J. Virol., 66, 632 - 640); the second region is l ocated downstream of the hairpins. We showed further that the double h airpin structure is required for efficient SNV RNA encapsidation. Our work is the first to demonstrate, via linker-scanning and site-directe d mutagenesis, that a specific RNA secondary structure is required for the encapsidation of retroviral RNA. Analysis of a series of mutation s within the E region indicates (i) that preserving the secondary stru cture of the two hairpins is important for efficient encapsidation and (ii) that the stem regions of the hairpins contain specific sequences critical for encapsidation. Within the hairpins, the presence of at l east one of the two conserved GACG four-residue loops, but not the mod erately conserved bulge sequence of the first hairpin, is crucial for function. The function of the hairpins is independent of the relative order of the two hairpins. However, the two hairpins are not redundant and are not functionally identical. Replacement of SNV double hairpin sequences with those of Moloney murine leukemia virus (M-MLV) has no detectable effect on the replication of SNV-based retrovirus vectors w ith reticuloendotheliosis virus strain A (REV-A) helper virus. Further more, replacement of the entire E sequence of SNV with that of Moloney murine sarcoma virus (M-MSV) and M-MLV results in retroviral vectors that replicate as well as SNV vectors with wild type SNV E. This resul t indicates that the encapsidation sequences of M-MSV/M-MLV and SNV ar e not virus specific and that, during packaging of SNV and MLV RNA wit h viral proteins from REV-A, the encapsidation sequences are recognize d largely by their secondary or tertiary structures.