Genesis of Sindbis virus by in vivo recombination of nonreplicative RNA precursors.

Genetically engineered RNA transcripts coding for various Sindbis virus (SIN) genes were used to study structure and sequence requirements of RNA recombination in BHK cells. Three different groups of RNA transcripts were made: (i) RNAs which retain the ability to replicate and which carry sequences coding for either viral polymerase or viral structural proteins; (ii) RNAs which lack the complete 3' end of the SIN genome and thus are incapable of replicating; and (iii) RNAs which lack the complete 5' end of the SIN genome and also are incapable of replicating. BHK cells were transfected with specific combinations of these precursor RNAs, and virus production and RNA synthetic abilities of the released virus were determined. We demonstrate in vivo generation of infectious SIN by fusion of (i) replicative RNAs to nonreplicative RNAs and (ii) two nonreplicative RNA precursors. Both homologous and nonhomologous types of recombinations were observed. In the homologous type of recombination, a 694-nucleotide overlap at the crossover region of the first pair of precursors resulted in the addition of an A residue converting the UAG stop codon of nonstructural protein P4 to a UAA stop codon. In the nonhomologous type of recombination, the crossover sites contained deletion of up to 76 nucleotides from one of the precursors and complete preservation of junction sequence from the other precursor. This is also the first report that a cytoplasmic RNA virus can be generated from biologically nonreplicative RNA precursors. These results have implications for initiation of viral RNA synthesis and recombination between RNA viral genomes in general. We favor template switching as a mechanism for the fusion events described here and suggest inclusion of polymerase scanning of diverse nonreplicative RNAs as an inherent feature of the copy choice model of RNA recombination. Very importantly, the facile nature of RNA recombination occurring between nonreplicative RNA precursors should speed up the production and analysis of targeted mutants of SIN and possibly other RNA viruses.