Common sequence elements in structurally unrelated genomes of defective interfering Semliki Forest virus.

The polymerase chain reaction was used to amplify defective interfering (DI) Semliki Forest virus (SFV) genomes from tissue culture preparations which had in vitro interfering and in vivo mouse-protecting activity. The products were molecularly cloned and sequenced (pSFVDI-6, 2146 nt and pSFVDI-19, 1244 nt). Comparison with the sequence of the SFV genome showed that both were derived from three noncontiguous regions: the 5' terminus including the start of the nsP1 coding region, part of the nsP2 coding region, and the 3' untranslated region. Both clones possessed open reading frames, including one with the potential of encoding truncated versions of nsP1. Transcribed RNAs from pSFVDI-6 and pSFVDI-19 were translated in an in vitro system in polypeptides of M(r) 27,000 and 25,000, respectively. Although pSFVDI-6 and pSFVDI-19 did not possess the rearranged sequences or extensive repeat regions of previously cloned DI SFV genomes (Lehtovaara et al., Proc. Natl. Acad. Sci. USA 78, 5353-5357, 1981; Lehtovaara et al., J. Mol. Biol. 156, 731-748, 1982), they were derived from similar regions of the SFV genome, suggesting a common sequence requirement for DI SFV particles.