The nucleotide sequences of infectious cloned DNAs 1 and 2 of a Kenyan isolate of cassava latent virus (CLV) have been determined. Five virus‐specific polyadenylated transcripts have been identified and mapped either to the viral or complementary sense DNAs of both components of the CLV genome, confirming that transcription is bidirectional on both DNAs. A major mRNA has been translated in vitro to yield a 30 000 mol. wt. product, which is precipitated by antibodies raised against whole virus, and has been mapped by both the S1 nuclease procedure and hybrid‐arrested translation to the long open reading frame (ORF) in the viral sense of DNA 1 which encodes the coat protein. Other transcripts were of sufficient size and appropriate origin to encode at least five potential products.

SUMMARYInfective clones of the Nigerian isolate of cassava latent virus (CLV) have been obtained. The apparent molecular weight of the capsid protein of this isolate is slightly higher than that produced in plants infected with cloned DNAs of the Kenyan isolate of CLV. Pseudorecombinant experiments using heterologous combinations of cloned DNAs have confirmed that the physical properties of the capsid protein are encoded on DNA 1 and at least some determinants of symptom induction are also located on this DNA. Comparison between the nucleotide sequences of the open reading frames encoding the two capsid proteins shows several nucleotide differences which affect the amino acid composition but which do not significantly alter the potential molecular weight of the product. In vitro translation of poly(A) ÷ RNAs shows that the differences in electrophoretic mobility are not due to differences in host-directed post-translational processing.

No abstract