Characterization of cassava vein mosaic virus: a distinct plant pararetrovirus

Cassava vein mosaic virus (CVMV) was found to be widespread throughout the north-eastern region of Brazil. The complete sequence of CVMV was determined, and the genome was 8158 bp in size. A cytosolic initiator methionine tRNA (tRNA met1)-binding site that probably acts as a primer for minus-strand synthesis was present. The genome contained five open reading frames that potentially encode proteins with predicted molecular masses of 186 kDa, 9 kDa, 77 kDa, 24 kDa and 26 kDa. The putative 186 kDa protein had regions with similarity to the zinc finger-like RNA-binding domain that is a common element in the capsid proteins and similarity to the intercellular transport domain of the plant pararetroviruses. The predicted 77 kDa protein had regions with similarity to aspartic proteases, reverse transcriptase and RNase H of pararetroviruses. This gene order was confirmed by the amplification of similar PCR products from total DNA extracted from CVMV-infected cassava plants. The genomic organization of CVMV was different from the organization of either the caulimoviruses or badnaviruses. In comparisons of the regions with the reverse transcriptase motif, CVMV was grouped between the caulimoviruses and badnaviruses. It appears that CVMV is distinct from the other well-characterized plant pararetroviruses.