Cloning of the maize rough dwarf virus genome: molecular confirmation of the plant-reovirus classification scheme and identification of two large nonoverlapping coding domains within a single genomic segment.

The segmented double-stranded RNA genome of maize rough dwarf virus, a plant-infecting reovirus of the genus Fijivirus, was cloned and partially characterized. Nucleotide sequence analysis of full-length cDNA clones corresponding to genomic segments S6, S7, and S8 revealed each segment to contain the conserved terminal oligonucleotide sequences (+) 5' AAGUUUUUU------UGUC 3' and adjacent, segment-specific, regions of inverted complementarity (inverted repeats), a structural motif previously reported for members of the genus Phytoreovirus. Genomic segment S6 was completely sequenced and found to consist of 2193 base pairs. Computer analysis indicated that the coding strand contained two large nonoverlapping open reading frames consisting of 363 and 310 codons and located in the 5'- and 3'-terminal domains, respectively. This was confirmed by cell-free translation studies with synthetic transcripts and denatured genomic RNA. However, only the product of the 5'-proximal open reading frame, a 40-kDa polypeptide, was efficiently expressed in vitro from the full-length S6 coding strand. This represents the first case in which a reovirus genomic segment was found to contain two large open reading frames in a nonoverlapping configuration, suggesting possible alternative strategies for regulation of gene expression by members of this genus. The combined results provide a molecular confirmation of the current classification scheme for plant-infecting reoviruses. Furthermore, the fact that the same terminal structural motif is conserved across genera provides additional evidence that these elements serve an important functional role during genome transcription or replication.