Sequences at both termini of the 10 genes of reovirus serotype 3 (strain dearing)

Abstract

We have sequenced 60–90 residues at the 5′- and 3′-termini of the 10 genes of the Dearing strain of reovirus serotype 3. All 10 genes share a common tetranucleotide GCUA at the 5′-terminus and a pentanucleotide UCAUC at the 3′-terminus of their plus strands. All genes contain a translation initiation codon within 13–33 residues of the 5′-terminus; an initiation codon starting at position 6 in one of the genes is inoperative as its reading frame leads quickly to a termination codon. Several genes possess additional initiation codons, either in phase or out of phase, further downstream. All genes possess multiple termination codons in positions ranging from 80 to less than 10 residues from the 3′-terminus. Computer searches were carried out to discover homologies, symmetries, dyad symmetries, dinucleotide frequencies, Shine-Dalgarno sequences and other features relating to questions concerning the presence of possible protein binding sites (such as for encapsidation or for transcriptase recognition), RNA molecule assembly during the initial stages of reovirus morphogenesis, and control of frequency of translation. It was found that plus-stranded transcripts of reovirus genes, that is, the messenger RNA molecules in infected cells, show little or no tendency for stable hairpin loop formation. The feature that correlates best with high frequency of translation is possession of AG-rich regions upstream from and around the initiation codon. No dyad symmetries (ability to associate by base pairing) with sufficient specificity to account for specific gene assembly could be detected. The genes contain about 20 regions 20 to 30 residues long that exhibit highly significant homology or symmetry (that is, similarity in the same or opposite sense) with sequences in some other gene. Remarkably, some of these sequences, which are present in all genes except gene S1, are very similar to not only one other sequence, but to two or even three others; and some exhibit similarity to both same and opposite sense segences. However, apart from the terminal tetra- and pentanucleotide, the ends of the 10 genes possess no common sequences sufficiently similar for them to be interpreted as representing common protein-binding sites. If such sites exist, they must share features other than similarity of base sequence.