Sequences of the S1 genes of the three serotypes of reovirus.

Abstract
The S1 genes of the 3 serotypes of reovirus were cloned and sequenced. The S1 genes encode protein .sigma.1, the protein against which serotype-specific neutralizing antibodies are directed; it is also the reovirus hemagglutinin and cell-attachment protein and is a major determinant of host range/tissue specificity and of the nature of the interaction of reovirus with cells of the immune system. The S1 genes of serotypes 1, 2 and 3 are 1458, 1442 and 1416 nucleotides long, respectively. They possess untranslated regions 13, 13 and 12 nucleotides long at their 5'' termini and 188, 229 and 36 nucleotides long at their 3'' termini. They possess 2 open reading frames. The 1st starts with a weak initiation codon and extends for 418, 399 and 455 codons, respectively; this is the size expected for the .sigma.1 proteins. The other reading frame starts at a strong initiation codon .apprx. 70 residues downstream from the 5'' terminus but extends for only .apprx. 120 codons, being terminated by 3 in-phase termination codons in all 3 genes. The proteins encoded by these short open reading frames are basic. The serotype 1 and 2 S1 genes are much more closely related to each other (28% homology) than to the serotype 3 S1 gene (5% and 9% homology, respectively). These figures are based on direct homology calculations, adjusted for 25% random coincidence. Serologic evidence and hydrophobicity profiles agree that the .sigma.1 proteins of serotypes 1 and 2 are much more closely related to each other (.apprx. 40% homology) than to that of serotype 3 (only .apprx. 20% homology). The fact that the serotype 1 and 2 S1 genes are much more closely related to each other than to the serotype 3 S1 gene is remarkable since for all other 9 reovirus genes the serotype 1 and 3 genes are much more closely related to each other than to the serotype 2 gene. Mechanisms that may effect this remarkable evolutionary pattern are discussed.