Evolution of reovirus genes: a comparison of serotype 1, 2, and 3 M2 genome segments, which encode the major structural capsid protein mu 1C.

The sequences of the M2 genome segments of reovirus serotypes 1 and 2 are presented and compared with that of the cognate genome segment of reovirus serotype 3. Genome segment M2 encodes protein mu 1, a cleavage product of which, mu 1C, is the major constituent of reovirus particles. The three M2 genome segments exhibit a serotype 1:3 relatedness pattern: the serotype 1 and 3, 1 and 2, and 2 and 3 genome segment pairs exhibit 15, 23, and 23% nucleotide mismatches, respectively. The vast majority of these mismatches (about 87%) occur in third base codon positions that do not cause amino acid changes; as a result the three mu 1 proteins are very highly related (about 97%). The mu 1 proteins are acidic proteins, low in cysteine, histidine, and methionine, and rich in proline; and they possess a rather low predicted alpha-helix content of 27%. The site where protein mu 1 is cleaved to mu 1C is predicted to be between residues 42 and 43 in a highly conserved portion of the molecule. The three M2 genome segments are related significantly more closely than the three S3 genome segments, and much more closely than the three S1 genome segments. We have analyzed the evolutionary divergence patterns of these three genome segments. The rate of mismatch accumulation in third base codon positions is roughly the same for all three genome segments, but the rates of mismatch accumulations in first, and particularly in second, base codon positions are quite different. For the S1 genome segments there is little difference between the rates of mismatch accumulations in all three codon positions, which indicates that retention of function is compatible with very extensive structural flexibility. By contrast, the rates of mismatch accumulations in first and second base codon positions are far less for the S3 genome segments, which encode the nonstructural protein sigma NS, and much lower still for the M2 genome segments, which suggests the existence of very stringent structural constraints for protein mu 1C, which might be expected of a protein that is a capsomer component.