On the information content of cytochrome c.

Abstract A previously published mathematical method, utilizing functionally equivalent or synonymous residues, is applied to the calculation of the information content of the genetic system required to specify at least one residue synonymous at a given site in cytochrome c . Utilizing only those residues known to be synonymous, the information content of the genetic message for the 101 sites is 373·83 bits or 3·701 bits per residue. A prescription has been written, based on a paper by Grantham (1974) which allows one to include those residues which are predicted to be synonymous due to a similarity in composition, polarity and volume. Including such residues, one finds that the information content of the genetic message for 101 sites is 298.21 bits or 2.953 bits/residue. These figures are substantially less than the maximum information content of a doublet genetic code or 4·00 bits/residue. A doublet code can code more than 16 residues but we find that cytochrome c could be coded with only 15. The suggestion of Jukes (1965) that a doublet code may have been used by vanished forms of life must be taken seriously since cytochrome c could be represented in such a way. We suggest a primeval doublet code similar to that of Jukes and show how it could have evolved to the modern degenerate triplet code. A triplet code which is non-degenerate is more subject to error than a doublet code. The redundance of a degenerate triplet code provides some error protection and therefore degeneracy must have evolved simultaneously. The vulnerability of the human and yeast cytochrome c sequences to error by single base interchange is calculated including the redundance due to the synonymous residues. The error protection due to genetic code degeneracy and to the synonymous residues was available from the earliest stages in the origin of life. These factors prove to be very much weaker than modern biological factors such as diploidy and repair processes. It is concluded that the modern triplet code was fully fixed before the appearance of these purely biological error correction mechanisms.