Sequence Analysis Of Proteins With Symmetric Tertiary Structures

On the basis of studies on ribonuclease, in the mid 1950's Anfinsen proposed that the tertiary structure of a protein is encoded in its amino acid sequence. However, we found that in most cases symmetries of the tertiary structures of many proteins can only be seen on the structural level. The common methods barely discover or visually show the symmetry on sequence level. Based on the fact that complex protein folds may be evolved from short peptide ancestors via a series of duplications and the recurrence method, we proposed a modified recurrence analysis method. By redefining the degree of sequence similarity s, the modified recurrence plots reveal the hidden repeats directly from sequences. We analyzed the typical symmetric protein family: beta-trefoil, and got good pseudo three-fold sequence symmetries characteristic of their structures. To prove the generality of this rule, we calculated the modified recurrence plots of other proteins with symmetric tertiary structures in αclass, βclass and αβclass. We found that most of their primary sequences show the same symmetries as their structures: from pseudo 2-fold to pseudo 5-fold symmetry. Moreover, we studied a representative sample of small αprotein and found the same internal sequence repetitions as their tertiary structures and their key amino acid residues. Considering that different amino acid residues share some similar physicochemical properties, we simplified and classified the 20 kinds of amino acid residues. We re-calculated the proteins of beta-trefoil family and four-propellor family based on their frequencies of amino acids in β?strands, and all their primary sequences show clearer pseudo 3-fold and 4-fold symmetry, respectively. Finally, we predicted the tertiary structure of a protein with unknown structure using this method together with the network prediction way and we initially presume it as a(βα)8-barrel architecture.