Minimum information content and formation of interacting ribonuclease fragment complexes

The degree to which amino acid sequence can be simplified with retention of conformational and functional properties has been investigated by semisynthesis using non-covalent fragment complexes of bovine pancreatic ribonuclease as test cases. Based on the ribonuclease S system, a set of synthetic model sequences was defined for the S-peptide (1-20) region which interacted productively with native S-protein (21-124). The most simple sequence, an eicosapeptide containing helix-favoring Ala residues at all positions except Glu 1 and 14, Phe 8, His 12, and Met 13, effected at least 15% of ribonuclease catalytic activity (versus native ribonuclease S) when added to S-protein in saturating amounts. The data for model S-peptides define an alpha-helical framework and specific side chains at positions 8, 12, and 13 as the core of sequence information necessary for S-peptide to effect a productive non-covalent complex with S-protein. Previous ribonuclease fragment studies also were used as a basis for making the productive, non-overlapping complex, (1-15):(21-111):(116-124). Addition of synthetic (1-15) and (116-124) to (21-111) led to a 3 degrees increase in Tm and 4% (versus ribonuclease A) catalytic activity. The three-fragment complex, with the beta-bend residues 112-115 deleted, exhibited significantly lower stability to thermal denaturation than did related two-fragment complexes. The potential use of three-fragment complexes related to the above is discussed for semi-synthetic sequence modeling concomitantly in the N- and C-terminal regions of ribonuclease.