Directed evolution of proteins is a powerful aid in enzyme design.  The usual objectives of directed evolution are to confer properties on the targeted proteins that do not exist in naturally evolved ones.  Typical goals include increased thermostability, catalytic activity in non-aqueous solvents, or acquisition of an enzymatic activity that does not exist naturally.  E. coli aspartate aminotransferase (AATase) and tyrosine aminotransferase (TATase) are 43% identical in sequence, and undoubtedly had a common progenitor.  Our initial objective was to understand how the amino acid substitutions elicited from AATase under laboratory directed evolution conditions compare with those evolved naturally.  The experiments were analyzed using Venn diagrams that indicated that the laboratory changes were found in a set of amino acids that are conserved in AATase, but not in TATase1. Venn diagrams constructed from more distantly related enzymes reveal an additional set in which amino acid positions are conserved in both the initiating and target protein but as different residues.  We suggest that this “strong forcing set” contains the amino acid substitutions that are most influential for substrate specificity.  This strategy has now been applied to convert E. coli malate dehydrogenase (MDH) to a lactate dehydrogenase (LDH). Five substitutions were incorporated into E. coli MDH.  The new variant had > 105 reduced activity for oxalacetate and > 109 increased activity with pyruvate.  The specificity is thus changed by ca. 1014.
1 Rothman, S. C. and Kirsch, J. F.: J. Mol. Biol. 327, 593-608 (2003)