DETERMINATION OF THE AMINO-ACID-REQUIREMENTS FOR A PROTEIN HINGE IN TRIOSEPHOSPHATE ISOMERASE

We have determined the sequence requirements for a protein hinge in tr iosephosphate isomerase. The codons encoding the hinge at the C-termin us of the active-site lid of triosephosphate isomerase were replaced w ith a genetic library of all possible 8,000 amino acid combinations. T he most active of these 8,000 mutants were selected using in vivo comp lementation of a triosephosphate isomerase deficient strain of E. coli , DF502. Approximately 3% of the mutants complement DF502 with an acti vity that is above 70% of wild-type activity. The sequences of these h inge mutants reveal that the solutions to the hinge flexibility proble m an varied. Moreover, these preferences are sequence dependent; that is, certain pairs occur frequently. They fall into six families of sim ilar sequences. In addition to the hinge sequences expected on the bas is of phylogenetic analysis, we selected three new families of 3-amino -acid hinges: X(A/S)(L/K/M), X(aromatic/beta-branched)(L/K), and XP(S/ N). The absence of these hinge families in the more than 60 known spec ies of triosephosphate isomerase suggests that during evolution, not a ll of sequence space is sampled, perhaps because there is no neutral m utation pathway to access the other families.