Converting trypsin to chymotrypsin: ground-state binding does not determine substrate specificity

Rat trypsin II has been converted to a protease with chymotrypsin-like substrate specificity [Hedstrom, L., et al. (1994) Biochemistry (preceding paper in this issue)]. The key alteration in this conversion is the exchange of two surface loops for the analogous loops of chymotrypsin. k(inact)/Ki for the inactivation of chymotrypsin, trypsin, a trypsin mutant with poor activity (D189S), and the chymotrypsin-like mutants Tr-->Ch[S1+L1+L2] and Tr-->Ch[S1+L1+L2+Y172W] by Suc-Ala-Ala-Pro-Phe-chloromethylketone correlates with kcat/Km for hydrolysis of Suc-Ala-Ala-Pro-Phe-AMC. k(inact)'s for the inactivation of Tr-->Ch[S1+L1+L2] and Tr-->Ch[S1+L1+L2+Y172W] are comparable to that of chymotrypsin, while Ki's were much higher. Ki for the inhibition of these enzymes by the transition-state analog MeOSuc-Ala-Ala-Pro-boro-Phe also correlates with kcat/Km for hydrolysis of Suc-Ala-Ala-Pro-Phe-AMC. These results suggest that the surface loops stabilize the transition state for hydrolysis of chymotrypsin substrates by improving the orientation of bound substrates relative to the catalytic residues. Lastly, trypsin and chymotrypsin have comparable affinities for proflavin, while the Kd for the Tr-->Ch[S1+L1+L2+Y172W]-proflavin complex is 10-fold higher. No proflavin binding could be observed for either D189S or Tr-->Ch-[S1+L1+L2], which suggests that the S1 binding pockets of these two mutant enzymes are deformed. This work confirms that enzyme specificity is expressed in the chemical steps of the reaction rather than in substrate binding.