An approach to optimizing the active site in a glutathione transferase by evolution in vitro.

A glutathione transferase (GST) mutant with four active-site substitutions (Phe(10)-->Pro/Ala(12)-->Trp/Leu(107)-->Phe/Leu(108)-->Arg) (C36) was isolated from a library of active-site mutants of human GST A1-1 by the combination of phage display and mechanism-based affinity adsorption [Hansson, Widersten and Mannervik (1997) Biochemistry 36, 11252-11260]. C36 was selected on the basis of its affinity for the transition-state analogue 1-(S-glutathionyl)-2,4, 6-trinitrocyclohexadienate. C36 affords a 10(5)-fold rate enhancement over the uncatalysed reaction between reduced glutathione and 1-chloro-2,4-dinitrobenzene (CDNB), as evidenced by the ratio between k(cat)/K(m) and the second-order rate constant k(2). The present study shows that C36 can evolve to an even higher catalytic efficiency by an additional site-specific mutation. Random mutations of the fifth active-site residue 208 allowed the identification of 18 variants, of which the mutant C36 Met(208)-->Cys proved to be the most active form. The altered activity was substrate selective such that the catalytic efficiency with CDNB and with 1-chloro-6-trifluoromethyl-2,4-dinitrobenzene were increased 2-3-fold, whereas the activity with ethacrynic acid was decreased by a factor of 8. The results show that a single-point mutation in the active site of an enzyme may modulate the catalytic activity without being directly involved as a functional group in the enzymic mechanism. Such limited modifications are relevant both to the natural evolution and the in vitro redesign of proteins for novel functions.