Glutathione S-transferase A1-1: Affinity labeling of the nonsubstrate site and probing interface interactions using mutagenesis

The affinity label 17beta-iodoacetoxy-estradiol-3-sulfate was used to elucidate binding determinants of the nonsubstrate steroid binding site of glutathione S-transferase Al-1. This affinity label inactivated the enzyme to 60% residual activity with an incorporation of 1 mol compound per mol enzyme dimer. Only Cys18, on a single subunit, is modified. Molecular modeling studies show support for the binding site being located in the cleft between the subunits with the sulfate moiety directing the binding of 17beta-iodoacetoxy-estradiol-3-sulfate.; Subunit interface residues of glutathione S-transferase A1-1 were mutated to study the importance of specific residues at the interface and their role in dimerization. Mutations were made in the hydrophobic ball and socket and the Arg/Glu region. Of the residues studied, it was found that Phe52 and Arg69 are the two most important residues at the interface. The dissociation constants (Kd) for mutant enzymes F52A and R69E are 10 fold higher than that of wild type, reflecting a shift in the equilibrium towards monomer. Using the Kds (determined in the presence of substrate) and the concentration in terms of subunits, we can calculate the monomer. In the case of R69E, the monomer appears to retain partial activity (8 mumol/min/mg).; Heterodimers of wild type A1 subunits and A1 subunits with mutations at the subunit interface were produced and used to determine whether the activity of one subunit will affect the activity of the other subunit. Heterodimers were formed between wild type and R69E or R69Q or E97Q or F136A, but no heterodimer was recovered for wild type and F52A. Heterodimers of wild type/R69Q and wild type/F136A had an experimental activity very close to that expected for independent active sites. However, wild type/R69E and wild type/E97Q heterodimers, the activity was much less than that expected for independent active sites. In both cases, this may be due to a small distortion of the interface that is transmitted to the wild type active site, resulting in a lower activity. Based on all the cases, we conclude that the nature of the interactions at the interface is one determinant of whether the active sites function independently or affect one another.