Two-electron reduced glutathione reductase from yeast reacted with iodoacetamide is alkylated almost exclusively in the nascent thiol nearer the amino terminus of the protein. The charge-transfer absorbance, maximal at 530 nm, characteristic of the two-electron reduced enzyme is not lost as the alkylation proceeds, and the product has a spectrum virtually identical with that of the two-electron reduced enzyme. This observation demonstrates that the thiol alkylated is not the charge-transfer-donor thiolate which interacts with the FAD. The spectrum of the monoalkylated derivative is stable in the presence of oxidized glutathione, indicating that the charge-transfer-donor thiol is not involved in interchange with the substrate in the native enzyme. Thus, the nascent thiols produced upon two-electron reduction of glutathione reductase have distinct functions, interchange with the substrate and interaction with the FAD. Treatment of the monoalkylated derivative with the apolar phenylmercuric acetate eliminates the charge-transfer interaction. The spectrum of the resulting species is similar to that of the oxidized enzyme but less resolved and blue shifted by 10 nm. The dependence on pH of the absorbance associated with the thiolate to FAD charge-transfer interaction in native two-electron reduced glutathione reductase is biphasic, with pK values at approximately 4.8 and 7.4. By analogy with glyceraldehyde-3-phosphate dehydrogenase and papain, these data indicate that the thiolate is stabilized by an adjacent basic residue. The pK 7.4 is associated with the titration of the base to give the ion pair, and the pK of 4.8 is associated with the titration of the thiolate. Unlike lipoamide dehydrogenase, glutathione reductase is sufficiently stable to allow titration with dithionite at pH 3.7. The spectrum at this pH is essentially the same as that of the monoalkylated derivative treated with phenylmercuric acetate. The changes with pH are completely reversible.

Tyr114 and Tyr197 are highly conserved residues in the active site of human glutathione reductase, Tyr114 in the glutathione disulfide (GSSG) binding site and Tyr197 in the NADPH site. Mutation of either residue has profound effects on catalysis. Y197S and Y114L have 17% and 14% the activity of the wild-type enzyme, respectively. Mutation of Tyr197, in the NADPH site, leads to a decrease in Km for GSSG, and mutation of Tyr114, in the GSSG site, leads to a decrease in Km for NADPH. This behavior is predicted for enzymes operating by a ping-pong mechanism where both half-reactions partially limit turnover. Titration of the wild-type enzyme or Y114L with NADPH proceeds in two phases, Eox to EH2 and EH2 to EH2-NADPH. In contrast, Y197S reacts monophasically, showing that excess NADPH fails to enhance the absorbance of the thiolate-FAD charge-transfer complex, the predominant EH2 form of glutathione reductase. The reductive half-reactions of the wild-type enzyme and of Y114L are similar; FAD reduction is fast (approximately 500 s-1 at 4 degreesC) and thiolate-FAD charge-transfer complex formation has a rate of 100 s-1. In Y197S, these rates are only 78 and 5 s-1, respectively. The oxidative half-reaction, the rate of reoxidation of EH2 by GSSG, of the wild-type enzyme is approximately 4-fold faster than that of Y114L. These results are consistent with Tyr197 serving as a gate in the binding of NADPH, and they indicate that Tyr114 assists the acid catalyst His467'.

The flavoenzyme glutathione reductase catalyzes the NADPH-dependent reduction of glutathione disulfide, yielding two molecules of glutathione. The oxidation-reduction potentials, Eox/EH2 (two-electron reduced enzyme), for yeast, Escherichia coli, and human glutathione reductase have been determined between pH 6.0 and 9.8 relative to the nonphysiological substrate couple NAD+/NADH and were found to be -237, -243, and -227 mV (+/-5 mV) at pH 7.0 and 20 degreesC, respectively. The potential as a function of pH demonstrated slopes of -51, -45, and -42 mV/pH unit, respectively, at low pH and -37, -31, and -34 mV/pH unit, respectively, at high pH. The change in slope indicated pKa values of 7.4, 8.5, and 7.6, respectively. The slopes indicate that two protons are associated with the two-electron reduction of Eox at low pH and that only one proton is involved with the two-electron reduction of Eox at high pH, provided that the effects of nearby titratable residues are considered in the data analysis. The influence of four such groups, Cys50, Cys45, His456', and either Tyr107 or the flavin-(N3), has been included (residue numbering refers to the yeast sequence). The enzyme loses activity upon deprotonation of the acid-base catalyst at high pH. Since the pKa ascribed to the EH2-to-EH- ionization is lower than the pKa of the acid-base catalyst, both the EH2 and EH- forms of glutathione reductase must be catalytically active, in contrast to the closely related enzyme lipoamide dehydrogenase, for which only EH2 is active.