Inhibitors of Glutathione Reductase as Potential Antimalarial Drugs Kinetic Cooperativity and Effect of Dimethyl Sulphoxide on Inhibition Kinetics

Abstract

We have developed inhibitors of glutathione reductase that improve on the inhibition of literature lead compounds by up to three orders of magnitude. Thus, analogues of Safranine O and menadione were found to be strong, reversible inhibitors of yeast glutathione reductase. Safranine O exhibited partial, uncompetitive inhibition with Ki and alpha values of 0.5 mM and 0.15, respectively. Thionine O was a partial (hyperbolic) uncompetitive inhibitor with Ki and alpha values of 0.4 microM and 0.15, respectively. LY83583 and 2-anilino-1,4-naphthoquinone also showed (hyperbolic) partial, uncompetitive inhibition with micromolar Ki values. For Nile Blue A a model for two-site binding with (parabolic) uncompetitive inhibition fitted the data with a Ki value of 11 microM and a kinetic cooperativity between the sites of 0.12, increased to 0.46 by preincubation of the enzyme and Nile Blue A in the presence of glutathione disulphide. Analysis of the effects of preincubation on the kinetics and cooperativity indicated the possibility of a slow conformational change in the homodimeric enzyme, the first such indication of kinetic cooperativity in the native enzyme to our knowledge. Further evidence of conformational changes for this enzyme came from studies of the effects of dimethyl sulphoxide which indicated that this co-solvent, which at low concentrations has no apparent effect on initial velocities under normal assay conditions, induced a slow conformational change in the enzyme. Thionine O, Nile Blue A and LY83583 were redox-cycling substrates producing superoxide ion, detectable by means of cytochrome c reduction, but leading to no loss of glutathione reductase activity, under aerobic or anaerobic conditions. The water-soluble Safranine analogues Methylene Blue, Methylene Green, Nile Blue A and Thionine O (5 mg/kg i.p. x 5) were effective antimalarial agents in vivo against P. berghei, but their effect was small and a higher dose (50 mg/kg i.p. x 1) was toxic in mice. Comparison was made with human glutathione reductase and its literature-reported interactions with several tricyclic inhibitors as studied by X-ray diffraction. It is possible that the conformational changes detected in the present study from alterations in detailed kinetic inhibition mechanisms may shed light on information transfer through the glutathione reductase molecule from the dimer interface ligand pocket to the active-site.