Heterologously expressed arsenite oxidase: a system to study biogenesis and structure/function relationships of the enzyme family.

Studies of native arsenite oxidases from Ralstonia sp. S22 and Rhizobium sp. NT-26 raised two major questions. The first one concerns the mode of the enzyme's membrane-association. It has been suggested that a hypothetical not conserved protein could account for this variable association. Expression of the wild type arsenite oxidase in Escherichia coli allowed us to study the cellular localization of this enzyme in the absence of such a hypothetical partner. The results with the Ralstonia sp. S22 enzyme suggest that no additional protein is required for membrane association. The second question addresses the influence of the disulfide bridge in the small Rieske subunit, conspicuously absent in the Rhizobium sp. NT-26 enzyme, on the properties of the [2Fe-2S] center. The disulfide bridge is considered to be formed only after translocation of the enzyme to the periplasm. To address this question we thus first expressed the enzyme in the absence of its Twin-arginine translocation signal sequence. The spectral and redox properties of the cytoplasmic enzyme are unchanged compared to the periplasmic one. We finally studied a disulfide bridge mutant, Cys106Ala, devoid of the first Cys involved in the disulfide bridge formation. This mutation, proposed to have a strong effect on redox and catalytic properties of the Rieske protein in Rieske/cytb complexes, had no significant effect on properties of the Rieske protein from arsenite oxidase. Our present results demonstrate that the effects attributed to the disulfide bridge in the Rieske/cytb complexes are likely to be secondary effects due to conformational changes.