Thioredoxins are small (12-14 kDa) ubiquitous proteins that control the redox state of diverse target proteins by the mediation of thiol-disulfide exchanges (1, 2). The redox-active sulfhydryls of thioredoxin are located in the highly conserved active-site sequence -Trp-Cys-Gly-Pro-Cys-. The pathway for the reduction of a protein disulfide by thioredoxin entails nucleophilic attack by one of the active-site sulfhydryls to form a protein-protein mixed disulfide followed by intramolecular displacement of the reduced target protein with concomitant formation of oxidized thioredoxin (Fig. 1). As the mixed disulfide is disfavored thermodynamically, its direct characterization has not been possible. However, the active-site cysteinyl residue of thioredoxin that is nearest the N terminus (Cys-32 of the Escherichia coli protein) is well established as the primary nucleophile. Originally, this assignment was deduced from the high chemical reactivity and the low pK a (6.7) of Cys-32 (3-6) and was subsequently supported by the three-dimensional structure of E. coli thioredoxin, in which Cys-32 is solventaccessible and Cys-35 is inaccessible (7). More recently, Cys-32 of E. coli thioredoxin, the corresponding residue of human thioredoxin, and the corresponding residue of the closely related protein glutaredoxin were shown to engage in mixed disulfide bond formation with low molecular weight thiols or peptides (8 -11). Finally, the C49S mutant of chloroplastic thioredoxin f (Trx) 1 retains the capacity to activate chloroplastic fructose-1,6-bisphosphatase, whereas the C46S mutant is totally inactive in this regard. Thus, Cys-46 of Trx (analogous to Cys-32 of E. coli thioredoxin) is verified as the primary nucleophile in the reduction of in vivo target proteins, thereby minimizing the possibility that protein-protein interactions might alter the relative accessibility and reactivity of the active-site sulfhydryls of thioredoxin.In contrast to the rigorous proof that Cys-46 of Trx participates in intermolecular disulfide bond formation, the identity of the pairing residue in any target protein has not been established heretofore. We have addressed this issue with PRK by examining the potential of site-directed mutants, which lack either one of the two redox-active sulfhydryls (Cys-16 or Cys-55), to form stable mixed disulfides with the C49S mutant of Trx. This approach, which should be applicable to other target proteins, clearly identifies Cys-55 as the participant in the intermolecular mixed disulfide.
EXPERIMENTAL PROCEDURESMaterials-The following chemicals and biologicals were procured from the indicated vendors: Bicine and DTT, Research Organics, Inc.; components for kinase assays, Sigma; cupric sulfate, Fisher Scientific; DTNB, Pierce. D-Ribulose 5-phosphate was prepared from D-ribose 5-phosphate by the action of phosphoribose isomerase, and its concen-