S-glutathiolated hepatocyte proteins and insulin disulfides as substrates for reduction by glutaredoxin, thioredoxin, protein disulfide isomerase, and glutathione.

The disulfide-reducing activities of glutaredoxin, thioredoxin, protein disulfide isomerase, glutathione, and cysteine were directly compared with a mixture of hepatocyte 35S-glutathiolated proteins as the substrate. Dethiolation of individual 35S-labeled protein bands from the mixture of hepatocyte proteins was analyzed by SDS-PAGE. All of the 35S-labeled protein bands could be completely dethiolated by glutaredoxin, thioredoxin, protein disulfide isomerase, glutathione, or cysteine. On a molar basis glutaredoxin was over 10 times more effective than either thioredoxin or protein disulfide isomerase. Dethiolation rates of individual proteins varied in minor ways. For example, glutaredoxin dethiolated the 15-, 30-, and 48-kDa protein bands 3 to 4 times faster than the 27-, 28-, and 77-kDa bands. Glutaredoxins from pig liver or from bovine heart had the same specificity and similar activity. The rate of dethiolation by glutathione alone was low compared to the glutaredoxin-catalyzed process, but all 35S-labeled protein bands could be reduced by glutathione, cysteine, or dithiothreitol. Glutathione was clearly more effective than cysteine when these two thiols were compared on the basis of the concentration of thiolate anion available at neutral pH. Therefore, glutathione is a more specific reductant of S-glutathiolated proteins than is cysteine but it is much less effective than glutaredoxin. Since glutaredoxin activity in cells is 10 times higher than the concentration used in these experiments, ample activity is available to account for substantial rates of dethiolation in vivo. Thioredoxin is quite inefficient as a reductant of S-glutathiolated proteins, but it was reasoned that it might first reduce glutaredoxin, which then could reduce the S-glutathiolated protein. A combination of thioredoxin and glutaredoxin was effective. It is proposed that glutaredoxin is the principal agent responsible for protein dethiolation in vivo. The effectiveness of glutaredoxin, thioredoxin, and protein disulfide isomerase as reductants for protein disulfide bonds was examined with insulin as the substrate. Protein disulfide isomerase was very effective and thioredoxin was nearly as effective. Human thioredoxin was similar to Escherichia coli thioredoxin in reactivity and specificity. Glutaredoxin did not facilitate insulin reduction at equal concentrations. Thus, protein disulfide isomerase and thioredoxin are more effective than glutaredoxin as reductants of insulin protein disulfides. The apparent reduction potential of pig liver glutaredoxin (-0.159 +/- 0.004 V) was determined by measuring the amount of reduced glutaredoxin in equilibrium with mixtures of glutathione and glutathione disulfide. Glutaredoxin was a weaker reductant than E. coli thioredoxin (-0.260 V) and was similar to protein disulfide isomerase (-0.11 to -0.19 V). The role of these proteins as disulfide reductants is not determined solely by thermodynamic considerations. A glutathione binding site at the dithiol region of glutaredoxin may be of primary importance for its function in protein dethiolation, while a different specific peptide binding site in thioredoxin may be more suited to certain protein disulfide structures.