Report

Reversing the Inactivation of Peroxiredoxins Caused by Cysteine Sulfinic Acid Formation

See allHide authors and affiliations

Science  25 Apr 2003:
Vol. 300, Issue 5619, pp. 653-656
DOI: 10.1126/science.1080273

Abstract

The active-site cysteine of peroxiredoxins is selectively oxidized to cysteine sulfinic acid during catalysis, which leads to inactivation of peroxidase activity. This oxidation was thought to be irreversible. However, by metabolic labeling of mammalian cells with 35S, we show that the sulfinic form of peroxiredoxin I, produced during the exposure of cells to H2O2, is rapidly reduced to the catalytically active thiol form. The mammalian cells' ability to reduce protein sulfinic acid might serve as a mechanism to repair oxidatively damaged proteins or represent a new type of cyclic modification by which the function of various proteins is regulated.

Some proteinaceous cysteine residues are sensitive to oxidation by H2O2 because their environment promotes ionization of the thiol (Cys–SH) group, even at a neutral pH, to the thiolate anion (Cys–S), which is more readily oxidized to sulfenic acid (Cys–SOH) than is Cys–SH (1, 2). The sulfenic acid group is usually unstable and either reacts with any accessible thiol to form a disulfide or undergoes further oxidation to sulfinic acid (Cys–SO2H); the disulfide group is stable and resistant to further oxidation (3, 4).

Members of the peroxiredoxin (Prx) family of proteins contain such an H2O2-sensitive Cys residue (57). These peroxidases, which exist as homodimers, catalyze the reduction of H2O2 by using reducing equivalents that are provided by thioredoxin (Trx) (8, 9). An H2O2-sensitive Cys, corresponding to Cys51 in mammalian Prx I and Prx II, is conserved in all Prx enzymes. The conserved Cys51–SH is selectively oxidized by H2O2 to Cys–SOH, which then reacts with Cys172–SH of the other subunit to form an intermolecular disulfide. The disulfide is then specifically reduced by Trx (6, 8). However, the sulfur atoms of Cys51 and Cys172 are relatively far apart (∼13 Å) and the formation of an intermolecular disulfide between these residues is a slow process (10, 11). Thus, the Cys51–SOH intermediate is occasionally oxidized to Cys–SO2H before it is able to form a disulfide (8, 1214). Given that sulfinic acid is not susceptible to reduction by thioredoxin, Prx enzymes with a Cys–SO2H group are catalytically inactive.

Hydrogen peroxide, which is generated as a result of respiration as well as in response to a variety of extracellular stimuli, including cytokines and peptide growth factors, serves as an intracellular messenger at low concentrations but induces cell death at higher concentrations (1, 15). By removing H2O2, Prx enzymes modulate various receptor signaling pathways and protect cells from oxidatively induced death (1618). The proportion of Prx enzymes in the sulfinic state (19), which is small in cells cultured under basal conditions, increases markedly in cells exposed to H2O2 or tumor necrosis factor–α (1214). Hydrogen peroxide also oxidizes exposed methionine residues to methionine sulfoxide, and mammalian cells contain two types of methionine sulfoxide reductase that reverse the oxidation reaction using reducing equivalents from Trx (20). Proteins that have been modified oxidatively at other amino acid residues are usually not repaired, however, and are removed by proteolysis (21).

We investigated the fate of Prx I that contained the Cys51–SO2H group. Raw 264.7 mouse macrophage cells were metabolically labeled with tracer amounts of a mixture of [35S]cysteine and [35S]methionine in a medium lacking nonradioactive sulfur-containing amino acids. They were then washed with Hank's balanced salt solution (HBSS), exposed for 10 min to 100 or 500 μM H2O2 in HBSS, and then incubated in culture medium supplemented with unlabeled cysteine and methionine, each at a concentration of 200 μg/ml (22). At various time points, cellular proteins were analyzed by two-dimensional (2D) gel electrophoresis and autoradiography, and the Prx I spots on the 2D gels were identified by immunoblot analysis (23) (Fig. 1, A and B). Oxidized Prx enzymes are detected as the more acidic satellite spots of the spots corresponding to the reduced enzymes on 2D gels (1214).

Fig.1.

Reversible oxidation of Prx I. 35S-labeled Raw 264.7 cells were exposed for 10 min to 100 μMH2O2 (A) or 500 μM H2O2 (B) in HBSS, washed, and incubated for various times in DMEM, supplemented with 10% FBS and with cysteine and methionine, each at a concentration of 200 μg/ml. Cellular proteins were then analyzed by 2D gel electrophoresis followed by either autoradiography or immunoblot analysis with antibodies to Prx I. The regions of the autoradiograms and immunoblots corresponding to molecular sizes (vertical) of 22 to 28 kD, and isoelectric points (horizontal) of 7.6 to 8.2 are shown. The positions of oxidized (Ox) and reduced (Re) Prx I are indicated. The times shown on the left side of the gels represent total time elapsed, beginning with the 10-min period of H2O2 treatment. The percentage of the reduced form was determined from the autoradiograms or immunoblots and is shown plotted against total time elapsed, as indicated. Quantitative data are means ± SEM of values from three independent experiments. (C) Electrospray ionization mass spectra of reduced Prx I (upper panels) and oxidized Prx I (lower panels), both of which were purified from H2O2-treated HeLa S3 cells (30) (fig. S2). The spectra of multiply charged ions with mass/charge ratio (m/z) values and corresponding total charges (left panels) and deconvoluted spectra derived from these component ions with calculated masses (right panels) are shown.

Both the autoradiogram and the immunoblot showed that Prx I exists almost exclusively in the reduced form in cells not treated with H2O2, as has previously been demonstrated (1214). In contrast, the proportion of the reduced enzyme decreased to ∼25% of the total Prx I in cells treated with 100 μM H2O2 for 10 min (Fig. 1A). After the removal of H2O2, the amount of reduced enzyme increased rapidly and that of the oxidized enzyme decreased concomitantly. However, the regeneration of the reduced enzyme reached a plateau at ∼70 to 80% recovery in cells treated with 100 μM H2O2, suggesting that a proportion of the oxidized Prx I molecules is refractory to reduction—probably as a result of damage caused by further oxidation of Cys51–SO2H to cysteic acid or by the oxidation of other amino acid residues. Almost 100% of the Prx I was oxidized in cells treated with 500 μM H2O2 for 10 min (Fig. 1B). The recovery of the reduced enzyme was slower in these cells than in those treated with 100 μM H2O2, possibly reflecting a diminished reducing power of cells exposed to the higher concentration of oxidant. The pattern of redox changes for Prx II was similar to that for Prx I (fig. S1).

The exposure of HeLa cells to H2O2 also elicited an acidic shift of Prx I similar to that observed in Raw 264.7 cells (14). To identify the modification responsible for the acidic shift, reduced and oxidized Prx I were purified from H2O2-treated HeLa cells (fig. S2), and the molecular masses of the two forms were determined by electrospray ionization mass spectrometry (Fig. 1C). The difference of 32 mass units between the reduced and oxidized enzymes suggests the presence of two additional oxygen atoms in the oxidized species. Furthermore, in-gel tryptic digestion of Prx I in the normal and acidic spots and mass spectral analysis of the resulting peptides showed that the acidic shift was due specifically to the formation of the sulfur oxyacid exclusively at Cys51, although Prx I contains three additional Cys residues (fig. S3).

It is unlikely that the increase in reduced Prx I was due to de novo synthesis because newly synthesized protein should not be detectable by autoradiography. The cells were labeled with only trace amounts of [35S]amino acids and were subsequently washed and supplemented with a large excess of unlabeled sulfur-containing amino acids. In addition, analysis of the samples used to generate Fig. 1A by 1D gel electrophoresis and immunoblotting revealed that the total amount of Prx I did not change during exposure of the cells to H2O2 and subsequent incubation for up to a total of 4 hours (24). Thus, the increase in the amount of reduced Prx I after removal of H2O2 reflected conversion of the oxidized enzyme rather than de novo synthesis.

To confirm that de novo protein synthesis did not contribute to our results, we labeled Raw 264.7 cells with [35S]methionine and [35S]cysteine, exposed them to 500 μM H2O2 for 10 min, and then monitored the conversion of oxidized Prx I to the reduced form in the presence of the protein synthesis inhibitor cycloheximide. Incubation of the cells with cycloheximide for 2 or 3 hours resulted in the reappearance of the reduced enzyme in amounts corresponding to ∼25 and 40%, respectively, of the total Prx I (Fig. 2). This rate of recovery was similar to that observed in the absence of the protein synthesis inhibitor (Fig. 1B). It was not possible to monitor the recovery of reduced Prx I for periods of more than 3 hours, because the oxidatively stressed cells appeared unhealthy and began to die in the presence of cycloheximide.

Fig.2.

Effect of cycloheximide on the apparent reversibility of Prx I oxidation in Raw 264.7 cells. 35S-labeled cells were exposed for 10 min to 500 μMH2O2 in HBSS, washed, and incubated for 2 or 3 hours in DMEM supplemented with 10% FBS, with cycloheximide (10 μg/ml), and with cysteine and methionine, each at a concentration of 200 μg/ml. Cellular proteins were then analyzed by 2D gel electrophoresis and autoradiography. The regions of the autoradiograms containing the Prx I spots (oxidized and reduced) are shown. The panels, top to bottom, correspond to cells before treatment with H2O2, cells exposed to H2O2 for 10 min, and cells allowed to recover for 2 or 3 hours, respectively.

Prokaryotes are able to convert various organic sulfonic acids to the corresponding aldehyde and sulfite by the action of alkanesulfonate monooxygenase (25). Although an equivalent enzyme has not been described in eukaryotes, we considered the possibility that “desulfurization” caused the reversal of the acidic shift of Prx. We therefore subjected 35S-labeled Raw 264.7 cells to repeated cycles of oxidation and recovery (Fig. 3). Recovery for 4 hours after the first 10-min exposure to 500 μM H2O2 resulted in the accumulation of ∼65% of Prx I in the normal position on the autoradiogram. The second treatment with 500 μM H2O2 resulted in the complete shift of Prx I to the acidic spot. Although a substantial proportion (10 to 20%) of the cells appeared unhealthy and detached from the culture dish during the second recovery period of 8 hours, ∼80% of Prx I molecules returned to the normal position on the autoradiogram. These results suggest that the shift of Prx I between the normal and acidic positions was due to the reversible oxidation of Cys51, not by desulfurization or another modification that compensates for the acidic charge.

Fig.3.

Repeated interconversion between reduced and oxidized forms of Prx I in Raw 264.7 cells. 35S-labeled cells were exposed to 500 μMH2O2 for 10 min, washed, and allowed to recover for 4 hours. The cells were then reexposed to 500 μMH2O2 for 10 min, washed, and allowed to recover for 8 hours. At various stages of this protocol, cellular proteins were analyzed by 2D gel electrophoresis and autoradiography. The regions of the autoradiograms containing the Prx I spots (oxidized and reduced) are shown. The panels, top to bottom, correspond to cells before the first exposure to H2O2, cells exposed to H2O2 for 10 min, cells allowed to recover for 4 hours, cells reexposed to H2O2 for 10 min, and cells allowed to recover for 8 hours.

We also examined whether the regeneration of reduced Prx I occurred in three additional cell lines: A549 and WI26 human pulmonary epithelial cells and HeLa human cervical carcinoma cells (Fig. 4). The regeneration of Prx I was apparent in all three cell lines, although the rate was slightly reduced in A549 cells and substantially reduced in WI26 and HeLa cells, compared with that observed in Raw 264.7 cells.

Fig.4.

Reduction of oxidized Prx I in various cell types. A549, HeLa, and WI26 cells were labeled with 35S, exposed to 500 μMH2O2 for 10 min, and allowed to recover for 1, 2, or 4 hours. Cellular proteins were analyzed by 2D gel electrophoresis and autoradiography. The regions of the autoradiograms containing the Prx I spots (oxidized and reduced) are shown. For each cell type, the panels top to bottom correspond to cells before exposure to H2O2, cells exposed to H2O2 for 10 min; and cells allowed to recover for 1, 2, or 4 hours, respectively.

Oxidation of proteinaceous Cys residues to the sulfinic state has been considered to be irreversible (4, 11, 26). In human hepatoma cells, methylselenenic acid (CH3SeO2H) is converted to methylselenol (CH3SeH) by glutathione (27). Attempts to reduce the sulfinic form of Prx I with glutathione in vitro have been unsuccessful (24). Furthermore, treatment of Raw 264.7 cells with buthionine sulfoximine, an inhibitor of glutathione biosynthesis, did not retard the regeneration of Prx I (24). The reduction of protein sulfinic acid might therefore require specific enzymes, as is the case for methionine sulfoxide. The proposed roles for the reversible oxidation of methionine include repair of damaged proteins, regulation of protein function, and elimination of oxidants (20). By analogy, the reduction of cysteine sulfinic acid could likewise serve these cellular functions.Recent studies suggest that proteins with Cys-SO2H residues are not uncommon (26) and that the formation of Cys-SO2H leads to the activation of promatrix metallo-proteinases (28) and nitrile hydratase (29).

Supporting Online Material

www.sciencemag.org/cgi/content/full/300/5619/653/DC1

Materials and Methods

SOM Text

Figs. S1 to S3

References

  • * These authors contributed equally to this work.

References and Notes

View Abstract

Navigate This Article