Peroxynitrite Increases the Degradation of Aconitase and Other Cellular Proteins by Proteasome*

We report that exposure of aconitase to moderate concentrations of peroxynitrite, 3-morpholinosydnonimine (SIN-1; a superoxide- and nitric oxide-liberating substance), or hydrogen peroxide, inhibits the enzyme and enhances susceptibility to proteolytic digestion by the isolated 20 S proteasome. Exposure to more severe levels of oxidative stress, from these same agents, causes further inhibition of the enzymatic activity of aconitase but actually decreases its proteolytic breakdown by proteasome. It should be noted that the superoxide and nitric oxide liberated by SIN-1 decomposition react to form a steady flux of peroxynitrite.S-Nitroso-N-acetylpenicillamine, a compound that liberates nitric oxide alone, causes only a small loss of aconitase activity (25% or less) and has no effect on the proteolytic susceptibility of the enzyme. Proteasome also seems to be the main protease in cell lysates that can degrade aconitase after it has been oxidatively modified by exposure to peroxynitrite, SIN-1, or hydrogen peroxide. Using cell lysates isolated from K562 cells treated for several days with an antisense oligodeoxynucleotide to the initiation codon region of the C2 subunit of proteasome (a treatment which diminishes proteasome activity by 50–60%), the enhanced degradation of moderately damaged aconitase was essentially abolished. Other model proteins as well as complex mixtures of proteins, such as cell lysates, also exhibit enhanced proteolytic susceptibility after moderate SIN-1 treatment. Therefore we conclude that peroxynitrite reacts readily with proteins and that mild modification by peroxynitrite results in selective recognition and degradation by proteasome.

A close relationship between protein oxidation and susceptibility to proteolysis has been well demonstrated in a variety of biological systems (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16). A series of publications have demonstrated the enhanced degradation of proteins oxidized by exposure to the superoxide anion radical (O 2 . ), to hydrogen peroxide (H 2 O 2 ), or to the hydroxyl radical ( ⅐ OH), in isolated systems as well as in various cell types (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16). These studies were performed in erythrocytes and reticulocytes of various species (3,6), in Escherichia coli (7,9), in rat muscles in vitro (1,4), in primary cultures of rat hepatocytes (13,14), and recently in dividing cell lines (15,16). Most of these studies revealed that the degradation of mildly oxidized proteins is a normal cellular function, whereas extensively oxidized proteins are poor substrates for proteases and may accumulate and, therefore, contribute to diseases or aging processes (1,2,6,17). In these studies using cells, cell extracts and isolated proteins strong evidence has been presented for the key role of a single proteolytic complex, the 20 S proteasome, in the selective recognition, degradation, and removal of oxidized proteins (1,15,16). Evidence has also been documented for the mechanism of recognition based on a partial unfolding of the oxidized protein with consequent exposure of hydrophobic moieties on the protein surface (18 -21). Over the past several years a number of publications have dealt with the reactions of new oxidants with proteins: nitric oxide (NO) 1 and peroxynitrite (ONOO Ϫ ). Nitric oxide and other nitrogen-based oxidants are produced by a variety of mammalian cells, including endothelial cells, neutrophils, and macrophages (22)(23)(24). While nitric oxide acts as a signal-transduction molecule, it exhibits, together with other nitrogen based oxidants, cytotoxic properties (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35). In particular the reaction of nitric oxide with mitochondrial aconitase in various cells, and the reaction with cytosolic aconitase, also called the iron-responsive element-binding protein, or iron-responsive protein-1, have been intensively investigated (30,31,34). Although NO is a relatively weak oxidant, it is able to react very rapidly with superoxide anion radical to form the strongly oxidizing product peroxynitrite (36). Peroxynitrite reacts with tyrosine residues of proteins, with sugars, and with DNA, and can also initiate membrane lipid peroxidation (25,26,(37)(38)(39)(40). In recent papers it has been demonstrated that aconitase, mitochondrial electron transport components, and mitochondrial ATPase are damaged by ONOO Ϫ (41). Whether NO also exhibits some effects on aconitase activity is not completely clear, because Hausladen and Fridovich (30) and Castro et al. (31) found aconitase to be stable toward NO, whereas Drapier et al. (34) found an inactivation by NO. Stadler et al. (33) described an inactivation of mitochondrial aconitase, whereas they found the cytosolic enzyme to be stable to NO treatment.
We undertook the present investigation to test whether inhibition of aconitase by various nitrogen related oxidants is also accompanied by an increase in proteolytic degradation by proteasome. Therefore we exposed aconitase to various oxidants and investigated both the enzymatic activity and the proteolytic susceptibility of the oxidized protein. To test whether oxidation by nitrogen related oxidants, and the resulting enhanced proteolytic susceptibility of oxidized proteins is a more general process, we also used various other purified proteins as substrates for protein oxidation and proteolysis. Finally we exposed the numerous proteins in centrifuged cell lysates to nitrogen-based oxidants and measured the proteolytic susceptibility of the oxidatively modified protein mixture.

EXPERIMENTAL PROCEDURES
Materials-Porcine heart aconitase (essentially mitochondrial), catalase, ovalbumin, myosin, myoglobin, ferritin, and carbonic anhydrase were obtained from Sigma. Collagenase and superoxide dismutase were from Boehringer Mannheim. Trinitrobenzenesulfonic acid was purchased from Fluka. S-Nitroso-N-acetylpenicillamine (SNAP) and 3-morpholinosydnonimine (SIN-1) were obtained from Alexis. Peroxynitrite was synthesized according to Beckman et al. (25). Radiolabeled formaldehyde was a product of Amersham Pharmacia Biotech. All other chemicals were of highest purity available.
Proteasome Isolation-Proteasome was isolated as previously (42), using outdated blood from the Charité Hospital, Humboldt-University, Berlin. After cell lysis the lysate was frozen to allow the 26 S proteasome complex to dissociate. The final step was a re-chromotography on a Mono Q-fast protein liquid chromatography column, which gave a single peak. Fractions with the highest activities were pooled and stored at Ϫ20°C until needed.
Cells, Cell Culture, and Cell Lysates-K562 cells (chronic myelogenous leukemia, human) were obtained from American Tissue and Cell Culture (ATCC CCL 243). The cells were cultured in 90% RPMI 1640 medium, supplemented with 10% fetal bovine serum. Cells were initially seeded at a density of 5 ϫ 10 6 cells/ml. Cells were harvested and washed three times. Afterward the cells were lysed for 1 h in 1 mM dithiothreitol followed by a centrifugation at 14,000 ϫ g of the lysate.
Exposure of Proteins to Oxidative Stress-Isolated proteins or cell lysates (all at a concentration of 0.33 mg/ml) were exposed to various concentrations of different oxidants, in a rotating bath, for 2 h at 37°C. Incubations were carried out at 50 rotations/min. Afterward any remaining reactive oxygen or nitrogen species were removed by addition of catalase in the case of hydrogen peroxide and by addition of glutathione and superoxide dismutase in the case of SNAP, SIN-1, and ONOO Ϫ . In several control experiments both (oxidant) treated and nontreated samples were next dialyzed extensively for 4 h (at 4°C) against 200 volumes of appropriate buffer, with two changes of dialysis medium, in order to test the possible effects of residual hydrogen peroxide, ONOO Ϫ , SNAP, SIN-1, or their reaction products. In additional control experiments, the possible influence of substances other than nitric oxide or peroxynitrite originating from SIN-1 or SNAP was also investigated by "reversed order addition experiments" (31) involving addition of SIN-1 or SNAP after decomposition of purified NO or ONOO Ϫ .
Proteolysis Measurements-Proteolysis studies involved either the degradation of radiolabeled aconitase, followed by liberation of acidsoluble radioactive counts; the degradation of unlabeled proteins, measured by the liberation of acid soluble primary amines; or the degradation of the fluorogenic peptide s-LLVY-MCA, measured by liberation of the free fluorescent MCA moiety.
Aconitase Degradation-Aconitase was radiolabeled by reductive methylation using [ 3 H]formaldehyde as described by Jentoft and Dearborn (44). The degradation of aconitase was assessed following addition of isolated proteasome (30 g of proteasome/mg of protein) to the damaged or undamaged protein. A proteolysis buffer consisting of 10 mM HEPES, 100 mM KCl, 5 mM MgCl 2 , and 1 mM dithiothreitol (pH 7.2) was used. Degradation was followed over a 2-h incubation period conducted in a rotating bath set at 37°C and 50 rpm. Afterward the proteolysis was stopped by the addition of ice-cold trichloroacetic acid as described in Ref. 4. A 3% albumin solution was used as carrier. The percentage of degradation was determined as described previously (4), by liquid scintillation counting, using the formula: (acid soluble counts Ϫ background counts)/(initial acid precipitable counts Ϫ background counts) ϫ 100.
Degradation of Nonlabeled Proteins-Proteasome addition and incubation was carried out as described above. Reactions were stopped by addition of an equal volume of trichloroacetic acid (20%). The samples were centrifuged for 15 min at 14,000 ϫ g, and the supernatants were neutralized by addition of 1 M NaOH. After treatment with N-ethylmaleimide (0.5 mM final concentration), twice the volume of trinitrobenzenesulfonic acid (0.33 mg/ml in 0.1 M borate buffer (pH 9)) was added. The samples were next incubated for 20 min at 40°C in a shaking water bath. Afterward the reaction was stopped by addition of 10 volumes of 1.0 N HCl. The formation of primary amines was measured spectrophotometrically at 334 nm, using leucine as a standard.
Degradation of Fluorogenic Peptide-The fluoropeptide s-LLVY-MCA was used as a model substrate for the proteasome, as described previously (4,15,16). The s-LLVY-MCA was dissolved in 10% dimethyl sulfoxide and used in a final concentration of 200 M. The peptide was incubated for 1 h with cell lysates at 37°C in a shaking water bath using the proteolysis buffer. The reaction was terminated by addition of 0.1 M borate/ethanol (9:1). Fluorescence of the proteolytically released MCA fluor was measured at 365 nm excitation/460 nm emission. A pure MCA standard was used for quantification.

Inactivation of Aconitase by Oxidative Stress-
The enzymatic activity of purified aconitase was decreased by various oxidants (Fig. 1). Whereas hydrogen peroxide inhibited the enzyme almost completely at concentrations of 30 mM, SNAP (a liberator of NO) was not able to decrease the activity by more then 25%. On the other hand peroxynitrite, in a bolus addition, significantly inhibited aconitase at much lower concentrations FIG. 1. Inactivation of aconitase by oxidants. Aconitase was incubated with the oxidants indicated, at the concentrations shown, for 2 h (for full conditions see "Experimental Procedures"). Oxidant reactions were then stopped, and aconitase or oxidized aconitase was incubated in a shaking water bath at 25°C, in 100 mM Tris (pH 7.6) containing 0.5 mM cis-aconitate, for measurements of enzymatic activity. The reaction was monitored at 240 nm (⑀ ϭ 3.6 mM Ϫ1 cm Ϫ1 ) for 5 min (43). All values represent the means of eight measurements from four independent experiments, for which S.E. values were always less than 10%. than required for H 2 O 2 and completely inhibited activity at a concentration of 1 mM.
To further investigate the effects of ONOO Ϫ we used SIN-1 a compound that liberates both NO and O 2 . , which then react with each other to form peroxynitrite. With SIN-1 exposures, therefore, we were able to produce a continuous flux of peroxynitrite. SIN-1 exposure caused approximately 80% inhibition of aconitase activity at 2.0 mM and complete inhibition at 30 mM (Fig. 1). Therefore, the inactivation of enzyme activity strongly depends on the oxidizing properties of the oxidant used.
In control experiments extensive dialysis of H 2 O 2 -treated, ONOO Ϫ -treated, SNAP-treated, or SIN-1-treated aconitase had no effect on enzyme activity. Furthermore addition of either SNAP or SIN-1 to aconitase after decomposition of purified NO or ONOO Ϫ (reversed order experiments) also had no additional measurable effects on aconitase activity (data not shown). These results indicate that the effects of each oxidant, or oxidant-generating system, used in this study are not due to confounding consequences or side effects of the reagents used.
Oxidative Stress Affects the Degradation of Aconitase by Purified Proteasome-The proteolytic susceptibility of tritiated aconitase was strongly affected by exposure of the protein to various oxidants (Fig. 2). Proteolysis of the enzyme gradually increased after treatment with hydrogen peroxide concentrations up to 2 mM. At this concentration an almost 9-fold increase in the rate of proteolysis, in comparison with controls, was observed. At higher concentrations of H 2 O 2 the protein became a poor substrate for the proteasome, probably because it was too seriously damaged, as previously reported for other proteins (1-6, 15, 16, 18, 19). Fig. 2 also demonstrates the effects of various oxides of nitrogen on the proteolytic susceptibility of aconitase. Whereas SNAP (the NO-liberating substance) did not show any significant influence on the proteolysis of aconitase at concentrations up to 30 mM, bolus addition of peroxynitrite significantly increased the proteolytic susceptibility, even at very low concentrations. An ONOO Ϫ concentration of 1 mM, for example, increased the proteolysis rate up to 6-fold in comparison with untreated controls. Higher ONOO Ϫ concentrations, from 2 to 5 mM, decreased proteolysis rates to below control levels.
Exposure to the continuous flux of ONOO Ϫ generated by SIN-1 caused a maximal increase in the proteolytic susceptibility of aconitase at concentrations of 1-2 mM; higher concentrations causing significant decreases in proteolytic susceptibility, in good agreement with direct ONOO Ϫ exposure (Fig. 2).
The results of Fig. 2 may, therefore, be summarized as follows. First, both H 2 O 2 and ONOO Ϫ increased the susceptibility of aconitase to degradation by the proteasome, at the moderate concentrations that caused variable loss of enzymatic activity in Fig. 1; second, at higher concentrations, both H 2 O 2 and ONOO Ϫ caused significant decreases in proteolytic susceptibility; and third, nitric oxide had no significant effect whatsoever on proteolytic activity Comparison of Figs. 1 and 2 demonstrates that there is no simple relationship between oxidative inactivation of aconitase and susceptibility to degradation by the proteasome. In the case of hydrogen peroxide, proteolytic susceptibility was significantly enhanced with only a 10% loss of aconitase activity. In contrast, ONOO Ϫ (SIN-1) exposure caused an almost 60% loss of aconitase activity before it increased the proteolytic susceptibility of the enzyme. Interestingly, both H 2 O 2 and SIN-1 exerted similar effects on proteolytic susceptibility at similar absolute concentrations (Fig. 2), in contrast SIN-1 (ONOO Ϫ ) was much more potent as an inactivator of aconitase (Fig. 1).
In control experiments extensive dialysis of H 2 O 2 -treated, ONOO Ϫ treated, SNAP-treated, or SIN-1-treated aconitase had no effect on aconitase degradation by proteasome. Furthermore addition of either SNAP or SIN-1 to aconitase after decomposition of purified NO or ONOO Ϫ (reversed order experiments) also had no additional measurable effects on aconitase degradation (data not shown). These results indicate that the effects of each oxidant, or oxidant-generating system, used in this study are not due to confounding consequences or side effects of the reagents used.
Proteasome Is Mainly Responsible for Aconitase Degradation in Cell Lysates-To begin exploring whether the proteasome is also mainly responsible for degradation of aconitase exposed to oxidative stress in intact cells, we next replaced the purified proteasome used in the experiments of Fig. 2 with cell lysates from K562 cells, containing various levels of proteasome. In order to manipulate the proteasome content of K562 cells we used an antisense oligodeoxynucleotide directed against the initiation codon region of the proteasome C2 subunit gene, exactly as described previously (15,16), to deplete the cells (and the lysates prepared from them) of proteasome. Fig. 3 reports the degradation of tritiated aconitase by cell lysates. As shown in the upper part of Fig. 3, moderate concentrations of H 2 O 2 increased the degradation of aconitase by control or sense-oligonucleotide-treated K562 cell lysates up to 3-fold, whereas high concentrations failed to increase the proteolysis rates. With lysates prepared from cells treated with the anti-FIG. 2. Degradation of aconitase by isolated proteasome after treatment with various oxidants. Aconitase was labeled by reductive methylation as described by Jentoft and Dearborn (44). The protein (0.33 mg/ml) was then incubated for 2 h alone, or with one of the oxidants, as described under "Experimental Procedures." Next, 30 g of isolated human 20 S proteasome/mg of aconitase was added and incubated for 2 h. Termination of the proteolysis and calculation of percent degradation were also as described under "Experimental Procedures." All values represent the means of eight measurements from four independent experiments, for which S.E. values were always less than 10%. sense oligonucleotide, no H 2 O 2 concentration tested was able to significantly stimulate proteolytic susceptibility, and high H 2 O 2 concentrations were strongly inhibitory.
In further experiments we tested the effect of proteasome depletion of cell lysates on the degradation of NO-or ONOO Ϫtreated aconitase. As in the case of isolated proteasome, none of the tested concentrations of SNAP was able to increase the proteolytic susceptibility of aconitase (Fig. 3). Relatively low concentrations of ONOO Ϫ , on the other hand, did increase the proteolysis rates in cell lysates from control or sense oligonucleotide treated cells, both after a direct bolus addition (ONOO Ϫ ) and after treatment with a continuous flux (SIN-1). In both cases (as with H 2 O 2 ) the increased degradation of oxidized aconitase was blocked in cell lysates from antisense oligonucleotide treated cells. In agreement with the results of exposure to high H 2 O 2 concentrations, we also found that treatment with high ONOO Ϫ concentrations, or exposure to high ONOO Ϫ flux rates (SIN-1), either failed to increase degradation or actually caused a significant decrease in proteolysis (Fig. 3). Taken together the results of Fig. 3 provide strong evidence that proteasome is responsible for the degradation of oxidized aconitase. This finding is completely consistent with the observation that proteasome is responsible for the intracellular degradation of many other oxidized proteins (1-6, 13-16).
In control experiments extensive dialysis of H 2 O 2 -treated, ONOO Ϫ -treated, SNAP-treated, or SIN-1-treated aconitase had no effect on aconitase degradation by cell-free lysates (using the methods used for Fig. 3). Furthermore addition of either SNAP or SIN-1 to aconitase after decomposition of purified NO or ONOO Ϫ (reversed order experiments) also had no additional measurable effects on aconitase degradation by cell lysates (data not shown). These results indicate that the effects of each oxidant, or oxidant-generating system, on proteolysis are not due to confounding consequences or side effects of the reagents used.
To make sure that a real proteasome depletion actually occurred during the antisense oligodeoxynucleotide treatments of Fig. 3, we measured the peptidase activity of the cell lysates toward the fluorogenic peptide s-LLVY-MCA. In the case of antisense-treated cells the proteolytic cleavage of the peptide in the lysates was decreased by some 50%, whereas sense-treated cells or cells treated with scrambled oligonucleotides had no significant effect. 2 This is comparable with the results published earlier by our group (16) on this same cell line and represents the contribution of the proteasome to the overall s-LLVY-MCA peptidase activity of cell lysates, whereas other peptidases are most likely responsible for the remaining 50% of activity. We also performed Western blots for proteasome and again found, as previously reported with the same procedure (16), a significant decline in the cellular proteasome content (confirmatory data not shown).
Peroxynitrite Induces Proteolytic Degradation of Many Proteins-Since the increased proteolytic susceptibility of ONOO Ϫ exposed aconitase to degradation by proteasome has not been previously reported, we thought it important to test whether the proteolytic susceptibility of other proteins is also increased after ONOO Ϫ treatment. Therefore, we exposed a number of different proteins to SIN-1, dialyzed them, and then incubated them with purified proteasome in vitro. Although the stimulation of proteolysis induced by ONOO Ϫ varied from 2-to 12-fold, a significant increase in proteolysis was found for every protein tested after treatment with 2 mM SIN-1 (Fig. 4).
To further test if the increased proteolytic susceptibility of proteins after peroxynitrite treatment is a general phenomenon, we used ultrafiltrated cell lysates of K562 cells in which the proteins were oxidized by exposure to 2 mM SIN-1, dialyzed, and afterward incubated with isolated proteasome. An almost 4-fold increase in proteolysis (liberation of primary amines) was observed after the SIN-1 treatment in comparison with untreated controls (Fig. 5). Unlike the liberation of acid-soluble counts from metabolically labeled proteins, measurement of protein degradation by liberation of primary amines (as performed for Fig. 5) is a rather insensitive procedure to which all cellular proteins can contribute. Therefore, the ability of peroxynitrite to increase proteolytic susceptibility seems to be a general effect, shared by a large number of cell proteins. DISCUSSION A number of studies have reported the inactivation of aconitase by various forms of activated oxygen, such as the superoxide anion radical and hydrogen peroxide (31,(45)(46)(47). Our own data show an inhibition of aconitase by moderately high concentrations of hydrogen peroxide, which is in agreement with the results of Gardner et al. (47) and Verniquet et al. (48). These authors found an inhibition of mitochondrial aconitase at between 100 and 500 M H 2 O 2 . We further find that oxidative modification is followed by an enhanced proteolytic susceptibility, which results in increased degradation of aconitase (as well as other proteins) by proteasome. FIG. 3. Aconitase degradation in cell lysates from K562 cells treated with antisense oligodeoxynucleotide to the proteasome C2 subunit gene. The proteolysis assay was performed as described in the legend to Fig. 2, with the exception that, instead of isolated proteasome, cell lysates of K562 cells were used. Cell lysate preparation procedures are described under "Experimental Procedures." Cells were either untreated with oligodeoxynucleotides (Control) or were exposed for 7 days to daily additions of either sense or antisense oligodeoxynucleotide (1 nmol/ml). The oligonucleotides used were 15-mers: 5Ј-CACCATGTTTCGAAA-3Ј as sense and 5Ј-TTTCGAAACATGGTG-3Ј as antisense oligonucleotide, both were synthesized on an ABI391 DNA synthesizer (Applied Biosystems) using the standard kits for synthesis and purification. The antisense oligonucleotide is directed against the initiation codon region of the proteasome C2 subunit gene (15,16). Percent degradation was calculated after counting the acid-soluble counts as described under "Experimental Procedures." The values reported represent the means and the S.E. of six measurements from three independent experiments.
The role of proteasome in the degradation of oxidized proteins has been demonstrated in a large number of studies from our laboratory (1-7, 15, 16, 18, 19) and others (8 -14). Recently we were able to demonstrate that dividing mammalian cells in culture are able to degrade oxidatively damaged proteins preferentially and that the proteasome is mainly responsible for this activity (15, 16). We, therefore, felt we were able to answer the doubts raised Dean et al. (49,50) in this regard. The present results even further extend our examples of both oxidants and protein substrates that appear to be involved in proteasomedirected protein turnover (recently reviewed in Ref. 51). Studies of various model proteins have demonstrated in isolated systems that mildly oxidized proteins are preferentially degraded by the isolated 20 S proteasome, also called macroxyproteinase earlier by our group (4,5). Because this multicatalytic proteinase complex exists in both an ATP-and ubiquitin-stimulated 26 S form, and an ATP-and ubiquitinindependent 20 S form, it is important to note that oxidized proteins are preferentially degraded only by the ATP-and ubiquitin-independent form. In fact ATP actually exhibits a 10 -20% inhibition of the degradation of oxidized proteins in cell extracts, and direct comparisons reveal that the 26 S proteasome (in the presence of ubiquitin and a fully competent conjugating system) appears unable to recognize the oxidized forms of various protein substrates (4,5,15,18). Our current studies reveal, for the first time, that both ONOO Ϫ -and H 2 O 2oxidized aconitase is also a preferred substrate for proteolytic degradation by the 20 S proteasome.
As reported previously (51), we suggest that polyubiquitin serves (at least in part) as a hydrophobic "tag" to target many soluble proteins to the 26 S proteasome and that ATP serves partly to de-ubiquitinate substrates and partly to activate their unfolding prior to degradation. Since oxidative inactivation generally causes some rearrangement of secondary and tertiary structure (partial protein unfolding) with consequent exposure of normally buried, hydrophobic domains (18 -21, 51), we suggest that oxidation essentially "replaces" both ubiquitin and ATP in activating substrates for degradation by the 20 S proteasome complex. As observed previously with other proteins exposed to other oxidants, aconitase is degraded most efficiently after treatment with moderate oxidative stress, whereas greater oxidative damage actually leads to decreased proteolytic susceptibility.
The effects of nitric oxide and peroxynitrite on aconitase activity have been investigated by others (30 -35). More recently Yan et al. (52) reported the exciting new observation that oxidative inactivation of mitochondrial aconitase can contribute to age-related deterioration of mitochondrial function in houseflies. Furthermore, by experimentally inactivating or inhibiting mitochondrial aconitase, Yan et al. (52) were actually able to decrease housefly lifespans. These results indicate that oxidative inactivation of mitochondrial aconitase is an important biological process and suggest (to us) that efficient removal/degradation of the damaged enzyme may be an important first step in coping with the problem, assuming that de novo synthesis of replacement aconitase can then occur.
It has been reported that aconitase is one of the major intracellular targets of nitric oxide, and the loss of activity observed by some authors has been attributed to the reaction of nitric oxide with the iron-sulfur cluster of the enzyme. However, two recent independent publications by Hausladen and Fridovich (30) and by Castro et al. (31), as well as the data presented here, demonstrate that nitric oxide itself does not exert a major effect on the activity of aconitase. We now report that nitric oxide causes only a mild decrease in enzymatic activity and produces no significant increase in susceptibility to degradation by the purified proteasome.
Bouton et al. (53) recently demonstrated both disulfide bond formation at Cys 437 and iron-sulfur cluster disruption in the peroxynitrite-mediated inactivation of cytosolic aconitase or iron-responsive protein-1. Earlier, the work of Hausladen and Fridovich (30) and that of Castro et al. (31) conclude that peroxynitrite exerts a strong inactivatory effect on aconitase. These results are in agreement with our own investigation of aconitase activity and our observation that aconitase is de- and only the high molecular weight portion (Ͼ30,000) was used for the experiments. The washed cell lysates were then treated with SIN-1, dialyzed again, and incubated with isolated proteasome as described under "Experimental Procedures." The proteolysis assay with trinitrobenzenesulfonic aid was used. Leucine was used as the standard for calculation of primary amines (4). The values are the means and S.E. of three measurements from nine independent experiments. graded more rapidly after SIN-1 or ONOO Ϫ treatment than is the untreated enzyme. Castro et al. (31) reported a 50% inhibition of aconitase activity at 60 M ONOO Ϫ (30), which is in good agreement with our own results (I 50 of 75 M). Castro et al. (31) also found that ONOO Ϫ attacks the 4iron-4sulfur complex of aconitase, leading to cluster disruption, and that this process is completely reversible due to dissociation of iron-sulfur and reincorporation of iron. However, either the changes in the prosthetic group are a relevant signal for a protease to recognize the protein, or further changes happened in the protein structure, perhaps similar to the disulfide formation observed by Bouton et al. (53) with iron-responsive protein-1. Because of the increased proteolytic susceptibility in the case of mild oxidative damage, and the decreased proteolysis rates after severe damage, it is likely that the proteasome recognizes damaged proteins due to the exposure of hydrophobic moieties on the protein surface, as reported previously (18 -21). Loss of aconitase activity of 15% in the case of H 2 O 2 or even 50% with ONOO Ϫ (SIN-1) is not accompanied by increased proteolytic susceptibility, but after further oxidative modification the proteolytic susceptibility rises. From the fact that the proteolytic susceptibility of aconitase starts to increase at different levels of remaining activity one can conclude that H 2 O 2 and ONOO Ϫ attack aconitase differently. It further appears likely that the oxidized moieties recognized by proteasome do not necessarily include the active center of the enzyme. This phenomenon was discussed previously by our group (1-6, 18, 19) and more recently in detail by the group of Stadtman (20,21), in reference to the recognition and degradation of glutamine synthetase by proteasome.
All peroxynitrite-treated proteins studied in the current investigation exhibited an enhanced proteolytic susceptibility toward degradation by proteasome. This was independent of the presence of iron-containing prosthetic groups in a particular protein. That the proteolytic degradation of peroxynitrite damaged proteins by proteasome is a general phenomenon may be inferred from our results of exposure of a complex protein mixture, such as cell lysates, to peroxynitrite.
The production of superoxide anion radical and nitric oxide is a well characterized process, especially in macrophages, neutrophils, and endothelial cells (22)(23)(24). Because of the fast reaction of NO with O 2 . , it is very likely that ONOO Ϫ is formed in vivo. Due to its high reactivity, ONOO Ϫ presumably reacts with numerous cellular proteins. Therefore we postulate that the reaction of peroxynitrite with proteins enhances their proteolytic susceptibility and reflects a general signal for proteolytic degradation by the 20 S proteasome complex, an efficient mechanism of oxidative protein turnover.