Dopamine prevents nitration of tyrosine hydroxylase by peroxynitrite and nitrogen dioxide: is nitrotyrosine formation an early step in dopamine neuronal damage?

Peroxynitrite and nitrogen dioxide (NO2) are reactive nitrogen species that have been implicated as causal factors in neurodegenerative conditions. Peroxynitrite-induced nitration of tyrosine residues in tyrosine hydroxylase (TH) may even be one of the earliest biochemical events associated with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-induced damage to dopamine neurons. Exposure of TH to peroxynitrite or NO2 results in nitration of tyrosine residues and modification of cysteines in the enzyme as well as inactivation of catalytic activity. Dopamine (DA), its precursor 3,4-dihydroxyphenylalanine, and metabolite 3,4-dihydroxyphenylacetic acid completely block the nitrating effects of peroxynitrite and NO2 on TH but do not relieve the enzyme from inhibition. o-Quinones formed in the reaction of catechols with either peroxynitrite or NO2 react with cysteine residues in TH and inhibit catalytic function. Using direct, real-time evaluation of tyrosine nitration with a green fluorescent protein-TH fusion protein stably expressed in intact cells (also stably expressing the human DA transporter), DA was also found to prevent NO2-induced nitration while leaving TH activity inhibited. These results show that peroxynitrite and NO2 react with DA to form quinones at the expense of tyrosine nitration. Endogenous DA may therefore play an important role in determining how DA neurons are affected by reactive nitrogen species by shifting the balance of their effects away from tyrosine nitration and toward o-quinone formation.

Peroxynitrite (ONOO Ϫ ) 1 is formed in the chemical reaction between nitric oxide and superoxide (1). ONOO Ϫ is a powerful oxidant that can modify proteins and cell organelles, damage DNA, and cause lipid peroxidation, properties that are thought to underlie its cytotoxicity (2,3). ONOO Ϫ nitrates free tyrosine and tyrosine residues in proteins, modifications that are used as markers of ONOO Ϫ action under conditions of cellular damage and in numerous diseases (4,5). ONOO Ϫ has been implicated as a causal factor in dopamine (DA) neuronal damage that occurs after exposure to the neurotoxic amphetamines (6 -8) and MPTP (9,10). It is also suspected to play a role in the etiology of idiopathic Parkinson's disease (11,12).
The participation of ONOO Ϫ in cellular or neuronal toxicity is an evolving and complicated issue. Real-time measures of intracellular nitration indicate that ONOO Ϫ may not cross cell membranes in sufficient amounts to cause intracellular tyrosine nitration (13). This may be a reflection of preferential nitration of hydrophobic, transmembrane tyrosines by ONOO Ϫ as compared with tyrosines in the aqueous phase (14). Mayer and colleagues (15) have argued on chemical and kinetic grounds that ONOO Ϫ is altogether ineffective as a tyrosine nitrating species in vivo. ONOO Ϫ is by no means the only nitrating species and a strong case can be made for nitrogen dioxide (NO 2 ) as a more relevant nitrating reagent (13,16).
The tyrosine-nitrating properties of ONOO Ϫ and NO 2 are not often considered within a context of cellular phenotype, but this could be extremely important in the case of DA neurons. ONOO Ϫ and NO 2 react with DA to form o-quinones and various nitro-catechols (17)(18)(19)(20)(21). DA-mediated neurotoxicity is associated with increased formation of catechol-quinones, and quinones are known to modify cysteine residues in proteins (22)(23)(24), including TH (25). Catechol-quinones have also been implicated in Parkinson's disease (26 -30).
The interaction of DA with reactive nitrogen species could have important consequences in DA neurons by determining the pathway of toxicity, yet the influence of DA on the proteinnitrating properties of ONOO Ϫ and NO 2 has not been considered. Given that tyrosine nitration of TH may be an early biochemical event in the DA neurodegeneration associated with Parkinson's disease (9), we have studied the effects of DA on nitration of TH. It is found presently that catechols prevent nitration of TH by ONOO Ϫ and NO 2 . Using intact cells expressing the human DA transporter (31,32) along with a green fluorescent protein-TH fusion protein as a reporter of real-time nitration (13), we also observe that intracellular tyrosine nitration is prevented by DA. These findings suggest that nitrotyrosine formation may be suppressed in DA neurons as long as catechol synthesis and storage are intact and point to catecholquinones as early participants in DA neuronal damage.

EXPERIMENTAL PROCEDURES
Materials-Tetrahydrobiopterin was obtained from Dr. B. Schircks Laboratories (Jona, Switzerland). Diethylenetriaminepentaacetic acid (DTPA), glutathione-agarose, DA, DOPA, DOPAC, sodium nitrite, sodium periodate, myeloperoxidase, and horseradish peroxidase were obtained from Sigma. Calbiochem was the source of 2-phenyl-4,4,5,5tetramethylimidazole-1-oxyl 3-oxide (PTIO). Catalase and a monoclonal antibody against TH were products of Roche Applied Science. Thrombin and pGEX fusion protein vectors were obtained from Amersham Biosciences. The mammalian cell expression vector pEGFP-C3 was purchased from Clontech. Zeocin, pCMV/Zeo, and LipofectAMINE 2000 were products of Invitrogen. Nitro blue tetrazolium (NBT) was purchased from Aldrich Chemical Co. A monoclonal antibody against nitrotyrosine and PAPA/NO was purchased from Cayman Chemical Co. (Ann Arbor, MI), and horseradish peroxidase-linked goat-anti-mouse IgGs were products of Cappel. Polyethylene oxide-maleimide-activated biotin (PMAB) and horseradish peroxidase-linked streptavidin were purchased from Pierce. Enhanced chemiluminescence (ECL) reagents were products of PerkinElmer Life Sciences, and Bio-Max MR film was from Kodak. All other reagents were obtained from commercial sources in the highest possible purities.
Cloning and Assay of TH-TH was expressed as a glutathione Stransferase fusion protein. The recombinant protein was purified by glutathione-agarose affinity chromatography and the glutathione Stransferase fusion tag was removed by thrombin cleavage, resulting in a highly purified TH preparation (Ͼ95% pure) as previously described (25,33,34). TH catalytic activity was assessed by the tritium release method as described by Lerner et al. (35). Protein was measured using the method of Bradford (36).
Preparation of ONOO Ϫ and NO 2 and Treatment of TH-ONOO Ϫ was synthesized by the quenched-flow method of Beckman et al. (37), and its concentration was determined by the extinction coefficient ⑀ 302 ϭ 1670 M Ϫ1 cm Ϫ1 . The hydrogen peroxide contamination of ONOO Ϫ solutions was removed by manganese dioxide chromatography and filtration (37). The typical concentration of stock ONOO Ϫ solutions ranged between 300 and 400 mM. ONOO Ϫ was added to TH with vigorous mixing in 50 mM potassium phosphate buffer, pH 7.4, containing 100 M DTPA, and incubations were carried out for 15 min at 30°C. The volume of ONOO Ϫ added to the enzyme samples was always less than 1% (v/v) and did not influence pH. When tested, catechols were added immediately prior to ONOO Ϫ when tested. Upon completion of incubation with ONOO Ϫ and other additions, enzyme samples were diluted with 10 volumes of 50 mM potassium phosphate, pH 6, and assayed for catalytic activity or posttranslational modification (nitration or quinolation) after SDS-PAGE and transfer to nitrocellulose (see below). NO 2 was produced by reacting horseradish peroxidase or myeloperoxidase (specified below) with hydrogen peroxide (100 M) in the presence of sodium nitrite (10 -500 M) as described by Espey et al. (13). TH was exposed to NO 2 -generating conditions with or without catechols for 60 min at 30°C after which the enzyme was diluted with 10 volumes of 50 mM potassium phosphate, pH 6, and assayed for catalytic activity or post-translational modification as described above for ONOO Ϫ .
Post-translation Modification of TH by Reactive Nitrogen Species and Catechol-quinones-Following treatment with ONOO Ϫ or NO 2 with or without catechols, TH was subjected to SDS-PAGE (62). Proteins were transferred to nitrocellulose, blocked in Tris-buffered saline containing Tween 20 (0.1% v/v) and 5% nonfat dry milk, and probed with a monoclonal antibody specific for nitrotyrosine (1:2000 dilution). After incubations with primary antibodies, blots were incubated with a goat anti-mouse secondary antibody conjugated with horseradish peroxidase, and immunoreactive protein bands were visualized with enhanced chemiluminescence. Catechol-quinone modification of TH was assessed in separate experiments by staining blots with NBT in the presence of 2 M potassium glycinate buffer pH 10 as described previously (38).
Modification of Cysteine Residues in TH by ONOO Ϫ and NO 2 -The effect of ONOO Ϫ and NO 2 with or without catechols on TH cysteine residues was determined with the use of the thiol-reactive biotinylation reagent PMAB as described previously (33). PMAB reacts selectively with reduced cysteines in proteins and does not react with cysteines that have been oxidized. This probe is not quantitative but allows a relative measure of the extent to which cysteine residues have been modified. Untreated TH or enzyme exposed to ONOO Ϫ or NO 2 with or without catechols as described above was diluted with 100 mM Tris-HCl, pH 8.5, for subsequent labeling with PMAB (50 M). Proteins were labeled for 60 min at room temperature in the dark after which they were subjected to SDS-PAGE and blotting to nitrocellulose. PMAB reactivity was detected with horseradish-conjugated streptavidin and visualized with ECL.
Direct Real-time Evaluation of Tyrosine Nitration with Enhanced Green Fluorescent Protein-A fusion protein comprised of enhanced green fluorescent protein (eGFP), and TH was constructed by cloning the full-length cDNA of TH into the vector pEGFP-3C at its XhoI/ BamHI restriction sites in the multiple cloning site. In this orientation, the eGFP fusion tag was upstream of the TH amino terminus, and the entire fusion protein complex had a molecular mass of 87 kDa. The pEGFP/TH fusion vector was stably transfected into HEK293/EM4 cells already stably expressing the human DA transporter (hDAT) as described previously (31,32). Because hDAT-expressing cells were resistant to hygromycin and Geneticin (G418), cells were transfected with pEGFP/TH and pCMV/Zeo in a ratio of 10:1 (in LipofectAMINE 2000), and selection was carried out in zeocin (40 g/ml) and by observing eGFP fluorescence with a fluorescence microscope. Both elements of the eGFP/TH fusion protein retained their respective functionality (i.e. fluorescence and TH catalytic activity). Cells were maintained in Dulbecco's modified Eagle's medium (high glucose) containing 10% fetal calf serum in an atmosphere of 5% CO 2 . The resulting stable transformants expressed eGFP/TH (for evaluation of tyrosine nitration in intact cells) and hDAT (for transport of DA into cells). Real-time evaluation of eGFP/TH tyrosine nitration was carried out as described by Espey et al. (13). Intact cells (1 ϫ 10 6 ) were washed into phosphate-buffered saline (pH 7.4) containing DTPA (100 M) and exposed to NO 2 via incubation with PTIO and PAPA/NO (39) at 37°C with or without pre-loading of cells with DA as previously described (32). The extent of intracellular tyrosine nitration caused by NO 2 was monitored through measures of reductions in eGFP/TH fluorescence at 488 em /512 ex nm (13,39) in an Aminco-Bowman Series 2 fluorescence spectrometer. Immediately after measures of fluorescence, intact cells were placed on ice, washed 3ϫ with ice-cold phosphate-buffered saline, and sonicated in 60 l of potassium phosphate buffer (pH 6). TH activity was subsequently measured in the cell supernatant after sedimentation of membranes by centrifugation at 40,000 ϫ g for 15 min at 4°C.

RESULTS
ONOO Ϫ caused a concentration-dependent inactivation of TH activity as previously reported (33,34), and, at a concentration of 100 M, TH activity was reduced to 50% of control. Fig. 1A presents data showing the effects on TH activity of adding DA in the presence of ONOO Ϫ . It can be seen that DA had little effect on the ONOO Ϫ -induced inhibition of TH. Concentrations up to 100 M DA plus ONOO Ϫ did not change the effect on TH catalytic activity caused by ONOO Ϫ alone. The effect of ONOO Ϫ on TH activity was statistically significant (p Ͻ 0.05, ANOVA), but the effect of DA on ONOO Ϫ -induced inhibition of TH was not. Additional catechols were tested for their influence on the ONOO Ϫ -induced inactivation of TH. Fig.  1B shows that equimolar concentrations (20 M) of DOPA, and DOPAC, in the presence of ONOO Ϫ , caused varying degrees of inhibition of TH. The effect of DOPAC was similar to that of DA in that it did not change the extent of inhibition caused by ONOO Ϫ alone. In contrast, DOPA decreased the inhibition caused by ONOO Ϫ . The effects of DOPA were significantly different from control and ONOO Ϫ alone (p Ͻ 0.05, Bonferroni's test).
In view of the fact that catechols prevent the nitration of free tyrosine caused by ONOO Ϫ (20, 40), we tested their effects on ONOO Ϫ -induced nitration of tyrosine residues in TH. Fig. 2A shows that ONOO Ϫ (100 M) caused extensive nitration of tyrosine residues in TH as measured by immunoblotting with a monoclonal antibody against nitrotyrosine. Catechol compounds were tested at a concentration of 20 M, and each completely prevented the ONOO Ϫ -induced nitration of TH. Although a careful titration of the concentration effects of the catechols on ONOO Ϫ -induced nitration was not carried out, we have observed that molar ratios of about 1:5 (DA:ONOO Ϫ ) are sufficient to block the TH-nitrating properties of ONOO Ϫ . The last lane in Fig. 2A shows that the DA aminochrome (formed by reacting DA with sodium periodate), like DA itself, also blocked ONOO Ϫ -induced tyrosine nitration in TH. When a blot similar to the one in Fig. 2A was exposed to redox cycling staining, it was observed that the catechol compounds in the presence of ONOO Ϫ converted TH to a redox cycling protein. Fig. 2B shows that DA and DOPAC produced the strongest redox cycling staining in TH after exposure to ONOO Ϫ . DOPA was somewhat less potent than DA and DOPAC in this regard. These results with ONOO Ϫ -DA interactions agree with previous studies showing that chemical or enzymatic conversion of DA to its aminochrome or o-quinone, respectively, modifies TH, converting the enzyme to a redox cycling quinoprotein (25).
Because the effects on TH of tyrosine nitrating species other than ONOO Ϫ have not been investigated, we used increasing concentrations of sodium nitrite to generate a range of concentrations of NO 2 in the presence of constant levels of horseradish peroxidase (25 units) and hydrogen peroxide (100 M). Fig. 3A shows that TH was quite sensitive to inhibition by NO 2 . TH was inhibited by 40 -50% at a nitrite concentration of 200 M; when the nitrite concentration reached 500 M, TH activity was inactivated by 60 -70%. Omission of any one or two of the components needed to generate NO 2 prevented inhibition of TH, indicating that the enzyme was not inhibited by the peroxidase, nitrite, or hydrogen peroxide. Substitution of myeloperoxidase for horseradish peroxidase produced the same inhibition of TH catalytic activity (data not shown). The effects of NO 2 on TH activity were statistically significant (p Ͻ 0.01, ANOVA). The effects of DA on the NO 2 -induced inhibition of TH are presented in Fig. 3B. Low concentrations of DA (5-20 M) slightly enhanced the inhibition of TH activity caused by NO 2 . For example, TH activity was reduced to about 50% of control by NO 2 in the absence of DA, whereas 20 M DA increased the inhibitory effects of NO 2 on TH to 40% of control. Higher concentrations of DA (50 -100 M) did not further alter the NO 2 -induced reduction in TH activity. The overall effect of DA on the inhibition of TH by NO 2 was statistically significant (p Ͻ 0.05, ANOVA). Equimolar concentrations of DOPA and DOPAC were also tested for effects on NO 2 -induced inhibition of TH, and the results are included in Fig. 3C. The inhibitory effects of NO 2 were not altered by DA or DOPA but were significantly increased by DOPAC. Whereas DA and DOPA increased the inhibition of TH by NO 2 from 50% to about 65%, DOPAC plus NO 2 resulted in a near-total inactivation of TH (i.e. 95% inhibition).
The effect of catechols on nitration of TH by NO 2 was tested, and the results are presented in Fig. 4A. The NO 2 -generating conditions that caused inhibition of TH catalytic activity (Fig. 3 above) resulted in the nitration of tyrosine residues in the TH monomer (60 kDa). Omission of any one of the components needed to produce NO 2 prevented tyrosine nitration in TH (data not shown). It can be seen in Fig. 4A that equimolar concentrations (20 M) of DA, DOPA, and DOPAC prevented NO 2 -induced nitration of tyrosine residues in TH. In agreement with results using ONOO Ϫ as the nitrating species (Fig.  2), exposure of TH to NO 2 in the presence of any of the catechols converted the enzyme to a redox cycling quinoprotein, as shown in Fig. 4B.
In view of results showing that catechol prevention of TH nitration by ONOO Ϫ or NO 2 did not relieve the enzyme of inhibition, we tested the effects of these treatments on the status of cysteine residues in TH. In agreement with previous studies (33), ONOO Ϫ lowered PMAB labeling of TH as shown in Fig. 5 (middle lane of each nitration condition), an indication of cysteine modification. NO 2 caused similar reductions in PMAB labeling as seen with ONOO Ϫ . When ONOO Ϫ or NO 2 was combined with DA, the reduction in PMAB labeling was still observed. DOPA and DOPAC produced the same effects on PMAB labeling as seen with DA in combination with ONOO Ϫ or NO 2 (data not shown). Digital scans of the data in Fig. 5 indicate that the nitrating species reduced PMAB labeling by ϳ95% and the DA-quinones also reduced labeling to roughly the same extent.
Espey et al. (13) recently introduced a method to measure intracellular tyrosine nitration directly and in real-time based on the sensitivity of eGFP to nitration-induced reductions in fluorescence. We created stable transformants expressing an eGFP/TH fusion protein in HEK293/EM4 cells bearing the All treatments were carried out for 60 min at 30°C after which TH was diluted and assayed for remaining activity. Results in all panels are expressed as percent control TH activity and are the mean Ϯ S.E. four to five experiments carried out in duplicate. The overall effect of NaNO 2 in A was statistically significant (p Ͻ 0.05, ANOVA) as was the effect of DA in B. The effects of DA and DOPA on TH activity were significantly different from control (*, p Ͻ 0.05, Bonferroni's test) but did not differ from those of NO 2 alone. The effect of DOPAC was significantly different from control and all other treatment conditions (**, p Ͻ 0.05, Bonferroni's test).

FIG. 4. Post-translational modification of TH after treatment with NO 2 in the presence of DA and other catechols.
A, TH (10 M) was treated with NO 2 -generating conditions as described in the legend to Fig. 3. The indicated catechols (20 M each) were added just prior to NO 2 , and, after a 60-min incubation at 30°C, samples were prepared for SDS-PAGE and immunoblotting with a monoclonal antibody against nitro-tyrosine (diluted 1:2000). Immunoreactivity was visualized by ECL. B, TH treated under the conditions described in A was analyzed by redox cycling staining after SDS-PAGE and electroblotting to nitrocellulose. Gels in A and B contained 10 g of TH protein per lane. hDAT (31). This cell line was chosen for use presently because of its low endogenous content of DA and because the hDAT could be used to transport DA into intact cells. Exposure of these cells to NO 2 via treatment with PTIO-PAPA/NO caused a significant reduction in eGFP fluorescence as shown in Fig. 6. When cells were preloaded with DA before exposure to NO 2 , the nitration-induced reduction in fluorescence was largely prevented (Fig. 6). The effect of DA on the NO 2 -induced reduction in eGFP fluorescence was significant (p Ͻ 0.05, Bonferroni's test). TH activity was almost completely inhibited after exposure of intact cells to NO 2 . It can also be seen in Fig. 6 that DA provided partial protection against NO 2 -induced inhibition of TH. In agreement with Pfeiffer et al. (41)(42)(43) and Espey et al. (13), ONOO Ϫ in concentrations up to 1000 M, added as a bolus or by slow decomposition of SIN-1, did not cause intracellular tyrosine nitration as measured by reductions in eGFP/TH fluorescence, nor did it inhibit TH activity (data not shown). Finally, to test if ONOO Ϫ could be playing a role in NO 2mediated nitration reactions, as a downstream by-product, cells were incubated with the ONOO Ϫ scavenger methionine (20 mM) during PTIO-PAPA treatment. Fig. 6 shows that methionine did not prevent NO 2 -induced nitration, suggesting that ONOO Ϫ was not playing a role under the present treatment conditions. DISCUSSION ONOO Ϫ has been implicated as a causal factor in the DA neuronal damage that occurs after treatment of animals with MPTP (9,10,44,45), and it is suspected of playing a role in Parkinson's disease (11,12,46). The damaging effects of the neurotoxic amphetamines (i.e. methamphetamine) on DA nerve endings have also been ascribed, in part, to ONOO Ϫ (6 -8). Methamphetamine and MPTP each cause reductions in TH activity in vivo (9,47), and ONOO Ϫ inhibits TH in vitro (9,33,34). Based on these findings, it is possible that ONOO Ϫ mediates the detrimental effects of MPTP and the neurotoxic amphetamines on DA neurons. Although the relevance of ONOO Ϫ as an in vivo nitrating species has been questioned on chemical and kinetic grounds (15,(41)(42)(43), it is not the only reactive species that can lead to tyrosine nitration. Espey et al. (13) monitored real-time tyrosine nitration of green fluorescent protein in intact human MCF-7 cells and observed that ONOO Ϫ caused little if any intracellular nitration whereas NO 2 did (13). The effect of NO 2 on the functional aspects of TH or DA neurons has not been investigated.
TH is quite sensitive to inhibition by ONOO Ϫ and NO 2 . Both reactive nitrogen species cause concentration-dependent reductions in TH catalytic function. The enzyme is also modified postranslationally by ONOO Ϫ and NO 2 as evidenced by tyrosine nitration and cysteine oxidation. Although it has been suggested that the cytotoxicity associated with ONOO Ϫ and NO 2 can be mediated by tyrosine nitration, the effects of these reactive nitrogen species are not often considered within the context of cellular phenotype. One case where this is particularly important is the DA neuron. These neurons are characterized, obviously, by their selective and high content of DA. Catechols react with ONOO Ϫ and NO 2 to form o-quinones and other radical species (17) and, in the process, inhibit the nitration of free tyrosine (20,40). As an initial step in assessing the influence of the DA phenotype on protein nitration, it was important to determine if catechols could modify the nitration of TH caused by reactive nitrogen species. We observed that DA, its precursor DOPA, and its metabolite DOPAC prevented ONOO Ϫ -and NO 2 -induced nitration of tyrosine residues in TH. Despite prevention of tyrosine nitration in TH, the catechols did not relieve the enzyme of inhibition. Considering that ONOO Ϫ -induced nitration of TH has been linked specifically to the loss of TH activity (9,48), it is interesting that TH could be inhibited in the presence of ONOO Ϫ or NO 2 , despite the prevention of tyrosine nitration. In view of results showing that catechols prevent tyrosine nitration caused by ONOO Ϫ or NO 2 without relieving the enzyme from inhibition, the possibility that another post-translational modification was mediating TH inhibition was investigated.
Catechol-quinones are known to attack protein cysteinyls (22,30) and form redox cycling sites after binding (38,49). TH (25), tryptophan hydroxylase (50,51), and the dopamine transporter (32,53) are examples of proteins that can be modified by DA-quinones and aminochromes. Quinone modification of each of these important proteins has the added effect of reducing their functional activity. Either ONOO Ϫ or NO 2 , when combined with DA, DOPA, or DOPAC, modified TH to a redox cycling quinoprotein. However, subtle differences were noted between ONOO Ϫ -and NO 2 -generated quinones and their impact on TH. For instance, DA did not alter the ONOO Ϫ -induced inhibition of TH but slightly increased the inhibition caused by NO 2 . DOPA provided some protection against ONOO Ϫ -induced inhibition of TH but was without effect on NO 2 . Finally, DOPAC did not change TH inhibition by ONOO Ϫ but significantly increased the effects of NO 2 on TH activity. These varying effects probably reflect differences in the chemical interactions between ONOO Ϫ or NO 2 and individual catechols. Any such differences in this regard do not mitigate the importance of the common property shared by catechols, the ability to prevent nitration of tyrosine residues in proteins while causing quinolation of cysteines.
The quinone of DOPAC caused the greatest amount of redox cycling in TH, especially when generated by ONOO Ϫ , and the DOPA quinone resulted in the lowest amount. The relationship between enzyme inhibition and redox cycling does not appear FIG. 6. Effects of DA on NO 2 -induced reductions in eGFP/TH fluorescence in intact cells. HEK293/EM4 cells (1 ϫ 10 6 ) stably expressing an eGFP/TH fusion protein as well as the hDAT were treated with PTIO-PAPA/NO to generate NO 2 as described by Espey et al. (13). Cells were treated for 30 min at 37°C after which the fluorescence of eGFP/TH was measured at 488 em /512 ex nm. Where indicated, cells were incubated with DA (20 M for 10 min) as previously described (32) or with the ONOO Ϫ -scavenger methionine (Met; 20 mM), and fluorescence was monitored in washed cells. After fluorescence measures, cells were washed 3ϫ in ice-cold phosphate-buffered saline and lysed for measures of soluble TH catalytic activity. Striped bars represent results from fluorescence measures, and solid bars represent results from TH activity measures. Results are presented as percent control for each measure and are the mean Ϯ S.E. of four to six experiments run in duplicate. The effect of NO 2 on eGFP fluorescence and TH activity was significant in the absence or presence of methionine (*, p Ͻ 0.01 for each measure, Bonferroni's test). The effect of DA on NO 2 -induced reductions in fluorescence and TH activity was also significant (**, p Ͻ 0.05 for each measure, Bonferroni's test).
to be directly correlated. Redox cycling by substituted quinones is a very complex chemical process and is difficult to use as a direct index of cysteine modification in TH. The o-quinones of DA (and other catechols as well) are extremely volatile, and the reactivity of any particular catechol-quinone will be determined by its access to sulfhydryls and by its electrophilicity (54,55). The redox potential of substituted quinones is also very difficult to predict from their structures (49,56), and, as an illustration, it has been shown that the redox cycling potential of DOPA-quinone is lower than that of many other proteinbound quinones (38). Thus, it appears that the total number of cysteines that are modified by catechol-quinones is a better predictor of the extent of TH inhibition than is the extent of redox cycling caused by a bound quinone.
We have argued recently that the ONOO Ϫ -induced inhibition of tryptophan hydroxylase (57) and TH (33,34) is mediated by cysteine modification, not tyrosine nitration. The effects of NO 2 on TH have not been investigated, so we tested it along with ONOO Ϫ for effects on cysteine residues in TH using PMAB labeling. This sulfhydryl-specific reactant labels reduced cysteine residues in proteins and its reactivity is diminished by treatment of proteins, including TH, with cysteine oxidants like ONOO Ϫ (33). In agreement with the effects of ONOO Ϫ , NO 2 also reduced PMAB labeling of TH, indicative of cysteine modification. What is more, treatment of TH with ONOO Ϫ or NO 2 , in the presence of DA (conditions preventing tyrosine nitration), resulted in reduced PMAB labeling as well. These data reinforce results with redox cycling and establish that catechol-quinones derived from ONOO Ϫ or NO 2 attack cysteine residues in TH. This post-translational modification appears to be the mechanism by which TH is inhibited when tyrosine nitration has been prevented by the catechols.
Evidence for nitration of TH by ONOO Ϫ in intact cells has been difficult to obtain. Ara et al. (9) treated PC12 cell lysates, not intact cells, with ONOO Ϫ and showed that TH was nitrated at selected tyrosine residues. We have not been able to establish that TH is nitrated after treatment of intact PC12 cells with ONOO Ϫ . Several factors could account for this failure and led us to consider an alternative approach to the problem. First, it does not appear that ONOO Ϫ penetrates the plasma membrane of intact cells in sufficient concentrations to cause tyrosine nitration in cytoplasmic proteins (13). Second, ONOO Ϫ is formed de novo from the reaction of nitric oxide with superoxide and high concentrations of these reactants must be maintained at or near a 1:1 stoichiometry to avoid secondary reactions that form species incapable of tyrosine nitration. An imbalance in this stoichiometry can lead to a quenching of nitration and oxidation reactions or may even lead to the formation of nitrosating species (15,58). Third, intact PC12 cells contain very high catecholamine concentrations that could also quench ONOO Ϫ -induced tyrosine nitration. Fourth, it is possible that only a small number of the TH molecules in PC12 cells are nitrated, and immunoprecipitation and immunoblotting are too insensitive to detect TH nitration. We used the method of Espey et al. (13) to monitor tyrosine nitration in intact cells by NO 2 through measures of fluorescence reductions in an eGFP/TH fusion protein stably expressed in hDAT-bearing HEK293/EM4 cells. We chose this cell line because of its extremely low endogenous DA content and because the intracellular content of DA could be increased substantially via the hDAT. The use of fluorescence is also a far more sensitive way of measuring nitration than immunoblotting. It was observed that NO 2 caused a significant reduction in eGFP/TH fluorescence and TH catalytic activity in intact cells. It does not appear that ONOO Ϫ plays a role, possibly as a downstream by-product of NO 2 generation, because the ONOO Ϫ scavenger methionine did not prevent reductions in eGFP fluorescence or TH activity. The magnitude of the reduction in TH activity (near-total) was greater than the reduction in eGFP fluorescence (about 50%) and also stands in contrast to in vitro results where TH activity was inhibited by 50% upon exposure to NO 2 . The reasons for this difference are not immediately evident but could result from use of different methods of NO 2 production (i.e. chemical versus enzymatic) or an attack on cellular TH by nitric oxide generated through PAPA/NO decomposition. Nitric oxide would not alter eGFP fluorescence (13) but could inhibit TH activity in vitro as well as in intact cells. The possibility that TH is modified by nitric oxide is currently under investigation. Pre-loading of cells with DA largely prevented the reduction in eGFP/TH fluorescence caused by NO 2 . Although DA provided partial protection against NO 2 -induced inhibition of TH activity, the enzyme remained inhibited. These results indicate that cellular DA can modulate tyrosine nitration. Furthermore, the loss of TH catalytic activity after exposure of intact cells to PTIO-PAPA/NO establishes that both elements of the eGFP/TH fusion protein had been modified by NO 2 (i.e. eGFP fluorescence and TH activity, respectively).
The present results provoke a re-consideration of ONOO Ϫmediated tyrosine nitration as an early event in the neurotoxic process in DA neurons. It appears that DA, its precursor DOPA, and its metabolite DOPAC shift the balance of influence of ONOO Ϫ and NO 2 toward the formation of o-quinones at the expense of tyrosine nitration. Although most intracellular DA is sequestered within synaptic vesicles, where it might be protected from attack by ONOO Ϫ or NO 2 , TH is a cytoplasmic enzyme, and newly synthesized DA appears in the cytoplasm before it is transported into vesicles. The cytoplasmic availability of DA is not determined solely by TH, particularly under conditions thought to cause the production of ONOO Ϫ in vivo. For example, methamphetamine (59) and MPTP (60, 61) cause extensive redistribution of DA from storage vesicles into the cytoplasm and extracellular space. Thus, ONOO Ϫ -or NO 2induced tyrosine nitration in DA neurons could be suppressed by endogenous catechols, reducing the likelihood that this posttranslational modification is an early occurring event in DA neurodegeneration.
Hastings and colleagues have shown that intrastriatal injections of DA (22,23) or systemic injections of methamphetamine (8), both of which result in damage to DA nerve endings, cause substantial increases in the levels of cysteinyl-DA adducts in proteins. Postmortem Parkinson's tissue contains elevated levels of cysteinyl-catechol species (29,52), and cerebrospinal fluid from individuals with Parkinson's disease contain antibodies that recognize quinone-modified proteins (28). We have shown that TH that has been modified by DA-quinones acquires the ability to cause redox cycling of iron (25), reinforcing the cytotoxic potential of catechol-quinones that was established by the influential studies of Graham (52,54,55). Taken together, evidence is mounting that DA, and its catechol precursors and metabolites, in the form of their o-quinones, may play an early and influential role in determining the viability of DA neurons under conditions of oxidative or nitrosative stress.