Endogenous Nitration of Iron Regulatory Protein-1 (IRP-1) in Nitric Oxide-producing Murine Macrophages

Iron regulatory protein-1 (IRP-1) is a bifunctional [4Fe-4S] protein that functions as a cytosolic aconitase or as a trans-regulatory factor controlling iron homeostasis at a post-transcriptional level. Because IRP-1 is a sensitive target protein for nitric oxide (NO), we investigated whether this protein is nitrated in inflammatory macrophages and whether this post-transcriptional modification changes its activities. RAW 264.7 macrophages were first stimulated with interferon-γ and lipopolysaccharide (IFN-γ/LPS) and then triggered by phorbol 12-myristate 13-acetate (PMA) in order to promote co-generation of NO. and \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{{\bar{{\cdot}}}}\) \end{document}. IRP-1 was isolated by immunoprecipitation and analyzed for protein-bound nitrotyrosine by Western blotting. We show that nitration of endogenous IRP-1 in NO-producing macrophages boosted to produce \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{{\bar{{\cdot}}}}\) \end{document} was accompanied by aconitase inhibition and impairment of its capacity to bind the iron-responsive element (IRE) of ferritin mRNA. Lost IRE-binding activity was not recovered by exposure of IRP-1 to 2% 2-mercaptoethanol and was not due to protein degradation. Inclusion of cis-aconitate with cell extract to stabilize the [4Fe-4S] cluster of holo-IRP-1 rendered protein insensitive to nitration by peroxynitrite, suggesting that loss of [Fe-S] cluster and subsequent change of conformation are prerequisites for tyrosine nitration. IRP-1 nitration was strongly reduced when IFN-γ/LPS/PMA-stimulated cells were incubated with myeloperoxidase inhibitors, which points to the contribution of the nitrite/H2O2/peroxidase pathway to IRP-1 nitration in vivo. Interestingly, under these conditions, IRP-1 recovered full IRE binding as assessed by treatment with 2% 2-mercaptoethanol. Peroxidase-mediated nitration of critical tyrosine residues, by holding IRP-1 in an inactive state, may constitute, in activated macrophages, a self-protecting mechanism against iron-induced toxicity.

Iron regulatory protein-1 (IRP-1) is a bifunctional [4Fe-4S] protein that functions as a cytosolic aconitase or as a trans-regulatory factor controlling iron homeostasis at a post-transcriptional level. Because IRP-1 is a sensitive target protein for nitric oxide (NO), we investigated whether this protein is nitrated in inflammatory macrophages and whether this post-transcriptional modification changes its activities. RAW 264.7 macrophages were first stimulated with interferon-␥ and lipopolysaccharide (IFN-␥/LPS) and then triggered by phorbol 12-myristate 13-acetate (PMA) in order to promote co-generation of NO ⅐ and O 2 . . IRP-1 was isolated by immunoprecipitation and analyzed for protein-bound nitrotyrosine by Western blotting. We show that nitration of endogenous IRP-1 in NO-producing macrophages boosted to produce O 2 . was accompanied by aconitase inhibition and impairment of its capacity to bind the iron-responsive element (IRE) of ferritin mRNA. Lost IRE-binding activity was not recovered by exposure of IRP-1 to 2% 2-mercaptoethanol and was not due to protein degradation. Inclusion of cis-aconitate with cell extract to stabilize the [4Fe-4S] cluster of holo-IRP-1 rendered protein insensitive to nitration by peroxynitrite, suggesting that loss of [Fe-S] cluster and subsequent change of conformation are prerequisites for tyrosine nitration. IRP-1 nitration was strongly reduced when IFN-␥/LPS/PMA-stimulated cells were incubated with myeloperoxidase inhibitors, which points to the contribution of the nitrite/H 2 O 2 /peroxidase pathway to IRP-1 nitration in vivo. Interestingly, under these conditions, IRP-1 recovered full IRE binding as assessed by treatment with 2% 2-mercaptoethanol. Peroxidase-mediated nitration of critical tyrosine residues, by holding IRP-1 in an inactive state, may constitute, in activated macrophages, a self-protecting mechanism against iron-induced toxicity.
In mammalian cells, iron regulatory proteins (IRP-1 and -2) 1 marshal iron trafficking, storage, and availability by modulat-ing ferritin and transferrin receptor expression at a post-transcriptional level (1). They operate by interacting with one or several specific stem-loop RNA structures called iron-responsive elements (IREs), which are located in untranslated regions (UTR) of several mRNAs. At low intracellular iron concentration, IRPs bind to the IRE of ferritin mRNA at its 5Ј-UTR and block translation, whereas they stabilize transferrin receptor mRNA through direct interactions with several IRE motifs in the 3Ј-UTR. One critical discrepancy between the two IRPs is that only IRP-1 functions as a cytosolic aconitase when holding a [4Fe-4S] cluster, which masks the IRE-binding domain. In this case, holo-IRP-1 catalyzes the interconversion of citrate into isocitrate via the intermediate formation of cis-aconitate, as does its mitochondrial counterpart in the Krebs cycle (2).
It is well documented that nitric oxide (NO ⅐ ) also stimulates the trans-regulatory activity of IRP-1 and consequently disturbs iron homeostasis in various cellular systems exposed to inflammatory conditions (3). More specifically, NO ⅐ converts IRP-1 from an aconitase to a trans-regulatory factor by directly targeting its [Fe-S] cluster leading to its complete disassembly (4 -6). In this way, the apoprotein generated by NO ⅐ , in a reducing environment, tightly binds to IRE(s) and exerts its iron-regulatory function (7,8). In the early 1990s, it was reported that simultaneous production of NO ⅐ and superoxide (O 2 . ) by activated macrophages led to intracellular formation of peroxynitrite (9). Since NO ⅐ can turn into a powerful oxidizing and nitrating agent such as peroxynitrite in vivo, its reactivity toward key enzymes of cellular metabolism has become an active area of investigation (10,11). Regarding IRP-1, in vitro studies have demonstrated that peroxynitrite causes the direct inhibition of its aconitase activity by promoting complete disruption of the [4Fe-4S] cluster but, unlike NO ⅐ , without stimulating the IRE-binding activity of IRP-1 (5)(6)12). These data suggest that peroxynitrite might generate additional posttranslational modifications of IRP-1 in comparison with NO ⅐ . Interestingly, recent biochemical and resonance Raman studies have allowed the detection of tyrosine nitration on pure recombinant IRP-1 by synthetic peroxynitrite or by SIN-1, an NO ⅐ /O 2 . donor (6,13). However, in vivo studies of IRP-1 nitration have never been undertaken. Currently, the physiological mechanism of protein nitration in vivo is much debated, and whether peroxynitrite is really the reactive precursor of nitrating species in biological systems has been questioned. Indeed, alternative mechanisms, such as the one mediated by the nitrite/H 2 O 2 /myeloperoxidase pathway, have now been identified, particularly in activated neutrophils and certain macrophages (14,15). In this study, we have therefore explored whether nitration of IRP-1 could occur in a whole-cell (physiological) context. To address this issue, we used the murine RAW 264.7 macro-phages and favored the generation of endogenous nitrating species by stimulating them with inflammatory and/or pharmacological stimuli, thus seeking IRP-1 tyrosine nitration. We then investigated the biological significance of this endogenous nitration by measuring both IRP-1 functions in parallel, in the presence or absence of cis-aconitate, a substrate that protects the holo-form of IRP-1. Finally, we also attempted to identify the nitrating pathway that mediates IRP-1 nitration in activated macrophages.
Cell Culture and Treatment-The macrophage RAW 264.7 cell line was cultured in high glucose Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum (Eurobio, Les Ulis, France) at 37°C in a 5% CO 2 atmosphere. Cells (5 ϫ Nitrite Measurement-The nitrite content in cell culture supernates was determined using the Griess reagent containing 0.5% sulfanilamide and 0.05% N-(1-naphthyl)ethylenediamine hydrochloride in 45% acetic acid. Nitrite reacts with Griess color reagent to give a red-violet diazo-dye, which is measured spectrophotometrically at 543 nm; nitrite concentration is calculated from a sodium nitrite standard curve.
Cytoplasmic Extracts-After each treatment, cells were harvested in 250 l of 0.25 M sucrose with 100 mM HEPES, pH 7.4, and then treated with 0.007% digitonin on ice. After 10 min, cell lysate was ultracentrifuged at 100,000 ϫ g for 1 h at 4°C. The cytosolic fraction (supernatant) was tested for protein concentration, and aconitase activity was measured before storage at Ϫ80°C.
In Vitro Treatment with a Bolus of Peroxynitrite-Cytosols (100 g) of control cells were treated with increasing concentrations of peroxynitrite added as a bolus in 50 mM phosphate buffer, pH 7.5, containing 100 mM sodium bicarbonate to favor formation of the nitrosoperoxocarboxylate intermediate, which causes more nitration and less sulfhydryl oxidation (17). In some experiments, cytosols were pretreated with increasing concentrations of cis-aconitate prior to incubation with a bolus of peroxynitrite. Concentration of peroxynitrite stock solution was determined at 302 nm (⑀ ϭ 1670 M Ϫ1 ϫ cm Ϫ1 ) in 0.1 M NaOH solution before use. The pH of each reaction mixture was carefully checked after peroxynitrite addition and was maintained at 7.5 in all experiments.
Aconitase Activity-Aconitase activity was measured spectrophotometrically by following the disappearance of the cis-aconitate at 240 nm (18). Fifty micrograms of cytosolic extracts was preheated at 37°C in 0.1 M Tris-HCl buffer at pH 7.4 before starting the enzymatic reaction with 0.3 mM cis-aconitate. Units correspond to nmol of substrate consumed/min, and ⑀ 240 nm ϭ 3.6 mM Ϫ1 ϫ cm Ϫ1 .
Electrophoretic Mobility Shift Assay-IRP-1/IRE interactions were visualized as described previously (19). Briefly, 2 g of protein was incubated with an excess of 32 P-labeled H-chain ferritin IRE in 20 l of 10 mM HEPES, pH 7.6, 40 mM KCl, 3 mM MgCl 2 , and 5% glycerol (buffer A). In parallel experiments, samples were treated with 2% 2-mercaptoethanol (2-ME) before the labeled RNA probe was added. This treatment allows full expression of IRP-1 IRE-binding activity. After 15 min at room temperature, 1 unit of RNase T1 was added. After 10 min, 5 mg/ml heparin was added for an equivalent additional time. IRP-1⅐IRE complexes were resolved in 6% nondenaturing polyacrylamide gels, and radioactive bands were quantified with ImageQuant software (Amersham Biosciences).
Immunoprecipitation and Western Blot Analysis-Cytosolic extracts from control and treated cells were immunoprecipitated using colloidal superparamagnetic microbeads of protein A (Miltenyi Biotec, Bergish Gladbach, Germany). One l of rat liver IRP-1 antiserum (kindly provided by Dr R. S. Eisenstein) was added to 200 l of cytosolic extract (0.5 mg/ml) in 700 l of 100 mM HEPES, pH 7.4, and mixed. Then, 100 l of protein A-microbeads was added to the suspension to magnetically label the immune complex. After 2 h on ice and in the dark, the magnetically labeled immune complex was bound to a column placed in a magnetic field, and after several washings the immunoprecipitated IRP-1 was eluted from the column with 50 l 1ϫ SDS gel loading buffer. Fifteen l of the eluate was resolved on 8% SDS-polyacrylamide gel and transferred to nitrocellulose membranes. The membranes were then blocked in 3% milk, phosphate-buffered saline for 20 min at room temperature with constant agitation and then incubated with polyclonal anti-nitrotyrosine antibody (diluted 1:5000; Cayman Chemical) in blocking buffer overnight at 4°C. After several washings, the blots were incubated with a peroxidase-conjugated goat anti-rabbit secondary antibody (Dako, Trappes, France) for 1 h at room temperature. The immunoreactive bands were detected by using the enhanced chemiluminescence Western blotting detection system (Amersham Biosciences).
Enzyme-linked Immunosorbent Assay for 3-Nitrotyrosine-Recombinant purified human IRP-1 was incubated with 5 mM cis-aconitate for 10 min at room temperature before adding 500 M SIN-1 for an additional 30 min at 37°C in phosphate-buffered saline, pH 7.6. No pH variation was detected at any time during incubation. 3-Nitrotyrosinemodified IRP-1 was quantified in 96-well plates using 0.5 g IRP-1. Hybridization was carried out with the mouse monoclonal nitrotyrosine antibody (Cayman Chemical) for 2 h at 37°C followed by several washings with phosphate-buffered saline, 0.05% Tween 20. Sequential incubations were then performed with horseradish peroxidase-coupled antimouse IgG (1:2000) for 1 h at 37°C. Absorbance was measured at 490 nm after color development was initiated by substrate addition.
Electron Paramagnetic Resonance (EPR) Spectroscopy and Spin Trapping-RAW 264.7 macrophages were stimulated or not with 20 units/ml IFN-␥ and 50 ng/ml LPS for 18, 20, and 24 h. Spin-trapping measurements of NO ⅐ were performed with 2.5 ϫ 10 6 cells/ml in Dulbecco's modified Eagle's medium containing the Fe 2ϩ complex of Nmethyl-D-glucamine dithiocarbamate (1:5). EPR spectra were recorded at room temperature and at 77 K using a Varian E 109 spectrometer (microwave frequency 9.32 GHz, modulation amplitude 5 gauss, and time constant 0.5 s). The magnetic field was calibrated with the stable radical 1-1-diphenyl-2-picrylhydrazyl used as a marker with g value ϭ 2.0036.

RESULTS
Tyrosine Nitration of Cytosolic IRP-1 by Peroxynitrite-Our previous work concerning the mechanism of action of peroxynitrite on IRP-1 has shown that the molsidomine metabolite SIN-1, which spontaneously decomposes to yield both NO ⅐ and O 2 . concomitantly, thus forming a steady flux of peroxynitrite, was able to nitrate pure recombinant human IRP-1 in vitro (6).
To address whether cellular endogenous IRP-1 might also be nitrated in response to nitrogen-derived species, we first exposed the cytosolic fraction of resting RAW 264.7 macrophages to increasing concentrations of synthetic peroxynitrite. IRP-1 was then immunoprecipitated from cytosols, and tyrosine nitration on IRP-1 was detected by Western blot analysis using a polyclonal nitrotyrosine antibody. As shown in Fig. 1, IRP-1 from untreated cytosol exhibited high aconitase activity (50 units/mg defined as 100% activity in this assay) and nonsignificant nitration. In contrast, the addition of peroxynitrite to resting cell cytosol led to IRP-1 nitration, which was rapidly detectable in the low micromolar range (Fig. 1, B and C). It is noteworthy that the appearance of nitration after a short exposure of 10 M peroxynitrite was accompanied by a significant decrease in cytosolic aconitase activity (76% of the control value) (Fig. 1A). Increasing concentrations of peroxynitrite, from 25 to 500 M, progressively rendered IRP-1 nitration more intense and cytosolic aconitase less active (Fig. 1, A and B). In parallel, we also observed that IRP-1 IRE-binding activity in cytosols treated with peroxynitrite was not stimulated (data not shown).

Endogenous Nitration of IRP-1 in IFN-␥/LPS-and IFN-␥/ LPS/PMA-activated Macrophages-
To address the question of whether endogenous IRP-1 might be nitrated in a whole-cell physiological context, murine RAW 264.7 macrophages were incubated with IFN-␥ and LPS for 18 h to stimulate NO ⅐ production through NO synthase-2 induction ( Fig. 2A). Then, activated cells were further incubated with or without 50 nM PMA for an additional 3 or 6 h to stimulate O 2 . production (20).
We checked that during this time frame nitrite accumulated linearly in the cell culture medium (65, 80, and 95 M at times 0, 3, and 6 h after PMA addition, respectively) and that stimulated cells produced NO ⅐ using an EPR spin-trapping assay (data not shown). We then harvested the cells and extracted their cytosols. IRP-1 was isolated from cytosols by immunoprecipitation, and its nitration was detected by Western blot analysis using a polyclonal nitrotyrosine antibody. As shown in Fig.  2B, constitutively expressed IRP-1 was not nitrated in intact resting macrophages, whereas its cytosolic aconitase activity was high and nitrite production was insignificant. However, after 24 h of IFN-␥/LPS stimulation, the protein exhibited tyrosine nitration. Interestingly, the intensity of IRP-1 nitra-tion increased further in a time-dependent manner when IFN-␥/LPS-treated macrophages were co-stimulated with PMA. Our data also showed that IRP-1 nitration in activated macrophages was accompanied by complete inhibition of cytosolic aconitase activity of IRP-1 (Fig. 2B). In another set of experiments, we investigated the direct involvement of the NO-signaling pathway in endogenous nitration of IRP-1. RAW 264.7 macrophages were stimulated for 24 h with IFN-␥/LPS or IFN-␥/LPS/PMA, as described in Fig. 2, in the presence of L-NMA and EIT, two inhibitors of inducible NO synthase. We showed in Fig. 3 (A and B) that the L-NMA/EIT combination prevented both nitrite production and inhibition of aconitase activity in stimulated cells. Importantly, these two events were correlated with a significant decline in IRP-1 tyrosine nitration (Fig. 3C,  compare lanes 5 and 6 with lanes 2 and 3). We also noticed that treatment of PMA alone triggers neither nitration of IRP-1 nor inhibition of cytosolic aconitase activity. Inhibition of IRP-1 Activities after Optimal Tyrosine Nitration-To investigate further the relationship between IRP-1 nitration and its functions, in parallel experiments we measured IRP-1 IRE-binding activity in control and activated macrophages. As expected from earlier studies, binding of IRP-1 to the IRE of ferritin mRNA was stimulated in IFN-␥/ LPS-treated macrophages (Fig. 4A). However, when PMA was added along with IFN-␥/LPS, IRP-1 was far less potent in binding IRE (Fig. 4A, upper panel, compare lane 2 with lane 3).
Remarkably, this impairment of IRP-1 in binding the IRE motif correlated with the strongest IRP-1 nitration (compare Fig. 3C with Fig. 4A). We also performed the binding assay in the presence of 2% 2-ME prior the IRE probe addition (Fig. 4A,  lower panel). It is well known that this treatment allows full expression of IRP-1 IRE-binding activity (21). Under these conditions, the IRP-1 of control cells was totally converted into its IRE-binding form (Fig. 4A, lower panel, lane 1). As previously shown, full IRP-1 IRE-binding capacity of IFN-␥/LPS- Cells were then harvested, and cytosolic fractions were isolated as described under "Experimental Procedures," and aconitase activity was measured. IRP-1 was immunoprecipitated from cytosols, and its nitration was revealed by Western blot analysis with a polyclonal anti-nitrotyrosine antibody. The experiments were performed at least three times, and a representative result is shown. treated cells measured in the presence of 2% 2-ME was half that of control (Fig. 4A, lower panel, compare lanes 1 and 2, and  Fig. 4B). This was previously explained by a 50% decrease in IRP-1 protein level in response to NO ⅐ (22) and illustrated here in Fig. 4C (compare lanes 1 and 2). Remarkably, when IFN-␥/ LPS-treated cells were co-incubated with PMA, the full IREbinding activity of IRP-1 (in the presence of 2% 2-ME) was further reduced (Fig. 4A, lower panel, compare lanes 1, 2, and  3, and Fig. 4B) to a residual value of 30% of control. This 2-ME-resistant reduction of IRP-1 IRE-binding activity was not due to IRP-1 protein loss, because IRP-1 level was not further reduced during co-stimulation of IFN-␥/LPS-treated cells with PMA (Fig. 4C, compare lanes 2 and 3). These results show that PMA-induced loss of IRE-binding capacity of IFN-␥/LPS-stimulated cells results from a post-translational modification of IRP-1.
We then asked whether IRP-1 nitration precedes or follows [Fe-S] cluster disruption. To solve this issue, we preincubated the cytosols of untreated cells with increasing concentrations of cis-aconitate before adding 25 M peroxynitrite, a concentration that significantly enhanced nitration of IRP-1 in vitro and considerably reduced aconitase activity (Fig. 1). In this set of experiments, we used cis-aconitate because this aconitase substrate is known to protect IRP-1 through stabilization of the [4Fe-4S] cluster of IRP-1 (12, 23). After cis-aconitate/peroxynitrite treatment, aconitase activity was determined in cytosols, and nitration of IRP-1 was sought after immunoprecipitation and Western blot analysis. As shown in Fig. 5, A and B, untreated cytosols exhibited high aconitase activity of IRP-1 and no significant IRP-1 nitration. As expected, peroxynitrite alone led to a 93% inhibition of aconitase activity, which was accompanied by pronounced tyrosine nitration on IRP-1. Notably, IRP-1 nitration by peroxynitrite was progressively reduced to the basal level in cytosols preincubated with increasing concentrations of cis-aconitate. This reduction was accompanied by progressive protection of IRP-1 aconitase activity (Fig.  5B). Parallel control studies demonstrated that 1 mM cisaconitate did not inhibit nitration of bovine serum albumin induced by 25 M peroxynitrite. We also observed the same protective effect of cis-aconitate on purified recombinant human IRP-1 as regards both nitration and aconitase inhibition by SIN-1, which spontaneously released a low, steady flux of peroxynitrite at physiological pH (data not shown). These data thus show that the [4Fe-4S] aconitase form of IRP-1, when protected by cis-aconitate, is not sensitive to in vitro nitration by peroxynitrite.
Involvement of a Nitrite/H 2 O 2 /Myeloperoxidase Pathway in IRP-1 Tyrosine Nitration-In biological systems, endogenous production of peroxynitrite, which results from the simultaneous production of O 2 . and NO ⅐ , requires very specific conditions to nitrate proteins efficiently (24). Alternative mechanisms have recently been identified in endogenous tyrosine nitration (14). In particular, it has been reported that nitrite, the primary metabolic end-product of NO ⅐ , can be oxidized by my-

FIG. 3. Effects of NO synthase inhibitors on IRP-1 tyrosine nitration by IFN-␥/LPS-and IFN-␥/LPS/PMA-activated macrophages.
Cells were stimulated for 18 h with 20 units/ml IFN-␥ plus 50 ng/ml LPS. Then, cells were incubated for an additional 6 h with or without 50 nM PMA. In parallel, stimulated cells were also exposed to NO synthase inhibitors (1 mM L-NMA plus 100 M EIT) throughout the incubation. A, nitrite production was determined in the cell culture medium. B, cytosols were then extracted and used to measure aconitase activity. C, in parallel, cytosols were immunoprecipitated with anti-IRP-1 antibody, and the resulting samples were analyzed by Western blot analysis using a specific anti-nitrotyrosine antibody. The experiments were performed at least three times, and a representative result is shown.

FIG. 4. Regulation of IRE-binding activity and expression level of IRP-1 by PMA in IFN-␥/LPS-treated cells.
RAW 264.7 macrophages were stimulated for 18 h with 20 units/ml IFN-␥ plus 50 ng/ml LPS. Cells were then incubated for an additional 6 h with or without 50 nM PMA. Cytosolic extracts were then prepared. A, IRP-1 IRE-binding activity was measured by electrophoretic mobility shift assay in the presence or absence of 2% 2-ME. B, IRE binding by IRP-1 in the presence of 2% 2-ME quantified by phosphorimaging and expressed as percent of control value. C, cytosols (15 g) from control-, IFN-␥/LPSor IFN-␥/LPS/PMA-treated cells were tested for IRP-1 protein level by Western blot analysis using a chicken polyclonal anti-IRP-1 antibody. The experiments were performed four times, and a representative result is shown. eloperoxidase in the presence of H 2 O 2 to yield NO 2 ⅐ , the main contributor to tyrosine nitration (25). Moreover, involvement of this pathway has been outlined in RAW 264.7 cells (15). We therefore investigated the molecular mechanism of intracellular IRP-1 nitration by incubating IFN-␥/LPS-or IFN-␥/LPS/ PMA-treated RAW 264.7 macrophages with the potent and selective myeloperoxidase inhibitors ABAH, pHBAH, and SHA (14,26,27) or the peroxynitrite scavengers epicatechin and selenomethionine (28,29). After 24 h, nitrite production was measured in cell culture medium, and aconitase activity, as well as nitration of IRP-1, was determined in cell cytosols. We first report, in Figs. 6A and 7A, that peroxynitrite scavengers and myeloperoxidase inhibitors did not prevent inducible NO synthase expression and activity in macrophages stimulated by IFN-␥/LPS in the presence or not of PMA, as indicated by unaffected nitrite production. As expected, IRP-1 aconitase activity was fully inhibited in all cases except for control and PMA-treated cells, which did not produce NO ⅐ . In parallel, we showed that incubation of activated cells with either epicatechin, selenomethionine, or the three myeloperoxidase inhibitors completely abrogated IRP-1 tyrosine nitration in IFN-␥/ LPS-and IFN-␥/LPS/PMA-activated macrophages (Figs. 6B and 7B). We also considered the possibility that nitrite, as a myeloperoxidase substrate, could induce nitration of IRP-1 in RAW 264.7 cells cultured in absence of stimulation. These cells did not display alteration of the IRP1 [Fe-S] cluster as testified by high aconitase activity, and no IRP1 nitration was detectable by immunoblotting (data not shown).
We then investigated whether impairment of IRE binding by IRP-1 in IFN-␥/LPS/PMA-treated macrophages, shown in Fig.  4A, was relieved by 2% 2-ME when nitration was prevented. As shown in Fig. 8, in the presence of SHA or pHBAH, full IREbinding activity of IRP-1 regained the same level as in macrophages activated with IFN-␥/LPS without co-stimulation with PMA (compare lanes 4 and 5 with lanes 2 and 3). DISCUSSION IRP-1, the most abundant pool of potential IRE-binding activity in mammalian tissues, is equipped with a redox-sensitive [Fe-S] cluster and is therefore liable to respond to NO ⅐ and congeners (30). Recently, we have shown that IRP-1 is also a potential protein target of nitration by peroxynitrite in vitro (13). However, important questions still remain regarding the physiological relevance of IRP-1 nitration and how this process is triggered in vivo. In this study, we investigated IRP-1 tyro- A, nitrite production was determined in the cell culture medium, and cytosolic fractions of control or treated cells were prepared and used for cytosolic aconitase activity measurement. B, cytosols were also immunoprecipitated with anti-IRP-1 antibody, and the resulting samples were analyzed by Western blot analysis using a specific anti-nitrotyrosine antibody. Recombinant human IRP-1 (rhIRP1) and bovine serum albumin (BSA) were exposed to synthetic peroxynitrite (ONOO Ϫ ) and served as positive nitrated controls. The experiments were performed twice. n.d., not detectable. To ensure that these two species were produced at once, we treated macrophages in a two-step manner. We first stimulated macrophages with IFN-␥/LPS for several hours to allow NO synthase-2 to be both expressed and functional and then we exposed NO-producing macrophages to PMA to stimulate O 2 .
generation via protein kinase C activation. Under these conditions, which mimic an in vivo inflammatory situation, we show for the first time that cellular IRP-1 is sensitive to nitration in vivo. IRP-1 tyrosine nitration in stimulated macrophages was dependent on enzymatically produced NO ⅐ , because the combination of the NO synthase inhibitors L-NMA and EIT, which fully blocked NO ⅐ production, prevented endogenous IRP-1 nitration almost completely. Nitrogen dioxide (NO 2 ⅐ ), which is the major nitrating species, can be formed by rapid decomposition of peroxynitrite at physiological pH or by the oxidation of nitrite (the stable endproduct of NO metabolism) by peroxidases in the presence of H 2 O 2 (31). Nonetheless, the question of which of these two proposed mechanisms promotes in vivo protein nitration has been fiercely debated. IRP-1 nitration by IFN-␥/LPS-or IFN-␥/LPS/PMA-treated macrophages was fully prevented by both peroxynitrite scavengers and myeloperoxidase inhibitors, suggesting that peroxynitrite may operate along with the nitrite/ H 2 O 2 /peroxidase activities to nitrate IRP-1. However, as prevention was complete with both types of compounds, this scenario is unlikely. We therefore wondered about the actual specificity of the peroxynitrite scavengers. First of all, it is worth recalling that most peroxynitrite scavengers also react with O 2 . , which is the precursor of H 2 O 2 , the substrate of my-  (32). Finally, it has been proposed that peroxidases may catalyze peroxynitrite-dependent nitration via a two-electron oxidation reaction (33)(34)(35). Altogether, these issues may explain why IRP-1 nitration was prevented both by the two presumed peroxynitrite scavengers and by myeloperoxidase inhibitors, casting doubt on the participation of the peroxynitrite pathway in peroxidase-independent IRP-1 endogenous nitration. Taken together, these data and considerations are in favor of the nitrite/H 2 O 2 /peroxidase pathway as the major IRP-1 nitrating pathway in physiologically activated macrophages. IRP-1 is a bifunctional metalloprotein in which the aconitase and IRE-binding activities are mutually exclusive, depending on the presence or absence of its [4Fe-4S] cluster (1). In the present study, we show that endogenous nitration of IRP-1 in activated macrophages boosted with PMA is associated with impairment of both IRP-1 functions. We found that lack of IRP-1 binding to IRE in cells co-generating NO ⅐ and O 2 . was not due to IRP-1 protein degradation but rather to a post-translational modification that was not reversible by high 2-ME concentrations. Importantly, the recovery of IRP-1 IRE binding under those reducing conditions occurred only when nitration was prevented by myeloperoxidase inhibitors. A number of reports have indicated that the bulky nitro group can prevent protein activity by steric hindrance (36,37). Therefore, we postulate that nitration, a fairly stable modification, arose at a strategic location(s) in the backbone of the IRP-1 apo-protein, blocking its binding to the IRE motif. Similarly, it has been reported that endogenous nitration of peroxisome proliferatoractivated receptor-␥ inhibited its ligand-dependent translocation from cytosol to nucleus in LPS-stimulated RAW 264.7 macrophages (38).
To determine whether nitration of IRP-1 follows or precedes aconitase loss, we exposed IRP-1-containing cytosols to a concentration of peroxynitrite that is efficient for nitration in the presence of cis-aconitate. We took advantage of the capacity of cis-aconitate to interact directly with [4Fe-4S] clusters of aconitases, which results in holo-IRP-1 stabilization (23). We observed that cis-aconitate dose-dependently prevented IRP-1 nitration by peroxynitrite or SIN-1, as well as inhibiting its aconitase activity. These results point to nitration as an event downstream of [Fe-S] disruption and therefore of aconitase loss. Accordingly, nitrite, a substrate for myeloperoxidase, which can trigger global nitration in RAW 264.7 macrophages (15), led neither to inactivation of cytosolic aconitase nor to subsequent endogenous IRP-1 nitration. Since the aconitase form of IRP-1 was insensitive to nitration, we can presume that potentially reactive tyrosine(s) of holo-IRP-1 is not accessible to the nitrating agent. It is held that IRP-1, devoid of its cluster, gains a more relaxed conformation than the [4Fe-4S] clustercontaining IRP-1, rendering its regulatory binding site accessible to IRE (39,40). As peroxynitrite and NO ⅐ disrupt the [Fe-S] cluster of IRP-1 in vitro (13), we propose that complete loss of [Fe-S] cluster is a prerequisite to allow one or several tyrosine residues initially buried in holo-IRP-1 to become accessible to nitrating species. In our cellular model, it is worth noting that moderate nitration of IRP-1 mediated by IFN-␥/LPS stimulation was not associated with any loss of IRP-1 IRE-binding activity. Under these conditions, nitration of IRP-1 would modify tyrosines not critical for adequate binding to IRE. In contrast, endogenous (without 2-ME) and full (with 2-ME) IRE-binding activities were reduced when IFN-␥/LPS-stimulated cells were also exposed to PMA. Generation of an oxidative burst upon PMA would therefore contribute to nitration of specific tyrosines close to the IRE-binding domain. This may simply result from strengthened overall IRP-1 nitration. Alternatively, based on a previous report (41), it was tempting to consider that PMAboosted production of reactive oxygen species, associated with iron knocked out from an IRP-1-disrupted [Fe-S] cluster, would favor local nitration through Fenton chemistry. Indeed, in vitro studies previously showed that redox metals, particularly iron, can catalyze protein tyrosine nitration promoted by peroxynitrite (42) or by nitrite/H 2 O 2 (41,43). However, our results showing that myeloperoxidase inhibitors completely abrogated nitration strongly suggest that "autocatalytic" nitration of IRP-1 by its own iron was not relevant under these conditions. In IRP-1, seven conserved tyrosines are located in one part of the IRE-binding domain, including the [Fe-S] cluster-proximal tyrosine 501 (44), and are thus good candidates for nitration. Future studies, including mass spectrometry experiments, are FIG. 8. Effect of salicylhydroxamic acid and p-hydroxybenzoic acid hydrazide on IRP-1 IRE-binding in IFN-␥/LPS/PMA-stimulated macrophages. Cells were activated with 20 units/ml IFN-␥ and 50 ng/ml LPS with or without 50 nM PMA as described for Fig. 2. In parallel, stimulated macrophages were also incubated with 50 M SHA or pHBAH. After 24 h, cytosols were prepared, and IRE-binding activity was determined by electrophoretic mobility shift assay in the presence of 2% 2-ME. The experiments were performed three times, and a representative result is shown. required to localize nitrated tyrosines, but we have already demonstrated that endogenous tyrosine nitration can persistently hamper IRP-1 functioning.
The biological significance of these findings remains debatable, but some clues may be foreseen in the particular metabolism of macrophages, which are prone to produce both reactive oxygen species and nitrogen-derived species. In pathophysiological situations such as inflammatory diseases, target cells adjacent to NO-producing macrophages undergo NO/IRP-1 regulation (45). Upon IRP-1 activation, ferritin is repressed and transferrin receptor is up-regulated, resulting in a toxic iron overload (46). Moreover, recent evidence indicates that nitration of tyrosine-bound proteins is a dynamic process sensitive to oxygen tension (47). Accordingly, we speculate that prevention of IRP-1 activation by nitration in macrophages during inflammatory processes might be an intrinsic protective mechanism against the noxious effect of iron through Fenton-like reactions.