Tyrosine nitration in prostaglandin H 2 synthase

In this study, we investigated the effects of various nitrogen oxide (NO x ) species on the extent of prostaglandin H 2 synthase-1 (PGHS-1) nitration in puriﬁed protein and in vascular smooth muscle cells. We also examined PGHS-1 activity under these conditions and found the degree of nitration to correlate inversely with enzyme activity. In addition, since NO x species are thought to invoke damage during the pathogenesis of atherosclerosis, we examined human atheromatous tissue for PGHS-1 nitration. Both peroxynitrite and tetranitromethane induced Tyr nitration of puriﬁed PGHS-1, whereas 1-hydroxy-2-oxo-3-( N -methyl-aminopropyl)-3-methyl-1-triazene (NOC-7; a nitric oxide-releasing compound) did not. Smooth muscle cells treated with peroxynitrite showed PGHS-1 nitration. The extent of nitration by speciﬁc NO x species was determined by electrospray ionization mass spectrometry. Tetranitromethane was more effective than peroxynitrite, NOC-7, and nitrogen dioxide at nitrating a tyrosine-containing peptide (12%, 5%, 1%, and (cid:2) 1% nitration, respectively). Nitrogen dioxide and, to a lesser extent, peroxynitrite, induced dityrosine formation. Using UV/Vis spectroscopy, it was es-timated prolonged oxidative tyrosine


Supplementary key words peroxynitrite • atherosclerosis • cyclooxygenase
The rapid, diffusion-limited reaction of superoxide anion and nitric oxide (NO • ) (1) leads to the formation of peroxynitrite (ONOO Ϫ ), which can initiate a number of toxic reactions in vivo. For example, ONOO Ϫ is capable of initiating lipid peroxidation (2), reacting with sulfhydryl groups (3), and causing oxidative damage by introducing tyrosine nitration (4). When produced by inflam-matory cells during the immune response, ONOO Ϫ has a beneficial role by destroying invading microorganisms, but under pathological conditions, ONOO Ϫ assumes a deleterious role, and may contribute to vascular disease (5).
We have previously demonstrated that specific nitrogen oxides (NO x ) have different modulatory effects on the enzyme prostaglandin H 2 synthase (PGHS; also known as cyclooxygenase) (6). This enzyme, which exists in both the constitutive (prostaglandin H 2 synthase-1; PGHS-1) and inducible (prostaglandin H 2 synthase-2; PGHS-2) forms, plays an important role in regulating vascular homeostasis. While PGHS-1 mediates housekeeping functions (7), PGHS-2 is a mediator of inflammation and has been closely associated with tumorigenesis and atherosclerosis (8,9). Both isozymes function by rapidly converting arachidonic acid to the cyclic endoperoxide prostaglandin H 2 (PGH 2 ), via a PGG 2 intermediate. PGH 2 can subsequently be metabolized to prostacyclin (PGI 2 ) by prostacyclin synthase. PGI 2 is known to regulate processes such as vasodilation and inhibition of both platelet aggregation and smooth muscle cell proliferation (10,11). Investigations of the effect of NO • on purified PGHS-1 enzyme activity have been controversial with reports of activation (6,12,13) as well as inhibition (14)(15)(16). The reason for these reported differences may be due, in part, to the complexity of NO • chemistry, with the result that various NO x species are formed, that include (but are not limited to): nitrosonium ion (NO ϩ ), nitroxyl anion (NO Ϫ ), nitrite (NO 2 Ϫ ), nitrate (NO 3 Ϫ ), and peroxynitrite (ONOO Ϫ ) (17). While ONOO Ϫ activates PGHS-1 in the presence of arachidonic acid substrate (6,13,18), NO • has an inhibitory effect (6). Peroxynitrite most likely activates PGHS-1 in a manner similar to that described for peroxide activation of PGHS-1 (19). Thus, in the process of activation, the Fe(III) porphyrin in PGHS is converted to an Fe(IV) ϭ O porphyrin cation radical. Indeed, ONOO Ϫ reacts with Fe(III) porphyrins on a very fast timescale (5 ϫ 10 7 M Ϫ 1 s Ϫ 1 ) (20).
While ONOO Ϫ may stimulate PGHS activity via its reaction with the heme moiety, its presence may also lead to tyrosine nitration, a reaction that is up to 100-fold slower than its reaction with Fe of hemoproteins (18). In this regard, we have examined the nitrating ability of various NO x reagents and have measured the extent of purified PGHS-1 nitration by ONOO Ϫ . We have also examined the effect of ONOO Ϫ on PGHS-1 nitration in smooth muscle cells. Further, since ONOO Ϫ is thought to cause tissue damage during atherosclerosis (21,22), we investigated the hypothesis that elevated levels of nitrated PGHS exist in human atheromatous tissue.

Isolation and purification of PGHS-1 enzyme from ram seminal vesicles
The isolation and purification of PGHS-1 from ram seminal vesicles was performed according to previously published literature methods (23).

Measurement of cyclooxygenase activity by oxygen monitor
Cyclooxygenase activity was measured using an oxygen electrode (YSI Model 53) attached to a recorder (Hitachi 561) according to previously published methods (24). One unit of cyclooxygenase activity represents a peak velocity of 1 nmol of oxygen/min.

Preparation of samples for mass spectrometric and UV/Vis spectroscopic experiments
The reaction of several NO x compounds ( 15 NO 2 • , ONOO Ϫ , TNM, and NOC-7) with a Tyr-containing peptide (Thr-Tyr-Ser; mw ϭ 369.4) was monitored by mass spectrometry and UV/Vis spectroscopy. These reactions were conducted in ammonium bicarbonate buffer (20 mM; pH 7.4) at ambient temperature for 1 h. For the reaction with 15 NO 2 • , the peptide was deaerated in a Thunberg cuvette using a vacuum/argon gas manifold and exposed to 15 NO 2 • gas (1 atm). When using ONOO Ϫ , we added ONOO Ϫ from a stock solution directly to the peptide; for experiments using TNM, we prepared a stock solution (200 mM) in ethanol; and when using NOC-7, we prepared a stock solution (10 mM) in deionized water. Unless otherwise indicated, 5 mM peptide was reacted with equimolar concentrations of the NO x compound (in the case of 15 NO 2 • , we exposed 5 mM peptide to 1 atm 15 NO 2 • gas). Following the addition of either ONOO Ϫ or NOC-7 to the peptide, the pH of the solution was adjusted to pH 7.4, if necessary.
Preparation of PGHS-1 samples for UV/Vis spectrometric analysis involved adding ONOO Ϫ (1 mM) to purified PGHS-1 (7-13 M) in ammonium bicarbonate for 1 h at ambient temperature. The mixture contained residual Tris buffer from the protein purification process. Final concentrations of Tris and ammonium bicarbonate were typically 5 mM and 15 mM, respectively.

Mass spectrometry
Electrospray ionization mass spectra were obtained on a Quattro II triple quadrupole mass spectrometer (Micromass, Beverly, MA) fitted with a standard electrospray probe. Samples from the nitration experiments (described above) were diluted ( ‫ف‬ 50 M) in 50:50:2 (v/v/v) methanol-water-acetic acid mixtures and infused (5 l/min) via a syringe pump (Harvard Apparatus, Natick, MA). The capillary voltage, cone voltage, and source temperature were kept at 3.5 kV, 18 V, and 80 Њ C, respectively. In tandem mass spectrometric experiments, argon (3.6 ϫ 10 Ϫ 4 mbar) was used as the collision gas (20 V collision energy). To take advantage of the selective detection capability of the tandem mass spectrometer, each sample was analyzed in the Neutral Loss scanning mode as well as in the Single Quadrupole scanning mode of the instrument. In the collision induced decomposition (CID) of the singly protonated Thr-Tyr-Ser peptide ion, a facile channel is the cleavage of the peptide bond between Tyr and Ser residues. The loss of a neutral serine molecule results in a b 2 ion (25), which is 105 Da lighter than its precursor. When the two quadrupoles bracketing the collision chamber are scanned in tandem offset by 105 Da (Neutral Loss 105 mode), the Thr-Tyr-Ser peptide and species that are potentially modified on Tyr are selectively detected. Entirely similar results were observed in Single Quadrupole scans (spectra not shown), but the product peaks were difficult to interpret at low conversions when not well separated from unavoidable background.

UV/Vis spectroscopy
Experiments were performed using a Perkin Elmer Lambda 20 spectrophotometer. Matching quartz suprasil ® precision cells (Hellma, 1 cm pathlength; 1 ml or 100 l volume) were used. The concentration of ONOO Ϫ was determined by measuring its absorbance at 302 nm ( ϭ 1,670 M Ϫ1 cm Ϫ1 ) (26). The extent of nitration of the Thr-Tyr-Ser peptide by specific NO x reagents was calculated by measuring the absorption at 428 nm ( ϭ 4100 M Ϫ1 cm Ϫ1 ) (27) at pH 9 after subtracting the unreacted peptide. The extent of apo-and holo-PGHS-1 nitration was calculated by measuring the absorption at 428 nm at pH 9, and subtracting any contribution at pH 6. Holo-PGHS-1 was prepared by optically titrating apo-PGHS-1 with hematin. A 1:1 hematin-protein complex was created by monitoring the Soret Band of PGHS-1 at 410 nm.

Treatment of purified PGHS-1 with nitrating agents
Purified holo-PGHS-1 (0.1 M Tris buffer) was exposed to a nitrating agent for 1 h at 37ЊC as follows. For experiments using ONOO Ϫ , the appropriate amount of stock ONOO Ϫ (38-48 mM) was added directly to the holo-PGHS-1 solution. SIN-1 was prepared as an anaerobic stock solution (10 mM) in a Schlenk tube as described previously (6). The appropriate amounts of SIN-1 were directly added to the protein via a Hamilton syringe. For experiments using TNM, fresh stock solutions (200 mM) were prepared in ethanol. For the treatment of PGHS-1 with NOC-7 or nitrite, stock solutions (10 and 20 mM, respectively) were prepared in deionized water.

Smooth muscle cell culture
Smooth muscle cells were propagated from explants of rat thoracic aorta and cultured as previously described (6). Rat cells were maintained in DMEM supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) glutamine as described previously (6). Cells between passage 8-15 were used in experiments. We did not see significant variation in our results in response to passage number.

Treatment of smooth muscle cells with ONOO Ϫ , tetranitromethane, or NOC-7
Rat smooth muscle cells were grown to near-confluence in 10 cm diameter dishes (containing approximately 8.8 ϫ 10 6 cells/ dish). The cells were quiesced for 24 h using serum-free DMEM (quiescent medium) containing 1% (v/v) ITS to suppress PGHS-2 expression (6). The cells were then exposed to either quiescent medium or quiescent medium containing ONOO Ϫ , TNM, or NOC-7 for 1 h at 37ЊC, as described below.
For experiments in which cells were exposed to ONOO Ϫ , we added the appropriate amount of ONOO Ϫ from a stock solution (38-48 mM) directly to the dish containing quiescent DMEM to yield the desired concentrations (6). For experiments using TNM, we prepared fresh stock solutions of TNM (200 mM) in ethanol and added the appropriate amounts to quiescent medium within 1 min of preparation. The final ethanol volumes (0.05Ϫ0.25%) did not affect cell viability. For the treatment of cells with NOC-7, we prepared a 10 mM stock solution in deionized water (6).

Preparation of smooth muscle cells for Western blotting
Following treatment with ONOO Ϫ , TNM, or NOC-7, the cells were washed once with cold PBS (7 ml; without Mg 2ϩ or Ca 2ϩ ), and then ice-cold lysis buffer (500 l; containing 0.15 M NaCl, 100 mM Tris pH 8.0, 1% Tween-20, 1 mM EDTA, 1 mM PMSF, and 10 g/ml aprotinin and 10 g/ml leupeptin) was added to each dish. The cells were scraped, put on ice for 30 min, briefly sonicated (Branson Sonicator), and clarified by centrifugation for 10 min (9,500 g) at 4ЊC. The protein content of the supernatants was measured by a modified Lowry method (28).

Detection of 3-nitrotyrosine in NO x -treated cells or in NO x -treated purified PGHS-1 by Western blotting
Protein (5-40g) was separated on a 10% acrylamide gel and transferred onto a nitrocellulose membrane. The membrane was immunoblotted with either a mouse monoclonal or a rabbit polyclonal 3-nitrotyrosine antibody according to manufacturer's instructions (Upstate Biotechnology). The bands were visualized using enhanced chemiluminescence (ECL or ECL ϩ ). Blots were stripped by agitating at 50ЊC for 30 min in Tris buffer (1 M, containing 20% SDS and 100 mM ␤-mercaptoethanol at pH 6.7) and reprobed with an anti-mouse PGHS-1 antibody.

Immunoprecipitation of nitrated PGHS-1 from rat smooth muscle cells
Rat smooth muscle cells were lysed as above or in a RIPA lysis buffer containing protease inhibitors. Polyclonal 3-nitrotyrosine antibody (8 g) was added to the lysate (1 mg/ml), which was incubated on a nutator at 4ЊC for 1 h. Protein A-agarose beads (40 l of a 50% slurry in PBS) were added to the lysates, which were again incubated on a nutator at 4ЊC overnight. The beads were washed three times with RIPA buffer (1 ml), centrifuged (30 s at 16,000 g), aspirated, and finally boiled in Laemmli sample buffer (30 l). Following a final centrifugation (30 s at 16,000 g), the samples were separated on an 8% SDS-PAGE gel and immunoblotted with a PGHS-1 monoclonal antibody.

Purification of atherosclerotic tissue obtained from endarterectomy patients for Western blotting
Human tissue was obtained from carotid endarterectomy procedures (five patients) under an IRB-approved protocol, and was dissected into sections representing the media, the fibrous cap, and the atheromatous necrotic core. The tissue sections were subjected to a series of purification steps equivalent to those used when isolating PGHS from sheep seminal vesicles, except that the buffers contained protease inhibitors (23).

Immunoprecipitation of nitrated PGHS-1 from human atherosclerotic tissue
Atheromatous tissue (three patients), surrounding media (two patients) and uninvolved tissue (two patients) were homogenized using a Tissue Tearor (Dremel) and sonicated for 30 s. The homogenate was centrifuged (16,000 g for 15 min). The resulting lysate (1 mg/ml in PBS) was subjected to pre-clarification with Protein G-agarose beads (50 l of a 50% slurry in PBS for 30 min) to reduce non-specific binding. The lysate/bead mixture was centrifuged (5,000 g for 5 min, followed by 16,000 g for 1 min) to remove the beads. The clarified lysate was incubated with polyclonal nitrotyrosine antibody (3.6 g overnight at 4ЊC) and then with Protein G-agarose beads (75 l of a 50% slurry overnight at 4ЊC). The bound agarose beads were collected by centrifugation (5,000 g for 10 min, followed by 16,000 g for 1 min), washed three times with PBS and resuspended in Laemmli sample buffer (75 l). The bead-immunocomplex was boiled (5 min), the sample (25 l) separated by SDS-PAGE (10%) and immunoblotted with monoclonal PGHS-1 antibody as described above.

Miscellaneous
Cell viability was assessed by in-situ staining with trypan blue and cell number was determined by hemocytometer. No cell injury was observed using 200 and 500 M ONOO Ϫ , although some cell death was associated with using 1,000 M ONOO Ϫ . Scion Image software was used to quantify Western blot band densities (Scion Corporation).

Nitrating efficiencies of NO x compounds
In order to establish the nitrating efficiency of specific NO x agents, a Tyr-containing peptide (Thr-Tyr-Ser, mw ϭ 369.4) was reacted with 15 NO 2 • , ONOO Ϫ , TNM (27), and the NO • -generator, NOC-7 (29), and analyzed by electrospray ionization mass spectrometry (ESI-MS) (Fig. 1) 1D) led to 4.5% and 11.5% Tyr nitration, respectively. The use of 15 NO 2 • instead yielded a dipeptide (7.9%; m/z ϭ 373.3), which is the coupling product of two Tyr radicals (30), and was also observed for ONOO Ϫ and TNM (4.8% and 1.7%, respectively) (data not shown). For ONOO Ϫ and TNM, low amounts of nitrated dipeptide (0.7% and 0.4%, respectively) were observed. From these data, one may generate a hierarchy of nitrating efficiency that may be tested in a biological setting.

ONOO Ϫ nitrates purified PGHS-1
Purified PGHS-1 was exposed to ONOO Ϫ , SIN-1 (an ONOO Ϫ generator) (31), or TNM, and the presence of 3-nitrotyrosine was probed by Western blotting (Fig. 2). All three reagents caused nitration of PGHS-1, although in the case of SIN-1 (Fig. 2B), high concentrations were required (500 M and above). Nitration was not observed in the presence of nitrite (200-5,000 M), a degradation byproduct of ONOO Ϫ , or NOC-7 (data not shown). As predicted by those agents tested by mass spectrometry, these data demonstrate that ONOO Ϫ , SIN-1, and TNM are effective PGHS-1 nitrating agents whereas NO • and nitrite are not. Fig. 1. ONOO Ϫ nitrates a simple Tyr-containing peptide more efficiently than 1-hydroxy-2-oxo-3-(N-methyl-aminopropyl)-3-methyl-1-triazene (NOC-7) and NO 2 • , but less than tetranitromethane (TNM). A Tyr-containing peptide (Thr-Tyr-Ser; mw ϭ 369.4) was reacted with specific NO x compounds for 1 h in ammonium bicarbonate buffer (20 mM; pH 7.4) at ambient temperature, and then analyzed by electrospray ionization mass spectrometry (ESI-MS). The samples analyzed included: (A) the Tyr-containing peptide (labeled "control"; 5 mM) and the Tyr-containing peptide (5 mM) treated with either (B) 15 N-labeled NO 2 • (one atmosphere, under anaerobic conditions), (C) ONOO Ϫ (5 mM), (D) TNM (5 mM), or (E) NOC-7 (5 mM). Each spectrum was normalized to the peak of unreacted peptide (m/z ϭ 370). Note that while the lower mass range of the spectra in each case is magnified ϫ 16, the higher mass range, where nitration is expected to appear is magnified by different factors for individual samples. Quantification of tyrosine nitration By ONOO Ϫ in purified PGHS-1 UV/Vis spectroscopy was used to quantify the extent of Tyr nitration in peptide and purified PGHS-1 by ONOO Ϫ . Tyr nitration is characterized by an absorbance in the visible region of the UV/Vis spectrum which at pH Ͼ 8 occurs above 400 nm, but at pH 6 shifts to lower wavelengths (32). Spectra were generated at pH 6 and 9 upon the treatment of peptide, apo-PGHS-1 (heme-free), or holo-PGHS-1 (heme-containing) with ONOO Ϫ (Fig. 3). Under these conditions, we calculate ‫%5ف‬ nitration of the peptide (Fig. 3A), which was also observed by mass spectrometry (Fig. 1). Further, it was calculated that, approximately 2 Tyr/apo-PGHS-1 subunit are nitrated (Fig. 3B). Nitration of holo-PGHS-1 by ONOO Ϫ was also measured by this method. Since heme absorbs very close to the region where nitrated tyrosines absorb, it is not very informative to show a direct comparison of the spectra obtained for nitrated apo-and holo-PGHS-1. It is essential to employ difference spectra to minimize absorbance contributions arising from the Soret band of the heme group. Figure 3C shows difference spectra with maxima at 438 nm for the reaction between ONOO Ϫ and apo-and holo-PGHS-1. The peak intensities reveal that 2.4 and 1.5 Tyr per sunbunit were nitrated for apo-PGHS-1 and holo-PGHS-1, respectively. The 38% reduction in nitration of holo-PGHS-1 may be attributed to the scavenging reaction of heme with ONOO Ϫ . Regardless, incubation of both apo-or holo-PGHS-1 with excess ONOO Ϫ led to changes in the UV/ Vis spectrum consistent with Tyr nitration.

ONOO Ϫ nitrates PGHS-1 in vascular smooth muscle cells
Immunoprecipitation experiments were performed to determine the ability of specific NO x reagents to nitrate PGHS-1 in vascular smooth muscle cells. After their exposure to ONOO Ϫ or TNM, proteins containing 3-nitrotyrosine were immunoprecipitated from cells using a polyclonal 3-nitrotyrosine antibody. The immunoprecipitated samples were loaded onto a gel and immunoblotted with a PGHS-1 monoclonal antibody (Fig. 4). While very little or no PGHS-1 was detected in the absence of ONOO Ϫ , it was detected in the presence of ONOO Ϫ , indicating that it is nitrated (Fig. 4A). TNM also induced a dose-dependent increase in PGHS-1 nitration (Fig. 4B). PGHS-1 nitration was not observed when using NOC-7 (data not shown). In parallel experiments, we determined the activity of PGHS-1 following ONOO Ϫ treatment. After incubation with different concentrations of ONOO Ϫ , cells were treated with arachidonic acid and the supernatant media analyzed for PGE 2 production (Fig. 4A, top). The data indicate that PGHS-1 activity is inversely related to PGHS-1 nitration.
In separate experiments, Western blots of cell lysates obtained from treatment of smooth muscle cells with sodium nitrite (200-1,000 M) revealed no nitration at 70 kDa. These data indicate that nitrite, a degradation byproduct of ONOO Ϫ , is not the cause of nitration in cells under our experimental conditions. The 70 kDa band observed as a result of nitrated PGHS-1 was not visible when Fig. 3. UV/Vis spectroscopy demonstrates that ONOO Ϫ nitrates a simple Tyr-containing peptide and PGHS-1. A: A Tyr-containing peptide (Thr-Tyr-Ser; mw ϭ 369.4, 0.1 mM) was exposed to ONOO Ϫ (0.1 mM) in ammonium bicarbonate (0.1 M; pH 7.4) at ambient temperature for 1 h, and then analyzed by UV/Vis spectroscopy. The pH of the solution was changed to six (by careful addition of 5N HCl) and then to nine (by careful addition of 5N NaOH). After each pH change, a spectrum was recorded. B: Apo-PGHS-1 (7 M) was exposed to ONOO Ϫ (1 mM) in ammonium bicarbonate buffer (15 mM) (containing 5 mM Tris buffer) for 1 h at ambient temperature, and analyzed by UV/Vis spectroscopy at pH 6 and pH 9. C: Apo-PGHS-1 (13.4 M) and holo-PGHS-1 (13.4 M) were each exposed to ONOO Ϫ (1 mM) in Tris buffer (100 mM, pH 7.4) for 1 h at ambient temperature, and UV/Vis spectra were recorded at pH 6 and pH 9. The spectra shown represent difference spectra that were obtained upon subtracting the spectrum at pH 6 from that obtained at pH 9 and provide a direct comparison of the amount of ONOO --induced Tyr nitration in apo-PGHS-1 and holo-PGHS-1, respectively. These data are representative of one experiment performed three times. by guest, on July 19, 2018 www.jlr.org Downloaded from using either secondary antibody alone or 3-nitrotyrosine antibody that had been rendered inactive by binding to 3-nitrotyrosine (data not shown). In summary, these data demonstrate that both ONOO Ϫ and TNM are effective at nitrating PGHS-1 in vascular smooth muscle cells.

Human atherosclerotic tissue contains nitrated PGHS-1
The media, necrotic core, and fibrous cap from human atherosclerotic lesions were each purified and probed by Western blotting for 3-nitrotyrosine, PGHS-1 and PGHS-2 (Fig. 5). A nitrated 70 kDa protein was present in the ne-crotic core and fibrous cap, with less in the surrounding media. In contrast, PGHS-1 levels were similar in all three sections. PGHS-2 was induced in the necrotic core and fibrous cap, with less appearing in the media (9). Thus, the immunoblotting data in Fig. 5 show that a 70 kDa protein (possibly PGHS-1 or PGHS-2) is nitrated in the fibrous cap and necrotic core of human atherosclerotic lesions.
Next, immunoprecipitation experiments were performed to determine whether nitrated PGHS-1 is found in human atherosclerotic tissue (Fig. 6). Nitrated proteins from atherosclerotic and uninvolved control tissue (from human mammary arteries) were immunoprecipitated using a polyclonal 3-nitrotyrosine antibody. The immunoprecipitated samples were separated by gel electrophoresis and immunoblotted with a PGHS-1 monoclonal antibody (Fig. 6A). While very little PGHS-1 nitration was detected in uninvolved tissue, substantial PGHS-1 nitration was present in atherosclerotic lesions. Next, we compared the degree of PGHS-1 nitration in the media proximate to the atherosclerotic lesion to the degree of nitration in the lesion (Fig. 6B). While minimal nitration was observed in the media, significant nitration was observed in the lesion (P ϭ 0.01).

DISCUSSION
Since NO • and PGI 2 perform similar functions and their production can be stimulated by the same inflamma- Fig. 4. ONOO Ϫ and TNM nitrate PGHS-1 in smooth muscle cells. A: Aortic smooth muscle cells were grown to near confluence in 10 cm dishes, quiesced for 24 h, and treated with serum-free DMEM containing ONOO Ϫ (0, 200 M, 500 M, and 1,000 M). After incubating at 37ЊC for 1 h, the media was removed. Top: PGHS-1 activity was determined by measuring PGE 2 production in the supernatant media after arachidonic acid treatment (10 M for 30 min at 37ЊC). Middle: Nitration of PGHS-1 was determined by immunoprecipitation using a polyclonal anti-rabbit 3-nitrotyrosine antibody and immunoblotting with a PGHS-1 monoclonal antibody. Bottom: Nitration was quantified by measuring band densities. B: Cells were grown as above and treated with TNM (0, 200 M and 500 M). Top: Nitration of PGHS-1 was determined by immunoprecipitation as above. Bottom: Nitration was quantified by measuring band densities and normalizing the control bands to one. These data are representative of one experiment performed twice.

Fig. 5.
Western blot of human atherosclerotic lesion comparing nitration levels in the media, fibrous cap, and necrotic core. Human atherosclerotic tissue (n ϭ 5) was sectioned into the media, fibrous cap, and necrotic core, and subjected to a purification procedure designed to isolate the PGHS-rich microsomes (described in Experimental Procedures). Western blots of purified protein from each tissue section were probed with monoclonal 3-nitrotyrosine, PGHS-1, and PGHS-2 antibodies. Data for relative nitration was obtained by calculating a 3-nitrotyrosine/PGHS-1 band density ratio. The band density ratio was normalized to one in the control case. by guest, on July 19, 2018 www.jlr.org Downloaded from tory cytokines (33), it has previously been proposed that there is "cross talk" between the NO • and arachidonic acid pathways. A form of NO x that activates PGHS-1 and PGHS-2 is ONOO Ϫ (6,13,18). A necessary requirement for PGHS activation by ONOO Ϫ is the simultaneous addition of arachidonic acid and ONOO Ϫ . If ONOO Ϫ is added prior to arachidonic acid, enzyme deactivation occurs which has been correlated with an increase in 3-nitrotyrosine levels (18). Taken together, these studies indicate that ONOO Ϫ has the potential to either stimulate or inhibit PGHS activity depending on the availability of arachidonic acid and concentration of ONOO Ϫ . This is the first study to demonstrate that, in the absence of exogenously-added arachidonic acid, ONOO Ϫ is capable of nitrating PGHS-1 in vascular smooth muscle cells. Immunoprecipitation experiments unequivocally demonstrate that PGHS-1 in smooth muscle cells is nitrated by ONOO Ϫ and TNM. In contrast, the NO-releasing compound, NOC-7, did not lead to any observable nitration. Indeed, NO • cannot nitrate Tyr residues directly, but can couple with Tyr cation radicals, resulting in the eventual formation of 3-nitrotyrosine (34). Nitrite, which has been implicated in tyrosine nitration in a peroxidasedependent pathway (35,36), also had no effect. Similar results were observed with purified PGHS-1. Although SIN-1 was effective at nitrating PGHS, high doses were required. In light of SIN-1's limited ability to donate O 2 Ϫ in the presence of biological electron acceptors (e.g., heme proteins) (37) Ϫ ), heterolytic cleavage of which yields NO 2 ϩ , a highly reactive nitrating species (39). UV/Vis spectroscopic data corroborated ESI-MS results, indicating that UV/Vis spectroscopy is a reliable technique for monitoring nitration. By UV/Vis spectroscopy, ONOO Ϫ was demonstrated to nitrate ‫2ف‬ Tyr residues per PGHS-1 subunit (18). TNM has previously been found to nitrate up to three Tyr residues per PGHS-1 subunit (using a 43:1 molar ratio of TNM-PGHS-1) (40).
It is not yet known which two Tyr residues in PGHS-1 are nitrated by ONOO Ϫ . Ovine and human PGHS-1 posess 27 Tyr residues, and yet the significance of only a few is known. In particular, studies have been conducted to examine the importance of Tyr 355, Tyr 348, and Tyr 385. While the mutation of either Tyr355 or Tyr348 to Phe results in little or no effect, the mutation of Tyr385 results in a complete loss of cyclooxygenase activity (41,42). As we have found that the degree of Tyr nitration in PGHS-1 inversely correlates with enzyme activity, it is possible that Tyr385 is nitrated. However, more conclusive studies must be conducted in order to corroborate this speculation. It is also important to consider that concomitant cysteine oxidation by ONOO Ϫ may contribute to PGHS-1 deactivation.
Levels of NO x accumulation have been reported in some disease states to be as high as 36 M (43). The concentrations of ONOO Ϫ employed in the study, although high, cannot be used as a direct measure of the amount of active species. It is important to recognize that the half-life of ONOO Ϫ is ‫1ف‬ s under physiological conditions (44). This means that 500 M ONOO Ϫ will decay to nanomolar concentrations in seconds (i.e., 500 nM in 10 s, and to 16 nM in 15 s). In addition, ONOO Ϫ is consumed in reactions other than nitration that will deplete the stock of added ONOO Ϫ . Thus, in order to generate ONOO Ϫ concentrations found in local environments, it is necessary to add excessive amounts.
Extensive protein nitration has been observed in atherosclerotic lesions (21,22). To test the hypothesis that Fig. 6. PGHS-1 is nitrated in human atherosclerotic tissue. A: Atheromatous (n ϭ 3) and uninvolved tissue (n ϭ 2) was homogenized as described in Materials and Methods. Proteins containing 3-nitrotyrosine were immunoprecipitated using a polyclonal 3-nitrotyrosine antibody and then immunoblotted with a PGHS-1 monoclonal antibody. IgG control refers to rabbit anti-nitrotyrosine that was subjected to the same protocol as both the control and atheromatous tissue samples to ensure that the 70 kDa PGHS-1 immunoreactive band was not the result of secondary interactions. B: Atheromatous tissue (n ϭ 3) and the surrounding media (n ϭ 2) were homogenized and immunoprecipitated with polyclonal 3-nitrotyrosine antibody as described in Materials and Methods. Samples were separated by SDS-PAGE and immunoblotted with a monoclonal PGHS-1 antibody. *P Ͻ 0.05. PGHS nitration occurs, we obtained human atherosclerotic tissue from several endarterectomy procedures and purified it prior to immunoblotting with both 3-nitrotyrosine and PGHS antibodies. Our results show that a 70 kDa protein is indeed nitrated, which may be PGHS-1 or PGHS-2. Our results also demonstrate that inducible PGHS (PGHS-2) expression is augmented in human atherosclerotic lesions (9). Immunoprecipitation results showed conclusively that PGHS-1 is nitrated in atherosclerotic tissue. PGHS-2 was not detected by immunoprecipitation, and was only minimally detected by Western blotting. Thus, PGHS-2 nitration in atheromatous lesions cannot be ruled out.
While ONOO Ϫ has previously been invoked as the initial oxidant involved in the pathogenesis of atherosclerosis (45), other mechanisms may be involved. For instance, peroxidase enzymes (e.g., myeloperoxidase) can catalyze Tyr nitration in proteins when utilizing nitrite (NO 2 Ϫ ) and hydrogen peroxide (H 2 O 2 ) as substrates (46). In addition, the combination of hypochlorous acid (HOCl) and NO 2 Ϫ , which also yields 3-chlorotyrosine (47), can lead to nitrotyrosine formation in proteins (48). Thus, in atherosclerosis the nature of the species causing nitration is not known, but could involve ONOO Ϫ .
Under atherosclerotic conditions, both NO • and superoxide production are elevated (49) providing favorable conditions for ONOO Ϫ formation. While NADPH oxidation may provide a source of superoxide (50), recent data suggest that iNOS itself generates superoxide anion simultaneously with NO • (51), indicating that iNOS alone is capable of generating ONOO Ϫ (52). Interestingly, iNOS has been found to co-localize with both PGHS-2 and nitrated proteins in macrophages/foam cells (8). This increases the probability that PGHS is affected by NO x in atherosclerotic tissue.
In a biological setting, it is possible that peroxidase enzymes, such as PGHS, may initially play an important role in detoxifying ONOO Ϫ , much in the way that PGHS detoxifies hydroperoxides. The fast reaction of the heme moiety with ONOO Ϫ forms a less harmful NO x product, NO 2 Ϫ . As a consequence, eicosanoids are produced in the presence of arachidonic acid substrate that may or may not be beneficial to restoring vascular reactivity depending on the availability of downstream prostacyclin synthase. Indeed, prostacyclin synthase is nitrated and its activity is impaired in early atherosclerosis (53). While PGHS activity remains unaffected initially, it results in elevated PGH 2 and PGE 2 production. Since reduced prostacyclin synthase activity in arteries prevents the rapid use of PGH 2 , it accumulates and acts as an agonist on the vasoconstrictive thromboxane receptor (53). In the presence of molar excesses of ONOO Ϫ , we propose that nitration of PGHS will occur, leading to loss of function. Tyr dimer formation may also occur during atherosclerosis, but this has yet to be determined. Elucidation of the nitrating species, together with the identification of the relative amounts of PGH 2 and PGI 2 that are formed as the disease progresses, will lead to a better definition of the molecular basis of this vascular disease.