Pro-thrombotic State Induced by Post-translational Modification of Fibrinogen by Reactive Nitrogen Species*

Formation of nitric oxide-derived oxidants has been linked to development of atherosclerosis and associated thrombotic complications. Although systemic levels of protein nitrotyrosine predict risk for coronary artery disease, neither specific proteins targeted for modification nor functional consequences that might contribute to disease pathogenesis have been defined. Here we report a selective increase in circulating levels of nitrated fibrinogen in patients with coronary artery disease. Exposure of fibrinogen to nitrating oxidants, including those produced by the myeloperoxidase-hydrogen peroxide-nitrite system, significantly accelerates clot formation and factor XIII cross-linking, whereas exposure of fibrinogen to non-nitrating oxidants decelerates clot formation. Clots formed with fibrinogen exposed to nitrating oxidants are composed of large bundles made from twisted thin fibrin fibers with increased permeation and a decrease in storage modulus G′ value, suggesting that these clots could be easily deformed by mechanical stresses. In contrast, clots formed with fibrinogen exposed to non-nitrating oxidants showed decreased permeation with normal architecture. Fibrinogen modified by exposure to physiologic nitration systems demonstrated no difference in the rate of plasmin-induced clot lysis, platelet aggregation, or binding. Thus, increased levels of fibrinogen nitration may lead to a pro-thrombotic state via acceleration in formation of fibrin clots. The present results may account, in part, for the association between nitrative stress and risk for coronary artery disease.

Formation of nitric oxide-derived oxidants has been linked to development of atherosclerosis and associated thrombotic complications. Although systemic levels of protein nitrotyrosine predict risk for coronary artery disease, neither specific proteins targeted for modification nor functional consequences that might contribute to disease pathogenesis have been defined. Here we report a selective increase in circulating levels of nitrated fibrinogen in patients with coronary artery disease. Exposure of fibrinogen to nitrating oxidants, including those produced by the myeloperoxidase-hydrogen peroxide-nitrite system, significantly accelerates clot formation and factor XIII cross-linking, whereas exposure of fibrinogen to non-nitrating oxidants decelerates clot formation. Clots formed with fibrinogen exposed to nitrating oxidants are composed of large bundles made from twisted thin fibrin fibers with increased permeation and a decrease in storage modulus G value, suggesting that these clots could be easily deformed by mechanical stresses. In contrast, clots formed with fibrinogen exposed to non-nitrating oxidants showed decreased permeation with normal architecture. Fibrinogen modified by exposure to physiologic nitration systems demonstrated no difference in the rate of plasmin-induced clot lysis, platelet aggregation, or binding. Thus, increased levels of fibrinogen nitration may lead to a pro-thrombotic state via acceleration in formation of fibrin clots. The present results may account, in part, for the association between nitrative stress and risk for coronary artery disease.
Epidemiological studies have indicated that increased levels of circulating fibrinogen is an independent predictor of coronary heart disease and in some cases of premature death from cardiovascular disease, although a causative relationship between high levels of fibrinogen and cardiovascular disease has not been firmly established (1)(2)(3)(4). Fibrinogen is a multifunctional protein essential for hemostasis. It is a 340-kDa glycoprotein, consisting of three non-identical peptide chains A␣, B␤, and ␥, which are linked together by 29 disulfide bonds (5). During coagulation, the soluble fibrinogen is converted to insoluble fibrin polymers. The process is initiated by thrombin, a serine protease, which catalyzes the cleavage first of two fibrinopeptides from the amino termini of the A␣ chains and then two fibrinopeptides from the amino termini B␤ chains. Upon release of the fibrinopeptides, the remaining fibrin monomers aggregate spontaneously to form ordered fibrin polymers (5). The clot is stabilized by the formation of covalent bonds introduced by the action of a transglutaminase, factor XIII (6). Under physiological conditions, fibrinolysis is dependent on the binding of circulating plasminogen and tissue-type plasminogen activator (tPA) 1 to fibrin clots. Urokinase and tPA convert plasminogen to the active protease plasmin, which then cleaves fibrin polymers to soluble fragments completing the coagulation and clot resolution cycle.
A major cause of vascular injury leading to the development of atherosclerosis is oxidative stress (7)(8)(9). Proteins are major targets for reactive species, and nitration of tyrosine residues is a selective protein modification induced by reactive nitrogen species in human disorders as well as animal and cellular models of disease (10,11). Nitrated proteins have been detected in atherosclerotic lesions (12)(13)(14)(15)(16), and recent studies (17) show that the levels of nitrated plasma proteins independently predict risk for coronary artery disease. In this study we identified and quantified nitrated fibrinogen in the plasma of patients with documented coronary artery disease. In a series of experiments the effects of nitration on the kinetics of fibrin formation, factor XIII cross-linking, fibrin architecture and rheology, platelet aggregation and binding, and lysis by plasmin were determined and contrasted with oxidized and control fibrinogen.

MATERIALS AND METHODS
Human Studies-Sequential patients presenting to the Cardiology Section of the Cleveland Clinic Foundation or responding to local advertisements were enrolled. To be classified as having coronary artery disease, patients had to have a documented history of myocardial infarction, coronary artery bypass graft surgery, percutaneous coronary intervention, or a stenosis of 50% or greater in one or more major coronary vessels demonstrated by coronary angiography. To be considered a control, subjects had to have no clinical history of coronary artery disease, no known peripheral artery disease, or history of symptoms suggestive of angina pectoris or congestive heart failure. All patients gave written, informed consent, and the Institutional Review Board of the Cleveland Clinic Foundation approved the study protocol.
Immunoprecipitation-Human plasma (225 g) was pre-cleared with protein G-Sepharose fast flow beads (Amersham Biosciences) in lysis buffer (20 mM Tris, 150 mM NaCl, 10% glycerol, 1.0% Triton X-100, 4.0 mM EGTA) containing the protease inhibitor mixture (Sigma) for 1 h at 4°C in order to remove the nonspecific proteins. The mixture was centrifuged for 3 min at 13,000 rpm, and 40 g of monoclonal antinitrotyrosine antibody (18) was added to the supernatant and incubated overnight at 4°C while rotating. The samples were then incubated at 4°C for 1 h with protein G-Sepharose while rotating and then centrifuged at 13,000 rpm for 3 min. The beads were washed three times for 2 min each with freshly prepared lysis buffer containing the protease inhibitor mixture. The samples were pelleted by centrifugation for 3 min at 13,000 rpm, and 30 l of 2ϫ sample buffer containing SDS and 2-mercaptoethanol was added to the pellet. The beads were boiled for 10 min and centrifuged at 13,000 rpm for 3 min, and the supernatant was then analyzed by SDS-electrophoresis. The proteins were separated on 6% SDS-PAGE and transferred to nitrocellulose paper (Schleicher & Schuell). The blot was blocked with dry milk solution and then incubated with rabbit anti-human fibrinogen-horseradish peroxidase-conjugated antibody (Dako, Carpinteria, CA). After antibody incubation, the blot was developed with ECL Western blotting detection reagent (Amersham Biosciences).
Affinity Purification of Fibrinogen from Human Subjects Plasma-An ImmunoPure Protein A IgG Orientation Kit (Pierce) was utilized to affinity-purify fibrinogen from coronary artery disease and age-matched control patient plasma. Briefly, rabbit anti-human fibrinogen antibody (Dako Corp., Carpinteria, CA) was bound to protein A and cross-linked with dimethylpimelimidate. For sample application, the patient plasma was diluted in 10 mM Tris, pH 7.5, and applied to the column. Unbound protein fractions were eluted with 10 mM Tris, pH 7.5. Fractions of bound fibrinogen were eluted with 100 mM glycine-HCl, pH 2.8, and collected in tubes containing 1 M Tris-HCl, pH 8.0. All fractions were dried down to a small volume using a Savant Instrument SpeedVac Concentrator (Savant Instruments Inc., Holbrook, New York). Protein in the fractions was monitored by absorbance at 280 nm. All fractions containing antigen or unbound protein were pooled together, respectively, and analyzed by SDS-PAGE as described above. Nitrotyrosine levels were determined by HPLC with on-line electrospray ionization/tandem mass spectrometry using stable isotope dilution methodology and an ion trap mass spectrometer, as described previously (17,19).
Nitration and Oxidation of Fibrinogen-Freeze-dried human fibrinogen (American Diagnostica, Inc., Greenwich, CT), which was free of plasminogen, factor XIII, and fibronectin, with coagulability greater than 95%, was solubilized in filtered de-ionized water. The fibrinogen solution was then eluted through a PD-10 desalting column (Amersham Biosciences). Stock solutions of fibrinogen solutions were divided equally into two tubes, and an equal volume of buffer (100 mM potassium phosphate buffer, 50 mM sodium bicarbonate, pH 7.4) was added to each tube. Fibrinogen was exposed to 60 nM myeloperoxidase (MPO) plus 100 M H 2 O 2 in the absence or presence of 100 M nitrite for 1 h. In the absence of nitrite and in physiological concentrations of chloride, MPO catalyzes the formation of hypochlorous acid (HOCl), a strong oxidant and chlorinating agent. In the presence of nitrite in addition to HOCl, MPO also catalyzes the formation of nitrogen dioxide, an oxidant and nitrating agent (19 -21). As a control fibrinogen was also exposed to 100 M reagent HOCl. Fibrinogen was also exposed to 3-morpholinosydnonimine, HCl (SIN-1) (Calbiochem-Novabiochem). SIN-1 is dissolved in 50 mM potassium phosphate buffer, pH 5.0, and purged with nitrogen gas before and after each use. SIN-1 undergoes base hydrolysis to release nitric oxide and superoxide in solution (for 100 M SIN-1 the rate is ϳ1 M per min), which they react with a rate constant of 10 10 M Ϫ1 s Ϫ1 to form peroxynitrite. The samples were incubated at 37°C for 1 h, which is approximately twice the half-life of SIN-1, and vortexed every 10 min to preserve the dissolved oxygen levels in the mixture. As a control, fibrinogen was exposed to SIN-1 in the presence of Cu,Zn superoxide dismutase (Calbiochem-Novabiochem) at 0.3 mg/ml to compete nitric oxide for superoxide. Under these conditions, fibrinogen is exposed to nitric oxide plus H 2 O 2 .
After incubation with the different chemical systems, the samples were applied to a PD-10 desalting column and were eluted with TBS (50 mM Tris base, 150 mM sodium chloride, pH 7.4). Following the different exposures, the nitration of fibrinogen was confirmed by Western blotting with anti-nitrotyrosine antibodies and by liquid chromatography/ electrospray ionization/mass spectrometry analysis (Table I), whereas oxidation was assessed by the formation of carbonyls (Table I) as described previously (22). Western blotting confirmed that exposure to SIN-1 and MPO/H 2 O 2 /NO 2 Ϫ but not in the other controls results in nitration of tyrosine residue(s) in all three fibrinogen chains. The nitration of fibrinogen did not result in any significant change in fibrinogen structure as determined by the CD spectrum of the control and nitrated fibrinogen (data not shown). Under our experimental conditions, exposure of fibrinogen to either SIN-1 or MPO/H 2 O 2 /NO 2 Ϫ did not result in covalent protein cross-linking.
Thrombin-catalyzed Fibrin Clot Polymerization-Polymerization was initiated by the addition of 0.1 unit/ml of human thrombin (American Diagnostica Inc., Greenwich, CT) to fibrinogen (1 mg/ml) in TBS, pH 7.4 (50 mM Tris base, 150 mM sodium chloride). Clot formation was monitored at 25°C as the increase in absorbance at 350 nm over time using a Hewlett-Packard diode array spectrophotometer. Clot formation in human plasma samples was initiated by simultaneous addition of CaCl 2 (12 mM final concentration) and rabbit brain thromboplastin (Plastinex® from Bio/Data Co.). The concentration of thromboplastin in plasma was 1.4 mM relative to stock. Turbidity was measured in glass cells with a 1-mm optical path length. The formation of fibrinopeptides A and B after addition of thrombin was followed over time by HPLC separation and UV detection. factor XIII fibrinogen and fibrin crosslinking was performed as described previously (23).
Scanning Electron Microscopy-Fibrin clots made at room temperature using 0.1 unit/ml thrombin were processed for scanning electron microscopy as described previously (23). All samples were prepared at least in duplicates, and for all clots at least three high resolution images from different regions were obtained.
Permeation and Viscoelastic Measurements-The Darcy constant, which represents the surface of the gel allowing flow through a network, and thus provides information on the pore structure and fiber diameter, was calculated from the flow measurements, pressure, and geometric parameters of the clot as described in detail previously (24). Viscoelastic measurements were performed in clots prepared between 12-mm diameter glass coverslips in a Plazek torsion pendulum described above for permeation studies (24). Each clot was measured three times, and three clots were prepared for each sample on 2 different days.
Lysis of Fibrin Clot-Clot lysis by 6 nM of recombinant tPA (American Diagnostica, Greenwich, CT) was monitored as described (25,26). The percent lysis was calculated after adjusting for value of the starting absorbance. The plasmin protection assay and the plasmin digestion of fibrin clots were performed as described previously (25). The previous published protocols for the preparation and lysis of 125 I-fibrin microparticles was utilized as described in detail previously (27,28).
Platelet Aggregation and Adhesion-Aggregation of gel-filtered platelets was measured on a Dual Channel Aggregometer (Chrono-log Corp., Havertown, PA) in the presence of 150 g/ml fibrinogen, 0.5 mM CaCl 2 , and different concentrations of ADP varying from 2 to 20 M (34). For platelet adherence to control and nitrated fibrinogen Immulon 2HB microtiter plates were employed, and adherence was measured by the phosphatase assay (29) utilizing a universal microplate reader (Bio-Tek Instruments, Inc., Winooski, VT).

Detection of Nitrated Fibrinogen in Patients with Coronary
Artery Disease-Plasma of coronary artery disease patients was immunoprecipitated with a monoclonal anti-nitrotyrosine antibody. Probing the immunoprecipitated proteins with a polyclonal anti-fibrinogen antibody revealed that a fraction of fibrinogen was nitrated in coronary artery disease patient plasma (Fig. 1A). In order to confirm and quantify the presence of nitrated fibrinogen in coronary artery disease and agematched control patient plasma, fibrinogen was purified using affinity chromatography in which the Fc portion of a polyclonal anti-fibrinogen antibody was coupled to protein A. The ability of the anti-fibrinogen column to capture nitrated fibrinogen was confirmed by using in vitro nitrated fibrinogen (not shown). The affinity column was efficient in capturing all the fibrinogen for the input plasma (Fig. 1A). The eluted fibrinogen was then hydrolyzed in order to quantify the protein 3-nitrotyrosine levels by HPLC with on-line electrospray ionization/ tandem mass spectrometry using stable isotope dilution methodology and an ion trap mass spectrometer (17,19). The values were normalized to the levels of tyrosine to avoid changes in fibrinogen levels among patients. A 30% increase (p Ͻ 0.001) in the levels of nitrated fibrinogen was found in coronary artery disease patients (n ϭ 30) as compared with age-matched controls (n ϭ 26) (Fig. 1C).
Effect of Fibrinogen Nitration on Fibrin Clot Formation-The polymerization of fibrin catalyzed by thrombin monitored by changes in turbidity at 350 nm in human plasma after the addition of 12 mM calcium from coronary artery disease patients and controls is depicted in Fig. 2A. The thrombin-induced polymerization of fibrinogen recovered from patients with coronary artery disease showed a shorter lag phase, a rapid rise in the initial velocity, and increased final turbidity as compared with fibrinogen recovered from plasma of controls. The changes in fibrinogen polymerization were not due to fibrinogen concentration in plasma since the mole content of tyrosine in the purified fibrinogen for these samples was 33.6, 25.4, and 24.3 for the coronary disease patients samples (curves 1-3 in Fig. 2A) and 35.7 and 35.1 (curves 4 and 5, respectively) for the control plasma. However, the nitrotyrosine content in these samples (expressed as mol of 3-nitrotyrosine/mol tyrosine) varied: 67, 38.4, and 52 for the coronary artery disease patients and 25.2 and 25.4, for the controls, respectively.
To investigate the potential effect of fibrinogen nitration on turbidity changes and other properties of the protein, nitrated and oxidized fibrinogens were prepared by chemical treatments described in detail under "Materials and Methods." The yield of nitration after exposure to either 60 nM MPO plus 100 M H 2 O 2 and 100 M NO 2 Ϫ or 100 M SIN-1, was 65 Ϯ 8 and 46 Ϯ 4 mol 3-nitrotyrosine/mol tyrosine (n ϭ 3-5), respectively, as as-sessed by liquid chromatography/electrospray ionization/tandem mass spectrometry (17,19). These values were in close proximity to more than half of the coronary disease patient values that exceeded the mean value of 38 Ϯ 14 mol of 3-nitrotyrosine/mol tyrosine. Data in Fig. 2B reveal that the polymerization of fibrin catalyzed by thrombin monitored by changes in turbidity at 350 nm show a rapid rise in the sample of fibrinogen treated with MPO/H 2 O 2 /NO 2 Ϫ as compared with the protein oxidized by MPO and H 2 O 2 . The same increase in turbidity was also observed with fibrinogen exposed to nitration conditions using SIN-1 in the presence of CO 2 as compared with control or fibrinogen exposed to SIN-1 in the presence of superoxide dismutase (Fig. 2C). In these experiments similar to the observations in coronary artery disease plasma, the lag phase in fibrin polymerization decreased, the maximum rate of turbidity for the first 500 s after the lag phase increased (Table  I), and the final turbidity was greater than in the control curves, suggesting alterations of clot structure. Moreover, the maximum rate of turbidity for the first 500 s after the lag phase corrected for the final turbidity increased as a function of the magnitude of fibrinogen nitration (inset, Fig. 2B). The increase in the maximum rate of turbidity of fibrinogen exposed to nitration conditions was sustained even in the presence of oxidized fibrinogen. Fibrinogen that was exposed first to SIN-1 and further oxidized by addition of 100 M HOCl or fibrinogen that was exposed first to 100 M HOCl and then to SIN-1 showed the same increase in maximum rate of turbidity as fibrinogen exposed only to SIN-1 (Fig. 3D). In contrast fibrinogen exposed to either MPO plus H 2 O 2 or nitric oxide plus H 2 O 2 or reagent HOCl failed to show increases in the maximum rate of turbidity (Fig. 2).
The significant increase in the maximum rate of turbidity was not due to acceleration in thrombin cleavage, since quantification of fibrinopeptide A and fibrinopeptide B released by thrombin proteolysis was the same between control and nitrated fibrinogen (Fig. 3).
Factor XIII Cross-linking-Fibrinogen cross-linking factor XIII cross-links both fibrinogen and fibrin. Fibrinogen crosslinking by factor XIII is not different between control and nitrated fibrinogen (not shown). However, factor XIII crosslinking of fibrin was accelerated in the nitrated fibrinogen than in control as evident by the disappearance of the A␣ and ␥ chains of fibrinogen (Fig. 4).
Effect of Nitration on Fibrin Clot Architecture, Permeation, and Viscoelastic Properties-The architecture of the fibrin clots formed by control, nitrated, and oxidized fibrinogens was examined by scanning electron microscopy as described previously (23). Fibrin clots formed by fibrinogens exposed to nitrating agents are strikingly different from control or oxidized fibrinogens under identical experimental conditions (Fig. 5). The clots made from fibrinogens exposed to either MPO/H 2 O 2 / NO 2 Ϫ or SIN-1 are made up of thinner fibers, even though the higher turbidity suggested that the fibers would be thicker. However, the fibrin clot made by nitrated fibrinogens is composed of large bundles of the thin fibrin fibers, which scatter light like thick fibers, accounting for the greater turbidity. Higher magnification images also reveal the existence of twisted fibers in nitrated fibrin. A number of large pores are evident in fibrin clots made of nitrated fibrinogens in contrast to the more dense and uniform fibrin network of the oxidized fibrinogen (Fig. 5).
These changes in the architecture of the fibrin clots would be expected to have certain consequences on the physical properties of the clot. Therefore, quantitative changes in the permeation coefficient or Darcy constant were determined by flow measurements made under a constant pressure applied and accounting for geometric parameters of the clot (24). The permeation coefficient of clots made from fibrinogens exposed to either MPO/H 2 O 2 /NO 2 Ϫ or SIN-1 was significantly higher than the control (Table II). This increase in permeation is accounted for by the presence of large pores, as seen by scanning electron microscopy. A decrease in the Darcy constant was noted in the fibrin clots made of oxidized fibrinogens (Table II), since the pore size in these clots was decreased as compared with control.
Clot rigidity was then evaluated by measurement of the viscoelastic properties of the fibrin clots made from nitrated or oxidized fibrinogen. Both fibrinogens exposed to nitrating and oxidized conditions produced clots that were considerably less stiff than control reflected by the decrease in storage modulus GЈ value that is directly related to the stiffness of the clot (Table II). There were no apparent changes in the loss modulus GЉ or inelastic deformation, but the ratio of inelastic to elastic deformation (tan␦) was considerably greater for clots made from both fibrinogens exposed to nitrating and oxidizing conditions (Table II).
Plasmin Protection Assay and Clot Lysis-Incubation of fibrinogen with plasmin in the presence of EDTA results in the digestion of fibrinogen to D1, D2, D3, and E fibrinogen fragments (25,26). No difference in the plasmin-induced fragmentation between control and fibrinogen exposed to nitrating conditions was observed (Fig. 6A). Incubation of fibrinogen with 5 mM CaCl 2 retards equally the plasmin-induced digestion in Ϫ (tracing 1) and fibrinogen exposed to 60 nM MPO plus 100 M H 2 O 2 (tracing 2). Inset, linear relationship between levels of in vitro nitration of fibrinogen and ratio of V 0 /A maximum. C, turbidity changes for fibrinogen exposed to 100 M SIN-1 (tracing 1), 100 M SIN-1 in the presence of superoxide dismutase (tracing 2), and control fibrinogen (tracing 3). D, changes in fibrinogen turbidity after addition of thrombin to fibrinogen that was exposed first to 100 M SIN-1 and then oxidized by the addition of 100 M HOCl (tracing 1) and to fibrinogen that was first oxidized by 100 M HOCl and then exposed SIN-1 (tracing 2). Tracing 3, control fibrinogen; tracing 4, fibrinogen exposed to 100 M HOCl. Extensive oxidation of fibrinogen by the addition of more than 100 M HOCl results in extensive protein cross-linking and renders the fibrinogen incapable of forming clots.

TABLE I
Rate of clot formation of control fibrinogen and fibrinogens exposed to nitrating and oxidizing conditions The rate of clot formation (V 0 ) starting 500 s after the initiation of the reaction by the addition of thrombin was measured by turbidity changes at 350 nm (n ϭ 3-5). Following the different exposures, the nitration of fibrinogen was measured by liquid chromatography/electrospray ionization/mass spectrometry analysis, whereas oxidation was assessed by the formation of carbonyls as described previously (25). ND, not detected; SOD, superoxide dismutase. both the control and nitrated fibrinogen. Moreover, plasmin digestions of fibrin clots made from control or fibrin made from fibrinogen exposed to nitrating conditions were cleaved to the same products yielding the predicted YY, DY, DD, and E fragments (Fig. 6B). The lysis rate between the control and fibrin clots made from fibrinogen exposed to nitrating conditions in vitro was the same after normalizing the data to the same starting turbidity and expressing it as percent lysis over time.
The value for control clots was 0.10 Ϯ 0.03 ϫ 10 Ϫ2 s Ϫ1 (n ϭ 5) and 0.12 Ϯ 0.02 ϫ 10 Ϫ2 s Ϫ1 (n ϭ 5) for clots made from fibrinogen exposed to nitrating agents, suggesting that the alterations in the physical properties of the fibrin clots induced by nitration do not significantly impact the proteolytic cleavage of the clots. The in vitro data were further validated in vivo by measuring the rate of clearance of microemboli made from control and nitrated fibrinogens after injection of radiolabeled microemboli ( 125 I-ME) with a mean diameter 2-5 m in the tail vein of anesthetized mice (Fig. 6). As described previously, 125 I-ME rapidly lodge in the pulmonary vasculature after in-travenous injection in mice and rats alike and, nearly 30 min after pulmonary deposition, undergo spontaneous dissolution that is completed within 5 h (25,26). Experiments in mice with genetically altered background revealed that both endogenous tissue type and urokinase plasminogen activator contribute to spontaneous dissolution of 125 I-ME lodged in the pulmonary vessels. Injection of exogenous plasminogen activators (e.g. activase and tPA) markedly accelerates dissolution of 125 I-ME deposited in rat and murine lungs and compensates for genetic ablation of the corresponding plasminogen activator (27,28). There was no significant difference in pulmonary deposition and rate of spontaneous dissolution of 125 I-ME prepared from intact versus nitrated fibrinogen in mice (Fig. 5C). However, 125 I-ME prepared from nitrated fibrinogen displayed markedly higher susceptibility to "therapeutic" fibrinolysis (inset in Fig.  6C). Thus, intravenous injection of a marginally effective dose of tPA (30 g/kg), which had no effect on the rate of dissolution of normal 125 I-ME, induced a detectable acceleration of dissolution of the nitrated 125 I-ME (p Ͻ 0.05). These data imply that fibrinogen nitration does not alter susceptibility of blood clots to spontaneous fibrinolysis, yet may be more amiable to fibrinolytic therapy.
Platelet Aggregation and Binding-The effects of control and nitrated fibrinogens upon the rate of human platelet aggregation and platelet binding were examined utilizing protocols established previously (29). The rate of ADP-induced platelet aggregation was the same for control and fibrinogens exposed to 100 M SIN-1 (data not shown). Similarly the adherence of platelets to immobilized fibrinogen on multiwell plates was the same for control and nitrated fibrinogens (data not shown). DISCUSSION A number of studies have provided evidence for a strong association between the pro-thrombotic state and risk for adverse outcomes such as myocardial infarction, sudden death, and stroke in coronary artery disease patients (1-4). Induction of a pro-thrombotic state has been associated with increased circulating levels of fibrinogen and other hemostatic proteins (1)(2)(3)(4). Recently, it was shown that the plasma levels of nitrotyrosine, a protein marker of nitric oxide-derived oxidants, are enhanced in the plasma of coronary artery disease patients and independently predict cardiovascular risks and atherosclerotic burden (17). The present studies provide a possible biochemical link between thrombosis and inflammation by showing both increased levels of nitrated fibrinogen in coronary artery dis- were mixed with thrombin and factor XIII, and after 60 min of incubation fibrinogen was separated and blotted with an anti-fibrinogen antibody. Lanes 1 and 3 are control and exposed fibrinogen, respectively, immediately after addition of thrombin and factor XIII. Lanes 2 and 4 are the corresponding fibrinogen 60 min after the addition of thrombin and factor XIII. The band intensity of the A␣, B␤, and ␥ chains of fibrinogen were quantified by densitometry. Values represent means Ϯ S.D. for three independent determinations. *, p Ͻ 0.05 between the control and exposed fibrinogen after analysis of variance using Tukey's post hoc test. ease subjects and functional alterations in fibrinogen/fibrin consistent with generation of a pro-thrombotic state. The present studies thus suggest a previously unrecognized link between enhanced nitrative stress and potential for pro-thrombotic state in patients with coronary artery disease.
Previous studies (18,30) have identified nitrated fibrinogen in patients with clinically documented acute respiratory distress syndrome (ARDS) and lung cancer. Fibrin deposits are abundant in the lungs of patients with ARDS, a common complication of hemorrhagic injury and sepsis, and likely the result of abnormal clotting rather than failure in the fibrinolytic pathways, which apparently function normally in ARDS patients (18). The findings in the human ARDS are in part con-firmed by proteomic identification of nitrated fibrinogen in the lungs of rats exposed to lipopolysaccharide (31). The data presented here in coronary artery disease patients and previously in ARDS and cancer patients indicate that fibrinogen is a target for modification by reactive nitrogen species in vivo, consistent with the potential contribution of these processes in the pathogenesis of atherosclerosis and acute lung injury.
It is rather difficult to attribute the changes in fibrinogen properties exclusively to tyrosine nitration because nitration of proteins quite often occurs in conjunction with oxidation of other amino acid residues. For example, fibrinogen immunoprecipitated from the plasma of cancer patients demonstrated increased content of carbonyl adducts, suggesting that fibrinogen was either oxidized or modified by the addition of aldehydic oxidized lipid species (30). Indeed, oxidation of fibrinogen by MPO ϩ H 2 O 2 or HOCl resulted in the formation of 1.54 mol of carbonyl adduct per mol of protein and induced a decrease in the rate of fibrin polymerization producing fibrin clots with decreased permeation properties. This is consistent with previous findings that indicated that formation of 1.8 mol of carbonyl adduct per mol of protein in fibrinogen (22) or oxidation of histidine, tryptophan, and methionine residues resulted in the inhibition of thrombin-induced polymerization of fibrinogen despite a normal rate of fibrinopeptide release (32)(33)(34)(35). In contrast, exposure of fibrinogen to nitrating conditions such as MPO/H 2 O 2 /NO 2 Ϫ , which resulted in both oxidation and nitration of the protein (Table I), resulted in a significant increase in the maximum turbidity, generating fibrin clots with distinct architecture and increased permeation as compared with oxidized fibrinogen (Fig. 5). It is interesting to note that changes in the properties of fibrinogen and fibrin clots observed after treatment of the protein with nitrating agents occurred in the absence of dityrosine cross-linking and without alterations in the secondary structure of fibrinogen, as assessed by CD spectroscopy (data not shown). Fibrinogen exposed to nitrating agents was able to maintain the ability to aggregate and adhere to platelets. These findings are consistent with a previous report showing that oxidation of tyrosine residues to form dityrosine, as well as conformational changes detected by 8-anilino-1 naphthalenesulfonic acid binding in fibrinogen, resulted in a significant decline in platelet adhesion and ADP-stimulated platelet aggregation (34). Taken together the dramatically opposing effect of nitrating compared with nonnitrating oxidants strongly suggests that nitration of fibrinogen tyrosyl residues is responsible for the alterations observed in selective fibrinogen/fibrin properties and functions.
The decrease in the lag period of turbidity measurements in nitrated fibrinogen (Fig. 2) indicates an increase in the rate of small oligomer formation, whereas the dramatic increase in the maximal rate of turbidity indicates a significant augmentation in the rate of lateral aggregation. Furthermore, the nature of the effects on clot structure, particularly the presence of many fiber ends and large pores in the clots, suggests that nitrated  6. A, plasmin protection assay. Control and exposed fibrinogen (lanes 1 and 4) were incubated with plasmin (lanes 2 and 3 and 5 and 6) in the presence of EDTA (lanes 2 and 5) or CaCl 2 (lanes 3 and 6). Proteins were separated on a 7% SDS-PAGE and blotted with an anti-fibrinogen antibody. B, products of fibrin degradation. Control (lane 1) and fibrin clots made from fibrinogen exposed to 60 nM MPO plus 100 M H 2 O 2 and 100 M NO 2 Ϫ (lane 2) were digested for 1 h, and the digested materials indicating the location of different fragments were separated on 4 -12% gradient gels stained with colloidal blue. C, kinetics of microemboli dissolution in mice following injection of 125 Ifibrin microparticles. Inset, dissolution of microparticles after administration of tPA (values indicate the number of particles in the lung at initiation and 60 min after addition of tPA (white bars) or saline (hatched bars)). Values represent mean Ϯ S.D. of n ϭ 3-5 independent determinations. molecules may "cap" the ends of growing protofibrils, preventing or retarding further growth. Kinetic modeling studies of changes in the rate constants resulting from capping that we have carried out are consistent with the observed effects on the polymerization curves (23). In addition, such effects could be present with only a few modified molecules as is the case with nitrated fibrinogen.
The viscoelastic properties of fibrin clots have been associated with complications of coronary heart disease (36 -39). Higher fibrinogen levels have been associated with formation of a less deformable fibrin clot (clot with increased stiffness), which is more likely to occlude blood flow than a more deformable clot (36). Two studies by Fatah et al. (38,39) described the in vitro formation of fibrin gels, made from patient plasma with documented heart disease, with abnormal gel network characterized by less porous and resilient and more space-filling structures. The biochemical reasons for the formation of these abnormal fibrin gels is not known but was not related to fibrinogen concentration and was postulated to result from posttranslational modifications such as deglycosylation or other unidentified alterations of fibrinogen (40). The architecture of these fibrin gels is different from the clots obtained from nitrated fibrinogen. The subjects of the aforementioned studies had experienced adverse effects such as myocardial ischemia before the age of 45, and it is possible that in these individuals the majority of the nitrated fibrinogen has been deposited in the lesions (41,42). Moreover, the viscoelastic properties of fibrin made from fibrinogen exposed to nitration conditions are consistent with the observations of clot structure determined by scanning electron microscopy. Fibrin clots made from nitrated fibrinogen are formed primarily of thin fibers that give rise to high turbidity because of their bundling into groups. The decrease in the storage modulus (GЈ) or stiffness of the clots made from nitrated fibrinogen is consistent with the observed structure, since these clots are composed of thinner fibers with many free fiber ends. Clots made up of thin fibers that have many branch points are generally stiffer than clots with thicker fibers and fewer branch points. However, clots with bundled thin fibers with many free ends would behave more like clots with very thick fibers. The free fiber ends allow fibers to move past each other, giving rise to an increase in the tan␦ or ratio of non-elastic to elastic components. The mechanical properties of these clots suggest a connection to pathological conditions. The weaker clots made from nitrated fibrinogen may fragment more easily upon mechanical stresses and thus increase the potential risk for microemboli formation. These observations are similar to reported effects of tyrosine acetylation (43). Acetylation of two tyrosine residues in fibrinogen produced a fibrin clot with a 50% reduction in clot strength (43).
Collectively, the present data indicates that tyrosine nitration selectively alters fibrinogen function, promoting clot acceleration, specific alterations in clot structure, and viscoelastic properties. Although all of the in vivo implications of these effects are not completely understood, the increased rate of polymerization, leading to an imbalance in the dynamic equilibrium between clotting and lysis and the risk of clot fragmentation, may account for the association between nitration of fibrinogen and the incidence of adverse effects in coronary artery disease.