Inducible nitric oxide synthase provides protection against injury-induced thrombosis in female mice

Abstract

Nitric oxide (NO) is an important vasoactive molecule produced by three NO synthase (NOS) enzymes: neuronal (nNOS), inducible (iNOS), and endothelial NOS (eNOS). While eNOS contributes to blood vessel dilation that protects against the development of hypertension, iNOS has been primarily implicated as a disease-promoting isoform during atherogenesis. Despite this, iNOS may play a physiological role via the modulation of cyclooxygenase and thromboregulatory eicosanoid production. Herein, we examined the role of iNOS in a murine model of thrombosis. Blood flow was measured in carotid arteries of male and female wild-type (WT) and iNOS-deficient mice following ferric chloride-induced thrombosis. Female WT mice were more resistant to thrombotic occlusion than male counterparts but became more susceptible upon iNOS deletion. In contrast, male mice (with and without iNOS deletion) were equally susceptible to thrombosis. Deletion of iNOS was not associated with a change in the balance of thromboxane A₂ (TxA₂) or antithrombotic prostacyclin (PGI₂). Compared with male counterparts, female WT mice exhibited increased urinary nitrite and nitrate levels and enhanced ex vivo induction of iNOS in hearts and aortas. Our findings suggest that iNOS-derived NO in female WT mice may attenuate the effects of vascular injury. Thus, although iNOS is detrimental during atherogenesis, physiological iNOS levels may contribute to providing protection against thrombotic occlusion, a phenomenon that may be enhanced in female mice.

thrombotic events affect 20 million people worldwide each year (44). To date, certain risk factors and acquired or inherited states have been uncovered that lead to an altered hemostatic equilibrium or vascular injury (9). Interestingly, men and women are not equally susceptible to thrombosis. For instance, the risk of stroke in women under age 55 is lower than in men (3). While the protective effect in females may be ascribed to reproductive hormones, the use of estrogen in preventing or treating vascular injury is controversial since oral contraceptives are associated with an increased risk of thrombosis (35). Thus a knowledge gap persists concerning the mechanisms underlying sex differences in thrombosis (7, 26).

The study of thrombosis has been greatly enhanced by the availability of in vivo models coupled with modern intravital techniques (33). An experimental model of arterial damage can be induced by chemical injury using ferric chloride (FeCl₃; Ref. 18). The FeCl₃ model represents a mechanical injury model that could be analogous to catheter-related injury, since FeCl₃ causes denudation of the endothelial cells (43). However, FeCl₃ is an oxidizing agent and chemical injury involving iron and oxygen species can also occur. While the exact mechanism of thrombus formation is not fully elucidated (33), ferric oxide aggregates have been found near the developing thrombus implicating the transmigration of iron particles (37). The morphology of the resultant thrombi following FeCl₃ injury, however, is similar to that found in human acute coronary syndromes (8, 18). When coupled with knockout mice, this model is important in characterizing components of the coagulation system and in the development of effective inhibitors (16). Thus the FeCl₃-induced vascular injury model represents a reproducible method that can increase our understanding of the complex interplay of components in thrombus formation (43).

Nitric oxide (NO) is produced by a family of three NO synthase (NOS) enzymes and is an important regulator of vasomotor tone and platelet aggregation. Indeed, a deficiency in NO is associated with an increased potential for thrombosis. The endothelial (eNOS) and neuronal (nNOS) isoforms are constitutively expressed, whereas the inducible form (iNOS) is stimulated by proinflammatory cytokines (2, 22). Unlike eNOS and nNOS, iNOS produces micromolar amounts of NO that are beneficial in destroying invading pathogens (12, 42) but deleterious in atherogenesis (5, 17), causing oxidative and nitrative stress (38, 39). Furthermore, iNOS may modulate vasoregulatory pathways by its interaction with the cyclooxygenase (COX) pathway, which gives rise to prostanoids that are both vasodilatory and vasoconstrictive (19). In this regard, iNOS deletion alters levels of COX-derived urinary and serum prostanoids (24). The availability of iNOS-derived NO may be linked with the female hormone estrogen. Although clinical results using estrogen are conflicting (10, 13, 32), estrogen has positive effects in conferring atheroprotection in female mice (6) and in preventing the vascular injury response (14, 31). In fact, estrogen has been shown to attenuate vasoconstriction by a pathway involving iNOS (48).

Herein, the FeCl₃-induced injury model was used to explore the hypothesis that male and female wild-type (WT) mice are not equally susceptible to thrombus formation and that this difference may be correlated with iNOS-dependent NO production.

MATERIALS AND METHODS

Reagents.

FeCl₃·6H₂O and citrate-dextrose solution [ACD; containing 22.0 g/l citric acid, trisodium salt, dihydrate, 7.3 g/l citric acid, anhydrous, and 24.5 g/l d-(+)-glucose] were obtained from Sigma-Aldrich. Avertin (2,2,2 tribromoethanol; 97%), from Sigma-Aldrich, was prepared in 2-butanol (125 mg in 125 μl) by vortexing for several minutes in a foil-wrapped tube, followed by further dilution with sterile saline (6 ml). The working solution was refrigerated in a dark sealed bottle and used within a 2-wk period.

Mice.

The animal protocol used in these studies was reviewed and approved by the Weill Cornell Medical College Institutional Animal Care and Use Committee. Male WT and iNOS^−/− mice and female WT and iNOS^−/− mice were used in this study (6 mo in age, fed a regular chow diet). The animals (on a C57BL6/J background) were obtained from Jackson Laboratories and bred to generate subsequent generations in our animal facility. Immediately before FeCl₃-induced carotid artery injury, urine was collected from individual mice. Urine was collected over 24 h, with and without the preservative butylated hydroxytoluene (0.005%), by housing individual mice in metabolic cages. Following FeCl₃-induced injury, blood was drawn for isolation of plasma. Hearts were isolated and weighed from euthanized mice.

FeCl₃ model of carotid artery injury.

FeCl₃ solutions were prepared by dissolving 10% FeCl₃·6H₂O in deionized water. UV/Vis spectra of FeCl₃ (a 1:1,000 dilution of a 10% FeCl₃·6H₂O solution) were recorded using a Perkin Elmer Lambda 20 spectrophotometer at 15 min, 60 min, and 5 days after dissolving in water. Male and female WT and iNOS^−/− mice were anesthetized with avertin (150–200 mg/kg). The left carotid artery was exposed by surgical incision and careful removal of surrounding tissue. Two silk threads, ∼3–4 mm apart, were used to lift the artery onto a miniature ultrasound flow probe (0.5 VB; Transonic Systems, Ithaca, NY). A baseline measurement of blood flow was obtained for 3–5 min using a Transonic T106 flowmeter and WinDaq data acquisition software (DataQ Instruments, Akron, OH). The ultrasound probe was removed, and, with the use of small forceps, a piece of filter paper (Whatman No. 1; 2 × 1 mm) soaked in aged 10% FeCl₃·6H₂O was placed on the carotid artery. After 1 min, the filter paper was removed and the ultrasound probe was once again positioned underneath the artery, and blood flow was monitored for 30 min. Results at different time points are reported as a mean percent change from the initial baseline value (established immediately before FeCl₃ injury) and also as the mean total time in which occlusion was observed over the 30-min period that blood flow was monitored.

Clot lysis studies.

Plasma clotting was examined using a protocol described previously (28). Briefly, whole blood was collected by cardiac puncture using a syringe coated with citrate-dextrose solution (50 μl). The whole blood/citrate mixture (9:1 ratio) was expelled into a tube (after removal of the syringe needle to prevent hemolysis), placed on ice (10 min), and centrifuged (17,000 g at 4°C; 10 min). Plasma aliquots were stored at −80°C. For the clotting and clot lysis study, plasma (15 μl) was mixed with buffer (60 μl; 0.04 M HEPES pH 7, 0.15 M NaCl, and 0.01% Tween 80) and deionized water (30 μl) and allowed to equilibrate in a 96-well plate (room temperature; 5 min). To measure the clotting time, a mixture (15 μl total) containing human thrombin (4 μl; 75 NIH U/ml; American Diagnostic), CaCl₂ (2 μl; 1 M), and deionized water (9 μl) was added to the plasma sample in the 96-well plate. For the measurement of half clot lysis times, a mixture (15 μl total) containing human thrombin (4 μl; 75 NIH U/ml), CaCl₂ (2 μl; 1 M), single-chain recombinant tissue plasminogen activator (1.05 μl; 2.86 μM; Genentech) and deionized water (7.95 μl) was added to plasmin. The absorbance at 405 nm was recorded in 30-s intervals for 30 min at room temperature using an iMark microplate absorbance reader (Bio-Rad). The half clot lysis time was defined as the time at which the absorbance is halfway between the plateau reached after clotting and the baseline value attained upon complete lysis.

Eicosanoid, nitrite/nitrate, and 17β-estradiol analyses.

Enzyme immunoassay kits to measure 6-keto-PGF_1α, 2,3-dinor-6-keto-PGF_1α, PGE₂ (GE Healthcare), thromboxane (TxB₂; Assay Designs), 8-isoprostane (Oxford Biomedical Research), 11-dehydro-TxB₂, PGE-M, creatinine, nitrite/nitrate, and 17β-estradiol (Cayman Chemical) were used according to the manufacturer's instructions.

Western blotting of homogenized hearts and aortas from WT and iNOS^−/− mice.

Hearts (n = 3 for each group) and aortas (n = 6 for each group) from WT and iNOS^−/− mice fed a regular chow diet for 24 wk were homogenized, and proteins in the tissue supernatants were separated via SDS/PAGE (7.5 and 10% polyacrylamide gels) and transferred onto 0.2-μm Immun-Blot PVDF membranes (162–0177; Bio-Rad). Membranes were then probed and visualized for proteins that included iNOS at ∼130 kDa (sc-560; polyclonal rabbit iNOS antibody; Santa Cruz), COX-1 at ∼70 kDa (160110; mouse monoclonal antibody; Cayman), COX-2 at ∼72 kDa (160126; polyclonal rabbit antibody; Cayman), and for GAPDH at ∼37 kDa (sc-20357; polyclonal goat antibody; Santa Cruz) in a manner described previously (4, 39). Of several iNOS antibodies tested in this application, the best antibody (with regard to reproducible iNOS detection) also resulted in the appearance of nonspecific bands at higher and lower molecular weights than iNOS, an effect that has also been reported previously (34, 45). These nonspecific interactions did not interfere with our assessment and visualization of iNOS protein levels.

LPS/IFNγ treatment of ex vivo tissue.

In some experiments, hearts and aortas were exposed ex vivo to LPS (from Escherichia coli 026:B6; 10 μg/ml; Sigma) and recombinant mouse IFNγ (100 U/ml; Calbiochem) for 24 h to induce iNOS. Hearts were removed surgically as described above. Aortas were removed surgically from the cardiac origination to the iliac bifurcation. With the use of a dissecting microscope (SMZ-1B; Nikon), the aortas were cleared of fat, connective tissues, adventitia, and denudation of the endothelial layer was performed by gently scraping with a scalpel. The tissue was dissected into smaller pieces and incubated in DMEM containing LPS/IFNγ. Following incubation for 24 h at 37°C in 5% CO₂ in air, the tissue was homogenized (50 mM Tris buffer pH 8, 10 mM ethylenediaminetetraacetic acid, 1% Tween 20, 1 mM PMSF, 1 mM sodium vanadate, and 10 μl/ml protease inhibitor cocktail set III; Calbiochem) and subjected to Western blotting, while the supernatants were analyzed for nitrite/nitrate levels.

Statistical analysis.

Results are presented as means SE with significant differences determined by t-test or two-way ANOVA, with P < 0.05 defined as statistically significant. Image J (version 1.36b: NIH) was used to quantify Western blot band densities.

RESULTS

Female iNOS^−/− mouse hearts are significantly heavier than hearts isolated from female WT mice.

The body weights of female WT vs. female iNOS^−/− mice were not significantly different (Table 1). In addition, the body weights of male WT and male iNOS^−/− mice were not significantly different from each other. Interestingly, female iNOS^−/− mouse hearts were significantly heavier than hearts isolated from female WT mice. Furthermore, when heart weights were normalized to body weight, there was still a significant difference for female iNOS^−/− hearts compared with WT female mice. In contrast, male WT and iNOS^−/− heart weights normalized to body weight were not significantly different.

Table 1.

Comparison of body weights and heart weight-to-body weight ratios for female and male WT and iNOS^−/− mice

	Body Weight, g	Heart weight, g	Heart Weight-to-Body Weight Ratio
Female WT (n = 9)	23.0 ± 0.4	0.104 ± 0.002	45.4 × 10−4 ± 1.1 × 10−4
Female iNOS−/− (n = 10)	23.9 ± 0.5	0.116 ± 0.003*	48.4 × 10−4 ± 0.7 × 10−4*
Male WT (n = 15)	31.8 ± 0.5	0.163 ± 0.005	51.3 × 10−4 ± 1.4 × 10−4
Male iNOS−/− (n = 15)	31.0 ± 0.4	0.154 ± 0.004	49.7 × 10−4 ± 1.4 × 10−4

Values are means ± SE; n = no. of mice. iNOS, inducible nitric oxide synthase; WT, wild type.

^*P < 0.05, compared with female WT mice.

Female WT mice exhibit decreased occlusion following FeCl₃-induced thrombosis compared with female iNOS^−/− mice.

We obtained reproducible and consistent results with the FeCl₃ model of carotid artery injury using aged FeCl₃ solutions (>5 days old). Figure 1A illustrates changes in the UV/Vis spectrum of FeCl₃. The changes in the UV/Vis spectrum are likely associated with hydrolysis, giving rise to hydroxide and ultimately iron oxide complexes (11).

Fig. 1.

Male and female response to FeCl₃-induced thrombosis in wild-type (WT) and inducible oxide synthase mice (iNOS^−/− mice). A: UV/Vis spectra monitoring changes in aqueous ferric chloride (FeCl₃) over time. UV/Vis spectra of FeCl₃ (1:1,000 dilution of a 10% FeCl₃·6H₂O aqueous solution) recorded at 0 min, 15 min, 60 min, and 5 days. B: representative blood flow traces following FeCl₃ application to the carotid artery (1 min) illustrating no occlusion (top trace) and occlusion ∼7 min after FeCl₃-induced injury (bottom trace). ^#Artifact introduced upon repositioning the probe. Female WT (n = 10) and iNOS^−/− mice (C; n = 10) and male WT (n = 14) and iNOS^−/− mice (D; n = 15) were subjected to FeCl₃-induced thrombosis, and blood flow was measured for 30 min following injury. Data represent the average %blood flow measured for the different groups at 0, 5, 10, 20, and 30 min. Note that 100% blood flow represents the maximal blood flow measured during a 5-min period immediately before application of FeCl₃. *P < 0.05, compared with female WT mice in both C and D. E: total time that carotid is occluded during 30 min immediately following FeCl₃-induced injury. Data represent the mean total time that blood flow was blocked during 30 min for the different groups immediately following FeCl₃ application. *P < 0.05, for female iNOS^−/− mice compared with female WT mice. +P < 0.05, for male WT mice compared with female WT mice. All mice were 6 mo in age and were fed a regular chow diet.

The response to FeCl₃-induced injury in the carotid artery was measured in four groups of mice: female WT and iNOS^−/− mice as well as male WT and iNOS^−/− mice. Figure 1B shows representative traces in which no occlusion occurs (top trace) and in which total occlusion is observed, i.e., the complete loss of blood flow (bottom trace). Figure 1, C-D, shows the average blood flow measured for female mice and male mice immediately following FeCl₃-induced injury. The data indicate that following FeCl₃ application a minimal change in blood flow was observed in female WT mice, but blood flow was impaired in iNOS^−/− female mice, indicating significant occlusion. There were no differences in the responses of WT and iNOS^−/− male mice, with both groups showing a similar response to that of female iNOS^−/− mice. The average amount of thrombotic occlusion for each of the four groups, as measured by the total time that blood flow was blocked (i.e., blood flow = 0 ml/min) during 30 min immediately following FeCl₃-induced injury, is shown in Fig. 1E. Thrombotic occlusion was significantly less in female WT mice compared with female iNOS^−/− mice or their male counterparts.

Plasma clotting and clot lysis times are similar in female and male WT and iNOS^−/− mice.

Plasma clotting and clot lysis times were measured in female and male WT and iNOS^−/− mice. Figure 2 shows that there were no appreciable differences between the groups in the mean plasma clotting times (WT female: 5.2 ± 0.3 min; WT male: 4.2 ± 0.5 min; iNOS^−/− female: 4.6 ± 0.6 min; iNOS^−/− male: 4.4 ± 0.4 min) or in the times to half-maximal clot lysis (WT female: 7.7 ± 0.7 min; WT male: 7.0 ± 0.6 min; iNOS^−/− female: 7.8 ± 0.5 min; iNOS^−/− male: 7.3 ± 0.2 min). These data indicate that plasma-based components of the coagulation and fibrinolytic systems functioned similarly in all four groups and suggested that observed differences in FeCl₃-induced thrombosis reflected alterations in either the blood vessel itself or possibly platelet reactivity.

Fig. 2.

Plasma clotting and clot lysis times for female and male WT and iNOS^−/− mice. Plasma clotting (−t-PA) and clot lysis (+t-PA) profiles were obtained for female WT (A), male WT (B), female iNOS^−/− (C), and male iNOS^−/− mice (D) as described in the materials and methods section. Traces represent an average of n = 3–5 mice for each group.

Female mice produce elevated levels of eicosanoids.

We examined urinary levels of the following: 6-keto-PGF_1α (A), 2,3-dinor 6-keto-PGF_1α (B), TxB₂ (C), 11-dehydro-TxB₂ (D), PGE₂ (E), PGE-M (F), and 8-isoprostane (G) collected from female and male WT and iNOS^−/− mice, as shown in Fig. 3. While no differences were observed between female WT and iNOS^−/− mice or male WT and iNOS^−/− mice, significant differences were observed between the female and male groups. Notably, female WT and iNOS^−/− mice produced significantly higher levels of eicosanoids than their male counterparts. Creatinine production by male or female mice was not different, indicating that the measured disparity in eicosanoid levels between male and female groups is not an artifact resulting from creatinine normalization. Despite the fact that urinary eicosanoid production was significantly higher in females compared with males, the ratio of urinary prothrombotic TxB₂ and antithrombotic 6-keto PGF_1α was similar in all groups. Thus urinary TxB₂-to-6-keto PGF_1α ratios were calculated to be 3.1 ± 0.3 in female WT, 2.6 ± 0.4 in male WT, 2.7 ± 0.3 in female iNOS^−/−, and 3.0 ± 0.5 in male iNOS^−/− mice (n = 10–12 mice).

Fig. 3.

Urinary levels of eicosanoids in WT and iNOS^−/− mice. Urinary levels of 6-keto-PGF_1α (A), 2,3 dinor 6-keto-PGF_1α (B), thromboxane (TxB₂; C), 11-dehydro-TxB₂ (D), PGE₂ (E), PGE-M (F), and 8-isoprostane (G) were measured in male and female WT and iNOS^−/− mice (n = 8–12 for each group) and normalized to creatinine. *P < 0.0001 and αP < 0.006, between female and male groups. By two-way ANOVA analysis, the sex effect was significant for all eicosanoids measured (P < 0.0001).

COX-1, prostacyclin synthase (PGI₂S), and iNOS protein levels were examined in the hearts and aortas of female and male WT and iNOS^−/− mice (Fig. 4). iNOS protein was detected in hearts from male and female WT mice but was absent, as expected, upon iNOS deletion. COX-1 and PGI₂S protein levels were elevated in female hearts from WT and iNOS^−/− mice compared with males. Importantly, COX-1 protein was also elevated in aortas homogenized from female WT mice compared with male WT mice. COX-2 and eNOS protein levels, however, were not significantly altered in the different groups (data not shown).

Fig. 4.

Cyclooxygenase-1 (COX-1) and iNOS protein expression in female and male WT and iNOS^−/− mice. A: homogenized hearts from female and male WT and iNOS^−/− mice were analyzed by Western blotting for COX-1, prostacyclin synthase (PGI₂S), iNOS, and GAPDH protein. B: Western blot band densities were quantified and are represented as either a COX-1-to-GAPDH or PGI₂S-to-GAPDH ratio. (n = 3 for each group; *P < 0.05). C: homogenized aortas from female and male WT and iNOS^−/− mice were analyzed by Western blotting for COX-1 and GAPDH protein. For each of the female and male WT and iNOS^−/− groups, 6 mice were used, with each lane representing the results from aortas homogenized from 2 mice. D: Western blot band densities were quantified and are represented as a COX-1-to-GAPDH ratio. (*P < 0.05; ^#P < 0.05 for male iNOS^−/− mice compared with male WT mice).

Urinary nitrite/nitrate and estrogen levels are higher in female WT and iNOS^−/− mice, respectively, compared with male cohorts.

Urinary levels of nitrite (NO₂⁻) and nitrate (NO₃⁻) were significantly greater in female WT mice than in other groups (Fig. 5A). While an increase in urinary levels of nitrite and nitrate might be expected compared with the iNOS^−/− cohorts, it is notable that female WT mice produce significantly higher levels than male WT mice. These findings are consistent with an elevation in iNOS-derived NO_x in female mice. Urinary 17β-estradiol levels were also measured and were significantly higher in female than in male mice. Interestingly, significantly increased levels were observed in female iNOS^−/− mice compared with female WT cohorts (Fig. 5B).

Fig. 5.

Urinary nitrite/nitrate and 17β-estradiol levels. Urinary nitrite/nitrate (A) and 17β-estradiol levels (B) were measured for male and female WT and iNOS^−/− mice (n = 9–10 for all other groups; *P < 0.05).

Female mice exhibit increased sensitivity to LPS/IFNγ-induced iNOS and nitrite/nitrate production.

Hearts and endothelium-denuded aortas from female and male WT mice were exposed ex vivo to LPS/IFNγ for 24 h. Figure 6A shows that LPS/IFNγ upregulated iNOS protein to a greater extent in female vs. male WT hearts. Consistent with this observation, Fig. 6B demonstrates that LPS/IFNγ-induced NO production was higher by endothelium-denuded aortas isolated from female vs. male WT mice.

Fig. 6.

Ex vivo LPS/IFNγ treatment of tissue from female and male WT mice. Hearts (A) and endothelium-denuded aortas (B) from female and male WT mice were exposed ex vivo to LPS (10 μg/ml) and IFNγ (100 U/ml) for 24 h to induce iNOS, as described in materials and methods. Hearts (A) were homogenized, subjected to Western blotting, and probed for iNOS and actin, and supernatants from endothelium-denuded aortas (B) were analyzed for nitrite/nitrate levels. (n = 3 for all groups; *P < 0.05).

DISCUSSION

Results reported herein demonstrate that female WT mice are less susceptible to FeCl₃-induced injury than male WT mice and that this effect is iNOS dependent (Fig. 1). While the presence of iNOS is correlated with decreased thrombotic occlusion in females, the same does not appear to be true for male WT mice. The effect could not be ascribed to differences in protein components of the plasma, as thrombin-dependent clotting and tissue plasminogen activator-dependent fibrinolytic activities were identical in the different groups of mice and not influenced by sex (Fig. 2). However, it is possible that platelet activation is altered in female mice compared with males, since it has recently been reported that 17β-estradiol inhibits platelet activation via pathways that increase NO (46).

Expression of iNOS traditionally plays a role in inflammation and the immune response, and in this regard, iNOS is activated by IFNγ and other cytokines (15). Our study, however, revealed the presence of basal levels of iNOS in the hearts of both male and female WT mice, which suggests a beneficial physiological role for a “constitutive” iNOS protein under nonpathological conditions (Fig. 4). NO released from iNOS under these conditions may favorably function as a vasodilator and may also prevent platelet aggregation. These are roles traditionally ascribed to eNOS-dependent NO (27). iNOS protein has been detected previously at low levels in rat carotid arteries (47) and in aortas and hearts of ApoE^−/− mice (30, 39). Basal levels of iNOS play a part in the control of heart rate (23). Our studies, however, demonstrate that female WT hearts possess a greater propensity for further iNOS induction and activity over basal levels than male WT hearts by LPS/IFNγ treatment (Fig. 6). Significantly greater levels of urinary nitrite and nitrate were also produced by female WT mice compared with cohorts (Fig. 5A), again indicating a greater role for iNOS in female WT mice than in male WT mice. Interestingly, male WT and iNOS^−/− mice produced similar levels of urinary nitrite and nitrate indicating very little iNOS involvement, and this result correlated with the observation that there was no significant difference in their response to FeCl₃-induced injury (Fig. 1D).

Based on the ability of NO and related NO_x species to alter eicosanoid profiles (40), we measured urinary levels of eicosanoids but did not observe a difference within the same-sex groups upon iNOS deletion (Fig. 3). Urinary isoprostane levels (generated by COX-independent mechanisms relying on free-radical driven peroxidation of arachidonic acid) were also similar between WT and iNOS^−/− mice of the same sex. Thus iNOS-derived NO does not play a role in urinary levels of eicosanoids. In contrast to these results, iNOS-deficient male mice have previously been reported to excrete 78% less PGE₂ than WT mice (24), suggesting that iNOS-derived NO species (possibly ONOO⁻) provide a stimulatory effect on COX activity (20, 41). In this previous study (24), iNOS-deficient males also produced less urinary isoprostanes than WT mice, indicating that iNOS contributes to oxidant stress. It is notable that in our study we selected 6-mo-old mice (on a C57BL/6J background) compared with 5-wk-old male mice (on mixed C57BL/6J and 129Sv/Ev backgrounds) in the previous study. Although our results did not indicate an effect of iNOS, we observed that eicosanoid production was dependent on sex, with female mice producing higher levels of urinary eicosanoids. We also noted higher levels of COX-1 and PGI₂S enzymes in homogenized hearts and aortas of female mice compared with male mice, although COX-2 levels were not altered (Fig. 4). Our results indicate a role for constitutive COX-1 in females, rather than COX-2. Interestingly, 17β-estradiol elevates COX-1 and PGI₂S proteins (21, 29) and women have elevated urinary isoprostane metabolites compared with men (36). Although eicosanoid production was different in male and female mice, the ratio of thromboxane to prostacyclin remained similar across all groups, indicating that the balance of prothrombotic vs. antithrombotic eicosanoids was not sex specific.

Urinary 17β-estradiol levels were significantly increased in female mice compared with male mice. However, an unexpected finding was that urinary 17β-estradiol levels were significantly higher in female iNOS^−/− mice compared with female WT cohorts (Fig. 5B). The relationship between iNOS and estrogen production remains to be determined. Estrogen-regulated iNOS may well provide beneficial levels of NO for the circulation and for preventing damage during vascular injury or ischemia in the myocardium (1). Estrogen has been shown to protect against vascular injury in mice bearing genetically disrupted estrogen receptors (14, 25, 48), but the use of oral contraceptives containing estrogen-related hormones remains controversial (35).

In summary, our results indicate that FeCl₃-induced thrombosis in female WT mice is less than in male WT mice and that the reduction is related to increased iNOS activity.

GRANTS

Support for this study was provided by National Heart, Lung, and Blood Institute Grants PO1-HL-046403 (to D. P. Hajjar and K. A. Hajjar) and HL-042493 and HL-090895 (to K. A. Hajjar), an Alice Bohmfalk Charitable Trust Award (to R. K. Upmacis), a Hartwell Foundation Award (to B. D. Lamon), American Heart Association Grant-in-Aid AHA655783T (to R. S. Deeb), and the Julia and Seymour Gross Foundation (to D. P. Hajjar).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).