A novel nitric oxide-releasing statin derivative exerts an antiplatelet/antithrombotic activity and inhibits tissue factor expression1

Authors


  • 1

    Presented in part at the Scientific Sessions 2003, American Heart Association (Circulation 2003; 108: 759).

Paolo Gresele, Department of Internal Medicine, Section of Internal and Cardiovascular Medicine, Via Enrico dal Pozzo, 06126 Perugia, Italy.
Tel.: +39 075 5783989; fax: +39 075 5716083; e-mail: grespa@unipg.it

Abstract

Summary. Background: NO-releasing statins are new chemical entities, combining HMG-CoA reductase inhibition and slow NO release, that possess stronger anti-inflammatory and antiproliferative activities than the native statins. Objective: We evaluated the antithrombotic effects of nitropravastatin (NCX-6550) by assessing its activity on platelet activation and tissue factor (TF) expression by mononuclear cells in vitro and in vivo. Methods and results:In vitro, NCX-6550 inhibited (1) U46619- and collagen-induced platelet aggregation in buffer and plasma; (2) collagen-induced P-selectin expression in whole blood and (3) platelet adhesion to collagen-coated coverslips under high shear stress. These effects were displayed at concentrations of NCX-6550 ranging from 25 to 100 μm, and were totally reverted by the guanylylcyclase inhibitor ODQ (10 μm). Equimolar concentrations of pravastatin had no influence on these parameters of platelet function. LPS- and PMA-induced TF expression by blood mononuclear cells was also inhibited by NCX-6550 (IC50 13 μm), but not by pravastatin, as assessed by functional and immunological assays and by real-time PCR. In a mouse model of platelet pulmonary thromboembolism, induced by the i.v. injection of collagen plus epinephrine, pretreatment with NCX-6550 (24–48 mg kg−1) significantly reduced platelet consumption, lung vessel occlusion and mortality. Moreover, nitropravastatin markedly inhibited the generation of procoagulant activity by spleen mononuclear cells and peritoneal macrophages in mice treated with LPS. In these in vivo models too, pravastatin failed to affect platelet activation and monocyte/macrophage procoagulant activity. Conclusions: Our results show that nitropravastatin exerts strong antithrombotic effects in vitro and in vivo, and may represent an interesting antiatherothrombotic agent for testing in acute coronary syndromes.

Introduction

Hypercholesterolemia is the main trigger of arterial atheroma formation. On the other hand, hypercholesterolemia and platelet hyper-reactivity are associated conditions and closely participate in the development of atherosclerosis-related thrombotic events. Ischemic cardiovascular events typically arise from the disruption of atherosclerotic plaques, especially those with inflammatory characteristics, with the exposure of platelet-activating substances and of tissue factor (TF) and the consequent formation of a platelet-rich thrombus [1,2].

Inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, statins, are widely used for the treatment of hypercholesterolemia and several clinical trials have shown that they efficiently reduce LDL-cholesterol thereby decreasing cardiovascular morbidity and mortality in hypercholesterolemic patients, in both primary and secondary prevention. However, statins exert protection against cardiovascular mortality also in normocholesterolemic subjects [3], suggesting that mechanisms beyond lipid lowering might contribute to their beneficial effects. Among these, the anti-inflammatory and antiproliferative actions appear of relevance because they may act in concert with lipid reduction to favorably influence the composition, structure and stability of atherosclerotic plaques [1,4].

Statins may also influence the thrombogenic properties of the plaque, among which the expression of TF is a prominent one. Indeed, functionally active TF is expressed by all the major cell types within the plaque, particularly by macrophages, and is abundant also in the lipid core, mainly in the form of microparticles of macrophage origin [2,5,6]. Its clinical relevance is highlighted by the fact that it is detected more frequently and at higher levels in vascular lesions associated with unstable angina and myocardial infarction than in those associated with stable angina [7]. Statins have been shown to reduce TF expression by cultured human macrophages and experimental atherosclerotic lesions [4,8,9]. The anti-inflammatory and antithrombotic effects of statins might be beneficial especially in conditions like the acute phase of coronary syndromes, characterized by an intense inflammatory response and thrombus formation. In most of the clinical and experimental studies, statins have displayed a protective action against ischemic cardiovascular events only after medium- to long-term treatment [10–13]. However, it is within the early period of an acute coronary syndrome (ACS) that patients experience the highest rate of death and recurrent ischemic events. To date, it has not been determined with certainty whether initiation of treatment with a statin soon after an ACS can reduce the occurrence of these early events. Three large clinical trials, the MIRACL [14], the PROVE IT-TIMI 22 [15] and the A to Z [16] studies, so far have tested the hypothesis of whether an early treatment with a statin can reduce recurrent ischemic events and death in the early period after an ACS. However, even a highly effective statin (atorvastatin 80 mg day−1) did not attain significant cardiovascular protection within the first hours or days of treatment [14,15] despite some recent reports showing that some statins, including atorvastatin, reduce platelet activation as early as 18 h after intake [17,18].

Nitric oxide, an endothelial- and platelet-derived mediator, inhibits platelet activation, leukocyte activation, smooth muscle cell proliferation and TF expression by endothelial cells thus exerting potentially antiatherothrombotic activities [19,20]. One mechanism of statins attracting interest is the enhancement of nitric oxide biosynthesis or activity [21]. Statins increase endothelial NO synthesis by stabilizing eNOS-mRNA, with an increase in eNOS protein, and by promoting eNOS activation via the serine/threonine kinase Akt-mediated phosphorylation [22].

Recently, a new class of drugs combining a statin with an NO-donating moiety, nitrostatins, has been described and found to exert enhanced in vitro antiproliferative and anti-inflammatory properties when compared with the native statins [23]. In order to assess whether the incorporation of an NO-donating moiety potentiates also the antithrombotic effects of statins, we investigated the influence of NO-pravastatin (NCX-6550) (NicOx Research Institute, Milan, Italy) on platelet activation and monocyte TF expression in vitro and in vivo.

Methods

In vitro studies

Reagents  U46619 (9,11-dideoxy-11α,9α epoxymethano-prostaglandin F), ODQ (1H-[1,2,4]oxadiazolo-[4,3-a]quinoxalin-1-one), PMA (phorbol-12-myristate-13-acetate), and SNP (sodium nitroprusside dihydrate) were from Sigma Chemicals (Milan, Italy); carboxy-PTIO (2[4-carboxyphenyl]-4,4,5,5-tetramethyl-imidazoline-1-oxyl-3-oxide) and FK409 ((±)-(E)-Ethyl-2-[(E)-hydroxyimino]-5-nitro-3-hexeneamide) were from Alexix Biochemicals (Vinci, Florence); epinephrine bitartrate, a 5 mm solution in Tris buffer, was from Mascia Brunelli (Milan); equine tendon collagen in suspension was from Hormon Chemie (Munich, Germany); FITC-conjugated mouse antihuman P-selectin (CD62P FITC) clone CLB-Thromb/6 and PE-conjugated mouse antihuman glycoprotein IIb/IIIa (CD41PE) clone P2 were from Immunotech (Marseille, France); lipopolysaccharide (Escherichia coli LPS) and thioglycollate solution were from Difco Laboratories (Detroit, MI, USA); RPMI 1640 was from Euroclone (Milan); Fycoll-Hypaque (Lympholyte-H) was from Cederline (Hornby, Ontario, Canada); NCX-6550 (1S-[1α(βS*,δS*), 2α,6α,8β-(R*),8aα]-1,2,6,7,8,8a-hexahydro-β,δ,6-trihydroxy-2-methyl-8-(2-methyl-1-oxobutoxy)-1-naphtaleneheptanoic acid 4-(nitrooxy)butyl ester), pravastatin and fluvastatin were provided by NicOx Research Institute, Milan. They were dissolved in DMSO and then diluted in buffer or culture medium at the desired concentration. DMSO final concentration never exceeded 0.5%.

Blood collection and cell preparations  Blood was collected from drug-free, fasting healthy human volunteers by venipuncture into 3.8% trisodium citrate (9 vol. of blood + 1 vol. of citrate). Platelet-rich plasma (PRP) was obtained by centrifugation at 150 g for 10 min, and adjusted to 250 × 109 L−1. Gel-filtered platelets (GFP) were prepared by passage through a Sepharose 2B column and adjusted to 100 × 109 L−1 before use [24]. Mononuclear cells were isolated by the Fycoll-Hypaque method as described [25], and resuspended in serum-free RPMI medium at the concentration of 3 × 109 L−1.

Platelet studies  Platelet aggregation was performed in PRP or GFP, according to Born, using a four-channel aggregometer (Aggregocorder, Menarini, Florence, Italy), as described [26]. U46619 (0.2 μm), a thromboxane (TxA2) analog, and collagen (2 μg mL−1) were used as aggregating agents. Test drugs were incubated with platelets at 37 °C for 30 min before stimulation.

Intraplatelet cGMP was measured by ELISA (R&D System, Minneapolis, MN, USA), as previously described [27].

Platelet surface P-selectin expression was analyzed in human whole blood stimulated with collagen (30 μg mL−1 for 30 min) by flow cytometry (EPICS XLMCL, Instrumentation Laboratory, Milan), as reported [28].

Platelet adhesion to a collagen-coated surface was studied as previously described [29]. Briefly, blood (5 mL) was passed through a rectangular parallel plate perfusion chamber, over a plastic coverslip sprayed with collagen from equine tendon (∼30 μg cm−2), at wall shear rates of 800 s−1 and 3000 s−1. Afterwards, the chamber was perfused with BSA-containing saline to remove all residual blood, and the coverslip was harvested, gently washed with 10 mmol L−1 HEPES, and fixed with 0.25% glutaric-dialdehyde in PBS. Adherent platelets were stained with May-Grünwald/Giemsa and observed under an optical microscope. The area covered by platelets was measured with computerized image analyzer (NIH Scion Image, Frederick, MD, USA). Drugs (100 μm) were added to blood and incubated at 37 °C for 30 min prior to the perfusion experiment. Control blood samples were run at the beginning and at the end of each experimental session to check for reproducibility.

TF expression by mononuclear cells  Washed mononuclear cells (3 × 109 L−1 in RPMI) were supplemented with test drug or vehicle and incubated for 3 h at 37 °C with either 1 μg mL−1 LPS or 10−9 m PMA. At the end of incubation, TF production was assessed by functional and immunological assays. Under all experimental conditions, cell viability, assessed by the trypan blue exclusion test, was always >90%.

Tissue factor activity was determined on intact cells by a one-stage clotting assay and expressed in arbitrary units, as reported [25]. Tissue factor antigen was assayed by an ELISA method (Imubind TF, Instrumentation Laboratory) on 0.25% Triton X-100 cell extracts, prepared as previously described [30], and expressed as ng mg−1 protein.

Expression of TF and β2-microglobulin genes in LPS-stimulated cells was determined by real-time PCR. RNA extraction and cDNA preparation were carried out as reported [30]. Real-time PCR of TF and β2-microglobulin cDNAs was performed in triplicate on the iCycler Real-Time PCR Detection System (BioRad, Segrate, Italy), using Pre-Developed Taqman® assay kits from Applied Biosystems, Monza, Italy (assay ID: Hs00175225_m1 and Hs99999907_m1 for TF and β2-microglobulin, respectively). PCR annealing and amplification temperatures were as detailed in the manufacturer's instructions. Relative quantification of TF expression was carried out by standard 2−ΔΔCT method [31] using β2-microglobulin as internal control gene and LPS-stimulated cells, in the absence of drug, as reference sample.

In vivo studies

Animals  Male CD-1 mice (Charles Rivers, Calco, Como, Italy), weighing 20–25 g, were used. The study was approved by the Committee on Ethics of Animal Experiments of the University of Perugia and by the Italian Ministry of Public Health (Authorization 21/2000-B). The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).

Plasmatic nitrites and nitrates  Nitric Oxide formation following the i.p. administration of NCX-6550 was monitored by measurement of circulating nitrites and nitrates (NOx). At fixed time points after drug administration, blood was collected by cardiac puncture from anesthetized mice into 3.8% trisodium citrate (9 vol. blood + 1 vol. citrate). Plasma was prepared by centrifugation at 1000 g for 10 min and NOx were measured by a colorimetric, non-enzymatic method (Oxford Biomedical Research, Rochester Hills, MI, USA) [32].

Pulmonary thromboembolism  Pulmonary thromboembolism was induced by a previously described method [33]. The drugs or their vehicles were administered i.p. or, in some experiments, by gavage in a fixed volume of 100 μL, 3 h before i.v. injection of a mixture of collagen (12.5 μg kg−1) and epinephrine (0.075 μg kg−1) or of U46619 (0.2 mg kg−1). The cumulative end point to be overcome was death of the animal within 15 min or prolonged paralysis of the hind limbs (for more than 15 min). In some experiments, we used a model of non-thrombotic disseminated pulmonary microembolism, which is sensitive to vasodilators but not to antiplatelet agents, induced by the i.v. injection of a suspension of swollen, hardened rat red blood cells (300 μL, hematocrit 12.5%) [34]. No anesthesia was given during the pulmonary embolism experiments because of the short duration and because anesthesia interferes with thromboembolism in this model [35].

In mice challenged with collagen plus epinephrine, we also determined platelet consumption and lung vessel occlusion. Blood and lungs were collected 2 min after the thrombotic challenge and processed as reported [34]. Platelet count was carried out by standard techniques. Hematoxylin and eosin-stained lung sections were examined under a light microscope (Wild-Leitz, Heerbrugg, West Germany) by a pathologist unaware of the treatment administered to animals. At least 10 fields, at a magnification of 400×, were observed for every specimen. The total number of identifiable lung vessels per field was counted and the percentage of them occluded by platelet thrombi was annotated.

Blood pressure  Systolic blood pressure was determined in trained conscious mice using a non-invasive computerized tail-cuff system (BP-2000, Visitech System, Apex, NC, USA).

Induction of procoagulant activity in spleen mononuclear cells and peritoneal macrophages  In a set of experiments, mice were given LPS (4 mg kg−1 i.v.) and 4 h later they were killed and the spleen was surgically removed and finely minced with scissors under sterile conditions. Mononuclear cells were isolated by density gradient centrifugation as reported [36] and suspended in RPMI at the concentration of 109 L−1. The percentage of monocytes was assessed by May-Grunwald staining of cytospin. To study peritoneal macrophages, mice were first injected i.p. with 1 mL endotoxin-free 10% thioglycollate solution. Three days later, the animals received an i.p. injection of LPS (4 mg kg−1) and, after an additional 4 h, macrophages were harvested by peritoneal lavage under anesthesia. Cells were washed three times by centrifugation (15 min at 500 g), under sterile conditions, and finally suspended in RPMI at the concentration of 109 L−1. Cell procoagulant activity was determined by a one-stage clotting assay as reported [25], using human plasma. In all instances, drugs were injected i.p. 15 min before LPS challenge.

Statistical analysis

Data are expressed as mean ± SEM. Differences among groups were tested by one-way anova using the Tukey's multiple comparison test as post test. Differences in mortality were assessed by the Fisher's exact test. All analyses were performed with the GraphPad Prism Software (San Diego, CA, USA).

Results

In vitro studies

Effect on platelet activation  The antiplatelet activity of NCX-6550 was investigated by studying different platelet functions. Firstly, we evaluated the effect of NCX-6550 on platelet aggregation induced by physiologically relevant agonists. As shown in Fig. 1, NCX-6550, unlike pravastatin, inhibited dose-dependently the aggregation of GFP induced by either collagen (Fig. 1A) or the TxA2 analog U46619 (Fig. 1B), with an IC50 of 25 μm and 66 μm, respectively. Similar results were obtained when platelet aggregation was studied in PRP (not shown), indicating that plasma proteins do not affect the pharmacologic effects of the compound.

Figure 1.

Effects of NCX-6550 on platelet activation. Panels A and B, effect on platelet aggregation. Human GFP were supplemented with increasing concentrations of drug and incubated at 37 °C for 30 min, after which platelet aggregation was induced by collagen (2 μg mL−1, panel A) or the TxA2 analog U46619 (0.2 μm, panel B). Results are expressed as maximal amplitude of aggregation. Panel C, effect on P-selectin expression. Human whole blood was incubated with increasing concentrations of drug for 30 min at 37 °C after which collagen (30 μg mL−1) was added in order to induce the expression of P-selectin. After an additional 30 min incubation, P-selectin-positive platelets were quantified by flow cytometry and expressed as percent of total platelets. Panel D, effect on platelet adhesion to collagen. Human blood was incubated with test drug (100 μm) or vehicle for 30 min at 37 °C and then passed through a perfusion chamber, over a collagen-coated coverslip, at wall shear rate of 800 s−1 (dashed bars) and 3000 s−1 (filled bars). The area covered by platelets was measured with a computerized image analyzer. Under all experimental conditions, the concentration of NCX-6550 giving the maximal inhibitory effect (100–200 μm) was also tested in the presence of ODQ (10 μm). Data are the mean ± SEM of 5 (panel D) or 10 experiments. *, P < 0.05 vs. vehicle.

P-selectin is an adhesion molecule stored in platelet α-granules and is translocated to platelet surface upon activation. In view of the key role played by this adhesion molecule in thrombus formation [37], we tested the effect of nitropravastatin on P-selectin exposure. To that purpose whole blood was stimulated with collagen, in the absence and in the presence of test drug, and P-selectin positive platelets were assessed by flow cytometry. As shown in Fig. 1C, NCX-6550, but not pravastatin, inhibited P-selectin expression in a concentration-dependent fashion.

Next, we evaluated the effect of NCX-6550 on platelet adhesion to collagen-coated coverslips perfused with whole blood at two different shear-rates (800 s−1 and 3000 s−1) [38]. Nitropravastatin (100 μm) significantly reduced platelet adhesion under both shear rate conditions, whereas pravastatin was ineffective (Fig. 1D).

The inhibition of platelet activation by NCX-6550 was totally abolished by a specific inhibitor of soluble guanylylcyclase (ODQ), indicating an NO-mediated mechanism (Fig. 1A–D). Accordingly, the incubation of platelets with NCX-6550 (100 μm) for 30 min resulted in a marked increase of intraplatelet cGMP (from 0.18 ± 0.02 to 1.31 ± 0.34 μm 10−8 platelets, n = 3, P < 0.01), an effect totally reversed by ODQ.

Effect on TF expression  Exposure of human blood mononuclear cells to LPS for 3 h induced strong TF activity (from <1 U 10−6 cells to 198 ± 41 U 10−6 cells). Addition of NCX-6550 immediately before LPS challenge reduced TF activity in a concentration-dependent manner, with an IC50 of 13 μm (Fig. 2). Under the same experimental conditions, pravastatin (up to 500 μm) had no influence on TF expression. Considering that the introduction of an NO-donating moiety in pravastatin renders the molecule more lipophilic, thus increasing its penetration into cells, we also tested fluvastatin, which has a degree of lipophilicity comparable with NCX-6550 [23]. As shown in Fig. 2, fluvastatin indeed inhibited TF activity, but at concentrations markedly higher than the nitropravastatin. To assess the contribution of NO to the TF-inhibiting activity of NCX-6550, we evaluated the effect of ODQ (10 μm) and of an NO scavenger (carboxy-PTIO, 50 μm). Neither compound attenuated to any significant extent the effect of the nitropravastatin (Fig. 2). We also tested whether pure NO-donors were able to reduce the generation of TF activity in LPS-stimulated mononuclear cells. Both SNP (10 μm) and FK409 (25 μm) failed to influence TF activity (not shown). SNP, moreover, also failed to enhance the TF inhibitory activity of fluvastatin (Fig. 2).

Figure 2.

Effect of NCX-6550, pravastatin and fluvastatin on LPS-induced TF activity of human blood mononuclear cells. Test compounds were added to cells and then incubated at 37 °C in the presence of LPS (1 μg mL−1). After 3 h, TF activity was measured on intact cells by clotting assay. Since the compounds could not be tested on the same cell preparations and because of the large interindividual variability in TF response to LPS, TF activity is expressed as percent of the relative control tested in parallel. NCX-6550 (50 μm) was also tested in the presence of ODQ (10 μm) or carboxy-PTIO (25 μm). Fluvastatin was also tested in the presence of SNP (10 μm). Data are the mean ± SEM of five experiments.

Qualitatively similar results were obtained when PMA was used as TF inducer instead of LPS (not shown).

We also evaluated the effect of NCX-6550 on TF antigen and mRNA in cells stimulated for 3 h with LPS: both TF protein and mRNA were strikingly lower in cells treated with NCX-6550 (50 μm), whereas they were unchanged in pravastatin- and SNP-treated cells (Fig. 3).

Figure 3.

Effect of NCX-6550 (50 μm), pravastatin (50 μm), and SNP (10 μm) on TF antigen and mRNA expression. Blood mononuclear cells were supplemented with test compound or vehicle and then stimulated with LPS for 3 h at 37 °C. TF antigen (dashed columns) was assayed by ELISA on Triton X-100 cell extracts. TF mRNA (open columns) was determined by real-time PCR, using β2-microglobulin as house-keeping gene, and expressed as fraction of vehicle. Data are the mean ± SEM of five (antigen) or three experiments. *, P < 0.01 when compared with vehicle.

In vivo studies

Plasmatic nitrites and nitrates  To gain information on the in vivo release of NO from NCX-6550, we measured the circulating levels of nitrites and nitrates (NOx) after a single i.p. administration of the drug. The injection of 24 mg kg−1 of NCX-6550 produced a rise of plasmatic NOx that peaked at 3 h and returned to baseline at 6 h (basal = 26.7 ± 5.9; 1 h = 53 ± 18.3; 3 h = 63.9 ± 16.7; 6 h = 27.6 ± 4.4 μm, n = 4). An equimolar dose of pravastatin (20 mg kg−1) failed to modify plasmatic NOx (1 h = 17.8 ± 6.9; 3 h = 17.6 ± 5.38; 6 h = 27.5 ± 7.2 μm, n = 4). Based on these findings, subsequent experiments on the antiplatelet effect of NCX-6550 were carried out by injecting the drug 3 h before starting the experiment.

Effect on platelet activation  To study the antiplatelet activity of nitropravastatin in vivo, we used a model of platelet pulmonary thromboembolism induced by the i.v. injection of collagen plus epinephrine and evaluated the effect of the drug on (1) mortality; (2) platelet consumption and (3) lung vessel occlusion. The i.p. injection of NCX-6550, given 3 h before thrombogenic challenge, reduced mortality in a fairly dose-dependent manner whereas pravastatin was ineffective at all tested doses (Fig. 4).

Figure 4.

Effect of NCX-6550 and equimolar pravastatin on collagen plus epinephrine-induced mortality in mice. Test drug or vehicle was injected i.p. 3 h before the thrombogenic challenge and mortality was assessed as detailed in methods. Data are expressed as percent mortality. *, P < 0.025; **, P < 0.01 vs. vehicle, n = 20–24.

An NCX-6550 treatment reduced also platelet consumption. Indeed, residual circulating platelets, 2 min after the thrombogenic stimulus, were 64.0 ± 7.5 × 109 L−1 in the vehicle group, 214.7 ± 45.7 × 109 L−1 in the NCX-6550 group (48 mg kg−1, P < 0.01), and 79.7 ± 16.1 × 109 L−1 in the pravastatin group (40 mg kg−1) (n = 8 per group).

Histologic examination of lung slices, obtained from control animals 2 min after the injection of collagen plus epinephrine, showed that 90.9 ± 1.8% of lung vessels were totally or partially occluded by platelet thrombi. In NCX-6550-treated animals (48 mg kg−1, i.p.), the number of occluded lung vessels was slightly but significantly reduced (72.5 ± 3.6%, P < 0.01) while in pravastatin-treated mice (40 mg kg−1) it was unchanged (88.5 ± 3.6 %, n = 5 per group).

NCX-6550 reduced also platelet pulmonary thromboembolism mortality induced by the i.v. injection of the stable TxA2 analog U46619 (not shown), an activity characteristic of nitric oxide donating agents [39].

To see whether the vasodilatory activity of NCX-6550 contributed to the protection of mice against mortality, we tested the drug in a model of mechanical pulmonary microembolism, induced by the i.v. injection of swollen, hardened rat red blood cells, which is sensitive to vasodilating agents but not to antiplatelet drugs [34]. In this model too, pretreatment with NCX-6550 (48 mg kg−1, i.p. 3 h before) reduced mortality from 80% to 35% (P < 0.01, n = 20) whereas pravastatin (40 mg kg−1) was ineffective (70% mortality). These data suggest that lung microvascular vasodilatation contributed to the antithrombotic effect of NCX-6550.

Finally, to determine whether nitropravastatin injection caused a systemic hypotensive reaction, we measured systolic blood pressure 3 h after i.p administration of the drug, i.e. when the circulating NOx reached the maximum level. Neither NCX-6550 (48 mg kg−1) nor pravastatin (40 mg kg−1) caused any noticeable change in systolic blood pressure (75.4 ± 3.3 and 80.3 ± 4.3 mmHg, respectively, when compared with 72.6 ± 4.97 mmHg in controls, n = 5).

Procoagulant activity of spleen mononuclear cells and peritoneal macrophages  In order to investigate the effect of nitropravastatin on TF production in vivo, we assayed the procoagulant activity of spleen mononuclear cells and peritoneal macrophages harvested from mice injected with LPS. Treatment with LPS caused a marked enhancement of the procoagulant activity of both cell types as indicated by their ability to shorten plasma clotting time. For example, peritoneal macrophages from control mice caused plasma to clot in 269 ± 29.3 s whereas macrophages from LPS-treated mice caused plasma to clot in 103 ± 7.2 s. NCX-6550 (48 mg kg−1, i.p.), given 15 min before LPS challenge, significantly reduced the procoagulant activity of both spleen cells and peritoneal macrophages (Fig. 5), whereas pravastatin was totally ineffective. This effect was not associated with any appreciable change in monocyte/lymphocyte ratio in spleen cell preparations or in the percentage of macrophages in peritoneal cells (not shown). When cell procoagulant activity was expressed in TF equivalent units using human thromboplastin as reference, it was calculated that NCX-6550 treatment reduced mean cell procoagulant activity by 71% and 59% in peritoneal macrophages and spleen cells, respectively.

Figure 5.

Effect of NCX-6550 (48 mg kg−1) and pravastatin (40 mg kg−1) on the expression of procoagulant activity by spleen mononuclear cells (panel A) and peritoneal macrophages (panel B) obtained from LPS-treated mice. Test drug or vehicle was given i.p. 15 min before LPS challenge. Cells were harvested 4 h later and tested for procoagulant activity by clotting assay. Data are presented as clotting time. *, P < 0.05; **, P < 0.001 vs. vehicle.

Discussion

This study shows that a novel NO-donating (NCX-6550) derivative of pravastatin displays a marked antiplatelet and TF-suppressing activity. Indeed, it inhibited human platelet aggregation induced by U46619 or collagen, both in buffer and plasma. Moreover, it inhibited collagen-induced platelet P-selectin expression in whole blood as well as platelet adhesion to collagen-coated coverslips under flow conditions. The latter effect was observed under two different shear rate conditions, including those encountered in stenotic coronary arteries [38], and appears to be particularly relevant considering that a conventional antiplatelet agent, such as aspirin, does not affect platelet activation under these conditions [40]. The antiplatelet activity of NCX-6550 was associated with an increase of intracellular cGMP and was abolished by the guanylylcyclase inhibitor ODQ, indicating that it was mediated by NO release. This might explain why rather high concentrations of NCX-6550 were required to inhibit platelet functions. Indeed, as previously shown [23], the in vitro release of NO from nitropravastatin is a slow process that lasts for several hours. The effect of NCX-6550 on platelet activation in vivo was studied using a model of platelet pulmonary thromboembolism induced by the i.v. injection of collagen plus epinephrine. This model is characterized by the massive activation of circulating platelets and the widespread formation of platelet thrombi in the microcirculation of the lungs leading to disseminated pulmonary microembolism and death of the animal [33,34]. The i.p. injection of NCX-6550, given 3 h before the thrombogenic challenge, an interval required to reach the peak level of nitrites and nitrates in blood, reduced thromboembolic mortality markedly and in a fairly dose-dependent manner. This effect was associated with a significant reduction of both platelet consumption and lung vessel occlusion, confirming the in vivo antiplatelet activity of NCX-6550. In the same model, thromboembolic mortality induced by the i.v. injection of the TxA2-analog U46619 was also prevented, an activity typical of NO-donating agents [34]. Moreover, the observation that nitropravastatin also enhanced survival in a model of non-thrombotic, mechanical pulmonary microembolism, which is responsive to vasodilators but not to antiplatelet drugs [33,34], suggests that a vasodilatory activity on lung blood vessels may also play a role. This activity is displayed without any significant hypotensive reaction, similarly to what previously observed with another NO-donating agent, nitroaspirin, most likely because the in vivo hydrolysis of these nitrocompounds is not a fast phenomenon, requires metabolism, probably at cellular level, and produces discrete but relatively long-lasting plasma levels of NO [41]. As compared with nitroaspirin [34], however, NCX-6550 appears to be approximately fivefold more potent in protecting against platelet pulmonary thrombembolism.

The second goal of our study was to evaluate the effect of NCX-6550 on TF expression. In in vitro studies, we used human blood mononuclear cells which, when challenged with LPS or PMA, synthesize large amounts of TF thus increasing their procoagulant potential by several orders of magnitude. Using a concentration of LPS that produces maximal stimulation of mononuclear cells [25], we found that NCX-6550 inhibited the expression of TF activity in a concentration-dependent manner with an almost complete suppression at 50 μm. NCX-6550 also inhibited the generation of TF activity by cells stimulated with PMA, suggesting a broad effect on TF expression, independent of the stimulus. The decrease in TF activity was accompanied by a parallel reduction of total TF antigen and mRNA content, indicating inhibition of TF synthesis. Although pravastatin has been reported to inhibit monocyte NFkB activation and TF expression in vitro [42], we did not observe any appreciable effect of this statin on TF production, which is in agreement with the data of Colli et al. [8] and with the rather hydrophobic nature of this molecule that can hardly penetrate cell membranes. The striking difference in TF-inhibiting activity between NCX-6550 and pravastatin might be due to the physicochemical changes caused by the introduction of the NO-donating moiety and/or to the release of NO. NCX-6550 has a greater lipophilicity than pravastatin and thus it is more prone to penetrate into the cell [23]. As a matter of fact, fluvastatin, which has a lipophilicity comparable with NCX-6550, did inhibit TF expression by blood mononuclear cells, though at concentrations markedly higher than the nitrocompound. A number of data, instead, suggest that NO release from NCX-6550 does not play any significant role in TF inhibition: (1) neither ODQ nor the NO scavenger carboxy-PTIO attenuated the effect of NCX-6550; (2) two different NO donors, SNP and FK409, failed to inhibit LPS-induced TF activity; (3) the NO donor SNP failed to enhance the TF-inhibiting activity of fluvastatin. Therefore, as far as TF inhibition is concerned, one major advantage deriving from the introduction of the NO donating moiety is a greater ability to penetrate into the cell and to inhibit the signaling pathways triggered by HMG-CoA reductase.

NCX-6550 was also able to inhibit the in vivo generation of procoagulant activity by spleen mononuclear cells and peritoneal macrophages in LPS-treated mice. However, it should be noted that, although mouse monocyte/macrophages have been shown to synthesize TF upon LPS challenge [43], it cannot be excluded that other procoagulants, like prothrombinase [44], may have contributed to the shortening of the clotting test in our experiments. Whatever the nature of cell procoagulant activity(ies), our findings show that NCX-6550 reduces the prothrombotic properties of activated mononuclear phagocytes in vivo, underscoring the potential of this drug as antithrombotic agent.

Arterial thrombosis usually occurs on disrupted atherosclerotic plaques and involves multiple mechanisms. Following plaque disruption, the exposure of flowing blood to platelet-activating substances present in the lesions, especially collagen, leads to platelet adhesion and aggregation whereas TF expressed by the major inflammatory cell types within the plaque, particularly macrophages, triggers blood coagulation thus initiating the formation of a platelet-rich thrombus. Platelet activation also causes the translocation of P-selectin from α-granules to the cell surface and recent evidence indicates that platelet-exposed P-selectin in the nascent thrombus plays a crucial role in thrombus growth by favoring leukocyte recruitment, notably monocytes, and supporting the binding of circulating TF-expressing microparticles (mainly of monocyte origin) [37]. NCX-6550, by inhibiting platelet functions involved in thrombus formation, including those insensitive to aspirin, and by suppressing TF production by monocyte/macrophages, represents a new pharmacologic entity with an improved antithrombotic profile. The antithrombotic effects are displayed within a few hours after single-dose administration and are, thus, unrelated to cholesterol lowering, an activity which is fully retained by NCX-6550 as observed in hypercholesterolemic animals following prolonged administration of the nitrostatin (Gresele P, University of Perugia, Perugia, Italy, unpublished data). A possible draw-back of NO-donating drugs, particularly when administered chronically, is the formation of peroxynitrite as a consequence of the interaction of NO with superoxide released by different cell types upon activation [45]. Peroxynitrite is highly reactive and may cause lipid peroxidation and protein nitration, including oxidation of LDL, which has strong proatherogenic effects [46]. However, the likelihood of peroxynitrite formation following adminidtration of NCX-6550 is very low, if not negligible, as suggested by the finding that lipid peroxidation was reduced in hypercholesterolemic mice chronically treated with NCX-4016 [47], a nitroaspirin which has a similar NO-donating pattern than nitropravastatin. In conclusion, our data suggest that NCX-6550 can be useful in the acute treatment of patients with atherothrombosis associated with an inflammatory component, such as in the first few hours or days of the ACS, and warrants testing in clinical trials.

Acknowledgement

The help of Dr C. Rotunno with some of the analyses is gratefully acknowledged.

Disclosure of conflict of interest

Part of the work was supported by a grant from NicOx Research Institute, Milan, Italy (MC, PG).

PG and MC designed the study and wrote the manuscript after a critic evaluation of the results. NS supervised part of the experiments, participated in discussing the results and revised critically the manuscript; MRR, SM, SG, GG and RC carried out the experiments and the statistical analysis of the data obtained. AM and EO revised critically the manuscript.

All authors gave the final approval for submission of the paper.

Ancillary