SEARCH

SEARCH BY CITATION

Keywords:

  • carbon nanotubes;
  • inflammation;
  • leukocytes;
  • P-selectin;
  • thrombosis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

Summary. Background: Inhaled ultrafine particles trigger peripheral thrombotic complications. Methods: We have analyzed the systemic prothrombotic risk following lung inflammation induced by pulmonary carbon nanotubes (CNTs). Results: Intratracheal instillation in Swiss mice of 200 and 400 μg of multiwall ground CNTs triggered substantial lung neutrophil, but not macrophage influx, 24 h later. The detection of circulating platelet–leukocyte conjugates exclusively 6 h after CNT instillation pointed to early but transient activation of circulating platelets. At 24 h, elevated plasma procoagulant microvesicular tissue factor activity was found in CNT-exposed but not in saline-exposed mice. However, at 24 h, both the tail and jugular vein bleeding times were prolonged in CNT-exposed but not in saline-exposed mice, arguing against strong CNT-induced platelet activation at this point. Nevertheless, at 24 h, enhanced peripheral thrombogenicity was detected in CNT-exposed but not in saline-exposed mice, via quantitative photochemically induced carotid artery thrombosis measurements. P-selectin neutralization abrogated platelet–leukocyte conjugate formation and microvesicular tissue factor generation, and abolished the CNT-induced thrombogenicity amplification. In contrast, the weak vascular injury-triggered thrombus formation in saline-treated mice was not affected by P-selectin neutralization at 24 h. Conclusions: The mild CNT-induced lung inflammation translates via rapid but mild and transient activation of platelets into P-selectin-mediated systemic inflammation. Leukocyte activation leads to tissue factor release, in turn eliciting inflammation-induced procoagulant activity and an associated prothrombotic risk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

Platelets are key players in atherothrombosis at the time of plaque rupture, but they are also crucial for the progression of atherosclerotic lesions [1,2]. P-selectin holds a central position in linking inflammation and thrombosis, because it mediates the adhesion of activated platelets to circulating leukocytes, a process contributing to their activation [3]. Activation of circulating monocytes results in the production and release of endogenous tissue factor [4]. More importantly, monocyte activation is accompanied by the production of circulating tissue factor-positive microvesicles [5], which concentrate on activated platelets in a growing thrombus in a P-selectin-dependent manner, thus leading to the accumulation of circulating tissue factor on negatively charged surfaces and stimulating thrombin formation and coagulation [6]. Moreover, elevated microvesicles and activation of platelets and leukocytes have now also been associated with venous thromboembolism [7].

The risk of myocardial infarction has been found to increase substantially with acute lower respiratory tract infections, suggesting a link between lung inflammation and thrombogenicity [8]. Epidemiologic associations have shown a strong relationship also between myocardial infarction and lung exposure to particulate air pollution [9–11]. Experimental studies revealed that particulate-induced pulmonary inflammation leads to platelet activation in a time frame of 3–24 h, resulting in the development of a prothrombotic tendency, measurable experimentally 24 h after induction of mild controlled vascular injury in small animals [12–14]. Interestingly, a recent study, performed in the heavily polluted area around Milan and the remainder of Lombardy [15], found an association between air pollution and shortening of prothrombin times in exposed individuals, whereas such an association was lacking for the activated partial thromboplastin time. This study suggests the occurrence of hypercoagulability during air pollution, potentially linked to increased circulating tissue factor activity and/or factor VIIa activity [15,16].

In the present study, we therefore investigated how pulmonary inflammation caused by air pollutants would translate into a circulating prothrombotic tendency, enhancing the risk for thrombosis, in cases of peripheral vascular lesions. Ultrafine particles present in polluted air, such as diesel particles [17] and silica [14], are responsible for lung inflammation, but positively charged model particles [18] can also trigger lung inflammation. In the present study, rather than administering diesel exhaust particles, causing both direct and indirect platelet activation [12,17], we chose to intratracheally instill multiwall carbon nanotubes (CNTs) to induce mild and selective pulmonary inflammation, on the basis of recent toxicity studies [19,20]. CNTs are cylindrical carbon molecules that exhibit unusual strength, display unique electrical properties, and are efficient conductors of heat. They can be packed with DNA molecules or peptides, after which they can deliver their contents to specific target sites in various tissues for therapeutic purposes [21,22]. They are not, strictly speaking, air pollutants, but allow the isolated study of the relationship between lung inflammation and peripheral thrombogenicity.

We have investigated the role of P-selectin in this relationship. Leukocyte-derived microvesicle function was analyzed and platelet activation was studied in wild-type mice. We found that lung inflammation triggers rapid and transient circulating platelet activation, platelet–leukocyte conjugate formation, and the release by leukocytes of tissue factor-positive microvesicles. These findings therefore explain the transfer of pulmonary to systemic inflammation and provide an insight into the development of a thrombogenicity risk. They further suggest that the effects on blood coagulation after air pollution [15] rely, indeed, on elevated circulating tissue factor activity.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

Carbon nanotubes

Ground multiwall CNTs, consisting of 15 carbon layers on average, were a gift from J. B. Nagy (Nuclear Magnetic Resonance Laboratory at the Facultés universitaires Notre-Dame de la Paix in Namur, Belgium) and D. Lison (Unit of Industrial Toxicology and Occupational Medicine, Université Catholique de Louvain, Louvain, Belgium). Whereas the lengths of individual nanotubes were significantly reduced by grinding (0.7 ± 0.07 vs. 5.9 ± 0.05 μm), other characteristics of the material, such as average inner diameter (5.1 ± 2.1 vs. 5.2 ± 1.5 nm), average outer diameter (11.3 ± 3.9 vs. 9.7 ± 2.1 nm), specific surface area (307 ± 15 vs. 378 ± 20 m2 g−1), oxidized forms (13.1 ± 0.7 vs. 13.7 ± 0.7 atomic %) and carbon content (98.0% ± 0.2% vs. 97.8% ± 0.2%), were not affected by the grinding process.

CNTs were suspended in sterile pyrogen-free saline (NaCl 0.9%) containing Tween-80 (0.1%). To minimize their aggregation, CNT suspensions were always sonicated (Branson 1200, VEL, Leuven, Belgium) for 15 min and vortexed immediately (< 1 min) before their dilution and prior to intratracheal administration. Samples of the CNT suspension were also applied, immediately, to formvar/carbon-coated nickel grids (300 mesh). The liquid was then removed, and the grids were air-dried before examination in a Zeiss 902A electron microscope. Ground CNTs were also suspended in absolute ethanol, sonicated for 30 min, and examined.

Animals and intratracheal instillation of CNTs

This project was reviewed and approved by the Institutional Review Board of the University of Leuven, and experiments were performed in accordance with protocols approved by the Institutional Animal Care and Research Advisory Committee. Male and female Swiss mice weighing 45–50 g were used. They were anesthetized with sodium pentobarbital (60 mg kg−1, i.p.), and placed supine with neck extended on an angled board. A Becton Dickinson 24 gauge cannula was inserted via the mouth into the trachea. The CNT suspensions (200 or 400 μg per mouse) or vehicle only were instilled (40 μL) via a sterile syringe and followed by an air bolus of 50 μL.

Bronchoalveolar lavage (BAL) fluid analysis

Twenty-four hours following the intratracheal instillation of CNT or vehicle, mice were killed with an overdose of sodium pentobarbital. The trachea was cannulated, and lungs were lavaged three times with 0.7 mL (a total volume of 2.1 mL) of sterile NaCl 0.9%. The recovered fluid aliquots were pooled. No differences in the volume of collected fluid were observed between the different groups. BAL fluid was centrifuged (1000 × g for 10 min at 4 °C). Cells were counted in a Thoma hemocytometer after resuspension of the pellets and staining with 1% gentian violet. The cell differentials were microscopically performed on cytocentrifuge preparations fixed in methanol and stained with Diff Quick (Dade, Brussels, Belgium). The supernatant was stored at −80 °C until further analysis.

Granulocyte–platelet heteroconjugate detection

In separate animals, after intratracheal instillation with saline or saline containing CNTs (400 μg), blood samples were collected from the retroorbital sinus on citrate (0.4%), containing hirudin (20 μg mL−1), G4120 (1 μg mL−1), and prostin (1 μg mL−1). To determine the existence of heteroconjugates between granulocytes and platelets via flow cytometry, 1 mL of Cell Fix (Becton Dickinson, Franklin Lakes, NJ, USA) was immediately added to 200 μL of blood, and this was followed by vortexing. After 1 h of fixation at room temperature, samples were centrifuged for 5 min at 150 × g. One milliliter of phosphate-buffered saline (PBS) was added to the pellet, and after gentle vortexing, cells were centrifuged for 5 min at 150 × g to eliminate fixative. Thereafter, 1 mL of lysis buffer (0.15 m NH4Cl, 10 mm KHCO3, 0.1 mm Na2EDTA, pH adjusted between 7.2 and 7.4 with 1 m HCl) was added to the pellet, and after gentle vortexing, the cells were centrifuged for 5 min at 90 × g. The lysis step was then repeated. A washing step was performed by adding 1 mL of PBS to the pellet, and then centrifuging for 5 min at 150 × g. The pellet was then resuspended in 50 μL of PBS, and 50 μL of rat serum was added, together with 5 μL of the leukocyte common antigen Ly-5 marker, R-phycoerythrin-conjugated monoclonal rat antibody 30-F11 (Pharmingen, San Diego, CA, USA) ,and 5 μL of platelet αIIbβ3 fluorescein isothiocyanate-labeled rat anti-integrin CD41/61 antibody MO20 (Emfret, Würzburg, Germany). Following incubation in the dark at room temperature for 20–40 min, 0.5 mL of PBS was added, and sample acquisition on a FACSCalibur (Becton & Dickinson, San José, CA, USA) was performed immediately. Live gating was performed on leukocyte-sized events to exclude single platelets; granulocytes were selected on the basis of their forward and side scatter characteristics, as well as CD45 positivity. Gated CD45-positive granulocytes were then analyzed for the presence of bound platelets in histogram plots of platelet integrin intensity.

Thrombin generation and microvesicle activity detection

Thrombin generation curves were measured in 96-well microtiter plates via calibrated automated thrombinography (Thrombinoscope BV, Maastricht, The Netherlands). Each well contained 80 μL of resuspended microvesicles (see below), 10 μL of neutralizing monoclonal anti-FVIII antibody, BO2C11 (50 μg mL−1) [23], 10 μL of prothrombin complex concentrate, PPSB (prothrombin = FII, proconvertin = FVII, Stuart–Prower factor = FX, and antihemophilic globulin B = FIX; CLB, Amsterdam, The Netherlands), 10 μL of a synthetic phosphatidylserine (PS)/phosphatidylcholine (PC) (Avanti Polar Lipids) mixture (see below), and 20 μL of fluorogenic substrate Z-Gly-Gly-Arg-AMC (Bachem, Bubendorf, Switzerland), solubilized in pure dimethylsulfoxide. The fluorescence from cleaved amidomethyl coumarin was followed over time, and the first derivatives of these curves were calculated and converted to nanomolar thrombin concentrations by using internal calibrators [24]. PC and PS were dissolved in chloroform. These solutions (25 mg mL−1) were mixed in a PC/PS ratio equal to 7:3 and dried overnight in a glass tube. Thereafter, the phospholipid mixture was resuspended in tris-buffered saline (TBS), sonicated for 4 h, fractionated, and then stored at –20 °C until analysis. Synthetic phospholipids were found to stimulate the reaction, and were therefore always added (50 μm) to standardize the phospholipid content.

Microvesicles were isolated from murine plasma by mixing 400 μL of mouse plasma (pool of four mice, 100 μL each) per group and 3600 μL of previously ultracentrifuged, i.e. microvesicle-depleted, human plasma (4 h at 230 000 × g), and the mixed plasma was then ultracentrifuged at 4 °C for 4 h at 230 000 × g. The pellet consisting of murine microvesicles was then resuspended in 400 μL of Tyrode’s buffer (12 mm NaHCO3, 0.42 m NaH2PO4, 137 mm NaCl, 2.7 mm KCl, 1 mm MgCl2, 5 mm HEPES, 5.5 mm glucose, and 0.35% human serum albumin), and diluted 1- to 4-fold in the same buffer. Thrombin generation curves were measured for plasma pools, derived from all four study groups, i.e. instilled with saline or saline containing CNTs (400 μg per animal), including groups pretreated with the anti-P-selectin antibody RB.40.34 (60 μg per mouse i.v.). The reference thrombin generation curves were constructed with human recombinant tissue factor (Innovin, Dade Behring, Heidelberg, Germany). Tissue factor inhibition was performed with a polyclonal rabbit antitissue factor antibody (American Diagnostics, Hauppauge, NY, USA), which was found to selectively neutralize the tissue factor-mediated thrombin generation in our assays.

Whole murine blood platelet aggregation assays

Following induction of anesthesia, mice were bled from the retroorbital plexus. Blood was collected on citrate (0.4%) and pipetted into 48-well microtiter plates at 250 μL per well. Plates were then rotated at 900 r.p.m. in an enzyme-linked immunosorbent assay microtiter plate incubator, and, at fixed time intervals (0–3 min), 25-μL samples were removed and diluted tenfold in Cell Fix, upon which platelets were counted in a Cell Dyn 3000 (Abbott, Abbott Park, IL, USA). The percentage aggregation was then calculated, in comparison to the platelet count in non-agitated blood.

In vivo P-selectin neutralization

To assess the role of platelet P-selectin in CNT-induced lung inflammation and enhancement of peripheral thrombogenicity, mice were injected i.v. through the tail vein, with 60 μg of the anti-P-selectin blocking monoclonal antibody (RB40.34; Becton Dickinson, San Diego, CA, USA), and 10 min later, saline or saline containing CNTs (400 μg) was instilled intratracheally. Leukocyte–platelet conjugate detection, arterial thrombosis in vivo and polymorphonuclear neutrophil (PMN) influx into BAL were assessed as outlined above.

Experimental arterial thrombosis model

Twenty-four hours after intratracheal instillation of CNTs or saline, in vivo thrombogenesis was assessed using a technique employed previously in the femoral vessels of hamsters [18]. Following induction of anesthesia, mice were placed in a supine position on a heating pad at 37 °C, tracheotomized, and artificially ventilated (Hugo Sachs Apparatus Minivent type 845 respirator; Hugo Sachs Elektronik-Harvard Apparatus GmbH, March-Hugstetten, Germany). A 2 F venous catheter (Portex, Hythe, UK) was inserted into the right jugular vein for the administration of Rose Bengal. Thereafter, the left carotid artery was exposed from the surrounding tissue and mounted on a transilluminator. After i.v. administration of Rose Bengal (20 mg kg−1), the segment of the carotid artery was irradiated with green light (540 nm) for 90 s, using an optic fiber mounted on a micromanipulator located 5 mm above the artery. The thrombus was monitored under a microscope at × 40 magnification. The change over time in light transmission through the blood vessel at the site of the trauma was recorded using a microscope-attached camera. Images were recorded at intervals of 10 s over a period of 40 min, and thrombus intensity was quantified via image analysis. The size of the thrombus was expressed in arbitrary units (AUs) as the total area under the curve for plots of light intensity vs. time. The mice were killed at the end of the recording.

Tail and jugular vein bleeding time measurements

Following induction of anesthesia, as indicated above, 8- to 12-week-old mice were placed in a supine position, and a standardized fragment of the tail (0.5 cm) was clipped, following which the amputated tail was immersed in saline at 37 °C and the time till arrest of bleeding was determined. Alternatively, bleeding times were recorded in punctured jugular veins, which represent a milder model of vessel wall injury, independent of vessel dilatation. For these purposes, both jugular veins were surgically exposed and punctured with a 29 G needle at an angle of about 30°; the bleeding time was recorded by blotting leaking blood onto filter paper. When indicated, mice were i.p. injected with the NO donor sodium nitroprusside (SNP) (0.42-0.72 mg kg−1, i.p. injection at –10 min), arginine (Arg) (80 μmol kg−1, i.p. injection at –10 min), the NO inhibitor l-NAME (80 μmol kg−1, i.p. injection at –10 min), or aspirin (Aspégic, Synthélabo, Brussels, Belgium) (40 μmol kg−1, i.p. injection, prior to the jugular vein bleeding time measurement, as specified). Contributions to the recorded bleeding time of P-selectin were evaluated by the separate i.p. injection of 15 mg kg−1 of the flow-dependent P-selectin inhibitor gallic acid (GA) [25] (Sigma-Aldrich, St Louis, MO, USA) 5 min prior to the jugular bleeding time measurement.

Statistics

Data are expressed as means ± SEM. Comparisons between groups were performed by one-way analysis of variance (anova), followed by Newman–Keuls or Tukey–Kramer multiple range tests, or by 2 × 2 column comparison, via a non-parametric Mann–Whitney two-tailed test. P-values < 0.05 were considered to be significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

CNTs trigger mild lung inflammation

Most individual CNTs were < 10 μm long, and approximately 10 nm in diameter (Fig. 1A,B). About 5% of them were approximately 5 nm in diameter, and some fragments of these CNTs were as short as 70 nm; these dimensions preclude migration across the epithelial barrier. Cells found in BAL, 24 h after CNT administration, were primarily macrophages and PMNs (Fig. 1C,D). Lymphocytes were absent in control mouse BAL. Furthermore, Diff-Quik staining of the cells in BAL fluid revealed that CNTs are taken up by macrophages, but aggregates of non-ingested CNTs were also found (Fig. 1D). At 200 and 400 μg per mouse, the CNT intratracheal instillation resulted in a clear cellular influx in the lung. Whereas macrophage numbers did not increase (Fig. 1E), PMN numbers increased 8-fold at 200 μg per mouse (P < 0.05) and twelvefold at 400 μg per mouse (< 0.05) (Fig. 1F). These findings showed that CNTs trigger mild lung inflammation, restricted to pulmonary neutrophil infiltration, when analyzed at 24 h, characteristic of sustained lung inflammation in previous studies of ultrafine particle-induced lung inflammation. This mild inflammation model therefore facilitates the study of systemic monocyte activation in the absence of major macrophage accumulation in the lung, such as encountered in these studies [12]. Therefore, the relationship between lung inflammation and its thrombogenic consequences was studied in this particulate matter lung model.

image

Figure 1.  Carbon nanotube (CNT)-induced lung inflammation. Transmission electron microscopy of a sample of ground CNTs, dispersed in ethanol and dried-down onto a formvar/carbon film (bar = 250 nm) (A, B) Diff-Quik stain of the cells recovered in bronchoalveolar lavage fluid (BAL) of mice 24 h after intratracheal instillation of saline (C) or CNTs (400 μg per mouse) (D); cells in the saline control group consist mainly of macrophages; in the CNT group, CNT-laden macrophages (black arrowheads), cell-free aggregates of CNTs (red arrowheads) and polymorphonuclear neutrophils (PMNs) (arrows) are found. (E) Numbers of macrophages and (F) numbers of PMNs in BAL fluid 24 h after intratracheal instillation of the indicated doses of CNTs. Data are mean ± SEM (n = 5–6). Statistical analysis is by one-way anova followed by Newman–Keuls multiple-comparison test.

Download figure to PowerPoint

Lung inflammation and bleeding time measurements

The intratracheal administration of diesel exhaust [12,17] or silica [14] particles leads to platelet activation. Compared to non-instilled control mice, the intratracheal instillation of saline hardly affected the tail bleeding time, measured 3, 6 and 24 h later (Fig. 2A). The instillation of CNTs transiently prolonged the bleeding time at 3 h, but prolonged it more pronouncedly at 24 h (Fig. 2A). Tail bleeding times depend on both platelet and vascular smooth muscle cell function; we therefore also determined jugular vein bleeding times, which are characterized by milder injury and are less dependent on vasoconstriction or dilatation. Fig. 2B shows how elevation of NO (chemically via SNP or biologically via Arg) partially inhibits platelet function, i.e. slows bleeding arrest. Conversely, l-NAME accelerated bleeding arrest, as a result of enhanced platelet activation. The P-selectin inhibitor GA, by shielding P-selectin-mediated platelet–endothelium interactions, also retards bleeding arrest (Fig. 2B). Aspirin, injected 2 or 7 h (no difference was seen between 2 and 7 h; not shown) prior to the bleeding time measurement prolonged the bleeding time from a mean of 14.6 ± 2.4 s (control) to 51.4 ± 8.9 s, showing that the jugular vein bleeding arrest strongly depends on platelet thromboxane A2 (TxA2).

image

Figure 2.  Bleeding times after carbon nanotube (CNT) administration. Clipped tail (A) and jugular vein (C, D) bleeding time upon intratracheal instillation in mice of saline or 400 μg per mouse CNTs, measured at 3, 6 and 24 h, as indicated. (B) Jugular vein bleeding time of mice treated with sodium nitroprusside (i.p. bolus of 0.24 mg kg−1 at – 10 min), Arg (i.p. bolus of 80 μmol kg−1 at – 10 min), gallic acid (i.p. bolus of 15 mg kg−1 at – 10 min), aspirin (i.p. bolus of 40 μmol kg−1 at – 2 h), and l-NAME (i.p. bolus of 80 μmol kg−1 at – 10 min). (D) Jugular vein bleeding time upon intratracheal instillation in mice of saline or 400 μg per mouse CNTs, measured under aspirin protection, i.e. after i.p. injection with 40 μmol kg−1 aspirin, 15 min before the bleeding time measurement. C: Non-treated control tail (A) and jugular vein (C) bleeding times; statistical analysis by 2 × 2 column comparison via a two-tailed Mann–Whitney assay.

Download figure to PowerPoint

Fig. 2C shows that the jugular vein bleeding time is transiently elevated at 3 and 6 h, for both saline and nanotube administration. At 24 h, the bleeding time normalizes in the case of saline (to 16 ± 2.5 s), but is further prolonged for the nanotube administration (to 42.6 ± 5.4 s). Therefore, at 24 h, both the prolonged tail and jugular vein bleeding times reflect CNT-altered platelet–vessel wall interactions, but they show no sign of platelet activation, i.e. shortened bleeding times. However, in view of the major role of TxA2 in the jugular vein bleeding, we have also performed measurements under protection with aspirin, eradicating this pathway. Fig. 2D shows that, when the bleeding time was measured under aspirin protection (injected 15 min before recording of the bleeding time), saline prolonged the bleeding time to the expected values (to 46.1 ± 10.4 s at 3 h, and to 42.9 ± 7.6 s at 6 h). The instillation of CNTs shortened these bleeding times (to 33.5 ± 12 s), slightly but significantly, when measured at 3 h, but not at 6 h (40 ± 9.6 s). This in vivo analysis showed that CNT-induced lung inflammation gives rise to a mild, systemic inflammatory condition, but, at best, causes only mild and transient platelet activation.

Additional ex vivo whole blood platelet aggregations, induced by 0.5 μm adenosine 5′-diphosphate (ADP), yield 14.5% ± 11% aggregation (n = 3) at 1 min, for blood taken from non-treated Swiss mice. The instillation procedure, however, influenced aggregation 24 h later: upon intratracheal saline administration, aggregation by 0.5 μm ADP, at 1 min, was 72% ± 22% (n = 6); upon intratracheal CNT administration, aggregation by 0.5 μm ADP, at 1 min, was 60% ± 27% (n = 5). Hence, ex vivo platelet aggregation was not more informative with regard to platelet activation status in vivo than the bleeding time.

Activated platelets form platelet–granulocyte heteroconjugates

To more convincingly demonstrate platelet activation, the flow cytometric detection of circulating conjugates between platelets and granulocytes was carried out. These conjugates have a longer half-life than free platelets and are detected more easily. Fig. 3A,B shows how granulocytes were gated in the forward scatter (FSc) vs. side scatter (SSc) and the SSc vs. CD45 scatter diagrams for murine blood samples. The cross-section of R1 (Fig. 3B) and R2 (Fig. 3A) was defined to select CD45-positive granulocytes, excluding monocytes and free platelets. Fig. 3C shows the CD41/61 distribution in the selected granulocyte cross-sectional gate, and illustrates the heteroconjugates between platelets and granulocytes. Heteroconjugates were only found 6 h after CNT administration (Fig. 3D). Indeed, in comparison to the numbers of conjugates in non-treated controls, set to 100%, at 3 h, conjugate numbers equaled 104% ± 48% of the control value in the saline control group, and 109% ± 43% in the CNT group (n = 3–4). Likewise, at 12 h, these values were 110% ± 30% in the saline control group, and 100% ± 30% in the CNT group (n = 4); at 24 h, they were 106% ± 23% in the saline control group, and 98% ± 33% in the CNT group (n = 3). Fig. 3D shows that 6 h after instillation, platelet–leukocyte conjugate numbers were 130% ± 9.4% in the saline control group and 282% ± 55% in the CNT group (P < 0.05, n = 6). These numbers confirm gradual platelet activation and conjugate formation in the hours preceding the 6-h analysis point. This phenomenon is itself transient, because activated platelets and their leukocyte conjugates have disappeared from the circulation 12 h after CNT administration.

image

Figure 3.  Granulocyte–platelet conjugate induction by carbon nanotubes (CNTs): role of P-selectin. (A) Forward scatter (FSc) vs. side scatter (SSc) distribution during flow cytometry of blood cells from mice intratracheally instilled with 400 μg of CNTs. (B) SSc vs. CD45 scatter diagram; CD45-positive granulocytes were back-gated via the cross-section of R1 and R2. (C) CD41/61 distribution in the selected gate shows the presence of heteroconjugates between granulocytes and platelets. (D) Percentage heteroconjugates in non-treated mice (range between 0.52% and 0.86%; the mean value of 0.71 ± 0.07 is set to 100%), assessed 6 h after intratracheal instillation of saline or CNTs (400 μg per mouse); groups treated with the P-selectin-neutralizing antibody RB.40.34 are as indicated. Data are mean ± SEM. Statistical analysis: unpaired Student’s t-test.

Download figure to PowerPoint

Fig. 3D also shows that treatment of mice with a neutralizing anti-P-selectin antibody (RB.40.34), given at the start, was capable of abrogating platelet–leukocyte conjugate formation in the CNT group 6 h later, and of reducing heteroconjugate formation, even in the saline-treated groups, to levels below those found in non-treated controls. These findings showed that heteroconjugate formation in the mouse circulation was platelet P-selectin-mediated.

CNTs enhance experimental arterial thrombosis: role of P-selectin

The development of a systemic prothrombotic tendency upon CNT administration was analyzed 24 h later in a mouse thrombosis model. Fig. 4A illustrates a prothrombotic tendency via an increase in the cumulative mass of thrombus formed in vivo over a 40-min interval in the photochemically injured carotid artery. CNT stimulated arterial thrombosis, both at 200 μg per animal (+ 245%, P < 0.05) and 400 μg per animal (+ 350%, P < 0.01) (Fig. 4A). The antibody-mediated P-selectin neutralization in mice did not affect the number of platelets, either in saline-treated mice (659 200 ± 55 766 μL−1 in antibody-treated mice vs. 641 250 ± 74 137 μL−1 in non-treated mice), or in CNT-treated mice (718 800 ± 66 917 μL−1 in antibody-treated mice vs. 715 000 ± 158 406 μL−1 in non-treated mice). Furthermore, in our model, 24 h after injection at 60 μg per mouse, the anti-P-selectin antibody did not affect the weak lesion-induced acute thrombotic response recorded in saline-treated mice, which is indicative of a minor role for platelet P-selectin in these mild injury conditions, i.e. with little P-selectin-mediated recruitment of circulating tissue factor-positive microvesicles (see below). In contrast, the CNT amplification of thrombogenesis was largely eliminated (Fig. 4A). These findings illustrate the role of platelet P-selectin in the systemic prothrombotic sensitization by CNTs.

image

Figure 4.  P-selectin in carbon nanotube (CNT)-induced thrombosis and lung inflammation. Cumulative thrombus size, expressed as total light intensity over 40 min [in arbitrary units (AUs)], after mild photochemical damage to the endothelium of the left carotid artery, 24 h after intratracheal instillation of saline or CNTs (200 or 400 μg per mouse, as indicated). (A) Mice instilled with saline or CNTs (400 μg per mouse) were treated with the anti-P-selectin antibody RB.40.34 (60 μg per mouse i.v.) as indicated. (B) Numbers of polymorphonuclear neutrophils in bronchoalveolar lavage fluid 24 h after intratracheal instillation of saline or CNTs (400 μg per mouse) in mice pretreated with the anti-P-selectin antibody RB.40.34 (60 μg per mouse i.v.) as indicated. Data are mean ± SEM (n = 4–8 in each group). Statistical analysis: one-way anova followed by Newman–Keuls multiple-comparison test.

Download figure to PowerPoint

Systemic inflammation is accompanied by microvesicle formation: role of tissue factor

We next investigated whether P-selectin neutralization by RB.40.34 could influence lung inflammation 24 h later. P-selectin neutralization had no effect on the number of macrophages in BAL (after saline, 8.4 ± 2.7 × 105 mL−1; after saline + RB.40.34, 10.4 ± 1.5 × 105 mL−1; after CNT, 8.5 ± 2.7 × 105 mL−1; after CNT + RB.40.34, 9.9 ± 3.2 × 105 mL−1). Although granulocyte inflammation showed a trend towards lower numbers in RB.40.34-treated animals (Fig. 4B), granulocyte inflammation was largely P-selectin-independent (see also Discussion).

To investigate tissue factor production during platelet-induced systemic inflammation, murine plasma from control mice and CNT-treated mice was subjected to ultracentrifugation and plasma microvesicles were collected. Plasma was also collected from saline-treated and CNT-treated mice subjected to P-selectin neutralization. Fig. 5A shows thrombin generation curves as a function of the concentration of murine plasma microvesicles, resuspended in a coagulation factor concentrate, in the presence of negatively charged phospholipids and in the additional presence of a human antibody (BO2C11) that completely neutralizes both human and murine FVIII. Therefore, the thrombin generation curves shown reflect extrinsic coagulation initiated by microvesicular tissue factor in the absence of contributions from the intrinsic coagulation cascade. The further elimination of the BO2C11-resistant thrombin generation by a neutralizing antitissue factor antibody confirmed the specificity of microvesicular tissue factor during thrombin generation (not shown). During the measurement of thrombin generation in the presence of PC/PS, PPSB and BO2C11 for murine resuspended microvesicles, we found that the CNT (400 μg per animal) treatment elevated the circulating microvesicular tissue factor as compared to saline-treated mice. P-selectin neutralization did not affect thrombin generation in saline-treated mouse-derived plasma, but the CNT enhancement was abrogated (Fig. 5B). These findings confirmed enhanced production of microvesicular tissue factor by monocytes, and potentially also by P-selectin-stimulated granulocytes [26].

image

Figure 5.  Microvesicular tissue factor-induced thrombin generation. Individual thrombin generation curves, upon recalcification, in PPSB and standardized phosphatidylcholine (PC)/phosphatidylserine (PS) phospholipid mixtures, after addition of decreasing amounts of murine plasma microvesicular tissue factor (TF), isolated via ultracentrifugation from fresh murine plasma and resuspension in an equal volume of Tyrode buffer. (A) The formal microvesicle (MV) dilution ranged from 3 to 12, as indicated; the FVIII-neutralizing antibody BO2C11 (50 μg mL−1) was added in all conditions. (B) Microvesicular murine TF-induced thrombin generation for frozen plasma isolated microvesicles, sampled from mice 24 h after intratracheal administration of saline or carbon nanotubes (CNTs) (400 μg per mouse) and treated with the anti-P-selectin antibody RB.40.34 as indicated; formal plasma microvesicle dilutions are 6-fold. Baseline indicates recalcification without addition of microvesicles.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

In this study, we found that pulmonary inflammation causes rapid, but mild and transient, circulating platelet activation, triggering P-selectin-mediated platelet–leukocyte conjugate formation. We observed leukocyte activation and microvesicular tissue factor production. Our study therefore shows that particulate pollutant lung inflammation increases the risk for thrombosis, via platelet P-selectin exposure and systemic leukocyte activation.

The industrial use of CNTs comprises applications such as inclusion in polymers or other matrices, applications for which ground instead of intact nanotubes are used. Grinding affects the length of individual nanotubes, but other characteristics, such as average diameter, specific surface area, oxidation state, and carbon content, are not affected. Because ground CNTs, like intact CNTs, cause dose-dependent and diffuse lung inflammation and fibrosis lasting up to 60 days [20], we determined whether CNTs give rise to pulmonary and systemic inflammation, and whether this leads to an enhanced risk for arterial thrombosis, as we had previously found for exposure to diesel exhaust particles [17] and silica [14]. In mice, CNTs are taken up by macrophages and cause a dose-dependent increase of neutrophil influx in BAL fluid, similar to what has been reported in rats, 3 days after intratracheal administration of intact or ground CNTs [20]. We found, however, that CNTs in mice were not markedly toxic. Yet, even though CNTs do not constitute a strict particulate air pollutant themselves, they are useful for the study of the extrapulmonary effects of such agents, because CNTs trigger mild and selective neutrophil inflammation in the lung, without direct macrophage activation, and do not show transepithelial passage into the circulation, as observed for other ultrafine particles, thus avoiding direct interactions with circulating cells.

The present study has enabled us to understand the kinetics of the transition of pulmonary inflammation to the systemic circulation in a mouse model of thrombosis. Potentially as early as 3 h following induction of lung inflammation by CNTs, and definitely at 6 h, platelets are activated mildly and transiently to a sufficient extent to trigger the formation of platelet–leukocyte conjugates that are absent in saline-treated controls. Hence, mild systemic inflammation induction occurred with kinetics similar to those reported after lung infiltration with diesel exhaust particles, assessed by the appearance of histamine in the circulation [12]. Recent evidence suggests that exposure of endothelial cells to PM10 (particles with a size limit of 10 μm) is capable of inducing E-selectin expression, as part of a proinflammatory cascade leading to leukocyte recruitment [27]. We have shown before that platelet rolling over activated endothelium is associated with mild platelet activation and adhesion, in part (but not exclusively) mediated via P-selectin and von Willebrand factor released from the Weibel–Palade bodies of activated endothelial cells [28]. It is therefore not surprising that pulmonary endothelial cell activation enables the CNT-induced recruitment of neutrophils, but also stimulates platelet rolling over the activated pulmonary endothelium [28], events leading to rapid but transient platelet activation, with an optimum between 3 and 6 h after the start of inflammation. The apparent reversibility in the formation of platelet–leukocyte conjugates, with a maximum around 6 h but an absence after 12 h, could be the result of margination of these complexes and/or of P-selectin cleavage from the platelet membrane [3]. Indeed, activated neutrophils release enzymes such as elastase and cathepsin G, which can degrade P-selectin glycoprotein ligand-1 (PSGL-1) [29], contributing to platelet dissociation from leukocytes. Irrespective of their observed disappearance, platelet–leukocyte conjugates reached higher numbers in CNT-treated mice than in saline-treated controls, in a time window between 3 and 6 h, which has been shown in other studies to convert lung inflammation into a systemic reaction, and to be coupled to generation of a prothrombotic effect [12,13,18].

P-selectin is found in storage α-granules of platelets and Weibel–Palade bodies of endothelial cells, from where it can be released upon cell activation, ending up in the outer membrane. Endothelial P-selectin initiates the capture and rolling of circulating leukocytes at sites of inflammation and atherosclerosis [30,31]. Platelet P-selectin is responsible for the initial rolling of leukocytes along the inflamed endothelium [28,31–34], interactions largely mediated by binding to PSGL-1 expressed on leukocytes [32,35]. Our findings illustrate that the development of a prothrombotic tendency in the circulation is entirely dependent on the availability of P-selectin. The functional neutralization of P-selectin not only abrogated platelet–leukocyte conjugate formation, but also, by eliminating the systemic inflammation, caused the CNT-induced thrombogenicity to disappear.

In the context of allergic reactions, in vivo models of inflammatory reactions in the skin [36] and the peritoneal cavity [37] demonstrated that P-selectin plays an important role in allergic inflammation, sooner (3–12 h) rather than later (20–24 h) after allergen challenge. In contrast, Pitchford et al. [38] recently reported a role for P-selectin at both 8 and 24 h after allergen challenge. In the present study, PMN influx in BAL was only slightly, if at all, affected by P-selectin neutralization. Whereas P-selectin on the platelet membrane is a major requirement for pulmonary eosinophil and lymphocyte recruitment, allowing circulating platelets to bind to and stimulate leukocytes for endothelial attachment [38], the cells presently identified comprised neutrophils exclusively: no lymphocytes or eosinophils were found in BAL of mice exposed to CNTs. This may explain the minor role of P-selectin in lung inflammation in the present study.

Previous work had shown that anti-inflammatory treatment could prevent particulate matter-induced lung inflammation, thus abrogating the development of a prothrombotic tendency [12,13]. Therefore, the present elimination of arterial thrombogenicity amplification by CNTs, at 24 h, by P-selectin neutralization, while the expected CNT-induced lung inflammation was maintained, confirmed a causative role for P-selectin in thrombogenicity induction at a later stage than that of lung inflammation itself. The presence of P-selectin on activated circulating platelets between 3 and 6 h after CNT administration, the subsequent P-selectin-mediated inflammatory response in platelet–leukocyte conjugates and the abolition of thrombogenicity amplification following P-selectin neutralization therefore suggest a causative link between these events.

In previous studies, particulate pollutant-induced thrombogenicity induction was mostly assessed by detection of activated platelets. In the present, mild, model, and despite clearly upregulated thrombogenicity 24 h after CNT challenge, no clear evidence was found for an activated circulating platelet pool: at 24 h, platelet–leukocyte conjugates were absent, and tail bleeding times were not shortened, but prolonged. Although we cannot rule out modest platelet activation at this stage, it seems that the inflammatory microvesicular tissue factor itself is responsible for enhancing thrombogenesis [7]. Indeed, tissue factor-positive microvesicles can circulate in an inactive form, i.e. with encrypted tissue factor, without activating circulating platelets, but they can quickly transfer to a growing thrombus by binding onto platelet P-selectin in a thrombus, where tissue factor will be decrypted upon membrane fusion [39]. As a consequence, thrombin formation may participate in additional local platelet activation and coagulation induction, both of which processes enhance thrombosis.

In conclusion, our data provide novel evidence that lung inflammation induced by CNTs rapidly activates platelets transiently, thus translating the inflammatory reaction to the circulation, via P-selectin-mediated platelet–leukocyte conjugation. The resulting thrombogenicity amplification involves P-selectin-dependent mechanisms, comprising platelets and tissue factor-bearing microparticles, whose contributions are of different magnitudes when analyzed at different time points. Microvesicular tissue factor may, then, be responsible for the link between air pollution and blood coagulation [15].

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

We thank J. B. Nagy, A. Fonseca and N. Moreau for providing the CNTs. The technical assistance of S. Van kerckhoven in platelet aggregation studies, K. Cludts in bleeding time measurements, H. Vanhooren in instillation and T. Smith in the preparation of electron microscopy samples is highly appreciated. J. Vermylen, D. Lison and J. B. Nagy are acknowledged for critically reading the manuscript.

Disclosure of Conflict of Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References

This work was supported by K. U. Leuven projects OT/02/45 and GOA/2004/09, and by the Fund for Scientific Research Flanders (G.0165.03). The Center for Molecular and Vascular Biology is supported by the ‘Excellentie financiering KU Leuven’ (EF/05/013).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interests
  9. References
  • 1
    Huo Y, Schober A, Forlow SB, Smith DF, Hyman MC, Jung S, Littman DR, Weber C, Ley K. Circulating activated platelets exacerbate atherosclerosis in mice deficient in apolipoprotein E. Nat Med 2003; 9: 617.
  • 2
    Gawaz M, Langer H, May AE. Platelets in inflammation and atherogenesis. J Clin Invest 2005; 115: 337884.
  • 3
    Wagner DD, Burger PC. Platelets in inflammation and thrombosis. Arterioscler Thromb Vasc Biol 2003; 23: 21317.
  • 4
    Giesen PL, Rauch U, Bohrmann B, Kling D, Roque M, Fallon JT, Badimon JJ, Himber J, Riederer MA, Nemerson Y. Blood-borne tissue factor: another view of thrombosis. Proc Natl Acad Sci USA 1999; 96: 23115.
  • 5
    Muller I, Klocke A, Alex M, Kotzsch M, Luther T, Morgenstern E, Zieseniss S, Zahler S, Preissner K, Engelmann B. Intravascular tissue factor initiates coagulation via circulating microvesicles and platelets. FASEB J 2003; 17: 4768.
  • 6
    Furie B, Furie BC. Role of platelet P-selectin and microparticle PSGL-1 in thrombus formation. Trends Mol Med 2004; 10: 1718.
  • 7
    Chirinos JA, Heresi GA, Velasquez H, Jy W, Jimenez JJ, Ahn E, Horstman LL, Soriano AO, Zambrano JP, Ahn YS. Elevation of endothelial microparticles, platelets, and leukocyte activation in patients with venous thromboembolism. J Am Coll Cardiol 2005; 45: 146771.
  • 8
    Smeeth L, Thomas SL, Hall AJ, Hubbard R, Farrington P, Vallance P. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004; 351: 26118.
  • 9
    Peters A, Dockery DW, Muller JE, Mittleman MA. Increased particulate air pollution and the triggering of myocardial infarction. Circulation 2001; 103: 28105.
  • 10
    Peters A, Von Klot S, Heier M, Trentinaglia I, Hormann A, Wichmann HE, Lowel H. Exposure to traffic and the onset of myocardial infarction. N Engl J Med 2004; 351: 172130.
  • 11
    Vermylen J, Nemmar A, Nemery B, Hoylaerts MF. Ambient air pollution and acute myocardial infarction. J Thromb Haemost 2005; 3: 195561.
  • 12
    Nemmar A, Nemery B, Hoet PH, Vermylen J, Hoylaerts MF. Pulmonary inflammation and thrombogenicity caused by diesel particles in hamsters: role of histamine. Am J Respir Crit Care Med 2003; 168: 136672.
  • 13
    Nemmar A, Hoet PH, Vermylen J, Nemery B, Hoylaerts MF. Pharmacological stabilization of mast cells abrogates late thrombotic events induced by diesel exhaust particles in hamsters. Circulation 2004; 110: 16707.
  • 14
    Nemmar A, Nemery B, Hoet PH, Van Rooijen N, Hoylaerts MF. Silica particles enhance peripheral thrombosis: key role of lung macrophage–neutrophil cross-talk. Am J Respir Crit Care Med 2005; 171: 8729.
  • 15
    Baccarelli A, Zanobetti A, Martinelli I, Grillo P, Hou L, Giacomini S, Bonzini M, Lanzani G, Mannucci PM, Bertazzi PA, Schwartz J. Effects of exposure to air pollution on blood coagulation. J Thromb Haemost 2007; 5: 25260.
  • 16
    Vermylen J, Hoylaerts MF. The procoagulant effects of air pollution. J Thromb Haemost 2007; 5: 2501.
  • 17
    Nemmar A, Hoet PH, Dinsdale D, Vermylen J, Hoylaerts MF, Nemery B. Diesel exhaust particles in lung acutely enhance experimental peripheral thrombosis. Circulation 2003; 107: 12028.
  • 18
    Nemmar A, Hoylaerts MF, Hoet PH, Dinsdale D, Smith T, Xu H, Vermylen J, Nemery B. Ultrafine particles affect experimental thrombosis in an in vivo hamster model. Am J Respir Crit Care Med 2002; 166: 9981004.
  • 19
    Shvedova AA, Kisin ER, Mercer R, Murray AR, Johnson VJ, Potapovich AI, Tyurina YY, Gorelik O, Arepalli S, Schwegler-Berry D, Hubbs AF, Antonini J, Evans DE, Ku BK, Ramsey D, Maynard A, Kagan VE, Castranova V, Baron P. Unusual inflammatory and fibrogenic pulmonary responses to single-walled carbon nanotubes in mice. Am J Physiol Lung Cell Mol Physiol 2005; 289: L698708.
  • 20
    Muller J, Huaux F, Moreau N, Misson P, Heilier JF, Delos M, Arras M, Fonseca A, Nagy JB, Lison D. Respiratory toxicity of multi-wall carbon nanotubes. Toxicol Appl Pharmacol 2005; 207: 22131.
  • 21
    Guo M, Chen J, Liu D, Nie L, Yao S. Electrochemical characteristics of the immobilization of calf thymus DNA molecules on multi-walled carbon nanotubes. Bioelectrochemistry 2004; 62: 2935.
  • 22
    He P, Bayachou M. Layer-by-layer fabrication and characterization of DNA-wrapped single-walled carbon nanotube particles. Langmuir 2005; 21: 608692.
  • 23
    Jacquemin MG, Desqueper BG, Benhida A, Vander Elst L, Hoylaerts MF, Bakkus M, Thielemans K, Arnout J, Peerlinck K, Gilles JG, Vermylen J, Saint-Remy JM. Mechanism and kinetics of factor VIII inactivation: study with an IgG4 monoclonal antibody derived from a hemophilia A patient with inhibitor. Blood 1998; 92: 496506.
  • 24
    Nieuwenhuys CM, Feijge MA, Beguin S, Heemskerk JW. Monitoring hypocoagulant conditions in rat plasma: factors determining the endogenous thrombin potential of tissue factor-activated plasma. Thromb Haemost 2000; 84: 104551.
  • 25
    Appeldoorn CC, Bonnefoy A, Lutters BC, Daenens K, Van Berkel TJ, Hoylaerts MF, Biessen EA. Gallic acid antagonizes P-selectin-mediated platelet–leukocyte interactions: implications for the French paradox. Circulation 2005; 111: 10612.
  • 26
    Maugeri N, Brambilla M, Camera M, Carbone A, Tremoli E, Donati MB, De Gaetano G, Cerletti C. Human polymorphonuclear leukocytes produce and express functional tissue factor upon stimulation. J Thromb Haemost 2006; 4: 132330.
  • 27
    Alfaro-Moreno E, Lopez-Marure R, Montiel-Davalos A, Symonds P, Osornio-Vargas AR, Rosas I, Clifford Murray J. E-Selectin expression in human endothelial cells exposed to PM(10): the role of endotoxin and insoluble fraction. Environ Res 2007; 103: 2218.
  • 28
    Theilmeier G, Michiels C, Spaepen E, Vreys I, Collen D, Vermylen J, Hoylaerts MF. Endothelial von Willebrand factor recruits platelets to atherosclerosis-prone sites in response to hypercholesterolemia. Blood 2002; 99: 448693.
  • 29
    Gardiner EE, De Luca M, McNally T, Michelson AD, Andrews RK, Berndt MC. Regulation of P-selectin binding to the neutrophil P-selectin counter-receptor P-selectin glycoprotein ligand-1 by neutrophil elastase and cathepsin G. Blood 2001; 98: 14407.
  • 30
    Ramos CL, Huo Y, Jung U, Ghosh S, Manka DR, Sarembock IJ, Ley K. Direct demonstration of P-selectin- and VCAM-1-dependent mononuclear cell rolling in early atherosclerotic lesions of apolipoprotein E-deficient mice. Circ Res 1999; 84: 123744.
  • 31
    Robinson SD, Frenette PS, Rayburn H, Cummiskey M, Ullman-Cullere M, Wagner DD, Hynes RO. Multiple, targeted deficiencies in selectins reveal a predominant role for P-selectin in leukocyte recruitment. Proc Natl Acad Sci USA 1999; 96: 114527.
  • 32
    Theilmeier G, Lenaerts T, Remacle C, Collen D, Vermylen J, Hoylaerts MF. Circulating activated platelets assist THP-1 monocytoid/endothelial cell interaction under shear stress. Blood 1999; 94: 272534.
  • 33
    Yokoyama S, Ikeda H, Haramaki N, Yasukawa H, Murohara T, Imaizumi T. Platelet P-selectin plays an important role in arterial thrombogenesis by forming large stable platelet–leukocyte aggregates. J Am Coll Cardiol 2005; 45: 12806.
  • 34
    Diacovo TG, Roth SJ, Buccola JM, Bainton DF, Springer TA. Neutrophil rolling, arrest, and transmigration across activated, surface-adherent platelets via sequential action of P-selectin and the beta 2-integrin CD11b/CD18. Blood 1996; 88: 14657.
  • 35
    McEver RP, Cummings RD. Role of PSGL-1 binding to selectins in leukocyte recruitment. J Clin Invest 1997; 100: S97103.
  • 36
    Teixeira MM, Hellewell PG. Contribution of endothelial selectins and alpha 4 integrins to eosinophil trafficking in allergic and nonallergic inflammatory reactions in skin. J Immunol 1998; 161: 251623.
  • 37
    Jia GQ, Gonzalo JA, Hidalgo A, Wagner D, Cybulsky M, Gutierrez-Ramos JC. Selective eosinophil transendothelial migration triggered by eotaxin via modulation of Mac-1/ICAM-1 and VLA-4/VCAM-1 interactions. Int Immunol 1999; 11: 110.
  • 38
    Pitchford SC, Momi S, Giannini S, Casali L, Spina D, Page CP, Gresele . Platelet P-selectin is required for pulmonary eosinophil and lymphocyte recruitment in a murine model of allergic inflammation. Blood 2005; 105: 207481.
  • 39
    Furie B. P-selectin and blood coagulation: it’s not only about inflammation any more. Arterioscler Thromb Vasc Biol 2005; 25: 8778.