- Email: [email protected]
Lipopolysaccharide augments venous and arterial thrombosis in the mouse
Thrombosis Research (2008) 123, 355–360
www.elsevier.com/locate/thromres
REGULAR ARTICLE
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Xinkang Wang ⁎ Discovery Translational Medicine, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA Received 31 October 2007; received in revised form 11 February 2008; accepted 13 March 2008 Available online 29 April 2008
KEYWORDS Arterial thrombosis; Venous thrombosis; LPS; Inflammation; Mice
Abstract Background: Animal models of diseases are essential for therapeutic target validation, drug discovery and development. Increasing evidence has demonstrated the importance of inflammation in thrombosis. Here, murine models of vena cava thrombosis and carotid arterial thrombosis augmented by lipopolysaccharide (LPS) were established and characterized to study the association between inflammation and thrombosis. Materials and methods: Murine (C57BL/6 mice) models of ferric chloride (FeCl3)induced carotid arterial and vena cava thrombosis were established. Thrombus formation was measured indirectly by Doppler blood flow (i.e., clot functional interference with blood flow) in the arterial thrombosis model and directly by protein content of the clot in the venous thrombosis model. An optimal concentration of FeCl3 was defined to induce thrombus formation and used to study the effects of LPS (i.e., a well-known inflammatory stimulus under these conditions). Real-time polymerase chain reaction (PCR) was used to examine the effect of LPS on TNFα and IL-1β mRNA expression in thrombus formation. Results: Dose-dependent analysis demonstrated that 2 mg/kg, i.p., LPS provided a maximal prothrombotic effect in 2.5% ferric chloride-induced vena cava thrombosis, with a 60% increase in thrombus size (n = 8, p b 0.05) compared to vehicle treatment. In contrast, 2 mg/kg LPS had no significant effect on thrombus formation in a more severe, 3.5% FeCl3-induced vena cava thrombosis. A similar prothrombotic effect was observed for LPS in 2.5% FeCl3-induced carotid arterial thrombosis model. Treatment of 2 mg/kg LPS significantly augmented arterial thrombosis immediately (between 5– 30 minutes) following FeCl3 injury as assessed by change of Doppler blood flow (n = 8, p b 0.05). Real-time PCR demonstrated significant induction of TNFα and IL-1β mRNA expression in the thrombus formation in the vessels in response to LPS challenge.
⁎ Tel.: +1 484 865 9906; fax: +1 484 865 9402. E-mail address: [email protected]. 0049-3848/$ - see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2008.03.015
356
X. Wang Conclusion: These data demonstrate that LPS augments thrombus formation in acute vascular injury and that LPS-augmented thrombosis might be a useful tool to study the relationship between inflammation and thrombosis. © 2008 Elsevier Ltd. All rights reserved.
Introduction Formation of a thrombus in the coronary or cerebral arteries is a major cause of morbidity and mortality throughout the world. Inflammation is recognized as one of the key risk factors contributed to the initiation, development, and eventually rupture or erosion of the atherosclerotic plaque that triggers thrombus formation and vascular occlusion [1]. To date, sufficient evidence has suggested a direct association between inflammation and thrombosis, especially in the case of sepsis. Sepsis is caused by the infection of gram-negative bacteria and the consequence of bacterial endotoxin (lipopolysaccharide, LPS) that induces a variety of systemic inflammatory responses. Severe sepsis is always associated with activation of the coagulation system, resulting in microthrombosis and subsequent organ failure. A number of animal models of endotoximia and disseminated intravascular coagulation (DIC) have been developed to mimic the human pathophysiology of sepsis. While these disease models may provide a vehicle to study the association and therapeutic intervention between inflammation and thrombosis, thrombus formation in these models are mainly limited within small and, to a lesser degree, medium-sized vessels [2]. In contrast, the most significant cardiovascular events caused by thrombosis most often occur in large vessels. Therefore, the present research was conducted to establish models of LPS-augmented thrombosis in the vena cava and carotid artery to study the association between inflammation and thrombosis. Previously, we demonstrated that ferric chloride (FeCl 3 ) produced a dose-dependent increase in venous or arterial thrombosis in mice. This was evaluated by the thrombus volume (in venous thrombosis) and by Doppler blood flow (in arterial thrombosis), including a characterization of the sensitivity of thrombus formation under these conditions in response to various antithrombotic agents [3–5]. In the present study we extended these thrombosis models to include inflammation. Therefore, the effects of LPS on thrombus formation were assessed using FeCl3 to induce both vena cava and carotid arterial thrombosis in mice. To study the
effect of LPS in inflammatory response and thus in thrombus formation, the expression of two key inflammatory cytokines, tumor necrosis factor α (TNFα) and interleukin-1β (IL-1β) mRNA was examined in the thrombus-containing vessels in response to LPS challenge using our previously established real-time polymerase chain reaction (PCR) method [6]. Materials and methods Mice C57BL/6 mice, purchased from Charles River Laboratories (Wilmington, MA), were used for the present study. Mice were housed in micro-isolator cages on a constant 12-hour light/dark cycle with controlled temperature and humidity and given access to food and water ad libitum. Mice were housed and cared for in accordance with the Guide for the Care and Use of Laboratory Animals [DHEW (DHHS) Publication No. (NIH) 85–23, revised 1996, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20205]. The Institutional Animal Care and Use Committees approved all procedures using mice.
FeCl3-induced venous thrombosis Mice (22–26 gm) were anesthetized with sodium pentobarbital (50 mg/kg, ip). The vena cava was exposed via a midline abdominal incision and the surface was cleared by blunt dissection between the renal and iliolumbar veins. Filter paper (2 mm × 4 mm; Gel Blot Paper, GB003, Schleicher & Schuell, Keene, NH, USA) pre-saturated with a FeCl3 (Sigma, St. Louis, MO, USA) solutions (2.5 or 3.5% in water) was placed on the vena cava for 3 minutes and then removed. Thirty min after initial filter paper application, the vena cava containing the thrombus was dissected free and a section of vessel (~ 15 mm in length) was cut out. The clot was dissected out from the vessel in saline. Based on our previous study that protein content provides a better assessment of thrombus size than thrombus wet weight [4,5], thrombus size was determined by protein content. Thrombus was digested overnight in 200 μl buffer (100 mM Tris, pH 7.5) containing 400 μg proteinase K (Invitrogen, Carlsbad, CA, USA) at 50 °C. The digested protein contents (amino acids and small peptides) were measured at OD280 calibrated with the digestion buffer in the presence of proteinase K.
FeCl3-induced arterial thrombosis The 2.5 and 3.5% FeCl3-induced carotid artery thrombosis model was carried out as described previously [3]. Mice (20–25 g) were anesthetized with pentobarbital (50 mg/kg, i.p.). An incision was made with scissors directly over the right common carotid artery, and a segment of the artery was exposed using blunt dissection. A miniature Doppler flow probe (Model 0.5 VB, Transonic System
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Inc., Ithaca, NY, USA) was attached to the carotid artery to monitor blood flow. Thrombus formation was induced by applying two pieces of filter paper (Schleicher & Schuell) (1 × 1.5 mm) saturated with 2.5% or 3.5% FeCl3 solution in water. Our previous study demonstrated that 2.5% FeCl3 only produced a subthreshold injury to the carotid artery whereas 3.5% FeCl3 resulted in consistent vascular occlusion [3]. The pieces of filter paper were placed on top of and beneath the carotid artery in contact with the adventitial surface of the vessel. After three minutes of exposure, the filter papers were removed and the vessel was washed with sterile physiological saline. Carotid blood flow was continuously monitored for thirty minutes after ferric chloride application, and flow rates were illustrated prior to (time 0), and 5, 10, 20 and 30 min after FeCl3 application for data processing.
LPS administration LPS was purchased from Sigma (Catalog # L-7011; St. Louis, MO, USA) and prepared in saline for intraperitoneal (i.p.) administration at various doses of 0.8, 2 and 5 mg/kg in a final volume of 10 μl/gm. Pilot studies were carried out for FeCl3-induced vena cava thrombosis model comparing 30 min and 2 hr after 2 mg/kg LPS treatment. Since no obvious difference was observed on thrombus formation following these two time-points, most studies described here for both vena cava and carotid arterial thrombosis were performed at 30 min after LPS or vehicle (saline) administration.
Real-time RT-PCR assay For RNA extraction, the carotid artery or vena cava containing thrombus was isolated 30 min following 2.5% FeCl3-treatment and immediately placed in RNAlater (Ambion, Austin, TX) solution and store at 4 °C. Total cellular RNA was prepared by grinding the tissue in liquid nitrogen, followed by homogenization in Trizol solution (Invitrogen, Carlsbad, CA) and extracted with chloroform according to the manufacturer's protocol. RNA was then treated with DNase using the Turbo DNA-free kit (Ambion) to remove potentially contaminated DNA. PCR primers for TNFα, IL-1β and a house-keeping gene ribosomal protein L32 (rpL32) were used as described previously [7]. Two-step RT- PCR was performed as described in detail previously [6]. Specifically, 2 μg of total RNA was reversely
357
transcribed in a total volume of 20 μl using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). This RT reaction was conducted at 25 °C for 5 min, 42 °C for 30 min and terminated at 85 °C for 5 min. Real-time PCR was then performed using iTaq SYBR green supermix with rox kit (Bio-Rad) with 0.4 μl of RT products (~ 40 ng total RNA) plus 500 nM primers in a total volume of 20 μl. PCR was preformed at 50 °C for 2 min, 95 °C for 10 min and then run for 40 cycles at 95 °C for 15 sec, 60 °C for 1 min and followed by melting curve program at 95 °C for 15 sec, 60 °C for 15 sec and 95 °C for 15 sec on the ABI PRISM 7900 Detection System. Since SYBR green is not selective for DNA species, it was essential to run a melting curve program to ensure that all the PCR products for a particular primer pair have the same melting temperature so as to eliminate any potential problems such as contamination, mispriming or primer-dimerization. The experiments confirming the melting curve of primers were conducted according to manufacturer's specification. The delta — delta method was used as described by PerkinElmer Applied Biosystems to determine the relative levels of mRNA expression between experimental samples and controls as reported previously [8,9]. It assumes that the real-time PCR achieves optimal amplification efficiencies (E) for both the target and reference genes, i.e., E = 2. Under these conditions we can define CT as the threshold cycle value of the PCR reaction and use this to calculate the amplification ratio = 2− (Δ CTsample − Δ CTcontrol), which can be further simplified as the ratio = 2− ΔΔCT. This CT value is inversely proportional to the logarithmic scale of the starting quantity of template DNA. Therefore, samples containing low concentrations of target molecules require more PCR cycles (i.e., higher CT) to amplify enough copies to reach the detected threshold amplification. The relative quantification is based on a reference mRNA, which has a consistent level of mRNA expression under existing experimental sample and control sample conditions. In the present study, rpL32 was selected as the reference gene for the normalization of inter-PCR variation [6,7]. The delta CT sample (ΔCT sample) is the difference of CT values between the target gene and the house-keeping reference gene, rpL32, present in either the thrombus samples. The delta CT control (ΔCT control) is the difference of CT values for the target genes between the experimental thrombus containing vessels and the naïve vessels. The delta-delta CT values (ΔΔCT) represents the CT values of the target genes (after normalized with the housekeeping genes for any potential loading difference) subtracted from the CT values of control samples (i.e., naive vessels). In
Figure 1 Effects of LPS on 2.5 and 3.5% FeCl3-induced vena cava thrombosis in mice. C58BL/6 mice (n = 8 each) were subjected to 2.5% (A) or 3.5% (B) FeCl3 topical induction of vena cava thrombosis 30 min following i.p. administration of either LPS or vehicle (saline) as indicated. Thrombus size was determined using proteinase K-digestion and measured for the total protein content at OD280. The value of OD280 was similar under control surgery conditions (e.g., control surgery similar to saline-soaked filter paper; data not shown). ⁎ = p b 0.05 reflect LPS differences from the saline-treated control group.
358
X. Wang effect at 5% LPS (27% increase in thrombus size, p = 0.34) (Fig. 1A). In contrast, the most effective dose of LPS (2 mg/ kg) produced no statistical difference in 3.5% FeCl3-induced vena cava thrombosis in mice (17% increase in thrombus size compared to vehicle, n = 8; Fig. 1B). As described previously [4,5], control experimental conditions such as surgery and water-soaked filter paper on the vessel did not produce vena cava thrombosis (OD280 = 0) in the presence and absence of 2 mg/kg LPS.
Effects of LPS on FeCl3-induced carotid arterial thrombosis
Figure 2 Effects of LPS on 2.5 and 3.5% FeCl3-induced carotid arterial thrombosis in mice. C57BL/6 mice (n = 8 each) were subjected to 2.5% (upper panel) or 3.5% (lower panel) FeCl3-induced arterial thrombosis and the carotid artery Doppler flow was monitored. The thrombosis study was induced 30 min following i.p. administration of either LPS (2 mg/kg) or vehicle (saline). ⁎ = p b 0.05 and ⁎⁎ = p b 0.01 reflect LPS differences from the saline-treated control group at the same time point.
The potential effect of LPS on arterial thrombosis was evaluated using a 2.5% or 3.5% FeCl3 induced carotid artery injury model. As demonstrated previously [3], 2.5% FeCl3 produced only a sub-threshold stimulation to induce carotid thrombosis while 3.5% FeCl3 resulted in consistent thrombus formation as measured by Doppler blood flow (Fig. 2). Administration of 2 mg/kg LPS significantly augmented the arterial thrombus formation immediately following 2.5% FeCl3 injury (from 5–30 min illustrated; Fig. 2 upper panel). In contrast, no significant effect was observed for LPS on thrombus formation in the 3.5% FeCl3 injury model (Fig. 2 lower panel). As described previously [3], no change was observed in Doppler flow under control experimental conditions such as surgery and water-soaked filter paper treatment in the presence and absence of 2 mg/kg LPS (data not shown).
other words, the ΔΔCT values reflect the overall specific gene expression under a particular set of experimental conditions, in this case thrombus-containing vessels from the FeCl3-injury models [6].
Statistical analysis Statistical comparisons were made by analysis of variance (ANOVA, with mean differences determined using Fisher's protected least squares difference) test and values were considered to be significant when p b 0.05.
Results Effects of LPS on FeCl 3-induced vena cava thrombosis Concentration-dependent FeCl3-induced vena cava thrombosis was established in C57BL/6 mice as described previously [4]. 2.5% FeCl3 induced a thrombus containing 0.81 + 0.11 (n = 8) protein content detected at OD280, while 3.5% FeCl3 resulted in a thrombus of 1.22 + 0.14 (n = 8) OD280 reading, showing 50.6% increase in thrombus size (p b 0.05) compared to that of 2.5% FeCl3 injury model (Fig. 1A&B). A dosedependent effect was observed for LPS on vena cava thrombosis induced by 2.5% FeCl3, showing no effect at 0.8 mg/kg LPS, the most robust effect at 2 mg/kg LPS (60% increase in thrombus over vehicle, p b 0.05), and moderate
Figure 3 Real-time PCR analysis for the effect of LPS on TNFα and IL-1β mRNA expression in 2.5% FeCl3-induced carotid arterial thrombosis and vena cava thrombosis in mice. C57BL/6 mice (n = 6 each) were treated with 2 mg/kg LPS, i.p., for 30 min and subjected to either 2.5% FeCl3-induced arterial thrombosis (A) or 2.5% FeCl3-induced vena cava thrombosis (B). The vessels were removed 30 min after thrombus induction and subjected to RNA extraction and real-time PCR as described in detail in Methods. The data are illustrated as the 2- ΔΔCT values (normalized with both the house-keeping gene rpL32 and the levels of gene expression in naïve vessels). ⁎ = P b 0.05 reflect LPS differences from the saline-treated control group.
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Effects of LPS on inflammatory cytokine mRNA expression in thrombus To confirm the effect of LPS on inflammatory response and thus thrombus formation in the vessels, real-time PCR was used to evaluate the expression of TNFα and IL-1β mRNA in both carotid artery and vena cava in response to LPS challenge. As illustrated in Fig. 3, 3.3- and 2.6-fold increase in TNFα mRNA induction (P b 0.05, n = 6) was observed in the thrombus-containing carotid artery and vena cava, respectively, as determined by the ΔΔCT values following 2 mg/kg treatment relevant to vehicle treatment. The expression of IL-1β mRNA was increased for 2.3- and 2.7-fold (p b 0.05, n = 6) for the thrombus-containing carotid artery and vena cava, respectively, in response to 2 mg/kg LPS treatment.
Discussion While it has been well established that LPS-induced endotoxemia accelerates microvascular thrombus formation in vivo, its effect on thrombosis in large vessels has not been previously reported. Sub-threshold FeCl3 stimulation was determined as described previously [3–5] and used to study in detail the effects of LPS on thrombus formation in large vessels. The present study is the first to provide direct evidence that LPS augments thrombus formation in both vena cava and carotid arterial thrombosis in response to the sub-threshold concentrations of FeCl3. Different from previously reported animal models of thrombosis induced/augmented by LPS, such as severe sepsis complicated by DIC [10] or coadministration of LPS and soluble TF [2], the present models assess the effect of LPS on thrombosis in large vessels. It is critical to select an appropriate thrombosis model to study the effect of LPS on thrombosis. A large number of thrombosis models have been developed [11], of which FeCl3 has been widely used to induce thrombosis produced by oxidative damage to vessel walls. The degree of vessel injury produced by different concentrations of FeCl3 applied onto the vessel are critical to establishing the sensitively necessary to detect the effect of LPS on thrombus formation. As shown in the present study, the synergy between LPS and FeCl3 treatment was observed only at a subthreshold concentration of FeCl3 (2.5%) but not at a higher concentration (3.5%) (e.g., note that at the higher FeCl3-induced thrombosis was consistently induced [3,4]). In agreement with the present study, coadministration of low-dose LPS and soluble TF was shown to induce thrombosis of irradiated vessels in a rat model of arteriovenus malformation [2]. LPS has a multitude of prothrombotic and proinflammatory effects. The prothrombotic effects for LPS are thought to be mediated through generalized endothelial activation and increased platelet reactiv-
359
ity since LPS was shown to stimulate increased expression of tissue factor (TF), endothelial P-selectin and plasminogen activator inhibitor (PAI)-1 [12–14]. The proinflammatory effects of LPS are mediated by the induction of cytokines such as TNF-α, IL-1 and -6 [15]. While the present study demonstrated the prothrombotic effect of LPS in both arterial and venous thrombosis, it remains to be explored if LPS mechanisms involve direct effects on both endothelial and platelet activation to produce thrombus formation in the current models. Likewise, the relative contributions of platelet and endothelial cell derived microparticles to the thrombotic events remain to be investigated. Using a sensitive real-time PCR method to monitor the expression of TNF-α and IL-1β in the thrombus-containing vessels, the present study suggested the proinflammatory effects of LPS, which appear to be involved in augmented thrombus formation in vivo. The specific mechanisms for the effect of LPS on thrombus formation are unknown, which are critical to interpret the bell-shaped dose-response to LPS as observed in the current venous thrombosis model. The bell-shaped dose response to LPS in thrombus formation may reflect the potential effect of higher doses of LPS to facilitate endogenous thrombolysis. However, a previous report excluded this possibility by showing that LPS was not sensitive to endogenous thrombolysis in ferrous chloride-induced carotid artery thrombosis in rats [16]. The curve might also be due to LPS-induced thrombocytopenia [17] and/or neutropenia [18,19], or their interactions with the prothrombotic and proinflammatory effects of LPS in FeCl3-induced thrombosis. In any event, the exact mechanisms related to the doseresponse to LPS remain to be explored. The specific effect of LPS on thrombocytopenia and neutropenia has not been examined in the present study. The cellular receptors responsible for LPS have been identified to be the toll-like receptor (TLR)-4 and its complex with CD14 and the adaptor molecule MD2 [20]. Upon binding to this complex, LPS activates the signaling pathways through NF-κB and p38 mitogen-activated protein kinase, leading to subsequent induction of various proinflammatory cytokines and adhesion molecules [21]. The effects of LPS in the current experimental models of thrombosis are likely to reflect the initial innate immune response according to the previous studies of the endotoxemia models induced by a single infusion of LPS [22–24]. This assumption can be explored in the future in the present LPS-augmented thrombosis models. In summary, this work describes two murine models LPS-augmented thrombosis (e.g., vena cava thrombosis and carotid arterial thrombosis)
360 induced by FeCl3. In the future, these models should provide useful tools to study the association between inflammation and thrombosis, two key components of atherothrombosis that ultimately leads to the major cardiovascular events.
Acknowledgement The author thanks Patricia Smith for technical assistance at the beginning of the research work and Frank Barone for critical review and comments of the manuscript.
References [1] Libby P. Coronary artery injury and the biology of atherosclerosis: inflammation, thrombosis, and stabilization. Am J Cardiol 2000;86:3J–8J. [2] Storer K, Tu J, Karunanayaka A, Smee R, Short R, Thorpe P, et al. Coadministration of low-dose lipopolysaccharide and soluble tissue factor induces thrombosis after radiosurgery in an animal arteriovenous malformation model. Neurosurgery 2007;61:604–10. [3] Wang X, Cheng Q, Xu L, Feuerstein GZ, Hsu MY, Smith PL, et al. Effects of Factor IX or Factor XI Deficiency on Ferric ChlorideInduced Carotid Artery Occlusion in Mice. J Thromb Haemost 2005;3:695–702. [4] Wang X, Smith PL, Hsu MY, Ogletree ML, Schumacher WA. Murine model of ferric chloride-induced vena cava thrombosis: evidence for effect of potato carboxypeptidase inhibitor. J Thromb Haemost 2006;4:403–10. [5] Wang X, Smith PL, Hsu MY, Gailani D, Schumacher WA, Ogletree ML, et al. Effects of factor XI deficiency on ferric chloride-induced vena cava thrombosis in mice. J Thromb Haemost 2006;4:1982–8. [6] Wang X, Hsu MY, Steinbacher TE, Monticello TM, Schumacher WA. Quantification of platelet composition in experimental venous thrombosis by real-time polymerase chain reaction. Thromb Res 2007;119:593–600. [7] Wang X, Xu L, Wang H, Young PR, Gaestel M, Feuerstein GZ. Mitogen-activated protein kinase-activated protein (MAPKAP) kinase 2 deficiency protects brain from ischemic injury in mice. J Biol Chem 2002;277:43968–72. [8] Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29:e45. [9] Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402–8. [10] Slofstra SH, Spek CA, ten Cate H. Disseminated intravascular coagulation. Hematol J 2003;4:295–302.
X. Wang [11] Westrick RJ, Winn ME, Eitzman DT. Murine models of vascular thrombosis (Eitzman series). Arterioscler Thromb Vasc Biol 2007;27:2079–93. [12] Drake TA, Cheng J, Chang A, Taylor Jr FB. Expression of tissue factor, thrombomodulin, and E-selectin in baboons with lethal Escherichia coli sepsis. Am J Pathol 1993;142:1458–70. [13] Hara S, Asada Y, Hatakeyama K, Marutsuka K, Sato Y, Kisanuki A, et al. Expression of tissue factor and tissue factor pathway inhibitor in rats lungs with lipopolysaccharide-induced disseminated intravascular coagulation. Lab Invest 1997;77:581–9. [14] Lindenblatt N, Menger MD, Klar E, Vollmar B. Systemic hypothermia increases PAI-1 expression and accelerates microvascular thrombus formation in endotoxemic mice. Crit Care 2006;10:R148. [15] Guha M, Mackman N. LPS induction of gene expression in human monocytes. Cell Signal 2001;13:85–94. [16] Wang YX, Dong N, Wu C, Martin-McNulty B, Fitch RM, da Cunha V, et al. Lipopolysaccharide attenuates thrombolysis in batroxobin-induced lung vasculature fibrin deposition but not in ferrous chloride-induced carotid artery thrombus in rats: role of endogenous PAI-1. Thromb Res 2003;111: 381–7. [17] Aslam R, Speck ER, Kim M, Crow AR, Bang KW, Nestel FP, et al. Platelet Toll-like receptor expression modulates lipopolysaccharide-induced thrombocytopenia and tumor necrosis factor-alpha production in vivo. Blood 2006;107: 637–41. [18] Wagner JG, Harkema JR, Roth RA. Pulmonary leukostasis and the inhibition of airway neutrophil recruitment are early events in the endotoxemic rat. Shock 2002;17:151–8. [19] Short A, Wong AK, Finch AM, Haaima G, Shiels IA, Fairlie DP, et al. Effects of a new C5a receptor antagonist on C5a- and endotoxin-induced neutropenia in the rat. Br J Pharmacol 1999;126:551–4. [20] Pierangeli SS, Vega-Ostertag ME, Raschi E, Liu X, RomayPenabad Z, De Micheli V, et al. Toll-like receptor and antiphospholipid mediated thrombosis: in vivo studies. Ann Rheum Dis 2007;66:1327–33. [21] Kawai T, Akira S. Toll-like receptor downstream signaling. Arthritis Res Ther 2005;7:12–9. [22] Schoenmakers SH, Versteeg HH, Groot AP, Reitsma PH, Spek CA. Tissue factor haploinsufficiency during endotoxin induced coagulation and inflammation in mice. J Thromb Haemost 2004;2:2185–93. [23] Corral J, Yelamos J, Hernandez-Espinosa D, Monreal Y, Mota R, Arcas I, et al. Role of lipopolysaccharide and cecal ligation and puncture on blood coagulation and inflammation in sensitive and resistant mice models. Am J Pathol 2005;166:1089–98. [24] Camerer E, Cornelissen I, Kataoka H, Duong DN, Zheng YW, Coughlin SR. Roles of protease-activated receptors in a mouse model of endotoxemia. Blood 2006;107:3912–21.
www.elsevier.com/locate/thromres
REGULAR ARTICLE
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Xinkang Wang ⁎ Discovery Translational Medicine, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA Received 31 October 2007; received in revised form 11 February 2008; accepted 13 March 2008 Available online 29 April 2008
KEYWORDS Arterial thrombosis; Venous thrombosis; LPS; Inflammation; Mice
Abstract Background: Animal models of diseases are essential for therapeutic target validation, drug discovery and development. Increasing evidence has demonstrated the importance of inflammation in thrombosis. Here, murine models of vena cava thrombosis and carotid arterial thrombosis augmented by lipopolysaccharide (LPS) were established and characterized to study the association between inflammation and thrombosis. Materials and methods: Murine (C57BL/6 mice) models of ferric chloride (FeCl3)induced carotid arterial and vena cava thrombosis were established. Thrombus formation was measured indirectly by Doppler blood flow (i.e., clot functional interference with blood flow) in the arterial thrombosis model and directly by protein content of the clot in the venous thrombosis model. An optimal concentration of FeCl3 was defined to induce thrombus formation and used to study the effects of LPS (i.e., a well-known inflammatory stimulus under these conditions). Real-time polymerase chain reaction (PCR) was used to examine the effect of LPS on TNFα and IL-1β mRNA expression in thrombus formation. Results: Dose-dependent analysis demonstrated that 2 mg/kg, i.p., LPS provided a maximal prothrombotic effect in 2.5% ferric chloride-induced vena cava thrombosis, with a 60% increase in thrombus size (n = 8, p b 0.05) compared to vehicle treatment. In contrast, 2 mg/kg LPS had no significant effect on thrombus formation in a more severe, 3.5% FeCl3-induced vena cava thrombosis. A similar prothrombotic effect was observed for LPS in 2.5% FeCl3-induced carotid arterial thrombosis model. Treatment of 2 mg/kg LPS significantly augmented arterial thrombosis immediately (between 5– 30 minutes) following FeCl3 injury as assessed by change of Doppler blood flow (n = 8, p b 0.05). Real-time PCR demonstrated significant induction of TNFα and IL-1β mRNA expression in the thrombus formation in the vessels in response to LPS challenge.
⁎ Tel.: +1 484 865 9906; fax: +1 484 865 9402. E-mail address: [email protected]. 0049-3848/$ - see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2008.03.015
356
X. Wang Conclusion: These data demonstrate that LPS augments thrombus formation in acute vascular injury and that LPS-augmented thrombosis might be a useful tool to study the relationship between inflammation and thrombosis. © 2008 Elsevier Ltd. All rights reserved.
Introduction Formation of a thrombus in the coronary or cerebral arteries is a major cause of morbidity and mortality throughout the world. Inflammation is recognized as one of the key risk factors contributed to the initiation, development, and eventually rupture or erosion of the atherosclerotic plaque that triggers thrombus formation and vascular occlusion [1]. To date, sufficient evidence has suggested a direct association between inflammation and thrombosis, especially in the case of sepsis. Sepsis is caused by the infection of gram-negative bacteria and the consequence of bacterial endotoxin (lipopolysaccharide, LPS) that induces a variety of systemic inflammatory responses. Severe sepsis is always associated with activation of the coagulation system, resulting in microthrombosis and subsequent organ failure. A number of animal models of endotoximia and disseminated intravascular coagulation (DIC) have been developed to mimic the human pathophysiology of sepsis. While these disease models may provide a vehicle to study the association and therapeutic intervention between inflammation and thrombosis, thrombus formation in these models are mainly limited within small and, to a lesser degree, medium-sized vessels [2]. In contrast, the most significant cardiovascular events caused by thrombosis most often occur in large vessels. Therefore, the present research was conducted to establish models of LPS-augmented thrombosis in the vena cava and carotid artery to study the association between inflammation and thrombosis. Previously, we demonstrated that ferric chloride (FeCl 3 ) produced a dose-dependent increase in venous or arterial thrombosis in mice. This was evaluated by the thrombus volume (in venous thrombosis) and by Doppler blood flow (in arterial thrombosis), including a characterization of the sensitivity of thrombus formation under these conditions in response to various antithrombotic agents [3–5]. In the present study we extended these thrombosis models to include inflammation. Therefore, the effects of LPS on thrombus formation were assessed using FeCl3 to induce both vena cava and carotid arterial thrombosis in mice. To study the
effect of LPS in inflammatory response and thus in thrombus formation, the expression of two key inflammatory cytokines, tumor necrosis factor α (TNFα) and interleukin-1β (IL-1β) mRNA was examined in the thrombus-containing vessels in response to LPS challenge using our previously established real-time polymerase chain reaction (PCR) method [6]. Materials and methods Mice C57BL/6 mice, purchased from Charles River Laboratories (Wilmington, MA), were used for the present study. Mice were housed in micro-isolator cages on a constant 12-hour light/dark cycle with controlled temperature and humidity and given access to food and water ad libitum. Mice were housed and cared for in accordance with the Guide for the Care and Use of Laboratory Animals [DHEW (DHHS) Publication No. (NIH) 85–23, revised 1996, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20205]. The Institutional Animal Care and Use Committees approved all procedures using mice.
FeCl3-induced venous thrombosis Mice (22–26 gm) were anesthetized with sodium pentobarbital (50 mg/kg, ip). The vena cava was exposed via a midline abdominal incision and the surface was cleared by blunt dissection between the renal and iliolumbar veins. Filter paper (2 mm × 4 mm; Gel Blot Paper, GB003, Schleicher & Schuell, Keene, NH, USA) pre-saturated with a FeCl3 (Sigma, St. Louis, MO, USA) solutions (2.5 or 3.5% in water) was placed on the vena cava for 3 minutes and then removed. Thirty min after initial filter paper application, the vena cava containing the thrombus was dissected free and a section of vessel (~ 15 mm in length) was cut out. The clot was dissected out from the vessel in saline. Based on our previous study that protein content provides a better assessment of thrombus size than thrombus wet weight [4,5], thrombus size was determined by protein content. Thrombus was digested overnight in 200 μl buffer (100 mM Tris, pH 7.5) containing 400 μg proteinase K (Invitrogen, Carlsbad, CA, USA) at 50 °C. The digested protein contents (amino acids and small peptides) were measured at OD280 calibrated with the digestion buffer in the presence of proteinase K.
FeCl3-induced arterial thrombosis The 2.5 and 3.5% FeCl3-induced carotid artery thrombosis model was carried out as described previously [3]. Mice (20–25 g) were anesthetized with pentobarbital (50 mg/kg, i.p.). An incision was made with scissors directly over the right common carotid artery, and a segment of the artery was exposed using blunt dissection. A miniature Doppler flow probe (Model 0.5 VB, Transonic System
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Inc., Ithaca, NY, USA) was attached to the carotid artery to monitor blood flow. Thrombus formation was induced by applying two pieces of filter paper (Schleicher & Schuell) (1 × 1.5 mm) saturated with 2.5% or 3.5% FeCl3 solution in water. Our previous study demonstrated that 2.5% FeCl3 only produced a subthreshold injury to the carotid artery whereas 3.5% FeCl3 resulted in consistent vascular occlusion [3]. The pieces of filter paper were placed on top of and beneath the carotid artery in contact with the adventitial surface of the vessel. After three minutes of exposure, the filter papers were removed and the vessel was washed with sterile physiological saline. Carotid blood flow was continuously monitored for thirty minutes after ferric chloride application, and flow rates were illustrated prior to (time 0), and 5, 10, 20 and 30 min after FeCl3 application for data processing.
LPS administration LPS was purchased from Sigma (Catalog # L-7011; St. Louis, MO, USA) and prepared in saline for intraperitoneal (i.p.) administration at various doses of 0.8, 2 and 5 mg/kg in a final volume of 10 μl/gm. Pilot studies were carried out for FeCl3-induced vena cava thrombosis model comparing 30 min and 2 hr after 2 mg/kg LPS treatment. Since no obvious difference was observed on thrombus formation following these two time-points, most studies described here for both vena cava and carotid arterial thrombosis were performed at 30 min after LPS or vehicle (saline) administration.
Real-time RT-PCR assay For RNA extraction, the carotid artery or vena cava containing thrombus was isolated 30 min following 2.5% FeCl3-treatment and immediately placed in RNAlater (Ambion, Austin, TX) solution and store at 4 °C. Total cellular RNA was prepared by grinding the tissue in liquid nitrogen, followed by homogenization in Trizol solution (Invitrogen, Carlsbad, CA) and extracted with chloroform according to the manufacturer's protocol. RNA was then treated with DNase using the Turbo DNA-free kit (Ambion) to remove potentially contaminated DNA. PCR primers for TNFα, IL-1β and a house-keeping gene ribosomal protein L32 (rpL32) were used as described previously [7]. Two-step RT- PCR was performed as described in detail previously [6]. Specifically, 2 μg of total RNA was reversely
357
transcribed in a total volume of 20 μl using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). This RT reaction was conducted at 25 °C for 5 min, 42 °C for 30 min and terminated at 85 °C for 5 min. Real-time PCR was then performed using iTaq SYBR green supermix with rox kit (Bio-Rad) with 0.4 μl of RT products (~ 40 ng total RNA) plus 500 nM primers in a total volume of 20 μl. PCR was preformed at 50 °C for 2 min, 95 °C for 10 min and then run for 40 cycles at 95 °C for 15 sec, 60 °C for 1 min and followed by melting curve program at 95 °C for 15 sec, 60 °C for 15 sec and 95 °C for 15 sec on the ABI PRISM 7900 Detection System. Since SYBR green is not selective for DNA species, it was essential to run a melting curve program to ensure that all the PCR products for a particular primer pair have the same melting temperature so as to eliminate any potential problems such as contamination, mispriming or primer-dimerization. The experiments confirming the melting curve of primers were conducted according to manufacturer's specification. The delta — delta method was used as described by PerkinElmer Applied Biosystems to determine the relative levels of mRNA expression between experimental samples and controls as reported previously [8,9]. It assumes that the real-time PCR achieves optimal amplification efficiencies (E) for both the target and reference genes, i.e., E = 2. Under these conditions we can define CT as the threshold cycle value of the PCR reaction and use this to calculate the amplification ratio = 2− (Δ CTsample − Δ CTcontrol), which can be further simplified as the ratio = 2− ΔΔCT. This CT value is inversely proportional to the logarithmic scale of the starting quantity of template DNA. Therefore, samples containing low concentrations of target molecules require more PCR cycles (i.e., higher CT) to amplify enough copies to reach the detected threshold amplification. The relative quantification is based on a reference mRNA, which has a consistent level of mRNA expression under existing experimental sample and control sample conditions. In the present study, rpL32 was selected as the reference gene for the normalization of inter-PCR variation [6,7]. The delta CT sample (ΔCT sample) is the difference of CT values between the target gene and the house-keeping reference gene, rpL32, present in either the thrombus samples. The delta CT control (ΔCT control) is the difference of CT values for the target genes between the experimental thrombus containing vessels and the naïve vessels. The delta-delta CT values (ΔΔCT) represents the CT values of the target genes (after normalized with the housekeeping genes for any potential loading difference) subtracted from the CT values of control samples (i.e., naive vessels). In
Figure 1 Effects of LPS on 2.5 and 3.5% FeCl3-induced vena cava thrombosis in mice. C58BL/6 mice (n = 8 each) were subjected to 2.5% (A) or 3.5% (B) FeCl3 topical induction of vena cava thrombosis 30 min following i.p. administration of either LPS or vehicle (saline) as indicated. Thrombus size was determined using proteinase K-digestion and measured for the total protein content at OD280. The value of OD280 was similar under control surgery conditions (e.g., control surgery similar to saline-soaked filter paper; data not shown). ⁎ = p b 0.05 reflect LPS differences from the saline-treated control group.
358
X. Wang effect at 5% LPS (27% increase in thrombus size, p = 0.34) (Fig. 1A). In contrast, the most effective dose of LPS (2 mg/ kg) produced no statistical difference in 3.5% FeCl3-induced vena cava thrombosis in mice (17% increase in thrombus size compared to vehicle, n = 8; Fig. 1B). As described previously [4,5], control experimental conditions such as surgery and water-soaked filter paper on the vessel did not produce vena cava thrombosis (OD280 = 0) in the presence and absence of 2 mg/kg LPS.
Effects of LPS on FeCl3-induced carotid arterial thrombosis
Figure 2 Effects of LPS on 2.5 and 3.5% FeCl3-induced carotid arterial thrombosis in mice. C57BL/6 mice (n = 8 each) were subjected to 2.5% (upper panel) or 3.5% (lower panel) FeCl3-induced arterial thrombosis and the carotid artery Doppler flow was monitored. The thrombosis study was induced 30 min following i.p. administration of either LPS (2 mg/kg) or vehicle (saline). ⁎ = p b 0.05 and ⁎⁎ = p b 0.01 reflect LPS differences from the saline-treated control group at the same time point.
The potential effect of LPS on arterial thrombosis was evaluated using a 2.5% or 3.5% FeCl3 induced carotid artery injury model. As demonstrated previously [3], 2.5% FeCl3 produced only a sub-threshold stimulation to induce carotid thrombosis while 3.5% FeCl3 resulted in consistent thrombus formation as measured by Doppler blood flow (Fig. 2). Administration of 2 mg/kg LPS significantly augmented the arterial thrombus formation immediately following 2.5% FeCl3 injury (from 5–30 min illustrated; Fig. 2 upper panel). In contrast, no significant effect was observed for LPS on thrombus formation in the 3.5% FeCl3 injury model (Fig. 2 lower panel). As described previously [3], no change was observed in Doppler flow under control experimental conditions such as surgery and water-soaked filter paper treatment in the presence and absence of 2 mg/kg LPS (data not shown).
other words, the ΔΔCT values reflect the overall specific gene expression under a particular set of experimental conditions, in this case thrombus-containing vessels from the FeCl3-injury models [6].
Statistical analysis Statistical comparisons were made by analysis of variance (ANOVA, with mean differences determined using Fisher's protected least squares difference) test and values were considered to be significant when p b 0.05.
Results Effects of LPS on FeCl 3-induced vena cava thrombosis Concentration-dependent FeCl3-induced vena cava thrombosis was established in C57BL/6 mice as described previously [4]. 2.5% FeCl3 induced a thrombus containing 0.81 + 0.11 (n = 8) protein content detected at OD280, while 3.5% FeCl3 resulted in a thrombus of 1.22 + 0.14 (n = 8) OD280 reading, showing 50.6% increase in thrombus size (p b 0.05) compared to that of 2.5% FeCl3 injury model (Fig. 1A&B). A dosedependent effect was observed for LPS on vena cava thrombosis induced by 2.5% FeCl3, showing no effect at 0.8 mg/kg LPS, the most robust effect at 2 mg/kg LPS (60% increase in thrombus over vehicle, p b 0.05), and moderate
Figure 3 Real-time PCR analysis for the effect of LPS on TNFα and IL-1β mRNA expression in 2.5% FeCl3-induced carotid arterial thrombosis and vena cava thrombosis in mice. C57BL/6 mice (n = 6 each) were treated with 2 mg/kg LPS, i.p., for 30 min and subjected to either 2.5% FeCl3-induced arterial thrombosis (A) or 2.5% FeCl3-induced vena cava thrombosis (B). The vessels were removed 30 min after thrombus induction and subjected to RNA extraction and real-time PCR as described in detail in Methods. The data are illustrated as the 2- ΔΔCT values (normalized with both the house-keeping gene rpL32 and the levels of gene expression in naïve vessels). ⁎ = P b 0.05 reflect LPS differences from the saline-treated control group.
Lipopolysaccharide augments venous and arterial thrombosis in the mouse Effects of LPS on inflammatory cytokine mRNA expression in thrombus To confirm the effect of LPS on inflammatory response and thus thrombus formation in the vessels, real-time PCR was used to evaluate the expression of TNFα and IL-1β mRNA in both carotid artery and vena cava in response to LPS challenge. As illustrated in Fig. 3, 3.3- and 2.6-fold increase in TNFα mRNA induction (P b 0.05, n = 6) was observed in the thrombus-containing carotid artery and vena cava, respectively, as determined by the ΔΔCT values following 2 mg/kg treatment relevant to vehicle treatment. The expression of IL-1β mRNA was increased for 2.3- and 2.7-fold (p b 0.05, n = 6) for the thrombus-containing carotid artery and vena cava, respectively, in response to 2 mg/kg LPS treatment.
Discussion While it has been well established that LPS-induced endotoxemia accelerates microvascular thrombus formation in vivo, its effect on thrombosis in large vessels has not been previously reported. Sub-threshold FeCl3 stimulation was determined as described previously [3–5] and used to study in detail the effects of LPS on thrombus formation in large vessels. The present study is the first to provide direct evidence that LPS augments thrombus formation in both vena cava and carotid arterial thrombosis in response to the sub-threshold concentrations of FeCl3. Different from previously reported animal models of thrombosis induced/augmented by LPS, such as severe sepsis complicated by DIC [10] or coadministration of LPS and soluble TF [2], the present models assess the effect of LPS on thrombosis in large vessels. It is critical to select an appropriate thrombosis model to study the effect of LPS on thrombosis. A large number of thrombosis models have been developed [11], of which FeCl3 has been widely used to induce thrombosis produced by oxidative damage to vessel walls. The degree of vessel injury produced by different concentrations of FeCl3 applied onto the vessel are critical to establishing the sensitively necessary to detect the effect of LPS on thrombus formation. As shown in the present study, the synergy between LPS and FeCl3 treatment was observed only at a subthreshold concentration of FeCl3 (2.5%) but not at a higher concentration (3.5%) (e.g., note that at the higher FeCl3-induced thrombosis was consistently induced [3,4]). In agreement with the present study, coadministration of low-dose LPS and soluble TF was shown to induce thrombosis of irradiated vessels in a rat model of arteriovenus malformation [2]. LPS has a multitude of prothrombotic and proinflammatory effects. The prothrombotic effects for LPS are thought to be mediated through generalized endothelial activation and increased platelet reactiv-
359
ity since LPS was shown to stimulate increased expression of tissue factor (TF), endothelial P-selectin and plasminogen activator inhibitor (PAI)-1 [12–14]. The proinflammatory effects of LPS are mediated by the induction of cytokines such as TNF-α, IL-1 and -6 [15]. While the present study demonstrated the prothrombotic effect of LPS in both arterial and venous thrombosis, it remains to be explored if LPS mechanisms involve direct effects on both endothelial and platelet activation to produce thrombus formation in the current models. Likewise, the relative contributions of platelet and endothelial cell derived microparticles to the thrombotic events remain to be investigated. Using a sensitive real-time PCR method to monitor the expression of TNF-α and IL-1β in the thrombus-containing vessels, the present study suggested the proinflammatory effects of LPS, which appear to be involved in augmented thrombus formation in vivo. The specific mechanisms for the effect of LPS on thrombus formation are unknown, which are critical to interpret the bell-shaped dose-response to LPS as observed in the current venous thrombosis model. The bell-shaped dose response to LPS in thrombus formation may reflect the potential effect of higher doses of LPS to facilitate endogenous thrombolysis. However, a previous report excluded this possibility by showing that LPS was not sensitive to endogenous thrombolysis in ferrous chloride-induced carotid artery thrombosis in rats [16]. The curve might also be due to LPS-induced thrombocytopenia [17] and/or neutropenia [18,19], or their interactions with the prothrombotic and proinflammatory effects of LPS in FeCl3-induced thrombosis. In any event, the exact mechanisms related to the doseresponse to LPS remain to be explored. The specific effect of LPS on thrombocytopenia and neutropenia has not been examined in the present study. The cellular receptors responsible for LPS have been identified to be the toll-like receptor (TLR)-4 and its complex with CD14 and the adaptor molecule MD2 [20]. Upon binding to this complex, LPS activates the signaling pathways through NF-κB and p38 mitogen-activated protein kinase, leading to subsequent induction of various proinflammatory cytokines and adhesion molecules [21]. The effects of LPS in the current experimental models of thrombosis are likely to reflect the initial innate immune response according to the previous studies of the endotoxemia models induced by a single infusion of LPS [22–24]. This assumption can be explored in the future in the present LPS-augmented thrombosis models. In summary, this work describes two murine models LPS-augmented thrombosis (e.g., vena cava thrombosis and carotid arterial thrombosis)
360 induced by FeCl3. In the future, these models should provide useful tools to study the association between inflammation and thrombosis, two key components of atherothrombosis that ultimately leads to the major cardiovascular events.
Acknowledgement The author thanks Patricia Smith for technical assistance at the beginning of the research work and Frank Barone for critical review and comments of the manuscript.
References [1] Libby P. Coronary artery injury and the biology of atherosclerosis: inflammation, thrombosis, and stabilization. Am J Cardiol 2000;86:3J–8J. [2] Storer K, Tu J, Karunanayaka A, Smee R, Short R, Thorpe P, et al. Coadministration of low-dose lipopolysaccharide and soluble tissue factor induces thrombosis after radiosurgery in an animal arteriovenous malformation model. Neurosurgery 2007;61:604–10. [3] Wang X, Cheng Q, Xu L, Feuerstein GZ, Hsu MY, Smith PL, et al. Effects of Factor IX or Factor XI Deficiency on Ferric ChlorideInduced Carotid Artery Occlusion in Mice. J Thromb Haemost 2005;3:695–702. [4] Wang X, Smith PL, Hsu MY, Ogletree ML, Schumacher WA. Murine model of ferric chloride-induced vena cava thrombosis: evidence for effect of potato carboxypeptidase inhibitor. J Thromb Haemost 2006;4:403–10. [5] Wang X, Smith PL, Hsu MY, Gailani D, Schumacher WA, Ogletree ML, et al. Effects of factor XI deficiency on ferric chloride-induced vena cava thrombosis in mice. J Thromb Haemost 2006;4:1982–8. [6] Wang X, Hsu MY, Steinbacher TE, Monticello TM, Schumacher WA. Quantification of platelet composition in experimental venous thrombosis by real-time polymerase chain reaction. Thromb Res 2007;119:593–600. [7] Wang X, Xu L, Wang H, Young PR, Gaestel M, Feuerstein GZ. Mitogen-activated protein kinase-activated protein (MAPKAP) kinase 2 deficiency protects brain from ischemic injury in mice. J Biol Chem 2002;277:43968–72. [8] Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29:e45. [9] Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402–8. [10] Slofstra SH, Spek CA, ten Cate H. Disseminated intravascular coagulation. Hematol J 2003;4:295–302.
X. Wang [11] Westrick RJ, Winn ME, Eitzman DT. Murine models of vascular thrombosis (Eitzman series). Arterioscler Thromb Vasc Biol 2007;27:2079–93. [12] Drake TA, Cheng J, Chang A, Taylor Jr FB. Expression of tissue factor, thrombomodulin, and E-selectin in baboons with lethal Escherichia coli sepsis. Am J Pathol 1993;142:1458–70. [13] Hara S, Asada Y, Hatakeyama K, Marutsuka K, Sato Y, Kisanuki A, et al. Expression of tissue factor and tissue factor pathway inhibitor in rats lungs with lipopolysaccharide-induced disseminated intravascular coagulation. Lab Invest 1997;77:581–9. [14] Lindenblatt N, Menger MD, Klar E, Vollmar B. Systemic hypothermia increases PAI-1 expression and accelerates microvascular thrombus formation in endotoxemic mice. Crit Care 2006;10:R148. [15] Guha M, Mackman N. LPS induction of gene expression in human monocytes. Cell Signal 2001;13:85–94. [16] Wang YX, Dong N, Wu C, Martin-McNulty B, Fitch RM, da Cunha V, et al. Lipopolysaccharide attenuates thrombolysis in batroxobin-induced lung vasculature fibrin deposition but not in ferrous chloride-induced carotid artery thrombus in rats: role of endogenous PAI-1. Thromb Res 2003;111: 381–7. [17] Aslam R, Speck ER, Kim M, Crow AR, Bang KW, Nestel FP, et al. Platelet Toll-like receptor expression modulates lipopolysaccharide-induced thrombocytopenia and tumor necrosis factor-alpha production in vivo. Blood 2006;107: 637–41. [18] Wagner JG, Harkema JR, Roth RA. Pulmonary leukostasis and the inhibition of airway neutrophil recruitment are early events in the endotoxemic rat. Shock 2002;17:151–8. [19] Short A, Wong AK, Finch AM, Haaima G, Shiels IA, Fairlie DP, et al. Effects of a new C5a receptor antagonist on C5a- and endotoxin-induced neutropenia in the rat. Br J Pharmacol 1999;126:551–4. [20] Pierangeli SS, Vega-Ostertag ME, Raschi E, Liu X, RomayPenabad Z, De Micheli V, et al. Toll-like receptor and antiphospholipid mediated thrombosis: in vivo studies. Ann Rheum Dis 2007;66:1327–33. [21] Kawai T, Akira S. Toll-like receptor downstream signaling. Arthritis Res Ther 2005;7:12–9. [22] Schoenmakers SH, Versteeg HH, Groot AP, Reitsma PH, Spek CA. Tissue factor haploinsufficiency during endotoxin induced coagulation and inflammation in mice. J Thromb Haemost 2004;2:2185–93. [23] Corral J, Yelamos J, Hernandez-Espinosa D, Monreal Y, Mota R, Arcas I, et al. Role of lipopolysaccharide and cecal ligation and puncture on blood coagulation and inflammation in sensitive and resistant mice models. Am J Pathol 2005;166:1089–98. [24] Camerer E, Cornelissen I, Kataoka H, Duong DN, Zheng YW, Coughlin SR. Roles of protease-activated receptors in a mouse model of endotoxemia. Blood 2006;107:3912–21.