- Email: [email protected]
Electric injury model of murine arterial thrombosis
Thrombosis Research (2007) 121, 103–106
intl.elsevierhealth.com/journals/thre
REGULAR ARTICLE
Electric injury model of murine arterial thrombosis Akiko Kusada b , Noritaka Isogai b , Brian C. Cooley a,⁎ a
Medical College of Wisconsin, Department of Orthopaedic Surgery, 8701 Watertown Plank Road, 53226, Milwaukee, Wisconsin, United States b Department of Plastic Surgery, Kinki University, Osaka, Japan Received 3 July 2006; received in revised form 16 January 2007; accepted 8 March 2007 Available online 25 April 2007
Abstract Background and Purpose: Murine models of arterial thrombosis have gained utility with applications in genetically manipulated mice. Implementation of current models requires specialized equipment and provides limited outcome measures. A new murine model of continuously monitored arterial thrombosis was created. Methods: An electric injury was delivered to the exterior surface of the common carotid artery using the flat end of a 140-μm steel microsurgery needle connected to the anode of a 3-V battery source. Direct current was delivered for 30 s. The developing thrombus was apparent as a white, platelet-dominated region at the site of injury. This region was continuously monitored and recorded by videotape for 30 min. Subsequently, the thrombus area was measured directly on the TV monitor, generating a time course for thrombogenesis. In a further evaluation of the model, three pharmacologically treated groups of mice were evaluated, with drug infusion immediately before thrombus induction: (1) saline (control), (2) heparin (60 units/kg), and (3) GR144053, a GPIIb/IIIa-specific antagonist (10 mg/kg). Results: The basic model showed consistent thrombus development by 7–9 min, occasionally forming an occlusive thrombus. Most of the thrombi underwent one or more cycles of embolization and thrombus regrowth. In the experimental series, the heparin-treated group had a significantly decreased thrombus area versus controls (p b 0.0001); the GR144053-treated mice had no apparent thrombus, supporting a dominant role of platelet aggregation in arterial thrombogenesis. Conclusion: This new model is simple to do, uses readily available instrumentation, and provides a continuously recorded quantifiable measure of thrombogenesis. © 2007 Elsevier Ltd. All rights reserved.
Introduction
⁎ Corresponding author. Tel.: +1 414 456 8729; fax: +1 414 456 6488. E-mail address: [email protected] (B.C. Cooley).
Arterial thrombosis has many clinical manifestations for which a better understanding of the causes and therapeutic options is needed. Murine models of
0049-3848/$ - see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2007.03.005
104 arterial thrombosis have been developed to take advantage of the ever increasing numbers of genetically manipulated mouse lines by which the isolated study of single-gene effects on thrombotic outcomes can be undertaken. The most frequently used murine models of large-artery (e.g., carotid) thrombosis are the ferric chloride (filter-strip) model, originally developed in rats by Kurz and associates, [1] and the Rose Bengal model [2,3]. Both of these models entail free-radical-generated induction of vascular injury, with monitoring of flow to record the “time to occlusion” as the quantitative end-point measure. Both models have shown utility in a number of studies [4–8], but they require investment in a rather expensive and limited piece of equipment, an ultrasonic flowmeter. The Rose Bengal model also requires illumination with an intense light or laser light source for photo-activated free-radical generation. We undertook the development of a simpler model which makes use of older but functional equipment often available in small animal surgery labs — an operating/dissecting microscope and standard (nondigital) video-camera and recorder. The model uses a unique but inexpensive method for delivering a small, precise injury to the vessel surface to stimulate thrombogenesis. Furthermore, rather than one endpoint (time to occlusion), the evolution of a nonocclusive thrombus can be monitored and measured continuously, yielding a more detailed, quantifiable time profile of thrombogenesis.
A. Kusada et al. The site of electric injury/thrombus induction was monitored and recorded for 30 min at high magnification through a C-mount-coupled analog video-camera
Materials and methods Adult female CD1 mice (2–4 months old) were used following NIH and institutional guidelines for the care and use of laboratory animals. Anesthesia was induced with intraperitoneal pentobarbital (50 mg/kg). The left common carotid artery was exposed through a midline cervical incision. All procedures were done with a Wild/Leica operating microscope, using a 100-mm objective lens. A marking stitch of 10-0 nylon was placed upstream of the intended injury site. The nylon suture of a 140-μm stainless steel microsurgery needle (Product #AK-0105, Surgical Specialties Inc., Reading, PA) was removed. The blunt/flat end of this needle (where the suture was previously swaged) was held against the top exterior surface of the carotid artery; a direct, 3-V electric current was applied to the needle with two 1.5-V C batteries, connecting the positive terminal (anode) to the needle and completing the circuit with the cathode connected to a wire in contact with the cervical musculature. Current was delivered for 30 s and the needle contact with the vessel was subsequently terminated.
Figure 1 Micro-photographs (taken through an operating microscope) of the electric injury site and developing thrombus. (a) Normal carotid artery; (b) discolored/blanched site of electric injury (between arrows) 3 min after injury; developing thrombus, seen at (c) 8 min and (d) 15 min; (e) embolized thrombus (arrowhead) breaking off; f) occlusive thrombus with evident thrombi blocking flow (arrows) and with decreased vessel diameter on right. Flow is from left to right in all images; s = marking suture proximal to electric injury site.
Electric injury model of murine arterial thrombosis (COLCAM NTSC-1, World Precision Instruments, Sarasota, Florida, USA) connected to a standard VHS videotape recorder. The videotape time display was sufficient (in subsequent playback mode) to provide timing of events. A millimeter ruler was recorded at the same magnification to provide a calibrated measure for subsequent quantitation of on-screen images. Upon creating a 30-min recording of thrombogenesis, the videotape was analyzed with on-screen morphometric measurements. The area of thrombus as captured by the videotape was clearly identifiable by its characteristic white appearance, in contrast to the red blood cell-dominated flowing portion of the artery. This area was measured at 1-min intervals, using a calibrated acetate overlay on the TV monitor to measure the thrombus dimensions and generating a time profile for thrombogenesis of each sample. After pilot studies to ascertain the above parameters and methodology, a simple pharmacologic experiment for inhibiting thrombosis was undertaken. Three treatment groups of mice (n = 10/group) were randomly given one of the following: heparin at a dose of 60 units/kg; GR144053 (Sigma Chemical Co., St. Louis, MO), a specific platelet glycoprotein IIb/IIIa (GP IIb/IIIa) inhibitor, at a dose of 10 mg/kg; and normal saline (control). All agents were infused into the jugular vein in a 0.1-ml volume immediately prior to thrombus induction.
Results The 30-s application of electric injury caused an immediate yellowing discoloration at the contact point with the carotid artery (Fig. 1a and b). A white thrombus was seen to begin growing at the electricinjury zone approximately 5 min later, with a downstream progression over time (Fig. 1c–e). Partial or near-complete embolization of the thrombus was usually seen, followed by regrowth of the thrombus; this embolization/regrowth often occurred several
Figure 2 Three representative traces of controls showing thrombus areas over time, with cyclic patterns of growth, embolization (asterisks), and regrowth.
105
Figure 3 Line graph of mean thrombus areas over time for mice treated with saline infusion (solid squares), heparin (60 units/kg) (open diamonds), or GR144053 10 mg/kg) (open circles); infusion of the respective agents was given immediately before thrombus induction. Error bars = SEM; n = 10 per group.
times for a given sample. The growth and cyclic embolization were reflected in the time course of measured thrombus area (Fig. 2). Occasionally, an occlusive thrombus was seen, which was evident by a lack of distal pulsation, relative reduction in vessel diameter, and stabilization of the thrombus size (Fig. 1f). In the pharmacologic-treatment series, the average area of thrombus increased over the first 7–9 min in the saline-and heparin-treated groups, achieving an average stable size for 10–30 min (Fig. 3) that was significantly larger in the saline-versus heparin-treated group (p b 0.0001; ANOVA). The heparin dose used in this study (60 units/kg) gave a consistently prolonged activated partial thromboplastin time (aPTT, of greater than 120 s; data not shown), indicated full anticoagulation. None of the GR144053treated vessels showed a measurable thrombus.
Discussion The model presented herein is a simple, inexpensive method for continuous quantitative monitoring of an acute arterial thrombus. The induction of thrombosis is similar to our previously described murine venous thrombosis model [9,10] which uses a smaller microsurgical needle (75-μm diameter) placed inside the vein and a lower voltage (1.5-V) electric stimulus delivered over a longer (60-s) time. Our initial efforts to place a needle through the wall and inside the artery were frustrated by high amounts of bleeding. Thus, we developed an outer-surface injury which was highly reproducible and technically easy to duplicate. In other pilot work, we found that the use of noniron-containing wires (zinc, copper, gold-coated) with the same electric current delivery to the carotid artery caused a similarly evident “heat” injury (with
106 discoloration of the vessel wall), but did not result in a thrombus (data not shown). This strongly suggests a free-radical mechanism for thrombus induction, generated by ferric ion electrolytic delivery to the vessel by an iron-containing wire/needle. Lucchesi and colleagues found a similar effect in their original development of a canine coronary artery thrombosis model with intralumenal steel coils [11]. Thus, our new model has similarities in the mechanisms of thrombus induction to both the ferric chloride [1] and the Rose bengal [2,3] models of free-radical-mediated thrombogenesis. An advantage of our model is its direct quantitation of the thrombus, in comparison to the less direct monitoring of blood flow for onset of occlusion. Our electric-injury model is similar to the thrombus observation of a rat thrombogenic arterial injury model, originally developed by Acland and colleagues, [12] in which transillumination of the iliac artery revealed a developing white thrombus and by which thromboembolic events could be observed. In our model, transillumination was not necessary, as the white thrombus could be clearly seen growing, remodeling, and embolizing with the incident microscope light. Pierangeli and colleagues [13,14] used a “pinch” model of venous injury in mice, measuring the area of the developing thrombus, as was done in our study. This method of evaluation, like ours, is not a true thrombus volume, but gives a reasonable estimate. We found a profound inhibition of the arterial thrombus with infusion of the GP IIb/IIIa antagonist, GR144-53, in comparison to heparin-treated and control mice. This strongly implicates the dominance of platelets in a developing arterial thrombus. In contrast, heparin infusion, at a dose that has shown effectiveness in other models, only had a moderate effect (Fig. 3), despite a high anticoagulated state achieved with this dose of heparin. Thus, our electricinjury model may provide a better means for evaluating platelet-dominated arterial thrombosis in mice.
References [1] Kurz DK, Main BW, Sandusky GE. Rat model of arterial thrombosis induced by ferric chloride. Thromb Res 1990;60: 269 – 80.
A. Kusada et al. [2] Kawasaki T, Kaida T, Arnout J, Vermylen J, Hoylaerts MF. A new animal model of thrombophilia confirms that high plasma factor VIII levels are thrombogenic. Thromb Haemost 1999;81: 306 – 11. [3] Matsuno H, Kozawa O, Niwa M, Ueshima S, Matsuo O, Collen D, et al. Differential role of components of the fibrinolytic system in the formation and removal of thrombus induced by endothelial injury. Thromb Haemost 1999;81:601 – 4. [4] Weiler H, Lindner V, Kerlin B, Isermann BH, Hendrickson SB, Cooley BC, et al. Characterization of a mouse model for thrombomodulin deficiency. Arterioscler Thromb Vasc Biol 2001;21: 1531– 7. [5] Kerlin B, Cooley BC, Isermann BH, Hernandez I, Sood R, Zogg M, et al. Cause–effect relation between hyperfibrinogenemia and vascular disease. Blood 2004;103:1728 – 34. [6] Falati S, Patil S, Gross PL, Stapleton M, Merrill-Skoloff G, Barrett NE, et al. Platelet PECAM-1 inhibits thrombus formation in vivo. Blood 2006;107:535 – 41. [7] Wilson KM, Lynch CM, Faraci FM, Lentz SR. Effect of mechanical ventilation on carotid artery thrombosis induced by photochemical injury in mice. J Thromb Haemost 2003;1: 2669 – 74. [8] Danenberg HD, Szalai AJ, Swaminathan RV, Peng L, Chen Z, Seifert P, et al. Increased thrombosis after arterial injury in human C-reactive protein-transgenic mice. Circulation 2003;108: 512 – 5. [9] Cooley BC, Szema L, Chen CY, Schwab JP, Schmeling G. A murine model of deep vein thrombosis: characterization and validation in transgenic mice. Thromb Haemost 2005;94: 498 – 503. [10] Cooley BC, Chen CY, Schmeling G. Increased venous versus arterial thrombosis in the Factor V Leiden mouse. Thrombosis Res 2007;119:747 – 51. [11] Romson JL, Haack DW, Lucchsei BR. Electrical induction of coronary artery thrombosis in the ambulatory canine: a model for in vivo evaluation of anti-thrombotic agents. Thromb Res 1980;17:841 – 53. [12] Barker JH, Gu JM, Anderson GL, O’Shaughnessy M, Pierangeli S, Johnson P, et al. The effect of heparin and dietary fish oil on embolic events and the microcirculation downstream from a small-artery repair. Plast Reconstr Surg 1993;91: 335 – 43. [13] Pierangeli SS, Barker JH, Stikovac D, Ackerman D, Anderson G, Barquinero J, et al. Effect of human IgG antiphospholipid antibodies on an in vivo thrombosis model in mice. Thromb Haemost 1994;71:670 – 4. [14] Pierangeli SS, Liu XW, Barker JH, Anderson G, Harris EN. Induction of thrombosis in a mouse model of IgG, IgM and IgA immunoglobulins from patients with the antiphospholipid syndrome. Thromb Haemost 1995;74:1361 – 7.
intl.elsevierhealth.com/journals/thre
REGULAR ARTICLE
Electric injury model of murine arterial thrombosis Akiko Kusada b , Noritaka Isogai b , Brian C. Cooley a,⁎ a
Medical College of Wisconsin, Department of Orthopaedic Surgery, 8701 Watertown Plank Road, 53226, Milwaukee, Wisconsin, United States b Department of Plastic Surgery, Kinki University, Osaka, Japan Received 3 July 2006; received in revised form 16 January 2007; accepted 8 March 2007 Available online 25 April 2007
Abstract Background and Purpose: Murine models of arterial thrombosis have gained utility with applications in genetically manipulated mice. Implementation of current models requires specialized equipment and provides limited outcome measures. A new murine model of continuously monitored arterial thrombosis was created. Methods: An electric injury was delivered to the exterior surface of the common carotid artery using the flat end of a 140-μm steel microsurgery needle connected to the anode of a 3-V battery source. Direct current was delivered for 30 s. The developing thrombus was apparent as a white, platelet-dominated region at the site of injury. This region was continuously monitored and recorded by videotape for 30 min. Subsequently, the thrombus area was measured directly on the TV monitor, generating a time course for thrombogenesis. In a further evaluation of the model, three pharmacologically treated groups of mice were evaluated, with drug infusion immediately before thrombus induction: (1) saline (control), (2) heparin (60 units/kg), and (3) GR144053, a GPIIb/IIIa-specific antagonist (10 mg/kg). Results: The basic model showed consistent thrombus development by 7–9 min, occasionally forming an occlusive thrombus. Most of the thrombi underwent one or more cycles of embolization and thrombus regrowth. In the experimental series, the heparin-treated group had a significantly decreased thrombus area versus controls (p b 0.0001); the GR144053-treated mice had no apparent thrombus, supporting a dominant role of platelet aggregation in arterial thrombogenesis. Conclusion: This new model is simple to do, uses readily available instrumentation, and provides a continuously recorded quantifiable measure of thrombogenesis. © 2007 Elsevier Ltd. All rights reserved.
Introduction
⁎ Corresponding author. Tel.: +1 414 456 8729; fax: +1 414 456 6488. E-mail address: [email protected] (B.C. Cooley).
Arterial thrombosis has many clinical manifestations for which a better understanding of the causes and therapeutic options is needed. Murine models of
0049-3848/$ - see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2007.03.005
104 arterial thrombosis have been developed to take advantage of the ever increasing numbers of genetically manipulated mouse lines by which the isolated study of single-gene effects on thrombotic outcomes can be undertaken. The most frequently used murine models of large-artery (e.g., carotid) thrombosis are the ferric chloride (filter-strip) model, originally developed in rats by Kurz and associates, [1] and the Rose Bengal model [2,3]. Both of these models entail free-radical-generated induction of vascular injury, with monitoring of flow to record the “time to occlusion” as the quantitative end-point measure. Both models have shown utility in a number of studies [4–8], but they require investment in a rather expensive and limited piece of equipment, an ultrasonic flowmeter. The Rose Bengal model also requires illumination with an intense light or laser light source for photo-activated free-radical generation. We undertook the development of a simpler model which makes use of older but functional equipment often available in small animal surgery labs — an operating/dissecting microscope and standard (nondigital) video-camera and recorder. The model uses a unique but inexpensive method for delivering a small, precise injury to the vessel surface to stimulate thrombogenesis. Furthermore, rather than one endpoint (time to occlusion), the evolution of a nonocclusive thrombus can be monitored and measured continuously, yielding a more detailed, quantifiable time profile of thrombogenesis.
A. Kusada et al. The site of electric injury/thrombus induction was monitored and recorded for 30 min at high magnification through a C-mount-coupled analog video-camera
Materials and methods Adult female CD1 mice (2–4 months old) were used following NIH and institutional guidelines for the care and use of laboratory animals. Anesthesia was induced with intraperitoneal pentobarbital (50 mg/kg). The left common carotid artery was exposed through a midline cervical incision. All procedures were done with a Wild/Leica operating microscope, using a 100-mm objective lens. A marking stitch of 10-0 nylon was placed upstream of the intended injury site. The nylon suture of a 140-μm stainless steel microsurgery needle (Product #AK-0105, Surgical Specialties Inc., Reading, PA) was removed. The blunt/flat end of this needle (where the suture was previously swaged) was held against the top exterior surface of the carotid artery; a direct, 3-V electric current was applied to the needle with two 1.5-V C batteries, connecting the positive terminal (anode) to the needle and completing the circuit with the cathode connected to a wire in contact with the cervical musculature. Current was delivered for 30 s and the needle contact with the vessel was subsequently terminated.
Figure 1 Micro-photographs (taken through an operating microscope) of the electric injury site and developing thrombus. (a) Normal carotid artery; (b) discolored/blanched site of electric injury (between arrows) 3 min after injury; developing thrombus, seen at (c) 8 min and (d) 15 min; (e) embolized thrombus (arrowhead) breaking off; f) occlusive thrombus with evident thrombi blocking flow (arrows) and with decreased vessel diameter on right. Flow is from left to right in all images; s = marking suture proximal to electric injury site.
Electric injury model of murine arterial thrombosis (COLCAM NTSC-1, World Precision Instruments, Sarasota, Florida, USA) connected to a standard VHS videotape recorder. The videotape time display was sufficient (in subsequent playback mode) to provide timing of events. A millimeter ruler was recorded at the same magnification to provide a calibrated measure for subsequent quantitation of on-screen images. Upon creating a 30-min recording of thrombogenesis, the videotape was analyzed with on-screen morphometric measurements. The area of thrombus as captured by the videotape was clearly identifiable by its characteristic white appearance, in contrast to the red blood cell-dominated flowing portion of the artery. This area was measured at 1-min intervals, using a calibrated acetate overlay on the TV monitor to measure the thrombus dimensions and generating a time profile for thrombogenesis of each sample. After pilot studies to ascertain the above parameters and methodology, a simple pharmacologic experiment for inhibiting thrombosis was undertaken. Three treatment groups of mice (n = 10/group) were randomly given one of the following: heparin at a dose of 60 units/kg; GR144053 (Sigma Chemical Co., St. Louis, MO), a specific platelet glycoprotein IIb/IIIa (GP IIb/IIIa) inhibitor, at a dose of 10 mg/kg; and normal saline (control). All agents were infused into the jugular vein in a 0.1-ml volume immediately prior to thrombus induction.
Results The 30-s application of electric injury caused an immediate yellowing discoloration at the contact point with the carotid artery (Fig. 1a and b). A white thrombus was seen to begin growing at the electricinjury zone approximately 5 min later, with a downstream progression over time (Fig. 1c–e). Partial or near-complete embolization of the thrombus was usually seen, followed by regrowth of the thrombus; this embolization/regrowth often occurred several
Figure 2 Three representative traces of controls showing thrombus areas over time, with cyclic patterns of growth, embolization (asterisks), and regrowth.
105
Figure 3 Line graph of mean thrombus areas over time for mice treated with saline infusion (solid squares), heparin (60 units/kg) (open diamonds), or GR144053 10 mg/kg) (open circles); infusion of the respective agents was given immediately before thrombus induction. Error bars = SEM; n = 10 per group.
times for a given sample. The growth and cyclic embolization were reflected in the time course of measured thrombus area (Fig. 2). Occasionally, an occlusive thrombus was seen, which was evident by a lack of distal pulsation, relative reduction in vessel diameter, and stabilization of the thrombus size (Fig. 1f). In the pharmacologic-treatment series, the average area of thrombus increased over the first 7–9 min in the saline-and heparin-treated groups, achieving an average stable size for 10–30 min (Fig. 3) that was significantly larger in the saline-versus heparin-treated group (p b 0.0001; ANOVA). The heparin dose used in this study (60 units/kg) gave a consistently prolonged activated partial thromboplastin time (aPTT, of greater than 120 s; data not shown), indicated full anticoagulation. None of the GR144053treated vessels showed a measurable thrombus.
Discussion The model presented herein is a simple, inexpensive method for continuous quantitative monitoring of an acute arterial thrombus. The induction of thrombosis is similar to our previously described murine venous thrombosis model [9,10] which uses a smaller microsurgical needle (75-μm diameter) placed inside the vein and a lower voltage (1.5-V) electric stimulus delivered over a longer (60-s) time. Our initial efforts to place a needle through the wall and inside the artery were frustrated by high amounts of bleeding. Thus, we developed an outer-surface injury which was highly reproducible and technically easy to duplicate. In other pilot work, we found that the use of noniron-containing wires (zinc, copper, gold-coated) with the same electric current delivery to the carotid artery caused a similarly evident “heat” injury (with
106 discoloration of the vessel wall), but did not result in a thrombus (data not shown). This strongly suggests a free-radical mechanism for thrombus induction, generated by ferric ion electrolytic delivery to the vessel by an iron-containing wire/needle. Lucchesi and colleagues found a similar effect in their original development of a canine coronary artery thrombosis model with intralumenal steel coils [11]. Thus, our new model has similarities in the mechanisms of thrombus induction to both the ferric chloride [1] and the Rose bengal [2,3] models of free-radical-mediated thrombogenesis. An advantage of our model is its direct quantitation of the thrombus, in comparison to the less direct monitoring of blood flow for onset of occlusion. Our electric-injury model is similar to the thrombus observation of a rat thrombogenic arterial injury model, originally developed by Acland and colleagues, [12] in which transillumination of the iliac artery revealed a developing white thrombus and by which thromboembolic events could be observed. In our model, transillumination was not necessary, as the white thrombus could be clearly seen growing, remodeling, and embolizing with the incident microscope light. Pierangeli and colleagues [13,14] used a “pinch” model of venous injury in mice, measuring the area of the developing thrombus, as was done in our study. This method of evaluation, like ours, is not a true thrombus volume, but gives a reasonable estimate. We found a profound inhibition of the arterial thrombus with infusion of the GP IIb/IIIa antagonist, GR144-53, in comparison to heparin-treated and control mice. This strongly implicates the dominance of platelets in a developing arterial thrombus. In contrast, heparin infusion, at a dose that has shown effectiveness in other models, only had a moderate effect (Fig. 3), despite a high anticoagulated state achieved with this dose of heparin. Thus, our electricinjury model may provide a better means for evaluating platelet-dominated arterial thrombosis in mice.
References [1] Kurz DK, Main BW, Sandusky GE. Rat model of arterial thrombosis induced by ferric chloride. Thromb Res 1990;60: 269 – 80.
A. Kusada et al. [2] Kawasaki T, Kaida T, Arnout J, Vermylen J, Hoylaerts MF. A new animal model of thrombophilia confirms that high plasma factor VIII levels are thrombogenic. Thromb Haemost 1999;81: 306 – 11. [3] Matsuno H, Kozawa O, Niwa M, Ueshima S, Matsuo O, Collen D, et al. Differential role of components of the fibrinolytic system in the formation and removal of thrombus induced by endothelial injury. Thromb Haemost 1999;81:601 – 4. [4] Weiler H, Lindner V, Kerlin B, Isermann BH, Hendrickson SB, Cooley BC, et al. Characterization of a mouse model for thrombomodulin deficiency. Arterioscler Thromb Vasc Biol 2001;21: 1531– 7. [5] Kerlin B, Cooley BC, Isermann BH, Hernandez I, Sood R, Zogg M, et al. Cause–effect relation between hyperfibrinogenemia and vascular disease. Blood 2004;103:1728 – 34. [6] Falati S, Patil S, Gross PL, Stapleton M, Merrill-Skoloff G, Barrett NE, et al. Platelet PECAM-1 inhibits thrombus formation in vivo. Blood 2006;107:535 – 41. [7] Wilson KM, Lynch CM, Faraci FM, Lentz SR. Effect of mechanical ventilation on carotid artery thrombosis induced by photochemical injury in mice. J Thromb Haemost 2003;1: 2669 – 74. [8] Danenberg HD, Szalai AJ, Swaminathan RV, Peng L, Chen Z, Seifert P, et al. Increased thrombosis after arterial injury in human C-reactive protein-transgenic mice. Circulation 2003;108: 512 – 5. [9] Cooley BC, Szema L, Chen CY, Schwab JP, Schmeling G. A murine model of deep vein thrombosis: characterization and validation in transgenic mice. Thromb Haemost 2005;94: 498 – 503. [10] Cooley BC, Chen CY, Schmeling G. Increased venous versus arterial thrombosis in the Factor V Leiden mouse. Thrombosis Res 2007;119:747 – 51. [11] Romson JL, Haack DW, Lucchsei BR. Electrical induction of coronary artery thrombosis in the ambulatory canine: a model for in vivo evaluation of anti-thrombotic agents. Thromb Res 1980;17:841 – 53. [12] Barker JH, Gu JM, Anderson GL, O’Shaughnessy M, Pierangeli S, Johnson P, et al. The effect of heparin and dietary fish oil on embolic events and the microcirculation downstream from a small-artery repair. Plast Reconstr Surg 1993;91: 335 – 43. [13] Pierangeli SS, Barker JH, Stikovac D, Ackerman D, Anderson G, Barquinero J, et al. Effect of human IgG antiphospholipid antibodies on an in vivo thrombosis model in mice. Thromb Haemost 1994;71:670 – 4. [14] Pierangeli SS, Liu XW, Barker JH, Anderson G, Harris EN. Induction of thrombosis in a mouse model of IgG, IgM and IgA immunoglobulins from patients with the antiphospholipid syndrome. Thromb Haemost 1995;74:1361 – 7.