Lack of P-selectin glycoprotein ligand-1 protects mice from thrombosis after collagen/epinephrine challenge

Page 1

Thrombosis Research 127 (2011) 228–234

Contents lists available at ScienceDirect

Thrombosis Research

j ourna l homepage: www.e lsev ie r.com/ locate / th romres

Regular Article

Lack of P-selectin glycoprotein ligand-1 protects mice from thrombosis aftercollagen/epinephrine challenge

Kornél Miszti-Blasius a, Ildikó Beke Debreceni a, Szabolcs Felszeghy b, Balázs Dezső c, János Kappelmayer a,⁎a Department of Clinical Biochemistry and Molecular Pathology, Hungaryb Department of Anatomy, Histology and Embryology, Hungaryc Department of Pathology, Medical and Health Science Center, University of Debrecen, Hungary

⁎ Corresponding author. Department of ClinicalPathology, Medical and Health Science Center, Univers+36 52 340 006; fax: +36 52 417 631.

E-mail address: [email protected] (J. Ka

0049-3848/$ – see front matter © 2010 Elsevier Ltd. Aldoi:10.1016/j.thromres.2010.11.022

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 4 July 2010Received in revised form 27 October 2010Accepted 26 November 2010Available online 14 January 2011

Keywords:PSGL-1ThrombosisFibrinCollagenMouse

Introduction: In thrombotic processes, during the association of leukocytes with platelets and endothelialcells, P-selectin glycoprotein ligand-1 (PSGL-1) binds to P-selectin, expressed on activated platelets andendothelial cells. Our aim was to establish the role of PSGL-1 in thrombus formation by evaluating theresponse to thrombotic stimuli in wild type and PSGL-1 knockout mice.Materials and methods: Mice were challenged by tail vein injection of (i) 15 μg collagen plus 3 μg epinephrine(coll/epi) (ii) 7.5 μg collagen plus 1.5 μg epinephrine or (iii) saline. Retro-orbital blood samples were collectedin ACD anticoagulaed tubes and platelet and leukocyte counts were measured. In addition, kidneys, liver,spleen and lungs were investigated for fibrin deposition by immunohistochemistry and Western-blotting.Frozen sections were analysed for double labeling for platelet and leukocyte presence.Results: After coll/epi challenge, the number of platelets and leukocytes decreased significantly in bothgenotypes. Lower agonist concentration resulted in an attenuated platelet decrease in PSGL-1 knockout mice

compared to the controls, however changes in leukocyte and neutrophil counts were not significantlydifferent in the two strains. In knockout mice considerably less fibrin deposition has been observed in thelungs by Western-blotting and immunohistochemistry. After coll/epi challenge the lungs of the PSGL-1knockout animals contained both platelets and leukocytes but less thrombi has been detected than in wild-type mice.Conclusions: Our results indicate that the deficiency of PSGL-1 results in milder thrombocytopenia, less fibrindeposition and lower number of thrombosed blood vessels, suggesting that this molecule is essential formulticellular interactions during thrombus formation.

© 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Clot formation is critical for limiting posttraumatic blood loss, butthe excessive process can occlude diseased vessels leading to stroke,myocardial infarction or atherosclerosis which are the leading causesof death in the Western world. Several different molecules facilitatethe activation of platelets by primary stimuli, potentially transforminga normal hemostatic response into the formation of an occlusivethrombus [1,2]. The exposure of subendothelial collagen is a majorcontributor to such thrombotic disorders and along with otherphysiologially occurring platelet agonists like epinephrine, ADP andthrombin are responsible for platelet activation [3]. Activated plateletsrapidly mobilize P-selectin (CD62P, formerly known as PADGEM orGMP-140) to their surface from their alpha granules [4]. Via thesurface expressed P-selectin, platelets can form heterotypic interac-

Biochemistry and Molecularity of Debrecen, Hungary. Tel.:

ppelmayer).

l rights reserved.

tions with leukocytes [5,6]. The counter-receptor on the leukocytesurface is the P-Selectin Glycoprotein Ligand-1 (PSGL-1). Thissialomucin is the ligand for selectins which plays a central role inthe formation of primary thrombus formation and also mediatesleukocyte tethering and rolling [7–9]. This cell-cell interaction iscritical during coagulation, since platelets roll on activated endothelialcells, while leukocytes roll and arrest on activated platelets andendothelial cells [6,10–12]. Adhesion of leukocytes to platelets formsplatelet-leukocyte aggregates that are deposited at the site of vascularinjury [13–19]. The quantity of platelet-monocyte and platelet-neutrophil aggregates was found to be a sensitive marker for plateletactivation both in animal experiments [20] and in human diseases[21].

Several mouse models have been developed to mimic thepathophysiology of thrombosis by using laser, ferric chloride,lipopolysaccharide, collagen or ADP induction [19,22–30]. In thisstudy we aimed to administer collagen plus epinephrine via tail veinto PSGL-1+/+ and PSGL-1-/- mice to determine the role of PSGL-1 oncell counts, thrombus formation and survival rate in a large number ofanimals challenged with different doses of the agonists. Based on our

Page 2

Fig. 1. Exit rates of wild-type and knockout mice within 30 minutes after thromboticchallenge. The percentage of perished wild type mice (open bars) and knockout mice(solid bars) after administration of full dose and half dose collagen+epinephrine (Fulldose: 15 μg collagen+3 μg epinephrine, Half dose: 7.5 μg collagen+1.5 μg epineph-rine). Owerall 87 animals were investigated in the following distribution: PSGL-1+/+

full dose n=13, half dose n=30, PSGL-1-/- full dose n=15, half dose n=29.

229K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

results we feel confident to conclude that the lack of PSGL-1 isprotective during thrombus formation elicited by platelet activation.

2. Materials and methods

2.1. Laboratory animals

C57B6/126 PSGL-1WTandKOmicewerepreviously described.Micewere weaned at 4 weeks, maintained on a 12-hour light – 12-hour darkcycle at 21 °C, and fed water and standard rodent chow (VRF1 CharlesRiver, Germany) ad libitum. All procedures conformed with the recom-mendations of the Hungarian law, the study protocol was approved bythe local Ethical Committee (1/2006 DE MÁB) and tissue samples wereobtained in accordance with guidelines.

2.2. Antibodies and detection reagents

Polyclonal rabbit anti-human fibrinogen antibody - that crossreacts with mouse fibrin(ogen)- was from DAKO (Glostrup,Denmark); mAb 350 specific for the heptapeptide sequence exposedat the new amino-terminal of the ß-chain of fibrin after thrombincleavage of fibrinopeptide B was from American Diagnostica (Green-wich CT, USA). HRP-conjugated anti-rabbit IgG was obtained from Rand D Systems (Minneapolis, MN), anti-actin - developed in rabbits -was from Sigma (St Louis, MO), R-Phycoerytrin (R-PE)-ConjugatedRat Anti-Mouse CD162 (PSGL-1) monoclonal antibody was from BDPharmingen (Becton Dickinson Franklin Lakes, NJ). For visualisationthe EnVision Detection Systems (Dako, Ely, UK) was used. Anti-mouseCD14 PE, anti-mouse CD41 FITC and non-conjugated rat anti-mouseCD162 (clone 4RA10) were from (Becton Dickinson Franklin Lakes,NJ).

2.3. Fibrinogen and fibrin generation

Mouse fibrinogen was from Sigma (St Louis, MO), while mousefibrin was prepared by incubating 0.25 mg/mL fibrinogen in 150 mMNaCl – 10 mM TRIS at pH: 7.4 with 2 U/mL thrombin from humanplasma (Sigma, St Louis, MO) at 37 °C for 15 minutes. Aftercentrifugation (13 500 g, 4 °C, 35 minutes) the supernatant wasdiscarded, the sediment was dissolved in SDS PAGE sample buffer(62.5 mM TRIS-HCl, 2% SDS, 10% Glycerol, 0.1% Bromphenol blue, 5%Mercaptoethanol at pH: 6.5) and the mixture was boiled for10 minutes.

2.4. Models of systemic intravascular thrombosis

Wild-type (PSGL-1 +/+) and PSGL-1 deficient (PSGL-1 -/-) malemice matched for age (12-16 weeks) were anesthetized with a singledose of intraperitoneal ketamine. For „full dose” experiment 15 μgcollagen (Collagen Reagent Horm, Nycomed Munich, Germany) and3 μg epinephrine (Richter Gedeon, Budapest, Hungary) was injectedvia tail vein and for „half dose” experiment 7.5 μg collagen and 1.5 μgepinephrine was administered. Three minutes after the injectionblood samples were drawn by puncture of the retrobulbar venousplexus, the survival/exit rate was investigated up to the 30th minuteafter which living mice were euthanized by cervical dislocation.

2.5. Blood collection and cell counting

Mice were anesthetized with intraperitoneal ketamine (SelBruHaKft, Budapest, Hungary) injection (150 mg/kg), and 200 μL blood wascollected into a cup containing 40 μL ACD (from Vacutainer tubeBecton Dickinson Diagnostics-Preanalytical Systems Plymouth, UK)by puncture of the retrobulbar venous plexus with a 30 mm long glasscapillary. ACD anticoagulated whole blood was analyzed by SiemensAdvia-120 hematology analyzer (Deerfield, IL) [31].

2.6. Lung fibrin determination

Lungs were harvested and then rinsed at 4 °C in extraction buffercomposed of 150 mMNaCl, 10 mM EDTA, 1 mM phenylmethylsulfonylfluoride (Sigma), 10 U/mL aprotinin (Richter, Hungary), 100 U/mLheparin (Merckle GmBH, Ulm, Germany), 0.1 M E-aminocaproic acid(Pannonpharma, Pécsvárad, Hungary) and 10 mM Tris/HCl, pH 7.4.Samples were frozen in cryovials on the surface of liquid nitrogen andstored at -20 °C until further use. Lung tissuewas thawed by immersingthe cryovial in 37 °C waterbath for 5 minutes, minced, homogenized inextraction buffer at 0,5 mL buffer/100 mg tissue for 10 minutes, andincubated on ice for 4 hours. The pellet, obtained after centrifugation at16 000 g for 30 minutes at 4 °C, was washed twice, resuspended in200 μL sample buffer without bromphenol blue and mercaptoethanol.Samples were incubated for 18 hours at 37 °C, and recentrifuged at 16000 g for 30 minutes at room temperature. To the supernatantcontaining the extracted fibrin we added 5% mercaptoethanol and0.1% bromphenol blue, samples were separated by SDS-polyacrylamidegel electrophoresis on a 7.5% gel and transferred to Immobilon PTransfer membrane (Millipore, Bedford, MA) for immunoblotting.Fibrin(ogen) was detected by reacting membranes first with a 1:2000dilution of rabbit anti human fibrinogen (DAKO, Glostrup, Denmark),followed by horseradish peroxidase-conjugated antirabbit antibody.Bands were visualized with enhanced chemiluminescence. To quantifyfibrinogen and fibrin, the optical density of bands derived from samplescontainingequal amounts of lung tissuewere comparedandnormalizedfor actin signal by scanning densitometry.

2.7. Tissue processing for histology and immunohistochemistry

For comparative light microscopy analysis, tissue samplesobtained from both animal groups were fixed with 10% phosphate-buffered formalin, embedded in paraffin, and stained with hematox-ylin-eosin or Masson trichrome. For immunohistochemical analysis5 μm sections were deparaffinized, rehydrated followed by antigenunmasking in pressure cooker (120 °C, 3 mins) using antigen retreivalsolution (pH 6.0, DAKO, Glostrup, Denmark). Sections were thentreated with 1% H2O2 in methanol for 30 mins at room temperature toblock endogenous peroxidase. Following pretreatment with 3% non-fat dry milk in background reducing solution (DAKO), sections were

Page 3

Fig. 2. A: Changes in platelet count after applying various collagen doses. Collagen+epinephrine challenge was carried out in 2 doses as before. On these and subsequentfigures mean values and standard errors are displayed. Platelet counts decreasedsignificantly in a dose-dependent manner. PSGL-1+/+ (open bars) saline n=25, fulldose n=7, half dose n=20, PSGL-1-/- (solid bars) saline n=24, full dose n=8, halfdose n=28. B: Leukocyte counts after thrombotic challenge. Leukocyte counts in wild-type mice (open bars) and knockout mice (solid bars) after administration of saline orcollagen+epinephrine in two different doses. Animal counts are the same as in Fig. 2A.C: Neutrophil counts after thrombotic challenge. In a limited number of mice absoluteneutrophil counts were evaluated before and after thrombotic chgallenge. Both totalleukocyte and absolute neutrophil counts decreased in a similar manner in the twostrains. Results from 9 wild type and 6 knockout animals are displayed.

230 K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

incubated with mAb350 (1:40 dilution) at room temperature for 1hour. After washing with PBS, visualisation was carried out by the useof biotin-free EnVision detection kit (DAKO) utilizing HRP (horserad-ish-peroxidase) and VIP for chromogenic substrate as previously [32].To check the specificity of immunostaining, negative controls wereincluded where no antibody or irrelevant control monoclonal anti-body (DAKO)was used in place of the primary antibody which did notshow labeling. To quantitate the accumulation of thrombi in the lungtissue sections, the number of occluded vessels in 12 randomlyselected high magnification microscopic fields from 4 mice in eachtreatment group were analyzed.

2.8. Immunofluorescence staining on frozen tissue sections

Samples were snap frozen and cut transversally with a cryostatmicrotome (Leica Instruments, Nussloch, Germany). The 5 μm cryo-sections were fixed in methanol at -20 °C for 5 min. After blocking in1% BSA-PBS, the samples were further incubated in the presence ofalexa-labeled antibodies to CD61-AF488 and CD 45-AF647 forovernight at 4 °C. Following multiple rinsing in PBS, the sampleswere mounted on glass slides using Vectashield medium containingDAPI (Vector Laboratories Inc, Peterborough, UK) for DNA counter-stain. No signal was recorded from control sections incubated withnon-immune IgG instead of the primary antibody. Fluorescent imageswere captured by a Nikon Eclipse 800 microscope equipped with aSpot RT-slider (Diagnostic Instruments, Sterling Heights, MI, USA)CCD camera. Acquired and presented images were representative ofall the samples examined. For documentation, images were processedusing Adobe PhotoShop software (Version 5.5, Adobe Systems Inc.,CA).

2.9. Platelet-leukocyte aggregate formation

Fifty μl of ACD-anticoagulated mouse blood was simultaneouslylabeled by rat anti-mouse CD14 PE and rat anti-mouse CD41 FITC for15 minutes at room temperature. Leukocyte subsets were identifiedby the SSC-FL2 dot plots and heterotypic aggregates were measuredby the CD41 fluorescence values in respective leukocyte gates. TRAPstimulation at 37 °C for 15 minutes was used to enhance heterotypicaggregate formation using 10 μmol/L of the agonist. For blockingexperiments blood samples were preincubated by 2 μg/mL rat anti-

Fig. 3. Western-blot analysis of lung extracts. Lung extracts from mice were separated by7.5% SDS polyacrylamide gel electrophoresis and subjected to immunoblot analysis withanti-human fibrin(ogen) that crossreacts with mouse fibrin(ogen). Panel A. lane 1:Fibrinogen standard, lane 2: Fibrin standard, lane 3: PSGL-1+/+ saline, lane 4: PSGL-1+/+

collagen+epinephrine, lane 5: PSGL-1-/- saline, lane 6: PSGL-1-/- collagen + epinephrine.Panel B: After thrombotic challenge, twice asmuchfibrindepositswere found in PSGL-1+/+

mice (n=7)when compared to PSGL-1-/- mice (n=8). (1 Relative unit=optical density offibrin, in mice treated with saline).

Page 4

231K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

mouse PSGL-1 or rat IgG as negative control for 15 minutes. Afterwashing in PBS, cells were processed as above.

2.10. Statistical analyses

For comparison of exit rates of challenged mice results wereevaluated by the Chi-square test. One Way ANOVA was used for theevaluation of thrombosed vessels. Cell counts were compared byStudent's T-test.

3. Results

3.1. The effect of thrombotic challenges on survival rate

The survival rate of mice was evaluated in 87 animals, 30 minutesafter the thrombotic challenge. Upon administration of the full dose

Fig. 4. Histological evaluation of lung sections. Histology revealed increased accumulation ofas compared to PSGL-1-/- mice. Five μm thick sections were stained with haematoxylin-eosinthe pulmonary vasculature. Fewer thrombi were observed in PSGL-1-/- mice. Arrowhead showantibody to fibrin (mAb350) - that crossreacts with mouse fibrin - was used to detect fibrin dantibody - no signal was detected (data not shown). Thrombi seemed more prevalent in PS

collagen and epinephrine (15 μg and 3 μg, respectively) 8 mice out of13 (62%) died from the PSGL-1+/+ group and 7 of the 15 mice (47%)from the PSGL-1-/- strain within 30 minutes (p=0.476). Reducing theconcentration of the thrombogenic agents by 50% (7.5 μg collagen and1.5 μg epinephrine) resulted in a considerable difference in survivalrates since 15 animals of 30 (50%) in the PSGL-1+/+ group died whileonly 3 out of 29 (10%) in the PSGL-1-/- group (p=0.002) (Fig. 1).

3.2. Cell counts in survived animals

Collagen and epinephrine administration resulted in a dose-dependent thrombocytopenia. Platelet count decreased significantly(pb0.0001) in both groups after applying the full dose challenge andthe difference between the two strains was not significant(p=0.261). On the contrary, after administrating half dose of theprothrombotic agents, the platelet count reduction was significantly

thrombi in the lungs of PSGL-1+/+ mice injected with half dose collagen+epinephrine(panels A and B) Masson-trichrome staining (panels C, D). Arrows indicate thrombi ins a near-occlusive thrombus in a knockout mouse (panel D). A monoclonal anti-humaneposition in the lungs (panels E, F). In negative controls - sections without the primaryGL-1+/+ compared to PSGL-1-/- mice.

Page 5

232 K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

less in the PSGL-1-/- mice (p=0.0325) (Fig. 2A). The decrease inleukocyte count was similar in the two groups and was independentfrom the applied dose of collagen and epinephrine (Fig. 2B). Sinceperipheral blood neutrophil count is significantly elevated in theabsence of PSGL-1 mice, we investigated the changes in absoluteneutrophil counts before and after the thrombotic challenge in 6 and 9animals in PSGL-1+/+ and PSGL-1 -/- strains respectively. Beforechallenge the absolute neutrophil counts were 3-times higher inknockout animals, however the values decreased in a similar mannerin boths strains (Fig. 2C).

3.3. Thrombin generation and fibrin deposition

Plasma samples derived from blood - obtained by retroorbitalsampling - were studied for thrombin generation in untreated andchallenged mice by investigating thrombin-antithrombin complexes(TAT). Thrombotic challenge by large-dose collagen/epinephrineresulted in elevated TAT complex values with no difference betweenthese strains. For positive control thromboplastin (5x diluted Innovin)challenged mice were used (data not shown). Fibrin deposition inlungs was assayed by Western blot analysis of lung fibrin extractsusing the anti-human polyclonal antibody – that crossreacts withmouse proteins – to detect both fibrin and fibrinogen. Our resultsshowed that after the injection of thrombotic agents more fibrindeposits in the lungs of PSGL-1+/+ mice than in those of PSGL-1-/-

mice (Fig. 3).

3.4. Morphological analysis of thrombi

Our immunohistochemical (IHC) analysis demonstrated that inthe lung vasculature several thrombi were formed after thromboticchallenge. First, sections were analysed after Masson trichromestaining to study thrombosed vessels. The number of thrombosedvessels was more abundant in wild type animals (Fig. 4A, C) and thesethrombi also tended to occlude large vessels. In the PSGL-1-/- animalsless thrombi has been observed and they were occlusive in largevessels (Fig. 4B, D). Fibrin could be identified immunohistochemically inthese thrombi on sections derived from both strains by using mAb350antibody that detects only fibrin but not fibrinogen. (Fig. 4 E, F). Thepercent of the thrombosed vessels [28] was found to be higher in wildtype animals (Fig. 5). After IHC analysis of the cellular composition ofthrombi, the expression pattern of anti-CD61 and anti-CD45 indicatedthat leukocyte and platelet colocalisation does occur in both strains(Fig. 6).

Fig. 5. Percentage of thrombosed vessels in the lung. Percentage of thrombosed vesselswasidentified by both Masson-trichrome staining and mAb350 immunostaining. Mean andstandard deviations of 12 microscopic fields per lungs (40x) from 3-4 mice are displayed.Significantly less occluded vesels were observed in PSGL-1-/- by both techniques.

3.5. Platelet-leukocyte aggregate formation

Neutrophils from wild type mice showed high values of CD41-fluorescence that increased upon stimulationwith TRAP. This effectwascompletely blocked by the anti-PSGL-1 antibody that also preventedTRAP induced elevation of CD41 fluorescence (Fig. 7A). Similar resultswere obtained for monocytes (data not shown). No changes inlymphocyte associated fluorescence was observed. PSGL-1-/- micewere negative for CD41-related staining for all cell types (Fig. 7B).

4. Discussion

Interactions between leukocytes, platelets and the endotheliumplay an important role in thrombosis. Platelet reactions are mediatedby specific cell adhesion molecules that generate homotypic andheterotypic adhesive interactions with other platelets, or leukocytes,endothelial cells and the extracellular matrix. Earlier studies focusedmainly on PSGL-1 mediated cell rolling [8,9,33–35] and considerablyless is known about the function of PSGL-1 in platelet aggregation andthrombosis. Although tremendous progress has been made to studythe mechanism for leukocyte activation and transmigration across aninjured vascular surface [4,8,12,34–38], limited data are availableabout the role of selectin-ligand interaction during thrombosis.

The lack of P-selectin have previously been shown to attenuatethrombus formation and cell rolling [39,40] but this has not beenproven for PSGL-1 in case of a systemic thrombosis model.

To have a better insight into this molecular mechanism, we appliedthe collagen/epinephrine challenge method [23,41]. By using theconventionally applied dose, we found that the lack of PSGL-1 showeda non-significant protective effect in mice. Introducing 50% of thethrombotic stimuli, this protection was more evident and highlysignificant differences were observed in survival rates. With the smallerdose only 10% of the PSGL-1-/- mice died in 30 minutes while 50% of thewild type animals suffered fatal thrombosis. After the large dose stimulus62% of the wild type and 47% of the PSGL-1-/- mice died, these resultssuggest that the higher dose masks the protective effect of PSGL-1.

The difference in platelet count decrease was not significantbetween the two strains studied applying full dose challenge, howeverreducing the dose to half, the platelet count reductionwas significantlyattenuated in the PSGL-1-/- mice. These data provide evidence that theabsence of this ligand may play an important role during thrombusformation and platelet-leukocyte interaction. We found a significantand similar type of decrease in neutrophil percentage in bothwild typeand PSGL-1-/- mice and also confirmed the higher percentage ofmyeloid cells that have been described previously [35].

Platelets were activated directly with collagen plus epinephrinevia tail vein to initiate thrombus formation. In our preliminaryexperiments we observed that the microthrombi deposition wasmostly evident in the lung since that is the organ that posessesnumerous small arteries, capillaries and it also has a considerableblood flow. By immunoblotting, more fibrin was detectable in wild-type mice and this result was further confirmed by microscopicevaluations. In haematoxylin-eosin stained, as well as Masson-stainedsections and sections applying monoclonal antibody staining againstcross-linked fibrin, more thrombi were detectable in PSGL-1+/+ miceand these thrombi were occlusive or near occlusive. In contrast to thisphenomenon, in PSGL-1-null mice the lumen of the blood vessels waspartially obstructed but not occluded and the percent of thrombosedvessels was significantly lower in PSGL-1-/-mice. We suggest, that thelower fibrin deposition in PSGL-1-null mice is related to the formationof less thrombi in these animals. Platelets and leukocytes could beobserved in co-localisation in thrombi on sections derived from bothstrains.

To provide further evidence to the knockout approach we couldprove the role of PSGL-1 in thrombotic processes in wild type animals

Page 6

Fig. 6. Immunofluorescent staining on frozen tissue sections. Trombi in both strains contained platelets and leukocytes in the lung. Detection of CD61 (A, D) and CD45 (B, E)identified positive structures and the labelings were colocalized (C, F).

233K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

by using the heterotypic aggregate model, where a blocking antibodyto PSGL-1 completely abolished the platelet binding of neutrophils.

Genetically modified mice are useful to develop different modelsfor understanding, preventing and treating human diseases [9,14,42].Although mice have higher platelet count, different signal transduc-tion and different vascular rheology, the thrombosis model describedabove can provide helpful data about human thrombus generationand can help in development of drugs that block P-selectin-PSGL-1interaction in arterial thrombosis [29]. The expression of PSGL-1 iswidespread in the body and aside from the myeloid, lymphoid anddendritic cell lineage [11] endothelial cells [8] and platelets [2] havealso been shown to be a source of this protein.

Our findings indicate, that the lack of PSGL-1 can preventexperimental thrombosis. This effect may be related to the wide-spread expression of this mucin and to the multitude of intercellularreactions that aremediated via PSGL-1. In human practicemonoclonalantibodies against P-selectin, PSGL-1, L-selectin and E-selectin can beapplied to inhibit leukocyte rolling and protect patients from

myocardial reperfusion injury [42]. Platelet-leukocyte interactionsand the subsequent endothelium activation are important factors inthrombus formation and these clinical data are in line with ourexperimental observations and suggest the beneficial effect of PSGL-1inhibition during atherothrombotic disorders.

Conflict of interest statement

The authors declare no conflict of interest.

Acknowledgement

The authors would like to acknowledge Dr. Rodger McEver andLijun Xia (Oklahoma Medical Research Foundation, Oklahoma City,OK USA) for providing the PSGL-1 knockout mice. The excellenttechnical assistance of Katalin Orosz-Tóth, Tünde Terdik Pál andMáriaBessenyei are acknowledged. This study was supported by grant:OTKA K75199 (J.K.) and Marie Curie Reintegration Grant (Sz.F.).

Page 7

Fig. 7. Fluorescence intenisty of CD41 labeling in leukocytes. Leukocyte platelet aggregateswere identified by the CD41 signal in respective leukocyte gates in PSGL-1+/+ (panel A)and PSGL-1-/- (panel B) animals. Wild type neutrophils displayed high values that wasfurther augmented after TRAP stimulation. Signals could be completely eliminated by ananti-PSGL-1 blocking antibody. No CD41 fluorescence signal was observed in PSGL-1-/-

mice. White bars represent neutrophils, while black bars represent lymphocytes.

234 K. Miszti-Blasius et al. / Thrombosis Research 127 (2011) 228–234

Appendix A. Supplementary data

Supplementary data to this article can be found online atdoi:10.1016/j.thromres.2010.11.022.

References

[1] Cambien B,Wagner DD. A new role in hemostasis for the adhesion receptor P-selectin.Trends Mol Med 2004;10:179–86.

[2] Frenette PS, Denis CV, Weiss L, Jurk K, Subbarao S, Kehrel B, et al. P-Selectinglycoprotein ligand 1 (PSGL-1) is expressed on platelets and can mediate platelet-endothelial interactions in vivo. J Exp Med 2000;191:1413–22.

[3] Leon C, Alex M, Klocke A, Morgenstern E, Moosbauer C, Eckly A, et al. Platelet ADPreceptors contribute to the initiation of intravascular coagulation. Blood 2004;103:594–600.

[4] Kappelmayer J, Nagy Jr B, Miszti-Blasius K, Hevessy Z, Setiadi H. The emergingvalue of P-selectin as a disease marker. Clin Chem Lab Med 2004;42:475–86.

[5] Klinkhardt U, Dragutinovic I, Harder S. P-selectin (CD62p) and P-selectin glycopro-tein ligand-1 (PSGL-1) polymorphisms: minor phenotypic differences in theformation of platelet-leukocyte aggregates and response to clopidogrel. Int J ClinPharmacol Ther 2005;43:255–63.

[6] McEver RP. Adhesive interactions of leukocytes, platelets, and the vessel wallduring hemostasis and inflammation. Thromb Haemost 2001;86:746–56.

[7] Asaduzzaman M, Rahman M, Jeppsson B, Thorlacius H. P-selectin glycoprotein-ligand-1 regulates pulmonary recruitment of neutrophils in a platelet-independentmanner in abdominal sepsis. Br J Pharmacol 2009;156:307–15.

[8] da Costa Martins P, García-Vallejo JJ, van Thienen JV, Fernandez-Borja M, van GilsJM, Beckers C, et al. P-selectin glycoprotein ligand-1 is expressed on endothelialcells and mediates monocyte adhesion to activated endothelium. ArteriosclerThromb Vasc Biol 2007;27:1023–9.

[9] Dold S, Laschke MW, Zhau Y, Schilling M, Menger MD, Jeppsson B, et al. P-selectinglycoprotein ligand-1-mediated leukocyte recruitment regulates hepatocellulardamage in acute obstructive cholestasis in mice. Inflamm Res 2010;59:291–8.

[10] Kappelmayer J, KissA,Karászi E,VeszprémiA, Jakó J,KissC. Identificationof P-selectinglycoprotein ligand-1 as a useful marker in acute myeloid leukaemias. Br J Haematol2001;115:903–9.

[11] Laszik Z, Jansen PJ, Cummings RD, Tedder TF, McEver RP, Moore KL. P-selectinglycoprotein ligand-1 is broadly expressed in cells of myeloid, lymphoid, anddendritic lineage and in some nonhematopoietic cells. Blood 1996;88:3010–21.

[12] Sperandio M, Smith ML, Forlow SB, Olson TS, Xia L, McEver RP, et al. P-selectinglycoprotein ligand-1 mediates L-selectin-dependent leukocyte rolling in venules.J Exp Med 2003;197:1355–63.

[13] Théorêt JF, Chahrour W, Yacoub D, Merhi Y. Recombinant P-selectin glycoprotein-ligand-1 delays thrombin-induced platelet aggregation: a new role for P-selectinin early aggregation. Br J Pharmacol 2006;148:299–305.

[14] Fernandes LS, Conde ID, Wayne Smith C, Kansas GS, Snapp KR, Bennet N, et al.Platelet-monocyte complex formation: effect of blocking PSGL-1 alone, and incombination with alphaIIbbeta3 and alphaMbeta2, in coronary stenting. ThrombRes 2003;111:171–7.

[15] León B, Ardavín C. Monocyte migration to inflamed skin and lymph nodes isdifferentially controlled by L-selectin and PSGL-1. Blood 2008;111:3126–30.

[16] Patel KD, Cuvelier SL, Wiehler S. Selectins: critical mediators of leukocyterecruitment. Semin Immunol 2002;14:73–81.

[17] Polgar J, Matuskova J, Wagner DD. The P-selectin, tissue factor, coagulation triad.J Thromb Haemost 2005;3:1590–6.

[18] Slotta JE, BraunOO,MengerMD,ThorlaciusH.Capture of platelets to theendotheliumof the femoral vein is mediated by CD62P and CD162. Platelets 2009;20:505–12.

[19] Falati S, Gross P, Merrill-Skoloff G, Furie BC, Furie B. Real-time in vivo imaging ofplatelets, tissue factor and fibrin during arterial thrombus formation in the mouse.Nat Med 2002;8:1175–81.

[20] MichelsonAD, BarnardMR, Krueger LA, Valeri CR, FurmanMI. Circulatingmonocyte-platelet aggregates are a more sensitive marker of in vivo platelet activation thanplatelet surface P-selectin: studies in baboons, human coronary intervention, andhuman acute myocardial infarction. Circulation 2001;104:1533–7.

[21] Kaplar M, Kappelmayer J, Veszpremi A, Szabo K, Udvardy M. The possibleassociation of in vivo leukocyte-platelet heterophilic aggregate formation and thedevelopment of diabetic angiopathy. Platelets 2001;12:419–22.

[22] Hechler B, Lenain N, Marchese P, Vial C, Heim V, Freund M, et al. A role of the fastATP-gated P2X1 cation channel in thrombosis of small arteries in vivo. J Exp Med2003;198:661–7.

[23] DiMinno G, Silver MJ. Mouse antithrombotic assay: a simple method for theevaluation of antithrombotic agents in vivo. Potentiation of antithromboticactivity by ethyl alcohol. J Pharmacol Exp Ther 1983;225:57–60.

[24] Li TT, Larrucea S, Souza S, Leal SM, López JA, Rubin EM, et al. Genetic variationresponsible for mouse strain differences in integrin alpha 2 expression isassociated with altered platelet responses to collagen. Blood 2004;103:3396–402.

[25] Matsukawa A, Lukacs NW, Hogaboam CM, Knibbs RN, Bullard DC, Kunkel SL, et al.Mice genetically lacking endothelial selectins are resistant to the lethality in septicperitonitis. Exp Mol Pathol 2002;72:68–76.

[26] Olman MA, Mackman N, Gladson CL, Moser KM, Loskutoff DJ. Changes inprocoagulant and fibrinolytic gene expression during bleomycin-induced lunginjury in the mouse. J Clin Invest 1995;96:1621–30.

[27] RennéT, PozgajováM,GrünerS, SchuhK,PauerHU,BurfeindP,et al.Defective thrombusformation in mice lacking coagulation factor XII. J Exp Med 2005;202:271–81.

[28] Smyth SS, Reis ED, Väänänen H, Zhang W, Coller BS. Variable protection of beta3-integrin–deficient mice from thrombosis initiated by different mechanisms.Blood 2001;98:1055–62.

[29] Tsakiris DA, Scudder L, Hodivala-Dilke K, Hynes RO, Coller BS. Hemostasis in themouse (Mus musculus): a review. Thromb Haemost 1999;81:177–88.

[30] Wang X. Lipopolysaccharide augments venous and arterial thrombosis in themouse. Thromb Res 2008;123:355–60.

[31] Diaz-Ricart M, Brunso L, Pino M, Navalon F, Jou JM, Heras M, et al. Preanalyticaltreatment of EDTA-anticoagulated blood to ensure stabilization of the meanplatelet volume and component measured with the ADVIA counters. Thromb Res2010;126:30–5.

[32] Gogolak P, Rethi B, Szatmari I, Lanyi A, Dezso B, Nagy L, et al. Differentiation ofCD1a- and CD1a+monocyte-derived dendritic cells is biased by lipid environ-ment and PPARgamma. Blood 2007;109:643–52.

[33] Baïsse B, Galisson F, Giraud S, Schapira M, Spertini O. Evolutionary conservation ofP-selectin glycoprotein ligand-1 primary structure and function. BMC Evol Biol2007;7:166.

[34] Fischer EG, RiewaldM, Huang HY, Miyagi Y, Kubota Y, Mueller BM, et al. Tumor celladhesion and migration supported by interaction of a receptor-protease complexwith its inhibitor. J Clin Invest 1999;104:1213–21.

[35] Xia L, Sperandio M, Yago T, McDaniel JM, Cummings RD, Pearson-White S,et al. P-selectin glycoprotein ligand-1-deficient mice have impaired leukocytetethering to E-selectin under flow. J Clin Invest 2002;109:939–50.

[36] Miner JJ, Xia L, Yago T, Kappelmayer J, Liu Z, Klopocki AG, et al. Separablerequirements for cytoplasmic domain of PSGL-1 in leukocyte rolling and signalingunder flow. Blood 2008;112:2035–45.

[37] Yao L, Setiadi H, Xia L, Laszik Z, Taylor FB, McEver RP. Divergent inducible expressionof P-selectin and E-selectin in mice and primates. Blood 1999;94:3820–8.

[38] Zarbock A, Müller H, Kuwano Y, Ley K. PSGL-1-dependent myeloid leukocyteactivation. J Leukoc Biol 2009;86:1119–24.

[39] Yamada S, Mayadas TN, Yuan F, Wagner DD, Hynes RO, et al. Rolling in P-selectin-deficient mice is reduced but not eliminated in the dorsal skin. Blood 1995;86:3487–92.

[40] Myers DD, Hawley AE, Farris DM, Wrobleski SK, Thanaporn P, et al. P-selectin andleukocyte microparticles are associated with venous thrombogenesis. J Vasc Surg2003;38:1075–89.

[41] Momi S, Falcinelli E, Giannini S, Ruggeri L, Cecchetti L, Corazzi T, et al. Loss ofmatrix metalloproteinase 2 in platelets reduces arterial thrombosis in vivo. J ExpMed 2009;206:2365–79.

[42] Nigam A, Kopecky SL. Therapeutic potential of monoclonal antibodies inmyocardial reperfusion injury. Am J Cardiovasc Drugs 2002;2:367–76.