herbert laurence millet, jeffrey i. weitz, lieve moons, désiré collen

HerbertLaurence Millet, Jeffrey I. Weitz, Lieve Moons, Désiré Collen, Peter Carmeliet and Jean-Marc

Mieke Dewerchin, Jean-Pascal Hérault, Goedele Wallays, Maurice Petitou, Paul Schaeffer,Heparin-Binding Domain of Antithrombin

Life-Threatening Thrombosis in Mice With Targeted Arg48-to-Cys Mutation of the

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2003 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.0000103634.69868.4F2003;93:1120-1126; originally published online October 30, 2003;Circ Res.

http://circres.ahajournals.org/content/93/11/1120World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org/content/suppl/2003/11/24/93.11.1120.DC1.htmlData Supplement (unedited) at:

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

by guest on February 23, 2013http://circres.ahajournals.org/Downloaded from

http://circres.ahajournals.org/content/93/11/1120

http://circres.ahajournals.org/content/suppl/2003/11/24/93.11.1120.DC1.html

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/

http://circres.ahajournals.org/

Life-Threatening Thrombosis in Mice With TargetedArg48-to-Cys Mutation of the Heparin-Binding Domain

of AntithrombinMieke Dewerchin, Jean-Pascal Hérault, Goedele Wallays, Maurice Petitou, Paul Schaeffer,

Laurence Millet, Jeffrey I. Weitz, Lieve Moons, Désiré Collen, Peter Carmeliet, Jean-Marc Herbert

Abstract—Antithrombin (AT) inhibits thrombin and some other coagulation factors in a reaction that is dramaticallyaccelerated by binding of a pentasaccharide sequence present in heparin/heparan-sulfate to a heparin-binding site on AT.Based on the involvement of R47 in the heparin/AT interaction and the frequent occurrence of R47 mutations in ATdeficiency patients, targeted knock-in of the corresponding R48C substitution in AT in mice was performed to generatea murine model of spontaneous thrombosis. The mutation efficiently abolished the effect of heparin-like molecules oncoagulation inhibition in vitro and in vivo. Mice homozygous for the mutation (ATm/m mice) developed spontaneous,life-threatening thrombosis, occurring as early as the day of birth. Only 60% of the ATm/m offspring reached weaningage, with further loss at different ages. Thrombotic events in adult homozygotes were most prominent in the heart, liver,and in ocular, placental, and penile vessels. In the neonate, spontaneous death invariably was associated with majorthrombosis in the heart. This severe thrombotic phenotype underlines a critical function of the heparin-binding site ofantithrombin and its interaction with heparin/heparan-sulfate moieties in health, reproduction, and survival, andrepresents an in vivo model for comparative analysis of heparin-derived and other antithrombotic molecules. (Circ Res.2003;93:1120-1126.)

Key Words: coagulation � gene targeting � knock-in � heparin � antithrombin deficiency

Antithrombin (AT) is a major physiological coagulationinhibitor, primarily inhibiting thrombin and factor Xa.1

In the presence of heparin, a potent anticoagulant andantithrombotic agent, the moderate inactivation rate by AT isdramatically accelerated. Heparan-sulfate proteoglycans in-tercalated in the endothelial cell membrane are thought tohave a similar effect in vivo.2 The acceleration depends on aspecific pentasaccharide sequence within the heparin glycos-aminoglycan chain, binding of which induces a conforma-tional change in the AT molecule.3,4 The interaction with thepentasaccharide appears sufficient to enhance the inhibitionof factor Xa. However, full enhancement of thrombin orfactor IX inhibition requires heparin species with longerpolysaccharide chains.5–8 Biochemical and structural studieswith wild type and variants of AT have shown that theheparin/pentasaccharide-binding domain of AT involves twoclusters of basic amino acids (residues 41 to 49 and 107 to156).1,5,8–12

The importance of AT in maintaining normal hemostasis isemphasized by the increased incidence of thromboembolismin individuals with inherited deficiency.1,13 Among these areseveral mutations affecting the heparin-binding site (type II

HBS AT deficiency). Heterozygous type II HBS patientshave a low incidence of thrombosis, whereas all homozygouspatients display thrombotic disease that may include venousas well as arterial events.1 To date, 70 reports on 12 distinctHBS mutations have been published,1,13,14 35 of which affectR47 [substitution to cysteine (19 reports), histidine,15 orserine1; see also16].

Heparin-based pentasaccharides or mimetics, which act, atleast in part, by binding to the heparin-binding site in AT,have potent anticoagulant and antithrombotic properties.4,15,17

To develop a murine model of thrombosis suitable forevaluation of such anticoagulants, mutant mice with targetedArg48-to-Cys substitution (R48C; corresponding to the hu-man “Toyama” R47C mutation) were generated.

Materials and MethodsGeneration of the AT Mutant MiceA targeting vector, pPNT.ATneo, was constructed to replace thewild-type exon 2 with a mutated exon 2 encoding the R48Csubstitution and an additional silent mutation creating a diagnosticSacI site (for details, see the online data supplement and onlineFigure S1, available at http://www.circresaha.org). Electroporation

Original received February 25, 2003; resubmission received July 23, 2003; revised resubmission received October 16, 2003; accepted October 16, 2003.From the Center for Transgene Technology and Gene Therapy (M.D., G.W., L. Moons, D.C., P.C.), VIB, KULeuven Campus Gasthuisberg O&N,

Leuven, Belgium; Cardiovascular/Thrombosis Research Department (J.-P.H., M.P., P.S., L. Millet, J.-M.H.), Sanofi-Synthélabo, Toulouse Cedex, France;Hamilton Civic Hospitals Research Centre (J.I.W.), Hamilton, Ontario, Canada.

Correspondence to Mieke Dewerchin, PhD, Center for Transgene Technology and Gene Therapy, VIB, KULeuven Campus Gasthuisberg O&N,Herestraat 49, B-3000 Leuven, Belgium. E-mail [email protected]

© 2003 American Heart Association, Inc.

Circulation Research is available at http://www.circresaha.org DOI: 10.1161/01.RES.0000103634.69868.4F

1120 by guest on February 23, 2013http://circres.ahajournals.org/Downloaded from


of R1 embryonic stem cells, subsequent Cre-mediated excision of theloxP-flanked neo cassette, generation of mice, and genotyping andexpression controls were performed as detailed in the online datasupplement. All experiments were performed on littermates. Housingand procedures involving experimental animals were approved bythe Institutional Animal Care and Research Advisory Committee ofthe KULeuven, Belgium.

Developmental, Histological, andImmunohistochemical AnalysisAnimals and embryos were initially evaluated by visual inspectionand stereomicroscopy. Organs of dead or anesthetized saline- andfixative-perfused mice or embryos were processed for immunohis-tochemical analysis as described.18 Fibrin(ogen) staining was asdescribed.19 Histological staining for collagen (Sirius red) andreticulin (Gordon and Sweets silver staining) were performed usingstandard procedures.

Coagulation and Hematological ParametersWhole blood was collected from anesthetized animals by cardiacpuncture into 0.1 vol 3.2% trisodium citrate and centrifuged twice for10 minutes at 3000 rpm to obtain plasma. All blood collectionsproceeded fast and with instantaneous mixing with the citratesolution. Measurements were performed immediately after samplepreparation or after storage at �80°C without intermediate thawing.Antithrombin inhibitory activity was measured using the Coamaticamidolytic anti-Xa heparin cofactor assay (Chromogenix, Sweden).Progressive fXa inhibitory activity was measured analogously butusing heparin-free buffer and longer reaction times. Coagulantactivities were determined in one-stage clotting assays as de-scribed.19 A similar analysis was performed after preconditioning ofthe mice with lepirudin to temporarily neutralize a possible in vivoactivation of the intrinsic coagulation factors (see online datasupplement). Plasma fibrinogen was determined by a coagulationrate assay.20 Factor VIII plasma levels were additionally determinedusing the Coatest Factor VIII kit (Chromogenix, Sweden)21 and byELISA.22 Plasma AT antigen levels were determined by rocketimmunoelectrophoresis23 using a polyclonal rabbit anti-human ATantibody (Dako) that cross-reacts with murine AT. Blood cell countswere determined as described.18 Prothrombin time (PT) and acti-vated partial thromboplastin time (aPTT) was measured usingroutine assays. Plasma TAT levels were determined using theEnzygnost TAT kit (Dade Behring). Blood was collected, processed,and assayed for FPA as described.24

Thrombin Generation in Plasma and VenousThrombosis ModelPlatelet poor plasma was pooled from at least 3 mice, defibrinated,and 20 �L-aliquots were used for continuous monitoring of tissuefactor-induced thrombin generation25 (see online data supplement),with or without addition of the pentasaccharide fondaparinux (Arix-tra),17,26 the heparin mimetic SanOrg123781A,27 heparin, or hirudin(Sanofi-Synthélabo) as indicated. Thrombus formation in vivo wasmeasured in a thromboplastin-induced vena cava stasis model withor without administration of fondaparinux (see online datasupplement).

Statistical AnalysisData are mean�SD unless otherwise indicated. The statisticalsignificance of differences between groups was determined byunpaired t test, unless mentioned otherwise. A value of P�0.05 wasconsidered significant.

ResultsTargeted R48C Mutation of the AT GeneMice with an Arg48-to-Cys (R48C) substitution were gener-ated by targeted mutation using homologous recombinationin embryonic stem cells (see online Figure S1a). This muta-

tion corresponds to the most prevalent of the type II muta-tions affecting heparin binding in patients with hereditary ATdeficiency, which have an increased risk of thrombosis(Toyama type II HBS AT deficiency).14,28 RNA analysis(semiquantitative RT.PCR on liver mRNA; not shown), andplasma antigen determination showed normal expression ofthe mutant protein in heterozygous (AT�/m) and homozygous(ATm/m) mutant mice (plasma antigen levels by rocket immu-noelectrophoresis were 100�8% in AT�/m and 110�10% inATm/m mice, n�3 to 4, P�NS; expressed in percent ofwild-type control values). DNA sequencing of the entirecDNA prepared from ATm/m liver samples confirmed theintegrity and correct mutation of the expressed message (notshown).

Viability and SurvivalAmong 167 progeny of heterozygous parents, ATm/m off-spring were somewhat underrepresented at birth (P0: 16%instead of the expected 25%) and suffered frequent neonataldeath, often within 24 hours after birth, with only 60% ofthem reaching weaning age (Table 1). Spontaneous death ofAT�/m neonates was occasionally observed. Beyond weaningage, spontaneous death among wild-type and heterozygousanimals was rare, whereas survival of ATm/m mice remainedcompromised with death at various ages and only about 30%survival beyond 6 months. No underrepresentation of ATm/m

offspring was found at embryonal ages up to embryonal day(E) 18.5 (Table 1), suggesting that the early loss of ATm/m

pups occurred during or immediately after birth, likely withimmediate cannibalism by the mother (occasionally wit-nessed) and therefore without recovery of these dead pups.

Heparin Binding and Inhibitor Activity of theR48C AT Mutant ProteinATm/m plasma samples displayed low heparin cofactor activityin the Coamatic fXa inhibition assay (96�4.3% residual fXaactivity at 1.5 minutes versus 8.4�2% for wild type;mean�SD, n�6 to 8, P�0.001) but normal progressive fXainhibitory activity (37�7.8% residual fXa activity at 30

TABLE 1. Genotype Distribution Among Offspring ofHeterozygous (AT�/m�AT�/m) or Mixed Breeding Pairs(AT�/m Male�ATm/m Female) at Different Embryonal (E) orPostnatal (P) Ages

�/� �/m m/m Total

�/m�m/m Breeding pairs

E14.5 11 15 26

42% 58%

�/m��/m Breeding pairs

E18.5 13 26 13 52

25% 50% 25%

P0* 52 88 27 167

31% 52% 16%

P21* 66 120 20 206

32% 58% 10%

*Difference from expected Mendelian distribution statistically significant by�2 analysis.

Dewerchin et al Targeted AT Heparin-Binding Site Mutation in Mice 1121



minutes versus 42�10% for wild type; mean�SD, n�6 to 8,P�NS). Tissue factor–induced thrombin generation mea-sured in defibrinated plasma from ATm/m animals was com-parable to that in wild-type samples (71�10 versus 63�3mOD for wild type; mean�SD, n�3, P�NS) (Figure 1A).However, the R48C mutation totally prevented the action ofAT-mediated inhibitors added at concentrations up to 8-fold

higher than those affecting wild-type AT (heparin, pentasac-charide, or the heparin mimetic SanOrg123781A, whichcomprises both an AT and a thrombin-binding domain27),whereas the effect of direct inhibitors (hirudin) was notaffected (Figure 1A).

In vivo, using a stasis-plus-tissue factor-induced thrombo-sis model in the caval vein, the pentasaccharide inhibitedthrombus formation in wild-type mice (69% inhibition), butnot in ATm/m mice (only 5% inhibition; P�NS) (Figure 1B).These results confirm effective abolition of heparin interac-tion by the R48C mutation.

Spontaneous Thrombosis in ATm/m MiceHomozygosity for the R48C mutation was associated withspontaneous, often massive, thrombosis in the heart and, lessfrequently, lungs. Out of 16 adult ATm/m mice euthanized forhistology at different ages (2 to 17 months), 6 showedmassive thrombosis in the atria and/or ventricles, stainingpositive for fibrinogen/fibrin (Figures 2a and 2b), and oftenassociated with leukocyte infiltration (Figure 2b). Out of 14adult ATm/m mice, 3 showed fibrin deposition and/or vesselocclusion in the lungs (Figures 2c and 2d). In addition,although obstruction of hepatic blood flow was not directlyobserved, signs of portal hypertension were seen in themajority of adult ATm/m mice analyzed (12/15), characterizedby nodular regenerative hyperplasia (Figures 2e and 2f),dilatation of the sinusoids and formation of shunt vessels(Figures 2g and 2h). Abnormalities at the cellular levelincluded macrovesicular steatosis and the presence of neutro-phil clusters in the sinusoids and of phagocytosing macro-phages in the parenchyme indicating an inflammatory re-sponse (not shown). In 1-day-old ATm/m pups (2 of 4analyzed), infarcted zones in the liver with coagulativenecrosis were observed (Figures 3a through 3c), presumablyillustrating acute impaired blood flow in the liver andcontributing to the development of the liver pathology.Although liver/body weight ratios were normal in ATm/m

animals, they showed enlarged spleens (see online datasupplement), which is frequently seen in liver disease. Bothmales (8/31; 26%) and, more frequently, females (9/19; 47%)displayed severe degeneration of the eyes (Figures 2i and 2j),often with disruption of the retina and occasionally perfora-tion of the cornea (not shown), likely due to ocular veinocclusion (Figure 2k). No obvious thrombosis was observedin other organs nor in the larger vessels (caval vein, femoralartery and vein, brachial vein; not shown). However, inanimals used for breeding, severe thrombosis was observed inthe placenta of pregnant females, and in the penile veins ofsexually active males. Placental thrombosis in ATm/m females(Figures 2l and 2m) occurred irrespective of the genotype ofthe embryo (Table 1), and likely caused the decreased littersizes observed for ATm/m mothers (3.9�2.8 pups per litter,n�7, versus 8.6�2.9 for AT�/m mothers, n�44; mean�SD,P�0.005). Fifty percent of all sexually active males devel-oped irreversible priapism (9 out of 18 mated males) due toocclusion of the dorsal penile vein and impaired drainage andthrombosis of the corpora cavernosa (not shown).

All ATm/m neonates [6 at postnatal day (P) 1 or P2, one atP7] recovered dead and analyzed histologically, showed

Figure 1. Impaired heparin cofactor activity of the mutant anti-thrombin in vitro and in vivo. A, In vitro thrombin generation inplasma from wild-type (top) or ATm/m mice (bottom) showinglack of effect on mutant antithrombin of heparin and heparinderivatives [pentasaccharide fondaparinux (penta), heparinmimetic SanOrg123781A] at concentrations up to 8-fold higherthan those affecting wild-type antithrombin, whereas the effectof the direct thrombin inhibitor hirudin is not affected. Specificactivity of heparin and of the heparin mimetic was 180 anti-IIaunits/mg. Data represent mean�SD, n�3; *P�0.05 vs control;#P�0.05 vs same concentration of heparin or ofSanOrg123781A. B, In vivo thrombus formation, showing 69%inhibition by pentasaccharide (100 nmol/kg) in wild-type mice,but only 5% inhibition in ATm/m mice. C indicates control; penta,pentasaccharide. Mean value�SD, n�4 to 5; *P�0.05 vscontrol.

1122 Circulation Research November 28, 2003



massive thrombosis in the heart (Figures 3d and 3e). Threesuccumbed heterozygous AT�/m neonates (age P1) also re-vealed a small clot in the heart and an occluded lung vessel inone of them (not shown). Neonates euthanized immediatelyafter birth (P0) revealed, apart from the liver infarction zonesmentioned above, clots in the heart in 2 out of 4 ATm/m pups(not shown), whereas no abnormalities in the wild-type orAT�/m neonates were observed.

Histological analysis of ATm/m offspring at E18.5 revealedno abnormalities (7 embryos analyzed; not shown). However,although rare, thrombosis was observed at earlier embryonalage: in 1 out of 16 E14.5 ATm/m embryos, a thrombus wasfound in the atrium, and a second embryo showed bleedingand fibrin deposition in the myocardial tissue, whereas all of18 AT�/m and 4 AT�/� embryos appeared normal (not shown).

Hemostasis Parameters in Adult Wild-Type andMutant MiceHematological parameters (blood cell counts, hematocrit,platelets; not shown), plasma TAT, and fibrinogen levels(Table 2) in AT�/m and ATm/m blood samples were similar tothose in wild-type samples. When measured using one-stage

clotting assays, which are sensitive to preactivation of thecoagulation factors, intrinsic clotting factors were elevated(Table 2). However, at least for fVIII, this increase was notobserved in a preactivation-insensitive two-stage activityassay (Table 2).21 Similarly, fVIII antigen levels by ELISAwere comparable in AT�/� and ATm/m mice (500�100 ng/mLfor ATm/m versus 410�230 for AT�/� mice; mean�SD, n�4to 8, P�NS). These results suggested a low level continuousactivation of the coagulation system in ATm/m mice, consistentwith the observed thrombotic phenotype. This possibility wasfurther verified by one-stage clotting assay measurements onsamples from mice preconditioned with lepirudin to preventin vivo activation. Levels of intrinsic factors in lepirudin-treated ATm/m mice were reduced to values close, although notentirely comparable, to those of wild-type mice (for details,see online data supplement). A similar increase in intrinsicfactor levels is seen in mice with targeted truncation of tissuefactor. Presumably, this mutation results in the generation ofsoluble tissue factor. These mice exhibit severe thrombosis,but have a normal antithrombin molecule (Melis E, Moons L,Arnout J, Collen D, Carmeliet P, Dewerchin M, unpublisheddata, 2003). Taken together, these data suggest that low level

Figure 2. Spontaneous thrombosis in adult ATm/m animals. a and b, Fibrin(ogen)-stained section through the heart of a 4-month-oldATm/m male showing a large thrombus with leukocyte infiltration (arrowheads in b) in the right ventricle. c and d, Fibrin(ogen)-stainedsection of a normal lung of a wild-type mouse (c) and of a lung of a 5-month-old ATm/m male with thrombosis in a large vessel (d). ethrough h, Architectural changes in the liver of ATm/m mice typical of portal hypertension and indicative of impaired hepatic blood flow.Reticulin staining of wild-type (e) and ATm/m liver (f) showing nodular regenerative hyperplasia in the ATm/m liver. Hyperplasia of the he-patocytes with compression of atrophic hepatic plates (arrowheads in f) was observed. H&E staining of an ATm/m liver (h) showing sinu-soidal dilatation and formation of shunt vessels (arrow) in response to impaired blood flow compared with normal wild-type liver tissue(g). i and j, Transverse section through an intact wild-type (i) and a severely degenerated ATm/m eye (j). Posterior (pc) and anterior (ac)chambers are absent in the ATm/m eye, with blood accumulation in the front part of the eye (asterisk in j). Retinal, choroid, pigment epi-thelium, and sclera layers are no longer discernible. k, Severe thrombosis (asterisks) in ocular vessels in an ATm/m mouse. ac indicatesanterior chamber; c, cornea; i, iris; l, lens; pc, posterior chamber; pe, pigment epithelium; r, retina; and sc, sclera. l and m, Thrombosisin the placenta of a pregnant ATm/m female at day 14.5 of gestation with leukocyte infiltration (arrowheads in m) compared with ahealthy placenta from an AT�/m female (l). Magnification bars�50 �m in b through h; 100 �m in k through m; and 250 �m in a, i, and j.




activation of coagulation may be a feature of thrombosis-prone mice. Plasma FPA levels were normal (12�3.8 nmol/Lfor ATm/m mice versus 13�4.2 nmol/L for wild-type controlmice; mean�SD, n�4 to 5, P�NS). Prothrombin times (PT)

and activated partial thromboplastin times (aPTT) also werenormal in ATm/m mice (PT, 9.8�0.9 versus 9.8�1.9 secondsin AT�/� mice, mean�SD, n�9 to 11, P�NS; aPTT, 28�3.8versus 32�4 seconds in AT�/� mice, mean�SD, n�6,P�NS), although aPTT values tended to be slightly reduced.

DiscussionIn this study, knock-in of the R48C mutation in the heparin-binding site (HBS) of murine AT was shown to result inlife-threatening, spontaneous thrombosis during neonatal andadult life. The mutant AT was synthesized and present in theplasma at levels comparable to those of wild-type AT inwild-type mice. Similar to the corresponding human R47Cmutation,29 plasma from the ATm/m mice showed low heparincofactor activity but normal progressive fXa inhibitory activ-ity. The R48C mutation abolished the action of AT-mediatedinhibitors (heparin, heparin mimetic, or AT-binding pentasac-charide) in plasma, whereas the effect of direct thrombininhibitors was not affected. The less pronounced inhibition bythe pentasaccharide as compared with heparin (mimetic) inwild-type plasma (Figure 1A) shows that the effect of thelatter molecules on coagulation via AT not only results frominhibition of factor Xa but also of other coagulation enzymessuch as thrombin. However, the lack of inhibition of thrombingeneration by the pentasaccharide and heparin mimetic inATm/m plasma, indicates that the R48C mutation effectivelyabolishes the functional interaction that results in AT-mediated inhibition of factor Xa and/or thrombin by thesemolecules. The weak dose-dependent response observed withheparin may reflect an AT-independent effect, possibly re-lated to heparin cofactor II activation. These in vitro and invivo results, together with the spontaneous thrombotic phe-notype of the ATm/m mice, underline the importance ofheparin binding of AT for efficient inhibition of coagulationand show that normally functioning AT plays a key role withregard to other “natural” inhibitors such as heparin cofactor-IIin particular.

Thrombosis in the adult ATm/m mice involved the heart,lung, liver, and eyes, and the reproductive organs in sexuallyactive animals. Unlike in adult AT�/m animals, spontaneousdeath and thrombotic events did occur in AT�/m neonates,although less frequently and less severely than in ATm/m

neonates. This might be due to a critical imbalance betweenpro- and anticoagulation in a fraction of the AT�/m neonates,likely due to the presence of the mutation in one allele on topof presumed lower neonatal plasma AT levels.30 Spontaneousthrombosis was more frequent and more severe in homozy-gous neonates, accounting for the frequent early loss of ATm/m

offspring. As in humans, our R48C phenotype was less severethan the embryonic lethal phenotype of AT-null mice.31

However, we occasionally observed fibrin deposition in themyocardium in E14.5 ATm/m embryos, reminiscent of theAT-null findings,31 although not in older embryos (E18.5).Whether such affected E14.5 ATm/m embryos recover or donot survive remains unclear. Nevertheless, the normal geno-type distribution at these embryonal ages, the absence ofapparent abnormalities before birth (E18.5), and the obviousthrombotic problems from early neonatal life onwards, sug-

Figure 3. Thrombotic events during neonatal life. a through c,ATm/m pup euthanized at the day of birth (P0) showing infarctionzones in the liver with coagulative necrosis (b) and fibrin deposi-tion (c) compared with normal liver of a wild-type littermate (a).H&E staining in a and b, fibrin(ogen) staining in c. d and e, Lon-gitudinal section through the heart of an ATm/m neonate suc-cumbed on postnatal day 1, showing a blocking thrombusbetween the left atrium and ventricle. Magnification bars�100�m in a through c and e and 250 �m in d.

TABLE 2. Plasma Levels of Coagulant Factors ShowingIncreased Levels of the Intrinsic Factors VIII, IX, XI, and XIIWhen Measured in the Preactivation-Sensitive One-StageClotting Assay

AT�/� AT�/m ATm/m

TAT, ng/mL 6.9�3 ND 5.1�1

Fbg, g/L 1.1�0.32 1.4�0.3 1.1�0.2

fII, % 89�9 89�11 102�10

fV, % 90�13 89�13 102�18

fVII, % 85�6 84�3 110�19

fX, % 89�9 84�6 97�15

fVIII, % 85�35 108�17 250�53*

fVIII, ng/mL† 380�94 510�82 500�60

fIX, % 101�15 112�17 260�33*

fXI, % 95�8 115�26 320�47*

fXII, % 98�15 107�12 280�38*

Data represent mean�SD, n�3 to 10. Unless otherwise indicated, valuesare expressed in percent of those measured in a wild-type plasma pool. Fbgindicates fibrinogen; fII-XII, coagulation factor II-XII; TAT, thrombin-antithrombincomplex; and ND, not determined.

†Values determined in a preactivation-insensitive two-stage activity assay;*P�0.05 vs �/� and �/m.




gest that the trauma of birth represents a major trigger ofthrombosis in the mutant mice.

Unlike patients with homozygous type II HBS AT defi-ciency,1,29 ATm/m mice displayed no obvious large vesselthrombosis, but showed massive thrombosis in the heart,severe thrombotic eye disease, liver pathology consistent withportal hypertension, placental thrombosis, and priapism.However, ocular vein occlusion, portal vein thrombosis,placental thrombosis, and although rarely, priapism are ob-served in patients with Factor V Leiden, prothrombin muta-tion, or decreased protein S, protein C, or AT levels.32–36

Thrombosis in the heart has been described in neonates withproved or suspected AT deficiency, in addition to thrombusformation in the large vessels and intracranial venous sinus(see review37).

The ATm/m thrombotic phenotype showed signs of inflam-matory response (leukocyte infiltration in thrombi, neutrophilclusters, and phagocytosing macrophages in the liver). Thisresponse might contain a direct AT-dependent component,not only through impaired thrombin inhibition and increasedPAR-mediated cytokine production, but also by impairedbinding to cell surface heparan-sulfate proteoglycans(HSPG). Indeed, AT, via interaction with HSPG present onendothelial cells or on neutrophils, promotes the release ofantiinflammatory prostacyclin and blocks the activation ofthe proinflammatory NF-�B, thereby decreasing platelet andneutrophil activation, chemotaxis, and interaction with theendothelium,38–41 effects that are lost after chemically block-ing the heparin-binding domain of AT.39,40

No thrombotic phenotype was observed so far in mice withaltered heparan-sulfate (HS) moieties or deficiencies in theHSPG core proteins. 3-O-sulfation of the pentasaccharidecore sequence of heparin/heparan-sulfate is essential forinteraction with AT.42 This modification is thought to becatalyzed mainly by heparan-sulfate-3-O-sulfotransferase-1(3-OST-1), which consequently plays a major role in thesynthesis of AT-binding anticoagulant HS.43 However, micedeficient in 3-OST-1 did not display a procoagulant pheno-type, perhaps due to redundancy by other 3-OST isoforms.44

Of note is the phenotype of mice deficient in glycosaminylN-deacetylase/N-sulfotransferase-2, which lack endogenoussulfated heparin but show no obvious signs of thrombosis,indicating that endogenous heparin is not critically involvedin coagulation.45 On the other hand, mice deficient in theHSPG core protein syndecan-4 are healthy and fertile, butshow impaired coagulation in fetal vessels in the placentallabyrinth.46

Heterozygous AT knockout mice recently were reported todevelop thrombosis only after challenge, largely analogous toheterozygous AT type I deficiency patients in which one ATallele is not expressed giving low functional and immunolog-ical AT.47 Mice deficient in heparin cofactor II displayed ashorter time to thrombotic occlusion of the carotid artery afterendothelial denudation, but otherwise did not show sponta-neous thrombosis nor other morphological abnormalities.48 Incontrast, transgenic mice with inactivation or mutation ofplasminogen system components49 or of coagulation inhibi-tors31,50,51 display spontaneous thrombotic phenotypes that,however, are either mild, or more severe with early, some-

times embryonal, lethality, or characterized by additionalnonthrombotic abnormalities.

In conclusion, knock-in of an R48C substitution in theheparin-binding site of antithrombin in mice effectivelyabolished the effect of heparin or heparin derivatives oncoagulation inhibition in vitro and in vivo. Homozygousmutant mice displayed life-threatening thrombosis at differ-ent sites, most prominently in the heart, liver, and in ocular,placental, and penile vessels, and represent an in vivo modelfor spontaneous thrombosis suitable for the analysis ofheparin-like and other antithrombotic molecules.

AcknowledgmentsThe authors gratefully acknowledge the excellent assistance of C.Gaich, S. Grailly, E. Gils, A. Hubert, L. Kieckens, R.Lavend’homme, T. Vancoetsem, A. Van Nuffelen, M. Vanrusselt,the assistance with artwork by A. Vandenhoeck and S. Jansen, andthe help with statistics by N. Boussac-Marlière. We are grateful to P.Lollar (Emory University, Atlanta, Ga) for kindly supplying therabbit anti-murine fVIII antibody, to I. Stalmans for help with the eyephenotype, to T. Roskams (University Hospital, Leuven) for helpwith the liver pathology, and to J. Arnout, E. Conway, and M.Jacquemin for helpful advice and critical discussion.

References1. Lane DA, Kunz G, Olds RJ, Thein SL. Molecular genetics of anti-

thrombin deficiency. Blood Rev. 1996;10:59–74.2. Rosenberg RD. Role of heparin and heparinlike molecules in thrombosis

and atherosclerosis. Fed Proc. 1985;44:404–409.3. Casu B, Oreste P, Torri G, Zoppetti G, Choay J, Lormeau JC, Petitou M,

Sinay P. The structure of heparin oligosaccharide fragments with highanti- (factor Xa) activity containing the minimal antithrombin III-bindingsequence: chemical and 13C nuclear-magnetic-resonance studies. BiochemJ. 1981;197:599–609.

4. Choay J, Petitou M, Lormeau JC, Sinay P, Casu B, Gatti G. Structure-activity relationship in heparin: a synthetic pentasaccharide with highaffinity for antithrombin III and eliciting high anti-factor Xa activity.Biochem Biophys Res Commun. 1983;116:492–499.

5. Wu YI, Sheffield WP, Blajchman MA. Defining the heparin-bindingdomain of antithrombin. Blood Coagul Fibrinolysis. 1994;5:83–95.

6. Hoylaerts M, Owen WG, Collen D. Involvement of heparin chain lengthin the heparin-catalyzed inhibition of thrombin by antithrombin III. J BiolChem. 1984;259:5670–5677.

7. Herault JP, Gaich C, Bono F, Driguez PA, Duchaussoy P, Petitou M,Herbert JM. The structure of synthetic oligosaccharides in relation tofactor IXa inhibition. Thromb Haemost. 2002;88:432–435.

8. Schedin-Weiss S, Desai UR, Bock SC, Gettins PG, Olson ST, Bjork I.Importance of lysine 125 for heparin binding and activation of anti-thrombin. Biochemistry. 2002;41:4779–4788.

9. Arocas V, Bock SC, Olson ST, Bjork I. The role of Arg46 and Arg47 ofantithrombin in heparin binding. Biochemistry. 1999;38:10196–10204.

10. Whisstock JC, Pike RN, Jin L, Skinner R, Pei XY, Carrell RW, Lesk AM.Conformational changes in serpins, II: the mechanism of activation ofantithrombin by heparin. J Mol Biol. 2000;301:1287–1305.

11. Huntington JA, McCoy A, Belzar KJ, Pei XY, Gettins PG, Carrell RW.The conformational activation of antithrombin: a 2.85 A structure of afluorescein derivative reveals an electrostatic link between the hinge andheparin binding regions. J Biol Chem. 2000;275:15377–15383.

12. Olson ST, Bjork I, Bock SC. Identification of critical molecular inter-actions mediating heparin activation of antithrombin: implications for thedesign of improved heparin anticoagulants. Trends Cardiovasc Med.2002;12:198–205.

13. Franco RF, Reitsma PH. Genetic risk factors of venous thrombosis. HumGenet. 2001;109:369–384.

14. Lane DA, Bayston T, Olds RJ, Fitches AC, Cooper DN, Millar DS,Jochmans K, Perry DJ, Okajima K, Thein SL, Emmerich J. Antithrombinmutation database: 2nd (1997) update. For the Plasma Coagulation In-hibitors Subcommittee of the Scientific and Standardization Committeeof the International Society on Thrombosis and Haemostasis. ThrombHaemost. 1997;77:197–211.




15. Turpie AG, Gallus AS, Hoek JA. A synthetic pentasaccharide for theprevention of deep-vein thrombosis after total hip replacement. N EnglJ Med. 2001;344:619–625.

16. Antithrombin Mutation Database Homepage. Available at: http://www.med.ic.ac.uk/divisions/7/antithrombin/default.htm. AccessedOctober 31, 2003.

17. Hérault JP, Bernat A, Gaich C, Herbert JM. Effect of new syntheticheparin mimetics on whole blood thrombin generation in vivo and in vitroin rats. Thromb Haemost. 2002;87:238–244.

18. Melis E, Moons L, De Mol M, Herbert JM, Mackman N, Collen D,Carmeliet P, Dewerchin M. Targeted deletion of the cytosolic domain oftissue factor in mice does not affect development. Biochem Biophys ResCommun. 2001;286:580–586.

19. Dewerchin M, Liang Z, Moons L, Carmeliet P, Castellino FJ, Collen D,Rosen E. Blood coagulation Factor X deficiency causes partial embryoniclethality and fatal neonatal bleeding in mice. Thromb Haemost. 2000;83:185–190.

20. Jalbert L, Rosen ED, Moons L, Chan JCY, Carmeliet P, Castellino FJ,Collen D. Inactivation of the gene for anticoagulant protein C causeslethal perinatal consumptive coagulopathy in mice. J Clin Invest. 1998;102:1481–1488.

21. Rosén S, Andersson M, Blomback M, Hagglund U, Larrieu MJ, Wolf M,Boyer C, Rothschild C, Nilsson IM, Sjorin E, et al. Clinical application ofa chromogenic substrate method for determination of factor VIII activity.Thromb Haemost. 1985;54:818–823.

22. Doering CB, Parker ET, Healey JF, Craddock HN, Barrow RT, Lollar P.Expression and characterization of recombinant murine factor VIII.Thromb Haemost. 2002;88:450–458.

23. Laurell CB. Quantitative estimation of proteins by electrophoresis inagarose gel containing antibodies. Anal Biochem. 1966;15:45–52.

24. Conway EM, Pollefeyt S, Cornelissen J, DeBaere I, Steiner-Mosonyi M,Weitz JI, Weiler-Guettler H, Carmeliet P, Collen D. Structure-functionanalyses of thrombomodulin by gene-targeting in mice: the cytoplasmicdomain is not required for normal fetal development. Blood. 1999;93:3442–3450.

25. Hemker HC, Wielder S, Kessels H, Beguin S. Continuous registration ofthrombin generation in plasma, its use for the determination of thethrombin potential. Thromb Haemost. 1993;4:617–624.

26. Bauer KA, Hawkins DW, Peters PC, Petitou M, Herbert JM, Van BoeckelCA, Meuleman DG. Fondaparinux, a synthetic pentasaccharide: the firstin a new class of antithrombotic agents: the selective factor Xa inhibitors.Cardiovasc Drug Rev. 2002;20:37–52.

27. Herbert JM, Hérault JP, Bernat A, Savi P, Schaeffer P, Driguez PA,Duchaussoy P, Petitou M. SR123781A, a synthetic heparin mimetic.Thromb Haemost. 2001;85:852–860.

28. Sakuragawa N, Takahashi K, Kondo S, Koide T. Antithrombin IIIToyama: a hereditary abnormal antithrombin III of a patient withrecurrent thrombophlebitis. Thromb Res. 1983;31:305–317.

29. Bauters A, Zawadzki C, Bura A, Thery C, Watel A, Subtil D, Aiach M,Emmerich J, Jude B. Homozygous variant of antithrombin with lack ofaffinity for heparin: management of severe thrombotic complicationsassociated with intrauterine fetal demise. Blood Coagul Fibrinolysis.1996;7:705–710.

30. Andrew M, Paes B, Johnston M. Development of the hemostatic systemin the neonate and young infant. Am J Pediatr Hematol Oncol. 1990;12:95–104.

31. Ishiguro K, Kojima T, Kadomatsu K, Nakayama Y, Takagi A, Suzuki M,Takeda N, Ito M, Yamamoto K, Matsushita T, Kusugami K, MuramatsuT, Saito H. Complete antithrombin deficiency in mice results inembryonic lethality. J Clin Invest. 2000;106:873–878.

32. Fegan CD. Central retinal vein occlusion and thrombophilia. Eye. 2002;16:98–106.

33. Janssen HL, Meinardi JR, Vleggaar FP, van Uum SH, Haagsma EB, vanDer Meer FJ, van Hattum J, Chamuleau RA, Adang RP, Vandenbroucke

JP, van Hoek B, Rosendaal FR. Factor V Leiden mutation, prothrombingene mutation, and deficiencies in coagulation inhibitors associated withBudd-Chiari syndrome and portal vein thrombosis: results of a case-control study. Blood. 2000;96:2364–2368.

34. Fisher NC, Wilde JT, Roper J, Elias E. Deficiency of natural antico-agulant proteins C, S, and antithrombin in portal vein thrombosis: asecondary phenomenon? Gut. 2000;46:534–539.

35. Ogueh O, Chen MF, Spurll G, Benjamin A. Outcome of pregnancy inwomen with hereditary thrombophilia. Int J Gynaecol Obstet. 2001;74:247–253.

36. Melissari E, Kakkar VV. Congenital severe protein C deficiency in adults.Br J Haematol. 1989;72:222–228.

37. Seguin J, Weatherstone K, Nankervis C. Inherited antithrombin III defi-ciency in the neonate. Arch Pediatr Adolesc Med. 1994;148:389–393.

38. Opal SM, Kessler CM, Römisch J, Knaub SK. Antithrombin, heparin, andheparan sulfate. Crit Care Med. 2002;30:S325–S331.

39. Oelschläger C, Römisch J, Staubitz A, Stauss H, Leithäuser B, TillmannsH, Hölschermann H. Antithrombin III inhibits nuclear factor �B acti-vation in human monocytes and vascular endothelial cells. Blood. 2002;99:4015–4020.

40. Hoffmann JN, Vollmar B, Römisch J, Inthorn D, Schildberg FW, MengerMD. Antithrombin effects on endotoxin-induced microcirculatory dis-orders are mediated mainly by its interaction with microvascular endo-thelium. Crit Care Med. 2002;30:218–225.

41. Dunzendorfer S, Kaneider N, Rabensteiner A, Meierhofer C, Reinisch C,Römisch J, Wiedermann CJ. Cell-surface heparan sulfate proteoglycan-mediated regulation of human neutrophil migration by the serpin anti-thrombin III. Blood. 2001;97:1079–1085.

42. Atha DH, Lormeau JC, Petitou M, Rosenberg RD, Choay J. Contributionof 3-O- and 6-O-sulphated glucosamine residues in the heparin-inducedconformational change in antithrombin III. Biochemistry. 1987;26:6454–6461.

43. Liu J, Shworak NW, Fritze LM, Edelberg JM, Rosenberg RD. Purifi-cation of heparan sulfate D-glucosaminyl 3-O-sulfotransferase. J BiolChem. 1996;271:27072–27082.

44. HajMohammadi S, Enjyoji K, Princivalle M, Christi P, Lech M, Beeler D,Rayburn H, Schwartz JJ, Barzegar S, de Agostini AI, Post MJ, RosenbergE, Shworak NW. Normal levels of anticoagulant heparan sulfate are notessential for normal hemostasis. J Clin Invest. 2003;111:989–999.

45. Forsberg E, Pejler G, Ringvall M, Lunderius C, Tomasini-Johansson B,Kusche-Gullberg M, Eriksson I, Ledin J, Hellman L, Kjellen L. Abnormalmast cells in mice deficient in a heparin-synthesizing enzyme. Nature.1999;400:773–776.

46. Ishiguro K, Kadomatsu K, Kojima T, Muramatsu H, Nakamura E, Ito M,Nagasaka T, Kobayashi H, Kusugami K, Saito H, Muramatsu T.Syndecan-4 deficiency impairs the fetal vessels in the placental labyrinth.Dev Dyn. 2000;219:539–544.

47. Yanada M, Kojima T, Ishiguro K, Nakayama Y, Yamamoto K,Matsushita T, Kadomatsu K, Nishimura M, Muramatsu T, Saito H.Impact of antithrombin deficiency in thrombogenesis: lipopolysaccharideand stress-induced thrombus formation in heterozygous antithrombin-deficient mice. Blood. 2002;99:2455–2458.

48. He L, Vicente CP, Westrick RJ, Eitzman DT, Tollefsen DM. Heparincofactor II inhibits arterial thrombosis after endothelial injury. J ClinInvest. 2002;109:213–219.

49. Carmeliet P, Collen D. Development and disease in proteinase-deficientmice: role of the plasminogen, matrix metalloproteinase and coagulationsystem. Thromb Res. 1998;91:255–285.

50. Eitzman DT, Westrick RJ, Bi X, Manning SL, Wilkinson JE, Broze GJ,Ginsburg D. Lethal perinatal thrombosis in mice resulting from theinteraction of tissue factor pathway inhibitor deficiency and factor VLeiden. Circulation. 2002;105:2139–2142.

51. Hogan KA, Weiler H, Lord ST. Mouse models in coagulation. ThrombHaemost. 2002;87:563–574.




by guest on February 23, 2013http://circres.ahajournals.org/

Dow

nloaded from


herbert laurence millet, jeffrey i. weitz, lieve moons, désiré collen

Documents