Upload
independent
View
0
Download
0
Embed Size (px)
Citation preview
Oral Communication Abstracts
Oral communications 1; Liver andNeuronal BiologyO1.1.01Interplay between Kupffer cells and non-Kupffer cellTNF signalling in liver PAI-1 productionScroyen I1, Bastelica D1, Poggi M1, Simonin Y2, Hibner U2, VanRooijen N3 and Alessi MC1
1Universite de la Mediterranee, Marseille; 2Institut de Genetique
Moleculaire de Montpellier, Montpellier, France; 3Vrije
Universiteit, Amsterdam, Netherlands
Objectives: Increased liver Plasminogen Activator Inhibitor-1 (PAI-1)
expression has been implicated in fatty liver disease development. In
the present study we used the macrophage depletion approach to
study the role of Kupffer cells (KC) and TNF in PAI-1 production.
Methods: A KC depletion approach with liposomal clodronate
(ClodL) in Wild Type (WT), Toll-like receptor-4 (TLR4)/)), myeloid
differentiation factor 88 (MyD88)/)), TNF)/) and TNF receptors
(TNFR)/)) deficient mice was used before LPS challenge or in a model
of steatohepatitis induced after methionine and choline deprivation.
Results: KC depletion diminished TNF, IL6 and TGF-a expression in
liver. However, it did not decrease PAI-1 expression neither in the basal
state nor after LPS challenge and even increased liver PAI-1 expression
in steatohepatitis. Overall these results indicate that KC are not respon-
sible for liver PAI-1 production. Hepatocytes may represent a relevant
PAI-1 source as freshly isolated hepatocytes from KC-depleted mice
challenged with LPS, expressed higher PAI-1 and TNF mRNA as com-
pared to non-LPS treated mice. Increased LPS-induced liver PAI-1
production was mediated through the TLR4 and MyD88 pathway, but
was independent of LPS-induced TNF as it was not blunted in TNF)/)
or TNFR)/) mice. Also liver specific PAI-1 expression was increased in
WT mice after methionine and choline deprivation but this phenome-
non was blunted upon TNF deficiency. Remarkably, in TNF)/) and
TNFR)/) mice, ClodL treatment attenuated both LPS and steatosis-
induced liver PAI-1 production.
Conclusions: TNF pathway deficiency revealed the role of KC deple-
tion on liver PAI-1 production during steatohepatitis and after LPS
challenge, indicating a complex interplay between a non-KC TNF
pathway that may compensate KC deprivation.
O1.1.02U-PA plays a critical role in the induction ofmacrophage polarity during liver repairKawao NK1, Nagai NN2 and Matsuo OM3
1Department of Physiology and Regenerative Medicine, Kinki
University Faculty of Medicine; 2Department of Animal
Bioscience, Faculty of Bioscience, Nagahama Institute of
Bio-Science and Technology, Nagahama, Japan; 3Kinki
University Faculty of Medicine, Osakasayama, Japan
Objective: Urokinase-type plasminogen activator (u-PA) plays an
important role in liver repair through the activation of plasminogen,
but its cellular mechanisms still remain unknown. Here, we investi-
gated accumulation of macrophages and their phenotypes at the bor-
der region of damaged area in u-PA-deficient and its wild-type mice.
Methods: After liver injury was induced by photochemical reaction,
the expression of phenotypic markers in activated macrophages and
their ultra-structural changes were sequentially analyzed by fluores-
cence and transmission electron microscopy.
Results: In wild-type mice, macrophages accumulated at the border
region of damaged area. The macrophages that accumulated at the edge
of damaged tissue expressed CD206, a phagocytic receptor, on day 7.
These accumulated macrophages clearly engulfed the cellular debris by
membrane ruffling. Furthermore, the distribution of the Golgi complex in
these macrophages was biased towards the direction of the damaged tis-
sue, indicating the extension of their pseudopodia in this direction. The
macrophages adjacent to CD206-positive macrophages expressed induc-
ible nitric oxide synthase (iNOS), a pro-inflammatory marker. In contrast,
such accumulation of macrophages was impaired in u-PA-deficient mice.
Phagocytosis of cellular debris by the macrophages with ruffled mem-
branes was rarely observed in u-PA-deficient mice. Neither CD206 nor
iNOS were expressed in the accumulated macrophages. Furthermore, the
Golgi complex was randomly distributed in these macrophages.
Conclusions: These data indicate that u-PA mediates primarily the
accumulation of macrophages and the induction of their polarity at
the edge of damaged tissue during liver repair.
O1.1.03Development of ELISAs to detect full-length andC-terminally intact alpha-2-antiplasminUitte de Willige S, Leebeek FWG and Rijken DCErasmus Medical Center, Rotterdam, the Netherlands
Background: Full-length alpha-2-antiplasmin (a2AP) consists of 464
amino acids, with an N-terminal Methionine (Met). Due to N- and
C-terminal proteolytic cleavages, a2AP circulates in functionally dif-
ferent molecular forms. N-terminally cleaved a2AP with an N-termi-
nal Asparagine (Asn) increases fibrinolytic resistance of a plasma clot
by increasing the crosslinking efficacy of a2AP to fibrin by activated
factor XIII. With its C-terminus, a2AP binds to plasminogen (plas-
minogen-binding-a2AP (PB-a2AP)). C-terminally cleaved a2AP
looses this ability, and is therefore a less potent plasmin inhibitor.
Objective: To develop ELISAs to detect full-length Met-PB-a2AP
and total PB-a2AP antigen levels, and investigate the range of N-ter-
minal variation in healthy controls.
Methods: To measure Met-PB-a2AP antigen levels, a custom made
rabbit antibody directed to the first 12 residues of the N-terminus
was used for coating. To measure total PB-a2AP antigen levels, a
commercially available goat a2AP antibody was used for coating. In
both assays, the primary antibody was a commercially available
monoclonal antibody directed to the C-terminus with Goat-anti-
Mouse-IgG-HRP as secondary antibody.
Results: The Met-PB-a2AP ELISA is specific for N-terminally intact
a2AP levels, as we found no response to recombinant Asn-a2AP. The
response of the total PB-a2AP ELISA to recombinant Met-a2AP and
recombinant Asn-a2AP was equal, indicating no discrimination
between the N-terminal variants. We applied both ELISAs to the indi-
vidual citrated plasma samples of 22 healthy controls. To investigate
the range of N-terminal variation, we calculated the ratio of Met-PB-
a2AP over total PB-a2AP. Compared to normal pool plasma (by defi-
nition 100%) N-terminal variation ranged between 57% and 156%.
Conclusions: We developed ELISAs to measure Met-PB-a2AP and
total PB-a2AP that will be suitable to determine the antigen levels of
these a2AP variants in plasma samples from patients with venous or
arterial thrombosis.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
ABSTRACTS____________________________________________________________________________________
O1.1.04Fibrinogen induces autoimmunity and inflammatorycell recruitmentKyu Ryu J1, Baeten K1, Petersen MA1, Murray SG1, Bedard C1,Davalos D1, Prod’homme T2, Lassmann H3, Degen JL4, ZamvilSS2 and Akassoglou K1
1Gladstone Institute of Neurological Disease; 2Department of
Neurology and Program in Immunology, University of California,
San Francisco, CA, USA; 3Centre for Brain Research, Medical
University of Vienna, Vienna, Austria; 4Cincinnati Children’s
Hospital Research Foundation and University of Cincinnati
College of Medicine, Cincinnati, OH, USA
Multiple Sclerosis (MS) is a neuroinflammatory disease of unknown
origin characterized by inflammatory demyelination in the brain lead-
ing to the loss of motor and sensory functions. MS lesions are pri-
marily perivascular and demyelination is preceded by blood-brain
barrier (BBB) disruption and local deposition of the plasma protein
fibrin(ogen), a proinflammatory coagulation factor. However,
whether BBB disruption directly contributes to the development of
autoimmunity and demyelination in MS remains unclear. Here we
report fibrinogen-induced encephalomyelitis (FIE) as a new mecha-
nism for the development of autoimmune inflammatory demyelina-
tion. Stereotactic fibrinogen injection in the corpus callosum of
healthy wild-type (WT) mice is sufficient to induce spontaneous
inflammatory demyelination, T cell and monocyte infiltration in the
CNS in vivo. Albumin or plasma derived from fibrinogen-deficient
mice does not induce inflammatory responses in the CNS, suggesting
that fibrinogen is specific among plasma proteins as a molecular
mediator of inflammatory demyelination upon BBB leakage. Demye-
lination upon fibrinogen injection is reduced in Rag2)/) mice, which
lack T and B cells. Furthermore, fibrin(ogen) promotes encephalito-
genic T-cell responses that result in exacerbated inflammatory demye-
lination, as shown in vivo upon injection of fibrinogen in 2D2
transgenic mice, which express T-cell antigen receptors specific for
myelin oligodendrocyte glycoprotein. Thus, our data suggest provi-
sional fibrin matrices within the CNS are a fundamental determinant
of adaptive immune-driven neuroinflammatory disease and introduce
FIE as a novel animal model for MS. This work was supported by
the National Multiple Sclerosis Society Postdoctoral Fellowships to
J.K.R. and D.D., American Heart Association Fellowship to J.K.R.,
the Howard Hughes Medical Student Research Fellowship to
S.G.M., the NIH/NHLBI grant HL096126 to J.L.D., and the NIH/
NINDS R01NS051470 and R01NS052189 grants to K.A.
O1.1.05Activated protein C prevents the thrombin-inducedactivation of astrocytesGorbacheva L1, Ivanova A1, Pinelis V2, Reiser R3, Ishiwata S4 andStrukova S1
1The Lomonosov Moscow State University; 2Scientific Centre for
Children‘s Health, RAMS, Moscow, Russian Federation;3Institute for Neurobiochemistry, Otto-von-Guericke University,
Medical Faculty, Magdeburg, Germany; 4Faculty of Science and
Engineering, Waseda University, Tokyo, Japan
Background: Recently neuroprotective effects of activated protein C
(APC) on stressed neurons, hypoxic brain endothelium has been
found and APC may be useful in therapy of stroke (Gorbacheva et
al. 2009, Zlokovic et al. 2005). However, the participation of APC in
regulation of astrocyte function is not clear. Thrombin (Th) promotes
cell proliferation, which correlated with astrogliosis (Nicole et al.
2005). S100B protein express in high abundance and release by astro-
cytes (Sen and Belli, 2007), and abnormally elevated levels of S100B
contribute to prominent reactive gliosis. The influences of APC on
the thrombin-induced re-activation of astrocytes are in a focus of our
research.
Methods: Confocal microscopy, Western blot methods and MTT-
assay of the cultured astrocytes at Th-induced toxicity and pre-treat-
ment with APC were used.
Results: Fluorescence immunostaining revealed that expression of
S100B was higher in cells that received the continuous application of
50 nM Th (20 h) than in untreated control astrocytes. Pre-treatment
and co-incubation of astrocytes with 1 nM APC led to decrease of
S100B level. Staining with fluorescence-conjugated phalloidin revealed
that continuous application of 50 nM Th caused actin cytoskeletal
rearrangements. In agreement with the previous studies, an MTT
assay indicated that cells treated with 50 nM Th (20 h) proliferate
25% as much as cells grown in 10% fetal bovine serum (the positive
control). Pretreatment with 1 nM APC prevented Th-induced astro-
cytic proliferation.
Conclusions: Thus APC has not only neuroprotective effects but also
can prevent the activation of astrocytes and astrogliosis during path-
ological condition. Our results demonstrate new aspects of APC as a
protective agent for brain at trauma and neuropathology.
Oral communications 2; Stroke andThrombolysisO1.2.01Gp1b-alpha blockade restores cerebral arterial patencyafter occlusive thrombosisLe Behot A1, Gauberti M2, Montagne A2, Lemarchand E2,Maubert E2, Vivien D2 and Orset C2
1Bd H Becquerel; 2Inserm UMR-S U919 SP2U, Caen, France
Background: Arterial thrombosis is the leading cause of mortality
and disability worldwide. Although Glycoprotein (Gp) IIb/IIIa inhib-
itors efficiently prevent thrombus development, they are ineffective to
recanalize occluded arteries, suggesting that occlusive thrombosis
involves specific and yet unidentified mechanisms.
Methods: We first characterized an experimental model of ischemic
stroke in mice, involving high shear occlusive thrombosis of the mid-
dle cerebral artery. Then, using non-peptidic inhibitors and monoclo-
nal antibodies, we investigated the respective roles of Gp IIb/IIIa and
GpIba-von Willebrand Factor (VWF) interactions in occlusive
thrombus development by laser Doppler, immunohistological and
MRI analyses. Thereafter, we investigated the recanalization effi-
ciency of GPIIb/IIIa and GPIba-VWF interaction inhibitions, in con-
junction or not with tissue-type plasminogen activator (tPA)
mediated thrombolysis.
Results: We demonstrated that occlusive thrombus formation is a
two-step process: first, platelets aggregate through involvement of
their GpIIb/IIIa receptors, resulting in partial occlusion of the blood
vessel and in a locally increased shear rate. Subsequently, when the
shear stress becomes elevated, platelet aggregation to the developing
thrombus becomes GpIba-VWF dependent, until closure of the vessel
lumen. Remarkably, inhibition of GpIba-VWF interaction by mono-
clonal antibodies or pharmacological inhibitors efficiently disaggre-
gated thrombi that were resistant to tPA and GpIIb/IIIa inhibitors.
Conclusion: These results suggest that disruption of GpIba-VWF inter-
action represents a promising approach to restore arterial patency in
patients with acute ischemic stroke or myocardial infarction.
O1.2.02MMP-10 as a new profibrinolytic agent inexperimental stroke through TAFI-mediatedmechanismOrbe J1, Rodriguez JA1, Orset C2, Barrenetxe J1, Angles-Cano E2,Vivien D2 and Paramo JA1
1CIMA, University of Navarra, Pamplona, Spain; 2INSERM
U9192 Serine Proteases and Pathophysiology of the
Neurovascular Unit (SP2U), Caen, France
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e11____________________________________________________________________________________
Background: The fibrinolytic and matrix metalloproteinase (MMP)
systems cooperate in thrombus dissolution and extracellular matrix
proteolysis. MMP-10 is induced by thrombin in endothelial cells and
has been implicated in inflammatory/thrombotic processes and vascu-
lar integrity. We hypothesised that MMP-10 could have a profibrino-
lytic effect and represent a promising thrombolytic agent is unknown.
Methods: The effect of MMP-10 on fibrinolysis was studied in vitro
and in vivo, in MMP-10-null mice (Mmp10)/)), by using two different
models of arterial thrombosis: laser-induced carotid injury and ische-
mic stroke. To analyze the implication of the thrombin activatable
fibrinolysis inhibitor (TAFI), an experimental stroke model was per-
formed in TAFI)/) mice receiving recombinant human MMP-10
(rhMMP-10) intravenously (6.5 lg/kg).Results: In vitro, we showed that MMP-10 was capable of enhancing
t-PA-induced fibrinolysis via TAFI inactivation-mediated mechanism.
In vivo, delayed fibrinolysis observed after photochemical carotid
injury in Mmp10)/) mice was reverted by active rhMMP-10. In a
thrombin-induced stroke model, the time to reperfusion and the
infarct size in sham or t-PA treated animals were severely impaired in
Mmp10)/) mice. In this model, administration of active rhMMP-10
to wild type animals significantly reduced reperfusion time and
infarct size (P < 0.05) compared to controls to the same extent as t-
PA but associated with shorter bleeding time (P < 0.01) and no in-
tracraneal haemorrhage. This effect was not observed in TAFI)/)mice, suggesting TAFI inactivation as one mechanism involved in
MMP-10 profibrinolytic effect.
Conclusions: A novel profibrinolytic role for MMP-10 in experimen-
tal ischemic stroke involving TAFI is described, opening new path-
ways for innovative fibrinolytic strategies in arterial thrombosis.
O1.2.03Molecular requirements for a safer generation ofthrombolytics by bioengineering the tissueplasminogen activator (TPA) A chainParcq J, Bertrand T, Baron AF, Hommet Y, Angles-Cano E andVivien DINSERM U919, Caen, France
Background: Since the NINDS trial in 1995, thrombolysis using
recombinant tissue plasminogen activator (rtPA) remains the only
treatment approved by the authorities for acute ischemic stroke.
Although rtPA is efficient in the blood circulation as a clot lysis
enzyme, rtPA may also produce damages at the neurovascular unit
including a risk of hemorrhagic transformations and neurotoxicity.
Based on our knowledge of the mechanism of action of tPA on the
glutamatergic signaling and subsequent neurotoxicity, we aim at pro-
posing safer thrombolytics.
Methods: We have first generated, produced and purified original
rtPA-related mutants including a non-cleavable single-chain rtPA
form (sc*-tPA) and a rtPA containing a degenerated kringle 2
lysine-binding site (K2*-tPA). Then, both fibrinolytic activity and
neurotoxicity of each of these muteins have been tested in vitro
on euglobulin clot lysis time model and on primary neuronal cul-
tures and in vivo on ischemic and NMDA-mediated neurotoxicity
models.
Results: In contrast to the wild-type tPA (wt-tPA), both sc*-tPA
and K2*-tPA fail to promote NMDA receptors-mediated neurotox-
icity. Moreover, both rtPA mutants show specific properties: K2*-
tPA has an increased fibrinolytic activity when compared to wt-tPA
and sc*-tPA has a regained zymogenicity when tested in absence of
fibrin(ogen).
Conclusions: We draw the molecular basis of a new generation of
thrombolytics for the management of stroke treatment. These data
call for in vivo studies in stroke (ongoing) but also in other neurode-
generative thrombosis.
O1.2.04On the efficacy of direct thrombolyticsNovokhatny V, Chamberlain D, Bromirski J, Nixon J andScuderi PGrifols Inc, Durham, NC, USA
The concept of direct thrombolysis has recently received attention as
a safe alternative to plasminogen activators. Several direct-acting
thrombolytics are currently being tested in the clinic. They include
plasmin (Pm), microplasmin (lPm), and alfimeprase, a recombinant
version of the snake enzyme fibrolase. These agents dissolve thrombi
directly in contrast to tPA which requires endogenous plasminogen
for its efficacy. Since each of these molecules is different biochemi-
cally, their in vitro fibrinolytic performance was compared to deter-
mine parameters which are important for the efficacy of this novel
class of thrombolytics. Pm was a clinical-grade material derived from
plasma (Grifols). Truncated Pm species (delta-plasmin, lacking mid-
dle portion of the molecule; lPm, comprising just the serine protease
domain) and fibrolase were produced in E. coli. Several in vitro
thrombosis models were used in the study. In a system free of
plasma, Pm degrades fibrinogen and fibrin to a greater extent than
lPm and fibrolase and does it with faster kinetics. In plasma, signifi-
cantly less Pm (approximately 1 lM) was required to initiate clot
lysis in comparison with lPm (approximately 3.6 lM). Fibrolase
activity in the later system was not registered at all at 2 h, but it pro-
gressively increased after 4 h. Under flow conditions, the efficacy of
each agent was directly proportional to the fibrin affinity. The degree
of clot lysis was found to be dependent on the catheter placement in
the clot. Almost twofold less clot lysis was achieved when the infu-
sion segment of the catheter was close to the proximal part of the
clot. Several factors critical for effective clot dissolution by direct
thrombolytics have been identified. Among them are their ability to
bind fibrin, the rate of inhibition by plasma inhibitors, the propensity
of these enzymes for autodegradation and the method of delivery to
the thrombus site.
O1.2.05Statins enhance vein recanalisation and reduce veinwall inflammation following venous thrombosisPremaratne SP, Grover S, Nuthall K, Saha P, Patel A, Modarai B,Waltham M and Smith AKing’s College London, BHF Centre of Research Excellence &
NIHR Biomedical Research Centre at Kings Health Partners,
London, UK
Background: Statins (3-hydroxy-3-methylglutaryl co-enzyme-A reduc-
tase inhibitors) exhibit anti-inflammatory, pro-angiogenic and pro-
fibrinolytic effects that may affect thrombus recanalisation and orga-
nisation.
Methods: Venous thrombi were induced in the vena cava of BalbC
mice by a combination of reduced flow and endothelial injury. One
day after thrombus induction, mice were randomised to three groups
(n = 7/gp). Atorvastatin (30 or 3 mg/kg) or vehicle (methyl-cellulose)
was given daily for 7 days by gavage. On day 7 thrombi were har-
vested and paraffin sections obtained at defined intervals. Vein recan-
alisation, thrombus volume and nucleated cell counts in thrombus
and vein wall were measured by image analysis of stained sections.
Immuno-histochemical staining was carried out to estimate percent-
age area of thrombus or vein wall containing macrophages (MAC-2)
and neutrophils (NIMP-R14).
Results: Vein recanalisation was greater following treatment with
high dose Atorvastatin (0.50 � 0.13 mm3) compared with low dose
or vehicle (0.29 � 0.11 mm3, 0.27 � 0.13 mm3, P = 0.002 ANOVA).
Neovascular channel number within the thrombus was significantly
higher in both treated groups (5.00 � .44 [high dose]; 5.14 � .5 [low
dose], vs. 3.14 � .40 [vehicle], P = 0.009). Thrombus volume, nucle-
ated cell count, MAC-2 and NIMP-R14 staining was similar for all
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e12 Abstracts____________________________________________________________________________________
three groups. Vein wall nucleated cell count was lower in treatment
groups (637 � 74 [high dose]); 649 � 75 [low dose] vs. control
(1023 � 60, P = 0.001). MAC-2 (0.41% � 0.04, 0.45% � 0.04) and
NIMP-R14 (3.92% � 0.38, 3.76% � 0.48) staining was significantly
lower in the vein wall of statin treated groups compared with vehicle
(0.97% � 0.05, P < 0.001; 7.33% � 0.36, P < 0.001).
Conclusions: Atorvastatin enhanced recanalisation and inhibited vein
wall inflammation associated with wall fibrosis. Statin use in combi-
nation with existing therapies may help to reduce post-thrombotic
syndrome.
Oral communications 3; Infection,Inflammation and ComplementO1.3.01Urokinase-type plasminogen activator andplasminogen activator inhibitor type-1 is increased byinterleukin-33 in human endothelial cellsDemyanets S1, Stojkovic S1, Kaun C1, Lemberger CE1, de MartinR1, Maurer G1, Huber K2 and Wojta J1
1Medical University of Vienna; 2Wilhelminen hospital, Vienna,
Austria
Background: Interleukin (IL)-33 is a novel member of the IL-1 cyto-
kine family and is a ligand of the ST2 receptor. IL-33 has been impli-
cated in the pathogenesis of atherosclerosis. It was shown to induce
vascular permeability and angiogenesis and the production of inflam-
matory cytokines in endothelial cells (ECs). Here we aimed to study
a possible regulation of u-PA and PAI-1 by IL-33 in human ECs.
Methods: HUVEC and human coronary artery EC (HCAEC) were
treated with IL-33. Specific mRNA levels for u-PA and PAI-1 were
determined by RT-PCR and u-PA and PAI-1 antigen and activity
were measured by ELISAs.
Results: u-PA mRNA was up-regulated up to fivefold in HUVEC
and up to 2.4-fold in HCAEC when these cells were treated with
100 ng/mL IL-33 for 9 h whereas PAI-1 mRNA increased up to 2.5-
fold and up to twofold, respectively. The increase in u-PA and PAI-1
antigen was concentration-dependent. PAI-1 activity was also
increased after the incubation with IL-33. sST2 Fc abrogated the IL-
33-induced increase in u-PA and PAI-1 antigen suggesting that these
effects of IL-33 are ST2-mediated. Overexpression of dominant nega-
tive form of IkappaB kinase2 in HUVEC abolished IL-33-induced
u-PA mRNA expression. IL-1 receptor antagonist had no effect on
IL-33-induced increase in u-PA and PAI-1, which suggests that these
effects are IL-1-independent. Simvastatin at concentrations ranging
from 0.5 to 2.5 lM abrogated IL-33-induced increase of u-PA and
PAI-1 antigen, thus providing further evidence that statins have
effects beyond reduction of cholesterol.
Conclusion: Via induction of u-PA and PAI-1 in endothelial cells, IL-
33 could contribute to the modulation of endothelial cell-mediated
extravascular proteolysis in processes such as neovascularization and
vascular remodeling. By modulating these processes IL-33 could
affect plaque angiogenesis thereby impacting on the stability of these
vascular lesions in atherosclerosis.
O1.3.02C1 esterase inhibitor misfolding induces powerfulactivation of the contact pathwayMadsen DE1, Gram J2 and Sidelmann JJ11Institute of Public Health, University of Southern Denmark;2Department of Clinical Biochemistry, Hospital of Southwest
Denmark, Esbjerg, Denmark
Objectives: C1 esterase inhibitor (C1-inh) is the primary inhibitor of
the contact pathway. Mutations in the C1-inh gene result in misfold-
ing and polymerization of the inhibitor as observed in patients suffer-
ing from hereditary angioedema (HAE). Here we study the inhibitory
and activating properties of polymerized C1-inh in the contact
pathway.
Methods: Polymers of C1-inh were generated by heating of native
C1-inh. Native PAGE was used to demonstrate C1-inh polymeriza-
tion and as electrophoretic gel mobility shift analysis for evaluation
of the mobility of proteins. The inhibitory potential of C1-inh poly-
mers was elucidated with a complement C1s activity assay. The abil-
ity of C1-inh polymers to activate the contact pathway was assessed
in a chromogenic assay system consisting of purified FXII, prekallik-
rein and a kallikrein sensitive substrate.
Results: Native PAGE demonstrated that stable polymers of C1-inh
were generated upon heating, whereas the C1s activity assay showed
that the inhibitory potential of C1-inh was lost upon polymerization.
Furthermore, incubation of FXII and polymerized C1-inh induced a
significant change in the electrophoretic mobility of FXII demonstrat-
ing an interaction between C1-inh polymers and FXII. Finally the
activating effect of C1-inh polymers was studied in the chromogenic
assay system. These experiments demonstrate that C1-inh polymers
induce a powerful generation of kallikrein when incubated with FXII
and prekallikrein and that the amount of kallikrein generated is
dependent on the concentration of C1-inh-polymers. C1-inh polymers
are not able to activate prekallikrein directly as no kallikrein activity
is recorded in absence of FXII.
Conclusion: Polymerization changes C1-inh from a major inhibitor to
a potent factor XII-dependent activator of the contact pathway. The
activation is modulated by binding of FXII to C1-inh polymers and
this mechanism may be significantly involved in the pathophysiology
of HAE-attacks.
O1.3.03Polyphosphate stabilises factor XIIa and acts as acofactor in factor XIIa-mediated plasminogenactivationBrain C and Mutch NJInstitute of Medical Sciences, University of Aberdeen, Aberdeen,
UK
Background and objectives: Factor XII (FXII) circulates in plasma as
an inactive zymogen. Cleavage at Arg353 generates a two-chain
active form, aFXIIa (80 kDa). aFXIIa can be further proteolysed by
cleavage at Arg334 to form bFXIIa (28 kDa), an active derivative
that is comprised only of the catalytic domain. Polyphosphate
(polyP) is an anionic chain of orthophosphate residues that is
secreted during platelet activation. PolyP is known to bind and acti-
vate FXII, thereby stimulating contact-mediated coagulation. This
study analyses the interaction of polyP with aFXIIa and its ability to
modulate plasminogen activation.
Methods: Binding of polyP to aFXIIa was analysed using native gel
electrophoresis. aFXIIa and plasmin activity was monitored using
specific chromogenic substrates. Proteolytic degradation of FXIIa
was monitored over time by SDS-PAGE. Plasminogen activation was
detected on Western blots with an HRP-linked antibody to plas-
min(ogen).
Results: Migration of aFXIIa on a native gel was retarded in the
presence of polyP, indicative of binding of the polymer. Incubation
of aFXIIa revealed the presence of several degradation products over
a 6 h time period. The appearance of these degradation products was
delayed in the presence of polyP; illustrating that binding protects
aFXIIa from autodegradation. Analysis of aFXIIa activity by a chro-
mogenic substrate revealed that polyP enhanced the intrinsic activity
of aFXIIa (P < 0.0001). aFXIIa-mediated plasminogen activation,
monitored by chromogenic assay and western blot, was significantly
enhanced in the presence of polyP.
Conclusions: PolyP protects aFXIIa from autodegradation and stabi-
lises its catalytic activity. Binding of polyP to aFXIIa stimulates
aFXIIa-mediated plasminogen activation. These data indicate that
platelet-derived polyP may regulate aFXIIa activity in vivo and func-
tion as a cofactor in local plasminogen activation.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e13____________________________________________________________________________________
O1.3.04Exploring the relationship between streptococcalvirulence and plasminogen activation efficiencyHuish-Williams S, Longstaff C and Thelwell CNational Institute for Biological Standards and Control (NIBSC),
South Mimms, UK
Objectives: Streptokinase (SK) contributes to streptococcal virulence
through plasminogen (Pg) binding, activation and fibrinolysis to
facilitate bacterial dissemination. S. pyogenes, group A streptococcus
(GAS) is a strict human pathogen and subgroups are known to have
different Pg activation characteristics. Unlike therapeutic SK from
non-pathogenic group C streptococci, little is known about the struc-
ture/function relationship of SK from GAS. The aim of this work is
to investigate the mechanism of Pg activation and fibrinolysis by
GAS SK.
Methods: Recombinant SK from group C (rSK-H46a) and GAS
(rSK-M1GAS) were cloned and expressed in E. coli. Glu- and Lys-
Pg activation rates were measured with Pg activation assays using a
chromogenic method, with or without fibrin, and a global assay was
used to determine fibrinolysis rates. Kinetic parameters kcat and Km
were determined for plasmin and the different SK-plasmin complexes
against the chromogenic substrate S-2251.
Results: Specific Pg-activation rates for rSK-H46a were 10-fold
higher than rSK-M1GAS in a chromogenic assay relative to a puri-
fied SK from streptococcus. No difference in activity was found
between rSK-H46a and rSK-M1GAS plasmin complexes, with both
increasing plasmin enzyme efficiency, kcat/Km, twofold. Pg-activation
rates were increased several-fold for rSK-M1GAS in fibrin however
no increase was observed for rSK-H46a. The same trend was
observed for fibrinolysis rates.
Conclusions: The fibrin-dependent stimulation of GAS SK activity
identified here may at least in part explain the differences between
group C and GAS pathogenicity. A detailed understanding of this
mechanism could be exploited to help identify new drug targets
against streptococcal diseases, and may contribute to the rational
design of new thrombolytic drugs.
O1.3.05Fibrinolytic activity of microvesicles: a compensatorymechanism in clot formation/dissolutionMartinez de Lizarrondo S1, Plawinski L1, Parcq J1, Orbe J2,Paramo JA2, Vivien D1 and Angles-Cano E1
1INSERM, U919, Serine Proteases and Pathophysiology of the
Neurovascular Unit, Caen, France; 2CIMA, University of
Navarra, Pamplona, Spain
Background: An early sign of the haemostatic response is plasma
membrane blebbing and shedding of membrane fragments, known as
microvesicles (MVs). Recent data indicate that besides their procoag-
ulant components, MVs may be a major source of plasminogen acti-
vators (PAs) and plasmin generation.
Objective: The purpose of this study was to investigate the role of
cellular MVs in clot triggering and dissolution and to determine their
balanced activity leading to efficient clot lysis.
Methods and results: MVs obtained from stimulated human micro-
vascular endothelial cells (HMEC-1, bearing uPA), from human
tPA-transfected HEK293 cells, as well as from platelets were iso-
lated by a sequential ultracentrifugation procedure and character-
ized. The plasminogen activator activity of these MVs was
identified by zymography and using a plasminogen-dependent chro-
mogenic assay. Their impact in clot formation and lysis was deter-
mined using a pool of human plasma (n = 16) and the
corresponding euglobulin fraction. Different concentrations of
fibrinolytic MVs (bearing either uPA or tPA) were added to the
calcium-induced clot system. Lysis efficiency was indicated by the
time to decrease by 50% the maximal turbidity of the clot. The
fibrinolytic capacity of MVs was calculated by reference to known
amounts of PAs. All cell type derived-MVs reduced the clotting
time in a dose-dependent manner. In contrast, only PA-bearing
MVs were able to reduce the time for clot lysis as a function of
their concentration. In a plasma system, this fibrinolytic effect of
MVs was twofold more effective than the corresponding amount
of soluble non-vesiculated PA. Furthermore, PAs of MV origin
were more resistant to soluble inhibitors.
Conclusions: In summary, PA-bearing MVs develop higher and more
efficient fibrinolytic capacity than soluble PAs. Thus, MVs seem to
participate in localised clot dissolution, as carriers of fibrinolytic
agents that are protected from neutralisation.
D. Collen Prize Symposium
DC.01Urokinase-type plasminogen activator receptor (uPAR)regulates angiogenesis via interaction with lowdensity lipoprotein receptor-(LDLR-) like proteinsPrager G, Unseld M, Poettler M, Novotny R, Kalinowska W,Kalinowska W, Geiger M and Zielinski CCMedical University of Vienna, Vienna, Austria
Tumor angiogenesis is induced when the net balance of pro- and an-
tiangiogenic molecules is tipped in favor of angiogenesis, the so called
‘angiogenic switch’. Recently, we described a mechanism how VEGF
induces pro-uPA activation, which led to uPAR-complex formation
and internalization of beta-1 integrins into the endosomal compart-
ment via LDLR-family members. Thereby, uPAR plays a central role
for VEGF-induced endothelial cell migration. By the use of affinity
chromatography or co-immunoprecipitation experiments, we have
identified a direct binding motif on uPAR domain 3 for LDLR-like
protein interaction. To proof a functional relevance of direct uPAR/
LDLR protein interaction, we reconstituted either uPAR mutants
(mutL3/uPAR), lacking the binding site for LDLR-proteins, or wild
type uPAR in uPAR deficient HEK 293 cells or in endothelial cells
derived from uPAR)/) mice. Notably, the mutants had no effect on
uPA binding to its receptor. Reconstitution of mutL3/uPAR was
incapable to rescue the impaired cell spreading behavior on integrin-
adhesive matrix as well as to rescue the diminished migratory
response upon VEGF stimulation. Furthermore, pY576 FAK phos-
phorylation upon cell adhesion was diminished, suggesting a central
role of uPAR-induced integrin redistribution in adhesion induced sig-
nal transduction. The functional importance of uPAR/LDLR interac-
tion was further reflected by an inhibitory peptide (P1), which
interfered with uPAR/LDLR-protein interaction, and functionally
reverted full length uPAR reconstitution. From our data we conclude
that the hitherto unknown uPAR/LDL-R protein interaction plays a
central role in effective pro-angiogenic endothelial cell behavior such
as endothelial cell spreading, migration and invasion. Thus, uPAR/
LDL-R interaction might represent a novel therapeutic target in
angiogenesis-related diseases.
DC.02Establishment of method and image analysis systemsto monitor clot formation and FXIII activity in realtime using a murine in vivo model of thrombosisAli M1, Ridger V2, Pease R1, Howells G1, Grant P1, Ariens R1 andPhilippou H1
1University of Leeds, Leeds; 2University of Sheffield, Sheffield,
UK
Background: Investigating in vivo thrombus formation in real time is
a practical tool for understanding thrombotic mechanisms and to
assess novel anti-thrombotic therapeutics. Whilst in vivo murine
thrombosis models exist, and FXIIIa activity has been assessed in
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e14 Abstracts____________________________________________________________________________________
vivo, we have extended these methods to enable real time assessment
of clot size/growth monitoring FXIII activity simultaneously.
Methods: Using intravital microscopy as an imaging tool, we have
established an in vivo murine model to study real-time thrombus for-
mation in the femoral vein of FXIII)/) and CBA/129 (control) mice
following 10% FeCl3 injury. Employing an infusion of 5% FITC-
labelled fibrinogen and an Alexa680 labelled peptide substrate for
FXIII based on the N-terminal sequence of alpha-2-antiplasmin
(A15) we have been able to perform simultaneous 488 and 680 nm
imaging. This has allowed us to monitor FXIII activity alongside the
initial stages of thrombus formation and its propagation in real time.
In addition, automated Image J macros have been developed and
commercial image analysis software IMARIS and Slidebook have
been used to allow accurate measurements of fibrin incorporation
and FXIII activity (by monitoring co-localisation of FITC-fibrin and
A15 peptide) in this system.
Results and conclusion: Our data demonstrates that FXIII incorpora-
tion of A15 takes place within 3 min of inducing clot formation with
FeCl3, with only a moderate increase in further incorporation over a
period of 60 min. In contrast FXIII)/) shows only a small back-
ground co-localisation of FITC and Alexa signal, as expected. Using
purpose built image analysis macros as well as commercial image
analysis software (IMARIS and Slidebook) we demonstrate that
appropriate background intensity adjustments are required to pro-
duce accurate and representative analysis of clot formation and real
time growth. Our data highlights for the first time that FXIII activity
plays an important role in thrombus size in the first 10 min of clot
formation. Our model enables investigation of mechanisms involved
in clot formation in real time in vivo.
DC.03Antibody therapy targeting tPA’s neurotoxicity: a newhope for the treatment of strokeMacrez R, Vivien D and Ali CInserm U919, Caen, France
Background and purpose: Tissue-type plasminogen activator (tPA) is
the only drug approved for the acute treatment of ischemic stroke,
but with two faces in the disease: beneficial fibrinolysis in the vascula-
ture and damaging effects on the neurovascular unit and brain paren-
chyma. To improve this profile, we developed a novel strategy,
relying on antibodies targeting the pro-neurotoxic effects of tPA.
Methods: After production and characterization of antibodies
(aATD-NR1) that specifically prevent the interaction of tPA with the
ATD-NR1 of NMDA receptor, we have evaluated their efficacy in a
model of murine thrombo-embolic stroke with or without rtPA-
induced reperfusion, coupled to MRI, near-infrared fluorescence
imaging, and behaviour assessments.
Results: In vitro, aATD-NR1 prevented the pro-excitotoxic effect of
tPA, without altering basal NMDA neurotransmission. In vivo, after
a single administration alone or with late rtPA-induced thrombolysis,
antibodies dramatically reduced brain injuries and blood-brain bar-
rier leakage, thus improving long-term neurological outcome.
Conclusions: Our strategy limits ischemic damages, and extends the
therapeutic window of tPA-driven thrombolysis. Thus, the prospect of
this immunotherapy is an extension of the range of treatable patients.
DC.04Increased zymogen activity of thrombin activatablefibrinolysis inhibitor prolongs clot lysisMishra N, Buelens K, Theyskens S, Compernolle G, Gils A andDeclerck PJKatholieke Universiteit Leuven, Leuven, Belgium
Introduction: Thrombin Activatable Fibrinolysis Inhibitor (TAFI) is
a zymogen, which on proteolytic activation into TAFIa removes car-
boxy-terminal lysine-residues from partially degraded fibrin and thus
attenuates fibrinolysis. Recent studies demonstrated that the zymogen
also exerts an intrinsic enzymatic activity.
Objective: To characterize the functional properties of zymogen-stim-
ulatory nanobodies identified in a panel of nanobodies raised against
a stable TAFI variant (TAFI-AI-CIYQ).
Methods and results: Nanobodies Vhh-TAFI-a51 and Vhh-TAFI-i103
were able to stimulate the zymogen activity 10- to 21-fold compared
to the baseline zymogen activity of TAFI and had no effect on TAFI
activation or TAFIa activity. The increase in catalytic efficiency can
be attributed mainly to an increased catalytic rate. The half-life of
the stimulated zymogen activity was longer than the half-life of TA-
FIa activity (30–40 min vs. 15 min at 37�C). The susceptibility
towards PTCI and GEMSA and the kinetics of the stimulated zymo-
gen activity differ significantly from those of TAFIa activity. Epitope
mapping revealed that both Asp75 and Thr301 are major determi-
nants in the binding of these nanobodies to TAFI. Localization of
the epitope strongly suggests that stimulation of zymogen activity is
due to the translocation of the activation peptide, leading to an
increased accessibility of the catalytic site. In TAFI-depleted plasma
reconstituted with a non-activatable variant of TAFI (TAFI-R92A)
clot lysis was prolonged either by increasing the concentration of
TAFI-R92A (thereby increasing the concentration of intrinsic zymo-
gen activity) or by stimulating its zymogen activity through addition
of the stimulatory nanobodies.
Conclusion: We identified two nanobodies that stimulate the zymogen
activity of TAFI. The stimulated zymogen activity differs significantly
from TAFIa activity and increasing the zymogen activity of TAFI
results in an antifibrinolytic effect.
DC.05Platelet factor XIII stabilises thrombi againstpremature fibrin degradationMutch NJ, Fraser SR and Booth NAUniversity of Aberdeen, Aberdeen, UK
Background and Objectives: We recently showed that factor XIII
(FXIII) exerts its antifibrinolytic function by cross-linking functional
a2antiplasmin to fibrin. Approximately 50% of plasma FXIII is
required for maximal stabilization of thrombi. Platelets play a pivotal
role in each stage of haemostasis, including fibrin degradation by the
fibrinolytic system. This study addresses the importance of platelet
FXIII in fibrinolysis.
Method: Platelet-poor (PPP) and platelet-rich plasma (PRP) was
obtained by centrifugation of whole blood. Washed isolates of plate-
lets were also prepared. Model thrombi were generated from PPP,
PRP and FXIII deficient plasma � isolated platelets. In some cases a
TG inhibitor (1 mM) was included. Incorporation of FITC-labelled
fibrinogen allowed lysis to be quantified as fluorescence release.
FXIII activity was analysed by an in-house activity assay.
Results: FXIII activity was significantly higher in PRP (1.5 � 0.05 IU/
mL) than PPP (0.9 � 0.12 IU/mL; P < 0.005, n = 4). FXIII activity
was below the limit of detection in FXIII depleted plasma, but could be
restored to normal plasma levels by the addition of 2 · 109/mL plate-
lets (1.4 � 0.003 IU/mL). Lysis of thrombi formed with FXIII-
depleted plasma was increased 11.8-fold relative to normal plasma
(184 � 6.87 FU/min compared to 15.6 � 6.29 FU/min). Addition of
0.4 U/mL Fibrogammin P or 0.5 · 108/mL platelets to FXIII-depleted
plasma decreased lysis several fold to 24 � 1.33 FU/min and
15.1 � 1.86 FU/mL, respectively. To assess the contribution of platelet
FXIII to thrombus stabilisation TG inhibitor was added to thrombi
formed from FXIII-depleted plasma + 0.5 · 108/mL platelets. This
resulted in a 10.1-fold increase in lysis (P < 0.005; n = 4). This indi-
cates that platelet FXIII accounts for approximately 80% of the stabil-
ity conferred to thrombi by platelets.
Conclusions: We have shown that platelets supply a pool of FXIII that
confers antifibrinolytic function and can compensate for plasma FXIII.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e15____________________________________________________________________________________
Oral communications 4; Cell Biology,Lysis and ReceptorsO2.4.01PAR1-dependent regulation of mast cell functions inacute inflammation in ratsGorbacheva L1, Rusanova A1, Piskunov A1, Erukhimovich A1,Koroban N1, Bespalova Z2, Feldman B3, Rumsh L4, Strukova S1
and Vorozhtsov B3
1Lomonosov Moscow State University; 2Russian Cardiology
Research Complex; 3NIOPIK; 4Shemyakin & Ovchinnikov
Institute of Bioorganic Chemistry, Moscow, Russian Federation
Objectives: Mast cells (MC) are involved in the process of atherogen-
esis, the development of coronary and acute cerebrovascular diseases,
which closely related to the processes of inflammation. MC release a
wide range of proinflammatory, vasoactive, proteolytic, fibrinolytic
mediators in response to activation by liberators. PAR (protease acti-
vated receptor) activation at inflammation is associated with regula-
tion of mast cell functions. Thrombin-agonist PAR-1 accelerates, but
activated protein C (APC) inhibits MC degranulation. The role of
MC in the anti-inflammatory activity of APC also as the effect of
PAR1 antagonist (aPAR1) on MC activated in acute inflammation
remains unknown.
Methods and Results: The peptide derivative of aPAR-1 was synthe-
sized and its and APC effects on MC activated in models of acute
inflammation (peritonitis) in rats were studied. We have shown that
APC at intraperitoneal injection into rats inhibits the production of
IL-1b and IL-6 (ELISA kit and RT-PCR) by MC; APC protects MC
from death (MTT-test), induced by inflammation; APC, also as
aPAR1, inhibits the translocation of NF-kappaBp65 into MC
nucleus (ELISA kit) after induction of inflammation in rats. PAR-1
blockade before APC action did not produce significant differences
in MC survival and NF-kBp65 translocation into nucleus. aPAR1
caused a decrease of histamine secretion by inflammatory MC.
Conclusions: Analysis of the data suggests that aPAR1 also as well
as APC is able to inhibit the development of inflammation in rats by
blocking MC activation in acute inflammation. Regulation of MC
functions with aPAR1 or APC may be a potential therapeutic avenue
to prevent inflammatory damage to the neurovasculature, especially
in conjunction with thrombolytics.
O2.4.02Plasminogen receptor expressed on the growth-promoted hepatocytes as a putative TAFI substrateduring liver regenerationSeki T, Arima M, Okumura N, Ishii K, Shigematsu H, Hosono T,Ogihara J and Ariga TNihon University, Fujisawa, Japan
Objective: Thrombin activatable fibrinolysis inhibitor (TAFI) is a
plasma procarboxypeptidase B secreted from the liver. We have dem-
onstrated that, during liver regeneration, the expression of TAFI is
suppressed under growth promoted conditions. This down-regulation
leads to high population of cell surface plasmin(ogen) and promotes
hepatocyte proliferation. In the present study, we have investigated
the plasminogen receptor (PlgR) expressed on the growth-promoted
hepatocytes as a putative TAFI substrate during liver regeneration.
Methods: Hepatocytes were isolated from rat liver by collagenase perfu-
sion method. The isolated hepatocytes were cultured on collagen-coated
culture dishes in the presence of EGF. The plasminogen-binding capac-
ity on hepatocytes was determined by the binding assay using HRP-
labeled plasminogen. Plasma membrane protein fraction of the hepato-
cytes was prepared by ultra centrifugation. The PlgR candidate was
detected by ligand blotting followed by LC-MS/MS analysis.
Results: The plasminogen binding capacity was upregulated in the
growth-promoted rat hepatocytes in primary culture. Tranexamic
acid completely cancelled the binding, but antibodies against known
PlgR such as a-enolase and annexin A2 did not show any effect on
the plasminogen binding. Several plasminogen binding proteins were
detected by ligand blotting, but those became undetectable in the
presence of tranexamic acid suggesting localization of plasminogen
bound on hepatocytes via lysine binding site. The 37 kDa PlgR was
strongly upregulated by EGF and it was identified to be malate dehy-
drogenase (MDH2) by peptide mass mapping. MDH2 expression was
also upregulated in the plasma membrane fraction in growth pro-
moted hepatocytes in comparison with quiescent control hepatocytes.
Conclusion: Malate dehydrogenase was identified as a novel plasmin-
ogen receptor on hepatocytes and it may play important role(s) in
the liver regeneration.
O2.4.03Accelerated fibrinolysis on vascular endothelial cellstriggered by membrane-retained secreted TPASuzuki Y and Urano THamamatsu University School of Medicine, Hamamatsu, Japan
Vascular endothelial cells (VECs) secret tissue plasminogen activator
(tPA) as an active form, and thus its facilitated secretion directly
enhances fibrinolytic activity. Recently, we succeeded to visualize the
secretory dynamics in GFP-tagged tPA expressing VECs using total
internal fluorescence microscopy. tPA-GFP appeared to stay on the cell
surface after secretion in a heavy chain dependent manner. PA inhibi-
tor-1 (PAI-1) facilitated dissociation of cell surface-retained tPA-GFP
by forming a high molecular weight complex. Here we analyzed how
retained active tPA modifies fibrinolytic activity on tPA-GFP express-
ing VEC surface. Alexa Fluor 568-labeled plasminogen (plg-568)
appeared to accumulate on cell surface at tPA-GFP retained spots as
well as intercellular/matrix adhesive area. Either addition of epsilon-
aminocaploic acid (EACA), a lysine analog, or pre-incubation with
carboxypeptidase B to remove carboxyl-terminal Lys residues attenu-
ated accumulation of plg-568. Alexa Fluor 568-labeled mini-plasmino-
gen, lacking four of five kringle domains of native plg, rarely
accumulated, suggesting that the binding of plg was LBSs-dependent.
Either employment of catalytically inactive mutant of tPA-GFP, or
supplementation of inhibitors such as PAI-1, aprotinin or a2-antiplas-min, also attenuated accumulation of plg-568, indicating that genera-
tion of plasmin on cell surface seems to be essential. Collectively
generation of plasmin by secreted and retained active tPA on cell sur-
face effectively facilitate further accumulation of plasminogen through
degradation of surface or pericellular proteins at Lys- or Arg- site and
thus exposure of additional carboxyl-terminal Lys residues. This pro-
posed positive feedback loop in plasmin generation on cell surface is
similar to that taking place on another solid-structure of fibrin.
O2.4.04Synergistic effect of thrombin and CD40l onendothelial MMP-10 expression and microparticlegeneration in vitro and in vivoMartinez de Lizarrondo S1, Roncal C1, Varo N1, Purroy A1,Lorente L2, Rodriguez JA1, Doeuvre L3, Angles-Cano E3,Paramo JA1 and Orbe J1
1CIMA, University of Navarra, Pamplona; 2Intensive Care Unit,
University Hospital of Canarias, La Laguna, Santa Cruz de
Tenerife, Spain; 3INSERM, U919, Serine Proteases and
Pathophysiology of the Neurovascular Unit, Caen, France
Objective: Thrombin induces CD40-ligand (CD40L) and matrix metallo-
proteinases (MMP) in inflammatory/prothrombotic conditions. Throm-
bin and CD40L could modulate endothelial MMP-10 expression in vitro,
in vivo and in clinical conditions associated with thrombin generation.
Methods: Human umbilical vein and aortic endothelial cells (HU-
VEC and HAEC) were stimulated with thrombin (0.1–5 U/mL),
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e16 Abstracts____________________________________________________________________________________
CD40L (0.25–1 lg/mL), or their combination (thrombin/CD40L) to
assess MMP-10 mRNA expression (real time-PCR), protein secretion
(ELISA), and microparticle generation (MPs). Experiments were also
performed in the presence of PAR-1 agonists, anti-PAR-1, and anti-
CD40/CD40L specific antibodies to assess the pathways implicated.
The induction of MMP-10 and MPs were determined in mice injected
with thrombin/CD40L and in endotoxemic mice model. Circulating
MMP-10 levels were also measured in septic patients (n = 60) with
systemic inflammation and enhanced thrombin generation.
Results: Endothelium exposed to thrombin/CD40L elicited higher
MMP-10 mRNA (fivefold; P < 0.001) and protein levels (4.5-fold;
P < 0.001) than either stimulus alone. This effect was mimicked by a
PAR-1 agonist peptide and antagonized by hirudin, a-PAR-1,a-
CD40L, and a-CD40 antibodies. Endothelial activation by these stim-
uli also increased MPs harbouring active MMP-10. Co-incubation of
thrombin with CD40L from platelet-derived MPs origin enhanced
MMP-10 expression. In mice, both stimuli independently induced
MMP-10, but their combination elicited a further increase of aortic
MMP-10 expression and generation of MPs carrying MMP-10.
Moreover, in septic patients, showing enhanced thrombin generation,
exhibited higher MMP-10 (P < 0.001) and sCD40L (P < 0.05) lev-
els associated with adverse clinical outcome.
Conclusions: Thrombin and CD40L elicited a strong synergistic effect
on endothelial MMP-10 expression in vitro and in vivo, which may
represent a new link between inflammation/thrombosis with prognos-
tic implications.
O2.4.05Effect of ageing on the murine venous circulationHemmeryckx B1, Emmerechts J1, Bovill EG2, Hoylaerts MF1 andLijnen HR1
1Center for Molecular and Vascular Biology, KU Leuven,
Leuven, Belgium; 2Department of Pathology and Laboratory
Medicine, University of Vermont, Burlington, VT, USA
Objectives: The effect of ageing on the morphology of veins, venous
valves and arteries was investigated in male wild-type C57BL/6J mice
in order to reveal potential age-associated vascular changes that may
explain the increased risk for deep vein thrombosis in the elderly.
Methods: Following injection of a silicone polymer Microfil� (that
preserves morphology of the vasculature) arteries, veins and venous
valves in the hindlimb of 10 weeks and 24 months old mice were
visualized and processed.
Results: Throughout the hind limb the arterial, but not the venous,
lumen area and wall thickness were significantly greater in 24 months
as compared to 10 weeks old C57BL/6 mice. Venous valves were
most frequently located at the sapheno-femoral vein junction in the
lower extremities, and appeared thicker at the base supported by
structurally intact collagen fibers, and thinner towards the proximal
end of the valve leaflet, with less organized collagen. Overall, valves
were less supported by structurally intact collagen at 24 months as
compared to 10 weeks (score: 0.43 � 0.14 vs. 1.35 � 0.20; P < 0.05).
Endothelial expression of CD31, endothelial protein C receptor or
von Willebrand Factor (VWF) was not affected by age, while throm-
bomodulin expression was lower in aged vs. young arteries. At both
ages, expression of VWF was lower at venous valves vs. veins. Evalu-
ation of the blood coagulation profile revealed that aged mice had
shortened prothrombin time, elevated plasma levels of factor (F)VII,
FVIII and VWF and increased neutrophil and platelet counts.
Conclusion: Our data indicate that in mice with ageing, venous valves
become more fragile, in association with a procoagulant and inflam-
matory blood phenotype. This procoagulant state is accompanied by
mild vascular changes.
Oral communications 5; Proteolysis andCancerO2.5.01Novel role of the urokinase receptor in regulatingPTEN levels and its role in angiogenesisUnseld M1, Schabbauer GS1 and Prager GP2
1Department of Physiology, Medical University of Vienna;2Department of Medicine I and Cancer Center, Vienna, Austria
Objective: he uPAR/uPA system mediates endothelial cell behaviour
either via cell proteolysis or intracellular signal transduction, thereby
affecting endothelial cell migration, survival as well as capillary-like
tube formation. Previously, it was shown that stimulation of endothe-
lial cells by VEGF activates uPAR bound uPA in a PI3K dependent
manner. Since the tumor suppressor protein PTEN is a major inhibi-
tor of the PI3K/Akt-pathway, we examined the probable correlation
between uPAR and PTEN.
Methods and results: Results showed that uPAR expression indirectly
correlates with PTEN protein levels. Analysis of the physiological sig-
nalling consequence for the PKB/Akt pathway revealed that uPAR
overexpression led to decreased PTEN and highly elevated pAKT
levels. Therefore we explored the effects and regulatory role of this
new finding in vitro and in vivo. We found that uPAR)/) endothelial
cells have lower migratory response toward growth factors such as
VEGF, FGF-2, HGF and EGF in vitro and in vivo. Crossbreads of
low PTEN expressing PTEN +/) and uPAR)/) were generated and
analyzed for rescuing the diminished migratory phenotype.
Conclusion: Since uPAR and PTEN are major regulators of angio-
genic endothelial cell behaviour, interfering with this axis might pro-
vide a novel therapeutic target for anti-angiogenic therapy.
O2.5.02PCI/PAI-3: a plasminogen activator inhibitor or aserpin with other (intracellular) functions?Geiger M, Einfinger K, Furtmueller M, Rieger D, Sokolikova Band Wahlmueller FMedical University of Vienna, Wien, Austria
Background: Protein C inhibitor (PCI, PAI-3) is a serpin with broad
protease reactivity. It binds glycosaminoglycans and certain phospho-
lipids, which modulate its protease inhibitory activity. PCI can be
internalized by cells in a phosphatidylethanolamine-dependent man-
ner. Intracellular functions as well as the biological role of phospho-
lipid-associated PCI are unknown.
Methods: Analysis of phospholipid vesicles (microparticles; MPs) by
flow cytometry, co-immunoprecipitation, mass spectrometry, and
functional assays (protease inhibition, complex formation). Identifica-
tion and characterization of intracellular proteins interacting with
endogenous, overexpressed, and internalized PCI in leukocytes and
prostate (cancer) cells: Yeast-2-hybrid screening, subcellular fraction-
ation and co-immunoprecipitation (anti-PCI-IgG), Western blot anal-
ysis of precipitated proteins, FRET-analysis of cells overexpressing
tagged PCI and differently tagged putative binding partners.
Results and conclusions: Plasma derived MPs from healthy donors
are mainly derived from platelets or megakaryocytes. They expose
phosphatidylserine (binding of Annexin V) and contain PCI, which is
inactive, but not in complex with a protease. By co-immunoprecipita-
tion of MP-lysates with anti-PCI-IgG and mass spectrometry we
identified complement proteins as interaction partners of PCI in/on
MPs. Functional assays revealed that PCI is a substrate for C1s. By
yeast-2-hybrid screening JFC-1 (= synaptotagmin-like protein
1;SLP1) was identified as intracellular protein interacting with PCI.
Both, PCI and JFC-1, bind phosphoinositides and colocalize in the
vicinity of the plasma membrane. By co-immunoprecipitation of sub-
cellular fractions we identified CSN6 and CSN5 (= JAB1), the active
enzyme of the COP9 signalosome, as proteins interacting with PCI in
the nucleus of peripheral blood lymphocytes.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e17____________________________________________________________________________________
O2.5.03Hepatocyte growth factor activator inhibitor-2 residesprimarily intracellularly to facilitate cell surfacetransport of matriptaseLarsen BR1, Steffensen SDR1, Nielsen NVL2, Friis S1, Godiksen S1,Bornholdt J1, Søndergaard C1, Andersen MN1, Poulsen SS1,Szabo R3, Bugge TH3, Lin CY4, Skovbjerg H1, Jensen JK5 andVogel LK1
1University of Copenhagen, Copenhagen; 2University of Aarhus,
Aarhus, Denmark; 3NIH, Bethesda, MD; 4University of Maryland,
Baltimore, MD, USA; 5Arhus Universtitet, Arhus, Denmark
Hepatocyte growth factor activator inhibitor-2 (HAI-2) is an inhibitor
of the membrane bound serine protease, matriptase in vitro. Studies of
knock-out mice have shown that HAI-2 is essential for placental devel-
opment and embryonic survival only in mice expressing matriptase.
Previous studies have shown that co-expression with another HAI,
HAI-1, is a requirement for detectable levels of matriptase to reach the
plasma membrane. In the present study we show that co-expression of
matriptase together with HAI-2, like HAI-1, influence the level of ma-
triptase and its subcellular localization. First, the matriptase level in
the cells is significantly higher if co-expressed with HAI-2. Secondly,
matriptase is transported to the plasma membrane, only when HAI-2 is
present, where it becomes activated as judged by cleavage at Arg614
and increase in peptidolytic activity of the cell extracts. Furthermore,
our results show that the vast majority of HAI-2 is located in the endo-
plasmic reticulum, although a minor fraction is located as a stable pool
not undergoing endocytosis on the apical plasma membrane. We detect
no co-localization of HAI-2 and active matriptase. Our studies suggest
that HAI-2 despite its ability to inhibit matriptase proteolysis in vitro,
play a non-conventional chaperone-like role towards matriptase in the
secretory pathway involving direct interaction between the Kunitz
domain 1 of HAI-2 and matriptase.
O2.5.04GSK-3b regulates PAI-1 expression, cell migration andangiogenesis via FBW7 and USP-28-dependentdegradation of HIF-1aKietzmann TDepartment of Biochemistry, Oulu, Finland
Increased plasminogen activator inhibitor-1 (PAI-1) levels indicate a risk
of ischemic/hypoxic cardiovascular events and a poor prognosis. Expres-
sion of PAI-1 can be induced by hypoxia via the hypoxia-inducible tran-
scription factor-1a (HIF-1a) which is a major regulator of angiogenesis
and carcinogenesis. Therefore, the identification of critical players regu-
lating HIF-1a is important for understanding PAI-1 expression and car-
cinogenesis. Here we report a novel mechanism by which PAI-1
expression can be affected after HIF-1a is degraded by glycogen synthase
kinase-3 (GSK-3) induced phosphorylation and recruitment of the
tumor suppressor Fbw7. Experiments with GSK-3b and Fbw7-deficient
cells revealed that GSK-3b and Fbw7-dependent HIF-1a degradation
can be antagonized by the ubiquitin specific protease-28 (USP28). In
line, Fbw7 and USP28 reciprocally regulated PAI-1 and cell migration
in a HIF-1a-dependent manner. Together, we identified a new pathway
which could be targeted at the level of GSK-3, Fbw7 or USP28 to influ-
ence PAI-1 linked processes like carcinogenesis.
O2.5.05Elucidating novel urokinase-type plasminogenactivator inhibitorsSmith E1, Spencer J2, Ali M1, Abdinejad M2, Kankanala J1,Fishwick C1 and Philippou H1
1University of Leeds, Leeds; 2University of Greenwich, London, UK
Background: Urokinase-type plasminogen activator (uPA) is a key
mediator, either directly or through its activation of plasminogen, in
a range of processes, including tumour growth and metastasis, angio-
genesis, and tissue remodeling. Therefore, selective inhibition of uPA
is of therapeutic interest for cancer and wound healing.
Methods and results: Thiouronium-substituted arylboronic acids (1, 2)
or pinacol esters (3) were synthesized as described in the prior art
(Deadman, Spencer et al. WO 02/057273, Trigen Ltd.). All three com-
pounds showed inhibitory effects against uPA in the low lM range.
These compounds were tested at 50 lM for selectivity against tPA,
plasmin, thrombin, factors VIIa, IXa, Xa, XIa, XIIa, and trypsin, with
compound 1 found to be the most selective inhibitor, only decreasing
factor XIa activity slightly (by 10%). Compounds 2 and 3 both showed
an inhibitory effect on all the enzymes tested, to varying degrees. In a
turbidity-lysis assay, compound 1 showed the most potent and selective
inhibition of clot lysis and did not interfere in clot formation, unlike
compounds 2 and 3. Molecular modelling of the inhibitors into the
active site within the uPA crystal structure has been performed in order
to rationalize the trends in affinity and selectivity for uPA found for
these inhibitors and to rationally design more selective and potent uPA
inhibitors. Finally, compound 1 shows the ability to reverse the effect
of uPA in vivo using a ferric chloride induced thrombosis murine mode.
When compound 1 was injected with uPA the rate of clot formation
and size was similar to that of control mice (using saline or compound
1 alone) compared with injection of uPA which led to reduced throm-
bus size due to fibrinolytic activity.
Conclusions: We have demonstrated that compound 1 has in vivo effi-
cacy and good selectivity for uPA, making this a good basis for the
design of future inhibitors.
Oral communications 6; Clot Stabilityand TAFIO3.6.01The effect of DNA and histones on fibrin structureand the regulation of plasminogen activation andfibrinolysisLongstaff C1, Varju I2, Szabo L3, Thelwell C1 and Kolev K2
1NIBSC, South Mimms, UK; 2Department of Medical
Biochemistry, Semmelweis University; 3Clinical Research Centre,
Hungarian Academy of Sciences, Budapest, Hungary
Objectives: Neutrophils release bactericidal extracellular traps (NETs)
of DNA, histones and enzymes at infection sites. NETs provide a
link between infection and coagulation and are associated with
increased mortality in sepsis. We have studied the impact of DNA,
histones and heparin on fibrinolysis using a variety of techniques.
Methods: Ranges of DNA, histones and heparin were added to fibrin
and effects measured in high throughput clot lysis and plasminogen
(Pgn) activation assays and supported by thromboelastography (TEM)
studies. Physical methods including isothermal titration calorimetry
(ITC) and fibrin rheology were applied to investigate fibrin binding and
mechanical properties under flow. Electron and confocal microscopy
studies, using engineered fluorescent tPA variants were used to study
the effect of NET components on fibrin structures, binding and lysis.
Results: DNA up to 1 mg/mL changed clot structures and slowed
fibrinolysis, especially time to 100% lysis with several Pgn activators
(tPA, uPA and streptokinase) despite DNA stimulation of Pgn acti-
vation in solution. DNA had no effect on plasmin activity. Histones
also prolonged clot lysis times and 0.01–0.1 mg/mL DNA + histones
slowed clot lysis up to threefold in microtitre plate clot lysis assays
and twofold in the TEM system. Confocal microscopy using fluores-
cent fibrin showed DNA inhibiting dissociation of fibrin degradation
products (FDP) and ITC confirmed that DNA binds large but not
small FDP with high affinity, Kd approximately 40 nM. Heparin
appears to replace DNA in complexes with histones.
Conclusions: DNA � histones can stimulate activation of Pgn (both
glu- and lys-) but changes clot structure making fibrin more resistant to
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e18 Abstracts____________________________________________________________________________________
lysis, noticeably in the final stages suggesting DNA holds FDP associ-
ated with the clot. Negatively charged heparin displaces DNA from hi-
stones and may be investigated as a therapeutic to mitigate the effects
of DNA-histone complexes on coagulation and fibrinolysis in sepsis.
O3.6.02Bispecific inhibition of TAFI and PAI-1 as afibrinolytic enhancerWyseure T, Vercauteren E, Gils A and Declerck PJKU Leuven, Leuven, Belgium
Objective: Investigation of simultaneous inhibition of Thrombin Acti-
vatable Fibrinolysis Inhibitor (TAFI) and Plasminogen Activator
Inhibitor-1 (PAI-1) as a robust enhancer of fibrinolysis.
Methods: Diabody Db-TCK26D6 · 33H1F7 was constructed by clon-
ing and assembling the variable domains of monoclonal antibodies (MA)
MA-TCK26D6 (which prevents plasmin-mediated activation of TAFI)
and MA-33H1F7 (which inhibits PAI-1), followed by expression in bac-
teria and purification. Inhibitory properties against TAFI and PAI-1
were determined by appropriate chromogenic functional assays. The
profibrinolytic properties were evaluated on a ROTEM analyzer using
human whole blood in which coagulation was initiated by tissue factor
and fibrinolysis was induced by a suboptimal concentration of rt-PA.
Results: Db-TCK26D6 · 33H1F7 was successfully produced and puri-
fied and was demonstrated to fully retain the inhibitory properties of
the corresponding parental MAs as confirmed by functional assays.
Viscoelastic measurements (ROTEM) revealed an enhanced profibrino-
lytic effect of the combined addition of MA-TCK26D6 (0.09 lM) and
MA-33H1F7 (0.09 lM) compared to the respective single addition
(based on the lysis index after 45 min (LY45), P < 0.04, n ‡ 5). Db-
TCK26D6 · 33H1F7 (0.18 lM, corresponding to an equal amount of
binding sites as the combined use of 0.09 lM of both MAs) caused a
significant increase in fibrinolysis compared to the combination of the
parental MAs (0.09 lM) (based on LY45, P = 0.05, n = 9).
Conclusion: Dual inhibition of TAFI and PAI-1 by Db-
TCK26D6 · 33H1F7 leads to an enhanced fibrinolysis. Its cross-reac-
tivity with mouse TAFI and mouse PAI-1 will allow its in vivo evalu-
ation as a profibrinolytic agent in various mouse models.
O3.6.03A TAFI-derived peptide enhances TAFI activation bythe serine proteases thrombin, plasmin and trypsinPlug T, Marx PF and Meijers JCMAcademic Medical Center, Amsterdam, the Netherlands
Objectives: Thrombin-activatable fibrinolysis inhibitor (TAFI) is a pro-
enzyme that, once activated, cleaves C-terminal lysine residues from
fibrin and thereby acts as an inhibitor of fibrinolysis. TAFI is proteolyti-
cally activated by thrombin, the thrombin/thrombomodulin complex,
plasmin, trypsin or neutrophil-derived elastase. The aim of this study was
to engineer peptides that affect activation of TAFI by these enzymes.
Methods: A peptide based on the TAFI sequence GGDDW
IYDLGIKYSFTIELR (G346-R365), two partly overlapping peptides,
S337-I356 and K357-E376 and mutants thereof in which one or more
aspartic acids were replaced for asparagines were synthesized. The
effects of these peptides on TAFI activation by thrombin, plasmin,
trypsine, or the thrombin-thrombomodulin complex were determined.
Results: The G346-R365 peptide increased TAFI activation by
thrombin, plasmin and trypsin respectively four, five and sixfold. No
effect was observed with activation by the thrombin/thrombomodulin
complex. The overlapping peptide S337-I356 gave a twofold increase
but the peptide K357-E376 showed no effect at all, so the amino
acids GGDDWIYDLGI are likely to cause the stimulating effects.
When one aspartic acid of this highly negatively charged peptide was
replaced by asparagine, the stimulating effect was nearly completely
lost. When two or three aspartic acids were replaced by asparagines,
the effect was gone indicating that the stimulatory effect was caused
by the negative charge.
Conclusions: TAFI-derived peptide G346-R365 selectively increases
TAFI activation and therefore this peptide is a useful tool to investi-
gate the role of the various TAFI activators in TAFI-mediated pro-
cesses such as fibrinolysis and infection.
O3.6.04TAFI deficiency plays a predominant role in thehyperfibrinolytic state observed in mice withcombined TAFI and PAI-1 gene deficiencyVercauteren E1, Peeters M1, Hoylaerts MF1, Lijnen HR1,Meijers JCM2, Declerck PJ1 and Gils A1
1KU Leuven, Leuven, Belgium; 2Academic Medical Center,
Amsterdam, the Netherlands
Objective: To unravel the relative function and importance of the an-
tifibrinolytic proteins Thrombin Activatable Fibrinolysis Inhibitor
(TAFI) and Plasminogen Activator Inhibitor-1 (PAI-1) through the
generation and characterization of mice with combined TAFI and
PAI-1 gene deficiency.
Methods and results: Mating of TAFI knock-out (KO) with PAI-1
KO mice resulted in the production of TAFI/PAI-1 double KO mice
that were viable, fertile and developed normally. In a tail vein bleed-
ing model, the bleeding time of the TAFI/PAI-1 double KO mice
(263 s) did not deviate significantly from that of the single KO mice
(170 and 168 s for TAFI KO and PAI-1 KO mice, respectively,
P > 0.05) nor of the wild-type (WT) counterparts (270 s, P > 0.05,
n = 10–28). Interestingly, in the presence of a suboptimal tPA con-
centration, rotational thromboelastometry (ROTEM�) measurements
revealed an enhanced lytic activity in whole blood samples of TAFI
KO and TAFI/PAI-1 double KO (complete lysis, P < 0.0001 com-
pared to WT mice, n = 5) in contrast to PAI-1 KO mice (hardly any
lysis, P > 0.05 compared to WT mice, n = 4) and WT mice. Fur-
thermore, the enhanced profibrinolytic effect was confirmed in vivo in
a mouse thromboembolism model as derived from the decreased
fibrin deposition in the lungs of TAFI KO mice (17 � 15%,
P < 0.05) and TAFI/PAI-1 double KO mice (26 � 20%, P < 0.05)
compared to WT mice (100 � 32%, n = 5) and PAI-1 KO mice
(97 � 57%, P > 0.05).
Conclusion: This study demonstrates that in TAFI and PAI-1 dou-
ble gene deficient mice, predominantly TAFI gene inactivation con-
tributes to the increased fibrinolytic capacity. Importantly, no
bleeding complications arise upon combined TAFI and PAI-1 gene
deficiency.
O3.6.05More on the effect of oral anticoagulant treatment onTAFI/TAFIa levels and clot resistance to fibrinolysisIncampo F1, Carrieri C1, Galasso R1, Di Serio F2, Scaraggi FA2,Woodhams B3, Semeraro N1 and Colucci M1
1Aldo Moro University; 2Policlinico, Bari, Italy; 3Stago,
Gennevilliers, France
We showed that plasma clots from warfarin-treated patients lyse
faster due to a reduced TAFI activation. Because blood cells influ-
ence fibrinolysis we evaluated the effect of oral anticoagulation
(OAT) on whole blood clot lysis and further investigated the mech-
anism behind the enhancement of fibrinolysis. Two hundred and
twenty-one consecutive patients on stable OAT (PT-INR 2.8 � 0.7)
and 132 controls were studied. Fibrinolysis resistance of plasma
clots (turbidimetric assay) and blood clots (thromboelastography)
was calculated as the lysis time of TF-induced clots exposed to 40
and 100 ng/mL t-PA, respectively. PAI-1 and TAFI levels were sim-
ilar in the two groups. OAT blood clots lysed significantly faster
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e19____________________________________________________________________________________
than control clots (37 � 13 vs. 47 � 13 min, P = 0.0002). Upon
addition of the TAFIa inhibitor PTCI (25 lg/mL), lysis time was
reduced by 17% in OAT and by 40% in controls, and the differ-
ence between the two groups disappeared. Similar data were
obtained with plasma clots. Thrombin generation (chromogenic
assay) and TAFI activation (ELISA) in OAT plasma amounted to
roughly 40% of controls, supporting a reduced thrombin-dependent
TAFI activation. Clot resistance of OAT plasma was normalized by
prothrombin replenishment but not by rFVIIa (100 U/mL) or by
the combination of factors VII, IX and X. The lysis time of OAT
clots was significantly correlated with thrombin generation parame-
ters but not with PT-INR. Surprisingly, the circulating levels of TA-
FIa/ai were higher in OAT patients (29 � 9 vs. 22 � 5 ng/mL,
P < 0.0001) despite a lesser in vivo clotting activation as indicated
by the reduced levels of VIIa/AT complex (75 � 10 vs.
110 � 39 pM, P < 0.0001). Our data indicate that OAT enhances
both plasma and blood fibrinolysis by reducing thrombin-dependent
TAFI activation, a phenomenon largely determined by the low pro-
thrombin level. The higher circulating levels of TAFIa in OAT sug-
gest that TAFI activation in vivo may be induced by enzymes other
than thrombin.
Oral communications 7; Fibrinolysis,Markers and MechanismsO3.7.01Fibrinolysis alterations in infertile women duringcontrolled ovarian stimulation: influence of BMI andgenetic componentsSticchi E, Romagnuolo I, Cellai AP, Lami D, Fedi S, Prisco D, NociI, Abbate R and Fatini CAOU-Careggi, Florence, Italy
Objectives: Women undergoing ovarian stimulation procedures may
exhibit alterations in haemostasis, thus determining both a hyperco-
agulable and hypofibrinolytic status which, in turn, may induce a
prothrombotic phenotype. We investigated fibrinolytic changes to
ovarian stimulation through a global test (CLT) and PAI-1 and
TAFI concentrations, and the influence of fibrinolysis polymorphisms
in genes encoding for fibrinogen chains (FGA, FGB, FGG), t-PA
(PLAT), TAFI (CBP2), FXIII (FXIIA1, FXIIIB), plasminogen
(PLG) and PAI-1 (PAI1) on these parameters.
Methods: We performed a prospective cohort study on 110 infertile
women undergoing ovarian stimulation procedure, such as in vitro
fertilization (IVF) or intracytoplasmic sperm injection (ICSI), by
evaluating fibrinolysis biohumoral and genetic parameters.
Results: Significant changes in fibrinolysis during ovarian stimulation
were found (CLT P = 0.003; TAFI P = 0.009 and PAI-1 P = 0.003).
CLT values, and TAFI and PAI-1 concentrations significantly
increased from baseline to T1 (P < 0.0001, P = 0.01, P = 0.005,
respectively), and decreased at T2, although remained higher than at
baseline. At T0 overweight women showed longer CLT, and higher
TAFI and PAI-1 concentrations than non-overweight women, and at
T1 twofold longer CLT and higher PAI-1 concentrations were
observed (P = 0.001 and P = 0.05, respectively). Significant differ-
ences of TAFI and PAI-1 concentrations during ovarian stimulation
according to TAFI and PAI1 polymorphisms were observed.
Conclusions: This study shows alterations of fibrinolysis, in particular
in overweight women, and the contribution of TAFI and PAI1 genes in
modulating fibrinolysis changes throughout ovarian stimulation cycle.
O3.7.02Difference in fibrinolytic potential between youngpatients with venous or arterial thrombosis is notexplained by PAISkov J, Sidelmann JJ, Gram J and Jespersen JUniversity of Southern Denmark, Esbjerg, Denmark
Objectives: In recent years, much debate has circulated around the poten-
tial causal relation between venous and arterial thrombotic disorders,
indicated by several shared risk factors. The present study investigates dif-
ferences in fibrinolysis between two groups of young patients, diagnosed
with arterial disease or venous thromboembolism (VTE), respectively.
Methods: Consecutive patients (n = 653), mean age 33 year, with
either VTE (n = 284) or arterial thrombosis (n = 369) were referred
to coagulation testing. Anthropometric measures, family history and
adherence to healthy lifestyle were registered and hereditary thrombo-
philic factors were determined. The fibrinolytic potential was mea-
sured as the activity of the euglobulin fraction of patient plasma
measured on bovine plasminogen-rich fibrin plates. PAI-1 activity
was determined with an enzymatic procedure.
Results: Patients with VTE were younger (32.2 vs. 33.8 years,
P < 0.05) and had higher BMI (27.4 vs. 25.6 kg/m2 P < 0.05) than
patients with arterial thrombosis. The percentage of men, smokers
and users of oral contraceptives was similar in the two groups. Fam-
ily history of VTE, BMI ‡ 30 kg/m2, Factor V Leiden and the Pro-
thrombin 20210A mutation were more prevalent in patients with
VTE than arterial thrombosis. PAI-1 activity did not differ signifi-
cantly between patients with VTE or arterial disease (median
12.3 IU/mL vs. 11.5 IU/mL, P = 0.1), but the fibrinolytic potential
was significantly lower for patients with VTE (P < 0.05). The latter
finding was repeated in the subgroup of patients with PAI-1 activity
within the normal range (n = 464).
Conclusions: The fibrinolytic potential is reduced in young patients
with VTE compared with patients with arterial disease. This finding
could not be explained by differences in PAI-1 activity, suggesting
that other determinants of the fibrinolytic potential are involved in
the pathophysiology of VTE.
O3.7.03Biological mechanisms responsible for MRI T1relaxation in venous thrombosisSaha P1, Andia ME2, Blume U2, Wiethoff A2, Schaeffter T2,Botnar RM2, Evans CE1, Ahmad A1, Patel AS1, Humphries J1,Grover S1, Nuthall K1, Modarai B1, Waltham M1 and Smith A1
1Academic Department of Surgery, St. Thomas’ Hospital; 2Division
of Imaging Sciences, King’s College London, London, UK
Background: Absolute quantification of Magnetic Resonance (MR)
longitudinal (T1) relaxation time (T1-RT) has been postulated as a
method to characterise the age and structure of venous thrombosis.
The biological mechanisms that affect MR signal during thrombus
propagation and its subsequent resolution are poorly understood. In
this study, we used an established model of thrombosis to investigate
the mechanisms that affect thrombus T1-RT in vivo.
Methods: An MRI 3D T1-mapping protocol was used to image vena
caval thrombi induced in male BALB/C, iNOS)/), and CCR2)/) mice
(n = 88). T1-RT and water diffusibility (ADC sequence) were quantified
1–28 days after thrombus induction. The total iron (mass spectrometry)
and Fe3+ content (Quantichrome) of thrombus was analysed. Fibrin
and red cell content was assessed by MSB histology and scanning elec-
tron microscopy. Monocyte phenotype was characterised by flowcytom-
etry (FACS) and in vivo activity visualized by intravital microscopy
(IVM) using Rag2)/)/cc)/)/CX3CR1+/GFP mice.
Results: Mean T1-RT change as the thrombus resolves. Total iron
content was greatest at 1 day (P < 0.001), which reflected early red
cell trapping by fibrin. As red cells lysed, the concentration of para-
magnetic Fe3+ increased until maximal at 7 day (P < 0.001). T1-
RTs were shortest when the levels of Fe3+ and water diffusibility
were greatest. T1-RTs were significantly longer in iNOS)/) mice
than wild-type controls (P < 0.001), and remained persistently short
in CCR2)/) mice (P < 0.001), which have an absence of inflamma-
tory monocytes. FACS analysis revealed macrophage heterogeneity
during thrombus resolution. IVM showed an exponential increase in
thrombus macrophage numbers with time.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e20 Abstracts____________________________________________________________________________________
Conclusions: This is the first study to show that T1-RT depends on both
accumulation of Fe3+ and water diffusibility. iNOS and CCR2 positive
macrophages appear to regulate iron metabolism and affect T1-RT.
O3.7.04Hemostasis, adhesion and clot stability characteristicsof human fibrinogen produced in the milk oftransgenic cowsCalcaterra J1, Carlson MA2, Pipinos II2, Johanning JM2, CordesCM2, van Veen H3, Nelson K4 and Velander WH1
1University of Nebraska, Lincoln, NE; 2University of Nebraska
Medical Center, Omaha, NE, USA; 3Pharming Technologies
B.V., Leiden, the Netherlands; 4Pharming Healthcare, Inc.,
DeForest, WI, USA
Objective: In the future, resuscitative and topical therapies for restor-
ing hemostasis disrupted by trauma will likely include recombinant
human fibrinogen (rFI). Previously, we found that rFI made in the
milk of transgenic cows possessed many molecular properties similar
to plasma-derived fibrinogen (pdFI). Here, we evaluate the wound
adhesion and hemostatic behavior of rFI in a lethal pig liver injury
model. We also evaluate the clot formation and stability of rFI in
normal human blood ex vivo.
Methods: The effects of rFI on time to clot initiation and clot strength
were investigated by thromboelastography (TEG) when added to normal
human blood at near hyperfibrinogenemic levels. Plasmin degradation
of rFI and pdFI clots created with human thrombin and factor XIIIa
(FXIIIa) were studied by TEG. The ability of rFI to covalently bind to
a2-antiplasmin (A2AP) by FXIIIa was studied by SDS-PAGE and wes-
tern blotting. The hemostatic efficacy of rFI as a component in a fibrin
sealant was investigated in vivo on lethal, Grade V stellate liver lacera-
tions in a swine model (N = 9). Adherence of the human fibrin made
from rFI to wounded tissue was examined by immunohistochemistry.
Results: The addition of rFI significantly increased thromboelasto-
graphic clot strength when added to normal human blood. Purified
rFI, made in the milk of cows, did not have detectable A2AP protein
or activity. As a result, plasmin (8–84 mU/mL) resulted in clot lysis
of rFI but not pdFI (9 mg/mL). In the presence of FXIIIa, A2AP
was found to covalently attach to rFI. rFI sealant resulted in a sta-
ble, wound adherent clot that stopped moderate and severe bleeding
in swine hemorrhage models.
Conclusions: These in vivo, ex vivo and in vitro studies indicate that
purified rFI produced in the milk of transgenic cows has potentially
desirable attributes for use in fibrinogen replacement therapy and as
a component of fibrin sealants.
O3.7.05Remodeling of clots without proteolytic digestionChernysh IN, Purohit PK, Nagaswami C and Weisel JWUniversity of Pennsylvania School of Medicine, Philadelphia,
PA, USA
Fibrin polymerization is a necessary part of hemostasis but clots can
obstruct blood vessels and cause heart attacks and strokes. The coagu-
lation cascade results in the cleavage of fibrinopeptides from fibrino-
gen, such that the resulting fibrin monomers rapidly polymerize to
form a gel, made up of a three-dimensional network of fibers. The poly-
merization reactions are specific and controlled, involving strong bonds
via knob-into-hole interactions to convert highly soluble fibrinogen
into insoluble fibrin. It has long been assumed that fibrin clots and
thrombi are stable structures until proteolytic digestion. Here we dem-
onstrate, using the technique of fluorescence recovery after photoble-
aching, that there is turnover of fibrin in an uncrosslinked clot. We also
show that a peptide representing the knobs involved in fibrin polymeri-
zation can compete for the holes and dissolve a preformed fibrin clot,
or increase the fraction of soluble oligomers, with striking rearrange-
ments in clot structure. Thus, fibrin is an equilibrium polymer, and
monomers or oligomers can dissociate and re-associate, primarily from
the surface of the fibers. These results imply that in vivo clots or
thrombi are more dynamic structures than has been previously
believed. This mechanism may account for some embolization that
commonly occurs in patients, leading to serious clinical consequences.
Furthermore, clots in the vasculature may be remodeled as a result of
local environmental conditions, suggesting a target for therapeutic
intervention of thrombosis in which the flow of blood is occluded.
Oral communications 8; Fibrinolysis andThrombolysis, Clinical AspectsO3.8.01Fibrinolysis at the interface of blood vessel wall andthrombiKolev K1, Rottenberger Z1, Komorowicz E1, Szabo L2 andMachovich R1
1Semmelweis University; 2Chemical Research Center, Hungarian
Academy of Sciences, Budapest, Hungary
Objectives: At sites of blood vessel wall injury neutrophil leukocytes
are recruited and through release of proteases (neutrophil elastase,
matrix metalloproteinases MMP-8 and MMP-9) they modify the ves-
sel wall structure. The neutrophil-mediated proteolysis generates col-
lagen fragments, releases glucoseaminoglycans (chondroitin sulfate
CS, dermatan sulfate DS) and exposes extracellular matrix proteins
(e.g. decorin) at sites of fibrin formation. Here we characterize the
effect of these vessel wall components on the lysis of fibrin.
Methods: MMP-8 digested collagen fragments, isolated CS, DS, gly-
cated decorin and its core protein were used to prepare mixed matri-
ces with fibrin. The structure of fibrin was examined with scanning
electron microscopy and images were analyzed morphometrically.
Fibrin matrices containing plasminogen were used to monitor the
plasmin formation initiated by tissue plasminogen activator (tPA)
applied to the clot surface. Fibrin formation and dissolution were
measured with turbidimetry.
Results: The presence of modifiers (at 10-fold lower mass concentra-
tion than fibrinogen) resulted in increase in fiber diameter from
85 nm in pure fibrin to 187 nm in the presence of glycated decorin.
Plasminogen activation on the surface of composite clots was slowed
down to a variable degree (10–30%). The lytic susceptibility of the
modified fibrin structures was increased; the time to 50% lysis by
plasmin was reduced approximately twofold for all examined modifi-
ers (only the core protein of decorin produced a moderate reduction
of the lysis time by 25%). The examined modifiers had no significant
effect on the plasmin and thrombin activity measured on synthetic
chromogenic substrates and neither was plasminogen activation
affected in the absence of fibrin.
Conclusion: Products of proteolysis in the blood vessel wall modify
the structure of fibrin and render it poor cofactor in tPA-dependent
plasminogen activation, but more susceptible to lysis by plasmin.
O3.8.02Decreased fibrinolytic resistance of plasma clotscontaining red blood cellsCarrieri C, Incampo F, Ammollo CT, Semeraro F, Semeraro Nand Colucci MAldo Moro University, Bari, Italy
It has recently been shown that red blood cells (RBC) confer lytic
resistance to clots derived from purified fibrinogen, which is in con-
trast with some older studies. Here, we investigated the effect of
RBC on fibrinolysis under more physiological conditions. Washed
RBC (10–40% v/v) were added to plasminogen-containing fibrinogen
(3 mg/mL) or autologous plasma. Following t-PA addition (20–
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e21____________________________________________________________________________________
150 ng/mL of extracellular volume) clotting was induced by CaCl2-
thrombin or CaCl2-tissue factor (TF). Fibrinolysis was evaluated by
thromboelastography (lysis time) or by the release of FITC-labeled
FDPs (% lysis). When clots were prepared from purified fibrinogen,
RBC inhibited fibrinolysis in a concentration-dependent manner
(33% lysis time prolongation at 40% RBC). The effect disappeared
in the presence of the GPIIb/IIIa inhibitor abciximab but not when
the TAFIa inhibitor PTCI was added. On the contrary, when clots
were prepared from plasma, RBC accelerated fibrinolysis (21% lysis
time reduction at 40% RBC). The effect of RBC on clot lysis was (i)
concentration-dependent; (ii) maintained in the presence of platelets
(3 · 105/lL); (iii) insensitive to abciximab; (iv) abolished by PTCI,
suggesting a TAFI-mediated mechanism. Replacement of RBC with
40% (v/v) polystyrene microbeads (6 lm diameter) did not apprecia-
bly influence clot lysis, ruling out an effect due to volume occupancy.
Qualitatively similar results were obtained with FITC-fibrinogen clot
lysis model. Thrombin generation (two-stage clotting assay) and
accumulation of TAFIa/ai (ELISA) in defibrinated plasma activated
with TF-Ca++ were 63% and 27% higher in the presence of 40%
RBC. However, by functional assay (using fibrin as the substrate) the
amount of TAFIa activity detected in RBC-containing plasma was
30% lower than in cell-free plasma, suggesting that RBC inhibit TA-
FIa activity (towards fibrin). These data unveil a new function of
RBC and underscore the potential role of these cells in fibrinolysis
modulation.
O3.8.03Fibrin-targeted molecular MRI identifies venousthrombi susceptible to thrombolysisSaha P1, Andia M2, Grover S1, Wiethoff AJ2, Schaeffter T2, PatelAS1, Humphries J1, Nuthall K1, Modarai B1, Waltham M1, BotnarRM2 and Smith A1
1Academic Department of Surgery, St. Thomas’ Hospital;2Division of Imaging Sciences, King’s College London, London,
UK
Objective: Venous thrombus composition determines the success of
thrombolysis and treatment is based on clinical judgment of throm-
bus age. Thrombolysis is associated with significant morbidity (bleed-
ing and stroke) and needs to be better targeted. This study aims to
investigate a fibrin-specific MRI contrast agent (EP-2104R) to stage
venous thrombus organisation and assess suitability for thrombolysis.
Methods: Venous thrombi were induced in BALB/C mouse vena
cava (n = 72) and imaged by MRI (3T Philips Achieva) at days 2, 4,
7, 10, 14, 21 after thrombus induction (n = 12/gp). Each group was
scanned pre- and 2 h post-injection of EP-2104R (EPIX Pharmaceuti-
cals, 8.0 lmol/kg). An inversion recovery 3D segmented gradient
echo (TFE) sequence was performed and T1 maps of the thrombus
calculated using custom-made software implemented in MATLAB.
Fibrin contrast uptake in the thrombus was correlated with fibrin
content as assessed by histology using Martius Scarlet Blue (MSB)
trichrome (n = 6/gp). A separate group underwent systemic venous
thrombolysis (10 mg/kg of tissue plasminogen activator [Actilyse, Bo-
ehringer Ingelheim, Germany]) at each time point (n = 6/gp).
Twenty-four hours after thrombolytic treatment, mice were re-
scanned to examine restoration of caval blood flow using a phase
contrast sequence.
Results: After injection of EP-2104R, large areas with high signal
intensity and short T1 relaxation times were observed. A larger visu-
alised thrombus enhancement volume in post-contrast images was
demonstrated in younger thrombi. Contrast uptake positively corre-
lated with the fibrin content of the thrombus (R2 = 0.97, P < 0.01).
ROC curve analysis showed that a mean thrombus T1 relaxation
time < 630 ms on post contrast images had sensitivity of 94% and
specificity of 99% to predict successful thrombolysis (AUC 0.993
[CI95%: 0.98–1.00]).
Conclusions: Fibrin-targeted MRI can be used to stage venous
thrombus organisation and allow accurate stratification of thrombi
amenable to lysis.
O3.8.04Diabetes modulates the fibrinolytic properties ofaspirin without altering the platelet inhibitory actions:a possible mechanism for aspirin treatment failureKurdee Z, Mamaniat A, Phoenix F, Rice P, Grant PJ and Ajjan RADivision of Cardiovascular and Diabetes Research, Faculty of
Medicine and Health, University of Leeds, Leeds, UK
Aspirin, which inhibits platelet function and modulates fibrin clot
lysis, has been used for prevention from atherothrombotic disease in
diabetes, but recent evidence suggests reduced clinical efficacy by
unclear mechanisms. We hypothesised that hyperglycaemia in diabe-
tes interferes with aspirin action, resulting in reduced cardiovascular
protection. Therefore, we investigated the effects of ex vivo addition
of aspirin to whole blood on platelet function and plasma clot lysis
in healthy controls and in type 1 diabetes (T1DM) subjects devoid of
complications and on insulin only therapy. Platelet function was
monitored by multiplate assay and fibrinolyis was tested using turbi-
dimetric analysis.
Platelet aggregation to arachidonic acid (AA) after treatment with 0,
1 and 10 mg/L aspirin in 24 T1DM subjects was 60.2 � 3.2,
51.2 � 3.5 and 32.0 � 2.9 AU, respectively (P < 0.05). Similar find-
ings were documented for healthy controls (65.3 � 3.9, 60.2 � 3.5
and 27.7 � 3.5 AU, respectively; P < 0.05). Fibrinolysis in 15
T1DM subjects showed no difference in the presence of 0, 1 and
10 mg/L aspirin (919 � 110, 932 � 113 and 812 � 73 s, respectively;
P > 0.1). In contrast, fibrin clot lysis was affected by aspirin treat-
ment in healthy controls (1009 � 112, 577 � 44 and 685 � 57 s,
respectively and P < 0.05).
Our data indicate that diabetes has little effect on platelet inhibition
by aspirin following AA stimulation. However, enhanced fibrin clot
lysis by aspirin is lost in the presence of diabetes and may be the
main mechanism for aspirin treatment failure in this condition.
Future work is warranted to investigate the relationship between clin-
ical aspirin treatment failure and fibrin clot lysis in diabetes subjects
on aspirin therapy.
O3.8.05On the opposing effects of factor XIIa on clot stabilityKonings J1, Govers-Riemslag JWP1, Philippou H2, Ariens RAS2
and ten Cate H1
1Laboratory for Clinical Thrombosis and Haemostasis,
Department of Internal Medicine, Cardiovascular Research
Institute Maastricht, Maastricht University Medical Center,
Maastricht, the Netherlands; 2Division of Cardiovascular and
Diabetes Research, Section on Mechanisms of Thrombosis,
Faculty of Medicine and Health, University of Leeds, Leeds, UK
There is strong homology between proteins of the contact and fibri-
nolytic systems. We have previously shown that FXII and a-FXIIa
bind with high affinity to fibrin(ogen) (range Kd = 2.5–4.6 nM) and
a-FXIIa makes the clot denser, thereby increasing resistance to fibri-
nolysis (Blood 2011, PMID = 21828145). Conversely, activated FXII
(FXIIa) has been reported to directly convert plasminogen into plas-
min. Here, we investigated the relative contribution of FXIIa to clot
stability by either mechanism.
We determined plasminogen activation by tissue-type plasminogen
activator (tPA) by monitoring the amidolytic activity of plasmin and
clot lysis by turbidity in the presence of a-FXIIa. Plasminogen
(32.5 nM) and a-FXIIa (0–60 nM) were added to a 96-well plate
coated with fibrin. Formation of plasmin was started by adding tPA
(12 nM) and monitored using S-2251. Clot lysis time was assessed
using purified fibrinogen (3 lM), in the presence of a-FXIIa (0 or
125 nM), plasminogen (300 nM) and tPA (15, 150 or 1500 pM) clot-
ted with thrombin (2.5 nM) and CaCl2 (5 mM).
We found that a-FXIIa is able to activate plasminogen: the forma-
tion of plasmin increased from 13.9 mOD/min in the absence to
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e22 Abstracts____________________________________________________________________________________
20.3 mOD/min in the presence of 60 nM a-FXIIa. Furthermore, a-FXIIa reduces the clot lysis time at all tPA concentrations tested.
We conclude that FXIIa is able to activate plasminogen and thereby
increase the rate of fibrinolysis. Since FXIIa also increases the fibrin
clot density, the question remains which of these two mechanisms of
a-FXIIa plays a more important role in vivo.
Oral communications 9; Protein Structureand FunctionO4.9.01Investigating sources of factor XIIIa specificity forglutamine-containing substratesMaurer MC, Doiphode PG and Malovichko MVUniversity of Louisville, Louisville, KY, USA
Objective: Factor XIIIa (FXIIIa) catalyzes formation of gamma-glut-
amyl-epsilon-lysinyl crosslinks in the fibrin network. Further knowl-
edge is needed on the characteristic properties that define FXIIIa
specificity for its substrates. Glutamine-containing peptides based on
a2-antiplasmin (a2AP), S. Aureus fibronectin binding A (FnbA), and
thrombin activatable fibrinolysis inhibitor (TAFI) can serve as effec-
tive substrate models.
Methods: A MALDI-TOF mass spectrometry assay was used to
directly follow loss of reactive substrate and formation of product. A
Q-containing peptide substrate interacted with FXIIIa with release of
ammonia. Glycine ethylester then served as the lysine-like mimic to
complete the reaction. Substrate loss was monitored as a function of
quenched time points and velocities calculated. Kinetic parameters
were determined along with standard errors of the mean.
Results and conclusions: Studies with a2AP (1 NQEQVSPLT
LLKLGN 15) showed that Q2 at the P1 position was reactive. Q4 at
the P-2 and K12 at the P-10 positions played roles in supporting sub-
strate binding. Assays with S. Aur FnbA (100 SGDQRQVDLIP 110
and 100 SGDQRQVDLIPKKAT 114) revealed that K111, K112 at
the P-8, P-9 positions enhanced FXIIIa binding interactions and may
serve as an extra anchor. When examining the P-1 position, the a2AP
(1–15, E3R) mutation could be accommodated whereas S. Aur FnbA
(100–114 R104E) could not. Excess acidic character may be problem-
atic for the FXIIIa active site. TAFI (1 FQSGQVLAALPRTSR 15)
has one reactive glutamine but showed weaker kinetic properties
likely due to the second Q being shifted to the P-3 position. Individ-
ual substrate residues, the FXIII active site, and distant binding sites
function to create an environment for promoting cross-linking reac-
tions. (NIH HL068440).
O4.9.02Kringle domains of plasmin are essential for efficientdegradation of fibrin but not fibrinogenKim PY, Stafford AR, Fredenburgh JC, Gross PL and Weitz JIThrombosis and Atherosclerosis Research Institute, McMaster
University, Hamilton, ON, Canada
The kringle (K) domains of plasmin (Pn) mediate its Lys-dependent
interaction with fibrin (Fn) and are essential for coordinated Fn deg-
radation. However, the contribution of the K domains for Pn-medi-
ated fibrinogen (Fg) degradation is unknown. This is an important
question because Micro-Pn, a truncated derivative that lacks all five
K domains, is under investigation as a fibrinolytic agent for catheter-
directed therapy. To explore the role of the K domains in Fg degra-
dation we compared Micro-Pn and Lys-Pn in terms of their capacity
to degrade Fg and Fn in the absence or presence of �ACA. Both
Micro-Pn and Lys-Pn shortened the Fn lysis time (the time to half
maximal decrease in turbidity) in a concentration-dependent fashion.
On average, Micro-Pn lysis times were 14-fold higher. With �ACA
addition, the lysis times with Lys-Pn were prolonged 11-fold, and
were comparable to those with Micro-Pn without �ACA. Although,
�ACA prolonged lysis times with Micro-Pn twofold, �ACA had no
effect on the time course of degradation as determined by SDS-
PAGE analysis. When Fg degradation was analyzed by SDS-PAGE,
�ACA addition rendered the time course of degradation with Lys-Pn
similar to that observed with Micro-Pn. �ACA, however, did not
affect Micro-Pn. Even with �ACA, however, Micro-Pn and Lys-Pn
were both efficient in cleaving Fg; a-chain was consumed by 10 and
15 min, b-chain by 30 and 45 min for Lys-Pn and Micro-Pn, respec-
tively. Our results demonstrate that the K domains are critical for
efficient lysis of Fn, but not Fg. Because Micro-Pn is resistant to
inhibition, it may affect fibrinolysis by reducing the local Fg concen-
tration, thereby attenuating thrombus accretion.
O4.9.03The X-ray crystal structure of full-length humanplasminogenLaw RHP1, Caradoc-Davies T2, Cowieson N2, Horvath AJ3, QuekAJ1, Amarante Encarnacao J1, Steer D1, Cowan A1, Zhang Q1, LuBGC3, Pike RN1, Smith AI1, Coughlin PB2 and Whisstock JC1
1Department of Biochemistry and Molecular Biology, Monash
University; 2Australian Synchrotron, Clayton; 3Australian Centre
for Blood Diseases, Monash University, Prahran, Melbourne, FL,
Australia
Plasminogen is the pro-enzyme of plasmin, an essential protease
that mediates fibrinolysis, wound healing, tumour cell mobility and
bacterial invasiveness. Circulating plasminogen, which comprises
an N-terminal Pan-apple (PAp), five kringle (KR1-5) and a C-ter-
minal serine protease (SP) domains, adopts a closed, activation
resistant conformation. The kringle domains bind to lysine residues
present in receptors and substrates. These interactions govern pro-
enzyme recruitment to clots and cell surface and concomitantly
trigger a conformational re-arrangement of the molecule to an
open form that can be cleaved and converted to plasmin by tissue-
or urokinase-type Plasminogen Activator (tPA and uPA). Despite
extensive study, a mechanistic understanding of the plasminogen
activation system remains limited through a lack of structural
insights.
Here, the structure of closed plasminogen reveals that the PAp and
SP domains maintain the closed conformation through interactions
made throughout the kringle array. Chloride ions bridge the PAp/
KR4 and SP/KR2 interfaces, explaining the physiological role of
chloride in stabilizing the closed conformer. Differences in glycosyl-
ation alter the position of KR3, illuminating reported functional
differences between the two plasminogen glycoforms. Further, access
to the activation loop targeted for cleavage by tPA and uPA is
blocked through the position of the KR3/KR4 linker sequence.
Inter-domain interactions also block all kringle ligand-binding sites
apart from that of KR-1, suggesting that the latter domain governs
pro-enzyme recruitment to targets. Strikingly, analysis of an inter-
mediate structure suggests that plasminogen conformational change
to the open form is initiated through KR-5 peeling away from the
PAp domain.
O4.9.04Retarded tissue plasminogen activator and facilitatedplasmin action in fibrin modified by carboxypeptidaseBKolev K1 and Longstaff C2
1Department of Medical Biochemistry, Semmelweis University,
Budapest, Hungary; 2Biotherapeutics, Haemostasis Section,
National Institute for Biological Standards and Control, South
Mimms, UK
Removal of C-terminal lysine residues that are continuously exposed
in the course of fibrin digestion by plasmin is an established anti-
fibrinolytic mechanism dependent on the plasma carboxypeptidase
TAFIa. Because it removes not only lysine, but also arginine residues
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
Abstracts e23____________________________________________________________________________________
that are exposed at the time of fibrinogen clotting by thrombin, here
we address its potential effects on fibrin structure using the stable
carboxypeptidase B (CPB), which shows the same substrate specific-
ity. If 0.5–2 g/L fibrinogen is clotted in the presence of 8 U/mL CPB,
a denser fibrin network is formed with thinner fibers (the median
fiber diameter decreases from 138–144 to 89–109 nm as established
with scanning electron microscopy). If clotting is initiated in the pres-
ence of 5–10 lM arginine, a similar decrease in fiber diameter (85–
95 nm) can be measured. Because fine fibrin structure is known to
confer resistance to tissue plasminogen activator (tPA)-induced fibri-
nolysis, such modification of the fibrin architecture supports the anti-
fibrinolytic role of TAFI, which is confirmed by the slower progress
of fluorescent tPA induced lysis of CPB-treated fibrin monitored with
confocal laser microscopy. However, if lysis is initiated with plasmin
and the stable CPB is continuously present in the fibrin, the rate of
dissolution monitored with turbidimetry is accelerated (half-lysis time
decreases from 42 to 26 min) which may be explained by elimination
of non-productive plasmin binding sites. Thus, in view of this surpris-
ing finding the short life-span of TAFIa appears to contribute to the
restriction of the plasmin action.
O4.9.05Impact of fibrin structure on fibrinolysis by differentplasminogen activators: studies using fluorescentconfocal microscopyWhyte CS and Mutch NJInstitute of Medical Sciences, University of Aberdeen, Aberdeen,
UK
Fibrin structure influences the binding of fibrinolytic proteins and
determines its susceptibility to lysis by plasminogen activators (PA).
Polyphosphate (polyP), a biomolecule found in platelet dense granules,
down-regulates tPA-mediated fibrinolysis by decreasing binding of
plasminogen and tPA to fibrin.1 Here we use fluorescent confocal
microscopy to visualise fibrinolysis. Fibrinogen and plasminogen were
fluorescently-labelled with DyLight488 and DyLight633, respectively.
Clots were prepared with a proportion of the total fibrinogen
(2.65 lM) and plasminogen (1.25 lM) as fluorescently-labelled protein
(1% and 20%, respectively) and clotting initiated with thrombin
(0.25 U/mL) and CaCl2 (5 mM). Clear visualisation of the fibrin
network was achieved in the presence of 488-fibrinogen. Addition of
polyP65 (5 lM) during fibrin formation generated shorter protofibrils
interspersed with aggregates of fibers, consistent with our previous
observations.1 Incorporation of 633-plasminogen produced a weak dis-
persed signal throughout the fibrin network. When exogenous PA
(75 nM) was applied a dramatic accumulation of plasminogen was
observed at the lysis front. uPA-mediated clot lysis illustrated a
broader zone of plasminogen accumulation and more rapid fibrinoly-
sis. This may reflect the ability of uPA to penetrate deeper into the
fibrin network. Surprisingly, the changes in fibrin structure induced by
polyP did not delay uPA-mediated lysis, in fact, lysis appeared to be
enhanced. Confocal microscopy using fluorescently-labelled proteins
provides an excellent tool for analysing the different modes of action of
tPA and uPA. These studies highlight the direct relationship between
fibrin structure and its resistance to fibrinolysis by different PAs.
Reference1. Mutch NJ, Engel R, Uitte de Willige S, Philippou H, Ar-
iens RA. Polyphosphate modifies the fibrin network and down-regu-
lates fibrinolysis by attenuating binding of tPA and plasminogen to
fibrin. Blood 2010;115(19): 3980–88.
ª 2012 International Society on Thrombosis and Haemostasis 10 (2012) e10–e24
e24 Abstracts____________________________________________________________________________________