From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
Feedback Activation of Factor XI by Thrombin in Plasma Results in
Additional Formation of Thrombin That Protects Fibrin
Clots From Fibrinolysis
By Peter A. Kr. von dern Borne, Joost C.M. Meijers, and Bonno N. Bourna
Recently, an alternative pathway forfactor XI activation has
been described in which factor XI is activated by thrombin.
Patients with a factor XI deficiency bleed mostly fromtissues
with high local fibrinolytic activity. Therefore, the role of
thrombin-mediated factor XI activation in both fibrinformation and fibrinolysis was studied in a plasma system. clotting was induced by the
addition of tissue factoror thrombin
t o recalcified plasma in the presence or absence of tissuetype plasminogen activator, after which clot formationand
lysis were measured using turbidimetry. Thrombin-mediated activationof factor XI was found t o take place in plasma
under physiologic conditions in the absence of a dextran
sulfate-like cofactor. At hightissue factorconcentrations, no
effect of factor XI was seen on the rateof fibrin formation.
Decreasing amounts of tissue factor resulted in a gradually
increasing contribution of factor XI to therate of fibrin formation. In addition, thrombin-mediated factor XI activation
resulted in an inhibition of tissue-type plasminogen activator-induced lysis of the clot. This inhibition occurred even
at tissue factor concentrations at which no effect of factor
XI was observed on fibrin formation. Trace amounts of activated factor XI (1.25 pmol/L, representing 0.019’0 activation)
were capable of completely inhibiting fibrinolysis in our system. The inhibitory effect was found t o be mediated by
thrombin that is
additionally generated in a factor Xl-dependent manner via the intrinsic pathway andiscapable of protecting the clotagainst lysis. We also observed that formation ofadditional thrombin continued after the clothad been
formed. We conclude that thrombin-mediated factor XI activation can take place in plasma. The presence of factor XI
during coagulation results in the formation of additional
thrombin within the clotcapable of protecting this clot from
fibrinolytic attack. The large amounts of thrombin that are
formed by the intrinsic pathway via factor XI may play an
important rolein the procoagulant and thrombogenic state
of clotsand may therefore have important clinical and therapeutic implications.
0 7995 by The American Societyof Hematology.
F
mediated factor XI activation takes place in plasmaand plays
a role in both coagulation and fibrinolysis.
ACTOR XI PARTICIPATES in the contact phase of
blood coagulation. A deficiency of one of the other
proteins of the contact system (factor XII, prekallikrein, or
high molecular weight kininogen) does not result in a bleeding disorder. For this reason, activation via this system is
not believed to be physiologically important. In contrast,
patients deficient in factor XI do suffer from variable bleeding abnormalities. Recently, an alternative pathway for factor
XI activation has been described. Factor XI was found to be
activated by thrombin in the presence of dextran sulphate,
which acts as a surface or template for activation.’,’ The
relevance of this discovery was questioned when it was
shown that high molecular weight kininogen, to which factor
XI is bound in plasma, and fibrinogen almost completely
inhibited a~tivation.~
However, factor XI activation by
thrombin was shown to take place in plasma in the presence
of sulfatides as a s ~ r f a c eFurthermore,
.~
in an invitro system
using purified coagulation factors it was shown that this
activation can take place in the absence as well as the presence of a surface, whereas the inhibitory effect of high molecular weight kininogen mainly occurred in the presence of
a surface.5
Despite these findings, the actual occurrence and importance of thrombin-mediated factor XI activation in plasma
has yet to be proven. Sulfatides are normally not present in
the vascular system, while the role of other potential surfaces
such as glycosaminoglycans is uncertain because they can
also exert a heparin-like inhibitory effect on coagulation.6
Patients with a deficiency of factor XI are prone to bleeding from tissues with a high local fibrinolytic activity (urinary tract, nose, oral cavity, or tonsils).’ The mechanism
behind this clinical observation is unclear. We hypothesized
that factor XI activation by thrombin may play an important
role in this process. We therefore decided to study the effect
of thrombin-mediated factor XI activation on fibrin formation and tissue-type plasminogen activator (t-PA)-mediated
fibrinolysis in plasma. We were able to show that thrombinBlood, Vol 86, No 8 (October 15). 1995: pp 3035-3042
MATERIALS AND METHODS
Materials. Fresh frozen plasma and citrated whole blood were
obtained from the local blood bank. Bovine serum albumin (BSA;
fraction V) was purchased from Sigma (St Louis, MO). Enzygnost
F1+2 kits were obtained from Behring (Marburg, Germany). A
monoclonal antibody (MoAb) capable of blocking activated factor
XI1 (OT2) was a kind gift of Dr Erik Hack (Central Laboratory of
the Netherlands Red Cross Blood Transfusion Services, Amsterdam,
The Netherlands). An MoAb against factor VI1 (2G3-2c) has been
described.*MoAbs against factor IX (2G4) and factor X (10F1) and
rabbit polyclonal antibodies against factor I1 were prepared according to standard procedures. The rabbit polyclonal antiserum was
immunopurified on a Sepharose column to which purified prothrombin was immobilized. Other chemicals obtained were the best grade
available.
Proteins. Factor XI was purified from fresh frozen plasma using
murine anti-factor XI MoAb (XI-5)immobilized on CNBr-activated
Sepharose, as described earlier.5Factor M and X were isolated from
fresh frozen plasma.’ All preparations appeared homogeneous using
sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
From
the
Department of Haematology, University Hospital
Utrecht, Utrecht, The Netherlands.
Submitted March 13, 1995; accepted June 22, 1995.
Supported in part by a fellowship
of the Royal Netherlands Academy of Arts and Sciences.
Address reprint requests to Peter A. Kr. von dem Borne, MD,
Department of Haematology G.03.647, University Hospital Utrecht,
PO Box 85500, 3508 GA Utrecht, The Netherlands.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
‘‘advertisement’’in accordance with 18 U.S.C. section I734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
0006-4971/95/8608-0030$3.00/0
3035
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3036
PAGE). The specific coagulant activities for factor IX, X and XI
were 210, 90, and 235 U/mg, respectively. Factor XIa was prepared
from factor X1 using P-factor XIIa, as previously described.'" Factor
IXwas activated using factor XIa," after which factor XIawas
removed from the factor IXa preparation using Q-Sepharose chromatography. Factor X was activated with Russell's Viper Venom."
Recombinant human tissue factor (Innovin) was obtained from Baxter (Unterschleil3heim, Germany). t-PA was purchased from Chromogenix (Molndal, Sweden). Purified human thrombin was a kind
gift from Dr Walter Kisiel (University of New Mexico, Albuquerque,
NM). Recombinant hirudin (1 2,000 U/mg) wasa generous gift from
R. Wallis (Ciba Geigy, Horsham, UK).
Dejicient plasmas. Factor XII-deficient plasma was obtained
from a congenitally deficient patient. Plasma deficient in both factor
XI1 and factor XI was prepared by passing this factor XII-deficient
plasma over a Sepharose column to which a murine anti-factor XI
antibody (XI-5) was immobilized. Less than 0.01 % of factor XI was
present in the factor XI and XI1 double-deficient plasma as determined using enzyme-linked immunosorbent assay (ELISA). Plasmas
deficient in either factor VII, IX, or X were obtained by passing
plasma from a healthy donor over Sepharose columns towhich
murine MoAbs against these factors were coupled. Plasma deficient
in factor I1 was obtained by passing plasma from a healthy donor
over a Sepharose column to which immunopurified rabbit polyclonal
antibodies against factor I1 were coupled. The undilutedflowthrough fractions of the columns were pooled and used as deficient
plasma. The deficient plasmas were tested with clotting assays and
found to contain less than 1% of the deficient factor, whereas the
other coagulation factors remained normal.
Turbidimetry. Thrombin-mediated factor X1 activation in
plasma was studied by monitoring the change in turbidity during
fibrin formation and lysis at 405 nmin a microplate reader.I3 An
increase in turbidity indicated fibringel assembly, whereas a decrease in turbidity indicated clot lysis.I4 All experiments were performed in citrated plasma recalcified with CaCI, (finalconcentration,
17 mmol/L) resulting in a free Ca2+concentration of 2.3 mmoUL.
A mixture of tissue factor (different dilutions of Innovin) or thrombin
(20 nmol/L) and calcium necessary for recalcification was added to
75 p L of plasma to initiate clotting. The volume was adjusted to
150 p L with HEPES buffer (25 mmol/L HEPES, 137 m m o K NaCI,
3.5 m m o m KCI, 3 mmoUL CaCI,, and 0.1 % BSA), resulting in a
final plasma concentration of 50%. After mixing, 100 p L of the
reaction mixture was pipetted into a microplate and then incubated
at 37"C, during which turbidity wasmeasuredat
different time
points. To study the effect of factor XI on fibrin formation, plasma
deficient in both factor XI and XI1 was reconstituted with factor XI
and preincubated for 30 minutes at room temperature before use.
The final concentration of factor XI was 12.5 nmol/L (approximately
0.5 U/mL). In some experiments, a murine MoAb (XI-l) that blocks
the activation of factor IXby factor XIa" was preincubated with
the plasma. In addition, experiments were performed in minimally
diluted plasma with a final plasma concentration of 94%.
Fibrinolysis was studied using the same assay as described above
except that 30 U/mL of t-PA was added. To studythe effect of
factor XI we used plasma deficient in both factor XI and XI1 that
was reconstituted with factor XI and normalplasma that was preincubated for 30 minutes at room temperature with the murine MoAb
(XI-l) that blocks factor XIa activity. In several experiments, factor
IXa, Xa, or XIa or additional thrombin was also included, as described in the legends to the figures. The concentrations of the components are indicated as final concentrations.
F1 +2 assay. Fibrin formation was induced in plasma by thrombin in the presence of t-PA, as described above. The reaction mixture
was incubated at 37°C in polypropylene tubes instead of microtiter
plates. At different time intervals, 100 pL of HEPES buffer con-
VONDEM
BORNE, MEIJERS, ANDBOUMA
taining 10 mmol/L of EDTA and 50 U/mL of hirudin was added to
the fibrin clot. After thorough mixing for 1 minute and centrifugation
(2 minutes at 15,OOOg), the amount of prothrombin activation was
determined in the supernatant by measuring the concentration of
fragment F1 +2 with a commercial ELISA.
RESULTS
The role of factor XI in tissue factor inducedjbrinf o m tion in plasma. The role of factor XI in tissue factor-induced coagulation was studied using a fibrin formation assay.
In this assay, tissue factor was added to recalcified plasma
and the rate of fibrin formation was followed by measuring
the increase in turbidity. Tissue factor-induced fibrin formation was measured in plasma deficient in both factor XI and
factor XI1 to avoid any interference of contact activation.
To study the influence of factor XI on the rate of fibrin
formation, this plasma was reconstituted with factor XI. At
high tissue factor concentrations, no difference in fibrin formation was observed in the presence or absence of factor
XI (Fig IA). With decreasing amounts of tissue factor, a
gradual increase in the rate of fibrin formation was observed
in the presence of factor XI compared with the rate of fibrin
formation in the absence of factor XI. Because the recombinant tissue factor preparation contains phosphatidylserine/
phosphatidylcholine phospholipids that might contribute to
the rate of fibrin formation in this assay, control experiments
were performed in which extra phospholipids were added,
showing identical results as those depicted in Fig IA (data
not shown). Preincubation of factor XIwiththeantibody
that blocks factor XIa activity (XI-l) resulted in a complete
abrogation of the effect of factor XIon fibrin formation
(Table l), indicating that this effect is caused by factor XI
and not by a contaminant in the factor XI preparation. Control experiments were performed with minimally diluted
plasma (94% plasma concentration) to investigate the effect
of varying concentrations of HK, fibrinogen,3or other inhibitory factors in plasma. An identical effect of factor XI on
fibrin formation was observed; fibrin formation rates in the
presence or absence of factor XI were 56.2 t 3.7 and 41.2
2 2.4 mOD/min, respectively. From these experiments it
was concluded that factor XI contributes to the formation of
thrombin at low tissue factor concentrations.
The effect of factor X I onthe lysis of a tissue factorinduced clot. To study the effect of factor XI on clot lysis,
we included t-PA in the tissue factor-induced fibrin formation assay. Formation and subsequent lysis of the fibrin clot
was determined by following the increase and decrease in
turbidity. In separate experiments, we studied the effect of
the t-PA concentration on the clot lysis time. A t-PA concentration of 30 U/mL was chosen because it induced complete
lysis of the fibrin clot in a reasonable amount of time (60 to
90 minutes). At the highest tissue factor concentration tested,
almost no clot lysis was observed in 80 minutes, either in
the presence or absence of factor XI (Fig 1B). Decreasing
concentrations of tissue factor resulted in an increase of the
rate of clot lysis in the absence of factor XI, whereas in the
presence of factor XI only a minor effect on clot lysis was
observed. This finding indicates that clot lysis is inhibited
in the presence of factor XI. Effects of factor XI on clot
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
FACTOR XI ACTIVATION BY THROMBIN IN PLASMA
3037
A 0.80
8
0.60
I
0
v)
2
0.40
0.20
4
0
12
8
16
20
Time (min)
0.80
0.60
Fig 1. The role of factor
XI in tissue factor-induced
in plasma. Plasma deficient
fibrin formation and lysis
in both factor XI and X11 was preincubated in the
presence (solid symbols) orabsence lopen symbols)
of factor XI (12.5nmollL). (AI Coagulation was initiated by adding tissue factor and calciumnecessary
for recalcification. Fibrin formation was
measured in
time as the change in t u r b i d w at 405 nm. For clarification, the areas between the curves representing
identical tissue factor dilutions wereshaded. (B1 To
study the influence of fibrinolysis, t-PA (30 UlmL)
was included in the incubation mixture used in (A).
Fibrin formation and subsequent lysis were measured in time as the change in turbidity at 405 nm.
Tissue factor (dilutions ofInnovin): lo3 (a, O), 10' (A,
AI, 3 x 10' (V,VI, lo5 (+, 0).3 x lo5 (*,h),buffer
(M, U). Each point represents the mean 2 SEM of
three independent measurements.
0.40
0.20
lysis could be observed at tissue factor concentrations at
which factor XI had no effect on fibrin formation.
The effect of factor XI on the lysis of a thrombin-induced
clot. The use of varying tissue factor concentrations results
in formation of different amounts of thrombin. Thrombin
can activate not only factor XI but also factor V and VIII,
resulting in amplification of the intrinsic pathway. Therefore,
we repeated the experiments using a fixed concentration of
thrombin (20 nmol/L) to induce clotting. Clot formation was
observed within a few minutes and lysis of the clot was
studied in the presence and absence of factor XI. Again, an
0
20
40
60
80
Time (min)
antifibrinolytic effect of the plasma concentration of factor
XI was observed (Fig 2A). Essentially the same results were
obtained with minimally diluted plasma (94% plasma concentration) instead of the normally used50% plasma concentration (data not shown).
The ant$brinolytic efect qf factor X I results from factor
Xla generated by thrombin-mediatedactivation. Traces of
factor XIa in our factor XI preparation instead of factor Xla
generated by thrombin-mediated factor XI activation might
be responsible for the observed antifibrinolytic effect of factor XI. Therefore, we repeated the experiments with different
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3038
VON DEM BORNE,MEIJERS,
Table 1. Tissue Factor-Induced Fibrin Formation in the Presence
and Absence of Factor XI: Effect of an Antibody That Blocks Factor
Xla Activity (XL11
Factor XI
-
+
+
XI-l Antibody
~
+
-
+
Fibrin Formation Rate
imOD/min)
39.0 .C 3.0
38.1 _f 1.5
58.4
40.3
_f
It
0.7
1.3
Plasma deficient in both factor XI and XI1 was preincubated in the
presence (+) or absence ( - ) of factor XI or MoAb XI-l. Coagulation
was initiated by adding tissue factor (dilution of Innovin, lo5) and
calcium necessary for recalcification. Fibrin formation was measured
in time as the change in turbidityat 405 nm. Results are depicted as
fibrin formation rate, which is defined as the slope of the linear part
of the fibrin formationcurve. The means t SD of three independent
experiments are shown.
concentrations of factor XIa to titrate the concentration of
factor XIa necessary for inducing an antifibrinolytic effect.
In these experiments, thrombin (20 nmol/L) was added to
recalcified plasma in the presence of t-PA and factor XIa.
Picomoles of factor XIa could completely inhibit fibrinolysis
(Fig 2A). An effect corresponding to 0.125 to 1.25 pmol/L
factor XIa was found with 12.5 nmoYL of factor XI.
To measure the amount of factor XIa present in our factor
XI preparation, weused the fibrin formation assay. This
assay is extremely sensitive for factor XIa.Recalcified
plasma was clotted in the absence of thrombin or tissue
factor by adding different concentrations of factor XIa or
12.5 nmol/L of factor XI. The clotting time induced by factor
XI was similar to the clotting time induced by 1.25 to 12.5
fmol/L of factor XIa, indicating that the factor XIa concentration in our factor XI preparation was in that range (Fig
2B). This amount is considerably lower than the amount of
factor XIa needed to induce a direct antifibrinolytic effect,
indicating that the observed antifibrinolytic effect is the result of newly formed factor XIa generated by thrombinmediated activation.
Lysis of a thrombin-induced clot: eflect of an antibody
blocking factor XIa activity. The experiments described
above were performed with plasma from a single factor XIIdeficient donor. To exclude donor-specific effects, experiments were performed with citrated plasma from a pool of
40 healthy volunteers that was preincubated with a murine
anti-factor XI antibody (XI-l) that blocks activation of factor
IX by factor XIa.” In addition, an MoAb capable of blocking
factor XIIa activity (OT2) and a control antibody against
protein C inhibitor (API-93) were also included. After preincubation with the antibodies, the plasma was recalcified,
thrombin and t-PA were added, and turbidity was measured.
An increased rate of clot lysis was observed in the presence
of XI-l (Fig 3), again confirming the antifibrinolytic effect of
factor XI. The presence or absence of OT2 had no influence,
suggesting that factor XII, which is known to play a role
in fibrinolysis,’6 apparently has no part in the effects we
observe.
Inhibition ofjibrinolysis by factor H ( a ) is dependent on
the intrinsic pathway of coagulation. The antifibrinolytic
ANDBOUMA
effect of factor XI can be caused either by a direct effect of
activated factor XI or byan indirect effect caused by the
generation of another activated clotting factor in the intrinsic
pathway. Therefore, we studied the effect of addition of
activated factor IX (50 pmol/L, representing 0. I % activation)
to factor XI-deficient plasma on clot lysis. Factor IXa was
also found to inhibit fibrinolysis (Fig 4A), indicating that
the presence of factor XIa itself is not a prerequisite for this
effect.
The lysis of a thrombin-induced clot was also studied in
plasmas deficient in either factor 11, VII, IX, or X in the
presence or absence of XI-l (Fig 4B). The antibody did not
influence clot lysis in plasmas deficient in factor 11, IX, or
X. Only in factor VII-deficient plasma was an effect of the
antibody found. Itwas concluded that an intact intrinsic
pathway is necessary for the inhibition of fibrinolysis by
factor XIa.
To determine which factor of the intrinsic pathway was
responsible for inhibition of fibrinolysis, lysis of a thrombin
induced clot was studied in prothrombin-deficient plasma
to which also factor IXa, Xa, or additional IIa was added.
Concentrations of activated factor IXand X representing
complete activation had noinfluence on fibrinolysis, whereas
a concentration-dependent antifibrinolytic effect of factor Ila
was observed (Fig 4C). We concluded that formation of
additional thrombin by activation of the intrinsic pathway
of coagulation is responsible for the inhibition of fibrinolysis.
F1 +2formation during coagulation andjibrinolysis. Because the inhibitory effect on fibrinolysis was found to be
mediated by additionally generated thrombin, the activation
of prothrombin during clotting and clot lysis was studied in
our system by measuring the activation fragment 1 +2. After
preincubation with or without XI-l, normal plasma was recalcified and both thrombin and t-PAwere added. After
the addition of EDTA and hirudin, fibrin was removed by
centrifugation and F1+2 was measured in the supernatant.
Significant differences in prothrombin activation were observed in the presence or absence of the antibody, indicating
an important contribution of the intrinsic pathway (Fig 5).
Prothrombin activation was found to continue after clot formation, indicating ongoing thrombin generation within the
clot. The presence of t-PA did not affect prothrombin activation, showing that this activation is not influenced by clot
lysis.
DISCUSSION
In this report, we show that factor XI activation by thrombin can take place in plasma under physiologic conditions
in the absence of a dextran sulfate-like cofactor. This activation was studied by monitoring tissue factor-induced fibrin
formation in plasma using turbidimetry. At high tissue factor
concentrations, no effect of the presence or absence of factor
XI was seen on fibrin formation, probably because sufficient
amounts of thrombin had already been formed via the extrinsic pathway to induce rapid fibrin formation. At low tissue
factor concentrations, thrombin formation was probably limited and additional fibrin formation could now be provided
by thrombin generated via the intrinsic pathway initiated by
thrombin-mediated factor XI activation. Only differences in
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3039
FACTOR XI ACTIVATION BY THROMBIN IN PLASMA
A
0.70
0.60
a
0
c
a
0.50
e0
v)
n
a
0.40
0.30
0.20
10
20
30
40
50
60
40
50
Time (min)
Fig 2. The antifibrinolytic effh
of factor XI res u b from factor Xla generated by thrombin-mediated activation. (A) To determine the concentration
of factor Xla necessary for an antifibrinotytic effect,
plasma d M e n t in both factor XI and XI1 was praincubated with factor XI (12.5 nmol/L) or buffer. h a g
ulation was initiated by adding cald?m necessary
for recakifidon, thrombin (20 nmol/L), t-PA(30 U!
mL), and factor Xla t o the plasma. Fibrin formation
and lysis were measured in time as the change in
turbid* at 405 nm. Factor XI (m); buffer (Cl); and
factor Xla at 1.25 pmol/L (A), 125 fmol/L (01, 12.5
fmol/L (01. or 1.25 fmol/L (V).(B1 To determine the
amount of factor Xla in our factor XI preparation,
plasma deficient in both factor XI and XI1 was preincubated with factor XI (12.5 nmol/L). Calcium necessary for recalcificationand factor Xla ware added to
tho plasma. Fibrin formation was measured in time
as the change in turbidm at 405 nm. The same symbola aa depicted in (A) were used in (B). The means
SD of three independent experimentsare shown.
B
0.80
a
0.60
0
c
cp
e0
U)
n
a
0.40
0.20
*
the rate of fibrin formation were seen, whereas virtually no
effect on the lag time of fibrin formation was observed,
illustrating the accessory role of factor XI on thrombin formation, but not on the initiation of fibrin formation. Interestingly, this phenomenon had already been observed in an in
vitro system with purified coagulation factors.”
In contrast to our results, no significant activation of factor
XI by thrombin in plasma could be measured in an earlier
study.I8 However, in this study, autoradiography was used
to detect factor XI activation. This technique is not sensitive
enough to detect the small amounts of activation that were
observed in this study.
0
10
20
30
Time (min)
Coagulation is thought to be initiated by exposure of blood
to tissue factor. While thrombin and fibrin are being formed
via the extrinsic pathway, tissue factor pathway inhibitor
quickly turns off the activation by inhibiting the factor VTIatissue factor complex. Continuation of coagulation is now
dependent on the formation of additional thrombin via the
intrinsic pathway.Ig According to our findings, maintenance
of thrombin formation can take place in plasma as a result of
thrombin-mediated factor XI activation leading to additional
fibrin formation.
The effect of factor XI on clot lysis was studied by adding
a fibrinolytic component (t-PA) to our plasma system. Clot
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3040
VON DEM BORNE, MEIJERS, AND BOUMA
0.70
0.60
Q)
0
c
lu
e0
2
a
0.50
0.40
0.30
0.20
10
20
30
40
50
Time (min)
11
VI1
IX
X
def
def
def
def
20 nM
40 nM
80 nM
160 nM
0.00 0.10 0.20 0.30 0.40
Absorbance change
Fig 4. Inhibition of fibrinolysis by factor Xl(a) is dependent on the
was initfated by adding
intrinsic pathway of coagulation. Coeguiation
calcium necessaryfor recakification, thrombin (20 nmoVU, and t-PA
(30 UlmL). Fibrin formation and lysis were measured in time es the
change in turbidity at 405 nm. In this
figure the amount of fibrinolysis
is shown and is dofind as the change in abrwbance between 10 and
80 minutes. Eech bar
represents the mean f SD of three i W n & n t
experiments. (A) Factor Xl-ddcient plasma with or without factor
IXa ( 5 0 pmol/L). (B) Factor II-,VII-, IX-, or X-dsfcient plasma preincubated with Xi-l (250 nmolll; solid bars) or buffer (open bars). (C)
Factor il-deficient plasma with factor IXa ( 4 5 nmol/L), Xa (75 nmoll
L), or Ila (20 to 160 nmoilL).
60
Fig 3. Lysis of a thrombin induced clot: effect of
an antibody blocking factor XIa activity. Normal
plasma was preincubatd with Xi-l(250 nmol/L; A,
A), API-93 (250 nmollL; V, V),
or buffer (0.0)
in the
presence (solid symbols) or absence (open symbols)
of OT2 (675 nmol/L). Coagulation was initiated by
adding calcium necessaryfor recalcification, thrombin (20 nmol/L), and t-PA
(30 U/mL). Fibrinformation
and lysis were measured in time as the change in
turbidity at 405 nm. The means 2 SD of three independent measurements are given.
lysis was found to be inhibited by the presence of factor XI,
even at tissue factor concentrations at which factor XI had
no effect on fibrin formation. We also observed that the
inhibitory effect of factor XI on fibrinolysis was caused by
the generation of additional thrombin. Additional thrombin
formation continued after clot formation had taken place. In
the presence of factor XI, the intrinsic loop of coagulation
could generate large amounts of thrombin within the clot
that were capable of inhibiting fibrinolysis. So, on the one
hand onlysmall amounts of thrombin provided by the extrinsic pathway were sufficient for clot formation, whereas on
the other hand large amounts of thrombin formed by the
intrinsic pathway were necessary for protecting the clot
against lysis.
By directly adding activated factor XI to the plasma, we
could show that even a very small amount of factor XIa
(1.25 pmol/L, representing 0.01% activation) was capable of
completely inhibiting fibrinolysis. This finding corresponds
very well with our earlier findings that trace amounts of
factor XIa can play a role in blood coagulation because of
the amplification potential of the coagulation cascade.' This
amplification could also easily overcome any inhibitory effect of high molecular weight kininogen and fibrinogen on
the thrombin-mediated factor XI a~tivation.~
At this moment it is not clear how thrombin exerts its
inhibitory effect on fibrinolysis. Several possible explanations exist, such as a direct effect of thrombin on the quantity
or structure of the formed fibrin or enhanced cross-linking
of the fibrin clot by activation of factor XIII.20 Activated
protein C is a known accelerator of clot lysis in vitro. It was
recently proposed that enhancement of clot lysis by activated
protein C is mediated by downregulation of the generation
of thrombin and factor XIIIa:' which is in agreement with
our findings and suggests an important role for factor XIII.
Further investigations are needed to elucidate the mechanism
of inhibition.
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3041
FACTOR XI ACTIVATION BY THROMBIN IN PLASMA
5 00
0.60
400
8
0.50
h
300
S
la
Fig 5. F1+2 formationduringcoagulationand
fibrinolysis.Normal plaama was preincubatedwith XI1 (250 nmol/LI orbuffer.Clottingwasinitiatedby
adding thrombin (20 nmol/L) and calcium required
recalcification
for
symbols)
in(solid
the presence
or
absence (opensymbols1 of t-PA (30 UlmL). (Left axis)
Fibrin formation and lysis ware measured
in time as
turbidity the
in change
at 405 nm. XI-l IO); buffer
(0).(Right axis) At differenttime points, hirudin 150
U/mLlEDTA
and
(20 mmol/LI were
added.
After
the
removal of fibrin, prothrombin fragment
1+2 was
measured in the supernatant. XI-l (A,A); buffer (V,
V).
e
g
c
v
(U
0.40
200
a
+
l-
a
0.30
100
0.20
0
The findings in this report may explain the clinical manifestations of factor XI deficiency. Bleeding is mostly seen
after surgery or trauma, often with a delay.” The formation
of the hemostatic plug is probably not inhibited, but the
fibrin that replaces the platelet plug after fibrinous transformation may be more susceptible to fibrinolysis. This susceptibility is most important in places with highlocal fibrinolytic
activity, ie, at those sites where bleeding in patients with a
factor XI deficiency is seen. As might be expected, fibrinolytic inhibitors such as tranexamic acid and c-aminocaproic
acid are successfully being used for the treatment of bleeding
in patients with a factor XI d e f i c i e n ~ y . ~ , ~ ~
Our findings may have some important implications.
Thrombin is believed to play a very important role in both
venous and arterial t h r o m b o s i ~Thrombin
.~~
bound to a clot
retains coagulant activity, although it is relatively inaccessible to inhibition by heparin-antithrombin 111,24 turning the
clot into a powerful thrombogenic surface. For this reason,
inhibitors of thrombin (heparin or hirudin) are used alone as
antithrombotics or as adjuvants in fibrinolytic treatment of
these thrombotic ~tates.’~.’~
A major disadvantage of these
agents is that their local antithrombotic effect parallels a
systemic anticoagulant effect. Therefore, treatment is presently hampered by bleeding complications, although failure
to achieve reperfusion and occurrence of subsequent reocclusion demand higher doses of heparin or hirudin.
Activation of the intrinsic pathway by thrombin-mediated
factor XI activation may play an important role in venous
and arterial thrombosis. The factor XI-dependent thrombin
generation that was observed to continue within the clot
is especially of interest, because itmay contribute to the
thrombogenic state of the clot. Inhibition of thrombin-mediated factor XI activation, eg, with antibodies or peptides,
may result in clots that are less thrombogenic and more
susceptible to fibrinolysis. Furthermore, the risk of bleeding
complications may be low, because this therapy would not
be accompanied by strong anticoagulant effects. Thrombindependent coagulation at the site of the clot would be inhibited, whereas tissue factor-dependent hemostasis elsewhere
0
10
20
30
40
50
60
70
Time (rnin)
would be less affected. Inhibition of coagulation may turn
out to be more effective at the start of the coagulation cascade by interfering with thrombin-mediated factor XI activation than at the end by using heparin or hirudin.
NOTE ADDED IN PROOF
Recently, Bajzk et alZ7published an article describing
the existence of a thrombin-activatable fibrinolysis inhibitor
(TAFI),which may be a plausible explanation for the thrombin-dependent inhibition of fibrinolysis in our system.
ACKNOWLEDGMENT
The authors thank the personnel of the Red Cross Blood Bank of
Utrecht for providing us with the donor plasma. We thank Dr H.
Karel Nieuwenhuis for referring the factor XU-deficient patient to
us. We also thank Drs Walter Kisiel, R. Wallis, and Erik Hack for
the generous gifts of thrombin, recombinant hirudin, and anti-factor
XI1 antibodies, respectively. We are grateful to Dr Jan J. Sixma for
critically reading the manuscript.
REFERENCES
1. Naito K, Fujikawa K: Activation of human blood coagulation
factor XI independent of factor XII. Factor XI is activated by thrombin and factor XIa in the presence of negatively charged surfaces. J
Biol Chem 266:7353, 1991
2. Gailani D, Broze GJ: Factor XI activation in a revised model
of blood coagulation. Science 253:909, 1991
3. Scott CF, Colman R W Fibrinogen blocks the autoactivation
and thrombin-mediated activation of factor XI on dextran sulfate.
Roc Natl Acad Sci USA 89:11189, 1992
4. Gailani D, Broze GJ:Factor XII-independent activation of factor XI in plasma:Effects of sulfatides on tissue factor-induced coagulation. Blood 822313, 1993
5. von dem Borne PAKr, Koppelman SJ, Bouma BN, Meijers
JCM: Surface independent factor XI activation by thrombin in the
presence of high molecular weight kininogen. Thromb Haemost
72:397, 1994
6. Marcum JA, Rosenberg RD: Anticoagulantly active heparinlike molecules from vascular tissue. Biochemistry 23:1730, 1984
7. Asakai R, Chung DW, Davie EW, Seligsohn U: Factor XI
deficiency in Ashkenazi Jews in Israel. N Engl J Med 325:153, 1991
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
3042
8. van 't Veer C, Hackeng TM, Delahaye C, Sixma JJ, Bouma
BN: Activated factor X and thrombin formation triggered by tissue
factor on endothelial cell matrix in a flow model: Effect of tissue
factor pathway inhibitor. Blood 84: 1132, 1994
9. Miletich JP, Jackson CM, Majerus PW: Properties of the factor
Xa binding site on human platelets. J Biol Chem 253:6908, 1978
10. Van der Graaf F, Greengard JS, Bouma BN, Kerbiriou DM,
Griffin JH: Isolation and functional characterization of the active
light chain of activated human blood coagulation factor XI. J Biol
Chem 258:9669, 1983
1 I . Ahmad SS, Rawala-Sheikh R, Walsh PN: Comparative interactions of factor IX and factor IXa withhuman platelets. J Biol
Chem 264:3244, 1989
12. Meijers JCM, Tijburg PNM, Bouma BN: Inhibition of human
blood coagulation factor Xabycy,-macroglobulin.
Biochemistry
26:5932, 1987
13. Bajzir L, Fredenburgh JC, Nesheim M: The activated protein
C-mediated enhancement of tissue-typeplasminogenactivator-induced fibrinolysisin a cell-free system. J Biol Chem 265: 16948,1990
14. Carr ME, Powers PL, Jones MR: Effects of poloxamer 188
on the assembly, structure and dissolution of fibrin clots. Thromb
Haemost 66:565, 1991
15. Meijers JCM: Regulation of the Blood Coagulation Mechanism by Plasma Proteinase Inhibitors. PhD Thesis, State University
of Utrecht, Utrecht, The Netherlands, 1988, p 93
16. Kluft C, Dooijewaard G, Emeis JJ: Role of the contact system
in fibrinolysis. Semin Thromb Hemost 13:50, 1987
17. Lawson JH, Kalafatis M, Stram S, Mann KG: A model for
the tissue factor pathway to thrombin. 1. An empirical study. J Biol
Chem 269:23357, 1994
18. BrunnCe T, La Porta C, Reddigari SR, Salerno VM, Kaplan
VONDEMBORNE,MEIJERS,
AND BOUMA
AP, Silverberg M: Activation of factor XI in plasma is dependent
on factor XII. Blood 81:580, 1993
19. Davie EW, Fujikawa K, Kisiel W: The coagulation cascade:
Initiation, maintenance and regulation. Biochemistry 30:10363, 1991
20. McDonagh J: Structure and function of factor XIII, in Colman
RW, Hirsch J, Marder VJ, Salzman EW (eds): Hemostasis and
Thrombosis, Basic Principles and Clinical Practice. Philadelphia,
PA, Lippincott, 1987, p 289
21. Gruber A, Mori E, del Zoppo GJ, Waxman L, GriffinJH:
Alteration of fibrin network by activated protein C. Blood 83:2541,
1994
22. Kitchens CS: Factor XI: A review of its biochemistry and
deficiency. Semin Thromb Hemost 1755, 1991
23. Hemker HC: Thrombin generation, an essential step in haemostasis and thrombosis, in Bloom AL, Forbes CD, Thomas DP,
Tuddenham EGD (eds): Haemostasis and Thrombosis, v01 1 (ed 3).
Edinburgh, UK, Churchill Livingstone, 1994, p 477
24. Hogg PJ, Jackson CM: Fibrin monomer protects thrombin
from inactivation by heparin-antithrombin 111: Implications for heparin efficacy. Proc Natl Acad Sci USA 86:3619, 1989
25. Hirsh J: Heparin. N Engl J Med 324:1565, 1991
26. Cannon CP, McCabe CH, Henry TD, Schweiger MJ, Gibson
RS, Mueller HS, Becker RC, Kleiman NS, Haugland JM, Anderson
JL, Sharaf BL, Edwards SJ, Rogers WJ, Williams DO, Braunwald
E: A pilot trial of recombinant desulfatohirudin compared with hirudin inconjunction with tissue-type plasminogen activator and aspirin
for acute myocardial infarction: Results of the thrombolysis in myocardial infarction (TIMI) 5 trial. J Am Col1 Cardiol 23:993, 1994
27. Bajzk L, Manuel R, Nesheim ME: Purification and characterization of TAFI, a thrombin-activatable fibrinolysis inhibitor. J Biol
Chem 270:14477, 1995
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
1995 86: 3035-3042
Feedback activation of factor XI by thrombin in plasma results in
additional formation of thrombin that protects fibrin clots from
fibrinolysis
PA von dem Borne, JC Meijers and BN Bouma
Updated information and services can be found at:
http://www.bloodjournal.org/content/86/8/3035.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.