Keywords
coagulation - vitamin K-dependent proenzyme - clotting factors - prothrombin - prethrombin-1
Introduction
Plasmatic hemostasis is the result of a cascade-like proteolytic activation of inactive
zymogens.[1] In its penultimate stage, the concerted action of activated coagulation factors
and their respective cofactors leads to the transformation of prothrombin into thrombin.[2] When present in sufficient concentration, thrombin causes the fibrinogen in blood
plasma to coagulate in the wound area and, to form, jointly with activated platelets,
an insoluble fibrin matrix. Thrombin also plays an important role in the extrinsic
pathway, where platelets and the clotting factor cofactors V and VIII are activated.[3]
Thrombin generation from prothrombin is a complex and rigorously controlled enzymatic
process.[1] To prevent it from shifting toward thrombogenicity, inhibitors, additional coagulation
factors, and cofactors such as factor Xa and cofactor Va as well as platelets and
certain endothelial factors are required.[4]
[5]
[6]
Prothrombin is a vitamin K-dependent proenzyme and has a complex structure, characterized
by an N-terminal Gla-domain and the Kringle 1-domain (fragment 1), the Kringle 2-domain
(fragment 2), and the serine protease domain. For the generation of thrombin, prothrombin
needs to be cleaved at two cleavage sites: R271 and R320.[7]
[8] Cleavage at R320 results in the formation of the enzymatically active meizothrombin.
Subsequent cleavage at R271 results in the release of thrombin. When R271 is cleaved
first, the enzymatically inactive prethrombin-2 is generated, which can be transformed
into thrombin by further cleavage at R320.[9] In the presence of cofactor Va, cleavage occurs at R320, and meizothrombin is generated.
In the absence of factor Va, on the other hand, cleavage occurs at R271, and prethrombin-2
is generated.[10]
Lanchantin et al[11] recognized early on that in a coagulation test using prothrombin deficient plasma,
prothrombin activity is lost, due to the action of thrombin on prothrombin. The two
split products resulting from the action of thrombin on prothrombin were ultimately
designated prethrombin-1 and fragment 1.
Prethrombin-1 is an enzymatically inactive split product which arises after cleavage
of fragment 1 from prothrombin by thrombin in a feedback reaction.[12] Prethrombin-1 is not recognized as a component within the activation pathway leading
from prothrombin to thrombin. Prethrombin-1 and prethrombin-2 do not contain the Gla-domain
necessary for membrane binding capacity and are therefore not efficiently activated
by prothrombinase bound to phospholipid vesicles.[13]
Quantitatively, prothrombin is the most prominent procoagulant factor in plasma. From
the work of Mann and coworkers[14] we know that high prothrombin concentrations are a risk for thrombosis. Prothrombin
has a relatively long half-life in blood, which implies that the repeated administration
of nonactivated and activated prothrombin complex concentrates (PCCs) increases the
prothrombin concentration in plasma to twice or three times the normal value. In clinical
use, prothrombin was identified as a major thrombogenic agent in PCCs.[15]
Highly purified recombinant prothrombin has been shown to be effective in achieving
hemostasis.[16] A clinical study to evaluate the safety, toxicity, and pharmacodynamics of recombinant
human prothrombin was discontinued prematurely for reasons that were not disclosed.[17]
Bleeding is considered a leading cause of mortality. If untreated, severe or chronic
hemorrhaging might lead to organ failure, seizures, coma, joint damage, and eventually
death. Even with treatment, severe bleeding is often fatal. The risks of fatal bleeding,
major bleeding, and intracranial and gastrointestinal hemorrhage are often underestimated.
Treatment options to arrest bleeding include activated or nonactivated PCCs, and recombinant
factor VIIa. PCCs are produced from human plasma.[18] Their composition is complex as they contain clotting factors such as prothrombin,
factor VII, factor IX, and factor X as well as the anticoagulant factors protein C
and protein S. Activated prothrombin complexes such as FEIBA (Takeda) have an even
more complex composition, as they additionally contain activated factor VII and activated
factor X.[19]
In addition to bleeding in congenital clotting disorders like hemophilia, bleeding
in acquired disorders often connected to treatment with direct oral anticoagulants
(DOACs) continues to be a key complication, affecting 2 to 4% of DOAC-treated patients.[20]
[21]
[22]
In the present preclinical studies, we aimed to identify and explore the potential
of prethrombin-1 as a new treatment option for patients who bleed or are at a risk
of bleeding.
Materials and Methods
Materials
Prethrombin-1 as well as prothrombin were manufactured in the laboratory of Biomedizinische
Forschung & Bio-Produkte AG, Vienna, Austria. FEIBA was purchased from Baxter Healthcare
Corporation (Westlake Village, California, United States), and Beriplex from CSL Behring
GmbH (Marburg, Germany). Rivaroxaban was purchased from Nschem Shanghai, China, dissolved
in DMSO (1000 ug/mL) and diluted with NaCl 0.9%. Sheep polyclonal antibody to human
coagulation factor VIII (FVIII) was procured from Haematologic Technologies, Essex
Junction, Vermont, United States.
Animals
Animals were obtained from Charles River Laboratories, Sulzfeld, Germany.
All experiments were approved by the animal care committee of the Center for Biomedical
Research and Translational Surgery, Medical University of Vienna, Austria.
Study Design
Prethrombin-1 was tested for hemostatic efficacy in rabbits and mice, and for thrombogenicity
in rabbits, both species rendered hemophilic by FVIII antibody administration. Studies
included dose escalation and comparison with prothrombin, Beriplex, and FEIBA. In
one study, the hemostatic efficacy of prethrombin-1 was investigated in mice after
rivaroxaban treatment.
Statistics
Descriptive statistics were used in all studies.
Analysis of 2-between group data was performed using t-statistics or the Mann–Whitney U-test. For comparison between treatment regimens
of more than two groups, the Kruskal–Wallis test was used. Post hoc comparisons were
performed with Bonferroni correction. A value of p <0.05 was considered to indicate statistical significance.
Fig. 1 SDS-PAGE of prethrombin 1 obtained from incubation of prothrombin with thrombin.
Lane 1: prothrombin purified with chromatography from human plasma (15 µL, diluted
1:400, nonreduced). Lane 2: prothrombin after incubation with thrombin (15 µL, diluted
1:200, reduced). Lane 3: prothrombin and prethrombin-1 fragment separated with chromatography
from incubation mixture (15 µL, diluted 1:32, reduced). Lane 4: molecular weight standard
BioRad Precision Plus Protein (5 µL). Lanes 5–7: prethrombin-1 purified with chromatography
from incubation mixture (5, 10, 15 µL, diluted 1:300, reduced). Lanes 8–10: prethrombin-1
purified with chromatography from incubation mixture (5, 10, 15 µL, diluted 1:300,
nonreduced). Samples were heated with 4 × sample buffer (BioRad, 4x Laemmli Sample
Buffer) and reduced with dithiothreitol DTT, loaded on gradient gel (BioRad, 4–20%
Mini-PROTEAN TGX Precast) and stained with Coomassie dye (BioRad, Bio-Safe Coomassie
Stain).
Fig. 2 Blood loss in FVIII inhibitor rabbits after i.v. treatment with different doses of
prethrombin-1 or sodium chloride. Individual values are marked with an “x” and means
with a horizontal line (-). Median blood loss untreated is the median of all animals
before prethrombin-1 treatment. i.v., intravenous.
Fig. 3 Clot weight (Wessler test) in FVIII-inhibitor rabbits after i.v. treatment with prethrombin-1,
prothrombin, Beriplex, and FEIBA. Individual values are marked with an “x” and means
with a horizontal line (-). Standard deviations are presented in bars. i.v., intravenous.
Fig. 4 Blood loss in FVIII-inhibitor mice after i.v. treatment with prethrombin-1 compared
with untreated animals. Individual values are marked with an “x” and means with a
horizontal line (-). Standard deviations are presented in bars. NaCl means physiological
saline 0.9%. Two concentrations of prethrombin-1 are displayed. The number of animals
per group (n) is shown. i.v., intravenous.
Fig. 5 Blood loss in mice after rivaroxaban administration and after reversal with prethrombin-1.
Individual values are marked with an “x” and means with a horizontal line (-). Standard
deviations are presented in bars.
Manufacturing of Prethrombin-1
All chemicals and reagents were purchased from Merk Millipore (Burlington, Massachusetts,
United States) and Thermo Fisher Scientific (Waltham, Massachusetts, United States).
Protein standards were purchased from CoaChrom (Maria Enzersdorf, Austria). Chromatography
resins and equipment were purchased from Cytiva (Marlborough, Massachusetts, United
States) and BioRad (Hercules, California, United States). Ultra- and dia-filtration
(UF/DF) units were purchased from Sartorius (Goettingen, Germany). Coagulation assay
reagents and devices were purchased from Werfen (Barcelona, Spain).
Starting material for the purification of prothrombin (FII) was an intermediate eluate
from a human plasma pool containing coagulation factors II, VII, IX, and X (Sanquin,
Amsterdam, The Netherlands). The frozen solution was melted to room temperature in
a water bath, and after dilution with deionized water was adsorbed batchwise on Sephadex
DEAE A50. Desorption of bound protein was performed stepwise with increasing concentration
of NaCl in citrate. Buffers used for this scope and all following buffers used during
purification of FII before UF/DF contained up to 2 mM benzamidine hydrochloride to
preserve the nonactivated form of coagulation factors. Concentrations of FII in samples
were monitored in IU with a standard coagulation assay with FII depleted plasma on
the ACL TOP Testing System. The fraction containing major amounts of FII from the
raw purification step was further processed consecutively on the AEKTA Pure system
with a weak anion exchanger (MacroPrepDEAE) and a multi-modal anion exchanger (Capto
ADHERE ImpRes). Desorption of bound FII was performed with increasing chloride concentration
and adding propylene glycol and arginine to the elution buffer. For polishing, FII
was first concentrated in a centrifugal UF/DF device with 30 kDa molecular weight
cutoff, and buffer was exchanged on a size exclusion resin (Sephacryl S300 PrepGrade).
The FII preparation was stored at −20°C before use for analysis or further processing.
Preparation of prethrombin-1 (Pre-1) started with purified FII to which α-thrombin
(FIIa) with a ratio of 0.5 IU FIIa per 1 IU FII. After an incubation period of 24 hours
in the mixture obtained, Pre-1 was separated from residual FII and FIIa using chromatography
in binding mode with the multi-modal resin (Capto ADHERE ImpRes) and in nonbinding
mode with an affinity gel (Benzamidine Sepharose FF). In the nonbinding mode, any
protein possessing an accessible serine protease active site will bind to the resin,
while the zymogen prethrombin-1, lacking such binding capability, undergoes elution
from the column, resulting in a higher degree of purification. Purity and conversion
rates from FII to Pre-1 were investigated preliminarily by protein electrophoresis
on a gradient polyacrylamide gel (SDS-PAGE) and Coomassie blue staining ([Fig. 1]). Concentration was determined from theoretical extinction coefficient using UV
absorption at 280 nm and a standard regression of bovine serum albumin in a Bradford
assay.
Analysis of purity, identity, and integrity of the intermediate preparation (FII)
and the product (Pre-1) were commissioned to the protein core facility of the Medical
University of Innsbruck. Identification of the first six amino acids after N-terminal
Edman degradation confirmed the expected sequence known from databases (e.g., UniProt,
EBI, Hinxton, United Kingdom). From mass spectrometry analysis coupled with high-pressure
liquid chromatography, the main abundant protein (86%) in the FII sample had a measured
mass of 70 kDa with 58% coverage with mature Prothrombin OS = Homo sapiens and the
main abundant protein (94%) in the Pre-1 sample had a measured mass of 48 kDa with
79% coverage with Prethrombin-1 OS = Homo sapiens.
In a prothrombin assay with prothrombin-deficient plasma, the prethrombin-1 obtained
had only approximately 1% of the activity of prothrombin. With an enzymatic complex
of factor Xa and its cofactor Va, one prothrombin molecular equivalent unit of prethrombin-1
generated approximately 200 NIH units of thrombin.
Hemostatic Efficacy of Prethrombin-1 in a Rabbit Nail Clipping Model
A rabbit hemostasis model was developed using 2.5 to 3.5 kg female New Zealand white
rabbits. Anesthesia was induced with Ketamine/Xylazine (40 mg/kg and 8 mg/mL; subcutaneous)
in the stable area. Then, the transfer to the operating room took place. After orotracheal
intubation, the animals were ventilated with 40% oxygen and 2% isoflurane in a volume-controlled
manner. As an analgesic, 0.3 mg/kg Buprenorphine was administered. Arterial blood
pressure monitoring was performed invasively through the ear artery. The animals received
Ringer's solution for fluid substitution (10 mL/kg/h). To simulate a temporary hemophilic
state, a FVIII antibody was intravenously infused at a dose of approximately 20,000
Bethesda units (BU) per kilogram of body weight (BW). The administration of the FVIII
antibody aimed to replicate the development of a FVIII inhibitor status observed in
humans.
Following a minimum 10-minute incubation period, hemophilic status was confirmed by
nail clipping. To this end, the cuticle of the nail was excised using a nail clipping
device, and blood loss was assessed by immersing the clipped cuticle in a preweighed
Falcon tube filled with 37°C 0.9% saline solution. Bleeding was monitored over a 15-minute
period. Gravimetric determination of blood loss was performed by differentially weighing
the tubes. Then, either the test or the reference substance was administered intravenously.
After an additional waiting period of more than 20 minutes, another nail clipping
test was performed as described above. Subsequently, the animals were euthanized through
an overdose of pentobarbital in anesthesia.
Modified Rabbit Wessler Model
A modification of the Wessler model[23]
[24] was employed to conduct jugular vein thrombosis experiments on female New Zealand
white rabbits weighing 2.5 to 3.5 kg. The animals underwent anesthesia as described
above. The contralateral vena jugularis communis was prepared, with all side branches
ligated with 8/0 Prolene or coagulated, including smaller outlets. Subsequently, the
rabbits were induced into a temporary hemophilic state, following the procedure outlined
above.
Ten minutes after the intravenous (i.v.) administration of the test or reference substance,
a segment of the vein was ligated both proximally and distally at a distance of 1.5 cm.
After a 20-minute interval, the ligated vein segment was excised, placed in a sodium
citrate buffer, and dissected open. Macroscopic examination of the luminal vein surface
was conducted. Any observed thrombi were extracted and weighed. The assessment of
thrombi utilized a scale (Wessler score) ranging from 0 to 3: 0 denoted an absence
of thrombus, 1 represented one thrombus or several small thrombi (<2 mg), 2 indicated
one or more non-occluding thrombi, and 3 signified a large, completely occluding thrombus.
Finally, the animals were euthanized by an overdose of pentobarbital in anesthesia.
Hemostatic Efficacy of Prethrombin-1 in a FVIII Inhibitor Mouse Model
Hemostasis experiments were performed in 20 to 30 g female FVB mice. To determine
the hemostatic efficacy of prethrombin-1 in a FVIII inhibitor mouse model, FVB mice
were anesthetized with Ketamine (100 mg/kg)/Xylazine (5 mg/kg) and Dormicum (1 mg/kg)
i.p. and VIII antibody at a dose between 25,000 and 30,000 BU/kg BW was injected into
the tail vein to ensure increased bleeding after tail resection. In the following
10 minutes, the vena femoralis was exposed for injection of the test substance. Fifteen
minutes after administration of the test substance via the vena femoralis and electro-cauterization
of the injection site, the tip of the tail was resected using a scalpel at a distance
of 3 mm. The tail was suspended in a pre-weighed Eppendorf tube filled with 37°C 0.9%
saline solution and bleeding was observed for 30 minutes, at which time the experiment
was stopped. Blood loss was measured gravimetrically by differential weighing.
Hemostatic Efficacy of Prethrombin-1 in a Rivaroxaban Mouse Model
The procedure was the same as the one above except for the use of 2 µg Rivaroxaban/mouse
instead of FVIII antibody. Rivaroxaban was dissolved in DMSO (1,000 µg/mL) and diluted
in 0.9% saline solution.
Results
Rabbit Nail Clipping Study
The rabbit nail clipping model was used to compare the hemostatic efficacy of five
doses of prethrombin-1 (4.7, 16, 47, 157, and 469 nmol/kg) against 2 mL of sodium
citrate as shown below.
Although the spreads in blood loss observed values among both prethrombin-1 and the
control samples appear to be quite high, it must be taken into account that animal
models are complex biological systems and variances are inherent.
Blood loss from rabbit nail clipping experiments is shown in [Table 1] and [Fig. 2]. At a dose of 4.7 nmol prethrombin-1/kg BW, blood loss was above the median blood
loss of untreated animals after anticoagulation by FVIII antibody infusion and comparable
to sodium chloride. At higher doses, blood loss after i.v. treatment with prethrombin-1
remained below the median blood loss of untreated animals. The difference between
sodium chloride and 157 nmol prethrombin-1 was statistically significant at p < 0.05.
Table 1
Blood loss in FVIII-inhibitor rabbits after i.v. treatment with different doses of
prethrombin-1 or sodium chloride
|
Animal #[a]
|
Dosing of prethrombin-1 or physiological saline solution as a control[b]
|
Blood loss[c]
[mg]
|
|
367
|
2 mL NaCl 0.9%
|
106.3
|
|
368
|
2 mL NaCl 0.9%
|
240.5
|
|
378
|
2 mL NaCl 0.9%
|
43.5
|
|
379
|
2 mL NaCl 0.9%
|
100.4
|
|
383
|
2 mL NaCl 0.9%
|
36.7
|
|
369
|
4.7 nmol/kg prethrombin-1
|
189.2
|
|
370
|
4.7 nmol/kg prethrombin-1
|
191.6
|
|
371
|
4.7 nmol/kg prethrombin-1
|
102.3
|
|
335
|
16 nmol/kg prethrombin-1
|
0
|
|
336
|
16 nmol/kg prethrombin-1
|
6
|
|
332
|
16 nmol/kg prethrombin-1
|
23.7
|
|
366
|
16 nmol/kg prethrombin-1
|
45.5
|
|
374
|
16 nmol/kg prethrombin-1
|
0
|
|
328
|
47 nmol/kg prethrombin-1
|
0
|
|
330
|
47 nmol/kg prethrombin-1
|
8.5
|
|
331
|
47 nmol/kg prethrombin-1
|
45.7
|
|
365
|
47 nmol/kg prethrombin-1
|
17.1
|
|
375
|
47 nmol/kg prethrombin-1
|
0
|
|
326
|
157 nmol/kg prethrombin-1
|
0
|
|
327
|
157 nmol/kg prethrombin-1
|
0
|
|
334
|
157 nmol/kg prethrombin-1
|
0
|
|
352
|
157 nmol/kg prethrombin-1
|
0
|
|
354
|
157 nmol/kg prethrombin-1
|
7.3
|
|
358
|
157 nmol/kg prethrombin-1
|
0
|
|
329
|
469 nmol/kg prethrombin-1
|
0
|
|
345
|
469 nmol/kg prethrombin-1
|
4.3
|
|
347
|
469 nmol/kg prethrombin-1
|
0
|
Abbreviation: FVIII, factor VIII.
a Animal consecutive numbers in the approved animal experimentation protocol.
b Dosing of prethrombin-1 or physiological saline solution as control.
c Blood loss in [mg] from gravimetric measurements.
Modified Rabbit Wessler Study
Animals receiving the highest dose of prethrombin-1 from the dose escalation study
above were compared by Wessler score and clot weight with equimolar doses of prothrombin
as well as Beriplex and FEIBA at doses typical for clinical use. The results are shown
in [Table 2] and [Fig. 3] below.
Table 2
Wessler score and clot weight in FVIII-inhibitor rabbits after i.v. treatment with
prethrombin-1, prothrombin, Beriplex, and FEIBA
|
Rabbit #[a]
|
Substance[b]
|
Dose[c]
|
Wessler score[d]
|
Clot weight[e]
|
|
329
|
Prethrombin-1
|
469 nmol/kg
|
0
|
No clot
|
|
345
|
Prethrombin-1
|
469 nmol/kg
|
0
|
No clot
|
|
347
|
Prethrombin-1
|
469 nmol/kg
|
0
|
No clot
|
|
342
|
Prothrombin
|
469 nmol/kg
|
3
|
13.9 mg
|
|
344
|
Prothrombin
|
469 nmol/kg
|
3
|
7.5 mg
|
|
351
|
Prothrombin
|
469 nmol/kg
|
1
|
No clot
|
|
164
|
Beriplex
|
70 U/kg
|
2
|
2.4 mg
|
|
191
|
Beriplex
|
70 U/kg
|
2
|
2.8 mg
|
|
222
|
Beriplex
|
70 U/kg
|
3
|
10.2 mg
|
|
165
|
FEIBA
|
70 U/kg
|
3
|
5.0 mg
|
|
227
|
FEIBA
|
70 U/kg
|
3
|
12.3 mg
|
|
298
|
FEIBA
|
70 U/kg
|
3
|
14.3 mg
|
Abbreviations: FVIII, factor VIII; i.v., intravenous.
a Animal consecutive numbers in the approved animal experimentation protocol.
b Substance applied by i.v. administration.
c Dose in nmol/kg or FIX U/kg.
d Wessler score as described in the Materials and Methods section.
e Wet clot weight in [mg] from gravimetric measurements.
With prethrombin-1, no detectable clots were observed. By comparison, prothrombin
showed Wessler scores of 1 to 3 at the same dose. FEIBA gave Wessler scores of 3 and
Beriplex a score of 2, both with doses of 70 U/kg BW. Analysis of clot weights gave
a significant result for the comparison between prethrombin-1 and FEIBA at p < 0.05.
Hemostatic Efficacy of Prethrombin-1 in a FVIII Inhibitor Mouse Study
The hemostatic efficacy of two doses of prethrombin-1 (224 and 671 nmol/kg) was tested
in tail clipping studies in mice treated with FVIII inhibitor. Sodium chloride was
used as a control to demonstrate the anticoagulant status induced by the FVIII inhibitor.
The results are shown in [Tables 3], [4], [5], and [Fig. 4] below.
Table 3
Blood loss in FVIII-inhibitor mice treated with sodium chloride
|
NaCl 0.9%
|
Ab[c] [µL]
|
BU[d]/kg
|
NaCl[e] [µL]
|
Blood loss[f]
[mg]
|
|
Mouse[a]
|
BW[b] [g]
|
|
MH300
|
23.7
|
40
|
31,472
|
144
|
504.4
|
|
MH301
|
23.8
|
40
|
31,339
|
144
|
603.0
|
|
MH302
|
24.3
|
40
|
30,695
|
144
|
424.8
|
|
MH303
|
24.7
|
40
|
30,198
|
144
|
823.6
|
|
MH304
|
24.6
|
40
|
30,320
|
144
|
740.3
|
Abbreviation: FVIII, factor VIII.
a Animal consecutive numbers in the approved animal experimentation protocol.
b Body weight.
c The antibody solution had a concentration of 13.8 mg/mL with an activity of 3,820
BU/mg. For the mouse studies, the solution was diluted 1:3 prior to application.
d Bethesda units per kg body weight.
e Physiological saline administered by i.v. administration.
f Blood loss in [mg] from gravimetric measurements.
Table 4
Blood loss in FVIII-inhibitor mice after i.v. treatment with 224 nmol/kg prethrombin-1
|
Prethrombin-1 (25.6 nmol/mL)
|
Dose Pre-1[e]
[nmol/kg]
|
Blood loss[f]
[mg]
|
|
Mouse[a]
|
BW[b] [g]
|
Ab[c] [µl]
|
BU[d]/kg
|
Pre-1[e] [µl]
|
|
MH626
|
28.4
|
40
|
24,749
|
48.9
|
224
|
0
|
|
MH630
|
26.1
|
40
|
26,930
|
45.1
|
224
|
3.7
|
|
MH631
|
29.2
|
40
|
24,071
|
50.6
|
224
|
0
|
|
MH646
|
27.7
|
40
|
25,375
|
47.4
|
224
|
0
|
|
MH648
|
25.7
|
40
|
27,349
|
45.4
|
224
|
10.1
|
Abbreviations: FVIII, factor VIII; i.v., intravenous.
a Animal consecutive numbers in the approved animal experimentation protocol.
b Body weight.
c The antibody solution had a concentration of 13.8 mg/mL with an activity of 3,820
BU/mg. For the mouse studies, the solution was diluted 1:3 prior to application.
d Bethesda units per kg body weight.
e Prethrombin-1 administered by i.v. administration in [µL] and nmol/kg body weight.
f Blood loss in [mg] from gravimetric measurements.
Table 5
Blood loss in FVIII-inhibitor mice after i.v. treatment with 671 nmol/kg prethrombin-1
|
Prethrombin-1 (25.6 nmol/mL)
|
Dose Pre-1[e]
[nmol/kg]
|
Blood loss[f]
[mg]
|
|
Mouse[a]
|
BW[b] [g]
|
Ab[c] [µL]
|
BU[d]/kg
|
Pre-1[e] [µL]
|
|
MH633
|
28.1
|
40
|
25,014
|
146.6
|
671
|
51.4
|
|
MH634
|
26.2
|
40
|
26,827
|
136.1
|
671
|
0
|
|
MH643
|
27.6
|
40
|
25,467
|
146.6
|
671
|
0
|
|
MH644
|
26
|
40
|
27,034
|
136.1
|
671
|
0
|
|
MH647
|
22.3
|
35
|
27,579
|
115.2
|
671
|
0
|
|
MH664
|
26
|
40
|
27,034
|
136.1
|
671
|
0
|
Abbreviations: FVIII, factor VIII; i.v., intravenous.
a Animal consecutive numbers in the approved animal experimentation protocol.
b Body weight.
c The antibody solution had a concentration of 13.8 mg/mL with an activity of 3,820
BU/mg. For the mouse studies, the solution was diluted 1:3 prior to application.
d Bethesda units per kg body weight.
e Prethrombin-1 administered by i.v. administration in [µL] and nmol/kg body weight.
f Blood loss in [mg] from gravimetric measurements.
Blood loss in the group receiving sodium chloride amounted to between 424.8 and 823.6 mg.
When prethrombin-1 was administered at a dose of 224 nmol/kg BW, blood loss was greatly
reduced in all animals. In three animals, no blood loss was observed and in two more
animals blood loss was 3.7 and 10.1 mg ([Table 4]). When the prethrombin-1 dose was increased to 671 nmol/kg BW, bleeding compared
with the control group was also reduced: in five animals no blood loss was seen at
all, in one animal it amounted to 51.4 mg ([Table 5]).
The experiments show that with both, a prethrombin-1 dose of 224 nmol/kg BW and a
prethrombin-1 dose of 671 nmol/kg BW, blood loss was reduced compared with the control
group. The result is statistically significant at p < 0.05.
Hemostatic Efficacy of Prethrombin-1 in a Rivaroxaban Mouse Study
Female FVB mice were treated with 2 µg rivaroxaban to induce an anticoagulant state.
After rivaroxaban administration, the tip of the tail was resected, and blood loss
was recorded. [Table 6] shows the data obtained from five mice.
Table 6
Blood loss in mice after rivaroxaban administration
|
Mouse #[a]
|
Blood loss[b] after rivaroxaban administration [mg]
|
|
736
|
35.8
|
|
737
|
44.6
|
|
738
|
28.6
|
|
739
|
45.7
|
|
740
|
38.9
|
a Animal consecutive numbers in the approved animal experimentation protocol.
b Blood loss in [mg] from gravimetric measurements.
Prethrombin-1 at a dose of 224 nmol/kg was used to reverse the anticoagulant effect.
Again, tail-clipping was used to record blood loss after reversal of anticoagulation.
The results are shown in [Tables 6] and [7]. See also [Fig. 5].
Table 7
Blood loss in mice after rivaroxaban reversal with 224 nmol/kg prethrombin-1
|
Mouse #[a]
|
Blood loss[b] after rivaroxaban reversal [mg]
|
|
741
|
0
|
|
742
|
0
|
|
743
|
0
|
|
744
|
0
|
|
745
|
0
|
a Animal consecutive numbers in the approved animal experimentation protocol.
b Blood loss in [mg] from gravimetric measurements.
The results demonstrate that the anticoagulant effect of rivaroxaban is reversed by
prethrombin-1. The result is significant at p < 0.05.
Discussion
It has been known for quite some time that thrombin cleaves fragment 1 from prothrombin,
resulting in the formation of prethrombin-1.[25] However, the potential physiological significance of this phenomenon has remained
unexplored.
One may be tempted to assume that a missing fragment 1 largely inactivates prothrombin.
The fact that prethrombin-1 is indeed able to arrest bleeding in several conditions
such as hemophilia or anticoagulation caused by direct oral anticoagulants is remarkable
and unexpected.
The exact mode of action by which prethrombin-1 promotes hemostasis will require further
exploration. It appears that prethrombin-1 intervenes when thrombin is generated by
activated factor X independently of activated platelet surfaces. This is the case
in the extrinsic phase when such surfaces are generated.
Prothrombin and prethrombin-1 are processed to thrombin with a low but similar kinetic
rate when no activated platelet surfaces are available. One might speculate that in
the extrinsic phase, when only minute amounts of thrombin are generated, its precursor,
the thrombin-zymogen concentration plays an important role in the final concentration
of generated thrombin. At this stage, it does not matter whether the zymogen form
is prothrombin or prethrombin-1.[26]
The thrombin generated at the extrinsic stage is used to activate factors V and VIII
as well as to activate platelets. It would be plausible that increased factor V activation
in the extrinsic phase is the reason for the macroscopic effect to arrest bleeding.
Other modes of action are also conceivable, and it is necessary to study the kinetic
effects in greater detail and with well thought-through approaches.
Elevated concentrations of prothrombin may escalate the risk of blood clots, including
those from deep vein thrombosis and pulmonary embolism. In our current preclinical
studies, we investigated the potential benefit of prethrombin-1 as an alternative
to administering prothrombin or prothrombin-containing PCCs in cases of acute bleeding.
Our findings suggest that prethrombin-1 effectively controls bleeding and exhibits
a lower thrombogenic potential compared with prothrombin and prothrombin-containing
PCCs. This unexpected discrepancy may be attributed to prethrombin-1's inability to
participate in the prothrombinase process, which converts the entire prothrombin in
the proximity of the enzyme complex into thrombin.
We postulate that prethrombin-1 facilitates the separation of processes reliant on
prothrombin as a zymogen in the natural system—specifically, the formation of minute
amounts of thrombin through the extrinsic pathway and the primary massive generation
of thrombin by prothrombinase in the common pathway, where thrombin cleaves fibrinogen
into fibrin. The detachment of fragment 1 and subsequent loss of substrate suitability
for prothrombinase may preserve zymogen for the extrinsic process.
Interestingly, prethrombin-2 does not promote hemostasis. The major difference between
prethrombin-1 and prethrombin-2 is the missing factor Va-binding site.[27] This implies that the factor Va-binding site has significant importance for the
hemostatic efficacy of prethrombin-1.
Our results from preclinical research have shown the potential of prethrombin-1 for
human therapeutic use and suggest that its thrombogenic potential is low, even at
high doses. However, only testing in humans will be able to reveal a therapeutic window
for prethrombin-1 in different bleeding situations. Although prethrombin-1 for our
preclinical studies was sourced from human plasma, it can alternatively be produced
relatively easily by recombinant DNA technology. This might be another major advantage
over activated or nonactivated PCCs.
What is known about this topic?
-
It has been known for quite some time that thrombin cleaves fragment 1 from prothrombin,
resulting in the formation of prethrombin-1.
-
The potential physiological significance of this phenomenon has remained unexplored.
What does this paper add?
-
The thrombin zymogen prethrombin-1 is promoting hemostasis with reduced risk of thrombosis.
-
In the present preclinical studies, we aimed to identify and to explore the potential
of prethrombin-1 as a new treatment option for patients who bleed or are at a risk
of bleeding.
