Keywords
coagulation - thrombosis - contact activation - polyphosphate - factor XII
Classical Anticoagulants
Thromboembolic diseases have remained the major cause of death and disability in the
Western world. Thrombosis is the formation of blood clots that occlude vessels and
terminate tissue perfusion leading to ischemia. Thrombosis occurs in both arterial
and venous vessels resulting in myocardial infarction, ischemic stroke, and deep-vein
thrombosis that eventually results in pulmonary embolism or peripheral vascular disease,
respectively. Cumulatively, thromboembolic diseases represent a major medical burden
causing around one in four deaths worldwide. The incidence of thrombosis is constantly
increasing due to the aging population. Therefore, novel therapeutic strategies for
prophylactic and acute thrombosis treatments are needed.
Occlusive clots are formed by activated platelets trapped by fibrin fibers that are
produced by the blood coagulation system. Antithrombotic therapies that interfere
with pathologic thrombus formation either target platelets or the blood coagulation
cascade. Widely used clinical platelet inhibitors include (1) cyclooxygenase (COX)
inhibitors, (2) purinergic receptor (P2Y) antagonists, and (3) antagonists of glycoprotein
IIb/IIIa (GPIIb/IIIa).[1]
[2] In contrast, anticoagulant drugs target single or combinations of coagulation proteins
(proteases) thereby reducing thrombin and fibrin formation. The majority of anticoagulants
have been known and used in patient care for decades. The polysaccharide heparin is
the most commonly used classical anticoagulant drug. Following injection, heparins
amplify the activity of the endogenous anticoagulant antithrombin that mainly inhibits
active thrombin and factor X (FXa). The anticoagulant activity of heparin unfolds
shortly after administration but depends on patients' antithrombin levels. Another
class of classical anticoagulants is vitamin K antagonists (VKA; e.g., phenprocoumon
and warfarin). These orally administered drugs prevent the posttranslational modification
of coagulation factors II, VII, IX, and X through the blockade of vitamin K–dependent
γ-glutamyl carboxylase, which catalyzes γ-carboxylation of glutamate residues within
Gla-domains which is required for coagulation factor binding to cell surfaces. Defective
local assembly (lateral diffusion) of the thrombin-forming enzymes largely impairs
the enzymatic turnover and fibrin formation. VKAs are potent anticoagulant drugs and
commonly used for prophylaxis and treatment of thrombosis; however, VKAs are indirect
and unspecific inhibitors that interfere with both pathologic thrombus formation and
physiologic hemostasis. Thus, VKA therapy is associated with a significant bleeding
risk and regular monitoring is required to avoid hemorrhagic diathesis due to overdosing.
While heparin and VKA target several coagulation factors, hirudin and recombinant/synthetic
hirudin derivatives inhibit activated thrombin alone. Independent of their mode of
action, classical antithrombotic agents are related with therapy-associated bleeding
events. Direct oral anticoagulants (DOACs) selectively target thrombin or FXa. Due
to their therapeutic range, DOACs do not require regular monitoring and are therefore
predisposed for long-term anticoagulation therapy. Large studies showed that the incidence
for hemorrhagic side effects is lower in antithrombotic treatments with DOACs in comparison
to VKA; yet, an association with a significant bleeding risk still persists.[3] Indeed, in the last years, specific antibodies (Idarucizumab) and recombinant modified
proteins (Andexanet alfa) have been approved to block DOACs for the treatment of drug-associated
hemorrhages. Although antagonizing DOACs in bleeding episodes seems feasible, there
is still a significant number of drug-associated bleeding diathesis that can eventually
become life-threatening, such as for intracerebral hemorrhage.[3] Taken together, currently used anticoagulants provide thrombo-protection, but confer
another risk: all currently available anticoagulants increase the risk of bleeding
that at least partially offsets the therapeutic benefits of reduced thrombotic risk
([Table 1]). Thus, the development of novel and safe anticoagulants for the prevention and
treatment of thrombotic diseases is required.
Table 1
Overview of traditional antithrombotic therapies
Antiplatelet therapy
|
Administration
|
Mechanism
|
Indication
|
Monitoring
|
Reversal
|
Aspirin
|
Oral
|
Inhibition of cyclooxygenase
|
Secondary prophylaxis for myocardial infarction or ischemic stroke
|
None, but consider platelet function analysis to rule out nonresponder
|
Nonspecific, e.g., platelet transfusion
|
Thienopyridine
|
Oral
|
P2Y receptors antagonist
|
Secondary prophylaxis for myocardial infarction or ischemic stroke
|
None, but consider platelet function analysis to rule out nonresponder
|
Nonspecific, e.g., platelet transfusion
|
GPIIb/IIIa antagonists
|
Intravenous
|
Glycoprotein IIb/IIIa receptors antagonist
|
Percutaneous transluminal coronary angioplasty
|
None
|
Nonspecific, e.g., platelet transfusion
|
Anticoagulant therapy
|
Administration
|
Mechanism
|
Indication
|
Monitoring
|
Reversal
|
Warfarin/Phenprocoumon
|
Oral
|
Inhibition of vitamin K–dependent carboxylation of factor II, VII, IX, and X
|
Prevention and treatment of venous and arterial thrombosis/thromboembolism
|
INR
|
Vitamin K, prothrombin complex concentrate
|
Unfractionated Heparins/low-molecular-weight heparins
|
Intravenous subcutaneous
|
Activation of antithrombin
|
Prevention and treatment of venous and arterial thrombosis/thromboembolism
|
aPTT/anti-Xa assays
|
Protamine sulfate
|
Direct oral anticoagulants (direct thrombin or factor Xa inhibitors)
|
Oral
|
Inhibition of thrombin or factor Xa
|
Prevention and treatment of venous and arterial thrombosis/thromboembolism
|
None, but anti-Xa assays can be used to rule out substantial residual effect
|
Idarucizumab or andexanet alfa
|
Abbreviations: aPTT, activated partial thromboplastin time; INR, international normalized
ratio.
The Factor XII–Driven Contact System
The Factor XII–Driven Contact System
Factor XII (FXII) is the zymogen form of the plasma serine protease factor XIIa (FXIIa).
Zymogen FXII is activated to FXIIa by limited proteolysis and proteolytic cleavage
is due to enzymatic activation (“fluid phase activation”) or by its unique property
to undergo autoactivation following binding (“contact”) to negatively charged artificial
or biologic surfaces (“contact activation”). FXII surface-induced contact activation
provides the mechanistic basis for a commonly used diagnostic coagulation assay, for
example, the activated partial thromboplastin time (aPTT) that is used as a clinical
parameter to measure the potential of the global plasma coagulation.[4]
[5] FXII contact activation induced by white clay materials, celite and kaolin or the
phenol antioxidant ellagic acid, is used to test the aPTT, a standard diagnostic coagulation
assay that is measured more than half a billion times annually.[6] On aged erythrocytes, FXII seems to be activated by yet to be identified protease
activities.[7] In blood vessels FXII binds to proteoglycans located on the surface of endothelial
cells and this a locally produced FXIIa initiates the intrinsic pathway of coagulation
via activation of its substrate factor XI (FXI). FXIIa formation also leads to the
liberation of the proinflammatory mediator bradykinin (BK) by activating plasma prekallikrein
to plasma kallikrein (PKa), which drives the cleavage of high-molecular-weight kininogen
(HK).[9]
[10]
[11] Serpin C1 esterase inhibitor (C1INH) is the major plasma inhibitor of FXIIa and
PKa. Deficiency or a dysfunctional C1INH is associated with a BK-mediated life-threatening
inherited swelling disorder, hereditary angioedema (HAE) type I or II, respectively.[12]
[13] The proteases FXIIa, PKa, their endogenous inhibitor C1INH, FXIa, and the BK precursor
HK together form the plasma contact system. Despite its importance for fibrin formation
in clotting assays, FXII had been considered to have no function in vivo and research
in this field was dormant for approximately 60 years. Nevertheless, studies on FXII
experienced a revival following the discovery of thrombo-protection in FXII-deficient
mice.[14]
Novel Concepts in Antithrombotic Therapy
Novel Concepts in Antithrombotic Therapy
Ideally, an antithrombotic drug should have strong antithrombotic efficacy both in
preventing and treating thrombosis. However, in sharp contrast to all currently used
anticoagulants, its antithrombotic activity should not be associated with an increased
bleeding tendency. However, in the classical textbook scheme, thrombosis (coagulation
in pathologic vascular occlusions) and hemostasis (coagulation at the injury site
to terminate bleeding) present the opposite poles of a coagulation balance. Following
this fundamental concept, safe anticoagulants, that protect from pathologic thrombus
formation, but spare physiologic hemostasis and do not increase bleeding, might not
be possible. In 2005, we have originally challenged this concept of the coagulation
balance. Our laboratory has generated the first FXII-deficient (F12−/−) mouse model. These mice are protected from occlusive thrombus formation triggered
by various chemical and mechanical injury types both in arterial and venous vessels.
However, similar to FXII-deficient individuals, the hemostatic potential of F12−/− mice was completely normal and exhibited no signs of excessive bleeding at injury
sites or defective hemostasis.[15] Furthermore, reconstitution of F12−/− mice with human FXII protein fully restored occlusive thrombus formation, suggesting
that FXII could be a target for thrombo-protection that lacks the risk of bleeding.[16] In addition to experimental vascular injury models, impaired thrombus formation
in the absence of FXII has also been shown in atherothrombosis[17] and ischemic stroke.[18]
Targeting Factor XII in Experimental Thrombosis Models
Targeting Factor XII in Experimental Thrombosis Models
The importance of FXII in pathologic thrombosis is not restricted to murine models.
We and others have shown a critical role of the FXIIa-driven intrinsic pathway of
coagulation for experimental thrombosis in large animals such as rats,[19] rabbits,[20] or primates,[21] and recently in humans.[22] To address the question which preclinical models have a more predictive value for
human therapy, we had screened for antibodies against FXIIa using phage display and
demonstrated that recombinant fully humanized antibody 3F7 inhibits FXIIa enzymatic
activity, interferes with FXII-mediated coagulation, and blocks experimental thrombosis
in mice and rabbits ([Fig. 1]). In an adapted extracorporeal membrane oxygenation (ECMO) cardiopulmonary bypass
system used for infant therapy, the clinical applicability of 3F7 was tested in rabbits.
3F7 provided thrombo-protection as efficiently as heparin; however, unlike heparin,
3F7 treatment did not impair the hemostatic capacity of animals and did not increase
the bleeding risk. Our data provide the first clinically relevant antithrombotic strategy
that is not complicated by excess bleeding.
Fig. 1
Regulation of coagulation pathways by antithrombotic agents in hemostasis and thrombosis. Vitamin K antagonists (VKA) inhibit synthesis of functional coagulation factors VII
(FVII), IX (FIX), X (FX), and prothrombin, thereby limiting thrombin generation through
the activated complexes tissue factor (TF):FVIIa and factor VIIIa (FVIIIa):IXa. Heparin
potentiates inhibition of FXa and thrombin by binding to antithrombin (AT) III. Direct
oral anticoagulants (DOACs) antagonize FXa and thrombin, while hirudin blocks specifically
thrombin. Thrombin-mediated fibrinogen activation to form insoluble fibrin fibers
by the extrinsic coagulation pathway (blue arrows) in hemostasis is exceeded by initiation
of the intrinsic coagulation pathway (red arrows). Negatively charged surfaces activate
factor XII (FXII) to FXIIa, which is enhanced by a positive feedback loop with FXIIa-mediated
activation of prekallikrein (PK) to plasma kallikrein (PKa). FXIIa and PKa are inhibited
by C1 esterase inhibitor (C1INH) and peptide-based inhibitor (D-Pro-Phe-Arg chloromethyl
ketone, PCK). Recombinant Ixodes ricinus inhibitor (Ir-CPI) blocks FXIIa and its activated substrate factor XI (FXI) in addition
to PKa. Fully humanized monoclonal antibody (mAB 3F7) and infestin-4 from Triatoma infestans fused to recombinant human albumin (rHA-infestin-4) specifically target FXIIa and
inhibit the FXIIa-mediated coagulation cascade.
In addition to 3F7, more FXII inhibitors have been developed. rHA-Infestin-4 is a
recombinant inhibitor of FXIIa that is composed of the insect protein infestin-4 that
has been fused to human recombinant albumin to increase its half-life and solubility.
rHA-Infestin-4 therapy protected from experimental arterial and venous thrombosis
and associated thromboembolic diseases such as pulmonary embolism, atherothrombosis,
and ischemic strokes.[23] Similar to 3F7 and despite its potent anticoagulant activities, physiological hemostasis
remained unaffected in rHA-infestin-4-treated animals.[23]
Inhibition of FXIIa with rHA-infestin-4 protected mice from FeCl3-induced thrombosis in veins and arteries.[24] In a model of silent brain ischemia in mice induced by intra-arterial injections
of microbeads or clot material, rHA-infestin-4 also reduced cerebral microinfarcts
but was inactive in reducing ischemia-induced inflammatory reactions.[25] Consistent to the infestin-4 data, mice treated with Ir-CPI, an inhibitor of FXIIa,
activated FXI and PK, were protected from lethal pulmonary embolism and stasis-induced
venous thrombosis.[26] Similarly, FXIIa-driven fibrin formation contributed to stasis-associated vena cava thrombosis induced by restriction of blood flow.[27] H-D-Pro-Phe-Arg-chloromethyl ketone (PCK), that inhibits FXIIa and PK, interfered
with lethal pulmonary embolism in murine models.[28]
Consistent with the genetic and pharmacologic models, antisense oligonucleotide (ASO)-mediated
“knockdown” of F12 gene expression attenuated experimentally induced arterial and
venous thrombosis.[29] However, in contrast to “classical” anticoagulants, a significant reduction in FXII
was required to achieve thrombo-protection. Thrombosis is normal in FXII heterozygous
mice (F12+/−) having 50% of normal circulating FXII and FXII needs to be reduced to less than
25% to provide safe anticoagulant activities. Supporting the key role of the FXII-driven
contact system in thrombosis, mice with deficiencies in PK[30]
[31] or HK[32] are also protected from thrombosis. More comprehensive overviews about targeting
FXII and contact system proteins in various experimental thrombosis models have been
presented in recent reviews.[33]
Factor XI in Thrombosis
Similar to F12−/− mice, animals deficient in its downstream substrate of the intrinsic pathway, coagulation
factor XI (FXI), were protected from FeCl3-triggered arterial thrombosis.[34] However, in contrast to FXII deficiency, low FXI activity is associated with bleeding
termed “hemophilia C.” Bleeding in FXI deficiency is mild; however, it is rather unpredictable,
as it is not linked to circulating FXI antigen levels and biomarkers to assess vascular
disease and bleeding risk in these patients are currently lacking.[35] Despite the striking phenotype in FXI-deficient (F11−/−) mice, the discovery of impaired thrombus formation in these animals did not raise
much attention most probably because the underlying mechanism was misinterpreted.
Originally it was believed that impaired thrombosis in F11−/− mice resulted from defective FXI activation by thrombin that in turn was produced
by tissue factor (“feedback activation loop”).[36] However, later studies challenged this hypothesis. Occlusive thrombus formation
was similarly defective in F12−/− and F11−/− mice, as well as in animals with combined deficiency in both clotting factors (F12−/−/F11−/−). Similar protection from thrombosis argues against a role of FXI “feedback activation”
in thrombotic reactions in vivo but indicates that FXI is mainly activated by FXIIa
in pathological platelet-mediated thrombosis. Genetic and pharmacologic targeting
of FXI activity has been shown to provide thrombo-protection in an array of murine
and large animal models and clinical studies on FXI inhibitors in humans are well
advanced. In contrast to FXII that triggers thrombo-inflammatory reactions, the role
of FXI is restricted to the coagulation system. Thus, targeting FXII provides additional
beneficial effects with clinical implications beyond thrombosis.
Targeting Factor XII in Allergy and Inflammation
Targeting Factor XII in Allergy and Inflammation
In addition to its critical function for thrombosis, the FXII-driven contact system
contributes to allergic disease states.[37] Upon allergen-challenge mast cells release, the negatively charged polysaccharide
heparin that acts as contact activator efficiently activates FXII. Mast cell heparin-driven
contact system activation culminates in BK formation that increases vascular permeability
through cytoskeleton rearrangements in endothelial cells.[38]
[39] FXII contact activation by mast cell heparin has been shown to play a critical role
for anaphylactic and allergic diseases both in genetically modified mouse models and
patients. Thus, targeting BK generation and signaling provides a novel and alternative
treatment strategy for anaphylactic attacks.
Recent findings revealed a role of FXII for SARS-CoV-2-associated lung pathology,[40] suggesting that targeting FXII might help reduce severity and thromboembolic complications
in COVID-19 patients.[41] In contrast to FXII, HK has recently been shown to have a role in experimental liver
inflammation.[42] Together the data reveal differential roles of contact system proteins in distinct
inflammatory disease states.[43]
Notably, plasma contact system deficiencies delay the aPTT, demonstrating the dichotomous
role of FXIIa for coagulation and inflammation. Patients with deficiencies in functional
C1INH develop the rare life-threatening swelling disorder, HAE type I and type II
(MIM #106100). C1INH deficiency aggravates activated mast cell–triggered edema with
implications for swelling attacks in HAE. Besides the two classic C1INH-dependent
HAE types, a third variant, HAE type III, exists in patients who have completely normal
C1INH but similarly suffer from edema attacks. HAE type III (HAE with normal C1INH;
FXII-HAE) is associated with missense mutations in the FXII gene resulting in a single
point mutation (Thr309Lys or Thr309Arg) that leads to increased FXII enzymatic activity
(“gain-of-function”). The underlying mechanism for excessive contact activation remained
unknown until it was recently shown that a HAE type III–associated mutant FXII is
defective in a mucin-type Thr309-linked glycosylation. Loss of glycosylation led to
increased contact-mediated autoactivation of zymogen FXII, resulting in excessive
formation of BK. Intravital microscopy imaging revealed that a humanized HAE type
III mouse model with inducible liver-specific expression of Thr309Lys-mutated FXII
exhibited increased contact-driven microvascular leakage. The FXIIa-neutralizing antibody
3F7 and clinically used PK inhibitors blunted edema in these mice and abolished BK
generation in HAE type III patient plasma, suggesting that FXIIa inhibition provides
a potential therapeutic strategy to interfere with excessive vascular leakage in HAE
and potentially other disease states.[44]
Moreover, application of rHA-infestin-4 was also protective in mouse models of autoimmune
encephalomyelitis[45] and anaphylactic shock,[46] illustrating that targeting FXIIa has anti -inflammatory potential in addition to
providing antithrombotic protection. Together, the data indicate that FXIIa inhibitors
have broader clinical applications exceeding their use as antithrombotic drugs.
Polyphosphate: The Natural Factor XII Contact Activator
Polyphosphate: The Natural Factor XII Contact Activator
The in vivo activation of FXII in the initiating steps of pathologic thrombus formation
has been an enigma for many years.[4] More specific, the natural surface that induces FXII contact activation in vivo
had been unknown.[47] But for decades it has been known that activated platelets induce plasma clotting
in a FXII-dependent manner,[48] indicating that platelets expose FXII-activating agents. Following the discovery
that the linear platelet-derived polymer polyphosphate (polyP) serves as the FXII-activating
surface,[28] research elucidated the role of polyP for FXIIa-mediated thrombosis and inflammation[5]
[49] ([Fig. 2]). Inorganic polyP is a polyanion consisting of up to several hundred phosphates
(Pi) linked by energy-rich phosphoanhydride bonds. The chain length of synthetic polyP
determines its solubility and FXII activation capacity in plasma,[50] while natural platelet polyP forms insoluble Ca2+ ion-rich nanoparticles independently of the chain length of the polyP molecule that
are retained on the surfaces of procoagulant platelets.[51] In addition to FXII activation in vivo, polyP has also been reported to modulate
several other coagulation reactions in vitro; however, a potential role of these pathways
remains to be shown.[50] Similar to targeting FXIIa, interference with polyP has emerged as a novel strategy
for safe inhibition of thrombosis but sparing hemostasis. To interfere with polyP
activities, we developed in 2016 the first specific polyP inhibitors, based on recombinant
Escherichia coli exopolyphosphatase (PPX) mutants.[52] Supporting a specific function of polyP in thrombosis, targeting polyP provided
thrombo-protection in vivo. Moreover, polyP was visualized as insoluble nanoparticles
on the surface of procoagulant platelets in vivo.[51] The fact that polyPs are insoluble calcium-rich particles that are deposited on
activated platelet surfaces explains that procoagulant platelets trigger thrombosis
in a FXIIa-dependent manner. In addition to activating FXII, polyP binding to platelet
factor 4 (PF4) generates neoantigens that lead to vaccine-induced immune thrombotic
thrombocytopenia (VITT).[53] Clearly, polyP is a new and exciting player in thrombosis that awaits clinical studies.
Indeed, recently a fluorescence-activated cell-sorting (FACS) assay was established
to measure polyP in patients using the polyP-binding domain of PPX (PPX_Δ12) as a
polyP probe.[52] Furthermore, polyP levels can be quantified by chromogenic malachite green assays
that measure the formation of phosphate complexes in vitro. Recently, the first regulator
for polyP in mammalian systems was identified. The phosphate transporter XPR1 controls
polyP in platelets in vivo, and interference with XPR1 modulates thrombosis but not
hemostasis both in mouse models and human samples.[54] Furthermore, the structural basis for polyP-mediated FXII activation that provides
the start of polyP-triggered thrombosis and proinflammatory reactions has been elucidated.[55] A specific segment of the FXII C-terminal proline-rich region (termed PR-III) of
the heavy chain is required for zymogen contact activation. A FXII variant lacking
this sequence can be activated by PKa; however, it is completely defective in undergoing
contact activation. The identification of the FXII contact activation side opens perspectives
for new coagulation assays. Antibodies directed to the PR-III region induce FXII zymogen
“contact” activation fully in solution and in a stoichiometric controllable manner
allowing for improved antibody-activated PTT (aaPTT) assays. In contrast to standard
aPTT systems, aaPTT assays allow the very precise measurement of FXI levels in the
therapeutic range of an emerging target for safe antithrombotic therapy.
Fig. 2
Regulation of procoagulant platelet polyphosphate. The phosphate exporter Xenotropic and polytropic retrovirus receptor 1 (XPR1) reduces
free phosphate (Pi) concentration in platelets, which decreases the level of procoagulant polyphosphate
(polyP) with implications for factor XII (FXII) contact activation at the membrane
of activated platelets.
Key Points
-
Current anticoagulant drugs reduce risk of thrombosis but concomitantly increase bleeding.
-
Genetic and pharmacologic inhibition of contact system factors XII, XI, PK, and HK
provides potent thrombo-protection without affecting physiological hemostasis.
-
Neutralization of factor XII confers additional anti-inflammatory activities.