TREATMENT

FIELD OF THE INVENTION:

The invention relates to the field of hematology; which includes the study, diagnosis, treatment, and prevention of diseases related to blood.

In particular, the invention relates to the diagnosis, treatment, and prevention of diseases in a subject under anticoagulant therapy.

BACKGROUND OF THE INVENTION:

There is a general need for novel compounds and methods for preventing or reducing bleeding in a subject undergoing anticoagulant therapy.

Following decades of vitamin K antagonists (VKA) prescription as unique oral anticoagulants bridging therapies with unfractionated heparin (UFH) or low molecular weight heparins (LMWH), Direct Oral AntiCoagulants (DOAC) have been developed targeting thrombin or coagulation factor Xa (FXa) overcoming many of the limitations of conventional anticoagulants.

VKA interfere with the cyclic interconversion of vitamin K epoxide resulting in non- functional vitamin K dependent coagulation factors II, VII, IX and X (Ageno et al. 2012; Oral anticoagulant therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest; 141 :e44S-88S). UFH and LMWH catalyze thrombin and/or FXa inhibition by antithrombin. These are administered via parenteral injection. DOAC are small direct inhibitors neutralizing their target either free in solution or bound to prothrombinase complex or to fibrin. In a non- limitative manner, three anti-FXa DOAC (rivaroxaban, apixaban, edoxaban) and a single anti-thrombin DOAC (dabigatran) are readily licensed in various countries for stroke prevention in atrial fibrillation, and treatment and/or prevention of venous thromboembolic events. Other DOAC are currently in development or clinical trials. Ongoing studies are evaluating DOAC potential in heart failure, coronary or peripheral artery disease as well as embolic stroke of unknown origin. Overall millions of patients are candidates for DOAC medication during their lifetime. As prescriptions of DOAC expand, experience now underlines that rare bleeding complication may occur and could be devastating, as with any other anticoagulant drug; thus requiring efficient monitoring of patients under anticoagulant therapy.

Indeed, treated patients may experience clinical scenarios, such as massive haemorrhage, trauma, thrombotic events, overdose, hypo- or hyper-coagulation states, stroke requiring thrombolysis or urgent surgery. These scenarios require evaluating without delay whether an anticoagulant (i.e. factor Xa inhibitor, heparin and derivatives thereof, thrombin inhibitors) exceedingly threatens the patient.

Currently these are real challenges to both physicians and laboratory staff. Specific quantification of the anti-coagulants (such as DOAC) is feasible, but can be out of reach in emergency situation. Furthermore routine tests available fail to accurately detect all anticoagulant drugs, in particular DOAC, especially apixaban.

A plurality of hemostasis tests have been reported in the Art. Yet, those are not fully satisfactory. For instance, those tests may not be rapid or accurate enough. Also, they may not necessarily be all performed in whole blood (WB), and thus require prior plasma separation; which may be cost-effective and/or be associated with a significant delay which is highly prejudicial to the subject in need of emergency treatment.

In particular, widely available hemostasis tests, namely prothrombin time (PT) and activated partial thromboplastin time (aPTT), have variable sensitivity to DOAC with the result that they may fail to detect even therapeutic amounts of drug.

When DOAC titration becomes necessary, target-specific assay (anti-Xa activity and diluted thrombin time or ecarin clotting time for anti-thrombin activity), or alternatively mass spectrometry, are recommended.

Laboratory titration only solves part of the problem because of the turnaround time required, unacceptable in the context of emergency. Furthermore, titration is neither widely available nor routinely performed or standardized. Other high sensitivity tests allow detection of anticoagulants yet are neither routinely available nor standardized. For instance dilute (dPT) or modified (mPT) prothrombin times, dilute Fiix-PT (dFiix-PT) as well as thrombin generation assay (TGA) or clot waveform assay may detect as little as 30 ng/ml DOAC. These tests are performed in specialised haemo stasis laboratories and triggered with low amounts of tissue factor (TF) probably accounting for their sensitivity. Still, these tests would require an unacceptable turnaround time in the context of emergency and, while indeed accurately detecting anticoagulant in plasma, they fail to rule out possible concomitant coagulopathy.

A common bottleneck of standard as well as specialized coagulation tests is the unavoidable time required for sample transportation and platelet-poor plasma (PPP) preparation. One of the few hemostasis tests not requiring prior plasma separation, thus performed in whole blood (WB) and without delay, relies on viscoelasticity measurement of the forming clot. Specifically thromboelastometry and thromboelastography are point of care devices providing results within minutes. To date however, available reactants, such as Extern™ of Ro tern™, have failed to detect accurately, or discriminate within a sample, all types of anti-coagulants (i.e. DOAC). Other potential devices used for WB evaluation include the chronometric approaches based methods such as (for instance) the Stago STart 4 Hemostasis Analyser or the CoaguChek INRange system measuring the International Normalized Ratio (INR). As with the Rotem™, these devices failed to detect accurately, or discriminate within anticoagulated samples (i.e. non-VKA anticoagulants such as DOAC) with the available reactants.

Another bottleneck is that, on the whole, current haemostasis assays affected by anticoagulant are not immune from false positive (as well as negative) results. For instance DOAC interfere in haemostasis tests designed for lupus anticoagulants or thrombophilia testing.

Thus, there is a need to develop assays that rapidly distinguish if a defect in hemostasis originates from an anticoagulant treatment and/or is associated to a coagulopathy resulting for instance from hepatic failure, factor consumption, hemodilution, active cancer or acquired hemophilia.

In particular, there is a need to develop assays permitting rapid and reliable detection of a clotting anomaly and simultaneous identification of its origin.

Thus, the problem is here three-fold:

- the first is to rapidly and reliably evaluate the patient hemostasis;

- the second is to identify its origin if an anomaly is detected;

- the third is to make a decision knowing that antidotes to anti-coagulants such as anti-FXa DOAC are not readily available or even lacking.

This is specially threatening with anti-FXa DOAC due to the lack of proven available reversal agent or antidote (Greinacher et al. 2015; Reversal of anticoagulants: an overview of current developments. Thromb Haemost; 115:931-42). One example of antidote is andexanet alpha, which is a modified catalytically inactive factor Xa which lacks the membrane-binding γ-carboxyglutamic acid domain (Gla domain) of native FXa, but which retains the ability to bind FXa inhibitors.

Yet, an antidote compatible with differential diagnosis should comply with at least two requirements:

- It should neutralize the targeted drugs;

- It should not interfere with coagulation (except for neutralization of the targeted drug).

Specifically, an optimal antidote must be strictly neutral neither pro- nor anticoagulant meaning that in a normal situation (without drug) coagulation in the presence of the antidote should match exactly coagulation in its absence. On the whole antidote must sequester, neutralize and even eliminate (preferably any) anticoagulant drugs while not interfering with normal coagulation. Antibodies are typically ideal antidote but they are specific for a given drug as for instance idarucizumab which is specific for dabigatran.

Accordingly, WO 2009/042962 teaches a FXa inhibitor antidote which may consist of a modified factor Xa protein, having reduced or lacking procoagulant activity. It also suggests producing antibodies binding to said antidote. Yet, antibodies directed against anti-FXa DOAC have not been described to date. Furthermore a universal antidote would have to be constituted of a plurality of antibodies, each one targeting a single drug.

Thus, there is still a need for novel antidotes allowing to assess the effect of anticoagulant drugs on hemostasis through efficient and reliable assays, as discussed above. In particular, there is a need for novel antidotes allowing differential diagnosis.

The invention has for purpose to meet this need.

SUMMARY OF THE INVENTION:

This invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising a complex between a Factor Xa (FXa) or a modified catalytically active FXa, and a-2 macroglobulin (a2M); b) determining clot formation in the sample from step a) to obtain a test value; and c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject. This invention relates to an isolated complex between a catalytically active Gla- Domain deficient Factor Xa (Gla-deficient FXa) and a-2 macroglobulin (a2M).

This invention relates to a pharmaceutical composition comprising:

(i) a complex between a catalytically active Gla-Domain deficient Factor Xa (Gla- deficient FXa) and a-2 macroglobulin (a2M); and

(ii) a pharmaceutically acceptable carrier.

This invention relates to a complex between Factor Xa (FXa) or a modified catalytically active FXa, and a-2 macroglobulin (a2M); for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; and in particular for discriminating a defect in haemo stasis due to a non- vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy and/or a VKA therapy in said subject.

This invention relates to a complex between Factor Xa (FXa) or a modified catalytically active FXa, and a-2 macroglobulin (a2M); for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

This invention relates to a kit comprising:

a) a first container containing an isolated catalytically active Gla-domain deficient Factor Xa (Gla-deficient FXa); and

b) a second container containing an isolated a-2 macroglobulin.

This invention relates to a kit comprising:

a) a first container containing an isolated Factor Xa or modified catalytically active Factor Xa ; and

b) a second container containing an isolated a-2 macroglobulin;

for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; and in particular for discriminating a defect in haemostasis due to a non- vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy or a VKA therapy in said subject.

This invention relates to a kit comprising:

a) a first container containing an isolated Factor Xa or modified catalytically active Factor Xa ; and

b) a second container containing an isolated a-2 macroglobulin; for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

This invention relates to a method for preparing a complex between a catalytically active Gla-domain deficient Factor Xa and a-2 macroglobulin (a2M) comprising the steps of:

a) providing an isolated catalytically active Gla-domain deficient Factor Xa (Gla- deficient fXa);

b) providing an isolated a2M;

c) bringing into contact the said Gla-domain deficient FXa and the said a2M in a reaction medium in a manner suitable for the Gla-deficient FXa to bind to a2M; wherein said a2M is present in the reaction medium in a molar excess of at least 1 :2, and preferably in a [Gla-deficient FXa : a2M] stochiometric molar ratio of about 1 :4;

cl) optionally incubating the reaction medium for at least 10 minutes;

c2) optionally adding a calcium chelating agent to the reaction medium;

d) recovering the complex between a Gla-domain deficient FXa and a2M from the reaction medium obtained at steps c), cl) or c2). In particular, the invention is defined by claims.

DETAILED DESCRIPTION OF THE INVENTION:

The inventors were of the opinion that a differential diagnosis can be achieved by comparing the results of a point of care test sensitive to anticoagulants with results obtained on the same sample containing an antidote/neutralizing reagent. The inventors were also of the opinion that an antidote should neutralise the drug while being strictly neutral with regard to haemostasis mechanisms not involving the targeted drugs.

Accordingly, the inventors have sought for a novel antidote allowing to provide a differential diagnosis in a rapid and sensitive manner, including in Whole Blood (WB) samples, or a fraction thereof. It has now been found that a complex between a Factor Xa or a modified catalytically active Factor Xa (i.e. a catalytically active Gla-domain deficient FXa or even a catalytically active Gla-domainless FXa), and a-2 macroglobulin (a2M) could be efficient, as an antidote, both for therapeutic and diagnostic purposes.

Accordingly, the inventors have found that said complex is efficient for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; in particular for discriminating a defect in hemostasis due to a non-vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy and/or a VKA therapy in said subject.

Also, the inventors have found that said complex is efficient for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject. As used herein, "preventing" encompasses "reducing the likelihood of occurrence".

Without wishing to be bound by any particular theory, the inventors are of the opinion that, within a complex between a ₂ M and FXa or a modified catalytically active FXa (i.e. a catalytically active Gla-domain deficient FXa or a catalytically active Gla- domainless FXa), the active site of FXa is accessible to small molecules such as DOAC while protected from antithrombin or tissue factor pathway inhibitor (TFPI) due to steric hindrance. Within the clotting cascade, FXa mainly interacts with prothrombin (within the prothrombinase complex), antithrombin, and TFPI. These functions involve macromolecule interactions which thus are fully impeded in the FXa-a ₂ M complex.

Also without wishing to be bound by any particular theory, the inventors are of the opinion that any pro-coagulant activity of FXa or a modified catalytically active FXa is abolished when in complex with a ₂ M, whether the said FXa or modified FXa still has a functional Gla-domain or not. Accordingly, even though the Examples provided herein focus on the properties of a GDFXa- a ₂ M as an antidote, those results are also applicable to complexes comprising a FXa or modified catalytically active FXa that is not a GDFXa construct.

Also, as shown in Example 6, the inventors have found that the complex between FXa or a modified catalytically active FXa (i.e. a catalytically active Gla-domain deficient FXa or even a catalytically active Gla-domainless FXa) and a ₂ M was not only able to neutralise anti-FXa DOAC, but could also neutralise other anticoagulant treatments, such as anti-thrombin DOAC and heparin or its derivatives. Thus, determining clot formation (i.e. using Rotem™ or clot waveform assays) in the presence of a complex between FXa or a modified FXa and a ₂ M, permits a differential diagnosis between coagulopathies and the impact of anticoagulant therapy (i.e. anti-FXa DOAC, anti-thrombin DOAC, UFH, or LMWH). In emergency this complex has in addition the potential of being a last resort almost universal in vivo anticoagulants antidote. As used herein, the term "antidote" (i.e. "antidote to a factor Xa inhibitor") refers to molecules, such as factor Xa (FXa) or a modified catalytically active FXa, which can substantially neutralize or reverse the coagulation inhibitory activity of a FXa inhibitor. Examples of the antidotes of this invention are FXa derivatives with normal or reduced phospholipid membrane binding, such as des-Gla FXa or Gla-deficient FXa, and FXa derivatives retaining a normal or substantially normal catalytic activity.

As used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. For example, the term "a pharmaceutically acceptable carrier" includes a plurality of pharmaceutically acceptable carriers, including mixtures thereof. As used herein, « a plurality of » may include « two » or « two or more ».

As used herein, « comprising}} may include « consisting of ».

A "subject" of diagnosis or treatment is a cell or a mammal, including a human. Non-human animals subject to diagnosis or treatment may include rodents, such as rats^ and mice, canine, such as dogs, leporids, such as rabbits, livestock, sport animals, and pets.

As used herein, a « blood sample » may refer either to Whole Blood (WB) samples, or a fraction thereof, which may thus consist of any sample obtainable or obtained through blood plasma fractionation for which clot formation can be determined; which may thus also include blood Plasma, Platelet-Rich Plasma and Platelet-Poor Plasma (PPP).

"Anticoagulant agents" or " anticoagulants" may consist of "Non-VKA Anticoagulant agents" or "Vitamin K antagonists (VKA) ". "Non-VKA Anticoagulant agents" or "Non-VKA anticoagulants" are defined herein as agents that inhibit blood clot formation, but which are not Vitamin K antagonists (VKA). Examples of such anticoagulant agents include, but are not limited to, specific inhibitors of thrombin, factor IXa, FXa, factor XIa, factor Xlla or factor Vila, heparin and derivatives, and anti-tissue factor antibodies. Non-VKA (Vitamin K Antagonists) anticoagulants are particularly considered; which includes non-VKA oral anticoagulants (NOAC) such as direct-oral anticoagulants (DOAC). Examples of specific inhibitors of thrombin include hirudin, bivalirudin (Angiomax®), argatroban, and lepirudin (Refludan®). Examples of heparin and derivatives include unfractionated heparin (UFH), low molecular weight heparin (LMWH), such as enoxaparin (Lovenox®), dalteparin (Fragmin®), tinzaparin (Innohep®), nadroparine (Fraxiparine® or Fraxodi®); and synthetic pentasaccharide, such as fondaparinux (Arixtra®). Examples of DOAC anticoagulants include rivaroxaban (Xarelto®), apixaban (Eliquis®), edoxaban (Lixiana®), and dabigatran (Pradaxa®). In one preferred embodiment, the anticoagulant is an inhibitor of FXa. "Vitamin K antagonists (VKA) " include, in a non- limitative manner, warfarin (Coumadin®), phenocoumarol, acenocoumarol (Sintrom®), clorindione, dicoumarol, diphenadione, ethyl biscoumacetate, phenprocoumon, phenindione, fluindione (Previscan®) and tioclomarol.

"Anticoagulant therapy" refers to a therapeutic regime that is administered to a patient to prevent undesired blood clots or thrombosis. An anticoagulant therapy comprises administering one or a combination of two or more anticoagulant agents (i.e. non- VKA anticoagulants) or other agents at a dosage and schedule suitable for treating or preventing the undesired blood clots or thrombosis in the patient.

As used herein, a « coagulopathy » refers to any condition in which the blood's ability to coagulate (form clots) is impaired. Examples of coagulopathies may include genetic disorders, such as hemophilia and Von Willebrand's disease. They may also be caused by the reduced levels or absence of functional blood-clotting proteins. They may also occur as a result of dysfunction or reduced levels of platelets. Accordingly, coagulopathies, as used herein, may encompass coagulopathies resulting from hepatic or renal failure, factor consumption, haemodilution, active cancer or acquired hemophilia. They may also be induced by the consumption of drugs, which may thus include anticoagulants.

As used herein, "clot formation" refers broadly to thrombosis, which is associated to the clotting cascade triggered by exposure of tissue factor following injury. Alternatively activation of coagulation factor XII, historically known as Hageman factor. Activation of factor XII occurs when it comes into contact with negatively charged surfaces specifically in vitro and in several pathological conditions such as bacterial infection in vivo.

As used herein, "determining clot formation" refers to the determination of any variable associated with clot formation or any sub-phase thereof; which includes the time from initiation of the reaction to the point of clot formation; which includes pre- coagulation, coagulation and post-coagulation phases. Said determination can be done by any method known in the Art; which includes both conventional clotting tests such as prothrombin time (PT), thromboplastin time (PTT) and activated partial thromboplastin time (aPTT), but also non-conventional clotting tests. In order to remove the inherent variability associated with the subjective endpoint determinations of manual techniques, instrumentation has been developed to measure clot time, based on (1) electromechanical properties, (2) clot viscoelasticity, (3) light scattering, (4) fibrin adhesion, (5) impedance, and (6) chromogenic or fluorogenic signals. For light scattering and chromogenic methods, data is gathered that represents the transmission of light through the specimen as a function of time (an optical time-dependent measurement profile). One type of analysis may consist of the determination of a clot waveform (or clot waveform assay), which is defined by a change in light transmittance that occurs during the process of clot formation. Accordingly, clot formation may be categorized by clot waveform assay in at least three parts: the pre- coagulation, coagulation and post-coagulation phases. Pre-coagulation is described as the first segment of the trace, from the beginning of light transmittance (the "signal") to the onset of coagulation. After the onset of coagulation, light transmittance or absorbance tends to stabilize and is characterized by a linear segment. If fibrinolysis is enhanced (i.e. due to acquired or congenital coagulopathies), light transmittance (the "signal") may change in the coagulation as well as post-coagulation phases.

All numerical designations, e.g., pH, temperature, time, concentration, and molecular weight, including ranges, are approximations which are varied ( + ) or ( - ) by increments of 0.1. It is to be understood, although not always explicitly stated that all numerical designations are preceded by the term "about". It also is to be understood, although not always explicitly stated, that the reagents described herein are merely exemplary and that equivalents of such are known in the art.

"Factor Xa" or "FXa" or "FXa protein" refers to a serine protease in the blood coagulation pathway, which is produced from its inactive precursor factor X (FX). FXa is activated by either factor IXa with its cofactor, factor Villa, in a complex known as intrinsic Xase, or by factor Vila with its cofactor, tissue factor, in a complex known as extrinsic Xase. FXa forms with its cofactor, factor Va, a complex known as prothrombinase and is the active component in the prothrombinase complex that catalyzes the conversion of prothrombin to thrombin. Thrombin is the enzyme that catalyzes the conversion of fibrinogen to fibrin, which ultimately leads to blood clot formation. Unless specified otherwise, ' Xa" refers to a polypeptide corresponding to a naturally occuring, or wild-type, FXa, in an unmodified and catalytically active form. For reference, the nucleotide sequence coding human factor X ("FX") can be found in GenBank, under accession number M57285 (Human coagulation factor X (F10) mRNA, complete cds). The domain structure of mature FX is also described in Venkateswarlu, D. et al, Biophysical Journal, 2002, 82: 1190-1206. It is composed of a light chain (SEQ ID N°l) linked to a heavy chain (SEQ ID N°2) by a disulfide bond. FX activation into FXa follows release of the first 52 residues (SEQ ID N°3) from SEQ ID N°2. FXa is thus composed of a light chain (SEQ ID N°l) linked by a disulfide bond to a shorter heavy chain (SEQ ID N°4). The nucleotide sequence of the gene encoding the zymogen form of the protein can be found in the Genbank under ID 2159. The amino acid sequence of the human precursor can be found in the UniProtKB/Swiss-Prot under ID P00742. The first 44 amino acids (residues 1-44 of SEQ ID N°l of the light chain are called the Gla-domain because it contains post-translationally modified γ-carboxyglutamic acid residues (GLA). It also contains a short aromatic stack sequence (residues 40-44 of SEQ ID N°l . GLA-domain proteins comprise an N-terminal portion called propeptide which is recognized by a vitamin K-dependent carboxylase. Chymotrypsin digestion selectively removes the 1-44 residues resulting in Gla-domainlessFXa (SEQ ID N°5). The serine protease catalytic site of FXa locates within the heavy chain (SEQ ID N°4). The heavy chain of FXa is highly homologous to other serine proteases such as thrombin, trypsin, and activated protein C.

"Normal FXa" or "wild-type FXa" refers to the activated form of FX naturally present in plasma or being isolated in its original, unmodified form, which processes the biological activity of activating prothrombin therefore promoting formation of blood clot. The term includes naturally occurring polypeptides isolated from blood as well as recombinantly produced FXa. An "active FXa " refers to FXa having the biological activity of activating prothrombin. A "catalytically active FXa" further refers to FXa having a functional catalytic domain, which retains after activation all or a substantial part of its serine protease activity over a reference activity from another substantially identical wild- type FXa, such as human FXa. The serine protease activity may be assessed, for instance, by bringing into contact a candidate modified FXa with a reference substrate of FXa, such as S-2765 which is chromogenic substrate for factor Xa, as disclosed in Example 2. In a particular embodiment, the "normal FX" has the domain GLA. In another embodiment, the "normal FXa" has the domain EGF. In a particular embodiment, the "normal FXa" has the domains GLA and EGF. Alternatively, the catalytic activity of a FXa polypeptide may be assessed by detecting the irreversible covalent binding of said FXa polypeptide to Diisopropyl- FluoroPhosphate (DFP), to chloromethyl-ketones (PPACK, DEGRCK), to PMSF and/or to other titration reagents such as P-nitro-phenyl p-guanidino benzoate (NPGB) which allow to characterize a catalytically active protease from a catalytically inactive protease. Mutations having the effect of reducing, or even abolishing, the catalytic activity of a FXa polypeptide are known in the Art. In particular, a "catalytically active FXa" must retain a functional catalytic triad, which is generally defined in human Factor Xa by a set of three residues consisting of D 102, H57 and SI 95 according to chymotrypsin numbering. Also, a catalytically active FXa generally comprises a funtional D194 residue, according to chymotrypsin numbering. In view of the above, a FXa or a modified catalytically active FXa must include a functional catalytic triad, in particular a functional SI 95 residue. Preferably, a "catalytically active FXa" comprises a native catalytic site of sequence SEQ ID N°2 or 4. Accordingly, a modified FXa such as des-Gla FXa-S379A (SI 95 A in chymotrypsin numbering) by mutagenesis, as described in WO 2009/042962, is not considered as a catalytically active modified FXa in the sense of the invention.

Indeed, it has been observed that only FXa and active forms of FXa (modified) derivatives are able to form a functional complex with a2M (and thus to retain all or part of the affinity towards non-VKA anticoagulants such as anti-FXa DOAC), in contrast to andexanet-a (unable to form a complex with a2M) and for which the affinity towards anti- FXa DOAC is inferior in comparison. Advantageously, the complex that is formed, which comprises a FXa or modified catalytically active FXa according to the invention is neither pro- nor anti-coagulant.

"FXa Derivatives" or "modified FXa" or "derivatives of a factor Xa protein" are terms used interchangeably herein and refer broadly to FXa proteins that have been modified. A "modified FXa" may consist of a " Gla-deficient FXa", a "Gla-domainless FXa", or "Gla and first EGF domain-deficient FXa". "Gla-deficient FXa" refers to FXa with reduced number of free side chain γ-carboxyl groups in its Gla-domain. Like GDFXa, Gla-deficient FXa can also bear other modifications. Gla-deficient FXa includes uncarboxylated, undercarboxylated and decarboxylated FXa. "Uncarboxylated FXa" or " decarboxylated FXa" refers to FXa derivatives that do not have the γ-carboxy groups of the γ-carboxyglutamic acid residues of the Gla domain, such as FXa having all of its Gla domain γ-carboxyglutamic acid replaced by different amino acids, or FXa having all of its side chain γ-carboxyl removed or masked by means such as amination, esterification, etc. For recombinantly expressed protein, uncarboxylated FXa is, sometimes, also called non- carboxylated FXa. " Under carboxylated FXa" refers to FXa derivatives having reduced number of γ-carboxy groups in the Gla domain as compared with wild-type FXa, such as FXa having one or more but not all of its Gla domain γ-carboxyglutamic acids replaced by one or more different amino acids, or FXa having at least one but not all of its side chain γ- carboxyl removed or masked by means such as amination and esterification. A "Gla- domainless FXa" is a more specific form of ^' " Gla-deficient FXa". A "Gla-domainless FXa" or "des-Gla FXa" or GDFXa refers to FXa that does not have a Gla-domain and encompasses FXa derivatives which may also bear other modification(s) in addition to the removal of the Gla-domain. Examples of GDFXa in this invention include, but are not limited to, FXa derivative lacking the 1-44 amino acids of SEQ ID N°l; FXa derivative lacking the 1-44 amino acid residues of SEQ ID N°l, corresponding to des-Gla FXa after chymotryptic digestion of human FXa (with a light chain corresponding to SEQ ID N°5 and a heavy chain corresponding to SEQ ID N°4; and FXa derivative lacking the entire 1- 44 Gla-domain of SEQ ID N°l as described in Padmanabhan et al, (1993; Structure of human des(l-45) factor Xa at 2.2 A resolution. J Mol Biol; 232;947-66). The domain structure of human GDFXa may be found in Padmanabhan et al. (1993) which is hereby incorporated by reference in its entirety. A "modified FXa" may also consist of a Gla and first domain EGF deficient FXa. Alpha2-Macroglobulin (a ₂ M) is a broad-spectrum inhibitor targeting a number of endopeptidases including serine, cysteine, aspartate, and metallo- proteinases. Blood a ₂ M mainly targets thrombin, FXa and plasmin. Inhibitor (a ₂ M) circulates at 2 to 4 mg/ml with higher concentration in newborns and children. The native molecule is a tetramer composed of 185 kDa identical subunits disulphide bonded in pairs. A bait region (residues 666-706) in each subunit is susceptible to proteolytic cleavage by the various targeted proteases, but each a ₂ M tetramer binds two proteases at the most. Cleavage of a ₂ M triggers a major conformational change which unmasks γ- carboxyl groups esterified in the native molecule to sulfhydryl of free thiol groups. Unmask γ-glutamyl reacts with NH ₂ ε-lysyl groups of the proteases. As a result, the targeted protease is covalently bound to a ₂ M. Within the complex the active site of the sequestered protease is not neutralised such that protease remains functional, pending that substrate may access to its active site. Indeed, sequestered protease still cleaves small peptidyl substrates and is neutralised by small peptidyl chloromethyl ketone even if compared with the free protease the catalytic (k _ca t) and association rate (k _on ) constants are lower. On the contrary most macromolecule substrates or inhibitors are prevented from accessing the active site of the protease due to the steric hindrance resulting from the covalently bound a ₂ M. Specifically thrombin and FXa sequestered by a ₂ M loose their procoagulant functions yet still cleave for instance synthetic substrate H-D-Phe-Pip-Arg- pNA (S2238) and Z-D-Arg-Gly-Arg-pNA (S2765), respectively. For reference, the nucleotide sequence coding human a-2 macroglobulin (a ₂ M) can be found in GenBank, under accession number Ml 1313, and the precursor protein in the UniProtKB/Swiss-Prot under ID P01023 and is listed in SEQ ID N°6.

According to a first embodiment, the invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising a complex between a

Factor Xa (FXa) or a modified catalytically active FXa, and a-2 macroglobulin (a2M); b) determining clot formation in the sample from step a) to obtain a test value; and c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

The above-mentioned steps a), b) and c) may be preceded by a step of determining clot formation in a blood sample from said subject, which does not necessarily contain the said complex.

According to said particular embodiment, the invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising or not an anticoagulant; b) determining clot formation in the sample from step a) to obtain a first test value; and

c) adding a complex between FXa or a modified catalytically active FXa and a-2 macroglobulin (a2M) in the sample from step a); d) determining clot formation in the sample from step c) to obtain a second test value; wherein a variation of the second test value over the first test value is indicative of hemostasis in said individual.

According to said particular embodiment, the invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising or not an anticoagulant; b) determining clot formation in the sample from step a) to obtain a test value; and c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject;

d) adding a complex between FXa or a modified catalytically active FXa and a-2 macroglobulin (a2M) in the sample from step a);

e) determining clot formation in the sample from step d) to obtain a test value; and indicative of the presence of a DOAC in a sample. Such a reference value may have variation of said test value is indicative of the specific contribution of the anticoagulant to hemostasis in said subject.

In particular, according to said methods, a variation of said test values may be indicative of the presence of an anticoagulant (i.e. a VKA or a non-VKA antagonist) in said subject.

According to a second embodiment, the invention relates to an isolated complex between a catalytically active Gla-Domain deficient Factor Xa and a-2 macroglobulin (a2M). In particular, said catalytically active Gla-Domain deficient Factor Xa may be a Gla-Domainless Factor Xa.

According to a third embodiment, the invention relates to a pharmaceutical composition comprising:

(i) a complex between a catalytically active Gla-domain deficient Factor Xa and a-2 macroglobulin; and

(ii) a pharmaceutically acceptable carrier.

According to a fourth embodiment, the invention relates to a method for preparing a complex between a catalytically active Gla-domain deficient Factor Xa (Gla-deficient FXa) and a-2 macroglobulin (a2M) comprising the steps of:

a) providing an isolated catalytically active Gla-deficient FXa; b) providing an isolated a2M;

c) bringing into contact the said Gla-deficient FXa and the said a2M in a reaction medium in a manner suitable for the Gla-deficient FXa to bind to a2M; wherein said a2M is present in the reaction medium in a molar excess of at least 1 :2, in particular of more than 1 :2 and preferably in a [Gla-deficient FXa : a2M] stochiometric molar ratio of about 1 :4, or more than about 1 :4;

cl) optionally incubating the reaction medium for at least 10 minutes;

c2) optionally adding a calcium chelating agent to the reaction medium;

d) recovering the complex between Gla-deficient FXa and a2M-from the reaction medium obtained at steps c), cl) or c2).

According to a fifth embodiment, the invention relates to a kit comprising:

a) a first container containing an isolated catalytically active Gla-domain deficient Factor Xa (Gla-deficient FXa); and

b) a second container containing an isolated a-2 macroglobulm.

According to a sixth embodiment, the invention relates to a kit comprising:

a) a first container containing an isolated Factor Xa or modified catalytically active Factor Xa ; and

b) a second container containing an isolated a-2 macroglobulin;

for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; and in particular for discriminating a defect in haemostasis due to a non- vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy and/or a VKA therapy in said subject. According to a seventh embodiment, the invention relates to a kit comprising:

a) a first container containing an isolated Factor Xa or modified catalytically active Factor Xa ; and

b) a second container containing an isolated a-2 macroglobulin;

for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

The invention also relates to a complex between Factor Xa (FXa) or a modified catalytically active Factor Xa (FXa), and a-2 macroglobulin (a2M); for use as a diagnostic reagent. Thus, according to an eighth embodiment, the invention relates to a complex between Factor Xa (FXa) or a modified catalytically active Factor Xa (FXa), and a-2 macroglobulin (a2M); for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; and in particular for discriminating a defect in haemostasis due to a non-vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy and/or a VKA therapy in said subject.

Typically, the normal FXa nM GDFXa-a2M is analysed in the presence or not of a large excess of antithrombin and enoxaparin. Activity of the same amount of unreacted GDFXa (fully neutralised by antithrombin/enoxaparin) is compared to the activity of the complex to assess the effective concentration of GDFXa sequestered.

The invention also relates to a complex between Factor Xa (FXa) or a modified catalytically active Factor Xa (FXa), and a-2 macroglobulin (a2M); for use as a medicament. Thus, according to a ninth embodiment, the invention relates to a complex between Factor Xa (FXa) or a modified catalytically active Factor Xa (FXa), and a-2 macroglobulin (a2M); for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

The invention will be further detailed hereafter.

Complexes & Kits & Methods for preparing said complexes

Methods for preparing a complex between FXa and a2M have been reported in Meijers et al. (1987; Inhibition of human blood coagulation factor Xa by a2- macroglobulin. Biochemistry; 26:5932-7).

Yet, those methods are not fully satisfactory for providing compositions (i.e. pharmaceutical compositions) comprising a complex between a modified catalytically active FXa and a2M in a soluble form, and having undetectable or reduced-levels of said modified FXa in a non-complexed form. In particular, it has been found that the protocols taught in Meijers et al. (1987) were not optimal for preparing specifically a complex between FXa or a modified catalytically active FXa and a2M.

Indeed, and as previously described, for the complex to be strictly neutral with respect to haemostasis (neither pro- nor anti-coagulant) it is preferable that the composition should be exempt of any trace of free FXa or modified catalytically active FXa.

As taught in Example 4 and the Material & Methods, the inventors provide herein optimized methods for preparing said complexes, and compositions thereof. In particular the inventors provided methods for preparing a complex between modified catalytically active FXa and a2M; wherein the modified FXa is a Gla-deficient FXa (Gla-deficient FXa) or a Gla-domainless FXa (GDFXa).

Kits according the invention are also defined herein, comprising:

a) a first container containing an isolated catalytically active Gla-domain deficient

Factor Xa (Gla-deficient FXa); and

b) a second container containing an isolated a-2 macroglobulin.

Kits of the invention comprising FXa or a catalytically active Gla-domain deficient Factor Xa (Gla-deficient FXa), and an isolated a-2 macroglobulin, are also suitable for the diagnostic and/or therapeutic uses & methods which are disclosed herein.

Diagnosis uses & methods

The present invention also relates to a complex between FXa or a modified catalytically active FXa, and a2M; for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy; and in particular for discriminating a defect in hemostasis due to a non-VKA anticoagulant therapy from a coagulopathy and/or a VKA therapy in said subject.

According to the invention, the expressions « an individual » and « a subject)) includes a human or a non-human mammal, which also includes a primate or a rodent. As used herein, the expression "an individual" preferably means a human mammal.

This invention also pertains to in vitro methods for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, which methods make use of a complex between FXa or a modified catalytically active FXa, and a2M as a diagnostic reagent. Such in vitro methods comprise a step of determining clot formation in a blood sample previously collected from a patient to obtain a test value. Such methods also comprise a further step of comparing the test value which has been previously measured to a reference value.

This invention also pertains to in vitro methods for discriminating a defect in hemostasis due to a non- vitamin K antagonist (non-VKA) anticoagulant therapy from a coagulopathy in said subject, which methods make use of a complex between FXa or a modified catalytically active FXa, and a2M as a diagnostic reagent. Such in vitro methods comprise a step of determining clot formation in a blood sample previously collected from a patient to obtain a test value. Such methods also comprise a further step of comparing the test value which has been previously measured to a reference value.

In preferred embodiments, the test value is selected in a group comprising (i) the clotting time (CT), (ii) the clot formation time, (iii) the alpha angle value, and (iv) the maximum clot firmness, as these test values are provided for example by performing the Rotem™ test which is well known from the one skilled in the art.

As used herein, the clotting time (CT) is the latency time from adding the start reagent to a blood sample until the clot start to form, i.e. the time upon for which a 2 mm amplitude is reached.

As used herein, the clot formation time (CFT) is the time upon which the clot starts to form until a clot firmness of 20 mm point has been reached. The clot formation time is the time to reach 20 mm amplitude from the time upon which a 2 mm amplitude was reached.

As used herein, the alpha-angle value is the angle of tangent between the middle axis and the tangent to the clotting curve through the 2 mm amplitude point. The alpha angle value is indicative of the kinetic of clotting.

As used herein, the maximum clot firmness (MCF) value corresponds to the maximum amplitude (generally expressed in mm) which is reached. The MCF value provides information on the firmness of clot.

As already specified previously herein, this invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising a complex between Factor Xa or a modified catalytically active Factor Xa, and a-2 macroglobulin (a2M); b) determining clot formation in the sample from step a) to obtain a test value; and c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

The above-mentioned steps a), b) and c) may be preceded by a step of determining clot formation in a blood sample from said subject, which does not necessarily contain the said complex. According to said particular embodiment, the invention relates to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) providing a blood sample from said individual comprising or not an anticoagulant; b) determining clot formation in the sample from step a) to obtain a test value; and c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject;

d) adding a complex between FXa or a modified catalytically active FXa and a-2 macroglobulin (a2M) in the sample from step a);

e) determining clot formation in the sample from step d) to obtain a test value; and f) comparing the test value from step b) to the test value from step e) wherein a variation of said test value is indicative of the specific contribution of the anticoagulant to hemostasis in said subject.

This invention also pertains to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) determining clot formation in a blood sample from said individual comprising a complex between Factor Xa or a modified catalytically active Factor Xa, and a-2 macroglobulin (a2M); whereby a test value is obtained; and

b) comparing the test value from step a) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

In particular, this invention also pertains to a method for monitoring hemostasis in a subject who may or may not be under anticoagulant therapy, comprising the steps of:

a) determining clot formation in a blood sample from said individual comprising a complex between FXa or a modified catalytically active FXa, and a2M; whereby a test value is obtained; and

b) determining clot formation in the said blood sample from said individual but not comprising a complex between FXa or a modified catalytically active FXa and a2M.

c) comparing the test value from step a) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

d) comparing the test value from step a) to the test value from step b) to evaluate if an anticoagulants interfere or not with clot formation in the said individual.

In some embodiments of the method, the said method comprises the steps of: a) providing a blood sample from said individual,

b) adding to the said blood sample a complex between FXa or a modified catalytically active FXa, and a2M;

c) determining clot formation in the sample from step b) to obtain a test value; and d) comparing the test value from step c) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

In some embodiments of the method, the said method comprises the steps of:

a) adding a complex between FXa or a modified catalytically active FXa, and a2M to a blood sample previously collected from an individual;

b) determining clot formation in the sample obtained at the end of step a); whereby a test value is obtained; and

c) comparing the test value from step b) to a reference value, wherein a variation of said test value over said reference is indicative of hemostasis in said subject.

Preferably the blood sample which is provided is a whole blood sample.

In some embodiments of the method, the catalytically active modified FXa is a Gla- domain deficient Factor Xa (Gla-deficient FXa).

In some embodiments of the method, the anticoagulant therapy is selected from the group consisting of: a direct FXa or direct thrombin inhibitor (such as DOAC), or an indirect FXa or indirect thrombin inhibitor (such as heparin and derivatives thereof).

In some embodiments of the method, the anticoagulant therapy is selected from the group consisting of: rivaroxaban, apixaban, edoxaban, dabigatran, heparin and heparin derivatives selected from the group consisting of: tinzaparin, enoxaparin, nadroparin, and dalteparine. rivaroxaban, apixaban, edoxaban, dabigatran, heparin and heparin derivatives selected from the group consisting of: tinzaparin, enoxaparin, nadroparin, dalteparine dalteparin, reviparin, bemiparin certoparin, or parnaparin, as well as heparinoids such as danaparoid and chondroitin or dermatan sulfates.

In preferred embodiments of the method, the test value is measured by using a test allowing detection of an anticoagulant therapy, for instance the Rotem™ method.

In preferred embodiments, the test value is selected from a group comprising (i) the clotting time (CT), (ii) the clot formation time, (iii) the alpha angle value, and (iv) the maximum clot firmness. In some embodiments of the method, the reference value is a value which is indicative of the absence of a DOAC in a sample. Such a reference value may have been previously obtained by measuring the said value in a blood sample previously collected from an individual who has not been administered with a DOAC, which includes a blood sample previously collected form an individual which has not been administered with any anticoagulant compound.

In some other embodiments of the method, the reference value is a value which is indicative of the presence of a DOAC in a sample. Such a reference value may have been previously obtained by measuring the said value in a blood sample previously collected from an individual who has been administered with a DOAC, which includes an individual who has been treated with a DOAC selected in a group comprising rivaroxaban or apixaban, or a combination thereof.

In some embodiments, the said method may also allow quantifying the DOAC which is eventually present in a blood sample to be tested. According to these embodiments, a blood sample collected from an individual is added with a given amount of a complex between FXa or a catalytically active modified FXa and a2M and a test value is measured. Then, the test value thus measured is compared to a calibration curve generated with a known amount of said complex and a known amount of a DOAC, so as to determine the amount of DOAC present in the blood sample which is tested.

Therapeutic uses & methods

This invention also pertains to a complex between FXa or a modified catalytically active FXa, and a2M; for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

The present invention also concerns the use of a complex between FXa or a modified catalytically active FXa, and a2M; for preparing a medicament for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in said subject.

This invention also relates to a method for preventing or reducing bleeding in a subject undergoing anticoagulant therapy; and/or for binding and inhibiting an exogenously administered anticoagulant in an individual in need thereof, comprising a step of administering to a subject in need thereof a complex between FXa or a modified catalytically active FXa, and a2M, or a composition comprising a complex between FXa or a modified FXa, and a2M.

In some embodiments, a complex between FXa or a modified catalytically active FXa, and a2M, or a composition comprising a complex between FXa or a modified catalytically active FXa, and a2M is administered to an individual who has already been tested for a treatment with a DOAC by performing an in vitro methods including those described herein.

The pharmaceutical compositions may contain more particularly an effective dose of comprising a complex between FXa or a modified catalytically active FXa, and a2M.

An "effective dose" means an amount sufficient to induce a positive modification in the condition to be regulated or treated, but low enough to avoid serious side effects. An effective amount may vary with the pharmaceutical effect to obtain or with the particular condition being treated, the age and physical condition of the end user, the severity of the condition being treated/prevented, the duration of the treatment, the nature of other treatments, the specific compound or composition employed, the route of administration, and like factors.

A composition comprising a complex between FXa or a modified catalytically active FXa, and a2M may be administered in an effective dose by any of the accepted modes of parenteral administration in the art.

In one embodiment, a pharmaceutical composition comprising a complex between

Factor Xa or a modified Factor Xa, and a-2 macroglobulin (a2M), and/or compositions in the form of a kit according to the invention, are suitable for administration by the oral, nasal, sublingual, aural, ophthalmic, topical, rectal, vaginal, urethral, or parenteral injection route. Most preferably, the administration route is a parenteral injection route. Most preferably, compositions of the invention are sterile and apyrogenic. The route of administration and the galenic formulation will be adapted by one skilled in the art pursuant to the desired pharmaceutical effect.

One of ordinary skill in the art of therapeutic formulations will be able, without undue experimentation and in reliance upon personal knowledge, to ascertain a therapeutically effective dose of a compound of the invention for a given indication. A pharmaceutical composition of the invention may be formulated with any known suitable pharmaceutically acceptable excipients according to the dose, the galenic form, the route of administration and the likes.

As used herein, "pharmaceutically acceptable excipients" include any and all solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic and absorption delaying agents and the like. Except insofar as any conventional excipient is incompatible with the active compounds, its use in a medicament or pharmaceutical composition of the invention is contemplated.

A medicament, kit, or composition of the invention may be in the form of tablets, pills, powders, lozenges, sachets, cachets, elixirs, suspensions, emulsions, solutions, syrups, aerosols, sprays, ointments, gels, creams, sticks, lotions, pastes, soft and hard gelatine capsules, suppositories, sterile injectable solutions, sterile packages powders and the like.

The term "unit dosage form", when used herein, describes physically discrete units, each unit containing a predetermined quantity of one or more desired active ingredient(s) calculated to produce the desired therapeutic effect, in association with at least one pharmaceutically acceptable carrier, diluent, excipient, or combination thereof.

Herein, the phrases "physiologically acceptable carrier" and "pharmaceutically acceptable carrier", which are used interchangeably, describe a carrier or a diluent that does not cause significant irritation to the subject and does not abrogate the biological activity and properties of the active ingredient(s). As used herein, the term "carrier" refers to a diluent, adjuvant, excipient, or vehicle with which the active ingredient(s) is(are) administered. Herein the term "excipient" refers to an inert substance added to a pharmaceutical composition to further facilitate administration of the active ingredient(s).

A pharmaceutical composition may be formulated, for example, as sachets, pills, caplets, capsules, tablets, dragee-cores or discrete (e.g., separately packaged) units of powder, granules, or suspensions or solutions in water or non-aqueous media. Thickeners, diluents, flavorings, dispersing aids, emulsifiers or binders may be desirable.

Pharmacological preparations for oral use can be made using a solid excipient, optionally grinding the resulting mixture, and processing the mixture of granules, after adding suitable auxiliaries if desired, to obtain tablets or dragee cores. Suitable excipients are, in particular, fillers such as sugars, including lactose, sucrose, mannitol, or sorbitol; cellulose preparations such as, for example, maize starch, wheat starch, rice starch, potato starch, gelatin, gum tragacanth, methyl cellulose, hydroxypropylmethyl-cellulose, sodium carboxymethylcellulose; and/or physiologically acceptable polymers such as polyvinylpyrrolidone (PVP). If desired, disintegrating agents may be added, such as cross- linked polyvinyl pyrrolidone, agar, or alginic acid or a salt thereof such as sodium alginate.

Composition unit dosage forms that are used according to the present embodiments may, if desired, be presented in a pack or dispenser device, such as an FDA (the U.S. Food and Drug Administration) approved kit, which may contain one or more unit dosage forms containing the active ingredient(s). The pack or dispenser device may, for example, comprise metal or plastic foil, such as, but not limited to a blister pack. The pack or dispenser device may be accompanied by instructions for administration. The pack may also be accompanied by a notice associated with the container in a form prescribed by a governmental agency regulating the manufacture, use or sale of pharmaceuticals, which notice is reflective of approval by the agency of the form of the compositions for human administration. Such notice, for example, may be of labeling approved by the U.S. Food and Drug Administration for prescription drugs or of an approved product insert.

Sterile injectable forms of the compositions of this invention may be aqueous or oleaginous suspension. These suspensions may be formulated according to techniques known in the art using suitable dispersing or wetting agents and suspending agents. The sterile injectable preparation may also be a sterile injectable solution or suspension in a nontoxic parenterally acceptable diluent or solvent, for example as a solution in 1,3- butanediol. Among the acceptable vehicles and solvents that may be employed are water, Ringer's solution and isotonic sodium chloride solution. In addition, sterile, fixed oils are conventionally employed as a solvent or suspending medium. For this purpose, any bland fixed oil may be employed including synthetic mono- or di-glycerides. Fatty acids, such as oleic acid and its glyceride derivatives are useful in the preparation of injectables, as are natural pharmaceutically-acceptable oils, such as olive oil or castor oil, especially in their polyoxyethylated versions. These oil solutions or suspensions may also contain a long- chain alcohol diluent or dispersant, such as carboxymethyl cellulose or similar dispersing agents which are commonly used in the formulation of pharmaceutically acceptable dosage forms including emulsions and suspensions. Other commonly used surfactants, such as Tweens, Spans and other emulsifying agents or bioavailability enhancers which are commonly used in the manufacture of pharmaceutically acceptable solid, liquid, or other dosage forms may also be used for the purposes of formulation. Compounds may be formulated for parenteral administration by injection such as by bolus injection or continuous infusion. A unit dosage form for injection may be in ampoules or in multi-dose containers.

The invention will be further illustrated by the following figures and examples. However, these examples and figures should not be interpreted in any way as limiting the scope of the present invention.

FIGURES:

Figure 1: Purification of a ₂ M by iminodiacetic acid-Sepharose and Superose

6B— chromatographies. Left panel represents a typical iminodiacetic acid-Sepharose chromatogram of the resuspended PEG precipitate during a ₂ M purification. Column (25 x 160 mm, flow rate 3 ml/min) equilibrated with zinc was washed in 20 mM sodium phosphate pH 6.0 to eliminate unbound material and fraction enriched in a ₂ M was eluted by 0.1 M EDTA pH 8.0. Right panel represents a typical filtration on Superose 6B chromatogram of the fraction enriched in a ₂ M. Column (16 x 500 mm, flow rate 1 ml/min) equilibrated in TBS was loaded with 1 ml fraction at a time. a ₂ M eluted in the second bump in front of the main peak. Fractions between 48 and 63 ml were collected.

Figure 2: Purified a ₂ M and GDFXa-a ₂ M complex are neither pro- nor anti- coagulant (analysis by clot waveform assay). The lack of pro- or anti-coagulant activity of purified a ₂ M and of the GDFXa-a ₂ M complex was assessed by clot waveform assay. A405 increase was recorded in the presence (open symbols) or absence (closed symbols) of 0.75 mg/ml (1.0 μΜ) α ₂ Μ (panel A) or of 1.7 μΜ GDFXa-a ₂ M (panel B; lyophilized material dissolved in PPP). Assay was triggered by adding 5 pM tissue factor, 4 μΜ phospholipids and 18 mM CaCl ₂ . The clot waveform assay was performed with PPP alone (control, n=6) or containing an aliquot of one of the various a ₂ M purifications (n=6) or of the various GDFXa-a ₂ M complexes prepared (n=6). The mean values observed for the lag time (LT or Lag Time, defined as the time needed to reach 15% of the maximum) and the standard errors of the mean (SEM) were 49 ± 0.1 1, 51 ± 0.16 and 54 ± 0.22 s with control, a ₂ M and GDFXa-a ₂ M, respectively. Purified a ₂ M as well as the GDFXa-a ₂ M complex had no significant effect on LT of fibrin polymerization. Overall neither a ₂ M nor GDFXa-a ₂ M hampered the clot waveform assay: neither exhibited pro- or anti-coagulant activity. Figure 3: Catalytic activity of GDFXa and GDFXa-a ₂ M in the presence or not of antithrombin/enoxaparin. The graph represents progress of S2765 hydrolysis by 1 nM free GDFXa (closed circles) or sequestered in the GDFXa-a ₂ M complex (closed squares).

Corresponding open symbols were obtained in the presence of 2.6 μΜ antithrombin and 1IU anti-Xa/ml enoxaparin representing a large excess with respect to free or sequestered GDFXa. Catalytic activity of GDFXa-a ₂ M was unaffected by antithrombin/enoxaparin, whereas free GDFXa was, on the contrary, completely inhibited. It can be concluded that GDFXa added to purified a ₂ M had been fully sequestered and that no free GDFXa remained in the preparation.

Figure 4: WB DOAC titration. Graphs represent the velocity of Z-Gly-Gly-Arg-7 amino-4-methylcoumarin cleavage by 200 nM GDFXa-a ₂ M (panel A) or 125 nM equivalent thrombin activity from the thrombin-a ₂ M TGA calibrator (panel B) as a function of the amount of DOAC added. In both panel, DOAC was spiked in WB at the indicated concentration with rivaroxaban (closed squares), apixaban (close circles), or dabigatran (close diamonds) and diluted 1/4 before test. The GDFXa-a ₂ M complex allowed detecting as little as 30 ng/ml apixaban or rivaroxaban but was not inhibited by dabigatran. In contrast the thrombin-a ₂ M complex was inhibited by dabigatran above 100 ng/ml but unaffected by apixaban or rivaroxaban. That the catalytic activity of GDFXa-a ₂ M was unaffected by dabigatran suggested that the mechanism of dabigatran neutralisation did not involve the active site of GDFXa (see figure 8 panel A). Noteworthy GDFXa-a ₂ M cleaved the fluorogenic substrate rendering TGA unsuitable to study anti- FXa DOAC neutralisation by the GDFXa-a ₂ M complex.

Figure 5: Differential diagnosis of anti-FXa DOAC in WB (Rotem assay). The graph represents the CT values obtained through Rotem™ assays in the presence or not of GDFXa-a ₂ M as a function of the effective concentration of anti-FXa DOAC measured in PPP. The reference range in the absence of inhibitor, taking into account the intra- and inter- variability, is 102-178 s and is indicated by the horizontal dashed lines. CT values were within the reference range in the absence of inhibitor whether or not GDFXa-a ₂ M had been added. In the absence of GDFXa-a ₂ M the CT values increased with increasing amount of rivaroxaban (closed squares) as well as apixaban (closed circles) and values all exceeded the upper limit of the reference range. In contrast, the same assays performed in the presence of 1.7 μΜ GDFXa-a ₂ M (rivaroxaban open squares; apixaban open circles) resulted in CT values which were all within the reference range. Overall Rotem™ assay suggested that 1.7 μΜ GDFXa-a ₂ M was devoid of pro- or anti-thrombotic effect and neutralised up to 600 ng/ml (at least 1 μΜ) anti-FXa DO AC.

Figure 6: GDFXa-a ₂ M fully neutralises anti-FXa DOAC (differential diagnosis by clot waveform assay)i-Graph represents the LT ratios (with respect to control value) of the clot waveform assay as a function of the anticoagulant drug added. Clot waveforms assays were triggered by adding 5 pM tissue factor, 4 μΜ phospholipids and 18 mM CaCl ₂ . Assays were repeated 3 times, and for each anticoagulant concentration, median, maximum and minimum values of the LT ratio obtained are represented. Inhibitor was added to PPP at the indicated concentration (in ng/ml) and clot waveform assay triggered in the presence or absence of 1.7 μΜ GDFXa-a ₂ M complex. PPP was spiked with apixaban (circles) or rivaroxaban (squares). Closed symbols data were obtained in the absence and open symbol data in the presence of GDFXa-a ₂ M. LT ratios increased with the anti-FXa DOAC which were fully neutralised (up to at least 600 ng/ml) by the GDFXa- a ₂ M complex.

Figure 7: GDFXa-a ₂ M complex does not modify the LT ratio of VKA-treated patients (clot waveform assay). Graph represents the LT ratio (with respect to a reference PPP; see Figure 6) as a function of the patient's INR. Clot waveform assays were performed as in Figure 6 on PPP from six VKA treated patients in the absence (closed squares) or presence of GDFXa-a ₂ M complex (open squares). With or without added complex LT ratios were comparable. Therefore GDFXa-a ₂ M complex has no detectable effect on the LT ratio of VKA treated patients.

Figure 8A-D: GDFXa-a ₂ M neutralises in part dabigatran, UFH and LMWH (differential diagnosis by clot waveform assay). Graphs represent the LT ratio (with respect to reference PPP; see Figure 6) as a function of the anticoagulant drug added. Clot waveform assays (repeated 3 times) were performed as in Figure 6. For each anticoagulant concentration, median, maximum and minimum values of the LT ratio obtained are represented. In each panel the inhibitor was added to PPP at the indicated concentration (in ng/ml or IU/ml) and clot waveform assay triggered in the presence or absence of 1.7 μΜ GDFXa-a ₂ M complex. In panel A dabigatran was spiked in PPP containing (open diamonds) or not GDFXa-a ₂ M (closed diamonds). Surprisingly, the GDFXa-a ₂ M complex neutralises at least 80% of up to 600 ng/ml dabigatran. GDFXa-a ₂ M also neutralises at least 60% of up to l .O IU/ml UFH (open squares in panel B) or 1.5 IU anti-Xa/ml tinzaparin (open circles in panel C) and about 40% of up to 1.5 IU anti-Xa/ml enoxaparin (open triangles in panel D).

Figure 9: Neutralization of rivaroxaban by GDFXa-a2M in mice. Mice were randomized into 4 groups: group I (control) was force-fed with 10 mM HCl and injected with vehicle (50 mM Tris pH 7.5 containing 0.1 M NaCl and 5 mM Mn2+). Group II and III were force-fed with rivaroxaban (50 mg/kg) and injected with vehicle or GDFXa-a2M (200 μΐ, 3.6 μΜ), respectively. Group IV was force-fed with 10 mM HCL and received the same amount of GDFXa-a2M. Graphs represent secondary BT (left panel) and BL (right panel) in each group of 8 to 10 mice. Values on the Y axe are the sum of bleeding recorded after challenge 15, 30 and 45 min following the initial tail vein transection. BT and BL were significantly increased only in group II suggesting that GDFXa-a2M neutralized in vivo rivaroxaban- induced bleeding (group III).

Figure 10: Neutralization of dabigatran by GDFXa-a2M in mice. Twelve mice were randomized into 2 additional groups (V and VI): both were force-fed with dabigatran etexilate (20 mg/kg). After 90 min 100 μΐ vehicle (group V) or GDFXa-a2M (group VI) were injected in each retro-orbital plexus. Graphs represent secondary BT (left panel) and BL (right panel) in each group, as in figure 10. Control values of groups I and II (see figure 10) are included for the purpose of comparison. BT and BL were significantly increased only in group V suggesting that GDFXa-a2M in vivo neutralized dabigatran-induced bleeding (group VI).

Figure 11: Safety of GDFXa-a2M in mice. PPP from mice injected with either vehicle (control) or GDFXa-a2M were analyzed for d-dimers and thrombin-antithrombin complexes (TAT) content as well as by clot waveform assay. Neither d-dimers (left panel), TAT (middle panel), nor LT in the clot waveform assay (right panel) differed significantly between the two groups. Accordingly, GDFXa-a2M did not increase the basal prothrombotic potential thus appeared safe in mice.

Figure 12: Half-life of GDFXa-a2M in mice. GDFXa-a2M was injected in each retro-orbital plexus of mice. Two min after injection 80 μΐ WB were initially collected by tail vein puncture. WB was later collected by cardiac puncture at the indicated time. PPP was prepared and initial steady-state velocities (vs) of S2765 cleavage were measured in each sample. Graph represents the normalized values of each vs (% residual activity compared to the initially collected sample) as a function of time after GDFXa-a2M injection. Each vs value represents the mean ± SEM of 3 evaluations for each time point. Non-linear regression analysis using a one phase exponential decay equation suggests that when injected at high concentration the apparent half-life of GDFXa-a2M in mice is 85 min (R2 = 0.90).

EXAMPLES

Material & Methods

Whole Blood (WB) and Platelet-Poor Plasma (PPP) samples

Whole Blood (WB) samples were collected by venipuncture in tubes containing (1/10 v/v) 0.105 M buffered trisodium citrate (BD Vacutainer, Plymouth, UK) at the local blood bank (Etablissement Francais du Sang, Paris, France; convention C CPSL UNT n°13/EFS/064) from healthy volunteers who gave their written informed consent. None had history of coagulopathy or received treatment that could interfere with haemostasis in the previous three weeks. WB was stored at room temperature (RT) and processed within 4 h from blood collection. Before use, tubes were gently inverted ten times to re-suspend sedimented cells. WB or Platelet-Poor Plasma (PPP) was added directly to lyophilized GDFXa-a ₂ M complex when required. Corresponding PPP was obtained by centrifugation of citrated WB at 2400 g for 10 min. Pooled normal plasma (Cryocheck) was purchased from Cryopep (Montpellier, France). Blood samples from six VKA treated patients who gave their written informed consent were collected (CRB BB-0033-00064-Hopital Lariboisiere). PPP were prepared as above within 30 min of collection. International Normalized Ratios (INR) were measured on STA-R (Stago, Asniere, France) using Neoplastin CI Plus® as thromboplastin reagent. Fibrinogen level was within normal range and anti-Xa activity undetectable.

Drugs, proteins and reagents

Apixaban and rivaroxaban were kindly provided by Bristol-Myers Squibb/Pfizer (Princeton, NJ, USA), and Bayer Healthcare AG (Leverkusen, Germany), respectively. Dabigatran was purchased from Euromedex (Souffelweyersheim, France). Apixaban and rivaroxaban were dissolved in pure dimethyl sulfoxide (DMSO) at a concentration of 4.30 and 3.75 mg/ml, respectively, and stored at -80°C while dabigatran was dissolved in 60% DMSO 40 mM hydrochloric acid at a concentration of 5 mg/mL and stored at -80°C. Just prior to use, aliquots of rivaroxaban and dabigatran were rapidly diluted 1/100 in H ₂ 0. Further dilutions were performed in 20 mM tris(hydroxymethyl)aminomethane (Tris) pH 7.5 containing 0.15 M NaCl and 1% DMSO. Initial sudden dilution in H ₂ 0 was not required for apixaban. DO AC were spiked in WB or PPP. Their effective (final) concentrations were systematically verified in PPP by anti-Xa activity measurement on STA-R (Stago) using the STA-Liquid Anti-Xa assay and specific set of calibrators for rivaroxaban and apixaban (Stago) or by diluted thrombin time measured on STA-R (Stago) using the Hemoclot Thrombin Inhibitor assay (Hyphen BioMed, Neuville-sur-Oise, France) and specific set of calibrator for dabigatran (Hyphen BioMed). Heparin and derivatives were added directly to PPP. Enoxaparin was purchased from Sanofi Aventis (Gentilly, France), tinzaparin from Laboratoire Leo (Vernouillet, France) and sodic heparin from Panpharma (Boulogne-Billancourt, France). Dilutions were performed in 20 mM Tris pH 7.5 containing 0.15 M NaCl. Human GDFXa expressed in bacteria was purchased from Cambridge Protein Works (Cambridgeshire, UK). Antithrombin was purchased from LFB (Aclotine, Courtaboeuf, France). Recombinant human TF (Innovin) was purchased from Dade Behring (Marburg, Germany). Phospholipid vesicles were prepared by sonication (2 min in pulse mode 0.15 s ^"1 , 80 W, 4°C) of a 1 mg/ml mixture of L-a- phosphatidylcholine (66%, w/w) with L-a-phosphatidylserine (33%, w/w), both from Avanti Polar Lipids (Alabaster, AL., U.S.A.) as previously described (Le Bonniec et al. 1992; The role of calcium ions in factor X activation by thrombin E192Q. J Biol Chem; 267:6970-6). Chromogenic substrate N-a-benzyloxycarbonyl-d-Arg-Gly-Arg-pNA (S2765) and Phe-Pro-Arg-chloromethyl ketone (PPACK) were purchased from Cryopep (Montpellier, France) and Z-Gly-Gly-Arg-7 amino-4-methylcoumarin fluorogenic substrate (FluCa-Kit) from Stago. Phenylmethylsulfonyl fluoride (PMSF) was from Sigma Aldrich (Chemie Gmbh, Steinheim, Germany) and aprotinin from Nordic Group Pharmaceuticals (Paris, France).

Human (I2M purification

Human a ₂ M purification was adapted from published protocols in: Harpel (1970; Human plasma alpha-2-macroglobulin, an inhibitor of plasma kallikrein. J Exp Med; 132:329-52) and Hoogendoorn et a . (1991; a ₂ -Macroglobulin Binds and Inhibits activated Protein C. Blood; 78:2283-90). Briefly, fresh frozen citrated human plasma was thawed at 37°C and a mixture of inhibitors was added (1 μΜ PPACK, 2 mM PMSF, 2 mM EDTA and 100 KlU/ml aprotinin). Plasma was filtered on whatman paper then on 0.45 μιη polyethersulfone membrane (Millipore Stericup filter units, Sigma Aldrich). Plasminogen was removed by chromatography on Lysine-Sepharose 4B and vitamin K dependent factors by 80 mM barium chloride precipitation (30 min, 4°C) followed by centrifugation (20 min, 4°C, 4500 g). Supernatant was brought to 4% polyethylene glycol (PEG-6000) by slow addition of a 50% solution in water (w/v), stirred for 30 min at RT and centrifuged as above. Supernatant was brought to 12% PEG-6000, stirred for 1 h at RT and centrifuged (30 min, RT, 1500 g). Pellet was resuspended in 1/10 of the initial plasma volume in 20 mM sodium phosphate, pH 6.0. a ₂ M was then adsorbed onto iminodiacetic acid- Sepharose column (25 x 160 mm, 3 ml/min) equilibrated in 20 mM sodium phosphate pH 6.0 following saturation by 50 mM zinc acetate. Column was washed with 20 mM sodium phosphate pH 6.0 containing 0.15 M NaCl and bound proteins were eluted with 0.1 M EDTA pH 8.0. Eluted material was concentrated by ultrafiltration (Amicon Ultra- 15 100K, Merck KGaA, Darmstadt, Germany) and traces of active protease inhibited by adding 150 μΜ PPACK. a ₂ M was further purified by gel filtration on Superose 6B column (16 x 500 mm, 1 ml/min) equilibrated in Tris 50 mM pH 7.5 containing 0.1 M NaCl (TBS). Fractions containing a ₂ M were pooled and concentrated by ultrafiltration up to 15 mg/ml estimated by immunonephelometry analysis (BN II Siemens, Dr. N Neveu, Laboratoire de biologie medicale et physiologie, Pr. L Cynober, CHU Cochin, Paris, France). Purity of a ₂ M was evaluated by polyacrylamide gel electrophoresis using NuPAGE-Bis-Tris Mini Gels 4-12% (Life technologies, Bleiswijk, Netherlands) and SeeBlue Plus2 pre-stained protein standard (Life technologies).

GDFXa-a.2M complex formation

Complex formation was induced by incubation of 2 μΜ GDFXa with 8 μΜ α2Μ for 20 min at 37°C in TBS containing 5 mM MnCl ₂ . The protocol was adapted to GDFXa constructs from Meijers et al. (1987; Inhibition of human blood coagulation factor Xa by a2-macroglobulin. Biochemistry; 26:5932-7) and Heeb et al. (1991; Identification of divalent metal ion-dependent inhibition of activated protein C by a ₂ -Macroglobulin and a ₂ - Antiplasmin in blood and comparisons to inhibition of factor Xa, thrombin and plasmin. J Biol Chem 1991 ; 266: 17606-12). Reaction was stopped by adding 11 mM EDTA to avoid precipitation/aggregation. Completion of the reaction was evaluated by comparing the rate of 400 μΜ S2765 hydrolysis by 1 nM GDFXa-a2M in the presence or not of a large excess of antithrombin and enoxaparin (2.6 μΜ antithrombin and 1 IU anti-Xa/ml, respectively) enoxaparin. Activity of the same amount of unreacted GDFXa (fully neutralised by antithrombin/enoxaparin) was compared to the activity of the complex to assess the effective concentration of GDFXa sequestered. GDFXa-a2M complex was dialysed against 100 mM ammonium acetate pH 6.5 and lyophilized.

WB DOAC titration

WB containing or not anti-FXa DOAC was diluted 1/4 (v/v) in TBS containing 20 mM EDTA and 200 nM GDFXa-a ₂ M. One hundred μΐ of the dilution were dispensed in a well of a flat bottom microtitre plate prewarmed at 37°C. Reaction was initiated in a Tecan Infinite M200 Pro reader by automated injection of 20 μΐ Z-Gly-Gly-Arg-7 amino-4- methylcoumarin. Progress of fluorescent substrate hydro lyses (excitation 340 nm, emission 440 nm, Z-position 20000 μιη) was recorded for 15 min every 5 s at 37°C. Microtitre plate was shacked between each reading to maintain red blood cells in suspension (5 s, orbital, amplitude 6). Data were exported to GraphPad Prism software and steady state rate of hydrolysis deduced by linear regression analysis. Dabigatran titration was performed similarly except that the thrombin-a ₂ M complex commercialized by Stago (calibrator for TGA) was substituted for the GDFXa-a ₂ M complex.

Rotational thromboelastometry (Rotem)

Rotational thromboelastometry was performed on a Rotem delta in 340 μΐ cup (Pentapharma GmbH, Munich, Germany). Sensitized Rotem assay was triggered by adding 2.5 pM tissue factor and 10 μΜ phospholipid vesicules. Coagulation was initiated by adding 300 μΐ prewarmed WB containing or not DOAC and/or the GDFXa-a2M complex to a mixture of CaCl ₂ and triggering solution. Cup holder was promptly placed to the measuring position of the instrument and viscosity recording started immediately. All assays (with or without DOAC) were performed in the presence of 0.05% DMSO. Clotting Time (CT), clot formation time, alpha angle, and maximum clot firmness were provided by the Rotem software analysis. CT was the most relevant parameter to detect anti-FXa DOAC and the unique parameter considered herein. Aliquot of each blood samples was centrifuged at 2400 g for 10 min and the resulting PPP was used to determine the exact DOAC amount contained in each sample by anti-Xa activity measurement.

Clot waveform assay

Kinetic of fibrin polymerization was studied in PPP by turbidimetry. Clot waveform assay was analyzed by differentiation of raw data from turbidity kinetics. Briefly A405 increase was recorded in the presence of the GDFXa-a ₂ M complex and of 1/20 (v/v) vehicle (1% DMSO in TBS) containing or not DOAC. Final amount of DMSO was 0.05% which was shown not to interfere with clot formation. Clot waveform assay was also used to assess the neutrality of purified a ₂ M compared with TBS (10 mg/ml, 1/20, v/v). Twenty five μΐ PPP-Reagent (Stago) containing 5 pM tissue factor and 4 μΜ phospholipid vesicles were added to 100 μΐ of each PPP mixture and 100 μΐ of this mixture were dispensed in a well of a flat bottom microtitre plate pre-warmed at 37°C. Coagulation was initiated in a thermostated Tecan Infinite M200 Pro reader by automated injection of 20 μΐ Hepes pH 7.35 containing 60 mg/ml bovine serum albumin and 100 mM CaCl ₂ . A405 was recorded every 8 s for 30 min at 37°C. Data were exported to GraphPad Prism software and clot waveform was deduced from the first-derivative of the turbidity kinetic. Data were normalized and the time needed to reach 15% of the maximum was used to define the lag time (LT) in each data set, as previously described in Hantgan et al. (1979; Assembly of fibrin. A light scattering study. J Biol Chem; 254:11272-81).

Animals

C57Bl/6JPvj male mice weighing between 25 and 30 g were purchased from Janvier Labs (Le Genest-Saint-Isle, France) and used in all in vivo assays. They were acclimatized for at least 7 days at the CRP2 Animal facility of Faculte de Pharmacie, Universite Paris Descartes, France) in standard conditions of dark/light cycle, temperature (20°C), humidity (40-60 %), watered and fed diet. At the end of each experiment, anesthetized mice were euthanized by cervical dislocation. All animal experiments were approved by the Ethic Committee on Animal Resources of Universite Paris Descartes (registration number 201506151109793 - V5 APAFiS #2677).

Mouse tail vein transection bleeding model

Mice bleeding model sensitive to pharmacological intervention was adapted from

Johansen et al. (2016; Development of a tail vein transection bleeding model in fully anaesthetized haemophilia A mice - characterization of two novel FVIII molecules. Haemophilia 22:625-31). Using a suitable intubation cannula, mice were forced feed (100 μ1/10 g) with 50 mg/kg rivaroxaban, 20 mg/kg dabigatran etexilate or 10 mM HC1. Two hours after rivaroxaban administration (90 min for dabigatran etexilate), mice were anesthetized by intraperitoneal injection of a ketamine (80 mg/kg) and xylazine (16 mg/kg) mixture. Seven min later, GDFXa-a2M complex (3.6 μΜ) or vehicule (50 mM Tris-HCl containing 0.1 M NaCl and 5 mM Mn2+) was injected (100 μΐ) in each retro-orbital plexus. Mouse tail was soaked in a mixture of NaCl (0.15 M) and EDTA (2 mM) at 37°C in a collection tube and mouse was placed on a 37°C heating pad. After 2.5 min, mouse was positioned on its right side and tail introduced in a home made device enabling positioning at precisely its 2.5 mm diameter section. Using a mechanical linear guide the left lateral tail vein was transected by a 0.5 mm deep incision using a # 23 surgical blade (Swann-Morton, Sheffield, England). Following incision, mouse tail was replaced into the 37°C soaking mixture and initial bleeding monitored. Fifteen, 30 and 45 min post-injury, wound was challenged by gently wiping it twice with a 37°C saline wetted gauze swab in the distal direction and bleeding was monitored again. Collection tube was changed after each challenge. Red blood cells were collected (1500 g 20°C, RT) and resuspended in 5 ml 20 mM Tris pH 7.5 to lyse the cells. A416 was recorded and transcribed into μΐ blood loss (BL) in reference to a titration curve and assuming an average 60% hematocrit in C57Bl/6JRj male mice. Initial bleeding time (BT) was defined as the time to first bleeding cessation following incision and secondary BT as the sum of the three BT following challenges. Primary and secondary BL were defined accordingly.

Evaluation of GDFXa-a2M safety in mice

Mice were anesthetized and GDFXa-a2M injected as above. Mice were placed on a 37°C heating pad. Thirty min after GDFXa-a2M injection WB was collected in 0.11 M buffered trisodium citrate 9/1 (v/v) by cardiac puncture (23 G citrated needle). D-dimers and thrombin-antithrombin complexes (TAT) were measured in PPP by Elisa according to the manufacturer instructions using Mouse D-Dimer (D2D) Elisa Kit (Cusabio, anticorps- enligne.fr) and TAT complexes Mouse Elisa Kit (Abeam, Paris, France), respectively. Prothrombotic potential of GDFXa-a2M was also evaluated by clot waveform assay performed as described above.

Apparent Half-life and persistence of GDFXa-a2M in mice

Mice were anesthetized and high amount GDFXa-a2M injected as above. Two min after GDFXa-a2M injection 80 μΐ WB was collected by tail vein puncture and added to an excess buffered trisodium citrate (20 μΐ 0.11 M). Other time points (at 2, 3, 4, 5 and 6 hours post injection) were collected by cardiac puncture and added to anticoagulant in excess. Mouse PPP was prepared as described above for human PPP. Residual GDFXa- a2M complex was evaluated in PPP by measuring the catalytic activity on S2765 substrate cleavage. Practically, 4 μΐ PPP (taking into account the buffered trisodium citrate in excess) were added to 146 μΐ TBS in a well of a flat bottom microtiter plate pre-warmed at 37°C. Reaction was initiated in a thermostated Tecan Infinite M200 Pro reader by automated injection of 50 μΐ, 1600 μΜ S2765. Progress of S2765 hydrolysis was recorded every 6 sec for 5 min at 37°C. Data were exported to GraphPad Prism Software and initial steady-state rate of hydrolysis (vs) deduced by linear regression analysis. The dependence of vs on time was analyzed by non-linear regression analysis using one phase exponential decay equation to estimate the half-life of GDFXa-a2M in vivo.

Statistical analyses

Non- linear regression analysis and the resulting coefficient of determination (R2) for the KI and half-life estimations were computed using GraphPad Prizm Software. KI values are expressed as mean ± SD and vs values for half-life determination as mean ± SEM. BT and BL were compared all together by Kruskal- Wallis test followed by Dunn tests performed for pair-wise comparisons. D-dimer, TAT values and LT of clot waveform assays were compared between control and antidote groups using two-tailed Mann Whitney test. Statistical significance was accepted when the p-value was below 0.05.

Example 1: Human a ₂ M purification and characterization

Human a ₂ M was purified by a combination of PEG precipitation, affinity chromatography and gel filtration as described in the Material and Methods (figure 1). Typically, starting from 300 ml PPP 50-100 mg a ₂ M were obtained representing about 15% yield. Concentration of the purified a ₂ M estimated by immunonephelometry was 15- 20 μΜ. Purified a ₂ M appeared homogenous on polyacrylamide gel electrophoresis with a major band visible by Coomassie blue staining at the expected 360 kDa molecular weight. Purified a ₂ M was devoid of anti- or pro-coagulant activity as attested by clot waveform assay: a ₂ M had no detectable effect on the lag time of fibrin polymerization (figure 2).

Example 2: GDFXa-a ₂ M complex formation and characterization

To optimize the anti-FXa DOAC neutralisation potential we first attempted incubating GDFXa with a ₂ M in a 2: 1 stochiometric molar ratio to form the GDFXa-a ₂ M complex, as disclosed in Meijers et al. (1987) for complex formation with FXa or Heeb et al. (1991) for complex formation with other proteases including thrombin or plasmin. However, prolonged incubation resulted in an insoluble precipitate whereas stopping the reaction after 20 min incubation through EDTA addition resulted in incomplete sequestration of GDFXa. For the complex to be strictly neutral with respect to haemostasis (neither pro- nor anti-coagulant) it was of upmost importance that product be exempt of any trace of free GDFXa. Unreacted GDFXa turned out difficult to separate from the formed complex. Even further filtration over superose 6B column failed to totally remove free GDFXa. S2765 hydrolysis by free GDFXa is rapidly inhibited by antithrombin in the presence of enoxaparin (figure 3). Thus selective neutralisation of the remaining free GDFXa was feasible through antithrombin and enoxaparin addition but required subsequent eviction of the GDFXa-antithrombin complex formed as well as excess antithrombin and enoxaparin. Our attempts to separate in the reaction mixture GDFXa- antithrombin complex, antithrombin and enoxaparin from the GDFXa-a ₂ M complex also failed. Actually this observation is consistent with the hypothesis that the GDFXa-a ₂ M complex binds LMWH with relatively high affinity (see below). Consequently, we used a 1 :4 stochiometric molar ratio (instead of the 2: 1) to form the GDFXa-a ₂ M complex, and a 20 min incubation period stopped by EDTA addition resulting in complete sequestration of GDFXa without detectable free GDFXa activity whereas in these conditions, hydrolysis by the GDFXa-a ₂ M complex was unaffected following addition of a large excess of antithrombin and enoxaparin suggesting absence of free GDFXa (figure 3). Comparison of the catalytic activity of GDFXa-a ₂ M complex in the presence or absence of antithrombin/enoxaparin therefore allowed to access whether traces of free GDFXa remained in the preparation. It is to mention that in this assay, the concentration of GDFXa-a ₂ M (1 nM) was well below that of enoxaparin. Thus even if complex bound a small amount of enoxaparin (see below) most remained free and ample enough to fully catalyze neutralisation by antithrombin of any free GDFXa. Assay permitted estimation of the GDFXa-a ₂ M active concentration based on its amido lytic activity by reference to a titration curve. The amidolytic activity of sequestered GDFXa within the complex was less than that of free GDFXa. Thus overall catalytic activity of the complex was somewhat lower than that of free GDFXa. We measured the KI of apixaban and rivaroxaban for GDFXa-a2M and found that they were slightly higher than for FXa. Estimated by nonlinear regression analysis, values of 2.41 ± 0.22 nM (R2 = 0.84) and 1.29 ± 0.13 nM (R2 = 0.87) were obtained for apixaban and rivaroxaban, respectively. Thus KI too were slightly higher than those reported for FXa (0.47 ± 0.02 nM and 0.74 ± 0.03 nM; Jourdi et al 2015; Association rate constants rationalise the pharmacodynamics of apixaban and rivaroxaban. Thromb Haemost; 114:78-86.). Each pair of a ₂ M subunits may bind one GDFXa molecule which additionally may crosslink an adjacent a ₂ M subunit. High molecular weight species may also develop through intermolecular crosslinking. Analysed by polyacrylamide gel electrophoresis in nonreducing conditions the GDFXa-a ₂ M complex exhibited four major bands consistent with the various expected molecular species (data not shown). Preparations were kept at 4°C until use, alternatively dialyzed against ammonium acetate, aliquoted, and lyophilized. PPP or WB was added directly on top of lyophilized powder allowing reaching in PPP or WB up to 1.7 μΜ a ₂ M-sequestered GDFXa.

Example 3: WB DOAC titration

We confirmed that the catalytic activity of thrombin within the thrombin-a ₂ M complex used as calibrator in the standard Thrombin Generation Assay (TGA) was inhibited by dabigatran but not by anti-FXa DOAC (figure 4). A major advantage of fluorescence measurement at 440 nm is that emitted signal is only partially quenched by red blood cells thus opening the prospect of WB quantification of dabigatran. We reasoned that similarly GDFXa-a ₂ M might be used to titrate the anti-FXa DOAC. To test this hypothesis we compared the rate of Z-Gly-Gly-Arg-7 amino-4-methylcoumarin cleavage by the thrombin-a ₂ M or GDFXa-a ₂ M complexes in the presence or not of dabigatran, apixaban or rivaroxaban. WB dilution 1/4 did not preclude detection of as little as 30 ng/ml apixaban or rivaroxaban (figure 4). This approach permitted therefore differential diagnosis of the anti-FXa DOAC inhibiting the catalytic activity of GDFXa within the GDFXa-a ₂ M complex and not the thrombin activity of thrombin within the thrombin- ₂ M complex. Inversely the thrombin- ₂ M complex permitted differential diagnosis of anti- thrombin DOAC not inhibiting GDFXa within the GDFXa-a ₂ M. Thus a ₂ M complexes would permit a rapid, point of care, differential diagnosis of DOAC.

Example 4: Neutralisation of anti-FXa DOAC by the GDFXa-a ₂ M complex Sensitized Rotem™ and clot waveform assay are two coagulation-based assays allowing accurate detection of DOAC and were used to evaluate neutralisation of the anti- FXa DOAC by GDFXa-a ₂ M. Clot waveform assays in PPP were triggered in the presence of 0, 30, 200, or 600 ng/ml apixaban or rivaroxaban corresponding to the level expected three half-lives past the average steady-state, the peak therapeutic concentration, and an overdose, respectively. For Rotem assay, WB of 3 healthy volunteers was spiked with rivaroxaban or apixaban and effective plasmatic concentrations were measured by anti-Xa assay. Each sample contained or not a flat 1.7 μΜ GDFXa-a ₂ M. Assays were repeated at least 3 times. Clotting Time (CT) values for the Rotem assay and Lag Time (LT) values for the clot waveform assay were recorded and used to evaluate efficacy of the GDFXa-a ₂ M complex. When neither GDFXa-a ₂ M nor anti-FXa DOAC was added, CT values obtained in WB through the sensitized Rotem assay were within the reference range. We evaluated that the intra- and inter- variability of the CT values in 10 healthy volunteers was 140 ± 38 s (mean plus 1.96-fold SD) for the sensitized Rotem. Adding apixaban or rivaroxaban dose-dependently increased the CT value which was higher than the reference range with concentrations above 30 ng/ml (figure 5). The same was not true following further addition of 1.7 μΜ GDFXa-a ₂ M: all CT values where within the reference range irrespective of the apixaban or rivaroxaban amount up to at least 600 ng/ml (figure 5). Similarly, LT values of the clot waveform assay dose-dependently increased with the concentration of apixaban or rivaroxaban which were already detectable at 30 ng/ml, whereas LT values were unaffected in the presence of 1.7 μΜ GDFXa-a ₂ M (figure 6). We concluded that 1.7 μΜ GDFXa-a ₂ M was sufficient to fully neutralise supratherapeutic amounts of apixaban or rivaroxaban.

Example 5: Differential diagnosis in VKA-treated patients

Projected differential diagnosis requires that GDFXa-a ₂ M neutralises anti-FXa DOAC and restores normal values in haemo stasis tests preferably only if the sample contains the drug: without drug test results must be unaffected. Thus tests affected by a coagulopathy must provide identical results, whether GDFXa-a ₂ M had been added or not. In VKA-treated patients the vitamin K dependent coagulation factors are more or less γ- carboxylated with the result that blood clotting is more or less impaired. The defect is quantified by using the so called International Normalized Ratio (INR). PPP from VKA- treated patients do not normally contain anti-FXa or anti-thrombin DOAC nor heparin and derivatives thus GDFXa-a ₂ M should not affect their clotting time. To verify the neutrality of the GDFXa-a ₂ M complex we tested PPP from six patients having INR between 1.0 (normal) and 7.0 (revealing VKA over medication). The LT of clot waveform assay in the presence or not of the GDFXa-a ₂ M complex was measured and LT ratios calculated according to LT obtained with the reference PPP. Clearly, adding 1.7 μΜ GDFXa-a ₂ M did not influence the LT ratio irrespective of the INR (figure 7). We concluded that GDFXa- a ₂ M had no effect on the coagulation abnormality induced by VKA thus that our differential diagnosis would permit to rule out that the coagulation defect in a VKA treated patient originates from a DO AC.

Example 6: Unexpected neutralisation of dabigatran, UFH and LMWH by GDFXa-a ₂ M

GDFXa-a ₂ M complex was designed to neutralise anti-FXa DOAC and therefore was not initially expected to neutralise the anti-thrombin DOAC and heparin or its derivatives. Indeed, it was initially for the purpose of validating the neutrality of the GDFXa-a ₂ M complex that we measured the LT ratio of PPP containing dabigatran expecting that, as with PPP from VKA-treated patients, the GDFXa-a ₂ M complex would not affect the LT values. Clot waveform assays were triggered in PPP spiked with 0, 30, 200, or 600 ng/ml dabigatran in the presence or not of 1.7 μΜ GDFXa-a ₂ M. LT values of the clot waveform assay dose-dependent ly increased with the concentration of dabigatran, yet to our surprise LT values were little affected by dabigatran in the presence of 1.7 μΜ GDFXa-a ₂ M (figure 8). As already reported for the WB DOAC titration assay dabigatran does not inhibit FXa or GDFXa and indeed, we confirmed that even 5 μΜ dabigatran was not inhibiting the catalytic activity of GDFXa whether free or sequestered in the GDFXa- a ₂ M complex. Following the same initial expectation of neutrality, clot waveform assays were triggered in PPP containing 0.25, 0.5 or 1.0 IU/ml UFH, 0.5, 1.0 or 1.5 IU anti-Xa/ml enoxaparin or tinzaparin. Each assay was performed in the presence or not of 1.7 μΜ GDFXa-a ₂ M. LT values dose-dependently increased with the concentration of heparin or derivatives but again LT values were much less affected in the presence of 1.7 μΜ GDFXa-a ₂ M (figure 8). We concluded that GDFXa-a ₂ M neutralised at least in part dabigatran and heparin or derivatives. Thus including through Rotem or clot waveform assays, GDFXa-a ₂ M would permit a differential diagnosis between coagulopathies and anti-FXa DOAC, anti-thrombin DOAC, UFH, or LMWH, opening therefore an avenue in the search for the Grail of an almost universal anticoagulant antidote.

Example 7: Rivaroxaban and dagigatran dabigatran neutralization by GDFXa-a2M in mice

The ability of GDFXa-a2M to neutralize rivaroxaban in vivo was evaluated in a mouse bleeding model. Mice were force-fed with rivaroxaban or 10 mM HCl. Two hours later 100 μΐ GDFXa-a2M or vehicle were injected in each retro-orbital plexus. Primary end points were BT and BL following lateral tail vein transection. Four groups of 8 to 10 mice were studied. Group I (control) was force-fed with 10 mM HCl and injected with vehicle. Group II and III were force-fed with rivaroxaban and injected with vehicle or GDFXa- α2Μ, respectively. Group IV was force-fed with 10 mM HCL and received GDFXa-a2M to evaluate safety of the antidote. Initial BT and BL were not significantly different between the 4 groups (p > 0.05). While rivaroxaban alone (group II) did not increased initial bleeding, secondary BT (p < 0.001) as well as BL (p < 0.01) increased significantly following the challenges (figure 9). Furthermore compared to group II, GDFXa-a2M administration in group III significantly reduced this increased secondary BT and BL (p < 0.001 and p < 0.01, respectively). In fact, secondary BT and BL of group III were comparable to those observed in group I as well as in group IV (p > 0.05). The ability of GDFXa-a2M to neutralize dabigatran in vivo was evaluated similarly. Two groups (V and VI) of 6 mice were force-fed with dabigatran etexilate. Ninety min later 100 μΐ vehicle (group V) or GDFXa-a2M (group VI) were injected in each retro-orbital plexus. As with rivaroxaban the primary BT and BL were not significantly increased by dabigatran (p > 0.05). Dabigatran etexilate was nevertheless hydrolyzed to dabigatran by mouse esterases as force-feeding increased significantly secondary BT and BL compared to control group I (p < 0.001 and p < 0.05 respectively). Above all GDFXa-a2M administration (group VI) significantly reduced the higher secondary BT and BL recorded for group V (p < 0.05 and p < 0.001 respectively; figure 10). We concluded that GDFXa-a2M effectively neutralized rivaroxaban and dabigatran in vivo. Our data also suggested that GDFXa-a2M alone had little or no adverse effect on mouse haemostasis.

Example 8: Safety and apparent half-life of GDFXa-a2M in mice

Above data suggested that GDFXa-a2M was devoid of in vivo adverse effect. To rule out more precisely the in vivo prothrombotic potential of GDFXa-a2M we compared ex vivo PPP of control mice with those having received GDFXa-a2M using clot waveform analysis as well as measurement of the fibrin d-dimers and TAT complexes formed. Whether or not mice received GDFXa-a2M, LT of clot waveform were comparable (p > 0.05; figure 11). ELISA titration also suggested that amounts of d-dimers and TAT complexes were comparable (p > 0.05; figure 1 1). Finally we evaluated the half-life of GDFXa-a2M in vivo. GDFXa-a2M was injected in each retro-orbital plexus and different time points were collected and residual GDFXa-a2M activity upon time was measured. An apparent Half-life in mice of 85 min (R2 = 0.90) was evaluated for GDFXa-a2M (figure 12). This relatively long half-life precluded evaluation of the potential DOAC elimination by mice because more than half GDFXa-a2M was still present one hour post injection meaning that DO AC titration would reflect the sum of the residual free as well as GDFXa- a2M sequestered DOAC. Nevertheless, we concluded that GDFXa-a2M was safe (devoid of procoagulant properties) and efficiently neutralized rivaroxaban and dabigatran in vivo.

In conclusion, inventors verified in vivo in a mouse bleeding model that GDFXa- a2M had indeed the ability to neutralize rivaroxaban and dabigatran. They also verified that GDFXa-a2M had little or no adverse effect on mouse haemostasis, thus appeared safe in vivo. Finally, they evidenced that, probably due to saturation of the clearance mechanisms, GDFXa-a2M had a relatively long actual persistence in mice (85 min) compared to the expected 2-4 min expected and previously reported (Imber et al. 1981 ; Clearance and binding of two electrophoretic "fast" forms of human a2-macroglobulin. J Biol Chem; 256:8134-9). Accordingly, GDFXa-a2M might still neutralize DOAC several hours post injection thus may constitute a long lasting antidote for DOAC without the need for continuous perfusion or multiple injections.

Accordingly, the GDFXa- a2M complex of the invention constitutes a unique tool to rapidly classify haemostasis deficiencies between pathologic and anticoagulant medication (excluding VKA) through in vitro and ex vivo differential diagnosis. In emergency GDFXa- a2M has the potential of being a last resort almost universal in vivo antidote to most anticoagulants.

SEQUENCE LISTING

SEQ ID N°l (FX light chain):

ANSFLEEMK GHLERECMEE TCSYEEAREV FEDSDKTNEF WNKYKDGDQC ETSPCQNQGK CKDGLGEYTC TCLEGFEGKN CELFTRKLCS LDNGDCDQFC HEEQNSVVCS CARGYTLADN GKACIPTGPY PCGKQTLER

SEQ ID N°2 (FX heavy chain):

SVAQATSSSG EAPDSITWKP YDAADLDPTE NPFDLLDFNQ TQPERGDNNL TRIVGGQECK DGECPWQALL INEENEGFCG GTILSEFYIL TAAHCLYQAK RFKVRVGDRN TEQEEGGEAV HEVEVVIKHN RFTKETYDFD IAVLRLKTPI TFRMNVAPAC LPERDWAEST LMTQKTGIVS GFGRTHEKGR QSTRLKMLEV PYVDRNSCKL SSSFIITQNM FCAGYDTKQE DACQGDSGGP HVTRFKDTYF VTGIVSWGEG CARKGKYGIY TKVTAFLKWI DRSMKTRGLP KAKSHAPEVI TSSPLK SEQ ID N°3 (52 first residues of FX heavy chain released through activation in FXa) SVAQATSSSG EAPDSITWKP YDAADLDPTE NPFDLLDFNQ TQPERGDNNL TR

SEQ ID N°4 (FXa heavy chain): IVGGQECKDG ECPWQALLIN EENEGFCGGT ILSEFYILTA AHCLYQAKRF KVRVGDR TE QEEGGEAVHE VEVVIKHNRF TKETYDFDIA VLRLKTPITF RMNVAPACLP ERDWAESTLM TQKTGIVSGF GRTHEKGRQS TRLKMLEVPY VDRNSCKLSS SFIITQNMFC AGYDTKQEDA CQGDSGGPHV TRFKDTYFVT GIVSWGEGCA RKGKYGIYTK VTAFLKWIDR SMKTRGLPKA KSHAPEVITS SPLK

SEQ ID N°5 (Light chain of Gla domain less FXa (GDFXa) resulting from chymotrypsin digestion):

KDGDQCETSP CQNQGKCKDG LGEYTCTCLE GFEGK CELF TRKLCSLDNG DCDQFCHEEQ NSVVCSCARG YTLADNGKAC IPTGPYPCGK QTLER

SEQ ID N°6 (Precursor Protein of human a2M):

MGKNKLLHPS LVLLLLVLLP TDASVSGKPQ YMVLVPSLLH TETTEKGCVL LSYLNETVTV SASLESVRGN RSLFTDLEAE NDVLHCVAFA VPKSSSNEEV MFLTVQVKGP TQEFKKRTTV MVK EDSLVF VQTDKSIYKP GQTVKFRVVS MDENFHPLNE LIPLVYIQDP KGNRIAQWQS FQLEGGLKQF SFPLSSEPFQ GSYKVVVQK SGGRTEHPFT VEEFVLPKFE VQVTVPKIIT ILEEEMNVSV CGLYTYGKPV PGHVTVSICR KYSDASDCHG EDSQAFCEKF SGQLNSHGCF YQQVKTKVFQ LKRKEYEMKL HTEAQIQEEG TVVELTGRQS SEITRTITKL SFVKVDSHFR QGIPFFGQVR LVDGKGVPIP NKVIFIRGNE ANYYSNATTD EHGLVQFSIN TTNVMGTSLT VRVNYKDRSP CYGYQWVSEE HEEAHHTAYL VFSPSKSFVH LEPMSHELPC GHTQTVQAHY ILNGGTLLGL KKLSFYYLIM AKGGIVRTGT HGLLVKQEDM KGHFSISIPV KSDIAPVARL LIYAVLPTGD VIGDSAKYDV ENCLANKVDL SFSPSQSLPA SHAHLRVTAA PQSVCALRAV DQSVLLMKPD AELSASSVYN LLPEKDLTGF PGPLNDQDNE DCINRHNVYI NGITYTPVSS TNEKDMYSFL EDMGLKAFTN SKIRKPKMCP QLQQYEMHGP EGLRVGFYES DVMGRGHARL VHVEEPHTET VRKYFPETWI WDLVVVNSAG VAEVGVTVPD TITEWKAGAF CLSEDAGLGI SSTASLRAFQ PFFVELTMPY SVIRGEAFTL KATVLNYLPK CIRVSVQLEA SPAFLAVPVE KEQAPHCICA NGRQTVSWAV TPKSLGNVNF TVSAEALESQ ELCGTEVPSV PEHGRKDTVI KPLLVEPEGL EKETTFNSLL CPSGGEVSEE LSLKLPPNVV EESARASVSV LGDILGSAMQ NTQNLLQMPY GCGEQNMVLF APNIYVLDYL NETQQLTPEI KSKAIGYLNT GYQRQLNYKH YDGSYSTFGE RYGRNQGNTW LTAFVLKTFA QARAYIFIDE AHITQALIWL SQRQKDNGCF RSSGSLLNNA IKGGVEDEVT LSAYITIALL EIPLTVTHPV VRNALFCLES AWKTAQEGDH GSHVYTKALL AYAFALAGNQ DKRKEVLKSL NEEAVKKDNS VHWERPQKPK APVGHFYEPQ APSAEVEMTS YVLLAYLTAQ PAPTSEDLTS ATNIVKWITK QQNAQGGFSS TQDTVVALHA LSKYGAATFT RTGKAAQVTI QSSGTFSSKF QVDNNNRLLL QQVSLPELPG EYSMKVTGEG CVYLQTSLKY NILPEKEEFP FALGVQTLPQ TCDEPKAHTS FQISLSVSYT GSRSASNMAI VDVKMVSGFI PLKPTVKMLE RSNHVSRTEV SSNHVLIYLD KVSNQTLSLF FTVLQDVPVR DLKPAIVKVY DYYETDEFAI AEYNAPCSKD LGNA