ANIONIC LINEAR POLYGLYCEROL DERIVATIVES, A METHOD FOR MANUFACTURING AND APPLICATIONS

Abstract

The invention relates to a linear polyglycerol compound, comprising a backbone of linearly linked glycerol residues. This compound is characterized in that it carries a plurality of substituents in the nature of covalently bound sulfates, wherein a degree of substitution of the backbone is preferably between 30 and 100%. A method of manufacturing the compound as well as uses of this compound and similar compounds are also disclosed.

Claims

1. A linear polyglycerol compound, comprising an optionally terminally substituted backbone of linearly linked glycerol units, wherein it carries a plurality of substituents in the nature of sulfates, wherein a degree of substitution of the backbone is between 10 and 100%.

2. The linear polyglycerol compound according to claim 1, wherein the degree of substitution of the backbone is between 30 and 100%.

3. The linear polyglycerol compound according to claim 1, wherein the glycerol residues of the backbone are 1,2-linked or 1,3-linked to each other.

4. The linear polyglycerol compound according to claim 1, wherein it has one the following formulae: ##STR00003## with n=5 to 1000, X, Y=independently from each other any organic residue with a functional group chosen from the group consisting of alcohol, amine, thiol, azide, alkyne, alkene, carboxylic acid, aldehyde, ketone, halogen, isocyanate, isothiocyanate, Michael acceptor/donor group, Z=independently from other residues Z in the same polyglycerol molecule a negatively charged residue in the nature of a sulfate, wherein at least 30% of all residues Z in the polyglycerol bear a negative charge.

5. The linear polyglycerol compound according to claim 1, wherein the backbone has a number average molecular weight of 1 kDa to 100 000 kDa.

6. The linear polyglycerol compound according to claim 1, wherein the backbone has a number average molecular weight of 2 kDa to 6 kDa and that the degree of substitution of the backbone is between 85 and 100%.

7. The linear polyglycerol compound according to claim 1, wherein the backbone has a number average molecular weight of 3 kDa to 5 kDa and that the degree of substitution of the backbone is between 85 and 100%.

8. The linear polyglycerol compound according to claim 1, wherein it comprises an end functionalization with a functional group chosen from the group consisting of azide, bromine, chlorine, iodine, fluorine, primary amine, secondary amine, tertiary amine, carboxylic acids, thiols, disulfides, ketals, acetals, aldehydes, hydroxyl, isocyanate, isothiocyante, unsaturated carbon-carbon units having double or triple bonds, a Michael acceptor, such as an α,β-unsaturated aldehyde, ketone, ester, carboxylic acid amide, carboxylic acid imide, such as maleimide, or an α,β-unsaturated nitrile, and a Michael donator, such as organic copper compounds, amines, thiols, phenolate ions, cyanides and acrylates.

9. The linear polyglycerol compound according to claim 1, wherein it is bound to a surface of an article.

10. (canceled)

11. (canceled)

12. A method for manufacturing a linear polyglycerol compound according to claim 1, comprising the following steps: a) providing a linear polyglycerol compound, comprising a backbone of linearly linked glycerol residues bearing hydroxyl groups or other functional groups chosen from the group consisting of azides, alkynes, alkenes, thiols, halogens, primary or secondary amines, carboxylic acids, aldehydes, ketons and any Michael donor or acceptor for conjugation of anionically charged entities, and b) causing a reaction of at least some of these hydroxyl groups or these other functional groups of the linear polyglycerol compound provided in step a) with a compound introducing a negatively charged group into the linear polyglycerol compound, this compound being at least one chosen from the group consisting of sulfuric acid, sulfonic acid, a sulfuric acid derivative, and a sulfonic acid derivative.

13. A gel, comprising a plurality of linear polyglycerol compounds according to claim 1, wherein it further comprises a crosslinker, wherein statistically each crosslinker molecule is covalently bound to one or more molecules of the linear polyglycerol compound.

14. (canceled)

15. Method for using of a universal heparin binding compound or protamine as antidote for a linear polyglycerol compound comprising an optionally terminally substituted backbone of linearly linked glycerol units, wherein the linear polyglycerol compound carries a plurality of substituents in the nature of covalently bound negatively charged groups chosen from the group consisting of sulfates, sulfonates, phosphates, phosphonates, bisphosphonates, carboxylates and combinations thereof, wherein a degree of substitution of the backbone is between 10 and 100%.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Aspects of the invention will now be explained in more detail by examples and accompanying figures.

[0057] FIG. 1 shows the molecular weight distribution of UFH and LMWH in daltons.

[0058] FIG. 2 shows the mode of action of UFH (A), LMWH (B) and the ATIII specific pentasaccharide unit of UFH (C) in ATIII mediated thrombin (FIIa) and FXa inhibition.

[0059] FIG. 3 shows a typical synthesis scheme for lPGS.

[0060] FIG. 4 shows concentration dependent aPPT plasma clotting times with lPGS of different molecular weight backbone.

[0061] FIG. 5 shows aPTT plasma clotting times of selected lPGS samples at very low concentrations.

[0062] FIG. 6 shows concentration dependent aPTT clotting times of lPGS with constant molecular weight of the backbone (5 kDa and 25 kDa) but varying degree of sulfation (DS).

[0063] FIG. 7 shows aPTT plasma clotting times of dPGS samples with high degree of sulfation but varying molecular weight of the dendritic core.

[0064] FIG. 8 shows representative thromboelastography (TEG) traces of lPG.sub.5kDaS.sub.60 and dPG.sub.6kDaS.sub.64 in fresh whole blood.

[0065] FIG. 9 shows representative TEG curves of whole blood with UFH (2 IU/ml) and lPG.sub.5kDaS.sub.60 (0.05 mg/ml) in the presence and absence of protamine.

[0066] FIG. 10 shows dilute prothrombin (PT) plasma clotting time of lPGS, dPGS, tinzaparin and UFH at a concentration of 0.05 mg/ml and 2 IU/ml, respectively.

[0067] FIG. 11 shows concentration dependent inhibition of FXa by UFH, lPGS and dPGS in the presence of ATIII via activity measurement of FXa in a chromogenic FXa assay.

[0068] FIG. 12 shows concentration dependent inhibition of FIIa by UFH, lPGS and dPGS in the presence of ATIII via thrombin activity measurement in a chromogenic FIIa assay.

[0069] FIG. 13 shows concentration dependent inhibition of FIIa by UFH, lPGS and dPGS in the absence of ATIII via thrombin activity measurement in a chromogenic FIIa assay.

[0070] FIG. 14 shows in vitro cell toxicity of HUVECs after 24 h exposure to lPGS with different molecular weights.

[0071] FIG. 15 shows in vitro cell toxicity of HUVEC's after 24 h exposure of lPGS (backbone=5 kDa) with different degree of sulfation.

[0072] FIG. 16 shows representative synthetic schemes of lPGS with an anchor group for surface immobilization.

[0073] FIG. 17A shows a structure of an lPGS which is well suited for coating surfaces.

[0074] FIG. 17B shows a reaction scheme of a coating process.

[0075] FIG. 18A shows the results of coating experiments in which a glass surface was coated by the lPGS of FIG. 17A.

[0076] FIG. 18B shows the results of coating experiments in which a polystyrene surface was coated by the lPGS of FIG. 17A.

[0077] FIG. 18C shows the results of coating experiments in which a tissue culture polystyrene surface was coated by the lPGS of FIG. 17A.

[0078] FIG. 19 shows a comparison between the anticoagulation effect of lPGS and of a linear polyglycerol phosphate.

[0079] FIG. 20 shows aPPT plasma clotting times of different linear compounds.

DETAILED DESCRIPTION

[0080] So far unpublished data of the inventors indicated that there is no difference between dPGS and lPGS in their anti-inflammatory potential. However, there was an unexpected and significant difference of dPGS and lPGS in terms of their anti-coagulant effect on blood in vitro. Within the following section an aspect of the invention relating to lPGS as a specific linear polyglycerol compound will be explained in detail.

[0081] LPGS is a fully synthetic, structurally defined, non-carbohydrate based polymer. It can be used for anti-coagulant therapy. The lPGS molecules showed surprisingly potent anticoagulant activity similar to unfractionated heparins (UFH) and low molecular weight heparins (LMWH). Unlike heparins they are fully synthetic and homogenous in structure. The in vivo activity and circulation time of lPGS can be adjusted and fine-tuned by the molecular weight which in turn can be precisely controlled via the underlaying anionic polymerization mechanism. Due to their synthetic nature, there will not be any fear of biological contamination (such as contamination by bacteria, viruses, immunomodulatory carbohydrates etc.).

[0082] LPGS is defined as a linear polyether based on a linear polyglycerol backbone which can be synthesized with adjustable molecular weight via controlled anionic polymerization of e.g., a side chain acetal protected glycerol monomer such as ethoxyethyl gycidyl ether (EEGE).[17-18, 23-24] Achieved polydispersities are typically below 1.3 and often lower than 1.1. Control over the end group of the synthesized polymer can be achieved via the applied initiator of the polymerization.[25] In addition, the acetal protecting groups in the side chain of the obtained polymer (PEEGE) allow selective post-functionalization of the terminal hydroxyl functionality. Acidic treatment and hydrolysis of the polymer's acetal side chains generates end group functionalized linear polyglycerol (lPG). Sulfation via e.g. sulfur trioxide pyridine complex yields linear polyglycerol sulfate (lPGS) with adjustable degree of sulfation (DS) analogously to dPGS.

[0083] A suited synthesis sequence for terminally monobromo or monoazido functionalized lPGS is outlined in FIG. 3 which show a suited synthetic scheme for lPGS using monomer activated anionic polymerization of EEGE to PEEGE, acidic hydrolysis to lPG and subsequent sulfation to yield lPGS. Functional or reactive terminal groups are introduced via application of the respective tetraalkyl ammonium salt during polymerization, e.g., to yield terminal monobromo functionalized polymers (A) or terminal monoazido functionalized polymers for use in surface modification (B) (cf. also FIG. 16). Control over the functional end-group of the polymer is simply achieved by applying the respective tetraalkylammonium salt as an initiator of the anionic polymerization. Terminal monoazide functionalized lPGS can immediately be applied for azide/alkyne click chemistry surface immobilization or for surface immobilization by suitable linkers for the generation of anti-thrombotic surfaces (cf. FIG. 16).

[0084] Analytical specifications in terms of molecular weight and polydispersity index (PDI) of synthesized lPGS samples determined via gel permeation chromatography multi-angle light scattering (GPC-MALLS) of their respective lPG backbones and degrees of sulfation, determined via .sup.1H NMR and combustion analysis, are summarized in Table 1. Nomenclature of lPGS and dPGS samples is as such that the molecular weight (MW) of the lPG backbone or dPG core is given in kDa in subscripts after PG and the number of sulfate groups is given in subscripts after S. Accordingly, a sample lPG.sub.5kDaS.sub.45 encodes a lPGS polymer with a lPG backbone of 5 kDa and 45 sulfate groups per molecule which equals a degree of sulfation of DS=67%.

TABLE-US-00001 TABLE 1 Analytical specification of lPGS and dPGS samples as well as commercial UFH and Tinzaparin as control samples for biological evaluation. M.sub.n of number of l/dPG sulfate backbone/ M.sub.n of groups per DS Sample name core PDI sample polymer in % lPG.sub.2 kDaS.sub.17 1.9 1.15 3.6 17 63 dPG.sub.2 kDaS.sub.27 2.0 1.60 4.7 27 100 lPG.sub.3 kDaS.sub.44 3.2 1.09 7.7 44 100 dPG.sub.3 kDaS.sub.6 3.0 1.60 3.6 6 15 dPG.sub.3 kDaS.sub.12 3.0 1.60 4.2 12 30 dPG.sub.3 kDaS.sub.38 3.0 1.60 6.7 38 94 lPG.sub.5 kDaS.sub.0 5.0 1.07 5.0 0 0 lPG.sub.5 kDaS.sub.19 5.0 1.07 7.2 19 28 lPG.sub.5 kDaS.sub.36 5.0 1.07 9.3 36 54 lPG.sub.5 kDaS.sub.60 5.0 1.07 12.1 60 90 dPG.sub.5 kDaS.sub.64 5.0 1.60 11.5 65 100 dPG.sub.6 kDaS.sub.65 6.0 1.60 12.5 64 80 lPG.sub.10 kDaS.sub.135 10 1.09 25.9 135 100 lPG.sub.25 kDaS.sub.94 25 1.11 36.1 94 28 lPG.sub.25 kDaS.sub.162 25 1.11 44.1 162 48 lPG.sub.25 kDaS.sub.337 25 1.11 64.8 337 100 dPG.sub.25 kDaS.sub.88 25 1.33 34.0 88 26 dPG.sub.25 kDaS.sub.145 25 1.33 39.8 145 43 dPG.sub.25 kDaS.sub.306 25 1.33 56.2 306 91 lPG.sub.54 kDaS.sub.625 54 1.31 118 625 85 dPG.sub.120 kDaS.sub.1264 120 1.60 249 1264 78 dPG.sub.250 kDaS.sub.2667 250 1.60 522 2667 79 dPG.sub.480 kDaS.sub.5185 480 1.60 1009 5185 80 UFH n.a. 1.3-1.4* 12-16* Tinzaparin n.a. — 6.5* n.a. not applicable, *data taken from literature[26]

[0085] Within a first screening of the anticoagulant efficiency, the polysulfates were evaluated in an in vitro activated partial thromboplastin time (aPTT) coagulation assay using fresh human platelet poor plasma (PPP) from healthy, consented donors. Via aPTT the intrinsic and common pathway of blood coagulation are tested with PPP which is the standard in vitro assay to monitor patients on heparin therapy. Blood samples were collected in citrated vacutainer tubes from BD (9:1 v/v blood to buffered sodium citrate solution) in order to prevent contact activated coagulation. Sodium citrate complexes Ca.sup.2+ ions in the blood sample which are mandatory for coagulation enzymes to work. Thus, coagulation is hindered in the presence of sodium citrate when no free calcium ions are available. The anticoagulant effect of sodium citrate can easily be reversed by addition of more calcium ions, e.g., added as calcium chloride solution, in order to saturate the calcium complexing sites of citrate and provide free uncomplexed calcium ions for the coagulation enzymes. After blood collection the samples are centrifuged (20 minutes, 1200×g) at room temperature in order to separate cellular components of blood including platelets from plasma. The obtained clear yellow upper phase (PPP) in the tube after centrifugation is separated and used immediately for the aPTT assay applying a ST4 automated hemostasis coagulation analyzer (Diagnostic Stago, Inc.) with mechanical endpoint determination. Therefore, PPP (180 μl) was mixed with polysulfate stock solutions (20 μl) in saline of various concentration at room temperature. Then the anticoagulated plasma was mixed with the partial thromboplastin reagent (Actin®FSL from Dade Behring) in a 1:1 ratio (200 μl each), 100 μl of this mixture was transferred into a well of cuvette strips, respectively and incubated for three minutes at 37° C. Subsequently, coagulation was initiated by addition of prewarmed (37° C.) calcium chloride solution (0.025 mM, 50 μl) into each well and time was recorded until a fibrin clot was formed at 37° C. Coagulation times of the polyglycerol sulfates were evaluated in comparison with saline treated PPP (untreated control) as well as UFH or Tinzaparin, a clinically used LMWH, treated PPP. Concentration dependent clotting times (in seconds) from at least three repeat experiments are graphically illustrated in FIG. 4. Specifically, FIG. 4 shows concentration dependent aPPT plasma clotting times with lPGS of different molecular weight backbones and high degree of sulfation, respectively. UFH and Tinzaparin plasma clotting times are given aside for comparison as well as a saline control. Clotting times are the average of at least three replicate measurements using plasma from different blood donors with the error bars displaying the standard deviation. Concentrations of lPGS are given in mg/ml and concentrations of UFH and Tinzaparin are given in the corresponding amount of international units (IU/mg). Plasma with clotting times >500 seconds is indicated with a clotting time of 500 seconds, but without error bars. Results show that lPGS have similar activity as compared to heparins.

[0086] From FIG. 4 it becomes obvious that lPGS with a linear polyglycerol backbone of 3 kDa and high degree of sulfation (DS=100% equals 44 sulfate groups per molecule, lPG.sub.3kDaS.sub.44) and an overall molecular weight of 7.7 kDa has the same concentration dependent anti-coagulant profile in the aPTT assay as Tinzaparin with a molecular weight of 6.5 kDa. Similarly, lPG.sub.5kDaS.sub.60 with an overall molecular weight of 12.1 kDa exhibits a comparable aPTT profile as UFH, with a similar molecular weight on average but much broader polydispersity. As a general trend the anticoagulant properties of lPGS increase with increasing molecular weight of the backbone linear polyglycerol and constant high degree of sulfation. Thereby, the anticoagulant properties of lPGS remain evident as down to a concentration of 1.5 μg/ml. This is shown for three lPGS samples of various molecular weight of the backbone lPG in FIG. 5 via their respective concentration dependent aPTT times. Although at that low concentrations the anticoagulant properties as measured via aPTT are not strong, they are still detectable in vitro.

[0087] FIG. 5 shows aPTT plasma clotting times of selected lPGS samples at very low concentrations. Presented data are average values of at least triplicates with plasma from varying donors. Error bars illustrate standard deviations. Concentrations of 0 mg/ml correspond to the saline control. Data suggest the potent, concentration-dependent anticoagulation activity of lPGS molecules.

[0088] As illustrated in FIG. 6 the anticoagulant properties of lPGS with constant molecular weight of the backbone but varying number of sulfate groups per polymer, as exemplified with a series on a 5 and 25 kDa lPG backbone, increase with the DS. This clearly demonstrates the higher anticoagulant potential with augmented charge density on the polymer.

[0089] Specifically, FIG. 6 shows concentration dependent aPTT clotting times of lPGS with constant molecular weight of the backbone (5 kDa and 25 kDa) but varying degree of sulfation (DS) in comparison with UFH and Tinzaparin (n.a.=not applicable). Plasma with clotting times >500 seconds is indicated with a clotting time of 500 seconds, but without error bars.

[0090] In accordance with previous results, the anticoagulant properties of dendritic polyglycerol sulfate were found to be weaker than that of UFH for all molecular weights of dPGS evaluated in the aPTT plasma coagulation assay.[13] In FIG. 7 the aPTT times of different dendritic polyglycerol sulfates with a high degree of sulfation (78-100%), respectively, but varying molecular weight of the dendritic core are plotted for a fixed anticoagulant concentration of 0.05 mg/ml dPGS. Plasma with clotting times >500 seconds is indicated with a clotting time of 500 seconds, but without error bars. In contrast to lPGS, for which the anticoagulant effect increases or at least remains static with increasing molecular weight of the backbone (cf. FIG. 4), dPGS shows decreasing aPTT times with increasing molecular weight. At a molecular weight of the dendritic core of 120 kDa almost no anticoagulant properties (comparable to saline control) are found via aPTT for dPGS.

[0091] In order to better mimic the in vivo situation of coagulation whole blood is generally the better test medium than PPP since cellular components of blood, in particular platelets, play a crucial role in the process of clot formation. Thromboelastography (TEG) allows the tracking of blood clot formation and subsequent fibrinolysis of the formed clot from whole blood. In aPTT by default unnaturally fast contact activated blood coagulation times are measured (25-33 seconds for the untreated control) due to the use of aPTT reagents with unphysiological concentrations of phospholipids to mimic the membrane of platelets and ellagic acid activator for contact activation of the intrinsic pathway.[27]

[0092] In contrast, thromboelastography is closer to the physiological conditions since only freshly donated, citrated whole blood is used which is recalcified by the addition of calcium chloride solution in order to initiate clotting. In FIG. 8 representative thromboelastography (TEG) traces of lPG.sub.5kDaS.sub.60 and dPG.sub.6kDaS.sub.64 in fresh whole blood in comparison with UFH and a saline control over the time course of one hour. The anticoagulant concentration was 0.05 mg/ml or 2 IU, respectively. Characteristic parameters like the R value (reaction time from the initiation of coagulation via recalcification to the first indication of clot formation), K value which is a measure for the speed of clot formation (the time from initial clot buildup until it reaches 20 mm), and the shape (□-angle and progression) of the curve as well as the maximum amplitude (MA) which is a measure of clot strength are indicated in FIG. 8. For the saline treated (considered as untreated) control blood a couple of minutes (strongly donor dependent, usually 7-13 minutes) after recalcification a symmetric deviation of the curve from the straight initial time line is observed which indicates the buildup of a blood clot. At some point of the curve the amplitude between the symmetrical curve is not growing any more but starts to decrease again (indicated with the MA arrow), which indicates endogenous fibrinolysis of the formed clot. UFH at a concentration of 2 IU/mg which roughly corresponds to 0.05 mg/ml polyglycerol sulfate showed no indication of clot formation during the time course of one hour in TEG since no typical curve shape development but only a straight line was observed. The same was true for lPG.sub.5kDaS.sub.60 at concentrations of 2 IU/mg which is why the two curves of UFH and lPGS in FIG. 8 overlap. When the TEG experiment was traced longer up to three hours UFH and lPG.sub.5kDaS.sub.60 at that concentration eventually showed initiated clotting with very slow rate of clot formation and very low amplitude, indicating very weak clot strength.

[0093] In strong contrast, dPG.sub.6kDaS.sub.64 at the same concentration revealed clear clot formation in whole blood and typical fibrinolysis after MA is reached (FIG. 8). Comparison of the later curve with the saline control shows that blood coagulation is slightly delayed by a couple of minutes (R value) as is the rate of clot formation (K value, □-angle). Clot strength (MA) with dPGS treated whole blood is a little weaker and the initiation of fibrinolysis is slightly delayed as compared to the saline control.

[0094] Hence, a TEG measurement is particularly well suited to reveal an authentic picture of the anticoagulant potential of lPGS and dPGS. Representative and characteristic parameters of TEG measurements with lPG.sub.3kDaS.sub.44, lPG.sub.5kDaS.sub.60, dPG.sub.6kDaS.sub.64 and UFH in whole blood are summarized in Table 2.

TABLE-US-00002 TABLE 2 Representative and characteristic parameters of lPG.sub.3 kDaS.sub.44, lPG.sub.5 kDaS.sub.60, dPG.sub.6 kDaS.sub.64 and UFH in TEG with whole blood. R value K value MA □-angle sample concentration [min] [min] [mm] [°] saline — 9.4 ± 2.7 2.7 ± 0.7 59 ± 4 54 ± 6  lPG.sub.3 kDaS.sub.44 0.05 mg/ml 98 ± 48 27 ± 12 43 ± 3 20 ± 18 lPG.sub.5 kDaS.sub.60 0.05 mg/ml 154 ± 27  32 ± 11 34 ± 4 5.7 ± 2.1 dPG.sub.6 kDaS.sub.64 0.05 mg/ml 16 ± 3  2.8 ± 0.9 58 ± 8 53 ± 4  UFH 2 IU/ml >120 — — — Tinzaparin 2 IU/ml >120 — — — Stated values are the mean average ± standard deviation of at least two independent measurements with blood from different donors. For UFH no indication of coagulation is observed for up to 120 min, thus measurements were usually stopped after 120 minutes.

[0095] Form this table and the aPTT times given in FIGS. 4 and 7 it becomes obvious that even though lPG.sub.3kDaS.sub.60 and dPG.sub.6kDaS.sub.64 have similar clotting times between 200 and 300 seconds in the aPTT assay at a concentration of 0.05 mg/ml, clotting times and curve profiles from TEG measurements of the two samples in whole blood at the same concentration are vastly different. This clearly demonstrates the importance of TEG measurements in whole blood in order to verify the obtained results from conventional in vitro plasma coagulation assays that require additional activating agents.

[0096] The majority of polysulfated or polyanionic polymers with known anticoagulant properties in vitro, like commercial polyvinyl sulfate sodium (PVS), polystyrene sulfonate sodium (PSS) and others were often identified only by conventional aPTT coagulation assays with PPP and are not very promising for in vivo application due to unknown toxicity profiles or biocompatibility issues.

[0097] Hence, lPGS as a surprisingly effective anticoagulant (similarly effective as Tinzaparin or UFH) in whole blood. It is based on a highly biocompatible linear polyglycerol backbone, thus the chance for biocompatibility issues are low. In fact, all studies concerning hemo- or biocompatibility of lPGS such as red blood cell aggregation, in vitro cell toxicity or a functional platelet assay did not show any indication for concern (unpublished results). The fully synthetic nature of lPGS via a controlled polymerization mechanism make them superior to UFH and LMWHs. It eliminates the risk of disease transmission from animals to humans and yields structurally homogenous polymers with high degree of reproducibility and thus constant, adjustable, and reliable anticoagulant activity.

[0098] With respect to heparin induced thrombocytopenia (HIT) the molecular weight of lPGS or the chain length of the backbone, respectively, should be as short as possible to reduce the risk for HIT. HIT is a fatal immunogenic response which occurs very occasionally after heparin administration. It is induced by the binding of heparin to plasma circulation platelet factor 4 (PF.sub.4) and a subsequent conformational change of PF.sub.4 that leads to antibody recognition. For this conformational change to happen in PF.sub.4, heparin must span the whole tetrameric protein equatorially which in turn requires a certain chain length of the polymer. By reducing the chain length of lPGS as far as possible but still keeping the anticoagulant properties, the risk for HIT development after lPGS administration can be reduced. Thus, out studies focused on lPGS samples with the shortest chain length but still good anticoagulant properties, i.e. lPG.sub.3kDaS.sub.44 and lPG.sub.5kDaS.sub.60.

[0099] A requirement for a safe drug is the existence of a fast and effective antidote that quenches the effect of the drug in case of accidental overdosing or adverse, unwanted response of the patient to the drug. Protamine sulfate is an FDA approved, effective antidote for UFH and can also reverse the action of lPGS. This can be seen from FIG. 9 showing representative TEG curves of whole blood with UFH (2 IU/ml) and lPG.sub.5kDaS.sub.60 (0.05 mg/ml) in the presence and absence of protamine. The plot shows recorded curves up to 150 minutes. Addition of protamine to UFH or lPGS treated whole blood reverses the respective anticoagulant (flat, straight TEG curves) effect and induces regular clot formation similar to the saline control. Below the curve characteristic parameters of these TEG experiments are listed.

[0100] The strongly anticoagulated whole blood (UFH and lPG.sub.5kDaS.sub.60) as indicated by the overlapping flat, straight lines in the TEG trace was brought back into a clotting state after mixing with protamine in a 2:1 or 2.5:1 ratio, respectively. The corresponding TEG traces of the protamine treated anticoagulated whole blood samples show a similar shape as the saline control and hence indicate successful reversal of the anticoagulant effect for both UFH and lPGS. Characteristic parameters of these TEG curves are given in the table 3 below.

TABLE-US-00003 TABLE 3 Characteristic parameters of the TEG curves of FIG. 9. anti- R K coagu- val- val- sample lant ue ue MA □- saline — 11.3 3.3 55.6 48.9 lPG.sub.5 kDaS.sub.60 0.05 mg/ml 140.2 36.9  n.a. n.a. lPG.sub.5 kDaS.sub.60:protamine 0.05 mg/ml 37.9 8.2 49.7 27.4 (2.5:1) UFH 1 IU/ml 191.1 n.a. n.a. n.a. UFH:protamine (2:1) 1 IU/ml 87.7 18.8  42.9  9.5 n.a. = not applicable since the curve shape does not allow the measurement indicates data missing or illegible when filed

[0101] The neutralizing effect of protamine on lPGS can also be verified via aPTT measurements in plasma as shown in Table 4. Similarly to UFH the effective amount of protamine to quench the charges of the polysulfate and thus neutralize the anticoagulant effect strongly depends on the respective donor blood and needs to be determined via titration experiments as demonstrated in Table 4. Protamine when overdosed (overtitrated) has a strong anticoagulant effect itself which is one of the major disadvantageous of protamine as an antidote for heparins in clinical settings.

TABLE-US-00004 TABLE 4 Summary of clotting times of lPG.sub.5 kDaS.sub.60 treated plasma in the presence and absence of protamine via aPTT. anticoagulant AC/P clotting time sample concentration ratio * [s] saline — — 31.2 ± 0.6 lPG.sub.5 kDaS.sub.60 0.05 — >500 lPG.sub.5 kDaS.sub.60:protamine 0.05 2:1 250 ± 10 lPG.sub.5 kDaS.sub.60:protamine 0.05 1:1 111 ± 4  lPG.sub.5 kDaS.sub.60:protamine 0.05 1:2 45.6 ± 1.1 lPG.sub.5 kDaS.sub.60:protamine 0.05 1:4 >500 Stated values are average ± standard deviation from two independent experiments with plasma from different donors. * anticoagulant/protamine ratio

[0102] Table 5 summarizes the neutralizing effect of the universal heparin binding polymer (UHBP) as described in reference [22] on lPGS with and demonstrates its efficiency as antidote for lPGS.

TABLE-US-00005 TABLE 5 Summary of clotting times of lPG.sub.3 kDaS.sub.44 and lPG.sub.5 kDaS.sub.60 treated plasma in the presence and absence of UHBP-1 as a neutralizing agent measured via aPTT. anticoagulant concentration AC/U clotting time Sample [mg/ml] ratio* [s] saline — — 29.7 ± 1.3 lPG.sub.3 kDaS.sub.44 0.05 — 339 ± 57 lPG.sub.3 kDaS.sub.44:UHBP-1 0.05 2:1 104 ± 26 lPG.sub.3 kDaS.sub.44:UHBP-1 0.05 1:1 62 ± 2 lPG.sub.3 kDaS.sub.44:UHBP-1 0.05 1:2 30.2 ± 1.3 saline — — 30.3 ± 1.4 lPG.sub.5 kDaS.sub.60 0.05 — >500 lPG.sub.5 kDaS.sub.60:UHBP-1 0.05 2:1 244 ± 85 lPG.sub.5 kDaS.sub.60:UHBP-1 0.05 1:1  67 ± 12 lPG.sub.5 kDaS.sub.60:UHBP-1 0.05 1:2 30.0 ± 1.2 Stated values are average ± standard deviation from at least three independent experiments (in double determination) with PPP from different donors. *anticoagulant/UHBP ratio

[0103] The used UHBP-1 is based on a 23 kDa dendritic polyglycerol statistically modified by 8 heparin binding groups. These groups consist of tertiary amine clusters which are protonated under physiological pH and hence yield a polymer with 8 positively charged amino clusters. The overall amount of positively charged tertiary amine groups per polymer at physiological pH was 24 for the UHBP which was used in the study summarized in Table 5. The design of this UHBP is as such that an mPEG (450 Da) shell on the surface of the dendritic polyglycerol was installed in order to improve biocompatibility of the positively charged polymer.

[0104] A similar neutralizing effect for the anticoagulant properties of lPG.sub.3kDaS.sub.44 on PPP was observed with a UHBP which had a lower molecular weight of the dendritic core (10 kDa) and lower amount of tertiary amine groups per molecule (7 and 5, respectively). The one corresponding to 7 amine groups denoted as UHBP-2 and the one with 5 amine groups per molecule denoted as UHBP-3. Here, UHBP-2 proofs to be slightly more efficient in its neutralizing efficiency for lPG.sub.3kDaS.sub.44 than UHBP-3. When comparing clotting times of the respective ratios of lPGS to UHBP (Table 5) for UHBP-2 and UHBP-3, the former one consistently yields clotting times closer to the one of the saline control. Although this trend is minimal, it is in line with the composition of the respective UBHP, since UBHP-2 bears more amine groups per polymer than UHBP-3 and hence has higher capacity to quench the anionic charges of lPGS. In accordance with its structural feature of presenting the highest amount of positive charges per molecule among the three UHBPs (Table 5 and 6), UHBP-1 is identified as the most effective neutralizing agent for lPGS in the aPTT assay.

TABLE-US-00006 TABLE 6 Summary of aPTT clotting times of lPG.sub.3 kDaS.sub.44 treated plasma in the presence and absence of UHBP-2 and UHBP- 3 as a neutralizing agent at different ratios. anticoagulant concentration AC/U clotting time Sample [mg/ml] ratio* [s] saline — — 30.7 ± 1.7 lPG.sub.3 kDaS.sub.44 0.05 — 362 ± 70 lPG.sub.3 kDaS.sub.44:UHBP-2 0.05 2:1 122 ± 14 lPG.sub.3 kDaS.sub.44:UHBP-2 0.05 1:1  62 ± 10 lPG.sub.3 kDaS.sub.44:UHBP-2 0.05 1:2 34.8 ± 2.2 saline — — 31.0 ± 1.4 lPG.sub.3 kDaS.sub.44 0.05 — 369 ± 77 lPG.sub.3 kDaS.sub.44:UHBP-3 0.05 2:1 156 ± 21 lPG.sub.3 kDaS.sub.44:UHBP-3 0.05 1:1 97 ± 9 lPG.sub.3 kDaS.sub.44:UHBP-3 0.05 1:2 43.3 ± 3.0 Stated values are average ± standard deviation from at least five independent experiments (in double determination) with PPP from different donors. *anticoagulant/UHBP ratio

[0105] In order to verify the aPTT based results concerning the antidote efficiency and to make sure that UHBP-1 will also neutralize the anticoagulant effect of lPGS in whole blood TEG measurements were performed with lPG.sub.5kDaS.sub.60 treated whole blood and UHBP-1 at different ratios. A summary of the characteristic TEG curve parameters in the presence and absence of UHBP-1 is given in Table 7. First of all the TEG measurements confirm the efficiency and safety of UHBP-1 as an antidote to lPGS also in whole blood. Comparison with analogous data in PPP via aPTT reveal a ratio of lPGS:UHBP-1 of 1:1 as sufficient for complete reversal of the anticoagulant properties of lPG.sub.5kDaS.sub.60 in whole blood, while data obtained from aPTT measurements with PPP would suggest a ratio 1:2 for exhaustive reversal. In addition, this TEG data confirms the previous finding that UHBP is superior to protamine sulfate since overdosing (or overtitrating as for lPGS:UHBP-1 of 1:1.5 and 1:2) did not lead to any anticoagulant effect on whole blood and thus could be the safer antidote in clinical settings.[28]

TABLE-US-00007 TABLE 7 Summary of characteristic and representative TEG curve parameters for lPG.sub.5 kDaS.sub.60 (0.05 mg/ml) treated whole blood in the presence and absence of UHBP-1 with various anticoagulant to antidote ratios. AC/U R value K value MA □-angle Sample ratio* [min] [min] [mm] [°] saline — 9.4 ± 2.7 2.7 ± 0.7 59 ± 4 54 ± 6 lPG.sub.5 kDaS.sub.60 — 154 ± 27  32 ± 11 34 ± 4  5.7 ± 2.1 lPG.sub.5 kDaS.sub.60:UHBP-1 2:1 41 ± 8  14.5 ± 4.8  41 ± 3 15 ± 2 lPG.sub.5 kDaS.sub.60:UHBP-1 1:1 9.0 ± 1.7 2.8 ± 0.9 56 ± 3 52 ± 6 lPG.sub.5 kDaS.sub.60:UHBP-1 .sup. 1:1.5 8.4 ± 0.6 2.6 ± 0.4 61 ± 3 57 ± 3 lPG.sub.5 kDaS.sub.60:UHBP-1 1:2 7.6 ± 0.4 2.2 ± 0.sup.  60 ± 0 60 ± 1 Stated values are average ± standard deviation from at least two independent experiments with whole blood from different donors. *anticoagulant/UHBP ratio

[0106] Blood coagulation is generally described by the cascade like waterfall diagram of coagulation enzymes and zymogens which is divided into the intrinsic or contact activation pathway, the extrinsic or tissue factor pathway and the common pathway.[29-31] Heparins or their anticoagulant efficiency, respectively in blood or plasma is usually tested via the aPTT assay in PPP (see FIG. 3). This assay tests for the intrinsic and common pathway of coagulation. Via the prothrombin time (PT assay) the extrinsic pathway can be analyzed. However, when working with polyanionic anticoagulants it is important to use a thromoplastin reagent for this assay that is sensitive for heparins. Since the latter two in vitro assays are standard in clinics where heparinized blood is commonly analyzed via the aPTT assay, many PT reagents contain additional protamine sulfate. This will neutralize any heparin present in the sample and allow the study of the extrinsic pathway independent of heparin. Thromboplastin reagents contain phospholipids and tissue factor as the active components which initiate the extrinsic pathway of coagulation. Since heparins only interfere with FXa within this pathway only moderate effects of heparins on PT coagulation times are observed, provided that the thromboplastin reagent is sensitive towards heparins, otherwise no effect at all will be detected from heparins. The conventional PT assay activates coagulation with a very high concentration of tissue factor and phospholipids in the thromboplastin PT reagent which is far of any physiological relevance. Hence, typical clotting times in the PT assay are fast compared to clotting times in the aPTT assay. Normal, human PPP from healthy donors usually has clotting times around 10 seconds. In order to mimic more physiological conditions variants of this PT assay have been developed that use diluted thromboplastin agent in order to end up at more physiological concentrations of phospholipids and tissue factor which consequently enhances the observed coagulation times.[32] Such a variation of the PT assay (e.g. 100 times dilution of the thromboplastin agent added) is generally termed dilute PT assay and aims at coagulation times of 30-60 seconds with normal human PPP.

[0107] A compilation of the dilute PT clotting times of lPGS, dPGS, tinazaparin and UFH treated human PPP at a concentration of 0.05 mg/ml and 2 IU/ml, respectively, generated with a heparin sensitive thromboplastin reagent is given in FIG. 10. Stated values are average±standard deviation from at least three independent experiments using whole blood from different donors. LPGS shows a moderate increase in clotting time with increasing molecular weight of the lPG backbone and a constant high degree of sulfation, respectively. Just at high molecular weight of the lPG backbone of 25 kDa and higher a similar effect on blood coagulation as the one from LMWH tinzaparin is observed in the dilute PT assay. In contrast, independent of molecular weight almost no influence on PT clotting time is observed for dPGS treated PPP when compared to the saline control. In strong contrast, UFH with an average molecular weight of 12 kDa exhibits a strong anticoagulant effect in the extrinsic pathway of blood coagulation compared to polyglycerol sulfates. This can be interpreted as a first indication for a different and completely unexpected anticoagulant mechanism or mode of action of lPGS compared to heparins.

[0108] For elucidation of the mode of action of lPGS chromogenic assays with purified enzymes in buffer and a suitable chromogenic substrate for the respective enzyme are used in the presence and absence of the anticoagulant. Thereby, the rate of dye development from enzymatic cleavage of the chromogenic substrate is measured on a UV plate reader at the respective wavelength at fixed concentration of the enzyme, excess substrate and various concentrations of the inhibitor of the enzyme, i.e. the anticoagulant. Since heparins (UFH and LMWH) mainly inhibit FXa and Thrombin (FIIa) in an indirect fashion via activation of circulating, endogenous ATIII during blood coagulation, purified ATIII has to be added in sufficient amounts to the chromogenic assay mixture when heparins efficiency as an inhibitor to FXa and FIIa is assayed. FIG. 11 shows the respective concentration dependent inhibition of FXa with lPGS, dPGS and UFH at a fixed ATIII concentration. lPGS of low and high molecular weight at a high degree of sulfation (lPG.sub.5kDaS.sub.60 and lPG54.sub.kDaS.sub.625), respectively, as well as dPG.sub.5kDaS.sub.65 reveal no effect at all on FXa. In strong contrast, UFH shows a concentration dependent inhibition of FXa as expected. This finding explains the observed differences in PT clotting time between polyglycerol sulfates and UFH, since the former does not inhibit FXa which is the main factor in the extrinsic pathway of blood coagulation.

[0109] FIG. 11 shows the concentration dependent inhibition of FXa by UFH, lPGS and dPGS in the presence of ATIII via activity measurement of FXa in a chromogenic FXa assay. The y-axis on top refers to the UFH concentration in IU/ml the one on the bottom to the corresponding polyglycerol sulfate concentration in mg/ml. Data are average of three replicates±standard deviation.

[0110] A similar assay with thrombin instead of FXa and a different chromogenic substrate with high specificity for thrombin revealed that thrombin is inhibited by lPGS and dPGS to a similar degree as by UFH in a purified buffer system with ATIII added. Compared with TEG and aPTT data in whole blood and PPP this finding suggests that dPGS is not as available as lPGS in plasma which is why in plasma lPGS is a much better anticoagulant than dPGS even though affinity for thrombin in the purified system is comparable and in the same range as heparins ATIII mediated affinity for thrombin.

[0111] The result are depicted in FIG. 12 that shows concentration dependent inhibition of FIIa by UFH, lPGS and dPGS in the presence of ATIII via thrombin activity measurement in a chromogenic FIIa assay. The y-axis on top refers to the UFH concentration in IU/ml the one on the bottom to the corresponding polyglycerol sulfate concentration in mg/ml. Data are average of three replicates±standard deviation.

[0112] An alternative scientific explanation for the differences in anticoagulation of lPGS and dPGS could be an additional target within the coagulation cascade that is only addressed by the linear but not the dendritic polyglycerol sulfate and hence results in the vastly different anticoagulant efficiency of the latter two in vitro. The thrombin activity in the presence of UFH, lPGS and dPGS in an identical chromogenic assay as described above but without addition of ATII to the purified system yields the expected result for UFH as illustrated in FIG. 13. Specifically, FIG. 13 shows concentration dependent inhibition of FIIa by UFH, lPGS and dPGS in the absence of ATIII via thrombin activity measurement in a chromogenic FIIa assay. The y-axis on top refers to the UFH concentration in IU/ml the one on the bottom to the corresponding polyglycerol sulfate concentration in mg/ml. Data are average of three replicates±standard deviation.

[0113] Since heparins require ATIII in order to inhibit FXa and FIIa no effect on thrombin activity is observed for UFH in the absence of ATIII. Surprisingly, identical thrombin inhibition is observed for lPGS and dPGS in the presence and absence of ATIII which indicates that the anticoagulant effect of polyglycerol sulfates is due to a direct inhibition of thrombin rather than an indirect ATIII mediated heparin analog inhibition mechanism.

[0114] This mechanistic hypothesis for lPGS and dPGS is supported by the measured K.sub.d values for thrombin and ATIII via microscale thermophoresis (MST) measurements, a label free technique for the determination of dissociation constants in solution. The experiments were performed with a Monolith NT.LabelFree instrument (NanoTemper, Munich, Germany) which allows to determine dissociation constants in solution. Since this device can only detect K.sub.d values in the medium affinity range but not very high and very low K.sub.d values in the high mM range were assigned as not binding (n.b.) in Table 7. As expected non-sulfated linear and dendritic polyglycerol does not exhibit any affinity for the blood coagulation enzymes ATIII or thrombin, which is in line with their perfect haemocompatibilty as published earlier.[16-17] Also in agreement with the data shown in FIG. 11 lPGS and dPGS does bind to FIIa, while lPGS (high nM range) shows a much higher affinity for Thrombin than dPGS (high nM vs. low μM range). In contrast, heparin shows a significantly higher affinity to ATIII.

TABLE-US-00008 TABLE 8 K.sub.d values of lPG.sub.5 kDaS.sub.60, dPG.sub.5 kDaS.sub.65 and UFH (14 kDa) as well as for the non-sulfates controls lPG.sub.5 kDa and dPG.sub.5 kDa for purified human thrombin and ATIII at 25° C. in Dulbeccos phosphate buffered saline, pH 7.4. Kd UFH lPG.sub.5 kDaS.sub.60 lPG.sub.5 kDa dPG.sub.5 kDaS.sub.60 dPG.sub.5 kDa ATIII 0.25 ± 0.15 n.b. n.b. n.b. n.b. [μM] FIIa 23 ± 5  0.9 ± 0.5 n.b. 9 ± 8 n.b. [μM] n.b. = no binding or > mM

[0115] In order to evaluate the cellular compatibility of lPGS dependent on the molecular weight (FIG. 14) and the degree of sulfation (FIG. 15), an MTS cytotoxicity assay with human umbilical vein endothelial cells (HUVECs) was conducted. This calorimetric assay is based on the enzymatic conversion of a dye and quantitatively describes the number of viable cells present. As can be seen in FIG. 14, up to a concentration of 2.5 mg/mL the cell viability is, independent of the molecular weight, not affected (≧80%). At a concentration of 5 mg/mL a reduction of the viability to 60% was found, however, the highest concentrations applied in this assay are not relevant for the application of lPGS. Specifically, FIG. 14 shows normalized values on in vitro cell toxicity of HUVECs after 24 h exposure to lPGS with different molecular weights. 100%: media only. Negative control: DMSO. Furthermore, the cell viability is independent (≧80%) of the degree of sulfation (5 kDa scaffold) up to a concentration of 2.5 mg/mL (FIG. 15) which confirms the good cell compatibility of lPGS.

[0116] FIG. 15 shows normalized values on the in vitro cell toxicity of HUVEC's after 24 h exposure of lPGS (backbone=5 kDa) with different degree of sulfation.100%: media only. Negative control: DMSO.

[0117] The in vivo efficacy of lPGS was proven for lPG.sub.3kDaS.sub.44 and lPG.sub.5kDaS.sub.60 by the determination of the aPTT clotting time of platelet poor plasma (PPP) samples obtained from rats after i.v. bolus injection of 0.0625 mg/mL at indicated time points (Tables 9 and 10). The given data are from PPP of freshly drawn blood which were measured at the same day. As can be seen from Table 8, a prolongation of the aPTT in the first hour after i.v. administration was observed when treated with lPG.sub.3kDaS.sub.44. Reduction of the aPTT to control levels (−15 min.) was observed 4 hours after the injection. In contrast, lPG.sub.5kDaS.sub.60 had a more pronounced anticoagulant effect for longer than one hour (Table 10). These profiles clearly show that a reproducible anticoagulant effect is obtained in vivo and that the temporal evolution can be influenced by the molecular weight of the lPGS.

TABLE-US-00009 TABLE 9 aPTT of platelet poor plasma at given time points after i.v. injection of lPG.sub.3 kDaS.sub.44. aPTT [s] aPTT [s] aPTT [s] aPTT [s] Time [min] Animal 1 Animal 2 Animal 3 Animal 4 −15 21.2 28.4 26.9 25.4 5 >500 206.7 104.7 228.5 10 202.3 146.0 162.0 175.7 30 91.5 52.5 71.3 54.2 60 30.0 32.2 46.2 — 240 18.3 20.5 18.7 19.3 −15 minutes refers to control plasma drawn 15 minutes prior to bolus injection. n.a. = data not available (clotting of blood during blood draw or clotting before reconstitution even without Ca addition).

TABLE-US-00010 TABLE 10 aPTT of platelet poor plasma at given time points after i.v. injection of lPG.sub.5 kDaS.sub.60. aPTT [s] aPTT [s] aPTT [s] aPTT [s] Time [min] Animal 1 Animal 2 Animal 3 Animal 4 −15 19.6 15.0 26.9 25.4 5 >500 >500 >500 >500 10 >500 >500 >500 397.6 30 496.4 406.5 >500 192.1 60 116.0 113.5 163.1 110.1 240 14.8 14.5 — — −15 minutes refers to control plasma drawn 15 minutes prior to bolus injection. n.a. = data not available (clotting of blood during blood draw or clotting before reconstitution even without Ca addition).
Surface Immobilization of lPGS

[0118] Preventing the activation of the coagulation pathway and hence thrombus formation is a major issue in many medical applications where devices come in contact with blood. However, coating of a material with a protective layer that can be optimized with respect to its anticoagulant properties still remains an unsolved problem. In addition, the device to be functionalized can be made up of a variety of materials which requires a universal approach to efficiently coat the device. Compared to non-covalent approaches where an anticoagulant layer, for example heparin, is physically adsorbed to the surface, covalent immobilization prevents it from being removed, thereby entering the bloodstream, and possibly exhibiting uncontrolled biological activity. In case of covalently immobilized heparin it is believed that the undefined linkage to the surface influences the anticoagulant activity to a major part. Since heparin is poorly homogeneous and has multiple reactive groups at different sites of the polymer only a statistical functionalization of surface immobilizing moieties (anchors) can be realized and by this the anticoagulant properties can hardly be adjusted.

[0119] In contrast, lPGS bearing one single functional group, such as an azide function, located at the terminus of the polymer chain, can be used to conjugate one single anchor to the end of the polymer. This is shown in FIG. 16 depicting representative synthesis schemes of lPGS with an anchor group for surface immobilization. The terminal azide of the lPGS allows mono functionalization of the polymer scaffold (FIG. 16A). Conjugation is realized by azide-alkyne “click” chemistry with either Cu(I) as catalyst (R═H) or catalyst-free by strain promoted cyclooctyne-alkyne click conjugation. The gold surface is representative only.

[0120] By this, control over the localization of the anchor, constitution and length of the linker, as well as the type of anchor which determines the surface to be functionalized, is guaranteed. FIG. 16B representatively illustrates how a gold surface is covalently functionalized with lPGS by using a dithol as the anchor group. This modular approach allows the design of the linker and lPGS separately and offers remarkable advantages compared currently applied covalent and non-covalent approaches. The material to be functionalized can be either a macroscopic material like a medical device or a particle and the surface can be made of any material. The most common materials are glass, plastic, or metal.

[0121] The above described immobilization of lPGS was established on gold nanoparticles in order to confirm the anticoagulant properties of a modified surface via TEG.

[0122] In summary, lPGS is a fully synthetic, structurally homogenous polymer with high anticoagulant properties. Due to the fully synthetic nature, and the anionic polymerization method molecular weights can easily be adjusted to the desired range from a couple of 150 Da to several 100,000 kDa with control over the end group and low polydispersity. End-group control with reactive entities at the terminal side of the linear polymer chain allows covalent modification of surfaces with lPGS for the fabrication of anti-thrombotic surfaces. In addition the fully synthetic nature of lPGS eliminates the risk of disease transmission from animals. The structural homogeneity of lPGS guarantees predictability and adjustability of the anticoagulant properties via the molecular weight (chain length of the lPG precursor) and the degree of sulfation. Moreover, in contrast to fondaparinux or LMWHs a 100% effective and FDA approved antidote for lPGS is available. The short chain length at which lPGS shows already high anticoagulant properties in whole blood are promising to eliminate adverse immune response from PF.sub.4 binding.

[0123] Applications of lPGS are in the clinical, therapeutic and diagnostic area as new fully synthetic and thus safer, structurally homogeneous anticoagulants for in vitro and in vivo use. Opportunities are emerging as a new drug but also as a tool for in vitro blood diagnostics. In addition, the convenient chance for end group functionalization of these polymers offer possibilities for chemical, covalent surface modification with these polysulfates for the generation of anti-thrombotic surfaces on implants and devices such as for blood storage bags, catheters, blood pumps (artificial hearts), blood collection vials and anti-thrombotic blood diagnostic single-use plastic, metal or glass ware or use in regenerative medicine in the delivery of drugs or growth factors by surface grafted linear polysulfates.

[0124] FIG. 17A shows the structure of an lPGS compound that is well suited for such coating applications. This lPGS compound is a terminally substituted lPGS, namely an lPGS-amine compound comprising poly allyl glycidyl ether (PAGE) blocks carrying amine groups.

[0125] These amine groups were used to covalently bond the lPGS to a layer of polydopamine (PDA) previously applied to a surface. An according reaction scheme is depicted in FIG. 17B. The surface coating by PDA was carried out according to a method described in literature (O. Pop-Georgievski et al, Langmuir 2012, 28, 14273-83). “MOPS” stands for 3-(N-morpholino)propanesulfonic acid, and “r.t.” stands for room temperature.

[0126] Three different surfaces (namely, glass, polystyrene, and tissue culture polystyrene) were used for the coating experiments by the lPGS illustrated in FIGS. 17A and 17B. The results of these coating experiments are depicted in FIGS. 18A to 18C. A glass surface activated by piranha solution (piranha-activated glass surface) was used as control glass surface. Piranha solution is a mixture of sulfuric acid and hydrogen peroxide and is used to clean organic residues off substrates. The mixture is a strong oxidizing agent. It will remove most organic matter, and it will also hydroxylate most surfaces, making them highly hydrophilic.

[0127] The characterization of the surfaces was performed by measuring the water contact angle (CA) before the coating process, after coating with polydopamine (PDA) or after additional coating with lPGS. Thereby, two replicates were analyzed per surface (S1/S2) with n=3 CA per surface. Since a contact angle of below 10° cannot be measured precisely, contact angle values below 10° are indicated with a downward facing arrow in FIGS. 18A to 18C.

[0128] FIG. 18A shows the results of coating a glass surface. LPGS reduces the water contact angle in a similar way like piranha solution, but even stronger than piranha solution. The reduction of the contact angle with respect to a PDA coated surface is even stronger. Thus, lPGS is very well suited to hydrophilize a surface.

[0129] FIG. 18B shows the results of coating a polystyrene surface. While an untreated polystyrene surface has a water contact angle of approximately 80° (strongly hydrophobic), this water contact angle can be decreased by a PDA coating to values between 30° to 40°. In contrast, lPGS is able to reduce the water contact angle of values below 10°. Thus, lPGS is also well suited to hydrophilize a polystyrene surface.

[0130] The same holds true for a surface of tissue culture polystyrene. According results are depicted in FIG. 18C. Untreated tissue culture polystyrene surface has a water contact angle of approximately 65°. This water contact angle can be decreased by a PDA coating to approximately 35°. However, with an lPGS coating, the water contact angle is decreased to values below 10°. Thus, an lPGS coating makes a tissue culture polystyrene surface highly hydrophilic.

[0131] The stability of an according lPGS coating was tested on lPGS-coated glass surfaces by long-term incubation in different aqueous media.

[0132] In a first experiment, the lPGS-coated surface was incubated in H.sub.2O for 7 days and the water contact angle was measured afterwards.

[0133] In a second experiment, the same lPGS-coated surface was incubated in phosphate-buffered saline (PBS) for 7 days and the water contact angle was measured afterwards.

[0134] The results are depicted in the following table 11. Thereby, the number of measured droplets is given by n.

TABLE-US-00011 TABLE 11 Water contact angles of lPGS-coated glass surface Immediately after coating After 7 days in H.sub.2O After 7 days PBS <10°(n = 3) <10°(n = 3) <10°(n = 3)

[0135] These stability experiments show that the lPGS coating is very stable, even after prolonged incubation in an aqueous medium. Therewith, the lPGS coating is well suited to modify and therewith hydrophilize different surfaces.

[0136] FIG. 19 shows the anticoagulation effect of lPGS in comparison to a linear polyglycerol phosphate. Specifically, FIG. 19 depicts the aPTT coagulation time of human platelet poor plasma with linear polyglycerol sulfate LPG.sub.25kDaS.sub.337 and linear polyglycerol phosphate LPG.sub.25kDaP.sub.280 added at 0.05 and 0.005 mg/ml final concentration in plasma.

[0137] While the linear polyglycerol phosphate LPG.sub.25kDaP.sub.280 does not show any significant anticoagulative effect at the two concentrations, LPG.sub.25kDaS.sub.337 is able to strongly increase the coagulation time in the aPTT assay (for details of this assay, see above) already at a concentration of 0.005 mg/ml. At a concentration of 0.05 mg/ml the coagulation time is strongly increased by a factor of approximately 15 with respect to saline control or LPG.sub.25kDaP.sub.280, respectively. This data clearly shows the superiority of sulfate as substituent of the linear polyglycerol over other negatively charged groups such as phosphate.

[0138] The superiority of linear sulfated polyglycerols with respect to other linear sulfated compounds regarding coagulation inhibition can be seen in the results from an aPTT assay depicted in FIG. 20.

[0139] Sulfated polyserine (degree of sulfation dS=100%) has been subjected to an aPTT assay in PPP (details see above) in comparison to two different linear polyglycerol sulfates and to a saline control (negative control).

[0140] The non-sulfated precursor molecule polyserine has a molecular weight of approximately 2 kDa and can be described by the following formula:

##STR00002##

[0141] Upon sulfatation, the molecular weight increases to approximately 4 kDa.

[0142] FIG. 20 shows the aPTT coagulation times of PPP in the presence of sulfated polyserin, LPG.sub.2kDaS.sub.17 and LPG.sub.3kDaS.sub.44 at concentrations of 0.005 and 0.025 mg/ml with saline as negative control.

[0143] LPG.sub.2kDaS.sub.17 already shows in concentrations of 0.005 mg/ml an anti-coagulant effect with respect to the negative control. The anti-coagulant effect is even stronger for LPG.sub.3kDaS.sub.44. For both lPGS, the anti-coagulant effect increases with an increase in concentration to 0.025 mg/ml. In contrast, linear sulfated polyserine does not show an anti-coagulant effect. The coagulation time of PPP is not altered by sulfated polyserine, regardless if being used at a concentration of 0.005 mg/ml or 0.025 mg/ml.

[0144] The results of this experiment clearly show that a linear structure of a molecule and multivalently presented sulfate groups as such are not sufficient to obtain a compound having heparin-like properties. However, the claimed linear polyglycerol sulfates surprisingly have such heparin-like anti-coagulative properties and are well suited to be used as coagulation inhibitors.

[0145] But even in case of promising results of an aPTT assay indicating a good coagulation inhibition in PPP, an effective coagulation inhibition in whole blood can often not be foreseen by a person skilled in the art. In contrast to dendritic polyglycerol sulfates, linear polyglycerol sulfates do also show a favorable coagulation inhibition of whole blood, as shown in FIGS. 8 and 9. Therewith, the surprising properties of the claimed linear polyglycerol sulfates were not derivable by theoretic considerations but only by experiments such as those performed by the inventors.

PRIOR ART REFERENCES CITED IN THE PRESENT DOCUMENT

[0146] [1] J. Hirsh, S. S. Anand, J. L. Halperin, V. Fuster, Circulation 2001, 103, 2994-3018. [0147] [2] R. De Caterina, S. Husted, L. Wallentin, G. Agnelli, F. Bachmann, C. Baigent, J. Jespersen, S. D. Kristensen, G. Montalescot, A. Siegbahn, F. W. A. Verheugt, J. Weitz, Eur. Heart 1 2007, 28, 880-913. [0148] [3] J. W. Wilson, J. Extra-Corp. Technol. 1974, 6, 207-213. [0149] [4] I.-K. Jang, M. J. Hursting, Circulation 2005, 111, 2671-2683. [0150] [5] G. P. Visentin, M. Moghaddam, S. E. Beery, J. G. McFarland, R. H. Aster, J. Lab. Clin. Med. 2001, 138, 22-31. [0151] [6] D. Hawkins, J. Evans, Expert Opin. Drug Saf. 2005, 4, 583-590. [0152] [7] M. Pai, M. A. Crowther, Handb. Exp. Pharmacol. 2012, 207, 265-277. [0153] [8] A. S. Fauci, E. Braunwald, D. L. Kasper, S. L. Hauser, D. L. Longo, L. J. Jameson, J. Loscalzo, Harrison's Principles of Internal Medicine, 17th Edition ed., Mc Graw Hill Medical, 2008. [0154] [9] J. Hirsh, R. Raschke, CHEST J. 2004, 126, 188S-203S. [0155] [10] M. Di Nisio, S. Middeldorp, H. R. Buller, New Engl. J. Med. 2005, 353, 1028-1040. [0156] [11] S. M. Nimjee, C. P. Rusconi, R. A. Harrington, B. A. Sullenger, Trends Cardiovas. Med. 2005, 15, 41-45. [0157] [12] D. Coombe, W. Kett, in Heparin—A Century of Progress, Vol. 207 (Eds.: R. Lever, B. Mulloy, C. P. Page), Springer Berlin Heidelberg, 2012, pp. 361-383. [0158] [13] H. Türk, R. Haag, S. Alban, Biocon. Chem. 2004, 15, 162-167. [0159] [14] M. Calderon, M. A. Quadir, S. K. Sharma, R. Haag, Adv. Mater. 2010, 22, 190-218. [0160] [15] R. K. Kainthan, D. E. Brooks, Biomaterials 2007, 28, 4779-4787. [0161] [16] R. K. Kainthan, S. R. Hester, E. Levin, D. V. Devine, D. E. Brooks, Biomaterials 2007, 28, 4581-4590. [0162] [17] R. K. Kainthan, J. Janzen, E. Levin, D. V. Devine, D. E. Brooks, Biomacromolecules 2006, 7, 703-709. [0163] [18] M. Weinhart, I. Grunwald, M. Wyszogrodzka, L. Gaetj en, A. Hartwig, R. Haag, Chem.—Asian J. 2010, 5, 1992-2000. [0164] [19] R. Haag, M. Wyszogrodzka, M. Weinhart, Production of linear, methylated polyglycerol derivatives for functionalization of surfaces in medical applications 2007, DE 102006027125. [0165] [20] J. Dernedde, A. Rausch, M. Weinhart, S. Enders, R. Tauber, K. Licha, M. Schirner, U. Zuegel, A. von Bonin, R. Haag, P. Natl. Acad. Sci. USA 2010, 107, 19679-19684. [0166] [21] R. Haag, J. Dernedde, R. Tauber, B. Gesche, S. Enders, H. Weinhart, Dendritic polyglycerol sulfates and sulfonates and their use for inflammatory diseases 2008, WO2007-EP6889. [0167] [22] J. N. Kizhakkedathu, R. A. Shenoi, C. J. Carter, D. E. Brooks, Polymers for reversing heparin-based anticoagulation 2012, U.S. Pat. No. 8,637,008 B2. [0168] [23] M. Gervais, A.-L. Brocas, G. Cendejas, A. Deffieux, S. Carlotti, Macromol. 2010, 43, 1778-1784. [0169] [24] M. Gervais, A. Labbe, S. Carlotti, A. Deffieux, Macromol. 2009, 42, 2395-2400. [0170] [25] A. Thomas, S. S. Müler, H. Frey, Biomacromolecules 2014, 15, 1935-1954. [0171] [26] E. Gray, B. Mulloy, T. W. Barrowcliffe, Thromb. Haemostasis 2008, 99, 807-818. [0172] [27] P. E. Bock, K. R. Srinivasan, J. D. Shore, Biochemistry 1981, 20, 7258-7266. [0173] [28] R. A. Shenoi, M. T. Kalathottukaren, R. J. Travers, B. F. L. Lai, A. L. Creagh, D. Lange, K. Yu, M. Weinhart, B. H. Chew, C. Du, D. E. Brooks, C. J. Carter, J. H. Morrissey, C. A. Haynes, J. N. Kizhakkedathu Sci. Transl. Med. 2014 6, 260ra150. [0174] [29] E. W. Davie, J. Biol. Chem. 2003, 278, 50819-50832. [0175] [30] E. W. Davie, O. D. Ratnoff, Science 1964, 145, 1310-1312. [0176] [31] K. A. Tanaka, N. S. Key, J. H. Levy, Anesth. Analg. 2009, 108, 1433-1446. [0177] [32] T. Liu, C. D. Scallan, G. J. Broze, Jr., S. Patarroyo-White, G. F. Pierce, K. W. Johnson, Thromb. Haemostasis 2006, 95, 68-76.