NL8701914A - Adsorbents, esp. for blood clotting factors - comprising, amino-butyl or amino-pentyl polymer derivs. - Google Patents

Abstract

Adsorbents (I) comprise prim., sec. tert. or quat. amino ligand gps. linked to a polymer carrier via spacer gps. of formula (CH2)n, where n=4 or 5. The carrier is a polymer contg. OH gps., esp. agarose, and is pre-activated with 1,1'-carbonyldiimidazole (CDI) to form CONH bridging gps. between the polymer and the (CH2)n gps. The ligand gps. are NH2. The ligand density is more than 30 (esp. 50-70) micromoles per ml of the swollen polymer.

Description

• è -1- VO 9196

Adsorbent material and use thereof for the isolation of coagulation factors.

The invention relates to an adsorbent material, which will also be referred to herein by the terms affinity matrix, hydrogel or gel, which adsorbent material comprises a polymeric support material to which ligands are bonded consisting of terminal primary, secondary, tertiary or quaternary amino groups.

The invention also relates to a method for isolating coagulation factors, such as in particular the blood coagulation factors FII, FVII, FVIII, FIX, FX and the Von Willebrand factor (vWF, also referred to in the literature as FVIIIR) from a liquid starting material. such as blood plasma, plasma products, cell lysates and fermentation or culture media by liquid chromatography.

The blood of humans and animals contains small amounts of various substances that are involved in the process of blood coagulation or blood clotting. For various reasons, there is a need for methods by which these substances can be isolated from the blood (or other suitable starting material) in the highest possible yield and highest purity. For example, e.g. the protein plasma VIII occurring in blood plasma (also referred to in the literature as FVIII), one of the thirteen clotting factors necessary for the proper functioning of blood clotting. In about 1000 people (haemophilia A patients) in the Netherlands, this coagulation factor is completely or partially absent or not functional, so that their blood cannot clot normally. Treatment of this bleeding disorder is possible with factor VlII preparations 8701914-2-formates, which can be prepared from human donor blood. Because the yields of the existing preparation methods are very low (<^ 40%) and the demand for preparations is increasing, there is a shortage of factor VIII all over the world.

Various methods are already known by which FVIII and / or other clotting factors and proteins can be isolated from a blood-based starting material by liquid chromatography.

Some of these known methods use conventional anion exchangers for this purpose, such as DEAE cellulose and DEAE-Sephadex, others use certain polyelectrolytes and still others use modified Sepharoses. These various known methods will be discussed in more detail below. Usual anion exchangers

Conventional anion exchangers such as DEAE-cell-lulose and DEAE-Sephadex were used several times in the purification of factor 20 VIII concentrates. These are (more or less outdated) matrices to which a substituted diamine (diethyl-aminoethylamine) is coupled using known, frequently used coupling methods. However, the use of these materials for factor VIII purification and isolation has not continued due to a number of disadvantages of these gels, so that they have been used little in recent years.

literature: Thromb. Diath. Haemorrh. 3 (1959) 572; 4 (1960) 211; 6 (1961) 282; 16 (1966) 738-751.

30 Polyelectrolytes EMA PE-5 and PE-100

The EMA polyelectrolytes introduced by Johnson in the 1970s also consist of a polymeric material to which a diamine (dimethylamino-propylamine) is coupled and therefore also belong to the anion exchangers. The ligand density is quite high and differs per polyelectrolyte. The chosen 8701914 «·» -3- matrix (a copolymer of ethylene and maleic anhydride) is quite unusual. Essential in the chemical structure is the N, N-dimethylaminopropylamide group, which is partially positively charged at 5 pH 6-7 due to its weak basic character (pKa = 5.4). Two different polyelectrolytes are used: 1) PE-100 - in which substantially 100% of the maleic anhydride units are substituted with the amine.

At pH 8.0 the clotting factors 10 II, VII, IX and X can hereby be adsorbed from plasma and eluted after washing the polyelectrolyte at pH 6 with 1.5 M sodium chloride. The supernatant is contacted with PE-5.

2) PE-5 - wherein 5% is substituted.

At pH 6, the FVIII: C is adsorbed while vWF remains in solution. FVIII: C can be eluted after washing the PE-5 at pH 7.4 with 1.0 M lysine / 1.5 M sodium chloride in 43% yield and 52-fold purification. By re-contacting the supernatant (other conditions) with PE-100 (pH 6.0), vWF can also be isolated with 43% yield and 71-fold purification.

From the experiments it was concluded that FVIII initially binds to the MA PE by electrostatic interaction, after which hydrophobic interaction enhances the binding. This makes elution more difficult, so that the yield is relatively low. Furthermore, it appears that lysine has a stabilizing effect, that the yield can be increased by adding heparin and that the Hepatitis B virus titer is lowered during the process.

Since neither the blood contact properties nor the mechanical properties of these EMA polyelectrolytes are good (coagulation activation, high nonspecific adsorption, poor flow and difficult suspension), they have mainly been used for the isolation 8701914 19 * -4- from pig plasma or the purification of cryoprecipitates. literature: J. Lab. Clin. Med. 92 (1978) 194-210; US-A-4,397,841; US-A-4,471,112.

Modified Sepharoses 5 The AH-Sepharose (aminohexyl-Sepharose) of the general formula Sepharose-0-C (= NH) -NH- (CH2) 6 "NH2 was initially introduced by Pharmacia for the immobilization of proteins by covalent bonding of the amine with carboxyl groups of the protein Due to the strongly basic nature of this primary amine (pKa = 9.9), the gel can also be classified among the anion exchangers and behave as such.

For the separation of factor VIII, the latter is apparent from a study published by Austen. With 8.6 ml of AH-Sepharose 4B, the FVIII can be purified from FVIII concentrate (600 IU) 70 times (1.8 IU / mg) in about 35% yield. Adsorption takes place in 0.1 M acetate / lysine pH 5.5 and elution in this buffer with 0-5% sodium chloride. A separation between FVIII: C and vWF can be achieved. The AH-

Sepharose was prepared by coupling 1,6-hexanediamine to CNBr-activated Sepharose 4B by the Axen method to achieve a ligand density of about 7 µmol / ml. The binding of FVIII to the gel appears to be electrostatic, although additional hydrophobic interaction and steric effects cannot be excluded. It was also found that the Hepatitis B virus titer after purification was 1/60 of the initial value.

Simultaneously, butyl 30 Sepharose was tested by Vukovich et al. With the general formula

Sepharose-0-C (= NH) -NH- (CH 2) 4 -H. This was prepared by coupling 1-butanamine to CNBr-activated Sepharose 4B. Plasma (10 ml) was diluted 5 times in 20 mM tris / 1 mM citrate buffer, pH 7.4 and applied to the butyl Sepharose column 35 (20 ml). The clotting factors II, V, VII, VIII, IX and X were hereby bound quantitatively. Elution 8701914-5% took place * in the buffer with 0-2 M sodium chloride.

FVIII eluted at 0.3-0.6 M sodium chloride in about 60% yield and 23-fold purification, separated from the other clotting factors. Based on the results 5, hydrophobic interaction of FVIII with butyl-Sepharose was concluded and electrostatic interaction was completely excluded.

Subsequently, Morgenthaler conducted systematic investigations to elucidate the binding character of FVIII with various substituted Sepharoses.

A homologous series of aminoalkyl Sepharoses was made and tested under the same conditions (pH 5.5) as those of Asten. A series of alkyl Sepharoses was also prepared, which was tested under both these and hydrophobic interaction conditions. Both series were prepared by coupling the respective amine to CNBr-activated Sepharose 4B.

More FVIII: C and protein were bound to the aminoalkyl Sepharoses with increasing chain length and eluted at higher sodium chloride concentrations.

No FVIII: C could be eluted from aminoheptane and aminononane Sepharose. The adsorption on the alkyl Sepharoses did not differ significantly from that on the aminoalkyl Sepharoses. With an ethylene glycol gradient 25 (0-50%), only little protein, but no FVII: C

are eluted. By increasing the salt concentration, some protein could be eluted from the shorter chains with FVIII: C, while the longer chains failed. Hydrophobic interaction was thought to be responsible for the adsorption, while the series of alkyl-Sepharoses even completely excluded the electrostatic interaction.

Both Vukovich and Morgenthaler assumed that the interaction of FVIII with the alkyl Sepharoses used is based on hydrophobic interaction. This is in contradiction with the fact that they could effect elution at elevated sodium chloride concentration.

87 01 914 «o -6-

In the meantime, both literature and our own research have shown that electrostatic interaction with the alkyl Sepharoses prepared via CNBr activation cannot be ruled out, because the coupling of the amine to the CNBr activated Sepharose results in an isourea bond. , which is weakly basic (pKa = 8).

In summary, it can be stated that FVIII: C interacts with (amino) alkyl ligands, but that the character of this interaction is not entirely clear. Furthermore, it is unknown whether the stronger binding to longer chains is caused by the fact that these chains are not only longer hydrophobic, but only because they are more accessible to the large FVIII: C molecule due to less steric hindrance.

It has also never been investigated what influence the "ligand density" has on the binding of FVIII: C and other plasma proteins.

In addition to the importance of the knowledge of the binding character, which could possibly lead to the development of "better" binders, there is the importance of practical utility. Some literature has since been published on this, mainly aimed at AH-Sepharose.

In general, it can be said that the pH of the buffer used influences both the yield and the purity. Most commonly used was pH 5.5.

Because at low pH the proteins generally have less negative charge, while the positive charge on the gel does not change, less protein will adsorb than at higher pH. Because this appears to be less true for FVIII (low pi), the specificity of the gel can be increased (greater purification). However, the FVIII: C is less stable at low pH, which reduces the yield.

Faure et al. Has sought the solution to this problem by adding stabilizers. Sucrose was added to prevent complexation of 8701 9 1..4 * -7-FVIII with FIXa and FX and thus proteolysis.

However, this had little effect in plasma. The yield from 224 ml plasma to 30 ml AH-Sepharose was approx. 407th with a 9-fold purification.

5 Amphlett has described an isolation from plasma at pH 6.8. Without stabilizer addition, the yield was 39% with a 32-fold purification. It was adsorbed in 20 mM tris / 0.1 M lysine with 0.3 M sodium chloride, pH 6.8 and eluted in this buffer with 0.5 M calcium chloride. When 0.5 M benzamidine / 0.1% trasylol was added - potentially resulting in less proteolytic hydrolysis - the yield was 90% with a 48-fold purification.

literature: Brit. J. Haematol. 43 (1979) 669-674; Thromb.

15 Haemost. 48 (1982) 46-48; Folia Haematol., Leipzig 107 (1979) 148-151; Thromb. Haemostas. 47 124-127 (1982); J. Chromat. 257 (1983) 387-391; US-A-4,508,709.

As appears from the discussion of the known methods above, all known methods have one or more disadvantages, such as e.g. disappointing yield, low purity and / or low stability of the isolated products.

The invention now provides a new adsorbent material capable of achieving significant improvements in the isolation of coagulation factors, in particular improvement in yield, purity, and / or stability, and also provides a new method of isolating coagulation factors from a suitable liquid starting material using the new adsorbent material.

The novel adsorbent material according to the invention comprises a polymeric support material to which ligands of terminal primary, secondary, tertiary or quaternary amino groups are bound via spacers and characterized in that the spacers consist of 4 or 5 linked methylene groups according to the 8701914 * -8- general formula 1:

- (CH2) n- (D

where n represents an integer with a value of 4 or 5.

Furthermore, according to the invention it is preferred that the support material is a hydroxyl-containing polymeric material to which the spacers are bonded via coupling groups of the formula 2: -CO-NH- (2). The most preferred adsorbent material according to the invention can be represented of the general formula 3: carrier -O-CO-NH- (CH2) n-NR1R2 (3), wherein n is equal to 4 or 5 and and R2 each, independently of one another, a hydrogen atom or an unsubstituted hydrocarbon group such as alkyl, cycloalkyl, aryl, aralkyl, etc.

It is preferred that the support material is an agarose, such as e.g. Sepharose CL 4B.

It is also preferred that the ligands consist of primary amino groups, i.e., that R 1 and R 2 in the above general formula 3 represent hydrogen atoms.

Furthermore, according to the invention it is preferred that the ligand density is greater than 30 µmol / ml swollen matrix, most preferably 60-80 µmol / ml swollen matrix.

The novel method of the invention for isolating clotting factors, such as FII, FVII, FVIII, FIX, FX, and vWF, from a starting liquid material such as blood plasma, plasma products, cell lysates, and fermentation or culture media using liquid chromatography. characterized in that the new adsorbent material is used.

It is preferred in this method that an equilibration buffer and liquid starting material with a specific conductivity below 20 mS / cm, most preferably 13-16 mS / cm, be used in the adsorption step.

8701914 * -9-

It is further preferred that a matrix: starting material volume ratio of less than 1: 8, most preferably about 1:10, is used.

It is also preferred that elution is achieved by increasing the conductivity of the buffer, preferably using sodium chloride.

Regarding the pH, it is preferred that a pH in the range of 6-8, preferably about 7, is maintained.

The advantages of the invention in the isolation of coagulation factors are the unexpected consequence of the use of spacers between the amino groups acting as ligands and the support material, which spacers are slightly shorter than those of the known aminohexyl-Sepharose and therefore less be hydrophobic.

By using a relatively long spacer, compared to the known materials such as DEAE-Sephadex and DMAP-EMA (Johnson), the adsorption capacity of the FVIII-vWF is at the same ligand density.

20 complex bigger. As a result, the adsorption of the desired proteins is more specific and a purer product can be obtained.

By using a relatively hydrophilic spacer, compared to e.g. AH-Sepharose, at the same adsorption capacity (i.e., a higher ligand density), the adsorption is more specific, while also the elution of FVIII-vWF is better (higher yield, more reproducible). As a result, a purer product and also a higher yield can be obtained.

The carrier in the adsorbent material according to the invention is preferably an agarose such as Sepharose CL 4B, but is not limited to such a polymeric material. Preferably, the carrier satisfies the following desires: 1. The carrier generally causes only minimal interaction with the plasma proteins. This applies both during the adsorption to promote the affinity matrix specificity, and during elution to increase the yield; 2. the carrier has a very wide-pore pore structure 5 with a large internal surface. This is conducive to a high adsorption capacity; 3. the support is very stable both chemically and physically.

This is desirable because of the ligand coupling, the potential for regeneration for reuse, sterilization, column elution, storage, etc .; 4. the carrier is not too expensive to be economically acceptable.

Since no carrier material 15 is currently known that optimally meets all the stated requirements, a choice will have to be made from a number of less ideal materials. In principle, all water-insoluble carrier materials are suitable to a greater or lesser extent for the FVIII isolation process.

Some examples of this can be mentioned.

This number is not exhaustive: 1. cross-linked agaroses; 2. cross-linked ethylene / maleic anhydride (EMA) copolymers; 3. cellulose; 4. trisacryl; 5. fractogel TSK; 6. polystyrene; 7. cross-linked polyvinyl alcohol; 8. polyacrylamide; 9. silica gel; 10. Cross-linked polymers based on acrylic acid as a monomer (PMMA, poly-HEMA, etc.); 11. cross-linked dextrans (Sephadex).

8701314 -11-

For the FVIII isolation illustrated in the examples, cross-linked agaroses were chosen as the carrier material in the form of Sepharose CL 4B.

The ligand in the adsorbent material according to the invention is preferably the primary amino group NH2, which has a relatively high pK value, so that at a relatively high pH of 6-8 a larger fraction of the ligands is positively charged than e.g. this is the case with DEAE-Sephadex, DEAE cellulose and DMAP-EMA (Johnson). Therefore, to achieve the same charge density (adsorption capacity), a lower ligand density is required. As a result, the adsorption of the desired proteins is more specific and a purer product can be obtained. However, the ligand is not limited to the primary amino group 15. In general, all amines will be more or less suitable for the intended purpose, so that the ligands can also consist of secondary, tertiary or quaternary amino groups.

The manner in which the ligand is coupled to the support material via the spacer is also not limited to one specific coupling group and method. In principle, any method with which an amino group can be immobilized to a carrier material via a spacer is eligible. Preferably, however, the coupling method meets the following requirements: 1. the coupling can be carried out in a simple, fast and reproducible manner; 2. a high degree of substitution (ligand density) is achievable without crosslinking; 3. the reaction conditions must of course not be destructive to the matrix; 4. The resulting bond is very stable, so that no ligand leakage will occur.

870 1 91 4 -12-

When the support material is a polymeric material with free hydroxyl groups, such as e.g. in case of agaroses, a coupling e.g. Established using the following coupling reagents: 5 a. Cyanogen halides; b. triazines; c. periodate; d. (bis-) epoxides; e. carboxyalkyl halides; 10 f. divinyl sulfone; g. tresyl and tosyl chloride; h. carboxydiimidazole.

One of the most commonly used coupling methods uses cyanogen bromide. The aforementioned alkyl and aminoalkyl Sepharoses were also prepared with this. However, this method does not meet the requirements with regard to the degree of substitution and the absence of extra cargo and is therefore less suitable.

The stability is also not great. Recently, Bethell et al (J. Biol. Chem. 254 (8) (1979) 2572-2574; J. Chrom. 185 (1979) 463-470; J. Chrom. 218 (1981) 509-518) developed a method that seems to meet the requirements. 1,1'-Carbonyl-diimidazole (CDI) is used as the coupling reagent. In a first reaction step the matrix is activated and in a second step e.g. an amine.

The activation takes place in anhydrous organic solvent such as acetone, dioxane or dimethylformamide (DMF). The hydrogel must be transferred herein by solvent exchange. Hydrolysis of the activated gel with sodium hydroxide produces imidazole and carbon dioxide, which makes it easy to determine the degree of activation. The activated gel reacts both in organic solvent and in water with, among other things, amines · in high efficiency, so that all kinds of amine-containing ligands can be easily linked.

8701914 -13-

The coupling is easily reproducible and allows a high ligand density. The resulting carbamate (urethane) bond is very stable against hydrolysis and uncharged at a pH higher than 4.

Accordingly, this coupling method with CDI as coupling reagent and the resulting coupling group -CO-NH- is preferred according to the invention.

Affinity matrices according to the invention of the general formula 3 can be prepared by coupling a diamine to the support by CDI activation.

The ligand density of the affinity matrices of the invention is preferably greater than 30 µmol / ml, e.g. 50-100 µmol / ml swollen gel, most preferably 50-70 µmol / ml swollen gel. Using the CDI activation method, a high ligand density, even up to about 200 µmol / ml, can be easily and reproducibly achieved.

A high ligand density leads to a high adsorption capacity. This allows a lower gel / plasma ratio 20, which ensures lower costs.

The liquid starting material containing the clotting factor or factors to be isolated and which can be used in the method according to the invention will preferably consist of blood plasma, most preferably fresh or freshly frozen human blood plasma. No pre-purification and adjustments to the isolation method are required.

However, the invention is not limited to use in blood plasma and may in particular also be used to isolate the desired substances from plasma products, such as FVIII concentrate and cryoprecipitate, or from cell lysates and fermentation or culture media used in a biotechnological production of FVIII or other clotting factors are obtained.

The isolation method of the invention, which comprises an adsorption step and an elution step, is preferably performed at normal pressure and temperature, although variations are allowable. Thus, the temperature can be selected in the range of 0-60 ° C without serious objection, but for practical reasons room temperature is preferred. Both the adsorption step and the elution step can be carried out batchwise or in a column. Column treatment, however, is preferred, especially with regard to the elution step. As for the adsorption step, batch treatment 10 may be preferred.

There are also no special restrictions with regard to the duration. The adsorption step will e.g. may take from 10 min to 4 hours, but usually the adsorption step will take 45-60 min.

The affinity matrix is equilibrated using a buffer before use. This buffer preferably has the same pH and the same conductivity as the liquid starting material containing the clotting factors to be isolated. It has been found advantageous in the method according to the invention that the pH is kept at a value of 6-8, preferably about 7 during the process, and that the specific conductivity of both the equilibration buffer and the liquid starting material is less than 20 mS / cm, preferably 13-16 mS / cm. These process conditions make it possible to use blood plasma as such as starting material and are conducive to a relatively high yield. Also, these process conditions, especially the use of an approximately 30 physiological pH, are beneficial for large scale application (setting the pH of large amounts of plasma to 5.5 is quite inconvenient) and for the stability of the isolated clotting factors such as FVIII : C.

Regarding the gel / plasma ratio, lower values are preferred because of lower costs and less nonspecific adsorption. In the method according to the invention, therefore, a gel / plasma ratio G / P less than 1/8, most preferably about 1/10, is used. Assuming 5 plasma containing about 1 unit / ml, a G / P ratio of 1/10 means that about 10 units of clotting factor are offered per ml of gel.

The elution of the adsorbed clotting factors takes place by increasing the conductivity of the buffer 10. This is preferably achieved by adding sodium chloride to the buffer (e.g. up to 0.5 or 1.0 M), but other means can of course also be used for this purpose.

Due to the combination of optimal process conditions and plasma as the starting medium, the clotting factors FtI, FVII, FIX and FX as well as the FVIII-vWF complex are isolated in relatively high yield and purity. Separation of the FVIII-VWF complex from the rest can be performed in another process step if desired.

The invention will be explained in more detail below with reference to examples.

Examples related to the preparation of adsorbent materials.

According to the experiments described below, derivatives of Sepharose CL 4B were prepared using various amines which could be used as adsorbent material (affinity matrix) for the isolation of coagulation factors, in particular from citrated human blood plasma.

Sepharose CL 4B was used in the form of an aqueous suspension. For the syntheses, the water had to be effectively replaced with an organic solvent, such as Ν, Ν-dimethylformamide (DMF). This could be accomplished batchwise at relatively small amounts on a glass filter, or carried out in a larger batch column. After the solvent exchange, the Sepharose matrix was activated with 1,1'-carbonyldiimidazole (CDI) according to the procedure described by Bethell (J. Biol. Chem. 254 (1979), 2572-2574).

Then an alkanediamine can be directly coupled to the activated gel to yield an aminoalkyl Sepharose.

Since all Sepharose derivatives have basic amino groups as ligands, the ligand density can be easily determined by acid-base titration or by nitrogen elemental analysis.

The solvent exchange in which water is replaced with dimethylformamide or another organic solvent was carried out in two different ways depending on the amount of hydrogel used: 1. Generally, 12 ml of gel suspension (10 ml of moist cake) was washed three times on a glass filter. (G3) with distilled water and vacuumed nearly dry, and washed successively once with distilled water / DMF 70:30 (v / v), once with 30:70 and three times with DMF. This procedure was found to be effective for replacing water in small amounts of hydrogel.

2. In other cases, up to 400 ml of the gel (480 ml of suspension) was poured into a glass column (26x750 mm) and the column was washed with two column volumes of distilled water. Then a linear water / DMF gradient (400 ml) of 0-100% DMF was applied to replace the water and finally the gel was washed with about 1 volume by volume DMF until the effluent was completely anhydrous (which became by water analysis of the effluent checked).

The CDI activation was performed by suspending the gel in an equal amount of DMF (v / v) and adding a known amount of CDI based on 8701914 * -17- on the original gel volume in water (eg 0.1 mmol / ml gel). The suspension was carefully stirred or shaken at room temperature for at least 20 min and the activated gel was washed three times on a glass filter 5 with the same amount of DMF.

The amine coupling was performed in the organic solvent (DMF, dioxane, acetone) or in aqueous solution, depending on the solubility of the amine and the properties of the gel. To this end, a known amount (generally at least five times the desired total ligand density (TLD)) of the amine was dissolved in a volume amount of the solvent and then added to the same volume amount of activated gel, then carefully at room temperature for the desired time. stirred or shaken.

To remove the excess amine after coupling, the gel was washed with the solvent, 0.1 N hydrochloric acid, 0.1 N sodium hydroxide and distilled water.

The hydrogels were stored at 4 ° C in phosphate buffered saline (PBS) containing 0.01% thimerosal, or in adsorption buffer instead of PBS.

Titration was normally performed with 4-8 ml of the hydrogel. Gel volume was determined after centrifugation at 1500 x g for 10 min and waiting time for 30 min. The gel was equilibrated with 0.05 N sodium hydroxide by washing three times with it, sucked dry on a glass filter, suspended in 50 ml of distilled water and stirred. 0.100 N hydrochloric acid was then added at a flow rate of 0.125 ml / min. and the pH and conductivity were measured and recorded. The ion exchange capacity (for charged ligands this is equal to the total 8701914-18 ligand density (TLD)) could be calculated from the conductivity curve (which clearly shows the start and end points of the titration) and the gel volume ; the apparent pK value was obtained from the pH curve 5 at half the titration at a conductivity of 5 mS / cm. Ligand density (LD) can be calculated for a known gel volume.

The gel volume (GV) of the gel derivatives was normally determined in PBS unless otherwise indicated.

In a 15 ml polystyrene tube with a ml graduation, about 7 ml of the gel suspension was centrifuged at 2000 revolutions per minute for 10 min.

The precipitated gel was then allowed to stand at room temperature for at least 15 min and the volume 15 (+ 0.2 ml) was read. When the gel is in an organic solvent (such as DMF), the original volume is normally given in an aqueous solution.

Primary aminoalkyl Sepharoses with different alkyl-20 chain lengths

A series of aminoalkane Sepharoses (An-CDI-SEPH) of different alkyl chain lengths was prepared by the activation of 10 ml Sepharose CL 4B with 0.32 g CDI (0.2 mmol / ml) in DMF (activation reaction time (ART) equal to 20 min.) And coupling with 5 mmol amine in DMF (coupling reaction time (CRT) equal to 40 min.). Solvent exchange was performed on a glass filter. The results of this synthesis are shown in Table 1. This series of gels (with a ligand density of 52 + 5 µmol / ml) was used to investigate the influence of the alkyl chain length on factor VIII adsorption and elution. The coupling yield compared to the added amount of CDI for this series was 17 + 2% (TLD = 328 + 43 µmol; GV = 6.3 + 0.6 ml, 35 from 8 experiments).

870 191 4 -19-

Table 1. Properties of Aminoalkyl Sepharoses (An-CDI-SEPH).

GEL CODE TLD GV LD 1 pKa XN DS DSC LD 2 umol ml umol / ml mg / g mol / mol mg / ml umol / ml A2-11.10.85 371 7.0 53 6.0 26.5 0.315 53.2 50 5 A3-09.10.85 290 6.3 46 8.0 23.2 0.276 51.4 43 A4-09.10.85 347 6.3 55 8.4 14.7 0.171 45.2 24 A5-10.10.85 413 7.0 59 8.9 30.0 0.380 52.9 57 A6-10.10.85 318 6.0 53 9.1 7.2 0.082 55.2 14 A7-16.10.85 294 7.0 42 8.7 27.3 0.352 48.6 47 10 A8-11.10.85 275 5.4 51 8.6 22.7 0.288 67.2 55 A9-16.10.85 314 5.6 56 8.0 29.2 0.395 65.5 68 1 derived from the titration result 2 derived from elemental analysis

Aminobutyl-Sepharoses with different ligand densities 15 To get the optimal ligand density for factor VIII

to determine isolation with aminobutyl-Sepharose (A4-CDI-SEPH), a series of gels was prepared using different amounts of CDI for the activation of 10 ml Sepharose CL 4B in DMF (ART = 20 min.) and coupling with 1, 4-butanediamine in DMF (CDI / amine = 2/5; CRT-40 min.). The solvent exchange was performed on a glass filter. The results of this synthesis are shown in Table 2.

8701914

Table 2. Synthesis of Aminobutyl Sepharose (A4-CDI-SEPH) with Different Amounts of CDI: Glass Filter-Solvent Exchange -20-

GEL CODE CDI TLD YIELD GV LD 1 pK

umol umol% ml umol / ml 5 A4-14.11.85.1 154 63 41 9.0 7 A4-12.11.85.1 308 93 30 9.3 10 9.3 A4-12.ll.85.2 617 128 21 8.0 16 9.0 A4-01.11.85.1 925 198 21 7.9 25 9.0 A4-11.12.85.1 1230 270 22 7.1 38 8.5 10 A4-11.12.85.2 1540 333 22 6.8 49 8.3 A4-09.10.85 1973 380 19 6.9 55 7.3 A4-13.02.86.1 2160 409 19 6.7 61 A4-13.02 .86.2 2470 494 20 6.5 76 A4-01.ll.85.2 3700 557 15 5.3 105 8.9 15 A4-14.ll.85.2 7400 681 9 4.1 166 9.2 1 mean value from 2 determinations

Also some larger amounts of aminobutyl Sepharose were prepared after solvent exchange in a column. The water content in the DMF during the activation with CDI was about 0.137% (w / w).

The 1,4-butanediamine was coupled in DMF (CDI / aminea = 2/5). The data are listed in Table 3.

8701914 ί

Table 3. Synthesis of aminobutyl Sepharose (A4-CDI-8EPH) with different amounts of CDI: solvent exchange in a column * -21-

GEL CODE SEPH. CDI ART GV 1 CRT TLD OPBR. GV LD

ml mmol hours ml hours mmol% ml umol / ml 5 A4-06.03.86 100 18.5 0.3 100 0.6 8.7 47 67 130 A4-13.03.86.1 100 8.6 0.3 75 17 3.2 50 47 68 A4-13.03.86.2 12.5 0.6 0.46 43 9.1 50 A4-13.03.86.3 12.5 not linked A4-03.06.86.1 350 30.2 16 10 1 0.19 22 5.8 32 10 A4-03.06.86.2 10 4 0.20 23 5.1 39 A4-03.06.86.3 330 17 9.2 32 230 40 A4-23.10. 86 2 100 10.9 22 100 7 4.9 45 64 77 A4-15.12.86.2 3 700 84 1.5 440 0.2 28.9 55 390 74 A6-15.12.86.2 3 71 0.2 4.4 52 59 75

15 i corrected for gel shrinkage in DMF

2 from SEPH.25.09.86 3 CD1 / amine ratio 1/25

Influence of water on the CDI activation and the amine coupling To investigate the influence of water during the CDI activation, the water was mixed in a batch of 400 ml

Sepharose CL 4B in a column replaced with DMF (SEPH.25.09.86).

The water content in the above DMF was 0.13%.

Then to 10 ml portions of the batch a known amount of water and DMF was added in a total volume of 45 ml, the water content in the above DMF (35 ml) was determined again, the gel suspension (10 ml) was activated with 2, 2 mmol CDI (ART = 180 min.) And coupled with 5.5 mmol 1,4-butanediamine (CRT-60 min.).

The results are shown in Table 4.

8701914 -22-

Table 4. Synthesis of aminobutyl Sepharose (A4-CDI-SEPH.) With different amounts of water during activation.

GEL CODE WATER 1 WATER 2 TLD YIELD GV LD

%% umol% ml umol / ml 5 'A4-02.10.86.1 0.00 0.13 869 40 4.8 183 A4-02.10.86.2 0.11 0.26 533 24 5.8 92 A4-02.10.86.3 0.22 0.36 435 20 6.0 72 A4-02.10.86.4 0.67 0.83 285 13 8.0 36 A4-02.10.86.5 1.11 1.33 193 9 8.8 22 10 1 added amount of water 2 set amount of water

Two batches of Sepharose CL 4B (SEPH 25.09.86) of 50 ml each were activated with 5.43 mmol GDI (ART = 180 min.) And used to determine the influence of water in DMF during the amine coupling.

To 7.1 ml CDI-SEPH portions (corrected for DMF shrinkage), a known amount of water was added prior to the reaction with 2 mmole 1,4-butanediamine (CRT = 40 min.). The effect of the water content 20 is shown in Table 5. Different CDI / amine ratios were also used as shown in Table 6.

8701 9 1 4 ί

Table 5. Synthesis of aminobutyl Sepharose (A4-CDI-SEPH.) With different amounts of water during the amine coupling.

-23-

GEL CODE WATER 1 WATER 2 TLD YIELD GV LD

%% umol% ml umol / ml 5 A4-12.10.86.1 0.0 0.13 303 39 4.3 70 A4-12.10.86.2 0.0 0.13 317 41 4.6 69 A4-12.10.86.3 3.2 313 41 4.9 64 A4-12.10.86.4 6.3 299 39 5.0 60 A4-12.10.86.5 57.0 190 25 6.5 29 10 A4-17.ll.86.1 0.0 0.13 298 39 4.1 73 A4-17.ll.86.2 17.0 3Q5 40 6.4 49 A4-17.ll.86.3 29.0 210 27 5.7 37 A4 -17.11.86.4 35.0 230 30 6.1 38 A4-17.ll.86.5 40.0 222 29 6.1 36 15 A4-17.11.86.6 50.0 209 27 6.0 35 1 added amount of water 2 set amount of water

Table 6. Synthesis of aminobutyl Sepharose (A4-CDI-SEPH.) At different CDIiamine ratios.

20 GEL CODE TLD GV LD 1 GDI: XN DS DSC LD 2 umol ml umol / ml amine mg / g mol / mol mg / ml umol / ml A4-12.10.86.1 303 4.3 70 2: 5 35.4 0.45 53.8 68 A4 -12.10.86.6 344 5.0 69 1:10 39.3 0.51 50.5 71 A4-17.ll.86.1 298 4.1 73 2: 5 34.6 0.44 59.1 73 25 A4-17.11.86.7 383 5.1 75 1:28 43.1 0.57 46.8 72 1 derived from the titration result 2 derived from element analysis 8701914 -24-

Secondary and tertiary amines as a ligand.

Several secondary and tertiary diamines were also coupled to Sepharose CL 4B. The solvent exchange was performed on a glass filter, 5 Sepharose (250 ml) was activated with CDI (44.8 mmol, ART = 60 min.) In DMF and 25 ml aliquots of the activated gel (CDI-SEPH 08.07.86) were coupled with 12 mmol of amine in DMF (CRT = 180 min.). The results of this synthesis are shown in Table 7. The water content e 10 in DMF during activation was 2.67o (w / w). The coupling yield was 22 ± 5%; LD = 52 + 11 µmol / ml; GV = 19 + 1 ml (9 experiments).

Table 7. Properties of secondary and tertiary aminoalkyl Sepharoses

GEL CODE CDI TLD YIELD GV LD pK

15 mmol mmol% ml umol / ml DBA3-SEPH.15.07.86 4.48 1.18 26 21.4 55 8.5 DEA2-SEPH.15.07.86 4.48 1.05 23 19.1 55 8.1 DEA3-SEPH.15.07.86 4.48 1.01 23 18.0 56 8.9 DEA5-SEPH .15.07.86 4.48 0.50 11 19.2 26 9.7 20 DMA2-SEPH.15.07.86 4.48 1.04 23 18.2 57 6.4 DMA3-SEPH.15.07.86 4.48 1.18 26 19.3 61 8.3 EA2-SEPH.15.07.86 4.48 1.04 23 18.0 58 8.3 DMA6-SEPH.15.07.86 4.48 0.67 15 17.5 38 9.5 MA3-SEPH.15.07.86 4.48 1.27 28 20.1 63 9.1 25 CDI-SEPH.08.07.86 4.48 2.65 59 23.0 115

Dimethylaminopropyl Sepharoses with different ligand densities.

To determine the optimal ligand density for factor VIII isolation on dimethylaminopropyl-Sepharose (DMA3-CDI-30 SEPH), a series of gels was prepared using different amounts of CDI for the activation of 25 ml Sepharose CL 4B in DMF (ART = 90 min.) And 8701914 -25 coupling with 30 mmol N, N-dimethyl-3-aminopropylamine in DMF (CRTs 90 min.) Were used. The solvent exchange was performed on a glass filter.

The water content in DMF during activation was 3.1% (w / w). The results of this synthesis are shown in Table 8 (GV * = 20.6 + 7 ml (10 experiments).

Table 8. Synthesis of dimethylaminopropyl-Sepharose (DMA3-CDI-SEPH.) With different amounts of CDI.

GEL CODE CDI TLD YIELD GV LD

10 mmol mmol% ml umol / ml DMA3-09.07.86.1 0.62 0.27 44 20.5 13 DMA3-09.07.86.2 1.23 0.38 31 20.0 19 DMA3-09.07.86.3 1.85 0.53 29 21.0 25 DMA3-09.07.86.4 2.47 0.53 21 21.0 25 15 DMA3 -09.07.86.5 3.08 0.74 24 21.0 35 DMA3-09.07.86.6 3.70 0.84 23 20.5 41 DMA3-Q9.07.86.7 4.62 0.88 19 19.5 45 DMA3-09.07.86.8 6.17 1.21 20 22.0 55 DMA3-09.07.86.9 9.25 1.41 15 21.0 67 20 DMA3-09.07.86.10 12.33 1.50 12 19.5 77

The gel codes specify the nature of the gel and its preparation date. The abbreviations used have the following meanings: A2 - aminoethyl 25 A3 - aminopropyl A4 - aminobutyl A5 - aminopentyl A6 - aminohexyl A7 - aminoheptyl 30 A8 - aminooctyl A9 - aminononyl 8701914% -26- DBA3 - Ν, Ν-dibutylaminopropyl DEA2 Ν-diethylaminoethyl DEA3 - Ν, Ν-diethylaminopropyl DEA5 - N, N-diethyl-2-aminopentyl 5 DMA2 - Ν, Ν-dimethylaminoethyl DMA3 - Ν, Ν-dimethylaminopropyl EA2 - N-ethylaminoethyl DMA6 - N, N'- 1,6-aminohexyl MA3 - N-methylaminopropyl Examples related to factor VIII isolation.

In the experiments described below, an adsorbent material (also referred to as an affinity matrix or gel) was used which was equilibrated with and suspended in adsorption buffer. Thereafter, in simple adsorption experiments, the gel and a normal plasma stock (NPP) were incubated either columnwise or batchwise at different gel / plasma ratios under certain conditions.

Gels exhibiting high factor VIII adsorption (above 75% at a G / P ratio of 1: 8) were incubated again and then eluted columnwise with 1 M sodium chloride in buffer. The factor VIII content in the eluate and the total protein concentration were determined taking the normal plasma stock as reference. From this data, the factor VIII recovery and the purification factor (PUR) could easily be obtained. In case a standard was used as reference (for FVIII: C the CLB-PS41 and for the protein human serum albumin), the specific activity could be determined (Table 12).

For gels with promising results, the recovery and specific activity in the eluate of factor II, VII, IX and X were also determined because these coagulation factors can also be adsorbed on the gel and eluted.

8701914 τ -27-

Normal plasma stock (NPP) was obtained by pooling freshly citrated plasma (CPDA-1 plasma) from at least 5 donors (human serum albumin HSA = 39.6 1.0 mg / ml; IgG = 11.8 + 0.4 mg / ml; 3 stocks).

Before freezing, if necessary, the pH was adjusted to 7.4 with 1 N acetic acid. In the case of pH 5.5, acetic acid was added to the plasma shortly before use to obtain the desired pH.

Batch adsorption.

The hydrogel to be tested was equilibrated on a glass filter with a 0.1 M acetate / lysine buffer at a desired pH and suspended in 1.2 times the gel volume. A known amount of this suspension was incubated "end over end" in a plastic human plasma tube (NPP) at the same pH for 5-120 min at room temperature.

The 0.1 M acetate / lysine buffer used contained 8.20 g of sodium acetate, 18.27 g of L-lysine monohydrochloride, 2.94 g of trisodium citrate, 1.0 ml of calcium M chloride solution of 1.0 M, and 1 N sodium hydroxide or 1 N hydrochloric acid to adjust the pH.

To investigate the influence of the conductivity on the FVIIIrC adsorption, an experiment was conducted at different sodium chloride concentrations.

To decrease conductivity, 0.6 ml of the gel suspension was mixed with 4.0 ml of citrate buffer (10 mM) containing an appropriate amount of sodium chloride.

An increase in conductivity was obtained by mixing 0.6 ml of the gel suspension with 4.0 ml of acetate / 30 lysine buffer containing a desired amount of sodium chloride. This mixture was then incubated with 4.0 ml NPP at room temperature for 30 min.

To get an impression of the influence of the pH on the FVIII: C adsorption, an incubation was carried out at a pH between 7.4 and 5.5. To adjust the pH of plasma and buffer, 1.0 M acetic acid was slowly added.

8701914 -28-

After incubation, the mixture was centrifuged at full speed for 2 min and generally the supernatant plasma (supernatant plasma SUP) was kept on ice for factor VIII determination.

When elution was required, the hydrogel was poured into a column (Pharmacia C10 / 10) and eluted as described below. The buffer was collected in the supernatant plasma during washing. This fraction was believed to contain the unbound proteins (unadsorbed protein fraction NADS).

Columnar adsorption.

About 5 ml of the hydrogel was used for column adsorption. The gel was poured into a 15 column (Pharmacia C10 / 10) and equilibrated with the 0.1 M acetate / lysine buffer at pH 7.0 or 5.5. A known amount (1-30 ml) of human plasma (NPP) with the same pH and conductivity as the buffer was applied to the column at room temperature, while 1-5 ml of NPP was kept on ice during the experiment ( REF). The flow rate was kept constant at 0.125-1.0 ml / min. using a peristaltic pump (Pharmacia P-1). UV absorbance at 280 nm was monitored in the column eluate using a Pharmacia Single Path Monitor UV-1. As soon as protein was eluted from the column (after 2-3 ml, depending on the hydrogel), the eluate was collected to a certain volume (plasma + wash buffer) (NADS) and stored on ice.

30 Column Elution

Columnwise elution started after washing with buffer (normally 10-20 ml) until no more protein could be detected. Elution of the bound protein was then performed with the same buffer containing 1.0 M sodium chloride. The eluate showing the protein 8701914 t -29 peak (about 10 ml) was collected and stored on ice (the eluted protein fraction ELU).

To correct the high sodium chloride concentration for FVIII: C APTT assay (the activated partial thromboplastin time), the ELU was desalted on a Sephadex G-10 column (Pharmacia C16 / 20) equilibrated with the original 0.1 M acetate / lysine buffer , or the ELU was diluted appropriately with APTT buffer containing less sodium chloride when the FVIII: C10 content was sufficiently high.

Protein concentrations were determined using the Bradford Protein Assay, BioRad (ΒΡΑ), based on the shift of the absorption maximum of Coomassie Brilliant Blue G-250 when binding to proteins occurs, or were determined using the BCA Protein Assay,

Pierce (BCA) at 60 ° C, which is based on the biuret reaction, using the conditions described by the manufacturer in both cases. Duplicate samples were diluted appropriately and analyzed according to the manufacturer's instructions against a standard curve of normal plasma (the BCA test gave slightly higher values for the same samples than the ΒΡΑ test) or human serum albumin (54.1 mg / ml ).

Antigen concentrations (FVIII: Ag and vWF: Ag) were determined using an Enzyme Linked Immuno Sorbent Assay (ELISA) based on the monoclonal antibodies CLB CAg-A and CLB CAg-117 for FVIII: Ag and CLB RAg-20 and CLB RAg-35 for vWF: Ag. The abbreviations used have the following meanings: FVIII = the blood coagulation factor VIII; VWF = the von Willebrand factor which is also referred to in the literature as FVIIIR; FVIII: C = the coagulation activity of factor VIII; FVIII: Ag = factor VIII antigen; and 35 VWF: Ag = von Willebrand factor antigen, which is also referred to in the literature as FVIIIR: Ag.

The clotting factor activities were determined with a single-stage APTT assay as described by Hardisty and McPherson, Thromb. Diath. Haemorr. 7 (1962), 215-229.

The assay was performed on a LODE LC 6 Coagulometer with artificial substrate for FII and FX, congenital deficient substrate for FVII and FVIII, and immunodepleted substrate for FIX. The reference was prepared by serial dilution (1/10 to 1/320) of a normal plasma standard.

The FVIII Coatest for the measurement of FVIII: C10 activity on a chromogenic substrate was performed according to the manufacturer's instructions (Kabi Vitrum).

The results were in good agreement with the FVIII: C APTT. Serial dilutions (1/160 to 1/800) of normal plasma (CLB-PS41, FVIII: C = 0.78 U / ml) were taken as reference.

Albumin and immunoglobulin G were measured by a fake helometric method. Fibrinogen (Fb) and fibronectin (Fn) were determined by Mancini radial immunodiffusion.

20 Charge-type adsorption experiments.

To obtain information about the adsorption of FVIII: C under different conditions, small-scale batch adsorption experiments were performed with almost all synthesized gels. About 1 ml of plasma was incubated with gel at a given gel / plasma ratio (pH = 7.4; AT = 30 min.) And the supernatant plasma (supernatant plasma SUP) was diluted 1:10 and analyzed in duplicate with APTT for FVIII : C; this was compared to the same experiment performed with the matrix starting material (Sepharose CL 4B).

Influence of the alkyl chain length on adsorption.

The results for a series of aminoalkyl Sepharoses (LD = 52 + 5 µmol / ml) at three different gel / plasma ratios are shown in Figure 1. Sepharose CL 4B 35 was taken as reference adsorbent.

Influence of the terminal amino group on adsorption.

A series of secondary and tertiary aminoalkyl- 8701914> -31-

Sepharoses (An-CDI-SEPH 15.07.86; LD = 52 µmoles / ml) was used for FVIII adsorption measurements. At a gel / plasma ratio of 1:10, the FVIII: C adsorption was 61 + 8% (9 experiments). Deviations from this percentage in the case of two gels were mainly due to other ligand densities. No direct relationship could be established between the apparent pK and the FVIII adsorption.

Column-wise adsorption experiments.

10 Influence of the ligand density on adsorption and elution.

A series of aminobutyl Sepharoses was tested to investigate the influence of the ligand density on FVIII adsorption during column adsorption.

Plasma (1 ml NPP) was applied to the column (4 ml; G / P = 4; flow rate FR = 0.25 ml / min; pH 7.0), the column was washed with buffer (NADS = 20 ml ) and eluted with 1.0 M sodium chloride (ELU = 10 ml). The results of the FVIII: C, FVIII: Ag and vWF: Ag analysis 20 are shown in Table 9. Protein adsorption data are not presented because in most cases they were too low to be determined.

Table 9. Column adsorption and column elution on aminobutyl-Sepharose 25 (A4-CDI-SEPH) with different ligand densities.

GEL CODE LD% VIIIïC% VIII: Ag% vWF: Ag% PROTEIN PÜR

µmol / ml ADS ELU ADS ELU ADS ELÜ ELU

A4-12.11.85.1 911 18 0 70 00 A4-12.11.85.2 16 22 2 21 0 4 16 0 0 30 A4-01.11.85.1 25 46 8 42 0 25 20 0 0 A4-Ll.12.85.1 38 61 20 66 12 35 45 0.8 25 A4-11.12.85.2 51 100 60 100 62 100 56 2.7 22 A4-09.10.85 55 98 77 100 70 100 90 4.6 17 A4-01.ll.85.2 105 97 44 100 37 100 60 4.8 9 35 A4-14.ll.85.2 166 99 63 100 51 100 86 11 6 870ΤΠΤ -32-

Columnwise elution.

Influence of the alkyl chain length on the column elution.

The influence of the alkyl chain length on FVIII interaction after column and batch adsorption was investigated. Plasma (10 ml NPP) was incubated batch with gel (pH 7.4; G / P = 0.125; adsorption time AT = 30 min.), Washed with buffer (FR = 1.0 ml / min .; NADS = 40 ml ) and eluted with 1.0 M sodium chloride (FR = 0.5 ml / min; ELU = 10 ml). FVI1I: C (APTT) and vWF: Ag 10 were analyzed in duplicate at three dilutions.

The data are shown in Table 10.

Table 10. Batch adsorption and columnar elution to aminoalkyl Sepharoses (An-CDI-SEPH) at a gel / plasma ratio of 1: 8.

GEL CODE LD% FVIII: C% vWF: Ag% PROTEIN 1 PUR

15 µmol / ml ADS ELU ELU / ADS ADS ELU ELU REF / ELU

DMA3-09.07.86.10 77 99 65 65 99 83 1.58 63 41 A4-13.03.86.2 50 75 38 51 80 78 0.27 370 141 A5-10.10.85 59 98 48 49 93 67 0.84 119 57 A6-10.10.85 53 94 43 46 88 36 0.81 123 53 20 ------------------------------------------- -------------------------

1 determined with BPA

Influence of the ligand density on the column elution.

The influence of the ligand density on the FVIII interaction during batch adsorption was also investigated under more practical conditions.

Plasma (10 ml NPP) was incubated batchwise with gel (G / P = 0.125; pH = 7.4; AT = 30 min.), Washed with buffer (FR1 / 10 ml / min .; NADS = 40 ml) and eluted with 1.0 M sodium chloride (FR = 0.3 ml / min; ELU = 10 ml).

FVIII: C (APTT) and VWF: AG were analyzed in duplicate at three dilutions. The data are shown in Table 11.

¢ 701914 <5 -33-

Table 11. Batch adsorption and column elution using aminobutyl Sepharoses (A4-CDI-SEPH) with different ligand densities at a gel / plasma ratio of 1: 8.

GEL CODE LD% FVIII: C Z PROTEIN 1 PÜR

5 umol / ml ADS ELÜ ELU / ADS ELU REF / ELÜ A4-01.11.85.1 25 000 0.022 4600 0 A4-11.12.85.1 38 13 3 23 0.063 1600 48 A4-13.03.86.2 50 75 38 51 0.275 363 138 A4-13.02 .86.1 61 94 53 56 0.43 232 123 10 A4-13.03.86.1 68 93 62 67 0.54 185 115 A4-12.10.86.6 2 69 100 77 77 0.74 136 105 A4-12.10.86.2 69 96 57 59 0.53 189 108 A4-13.02 .86.2 76 94 56 60 0.72 139 78 A4-01.ll.85.2 105 99 58 58 2.22 45 26 15 A4-06.03.86 130 94 64 68 2.22 45 29 A4-14.11.85.2 166 97 79 81 3.03 33 26

1 determined with BPA

2 this gel was prepared with a large excess of amine

The FVIIIrC adsorption and elution are plotted in Fig. 2 for clarification as a function of the ligand density.

Furthermore, in Fig. 3 the purification factor (PUR) is plotted as a function of the ligand density, it being found that an optimum is at a ligand density of 60 µmol / ml.

Similar experiments as with aminobutyl-Sepharose were performed with some dimethylaminopropyl Sepharoses with different ligand densities. The data from these experiments were approximately the same and are not shown here.

Reproducibility with promising matrices.

To investigate the reproducibility of the affinity chromatography method, two hydrogels were used for batch adsorption and column elution experiments. This concerns the promising 8701914 Λ -34-aminobutyl-Sepharose and the well-known aminohexyl-Sepharose. Plasma (20 ml NPP) was incubated with gel (G / P = 0.1;

AT = 45 min .; pH = 7.4), washed with buffer (FR = 1.0 ml / min; NADS = 32 ml) and eluted with 1.0 M sodium chloride 5 (FR = 0.5 ml / min; ELU = 20 ml ). FVIII: C in NPP, NADS

and ELU was determined with the FVIII Coatest against a plasma standard. The experiment was repeated n times for each gel. The mean values and standard deviation for FVIII and total protein are shown in Table 12.

Table 12. Batch adsorption and column elution of 20 ml NPP on 2 ml affinity gel: FVIII and total protein. (average from 3 experiments)

GEL CODE LD NPP% FVIII: C YIELD% PROTEIN 1 FVIII: C

U / ml ADS ELU U / kg NPP ELU E / mg A4-15.12.86.2 77 0.92 95 + 2 60 + 13 560 + 14 1.13 0.61 + .16 15 A6-15.12.86.2 79 0.90 94 + 2 65 + 11 570+ 7 1.61 0.36 +. 17

1 determined by BCA

Tables 13 and 14 show the results of the determination of coagulation factors and other proteins for the same experiments. The values for albumin 20 in the eluate were below 0.47, that for immunoglobulin G below 0.77.

Table 13. Batch adsorption and column elution of 20 ml NPP on 2 ml affinity gel: clotting factors FII, FVII, FIX and FX.

GEL CODE% FII: C% FVII: C% FIX: C% FX: C FIX

25 ADS ELU ADS ELU ADS ELU ADS ELU E / mg A4-15.12.86.2 99 67 92 90 99 67 99 78 0.65 A6-15.12.86.2 99 56 94 100 98 67 99 67 0.38 8701914 t -35-

Table 14. Batch adsorption and column elution of 20 ml NPP on 2 ml affinity gel: vWF, FVIII: Ag, fibrinogen and fibronectin.

GEL CODE% vWF: Ag FVIIIrC ELU% Fb: Ag% Fn: Ag

ADS ELU REC / VIII: Ag / vWF: Ag ADS ELU REC ADS ELU REC

5 A4-15.12.86.2 96 65 69 0.97 0.93 0 1.4 100 60 57 97 A6-15.12.86.2 93 30 37 0.91 2.53 14 8.4 94 96 77 81 8701914

Claims (10)

Adsorbent material, comprising a polymeric support material to which ligands are bonded via spacers consisting of terminal primary, secondary, tertiary or quaternary amino groups, characterized in that the 5 spacers consist of 4 or 5 concatenated methylene groups of the general formula 1: - (CH2) n- (1) where n represents an integer with a value of 4 or 5. The adsorbent material according to claim 1, characterized in that the support material is a hydroxyl group-containing polymeric material to which the spacers are bonded via coupling groups of the formula 2: 15 -C0-NH- (2) Adsorbent material according to any one of claims 1-2, characterized in that the carrier material is an agarose. Adsorbent material according to any one of claims 1-3, 20, characterized in that the ligands consist of primary amino groups. Adsorbent material according to any one of claims 1 to 4, characterized in that the ligand density is greater than 30 µmol / ml swollen matrix, most preferably 50-70 µmol / ml swollen matrix. 6. A method of isolating coagulation factors from a liquid starting material such as blood plasma, plasma products, cell lysates and fermentation or culture media by liquid chromatography, characterized in that an adsorbent material according to any one of claims 1-5 is used. A method according to claim 6, characterized in that an equilibration buffer and liquid starting material with a specific conductivity below 20 mS / cm, most preferably 13-16 mS / cm, are used in the adsorption step. Method according to claim 6 or 7, characterized in that a volume ratio matrix: starting material 5 less than 1: 8, preferably about 1:10, is used. Process according to any one of claims 6-8, characterized in that elution is effected by increasing the conductivity of the buffer, preferably with the aid of sodium chloride. Method according to any one of claims 6-9, characterized in that a pH in the range of 6-8, preferably about 7, is maintained. 8701914 -3fc- FIGURE DESCRIPTION Fig. 1 shows the FVIIIrC adsorption versus the alkyl chain length N of aminoalkyl Sepharoses (An-CDI-SEPH) at different gel / plasma ratios; Figure 2 shows the FVIIIrC adsorption and elution versus the ligand density using aminobutyl-Sepharose (A4-CDI-SEPH) at a gel / plasma ratio of 1: 8; Figure 3 shows the purification factor versus the ligand density using aminobutyl-Sepharose (A4-CDI-SEPH) at a gel / plasma ratio of 1: 8. i S7019 1 4