<U>Chemical Process</U> This invention concerns an improved method for the purification or enrichment of vitamin R- dependent blood clotting factors in fractions containing such factors.
Plasma obtained from blood of human or animal origin contains a number of valuable physiologically active substances, notably the various blood clotting factors, some of which are used in medicine for the treatment of individuals lacking one or more of such substances. The blood clotting factors, i.e. the members of the so-called blood clotting cascade, comprise a group.of inactive and active proteolytic enzymes and modified proteins related in the following manner:
EMI0001.0005
The subscript "a" means that the enzyme is in the activated form. Further significant plasma proteins, which may be considered to fall within the general category of blood clotting factors, include Protein C and Protein S.
Protein C is an inactive zymogen of theproteolytic enzyme Activated Protein C (APC,PCa).
Unlike the above-mentioned clotting factors, APC has an anticoagulant effect which functions through the proteolysis of Factor V and Factor VIII which are thus inactivated. Activation of Protein C appears to be by a feedback process involving thrombin and an endothelial-bound protein, thrombomodulin, which both function during the coagulation process. The activity of APC is then enhanced by the presence of Protein S, which seems to act as a cofactor.
A number of the above factors, namely II, VII, IX, X, and also Protein C and Protein S, have been shown to be vitamin K-dependent. These factors, both in the active and inactive form (apart from activated Factor II), carry terminal gamma-carboxy glutamic acid residues which are introduced by an efizymic process mediated by vitamin K. Additionally, Protein C, Protein S and Factors X and IX, but not prothrombin (Factor II), have been found to possess a modified aspartic acid residue, namely a (3-hvdroxyaspartic acid residue.
Thus, the vitamin K-dependent blood clotting factors represent a group of related acidic proteins having closely similar physicochemical properties. The acidic nature of these substances has enabled them previously to be isolated as a group though their structural and chemical similarities have made it difficult to separate the individual proteins using conventional techniques such as anion exchange chromatography.
Apart from Factor VIII, which is not part of the prothrombin complex, the most frequently administered blood clotting factor is Factor IX. Traditionally, Factor IX has been provided as a concentrate of all the factors of the prothrombin complex, II, VII, IX and X, although Factor VII is sometimes provided separately from IX, II and X.
Following infusions of such Factor IX concentrates, there have been a few reported cases of venous thrombosis and disseminated intravascular coagulation (DIC). These undesirable responses might be ascribed to a hypercoagulable state encouraged by high levels of unnecessary clotting factors infused with the Factor IX and/or by unwanted thrombogenic components in the concentrate. Intense and lengthy administration of Factor IX may lead to much higher than normal levels of Factor II, which is infused in the equivalent amounts and has a longer half-disappearance time in plasma.
Some hypotheses based on<U>in vitro</U> studies or animal models have also suggested that a combination of factors might be more thrombogenic than Factor IX alone. Indeed, present dosage levels of Factor IX are in part limited by the need to avoid excessive levels of Factor II.
Furthermore, following treatment with factor VIII concentrates, some haemophilia A patients show signs of non-viral immunosuppression. A similar i:Rmunosuppression in haemophilia B patients after treatment with Factor IX concentrate suggests that the response in both cases may be due to an unwanted contaminant in the concentrates.
Donor plasma pools from which clotting factor concentrates are prepared contain the causative agent (probably viral) of non-A non-B hepatitis and must be assumed to carry the risk of contamination with viruses (e.g. hepatitis B and HIV). To improve product safety, the concentrates must undergo virucidal treatments, which may be complicated by the presence of contaminant proteins. For example, this virucidal treatment may be easier to devise in a more highly purified concentrate in which interference from unnecessary proteins can be minimised.
In general, it is preferable in principle to administer a blood clotting factor intended to replace or supplement a specific clotting factor in a human patient, without simultaneously raising the level of other factors or_, indeed, any contaminants which might produce unwanted physiological effects. There is thus a need for an improved method of at least partially separating the vitamin K- dependent clotting factors in order to reduce or even eliminate some of the above problems. It has now been found that the vitamin K-dependent blood clotting factors can be at least partially separated from each other by metal chelate chromatography.
According to the present invention, there is provided a method for at least partially separating vitamin K-dependent blood clotting factors from a mixture containing at least one such factor, characterised in that said mixture is adsorbed on to a chelate of a polyvalent metal immobilised on an inert support, followed by elution to yield one or more fractions enriched in respect of one of said factors.
In general said mixture will contain at least two vitamin K-demendent blood clotting factors.
Thus, although the invention includes purification or enrichment of fully functionalised clotting factors derived from microbiological cultures using recombinant DNA technology, the source of the clotting factors will commonly be blood plasma, particularly human plasma.
The mixture of factors to be treated in accordance with the invention will most commonly be a plasma fraction concentrate from which some plasma factors have already been removed.
In obtaining the mixed factors from plasma, typical initial fractions containing the desired components include cryoprecipitate supernatant (the supernatant remaining after cryoprecipitation of Factor VIII concentrate), Fraction I supernatant and fractions obtained after adsorption of the plasma with various affinity reagents, e.g.
heparin Sepharose. In general, the factors will have been further concentrated by adsorption on to an anion exchange resin such as diethylaminoethyl (DEAF) cellulose, DEAF Sephadex or DEAF Sepharose to produce a prothrombin complex concentrate.
It should be noted that under optimal conditions for Factor IX preparation, DEAE cellulose does not bind Factor VII, although higher capacity ion-exchangers (e.g. DEAE Sephadex and DEAE Sepharose) bind Factor VII more efficiently under accepted industrial processing conditions. Thus the prothrombin complex concentrate from DEAE cellulose will not contain Factor VII. This difference in behaviour enables a concentrate containing predominantly Factor VII to be obtained, although this can also benefit from purification and enrichment according to the invention.
While the anion exchange resin concentrates may beneficially be treated in accordance with the invention, other combinations of the factors may also advantageously be treated to enhance the concentration of a desired factor. In particular, an enriched fraction produced in accordance with the invention, which may contain only two of the factors, may be subjected to a second treatment to produce still further enrichment.
Existing purification systems have several inherent disadvantages. Heparin-Sepharose gives poor resolution and clotting factor activities when used on an industrial scale. Sulphated dextran produces Factor<B>IX</B> with a very short NAPTT (see below) which could lead to thrombotic episodes in clinical usage and dextran sulphate is cytotoxic and any leakage from such an affinity matrix may seriously contaminate the product. The present treatment has advantages over such purification systems in a11 these respects.
While some benefit can be obtained by a batchwise treatment of the initial mixture with the immobilised metal chelate followed by simple washing to remove the most readily eluted factor(s), the most preferred mode of operation is column chromatography. The mixture of factors can thus be applied to the top of the column and an eluant run through to provide fractions containing the separated and partially separated factors. Alternatively, the initial mixture can be-applied batchwise to the immobilised metal chelate prior to loading the column.
The chelated polyvalent metals for use in the process of the invention are generally in the divalent state, Cu 2+ being particularly preferred.
The immobilised chelating agent may comprise a conventional backbone or skeletal support (such as a support based on cellulose, polystyrene, acrylamide, silica, fluorocarbons, cross-linked dextran or cross-linked-agarose) carrying groups capable of chelating polyvalent metals. Such groups are well known and can be found in text books such as Martell and Calvin, Chemistry of the Metal Chelate Compounds, Prentice Hall, Inc. New York (I952).
Iminodiacetic acid groups, -N(CH2COOH)2, are particuarly preferred chelating groups. The chelating groups will generally be spaced from the backbone or skeleton of the support by linear "spacer" groupings such as optionally substituted hydrocarbon or carbohydrate chains, having advantageously between 8 and 16 atoms, e.g. about 12 atoms, between the chelating group and the support. Such spacer groupings may be introduced into appropriate substrates by reaction with a divalent reagent of desired chain length.
The preferred metal binding matrix is Chelating Sepharose which is a beaded 6% agarose product sold by Pharmacia AB which carries iminodiacetic acid groups, linked to the sepharose by a chain:
EMI0007.0012
The binding of proteins to the chelating support is dependent on at least four considerations; <B>(1)</B> ionic strength, where increased-ionic strength apparently minimises non-specific binding;
(ii) the presence of buffer ions both in the sample to be loaded and in the ecruilibrated metal-chelate gel; (iii) loading capacity of the matrix: this is a finite quantity which can be fully saturated by the more tightly binding proteins at the expense of the weaker binding proteins. The capacity of the matrix for a particular protein is thus dependent on the composition and the concentration of the protein mixture with which it is challenged during loading and the adsorption contact time; (iv) with both binding and elution, pH is a recognised effector.
It has been found that at relatively high ionic strengths, the binding to the chelate of prothrombin (Factor II) and thrombin is reduced, while that of Factors IX and X and Protein C is increased. Since, in general, it is desirable to reduce the prothrombin level in concentrates of the other blood clotting factors, in order to avoid the possibility of thrombin formation, this finding is particularly useful.
It is thus preferred to load the metal chelate in an aqueous solution of relatively high ionic strength, for example containing 0.4 to<B>1.0</B> '.M, preferably 0.4 to 0.6M, most preferably about 0.5 M, sodium choride and/or one or more other electrolytes providing a solution of equivalent ionic strength.
Different vitamin R-dependent clotting factors have different pH optima for binding to a chelating gel and consequently selection of the pH of the loading buffer solution provides further means for promoting preferential binding to the chelating gel of one or more vitamin R-dependent clotting factors over another such factor. Thus, for example, in the case of Cu 2+-primed Chelating Sepharose, optimum binding of Factors II, IX and X occurs at pH 6.0- 7.0, above pH 7.0 and pH 6.5-7.5 respectively.
Consequently, for example, to enhance differential binding to such a gel of Factor IX over prothrombin, the loading pH will desirably be pH 7.0 or above.
Further, where it is desired to deplete prothrombin from one or more other vitamin R-dependent clotting factors, decreased binding of prothrombin in favour of other blood clottina factors, e.g. Factor IX, may be promoted by appropriate selection of the adsorption contact time and the loading challenge.
Thus, for optimization of separation of Factor Ix from Factor II on Cu 2+-primed Chelating Sepharose starting with a prothrombin complex concentrate, the adsorption contact time will advantageously be over 20 minutes, more preferably about 40 minutes or longer, and the loading challenge will desirably be such that a Factor IX loading of about 250-450 F.IX:Ag units per ml. of gel is achieved.
Column chromatography is preferred. The initial mixture of factors is preferably applied to the top of a column, but it is also possible to load the initial mixture on to the immobilised chelate batchwise. After washings, mostly at high ionic strength, the remaining factors can be separated either by batch elution or by transferring the gel to a column. for chromatographic elution.
Loading is desirably directly followed by washing with an aqueous solution of the same or similar ionic strength and pH to remove prothrombin and any thrombin which may be present. Advantageously, an acidic pH washing step may additionally be carried out to remove from the chelating gel inter-a-trypsin inhibitor. This protein is recognised as a major unwanted contaminant of Factor IX concentrates prepared by prior art techniques, making up in some cases as much as 30% of the total protein.
It has been found, however, that inter-a-trypsin inhibitor can be eluted from Cu 2+-primed Chelating Sepharose at about pH 4.5 - 5.5 without appreciable loss of Factor IX from the qel. In order to reduce electrolyte levels in the final product, it is convenient to include at least a final washing step at low ionic strength, e.g. equivalent to 100-200 mM NaCl.
The washing procedure allows for some removal of virus by additional washing steps while the protein is immobilised. Furthermore, the immobilised factors may be treated with a virucide such as a detergent at this stage and the virucide can then be removed by further washing prior to the elution step.
It should be noted that the metal chelate material readily binds the protein factors of interest in the presence of detergents and/or solvents used as virucides prior to the initial adsorption according to the invention and can retain these factors when the virucide is removed by washing or indeed, when a virucidal detergent wash is applied prior to elution of the factors. This contasts with many prior affinity columns used to adsorb such proteins, where detergent tends to affect elution of the proteins.
The clotting factors remaining on the column after removal of Factor II or, indeed, such factors applied to the column in the substantial absence of Factor II, may be eluted by a change of pH and/or use of a displacing agent selected from imidazole and amino acids, desirably at significantly lower ionic strength than the loading buffer solution in order to reduce or avoid the need to remove salts from the resulting enriched fractions. Suitable preferred ionic strengths for such elution are in the range 100-200 mM.
There elution of desired blood clotting factors depends on a pH gradient, this may be an increasing pH gradient or a decreasing pH gradient and in either case inter-a-trypsin inhibitor may desirably be depleted from the chelating gel prior to collection of a Factor fraction.
Thus, in the case of use of an increasing pH gradient for elution, it will be appreciated that the starting pH may be as low as about pH 4.5-5.5 to achieve removal of inter-a-trypsin inhibitor, but this will generally he followed by a substantial step upward in pH for elution of desired blood clotting factors.
In particular, for example, it has been found that Factor IX- and Factor X-enriched fractions may be desirably eluted from Cu2+-primed Chelating Sepharose in the presence of a non-amino acid or imidazole containing buffer system at relatively low ionic strength, e.g. 100mM NaCl, by varying the pH linearly from about pH 7.0 to about pH 8.0.
In the case of use of a decreasing pH gradient to elute for example from the same type of gel blood clotting factors of a prothrombin complex concentrate, following washing of the gel at above pH 7.0 at high ionic strength.e.g. 500 mM NaCl, to remove Factor <B>r</B> II, elution of Factor X- and Factor IX-enriched fractions substantially free from inter-a-trypsin may desirably be achieved at relatively low ionic strength, e9. 100 mM NaCl, by stepwise decrease of the pH of the elution buffer in the acidic pH range down to about pH 4.0.
In this case, the following order of protein elution is observed: Factor X/Factor V (pH about 5-5.6) inter- a-trypsin inhibitor/protein C (pH about 4.6), Factor IX (pH about 4.1).
If it is desired to produce purified or enriched Factor II, this may also preferably be eluted from a metal chelate column at low ionic strength in the absence of amino acids.
Where a displacing agent is employed, the concentration of imidazole or amino acid (in particular, glycine, methionine or glutamic acid) in the eluant is preferably in the range 5 to 70 mM. Suitable amino acids for use in such elution include, for example, alanine, phenylalanine, valine, lysine, glycine, methionine and glutamic acid. The last three named amino acids are particularly preferred for this purpose.
A displacing agent containing eluant is preferably buffered to a pH in the range 4 to 9, more preferably 6 to 8, e.g. about 7. Suitable buffer systems include amino alcohol buffers such as Tris and citrate-phosphate buffer. It is found that Tris enables elution of a11 the factors at lower concentrations of displacing agent, i.e. imidazole or amino acid.
In general, whichever eluant system is used, the order of elution, which appears to reflect binding affinity, appears to be Factor II/IIa, Factor X, Factor IX/IXa/Protein C.
In general, elution is preferably carried out in such a way that the eluted protein specific activity (purity and/or potency) is optimised.
In general the concentration of the principal factor of interest, e.g. Factor IX, is preferably at least 30 iu/ml.
After elution, the solution may be freeze dried. For inactivation of virus infections, the freeze dried concentrate may be heated at, for example, 800C. We have found that the purified concentrates according to the invention substantially survive such heating, avoiding significant activation of the respective factors to a greater extent than previous concentrates.
Using the methods of the invention, it has proved possible to prepare for the first time: (a) Factor IX substantially free of Factor II and inter-a-trypsin inhibitor; (b) Protein C substantially free from Factors II. IX and X: (c) Factor X substantially free of Factor II and having specific Factor X activity of at least 13 iu/mg protein; and (d) Factor VII substantially free from Factor II and having a specific Factor VII activity of greater than one unit per mg protein.
<U>Methods of testing</U> The reported incidents of thrombotic episodes following infusion of prior prothrombin complex concentrate has resulted in<U>in vitro</U> and<U>in vivo</U> tests of products for their potential thrombogenic effect. The<U>in vivo</U> tests have been in animals (rabbits, pigs and dogs), using various criteria to asses thrombosis after infusion.<U>In vitro</U> tests are based upon the ability of the concentrate to reduce the clotting time of substrate plasma or fibrinogen. The significance of these tests has not been shown, but great weight is placed upon them in terms of product Quality control.
The three main tests are: NAPTT: Measures level of activated clotting factors in the sample, with an arbitrary lower limit clotting time of 150 seconds. The NAPTT cannot be related conclusively to the concentration of any particular activated clotting factor but indicates the presence of F.IXa and F.Xa.
FCT: Measures the time taken for a fibrinogen solution to clot in the presence of test material. This is a direct measure of thrombin (the penultimate protein in the blood clotting "cascade"). The lower limit clotting time is three hours, which corresponds to 5 x<B>10-3</B> iu/ml or 2 ng/ml of thrombin. TGt <U>50</U>: Measures the time taken to generate a known amount of thrombin. It is a measure of the activation of clotting factors which operate earlier in the process.
The following Examples are given by way of illustration only.
<U>Tests used in the Examples for Quantification of</U> <U>proteins</U> Both functional biological activity and antigenic activity were used to quantitate the proteins of interest.
Biological activity of Factors II, IX and X was measured by established clotting assays. Unitage of clotting activity (:C) was defined using a Working Standard 87/532 which had been calibrated against the 1st International Standard for Factors II, IX and X Concentrates, code 84/681.
Biological activity of Factor<B>VII</B> was measured by established clotting assay or by chromogenic assay using a synthetic substrate. In both cases, unitage was assigned using an in-house Factor VII concentrate working standard which had been calibrated against a human plasma pool of greater than thirty donors, defined as having a Factor VII activity of 1.0 F.VII:C u/ml.
A11 antigen activities(:Ag) for Factors II, V, IX and X, Protein C and Inter-a-Trypsin Inhibitor were measured by Immune Electrophoresis (Laurell's Rocket method) using Normal Pooled Plasma as the standard. In each case, the standard had been assigned a potency of 1.0 unit per ml and sample activities were expressed as plasma equivalent units per ml.
In the present specification , clotting activity or "Factor -:C" is given,in terms of international units per ml or iu/ml. Antigen activity or "Factor -:Ag" is given as units per ml (u/ml) or plasma equivalent units per ml (p.e.u/ml).
As the clotting and antigen activities are measures of different aspects of the protein chemistry, the international unit of clotting activity is not the same as the plasma equivalent unit. Typically, the ratio of iu : p.e.u. for Factor II, Factor TX and Factor X in partially purified protein concentrates is 0.86, 0.6 and 0.63 respectively.
<U>Example 1</U> <U>Preparation of</U> Prothrombin <U>from</U> Prothrombin <U>Complex</U> <U>Concentrate</U> (PCC) Chelating Sepharose was primed with copper ions by passing copper sulphate solution through it. The gel was then washed with citrate-phosphate buffer containing 500 MM NaCl at pH 7.0. PCC containing 500 ?nM NaCl was applied to the column at a loading of 100-200 Factor IX iu per ml gel.
The protein eluate was monitored and the breakthrough protein collected. This contained prothrombin at potencies of 5-25 iu/ml with a specific activity of 5 iu/mg protein.
This material was at least<B>70%</B> prothrombin. <U>Example 2</U> <U>Effect of ionic strength on binding of Factors</U> <U>II, IX, X and Protein C to</U> copper-chelate <U>gel.</U> PCC was dialysed against citrate-phosphate buffer pH 7.0 containing 100, 250, 500 or 1000 mM NaCl. This was then loaded on to a Cu 2+-primec Chelating Sepharose column, which had been washed with the same ionic strength buffer.
After loadii at 430 units per ml of gel, the column was further washed with the same buffer and the eluted unbounc protein assayed for Factors II, IX, X and Protein C. Table I below shows the results.
EMI0016.0012
<U>Table <SEP> T_</U>
<tb> Adsorption <SEP> Unbound <SEP> units <SEP> per <SEP> ml <SEP> of <SEP> gel
<tb> [NaClI <SEP> (MM)
<SEP> F.II <SEP> F.IX <SEP> F.X <SEP> PC
<tb> 100 <SEP> 162 <SEP> 64 <SEP> 82 <SEP> 14
<tb> 250 <SEP> 155 <SEP> 64 <SEP> 87 <SEP> 13
<tb> 500 <SEP> 138 <SEP> 49 <SEP> 74 <SEP> 10
<tb> 1000 <SEP> 95 <SEP> 35 <SEP> 55 <SEP> 7 Maximum differential binding of Factor IX over Factor II was observed at 500 mM NaCl. <U>Example 3</U> <U>Effect of adsorption pH on binding of Factors II,</U> <U>IX and X to</U> copper-chelate <U>gel.</U>
Chelating Sepharose was first primed with copper ions and then washed with a citrate-phosphate buffer system containing 500 mM NaCl with pH adjusted to the indicated level. PCC, containing 500 mM NaCl, was titrated to the same pH and then loaded on to the gel at about 300 Factor IX units per ml of gel. The column was further washed with the same buffer and the breakthrough/unbound protein collected and measured for Factors II, IX and X. Table II below shows the results.
EMI0017.0012
<U>Table <SEP> II</U>
<tb> Unbound <SEP> units <SEP> per <SEP> ml <SEP> of <SEP> gel
<tb> Adsorption <SEP> pH <SEP> F.II <SEP> F.IX <SEP> F.X
<tb> 5.5 <SEP> 231 <SEP> 69 <SEP> 121
<tb> 6.0 <SEP> 227 <SEP> 64 <SEP> 93
<tb> 6.5 <SEP> 189 <SEP> 64 <SEP> 74
<tb> 7.0 <SEP> 234 <SEP> 71 <SEP> 69
<tb> 7.5 <SEP> 285 <SEP> 43 <SEP> 75
<tb> 8.2 <SEP> 328 <SEP> 59 <SEP> 155 Optimum binding of Factor II was achieved between pH 6.0-7.0.
Optimum binding of Factor IX was achieved above pH 7.0.
Optimum binding of Factor X was achieved between pH 6.5 and pH 7.5. Adsorption pH can therefore be used to optimise the binding of a preferred clotting factor to the gel. <U>Example</U> 4- Preparation <U>of a concentrate of Factor X and Factor</U> <U>IX with Reduced</U> Prothrombin A Chelating Sepharose column was prepared and loaded as in Example 1. After application of the PCC, the column was washed with citrate- phosphate buffer containing 500 mM NaCl at pH 7.0.
The ionic strength was then reduced by washing with citrate-phosphate buffer containing 100 mM NaCl. The Factor X fraction was then eluted with citrate-phosphate buffer containing 100 mM NaCl and 5 mM glycine. This contained Factor X at 15 40 Factor X iu/ml and less than 0.01 unit prothrombin per unit of Factor X. The Factor X specific activity was 13 iu/mg protein and was therefore<B>7%</B> pure.
This fraction also contained Factor IX in approximately equimolar amounts ( 1 iu Factor IX per unit Factor X). <U>Example 5</U> <U>Preparation of Factor X with reduced</U> Prothrombin <U>and Factor IX</U> A Chelating Sepharose column was primed and loaded as described in Example 1. After application of the sample, the column was washed with citrate- phosphate buffer containing 500 mM NaCl at pH 7.0.
The Factor X_ was then eluted with citrate-phosphate buffer at pH 7.0 containing 500 mM NaCl and 5 mM glycine. This material contained 15-30 Factor X iu/ml with a specific activity of 14 iu/mg protein. This material differed from that obtained in Example 4 by having higher prothrombin activity and lower Factor IX activity. Here, 0.25 iu/unit Factor X was obtained for both prothrombin and Factor IX.
Therefore, the prothrombin still constituted about 40-50% by weight of the total protein.
<U>Example 6</U> <U>Preparation of a concentrate of Factor IX and Factor</U> <U>X with reduced</U> Prothrombin A Chelatinq Sepharose column was primed and loaded as described in Example 1. It was then washed sequentially with citrate-phosphate buffer containing first 500 mM and then<B>100</B> _MM NaCl at pH 7.0.
Both Factor IX and Factor X were then eluted with Tris-glycine buffer containing<B>100</B> mM NaCl, at potencies of 30 iu/ml. Prothrombin was reduced to 0.03 iu per unit of Factor IX or Factor X.
<U>Example 7</U> <U>Preparation of a concentrate of Factor IX and Protein</U> <U>C with reduced</U> Prothrombin <U>and Factor X</U> A Chelating Sepharose column was prepared and washed as described in Example 4. After elution of protein in 5 mM glycine, Tris-glycine buffer was used to elute Factor IX.
It was found that Tris was suitably in the range 10-30 mM, glycine in the range 40-70 mTA and NaCl 100 mM. This eluted Factor IX at potencies of 30-50 iu/ml. Prothrombin was undetectable and Factor X showed less than 0.01 unit per unit of Factor IX (less than 0.050 by weight). Protein C was also eluted at concentrations of 25-40 plasma equivalent units per ml.
<U>Example 8</U> <U>Separation of Protein C from</U> Prothrombin <U>and Factor X</U> Purification of Protein C is complicated by co-elution of prothrombin and Factor X in conventional chromatographic systems such as heparin-Sepharose or Dextran-Sulphate-Sepharose. However, in these systems, Factor IX contamination is minimal due to tighter binding of that protein. The metal chelate system described here can therefore be used as a step in the purification of Protein C.
The preparation. and washing of a Chelating Sepharose column was as described in Example 7. As protein binding is concentration dependent, loading was as high as possible, up to 160 Factor X iu per ml gel or 240 prothrombin iu per ml of gel. The column was then washed and eluted as described in Example 7, with the Protein C eluting with the final Tris-glvcine buffer eluant, separated from prothrombin and Factor X contaminants.
<U>Example 9</U> <U>Effect of adsorption contact time on binding and</U> <U>recovery of Factor II, Factor IX, Factor X and</U> <U>Protein C from Cu</U> 2+-primed ChelatinQ Sepharose. Chelating Sepharose was packed into columns, primed with copper ions and then washed with citrate- phosphate buffer pH 7.0 containing 500 mM NaCl.
PCC, containing 500 mM NaCl, was loaded on to each column at a controlled flow rate so that the contact time with the gel could be varied in different columns. The columns were loaded at 430 Factor IX units per ml gel, washed with the 500 mM NaCl buffer, then with 100 mM NaCl buffer and then with lOmM Tris, 70 mM glycine, 100 mM NaCl buffer pH 7.0.
The breakthrough and the Tris-glycine eluates were measured for Factor II, Factor IX, Factor X and Protein C. The.results are shown in Table III below.
EMI0021.0028
<U>Table <SEP> III</U>
<tb> Contact <SEP> Time <SEP> Unbound <SEP> units <SEP> per <SEP> Eluted <SEP> units
<tb> (mins) <SEP> ml <SEP> gel <SEP> per <SEP> ml <SEP> gel
<tb> F.II <SEP> F.IX <SEP> F.X <SEP> PC <SEP> F.
<SEP> II <SEP> F.IX <SEP> F.X <SEP> PC
<tb> 10 <SEP> 172 <SEP> 156 <SEP> 153 <SEP> 27 <SEP> 91 <SEP> 231 <SEP> 135 <SEP> 66
<tb> 20 <SEP> 198 <SEP> 151 <SEP> 126 <SEP> 26 <SEP> 64 <SEP> 216 <SEP> 91 <SEP> 68
<tb> 40 <SEP> 189 <SEP> 63 <SEP> 66 <SEP> 13 <SEP> 39 <SEP> 242 <SEP> 160 <SEP> 78
<tb> 70 <SEP> 210 <SEP> 57 <SEP> 57 <SEP> 11 <SEP> 37 <SEP> 227 <SEP> 142 <SEP> 77
<tb> 110 <SEP> 247 <SEP> 72 <SEP> 105 <SEP> 18 <SEP> 23 <SEP> 266 <SEP> 126 <SEP> 93 Binding of protein to the gel is therefore dependent on contact time. This can be optimised for Factors IX, X and Protein C at contact times of forty minutes or greater. Increased contact time results in less Factor II binding to the gel.
<U>Example 10</U> <U>Effect of</U> loading <U>challenge on the binding and recovery</U> <U>of Factor II, Factor IX, Factor X and Protein C with</U> <B><U>Cu</U></B> 2+-primed Chelating Sepharose.
Chelating Sepharose was packed into several columns, primed with copper ions and washed with citrate phosphate buffer pH 7.0 containing 500 mM NaCl. PCC, containing 500 m\i NaCl, was loaded on the columns in different amounts, the contact time for each column being maintained at 40 minutes by varying the flow rate through the column. The columns were then washed with buffers as described in Example 6. The results are shown in Table IV below.
EMI0022.0013
<U>Table <SEP> IV</U>
<tb> Loading <SEP> Unbound <SEP> units <SEP> Eluted <SEP> units
<tb> F.IX:Ag <SEP> units <SEP> per <SEP> ml <SEP> gel <SEP> per <SEP> ml <SEP> gel
<tb> per <SEP> ml <SEP> gel <SEP> F.II <SEP> F.IX <SEP> F.X <SEP> PC <SEP> F.II <SEP> F.IX <SEP> F.X <SEP> PC
<tb> 43 <SEP> 6 <SEP> 0 <SEP> 0 <SEP> 0 <SEP> 13 <SEP> 30 <SEP> 23 <SEP> 9
<tb> 88 <SEP> 13 <SEP> 1 <SEP> 2 <SEP> 0 <SEP> 23 <SEP> 76 <SEP> 46 <SEP> 19
<tb> 173 <SEP> 54 <SEP> IS <SEP> 18 <SEP> 2 <SEP> 42 <SEP> 112 <SEP> 85 <SEP> 37
<tb> 258 <SEP> 115 <SEP> 49 <SEP> 43 <SEP> 11 <SEP> 45 <SEP> 165 <SEP> 100 <SEP> 52
<tb> 344 <SEP> 185 <SEP> 55 <SEP> 52 <SEP> 11 <SEP> 31 <SEP> 186 <SEP> 115 <SEP> 62
<tb> 430 <SEP> 189 <SEP> 63 <SEP> 66 <SEP> 13 <SEP> 39 <SEP> 242 <SEP> 160 <SEP> 78
<tb> 513 <SEP> 305 <SEP> 175 <SEP> 128 <SEP> 42 <SEP> 26 <SEP> 256 <SEP> 136 <SEP> 79
Increased loading displaces Factor II preferentially. Loading challenge can be used to enhance the binding and recovery of one particular protein relative to the others. Improved separation of Factor IX from Factor II was achieved at loadings of 258-430 F.IX:Ag units per ml of gel (equivalent to approximately 150-250 F.IX:C iu per ml of gel).
<U>Example 11</U> <U>Effect of Factor IX concentration on binding and</U> <U>recovery of Factor II, Factor IX, Factor X and Protein</U> <U>C</U> a=ith Cu2+-primed Chelatina Sepharose.
Chelating Sepharose was primed with copper ions and washed as described in previous examples. PCC containing 500 mM NaCl was diluted in citrate- phosphate buffer pH 7.0 containing 500 mM NaCl. Columns were loaded at different dilutions and washed with buffers as described in Example 6. Total loading challenge (approx. 250 units per ml of gel) and contact times were kept constant. The results are shown in Table V below.
EMI0023.0017
<U>Table <SEP> V</U>
<tb> Loading <SEP> Unbound <SEP> units <SEP> per <SEP> Eluted <SEP> units <SEP> per
<tb> potency <SEP> ml <SEP> gel <SEP> ml <SEP> gel
<tb> F. <SEP> IX: <SEP> AcT <SEP> F. <SEP> II <SEP> F. <SEP> IX <SEP> F. <SEP> X <SEP> PC <SEP> F. <SEP> II <SEP> F. <SEP> IX <SEP> F. <SEP> X <SEP> PC
<tb> u/ml <SEP> gel
<tb> 173 <SEP> 180 <SEP> 75 <SEP> 140 <SEP> 17 <SEP> 15 <SEP> 200 <SEP> 142 <SEP> 70
<tb> 86 <SEP> 195 <SEP> 80 <SEP> 145 <SEP> 20 <SEP> 10 <SEP> 170 <SEP> <B>125</B> <SEP> 65
<tb> 43 <SEP> 245 <SEP> 140 <SEP> 185 <SEP> 35 <SEP> 7 <SEP> 185 <SEP> 115 <SEP> 55 Higher concentrations of the loaded material are preferred for binding to the chelate to occur.
<U>Example 12</U> <U>Effect of</U> chelating <U>gel on Factor IX binding and</U> <U>recovery</U> Equal volumes of five chelating gel types were packed into five columns, primed with Cu2+ions and washed with citrate-phosphate buffer pFt 7.0 containing 500 mm NaCl. PCC, adjusted to contain 500 PM NaCl, was loaded in equal amounts on to each column, followed by washing sequentially with the same buffer,
buffer containing 100 mM NaCl and then elution with 10 mt'@. Tris, 70 mM glycine buffer pH 7.0. Factor IX was measured in the breakthrough, wash and eluate fractions.
Gel types used were: 1. Chelating Sepharose comprising iminodiacetic acid attached to SeLharOSe by a 12 atom (10 carbon) spacer molecule [Pharmacia].
2. As for 1 but with additional cross-linking for Fast Flow Sepharose.
3. Agarose carrying iminodiacetic acid attached by a 7:2 atom (10 carbon) spacer molecule [Sigma] 4. Iminodiacetic acid directly attached to a styrene-divinyl benzene cross-linked polymer [Chelex 100, BioRad].
5. As 4 but with a coarse mesh, non-cross-linked matrix [Chelex 20, BioRad]. Table VI below shows the results.
EMI0025.0001
<U>Table <SEP> VI</U>
<tb> Factor <SEP> IX <SEP> % <SEP> of <SEP> loaded
<tb> Gel <SEP> Type <SEP> Breakthrough <SEP> Wash <SEP> Eluate
<tb> 1 <SEP> 8.9 <SEP> 4.6 <SEP> 60.2
<tb> 2 <SEP> 14.7 <SEP> 13.3 <SEP> 43.4
<tb> 3 <SEP> 12.1 <SEP> 60.2 <SEP> 20.9
<tb> 4 <SEP> 76.0 <SEP> 0.2 <SEP> 0.2
<tb> 5 <SEP> 93.0 <SEP> 2.1 <SEP> 0.4 Factor IX binding to Cu 2+-primed iminodiacetic groups is increased by provision of a spacer molecule between the matrix and the chelating group and by the use of
non-cross-linked Chelating Sepharose. <U>Example 13</U> <U>Preparation of Factor IX using Cu</U> 2+-primed Chelating Sepharose <U>in a batch adsorption.</U>
Chelating Sepharose was washed with water, then suspended in a copper sulphate solution (5mg/ml) and mixed for 20-30 minutes. The Cu 2+-primed gel was recovered from the mixture by centrifugation and washed extensively with citrate-phosphate buffer pH 7.0 containing 500 -T-M. NaCl. The gel was recovered as above, resuspended in a solution of PCC containing 500 mM NaCl and mixed for 20-30 minutes.
The ratio of PCC to gel was approximately 150 Factor IX:C units per ml of gel. The gel was recovered by centrifugation or gravity settling and the supernatant removed. Sequential washes with citrate-phosphate buffer pH 7.0 containing 500 mM NaCl, then 100 mM NaCl, removed weakly-bound protein and reduced the ionic strength. Factor X could then be removed by washing with 10 mM Tris, 5 mM glycine, 100 mM Nacl pH 7.0.
Factor IX was subsequently eluted by washing the gel with 10 mM Tris, 70 mM glycine, 100 mM NaCl, pH 7.0.
<U>Example 14</U> <U>Recovery of Factor IX from</U> Cu2+-primed Chelating Sepharose <U>using different amino acid</U> elution <U>buffers</U> Columns of Chelating Sepharose were primed with Cu2+ ions and washed as described in previous examples.
PCC (containing 500 mM NaCl) was loaded on to each column, followed by washing with citrate-phosphate buffer pH 7.:0. Factor IX was recovered by eluting with 10 mM Tris, pH 7.0, containing at 70 mM an amino acid selected from (a) non-polar amino acids (e.g.
alanine, valine, phenylalanine, methionine), (b) polar uncharged amino acids (e.g. glycine, serine, tyrosine, glutamine) , (c) polar negatively charged amino acids (e.g. glutamic acid) or (d) polar positively charged amino acids (e.g.
lysine, arginine). Methionine, glycine and glutamic Acid were found to be particularly advantageous, these giving Factor IX yields of areater than 80% and Factor IX specific activities of greater than 5 iu/mg. <U>Example 15</U> <U>Elution of Factor X and Factor IX from</U> Chelating Sepharose <U>using</U> Methionine A column of Cu 2+-primed Chelating Sepharose was washed with citrate phosphate buffer pH 7.0 containing
500 mM NaCl. PCC was applied to the column, which was then washed with the same buffer before reducing the ionic strength by washing with a buffer containing 100 mM NaCl. Factor X was then eluted with Tris buffer pH 7.0 containing 2 mM - 5 mM methionine.
The Factor X potency was greater than 15 u/ml and the specific activity was approximately 30 units/mg. Factor IX was reduced in the Factor X fraction to about 0.5 unit Factor IX per unit Factor X and prothrombin was present at less than 0.1 unit per unit Factor X.
Factor Ix was found to be eluted by washing with Tris buffer containing 10-15 mm methionine. The Factor IX potency was greater than 15 u/ml and it had a specific activity of approximately 20 units/mg. Prothrombin was undetectable in the Factor IX fraction and Factor X was present at less than 0.1 Factor X unit per unit of Factor IX.
<U>Example 16</U> Elution <U>of Factor X and Factor IX from</U> Chelating Sepharose <U>using</U> Glutamic <U>Acid</U> A Cu 2+-primed Chelating Sepharose column was prepared, washed and loaded with PC.C and washed as described for methionine elution (Example 15).
Factor X was eluted with 10 mM Tris buffer containing 5 mM - 10 mM glutamic acid. The Factor X had a specific activity of 16-22 units/mg.
Factor IX, with a specific activity of approximately 15u/mg was then eluted by raising the glutamic acid concentration in the eluting buffer to 40 mM. <U>Example 17</U> <U>Preparation of Factor</U> IX <U>and Factor X using an</U> increasing pH <U>Gradient to</U> elute <U>from</U> Chelating Sepharose A column of Chelating Sepharose was primed with Cu 2+ ions and washed with buffer at pH 7.0 containing 500 <RTI
ID="0028.0014"> mM NaCl. PCC containing 500 mM NaCl was loaded on to the column, which was then washed with buffer. Ionic strength was reduced to 100 mM NaCl using low salt buffer at pH 7.0. Some Factor X was eluted by this buffer with a specific activity of 10.5 u/mg.
The column was progressively eluted by a pH gradient from pH 7.0 to pH 8.5. Factor IX and Factor X were eluted between pH 7.1 and pH 7.9 with specific activities of 27 u/mg and 6 u/mg respectively.
<U>Example 18</U> <U>Preparation of Factors IX, X and Protein C using</U> pH <U>and</U> alvcine <U>to elute</U> A column of Chelating Sepharose was primed with Cu 2+ ions and washed with citrate-phosphate buffer, pH 7.5. PCC containing<B>500</B> mM NaCl was adjusted to pH 7.5 and loaded on to the column, which was washed with buffer.
Ionic strength was then reduced by washing with citrate-phosphate buffer pH 7.5 containing 100 m_M NaCl. This eluted Factor X with a specific activity of 18 u/mg in a mixture with Factor IX, which had a specific activity of 9 u/mg. A combined Protein C (15 u/ml; 6.5 units per mg protein) and Factor IX (12 u/ml; 5 units per mg protein) fraction was then collected by washing with 10 mM Tris, 5 mm glycine, 100 mm NaCl buffer, pH 8.0.
<U>Example 19</U> <U>Elution of protein components from</U> Chelatina Sepharose <U>using a decreasing pH gradient</U> A column of Chelating Sepharose was primed with <B>Cu</B> 2+ ions and washed with citrate-phosphate buffer pH 7.5 containing<B>500</B> mM NaCl. PCC was then applied after adjustment of pH to 7.5 and NaCl to 500 mM. This was followed by use of the above buffer as a wash to remove Factor II. 30% of the Factor V is the starting material was also removed in this wash.
Ionic strength was then reduced by washing with buffer containing 100 mM NaCl. This also removed Factor X at a specific activity of abour 15 units per mg protein. Using citrate-phosphate buffers, a pH gradient was generated during washing of the column from pH 7.5 to pH 4.0. It was found that fractions could be collected which were rich in selected proteins. Thus, for example, at about pH 5.5, Factor X was further removed while between pH 5 and pH 5.6, a further 40% of the bound Factor V was removed from the column at about five plasma equivalent units per ml.
In this pH range, Inter-a-Trypsin Inhibitor also started to elute, as did Protein C. As the buffer pH fell to pH 4.6, Inter-a-Trypsin Inhibitor and Protein C elution continued with specific activities of 3.0 and 4.6 plasma equivalent units per mg protein respectively. Only small amounts of Factor IX were eluted during these segments of the descending pH gradient.
When the pH was then reduced further, for example to pH 4.1, Factor IX was eluted at potencies greater than 40 units per ml and specific activities of greater than 30 units per mg protein. This fraction contained no detectable Factor II or Factor X. Protein C was present at abour 0.01 units per unit of Factor IX, Factor V was present at 0.06 units per unit of Factor IX and Inter-a-Trypsin Inhibitor was present at 0.03 units per unit of Factor IX.
In terms of their contamination of one unit of Factor IX, this represents a reduction in Protein C, Factor V and Inter-a-Trypsin Inhibitor of 95$, 63% and 90$ respectively over the starting material (PCC) .
<U>Example 20</U> <U>Use of pH and amino acid to separate protein components</U> from Chelating Sepharose A column of Chelating Sepharose was primed with Cu 2+ ions and washed with citate-phosphate buffer pH 7.5 containing 500 mM NaCl. PCC, adjusted to contain 500 mM NaCl and with a pH of 7.5, was loaded on to the column, which was then washed with the buffer described above.
A further wash with citrate-phosphate buffer pH 7.5 containing 100 mM NaCl reduced the ionic strength and removed Factor X, which could be collected at a potency of at least 6 u/ml and a specific activity of at least 15 units per mg protein. The pH was then reduced by washing with a third buffer, typically citrate-phosphate pH 4.5 containing 100 mM NaCl. This removed at least 55% of the remaining bound Inter-a-Trypsin Inhibitor and 24% of the remaining bound Protein C, but only<B>5%</B> of the bound Factor IX.
The column was then re-equilibrated in a buffer having a pH and salt concentration suitable for Factor IX elution, typically pH 7.0, 100mM NaCl.
The Factor IX was then eluted with an amino acid eluant buffer, e.g. glycine in Tris buffer containing 100 mM NaCl as described in previous examples.
This yielded Factor IX at 200 u/ml with a specific activity of greater than 50 units per mg protein. Protein C was also present with a specific activity of 20 plasma equivalent units per mg of protein.
Compared to the starting PCC material, the Factor IX fraction contained s?gnificantly reduced amounts of contaminating proteins. Factor II, Factor X, Factor V and Inter-c-Trypsin Inhibitor were present per unit of Factor IX at levels of 0.0005, 0.005, 0.002 and 0.04 units respectively. This compared with values of 1.0, 1.0, 0.2 and 0.4 respectively in the PCC starting material.
<U>Example 21</U> <U>Preparation of Factor VII on Cu</U> 2+-primed Chelating Sepharose A column of Chelating Sepharose was charged with Cu 2+ ions, then washed with citrate-phosphate buffer pH 7.0, containing 500 m_M NaCl. Factor VII concentrate, prepared by elution of DEAE-Sepharose adsorbed with cryosupernatant or Factor IX-depleted supernatant,
was adusted to contain 500 mM NaCl and loaded on to the column at 50-150 Factor VII units per ml of gel. After washing with the above buffer, the ionic strength was reduced by washing with citrate-phosphate buffer pH 7.0, containing 100 mM NaCl. The Factor VII was recovered with a specific activity of greater than one unit per mg by washing the column with 10 mM Tris,
60 mM glycine at pH 7.0 containing 100 mM NaCl. <U>Example 22</U> <U>Detergent treatment of blood clotting factors</U> <U>while adsorbed on</U> Chelatina Sepharose Chelating Sepharose was charged with Cu 2+ ions washed and then loaded as described in Example 1.
The column was then washed with citrate-phosphate buffer at pH 7.0 containing 500 mM NaCl. The column was then washed with at least four bed volumes of the same buffer containing 1% by weight of sodium docecyl sulphate to solubilise lipid components bound to the gel. Next, the-detergent was washed from the gel with citrate-phosphate buffer containing 100 mM NaCl. This stage also served to reduce the ionic strength by desalting.
Factor IX was then eluted from the gel with glycine in Tris buffer as described in Example 7. Using the same procedure, the sodium dodecyl sulphate may be replaced by the same weight of cholic acid or polyoxyethylene (20) mono-oleate (Tween 80). <U>Example 23</U> <U>Freeze-drying and subsequent heat-treatment of</U> <U>the final fractions of Examples l and 4-7</U> Eluates of the above specified Examples were freeze-dried without loss of activity, and then heated at 800C for 72 hours to reduce potential viral infectivity.
Unlike existing PCC, no thrombin generation was observed upon heating. Factors II, IX and X showed different stabilities to heating in different media. If eluates were diluted into citrate and water, heating yields of 65-90% of the unheated activity were obtained. Dilution into Tris resulted in much greater loss of activity, which was greater for Factor X than for prothrombin and greater for Factor IX than for Factor X. This general effect can be used to advantage as described below in Examples 24 and 25.
<U>Example 24</U> <U>Reduced Factor IX activity in a concentrate of</U> <U>Factor X</U> Example 4 and Example 6 above describe preparation of a Factor X concentrate with Factor IX also present. The Factor IX activity can be preferentially reduced by dilution of the relevant Factor X fractions into 50 mM Tris buffer pH 7.0 prior to freeze-drying. Alternatively, to avoid dilution of the eluate Factor X potency, the Tris in the eluting buffer can be increased to 50mM so that the Factor X is eluted in its formulation buffer. After freeze- drying, this material can be heated at 80 C for 72 hours.
Factor IX activity will be reduced more than Factor X so that the product may be used with greater specificity as a Factor X concentrate. <U>Example 25</U> <U>Heat-treatment of high</U> purity <U>Factor IX concentrate</U> In order to attain the maximum possible yields of Factor IX activity after heating, the Factor IX eluate described in Example 7 was diluted to the desired final potency in citrate or citrate- phosphate buffers or in water, freeze-dried then heat-treated at 800C for 72 hours.
The eluting buffer contained Tris, but the effect of this was minimised by selection of fractions with sufficiently high Factor IX potency to permit the effective diluting out of the Tris component into the final buffer. Alternatively the formulation of the eluting buffer may be modified to allow direct freeze-drying of the undiluted eluate.