Sulfated cellulose hydrate membrane, method for producing same, and use of the membrane as an adsorption membrane for a virus purification process

Abstract

The present invention relates to a sulfated cellulose hydrate membrane, a method for the preparation thereof and the use of the membrane as adsorption membrane for the purification of viruses.

Claims

1. A sulfated cellulose hydrate membrane which is a sheetlike adsorption membrane defining two outer main surfaces, the sulfated cellulose hydrate membrane comprising a crosslinked cellulose hydrate matrix with pores which extend from one outer main surface to an opposing outer main surface of the membrane, wherein the cellulose hydrate membrane has sulfate ligands on its inner surfaces and on its outer main surfaces for adsorptive substance separation wherein the degree of sulfation of the cellulose hydrate matrix is more than 20% by weight.

2. The sulfated cellulose hydrate membrane a claimed in claim 1, wherein the mean pore size of the membrane is between 0.5 and 5.0 ?m.

3. The sulfated cellulose hydrate membrane as claimed in claim 1, wherein the degree of crosslinking of the cellulose hydrate matrix is 0.05 to 0.5.

4. A method for preparing a sulfated cellulose hydrate membrane as claimed in claim 1, comprising the steps of: providing a cellulose membrane with a pore size of 0.1 to 20 ?m; crosslinking the cellulose hydrate matrix using a crosslinker having at least two functional groups in the molecule which react with the hydroxyl groups of the cellulose hydrate matrix; and sulfating the crosslinked cellulose hydrate matrix.

5. The method as claimed in claim 4, wherein the crosslinker is selected from the group consisting of diepoxide compounds, diisocyanates, epichlorohydrin, epibromohydrin, dimethylurea, dimethylethyleneurea, dimethylchlorsilan, bis (2-hydroxyethylsulfone), divinylsulfone, alkylene dihalides, hydroxyalkylene dihalides and diglycidyl ethers or a mixture thereof.

6. The method as claimed in claim 4, wherein 1, 4-butanediol diglycidyl ether or ethylene glycol diglycidyl ether is used as crosslinker.

7. The method as claimed in claim 4, wherein the concentration of the crosslinker in the crosslinking solution of 10 to 30% by weights.

8. The method as claimed in claim 4, wherein the sulfation is effected by reacting the crosslinked cellulose hydrate matrix having a Lewis base-SO.sub.3 complex.

9. The method as claimed in claim 4, wherein the sulfation is effected by reaction with a SO.sub.3-pyridine complex.

10. The method as claimed in claim 9, wherein the concentration of SO.sub.3-pyridine-complex in the sulfation solution is 1 to 40% by weight.

11. The method as claimed in claim 4, wherein the sulfation is effected at a temperature of 20 to 90? C.

12. A method of purifying viruses or virus fragments comprising: preparing a sulfated cellulose hydrate membrane as claimed in claim 1; and contacting the membrane with a solution containing the viruses or virus fragments.

13. The method of claim 12 wherein the viruses have a molecular mass of greater than 10.sup.7 Da.

14. The method of claim 12 wherein the viruses are influenza viruses.

Description

DESCRIPTION OF THE DRAWINGS

(1) The present invention is illustrated in detail by means of the following working examples and FIGS. 1 to 3. These show

(2) FIG. 1 the influence of the degree of crosslinking on the degree of sulfation of the membrane according to the invention, on the permeability for RO water and on the binding capacity for lysozyme,

(3) FIG. 2 the dynamic binding capacity for influenza viruses and lysozyme, and

(4) FIG. 3 breakthrough curves of sulfated polysaccharide gels from the prior art and of an inventive sulfated cellulose membrane loaded with HCP solution (host cell protein solution).

DESCRIPTION OF THE INVENTION

Working Examples

(5) Five different membranes were functionalized in order to illustrate the influence of crosslinking. Two different sulfation methods were applied both to crosslinked and non-crosslinked cellulose hydrate membranes (comparative examples) with the same pore size (samples 1 to 4).

(6) A further sample was prepared which was crosslinked according to example 1 and sulfated according to example 2 (sample 5). This sample differs from samples 1 to 4 in that a narrower starting membrane with a mean pore size of 1.2 ?m was used for the sulfation. All membrane samples were characterized according to examples 5 to 8. The results are summarized in table 2. The Sartobind? S sample is a cellulose hydrate membrane with sulfonic acid ligands.

(7) FIG. 1 shows that the increase in the degree of crosslinking leads to an increase in the degree of sulfation but also to an increase in the flow rate of RO water. In addition, the sulfated membranes crosslinked according to the invention have a higher selectivity for influenza viruses compared to contaminants, since they show experimentally a significantly lower dynamic binding capacity for small negatively charged proteins such as lysozyme. The degree of sulfation in FIG. 1 is stated in % by weight. Table 1 comprises the details of the composition of the crosslinking solutions which were used for preparing the membranes of FIG. 1.

(8) FIG. 2 shows that the membranes of samples 2, 4 and 5 according to the invention have distinctly better binding properties, since they have a higher dynamic binding capacity for influenza viruses, than the membranes of samples 1 and 3 which were selected as comparative examples. The membrane of sample 5 has a higher dynamic binding capacity for influenza viruses by at least a factor of 10, compared to samples 1 and 3. In addition, in the membranes of samples 2, 4 and 5, the binding capacity for small, positively charged protein contaminants such as lysozyme is only ? of that of the non-crosslinked samples 1 and 3 and the Sartobind? 5 membrane which has cation exchanging sulfonate groups as ligands. The membrane of sample 5 according to the invention therefore has the highest selectivity for binding of influenza viruses of all samples investigated.

(9) The descriptor E+12 in FIG. 2 means ?10.sup.12.

(10) The breakthrough curves with HCP solution (host cell protein solution) presented in FIG. 3 show that the sulfated membrane of sample 4, unlike conventional sulfated polysaccharide gels, practically does not bind any contaminants such as host cell proteins (HCP). In this case, Cellufine? Sulfate from JNC Corporation and Capto? DeVirS from GE Healthcare were used for comparison. In the membrane of sample 4 according to the invention, an immediate breakthrough of host cell proteins (HCP) occurs, which proves that this membrane does not bind any HCP. In the sulfated gels regarded as comparative examples, the UV signal of the HCP solution originally used is only achieved again after 100 ml colume volume (CV [ml]=column volume in ml), which proves that the sulfated gels bind these contaminants to a considerable extent. The property of the membranes according to the invention described above is of major advantage in the purification of viruses, since contaminants such as HCP cannot bind to the membrane material and, after elution of the virus, are therefore not found in the eluate. For this reason, further contamination depletion steps can be omitted and the total costs of the purification process are significantly reduced. It is also clearly apparent here that the higher sulfate ligand density in the membrane according to the invention has no negative effect on the binding of contaminants and, in a desired manner, selectively improves only the binding capacity for influenza viruses.

Example 1: Crosslinking of Cellulose Microfilter Membranes

(11) The crosslinking reaction takes place at 20 to 25? C. 10% aqueous sodium hydroxide solution is added to a precharged amount (in g) of reverse osmosis water and stirred. 1,4-Butanediol diglycidyl ether (Sigma-Aldrich, Cat. No.: 220892) is then added and stirred until a homogeneous solution is formed. The degree of crosslinking can be controlled by the concentration of 1,4-butanediol glycidyl ether in the crosslinking solution. The respective amounts of reverse osmosis water, aqueous sodium hydroxide solution and 1,4-butanediol diglycidyl ether (BUDGE) in the crosslinking solutions prepared here and the resulting degree of crosslinking of the membrane according to the invention are summarized in table 1.

(12) TABLE-US-00001 TABLE 1 Composition of the crosslinking solutions and resulting degree of crosslinking 10% NaOH/ ROW/ BUDGE Degree of [g] [g] [g] crosslinking 6.0 94.0 0.0 0 6.0 93.0 1.0 0.02 6.0 92.0 2.0 0.03 6.0 89.0 5.0 0.09 6.0 84.0 10.0 0.12 6.0 79.0 15.0 0.19

(13) A DIN A4 sheet of dry cellulose microfilter membrane from Sartorius Stedim Biotech GmbH (type 10142V) is completely wetted with the crosslinking solution and then stored for seven days in an airtight polyethylene bag. After completion of the crosslinking reaction, the microfilter membrane is rinsed with running reverse osmosis water for 10 min and then dried in a convection oven at 80? C. for 10 to 12 minutes.

(14) To determine the degree of crosslinking as defined above, the following procedure is carried out: the non-crosslinked cellulose membranes of type 10142V are dried in a Sartorius moisture analyzer balance (type MA100) at 110? C. for 5 min and their weight m.sub.A is determined. Then, the membranes are wetted with the crosslinking solution. An additional blank sample, which is not treated with crosslinker, is weighed to determine the weight loss due to leachable constituents. After a crosslinking time of 7 days, the membranes are rinsed for 10 minutes with running RO water and then pre-dried in a convection oven at 80? C. for 10 min and finallyas the starting membranesdried on the moisture analyzer balance for 5 min and m.sub.V is determined. By means of the blank sample untreated with crosslinker, the mass of leachable constituents m.sub.B is determined. The percentage increase in mass ?M, required for the value defined, arises as follows:
?M=100%*((m.sub.V?m.sub.A?m.sub.B)/m.sub.A?m.sub.B)),
where
m.sub.V mass of the crosslinked membrane
m.sub.A mass of the non-crosslinked membrane and
m.sub.B mass of the leachable constituents of the blank sample.

Example 2: Sulfation with 2-Pyrrolidone and Pyridine-Sulfur Trioxide Complex

(15) 2 g of pyridine-sulfur trioxide complex (Sigma-Aldrich, Cat. No. 84737-500g) are added to 8 g of 2-pyrrolidone with stirring. The reaction vessel is tightly sealed. The mixture is then temperature controlled at 70? C., whereupon the pyridine-sulfur trioxide complex dissolves and an ochre-colored sulfation solution is formed.

(16) Into the base of an 80 ml weighing bottle with ground glass lid heated to 70? C. are pipetted 3.8 g of freshly prepared sulfation solution heated to 70? C. Circular dry cellulose microfilter membrane discs (diameter 70 mm) of the non-crosslinked membrane (comparative example, starting membrane of example 1 of type 10142V) and the membrane crosslinked according to example 1 are placed successively in the sulfation solution; they are wetted spontaneously in this case. The reaction vessel is then immediately tightly sealed.

(17) The reaction vessel is then stored for 4 hours at 70? C. in the convection drying cabinet. The sulfated celluslose microfilter membranes are subsequently rinsed with running reverse osmosis water for 10 min, shaken in 100 g of 1M NaCl solution for 5 min and then rinsed under running reverse osmosis water for 5 min. The membranes are then shaken in 100 g of a solution of 30% by weight glycerol and 70% by weight water for 5 min and subsequently dried at 80? C. for 10 min.

Example 3: Sulfation with Pyridine and Chlorosulfonic Acid

(18) Pyridine (Sigma-Aldrich, Cat. No. 270970-1L) and chlorosulfonic acid (Sigma-Aldrich, Cat. No. 571024-100G) are cooled to ?18? C. prior to use. With vigorous stirring, 60 ml of chlorosulfonic acid are added dropwise to 1000 ml of pyridine over 15 min via a pressure-equalizing dropping funnel and drying tube, wherein a temperature of 5? C. must not be exceeded; a white solid precipitates here.

(19) The mixture is then tightly sealed and heated to 65? C. under stirring in a water bath until the solid has completely dissolved. The sulfation solution thus obtained is then cooled to 40? C. over one hour.

(20) Non-crosslinked membrane strips (comparative example from example 1) and membrane strips from example 2 having a width of 3.3 cm and a length of 21 cm, 18 strips in total, are wound with coarse PP fabric and transferred to a 500 ml polypropylene screw-cap jar. The sulfation solution which is temperature-controlled to 40? C. is transferred to the screw-cap jar. The screw-cap jar is tightly sealed.

(21) The reaction vessel is shaken at 40? C. for 20 hours. The membrane filter strips are then rinsed and impregnated as follows: 10 min running reverse osmosis water, 10 min shaking with 1000 ml of 1M NaCl, 10 min running reverse osmosis water, 10 min in a solution of 30% by weight glycerol and 70% by weight water.

(22) These are then dried at 80? C. for at least 12 hours.

Example 4: Permeability Determination of Microfilter Membranes

(23) 47 mm discs are punched out from sulfated cellulose microfilter membranes according to example 3. These are wetted with reverse osmosis water and rinsed for 5 min under running reverse osmosis water. A punched blank is incorporated in a Sartorius Type 16249 pressure filter holder. The measurement is carried out at 20 to 25? C. and 0.1 bar positive pressure. The time taken for 100 g of medium to flow through the membrane filter is measured. The unit of flow thus determined is stated as ml/(cm.sup.2.Math.min.Math.bar). Both reverse osmosis water and 10 mM potassium phosphate buffer with pH 7 are used as media.

(24) The discs may be used for further investigations after determination of the flow.

Example 5: Quantitative Determination of the Degree of Sulfation by Titration

(25) 30 mm discs are punched out from the sulfated cellulose microfilter membranes using a round punch. A 30 mm punched blank is incorporated in a dead-volume optimized filter holder. The active filter surface is 5.7 cm.sup.2. The filter holder is filled with reverse osmosis water with exclusion of air bubbles. The filter holder is connected to a multichannel cartridge pump from Watson Marlow (Type 205 U), of which the lines are free of air bubbles. The cartridge pump output is circa 5 ml per minute. The following media are then pumped in series through the membrane filters, each for a 4 min period: 1M NaCl, 1M HCl, 1 mM HCl, reverse osmosis water.

(26) A 40 ml glass beaker with magnetic stirrer bars is then placed under the output side of the filter holder. 1M NaCl is then supplied for a further 4 min. The eluate collected is then titrated by means of potentiometirc indication. The titrant is 5.0 mM aqueous sodium hydroxide solution. To this end, the consumption of a 5 mM aqueous sodium hydroxide solution is determined up to the equivalence point at pH 7.0. The quantitative amount of aqueous sodium hydroxide solution consumed is directly proportional to the quantitative amount of sulfate groups on the membrane. From the active surface of the membrane available and the consumption of aqueous sodium hydroxide solution and also the mass of cellulose per cm.sup.2, the degree of sulfation, which is the proportion by mass of sulfate groups on the sulfated cellulose, can be calculated.

(27) Formula:

(28) Sulfate Group Density:
n.sub.sulfate (?mol/cm.sup.2)=c(NaOH)?t(NaOH)?V(NaOH)?1000/A
where
c(NaOH) quantitative concentration of aqueous sodium hydroxide titrant in mol/l
V(NaOH) volume in ml of titrant consumed at the equivalence point (pH 7)
t(NaOH) correction factor of titrant
1000 conversion factor from mol/I to ?mol/ml
A active filter surface in cm.sup.2
and where
M(SO.sub.3) molar mass of SO.sub.3 in ?g/?mol.
Formula:
Degree of sulfation in % by weight=(n.sub.sulfate?M(SO.sub.3)/(m.sub.cellulose(n.sub.sulfate?M(SO.sub.3)))?100%
where
m.sub.cellulose mass of cellulose in ?g per cm.sup.2 of punched blank.

(29) The results of the degree of sulfation of samples 1 to 5 are summarized in Table 2.

Example 6: Determination of the Binding Capacity for Lysozyme

(30) Membrane samples each having an active membrane surface of 17.6 cm.sup.2 are shaken three times for 5 minutes in 35 ml of 10 mM potassium phosphate buffer (KPi), pH 7.0, at about 80 revolutions per minute (rpm) and then placed in 35 ml of a solution of 2 mg/ml lysozyme in 10 mM Kpi, pH 7.6, at 20 to 25? C. for 12 to 18 hours. The membrane samples are then rinsed for 2?15 minutes in 35 ml of 10 mM KPi buffer (pH 7.0) each time. The membrane samples are then shaken in 20 ml of 10 mM KPi buffer (pH 7.0) and in 1M aqueous NaCl solution. The amount of protein eluted is determined by measuring the optical density (OD) at 280 nm.

(31) The results of the binding capacity for lysozyme for samples 1 to 5 and Sartobind? S are summarized in Table 2.

Example 7: Binding of Host Cell Proteins (HCP)

(32) To determine the binding of host cell proteins, a 10-fold concentrated HCP solution in PBS buffer (phosphate-buffered saline solution, pH 7.4) is used without antibody, prepared, in the case of a contract producer, in a mock run (culturing of a cell line without antibody production) of a Chinese hamster ovarian cell line. The HCP solution is diluted 1:10 in 20 mM TRIS/HCl buffer (pH 7.4) and the conductivity is adjusted to 10 mS/cm by the addition of NaCl. 1000 ml of the diluted HCP solution are used to load the membranes. The HCP concentration is determined by means of ELISA (enzyme-linked immunosorbent assay ELISA Cygnus CM015) according to the procedure specification. The concentration of host cell proteins (HCP) is 7 ?g/ml.

(33) 3 membrane layers (sample 4) are fixed in a membrane holder. The membrane stack in the membrane holder has a membrane surface of 15 cm.sup.2, an in-flow surface of 5 cm.sup.2 and a bed height (thickness of the membrane stack) of 750 ?m. The membranes in the membrane holder are flushed with 20 mM TRIS/HCl buffer (pH 7.4) to displace the air and then connected to an ?kta? Explorer 100 chromatography system from GE Healthcare.

(34) The membranes or the membrane stack are then investigated with respect to the HCP binding using a test program comprising four steps. The four steps of the test program are specified below: 1. Equilibrating the membrane with 10 ml of 20 mM TRIS/HCl buffer (pH 7.4) having a conductivity of 10 mS/cm, 2. Loading of the membrane with 100 ml of HCP solution, 3. Washing with 10 ml of 20 mM TRIS/HCl (pH 7.4, conductivity 10 mS/cm) and 4. Eluting with 10 ml of 1M NaCl in 20 mM TRIS/HCl buffer (pH 7.4).

(35) All steps are carried out at a flow rate of 10 ml/min. In all steps, the absorption is measured in the detector at 280 nm following on from the membrane unit. The breakthrough curves are shown in FIG. 3.

Example 8: Determination of the Binding Capacity for Small, Inactivated Influenza Viruses

(36) Influenza A Puerto Rico/8/34, H1N1 is produced in adhering MDCK cells (GMEM medium) (cf. Y. Genzel et al., Vaccine 22 (2004), 2202-2208). After culturing, the culture broth is filtered via two filtration steps in succession (pore size 5 ?m and 0.65 ?m filter depth, GE Water & Process Technologies) and is subsequently chemically inactivated with 3 mM ?-propiolactone at 37? C. for 24 hours. After inactivation, the solution is again clarified (0.45 ?m membrane filter, GE Water & Process Technologies) and subsequently concentrated 20-fold via cross flow filtration (750 kDa MWCO, GE Healthcare) (cf. B. Kalbfuss et al., Biotechnol. Bioeng. 97 (1), 2007, 73-85). The concentrated samples are stored at ?80? C. until required for use.

(37) The frozen virus aliquots (3?2 ml) are thawed in a water bath, mixed and centrifuged at 9000 g for 10 min. The supernatant is subsequently diluted 1:3 with the binding buffer (10 mM Tris, pH 7.4 and 50 mM NaCl).

(38) The solution thus prepared is used for the determination of the dynamic binding capacity. All experiments were carried out on an ?kta? Explorer 100 chromatography system from GE Healthcare. The units tested for the experiments consist of 15 membrane layers with an in-flow surface of 0.36 cm.sup.2 and a bed volume of 0.14 ml. The equilibration is carried out with 10 mM Tris, 50 mM NaCl pH 7.4. After equilibration, 26 ml of the virus solution is loaded onto the unit and 2 ml fractions are collected continuously. The flow rate during loading is 1 ml/min. The hemagglutinin (HA) activity is quantified in the various fractions. After loading, the membrane units are rinsed with equilibration buffer until the baseline is reached and are subsequently eluted with 10 mM Tris, 2.0 M NaCl pH 7.4. The eulate and the rinse fractions are also collected and their HA activity also analyzed.

(39) The virus concentration is determined in the continuous flow fractions and are quantified via the hemagglutination assay (cf. B. Kalbfuss et al., Biologicals 36 (2008), 145-161). The binding capacity for influenza viruses is measured for all samples and the capacity is calculated at 10%, 25% and 50% breakthrough, i.e. the amount of viruses bound per volume of separation material are attained up to 10% or 25% or 50% of the virus concentration of the starting solution. The dynamic binding capacity was calculated as follows:

(40) C.sub.0: virus concentration in the starting solution (virus particles/ml=part./ml)

(41) V.sub.x: loading volumes which were reached up to x % breakthrough (ml)

(42) A.sub.x: number of virus particles in the continuous flow which were reached up to x % breakthrough

(43) V.sub.B: bed volumes of the unit tested
Dynamic binding capacity up to x % breakthrough (part/ml)=(V.sub.x?C.sub.0?A.sub.x)/V.sub.B

(44) The results of the binding capacities of samples 1 to 5 and for Sartobind? S are summarized in Table 2.

(45) TABLE-US-00002 TABLE 2 Results summary Sample 1 2 3 4 5 Sartobind? S Pore size of cellulose starting membrane (?m) 3 3 3 3 1.2 Crosslinking according to none Example 1 none Example 1 Example 1 Not applicable Sulfation method according to Example 3 Example 3 Example 2 Example 2 Example 2 Not applicable Degree of sulfation in % by weight 0.64 19.37 1.26 10.34 16.65 Binding capacity with influenza part./ml Up to 10% breakthrough 0.0 5.0 .Math. 10.sup.11 0.0 3.4 .Math. 10.sup.12 6.4 .Math. 10.sup.12 4.9 .Math. 10.sup.11 Up to 25% breakthrough 4.3 .Math. 10.sup.11 3.1 .Math. 10.sup.12 4.0 .Math. 10.sup.11 4.3 .Math. 10.sup.12 Not reached* 4.9 .Math. 10.sup.11 Up to 50% breakthrough 4.3 .Math. 10.sup.11 4.4 .Math. 10.sup.12 4.0 .Math. 10.sup.11 5.7 .Math. 10.sup.12 Not reached* 2.3 .Math. 10.sup.12 Binding capacity with lysozyme (mg/mL) 32.6 7.0 30.0 4.8 8.1 37.0 *In sample 5, owing to the very high binding capacity of this membrane for influenza viruses, no 25% or 50% breakthrough were able to be reached with the loading volume of 26 ml of virus solution mentioned in Example 8.