CHROMATOGRAPHIC MATERIAL AND METHOD OF PRODUCING SAME

Abstract

The present invention relates to a chromatographic material comprising a polymer network material-based self-supporting bi-continuous separation matrix for adsorptive material separation in liquid media, and a method of producing the chromatographic material comprising a polymer network material-based self-supporting bi-continuous separation matrix.

Claims

1. A chromatographic material comprising a polymer network material-based self-supporting bi-continuous separation matrix comprising a first continuous phase and a second continuous phase, the chromatographic material optionally comprising a third inert phase, wherein the first continuous phase is a portion of the matrix which is formed by a self-gelled polysaccharide and wherein the second continuous phase is a portion of the matrix which defines non-cylindrical large diameter continuously connected convective pores in-between the first continuous phase, wherein the non-cylindrical large diameter continuously connected convective pores have a median pore diameter of from 0.1 m to 6.0 m, and wherein the third inert phase does not constitute the structure of the second continuous phase.

2. The chromatographic material according to claim 1, wherein the self-gelled polysaccharide is one or more selected from the group consisting of agarose, agar, agaropectin, kappa-carrageenan, iota-carrageenan, lambda-carrageenan, gellan gum, amylose, curdlan, alginate and rhamsan gum.

3. The chromatographic material according to claim 1, wherein a median bridge diameter of the first continuous phase is from 0.4 to 4 times larger than the median pore diameter of the non-cylindrical large diameter continuously connected convective pores.

4. The chromatographic material according to claim 1, wherein when a convective porosity is a ratio of a volume of the second continuous phase to a total volume of the first and second continuous phase, the convective porosity is from 0.05 to 0.7.

5. The chromatographic material according to claim 1, which has a permeability of at least 0.31(100) mD, wherein is a convective porosity.

6. The chromatographic material according to claim 1, wherein a coefficient of variance of the median pore diameter of the non-cylindrical large diameter continuously connected convective pores is at most 0.8.

7. The chromatographic material according to claim 1, wherein a coefficient of variance of a median bridge diameter of the first continuous phase is at most 0.7.

8. The chromatographic material according to claim 1, further comprising the third inert phase.

9. The chromatographic material according to claim 8, wherein a compressibility of the third inert phase is at least 85%, as determined according to the method indicated in the description.

10. The chromatographic material according to claim 8, wherein a ratio of a volume of the third inert phase to a total volume of the chromatographic material including the third inert phase is from 0.03 to 0.60.

11. A method of producing the chromatographic material according to claim 1, the method comprising: (a) preparing a solution (A) comprising a self-gelling polysaccharide and a first solvent; (b) preparing a solution (B) comprising at least one surfactant and a second solvent which is an immiscible solvent to solution (A); (c) combining the solution (A) and the solution (B) to produce a combined solution (C); (d) emulsifying the combined solution (C) at a condition to allow the self-gelling polysaccharide to remain in solution in the first solvent as solution (A) to obtain an emulsion which is in form of a solution (B)-in-solution (A)-emulsion; and (e) allowing the solution (A) containing the self-gelling polysaccharide to solidify by gelation at a certain condition to form the chromatographic material which comprises the separation matrix comprising the self-gelled polysaccharide as the first continuous phase.

12. The method according to claim 11 in which the self-gelling polysaccharide is agarose, the method comprising: (a) preparing the solution (A) comprising agarose and water by heating an agarose-in-water-suspension to at least 90 C. for at least 15 minutes to solve the agarose in the water; (b) preparing the solution (B) comprising the at least one surfactant and the second solvent which is a water-immiscible organic solvent and heating solution (B) to a temperature above a gelation temperature of the used agarose; (c) combining the solution (A) and the solution (B) at a temperature of above the gelation temperature of the used agarose to produce a combined solution (C); (d) emulsifying the combined solution (C) at a temperature above the gelation temperature of the used agarose to allow the agarose to remain in solution in the first solvent as solution (A) to obtain an emulsion which is in form of a solution (B)-in-solution (A)-emulsion; and (e) allowing the solution (A) containing the self-gelling agarose to solidify by gelation at a temperature below the gelation temperature of the used agarose to form the chromatographic material which comprises the separation matrix comprising the self-gelled agarose as the first continuous phase.

13. The method according to claim 11, wherein the at least one surfactant has an HLB value of from 8 to 16.

14. The method according to claim 11, wherein the second solvent is one or more selected from C.sub.4-12 alcohols, C.sub.4-12 isoalcohols, alkanes, silicone oils having a viscosity of from 4 to 200 cP, natural oils having a viscosity of from 4 to 200 cP, and mixtures thereof.

15. The method according to claim 11, wherein the at least one surfactant is selected from polysorbates and/or C.sub.2-150 fatty acid esters of sorbitol.

16. The method according to claim 11, wherein the concentration of the polysaccharide in the solution (A) is from 0.5 to 8 wt %.

17. The method according to claim 11, wherein a volume ratio of the second solvent in the combined solution (C) in (c) is from 10 to 70 vol %.

18. The method according to claim 11, wherein (e) is performed by pouring the emulsion on a casting form and cooling the emulsion below the gelling point.

Description

[0084] The Figures show:

[0085] FIG. 1 shows a comparison of the convective and diffusional flow paths in polysaccharide particles and polysaccharide membranes.

[0086] FIG. 2 shows the relationship between the permeability and the median diameter of the convective pores (at =0.33 convective porosity) for bi-continuous and closed-cell foam structures (Example 1).

[0087] FIG. 3 shows the relationship of the median pore diameter and the used surfactant concentration (at =0.33 convective porosity) in the preparation of chromatographic materials in accordance with the reaction conditions (process parameters) described in U.S. Pat. No. 5,723,601 A (Example 1).

[0088] FIG. 4 shows the proportionality of the median bridge diameter and the median pore diameter (at =0.33 convective porosity) of the non-cylindrical large diameter continuously connected convective pores for selected examples of Table 2 (Example 2).

[0089] FIG. 5 shows a device for determining DBC 10%.

[0090] FIG. 6 shows a relationship between the median bridge diameter of the separation matrix and the dynamic binding capacity (at =0.33 convective porosity) (Example 3).

[0091] FIG. 7 shows a relationship between the coefficient of variance of the median bridge diameter of the first continuous phase and the dynamic binding capacity (at =0.33 convective porosity) (Example 4).

[0092] FIG. 8 shows a relationship between specific interface between the first and the second continuous phases and the dynamic binding capacity (at =0.33 convective porosity) (Example 5).

[0093] FIG. 9 shows DBC 10% in dependence of the residence time (Example 8).

[0094] FIG. 10 shows a comparison of DBC 10% of the inventive chromatographic material and state-of-the-art materials as a function of the residence time (Example 8).

[0095] FIG. 11 shows the trans-device pressure during cycling with different cleaning strategies (load: mAb in clarified cell culture at an mAb concentration of 2.8 mg/mL) (Example 9).

[0096] FIG. 12 shows overlayed chromatograms of a single bind & elute cycle for three different chromatography modules (Example 10).

[0097] FIG. 13 shows an overlay of 200 cycles with mAb A, UV signal at 280 nm (Example 11).

[0098] FIG. 14 shows an overlay of 200 cycles with mAb B, UV signal at 280 nm (Example 11).

[0099] FIG. 15 shows an overlay of 200 cycles with mAb C, UV signal at 280 nm (Example 11).

[0100] FIG. 16 shows an overlay of 200 cycles with mAb A, delta pressure [MPa] (Example 11)

[0101] FIG. 17 shows an overlay of 200 cycles with mAb B, delta pressure [MPa] (Example 11).

[0102] FIG. 18 shows an overlay of 200 cycles with mAb C, delta pressure [MPa] (Example 11).

[0103] FIG. 19 shows the yield vs. cycle number for mAbs A, B and C (Example 11).

[0104] FIG. 20 shows the HCP removal vs. cycle number for mAbs A, B and C (Example 11).

[0105] FIG. 21 shows the DNA removal vs. cycle number for mAbs A, B and C (Example 11).

[0106] FIG. 22 shows the ProA ligand leaching vs. cycle number for mAbs A, B and C (Example 11).

[0107] FIG. 23 shows the elution peak width vs. cycle number for mAbs A, B and C (Example 11).

[0108] FIG. 24 shows the pool concentration vs. cycle number for mAbs A, B and C (Example 11).

[0109] FIG. 25 shows the aggregate concentration vs. cycle number for mAbs A, B and C (Example 11).

[0110] FIG. 26 shows the breakthrough behavior of an inventive 3-layer and 12-layer chromatographic device (Example 12).

[0111] FIG. 27 shows the elution peak shape of an inventive 3-layer and 12-layer chromatographic device (Example 12).

[0112] FIG. 28 shows DBC 10% results as a function of the residence time for CIEX materials (Example 13).

[0113] FIG. 29 shows DBC 10% results as a function of the residence time for MM materials (Example 13).

[0114] The present invention will be further illustrated in the following examples without being limited thereto.

EXAMPLES

General Preparation Example:

[0115] A self-gelling polysaccharide solution (e.g. 3% agarose in water) was prepared and stored at elevated temperature (e.g. 95 C.) until dissolved completely. Decanol (e.g. 27.39 g), Tween80 (e.g. 2.48 g) and Span80 (e.g. 0.98 g) were mixed and stirred at elevated temperature (e.g. 60 C.) to form an organic phase. The above self-gelling polysaccharide solution (e.g. agarose solution in an amount of 63.72 g) was poured into a beaker tempered in a water tub at 50 C., and the organic phase was added slowly while stirring. The emulsion was rigorously stirred with an Ultraturrax (preferably) at 20.000 rpm and 70-75 C. for 10 min. Alternatively, the emulsion was rigorously stirred with an EUROSTAR 40 digital from IKA with a dispersing disk of type R1303 Dissolver Stirrer (preferably at 2.000 rpm). The temperature was observed during the emulsion process and the water tub cooled or warmed if needed to keep the emulsion temperature constant. In the meantime, the casting platform was prepared. For this, optional a sheet of fleece (third inert phase) acting as a reinforcing material was taped at both ends of the casting platform which was tempered at 40 C. by a stream of warm water from below. The same process also works without third inert phase (reinforcing material/fleece). The emulsion was poured onto the one end of the casting platform (in presence of fleece on the top of the fleece) and was spread with a scraper to obtain a layer of emulsion with a width of e.g. 300 m. The casting platform was immediately cooled down to 15 C. using cold tap water from below. By this, gelation of the agarose was induced. After waiting for 1 to 2 minutes, the emulsion became a firm film (optional attached to the fleece). The membrane with or without reinforcing material/fleece was removed carefully from the casting platform and placed into a tub with running water to wash out the organic phase. After washing the membrane with running water for 20 min, it was gently shaken in isopropanol (210 min) and water/isopropanol mixture (50:50, 210 min), each step including a change of washing medium. Finally, the membrane was washed in running water for another min.

[0116] For confocal microscopy analysis samples without reinforcing material/fleece proved to be the easiest way to obtain good microscope images. For this, a small hole with a diameter of 13 mm was cut into the fleece before casting the emulsion. The casting procedure was the same as stated above. Before the washing procedure, the membrane was cut out at that same hole and a membrane without fleece was obtained.

[0117] Ligand immobilization for an affinity ligand (Protein A) was carried out in a two-step process. In a first step, the activation is carried out with a bisoxirane molecule in case the matrix has functional groups that reacts with oxirane so that at least one oxirane group of the bisoxirane reacts with the matrix. The so created unreacted oxiranes on the matrix surface are then available for further surface chemistry.

[0118] The second step is carried out by adding the ligand to the activated matrix which has at least one reactive side with respect to the oxirane. This or similar possible methods can be found in A. Krishna Mallia, Paul K. Smith, Greg T. Hermanson, Immobilized Affinity Ligand Techniques, Elsevier Science, 1992.

Example 1

Preparation and Characterization of Chromatographic Materials According to the Present Invention and of the Prior Art

[0119] According to the above general preparation example, chromatographic material Sample Nos. 1 to 33 were prepared under the conditions indicated in Table 1. The affinity ligand was introduced by dissolving the used protein A Ligand in 1M KPI buffer with 10 mg/mL in concentration at pH 7. After preparing this coupling solution the separation matrix was inserted in this coupling solution at room temperature for at least 16 h. After this, the separation matrix was washed with 0.1M 1PBS buffer at pH 7 and stored in this until usage.

[0120] In this Example and in the following Examples, the chromatographic materials were characterized with respect to [0121] Ligand density: Protein A Material: BCA assay, Literature: C. M. Stoscheck: Quantitation of protein. In: Methods in enzymology. Band 182, 1990, S. 50-68, PMID 2314256 IEX Material: Titration with e.g. NaOH or HCl as known to persons skilled in the art. [0122] Determination of SBC (static binding capacity): The SBC value is measured by a 12 h incubation of the target molecule in binding buffer conditions and an elution step after washing in elution condition buffer for the target molecule for at least 1 h. The concentration in the elution buffer is measured by UV-Vis at 280 nm within a calibration curve measured prior to the SBC measurement. [0123] Determination of DBC 10% (Dynamic binding capacity at 10% breakthrough): The DBC 10% is measured with an kta Avant150 chromatography System from GE-Healthcare. At first the described material is placed into an in-house measurement device (LP15, FIG. 5) and connected to the kta Avant150 via luer-lock. Then at least 20 separation-matrix volumes (MV) of binding buffer were purged through the separation matrix. Then binding buffer with solved target protein (1 mg/mL) is purged through the separation matrix until 10% of the concentration of the initial target protein solution is reached (0.1 mg/mL). This until then purged volume in mL is due to the concentration of 1 mg/mL equivalent with the mass in mg and gives the DBC 10% by dividing by the separation matrix volume to derive the DBC 10% in mg/mL. [0124] Determination of DBC 100% (Dynamic binding capacity at 100% breakthrough): The DBC 100% is measured with an kta Avant150 chromatography System from GE-Healthcare. At first the described material is placed into an in-house measurement device (LP15, FIG. 5) and connected to the kta Avant150 via luer-lock. Then at least 20 separation-matrix volumes (MV) of binding buffer were purged through the separation matrix. Then binding buffer with solved target protein (1 mg/mL) is purged through the separation matrix until 100% of the concentration of the initial target protein solution is reached (1 mg/mL). The area under the so derived breakthrough curve can then be integrated and the DBC 100% can then be determined by the measured area, the needed volume to reach 100% of the initial concentration and the separation matrix volume in mg/mL.

[0125] Results of the characterization of the separation matrices are indicated in Table 2.

TABLE-US-00001 TABLE 1 Stirrer Agarose Aqueous Organic Surfactants Tween Span Sample Structure Emulsified Speed Organic concentration phase phase total 80 80 HLB No. Type by [1/min] phase [wt. %] [v %] [v %] [v %] [v %] [v %] [] 1 Bicontinuous (A) 2000 Cyclohexane 3 65.34 32.70 1.96 1.96 0.00 15.00 2 Bicontinuous (A) 2000 Cyclohexane 3 64.02 32.70 3.28 3.28 0.00 15.00 3 Bicontinuous (A) 2000 Cyclohexane 3 65.63 32.70 1.67 1.67 0.00 15.00 4 Bicontinuous (A) 2000 Cyclohexane 3 66.63 32.70 0.67 0.67 0.00 15.00 5 Bicontinuous (B) 10000 Cyclohexane 3 65.34 32.70 1.96 1.96 0.00 15.00 6 Bicontinuous (A) 2000 Cyclohexane 5 65.34 32.70 1.96 1.96 0.00 15.00 7 Bicontinuous (A) 2000 Cyclohexane 5 64.02 32.70 3.28 3.28 0.00 15.00 8 Bicontinuous (A) 2000 Cyclohexane 5 65.63 32.70 1.67 1.67 0.00 15.00 9 Bicontinuous (A) 2000 Cyclohexane 6 65.34 32.70 1.96 1.96 0.00 15.00 10 Bicontinuous (A) 2000 Cyclohexane 8 65.34 32.70 1.96 1.96 0.00 15.00 11 Foam (A) 2000 Cyclohexane 3 66.97 32.70 0.33 0.33 0.00 15.00 12 Foam (A) 2000 Cyclohexane 6 66.97 32.70 0.33 0.33 0.00 15.00 13 Foam (A) 2000 Cyclohexane 5 66.63 32.70 0.67 0.67 0.00 15.00 14 Foam (A) 2000 Cyclohexane 5 66.97 32.70 0.33 0.33 0.00 15.00 15 Foam (A) 2000 Cyclohexane 8 66.97 32.70 0.33 0.33 0.00 15.00 16 Bicontinuous (A) 2000 1-Decanol 3 63.72 33.00 3.28 2.30 0.98 11.97 17 Bicontinuous (A) 2000 1-Decanol 3 61.98 33.00 5.02 3.52 1.50 11.97 18 Bicontinuous (A) 2000 1-Decanol 4 63.72 33.00 3.28 2.30 0.98 11.97 19 Bicontinuous (A) 2000 1-Decanol 4 61.98 33.00 5.02 3.52 1.50 11.97 20 Bicontinuous (A) 2000 1-Decanol 4 65.33 33.00 1.67 1.17 0.50 11.97 21 Bicontinuous (B) 20000 1-Octanol 3 63.72 33.00 3.28 2.30 0.98 11.97 22 Bicontinuous (B) 20000 Decane 3 68.20 30.00 1.81 1.30 0.54 12.03 23 Bicontinuous (B) 20000 Octane 3 67.90 30.00 2.11 1.48 0.63 11.97 24 Bicontinuous (B) 20000 1-Hexanol 3 60.30 33.00 6.69 4.70 2.00 11.98 25 Bicontinuous (B) 20000 2-Heptanol 3 62.00 33.00 5.02 3.50 1.50 11.96 26 Bicontinuous (B) 20000 Silicone oil 3 63.70 33.00 3.28 2.30 0.98 11.97 (PDMS) 20cP Wacker 27 Bicontinuous (B) 20000 1-Decanol 4 65.33 33.00 1.67 1.17 0.50 11.97 28 Foam (A) 2000 1-Decanol 4 66.33 33.00 0.67 0.47 0.20 11.97 29 Foam (A) 2000 1-Decanol 4 66.67 33.00 0.33 0.23 0.10 11.97 30 Foam (B) 20000 1-Decanol 4 66.33 33.00 0.67 0.47 0.20 11.97 31 Foam (B) 20000 1-Decanol 4 66.67 33.00 0.33 0.23 0.10 11.97 32 Foam (A) 2000 1-Decanol 3 66.33 33.00 0.67 0.47 0.20 11.97 33 Foam (A) 2000 1-Decanol 3 66.67 33.00 0.33 0.23 0.10 11.97 (A) R1303 Dissolver Stirrer; (B) T25 digital ULTRA-TURRAX. Model T25D. Firma IKA

TABLE-US-00002 TABLE 2 Median Coefficient Specific Interface Pore Coefficient of of Variance between First and DBC 10% Diameter Variance of Median Median of Median Second Protein A- (Convective Pore Diameter Bridge Bridge continuous Membrane. IgG Sample Permeability Pores) (Convective Pores) Diameter Diameter phases Tortuosity binding 5MV/min No. [mD] [m] [] [m] [] [m.sup.2/mL] [] [mg/ml] 1 1123.8 15.57 0.62 22.31 0.55 0.03 2.18 2 1293.4 16.23 0.64 25.11 0.54 0.03 3.05 3 826.0 12.30 0.56 18.31 0.50 0.04 2.01 4 393.0 7.80 0.60 12.68 0.53 0.06 1.65 5 563.8 10.26 0.51 17.93 0.49 0.01 1.67 6 413.9 9.64 0.55 15.56 0.52 0.02 1.42 7 912.8 15.01 0.63 21.02 0.58 0.03 1.49 8 259.0 9.95 0.67 14.71 0.57 0.05 1.53 9 112.2 6.85 0.49 10.07 0.44 0.02 1.79 10 69.7 6.50 0.57 9.68 0.47 0.07 1.73 11 10.0 5.13 0.58 8.12 0.50 0.10 1.57 12 1.0 2.05 0.50 3.28 0.44 0.26 2.79 13 6.2 2.88 0.46 4.78 0.45 0.20 2.86 14 9.3 4.67 0.58 7.87 0.53 0.06 1.35 15 1.0 1.88 0.52 2.80 0.39 0.29 3.28 16 95.4 2.88 0.46 3.96 0.37 0.23 3.43 17 150.0 3.54 0.43 4.53 0.43 0.20 1.60 18 132.3 3.39 0.55 4.59 0.40 0.20 2.76 32.16 19 263.7 4.99 0.42 6.26 0.40 0.15 1.60 29.68 20 12.0 2.90 0.45 4.25 0.41 0.22 2.62 39.49 21 62.7 2.52 0.41 3.64 0.43 0.26 2.20 33.6 22 59.6 2.95 0.73 5.99 0.56 0.14 1.49 36.3 23 150.7 3.76 0.45 6.18 0.45 0.16 4.37 26.85 24 21.3 1.05 0.35 1.39 0.35 0.37 1.89 29.61 25 114.0 2.76 0.43 3.76 0.45 0.25 1.95 31.06 26 175.2 2.93 0.44 4.40 0.44 0.22 1.71 29.3 27 40.0 3.30 0.52 4.13 0.39 0.21 3.29 44.86 28 6.4 3.86 0.64 5.52 0.41 0.16 4.05 29 3.8 4.98 0.64 5.70 0.40 0.14 27.94 30 9.9 3.11 0.53 4.61 0.42 0.21 2.70 9.65 31 8.2 6.46 0.72 7.27 0.38 0.10 4.83 5.96 32 3.5 3.32 0.64 4.40 0.38 0.20 3.03 11.38 33 2.0 4.80 0.70 6.44 0.42 0.14 3.60 9.93

[0126] Sample Nos. 1 to 15 are chromatographic materials prepared in accordance with the reaction conditions (process parameters) described in U.S. Pat. No. 5,723,601 A. Sample Nos. 1 to 10 are bi-continuous separation matrices, but the non-cylindrical large diameter continuously connected convective pores of each of the separation matrices have a median pore diameter of more than 6 m. As can be seen from Sample Nos. 1 to 5, less surfactant (at constant polysaccharide concentration) leads to smaller pore diameters of the convective pores. However, in Sample Nos. 11 to 15, in which the concentration of the surfactant falls under a certain threshold, a foam structure is obtained instead of a bi-continuous structure. The foams consist of partially more or less connected globular structures as convective phase (macropores) between the diffusive phase, i.e. the macropores are not interconnected. The inherently resulting poorer connectivity results in structures with a very low permeability at comparable pore size (i.e. the median diameter of the convective pores), since there is no through flow via these pores. FIG. 2 shows the relationship between the permeability and the median diameter of the convective pores for bi-continuous and foam structures. Therefore, foam structures can only be traversed in a very inhomogeneous and inadequate manner, which results in disadvantages in the homogeneous distribution of fluid in the chromatographic medium and, in addition to the lower permeability, disadvantages in the dynamic binding due to inhomogeneous flow and thus inhomogeneous binding site accessibility.

[0127] FIG. 3 shows the relationship of the median pore diameter and the used surfactant concentration in the preparation of chromatographic materials in accordance with the reaction conditions (process parameters) described in U.S. Pat. No. 5,723,601 A. With the reaction conditions described in U.S. Pat. No. 5,723,601 A, it is not possible to prepare a bi-continuous separation matrix with non-cylindrical large diameter continuously connected convective pores having a median pore diameter of from 0.1 m to 6 m.

[0128] Sample Nos. 16 to 27 are chromatographic materials according to the present invention, comprising bi-continuous separation matrices with non-cylindrical large diameter continuously connected convective pores having a median pore diameter of from 0.1 m to 6 m. The separation matrices show an excellent permeability, particularly when compared to the foams of Sample Nos. 11 to 15 which have similar median pore diameters of the convective pores. The chromatographic material of the present invention has excellent accessibility of the binding sites and dynamic binding capacity at 10% breakthrough.

[0129] Sample Nos. 28 to 33 are chromatographic materials prepared as Comparative Examples, in which the same solvent as in (inventive) Sample Nos. 16 to 20 and 27 was used, but the concentration of surfactant used in the preparation has been lowered below a certain threshold such that foam structures are formed. Sample Nos. 28 to 33 have convective pores having a median pore diameter similar to Sample Nos. 16 to 27, but the convective pores of foam structure Sample Nos. 28 to 33 are not continuously connected (interconnected) like the convective pores of Sample Nos. 16 to 27. Accordingly, when comparing the foam structures of Sample Nos. 28 to 33 with the bi-continuous structures of Sample Nos. 16 to 27, the foam structures show a reduced permeability at comparable convective pore diameters. Moreover, the foam structures of Samples Nos. 30 to 33 show an inferior dynamic binding capacity at 10% breakthrough when compared with the bi-continuous structures of Sample Nos. 18 to 27.

Example 2

Relationship Between the Mean Bridge Diameter and the Mean Pore Diameter of the Large Diameter Convective Pores

[0130] Separation matrices comprising a self-gelled porous polysaccharide as the first continuous phase were prepared and analyzed with respect to their mean bridge diameter and mean pore diameter of the large diameter convective pores. The results are depicted in FIG. 4, which clearly shows the proportionality of the mean bridge diameter and the mean pore diameter of the large diameter convective pores. In particular, the bigger the large diameter convective pores become with a constant pore volume fraction (convective porosity), the bigger the bridges become because the bridges just represent the space between the large diameter convective pores.

Example 3

Relationship Between DBC 10% of Separation Matrices and Their Median Bridge Diameter

[0131] Separation matrices comprising a self-gelled porous polysaccharide as the first continuous phase were prepared and analyzed with respect to their median bridge diameter and DBC 10% (dynamic binding capacity 10% breakthrough (ligand: affinity protein A; buffer: phosphate-buffered saline (PBS) pH 7.3; analyte 1 mg/mL monoclonal antibody (mAb) in PBS; velocity: 5 MV/min=12 sec residence or contact time with or in the medium)). All data were derived by using flat sheet materials inside a device like shown in FIG. 5 (called LP15) and were measured with an kta Avant150.

[0132] The results are depicted in FIG. 6, which clearly shows that when keeping the process parameters constant (contact and process time). the larger the median bridge diameter of the separation matrix is. the smaller also the dynamic binding capacity is.

Example 4

Relationship Between DBC 10% of Separation Matrices and Their Coefficient of Variance of the Median Bridge Diameter of the First Continuous Phase

[0133] Separation matrices comprising a self-gelled porous polysaccharide as the first continuous phase were prepared and analyzed with respect to their variance of the median bridge diameter of the first continuous phase and DBC 10% (ligand: affinity protein A; buffer: phosphate-buffered saline (PBS) pH 7.3; analyte 1 mg/mL monoclonal antibody (mAb) in PBS; velocity: 5 MV/min=12 sec residence or contact time with or in the medium).

[0134] The results are depicted in FIG. 7, which clearly shows that, when keeping the process parameters constant (contact and process time), the larger the coefficient of variance of the median bridge diameter is, the smaller also the dynamic binding capacity is.

Example 5

Relationship Between DBC 10% of Separation Matrices and Their Specific Interface Between the First and the Second Continuous Phases

[0135] Separation matrices comprising a self-gelled porous polysaccharide as the first continuous phase were prepared and analyzed with respect to their specific interface between the first and the second continuous phases and DBC 10% (ligand: affinity protein A; buffer: phosphate-buffered saline (PBS) pH 7.3; analyte 1 mg/mL monoclonal antibody (mAb) in PBS; velocity: 5 MV/min=12 sec residence or contact time with or in the medium). The separation matrices in this Example are designed to have the maximum possible static binding capacity accessible via ligand immobilization.

[0136] The results are depicted in FIG. 8, which clearly shows that, when keeping the process parameters constant (contact and process time), the larger the specific interface between the first and the second continuous phases is, the faster the mass transport can take place and the larger also the dynamic binding capacity is.

Example 6

[0137] Inventive chromatographic membranes (Sample Nos. 34 and 35) comprising a self-gelled porous agarose as the first continuous phase were prepared according to the above-described general preparation example under the following conditions indicated in Table 3.

[0138] In particular, a self-gelling agarose solution was prepared and stored at elevated temperature (e.g. 95 C.) until dissolved completely. 1-Decanol, Tween80 and Span80 are mixed and stirred at elevated temperature (e.g. 80 C.) to form an organic phase. The above self-gelling agarose solution was poured into a vessel tempered at elevated temperature (e.g. 80 C.) and the organic phase was added slowly while stirring. The emulsion was rigorously dispersed with a Disperser YSTRAL Inline Z66 (C) at e.g. 5000 rpm. The third inert phase (e.g. fleece) acting as a reinforcing material was attached to casting platform which was kept a temperature below the gelling temperature (e.g. 30 C.). The emulsion was cast as a thin film onto the casting platform, soaking the fleece. Due to the temperature of the casting platform being below the gelling temperature of the agarose, the agarose began immediately to form the gelled inventive chromatographic membrane. The inventive chromatographic membrane with reinforcing material was removed from the casting platform, winded on a roll and rinsed with isopropanol and finally with water.

TABLE-US-00003 TABLE 3 Components and recipe used for sample preparation Agarose Aqueous Organic Surfactants Tween Sample Structure Emulsified Organic concentration phase phase total 80 Span 80 No. Type by phase [wt %] [v %] [v %] [v %] [v %] [v %] HLB 34 Bicontinuous YSTRAL 1-Decanol 2.5 61.70 35.20 3.20 2.70 0.50 13.50 Inline Z66 35 Bicontinuous YSTRAL 1-Decanol 5.0 64.00 33.00 3.38 2.43 0.98 12.00 Inline Z66

Example 7

Impact of the Compression of the Third Inert Phase (Reinforcing Material) on the Permeability of the Inventive Chromatographic Membrane Under Typical Operating Pressures

[0139] Different flat sheet reinforcing materials were used as third inert phase for the preparation of inventive chromatographic membranes according to Example 6, Sample Nos. 34 and 35. The used reinforcing materials were fleece materials A, B and C with different compressibility measured by compression as calculated by the above-mentioned equation. For a compression measurement, the thickness is measured before and after the compression by a weight of 1 kg that is placed on top of the push rod while monitoring the sample's thickness after 20 sec.

[0140] The permeability of the resulting inventive chromatographic membrane was determined as described in scientific publication Ley et. al, Journal of Membrane Science, Volume 564, 2018, Pages 543-551 (page 547: description; page 545: measurement). The permeability at 100 mbar trans membrane pressure (TMP) of a single layer inventive chromatographic membrane was compared to the permeability of a stack of 3 layers of inventive chromatographic membrane at a TMP of 100 mbar. The permeabilities are normalized to the different thicknesses by the following calculation:

[00008]$\left[\mathrm{mD}\right]=v*\frac{10^{-6}}{10^{-4}*10^5*60}*x*10^{-6}**10^{-3}*\frac{1}{9.869*10^{-13}}*10^3$

[0141] Here, is the flow rate of water passing through the membrane single layer or stack

[00009]$\left(\mathrm{in}\frac{{\mathrm{cm}}^3}{\min *\mathrm{bar}*{\mathrm{cm}}^2}\right),$

the viscosity of water (in mPa.Math.s) and x the single layer or stack thickness (in m).

[0142] The relative permeability of inventive chromatographic membranes is calculated by

[00010]$\mathrm{relative}\mathrm{permeability}\left[\%\right]=\frac{\mathrm{permeability}3-\mathrm{layer}\mathrm{membrane}\mathrm{stack}}{\mathrm{permeability}\mathrm{single}\mathrm{layer}\mathrm{membrane}}.Math.100$

TABLE-US-00004 TABLE 4 Relative membrane Relative membrane Relative membrane permeability of permeability of permeability of stack of 3 layers stack of 3 layers stack of 3 layers of inventive of inventive of inventive chromatographic chromatographic chromatographic Compression membrane at membrane at membrane at of third 100 mbar [%] 100 mbar [%] 1000 mbar [%] inert phase according to according to according to [%] Sample 34 Sample 35 Sample 35 Third inert phase A 90 94 85 65 Third inert phase B 87 91 86 79 Third inert phase C 82 81 48 37

[0143] The compression of the third inert phase has a strong effect on the permeability of the inventive chromatographic membrane under typical operating TMPs. The higher the compression of the third inert phase, the higher is the relative permeability of a 3-layer stack of inventive chromatographic membrane under typical operating pressures.

Example 8

Performance Characteristics of the Inventive Chromatographic Membrane Functionalized with Affinity Protein A Ligand

[0144] A reinforced inventive chromatographic membrane (=Sample A) was generated according to Example 6, Sample no. 34, having a third inert phase C according to Example 7, being modified according to Example 1. The generated inventive chromatographic membrane Sample A exhibits the following material and performance characteristics: thickness of 280 m, permeability of 66 mD, ligand density 19.8 g/L. Inventive chromatographic membrane Sample A was cut into circular coupons and assembled as a 4-layer circular membrane stack in a reusable filtration device with a polypropylene filter housing (filter table and filter cover) with luer lock connectors and mechanically supported by a stainless steel holder which forms a mechanically stable chromatographic unit out of the filter cover, membrane stack and the filter table. The frontal area of the membrane stack was 5 cm.sup.2 and the thickness 1.12 mm, which resulted in a total membrane volume (MV) of 0.56 mL.

[0145] The dynamic binding capacity of the inventive chromatographic membrane device for a purified monoclonal antibody at different residence times was determined using an kta Avant 150 chromatography system (Cytiva, 28976337). Equilibration and wash buffers used in the performance characterization were made from phosphate buffered saline (0.1 M pH 7.4, conductivity 16 mS/cm); As an elution buffer, acetic acid at 0.1 M concentration at pH 2.9 was used. The monoclonal antibody (MW 140 kDa) was generated by CHO cell fermentation. Clarification of the fermentation broth was done by depth filtration (Sartoclear DL20 0.8 m.sup.2 29XDL20-FCC; Sartoclear DL90 0.8 m.sup.2 29XDL90-FCC) and sterile filtration through a 0.2 m sterile filter capsule (Sartopore 2 XLG 30 5447307G3-SS). The antibody was purified by using a Protein A MabSelectSuRe column (MabSelectSuRe GE 17-5438; XK 50/20 GE). The load solution used for the performance characterization of the membrane was made from equilibration buffer containing 1 mg/mL of the purified antibody. The flow rates for the dynamic binding capacity determination were 1, 3, 5 and 10 MV/min, which corresponds to residence times of 0.1, 0.2, 0.3 and 1 min. The concentration of the antibody in the flow-through fraction was determined by an in-line UV detector at 280 nm with a known extinction coefficient of the antibody of 1.42. The dynamic binding capacity was calculated for a 10% breakthrough of load solution based on the UV signal (=DBC 10%; cf. FIG. 9).

[0146] A 0.4 mL unit Fibro HiTrap PrismA (Cytiva 17549856 lot 17103843) was used as a benchmark chromatographic material representing a state-of-the-art purely convective protein A affinity chromatography material, exhibiting no diffusive phase. The material is made from cellulose nano fibers, where the ligand is immobilized onto the surface of the fibers and therefor directly accessible by the convective flow through the fibrous material eliminating any relevant diffusional transport limitations.

[0147] A 1 mL MabSelectSure column (Cytiva, 11003493) was used as an additional benchmark material representing a state-of-the-art protein A affinity resin material. The resin is made from crosslinked agarose and has an average bead size of 85 m (GE Instructions 71-5020-91 AC), representing the diffusive phase.

[0148] The two state-of-the-art protein A affinity materials and the inventive protein A chromatographic material were characterized with regard to their chromatographic performance. Data were determined as described before. Equilibration and wash buffers used in the performance characterization were made from phosphate buffered saline (0.1 M pH 7.4, conductivity 16 mS/cm). As an elution buffer, acetic acid at 0.1 M concentration at pH 2.9 was used. The monoclonal antibody (mAb) (MW 140 kDa) was generated by CHO cell fermentation. Clarification of the fermentation broth was achieved by depth filtration (Sartoclear DL20 0.8 m.sup.2 29XDL20-FCC; Sartoclear DL90 0.8 m.sup.2 29XDL90-FCC) and followed by sterile filtration through a 0.2 m sterile filter capsule (Sartopore 2 XLG 30 5447307G3-SS). The antibody was purified by using a ProA MabSelectSuRe column (MabSelectSuRe GE 17-5438; XK 50/20 GE).

[0149] The load solution used for the performance characterization of the inventive chromatographic membrane was made from equilibration buffer containing 1 mg/mL of the purified mAb. The concentration of the mAb in the flow-through fraction was determined by an in-line UV detector at 280 nm with a known extinction coefficient of the antibody of 1.42. The dynamic binding capacity was calculated for a 10% breakthrough of load solution based on the UV signal (=DBC 10%).

[0150] As depicted in FIG. 10, the inventive chromatographic membrane shows a markedly higher DBC 10% vs. residence time performance than a state-of-the-art agarose resin material, enabling highly productive and fast processes for antibody capture. The dynamic binding performance is very similar to the state-of-the-art convective membrane material. For very short residence times (<0.2 min), the convective material shows slightly higher DBC 10% values.

Example 9

Low Fouling Propensity of the Inventive Chromatographic Membrane Compared to Purely Convective Material of Similar Binding Capacity

[0151] A reinforced inventive chromatographic membrane was generated according to Example 6, Sample No. 34, having a third inert phase C according to Example 7, being modified according to Example 1. The fouling propensity of a freshly prepared membrane, assembled into a 4-layer chromatographic device configuration as described in Example 8, was compared to the fouling propensity of an unused, purely convective benchmark material Fibro (Cytiva 17549856 lot 17103843), described in Example 8 in order to compare the fouling propensity of these different chromatographic materials. Multiple bind & elute cycles were performed using a clarified mAb feed stream produced from CHO cells as described in Example 8, observing the changing permeability of the materials indicated by observing the change of the trans-device pressure during the consecutive bind & elute cycles.

[0152] The antibody concentration in the feed stream was 2.8 mg/mL. Before executing the cycling experiment, the dynamic binding capacity of the used devices was determined with purified mAb as generated in Example 8. It is important to note that both materials have a similar binding capacity of around 40 mg/mL at a residence time of 0.2 min (cf. FIG. 10). The load of the cycling study was then set to 80% of the determined DBC 10% value. The different phases in the bind & elute cycle are the following: 1) equilibration with 8 MV equilibration buffer at 10 MV/min; 2) loading to 80% of DBC 10% with clarified feed stream at 5 MV/min (i.e. 7.5 mL load for Cytiva HiTrap 0.4 mL, and 6 mL load for the inventive chromatographic membrane); 3) washing with 8 MV wash buffer at 10 MV/min; 4) elution with 10 MV at 5 MV/min; and 5) washing with 5 MV wash buffer at 10 MV/min. Then the next cycle was started by equilibration (step 1), or a Cleaning-In-Place (CIP) step was performed before restarting, using 10 MV of 0.1 M sodium hydroxide at 3 MV/min, which results in a 3 min duration of one CIP step, followed by an additional wash step with equilibration buffer until pH 7.4 is reached using a flow rate of 10 MV/min.

[0153] The frequency of CIP steps was diminished as the cycling experiment was progressing in order to reduce the cleaning intensity of the CIP step. For the first 20 cycles, a CIP step was performed after each cycle. For the cycles number 21-40, a CIP step was performed only after every 5.sup.th cycle. For cycle number 41-70, a CIP step was performed only after every 10.sup.th cycle. For all further cycles starting from the 71.sup.st cycle, no CIP step was performed anymore.

[0154] FIG. 11 shows the trans-device (delta) pressure as observed at each cycle during the equilibration phase at a flow rate of 10 MV/min for both devices.

[0155] First, it is important to realize that despite the very similar configuration of both materials (bed height is for both materials approx. 1 mm) the trans-device pressure at the same flow rate of 10 MV/min is much higher for the convective material (approx. 0.07 MPa) compared to the inventive chromatographic membrane (0.01 MPa). This is due to the smaller pore size observed for the convective Fibro material from Cytiva that is a prerequisite for attaining a binding capacity similar to the binding capacity of the inventive chromatographic membrane.

[0156] The convective Fibro material from Cytiva shows stable trans-device pressure during the first 20 cycles, were a CIP step followed after each bind & elute cycle. However, as the CIP frequency was reduced after cycle 21.sup.st to a CIP step only after every 5.sup.th cycle, the Fibro material starts to exhibit rapidly increasing trans-device pressures. At the 31.sup.st cycle, the trans-device pressure reaches 0.76 MPa, which well exceeds typical processing pressures of up to 0.4 MPa in related industrial applications of such materials.

[0157] The inventive chromatographic membrane Sample A shows a different behavior regarding the trans-device pressure development over the consecutive cycles: The trans-device pressure remains at the low level of 0.01 MPa for almost 79 cycles, despite the fact that the CIP frequency was continuously reduced during the 79 cycles until no CIP step was performed anymore after the 70.sup.th cycle. Slow increase in trans-device pressure is observed starting after the 79.sup.th cycle, reaching approx. MPa at cycle 150. This behavior is markedly different for those two materials despite the fact that the binding capacity is on the same level.

[0158] The data clearly shows the markedly different fouling propensity of the two separation materials. This behavior is most likely due to the significantly larger convective pore size of the inventive chromatographic membrane (approx. 4 m) compared to the purely convective cellulose fiber based membrane material (approx. 0.4 m).

Example 10

Integration into a Scalable Device Family

[0159] Inventive chromatographic membrane Sample A (cf. Example 8) was manufactured in roll format using a continuous coating process where the emulsion is coated onto the third inert phase. The emulsion is cooled down below the gelling point in order to form a continuous roll of membrane.

[0160] The inventive chromatographic membrane roll material containing surfactants and solvent is extracted by a continuous extraction and washing process using alcohol and water. Subsequently the roll material is crosslinked and simultaneously activated by impregnation with a bisoxirane solution containing sodium hydroxide. After washing in a continuous washing process, coupling of Protein A is performed in a jigger by winding the activated membrane roll between two winding rolls, thereby passing the reaction solution containing Protein A (10 mg/mL) dissolved in phosphate buffer (1 M, pH8.5). Subsequently, the Protein A membrane is dried from a humectant in a continuous rinsing/drying process to yield a Protein A membrane in roll format.

[0161] From these membrane rolls, spiral-wound chromatography modules with three different bed volumes at constant bed height could be generated. By changing the number of layers of the stacked inventive chromatographic membrane, various bed volumes and bed dimensions can be established based on targeted values like permeability, binding capacity or axial dispersion. In these modules, the membrane was rolled together with a spacer fleece in an alternating fashion.

[0162] Membrane layers were connected in a fluid-tight fashion to the housing by a melt process. Edges of the membrane bed were sealed in a previous step before integration into the respective housings. An impinging flow channel is given upstream from the membrane bed to the inlet and an outflow channel is given downstream from the outlet.

[0163] Table 5 shows the characteristics of the three modules differing in bed volume from 1.2 to 70 mL, which corresponds to a scaling factor of 58.3. The bed height is kept constant at 4 mm, ensuring almost identical permeability for all three modules of 7.6-8.3 MV/min*bar despite the different bed volumes.

[0164] Using state-of-the-art chromatographic systems, a single chromatographic bind & elute cycle of a clarified mAb feed stream was run as described in Example 9 with each of the three modules in order to evaluate the level of similarity of the chromatographic performance. For the 1.2 mL Nano device an kta 150 Avant system was used as described in Example 8. For the 10 mL Mini Capsule and the 70 mL 5-Inch Capsule, a multi-use Membrane Chromatography system (Sartorius, MCSCEB46NAK) was employed. The system is a liquid chromatography system intended for production and process development consisting of a configuration for the implementation of membrane chromatography based on Rapid Cycling Chromatography (RCC).

[0165] The generated UV chromatograms of the respective bind & elute cycles, normalized to the membrane volume permeated through the respective module (x-Axis), are depicted in FIG. 12. The chromatograms of the three devices are almost identical, which is a strong indication for a scalable performance, i.e. very similar chromatographic performance despite the wide range of module sizes, i.e. the wide range of bed volumes employed.

TABLE-US-00005 TABLE 5 Characteristics of the employed chromatography modules Nano Mini 5 Inch Capsule Capsule Capsule Bed volume [mL] 1.2 10 70 Bed height [mm] 4 4 4 Permeability [MV/min*bar] 7.6 8.1 8.3 Elution volume [mL] 5.0 5.3 5.3

Example 11

Robust Cycling Behavior for Different Feed Streams

[0166] In order to demonstrate the versatility and broad applicability of the inventive chromatographic membrane, three different clarified mAb fermentation broths that contained three different mAbs were purified in a rapid cycling process with a reinforced inventive chromatographic membrane, generated according to Example 6, Sample No. 34, having a third inert phase A according to Example 7, being modified according to Example 1. The inventive chromatographic membrane was integrated into a downscale device with 1.2 mL bed volume as described in Example 10. For each cycling study with 200 cycles a fresh device was used.

[0167] Table 6 gives an overview of the performed cycling study for the three different monoclonal antibodies. After equilibration, the device was loaded with feed to 80% of the earlier determined DBC 10% value, followed by a wash, the elution of the mAb, a CIP cycle and the re-equilibration. In column 3 of Table 6, the applied volumes for the different steps are indicated. For the CIP cycle, the volume was variable, as the device was flushed with 0.2 M NaOH until the pH of the effluent of the device reached pH 12.5 and flushing another 3.3 MV after reaching this pH value. Also for the Re-equilibration step, the applied PBS buffer volume was variable, as the device was flushed until the effluent reached a pH of 7.5. During all cycling studies an inline pre filtration of the feed was performed by using a Sartopore2 XLG 0.8/0.2 m (5441307G4-SS, 210 cm.sup.2), connected between sample pump and injection valve.

[0168] Cycling at an kta avant 150 chromatography system:

TABLE-US-00006 TABLE 6 Cycling method volume time step [MV] MV/min [min] equilibration 1x PBS pH 7.4 4.16 8.33 0.5 Feed ~80% mAb A (3.1 mg/ml) 8.75 4.16 2 of DBC 10% mAb B (4.3 mg/ml) 7.5 2 mAb C (2.3 mg/ml) 10 1.2 12.1 3 wash 1x PBS pH 7.4 8.33 8.33 1.2 elution 0.1M Acetate pH 2.9 + 8.33 4.16 2.4 150 mM sodium chloride CIP each cycle 0.2M NaOH until pH >12.5 + 4.16 ~2 3.3 MV (~10 MV) Re equilibration 1x PBS pH 7.4 until pH 7.5 (~21 MV) 8.33 ~2.5

[0169] FIGS. 13, 14, and 15 show the overlayed UV chromatograms of the respective 200 cycles for each antibody as a function of permeated volume. All three chromatograms show very similar and consistent behavior over all 200 cycles, indicating that the inventive chromatographic membrane did not change its characteristics or performance over the course of the cycling study. This can also be seen from FIGS. 16, 17, and 18 showing the overlayed trans-device pressures of the respective 200 cycles for each antibody/cycling study respectively. All three pressure traces show very similar and consistent behavior over all 200 cycles, again, indicating that the inventive chromatographic membrane did not change its characteristics or performance over the course of the cycling study. A slight decrease in pressure can be observed over the course of the 200 cycles. This drop in trans-device pressure can be attributed to a slow heat-up of the feed solution from 4 C. in the beginning of the scaling study to room temperature during the study decreasing the viscosity of the feed solution and therefore generating a reduced trans-device pressure.

[0170] The elution peaks (second peak in the UV chromatograms, starting at 30-35 mL volume) were fractionated from 100 mAu (first time signal surpasses this threshold) 100 mAu (signal falls below this threshold), and yield, elution volume, elution concentration, contaminant removal (HCP, DNA, aggregates) as well as levels of leached ProA ligand were determined. The methods for detection of yield, elution concentration and aggregates are size exclusion chromatography with a Yarra SEC 3000 column from Phenomenex, using a HPLC system of Thermo fisher scientific. HCP analyses were carried out with HCP ELISA Cygnus technologies F #550 3.sup.rd generation. DNA was analysed by using PICOGREEN assay P11496. ProA ligand concentration in the eluate fraction was detected with an ELISA of Repligen 9000/1, however, only for mAb A eluate fractions.

[0171] FIGS. 19 to 25 show the data sets for mAb A, mAb B and mAb C. It becomes apparent that all relevant process characteristics (yield, HCP removal, DNA removal, ProA ligand leaching, elution peak width, pool concentration, aggregate concentration) stayed at a constant level over the course of the 200 cycles for all three mAbs. This confirms the stable performance behavior of all three devices for different mAbs and respective feedstreams, despite the fact that only one generic cycling recipe was utilized. These results show the very robust and feed stream independent behavior of the inventive chromatographic membrane.

Example 12

High Chromatographic Performance Due to High Bed Heights

[0172] A reinforced inventive chromatographic membrane was generated according to Example 6, Sample No. 34, having a third inert phase B according to Example 7, being modified according to Example 1. In order to demonstrate the effect of bed height on chromatographic performance, the generated inventive chromatographic membrane was built into a chromatographic device with two different bed heights by stacking different number of membrane layers into the device, namely 3 and 12 layers. The flow distribution in front of the chromatographic bed as well as after the chromatographic bed was identical for both devices, independent of the bed height. The two chromatographic devices, only differentiated by the number of integrated inventive chromatographic membrane layers, had the following characteristics:

TABLE-US-00007 TABLE 7 Bed Trans-device pressure height drop at 5 MV/min 3 layers 0.78 mm 0.007 MPa 12 layers 3.1 mm 0.043 MPa

[0173] The chromatograms in FIGS. 26 and 27 show the different chromatographic behavior of the device differing only in bed height. Both the breakthrough curve and the elution peak show marked differences. For the low bed height device with only 3 layers of inventive chromatographic membrane, early breakthrough, a slow increase of the breakthrough curve and a wide elution peak can be seen, whereas for the high bed height device with 12 layers of inventive chromatographic membrane, late breakthrough with a relatively sharp increase of the breakthrough and a narrow elution peak are observed. These effects lead to higher binding capacity and faster elution, resulting in higher productivity, lower buffer consumption, and higher elution pool concentration. These effects can be attributed to the improved chromatographic behavior of the device with the higher bed height, due to lower axial dispersion.

Example 13

Materials and Performance Characteristics for Inventive Chromatographic Membranes Functionalized with Cation-Exchange or Mixed Mode Ligands

[0174] In order to demonstrate the versatility and platform character of the inventive chromatographic membrane, different functional groups were immobilized onto an inventive chromatographic membrane generated according to Example 6, Sample Nos. 34 and 35, described in the above-described general preparation example.

CIEX Functionalization:

[0175] CIEX ligand functionalization was carried out in a two-step process. In a first step, the activation of the inventive chromatographic membrane according to Example 6, Sample No. 35 is carried out with a bisoxirane molecule so that at least one oxirane group of the bisoxirane reacts with the membrane to enhance the mechanical stability of the matrix material. The second step is carried out by adding the ligand to the activated inventive chromatographic membrane at a basic pH. For this example, 3-bromopropanesulfonate (3-BPS) was immobilized onto the inventive chromatographic membrane with a concentration of the reaction solution of 20 wt % at pH=14 over a 24 hour period. Table 8 lists the properties of the generated cation ion exchange membrane.

TABLE-US-00008 TABLE 8 Membrane SBC volume in Ligand (- Sample Functiona- Thickness device Permeability density globulin) No. lization [m] [mL] [mD] [mol/mL] [mg/ml] 36 CIEX 260 0.52 297 92.6 79.4

[0176] In order to characterize the bind & elute performance of the CIEX functionalized inventive chromatographic membrane generated according to Example 13, sample No. 36, the membrane was assembled as a 4-layer membrane stack in a reusable filtration device with PP filter housing (filter table and filter cover) and stainless-steel holder as described in Example 8.

[0177] DBC 10% values of the inventive chromatographic membrane were determined at various flow rates to obtain different residence times (RT) of the target protein -globulin (Sigma-Aldrich, G5009) according to the method described in Example 1. The target protein -globulin (0.88 g/mL) was loaded onto the inventive chromatographic membrane in binding buffer conditions (sodium acetate, 20 mM, pH=5, 1.5 mS/cm) at flow rates of 15, 10, 7,5, 5, 2.5 and 1 MV/min, which result in residence times of 4, 6, 8, 12, 24 and 60 s. Sartobind S (Sartorius catalog, 96IEXS42EUC11-A) was examined under the same conditions, representing a state-of-the-art membrane material.

[0178] As a benchmark material representing a state-of-the-art CIEX resin material a 1 mL HiTrap Capto S column (Cytiva catalog, 29400458) was purchased from Cytiva.

[0179] The resins have an average bead size of 90 m, representing the diffusive phase. The column filled with Capto S resins was characterized with regard to the chromatographic performance, by determining the DBC 10% value for -globulin (4.74 g/mL) in the same binding buffer conditions at flow rates of 5, 1, 0.5 and 0.2 MV/min which is equivalent to residence times of 0.2, 1, 2 and 5 min as described above for the membrane material.

[0180] FIG. 28 shows the DBC 10% results as a function of the residence time for the CIEX materials.

[0181] The results show an increase in DBC 10% of 47% for the inventive chromatographic CIEX membrane in comparison to the respective state-of-the-art resin, and of 62% compared to the state-of-the-art membrane while having applied a 90% shorter or the same residence time of the target molecule in the chromatographic material, respectively.

Mixed-Mode (MM) Functionalization:

[0182] Ligand immobilization for a mixed mode ligand with a permanent positive charge was conducted on an inventive chromatographic membrane generated according to Example 6, Sample No. 34. Activation using a bisoxirane molecule according to the above-described general preparation example was followed by allylation with allylic bromide (20 wt %, 2 mol/L NaOH, 24 h). Bromination and the final coupling of the ligand (N-Benzyl-N-methylethanolamine, BMEA) were conducted according to U.S. Pat. No. 8,895,710 B2.

[0183] The inventive chromatographic MM membrane was characterized with respect to ligand density, SBC and DBC 10% according to Example 1.

[0184] Ligand density for the inventive chromatographic MM membrane was determined according to CIEX material titration with HCl as known to persons skilled in the art. SBC was determined with -globulin as target (3 mg/mL) in binding buffer conditions (Glycin-NaOH, 20 mM, pH=8, 40 S/cm). The elution step after washing was conducted in elution condition buffer (Glycin-HCl, 20 mM, pH=2.5, 2.1 mS/cm).

[0185] Results of the characterization of the inventive chromatographic MM membrane are indicated in Table 9.

TABLE-US-00009 TABLE 9 Number of layers in Membrane membrane volume in Ligand SBC (- Sample Functiona- Thickness stack device Permeability density globulin) No. lization [m] [] [mL] [mD] [mol/mL] [mg/ml] 37 MM 250 4 0.50 130 75 99

[0186] In this Example, the inventive chromatographic membrane according to Example 13, Sample No. 37, were assembled as a membrane stack in a reusable filtration device with PP filter housing (filter table and filter cover) and stainless steel holder and characterized with respect to the methods described in Example 1.

[0187] Determination of DBC 10% was conducted with -globulin as target (1 mg/mL) and with conditions used for SBC determination. DBC 10% values were determined at various flow rates to obtain different residence times (RT) of the target molecule in the chromatographic material. The flow rates have been adjusted to 1, 5, 10 and 20 MV/min, which corresponds to residence times of 1.0, 0.2, 0.1 and 0.05 min.

[0188] A 1 mL HiTrap Capto Adhere column was purchased from Cytiva (Cytiva catalog, 28405844) and used as a benchmark material representing state of the art mixed-mode resin material. The resin has an average bead size of 75 m, representing the diffusive phase. This state-of-the-art mixed-mode resin material was characterized with regard to the chromatographic performance, by determining the 10% DBC value for a -globulin (Sigma-Aldrich, catalog G5009) at residence times of 2 and 4 min which is equivalent to a flow rate of 2 and 4 MV/min as described above for the CIEX-membrane.

[0189] The results of DBC 10% as function of residence time are depicted in FIG. 29. The results show an increase in DBC 10% of 280% for the inventive chromatographic MM membrane in comparison to the respective state-of-the-art MM resin while having applied a 95% shorter residence time of the target molecule in the inventive chromatographic MM membrane.

[0190] The performance data with the different ligands shows the versatility of the inventive chromatographic membrane, and its applicability for different chromatographic tasks. Due to the ease of functionalization of polysaccharids, a broad variety of functional ligands can be immobilized to the inventive chromatographic membrane, generating respective chromatographic materials.