HK40007639A - The use of a polymeric mesh for the purification of macromolecules - Google Patents

Description

Use of polymeric networks for purification of macromolecules

Technical Field

The present invention relates to the use of a polymeric network comprising a polymer gel or composite for separating unwanted compounds from a solution or suspension containing the unwanted compounds and a target compound. The invention also relates to a specific polymeric network suitable for this purpose, and to a process for the preparation of a composite material comprising said polymeric network. The invention also relates to a method for recovering a protein of interest from a fermentation broth using the polymeric network. In particular, the invention relates to the isolation of recombinant proteins, preferably antibodies and antibody fragments, by using said composite material.

Background

Purified soluble macromolecules are very important substances throughout the industry. The increasing demand for biopolymer substances has been reported mainly in the pharmaceutical and medical fields, mainly for therapeutic and diagnostic purposes, but also for tissue engineering and other techniques. Among available chromatographic methods, Size Exclusion Chromatography (SEC) is considered unsuitable for large scale operations other than purification (refining) purposes due to the well-known low productivity resulting from low capacity, low resolution and low speed. In particular, the loading capacity of SEC is often very limited, since separation of molecules according to their size occurs within at most one total liquid volume within a packed column. Thus, the ratio between the interstitial volume and pore volume of a particular stationary phase and the pore size (pore size) distribution are the main characteristics of the limited sample volume leading to SEC.

Due to the large number of different impurities present in the original solutions of macromolecules, e.g. in crude extracts from almost all biological starting materials, in particular from living or dead tissues, tissues of various culture techniques and cell cultures, the first step of conventional chromatographic purification processes usually involves binding ("capture") of the target compound, whereas most of the undesired products are either completely unbound or can be separated from the target by selective elution steps, in which the bound impurities are released before or after the target substance.

However, it is very advantageous in terms of product recovery and overall process flow that most of the impurities are bound in the first step, while the purified target compound remains unbound in solution. Mainly for the purpose of antibody purification, the method should allow binding of most of the accompanying proteins present in the original raw material solution, which exhibit a molecular weight between about 10,000Da and 100,000 Da. Proteins of this molecular weight range correspond to a hydrodynamic radius R of about 1.5nm to 5nm_hThe molecular size of (a).

The prior art does not generally solve the corresponding problem in this way.

For the purification of antibodies and other proteins, some methods have been reported using complex adsorbents as separating agents, which comprise various support materials and amino polymers. The support material is a particle, a filter medium or a membrane. Polyethyleneimine, poly (allylamine) and poly (vinylamine) are preferred functional amino polymers, either covalently attached to the support or attached to the support after crosslinking.

For this purpose, the chromatographic method used is essentially characterized in that the filter material or the packed bed with the particles is perfused with various buffers for binding and elution, while retaining certain compounds of the starting material.

With respect to particle-based composites, the patent application families of WO 2013/007793 a1, WO 2013/007799 a1, WO 2013/037991 a1, WO2013/037992 a1, WO 2013/037993 a1, WO 2013/037994 a1, WO 2013/037995a1 ("WO 994" family) include coatings of spherical silica and polystyrene sulfonate carriers with polyvinylamines of unknown origin. Often these complexes have been further derivatized in an effort to selectively separate a variety of pharmaceutical compounds. Adhesion of poly (vinylamine) coatings is typically achieved by a two-step procedure: the support material is soaked with an aqueous polymer solution comprising pores and the material is then dried. In a second step, crosslinking is carried out in suspension after dissolving the crosslinking agent in a suitable organic solvent which no longer dissolves the precipitated polymer.

WO 90/14886 discloses a composite separation medium for protein separation comprising a substrate carrying a plurality of polyamines which are covalently linked to the substrate.

WO 95/25574 relates to a method of removing contaminants from a biological fluid comprising contacting the biological fluid with a cross-linked hydrophobic polymeric network which covers but is not covalently bound to a porous mineral oxide matrix, the porous volume within the porous mineral oxide matrix being substantially filled with the hydrophobic network, thereby removing hydrophobic and amphiphilic molecules having an average molecular weight below 10,000 daltons.

US 6,783,962B 1 describes a particulate material consisting of a non-porous core and a polymeric matrix (e.g. dextran) comprising pendant groups that can be charged or affinity ligands for binding biomolecules. The pendant chargeable group is polyethyleneimine or modified polyethyleneimine, and may form a tentacle-like structure. The material is used for isolating biological macromolecules, such as DNA.

WO 2004/073843 discloses a composite material comprising a support member having a plurality of pores and a macroporous cross-linked gel filling the pores of the support member.

US 2010/0200507 a1 relates to the purification of biological samples using a cross-linked polyamine coating immobilized on a membrane.

Most attempts at complex formulations involve polymerizing the monomer within the empty spaces of the volume or surface of the carrier material. These methods may also include covalent attachment of polymer chains produced on the surface. The present invention preferably does not relate to polymerization but describes the use of preformed polymers.

For the synthesis of composite materials comprising a preformed polymer permanently fixed to a carrier material, basically two methods are available:

covalent attachment of the Polymer to the surface of the support Material

Fixing the polymer on the surface or in the pores by crosslinking.

Covalent attachment is more laborious and expensive and is additionally limited by the layer thickness attached to the surface. On the other hand, if the support surface is flat or uneven, this is the preferred method, simply for stability reasons, for example with films, fabrics or tissues.

Crosslinking requires less work and is suitable for porous materials, since the resulting network is well captured and can form thick layers. On the other hand, coating porous particles in this way presents significant problems as they may inadvertently stick together and/or the pores may become blocked.

Disclosure of Invention

In order to attach the polymer to the pores of the particulate or porous monolithic support material by cross-linking, two synthetic routes are again provided:

first precipitating the polymer onto the inner surface of the support, followed by cross-linking the precipitated material by: the crosslinking agent dissolved in the solvent (which is no longer capable of dissolving the polymer) is added and the reaction between these components is then initiated until the desired degree of crosslinking has been established.

This procedure definitely avoids pore clogging and particle attachment. On the other hand, as shown in the present invention (see fig. embodiments 1.1 to 1.4 and table 2), a different and less advantageous morphology is obtained in terms of accessibility of the polymer.

The polymer and the crosslinking agent are added simultaneously in order to fill the pores of the support with the mixture and then to initiate the reaction between these components. The reaction is continued until the desired degree of crosslinking is established, thereby mediating immobilization of the resulting polymeric network without intermediate physical manipulation.

The skilled person will avoid introducing large amounts or volumes of dissolved preformed polymer and crosslinking agent simultaneously into the porous support material for the reaction, since there is a risk of clogging the pores of the support material, thus resulting in poor mass transfer (masstransfer) of the modified product.

However, it was unexpectedly found that neither plugging nor ligation occurred under the synthesis conditions of the present invention. In addition, it is generally not recommended to premix the functional polymer and the reactive crosslinking agent. However, if the crosslinking reaction is initiated at high temperatures, the relevant method as used in the present invention is preferred.

Basically, the prior art avoids "one-step synthesis" of the composite, always with intermediate or subsequent drying of the unfinished composite. Moreover, the prior art avoids the presence of any reactive solution outside the pore volume of the support. The clogging of the pores of the carrier by the polymer and the unintentional attachment of the particles seems to be an insurmountable problem, which has now been solved by the present invention.

Therefore, upstream drying prior to completion of the manufacturing process seems to be the best way to prevent this risk in the past.

The concept of avoiding the step of capturing the target compound can be defined as a "negative chromatography" characterized in that most of the unwanted compounds are adsorbed "positively" and thus removed. The strategy to achieve this goal is outlined in steps I, II and III below. It is important to distinguish any method by which the "do not bind by exclusion" is rejected by the adsorbent from the target compound simply due to the same charge, e.g., of the surface and molecules. The following three discrimination mechanisms, the strategies implemented using the materials and methods of the present invention, are characteristic features:

I. removing substances in the recombinant antibody having a molecular weight lower than that of the target substance, e.g., most host cell proteins and BSA, wherein the impurities may be derived from the host cells and cell culture media used in the corresponding fermentation process.

Simultaneous removal of high molecular weight compounds and nanoparticles such as nucleic acids, viruses and fragments thereof.

Recovering the target compound, typically a molecule, e.g. a protein, that is too large to enter the volume of the pores of the network, while avoiding (strong) binding of the molecule(s) to the outer surface of the medium (media).

It is therefore an object of the present invention to provide methods and materials that enable purification methods to achieve the goals of I, II and III.

The "negative" separation step is achieved by steric exclusion of the target compound outside the specific mesh pore volume into which smaller molecular size impurities enter, where they may be retained by covalent binding or adsorption to the inner surface, or partition to the mesh volume due to partitioning mechanisms or simple inclusion. Adsorption is preferably achieved by non-covalent interactions, mainly including ionic, hydrogen bonding, dispersive, van der waals and coordination (ligand exchange, donor-acceptor) forces.

Thus, the material of the invention in principle incorporatesAdsorption,Partition chromatography and size exclusion chromatographyHas the advantage of allowing the target compound to be excluded while retaining the impurities within the polymeric network.

Accordingly, the present invention provides a process for recovering a target compound from a feedstock comprising at least one target compound and at least one impurity, characterised in that a filtered or virgin feedstock is contacted with at least one polymeric network for a sufficient time while at least one impurity is retained by the polymeric network, the polymeric network is then separated from a purified feedstock containing the at least one target compound, and optionally, the at least one target compound is separated from the feedstock.

Furthermore, the relatively small contribution of the outer particle surface of the porous material relative to the entire particle surface significantly contributes to the desired negative adsorption effect, since for undesired adsorption of the target compound to the outer surface a very small area is left.

At the same time, the significantly larger inner surface provides high binding capacity for a greater amount of low molecular size impurities that can enter the inner pore volume. In addition, certain low concentration but highly charged and high molecular weight contaminants, due to their molecular size, are also excluded from the internal pore volume, such as nucleic acids and viral particles and fragments thereof, which can be effectively captured on the external surface of the media, thereby acting as adsorption competitors for the sterically excluded target compounds.

It is important to underline that the present invention is capable of incorporating inside said polymeric network macromolecules having a molecular weight of up to about 100,000Da, equal to the hydrodynamic radius R_hBelow about 5nm, and not just small molecules such as drugs or drug metabolites. Furthermore, the target compound is not bound to the pore volume of the polymeric network, but remains excluded. This method is in direct contrast to the limited access method, where the unwanted compounds remain unbound and often little of the target substance penetrates the porous space where it is adsorbed.

The purification strategy of the present invention is unusual so far, but generally implies a valuable approach to the separation problem. Since the amount of impurities in the original solution (e.g. plant extract or cell culture supernatant) is in the range of up to several thousand fractions, containing many different structures, the composite material must provide very high generic affinity in order to reach a > 90% removal level in one step. In general, the present invention provides the following teachings:

a method for the efficient purification of macromolecules, characterized in that the target molecule is excluded from the pores of the polymeric network of the stationary phase or adsorbent, while the majority of impurities are retained by the internal pores. The interior bore includes the interior and volume between fixed polymer coils.

A rapid, simple, inexpensive and durable (rugged) synthesis of composites comprising a polymeric network, characterized in that a dissolved mixture of polymer and crosslinking agent in solvent A is added to a support material, which after reaction and swelling in solvent B yields a nanoporous network, the volume of the reaction mixture preferably exceeding the pore volume of the suspended support material.

Material design of a polymeric network (soft gel or composite) comprising a crosslinked polymer, characterized in that the network exhibits pore sizes in a specific solvent, in particular an exclusion hydrodynamic radius of at least R_h1While the hydrodynamic radius is lower than R_h1Is retained by the internal pores after penetrating the network.

The most important field of application in connection with the present invention is in fact biotechnology. In addition, other polymer purification problems are also applicable to the present invention.

In biological separations, for example starting from fermentation broths or body fluids containing proteins of interest, such as antibodies, the aim is to remove substances, such as DNA, RNA, Host Cell Proteins (HCP), abundant protective or raw material proteins such as BSA, transferrin (naturally occurring or introduced as a component of the culture medium for cell growth) and endotoxins and pathogenic bacteria or fragments thereof. Furthermore, detergents that are often added to achieve better cell growth or preservation should preferably be removed.

A general task in bioseparations is therefore to remove the substances listed under a) to g). In fact, there are no methods and materials available to solve the associated separation problems at low cost levels in one or two steps.

When a fermentation broth (before or after filtration) or a Cell Culture Supernatant (CCS) is used as a raw material containing a target compound, the target compound is a recombinant protein, preferably an antibody, and the raw material contains the following classes of compounds as impurities:

a) DNA, RNA, other nucleic acids, proteins, and organic substances having a molecular weight of at least 100,000 daltons;

b) host Cell Proteins (HCPs) including proteases with molecular weights below 100,000 daltons

c) Albumin (BSA, HAS, ovalbumin);

d) other proteins present in the cell culture medium and substances of various molecular weights derived from nutrients or cell metabolism;

e) an endotoxin;

f) a detergent; and

g) bacteria and microorganisms such as viruses or fragments thereof.

It is therefore an object of the present invention to provide a purification process, a separation material suitable for this purpose and a process for the synthesis of said separation material, which show the improvements and advances described below.

It is an object of the present invention to provide a separation and purification process which simultaneously removes several impurities, preferably substances belonging to structurally different classes, from a solution (starting material) while at least one target compound remains unbound and recovered in high yield.

It was unexpectedly found (see examples) that the polymeric network of the present invention is capable of removing host cell proteins, nucleic acids and nutrient proteins simultaneously to a very high degree, while a complex mixture of antibodies from human plasma is recovered in high yield.

The present invention therefore relates to the use of a polymeric network which is a soft gel or a composite. By soft gel is meant a network synthesized from monomers and cross-linking agents or preferably from preformed polymers and cross-linking agents, thus producing a porous solid material comprising linked polymer chains or coils, which are immobilized by covalent or non-covalent cross-linking.

In a preferred embodiment, in combination with any of the above or below embodiments, the corresponding soft gel comprises a cross-linked functional polymer (preferably a cross-linked amino group-containing polymer) and a plurality of cross-linking agents. In a more preferred embodiment, in combination with any of the above or below embodiments, the composite material preferably comprises a crosslinked amino group-containing polymer that is immobilized on a porous support material to form a composite.

In a different embodiment, in combination with any of the above or below embodiments, the polymer may be covalently attached to the surface of the support material, and optionally additionally crosslinked.

Both soft gels and composites can be used to separate unwanted compounds from solutions (raw materials) containing both the unwanted compounds and the target compounds.

The target compounds of the present invention include polymers, preferably biopolymers, more preferably proteins, and most preferably antibodies.

The raw material solution or suspension comprises at least two dissolved substances of synthetic or natural origin, preferably a fermentation broth, which is filtered (cell culture supernatant) or still contains solid debris such as cells and debris.

An advantageous and improved separation technique should allow:

1. impurities are simultaneously separated from the purified dissolved target compound to a high degree.

2. Capture of the target compound is avoided because binding to the resin and subsequent elution increases overall process costs and may also reduce product yield. Instead of the usual capture procedure, irreversible binding or at least strong binding of most impurities is opposed.

3. The target compound is rapidly and inexpensively purified from a feedstock solution or suspension (e.g., from a fermentation broth) by reducing the need for expensive equipment and materials, as well as the loss and degradation of valuable target compound due to time-consuming operations.

4. Preferably, a chromatography step is avoided, which is characterized by a flow through the separation material. In particular, flow-through or perfusion is disadvantageous for removing less strongly bound impurities, since they can co-elute with the target compound.

5. A clear solution containing unbound, purified target compound is obtained quickly in high yield and without adsorbing a large portion of this valuable compound onto the composite.

6. The separated material is preferably treated after one cycle without regeneration, thereby avoiding tedious validation of the desired number of process cycles.

According to the invention, there is provided a fluid containing fluid having a hydrodynamic radius below R_h1And a hydrodynamic radius R_h1Or higher than R_h1Solves the above technical problems 1 to 6, preferably problems 1, 2 and 5, and achieves the object by providing a process for recovering a target compound from a starting material of at least one target compound, comprising the steps of:

contacting said feedstock with a polymeric network comprising at least one aminopolymer for a sufficient time and said hydrodynamic radius is less than R_h1At least one compound of (A) isThe polymeric network is retained and subsequently the polymeric network is contacted with a fluid having a hydrodynamic radius R_h1Or higher, and optionally separating the target compound from the feedstock.

R_h1Is defined as the "exclusion limit" and ranges between 1nm and 20nm, preferably between 3nm and 10nm, most preferably between 4nm and 6 nm.

Isolation refers to obtaining the purified compound of interest as a solid or extract by extraction, evaporation, freeze-drying or other known methods.

In a preferred embodiment, in combination with any of the above or below embodiments, the polymeric network comprises at least one amino polymer.

Specifically, according to the present invention, there is provided a method of recovering a protein of interest from a starting material (solution or suspension) comprising at least one protein of interest and impurities comprising Host Cell Protein (HCP), DNA, RNA or other nucleic acids, and optionally albumin, endotoxin detergent and microorganisms or fragments thereof, comprising the steps of:

i) contacting the feedstock with a polymeric network comprising at least one amino polymer for a sufficient period of time while at least one impurity compound is retained;

ii) subsequently, separating the polymeric network from the purified feedstock containing the at least one protein of interest;

iii) optionally, isolating the target protein from the feedstock;

in a preferred embodiment, in combination with any of the above or below embodiments, the target compound is dissolved; in a different embodiment, the target compound is at least partially suspended. The impurities may also be at least partially suspended.

In a preferred embodiment, in combination with any of the above or below embodiments, the solution or suspension is preferably a filtered or raw fermentation broth.

At least one retained compound is a subset of a portion comprising Host Cell Proteins (HCPs), DNA, RNA or other nucleic acids, albumin, endotoxins, detergents, and microorganisms.

In a preferred embodiment, in combination with any of the above or below embodiments, at least one target compound remaining in the purified feedstock according to step ii) is excluded from the polymeric network volume.

Optionally, in step iv), the polymeric network is washed with a weak solvent and the resulting solution is collected for further processing.

In a preferred embodiment, in combination with any of the above or below embodiments, wherein the amino polymer is poly (vinylamine) orPoly (vinylformamide-co-vinylamine)。

In a preferred embodiment, in combination with any of the above or below embodiments, the above feedstock is contacted with the polymeric network comprising at least one aminopolymer for a sufficient period of time to provide a hydrodynamic radius R_h2At least one impurity compound below 4nm is retained by the polymeric network comprising the amino polymer while having a hydrodynamic radius R_h1At least one target protein of 4nm or more remains in the purified raw material.

By definition, the pore volume of a polymeric network is considered to be the specific volume of the interior of the network spanned by the immobilized polymer coils. The outer web surface and the outer surface of the carrier material do not contribute to the volume of the web.

By polymeric network retained is meant removal within the pores of the network by any non-covalent or covalent binding mechanism, such as adsorption, or by partitioning, size exclusion or extraction mechanisms.

In a preferred embodiment, in combination with any of the above or below embodiments, the polymeric network comprises a composite material.

Accordingly, the present invention provides a method for the synthesis of a composite material for separation corresponding to step i), wherein a porous support material is filled with a cross-linked functional polymer, preferably with a cross-linked functional polymerPoly (vinylformamide-co-vinylamine)Linear or branched poly (vinylamine), poly (allylamine), poly (ethylenimine) or polylysine, or copolymers containing such amino polymers.

Furthermore, according to the present invention, the composition of poly (vinylformamide-co-vinylamine) comprises 5% to 80% of poly (vinylformamide), preferably 10% to 40%, more preferably 10% to 20%.

Furthermore, for any contact according to step i) or equivalent to step i), the polymeric network comprising the aminopolymer is pre-equilibrated to a pH below 8, preferably between pH 3 and 7.5, more preferably between pH 4 and 7, most preferably between pH6 and 6.8, by mixing with an aqueous buffer or a salt solution, preferably of a monobasic acid. The balance is also a rule for handling soft gels. The buffer or salt concentration in the polymeric network after equilibration is less than 500mM, preferably between 10mM and 200mM, more preferably between 20mM and 100 mM. Before use, at least the equilibrated polymeric network is wetted with the buffer or salt solution, preferably at least the pore volume is filled.

Preferred monoacids are formic acid, acetic acid, sulfamic acid, hydrochloric acid, perchloric acid or glycine. Preferred cations are ammonium, alkylammonium, sodium and potassium.

In a preferred embodiment, in combination with any of the above or below embodiments, the salt is ammonium acetate.

Optionally, when the target compound is basic, the pH of the starting material is adjusted to a range of 4 to 7.

The following are preferred embodiments of the process for isolating a compound of interest according to the invention:

the separation process of the invention preferably involves a feedstock, e.g. a fermentation broth, which represents a filtered solution or a feedstock suspension, which still contains, for example, cells and cell debris.

In a preferred embodiment in combination with any of the above or below embodiments, a solution or suspension comprising the recombinant protein as the target compound and comprising host cell protein, DNA and BSA as impurities is contacted with the polymeric network comprising the amino polymer and impurities from the starting materials are simultaneously removed.

In a preferred embodiment, in combination with any of the above or below embodiments, the polymeric network comprising the amino polymer is used to completely remove BSA from a 5% (w/v) solution.

Unexpectedly, it was found that with the polymeric network of the invention, impurities with a pI (isoelectric point) value of 7 and higher than 7 were removed by at least 95% (examples). Binding to isoelectric focusing using a host cell specific ELISA showed that most of the neutral and basic compounds removed were host cell proteins (see methods).

Thus, in a preferred embodiment, in combination with any of the following embodiments, host cell proteins having a pI of 7 or greater than 7, or exhibiting basic properties, capable of ionic interaction, are removed by 50%, preferably 80%, most preferably at least 90%, when using a positively charged polymeric network comprising an amino group containing polymer.

Thus, the present invention provides a method for removing compounds having a pI of 7 or greater, using a positively charged polymeric network, preferably a complex, containing amino groups, most preferably comprising poly (vinylamine) or morePoly (vinylformamide-co-poly- Vinylamine)The complex of (1).

The polymeric network is positively charged after equilibration with a solution having a pH below 8, preferably below 7.

In a preferred embodiment, in combination with any of the following embodiments, the hydrodynamic radius R is the radius of the solvent used_h1Target compounds above 5nm, preferably above 4nm, and calibrated test substances are sterically excluded from the meshOutside the pore volume of the cake and thus separated from the other components. Thus, the present invention provides a method of recovering a hydrodynamic radius R_h1A method for the production of compounds above 4nm, using a positively charged polymeric network containing amino groups, while at least 80%, preferably 90%, of said compounds are retained in the liquid phase.

See also the methods section for the relationship between molecular weight and hydrodynamic radius.

In a preferred embodiment, in combination with any of the above or below embodiments, a one-step batch adsorption process is used instead of chromatography in steps i), ii) and iii), characterized in that the purified starting solution is removed by precipitation or centrifugation from the impurity-loaded composite adsorbent.

It is emphasized that batch removal (batch removal) is based on a diffusion mechanism. Preferably, no columns or other devices requiring a convective transport mechanism (e.g., flow-through) are employed.

The person skilled in the art would not expect to obtain a satisfactory purification using only one unit operation (corresponding to any one theoretical plate separation step), since in this method there is usually not sufficient selectivity for complex sample mixtures. A general theory is that due to the wide distribution of binding constants of the various compounds in the composite feedstock, successful separation invariably requires the use of high plate numbers in chromatographic columns or selective gradient elution techniques. However, for the present invention, selectivity for a wide variety of impurities is created for high molecular weight target species by size exclusion mechanisms. Thus, by combining the adsorption and size exclusion mechanisms designed with the materials and methods of the present invention, a high level of purification can be achieved.

In a preferred embodiment, in combination with any of the above or below embodiments, the duration of step i) is from 5 to 30 minutes.

In a preferred embodiment, in combination with any of the above or below embodiments, effective removal of impurities can be achieved by a single purification step.

In a further preferred embodiment, in combination with any of the above or below embodiments, the porous support material is a particulate material having an average particle size of from 3 μm to 10mm, preferably between 20 μm and 500 μm, most preferably between 35 μm and 200 μm.

In a preferred embodiment, in combination with any of the above or below embodiments, the compound of interest is a polymer, preferably a biopolymer, more preferably a protein. Biopolymers include peptides, proteins, glycoproteins, lipoproteins, nucleic acids and any other compounds produced by living organisms with a molecular weight greater than 500 Da.

In a further preferred embodiment, in combination with any of the embodiments described below, the protein is an antibody, a pegylated antibody, or another derivative of an antibody or an antibody fragment.

The invention therefore relates to a purification method comprising steps i), ii) and optionally iii), characterized in that the compound of interest is an antibody.

Antibody herein refers to any immunoglobulin of human or other origin, either from recombinant proteins of any kind of cell culture or cell-free system for protein synthesis or isolated from biological fluids or tissues.

In a preferred embodiment, in combination with any of the above or below embodiments, it was unexpectedly found that no aggregates are formed in the polyclonal antibody (hIgG) mixture even after 20 minutes after contact with the polymeric network comprising the aminopolymer. In contrast, many conventional capture methods suffer from yield loss after non-covalent binding on the resin surface. Furthermore, due to the formation of aggregates during the elution step, a large part of the antibody is lost even from the affinity column.

Detailed Description

The combination of typical essential technical features allows the design of a large number of possible embodiments. Without any comprehensive requirement, the following items are considered to be important embodiments according to the invention.

In a preferred embodiment, in combination with any of the following embodiments, the undesired compound is selected from the group consisting of DNA, RNA, albumin, Host Cell Protein (HCP), endotoxin, detergent, bacteria and virus. Fragments of the undesirable compounds, such as coat proteins, S-layers, cell debris or fragments, are also within the scope of this embodiment.

In a preferred embodiment, in combination with any of the above or below embodiments, the compound of interest is an antibody and only the above listed impurities a), b) and c) are removed from the solution. In another preferred embodiment, in combination with any of the above or below embodiments, the compound of interest is an antibody and only the above listed impurities a) and b) are removed from the solution. In another preferred embodiment, in combination with any of the above or below embodiments, the compound of interest is an antibody and only DNA and host cell proteins are removed from the solution as impurities (unwanted compounds).

Preferably, the contaminants or impurities are removed from the feedstock (e.g. a biological fluid, a supernatant of a fermentation process or a fermentation broth before filtration) to an extent of > 90%, > 95%, > 99% of their respective total amount in the feedstock, the concomitant binding not exceeding 10%, preferably 5%, more preferably 1% of the total amount of the target substance.

The invention therefore relates to a purification process comprising steps i), ii) and iii), characterized in that impurities are removed by at least 90% and the target protein is recovered by at least 90%.

In a preferred embodiment, in combination with any of the above or below embodiments, host cell proteins are removed in an amount of at least 90%, preferably at least 95%, more preferably at least 99%.

Accordingly, the present invention relates to a purification process wherein host cell proteins are removed from the feedstock at least 90% of their initial concentration.

In a preferred embodiment, in combination with any of the above or below embodiments, a volume of a particular feedstock (e.g., from a fermentation process, before or after removal of solid material such as cell culture or supernatant) is contacted with a sufficient amount of a suspension of the amino polymer-containing composite material. After stirring or shaking for a suitable time, the composite material is separated from the removed raw material solution by, for example, suction, filtration or, preferably, sedimentation or centrifugation.

When the starting material is a suspension, e.g., containing cells, cell debris or tissue, these solids are removed along with the composite material. The remaining centrate, filtrate or fraction contains the purified target compound.

When a particulate composite material is used, some of the target compound may remain in the interstitial volumes between the particles. In this case, a very weak solvent of external void volume is applied to displace a major part of the solution containing the target. Depending on purity and yield, this additional volume may be combined with the target major fraction, or may be dedicated to another purification step. In the case of purifying the antibody from the cell culture supernatant, the solvent used for the replacement may be a weak buffer containing little water or being far different from the target protein, preferably the antibody pI.

The weak solvent or weak buffer preferably exhibits no chromatographic elution capability and is characterized by low solubility for most of the impurity compounds of the starting material.

In a preferred embodiment, in combination with any of the above or below embodiments, the ratio of raw materials to composite material ranges between 5 and 100 liters per kg, with a preferred contact time of 5 to 60 minutes.

As long as the target compound remains unbound to an acceptable degree in the liquid phase, there is generally no need for subsequent elution of the bound substances, since the composite materials are preferably handled at the end of the process cycle, i.e. they are designed for one-time use. However, elution and isolation of any bound compound is still within the scope of the present invention and may be achieved by applying any known elution method.

In order to meet the stringent quality requirements of an API (active pharmaceutical ingredient), a target compound purified according to new technical knowledge may require one or two additional purification steps. This may be the case if removal below the limit of detection is required, or complex heterogeneous by-products or impurities (such as host cell proteins) must be removed to levels below 10ppm, depending on the quality of the final API.

Thus, the sequential use of any polymeric network of any of the following or above embodiments with any other purification steps is the subject of the present invention. In combination with any of the above or below described embodiments, their use before or after an ion exchanger or affinity chromatography step, or any other purification step, is within the scope of the present invention, especially if affinity based separation steps are contemplated, e.g. selective adsorption of the target compound onto any kind of separation medium containing protein a, protein G or a combination of both. Furthermore, any combination with membrane filtration, depth filtration or application of integral separating agents is considered to be within the scope of the present invention. In a more preferred embodiment, in combination with any of the above or below embodiments, the polymeric network is used before or after an ion exchanger or affinity chromatography step or other purification step.

The present invention therefore relates to a combination with one or more further separation steps, characterized in that the above-mentioned steps i), ii) and iii) are carried out with the original starting material suspension or solution before any further chromatographic or non-chromatographic purification steps are carried out.

Although the target compound is excluded from the polymeric network, small amounts may be adsorbed on its outer surface. In order to obtain sufficient yields and/or recoveries of the target compound, it is necessary to avoid binding of large amounts of the target compound to the outer surface of the polymeric network. Due to the inevitable distribution of the pore diameters of the polymeric network and due to potential interactions with the outer surface, it is generally not possible to completely avoid the loss of small amounts of the target substance. Since these losses may reduce the overall recovery of valuable materials, additional strategies have been incorporated into the present invention to minimize such losses.

Together with the exclusion effect provided by the specific porosity of the polymeric network, the present invention therefore also employs the following general design principles:

A1. for use in an aqueous environment, the portion of the polymer in contact with the mobile phase should be polar, approaching the polarity and/or charge of the target compound. Thus, the solvated target compound is remotely separated from the composite and is not bound thereto.

A2. Thus, in organic solvents, lipophilic targets are repelled by hydrophilic polymers and vice versa.

B1. When the polymer is charged, the pH of the binding buffer is below or above the isoelectric point (pI) +/-1 units of the target compound, provided that it is an ionic species, to keep the molecule almost uncharged.

B2. More preferably, the polymeric network is equilibrated with a buffer or solvent to a pH below or above the isoelectric point (pI) +/-1 unit of the protein of interest, most preferably a pH where binding of the compound of interest is minimized, prior to contact with the starting material.

Preferably, the pH conditions of the preliminary network equilibrium are kept sufficiently constant during the contact time of less than 20 minutes to avoid said undesired interaction with the target compound.

C. High salt concentrations (preferably >100mM NaCl equivalent or conductivity >10mS/cm) reduce polar binding forces (mainly ionic charge interactions on the outer surface of the polymeric network) and are therefore less attractive to target compounds. However, strong binding forces (e.g. multivalent binding forces) within the polymeric network are also capable of adsorbing incoming molecules under high salt loading.

D. The multivalent strong binding of large macromolecular chain molecules on the outer surface (such as small amounts of DNA during fermentation) will hinder or inhibit the binding of the target molecule by competitive displacement, as long as it does not interact with the DNA itself.

Thus, the use of the polymeric networks of the present invention includes the preliminary steps of washing and equilibrating the polymeric network with a solvent or buffer, adjusting the pH and ionic strength to avoid or minimize binding of the target compound, or preferably using a solvent that elutes less strongly than the solvent of the starting material. This measure also improves the binding of impurities with lower binding constants.

In combination with any of the above or below embodiments, steps i), ii) and optionally iii) of the present invention are combined with one or more of the measures described above under a.

By "target compound" is meant any substance of value which is subject to purification according to the present invention.

It is also an object of the present invention to provide a method for preparing an inexpensive composite material. The composite material enables multiple substance separation or purification processes, such as the simultaneous removal of several structurally different classes of substances from a solution (feedstock).

The object is achieved by a method comprising:

at least the pore volume of the porous support material is filled with a solution (reaction mixture) of at least one functional polymer or copolymer and at least one crosslinking agent and the functional polymer is fixed in situ by crosslinking, while the support material is in the form of particles, films or monoliths.

By thin film is meant a solid particle or material coated with a porous layer.

Monolithic means a uniformly porous sheet of carrier material having a thickness of at least 0.5 mm.

In a preferred embodiment, in combination with any of the embodiments described below, the pores are filled with a mixture of the functional polymer and the crosslinking agent and reacted in a one-step process without prior or intermediate drying. In another embodiment, in combination with any of the embodiments described below, after filling the pores, the solvent may be removed, either completely or partially, before the crosslinking reaction begins. When using an epoxide crosslinker or a crosslinker that is not reactive at ambient temperature, the solvent reduction is preferably achieved by evaporation at a temperature below 30 ℃. When the desired portion of the solvent is removed, the crosslinking reaction begins at a temperature of 50 ℃ or above 50 ℃. In another embodiment, in combination with any of the embodiments described below, the empty space may be partially or completely filled with a different solvent after partial or complete removal of the solvent.

In all of these above cases, the cross-linking agent is already present during the initial pore-filling step, and the reaction proceeds at the interior or interface of the porous support. If the carrier material is a component, e.g., a stack, of granular or monolithic articles, the process optionally includes:

filling at least the settled volume of support material (see method) (pore volume and interstitial volume between particles or layers) with the reaction mixture; or

Optionally applying an excess of reaction mixture containing up to 120% of the settled volume of reaction mixture.

In a preferred embodiment, in combination with any of the following embodiments, the support material is filled with the reaction solution, allowing the liquid to spontaneously soak into the pores. Any other hole filling method known in the art is also suitable.

However, it is difficult to accurately fill the entire pore volume of the porous particulate support material using the soaking technique. Also, the pore volume measurement (comparative method) always implies a certain error, and it becomes more difficult to accurately determine the required volume of the reaction solution to be applied. It is therefore difficult to avoid that a large part of the particles will be slightly overloaded with liquid on the outer surface, e.g. simply because of the surface tension of the liquid. As a result, when the volume of the reaction mixture added is only equal to the measured pore volume, the pore portion of the other particles cannot be completely filled.

The problem of uneven particle packing becomes even more severe, especially when handled in large quantities, since it is substantially impossible to contact all particles with liquid at the same time during the manufacturing process.

For special applications, it is absolutely necessary to completely cover the accessible surface of the support material with the crosslinked polymer. Therefore, the above-mentioned types and the degree of inaccuracy are not negligible. The part of the surface of the support body that is not covered by the polymer will have a negative influence on the selectivity and mainly on the recovery during the separation. The target compounds may be more strongly adsorbed at these sites, particularly proteinaceous target compounds on polar support materials such as silica or other polar media.

The use of a sufficient excess of reaction mixture volume to enable complete wetting and coverage of the polymer over the entire support surface can avoid the problems described. In this case, however, it is necessary to prevent any crosslinking reaction outside the volume of the particles. Furthermore, a rapid reaction of the polymer and the crosslinking agent within a sufficient pot life after the preparation of the reaction mixture is also unacceptable.

It has surprisingly been found that even when the interstitial volume between the particles is partially or completely comprised of the reaction solution, no particles are fused together. Without being limited to any explanation, in this case the cross-linking agent may be adsorbed by the porous support material (example 1). Thus, if an amount of excess polymer crosslinker solution is applied, complex preparation is not adversely affected. These unexpected findings allow to simplify the manufacturing process according to the invention, especially when producing large quantities of composite material, since the reaction is preferably carried out with the settled support material without stirring, shaking or other movements.

In combination with any of the above or below embodiments, at least the pore volume of the support material is filled with the reaction solution, preferably an excess solution is added in relation to the pore volume, more preferably a sediment volume, most preferably a slight excess sediment volume. Thus, in combination with any of the above or below embodiments, there is provided functionalityPolymers, preferably poly (vinylamine) orPoly (vinylformamide-co-poly- Vinylamine)With a bis-epoxide, preferably ethylene glycol, propylene glycol, butylene glycol or hexane diglycidyl ether, in an amount of at least the pore volume, preferably the sediment volume, most preferably 110% to 120% of the total sediment volume, while the pores of the support material become completely filled. Unexpectedly, no polymer gel formed outside the pores at the end of the reaction, nor did the particles stick together.

In a preferred embodiment, also in combination with any of the above or below embodiments, an excess of the functional polymer solution containing the crosslinking agent is added, preferably 110% to 120% of the settled volume of the support material, so that the interstitial volumes between the particles are completely filled with liquid and a thin liquid film of the reaction solution is covered on top of the settled solids.

In a more preferred embodiment, in combination with any of the above or below embodiments, the cross-linking agent is applied in water or an aqueous solution with the cross-linkable polymer. Although even amounts of cross-linker below 2% (v/v) are not completely soluble in water, preferably the diepoxides described above, most preferably hexanediol diglycidyl ether, are used, the resulting emulsion is surprisingly distributed within the pores of the support material, resulting in a stable cross-linked polymeric network.

For the synthesis process and the subsequent washing and equilibration, it is advantageous to use only an aqueous medium.

For reactions involving preliminary pore filling, the crosslinkable polymer or copolymer is preferably dissolved in a solvent or buffer that will shrink the polymer. Thus, the molecular volume of a single polymer coil or body will be minimized, allowing for the introduction of the maximum amount of polymer into the narrow pores.

In the case of polyacrylates or other acidic polymers, swelling is inhibited in the acidic pH range, resulting in a non-dissociated configuration. In the case of amino-containing polymers, an alkaline pH produces this desired molecular shrinkage. Neutral polymers, such as polyvinyl alcohol, are preferably dissolved in an aqueous mixture near the theta point, such as a water-propanol mixture.

In a preferred embodiment, in combination with any of the following embodiments, the present invention therefore relates to a well filling step, wherein the reaction solution is prepared with a non-swelling solvent, solvent mixture or buffer.

In a preferred embodiment, in combination with any of the embodiments described below, the reaction mixture comprises the functional polymer or copolymer, the crosslinking agent and optionally auxiliary substances such as buffer compounds, salts or a plurality of by-products, which originate from the original reactants applied, dissolved, suspended or emulsified together in a solvent or solvent mixture.

The invention also provides a polymeric network comprising a crosslinked preformed polymer which exhibits a pore size distribution in a particular solvent or buffer characterized in that

The upper pore size limit of the fully swollen polymeric network is determined by exclusion of a hydrodynamic radius R_h1(nm) as defined by a polymer standard molecule and at R for a hydrodynamic radius_h1To R_h2(R_h1>R_h2) With the polymer standard molecule in between, a pre-selected defined fraction of the total pore volume remains accessible.

Preferred polymer standards for use in aqueous solutions are polyethylene oxide, dextran and pullulan.

In combination with any of the above or below embodiments, the polymeric network is capable of retaining at least one permeant within the accessible pore volume.

The exclusion properties are not limited to complex designs, but are also applicable to soft gel embodiments.

R_h1Is preferably 5nm, more preferably 4 nm. R_h2Is preferably 2nm, more preferably 1nm, most preferably 0.2 nm.

Having R_h1And R_h2The preferred range of the limit is between 1.5nm and 5nm, since this range of pore sizes is essentially accessible for proteins with molecular weights between 10,000Da and 100,000 Da.

The following are preferred embodiments of the preparation of the polymeric network according to the invention:

copolymers, polycondensation products (e.g., polyamides), and oligomers or molecules having at least four identical or different repeat units are considered to be within the definition of polymer herein.

In a preferred embodiment, in combination with any of the following embodiments, the individual crosslinkable polymer or copolymer chains comprise at least one functional group ("functional polymer").

Basically, the functional polymer can be any kind of polymer comprising at least one or more of the same or different functional groups.

In fact, the functional polymer bears at least one OH-, SH-, COOH-, SO₃H or an amino group.

Preferred hydroxyl-containing functional polymers are poly (vinyl alcohol), agarose and cellulose. Preferred carboxyl-containing polymers are poly (acrylates) or poly (methacrylates).

If the fixed polymer or copolymer is functionalized, it exhibits at least one crosslinkable group per molecule.

In a preferred embodiment, in combination with any of the above or below embodiments, the functional polymer is an amino group-containing polymer, or an oligomer having at least four amino groups, more preferably a polyamine. The amino groups are primary and secondary amino groups.

Polymers derivatives containing amino groups, such as polyvinyl alcohol or polysaccharides with amino groups, are within the scope of the invention. In a further preferred embodiment, in combination with any of the above or below embodiments, the polyamine is a poly (vinylamine) or a poly (vinylformamide-co-vinylamine).

In a further preferred embodiment, in combination with any of the above or below embodiments, the functional polymer is soluble in water.

In a further preferred embodiment, in combination with any of the above or below embodiments, the reaction compound is soluble in water, or at least emulsified.

Another object of the present invention is to provide an efficient general synthesis process of composite materials, which comprises only one operating step at moderate temperature, without the need for pre-treatments, e.g. purification of any starting materials or intermediates.

Furthermore, in the most preferred embodiment, the synthesis and washing and equilibration of the composite material avoids the use of organic solvents.

One step and in situ means that all reactants are mixed, reacted, and the composite is washed in one processing operation to obtain the desired product. The immobilization is achieved mainly by crosslinking at the time of application of the complete reaction mixture or immediately thereafter.

In particular, no drying step is performed before the addition of the crosslinking agent.

In a preferred embodiment, in combination with any of the embodiments described below, the crosslinking reaction is not already initiated during the pore filling, but is subsequently initiated, preferably at an elevated temperature or a change in pH. Thus, crosslinking with epoxide crosslinkers starts at temperatures preferably above 50 ℃ without visible gelling after 30 minutes at room temperature even after 2 hours.

Crosslinking of the amino group containing polymer with a reactive crosslinking agent such as carbonyldiimidazole is inhibited at pH values below 7, preferably below 6, and starts after adjusting the pH to preferably above 7, more preferably 8, because the reaction rate of the protonated amino groups is very low. In another embodiment, in combination with any of the embodiments described below, the crosslinking agent is first applied to the pores of the carrier material, the resin is optionally at least partially dried, and finally the polymer solution is introduced and crosslinked.

The present invention also provides a method of making a composite material comprising:

at least the pore volume of the porous support material is filled with a solution (reaction mixture) of at least one functional polymer or copolymer and at least one crosslinking agent and the functional polymer is fixed by crosslinking, while the reaction mixture contains salts, buffers and/or other compounds which are not incorporated into the composite product.

In a preferred embodiment, in combination with any of the above or below embodiments, the original polymer or polymer solution and industrial quality cross-linking agent are used in the complex synthesis, still containing salts and other by-products from the manufacturing process of the respective reagents.

Surprisingly, it has been found that any by-products, mainly salts or impurities, do not interfere with the reproducible crosslinking process. After simultaneous immobilization of the reactants, they are removed from the complex by a subsequent washing step together with unreacted, e.g. excess, compound. This treatment saves the overall effort and cost of pre-concentrating or cleaning the reactive compounds. Thus, inexpensive commercial quality polymers and crosslinkers can be used in place of expensive high purity grade chemicals.

In a preferred embodiment, in combination with any of the embodiments described below, the pristine polyamine is used for complex synthesis, more preferably pristine poly (vinylamine) orPoly (vinylformamide-co-vinylamine)Solution containing salts from the polymer manufacturing process, sodium hydroxide, sodium formate and other by-products (example 1).

Preferred are support materials having an average pore size of from 10nm to 5mm, more preferred pore sizes are from 20nm to 500nm, and the most preferred range is from 20nm to 100 nm.

The form of the porous support material is not particularly limited and may be, for example, a membrane, a nonwoven fabric, a monolith, or a particulate material. Particulate and monolithic porous materials are preferred as supports. Thin film materials are also within the scope of the present invention. The shape of the particulate porous support material may be irregular or spherical. In combination with any of the above or below embodiments, the porous support material preferably has a substantially irregular shape.

In another preferred embodiment, in combination with any of the above or below embodiments, the porous support material is a monolithic or particulate material.

In another preferred embodiment, in combination with any of the above or below embodiments, the porous support material is comprised of a metal oxide, a semi-metal oxide, a ceramic material, a zeolite, or a natural or synthetic polymeric material.

In another preferred embodiment, in combination with any of the above or below embodiments, the porous support material is a porous silica gel.

In a further preferred embodiment, in combination with any of the above or below embodiments, the porous support material is porous cellulose, chitosan or agarose.

In another preferred embodiment, in combination with any of the above or below embodiments, the porous support material is a porous polyacrylate, polymethacrylate, polyetherketone, polyalkyl ether, polyarylether, polyvinylalcohol, or polystyrene.

In a further preferred embodiment, in combination with any of the above or below embodiments, the porous support material is a particulate material having an average particle size of 1 to 500 μm.

In another preferred embodiment, in combination with any of the above or below embodiments, the monolithic support material is a disk, ring, cylinder, or hollow cylinder having a height of at least 0.5mm and any diameter.

In a further preferred embodiment, in combination with any of the above or below embodiments, the support material is silica, alumina or titania, on averagePore diameter (diameter) between 20nm and 100nm (analysis by mercury porosimetry according to DIN 66133) and surface area of at least 100m²G (BET surface area according to DIN 66132).

In another preferred embodiment, in combination with any of the above or below embodiments, the support material is irregularly shaped silica, alumina or titania having a surface area of at least 150m²/g。

Even more preferred are irregularly shaped silica gel materials having an average pore diameter of 20-30nm which allow a hydrodynamic radius R when dissolved and measured in 20mM ammonium acetate at pH6 under the iesec conditions (see fig. 1, which consists of fig. embodiments 1.1, 1.2, 1.3 and 1.4) to be achieved_hEntry of polymeric pullulan standards below or equal to 6 nm.

Most preferred are irregular silicas having a BET surface area of at least 150m²Per g, preferably 250m²(ii) a pore volume (mercury intrusion) of at least 1.5ml/g, preferably 1.8 ml/g. The use of silica and silica derivatives for preparative protein purification purposes is not common and average pore sizes above 50nm are the preferred range in the prior art, typically achieved with organic support materials.

The inorganic support material can also be obtained in a dry state, enabling the introduction of the dissolved reagent by simply filling the pores, without the need for initial drying and polymer purification steps.

In a most preferred embodiment, in combination with any of the above or below embodiments, the silica gel having an average pore diameter of 25nm, preferably the pores of the support material Davisil 250, is filled with an aqueous solution containing a mixture of a diepoxide crosslinker and polyvinylamine Lupamin 50/95 (average molecular weight 50.000Da, hydrolyzed to about 95%) or Lupamin 45/70 (average molecular weight 45.000Da, hydrolyzed to about 70%), or a material having the same specifications, at a pH between 9 and 11, and reacted at a temperature of 50-60 ℃ for 24 to 48 hours.

The invention therefore relates to a process for the synthesis of a composite material comprising a porous support material and an amino group-containing polymer, characterized in that the support material is a silica gel having an average pore diameter of 20nm to 100nm and the pores are filled with a mixture of amino group-containing polymer and crosslinker and reacted in a one-step process, preferably without predrying or intermediate drying, to fix the amino group-containing polymer by crosslinking.

If the (functional) polymers are immobilized within the pores of the support material, they do not exhibit any observable network porosity in the dry state. After drying of this composite, using established methods such as BET nitrogen adsorption or mercury intrusion porosimetry, the pore size distribution of the basic support material is again found, at least as long as the degree of crosslinking remains below 25%. This behavior may be attributed to strong adhesion inside the polymer coil, causing the mesh to shrink to a value close to the excluded polymer volume. If the functional group is itself charged, for example with a polyacrylate or polyamine, the resulting excluded volume may be slightly larger.

The polymer structure shows a fundamentally different morphology if wetted by a solvent or suspension. Providing sufficient solvation, the polymeric network expands until the maximum possible volume is reached, spanning the classical hydrogel structure. In this case, the resulting porosity of the polymeric network depends on the nature (polarity, etc.) of the solvent, the pH, the ionic strength and the concentration of auxiliaries, such as detergents.

When treating functional polymers, in particular charged polymers, according to the invention, it is important to distinguish between the degree of filling of the pores of the carrier material during the immobilization of the polymer, and the "filled or occupied pores" when using the composite for separation.

In the case of first polymer attachment, the entire carrier pore volume is filled with reagent solution. In the latter case, the pores of the network are filled, i.e. a certain hydrodynamic radius R, due to the swelling behaviour of the crosslinked polymer and the resulting network in the chosen solvent_h1Is no longer accessible. In each particular caseThe potential swelling behaviour can be estimated from the available polymer literature. Thus, the degree of pore filling can be achieved, adjusted and controlled by selecting the appropriate solvent and pH.

Suitable solvents are those which, as is known to the skilled worker, are capable of swelling the polymeric network according to the rules of solvation of the polymer, see in particular H. -G.Elias, Makromolek ü le, H ü thig & Wepf, Basel, Bd.1(1990), pp.145-.

For composite materials, but also for the pressure stable polymer gels of the present invention, iSEC is a method of determining the pore volume and pore volume fraction. The solution for mercury intrusion or BET-nitrogen adsorption for rigid porous materials is not applicable here because the mesh collapses after drying.

In a preferred embodiment, in combination with any of the embodiments described below, the amino group containing polymer is introduced into the support material in a contracted state, preferably above pH 8.5, more preferably between 9 and 12, most preferably between 10 and 11, thus allowing for the maximum density of dissolved polymer under pore filling conditions. After crosslinking and swelling at a pH below 8, the space occupied by the polymeric network within the initial support pore volume will increase and eventually reach a maximum at acidic pH.

In a preferred embodiment, the object of the invention is achieved in combination with any of the following embodiments by: reacting at least one contracting crosslinkable polymer with at least one crosslinking agent to form a network that selectively swells or contracts in certain solvents or buffers.

The degree of pore filling can be adjusted to the desired level by selecting an appropriate solvent or solvent mixture under the conditions of use. By definition, if it has a selected and well-defined hydrodynamic radius R_h1Is no longer able to enter the pores of the network, the pores of the polymeric network are considered to be filled. In the present invention, this degree of swelling is calibrated and adjusted using a method of reverse size exclusion chromatography (iSEC), as described in the methods section and in FIG. 1(FIG. embodiments 1.1 to 1.4) and further control during the purification process while maintaining the corresponding swollen state by the presence of the selected buffer.

Essentially, steric exclusion of molecules with a defined minimum ("or critical") hydrodynamic radius occurs in a specific pore volume fraction, as evidenced by comparison with pullulan molecular weight standards used as model target compounds according to the figure embodiments 1.1 to 1.4 and examples.

For the complexes containing poly (vinylamine) or poly (vinylformamide-co-vinylamine) in the examples, the degree of swelling was adjusted using 20mM to 200mM ammonium acetate solution. For subsequent removal of impurities, the complex is preferably equilibrated with 50mM ammonium acetate buffer at a pH below 7, more preferably 3 to 7.

Thus, in a preferred embodiment, and in combination with any of the above or below embodiments, the extent of polymer swelling is determined by reverse size exclusion chromatography, selecting a polymer standard of defined molecular size for calibration and concomitantly adjusting the polymeric network by addition of an appropriate solvent or solvent mixture.

According to the invention, the accessible network pore volume increases in the swelling condition and decreases in the shrinking condition in the presence of a suitable solvent. The mesh pore size volume and mesh size distribution are always related to the space within or between the specifically attached polymer coils or globules, rather than to the space initially available in the support material or space ultimately remaining in the support material.

The amount of polymer introduced into the support material and fixed is preferably controlled by the polymer concentration in the respective reaction solution.

In contrast, the degree of filling of the carrier pores and the network size distribution under the application conditions are controlled by the solvent-dependent swelling of the polymer and its overall fixed amount. The two parameters are combined together, the total amount of polymer immobilized and the degree of swelling allowing the percentage of the total pore volume filled with polymer to be adjusted.

In another preferred embodiment, in combination with any of the above or below embodiments, the degree of pore filling of the support and the network size distribution under the application conditions are achieved and determined by introducing and fixing different amounts of polymer and by subsequently measuring the pore size distribution. The amount of polymer to be fixed is preferably adjusted by the polymer concentration in the reaction solution. Thus, for the purposes described, the maximum possible amount of polymer that can be fixed can be easily elucidated.

The filling degree is precisely determined and standardized by weighing the wet and dry materials before and after introduction of the polymer-crosslinker solution.

It is advantageous for the separation or purification of a dissolved target polymer (e.g. a protein) to simultaneously achieve retention of various impurities and steric exclusion of at least one target compound.

In combination with any of the above or below embodiments, the present invention provides materials and methods for using polymeric networks, preferably using composites that achieve the simultaneous removal of several structurally distinct classes of substances from a solution (preferably the starting material) while at least one target compound remains substantially unbound and is recovered in high yield. The yield of the target compound is preferably 80%, more preferably 90%, most preferably above 95%.

The above and the following further objects of the invention are also achieved according to the embodiments outlined below.

In combination with any of the above or below embodiments, the present invention provides methods for synthesizing and using polymeric networks that exhibit variable upper pore sizes, R, when equilibrated with a suitable solvent_hiThus enabling a large number of hydrodynamic radii to be kept below the exclusion limit R_hi(nm) the compound remains in the pore volume, preferably 50%, more preferably 80%, most preferably 50% of the initial content>90% and a hydrodynamic radius R_hiOr higher than R_hiAt least one target compound ofIs able to enter the pores of the polymeric network, thus allowing the recovery of the target compound in solution, preferably in purified starting materials. This exclusion limits R according to the solvent-dependent swelling of the network_hiIs of variable size. Controlling R in addition to solvent strength and pH_hiThe main parameters of (a) are the structure of the functional polymer, the nature of the crosslinking agent, the degree of crosslinking, and in the case of composites, the pore size distribution of the support material.

The term R_hiA series of different sizes obtained as a function of the degree of swelling are indicated. In contrast, R_h1And R_h2Indicating a fixed pitch for a particular application.

The stated object of combining adsorption, partitioning and size exclusion is preferably achieved by using a composite material comprising:

a porous support material having an average pore diameter of from 5nm to 5mm,

wherein the total pore volume of the porous carrier material is filled with a polymer which is cross-linked and thus forms a network which excludes the hydrodynamic radius R when equilibrated with a suitable solvent_hi(nm) standard molecule and thus a hydrodynamic radius R_hiOr higher than R_hi(nm) synthetic and natural macromolecules provide exclusion limits.

If the target compound is an antibody, if the hydrodynamic radius R_h1Molecules above 5nm, preferably above 4nm, cannot enter the network, and the exclusion effect is achieved.

In further combination with any of the above or below embodiments, the present invention provides for the synthesis and use of soft gels or composites having a defined network pore volume capable of retaining a substantial amount of the hydrodynamic radius R_hCompounds below 4nm, preferably 50%, more preferably 80%, most preferably of the initial content>90% and antibody cannot enter this part of the pore volume.

In the context of the present invention, pore volume refers to the integration or sum of the parts of the entire specified pore volume, wherein each part is defined by a lower pore diameter and an upper pore diameter.

In further combination with any of the above or below embodiments, the present invention provides for the synthesis and use of soft gels or composites having a defined pore size distribution capable of retaining a large number of hydrodynamic radii R_h2The compounds below 4nm remain in their network pore volume, preferably 50%, more preferably 80%, most preferably of the initial content>90%, and R_h1Target compounds (e.g. antibodies) at or above 4nm cannot enter this part of the pore volume, while another part of the unwanted products with higher molecular weight bind to the outer surface.

The said other undesired product is preferably a nucleic acid and/or host cell protein with a molecular weight of more than 100,000 Da.

The above-mentioned object of protein purification is preferably achieved by using a composite material comprising: a porous support material having an average pore diameter of 20nm to 5mm, wherein the total pore volume of the porous support material is filled with a crosslinked amino group-containing polymer under application conditions of a pH below 8,

the composite is characterized by a pore size distribution wherein the hydrodynamic radius R_hIs 4nm and above 4nm, especially 21.7kDa and R_hCalibrated pullulan standards of 3.98nm are excluded from at least 90% of the pore volume, and

wherein at least 35% of the total pore volume is defined by the radius enabling hydrodynamic radius R_hWells entered at 6.2kD as standard pullulan 2.13nm are indicated.

Optionally, at least 15% of the total pore volume of the above composite material may be represented by pores capable of admitting a pullulan standard 10.0kD having a hydrodynamic radius of 2.7nm

In combination with any of the above or below embodiments, the above amino group-containing polymer preferably comprises poly (vinylamine) or poly (vinylformamide-co-vinylamine).

These increased pore volume fractions of 35% and 15%, respectively, are advantageous in order to provide significant binding capacity for proteins having, for example, a molecular weight between about 10,000Da and 100,000Da, which essentially represents the molecular hydrodynamic radius R in the selected solvent or buffer_hBetween about 2nm and 4 nm.

In combination with any of the above or below embodiments, the present invention provides the use of a suspension polymer network comprising poly (vinylamine) or poly (vinylformamide-co-vinylamine) and a cross-linking agent, characterized in that the hydrodynamic radius is higher than R_hPullulan standards, 3.98nm, were substantially excluded from the pore volume, thus defining the upper pore diameter in the corresponding solvent, while the network volume allowed for a hydrodynamic radius R_h22.13nm pullulan standard 6.2kD entry.

By substantially is meant that at least 90% of the pore volume is inaccessible.

The polymeric network comprising the amino polymer is preferably obtained by crosslinking the corresponding polymer in aqueous solution, while the pH is between 8 and 13, preferably between 9 and 12, most preferably between 10 and 11.

Well accessibility and exclusion limits were always determined by iSEC using pullulan standards in 20mM ammonium acetate buffer at pH6 (see FIG. 1 and methods).

The feedstock solution comprises a mixture of synthetic or natural origin. Preferably, the starting material is a fermentation broth, which is filtered (cell culture supernatant) or crude, which still contains solids such as cells and cell debris.

In combination with any of the above or below embodiments, the average molecular weight of the functional polymer is preferably from 2,000 to 2,000,000 daltons, more preferably from 10,000 to 1,000,000 daltons, even more preferably from 15,000 to 200,000 daltons, most preferably from 20,000 to 100,000 daltons.

In combination with any of the above or below described embodiments, any cross-linking agent known in the art is suitable for immobilizing the polymer according to the present invention.

In combination with any of the above or below embodiments, the crosslinking agent is preferably a bis-oxirane or a bis-aldehyde, such as succinic acid or glutaraldehyde, provided that the polymer contains amino groups. If a dialdehyde is used as crosslinking agent, a subsequent reduction step is advantageous for stabilization purposes.

Crosslinkers having more than two reactive groups are also suitable.

Preferably, the cross-linking agent should represent a chemical activating agent in the formation of the polymeric network.

Alternatively, the polymer may be introduced as a chemical activation partner using reagents and methods known in the art, particularly in peptide synthesis.

The polymer may also be inherently reactive. In this case, the functional groups of the polymer may be generated during the crosslinking process itself or subsequently, using reactive or activated polymers, such as anhydrides from poly (maleic acid) or polyethylene oxide.

Both the crosslinking agent or the polymer may also be activated using prior art carbodiimide reagents, preferably water soluble carbodiimides, so that the entire reaction is carried out under non-aqueous conditions.

In combination with any of the above or below embodiments, the acrylic polymer is thus crosslinked with a diamine, diol, or disulfide. Alternatively, activated dicarboxylic acids are used to crosslink amino or hydroxyl or thiol containing polymers.

In combination with any of the above or below embodiments, the degree of crosslinking is preferably from 5% to 30%, more preferably from 7% to 20%, most preferably from 10% to 15%.

The synthesis can be carried out using any solvent which does not react or only slowly reacts with the crosslinking agent and the crosslinkable polymer under the preparation conditions and preferably dissolves the reactants to at least a 1% (w/v) solution. In this context, slowly means that no visible gelation occurs at the selected temperature until at least 30 minutes using only the polymer crosslinker solution shown in comparative example 1.

In combination with any of the embodiments outlined above or below, the amount of cross-linked polymer immobilised within the pores of the support material is preferably at least 1% w/w (weight of polymer and cross-linking agent/weight of dried composite), more preferably between 5% and 10% w/w, preferably less than 25% w/w.

The temperature range of the synthesis process is preferably between 20 ℃ and 180 ℃, more preferably between 40 ℃ and 100 ℃, most preferably between 50 ℃ and 70 ℃.

The relevant reaction time is preferably between 1 hour and 100 hours, more preferably between 8 hours and 60 hours, most preferably between 18 hours and 48 hours.

Abbreviations and Definitions

The partial volume (. mu.l) is necessary to obtain porosity data of the adsorbent by injection with a defined hydrodynamic radius R_hThe molecular standards of (a) are measured using a packed chromatography column. These volumes are determined by multiplying the signal time by the flow rate.

V_e

When the additional column volume of the chromatography system is subtracted from the total elution volume, a net elution volume V is obtained_e。V_eTotal void volume V with column_oThe same is true. V_enIs the elution volume of a single standard n.

V_o

The total void volume of the column is the pore volume V_pAnd the gap volume V_iThe sum of (a) and (b).

V_i

Volume of gap V_iIs the volume between the particles.

V_p

Pore volume V of the adsorbent_pIncluding the total porous space.

hIgG polyclonal human immunoglobulin G.

S layer protein, cell surface layer

Material

Cell Culture Supernatant (CCS)

CH0-K1, Invivo, Berlin, batch RP SZ 292/01

Simulated culture medium of cell line CHO-K1

Conductivity: 15mS/cm, pH 7.0

The HCP content was 120. mu.g/ml and the DNA content was 1000 ng/ml.

Removal experiments for tables 1.1 and 1.2

The mock medium of CCS BHK-21, Invivo, Berlin, batch RP SZ 352/01 cell line BHK-21 contained 5% BSA for Bovine Serum Albumin (BSA) removal as shown in FIG. 3.

Raw materials

A specific CCS incorporates 2mg/ml hIgG from human plasma (Octagam, 10% solution, Octapharma, Vienna).

Octagam contained 2.5% to 3.5% aggregates according to SEC analysis.

Carrier material

Silica Gel Davisil LC250 (w.r.grace), mean nominal pore sizeThe particle size was 40-63 μm (batch No. 1000241810).

Eurosil Bioselect 300-5，5μm,Knauer WissenschaftlicheBerlin, germany.

Polymer and method of making same

Aqueous poly (vinylformamide-co-polyvinylamine) solution, supplied from Lupamin 45-70 (BASF): BTC Europe, Mongham, Germany, was partially hydrolyzed by heating 1000g of Lupamin 45-70 with 260g of sodium hydroxide (10% w/v) at 80 ℃ for 5 hours for the embodiment of example 1. Finally the pH was adjusted to 9.5 with 170g of 10% hydrochloric acid. For example 1a, untreated Lupamin 45-70 solution was used without sodium hydroxide hydrolysis and hydrochloric acid pH adjustment.

Crosslinking agent

Hexanediol diglycidyl ether, Ipoxx RD 18, Ipox chemicals, Laupheim (Germany) -batch No. 16092)

Method and measurement

Size Exclusion Chromatography (SEC)

The concentration of hIgG in the starting material and the recovery of hIgG in the purified solution were determined by SEC under the following conditions:

column: tosoh TSK G3000 SWXL 7.8 mm (particle size 5 μm)

Mobile phase: 10mM sodium phosphate buffer, pH7.2+150mM NaCl

Injection amount: 100 μ L-sample diluted with mobile phase.

Flow rate: 1 mL/min.

A detector: DAD 280nm, hIgG solution (Octapharma) was used as external standard.

Temperature: 20 ℃ C. +/-1 ℃ C

And (4) measuring dry mass.

The complex was washed five times with 5 bed volumes of water each time, then five times with five bed volumes of methanol each time. After drying a 0.5g sample at 80 ℃ for 12 hours under reduced pressure (0.1mbar), the dry mass of the support material and of the composite is determined and then measured 2-3 times over 2 hours until the weight is constant.

Pore size distribution and pore volume fraction of multiple composite adsorbents

The accessible pore volume fraction related to the pore diameter and the exclusion limit of polymer molecules with various hydrodynamic radii was determined using reverse size exclusion chromatography (iesc). For this purpose, the composite material is loaded into a 1ml (50X 5mM) column, equilibrated with 20mM ammonium acetate buffer (pH6), and purified by applying two low molecular weight standards and selecting six of the known defined average molecular weights M_wCalibration of commercially available pullulan polymer standards (PPS, MainzGermany, see figure embodiments 1.1-1.4 for details).

M on pullulan standards was achieved by SEC in PSS with water, sodium azide 0.005% as mobile phase at 30 ℃ at a flow rate of 1ml/min_wAnd (4) measuring. Three analytical columns, each 8X 300mm (PSS SUPREMA 10 μm)) It has been used in combination with an 8X 50mm pre-column (PSS SUPREMA 10 μm) in series. The sample concentration was 1g/l and the volume per injection was 20. mu.l. Detection was achieved using a Refractive Index (RI) monitor (Agilent RI D) connected to the PSS WinGPC data acquisition system.

By evaluating the net elution volume V_en(mul) obtaining the fraction of pore volume K accessible to a particular standard in a particular composite_av。

Thus, K_avHaving a given hydrodynamic radius R is described_hFraction of total pore volume that a particular standard can enter. Determination of the Total liquid volume V Using methanol_t＝V_e＝V₀Is represented by K_avThe value is 1. Determination of filling Using 210,000Da pullulan standardsInterstitial volume V between composite particles_iDenotes the volume of liquid outside the particle, since it has been excluded from the pores (see FIG. 1), and therefore represents K_avIs 0 (0% of pore volume). V_oAnd V_iThe difference between is the pore volume V_p。

Partial pore volumes are defined as the respective volume fractions in the composite adsorbent, which can be taken in by the unretained pullulan polymer standards as well as by the unretained smaller molecules. By not retained is meant that no interaction or binding of the respective standard occurs on the surface of the stationary phase in order to determine only the pore volume fraction. This is the case for the carrier materials and composites of the invention, for alcohols and hydrophilic carbohydrates, preferably pullulan, which exhibit a known hydrodynamic radius (R) in aqueous solvent systems_h)。

According to empirical equation R_h＝0.027Mw^0.5(I.Tatarova et al, J.Chromatogr.A 1193(2008), page 130) from molecular weight M_wCalculating R of pullulan_hThe value is obtained.

R of IgG_hValues were taken from the literature (k. ahrer et al, j. chromanogr.a 1009(2003), page 95, fig. 4).

Distribution of isoelectric points. pI values of host cell proteins in CCS CHO K1

Determined by isoelectric focusing (IEF) (see figure 2 for details), calibrated with standard proteins of known pI, as shown in the table (example 4).

DNA assay

After DNA extraction using the DNA extraction kit, Cygnus Technologies, Southport (USA), DNA quantification was done using the Quant-iT PicoGreen dsDNA kit, Life Technologies, Darmstadt (Germany).

Host Cell Protein (HCP) assay

HCP quantification was performed using the Cygnus HCP ELISA kit, CHO host cell protein third generation (F550), from Cygnus Technologies, southport (usa).

Examples

Example 1

Preparation of composite adsorbent batch 07 (Table 1)

Mu.l (658mg) of hexanediol diglycidyl ether (Mw 230.2, d 1.07g/ml) crosslinker were dissolved in 42ml of water. This crosslinker solution was added to 15ml of an aqueous solution of poly (vinylformamide-co-polyvinylamine) (Lupamin 45-70, partially hydrolyzed, see materials). After mixing, the pH was adjusted to 11 with 3ml of 0.5M NaOH.

10g of Silica Gel Davisil LC250, 40-63 μm (W.R.Grace), dry powder, was settled into a flat bottom stainless steel plate of 8cm diameter. The height of the bed (bed) is 8 mm. 39.5g of the polymer-crosslinker solution was added and distributed evenly over the silica, while the solution was quickly soaked in the pores. The resulting paste was shaken on a rotary shaker at 600rpm for 1 minute to obtain a homogeneous mass with a smooth surface, covered with a 1-3mm liquid film. After covering the petri dish with a stainless steel lid, the paste was heated in a drying oven at 60 ℃ for 48 hours without further mixing or movement, yielding 49.6g of a wet composite.

Subsequently, 41.3g of this still moist paste were washed 5 times with 25ml of water on a glass frit (frit). The composite filter cake was then suspended in 31.6ml of 10% sulfuric acid and treated with smooth shaking at ambient temperature over 2 hours to hydrolyze unreacted epoxy groups. Finally, the product was washed 5 times with 25ml of water on a glass frit and then stored in 20% ethanol/water.

Any other lot of table 1 was prepared in this way, only the amount (volume) of the crosslinking agent was changed according to the target degree of crosslinking.

Reference example 1

(preparation of crosslinked polyvinylamine gel)

To examine the reaction without support material, 3ml of the polymer-crosslinker solution from example 1 were heated at 50 ℃ for 24 hours. After six hours, gelation was observed. After 24 hours, a transparent solid elastic gel was obtained.

Example 1a

The composite adsorbent is prepared using a small particle support material.

1ml (935mg) hexanediol diglycidyl ether (Mw 230.2, d 1.07g/ml) crosslinker was shaken with 59ml water to form a homogeneous emulsion. This crosslinker solution was added to 21ml of an aqueous solution of poly (vinylformamide-co-polyvinylamine) (Lupamin 45-70, virgin and untreated).

After mixing, the pH was adjusted to 10 with 0.5M NaOH.

25g of Silica Eurosil Bioselect 300-5, 5 μm, dry powder, was settled into a flat bottom stainless steel pan 12cm in diameter. The bed height is about 15 mm. 46g of the polymer-crosslinker solution was added and distributed evenly over the silica, and the solution was soaked in the pores to form a viscous mass. After addition of 1.5ml portions of the polymer-crosslinker solution and finally 4ml of the diluted polymer (1 ml of poly (vinylformamide-co-polyvinylamine) diluted with 3ml of water), the suspension became smooth and homogeneous. The resulting paste was covered with a film of liquid about 1mm high. After covering the petri dish with a stainless steel lid, the batch was heated in a drying oven at 65 ℃ for 21 hours without further mixing or movement, yielding 72g of a wet composite.

Subsequently, the paste was diluted with distilled water to a volume of 150ml and the resulting suspension was pumped into a 250 x 20mm HPLC column using a preparative HPLC pump. The packed composite bed was then washed with 250ml of water. To hydrolyze unreacted epoxy groups, 100ml of 2n hydrochloric acid was pumped into the column and allowed to dwell at ambient temperature for more than 2 hours. As the back pressure increased in this step and subsequently rinsed with water, the filled composite was finally washed with 300ml of ethanol, while the pressure was reduced to 5 bar at a flow rate of 10 ml/min. The product was removed from the column and dried at ambient temperature. The nitrogen content was determined to be 1.18% and the carbon content was determined to be 2.99%.

Reference/comparative example 2

Preparation of composite batch 19 according to WO 2013/037994 (prior art)

10g of Silica GelDavisil LC250, 40-63 μm pores have been completely impregnated with the poly (vinylformamide-co-polyvinylamine) of example 1. The intermediate product of step 1 was dried at 50 ℃ until the weight was constant. The dried adsorbent was then suspended in isopropanol containing ethylene glycol diglycidyl ether (121mg in 30ml isopropanol) and stirred at 55 ℃ for 5 hours.

The product of step 2 was then filtered and washed with isopropanol, 0.5M trifluoroacetic acid, water and methanol.

As shown in figure embodiments 1.3 and 1.4, the pore volume and pore size distribution differed from the performance obtained with composite batch 07. See table 2 for details.

Table 2: pore volume distribution according to fig. embodiments 1.2 and 1.4, measured with pullulan standards; support material Davisil LC250 and composite batches 07 and 19

R_h2nm and R_hVolume fraction between 4nm as molecular weight below 100,000DaThe capture space for the protein is important. Therefore, it is advantageous to generate as much volume as possible in this range in order to provide a high binding capacity for host cell proteins. Batch 07 exhibited about 44% of the total volume within the range, while the product of the two-step synthesis batch 19 exhibited only about 19% of the pore volume fraction. In addition, batch 19 has an R between 2.7nm and 4nm_aIn the range only 4% of the pore volume fraction.

Basically, the data for batches 07 and 19 show that different synthesis regimes result in different pore size distributions and hence different morphologies of the resulting composites.

Example 2

Analysis of the separation capacity of the composite material prepared according to example 1 listed in table 1.1 and the comparative commercial material listed in table 1.2.

In order to measure the separation ability of the composite material, the degree of removal (separation) of impurities or unwanted compounds from the target substance is determined. For this purpose, selective assays are used to determine the concentration of individual components or substance classes in the starting material. After the separation step, the concentration measurement was repeated with the purified fraction. Thus, purity and recovery can be calculated from these concentrations and related volumes.

For Host Cell Protein (HCP) removal, the purification efficiency was determined by comparing the original feedstock solution to the removed supernatant fraction after a specific time of contact with the new composite using the Cygnus CHO HCP third generation Elisa assay.

dsDNA was determined using the Quant-iT PicoGreen dsDNA kit. hIgG recovery was determined by quantitative SEC.

General removal procedure

The starting material was untreated and undiluted cell culture supernatant CHO-K1, spiked with 2mg/ml hIgG (polyclonal antibody Octagam, see materials for details). 400mg of wet, equilibrated adsorbent was incubated with 2ml of starting material using Falcon tubes or centrifuge tubes. After gentle shaking for 5 minutes, the supernatant was separated by centrifugation for subsequent analysis.

Recovery of hIgG was determined by quantitative SEC (for equipment, see methods and measurements). The main peak (97-98%) in the chromatogram was associated with the monomer, and the early eluting peak (2-3%) was associated with immunoglobulin aggregates already present in the original higg (octagram) preparation. Calibration and recovery the reference main peak was determined.

Since the adsorbent is in a wet state prior to contact with the feedstock, the associated void volume increases the total liquid volume. The concentration of the substance will decrease. The void volume is typically 70% to 90% by weight of the resin. The final substance concentration is thus corrected after the removal step by the corresponding dilution factor.

During this purification, the amount of aggregates was found to be constant. In contrast, aggregate formation is typically observed using prior art adsorbents and separation schemes.

Macromolecular DNA is much larger than hIgG and is therefore usually excluded from the complex pore system. hIgG was also excluded as shown in example 3.

Results

The results of the composite adsorbents obtained after pore-filling the support material Silica Gel Davisil LC 25040-63 μ with hydrolyzed poly (vinylformamide-co-polyvinylamine) Lupamin 45-70(BASF) and various amounts of the crosslinking agent hexanediol diglycidyl ether are shown in table 1.

The volume of the crosslinking agent is varied to achieve a specified degree of crosslinking. The volume of 704. mu.l used to prepare batch 07 of the composite represents a degree of crosslinking of 10%. All other conditions were as described in example 1.

The ratio of the volume of the raw material to the mass of the adsorbent is 5: 1(2ml of starting material: 0.4g of adsorbent).

All adsorbents were equilibrated with 50mM ammonium acetate buffer (pH6) prior to contact with the feed.

Table 1.1: impurity removal as a function of crosslinking

DNA analysis was performed after DNA extraction using the Quant-iT PicoGreen assay (see methods).

HCP analysis was performed using Cygnus HCP ELISA, CHO 3 rd generation (F550), see methods

Comparative example 3

Analysis of the separating ability of commercially available amino group-containing adsorbents

The separation capacity of a commercial amino group-containing adsorbent, which exhibits positively charged groups under the specified test conditions, was measured using the same removal procedure as described in example 2.

Table 1.2 shows the commercial Amino-containing anion exchange adsorbent, Toyopearl AF Amino 650M, Tosohbioscience, Griesheim (Germany); toyopearl DEAE 650M, Tosoh; and Q Sepharose FF, GEHealthcare, Little Chalfot (UK).

The ratio of the volume of the raw material to the mass of the adsorbent is 5: 1(2ml of starting material: 0.4g of adsorbent). All adsorbents were equilibrated with 50mM ammonium acetate buffer (pH 6.5) prior to contact with the feed.

Comparative examples	Carrier material	HCP removal (%)	DNA removal (%)
				Toyopearl AF Amino 650M	Toyopearl HW 65	47.9	92.0
Toyopearl DEAE 650M	Toyopearl HW 65	53.4	93.4
				Q Sepharose FF	Sepharose	81.8	95.6

Table 1.2: removal of impurities from feedstocks using commercially available amino-containing adsorbents

DNA analysis was performed after DNA extraction using the Quant-iT PicoGreen assay (see methods). HCP analysis was performed using the cygnus HCP ELISA, CHO third generation (F550), see methods.

Results

As can be seen from table 1.1, between 80% and 96% recovery of hIgG was achieved, as well as the ability of the composite adsorbent of the invention to remove both HCP and DNA to a high degree (over 92%) simultaneously.

It has been shown by reverse-phase size exclusion chromatography (iesc),>90% of the hydrodynamic radius R_h>The 4nm pullulan polymer standards were excluded by the pores of the composite (see figure embodiment 1.2). IgG is characterized by R_hFrom 4.5 to 5nm and are therefore also excluded from these wells (see example 3 for details).

In addition, the dynamic binding energy of DNA (sodium salt, from calf thymus, type 1, fiber, Sigma) has been shownThe force was about 1.2mg/ml of complex. The subsequently injected DNA fraction is eluted from the 1ml column in a gap volume V of about 0.5ml_i(volume between packed particles) while smaller polymer standards and methanol can still enter the internal pore volume. This indicates that the macromolecular DNA binds only to the outer surface of the composite adsorbent. Macromolecular DNA is much larger than hIgG and is therefore usually excluded from the complex pore system. hIgG was also excluded as shown in example 3.

Example 3

Determination of the binding Capacity of hIgG

The purpose of this experiment was to show that hIgG was excluded from the adsorbent pores, while only very small amounts of hIgG bound to the exterior of the adsorbent.

About 1ml of adsorbent batch 07 (example 1 and table 1.1) was packed into a 1ml column (50 x 5mM) and equilibrated with 20mM ammonium acetate buffer (pH 6).

An hIgG test solution (c ═ 2mg/ml) was prepared from a 10% octagram stock solution by dilution with 50mM ammonium acetate buffer (pH 6.5) (buffer a).

The flow rate during the loading step was 0.2mL/min and during the wash was 1 mL/min.

The optical density in the eluate was monitored at 280 nm.

The total elution volume consists of two contributions of the volume in the column, i.e.the liquid volume V of the packed column_t＝V_ePlus additional column volume of the chromatography system.

Determination of the total liquid volume V of the packed column by running a methanol sample_t972. mu.l, the interstitial volume V was determined using pullulan 48.800Da_i523 μ l. Pore volume V of the adsorbent_pIs a difference and is therefore 449. mu.l.

hIgG elution signal started in a volume of 630. mu.l (FIG. 4.1). 44 μ l was subtracted from this total elution volume (representing an additional chromatography system)Column volume). Thus, the net retention volume of hIgG was 586. mu.l. This volume is significantly less than the column liquid volume V of 972. mu.l_tIndicating that hIgG cannot enter the pore system.

At the moment of penetration (breakthrough), an interstitial volume V of 523. mu.l_iFilled with unbound hIgG. Thus, the amount of hIgG that had been initially contained in the delta volume of 586 μ Ι -523 μ Ι -63 μ Ι bound to composite batch 07 upon penetration. Thus, 126. mu.g bound to 1ml of the complex. This small adsorption capacity may be associated with pI<6, which binds to the low-area outer surface of the particle under conditions of actual salt concentration.

Additional injections of material were performed after saturating the complex with hIgG.

After saturating the column packed with composite batch 07 with hIgG, 5 equal injections of 50. mu.g of hIgG solution were performed under the same buffer conditions, but at a flow rate of 1ml/min (FIG. 4.2).

All chromatograms were identical. Penetration had occurred after 0.45 minutes with peak maxima of 0.572 minutes or 572 μ l. Subtract 44. mu.l of additional column volume to give a 528. mu.l volume, and the interstitial void volume V in the column_iA good match.

This demonstrates that hIgG does not otherwise bind to the complex after saturation of the complex with hIgG, and demonstrates that hIgG is blocked outside the pore system.

Thus, a small amount of adsorbed hIgG binds only to the outer surface of the particle, whereas the hydrodynamic radius R_hThe major part of hIgG between 4.5 and 5.5nm remains unbound and cannot enter the pores of the complex, which are below R of 4nm in size_h(embodiment of fig. 1).

From FIG. 1, it can be concluded that R_aMolecules larger than 4nm are typically excluded from at least 90% of the pores of the complex.

Example 4

CCS CHO-K1 and removal of basic (positively charged) host cell proteins were characterized by isoelectric focusing.

Isoelectric focusing (IEF) was used to determine the charge heterogeneity of proteins in CCS and to demonstrate the presence of basic proteins with pI higher than 7.5.

The pH range of the IEF strip was calculated using a commercial kit containing proteins of known isoelectric point. The stained gel of the mock CCS studied under the same conditions showed a broad band distribution between pI 3 and pI 10 and a fraction of the proteins had a high pI higher than 8. Thus, a substantial amount of Host Cell Proteins (HCPs) in CCS are basic (fig. 2).

According to the quantitative HCP ELISA, hIgG purified from hsgg-supplemented CCS showed HCP removal of e.g. 98.7% (table 1.1, batch 07), demonstrating that the major part of the alkaline host cell proteins present in the crude CCS were thus removed. A single IEF run was calibrated using nine protein standards of known pI (Biorad), as listed below.

Component proteins of IEF standards used to calibrate pI values:

conditions for isoelectric focusing (IEF) K1 (gel: 4-15%TGX^TMGel (Biorad), 7cm IPG/preparation well, 250. mu.l, strip: linear gradient pH 3 to pH 10 for 7cm IPG band, staining method: coomassie brilliant blue R-250 from CCS CHO) showed a broad band distribution between pI 4 and pI 10 (see fig. 2) and a substantial part of the protein was higher than pI 7.5.

And (4) conclusion: unexpectedly, the amino group-containing complexes bind not only to anionic species (pI 2-6), but also to cations (pI 8-11) and neutral compounds, most importantly to groups from HCPs. These results were obtained using positively charged (as it is usually) immobilized polyvinylamine when complex equilibration was performed using 50mM ammonium acetate buffer at pH < 7.5. If equilibrated with 50mM ammonium acetate buffer at pH 6.5, the complex removed up to 99.8% HCP in CCS CHO K1 (Table 1.1, batch 06). Thus, it was surprisingly found that the removal of HCPs and other impurities is largely independent of their isoelectric point.

In contrast to the conventional strong anion exchanger Q Sepharose (see table 1.2) which removes only 81.8% of HCP under the same conditions, the poly (vinylamine) -containing silica adsorbent material of the present invention removed up to 99.8% of the total HCP inventory in the presence of 2mg/ml polyclonal hIgG, with up to 96% recovery of polyclonal hIgG.

Description of example 5 and FIG. 3

Bovine Serum Albumin (BSA) was removed from cell culture supernatant CCS BHK-21 (see materials) after treatment with two complex batches 07 and 08 at two pH conditions, with or without salt, according to example 2.

FIG. 3 shows a starting material consisting of CCS Invivo BHK-21 (lane 9) containing approximately 50mg/mL BSA as a nutrient and spiked with hIgG (2mg/mL), which was removed to detectable levels in a batch step (lanes 3-8). Removal was performed in six independent experiments according to example 2 with compound batches 07 and 08 of table 1.1.

In addition to the usual equilibration with 50mM ammonium acetate pH 6.5 (lanes 5 and 8), the complexes were also equilibrated with 50mM ammonium acetate pH 6.5 containing 150mM NaCl (lanes 4 and 7). In addition, both complexes were also equilibrated with 50mM ammonium acetate, 150mM NaCl, pH 7.4 (lanes 3 and 6).

Thus, it was demonstrated that quantitative removal of BSA also works in the presence of salt and at higher pH.

Purified samples were applied to SDS PAGE in specific lanes 3-8 and starting material was applied to lane 9

Marker Protein (Bio Rad, Precision Plus Protein)

2, blank space

3, purification with batch 08 (complex equilibrated with 50mM ammonium acetate in water pH 7.4, 150mM sodium chloride)

4, purification with batch 08 (complex equilibrated with 50mM ammonium acetate in water pH 6.5, 150mM sodium chloride)

Batch 08 (complex equilibrated with 50mM ammonium acetate in water pH 6.5 (no salt)) purification

Batch 07 (complex equilibrated with 50mM ammonium acetate in water pH 7.4, 150mM sodium chloride) purification

Batch 07 (complex equilibrated with 50mM ammonium acetate in water pH 6.5, 150mM sodium chloride) purification

Purification with batch 07 (complex equilibrated with 50mM ammonium acetate in water pH 6.5 (no salt)) 8

9 CCS Invivo BHK-21 (batch RP _ SZ _352/01) + hIgG 2mg/mL, containing 5% (w/v) Bovine Serum Albumin (BSA)

FIG. examples 1.1

Composite adsorbent batch 07. With known different hydrodynamic radii (R)_hi) The net elution volumes V of methanol, ethylene glycol and six pullulan standards_e(. mu.l) with R_hiThe figure (a).

By iSEC (Chart V)_e) The pore volume V of the adsorbent was determined using a packed column of 1ml (50X 5mm) of nominal resin volume_pAnd interstitial volume V between particles_i. The total liquid volume V has been measured in a column filled with the carrier material Davisil LC250 and various composite materials_t＝V_e(V_eIs the determined net elution volume, the additional column volume has been subtracted) is between 965 μ l and 998 μ l, the smallest standard methanol is fully accessible. The interstitial volume V between the particles has been determined_iBetween 450 and 530. mu.l. The deviation in specific volume fractions is due to small differences in the amount of packing material and packing density in the individual columns. R_h>The 9nm standard did not enter the pores of the silica Davisil LC250 and passed through a gap volume V of 449. mu.l_iThen eluted only after migration in the same volume. For example, the total pore volume V of, for example, Davisil LC250 silica in the column of embodiment 1.1_pThe difference was 998. mu.l-449. mu.l-549. mu.l. Calibrated pullulan standards according to their specific hydrodynamic radius R_hFractional volume penetration. The volume ratio of the various composite materials was measured in the same manner.

FIG. embodiment 1.2

Composite adsorbent batch 07. Distribution coefficient (K)_avThe value, i.e. pore volume distribution fraction, see methods; k_avCorresponding to the pore volume fraction available for a single species) the hydrodynamic radius R of the same test species as in figure embodiment 1.1_hiThe figure (a).

Distribution coefficient K_avPore volume fraction V, defined as the available pore volume fraction of a particular molecular standard n above a certain pore diameter_enI.e. K_av＝V_en–V_i/V_e–V_i. The upper iSEC curve (silica 250) shows the pore size distribution of the support material Davisil LC250 with an exclusion limit of R_hR at 9nm and 4nm_hLower, accessible pore volume fraction K_av0.36 (36% of the total pore volume between the hydrodynamic radii of 4nm to 9nm of the polymer standard). This means that R_hMolecules at 4nm can enter 36% of the pore volume.

The three lower curves show the porosity of embodiment batch 07 (table 1.1) obtained by repeated runs. After fixing the polymer, only<5％(K_av0.05) exhibits a value of 4nm or more。

This is that the hole is covered under the condition of useFilling/filling or occupyingWith respect to the accessibility of molecules of a specific diameter:

however, in the starting material Davisil LC250, more than 36% of the pores were found in the range between 4 and 9nm, and after fixation of the crosslinked polymeric network, more than 30% of the corresponding pore volume was absent in product batch 07. This is apparently due to the space taken up by the polymer network and the mere distribution of this volume.

In other words: the > 30% of the pore volume of Davisil LC250 (initially 36% of the total pore volume) between 4nm and 9nm disappeared, since pores of this size were occupied by the polymeric network, showing significantly smaller pores. All smaller support pores also contain the polymeric network. Thus, the porosity of the composite is determined by the internal pores of the polymeric network (e.g., microsponge) in the swollen state at pH 6.

However, the low molecular weight standard methanol enters the entire pore volume of the support material and the entire pore volume of the composite. Thus, the slope of the composite porosity curve is significantly steeper than the slope of the Davisil LC250 curve.

K of the composite if only the wall of Davisil LC250 is coated_avThe curve will be parallel to the Davisil LC250 curve, at least at an R of 4nm to 9nm_hInsofar as there will always be a gap left in the centre of each hole.

The steric exclusion effect described is responsible for the purification ability of the polymeric network within the complex, e.g. in the outer volume of the new composite material filled or suspended, the antibodies remain largely unbound, while the lower molecular weight components, e.g. host cell proteins, enter the pores of the immobilized polymer where they may be captured.

Of utmost importance, for R_hBetween 2nm and 4nm (including the hydrodynamic radius of most proteins with molecular weights below 100.000), about 40% of the complex pores are accessibleIn (1). Within this section of the well, most of the host cell proteins are captured, including the pI>7.0, although the polymer is also positively charged under normal working conditions (see example 3 and FIG. 2).

FIG. embodiment 1.3

Composite adsorbent batch 19, comparative example and composite adsorbent batch 07, examples of the present invention. Methanol, ethylene glycol and a mixture of known different hydrodynamic radii (R)_hi) The net elution volume V of six pullulan standards_e(. mu.l) with R_hiThe figure (a).

FIG. embodiment 1.4

Composite adsorbent batch 19, comparative example and composite adsorbent batch 07, examples of the present invention. Distribution coefficient (K)_avThe value, i.e. pore volume distribution fraction, see methods; k_avCorresponding to the pore volume fraction available for a single species) the hydrodynamic radius R of the same test species as in figure embodiment 1.1_hiThe figure (a).

In contrast to composite batch 07, batch 19 of the two-step synthesis (including preliminary drying after the first step) showed very different porosities in the nanometer range. Only 4% of the pore volume was available between hydrodynamic radii of 2.7nm and 4nm, whereas batch 07 provided a volume fraction of 22% in the same range. See table 2 for details.

FIG. 2

According to example 4 isoelectric focusing (ief) of concentrated solutions of Cell Culture Supernatants (CCS) from the CHO K1 cell line was typically used for visualization of pI spectra of Host Cell Proteins (HCPs) in such samples. A large number of neutral and alkaline host cell proteins were found, which were tightly focused in the pI range of 7 to 10.

Isoelectric focusing (IEF) was used to determine the charge heterogeneity of proteins in CCS and to demonstrate the presence of basic proteins with pI higher than 7.5.

The pH range of the IEF strip was calculated using a commercial kit containing proteins of known isoelectric point. The stained gel of the mock CCS studied under the same conditions showed a broad band distribution between pI 3 and pI 10 and a fraction of the proteins had a high pI higher than 8. Thus, a substantial amount of Host Cell Proteins (HCPs) in CCS are basic.

FIG. 3

Bovine Serum Albumin (BSA) was removed from the cell culture supernatant CCS BHK-21 after treatment with two complex batches 07 and 09 at two pH conditions, with or without salt, according to example 5.

FIG. 4

According to example 3, hIgG is excluded from the adsorbent pores and only a small fraction binds to the outer surface of the adsorbent.

For the description of FIG. 4, see also example 3

The purpose of this experiment was to show that hIgG was excluded from the adsorbent pores, while only very small amounts of hIgG bound to the exterior of the adsorbent.

Under the conditions of dynamic Capacity measurement, at the moment of penetration (FIG. 4.1), a gap column volume V of 523. mu.l_iFilled with unbound hIgG. Thus, the amount of hIgG initially contained the difference between the elution volume and the void (586 μ Ι -523 μ Ι 63 μ Ι) bound to composite batch 07 upon penetration. Thus, 126. mu.g bound to 1ml of the complex.

After saturating the column packed with composite batch 07 with 126 μ g of hIgG, 5 equal injections of 50 μ g of hIgG solution were performed under the same buffer conditions, but at a flow rate of 1ml/min (fig. 4.2).

All chromatograms were identical. Penetration had occurred after 0.45 minutes with peak maxima of 0.572 minutes or 572 μ l. Subtract 44. mu.l of additional column volume to give a 528. mu.l volume, and the interstitial void volume V in the column_iA good match.

FIGS. 5 to 12 summarize the objects, general concepts, working principles, batch removal material for one potential process, batch results for recovery of hIgG, batch results and summaries in removal of HCP and DNA of the present invention.

In another aspect, the invention relates to the following items:

1. a method of recovering a protein of interest from a feedstock in the form of a solution or suspension comprising at least one protein of interest and at least one impurity compound selected from Host Cell Proteins (HCPs), DNA, RNA or other nucleic acids or a combination of two or more thereof, and optionally comprising albumin, an endotoxin detergent and a microorganism or a fragment thereof or a combination of two or more thereof, the method comprising the steps of:

i) contacting the feedstock with a polymeric network comprising at least one aminopolymer for a sufficient period of time wherein at least one impurity compound is retained;

ii) subsequently, separating the polymeric network from the purified feedstock containing the at least one protein of interest;

iii) optionally, isolating the protein of interest from the feedstock.

2. The method of item 1, wherein the hydrodynamic radius R of the at least one impurity compound retained by the polymeric network comprising at least one aminopolymer_h1Below the hydrodynamic radius of the target protein remaining in the purified feedstock,

preferably, the hydrodynamic radius R of the at least one impurity compound retained therein by the polymeric network comprising at least one amino polymer_h1Less than 4nm and wherein the hydrodynamic radius R of at least one target protein remaining in the purified starting material_h1Is 4nm or more than 4 nm.

3. The process according to any one of the preceding claims, wherein the polymeric network comprising at least one aminopolymer is equilibrated to a pH lower than 8 prior to contacting with the feedstock.

4. The method according to any one of the preceding claims, wherein the protein of interest is an antibody.

5. A method according to any one of the preceding claims, wherein a compound having a pI of 7 or greater than 7 is removed from at least 50% of its initial concentration by a polymeric network comprising at least one aminopolymer which has been equilibrated to a pH below 8.

6. The method of item 5, wherein the impurity compound is a host cell protein.

7. The method of any one of the preceding claims, wherein the host cell protein is removed from the feedstock by at least 90% of its initial concentration.

8. The method according to any one of the preceding claims, wherein the feedstock is a fermentation broth suspension.

9. The process according to any of the preceding claims, wherein in steps i) and ii) a one-step batch adsorption process is used, characterized in that no convective transport is applied.

10. The method of any preceding claim, wherein the polymeric mesh is part of a composite material.

11. The method according to any one of the preceding claims, wherein the amino polymer is poly (vinylamine) or poly (vinylformamide-co-vinylamine), or a mixture thereof.

12. A method of synthesizing a composite material comprising the steps of:

i) filling at least the pore volume of the support material with a solution of at least one crosslinkable polymer or copolymer and at least one crosslinking agent,

ii) and fixing the crosslinkable polymer in situ by crosslinking, wherein the support material is in the form of particles, films or monoliths.

13. The method of item 12, wherein the crosslinkable polymer is poly (vinylformamide-co-vinylamine) or poly (vinylamine), or a mixture thereof.

14. A composite material prepared according to item 12 or 13.

Claims (19)

1. A method of recovering a protein of interest from a feedstock comprising said protein of interest and at least one impurity compound selected from a Host Cell Protein (HCP), DNA, RNA or other nucleic acid or a combination of two or more thereof, and optionally further comprising albumin, endotoxin, a detergent and a microorganism or a fragment thereof or a combination of two or more thereof, said feedstock being in the form of a solution or suspension, and the protein of interest being characterized by a hydrodynamic radius R_h1And the impurity compound is characterized by a hydrodynamic radius R_h2Wherein R is_h1>R_h2(iii) the process comprises the following steps (i) to (iv) and optionally step (v):

(i) providing a polymeric network comprising at least one crosslinked polymer containing positively charged amino groups, said polymer being characterized by a pore size exclusion limit R_hiSaid R is_hiCan be variably set;

(ii) according to hydrodynamic radius R_h1And R_h2Adjusting the variable pore size exclusion limit R of a polymeric network_hiSo that R is_h2<R_hiAnd R is_h1>R_hi；

(iii) Contacting the polymeric network with the feedstock for a time sufficient to retain impurity compounds in the polymeric network and to exclude the protein of interest from the polymeric network;

(iv) separating the polymeric network containing the retained impurity compounds from the feedstock containing the excluded target protein to obtain a purified feedstock;

(v) optionally, the target protein is isolated from the purified starting material.

2. A method of recovering a protein of interest from a feedstock comprising said protein of interest and at least one impurity compound selected from a Host Cell Protein (HCP), DNA, RNA or other nucleic acid or a combination of two or more thereof, and optionally further comprising albumin, endotoxin, a detergent and a microorganism or a fragment thereof or a combination of two or more thereof, said feedstock being in the form of a solution or suspension, and the protein of interest being characterized by a hydrodynamic radius R_h1And the impurity compound is characterized by a hydrodynamic radius R_h2Wherein R is_h1>R_h2The process comprising the following steps (i) and (iii) to (iv) and optionally step (v):

(i) providing a polymeric network comprising at least one crosslinked polymer containing positively charged amino groups, said polymer being characterized by a pore size exclusion limit R_hiTo makeTo obtain R_h2<R_hiAnd R is_h1>R_hi；

(iii) Contacting the polymeric network with the feedstock for a time sufficient to retain impurity compounds in the polymeric network and to exclude the protein of interest from the polymeric network;

(iv) separating the polymeric network containing the retained impurity compounds from the feedstock containing the excluded target protein to obtain a purified feedstock;

(v) optionally, the target protein is isolated from the purified starting material.

3. The method of claim 1, wherein the setting in step (i) or the adjusting in step (ii) or both are performed by one or more of: the method comprises the steps of varying the structure of the polymer, selecting a cross-linking agent for producing the polymeric network, selecting the degree of cross-linking of the polymeric network, controlling the degree of swelling of the polymeric network by varying the solvent used in the preparation and use of the polymeric network, in particular varying the pH of the solvent, thereby varying the degree of protonation of the polymeric network, and controlling the concentration and fixed amount of polymer in the case where the polymeric network is comprised in a complex comprising the polymeric network and a carrier.

4. The method of claim 1 or 3, comprising: (iv) equilibrating the polymeric network obtained in step (ii) to a pH below 8 prior to contacting in step (iii); or

The method of claim 2, comprising: (iv) equilibrating the polymeric network provided in step (i) to a pH of less than 8 prior to contacting in step (iii).

5. The method of claim 4, step (iii) comprising: neutral or positively charged compounds with a pI (isoelectric point) of 7 or greater than 7 are removed by the equilibrated polymeric network.

6. The method of claim 1 or any one of claims 3 to 5 as dependent on claim 1, wherein the variable pore size exclusion limit R of the polymeric mesh provided in step (i) or adjusted in step (ii) is adjusted_hiIs set or adjusted to 1 to 20 nm; or

The method of claim 2 or any one of claims 4 or 5 when dependent on claim 2, wherein the polymeric mesh provided in step (i) has a pore size exclusion limit R_hiIn the range of 1 to 20 nm.

7. The method of claim 1 or any one of claims 3 to 6 as dependent on claim 1, wherein the variable pore size exclusion limit R of the polymeric mesh provided in step (i) or adjusted in step (ii) is adjusted_hiThe range of (3) to (10) nm; or

The method of claim 2 or any one of claims 4 to 6 when dependent on claim 2, wherein the polymeric mesh provided in step (i) has a pore size exclusion limit R_hiIn the range of 3 to 10 nm.

8. The process according to any one of claims 6 or 7, wherein the impurity compound to be retained in the polymeric network is selected from the group consisting of hydrodynamic radius R_h2Impurity compounds below 4 nm.

9. The method of claim 8, wherein the impurity compound is a host cell protein.

10. The method of claim 9, wherein the protein of interest is an antibody.

11. The method of any preceding claim, wherein the feedstock is a fermentation broth suspension.

12. The method according to any one of the preceding claims, wherein steps (iii) and (iv) are carried out as a one-step batch process without the application of convective transport.

13. The method of any of the preceding claims, wherein the polymeric mesh is part of a composite material.

14. The method of any preceding claim, wherein step (i) comprises: converting poly (vinylamine) or poly (vinylformamide-co-vinylamine) or mixtures thereof into said crosslinked polymer containing positively charged amino groups.

15. A method of synthesizing a composite material comprising the steps of:

i) filling at least the pore volume of the support material with a solution of at least one crosslinkable polymer or copolymer and at least one crosslinking agent,

ii) and fixing the crosslinkable polymer in situ by crosslinking, wherein the support material is in the form of particles,

Film-like or monolithic.

16. The method of claim 15, wherein the crosslinkable polymer is poly (vinylformamide-co-vinylamine) or poly (vinylamine), or a mixture thereof.

17. A composite material prepared according to claim 15 or 16.

18. A composite material comprising a porous support material having an average pore diameter of from 20nm to 5mm,

wherein the total pore volume of the porous support material is filled with a crosslinked polymer containing primary or secondary amino groups under application conditions at a pH below 8,

the composite is characterized by a pore size distribution as determined by reverse phase size exclusion chromatography (iSEC) using pullulan standards in 20mM ammonium acetate buffer at pH6,

wherein the hydrodynamic radius R_hIs 4nm and above 4nm, especially 21.7kDa and R_hCalibrated pullulan standards of 3.98nm are excluded from at least 90% of the pore volume, and

wherein at least 35% of the total pore volume is defined by the radius enabling hydrodynamic radius R_hWells entered at 6.2kD as standard pullulan 2.13nm are indicated.

19. The composite of claim 18, wherein the hydrodynamic radius R_hA pullulan standard 10.0kDa 2.7nm can enter at least 15% of the total pore volume.