Method of storing a separation matrix

TECHNICAL FIELD

The present invention relates to a method of storing separation matrices that comprise multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to a porous support. The present invention further relates to separation matrix products and the use of storage liquids comprising at least 50% by volume of an aqueous alkali metal hydroxide solution for the storage of separation matrices.

BACKGROUND ART

Immunoglobulins represent the most prevalent biopharmaceutical products in either manufacture or development worldwide. The high commercial demand for and hence value of this particular therapeutic market has led to the emphasis being placed on pharmaceutical companies to maximize the productivity of their respective immunoglobulin manufacturing processes whilst controlling the associated costs.

Affinity chromatography is used in most cases, as one of the key steps in the purification of these immunoglobulin molecules, such as monoclonal antibodies (mAbs) or polyclonal antibodies (pAbs). A particularly interesting class of affinity reagents is proteins capable of specific binding to invariable parts of an immunoglobulin molecule, such interaction being independent on the antigen-binding specificity of the antibody. Such reagents can be widely used for affinity chromatography recovery of immunoglobulins from different samples such as but not limited to serum or plasma preparations or cell culture derived feed stocks. An example of such a protein is staphylococcal protein A, containing domains capable of binding to the Fc and Fab portions of IgG immunoglobulins from different species. These domains are commonly denoted as the E-, D-, A-, B- and C-domains.

Staphylococcal protein A (SpA) based reagents have due to their high affinity and selectivity found a widespread use in the field of biotechnology, e.g. in affinity chromatography for capture and purification of antibodies as well as for detection or quantification. At present, SpA-based affinity medium probably is the most widely used affinity medium for isolation of monoclonal antibodies and their fragments from different samples including industrial cell culture supernatants. Accordingly, various matrices comprising protein A-ligands are commercially available, for example, in the form of native protein A (e.g. Protein A

SEPHAROSE™, GE Healthcare, Uppsala, Sweden) and also comprised of recombinant protein A (e.g. rProtein A SEPHAROSE™, GE Healthcare). More specifically, the genetic manipulation performed in the commercial recombinant protein A product is aimed at facilitating the attachment thereof to a support and at increasing the productivity of the ligand.

An ongoing trend in the biopharmaceutical industry is the use of versatile multi-product production facilities instead of single-product production facilities, allowing production-on- demand of biopharmaceuticals and a greater product variety, e.g. personalized or orphan biopharmaceuticals. Production campaigns in such multi-product facilities are shorter and there is a need to effectively store the affinity separation matrix between campaigns.

A common medium for storing separation matrices between campaigns is sodium hydroxide. According to the PDA Biotechology Cleaning Validation Committee, concentrations of 0.1 to 1.0 M sodium hydroxide are common for storing packed chromatography columns. However, such storage conditions are associated with exposing the matrix to solutions with pH-values above 13 for long periods. For many affinity chromatography matrices containing

proteinaceous affinity ligands such alkaline environment is a very harsh condition and consequently results in decreased capacity of the affinity separation matrix owing to instability of the ligand to the high pH involved.

An extensive research has therefore been focused on the development of engineered protein ligands that exhibit an improved capacity to withstand alkaline pH-values. For example, Gulich et al. (Susanne Gulich, Martin Linhult, Per-Ake Nygren, Mathias Uhlen, Sophia Hober, Journal of Biotechnology 80 (2000), 169-178) suggested protein engineering to improve the stability properties of a Streptococcal albumin-binding domain (ABD) in alkaline environments. Gulich et al. created a mutant of ABD, wherein all the four asparagine residues have been replaced by leucine (one residue), aspartate (two residues) and lysine (one residue). Further, Gulich et al. report that their mutant exhibits a target protein binding behavior similar to that of the native protein, and that affinity columns containing the engineered ligand show higher binding capacities after repeated exposure to alkaline conditions than columns prepared using the parental non-engineered ligand. Thus, it is concluded therein that all four asparagine residues can be replaced without any significant effect on structure and function.

Recent work shows that changes can also be made to protein A (SpA) to effect similar properties. US patent application publication US 2005/0143566, which is hereby incorporated by reference in its entirety, discloses that when at least one asparagine residue is mutated to an amino acid other than glutamine or aspartic acid, the mutation confers an increased chemical stability at pH-values of up to about 13-14 compared to the parental SpA, such as the B-domain of SpA, or Protein Z, a synthetic construct derived from the B-domain of SpA (US 5,143,844, incorporated by reference in its entirety). The authors show that when these mutated proteins are used as affinity ligands, the separation media as expected can better withstand cleaning procedures using alkaline agents. Further mutations of protein A domains with the purpose of increasing the alkali stability have also been published in US 8,329,860, JP 2006304633A, US 8,674,073, US 2010/0221844, US 2012/0208234, US 9,051,375, US

2014/0031522, US 2013/0274451 and WO 2014/146350, all of which are hereby incorporated by reference in their entireties. However, the currently available mutants are still sensitive to alkaline pH and the corresponding affinity separation matrices are therefore typically stored in 20% ethanol solution or 2% benzyl alcohol solution.

There is thus still a need in this field to obtain a separation matrix containing protein ligands having a further improved stability towards alkaline storage procedures. There is also a need for such separation matrices with an improved binding capacity to allow for economically efficient purification of therapeutic antibodies.

SUM MARY OF THE I NVENTION

The inventors of the present invention have recognised that alcohol solutions are suboptimal for the storage of affinity separation matrices. Alcohols are flammable, subject to regulation, and difficult to dispose of. The inventors have recognised that aqueous alkali metal hydroxide solutions have a number of advantages as storage solutions. Alkali metal hydroxide solutions are bactericidal or bacteriostatic depending on concentration. They can inactive most viruses, bacteria, yeasts, fungi and endotoxins. They are relatively cheap, easily disposed and removal from the separation matrix is simple to detect using pH and/or conductivity measurements.

It is therefore an object of the present invention to provide an affinity separation matrix for the purification of immunoglobulins that is stored in alkali metal hydroxide solution. It is a further object of the present invention to provide a method for storing an affinity separation matrix for the purification of immunoglobulins in an alkali metal hydroxide solution.

These objects are achieved by the method according to the appended claims of storing a separation matrix comprising multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to a porous support. The alkali-stabilized Protein A domains comprise mutants of a parental Fc-binding domain of Staphylococcus Protein A (SpA), as defined by, or having at least 80% such as at least 90%, 95% or 98% identity to, SEQ ID NO 51 or SEQ ID NO 52, wherein the amino acid residues at positions 13 and 44 of SEQ ID NO 51 or 52 are asparagines and wherein at least the asparagine residue at position 3 of SEQ ID NO 51 or 52 has been mutated to an amino acid selected from the group consisting of glutamic acid, lysine, tyrosine, threonine, phenylalanine, leucine, isoleucine, tryptophan, methionine, valine, alanine, histidine and arginine. The method comprises the steps of: a) providing a storage liquid comprising at least 50% by volume of an aqueous alkali metal hydroxide solution; b) permeating the separation matrix with the storage liquid; and c) storing the storage liquid-permeated separation matrix for a storage time of at least 5 days.

By using a separation matrix comprising multimers of immunoglobulin-binding alkali-stabilized Protein A domains as defined above, a highly alkali-stable separation matrix having a high dynamic binding capacity is obtained. The inventors of the present invention have observed that such separation matrices are stable upon immersion in aqueous alkali metal hydroxide solution for extended periods such as five days or more, and substantially retain dynamic binding capacity after such prolonged immersion. This means that such a separation matrices are suitable for storage in alkali metal hydroxide solutions. Further mutations to the immunoglobulin-binding alkali-stabilized Protein A domains may provide further enhancement of properties such as enhanced alkali stability. For example, the glutamine residue at position 1 of SEQ ID NO 51 or 52 may be mutated to an alanine; and/or the asparagine or glutamic acid residue at position 35 of SEQ. ID NO 51 or 52 may be mutated to an alanine.

The multimers of immunoglobulin-binding alkali-stabilized Protein A domains may be homomultimers selected from the group consisting of dimers, trimers, tetramers, pentamers, hexamers, heptamers, octamers or nonamers. By using an appropriate multimer, the immunoglobulin binding capacity and alkali stability of the separation matrix may be increased.

The multimers of immunoglobulin-binding alkali-stabilized Protein A domains may each comprise a C-terminal cysteine residue for covalent coupling to the porous support. The multimers of immunoglobulin-binding alkali-stabilized Protein A domains may be coupled to the porous support via thioether links. This provides a robust, alkali-stable and well-proven method of attaching the ligands to the solid support.

The separation matrix may comprise at least 11 mg/ml, such as at least 15 mg/ml, of the multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to the porous support. This ensures a separation matrix with a good binding capacity.

The porous support may comprise cross-linked polymer particles having a volume-weighted median diameter (d50,v) of 56-70 micrometers and a dry solids weight of 55-80 mg/ml. The porous support may for example be highly cross-linked agarose beads.

The aqueous alkali metal hydroxide solution used in the storage liquid may be sodium hydroxide solution, potassium hydroxide solution or a mixture thereof, preferably sodium hydroxide solution. Sodium hydroxide solution is relatively cheap, readily available and widely accepted for use as a storage solution. The aqueous alkali metal hydroxide solution may have a molarity of from 10 mM to 100 mM, such as from 30 mM to 50 mM. This ensures a solution with a stable pH and good bacteriostatic or bactericidal properties.

The storage liquid may in some instances further comprise a C ₂ - C ₇ alcohol, such as ethanol, isopropanol or benzyl alcohol. A storage liquid combining an alcohol and an alkali metal hydroxide may be more effective in inactivating certain microorganisms, such as some spore- forming bacteria.

The storage liquid may comprise at least 70% by volume aqueous alkali metal hydroxide solution, such as at least 90% by volume aqueous alkali metal hydroxide solution, preferably at least 99% by volume aqueous alkali metal hydroxide solution.

In some instances the storage liquid may consist of, or consist essentially of, aqueous alkali metal hydroxide solution.

The minimum storage time for the separation matrix may be as short a time as storage is required, such as at least 5 days, such as at least 10 days, such as at least 50 days, such as at least 100 days, or such as at least 200 days. The maximum storage time for the separation matrix may be as long a time as storage is required, such as up to 400 days, or such as up to 700 days.

Prior to storing, the separation matrix may be cleaned and/or sanitized with a cleaning fluid, wherein the cleaning fluid comprises at least 50% by volume of an aqueous alkali metal hydroxide solution and wherein the aqueous alkali metal hydroxide solution has a molarity of from 500 mM to 5 M, such as from 1 M to 2 M. The cleaning fluid may consist of, or consist essentially of, aqueous alkali metal hydroxide solution. Thus, the separation matrix may be cleaned, sanitized and stored with little or no requirement for using alcohols.

The separation matrix retains at least 80% of its original dynamic binding capacity, such as at least 90% of its original dynamic binding capacity, after step b), i.e. after prolonged storage. Thus, the separation matrix may be stored in aqueous alkali metal hydroxide solution without subsequent excessive negative impact on its ability to purify immunoglobulins.

According to a further aspect of the present invention, the objects of the present invention are achieved by use of a storage liquid as defined in the appended claims. That is to say, use of a storage liquid comprising at least 50% by volume of an aqueous alkali metal hydroxide solution for the storage of a separation matrix comprising multimers of immunoglobulin- binding alkali-stabilized Protein A domains covalently coupled to a porous support.

The storage liquid may be the same as the storage liquid previously described above in relation to the method of storing a separation matrix. For example, it may comprise, consist essentially of, or consist of, sodium hydroxide solution having a molarity of from 10 mM to 100 mM, such as from 30 mM to 50 mM.

According to another aspect of the present invention, the objects of the present invention are achieved by a separation matrix product according to the appended claims. The separation matrix product comprises a storage receptacle, a separation matrix and a storage liquid. The storage receptacle contains the separation matrix permeated with the storage liquid. The separation matrix comprises multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to a porous support, wherein the alkali-stabilized Protein A domains comprise mutants of a parental Fc-binding domain of Staphylococcus Protein A (SpA), as defined by, or having at least 80% such as at least 90%, 95% or 98% identity to, SEQ ID NO 51 or SEQ. ID NO 52, wherein the amino acid residues at positions 13 and 44 of SEQ ID NO 51 or 52 are asparagines and wherein at least the asparagine residue at position 3 of SEQ ID NO 51 or 52 has been mutated to an amino acid selected from the group consisting of glutamic acid, lysine, tyrosine, threonine, phenylalanine, leucine, isoleucine, tryptophan, methionine, valine, alanine, histidine and arginine. The storage liquid comprises at least 50% by volume of an aqueous alkali metal hydroxide solution.

Thus it is possible to package, store and transport separation matrices stored in aqueous alkali metal hydroxide solution. This avoids the requirement of storing in alcohol solution and thus avoids the need for using volatile and flammable components in the storage liquid. The storage receptacle may for example be a bottle, can or drum made from a liquid- impervious material such as plastic or glass. The storage receptacle may also be a pre-packable column, i.e. a separation column that is filled with separation matrix at the production site.

The storage liquid may be the same as the storage liquid previously described above in relation to the method of storing a separation matrix. For example, it may comprise, consist essentially of, or consist of, sodium hydroxide solution having a molarity of from 10 mM to 100 mM, such as from 30 mM to 50 mM.

Further objects, advantages and novel features of the present invention will become apparent to one skilled in the art from the following detailed description. DEFINITIONS

The terms "antibody" and "immunoglobulin" are used interchangeably herein, and are understood to include also fragments of antibodies, fusion proteins comprising antibodies or antibody fragments and conjugates comprising antibodies or antibody fragments. The terms an "Fc-binding polypeptide", "Fc-binding domain" and "Fc-binding protein" mean a polypeptide, domain or protein respectively, capable of binding to the crystallisable part (Fc) of an antibody and includes e.g. Protein A and Protein G, or any fragment or fusion protein thereof that has maintained said binding property.

The term "linker" herein means an element linking two polypeptide units, monomers or domains to each other in a multimer.

The term "spacer" herein means an element connecting a polypeptide or a polypeptide multimer to a support.

The term "% identity" with respect to comparisons of amino acid sequences is determined by standard alignment algorithms such as, for example, Basic Local Alignment Tool (BLASTTM) described in Altshul et al. (1990) J. Mol. Biol., 215: 403-410. A web-based software for this is freely available from the US National Library of Medicine at

http://blast.ncbi.nlm. nih.gov/Blast. cgi?PROGRAM=blastp&PAGE_TYPE=BlastSearch&LINK_LOC =blasthome . Here, the algorithm "blastp (protein-protein BLAST)" is used for alignment of a query sequence with a subject sequence and determining i.a. the % identity. As used herein, the terms "comprises," "comprising," "containing," "having" and the like can have the meaning ascribed to them in U.S. Patent law and can mean "includes," "including," and the like; "consisting essentially of or "consists essentially" likewise has the meaning ascribed in U.S. Patent law and the term is open-ended, allowing for the presence of more than that which is recited so long as basic or novel characteristics of that which is recited is not changed by the presence of more than that which is recited, but excludes prior art

embodiments. BRIEF DESCRIPTION OF THE FIGURES

For a fuller understanding of the present invention and further objects and advantages of it, the detailed description set out below should be read together with the accompanying figures, and in which: Fig. 1 shows an alignment of the Fc-binding domains as defined by SEO ID NO:l-7 and

51-52.

Fig. 2 shows results from Example 2 for the alkali stability of parental and mutated tetrameric Zvar (SEO ID NO 7) polypeptide variants coupled to an SPR biosensor chip. Fig. 3 shows results from Example 4 for the alkali stability (0.5 M NaOH) of parental and mutated tetrameric Zvar (SEO ID NO 7) polypeptide variants coupled to agarose beads.

Fig. 4 shows results from Example 4 for the alkali stability (1.0 M NaOH) of parental and mutated tetrameric Zvar (SEO ID NO 7) polypeptide variants coupled to agarose beads.

Fig. 5 shows results from Example 7 for the alkali stability (1.0 M NaOH) of agarose beads with different amounts of mutated multimer variants (SEO ID NO. 20) coupled. The results are plotted as the relative remaining dynamic capacity (Obl0%, 6 min residence time) vs. incubation time in 1 M NaOH. Fig. 6 shows results from Example 7 for the alkali stability (1.0 M NaOH) of agarose beads with different amounts of mutated multimer variants (SEO ID NO. 20) coupled. The results are plotted as the relative remaining dynamic capacity (Obl0%, 6 min residence time) after 31 h incubation in 1 M NaOH vs. the ligand content of the prototypes. Fig. 7 shows results from a pH gradient elution of polyclonal human IgG a) from the reference matrix MabSelect SuRe LX and b) a matrix according to the invention.

Fig. 8 shows analyses of the IgGl, lgG2 and lgG4 components in fractions from the chromatograms of Fig. 7. a) reference matrix and b) matrix according to the invention. For each fraction, the first bar (blue) represents IgGl, the second (red) IgG 4 and the third (green) IgG 2.

Fig. 9 shows results from accelerated alkali stability measurements with 1 M NaOH incubation for the reference matrix MabSelect SuRe LX (MSS LX) and a matrix according to the invention. The stability is expressed as the percentage of the 10% breakthrough capacity remaining after incubation.

Fig. 10 shows results from extended incubation with NaOH solutions having

concentrations up to 50 mM for the reference matrix MabSelect SuRe (MSS) and a separation matrix according to the invention (Inv. Ex). The dynamic binding capacity at 10% breakthrough is shown for the matrices prior to and after alkali incubation.

DETAILED DESCRIPTION

One aspect of the present invention concerns a method of storing a separation matrix comprising multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to a porous support.

Throughout this detailed description, two separate numbering conventions may be used.

Unless otherwise stated, the amino acid residue position numbering convention of Fig 1 is used, and the position numbers are designated as corresponding to those in SEO ID NO 4-7. This applies also to multimers, where the position numbers designate the positions in the polypeptide units or monomers according to the convention of Fig. 1, unless otherwise stated.

However, throughout the claims, summary of invention and on occasion in the detailed description, the position numbers corresponding to those of SEO ID NO 51 and 52 are used.

Note that position 1 of SEO ID NO 51 or SEO ID NO 52 corresponds to position 9 of SEO ID NO 4-7, and in this manner the different numbering conventions may be interconverted.

The immunoglobulin-binding alkali-stabilized Protein A domains of the invention, also termed herein as "the polypeptide", comprise, consist essentially of, or consist of mutants of a parental Fc-binding domain of Staphylococcus Protein A (SpA), as defined by, or having at least 80% such as at least 90%, 95% or 98% identity to, SEO ID NO 51 or SEO ID NO 52, wherein the amino acid residues at positions 13 and 44 of SEQ. ID NO 51 or 52 are asparagines and wherein at least the asparagine residue at position 3 of SEQ ID NO 51 or 52 has been mutated to an amino acid selected from the group consisting of glutamic acid, lysine, tyrosine, threonine, phenylalanine, leucine, isoleucine, tryptophan, methionine, valine, alanine, histidine and arginine.

SEQ ID NO 51 (truncated Zvar)

QQ NAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQ SEQ ID NO 52 (truncated C domain)

QQ NAFYEILHLP NLTEEQRNGF IQSLKDDPSV SKEILAEAKK LNDAQ Such immunoglobulin-binding alkali-stabilized Protein A domains may comprise, consist essentially of, or consist of mutants of a parental Fc-binding domain of Staphylococcus Protein A (SpA), as defined by, or having at least 90%, at least 95% or at least 98% identity to, SEQ ID NO: 1 (E-domain), SEQ ID NO: 2 (D-domain), SEQ ID NO:3 (A-domain), SEQ ID NO:22 (variant A- domain), SEQ ID NO: 4 (B-domain), SEQ ID NO: 5 (C-domain), SEQ ID NO:6 (Protein Z) , SEQ ID NO:7 (Zvar), SEQ ID NO 51 (Zvar without the linker region amino acids 1-8 and 56-58) or SEQ ID NO 52 (C-domain without the linker region amino acids 1-8 and 56-58) as illustrated in Fig. 1, wherein at least the asparagine (or serine, in the case of SEQ ID NO 2) residue at the position corresponding to position 11 in SEQ ID NO:4-7 has been mutated to an amino acid selected from the group consisting of glutamic acid, lysine, tyrosine, threonine, phenylalanine, leucine, isoleucine, tryptophan, methionine, valine, alanine, histidine and arginine, and wherein the asparagine residues corresponding to positions 21 and 52 in SEQ ID NO:4-7 (positions 13 and 44 of SEQ ID NO 51 or 52) are conserved.

A number of the Fc-binding domains listed above are shown aligned in Fig. 1. The parental, i.e. non-engineered, Staphylococcus Protein A (SpA) comprises five Fc-dining domains termed domain E (SEQ ID NO 1), D (SEQ ID NO 2), A (SEQ ID NO 3), B (SEQ ID NO 4) and C (SEQ ID NO 5). Protein Z (SEQ ID NO:6) is a mutated B-domain as disclosed in US5143844, hereby incorporated by reference in its entirety. SEQ ID NO 7 denotes a further mutated variant of Protein Z, here called Zvar, with the mutations N3A,N6D,N23T. SEQ ID NO:22 (not shown in Fig. 1) is a natural variant of the A-domain in Protein A from Staphylococcus aureus strain N315, having an A46S mutation, using the position terminology of Fig. 1. SEQ ID NO 51 is Zvar (SEO ID NO 7) without the linker region amino acids at positions 1-8 and 56-58. SEO ID NO 52 is the C-domain of protein A without the linker region amino acids 1-8 and 56-58) as illustrated in Fig. 1.

The mutation of Nil (N3 of SEO ID NO 51:52) in these domains, together with the

conservation of the asparagine residues N21 and N52 (N13 and N44 of SEO ID NO 51:52) confers an improved alkali stability in comparison with the parental domain/polypeptide, without impairing the immunoglobulin-binding properties. Hence, the polypeptide can also be described as an Fc- or immunoglobulin-binding polypeptide, or alternatively as an Fc- or immunoglobulin-binding polypeptide unit. Described in alternative language, the immunoglobulin-binding alkali-stabilized Protein A domains may comprise, consist essentially of, or consist of a sequence as defined by, or having at least 90%, at least 95% or at least 98% identity to SEO ID NO 53.

SEO ID NO 53

XiO X2AFYEI LX3LP NLTEEORX4X5F IXeXyLKDXsPSXg SX10X11X12LAEAKX13 X14NX15AO wherein individually of each other:

Xi=A, O or is deleted

X ₂ =E,K,Y,T,F,L,W,I,M,V,A,H or R

X ₃ =H or K

X ₄ =A or N

X ₅ =A, G, S,Y,0,T,N,F,L,W,l,M,V,D,E,H,R or K, such as S,Y, T,N,F,L,W,l,M,V,D,E,H,R or K

X ₆ =O or E

X ₇ =S or K

X ₈ =E or D

V or is deleted

Xio=K, R, A or is deleted

Xii=A, E, N or is deleted

Xi ₄ =L or Y

Specifically, the amino acid residues in SEQ. ID NO 53 may individually of each other be:

Xi = A or is deleted

X ₂ = E

X ₃ = H

X ₄ = N

Xe = Q

X ₇ = S

X ₈ = D

X ₉ = V or is deleted

Xio = K or is deleted

Xii = A or is deleted

Xi4 = L

In certain embodiments, the amino acid residues in SEQ ID NO 53 may be:

Xi=A, X ₂ = E, X ₃ = H, X ₄ = N, X ₆ = X7 = S, X ₈ = D, X ₉ = V, Xio = K, X = A, X ₁₂ = I, Xi ₃ = K, X _M = L. In some embodiments X ₂ = E, X ₃ = H, X ₄ = N, X ₅ =A, X ₆ = Q, X ₇ = S, X ₈ = D, X _i2 = I, Xi ₃ = K, X _i4 = L and Xi5=D and one or more of Xi, Xg, Xio and Xn is deleted. In further embodiments, Xi=A, X ₂ = E, X ₃ = H, X ₄ = N, X ₅ = S,Y,Q,T,N,F,L,W,I,M,V,D,E,H,R or K , Χε = Q, X ₇ = S, X ₈ = D, Xg = V, Xio = K, Xn = A, Xi2 = I, Xi3 = K, Xi4 = L and Xi ₅ =D, or alternatively Xi=A, X ₂ = E, X ₃ = H, X ₄ = N, X ₅ =A, X ₆ = Q, X ₇ = S, X ₈ = D, X ₉ = V, Xio = K, X = A, X _i2 = I, X _i3 = K, X _M = L and X _i5 = F,Y,W,K or R.

In some embodiments, the amino acid residues may individually of each other be: a) Xi = A or is deleted, X ₂ = E, X ₃ = H, X ₄ = N, X ₆ = Q, X ₇ = S, X ₈ = D, Xg = V or is deleted, Xio = K or is deleted, X = A or is deleted, X _i2 = I, X _i3 = K, X _i4 = L; b) Xi=A, X ₂ = E, X ₃ = H, X ₄ = N, X ₅ =A, X ₆ = Q, X ₇ = S, X ₈ = D, Xg = V, Xio = K, X = A, X _i2 = I, X _i3 = K, Xi ₄ = L and Xi ₅ =D; c) Xi is A, X ₂ = E, X ₃ = H, X ₄ = N, X ₆ = Q, X? = S, X ₈ = D, Xg = V, Xio = K, X = A, X ₁₂ = I, X13 = K, Xi ₄ = L and Xis=D; or d) Xi is A, X ₃ = H, X ₄ = N, X ₅ =A, X ₆ = Q, X7 = S, X ₈ = D, Xg = V, Xio = K, X11 = A, X12 = I, X13 = K, Xi ₄ = L and Xi ₅ =D. The Nil (X ₂ ) mutation (e.g. a N11E or N11K mutation) may be the only mutation or the polypeptide may also comprise further mutations, such as substitutions in at least one of the positions corresponding to positions 3, 6, 9, 10, 15, 18, 23, 28, 29, 32, 33, 36, 37, 40, 42, 43, 44, 47, 50, 51, 55 and 57 in SEQ. ID NO:4-7. In one or more of these positions, the original amino acid residue may e.g. be substituted with an amino acid which is not asparagine, proline or cysteine. The original amino acid residue may e.g. be substituted with an alanine, a valine, a threonine, a serine, a lysine, a glutamic acid or an aspartic acid. Further, one or more amino acid residues may be deleted, e.g. from positions 1-6 and/or from positions 56-58.

In some embodiments, the amino acid residue at the position corresponding to position 9 in SEQ. ID NO:4-7 (Xi) is an amino acid other than glutamine, asparagine, proline or cysteine, such as an alanine or it can be deleted. The combination of the mutations at positions 9 and 11 provides particularly good alkali stability, as shown by the examples. In specific embodiments, in SEQ ID NO: 7 the amino acid residue at position 9 is an alanine and the amino acid residue at position 11 is a lysine or glutamic acid, such as a lysine. Mutations at position 9 are also discussed in copending application PCT/SE2014/050872, which is hereby incorporated by reference in its entirety.

In some embodiments, the amino acid residue at the position corresponding to position 50 in SEQ ID NO:4-7 (X13) is an arginine or a glutamic acid.

In certain embodiments, the amino acid residue at the position corresponding to position 3 in SEQ ID NO:4-7 is an alanine and/or the amino acid residue at the position corresponding to position 6 in SEQ ID NO:4-7 is an aspartic acid. One of the amino acid residues at positions 3 and 6 may be an asparagine and in an alternative embodiment both amino acid residues at positions 3 and 6 may be asparagines.

In some embodiments the amino acid residue at the position corresponding to position 43 in SEQ ID NO:4-7 (Xn) is an alanine or a glutamic acid, such as an alanine or it can be deleted. In specific embodiments, the amino acid residues at positions 9 and 11 in SEQ. ID NO: 7 are alanine and lysine/glutamic acid respectively, while the amino acid residue at position 43 is alanine or glutamic acid.

In certain embodiments the amino acid residue at the position corresponding to position 28 in SEQ ID NO:4-7 (X5) is an alanine or an asparagine, such as an alanine.

In some embodiments the amino acid residue at the position corresponding to position 40 in SEQ ID NO:4-7 (X9) is selected from the group consisting of asparagine, alanine, glutamic acid and valine, or from the group consisting of glutamic acid and valine, or valine, or it can be deleted. In specific embodiments, the amino acid residues at positions 9 and 11 in SEQ ID NO: 7 are alanine and glutamic acid respectively, while the amino acid residue at position 40 is valine. Optionally, the amino acid residue at position 43 may then be alanine or glutamic acid.

In certain embodiments, the amino acid residue at the position corresponding to position 42 in SEQ ID NO:4-7 (X10) is an alanine, lysine or arginine or it can be deleted.

In some embodiments the amino acid residue at the position corresponding to position 18 in SEQ ID NO:4-7 (X3) is a lysine or a histidine, such as a lysine.

In certain embodiments the amino acid residue at the position corresponding to position 33 in SEQ ID NO:4-7 (X ₇ ) is a lysine or a serine, such as a lysine.

In some embodiments the amino acid residue at the position corresponding to position 37 in SEQ ID NO:4-7 (Xs) is a glutamic acid or an aspartic acid, such as a glutamic acid. In certain embodiments the amino acid residue at the position corresponding to position 51 in SEQ ID NO:4-7 (Xi ₄ ) is a tyrosine or a leucine, such as a tyrosine.

In some embodiments, the amino acid residue at the position corresponding to position 44 in SEQ ID NO:4-7 (X12) is a leucine or an isoleucine. In specific embodiments, the amino acid residues at positions 9 and 11 in SEQ ID NO: 7 are alanine and lysine/glutamic acid

respectively, while the amino acid residue at position 44 is isoleucine. Optionally, the amino acid residue at position 43 may then be alanine or glutamic acid.

In some embodiments, the amino acid residues at the positions corresponding to positions 1, 2, 3 and 4 or to positions 3, 4, 5 and 6 in SEQ ID NO: 4-7 have been deleted. In specific variants of these embodiments, the parental polypeptide is the C domain of Protein A (SEQ. ID NO: 5). The effects of these deletions on the native C domain are described in US9018305 and US8329860, which are hereby incorporated by reference in their entireties.

In certain embodiments, the mutation in SEQ. ID NO 4-7, such as in SEQ ID NO 7, is selected from the group consisting of:NllK; N11E; N11Y; NUT; N11F; N11L; N11W; Nlll; N11M;

N11V; N11A; N11H; N11R; N11E,Q32A; N11E,Q32E,Q40E; N11E,Q32E,K50R; Q9A,N11E,N43A;

Q9A,N11E,N28A,N43A; Q9A,N11E,Q40V,A42K,N43E,L44I; Q9A,N11E,Q40V,A42K,N43A,L44I;

N11K,H18K,S33K,D37E,A42R,N43A,L44I,K50R,L51Y; Q9A,N11E,N28A,Q40V,A42K,N43A,L44I;

Q9A,N11K,H18K,S33K,D37E,A42R,N43A,L44I,K50R,L51Y; NllK, H18K, D37E, A42R, N43A, L44I; Q9A, NllK, H18K, D37E, A42R, N43A, L44I; Q9A, NllK, H18K, D37E, A42R, N43A, L44I, K50R;

Q9A,N11K,H18K,D37E,A42R; Q9A,N11E,D37E,Q40V,A42K,N43A,L44I and

Q9A,N11E,D37E,Q40V,A42R,N43A,L44I. These mutations provide particularly high alkaline stabilities. The mutation in SEQ ID NO 4-7, such as in SEQ ID NO 7, can also be selected from the group consisting of NllK; N11Y; N11F; N11L; N11W; Nlll; N11M; N11V; N11A; N11H; N11R; Q9A,N11E,N43A; Q9A,N11E,N28A,N43A; Q9A,N11E,Q40V,A42K,N43E,L44I;

Q9A,N11E,Q40V,A42K,N43A,L44I; Q9A,N11E,N28A,Q40V,A42K,N43A,L44I;

N11K,H18K,S33K,D37E,A42R,N43A,L44I,K50R,L51Y;

Q9A,N11K,H18K,S33K,D37E,A42R,N43A,L44I,K50R,L51Y; NllK, H18K, D37E, A42R, N43A, L44I; Q9A, NllK, H18K, D37E, A42R, N43A, L44I and Q9A, NllK, H18K, D37E, A42R, N43A, L44I, K50R.

In some embodiments, the polypeptide comprises or consists essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of: SEQ ID NO 8, SEQ ID NO 9, SEQ ID NO 10, SEQ ID NO 11, SEQ ID NO 12, SEQ ID NO 13, SEQ ID NO 14, SEQ ID NO 15, SEQ ID NO 16, SEQ ID NO 23, SEQ ID NO 24, SEQ ID NO 25, SEQ ID NO 26, SEQ ID NO 27, SEQ ID NO 28, SEQ ID NO 29, SEQ ID NO 36,

SEQ ID NO 37, SEQ ID NO 38, SEQ ID NO 39, SEQ ID NO 40, SEQ ID NO 41, SEQ ID NO 42, SEQ ID NO 43, SEQ ID NO 44, SEQ ID NO 45, SEQ ID NO 46, SEQ ID NO 47, SEQ ID NO 48, SEQ ID NO 49 and SEQ ID NO 50. It may e.g. comprise or consist essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of: SEQ ID NO 8, SEQ ID NO 9, SEQ ID NO 10, SEQ ID NO 11, SEQ ID NO 16, SEQ ID NO 23, SEQ ID NO 24, SEQ ID NO 25, SEQ ID NO 26, SEQ ID NO 27, SEQ ID NO 28 and SEQ ID NO 29. It can also comprise or consist essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of: SEO ID NO 8, SEO ID NO 9, SEO ID NO 10, SEO ID NO 11, SEO ID NO 16, SEO ID NO 23, SEO ID NO 24, SEO ID NO 25, SEO ID NO 27, SEO ID NO 28, SEO ID NO 38, SEO ID NO 40; SEO ID NO 41; SEO ID NO 42; SEO NO 43, SEO ID NO 44, SEO ID NO 45, SEO ID NO 46, SEO ID NO 47 and SEO ID NO 48.

In certain embodiments, the polypeptide comprises or consists essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of SEO ID NO 54-70; comprises or consists essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of SEO ID NO 71-75; or it may comprise or consist essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of SEO ID NO 76-79. It may further comprise or consist essentially of a sequence defined by or having at least 90%, 95% or 98% identity to an amino acid sequence selected from the group consisting of SEO ID NO 89-95.

The polypeptide may e.g. be defined by a sequence selected from the groups above or from subsets of these groups, but it may also comprise additional amino acid residues at the N- and/or C-terminal end, e.g. a leader sequence at the N-terminal end and/or a tail sequence at the C-terminal end. SEO ID N0 8 Zvar(09A,NllE,N43A)

VDAKFDKEAO EAFYEILHLP NLTEEORNAF I SLKDDPS SAALLAEAKK LNDAOAPK

SEO ID NO 9 Zvar(09A,NllE,N28A,N43A)

VDAKFDKEAO EAFYEILHLP NLTEEORAAF IOSLKDDPSO SAALLAEAKK LNDAOAPK

SEO ID NO 10 Zvar( 9A,NllE, 40V,A42K,N43E,L44l)

VDAKFDKEAO EAFYEILHLP NLTEEORNAF I SLKDDPSV SKEILAEAKK LNDAOAPK

SEO ID NO 11 Zvar( 9A,NllE, 40V,A42K,N43A,L44l)

VDAKFDKEAO EAFYEILHLP NLTEEORNAF IOSLKDDPSV SKAILAEAKK LNDAOAPK SEQ ID NO 12 Zvar(NllE,Q32A)

VDAKFDKEQQ EAFYEILHLP NLTEEQRNAF lASLKDDPSQ SANLLAEAKK LNDAQAPK SEQ ID NO 13 Zvar(NllE)

VDAKFDKEQQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 14 Zvar(NllE,Q32E,Q40E)

VDAKFDKEQQ EAFYEILHLP NLTEEQRNAF IESLKDDPSE SANLLAEAKK LNDAQAPK

SEQ ID NO 15 Zvar(NllE,Q32E,K50R)

VDAKFDKEQQ EAFYEILHLP NLTEEQRNAF IESLKDDPSQ SANLLAEAKR LNDAQAPK

SEQ ID NO 16 Zvar(NllK)

VDAKFDKEQQ KAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 23 Zvar(NllK,H18K,S33K,D37E,A42R,N43A,L44l,K50R,L51Y)

VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQKLKDEPSQ SRAILAEAKR YNDAQAPK SEQ ID NO 24 Zvar(Q9A,NllE,N28A,Q40V,A42K,N43A,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 25 Zvar(Q9A,NllK,H18K,S33K,D37E,A42R,N43A,L44l,K50R,L51Y) VDAKFDKEAQ KAFYEILKLP NLTEEQRAAF IQKLKDEPSQ SRAILAEAKR YNDAQAPK

SEQ ID NO 26 Zvar(NllK, H18K, D37E, A42R, N43A, L44I)

VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRAILAEAKK LNDAQAPK

SEQ ID NO 27 Zvar(Q9A, N11K, H18K, D37E, A42R, N43A, L44I)

VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRAILAEAKK LNDAQAPK

SEQ ID NO 28 Zvar(Q9A, N11K, H18K, D37E, A42R, N43A, L44I, K50R) VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRAILAEAKR LNDAQAPK

SEQ ID NO 29 Zvar(Q9A,NllK,H18K,D37E,A42R)

VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRNLLAEAKK LNDAQAPK

SEQ ID NO 36 B(Q9A,N11E,Q40V,A42K,N43A,L44I)

ADNKFNKEAQ EAFYEILHLP NLNEEQRNGF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 37 C(Q9A,N11E,E43A)

ADNKFNKEAQ EAFYEILHLP NLTEEQRNGF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 38 Zvar(NllY)

VDAKFDKEQQ YAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK SEQ ID NO 39 Zvar(NllT)

VDAKFDKEQQ TAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 40 Zvar(NllF)

VDAKFDKEQQ FAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID N0 41 Zvar(NllL)

VDAKFDKEQQ LAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID N0 42 Zvar(NllW)

VDAKFDKEQQ WAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 43 Zvar(Nlll)

VDAKFDKEQQ IAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK SEQ ID NO 44 Zvar(NllM)

VDAKFDKEQQ MAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK SEQ ID NO 45 Zvar(NllV)

VDAKFDKEQQ VAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID N0 46 Zvar(NllA)

VDAKFDKEQQ AAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 47 Zvar(NllH)

VDAKFDKEQQ HAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK SEQ ID N0 48 Zvar(NllR)

VDAKFDKEQQ RAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SANLLAEAKK LNDAQAPK

SEQ ID NO 49 Zvar(Q9A,NllE,D37E,Q40V,A42K,N43A,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 50 Zvar(Q9A,NllE,D37E,Q40V,A42R,N43A,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SRAILAEAKK LNDAQAPK

SEQ ID NO 54 Zvar(Q9A,NllE, A29G,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNGF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 55 Zvar(Q9A,NllE, A29S,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNSF IQSLKDDPSV SKAILAEAKK LNDAQAPK SEQ ID NO 56 Zvar(Q9A,NllE, A29Y,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNYF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 57 Zvar(Q9A,NllE, A29Q,Q 0V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNQF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 58 Zvar(Q9A,NllE, A29T,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNTF IQSLKDDPSV SKAILAEAKK LNDAQAPK SEQ ID NO 59 Zvar(Q9A,NllE, A29N,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNNF IQSLKDDPSV SKAILAEAKK LNDAQAPK SEQ ID NO 60 Zvar(Q9A,NllE, A29F,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNFF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 61 Zvar(Q9A,NllE, A29L,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNLF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 62 Zvar(Q9A,NllE, A29W,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNWF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 63 Zvar(Q9A,NllE, A29I,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNIF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 64 Zvar(Q9A,NllE, A29M,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNMF IQSLKDDPSV SKAILAEAKK LNDAQAPK SEQ ID NO 65 Zvar(Q9A,NllE, A29V,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNVF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 66 Zvar(Q9A,NllE, A29D,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNDF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 67 Zvar(Q9A,NllE, A29E,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNEF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 68 Zvar(Q9A,NllE, A29H,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNHF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 69 Zvar(Q9A,NllE, A29R,Q40V,A42K,N43A,L44I) VDAKFDKEAQ EAFYEILHLP NLTEEQRNRF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 70 Zvar(Q9A,NllE, A29K,Q40V,A42K,N43A,L44I)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNKF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 71 Zvar(Q9A,NllE, Q40V,A42K,N43A,L44I,D53F)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNFAQAPK

SEQ ID NO 72 Zvar(Q9A,NllE, Q40V,A42K,N43A,L44I,D53Y)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNYAQAPK

SEQ ID NO 73 Zvar(Q9A,NllE, Q40V,A42K,N43A,L44I,D53W)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNWAQAPK SEQ ID NO 74 Zvar(Q9A,NllE, Q40V,A42K,N43A,L44I,D53K)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNKAQAPK

SEQ ID NO 75 Zvar(Q9A,NllE, Q40V,A42K,N43A,L44I,D53R)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNRAQAPK

SEQ ID NO 76 Zvar(Q9del,NllE, Q40V,A42K,N43A,L44I)

VDAKFDKE_Q EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK

SEQ ID NO 77 Zvar(Q9A,NllE, Q40del,A42K,N43A,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPS_ SKAILAEAKK LNDAQAPK

SEQ ID NO 78 Zvar(Q9A,NllE, Q40V,A42del,N43A,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV S_AILAEAKK LNDAQAPK SEQ ID NO 79 Zvar(Q9A,NllE, Q40V,A42K,N43del,L44l)

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKJLAEAKK LNDAQAPK SEQ I D NO 89 Zvar(D2del,A3del,K4del,Q9A,NllE,Q40V,A42K,N43A,L44l)

V FDKEAQ EAFYEILH LP N LTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPK

SEQ I D NO 90 Zvar(Vldel,D2del,Q9A,NllE,Q40V,A42K,N43A,L44l,K58del)

AKFDKEAQ EAFYEILH LP N LTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAP_

SEQ I D NO 91 Zvar(K4del,F5del,D6del,K7del,E8del,Q9A,NllE,Q40V,A42K,N43A,L 44l)

VDA AQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPK

SEQ I D NO 92 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l,A56del,P57del,K58del)

VDAKFDKEAQ EAFYEILH LP N LTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQ

SEQ I D NO 93 Zvar(Vldel,,D2del,A3del,Q9A,NllE,Q40V,A42K,N43A,L44l)

KFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPK

SEQ I D NO 94

Zvar(Vldel,D2del,A3del,K4del,F5del,D6del,K7del,E8del,Q9A,Nll E,Q.40V,A42K,N43A,L44l)

AQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPK SEQ I D NO 95 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l,K58_insYEDG)

VDAKFDKEAQ EAFYEILH LP N LTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPKYE DG

The separation matrix comprises multimers of the immunoglobulin-binding alkali-stabilized Protein A domains. Such multimers comprise, consist essentially of, or consist of a plurality of immunoglobulin-binding alkali-stabilized Protein A domains (polypeptide units) as defined by any embodiment disclosed above. The use of multimers may increase the immunoglobulin binding capacity and multimers may also have a higher alkali stability than monomers. The multimer can e.g. be a dimer, a trimer, a tetramer, a pentamer, a hexamer, a heptamer, an octamer or a nonamer. The multimer may be a homomultimer, where all the units in the multimer are identical or it can be a heteromultimer, where at least one unit differs from the others. Advantageously, all the units in the multimer are alkali stable, such as by comprising the mutations/conservations disclosed above. The polypeptides can be linked to each other directly by peptide bonds between the C-terminal and N-terminal ends of the polypeptides. Alternatively, two or more units in the multimer can be linked by linkers comprising oligomeric or polymeric species, such as linkers comprising peptides with up to 25 or 30 amino acids, such as 3-25 or 3-20 amino acids. The linkers may e.g. comprise or consist essentially of a peptide sequence defined by, or having at least 90% identity or at least 95% identity, with an amino acid sequence selected from the group consisting of APKVDAKFDKE, APKVDNKFNKE,

APKADNKFNKE, APKVFDKE, APAKFDKE, AKFDKE, APKVDA, VDAKFDKE, APKKFDKE, APK, APKYEDGVDAKFDKE and YEDG or alternatively selected from the group consisting of

APKADNKFNKE, APKVFDKE, APAKFDKE, AKFDKE, APKVDA, VDAKFDKE, APKKFDKE,

APKYEDGVDAKFDKE and YEDG. They can also consist essentially of a peptide sequence defined by or having at least 90% identity or at least 95% identity with an amino acid sequence selected from the group consisting of APKADNKFNKE, APKVFDKE, APAKFDKE, AKFDKE, APKVDA, VDAKFDKE, APKKFDKE, APK and APKYEDGVDAKFDKE. In some embodiments the linkers do not consist of the peptides APKVDAKFDKE or APKVDNKFNKE, or alternatively do not consist of the peptides APKVDAKFDKE, APKVDNKFNKE , APKFNKE, APKFDKE, APKVDKE or APKADKE.

The nature of such a linker should preferably not destabilize the spatial conformation of the protein units. This can e.g. be achieved by avoiding the presence of proline in the linkers. Furthermore, said linker should preferably also be sufficiently stable in alkaline environments not to impair the properties of the mutated protein units. For this purpose, it is advantageous if the linkers do not contain asparagine. It can additionally be advantageous if the linkers do not contain glutamine. The multimer may further at the N-terminal end comprise a plurality of amino acid residues e.g. originating from the cloning process or constituting a residue from a cleaved off signaling sequence. The number of additional amino acid residues may e.g. be 20 or less, such as 15 or less, such as 10 or less or 5 or less. As a specific example, the multimer may comprise an AQ, AQ.GT, VDAKFDKE, AQVDAKFDKE or AQGTVDAKFDKE sequence at the N- terminal end.

In certain embodiments, the multimer may comprise, or consist essentially, of a sequence selected from the group consisting of: SEQ. ID NO 80-87. These and additional sequences are listed below and named as Parent(Mutations)n, where n is the number of monomer units in a multimer. SEQ ID NO 17 Zvar(Q9A,NllE,N43A)4

AQGT VDAKFDKEAQ, EAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSQ SAALLAEAKK LNDAQAPKC

SEQ ID NO 18 Zvar(Q9A,NllE,N28A,N43A)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSQ SAALLAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSQ SAALLAEAKK LNDAQAPKC

SEQ ID NO 19 Zvar(Q9A,NllE,Q40V,A42K,N43E,L44l)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKEILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKEILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKEILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKEILAEAKK LNDAQAPKC

SEQ ID NO 20 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC SEQ ID NO 30 Zvar(NllK,H18K,S33K,D37E,A42R,N43A,L44l,K50R,L51Y)4

AQGT VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQKLKDEPSQ SRAILAEAKR YNDAQAPK VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQKLKDEPSQ SRAILAEAKR YNDAQAPK VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQKLKDEPSQ SRAILAEAKR YNDAQAPK VDAKFDKEQQ KAFYEILKLP NLTEEQRNAF IQKLKDEPSQ SRAILAEAKR YNDAQAPKC

SEQ ID NO 31 Zvar(Q9A,NllK,H18K,D37E,A42R)4 AQGT VDAKFDKEAQ, KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRNLLAEAKK LNDAQAPK VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRNLLAEAKK LNDAQAPK VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRNLLAEAKK LNDAQAPK VDAKFDKEAQ KAFYEILKLP NLTEEQRNAF IQSLKDEPSQ SRNLLAEAKK LNDAQAPKC

SEQ ID NO 32 Zvar(Q9A,NllE,N28A,Q40V,A42K,N43A,L44l)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRAAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 33 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)6

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC SEQ ID NO 34 Zvar(Q9A,NllE,D37E,Q40V,A42K,N43A,L44l)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SKAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 35 Zvar(Q9A,NllE,D37E,Q40V,A42R,N43A,L44l)4

AQGT VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SRAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SRAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SRAILAEAKK LNDAQAPK VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDEPSV SRAILAEAKK LNDAQAPKC

SEQ ID NO 80 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with D2, A3 and K4 in linker deleted VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 81 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with K58, VI and D2 in linker deleted VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAP AKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 82 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with P57, K58, VI, D2 and A3 in linker deleted

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAP AKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 83 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with K4, F5, D6, K7 and E8 in linker deleted

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK VDAAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 84 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with A56, P57 and K58 in linker deleted VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQ VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 85 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with VI, D2 and A3 in linker deleted VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK KFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 86 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with VI, D2, A3, K4, F5, D6, K7 and E8 in linker deleted

VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPK AQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LNDAQAPKC

SEQ ID NO 87 Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)2 with YEDG inserted in linker between K58 and VI VDAKFDKEAQ EAFYEI LHLP N LTEEQRNAF IQSLKDDPSV SKAI LAEAKK LNDAQAPK YEDG VDAKFDKEAQ EAFYEILHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPKC

SEQ I D NO 88 Zvar2

VDAKFDKEAQ EAFYEI LHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPK VDAKFDKEAQ EAFYEI LHLP NLTEEQRNAF IQSLKDDPSV SKAILAEAKK LN DAQAPKC

I n some embodiments, the polypeptide and/or multimer, as disclosed above, further comprises at the C-terminal or N-terminal end one or more coupling elements, selected from the group consisting of one or more cysteine residues, a plurality of lysine residues and a plurality of histidine residues. The coupling element(s) may also be located within 1-5 amino acid residues, such as within 1-3 or 1-2 amino acid residues from the C-terminal or N-terminal end. The coupling element may e.g. be a single cysteine at the C-terminal end. The coupling element(s) may be directly linked to the C- or N-terminal end, or it/they may be linked via a stretch comprising up to 15 amino acids, such as 1-5, 1-10 or 5-10 amino acids. This stretch should preferably also be sufficiently stable in alkaline environments not to impair the properties of the mutated protein. For this purpose, it is advantageous if the stretch does not contain asparagine. It can additionally be advantageous if the stretch does not contain glutamine. An advantage of having a C-terminal cysteine is that endpoint coupling of the protein can be achieved through reaction of the cysteine thiol with an electrophilic group on a support. This provides excellent mobility of the coupled protein which is important for the binding capacity.

The alkali stability of the polypeptide or multimer can be assessed by coupling it to a surface plasmon resonance (SPR) chip, e.g. to Biacore CM5 sensor chips as described in the examples, using e.g. N HS- or maleimide coupling chemistries, and measuring the immunoglobulin- binding capacity of the chip, typically using polyclonal human IgG, before and after incubation in alkaline solutions at a specified temperature, e.g. 22 +/- 2 °C. The incubation can e.g. be performed in 0.5 M NaOH for a number of 10 min cycles, such as 100, 200 or 300 cycles. The IgG capacity of the matrix after 100 10 min incubation cycles in 0.5 M NaOH at 22 +/- 2 °C can be at least 55, such as at least 60, at least 80 or at least 90% of the IgG capacity before the incubation. Alternatively, the remaining IgG capacity after 100 cycles for a particular mutant measured as above can be compared with the remaining IgG capacity for the parental polypeptide/multimer. In this case, the remaining IgG capacity for the mutant may be at least 105%, such as at least 110%, at least 125%, at least 150% or at least 200% of the parental polypeptide/multimer. The immunoglobulin-binding alkali-stabilized Protein A domains and/or multimers thereof may be encoded by a nucleic acid sequence, such as an RNA sequence or a DNA sequence encoding the polypeptide or multimer. A vector, such as a plasmid, which in addition to the coding sequence comprises the required signal sequences, may be used for expression of the polypeptide or multimer. The vector may comprise nucleic acid encoding a multimer as described above, wherein the separate nucleic acids encoding each unit may have

homologous or heterologous DNA sequences.

An expression system, which comprises a nucleic acid or a vector as disclosed above, may be used for expression of the polypertide or multimer. The expression system may e.g. be a gram- positive or gram-negative prokaryotic host cell system, e.g. E.coli or Bacillus sp. which has been modified to express the present polypeptide or multimer. Alternatively, the expression system may be a eukaryotic host cell system, such as a yeast, e.g. Pichia pastoris or

Saccharomyces cerevisiae, or mammalian cells, e.g. CHO cells.

The separation matrix comprises, consists essentially of, or consists of multimers of immunoglobulin-binding alkali-stabilized Protein A domains as described above, covalently coupled to a porous support.

The separation matrix may comprise at least 11, such as 11-21, 15-21 or 15-18 mg/ml Fc- binding ligands covalently coupled to a porous support, wherein : a) the ligands comprise multimers of alkali-stabilized Protein A domains, b) the porous support comprises cross-linked polymer particles having a volume-weighted median diameter (d50,v) of 56-70, such as 56-66, micrometers and a dry solids weight of 55-

80, such as 60-78 or 65-78, mg/ml. The cross-linked polymer particles may further have a pore size corresponding to an inverse gel filtration chromatography Kd value of 0.69-0.85, such as 0.70-0.85 or 0.69-0.80, for dextran of Mw 110 kDa. The multimers may e.g. comprise tetramers, pentamers, hexamers or heptamers of alkali-stabilized Protein A domains, such as hexamers of alkali-stabilized Protein A domains. The combination of the high ligand contents with the particle size range, the dry solids weight range and the optional Kd range provides for a high binding capacity, e.g. such that the 10% breakthrough dynamic binding capacity for IgG is at least 45 mg/ml, such as at least 50 or at least 55 mg/ml at 2.4 min residence time.

Alternatively, or additionally, the 10% breakthrough dynamic binding capacity for IgG may be at least 60 mg/ml, such as at least 65, at least 70 or at least 75 mg/ml at 6 min residence time.

The alkali-stabilized Protein A domain multimers are highly selective for IgG and the separation matrix can suitably have a dissociation constant for human lgG2 of below 0.2 mg/ml, such as below 0.1 mg/ml, in 20 mM phosphate buffer, 180 mM NaCI, pH 7.5. This can be determined according to the adsorption isotherm method described in N Pakiman et al: J Appl Sci 12, 1136-1141 (2012).

In certain embodiments the separation matrix comprises at least 15, such as 15-21 or 15-18 mg/ml Fc-binding ligands covalently coupled to a porous support, wherein the ligands comprise multimers of alkali-stabilized Protein A domains. These multimers can suitably be as disclosed in any of the embodiments described above or as specified below.

In some embodiments the separation matrix comprises 5 - 25, such as 5-20 mg/ml, 5 - 15 mg/ml, 5 - 11 mg/ml or 6 - 11 mg/ml of the polypeptide or multimer coupled to the support. The amount of coupled polypeptide/multimer can be controlled by the concentration of polypeptide/multimer used in the coupling process, by the activation and coupling conditions used and/or by the pore structure of the support used. As a general rule the absolute binding capacity of the matrix increases with the amount of coupled polypeptide/multimer, at least up to a point where the pores become significantly constricted by the coupled

polypeptide/multimer. Without being bound by theory, it appears though that for the Kd values recited for the support, the constriction of the pores by coupled ligand is of lower significance. The relative binding capacity per mg coupled polypeptide/multimer will decrease at high coupling levels, resulting in a cost-benefit optimum within the ranges specified above.

Such a separation matrix is useful for separation of immunoglobulins or other Fc-containing proteins and, due to the improved alkali stability of the polypeptides/multimers, the matrix will withstand highly alkaline conditions during cleaning, which is essential for long-term repeated use in a bioprocess separation setting. The alkali stability of the matrix can be assessed by measuring the immunoglobulin-binding capacity, typically using polyclonal human IgG, before and after incubation in alkaline solutions at a specified temperature, e.g. 22 +/- 2 °C. The incubation can e.g. be performed in 0.5 M or 1.0 M NaOH for a number of 15 min cycles, such as 100, 200 or 300 cycles, corresponding to a total incubation time of 25, 50 or 75 h. The IgG capacity of the matrix after 96-100 15 min incubation cycles or a total incubation time of 24 or 25 h in 0.5 M NaOH at 22 +/- 2 °C can be at least 80, such as at least 85, at least 90 or at least 95% of the IgG capacity before the incubation. The capacity of the matrix after a total incubation time of 24 h in 1.0 M NaOH at 22 +/- 2 °C can be at least 70, such as at least 80 or at least 90% of the IgG capacity before the incubation. The the 10% breakthrough dynamic binding capacity ( bl0%) for IgG at 2.4 min or 6 min residence time may e.g. be reduced by less than 20 % after incubation 31 h in 1.0 M aqueous NaOH at 22 +/- 2 C.

As the skilled person will understand, the expressed polypeptide or multimer should be purified to an appropriate extent before being immobilized to a support. Such purification methods are well known in the field, and the immobilization of protein-based ligands to supports is easily carried out using standard methods. Suitable methods and supports will be discussed below in more detail.

The porous support of the separation matrix may be of any suitable well-known kind. A conventional affinity separation matrix is often of organic nature and based on polymers that expose a hydrophilic surface to the aqueous media used, i.e. expose hydroxy (-OH), carboxy (- COOH), carboxamido (-CON H2, possibly in N- substituted forms), amino (-N H2, possibly in substituted form), oligo- or polyethylenoxy groups on their external and, if present, also on internal surfaces. The porosity of the support can be expressed as a Kav or Kd value (the fraction of the pore volume available to a probe molecule of a particular size) measured by inverse size exclusion chromatography, e.g. according to the methods described in Gel Filtration Principles and Methods, Pharmacia LKB Biotechnology 1991, pp 6-13. Kav is determined as the ratio (Ve-Vo)/(V _t -Vo), where Ve is the elution volume of a probe molecule (e.g. Dextran 110 kD), Vo is the void volume of the column (e.g. the elution volume of a high Mw void marker, such as raw dextran) and V _t is the total volume of the column. Kd can be determined as (V _e -Vo)/Vi, where V, is the elution volume of a salt (e.g. NaCI) able to access all the volume except the matrix volume (the volume occupied by the matrix polymer molecules). By definition, both Kd and Kav values always lie within the range 0 - 1. The Kav value can advantageously be 0.6 - 0.95, e.g. 0.7 - 0.90 or 0.6 - 0.8, as measured with dextran of Mw 110 kDa as a probe molecule. The Kd value as measured with dextran of Mw 110 kDa can suitably be 0.68-0.90, such as 0.68-0.85 or 0.70-0.85. An advantage of this is that the support has a large fraction of pores able to accommodate both the polypeptides/multimers of the invention and immunoglobulins binding to the polypeptides/multimers and to provide mass transport of the immunoglobulins to and from the binding sites.

The polypeptides or multimers may be attached to the porous support via conventional coupling techniques utilising e.g. thiol, amino and/or carboxy groups present in the ligand. Bisepoxides, epichlorohydrin, CNBr, N-hydroxysuccinimide (NHS) etc are well-known coupling reagents. Between the support and the polypeptide/multimer, a molecule known as a spacer can be introduced, which improves the availability of the polypeptide/multimer and facilitates the chemical coupling of the polypeptide/multimer to the support. Depending on the nature of the polypeptide/multimer and the coupling conditions, the coupling may be a multipoint coupling (e.g. via a plurality of lysines) or a single point coupling (e.g. via a single cysteine). In certain embodiments the polypeptides or multimers are coupled to the support via thioether bonds. Methods for performing such coupling are well-known in this field and easily performed by the skilled person in this field using standard techniques and equipment.

Thioether bonds are flexible and stable and generally suited for use in affinity

chromatography. In particular when the thioether bond is via a terminal or near-terminal cysteine residue on the polypeptide or multimer, the mobility of the coupled

polypeptide/multimer is enhanced which provides improved binding capacity and binding kinetics. In some embodiments the polypeptide/multimer is coupled via a C-terminal cysteine provided on the protein as described above. This allows for efficient coupling of the cysteine thiol to electrophilic groups, e.g. epoxide groups, halohydrin groups etc. on a support, resulting in a thioether bridge coupling.

In certain embodiments the support comprises a polyhydroxy polymer, such as a

polysaccharide. Examples of polysaccharides include e.g. dextran, starch, cellulose, pullulan, agar, agarose etc. Polysaccharides are inherently hydrophilic with low degrees of nonspecific interactions, they provide a high content of reactive (activatable) hydroxyl groups and they are generally stable towards alkaline cleaning solutions used in bioprocessing. In some embodiments the support comprises agar or agarose. The supports used in the present invention can easily be prepared according to standard methods, such as inverse suspension gelation (S Hjerten: Biochim Biophys Acta 79(2), 393-398 (1964). Alternatively, the base matrices are commercially available products, such as crosslinked agarose beads sold under the name of SEPHAROSE™ FF (GE Healthcare). In an embodiment, which is especially advantageous for large-scale separations, the support has been adapted to increase its rigidity using the methods described in US6602990 or US7396467, which are hereby incorporated by reference in their entireties, and hence renders the matrix more suitable for high flow rates.

In certain embodiments the support, such as a polymer, polysaccharide or agarose support, is crosslinked, such as with hydroxyalkyl ether crosslinks. Crosslinker reagents producing such crosslinks can be e.g. epihalohydrins like epichlorohydrin, diepoxides like butanediol diglycidyl ether, allylating reagents like allyl halides or allyl glycidyl ether. Crosslinking is beneficial for the rigidity of the support and improves the chemical stability. Hydroxyalkyl ether crosslinks are alkali stable and do not cause significant nonspecific adsorption. Alternatively, the porous support is based on synthetic polymers, such as polyvinyl alcohol, polyhydroxyalkyl acrylates, polyhydroxyalkyl methacrylates, polyacrylamides,

polymethacrylamides etc. In case of hydrophobic polymers, such as matrices based on divinyl and monovinyl-substituted benzenes, the surface of the matrix is often hydrophilised to expose hydrophilic groups as defined above to a surrounding aqueous liquid. Such polymers are easily produced according to standard methods, see e.g. "Styrene based polymer supports developed by suspension polymerization" (R Arshady: Chimica e L'lndustria 70(9), 70-75 (1988)). Alternatively, a commercially available product, such as SOURCE™ (GE Healthcare) is used. In another alternative, the porous support according to the invention comprises a support of inorganic nature, e.g. silica, zirconium oxide etc. In yet another embodiment, the solid support is in another form such as a surface, a chip, capillaries, or a filter (e.g. a membrane or a depth filter matrix).

As regards the shape of the matrix according to the invention, in one embodiment the matrix is in the form of a porous monolith. In an alternative embodiment, the matrix is in beaded or particle form that can be porous or non-porous. Matrices in beaded or particle form can be used as a packed bed or in a suspended form. Suspended forms include those known as expanded beds and pure suspensions, in which the particles or beads are free to move. In case of monoliths, packed bed and expanded beds, the separation procedure commonly follows conventional chromatography with a concentration gradient. In case of pure suspension, batch-wise mode will be used. The separation matrix as disclosed above has excellent alkali stability and may be stored in an alkaline storage liquid. The method of storing the separation matrix comprises the following steps: a) providing a storage liquid comprising at least 50% by volume of an aqueous alkali metal hydroxide solution; b) permeating the separation matrix with the storage liquid; and c) storing the storage liquid-permeated separation matrix for a storage time of at least 5 days.

By using a storage liquid comprising aqueous alkali metal hydroxide solution, a bacteriostatic or bactericidal solution may be obtained without requiring the use of alcohols such as ethanol, isopropanol, or benzyl alcohol. This means that a storage liquid may be used that is cheaper, subject to less regulatory burden, non-flammable and easier to dispose of than the storage solutions presently used for known protein A affinity separation matrices.

The separation matrix is a separation matrix as disclosed above, comprising multimers of immunoglobulin-binding alkali-stabilized Protein A domains covalently coupled to a porous support, wherein the alkali-stabilized Protein A domains comprise mutants of a parental Fc- binding domain of Staphylococcus Protein A (SpA), as defined by, or having at least 80% such as at least 90%, 95% or 98% identity to, SEO ID NO 51 or SEO ID NO 52, wherein the amino acid residues at positions 13 and 44 of SEO ID NO 51 or 52 are asparagines, and wherein at least the asparagine residue at position 3 of SEO ID NO 51 or 52 has been mutated to an amino acid selected from the group consisting of glutamic acid, lysine, tyrosine, threonine, phenylalanine, leucine, isoleucine, tryptophan, methionine, valine, alanine, histidine and arginine. The immunoglobulin-binding alkali-stabilized Protein A domains may comprise further mutations. For example, the glutamine residue at position 1 of SEO ID NO 51 or 52 may be mutated to an alanine; and/or the asparagine or glutamic acid residue at position 35 of SEQ. ID NO 51 or 52 may be mutated to an alanine.

The storage liquid comprises at least 50% by volume of an aqueous alkali metal hydroxide solution. The aqueous alkali metal hydroxide solution may comprise a single alkali metal hydroxide or a mixture of alkali metal hydroxides, such as sodium hydroxide, potassium hydroxide, or a mixture of sodium hydroxide and potassium hydroxide. The aqueous alkali metal hydroxide solution may have a molarity of from 10 mM to 100 mM, such as from 30 mM to 50 mM, expressed as the total combined concentration of alkali metal hydroxides if a mixture of alkali metal hydroxides is used. The storage liquid may essentially consist of, or consist of, the aqueous alkali metal hydroxide solution. However, the storage liquid may in some embodiments also comprise further components. Such further components may include alcohols, such as a C ₂ - C ₇ alcohol, such as ethanol, isopropanol or benzyl alcohol. Such further components may include salts, such as sodium chloride. The use of alcohols and/or salts in the storage liquid may increase the efficacy of the storage liquid in inhibiting or inactivating certain microorganisms, such as spore-forming bacteria.

Non-limiting examples of storage liquids include:

Sodium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M);

Potassium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M);

Sodium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 10-20% by volume ethanol; Sodium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 10-50% by volume isopropanol;

Sodium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 1-5% by volume benzyl alcohol;

Potassium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 10-20% by volume ethanol;

Potassium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 10-50% by volume isopropanol; or Potassium hydroxide solution (0.01 M, 0.03 M, 0.05 M, 0.1 M) with 1-5% by volume benzyl alcohol.

The separation matrix is permeated with the storage liquid prior to and during storage. By permeated with storage liquid, it is meant that the pores and interstices of the separation matrix are to a large extent filled with storage liquid. The separation matrix should be permeated with a quantity of storage liquid sufficient to inhibit growth of microorganisms in the stored separation matrix. The separation matrix may be impregnated, saturated with, or immersed in the storage liquid. Typically, a slurry of separation matrix in storage liquid suitable for storage may comprise about 50 % to 80% by weight of separation matrix, relative to the total weight of the slurry.

The separation matrix may be stored in the storage liquid for as extended a period as required. Typically, if the separation matrix is to be stored, this is for at least 5 days, often for at least 10 days, such as at least 50 days, or such as at least 100 days, or such as at least 200 days. The maximum storage time, or shelf life, of the separation matrix depends on the nature of the storage liquid used, i.e. alkali concentration, as well as the degree of capacity loss acceptable to the user, but may for example be up to 400 days or up to 700 days.

The mixture of storage liquid and separation matrix is contained in a suitable storage receptacle. The storage receptacle may be a bottle, can or drum made from a liquid- impervious material such as plastic, e.g. polyethylene, or glass. The separation matrix may be packaged in such storage receptacles for initial storage and distribution after production, or may be re-filled into such storage receptacles after use in purifying an immunoglobulin. The storage receptacle may alternatively be a pre-packed product for use in development or manufacturing of immunoglobulins. Such pre-packed products include filter plates, such as 96- well filter plates, and pre-packed columns. Such pre-packed columns include columns of all sizes known to the skilled person, from laboratory scale to process scale. Such columns can be shipped prepacked, qualified and sanitized, thus substantially reducing the time required for immunoglobulin purification processes.

The storage receptacle for storing the separation matrix may be open, vented or sealed. Since aqueous alkali metal hydroxide solutions are non-flammable and relatively non-volatile, no special considerations must be given regarding pressure build-up in the storage receptacle and ventilation of the storage room. In order to prevent dry-out of the separation media, it is preferable if the storage media is stored in an airtight receptacle.

The separation matrix may be stored at any temperature known in the art for storage of affinity media, such as from 1 °C to 30 °C, or from 10 to 20 °C. However, prolonged storage at elevated temperatures may degrade the capacity of the separation matrix, and therefore it is preferable if the separation matrix can be stored at a temperature of from 2 to 8 °C.

If the separation matrix has been previously used for purifying an immunoglobulin prior to storage, e.g. if it has been used in a production campaign that has now concluded, it is preferable that the separation matrix is cleaned and/or sanitized prior to storage. Cleaning liquids comprising at least 50% by volume of an aqueous alkali metal hydroxide solution and having a molarity of 0.5 M to 5 M may suitably be used to clean and/or sanitize the separation matrix.

Examples M utagenesis of protein

Site-directed mutagenesis was performed by a two-step PCR using oligonucleotides coding for the mutations. As template a plasmid containing a single domain of either Z, B or C was used. The PCR fragments were ligated into an E. coli expression vector. DNA sequencing was used to verify the correct sequence of inserted fragments. To form multimers of mutants an Acc I site located in the starting codons (GTA GAC) of the B, C or Z domain was used, corresponding to amino acids VD. The vector for the monomeric domain was digested with Acc I and phosphatase treated. Acc I sticky-ends primers were designed, specific for each variant, and two overlapping PCR products were generated from each template. The PCR products were purified and the concentration was estimated by comparing the PCR products on a 2% agarose gel. Equal amounts of the pair wise PCR products were hybridized (90°C -> 25°C in 45min) in ligation buffer. The resulting product consists approximately to ¼ of fragments likely to be ligated into an Acc I site (correct PCR fragments and/or the digested vector). After ligation and transformation colonies were PCR screened to identify constructs containing the desired mutant. Positive clones were verified by DNA sequencing.

Construct expression and purification

The constructs were expressed in the bacterial periplasm by fermentation of E. coli K12 in standard media. After fermentation the cells were heat-treated to release the periplasm content into the media. The constructs released into the medium were recovered by microfiltration with a membrane having a 0.2 μιη pore size.

Each construct, now in the permeate from the filtration step, was purified by affinity. The permeate was loaded onto a chromatography medium containing immobilized IgG (IgG Sepharose 6FF, GE Healthcare). The loaded product was washed with phosphate buffered saline and eluted by lowering the pH.

The elution pool was adjusted to a neutral pH (pH 8) and reduced by addition of dithiothreitol. The sample was then loaded onto an anion exchanger. After a wash step the construct was eluted in a NaCI gradient to separate it from any contaminants. The elution pool was concentrated by ultrafiltration to 40-50 mg/ml. It should be noted that the successful affinity purification of a construct on an immobilized IgG medium indicates that the construct in question has a high affinity to IgG.

The purified ligands were a nalyzed with RPC LC-MS to determine the purity and to ascertain that the molecular weight corresponded to the expected (based on the amino acid sequence).

Example 1

The purified monomeric ligands listed in Table 1, further comprising for SEQ. I D NO 8-16, 23-28 and 36-48 an AQGT leader sequence at the N-terminus and a cysteine at the C terminus, were immobilized on Biacore CM5 sensor chips (GE Healthcare, Sweden), using the amine coupling kit of GE Healthcare (for carbodiimide coupling of amines on the carboxymethyl groups on the chip) in an amount sufficient to give a signal strength of about 200-1500 RU in a Biacore surface plasmon resonance (SPR) instrument (GE Healthcare, Sweden) . To follow the IgG binding capacity of the immobilized surface lmg/ml human polyclonal IgG (Gammanorm) was flowed over the chip and the signal strength (proportional to the amount of binding) was noted. The surface was then cleaned-in-place (CIP), i.e. flushed with 500mM NaOH for 10 minutes at room temperature (22 +/- 2°C). This was repeated for 96-100 cycles and the immobilized ligand alkaline stability was followed as the remaining IgG binding capacity (signal strength) after each cycle. The results are shown in Table 1 and indicate that at least the ligands Zvar(NllK)l, Zvar(NllE)l, Zvar(NllY)l, Zvar(NllT)l, Zvar(NllF)l, Zvar(NllL)l, Zvar(NllW)l, ZN11I)1, Zvar(NllM)l, Zvar(NllV)l, Zvar(NllA)l, Zvar(NllHl), Zvar(NllR)l, Zvar(NllE,Q32A)l, Zvar(NllE,Q32E,Q40E)l and Zvar(NllE,Q32E,K50R)l,

Zvar(Q9A,NllE,N43A)l, Zvar(Q9A,NllE,N28A,N43A)l,

Zva r( Q9 A, N 11 E, Q40 V, A42 K, N 43 E, L441 ) 1, Zva r( Q9 A, N 11 E, Q40 V, A42 K, N43 A, L441 ) 1,

Zva r( Q9 A, N 11 E, N 28 A, Q40 V, A42 K, N43 A, L44I ) 1,

Zvar(NllK,H18K,S33K,D37E,A42R,N43A,L44l,K50R,L51Y)l,

Zvar(Q9A,NllK,H18K,S33K,D37E,A42R,N43A,L44l,K50R,L51Y)l, Zvar(NllK, H18K, D37E, A42R, N43A, L44l)l, Zvar(Q9A, NllK, H18K, D37E, A42R, N43A, L44l)l and Zvar(Q9A, NllK, H18K, D37E, A42R, N43A, L44I, K50R)1, as well as the varieties of

Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)l having G,S,Y,QJ,N,F,L,W,I,MN,D,E,H,R or K in position 29, the varieties of Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)l having F,Y,W,K or R in position 53 and the varieties of Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)l where Q9, Q40, A42 or N43 has been deleted, have an improved alkali stability compared to the parental structure Zvarl, used as the reference. Further, the ligands B(Q9A,N11E,Q40V,A42K,N43A,L44I)1 and C(Q9A,N11E,E43A)1 have an improved stability compared to the parental B and C domains, used as references.

Table 1. Monomeric ligands, evaluated by Biacore (0.5 M NaOH).

Ligand Sequence Capacity Reference Capacity after 96- capacity relative 100 after 96- to cycles 100 cycles reference

Zvar(Nl lE,Q32A)l SEQ ID 57% 55% 1.036

NO 12

Zvar(Nl lE)l SEQ ID 59% 55% 1.073

NO 13

Zvar(Nl 1E,Q32E,Q40E)1 SEQ ID 52% 51% 1.020

NO 14

Zvar(Nl 1E,Q32E,K50R)1 SEQ ID 53% 51% 1.039

NO 15

Zvar(Nl lK)l SEQ ID 62% 49% 1.270

NO 16

Zvar(Nl lY)l SEQ ID 55% 46% 1.20

NO 38

Zvar(Nl lT)l SEQ ID 50% 46% 1.09

NO 39

Zvar(Nl lF)l SEQ ID 55% 46% 1.20

NO 40

Zvar(Nl lL)l SEQ ID 57% 47% 1.21

NO 41

Zvar(Nl lW)l SEQ ID 57% 47% 1.21

NO 42

Zvar(Nl lI)l SEQ ID 57% 47% 1.21

NO 43

Zvar(Nl lM)l SEQ ID 58% 46% 1.26

NO 44

Zvar(Nl lV)l SEQ ID 56% 46% 1.22

NO 45

Zvar(Nl lA)l SEQ ID 58% 46% 1.26

NO 46

Zvar(Nl lH)l SEQ ID 57% 46% 1.24

NO 47

Zvar(Nl lR)l SEQ ID 59% 46% 1.28

NO 48

Zvar(Q9A,N 11 E,N43 A) 1 SEQ ID 70% 47% 1.49

NO 8 Zvar(Q9A,Nl 1E,N28A,N43A)1 SEQ ID 68% 47% 1.45

NO 9

Zvar(Q9A,N 11 E,Q40V,A42K,N43E,L44I) 1 SEQ ID 67% 47% 1.43

NO 10

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I) 1 SEQ ID 66% 47% 1.40

NO 11

Zvar(Q9 A,N 11 E,N28 A,Q40 V, A42K,N43 A,L44I) 1 SEQ ID 65% 48% 1.35

NO 24

Zvar(Nl lK,H18K,S33K,D37E,A42R,N43A,L44I,K50R,L51Y)l SEQ ID 67% 46% 1.46

NO 23

Zvar(Q9 A,N 11 K,H 18K,S33 K,D37E, A42R,N43 A,L44I,K50R,L51 Y) 1 SEQ ID 59% 46% 1.28

NO 25

Zvar(Nl IK, H18K, D37E, A42R, N43A, L44I)1 SEQ ID 59% 45% 1.31

NO 26

Zvar(Q9A, Nl IK, H18K, D37E, A42R, N43A, L44I)1 SEQ ID 63% 45% 1.40

NO 27

Zvar(Q9A, Nl IK, H18K, D37E, A42R, N43A, L44I, K50R)1 SEQ ID 67% 45% 1.49

NO 28

B(Q9A,N11E,Q40V,A42K,N43A,L44I)1 SEQ ID 39% 35% 1.11

NO 36

C(Q9A,N11E,E43A)1 SEQ ID 60% 49% 1.22

NO 37

Zvar(Q9A,Nl 1E,A29G,Q40V,A42K,N43 A,L44I) 1 SEQ ID 69% 48% 1.44

NO 54

Zvar(Q9A,Nl 1E,A29S,Q40V,A42K,N43 A,L44I) 1 SEQ ID 66% 48% 1.38

NO 55

Zvar(Q9A,Nl 1E,A29Y,Q40V,A42K,N43 A,L44I) 1 SEQ ID 61% 48% 1.27

NO 56

Zvar(Q9A,Nl 1E,A29Q,Q40V,A42K,N43 A,L44I) 1 SEQ ID 60% 47% 1.28

NO 57

Zvar(Q9A,Nl 1E,A29T,Q40V,A42K,N43 A,L44I) 1 SEQ ID 60% 47% 1.28

NO 58

Zvar(Q9A,Nl 1E,A29N,Q40V,A42K,N43 A,L44I) 1 SEQ ID 61% 47% 1.30

NO 59

Zvar(Q9A,Nl 1E,A29F,Q40V,A42K,N43 A,L44I) 1 SEQ ID 62% 46% 1.35

NO 60

Zvar(Q9A,Nl 1E,A29L,Q40V,A42K,N43 A,L44I) 1 SEQ ID 61% 46% 1.33

NO 61

Zvar(Q9A,Nl 1E,A29W,Q40V,A42K,N43 A,L44I) 1 SEQ ID 60% 46% 1.30

NO 62 Zvar(Q9A,Nl 1E,A29I,Q40V,A42K,N43 A,L44I) 1 SEQ ID 58% 47% 1.23

NO 63

Zvar(Q9A,Nl 1E,A29M,Q40V,A42K,N43 A,L44I) 1 SEQ ID 62% 47% 1.32

NO 64

Zvar(Q9A,Nl 1E,A29V,Q40V,A42K,N43 A,L44I) 1 SEQ ID 62% 47% 1.32

NO 65

Zvar(Q9A,Nl 1E,A29D,Q40V,A42K,N43 A,L44I) 1 SEQ ID 56% 47% 1.19

NO 66

Zvar(Q9A,Nl 1E,A29E,Q40V,A42K,N43 A,L44I) 1 SEQ ID 57% 47% 1.21

NO 67

Zvar(Q9A,Nl 1E,A29H,Q40V,A42K,N43 A,L44I) 1 SEQ ID 57% 47% 1.21

NO 68

Zvar(Q9A,Nl 1E,A29R,Q40V,A42K,N43 A,L44I) 1 SEQ ID 58% 46% 1.26

NO 69

Zvar(Q9A,Nl 1E,A29K,Q40V,A42K,N43 A,L44I) 1 SEQ ID 59% 46% 1.28

NO 70

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I,D53F) 1 SEQ ID 58% 46% 1.26

NO 71

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I,D53Y) 1 SEQ ID 59% 46% 1.28

NO 72

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I,D53W) 1 SEQ ID 62% 46% 1.35

NO 73

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I,D53K) 1 SEQ ID 65% 46% 1.41

NO 74

Zvar(Q9A,N 11 E,Q40V,A42K,N43 A,L44I,D53R) 1 SEQ ID 60% 46% 1.30

NO 75

Zvar(Q9 del,N 11 E,Q40 V,A42K,N43 A,L44I) 1 SEQ ID 60% 46% 1.30

NO 76

Zvar(Q9A,Nl lE,Q40del,A42K,N43 A,L44I) 1 SEQ ID 59% 46% 1.28

NO 77

Zvar(Q9A,N 11 E,Q40V,A42del,N43 A,L44I) 1 SEQ ID 57% 46% 1.24

NO 78

Zvar(Q9A,Nl lE,Q40V,A42K,N43del,L44I)l SEQ ID 55% 46% 1.20

NO 79

The Biacore experiment can also be used to determine the binding and dissociation rates between the ligand and IgG. This was used with the set-up as described above and with an IgGl monoclonal antibody as probe molecule. For the reference Zvarl, the on-rate (10 ⁵ M ¹ was 3.1 and the off-rate (10 ⁵ s ¹ ) was 22.1, giving an affinity (off-rate/on-rate) of 713 pM. For Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)l (SEQ. ID NO. 11), the on-rate was 4.1 and the off-rate 43.7, with affinity 1070 pM. The IgG affinity was thus somewhat higher for the mutated variant.

Example 2

The purified dimeric, tetrameric and hexameric ligands listed in Table 2 were immobilized on Biacore CM5 sensor chips (GE Healthcare, Sweden), using the amine coupling kit of GE Healthcare (for carbodiimide coupling of amines on the carboxymethyl groups on the chip) in an amount sufficient to give a signal strength of about 200-1500 RU in a Biacore instrument (GE Healthcare, Sweden) . To follow the IgG binding capacity of the immobilized surface lmg/ml human polyclonal IgG (Gammanorm) was flowed over the chip and the signal strength (proportional to the amount of binding) was noted. The surface was then cleaned-in-place (CIP), i.e. flushed with 500mM NaOH for 10 minutes at room temperature (22 +/- 2°C). This was repeated for 300 cycles and the immobilized ligand alkaline stability was followed as the remaining IgG binding capacity (signal strength) after each cycle. The results are shown in Table 2 and in Fig. 2 and indicate that at least the ligands Zvar(Q9A,NllE,N43A)4,

Zva r( Q9 A, N 11 E, N 28 A, N 43 A)4, Zva r( Q9 A, N 11 E, Q40 V, A42 K, N43 E, L441 )4 a n d

Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)4, Zvar(Q9A,NllE,D37E,Q40V,A42K,N43A,L44l)4 and Zvar(Q9A,NllE,D37E,Q40V,A42R,N43A,L44l)4 have an improved alkali stability compared to the parental structure Zvar4, which was used as a reference. The hexameric ligand

Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l)6 also has improved alkali stability compared to the parental structure Zvar6, used as a reference. Further, Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l) dimers with deletions of a) D2,A3,K4; b) K58,V1,D2; c) P57,K58,V1,D2,A3; d) K4,F5,D6,K7,E8; e) A56,P57,K58; V1,D2,A3 or f) V1,D2,A3,K4,F5,D6,K7,E8 from the linker region between the two monomer units have improved alkali stability compared to the parental structure Zvar2, used as a reference. Also Zvar(Q9A,NllE,Q40V,A42K,N43A,L44l) dimers with an insertion of YEDG between K58 and VI in the linker region have improved alkali stability compared to Zvar2. Table 2. Dimeric, tetrameric and hexameric ligands, evaluated by Biacore (0.5M NaOH).

Ligand SEQ Remaining Capacity Remaining Capacity Remaining Capacity

ID capacity relative capacity relative capacity relative

NO: 100 cycles to ref. 200 cycles to ref. 300 cycles to ref.

(%) 100 (%) 200 (%) 300 cycles cycles cycles

Zvar4 21 67 1 36 1 16 1

Zvar(Q9A,Nl 1E,N43A)4 17 81 1.21 62 1.72 41 2.56

Zvar(Q9A,Nl 1E,N28A,N 18 80 1.19 62 1.72 42 2.62 43A)4

Zvar(Q9A,Nl 1E,Q40V,A 19 84 1.25 65 1.81 48 3.00 42K,N43E,L44I)4

Zvar(Q9A,Nl 1E,Q40V,A 20 90 1.34 74 2.06 57 3.56 42K,N43A,L44I)4

Zvar(Q9A,Nl 1E,N28A,Q 32 84 1.24 Not tested Not Not tested Not 40V,A42K,N43A,L44I)4 tested tested

Zvar(Q9A,Nl 1E,Q40V,A 33 87 1.30 Not tested Not Not tested Not 42K,N43A,L44I)6 tested tested

Zvar(Q9A,Nl lE,D37E,Q 34 81 1.13 Not tested Not Not tested Not 40V,A42K,N43A,L44I)4 tested tested

Zvar(Q9A,Nl lE,D37E,Q 35 84 1.17 Not tested Not Not tested Not 40V,A42R,N43A,L44I)4 tested tested

Zvar(Q9A,Nl 1E,Q40V,A 80 70 1.27 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested D2, A3 and K4 in linker

deleted

Zvar(Q9A,Nl 1E,Q40V,A 81 76 1.38 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested K58, VI and D2 in linker

deleted

Zvar(Q9A,Nl 1E,Q40V,A 82 74 1.35 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested P57, K58, Vl, D2 and A3

in linker deleted

Zvar(Q9A,Nl 1E,Q40V,A 83 70 1.30 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested K4, F5, D6, K7 and E8 in

linker deleted Zvar(Q9A,Nl 1E,Q40V,A 84 68 1.26 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested A56, P57 and K58 in

linker deleted

Zvar(Q9A,Nl 1E,Q40V,A 85 75 1.39 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested VI, D2 and A3 in linker

deleted

Zvar(Q9A,Nl 1E,Q40V,A 86 62 1.13 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested VI, D2, A3, K4, F5, D6,

K7 and E8 in linker

deleted

Zvar(Q9A,Nl 1E,Q40V,A 87 72 1.31 Not tested Not Not tested Not 42K,N43A,L44I)2 with tested tested YEDG inserted in linker

between K58 and VI

Zvar2 88 55 1 Not tested Not Not tested Not tested tested

Example 3

Example 2 was repeated with 100 CIP cycles of three ligands using 1 M NaOH instead of 500 mM as in Example 2. The results are shown in Table 3 and show that all three ligands have an improved alkali stability also in 1M NaOH, compared to the parental structure Zvar4 which was used as a reference.

Table 3. Tetrameric ligands, evaluated by Biacore (1M NaOH).

Ligand Sequence Remaining Capacity

capacity 100 relative to cycles (%) ref. 100 cycles

Zvar4 SEQ ID NO 21 27 1

Zvar(Q9A,Nl 1E,N28A,N43A)4 SEQ ID NO 18 55 2.04

Zvar(Q9A,Nl lE,Q40V,A42K,N43E,L44I)4 SEQ ID NO 19 54 2.00

Zvar(Q9A,Nl lE,Q40V,A42K,N43A,L44I)4 SEQ ID NO 20 56 2.07 Example 4

The purified tetrameric ligands of Table 2 (all with an additional N-terminal cysteine) were immobilized on agarose beads using the methods described below and assessed for capacity and stability. The results are shown in Table 4 and Fig. 3. Table 4. Matrices with tetrametric ligands, evaluated in columns (0.5 M NaOH).

Activation

The base matrix used was rigid cross-linked agarose beads of 85 micrometers (volume- weighted, d50V) median diameter, prepared according to the methods of US6602990, hereby incorporated by reference in its entirety, and with a pore size corresponding to an inverse gel filtration chromatography Kav value of 0.70 for dextran of Mw 110 kDa, according to the methods described in Gel Filtration Principles and Methods, Pharmacia LKB Biotechnology 1991, pp 6-13. 25 mL (g) of drained base matrix, 10.0 mL distilled water and 2.02 g NaOH (s) was mixed in a 100 mL flask with mechanical stirring for 10 min at 25°C. 4.0 mL of epichlorohydrin was added and the reaction progressed for 2 hours. The activated gel was washed with 10 gel sediment volumes (GV) of water. Coupling

To 20 mL of ligand solution (50 mg/mL) in a 50 ml Falcon tube, 169 mg NaHC03, 21 mg Na ₂ C03, 175 mg NaCI and 7 mg EDTA, was added. The Falcon tube was placed on a roller table for 5-10 min, and then 77 mg of DTE was added. Reduction proceeded for >45 min. The ligand solution was then desalted on a PD10 column packed with Sephadex G-25. The ligand content in the desalted solution was determined by measuring the 276 nm UV absorption.

The activated gel was washed with 3-5 GV {0.1 M phosphate/1 mM EDTA pH 8.6} and the ligand was then coupled according to the method described in US6399750, hereby

incorporated by reference in its entirety. All buffers used in the experiments had been degassed by nitrogen gas for at least 5-10 min. The ligand content of the gels could be controlled by varying the amount and concentration of the ligand solution.

After immobilization the gels were washed 3xGV with distilled water. The gels + 1 GV {0.1 M phosphate/1 mM EDT A/10% thioglycerol pH 8.6} was mixed and the tubes were left in a shaking table at room temperature overnight. The gels were then washed alternately with 3xGV {0.1 M TRIS/0.15 M NaCI pH 8.6} and 0.5 M HAc and then 8-10xGV with distilled water. Gel samples were sent to an external laboratory for amino acid analysis and the ligand content (mg/ml gel) was calculated from the total amino acid content.

Protein

Gammanorm 165 mg/ml (Octapharma), diluted to 2mg/ml in Equilibration buffer. Equilibration buffer

PBS Phosphate buffer 10 mM + 0.14 M NaCI + 0.0027 M KCI, pH 7,4 (Medicago)

Adsorption buffer

PBS Phosphate buffer 10 mM + 0.14 M NaCI + 0.0027 M KCI, pH 7,4 (Medicago) Elution buffers

100 mM acetate pH 2.9

Dynamic binding capacity 2 ml of resin was packed in TRICORN™ 5 100 columns. The breakthrough capacity was determined with an AKTAExplorer 10 system at a residence time of 6 minutes (0.33 ml/min flow rate). Equilibration buffer was run through the bypass column until a stable baseline was obtained. This was done prior to auto zeroing. Sample was applied to the column until a 100% UV signal was obtained. Then, equilibration buffer was applied again until a stable baseline was obtained.

Sample was loaded onto the column until a UV signal of 85% of maximum absorbance was reached. The column was then washed with 5 column volumes (CV) equilibration buffer at flow rate 0.5ml/min. The protein was eluted with 5 CV elution buffer at a flow rate of 0.5 ml/min. Then the column was cleaned with 0.5M NaOH at flow rate 0.2 ml/min and re- equilibrated with equilibration buffer.

For calculation of breakthrough capacity at 10%, the equation below was used. That is the amount of IgG that is loaded onto the column until the concentration of IgG in the column effluent is 10% of the IgG concentration in the feed.

= 100% UV signal;

= absorbance contribution from non-binding IgG subclass;

A(V) = absorbance at a given applied volume;

= column volume;

V; app = volume applied until 10% breakthrough;

V. sys = system dead volume;

Co = feed concentration. The dynamic binding capacity (DBC) at 10% breakthrough was calculated. The dynamic binding capacity (DBC) was calculated for 10 and 80% breakthrough.

CIP - 0.5 M NaOH

The 10% breakthrough DBC (QblO) was determined both before and after repeated exposures to alkaline cleaning solutions. Each cycle included a CIP step with 0.5 M NaOH pumped through the column at a rate of 0.5/min for 20 min, after which the column was left standing for 4 h. The exposure took place at room temperature (22 +/- 2°C). After this incubation, the column was washed with equilibration buffer for 20 min at a flow rate of 0.5 ml/min. Table 4 shows the remaining capacity after six 4 h cycles (i.e. 24 h cumulative exposure time to 0.5 M NaOH), both in absolute numbers and relative to the initial capacity.

Example 5

Example 4 was repeated with the tetrameric ligands shown in Table 5, but with 1.0 M used in the CIP steps instead of 0.5 M. The results are shown in Table 5 and in Fig. 4.

Table 5. Matrices with tetrametric ligands, evaluated in columns - 1.0 M NaOH.

Ligand SEQ Ligand Initial Remainin Remainin Capaci

ID content IgG g lgG g lgG t

NO. (mg/ml) capacity capacity capacity retenti

QblO QblO after six 4 on

(mg/ml) after six 4 h cycles relative h cycles (%) to ref.

(mg/ml) after six 4 h cycles

Zvar4 21 12 60.1 33.5 56 1

Zvar(Q9A,Nl lE,Q40V,A42K,N43A,L44I)4 20 12.8 60.3 56.0 93 1.67

Zvar(Nl 1 K,H 18K,S33K,D37E,A42R,N43 A 30 9.7 62.1 48.1 77 1.44

,L44I,K50R,L51Y)4 Example 6 Base matrices

The base matrices used were a set of rigid cross-linked agarose bead samples of 59-93 micrometers (volume-weighted, d50V) median diameter (determined on a Malvern

Mastersizer 2000 laser diffraction instrument), prepared according to the methods of

US6602990 and with a pore size corresponding to an inverse gel filtration chromatography Kd value of 0.62-0.82 for dextran of Mw 110 kDa, according to the methods described above, using HR10/30 columns (GE Healthcare) packed with the prototypes in 0.2 M NaCI and with a range of dextran fractions as probe molecules (flow rate 0.2 ml/min). The dry weight of the bead samples ranged from 53 to 86 mg/ml, as determined by drying 1.0 ml drained filter cake samples at 105 °C overnight and weighing.

Table 6. Base matrix samples

Base matrix Kd d50v (μιη) Dry weight (mg/ml)

A18 0.704 59.0 56.0

A20 0.70 69.2 55.8

A27 0.633 87.2 74.2

A28 0.638 67.4 70.2

A29 0.655 92.6 57.5

A32 0.654 73.0 70.5

A33 0.760 73.1 55.5

A38 0.657 70.9 56.2

A39 0.654 66.0 79.1

A40 0.687 64.9 74.9

A41 0.708 81.7 67.0

A42 0.638 88.0 59.4

A43 0.689 87.5 77.0

A45 0.670 56.6 66.0

A52 0.620 53.10 63.70

A53 0.630 52.6 86.0

A54 0.670 61.3 75.3

A55 0.640 62.0 69.6

A56 0.740 61.0 56.0

A56-2 0.740 51.0 56.0

A62a 0.788 48.8 70.1

A62b 0.823 50.0 46.9

A63a 0.790 66.8 59.6

A63b 0.765 54.0 79.0

A65a 0.796 58.0 60.0

A65b 0.805 57.3 46.0

B5 0.793 69.0 84.4

CI 0.699 71.0 73.4

C2 0.642 66.5 81.1

C3 0.711 62.0 82.0 C4 0.760 62.0 82.0

H31 0.717 82.0 59.0

H35 0.710 81.1 61.0

H40 0.650 52.8 65.0

11 0.640 50.0 67.0

41 0.702 81.6 60.6

Coupling

100 ml base matrix was washed with 10 gel volumes distilled water on a glass filter. The gel was weighed (1 g = 1 ml) and mixed with 30 ml distilled water and 8.08 g NaOH (0.202 mol) in a 250 ml flask with an agitator. The temperature was adjusted to 27 +/- 2 °C in a water bath. 16 ml epichlorohydrin (0.202 mol) was added under vigorous agitation (about 250 rpm) during 90 +/- 10 minutes. The reaction was allowed to continue for another 80 +/- 10 minutes and the gel was then washed with >10 gel volumes distilled water on a glass filter until neutral pH was reached. This activated gel was used directly for coupling as below. To 16.4 mL of ligand solution (50 mg/mL) in a 50 ml Falcon tube, 139 mg NaHC0 ₃ , 17.4 mg Na ₂ C03, 143.8 mg NaCI and 141 mg EDTA, was added. The Falcon tube was placed on a roller table for 5-10 min, and then 63 mg of DTE was added. Reduction proceeded for >45 min. The ligand solution was then desalted on a PD10 column packed with Sephadex G-25. The ligand content in the desalted solution was determined by measuring the 276 nm UV absorption. The activated gel was washed with 3-5 GV {0.1 M phosphate/1 mM EDTA pH 8.6} and the ligand was then coupled according to the method described in US6399750 5.2.2, although with considerably higher ligand amounts (see below). All buffers used in the experiments had been degassed by nitrogen gas for at least 5-10 min. The ligand content of the gels was controlled by varying the amount and concentration of the ligand solution, adding 5-20 mg ligand per ml gel. The ligand was either a tetramer (SEQ. ID NO. 20) or a hexamer (SEQ. ID NO. 33) of an alkali-stabilized mutant.

After immobilization the gels were washed 3xGV with distilled water. The gels + 1 GV {0.1 M phosphate/1 mM EDTA/10% thioglycerol pH 8.6} was mixed and the tubes were left in a shaking table at room temperature overnight. The gels were then washed alternately with 3xGV {0.1 M TRIS/0.15 M NaCI pH 8.6} and 0.5 M HAc and then 8-10xGV with distilled water. Gel samples were sent to an external laboratory for amino acid analysis and the ligand content (mg/ml gel) was calculated from the total amino acid content.

Evaluation The QblO % dynamic capacity for polyclonal human IgG at 2.4 and 6 min residence time was determined as outlined in Example 4.

Table 7. Prototype results

Prototype Base matrix Ligand Multimer Qbl0% 2.4 Qbl0% 6 content min (mg/ml) min (mg/ml) (mg/ml)

Nl A38 7.45 tetramer 44.4 58.25

N2 A20 7.3 tetramer 45.12 57.21

N3 A42 6.72 tetramer 33.56 50.02

N4 A29 7.3 tetramer 36.34 51.8

N5 A28 7.9 tetramer 42.38 58.25

N6 A39 6.96 tetramer 41.88 54.67

N7 A27 7.5 tetramer 29.19 48.73

N8 A43 6.99 tetramer 33.43 49.79

N9 A38 11.34 tetramer 48.1 72.78

N10 A20 10.6 tetramer 50.66 70.07

Ni l A42 11.1 tetramer 32.25 57.78

N12 A29 11 tetramer 34.85 64.68

N13 A28 11.9 tetramer 39.92 63.75

N14 A39 10.48 tetramer 44.37 64.79

N15 A27 12.1 tetramer 24.8 55.56

N16 A43 10.51 tetramer 31.82 58.04

N17 A41 8.83 tetramer 38.5 56.8

N18 A41 8.83 tetramer 37.84 58.6

N19 A41 8.83 tetramer 35.06 57.23

N20 A41 5.0 tetramer 35.64 46.04

N21 A41 13.0 tetramer 34.95 62.23

N22 A40 13.15 tetramer 56.85 71.09

N23 A33 7.33 tetramer 48.69 55.76

N24 A40 11.03 tetramer 54.96 73.8

033A A38 7.5 tetramer 44 58

033B A38 11.3 tetramer 48 73

097A A20 7.3 tetramer 45 57

097B A20 10.6 tetramer 51 70 A A28 7.9 tetramer 42 58B A28 11.9 tetramer 40 64C A28 15.8 tetramer 37 67 A A39 7.0 tetramer 42 55B A39 10.5 tetramer 44 65

A40 13.2 tetramer 57 71

A33 7.3 tetramer 49 56 A A40 11.0 tetramer 55 74

A18 8.2 tetramer 52 59

A32 10.7 tetramer 40 57

A32 13.3 tetramer 37 66A B5 6.3 tetramer 32 38B B5 9.6 tetramer 45 47A CI 5.4 tetramer 38 47B CI 10.8 tetramer 43 60A C2 5.1 tetramer 39 46B C2 10.5 tetramer 42 57A H40 5.6 tetramer 47 52B H40 9.1 tetramer 52 67

A18 13.4 tetramer N/A 63

A20 12.8 tetramer 49 67

A33 9.4 tetramer 51 58

A40 6.1 tetramer 49 59A A62a 6.6 tetramer 41 46B A62a 14.8 tetramer 55 62

A62a 9.7 tetramer 48 60A A63a 6.6 tetramer 40 44B A63a 14.0 tetramer 48 56B A65a 13.1 tetramer 56 64A A62b 5.2 tetramer 40 N/AB A62b 11.1 tetramer 52 N/AA A65b 5.8 tetramer 42 46B A65b 8.8 tetramer 49 56 017A A65a 6.1 tetramer 40 44

387A A62a 6.4 tetramer 43 45

387B A62a 7.5 tetramer 47 56

432 A63a 6.1 tetramer 39 44

433A A65a 6.6 tetramer 42 47

433B A65a 13.6 tetramer 52 61

579A 11 6.1 tetramer 45 51

579B 11 11.2 tetramer 57 68

064A C3 5.9 tetramer 44 52

064B C3 9.0 tetramer 49 62

064C C3 14.3 tetramer 51 70

352A C4 10.1 tetramer 55 63

352B C4 14.4 tetramer 59 67

066A C3 6.8 hexamer 48 59

066B C3 11.9 hexamer 51 73

066C C3 15.1 hexamer 43 61

353A C4 11.2 hexamer 62 74

353B C4 15.2 hexamer 57 82

872A A62a 9.6 hexamer 56 72

872B A62a 14.5 hexamer 62 84

869A H40 6.9 hexamer 50 56

869B H40 14.3 hexamer 56 75

869C H40 23.0 hexamer 41 65

962A H35 6.8 hexamer 36 49

962B H35 12.3 hexamer 31 54

962C H35 20.3 hexamer 20 43

112A A56 7.9 hexamer 47 55

112B A56 12.4 hexamer 57 73

112C A56 19.2 hexamer 55 80

113A A56 7.1 hexamer 48 57

113B A56 12.4 hexamer 53 73

113C A56 15.2 hexamer 48 76

212A H31 6.5 hexamer 37 38 212B H31 10.4 hexamer 50 61

212C H31 20.0 hexamer 31 52

213A A33 6.5 hexamer 44 53

213B A33 10.9 hexamer 50 65

213C A33 11.1 hexamer 50 68

432A A20 6.4 hexamer 41 56

432B A20 12.4 hexamer 38 64

432C A20 21.1 hexamer 44 43

433A A38 5.9 hexamer 47 57

433B A38 11.6 hexamer 48 72

433C A38 15.8 hexamer 36 62

742A A54 6.7 hexamer 38 46

742B A54 12.6 hexamer 45 52

742C A54 21.1 hexamer 38 65

726A A63b 6.4 hexamer 42 46

726B A63b 10.6 hexamer 49 60

726C A63b 16.7 hexamer 53 69

793A A56-2 6.8 hexamer 50 58

793B A56-2 12.5 hexamer 59 72

793C A56-2 19.2 hexamer 61 82

Example 7

A series of prototypes, prepared as above, with different ligand content (tetramer, SEQ. ID NO:20) were incubated in 1 M NaOH for 4, 8 and 31 hours at 22 +/- 2 °C and the dynamic IgG capacity (Qbl0%, 6 min residence time) was measured before and after incubation. The prototypes are shown in Table 8 and the results in Figs. 5 and 6. It can be seen that the stability towards this harsh alkali treatment increases with increasing ligand content. Table 8. Samples for incubation in 1 M NaOH

Example 8

Two crosslinked agarose bead prototypes, prepared as above, with different ligand content (hexamer, SEQ. ID NO:33), median bead diameter (d50,v) 62 μιη and Kd 0.70 for dextran of Mw 110 kD, were evaluated with a real mAb feed. The ligand content of prototype A was 14.3 mg/ml and of prototype B 18.9 mg/ml. For comparison, the commercial product MabSelect SuRe ^® LX (GE Healthcare Life Sciences, with ligand SEQ. ID NO. 21) was used. The resins were packed in Tricorn columns (GE Healthcare Life Sciences) to bed heights of 10 cm, giving bed volumes of 2 ml and the columns were shown to have peak asymmetry within the 0.8-1.5 interval. The sample loaded was a clarified CHO cell supernatant with 4.9 mg/ml monoclonal IgGl antibody at physiological pH and the experimental conditions were as listed below in Table 9 (CV = column volumes, RT = residence time).

Table 9. Conditions for evaluation with real feed.

Equilibration: 3 CV 20 mM phosphate, 150 mM NaCl pH 7.4, RT=3.4 min

Sample loading: 43 mg mAb/ml resin, RT=6 min

Wash 1 : 5 CV 20 mM phosphate, 500 mM NaCl pH 7.4, 1.5 CV at RT=6min and

3.5 CV at RT=3.4 min

Wash 2: 1 CV 50 mM acetate pH 6.0, RT=3.4 min

Elution: 3 CV 50 mM acetate pH 3.5, RT=6 min, peak collected between 150 mAU- 150 mAU

Strip: 2 CV 100 mM acetate, RT=3.4 min

CIP: 3 CV 0.1 M NaOH, RT=6 min

Re-equilibration: 5 CV 20 mM phosphate, 150 mM NaCl pH 7.4, RT=3.4 min The imAb peak was collected using a UV watch function and the concentration of the mAb was determined by UV measurement at 280 nm (extinction coefficient 1.5). All absorbance detections were performed using a spectrophotometer, including the measurements for the yield calculations. Samples for HCP (host cell protein) analyses were prepared by adding 10% Preservation buffer (0.2 M NaH ₂ P0 ₄ *H ₂ 0 (5.3%), 0.2 M Na ₂ HP0 ₄ *12 H ₂ 0 (94.7%), 0.5% Tween 20, 1% BSA pH 8) to the samples directly after each run made (e.g. 50 μΙ preservation buffer to 450 μΙ sample). The HCP content was measured using commercial anti-CHO antibodies (Cygnus Technologies) and a Gyrolab (Gyros AB, Sweden) work station. The results are presented in Table 10 below and show that the performance of the prototypes is in the same range as for the commercial product. The HCP content in the feed was 331 000 ppm.

Table 10. Results from real feed evaluation

Example 9

A crosslinked agarose bead matrix prototype, prepared as above, with 14.5 mg/ml ligand (hexamer, SEQ. ID NO:33), median bead diameter (d50,v) 57.4 μιη, Kd 0.72 for dextran of Mw 110 kD and dry weight 70.3 mg/ml, was evaluated for elution pH with two real mAb feeds (mAbl 2.4 g/l and mAb2 4.9 g/l) IgGl, physiological pH, and a sample of polyclonal human IgG (Gammanorm, Octapharma). For comparison, the commercial product MabSelect SuRe ^® LX (GE Healthcare Life Sciences) was used. The resins were packed in Tricorn columns (GE Healthcare Life Sciences) to bed heights of 10 cm, giving bed volumes of 2 ml and the columns were shown to have peak asymmetry within the 0.8-1.5 interval. The samples loaded were clarified CHO cell supernatants with IgGl mAbs at physiological pH and the experimental conditions were as listed below in Table 11 (CV = column volumes, RT = residence time).

Table 11. Conditions for elution pH evaluation.

The results are shown below in Table 12 and indicate that the antibodies elute at simi levels as on the reference, although with some individual variation depending on the particular antibody-resin combination.

Table 12. Results from elution pH evaluation

Fractions from the pH-gradient elution of polyclonal IgG were also analysed with respect to content of IgGl, lgG2 and lgG4, using a Biacore SPR instrument (GE Healthcare Life Sciences) with antibodies against the four different IgG classes immobilized on a CM5 Biacore chip.

The chromatograms for polyclonal IgG on the reference and the prototype are shown in Fig 7 and the IgG class analyses are shown in Fig. 8. The data show that all three classes bind to both resins in a similar way and that the first peak predominantly contains lgG2, while IgGl and lgG4 elute mainly in the second peak. The anti-lgG3 antibodies cross-reacted with lgG4, so no reliable results for lgG3 were obtained. lgG3 is generally known to show no or only weak binding to Protein A.

Example 10 A crosslinked agarose bead matrix prototype, prepared as above, with 12.6 mg/ml ligand

(tetramer, SEQ. ID NO:20), 84.9 μιη median bead diameter (d50,v), Kd 0.71 for dextran Mw 110 kD and 62.2 mg/ml dry weight, was evaluated with respect to alkali stability, using the commercial product MabSelect SuRe LX as a reference. Tricorn 5 columns packed with the resins to 10 cm bed height were flushed with 3 column volumes of 1 M NaOH. The flow was then stopped for 240 minutes (corresponding to 16 normal CIP cycles of 15 min/cycle) before washing out the NaOH solution by 3 column volumes of PBS buffer. The dynamic binding capacity for polyclonal IgG (Gammanorm, Octapharma) was then measured and the process was repeated with another injection of 1 M NaOH. The dynamic capacity was measured after each 240 min incubation cycle with 1 M NaOH. In the capacity measurements, the columns were equilibrated with PBS buffer before the 2 mg/ml sample was loaded (residence time 6 min) until a UV signal of 85% of maximum absorbance was reached. Then the column was washed with PBS buffer, eluted with 500 mM acetic acid pH 3.0 and re-equilibrated. The dynamic binding capacity at 10% and 80% breakthrough was calculated as described above. The results are shown in Fig. 9 and they show that the prototype was significantly more stable than the commercial product.

Example 11

The long-term storage stability of a separation matrix according to the invention was assessed. The inventive example (Inv. Ex.) was compared to the commercial product MabSelect SuRe (MSS) (GE Healthcare Life Sciences, with ligand SEO ID NO. 21) as a reference. The separation matrix (Inv. Ex. or MSS) was incubated together with the storage liquid for a predetermined period (two weeks). The storage liquids tested were 20% ethanol solution; 0.01 M NaOH solution; 0.03 M NaOH solution and 0.05 M NaOH solution. After incubation with the storage liquid for the predefined period, the 10% breakthrough dynamic binding capacity at was determined using human polyclonal IgG as described in Example 4, using a residence time of 2.4 minutes. The 10% breakthrough capacity of the separation matrices prior to storage were also determined for comparison. The results are shown in Figure 10 and Table 13 below.

Table 10. Results from long term storage evaluation

It can be seen that the inventive example has a significantly higher initial dynamic binding capacity as compared to the commercial MabSelect SuRe. After four weeks storage in 50 mM NaOH solution MabSelect SuRe retains only approximately 52% of its original dynamic binding capacity, whereas the inventive example retains approximately 96% of its original dynamic binding capacity. Thus, it can be seen that the separation matrix of the inventive example is suitable for long-term storage in NaOH solution with concentrations up to at least 50 mM.