SE1451564A1 - Modified kappa light chain-binding polypeptides - Google Patents

Abstract

ABSTRACT The inVention discloses kappa light chain-binding polypeptide coniprising a niutated bindingdomain of Peptostreptococcus protein L, Wherein at least one asparagine residue of a parentaldomain defined by, or having at least 95% or 98% sequence honiology With, SEQ ID NO: 2-6 or 12 has been niutated to another an1ino acid residue Which is not asparagine, proline or cysteine.

Description

MODIFIED KAPPA LIGHT CHAIN-BINDING POLYPEPTIDES Technical field of the invention The present invention relates to the field of affinity chromatography, and more specif1cally topolypeptides comprising kappa light chain-binding domains of Protein L, Which are useful inaffinity chromatography of many types of immunoglobulins and immunoglobulin fragments. Theinvention also relates to separation matrices containing the polypeptides and to separation methods using such separation matrices.

Background of the invention Immunoglobulins and immunoglobulin fragments represent the most prevalentbiopharrnaceutical products in either manufacture or development Worldwide. The highcommercial demand for and hence value of this particular therapeutic market has led to theemphasis being placed on pharrnaceutical companies to maximize the productivity of their respective manufacturing processes Whilst controlling the associated costs.

Affinity chromatography, typically on matrices comprising staphylococcal Protein A orvariants thereof, is norrnally used as one of the key steps in the purif1cation of intactimmunoglobulin mo lecules. The highly selective binding of Protein A to the Fc chain ofimmunoglobulins provides for a generic step With very high clearance of impurities and contaminants.

For antibody fragments, such as Fab, single-chain variable fragments (scFv), bi-specificT-cell engagers (BiTEs), domain antibodies etc., Which lack the Fc chain but have a subclass 1,3or 4 kappa light chain, matrices comprising Protein L derived from Peptostreptococcus magnus(B Åkerström, L Björck: J. Biol. Chem. 264, 19740-19746, 1989; W Kastem et al: J. Biol.Chem. 267, 12820-12825, 1992; B H KNilson et al: J. Biol. Chem. 267, 2234-2239, 1992 andUS Pat. 6,822,075) show great promise as a purif1cation platform providing the high selectivityneeded. The Protein L disclosed in US Pat. 6,822,075 comprises the amino acid sequence SEQID NO: 1 plus an additional AVEN sequence at the N-terrninus.

SEQ ID NO: 1 (Protein L) KEETPETPETD SEEEVTIKAN LIFANGSTQT AEFKGTFEKA TSEAYAYADTLKKDNGEYTV DVADKGYTLN IKFAGKEKTPEE PKEEVTIKAN LIYADGKTQTAEFKGTFEEA TAEAYRYADA LKKDNGEYTV DVADKGYTLN IKFAGKEKTPEEPKEEVTIKAN LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKENGKYTVDVADKGYTLN IKFAGKEKTPEE PKEEVTIKAN LIYADGKTQT AEFKGTFAEATAEAYRYADL LAKENGKYTA DLEDGGYTIN IRFAGKKVDEKPEE Protein L matrices are commercially available as CaptoTM L from GE Healthcare Bio-SciencesAB, Sweden (Capto L data file 29-0100-08 AC, 2014) and can be used for separation of kappalight chain-containing proteins such as intact antibodies, Fab fragments, scFv fragments, domainantibodies etc. About 75% of the antibodies produced by healthy humans have a kappa lightchain and many therapeutic monoclonal antibodies and antibody fragments contain kappa light chains.

Any bioprocess chromatography application requires comprehensive attention to def1niteremoval of contaminants. Such contaminants can for example be non-eluted mo lecules adsorbedto the stationary phase or matrix in a chromatographic procedure, such as non-desiredbiomolecules or microorganisms, including for example proteins, carbohydrates, lipids, bacteriaand viruses. The removal of such contaminants from the matrix is usually performed after a firstelution of the desired product in order to regenerate the matrix before subsequent use. Suchremoval usually involves a procedure known as cleaning-in-place (CIP), Wherein agents capableof eluting contaminants from the stationary phase are used. One such class of agents often usedWith chromatography media is alkaline solutions that are passed over the matrix. At present themost extensively used cleaning and sanitizing agent is NaOH, and it is desirable to use it inconcentrations ranging from 0.05 up to e.g. 1 M, depending on the degree and nature ofcontamination. Protein L is however a rather alkali-sensitive protein compared to e.g. Protein Aand only tolerates up to about 15 mM NaOH over a large number of cycles. This means thatadditional, less desirable cleaning solutions, e.g. urea or guanidinium salts, may also have to be used in order to ensure sufficient cleaning.

An extensive research has earlier been focused on the development of engineered proteinA ligands that exhibit an improved capacity to Withstand alkaline pH-values. For example,WO2003/080655A1 disclo ses that Protein A domains With particular asparagine mutations areconsiderably more alkali stable than the native protein.

There is thus still a need in this field to obtain a separation matrix containing Protein L-derived ligands having an improved stability towards alkaline cleaning procedures.

Summarv of the invention One aspect of the invention is to provide a polypeptide With improved alkaline stability. This is achieved With a polypeptide as defined in claim 1.

One advantage is that the alkaline stability is improved over Protein L and the parentalpolypeptides. A further advantage is that the highly selective binding towards kappa light chain- containing proteins demonstrated for Protein L is retained in the polypeptides of the invention.

A second aspect of the invention is to provide a nucleic acid or a vector encoding a polypeptideor multimer With improved alkaline stability. This is achieved With a nucleic acid or vector as defined in the claims.

A third aspect of the invention is to provide an expression system capable of expressing apolypeptide or multimer With improved alkaline stability. This is achieved With an expression system as defined in the claims.A fourth aspect of the invention is to provide a separation matrix capable of selectively bindingkappa light chain-containing proteins and exhibiting an improved alkaline stability. This is achieved With a separation matrix as defined in the claims.

A fifth aspect of the invention is to provide an efficient and economical method of iso lating a kappa light chain-containing protein. This is achieved With a method as defined in the claims.

Further suitable embodiments of the invention are described in the dependent claims.

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 a “kappa light chain-binding polypeptide” and “kappa light chain-binding protein”herein mean a polypeptide or protein respectively, capable of binding to a subclass 1, 3 or 4kappa light chain of an antibody (also called VKI, VKHI and VKW, as in B H K Nilson et al: J. Biol.Chem. 267, 2234-2239, 1992), and include eg. Protein L, and any variant, fragment or fusion protein thereof that has maintained said binding property.

The term “kappa light chain-containing protein” is used as a synonym of “immunoglobulinkappa light chain-containing protein” and herein means a protein comprising a subclass 1, 3 or 4kappa light chain (also called VKI, VKHI and VKIV, as in B H K Nilson et al: J. Biol. Chem. 267,2234-2239, 1992) derived from an antibody and includes any intact antibodies, antibodyfragments, fusion proteins, conjugates or recombinant proteins containing a subclass 1, 3 or 4 kappa light chain.

Brief description of figures Fig. 1 shows an alignment of the five kappa light chain-binding domains of Protein L as described in US 6,822,075 and W Kastem et al: J Biol. Chem. 267, 12820-12825, 1992.

Fig. 2 shoWs the alkali stability of different kappa light chain-binding domains of Protein L.

Fig. 3 shoWs the alkali stability of mutated kappa light chain-binding domains of Protein L.

Fig. 4 shows the alkali stability of Protein L ligands comprising four domains.

Fig. 5 shoWs the alkali stability of mutated dimeric, tetrameric and hexameric kappa light chain- binding domains of Protein L in comparison With Protein L.

Detailed description of embodiments In one aspect the present invention discloses a kappa light chain-binding polypeptide comprisingor consisting essentially of one or more binding domains of Peptostreptococcus magnus ProteinL, Wherein each of these domains is selected from the group consisting of Domain 2, Domain 3and Domain 4. Domain 2 can have an amino acid sequence defined by SEQ ID NO:3 or SEQ IDNO: 12, or it can have at least 90%, such as at least 95%, sequence homology With SEQ IDNO:3 or 12. SEQ ID NO: 12 is a variant of SEQ ID NO: 3, With an alanine in position 31.4 Domain 3 can have an amino acid sequence defined by SEQ ID NO:4, or it can have at least90%, such as at least 95%, sequence homology With SEQ ID NO:4. Domain 4 can have anamino acid sequence defined by SEQ ID NO:5, or it can have at least 90%, such as at least 95%,sequence homology With SEQ ID NO:5.

In some embodiments of the polypeptide, each domain is selected from the group consisting ofDomain 3 and Domain 4, or each of the domains is Domain 3. Specifically, the polypeptide may comprise or consist essentially of a multimer of Domain 3.

In certain embodiments, at least tWo of the domains are selected from the group consisting of Domain 2, Domain 3 and Domain 4, or from the group consisting of Domain 3 and Domain 4.

In some embodiments, the polypeptide does not contain any Domain l of PeptostreptococcusProtein L. Domain l can have an amino acid sequence as defined by SEQ ID NO:2, or it can have at least 90%, such as at least 95% sequence homology With SEQ ID NO:2.

In certain embodiments of the polypeptide, at least the amino acid at the position correspondingto position 45 in SEQ ID NOi2-5 (e.g. the amino acid at position 45 in SEQ ID NO: 2-5 or l2) inone or more, such as all, of the binding domains has been mutated to an amino acid Which is not asparagine, proline or cysteine. The amino acid at position 45 can e.g. be mutated to an alanine.

In some embodiments of the polypeptide, at least the amino acid at the position corresponding toposition l0 in SEQ ID NOi2-5 (e.g. the amino acid at position l0 in SEQ ID NO: 2-5 or l2) inone or more, such as all, of the binding domains has been mutated to an amino acid Which is not asparagine, proline or cysteine. The amino acid at position l0 can e. g. be mutated to a glutamine.

In certain embodiments of the polypeptide, at least the amino acid at the position correspondingto position 60 in SEQ ID NOi2-5 (e.g. the amino acid at position 60 in SEQ ID NO: 2-5 or l2)in one or more, such as all, of the binding domains has been mutated to an amino acid Which is not asparagine, proline or cysteine. The amino acid at position 60 can e. g. be mutated to a glutamine.

Specifically, one or more, such as all, of the binding domains may have mutations selected from the group consisting of Nl0Q; N45A; N60Q; Nl0Q,N45A; N45A,N60Q, Nl0Q,N60Q and N10Q,N45A,N60Q, or altematively selected from the group consisting of N45A; N10Q,N45A;N45A,N60Q and N10Q,N45A,N60Q.

In some embodiments of the polypeptide, at least the amino acid at the position corresponding toposition 19 in SEQ ID NOi2-5 (e.g. the amino acid at position 19 in SEQ ID NO: 2-5 or 12) inone or more, such as all, of the binding domains has been mutated to an amino acid Which is notglutamine, asparagine, proline or cysteine. The amino acid at position 19 can e. g. be mutated to aglutamic acid or an alanine. Specifically, one or more, such as all, of the binding domains may have mutations selected from the group consisting of Q 19E and Q19A.

In certain embodiments of the polypeptide, one or more, such as all, of said binding domains hasan amino acid sequence selected from the group consisting of sequences defined by SEQ IDNO:7, SEQ ID NO:8, SEQ ID NO:9, SEQ ID NO:10, SEQ ID NO:11, SEQ ID NO:13 and SEQID NO: 14. One or more, such as all, of said binding domains can altematively have an aminoacid sequence selected from the group consisting of sequences defined by SEQ ID NO:7, SEQID NO:8, SEQ ID NO:9, SEQ ID NO: 10 and SEQ ID NO:11. The polypeptide may fiarther at theN-terrninus comprise a plurality of amino acid residues originating from the cloning process orconstituting a residue from a cleaved off signaling sequence. The number of additional aminoacid residues may e.g. be 15 or less, such as 10 or less or 5 or less. As a specific example, thepolypeptide may comprise an AQV sequence at the N-terrninus.

SEQ ID NO: 7 (Domain 3, N45A mutation)PKEEVTIKAN LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLN IKFAGKEKTPEE SEQ ID NO: 8 (Domain 3, N10Q,N45A mutation)PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLN IKFAGKEKTPEE SEQ ID NO: 9 (Domain 3,N45A,N60Q mutation)PKEEVTIKAN LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLQ IKFAGKEKTPEE SEQ ID NO: 10 (Domain 3, N10Q, N60Q mutation)PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKENGKYTVDVADKGYTLQ IKFAGKEKTPEE SEQ ID NO: 11 (Domain 3, N10Q,N45A,N60Q mutation)PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLQ IKFAGKEKTPEE SEQ ID NO: 12 (variant of Domain 2)PKEEVTIKAN LIYADGKTQT AEFKGTFEEA AAEAYRYADA LKKDNGEYTVDVADKGYTLN IKFAGKEKTPEE SEQ ID NO: 13 (Domain 3, Q19A mutation)PKEEVTIKAN LIYADGKTAT AEFKGTFEEA TAEAYRYADL LAKENGKYTVDVADKGYTLN IKFAGKEKTPEE SEQ ID NO: 14 (Domain 3, Q19E mutation)PKEEVTIKAN LIYADGKTET AEFKGTFEEA TAEAYRYADL LAKENGKYTVDVADKGYTLN IKFAGKEKTPEE In some embodiments, the polypeptide is a multimer comprising, or consisting essentially of, aplurality of mutated or non-mutated domains as defined by any embodiment disclo sed above.The multimer can e.g. be a dimer, a trimer, a tetramer, a pentamer or a hexamer. It can be ahomomultimer, 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 arealkali stable, such as by comprising the mutations disclosed above. The domains can be linked toeach other directly by peptide bonds between the C- and N-terrnini of the domains. Altematively,two or more units in the multimer can be linked by elements comprising oligomeric or polymericspecies, such as elements comprising up to 15 or 30 amino acids, such as 1-5, 1-10 or 5-10amino acids. The nature of such a link should preferably not destabilize the spatial conforrnationof the domains. This can e.g. be achieved by avoiding the presence of cysteine in the links.Furthermore, said link should preferably also be sufficiently stable in alkaline environments notto impair the properties of the domains. For this purpose, it is advantageous if the links do notcontain asparagine. It can additionally be advantageous if the links do not contain glutamine. Themultimer may further at the N-terrninus comprise a plurality of amino acid residues originatingfrom the cloning process or constituting a residue from a cleaved off signaling sequence. Thenumber of additional amino acid residues may e. g. be 15 or less, such as 10 or less or 5 or less.

As a specific example, the multimer may comprise an AQV sequence at the N-terrninus.

In certain embodiments, the multimer may comprise, or consist essentially, of a sequenceselected from the group consisting of: SEQ ID NO: 15, SEQ ID NO: 16, SEQ ID NO: 17 andSEQ ID NO: 18, such as a sequence selected from the group consisting of SEQ ID NO: 16, SEQID NO: 17 and SEQ ID NO: 18.

SEQ ID NO: 15 (Domain 3, tetramer) PKEEVTIKAN LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKENGKYTVDVADKGYTLN IKFAGKEKTPEE PKEEVTIKAN LIYADGKTQT AEFKGTFEEATAEAYRYADL LAKENGKYTV DVADKGYTLN IKFAGKEKTPEE PKEEVTIKANLIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKENGKYTV DVADKGYTLNIKFAGKEKTPEE PKEEVTIKAN LIYADGKTQT AEFKGTFEEA TAEAYRYADLLAKENGKYTV DVADKGYTLN IKFAGKEKTPEE SEQ 1D No; 16 Domain 3(N1oQ,N45A,N60Q)2 PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDvADKoYTLQ IKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQT AEFKGTFEEATAEAYRYADL LAKEAGKYTV DvADKoYTLQ IKFAGKEKTPEE SEQ ID NO: 17 Domain 3 (N10Q,N45A,N60Q)4 PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLQ IKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQT AEFKGTFEEATAEAYRYADL LAKEAGKYTV DVADKGYTLQ IKFAGKEKTPEE PKEEVTIKAQLIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTV DVADKGYTLQIKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADLLAKEAGKYTV DVADKGYTLQ IKFAGKEKTPEE SEQ ID NO: 18 Domain 3 (N10Q,N45A,N60Q)6 PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLQ IKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQT AEFKGTFEEATAEAYRYADL LAKEAGKYTV DVADKGYTLQ IKFAGKEKTPEE PKEEVTIKAQLIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTV DVADKGYTLQIKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADLLAKEAGKYTV DVADKGYTLQ IKFAGKEKTPEE PKEEVTIKAQ LIYADGKTQTAEFKGTFEEA TAEAYRYADL LAKEAGKYTV DVADKGYTLQ IKFAGKEKTPEEPKEEVTIKAQ LIYADGKTQT AEFKGTFEEA TAEAYRYADL LAKEAGKYTVDVADKGYTLQ IKFAGKEKTPEE In some embodiments, the polypeptide and/or multimer, as disclo sed above, furthercomprises at the C-terrninus or N-terrninus one or more coupling elements, selected from thegroup consisting of a cysteine residue, a plurality of lysine residues and a plurality of histidineresidues. The coupling element may e. g. be a single cysteine at the C-terrninus. The couplingelement(s) may be directly linked to the C- or N-terminus, or it/they may be linked via a linkercomprising up to l5 amino acids, such as 1-5, l-l0 or 5-l0 amino acids. This stretch shouldpreferably also be sufficiently stable in alkaline environments not to impair the properties of themutated protein. For this purpose, it is advantageous if the stretch does not contain asparagine. Itcan additionally be advantageous if the stretch does not contain glutamine. An advantage ofhaving a C- or N-terrninal cysteine is that endpoint coupling of the protein can be achievedthrough reaction of the cysteine thiol With an electrophilic group on a support. This providesexcellent 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 an SPR chip,e. g. to Biacore CM5 sensor chips as described in the examples, and measuring the kappa lightchain-binding capacity of the chip, using e.g. a specific kappa light chain-containing protein orpolyclonal human IgG (Where the maj ority of the IgG molecules have a kappa light chain),before and after incubation in alkaline solutions at a specified temperature, e. g. 22 +/- 2 °C. Theincubation can e. g. be perforrned in 0.1 M NaOH for a number of 10 min cycles, such as 50, 96or 100 cycles. The binding capacity of the matrix after 96-100 10 min incubation cycles in 0.1 MNaOH at 22 +/- 2 °C can be at least 40, such as at least 50, or at least 55 % of the bindingcapacity before the incubation. Altematively, the remaining binding capacity after 96-100 cyclesfor a particular mutant measured as above can be compared With the remaining binding capacityfor the parental polypeptide/multimer. In this case, the remaining binding capacity for the mutantmay be at least 105%, such as at least 110%, at least 125%, at least 150% or at least 200% ofthe parental polypeptide/multimer.

The invention also discloses a kappa light chain-binding polypeptide comprising at least onemutated binding domain of Peptostreptococcus Protein L, in Which at least one asparagineresidue of a parental domain defined by, or having at least 95% or 98% sequence homologyWith, SEQ ID NO: 2-6 or 12 has been mutated to another amino acid residue Which is notasparagine, proline or cysteine. The polypeptide may comprise at least the mutation N45A and/orthe mutation N60Q. In specific embodiments, the mutation(s) are selected from the groupconsisting of N45A; Nl0Q,N45A; N45A,N60Q, N10Q,N60Q and N10Q,N45A,N60Q, oraltematively selected from the group consisting of N45A; Nl0Q,N45A; N45A,N60Q andN10Q,N45A,N60Q. The alkali stability relative to a parental polypeptide can be improved and measured as disclosed above.

In some embodiments the polypeptide comprises or consists essentially of a plurality of mutatedbinding domains, such as 2, 3, 4, 5 or 6 domains, Wherein each domain comprises at least one ofthe mutations Nl0Q, N45A and N60Q, such as N45A and/or N60Q. Specifically, the mutation(s)in each domain can be selected from the group consisting of N45A; Nl0Q,N45A; N45A,N60Q,N10Q,N60Q and N10Q,N45A,N60Q, or altematively selected from the group consisting ofN45A; Nl0Q,N45A; N45A,N60Q and N10Q,N45A,N60Q. The domains can optionally be linked to each other by elements comprising up to 15 amino acids.

In a second aspect the present invention discloses a nucleic acid encoding a polypeptide ormultimer according to any embodiment disclo sed above. Thus, the invention encompasses allforms of the present nucleic acid sequence such as the RNA and the DNA encoding thepolypeptide or multimer. The invention embraces a vector, such as a plasmid, Which in additionto the coding sequence comprises the required signal sequences for expression of the polypeptideor multimer according the invention. In one embodiment, the vector comprises nucleic acidencoding a multimer according to the invention, Wherein the separate nucleic acids encoding each unit may have homo lo gous or heterologous DNA sequences.

In a third aspect the present invention disclo ses an expression system, Which comprises, anucleic acid or a vector as disclosed above. The expression system may e.g. be a gram-positiveor gram-negative prokaryotic host cell system, e.g. Bacíllus sp. or Escheríchía coli Which hasbeen modified to express the present polypeptide or multimer. In an altemative embodiment, theexpression system is a eukaryotic host cell system, such as a yeast, e.g. Píchea pastorís or Saccharomyces cerevísíae.

In a fourth aspect, the present invention discloses a separation matrix, Wherein a plurality ofpolypeptides or multimers according to any embodiment disclosed above have been coupled to asolid support. Such a matrix is useful for separation of kappa light chain-containing proteins and,due to the improved alkali stability of the polypeptides/multimers, the matrix Will Withstandhighly alkaline conditions during cleaning, Which is essential for long-terrn repeated use in abioprocess separation setting. The alkali stability of the matrix can be assessed by measuring thekappa light chain-binding capacity, using e.g. a specific kappa light chain-containing protein orpolyclonal human IgG, before and after incubation in alkaline solutions at a specifiedtemperature, e. g. 22 +/- 2 °C. The incubation can e. g. be performed in 0.l M NaOH for a numberof l5 min cycles, such as l00, 200 or 300 cycles. The binding capacity of the matrix after l00 l5min incubation cycles in 0.l M NaOH at 22 +/- 2 °C can be at least 80, such as at least 85, atleast 90 or at least 95% of the binding capacity before the incubation. Altematively, theincubation can be performed in 0.l M NaOH for a number of 4 h cycles, such as 6 cycles givinga total incubation time of 24 h. The binding capacity of the matrix after 24 h min total incubationtime in 0.l 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 binding capacity before the incubation.

As the skilled person Will understand, the expressed polypeptide or multimer should be purif1ed l0 to an appropriate extent before been immobilized to a support. Such purif1cation methods arewell known in the f1eld, and the immobilization of protein-based ligands to supports is easilycarried out using standard methods. Suitable methods and supports will be discussed below in more detail.

The solid support of the matrix according to the invention can be of any suitable well-knownkind. A conventional aff1nity separation matrix is often of organic nature and based on polymersthat expose a hydrophilic surface to the aqueous media used, i.e. expose hydroxy (-OH), carboxy(-COOH), carboxamido (-CONH2, possibly in N- substituted forms), amino (-NH2, possibly insubstituted form), oligo- or polyethylenoxy groups on their extemal and, if present, also onintemal surfaces. The solid support can suitably be porous. The porosity can be expressed as aKav or Kd value (the fraction of the pore volume available to a probe molecule of a particularsize) measured by inverse size exclusion chromatography, e.g. according to the methodsdescribed in Gel Filtration Principles and Methods, Pharrnacia LKB Biotechnology 1991, pp 6-13. By definition, both Kd and Kav values always lie within the range 0 - 1. The Kav value canadvantageously be 0.6 - 0.95, e.g. 0.7 - 0.90 or 0.6 - 0.8, as measured with dextran ofMw 110kDa as a probe molecule. An advantage of this is that the support has a large fraction of poresable to accommodate both the polypeptides/multimers of the invention and immunoglobulinsbinding 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 support via conventional couplingtechniques 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 beintroduced, which improves the availability of the polypeptide/multimer and facilitates thechemical coupling of the polypeptide/multimer to the support. Suitable spacers can be introducede.g. by activation of the support with epichlorohydrin, butanediol diepoxide, allyl glycidyl etheretc. Altematively, the polypeptide/multimer may be attached to the support by non-covalent bonding, such as physical adsorption or biospecific adsorption.

In some embodiments the matrix comprises 5 - 20, such as 5 - 15 mg/ml, 5 - 11 mg/ml or 8 - 11mg/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 11 coupling process, by the coupling conditions used and/or by the pore structure of the supportused. As a general rule the absolute binding capacity of the matrix increases With the amount ofcoupled polypeptide/multimer, at least up to a point Where the pores become significantlyconstricted by the coupled polypeptide/multimer. The relative binding capacity per mg coupledpolypeptide/multimer Will decrease at high coupling levels, resulting in a cost-benefit optimum Within the ranges specified above.

In some embodiments the polypeptides are coupled to the support via multipoint attachment.This can suitably be done by using such coupling conditions that a plurality of reactive groups inthe polypeptide react With reactive groups in the support. Typically, multipoint attachment caninvolve the reaction of several intrinsic reactive groups of amino acid residues in the sequence,such as amines in lysines, With the reactive groups on the support, such as epoxides, cyanateesters (e.g. from CNBr activation), succinimidyl esters (e. g. from NHS activation) etc. It ishowever also possible to deliberately introduce reactive groups at different positions in thepolypeptides to affect the coupling characteristics. In order to provide multipoint coupling vialysines, the coupling reaction is suitably carried out at a pH Where a significant fraction of thelysine primary amines are in the non-protonated nucleophilic state, e.g. at pH higher than 8.0, such as above 10.

In certain embodiments the polypeptides or multimers are coupled to the support via thioetherbonds. Methods for performing such coupling are Well-known in this field and easily performedby the skilled person in this field using standard techniques and equipment. Thioether bonds areflexible and stable and generally suited for use in aff1nity chromatography. In particular When thethioether bond is via a terminal or near-terrninal cysteine residue on the polypeptide or multimer,the mobility of the coupled polypeptide/multimer is enhanced Which provides improved bindingcapacity and binding kinetics. In some embodiments the polypeptide/multimer is coupled via aC-terrninal cysteine provided on the protein as described above. This allows for efficientcoupling of the cysteine thiol to electrophilic groups, e.g. epoxide groups, halohydrin groups etc.on a support, resulting in a thioether bridge coupling. The polypeptide/multimer can e. g. becoupled via single-point attachment, e. g. via a single cysteine or by directed multipointattachment, using e. g. a plurality of lysines or other coupling groups near a terrninus of the polypeptide/multimer. 12 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 nonspecif1c interactions, theyprovide 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 presentinvention can easily be prepared according to standard methods, such as inverse suspensiongelation (S Hj erten: Biochim Biophys Acta 79(2), 393-398 (1964). Altematively, the basematrices are commercially available products, such as crosslinked agarose beads sold under thename of SEPHAROSETM FF (GE Healthcare). In an embodiment, Which is especiallyadvantageous for large-scale separations, the support has been adapted to increase its rigidityusing the methods described in US6602990 or US7396467, Which are hereby incorporated by reference in their entirety, and hence renders the matrix more suitable for high flow rates.

In certain embodiments the support, such as a polysaccharide or agarose support, is crosslinked,such as With hydroxyalkyl ether crosslinks. Crosslinker reagents producing such crosslinks canbe e. g. epihalohydrins like epichlorohydrin, diepoxides like butanediol diglycidyl ether,allylating reagents like allyl halides or allyl glycidyl ether. Crosslinking is benef1cial for therigidity of the support and improves the chemical stability. Hydroxyalkyl ether crosslinks are alkali stable and do not cause significant nonspecif1c adsorption.

Altematively, the solid 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 divinyland monovinyl-substituted benzenes, the surface of the matrix is often hydrophilised to exposehydrophilic groups as defined above to a surrounding aqueous liquid. Such polymers are easilyproduced according to standard methods, see e.g. “Styrene based polymer supports developed bysuspension polymerization” (R Arshady: Chimica e L'Industria 70(9), 70-75 (1988)).Altematively, a commercially available product, such as SOURCETM (GE Healthcare) is used. Inanother altemative, the solid 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, 13 capillaries, or a filter (e. g. a membrane or a depth f1lter matrix).

As regards the shape of the matrix according to the invention, in one embodiment the matrix is inthe form of a porous monolith. In an altemative embodiment, the matrix is in beaded or particleform that can be porous or non-porous. Matrices in beaded or particle form can be used as apacked bed or in a suspended form. Suspended forms include those known as expanded beds andpure suspensions, in which the particles or beads are free to move. In case of monoliths, packedbed and expanded beds, the separation procedure commonly follows conventionalchromatography with a concentration gradient. In case of pure suspension, batch-wise mode will be used.

In a sixth aspect, the present invention discloses a method of iso lating a kappa light chain- containing protein, wherein a separation matrix as disclosed above is used.

In certain embodiments, the method comprises the steps of: a) contacting a liquid sample comprising a kappa light chain-containing protein with a separationmatrix as disclosed above, b) washing said separation matrix with a washing liquid, c) eluting the kappa light chain-containing protein from the separation matrix with an elutionliquid, and d) cleaning the separation matrix with a cleaning liquid.

The method may also comprise steps of, before step a), providing an affinity separation matrixaccording to any of the embodiments described above and providing a solution comprising akappa light chain-containing protein and at least one other substance as a liquid sample and of,after step c), recovering the eluate and optionally subj ecting the eluate to further separation steps,e.g. by anion or cation exchange chromatography, multimodal chromatography and/orhydrophobic interaction chromatography. Suitable compositions of the liquid sample, thewashing liquid and the elution liquid, as well as the general conditions for performing theseparation are well known in the art of aff1nity chromatography and in particular in the art ofProtein L chromatography. The liquid sample comprising a kappa light chain-containing proteinand at least one other substance may comprise host cell proteins (HCP), such as chinese hamsterovary (CHO) cell, E. coli or yeast cell proteins. Contents of CHO cell and E. coli proteins can conveniently be deterrnined by immunoassays directed towards these proteins, e. g. the CHO l4 HCP or E. coli HCP ELISA kits from Cygnus Technologies. The host cell proteins or CHOcell/E. colí/yeast proteins may be desorbed during step b).

The elution may be performed by using any suitable solution used for elution from Protein Lmedia. This can e.g. be a solution or buffer with pH 4 or lower, such as pH 2.5 - 4 or 2.8 - 3.5.In some embodiments the elution buffer or the elution buffer gradient comprises at least onemono- di- or trifunctional carboxylic acid or salt of such a carboxylic acid. In certainembodiments the elution buffer or the elution buffer gradient comprises at least one anionspecies selected from the group consisting of acetate, citrate, glycine, succinate, phosphate, and formiate.

In some embodiments, the cleaning liquid is alkaline, such as with a pH of 12 - 14. Such solutions provide efficient cleaning of the matrix, in particular at the upper end of the interval In certain embodiments, the cleaning liquid comprises 0.01 - 1.0 M NaOH or KOH, such as 0.05- 1.0 or 0.05 - 0.1 M NaOH or KOH. The high stability of the polypeptides of the invention enables the use of such comparatively strong alkaline solutions.

In some embodiments, steps a) - d) are repeated at least 10 times, such as at least 50 times or 50- 200 times. This is important for the process economy in that the matrix can be re-used many times.

Examples Mutagenesis of proteinMonomer constructs were designed from a Protein L disclosed in US6822075 (SEQ ID NO: 1), containing four kappa light chain-binding domains. These are numbered 1, 2, 3 and 4, startingfrom the N-terrninus (Fig.1). The DNA fragments were purchased from a DNA synthesizingcompany (DNA2.0). Four monomer constructs were prepared in a pJexpress201 cloning vector,each with an N-terrninal cysteine. For an overview of constructs, see SEQ ID NO: 2,4,5,12.Constructs were subcloned to expression vector pGO, containing E .coli GAP promoter andOmpA signal peptide sequence for periplasmatic localization of the target protein. The sequenceencoding the four domains was prepared by amplifying with oligonucleotides containing therestriction enzyme recognition sites for FspI and PstI on the 5” side and 3” side, respectively. The prepared DNA fragment encoding each domain was digested with FspI and PstI (New England Bio labs). Separately, expression Vector Was prepared With digestion With FspI and PstI andpurified by agarose gel electrophoresis and recovered. Both Were mixed and ligated With Quickligation kit (New England Biolabs). The ligated plasmid expressing each domain Was transforrned into a chemical competent E. coli K12 strain With a heat shock method.

Further mutations of amino acids N10, N45, Q19, and N60 in domain 3 Were prepared inexpression Vector pJexpress40l(DNA2.0) containing T5 promoter under a lac operon controlmechanism (SEQ ID NO: 7-11,13-14). Constructs Were designed With and Without OmpA signal peptide but Without a C-terrninal cysteine.Tetramers of domain 3, dimer, tetramer and hexamer of domain 3 With N45, N10 and N60mutations Were also prepared in pJexpress401, With and Without C-terrninal cysteine (SEQ ID No; 15-18).

Construct expression and purification The E. coli K12 recombinant cells Were cultured in shake flasks With LB-broth (10 g peptone, 5g yeast extract, 5 g NaCl) supplemented With 25 mg/l kanamycin at 37°C until optical density at600 nm reached 0.8. At this point protein expression Was induced With Isopropyl ß-D-l-thiogalactopyranoside (VWR Intemational) With final concentration of 1 mM. Upon inductionthe temperature Was loWered to 30°C and the cultures Were incubated for 5 hours. Thecultivation Was stopped and cells Were centrifuged for 15 minutes at 4000 x g and thesupematant Was discarded. Cells Were resuspended in 1/10 of culturing Volume With phosphatebuffered saline (PBS) and sonicated using pulse-sonication With an active time of 2 minutes. Thesonicated samples Were clarified from cell debris by centrifugation at 6000xg for 30 minutes, fo lloWed by micro f1ltration With a membrane having a 0.2 um pore size.

The purif1ed ligands Were analyzed With 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, in the cases of the non-mutated single domains, a cysteine at the C terrninus and an AQV sequence at the N-terrninus,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 16 groups on the chip) in an amount suff1cient to give a signal strength of about 1000RU in aBiacore instrument (GE Healthcare, Sweden) . To follow the IgG binding capacity of theimmobilized surface 1mg/ml human polyclonal IgG (Gammanorrn) was flowed over the chipand the signal strength was noted. The surface was then cleaned-in-place (CIP), i.e. flushed with100mM NaOH for 10 minutes at room temperature (22 +/- 2°C). This was repeated for 96 cyclesand the immobilized ligand alkaline stability was followed as the relative loss of IgG bindingcapacity (signal strength) after each cycle. The results for the non-mutated domains are shown inFig. 2 and indicate that Domain 1 has a distinctly lower alkali stability than the other domainsand that Domain 3 has the highest alkali stability. Results for single-domain asparagine mutantsof Domain 3 are shown in Fig. 3 and show an improved alkali stability for all the mutants in comparison with the parental Domain 3, which was used as a reference in parallel with the mutations.

Table 1.

Ligand Sequence Retained Ref. Sample/ref.capacity capacity ratioafter 96 (%)cycles (%) Domain 1 (D1) SEQ ID NO: 2 13 31 0.42 Domain 2 (D2) SEQ ID NO: 12 22 31 0.71 Domain 3 (D3) SEQ ID NO: 4 31 31 1.00 Domain 4 (D4) SEQ ID NO: 5 26 31 0.84 D3(N45A)1 SEQ ID NO: 7 44 31 1.42 D3(N10Q,N45A)1 SEQ ID NO: 8 48 31 1.55 D3(N45A,N60Q)1 SEQ ID NO: 9 59 31 1.90 D3(N10Q,N45A,N60Q)1 SEQ ID NO: 11 59 31 1.90 Domain 3 (D3) SEQ ID NO: 4 28 28 1.00 D3(Q19A)1 SEQ ID NO: 13 28 28 1.00 D3(Q19E)1 SEQ ID NO: 14 31 28 1.11 Example 2 The purified multidomain ligands listed in Table 2 were immobilized on Biacore CM5 sensorchips (GE Healthcare, Sweden), using the amine coupling kit of GE Healthcare (forcarbodiimide coupling of amines on the carboxymethyl groups on the chip) in an amountsuff1cient to give a signal strength of about 1000RU in a Biacore instrument (GE Healthcare,Sweden) . The Protein L had an additional AIHNRA sequence at the N-terrninus.To follow theIgG binding capacity of the immobilized surface 1mg/ml human polyclonal IgG (Gammanorrn) 17 was flowed over the chip and the signal strength was noted. The surface was then cleaned-in-place (CIP), i.e. flushed with 100mM NaOH for 10 minutes at room temperature (22 +/- 2°C).This was repeated for 96 cycles and the immobilized ligand alkaline stability was followed as therelative loss of IgG binding capacity (signal strength) after each cycle. The results are shown inTable 2 and Fig. 4 and show that the tetrameric Domain 3 has an improved alkali stability in comparison with Protein L which was run in parallel as a reference.

Table 2Ligand Sequence Retained Ref. Sample/ref. capacity capacity ratio after 96 (%) cycles (%)Protein L SEQ ID NO: 1 28 28 1.00Domain 3 SEQ ID NO: 15 38 28 1.36tetramerExample 3 The purified multidomain ligands listed in Table 3 were immobilized on Biacore CM5 sensorchips and evaluated by the methods used in Example 2. -cys at the end of the ligand designationindicates that the ligand has a C-terrninal cysteine in addition to the sequence defined by SEQ IDNO: 16-18. The results are shown in Table 3 and Fig. 5 and show that all the mutated Domain 3dimers, tetramers and hexamers have an improved alkali stability in comparison with Protein L which was run in parallel as a reference.

Table 3Ligand Sequence Retained Ref. Sample/ref.capacity capacity ratioafter 100 (%)cycles (%)Protein L SEQ ID NO: 1 23 23 1.00D3(N10Q,N45A,N60Q)2 SEQ ID NO: 16 59 23 2.56D3(N10Q,N45A,N60Q)2-cys SEQ ID NO: 16 59 23 2.56D3(N10Q,N45A,N60Q)4 SEQ ID NO: 17 60 23 2.61D3(N10Q,N45A,N60Q)4-cys SEQ ID NO: 17 54 23 2.35D3(N10Q,N45A,N60Q)6 SEQ ID NO: 18 58 23 2.52D3(N10Q,N45A,N60Q)6-cys SEQ ID NO: 18 56 23 2.43 Example 4The purified di-, tetra- and hexameric ligands of Table 4 were immobilized on agarose beads using the methods described below and assessed for capacity. The results are shown in Table 4. 18 Table 4 Ligand Sequence Ligand QB10 Fab QB10 Dabcontent mg/ml mg/ml D3(N10Q,N45A,N60Q)2 SEQ ID NO: 16 9.5 mg/ml 19.3 15.4D3(N10Q,N45A,N60Q)4 SEQ ID NO: 17 8.8 mg/ml 19.3 15.9D3(N10Q,N45A,N60Q)6 SEQ ID NO: 18 11.0 mg/ml 21.1 16.7Activation The base matrix used Was rigid cross-linked agarose beads of 85 micrometers (Volume-Weighted)median diameter, prepared according to the methods of US6602990 and With a pore sizecorresponding to an inverse gel filtration chromatography Kav Value of 0.70 for dextran of MW110 kDa, according to the methods described in Gel Filtration Principles and Methods,Pharrnacia 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 a100 mL flask With mechanical stirring for 10 min at 25°C. 4.0 mL of epichlorohydrin Was addedand the reaction progressed for 2 hours. The activated gel Was Washed With 10 gel sedimentVolumes (GV) of Water.

Coupling The activated gel Was Washed With 5 GV 0.2 M phosphate/1 mM EDTA pH 11.5 (couplingbuffer). 15 ml gel + 13 mg ligand/ml gel (5.0 ml) + 5.5 ml coupling buffer + 4.7 g sodiumsulfate Were mixed in a 50 ml flask and stirred at 30 °C for 18.5 h. The pH Was measured as10.8.

After immobilization the gels Were Washed 3xGV With distilled Water and then 5xGV With 0.1 Mphosphate/1 mM EDTA pH 8.5 . The gels + 1 GV {0.l M phosphate/1 mM EDTA/7.5 %thioglycerol pH 8.5} Was mixed and the flask Was stirred at 45 °C for 2h 20 min. . The gel Wasthen Washed altemately With lxGV 0.1 M HAc and lxGV {0.1 M TRIS/0.15 M NaCl pH 8.5}and and then 6xGV With distilled Water. Gel samples Were sent to an extemal laboratory foramino acid analysis and the ligand content (mg/ml gel) Was calculated from the total amino acidcontent. The coupling protocol used provides multipoint coupling, With several lysines of eachdomain attached to the gel. 2 ml of resin Was packed in TRICORNTM 5 100 columns.19 Protein a) Purified Fab prepared from a papain-digested IgG mAb , diluted to 1mg/ml in Equilibrationbuffer. b) Purified Dab prepared from heat-treated E. coli supematant, diluted to 1mg/ml inEquilibration buffer. The Dab contained solely a kappa light chain, Without any antigen-binding site.

Equilibration buffer APB Phosphate buffer 20 mM +0.15 M NaCl , pH 7,4 (Medicago) Adsorption buffer APB Phosphate buffer 20 mM +0. 15 M NaCl, pH 7.4 (Medicago).

Elution buffer 25 mM citrate pH 2.5 The breakthrough capacity Was deterrnined With an ÄKTAExplorer 10 system at aresidence time of 4 minutes. Equilibration buffer Was run through the bypass column until astable baseline Was obtained. This Was done prior to auto zeroing. Sample Was applied to thecolumn until a 100% UV signal Was obtained. Then, equilibration buffer Was applied again untila stable baseline Was obtained.

Sample Was loaded onto the column until a UV signal of 85% of maximum absorbanceWas reached. The column Was then Washed With equilibration buffer and eluted at pH 2.5 at a flow rate of 0.5 ml/min.For calculation of breakthrough capacity at 10%, the equation beloW Was used. That is the amount of Fab/Dab that is loaded onto the column until the concentration of Fab/Dab in the column effluent is 10% of the Fab/Dab concentration in the feed.

VGPP C A V - Aq10% I i Vapp _ Vsys _ * dvVc V A100% _ AsubsysA100% = 100% UV signal; Asub = absorbance contribution from non-binding proteins; A(V) = absorbance at a given applied Volume; VC = column Volume; Vapp = Volume applied until 10% breakthrough; Vsys = system dead Volume; CO = feed concentration.The dynamic binding capacity (DBC) at 10% breakthrough Was calculated and the appearance ofthe curve Was studied. The curve Was also studied regarding binding, elution and CIP peak. The dynamic binding capacity (DBC) Was calculated for 10 and 80% breakthrough.

This Written description uses examples to disclose the invention, including the best mode, andalso to enable any person skilled in the art to practice the inVention, including making and usingany devices or systems and performing any incorporated methods. The patentable scope of theinVention is defined by the claims, and may include other examples that occur to those skilled inthe art. Such other examples are intended to be Within the scope of the claims if they havestructural elements that do not differ from the literal language of the claims, or if they includeequivalent structural elements With insubstantial differences from the literal languages of theclaims. All patents and patent applications mentioned in the text are hereby incorporated by reference in their entireties, as if they Were indiVidually incorporated. 21

Claims (20)

1. A kappa light chain-binding polypeptide comprising at least one mutated binding domain ofPeptostreptococcus Protein L, in Which domain at least one asparagine residue of a parentaldomain defined by, or having at least 95% or 98% sequence homology With, SEQ ID NO: 2-6 or 12 has been mutated to another amino acid residue Which is not asparagine, proline or cysteine.

2. The polypeptide of claim 1, comprising at least the mutation N45A.

3. The polypeptide of claim 1 or 2, comprising at least the mutation N60Q.

4. The polypeptide of any preceding claim, Wherein the mutation(s) are selected from the groupconsisting of N10Q; N45A; N60Q; N10Q,N45A; N45A,N60Q, N10Q,N60Q andN10Q,N45A,N60Q or from the group consisting of N45A; N10Q,N45A; N45A,N60Q,N10Q,N60Q and N10Q,N45A,N60Q.

5. The polypeptide of any preceding claim, comprising or consisting essentially of a plurality ofmutated binding domains, such as 2, 3, 4, 5 or 6 domains, Wherein each domain comprises the mutation N45A and/or N60Q.

6. The polypeptide of claim 5, Wherein the mutation(s) in each domain are selected from thegroup consisting of N10Q; N45A; N60Q; N10Q,N45A; N45A,N60Q, N10Q,N60Q andN10Q,N45A,N60Q or from the group consisting of N45A; N10Q,N45A; N45A,N60Q,N10Q,N60Q and N10Q,N45A,N60Q.

7. The polypeptide of claim 5 or 6, Wherein the domains are linked by elements comprising up to 15 amino acids.

8. The polypeptide according to any preceding claim, Wherein the alkaline stability is improvedrelative to a parental polypeptide, as measured by the remaining binding capacity for kappa lightchain-containing proteins after 96-100 10 min incubation cycles in 0.1 M aqueous NaOH at 22 +/- 2 °C. 22

9. The polypeptide of any preceding claim, Wherein the parental domain is defined by, or has atleast 95% or 98% sequence homology With, an amino acid sequence selected from the group consisnng of SEQ 1D No; 3, SEQ 1D No; 4, SEQ 1D No; s and SEQ 1D No; 12.

10. The polypeptide of any preceding claim, Wherein the parental domain is defined by, or has atleast 95% or 98% sequence homology With SEQ ID NO: 4.

11. A nucleic acid or a Vector encoding a polypeptide or multimer according to any preceding claim.

12. An expression system, Which comprises a nucleic acid or Vector according to claim ll.

13. A separation matrix, Wherein a plurality of polypeptides according to any one of claims 1 - 10 have been coupled to a solid support.

14. The separation matrix according to claim 13, Wherein the polypeptides have been coupled to the solid support by multipoint attachment.

15. The separation matrix according to any one of claims 13-14, Wherein the binding capacity ofthe matrix for kappa light chain-containing proteins after 100 10 min incubation cycles in 0.1 MNaOH at 22 +/- 2 °C is at least 40, such as at least 50, or at least 55% of the binding capacity before the incubation.

16. A method of isolating a kappa light chain-containing protein, Wherein a separation matrix according to any one of claims 13-15 is used.

17. The method of claim 16, comprising the steps of: a) contacting a liquid sample comprising a kappa light chain-containing protein With a separationmatrix according to any one of claims 13-15, b) Washing said separation matrix With a Washing liquid, c) eluting the kappa light chain-containing protein from the separation matrix With an elutionliquid, and d) cleaning the separation matrix With a cleaning liquid. 23

18. The method of claim 17, Wherein the Cleaning liquid is alkaline, such as With a pH of 12 -14.

19. The method of claim 17 or 18, Wherein the Cleaning liquid comprises 0.01 - 1.0 M NaOH orKOH, such as 0.05 - 1.0 M or 0.05 - 0.1 M NaOH or KOH.

20. The method of any one of claims 17 - 19, Wherein steps a) - d) are repeated at least 10 times, such as at least 50 times or 50 - 200 times. 24