A recombinant polypeptide comprising a metal-binding motif

Abstract

The present disclosure generally relates to a recombinant polypeptide comprising an IgG-binding domain derived from Streptococcus Protein G (SpG) and a metal-binding motif, wherein the recombinant polypeptide has a first, target-binding configuration in the presence of metal ions, and a second, non-target-binding configuration in the absence of metal ions, wherein the recombinant polypeptide is configured to bind to the Fab-region of IgG in the first, target-binding configuration. The present disclosure also relates to a ligand comprising the recombinant polypeptide, a chromatography matrix comprising the ligand coupled to the matrix, and a method for purification of sample proteins utilizing the chromatography matrix. Furthermore, the present disclosure relates to a nucleic acid encoding the recombinant polypeptide and to an expression system comprising the nucleic acid.

Description

[0001] A RECOMBINANT POLYPEPTIDE COMPRISING A METAL-BINDING MOTIF

[0002] TECHNICAL FIELD

[0003] The present disclosure generally relates to a recombinant polypeptide comprising an IgG-binding domain derived from Streptococcus Protein G (SpG) and a metalbinding motif. The present disclosure further relates to a ligand comprising the recombinant polypeptide, a chromatography matrix comprising the ligand coupled to the matrix, and a method for purification of sample proteins utilizing the chromatography matrix. The present disclosure also relates to a nucleic acid encoding the recombinant polypeptide and to an expression system comprising the nucleic acid.

[0004] BACKGROUND

[0005] Antibodies, also known as immunoglobulins (Ig), are commonly used in the field of biological sciences, and represent an important class of biotherapeutic drugs. Antibodies may be used to treat various diseases, including allergy, autoimmune diseases, and cancer. An antibody is a relatively large, Y-shaped protein, and contains two antigenbinding fragments (Fab), and one crystallizable fragment (Fc) forming the trunk of the Y- shape.

[0006] Immunoglobulin G (IgG) is the main type of antibody found in blood and extracellular fluid and represents the majority of the antibody -based immunity against invading pathogens.

[0007] Bacteria such as staphylococci and streptococci have evolved to express surface proteins with an inherent affinity to immunoglobulins (Ig) with broad species specificity as a means of evading an immune response. As proven through several engineering efforts, these proteins can serve as model systems since they are well characterized in terms of structural properties, interactions, and thermodynamic stability.

[0008] Furthermore, such Ig-binding proteins are often used in affinity chromatography applications for recovery of immunoglobulins (Ig) from different samples.

[0009] Staphylococcus aureus protein A (SpA) is a widely used affinity chromatography ligand. SpA binds with high affinity to the Fc region of immunoglobulins.

[0010] To date, various efforts have been made to increase the binding capacity of protein A and domains thereof through recombinant technologies. Furthermore, several methods have been developed to render protein A less susceptible to alkaline pH, which is commonly used in the elution step in affinity chromatography. EP 3 523 318 Bl relates to an IgG-binding polypeptide mutant of domain Z derived from SpA domain B, which binds to the Fc-region of an immunoglobulin, and which comprises a metal-binding motif inserted between helix 2 and 3 of the Z domain. The IgG- binding polypeptide mutant allows for elution of IgG by removal of calcium at a pH of 5.5 to 7.0.

[0011] Streptococcal protein G (SPG) is another protein with affinity to immunoglobulins. SPG generally binds to the Fc fragment of IgGs from various mammalian species.

[0012] SPG consists of repetitively arranged domains, wherein the C-terminal domains (Cl, C2 and C3) are responsible for IgG-binding, and the N-terminal domains bind to human serum albumin.

[0013] The tertiary structure as well as the amino acid composition of streptococcal protein G (SPG) differs significantly from the tertiary structure and amino acid composition of protein A. Hence, these proteins have different binding characteristics. SPG may e.g. be used for the purification of IgGs, which do not bind well to protein A.

[0014] As the market for therapeutic antibodies is rapidly growing, the demand for new and improved ligands is also increasing. Hence, IgG-binding polypeptides derived from streptococcal protein G (SPG) will become important for recovery and isolation of immunoglobulins or fragments thereof. There is consequently a need to provide an improved SPG which binds selectively and with high affinity to immunoglobulins (IgG), and which may be used to recover and isolate IgGs from target samples.

[0015] SUMMARY

[0016] In view of the above-mentioned problems, it is an object of the present disclosure to provide improvements with respect to IgG-binding polypeptides derived from streptococcal protein G.

[0017] According to a first aspect of the present disclosure, there is provided a recombinant polypeptide comprising an IgG-binding domain derived from Streptococcus Protein G (SPG) and a metal-binding motif, wherein the recombinant polypeptide has a first, target-binding configuration in the presence of metal ions, and a second, non-target-binding configuration in the absence of metal ions, wherein the recombinant polypeptide is configured to bind to IgG in the target-binding configuration. The inventors have found that a metal-binding motif may be successfully introduced in an IgG-binding domain derived from SPG and have thereby provided a recombinant metal-dependent protein G-binding domain.

[0018] Upon binding of a metal ion, the recombinant polypeptide can switch from an inactive state; i.e. non-target binding configuration, to a functional, i.e. target-binding configuration.

[0019] The inventors have made an engineering effort and combined a semi-rational protein design and selection by Escherichia coli display followed by fluorescence-activated cell sorting and isolated several polypeptides where the inherent target affinity to antibodies and antibody fragments have been rendered metal ion dependent. Once depleted of metal ions, the target interaction of these proteins is lost.

[0020] Introduction of a metal-binding motif into a naturally occurring antibody-binding protein domain combined with directed evolution using bacterial display resulted in several recombinant polypeptides comprising switch-like binding properties with intact affinity to IgG solely in the presence of metal ions.

[0021] As used herein, the term “recombinant polypeptide” means a polypeptide formed by using recombinant DNA techniques. The recombinant polypeptide is derived from an IgG-binding domain of streptococcal protein G (SPG), into which a metal-binding motif is introduced. The recombinant polypeptide may also be referred to as an engineered polypeptide.

[0022] The term “IgG-binding domain” means the Cl, C2 or C3 domain of Streptococcus protein G (SPG).

[0023] The IgG-binding domain; i.e. the Cl, C2 or C3 domain may be intact, but the introduction of the metal-binding motif into the domain provides a recombinant polypeptide. It is also conceivable that the IgG-binding domain comprises one or a plurality of mutations.

[0024] The “metal-binding motif’ is typically a metal-binding loop. A loop is a peptide fragment generally void of secondary structure and which may impact protein folding. The metal-binding loop binds metal ions, and in doing so, the loop (and the recombinant polypeptide) switches configuration from an inactive, non-target binding configuration to a configuration in which the polypeptide binds to an immunoglobulin (IgG).

[0025] The recombinant polypeptide may have a different conformation in the first, target-binding configuration, and the second, non-target-binding configuration. For example, the folding of the metal-binding motif; i.e. metal-binding loop may be different. As mentioned hereinbefore, the IgG-binding domains of streptococcal protein G include domain Cl, C2, and C3. Accordingly, the recombinant polypeptide of the present disclosure may comprise domain Cl, C2 or C3.

[0026] The C domains of streptococcal protein G; i.e. domain Cl, C2, and C3 are small (55 residue) peptides with immunoglobulin-binding activity.

[0027] Domain Cl, C2, and C3 of streptococcal protein G contain one alpha helix and one beta sheet. The beta sheet is composed of four anti-parallel beta strands. The beta strands are arranged in a sheet-like structure that is stabilized by hydrogen bonds between them. The alpha helix is positioned in the middle of the structure, crossing over the beta sheet. This configuration positions the alpha helix on top of the beta sheet formed by the four beta strands.

[0028] Each C domain (Cl, C2, C3) contains a first beta strand, a second beta strand arranged C-terminally of the first beta strand, an alpha helix arranged C-terminally of the second beta strand, a third beta strand arranged C-terminally of the alpha helix, and a fourth beta strand arranged C-terminally of the third beta strand.

[0029] The C domain of streptococcal protein G may be defined by: [first beta strand]- [second beta strand]-[alpha helix]-[third beta strand] -[fourth beta strand].

[0030] The C domains have a similar structure but differ slightly in the amino acid sequences. Figure la illustrates the crystal structure of the C2 domain, wherein the insertion site for the metal-binding loop is marked in darker grey. In figure lb, an exemplary metalbinding loop (C2caV3) has been inserted.

[0031] The metal-binding motif is typically inserted N-terminally (upstream) of the alpha helix of the IgG-binding domain, i.e. domain Cl, C2 or C3.

[0032] Preferably, the metal-binding motif is inserted between the second beta strand and the alpha helix of the IgG-binding domain, i.e. domain Cl, C2 or C3 (see figure 1c).

[0033] Figure 1c illustrates the sequences of domain Cl, C2, and C3, respectively (wild type), and the metal -binding loop insertion site. Domain Cl differs from domain C2 in two positions located in the first beta sheet (see arrows in figure 1c). Domain C3 differs from domain C2 in three positions throughout the scaffold, one in the second beta sheet, one in the alpha helix, and one in the third beta sheet (see arrows in figure 1c).

[0034] Typically, the IgG-binding domain is domain Cl or domain C2.

[0035] In exemplary embodiments, the recombinant polypeptide is configured to bind to the Fab region of the immunoglobulin (IgG) in the first, target-binding configuration. Hence, the binding between the recombinant polypeptide and the IgG is highly selective and specific.

[0036] According to published crystal structures (PDB: 1FCC (1) and 1IGC (2)), the binding surface for Fab is mainly located to the second beta-sheet and the C-terminal end of the alpha helix.

[0037] As mentioned hereinbefore, the IgG-binding domains of protein G predominantly interact with the Fc region of IgG. The inventors have surprisingly found that the recombinant polypeptides of the present disclosure can selectively bind to the Fab-region of IgG with high specificity.

[0038] The recombinant polypeptides of the present disclosure may thus be used to isolate Fab fragments. Beyond therapeutic applicability, Fab fragments may be used for cocrystallization experiments to aid in protein structure determination and for visualization of tumors through radio imaging (3).

[0039] The metal-binding motif may be a calcium-binding motif, and wherein the metal ions are calcium ions.

[0040] The metal-binding motif is preferably a metal-binding loop comprising 10-14 amino acids, preferably from 12-13 amino acids.

[0041] Such a metal-binding loop is stable, and the loop structure is maintained intact. If the metal-binding loop is too small or too big, the stability and structure of the loop may be impaired. If the loop is too small, the folding of the loop, and the ability to bind metal-ions may be impaired.

[0042] The metal-binding motif may be defined by the following sequence:

[0043] X4X5X6X7X8X9XioXi 1X12X13X14X15X16 (SEQ ID NO: 1), wherein

[0044] X4 is aspartic acid (D),

[0045] X5 is alanine (A), arginine (R), glutamic acid (E), lysine (K), serine (S), threonine (T), leucine (L), or valine (V),

[0046] Xe is aspartic acid (D), or asparagine (N),

[0047] X7 is arginine (R), asparagine (N), glycine (G), or lysine (K),

[0048] Xs is asparagine (N), aspartic acid (D), or serine (S), preferably aspartic acid (D),

[0049] X9 is glycine (G),

[0050] X10 is arginine (R), glycine (G), lysine (K), phenylalanine (F), threonine (T), histidine (H), or tyrosine (Y), preferably phenylalanine (F), tyrosine (Y) or arginine (R), Xu is isoleucine (I), leucine (L), or valine (V),

[0051] X12 is asparagine (N), aspartic acid (D), glutamic acid (E), serine (S), or threonine (T), preferably asparagine (N) or aspartic acid (D),

[0052] X13 is alanine (A), phenylalanine (F), threonine (T), or tyrosine (Y),

[0053] Xu is aspartic acid (D), glutamic acid (E), or lysine (K)

[0054] X15 is aspartic acid (D), or glutamic acid (E), preferably aspartic acid (D),

[0055] Xi6 is alanine (A), leucine (L), phenylalanine (F), or valine (V).

[0056] Preferably, the metal-binding motif comprises a plurality of polar amino acids.

[0057] The plurality of polar amino acids may be selected from arginine (R), asparagine (N), glutamine (Q), glutamic acid (E), histidine (H), lysine (K), serine (S), and threonine (T). Polar side chains contain groups that are either charged at pH 7 (e.g. aspartic acid and glutamic acid) or groups that are able to participate in hydrogen bonding.

[0058] For example, the metal-binding motif may comprise a polar amino acid residue in position X4, Xe, X7, Xs, X10, X12, X14, and X15.

[0059] Preferably, at least some of polar amino acids are negatively charged. The polar and negatively charged amino acids are believed to coordinate calcium by pointing inwards to the loop.

[0060] Preferably, the metal-binding motif comprises at least three, preferably four amino acids selected from aspartic acid (D) or asparagine (N).

[0061] Preferably, the metal-binding motif contains aspartic acid (D) in position X4, position Xs, and position X15. The metal-binding motif may also contain aspartic acid (D) in position X12.

[0062] Preferably, the metal-binding motif contains phenylalanine (F), tyrosine (Y) or arginine (R) in position X10.

[0063] The inventors have found that an improved binding performance with less protein ending up in the flow-through is obtained if X10 is phenylalanine (F), tyrosine (Y) or arginine (R).

[0064] The recombinant polypeptide may further comprise a linker upstream of the metal-binding motif, wherein the linker comprises 1 to 4 amino acids, preferably 2 to 3 amino acids.

[0065] The linker prevents the protein structure from becoming disrupted. The linker is arranged N-terminally of the metal-binding motif, i.e. the metal-binding loop.

[0066] The linker preferably comprises 2 or 3 amino acids selected from serine (S), threonine (T), glycine (G) and / or proline (P). Serine and threonine impart polar characteristics, whereas glycine is associated with flexibility. Proline imparts rigidity to the metal-binding motif.

[0067] In exemplary embodiments, the linker contains at least one proline (P) residue. The metal-binding motif comprising the linker may be defined by:

[0068] XiX2X3X4X5X6X7X8X9XioXi 1X12X13X14X15X16 (SEQ ID NO: 2), wherein

[0069] Xi is serine (S), threonine (T), glycine (G), proline (P), or absent,

[0070] X2 and X3 are selected from serine (S), threonine (T), glycine (G) or proline (P).

[0071] In SEQ ID NO:2 hereinabove, X1X2X3 represents the linker.

[0072] The recombinant polypeptide of the present disclosure may be defined by: TTYKLILNGKTLKGETTTEXiX2X3X4X5X6X7X8X9XioXiiXi2Xi3Xi4Xi5Xi6EKVFKQYAN DNGVDGEWTYDDATKTFTVTE (SEQ ID NO:3);

[0073] TTYKLVINGKTLKGETTTEXiX2X3X4X5X6X7X8X9XioXiiXi2Xi3Xi4Xi5Xi6EKVFKQYA NDNGVDGEWTYDDATKTFTVTE (SEQ ID NO:4);

[0074] TTYKLVINGKTLKGETTTKXiX2X3X4X5X6X7X8X9XioXiiXi2Xi3Xi4Xi5Xi6EKAFKQYA NDNGVDGVWTYDDATKTFTVTE (SEQ ID NO: 5), or a corresponding sequence having at least 80%, preferably at least 90 %, more preferably at least 95% sequence homology with any one of SEQ ID NO: 3 -5.

[0075] SEQ ID NO:3 represents a recombinant polypeptide comprising domain Cl.

[0076] SEQ ID NO:4 represents a recombinant polypeptide comprising domain C2.

[0077] SEQ ID NO:5 represents a recombinant polypeptide comprising domain C3.

[0078] The metal -binding motif by a sequence selected from any one of SEQ ID NO: 6-19 or a corresponding sequence having at least 80%, preferably at least 90 %, more preferably at least 95% sequence homology with any one of SEQ ID NO:6-19.

[0079] Accordingly, the metal -binding motif may be selected from: DVNNDGFIDFEDA (SEQ ID NO: 6), DKDGDGRLNYKDA (SEQ ID NO: 7), DVNRDGKINAKDA (SEQ ID NO: 8), DTNKDGFVDFKDA (SEQ ID NO: 9), DVNKDGRVDFKDA (SEQ ID NO: 10), DTNNDGYVNFKDA (SEQ ID NO: 11), DTNRDGRVDYKDA (SEQ ID NO: 12), DTNRDGKVTFEDA (SEQ ID NO: 13), DKNNDGKINYKDA (SEQ ID NO: 14), DANRDGRIDAKDA (SEQ ID NO: 15), DEDGNGYINFEDA (SEQ ID NO: 16), DANKDGKINAKDA (SEQ ID NO: 17), DANRDGRINADDA (SEQ ID NO: 18), DANKDGKINAEDA (SEQ ID NO: 19), or a corresponding sequence having at least 80%, preferably at least 90 %, more preferably at least 95% sequence homology with any one of SEQ ID NO 6-19.

[0080] The term “sequence homology” may also be referred to as “sequence identity”. As used herein, the term "at least 80% sequence homology" refers to a comparison between two amino acid sequences, where 80% or more of the residues in one sequence are identical to the corresponding residues in the other sequence when aligned. Sequence homology (or sequence identity) is typically measured over the full length of the sequences being compared (e.g. across the length of the metal-binding motif, or of the recombinant polypeptide, where applicable).

[0081] In practice, this means that a sequence having at least 80% sequence homology with any one of SEQ ID NO: 6- 19 may differ in up to 20% of the amino acids.

[0082] The IgG-binding domain may comprise at least one mutation, wherein the mutation may be a replacement of glutamic acid (E) with glycine (G) in the first amino acid residue of the alpha helix of the IgG-binding domain.

[0083] The inventors have found that replacing the glutamic acid residue in this position; i.e. the first amino acid residue of the alpha helix of the IgG-binding domain impacts the ability of the metal-binding motif (and the recombinant polypeptide) to retain its calcium dependence.

[0084] The recombinant polypeptide of the present disclosure may be defined by a sequence selected from any one of SEQ ID NO:20-38, or a corresponding sequence having at least 80%, preferably at least 90%, more preferably at least 95% sequence homology with any one of SEQ ID NO:20-38 (see also figure 6).

[0085] According to another aspect, there is provided a ligand comprising a recombinant polypeptide as described hereinabove.

[0086] The ligand may be a multimeric version of the recombinant polypeptide, e.g. a dimer, trimer, tetramer, pentamer, hexamer, heptamer, or an octamer. Multimeric versions of the recombinant polypeptide may benefit from avidity effects, resulting in stronger binding interactions. As demonstrated in the Example section, the tetrameric version of a recombinant polypeptide of the present disclosure (C2caEP?) displays an enhanced binding capacity. Furthermore, as demonstrated in the Example section, the tetrameric version retains bound target proteins effectively during the washing steps of affinity chromatography such that loss of target protein into the flow-through is minimized.

[0087] A ligand comprising a multimer of the recombinant polypeptide may further comprise at least one linker. The linker connects the recombinant polypeptide domains and provides flexibility between the domains. The linker(s) may prevent steric hindrance, and secure that the overall polypeptide structure remains stable.

[0088] For example, the ligand may comprise the following structure:

[0089] [recombinant polypeptide]-[linker]-[recombinant polypeptide]-[linker]- [recombinant polypeptide]-[linker]-[recombinant polypeptide].

[0090] Accordingly, the ligand may be a tetrameric version of the recombinant polypeptide. The ligand comprises four recombinant polypeptide domains separated by a linker sequence.

[0091] The linker may e.g. be defined by SEQ ID NO:40.

[0092] In exemplary embodiments, the ligand comprises the following structure: [SEQ ID NO:20]-[SEQ ID NO:40]-[SEQ ID NO:20]-[SEQ ID NO: 40]-[SEQ ID NO:20]- [SEQ ID NO:40]-[SEQ ID NO:20]-[SEQ ID NO:40]-[SEQ ID NO:41],

[0093] SEQ ID NO:41 is used for coupling the ligand to a chromatography matrix. The three first amino acid residues of SEQ ID NO:41 act as a spacer before the C-terminal cysteine residue, which is used for coupling to the chromatography matrix.

[0094] The ligand may further comprise linking peptides and / or signal peptides between the monomer units. The ligand may also comprise a C-terminal cysteine for subsequent coupling to a chromatography resin of a chromatography matrix.

[0095] According to another aspect, there is provided a chromatography matrix comprising the ligand as described hereinbefore coupled to the matrix.

[0096] The chromatography matrix is not limited to a specific matrix, but any matrix known to the skilled person may be utilized. For example, the chromatography matrix may be of natural or synthetic origin. Preferably, the chromatography matrix is a solid matrix, e.g. a polysaccharide matrix.

[0097] The ligand may be coupled to the matrix by any coupling technique known to the skilled person. For example, amino acids and / or carboxy group present in the ligand may be utilized. Alternatively, the ligand may be coupled to the solid support of the matrix by thioether bonds. In this regard, the ligand may be provided with a terminal cysteine residue for subsequent use in the coupling.

[0098] Alternatively, a spacer molecule may be introduced between the matrix support and the ligand to facilitate the coupling of the ligand to the matrix.

[0099] It is also conceivable that the ligand is coupled to the matrix by non-covalent bonding, such as physical adsorption or biospecific adsorption.

[0100] According to yet another aspect, there is provided a method for purification of sample proteins comprising: a) providing a chromatography matrix as described hereinbefore, b) adding metal ions to the chromatography matrix, c) adding a sample comprising target proteins, d) eluting bound target proteins, preferably at a pH from 5 to 8.

[0101] The sample comprising target proteins may be antibodies or parts thereof, such as IgG or Fab-fusion proteins. The antibodies or parts thereof may be present in complex sample mixtures, e.g. culture supernatants.

[0102] The step d) of eluting bound target proteins may be carried out at a pH from 5 to 8, preferably from 5.5 to 7.

[0103] The step d) of eluting the bound target proteins typically comprises adding an agent which disrupts the bond between the metal ions and the metal-binding motif. For example, the agent may be a metal-binding molecule, a salt, such as sodium chloride (NaCl) and / or a chelating agent, e.g. EDTA, EGTA, glutamate diacetate or citrate glutamate.

[0104] According to another aspect, there is provided a nucleic acid encoding a recombinant polypeptide as described hereinabove.

[0105] Hence, the present disclosure also encompasses a DNA sequence that can be used in the production of the recombinant polypeptides by expression thereof in a recombinant host according to well-established biotechnological methods.

[0106] The DNA sequence may be defined by SEQ ID NO:39 or a corresponding sequence having at least 90 %, preferably at least 95%, more preferably at least 99% sequence homology with SEQ ID NO: 39.

[0107] SEQ ID NO:39 encodes the recombinant polypeptide defined by SEQ ID NO:20.

[0108] Hence, according to a further aspect, there is provided an expression system comprising a nucleic acid as described hereinabove. Further features of, and advantages with, the present disclosure will become apparent when studying the appended claims and the following description. The skilled addressee realizes that different features of the present disclosure may be combined to create embodiments other than those described in the following, without departing from the scope of the present disclosure.

[0109] BRIEF DESCRIPTION OF THE DRAWINGS

[0110] The various aspects of the present disclosure, including its particular features and advantages, will be readily understood from the following detailed description and the accompanying drawings, in which:

[0111] Figure la illustrates the crystal structure of the C2 domain of streptococcal protein G (SPG), wherein the grafting site for insertion of a metal-binding loop is shown in dark grey.

[0112] Figure lb illustrates the crystal structure of a recombinant polypeptide according to an exemplary embodiment of the present disclose. The metal-binding loop insertion site is shown in dark grey.

[0113] Figure 1c illustrates the sequences of the wild type domains Cl, C2 and C3, as well as the sequence of the linker and the metal-binding motif inserted between the second beta strand and the alpha helix of the C2 domain.

[0114] Figures 2a-n illustrates the target interaction, measured by surface plasmon resonance (SPR), in the presence (black) and absence (grey) of calcium for a plurality of recombinant polypeptides according to exemplary embodiments of the present disclosure; i.e. C2CaEP7 (SEQ ID NO:20) (Fig. 2a), C2CaEP2 (SEQ ID NO:26) (Fig. 2b), C2CaEPi5 (SEQ ID NO:27) (Fig. 2c), C2CaEP4 (SEQ ID NO:28) (Fig. 2d), C2CaEP5 (SEQ ID NO:29) (Fig. 2e), C2caEP8 (SEQ ID NO:30) (Fig. 2f), C2Cav3 (SEQ ID NO:31) (Fig. 2g), C2Cav2 (SEQ ID NO: 32) (Fig. 2h), C2cav4(SEQ ID NO:33) (Fig. 2i), C2Cav6(SEQ ID NO:34) (Fig. 2j), C2Cavi (SEQ ID NO:35) (Fig. 2k), C2Cav5(SEQ ID NO:36) (Fig. 21), C2Cav7(SEQ ID NO:37) (Fig. 2m), and C2Cav8 (SEQ ID NO: 38) (Fig. 2n).

[0115] Figures 3a-h illustrate the elution profiles for a plurality of recombinant polypeptides according to exemplary embodiments of the present disclosure; i.e. C2caEP7 (SEQ ID NO:20) (Fig. 3a), C2CaEP2 (SEQ ID NO:26) (Fig. 3b), C2CaEPi5 (SEQ ID NO:27) (Fig. 3c), C2caEP4 (SEQ ID NO:28) (Fig. 3d), C2CaEP5 (SEQ ID NO:29) (Fig. 3e), C2CaEP7_F28R (SEQ ID NO:21) (Fig. 3f), C2CaEP7_F28Y (SEQ ID NO:22) (Fig. 3g), C2CaEP7_F28H (SEQ ID NO:23) (Fig. 3h) when applying 0.5 mg of each variant onto a 0.5 ml IgG sepharose column under neutral pH (7.5) conditions with elution by addition of 100 mM chelator followed by an acidic control elution using 0.5 M Hac, pH 2.8.

[0116] Figure 4 illustrates the elution profile for a recombinant polypeptide according to an exemplary embodiment of the present disclosure; i.e. ClcaEP? (SEQ ID NO:25) when applying 0.5 mg of each variant onto a 0.5 ml IgG sepharose column under neutral pH (7.5) conditions with elution by addition of 100 mM chelator followed by an acidic control elution using 0.5 M Hac, pH 2.8.

[0117] Figures 5a-c illustrate SPR sensorgrams of monomeric and tetrameric C2caEP7 showing a retained calcium-dependent interaction with mlgGl at pH 7.5 and improved binding capacity when multimerizing. Figure 5a represents the binding analysis of the interaction between C2caEP7 (SEQ ID NO:20) and full-length mlgGl during a concentration series ranging from 125 nM-1000 nM in the presence (black) and absence (grey) of calcium. Figure 5b illustrates the immobilization of equimolar amounts of monomeric (solid line) and tetrameric (dashed line) C2caEP7 with the same injection series of mlgGl showing a higher obtained binding capacity for the multimer. Figure 5c illustrates the tetrameric version of C2caEP?in the presence (dashed line) and absence (grey) of calcium, displaying the same calcium-dependency as the monomer when allowed to interact with injected mlgGl (125 nM-1000 nM).

[0118] Figure 6 illustrates sequences of the recombinant polypeptide according to exemplary embodiments of the present disclosure, which are able to switch between a targetbinding configuration in the presence of calcium ions, and a non-target binding configuration in the absence of calcium ions.

[0119] Figure 7 illustrates the test purification on the C2caEP7 monomer with pure mouse IgGl, showing successful capturing at pH 7 and elution with calcium chelating buffer. Test purification was done on a monomeric version of C2caEP7 coupled to resin matrix via thiol coupling in a table top column.

[0120] Figure 8 illustrates the evaluation of salt-dependent antibody binding on monomeric and tetrameric versions of C2caEP7 using table top columns, wherein salt sensitivity was observed for the monomeric version (figure 8a), but not for the tetrameric version (figure 8b). Test purification was done on the monomeric and tetrameric versions of C2caEP7 coupled to resin matrix via thiol coupling in a table top column.

[0121] Figure 9 illustrates the citrate elution of pure mouse IgGl on a tetrameric version of C2caEP7. C2caEP7 tetramer was coupled to resin matrix via thiol coupling and packed into a table top column. Successful elution was observed using citrate as chelating agent instead of EDTA.

[0122] Figure 10a illustrates a chromatogram of AKTA purification of mouse IgGl in HEK293 cell supernatant enriched with fetal bovine serum using the tetrameric version of C2caEP7, demonstrating successful capturing and elution using citrate as chelating agent.

[0123] Figure 10b illustrates SDS-PAGE of flow through (FT), wash (W), elution (E) and control elution (HAc) with a protein marker (M) showing molecular weights in kDa.

[0124] Figure 11 demonstrates the importance of a glycine residue in the first position of the alpha helix of the recombinant polypeptide. The original amino acid residue glutamic acid (E) was exchanged to a glycine (G) in position 35 in the recombinant polypeptides C2caV3_E35G defined by SEQ ID NO:42 (fig. I la), C2CaEPi5_E35G defined by SEQ ID NO:43 (fig. 1 lb), C2caEP4 E35G defined by SEQ ID NO:44 (fig. 11c), and C2caEP5_E35G defined by SEQ ID NO:45 (fig. l id). Figure l ie illustrates the loss of calcium-dependent elution when the glycine (G) in position 35 of C2caEP7 was replaced with glutamic acid (E) (SEQ ID NO:46).

[0125] DETAILED DESCRIPTION

[0126] The present invention provides novel recombinant polypeptides derived from the IgG-binding domains Cl, C2, and C3 of streptococcal protein G (SPG). The target binding of the IgG-binding domains has been rendered metal ion dependent by the introduction of a metalbinding motif. Accordingly, the recombinant polypeptides display a switchable target affinity depending on the presence of metal ions.

[0127] The recombinant polypeptides may be used as ligands for purification of immunoglobulins (Ig) by affinity chromatography. Such ligands may circumvent the currently harsh elution conditions associated with immunoglobulin G purification.

[0128] Thus, the invention provides a modular way of tailoring the target affinity of several protein domains with inherent antibody affinity and new tools optimizable for gentle IgG purification, particularly Fab-fragment purification.

[0129] According to a first aspect of the present disclosure, there is provided a recombinant polypeptide comprising an IgG-binding domain derived from Streptococcus Protein G (SPG) and a metal-binding motif, wherein the recombinant polypeptide has a first, target-binding configuration in the presence of metal ions, and a second, non-target-binding configuration in the absence of metal ions, wherein the recombinant polypeptide is configured to bind to the Fab-region of IgG in the target-binding configuration. According to a second aspect, there is provided a ligand comprising the recombinant polypeptide described hereinbefore.

[0130] According to a third aspect, there is provided a chromatography matrix comprising the ligand coupled to the matrix.

[0131] According to a fourth aspect, there is provided a method for purification of sample proteins.

[0132] According to a fifth aspect, there is provided a nucleic acid encoding a recombinant polypeptide as defined hereinabove.

[0133] According to a sixth aspect, an expression system comprising the nucleic acid is provided.

[0134] Experimental section

[0135] Materials and methods

[0136] Library creation, first generation

[0137] The library was designed to consist of the C2 domain with a randomized stretch of 12 amino acids between the second beta strand and the alpha helix. Upstream of this stretch a linker region of 0-3 amino acids was included and oligonucleotides encoding this library was obtained using trinucleotide synthesis (Twist Bioscience, San Francisco, CA, US). 200 ng of DNA was PCR-amplified during 20 cycles before performing traditional cloning to insert it into the pBad2.2 E. coli display vector (10). The final ligation product was electroporated into BL21(DE3) cells (Lucigen, Middleton, WI, US) which were cultivated in 100 ml LB media supplemented with 100 g / ml carbenicillin overnight at 37°C, 150 rpm. Cells were harvested through centrifugation the following morning and the practical library size was calculated to be 1.5 x 108variants through replicate titrations. Stocks containing 15% glycerol was made with 1.5 x 109cells and stored at - 80°C.

[0138] Library creation, second generation

[0139] Four previous selection outputs originating from selection round five and six from the library described above, extracted from cultivated cells harboring the genotype with QIAprep Spin Miniprep Kit (Qiagen, Hilden, Germany) according to the supplied protocol, was obtained. The obtained plasmid pools were subjected to four consecutive rounds of error prone-PCR using 0.1 pg template and the GeneMorph II kit (Agilent, Santa Clara, CA, US). The fragments were purified and introduced into the pBad2.2 E. coli display vector by traditional cloning. The ligation products were electroporated into BL21(DE3) cells (Lucigen, Middleton, WI, US) which were cultivated in 100 ml LB media supplemented with 100 pg / ml carbenicillin overnight at 37°C, 150 rpm. Centrifugation was used to harvest the cells the following morning and the practical library sizes was estimated to range from 2 x 106- 9 x 107variants.

[0140] Target preparation

[0141] Mouse IgGl (mlgGl) and human polyclonal Fab fragment (hFab) was ordered from Jackson ImmunoResearch (West Grove, PA, US) and biotinylated using EZ-link Sulfo-NHS-LC- Biotin (Thermo Scientific, Waltham, MA, US) according to the manufacturer’s instructions. The biotinylation was stopped by removing non-reacted biotin using a NAP5 desalting column (Cytiva, Uppsala, Sweden).

[0142] Escherichia coli display selections

[0143] 1.5 x 109cells were cultivated in 100 ml LB containing 100 g / ml carbenicillin at 37°C, 150 rpm until the OD reached 0.5-0.8 and the culture was induced with 0.6% w / v L- arabinose and grown overnight at 25°C.

[0144] The induced E. coli cells were washed with TBSC (50 mM Tris, 150 mM NaCl, 1 mM CaCh, pH 7) before being resuspended in the same buffer but containing 150 nM of biotinylated mlgGl or 500 nM of biotinylated hFab and incubated on a rotamixer at room temperature (RT) for two hours. Next, the cells were washed with TBSC, resuspended in 150 nM human serum albumin (HSA)-Alexa 647 conjugate and 2 pg / ml streptavidin conjugated with R- phycoerythrin (SAPE) (Invitrogen, Waltham, MA, US) and incubated dark on ice for 30 minutes. The cells were subsequently washed and lastly resuspended in cold TBSC before being sorted using a MoFlo Astrios EQ cell sorter (Beckman Coulter, Brea, CA, US). The sort gate was set to capture cells displaying the library (detected library expression with (HSA)-Alexa-647 conjugate, X-axis) and binding to the target (detected target binding with SAPE, Y-axis).

[0145] Sorted cells were collected in a 15 ml tube containing LB media and was incubated during rotation for one hour at 37°C before being inoculated to 150 ml of LB media supplemented with 100 pg / ml carbenicillin and cultivated overnight at 37°C, 150 rpm. Six rounds of selection in solution were performed following this procedure, round 2-6 also included a negative selection step using magnetic activated cell sorting (MACS) prior to the positive fluorescent activated cell sorting (FACS) described. MACS was implemented to remove non-calcium dependent binders by pre-incubating target (150 nM mlgGl or 500 nM hFab) with streptavidin-coated Dynabeads™ MyOne™ Streptavidin Cl (Thermo Scientific) before allowing the cells to be incubated with the target-coated beads for 2 hours in MBSE (25 mM MES, 150 mM NaCl, 100 mM EDTA, pH 5.5). The magnetic beads were thoroughly washed, captured and the resulting supernatant was used as input for FACS.

[0146] For the second generation, a positive FACS round was performed as described prior to performing several MACS-rounds where the captured target-binding cells were eluted from the magnetic beads with MBSE, pH 7.

[0147] Subcloning, protein production & purification

[0148] To produce and characterize promising candidates, single clone genes were amplified from the display vector into expression vector pET-45b(+) (MilliporeSigma, Burlington, MA, US) using primers to introduce anN-terminal Hise-tag and the Xhol and Ncol restriction sites. The resulting plasmids were sequence-verified through Sanger sequencing by Eurofins (Luxembourg, Luxembourg). E. coli BL21(DE3) was used for production and the cells were subjected to sonication preparatory to purifying the supernatant using table-top columns packed with IgG Sepharose 6 Fast Flow resin (Cytiva).

[0149] The buffers used for purification were MBSC (25 mM MES, 150 mM NaCl, 1 mM CaCh, pH 6) for equilibration, 5 mM NH4AC supplemented with 1 mM CaCh, pH 6 for washing and 0.5 M HAc, pH 2.8 for elution. Upon elution the samples were buffer exchanged on PD-10 Desalting columns (Cytiva) to TBSC and the final purity was analyzed using SDS-PAGE.

[0150] Surface plasmon resonance (SPR)

[0151] The affinity to the native targets mlgGl, human polyclonal Fab- and human polyclonal Fc was determined using a Biacore T200 instrument (Cytiva). By amine coupling, each target was diluted in 10 mM NaAc, pH 4.5 and immobilized in separate flow cells on a Series S CM5-chip. Single-cycle kinetics was obtained at 25°C through injection of a two-fold dilution series (1000 nM, 500 nM, 250 nM, 125 nM) at a flow-rate of 30 pl / min. The experiment was performed both with TBSCT (TBSC, 0.05% (v / v) Tween20, pH 7) and TBSET (TBS, 100 mM EDTA, 0.05% (v / v) Tween20, pH 7) as running buffer and after each cycle the surfaces were regenerated with 10 mM HC1.

[0152] The responses obtained were double referenced against a blank flow cell and a corresponding buffer injection preceding determination of affinity constants applying a 1 : 1 Langmuir binding model. Affinity chromatography experiments

[0153] 0.5 ml of IgG Sepharose 6 FF (Cytiva) was packed in a column and pulsed with 5 CV 0.5 M HAc (pH 2.8) and 20 CV 1 * TBS-C, pH 7.5 (50 mM Tris, 150 mM NaCl, 1 mM CaC12). 0.5 mg of pure protein (each variant) was added followed by a wash with 10 CV TBS-C. 10CV of elution with 100 mM EDTA in TBS, pH 7.5, 0.5 ml fractions. After elution, a control elution with 3 CV of 0.5 M HAc, pH 2.8 was performed. All eluted fractions were measured at 280 nm.

[0154] Results

[0155] Design of a metal-coordinating library

[0156] The protein structure of C2 including its previously characterized antibody binding surfaces was investigated to find a suitable insertion site for a metal-coordinating loop that could render the inherent target affinity calcium-dependent. According to published crystal structures (PDB: 1FCC (1) and 1IGC (2)) the binding surface for Fab is mainly located to the second betasheet and the C-terminal end of the alpha helix while the interaction with Fc is identified to be present at the C-terminal end of the alpha helix, extending into the third beta-sheet. Hence, it was hypothesized that grafting of a calcium-binding loop preceding the alpha helix might infer a calcium-dependency without risking complete loss of functionality.

[0157] Measures to further prevent disrupting the protein structure were taken by including the option of a flexible linker N-terminally of the loop. Therefore, the library was designed to include four different linker lengths of 0-3 amino acids chosen to allow for different characteristics such as polarity (Ser / Thr), flexibility (Gly), and rigidity (Pro). Following the linker, the library was intended to cover a large number of possible calcium-binding loops consisting of 13 amino acids with designed randomizations in all but two positions (Figure 1).

[0158] Enrichment of calcium-dependency throughout the selection process

[0159] Directed evolution using bacterial display was performed towards full-length mouse IgGl and polyclonal human Fab-fragments in order to enrich for variants with calcium-dependent target interaction. The strategy consisted of alternating negative and positive selection pressures. The negative selection was performed in a calcium-free environment and aimed at eliminating variants that was still capable of interacting with the target independent of the presence of calcium. The positive selection was carried out with fluorescence-activated cell sorting to isolate variants with retained target affinity when supplemented with calcium. After six selection rounds, a targetbinding cell population seemingly calcium-dependent could be seen. Second generation library

[0160] To further improve the binders, the selection output from round 5 and 6 was used as starting material for error-prone PCR aiming to insert an average of two to four amino acid substitutions per variant, randomly distributed across the domain.

[0161] Bacterial display was continuedly used as the selection system with alternating negative and positive selection rounds. Although, this time the designed strategy included using magnetic-activated cell sorting (MACS) also to isolate variants which interacted with the inherent target in the presence of calcium but could be eluted from the magnetic beads by removal of calcium.

[0162] Inherent affinity for native targets evolved to be calcium-dependent

[0163] To investigate the conditional binding properties of the variants deeper, surface plasmon resonance was used. Full-length mlgGl and digested mFab and mFc fragment of the same antibody was immobilized on separate surfaces prior to single-cycle injections consisting of four concentrations of each variant (125 nM, 250 nM, 500 nM and 1000 nM), both in the presence and absence of calcium. All interesting variants as well as one of the variants that seem to have a calcium independent binding (C2caEP6) were included in the analysis. All variants except for C2caEP6 displayed a calcium-dependent binding at pH 7 both towards full-length mlgGl and mFab while no binding signals could be obtained towards the mFc surface. Interestingly, the binding capacity seemed to vary between the variants with higher obtained responses both towards mlgGl and mFab for C2CaEP2 (defined by SEQ ID NO:26), C2CaEP4 (defined by SEQ ID NO:28), C2CaEP5 (defined by SEQ ID NO:29), and C2caEPi5 (defined by SEQ ID NO:27) on par with the nonconditional variant C2caEP6 (Fig. 2).

[0164] It was further elucidated, using a similar experimental set-up, that C2cav3 (defined by SEQ ID NO:31) was showing an intact calcium-dependent human polyclonal Fab-binding while interaction with human polyclonal Fc-fragments could not be detected (Fig. 2g).

[0165] Additionally, the measured affinities for C2cav3 in comparison to C2wt seemed to be in the same pM-range or slightly improved towards both murine and human antibodies when measuring target interaction at neutral pH 7.0.

[0166] Evaluation of the calcium switch in a chromatography setting

[0167] When initially evaluating novel binders in the intended application, a purification set-up with neutral pH in every step from equilibrium till elution was utilized. The variant of interest was captured on an IgG Sepharose column and eluted by addition of a chelator. Here, a number of different variants were examined, both from the two selection rounds as well as additional variants to understand the importance of different amino acids within and around the calcium chelating loop. A control elution with acidic pH (2.8) was performed to evaluate residuals captured after chelator encounter (Fig. 3 and 4).

[0168] Interestingly, sequencing of C2caEP7 revealed one mutation in the scaffold located to the first position succeeding the inserted calcium-binding loop (i.e. the first amino acid residue of the alpha helix) showing a glutamate exchanged for a glycine. The importance of this evolved feature was investigated by back-mutating it to a glutamate residue in a version referred to as C2caEP7 GSSE. When this version was tested for calcium-dependent elution in the exact same gentle purification set-up as previously, a majority of the sample remained bound to the column until the control elution with low pH. Additionally, this mutation, changing the original glutamate in the first position of the alpha helix for a glycine was tested in C2cav3, C2caEPis, C2caEP4 and C2caEPis creating: C2caV3_E35G, C2caEPis_E35G, C2caEP4 E35G, and C2caEP5 E35G. Once again showing the importance of this evolved feature since when these versions were tested for calcium-dependent elution in the exact same gentle purification set-up as previously, a majority of the sample could now be eluted from the IgG- sepharose simply by removing calcium through the addition of a chelator. Further, the tolerance to different amino acids in position 25 and 28 was evaluated; i.e. position X? and Xio of the metal-binding loop. These recombinant polypeptides are defined by SEQ ID NO:21 (C2caEP7_F28R), SEQ ID NO:22 (C2caEP7_F28Y), SEQ ID NO:23 (C2caEP7_F28H), and SEQ ID NO:24 (C2caEP7_N25K).

[0169] The change to a positively charged arginine or the larger amino acid residue tyrosine in position 28 (Xio of the metal-binding loop) resulted in an improved binding capacity seen by less protein leaking through the column and ending up in the flow-through.

[0170] Hypothesizing that once a switch has been engineered it could be inserted into any available domains with similar prerequisites. Thus, the evolved calcium-binding loop, including subsequent position 35 (corresponding to the first amino acid of the alpha helix) of seemingly high importance, was inserted in the IgG-binding domain Cl of Protein G. Again, the mild purification procedure was evaluated with this variant (ClcaEP? defined by SEQ ID NO: 25) and it was observed that the incorporated calcium-dependency worked almost equally well in the Cl domain. Multimerization for increased binding capacity

[0171] Candidate C2caEP7 was evaluated further by multimerizing it to become a tetramer produced with a C-terminal cysteine alongside the monomeric version. Through thiol coupling, both monomer and tetramer C2caEP7 was immobilized to equimolar amounts on separate chip surfaces. A two-fold dilution series of full-length mlgGl (125 nM- 1000 nM) was injected and the calcium-dependent binding previously eluded for the monomer could be seen again in this reversed set-up (Figure 5). The tetramer demonstrated higher binding signals, showing that a multimeric version would have improved binding capacity. Moreover, the complete calciumdependency was still retained for the tetrameric version of C2caEP7.

[0172] Capturing and elution of pure mouse IgGl on C2caEP?

[0173] The performance of the C2caEP? monomer coupled on SulfoLink™ Coupling Resin via thiol coupling was tested with 1 mg pure mouse IgGl and elution at pH 7 with a chelating buffer (Figure 7). All fractions from sample application to elution were collected, and, through absorbance measurements at 280 nm, the percentage of protein in each fraction were estimated. The sample was applied in 4 ml of 5 mM NH4AC, pH 7 containing 10 mM CaCh. 10 CV of wash buffer with 5 mM NH4AC, pH 7 containing 1 mM CaCh were performed. No loss during flow through and washes was observed, indicating good capture performance. For elution, 15 CV of 5 mM NH4AC containing 100 mM EDTA, pH 7 was used followed by 10 CV of 0.5 M HAc pH 2.8 for control elution.

[0174] The applied sample could be completely eluted with the chelating buffer (figure 7), showing a successful elution on the monomeric version with the complete purification procedure performed at pH 7.

[0175] To optimize the buffer conditions and to be able to purify from crude supernatant, salt dependency of antibody binding was tested for the monomeric and tetrameric versions of C2caEP7 by washing with 5 mM NH4AC, pH 7 supplemented with 1 mM CaCh and eluting stepwise with 5 mM NH4AC, pH 7 supplemented with either 50 mM NaCl, 100 mM NaCl or 150 mM NaCl during 5 elution steps each (figures 8a-b).

[0176] The monomeric version showed elution of the antibody at a salt concentration of 100 mM NaCl (figure 8a). The tetrameric version showed no elution of antibody (figure 8b) for the tested salt concentrations of 50 mM, 100 mM and 150 mM of NaCl. In comparison, this indicates a stronger antibody binding of the tetrameric version than for the monomeric version of C2caEP7, hypothesizing that this would most likely be due to avidity effects. Based on these results, it was assumed that a salt containing buffer would work for sample application and washing of the tetrameric column.

[0177] Elution using citrate as chelating agent

[0178] To aim for an environmentally friendly purification, elution was performed with citrate instead of EDTA (figure 9). Using citrate as calcium chelating buffer 97 ± 3 % of the captured antibody was eluted from the tetrameric version. The sample was applied in 4 ml of TBS containing 10 mM CaCh. Washing buffer: TBSC, pH 7. Elution buffer: TBS, 100 mM citrate, pH 7. Control elution was done with 0.5 M HAc, pH 2.8. The tests were performed in triplicates. The percentage of antibody was determined by OD280 measurement of each sample.

[0179] In both cases, for the monomeric and the tetrameric version, the complete sample was eluted within elution fraction 2-7 resulting in a concentrated peak of 5 ml which was similar to the elution peak seen with EDTA. Altogether, this is a positive indication that an environmentally friendly and mild purification is possible.

[0180] Furthermore, additional tests were carried out where the C2caEP7 tetramer was coupled to resin matrix via thiol coupling and packed into an AKTA column with a CV of 1.5 ml.

[0181] Mimicking a real purification process, 35 ml of HEK293 cell supernatant was supplemented with 10 % fetal bovine serum (FBS) and 0.5 mg mouse IgG, and used as the sample to be purified. 10 CV washes were executed with TBSC and 15 CV elution was performed with TBS, 100 mM citrate, pH 7. Figure 10a shows a section of the corresponding chromatogram, zoomed into the elution peak. Even with crude sample, the column showed no loss of antibody during sample application (FT) and washes while still displaying a concentrated elution of antibody within 5 ml once calcium is depleted. A control elution was done with 0.5 M HAc, pH 2.8 afterwards (not included in the chromatogram) which showed that a negligible amount of antibody came with the acidic control elution. As illustrated in figure 10a, successful capturing and elution with citrate as chelating agent was observed when using the tetrameric version of C2caEP7.

[0182] The purity of the eluted antibody was verified by SDS-PAGE (figure 10b). As can be seen in figure 10b, the antibody appears only in the “elution (E)” fraction resulting in two bands corresponding to the heavy (50 kDa) and light (25 kDa) chain. Evaluating the importance of glycine in the first position of the alpha helix

[0183] In order to understand the molecular features as well as the different amino acids importance for the molecular switch, binding to release upon deprivation of calcium, a number of C2cavariants were constructed. Here, the impact of amino acid number 35 was evaluated.

[0184] The following recombinant polypeptides were tested: a) C2caV3 E35G defined by SEQ ID NO:42 (corresponding to SEQ ID NO:31, but with the glutamic acid in position 35 being replaced by glycine) b) C2CaEPi5 E35G defined by SEQ ID NO:43 (corresponding to SEQ ID NO:27, but with the glutamic acid in position 35 being replaced by glycine) c) C2caEP4 E35G defined by SEQ ID NO:44 (corresponding to SEQ ID NO:28, but with the glutamic acid in position 35 being replaced by glycine) d) C2caEP5 E35G defined by SEQ ID NO:45 (corresponding to SEQ ID NO:29, but with the glutamic acid in position 35 being replaced by glycine) e) C2ca EP7 G35E defined by SEQ ID NO:46 (corresponding to SEQ ID NO:20, but with the glycine in position 35 being replaced by glutamic acid)

[0185] As illustrated in figures 1 la-d, the replacement of the amino acid glutamic acid

[0186] (E) in position 35 with a glycine (G), i.e. in the first position of the alpha helix of the recombinant polypeptides a)-d) hereinabove, resulted in improved calcium-dependent elution. Furthermore, replacement of the amino acid glycine (G) in position 35 with glutamic acid (E) resulted in loss of calcium-dependent elution during purification (figure l ie).

[0187] Discussion

[0188] Naturally occurring proteins have evolved to incorporate and transmit signals or regulate their level of activity through induced conformational changes in the protein structure. Inspired by this fine-tuned control of protein functionality, the inventors made an engineering effort to equip an existing antibody-binding domain with metal-dependency using a combinatorial library displayed on bacterial cells. The library was designed to be introduced N-terminally of the alpha helix in the C2 domain. Further, this insertion site was believed to be in close enough proximity to both inherent binding surfaces of C2 thereby endowing calcium-dependent target affinity without risking destroying the delicate protein structure completely. By implementing cell display it was possible to early on observe that the protein structure of the library members seemed intact as they evidently could be expressed on cells and interact with the native target. The sequential utilization of MACS and FACS enabled directed removal of non-calcium dependent variants during the negative selection and focused enrichment of variants ensured to display target interaction in the presence of calcium.

[0189] Interestingly, sequencing of the screened variants showing promising calciumdependent characteristics indicated a preference for a linker length of two amino acids preceding the introduced loop although the number of sequenced variants was too low to be of significant statistical importance. Furthermore, rigid (Pro) and polar (Ser / Thr) amino acids were observed in the linker and the positively charged amino acid (Lys) was dominant in position eleven. Additionally, the design of the loop with residues (Asp / Asn) expected to coordinate calcium was seen in the sequencing data and demonstrated through the structural determination. By utilizing error-prone PCR to produce a second-generation library for another selection campaign, the inventors were able to further enhance the calcium dependency of the binders and also gain more information about the importance of the residues in and around the grafted loop. Notably, it was concluded that the glycine in position 35 is important to retain the calcium dependence.

[0190] Terms, definitions and embodiments of all aspects of the present disclosure apply mutatis mutandis to the other aspects of the present disclosure.

[0191] Even though the present disclosure has been described with reference to specific exemplifying embodiments thereof, many different alterations, modifications and the like will become apparent for those skilled in the art.

[0192] Variations to the disclosed embodiments can be understood and effected by the skilled addressee in practicing the present disclosure, from a study of the drawings, the disclosure, and the appended claims. Furthermore, in the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0193] References

[0194] 1. A. E. Sauer-Eriksson, G. J. Kleywegt, M. Uhlen, T. A. Jones, Pii: 80969-2126(01)00157- 5. Structure 3, 1-14 (2004). 2. J. P. Derrick, D. B. Wigley, The third IgG-binding domain from streptococcal protein G:

[0195] An analysis by X-ray crystallography of the structure alone and in a complex with Fab. J. Mol. Biol. 243, 906-918 (1994).

[0196] 3. F.L. Moffatt Jr., S.A. Gulac, A.N. Sarafini, G.N. Sfakanckis, R. Dop, D.S. Robinson, D. Fransceschi, J. Boggs, A.S. Livingstone, A thousand points of light or just a dim light bulb? Radiolabeled antibodies and colorectal cancer imaging, Cancer Investigation, 17 (1999), pp. 322-334.

Claims

CLAIMS1. A recombinant polypeptide comprising an IgG-binding domain derived from Streptococcus Protein G (SpG) and a metal-binding motif, wherein said recombinant polypeptide has a first, target-binding configuration in the presence of metal ions, and a second, non-target-binding configuration in the absence of metal ions, wherein said recombinant polypeptide is configured to bind to IgG in said first, target-binding configuration.

2. The recombinant polypeptide according to claim 1, wherein said IgG-binding domain is domain Cl or domain C2.

3. The recombinant polypeptide according to claim 1 or claim 2, wherein said metal -binding motif is inserted N-terminally of the alpha helix of said IgG-binding domain, preferably between the second beta strand and the alpha helix of said IgG-binding domain.

4. The recombinant polypeptide according to any one of the preceding claims, wherein said recombinant polypeptide is configured to bind to the Fab-region of said IgG in said first, target-binding configuration.

5. The recombinant polypeptide according to any one of the preceding claims, wherein said metal-binding motif is a calcium-binding motif, and wherein said metal ions are calcium ions.

6. The recombinant polypeptide according to any one of the preceding claims, wherein said metal-binding motif is a metal-binding loop comprising from 10 to 14 amino acids, preferably from 12-13 amino acids.

7. The recombinant polypeptide according to any one of the preceding claims, wherein said metal-binding motif is defined by the following sequence:X4X5X6X7X8X9XioXi 1X12X13X14X15X16 (SEQ ID NO: 1), whereinX4is aspartic acid (D),Xs is alanine (A), arginine (R), glutamic acid (E), lysine (K), serine (S), threonine (T), leucine (L) or valine (V),Xe is aspartic acid (D), or asparagine (N),X7 is arginine (R), asparagine (N), glycine (G), or lysine (K),Xs is asparagine (N), aspartic acid (D), or serine (S), preferably aspartic acid (D),X9 is glycine (G),X10 is arginine (R), glycine (G), lysine (K), phenylalanine (F), threonine (T), histidine (H), or tyrosine (Y), preferably phenylalanine (F), tyrosine (Y) or arginine (R), Xu is isoleucine (I), leucine (L), or valine (V),X12 is asparagine (N), aspartic acid (D), glutamic acid (E), serine (S), or threonine (T), preferably asparagine (N) or aspartic acid (D),X13 is alanine (A), phenylalanine (F), threonine (T), or tyrosine (Y),X14 is aspartic acid (D), glutamic acid (E), or lysine (K)X15 is aspartic acid (D), or glutamic acid (E), preferably aspartic acid (D), Xi6 is alanine (A), leucine (L), phenylalanine (F), or valine (V).

8. The recombinant polypeptide according to any one of the preceding claims, further comprising a linker upstream of said metal-binding motif, wherein said linker comprises 1 to 4 amino acids, preferably 2 to 3 amino acids.

9. The recombinant polypeptide according to claim 8 when dependent on claim 7, wherein said metal-binding motif comprising said linker is defined by:XiX2X3X4X5X6X7X8X9XioXi 1X12X13X14X15X16 (SEQ ID NO: 2), whereinXi is serine (S), threonine (T), glycine (G), proline (P), or absent,X2 and X3 are selected from serine (S), threonine (T), glycine (G) or proline.

10. The recombinant polypeptide according to any one of the preceding claims, wherein said metal-binding motif is defined by a sequence selected from any one of SEQ ID NO:6-19 or a corresponding sequence having at least 80%, preferably at least 90 %, more preferably at least 95% sequence homology with any one of SEQ ID NO:6-19.

11. The recombinant polypeptide according to any one of the preceding claims, wherein said IgG-binding domain comprises at least one mutation, wherein said at least one mutation is a replacement of glutamic acid (E) with glycine (G) in the first amino acid residue of the alpha helix of said IgG-binding domain.

12. A ligand comprising a recombinant polypeptide according to any one of claims 1-11.

13. A chromatography matrix comprising the ligand according to claim 12 coupled to said matrix.

14. A method for purification of sample proteins comprising: a) providing a chromatography matrix according to claim 13, b) adding metal ions to said chromatography matrix, c) adding a sample comprising target proteins, d) eluting bound target proteins, preferably at a pH from 5 to 8.

15. Nucleic acid encoding a recombinant polypeptide as defined in any one of claims 1-11.

16. An expression system comprising a nucleic acid according to claim 15.

Images