See also how to attach Multi-domain ligands to the resin

Berg (WO2006/065208) discloses a matrix to which antibody binding protein ligands such as Protein A, G and/or L have been immobilised. In a preferred embodiment, the ligands comprise a monomer, dimer or multimer of Protein A domains (e.g., one or more of Domain A, B, C, D and E domains or a dimer or multimer of Protein Z). The ligands may be immobilised using well known methods such as epoxi coupling. In a specific emobdiment, the ligand density is in the range of 5-10 mg/ml. 

Berg (US 2006/0134805) also discloses monomer, dimer or multimer of Protein A domains such as the A, B, C, D and E, preferably B or C domains. In one embodiment, such a dimer or multimer comprises Protein Z, which is the mutated form of Domain B.

Godavarti (WO 2006/138553 A2) teaches methods of purifying polypeptides having an Fc region such as antibodies or antibody fusions by adsorbing the polypeptides to Protein A or Protein G, followed by a wash with a divalent cation salt buffer to remove impurities and subsequent recovery (pH 2-4). In various emobdiments, the Fc binding protein comprises one or more Protein A and Protein G. 

Gore (WO92/09633) discloses a polypeptide which has at least 2 but not more than 4 binding domain which possess a high degree of sequence homology with the binding domains of SpA , each capable of binding to the Fc region of IgG.

Jendeberg (J. Molecular Recognition, 8, 270-278 (1995) disclsoes by increasing the valency of an analogue of SpA, going from one to two domains, the affinity for Fc1 increases by almost two orders of magnitudesbut no significant increases in affinity was detected going to five domains. Jendeberg describes four Z mutant proteins (L17D), (N28A), (I31A) and (K35A) which decreased affinity to human IgG1 and CD4-Fc1.

Kihira (US 5,580,788 and EP 0550771A1) examined in detail the relationship between the IgG binding ability of the repetitive protein subunits of protein A (E, D, A, B and/or C) such as repetitive units of the AB domain and found that the binding capacity of the repetive proteins for IgG increased in proportion with the increase in the number of repeat of the IgG binding domain (p. 5, lines 51-56; claims 1-4). Kihira also diclsoes that 1-6 number of such repeats (Table 1) and that the methods used may be used in the case of the other domains to easily produce the desired repetitive proteins between identical or different domains, and thus obtained repetitive proteins may be used to purify IgG to a high degree (Example 3, p. 6, lines 30-34). 

Lee (WO95/06125) discloses a peptide comprising binding domains of protein A and protein G (at least one binding domain from each protein). In one embodiment, a gene cassette which encodes for binding domains of both Protein A and Protein G is provided. 

Ljungberg (Mol Immunol. 1993, 30(14): 1279-85) discloses intact as well as SpA derived fragments cotnaining 1-5 IgG binding domains of different compositions and that while all the proteins bound to IgG, regardless of size or composition, the binding strengh differed significantly. Proteins containing 5 domains have a stronger affinity for IgG than those containing 1-or 2. 

Particular Domain Combinations

A-B Domains:

Kihira (US 5,580,788 and EP08632100 discloses linking the A and B domains of SpA which results in superior ability to purify IgG.

C Domain:

Bian (US 16/113403, pulbished as US 2018/0362595) disclsoes two or more C domains of SpA attached to a chromatography resin at more than one site on the resin where each C domain includes the WT C domain having a mutation to replace the glycine at postion 29 with an amino acid other than alanine or tryptophan that reduces Fab binding. 

Hall (WO2008/039141) discloses at least two Domain C from SpA as a multimeric chromatography ligand.

Majima (US 14/916,316, published as US 2016/0215027) also teaches a multimeric immunoglobulin binding protein having 6-10 total domains from SpA such as the C domain having the general structure (R1)n-(R2m) or (R2m-(R1)n where the R1 domain is an amino acid sequence in which a non-lysine amino acid has been replaced such as the lysine being substituted with a non-lysine amino acid at 1-3 of positions 4, 7, and 35 R2 is an immunoglbulin binding domain coccuring at the N-temrinus or the C-terminus of the protein and includes an amino acid residue that covalently bonds to an insoluble support and the R2 domain is based on an amino acid sequence in which the lysine residue(s) are substituted with a non-lysine amino acid only at position 35, or at posiiton 35 and one or more positions 4, 7 and 35. In one embbodiment the R2 domain includes substitutions of 1-6 amino acid reisdues with lysine at positions 40, 43, 46, 53, 54, and 56. Because only the (R2) domains are selectively immobilized on the support via a covalent bond, it is possible to acheive a highly selective immobilization reaction trhough for example a lysine or a cysteine residue. The R1 in trun has an amino acid sequence that does not contain an amino acid that is active to the chemical reaction used for immobilization. When the immobilization reaction used to immobilzie the prtoien on the support takes place via an amino group, an R2 domain can be produce by substituting the lysine residues contained in the amino acid sequence with non-lysine aino acids only in lysine residues occurring at positiosn that interfere with bidnign to an immunoglobulin upon immobilizing the protein on the support and by substituting some of the non-lysine amino acids not involved in bidning to an immunoglobulin with lysine. In the case of immobilization via a thiol group, a new cysteine residue can be added to the R2 domain. A support immobilization reaction using a disulfide bond or a maleimide group that is highly selective to the thiol group may be used for said immobilization.

–Commercially available: 

Toyopearl AF-rProtein A-650F: is a tetramer of the non-Fab binding C domain which exhibits DBC of greater than 30 g/L for human IgG at 2 minutes residence time. The resin is avialble in 45 um F grade particle size. It is availabe from TOSOH. 

Z domain:

–Z4: MabSelect SuRe (GE Healthcare) is a commercially available Protein A stationary phase that incorporates a ligand, designated Z4, comprising four repeats of the Z domain engineered to be alkaline resistant. A linker with a C terminal cysteine is added to the last repeat to facilitate coupling to the resin. (Pabst, J. of Chromatography A volume 1362, 2014)

Multimers with Asparagine Substitutions in the Domains

Hober (US2006/0194955 and WO 03/080655) teaches an immunoglobulin binding protein such as Staphylococcal protein A (SpA) derivative/analogue wherein at least one asparagine residue of the B-domain and protein Z has been mutated to amino acids other than glutamine or aspartic acid (i.e., such as lysine and leucine) resulting in a ligand with higher binding capacity under alkaline conditions (see above). Hober also teaches that such protein monomers can be combined into multimeric proteins, such as dimers, trimers, tetramers, pentamers etc.

Multimers with Lysine Deletions/Substitutions in the Domains

Majima (EP1992692A1 and US 12/280221, published as US 2010/0286373; also disclose that in the case of using an immunoglobulin-binding protein derived from protein A as an affinity chromatography ligand for an immunoglobulin, a multimer protein abtained by ligating two or more, desirably about 4 binding domains have been conventionaly produced and used. Majama further disclose a modified C-domain of SpA or Z-domain which has improved orientation for maintaining affinity for an immunoglobulin and improved chemical stability under acidic pH conditions because a ratio of the number of lysine at positions 39 onwards to the number of lysine at position. Majima (US 14/916,316, published as US 2016/0215027) also discloses six or ore domains of the C domain with lysine substitutions. 

–In the ligation sequence

Yoda (US 15/669498, published as US 2107/0333811) discloses one of any of the domains of Protein A (E, D, A, B, C and Z) wherein in the amino acid sequence of at least one domain one or more lysines are included and the C terminal lysine is deleted or substituted. In a particular emobdiment, one or more lysines are includes and lysine at position 4 and the C-terminal lysine is deleted or substituted with a hydrophilic amino acid. Specifically, then the C terminal of each domain and a sequence at positions 1-5 in the amino acid sequence of each domain are defined as a ligation site sequence, the C terminall lysine and lysine at position 4 in at least one of the ligation site sequences are dleted or substituted. The amino acid at position 4 is the fourth amino acid from the N teminus of each of the B, C and Z domains of Protein A is lysine. The E domain is treated by assuming that the amino acid residue at postiions 1 and 2 are deleted based on the region with a high homology with the other domains. The D domain is treated by assuming that the third to fifth amino acid reisdues from the N teminus are inserted based on the region with a high homoloy with the other domains. In other wordds, the lysine at position 4 of the domain closest to the N temrinus or the C terminal lysine of the daomin closest to the C terminus is deleted or substituted. By doing this, it was found that by forming a mutated ligation site sequence such that specific lysine (K) which is present in the ligation site sequence is not included, the cleavage of the protein by a serine protease can be avoided. Lysine (K) and arginine (R) present other than in the ligation site sequence, however, are less likely to be cleaved by a serine protease and thus may be present in each domain. 

Ligation of Domains (linkers/spacers):

Linkage of SpA domains:

Honda (US 9,382,297) discloses a modified protein of an extracellular domain of protein A which has the reduced ability to bind to immunoglobulin in an acidic region. Teh amino acid sequence of the modified protein may be a tandem type amino acid sequence in which the amino acid sequence and aribtrary linker sequences are alternately arranged as a plurality of repeats. Such a sequence might be for example (amino acid sequence a -linker sequence A – amino acid sequence a-Linker sequence B -amino aequence a or it could be amino acid sequence a- linker sequence C amino acid sequence b -linker sequence D – amino acid sequence c. 

Rodrigo (WO 2017/1494596) discloses an Fc binding polypeptide with improved alkli stability which include a mutant of a Fc binding domain of SpA. The ligand can be in the form of multimers such as a dimer, a trimer, a tetramer, pentamer, hexamer, heptamer, octamer or a nonamer, which have higher alkali stability than monomers. All units in the multimer can b identical or it can be a heteromultimer, where at least one unit differs form teh others. The polypeptides can be liniked to each other direclty by peptide bonds between the C-temrinal and N-temrinal ends of the polypepitdes. 

Qian (US 2016/0237124) discloses a multimer of a protein A mutant that includes at least two protein A mutants wherein the at least two protein A mutants includes the same amino acid sequence or different amino acid sequences. The multimer can be a dimer, trimer or tetramer. The one or more protein A mutatns are linked together via one or more linker comrpsing 4-10 amino acids. 

Linkage of Domain C units from SpA:

Hall (WO 2008/039141) discloses multimeric chromatography ligands from Domain C from SpA or a functional fragment or variant therof comprised of at least two Domain C nits. In one emboidment, the multimer is comprised of 2-8 units, such as 4-6 units. Linkers may be inserted between the multimer units. In one embodiment, the chromatogrpahy ligand includes no other SpA domains than Domain C. 

Lenght of Linker:

Honda (US 2016/0280744) also discloses domains dervied from protein G or protein A linking the plural domains having affinity for the protein comprising the Fc region of IgG. A polypeptide linker consisting of amino acids is preferable. The number of domains may be optionally selected depending on the use of the protein and the kinds of the domains. The protein can be a dimer, trimer, tetramer or pentamer. The linker has an extended lenght of 8-240 angstrom. The number of the amino acids constituting the peptide linkers is usually about 22-26, preferably about 24-60, dpending on the kind of the amino acids. For example, the peptide linker is preferably 4-10 times repeats of the unit of GlyGlySerGlyGlySer, or mroe preferably 6-10 times rpeats of the same unit. 

Particular Types of Linkers/Spacers:

Zong (US 16,485,854, published as US 2019/0375785) discloses multimers such as tetramers and pentamers of Protein A domains and the Z domain which are connected in series with a spacer. the spacer domain includes an alpha helix or a helix bundel. The alpha helix can contain about 8-32 amino cids, which can form about 2-8 helical turns and ranges from about 10-45A in lenght. The spacer domain can also be a domain of glucagon, a connecting helix of calmodulin, or a single alpha-helix domain. In some embodiments, the spacer is a small protein such as a Sumo domain or an EGF domain. It was discovered that each IgG binding domain of SpA may have a secondary minor Fc binding site in addition to the well-known major Fc binding site. The secondary interaction occurs when the adjacent immunoglobulin binding domain binds to the dominant Fc site. The major and secondary interactions contribue to an high affinity between SpA and IgG due to cooperative effects. In order to eliminate the secondary binding, the neighboring immunoglobulin binding domains can be separated by a spacer. domain. Zong exemplofies a helical domain spacer from the central segment of the connecting helix in ribosomal protein L9 from B. sterothermphilus. The L9 connecitng helix, which is conserved in a number of organissms, was shwon to be exceptionally stable. It has bout 18 residues (forming aobut 4 helical turns and having roughly the same lenght as the Z domain). This tetraZ-H had significantly higher elution pH than Tetra Z connected in eries without the spacers. 

Particular Substitutions

Position 23:

Johansson (US14/358821, published as US 2014/0329995) discloses six or more doamins of protein Z or the C domain of protein A. In one embodiment, the amino acid residue at position 23 is a threonine in all domains of the polypeptide. The mutation improves alkali stability. The domains are liniked by linkers which include up to about 15 mino acid reisudes. In certain emobdiment, the polypeptide is coupled to the solid support by single point attachment such as via thioether bonds.

–position 23 with other substitutions:

Bjorkman (US2013/0274451) teaches multimers of the various domains of protein A with at least one of the monomers having a substitution of the Asparagine at N28 of the B domain or Z domain which results in increase in elution pH. In one embodiment, the asparagine residue at position has also been mutated such as to a threonine.

Position 29:

Bian (US 12/653888 and EP2202310) discloses chromatography ligands comprising 2 or more B or Z domains of SpA attached to a chromatography resin at more than one site on the resin. In some embodiments, the domains differ from the parent amino acid sequence such as at position 29 which is replaced by an amino acid residue other than alanine or tryptophan.

Bjorkman (US 2013/0274451) discloses a ligand form SpA that is made of multimer copies of domain C, with Arginine residue at N28 mutated and optionally, the ligand also contains a G29A mutation.

–With deletion of AAs at N-terminal:

Spector (US9,018,305; US9,234,010; US8,754,196; 8,754,196 and US8,895,706; See also EP2157099)  teaches an affinity chromatogrpahy ligand attached to a solid support where the ligand is based on two or more B domains or two or more Z domains or two or more C domains os SpA, each domain having a deletion of at least 4 consecutive amino acids from the N-terminus starting at position 1 corresponding to the sild type domains and further having a mutation to reduce Fab binding such as a mutation at position 29.

Particular Insertions

Between Positions 3/4:

Yasuoka (US 15/561, 332, published as US 2018/0105560) discloses inserting at least one amino acid between the positions corresponding to positions 3 and 4 of the B, Z or C domain. This is particularly useful when suing repeated domains because steric hindrance causes a problem in producign an affinity chromatography carreir that employs a repeated strucure of the B, C or Z domain.