Related Application

The present application claims the benefit of provisional application serial number

US 62/432,807 filed December 12, 2016, entitled, "Compositions and Methods for Increasing the Immunoglobulin Binding Capacities of Immunoglobulin-Binding Polypeptides and Oligopeptides," inventors Rajiv Datar and Carole Marie Laine.

Technical Field

Compositions and methods of use are provided for derivatives of full-length truncated forms of Protein A in combination with IgG-binding oligopeptides.

Background

The number of monoclonal ant ibody (MAb) drugs continues to grow. In 2008, MAbs generated revenues in excess of US$15 billion (Leavy O. 2010. Nat. Rev. Immunol. 10: 297), making them, the highest earning category of all biotherapeutics. The world MAb market will reach $62.3 billion in 2015, with next-generation therapeutic antibody revenues reaching $2.3 billion in 2015 according to Vision gain reports published in September and Nov ember 201 1 (Visiongain 201 1 . London, UK,

www.visiongain.corn/Report/685/Therapeutic-Monoclonal-Ant ibodies-World-Market-2011- 2021; Visiongain 2011. London, UK; www.visiongain.com/Report/712/Next-Generation- Antibody-Therapies-Pipeline-and-Market-201 1 -2021 ). Biosimilar antibodies will also begin to enter established markets as regulatory authorities clear approval pathways for them. Most antibody drugs treat cancer and autoimmune diseases, and many others are used to treat orphan and infectious diseases. Unfortunately, antibodies are complex proteins in a variety of parameters, which complicates their purification and characterization, making it difficult for their developers to meet the rigid requirements for therapeutics.

Because of both the natural and engineered variations in therapeutic antibody structures, there is no "one-size-fits- all" when it comes to techniques for MAb purification. The method that most closely approximates universal use is Protein A affinity

chromatography, which has become the workhorse for antibody production. However, Protein A is expensive (with costs an order of magnitude over conventional chromatography l resins), susceptible to degradation by proteases (cleaved domains can adhere to a MAb product, problems that complicate separations) (Carter-Franklin, JN et al. 2007. J.

Chromatogr. A 1163, 2007 105-111), and is not fully stable to column washing and elution conditions. Further, Protein A generates an immunomodulation response and has limited capacity to accommodate the increasingly high titers found in modern upstream feeds.

Although the antibody purification field is advanced, among companies involved there has been some reluctance to inv est in and introduce new technologies and. or further advance purification technologies. New alternatives have bee described as "disruptive" (Low, D et ai. 2007. J. Ch.rom.atog. B 848(1) S48-S63) predicting that Protein A will continue to be used for commercial-scale MAb purification throughout the foreseeable future ( Low, D et al. 2007. J. Chromatog. B 848(1) S48-S63; Shukla, A A et al. 2007. J.

Chromatogr. B 848(1) S28-S39). There is a need in the industry to lower production costs and pass along those savings by making medications more affordable for patients. The emergence of biosimilars (or follow-on biologies) and a growing number of companies seeking to capitalize on such products creates a need for new approaches for IgG purification (Gagnon. P. 2012. J. Chromatogr. A 1221 , 57 70).

Summary of the embodiments

An aspect of the invention herein provides an engineered polypeptide that binds immunoglobulins or immunoglobulin-containing compounds, the polypeptide comprising at least one functional moiety of at least one naturally occurring or recombinant immunoglobulin binding protein selected from the group of a protein A, a protein G, a protein A/G, a protein L, and other immunoglobulin binding proteins, and a functional variant or portion thereof, the polypeptide being chemically conjugated or genetically fused with at least one synthetic functional immunoglobulin binding oligopeptide at a terminal amino acid residue or to an internal reside such as an internal lysine of the binding protein.

For example, the polypeptide contains a plurality of the functional moieties from the group and/or iterations of one of the functional binding protein moieties. The polypeptide in another embodiment further contains at least one linking element connecting at least two functional moieties. For example, the linking element has an amino acid sequence and contains fewer than about 1800 amino acids, or fewer than about ninety- five amino acids. For example, the linking element contains from about two to about fifty- four amino acids, or from about four to about ten amino acids. In an embodiment of the polypeptide, the binding protein functional moiety has an amino acid sequence selected from the group of: SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 11, 14, 15, 16 and 17, and functional variants and portions thereof. In an embodiment of the

polypeptide, the oligopeptide is selected from at least one of: amino acid sequence of: SEQ ID NOs: 8-10 and 18-23; functional variants and portions of these amino acid sequences; iterations having multiple copies of the sequences; and functional conservative amino acid substitutions of these sequences.

In various embodiments of the polypeptide, the functional moieties being chosen for binding targeted classes of IgG immunoglobulins or immunoglobulin-containing compounds, the polypeptide further contains a separation matrix medium in a large capacity capture bed that is characterized by fast flow rate kinetics.

In an engineered polypeptide that binds immunoglobulins or immunoglobulin- containing compounds, the polypeptide containing at least one functional moiety of a naturally occurring or a recombinant immunoglobulin binding protein or oligopeptide selected from the group of a protein A, a protein G, a protein A/G, a protein L, and other Ig binding proteins, and a functional variant or portion thereof, the improvement contains: at least one copy of the amino acid sequence or a portion thereof of the functional moiety or the functional variant or portion thereof, and further comprising at least one oligopeptide chemically conjugated to or genetically fused to a terminal amino acid residue or conjugated to or fused to within an internal reside such as an internal lysine.

An aspect of the invention provides a separation matrix that includes a polypeptide as described herein, coupled to a solid support. For example, the solid support preferably comprises a medical-grade porous polyvinyl chloride (PVC) medium having a form selected from the group of beads and sheets. In other examples, the PVC medium is embedded with or constitutes porous protein-adsorptive support surfaces, the medium having a bi-modal pore size distribution with the larger pore size ranging in average from about 0.5 - 5.0

micrometers and the smaller pore size ranging in average from about 0.003 - 0.3

micrometers. In various embodiments, the support surfaces material is selected from the group consisting of cellulose, agarose, nylon, porous metalloid oxides, porous metallic oxides, and porous mixed metallic oxides, silica particles, silica gel, controlled pore glass, alumina, stannia, titania, and zirconia. In various embodiments, the polypeptide and the solid support and the support surfaces material are coupled by single-point attachment. Alternatively, the separation matrix is generally the solid support and the support surfaces material are coupled by multi-point attachment.

The separation matrix aspects of the invention provided herein have very high binding capacities, for example, the immunoglobulin binding capacity of the polypeptide in mg per ml of bed volume is at least about 25, at least about 50, at least about 75. Further, the separation matrix has a scale up capacity which is linear and reproducible over a scale-up factor increase of at least about 500-fold, 1000-fold, 2000 fold, or at least about 3000-fold. The separation matrix in various embodiments includes at least one linker for attachment to at least one of the solid support and the support surfaces material. In general, the linker is selected from an amino acid sequence, a random amino acid polymer, a polyethylene glycol, covalently attached chemically or by genetic fusion to the polypeptide.

An aspect of the invention provides a composition comprising an oligopeptide having amino acid sequence selected from the group of: QPQMSHM (SEQ ID NO: 9); KPGKEDNN (SEQ ID NO: 10); CPSTHWK (SEQ ID NO: 18); NVQYFAV (SEQ ID NO: 19); ASHTQKS (SEQ ID NO: 20); TNIESLK (SEQ ID NO: 21); NCHKCWN (SEQ ID NO: 22); and, SHLSKNV (SEQ ID NO: 23). For example, an aspect of the invention provides a

polypeptide having an amino acid sequence according to SEQ ID NO: 16 or 17. Various embodiments include an amino acid sequence which is 85% identical, 90% identical, 95% identical or 98% identical. Other embodiments include a polypeptide or a protein containing at least one copy of any of these oligopeptides. Still other embodiments include a

chromatographic medium containing any of the oligopeptide or polypeptides compositions herein affixed to a support selected from the group of: a resin, a membrane, a filter, and a bead. These compositions include any with a change of up to 15% in amino acid sequence, resulting from one or more changes including due to deletion, an addition, or a conservative substitution.

An aspect of the invention provides a method of purifying an immunoglobulin from a biological sample, the method including steps of: contacting the sample to a separation matrix having a polypeptide as described herein, under conditions of ionic strength and pH for binding the immunoglobulins selectively and specifically to the matrix and passing other sample components into a flow through; and optionally washing the column and eluting the bound immunoglobulins from the matrix with an eluant buffer containing selected from the group characterized by in comparison to a loading buffer: decreased pH, increased pH, increased ionic strength, and presence of a competitive binding ligand, thereby purifying the immunoglobulin.

In an embodiment of this method, the separation matrix consists of a polypeptide, or a plurality of polypeptides, each of which includes at least one or a plurality of the oligopeptides, the respective polypeptides and oligopeptides being non- identical and having non-identical affinities for classes of immunoglobulins, and the method further includes purifying at least one or a plurality of antibody types selected from the group of: IgGi, IgG ₂ , IgG ₃ , IgG ₄ , IgM, IgA, IgE and IgD. In alternative embodiments, the method further includes purifying selectively only IgG species, for example, one or more of all of the antibody types IgGi, IgG ₂ , IgG ₃ , and IgG ₄ .

Brief Description of the Drawings

FIG. 1 is a schematic map of native Protein A domains, including the N-terminal end of the Protein A chain 1, with the signal sequence, 2, followed by the five IgG binding domains of E, D, A, B and C. The C-terminal half of the Protein A molecule consists of the X-domain, and the LPXTG motif, 3, and a hydrophobic region, 4.

FIG. 2 is an amino acid sequence (SEQ ID NO:l) of native Staphylacoccus aureus (NCTC strain 8325-4) Protein A (Lofdahl, S et a!. 1983. Proc. Natl. Acad. Sci. USA 80, 697- 701). N-terminal underlined sequence represents S. aureus signal peptide. C-terminal underlined sequence represents the X-domain.

FIG. 3 shows amino acid Protein A sequences of functional domains: IgG binding E domain (SEQ ID NO:2), IgG binding D domain (SEQ ID NO:3), IgG binding A domain (SEQ ID NO:4), IgG binding B domain (SEQ ID NO:5), IgG binding C domain (SEQ ID NO: 6) and X domain (SEQ ID NO:7). The X domain, SEQ ID NO: 7, in the underlined segment contains 12 repeated octapeptide units having amino acid sequence KPGKEDNN or variants of the two residues on the right which are NN for five of these repeats, and are NK or GN among the other repeats of the X region of SEQ ID NO: 7.

FIG. 4A is a schematic drawing of the structure formed by natural amino acids or by a combination of natural and unnatural amino acids having specific length "n" creating a linear polypeptide 5. The length of a linker comprising amino acid chain can be from one single amino acid up to 1,800 amino acids or more.

FIG. 4B are amino acid sequences of exemplary oligopeptides that have affinity to bind IgG protein. These oligopeptides can be engineered into longer proteins as monomeric units (single copy), or can be linked together to form longer chain polypeptides, or scattered at recommended positions in a protein in any manner or combination, for example with the elements shown in FIG. 3. It is known that Protein A, for example, does not bind human IgG ₃ (see Table 1 which shows that Protein A binds differentially to immunoglobulins IgG and IgM, and differentially to subclasses of IgG from humans, and from other mammalian species). Such peptides can confer additional specificity to various embodiments of the protein-polypeptide or polypeptide-polypeptide moieties, which are the subject of this invention, to engineer for capture of a broader range of immunoglobulins in one step.

FIG. 5 is a graphical representation of IgG binding capacity (ordinate) as a function of flow rate (abscissa). The upper left hand quadrant depicts the operational range for prior art packed columns containing chromatography resins or beads, and the lower right hand quadrant shows the operational area for membrane adsorbers. The ideal area to optimize capacity and speed is shown in FIG. 5. The difference in the "Y" component between optimum technology exceeds the difference in the "X" component offered by the packed column technologies.

FIG. 6 shows schematic examples of various protein-polypeptide moiety or protein domain and oligopeptide combinations, respectively that are possible to engineer from among any of the five IgG binding domains, depicted as 6, and 7, respectively to form a variety of combinations, illustrated as 8 and 9.

FIG. 7 is a schematic representative of IgG binding domains of Protein A - full length or truncated, linked to each other and to polypeptide moieties of Figs. 4A-4B. The IgG binding domains of Protein A may be repeat units of one single domain (e.g. a sequential oligomer of domain E, for example, E-E-E-..., or such an oligomer interspersed with one or more oligopeptides or with linkers) or to any combination with each other as may be desired for a particular IgG binding effect.

FIG. 8 is a depiction of polypeptides engineered herein, 7, bound to a support matrix 14 as described here and in the claims. Support matrix 14 is a solid support or is porous in nature. The composition 7 can be attached to- or grown on and within the surfaces of the support matrix 14.

FIG. 9 is a graphical representation of the data in Table 5, showing unexpected linearity of scale up over a range of at least three orders of magnitude, of binding of IgG as a function of the bed volume, using the methods and compositions herein. At the very low end of the X-axis the bed volume is 4 mL. At the high end, the bed volume is 10,458 mL. The scale-up factor with respect to the bed volume is 2,615-fold. At the low end, the quantity of IgG bound is 0.1 g, while at the upper end it is 264 g, yielding a scale up capability which is at least a factor of 2,640-fold. Examples with even greater bed volumes, double the high end, follow the same extent of linearity of scale-up.

FIG. 10 is an amino acid sequence, SEQ ID NO: 11, of a Protein A derivative engineered herein which was designed for production by cytosolic expression. The gene encoding this amino acid sequence was engineered to remove the portion encoding the 36 amino acid leader sequence shown underlined at the amino terminus in FIG. 2, and further engineered to remove genetic material encoding additional carboxy terminus 31 amino acids, resulting in expression of the protein within the producing cells. The gene encoding the amino acid sequence was designed with codons optimized for E. coli, was synthesized, and the sequence was verified by restriction enzyme analysis. Genes were cloned in a

commercially available standard vector.

FIG. 11 is an amino acid sequence, SEQ ID NO: 12, of a Protein A derivative engineered herein that was designed for bacterial periplasmic expression. The gene encoding SEQ ID NO: 12 was further engineered from that encoding SEQ ID: 11 to have a 20 amino acid amino terminus leader signal sequence causing the resulting Protein A derivative to be secreted into the periplasmic space and retained there. The gene encoding the amino acid sequence was designed with codons optimized for E. coli, was synthesized, and the sequence was verified by restriction enzyme analysis. Genes were cloned in the standard vector as described herein.

FIG. 12 is an amino acid sequence, SEQ ID NO: 13, of a Protein A derivative engineered herein which has been designed for extracellular expression, and contains the full leader sequence of native S. aureus Protein A shown in FIG. 2, and lacks the additional carboxy terminus 31 amino acids of FIG. 10. The gene encoding the amino acid sequence was designed with codons optimized for E. coli, was synthesized, and the sequence was verified by restriction enzyme analysis. Genes were cloned in the standard vector.

FIG. 13 displays amino acid sequences of two engineered versions of Protein A. A deleted version was designed having 415 amino acid residues, SEQ ID No: 14, and was constructed to remove the native signal sequence and to remove 35 amino acids from the carboxy terminus, and having a methionine initiating reside at the amino terminus. A 450 amino acid version, SEQ ID NO: 15, contains the S. aureus Protein A signal sequence to direct secretion and processing during production. The vertical bars between residues 327 and 328, and 332 and 333, respectively, indicate respective insertion points for design and construction of two engineered Protein A derivatives that contain oligopeptides and linkers inserted into the respective locations. The engineered polypeptide BIG_Hep4_linlopt4_ExC having amino acid sequence SEQ ID NO: 16 was constructed by insertion between the amino acid residues 327 and 328, and BIG_Hep4_lin3opt4_ExC amino acid sequence SEQ ID NO: 17 was constructed by insertion between 332 and 333.

FIG. 14 displays amino acid sequences of designed and synthesized Protein A derivatives containing inserts of linker sequences and multiple copies of engineered heptapeptide iterations of SEQ ID NO: 9, the protein being engineered for enhanced binding of immunoglobulins. Construct BIG_Hep4_linlopt4_ExC amino acid sequence SEQ ID NO: 16, 543 amino acids, contains an insert displayed with a gray highlight, from reside 333 to reside 425 with the designed linker and three embedded heptapeptide iterations, each having the amino acid sequence QPQMSHM (SEQ ID NO:9) shown in dark rectangles. Construct BIG_Hep4_lin3opt4_ExC amino acid sequence SEQ ID NO: 17, 533 amino acids, contains an insert from reside 328 to reside 410 with the designed linker and the three embedded heptapeptide iterations similarly indicated.

FIG. 15 is a restriction map of the 6919 bp vector for expression and secretion of the engineered Protein A construct BIG_Hep4_linlopt4_ExC, cloned in a background of pET30B+. The Protein A construct is expressed from the T7 promoter transcribed

counterclockwise.

FIG. 16 is a restriction map of the 6889 bp vector for expression and secretion of the engineered Protein A construct BIG_Hep4_lin3opt4_ExC, cloned in a background of pET30B+. The Protein A construct is expressed from the T7 promoter transcribed

counterclockwise.

FIG. 17 is a photograph of an SDS-PAGE analysis of 20 samples of supernatants of cells carrying a gene encoding engineered Protein A construct BIG_Hep4_linlopt4_ExC, cloned in a background of plasmid pET30B+, cultured in each of three media (LB, TB and M9 minimal medium), the samples taken at times indicated after induction of expression (each of Oh, 3h and 6h). The arrow indicates the protein expressed having the predicted molecular weight. Molecular weight standards (MW kDa) were applied to the left hand lane. Detailed Description

Protein A: Protein A is a 42-kDa protein anchored in the cell wall of Staphylococcus aureus (Sjoquist, J et al. 1972. Eur. J. Biochem. 30(1) 190-194) with the ability to selectively interact with immunoglobul ins (IgGs) ( Langone, J J. 1982. Adv. Immunol. 32, 157 252 ). It binds strongly to all classes of human IgGs except for IgG ( Langone, J J supra). Full-length Protein A consists of five homologous domains (referred to as E, D, A, B. and C, in order of their arrangement from the N-terminus) and one cell-wall associated domain ( Lofdahl et al. 1 983 supra; Guss, B, et al. 1984. Eur. J. Biochem. 138(2 ) 41 3 420).

Protein A was initially produced by cuituring the Cowan strain of S. aureus and extracting the protein from the bacterial cell walls (Sjoquist, J et al. 1972. Eur. J. Biochem. 29(3) 572-578). The sequence of native S. aureus Protein A is shown in FIG. 2 (SEQ ID NO: 1). A strain of S. aureus was later discovered that secretes Protein A into its culture supernatant ( Lindmark, R et al. 1 977. Eur. .1. Biochem. 74(3 ) 623 628 ). As recombinant DNA technology advanced. Protein A was expressed as a fragment without its ceil- wall domain using Escherichia coli as an expression host (Duggleby, CJ et al. 1983. Nuci. Acids Res. 1 1(10) 3065-3076;

Hammond, PM et al. 1 990. Ann. NY Acad. Sci. 61 3, 863 867: Cai. S et al. 1992. Chin. J. Biotechnol. 8(2) 93-98; Engel, H et al. 1992. Prot. Expr. Purif. 3(2) 1992: 108-1 13). Various engineered derivatives of Protein A are shown in FIGS. 10-12, engineered to be expressed and to be located intracel lularly, in periplasmic space, or extraceiiuiarly, respectively, in bacterial expression vectors and host cells.

IgG binds to Protein A at the IgG Fc region ( Lindmark, R et al. 1983. .1. Immunol. Meth. 62(1) 1983 : 1-13; Gouda, H et al. 1998. Biochem. 37(1 ) 129-136). The interaction is very specific and hydrophobic in nature. It involves some hydrogen bonds and two salt bridges. The high specificity enables Protein A affinity chromatography to remove greater than 98% of impurities from, complex solutions such as cell harv est media in a single

puri fication step ( Follman, D et al. 2004. .1. Chromatogr. A 1024: 79 85 ). One drawback of the well-known specificity of interaction of Protein A with IgGs is necessity of use of harsh conditions such as low pi 1 for elution. That can be problematic for some antibodies that are either unstable or tend to aggregate at low pi 1 lev els. In general, only a small amount of impurities— e.g., aggregates, residual host-cell proteins, DNA, and leached Protein A— will remain after this single starting unit of downstream process operation. These remaining impurities usually are removed in one or two additional chromatography steps. Affinity Supports: Protein A has been immobil ized to a large number of types of supports suited for l iquid chromatography of proteins ( Boschetti, E et al. 2000. Academic Press: San Diego, CA, 535; see Table 2). Initial ly, a popular product was Protein A immobilized on CNBr-activated Sepharose CL 4B from Amersham (now GE Healthcare) in Sweden. The medium was characterized as hav ing high selectivity and low nonspecific adsorption, but due to the nature of the agarose-based support, a packed bed was too soft to allow for high flow rates. For this reason that medium has been largely replaced by more highly cross-linked Sepharose for large-scale applications. Modern Protein A sorbents are based on controlled porous glass, coated porous polymer gel-filled mineral materials, and other supports (Hahn, sR et al 2003. J. Chromatogr. B 790 2003 : 35-51) using materials sufficiently rigid to allow for column operation at high flow rates. Exemplary support materials are shown in EP2902094 published May 8, 2015, Laine, C.

State of the Art: Since the first reports over 40 years ago involving use of immobilized Protein A for affinity purification of antibodies (Hjelm, H et al. 1972. FEBS Lett. 28(1) 1972 73-76; Kronvall G. 1973. J. Immunol. 1 1 1(5) 1401-1406), it has become the industrial standard for purification of clinical-grade MAbs (Gagnon P. 1996. Validated Biosystems Inc. : Tucson, AZ). Janssen Biotech's Muronomab (brand-name Orthoclone OK.T3 ) is a CD3-specific MAb that was approved by the US Food and Drug Administration (FDA) in 1986 (Becker, H. 2007. Handbook of Therapeutic Antibodies Volume III:

Approv ed Therapeutics. Dubel S, Ed. Wiley- VCH : Weinheim, Germany. 905-940) for use in treatment of acute transplant rejection, an early approved product made using Protein A as a capture step in its manufacturing process. Protein A capture serves as the key volume- reduction step in antibody downstream processing. The purification scheme for Protein A chromatography involves binding at neutral pH and elution at acidic pH . Ease of method development has caused Protein A affinity chromatography to be almost universally adopted in large-scale manufacturing processes, and almost universal applicability and the overall scheme of associated operating conditions lend themselves readily to a platform format (Shukla, AA et al. 2007. J. Chromatogr. B 848(1) S28-S39).

Economics: Cost is another very important factor as it relates to medical care costs. Some Protein A resins cost as much as $ 1 5,000 L. Table 2 shows prices from 2013, and it is expected that these are underestimates for current values. Even with larger numbers of players and increasing competition, prices continue to be quite high. Repligen resins cost about $6,000 L and tolerate only 0. IN NaOH. It has been argued that the main limitation of Protein A is capacity or productivity, rather than cost (Gagnon, P. 2012. J. Chromatogr. A 1 1 , 57 70), and that capacity issues result from, use of porous particles in a fixed bed. The

Protein A molecule per se occupies a large amount of intrapore space in porous media because of its size.

Table 2. Some commercially available Protein A sorbents for affinity chromatography

* 2013 list prices in US dollars (from websites or direct sales inquiries) listed as fair comparison without discounts (e.g., for large-volume orders).

Other Chromatographic Methods: A significant amount of effort has gone into further developing one type of fluidized chromatography technique known as cxpanded-bed absorption (EBA). It involves the use of adsorbent particles dispersed in a liquid media. One main benefit of EBA is a reduction in the number of steps required for antibody recovery due to direct capture of product from a ceil suspension. For antibodies, however, this application could not replace Protein A, which in a form bonded to agarose is used to capture the antibody in EBA.

Membrane adsorbers ("membrane chromatography") offer clear advantages over conventional, resins, both, in terms of disposability (which eliminates the need for cleaning and validation) and the ability to operate at high flow rates. Because of their lower surface area, however, most membrane adsorbers suffer from low binding capacity compared with an equivalent volume of porous particles. An exception to that limitation is membrane technology for example, developed by Natrix Bioseparations ( ucze vski M, et al. 201 1. Biotechnol. J. 6(1) 56-65), with membranes that consist of a polymeric hydrogel formed within a fle ible porous support matrix. That macroporous hydrogel polymer structure prov ides high binding-site density and a large surface area for binding and ra id mass transfer.

Kuczweski et al. supra developed a membrane-based, high- capacity, cation-exchange capture step for MAbs using a C membrane from Natrix. They reported a capacity of 55 mg of antibody per milliliter of membrane, which is about five times that of other membrane adsorbers (Gottschalk, U. 2005. Biopharm Int. 18 (6): 42 - 58). Table 3 shows binding capacities of some commercially available Protein A sorbents, in a range from. 30 mg mL to 45 mg mL. Because of low capacities and dilutions created by the void volume and variations in membrane thickness, most of these adsorbers are preferentially used in flow-through mode.

The words, "protein", "polypeptide" and "oligopeptide" as used herein all refer to polymers of amino acids. The amino acids are generally L-amino acids, and may be naturally occurring or non-naturally occurring. Generally a protein refers to a biologically expressed product found in correct conformation with respect to secondary, tertiary and quaternary structures. In contrast, the term "oligopeptide" refers to a short polymer, at least about four amino acids in length, generally less than about 50 amino acids in length, which has been synthesized by standard procedures using solid phase initiating residues and F-moc or T-boc protected reagents. The oligopeptides herein are useful individually as ligands for various target proteins, and are envisioned herein as moieties of a larger polypeptide in combination with one or more domains that are subsequences of naturally occurring proteins having affinity for immunoglobulins.

The term, "polypeptide" as used herein and in the claims designates a moiety which is generally designed, although it may contain naturally occurring amino acid sequences, and may be produced by peptide synthesis or produced recombinantly in transformed or transfected cells. A polypeptide is generally longer than an oligopeptide, for example, may be a polymer of several hundred amino acids. The useful compositions herein are designed and are recombinantly expressed from a synthetic gene encoding the polypeptide, or are chemically engineered by standard chemical coupling procedures. The polypeptides contain one or more subcomponent amino acid sequences so that both naturally occurring and synthetic designed novel amino acid sequences are, in various embodiments of the

polypeptide, present in various functional portions of the polypeptides.

The term polypeptide is used to distinguish this designed binding material from its targets, which are referred to herein as proteins as these are members of the immunoglobulin family. The target proteins can be made in vivo in a vertebrate animal and obtained from blood or serum, or can be produced in cultures from genetically engineered cells, e.g., expressed from recombinantly prepared vectors and cell lines in cell culture. In general the target proteins are oligomeric and are in native conformation. However target proteins include single chain immunoglobulin derivatives that are engineered from antibody proteins, including immunoglobulins that are the result of protein engineering and were isolated from mutagenized libraries of display vectors.

The designed and synthesized polypeptides of the invention herein can be genetically encoded and biologically expressed by a cell transformed or transfected with a vector containing an encoding gene. The choices in design of the polypeptide of types of components, and number of iterations of each of the components present in the polypeptide, is determined by the user, and the binding functions of each of the candidate polypeptides are determined with assays of affinity for each of a plurality of desired target immunoglobulins, and selection for those candidates with an optimum extent of affinity and a desired pattern of affinities for each among the classes of immunoglobulins.

Protein A is a 42 kDa surface protein originally found in the cell wall of the bacteria Staphylococcus aureus. It is encoded by the SpA gene. It has found extensive use in biochemical research and in the production of biological drugs because of its ability to bind immunoglobulins (IgG). It is composed of five homologous Ig-binding domains that fold into a three-helix bundle (see FIG. 1). Amino acid sequences of the intact, native

Staphylococcus aureus Protein A is shown in FIG. 2, while the amino acid sequences of the five individual Ig-binding domains and the anchoring region is shown in FIG. 3. Each domain is able to bind proteins from many mammalian species, most notably IgGs (see Table 1). It binds the heavy chain within the Fc region of most immunoglobulins and also within the Fab region in the case of the human VH3 family (Ljunberg et al., 1993).

Table 1 (A) Immunoglobulin affinity for recombinant Protein A

IgM Some

(B) Comparison of species specific immunoglobulin binding to Protein A and protein G

Recombinant Staphylococcal Protein A (SpA) is often produced in E. coli for use in immunology and other biological research. Protein A has variously been coupled to other molecules including a fluorescent dye, enzymes, biotin, colloidal gold or radioactive iodine without affecting the antibody binding site. Protein A is also widely utilized coupled to magnetic, latex, agarose beads and a host of other media.

Protein A is often immobilized onto a solid support and used as a reliable method for purifying total IgG from crude protein mixtures such as from serum, ascites fluid, fermentation- or bioreactor broths, or coupled with markers to detect the presence of antibodies. Immunoprecipitation studies with Protein A conjugated to beads are also commonly used to purify proteins or protein complexes indirectly through antibodies against the protein or protein complex of interest.

Because of the ability of Protein A to bind a large variety of IgGs, one of its most important uses is in the affinity chromatographic purification of antibodies and antibody fragments. Immunoglobulins represent the most prevalent biopharmaceutical products in either manufacture or development worldwide. The high commercial demand for and hence value of this particular therapeutic market has led to the emphasis being placed on

pharmaceutical companies to maximize the productivity of their respective monoclonal antibody (MAb) manufacturing processes whilst controlling the associated costs. Affinity chromatography is used in most cases, as one of the key steps in the purification of these immunoglobulin molecules, such as monoclonal or polyclonal antibodies. At present, SpA-based affinitymedium probably is the most widely used affinity medium for isolation of monoclonal antibodies and their fragments from a variety of different industrial feed stocks from cell cultures. Accordingly, various matrices and resins comprising Protein A-ligands are commercially available, for example, in the form of native Protein A (e.g.

Protein A SEPHAROSE™, GE Healthcare, Uppsala, Sweden), or containing recombinant Protein A (e.g. rProtein A SEPHAROSE™, GE Healthcare).

Major drawbacks of Protein A include the fact that it is expensive and unstable under typical column cleaning/sanitization conditions such as 1M NaOH (Costioli, M et al, 2010 Biopharm International 23 (6); Gagnon, P. 2012 J. Chromatogr A 1221 : 57 - 70).

Derivatives of Protein A have been shown to retain similar binding properties while showing an increase either in binding capacity or stability (Gulich, S et al, 2002 Protein Eng 15 (10): 835 - 842; Linhult, M et al, 2004 Proteins 55 (2): 407 - 416). A few alkaline- stabilized Protein A derivatives are currently marketed as chromatography resins such as the GE Healthcare (Pittsburgh, PA, USA) MAbSelect Sure™ resin which uses a modified tetrameric B binding domain, Protein A ceramic Hyper D F resin from Pall Corporation (Port Washington, NY, USA) and Tosoh Biosciences (South San Francisco, CA, USA) Toyopearl AF-Protein A-650F resin, which uses a tetrameric derivative of the C binding domain of Protein A.

However, Protein A derivatives still suffer from high costs associated with licensing fees and costs of producing and purifying the recombinant protein. This purification comes with an extremely high price tag (see Table 2). The price of Protein A resins can be as high as $17,000 or more per liter of the resin. This high cost adds directly to the cost-of-goods factor in the production of biological drugs.

To decrease the operating and replacement costs of Protein A affinity chromatography media, there is a need in the bioprocessing industry for Protein A media that can be used with increased number of cycles than are now available.

Modifications to the Protein A molecule, which have involved the sequencing and working with specific domains from within the Protein A molecule have been carried out to increase its binding capacity, but these have resulted only in incremental improvements (see Table 3). Table 3. Immunoglobulin binding capacities of some commercially available Protein A sorbents.

DBC: Dynamic Binding Capacity at a particular residence time. From websites and sales literature.

Mutated immunoglobulin-binding proteins having an Asn residue mutated to an amino acid other than Gin or Asp (using the three letter amino acid code) are shown in patent application WO2003080655A1 published Oct 2, 2003, inventor Hober, S. et al, and were found to have increased chemical stability at high pH. Hall et al. in WO2008039141 published April 3, 2008 showed that domain C of Protein A withstands harsh cleaning agents. Nakamura, S. et al. produced a protein having an amino acid sequence ATK or ASK (using the one letter amino acid code) to be used for isolating immunoglobulins in

WO2012086660A1 published June 28, 2012. Bjoerkman et al. WO2012087231 published December 20, 2010, produced an affinity chromatography matrix with one or more ligands of Protein A domains having asn or his resides at the HI 8 of the B domain, and observed increased elution pH compared to non-substituted Protein A. Kihira, Y. et al. EP0863210 published September 9, 1998 produced immunoglobulin-binding artificial protein having linked units. Caustic-stable chromatography ligands having two or more B or Z domains were found by Bian et al. to be alkali stable in EP2202310 published June 30, 2010. Similarly, Yoshida et al. engineered acid stable Protein A by replacing G resides with amino acids other than A, in EP2412809 published Feb 1, 2012. Novel Protein A-based ligands having a deletion of at least three amino acids with mutations at position 29 that replace G or A with K were observed by Spector et al. to reduce Fab binding, in EP2532672 published Dec 12, 2012. The ability of short peptides of specific length "n" that might mimic the affinity binding characteristics of Protein A, have been explored (Yang et al., J. Pept. Res. 6: 120- 137, 2006) in efforts to reduce the high cost of Protein A-based, large-scale affinity purifications (see FIG. 4).

The compositions provided herein are polypeptides that integrate the endogenous affinity characteristics of either the full-length or the truncated versions of Protein A, with the affinity characteristics of shorter oligopeptides, to dramatically enhance the IgG binding characteristics of the combined entity or entities. Such modifications are engineered either by design, synthesis, and ligation of genetically- fused nucleic acid moieties encoding full-length or truncated versions of recombinant Protein A and the oligopeptides, including possible peptide linkers for display of the high binding oligopeptides on outer surfaces of the designed polypeptides, or are engineered by chemical conjugation of these entities.

There are serious cost drawbacks to using existing forms of Protein A, bound to either resins or membrane supports to purify high-value, life-saving biological drugs. Downstream purification of these MAbs can account for almost 80% of the total cost of manufacturing (Gottschalk, U. 2005. Biopharm Int. 18 (6): 42 - 58), while purification costs with Protein A can amount to in excess of 40% of the total purification cost, suggesting opportunities for novel antibody purification technology.

There continues to be a need in this field to obtain protein ligands bound to a separation matrix that is able to optimize an increased binding capacity with an increased flow rate.

This invention, by its ability to substantially increase the "binding capacity- flow rate" combination paradigm of Protein A-based moieties (see FIG. 5), can dramatically reduce the cost of affinity purification, thereby reducing the cost-of-goods factor in the overall cost of high-value biological drugs.

Embodiments of the invention herein provide the use of methods for increasing IgG binding capacity of IgG binding proteins and peptides such as, but not limited to Protein A- based moieties. Protein A affinity chromato graphy is a very effective capture step for monoclonal antibody (MAb) purification due to its high selectivity, enabling high purity and relatively high concentration in a single affinity chromatographic step. However, Protein A has characteristics that limit its utility during MAb purification. Native and recombinant Protein A are not stable under high alkaline conditions, as a result of which the ideal cleaning/sanitization solution, 1M NaOH, cannot be used for Protein A cleaning (Jones, SCB et al. 2004. 3rd International Symposium on Downstream Processing of Genetically Engineered Antibodies and Related Molecules. Nice, France). Binding capacity of Protein A columns or Protein A membrane adsorbers is also limited by the kinetics of MAb-Protein A binding as well as the density of Protein A ligand obtainable (Saha, K et al. 2003 J. Anal. Chem. 75(4): 835 - 842; Sheth, B. Thesis. University College of London, 2009).

The production bottleneck with Protein A chromatography as a result of improved expression and production techniques, increasing yields of MAb titers have led to the need for ever larger Protein A columns and eventually expensive new hardware such as pumps and columns. Column diameters are limited by the footprint within existing plants (Palma, AD .2005. Cost-Drive Chromatograpy.

www.pharmacueticalmanufacturing.com/articles/2005). Increasing the column height is impractical since pressure in the resin or separation medium increases with bed height and higher pressure either compresses the resin or damages the pumps (Thillaivinayagalingam, P et al. 2007. Genet. Eng. Biotechn. N. 27(11).). Increasing bio-reactor titers by increasing the number of chromatography cycles, increases both total processing time and resources and consequently the total cost. As seen in Table 5, binding capacity of PVD-Si-Protein A remains constant throughout a scale-up that ranges more than three orders of magnitude making large-scale antibody production more efficient and economical.

Protein A instability is a major cause of the Protein A chromatography bottleneck and high resin cost. This instability, particularly occurring during resin cleaning and

sanitization, decreases the binding capacity over a number of cycles further lowering yields of the product throughput.

Small peptides have an advantage over larger ligands such as Protein A due to the much higher ligand density attainable on the pore surface (Yang, H et al. 2006. J. Pept. Res. 66: 120 - 137). On similar polymer porous beads as the Protein A resin MAbSelect®, MAb binding capacities on certain cation-exchange resins have reached over 100 mg/mL (Liu et al, 2011); though MAbSelect binding capacity is in the 30 - 45 mg/mL range, with typical 2 - 3 min column residence times (Ljunglof, A et al. 2011. Bioprocess Int. 9(7): 66 - 67).

Therefore, it is possible to bind MAb to a higher capacity than can be realized with Protein A resins prior to use of the engineered polypeptides herein. Table 4 shows data obtained in examples herein using recombinant Protein A or protein G ligands attached to porous PVC media embedded with silica particles. Different human serum antibody subclasses were observed to bind to each of these media to different extents. For example,

PVC-Si-Protein A was observed to have preferentially bound IgGi, IgG ₂ , IgG ₄ , IgM, IgA and

IgE. In contrast, PVC-Si-Protein G was observed to bind four IgG subclasses but not IgM,

IgA, IgE or IgD. Clearly the data in examples herein shown in FIG. 9 and Table 5 illustrate the non- identical and even unique affinities of each of Protein A and protein G to

differentially bind eight types or subclasses of human immunoglobulins.

Table 4. Specificities of porous PVC media embedded with silica particles with recombinant Protein A and protein G ligands for human serum proteins

Table 5. Example of IgG binding capacity of porous PVC media embedded with silica particles with native recombinant Protein A ligand

In an alternative embodiment provided herein for purification media compositions and methods of making such media, small peptides are "grown" on- and/or from the pore surface of matrix material such as PVC-Si via surface-initiated polymerization. A potential for much higher peptide densities compared to Protein A is realized from surface polymerization methods, similar to other chemical ligands (Bhut et al., 2008. J. Membr. Sci. 325(1): 176 - 183). Additionally, small peptides have greater stability than larger polypeptides and proteins under sanitization conditions, particularly alkaline conditions, because small peptides lack easily disrupted secondary and tertiary structures, and these structures are more readily reassembled under original conditions. Nevertheless, a concern arising from use of smaller ligands is that they may not have the same range of specificity as the Protein A molecule itself. However, incorporating the smaller oligopeptides provide herein into full or truncated forms of Protein A, or other immunoglobulin-binding moieties, provides compositions having the potential for increasing capacity by working around the issues of steric hindrance.

Present processes, systems and hardware supporting Protein A based affinity purification, and the prohibitive cost of the Protein A ligand itself, are not amenable for single-use technologies. It is a particular focus and subject of this invention to optimize the above-mentioned limiting parameters of Protein A based affinity purification and to bring the potential of single-use Protein A based purification technology and products into reality.

Nucleic acid sequences encoding various combinations of domain polypeptides of

SEQ ID NOs: 2-6 (FIG. 3) and the oligopeptides of SEQ ID NOs: 8-10 (FIG. 4B) were expressed to obtain the oligopeptides having the amino acid sequences in examples herein using display phage, by methods shown in U.S. patent 8,685,893, Sidhu et al, issued April 1, 2014, and references cited therein. Resulting libraries of displayed sequences were enriched to select for improved optimized affinity to a target which is a class of

immunoglobulin of choice, for example IgGi. Positive selection was combined with a negative selection, to reduce affinity for unwanted immunoglobulins such as IgE.

A series of oligopeptides were obtained herein using these phage display methods, and candidate oligopeptides bound to phage were screened for ability to bind to the immunoglobin subtypes. Oligopeptides obtained by this method were all found to bind to all of the tested immunoglobin subtypes. These amino acid sequences include, using the one letter amino acid code: CPSTHWK (SEQ ID NO: 18); NVQYFAV (SEQ ID NO: 19); ASHTQKS (SEQ ID NO: 20); TNIESLK (SEQ ID NO: 21); NCHKCWN (SEQ ID NO: 22); and, SHLSKNF (SEQ ID NO: 23).

Among the oligopeptides tested for affinity to immunoglobulin protein subclasses, the oligopeptide of SEQ ID NO: 9 was observed to have the greatest affinity for binding to subtypes IgGi and IgG ₃ . Accordingly, this oligopeptide was selected for insertion, with linkers, into the designed and engineered Protein A derivative compositions herein. Design and synthesis of Protein A polypeptides

Two polypeptide amino acid sequence combinations of portions of Protein A with linkers and oligopeptide inserts were designed for testing expression, secretion and production in the context of pET30b+ (commercially available from EMD/Sigma/Millipore) plasmid backbone in transformed cells of Escherichia coli. Theoretical design considerations are shown FIG. 6, in which various protein-polypeptide moiety or protein domain and oligopeptide combinations, respectively illustrate possible engineered derivatives chosen from among any of the five IgG binding domains of Protein A. These may be further linked linearly, as illustrated in FIG. 7. In these drawings, high binding oligopeptide domains and linkers are shown as wavy lines, and the Protein A core is illustrated as a circle. These engineered Protein A compositions may be bound to a support matrix as shown in FIG. 8.

The genes encoding the polypeptides having these amino acid sequences, SEQ ID NOs: 16 and 17, are shown in FIG. 14, were synthesized and ligated into the recipient plasmid, resulting in plasmids: BIG_Hep4_linlopt4_ExC; and, BIG_Hep4_lin3opt4_ExC, see FIGS. 15 and 16, respectively. Plasmids were purified and restriction digestion with Aval verified the expected size of restriction digestion fragments. Two single colony clones of each construct were chosen for analysis of expression and production optimization in cells of E. coli strain BL21 DE3 (commercially available from New England Biolabs).

Expression Optimization: time, temperature, media, additives.

Initial expression studies used shake flasks with TB complete medium, and growth of cells and production of polypeptide as analyzed as a function of time of culture. Control cells were the same strain carrying the empty vector without the polypeptide insert. Protein production of the engineered polypeptide was assessed by SDS-PAGE following induction of insert encoded protein by addition of IPTG to inactivate lac repressor and obtain the desired protein as a result of induction of synthesis of T7 polymerase, by methods well known to those of ordinary skill in recombinant plasmids and bacterial gene expression systems.

Production of the polypeptide was monitored in samples removed from each of the cultures at 3h intervals. The polypeptides which are the engineered Protein A derivatives were observed to have appeared in culture supernatants beginning at 3h after inoculation, and continuing to increase through 18 h of culture. No protein band having the same molecular weight was observed in the supernatants during growth for the same time periods of the control strain. Representative protein expression analytical data are shown in FIG. 17, which is a comparison of growth and production parameters for these strains in three standard E. coli media: LB (Luria-Bertani broth), TB (commercially available from Sigma), and M9 minimal media. Greatest polypeptide production was observed in TB, and the yield was estimated after 6 h of culture by SDS-PAGE to be at a level of at least about 250 mg/L. The minimal medium by comparison yielded about 180 mg/L. The SDS-PAGE data indicated that these supematants contained predominantly the engineered polypeptide of interest, and small amounts of additional bands were seen at lower molecular weights.

Expression was further compared at two temperatures of culture, and the result of that example was that production was observed to have been greater at 37 C than at 30 C. Effects of potential medium additives were analyzed for the agents Glycine, Triton X-100 and Tween-20 which were added to TB medium and were present during growth. Analysis of culture supematants by SDS-PAGE showed that each of these additives increased appearance of non-specific proteins in the supematants, in comparison to control cultures in TB medium with no additives. The non-specific proteins were observed to be of higher and also lower molecular weights than the molecular weight of the engineered polypeptide.

Batch fermentation and High-cell density Fed-Batch fermentation.

Cells in TB medium were induced for production at high densities (OD ₆ oo between 6 and 8) or very high density (OD ₆ oo 40-50) and fermentation was monitored for p0 ₂ , pH, temperature and growth profile. Samples were taken for determination of protein expression by SDS-PAGE analysis at time intervals after induction. Expression at high levels was observed in the supernatant in an almost pure form at a level equal to or greater than about 275 mg/L from batch fermentations, and at a level of about 1.2 g/L in the fed-batch fermentation.

Binding capacity of designed polypeptides.

Surface plasmon resonance is used to analyze binding to each of eight sub-classes of human serum immunoglobulins by each of the Protein A polypeptides herein. Applicants envision that the polypeptides of SEQ NOs: 16 and 17 specifically bind IgG proteins to a greater extent per weight or per molecular weight of polypeptide, than parent Protein A as a control. Further, it is envisioned that arrays of affixed polypeptides display more resistance to washing and rinsing agents than control Protein A.