WO2009135656A1

WO2009135656A1 - A method for the purification of antibodies using displacement chromatography

Info

Publication number: WO2009135656A1
Application number: PCT/EP2009/003230
Authority: WO
Inventors: Jim Davies; Lee Allen
Original assignee: Lonza Biologics Plc.
Priority date: 2008-05-06
Filing date: 2009-05-06
Publication date: 2009-11-12

Abstract

The present invention relates to purification of monoclonal antibodies for parenteral therapeutic use without using a Protein A affinity capture step. The process includes the use of displacement chromatography for the purification of monoclonal antibodies, preferably its use for the initial capture step, from complex feed streams such as mammalian cell culture supernatant. Displacement chromatography is combined with other unit operations to allow a process suitable for large scale manufacture to be used and to further increase the purity of the resulting product at a yield higher than could otherwise be achieved.

Description

A METHOD FOR THE PURIFICATION OF ANTIBODIES USING DISPLACEMENT CHROMATOGRAPHY

The present invention relates to purifying recombinantly expressed monoclonal antibodies using a range of unit operations to produce product of suitable purity for diagnostic as well as for therapeutic use via parenteral administration. More particularly, this invention relates to the purification of monoclonal antibodies (mAbs) using a set of defined unit operations that may be used in a combination suitable for a particular cell line and product. The sequence of steps and process parameters are designed to optimise process yield and plant operation. Example unit operations may include ion exchange chromatography (displacement mode) as the first bind and elute step in combination with other unit operations (before and/or after the ion exchange chromatography step) to maximise the recovery and purity of mAbs obtained from a complex feed stream (e.g. mammalian cell culture supernatant (CCS)).

In this specification a number of documents are cited, the entire disclosures of these references (including i. a. scientific articles, patents and patent applications) are hereby incorporated herein by reference for the purpose of describing at least in part the knowledge of those with ordinary skill in the art and for the purpose of disclosing the techniques, methods and materials used in ion exchange / displacement chromatography as well as in other unit operations (purification steps and methods).

Background of the Invention

There is an increasing demand for purified monoclonal antibodies. These highly specific antibodies have found uses in many fields, including agriculture, biochemistry, medicine, immunology, microbiology, pharmaceutics and virology. In addition, larger amounts (g to kg/year) are required, as diagnostic and therapeutic applications expand.

Downstream processing (DSP) of monoclonal antibodies (mAbs) from mammalian cell culture is required to purify the product to a state where impurities (for example, but not limited to, host cell proteins (HCP), host cell DNA, viruses, cellular debris, lipids, cell culture media components and product related impurities such as fragments, aggregates and free light chain) are reduced to levels defined in the product specification. This requirement is especially related to the use of mAbs in therapeutic as well as in diagnostic applications.

The upstream process includes fermentation and removal of cells by a combination of centrifugation and filtration. The resulting cell culture supernatant (CCS) is the starting point for the downstream process (purification process).

The current industry standard for the first step of mAb purification is to use Protein A affinity chromatography (Shukla et ai, (2007) J. Chromatogr. B., 848, 28-39 and Fahrner et a/., (1999) Biotechnol. Appl. Biochem, 30, 121-128). In this step the CCS is passed through a chromatography column packed with a resin that has recombinantly expressed S. aureus Protein A (PrA) or related molecules covalently attached. The monoclonal antibody specifically binds to the PrA molecule and other host cell impurities are washed away. The product is then eluted from the PrA by washing with a low pH buffer and collected.

Following PrA chromatography the product is then processed using a combination of the following:

Further chromatography steps (e.g. ion exchange, hydrophobic interaction, hydroxyapatite) to remove trace levels of impurities not removed by the PrA step. Ultrafiltration/diafiltration steps to concentrate product and exchange buffers.

Virus inactivation step; there are several methods some of which are covered by patents. Examples include a low pH hold (pH 3.60-3.80 for 60-90 minutes), addition of detergents and organic solvents, exposure to UV radiation.

Virus filtration step: use of a small pore size (typically 15-40 nm diameter) to remove any virus particles that may be present. In current DSP PrA chromatography is a key step for purification due to its excellent selectivity for mAb. However, PrA has a number of disadvantages: High cost. Modern Protein A type resins (e.g. MabSelect SuRe (GE Healthcare) which has a recombinant protein with Protein A domains as its ligand) can cost -$15,000 per litre. At large scale a column can use millions of dollars worth of resin. As the resin has a limited life span (-100 cycles), partly due to the low pH elution method, this is a significant cost when the resin has to be replaced.

Low dynamic binding capacity. At normal operational flow rates, the dynamic binding capacity (DBC, defined as the grams of product bound per litre of chromatography resin) for most Protein A resins is in the range 20-30 g/l. The DBC defines the size of column required and the number of cycles required to process the product produced in a fermentation. With the high cost of PrA resins the DBC has a direct effect on process economics. As titres increase due to upstream technical development the cost of PrA resin increases as DBC limits the productivity of each cycle.

Low pH Elution. The interaction between mAb and PrA ligand is very strong and using a low pH elution buffer (around pH 3.0-3.5) is the best way of eluting the bound product. However, this can damage the product resulting in increased aggregation - a significant problem as the aggregated product is immunogenic and has to be removed in later downstream steps.

PrA Ligand Leaching. The PrA ligand is immunogenic and can be co-eluted with the product from the column due to the low pH elution buffer. This leached ligand also has to be removed in later downstream steps.

Cleaning: Protein A resins are sensitive to cleaning agents therefore most resins have to be cleaned with chaotropic agents such as urea or guanidine hydrochloride which cause handling and disposal issues. Even specialized engineered resins such as MabSelect SuRe can only withstand cleaning with 0.1 M sodium hydroxide. Although Protein A affinity chromatography has a high selectivity for mAbs and purifies mAbs to a high purity there are a number of problems as summarised above. From a high titre manufacturing perspective, the cost of PrA resins is a significant problem that affects the performance of the business especially as product titres increase. This is of specific importance with regard to large scale manufacturing (kg/year) of mAbs.

Besides the industry standard to use Protein A affinity chromatography for the purification of mAbs the prior art relates to a number of purification methods that do not include a Protein A affinity step. A number of publications describe the use of ion exchange steps for monoclonal antibody production (e.g. WO 2007/117490, WO 2007/108955, WO 2006/110277) but these use elution mode rather than displacement mode for the separation.

The prior art also relates to the use of ion exchange chromatography in displacement mode, some of which is relevant to the purification of mAbs.

Luellau et al. (1998/J. Chromatogr. A., 796, 165 - 175) disclose the use of hydroxyapatite (ceramic beads) in a HPLC displacement system/mode to separate biologically active IgA dimers from inactive monomers. EGTA was used as a displacer molecule, examined at pH 7 to 8. Little separation of monomers and dimers was observed concluding that in this application the displacement mode could not be substituted for elution mode.

J. Chromatogr. 499, 47 - 54 discloses a new displacement procedure involving a displacer protein complex. The purification of mAbs by complex-displacement chromatography on CM-cellulose, more specifically, the purification of lgG2 mAb from mouse ascites fluid is described. A cation exchange column (CM cellulose) was used together with a displacer designed for an anion exchange column. Thus, the aim was to bind the displacer to the protein on the stationary phase rather than to the resin itself. The displacer/protein complexes were eluted from the resin. Displacement Chromatography has also been used for processing other biological substances such as proteins that are not antibodies, peptides and oligonucleotides, for example:

J. Chromatogr. A, 1068, 269 - 278: Displacement chromatography for separation of soybean phosphatidylcholine and phosphatidylethanolamine;

J. Chromatogr. A, 959, 85 - 93: Displacement chromatography of isomers and therapeutic compounds using benzethonium chloride as the displacer; J. Chromatogr. A, 923, 65 - 73: Purification of an oligonucleotide at high column loading by high affinity, low molecular mass displacers. J. Chromatogr. A., 1165, 109-115: Displacement chromatography of proteins on hydrophobic charge induction adsorbent column.

Furthermore, there are various documents in the prior art relating to improvements to displacement chromatography such as WO 95/26795 describing the use of two chromatography columns in series which may have the same resin packed in both or may have different resins packed in each column. The first packed bed chromatography column is to remove unwanted components from the feed stream that would bind to the second column more tightly than the product of interest. Product and impurities bound to the second column are then removed by addition of a displacer molecule. This process generally relates to the separation of a known material from a mixture of materials which may contain unknown materials.

Further documents describing specific features of how to operate displacement chromatography include US 6,379,554, US 5,851 ,400, US 6,576,134 and J. Chromatogr. A, 444, 349 - 362.

There is prior art disclosing displacer molecules and their synthesis, preferably low molecular weight displacers, for the separation and purification of biomolecules such as oligonucleotides and proteins: WO 99/47574, WO 03074148, NO 963420, WO 2007/055896 and WO 2007/064809. In addition, WO 2007/055896 and WO 2007/064809 also disclose general displacement chromatography processes using cation or anion exchange chromatography matrices including where the mixture loaded on to the column may include one or more recombinant antibodies.

The prior art work on displacement chromatography of proteins and other bioproducts has been performed generally on small scale HPLC columns using low flow rates and model systems which had fewer impurities than a complex feed stream such as CCS. Moreover, the displacement chromatography prior art fails to describe a complete process for producing a purified monoclonal antibody suitable for therapeutic use. Ion exchange chromatography in displacement mode is unable to produce products with the required high purity therefore the prior art in isolation is unable to provide a solution to the problem addressed by the invention described herein. Moreover, the displacement chromatography prior art fails to take factors that may affect manufacturing scale purification activities (e.g. precipitation of CCS components during conditioning, reduction of impurities) into account.

Thus, it was an object of the present invention to provide a method and means for the purification of mAbs from complex feed streams (such as for example CCS) to the high purity required for parenteral use which is especially applicable for large scale production of mAbs and which essentially avoids the disadvantages of the prior art solutions including Protein A affinity chromatography methods. A further object of the present invention is to specify a sequence of steps including the use of a high capacity displacement ion exchange chromatography step (bind/elute capture step) to (i) allow conditioning of CCS to the pH and conductivity required for displacement chromatography without causing significant precipitation, (ii) increase the yield possible from displacement ion exchange chromatography by removing impurities before or after the first bind/elute capture step and (iii) avoid the difficulties with Protein A affinity chromatography that have been summarized above Description of the Invention

This invention provides a process for purification of monoclonal antibodies for diagnostic and therapeutic parenteral use. One aspect of the invention is the use of displacement chromatography for the purification of rπAbs specifically, but these methods may be applied for the purification of other biomolecules such as proteins produced in mammalian or microbial cell culture.

Displacement chromatography as used for the purpose of this invention is a method of eluting species bound to a chromatography column (ion exchange resin) on the basis of competition between bound species and a molecule ("the displacer") that strongly binds to the column. The displacer binds directly to the resin to displace the bound molecule of interest and not to the molecule itself. Displacement chromatography is not a new chromatography technique, it was first defined by Tiselius in 1943, however it has not been regularly used in commercial protein chromatography due to a lack of suitable displacer molecules. In principle, displacement chromatography means operating a (known) chromatography column/assembly in a displacement mode (using a displacer molecule) instead of operating it in the elution mode (using an elution buffer / buffer solution).

The column is loaded with the mixture that is to be separated. Species in the mixture bind to the column with different affinities and bind approximately in order of affinity in diffuse bands. Displacement chromatography resolves these bands in to sharper bands in order of affinity improving separation of species on the basis of affinity for the resin. The displacer binds at the top of the column and displaces any high affinity product/impurities bound in this region which then re-bind further down the column in turn displacing other species with a slightly lower affinity for the resin. This cycle of binding/displacement/re-binding continues down the chromatography column as more displacer is loaded and results in different species being resolved in to distinct bands which are then eluted from the column in reverse order of affinity (weakest binding first). The technique allows high loading concentrations of product (~100g product per liter resin) to be used in contrast to current PrA elution mode chromatography (~30 g/l). The resins and displacer cost significantly less than the current preferred Protein A affinity chromatography resin. Displacement chromatography also has benefits compared to standard elution ion exchange chromatography where changes in conductivity or pH are used to elute bound proteins from the column; these include being able to operate at much higher loading concentrations and increased purity.

Displacement chromatography as described and defined above has been performed using chromatography media suitable for use in a large scale manufacturing process i.e. it has a large enough bead diameter to prevent pressure problems at scale and uses high flow rates. This contrasts with most of the work summarized above that used HPLC columns with much smaller resin particles at low flow rates that would only be suitable for high pressure analytical applications.

Operational loading concentration of product on to the column was found to be an important parameter for yield of high purity product (mAb). Low concentrations of product (e.g. 40 g/l) resulted in little purification while high loading concentrations close to the dynamic binding capacity of the resin resulted in good purification. A window of operating existed for loading concentration (80-110 g/l) where changing operational loading capacity did not have a large effect on purification, therefore suggesting that a minimum loading concentration is preferred. The ability of displacement chromatography to use high capacity protein concentrations is an improvement over the relatively low operational dynamic binding capacities used with

Protein A affinity chromatography. Moreover, with typical elution mode chromatography greater purification is achieved at lower loading capacities hence use of a displacement ion exchange step in the mAb purification process with high loading capacity is an improvement on the prior art.

The effect of pH was investigated as the chromatography matrix is an ion exchange resin. For the product examined (mAb cB72.3) a pH between 4.8 and 5.2 was found to be suitable for purification. Use of pH values outside of this range were found to result in greatly reduced amounts of protein being displaced (Figure 1). Displacer concentration was found to be an important variable for product purity. Lower concentrations of displacer were found to correlate with improved product purity and increased host cell protein clearance. To prevent excessive intermediate product volumes, a minimum displacer concentration was identified and is preferred.

Flow rate was found to have an effect on purification but at the minimum displacer concentration the effect was found to be relatively small.

Running displacement chromatography followed by an anion exchange flow through step resulted in further reduction in host cell impurities (e.g. total 2.25 log reduction). The anion exchange step was able to reduce HCP levels in displacement chromatography eluates that had been pooled with different yield criteria to similar levels.

DNA clearance over the displacement step was good and was close to a typical target DNA specification of <10 pg/mg.

Displacement chromatography reduced aggregate levels but increased levels of fragment. These product related impurities would be expected to be removed by other downstream steps.

The above findings primarily relate to the purification of mAbs by employing displacement chromatography and represent preferred embodiments which can be combined in any possible manner. These findings can as well be applied to the purification of other biomolecules by displacement chromatography such as for example peptides and oligonucleotides.

Especially with regard to purification of mAbs using displacement chromatography the preferred features and embodiments of its use and of the respective method of purification are specified in more details below, all of which can be combined in any suitable manner. Any specific combination being optimum for a specific mAb or biomolecule and/or for a specific feed stream can be found by routine experimentation using the knowledge as disclosed in the prior art.

Suitable and preferred chromatography media:

Suitable cation exchange chromatography media for displacement ion exchange may include (but not restricted to) Capto S (GE Healthcare), UNOsphere S (Bio-Rad), Fractogel EMD SO₃ ^" (M) (Merck) and Toyopearl Gigacap S-650M (Tosoh Biosciences). In general, cation ion exchange resins are the preferred chromatography media.

Operational loading concentration:

The loading capacity of the column should be in the range of 60-95% of the dynamic binding capacity (at 5% breakthrough).

pH-ranges:

The displacement chromatography step should be run at a pH value that will ensure a high dynamic binding capacity without causing damage to the protein structure. For cation exchange displacement chromatography using Capto S resin of a model lgG4, the optimal pH range was found to be approximately 2 pH units below the mean pi of the charged variants of the product. A window of operation of ±0.2 pH units was found to provide the maximum dynamic loading capacity of the column for target product. Variations outside this window were assumed to result in reduced process efficiency. The optimum pH operational window may vary depending on the product, resin chemistry and type of resin used. The operating pH may also be chosen to slightly reduce product binding but greatly reduce impurity binding hence increasing purification achieved by the step Displacer molecules:

Displacer molecules should have a high affinity for the stationary phase (i.e. the chromatography resin) so that displacement of other lower affinity compounds (i.e. the protein product and impurities) occurs. The displacer should have a relatively low molecular weight to facilitate removal in downstream steps such as ultrafiltration, the manufacturing/synthesis method should meet all the regulatory requirements for a compound used in cGMP manufacture of an injectable biopharmaceutical, the displacer should meet all relevant safety criteria (e.g. toxicity, teratogenicity).

An example of a commercially available displacer molecule is Expell SP1 (Sachem Inc, Texas, WO 2007/055896). Other families of potential displacer molecules have been invented, see WO 99/47574, WO 03074148, NO 963420 for examples.

Displacer concentration: Work using Expell SP1 (Sachem Inc, Texas) found that purification improved with decreasing displacer concentration but that elution volumes could be a significant problem for large scale manufacture, therefore a target displacer concentration of 2 to 5 mM was the optimal range. For a specific product the displacer concentration should be chosen to achieve the required purification with an eluate volume that is able to be handled by current biopharmaceutical plant engineering solutions.

Flow rate:

Flow rate was found to have a minor effect on separation under conditions tested therefore it was concluded that flow rate should be approximately the maximum of the recommended operational range for the chromatography resin used to optimize plant throughput.

Resins:

Resins should be suitable for large scale manufacturing in low pressure chromatography systems, e.g. resin mean particle diameter should be no less than 34 μm in diameter and would probably not exceed 200μm. Column sizes in manufacturing of between 10cm to 250cm diameter are most probable with bed heights from 10cm to 50cm being employed.

The above preferred embodiments are especially useful for the purification of mAbs. Purification means in another preferred embodiment the bind and elute capture step of antibodies and mAbs from CCS.

Displacement chromatography is used in combination with other pre- and/or post- displacement chromatography unit operations (purification and/or processing operations = orthogonal purification steps) to produce product of suitable purity for clinical use such as for example ultrafiltration, high performance tangential flow filtration, anion or cation exchange chromatography using packed bed, membranes or monolith supports, thiophilic adsorbtion chromatography, hydrophobic interaction chromatography, product or impurity precipitation, chemical treatment of cell culture supernatants. In a specific preferential embodiment displacement chromatography is followed by an anion-exchange step to further purify the desired mAbs. A further preferred embodiment would be to use chemical treatment of cell culture fermentation broths or supernatants and/or HP-TFF and/or charged dead end filters to reduce impurities that may precipitate during load conditioning for displacement chromatography. This combination of unit operations will allow the process to be performed without problems that may be encountered using methods in the prior art, for example precipitation caused by pH adjustment of cell culture supernatant is significantly reduced by use of a charged dead end filter.

Use of unit operations before displacement chromatography would also increase the yield of high purity product to be produced by displacement chromatography by reducing the level of impurities being loaded on to the chromatography column therefore combining the defined steps is one key aspect of the invention.

In the following preferred embodiments of the present invention and method, which can be combined in any sequence, are summarized, especially with regard to the use of displacement chromatography in combination with other purification steps/operations:

- The invention provides for a method for the purification of antibodies comprising using ion exchange chromatography in displacement mode (displacement chromatography) in combination with other pre- and/or post displacement chromatography purification steps.

- In this method displacement chromatography is used as the first bind and elute step. - The method is used for the purification of monoclonal antibodies from a feedstream that has been produced by mammalian cell culture

- The feedstream is a cell culture supernatant that has been chemically treated to remove impurities before displacement chromatography.

- A precipitating agent can be added to the feedstream to precipitate the product of interest or impurities, for example but not limited to host cell proteins (HCP) or host cell DNA, before displacement chromatography.

- The feedstream is passed through a charged dead end filter with the aim of reducing impurities before displacement chromatography.

- The feedstream is processed using ultrafiltration or high performance tangential flow filtration to reduce impurities before displacement chromatography.

- The feedstream is processed using anion exchange resin, membrane, or monolith chromatography to reduce impurities before displacement chromatography.

- A packed bed chromatography in displacement mode is used using a cation exchange chromatography resin that has a high dynamic binding capacity of >100 g/l at 5% breakthrough at a flow rate that produces a residence time of less than 3 minutes and low backpressure of <0.4 MPa with a resin bed height of 20 cm using a buffer with a viscosity similar to water at 20 ⁰C.

- An operating pH of less than the pi of the protein of interest preferably 2 pH units below the pi is used for displacement ion exchange chromatography such that dynamic binding capacity of the resin is high enough to allow formation of a displacement train. The pH may be chosen to ensure minimal impurity binding to the chromatography resin hence improving the purification process. - The pH and or conductivity of the equilibration buffer used in displacement chromatography is different to that of the displacer solution.

- In displacement chromatography a post load wash with a different pH or conductivity is used. - Displacement chromatography is followed by an anion-exchange step with or without any intermediate step.

- Displacement chromatography is followed by an thiophilic interaction step with or without any intermediate step.

- Displacement chromatography is followed by a hydrophobic interaction chromatography step with or without any intermediate step.

- Displacement chromatography is followed by mixed mode chromatography step with or without any intermediate step.

As will be further shown in the working examples the major advantages of using displacement chromatography (preferably in combination with other unit operations) can be summarized as follows:

High Cost Costs associated with PrA resins are due to the cost of the resin. Displacement chromatography uses a less expensive resin that works on the principle of ion exchange. Even given the cost of the displacer molecule, displacement chromatography is still significantly less expensive than PrA.

Low DBC. The use of ion exchange (IEX) resins allows choice of resin to maximize DBC. CaptoS (GE Healthcare) has been found to have a DBC of ~ 115 g/l for a recombinant monoclonal IgG₄, significantly higher than PrA. Moreover, displacement chromatography works better at higher loading concentrations which is an improvement over elution mode ion exchange methods published in the prior art. This allows greater efficiency in column productivity.

Low pH Elution. Displacement chromatography does not require a pH shift to elute bound product. A pH of ~ 5 reduces the risk of product damage. Protein A Ligand Leakage. As no Protein A ligand is involved in displacement chromatography this is not an issue.

Cleaning: ion exchange resins can typically be cleaned with 0.5 M sodium hydroxide which provides a more practical and stringent cleaning method than used for Protein A resins.

Using displacement chromatography in combination with other unit operations (e.g. Q anion exchange flow through, ultrafiltration) will allow purification of monoclonal antibodies avoiding the problems outlined above but producing product that meets the target specification. In a key aspect to this invention, use of unit operations before and after displacement chromatography allows a greater yield of high purity product that meets the required product specification. Another key aspect of the invention is that use of steps before and after a displacement ion exchange step overcomes problems associated with prior art solutions such as precipitation during conditioning of CCS and product produced by displacement chromatography not meeting the required purity specifications for a biopharmaceutical. The overall process aims to reduce costs and improve plant throughput compared to the current industry standard processes that use Protein A affinity chromatography.

Examples

Further characteristics of the invention, which can be employed in general, result from the following examples. In this context single characteristics of this invention alone or in any combination can be realized. The following examples are provided to illustrate preferred embodiments and are intended to be illustrative and not limitative of the scope of the invention.

Example 1 : Effect of pH

A 0.5 cm diameter x 15 cm bed height Tricorn column (bed volume 2.95 ml) was packed with CaptoS resin (GE Healthcare) in 0.2 M sodium chloride/20% ethanol at a flow rate of 750 cm/h (2.38 mi/min). The column was HETP tested with 5% CV 2.0 M sodium chloride and found to have an asymmetry of 1.09 and 1178 plates per meter. Equilibration buffers of 25 mM sodium acetate 25 mM sodium chloride were prepared and the pH adjusted using glacial acetic acid to the target (target pHs were 4.0, 4.5, 4.8, 5.0, 5.2 and 5.4) . The buffer was then adjusted to a conductivity of <5 mS/cm using deionised (Dl) water and the pH checked.

Load material was a cell culture supernatant (CCS) from a Lonza Biologies GS-CHO cell line LB01 expressing the chimeric monoclonal antibody cB72.3 at a titre of 2.95 g/l. The fermentation broth was harvested using a combination of centrifugation and diafiltration. The CCS was prepared for experimentation by adjustment to the target pH using glacial acetic acid and dilution to <5 mS/cm conductivity using Dl water.

Displacer solution was prepared by diluting 50 mM Expell SP1 Cation Displacer 10X Solution (Sachem Inc) to a final concentration of 2 rnM using 25 mM sodium acetate 25 mM sodium chloride of the appropriate pH.

Chromatography was performed at a product loading concentration of 80 mg/ml at a linear flow rate of 500 cm/h. All chromatography was performed using an AKTA Purifier system (GE Healthcare). The column was initially sanitized with 3 column volumes (CV) of 0.1 M sodium hydroxide and then equilibrated with 10 CV equilibration buffer of appropriate pH. Appropriately pH adjusted CCS was then loaded and then a 10 CV post load wash of equilibration buffer was used to remove any unbound molecules. A 20 CV displacement step using 2 mM Expell SP1 was then used to displace bound product from the column. Fractions (volume 1 CV) were collected. The column was then stripped using 2.0 M sodium chloride and sanitized with 3 CV 0.1 M sodium hydroxide.

Fractions eluted from each experiment were analyzed using pre-cast SDS polyacrylamide electrophoresis gels and the product purity in each fraction estimated by scanning laser densitometry. Chromatography profiles showed that pH 4.0, 4.5 and 5.4 were not suitable as a reduced amount of product binding was observed presumably due to product denaturation/damage or insufficient charge (Figure 1). pH 4.8, 5.0 and 5.2 were found to bind the expected amount of product . Based on quantification of SDS PAGE analysis it was concluded that within the range 4.8-5.2 pH did not affect the yield of high (>95%) purity product from this chromatography step (Figure 2).

Figure 1: Six pH values were trialed in displacement chromatography: pH 4.0 (1), pH 4.5 (2), pH 4.8 (3), pH 5.0 (4), pH 5.2 (5), pH 5.4 (6). All other parameters were constant: bed height = 15 cm, loading concentration = 80 mg/ml, flow rate 500 cm/h, 2 mM Expell SP1

Figure 2: Cumulative yield versus purity of each fraction (determined by densitometry of SDS PAGE gels) at pH 4.8 (3), pH 5.0 (4) and pH 5.2 (5)

Example 2: Displacement chromatography followed by anion-exchange chromatography.

An XK16 chromatography column packed with Capto S resin (GE Healthcare) was used to capture antibody from conditioned cell culture supernatant. The column was packed with over five column volumes of 0.1 M sodium hydroxide at 750 cm/h and had an internal diameter of 16 mm, a bed height of 203 mm and a column volume of 40.8 ml. An HETP test was performed at 100 cm/h using a tracer solution of 0.5 M sodium hydroxide. All chromatography was performed using an AKT-A Purifier system (GE Healthcare).

The column containing the Capto S cation-exchange resin was operated in displacement mode. Cell culture supernatant was pH adjusted to pH 5.0 +/- 0.10 with glacial acetic acid and then a four-fold dilution of the cell culture supernatant was performed with deionised water to reduce the conductivity to below 5 mS/cm. The diluted cell culture supernatant was then filtered with a 0.2 urn filter to remove precipitated proteins. The displacement column was operated at 500 cm/h as described in Table 1. The column was loaded with the conditioned cell culture supernatant to an antibody concentration of 110 mg/ml of resin. 2.0 mM Expell SP1 was used as the displaces Product fractionation during the displacement phase occurred when the absorbance at 280 nm of the column effluent was greater than 100 mAU. The volume of each fraction was 40.8 ml.

The antibody concentration of each fraction was determined by Protein A HPLC. The yield that could be achieved if each fraction was pooled with its previous fractions was calculated and then 10 ml samples of each fraction were taken and pooled to give mock pools equivalent to product step yields of 70%, 80% and 90%).

Table 1 : Operating conditions for displacement column

The three displacement chromatography pooled eluates were diafiltered using a Minim TFF system (Pall Corporation, USA) and a 30 kDa cut-off, 0.0093 m² area, PES membrane (GE Healthcare, UK). The diafiltration buffer was 10 mM Tris- HCI/ 50 mM sodium chloride pH 8.0 in each case. A Tricorn 5/200 chromatography column was packed with Capto Q resin to further purify the Mabselect SuRe (Protein A type resin; GE Healthcare, Sweden) eluate and the three displacement chromatography mock pools. The column was packed with over five column volumes of 0.1 M sodium hydroxide at 750 cm/h and had an internal diameter of 5 mm, a bed height of 199 mm and a column volume of 3.91 ml. An HETP test was performed at 100 cm/h using a tracer solution of 0.5 M sodium hydroxide.

The anion-exchange column was operated at 500 cm/h as described in Table 2. The column was loaded with either the Mabselect SuRe eluate or one of the three displacement chromatography mock pools to an antibody concentration of 80 mg/ml of resin. Product collection during the load and post load wash commenced when the absorbance at 280 nm of the column flow through was greater than 100 mAU and terminated when it fell below 200 mAU.

Table 2: Operating conditions for anion-exchange column

SDS PAGE was used to investigate the purity of products produced by the displacement and anion exchange process at different yields as compared to a process using MabSelect SuRe bind and elute chromatography followed by CaptoQ flow through chromatography (Figure 3). This analysis indicated that product purity after displacement/CaptoQ was >97% which is slightly lower than that observed with Protein A MabSelect SuRe chromatography.

A CHO host cell protein ELISA was used to estimate the cumulative log clearance of HCP over the two step processes (Figure 4). This showed that varying the yield from the displacement step affected the HCP clearance (lower yield on displacement resulted in higher HCP clearance) but that after the CaptoQ step the cumulative log clearance for all displacement samples was 2.5 log. This compares to a MabSelect SuRe/CaptoQ process with 3.5 log HCP clearance. However, other steps before or after displacement would further reduce HCP levels to the required level.

CHO DNA levels after each step in the displacement/CaptoQ were compared with a MabSelect SuRe/CaptoQ process using a qPCR assay (Figure 5). For the displacement/CaptoQ sample the CHO DNA was quantified as 35.1 pg/mg whereas for the MabSelect SuRe/CaptoQ process the DNA level was 3 pg/mg. However, as the target for DNA is typically <10 pg/mg the displacement/CaptoQ process would be expected to satisfy this target when other steps are incorporated, such as for example use of chemical treatment of cell culture supernatants and/or HP-TFF and/or charged dead end filtration and/or precipitation of product or impurities before displacement chromatography to reduce impurity load on to the capture step or hydrophobic interaction chromatography, and/or thiophilic interaction chromatography and/or mixed mode chromatography after the displacement/CaptoQ process.

These data demonstrate that using another step in combination with displacement chromatography can further improve the purity of mAb product. Use of a suitable combination of steps as defined herein would produce product that meets the required purity specifications.

Figure 3: Reducing SDS PAGE of samples produced by displacement and anion exchange chromatography

Figure 4: Graph of cumulative log clearance of host cell protein from three displacement/ CaptoQ chromatography processes with varying step yields compared to a MabSelect SuRe/CaptoQ process. (7) CCS conditioning, (8) capture, (9) ultrafiltration, (10) anion-exchange

Figure 5: CHO DNA clearance over a two step process using either displacement or Protein A affinity chromatography in combination with CaptoQ anion-exchange chromatography. (11) cell culture supernatant, (12) conditioned CCS, (13) capture step eluate, (14) ultrafiltration retentate, (15) anion-exchange eluate. Example 3: Displacement chromatography followed by thiophilic interaction chromatography Mab B72.3 (see Ex. 1) at a concentration of 3.08 g/L (pH 7.21 , conductivity 20.73 ms/cm) was harvested and filtered.

The material was concentrated 4-fold using a Millipore PrepScale spiral-wound TFF cartridge (2.5ft2, 3OkDa regenerated cellulose Cat No. CDUF002LT) and a Watson Marlow 504U peristaltic pump. The actual start volume of CCS was 2016.9ml. 1512.7ml of permeate was removed, and then 1506ml of lab grade water was added to reduce conductivity of the material. The final volume of the material was 2020.7ml, with a conductivity of 5.67mS/cm and a pH of 7.21.

10ml of 2M acetic acid was added in a dropwise manner whilst mixing to give a final pH of 4.99 and a conductivity of 5.70mS/cm. During this titration, significant precipitation occurred, and after 1 hour, this was removed by 0.22μm filtration using a Millipak Gamma Gold 60 capsule filter. The material was then further diluted to bring the conductivity to 2.86mS/cm. Material was prepared in a similar way for runs 2 and 3.

For displacement chromatography, Capto S (GE Healthcare, Uppsala, Sweden) was used as the chromatography matrix and Expell SP1 (Sachem Inc, Texas, USA) was used as the displacer molecule. The chromatography process is summarised in Table 3. Table 3: Capto S Displacement Chromatography run details

The eluate material from all 3 Capto S Displacement chromatography runs was pooled together and concentrated to 26.63 g/l. This was then diafiltered against 5OmM Sodium Phosphate buffer pH7.5. This material was adjusted with combination of 5OmM Sodium Phosphate buffer and 5OmM Sodium Phosphate, 1M ammonium sulphate (and titrated using either 10M NaOH or concentrated HCL) to achieve the required load conditions.

PlasmidSelect Xtra (GE Healthcare, Uppsala, Sweden) was used as a second purification step. The parameters used for this thiophilic interaction chromatography are summarised in Table 4.

Table 4 PlasmidSelect Xtra Evaluation Parameters

Eluate samples and the load product were then analysed for host protein using an ELISA assay method. As shown in Table 5, there was a reduction in host cell protein levels under all conditions tested. Reduction of HCP ranged between 0.38 log and 1.38 log with a mean clearance of 0.76 log. This exemplifies that combining displacement chromatography with an orthogonal purification unit operation allows further reduction of impurities. Table 5

_{* * * * *}

Claims

1. A method for the purification of antibodies comprising using ion exchange chromatography in displacement mode (displacement chromatography) in combination with other pre- and/or post displacement chromatography purification steps.

2. The method of claim 1 , wherein displacement chromatography is used as the first bind and elute step.

3. The method of claim 1 or 2 for the purification of monoclonal antibodies from a feedstream that has been produced by mammalian cell culture.

4. The method of claim 3, wherein the feedstream is a cell culture supernatant that has been chemically treated to remove impurities before displacement chromatography.

5. The method of any of claims 3 or 4, wherein a precipitating agent has been added to the feedstream to precipitate the product of interest or impurities before displacement chromatography.

6. The method of any of claims 3 to 5, wherein the feedstream is passed through a charged dead end filter before displacement chromatography.

7. The method of any of claims 3 to 6, wherein the feedstream is processed using ultrafiltration or high performance tangential flow filtration to reduce impurities before displacement chromatography.

8. The method of any of claims 3 to 7, wherein the feedstream is processed using anion exchange resin, membrane, or monolith chromatography to reduce impurities before displacement chromatography.

9. The method of any of claims 1 to 8, wherein a packed bed chromatography in displacement mode is used using a cation exchange chromatography resin that has a high dynamic binding capacity of >100 g/l at 5% breakthrough at a flow rate that produces a residence time of less than 3 minutes and low backpressure of <0.4 MPa with a resin bed height of 20 cm using a buffer with a viscosity similar to water at 20 ⁰C.

10. The method of any of claims 1 to 9, wherein displacement chromatography is followed by an anion-exchange step with or without any intermediate step.

11. The method of any of claims 1 to 9, wherein displacement chromatography is followed by an thiophilic interaction step with or without any intermediate step.

12. The method of any of claims 1 to 9, wherein displacement chromatography is followed by a hydrophobic interaction chromatography step with or without any intermediate step.

13. The method of any of claims 1 to 9, wherein displacement chromatography is followed by mixed mode chromatography step with or without any intermediate step.

14. The use of displacement chromatography for the purification of antibodies according to any of claims 1 to 13.

_{* * * * *}