FIELD OF INVENTION

The present invention relates to a rapid and efficient method for chromatographic fractionation of charge and/or conformation variants of a recombinant human Fc fusion protein, wherein the separation is high pressure liquid chromatography ("HPLC") based.

BACKGROUND OF INVENTION

Production of recombinant proteins is typically associated with multiple protein variants. These variants are predominantly charge and/or conformation based and can influence biological outcomes of a protein preparation.

Conformation based variants refer to protein species exhibiting structural heterogeneity in solution. Such variants may result from protein misfolding, aggregation, dimerization or multimerization.

Similarly, charge based variants are formed due to heterogenic distribution of charged entities in the molecule. Charged sugar moieties, deamidation or differential processing of amino acids, such as C-terminal lysine may contribute to such heterogeneity.

Etanercept (US5605690) (ENBREL® Immunex Corporation) is a recombinant human Fc containing fusion protein consisting of 934 amino acids, wherein the (p75) extracellular ligand-binding portion of the human tumor necrosis factor receptor (hTNF-R) has been linked to the Fc portion of a human IgGl antibody. It is an acidic protein, with isoelectric point ("pi") in the range of 4.5-5.5. The net negative charge contributing to the acidic pi is primarily due to sialylation of the molecule.

Production of recombinant etanercept typically leads to accumulation of conformation, and charge variants due to heterogeneous sialylation of the molecule.
However apart from sialylation, alteration of the protein conformation may also contribute to heterogeneous surface charge distribution of the protein.

Different techniques have been proposed for the separation of protein variants. These include two electrophoresis based techniques - isoeletric focusing ("IFE") and capillary zone electrophoresis ("CZE"). However, IEF and CZE are limited by their inability to effectively separate conformation variants. C. Jochheim et, al., (Chromatographia, 53,2001, S59-S65) has described capillary electrophoresis based method for separating charge variant/s present in an etanercept preparation.

Moreover, isoelectric focusing gel is also limited by its inability to measure relative proportion of different variants present in a preparation. However, while image capillary zone electrophoresis ("iCE") tries to address this problem, several constraints remain including- lengthy sample preparation steps, pretreatment, excipient removal, lengthy run time, lack of robustness, cost effectiveness, lengthy assay validation steps, lack of consistency and reproducibility.

Similarly, various chromatography based approaches have been tried. However, the chromatographic conditions including the material and mobile phase characteristics are dictated by the physico-chemical properties of the variant/s in question.

De La Calle Guntiiias et al., (Anal. Bioanal. Chem. 378,2004, 383) describe a HPLC based method, and suggest use of a Dinoex ProPac PA1 column for separating sialylated glycoform variants of the protein trasnferrin.

Podgornick et al., (J Chromatogr B. 799. 2004, 343) discloses a means of separating manganese peroxidase isozymes using a strong anion exchange monolithic column. Additionally use of shallow salt and pH gradient has been suggested. The salt concentration was maximally altered only about two folds over the binding buffer concentration.

Considerably low salt concentration favoring binding of monoclonal antibody isoforms, followed by fractionation of antibody isoforms by simultaneous linear increase of pH and reduction of salt concentration have been described in Kaltenbrunner et al., J. Chromatogr. 639,1993,41.

Size exclusion chromatography ("SEC"), has also been proposed for separating protein conformation variants, but is hindered by its inability to separate charge variant of the protein. In addition, use of large column volumes can be cumbersome.

Given the complexity of different variant/s present in recombinant human Fc fusion class of protein preparation, such as etanercept, there is need for a robust and efficient method for fractionating said variant/s.

The principal object of the invention is to provide a method of fractionating charge and/or conformation variants present in a recombinant human Fc fusion protein preparation, wherein the chromatography is high performance liquid chromatography ("HPLC") based. Another object of the invention is to describe a pH gradient, a salt gradient or a combination thereof for separating charge and/or conformation variant/s present in a recombinant human Fc fusion protein preparation by any HPLC. A further object of the invention is to reduce the total operation time for such separation, and in addition to provide a robust method capable of handling recombinant human Fc fusion protein preparations derived from different source or stages of preparation. The disclosed method can be used for separating said variants and measuring relative proportion of variant/s present in the recombinant human Fc fusion protein preparation.

SUMMARY OF INVENTION

This present invention discloses a method of chromatographic fractionation of charge and/or conformation variants of a recombinant human Fc fusion protein, present in a preparation comprising the same, wherein the chromatography is HPLC based. The method comprises, binding the variant/s to a strong ion exchange column, and differentially eluting the said variant/s by employing a mobile phase capable of establishing a pH gradient, a salt gradient or a combination thereof.

BRIEF DESCRIPTION OF DRAWINGS

Figure 1: Illustration of isoelectric focusing gel of sialylated and neuraminidase treated etanercept by the procedure described in Example 1

Figure 2: Illustration of isoelectric focusing gel of aggregated etanercept by the procedure described in Example 1

Figure 3: Illustration of strong anion exchange HPLC of etanercept variants by the procedure described in Example 3

Figure 4: Illustration of strong anion exchange HPLC of etanercept variants by the procedure described in Example 4

Figure 5: Illustration of strong anion exchange HPLC of etanercept variants by the procedure described in Example 5

DETAILED DESCRIPTION OF THE INVENTION

Charge and conformational heterogeneity of a recombinant protein preparation are primarily dictated by production conditions, such as cell culture, media composition, or purification steps. These heterogeneities are known to affect the biological outcomes of a therapeutic preparation. Fractionation of these variants may aid in measuring relative proportion of different variants present in the said preparation, and predict biological outcomes of the preparation.

In various embodiments, the disclosed invention provides a method of separating charge and/or conformation variant/s of a recombinant human Fc fusion protein, wherein;

a) the recombinant human Fc fusion protein preparation comprising the charge and/or conformation variant/s is bound to a strong ion exchange support, in presence of a suitable mobile phase buffer at a suitable pH and/or having a salt concentration, and

b) differential elution of the said charge and/or conformation variants from the said column by using either a pH gradient, a salt gradient or a combination thereof, wherein the fractionation is performed in a HPLC.

In one embodiment of the invention, a strong anion exchange support is used for an acidic recombinant human Fc fusion protein. Alternately a strong cation exchange support may be used for a basic recombinant human Fc fusion protein.

In another embodiment, the recombinant human Fc fusion protein is etanercept.

In a further embodiment of the claimed invention, a combination of pH and salt gradient is used simultaneously in order to differentially resolve charge and/or conformation variant of a recombinant human Fc fusion protein present in said preparation, wherein the fractionation is performed in a HPLC.

In another preferred embodiment, a salt gradient is used, wherein the fractionation is performed in a HPLC. In yet another embodiment, the salt gradient is used for separation of de-sialylated variant of a recombinant human Fc fusion protein.

In another embodiment, a pH gradient is used, wherein the fractionation is performed in a HPLC. In yet another embodiment, the pH gradient is used for separation of aggregate variant of a recombinant human Fc fusion protein.

In another embodiment of the invention, the total run time for the assay may be significantly less. In some non-limiting examples it may be less than or equal to one hour.

In another embodiment of the invention, the salt concentration in the gradient is varied more than about fivefold over the binding buffer concentration. The disclosed salt gradient can be used alone or in combination of a pH gradient, wherein the fractionation is performed in a HPLC.

In another embodiment of the invention, the pH of the elution buffer is altered such that it migrates towards the isoelectric point of the variant. The said pH gradient can be used alone or in combination with a salt gradient, wherein the fractionation is performed in a HPLC.

Several buffers are known in the art, some non-limiting examples include acetate buffers, amino acid buffers such as histidine buffer, Tris buffer, MOPS buffer. Different salts may be employed for generating a salt gradient, some non limiting examples include sodium chloride, potassium chloride, magnesium chloride.

Fractionation of the variants as per the disclosed invention can aid in measuring relative proportion of different variants present in the recombinant Fc fusion protein preparation with considerable efficiency and rapidity, and therefore predict biological outcomes of the preparation.

DEFINITIONS

"Conformational" or "conformation" variant/s refer to various structural variants of a protein, such as misfolded, aggregate, inclusive of dimers, trimers or multimers. The charge variants refer to various species resulting from charge heterogeneity. In some instances, variants resulting from conformation changes may also alter surface charge of the recombinant human Fc fusion protein, thereby contributing to charge heterogeneity.

The term heterogeneity can be used interchangeably with micro-heterogeneity. Human Fc fusion protein in context of the present invention refers to a recombinant fusion protein, wherein the human Fc region has been fused to a binding domain, capable of specific binding to a ligand. Some non limiting examples include etanercept, abatacept.

In context of the present invention, the term support refers to any immobilized or immobilizable matrix, including resins, monoliths, membranes or other form of strong ion exchange support known in the art, usable or compatible with HPLC based fractionation systems.

EXAMPLES

Example 1

Isoelectric focusing Isoelectric focusing was done in a vertical slab gel mode with commercially procured precast gels with pH gradient of 3-7. Sample buffer and electrophoresis buffers were procured commercially. Standard electrophoresis run conditions were modified so as to achieve optimal resolution between charge variants. Post run the gel was stained with Colloidal Blue Stain for detection of protein bands.

Figure 1 represents a typical charge variant/s profile of etanercept (lane 1 and lane 2), the load of etanercept is variable. Neuramindase treatment of etanercept altered the isoelectric point of the molecule as the enzyme is known to cleave terminal sialic acid residues. This is reflected by limited migration of the neuramindase treated etanercept in the isoelectric focusing gel (Figure 1; lane 3). Though charge variant/s were separated electrophoretically, the process was limited by its ability to measure relative proportion of different variant.

ENBREL® formulation was deliberately contaminated by adding 5-50% forced aggregated etanercept prior to loading in an IEF gel (Figure 2; lane 2-4). The reference drug substance (ENBREL®), devoid of aggregate variants was used as control (Figure 2; lane 1). Aggregate variant/s exhibit limited migration in the gel, additionally in spite of a ten-fold difference in the amount of aggregate variants between lane 2 and 4, the IEF profiles were almost indistinguishable, thus significantly highlighting the insufficiency of gel isoelectric focusing in resolving conformation variants, and ability to measure relative proportion of variants present in the preparation.

Example 2

Sample Preparation

For all HPLC based methods described herein, identical steps were followed. Samples analyzed were either of etanercept final drug substance formulation (ENBREL®), in-process sample, neuraminidase treated etanercept, or force aggregated etanercept. In-process etanercept samples were obtained by subjecting harvested cell culture broth to at least one round of Protein A affinity purification. Samples were diluted prior to injection in HPLC, when required, in appropriate background buffer such that the absorbance response was within the detector linear range.

Instrumentation and Blank

Waters Alliance HPLC system housing a strong anion exchange column was used. The differentially eluted variants were detected using a PDA detector by measuring UV light absorbance at 214nm. Additional or alternate detection systems measuring protein internal fluorescence may also be utilized. Detection of protein internal fluorescence may be additionally useful in assessing and/or detecting variant/s, especially of conformation type.

The HPLC profiles for the test samples were analyzed by integrating peaks after subtracting respective buffer blanks. Integration of the chromatographic profile was done so as to estimate relative proportion of different variant/s present in the preparation. The retention time and percentage area under the curve ("%AUC") was calculated.

Example 3

The sample preparation and instrumentation is described in Example 2. Simultaneous pH and salt gradient was used for separation of different charge and conformation variants. Etanercept final drug substance formulation (ENBREL®), in-process sample, neuraminidase treated etanercept, or force aggregated etanercept were analyzed as described herein.

Salt and pH Gradient

The pH- cum- salt gradient was achieved by using a system comprising three mobile phases. Mobile phases A & B were equimolar concentrations of di-sodium hydrogen phosphate and sodium di-hydrogen phosphate, respectively. Mobile phase C consisted of high salt. The pH during the gradient was varied between 9.2 and 4.4, while salt concentration was varied between 0- 0.5 M.

Figure 3 depicts the HPLC profile of (a) etanercept, (b) in-process etanercept, (c) de-sialylated etanercept and (d) forced aggregated etanercept. The pH was altered such that it migrated towards the isoelectric point of the variant. The entire run could be completed in less than an hour from start until equilibration of the column for subsequent run.

Lack of sialylation renders a very low net negative charge to molecule, consequently this variant was observed to elute first (Figure 3; C) followed by the hemi to appropriately-sialylated variants (Figure 3; A), while the aggregate variant eluted towards the end of the run (Figure 3; D). Figure 3, B; illustrates an in-process sample containing the different proportions of charge and conformation variants. The profile could be broadly divided in three main regions. Whereas region 1 and 3 corresponded to the minimally sialylated and aggregated variants respectively, region 2 corresponded to the hemi to appropriately-sialylated variant. The relative proportion of different variant present in the preparation was determined by calculating area under the curve as illustrated in table 1. Figure 3 further illustrates the robustness of the method as samples derived from different source or stages of preparation could be handled effectively.

Peak

Elution Time range

Percentage Abundance

Peakl; Minimally sialylated

4-10 min

19.4%

Peak 2;

Sailylated

Variant

10-24 min

75.8%

Peak 3;

24-26 min

4.8%

Aggregate

Variant

Table 1: Illustrates the elution conditions for different variant/s of etanercept when a simultaneous salt and pH gradient is employed. The variants are quantitated by calculating the percentage area under the curve (%AUC)

Example 4

Resolution of variants by a salt gradient was explored as described herein. Sample preparation and instrumentation used, are as explained in Example 2.

Salt Gradient

The salt gradient was achieved by using a system comprising three mobile phases; wherein the pH was maintained at > 6 for elution of all variants while salt concentration was changed from 0 - 0.5 M. Figure 4 illustrates the HPLC strong anion exchange profile of in-process etanercept by using a salt gradient. While the salt concentration of the mobile phase was varied, the pH was kept constant at about > 6. The profile could be primarily divided in three main regions (Figure 4). The region 1 corresponded to the minimally sialylated variants, whereas most of the other variant/s could be separated and eluted in the region 2. Region 3 represents non-proteinaceous peak. Resolution of minimally sialylated variants was possible by utilizing only a salt gradient.

Example 5

pH Gradient The pH gradient was achieved by using a system comprising three mobile phases; wherein salt concentration was maintained at 150 mM, while the pH was varied between 9.2 and 4.4. Figure 5 illustrates the strong anion exchange HPLC profile of etanercept. The variant/s were resolved by only altering pH while maintaining a constant salt concentration. The salt concentration was kept constant at about 150mM, while the pH was varied over time form 9.2 to 4.4 in the time span of 6-15 minutes. The profile could be primarily divided in two regions (Figure 5). Whereas the aggregated variant/s were resolved in region 2, non-aggregated variant/s resolved in the region 1. Efficient resolution of aggregated variants was possible by utilizing a pH gradient (Figure 5).

Therefore, whereas the salt gradient was found capable of separating charge based variants (Figure 4), the pH gradient was found to be capable of fractionating conformation based variants (Figure 5). Simultaneous use of salt and pH gradient exhibited synergism and both conformation and charge based variants could be fractionated. Additionally the process was found robust to handle recombinant human Fc fusion protein derived from different source or stages of preparation with significant rapidity.

WE CLAIM:

1. A method of separating charge and/or conformation variant/s of a recombinant human Fc fusion protein, wherein;

a) the recombinant human Fc fusion protein preparation comprising the charge and/or conformation variant/s is bound to a strong ion exchange support, in presence of a suitable mobile phase buffer at a suitable pH and/or having a salt concentration, and

b) differential elution of the said charge and/or conformation variants from the said column by using either a pH gradient, a salt gradient or a combination thereof, wherein the fractionation is performed in a HPLC.

2. A method according to claim 1, wherein a combination of pH and salt gradient is used in order to differentially resolve charge and conformation variant of a recombinant human Fc fusion protein present in a preparation, wherein the fractionation is performed in a HPLC.

3. A method according to claim 1, wherein a salt gradient is used for separation of de-sialylated variant of a recombinant human Fc fusion protein; wherein the fractionation is performed in a HPLC.

4. A method according to claim 5, wherein the pH gradient is used for separation of aggregate variant of a recombinant human Fc fusion protein; wherein the fractionation is performed in a HPLC.

5. A method according to claim 1, wherein the salt concentration in the gradient is varied more than about fivefold over the binding buffer concentration.

6. A method according to claim 1, wherein the pH of the elution buffer is altered such that it migrates towards the isoelectric point of the said variant, wherein the fractionation is performed in a HPLC.