EP2051989A1 - One step imac (mcac) purification of proteins - Google Patents

Abstract

This invention relates to a process for isolation and/or purification of a protein of interest and to the use of IMAC as a capture and the main separation step. In one embodiment, the invention relates to a process for the isolation and/or purification of a protein of interest by use of immobilized metal affinity chromatography (IMAC), comprising the steps of (a) applying a mixture or solution comprising said protein to an IMAC matrix, (b) separating contaminant(s) by use of metal ion affinity of said protein to said IMAC matrix, under conditions suitable to keep said protein bound to the matrix, (c) separating the protein of interest from further contaminant(s) by use of both a non-metal ion affinity interaction and a metal ion affinity interaction between said protein and said IMAC matrix.

Description

Title of the invention

ONE STEP IMAC (MCAC) PURIFICATION OF PROTEINS

Field of the invention

The invention relates to use of Immobilized Metal-Ion Affinity Chromatography (IMAC) for isolation and/or purification of a protein of interest. In particular, this invention relates to a process for isolation and/or purification of Granulocyte-Colony-Stimulating Factor (G-CSF).

The principle of the present invention can be applied to a wide range of heterologous and homologous proteins as well as fusion proteins, expressed in a variety of host cells.

Background of the invention

Whereas various techniques can be employed to produce high yields of a crude protein of interest, the isolation and purification of the protein is not a simple matter. A typical isolation procedure comprise multi-stage processes comprising several purification steps e.g. different chromatographies, chromatofocusing, filtrations etc. In addition, in the known processes for isolation and/or purification of proteins the number of additional substances are used that could potentially contaminate the final product and furthermore, a number of toxic substances representing a pollution burden for environment e.g. denaturants, detergents, organic solvents (waste water) are normally used.

In the known purification processes where IMAC is used as capture and the key separation step, the improvement is needed, especially for use in the industrial scale. Furthermore, several described processes for isolation and/or purification of proteins involve very expensive chromatographic media, are inappropriate for cleaning-in- place and sanitization procedures or might represent a bottleneck in large-scale purifications (e.g. size exclusion chromatography, affinity matrices with immobilized monoclonal antibodies etc.).

The nature of the prior art procedure may lead to lower purity of a protein of interest, to higher costs of the protein production, to the environmental pollution and can be time-consuming. Therefore, in the view of the processes described above, there is a strong need for a simplified process for protein purification in a high yielding, cost- effective and, environmental-friendly matter.

Such simplified process for isolation and/or purification would comprise minimal number of purification steps, which would in addition have optimized technological parameters, capacity of chromatographic media, number of cleanings-in-place, sanitations cycles and similar. Thus, in the simplified process for purification the potential contaminants would be kept as low as possible during the process. In addition, in the simplified process for protein purification toxic substances should be kept as low as possible.

In the purification processes where Immobilized IMAC is used as capture and the main separation steps, the improvement is needed. Simplified isolation and/or purification procedure shall be devoid of chromatographic steps that involve very expensive chromatographic media, are inappropriate for cleaning-in-place and sanitization procedures or might represent a bottleneck in large-scale purifications (e.g. size exclusion chromatography, affinity matrices with immobilized monoclonal antibodies etc.).

The present invention provides such a simplified process for isolation and/or purification of a protein of interest, in particular G-CSF.

The object of the present invention is to use IMAC, for the isolation and/or purification of a protein of interest, as both, a capture and the main separation step by combining two or more interaction and/or separation principles acting simultaneously. The approach of present invention can be applied to a wide variety of proteins.

In particular, the process for the isolation and/or purification of the present invention relates to isolation and/or purification of G-CSF. In this particular embodiment, IMAC as described above is used as the main chromatographic step. The process can additionally comprise a chromatographic step that serves as a polishing step and can be accomplished by various chromatographies. For final purity ultrafiltration / diafiltration can be applied.

The use of IMAC according to the present invention enables fast, environmental- friendly, good yield and cost-effective preparation of pharmaceutical-grade purity proteins.

In the case of G-CSF, the process of the present invention leads to over 98% RP- HPLC) purity of G-CSF and over 99% SE-HPLC purity of G-CSF with dimers and higher aggregates at the detection limit or below it.

Summary of the invention

It is an object of the present invention to use IMAC as both, a capture and the main separation step by combining two or more interaction/separation principles acting simultaneously. Such use enables simpler purification and/or isolation of proteins, and provides biologically active proteins in highly purified and active form, as well as a pharmaceutical composition comprising the same.

According to the present invention it was surprisingly found that a simplified process for isolation and/or purification of a protein of interest can be achieved by a combination of various separation principles, comprising IMAC as the main chromatographic step functioning as both, a capture and the main separation step. Said main separation step involves two phases: the first phase uses a classical separation principle based on the specific protein affinity to the IMAC matrix with immobilized metal ions including binding at high pH and elution with lower pH buffer. In the second phase, a separation of the bound proteins is achieved using a combination of pH/salt gradient. The combination of two different interaction principles acting simultaneously on the same matrix results in better resolution and higher purity of the target protein. In particular, the use of IMAC according to the present invention can be applied in the two-stage chromatographic process for isolation and/or purification of a protein of interest, e.g. G-CSF. The IMAC as described above can be used as the main chromatographic step leading to such a purity of a protein of interest, e.g. G-CSF that only one additional chromatographic step that serves as a polishing step can be used. This step can be accomplished by different chromatographies, e.g. affinity chromatography, hydrophobic interaction chromatography, reversed phase chromatography, anion exchange chromatography, cation exchange chromatography hydroxyapatite chromatography, and other chromatographies, depending on the nature of the target protein.

Ultrafiltration/diafiltration as the final isolation phase can be used for final buffer exchange and with a new function of final removing of protein dimers and other aggregates.

The use of IMAC according to the present invention enables preparation of pharmaceutical-grade purity proteins. In the case of G-CSF, the process for isolation and/or purification leads to over 98% RP-HPLC purity and over 99 % SE-HPLC purity of G-CSF with dimers and higher aggregates close to the detection limit or below it.

Detailed description of the invention and preferred embodiments thereof

In the context of the present invention it has been surprisingly found that a unique way of using IMAC can lead to simpler, cost effective process for isolation and/or purification of proteins leading to high purity of proteins. Although the concept of the present invention can be applied to a wide variety of homologous and heterologous proteins, the present invention is particularly developed for the production of G-CSF.

In the present invention IMAC is used and functions as both, a capture and the main separation step by combining two or more separation principles acting simultaneously. A process for isolation and/or purification of proteins by using IMAC was described in WO03051922, where IMAC was performed by using only one interaction principle with no additional separation principle acting simultaneously. In this case IMAC was performed by a step gradient elution and the whole chromatographic peak at 100% second elution buffer was pooled and used for further purification. In addition, IMAC was usually performed in such a way that binding and elution buffers contained high amounts of salt (usually NaCI or KCI) to reduce ionic interactions.

By using IMAC according to the present invention the stronger non-specific interactions than expected are surprisingly noticed. Specifically, G-CSF does not elute from the IMAC matrix by lowering pH if no salt is present in the elution buffer. As pure metal ion-protein affinity interactions are reduced at low pH and negligible at sufficiently low pH, this is probably due to residual non-specific interactions. In the prior art classical approaches these residual interactions are completely excluded by, e.g. using salt buffer. In the present invention it is unexpectedly found out that contrary to the conventional approach, residual interactions can be efficiently used for improving separation power of IMAC. One way to achieve this additional separation is to apply combined elution gradients. This might be even more important in resolving structurally very similar product-related substances, because they in general have very similar affinity regarding the IMAC principle but are sufficiently different to be separated due to residual interactions with the chromatographic matrix.

The improvement is especially important on the industrial scale because it contributes to faster, environmental-friendlier and cost-effective production of a protein of interest.

In the main aspect of the invention, the use of IMAC for the isolation and/or purification of a protein of interest comprises:

1. a classical separation principle based on the specific protein affinity to the IMAC matrix with immobilized metal ions including binding at high pH and elution with lower pH buffer or by the addition of competing substances, such as imidazol, histidine or ammonium salts, and acting simultaneously

2. a further separation of the bound proteins by using a combination of pH/salt gradient (decreasing pH with simultaneous rising salt concentration) or a combination of competing substance/salt gradient (rising concentration of the competing substance with simultaneous rising salt concentration).

Therefore, in the main aspect the invention relates to a use of IMAC for the isolation and/or purification of a protein, wherein said use comprises:

- a classical separation principle based on the specific protein affinity to the IMAC matrix with immobilized metal ions, and acting simultaneously

- a further separation of the bound proteins by using a combination of pH/salt gradient or a combination of competing substance/salt gradient

The combination of two different interaction principles acting simultaneously on the same matrix results in better resolution and higher purity of the target protein.

The competing substance is preferably selected from the group consisting of imidazol, histidine or ammonium salts.

The purity indicated herein refers to RP-HPLC and SE-HPLC purity.

The term 'RP-HPLC purity' as used herein refers to the protein purity which is determined by RP-HPLC.

The term 'SE-HPLC purity' as used herein refers to the protein purity which is determined by SE-HPLC.

Accordingly, in the main aspect of the present invention, there is provided a use of IMAC for the isolation and/or purification of a protein of interest, comprising: a) loading a solution or mixture, which comprises a protein, on the IMAC support b) selective binding of biologically active form of a protein to the IMAC support at high pH, optionally washing the IMAC column, c) eluting protein and non-protein contaminants from the IMAC column by lower pH buffer or by the addition of the competing substances such as histidine, imidazole or ammonium salts d) eluting the biologically active form of a protein of interest from the IMAC column by using combined gradient (pH/salt or competing substance/salt gradient) e) identification of fractions to be pooled or rejected (using SDA-PAGE, RP- HPLC and SE-HPLC) f) pooling of the fractions

Therefore, in the main aspect the invention relates to a use of IMAC for the isolation and/or purification of a protein comprising: - loading a solution or mixture, which comprises a protein, on the IMAC support

- selective binding of biologically active form of a protein to the IMAC support at high pH, optionally washing the IMAC column,

- eluting protein and non-protein contaminants from the IMAC column by lower pH buffer or by the addition of competing substances - eluting the biologically active form of a protein of interest from the IMAC column by using combined gradient

- identification of fractions to be pooled or rejected

- pooling of the fractions

The competing substance is preferably selected from the group consisting of imidazol, histidine or ammonium salts.

The process of the present invention can be used in case of purification and/or isolation of proteins, which are selected from the group consisting of granulocyte - colony stimulating factor (G-CSF) and G-CSF derivatives such as methionyl G-CSF (Met-G-CSF), glycosylated, enzymatically or chemically modified (e. g. pegylated) G- CSF, G-CSF analogues and the fusion proteins which comprise G-CSF, other fusion proteins (e.g. with albumin, with Fc), His Tagged proteins, interferons (IFN), such as INF-beta 1 b, IFN-beta 2b, IFN-gamma 1 b, interleukins (IL), such as IL-2 and IL-4, granulocyte macrophage-colony stimulating factor (GM-CSF), macrophage-colony stimulating factor, (M-CSF), epidermal growth factor (EGF), human serum albumin (HSA), deoxyhbonuclease (DNAse), fibroblast growth factor (FGF), tumour necrosis factor alpha (TNF alpha) and tumour necrosis factor beta (TNF-beta) and further comprises other proteins.

Preferably, the selected protein is G-CSF.

Accordingly, in the particular aspect of the present invention, there is provided a use of IMAC for the isolation and/or purification of G-CSF, comprising :

(1 ) loading a solution or mixture, which comprises G-CSF, on the IMAC support

(2) selective binding of biologically active form of G-CSF to the IMAC support at high pH, optionally washing the IMAC column,

(3) eluting protein and non-protein contaminants from the IMAC column by lower pH buffer or by the addition of the competing substances such as histidine, imidazole or ammonium salts,

(4) eluting the biologically active form of G-CSF of interest from the IMAC column by using combined pH and salt gradient (pH/salt or competing substance/salt gradient)

(5) identification of fractions to be pooled or rejected (using SDA-PAGE, RP- HPLC and SE-HPLC)

(6) pooling of the fractions

Therefore, in the main aspect the invention relates to a use of IMAC for the isolation and/or purification of G-CSF comprising:

- loading a solution or mixture, which comprises G-CSF, on the IMAC support

- selective binding of biologically active form of G-CSF to the IMAC support at high pH, optionally washing the IMAC column,

- eluting protein and non-protein contaminants from the IMAC column by lower pH buffer or by the addition of competing substances

- eluting the biologically active form of G-CSF from the IMAC column by using combined gradient - identification of fractions to be pooled or rejected

- pooling of the fractions In addition to the highly effective process of the present invention, the process can also be all along performed under native conditions. Thus renaturation steps can be omitted.

The term 'native conditions' used herein refers to the conditions by which the molecule (a protein, e.g. G-CSF) preserves the native conformation and the biological activity.

In one embodiment the use of IMAC according to the present invention is applied in the process for isolation and/or purification of G-CSF. In this case IMAC is applied as the first chromatographic step. After said first chromatographic step 95.5% RP-HPLC purity of G-CSF and over 99 % SE-HPLC purity of G-CSF with dimers and higher aggregates close to the detection limit or below it is achieved.

Accordingly, in a further purification process of a protein of interest only polishing and removal of dimers and other aggregates can be needed.

The process for isolation and/or purification of proteins of the present invention can additionally comprise further polishing. Depending on the nature of the protein of interest, further polishing of the protein according to the present invention can be accomplished by using a second chromatographic step. Said chromatographic step can be selected from the group consisting of an affinity chromatography, hydrophobic interaction chromatography, reversed phase, anion exchange chromatography, cation exchange chromatography or hydroxyapatite chromatography. In the case of isolation and/or purification of G-CSF, cation exchange chromatography can be preferentially applied as the second chromatographic step.

For final exchange of buffer and removal of protein dimers and other aggregates ultrafiltration or diafiltration can be used as the final isolation and/or purification step in the process of the present invention. Ultrafiltration or diafiltration can replace the gel filtration step which can be a bottle neck in the industrial preparation of proteins. Accordingly, in one aspect the present invention relates to a process for the two- chromatographic stage isolation and/or purification of a protein of interest, which process consists of the following chromatographic steps: (a) IMAC used as described above, and

(b) a chromatography

Accordingly, in one aspect the present invention relates to a process for the two- chromatographic stage isolation and/or purification of a protein of interest, which process consists of the following chromatographic steps:

(a) IMAC used as described above, and

(b) a chromatography, selected from the group consisting of an affinity chromatography, hydrophobic interaction chromatography, reversed phase, anion exchange chromatography, cation exchange chromatography or hydroxyapatite chromatography

Therefore, in one aspect the present invention relates to the use of IMAC as described above in the process for isolation and/or purification of a protein, wherein IMAC is used as the main chromatographic step.

Therefore, in one aspect the present invention relates to the use of IMAC as described above in the process for isolation and/or purification of a protein, wherein IMAC is used as the first chromatographic step.

Optinally, the two-chromatographic process of the present invention can further comprise:

(c) ultrafiltration of diafiltration

In another aspect, the process for isolation and/or purification of G-CSF of the present invention consists of:

(a) IMAC used as described above, and

(b) cation exchange chromatograpy, and (c) ultrafiltration or diafiltration

No additional steps (e.g. concentration, changing of buffers, filtrations, freezing, etc) are needed.

The process for purification and/or isolation of proteins e.g. G-CSF of the present invention can be transferred to industrial scale and to the production of biologically active proteins. In the case of G-CSF 98% RP-HPLC purity of G-CSF and over 99 % SE-HPLC purity of G-CSF with dinners and higher aggregates close to the detection limit or below it is achieved.

The process for purification and/or isolation of protein of the present invention results in the production of biologically active proteins which can be suitable for clinical use in medicine.

Exemplary, the process leads to the production of biologically active G-CSF suitable for clinical use in medicine.

The biologically active protein e.g. G-CSF obtained by the process for the purification and/or isolation of the present invention is suitable for the preparation of pharmaceutical composition, which comprises the therapeutically effective amount of biologically active protein e.g. G-CSF and is suitable for clinical use.

The term 'therapeutically effective amount' used herein refers to the amount of a biologically active protein e.g. G-CSF which has the therapeutic effect of biologically active protein e.g. G-CSF.

The possibility of maintaining the active form of a protein, e.g. G-CSF in a short purification and isolation process contributes not only to an improved yield, but also to an improved purity and effectiveness of the biologically active protein e.g. G-CSF and the pharmaceutical composition containing it. Suitable pharmaceutically acceptable auxiliary substances include suitable diluents, adjuvants and/or carriers useful in protein e.g. G-CSF therapy.

Exemplahly, biologically active G-CSF which was obtained by using the process of the present invention, particularly when performing the additional steps of cationic exchange chromatography can be used for preparation of medicaments, which are indicative for G-CSF.

The pharmaceutical composition containing the pure and biologically active G-CSF obtained by the process of the invention can thus be administered, in a manner known to those skilled in the art, to patients in a therapeutically amount which is effective to treat the above mentioned diseases.

The process for isolation and/or purification of proteins of the present invention is most preferably maintained under native conditions and is described in details in the following.

The process can begin with the sample preparation. The samples can be prepared by any one of the following processes: direct extraction of the whole cells expressing the target protein, homogenisation of the cells expressing the target protein either in the soluble form in the cytoplasm or in the form of inclusion bodies, denaturing solubilization of "classical" inclusion bodies (using chaotropes such as 6-8 M urea or 4-6 M guanidinium hydrochloride, high pH buffers, surfactants at high concentration, etc) and subsequent renaturation (performed by dilution, denaturant removal e. g. by dialysis, buffer exchange etc.). The samples can also be prepared by mild solubilization of "non-classical" inclusion bodies (containing a high percentage of correctly folded target protein or its precursor) under native conditions e.g. urea in non-denaturing concentrations (1-2 M, preferably in a buffer at a pH of below 10 and more preferably at a pH of about 8.0), N-lauroyl sarcosine in non-denaturing concentrations (0.05-0.25% (m/v)), low concentrations of Zwittergents, different non- detergent sulfobetaines (NDSB), betaine, sarcosine, carbamoyl sarcosine, taurine, dimethylsulfoxide (DMSO) and higher concentrations of buffers (preferably at a pH being maintained at below 10 and more preferably about 8.0), for example: HEPES, HEPPS, MES, ACES. Preferably, N-lauroyl sarcosine, NDSB and DMSO are used. Most preferably N-lauroyl sarcosine in the concentration range from 0.1 % to 0.25% (m/v) is used. In the case of cytoplasmatic expression, the soluble protein solution after solid/liquid phase separation of the cell homogenate can be directly used for chromatographic isolation/purification.

In one embodiment, the samples can be prepared in a buffer that can be directly used for the first chromatographic step IMAC.

In the case of G-CSF the mild extraction of 'non-classical inclusion bodies' with NDSB is preferably used. NDSB is non-denaturing for the protein and does not interfere with chromatographic matrix. Most preferably, NDSB in the concentration range from 0.1 % to 0.25% (m/v) is used.

The IMAC support comprises a solid phase material and a metal ion chelate bound to the solid phase material. Conventional solid phase materials can be suitably used, such as Sepharose, Fractogel and other gel support materials. The metal ion chelate bound to the IMAC support is suitably selected from metal ions preferably having two valencies, especially transition metal ions. Hg is less suitable in view of its toxicity and its tendency to be leached out of the IMAC column. Preferably, the metal ion chelates, being bound to the IMAC support, comprise: M(ll)-iminodiacetate (IDA), M(ll)-nithlotriacetic acid (NTA), M(ll)-carboxymethylaspartate etc., where M presents Zn, Cu, Co, Ni etc. Most preferably Zn(II) -IDA, Ni(II)-IDA and Ni(II)-NTA are used. The eluate resulting from the first elution from the IMAC column can also be used as loading solution or mixture for IMAC. The pH of the eluate should be adjusted, e.g. by addition of a NaOH solution or a high pH buffer solution, to the pH range from 6.5 to 9.0. The eluate can be then reloaded to the same IMAC support such as, e.g., Zn(II)- IDA, Ni(II)-IDA, Ni(II)-NTA or to a different IMAC support. The combination with other immobilised metal ions is also possible (e.g. Zn(II), Cu(II), Co(II), Ni(II) etc..) and enables better separation and removal of specific host proteins. Regardless to the preparation and the origin of the loading solution, the pH of the loading solution is in the range from 6.5 to 9.0. Preferred pH of loading solution is from 7.0 to 8.5, the mostly preferred pH is from 7.8 to 8.2.

Various buffers, which can maintain pH in the range from 6.5 to 9.0 can be used for loading and binding of G-CSF to the IMAC support and are selected from the group consisting of: phosphate, acetate, hydroxymethylaminomethan (Tris)/HCI, Tris/acetate, citrate and other buffers providing a pH of from 6.5 to 9.0. Preferably, Tris/HCI is used. Tris/HCI buffer is preferably used in the concentration range from 5 to 50 mM, most preferably in the range from 10 to 40 mM.

After binding to the support the process is continued by washing of the column and elution of proteins from the column. Elution can be performed by using either a discontinuous step gradient or linear gradient by descending pH or competitive elution at high pH (e.g. with imidazole, histidine, ammonium chloride and similar). The term 'linear gradient' used herein refers to washing of chromatographic column by a solution which composition is changed in a way that the proportion of one buffer (or one component of the buffer) is increased linearly, whereas the proportion of the other buffer (or another component of the buffer) is decreased linearly. The term 'discontinuous step gradient' used herein refers to washing of chromatographic column by a solution, which is composed of certain proportion of one buffer (or one component of the buffer) and certain proportion of another buffer (or another component of the buffer) for a determined time period. The proportions of both buffers are rapidly (suddenly) changed and the column is washed by another determined time period. The composition of the solution (change of buffer proportions or component proportions) is not changed linearly.

The term 'competitive elution' used herein refers to elution at the pH of the binding buffer where the competitor molecules, such as imidazole, histidine, ammonium chloride etc., in the elution buffer bind to the metal chelated matrix by themselves and thus displace the protein molecules. Preferably, discontinuous step gradient, resulting to elution at low pH is used. Namely, high pH could cause the activation of cysteine residues and the formation of dimers. Stability of G-CSF is also high at low pH.

Several elution buffers can be used for discontinuous step or linear washing and elution and are selected from the group consisting of: acetate, Tris/acetate, phosphate, citrate and other suitable buffers. The pH range for the elution can be from 3.0 to 5.0, preferably 3.5 to 4.5. By the discontinuous step gradient, the pH is rapidly changed from the order of the loading pH to the order of elution pH, such as from 7.0 to 4.0, and the isoelectric point is thus jumped over, avoiding the precipitation of the protein. Namely, environment with the pH of the protein isoelectric point can cause its precipitation.

The eluate obtained from the IMAC column can be loaded to different chromatographic columns, depending on the nature of protein, and are selected from the group consisting of affinity chromatography, hydrophobic chromatography, reversed phase chromatography, anion exchange chromatography, cation exchange chromatography and hydroxyapatite chromatography.

In one embodiment the eluate obtained from the IMAC column can be loaded directly to the cationic exchange chromatography column, without any additional intermediate steps being required. Various cation exchange chromatography supports can be used and may be selected from the group consisting of: SP Sepharose FF, SP

Sepharose HP, CM Sepharose FF, TSK gel SP-5PW, TSK gel SP-5PW-HR,

Toyopearl SP-650M, Toyopearl SP-650S, Toyopearl SP-650C, Toyopearl CM-650M, Toyopearl CM-650S, Macro-Prep High S support, Macro-Prep S support, Macro-Prep

CM support etc. Preferably, Toyopearl SP-650S or TSK gel SP-5PW are used. pH range of the loading solution for cationic exchange chromatography is in the range from 3.0 to 5.8, preferably in the range from 4.0 to 5.0.

The salt concentration in the loading solution for cationic exchange chromatography has to be low enough to enable the binding, which also depends on pH of the solution. Various buffers with the pH range from 3.0 to 5.8 can be used for loading and binding to the support for cationic exchange chromatography and may be selected from the group consisting of: acetate, citrate, Tris/HCI, Tris/acetate, phosphate, succinate, malonate, 2-(N-morfolinoethansulfonate) (MES) and other buffers. Preferably, acetate buffer is used.

Acetate buffer can be used in the concentration range from 10 to 60 mM, preferably in the concentration range from 10 to 30 mM.

In the cationic exchange chromatography, the column loading is followed by washing of the column and the elution of the proteins from the column. The elution occurs due to increased ionic strength after the addition of high concentration of salt in buffer solution. Discontinuous step gradient, linear gradient and a suitable combination of step and linear gradient can be used.

Elution buffers, which can be used for washing and elution, may be selected from the group consisting of: acetate, citrate, Tris/HCI, Tris/acetate, phosphate, succinate, malonate, MES and other suitable buffers with addition of salts such as NaCI or KCI. Ionic strength and salt concentration, by which the elution is achieved, depends on the pH of the buffer solution. The higher is pH of the buffer, the lower ionic strength is needed for the elution of the proteins from the column.

If desired, the eluate obtained from the IMAC or, preferably, after the consecutive cationic exchange chromatography column can be loaded directly to the ultrafiltration / diafiltration, without any additional intermediary steps being required.

In the eluate, monomeric, biologically active, correctly folded G-CSF is obtained with a RP-HPLC purity of G-CSF of greater than 98% and SE-HPLC purity of G-CSF of greater than 99% and biological activity of about 1x10⁸ IU/mg.

Further embodiments of the invention are listed below:

1. A use of IMAC for the isolation and/or purification of a protein, wherein said use comprises:

- a classical separation principle based on the specific protein affinity to the IMAC matrix with immobilized metal ions, and acting simultaneously - a further separation of the bound proteins by using a combination of pH/salt gradient or a combination of competing substance/salt gradient

2. A use of IMAC for the isolation and/or purification of a protein comprising: - loading a solution or mixture, which comprises a protein, on the IMAC support

- selective binding of biologically active form of a protein to the IMAC support at high pH, optionally washing the IMAC column,

- eluting protein and non-protein contaminants from the IMAC column by lower pH buffer or by the addition of competing substances - eluting the biologically active form of a protein of interest from the IMAC column by using combined gradient

- identification of fractions to be pooled or rejected

- pooling of the fractions

3. The use of IMAC according to embodiment 2 wherein the protein is G-CSF.

4. The use of IMAC according to embodiment 2 wherein the combined gradient is selected from the group consisting of pH/salt and competing substance/salt gradient.

5. The use of IMAC according to embodiments from 1 to 4 in the process for isolation and/or purification of a protein, wherein IMAC is used as the main chromatographic step.

6. The use of IMAC according to embodiments from 1 to 4 in the process for isolation and/or purification of a protein, wherein IMAC is used as the first chromatographic step.

7. The process for isolation and/or purification of a protein according to embodiment 6, wherein the process further comprises a chromatography.

8. The process according to embodiment 7 wherein the chromatography is selected from the group consisting of an affinity chromatography, hydrophobic interaction chromatography, reversed phase, anion exchange chromatography, cation exchange chromatography or hydroxyapatite chromatography

7. The process according to embodiment 8 wherein the chromatography is cation exchange chromatography.

8. The process according to embodiment 8 wherein the process further comprises ultrafiltration or diafiltration.

9. The process according to embodiments from 5 to 8 wherein the protein is G-CSF

10. A pharmaceutical composition obtained by the process according to any one of embodiments from 6 to 9.

Except where noted, all definitions and embodiments given before may apply also to the following embodiments and all definitions given below may also apply to any embodiment mentioned elsewhere in the specification.

In a further embodiment, the invention relates to a process for the isolation and/or purification of a protein of interest by use of immobilized metal affinity chromatography (IMAC), comprising the steps of a) applying a mixture or solution comprising said protein to an IMAC matrix, b) separating contaminant(s) by use of metal ion affinity of said protein to said IMAC matrix, under conditions suitable to keep said protein bound to the matrix, c) separating the protein of interest from further contaminant(s) by use of both a non-metal ion affinity interaction and a metal ion affinity interaction between said protein and said IMAC matrix.

In the context of this and the following embodiments, the term IMAC matrix is to be understood to relate to a support comprising chelated metal ions. Such supports comprise a metal chelating group, e.g. iminodiacetic acid, nitrilothacetic acid, or Tris- (carboxymethyl)ethylendiamine. Suitable metal ions have been described elsewhere in this specification. The terms IMAC and metal ion affinity are well known and understood by the person skilled in the art.

In step a), the mixture or solution comprising the protein of interest is applied to the IMAC matrix. The person skilled in the art is able to determine conditions under which the protein of interest will bind to the IMAC matrix (see also below). Preferably, conditions are chosen which allow for the binding of the protein of interest, but at the same time ensure that other contaminants are preferably not bound to the matrix. Thus, already step a) can act to remove many contaminants.

The mixture or solution is preferably applied at a pH of 6.5 to 9.0, more preferably at pH 7.5 to pH 8.5, e.g. approximately pH 8. The buffer may comprise any other ingredients considered appropriate. E.g., the buffer may comprise salt or no salt. If the buffer is salt-containing, a salt concentration corresponding corresponding to equal or more than 100 mM NaCI, preferably equal or more than 120 mM NaCI may be considered, e.g. a salt concentration corresponding to 120 mM to 1.5 M NaCI or a salt concentration corresponding to 100 to 200 mM NaCI or 120 to 180 mM NaCI.

According to step b) a first (set of) contaminant(s) is separated by use of metal ion affinity of said protein to said IMAC matrix, under conditions suitable to keep said protein bound to the matrix. The person skilled in the art is able to determine conditions under which the protein of interest is kept bound to the matrix. The person skilled in the art can test different buffers of differing pH and different ionic strength and analyze whether the protein of interest remains bound or is eluted by the buffer. As a starting point, the conditions of regular IMAC chromatography can be applied. Preferably, the pH should be equal or higher than pH 4.5, preferably equal or higher than pH 5.0. For example, one or more buffer(s) each having a pH of 6.0 to 9.0 and a salt concentration corresponding to 100 mM NaCI or higher may be used. Competing substances (e.g. imidazole, see below for further examples) may be added to the buffer (particularly if the pH of the buffer is rather at the higher end of the suggested range). Such competing substances may facilitate the removal of contaminants. However, if such competing substance is added, its amount should preferably be low enough to ensure that the protein of interest remains bound to the chromatography matrix. A starting point for such a suitable amount, one may choose a range of competing substance corresponding to 10 to 20 mM imidazole. The buffer may also contain a low amount of Tris (see also Examples), which can also be viewed as having properties of a competing substance. For further guidance concerning preferred buffers in step b), see the preferred embodiments as laid out further below.

Preferably, the separation according to step b) is carried out essentially or predominantly by use of metal ion affinity. The person skilled in the art is familiar with methods to ensure binding of the protein by metal ion affinity, particularly essentially or predominantly by metal ion affinity. For example pH neutral buffers with high ionic strength, such as buffers having a salt concentration corresponding to equal or more than 100 mM NaCI, preferably equal or more than 120 mM NaCI may be used, e.g. having a salt concentration corresponding to 120 mM to 1.5 M NaCI. For example, it is known that a salt concentration of 1 M NaCI avoids undesired ionic exchange affinity interactions between matrix and loaded proteins (see e.g. Lottspeich, F. and Engels (Eds.), Bioanalytik, 2nd edition 2006, page 231 , first full paragraph). Again, for further guidance concerning preferred buffers in step b), see the preferred embodiments as laid out further below. Usually, concentrations of 0.1 to 1 M are considered as useful to eliminate undesired ionic interactions.

In summary, step b) uses a classical IMAC separation principle to separate contaminant(s) from the protein of interest. Contaminants removed may include aggregated host cell proteins (e.g. host cell proteins remaining from fermentation in

Escherichia coli) as well as host cell oligonucleotides (such as RNA or DNA fragments). However, also aggregated and therefore undesired derivatives of the protein of interest may be removed in this step (such aggregates are being formed e.g. during expression of the protein in inclusion bodies and may not be fully solubilized and refolded). The term "corresponding to" as used in the context of salt concentrations is readily understood by the person skilled in the art. It is known by the person skilled in the art, that different salts can be used instead of e.g. NaCI or KCI, for example one or more salts from the group consisting of LiCI, LiBr, NaBr, KBr, Na2SO₄, and K₂SO₄. Therefore, the term "corresponding to" is preferably understood as relating to a salt concentration of a different salt which has the same effect concerning binding of the protein of interest as the corresponding concentration of the reference salt. More preferably, the term is understood as relating to a salt concentration having the same ionic strength as the reference salt concentration. The term "ionic strength" has a defined textbook meaning and is readily understood by the person skilled in the art.

In step c) the protein of interest is separated from further contaminant(s) by use of both a non-metal ion affinity interaction and a metal ion affinity interaction between said protein and said IMAC matrix. Optionally, the protein of interest is also eluted from the chromatography matrix during step c). Therefore step c) may also relate to eluting and separating the protein of interest from further contaminant(s) by use of both a non-metal ion affinity interaction and a metal ion affinity interaction between said protein and said IMAC matrix.

Unexpectedly, it has been found that a combination of metal ion affinity interaction and non-metal ion affinity interaction can serve to separate the protein of interest from further contaminants. Interestingly, it is particularly possible to separate impurities related to the protein of interest such as fragments, isoforms, dimers or other undesired derivatives having similar metal ion affinity as the protein of interest. Most notably, it has been possible to separate G-CSF monomers from disulfide linked dimers of G-CSF. Previously, it was typically necessary to remove such dimers by means of size exclusion chromatography, which is an expensive and economically unfavourable step. Thus, the process of the present invention provides a purified protein at low cost and in high yield. The advantage becomes particularly relevant in case of a industrial scale process designed to produce G-CSF in the scale of grams or more. Essentially, one can view the steps b) and c) as a switching of the interaction principle between first a metal ion affinity and then a metal ion affinity plus an ion exchange separation principle. Thus, it has been surprisingly been found that the same matrix can be used to carry out two separation steps, particularly with the protein remaining bound to the same column. Thus, by using a sequence of different conditions (particularly different buffers), it has become possible to achieve a purity previously only achievable by two independent chromatographies with different matrices and an intermediate elution step. Thus, the present invention provides a very economic and rapid purification principle with high yield, minimal handling and minimized loss of material.

A non-metal ion affinity interaction can be favored by a low salt concentration, e.g. no salt or a salt concentration corresponding to less than 100 mM NaCI, more preferably corresponding to less than 80 mM NaCI, more preferably corresponding to less than 50 mM NaCI, more preferably corresponding to less than 30 mM NaCI, most preferably corresponding to less than 10 mM NaCI. At the same time the pH of the buffer may be decreased in order to facilitate separation of the protein of interest from contaminants. Alternatively or additionally to decreasing the pH of the buffer, a competing substance may be added, preferably in increasing concentration. The increase of the salt concentration and the decrease of the pH (or increase of a competing substance) may be carried out in the form of step gradients, or linear or non-linear continuous gradients. Preferably the gradients are applied by mixing of appropriate buffers, preferably by mixing of only two buffers having different pH and different salt concentration (see also Examples). The person skilled in the art can determine when a relevant non-metal ion affinity interaction is achieved.

For example, at low pH, the histidine residues of a protein become protonated. Thus, depending on the degree of protonation, the protein of interest does not bind to the matrix predominantly by means of metal affinity interaction anymore, but by means of a non-metal affinity interaction. Similarly, if the amount of a competing substance is increased, the protein does not bind predominantly by means of metal affinity interaction anymore. In contrast, previous IMAC separation methods have attempted to minimize other interactions than metal affinity interactions (see e.g. Lottspeich, F. and Engels (Eds.), Bioanalytik, 2nd edition 2006, page 231 , first full paragraph).

Therefore, according to the present invention, the term "both a non-metal ion affinity interaction and a metal ion affinity interaction" relates to an interaction in which the non-metal ion affinity interaction is not minimized or wherein the non-metal affinity interaction is not negligible. For example, if the protein of interest remains bound to the IMAC matrix at low pH (e.g. at pH 4) or in the presence of a high concentration of competing substance (e.g. 100 mM imidazole) in the absence of salt or in the presence of only a low salt concentration (e.g. corresponding to 10 or 20 mM NaCI), one may safely assume that the interaction with the IMAC matrix is quite significantly mediated by means of a non-metal ion affinity.

Preferably, the chromatographic separation steps b) and c) according to the invention as well as any other chromatographic separation steps are carried out using suitable chromatographic columns as familiar to the person skilled in the art. Using suitable columns, particularly the application of buffers and buffer gradients are greatly facilitated.

The process according to the invention is particularly suited for proteins which are capable of remaining bound to the IMAC matrix by means of other interactions than metal affinity interactions. More particularly, the process according to the invention is suited for purifying a protein of interest which is capable of remaining bound to an IMAC matrix if a buffer consisting of 20 mM CH₃COOH/NaOH, having a pH of 4.0 is used on a chelating Sepharose fast flow matrix precharged with Zn(II) ions. Even more particularly, the process according to the invention is suited for purifying proteins which can be eluted by increasing the salt concentration if they are bound to an IMAC matrix under conditions of a buffer consisting of 20 mM CH₃COOH/NaOH, having a pH of 4.0 is used on a chelating Sepharose fast flow matrix precharged with Zn(II) ions. As such proteins can be eluted by increasing the salt concentration (e.g. to 50 to 100 mM NaCI), it has to be concluded that the main non-metal affinity interaction with the IMAC matrix is an ion exchange interaction (as opposed e.g. to a hydrophobic interaction).

As already mentioned, the process according to the invention is particularly suited for purifying G-CSF (granulocyte colony stimulating factor), particularly non-glycosylated G-CSF, most particularly non-glycosylated human met-G-CSF. All of these terms are well known and understood by the person skilled in the art. Human G-CSF is a glycosylated protein of 174 amino acids. Recombinant human G-CSF expressed in Escherichia coli consists of 175 amino acids (there is a an additional N-terminal methionine residue due to expression in E. coli) and is being marketed as a pharmaceutical under the international non-proprietary name (INN) Filgrastim. Due to the N-terminal methionine residue, the molecule is abbreviated as met-G-SCF.

In a further preferred embodiment, the process according to the invention comprises an additional chromatographic step, particularly chosen from the group consisting of anion exchange chromatography, cation exchange chromatography, and hydrophobic interaction chromatography. Such step may be used to further increase the purity of the protein obtained.

Advantageously, the IMAC purification according to the present invention can be used as a capture as well as a first purification step. The term "capture" is known to the person skilled in the art and preferably relates to the first chromatographic purification step in which bulk impurities are being removed. Such step may be preceded by other suitable steps (e.g. centhfugation) designed to remove major impurities such as host cell debris.

Further steps of purification and/or buffer exchange may be included as well, e.g. diafiltration and/or ultrafiltration.

As already mentioned, the present invention allows to highly purify the protein of interest without the need for size exclusion chromatography. Therefore, the invention relates also to a process according to the invention wherein no size exclusion chromatography is carried out. Such process is particularly advantageous if undesired derivatives, particularly fragments (particularly almost full-length fragments such fragments comprising at least 80%, more preferably at least 90% of the length of the protein of interest), isoforms, mutants, and dimers (particularly disulfide- bridged dimers) of the protein of interest are to be separated. The term "mutants" preferably relates to mistranslated protein of interest. For example, it is known that due to translational errors, protein mutants of G-CSF are generated during expression in E. coli. Particularly if such mutations affect the pi of the protein of interest or its histidine content, the corresponding mutants can be separated by the method according to the invention.

In a preferred embodiment, steps b) and c) are carried out separately, for example by subsequently applying suitable buffers.

In a further preferred embodiment, steps b) and c) are carried out on the same IMAC matrix. Carrying out the process on the same matrix allows to take full advantage of the protein of interest remaining bound to the matrix. Thus, there is no need for elution of the protein between each separation step b) and c). Thus, no additional handling is required and loss of material is minimized.

In a further preferred embodiment, step b) is carried out before step c). Advantageously, the bulk impurities can be removed by carrying out the step b) in the sense of a capture step. In step c), acting as an intermediate purification step, the protein of interest may be eluted. If step c) is carried out after step b), there is less need to fine-tune the separation conditions in step c), as the protein of interest can be conveniently collected in the eluate (e.g. by collecting the fractions eluting from a chromatographic column).

Preferably, the process comprises the additional step of formulating the purified protein into a pharmaceutical composition. Advantageously, the present invention allows to purify a protein of interest sufficiently in order to formulate it into a pharmaceutical composition. Many such protein pharmaceuticals are known to the person skilled in the art, e.g. erythropoetin, somatotropin, G-CSF, GM-CSF, interferon (e.g. interferon alpha, interferon beta, interferon gamma), interleukin-2, blood factors (e.g. Factor VIII, Factor IX), tissue plasminogen activator (tPA), insulin, antibodies or antibody fragments (e.g. trastuzumab, adalimumab, rituximab, bevacizumab, cetuximab, ranibizumab), and derivatives thereof. The step of formulating the protein into a pharmaceutical composition may also comprise a step of modifying the protein of interest. E.g. tags for purification (e.g. His-tags) and/or secretion of the protein (secrection signal sequences) may be removed. The protein may also be modified by adding or removing sugar chain residues (as may occur in glycosylated proteins), by adding polyalkylene glycol (PAG) residues (particularly polyethylene glycol (PEG) residues) or other water-soluble polymers (e.g. hydroxyethyl starch).

A suitable pharmaceutical composition comprises a therapeutically effective amount of the protein of interest. The term therapeutically effective amount is well understoof by the person skilled in the art. More particularly, the term refers to an amount which has the therapeutic effect of the protein of interest, preferably taking into account any possible undesired side-effects of amounts higher than necessary to achieve the desired effect. Such therapeutically effective amounts are well-known to the person skilled in the art, particularly for proteins of interest which are already being used as known pharmaceuticals (see e.g. the examples given above).

The pharmaceutical composition may comprise suitable pharmaceutically acceptable auxiliary substances such as diluents, stabilizers (e.g. amino acids), preservatives (e.g. phenol, cresol, benzyl alcohol) and detergents (e.g. polysorbate).

For many proteins, medical indications and uses are known. For example the indications include diabetes in the case of insulin, erythropenia in the case of erythropoetin, thrombosis in the case of tPA). In the case of G-CSF, the indications include neutropenia, mobilisation of hematopoietic progenitor cells, neutropenic and non-neutropenic infections, chronic inflammatory conditions, prevention of infection and infection-related complications in neutropenic and non-neutropenic patients.

In a preferred embodiment, step b) is carried out by or comprises applying one or more buffer(s) each having a pH higher than 5.5, preferably higher than 6.0, more preferably each having a pH of 6.0 to 9.0, most preferably each having a pH of 6.0 to

7.5. Preferably, the pH as measured in the eluate of the chromatography during step b) will be in the higher than pH 6.0, more preferably it will be in the range of 6.0 to

8.0, most preferably in the range of pH of 6.0 to 7.5. Such pH levels are still high enough to enable binding of the protein of interest to the IMAC matrix.

As already mentioned, step b) can be considered as a classical IMAC separation step. However, in a classical IMAC separation, one may typically conclude the step of separation of contaminants by a final elution step, e.g. by lowering the pH to less than pH 5.5, particularly less than pH 5.0, more particularly less than pH 4.5. Alternatively or additionally, such elution step may comprise the use of competing substances such as histidine, imidazole or ammonium salts. See also further below for appropriate conditions for elution of the protein.

However, such elution of the protein after step b) is not necessary according to the present invention. Instead, preferably conditions are chosen which ensure that the protein remains bound to the IMAC matrix. Therefore, the present invention also relates to a process according to the invention wherein no elution of the protein of interest is carried out between steps b) and c), particularly wherein no elution of the protein of interest is carried out after step b).

Preferably, the one or more buffer(s) according to step b) each have a salt concentration higher than corresponding to 100 mM NaCI, more preferably each have a salt concentration corresponding to 100 to 200 mM NaCI, most preferably each have a salt concentration corresponding to 120 to 180 mM NaCI. The salt serves to reduce undesired non metal affinity interactions with the IMAC matrix. Thus, the salt allows to take full advantage of the metal ion affinity separation principle.

The term "applying" in the context of the application of buffers relates to the buffer which actually will be brought into contact with the IMAC matrix. For example, different buffers may be prepared and applied only thereafter as a mixture to the IMAC matrix. In such case the definition of the buffer "applied" would relate to the said mixture of buffers being applied.

As already mentioned, step c) can be viewed as a step in which the separation principles of metal-ion affinity and non-metal ion affinity act simultaneously in order to remove further contaminants. One key difference to "classical" IMAC separation principles is the use of one or more buffer(s) containing no or only a small amount of salt. Such buffers have previously been avoided in the context of IMAC chromatography, as in such way usually a lower purity of the protein of interest is achieved: The protein of interest may be contaminated e.g. with host cell proteins that co-elute with the protein of interest although they have no specific affinity binding sites. Co-elution may happen as a consequence of non-specific interaction between matrix and those host cell proteins.

In contrast, the process according to the invention is advantageous in that such impurities which would ordinarily co-elute in an IMAC principle if a low salt containing elution buffer is used, can largely be removed in step b). This is particularly relevant if step b) is carried out before step c).

In a preferred embodiment, step c) is carried out by or comprises applying one or more buffer(s) each of which does not exclude ion exchange effects. As already mentioned, the buffer should be chosen in a manner which allows for a non-metal ion affinity interaction to occur between the protein of interest and the IMAC matrix. Particularly, said non-metal ion affinity interaction may be an ion exchange interaction. Such interaction can be favoured by choosing appropriate buffers which allow for a significant ion exchange interaction to occur between protein of interest and IMAC matrix. Based on the guidance given below, the person skilled in the art is able to choose suitable buffers or other conditions.

The term " not excluding ion exchange effects" will be understood by the person skilled in the art. More particularly, the term relates to conditions in which ion exchange effects are not negligible. Even more particularly, the term relates to conditions under which ion exchange effects contribute, particularly significantly contribute, to the separation properties of the IMAC matrix. Preferably, the determination of such contribution is performed with respect to the protein of interest, particularly with respect to G-CSF.

Preferably, step c) is carried out by or comprises applying one or more buffers each of which comprises less salt than corresponding to 100 mM NaCI, more preferably less than corresponding to 50 mM NaCI, more preferably less than corresponding to 20 mM NaCI, most preferably less than corresponding to 10 mM NaCI. The lower the salt concentration, the more favoured are non-metal ion affinity interactions with the protein of interest.

In a preferred embodiment, step c) comprises applying one or more buffer(s) which favour elution of the protein of interest. Elution of the protein of interest can be favoured by lowering the pH and/or by adding one or more competing substances (such as imidazole, see below for further examples for competing substances). Preferably, the composition of the buffer is chosen to just elute the protein of interest but to separate it as well as possible from further contaminants which are also bound to the chromatography matrix. Thus, separation of contaminants related to the protein of interest (e.g. disulfide-bridged dimers, but see above for further examples) can be carried out particularly well. Based on the guidance given below, the person skilled in the art is able to determine suitable buffer compositions.

Suitable conditions may e.g. be determined by applying a gradient of decreasing pH (e.g. from 7.0 to 4.0, preferably from 7.0 to 4.5) and/or a gradient of increasing concentration of a competing substance. Preferably such gradient is a continuous gradient. The person skilled in the art is able to collect fractions of the eluate and to determine which fractions of the eluate contain the protein of interest. The pH and the competing substance concentration of these fractions can be easily determined. However, the person skilled in the art may also choose any other conditions favouring elution of the protein of interest and being deemed appropriate.

In a preferred embodiment, step c) is carried out by or comprises applying one or more buffer(s) having a pH not higher than pH 5.0, preferably wherein said pH is not higher than pH 4.5.

In a preferred embodiment, step c) is carried out by or comprises applying one or more buffer(s) comprising a competing substance (such as imidazole, for further examples see below). Preferably, the concentration of the competing substance is sufficient to have an effect on separation. Preferably, the concentration of the competing substance should be sufficient to achieve an effect on separation at least comparable to the effect achieved by a pH of pH 5.0, more particularly to pH 4.5, most particularly to pH 4.0, if measured in comparison to a pH of 7.0. For example, a concentration of 100 mM imidazole may be considered as an appropriate starting point.

In preferred embodiment, step c) is carried out by or comprises applying one or more buffer(s) having a pH not higher than pH 5.0, preferably not higher than pH 4.5, and wherein said buffer(s) has/have a salt concentration of less than the salt concentration corresponding to a total of 100 mM of NaCI and/or KCI, preferably less than the salt concentration corresponding to a total of 80 mM of NaCI and/or KCI, more preferably of less than the salt concentration corresponding to a total of 50 mM of NaCI and/or KCI, more preferably less than the salt concentration corresponding to a total of 20 mM NaCI and/or KCI, most preferably less than the salt concentration corresponding a total of 10 mM NaCI and/or KCI, and/or wherein said buffer(s) comphse(s) a competing substance. In a preferred embodiment, the one or more competing substance is chosen from the group consisting of histidine, imidazole, ammonium salts (such as ammonium chloride or ammonium sulfate, 1 ,4-piperazinebis-(ethanesulfonic acid) (PIPES), N-(2- acetamido)iminodiacetic acid (ADA), N-(2acetamido)-2-aminoethanesulfonic acid (ACES), ethylenediamine tetraacetic acid (EDTA) and ethylene glycol tetraacetic acid (EGTA). In a further preferred embodiment, the one or more competing substance is chosen from the group consisting histidine, imidazole, ammonium salts, 1 ,4- piperazinebis-(ethanesulfonic acid) (PIPES), N-(2-acetamido)iminodiacetic acid (ADA), N-(2acetamido)-2-aminoethanesulfonic acid (ACES).

EDTA and EGTA elute not only proteins but also the chelated metal ions from the matrix. Therefore, EDTA and EGTA are less preferred if an undesired contaminant shall be separated by eluting the protein of interest while said contaminant shall remain bound to the IMAC matrix. However, EDTA and EGTA are useful for regenerating the matrix and for removing particularly strong bound proteins.

Preferably, said competing substance is chosen from the group consisting histidine, imidazole, and ammonium salts. Such competing substances have several advantages e.g. being reliable, easy to handle, and economically favoured.

In preferred embodiment, step c) is carried out by or comprises applying a gradient of decreasing pH (e.g. from 7.0 to 4.0, preferably from 7.0 to 4.5) and/or a gradient of an increasing concentration of a competing substance (e.g corresponding to a concentration of 0 to 100 mM imidazole or of 0 to 500 mM of imidazole).

In a preferred embodiment, a gradient of increasing salt concentration (e.g. from 0 to 100 mM NaCI) and simultaneously decreasing pH (e.g. from 7.0 to 4.0, preferably from 7.0 to 4.5) is applied. Preferably, the gradient said increasing salt concentration comprises the range of a salt concentration corresponding to 40 to 60 mM NaCI. Alternatively or additionally to the simultaneously decreasing pH, a gradient of an increasing concentration of competing substance may be used. Preferably such gradient is a continuous gradient. The person skilled in the art is able to collect fractions of the eluate and to determine which fractions of the eluate contain the protein of interest. The pH, the salt concentration, and (if applicable) the concentration of a competing substance for such fractions can be easily determined, as well as any other conditions giving causing elution of the protein of interest.

For determining suitable buffer compositions, preferably continuous gradients are applied as they allow to determine which buffer composition will yield the desired result. Furthermore, continuous gradients (particularly if sufficiently shallow), whill allow to collect fractions of the protein of interest with the highest purity and to discard fractions containing undesired amounts of contaminants. However, the use of one or more step gradient(s) may have the advantage that the process becomes more stable and reproducible. The person skilled in the art may choose the appropriate shape of a gradient according to such considerations as well as the particular application or separation problem encountered.

Since the fractions comprising the protein of interest can be easily determined and collected, it is possible to thus separate the protein from further contaminants. Such contaminants may either elute before or after the protein of interest or they may even remain bound to the IMAC matrix.

In a preferred embodiment, step c) is started by a step c1 ) of applying one or more buffers which do not favor the elution of the protein. For example, such buffer may have a pH sufficiently high to keep the protein bound to the IMAC matrix. For example the pH of such buffer(s) may be higher than pH 6.0, preferably higher than pH 6.5, more preferably higher than pH 6.8. More particularly, such buffer(s) may have a pH of 6.0 to 9.0, most preferably a pH of 6.0 to 7.5 , e.g. approximately pH 7.0. Furthermore such buffer will preferably not contain an amount of competing substance which would be sufficient to elute the protein. Again, as in the other buffers used in step c), such buffer according to step c1 ) should comprise less salt than corresponding to 100 mM NaCI, more preferably less than corresponding to 80 mM NaCI, more preferably less than corresponding to 50 mM NaCI, more preferably less than corresponding to 20 mM NaCI, most preferably less than corresponding to 10 mM NaCI. Starting step c) by means of such step c1 ) has the advantage of separating contaminants which elute at low salt concentration. Furthermore, such step c1 ) may be used to reduce the salt concentration in order to facilitate a rising salt gradient in step c).

Depending on the particular setup of step c), the protein of interest may be eluted already during step c) (see above) or it may be eluted separately.

In a further preferred embodiment, the invention relates to a process for the isolation and/or purification of G-CSF by use of immobilized metal affinity chromatography (IMAC), comprising the steps of a) applying a mixture or solution comprising said G-CSF to an IMAC matrix, b) separating contaminant(s) by use of metal ion affinity of said G-CSF to said IMAC matrix, under conditions suitable to keep said G-CSF bound to the matrix, the step comprising applying one or more buffer(s) having a pH higher than pH 5.5 and each having a salt concentration equal or higher than corresponding to 100 mM NaCI, c) separating further contaminant(s) by use of a non-metal ion affinity interaction between said G-CSF and said IMAC matrix, the step comprising applying one or more buffer(s) each having a salt concentration of less than a total of 80 mM NaCI and/or KCI, preferably each having a pH not higher than pH 5.0 and/or comprising a competing substance.

Of course, all preferred features and particularly buffer compositions as laid out elsewhere in this specification can also be applied to the above embodiment.

In a further preferred embodiment, the invention relates to a process for the isolation and/or purification of G-CSF by use of immobilized metal affinity chromatography (IMAC), comprising the steps of a) applying a mixture or solution comprising said G-CSF to an IMAC matrix, b) separating contaminant(s) under conditions suitable to keep said G-CSF bound to the matrix, the step comprising applying one or more buffer(s) having a pH higher than pH 5.5 and each having a salt concentration equal or higher than corresponding to 100 mM NaCI. c) separating further contaminant(s), the step comprising applying one or more buffer(s) each having a salt concentration of less than a total of 80 mM NaCI and/or KCI, preferably each having a pH not higher than pH 5.0 and/or comprising a competing substance.

Of course, all preferred features and particularly buffer compositions as laid out elsewhere in this specification can also be applied to the above embodiment.

Figures

Fig. 1 shows the chromatography profile according to Example 1. The straight lines indicate the buffer addition scheme (see also the indicated buffers P2 to P5 below the chromatogram). The diagram furthermore shows absorbance profile at 280 nm, with the major peaks by the boxed letters A, B, and C, as mentioned in Example 1.

The following examples are given for the purpose of illustrating various embodiments of the invention and are not meant to limit the present invention in any way.

Examples

Example 1 : IMAC (Chelating Sepharose FF, Pharmacia)

Inclusion bodies from G-CSF expressing E. coli are solubilized under native conditions (either by 0.2% NDSB (NDSB 195, NDSB 201 , NDSB 211 , NDSB 256) or O.2% N-lauroyl sarcosine (NLS) in P1. In the case of NLS solubilization, the detergent must be removed (for example using an ion exchange resin) prior to IMAC step while NDSBs are compatible with IMAC and do not interfere with separation, therefore, no removal step is needed. The obtained protein solution (approximately 150 mg) with an approximate concentration of 1.8 mg/ml is applied to the Amersham chromatographic XK 26/20 column loaded with Chelating Sepharose fast flow (h=10 cm, d=2.6 cm, V=50 ml) which was pre-charged with Zn(II) ions. After loading, the column is washed with P2 buffer for 20 minutes (2 Vc) followed by a 15 minute-wash (1.5 Vc) with a mixture of buffers P2 and P3 at the volume ratio of 65:35. Although the theoretical pH value of this P2:P3 (65:35) mixture is much lower, in 1.5 column volumes applied, the experimentally determined pH value (pH of the eluate) reached only approximately pH 6.8, which was still high enough to enable binding of the target protein. Afterwards, P4 buffer is pumped through the column for 20 minutes (2Vc), followed by a shallow linear gradient of P5 buffer (from 100% P4 to 40% P4 and 60% P5 in 47 minutes (4.7 Vc)). The protein of interest (here G-CSF) was eluted at a pH below 6.0. The pH values of consecutive fractions within this main chromatographic peak were between 6 and 4.7. Final elution is accomplished by 100% P5 in 25 min (2.5 Vc). During this final elution, also disulfide-linkes dimer of G-CSF elutes (peak C). The pH of this fraction has not been experimentally determined but it is presumably close to 4.0. The whole chromatographic procedure is performed at a constant flow rate of 5 ml/min (56 cm/h). The gradient used is

The typical chromatogram is composed of three major peaks (see Fig. 1 ):

Peak A in the flow-through represents mostly aggregated G-CSF and host cell proteins, which do not bind to IMAC matrix. Peak B eluted during the linear gradient belongs to the monomeric biologically active G-CSF. Careful pooling of the fractions on the basis of RP-HPLC, SE-HPLC and SDS-PAGE (under reducing and non-reducing conditions) analyses results to improved purity of the IMAC-pool (Table 1 ), which is directly used in the next, polishing chromatographic step. For further purification, the central part of peak B is pooled, based on the results of RP-HPLC, SE-HPLC and SDS-PAGE analyses.

Peak C eluted at 100% P5 is more than 90% pure disulfide-linked dimer of G-CSF. based on the Gradient No. 4 from the Table 1.

Table 1 : Step by step improvement of combined pH & salt gradient and achieved purity of the IMAC pool.

Composition of buffers:

P1 : 4OmM Tris/HCL, pH 8.0

P2: 20 mM Tris/HCI, 150 mM NaCI, pH 8.0

P3: 20 mM CH₃COOH/NaOH, 150 mM NaCI, pH 4.0

P4: 20 mM Tris/HCI, pH 7.0

P5: 20 mM CH₃COOH/NaOH, 100 mM NaCI, pH 4.0

Example 2: Process for the purification and/or isolation of G-CSF, including additional purification steps

IMAC pool (the central part of peak B) is diluted with the same volume of P6 buffer (20 mM CH₃COOH, pH 4,0) and directly applied to the chromatographic column

HR10/10 (Amersham), loaded with chromatographic support Toyopearl 650S (35 μm; TosoHaas) to the height of 10 cm (h=10 cm, d=1 ,0 cm, V= 8 ml). Loading of the sample and elution from the column are performed at a constant flow of 2 ml/min. After loading the sample, the column is washed with P7 buffer for 6 min (1.5 Vc). Elution is accomplished by a linear gradient of P8 buffer from 0% to 18% Pδbuffer in 26 min (6.5 Vc) ( from 100% to 82% P7 buffer). Afterwards, the column is washed with the mixture of P7 and P8 buffers in a volume ratio of 82:18 for 12 min (3 Vc) and then for 8 min (2Vc) with 100% P8 buffer. Fractions of the main chromatographic peak are analysed by SDS-PAGE, RP-HPLC and SE-HPLC and pooled on the basis of these analyses. The buffer is exchanged by ultrafiltation/diafiltration. The final, sterile-filtered G-CSF has over 99% SE-HPLC purity, over 98% RP-HPLC purity, and biological activity of approximately 1 x 10⁸ IU/mg, which corresponds to the biological activity of the standard. Composition of buffers: P6: 20 mM CH₃COOH, pH 4,0 P7: 20 mM CH₃COOH, pH 5,5 P8: 20 mM CH₃COOH, 500 mM NaCI, pH 5,5

Example 3: In vitro G-CSF biological activity assay Biological activity of G-CSF is determined by the method based on stimulation of cellular proliferation (NFS-60 cells) using the known method (Hammerling, U. et al. in J Pharm Biomed Anal 13, 9-20 (1995)) and the use of international standard Human recombinant G-CSF (88/502, yeast cell derived; NIBSC Potters Bar, Hertfordshire, UK; see Mire-Sluis,A.R. et al. v J Immunol Methods 179, 117-126 (1995)

Claims

1. A process for the isolation and/or purification of a protein of interest by use of immobilized metal affinity chromatography (IMAC), comprising the steps of a) applying a mixture or solution comprising said protein to an IMAC matrix, b) separating contaminant(s) by use of metal ion affinity of said protein to said IMAC matrix, under conditions suitable to keep said protein bound to the matrix, c) separating the protein of interest from further contaminant(s) by use of both a non-metal ion affinity interaction and a metal ion affinity interaction between said protein and said IMAC matrix.

2. The process according to claim 1 , wherein the protein of interest is capable of remaining bound to said IMAC matrix if a buffer consisting of 20 mM

CH₃COOH/NaOH, having a pH of 4.0 is used on a Chelating Sepharose Fast Flow matrix precharged with Zn(II) ions.

3. The process according to any of claims 1 to 2, wherein the protein of interest is G-CSF, particularly non-glycosylated human met-G-CSF.

4. The process according to any of claims 1 to 3, comprising an additional chromatographic step, particularly chosen from the group consisting of anion exchange chromatography, cation exchange chromatography, and hydrophobic interaction chromatography.

5. The process according to any of claims 1 to 4, wherein no size exclusion chromatography is carried out.

6. The process according to any of claims 1 to 5, wherein steps b) and c) are carried out on the same IMAC matrix.

7. The process according to any of claims 1 to 6, wherein step b) is carried out before step c).

8. The process according to any of claims 1 to 7, wherein the process comprises the additional step of formulating the purified protein into a pharmaceutical composition.

9. The process according to any of claims 1 to 8, wherein step b) comprises applying one or more buffer(s) each having a pH higher than pH 5.5, preferably each having a pH of 6.0 to 9.0, preferably each having a pH of 6.0 to 7.5.

10. The process according to any of claim 9, wherein said one or more buffer(s) each have salt concentration equal or higher than corresponding to 100 mM NaCI, preferably a salt concentration corresponding to 100 mM to 1.5 M NaCI.

11. The process according to any of claims 1 to 10, wherein step c) comprises applying one or more buffer(s) each of which does not exclude ion exchange effects.

12. The process according to any of claims 1 to 11, wherein step c) comprises applying one or more buffer(s) each having a salt concentration of less than the salt concentration corresponding to a total of 100 mM of NaCI and/or KCI.

13. The process according to any of claims 1 to 12, wherein step c) is comprises applying one or more buffer(s) each having a salt concentration of less than the salt concentration corresponding to a total of 80 mM of NaCI and/or KCI, preferably of less than the salt concentration corresponding to a total of 50 mM of

NaCI and/or KCI.

14. The process according to claim 14, wherein said salt concentration is less than a total of 20 niM NaCI and/or KCI, preferably less than a total of 10 mM NaCI and/or KCI.

5 15. The process according to any of claims 1 to 14, wherein step c) is comprises applying one or more buffer(s) which have a pH not higher than pH 5.0 and/or which comprise a competing substance.

16. The process according to claim 15, wherein said pH is not higher than pH 4.5.

10

^^ Jtf. lhe process according to any of claims 15 to 16, wherein said competing substance is chosen from the group consisting of histidine, imidazole, ammonium salts, 1 ,4-piperazinebis-(ethanesulfonic acid) (PIPES), N-(2- acetamido)iminodiacetic acid (ADA), N-(2acetamido)-2-aminoethanesulfonic acid

15 (ACES), ethylenediamine tetraacetic acid (EDTA) and ethylene glycol tetraacetic acid (EGTA), preferably chosen from the group consisting histidine, imidazole, ammonium salts, 1 ,4-piperazinebis-(ethanesulfonic acid) (PIPES), N-(2- acetamido)iminodiacetic acid (ADA), N-(2acetamido)-2-aminoethanesulfonic acid (ACES).

20 ^

^iS >9. The process according to any of claims 15 to Aβ, wherein said competing substance is chosen from the group consisting histidine, imidazole, and ammonium salts.

25 β8. A process for the isolation and/or purification of G-CSF by use of immobilized metal affinity chromatography (IMAC), comprising the steps of a) applying a mixture or solution comprising said G-CSF to an IMAC matrix, b) separating contaminant(s) by use of metal ion affinity of said G-CSF to said IMAC matrix, under conditions suitable to keep said G-CSF bound to the

30 matrix, the step comprising applying one or more buffer(s) having a pH higher than pH 5.5 and each having a salt concentration equal or higher than corresponding to 100 mM NaCI, c) separating further contaminant(s) by use of a non-metal ion affinity interaction between said G-CSF and said IMAC matrix, the step comprising applying one or more buffer(s) each having a salt concentration of less than a total of 80 mM NaCI and/or KCI, preferably each having a pH not higher than pH 5.0 and/or comprising a competing substance.

^2f. A process for the isolation and/or purification of G-CSF by use of immobilized metal affinity chromatography (IMAC)₁ comprising the steps of a) applying a mixture or solution comprising said G-CSF to an IMAC matrix, b) separating contaminant(s) under conditions suitable to keep said G-CSF bound to the matrix, the step comprising applying one or more buffer(s) having a pH higher than pH 5.5 and each having a salt concentration equal or higher than corresponding to 100 mM NaCI. c) separating further contaminant(s), the step comprising applying one or more buffer(s) each having a salt concentration of less than a total of 80 mM NaCI and/or KCI₁ preferably each having a pH not higher than pH 5.0 and/or comprising a competing substance.