Cation chromatography using predicted elution buffer salt concentration

Abstract

To provide a chromatography method of producing a target elution volume comprising first and second target proteins.SOLUTION: A method comprises: providing a cation exchange chromatography column; applying a protein solution on the column, the protein solution comprising a first target protein, a second target protein and optionally one or more further proteins; inputting an optimization criterion; computing chromatography simulations for computing an elution buffer salt concentration adapted to provide a target elution volume matching the optimization criterion best; computing the pooling borders of the target elution volume as a function of at least the computed salt concentration and the input optimization criterion; applying an elution buffer having the computed salt concentration on the chromatography column; performing the elution; and collecting the computed target elution volume.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] FIELD OF THEINVENTION The present invention relates to the field of chromatography, and more particularly to the field of cation chromatography for obtaining specific proteins. [Background technology]

[0002] Background and Related Technologies Cation exchange chromatography is a form of ion exchange chromatography commonly used to separate molecules based on their net surface charge. Negatively charged ion exchange resins that have an affinity for molecules with a net positive surface charge (e.g., proteins or peptides) are used for both preparative and analytical purposes.

[0003] The net surface charge of a protein varies with pH in a manner determined by the protein's isoelectric point ("pI"). Proteins with different pI values ​​have different degrees of charge at a given pH and therefore different affinities for the negatively charged surface groups on the particles of the ion exchange medium in the chromatography column. Thus, proteins with different pI values ​​bind with different strengths to the chromatography resin, facilitating their separation.

[0004] However, not all proteins and peptides that need to be separated have sufficiently different pI values ​​to allow a clear separation on a chromatographic column. In addition, the separation may further depend on the number of residues on the amino acid side that interact with the column: due to steric factors, not all charged groups of the protein can participate in binding. In particular, the degree of separation of proteins containing protein glycoforms and other types of post-translational modifications (PTMs) on a chromatographic column is a challenge.

[0005] Glycosylation is one of the most common post-translational modifications (PTMs) of proteins. Glycosylation involves the covalent attachment of oligosaccharides (glycans) to the amino acid backbone of proteins, specifically to serine / threonine (O-glycosylation) or asparagine (N-glycosylation) amino acid residues. Glycan residues have a major impact on the biological function of proteins, as they can affect protein stability, solubility, antigenicity, folding and serum half-life. Protein glycosylation and many other forms of PTM (e.g., phosphorylation, methylation, succinimidation, oxidation, N-terminal methionine loss, etc.) are created in a complex process involving various enzymes and substrates, and are dependent on the host organism, the production cell line, and the culture conditions.

[0006] Glycosylation and other types of PTMs are important quality attributes of therapeutic proteins. PTMs through conformational changes can modulate the therapeutic value (potency) of proteins and peptides, e.g., complement-dependent cytotoxicity (CDCC) and antibody-dependent cellular cytotoxicity (ADCC) activities of MAbs.

[0007] Thus, the problems of current chromatographic techniques used to separate protein variants that have undergone glycosylation and / or other forms of PTM are serious when it comes to selectively eluting proteins bearing specific PTMs or when it comes to obtaining an elution containing a desired ratio of two or more protein variants that differ from each other only with respect to the type, number and / or location of the PTMs. Summary of the Invention

[0008] overview The object of the present invention is to provide a method for obtaining a target elution volume containing two or more target proteins using ion exchange chromatography according to the independent claims and a corresponding chromatography control system. Embodiments of the invention are set out in the dependent claims. The embodiments of the invention may be freely combined with one another if they are not mutually exclusive.

[0009] In one aspect, the invention relates to a chromatography method for generating a target elution volume comprising a first and a second target protein, the method comprising: - providing a cation exchange chromatography column; - applying a protein solution to the column, the protein solution comprising at least a first target protein, a second target protein, and optionally one or more further proteins; - entering optimization criteria into the chromatography simulation software (the optimization criteria are desired properties of the target elution volume with respect to a first and a second target protein contained in the target elution volume, for example the optimization criteria can be quantitative and / or qualitative properties of two or more target proteins contained in the elution volume. The quantitative property can be, for example, the absolute amount (e.g. concentration) of the target protein in the elution volume or the relative amount (e.g. ratio) of the two or more target proteins, additionally or alternatively the criteria can be or include one or more qualitative properties, for example the purity of one or both of the target proteins, the optimization criteria can be a derivative score value calculated as a function of one or more qualitative properties and one or more quantitative properties of the two target proteins).

[0010] The chromatography simulation software is configured to calculate salt concentrations of elution buffers adapted to elute the first and second target proteins from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained, the calculation including calculating a plurality of chromatography simulations as a function of salt concentrations of a plurality of different elution buffers.

[0011] The chromatography simulation software is further configured to calculate pooling boundaries of target elution volumes as a function of at least the calculated salt concentrations and the inputted optimization criteria.

[0012] The method comprises: - applying an elution buffer having a calculated salt concentration to a chromatography column; - carrying out elution with an applied elution buffer; - collecting the calculated target elution volume as a separate fraction using the calculated pooling boundaries.

[0013] These features can be advantageous because even though two or more proteins of interest may have highly similar pI values ​​and may have overlapping elution profiles under some elution conditions, embodiments of the invention allow for obtaining a target elution volume containing these proteins in a predetermined amount, ratio and / or purity.

[0014] Without intending to be bound by any theory, the applicant assumes that this result is achieved because the salt concentration of the elution buffer is not a one-size-fits-all elution salt concentration. Rather, the salt concentration is a concentration that is specifically computationally predicted for the desired optimization criteria, in particular the absolute or relative amount and / or purity of two or more target proteins of interest. This may allow to avoid additional process steps to remove or reduce the amount of a specific target protein in the eluate, and / or to avoid additional process steps to concentrate a specific target protein of interest in order to recombine two or more separately eluted target proteins in a desired ratio.

[0015] Applicants have observed that in some pharmaceutical application scenarios, it is desirable to form a drug consisting of a mixture of two or more target proteins provided in a predetermined ratio and / or purity. For example, a drug may consist of a mixture of two PTM variants of a particular protein P, designated P1 and P2, such that the drug is most effective when variants P1 and P2 are provided in a particular ratio of amounts. Embodiments of the present invention may make it possible to perform a chromatographic step that may be performed anyway to extract and / or concentrate a particular protein of interest from the total protein of a cell, such that two or more proteins of interest are easily eluted in a desired ratio, amount and / or purity in one chromatographic run.

[0016] Chromatography simulation software can achieve this desired result by simulating multiple chromatographic runs of a protein applied to a column for various (assumed / candidate) elution buffer salt concentrations to identify the optimal salt concentration for an optimization criterion. This computationally identified salt concentration is then used to create or select the elution buffer used in the actual chromatography process, so that the salt concentration of the created or selected elution buffer is identical to the calculated salt concentration.

[0017] The elution buffers obtained according to embodiments of the present invention are not "standard" / "one size fits all" elution buffers, but rather elution buffers specifically adapted to optimization criteria which may be optionally flexibly defined.

[0018] According to an embodiment, the optimization criterion is: a) the desired ratio or range of ratios of the amounts of the first and second target proteins in the target elution volume (particularly, the ratio can be any ratio within the range of 5:95 to 1:1, particularly 1:95 to 1:1); b) a desired amount or range of amounts of a first target protein combined with a desired amount or range of amounts of a second target protein in the target elution volume (e.g., the amounts can be absolute amounts or concentrations); c) ■ a desired ratio of the amounts of the first and second proteins in the target elution volume; ■ the desired amount or range of amounts of the second target protein in the target elution volume; ■ the desired purity or purity range of the second target protein in the target elution volume; a desired purity or purity range of the first target protein in combination with; d) Combination of two or more of the above criteria. For example, a composite score calculation algorithm can be used to calculate a score value for an elution buffer having a match M1 for the desired target protein ratio and a match M2 for the desired purity of two or more target proteins, where M1 has about 70% impact on the total score and M2 has about 30% impact on the total score. According to another example, the optimization criteria is a combination of the required minimum absolute amount of one or more target proteins, the desired ratio range (e.g., 1:2.0 to 1:2.5), and the minimum purity score of one or more target proteins. The compound is selected from the group consisting of:

[0019] This allows the user to identify and use elution buffers with optimized salt concentrations for the respective use case scenario. For example, in a research environment, the protein purity aspect may be less important than in a clinical use case scenario. According to a preferred embodiment, the simulation software is configured to generate a GUI that allows the user to input and / or modify optimization criteria, e.g., for an individual chromatography run. This allows the chromatography process to be flexibly adapted to different use case scenarios, different target proteins and combinations of target proteins, as well as many other factors.

[0020] According to an embodiment, the applied protein solution comprises one or more additional proteins. The additional proteins may have elution profiles that overlap with the target protein and may contaminate the elution solution. By considering the purity of the target protein, in particular the amount and / or chemical properties of the target protein in the simulation and / or pooling boundary calculation, it is possible to identify and use the salt concentration of the elution buffer that allows the target protein and the contaminating additional proteins to elute, so that an acceptable level of purity of the target protein is achieved.

[0021] According to embodiments, the salt concentrations of the multiple different elution buffers are automatically created, for example, based on a predetermined program routine for creating salt buffer specifications having different salt concentrations but otherwise identical (e.g., with respect to pH value). These automatically created buffer salt concentrations may be referred to as "hypothetical" or "candidate" elution buffer salt concentrations. For example, the calculated salt concentration is one of multiple candidate elution salt concentrations computationally identified to best meet an optimization criterion. For example, the calculation of the elution buffer salt concentration may include predicting the amount and / or purity of one or more target proteins in a target elution volume as a function of multiple different candidate elution salt concentrations, and identifying as the elution buffer salt concentration one of the candidate elution buffer salt concentrations that is most suitable for collecting a target elution volume containing one or more target proteins in a desired amount and / or purity. This identified salt concentration is used as the "calculated" elution buffer salt concentration to select or create an elution buffer whose salt concentration is identical to the calculated and identified elution buffer salt concentration. This selected or created elution buffer is used to actually carry out the elution with the applied elution buffer.

[0022] The desired optimization criteria input into the chromatography simulation software can be specified as absolute amounts (e.g., concentration of each target protein) or relative amounts (e.g., concentration or mass ratio of two or more target proteins), depending on the embodiment.

[0023] According to an embodiment, the method is used for analytical purposes. According to a preferred embodiment, the method is used for preparative purposes.

[0024] According to a preferred embodiment, the calculated salt concentration adapted to elute the first target protein is one of a plurality of candidate salt concentrations that are most suitable for separating the first target protein and / or the second target protein from any one of the further proteins contained in the applied protein solution.

[0025] These features can be advantageous, since the applicant has observed that it is possible to simulate a chromatographic process sufficiently well based on a set of known or easily obtainable input data values ​​to predict the concentration of salt in an elution buffer suitable for eluting two or more proteins of interest ("target proteins"), such that these target proteins are contained in a target elution volume in a desired amount, ratio and / or purity. Preferably, the calculated salt concentration is best or approximately best suited for separating a first target protein from further proteins contained in the protein solution.

[0026] By eluting the applied protein solution with an elution buffer whose salt concentration corresponds to the predicted suitable or optimal salt concentration, it is possible to perform high-resolution protein chromatography that can separate proteins and protein variants with highly similar chemical properties, including highly similar polarity. Furthermore, the calculated salt concentration of the elution buffer itself can be used as an input for performing further predictions, in particular for predicting with great accuracy the pooling boundaries of the elution volume that will contain one or more target proteins in a desired amount. The salt concentration of the elution buffer predicted to be the suitable or optimal salt concentration can be calculated as one of a number of candidate elution salt concentrations predicted to provide two or more target proteins in a desired absolute or relative amount and / or a desired purity. Thus, based on the prediction of the pooling boundaries as a function of the salt concentration of the computationally identified "optimal" elution buffer, it is possible to specifically collect fractions of the total elution volume that contain one or more target proteins in a desired amount, ratio and / or purity.

[0027] This can be particularly useful when one or more target proteins need to be separated from one or more other highly similar proteins, such as other glycoforms, or when obtaining a mixture of two or more highly similar target proteins having a desired ratio (relative amount).

[0028] The method according to an embodiment of the invention may have the advantage that highly similar protein variants that could not be separated by cation chromatography before can now be separated in one chromatographic run: in a first step, the salt concentration of a suitable elution buffer is predicted, and then elution is carried out using an elution buffer with the predicted salt concentration. This may be highly advantageous in any situation where one protein needs to be separated from other chemically highly similar proteins, or where two or more proteins with highly similar chemical properties need to be obtained in specific relative amounts. For example, the glycosylation pattern of recombinant pharmaceutical proteins often has a strong influence on the quality and efficacy of those proteins. Often, a cell line used to produce a particular target protein, for example an antibody or antibody fragment with a particular glycosylation pattern, not only produces this particular glycoform, but also one or more other glycoforms, including those lacking glycosylation. By first predicting the salt concentration of an appropriate elution buffer, performing elution with an elution buffer having the predicted salt concentration, and then predicting and collecting pooling boundaries, embodiments of the present invention may be able to specify one or more glycoforms of interest and have these glycoforms already available at the appropriate concentrations.

[0029] Predicting the pooling boundaries of a target elution volume such that this target elution volume contains two or more target proteins in a desired amount, ratio and / or purity can be beneficial since additional steps to increase or decrease the concentration of the target protein(s) can be avoided. This accelerates the process of obtaining one or more target proteins at a desired concentration and also increases the quality of the target proteins since all additional processing steps can be avoided, reducing the risk of contaminating or otherwise damaging the target protein(s) contained in the target elution volume.

[0030] Embodiments of the invention are particularly suitable for separating therapeutic or diagnostic proteins or protein fragments, such as antibodies or antibody fragments.Embodiments of the invention may be used to ensure the quality, purity, safety and efficacy of biopharmaceuticals by using chromatographic methods according to embodiments of the invention to selectively obtain one or more target proteins in desired amounts.

[0031] According to an embodiment, the protein solution applied to the column is an aqueous solution comprising one or more salts and one or more proteins.

[0032] According to an embodiment, a plurality of chromatographic simulations are calculated as a function of the salt concentrations of a plurality of different elution buffers and as a function of the pH values ​​of a plurality of different elution buffers. The calculation of the salt concentrations includes calculating combinations of elution buffer salt concentrations and elution buffer pH values ​​that are combined and adapted to elute the first and second target proteins from the chromatographic column so as to obtain a target elution volume that best meets the optimization criteria. The chromatographic simulations are performed to identify combinations of elution buffer salt concentrations and elution buffer pH values. The pooling boundaries of the target elution volumes are calculated as a function of at least the calculated combinations of elution buffer salt concentrations and elution buffer pH values ​​and the input optimization criteria. The elution buffer applied to the column has a calculated salt concentration and a calculated pH value.

[0033] Applicant has observed that in addition to salt concentration, pH value also has a significant effect on the elution profile of various proteins. Like salt concentration, pH value can be easily set to a specific desired value. One or more models used to perform the simulation typically take into account both salt concentration and pH value of the buffer, and calculations of multiple chromatography runs and / or chromatographic elution steps based on combinations of different salt concentrations and different pH values ​​can be performed without the need to fit models or implement complex additional program routines. Furthermore, the complex interrelationships between salt concentration and pH value, and their combined effects on the elution profile can be considered in the simulation process. Moreover, simulations based on variations in both salt concentration and pH value can cover a vast data space, thus increasing the likelihood of identifying a solution buffer that can follow the optimization criteria. For example, if a simulation is performed for 10 different salt concentrations, and each salt concentration is simulated based on 10 different pH values, the simulation will include 10 x 10 = 100 different simulations and corresponding theoretical / candidate elution buffer specifications.

[0034] According to an embodiment, the method includes automatically or manually selecting or preparing an elution buffer, such that the selected or prepared elution buffer has the salt concentration and pH value identified in the simulation as best meeting the optimization criteria. The selected or prepared elution buffer is used as the buffer applied to the column to elute the protein.

[0035] Single target protein case scenario According to a further use case scenario described herein, the target elution volume contains only one (first) target protein and, optionally, one or more further (contaminating) proteins. The optimization criterion is the desired absolute amount of the desired target protein and / or the desired purity of the target protein. In particular, the desired amount of the first target protein can be specified in the form of a desired concentration of the first protein in the collected volume of elution buffer or its derivative value.

[0036] For example, if a desired concentration is provided, the target elution volume can be calculated as the elution volume containing the target protein at the desired concentration, and the pooling boundary is selected such that the average concentration of the eluted target protein in the target elution volume is the same as the desired concentration. During collection of the target elution volume, the first target protein concentration in the elution buffer exiting the column can be higher or lower than the desired concentration.

[0037] When a desired absolute amount is provided, the target elution volume can be calculated as the minimum elution volume that contains the first target protein in the desired amount.

[0038] Using embodiments of the present invention to calculate pooling boundaries of target elution volumes of single target proteins can be advantageous as it allows to collect the target protein such that the collected target elution volume already contains the (single) target protein at the desired concentration and / or absolute amount, avoiding additional steps to concentrate or dilute the target protein. Furthermore, the applicant has observed that the chromatographic method according to embodiments of the present invention allows to predict the pooling boundaries of the target elution volume very accurately, thereby providing the target protein in a pure and predictable manner, minimizing the risk of losing a significant portion of the target protein by starting to collect the elution buffer leaving the column too late or terminating collection too early.

[0039] In the case of multiple target proteins, general aspects According to an embodiment, the applied protein solution comprises a first target protein, a second target protein and zero, one or more further proteins. The optimization criteria are input into the chromatography simulation software in the form of relative amounts of the first and second target proteins, for example in the form of a ratio between the first target protein and the second target protein. The chromatography simulation software calculates pooling boundaries of the target elution volume as a function of at least the calculated salt concentration and the input desired ratio. The method is used to generate a target elution volume that includes a desired ratio between the first and second target proteins.

[0040] This can be particularly advantageous in use case scenarios where two proteins with highly similar chemical properties, e.g., two glycoforms, should be separated from all other proteins and obtained in a specific ratio. For example, it has been observed that some glycoforms of biomedically active compounds, such as antibodies or antibody fragments, are particularly effective when two or more different glycoforms are applied in a specific ratio. In prior art approaches, when two or more target proteins have highly similar chemical properties, obtaining a pure protein solution with defined relative amounts of these target proteins was often very time-consuming and sometimes infeasible. This was because traditional chromatographic approaches did not allow these proteins to be separated from each other. Rather, both proteins were obtained in the elution buffer in a ratio that changed as the elution progressed and could not be set to a desired value. An embodiment of the invention utilizes a chromatography simulation software configured to perform calculations and predictions of the elution behavior of a first target protein, to perform calculations and predictions of the elution behavior of a second target protein (any one of the further proteins applied to the column, if present, the purity of at least one or more target proteins should be considered), and to automatically calculate and predict pooling boundaries of target elution volumes containing the first and second target proteins in a desired ratio, amount and / or purity. For example, the respective concentrations of the first and second target proteins and, optionally, also the further proteins, if present, contained in the applied protein solution in the elution buffer exiting the column can be predicted for a series of future time points (e.g., every second or every minute) for both the first and second target proteins. The obtained predicted concentration values ​​can be interpolated and / or extrapolated to obtain absolute and / or relative amounts and / or purities of the two or more target proteins for various candidate pooling boundaries. Pooling boundaries defining target elution volumes containing the first and second target proteins in a desired amount and / or purity can then be output and used to start and stop collection of the target elution volumes according to the pooling boundaries.

[0041] According to embodiments, the first and second target proteins are proteins with a low resolution factor, in particular a resolution factor of less than 0.75, in some instances less than 0.6 or even less than 0.5. As used herein, resolution factor is a number such as 0.8, 1.0, or 3.0 that indicates the degree of peak separation of two types of molecules in the eluate of a chromatography column. In general, resolution is the ability to separate two signals. In terms of chromatography, this is the ability to separate two peaks of two different types of molecules. Resolution R is given by the formula: TIFF2025029052000002.tif9128 formula, t r1 and t r2 and w1 and w2 are the time t and width w of two directly adjacent peaks, respectively. If the peaks are close enough, this is a reasonable problem, and the width w is approximately the same for both peaks. Assuming a Gaussian distribution for both types of molecules, a resolution factor of 1 indicates that 2.5% of the area of ​​the first peak overlaps with the area of ​​the second peak. Similarly, a resolution of 1.5 indicates a retention time difference of 1.5 x 4 x σ, where σ is the standard deviation of the Gaussian distribution, corresponding to an overlap of 0.15%. In general, the higher the resolution factor, the better the separation of the two types of molecules. It has been observed that embodiments of the present invention can separate even proteins with very low resolution factors, for example, less than 0.5.

[0042] It has been observed that embodiments of the present invention are able to separate proteins even when they have a very low resolution factor. By also taking into account the chemical properties and amounts of the individual proteins loaded onto the column, by computationally predicting the optimal salt concentration of the elution buffer specifically for the type and amount of protein applied, and by performing the elution with an elution buffer having the predicted optimal salt concentration, it has been observed that embodiments of the present invention are able to separate even highly similar proteins from each other and / or accurately predict the target elution volume containing one or more target proteins in the exact desired amount and / or purity.

[0043] According to embodiments, the first and second target proteins are variants of a protein having the same amino acid sequence but different types, numbers or positions of PTMs. In particular, the first and second target proteins are glycoforms.

[0044] According to an embodiment, the first and second target proteins are antibody monomers having identical amino acid sequences and containing different numbers of glycosyl groups on the FAB fragment.

[0045] According to some embodiments, the elution process is a stepwise elution process in which multiple different elution buffers ("steps") are applied sequentially, where computational identification of the (most suitable) elution salt concentration, calculation of pooling boundaries of the target elution volume as a function of the calculated elution salt concentration, and application of an elution buffer with the calculated salt concentration are performed for each step.

[0046] For example, each of the plurality of chromatography simulations can be a simulation of a chromatography process using two or more elution steps, where each elution step uses an elution buffer having a different elution salt concentration. The calculated elution buffer salt concentrations are a series of elution buffer salt concentrations specific to the different elution steps. Applying the elution buffer with the calculated salt concentrations to the chromatography column includes stepwise applying a series of elution buffers with different salt concentrations according to the series of calculated salt concentrations specific to the elution steps.

[0047] According to an embodiment, the applied protein solution comprises each of the two or more target proteins and, if present, each of the one or more further proteins, in a respective concentration of at least 0.5% by weight, in particular at least 1% by weight, in particular at least 2% by weight.

[0048] The applicant has observed that the protein concentration influences the elution behavior of the protein, especially if it is within the range mentioned above. By taking into account that the concentration of each protein contained in the protein solution applied to the column exceeds the above threshold, also called the "minimum amount threshold", the accuracy of the prediction of the salt concentration of the appropriate elution buffer can be increased. The proteins applied to the column can interact with each other and with the stationary phase in a number of ways. Some proteins may block other proteins from binding to the column resin, may compete with other proteins for binding to the stationary phase, or may promote their binding. By taking into account the chemical properties and concentrations of all proteins contained in the protein solution (i.e. all target proteins and all further proteins, if any, whose concentrations exceed the above-mentioned minimum amount threshold), the accuracy of the calculations and predictions of the chromatography simulation software can be significantly increased.

[0049] Preferably, the protein solution applied to the column is a pre-purified protein solution that contains only one or more target proteins, typically also one or more non-target proteins. For example, first, whole proteins can be extracted from a cell culture, such as a cell culture modified to produce one or more target proteins of interest. In a next step, this cell culture protein extract is pre-purified to obtain only a small number of proteins, preferably only one or more target proteins, and as few other proteins as possible. However, in many cases, it is not possible to obtain one or more target proteins in pure form in the pre-purification step. For example, a cell culture protein extract may contain thousands of different types of proteins. In the pre-purification step, the protein extract is applied to an affinity chromatography column to obtain a limited set of proteins, for example all proteins with a particular protein sequence or with a particular epitope. Thus, the protein solution applied to the cation chromatography column can be a protein solution obtained by performing a pre-purification process, such as affinity chromatography, on a cell culture extract or other source.

[0050] For example, the proteins in the protein solution resulting from a preliminary purification procedure using an affinity chromatography column may consist of all glycoforms of a particular protein, e.g. a glycoform with no glycosyl groups, a second glycoform with one glycosyl group, a second glycoform with a different type of glycosyl group, a third glycoform with the same glycosyl group as the first glycoform but at a different position, etc. Optionally, the protein solution may further comprise further proteins that are not desired / interesting but failed to be removed during the preliminary purification step for some reason.

[0051] Typically, the number of protein types in the protein solution applied to the cation chromatography column is limited and much smaller than the number of proteins in the cell culture extract, for example, the number of protein types in the protein solution applied to the cation chromatography column is typically less than 10, and often less than 5.

[0052] According to an embodiment, the protein solution applied to the column comprises a first P1 and a second P2 target protein, and a further protein P3, considered as a contaminating protein. The affinity of the second target protein P2 and the further protein P3 to the stationary phase of the column is similar to the affinity of the first target protein to the stationary phase, resulting in overlapping elution behavior of P1, P2 and P3.

[0053] For example, proteins P1, P2 and P3 are glycoforms or other PTM variants of the same protein and therefore have similar affinity to the stationary phase. Separating P1 and P2 from P3 and providing P1 and P2 in a desired ratio was typically not feasible with prior art cation chromatography approaches. An embodiment of the present invention allows first to predict an appropriate elution salt concentration, and then, based on this optimal salt concentration and other parameter values, to predict pooling boundaries that allow a more accurate separation of the target protein from other proteins in a particularly accurate manner, or, if this is physically impossible and the peaks are strongly overlapping, to at least collect a target elution volume that contains the two target proteins P1 and P2 in an acceptable desired ratio (with respect to contamination by P3).

[0054] According to an embodiment, the total amount of protein in the protein solution applied to the column is equal to or smaller than the maximum protein loading capacity of the column, in particular within the range of 50% to 100%, for example 50% to 90%, of the maximum protein loading capacity.

[0055] The applicant has observed that within this range of amounts, the chromatographic procedure is particularly sensitive to the concentration and chemical properties of the proteins contained in the applied protein solution, and that by taking into account the amount of each protein contained in the applied protein solution, the prediction accuracy regarding the optimal elution salt concentration is significantly improved.

[0056] According to an embodiment, the calculated pooling boundaries of the target elution volume are specified in the form of an acquisition start time offset and an acquisition stop time offset. - continuously monitoring the time elapsed from the start of elution by an automated chromatography system comprising a chromatography column; - automatically starting collection of elution buffer by the chromatography system when the elapsed time is equal to the collection start time offset; and - Stopping collection of elution buffer eluted by the chromatography system when the elapsed time is equal to the collection stop time offset. Includes.

[0057] According to an embodiment, the method comprises inputting into the chromatography simulation software the respective amounts of the first and second target proteins contained in the applied protein solution and, optionally, if present, also the respective amounts of one or more further proteins contained in the applied protein solution.

[0058] The simulation is calculated as a function of a set of parameter values ​​including at least the following: - the dimensions of the cation exchange chromatography column provided (e.g., length, diameter and height or total volume may be provided; the dimensions may also be selected from one of a set of predefined column types known to the chromatography simulation software), and - the amount of the first and second target proteins and, optionally, the amount of one or more further proteins applied to the column. For example, the amount of the first and second target proteins and any further proteins contained in the protein solution can be determined prior to the calculation of the simulation by a suitable protein analysis, for example by UV measurement at 280 nm to measure the total protein amount, the molecular weights of the various proteins applied to the column, which differ slightly due to different glycosylation patterns, can be determined by cation exchange high performance liquid chromatography (CEX-HPLC). This molecular weight information can be used, for example, to calculate the number / molar concentration of protein molecules of a particular protein applied to the column and / or to verify whether the ratio of the target proteins in the target elution volume meets, for example, a required amount, ratio or purity.

[0059] The applicant has observed that the amount of protein applied to the column may vary greatly from case to case. By taking into account the amount of each individual protein applied to the column and the dimensions of the column, it may be possible to calculate and predict the most suitable salt concentration of the elution buffer for obtaining particularly precisely the target elution volume containing the target protein in the desired amount and / or purity. This particular computationally determined salt concentration is then used as the salt concentration of the elution buffer actually used to perform the chromatographic procedure. Furthermore, the computationally determined salt concentration of the elution buffer is used to predict the pooling boundaries of the target elution volume containing the target protein in the desired amount and / or purity. The computationally predicted pooling boundaries can be output to a human or automated elution buffer collection unit, thereby facilitating the obtaining of one or more target proteins in the desired amount and / or purity.

[0060] The elution buffer obtained according to the embodiments of the invention is therefore not a "standard" / "one-size-fits-all" elution buffer, but rather an elution buffer that is specifically adapted to the type and amount of proteins contained in the protein solution and adapted to the dimensions of the column. The Applicant has observed that by adapting in each case the elution salt concentration to the column used and the amount of protein applied to the column, the capacity of the chromatographic column to separate different proteins can be significantly increased.

[0061] According to an embodiment, the simulation is performed as a function of both a number of different salt concentrations (sometimes referred to as "candidate salt concentrations") and a number of different elution buffer pH values ​​(sometimes referred to as "candidate pH values"). As a result of the simulation, a particular salt concentration and pH value combination that best satisfies the optimization criteria is identified. This salt concentration and pH value pair is used as the salt concentration and pH value of the elution buffer actually used to elute the protein.

[0062] According to an embodiment, the set of parameters further comprises: - a pre-defined pH value of the elution buffer (for example, the pH value may be the pH value of an elution buffer that has already been successfully used to elute the target protein or a similar protein previously), - the flow rate of the elution buffer through the column, - The chemical properties of the proteins in the applied protein solution.

[0063] The flow rate of the elution buffer through the column can be a flow rate set by an elution buffer pump configured to pump the elution buffer through the column at a preset rate. The flow rate can be specified, for example, in ml / min. The flow rate can also be specified, for example, in mm / sec, since the column diameter is known, allowing the flow rate in ml / min to be derived from the mm / sec value.

[0064] In some embodiments, the flow rate generated by the pump can be set by a user by configuring the pump. In some embodiments, the user manually enters the elution buffer flow rate (which may be displayed on the pump's display) in the chromatography simulation software for each chromatography run. In other embodiments, the elution buffer flow rate is stored in a configuration file of the chromatography simulation software and is reused for multiple chromatography runs until the user changes the elution buffer flow rate in the chromatography simulation software configuration file and in the pump settings.

[0065] The accuracy of the simulations and predictions can be further improved by taking into account the amount as well as the chemical properties of the target protein, and optionally also one or more further proteins, that are applied to the column.

[0066] According to embodiments, the pooling boundaries actually output to the user may be converted to different formats for ease of understanding. For example, the collection start time offset and collection stop time offset may be specified as time offsets starting from the beginning of the "280 nm peak" (representing the peak of total protein concentration in the eluate), which may also be predicted and empirically compared to the optically measured 280 nm peak signal. Similarly, the collection start time offset and collection stop time offset may be specified in "column volumes".

[0067] Many different computational approaches exist for simulating cation chromatography runs and predicting the elution profile of a particular protein and can be used to perform simulations and predict target elution boundaries according to embodiments of the present invention.

[0068] For example, Wittkopp F, Peeck L, Hafner M, Frech C. "Modeling and simulation of protein elution in linear pH and salt gradients on weak, strong and mixed cation exchange resins applying an extended Donnan ion exchange model", J Chromatogr A. 2018;1545:32-47. doi:10.1016 describes the application of the Donnan equilibrium ion exchange (DIX) model for modeling and simulation of ion exchange chromatography of monoclonal antibodies in linear chromatography. This publication deals with the description of conditions in a chromatography column at low protein concentrations but does not concern the optimization of protein separations.

[0069] By way of another example, Muller-Spath T, Strohlein G, Aumann L et al., "Model simulation and experimental verification of cation-exchange IgG capture step in batch and continuous chromatography", J Chromatogr A. 2011;1218(31):5195-5204, doi:10.1016 describes the simulation of the cation-exchange capture step of a mAb purification process using single-column batch and multi-column continuous chromatography (MCSGP) based on a lumped kinetic model using model parameters such as void fraction, retention factor and mass transfer parameters of the purified monoclonal antibody (mAb).

[0070] According to an embodiment, the chromatography simulation software is configured to use a combination of mathematical models to calculate the simulation and / or to calculate pooling boundaries for the target elution volume. The models include: - a column model constructed to correlate the concentration of each protein in the elution buffer, the salt concentration and the pH value in the void volume of the column; and - a pore model configured to correlate the concentration, salt concentration and pH value of each protein in the elution buffer in the pore volume of the stationary phase of the column; and - a reaction model configured to correlate at least some of the concentrations of each protein in the stationary phase, the salt concentrations of the elution buffer, and the chemical properties of each (target and optionally further) protein in the protein solution.

[0071] A number of different column, pore and reaction models have been described in the literature. Many of these models are continually being improved. A list of references describing examples of models that may be used in the context of the present invention is disclosed, for example, in the description of FIG. 4C.

[0072] For example, the column model can be implemented according to Schmidt Traub et al., "Preparative Chromatography" 2012, Viley-VCH. The pore model can be implemented according to Schmidt Traub et al., "Preparative Chromatography" 2012, Viley-VCH. The reaction model can be implemented according to Kluters et al., "Application of linear pH gradients for the modeling of ion exchange chromatography: separation of monoclonal antibody monomer", J Sep Sci, February 2016, 93(4), 663-675. The reaction model parameters and / or the model itself can be modified with various levels of complexity to facilitate the explanation of experimental data. This is demonstrated, for example, in "Modeling of complex antibody elution behavior under high protein load densities in ion exchange chromatography using an asymmetric activity coefficient" by Huuk et al., Biotechnol.J., 12:1600336.doi:10.1002 / biot.201600336. This publication describes how generalized ion exchange isotherms can be used to model the separation of single protein species under high protein loading conditions.

[0073] The pore model, column model and reaction model are interconnected via salt concentration and protein concentration. The column model and pore model are The pore model formula can be inserted into the column model since they are interconnected via TIFF2025029052000003.tif7128.

[0074] The first equation of the reaction model is: The parameters that describe the binding strength, or chemical properties, of the protein can be inserted into the column model via the TIFF2025029052000004.tif7128 term. The equilibrium constant K eq and the number of binding sites, v. Both depend on pH via v (the second equation of the reaction model).

[0075] Thus, the salt concentration of the elution buffer required to elute a desired amount, ratio and / or purity of target protein can be calculated based on a set of parameters such as the pH value, the amount and chemical properties of the applied protein, the column dimensions, and the flow rate of the elution buffer defined by the pump.

[0076] The protein amount of the target protein, or any additional protein, is the integral over time of the concentration of this protein in the mobile phase exiting the chromatography column. The calculation software calculates the concentration of the protein in the mobile phase by the salt concentration c SALT and performing a prediction of the elution profile of the target protein, and optionally performing one or more further predictions as a function of salt concentration, to identify the salt concentration that provides the protein elution profile that best meets the optimization criteria.

[0077] Column Model According to an embodiment, the interrelationships specified in the column model include the following column model expressions: TIFF2025029052000005.tif22130

[0078] where i is the index of a particular protein, t is time, x is the position within the column along the vertical axis of the column, and c SALT is the concentration of free (unbound) salt ions (typically Na+ in cation exchange chromatography) in the column's mobile phase (i.e., the elution buffer phase containing a negligible amount of the applied protein solution), and c PROT_i is the concentration of protein i in the column mobile phase, and q PROT_i is the concentration of bound protein i in the stationary phase, and u intis the interstitial flow velocity, defined as the velocity obtained by flow through the column between the stationary phase particles, and ε col is the pore porosity (the open area in the cross section of the column, which is defined as the ratio of the pore volume to the column volume, where the pore volume is the volume of elution buffer in the column outside the particles), ε p is the particle porosity (defined as the internal porosity (pore volume) of the particle itself), D ax is the axial dispersion coefficient. The axial dispersion coefficient is a measure of the extent of spread of the inert trace material along the length of the column. D ax describes the axial dispersion as stated in Fick's law.

[0079] Porosity ε col , ε _p and the dispersion coefficient D ax are parameters that can affect the separation of the eluted protein peaks. The porosities are usually fixed and are determined by the chromatographic resin contained in the column. Preferably, they are stored as default values, e.g. in a configuration file, but can be changed if different stationary phase materials are used.

[0080] Thus, according to an embodiment of the invention, the dimensions of a cation exchange chromatography column are considered via the interstitial flow velocity (normalization of flow to column diameter in the pore model) and the position within the column as reflected by the vertical position x of the column. The combination of the length and diameter of the column defines the total volume of the column.

[0081] The interstitial flow rate is computationally derived from the flow rate of the elution buffer through the column.

[0082] For example, the user can input the flow rate of the elution buffer, "ebfr" in [mm / sec], into the chromatography simulation software. The chromatography simulation software will calculate the interstitial flow velocity u(int) [mm / sec] according to the following equation: TIFF2025029052000006.tif5128 in the formula, εcol is the pore column porosity (unitless).

[0083] A specific example of how the interstitial flow rate can be derived from the flow rate of the elution buffer (which can be defined by a pump configured to pump the elution buffer through the column) is described in Schmidt-Traub, Henner, eds., "Preparative chromatography: of fine chemicals and pharmaceutical agents", John Wiley & Sons, 2006.

[0084] Pore ​​Model According to an embodiment, the interrelationships specified in the pore model include the following pore model equations: TIFF2025029052000007.tif9128, where i is the index of a particular protein, t is time, x is the position in the column along the vertical axis of the column, and c SALT is the concentration of free (unbound) salt ions in the mobile phase of the column, and c PROT_i is the concentration of protein i in the column mobile phase, and q PROT_i is the concentration of bound protein i in the stationary phase, and ε col is the pore column porosity, and ε _p is the particle porosity and k eff is the effective mass transfer coefficient (describes the transfer of proteins from the interstitial mobile phase into the pore volume), z a value in the range of 2.7 to 3.3, in particular 3.0r_p is the radius of the chromatography beads. The parameter z depends on the geometrical aspects of the column material. Specific examples of how to obtain the parameter z are described in Schmidt-Traub, Henner (Eds.), Preparative chromatography: of fine chemicals and pharmaceutical agents. John Wiley & Sons, 2006, pp. 338-339.

[0085] According to one example, the interstitial flow rate can be related to the flow rate of the elution buffer through the column as follows: The flow rate through the column is the volumetric flow rate, u p Also called (mL / min), it describes the volume of mobile phase pumped through the column. To normalize the flow rate to the dimensions of the chromatography column, the column volume and length can be used to normalize the flow rate with respect to the column diameter. The units of flow rate u are changed to (mm / sec) according to the following formula: TIFF2025029052000008.tif14128The pore flow velocity is the pore column volume V interstitial We further normalize the flow rate with respect to the porosity of the chromatography beads by considering only the interstitial volume (i.e., only the volume between the porous chromatography beads). The ratio of the interstitial volume to the total column volume of the chromatography column used is given below as ε col As follows: TIFF2025029052000009.tif8128.

[0086] Therefore, the parameter ε col is the ratio of the interstitial volume to the total column volume.

[0087] The stationary phase of the column can be described as a matrix of beads. The total volume of all the beads in the column is the "bead volume", V beads The volume of each bead includes both the volume used by the stationary phase material of the bead and the volume of the pores contained in the bead. The void volume V of a chromatography column interstitial is the total volume of the column, V col From the total bead volume V of all beads contained in the column, beads According to an embodiment, the column model and the core model are combined. For example, according to an embodiment, the calculation of the target elution volume as a function of at least the calculated salt concentration is defined as the salt concentration c SALT and the concentration q of protein i in the stationary phase PROT_i This involves using the column model equation and the pore model equation to calculate the concentration of protein i in the void volume as a function of

[0088] According to an embodiment, calculating the pooling boundaries of the target elution volumes as a function of at least the calculated salt concentration includes using a column model equation and a pore model equation to calculate the concentration of the first target protein at a position x = column length at a plurality of different elution times t.

[0089] Reaction Model According to an embodiment, the interrelationships specified in the reaction model include: The equation of the first reaction model according to: TIFF2025029052000010.tif7128Equation of the second reaction model according to: TIFF2025029052000011.tif22128Equation of the third reaction model according to: TIFF2025029052000012.tif8128.

[0090] where i is the index of a particular protein, t is time, x is the position within the column along the vertical axis of the column, and c SALT is the concentration of free (unbound) salt ions in the mobile phase of the column, and c PROT_i is the concentration of protein i in the column mobile phase, and q PROT_i is the concentration of bound protein i in the stationary phase, and K kin is the kinetic rate (defined as the inverse of the desorption rate, describing the rate of reaction of a protein to displace ions from its ligands in the stationary phase), Keq is the equilibrium constant (i.e., the molar concentration of a particular protein i in the stationary phase divided by the molar concentration of protein i in the mobile phase), Λ is the ligand density (defined as the number of ligands in mol per column volume), and ν _iis the number of binding sites (i.e., the number of binding sites of protein i involved in protein binding to the stationary phase), σ_i is the shielding coefficient (i.e., the number of ligands per bound protein molecule of protein i that are shielded when protein i binds to the stationary phase), N_(carb,i) is the number of carboxy sites of protein i that interact with the stationary phase, N_(his,i) is the number of histidine sites of protein i that interact with the stationary phase, N_(N-term,i) is the number of N-terminal sites of protein i that interact with the stationary phase, N_(ami,i) is the number of amino sites of protein i that interact with the stationary phase, pH is the pH value of the elution buffer, TIFF2025029052000013.tif5128 is the standard Gibbs free enthalpy difference between the adsorbed and unbound states of protein i, TIFF2025029052000014.tif5128 is the standard Gibbs free enthalpy difference between the adsorbed and unbound states of the salt, TIFF2025029052000015.tif5128 is the average acid dissociation constant (Ka) of all carboxyl groups of protein i interacting with the stationary phase, TIFF2025029052000016.tif5128 is the average Ka of all histidine-amino acids of protein i that interact with the stationary phase, TIFF2025029052000017.tif5128 is the average Ka of all N-terminal α-amino groups of protein i that interact with the stationary phase, where TIFF2025029052000018.tif5128 is the average Ka of all amino groups of protein i, R is the gas constant, and T is the temperature of the column and elution buffer. The concentration of a particular protein "i" in the stationary phase ("q PROT (i)" uses the equations of the first and second reaction models to derive the equation of the third reaction model. PROT (i) as a function of the pH value of the elution buffer.

[0091] Salt concentration c SALTis the protein bound to the stationary phase, q, via the first reaction model equation PROT_i At low salt concentrations, proteins bind to the chromatography column, whereas at higher salt concentrations, proteins are displaced by salt ions and eluted. The reaction model is governed by the number of binding sites, ν, and the equilibrium constant, K eq Depending on the binding strength of the protein, which consists of , proteins elute at different salt concentrations. By choosing the appropriate salt concentration, it is possible to elute the target protein while leaving undesired proteins bound to the chromatography column. Some of these parameter values ​​are protein-specific and can be derived from the literature. Some of these parameters are protein- and model-specific. They may also be derived from the literature or can be obtained by performing some preliminary empirical calibration steps, usually described by the authors who published the respective model used. For example, parameters such as N_(carb,i), which indicates the number of carboxyl sites of protein i that interact with the stationary phase, are protein-specific but also depend on the model used to perform the simulation. The number of carboxyl sites that actually interact with the stationary phase is limited by steric effects described in the reaction model used. The number of the respective sites that interact with the stationary phase is determined using protein- and model-specific empirical tests. For some proteins, parameter values ​​can be readily derived from the literature. The Gibbs energy ΔG 0 P / RT and ΔG 0 s / RT may also be empirically determined for each protein in the protein solution or derived from the literature.

[0092] The pH value affects the number of binding sites v on a protein (equation 2). The number of binding sites v affects the equilibrium constant Keq (equation 3). This relates the pH value to the salt concentration in equation 1 of the reaction model.

[0093] Thus, the reaction model takes into account the contribution of the pH value and the chemical properties of the protein, for example with respect to the pKa of certain groups in the protein and the binding site v of the protein. The reaction model also takes into account the salt concentration, thereby correlating the salt concentration of the elution buffer with the chemical properties of the eluted protein. The salt concentration c SALT is part of the first formula and is directly linked to the protein binding properties via the number of binding sites v.

[0094] However, the above-mentioned reaction model is only one of many possible ways that the reaction model can be implemented.Since modeling and simulation of cation chromatography execution is currently an ongoing research field, various modifications of the formulas and models described herein can be used to perform simulations as well.It is also possible to use alternative column, pore and reaction models currently described in the literature, and in the future, new column, pore and reaction models may be developed and published that can be used to perform chromatography simulations as a function of at least some different elution buffer salt concentrations, and preferably as a function of additional parameters such as the amount and chemical properties of protein applied to the column.

[0095] According to one embodiment, the following model-specific chemical property values ​​for proteins P1 to P3, e.g. as described with reference to FIG. 7, are provided and used as inputs by the reaction model during calculation of appropriate elution buffer salt concentrations and pooling boundaries: TIFF2025029052000019.tif56144

[0096] Additionally, the following input parameter values ​​can be provided to the reaction model, for example via a GUI, a configuration file, or other data source: TIFF2025029052000020.tif26128

[0097] The shielding factor decreases from variant 1 to variant 3, which is consistent with the degree of reduction in glycosylation and therefore protein size from variant 1 to variant 3.

[0098] The parameter values ​​mentioned above may be provided by the user via a GUI and / or may be specified in a configuration file of the chromatography simulation software.

[0099] Various techniques exist to obtain and / or estimate the model parameters. For example, the research paper "Application of linear pH gradients for the modeling of ion exchange chromatography:Separation of monoclonal antibody monomer from aggregates" (https: / / doi.org / 10.1002 / jssc.201500994), first published on November 9, 2015, by Simon Kluters et al., describes a mechanistic model of ion exchange in a chromatography column and how the parameters of this model can be obtained.

[0100] In a further aspect, the present invention relates to a chromatography control system including a simulation software. For example, the chromatography control system can be used to guide manual control of a chromatography system or to provide fully automated or semi-automated control of the control system. The simulation software is configured to implement a method for obtaining pooling boundaries for target elution volumes that contain a desired first amount of a first target protein. The chromatography control system is configured for: - receiving an optimization criterion and inputting the optimization criterion into the chromatography simulation software, the optimization criterion being a desired characteristic of the target elution volume with respect to the first and second target proteins contained in the target elution volume; - calculating, using chromatography simulation software, salt concentrations of elution buffers adapted to elute the first and second target proteins from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained (calculating includes calculating a plurality of chromatography simulations as a function of salt concentrations of a plurality of different elution buffers); - calculating pooling boundaries of target elution volumes as a function of at least the calculated salt concentration and the input optimization criteria; - outputting the calculated salt concentrations and pooling boundaries; and / or - Controlling the elution volume collection unit such that the calculated target elution volume is automatically collected as separate fractions according to the calculated pooling boundaries.

[0101] According to an embodiment, the system is configured to: - receiving the dimensions of a cation exchange chromatography column; and - receiving amounts of the first and second target proteins, and optionally also amounts of one or more further proteins, that are applied to the column.

[0102] The simulation and / or pooling bounds are calculated as a function of a set of parameter values ​​including at least the following: ● the dimensions of the cation exchange chromatography column provided, and - The amounts of the first and second target proteins, and optionally, one or more further proteins, that are applied to the column.

[0103] According to an embodiment, the set of parameters further comprises: A given pH value of the elution buffer, - the flow rate of elution buffer through the column; • Chemical properties of the proteins in the applied protein solution.

[0104] According to an embodiment of the present invention, the desired first amount of the first protein is specified as a desired absolute amount of the desired first protein. In particular, the desired amount of the first target protein can be specified in the form of a desired concentration of the first protein in the collected volume of elution buffer or a derivative thereof.

[0105] According to another embodiment, the applied protein solution comprises a first target protein, a second target protein and zero, one or more additional proteins. A desired first amount of the first protein is input into the chromatography simulation software in the form of a relative amount. The relative amount is a ratio of the first target protein to the second target protein. The chromatography simulation software calculates a pooling boundary of the target elution volume as a function of at least the calculated salt concentration and the input desired ratio. The method is used to generate a target elution volume that includes a desired ratio of the first and second target proteins.

[0106] According to an embodiment, a plurality of chromatographic simulations are calculated as a function of the salt concentrations of a plurality of different elution buffers and as a function of the pH values ​​of a plurality of different elution buffers. The calculation of the salt concentrations includes calculating a combination of elution buffer salt concentrations and elution buffer pH values ​​that are combined and adapted to elute the first and second target proteins from the chromatographic column so as to obtain a target elution volume that best meets the optimization criteria. The chromatographic simulations are performed to identify a combination of elution buffer salt concentrations and elution buffer pH values. The calculation of the pooling boundaries of the target elution volume is calculated as a function of at least the calculated combination of elution buffer salt concentrations and elution buffer pH values ​​and the input optimization criteria. The calculated pH values ​​are output in addition to the calculated elution buffer.

[0107] According to an embodiment, the chromatography control system is configured to control the buffer mixing unit for automatically generating an elution buffer having the output elution salt concentration.

[0108] Additionally or alternatively, the chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations, the selected elution buffer having the output salt concentration.

[0109] Additionally or alternatively, the chromatography control system is configured to control the buffer mixing unit to automatically generate an elution buffer having both a salt concentration and a pH value calculated and output in combination.

[0110] Additionally or alternatively, the chromatography control system is configured to control an elution volume collection unit of the chromatography system such that the calculated target elution volumes are automatically collected as separate fractions according to the calculated pooling boundaries.

[0111] Additionally or alternatively, the chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations and different pH values, the selected elution buffer having both the calculated and outputted salt concentration and pH value in combination.

[0112] Additionally or alternatively, the chromatography control system is configured to control a buffer application unit configured to automatically apply an automatically generated or selected elution buffer to the chromatography column, the applied elution buffer having an output salt concentration or having a combination of both a calculated and output salt concentration and a pH value.

[0113] In a further aspect the present invention relates to a chromatography system comprising a chromatography control system and further comprising the above mentioned buffer mixing unit and / or elution buffer selection unit and / or buffer application unit and / or automatic elution volume collection unit.

[0114] In a further aspect, the present invention relates to a chromatography system comprising a chromatography control system and further comprising a buffer mixing unit adapted to automatically generate an elution buffer having an output elution salt concentration.

[0115] Additionally or alternatively, the chromatography system further comprises an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations, the selected elution buffer having the output salt concentration.

[0116] Additionally or alternatively, the chromatography system further comprises a buffer application unit configured to automatically apply an automatically generated or selected elution buffer to the chromatography column.

[0117] Additionally or alternatively, the chromatography system further comprises an automatic elution volume collection unit configured to automatically start and stop collecting fractions of elution buffer exiting the column according to the output pooling boundaries.

[0118] In a further aspect, the present invention relates to a computer program for managing a chromatography process such that a target elution volume comprising a first and a second target protein is obtained. The computer program comprises chromatography simulation software and is configured to: - receiving optimization criteria that are desired characteristics of a target elution volume for a first and a second target protein contained in the target elution volume; - calculating, using chromatography simulation software, salt concentrations of elution buffers adapted to elute the first and second target proteins from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained (calculating includes calculating a plurality of chromatography simulations as a function of salt concentrations of a plurality of different elution buffers); - calculating pooling boundaries of target elution volumes as a function of at least the calculated salt concentration and the input optimization criteria; and - outputting the calculated salt concentrations and / or pooling boundaries to enable a user to control a chromatography system such that a target elution volume comprising the first and second target proteins is obtained according to the optimization criteria, and / or using the calculated salt concentrations and / or pooling boundaries to automatically or semi-automatically control a chromatography system such that a target elution volume comprising the first and second target proteins is obtained according to the optimization criteria.

[0119] According to an embodiment, the computer program is further configured to generate control commands for automatically or semi-automatically controlling one or more units of the chromatography system.

[0120] According to an embodiment, the computer program is further configured to generate control commands for automatically or semi-automatically controlling one or more units of the chromatography system.

[0121] A "target protein," as used herein, is a protein or protein variant having a particular PTM that can be obtained by cation chromatography in a desired absolute or relative amount.

[0122] The "elution profile" of a protein, as used herein, is a measure of the concentration of the protein in an eluate over a period of time.

[0123] The one or more "additional proteins" contained in the applied protein solution may comprise one or more additional target proteins and / or one or more non-target proteins, the "non-target proteins" being proteins whose presence is undesirable, for example because they are considered contaminants to be removed from the one or more target proteins during a chromatography run.

[0124] "Target elution volume," as used herein, is the specific volume of target buffer that has passed through a chromatography column and then been collected after exiting the column. The target elution volume contains one or more target proteins in a desired absolute or relative amount. The target elution volume is determined by:

[0125] "Ion exchange chromatography", as used herein, is a chromatographic process that separates ions and polar molecules, such as proteins or peptides, based on their affinity for an ion exchanger. However, ion chromatography must be performed under conditions that are one unit away from the isoelectric point of the protein. Thus, separation of molecules with highly similar isoelectric points under given conditions is a major challenge.

[0126] The term "cation exchange chromatography" refers to a type of ion exchange chromatography that is used when the molecules of interest are positively charged or have a positive polarity. The molecules are positively charged or have a positive polarity because the pH of the chromatography is below the pI. In this type of chromatography, the stationary phase is negatively charged and loaded such that positively charged molecules are attracted to it. Cation exchange chromatography is used when the desired molecules to be separated are cations. The bound molecules can then be eluted from the column and collected using an elution solution, referred to herein as "elution buffer". One of the main advantages of using ion chromatography is that, in contrast to other separation techniques, only one interaction is involved during the separation.

[0127] "Chromatography simulation software", as used herein, is any type of software, e.g., an application program, script, software module of a larger software suite or software platform, that contains computer-interpretable code to simulate (predict) some aspect of a chromatography process, in particular the recommended salt concentration of an elution buffer and / or the recommended elution boundaries of an elution volume containing a desired amount of one or more target proteins.

[0128] "Elution buffer," as used herein, refers to an aqueous solution, also called an "eluent," that is applied to a chromatography column to elute one or more molecules bound to the column from the column, such that the eluted proteins are contained in the elution buffer that exits the column and can be collected separately from other proteins that exit the column in different fractions of the eluate.

[0129] A "calculated target elution volume" is an elution volume predicted to contain one or more target proteins at an input desired absolute or relative amount. For example, a calculated target elution volume can be specified via a predicted pooling boundary.

[0130] The predicted "pooling boundary" is a data value that indicates what portion of the target elution buffer volume leaving the chromatography column needs to be collected separately to obtain an elution volume containing the desired amount of one or more target proteins. According to some embodiments, the pooling boundary is specified in the form of a start time indicating the time when the collection of the target elution buffer volume should start, and a stop time indicating the time when the collection of the target volume should stop. According to other embodiments, the pooling boundary is specified in the form of a start elution buffer volume indicating the elution buffer volume that has crossed and left the column when the collection of the target volume should start, and a stop elution buffer volume indicating the elution buffer volume that has crossed and left the column when the collection of the target volume should stop. For example, if the pooling boundaries of the target elution volume are "3.5 column volumes" and "4.6 column volumes", the target elution volume can be obtained by selectively collecting the elution buffer that is leaving the column after 3.5 column volumes of elution buffer have already been applied to the column and have left the column, and stopping the collection when more than 4.6 column volumes have been applied to the column and are about to leave the column. Whether the expected pooling boundaries are specified as times or amounts is up to each implementation.

[0131] A "glycoform," as used herein, is any of several different forms of a glycoprotein (or glycopeptide or other biological glycoside) that have different glycans attached or that differ in the number or location of the glycans.

[0132] The "purity" of a protein indicates the degree to which this protein is not mixed with other things, especially not mixed with other proteins. For example, a certain protein P1 with a purity of 99% in the elution volume means that the elution volume consists of an elution buffer fraction and a non-elution buffer fraction, 99% of the non-elution buffer fraction consists of protein P1, and 1% of the non-elution buffer fraction consists of other substances, for example one or more others of proteins P2, P3, ... contained in the applied protein solution. If the concentrations of all proteins in the eluate collected within a given pooling boundary are known (for example predicted or empirically determined), the purity of the protein is implicitly given. Purity can be used as an optimization criterion for one or more target proteins. For example, in the ideal case, the purity of all target proteins obtained in the target elution volume is 100%, but in practice, the optimization criterion may require a minimum purity criterion that depends on the individual case, for example 99% or less. [The present invention 1001] 1. A chromatography method for producing a target elution volume (238) comprising a first target protein (P1) and a second target protein (P2), comprising: - providing (100) a cation exchange chromatography column (230); - applying (102) to said column a protein solution (202) comprising at least said first target protein (P1), said second target protein (P2) and optionally one or more further proteins (P3, P4, P5, P6); - inputting (106) optimization criteria into a chromatography simulation software (204), the optimization criteria being desired properties of the target elution volume with respect to the first target protein and the second target protein contained in the target elution volume; - calculating (108) by the chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute the first target protein and the second target protein from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained, the salt concentration (222) of the elution buffer comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) by the chromatography simulation software a pooling boundary of the target elution volume as a function of at least the calculated salt concentration and the inputted optimization criterion; - applying (112) an elution buffer (226) having the calculated salt concentration to the chromatography column; - performing elution using said applied elution buffer (114); - collecting (116) the calculated target elution volume as a separate fraction using the calculated pooling boundary; A method comprising: [The present invention 1002] The optimization criterion is a) a desired ratio or ratio range of the amounts of said first target protein and said second target protein in said target elution volume; b) a desired amount or range of amounts of said first target protein in combination with a desired amount or range of amounts of said second target protein in said target elution volume; c) ■ a desired ratio of the amounts of said first protein and said second protein in said target elution volume; ■ a desired amount or range of amounts of said second target protein in said target elution volume; ■ The desired purity or purity range of the second target protein in the target elution volume. a desired purity or purity range of said first target protein in combination with; d) A combination of two or more of the above criteria 1001. The chromatography method of claim 1001, wherein the compound is selected from the group consisting of: [The present invention 1003] Any of the above chromatography methods of the invention, wherein the applied protein solution (202) comprises the one or more further proteins (P3, P4, P5, P6). [The present invention 1004] - the plurality of chromatographic simulations are calculated as a function of salt concentration of the plurality of different elution buffers and as a function of pH value of the plurality of different elution buffers; Any of the chromatography methods of the present invention, wherein the method comprises: - said calculating (108) comprises calculating a combination of an elution buffer salt concentration (222) and an elution buffer pH value adapted in combination to elute said first target protein and said second target protein from said chromatography column so as to obtain a target elution volume that best meets said optimization criteria, said chromatographic simulation being performed to identify said combination of said elution buffer salt concentration (222) and said elution buffer pH value; - the pooling boundary of the target elution volume is calculated as a function of the calculated combination of at least the salt concentration of the elution buffer and the pH value of the elution buffer and the input optimization criterion; - the elution buffer (226) applied to the column has the calculated salt concentration and the calculated pH value. [The present invention 1005] Any of the methods of the above invention, wherein the first target protein and the second target protein are proteins having a resolution coefficient of less than 0.75 and / or less than 0.5. [The present invention 1006] Any of the methods of the above invention, wherein the first target protein and the second target protein are glycosylation variants of a protein having the same amino acid sequence. [The present invention 1007] The first target protein and the second target protein are Antibody monomers having the same amino acid sequence but containing different numbers of glycosyl groups on the FAB fragment Any of the methods of the present invention, [The present invention 1008] Any of the methods of the invention, wherein the protein solution applied comprises each of the target proteins (P1, P2) and, if present, each of the further proteins (P3 to P6), in a respective concentration of at least 0.5% by weight, particularly at least 1% by weight, in particular at least 2% by weight. [The present invention 1009] Any of the methods of the invention, wherein the second target protein (P2) and, if present, one or more of the further proteins (P3-P6) contained in the applied protein solution have an affinity for the stationary phase of the column similar to the affinity of the first target protein for the stationary phase, resulting in overlapping elution behavior. [The present invention 1010] Any of the methods of the present invention, wherein the total amount of protein in the protein solution applied to the column is the same as or smaller than the maximum protein loading capacity of the column, in particular within the range of 50% to 100%, for example 50% to 90%, of the maximum protein loading capacity. [The present invention 1011] the calculated pooling boundaries of the target elution volume are specified in the form of an acquisition start time offset and an acquisition stop time offset; The method further comprising: - continuously monitoring the time elapsed from the start of said elution by an automated chromatography system comprising said chromatography column; - automatically starting the collection of the elution buffer by the chromatography system when the elapsed time is equal to the collection start time offset; and - stopping the collection of the elution buffer by the chromatography system when the elapsed time is equal to the collection stop time offset. Including, Any of the methods of the present invention. [The present invention 1012] - each of said plurality of chromatography simulations is a simulation of a chromatography process using two or more elution steps, whereby in each elution step an elution buffer having a different elution salt concentration is used; - the calculated elution buffer salt concentrations are a series of elution buffer salt concentrations specific for different elution steps, - applying the elution buffer having the calculated salt concentration to the chromatography column comprises stepwise applying a series of elution buffers having different salt concentrations according to the series of salt concentrations specific for a calculated elution step. Any of the chromatography methods of the present invention. [The present invention 1013] The method further comprising: - inputting (104) into said chromatography simulation software (204) the amounts of each of said first and second target proteins (P1-P2) contained in said applied protein solution and, optionally, if present, also the amounts of each of one or more further proteins (P3-P6) contained in said applied protein solution; Includes; - the simulation comprises at least the dimensions of the cation exchange chromatography column provided; and the amount of said first target protein and said second target protein, and optionally the amount of said one or more further proteins, applied to said column; , calculated as a function of a set of parameter values ​​including Any of the chromatography methods of the present invention. [The present invention 1014] The set of parameters is - a predefined pH value of said elution buffer, - the flow rate of the elution buffer through the column; - the chemical properties of the proteins (P1, P2, P3 to P6) of the protein solution to be applied; The chromatography method of the present invention further comprises: [The present invention 1015] the chromatography simulation software is configured to use a combination of mathematical models to calculate the simulation r and / or to calculate the pooling boundaries of the target elution volume; The model is - a column model (206) configured to correlate the concentration of each of the proteins in the elution buffer in the void volume of the column (230), the salt concentration, and the pH value; and - a pore model (208) configured to correlate the concentration of each of the proteins in the elution buffer in a pore volume of the stationary phase of the column (230), the salt concentration, and the pH value; and - a reaction model (210) configured to correlate the concentration of each of the proteins in the stationary phase (228), the salt concentration of the elution buffer, and at least some of the chemical properties of each of the proteins in the protein solution. Any of the methods of the present invention comprising: [The present invention 1016] 1. A chromatography control system, comprising simulation software configured to implement a method for obtaining a pooling boundary for a target elution volume (238) comprising a first target protein (P1) and a second target protein (P2), the method comprising: - receiving (106) optimization criteria, which are desired characteristics of the target elution volume with respect to the first target protein and the second target protein contained in the target elution volume, and inputting the optimization criteria into the chromatography simulation software (204); - calculating (108) using the chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute the first target protein and the second target protein from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained, comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) the pooling boundaries of the target elution volumes as a function of at least the calculated salt concentrations and the input optimization criteria; - outputting the calculated salt concentrations and the pooling boundaries; A chromatography control system configured for: [The present invention 1017] The system further comprises: - receiving the dimensions of a cation exchange chromatography column (230); - receiving an amount of said first target protein and said second target protein applied to said column, and optionally an amount of said one or more additional proteins (105); further constituted for; - said simulation and / or said pooling bounds are at least the dimensions of the cation exchange chromatography column provided; and the amount of the first target protein and the second target protein, and optionally the amount of the one or more further proteins, applied to the column; , calculated as a function of a set of parameter values ​​including The chromatography control system of the present invention 1016. [The present invention 1018] - the plurality of chromatographic simulations are calculated as a function of salt concentration of the plurality of different elution buffers and as a function of pH value of the plurality of different elution buffers; The chromatography control system of the present invention 1016 or 1017, wherein the method further comprises: - calculating (108) comprises calculating a combination of an elution buffer salt concentration (222) and an elution buffer pH value adapted in combination to elute the first target protein and the second target protein from the chromatography column, such that a target elution volume that best meets the optimization criteria can be obtained, - calculating (110) the pooling boundary of the target elution volume is calculated as a function of the calculated combination of at least the salt concentration of the elution buffer and the pH value of the elution buffer and of the input optimization criterion; - The calculated pH value is output in addition to the calculated elution buffer (226). [The present invention 1019] the chromatography control system is configured to control a buffer mixing unit to automatically generate an elution buffer having the output elution salt concentration; and / or the chromatography control system is configured to control a buffer mixing unit to automatically generate an elution buffer having both the salt concentration and the pH value calculated and output in combination according to the present invention; and / or the chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations, the selected elution buffer having the outputted salt concentration; and / or the chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations and different pH values, the selected elution buffer having both the salt concentration and the pH value calculated and output in combination according to the present invention 1018; and / or the chromatography control system is configured to control a buffer application unit configured to automatically apply an automatically generated or selected elution buffer to the chromatography column, the applied elution buffer having the output salt concentration or having both the output salt concentration and the output pH value calculated in combination according to the present invention; and / or the chromatography control system is configured to control an elution volume collection unit of a chromatography system to automatically collect the calculated target elution volume as separate fractions according to the calculated pooling boundaries. The chromatography control system of any one of 1016 to 1018. [The present invention 1020] A chromatography system comprising the chromatography control system of the present invention 1019, and further comprising said buffer mixing unit and / or said elution buffer selection unit and / or said buffer application unit and / or said automated elution volume collection unit. [The present invention 1021] A computer program for managing a chromatography process to obtain a target elution volume (238) comprising a first target protein (P1) and a second target protein (P2), the computer program comprising chromatography simulation software; and - receiving optimization criteria (106), which are desired characteristics of a target elution volume for a first target protein and a second target protein contained in said target elution volume; - calculating (108) using the chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute the first target protein and the second target protein from the chromatography column such that a target elution volume that best meets the optimization criteria can be obtained, comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) the pooling boundaries of the target elution volumes as a function of at least the calculated salt concentrations and the input optimization criteria; and - outputting the calculated salt concentrations and / or the pooling boundaries to enable a user to control a chromatography system to obtain a target elution volume comprising the first target protein and the second target protein according to the optimization criterion, and / or using the calculated salt concentrations and / or the pooling boundaries to automatically or semi-automatically control a chromatography system to obtain a target elution volume comprising the first target protein and the second target protein according to the optimization criterion. A computer program configured for [The present invention 1022] The computer program of the present invention 1021 further configured to generate control commands for automatically or semi-automatically controlling one or more units of the chromatography system. [Brief description of the drawings]

[0133] In the following, embodiments of the invention will be described in more detail, purely by way of example, with reference to the drawings, in which: [Figure 1] 2 shows a flowchart of a method according to an embodiment of the present invention. [Diagram 2] 1 shows a chromatography system. [Diagram 3]1 shows another chromatography system. [Figure 4A] A GUI for selecting the model used by the chromatography simulation software is shown. [Figure 4B] The types of each model are further shown. [Figure 4C] The types of each model are further shown. [Figure 4D] A GUI for inputting data into the chromatography simulation software is shown. [Figure 4E] A GUI for inputting data into the chromatography simulation software is shown. [Figure 5A] 1 is an elution plot obtained for a first protein solution. [Figure 5B] 1 shows predicted results for the yield and purity of a target protein in a first protein solution and the combined purity-yield-objective parameter given a number of different candidate elution salt concentrations. [Figure 5C] 1 shows a plot of predicted results of the combined purity / yield-objective parameter of a target protein in a first protein solution given a number of different candidate elution salt concentrations. [Figure 6A] 1 is an elution plot obtained for a second protein solution. [Figure 6B] 1 shows predicted results for yield and purity of a first and second target protein in a second protein solution and for combined purity-yield-objective parameters given a number of different candidate elution salt concentrations. [Figure 6C] 13 shows a plot of predicted results of the combined purity / yield-objective parameter of a first and second target protein in a second protein solution given a number of different candidate elution salt concentrations. [Figure 6D] A comparison of predicted and measured protein ratios is shown. [Figure 6E] The resolution of proteins P1 and P2 is shown. [Figure 7]An example of two target proteins P1, P2 and a further protein P3 is shown. [Figure 8] Figure 8A is a plot showing the effect of protein loading on relative pool composition, and Figure 8B is a plot showing the effect of elution buffer salt concentration on relative pool composition. [Figure 9] 13 is a plot illustrating compensation of the effect of protein loading on target protein ratio by predicted elution buffer salt concentration. [Figure 10A] The effect of different salt concentrations on the elution profiles of two proteins is shown. [Figure 10B] The effect of different salt concentrations on the elution profiles of two proteins is shown. [Figure 10C] The effect of different salt concentrations on the elution profiles of two proteins is shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0134] FIG. 1 shows a flow chart of a chromatography method according to an embodiment of the present invention.

[0135] The method can be used to obtain a target elution volume that contains two or more target proteins with desired absolute or relative amounts and / or purities.For example, the method can be carried out to obtain target protein P1 and target protein P2 at a certain concentration or amount, respectively, and target protein P1 preferably has a minimum purity within the target elution volume.In addition or alternatively, the method can be used to obtain a target elution volume that contains two or more target proteins with a desired ratio.

[0136] In a first step 100, the method involves providing a cation exchange chromatography column 230, for example as shown in Figures 2 and 3. There are a number of commercially available chromatography resin materials that can be used to manufacture self-packed and pre-packed columns.

[0137] For example, Poros™ XS (Thermo Scientific™) can be used as the chromatography resin material in a self-packed column, such as a chromatography column purchased from KronLab. Alternatively, a pre-packed column, such as Repligen®, purchased from KronLab, can be used (e.g., having dimensions of length=5 cm and diameter=0.5 cm, or length=10 cm and diameter=0.5 cm). For large-scale runs, the column dimensions can be significantly different from those for smaller laboratory-scale runs. For example, the column dimensions of a column used for large-scale runs can be length=22.5 cm and diameter=25 cm.

[0138] The chromatography column can be placed in an automated or semi-automated chromatography system, such as the AKTA Avant 25 and 150 chromatography systems (GE Healthcare) equipped with a sample pump and internal fractionation.

[0139] Next, a protein solution containing two or more target proteins and, optionally, one or more non-target proteins (considered as impurities) is applied to the column in step 102. The protein solution contains a first target protein P1, a second target protein P2, and, optionally, one or more further proteins that are non-target proteins P3, P4, P5, P6.

[0140] One or more pretreatment and pre-purification steps are performed to create a protein solution that is applied to a chromatography column. First, the cell culture is harvested and the proteome of the harvested cells is obtained. The proteome contains a complex mixture of glycosylated antibodies and many other types of proteins. To selectively obtain the antibody fraction, the cell culture extract is applied to a first pre-purification column adapted to selectively bind antibodies and antibody fragments.

[0141] According to one example, the first pre-purification column contains a resin for monoclonal antibody (mAb) purification, e.g., Protein A resin, which is equilibrated before loading of the sample. The loading density of the first pre-purification column in this example and many other examples is about 23 g protein per liter of resin. The antibody fraction is eluted and the pH of the eluate is increased again after a viral inactivation step. The solution is incubated overnight at 4°C and filtered through a 0.2 μm sterile filter. The filtered Protein A eluate is used as the loading material for a second pre-purification column, e.g., a column packed with a mixed-mode chromatography resin. Mixed-mode chromatography is used to remove protein impurities. The column is equilibrated before loading. The loading density is 25 g protein per liter of resin and the flow rate is 150 cm / hr. The pH of the flow-through eluate is then lowered and the pool is filtered through a 0.2 μm sterile filter.

[0142] As a result of these pretreatment and pre-purification steps, a protein solution is obtained that is applied to the chromatography column 230. The protein solution contains only a few proteins whose chemical properties are usually well known due to the number and nature of the pre-purification steps used. The nature and number of pretreatment and pre-purification steps may depend on the type of cell culture and the cells used to provide the cell culture protein extract, and / or may depend on the protein(s) of interest. Typically, a series of multiple purification steps also does not fully purify the target protein(s), especially when the target protein(s) are glycoforms or other types of PTM-based protein variants whose chemical properties are highly similar to one or more other proteins in the cell culture extract.

[0143] Analytical tests such as mass spectrometry may also be used to determine the number as well as the amount of each protein species contained in the protein solution applied to column 230 .

[0144] Then, in step 104 , the amount of each protein in the protein solution that has been applied or will be applied to the column is entered into chromatography simulation software 204 .

[0145] Additionally, in step 106, the desired first amounts 214 of the first and second target proteins, and optionally, the respective amounts of non-target proteins, if present, are also entered by the user into the chromatography simulation software. Additionally, several other values ​​may be entered into the chromatography simulation software via a user interface or provided via a configuration file or other means. For example, these other values ​​may include the desired optimization criteria (amount and / or purity of the first and / or second target proteins), the pH value of the elution buffer, the flow rate of the elution buffer specified by the pump used to pump the elution buffer to the column, the chemical properties of each protein in the protein solution, the amounts of individual proteins in the protein solution already applied or to be applied to the column, and the dimensions of the column. The number and nature of the parameters may depend on the model used by the simulation software to simulate the chromatography runs at a number of different elution buffer salt concentrations.

[0146] Several input parameters, such as the amount of protein, column dimensions, pH value of the elution buffer, porosity of the stationary phase, flow rate of the elution buffer, etc., can be stored in a configuration file or entered via the GUI and adapted to the respective application scenario and protein solution. The calibrated model parameters (e.g., isotherms, N, ΔGp 0 , ΔGs 0 Some parameters such as the molecular weight, molecular weight (molecular weight), pKa, and ligand density can be reused and are usually stored in a configuration file.

[0147] This allows the chromatography simulation software to simulate concentration and / or purity values ​​of one or more target proteins as a function of salt concentrations in many different candidate elution buffers, for one or more different candidate elution volumes, based specifically on the amount and chemical properties of the protein to be loaded onto the column.

[0148] For example, the salt in the elution buffer for which the concentration is calculated may be a sodium salt, such as sodium acetate.

[0149] The recommended pH value of the elution buffer may be derived, for example, from the literature or from preliminary empirical tests. In protein chromatography in general, a pH range of the elution buffer between pH 2.5 and 10 is usual. For cation exchange chromatography, a pH range of the elution buffer between pH 4 and 9 is usual. Preferably, the pH value entered into the simulation software is a pH value known to work well for performing chromatographic separations or analyses of similar proteins similar to those currently separated. In the example described here, the pH values ​​entered into the chromatography simulation software, which are also the pH values ​​of the elution buffers actually used, are in the range of 5.30-5.70.

[0150] Typically, the pH value is not changed during the simulation. The pH value of the elution buffer actually used can be controlled, for example, using a pH meter, for example a WTW Sentix Mic probe.

[0151] The chemical properties of a protein can include properties independent of any predictive model used to simulate the chromatographic process for calculating the salt concentration of an appropriate elution buffer (e.g., pK or the number and nature of amino acid moieties).

[0152] The chemical properties of the protein may further include properties specific to the predictive model used to simulate the chromatographic process.

[0153] Chromatographic data (chromatograms) and corresponding off-line analytical data (eg, HPLC data) can be used to obtain model-specific chemical properties of the protein.

[0154] For example, different chromatography modeling workflows have been described in the literature to obtain specific characteristic values ​​for the chromatography model of a protein. When working at high protein concentrations in the nonlinear range of the isotherm, different models may affect the chromatogram in the same way. For example, the position of the peak maximum changes depending on the shielding coefficient, ν, or Keq. To estimate the parameters in the linear range of the isotherm, the Yamamoto equation can be applied, as described, for example, in M. Rudt, F. Gillet, S. Heege, J. Hitzler, B. Kalbfuss, B. Guelat, "Combined Yamamoto approach for simultaneous estimation of adsorption isotherm and kinetic parameters in ion-exchange chromatography", Journal of Chromatography A, 1413 (2015) 68-76 and S. Yamamoto, K. Nakanishi, R. Matsuno, T. Kamikuno, "Ion-exchange chromatography of proteins - Prediction of Elution Curves and Operating Conditions", I. Theoretic Considerations, Biotechnology and Bioengineering, 25 (1983) 1465-1483.

[0155] According to an embodiment, the experiments within the linear range of the isotherm are evaluated in one or more of the following three ways: 1) parameter estimation by curve fitting of the chromatogram (T. Hahn, T. Huuk, V. Heuveline, J. Hubbuch, Simulating and Optimizing Preparative Protein Chromatography with ChromX, Journal of Chemical Education, 92 (2015) 1497-1502); 2) logarithmic graph evaluation using the log(GH) / log(c(Na+)) plot described by Yamamoto et al. (T. Ishihara, T. Kadoya, H. Yoshida, T. Tamada, S. Yamamoto, Rational methods for predicting human monoclonal antibody retention in protein A affinity chromatography and cation exchange chromatography: Structure-based chromatography design for monoclonal antibodies, Journal of Chromatography A, 1093 (2005) 126-138), and 3) non-logarithmic graphical evaluation of the GH(c(Na+)) plot (M. Schmidt, M. Hafner, C. Frech, Modeling of salt and pH gradient elution in ion-exchange chromatography, Journal of Separation Science, 37 (2014) 5-13). If no good correlation is observed in this respect, the selected model may not be suitable for the application.

[0156] The next step is to add experiments with high protein, and the results of the evaluation of the linear gradient data can be used to improve the parameter estimation. If the curve fitting of the chromatograms results in a good correlation between the model and the experimental data, the parameter set can be tested by performing validation runs with combinations of important process parameters that are not included in the calibration data set. If no satisfactory model is found, the model equation needs to be expanded or the data set needs to be reduced to a smaller design space. In the following, the acquisition of some model-specific chemical properties of proteins according to embodiments of the present invention is described.

[0157] According to some embodiments, the Yamamoto evaluation was performed to obtain model-specific protein properties. For example, a graphical evaluation of log(GH) / log(c(Na+)) of linear gradient elution experiments was performed as described elsewhere (M. Schmidt, M. Hafner, C. Frech, Modeling of salt and pH gradient elution in ion-exchange chromatography, Journal of Separation Science, 37(2014)5-13; T. Ishihara, T. Kadoya, H. Yoshida, T. Tamada, S. Yamamoto, Rational methods for predicting human monoclonal antibody retention in protein A affinity chromatography and cation exchange chromatography:Structure-based chromatography design for monoclonal antibodies, Journal of Chromatography A, 1093(2005)126-138). The slope of the normalized gradient of the salt gradient, GH, can be calculated as follows: TIFF2025029052000021.tif11130

[0158] VG is the volume of the gradient, V C is the column volume, ε col is the pore column porosity, ε p is the bead porosity, and k D is the exclusion coefficient. K D was assumed to be 0.6 (F. Wittkopp, L. Peeck, M. Hafner, C. Frech, Modeling and simulation of protein elution in linear pH and salt gradients on weak, strong and mixed cation exchange resins applying an extended Donnan ion exchange model, J Chromatogr A, 1545 (2018) 32-47). The number of binding sites is log(GH) / log(c(Na + )) and subtract 1 from the slope (M. Schmidt, M. Hafner, C. Frech, Modeling of salt and pH gradient elution in ion-exchange chromatography, Journal of Separation Science, 37(2014)5-13). The equilibrium constant can be calculated using the same graph for monovalent salts. TIFF2025029052000022.tif12128

[0159] The Gibbs free energies of protein and salt (ΔG 0 P / RT and ΔG 0 s / RT) is calculated according to the following formula: eq The slope and y-intercept can be determined by plotting v vs. TIFF2025029052000023.tif6128

[0160] All calculations were performed in Microsoft Excel applying the software's internal linear regression function.

[0161] According to an embodiment, in a further step, GH / c(Na+) curve fitting was performed. The GH / c(Na+) curve was performed by calculating the differential equation according to previous publications (S. Kluters, F. Wittkopp, M. Johnck, C. Frech, Application of linear pH gradients for the modeling of ion exchange chromatography: Separation of monoclonal antibody monomer from aggregates, J Sep Sci, 39 (2016) 663-675; and M. Schmidt, M. Hafner, C. Frech, Modeling of salt and pH gradient elution in ion-exchange chromatography, Journal of Separation Science, 37 (2014) 5-13): TIFF2025029052000024.tif13128 formula, GH salt is the slope of the normalized salt gradient, and c SALT is the concentration of free (unbound) salt ions in the mobile phase of the column, and c SALT、elu is the concentration of salt in the elution buffer at the elution peak maximum, and q PROT_i is the concentration of bound protein i in the stationary phase, and K eq_i is the equilibrium constant (i.e., the molar concentration of a particular protein i in the stationary phase divided by the molar concentration of protein i in the mobile phase), Λ is the ligand density (defined as the number of ligands in mol per column volume), and ν i is the number of binding sites (i.e., the number of binding sites of protein i involved in protein binding to the stationary phase), and σ i is the shielding coefficient.

[0162] The above calculations can be performed, for example, in the Berkeley Madonna® software using, for example, the software's internal fourth-order Runge-Kutta algorithm and 0.01 mol / L Na+ This can be implemented by applying time stepping of the eigenvalues. Parameter estimates were chosen for the pre-determined parameters of the Yamamoto evaluation and allowed to vary until no further improvement was achieved.

[0163] According to the embodiment, in the next step, chromatogram curve fitting was performed. This step can be performed, for example, with ChromX from GoSilico. A lumped velocity transport dispersion model was applied, for example with a linear SUPG spatial discretization of 30 cells. The initial time stepping was set to 0.2 seconds using the IDAS function. For global optimization, the ASA algorithm with software standard values ​​was used. For deterministic optimization, the IOPT algorithm was used. The fractionation data were imported as a percentage of the absolute UV signal measured by the AKTA system. The time correspondence between the experimental data and the simulation was confirmed by correlating the experimental and simulated data of the salt. The limits of the parameter estimation were selected based on previous results and can be fitted when the optimization function calculated values ​​close to the boundaries. The model parameters were determined using the estimation function of the software. The pooled analytical measurements were considered using the optimization function with each pooling criterion. The influence of the different AKTA systems on the dispersion was considered by adding a continuous stirred tank reactor in front of the columns with different lengths. Latin hypercube sampling was performed with a population size of 1000 using the sampling function. All calculations can also be performed in other software packages, such as CADET.

[0164] In other cases, model-specific protein parameters can be taken directly from the literature and stored in the configuration file of e.g. the chromatography simulation software.

[0165] After providing all input parameters to the chromatography simulation software, e.g. via a GUI and / or via a configuration file or via a further data source, the chromatography simulation software calculates, in step 108, the salt concentration of an elution buffer adapted to elute the first target protein from the chromatography column.

[0166] In particular, according to embodiments of the invention, calculating the elution buffer salt concentration may include calculating the amount and / or purity of the one or more target proteins as a function for each of a plurality of candidate elution buffer salt concentrations, outputting the calculated amount and / or purity of the one or more target proteins for each of the candidate elution salt concentrations in association with the respective candidate elution salt concentration, and receiving from a user or from a software functionality a selection of one of the candidate elution salt concentrations in which the amount and / or purity of the one or more target proteins best matches the desired amount and / or purity of the one or more target proteins input in step 106.

[0167] The chromatography simulation software is configured to calculate the salt concentration as a function of a set of parameter values ​​including at least the following: a given pH value of the elution buffer, the dimensions of the cation exchange chromatography column provided, the load of the applied protein (P1-P6), and the chemical properties of the proteins of the applied protein solution. As mentioned above, these parameter values ​​can be provided via the GUI and / or via a configuration file or via other data sources.

[0168] After calculating the (most) suitable elution salt concentration to obtain an elution target volume containing the first and / or second target protein in a desired absolute or relative amount and / or purity, the chromatography simulation software calculates in step 110 pooling boundaries of target elution volumes containing one or more target proteins in a desired amount and / or desired purity as a function of at least the calculated salt concentration and the input desired amount of the first target protein. Preferably, the chromatography simulation program uses one or more chromatography models when calculating the optimal elution salt concentration based on the amount and purity of protein obtained for many different candidate elution salt concentrations and / or when calculating the target elution volume based on the elution salt concentration obtained in step 108.

[0169] Next, in step 112, a pump included in the automated or semi-automated chromatography system applies an elution buffer whose pH value is the same as the pH value entered into the chromatography simulation software and whose salt concentration is the same as the salt concentration calculated in step 108.

[0170] According to one example, "Elution Buffer A" is prepared according to the calculated salt concentration, the buffer consisting of 0.04 mol / L Na-acetate.

[0171] According to another example, "elution buffer B" is prepared according to the calculated salt concentration, the buffer consists of 1 mol / L Na acetate solution. The buffer can be prepared, for example, with 30% acetic acid (Merck Chemicals GmbH) and Na acetate*3H20 (Merck Chemicals GmbH), with pH values ​​of 5.30 and 5.70.

[0172] Chromatography is performed in step 114 by continuously applying elution buffer to the column and pumping it through the column at a rate used to predict the pooling boundary by the chromatography simulation software. Often, the volume of elution buffer that needs to be applied to reach the starting pooling boundary is many times the volume of the chromatography column.

[0173] Then, in step 116, a human user or an automated or semi-automated chromatography system collects the calculated target elution volume as a separate fraction using the pooling boundaries calculated in step 110. For example, a robotic arm holding an empty container can be configured to move the container under the outlet of the column to begin collecting eluate exiting the column at a start pooling boundary of the calculated target elution volume, and to remove the container from the outlet such that collection of eluate stops when a stop pooling boundary is reached. The eluate collected in the container is a target elution volume containing one or more target proteins in a desired amount and / or purity.

[0174] According to some embodiments, steps 104-110 are performed before the protein solution is actually applied to the column. In this case, it is possible to perform one or more optional steps. For example, the column 230 can be equilibrated with the elution buffer used in step 112 until the pH and conductivity readings are stable before the protein solution is actually applied to the sample. For example, equilibration of the column may require 3-5 column volumes of elution buffer applied for equilibration purposes before the actual protein solution is applied.

[0175] FIG. 2 shows a chromatography system including a computer system 200 with a chromatography simulation software 204 and a cation chromatography column 230. In the embodiment shown in FIG. 2, the chromatography system is used to selectively obtain a first target protein P1 and a second target protein P2. The target proteins are included in a protein solution 202 that includes additional proteins P3-P6, which are non-target proteins in the application scenario shown in this figure. Before applying the protein solution to the column, the concentration of each protein P1-P6 in the protein solution 202 is determined. Furthermore, some chemical properties of the individual proteins P1-P6 are derived from the literature or determined experimentally.

[0176] The chromatography simulation program 204 includes a GUI 212 that allows a user to specify desired optimization criteria, e.g., the amount and / or the desired purity of the first and second target proteins P1, P2, or derivatives of the amount and purity, e.g., via input fields 214. Furthermore, the GUI comprises one or more additional data input fields 215, 216 that allow a user to specify the nature and amount of each of the proteins P2-P6 contained in the protein solution 202, and to specify additional parameter values, such as the dimensions of the column 230 as well as the pH of the elution buffer, the chemical properties of each of the proteins P1-P6, the type of predictive model to be used to simulate the chromatography.

[0177] The chromatography simulation software may include a number of predictive models 206-210 configured to model one or more aspects of a chromatography process. Some of the parameters used by the simulation software 204 or the chromatography simulation models 206-210 may be specified in a configuration file 220.

[0178] The chromatography simulation software 204 is configured to predict the most suitable salt concentration 222 of the elution buffer to obtain the target protein P1, P2 in the target elution volume with the desired amount and / or purity. For example, the simulation software 204 can be configured to simulate the expected elution curve of each protein contained in the protein solution for each of a plurality of different candidate elution buffer salt concentrations, which implicitly provides the amount and purity profile of each protein during the entire elution process. The simulation can be based on a combination of a plurality of different models, each describing a different aspect of the chromatography process. Based on these simulations, the user or the software function can select one of the candidate elution buffer salt concentrations that is most suitable to elute the target protein P1 with the desired amount and / or purity.

[0179] According to one embodiment, the calculated elution salt concentration 222 is output to a user via the GUI 212, 218, who creates an elution buffer 226 having the pH inputted into the simulation software and the salt concentration 222 outputted by the software 204. Additionally, the GUI can output a pooling boundary 224 of a target elution volume predicted to contain the target protein in a desired amount and / or purity.

[0180] The elution buffer 226 is particularly suitable for separating the target proteins P1, P2 from all other proteins, since the salt concentration of the elution buffer was specifically calculated for the respective amounts of proteins P1-P6 applied to the column. The elution buffer is therefore not a "standard" / "one-size-fits-all" elution buffer, but rather an elution buffer specifically adapted to the type and amount of proteins contained in the protein solution. The applicant has observed that by adapting the elution salt concentration, in each individual case, to the nature and amount of proteins applied to the column (which may also depend on various aspects of the cell culture extract and the preliminary purification procedure), the capacity of the chromatography column 232 to separate different proteins can be significantly increased.

[0181] The user first applies the protein solution 202 to the column 230, then the elution buffer 226. The elution buffer 226 and the protein solution 202 percolate through the column 230. The molecules of the various proteins P1 to P6 thereby interact with the stationary phase 228 of the column 230 according to their chemical properties. The interaction of the protein molecules with the stationary phase and the elution buffer determines their elution profile, i.e. the amount of each protein that leaves the column at a given time. By changing the containers 232, 234, 236 used to collect the eluate that leaves the outlet 238 of the column, different fractions 238, 240 of the eluate can be obtained. According to a preferred embodiment, the exchange of containers is coordinated such that the target elution volume predicted to contain the target protein in the desired amount and / or purity is collected in the same container. This means that the pooling boundaries of the predicted target elution volume determine the time at which the collection is placed below the outlet of the column and removed from below.

[0182] The volume of the protein solution is much smaller than the volume of the applied elution buffer and is therefore neglected in many chromatography simulation models.

[0183] In some embodiments, the elution buffer 226 is actually a series of two or more different elution buffers with the same pH value but different salt concentrations. These different elution buffers are also referred to as "elution steps." For example, chromatography simulation software can be configured to calculate the most suitable salt concentrations for the different elution steps.

[0184] A further chromatography system is shown in Figure 3. The chromatography system shown in Figure 3 can be the same as that shown in Figure 2, where the software 204 is used not to obtain a desired amount of a single target protein, but rather to obtain a desired ratio of two target proteins P1, P2.

[0185] The components and features of the system shown in Figure 3 essentially correspond to the components and features of the system shown in Figure 2. Therefore, the explanations provided with respect to these features in the description of Figure 2 will not be repeated.

[0186] In contrast to the GUI shown in FIG. 2, the GUI 212 generated by the simulation software 204 and shown in FIG. 3 allows the user to specify the desired relative amounts of the two target proteins P1, P2 via a data entry field 304. Optionally, the user may specify a minimum purity level required for each of the target proteins P1, P2. Calculation of the most suitable elution buffer salt concentration 222 to provide a target elution volume containing the two target proteins in the desired ratio and optionally the desired purity includes simulating a number of chromatographic processes based on the salt concentrations of a number of different candidate elution buffers, determining the relative amounts and optionally also the purity of the two target proteins in the target elution volume, and outputting one of the candidate elution buffer salt concentrations that is most suitable to provide a target elution volume containing P1 and P2 in the desired relative amounts and purity. Each simulation may involve predicting the elution profile of each of proteins P1-P6 based on the salt concentration of a given candidate elution buffer, and then analyzing the elution profile of each of proteins P1-P6 to determine pooling boundaries of target elution volumes whose protein content meets user-defined requirements in terms of relative target protein amount and purity.

[0187] Figure 4A shows a GUI 400 that allows a user to select a number of different chromatography process simulation models for different aspects of a chromatography process, and Figure 4B shows the aspects of the chromatography process simulated by each model.

[0188] For example, the GUI allows the user to select the model "transport-dispersed" from multiple alternative column models. A "column model," as used herein, is a model that describes the interrelationships of the concentration of each protein in the elution buffer, the salt concentration, and the pH value in the void volume of the column.

[0189] The GUI 400 further allows the user to select a model "concentration rate" from a number of alternative pore models. "Pore model," as used herein, describes the interrelationship of the concentration of each protein in the elution buffer, the salt concentration, and the pH value in the pore volume of the stationary phase 228 of the column.

[0190] The GUI 400 further allows the user to select the model "IEC2015" from a number of alternative reaction models, also referred to as "isotherm models." A "reaction model," as used herein, describes the interrelationships between at least some of the concentrations of each protein in the stationary phase 228, the salt concentration of the elution buffer, and the chemical properties of each protein in the protein solution.

[0191] The various models can be derived from the literature and / or can be based on experimentally determined protein-specific model parameters.

[0192] Figure 4C illustrates in more detail the mathematical concepts behind each model type that may be selected via GUI 400 according to one embodiment of the present invention. However, alternative models have been published, and it is possible that at least some of the models shown in Figure 4B may be adapted and improved in the future. Thus, the present invention is not limited to any particular models or formulas that are described herein for illustrative purposes.

[0193] FIG. 4D shows a graphical user interface 402 generated by the chromatography simulation software 204 that allows a user to specify details of the chromatography column 230 and the stationary phase 228 contained therein. For example, the user may specify the length and volume of the column 230, as well as the interstitial column porosity, axial dispersion, bead radius and bead porosity of the stationary phase 228. A further graphical user interface 404 generated by the chromatography simulation software 204 allows the user to specify model-specific chemical properties, such as the ionic capacity of the stationary phase (corresponding to ligand density), acid dissociation constants of carboxyl groups, histidine groups, amino groups, etc. The data entry fields of the GUI may depend on the type of model selected by the user via the GUI 400.

[0194] FIG. 4E shows a GUI 406 that allows a user to specify the salt type of the elution buffer, the pH value of the elution buffer, the number and type of proteins contained in the protein solution, as well as the desired absolute or relative amounts specific to one or more proteins, the desired purity level, and model-specific or model-independent chemical property values.

[0195] 5A is an elution plot 500 showing a simulated elution profile of multiple proteins P1, P2, and P3 contained in a protein solution applied to a column. Proteins P1, P2, and P3 can be, for example, proteins P1, P2, and P3 described in more detail with reference to FIGS. 7A-7D.

[0196] Protein P1 is provided as the target protein, and the user specifies that P1 should be obtained at a given concentration and at a given purity. The other proteins P2 and P3 are non-target proteins. This means that P2 and P3 are considered as unwanted contaminants and their concentrations must be as low as possible to ensure sufficient purity of P1.

[0197] Plot 500 includes a curve 502 showing the candidate elution salt concentrations used in the simulation that provided plot 500. As can be derived from the plot, the candidate elution buffer is actually a series of multiple different candidate elution buffers with different salt concentrations ("elution steps"). As soon as the salt concentration of the elution buffer increases above 0.35 mol / l, the target protein P1 and the non-target protein P2 start to leave the column and become observable in the eluate. This can be derived from the elution profile 508 for protein P1 and the elution profile 506 for protein P2, respectively. When the salt concentration increases above 0.9 mol / l, P3 is also eluted, as can be derived from the elution profile 504.

[0198] The chromatography simulation software is configured to simulate (predict) the elution profiles 504-508 for each of the candidate elution salt concentrations at each step and determine a pooling boundary that includes an elution volume whose protein content meets the user-defined requirements for the amount and purity of the target protein. The determined pooling boundary is indicated by the dotted lines 512, 514. The two lines define a target elution volume that includes a peak of protein P1. The target elution volume also contains some amount of protein P2. However, by stopping the collection of eluate at the pooling boundary 514, one can ensure that the purity of P1 in the target elution volume meets the user-defined purity requirement, because after this point, the relative amount of non-target protein P2 contained in the target elution volume increases.

[0199] The plot shown in Figure 5A was described with reference to the optimization criteria for a single target protein P1. However, in other use case scenarios, protein P2 can also be considered as a target protein (see, for example, Figure 6A), and the optimization criteria can be criteria related to the amount and / or purity of the two target proteins P1, P2, while only P3 is considered as a contaminant. In this case, the pooling boundaries are adapted to the new optimization criteria.

[0200] FIG. 5B shows the predicted results of the yield and purity of target protein P1 in a first protein solution given multiple different candidate elution salt concentrations and the combined purity-yield-objective parameters. "Purity" indicates the predicted purity of P1 in the target elution volume given the candidate salt concentration. "Yield" indicates the predicted amount of target protein P1 in the target elution volume given the candidate salt concentration. "Objective" is an aggregate value derived from the combination of the predicted amount of P1 and the purity. For example, the "Objective" shown in FIG. 5B is a "combined objective" calculated as follows:

[0201] Single Objective SO=F×(Objective)Power+A, where F (the "coefficient") is a number, POWER is a number, and A (the "addend") is also a number.

[0202] Compound objectives can be generated by combining single objectives by multiplication or addition.

[0203] For example, the following parameters may be selected: POWER = 1, F = 1 and A = 0. However, each of these parameters may have different values.

[0204] Objective SO1, calculated in iteration 10 for an elution buffer salt concentration of 0.342544, is the predicted purity of protein P1 in the target elution volume, with an elution pool volume starting at 135 mL and ending at 147 mL. The predicted purity SO1 has a value of 0.751832. Objective SO2, calculated in iteration 10 for an elution buffer salt concentration of 0.342544, is the predicted yield of protein P1. The predicted yield SO2 has a value of 0.836734.

[0205] The mixed-purpose CO1 can be calculated by aggregating SO1 and SO2, for example by calculating SO1+SO2. According to another example, the mixed-purpose CO2 is calculated as SO1×SO2.

[0206] The results are presented in the form of a table 510. The first column of the table ("Iteration") contains an identifier of the salt concentration of the candidate elution buffer or a set of salt concentrations of the candidate elution buffer (in case of multiple steps) used as the basis for the respective chromatographic simulation. A further column ("Salt Concentration") indicates the salt concentration of the candidate elution buffer used in the particular elution step. A further column ("Purity Protein 1") indicates the predicted purity value for a given salt concentration of the candidate elution buffer. A further column ("Yield Protein 1") indicates the predicted amount (e.g. concentration) of protein P1 for a given salt concentration of the candidate elution buffer. A further column ("Objective") indicates the objective to be optimized (here, minimized) and calculated as a function of the predicted purity and predicted yield of the target protein P1.

[0207] Each row of table 510 corresponds to one candidate elution buffer salt concentration and the respective prediction. The rows of table 510 are sorted according to the objective value, and the row with the optimal (here, the smallest) objective, i.e. with the highest purity and yield, appears as the first row of the table. In the example shown in this figure, the lowest possible value for "purity" or "yield" means that the purity or yield is as high as possible. This means that the salt concentration of the candidate elution buffer of the simulation with "iteration" number 23, 0.345, is considered to be the optimal salt concentration for the given protein solution. As a result, the elution buffer actually used for the "real" elution is selected to have a salt concentration of 0.345. There are also simulations with some single values ​​that are higher, for example iteration 10. Here, the purity is higher than in iteration 23, but the yield is lower. The best compromise between purity and yield (Pareto principle) is iteration 23.

[0208] FIG. 5C shows a plot 518 of the predicted results of the combined purity / yield-objective parameters of the target protein in the first protein solution given a number of different candidate elution salt concentrations. Curve 520 represents the objective derived from the purity / yield described with reference to FIG. 5B. Thus, plot 518 of FIG. 5C shows the purity-yield objective obtained based on a number of different candidate elution buffer salt concentrations and their respective simulations in the form of a curve rather than a column as in table 510. Data point 516 indicates that the minimum value of this objective was obtained in iteration 23, and that the salt concentration of the candidate elution buffer used for this particular simulation #23 should be used as the actual elution buffer salt concentration to obtain the target elution volume. The salt concentration of the elution buffer for the elution buffer step corresponding to the elution target volume bounded by the pooling boundaries 512, 514 in plot 5A corresponds to the optimal salt concentration predicted in the simulation having the number #23.

[0209] Figure 6A is an elution plot 600 obtained for a second protein solution. The second protein solution also contains three proteins P1, P2 and P3. Proteins P1, P2 and P3 can be, for example, proteins P1, P2 and P3 described in more detail with reference to Figures 7A-7D.

[0210] In contrast to the situation described in Figure 5A, both proteins P1 and P2 are considered to be target proteins. The user specifies via the graphical user interface that they are interested in obtaining proteins P1 and P2 in a ratio of 55:45. Furthermore, the resulting target elution volume should be as pure as possible, which means that the target elution volume should contain as little as possible of P3 and any other proteins.

[0211] The simulated elution profile of protein P1 is shown as curve 608, the simulated elution profile of protein P2 is shown as curve 606, the simulated elution profile of protein P3 is shown as curve 604, and the elution buffer salt concentration steps are shown as curve 602. The objective simulation, derived from the amounts and purities of the two target proteins P1 and P2, and the amounts and purities for each of a plurality of different candidate elution buffer concentrations, reveals optimal elution buffer salt concentrations for each of a plurality of steps, and reveals pooling boundaries 612, 614 that define a target elution volume that contains the first and second target proteins P1, P2 in the desired relative amounts and meets the specified purity requirements. Additionally, the plot 600 includes a simulated total signal 610 that indicates the total amount of protein that can be used to compare the predicted protein peaks to the total protein peaks empirically obtained during the elution process.

[0212] 6B shows the predicted results of the yield and purity of the target proteins P1 and P2 in the second protein solution and the combined purity-yield-objective parameter when multiple different candidate elution salt concentrations are given. "Purity" indicates the predicted purity of each target protein P1, P2 in the target elution volume for each candidate elution buffer salt concentration.

[0213] The results are presented in the form of a table 617. The first column of the table ("Iteration") contains an identifier of the salt concentration of the candidate elution buffer or a set of salt concentrations of the candidate elution buffer (in case of multiple steps) used as the basis for the respective chromatographic simulation. A further column ("Salt Concentration") indicates the salt concentration of the candidate elution buffer used in the particular elution step. Further columns ("Purity Protein 1", "Purity Protein 2") indicate the predicted purity values ​​for each of the proteins P1, P2 for the given salt concentration of the candidate elution buffer. A further column ("Objective") indicates the objective to be optimized (here, minimized) and calculated as a function of the predicted purity and predicted relative amount of the target proteins P1 and P2.

[0214] The "objective" (also called "composite objective") may be, for example, an aggregate value derived from a combination of the predicted relative amounts of P1 and P2 and the desired purity levels of P1 and P2. For example, a "target" CO may be calculated from two single targets SO1, SO2 by adding or multiplying the two single targets. For example, CO may be calculated as CO=SO1+SO2 or CO=SO1×SO2.

[0215] Thereby, SO1 is the predicted purity of P1 in the target elution volume for a given candidate elution buffer salt concentration, e.g., 1:0.55, with the elution target volume predicted to start at 112 mL and end at 127 mL. SO2 is 2:0.45, which is the desired purity of P2 in the elution target volume, with the elution target volume predicted to start at 112 mL and end at 127 mL.

[0216] Optionally, the error between the calculated aim CO and the target value can be calculated using a "reference comparison" cost function that calculates the point-wise squared deviation between the aim and target values.

[0217] Each row of table 617 corresponds to one candidate elution buffer salt concentration and the respective prediction. The rows of table 617 are sorted according to the objective value, and the row with the optimal (here, the smallest) objective, i.e. with the highest purity and the best matching P1:P2 ratio, appears as the first row of the table. The smallest possible value of the "objective" parameter means that the difference between the single objective and the corresponding target value is as small as possible. This means that the salt concentration of the candidate elution buffer of the simulation with "iteration" number 35, 0.388, is considered to be the optimal salt concentration for the given protein solution. As a result, the elution buffer actually used for the "real" elution is selected to have a salt concentration of 0.388.

[0218] FIG. 6C shows a plot 618 of predicted results of the combined purity / yield-objective parameters of the target protein in the first protein solution given multiple different candidate elution salt concentrations. Curve 620 represents the objective derived from the purity / yield described with reference to FIG. 6B. Thus, plot 618 in FIG. 6C shows the resulting purity-yield objective based on multiple different candidate elution buffer salt concentrations and their respective simulations in the form of a curve rather than a column as in table 617. Data point 616 indicates that the minimum value of this objective was obtained in iteration 35, and that the candidate elution buffer salt concentration used for this particular simulation #35 should be used as the actual elution buffer salt concentration. The elution buffer salt concentration of the elution buffer step corresponding to the elution target volume bounded by the pooling boundaries 612, 614 in plot 6A corresponds to the optimal salt concentration predicted in the simulation having the number #35.

[0219] FIG. 6D shows a comparison of the predicted and measured protein ratios of P1 and P2 obtained for the second protein solution. The predicted concentration of protein P1 in the target elution volume bounded by the pooling boundaries 612, 614 is predicted to be 51% based on the concentration of the candidate elution buffer of 0.388 mol / liter used in simulation #35. When elution is performed with an elution buffer having a salt concentration of 0.388 mol / liter, the empirically measured concentration of P1 in the target elution volume is 56%. Similarly, the predicted concentration of protein P2 in the target elution volume is 49%. The empirically measured concentration of P2 in the target elution volume is 44%. Thus, the accuracy of the predicted concentrations in the target elution volume defined by the predicted pooling boundaries is very high.

[0220] FIG. 6E shows the empirically obtained resolution of proteins P1, P2 and P3. Plot 624 shows the optical peaks 626, 628, 630 induced by the elution of the respective proteins P1, P2, P3. As can be inferred from the plot, the peaks of P1 and P2 overlap significantly. It was observed that the resolution R of the two proteins P1, P2 at a protein load of 45 g / liter and a gradient of 40 column volumes is about 0.43, i.e. very low. The embodiments of the present invention can be used to obtain two or more proteins in a desired ratio, even when the two or more proteins have low resolution, without the need to first separate the two proteins and then recombine the proteins in the desired ratio.

[0221] FIG. 7 shows an example of two target proteins P1, P2 and a further protein P3 that is considered an unwanted contamination.

[0222] FIG. 7A shows a diagram of protein 700 "bsGant" consisting of a brain shuttle molecule, referred to as "BSM", linked to an antibody, referred to herein as "Gant". Brain shuttle molecules are molecules that can increase the penetration of large molecules, such as antibodies, into the brain. Access of large molecules to the brain is restricted by the blood-brain barrier (BBB), a gatekeeper between the blood and brain tissue that carefully filters the molecules that can enter the brain. Some antibodies engineered by the applicant are able to cross the blood-brain barrier by binding to one of the protein receptors located on its surface. Brain shuttle molecules can potentially transport one or more therapeutic molecules to the brain regardless of their inherent ability to cross the blood-brain barrier. The bsGant protein is described in detail in International Patent Application WO2017 / 055540, which is incorporated herein by reference in its entirety.

[0223] The "bsGant" antibody is a monomeric immunoglobulin G (IgG) Fc fusion protein. The cell culture used to produce the "bsGant" antibody produces this antibody as three different variants with different degrees of Fab glycosylation. Fab glycosylation is not complete, resulting in a mixture of non-glycosylated, mono-glycosylated and di-glycosylated antibodies.

[0224] According to an embodiment, the one or more target proteins are different glycoforms of a monomeric bsGant protein (as defined above).

[0225] According to one example, the one or more target proteins are different glycoforms of bsGant "antibody 0015" described in patent application WO2017 / 055540. This antibody is a bispecific antibody comprising a light chain having the amino acid sequence of SEQ ID NO: 01, a heavy chain having the amino acid sequence of SEQ ID NO: 02, a light chain having the amino acid sequence of SEQ ID NO: 03 and an antibody Fab fragment comprising the amino acid sequence of SEQ ID NO: 04 disclosed in WO2017 / 055540. SEQ ID NO: 01 relates to a 215 aa residue polypeptide corresponding to the non-domain-swapped light chain of the N-terminal fab, SEQ ID NO: 02 is a 455 aa residue polypeptide corresponding to the two heavy chains, SEQ ID NO: 03 is a 215 aa residue polypeptide corresponding to the domain-swapped light chain fab fragment of the additional C-terminal fab fragment and SEQ ID NO: 04 is a 229 aa residue fragment corresponding to the domain-swapped heavy chain fab fragment of the additional C-terminal fab fragment.

[0226] FIG. 7B shows a diglycosylated antibody variant, also referred to herein as "protein P1."

[0227] FIG. 7C shows a monoglycosylated antibody variant, also referred to herein as "protein P2."

[0228] FIG. 7D shows an aglycosylated antibody variant, also referred to herein as "protein P3."

[0229] The close elution conditions of the three variants pose unique challenges to the chromatographic protocols used to separate and / or further purify the desired antibody variants.

[0230] Differences in glycosylation patterns in the Fc region of antibodies can lead to differences in conformation, pharmacokinetics and binding properties. Glycosylation of the Fab region can affect ligand binding strength and tissue penetration. Therefore, glycosylation patterns are of particular interest for several antibodies used in the medical field. Glycosylation patterns may depend on the expression vector, cell line, cell culture mode, as well as cell culture conditions, purification process and other factors. Because glycosylation variants may only differ slightly in charge, separation using ion exchange resins is possible, but elution conditions need to be optimized.

[0231] Applicants have observed that the protein load density applied to the chromatography column is a further important process parameter that influences the product pool composition. Embodiments of the present invention may make it possible to obtain a desired ratio of antibody variants that are pharma- ceutical particularly effective, for example a P1:P2 ratio of 55:45.

[0232] The challenge of separating a protein solution containing three proteins P1, P2 and P3 is that the product pool ("target elution volume") should contain a ratio of P1 to P2 of approximately 55%:45%. The yield of this separation is limited due to this relative abundance criterion, since the protein solution contains more than 50% P2 (see Table 1). Protein P3 in this case is a product-specific impurity, i.e., non-glycosylated Protein 700 protein.

[0233] Applicants have observed that the relative and absolute amounts of proteins P1, P2 and P3 in the protein solution applied to the column can vary widely from case to case, which can pose an additional challenge as a chromatographic protocol that worked well with a particular protein preparation may not be able to separate proteins in a different preparation of these three protein variants.

[0234] For example, a protein solution containing the three proteins P1, P2 and P3 selectively can be obtained by harvesting a cell culture genetically engineered to produce protein 700 in three glycosylation variants. All glycoforms were captured on a Protein A resin equilibrated before loading. The loading density of the raw protein extract was 23 g protein per liter of resin. The antibodies were eluted and after a viral inactivation step, the pH of the eluate was increased again, the solution was incubated overnight at 4 °C and filtered through a 0.2 μm sterile filter. The filtered Protein A eluate was used as the loading material for a mixed-mode chromatography resin. A mixed-mode chromatography resin was used to remove protein impurities. The column was equilibrated before loading. The loading density was 25 g protein per liter of resin and the flow rate was 150 cm / h. The pH of the flow-through eluate was then reduced and the pool was filtered through a 0.2 μm sterile filter.

[0235] This procedure was repeated multiple times with six different cell culture extracts and the resulting protein solutions for each applied to the cation column are summarized below in Table 1. TIFF2025029052000025.tif51130

[0236] To isolate single glycosylated variants, the PorosXS pool was diluted with water and reprocessed with PorosXS resin in bind and elute mode at high protein loading density. The column was equilibrated with 376 mM sodium acetate pH 5.5, the loading density was 80 g per liter of resin. The flow-through was fractionated and analyzed. The fraction at the beginning of the flow-through contained variant 1 with a purity of 98.0% (protein solution 4). To elute variants 2 and 3, a gradient from 376 mM sodium acetate pH 5.5 to 616 mM sodium acetate pH 5.5 in 6.25 CV was used. The elution peak contained variant 2 with a purity of 97.6% (protein solution 5) and variant 3 with a purity of 96.5% (protein solution 6).

[0237] Glycosylation variant analysis was performed by injecting 100 μg of sample onto an analytical cation exchange chromatography column (Mono S 5 / 50 GL, GE Healthcare) with salt gradient elution at pH 5.3 and a flow rate of 1 ml / min. Preliminary peak identification was performed by mass spectrometry.

[0238] The system and resin parameters required for mechanistic modeling of the various model parameters for models 206-210 were determined by pulse experiments with different tracers as described in A. Osberghaus, S. Hepbildikler, S. Nath, M. Haindl, E. von Lieres, J. Hubbuch, Determination of parameters for the steric mass action model-a comparison between two approaches, Journal of Chromatography A, 1233 (2012) 54-65. Latin hypercube sampling of size 1000 was performed to study the influence of the process parameters (pH value, load composition, salt concentration of the elution step) on the impurity profile in the elution pool. To vary the load density, the injection volume was varied. Simulations were performed using ChromX with a time step of 5 seconds at a flow rate of 300 cm / h, 30 axial cells and a column length of 5 cm. For each in silico experiment, the impurity profile in the elution pool and the elution profiles of the individual proteins were calculated using a pooling determination with a fixed volume. The results were analyzed in MATLAB using linear regression analysis with the impurity pool concentration as the dependent variable and the process parameters as the independent variables.

[0239] 8A shows a plot 802 illustrating the effect of protein loading on the relative protein composition in the elution target volume. Assuming an elution buffer salt concentration of 40%, the relative amount of protein P1 in the eluate at different protein loads is shown in curve 802. The relative amount of protein P2 in the eluate is shown in curve 804.

[0240] 8B shows a plot 808 illustrating the effect of the salt concentration of the elution buffer on the relative protein composition in the elution target volume. Assuming a total protein load applied to the column of 45 g / L, the relative amount of protein P1 in the eluate at different salt concentrations is shown in curve 810. The relative amount of protein P2 in the eluate is shown in curve 812. Embodiments of the present invention allow for computationally identifying salt concentrations of the elution buffer that are specifically matched to the amount and type of individual protein loaded onto the column, thereby predicting the elution profile of each protein and significantly improving the ability to collect a target elution volume containing the desired protein in the desired amount and purity.

[0241] Empirical testing has shown that chromatographic methods according to embodiments of the present invention can predict desired pool compositions and thus product qualities that can be important for molecular characterization and definition of process design space.

[0242] FIG. 9 is a plot 900 showing the compensation of the effect of protein loading by adapting the salt concentration of the elution buffer. Chromatography simulation software was used to computationally predict the salt concentration 902 of the elution buffer that could provide a target elution volume containing the two target proteins P1, P2 in the desired ratio of 55:45 for each of several increasing amounts of protein loaded onto the column (x-axis). As can be derived from the plot 900, the predicted curves 904, 906 showing the relative amounts of the two target proteins P1, P2 are essentially identical to the empirically measured relative amounts 910, 908. The predicted elution salt concentration allowed to ensure that the target elution volume contained a constant desired ratio of the two target proteins P1 and P2, regardless of the amount of protein obtained by preparing cell culture protein extracts and prior purification steps (which may differ significantly in some cases).

[0243] Applicants have observed that by simulating and selecting the salt concentration of the elution buffer according to the protein load, the simulation accuracy can be significantly improved and salt concentrations that support the optimization criteria for many types of target proteins can be identified. Applicants have observed that both protein load density and salt molarity have a strong effect on the elution profile of the protein. By combining this knowledge, the salt concentration can be used as a process steering parameter that is specifically selected for each individual protein load density. Optimization of the elution salt molarity according to the protein load density may also make it possible to ensure a constant product quality.

[0244] 10A to 10C show the effect of the salt concentration of the elution buffer on the elution profiles of two proteins, P1 and P2.

[0245] For example, a first simulation can assume a low elution buffer salt concentration, as shown in Figure 10A. The elution buffer is not able to separate proteins P1 and P2 completely and according to the desired protein ratio provided as optimization criterion. Both proteins elute in the column cleaning (CIP) step. The eluate obtained in the "pooling" stage is essentially free of proteins P1 and P2.

[0246] A second simulation can be based on a high elution buffer salt concentration, as shown in Figure 10B. This elution buffer also cannot completely separate the two proteins, nor provide an eluate containing the two proteins in the desired ratio. The eluate obtained in the pooling step contains both proteins P1 and P2 in large amounts, but not in the desired ratio (assuming the desired ratio is other than 1:1).

[0247] The third simulation can be based on a medium salt concentration of the elution buffer, as shown in Figure 10C. This elution buffer completely separates the two proteins. The eluate obtained after the pooling step contains a large amount of protein P1 and essentially no protein P2.

[0248] If the desired ratio of P1:P2 is, for example, about 1:2, the simulation software can be configured to perform further simulations based on several different elution buffer salt concentrations lower than that shown in FIG. 10B but higher than that shown in FIG. 10C to obtain predicted elution profiles of proteins P1 and P2, and to identify one of the salt concentrations that provides a pooled elution volume whose protein ratio of P1:P2 best matches the desired ratio.

Claims

1. 1. A chromatography method producing a target elution volume (238) comprising a first target protein (P1) and a second target protein (P2), the first target protein (P1) and the ... second target protein (P2) being glycosylation variants of proteins having the same amino acid sequence, the method comprising the steps of: - providing (100) a cation exchange chromatography column (230) having a stationary phase; - applying (102) to said column a protein solution (202) comprising at least said first target protein (P1) and said second target protein (P2); inputting (106) optimization criteria into a chromatography simulation software (204), the optimization criteria being desired properties of the target elution volume with respect to the first target protein and the second target protein contained in the target elution volume; - calculating (108) by means of said chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute said first target protein and said second target protein from said chromatography column in order to obtain a target elution volume that best meets said optimization criteria, said salt concentration (222) of an elution buffer comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) by said chromatography simulation software pooling boundaries of said target elution volumes as a function of at least said calculated salt concentrations and said input optimization criteria; - applying (112) an elution buffer (226) having the calculated salt concentration to the chromatography column; - carrying out elution using said applied elution buffer (114); - collecting (116) the calculated target elution volume as a separate fraction using the calculated pooling boundaries; Including, The optimization criterion is a) a desired ratio or ratio range of the amounts of said first target protein and said second target protein in said target elution volume; b) a desired amount or range of amounts of said first target protein in combination with a desired amount or range of amounts of said second target protein in said target elution volume; c) a desired ratio of the amounts of the first target protein and the second target protein in the target elution volume; ■ a desired amount or range of amounts of said second target protein in said target elution volume; The desired purity or purity range of the second target protein in the target elution volume. a desired purity or purity range of said first target protein in combination with; d) A combination of two or more of the above criteria The method is selected from the group comprising:

2. Chromatography method according to claim 1, wherein the applied protein solution (202) further comprises one or more further proteins (P3, P4, P5, P6).

3. - said calculating (108) further comprises calculating for a plurality of different elution buffer pH values ​​a salt concentration (222) of an elution buffer adapted to elute said first target protein and said second target protein from said chromatography column such that a target elution volume that best meets said optimization criteria can be obtained, said plurality of chromatographic simulations being calculated for each combination of said elution buffer salt concentration (222) and said elution buffer pH, - the pooling boundary of the target elution volume is calculated as a function of the calculated combination of at least the salt concentration of the elution buffer and the pH value of the elution buffer and of the input optimization criteria; - the elution buffer (226) applied to the column has the calculated salt concentration and the calculated pH value; The method according to claim 1 or 2.

4. The method according to any one of claims 1 to 3, wherein the first target protein and the second target protein are proteins having a resolution coefficient of less than 0.75 and / or less than 0.

5.

5. The first target protein and the second target protein are Antibody monomers having the same amino acid sequence but containing different numbers of glycosyl groups on the FAB fragment. The method according to any one of claims 1 to 4, wherein

6. 6. The method according to any one of claims 1 to 5, wherein the applied protein solution comprises each of the target proteins (P1, P2) and, if present, each of the further proteins (P3 to P6), in a respective concentration of at least 0.5 wt.-%, in particular at least 1 wt.-%, in particular at least 2 wt.-%.

7. 7. The method according to any one of claims 1 to 6, wherein the second target protein (P2) and, if present, one or more of the further proteins (P3-P6) contained in the applied protein solution have an affinity for the stationary phase of the column similar to the affinity of the first target protein for the stationary phase, resulting in overlapping elution behaviour.

8. 8. The method according to any one of claims 1 to 7, wherein the total amount of protein in the protein solution applied to the column is equal to or smaller than the maximum protein loading capacity of the column, in particular within the range of 50% to 100%, such as 50% to 90% of said maximum protein loading capacity.

9. the calculated pooling boundaries of the target elution volume are specified as an acquisition start time offset and an acquisition stop time offset; The method further comprising: - continuously monitoring the time elapsed from the start of said elution by an automated chromatography system comprising said chromatography column; - automatically starting the collection of the elution buffer by the chromatography system when the elapsed time is equal to the collection start time offset; and - stopping the collection of the elution buffer by the chromatography system when the elapsed time is equal to the collection stop time offset. Including, The method according to any one of claims 1 to 8.

10. - each of said plurality of chromatography simulations is a simulation of a chromatography process using two or more elution steps, whereby in each elution step an elution buffer with a different elution salt concentration is used; - the calculated salt concentrations of the elution buffer are a series of salt concentrations of the elution buffer specific for different elution steps, - said applying said elution buffer with said calculated salt concentration to said chromatography column comprises stepwise applying a series of elution buffers with different salt concentrations according to said series of salt concentrations specific for a calculated elution step; A chromatography method according to any one of claims 1 to 9.

11. The method further comprising: - inputting (104) into said chromatography simulation software (204) the amounts of each of said first and second target proteins (P1-P2) contained in said applied protein solution and, optionally, if present, also the amounts of each of one or more further proteins (P3-P6) contained in said applied protein solution; Including; said plurality of chromatography simulations comprising at least the dimensions of the cation exchange chromatography column provided; and the amount of said first target protein and said second target protein, and optionally the amount of said one or more further proteins, applied to said column; , calculated as a function of a set of parameter values ​​including A chromatography method according to any one of claims 1 to 10.

12. The set of parameter values: - a predefined pH value of said elution buffer, - the flow rate of the elution buffer through the column, - the chemical properties of the proteins (P1, P2, P3 to P6) of the protein solution applied The chromatography method of claim 11 further comprising:

13. the chromatography simulation software is configured to use a combination of mathematical models to calculate the plurality of chromatography simulations and / or to calculate the pooling boundaries of the target elution volume; The model is a column model (206) configured to correlate the concentrations of each of said proteins in the elution buffer in the void volume of said column (230), said salt concentration and said pH value; and a pore model (208) configured to correlate the concentration of each of said proteins in the elution buffer in the pore volume of the stationary phase of said column (230), said salt concentration and said pH value; and a reaction model (210) configured to correlate the concentration of each of said proteins in said stationary phase (228), the salt concentration of said elution buffer, and at least some of said chemical properties of each of said proteins in said protein solution; The method of claim 12, comprising:

14. 1. A chromatography control system comprising simulation software configured to perform a method for obtaining pooling boundaries for target elution volumes (238) comprising a first target protein (P1) and a second target protein (P2), the first target protein (P1) and the second target protein (P2) being glycosylation variants of a protein having an identical amino acid sequence, the method comprising: receiving (106) optimization criteria that are desired properties of the target elution volume for the first target protein and the second target protein contained in the target elution volume, and inputting the optimization criteria into the chromatography simulation software (204); - calculating (108) using said chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute said first target protein and said second target protein from a chromatography column in order to obtain a target elution volume that best meets said optimization criteria, comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) the pooling boundaries of the target elution volumes as a function of at least the calculated salt concentrations and the input optimization criteria; - outputting the calculated salt concentrations and the pooling boundaries; It is configured for The optimization criterion is a) a desired ratio or ratio range of the amounts of said first target protein and said second target protein in said target elution volume; b) a desired amount or range of amounts of said first target protein in combination with a desired amount or range of amounts of said second target protein in said target elution volume; c) a desired ratio of the amounts of the first target protein and the second target protein in the target elution volume; ■ a desired amount or range of amounts of said second target protein in said target elution volume; The desired purity or purity range of the second target protein in the target elution volume. a desired purity or purity range of said first target protein in combination with; d) A combination of two or more of the above criteria A chromatography control system selected from the group comprising:

15. The system further comprises: - receiving the dimensions of the cation exchange chromatography column (230); - receiving an amount of said first target protein and said second target protein applied to said column, and optionally an amount of one or more further proteins (105); further configured for: said simulation and / or said pooling bounds comprise at least the dimensions of the cation exchange chromatography column provided; and the amount of said first target protein and said second target protein, and optionally the amount of said one or more further proteins, applied to said column; , calculated as a function of a set of parameter values ​​including 15. The chromatography control system of claim 14.

16. - said plurality of chromatographic simulations are calculated as a function of salt concentration of said plurality of different elution buffers and as a function of pH value of said plurality of different elution buffers; 16. The chromatography control system of claim 14 or 15, wherein the method further comprises: - said calculating (108) comprises calculating a combination of an elution buffer salt concentration (222) and an elution buffer pH value adapted in combination to elute said first target protein and said second target protein from said chromatography column so as to obtain a target elution volume that best meets said optimization criteria, - calculating (110) the pooling boundary of the target elution volume is calculated as a function of the calculated combination of at least the salt concentration of the elution buffer and the pH value of the elution buffer and of the input optimization criterion; - The calculated pH value is output in addition to the calculated salt concentration of the elution buffer.

17. the chromatography control system is configured to control a buffer mixing unit to automatically generate an elution buffer having the calculated elution buffer salt concentration; and / or The chromatography control system is configured to control a buffer mixing unit to automatically generate an elution buffer having both the salt concentration and the pH value that are calculated and output in combination according to claim 16; and / or the chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations, the selected elution buffer having the outputted salt concentration; and / or The chromatography control system is configured to control an elution buffer selection unit adapted to automatically select one of a plurality of available elution buffers having different salt concentrations and different pH values, the selected elution buffer having both the salt concentration and the pH value that are calculated and output in a combined manner according to claim 16; and / or the chromatography control system is configured to control a buffer application unit configured to automatically apply an automatically generated or selected elution buffer to the chromatography column, the applied elution buffer having the output salt concentration or having both the output salt concentration and the pH value calculated in a combined manner according to claim 16; and / or the chromatography control system is configured to control an elution volume collection unit of a chromatography system to automatically collect the calculated target elution volume as separate fractions according to the calculated pooling boundaries.

17. The chromatography control system of claim 16.

18. 20. A chromatography system comprising the chromatography control system of claim 17, further comprising the buffer mixing unit and / or the elution buffer selection unit and / or the buffer application unit and / or the automated elution volume collection unit.

19. A computer program for managing a chromatography process to obtain a target elution volume (238) comprising a first target protein (P1) and a second target protein (P2), the first target protein (P1) and the ... second target protein (P2) being glycosylated variants of a protein having an identical amino acid sequence, the computer program comprising a chromatography simulation software; - receiving optimization criteria (106), which are desired characteristics of a target elution volume for a first target protein and a second target protein contained in said target elution volume; - calculating (108) using said chromatography simulation software a salt concentration (222) of an elution buffer adapted to elute said first target protein and said second target protein from a chromatography column in order to obtain a target elution volume that best meets said optimization criteria, comprising calculating a plurality of chromatography simulations as a function of a plurality of different elution buffer salt concentrations; - calculating (110) the pooling boundaries of the target elution volumes as a function of at least the calculated salt concentration and the input optimization criteria; and - outputting said calculated salt concentrations and / or said pooling boundaries to enable a user to control a chromatography system such that a target elution volume comprising said first and second target proteins is obtained according to said optimization criteria, and / or using said calculated salt concentrations and / or said pooling boundaries to automatically or semi-automatically control a chromatography system such that a target elution volume comprising said first and second target proteins is obtained according to said optimization criteria. It is configured for The optimization criterion is a) a desired ratio or ratio range of the amounts of said first target protein and said second target protein in said target elution volume; b) a desired amount or range of amounts of said first target protein in combination with a desired amount or range of amounts of said second target protein in said target elution volume; c) a desired ratio of the amounts of the first target protein and the second target protein in the target elution volume; ■ a desired amount or range of amounts of said second target protein in said target elution volume; The desired purity or purity range of the second target protein in the target elution volume. a desired purity or purity range of said first target protein in combination with; d) A combination of two or more of the above criteria A computer program product selected from the group comprising:

20. 20. The computer program of claim 19, further configured to generate control commands for automatically or semi-automatically controlling one or more units of the chromatography system.

Images