Coupling of chromatography and precipitation for adjusting acidic variant content in a monoclonal antibody pool.

The acidic charge variants (av) of monoclonal antibodies (mAb) are often reported to have reduced therapeutic potency compared with the main (mv) and basic variants (bv), therefore reduction in the av content in mAb pools is often prioritized over reduction in the bv content. In previous studies we described two different methods for reducing the av content, which were based on either ion exchange chromatography or selective precipitation in polyethylene glycol (PEG) solutions. In this study, we have developed a coupled process, in which advantages of simplicity and ease in realization of PEG-aided precipitation and high separation selectivity of anion exchange chromatography (AEX) were exploited. The design of AEX was supported by the kinetic-dispersive model, which was supplemented with the colloidal particle adsorption isotherm, whereas the precipitation process and its coupling with AEX was quantified by simple mass balance equations and underlying thermodynamic dependencies. The model was used to assess the performance of the coupling of AEX and precipitation under different operating conditions. The advantage of the coupled process over the stand-alone AEX depended on the demand for the av reduction as well as the initial variant composition of the mAb pool, e.g., the improvement in the throughput provided by the optimized sequence of AEX and PREC varied from 70 to 600% for the initial av content changed from 35 to 50% w/w, and the reduction demand changed from 30 to 60%.

Separation of non-racemic mixtures of enantiomers by achiral chromatography.

The phenomenon of partial separation of enantiomeric mixtures in achiral chromatography (ACh) has already been documented for a wide variety of chiral compounds. It is attributed to the so-called effect of self-disproportionation of enantiomers (SDE). However, quantitative description of the SDE mechanism underlying adsorption of enantiomers on achiral surfaces is still incomplete, which hinders the application of that technique for large-scale separations. In this study, a mechanistic model for description of retention behavior of SDE-phoric compounds in silica-based ACh has been developed along with a procedure for fast determination of the model parameters. The model assumes formation of associates of chiral molecules, which occurs due to homo and hetero-chiral interactions in the adsorbed phase. The ability of the model to reproduce band profiles was verified for enantiomeric mixtures of three structurally different chiral compounds.

Preferential precipitation of acidic variants from monoclonal antibody pools.

Microheterogeneity of monoclonal antibodies (mAbs) can impact their activity and stability. Formation of charge variants is considered as the most important source of the microheterogeneity. In particular, controlling the content of the acidic species is often of major importance for the production process and regulatory approval of therapeutic proteins. In this study, the preferential precipitation process was developed for reducing the content of acidic variants in mAb downstream pools. The process design was preceded by the determination of phase behavior of mAb variants in the presence of different precipitants. It was shown that the presence of polyethylene glycol (PEG) in protein solutions favored precipitation of acidic variants of mAbs. Precipitation yield was influenced by the variant composition in the mAb feed solutions, the concentration of the precipitant and the protein, and the ionic strength of the solutions. To improve yield, multistage precipitation was employed, where the precipitate was recycled to the precipitation process. The final product was a mixture of supernatants pooled together from the recycling steps. Such an approach can be potentially used either instead or in a combination with chromatography for adjusting the acidic variant content of mAbs, which can benefit in improvement in throughput and reduction in manufacturing costs.

Separation of charge variants of a monoclonal antibody by overloaded ion exchange chromatography.

A procedure for adjusting the content of charge variants of monoclonal antibody by ion exchange chromatography has been developed. The band splitting phenomenon was utilized to split the protein load into two parts, i.e., the flowthrough and bound fractions, which were either enriched or depleted with some of variants. The phenomenon was triggered by thermodynamic effects resulting from oversaturation of the resin binding sites at high column loadings as well as from kinetic effects arising from limited rates of mass transport. Cation exchange chromatography (CEX) and anion exchange chromatography (AEX) separations were examined, with the reverse order of the variant elution: acidic, main, basic in CEX, and basic, main, acidic in AEX, and the corresponding reverse enrichment tendency in the collected fractions. The separations were performed by pH gradient, whose course was simplified to two stages: isocratic loading and washing at mild pH to load and partly elute the protein, followed by a rapid pH change towards non-binding conditions to desorb the remains of the protein load. To improve yield of the operation, possibility of recycling of waste fractions was considered. To predict the process performance, a dynamic model was developed, which accounted for both adsorption kinetics and thermodynamics.

A case study of the mechanism of unfolding and aggregation of a monoclonal antibody in ion exchange chromatography.

A mechanistic model for describing unfolding of a monoclonal antibody (mAb) in ion exchange chromatography has been developed. The model reproduced retention behavior characteristic for conformational changes of antibodies upon adsorption, including: multi-peak elution, aggregate formation, and recovery reduction. Two competitive paths in the adsorption mechanism of the unfolded protein were assumed: refolding in the adsorbed phase to the native form followed by its desorption, or direct desorption followed by instantaneous aggregation in the liquid phase. The reduction in recovery of the eluted protein was attributed to spreading of the unfolded protein on the adsorbent surface, which enhanced the binding affinity. The model was formulated based on the analysis of retention behavior of a model mAb that was eluted in pH gradients on a strong cation exchange resin. The pH profile was found to be distorted in the presence of the protein, which was ascribed to dissociation of ionizable groups of the protein in the adsorbed phase. Since the protein retention was strongly pH dependent, that phenomenon was also accounted for in mathematical modeling. A series of independent experiments was designed to evaluate the model parameters that quantified the process thermodynamics and kinetics: the Henry constants of the native, unfolded, spread and aggregated forms of the protein along with underlying kinetic coefficients. The model was efficient in reproducing the retention pattern of the protein and the aggregate content in eluting band profiles. After proper calibration, the model can potentially be used to quantify protein unfolding and elution in other ion exchange systems.

Effects of negative and positive cooperative adsorption of proteins on hydrophobic interaction chromatography media.

The adsorption behavior of the model proteins: alpha-Lactalbumin, Bovine Serum Albumin, Lysozyme, and a monoclonal antibody, in single component and in binary mixtures, was investigated on two different hydrophobic interaction chromatography resins using both static and dynamic methods. A kinetic model of the adsorption process was developed, which accounted for protein unfolding and intermolecular interactions in the adsorbed phase. The latter incorporated positive cooperative interactions, resulting from preferred and multilayer adsorption on the adsorbent surface, as well as negative cooperative interactions attributed to exclusion effects due to size exclusion and repulsion. Cooperative adsorption resulted in negative or positive deviations from the Langmuir model for both single and multicomponent isotherms. The model was used to assess possible contributions of different adsorption mechanisms of proteins and their structurally different forms to the overall adsorption pattern, as well as to simulate chromatographic band profiles under different loading conditions. For proteins with unstable structure, the overall adsorption isotherm was dominated by binding of unfolded species at low surface coverage and by positive cooperative adsorption at high surface coverage. Furthermore, regardless of structural stability, exclusion effects influenced strongly adsorption equilibrium, particularly at low surface coverages. In case of chromatographic elution, i.e. under dynamic conditions, unfolding, negative cooperative adsorption, and kinetic effects governed the retention behavior and determined peak shapes, whereas the effect of positive cooperative adsorption was negligible.

Mechanism of nutrition activity of a microgranule fertilizer fortified with proteins.

BACKGROUND: A microgranule fertilizer was designed for localized fertilization of soil with controlled release of nutrients. The microgranule matrix was fortified with proteins, which were obtained from food industry byproducts or waste, i.e., whey protein from milk serum, soy protein from soy isolate and egg white protein from chicken egg white powder. The mechanism of the protein decomposition and migration of micro and macromolecule compounds through two different model soil systems was investigated. The potential of the protein fortified fertilizer for localized fertilization of the potted maize seeds was evaluated. RESULTS: The study revealed that proteins slowly diffused through soil with simultaneous degradation, which was accompanied with release of ammonia ions. The highest concentration of proteins and degradation products was found in a close vicinity of the microgranule. The microgranules were used as a local fertilizer for maize seeds in the pot experiments. The experiments confirmed statistically significant improvement in root density of maize plant compared to control group. CONCLUSIONS: Byproducts or waste of food industry, such as the milk serum and soy can be used as a source of proteins that degrade in soil without a pretreatment. The degradation is accompanied with formation of ammonium ions, which can be utilized by plants as a nitrogen source. The fertilizer microgranule should be placed in a close vicinity to the plant seed, since the maximum of the protein concentration and ammonia ions is reached at a very close distance from the microgranule.

Scale up of a chromatographic capture step for a clarified bacterial homogenate - Influence of mass transport limitation and competitive adsorption of impurities.

A model-based approach for scaling up chromatographic capture step was developed. The purification of human basic fibroblast growth factor protein 2 (FGF2) from an E. coli homogenate on a cation exchange resin was selected as a case study. Non-ideal effects accompanying the capture operation were examined, including: reduction in the protein diffusivity in the presence of the homogenate, competitive adsorption between FGF2 and undefined impurities, and flow behavior in external column volumes. The viscosity of the homogenate was measured as a function of dilution degree and shear stress, and its contribution to the diffusivity reduction was quantified. A dynamic model was formulated which accounted for underlying kinetic and thermodynamic dependencies. The model parameters were determined for a lab scale system using a small 2-mL column. The model was successfully used to scale up the capture operation from the lab scale column to a preparative bench scale column of about 1 L volume.

Effect of flow behavior in extra-column volumes on the retention pattern of proteins in a small column.

Experimental and theoretical analysis of deformation of band profiles in extra-column volumes (ECV) was performed, and its influence on the retention pattern of proteins in a small chromatographic column was quantified. Several macromolecule and small-molecule compounds, and their mixtures were eluted from a chromatographic system in the absence and presence of the column. The peak deformation in ECV was attributed to non-uniform velocity distribution in the radial direction in connecting capillaries. The phenomenon enhanced with increasing molecular weight of the model compound, when radial diffusion dominated the mechanism of band spreading. The band shape was also affected by the geometry of the injection system used, i.e., an injection loop capillary or a superloop. The phenomenon vanished for a small molecule compound, for which plug flow conditions could be established. The difference in flow behaviour of the macromolecule and small-molecule compounds caused them to migrate with different velocities in ECV, which resulted in partial separation of their bands. The ECV effect influenced the retention behaviour of macromolecules in a small column; it caused tailing of peaks and asymmetry of breakthrough curves. To describe the elution profiles in ECV and in the column, a mathematical model was used which accounted for non-ideality of the flow pattern. The model reproduced accurately band profiles of macromolecules within a range of relatively low velocities, typical however for protein chromatography.

Retention Behavior of Polyethylene Glycol and Its Influence on Protein Elution on Hydrophobic Interaction Chromatography Media.

The retention behavior of polyethylene glycol (PEG) on different types of hydrophobic interaction chromatography (HIC) resins containing butyl, octyl, and phenyl ligands was analyzed. An incomplete elution or splitting of the polymer peak into two parts was observed, where the first one was eluted at the dead time of the column, whereas the second one was strongly retained. The phenomenon was attributed to conformation changes of the polymer upon its adsorption on hydrophobic surface. The effect enhanced with increasing molecular weight of the polymer and hydrophobicity of the HIC media. Addition of PEG to the mobile phase reduced binding of proteins to HIC resins, which was demonstrated with two model systems: lysozyme (LYZ) and immunoglobulin G (IgG), and their mixtures. In case of LYZ, the presence of PEG caused reduction in the protein retention, whereas for IgG-a decrease in efficiency of the protein capture. The effect depended on the adsorption pattern of PEG; it was pronounced in the systems in which conformational changes of the polymer were suggested to occur.

Prediction tool for loading, isocratic elution, gradient elution and scaling up of ion exchange chromatography of proteins.

An efficient mathematical tool for the design and scaling up of protein chromatography is suggested, in which the model parameters can be determined quickly over a wide operating space without large material investments. The design method is based on mathematical modelling of column dynamics and moment analysis. The accuracy of the dynamic models that are most frequently used for simulations of chromatographic processes is analyzed, and possible errors that can be generated using the moment analysis are indicated. The so-called transport dispersive model was eventually employed for the process simulations. The model was modified to account for the protein dispersion in void volumes of chromatographic systems. The manner of the model calibration was suggested, which was based on a few chromatographic runs and verified over a wide space of the operating parameters, including composition and flow rate of the mobile phase, column dimensions, residence time, and mass loading. The model system for the study was ion-exchange chromatography. The analysis was performed based on the elution profiles of basic fibroblast growth factor 2 and lysozyme, on two different IEX media.

Effect of mass overloading on binding and elution of unstable proteins in hydrophobic interaction chromatography.

Adsorption behavior of unstable proteins, i.e., bovine serum albumin and α-lactalbumin, has been studied on a hydrophobic interaction chromatography medium under mass overloading conditions at different kosmotropic salt concentrations in the mobile phase. A mechanistic model has been formulated and used to describe kinetics and thermodynamics of protein interactions with the adsorbent surface. The model assumed two-site binding adsorption and reversible protein unfolding, which allowed predicting the inhibition of protein unfolding at high column loadings. A simplified procedure for the determination of model parameters has been developed, which was based on the inverse method. The model was successfully used to reproduce the pattern of chromatographic elution as well as the course of breakthrough curves. The model formulation was supported by Nano Differential Scanning Fluorimetry measurements, which were exploited to determine the protein stability in the liquid and adsorbed phases at different column loadings and salt concentrations.

A shortcut method for evaluation of protein deposition onto the membrane surface in crossflow ultrafiltration.

In this study, a procedure for quantifying the surface deposition of proteins in crossflow ultrafiltration has been developed. The procedure consists of determining the protein adsorption behavior onto the membrane surface from a few dynamic measurements performed in a nonfiltration and a filtration mode, and evaluating the concentration polarization (CP) layer thickness based on the adsorption data. To predict the interdependence between the protein adsorption and CP, a simplified mathematical model has been formulated. The model was used to assess the protein adsorption and thus yield reduction in the ultrafiltration process at different protein concentration in the solution. As a case study, ultrafiltration of aqueous solutions of BSA and lysozyme (LYZ) was examined on a polyethersulfone membrane with the molecular weight cutoff of 10 or 100 kDa. The protein concentration in the solutions varied within a relatively low concentration range, i.e. below 10 mg mL-1, characteristic for solvent exchange between sequential operations of protein purification by chromatography and extraction. Both proteins markedly differed in the mechanism of surface deposition; for BSA hydrophobic interactions were suggested to be dominant, whereas in case of LYZ electrostatic interactions contributed the most to the deposition mechanism. The effect of additives of the protein solutions, i.e. inorganic salts, PEG, and urea depended on the adsorption mechanism and was also specific for each protein. Nevertheless, the proposed procedure performed well in the evaluation of surface deposition and yield reduction, regardless of the protein type and its solvent environment.

Protein separation in carousel multicolumn setup. Performance analysis and experimental validation.

To overcome limitations of periodic separations of proteins in batch chromatographic columns Carousel Multi-Column Setup (CMS) has been recently suggested and theoretically analyzed in a previous study (R. Bochenek, W. Marek, W. Piatkowski, D. Antos, J. Chromatogr. A, 1301 (2013) 60-72). In this system, feed and mobile phase streams are subsequently delivered through parallel columns to mimic their countercurrent movement with respect to the fluid flow. All fluxes in the system are synchronized to ensure continuous feed delivery, which however causes reduction in the size of the operating window compared to batchwise-operating systems. In this study to improve the performance of CMS, additional process variables have been considered, such as the flow rate gradient and feed concentration. Though altering both variables allowed improving the separation selectivity and extending the operating window, the feed concentration appeared to be the most influential parameter affecting the process performance. Moreover, a procedure for practical realization of protein separations in CMS has been developed, including hints about the process design, configuration of columns and detectors, and use of pumps. As the case study, the separation of a ternary mixture of proteins, i.e., cytochrome C, lysozyme and immunoglobulin G, on hydrophobic interaction columns was used. A target product was a protein with intermediate adsorption strength that was isolated out of a more and less strongly adsorbed compound.

Overcoming solubility limits in overloaded gradient hydrophobic interaction chromatography.

The impact of the solubility limits on the performance of gradient protein chromatography has been studied. As a case study elution of model protein, i.e., lysozyme, on hydrophobic interaction media has been selected. A dependence of the protein solubility and crystallization kinetics on the content of cosmotropic salt in the mobile phase has been determined. Moreover, adsorption properties of the protein versus the mobile phase composition have been quantified. A model of chromatographic column dynamics has been developed which incorporated the mass transport kinetics accompanying both adsorption and crystallization processes. The model was used to study the influence of operating parameters such as flowrate and concentration loading on the solubility pattern inside the column and the separation performance. The analysis performed indicated existence of supersaturation regions for which, due to slow kinetics of crystallization, chromatographic process could be performed under conditions of strong concentration overloading while avoiding undesirable effects of flow blockage in chromatographic systems.

Adsorption behavior of proteins on temperature-responsive resins.

The adsorption behavior of proteins on thermo-responsible resins based on poly(N-isopropylacrylamide) and its copolymer containing an anionic co-monomer has been investigated. The influence of the polymer composition, i.e., the content of the co-monomer and crosslinker on the thermo-sensitivity of the protein adsorption has been quantified. The properties of ungrafted polymer as well grafted onto the agarose matrix have been analyzed and compared. Batch and dynamic (column) experiments have been performed to measure the adsorption equilibrium of proteins and to quantify the phase transition process. As model proteins lysozyme, lactoferrin, α-chymotrypsinogen A and ovalbumin have been used. The adsorption process was found to be governed by ionic interactions between the negatively charged surface of resin and the protein, which enabled separation of proteins differing in electrostatic charge. The interactions enhanced with increase of temperature. Decrease of temperature facilitated desorption of proteins and reduced the salt usage in the desorption buffer. Grafted polymers exhibited markedly higher mechanical stability and, however, weaker temperature response compared to the ungrafted ones.

Isolation of monoclonal antibody from a Chinese hamster ovary supernatant. I: assessment of different separation concepts.

The performance of different non-affinity purification techniques commonly used for isolating CHO derived monoclonal antibodies has been investigated. Ion exchange chromatography (IEC), hydrophobic interaction chromatography (HIC), aqueous-two-phase extraction (ATPE) and their integration has been evaluated in terms of yield and purity of the product obtained. The integration of chromatographic techniques comprised two steps, in which the CHO supernatant was directly injected into the IEC column to capture monoclonal IgG1 and then the isolated fraction was processed using the HIC column. To reduce the influence of the feed media on retention of the target protein, the feed mixture was on-column diluted by use of the multi-injection technique. In the coupled process of extraction and chromatography the ATPE operation was used for the pre-purification of the supernatant as well as for buffer exchange. The bottom ATPE phase containing the target protein was further purified on the HIC column without feed dilution. The influence of operating conditions on the effectiveness of different purification processes has been evaluated. The best performance with respect to the product purity was achieved for the coupled process of IEC and HIC. The experimental data acquired were exploited in subsequent investigations for determining underlying kinetic parameters and for the process prediction and optimization.

Isolation of monoclonal antibody from a Chinese hamster ovary supernatant. II: dynamics of the integrated separation on ion exchange and hydrophobic interaction chromatography media.

Dynamics of the purification process of a CHO derived monoclonal antibody by ion exchange chromatography (IEC), hydrophobic interaction chromatography (HIC) and their integration has been investigated. To quantify the adsorption behavior of the target protein (IgG1) and impurities contained in the supernatant, their elution course on IEC and HIC columns has been analyzed versus pH and/or the salt concentration in the mobile phase. A short-cut method has been proposed for mathematical modeling and determining underlying kinetic and thermodynamic parameters. The accuracy of the model predictions has been verified by comparing the simulated and experimental band profiles recorded in both chromatographic processes. After verification, the model was used to optimize operating conditions for the column loading and chromatographic elution in the integrated process IEC/HIC. Two alternative loading techniques based on the upstream and downstream feed dilution were taken into account in the optimization routine. In the first one the feed stream was diluted with the loading buffer prior to the column loading, while in the latter one the feed dilution was realized inside the column using the multiple-injection technique. It was shown that the downstream dilution allowed significant reduction of the contact time between the protein and the loading buffer.

Evaluating the performance of different multicolumn setups for chromatographic separation of proteins on hydrophobic interaction chromatography media by a numerical study.

A theoretical study has been performed on the effectiveness of isolating a target component out of a multi-component protein mixture using different arrangements of chromatographic columns. Three continuous systems have been considered which were able to perform solvent gradient separations, such as: open loop simulated moving bed, countercurrent solvent gradient purification and carousel multicolumn setup. The performance of the continuous processes was examined with respect to productivity, yield and eluent consumption and compared to a single-column batch system. As a case study separation of a ternary mixture of proteins on HIC media has been selected. Two separation problems have been analyzed referring to the situation when the target component was the most strongly adsorbed as well as when it exhibited intermediate adsorption strength. A mathematical model has been used to simulate the process dynamics and to optimize operating conditions for the separation. The numerical study indicated that batch column arrangements can outperform SMB-based configurations regarding all performance indicators considered, which has been attributed to solvent mixing in the recycled streams and distortion of the gradient shape in SMB units. It has been concluded that the performance of complex multicolumn systems should be verified vs. batch column operations prior to the realization of the separation process.

Multiple-injection technique for isolating a target protein from multicomponent mixtures.

An integrated chromatographic process comprising ion exchange (IEC) and hydrophobic interaction chromatography (HIC) for isolating a target protein form multicomponent mixtures has been analyzed. The model mixture contained immunoglobulin G that was the key product of the separation process, cytochrome C and ovalbumin. The adsorption characteristics and the mass transport kinetics of the model proteins have been determined along with their dependencies on the operating variables such as pH, temperature and the salt concentration for IEC as well as HIC media. Limitations of the process efficiency resulting from kinetic effects, solubility constraints and the necessity of the mobile phase exchange between chromatographic steps have been discussed. To improve the performance of the integrated process the multiple-injection technique has been suggested. This technique consisted in loading feed mixtures dissolved in a good solvent onto the column by several small-volume injections under conditions of strong protein adsorption. It allowed diminishing interactions between the sample-solvent and protein and elimination of undesired effects such as band splitting and band broadening. For the process design and optimization a dynamic model has been used accounting for thermodynamics and kinetics of the process. The optimization results indicated superiority of the multiple-injection technique over standard isocratic injections in terms of the process yield and productivity.