Timothy M. Pabst

Protein separations with induced pH gradients using cation-exchange chromatographic columns containing weak acid groups

The behavior of chromatographic columns packed with resins containing both weak and strong cation-exchange groups is investigated in order to obtain protein separations by means of internally generated pH gradients in response to step changes in buffer composition. A local equilibrium model is developed to predict pH transitions using non-adsorbed buffers, i.e. containing neutral and negatively charged buffering species, based exclusively on the resin titration curve. In agreement with experimental results, the model predicts practical, fairly linear gradients between pH 5 and 7, which are formed using suitable mixtures of acetate and phosphate buffers. The separation of mixtures of ovalbumin, albumin, and transferrin is used as a model system, but, unlike most previous work, we consider preparative conditions. Near baseline resolution is obtained with protein loads as high as 10mg/mL and mobile phase velocities at high as 460 cm/h using porous, 70-microm diameter particles. The peaks obtained with this approach are much sharper than could be obtained isocratically or using externally generated, unretained gradients as a result of the peak compression caused by the axial pH gradient formed along the column. Moreover, separation is obtained at very low ionic strengths (2-3 mS/cm). The effects of flow velocity, mobile phase composition, time of injection, and protein load on retention and elution pH are investigated systematically demonstrating a range of ways in which the separation can be controlled and optimized.

Engineering of novel Staphylococcal Protein A ligands to enable milder elution pH and high dynamic binding capacity.

We describe novel Staphylococcal Protein A ligands that enable milder elution pH for use in affinity chromatography. The change in elution pH is the result of point mutations to the protein sequence. Two novel ligands are investigated in this study. The first, designated Z(H18S)4, represents a histidine to serine substitution single mutation. The second, designated Z(H18S, N28A)4, is a double mutant comprising histidine to serine and asparagine to alanine mutations. Both are compared against the unmutated sequence, designated Z4, which is currently utilized in a commercially available Protein A stationary phase for the purification of molecules containing Fc domains. The ligands are coupled to a chromatography support matrix and tested against a panel of antibodies and an Fc fusion protein for elution pH, dynamic binding capacity, step-wise elution, and capture from clarified culture media. Results demonstrate that the novel ligands result in milder elution pH, on average >0.5 pH units, when tested in a pH gradient. For step-wise elution at pH 4.0, the Z(H18S, N28A)4 ligand showed on average a greater than 30% increase in yield compared to Z4. Importantly, for the antibodies tested the mutations did not result in a decrease in dynamic binding capacity or other desirable attributes such as selectivity. A potential application of the novel ligands is shown with a pH sensitive molecule prone to aggregation under acidic conditions.

Adsorption equilibrium and kinetics of monomer–dimer monoclonal antibody mixtures on a cation exchange resin

Adsorption equilibrium and kinetics are determined for a monoclonal antibody (mAb) monomer and dimer species, individually and in mixtures, on a macroporous cation exchange resin both under the dilute limit of salt gradient elution chromatography and at high protein loads and low salt based on batch adsorption equilibrium and confocal laser scanning microscopy (CLSM) experiments. In the dilute limit and weak binding conditions, the dimer/monomer selectivity in 10mM phosphate at pH 7 varies between 8.7 and 2.3 decreasing with salt concentration in the range of 170-230mM NaCl. At high protein loads and strong binding conditions (0-60mM NaCl), the selectivity in the same buffer is near unity with no NaCl added, but increases gradually with salt concentration reaching high values between 2 and 15 with 60mM added NaCl. For these conditions, the two-component adsorption kinetics is controlled by pore diffusion and is predicted approximately by a dual shrinking core model using parameters based on single component equilibrium and kinetics measurements.

Development and scale‐up of a commercial fed batch refolding process for an anti‐CD22 two chain immunotoxin

We describe the development and scale‐up of a novel two chain immunotoxin refolding process. This work provides a case study comparing a clinical manufacturing process and the commercial process developed to replace it. While the clinical process produced high quality material, it suffered from low yield and high yield variability. A systematic approach to process development and understanding led to a number of improvements that were implemented in the commercial process. These include a shorter inclusion body recovery process, limiting the formation of an undesired deamidated species and the implementation of fed batch dilution refolding for increased refold titers. The use of a combination of urea, arginine and DTT for capture column cleaning restored the binding capacity of the capture step column and resulted in consistent capture step yields compared to the clinical process. Scalability is shown with data from 250 L and 950 L scale refolding processes. Compared to the clinical process it replaces, the commercial process demonstrated a greater than fivefold improvement in volumetric productivity at the 950 L refolding scale.

Process-scale purification and analytical characterization of highly gamma-carboxylated recombinant human prothrombin.

Prothrombin (coagulation Factor II) is a complex multidomain glycoprotein that plays a central role in blood coagulation. It is the zymogen precursor to the protease thrombin that catalyzes the formation of the fibrin clot and regulates a multitude of other cellular responses related to coagulation and hemostasis. For the biological activity of prothrombin, the vitamin K dependent posttranslational modification of glutamic acid residues to gamma-carboxylglutamic acid is of crucial importance. Prothrombin can be recombinantly expressed using mammalian cell culture. However, the product is a heterogeneous mixture of variants with different degrees of carboxylation, requiring separation of closely related charge isoforms. A second challenge for purification is the need to remove traces of the product-related impurity thrombin, a protease, to extremely low levels. In this work, we describe a purification strategy that provides solutions to both challenges and results in an efficient and robust process for active recombinant prothrombin. We also describe the analytical characterization of recombinant prothrombin by HPLC, LC-MS/MS, and complementary biochemical assays.

Separation of antibody monomer-dimer mixtures by frontal analysis

The removal of aggregates, particularly soluble dimers, from monoclonal antibodies (mAbs) remains a persistent challenge in downstream processing. In this work, we have examined the separation of an antibody monomer from its dimer on the cation exchange resin Nuvia HR-S (Bio-Rad Laboratories) using frontal analysis. In this process, a mixture of monomer and dimer is continuously fed to the column under conditions where the mixture is favorably bound, resulting in two breakthrough fronts whose monomer and dimer compositions are determined by the multi-component equilibrium and kinetics of the system. Experimentally, the selectivity for dimer was found to vary substantially with ionic strength, being lowest when conditions favor the strongest binding, and increasing to a maximum at intermediate ionic strengths where rapid exchange with the bound monomer can occur. A mechanistic model is developed to describe the competitive binding frontal analysis process, assuming pore diffusion and a significant kinetic resistance to binding as a function of ionic strength. The model was solved numerically and was able to describe both the frontal analysis processes and batch adsorption experimental data, accounting for process parameters such as feed composition and salt concentration. The resulting model can be used to optimize column operating conditions for yield and purity.

Liquid-liquid phase separation causes high turbidity and pressure during low pH elution process in Protein A chromatography

Turbid elution pools and high column back pressure are common during elution of monoclonal antibodies (mAbs) by acidic pH in Protein A chromatography. This phenomenon has been historically attributed to acid-induced precipitation of incorrectly folded or pH-sensitive mAbs and host cell proteins (HCPs). In this work, we propose a new mechanism that may account for some observations of elution turbidity in Protein A chromatography. We report several examples of turbidity and high column back pressure occurring transiently under a short course of neutral conditions during Protein A elution. A systematic study of three mAbs displaying this behavior revealed phase separation characterized by liquid drops under certain conditions including neutral pH, low ionic strength, and high protein concentration. These liquid droplets caused solution turbidity and exhibited extremely high viscosity, resulting in high column back pressure. We found out that the droplets were formed through liquid-liquid phase separation (LLPS) as a result of protein self-association. We also found multiple factors, including pH, temperature, ionic strength, and protein concentration can affect LLPS behaviors. Careful selection of process parameters during protein A elution, including temperature, flow rate, buffer, and salt can inhibit formation of a dense liquid phase, reducing both turbidity (by 90%) and column back pressure (below 20 pounds per square inch). These findings provide both mechanistic insight and practical mitigation strategies for Protein A chromatography induced LLPS.

Evaluation of recent Protein A stationary phase innovations for capture of biotherapeutics

We describe a comprehensive evaluation of 12 Protein A stationary phases for capture of biotherapeutics. We first examine the morphological properties of the stationary phases using a variety of orthogonal techniques including electron microscopy, particle sizing, pressure-flow behavior, and isocratic pulse response. A panel of nine proteins spanning a wide range of structures and biochemical properties was then used to assess equilibrium uptake, mass transport, dynamic binding capacity, and elution pH. Process performance and product quality were also examined under realistic bioprocess conditions using clarified mammalian cell culture broth. Equilibrium isotherms were found to be highly favorable, with equilibrium binding capacity for monoclonal and bispecific antibodies ranging from 47-100 mg/mL packed bed across all stationary phases tested. Effective pore diffusivities, De, were obtained by fitting the chromatography general rate model to breakthrough data. The fitted De values for monoclonal antibodies ranged from 1.1-5.7 × 10-8 cm2/s. The stationary phases had high dynamic binding capacities for the model proteins. The highest dynamic capacities for monoclonal and bispecific antibodies were seen with MabSelect SuRe pcc and MabSelect PrismA, which ranged from 58-74 mg/mL packed bed at 4 min residence times. Product capture using clarified cell culture broth as a feedstock showed high yields and elution pool volumes that ranged from 2-3 column volumes in most cases. Host cell protein, DNA, and aggregate levels in the elution pool were dependent on the specific nature of protein being purified, and levels were consistent between stationary phases. Lastly, we perform an analysis of bivariate correlations and discuss considerations for process design and optimization.

Cathepsin L Causes Proteolytic Cleavage of CHO Expressed Proteins During Processing and Storage: Identification, Characterization, and Mitigation

A stochastic approach of copurification of the protease Cathepsin L that results in product fragmentation during purification processing and storage is presented. Cathepsin L was identified using mass spectroscopy, characterization of proteolytic activity, and comparison with fragmentation patterns observed using recombinant Cathepsin L. Cathepsin L existed in Chinese hamster ovary cell culture fluids obtained from cell lines expressing different products and cleaved a variety of recombinant proteins including monoclonal antibodies, antibody fragments, bispecific antibodies, and fusion proteins. Therefore, characterization its chromatographic behavior is essential to ensure robust manufacturing and sufficient shelf life. The chromatographic behaviors of Cathepsin L using a variety of techniques including affinity, cation exchange, anion exchange, and mixed mode chromatography were systematically evaluated. Our data demonstrates that copurification of Cathepsin L on nonaffinity modalities is principally because of similar retention on the stationary phase and not through interactions with product. Lastly, Cathespin L exhibits a broad elution profile in cation exchange chromatography (CEX) likely because of its different forms. Affinity purification is free of fragmentation issue, making affinity capture the best mitigation of Cathepsin L. When affinity purification is not feasible, a high pH wash on CEX can effectively remove Cathepsin L but resulted in significant product loss, while anion exchange chromatography operated in flow‐through mode does not efficiently remove Cathepsin L. Mixed mode chromatography, using Capto™ adhere in this example, provides robust clearance over wide process parameter range (pH 7.7 ± 0.3 and 100 ± 50 mM NaCl), making it an ideal technique to clear Cathepsin L.

Systematic Interpolation Method Predicts Antibody Monomer-Dimer Separation by Gradient Elution Chromatography at High Protein Loads

A previously developed empirical interpolation (EI) method is extended to predict highly overloaded multicomponent elution behavior on a cation exchange (CEX) column based on batch isotherm data. Instead of a fully mechanistic model, the EI method employs an empirically modified multicomponent Langmuir equation to correlate two-component adsorption isotherm data at different salt concentrations. Piecewise cubic interpolating polynomials are then used to predict competitive binding at intermediate salt concentrations. The approach is tested for the separation of monoclonal antibody monomer and dimer mixtures by gradient elution on the cation exchange resin Nuvia HR-S. Adsorption isotherms are obtained over a range of salt concentrations with varying monomer and dimer concentrations. Coupled with a lumped kinetic model, the interpolated isotherms predict the column behavior for highly overloaded conditions. Predictions based on the EI method shows good agreement with experimental elution curves for protein loads up to 40 mg mL-1 column or about 50% of the column binding capacity. The approach can be extended to other chromatographic modalities and to more than two components.