A case study application of AQbD to the re-development and validation of an affinity chromatography analytical procedure for mAb titer quantitation.

Protein A (ProA) high-performance liquid chromatography (HPLC) is a common analytical procedure for measuring monoclonal antibody (mAb) titers due to its high specificity and efficiency. Accurate and reliable results of this procedure are imperative, as the quantitation of the total mAb present for in-process samples directly impacts downstream purification steps related to the removal of process-related impurities. This study aimed to improve a platform ProA HPLC analytical procedure which was previously developed using traditional approaches and was not always reliable. By retrospectively applying Analytical Quality by Design (AQbD) principles and statistical assessments of performance, a bias in the calibration standard due to protein-adsorption to common sample vial materials was identified. The inclusion of Tween® 20 into the mobile phase used as sample diluent was optimized to ensure procedure performance and improve analytical range. The resulting procedure robustness was evaluated using Design of Experiment (DoE) approaches and performance was verified against Analytical Target Profile (ATP) criteria as recommended by regulatory agencies. The resulting linearity displayed R2 values of 1.00 with intercept biases of 1.2 % (analyst 1) and 0.8 % (analyst 2), accuracy across all levels was reported at 99.2 % recovery, and intermediate precision was reported as 3.0 % RSD. Application of this new platform procedure has since reduced development timelines for new mAb products by 50 % and allowed for accurate titer determination to support >5 early phase product-specific process decisions without requiring extensive analytical procedure development. This work demonstrates the utility and relative ease of adopting AQbD concepts, even for established procedures, and supporting them with a lifecycle approach to managing procedure performance.

[Preparation technology comparison and performance evaluation of different protein A affinity chromatographic materials].

Protein A affinity chromatographic materials are widely used in clinical medicine and biomedicine because of their specific interactions with immunoglobulin G (IgG). Both the characteristics of the matrix, such as its structure and morphology, and the surface modification method contribute to the affinity properties of the packing materials. The specific, orderly, and oriented immobilization of protein A can reduce its steric hindrance with the matrix and preserve its bioactive sites. In this study, four types of affinity chromatographic materials were obtained using agarose and polyglycidyl methacrylate (PGMA) spheres as substrates, and multifunctional epoxy and maleimide groups were used to fix protein A. The effects of the ethylenediamine concentration, reaction pH, buffer concentration, and other conditions on the coupling efficiency of protein A and adsorption performance of IgG were evaluated. Multifunctional epoxy materials were prepared by converting part of the epoxy groups of the agarose and PGMA matrices into amino groups using 0.2 and 1.6 mol/L ethylenediamine, respectively. Protein A was coupled to the multifunctional epoxy materials using 5 mmol/L borate buffer (pH 8) as the reaction solution. When protein A was immobilized on the substrates by maleimide groups, the agarose and PGMA substrates were activated with 25% (v/v) ethylenediamine for 16 h to convert all epoxy groups into amino groups. The maleimide materials were then converted into amino-modified materials by adding 3 mg/mL 3-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS) dissolved in dimethyl sulfoxide (DMSO) and then suspended in 5 mmol/L borate buffer (pH 8). The maleimide groups reacted specifically with the C-terminal of the sulfhydryl group of recombinant protein A to achieve highly selective fixation on both the agarose and PGMA substrates. The adsorption performance of the affinity materials for IgG was improved by optimizing the bonding conditions of protein A, such as the matrix type, matrix particle size, and protein A content, and the adsorption properties of each affinity material for IgG were determined. The column pressure of the protein A affinity materials prepared using agarose or PGMA as the matrix via the maleimide method was subsequently evaluated at different flow rates. The affinity materials prepared with PGMA as the matrix exhibited superior mechanical strength compared with the materials prepared with agarose. Moreover, an excellent linear relationship between the flow rate and column pressure of 80 mL/min was observed for this affinity material. Subsequently, the effect of the particle size of the PGMA matrix on the binding capacity of IgG was investigated. Under the same protein A content, the dynamic binding capacity of the affinity materials on the PGMA matrix was higher when the particle size was 44-88 µm than when other particle sizes were used. The properties of the affinity materials prepared using the multifunctional epoxy and maleimide-modified materials were compared by synthesizing affinity materials with different protein A coupling amounts of 1, 2, 4, 6, 8, and 10 mg/mL. The dynamic and static binding capacities of each material for bovine IgG were then determined. The prepared affinity material was packed into a chromatographic column to purify IgG from bovine colostrum. Although all materials showed specific adsorption selectivity for IgG, the affinity material prepared by immobilizing protein A on the PGMA matrix with maleimide showed significantly better performance and achieved a higher dynamic binding capacity at a lower protein grafting amount. When the protein grafting amount was 15.71 mg/mL, the dynamic binding capacity of bovine IgG was 32.23 mg/mL, and the dynamic binding capacity of human IgG reached 54.41 mg/mL. After 160 cycles of alkali treatment, the dynamic binding capacity of the material reached 94.6% of the initial value, indicating its good stability. The developed method is appropriate for the production of protein A affinity chromatographic materials and shows great potential in the fields of protein immobilization and immunoadsorption material synthesis.

Direct Affinity Ligand Immobilization onto Bare Iron Oxide Nanoparticles Enables Efficient Magnetic Separation of Antibodies.

Magnetic separation is a promising alternative to chromatography for enhancing the downstream processing (DSP) of monoclonal antibodies (mAbs). However, there is a lack of efficient magnetic particles for successful application. Aiming to fill this gap, we demonstrate the suitability of bare iron oxide nanoparticles (BION) with physical site-directed immobilization of an engineered Protein A affinity ligand (rSpA) as an innovative magnetic material. The rSpA ligand contains a short peptide tag that enables the direct and stable immobilization onto the uncoated BION surface without commonly required laborious particle activation. The resulting BION@rSpA have beneficial characteristics outperforming conventional Protein A-functionalized magnetic particles: a simple, fast, low-cost synthesis, a particle size in the nanometer range with a large effective specific surface area enabling large immunoglobulin G (IgG) binding capacity, and a high magnetophoretic velocity advantageous for fast processing. We further show rapid interactions of IgG with the easily accessible rSpA ligands. The binding of IgG to BION@rSpA is thereby highly selective and not impeded by impurity molecules in perfusion cell culture supernatant. Regarding the subsequent acidic IgG elution from BION@rSpA@IgG, we observed a hampering pH increase caused by the protonation of large iron oxide surfaces after concentrating the particles in 100 mM sodium acetate buffer. However, the pH can be stabilized by adding 50 mM glycine to the elution buffer, resulting in recoveries above 85% even at high particle concentrations. Our work shows that BION@rSpA enable efficient magnetic mAb separation and could help to overcome emerging bottlenecks in DSP.

Rationally designed protein A surface molecularly imprinted magnetic nanoparticles for the capture and detection of Staphylococcus aureus.

Staphylococcus aureus (S. aureus), a commensal organism found on the human skin, is commonly associated with nosocomial infections and exhibits virulence mediated by toxins and resistance to antibiotics. The global threat of antibiotic resistance has necessitated antimicrobial stewardship to improve the safe and appropriate use of antimicrobials; hence, there is an urgent demand for the advanced, cost-effective, and rapid detection of specific bacteria. In this regard, we aimed to selectively detect S. aureus using surface molecularly imprinted magnetic nanoparticles templated with a well-known biomarker protein A, specific to S. aureus. Herein, a highly selective surface molecularly imprinted polymeric thin layer was created on ∼250 nm magnetic nanoparticles (MNPs) through the immobilization of protein A to aldehyde functionalized MNPs, followed by monomer polymerization and template washing. This study employs the rational selection of monomers based on their computationally predicted binding affinity to protein A at multiple surface residues. The resulting MIPs from rationally selected monomer combinations demonstrated an imprinting factor as high as ∼5. Selectivity studies revealed MIPs with four-fold higher binding capacity (BC) to protein A than other non-target proteins, such as lysozyme and serum albumin. In addition, it showed significant binding to S. aureus, whereas negligible binding to other non-specific Gram-negative, i.e. Escherichia coli (E. coli), Pseudomonas aeruginosa (P. aeruginosa), and Gram-positive, i.e. Bacillus subtilis (B. subtilis), bacteria. This MIP was employed for the capture and specific detection of fluorescently labeled S. aureus. Quantitative detection was performed using a conventional plate counting technique in a linear detection range of 101-107 bacterial cells. Remarkably, the MIPs also exhibited approximately 100% cell recovery from milk samples spiked with S. aureus (106 CFU mL-1), underscoring its potential as a robust tool for sensitive and accurate bacterial detection in dairy products. The developed MIP exhibiting high affinity and selective binding to protein A finds its potential applications in the magnetic capture and selective detection of protein A as well as S. aureus infections and contaminations.

An inert tracer: The binding site of a fluorescent dye on the antibody and its effects on Protein A chromatography.

Fluorescently labeled antibodies are widely used to visualize the adsorption process in protein chromatography using confocal laser scanning microscopy (CLSM), but also as a tracer for determination of residence time distribution (RTD) in continuous chromatography. It is assumed that the labeled protein is inert and representative of the unlabeled antibody, ignoring the fact that labeling with a fluorescent dye can change the characteristics of the original molecule. It became evident that the fluorescently labeled antibody has a higher affinity toward protein A resins such as MabSelect Sure. This can be due to slight differences in hydrophobicity and net charge, which are caused by the addition of the fluorescent dye. However, this difference is eliminated when using high salt concentrations in the adsorption studies. In this work, the site occupancy of two labeled antibodies, MAb1 (IgG1 subclass) and MAb2 (IgG2 subclass) conjugated with the fluorescent dye Alexa Fluor™ 488 was elucidated by intact mass spectrometry (MS) and peptide mapping LC-MS/MS, employing a sequential cleavage with Endoproteinase Lys-C and trypsin and in parallel with chymotrypsin alone. It was shown that the main binding site for the dye was a specific lysine in the heavy chains of the MAb1 and MAb2 molecules, in positions 188 and 189 respectively. Other lysine residues distributed throughout the protein sequence were labeled to a lot lesser extent. The labeled antibody had a slightly different affinity to MabSelect Sure although its primary binding site (to Protein A) was not affected by labeling, despite the secondary region responsible for binding to the protein A was partly labeled. Overall, the fluorescent-labeled antibodies are a good compromise as an inert tracer in residence time distribution and chromatography studies because they are much cheaper than isotope-labeled antibodies; However, the differences between the labeled and unlabeled antibodies should be considered.

CaptureSelect FcXP affinity medium exhibits strong aggregate separation capability.

Protein A affinity chromatography has been widely used for initial product capture in recombinant antibody/Fc-fusion purification. However, in general Protein A lacks the capability of separating aggregates (unless the aggregates are too large to enter the pores of resin beads or have their Protein A binding sites buried, in which case the aggregates do not bind). In the current work, we demonstrated that CaptureSelect FcXP affinity medium exhibited strong aggregate separation capability and effectively removed aggregates under pH or conductivity gradient elution in two bispecific antibody (bsAb) cases. For these two cases, aggregate contents were reduced from >16% and >22% (in the feed) to <1% and <5% (in the eluate) for the first and second bsAbs, respectively. While more case studies are required to further demonstrate FcXP's superiority in aggregate removal, findings from the current study suggest that FcXP can potentially be a better alternative than Protein A for product capture in cases where aggregate content is high.

Computer-aided design space identification for screening of protein A affinity chromatography resins.

The rapidly growing market of monoclonal antibodies (mAbs) within the biopharmaceutical industry has incentivised numerous works on the design of more efficient production processes. Protein A affinity chromatography is regarded as one of the best processes for the capture of mAbs. Although the screening of Protein A resins has been previously examined, process flexibility has not been considered to date. Examining performance alongside flexibility is crucial for the design of processes that can handle disturbances arising from the feed stream. In this work, we present a model-based approach for the identification of design spaces, enhanced by machine learning. We demonstrate its capabilities on the design of a Protein A chromatography unit, screening five industrially relevant resins. The computational results favourably compare to experimental data and a resin performance comparison is presented. An improvement on the computational time by a factor of 300,000 is achieved using the machine learning aided methodology. This allowed for the identification of 5,120 different design spaces in only 19 h.

Purification of monoclonal antibodies using novel 3D printed ordered stationary phases.

3D printing offers the unprecedented ability to fabricate chromatography stationary phases with bespoke 3D morphology as opposed to traditional packed beds of spherical beads. The restricted range of printable materials compatible with chromatography is considered a setback for its industrial implementation. Recently, we proposed a novel ink that exhibits favourable printing performance (printing time ∼100 mL/h, resolution ∼200 µm) and broadens the possibilities for a range of chromatography applications thanks to its customisable surface chemistry. In this work, this ink was used to fabricate 3D printed ordered columns with 300 µm channels for the capture and polishing of therapeutic monoclonal antibodies. The columns were initially assessed for leachables and extractables, revealing no material propensity for leaching. Columns were then functionalised with protein A and SO3 ligands to obtain affinity and strong cation exchangers, respectively. 3D printed protein A columns showed >85 % IgG recovery from harvested cell culture fluid with purities above 98 %. Column reusability was evaluated over 20 cycles showing unaffected performance. Eluate samples were analysed for co-eluted protein A fragments, host cell protein and aggregates. Results demonstrate excellent HCP clearance (logarithmic reduction value of > 2.5) and protein A leakage in the range of commercial affinity resins (<100 ng/mg). SO3 functionalised columns employed for polishing achieved removal of leaked Protein A (down to 10 ng/mg) to meet regulatory expectations of product purity. This work is the first implementation of 3D printed columns for mAb purification and provides strong evidence for their potential in industrial bioseparations.

Chromatography affinity resin with photosynthetically-sourced protein A ligand.

Green, photosynthesizing plants can be proficiently used as cost-effective, single-use, fully biodegradable bioreactors for environmentally-friendly production of a variety of valuable recombinant proteins. Being near-infinitely scalable and most energy-efficient in generating biomass, plants represent profoundly valid alternatives to conventionally used stationary fermenters. To validate this, we produced a plastome-engineered tobacco bioreactor line expressing a recombinant variant of the protein A from Staphylococcus aureus, an affinity ligand widely useful in antibody purification processes, reaching accumulation levels up to ~ 250 mg per 1 kg of fresh leaf biomass. Chromatography resin manufactured from photosynthetically-sourced recombinant protein A ligand conjugated to agarose beads demonstrated the innate pH-driven ability to bind and elute IgG-type antibodies and allowed one-step efficient purification of functional monoclonal antibodies from the supernatants of the producing hybridomas. The results of this study emphasize the versatility of plant-based recombinant protein production and illustrate its vast potential in reducing the cost of diverse biotechnological applications, particularly the downstream processing and purification of monoclonal antibodies.

A Multifunctionalized Potyvirus-Derived Nanoparticle That Targets and Internalizes into Cancer Cells.

Plant viral nanoparticles (VNPs) are attractive to nanomedicine researchers because of their safety, ease of production, resistance, and straightforward functionalization. In this paper, we developed and successfully purified a VNP derived from turnip mosaic virus (TuMV), a well-known plant pathogen, that exhibits a high affinity for immunoglobulins G (IgG) thanks to its functionalization with the Z domain of staphylococcal Protein A via gene fusion. We selected cetuximab as a model IgG to demonstrate the versatility of this novel TuMV VNP by developing a fluorescent nanoplatform to mark tumoral cells from the Cal33 line of a tongue squamous cell carcinoma. Using confocal microscopy, we observed that fluorescent VNP-cetuximab bound selectively to Cal33 and was internalized, revealing the potential of this nanotool in cancer research.

Helical Content Correlations and Hydration Structures of the Folding Ensemble of the B Domain of Protein A.

The B domain of protein A (BdpA), a small three-helix bundle, folds on a time scale of a few microseconds with heterogeneous native and unfolded states. It is widely used as a model for understanding protein folding mechanisms. In this work, we use structure-based models (SBMs) and atomistic simulations to comprehensively investigate how BdpA folding is associated with the formation of its secondary structure. The energy landscape visualization method (ELViM) was used to characterize the pathways that connect the folded and unfolded states of BdpA as well as the sets of structures displaying specific ellipticity patterns. We show that the native state conformational diversity is due mainly to the conformational variability of helix I. Helices I, II, and III occur in a weakly correlated manner, with Spearman's rank correlation coefficients of 0.1539 (I and II), 0.1259 (I and III), and 0.2561 (II and III). These results, therefore, suggest the highest cooperativity between helices II and III. Our results allow the clustering of partially folded structures of folding of the B domain of protein A on the basis of its secondary structure, paving the way to an understanding of environmental factors in the relative stability of the basins of the folding ensemble, which are illustrated by the structural dependency of the protein hydration structures, as computed with minimum-distance distribution functions.

Host cell protein networks as a novel co-elution mechanism during protein A chromatography.

Host cell proteins (HCPs) are process-related impurities of therapeutic proteins produced in for example, Chinese hamster ovary (CHO) cells. Protein A affinity chromatography is the initial capture step to purify monoclonal antibodies or Fc-based proteins and is most effective for HCP removal. Previously proposed mechanisms that contribute to co-purification of HCPs with the therapeutic protein are either HCP-drug association or leaching from chromatin heteroaggregates. In this study, we analyzed protein A eluates of 23 Fc-based proteins by LC-MS/MS to determine their HCP content. The analysis revealed a high degree of heterogeneity in the number of HCPs identified in the different protein A eluates. Among all identified HCPs, the majority co-eluted with less than three Fc-based proteins indicating a drug-specific co-purification for most HCPs. Only ten HCPs co-purified with over 50% of the 23 Fc-based proteins. A correlation analysis of HCPs identified across multiple protein A eluates revealed their co-elution as HCP groups. Functional annotation and protein interaction analysis confirmed that some HCP groups are associated with protein-protein interaction networks. Here, we propose an additional mechanism for HCP co-elution involving protein-protein interactions within functional networks. Our findings may help to guide cell line development and to refine downstream purification strategies.

pH inactivation of SARS-CoV-2 and SARS-CoV in virus spiked protein A eluates from a mAb purification process.

Biopharmaceutical manufacturing processes may include a low pH treatment step as a means of inactivating enveloped viruses. Small scale virus clearance studies are routinely performed using model enveloped viruses such as murine leukemia virus to assess inactivation at the pH range used in the downstream manufacturing process. Further, as a means of bioburden reduction, chromatography resins may be cleaned and stored using sodium hydroxide and this can also inactivate viruses. The susceptibility of SARS-CoV-2 and SARS-CoV to low pH conditions using protein A eluate derived material from a monoclonal antibody production process as well as high pH cleaning conditions was addressed. SARS-CoV-2 was effectively inactivated at pH 3.0, moderately inactivated at pH 3.4, but not inactivated at pH 3.8. Low pH was less effective at inactivating SARS-CoV. Both viruses were inactivated at a high pH of ca.13.4. These studies provide important information regarding the effectiveness of viral clearance and inactivation steps of novel coronaviruses when compared to other enveloped viruses.

Continuous multi-column capture of monoclonal antibodies with convective diffusive membrane adsorbers.

Downstream processing is the bottleneck in the continuous manufacturing of monoclonal antibodies (mAbs). To overcome throughput limitations, two different continuous processes with a novel convective diffusive protein A membrane adsorber (MA) were investigated: the rapid cycling parallel multi-column chromatography (RC-PMCC) process and the rapid cycling simulated moving bed (RC-BioSMB) process. First, breakthrough curve experiments were performed to investigate the influence of the flow rate on the mAb dynamic binding capacity and to calculate the duration of the loading steps. In addition, customized control software was developed for an automated MA exchange in case of pressure increase due to membrane fouling to enable robust, uninterrupted, and continuous processing. Both processes were performed for 4 days with 0.61 g L-1 mAb-containing filtrate and process performance, product purity, productivity, and buffer consumption were compared. The mAb was recovered with a yield of approximately 90% and productivities of 1010 g L-1 d-1 (RC-PMCC) and 574 g L-1 d-1 (RC-BioSMB). At the same time, high removal of process-related impurities was achieved with both processes, whereas the buffer consumption was lower for the RC-BioSMB process. Finally, the attainable productivity for perfusion bioreactors of different sizes with suitable MA sizes was calculated to demonstrate the potential to operate both processes on a manufacturing scale with bioreactor volumes of up to 2000 L.

Explainable deep learning enhances robust and reliable real-time monitoring of a chromatographic protein A capture step.

The application of model-based real-time monitoring in biopharmaceutical production is a major step toward quality-by-design and the fundament for model predictive control. Data-driven models have proven to be a viable option to model bioprocesses. In the high stakes setting of biopharmaceutical manufacturing it is essential to ensure high model accuracy, robustness, and reliability. That is only possible when (i) the data used for modeling is of high quality and sufficient size, (ii) state-of-the-art modeling algorithms are employed, and (iii) the input-output mapping of the model has been characterized. In this study, we evaluate the accuracy of multiple data-driven models in predicting the monoclonal antibody (mAb) concentration, double stranded DNA concentration, host cell protein concentration, and high molecular weight impurity content during elution from a protein A chromatography capture step. The models achieved high-quality predictions with a normalized root mean squared error of <4% for the mAb concentration and of ≈10% for the other process variables. Furthermore, we demonstrate how permutation/occlusion-based methods can be used to gain an understanding of dependencies learned by one of the most complex data-driven models, convolutional neural network ensembles. We observed that the models generally exhibited dependencies on correlations that agreed with first principles knowledge, thereby bolstering confidence in model reliability. Finally, we present a workflow to assess the model behavior in case of systematic measurement errors that may result from sensor fouling or failure. This study represents a major step toward improved viability of data-driven models in biopharmaceutical manufacturing.

Digital twin in high throughput chromatographic process development for monoclonal antibodies.

The monoclonal antibody (mAb) industry is becoming increasingly digitalized. Digital twins are becoming increasingly important to test or validate processes before manufacturing. High-Throughput Process Development (HTPD) has been progressively used as a tool for process development and innovation. The combination of High-Throughput Screening with fast computational methods allows to study processes in-silico in a fast and efficient manner. This paper presents a hybrid approach for HTPD where equal importance is given to experimental, computational and decision-making stages. Equilibrium adsorption isotherms of 13 protein A and 16 Cation-Exchange resins were determined with pure mAb. The influence of other components in the clarified cell culture supernatant (harvest) has been under-investigated. This work contributes with a methodology for the study of equilibrium adsorption of mAb in harvest to different protein A resins and compares the adsorption behavior with the pure sample experiments. Column chromatography was modelled using a Lumped Kinetic Model, with an overall mass transfer coefficient parameter (kov). The screening results showed that the harvest solution had virtually no influence on the adsorption behavior of mAb to the different protein A resins tested. kov was found to have a linear correlation with the sample feed concentration, which is in line with mass transfer theory. The hybrid approach for HTPD presented highlights the roles of the computational, experimental, and decision-making stages in process development, and how it can be implemented to develop a chromatographic process. The proposed white-box digital twin helps to accelerate chromatographic process development.

On-line PAT based monitoring and control of resin aging in protein A chromatography for COGs reduction.

Resin aging is a common occurrence in chromatographic processes and generally influenced by factors such as cleaning procedure and composition of the feed stream. Two major events occur along with protein fouling, one is the loss of protein A ligand and the other is non-specific, irreversible interactions of foulants with resin particles. Both these are responsible for resin aging. As a result, the performance of the resin suffers a fall, and this can be quantified through indicators like reduction in dynamic binding capacity, increased column pressure, or peak broadening. The number of reuse cycles of a resin has a major influence on the cost per batch. This is even more significant in the case of protein A resin, which is the primary cost driver for downstream processing. In this work, we first identify chromatogram characteristics that correlate to resin aging. Next, we propose a data monitoring-based tool for prediction of resin aging. Principal component analysis of the UV data of Mab 1 showed a deviation at 120th cycle and an out of specification at around 149th cycle, corroborating with yield decline. Batch level modelling could deliver a predictable trend for resin aging and was demonstrated for two different Mabs (Mab1 and Mab2). The results demonstrate that significant resin aging can be detected 20-25 cycles prior to observable yield decline. A control strategy has been suggested such that once the deviation has been detected, additional resin cleaning is triggered. Overall, a 50-100 Protein A cycle enhancement in resin lifespan could be achieved.

Factors affecting product association as a mechanism of host-cell protein persistence in bioprocessing.

Product association of host-cell proteins (HCPs) to monoclonal antibodies (mAbs) is widely regarded as a mechanism that can enable HCP persistence through multiple purification steps and even into the final drug substance. Discussion of this mechanism often implies that the existence or extent of persistence is directly related to the strength of binding but actual measurements of the binding affinity of such interactions remain sparse. Two separate avenues of investigation of HCP-mAb binding are reported here. One is the measurement of the affinity of binding of individual, commonly persistent Chinese hamster ovary (CHO) HCPs to each of a set of mAbs, and the other uses quantitative proteomic measurements to assess binding of HCPs in a null CHO harvested cell culture fluid (HCCF) to mAbs produced in the same cell line. The individual HCP measurements show that the binding affinities of individual HCPs to different mAbs can vary appreciably but are rarely very high, with only weak pH dependence. The measurements on the null HCCF allow estimation of individual HCP-mAb affinities; these are typically weaker than those seen in affinity measurements on isolated HCPs. Instead, the extent of binding appears correlated with the initial abundance of individual HCPs in the HCCF and the forms of the HCPs in the solution, i.e., whether HCPs are present as free molecules or as parts of large aggregates. Separate protein A chromatography experiments performed by feeding different fractions of a mAb-containing HCCF obtained by size-exclusion chromatography (SEC) showed clear differences in the number and identity of HCPs found in the protein A eluate. These results indicate a significant role for HCP-mAb association in determining HCP persistence through protein A chromatography, presumably through binding of HCP-mAb complexes to the resin. Overall, the results illustrate the importance of considering more fully the biophysical context of HCP-product association in assessing the factors that may affect the phenomenon and determine its implications. Knowledge of the abundances and the forms of individual or aggregated HCPs in HCCF are particularly significant, emphasizing the integration of upstream and downstream bioprocessing and the importance of understanding the collective properties of HCPs in addition to just the biophysical properties of individual HCPs.

The beneficial impact of kosmotropic salts on the resolution and selectivity of Protein A chromatography.

During the manufacturing of therapeutic antibodies, effective Protein A chromatography as initial column step is crucial to simplify the remaining purification effort for subsequent polishing steps. This is particularly relevant for molecules with high impurity content so that desired product purity can be attained. The present study demonstrates beneficial effects on impurity removal when applying kosmotropic salts, e.g., sodium sulfate or sodium chloride, in the elution phase. Initially, a screen using negative linear pH gradient elution evaluated the impact of the kosmotropic salts in comparison to no additive and chaotropic urea using three mAbs and three common resins. Retaining acceptable yield, the kosmotropic salts improved resolution of monomer and impurities and reduced the contents of process-related host cell proteins and DNA as well as of product-related low and high molecular weight forms, despite some resin- and mAb-dependent variations. Moreover, a decrease in hydrolytic activity measured by a new assay for polysorbase activity was observed. In contrast, urea was hardly effective. The findings served to establish optimized step elution conditions with 0.25 M of sodium sulfate for a challenging mAb with complex format (bispecific 2 + 1 CrossMab) displaying high relative hydrophobicity and impurity levels. With yield and purity both in the range of 90 %, the contents of all impurity components were reduced, e.g., low molecular weight forms by two-fold and polysorbase activity by four-fold. The study indicates the potential of kosmotropic salts to establish efficient and comprehensive impurity separation by Protein A for facilitated downstream processing and economic manufacturing of complex antibodies.

Acidic pH or salt treatment can convert soluble antibody aggregates in culture harvest into monomers and improve Protein A chromatography step yield.

While purifying a regular monospecific antibody, we found that the Protein A step yield was much lower than expected. Further studies revealed that the antibody formed large-size aggregates that did not bind to the Protein A resin, hence leading to dropped recovery. In an attempt to solve this low yield issue, we found that mildly acidic pH or ammonium sulfate treatment can partially convert the aggregates into monomers. In addition, when acidic pH treated culture harvest was processed by Protein A chromatography, the yield was restored to the normal range, suggesting that the monomers recovered from aggregates regained Protein A binding capability. Thus, low pH treatment of culture harvest can be potentially used as a general approach for improving Protein A step yield in cases where non-binding antibody aggregates are formed through noncovalent interactions.