Daniel Karst

Process performance and product quality in an integrated continuous antibody production process.

Continuous manufacturing is currently being seriously considered in the biopharmaceutical industry as the possible new paradigm for producing therapeutic proteins, due to production cost and product quality related benefits. In this study, a monoclonal antibody producing CHO cell line was cultured in perfusion mode and connected to a continuous affinity capture step. The reliable and stable integration of the two systems was enabled by suitable control loops, regulating the continuous volumetric flow and adapting the operating conditions of the capture process. For the latter, an at‐line HPLC measurement of the harvest concentration subsequent to the bioreactor was combined with a mechanistic model of the capture chromatographic unit. Thereby, optimal buffer consumption and productivity throughout the process was realized while always maintaining a yield above the target value of 99%. Stable operation was achieved at three consecutive viable cell density set points (20, 60, and 40 × 106 cells/mL), together with consistent product quality in terms of aggregates, fragments, charge isoforms, and N‐linked glycosylation. In addition, different values for these product quality attributes such as N‐linked glycosylation, charge variants, and aggregate content were measured at the different steady states. As expected, the amount of released DNA and HCP was significantly reduced by the capture step for all considered upstream operating conditions. This study is exemplary for the potential of enhancing product quality control and modulation by integrated continuous manufacturing. Biotechnol. Bioeng. 2017;114: 298–307.

Model based adaptive control of a continuous capture process for monoclonal antibodies production

A two-column capture process for continuous processing of cell-culture supernatant is presented. Similar to other multicolumn processes, this process uses sequential countercurrent loading of the target compound in order maximize resin utilization and productivity for a given product yield. The process was designed using a novel mechanistic model for affinity capture, which takes both specific adsorption as well as transport through the resin beads into account. Simulations as well as experimental results for the capture of an IgG antibody are discussed. The model was able to predict the process performance in terms of yield, productivity and capacity utilization. Compared to continuous capture with two columns operated batch wise in parallel, a 2.5-fold higher capacity utilization was obtained for the same productivity and yield. This results in an equal improvement in product concentration and reduction of buffer consumption. The developed model was used not only for the process design and optimization but also for its online control. In particular, the unit operating conditions are changed in order to maintain high product yield while optimizing the process performance in terms of capacity utilization and buffer consumption also in the presence of changing upstream conditions and resin aging.

Design and operation of a continuous integrated monoclonal antibody production process

The realization of an end‐to‐end integrated continuous lab‐scale process for monoclonal antibody manufacturing is described. For this, a continuous cultivation with filter‐based cell‐retention, a continuous two column capture process, a virus inactivation step, a semi‐continuous polishing step (twin‐column MCSGP), and a batch‐wise flow‐through polishing step were integrated and operated together. In each unit, the implementation of internal recycle loops allows to improve the performance: (a) in the bioreactor, to simultaneously increase the cell density and volumetric productivity, (b) in the capture process, to achieve improved capacity utilization at high productivity and yield, and (c) in the MCSGP process, to overcome the purity‐yield trade‐off of classical batch‐wise bind‐elute polishing steps. Furthermore, the design principles, which allow the direct connection of these steps, some at steady state and some at cyclic steady state, as well as straight‐through processing, are discussed. The setup was operated for the continuous production of a commercial monoclonal antibody, resulting in stable operation and uniform product quality over the 17 cycles of the end‐to‐end integration. The steady‐state operation was fully characterized by analyzing at the outlet of each unit at steady state the product titer as well as the process (HCP, DNA, leached Protein A) and product (aggregates, fragments) related impurities.

Modulation and modeling of monoclonal antibody N-linked glycosylation in mammalian cell perfusion reactors

Mammalian cell perfusion cultures are gaining renewed interest as an alternative to traditional fed‐batch processes for the production of therapeutic proteins, such as monoclonal antibodies (mAb). The steady state operation at high viable cell density allows the continuous delivery of antibody product with increased space‐time yield and reduced in‐process variability of critical product quality attributes (CQA). In particular, the production of a confined mAb N‐linked glycosylation pattern has the potential to increase therapeutic efficacy and bioactivity. In this study, we show that accurate control of flow rates, media composition and cell density of a Chinese hamster ovary (CHO) cell perfusion bioreactor allowed the production of a constant glycosylation profile for over 20 days. Steady state was reached after an initial transition phase of 6 days required for the stabilization of extra‐ and intracellular processes. The possibility to modulate the glycosylation profile was further investigated in a Design of Experiment (DoE), at different viable cell density and media supplement concentrations. This strategy was implemented in a sequential screening approach, where various steady states were achieved sequentially during one culture. It was found that, whereas high ammonia levels reached at high viable cell densities (VCD) values inhibited the processing to complex glycan structures, the supplementation of either galactose, or manganese as well as their synergy significantly increased the proportion of complex forms. The obtained experimental data set was used to compare the reliability of a statistical response surface model (RSM) to a mechanistic model of N‐linked glycosylation. The latter outperformed the response surface predictions with respect to its capability and reliability in predicting the system behavior (i.e., glycosylation pattern) outside the experimental space covered by the DoE design used for the model parameter estimation. Therefore, we can conclude that the modulation of glycosylation in a sequential steady state approach in combination with mechanistic model represents an efficient and rational strategy to develop continuous processes with desired N‐linked glycosylation patterns. Biotechnol. Bioeng. 2017;114: 1978–1990.

Continuous integrated manufacturing of therapeutic proteins

With the growth of our understanding of biopharmaceutical processes, a transition from classical batch to continuous integrated manufacturing of therapeutic proteins is taking place across laboratory, clinical and commercial scales. Encouraged by regulatory authorities, this transition is favoured by new emerging technologies as well as by the development of better simulation models. The current status of continuous cell culture and downstream processes and requirements for their successful integration are discussed in this article, with specific reference to product quality attributes.

Intracellular CHO Cell Metabolite Profiling Reveals Steady-State Dependent Metabolic Fingerprints in Perfusion Culture

Perfusion cell culture processes allow the steady‐state culture of mammalian cells at high viable cell density, which is beneficial for overall product yields and homogeneity of product quality in the manufacturing of therapeutic proteins. In this study, the extent of metabolic steady state and the change of the metabolite profile between different steady states of an industrial Chinese hamster ovary (CHO) cell line producing a monoclonal antibody (mAb) was investigated in stirred tank perfusion bioreactors. Matrix‐assisted laser desorption/ionization time of flight mass spectrometry (MALDI‐TOF‐MS) of daily cell extracts revealed more than a hundred peaks, among which 76 metabolites were identified by tandem MS (MS/MS) and high resolution Fourier transform ion cyclotron resonance (FT‐ICR) MS. Nucleotide ratios (Uridine (U)‐ratio, nucleotide triphosphate (NTP)‐ratio and energy charge (EC)) and multivariate analysis of all features indicated a consistent metabolite profile for a stable culture performed at 40 × 106 cells/mL over 26 days of culture. Conversely, the reactor was operated continuously so as to reach three distinct steady states one after the other at 20, 60, and 40 × 106 cells/mL. In each case, a stable metabolite profile was achieved after an initial transient phase of approximately three days at constant cell density when varying between these set points. Clear clustering according to cell density was observed by principal component analysis, indicating steady‐state dependent metabolite profiles. In particular, varying levels of nucleotides, nucleotide sugar, and lipid precursors explained most of the variance between the different cell density set points.

Microarray-based MALDI-TOF mass spectrometry enables monitoring of monoclonal antibody production in batch and perfusion cell cultures.

Cell culture process monitoring in monoclonal antibody (mAb) production is essential for efficient process development and process optimization. Currently employed online, at line and offline methods for monitoring productivity as well as process reproducibility have their individual strengths and limitations. Here, we describe a matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS)-based on a microarray for mass spectrometry (MAMS) technology to rapidly monitor a broad panel of analytes, including metabolites and proteins directly from the unpurified cell supernatant or from host cell culture lysates. The antibody titer is determined from the intact antibody mass spectra signal intensity relative to an internal protein standard spiked into the supernatant. The method allows a semi-quantitative determination of light and heavy chains. Intracellular mass profiles for metabolites and proteins can be used to track cellular growth and cell productivity.

Isotope labeling to determine the dynamics of metabolic response in CHO cell perfusion bioreactors using MALDI-TOF-MS

The steady‐state operation of Chinese hamster ovary (CHO) cells in perfusion bioreactors requires the equilibration of reactor dynamics and cell metabolism. Accordingly, in this work we investigate the transient cellular response to changes in its environment and their interactions with the bioreactor hydrodynamics. This is done in a benchtop perfusion bioreactor using MALDI‐TOF MS through isotope labeling of complex intracellular nucleotides (ATP, UTP) and nucleotide sugars (UDP‐Hex, UDP‐HexNAc). By switching to a 13C6 glucose containing feed media during constant operation at 20 × 106 cells and a perfusion rate of 1 reactor volume per day, isotopic steady state was studied. A step change to the 13C6 glucose medium in spin tubes allowed the determination of characteristic times for the intracellular turnover of unlabeled metabolites pools, τST (≤0.56 days), which were confirmed in the bioreactor. On the other hand, it is shown that the reactor residence time τR (1 day) and characteristic time for glucose uptake τGlc (0.33 days), representative of the bioreactor dynamics, delayed the consumption of 13C6 glucose in the bioreactor and thus the intracellular 13C enrichment. The proposed experimental approach allowed the decoupling of bioreactor hydrodynamics and intrinsic dynamics of cell metabolism in response to a change in the cell culture environment.

Development of a shake tube-based scale-down model for perfusion cultures: WOLF et al.

The use of benchtop bioreactors (BRs) for the development of mammalian cell perfusion cultures is expensive and time consuming, given its complexity in equipment and operation. Scale‐down models, going from liter to milliliter scale, are needed to support the rapid determination of suitable operating conditions in terms of viable cell density (VCD), perfusion rate, and medium composition. In this study, we compare the performance of steady‐state perfusion cultures in orbitally shaken tube and BR systems for a given Chinese hamster ovary cell line. The developed scale‐down model relied on a daily workflow designed to keep the VCD constant at specific target values. This includes: cell count, removal of excessive cells (bleeding), spin down of remaining cells, harvest of cell‐free supernatant, and resuspension in fresh medium. Steady‐state cultures at different VCD values, medium exchange rates and working volumes were evaluated. Shake‐tube perfusion cultures allowed the prediction of cell‐specific growth, glucose consumption, ammonia, and monoclonal antibody production rates for much larger BRs, but not lactate (LAC) production rates. Although charge variant profiles remained comparable, different glycosylation patterns were obtained. The differences in LAC production and glycosylation probably resulted from the discontinuous medium exchange, the poor carbon dioxide removal, and the deficient pH control. Therefore, if requested by the specific process to be developed, product quality has to be fine‐tuned directly in the BR system. Altogether, the developed strategy provides a useful scale‐down model for the design and optimization of perfusion cultures with strong savings in time and media consumption.