Cell culture process metabolomics together with multivariate data analysis tools opens new routes for bioprocess development and glycosylation prediction.

Multivariate latent variable methods have become a popular and versatile toolset to analyze bioprocess data in industry and academia. This work spans such applications from the evaluation of the role of the standard process variables and metabolites to the metabolomics level, that is, to the extensive number metabolic compounds detectable in the extracellular and intracellular domains. Given the substantial effort currently required for the measurement of the latter groups, a tailored methodology is presented that is capable of providing valuable process insights as well as predicting the glycosylation profile based on only four experiments measured over 12 cell culture days. An important result of the work is the possibility to accurately predict many of the glycan variables based on the information of three experiments. An additional finding is that such predictive models can be generated from the more accessible process and extracellular information only, that is, without including the more experimentally cumbersome intracellular data. With regards to the incorporation of omics data in the standard process analytics framework in the future, this works provides a comprehensive data analysis pathway which can efficiently support numerous bioprocessing tasks.

Proteomic analysis of micro-scale bioreactors as scale-down model for a mAb producing CHO industrial fed-batch platform.

The pharmaceutical production of recombinant proteins, such as monoclonal antibodies, is rather complex and requires proper development work. Accordingly, it is essential to develop appropriate scale-down models, which can mimic the corresponding production scale. In this work, we investigated the impact of the bioreactor scale on intracellular micro-heterogeneities of a CHO cell line producing monoclonal antibodies in fed-batch mode, using a 10 mL micro-bioreactor (ambr™) scale-down model and the corresponding 300 L pilot-scale bioreactor. For each scale, we measured the time evolution of the proteome, which enabled us to compare the impact of the bioreactor scale on the intracellular processes. Nearly absolute accordance between the scales was verified by data mining methods, such as hierarchical clustering and in-detail analysis on a single protein base. The time response of principal enzymes related to N-glycosylation was discussed, emphasizing major dissimilarities between the glycan fractions adorning the heavy chain and the corresponding protein abundance. The enzyme expression displayed mainly a constant profile, whereas the resulting glycan pattern changed over time. It is concluded that the enzymatic activity is influenced by the changing environmental conditions present in the fed-batch processes leading to the observed time-dependent variation.

Glycosylation flux analysis reveals dynamic changes of intracellular glycosylation flux distribution in Chinese hamster ovary fed-batch cultures.

N-linked glycosylation of proteins has both functional and structural significance. Importantly, the glycan structure of a therapeutic protein influences its efficacy, pharmacokinetics, pharmacodynamics and immunogenicity. In this work, we developed glycosylation flux analysis (GFA) for predicting intracellular production and consumption rates (fluxes) of glycoforms, and applied this analysis to CHO fed-batch immunoglobulin G (IgG) production using two different media compositions, with and without additional manganese feeding. The GFA is based on a constraint-based modeling of the glycosylation network, employing a pseudo steady state assumption. While the glycosylation fluxes in the network are balanced at each time point, the GFA allows the fluxes to vary with time by way of two scaling factors: (1) an enzyme-specific factor that captures the temporal changes among glycosylation reactions catalysed by the same enzyme, and (2) the cell specific productivity factor that accounts for the dynamic changes in the IgG production rate. The GFA of the CHO fed-batch cultivations showed that regardless of the media composition, galactosylation fluxes decreased with the cultivation time more significantly than the other glycosylation reactions. Furthermore, the GFA showed that the addition of Mn, a cofactor of galactosyltransferase, has the effect of increasing the galactosylation fluxes but only during the beginning of the cultivation period. The results thus demonstrated the power of the GFA in delineating the dynamic alterations of the glycosylation fluxes by local (enzyme-specific) and global (cell specific productivity) factors.

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. © 2017 Wiley Periodicals, Inc.

Robust factor selection in early cell culture process development for the production of a biosimilar monoclonal antibody.

This work presents a multivariate methodology combining principal component analysis, the Mahalanobis distance and decision trees for the selection of process factors and their levels in early process development of generic molecules. It is applied to a high throughput study testing more than 200 conditions for the production of a biosimilar monoclonal antibody at microliter scale. The methodology provides the most important selection criteria for the process design in order to improve product quality towards the quality attributes of the originator molecule. Robustness of the selections is ensured by cross-validation of each analysis step. The concluded selections are then successfully validated with an external data set. Finally, the results are compared to those obtained with a widely used software revealing similarities and clear advantages of the presented methodology. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 33:181-191, 2017.

Controlling the time evolution of mAb N-linked glycosylation, Part I: Microbioreactor experiments.

N-linked glycosylation is of key importance for the efficacy of many biotherapeutic proteins such as monoclonal antibodies (mAbs). Media components and cell culture conditions have been shown to significantly affect N-linked glycosylation during the production of glycoproteins using mammalian cell fed-batch cultures. These parameters inevitably change in modern industrial processes with concentrated feed additions and cell densities beyond 2 × 107 cells/mL. In order to control the time-dependent changes of protein glycosylation, an automated microbioreactor system was used to investigate the effects of culture pH, ammonia, galactose, and manganese chloride supplementation on nucleotide sugars as well as mAb N-linked glycosylation in a time-dependent way. Two different strategies comprising of a single shift of culture conditions as well as multiple media supplementations along the culture duration were applied to obtain changing and constant glycosylation profiles. The different feeding approaches enabled constant glycosylation patterns throughout the entire culture duration at different levels. By modulating the time evolution of the mAb glycan pattern, not only the endpoint but also the ratios between different glycosylation structures could be modified. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:1123-1134, 2016.

Controlling the time evolution of mAb N-linked glycosylation - Part II: Model-based predictions.

N-linked glycosylation is known to be a crucial factor for the therapeutic efficacy and safety of monoclonal antibodies (mAbs) and many other glycoproteins. The nontemplate process of glycosylation is influenced by external factors which have to be tightly controlled during the manufacturing process. In order to describe and predict mAb N-linked glycosylation patterns in a CHO-S cell fed-batch process, an existing dynamic mathematical model has been refined and coupled to an unstructured metabolic model. High-throughput cell culture experiments carried out in miniaturized bioreactors in combination with intracellular measurements of nucleotide sugars were used to tune the parameter configuration of the coupled models as a function of extracellular pH, manganese and galactose addition. The proposed modeling framework is able to predict the time evolution of N-linked glycosylation patterns during a fed-batch process as a function of time as well as the manipulated variables. A constant and varying mAb N-linked glycosylation pattern throughout the culture were chosen to demonstrate the predictive capability of the modeling framework, which is able to quantify the interconnected influence of media components and cell culture conditions. Such a model-based evaluation of feeding regimes using high-throughput tools and mathematical models gives rise to a more rational way to control and design cell culture processes with defined glycosylation patterns. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:1135-1148, 2016.

High-throughput profiling of nucleotides and nucleotide sugars to evaluate their impact on antibody N-glycosylation.

Recent advances in miniaturized cell culture systems have facilitated the screening of media additives on productivity and protein quality attributes of mammalian cell cultures. However, intracellular components are not routinely measured due to the limited throughput of available analytical techniques. In this work, time profiling of intracellular nucleotides and nucleotide sugars of CHO-S cell fed-batch processes in a micro-scale bioreactor system was carried out using a recently developed high-throughput method based on matrix-assisted laser desorption/ionization (MALDI) time-of-flight mass spectrometry (TOF-MS). Supplementation of various media additives significantly altered the intracellular nucleotides and nucleotide sugars that are inextricably linked to the process of glycosylation. The results revealed that UDP-Gal synthesis appeared to be particularly limiting whereas the impact of elevated UDP-GlcNAc and GDP-Fuc levels on the final glycosylation patterns was only marginally important. In contrast, manganese and asparagine supplementation altered the glycan profiles without affecting intracellular components. The combination of miniaturized cell cultures and high-throughput analytical techniques serves therefore as a useful tool for future quality driven media optimization studies.

Screening and assessment of performance and molecule quality attributes of industrial cell lines across different fed-batch systems.

The major challenge in the selection process of recombinant cell lines for the production of biologics is the choice, early in development, of a clonal cell line presenting a high productivity and optimal cell growth. Most importantly, the selected candidate needs to generate a product quality profile which is adequate with respect to safety and efficacy and which is preserved across cell culture scales. We developed a high-throughput screening and selection strategy of recombinant cell lines, based on their productivity in shaking 96-deepwell plates operated in fed-batch mode, which enables the identification of cell lines maintaining their high productivity at larger scales. Twelve recombinant cell lines expressing the same antibody with different productivities were selected out of 470 clonal cell lines in 96-deepwell plate fed-batch culture. They were tested under the same conditions in 50 mL vented shake tubes, microscale and lab-scale bioreactors in order to confirm the maintenance of their performance at larger scales. The use of a feeding protocol and culture conditions which are essentially the same across the different scales was essential to maintain productivity and product quality profiles across scales. Compared to currently used approaches, this strategy has the advantage of speeding up the selection process and increases the number of screened clones for getting high-producing recombinant cell lines at manufacturing scale with the desired performance and quality.

Pilot-scale verification of maximum tolerable hydrodynamic stress for mammalian cell culture.

Although several scaling bioreactor models of mammalian cell cultures are suggested and described in the literature, they mostly lack a significant validation at pilot or manufacturing scale. The aim of this study is to validate an oscillating hydrodynamic stress loop system developed earlier by our group for the evaluation of the maximum operating range for stirring, based on a maximum tolerable hydrodynamic stress. A 300-L pilot-scale bioreactor for cultivation of a Sp2/0 cell line was used for this purpose. Prior to cultivations, a stress-sensitive particulate system was applied to determine the stress values generated by stirring and sparging. Pilot-scale data, collected from 7- to 28-Pa maximum stress conditions, were compared with data from classical 3-L cultivations and cultivations from the oscillating stress loop system. Results for the growth behavior, analyzed metabolites, productivity, and product quality showed a dependency on the different environmental stress conditions but not on reactor size. Pilot-scale conditions were very similar to those generated in the oscillating stress loop model confirming its predictive capability, including conditions at the edge of failure.

Fingerprint detection and process prediction by multivariate analysis of fed-batch monoclonal antibody cell culture data.

This work presents a sequential data analysis path, which was successfully applied to identify important patterns (fingerprints) in mammalian cell culture process data regarding process variables, time evolution and process response. The data set incorporates 116 fed-batch cultivation experiments for the production of a Fc-Fusion protein. Having precharacterized the evolutions of the investigated variables and manipulated parameters with univariate analysis, principal component analysis (PCA) and partial least squares regression (PLSR) are used for further investigation. The first major objective is to capture and understand the interaction structure and dynamic behavior of the process variables and the titer (process response) using different models. The second major objective is to evaluate those models regarding their capability to characterize and predict the titer production. Moreover, the effects of data unfolding, imputation of missing data, phase separation, and variable transformation on the performance of the models are evaluated.

Tailoring recombinant protein quality by rational media design.

Clinical efficacy and safety of recombinant proteins are closely associated with their structural characteristics. The major quality attributes comprise glycosylation, charge variants (oxidation, deamidation, and C- & N-terminal modifications), aggregates, low-molecular-weight species (LMW), and misincorporation of amino acids in the protein backbone. Cell culture media design has a great potential to modulate these quality attributes due to the vital role of medium in mammalian cell culture. The purpose of this review is to provide an overview of the way both classical cell culture medium components and novel supplements affect the quality attributes of recombinant therapeutic proteins expressed in mammalian hosts, allowing rational and high-throughput optimization of mammalian cell culture media. A selection of specific and/or potent inhibitors and activators of oligosaccharide processing as well as components affecting multiple quality attributes are presented. Extensive research efforts in this field show the feasibility of quality engineering through media design, allowing to significantly modulate the protein function.

Determination of the maximum operating range of hydrodynamic stress in mammalian cell culture.

Application of quality by design (QbD) requires identification of the maximum operating range for parameters affecting the cell culture process. These include hydrodynamic stress, mass transfer or gradients in dissolved oxygen and pH. Since most of these are affected by the impeller design and speed, the main goal of this work was to identify a maximum operating range for hydrodynamic stress, where no variation of cell growth, productivity and product quality can be ensured. Two scale-down models were developed operating under laminar and turbulent condition, generating repetitive oscillating hydrodynamic stress with maximum stress values ranging from 0.4 to 420Pa, to compare the effect of the different flow regimes on the cells behavior. Two manufacturing cell lines (CHO and Sp2/0) used for the synthesis of therapeutic proteins were employed in this study. For both cell lines multiple process outputs were used to determine the threshold values of hydrodynamic stress, such as cell growth, morphology, metabolism and productivity. They were found to be different in between the cell lines with values equal to 32.4±4.4Pa and 25.2±2.4Pa for CHO and Sp2/0, respectively. Below the measured thresholds both cell lines do not show any appreciable effect of the hydrodynamic stress on any critical quality attribute, while above, cells responded negatively to the elevated stress. To confirm the applicability of the proposed method, the obtained results were compared with data generated from classical small-scale reactors with a working volume of 3L.

Modulation of mAb quality attributes using microliter scale fed-batch cultures.

A high-throughput DoE approach performed in a 96-deepwell plate system was used to explore the impact of media and feed components on main quality attributes of a monoclonal antibody. Six CHO-S derived clonal cell lines expressing the same monoclonal antibody were tested in two different cell culture media with six components added at three different levels. The resulting 384 culture conditions including controls were simultaneously tested in fed-batch conditions, and process performance such as viable cell density, viability, and product titer were monitored. At the end of the culture, supernatants from each condition were purified and the product was analyzed for N-glycan profiles, charge variant distribution, aggregates, and low molecular weight forms. The screening described here provided highly valuable insights into the factors and combination of factors that can be used to modulate the quality attributes of a molecule. The approach also revealed specific intrinsic differences of the selected clonal cell lines - some cell lines were very responsive in terms of changes in performance or quality attributes, whereas others were less affected by the factors tested in this study. Moreover, it indicated to what extent the attributes can be impacted within the selected experimental design space. The outcome correlated well with confirmations performed in larger cell culture volumes such as small-scale bioreactors. Being fast and resource effective, this integrated high-throughput approach can provide information which is particularly useful during early stage cell culture development.

Tools for high-throughput process and medium optimization.

Shaking technology in combination with small-scale disposable plastic vessels has become a notable bioprocess optimization tool widely exploited for cells grown in suspension. This chapter focuses on the two most accommodating culture systems: 50 mL centrifugation tubes and 96-deepwell plates. Used by many laboratories for routine passaging of suspension cultures and all types of optimization experiments as flexible culture system, the 50 mL shake tubes are the preferred vessels for manual manipulations, while microtiter plates are the ideal containers when automatic liquid handling systems are available. Both culture systems can offer to the cells a well-mixed environment which is close to the conditions found at larger scale in production bioreactors.

A high-throughput media design approach for high performance mammalian fed-batch cultures.

An innovative high-throughput medium development method based on media blending was successfully used to improve the performance of a Chinese hamster ovary fed-batch medium in shaking 96-deepwell plates. Starting from a proprietary chemically-defined medium, 16 formulations testing 43 of 47 components at 3 different levels were designed. Media blending was performed following a custom-made mixture design of experiments considering binary blends, resulting in 376 different blends that were tested during both cell expansion and fed-batch production phases in one single experiment. Three approaches were chosen to provide the best output of the large amount of data obtained. A simple ranking of conditions was first used as a quick approach to select new formulations with promising features. Then, prediction of the best mixes was done to maximize both growth and titer using the Design Expert software. Finally, a multivariate analysis enabled identification of individual potential critical components for further optimization. Applying this high-throughput method on a fed-batch, rather than on a simple batch, process opens new perspectives for medium and feed development that enables identification of an optimized process in a short time frame.

Lactate metabolism shift in CHO cell culture: the role of mitochondrial oxidative activity.

Lactate production is monitored in industrial processes as a crucial metabolite for cultured mammalian cells. Typically lactate is strongly produced during the exponential growth phase, while its net consumption is frequently observed when cells enter into the stationary phase. Such a metabolic shift is desirable because it seems to favor optimal process performance. However, this shift is neither generic nor can it be easily controlled, as the mechanisms modulating lactate production/consumption in cell culture are still under investigation. In this study different lactate profiles were observed in a chemically defined medium for the parental CHO-S cells and a non-recombinant subclone. The initial lactate production phase, which is typical for fast growing cells, was similar for both cell lines. After glutamine depletion the situation changed: the parental cell line promptly switched to net lactate consumption, whereas the subclone continued to produce lactate until glucose was depleted as well. We speculated that the extra lactate production would be ascribed to a different mitochondrial oxidative capacity in the subclone. Therefore, the mitochondrial membrane potential and oxygen consumption were measured for both cell lines. Indeed, a correlation between high lactate production and a reduced oxidative metabolism was found. Interestingly, this particular metabolic phenotype was also strongly influenced by the medium composition: both cell lines underwent a switch to lactate consumption when cultivated in a second medium, while a third one promoted continuous lactate production even for the parental CHO cells. Again, the correlation between lactate profile and oxidative metabolism was confirmed, pointing to a central role of mitochondria on lactate metabolism.

Effect of hydrocortisone on the production and glycosylation of an Fc-fusion protein in CHO cell cultures.

Glucocorticoids are known to modulate various cellular functions such as cell proliferation, metabolism, glycosylation, and secretion of many proteins. We tested the effect of hydrocortisone (HC) on cell growth, viability, metabolism, protein production, and glycosylation of an Fc-protein expressing Chinese hamster ovary (CHO) cell culture. HC extended cell viability but impaired cell growth. The inhibitory effect on cell growth was dose-dependent and decreased when the glucocorticoid addition was delayed. When HC was added after 2 or 3 days of culture, an increase in glutamate consumption was observed, which was reversed by the glucocorticoid receptor antagonist mifepristone (Mif). Titer and specific productivity increased in the presence of HC. The increase in titer was only slightly reversed by Mif. On the other hand, Mif by itself induced an increase in titer to a level comparable to or higher than HC. Protein glycosylation was altered by the glucocorticoid in a dose- and time-dependent manner, with a shift to more acidic bands, which correlated with an increase in sialic acid moieties. This increase, which was not linked to a decrease in extracellular sialidase activity in HC-treated cultures, was reversed by Mif. Predictive models based on design of experiments enabled the definition of optimal conditions for process performance in terms of viability and titer and for the quality of the Fc-fusion protein in terms of glycosylation. The data obtained suggest a use of glucocorticoids for commercial production of Fc-fusion proteins expressed in CHO cells.

Innovative, non-stirred bioreactors in scales from milliliters up to 1000 liters for suspension cultures of cells using disposable bags and containers--a Swiss contribution.

Innovative mixing principles in bioreactors, for example using the rocking of a platform to induce a backwards and forwards 'wave', or using orbital shaking to generate a 'wave' that runs round in a cylindrical container, have proved to be successful for the suspension cultures of cells, especially when combined with disposable materials. This article presents an overview of the engineering characteristics when these new principles are applied in bioreactors, and case studies covering scales of operation from milliliters to 1000 liters.

Use of orbital shaken disposable bioreactors for mammalian cell cultures from the milliliter-scale to the 1,000-liter scale.

Driven by the commercial success of recombinant biopharmaceuticals, there is an increasing demand for novel mammalian cell culture bioreactor systems for the rapid production of biologicals that require mammalian protein processing. Recently, orbitally shaken bioreactors at scales from 50 mL to 1,000 L have been explored for the cultivation of mammalian cells and are considered to be attractive alternatives to conventional stirred-tank bioreactors because of increased flexibility and reduced costs. Adequate oxygen transfer capacity was maintained during the scale-up, and strategies to increase further oxygen transfer rates (OTR) were explored, while maintaining favorable mixing parameters and low-stress conditions for sensitive lipid membrane-enclosed cells. Investigations from process development to the engineering properties of shaken bioreactors are underway, but the feasibility of establishing a robust, standardized, and transferable technical platform for mammalian cell culture based on orbital shaking and disposable materials has been established with further optimizations and studies ongoing.