Fermentanomics informed amino acid supplementation of an antibody producing mammalian cell culture

Fermentation Analytical Technology (FAT): Monitoring industrial E. coli fermentations using absolute quantitative 1H NMR spectroscopy

BACKGROUND

To perform fast, reproducible, and absolute quantitative measurements in an automated manner has become of paramount importance when monitoring industrial processes, including fermentations. Due to its numerous advantages - including its inherent quantitative nature - Proton Nuclear Magnetic Resonance (1H NMR) spectroscopy provides an ideal tool for the time-resolved monitoring of fermentations. However, analytical conditions, including non-automated sample preparation and long relaxation times (T1) of some metabolites, can significantly lengthen the experimental time and make implementation in an industrial set up unfeasible.

RESULTS

We present a high throughput method based on Standard Operating Procedures (SOPs) and 1H NMR, which lays the foundation for what we call Fermentation Analytical Technology (FAT). Our method was developed for the accurate absolute quantification of metabolites produced during Escherichia coli industrial fermentations. The method includes: (1) a stopped flow system for non-invasive sample collection followed by sample quenching, (2) automatic robot-assisted sample preparation, (3) fast 1H NMR measurements, (4) metabolites quantification using multivariate curve resolution (MCR), and (5) metabolites absolute quantitation using a novel correction factor (k) to compensate for the short recycle delay (D1) employed in the 1H NMR measurements. The quantification performance was tested using two sample types: buffer solutions of chemical standards and real fermentation samples. Five metabolites - glucose, acetate, alanine, phenylalanine and betaine - were quantified. Absolute quantitation ranged between 0.64 and 3.40 mM in pure buffer, and 0.71-7.76 mM in real samples.

SIGNIFICANCE

The proposed method is generic and can be straight forward implemented to other types of fermentations, such as lactic acid, ethanol and acetic acid fermentations. It provides a high throughput automated solution for monitoring fermentation processes and for quality control through absolute quantification of key metabolites in fermentation broth. It can be easily implemented in an at-line industrial setting, facilitating the optimization of the manufacturing process towards higher yields and more efficient and sustainable use of resources.

Complete Reduction to Amplitude Frequency Table (CRAFT)-A perspective

The CRAFT (Complete Reduction to Amplitude Frequency Table) technique, based on Bayesian analysis approach, converts FID and/or interferogram (time domain) to a frequency-amplitude table (tabular domain) in a robust, automated, and time-efficient fashion. This mini review/perspective presents an introduction to CRAFT as a processing workflow followed by a discussion of several practical 1D and 2D examples of its applicability and associated benefit. CRAFT provides high quality quantitative results for complex systems without any need for conventional preprocessing steps, such as phase and baseline corrections. Two-dimensional time domain data are typically truncated, particularly in the evolution dimension, and conventional processing after zero-filling and t_1max -matched apodization masks potentially available peak resolution. The line broadening introduced by extensive zero-filling and severe apodization functions leads to the lack of clear resolution of cross peaks. CRAFT decimation of interferograms, on the other hand, requires minimal or no apodization prior to extraction of the NMR parameters and significantly improves the spectral linewidth of the cross peaks along F₁ dimension compared to conventional (FT) processing. The tabular representation of the CRAFT2d cross peaks information can be visualized in a variety of frequency domain formats for conventional spectral interpretation as well as quantitative applications. A simple workflow to generate in silico oversampled interferogram (iSOS) is presented, and its potential benefit in CRAFT decimation of highly crowded 2D NMR is demonstrated. This report is meant as a collective thesis to present a potentially new paradigm in data processing that questions the need for hitherto unchallenged preprocessing steps, such as phase and baseline correction in 1D and zero-fill/severe apodization in 2D.

Computer-Aided Strategies for Determining the Amino Acid Composition of Medium for Chinese Hamster Ovary Cell-Based Biomanufacturing Platforms

Chinese hamster ovary (CHO) cells are used for the production of the majority of biopharmaceutical drugs, and thus have remained the standard industry host for the past three decades. The amino acid composition of the medium plays a key role in commercial scale biologics manufacturing, as amino acids constitute the building blocks of both endogenous and heterologous proteins, are involved in metabolic and non-metabolic pathways, and can act as main sources of nitrogen and carbon under certain conditions. As biomanufactured proteins become increasingly complex, the adoption of model-based approaches become ever more popular in complementing the challenging task of medium development. The extensively studied amino acid metabolism is exceptionally suitable for such model-driven analyses, and although still limited in practice, the development of these strategies is gaining attention, particularly in this domain. This paper provides a review of recent efforts. We first provide an overview of the widely adopted practice, and move on to describe the model-driven approaches employed for the improvement and optimization of the external amino acid supply in light of cellular amino acid demand. We conclude by proposing the likely prevalent direction the field is heading towards, providing a critical evaluation of the current state and the future challenges and considerations.

Real-time quantification and supplementation of bioreactor amino acids to prolong culture time and maintain antibody product quality

Real‐time monitoring of cell cultures in bioreactors can enable expedited responses necessary to correct potential batch failure perturbations which may normally go undiscovered until the completion of the batch and result in failure. Currently, analytical technologies are dedicated to real‐time monitoring of bioreactor parameters such as pH, dissolved oxygen, and temperature, nutrients such as glucose and glutamine, or metabolites such as lactate. Despite the importance of amino acids as the building blocks of therapeutic protein products, other than glutamine their concentrations are not commonly measured. Here, we present a study into amino acid monitoring, supplementation strategies, and how these techniques may impact the cell growth profiles and product quality. We used preliminary bioreactor runs to establish baselines by determining initial amino acid consumption patterns, the results of which were used to select a pool of amino acids which gets depleted in the bioreactor. These amino acids were combined into blends which were supplemented into bioreactors during a subsequent run, the concentrations of which were monitored using a mass spectrometry based at‐line method we developed to quickly assess amino acid concentrations from crude bioreactor media. We found that these blends could prolong culture life, reversing a viable cell density decrease that was leading to batch death. Additionally, we assessed how these strategies might impact protein product quality, such as the glycan profile. The amino acid consumption data were aligned with the final glycan profiles in principal component analysis to identify which amino acids are most closely associated with glycan outcomes.

Prediction of N-linked Glycoform Profiles of Monoclonal Antibody with Extracellular Metabolites and Two-Step Intracellular Models

Monoclonal antibodies (mAbs) are commonly glycosylated and show varying levels of galactose attachment. A set of experiments in our work showed that the galactosylation level of mAbs was altered by the culture conditions of hybridoma cells. The uridine diphosphate galactose (UDP-Gal) is one of the substrates of galactosylation. Based on that, we proposed a two-step model to predict N-linked glycoform profiles by solely using extracellular metabolites from cell culture. At the first step, the flux level of UDP-Gal in each culture was estimated based on a computational flux balance analysis (FBA); its level was found to be linear with the galactosylation degree on mAbs. At the second step, the glycoform profiles especially for G0F (agalactosylated), G1F (monogalactosylated) and G2F (digalactosylated) were predicted by a kinetic model. The model outputs well matched with the experimental data. Our study demonstrated that the integrated mathematical approach combining FBA and kinetic model is a promising strategy to predict glycoform profiles for mAbs during cell culture processes.

Applications of a metabolic network model of mesenchymal stem cells for controlling cell proliferation and differentiation

Mesenchymal stem cells (MSCs) can be isolated from several tissues of adults. In addition, MSCs have the potential of differentiation into several cell types. Therefore, MSCs are very useful in stem cell therapy and regenerative medicine. MSCs have also been used as gene or protein carriers. As a result, maintaining MSCs in a desirable metabolic state has been the subject of several studies. Here, we used a genome scale metabolic network model of bone marrow derived MSCs for exploring the metabolism of these cells. We analyzed metabolic fluxes of the model in order to find ways of increasing stem cell proliferation and differentiation. Consequently, the experimental results were in consistency with computational results. Therefore, analyzing metabolic models was proven to be a promising field in biomedical researches of stem cells.

In-line monitoring of amino acids in mammalian cell cultures using raman spectroscopy and multivariate chemometrics models

The application of PAT for in-line monitoring of biopharmaceutical manufacturing operations has a central role in developing more robust and consistent processes. Various spectroscopic techniques have been applied for collecting real-time data from cell culture processes. Among these, Raman spectroscopy has been shown to have advantages over other spectroscopic techniques, especially in aqueous culture solutions. Measurements of several process parameters such as glucose, lactate, glutamine, glutamate, ammonium, osmolality and VCD using Raman-based chemometrics models have been reported in literature. The application of Raman spectroscopy, coupled with calibration models for amino acid measurement in cell cultures, has been assessed. The developed models cover four amino acids important for cell growth and production: tyrosine, tryptophan, phenylalanine and methionine. The chemometrics models based on Raman spectroscopy data demonstrate the significant potential for the quantification of tyrosine, tryptophan and phenylalanine. The model for methionine would have to be further refined to improve quantification.

What NMR can do in the biopharmaceutical industry

Nuclear magnetic resonance (NMR) spectroscopy has a unique capability to probe the primary and higher order molecular structure and the structural dynamics of biomolecules at an atomic resolution, and this capability has been greatly fortified over the last five decades by an astonishing NMR instrumental and methodological development. Because of these factors, NMR has become a primary tool for the structure investigation of biomolecules, spawning a whole scientific subfield dedicated to the subject. This role of NMR is by now well established and broadly appreciated, especially in the context of academic research dealing with proteins that are purified and isotope-labeled in order to facilitate the necessary sophisticated multidimensional NMR measurements. However, the more recent industrial development, manufacturing, and quality control of biopharmaceuticals provide a different framework for NMR. For example, protein drug substances are not isotope-labeled and are present in a medium of excipients, which make structural NMR measurements much more difficult. On the other hand, biotechnology involves many other analytical requirements that can be efficiently addressed by NMR. In this respect the scope and limitations of NMR are less well understood. Having the non-expert reader in mind, herein we wish to highlight the ways in which modern NMR can effectively support biotechnological developments. Our focus will be on biosimilar proteins, pointing out certain cases where its use is probably essential. Based partly on literature data, and partly on our own hands-on experience, this paper is intended to be a guide for choosing the proper NMR approach for analytical questions concerning the structural comparability of therapeutic proteins, monitoring technology-related impurities, protein quantification, analysis of spent media, identification of extractable and leachable components, etc. Also, we focus on critical considerations, particularly those coming from drug authority guidelines, which limit the use of the well-established NMR tools in everyday practice.

Selection of chemically defined media for CHO cell fed-batch culture processes

Two CHO cell clones derived from the same parental CHOBC® cell line and producing the same monoclonal antibody (BC-G, a low producing clone; BC-P, a high producing clone) were tested in four basal media in all possible combinations with three feeds (=12 conditions) in fed-batch cultures. Higher amino acid feeding did not always lead to higher mAb production. The two clones showed differences in cell physiology, metabolism and optimal medium-feed combinations. During the phase transitions of all cultures, cell metabolism showed a shift represented by lower specific consumption and production rates, except for the specific glucose consumption rate in cultures fed by Actifeed A/B. The BC-P clone fed by Actifeed A/B showed a threefold cell volume increase and an increase of the specific consumption rate of glucose in the stationary phase. Since feeding was based on glucose this resulted in accumulation of amino acids for this feed, while this did not occur for the poorer feed (EFA/B). The same feed also led to an increase of cell size for the BC-G clone, but to a lesser extent.

Solution state nuclear magnetic resonance spectroscopy for biological metabolism and pathway intermediate analysis

Using nuclear magnetic resonance (NMR) spectroscopy in the study of metabolism has been immensely popular in medical- and health-related research but has yet to be widely applied to more fundamental biological problems. This review provides some NMR background relevant to metabolism, describes why ¹H NMR spectra are complex as well as introducing relevant terminology and definitions. The applications and practical considerations of NMR metabolic profiling and ¹³C NMR-based flux analyses are discussed together with the elegant 'enzyme trap' approach for identifying novel metabolic pathway intermediates. The importance of sample preparation and data analysis are also described and explained with reference to data precision and multivariate analysis to introduce researchers unfamiliar with NMR and metabolism to consider this technique for their research interests. Finally, a brief glance into the future suggests NMR-based metabolism has room to expand in the 21st century through new isotope labels, and NMR technologies and methodologies.

Selection of chemically defined media for CHO cell fed-batch culture processes

Two CHO cell clones derived from the same parental CHO^BC® cell line and producing the same monoclonal antibody (BC-G, a low producing clone; BC-P, a high producing clone) were tested in four basal media in all possible combinations with three feeds (=12 conditions) in fed-batch cultures. Higher amino acid feeding did not always lead to higher mAb production. The two clones showed differences in cell physiology, metabolism and optimal medium-feed combinations. During the phase transitions of all cultures, cell metabolism showed a shift represented by lower specific consumption and production rates, except for the specific glucose consumption rate in cultures fed by Actifeed A/B. The BC-P clone fed by Actifeed A/B showed a threefold cell volume increase and an increase of the specific consumption rate of glucose in the stationary phase. Since feeding was based on glucose this resulted in accumulation of amino acids for this feed, while this did not occur for the poorer feed (EFA/B). The same feed also led to an increase of cell size for the BC-G clone, but to a lesser extent.

Hybridoma cell-culture and glycan profile dataset at various bioreactor conditions

This is an “11 factor-2 level-12 run” Plackett-Burman experimental design dataset. The dataset includes 11 engineering bioreactor parameters as input variables. These 11 factors were varied at 2 levels and 23 response variables that are glycan profile attributes, were measured “A Design Space Exploration for Control of Critical Quality Attributes of mAb” (H. Bhatia, E.K. Read, C.D. Agarabi, K.A. Brorson, S.C. Lute, S. Yoon S, 2016) [2].

Effect of amino acid supplementation on titer and glycosylation distribution in hybridoma cell cultures-Systems biology-based interpretation using genome-scale metabolic flux balance model and multivariate data analysis

Genome-scale flux balance analysis (FBA) is a powerful systems biology tool to characterize intracellular reaction fluxes during cell cultures. FBA estimates intracellular reaction rates by optimizing an objective function, subject to the constraints of a metabolic model and media uptake/excretion rates. A dynamic extension to FBA, dynamic flux balance analysis (DFBA), can calculate intracellular reaction fluxes as they change during cell cultures. In a previous study by Read et al. (2013), a series of informed amino acid supplementation experiments were performed on twelve parallel murine hybridoma cell cultures, and this data was leveraged for further analysis (Read et al., Biotechnol Prog. 2013;29:745-753). In order to understand the effects of media changes on the model murine hybridoma cell line, a systems biology approach is applied in the current study. Dynamic flux balance analysis was performed using a genome-scale mouse metabolic model, and multivariate data analysis was used for interpretation. The calculated reaction fluxes were examined using partial least squares and partial least squares discriminant analysis. The results indicate media supplementation increases product yield because it raises nutrient levels extending the growth phase, and the increased cell density allows for greater culture performance. At the same time, the directed supplementation does not change the overall metabolism of the cells. This supports the conclusion that product quality, as measured by glycoform assays, remains unchanged because the metabolism remains in a similar state. Additionally, the DFBA shows that metabolic state varies more at the beginning of the culture but less by the middle of the growth phase, possibly due to stress on the cells during inoculation. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:1163-1173, 2016.

Improved Stability of a Model IgG3 by DoE-Based Evaluation of Buffer Formulations

Formulating appropriate storage conditions for biopharmaceutical proteins is essential for ensuring their stability and thereby their purity, potency, and safety over their shelf-life. Using a model murine IgG3 produced in a bioreactor system, multiple formulation compositions were systematically explored in a DoE design to optimize the stability of a challenging antibody formulation worst case. The stability of the antibody in each buffer formulation was assessed by UV/VIS absorbance at 280 nm and 410 nm and size exclusion high performance liquid chromatography (SEC) to determine overall solubility, opalescence, and aggregate formation, respectively. Upon preliminary testing, acetate was eliminated as a potential storage buffer due to significant visible precipitate formation. An additional 2⁴ full factorial DoE was performed that combined the stabilizing effect of arginine with the buffering capacity of histidine. From this final DoE, an optimized formulation of 200 mM arginine, 50 mM histidine, and 100 mM NaCl at a pH of 6.5 was identified to substantially improve stability under long-term storage conditions and after multiple freeze/thaw cycles. Thus, our data highlights the power of DoE based formulation screening approaches even for challenging monoclonal antibody molecules.

A control strategy to investigate the relationship between specific productivity and high-mannose glycoforms in CHO cells

The integration of physiological knowledge into process control strategies is a cornerstone for the improvement of biopharmaceutical cell culture technologies. The present contribution investigates the applicability of specific productivity as a physiological control parameter in a cell culture process producing a monoclonal antibody (mAb) in CHO cells. In order to characterize cell physiology, the on-line oxygen uptake rate (OUR) was monitored and the time-resolved specific productivity was calculated as physiological parameters. This characterization enabled to identify the tight link between the deprivation of tyrosine and the decrease in cell respiration and in specific productivity. Subsequently, this link was used to control specific productivity by applying different feeding profiles. The maintenance of specific productivity at various levels enabled to identify a correlation between the rate of product formation and the relative abundance of high-mannose glycoforms. An increase in high mannose content was assumed to be the result of high specific productivity. Furthermore, the high mannose content as a function of cultivation pH and specific productivity was investigated in a design of experiment approach. This study demonstrated how physiological parameters could be used to understand interactions between process parameters, physiological parameters, and product quality attributes.

Designed Amino Acid Feed in Improvement of Production and Quality Targets of a Therapeutic Monoclonal Antibody

Cell culture feeds optimization is a critical step in process development of pharmaceutical recombinant protein production. Amino acids are the basic supplements of mammalian cell culture feeds with known effect on their growth promotion and productivity. In this study, we reported the implementation of the Plackett-Burman (PB) multifactorial design to screen the effects of amino acids on the growth promotion and productivity of a Chinese hamster ovary DG-44 (CHO-DG44) cell line producing bevacizumab. After this screening, the amino acid combinations were optimized by the response surface methodology (RSM) to determine the most effective concentration in feeds. Through this strategy, the final monoclonal antibody (mAb) titre was enhanced by 70%, compared to the control group. For this particular cell line, aspartic acid, glutamic acid, arginine and glycine had the highest positive effects on the final mAb titre. Simultaneously, the impact of the designed amino acid feed on some critical quality attributes of bevacizumab was examined in the group with highest productivity. The product was analysed for N-glycan profiles, charge variant distribution, and low molecular weight forms. The results showed that the target product quality has been improved using this feeding strategy. It was shown how this strategy could significantly diminish the time and number of experiments in identifying the most effective amino acids and related concentrations in target product enhancement. This model could be successfully applied to other components of culture media and feeds.

A correction method for systematic error in (1)H-NMR time-course data validated through stochastic cell culture simulation

Background

The growing ubiquity of metabolomic techniques has facilitated high frequency time-course data collection for an increasing number of applications. While the concentration trends of individual metabolites can be modeled with common curve fitting techniques, a more accurate representation of the data needs to consider effects that act on more than one metabolite in a given sample. To this end, we present a simple algorithm that uses nonparametric smoothing carried out on all observed metabolites at once to identify and correct systematic error from dilution effects. In addition, we develop a simulation of metabolite concentration time-course trends to supplement available data and explore algorithm performance. Although we focus on nuclear magnetic resonance (NMR) analysis in the context of cell culture, a number of possible extensions are discussed.

Results

Realistic metabolic data was successfully simulated using a 4-step process. Starting with a set of metabolite concentration time-courses from a metabolomic experiment, each time-course was classified as either increasing, decreasing, concave, or approximately constant. Trend shapes were simulated from generic functions corresponding to each classification. The resulting shapes were then scaled to simulated compound concentrations. Finally, the scaled trends were perturbed using a combination of random and systematic errors. To detect systematic errors, a nonparametric fit was applied to each trend and percent deviations calculated at every timepoint. Systematic errors could be identified at time-points where the median percent deviation exceeded a threshold value, determined by the choice of smoothing model and the number of observed trends. Regardless of model, increasing the number of observations over a time-course resulted in more accurate error estimates, although the improvement was not particularly large between 10 and 20 samples per trend. The presented algorithm was able to identify systematic errors as small as 2.5 % under a wide range of conditions.

Conclusion

Both the simulation framework and error correction method represent examples of time-course analysis that can be applied to further developments in ¹H-NMR methodology and the more general application of quantitative metabolomics.

Electronic supplementary material

The online version of this article (doi:10.1186/s12918-015-0197-4) contains supplementary material, which is available to authorized users.

Bioreactor process parameter screening utilizing a Plackett-Burman design for a model monoclonal antibody

Consistent high-quality antibody yield is a key goal for cell culture bioprocessing. This endpoint is typically achieved in commercial settings through product and process engineering of bioreactor parameters during development. When the process is complex and not optimized, small changes in composition and control may yield a finished product of less desirable quality. Therefore, changes proposed to currently validated processes usually require justification and are reported to the US FDA for approval. Recently, design-of-experiments-based approaches have been explored to rapidly and efficiently achieve this goal of optimized yield with a better understanding of product and process variables that affect a product's critical quality attributes. Here, we present a laboratory-scale model culture where we apply a Plackett-Burman screening design to parallel cultures to study the main effects of 11 process variables. This exercise allowed us to determine the relative importance of these variables and identify the most important factors to be further optimized in order to control both desirable and undesirable glycan profiles. We found engineering changes relating to culture temperature and nonessential amino acid supplementation significantly impacted glycan profiles associated with fucosylation, β-galactosylation, and sialylation. All of these are important for monoclonal antibody product quality.

A framework for the systematic design of fed-batch strategies in mammalian cell culture

A methodology to calculate the required amount of amino acids (a.a.) and glucose in feeds for animal cell culture from monitoring their levels in batch experiments is presented herein. Experiments with the designed feeds on an antibody-producing Chinese hamster ovary cell line resulted in a 3-fold increase in titer compared to batch culture. Adding 40% more nutrients to the same feed further increases the yield to 3.5 higher than in batch culture. Our results show that above a certain threshold there is no linear correlation between nutrient addition and the integral of viable cell concentration. In addition, although high ammonia levels hinder cell growth, they do not appear to affect specific antibody productivity, while we hypothesize that high extracellular lactate concentration is the cause for the metabolic shift towards lactate consumption for the cell line used. Overall, the performance of the designed feeds is comparable to that of a commercial feed that was tested in parallel. Expanding this approach to more nutrients, as well as changing the ratio of certain amino acids as informed by flux balance analysis, could achieve even higher yields.

Profiling convoluted single-dimension proton NMR spectra: a Plackett-Burman approach for assessing quantification error of metabolites in complex mixtures with application to cell culture

Single-dimension hydrogen, or proton, nuclear magnetic resonance spectroscopy (1D-(1)H NMR) has become an attractive option for characterizing the full range of components in complex mixtures of small molecular weight compounds due to its relative simplicity, speed, spectral reproducibility, and noninvasive sample preparation protocols compared to alternative methods. One challenge associated with this method is the overlap of NMR resonances leading to "convoluted" spectra. While this can be mitigated through "targeted profiling", there is still the possibility of increased quantification error. This work presents the application of a Plackett-Burman experimental design for the robust estimation of precision and accuracy of 1D-(1)H NMR compound quantification in synthetic mixtures, with application to mammalian cell culture supernatant. A single, 20 sample experiment was able to provide a sufficient estimate of bias and variability at different metabolite concentrations. Two major sources of bias were identified: incorrect interpretation of singlet resonances and the quantification of resonances from protons in close proximity to labile protons. Furthermore, decreases in measurement accuracy and precision could be observed with decreasing concentration for a small fraction of the components as a result of their particular convolution patterns. Finally, the importance of a priori concentration estimates is demonstrated through the example of interpreting acetate metabolite trends from a bioreactor cultivation of Chinese hamster ovary cells expressing a recombinant antibody.

Comprehensive, quantitative bioprocess productivity monitoring using fluorescence EEM spectroscopy and chemometrics

Using fluorescence excitation-emission matrix spectroscopy and chemometric methods we demonstrate an effective and rapid method for quantitative monitoring of a mammalian cell culture based manufacturing process.