Flow-cytometry and cell sorting: an efficient approach to investigate productivity and cell physiology in mammalian cell factories

The performance of cell lines used for the production of biotherapeutic proteins typically depends on the number of cells in culture, their specific growth rate, their viability and the cell specific productivity (qP). Therefore both cell line development and process development are trying to (a) improve cell proliferation to reduce lag-phase and achieve high number of cells; (b) delay cell death to prolong the production phase and improve culture longevity; (c) and finally, increase qP. All of these factors, when combined in an optimised process, concur to increase the final titre and yield of the recombinant protein. As cellular performance is at the centre of any improvement, analysis methods that enable the characterisation of individual cells in their entirety can help in identifying cell types and culture conditions that perform exceptionally well. This observation of cells and their complexity is reflected by the term "cytomics" and flow cytometry is one of the methods used for this purpose. With its ability to analyse the distribution of physiological properties within a population and to isolate rare outliers with exceptional properties, flow cytometry ideally complements other methods used for optimisation, including media design and cell engineering. In the present review we describe approaches that could be used, directly or indirectly, to analyse and sort cellular phenotypes characterised by improved growth behaviour, reduced cell death or high qP and outline their potential use for cell line and process optimisation.     

Determination of Chinese hamster ovary cell line stability and recombinant antibody expression during long-term culture

Chinese hamster ovary (CHO) cell lines are frequently used as hosts for the production of recombinant therapeutics, such as monoclonal antibodies, due to their ability to perform correct post-translational modifications. A potential issue when utilizing CHO cells for therapeutic protein production is the selection of cell lines that do not retain stable protein expression during long-term culture (LTC). Instability of expression impairs process yields, effective usage of time and money, and regulatory approval for the desired therapeutic. In this study, we investigated a model unstable GS-CHO cell line over a continuous period of approximately 100 generations to determine markers of mechanisms that underlie instability. In this cell line, stability of expression was retained for 40-50 generations after which time a 40% loss in antibody production was detected. The instability observed within the cell line was not due to a loss in recombinant gene copy number or decreased expression of mRNA encoding for recombinant antibody H or L chain, but was associated with lower cumulative cell time values and an apparent increased sensitivity to cellular stress (exemplified by increased mRNA expression of the stress-inducible gene GADD153). Changes were also noted in cellular metabolism during LTC (alterations to extracellular alanine accumulation, and enhanced rates of glucose and lactate utilization, during the exponential and decline phase of batch culture, respectively). Our data indicates the breadth of changes that may occur to recombinant CHO cells during LTC ranging from instability of recombinant target production at a post-mRNA level to metabolic events. Definition of the mechanisms, regulatory events, and linkages underpinning cellular phenotype changes require further detailed analysis at a molecular level.     

Improvement of the efficiency and quality in developing a new CHO host cell line

Chinese hamster ovary (CHO) cells are a ubiquitous tool for industrial therapeutic recombinant protein production. However, consistently generating high-producing clones remains a major challenge during the cell line development process. The glutamine synthetase (GS) and dihydrofolate reductase (DHFR) selection systems are commonly used CHO expression platforms based on controlling the balance of expression between the transgenic and endogenous GS or DHFR genes. Since the expression of the endogenous selection gene in CHO hosts can interfere with selection, generating a corresponding null CHO cell line is required to improve selection stringency, productivity, and stability. However, the efficiency of generating bi-allelic genetic knockouts using conventional protocols is very low (<5%). This significantly affects clone screening efficiency and reduces the chance of identifying robust knockout host cell lines. In this study, we use the GS expression system as an example to improve the genome editing process with zinc finger nucleases (ZFNs), resulting in improved GS-knockout efficiency of up to 46.8%. Furthermore, we demonstrate a process capable of enriching knockout CHO hosts with robust bioprocess traits. This integrated host development process yields a larger number of GS-knockout hosts with desired growth and recombinant protein expression characteristics.     

Comparison of the production of a human monoclonal antibody against HIV-1 by heterohybridoma cells and recombinant CHO cells: A flow cytometric study

The production of human monoclonal antibodies for therapeutic use is of increasing importance for treatment of viral infections such as AIDS. As human x mouse heterohybridomas rarely reach the growth rates and cell specific production rates of mouse hybridomas the transfection of standard cell lines, such as CHO or BHK, is a promising alternative. This has the additional advantage that the IgG subtype can be changed to suit the desired application. However, the use of a cell line that has not originally developed to produce antibodies, as lymphocytes and myeloma cells have, might have unrecognised drawbacks. This will be especially significant in the case of antibodies as each molecule consists of 4 chains linked by disulphide bonds which require specific intracellular factors to be properly folded and processed (Heavy chain binding protein, Protein Disulfide Isomerase a.o.). In this study we have therefore compared two cell lines: a human x mouse heterohybridoma producing IAM-2F5, a human IgG₃ antibody specific for HIV-1 with neutralising properties and a Chinese Hamster Ovary cell transfected with dihydrofolate reductase and with the heavy and light chain genes of IAM-2F5 modified to IgG₁. From each cell line three subclones were selected with low, medium and high specific production rates. Batch cultures were performed and the following cellular parameters analysed by flow cytometry; 1) total RNA content (translational activity); 2) total protein content; 3) cell cycle phase distribution; 4) concentration of light and heavy chains; 5) concentration of helper proteins such as BiP and PDI. The production rate of heterohybridoma cells was best reflected in the intracellular concentration of kappa chain, while the gamma chain concentration was comparable for all three subclones. In the CHO cells the gamma chain expression and thus gene copy number appeared to be the limiting factor. The GRP78/BiP concentration in CHO remained unchanged in spite of a 5-fold higher concentration of gamma chain in the high producing subclone. The PDI concentration in CHO cells was much lower compared to the heterohybridoma cells, irrespective of production rates.Abbreviations PDI protein disulfide isomerase - GRP78/BiP Glucose regulated protein; Heavy chain binding protein     

Efficient selection of high-producing subclones during gene amplification of recombinant Chinese hamster ovary cells by flow cytometry and cell sorting

The screening procedure for high-producing cell lines is extremely time- and labor-intensive and costly, and is at present guided by an empirical approach based on individual experience. Flow cytometry and cell sorting, with its ability to analyze and separate single cells, an ideal method in the selection of such rare cells. The isolation of recombinant cell lines is especially difficult due to repeated gene amplification, which introduces high mutational variation into the population. We have established and evaluated a modification of a previous method that traps secreted product on the surface of the secreting cell, thus allowing direct analysis of single cell specific production rates. This method was used to select for high-producing subclones of a recombinant Chinese hamster ovary (CHO) cell line producing a human antibody against HIV-1 by repeated rounds of gene amplification and cell sorting. This cell line has been amplified in previous investigations, so that the amount of work and testing required by traditional methods can be compared with the protocol described herein. Forty-five 96-well plates were necessary to obtain a high-producing subclone by limited dilution methods, whereas only five plates were required when cell sorting was used. The specific production rate of the best clone obtained by sorting, however, was five times that of the clone obtained by traditional methods. In contrast to the clones obtained by limited dilution, which consisted of several populations of low- and high-producing cells even at high methotrexate concentrations (6.4 microM), the clones isolated by sorting were already homogeneous at 0.8 microM methotrexate.     

Towards dynamic metabolic flux analysis in CHO cell cultures

Chinese hamster ovary (CHO) cells are the most widely used mammalian cell line for biopharmaceutical production, with a total global market approaching $100 billion per year. In the pharmaceutical industry CHO cells are grown in fed-batch culture, where cellular metabolism is characterized by high glucose and glutamine uptake rates combined with high rates of ammonium and lactate secretion. The metabolism of CHO cells changes dramatically during a fed-batch culture as the cells adapt to a changing environment and transition from exponential growth phase to stationary phase. Thus far, it has been challenging to study metabolic flux dynamics in CHO cell cultures using conventional metabolic flux analysis techniques that were developed for systems at metabolic steady state. In this paper we review progress on flux analysis in CHO cells and techniques for dynamic metabolic flux analysis. Application of these new tools may allow identification of intracellular metabolic bottlenecks at specific stages in CHO cell cultures and eventually lead to novel strategies for improving CHO cell metabolism and optimizing biopharmaceutical process performance.     

Characterization and improvement of cell line performance via flow cytometry and cell sorting

The improvement of specific productivity is a continuous challenge for bioprocesses involving mammalian cells, and hence, high‐throughput methods and low‐cost strategies are needed for the selection of high producers. The aim of this study was the productivity improvement of the hybridoma cell line IV F 19.23. For this purpose, a cell surface affinity matrix assay was established to identify and select high producers. This assay is based on the binding of secreted monoclonal antibodies to an affinity matrix assembled on the outer cell membrane. A protein microarray approach was used to investigate and optimize the functionality of the affinity matrix. The protein microarray was particularly useful to identify critical steps of the staining method, such as unspecific binding, before it was applied to the hybridoma cell line. Secreting hybridomas were treated with the affinity matrix and then selected via flow‐cytometric cell sorting in four consecutive bulk sort rounds. The applied bulk strategy, allowing low screening costs, resulted in a 125% increase in specific productivity of the cell line in comparison to the initial population.     

Selection of CHO host cell subclones with increased specific antibody production rates by repeated cycles of transient transfection and cell sorting

Optimization of host cell lines both for transient and stable protein production is typically hampered by the inherent heterogeneity of cells within a population. This heterogeneity is caused not only by “hard fact” gene mutations, but also by subtle differences in the cellular network of regulation, which may include epigenetic variations. Taking advantage of this heterogeneity, we sorted for naturally occurring variants of CHO‐K1 and CHO‐S host cells that possess an improved cellular machinery for transient antibody production. The long‐term goal of this study was both to identify host cells that yield recombinant cell lines with on average higher productivity, but also to study the molecular differences that characterize such cells, independent of the site of gene integration or gene amplification. To identify such cells we optimized the procedure for transient transfection by electroporation to a degree that gave uniform transfer of plasmid DNA into nearly 100% of the cells and resulted in reproducible average productivities, with a standard deviation of 16% between independent experiments. Using this optimized protocol, the 1% of cells with the highest specific productivity was sorted and subcloned with a cold capture secretion assay. Upon re‐transfection, the resulting subclones showed the same specific productivity as their respective parental cell line. To enrich for cells with potentially stable improved properties, the 1% highest producers were sorted three times, 2 days after transient transfection each, and the enriched population was again sorted into microtiter plates for subcloning. For each of the two parental cell lines tested, three subclones were obtained that had a threefold higher specific productivity after transient transfection. This property was stable for approximately 3 months, indicating that the changes in productivity were regulatory and not mutational. Biotechnol. Bioeng. 2011;108: 386–394. © 2010 Wiley Periodicals, Inc.     

An improved phage display procedure for identification of lipopolysaccharide-binding peptides

We successfully implemented several modifications to the regular phage display procedure and significantly improved the lipopolysaccharides-binding properties of the peptides selected. Specifically, the number of biopannings was increased and peptides with consensus sequences were obtained. A dual selection procedure (referred to as subtractive panning) was used to simultaneously select for the desired target and deselect for an undesired target, thereby increasing the binding specificity. In addition, binding and washing conditions in the subtractive panning were also modified to favor the selection of peptides with higher binding strength. As a result, two peptides, ASFPPAF and SSHTISF, were identified with much improved binding properties compared to those selected with regular panning. The binding specificities of these two peptides, as measured by the ratio of phages bound to the desired and undesired targets, were severalfold higher than previously reported. These modifications could easily be implemented with many other target molecules, indicating the general applicability of the procedure.     

A 24-microwell plate with improved mixing and scalable performance for high throughput cell cultures

Multi-well plates are widely used in high throughput drug screening, cell clone development, media design and cell culture optimization in the biotechnology industry. The reproducibility and data quality of cell cultures in multi-well plates are greatly affected by mixing, aeration, and evaporation. A novel 24-microwell plate (MWP) with static mixers for improved mixing and aeration, and gas permeable lids for reduced evaporation was developed for cell cultures. Mixing, oxygen transfer, evaporation, and cell proliferation as affected by the static mixer, shape of the well and agitation rate were studied. The static mixer improved mixing pattern and reduced cell aggregation under orbital shaking conditions. Consequently, the static mixer also improved cell proliferation with a significantly higher specific growth rate in round wells. In general, consistent growth kinetics was observed for cells cultured on the plate. Overall, the MWP improved the data quality with smaller standard deviations and better reproducibility. Furthermore, CHO cells cultured in the MWP gave similar kinetics in glucose consumption, lactate production, cell growth and viability, and antibody production in a serum-free medium to those cultured in spinner flasks, demonstrating its scalable performance and potential application in high throughput screening for cell culture process development.     

Enhancing CHO cell productivity through a dual selection system using Aspg and Gs in glutamine free medium

The dominant method for generating Chinese hamster ovary (CHO) cell lines that produce high titers of biotherapeutic proteins utilizes selectable markers such as dihydrofolate reductase (Dhfr) or glutamine synthetase (Gs), alongside inhibitory compounds like methotrexate or methionine sulfoximine, respectively. Recent work has shown the importance of asparaginase (Aspg) for growth in media lacking glutamine—the selection medium for Gs-based selection systems. We generated a Gs/Aspg double knockout CHO cell line and evaluated its utility as a novel dual selectable system via co-transfection of Gs-Enbrel and Aspg-Enbrel plasmids. Using the same selection conditions as the standard Gs system, the resulting cells from the Gs/Aspg dual selection showed substantially improved specific productivity and titer compared to the standard Gs selection method, however, with reduced growth rate and viability. Following adaptation in the selection medium, the cells improved viability and growth while still achieving ~5-fold higher specific productivity and ~3-fold higher titer than Gs selection alone. We anticipate that with further optimization of culture medium and selection conditions, this approach would serve as an effective addition to workflows for the industrial production of recombinant biotherapeutics.     

CRISPR-interceded CHO cell line development approaches

For industrial production of recombinant protein biopharmaceuticals, Chinese hamster ovary (CHO) cells represent the most widely adopted host cell system, owing to their capacity to produce high-quality biologics with human-like posttranslational modifications. As opposed to random integration, targeted genome editing in genomic safe harbor sites has offered CHO cell line engineering a new perspective, ensuring production consistency in long-term culture and high biotherapeutic expression levels. Corresponding the remarkable advancements in knowledge of CRISPR-Cas systems, the use of CRISPR-Cas technology along with the donor design strategies has been pushed into increasing novel scenarios in cell line engineering, allowing scientists to modify mammalian genomes such as CHO cell line quickly, readily, and efficiently. Depending on the strategies and production requirements, the gene of interest can also be incorporated at single or multiple loci. This review will give a gist of all the most fundamental recent advancements in CHO cell line development, such as different cell line engineering approaches along with donor design strategies for targeted integration of the desired construct into genomic hot spots, which could ultimately lead to the fast-track product development process with consistent, improved product yield and quality.     

Hybridoma growth and antibody production as a function of cell density and specific growth rate in perfusion culture

Steady state metabolic parameters for hybridoma cell line H22 were determined over a wide range of cell densities and specific growth rates in a filtration based homogeneous perfusion reactor. Operating the reactor at perfusion rates of 0.75, 2.0, and 2.9 day(-1)(each at four different specific growth rates), viable cell densities as high as 2 x 10(7) cells/mL were obtained. For the cell line under investigation, the specific monoclonal antibody production rate was found to be a strong function of the viable cell density, increasing with increasing cell density. In contrast, most of the substrate consumption and product formation rates were strong functions of the specific growth rate. Substrate metabolism became more efficient at high cell densities and low specific growth rates. The Specific rates of metabolite formation and the apparent yields of lactate from glucose and ammonia from glutamine decreased at low specific growth rates and high cell densities. While the specific oxygen consumption rate was independent of the specific growth rate and cell density, ATP production was more oxidative at lower specific growth rate and higher cell density. These observed shifts are strong indications of the production potential of high-density perfusion culture. (c) 1995 John Wiley & Sons, Inc.     

Effect of specific growth rates on productivity in continuous open and partial cell retention animal cell bioreactors.

A clonal derivative of a transfectant of the SP2/0 myeloma cell line producing a chimeric monoclonal antibody was cultivated in both continuous open and continuous partially-closed bioreactors. Using an open system for the determination of kinetic parameters, we showed that the production of this chimeric mAb was growth associated. As such, the volumetric productivity increased linearly with increasing dilution rate up to the maximum dilution rate. Three continuous cultivations employing partial cell retention were conducted. In agreement with mathematical predictions, the product titer and volumetric productivity were independent of the degree of cell retention when the total dilution was held constant. When cells were maintained at a low specific growth rate, the product titer was independent of dilution rate and the volumetric productivity increased with increasing dilution rate, again in agreement with mathematical predictions. Since the partially-closed bioreactor could be operated at dilution rates in excess of the maximum specific cellular growth rate, volumetric productivities were greater than those achievable in the open bioreactor. However, when cells were maintained at a high specific growth rate, cell accumulation was limited and product titers decreased at high dilution rates. Therefore, the volumetric productivity in this latter case did not increase at higher dilution rates.     

Reverse engineering of protein secretion by uncoupling of cell cycle phases from growth

The demand for recombinant proteins both for biopharmaceutical and technical applications is rapidly growing, and therefore the need to establish highly productive expression systems is steadily increasing. Yeasts, such as Pichia pastoris, are among the widely used production platforms with a strong emphasis on secreted proteins. Protein secretion is a limiting factor of productivity. There is strong evidence that secretion is coupled to specific growth rate (µ) in yeast, being higher at higher µ. For maximum productivity and product titer, high specific secretion rates at low µ would be desired. At high secretion rates cultures contain a large fraction of cells in the G2 and M phases of cell cycle. Consequently, the cell design target of a high fraction of cells in G2 + M phase was achieved by constitutive overexpression of the cyclin gene CLB2. Together with predictive process modeling this reverse engineered production strain improved the space time yield (STY) of an antibody Fab fragment by 18% and the product titer by 53%. This concept was verified with another secreted protein, human trypsinogen. Biotechnol. Bioeng. 2011;108: 2403–2412. © 2011 Wiley Periodicals, Inc.     

An improved flow microfluorometer for rapid measurement of cell fluorescence

An improved instrument for the making of high-speed fluorescent measurements on single cells has been constructed and characterized. Instrumental optimization was accomplished with Chinese hamser cells (line CHO) stained by the fluorescent Feulgen procedure using either auramine-0 or acriflavine. Contributions of instrumental resolution and artifacts to cellular DNA distributions have been determined. The instrument introduces a total coefficient of variation of less than 2% for the case of CHO cells stained with bright dyes such as acriflavine, where photon-statistical effects are minimal. The design of the instrument is such that any of a number of cellular constituents or properties can be studied and quantitated on a single-cell basis at a rate of 50 000 cells/min.     

IL‐17F co‐ ;expression improves cell growth characteristics and enhances recombinant protein production during CHO cell line engineering

The generation of a high productivity cell line is a critical step in the production of a therapeutic protein. Many innovative engineering strategies have been devised in order to maximize the expression rate of production cells for increased process efficiency. Less effort has focused on improvements to the cell line generation process, which is typically long and laborious when using mammalian cells. Based on unexpected findings when generating stable CHO cell lines expressing human IL‐17F, we studied the benefit of expressing this protein during the establishment of production cell lines. We demonstrate that IL‐17F expression enhances the rate of selection and overall number of selected cell lines as well as their transgene expression levels. We also show that this benefit is observed with different parental CHO cell lines and selection systems. Furthermore, IL‐17F expression improves the efficiency of cell line subcloning processes. IL‐17F can therefore be exploited in a standard manufacturing process to obtain higher productivity clones in a reduced time frame. Biotechnol. Bioeng. 2013; 110: 1153–1163. © 2012 Wiley Periodicals, Inc.     

High density culture of HeLa cells in a CelliGen perfusion system

A hollow fiber cartridge may be used in an extraneous recycle loop to facilitate perfusion operation of a stirred tank bioreactor. Retention of cells while removing waste products and replenishment with fresh nutrients allows higher than normal cell densities obtained in batch or continuous culture systems. This system successfully propagated HeLa cells to over 11 million viable cells per milliliter. Much higher perfusion rates (up to 4 vessel volumes per day) were necessary for high density culture of HeLa cells compared to BHK or a hybridoma cell line because of a much higher specific cellular metabolic rate. Cell specific glucose consumption rate, lactate production and ammonia production rates are several times higher for HeLa cells. Reproducible high cell densities and viabilities can be repeatedly obtained after harvest and dilution of a HeLa cell culture by partial drainage and reconstitution in the bioreactor.     

Characterisation of recombinant CHO cell lines by investigation of protein productivities and genetic parameters

We have generated a recombinant CHO cell line expressing the fusion protein EpoFc. After selection and screening, protein expression, gene and mRNA copy numbers were analysed in order to gain more information on the influence of genetic parameters on the productivity and stability of production cells. Results from semi-quantitative blot methods were compared to quantitative PCR (qPCR) analyses, whose advantage mainly lies in their higher sensitivity, and the cheaper and faster methodology. We developed stable and high producing clones with low gene copy numbers, in contrast to other cell lines where multiple steps of methotrexate amplification have lead to hundreds of copies of inserts with the risk of karyotypic instabilities and decreased growth rates that overcome the benefits of increased productivities. When comparing genetic parameters to productivity, a good correlation of mRNA levels with specific productivity was observed, whereas high gene copy numbers were not always accompanied by high protein expressions. Based on our data derived from a typical example of a cell line development process, genetic parameters are useful tools for the selection of scalable production clones. Nevertheless, a wider range of cell lines has to be investigated in order to implement genetic analyses into a screening process.