The Super-Spinner: a low cost animal cell culture bioreactor for the CO2 incubator

A Call for a European Prohibition of Monoclonal Antibody Production by the Ascites Procedure in Laboratory Animals

Monoclonal antibodies (mAbs) are particularly valuable in therapeutics and research. Unfortunately, one of the most familiar methods of producing mAbs, the ascites induction method, causes pain and distress to the animals used. In most cases, non-animal or in vitro alternatives can be employed to reduce or eliminate the use of animals for mAb production. Prohibition of the use of animals in the production of mAbs is recommended, except when the replacement in vitro methods prove to be insufficient, and in a limited number of other well-documented cases, such as an exceptional need for an emergency therapeutic application. A total ban on the use of animals for mAb production is impractical and it is imperative that an appeals process should accompany the prohibition. The need for the establishment of core facilities for in vitro mAb production is emphasised.

Numerical simulation of scaling-up an inverted frusto-conical shaking bioreactor with low shear stress for mammalian cell suspension culture

Shear stress is one of the key factors affecting the large-scale culture of mammalian cells. In this study, numerical simulation based on computational fluid dynamics was used to conduct a flow-field analysis of 7, 50, 200, and 1200 L inverted frusto-conical shaking bioreactors. The results show that the shear rate, specific mass transfer area (a), and volumetric oxygen mass transfer coefficient (kLa) gradually decreased as the scale of the bioreactor increased. Through application of BHK21 and CHO cells in 7, 200, and 1200 L bioreactors, it was found that the cell density and antibody expression level increased as the volume of the bioreactor increased. Moreover, the antibody expression level in a 1200 L bioreactor was nearly 30% and 35% higher than that of 7 and 200 L bioreactors, respectively. The results demonstrate that the environment with a larger volume is more suitable for the growth and antibody expression of CHO cells, indicating shear stress might be the most critical factor affecting the scale-up of mammalian cells.

A membrane stirrer for product recovery and substrate feeding

During fermentation processes, in situ product recovery (ISPR) using submerged membranes allows a continuous operation mode with effective product removal. Continuous recovery reduces product inhibition and organisms in the reactor are not exposed to changing reaction conditions. For an effective in situ product removal, submerged membrane systems should have a sufficient large membrane area and an anti-fouling concept integrated in a compact device for the limited space in a lab-scale bioreactor. We present a new membrane stirrer with integrated filtration membranes on the impeller blades as well as an integrated gassing concept in an all-in-one device. The stirrer is fabricated by rapid prototyping and is equipped with a commercial micromesh membrane. Filtration performance is tested using a yeast cell suspension with different stirring speeds and aeration fluxes. We reduce membrane fouling by backflushing through the membrane with the product stream.

Effects of dissolved oxygen levels and the role of extra- and intracellular amino acid concentrations upon the metabolism of mammalian cell lines during batch and continuous cultures

The effects of dissolved oxygen and the concentration of essential amino acids upon the metabolism of two mammalian cell lines (rCHO producing human active (t-PA) and a mouse-mouse hybridoma) were investigated in batch, chemostat, and perfusion cultures. Intracellular amino acid concentrations were measured for both cell lines during repeated batch cultures and the K(S)-values for the essential amino acids were calculated using Monod equations via computer simulation. The K(S)-values were in the range of 10 mmol L(-1) and the pool of most intracellular amino acids remained constant at about 10-100 fold higher in concentration than in the medium. No significant differences were observed between the hybridoma and CHO cell. The specific nutrient uptake rates corresponded with the cell specific growth rate and the effects of reduced dissolved oxygen concentrations only became evident when the DO dropped below 5% of air saturation (critical concentration below 1%). Nevertheless, a correlation between nutrient concentration and specific oxygen uptake was detected.

Production of recombinant human antithrombin III on 20-L bioreactor scale: correlation of supernatant neuraminidase activity, desialylation, and decrease of biological activity of recombinant glycoprotein

Chinese hamster ovary (CHO) cells producing the recombinant glycoprotein human antithrombin III (rhAT III) were batch cultivated in a 20-L bioreactor for 13 days. Neuraminidase activity in cell-free supernatant was monitored during cultivation and free sialic acid was determined by HPLC. Neu5Acalpha(2-->3)Gal-specific Maackia amurensis and Galbeta(1-->4)GlcNAc-specific Datura stramonium agglutinin were used for determination of sialylated and desialylated rhAT III, respectively. A commercial test kit was used for evaluation of functional rhAT III activity. Supernatant neuraminidase as well as lactate dehydrogenase activity increased significantly during batch growth. The enhanced number of dead cells correlated with increased neuraminidase activity, which seemed to be principally due to cell lysis, resulting in release of cytosolic neuraminidase. Loss of terminally alpha(2-->3) linked sialic acids of the oligosaccharide portions of rhAT III, analyzed in lectin-based Western blot and lectin-adsorbent assays, correlated with a decrease of activity of rhAT III produced throughout long-term batch cultivation. Thus, structural oligosaccharide integrity as well as the functional activity of recombinant glycoprotein depend on the viability and mortality of the bioreactor culture, and batches with a high number of viable cells are required to guarantee production of glycoproteins with maximum biological activity. (c) 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 56: 441-448, 1997.

Cell size distribution as a parameter for the predetermination of exponential growth during repeated batch cultivation of CHO cells

The routine measurement of the cell size distribution of a Chinese hamster ovary (CHO) cell population during a repeated batch process enables the predetermination of exponential growth even 24 h before the population enters the log phase, due to a short but significantly increased cell size during the lag phase. A prolongation of the stationary phase causes to progressive limitation in asparagine, serine, and ethanolamine. Such extended limitation influences the duration of the following lag phase and obviously induces a synchronization of the cell population that can be monitored easily by a fast cell size analyzing technique. (c) 1997 John Wiley & Sons, Inc. Biotechnol Bioeng 55: 793-797, 1997.

Bag bioreactor based on wave-induced motion: characteristics and applications

Today wave-mixed bag bioreactors are common devices in modern biotechnological processes where simple, safe and flexible production has top priority. Numerous studies that have been published on ex vivo generation of cells, viruses and therapeutic agents during the last 10 years have confirmed their suitability and even superiority to stirred bioreactors made from glass or stainless steel for animal as well as plant cell cultivations. In these studies the wave-mixed bag bioreactors enabled middle to high cell density and adequate productivity in laboratory and pilot scale. This mainly results from low-shear conditions and highly efficient oxygen transfer for cell cultures, as demonstrated for the widely used BioWave((R)).Starting with an overview of wave-mixed bag bioreactors and their common operation strategies, this chapter delineates engineering aspects of BioWave((R)), which like Wave Reactor and BIOSTAT((R))CultiBag RM originates from the prototype of a wave-mixed bag bioreactor introduced in 1998. Subsequently, the second part of the chapter focuses on reported BioWave((R)) applications. Conditions and results from cultivations with animal cells, plant cells, microbial cells and nematodes are presented and discussed.

Serum-free production of a chimeric E-selectin-IgG protein from 1 to 100 l scale: Repeated batch cultivation versus continuous spin filter perfusion

On inflamed endothelium the cell surface protein E-selectin isexpressed which supports the initial process of attachment -capturing and rolling of leukocytes. A recombinant CHO cell linesecreting a soluble E-selectin-IgG chimera was cultivated competitively under serum free conditions in three different bioreactor systems: a 1 l Super-Spinner, a 2 l stirred tank bioreactor equipped with a spinfilter, and a 100 l stirred tankbioreactor. In the smallest system 25.4 mg E-selectin-IgG wereproduced in 62 days using a repeated batch process whileachieving a maximal viable cell density of 3.7 x 10(6) cells ml(-1). Using continuous perfusion mode a total amount of35.2 mg were produced with a maximal viable cell density of1.65 x 10(7) cells ml(-1) in the 2 l bioreactor within 29 days. Large scale cultivation in a 100 l stirred tankbioreactor yielded 105.6 mg in three batches with a maximal viable cell density of 9.7 x 10(5) cells ml(-1) within 15 days. After removal of the cells by continuous centrifugation and a depth filter clearance step, the supernatants were concentrated via ultra filtration. Purificationwas performed by affinity chromatography with rProtein A. Integrity of the E-selectin-IgG protein was checked with SDS PAGE. Its activity was verified in a cellular adhesion assay performed with HL-60 cells and a recombinant CHO cell line expressing membrane-anchored E-selectin constitutively, and E-selectin expressing HUVECs, respectively. Soluble E-selectin-IgG was used to block adhesion to these cell layerscompetitively. A concentation of 18.8 and 37.5 mug ml(-1)was sufficient to reduce the amount of adhering HL-60 cells to 50% on CHO and HUVEC layers, respectively.

Cost-effective engineering of a small-scale bioreactor

Several methods exist for increasing the scale of cell culture in the laboratory. While these methods provide significant increases in biomass, they are often prohibitively expensive for many laboratories. We have engineered a small-scale bioreactor with a novel means of introducing oxygen through the catalytic breakdown of hydrogen peroxide using a manganese oxide catalyst. We have also adapted and modified an existing assay for dissolved oxygen to be compatible with culture conditions. In this system we have been able to culture CHO cells at densities of up to 10(7) cells/mL without the use of automated feedback systems.

Preferential binding of the anticancer drug rViscumin (recombinant mistletoe lectin) to terminally alpha2-6-sialylated neolacto-series gangliosides

Production of biochemically defined recombinant mistletoe lectin was achieved by cloning and separate expression of the single catalytically active A-chain and the B-chain with carbohydrate binding properties in Escherichia coli, yielding an active heterodimeric protein named rViscumin (Eck et al. [1999] Eur. J. Biochem., 265, 788-797). Employing solid phase binding assays, rViscumin was shown to preferentially bind to terminally alpha2-6-sialylated neolacto-series gangliosides IV(6)Neu5Ac-nLc4Cer, VI(6)Neu5Ac-nLc6Cer, and VIII(6)Neu5Ac-nLc8Cer isolated from human granulocytes. Only marginal binding of rViscumin to galactose-terminated neutral GSLs was determined, whereas reinvestigation of ricin specificity demonstrated this lectin as a galactose-binding protein. Human promyelotic HL-60 cells exhibited an IC(50) value (half maximum cytotoxicity) of 1.16 pM and human bladder carcinoma 5637 cells of 12.1 pM rViscumin; CHO-K1 cells were resistant to rViscumin treatment up to a concentration of 5.26 nM tested. Quantification of the predominant receptor ganglioside IV(6)Neu5Ac-nLc4Cer by means of a specific anti-Neu5Acalpha2-6Galbeta1-4GlcNAc-R antibody revealed 3.68 x 10(6) and 1.54 x 10(6) receptor molecules per HL-60 and 5637 cell, respectively; CHO-K1 cells were negative, lacking alpha2-6-sialylated gangliosides. The data imply a direct correlation of rViscumin cytotoxicity and the expression of receptor ganglioside. Moreover, CHO-K1 cells were rendered susceptible toward rViscumin cytotoxicity after exogenous application of human granulocyte gangliosides. Thus, (1) rViscumin has to be considered as a sialic acid-specific rather than a galactose-specific type II ribosome-inactivating protein, and (2) neolacto-series gangliosides with Neu5Acalpha2-6Galbeta1-4GlcNAc-terminus are true functional and physiologically relevant rViscumin receptors.

Tyrosine sulfation and N-glycosylation of human heparin cofactor II from plasma and recombinant Chinese hamster ovary cells and their effects on heparin binding

The structure of post-translational modifications of human heparin cofactor II isolated from human serum and from recombinant Chinese hamster ovary cells and their effects on heparin binding have been characterized. Oligosaccharide chains were found attached to all three potential N-glycosylation sites in both protein preparations. The carbohydrate structures of heparin cofactor II circulating in blood are complex-type diantennary and triantennary chains in a ratio of 6 : 1 with the galactose being > 90% sialylated with alpha 2-->6 linked N-acetylneuraminic acid. About 50% of the triantennary structures contain one sLe(x) motif. Proximal alpha 1-->6 fucosylation of oligosacharides from Chinese hamster ovary cell-derived HCII was detected in > 90% of the diantennary and triantennary glycans, the latter being slightly less sialylated with exclusively alpha 2-->3-linked N-acetylneuraminic acid units. Applying the ESI-MS/ MS-MS technique, we demonstrate that the tryptic peptides comprising tyrosine residues in positions 60 and 73 were almost completely sulfated irrespective of the protein's origin. Treatment of transfected Chinese hamster ovary cells with chlorate or tunicamycin resulted in the production of heparin cofactor II molecules that eluted with higher ionic strength from heparin-Sepharose, indicating that tyrosine sulfation and N-linked glycans may affect the inhibitor's interaction with glycosaminoglycans.

An automatic system for the assessment of complex medium additives under cultivation conditions

Complex medium additives such as yeast extract or peptone are often used in industrial cell culture processes to prolong cell growth and/or to improve product formation. The quality of those supplements is dependent on the preparation method and can differ from lot to lot. To guarantee consistent production these different lots have to be tested prior to use in fermentation processes. Because a detailed qualitative and quantitative analysis of all components of such a complex mixture is a very difficult task, another assessment method has to be chosen. The best way to evaluate the effect of such supplements is to monitor cell activity during real cultivation conditions with and without the added supplement lot. A bioreactor-based test system has been developed to determine the oxygen requirement of the cells as a response to the addition of a supplement to be tested under standardized conditions. Investigations were performed with a mouse-mouse hybridoma cell line and yeast extracts as an example for complex medium additives. The results showed differences in the impact between different extract lots and between different concentrations of an extract.

Perfusion culture of baculovirus-infected BTI-Tn-5B1-4 insect cells: a method to restore cell-specific beta-trace glycoprotein productivity at high cell density

The impact of different cultivation-infection strategies on the productivity of baculovirus-infected BTI-Tn-5B1-4 (High Five) cells was investigated. Using beta-trace protein as the recombinant glycoprotein, the effects of multiplicity of infection (MOI) and time of infection (TOI) were studied on growth after infection as well as the degree of infection and recombinant protein productivity in batch culture. The highest productivities were found when infecting Tn5 cells at early exponential growth phase (EGP) (low cell density) using a high MOI. To increase the productive cell density of Tn5 cells after beta-trace-baculovirus infection, we performed studies infecting cells in the range of 1 to 5 x 10(6) cells/mL in fresh medium. Although the protein production was increased twofold, a strong negative cell density effect was still observed when maximal productive cell density exceeded 1 x 10(6) cells/mL. To verify whether the changing cell environment of the batch experiments was responsible for the decrease in protein productivity at increasing cell density at infection, several perfusion experiments were designed by infecting Tn5 cells at cell densities over 2 x 10(6) cells/mL under more steady-state conditions. The use of this experimental setup enabled successful infections at high cell densities with volumetric productivities of up to 1.2 g L(-1) day(-1) of beta-trace protein, which is very high for a glycoprotein expressed with the baculovirus expression vector system (BEVS). The cell specific protein productivity observed after infections at higher cell densities in perfusion mode was the same as in batch experiments at low cell concentrations, which clearly demonstrates that the cell density effect could be completely overcome with perfusion cultivation.

Estimation of cell damage in bench- and pilot-scale affinity expanded-bed chromatography for the purification of monoclonal antibodies

The first step in downstream processing of mammalian cell culture includes the separation of the cells without cell damage to avoid the release of intracellular enzymes, which could potentially cause proteolytic degradation of the target protein and may increase the impurities for further chromatographic steps. This especially includes the reduction of host DNA for therapeutic proteins. The aim of this investigation was to examine the extent of cell damage at the bench and pilot scale using a stabilized fluidized bed (expanded bed) for direct recovery of IgG from cell culture broth. For this purpose, Streamline-25 and -200 columns containing 75 mL and 5 L of rProtein A matrix, respectively, were used. The repeated batch cultivations resulted in high cell viabilities of about 90% prior purification. The pH was gently adjusted to pH 8 before the broth was applied to the gel. In bench scale, 1 to 6 L of unclarified feed was applied to the Streamline-25 column. In pilot scale, up to 95 L was processed using the Streamline-200 column. The antibodies from 95 L of unclarified feed were recovered after approximately 1.5 h. The possible cell damage, caused either by the equipment or by the cells' passage through the expanded bed, was detected by the following assays: microscopic count of the cells using trypan blue dye exclusion to determine viability; monitoring of intracellular components (i.e., DNA concentration); activity of lactate dehydrogenase (LDH); and, finally, the particle load in the flow through and the eluate. Despite the sensitivity of hybridoma cells to shear forces, neither the high flow rate (300 to 450 cm/h) nor the passage of the cells through the expanded bed caused any relevant cell damage or clogging of the gel. Excellent DNA depletion was observed.

Functional expression of a fragment of human dihydroorotate dehydrogenase by means of the baculovirus expression vector system, and kinetic investigation of the purified recombinant enzyme

Human mitochondrial dihydroorotate dehydrogenase (the fourth enzyme of pyrimidine de novo synthesis) has been overproduced by means of a recombinant baculovirus that contained the human cDNA fragment for this protein. After virus infection and protein expression in Trichoplusia ni cells (BTI-Tn-5B1-4), the subcellular distribution of the recombinant dihydroorotate dehydrogenase was determined by two distinct enzyme-activity assays and by Western blot analysis with anti-(dihydroorotate dehydrogenase) Ig. The targeting of the recombinant protein to the mitochondria of the insect cells was verified. The activity of the recombinant enzyme in the mitochondria of infected cells was about 740-fold above the level of dihydroorotate dehydrogenase in human liver mitochondria. In a three-step procedure, dihydroorotate dehydrogenase was purified to a specific activity of greater than 50 U/mg. Size-exclusion chromatography showed a molecular mass of 42 kDa and confirmed the existence of the fully active enzyme as a monomeric species. Fluorimetric cofactor analysis revealed the presence of FMN in recombinant dihydroorotate dehydrogenase. By kinetics analysis, Km values for dihydroorotate and ubiquinone-50 were found to be 4 microM and 9.9 microM, respectively, while Km values for dihydroorotate and decylubiquinone were 9.4 microM and 13.7 microM, respectively. The applied expression system will allow preparation of large quantities of the enzyme for structure and function studies. Purified recombinant human dihytdroorotate dehydrogenase was tested for its sensitivity to a reported inhibitor A77 1726 (2-hydroxyethyliden-cyanoacetic acid 4-trifluoromethyl anilide), which is the active metabolite of the isoxazole derivative leflunomide [5-methyl-N-(4-trifluoromethyl-phenyl)-4-isoxazole carboximide]. An IC50 value of 1 microM was determined for A77 1726. Detailed kinetics experiments revealed uncompetitive inhibition with respect to dihydroorotate (Kiu = 0.94 microM) and non-competitive inhibition with respect to decylubiquinone (Kic = 1.09 microM, Kiu = 1.05 microM). These results suggest that the immunomodulating agent A77 1726 (currently in clinical phase III studies for the treatment of rheumatoid arthritis) is a very good inhibitor of human dihydroorotate dehydrogenase.

Design of a bubble-swarm bioreactor for animal cell culture

A stationary bubble-swarm has been used to aerate a mammalian cell culture bioreactor with an extremely low gas flow rate. Prolonging the residence time of the gas bubbles within the medium improved the efficiency of the gas transfer into the liquid phase and suppressed foam formation. An appropriate field of speed gradients prevented the bubbles from rising to the surface. This aeration method achieves an almost 90% transfer of oxygen supplied by the bubbles. Consequently, it is able to supply cells with oxygen even at high cell densities, while sparging with a gas flow of only 0.22 x 10(-3) -1.45 x 10(-3) vvm (30-200 ml/h). The reactor design, the oxygen transfer rates and the high efficiency of the system are presented. Two repeated batch cultures of a rat-mouse hybridoma cell line are compared with a surface-aerated spinner culture. The used cell culture medium was serum-free, either with or without BSA and did not contain surfactants or other cell protecting agents. One batch is discussed in detail for oxygen supply, amino acid consumption and specific antibody production.