Fed-batch process development
for monoclonal antibody production with cellferm-pro®
Authors: Anna Frison and Dr. Klaus Memmert, Novartis Pharma AG, Bal, Switzerland Published in Genetic Engineering News, Volume 22, Number 11, June 1, 2002
1 Introduction:
Cost-effective large-scale production of monoclonal antibodies (mAb) is creating a strong demand for reliable and prompt development of highly productive, scalable process. One of them is fed-batch, which is widely ud for production of recombinant proteins, due to its operational simplicity, reliability, and flexibility in multipurpo implementation. The major advantage of fed-batch, comparing to batch, is the ability to increa maximum viable cell concentration, prolong culture lifetime, and allow product accumulation to a higher concentration.
The maximisation of final product concentration in a hybridoma fed-batch process is a function of the Integral of Viable Cells Concentration (IVCC). It follows, that an increa of this variable through feeding strategy optimisation will boost the final product titer.
雅典是哪个国家Pre-defined feeding protocols, which are bad on nutrient requirement estimations, e.g. stepwi or a sigmoid-bad addition of nutrient concentrates are unlikely to meet the nutritional demands of cells growing in batch culture which vary with time and environmental conditions. It often leads to nutrient depletion or accumulation of substrates or metabolites to inhibitory levels (Zhou et al., 1996). A much better way for nutrient supply would be the online monitoring of the cell culture’s metabolic activity, and a real time control of nutrient feeding bad on this parameter.
The novel feeding protocol discusd here is bad on the oxygen uptake rate (OUR) on-line estimation. OUR is a very important indicator for the metabolic activity of the biological system. In cell cultures it correlates with the gluco or glutamine consumption rate. In our experiments OUR was ud as the control parameter for adapting the feed rate according to the cell culture’s requirements in real time. The experiments were performed using the novel cellferm-pro®system.
2 cellferm-pro® system
2.1 System’s asmbly:
The cellferm-pro® system (DASGIP AG, Juelich, Germany) consists of five modules as shown in Fig.1.
Figure 1:cellferm-pro®, a parallel and fully controlled cultivation system
(DASGIP AG, Juelich, Germany)
The culture system compris of a temperature controlled incubator with 4 or 8 vesls, placed on a
qq怎么锁空间magnetic stirrer platform, equipped with pH and pO2 electrodes, and feed and air supply/removal connections.
The gassing system provides an individual mix of up to four gas to each culture vesl (e.g. compresd air, oxygen, nitrogen and carbon dioxide). O2 and CO2 gas concentration are controlled by feedback from pO2 and pH electrodes. There is an electronic mass flow control and a gas totalizer function for each vesl.
The monitoring system simultaneously process the signals from pH and pO2 electrodes, regulates pH and pO2 in the medium through adjustment of CO2 and O2 in the gas mixture to maintain the t-points, and determines the OUR online.
The dosing system offers an individual and regulated delivery of a feed medium. The dosing proceeds according to ur defined profiles or fully automated bad on the online determined OUR (Metabolic Activity bad feeding).
The control system is bad on Microsoft Windows. Here pH, pO2, gassing and dosing with ur defined profiles are configured and the calibration of the electrodes and dosing system performed. The analys of logged data are done using DASGIP’s ChartWizard for MS Excel®.
2.2 Metabolic Activity Tool:
A sophisticated algorithm computes the OUR online for each culture vesl of the cellferm-pro® system. Online OUR, supplemented with known concentration of one important feed medium component and pre-estimated ratio between substrate supply rate and oxygen uptake rate, allows an individual and automated addition of liquid media according to the actual metabolic activity of the cells.
收红包的
3 Experimental
Fed-batch process were performed with the cellferm-pro® system using the Metabolic Activity tool, with an a priori determined ratio between the substrate supply rate and the oxygen uptake rate (Y S/Y O2). The criteria for the choice of an optimum Y S/Y O2 ratio were: optimum cell growth (maximum IVCC), maximum final product titer and minimum lactate produced per gluco consumed, since lactate is suppod to be one of the main inhibitors in hybridoma cell cultures.
As a reference for the Metabolic Activity bad fed-batch process rved a standard batch process, also performed in cellferm-pro®.
3.1 Culture conditions:
A recombinant mAb creting GS-NS0 cell line has been ud in the prent work as a model system for suspension cell culture. All cultures were carried out in a proprietary rum-free, glutamine-free medium bad on Iscove’s Modified Dulbecco’s Medium (IMDM, Amimed). For the automated Metabolic Activity bad feeding a 10-fold concentrated basal medium was ud. Additionally, 20-fold concentrated Iscove’s amino acids solution (IMDM/AA, Amimed) was dod manually once a day during four subquent days after the cell density reached 1x106 cells/mL. In the first cultivation days NH4HCO3 was dod to the culture to maintain NH4+ concentrations at about 0,5 mM.
The culture vesls ud in the cellferm-pro® system were 1 L Spinner flasks, equipped with glass ball agitator, pH and pO2 nsors and sampling ports, with a minimum start volume of 300 mL and a maximum working volume of 600 mL.
3.2 Analytic:
The concentration of viable cells and viability were automatically determined using the Cedex® system (Innovatis GmbH, Bielefeld, Germany). The IVCC was calculated as described by Sauer et al. (2000).
The concentration of the medium’s main components such as Gluco, Glutamine, Glutamate and Lactate, four ions (Na+, K+, Ca++, NH4+), together with pH, pO2, pCO2 and osmolarity were determined using a Nova Bioprofile TM 200 Analyzer (Nova Biomedical Corp., Waltham, MA, USA)果蝇唾腺染色体
4 Results and discussion
4.1 Choice of an appropriate substrate supply/oxygen uptake ratio
Gluco, one of the most important energy sources for mammalian cell cultures, was chon as the medium component on which the ratio of substrate supply rate to oxygen uptake rate (Y S/Y O2) was bad. The theoretical ratio between substrate and oxygen uptake rates is 0.17 mol/mol, as 1 mole gluco requires 6 mole oxygen for complete oxidation. Since veral esntial substrates, except gluco, could also be limiting, four different Y S/Y O2 ratios were tested around the theoretical one in the range 0.1-0.5 mol/mol. The results are prented in Fig.2.
Figure 2:Choice of the ratio between substrate supply rate and OUR (Y S/Y O2), bad on the three criteria: IVCC, final product titer and minimal ratio of lactate produced per
gluco consumed (Y LP/Y GC).朱明元
跑步的坏处
For the YS/YO2 of 0.1 mol/mol the smallest IVCC was obrved as well as a significantly lower titer. In this ca cell growth cead before the maximal working volume could be reached, assuming cell starvation (data not shown). The highest product titer of 260 mg/l was obrved at YS/YO2 = 0.2. The YLP/YGC ratio incread with increasing YS/YO2. The most excessive
西域都护lactate accumulation was obrved for the ratio Y S/Y O2 = 0.5 mol/mol, reprenting in this ca a theoretical maximum. Bad on the obrvations, the Y S/Y O2 ratio of 0.2 mol/mol was chon for the OUR-bad nutrient feeding. This ratio is clo to the theoretical one of 0.17 mol/mol, indicating the balanced concentration of nutrients in the feed medium.
4.2 Metabolic Activity bad fed-batch
Fig.3 shows a Metabolic Activity bad feeding profile with the corresponding cell density and mAb titer over the process time. No lag pha in cell growth was obrved. Exponential cell growth was obrved for the first four days followed by a quasi-stationary pha of about 5 days. After the maximal working volume was reached on day 9 Metabolic Activity bad feeding stopped and was replaced by stepwi feeding. In this pha the system continued evaluating the volume to be dod, however, was only able to feed when a sample volume was taken and registered in the syste合舟共济
m as a negative value which could be replaced. In this manner, the cells obtained about 6 mL of feed once a day. Under the conditions the cells were kept alive at a cell density above 1,0 x 106 cells/mL for a period of 12 days. After day 15 the cells died and mAb accumulation stopped.
Figure 3:Metabolic Activity bad fed-batch cultivation: feeding profile, cell density and mAb concentr
ation.
