Skip to main content

Suppression of programmed cell death in industrial scale biological production systems

Deliverables

The objective target of these experiments was on the one hand the investigation of the cultivation behaviour of the cell lines in stirred suspension reactors. This reactor type is a standard type used in industry for animal cells up to scales of ~10.000 litres. The ability of the cell lines to provide good growth and product formation characteristics in this reactor type is therefore an essential subject. On the other hand, fed-batch cultivation mode, which in general yields improved results in comparison to batch mode, was tested. An adaptive, model-based process control originally set up for hybridoma cells was successfully transferred to these two new cell lines. The process control strategy models the course of important process states in the reactor. The fit of modelled and measured courses of process states is achieved by adaptation of model parameters. The well-adapted model is then used to predict the future course of the experiment. Based on this prediction, the feed of glucose-/ amino acid-/ vitamin concentrate is calculated with regard to an optimisation criterion (e.g. control of glucose at a certain target concentration). The established process control provided satisfactory results for fed-batch cultivation as well as for fed-batch cultivation using dialysis. The dialysis enables removal of toxic metabolites and improvement of cell growth and antibody production.
The model cell line NS0 6A1 bcl-2 has been cultivated in a high cell density 60 ml fixed bed in continuous operation. Afterwards, the focus has been shifted to a different type of high cell density culture, the dialysis cultivation. For the two model cell lines (NS0 6A1 bcl-2 and NS0 6A1 (100)3) high cell density dialysis cultivation allowed an increase of maximum cell density and product accumulation by a factor of 3 to 7. High cell density dialysis cultivation can be applied to standard batch and fed-batch processes widely used in industry. Focusing on high cell density dialysis cultivation in this work package allowed to investigate dialysis cultivation for bcl-2 cells, their control and the performance of cells over expressing bcl-2 on the one hand and allowed to account for the importance of such high cell density processes in industry on the other hand. The results have already been published: Frahm, B., Lane, P., Märkl, H., Pörtner, R. (2003): Improvement of a mammalian cell culture process by adaptive, model-based dialysis fed-batch cultivation and suppression of apoptosis, Bioprocess and Biosystems Engineering 26, 1-10.
The aim was the characterisation of the two cell lines NSO 6A1 (100)3 and NSO 6A1 bcl-2 regarding growth and product formation characteristics in different cultivation systems. Both cell lines were cultivated in T-flasks as well as in a 2 L foil suspension reactor (Bioengineering, Switzerland) and a 10 L dialysis membrane reactor (Bioengineering, Switzerland). The cultivations of the cell lines provided satisfactory results. Different cultivation modi were performed (batch, fed-batch). The desired characteristics of the cell lines could be observed. The cell line NSO 6A1 bcl-2 showed a prolonged growth phase, a higher maximum cell density and a prolonged stationary phase and apoptotic phase in suspension cultures. These characteristics are desired consequences of the bcl-2 over expression and represent a positive improvement of the cultivation characteristics.
All of the NSO puromycin resistant transfected populations (A-E pEF, A-E Y28A and A-C Bcl-2) were analysed for Bcl-2 content by western blot analysis. Of these NSO transfected populations only four (A, D and E Y28A; and C Bcl-2) were found to contain human Bcl-2 and subsequently those cell lines expressing Bcl-2 and there equivalent empty vector populations (that should not and did not contain Bcl-2) were isolated and grown to stock cultures. Western blotting results on the mixed populations of A, D and E Y28A and pEF did not show any significant difference in Bcl-2 expression levels between batches. Bcl-2 expression was either present in the populations or not and those where expression was found did not vary in the extent of expression level between populations of cells. Upon screening all of the populations in batch culture only population EY28A showed significant protection of apoptosis. Protection from apoptotic cell death was demonstrated in serum-deprived conditions and apoptosis induction by staurosporine compared to the control population (EpEF) and the ild type untransfected population (NSO WT). These populations of cells were then cloned and studied further. Cho cells transfected with the mutant Bcl-2 gene (Y28A bcl-2) offered no protection to apoptosis in similar conditions as those used for the NSO cells in any of the batches tested even though Bcl-2 expression was evident. However, CHO cells transfected with the wild type Bcl-2 (puromycin resistance) showed significant protection from apoptosis and also increased growth rate compared to the wild type and control cells. This population of cells was also sent for cloning to Lonza Biologics but work was suspended due to IP issues. CHO22H11 cells containing the IgG construct were transfected with the wild type neomycin resistance Bcl-2 gene. A control population was also constructed containing the neomycin vector without the Bcl-2. CHO22H11 wild type, CHO22H11 neo and CHO22H11 Bcl-2 populations were tested in batch culture in standard growth conditions (DMEM/F12 10%FCS). The Bcl-2 cell line showed significant protection from apoptosis in normal batch and staurosporine induced conditions compared to the wild type and control vector transfected populations. Strangely, as previously found with the NSO 6A1 populations, the control vector offered some protection in the death phase compared to the wild type population. As previously found with the NSO cell line, it appears that the total antibody production for the three cell lines is relatively equivocal and that towards the end of the batch when you would expect the Bcl-2 cell line to be producing more compared to the control it is in fact producing slightly less antibody. This may be due to metabolic overload of the cells by the Bcl-2 thus the cells are driving their energy resources towards cell division rather than protein production. In order to investigate the protection offered by the neomycin vector alone at the death phase, cell samoles were taken during the batch culture for Bcl-2 expression analysis. Although no Bcl-2 expression was observed for this cell line as the only levels present should be endogenous we observed an increase in Bcl-2 expression throughout the batch culture. This cl-2 expression is thought to be endogenous as it exhibits the traditional two band pattern not found with the transfected Bcl-2 cell lines. No such Bcl-2 expression is found in the wild type cell line. It appears that by an unknown method the neomycin vector is increasing the expression levels of the Bcl-2 within the cells thus offering the cells increased natural protection to the onset of apoptosis.
The NS0 cell line over-expressing Bcl-2 (result 2114) is being evaluated for its potential to act as a potential host cell line for expression of recombinant antibodies. The two cell lines are being taken, in parallel, through the generic cell line construction process used at Lonza Biologics. This is being done using a model antibody. The output from this work package is data detailing the antibody concentrations achievable. Data generated by these activities will form a key part of the data set used to evaulate the benefits of Bcl-2 over expressing cell lines. Both cell lines have been transfected with the genes for a model antibody. Currently, transfectants are going through the rounds of selection that are typically used to select highly productive cell lines. This work package is expected to be complete by mid Q1 2003. The NS0 cell line over-that are typically undertaken during the selection of a highly productive cell line, where increased robustness to apoptosis would, a priori, be expected to offer an advantage. No benefit was seen.
The impact of bcl-2 on culture viability, apoptosis, product yield / productivity, product recovery and other parameters could be demonstrated under non-optimised culture conditions (e.g. standard medium, feed). This has been performed at incubator scale (T-flasks, used for inoculum), standard fed-batch cultivation (at a 2 L scale) as well as for enhanced fed-batch cultivation using dialysis (2 L scale). As a model system the cell lines NS0 6A1 (100)3 (control cell line, not transfected with bcl-2) and NS0 6A1 bcl-2 (transfected with bcl 2) have been used. Consistent results have been obtained at all stages. The improvement of the bcl-2 over-expression can be clearly seen in the prolonged growth phase and a higher maximum cell density resulting in a higher antibody production. The final product has been analysed with respect to purity with particular attention to the degree of contamination with cellular DNA. Concerning the different processes, improved growth and product formation characteristics were obtained using dialysis-fed-batch cultivation in comparison to standard fed-batch cultivation. In addition, the successful performance of the developed fed-batch control could be demonstrated. The results have already been published: Frahm, B., Lane, P., Märkl, H., Pörtner, R. (2003): Improvement of a mammalian cell culture process by adaptive, model-based dialysis fed-batch cultivation and suppression of apoptosis, Bioprocess and Biosystems Engineering 26, 1-10.
All of the NSO transfected populations (A-E pEF, A-E Y28A and A-C Bcl-2) were analysed for Bcl-2 content by western blot analysis. Of these NSO transfected populations only four (A, D and E Y28A; and C Bcl-2) were found to contain human Bcl-2 and subsequently those cell lines expressing Bcl-2 and there equivalent empty vector populations (that should not and did not contain Bcl-2) were isolated and grown to stock cultures. Western blotting results on the mixed populations of A, D and E Y28A and pEF did not show any significant difference in Bcl-2 expression levels between batches. Bcl-2 expression was either present in the populations or not and those where expression was found did not vary in the extent of expression level between populations of cells.
An adaptive, model based process control for fed-batch suspension cultures has been successfully developed and validated for the model cell lines NS0 6A1 (100)3 and NS0 6A1 bcl-2. This control was used for the enhanced fed-batch cultivation of these cell lines. Besides, it served for more detailed process evaluation. The adaptive, model-based process control was successfully transferred from fed-batch control to the control of the dialysis fed-batch process. In addition, performed time-space yield considerations led to a deeper understanding of the processes. Comparison to hybridoma cell cultivations using the process control showed, how exact the control strategy reacts to different cell lines and culture conditions. Again, the improvement of the bcl-2 over-expression can be clearly seen in the prolonged growth phase and a higher maximum cell density resulting in a higher antibody production. The results have already been published: Frahm, B., Lane, P., Märkl, H., Pörtner, R. (2003): Improvement of a mammalian cell culture process by adaptive, model-based dialysis fed-batch cultivation and suppression of apoptosis, Bioprocess and Biosystems Engineering 26, 1-10 The fed-batch process control is also presented in a published book.: Frahm, B. (2003): Adaptive, modellgestützte Prozessführung von Suspensionskulturen tierischer Zellen, ISBN 3-8334-0111-7.
The NS0 cell line over-expressing Bcl-2 was evaluated for its potential to act as a host cell line for expression of recombinant antibodies. The cell line was taken through the generic cell line construction process used at Lonza Biologics using a model antibody. Antibody accumulation and growth kinetics were evaluated at different stages in the process. Finally, cell lines were evaluated in bioreactors using a scale-down model (10 L) of a process operated at 5000 L. No benefit was seen from using the Bcl-2 over-expressing cell lines: antibody concentrations were substantially lower than those seen with the control cell lines.
The behaviour of a GS-NS0 cell line expressing a model antibody was compared with that of the same cell line over-expressing the Bcl-2 protein (this cell line was generated in a previous EU project) in state-of-the-art antibody manufacturing process. The process was scaled down from a fermentation process used at the 5000 L scale for the commercial manufacture to GMP of therapeutic antibodies by Lonza Biologics. The expected increase in space-time yield of viable biomass was seen. However, no increase in antibody concentration was seen: this is in contrast with results generated with these cell lines by one of the contractors (Al-rubeai) in the previous EU-funded programme. The over-expression of Bcl-2 also appeared to adversely affect the operations to separate cell debris from the feed stream for the purification process. These data, which to the contractors' knowledge are the first time that Bcl-2 over-expressing cell lines have been evaluated in a state-of-the-art manufacturing process, show that introduction of such cell lines into current manufacturing processes will not simply be a case of making minor adjustments to existing processes. Work has been presented at a number of meetings where other manufacturers of therapeutic antibodies were present. This work package is now complete.
The behaviour of a GS-NS0 cell line expressing a model antibody was compared with that of the same cell line over-expressing the Bcl-2 protein (this cell line was generated in a previous EU project) in state-of-the-art antibody manufacturing process. The process was scaled down from a fermentation process used at the 5000 L scale for the commercial manufacture to GMP of therapeutic antibodies by Lonza Biologics. The expected increase in space-time yield of viable biomass was seen. However, no increase in antibody concentration was seen: this is in contrast with results generated with these cell lines by one of the contractors (Al-rubeai) in the previous EU-funded programme. The over-expression of Bcl-2 also appeared to adversely affect the operations to separate cell debris from the feedstream for the purification process. These data, which to the contractors' knowledge are the first time that Bcl-2 over-expressing cell lines have been evaluated in a state-of-the-art manufacturing process, show that introduction of such cell lines into current manufacturing processes will not simply be a case of making minor adjustments to existing processes. Work has been presented at a number of meetings where other manufacturers of therapeutic antibodies were present.