Accelerating the development time of recombinant mammalian cell lines producing high titers of protein therapeutics

The forthcoming arrival on the market of numerous protein therapeutics that require high clinical doses will foster the need for high-producing mammalian cell lines. Furthermore, pressures to reduce the development time for new therapeutics has meant more emphasis on efforts to accelerate the process that yields such cell lines. The objective of this thesis work was to develop methods to decrease the time needed to generate stable and high producing mammalian cell lines. This could be achieved by overcoming some of the major limitations that hamper the process of cell lines development. The host systems chosen for this study were Chinese hamster ovary (CHO DG44) and baby hamster kidney (BHK-21). The recombinant proteins used as reporters were the green fluorescent protein (GFP) and a human anti-Rhesus D IgG. One limitation hampering the development of cell lines producing high recombinant protein titers is the extensive variability of transgene expression within a transfected population as a result of the gene delivery process. This decreases the reproducibility of the generation of high producing cell lines. A direct comparison of two gene chemical delivery methods, polyethylenimine (PEI)- and calcium phosphate (CaPO4)-mediated transfection, with a mechanical method (microinjection) provided some insights into ways to overcome this limitation. Indeed, a correlation was observed between the amount of plasmid DNA delivered into cells and the specific productivity of the recovered recombinant cell lines. Hence, by increasing the amount of DNA delivered per cell, IgG-producing clones with specific productivities up to 22 pg/cell/day were recovered with a high efficiency. A second limitation is the time needed to assure the stability and clonality of cell lines by current analytical methods. This was overcome by using GFP as a reporter for the level of production of a co-expressed therapeutic protein. The co-transfection of plasmids individually carrying the GFP and IgG heavy (Hc)- and light chain (Lc) genes, resulted in integration of all three plasmids at the same site in the cell's genome. As a consequence, the stability of GFP expression correlated with that of IgG expression. This proved to be advantageous finding and eliminating unstable cell lines early in the process of selection. As a consequence, it was possible to derive in serum-free medium clones with a specific productivity almost 6 times higher than that of clones derived from transfection performed with the IgG Hc and Lc vectors only. A third limitation to the rapid recovery of stable cell lines is the time devoted to the cultivation of transfected cells in selective medium, normally one month. Here, the time of selection was shortened to one week by increasing the stringency of the selective pressure. The results from studies of plasmid integration kinetics by fluorescence in situ hybridization (FISH) showed that the increased stringency of selection decreased the time needed to enrich a transfected population with recombinant cells. Furthermore, stable clones were shown to be present in this population after one week of selection and could be uncovered by identifying patterns of GFP expression predictive of production stability. By limiting the screening to clones exhibiting these stable patterns of GFP expression, the number of clones needed in order to isolate stable cell lines producing high IgG titers could be dramatically reduced. In conclusion, these results combined together provided a new procedure to isolate within a week of the start of selection, stable and high-producing clones. 30-40% of the clones isolated using this procedure were stable for at least three months and their average specific productivity was 29 pg/cell/day.

Calcium phosphate transfection generates mammalian recombinant cell lines with higher specific productivity than polyfection

Sébastien Chenuet, Madiha Derouazi, David Hacker, Nicolas Jaccard, Florian Maria Wurm

Transfection with polyethylenimine (PEI) was evaluated as a method for the generation of recombinant Chinese hamster ovary (CHO DG44) cell lines by direct comparison with calcium phosphate-DNA coprecipitation (CaPO4) using both green fluorescent protein (GFP) and a monoclonal antibody as reporter proteins. Following transfection with a GFP expression vector, the proportion of GFP-positive cells as determined by flow cytometry was fourfold higher for the PEI transfection as compared to the CaPO4 transfection. However, the mean level of transient GFP expression for the cells with the highest level of fluorescence was twofold greater for the CaPO4 transfection. Fluorescence in situ hybridization on metaphase chromosomes from pools of cells grown under selective pressure demonstrated that plasmid integration always occurred at a single site regardless of the transfection method. Importantly, the copy number of integrated plasmids was measurably higher in cells transfected with CaPO4. The efficiency of recombinant cell line recovery under selective pressure was fivefold higher following PEI transfection, but the average specific productivity of a recombinant antibody was about twofold higher for the CaPO4-derived cell lines. Nevertheless, no difference between the two transfection methods was observed in terms of the stability of protein production. These results demonstrated the feasibility of generating recombinant CHO-derived cell lines by PEI transfection. However, this method appeared inferior to CaPO4 transfection with regard to the specific productivity of the recovered cell lines.

2008

Lentiviral Vector-Mediated Gene Delivery for the Generation of Chinese Hamster Ovary Cell Lines with Improved Productivity and Stability

Agata Oberbek

Recombinant cell line generation by standard transfection techniques is a time-consuming and labor-intensive process often leading to an unpredictable outcome as transgene integration is a rare, random event. Consequently, the population of cells obtained after standard gene delivery is heterogeneous in terms of specific growth and recombinant protein productivity. Therefore, genetic selection is usually applied to the population for up to 3 weeks to eliminate non-recombinant cells. A large number of the surviving cell lines must be screened to find a few which stably produce the recombinant protein at the desired level. These must be maintained in culture for 2-4 months in the absence of selection to assess the stability of protein production. In contrast, lentiviral vectors (LVs) deliver the recombinant gene to target cells through infection and a controlled integration process, which allows the inserted transgene to become a permanent genetic element of the transduced cell. Moreover, lentiviruses tend to integrate into transcriptionally active sites of the host cell genome. Hence, LV-mediated gene transfer was expected to increase the probability of generating a stable high-producing clone in comparison to the standard methods based on plasmid transfection. Therefore, the objective of this PhD thesis was to improve the efficiency of stable cell line generation with the use of LVs. For this purpose, LVs carrying the enhanced green fluorescent protein (eGFP) gene under the control of either the cytomegalovirus immediate early promoter/enhancer (CMV) or the elongation factor-1 alpha promoter (EF-1α) were generated and used to infect Chinese hamster ovary (CHO) cells. Two days after gene delivery at a multiplicity of infection (MOI) of 2, up to 98% of infected cells were eGFP-positive. Following isolation by fluorescence-activated cell sorting (FACS), CHO cell pools and clonal cell lines were analyzed for the stability of eGFP expression over time. In these studies, the EF-1α promoter was found to yield more stable and homogeneous protein expression compared to the CMV promoter. For comparison, cell pools and cell lines were generated by transfection with the LV transfer plasmid bearing the eGFP gene. The level and stability of eGFP expression was greater in LV-generated pools and cell lines than in those established by transfection. Subsequently, an LV expressing the tumor necrosis factor receptor-Fc (TNFR-Fc) fusion gene and the eGFP gene from a bicistronic mRNA under the control of the EF-1α promoter (LV_EF-TNFR) was generated and used to infect CHO cells. Based on selective FACS gating for high eGFP expression, stable CHO clones showing volumetric productivities of up to 100 mg/L of TNFR-Fc in a non-optimized batch process could be isolated within 2 days of infection at an MOI of 2. The level of expression of the recombinant protein remained constant for 3 months in serum-free suspension culture and in the absence of chemical selective agents. Next, CHO cells were infected at 3 higher MOIs with the LV_EF-TNFR and clonal cell lines with high eGFP-specific fluorescence were recovered by FACS at 2 d post-infection. These cell lines had volumetric TNFR:Fc productivities ranging from 100 to 400 mg/L in 4-day cultures with cell-specific secretion rates up to 100 pg/cell/day. Moreover, recombinant cell lines developed by the LV-based technology could be successfully cultured in bench-scale bioreactors, yielding up to 1.5 g/L of TNFR:Fc under batch mode operation. The increase in integrated TNFR:Fc copy number was found to correlate with the MOI, and resulted in improved protein productivity. However, this relationship was found to be non-linear: cell saturation in terms of protein productivity and integrated transgene copy number was observed at an MOI of 100. In order to determine the shortest time-frame necessary for the production of satisfactory quantities of a recombinant product, the high MOI-infection approach was tested on pools of transduced cells. Without a priori FACS isolation, recombinant pools of CHO cells expressing on average 500 mg/L of TNFR:Fc within 8 days of inoculation at 3 × 105 cells/ml could be obtained. This strategy provided a simple protocol for high titer recombinant protein production, executable within 4 weeks from LV batch preparation to product harvest. Since large-scale production of LVs would be needed for industrialized bioprocesses, a method for LV production in serum-free suspension culture was developed. However, the best titers generated by this system were still approximately 20-fold lower than the yields obtained regularly from the adherent cultures, thus leaving room for further optimization. In conclusion, LV-mediated gene transfer provided an efficient alternative to plasmid transfection for the generation of stable and high-producing recombinant cell lines. Moreover, it proved amenable to large-scale manufacture and could yield several hundred milligrams per liter of the recombinant product without the need for any kind of genetic selection. In addition, using the LV technology described here it seems possible to target the number of copies of the delivered transgene to a specific range which balances favorable growth with good productivity characteristics.

EPFL2010

Study of transposon-mediated cell pool and cell line generation in CHO cells

Sowmya Balasubramanian

The goal of this thesis was to evaluate the use of artificial transposon systems for the generation of recombinant cell pools and cell lines with Chinese hamster ovary (CHO) cells as the production host. Transposons are naturally occurring genetic elements present in the genome of most organisms. Several transposons of metazoan origin have been engineered to facilitate the integration of one or more recombinant genes (transgenes) into the genome of the host cell. In this thesis, three of the most commonly used transposons, piggyBac (PB), Tol2, and Sleeping Beauty (SB), were used for transgene delivery into CHO-DG44 cells. The transposon systems described here involves co-transfection of the host cells with a helper plasmid for the transient expression of the transposase, and a donor plasmid coding for the transgene and a gene for the selection, both positioned between two transposon repeat sequences that are required for DNA transposition. Through the optimization of various selection parameters we showed that PB-mediated cell pools can be generated with selection duration of as little as 5 days in the presence of puromycin. All three transposon systems, PB, Tol2, and SB, resulted in cell pools with similar volumetric TNFR:Fc productivities that were about 9 times higher than those generated by conventional plasmid transfection. Transposon-mediated cell pools had 10 - 12 transgene integrations per cell. However, we demonstrated that some the integration events occurred via DNA recombination rather than transposition. We also isolated clonal cell lines from cell pools. As expected, the average volumetric TNFR:Fc productivity of transposon-derived cell lines was higher than that of cell lines generated by conventional transfection. In 14-day fed-batch cultures, protein levels up to 900 mg/L and 1.5 g/L were obtained from transposon-mediated cell pools and cell lines, respectively. The stability over time of the volumetric productivity of cell pools was determined by maintaining the cells in culture for 3 months in the absence of selection. In general, the productivity decreased to 50 % its initial level over the first 7 weeks in culture and then remained constant for the following 5 weeks. In contrast, the volumetric protein yield from transposon-mediated cell lines remained constant for up to 4 months in the absence of selection. We also showed that the three transposon systems could be used for cell pool generation with CHO-K1 and CHO-S with similar volumetric productivities as observed with CHO-DG44 cells. Finally, we utilized the PB transposon system for generating cell pools co-expressing up to 4 different transgenes, enhanced green fluorescent protein (EGFP), secreted alkaline phosphatase (SEAP), and the light and heavy chains of an IgG1 monoclonal antibody, by simultaneous transfection of all four transgenes. We showed that PB-mediated cell pools had increased volumetric productivity of each of the proteins compared to those generated by conventional co-transfection. The use of the PB transposon system increased the percentage of cells in the pools that were co-expressing all four proteins as compared to the results with cell pools generated by conventional transfection. In conclusion, the transfection of CHO cells with the PB, Tol2 or SB transposon system is a simple, efficient, and reproducible approach to the generation of cell pools and cell lines for the rapid production of recombinant proteins.