Improving Protein Production in CHO Cells

June 2, 2008
Mark Stramaglia

Senior Product Manager at Invitrogen Corporation

,
Richard Fike

Principal Scientist, Cell Systems Division at Invitrogen Corporation

,
Borka Naumovich

Associate Scientist III at Invitrogen Corporation

,
David (Xiaojian) Zhao

Technical area manager of media development at Invitrogen Corporation

BioPharm International, BioPharm International-06-02-2008, Volume 2008 Supplement, Issue 5

Using chemically defined feeds with CHO cell lines not only eliminates the variability associated with using plant hydrolysates, but could also improve the productivity of biopharmaceutical protein manufacture and help move therapeutic proteins into clinical trials more rapidly.

Abstract

With the drive to remove serum and products of animal origin from cell-culture media during manufacture of protein-based biologics, plant hydrolysates have become popular as feeds in fed-batch cell culture to boost productivity. However, plant hydrolysates often contain undefined levels of nutrients, which can result in lot-to-lot variation in fed-batch performance and inefficiency. These are disadvantages when consistent protein yields are required from recombinant Chinese hamster ovary (CHO) cell lines. To achieve a more controlled process, one approach is to use chemically defined formulations as feeds containing only a carbon source, amino acids, vitamins, and trace elements. As a test of this approach, we compared cell growth and protein production characteristics of CHO cell lines grown in chemically defined media supplemented with either chemically defined nutrient feed supplements or plant hydrolysates alone, or with nutrient feed supplements.

Removing serum and products of animal origin from cell-culture media during production of therapeutic proteins from Chinese hamster ovary (CHO) cells offers improved biosafety and can make regulatory approval smoother. However, not having these supplements in the media can limit CHO cell growth and protein production. One method of improving growth is to add plant protein hydrolysates as a feed because they provide a highly pure source of soluble amino acids, peptides, vitamins, and essential elements for cell culture. Supplementing media with plant hydrolysates has been shown to improve protein production from engineered CHO cells.1,2

Invitrogen Corporation

The drawback of using plant hydrolysates is that they often contain undefined levels of nutrients and can result in lot-to-lot variation in fed-batch performance and purification inefficiency, which are disadvantages when consistent protein yields are required.3 To achieve a more controlled process, it is important to begin with a base medium for batch culture that does not make feeding strategies more complicated. An alternative approach to using nutrient-rich plant hydrolysates is to use a medium that is specifically designed to promote rapid cell growth and high levels of protein expression for CHO cells as a basal media, and then add chemically defined feeds containing only a carbon source, amino acids, vitamins, and trace elements to replenish the nutrients that are being depleted in the media.4 This article describes the experimental performance of chemically defined nutrient feed supplements compared to plant hydrolysates in a chemically defined medium with a recombinant CHO cell line.

Experimental Approach

Shake Flask Studies

In a first shake flask study (Figure 1), recombinant CHO cells engineered to express a proprietary antibody were inoculated at 3 x 105 cells/mL into 30-mL shake flasks containing CD OptiCHO Medium (Invitrogen, Grand Island, NY) as a base. The cells were incubated at 37 °C in 8% CO2, and agitated at 125 rpm. To determine the effect that plant hydrolysates or a chemically defined medium had on protein production, on days 3, 6, and 9, a 10% v/v solution of one of the plant hydrolysates derived from pea, wheat, soy, or rice (Kerry Bio-Science, Rochester, MN) was added at a stock concentration of 100 g/L to the cultures or the media were supplemented with a 10% v/v chemically defined feed, CHO CD EfficientFeed A (Invitrogen, Grand Island, NY). For each flask containing a different type of feed, there were three replicates made and samples were withdrawn for high performance liquid chromatography (HPLC) analysis of IgG concentration from each replicate every 24 hours on days 7–14. In those instances, when cell viabilities dropped below 20%, samples were no longer obtained.

Figure 1

In a second shake flask study to determine if protein production could be increased by using a combination of feeds (Figure 2), 3 x 105 cells/mL were inoculated into 30-mL shake flasks containing the same base medium. They were incubated at 37 °C in 8% CO2, and agitated at 125 rpm. On days 3, 6, and 9, a 10% v/v addition of the same solutions of pea, wheat, soy, or rice plant hydrolysates plus a 10% v/v solution of chemically defined feed A were added simultaneously, or a 10% v/v solution of chemically defined feed A was added alone to the cultures. For each flask containing a different feed combination, there were three replicates and samples were withdrawn for HPLC analysis of IgG concentration from each replicate every 24 hours on days 7–14.

Figure 2

Bioreactor Studies

To determine the effects different feeds had on cell numbers and protein expression, a 15-L stir-tank bioreactor (Applikon Biotechnology, Foster City, CA) containing CD CHO medium was inoculated with a different proprietary recombinant IgG CHO cell line at 3 x 105 cells/mL. The pH and dissolved oxygen were set at 6.9 and 50%, respectively, with agitation at 150–175 rpm. The CO2 and oxygen sparge rates were both set at 100 mL/minute. Glucose was added to 6 g/L if the concentration in the media in the bioreactor fell to less than 2 g/L. GS-Max (HyClone, Logan, UT) was fed at 1% of starting volume if the level of glutamate fell to less than 1.5 mM, as occurred in all bioreactors. Four different feed strategies were carried out over 10 days. In one bioreactor (Figure 3), 3% v/v addition of soy hydrolysate was added from a stock of 40 g/L every day for 10 days. In another bioreactor (Figure 4), a 3% v/v solution of chemically defined feed A, (CHO CD EfficientFeed A) was added on days 0–5 followed by addition on days 6–10 of daily supplementation of a 3% v/v solution of chemically defined feed B, (CHO CD EfficientFeed B). The additional two feed strategies determined the effects of higher concentrations of the chemically defined feeds but at fewer time points: 1) a 10% v/v solution of chemically defined feed A was added on days 0 and 3, followed by a 10% v/v solution of chemically defined feed B on days 6, 9, and 2) a 15% v/v solution of feed A was added on day 0 followed by addition of a 15% v/v solution of feed B on day 6. Both total cell counts and protein titers, as determined by ELISA over 13 days, are presented in the results (Figures 3 and 4).

Figure 3

Results

In the first shake flask study, addition of the plant hydrolysates produced protein yields of around 100–125 μg/mL (Figure 1). Culture viability using any of the hydrolysates did not extend beyond 12 days, while viabilities remained high with EfficientFeed A for the 14-day duration of the experiment. Interestingly, the soy and rice hydrolysates maintained viability for an extra day. This result is consistent with published data where the ability of soy hydrolysates to outperform other plant hydrolysates has been previously noted.5 The chemically defined feed attained protein yields of greater than 400 μg/mL, a four-fold yield improvement compared to any of the plant hydrolysates tested.

Figure 4

In the second shake flask study, addition of the plant hydrolysates and the chemically defined feed produced protein yields from around 175–375 μg/mL (Figure 2). Addition of chemically defined feed and plant hydrolysates together extended cell viability to 14 days with all plant hydrolysate combinations and resulted in increased protein yields by as much as three-fold when compared to the performance of plant hydrolysate alone. Soy and rice hydrolysate combinations with the chemically defined supplement generated the highest yield, while pea and wheat hydrolysate combined with chemically defined feed generated the lowest bioproductivity. However, when the chemically defined feed was added alone, it achieved higher protein yield than the feed and plant hydrolysate combinations. The lower yields obtained by adding plant hydrolysates and chemically defined feed together may be explained by osmolality increases, although this was not investigated, and therefore, cannot be proved in this study.

In the bioreactor study, after the daily addition of 3% v/v soy hydrolysates, a total cell count of 7 x 106 cells/mL was achieved at day 9, compared to 9–11 x 106 cells/mL with various combinations of defined supplement and feeding strategies (Figures 3 and 4). The addition of soy hydrolysate resulted in a lower cell growth rate up to day 9, when the soy bioreactor was terminated because of contamination. The growth rates of cells grown in the presence of 10% and 3% v/v chemically defined feeds were similar and both achieved a maximum cell count of ~11 x 106 cells at day 9, while those grown in 15% v/v chemically defined feeds were lower, with cells achieving a maximum cell density of ~9 x 106 at day 9.

Cell density does not necessarily have a direct relationship to protein expression. Cells supplemented with 15% feed A on day 0 and an addition of 15% feed B on day 6, which achieved the highest cell density, expressed comparable protein production values to those grown with other feed strategies. Cells grown in the presence of 3% v/v soy hydrolysate compared to the 3% v/v chemically defined feeds showed slightly less than half the cell density, yet protein production was only around 10% lower (0.7 mg/mL at day 9 with soy hydrolysate and 0.8 mg/mL at day 9 with chemically defined medium). Lower osmolality for the soy hydrolysate condition may have contributed to the lower productivity seen. Overall, the best bioproductivity resulted when using a counter-intuitive feeding strategy—feeding with a relatively large bolus of supplement, both at initiation of culture and at one more timepoint before the culture lag phase. Therefore, consideration of multiple defined nutrient supplement formulations is suggested when maximum protein expression is required.

Conclusion

Results demonstrated that for the recombinant CHO cell line studied, the addition of chemically defined feed to the basal medium produced approximately a four-fold greater protein yield than any of the four different plant hydrolysates used. Also, culture supplementation with different plant hydrolysates in combination with chemically defined feeds provided no advantage in terms of increased protein production over the addition of a chemically defined feed on its own. Supplementing media with plant hydrolysates may not always be the most efficient choice or may not offer the same increases in productivity available from chemically defined feeds. Therefore, when culturing CHO cells for maximum protein production, it may be worthwhile to compare different feed types. Bioreactor studies of a different CHO cell line also demonstrated that chemically defined feeds in the culture provided slightly higher protein yields than the addition of a soy hydrolysate when the two feeds were added at the same concentration. In addition, using a higher concentration of chemically defined feed A on day 0 followed by feed B on day 6 yielded superior bioproductivity and proved the value of including multiple, and sometimes counter-intuitive, feeding strategies when trying to maximize protein production.

In summary, using chemically defined feeds with CHO cell lines not only eliminates the variability associated with using plant hydrolysates, but could also improve the productivity of biopharmaceutical protein manufacture and help move therapeutic proteins into clinical trials more rapidly.

David (Xiaojian) Zhao is a technical area manager of media development, Richard Fike is a principal scientist, Cell Systems Division, Borka Naumovich is an associate scientist III, and Mark Stramaglia is a senior product manager, all at Invitrogen Corporation, Grand Island, NY, 716.774.6860, mark.stramaglia@invitrogen.com

References

1. Mols J, Peeters-Joris C, Wattiez R, Agathos SN, Schneider YJ. Recombinant interferon-gamma secreted by Chinese hamster ovary-320 cells cultivated in suspension in protein-free media is protected against extracellular proteolysis by the expression of natural protease inhibitors and by the addition of plant protein hydrolysates to the culture medium. In Vitro Cell Dev Biol Anim. 2005. Mar–Apr; 41(3–4):83–91.

2. Burteau CC, Verhoeye FR, Mols JF, Ballez JS, Agathos SN, Schneider YJ. Fortification of a protein-free cell culture medium with plant peptones improves cultivation and productivity of an interferon-gamma-producing CHO cell line. In Vitro Cell Dev Biol Anim. 2003 Jul–Aug;39(7):291–6.

3. Luo Y, Chen G. Combined approach of NMR and chemometrics for screening peptones used in the cell culture medium for the production of a recombinant therapeutic protein. Biotechnol Bioeng. 2007 Aug 15;97(6):1654–9.

4. Stein A. Decreasing variability in your cell culture. Biotechniques. 2007 Aug;43(2):228–9.

5. Chun BH, Kim JH, Lee HJ, Chung N. Usability of size-excluded fractions of soy protein hydrolysates for growth and viability of Chinese hamster ovary cells in protein-free suspension culture. Bioresour Technol. 2007 Mar;98(5):1000–5.