June 15, 2007 (Vol. 27, No. 12)
Michael Wathen Ph.D.
Mark Gerber Ph.D.
Kevin J. Kayser Ph.D. Director R&D Upstream and Systems MilliporeSigma
James S. Ross
Matthew V. Caple
Simultaneous Clone and Media Screening
Optimizing the mammalian cell culture process for the production of recombinant proteins such as humanized IgG represents one of the most critical factors in determining the economic feasibility of a biopharmaceutical product.
While selection of a stable, highly secreting clone is the first critical step in this process, media optimization, feed strategies, and bioreactor process parameters all play a critical role in maximizing recombinant protein production. Since productivity levels with a particular media formulation can vary greatly between different clones, we have developed strategies to integrate the clone and media selection steps for identifying the highest producing recombinant clones.
Standard clone-selection methodology involving limited-dilution cloning is a time-consuming, hit-or-miss process that often requires six months of intense labor to generate small cultures of a chosen clone. We have used a new clone-selection technology called Cell Xpress to isolate high-producing CHO cell clones.
The Cell Xpress clone-selection technology is a combination of high-speed image scanning and quantitation with high-speed laser manipulation that allows for in situ measurement of recombinant protein secretion on a single cell basis followed by laser-mediated destruction of the low-producing cells in a population.
Cell Xpress utilizes the LEAP™ (Laser-Enabled Analysis and Processing) platform and incorporates an immunofluorescence-based assay for identifying the highest secreting clones in either transfected or stable pools of cells. Cells are seeded into 384-well tissue culture plates in the presence of a matrix designed to capture secreted IgG molecules.
After overnight incubation, captured IgG is detected via a secondary reagent containing a conjugated fluorophore, and the viable cells are stained with live cell tracking dye. Following a short incubation, the captured IgG secretions and associated live cells are visualized using the LEAP instrument. The Cell Xpress software can then be used to identify and retain the highest producing cells in the population for cloning and expansion while removing the undesired cells from the population by laser-induced apoptosis.
The Cell Xpress system can rapidly screen and process a 384-well plate with 150–200 cells/well in less than one hour making it possible to evaluate over 40,000 individual cells for selecting the highest secreting clones (Figure 1A). In the current study, eight clones were selected from two separate 384-well plates and expanded to spinner flask cultures (60-mL volume) within a 10–12 week period.
Of the eight clones generated, five exhibited increased specific productivity (qP) ranging from 15–40 pg/cell/day compared to the initial cell line that had a specific productivity of 10–12 pg/cell/day under identical growth conditions. Four of these clones (Figure 1B) were chosen for media screening to identify the preferred clone for process development.
The CHO Media Library
We have collected a set of CHO media formulations from the SAFC Biosciences (www.safcbiosciences.com) historical library of formulations that have a proven track record in industrial applications. The CHO Media Library formulations were chosen based on growth capabilities with numerous cell lines, including parental CHO-K1, CHO-S, and CHO-DG44, and for high productivity of recombinant mAbs.
Performance of each formulation was evaluated based on peak cell density, integrated cell days, viability, culture longevity, and where applicable, specific and volumetric productivity. These media formulations do not require extensive cell line adaptation.
The 24 formulations in this screening library were selected from hundreds of formulations and are all animal-component free. Additionally, some of the media in this library are protein-free and/or chemically defined formulations. Several formulations were developed by employing Design of Experiment (DOE) to blend formulations exhibiting high growth rates together with formulations yielding high specific productivity. The statistical software can analyze multiple criteria simultaneously to determine synergistic responses, resulting from mixing the media formulations, and predict the optimal mixture for the desired outcome.
Small-scale Bioreactor Systems
Small-scale fed-batch systems are used to model performance of a clone in process development prior to advancing to production scale. Several process development groups have reported using small bioreactors known as a TubeSpin (Techno Plastic Products, www.tpp.ch) for modeling bioreactor and shake flask performance. Scaling down the size of cultures provides a tremendous advantage for screening cell clones in a media library.
TubeSpin tubes are the size of 50-mL conical centrifuge tubes and offer a useful alternative to shake flasks since they use smaller volumes of media (typically 30 mL) and conserve incubator space. We have compared the growth and productivity of an IgG-producing CHO cell line in TubeSpin tubes, shake flasks, and bioreactors. While TubeSpin cultures exhibited more rapid early growth, the overall viable cell densities for the entire culture period as measured by area under the curve and the cultures’ IgG productivity were equivalent to shake flasks. As expected, bioreactors produced higher levels of IgG than either TubeSpin tubes or shake flasks but the effects of different media observed in bioreactors were the same as in the other two vessels.
Screening Clones in the CHO Media Library
Using Cell Xpress, four clones were selected to be evaluated for improved growth and productivity in the CHO Media Library of 24 different formulations. The clones were adapted to each media formulation (two passages) before the start of the growth kinetics study. Viable cell density was determined by trypan exclusion utilizing the Cedex Automated Cell Culture Analyzer (Innovatis, www.innovatis.com). The clones were seeded at a density of 2 x 105 cells/mL in TubeSpin tubes containing 30 mL of media. The cultures were incubated in a shaking incubator (Adolf Kuhner, www.kuhner.com) with set points of 200 rpm, 6% CO2, and 36ºC. Samples for viable cell density determination were taken starting on day 3 of culture and concluding on day 14. IgG titers were determined on day 14 using a Protein G HPLC assay.
Growth and productivity data for the four clones in the media library are shown in Figures 2 and 3. Formulations that did not support growth during adaptation are not depicted in these figures. In general, media performed similarly for the four clones.
For instance, Media #9 showed the highest growth and productivity for three of the four clones and was ranked second best for the fourth clone. Media #9 is a chemically defined formulation resulting from the blending of a high-growth formulation with a high-productivity formulation using DOE as described above.
There were notable exceptions in media performance between the clones, however. Media #16 only performed well with Clone 1. Media #5 performed well with three of the clones but exhibited very poor growth with Clone 2. These exceptions help to illustrate the need for media screening during clone selection to match each clone with its optimal media. The combination of Clone 4 and Media #9 yielded the highest IgG titer in this study with a 2x improvement in productivity compared to the initial media formulation. However, this clone was not the highest producer in the customer’s original media formulation and may not have been prioritized without the additional media screening.
We have combined a novel clone selection technology with early media screening to rapidly identify high-quality clones for cell-line generation. The Cell Xpress platform provides a powerful approach to analyze therapeutic protein production on a single-cell basis and select highly expressing cells for clonal expansion.
Compared to traditional methods, single-cell clones derived from Cell Xpress have superior or comparable frequencies of high producing clones and can be developed in a shorter period of time. Incorporating media screening into the clone selection process provides the opportunity to assess the productivity of multiple clones in their individual preferred media, allowing the investigator to make more informed decisions in prioritizing clones for cell line development.
Michael Wathen, Ph.D., (E-mail [email protected]) is R&D manager—new technology, analytical & automation development, Bradley Dalton is R&D scientist, Mark Gerber, Ph.D., is senior scientist, Kevin J. Kayser, Ph.D., is R&D manager—cell engineering, James S. Ross is R&D manager—industrial cell culture, and Matthew V. Caple is director, R&D, all in cell sciences and development at SAFC Biosciences. Website: www.safcbiosciences.com. Phone: (913) 469-5580.