February 1, 2014 (Vol. 34, No. 3)

Ingrid Lange Scientific Lead Cobra Biologics
Kristina Lae Scientist Cobra Biologics
Ulrica Skoging-Nyberg, Ph.D. Service Delivery Manager Cobra Biologics

Novel Approach for Cell-Line Development and Improving Decisions Making at Early Stages

Requirements for rapid cell-line and process development have pushed an increasing number of development programs toward a higher throughput at smaller scale. Tight timelines are important when developing a platform approach for an antibody program from cell-line development to production of clinical material.

Evaluation of cell culture media and additives, as well as process parameter optimization, can be extremely time consuming. A high-throughput system that can handle several bioreactors at small scale has been shown to be useful in making upstream process development and cell-line development more efficient and improving decision-making processes. To that end, Cobra Biologics has looked into how it can integrate high-throughput development tools into its existing MaxXpress cell-line development platform.

A typical development program for the production of clinical material consists of cell-line development, process development, production material for toxicology studies, and further production of material for clinical trials. Traditionally, cell culture process development is performed after the final production clone has been selected during cell-line development.

However, the request for tight timelines not only demands the use of high-throughput development tools such as small-scale bioreactor systems but also the possibility of performing cell-line and process development in parallel.

Before a development program can progress to the production of clinical material the final production clone needs to be selected and the production process must be fully defined. Both selection of the cell culture process and production clone are critical decisions and need to be based on representative bioreactor data.

This article describes the implementation of a high-throughput system (ambr™ by TAP Biosystems) where several clones were taken to process development before the selection of the final production clone and the production process.

Titer Optimization

The microbioreactor system has also been used as a tool for optimizing titer in a fed-batch process. Figure 1 shows fed-batch optimization of a CHO-S cell line producing a mAb. Basal media was screened in batch mode before two basal media were selected to be taken further to the first feed screen in ambr.

For the second round, one basal media and one feed were selected. The design of experiment (DoE) setup with parameters like feed volume and feed start resulted in an increase of the mAb titer to around 2 g/L. In the third round process parameters like pH, temperature, and DO were evaluated in a DoE setup.

Amino acid analysis of cell culture samples led to the extra addition of depleted amino acids in the last round using ambr, and the titer was increased to 2.9 g/L. Due to the quick turnaround setup time and cleaning of the microbioreactor system, as well as the in-house analytical capability, the time elapsed between the ambr rounds was decreased to a few days. Product quality was confirmed by binding assay and cell-based potency assay.

Figure 1. Titer optimization of a fed-batch process for a CHO-S cell line in a micro-bioreactor system

Product Quality Optimization

Product quality and characterization are critical aspects to consider, especially for the development of biosimilars. Figure 2 illustrates a case study where the goal was to optimize product potency by altering the glycosylation profile. One round in the microbioreactor system was performed which included two different basal media, two different additives in several concentrations, and three different pH set points.

Figure 2A displays how the titer varied among the microbioreactors. Harvests from 17 bioreactors were partially purified by protein A and analyzed in a cell-based assay. The glycosylation profile was also analyzed (data not shown). The cell-based potency assay revealed that bioreactor numbers 2.3 and 2.4 had the highest potency.

However, the titer in bioreactor 2.3 was less than half of the titer in bioreactor 2.4 and hence the conditions in 2.4 were more favorable for both high potency and high titer. Thanks to the ability of testing 24 process conditions in one experiment and to utilizing in-house bioassay capability, it was possible to identify the basal media, pH, and additive most beneficial for high potency in just a few weeks.

Figure 2. Process optimization of a fed-batch process in the microbioreactor system to find the best setting for highest potency: (A) titer, (B) potency in cell-based assay


The cell culture process should give a consistent product and titer at various scales. During development it is beneficial to understand which process parameters are critical for the product as early as possible. This facilitates scaleup of the process and to making any adjustments that might be necessary to the setup of the production facility.

Including a microbioreactor system instead of shake flasks during cell-line development provides more reliable data and is representative for further scaleup of the bioreactor process.

Figure 3 shows scaleup of three clones from ambr to 5 L scale. The best performing clone was further scaled up to 250 L for production of clinical material. Cell growth performance was comparable between the different scales (Figure 3A). The titer obtained in 15 mL scale gave a meaningful prediction of what to expect when scaling up to 5 L or 250 L scale (Figure 3B).

A platform strategy with short timelines requires reliable expression system, high-throughput technology and short lead times for protein analysis. The implementation of a microbioreactor system enables more efficient cell-line and process development. Several clones also can be taken further for process development in bioreactor conditions without prolonging the development program.

More data can be generated in a shorter time, and the data is obtained from a bioreactor system, which gives information about the bioreactor performance of the different clones early in the program. The process performance in the microbioreactor system was comparable to 250 L scale, which is useful for a platform approach where an early estimation of process performance in production of clinical material is advantageous.

If the project requires optimization of titer or product potency the microbioreactor system has proven to be a useful system, enabling you to test many different parameters in a shorter time.

Figure 3. Scaleup of three clones (A, B, C) from 15 mL in a microbioreactor system to 5 L in a glass bioreactor and 250 L in a single-use bioreactor.

Ingrid Lange is scientific lead, cell culture services, Kristina Lae works as a scientist, cell culture services, and Ulrica Skoging-Nyberg, Ph.D., serves as service delivery manager, cell culture services at Cobra Biologics. For more information contact Lara Jupp at Lara.Jupp@cobrabio.com.