Umetrics Suite Blog

Using DOE With Micro Bioreactor Improves Process Control for Scaling up Cell Cultures

June 7, 2019

A challenge for the regenerative medicine industry is to develop cell culture processes that can be scaled up for high volume production. Finding a better way to scale up commonly used research cells like HEK293T (used for protein expression and the production of recombinant retroviruses or lentiviral vectors) would be beneficial for biologists in many fields of medicine. Dr. Franziska Bollmann, virus scientist at Sartorius Stedim Biotech in Germany, recently conducted two experiments to find out if micro bioreactor systems can help facilitate the transition from the traditional shake flask process to a more improved method optimizing process control.

A HEK293T suspension cell cultivation was optimized using design of experiments (DOE) in the ambr®15 micro bioreactor system

In the experiments, a HEK293T suspension cultivation was optimized using design of experiments (DOE) in the ambr®15 micro bioreactor system — a high throughput, automated bioreactor system from Sartorius. The experiments showed that a transition from shake flask to a scalable micro bioreactor system can be rapidly accomplished with improved process control. Furthermore, the viable cell count increased by almost 20%. Read on for a summary of the experiments.

Watch a recorded webinar with Dr. Bollmann >

Using the ambr®15 for process control and optimization

The Sartorius ambr®15 is a micro bioreactor that mimics the process control in larger scale bioreactors, but in volumes of 10-15 mL. It can be used for a range of different cell lines and applications.

The aim of the experiments was to identify optimal set points of three process parameters – stirring speed, DO and pH values – by using MODDE®, a state-of-the-art DOE software that is connected to the ambr®15 system. The responses monitored were viable cell count (VCC) and viability. The resulting values were compared to an optimized standard shake flask culture as a reference control.

Experiment 1: Finding optimal process parameters

Setting up a DOE in the MODDE software is an easy process. MODDE assists in entering the process parameters and responses and suggests a model and design best suited for the study. In the two experiments, a 2-level full factorial design with three center points was used. The ranges of the process parameters and the resulting bioreactor profiles are illustrated in the figure below.


In the top right graph, each line represents the growth profile of a micro-bioreactor vessel. In the response counter plot below, cell densities are shown in different colors, depending on the process parameters, where high cell concentrations are colored in red. The graph to the left shows that the validity of the DOE model is very good.

With the DOE evaluation in the MODDE software, Dr. F. Bollmann and her colleagues were able to identify optimal set points with a low probability of failure, using a stirring speed of 400 rpm, pH of 7.2 and DO of 50%, as shown in the figure below.

find optimal settings within design space rv

The design space shows the probability of failure, with low values colored in green and high values in red.

Get MODDE free trial >>

Experiment 2: Shake flask comparison

The first experiment indicated that a stirring speed lower than 400 rpm might yield an even better result. To test this hypothesis, a follow-up experiment was conducted, where the cultivation in ambr®15 was compared to an optimized shake flask cultivation.

The second experiment did not show any significant difference in cell growth between 300 and 400 rpm. With this new range of set points for rpm, a pH of 7.3 at 400 rpm stirring speed and a DO of 50% were determined to be optimal for cell growth. At these optimal set points, Dr. F. Bollmann and her colleagues were able to achieve a viable cell count (cells/mL) of 4.01x106 compared to 3.35x106 in the shake flask. The viability of the cells in ambr®15 was 98.4% compared to 97.5% in the shake flask.


The experiments demonstrate that the ambr®15 micro bioreactor system in combination with the MODDE software for DOE can be used for early process development and optimization. The transition from shake flask to the bioreactor system was fast and easy and also led to increased process control.

Get MODDE free trial >>

Want to know more?

Watch a recorded webinar in which Dr. Bollmann gives a detailed presentation of the experiments. (Register to watch the recording). 

Watch the webinar


Topics: Design of Experiments (DOE), MODDE, Medicine/Health

Marie Wensley

Written by Marie Wensley

Marketing Manager at Sartorius Stedim Data Analytics

Search the Blog

    Subscribe to the Blog

    View the:

    Data Analytics Glossary of Terms

    List of Webinars

    Get a free trial