Umetrics Suite Blog

Design of Experiments enables the optimization of transfection efficiency in line with QbD principles

November 29, 2018

Pressure to cut development costs and lower regulatory barriers while assuring product quality has stimulated the pharmaceutical industry to apply Quality by Design (QbD) to manage risk and gain process and product understanding. As a result, QbD is being widely promoted by regulatory authorities such as the Food and Drug Administration, and the International Conference on Harmonization.

bioprocess Transfection efficiency DOE ambr15

Bioprocess development using DOE leads the way for regulatory approval, while saving time and money.

Design of Experiments (DOE) has become a key tool in implementing QbD in bioprocess development and applying this approach to optimize bioprocesses early on leads to major benefits in product performance together with improved economics based on limited resources. Applying DOE was key for a team at SciLife Lab, Stockholm, Sweden in their optimization of Transient Gene Expression ready for scale up.

Improving the yield of transiently expressed proteins ready for scale up

Science for Life Laboratory, SciLifeLab, is a national center for molecular biosciences with focus on health and environmental research. The Protein Expression and Characterization facility at SciLifeLab provides support to research and development, including the expression, purification and characterization of biopharmaceutical candidates for proof of concept and mechanism-of-action studies. Their main objective in this study was to find the conditions needed to maximize the transfection efficiency of a cell line and deliver higher protein titers when transfected with a plasmid of interest.

Application of DOE to expression systems using microscale bioreactors

The team at the Protein Expression and Characterization Facility needed to determine the optimal settings for transient gene expression using the FectoPRO® Transfection reagent (Polyplus transfection) together with Expi293F™ human cells (ThermoFisher Scientific), derived from the HEK 293 cell line. To do this they used ambr® 15 (Sartorius Stedim), a microscale bioreactor system that replicates classical laboratory scale bioreactors.

The ambr 15 system provides parallel processing and control of 24 or 48 bioreactor experiments, which simplifies the application of QbD principles enabling large multifactorial experimental designs (DOE) in real time. The bioreactors are designed to support scalability by functionally translating into systems of 2000 L or larger.

Experimental design

Earlier DOE screening experiments had pointed to two key factors for optimizing the transfection efficiency: DNA amount, and the ratio FectoPRO reagent:DNA (Figure 1). The transfection efficiency was measured using two responses, % transfected cells and % cell viability, with the goal of achieving >60% transfected cells and >80% viability (Figure 2).

MODDE 12 was used to design 12 experiments that included 9 combinations and 3 duplicated points (low/low, mid/mid, high/high).

dna amount fectoPRODNA

Figure 1. The factors investigated were DNA amount and FectoPRO:DNA ratio.

transfected cells viabilitypng

Figure 2. The responses measured were Transfected cells (%) and Viability (%), with the aim of achieving transfected cells > 60% and Viability > 80%.

experimental design 12 experiments

Figure 3. The experimental design included 12 experiments, with 9 combinations and 3 duplicated points (low/low, mid/mid, high/high). Run number 6 was omitted due to a technical error.

The influence of the factors on the responses

Statistical analysis showed that the experiments provided very good data and reliable regression models were obtained. You can find out more about the statistical basis for judging the quality of data and models here: What tools make DOE data analysis faster and more accurate?

The response contour plot for % transfected cells showed a higher response at higher DNA amount and lower FectoPRO:DNA, while the %Viability was highest in the left region of the response contour plot (Figure 4).

response contour plots

Figure 4. Response contour plots.

These results could be combined in a Sweet Spot plot (Figure 5a) that indicated which combinations of the two factors fulfilled the minimum acceptance criteria of %viability and %transfected cells.

To home in on the must robust combination of the two factors, Monte Carlo simulations were used to generate a Design Space (Figure 5b). The most robust combination for obtaining stable, high amounts of transfected cells and high viability was a DNA amount of > 1.1 and a FectoPRO:DNA ratio of approximately 0.8.

The Design Space indicated that with these factor settings the risk of failure to comply with the response specifications was less than 2.5%, which gave the team the confidence to move forward, establish the protocol, and apply it to lead molecules for screening.

sweetspot plot

Figure 5. (a) Sweetspot plot shows possible region where acceptance criteria are fulfilled. (b) Design space plot shows low risk region (probability of failure <2.5%).

Finding the Design Space is key to QbD in bioprocess optimization

The limited resources available for investigating and optimizing bioprocesses have meant that DOE is one of our primary tools to achieve a rational and cost-effective approach to practical experimentation.

DOE is the backbone for efficient implementation of QbD, with the Design Space defining the final specifications for a region where all specifications are fulfilled. Finding this region is critical to success. Applying DOE therefore not only saves time and effort, it also enables the implementation of QbD to pave the way to regulatory approval.

Want to know more?

Find out more about creating an optimal design space in this free recorded webinar. Register to watch the previously recorded webinar:

Design of Experiments (DOE) and Quality by design (QBD) in Bioassay Optimization

Watch webinar


You may also like this blog:

Creating a design of experiments study to predict formula robustness

Or this one:

Using data analytics to optimize design space and setpoint conditions for bioreactors



Sartorius Stedim Data Analytics are grateful to Alexander Korpys (Research Scientist) and Anders Olsson (Head of Facility) at the Protein Expression and Characterization facility, SciLifeLab for sharing their data with us.




Topics: Design of Experiments (DOE), Quality by Design (QbD), Process Validation

Lennart Eriksson

Written by Lennart Eriksson

Sr Lecturer and Principal Data Scientist 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