Here, we present a detailed procedure to run a Design of Experiment in an automated micro-bioreactor followed by cell harvest and protein quantification using a Protein A column.
Optimization of bioprocesses to increase the yield of desired products is of importance in the biopharmaceutical industry. This can be achieved by strain selection and by developing bioprocess parameters. Shake flasks have been used for this purpose. They, however, lack the capability to control the process parameters such as pH and dissolved oxygen (DO). This limitation can be overcome with the help of an automated micro-bioreactor. These bioreactors mimic cultivation at a larger scale. One of the major advantages of this system is the integration of the Design of Experiment (DOE) in the software. This integration enables establishing a design where multiple process parameters can be varied simultaneously. The critical process parameters and optimum bioprocess conditions can be analyzed within the software. The focus of the work presented here is to introduce the user to the steps involved in process design in the software and incorporation of the DOE within the cultivation run.
The global biopharmaceutical market was worth more than US $250 billion in 2018 and has been continuously expanding1. Pharmaceutical companies are moving away from producing small molecular drugs to biotechnologically produced therapeutics such as recombinant proteins. These alone are responsible for a revenue of more than $150 billion1. Mammalian cells are now extensively used for the production of these pharmaceutical recombinant proteins. In the current period, among the 68 approved products produced by mammalian cells, 57 are produced by Chinese Hamster ovary cells (CHO)2. CHO cells are specifically used for the production of recombinant proteins that require post-translational modifications. These cells are preferred as they grow in a suspension and thereby enable reproducible results in a serum free chemically defined medium3,4. The other advantage of using CHO cells is that the glycan structure of the product resembles that of the human monoclonal antibody (mAb) and results in higher recombinant protein yield and specific productivity due to gene amplification5.
The yield of recombinant CHO (rCHO) cell culture has increased by a hundred-fold in the past two decades. This improvement is attributed to the optimization of the process parameters, feeding strategy and development of serum free chemically defined medium6. With the increase in requirements of the pharmaceutical products, the pressure increases on cost and time efficiency for the development of the production process7. To reduce the pressure while assuring product quality has redirected the focus of the pharmaceutical industry on Quality by Design (QbD). QbD is used to understand the product production as well as the process. A vital tool used in the ObD is the Design of Experiment (DOE). It helps increase understanding of the process by revealing the relationship between various input variables and resulting output data. Applying the DOE approach to optimize the bioprocess is beneficial during the early stages of the project in assimilating the process conditions and increasing the titer quantity and quality. This approach is beneficial when compared to the old-fashioned strategy: one-factor-at-a-time (OFAT). The statistical approaches to DOE using Classical, Shainin or Taguchi are far superior to the OFAT8.
The process and media optimization can be performed in shake flasks. The flasks are relatively inexpensive. However, it is not possible to control parameters such as temperature, pH and dissolved oxygen (DO). To overcome these drawbacks, multiuse bench-top bioreactors ranging from working volume of 0.5 L to 5 L can be used. The reactors provide an extensive on-line monitoring and process control. However, the use of the multiuse bioreactor is time and labor intensive. In order to overcome these disadvantages, a novel single-use bioreactor that combines the comprehensive process of monitoring the bench-top bioreactor and easy handling of the shake flask is used. The high throughput screening system and single-use technology have contributed to enhance the efficiency of process performance and development9.
In this article, the guidelines to load the recipe in the automated micro-bioreactor (AMBR) software are listed. The influence of different stirrer speeds and pH on the viable cell concentration (VCC) and titer is studied during the course of this experiment. The experimental result and analysis are carried out with design of experiment software MODDE 12. The product analytics are carried out in a high pressure liquid chromatography (HPLC) system with a Protein A column. It is based on the principle that the Fc region of the mAb binds to protein A with high affinity10,11. With this method, it is possible to identify and quantify the mAb. The quantification is carried out over the measured elution peak areas at 280 nm.
1. Preculture Procedure
NOTE: Recombinant CHO DG44 cells with a viable cell concentration of 1 x 107 cells/mL are used for this protocol.
2. Main Cultivation
3. Writing the Recipe in the Automated Micro-bioreactor Software
NOTE: There are two ways of writing a recipe in the AMBR cell culture software: it is created either by using a wizard or by adding each step manually. For the purpose of this protocol, steps using the wizard are shown.
4. Execution of Cultivation in the Automated Micro-bioreactor
NOTE: The following steps are executed by the user with the help of the protocol written in the aforementioned software. The steps are carried out by the user unless mentioned otherwise.
5. Measuring mAb Concentration
Figure 1: Protein A chromatogram, representing the different phases during a single run. Please click here to view a larger version of this figure.
An overview of the cultivation performed in this study is presented in Figure 2.
Figure 2: Schematic representation of the experimental conditions to test pH and stirrer speed profiles in the culture stations. The figure also represents the correct layout to place the vessels. Please click here to view a larger version of this figure.
The cell growth in the automated micro-bioreactors is comparable to the multi-use bioreactors. This is depicted in Figure 3. The cell concentration from the three different scales is compared and it is observed that the 15 mL automated micro-bioreactor mimics the 2 L glass bioreactor. The results from the shake flask are also compared to exhibit the benefit of the AMBR.
Figure 3: Comparison of the viable cell concentration at different scales. 15 mL micro-bioreactor, 150 mL shake flask, and 2 L multi-use glass bioreactor. Please click here to view a larger version of this figure.
The influence of different stirrer speed and pH is studied in the automated micro-bioreactors. Viable cell concentration (VCC) is one of the vital parameters to compare the cultivation. Figure 4 represents the comparison of the VCC and monoclonal antibody concentration in the different micro-bioreactors. Figure 5 represents the response contour plot of the two responses considered for the comparison, namely, VCC and mAb concentration. The values are comparable in the vessels with the same pH and different stirrer speed, indicating that the stirrer speed selected for this process has no significant influence on the process output. For future cultivations, the stirrer speed of 1050 rpm would be used in order to avoid foaming.
The pH, however, has a conspicuous impact on the process output data. The negative influence of pH 6.9 on the VCC can be observed in the Figure 4A. The growth of the cells improved significantly under the culture at pH 7.3 compared to pH 7.1. In Figure 5B, the monoclonal antibody concentration of the different cultivations is compared. The production of the mAb is slower in the vessels maintained at pH 7.3; however, the final product concentration is comparable to the vessel maintained at pH 7.1.
Figure 4: Result of the DOE experiment. (A) Viable cell concentration profile of the cultivation run to study the influence of pH and stirrer speed (B) Monoclonal antibody concentration profile over the fed-batch process in the respective cultivation conditions. Please click here to view a larger version of this figure.
Figure 5: Response contour plot indicating the influence of pH and stirrer speed on the maximum viable cell concentration and monoclonal antibody, respectively. Please click here to view a larger version of this figure.
One of the advantages of using the AMBR is the continuous monitoring and control of the cultivation. The monitoring of the pH can be observed in Figure 6. The pH is controlled at the set point using CO2. The bolus feeding from Day 3 is responsible for the spike in the values.
Figure 6: pH monitoring in the automated micro bioreactors for the cultivation, with set points of pH 6.9, 7.1 and 7.3 and a stirrer speed of 1050 rpm. Please click here to view a larger version of this figure.
The process parameters used for the future tests was narrowed down to a stirrer speed of 1050 rpm and pH of 7.1.
Optimization of the process to increase the yield is of crucial importance in the biopharmaceutical industry. Shake flasks could potentially be used for the screening of the strain; however, the monitoring of the process parameters such as pH and DO are unavailable in the flasks. The micro-bioreactors have an advantage as they allow continuous monitoring and control of the process. These control loops in the micro-bioreactor also provide a condition similar to those at larger scale and thus, deliver results that are comparable to the larger scale bioreactors. Another advantage of the micro-bioreactor is the wide range of conditions that can be tested within the same time frame and at a lower cost when compared to bench top bioreactors in terms of time and labor12,13 .The smaller size is also advantageous in terms of lower running costs due to the reduced amount of substrate and reduced space requirement for parallel operations; however, the cost of the vessels must be considered as it may be more expensive to run a the cultivation in the single use micro-bioreactors when compared to the bench top multiuse bioreactor.
While there are several advantages to the system, there are a few considerable disadvantages to using the micro-bioreactors. Changing the pH and the DO for each of the vessel within one culture station is possible; however, the stirrer speed and the temperature cannot be changed for an individual vessel. This increases the number of experiments carried out to establish the optimum stirrer speed and temperature. The online measurements are restricted to pH and DO. Real time process monitoring of the critical process parameters (CPP), such as temperature, pH, DO, would be beneficial to avoid the delay between sampling and analysis. It has the potential to be a useful tool for optimizing productivity14. Development in the field of spectroscopic measurements in micro-bioreactors may prove to be beneficial in the future models.
The advantages outweigh the disadvantages of using the automated micro-bioreactor. However, there are a few criteria to be considered when operating these systems. First and foremost is the designing of the script to successfully run the cultivation. It is of vital importance that the script written is inclusive of all the crucial steps such as definition of the plates along with the correct mapping of the plate and vessels. All the parameters must be checked before starting the experiment. Ensure no process steps are scheduled to occur at the same time. This will result in the workstation choosing one step over the other.
Another important criterion to be focused upon is the use of the clamp plates. The plates are autoclaved before every cultivation; this could lead to damage of the O-rings. Visually check the O-rings before autoclaving. Faulty O-rings could result in unexpected variation in the DO and pH control. The other factor for faulty readings could be a loose connection between the clamp plate and the vessels. Ensure the clamp plate and the stirrer plates are secured tightly to the stirring assembly. Use the tightening tool provided with the system.
During the cultivation, it is crucial to visually monitor the foam. A minimum of 20 µL of antifoam is required at the start of the cultivation. Foaming will result in liquid being present in the trough of the clamp plate. Excess of the foam in the clamp plate will lead to overflow, leading to the foam being collected at the bottom of the culture station, obscuring the pH and DO spots to be read by the sensor.
Another important factor to focus upon is the use of the DOE software. The integrated DOE software enables quick designing of the process incorporating the relevant conditions. The visualization of the vessels being used ensures that no factor has been overlooked. One of the advantages of DOE analysis is that the experiments can be configured via work packets, configuring each bioreactor with defined bioprocessing parameters. All the data generated is analyzed using MODDE. The software manages the large amount of data generated during the course of the experiment. A generalized subset design setup generates a sequence of reduced design set, thereby, solving the problem with multivariate calibration.
A detailed procedure to run a Design of Experiment in an automated micro-bioreactor is demonstrated in this article. The protocol was designed to focus on the feed batch process. The DOE software aided in establishing the optimum process parameters to increase the biomass and the titer concentration. The cultivation data were also compared to an experiment conducted in a shake flask and a 2 L bench-top bioreactor. The results demonstrate the reproducibility and scalability of the process. The goal of the protocol was to demonstrate the use of automated micro-bioreactors in a feed batch process and to analyze the bioprocessing parameters using the DOE software. It can be concluded that the automated micro-bioreactors are useful for the process development and these can be extended to a semiperfusion system15.
The authors have nothing to disclose.
The authors would like to thank the Bundesministerium für Bildung und Forschung (BMBF), the Federal Ministry of Education and Research, Germany, and the BioProcessing team of Sartorius Stedim Biotech GmbH, Germany, for their support.
1 mL disposable pipette tips, sterilized | Sartorius Stedim Biotech GmbH | A-0040 | |
200 mM L-glutamine | Corning, Merck | 25-005-CV | |
24 Well deep well plates | Sartorius Stedim Biotech GmbH | A-0038 | |
5 mL disposable pipette tips, sterilized | Sartorius Stedim Biotech GmbH | A-0039 | |
ambr 15 automated microbioreactor system | Sartorius Stedim Biotech GmbH | 001-2804 | |
ambr 15 Cell Culture 24 Disposable Bioreactors – Sparged | Sartorius Stedim Biotech GmbH | 001-1B86 | |
Antifoam C Emulsion | Sigma-Aldrich, Merck | A8011 | |
Bottle Top Sterile filter | Corning, Merck | CLS431474 | 0.1 μm pore size |
CEDEX Detergent (3% Mucosol) | Roche Innovatis AG | 05-650-658-001 | |
Cell counter | Roche Innovatis AG | 05-650-216-001 | CEDEX HiRes |
CHO DG44 cell line | Cellca, Sartorius Stedim Biotech GmbH | ||
CHOKO Feed Media A (FMA) | Sigma-Aldrich, Merck | CR80025 | |
CHOKO Feed Media B (FMB) | Sigma-Aldrich, Merck | CR80026 | |
CHOKO Production Medium | Sigma-Aldrich, Merck | CR80027 | |
CHOKO Stock Culture Meium | Sigma-Aldrich, Merck | CR80028 | |
Chromaster high pressure liquid chromatography system | VWR International | ||
Conical Centrifuge tube | Corning, Merck | SIAL0790 | |
Ethanol | Merck | 1070179026 | |
Glycine | Carl Roth | 56-40-6 | |
HPLC Vials | VWR International | SUPLSU860181 | |
PBS | Sigma-Aldrich,Merck | P4417 | |
Protein A Column | Thermo Fisher Scientific | 1502226 | POROS™ A 1.7 mL |
Sodium chloride | Sigma-Aldrich,Merck | 7647-14-5 | |
Sodium phosphate dibasic anhydrous | Sigma-Aldrich,Merck | 7558-79-4 | |
Trypan Blue | VWR International | VWRVK940 | |
YSI | YSI Inc | 2900D | YSI 2900 Select |