With new monoclonal antibody (mAb) drugs such as Kadcyla costing over €100,000 per year, the price per patient for these therapies is considerably more than small molecule drug treatments. One reason for this is that these therapies cost more to manufacture than small molecules.¹ Hence, using mini scale bioreactor models to determine optimum scale-up bioprocess parameters in fed-batch culture could speed up process development and reduce manufacturing costs.

Fed-batch culture, in which a base medium supports initial cell culture and a feed medium is added to prevent nutrient depletion is the commonly adopted biologicals manufacturing method used in many biopharmaceutical companies. Fed-batch culture can generate gram/liter amounts of mAb-based therapeutics² provided the optimum formulation of culture media and feeds, as well as the correct bioprocess parameters are used. Traditionally, shake flasks and benchtop bioreactors have been used to define optimal media, feed and bioprocessing conditions in fed-batch culture.

However, using benchtop bioreactors is resource-intensive and due to their scale, expense and limited throughput, researchers are restricted in the number that can be run in parallel. Therefore, the final bioprocessing parameters established can sometimes perform sub-optimally upon scale-up, adversely affecting outcomes such as yield and product quality. If a larger number of runs could be performed during process development under conditions representative of the bioreactor environment, then it could help to determine optimum process parameters for use in manufacturing and thus save thousands of Euros in production costs.
This need to conduct multiple experiments under benchtop bioreactor conditions has resulted in the development of miniaturized high-throughput culture technologies. The disadvantage with many of these approaches is that they have many complex connections, thus are labor intensive and time consuming to use due to set-up and decontamination regimes between runs.

Stirred, sparged, single-use mini bioreactors
With the need to replicate bioreactor parameters in a high throughput methodology in mind, ambr250 was introduced by TAP Biosystems, now part of the Sartorius Stedim Biotech Group, in 2013. This stirred, sparged mini bioreactor system is based on the established ambr15 microbioreactor technology and there is published data to show that Chinese Hamster Ovary (CHO) cells cultured in this microbioreactor match those grown in benchtop vessels, making this technology suitable for high-throughput development.^4,5

The mini bioreactor system has three components: disposable single-use bioreactor, workstation and software. Each bioreactor has a 100 to 250 mL working volume and an easy connect bioreactor integrated to the gas, liquid and sensor connections to allows fast turnaround between experiments with a bioreactor system being set up around three times faster than a single benchtop bioreactor (Table 1). Since the mini bioreactors are pre-sterilized and single-use this means up to 24 bioreactors can be ready for use in three hours compared to 1-2 days it takes to clean, assemble and sterilize the same number of stirred benchtop bioreactors. 

The key to ambr250 is that its mini bioreactors are designed using standard bench top bioreactor configurations, which allows for seamless scale up. Each mini bioreactor is connected via the workstation to liquid handling automation that facilitates aseptic culture set-up, inoculation and feeding, as well as sampling into vessels such as 24/96 well plates and ViCell cups. These operating protocols can be configured for individual control of up to 24 reactors operated in parallel. In each single-use mini bioreactor temperature, impeller speed, pH and DO are individually controlled allowing users to rapidly analyze and determine which of the multiple process parameters they are assessing produces optimum cell growth and product titre of their cells. The automated sampling and feeding allows for more intensive data capture and more complex feeding strategies than is practical using a benchtop bioreactor.

The system is supported by software in which the user sets up a protocol of steps that defines operating parameters such as DO/pH set points, stirrer speed and temperature as well as tasks such as inoculation, feeding and sampling. Protocol timings and details can be edited at any time if steps need to be added, modified or deleted, including while an experiment is running. Real-time data such as gas flow rates, volumes, pH/DO values are logged continuously during a run and external values such as cell counts, metabolites and product titres can also be imported into the software. Data can be used to generate graphs within the software or exported for production of graphs or tabular results in spreadsheet software.

Case study
To determine whether the ambr250 bioreactor could mimic fed-batch cell culture in a 7 L bioreactor, a CHO-S cell line expressing an Fc fusion protein was cultured in duplicate for 12 or 13 days in either 3L shake flasks, 7L benchtop bioreactors or in the single-use mini bioreactors and the cell growth, viability, metabolite concentrations, titre and protein quality profiles were compared. 
The CHO-S cell line showed comparable growth and viability profiles for the mini bioreactors and 7L bioreactors (Figure 1). 

The CHO-S clone cultured in the benchtop bioreactors and the single-use mini bioreactors had comparable harvest titre yields (380-425 mg/L); while the shake flask titre was slightly lower (300 mg/L) (Figure 2). The growth, viability, and titre data all suggest that the single-use mini bioreactors provide an easier to use, comparable model to benchtop bioreactors for replicating bioprocessing conditions.

In terms of protein quality, the CHO-S clones produced similar amounts of percentage monomers when cultured in single-use mini bioreactors, benchtop bioreactors and shake flasks (Table 2).

The amount of monomeric protein produced by a CHO-S cell line expressing an Fc fusion protein cultured in ambr250 mini bioreactors (AMBR1 and AMBR2), benchtop bioreactors (BX1 and BX2) or shake flasks (SF1 and SF2).

Conclusion
Stirred, sparged, single-use mini bioreactor technology can provide as accurate a prediction of growth, viability, titre profiles and MAb quality as a benchtop bioreactor with CHO clones. Setting up and running benchtop vessels is manually intensive, whereas since the mini bioreactor and its analysis components are single-use, each one requires minimal set-up and cleaning time. This means the mini bioreactor offers higher throughput than traditional bioreactors. Thus, process optimization can be performed more quickly and efficiently with the mini bioreactor, increasing the number of process parameters that can be evaluated. 

In summary, the ambr250 mini bioreactor system has the capability to be a high-throughput tool for developing optimum media and feed strategies, as well as fed-batch bioprocess parameters, thus reducing costs when scaling-up therapeutic MAb manufacturing processes. 

References

1. Biot J. et al.: Med Sci (Paris). 25 (12):1177-82 (2009)
2. Li F. et al. MAbs. 2(5): 466–477 (2010)
3. Ngibuini M. GEN 33 (1): 25-26 (2013)
4. Nienow A.W. et al.: Biochem. Eng. J. 76: 25–36 (2013)
5. Moses S. et al.: Adv Biosci Biotechnol. 3: 918-927 (2012)