System Automates Cell Culture Optimization

A new combined bioreactor and cell culture analyzer automates the process of optimizing cell culture conditions during cell and biopharmaceutical development and production. The system launched in June 2017, with a second system based on a smaller bioreactor launching later this year, says Matt McRae, biotechnology sales product line manager at Nova Biomedical, which makes the culture analyzer.

The combined systems point toward greater integration of growth and analysis tools as cell therapy begins to scale up for larger clinical tests and commercialization. It also shows an industry grappling with ways to quickly add automation to bioreactors. Patients receiving immunotherapy for cancer, autoimmune diseases, or neurological diseases get doses of antibodies that bind to diseased cells. This labels them so that the body can find and attack them.

Pharmaceutical companies produce those antibodies using mammalian cells genetically engineered to synthesize a desired protein. Typically, those cells are Chinese hamster ovary cells. It is easy to assume that the same cell line engineered to produce different proteins would require similar growing conditions. Yet producing a new protein changes the cells’ response.

“Every cell line requires a level of optimization,” McRae says.

During that optimization, scientists tailor nutrient and oxygen delivery, pH, and temperature to speed cell growth. These factors also help them maximize the amount of protein produced without sacrificing quality or consistency. The engineered protein needs to be properly folded, assembled, and decorated with molecules that enable the immune system to target it. All this must be done consistently, from doses delivered to the first clinical trials to batches released if a drug becomes commercially available.

But before cells can move to production-scale systems for larger clinical tests or commercialization, scientists must test a variety of growing conditions using 10 to 15 mL cultures to see how those cells would grow in larger bioreactors. One of them is the ambr 15 bioreactor, made by Sartorius Stedim Biotech (SSB). It is an automated microscale bioreactor designed to model upstream processes. The tabletop instrument handles 24 or 48 cell cultures, each with ports to add ingredients or remove samples for manual analysis.

The challenge for the ambr 15 and other bioreactors is finding a way to monitor growing conditions closely during cell growth. Removing and testing too many samples of nutrient broth from the cultures could affect cell growth and behavior. That means scientists using the ambr system often had limited data about those various culture experiments, McRae says.

Enter BioProfile Flex2, an instrument for automated analysis of cell cultures from research to commercial scales from Nova Biomedical for automated analysis of cell cultures from small to manufacturing scale. The system requires only one 265 µL sample to perform 16 tests in about 4.5 minutes, combining the work of many instruments into a single tabletop machine. The tests include monitoring pH, detecting cell density and viability, and measuring levels of nutrients (glucose, lactate) and toxic waste products (ammonium, carbon dioxide) produced as the cells grow.

At just over one-quarter of a milliliter, the sample volume needed by the Flex2 is too small to affect culture conditions in ambr bioreactors. On the other hand, they had to ensure that the fluid dynamics of samples traveling from the bioreactors through tubes to the sensor modules delivered the cells intact.

For example, fluids running through small tubes are dominated by surface interactions that can generate unwanted forces. During that trip, shear forces could lyse cells. Reduce the speed of the draw and cells could settle on the bottom of the lines. Cells that linger too long in a tube could consume all the oxygen in the lines and start producing carbon dioxide, leading to inaccurate measurements.

As a result, researchers needed to confirm that information provided by the automated cell culture analyzer was as accurate as a manual sample analyzed in a separate test.

“We worked really diligently to ensure that customers with an integrated system can be confident in the fact that the automated samples are going to give the same results as they would if the sample was taken manually and analyzed on the analyzer,” he says.

The combined system is designed for someone to set up and leave while it runs for a week. To do this, it must rely on automated feedback loops that send data from cell culture analysis to the bioreactor controls, which automatically adjust culture parameters, perhaps delivering more glucose or increasing temperature to encourage cell growth. This involved finding ways for the SSB and Nova system to communicate with one another without drastically overhauling their basic software, McRae says.

Finally, the scientists and engineers combining the systems worked to anticipate problems that could arise during a long, unattended run. They introduced multiple checks and balances to allow the system to adjust itself and continue operating independently. For example, the system generates a notification if it is set to run for five days but the Flex2 only has enough reagents for three days of analysis.

After building a prototype of the combined system, the company sent it to a lab at Massachusetts Institute of Technology and allowed students to put it through its paces. Then they made it commercially available in September 2017.

“There’s not been any other technology that’s been able to accomplish what we can,” McRae says. The next version of the combined system connects the Flex2 to the ambr250, a system of 12 or 24 bioreactors containing 100-250 mL of culture media.

Melissae Fellet is a science writer based in Missoula, MT.

Read more about Cell Therapy on AABME.org.