September 1, 2008 (Vol. 28, No. 15)
Successful biomanufacturing requires good technologies for assessing the performance of cell cultures.
Innovatis, a spin-off from the University of Applied Sciences, Bielefeld, Germany, develops, manufactures, and markets a range of instruments for automated cell counting and analysis.
According to the company, its Cedex was the first cell analyzer to automate the standard Trypan Blue method of cell counting. Cedex HiRes, the latest version, provides detailed images of a cell culture sample.
The Cedex HiRes image analysis system channels the cell suspension, which has been stained automatically, through a special flow cell that is equivalent to a standard hemacytometer. It then detects and counts all viable and dead cells, as well as differentiates them from debris and protein clumps. The 2-D image information of each individual cell is used to quantify further parameters like cell diameter or cell shape, which can be used to further determine the status of the culture.
Another innovatis product is Cellavista, which allows fully automated cellular analysis in microplates. It combines brightfield imaging with fluorescence capabilities and is applicable to many aspects of cell line development, cellular research, and drug discovery. The integrated analysis software’s flexibility allows users to focus on colonies, individual cells, or nuclei.
Innovatis hosted a “Seminar Week” in the U.K. recently to discuss key issues in cell counting and cell monitoring. Opening the meeting, Colin McGuckin, Ph.D., professor of regenerative medicine at Newcastle University, said that stem cells in cord blood could now treat more than 85 different conditions including type 1 diabetes.
Dr. McGuckin’s group processes cord blood under GMP conditions and has found cells with embryonic cell-like characteristics. So far, it has been able to generate more than 20 different types of tissue cells from cord blood, but the applicability of these is not yet certain. Dr. McGuckin believes the hepatocytes and biliary eplithelial cells generated from this program can be used to better predict toxicity.
The team is making 3-D “mini-livers” from these cells, using a specialized bioreactor derived from space research that relies on microgravity. “The best way to bring forward these cells is to use them in developing new drugs,” he added. “Stem cells are far from some clinical applications, closer for others—but they are already useful for pharma.”
Cell concentration is an important measurement, and there are many different methods of analysis, commented Professor Frank Gudermann of the University of Applied Sciences, singling out membrane integrity and functional and morphological assays.
Trypan Blue dye, which is absorbed by dead cells, is the basis of an important cell-counting technique because it allows cells to be stained selectively. There is an element of subjectivity involved, though, and it has been shown that different types of hemocytometers give different results. The approach is well accepted by the FDA, and it has been the only way of getting a viable cell count for a long time.
There are now ways of automating Trypan Blue cell counting, however, that involve image recognition. “An advantage of image recognition is that it gives more information than just the numbers,” he said. “Automation also gives more precision and accuracy.”
Monitoring Baculovirus Infection
Laura Sander, application specialist at Innovatis, described a new way of monitoring baculovirus infection of insect cells based on experiments she carried out with researchers at AstraZeneca. “Insect cells are increasingly being used for the production of recombinant proteins because they are cheaper and easier to culture than mammalian cells. They may have increased protein expression levels, and they can also carry out the post-translational modifications that yeast and bacteria cannot.” Sf9 and Hi-5 are the main cell lines used, with the latter becoming more important.
Although insect cells can be an extremely efficient way of producing protein, there are some critical issues that currently limit their use. One is proteolysis of the protein when it is produced at a late stage of the baculovirus infection—just when there is an increase in protease production too. This raises the question of when best to harvest the desired protein. Next, baculovirus is unstable. If it is stored between -20 and +4°C it loses its infectivity. Carrying out repeated plaque assays to assess infectivity is tedious. Sander explained that it is hard to know how well a particular baculovirus infection has worked because the pattern of infection may change on scale-up.
Insect cells swell on infection, and the scientists used the resulting change in diameter to monitor the production process. “There is a huge change in the morphology of the cells,” she continued. The researchers identified a relationship between the course of the infection and the pattern of morphological change. “The change in diameter is an easy parameter to check and there are many tools on the market for doing so.”
In further experiments, the group used various combinations of the baculovirus Autographa californica nuclear polyhedrosis virus expressing green fluorescent protein and Sf9 cells with the Cedex HiRes automated cell counter to determine cell counts and cell diameter. “We found a clear predictive relationship between peak average diameter and peak protein production that tells the best time to harvest the protein,” which turns out to be a day or two after peak cell diameter.
Repeating these experiments with Hi-5 gave different results. The highest productivity was found with the least amount of virus. “This was an unexpected connection,” she concluded. But there was still the same relationship as was found with Sf9 between peak cell diameter and peak production.
“This was a very consistent predictive element for both cell lines at all concentrations,” Sander added. The implication is that monitoring cell diameter can improve the predictability and productivity of an insect cell production process.
Path to Manufacturing
Richard Dennett, Ph.D., head of consultancy services at Eden Biodesign, talked about the cell as a focal point on the critical path to manufacture Phase I supplies. “Cell lines and expression systems are the crown jewels of a protein-production process. They are the core element. The correct choice of cell line is paramount for production and regulatory compliance, but this importance is often not realized until later on in clinical development.”
Many clients often have problems with their cell lines because they were acquired by accident or as a gift, so the in vitro cell age of the line is not known and the cell identity not necessarily certain. “It is best to establish what you want the cell line to do from the start—for example, the clinical indication product type and amount of productivity,” Dr. Dennett said. He went on to describe the ideal cell line, which would have a trace history with quality records from its origin.
European Collection of Cell Cultures and American Type Culture Collection are important sources of cell lines in this respect. Cells should be free of adventitious agents such as TSEs and mycoplasma, with the use of animal component-free media being especially important. Knowledge of the genetic construction of a cell line is also significant, as is the surrounding IP.
Performance characteristics should include robustness, ease of culture, known growth characteristics, and genetic and phenotypic stability. And finally, the ideal cell line should have the elements required for regulatory compliance including safety (including knowledge of tumorgenicity potential), and stability.
There is a balance on the critical path to Phase I between getting there as soon as possible and developing a process that will be both scalable and capable of meeting regulatory requirements, Dr. Dennett added. Critical path planning must therefore be modular and objective. The cell-expression system is central, so the process needs to be mapped onto the cells, remembering that every cell line is different. Medium development is intrinsic to cell performance. For Phase I, one should do basic, but meaningful, optimization, Dr. Dennett said. Off-line monitoring is fine, and the growth curve must be understood for cell expansion.
Eden Biodesign has been working with Strathclyde University on the application of near infra-red (NIR) for monitoring—with a probe in the fermentor. The data readout is being statistically modeled for use in the determination of predictive scale-up, thereby reducing development cycle time. NIR has the advantage of being fast, involving no sample preparation. It also scores by being both sensitive and precise.
Finally, James Mills, Ph.D., director of operations, Xenova Biomanufacturing, discussed the application of design of experiments (DoE) in mammalian cell culture process development. DoE allows for the variation of multiple parameters in an experimental setup.
“You can find effects that you would perhaps miss using traditional experimental design,” he explained. DoE has been used widely in industry but not much, until now, in biotech. “It has great potential for helping us move a product forward more quickly.” Xenova is using DoE to carry out manufacturing-relevant work, and Dr. Mills gave several examples of its application where yields had been optimized and production times cut. “It is important to gather as much quality data as you can upfront and design of experiments can be used for this.”