October 15, 2010 (Vol. 30, No. 18)
Angelo DePalma Ph.D. Writer GEN
Myriad of Enabling Strategies Must Be Deployed from Beginning to End of Product Life Cycle
Optimizing cell culture involves numerous strategies, undertaken from the earliest stages of cell-line engineering or selection, through development, and even during full-scale manufacturing.
In addition to higher yield, cell-culture optimization confers the benefit of improving downstream operations. For example, the fewer cells that die, the cleaner the purification stream. “Upstream drives downstream,” comments Scott Deeter, CEO of Invitria, which specializes in media and feed supplements. “If you don’t start with the right material at the front end, it’s difficult to produce high-quality product on the back end.”
Cell death causes the release of proteases, and glycosidases degrade therapeutic protein backbones and glycans, respectively. Deeter notes that the protein A capture step can be made more efficient by improving cell viability and thus reducing nonspecific binding to the column.
Invitria manufactures an array of feeds based on recombinant albumin (the principal serum protein), lactoferrin (a growth factor with anti-apoptosis activity), lysozyme (a cell lysis-inducing agent), and transferrin (iron transport). In addition, the company offers ZapCho, a multicomponent additive for CHO cultures, and Zap-Hybridoma, for hybridoma cells.
An aggressive media development and feed/supplementation program is the best way to ensure optimal productivity during production. Increasingly, bioprocessors are employing these methods at the earliest stages to assist in clone selection under best-case conditions and to assist cells in adapting to the desired culture medium. “Using supplements can shave many weeks off development time,” Deeter explains.
Speed and Versatility
Bioprocesses have historically been tightly controlled whether they occurred in glass, stainless steel, or single-use equipment. Scaling up introduces variables, however, as bench, development, pilot, and production scales differ not only in size, but in the operation of engineering phenomena such as mass transfer and oxygenation levels, and shear stress.
Roman Rodriguez, global product manager for bioreactors at ATMI, makes a good case for retaining the same bioreactor platform or geometry throughout development to minimize the effects of these differences.
ATMI specializes in single-use bioreactors and mixing systems, which, according to Rodriguez, enable customers to retain the bioreactor’s critical physical characteristics at successively larger scales. The process-contact components of mixing systems, acquired through the NewMix® and LevTech® brands, are fully disposable and scalable. The bioreactors provide the usual benefits of reduced cleaning and cleaning validation, as well as rapid changeover.
“Pressures on cell-culture developers are increasing from regulators and due to time-to-market considerations. The disposables approach permits less down-time and saves on cleaning and validation. This is what makes disposables so attractive during development.”
He argues that, unburdened with cleaning and limitations on appropriate cell-culture capacity, process developers can test more process scenarios in parallel than ever before during development and begin a new run immediately after the old run ends, with no worries about cross-contamination. And they can achieve optimized lab-, bench-, pilot-, and production-scale cell cultures—up to 1,000 L—in a familiar, fully characterized system without re-validation.
Disposables can also come to the rescue for processes scaled down in volume to accommodate rising productivity, Rodriguez says. “Everybody wants to do more in smaller volume,” but to achieve that smoothly, processors must fully understand the mixing characteristics of the vessel.
ATMI products are designed for mammalian cell culture, but Rodriguez says they should be compatible with fermentation as well. “That’s our next step. It will require more validation, testing, and perhaps some upgrades.”
Greater Impact Early On
Cell-culture optimization occurs throughout a product’s production life cycle, but options decrease over time. Processors have some leeway during late-stage development, even during clinical manufacturing, but the window of opportunity for enacting substantial changes begin to peter out during Phase II.
“That’s where things start to get locked down, but improvement continues throughout,” says Mindy Goldsborough, Ph.D., R&D director for advanced bioprocessing at BD Biosciences. “Companies are always interested in improving the efficiency and safety of their processes, and will even re-file if these benefits are sufficiently strong.”
BD’s cell-culture products include catalog media for growing CHO and other mAb-producing cells. The company has sold peptones (hydrolysates) for over 50 years and will soon launch chemically defined feeds/supplements.
“Chemically defined products take cell culture to the next logical step beyond animal component-free,” says Dr. Goldsborough. Adoption of animal-free ingredients was undertaken for safety reasons, chemically defined media for consistency.
BD also recently opened a media and supplement manufacturing facility in Miami, which is both animal-derived component-free and antibiotic-free, and controlled for animal-origin component raw materials to the tertiary level. It serves the cell-culture media needs of bioprocessors working with mammalian, microbial, or stem cell lines.
Process analytic technology (PAT), if widely implemented, could become the last link in the cell-culture process improvement chain, and one that operates in real time or nearly so.
Ultimately, processors would like to tweak processes as they proceed through PAT, but that is still a long way off for most biomanufacturers. Despite protestations from large manufacturers, regulators, and industry observers, PAT has been slow in its realization.
Friedrich Srienc, Ph.D., a professor at the University of Minnesota Biotechnology Institute, has developed a technique, based on flow cytometry, that is suitable for a PAT-based global assessment of cell-culture health.
Apoptosis and other cell cycle-indicating phenomena are useful in assessing the vitality and productivity of cultured cells, as well as their suitability for passaging. Investigators can assess culture density, protein-expression levels, and viability for both the culture and individual cells using appropriate off-the-shelf stains and standard flow-cytometry equipment.
Flow cytometry is traditionally thought of as an off-line analysis mode, which slows it down. Add to this the time and effort required to withdraw samples from the bioreactor and stain the cells. Making the process PAT-worthy would require speeding up or automating the individual operations.
Dr. Srienc has achieved that by automating the sampling, staining, and cytometry steps. Now, it is possible to quantify cell and culture parameters in a way that supports decision-making during the culture. “In my opinion you cannot get more detailed information on the properties of cells or a cell culture than with flow cytometry,” he says.
But despite the near-real-time capabilities of this technique, interest from industry has been mixed. “The stumbling block is that the system is not an off-the-shelf instrument that can be easily installed and deployed.” Potential biopharmaceutical customers all have engineers capable of putting such a system together, “but they have other priorities.”
Vaccine Cell Culture: A Special Case?
Optimization efforts differ for cell cultures that produce antibodies and those that make vaccines. MedImmune works on both. Its FluMist® live-virus influenza vaccine is currently manufactured in eggs, but the company is developing a cell culture-based process.
Vaccine-makers are interested in switching from eggs to cells because the former are more labor-intensive. Controlling process variability and product quality are also issues for eggs, whose characteristics change depending on the season.
MedImmune also makes Synagis®, a monoclonal antibody for preventing respiratory syncytial virus.
While antibody production can employ a range of cells, cells that grow viruses are limited to lines that are susceptible to the virus. MedImmune uses MDCK cells for this purpose.
Optimization for mAb manufacturing is an ongoing process, says Ben Machielse, executive vp of operations. “We want to be done with major optimization efforts before Phase III, but we continue to fine-tune the process afterward to assure there are no issues related to scale-up.”
Yield may be the key to optimizing a mAb process, but as Machielse points out, rising productivity creates issues of matching upstream and downstream operations and avoiding bottlenecks. “You don’t want the yield to be too high. You should optimize it to fit the plant, to strive for maximum utilization end to end.”
Optimizing cell cultures for viral vaccines is more straightforward despite the cell-line limitations. Purification steps are fewer for vaccines than for antibodies, and facility design constraints far less daunting. “Because vaccine doses are small to begin with, and FluMist doses less than 1% of the dose for an inactivated vaccine, we don’t need a big plant to produce millions of doses.”
Nature vs. Nurture
The question of nature (cell-line engineering) vs. nurture (media, feed, and culture strategies) inevitably arises during any discussion of cell-culture optimization. “You can do a lot of things in the way of nurturing cells, but you cannot change the fundamental nature of a cell once it’s set,” notes Eileen Skaletsky, Ph.D., CEO of QED Bioscience, a contract cell culture services company.
“If it’s an inherently low producer, there’s only so much you can do to ramp that up because the cell probably lacks the signaling machinery required to pump out lots of product.” Media and feed strategies can help, but turning a poor producer into a high producer is unlikely to succeed.
Despite the homage paid to nature, most experts—Dr. Goldsborough included—believe that nurture has thus far paid the most dividends. “Nurture, including media, supplements, and control over bioreactor conditions, is still the predominant activity today.” Nevertheless, researchers are looking into “nature” as well as into cell processes and signaling, which has the potential to harness all the cell’s activities and recruit them toward production. “So it’s clear the two, nature and nurture, have to work together,” Dr. Goldsborough says.
“Nature and nurture go hand in hand,” adds Machielse. While cells are initially selected based on growth and productivity, “you can do a lot once you have selected them through the right feeding strategy. And with refined media the exercise becomes a lot more manageable.”
Machielse believes that in the future cell phenotype and genotype will help processors predict the type of feeding strategies that are likely to improve productivity. And at some point, those characteristics may be designed in.
Characteristics of a Good Cell Line
A “good” cell line grows readily, is stable for the duration of the culture, is a high producer, and generates high-quality product. According to Eileen Skaletsky, Ph.D., CEO of QED Bioscience, there are three important characteristics of a good cell line:
1 Stability: The ability of cells to grow predictably over multiple passages, while producing consistently. Cells that don’t thrive in culture make poor candidates for long-term production.
2 Overall health or viability: How long will cells survive in culture? Some continue to reproduce indefinitely, while others cannot be kept alive for more than two weeks.
3 Productivity: How well do cells produce over the course of the culture? Production capacity is often negatively affected by limitations imposed by the client, for example on the use of animal component-derived ingredients such as serum. “Some cells adapt relatively easily to serum-free conditions, others don’t survive.”