January 15, 2017 (Vol. 37, No. 2)
Angelo DePalma Ph.D. Writer GEN
Downstream Operations Can Relieve the Pressure of Higher Titers and Cell Densities without Undue Disruption
Platform manufacturing keeps driving the industrialization of bioprocessing, both upstream and downstream. Upstream operations, risk averse as ever, are using the new platform approaches to do the usual: push raw production. And it must be said that downstream operations are hardly more daring. Yes, they sometimes seek innovation—but always prudently.
Reasonableness is evident in downstream’s approach to “process intensification,” a term that has generally been applied to processes where two or more steps or unit operations are combined. An alternative definition, however, is suggested by Peter Levison, Ph.D., senior marketing director for downstream processing at Pall. He says that process intensification encompasses not only combined operations, but also any operational innovations that enhance upstream efficiencies and increase cell densities and product titers.
Process intensification of this sort, warns Dr. Levison, can be something of a mixed blessing. “With higher titers and cell densities come more product, but also higher levels of impurities,” he points out. “Process intensification can pressure clarification and subsequent purification steps.”
As the pressures imposed by process intensification build, downstream operations may become so motivated to seek relief that they may consider radical solutions. For example, to keep up with the inflow of protein and impurities, downstream operations may start to favor precipitation and extraction strategies.
More commonly, however, downstream operations will prefer less disruptive options, should any be available. Ideally, such options would exploit currently accepted purification modalities.
Safety and regulatory risk have traditionally been arguments for evolutionary, rather than revolutionary, bioprocess modifications. Accordingly, over the last year and a half, Pall has concentrated on wringing additional benefit from familiar technologies.
“Continuous processing, for example, allows us to use the same buffers, skid principles, adsorbents, and membranes as in a batch process, but in a slightly different manner,” notes Dr. Levison. “It mitigates the discomfort that sometimes accompanies the choice between established and experimental technologies.”
Although Dr. Levison recognizes that continuous processing may seem merely practical in new operations, he does not trivialize the risks for established operations. “It will take a while,” he says, “for continuous processing to be adopted in existing processes and facilities.”
Pall works to accomplish smooth transitions to continuous downstream processing. For example, the company has developed the Cadence product family, which includes the BioSMB Process System, a multicolumn chromatography system that features a disposable flow path and is scalable to GMP manufacturing. The BioSMB Process System has been designed to facilitate conversion of processes from batch mode to continuous chromatography while maintaining sorbent, buffers, and product quality assays.
Utilization the Key?
For Jonathan Royce, bioprocess senior product manager at GE Healthcare Life Sciences, the prime issue facing downstream bioprocessing is maximizing utilization and productivity of existing manufacturing resources. “This issue spans the spectrum from small-scale through commercial manufacturing,” he insists. “Maximizing facility utilization offers the best return on investment.”
Optimizing utilization is about doing more with less, and its recognition in bioprocessing indicates industry-wide maturation and industrialization. This, in turn, often leads to second- and third-generation production technologies, an example of which is continuous downstream processing.
Utilization begins with process development but its true implementation occurs at pilot and production scale, which is where the benefits occur. Some have argued that the success of mammalian cell culture constrains the degree to which purification may “intensify.”
“Capacity mismatch is often exaggerated,” Royce asserts. “In the majority of cases, upstream and downstream output can be matched through good process development and a close working relationship between personnel in process development and their colleagues in facility design and manufacturing.”
Royce acknowledges that upstream titers have risen, which has led in some cases to higher masses for downstream purification, but in others the result is volumetric reductions upstream. In practice, to maintain product quality, many manufacturers limit bioreactor titer to about 5–10 g/L. These concentrations are not difficult to manage with modern downstream technology. “After all,” Royce observes, “the mass requirements for clinical trials and/or commercialization haven’t changed.”
Little Things Count
To Jim Powell, senior business development manager at Asahi Kasei Bioprocess America (AKBA), a disproportionate amount of downstream buzz is being generated by a handful of technologies: dilution-based automated buffer preparation systems, automated continuous processing systems, and single-use (SU) technologies for clinical or multiproduct facility production.
“AKBA inline buffer dilution (IBD) systems feature a patented technology, APAT, that makes buffer production from concentrates easy and precise while reducing buffer tank size requirements,” says Powell. “With competition from biosimilars ramping up, bioprocessors seek to squeeze the last drop of efficiency from existing processes, particularly toward product recovery.”
IBD prepares precise buffers from untitrated buffer concentrate, even if the buffer stock concentrate is made with ±10% accuracy. The high accuracy allows manufacturers to narrow their buffer specifications to increase product recovery. That can potentially pay for the system in relatively few batches. IBD also eliminates bottlenecks by eliminating the QA buffer hold as it records buffer quality information continuously.
AKBA has invested in new software technologies to address other aspects of buffer production, including the company’s ReagentTracker technology, which tracks raw materials to final product via barcode scanning. Powell confirms that IBD is already deployed in commercial and clinical applications, even supporting automated continuous processing technologies such as simulated moving bed (SMB) purification. No commercial proteins have yet been approved using SMB, but several biopharmaceutical companies have invested in it and other continuous process technologies.
Tried and True
Monoclonal antibodies (mAbs) currently occupy 7 of the top 10 spots on the list of highest grossing therapeutics. The current crop of 200 or so development-stage mAbs assure their therapeutic popularity, if not dominance, for decades.
“The great thing about mAbs is their manufacture lends itself to platform processing, which includes protein A capture,” says Oliver Hardick, Ph.D., CEO of Puridify. Yet competition from antibody fragments, biosimilars, and production within emerging markets threatens to disrupt the notion of platforming.
Hundredfold mAb culture titers have provided comparable productivity enhancements, but downstream operations have not kept pace, hence the much-discussed (and perhaps hyped) upstream/downstream “capacity mismatch.” Simply increasing chromatography column size is a strategy constrained by time, space, and the high cost of protein A resins. Thus, for given feed and protein concentration values, capture becomes the downstream bottleneck.
Various workarounds improve the economics of protein A utilization and operation. Puridify, for example, has developed a cellulosic nanofiber matrix for immobilizing protein A. Binding capacity compared with beads is not improved, but nanofiber-based columns run significantly faster, translating to a volume-based throughput upgrade of around 100, thus easily matching recent upstream productivity enhancements.
The matrix/adsorbent, FibroSelect, offers flexibility in column size and/or process time. “Bioprocessors can opt to run a standard-size column in a fraction of the time of a conventional protein A capture step,” asserts Dr. Hardick, or they can rapidly cycle a much smaller column in the same elapsed time as a traditional capture.”
A New Spin
In July 2016, Sartorius Stedim Biotech acquired kSep Systems, a specialist in advanced centrifugation for the purification of recombinant proteins, cell therapy products, and vaccines. When announcing the purchase, Sartorius Stedim Biotech described the kSep system as “providing a gentle processing environment” for cell separation.
Fritjof Linz, Ph.D., vice president of purification at the company, explained why the system’s gentleness was significant: “With conventional centrifugation, if you want to get rid of cells, you must not only sacrifice product recoveries with increasing cell density (since solids are discarded as slurry) but also sacrifice product quality from high amounts of host cell protein resulting from high shear.”
Independent of cell density, kSep centrifuges provide high recovery by completely recovering product from interstitial space between the cells prior to discharge. In addition, low shear force ensures lower amounts of host cell protein contamination.
Depth filtration and centrifugation go in and out of relative favor as cell removal techniques. The choice often comes down to scale and the specific application—whether retained cells are simply removed (and discarded) or reused.
Standard depth filters have very large surface areas, are not economical, and are limited in processing capability to about 40 million cells/mL, which is lower than top-density mammalian cultures today by a factor of five.
Bowl centrifuges are normally constrained by operating volume, which once reached require stopping the process, removing the cells and supernatant, and starting over. With continuous centrifugation, the cell mass can be pumped out as it forms so that productivity is limited by volumetric throughput and not centrifuge volume.
This kind of centrifugation is especially beneficial when cells are to be resuspended, or when cells themselves are the product. The cells remain fully viable after removal or harvest—a welcome bonus!
kSep facilitates cell-based therapies, which in fact is the principal market for kSep. Similarly, semicontinuous virus production from mammalian cells occurs through clean removal of the cells.
High-density cell culture for antibody production is another kSep application that Sartorius Stedim Biotech is eager to promote. The completely closed centrifugation technology supports up to 150 million cells/mL, and all product contact surfaces are single-use.
Thermo Fisher Scientific has been working on remedying the shortcomings of centrifugation at production scale, particularly for adeno-associated virus (AAV). For example, purifying AAV by density gradient ultracentrifugation can last several days, which limits large-scale application.
Kevin Tolley, senior field application scientist for purification at Thermo Fisher, notes that centrifuges capable of handling large volumes of AAV do not exist. Regardless, he is not inclined to despair, not in light of findings obtained by a group of investigators at ViroTek. The group demonstrated purification of AAV using a research-scale continuous flow centrifuge from Alfa Wassermann. Virus recovery amounted to 50%, suggesting—but not demonstrating—that large-scale continuous centrifugation may be feasible.
Similarly, AAV purification has been achieved using affinity, ion exchange, hydrophobic interaction, and size-exclusion chromatography, although none of these techniques has gained widespread adoption.
With respect to current large-scale purification strategies using chromatography, Thermo Fisher worked with customers to develop affinity resins specific for AAV serotypes. The company’s CaptureSelect technology comprises single domain antibody fragments, or ligands, immobilized on a chromatography backbone. These ligands can be directed to highly specific surface proteins on the respective viral capsid to achieve high purity with a scalable solution.
Thermo Fisher currently has two products available for AAV purification: POROS CaptureSelect AAV8 and AAV9 resins, which have high specificity toward AAV8 and AAV9 serotypes, respectively. “We’re beginning to see interest in AAV8 and AAV9 as gene therapy agents,” says Tolley. “These products have a very high specificity and capacity for those particular serotypes.”
New (Old) Resins for New Molecules
This column has covered immobilized metal affinity chromatography (IMAC) in the past as both a workable research-level protein purification modality, and for non-mAb recombinant proteins. IMAC resins contain immobilized electron acceptor metals that attract polyhistidine-tagged proteins or natural proteins containing suitable electron donor groups such as cysteine and histidine.
Their major drawback is requiring, more often than not, the introduction of polyhistidine tags through recombinant DNA technology, and a separate step to remove the tag proteolytically. From a process intensification perspective then, IMAC is a step backward.
Another minus for IMAC is the possibility that therapeutic proteins could be sensitive to nickel leaching from the columns. And yet another knock on IMAC stems from biopharma’s longstanding appreciation for custom, experience, and regulatory discretion: Bioprocessors strongly prefer tried-and-true affinity ligands instead of those that may work better in certain instances or cost less.
“Protein A has been around so long, and used so successfully,” says Xuemei He, Ph.D., R&D manager for chromatography media chemistry at Bio-Rad, that it remains attractive to bioprocessors despite its high cost.
Bio-Rad’s NUVIA IMAC resin uses nitrilotriacetic acid to chelate di- or trivalent metal ions such as Ni2+ or other transition metals, such as Zn2+ or Cu2+. Complexed ions bind to histidine tags on either the N- or C-terminus of the protein to be purified.
Vaccines, factor 8, and viral vectors are among the products that have been manufactured at large scale using IMAC. However, given the prevailing market and regulatory climate, and the fact that mAb manufacturers prefer platform purification methods, protein A becomes less relevant with the introduction of next-generation therapeutic molecules such as antibody fragments, bispecific antibodies, and antibody-drug conjugates.
“So many variations exist today,” remarks Dr. He, “that it has been impossible to come up with an affinity ligand for everyone’s protein.”