Explosive growth in the area of therapeutic antibodies in recent years has stimulated new demand for improved production and manufacturing techniques. Traditionally, antibodies are produced in mammalian cells, a time- and labor-intensive process. This step immediately opens into a true bottleneck—the purification process, which typically involves at least three chromatography steps. In addition, antibodies present new challenges for analysis, characterization, and storage.
At several recent bioindustry conferences, novel technologies streamlining the manufacturing process for new biologics were discussed.
One of the most important advances in antibody production is the effort to adapt microorganisms to produce them. Since the 1980s, there have been many unsuccessful efforts to produce antibodies in yeast. Alder Biopharmaceuticals (www.alderbio.com) has implemented a plug-and-play yeast expression system that uses two proprietary vectors. When the desired nucleic acid sequence is inserted into the expression vector, the system will produce antibody at a rate of 300–1,000 mg per liter, according to the company. Per-unit time, the yeast system represents a three- to fivefold advantage over mammalian systems, which take from 12 to 20 days per cycle.
Alder has produced up to 2,000 L from a single batch, and in the hands of others the system has reportedly been ramped up to 50,000–100,000 L.
Nonmammalian antibody production comes with drawbacks. Yeast produces different carbohydrates than a mammalian cell. Some mammalian-produced antibodies are designed to take advantage of the glycosylation for the antibody dependent cell-mediated cytotoxicity or complement dependent cytotoxicity mechanism. The Alder system is incompatible with antibodies designed to use this mechanism. Instead, the engineered process results in antibodies that do not need to be glycosylated to attack their targets. Newer-generation antibody therapeutics are getting away from carbohydrate-dependent mechanisms due to side effects.
“We have a list of good targets to work on for our methodology,” said John Latham, Ph.D., CSO of Alder, who spoke in a session on microbial hosts for antibody production at BIO’s annual conference in May. “Alder provides antibody production services for corporate partners, and also has its own therapeutic pipeline, with its first product entering clinical trials in the fall of 2007 and an anti-TNF program in development for 2008.”
TNO Quality of Life (www.tno.nl) is another organization developing microbial antibody production. Peter Punt, Ph.D., project leader in the microbiology department, described filamentous fungi as “champions” in protein production, producing up to tens of grams per liter of protein. The fungi have been developed as an antibody production platform by TNO.
TNO, which is neither a public nor a private entity, was established by an act of law in the Netherlands to support the industrial needs of Dutch companies primarily, worldwide companies with Dutch interests secondarily, and all other companies on a discretionary basis. Its development of protein-production systems began with chymosin, integral for cheesemaking. Based on the company’s results with chymosin, it expanded into more specialized proteins such as interleukins and plasminogen activator. According to Dr. Punt, the output of the fungi is far superior to typical mammalian (CHO cell) systems.
Filamentous fungi also have some unique obstacles. They produce a lot of protease, so much of the TNO development effort has been directed toward preventing proteolytic degradation. Current applications of the filamentous fungi system include not only pharma and biotech, but hydrolytic enzymes for bioenergy.
“This is not only in Europe and the U.S. Developing countries are now entering into that area because of the enormous interest in bioenergy,” said Dr. Punt. Future directions for TNO include systems biology and metabolomics. “We recently merged with another department where process design and engineering design is the stronghold. We also think we can combine systems biology with process design.”
Large-scale purification of antibodies represents another area of growing innovation. GE Healthcare (www.gehealthcare.com) provides platforms and column resins to support antibody purification. The process typically requires at least three steps: protein A affinity purification, low pH inactivation of viruses, and anion or cation exchange chromatography. Noninfectious endogenous and adventitious viruses must be removed from the final product.
Using a platform system to integrate purification steps expedites the downstream process of antibody production. GE provides resins and filters and helps customers set up purification platforms.
“One of the things we’ve seen is that the cell culture upstream has been more productive than the downstream,” noted Gail Sofer, director of regulatory compliance for the fast-track group at GE Healthcare. Sofer gave a presentation on platform technologies for monoclonal antibody production at Strategic Research Institute’s “Antibody World Summit.”
Complaints about bottlenecking in the purification process are largely based on older, existing facilities (See Debunking Downstream Bottleneck Myth on page 62). Sofer spoke about upgrades such as increased column size, multiple trains, and more productive resins for improving efficiency in purification.
“Over the years, the chromatography resins have become more productive. We also have resins that have higher capacity,” she pointed out.
The advent of microbial hosts eliminates one major purification step—viral inactivation. However, there may be unique separations added to the process based on the host that is chosen. For example, when E. coli is used, endotoxins must be removed.
Analysis and Characterization
It can be difficult to analyze the output of a process for production of a protein therapeutic, especially a glycoprotein. To that end, Waters (www.waters.com) recently published a poster that describes the combination of ion mobility separation and time-of-flight mass spectrometry (IM-TOF MS) in the Synapt™ HDMS™, which is used to characterize mAbs.
The instrument is capable of resolving ions by size, shape, and charge prior to mass detection. Ion mobility is essentially gas-phase electrophoresis, according to Waters. A series of pulses of electrical waves propel the ions forward through a neutral gas. The gas molecules offer frictional resistance, further differentiating the ions by size and shape. As a single instrument platform, IM-TOF MS can be used to analyze both glycoproteins and glycopeptides, says a Waters scientist
“What’s unique about us is that we are doing it in parallel, getting information on all ions coming out of ion mobility cells at the same time,” said John Gebler, Ph.D., director of biopharmaceutical sciences at Waters.
“With traditional ion traps, you have to analyze them sequentially. Synapt separates all of them because of ion mobility. This is an extremely valuable capability for the analysis of therapeutic antibodies. Synapt can reveal whether the antibody matches what is predicted by the primary amino acid sequence. It can tease our heterogeneity in a batch of antibodies, glycan details, amino acid variants, C-terminal lysine truncation, and heavy and light chains.”
IM-TOF MS can also be applied to PEG classification. PEGylation is a process of attaching the polymer PEG (polyethylene glycol) to biomolecules most typically peptides, proteins, or antibodies, which can help to meet the challenges of improving the safety and efficiency of many therapeutics.
“However, the chemical nature of PEG makes characterization by conventional methods extremely challenging,” noted Petra Olivova, a senior researcher at Waters. “Synapt offers unique, enabling functionalities for characterization of PEG and its derivatives. Ion mobility time-of-flight MS is capable of characterizing and differentiating PEGs by their sizes and conformations in a quick infusion experiment (a few minutes).
“In addition, the impurities and low molecular mass species in PEG material can be promptly observed from 3D heat maps by simple eyeview.”
Future directions for Waters’ Synapt HDMS technology include new methods for detecting conformational differences using ion mobility separation and implementing more top-down analyses of the 3-D data set generated by the instrument. “We’re making progress on utilizing top-down in a more comprehensive characterization of intact proteins,” says Dr. Gebler.
The challenges of antibody therapeutics do not end when the product has been produced, purified, and analyzed. The fragile molecules must be prepared for storage and distribution. Up to 90% of antibody and protein therapeutics will be lyophilized.
Lyophilization is accomplished by first freezing the batch to -40 or -50ºC, then subliming the ice away in a primary drying stage and finally removing residual ice through secondary drying. Once frozen, the product never reenters the liquid phase.
Traditional methods of lyophilization involve cooling the drying chamber with a compressor-based refrigeration system. However, the mechanical components of these systems are pushed to the edges of their operating ranges at the extremely low temperatures used in lyophilization. In the case of a mechanical failure, the entire batch could be lost. For biologics, a single batch might be worth hundreds of thousands of dollars, so reliability becomes crucial.
Cryogenic freezing has become increasingly popular for these applications. Cryogenic freezing reduces the environmental impact by eliminating refrigerants and using liquid nitrogen instead. Praxair (www.praxair.com) provides lyophilization systems for many biotechnology companies.
“In Europe, all the majors have both kinds of systems. Anyone who has a protein therapeutic has one of each unit,” noted Balasz Hunek, Ph.D., senior manager of technology for Praxair, who spoke in detail on lyophilization at New York’s “Interphex 2007.”
New technological improvements to cryogenic drying systems focus on automation. “Larger systems will fill the formulation into vials without anyone touching them. The production is aseptic and sterile,” he said. Praxair has also made improvements to the freezing cycle, shifting from a random freezing process to a controlled nucleation.
The large-scale production of therapeutic antibodies requires many supportive technologies. The challenges associated with antibodies are unlike those of small molecule therapeutics. Many of the supportive technologies will require a complete retooling, but there is enough growth and enthusiasm in this market to justify it.