February 15, 2006 (Vol. 26, No. 4)
Meeting the Demands of Product Manufacture by Moving One Step at a Time
Antibody processing is a complex field, demanding an understanding of bioengineering, equipment design, cell genetics, and cell culture technology. As demands for large-scale antibody production levels become more intense, companies are pursuing these disciplines from a variety of perspectives.
Many companies already have infrastructure and purification technologies up and running, so the latest improvements in antibody purification technology are the result of optimizing through incremental changes, rather than by the introduction of radical new technologies.
Purification with Hydroxyapatite
Crude preparations, whether generated from serum, bacterial cells, or culture media, are a complex brew of salts, ancillary proteins, endotoxins, and aggregates, which present special challenges to the protein purification specialist. Affinity chromatography with staphylococcal Protein A is widely used for both commercial and research levels of purification.
The widespread use of protein A as a purification ligand has driven much research into its molecular structure as companies seek to improve its performance. Sanchayita Ghose, Ph.D., and his collaborators at Amgen (www.amgen.com), have investigated elution pH differences during Protein A chromatography among several IgG1s, IgG2s, antibody fragments, and Fc-fusion molecules.
Variable region interactions determine elution pH for various antibody subclasses with traditional protein A chromatographic materials. The Amgen group have shown that a chopped down, engineered Protein A will respond to a single elution pH for a range of antibodies, mitigating problems associated with low pH induced aggregation.
The strong interaction between protein A and members of the IgG subclass ensures a powerful and specific binding, but elution of the antibody molecules can present difficulties. Protein A binds to the Fc region at the juncture of the Cd2 and Cd3 domains through hydrophobic interactions, hydrogen bonding, and salt bridges.
At low pH complementary histidine groups facing one another take on positive charges, resulting in an electrostatic repulsion. While this force is powerful enough to allow elution from a Protein A-packed column, these conditions can be harsh and cause aggregate formation and denaturation of the antibody molecules. Circumventing this molecular challenge has driven a number of independent companies to develop strategies before and after elution from Protein A.
Scientists at Bio-Rad Laboratories (www.bio-rad.com) have optimized the use of ceramic hydroxyapatite, according to Pete Gagnon, process applications and new technologies R&D manager, process chromatography division of the company. Gagnon adapted this material for antibody purification. Ceramic calcium hydroxyapatite, Ca10(PO4)6(OH)2, can be especially useful in removing aggregate contaminants. Protein aggregates are chronic problems in the case of proteins expressed in E. coli. Formed in inclusions bodies, they can prove to be highly persistent in bacterial protein expression systems.
Calcium participates in metal affinity interactions, while the phosphate portion is active in cation exchange exclusion reactions, Gagnon explains. CHT ceramic hydroxyapatite has the added quality of being stable down to pH 6.5 in the presence of 5 mM phosphate.
CHT ceramic hydroxyapatite fractionation of contaminants on linear sodium chloride gradients allowed separation of Protein A-purified human IgG1 from both large and small aggregates, as well as DNA, endotoxins, and lipopolysaccharides.
Aggregate removal was very efficient, greater than 99%. Leached protein Aremoval was greater than 90%, while DNA removal was reduced by three logs, and endotoxins was brought down by greater than four logs. Gagnon and his colleagues optimized the process for scale-up by adjusting the slope and amplitude of the salt gradient.
When CHT ceramic hydroxyapatite columns are eluted with a sodium chloride gradient at a low concentration of phosphate, we simultaneously achieve reductions in the levels of aggregates, leached protein A, DNA, and endotoxin, says Gagnon. The method is easily integrated into a two step platform with Protein A, or with an additional step to exploit alternative fractionation mechanisms.
Automated Mab Purification
Researchers at GE Healthcare (www. gehealthcare.com) developed several new options for multistep purification of Mabs. These include AKTAxpressMab and HiTrapMabSelectSuRe. These products simplify purification and screening through automation of the more tedious procedures. AKTAxpressMab (the name, KTA, comes from the Swedish word meaning true or real) is a modular system with specialized software for processing Mabs.
According to GE, it was designed to automate the purification, regeneration, and cleaning processes to allow convenient handling of multiple antibody samples.
The standard assembly consists of two tandem modules that may be operated independently with different purification protocols or alternatively, in series using a single protocol. The instruments that are robust enough to be run in a cold room can be expanded to as many as 12 units, linked together, controlled by a single computer. After a first affinity purification step using Protein G or A, the system can be programmed to run either desalting or size-exclusion fractionation.
The configuration includes a built-in, fixed-wavelength UV monitor for online monitoring of real-time conditions. The monitor allows the use of special, preprogrammed steps, so that following elution the protein peaks are collected in capillary loops.
The software package, referred to as Unicorn, which controls the separation process in AKTAxpress, is also used for AKTAexplorer (milligram amounts), as well as AKTApilot and AKTAprocess (gram and kilogram amounts).
This integrated system allows for scale-up, which helps avoid the failures and pitfalls that often occur when the transition to larger-scale is attempted by inserting different control devices into a small-scale purification scheme.
GE Healthcare combines the automated purification instrument with a prepacked column, HiTrap MabSelectSuRe (superior resistance). The medium is a rigid, high-flow agarose matrix bound to an alkali-stabilized, Protein A-related ligand. The material was designed for greater stability than conventional Protein A-based media, allowing numerous cleaning cycles with sodium hydroxide, an important consideration when multiple antibodies are being purified.
It was developed using protein engineering techniques, in which a number of alkali-sensitive asparagine residues were replaced in the B domain of Protein A, and a new ligand was created as a tetramer of four identical modified domains.
MabSelectSuRe has been developed for large-scale Mab production where cleaning with NaOH is an industry preference, according to Eric Grund, Ph.D., director of fast track biopharma research services at GEs Uppsala facility. Adequate cleaning between cycles or batches of feedstock affects process robustness, product safety, and quality issues. This product is also useful when applied to high-throughput purification of Mabs at the lab bench in HiTrap product.
We may have gotten more than we hoped for, states Grund. Not only do we have a product resistant to caustic cleaning, but there is also promise of milder elution conditions and decreased product aggregation.
GE Healthcare is continuing with its efforts to improve Mab purification processes. A whole process is no better than its weakest link, so by improving the steps before and after protein A capture we can obtain ameliorated overall performance, says Henrik Ihre, Ph.D., product manager, also of the GE Uppsala facility. After MabSelect, further purification is usually by ion exchange and modifications in this area are being integrated into revised protocols that will be introduced later in 2006.
At Human Genome Sciences (www. humangenomesciences.com) several antibodies have progressed to the stage of clinical development for treatments of autoimmune diseases (LymphoStat-B), cancer (HGS-ETR1, HGS-ETR2, and HGS-TR2J), and infectious diseases (ABthrax, CCR5 mAb004).
Yuling Li, senior director, purification sciencesprocess development, discussed the companys antibody purification strategies and the modifications required to meet a broad range of challenges.
The company designed a number of different approaches to generate human antibodies against targets derived from its genomics discovery program. LymphoStat-B is a Mab directed against B lymphocyte stimulator (BLyS). This factor is required for the maturation of B lymphocytes into antibody-producing plasma cells. Overexpression of BLyS may contribute to the production of autoantibodies that attack and destroy the bodys own tissue in various autoimmune diseases.
In 2005 HGS announced the results of two Phase II studies with LymphoStat-BTM. The first study involved 283 patients with rheumatoid arthritis who failed other therapies. The antibody was well tolerated with no serious adverse reactions and showed statistically significant improvements in primary endpoints.
The second study involved 449 patients with active systemic lupus erythematosus. Although the primary efficacy endpoints were not met, a statistically significant clinical effect was observed in seropositive patients, representing 75% of the patient population in the study. With these results there was a strong incentive to scale-up antibody production.
Li reviewed the strategies at various stages of downstream process development for clinical products. We developed and implemented a platform process for the purification of multiple monoclonal antibodies, she says.
The platform purification process employs multiple steps and conditions optimized over a wide range. According to Li, it was designed to ensure the capability to conform to a variety of Mabs and yield high purity products with ample recovery.
The platform development approach enabled the company to better focus resources on key areas to develop in-depth process knowledge, shorten process development cycle, simplify raw material inventory, and streamline documentation and technical transfer to manufacturing. Experience and knowledge gained from each product feeds back to refine the platform process. Later-stage process development focuses on optimizing conditions for product specific attributes and suitability for commercial manufacturing facilities.
For mammalian cell-derived products, viral safety is an important consideration. Use of the platform purification approach provides comparable unit operation steps between products. Information obtained from viral clearance studies from the same step with different products can be used to build a database for future matrix viral clearance validation.
We have found that the best way to optimize our procedures for large-scale antibody production is to move step-wise from T-flasks to 15-L reactors and then to 5070-L reactors. Beyond 70 L, there is very little difference in performance, says Robert Toso, Ph.D., director of process development at Baxter Laboratories (www.baxterlaboratories.com).
Baxter is primarily contract manufacturing, so it is frequently involved in projects that move from the small-scale, preclinical to midlevel, Phase I to larger volumes of greater than 100 L for Phase II investigations.
There are a number of devices available today for efficient production of mammalian cells in large volumes. Dr. Toso and his colleagues opted for a conventional bioreactor, fitted with an inner plastic liner. This permits interchangeability in as short a period as 30 minutes, facilitating changeover of the reactor to handle different cell lines.
Moreover, a conventional bioreactor behaves as a stir reactor with a rotating internal paddle, a configuration that Dr. Toso feels is superior to the back and forth sloshing performance of some methods.
The main disadvantage that Toso sees in initiating a protocol with 15 L, rather than a much smaller volume, is the extra cost of materials consumed during the pilot optimization phase. However, since the upscaling from 15 L proceeds with greater ease, the tradeoff more than compensates for the ease of moving production to the next level.
The early period of process development is the time to determine media constitution and other decisions affecting growth parameters. We have found that its important to resolve your performance issues early on, before you move to larger volume runs, Dr. Toso says.
One of our most significant issues is that of stability. So if youre making subtle changes in media and growth conditions, you may improve your production rates, but with an unacceptable loss in stability of the product. We decided to employ a wide variety of analytical techniques, which turned out to be a fortunate game plan
These include multiangle light scattering and other approaches, allowing the assessment of performance, and a close monitoring of antibody quality.
Engineering Improved Cell Lines
Selexis (www.selexis.com) develops tools and technologies for boosting recombinant protein production in eukaryotic cells.
Improving protein expression in cell lines is one of the routes to boosting productivity with resultant savings of cost and time. Traditional routes for augmenting protein synthesis rates have been haphazard and time-consuming, often based on random testing of many clonal isolates and the use of drug selection markers to drive the overexpressing phenotypes.
Hundreds of cells must be tested to identify one or two with the desired characteristics. Such cell lines may show only marginal increases and these modest gains are often lost on subsequent proliferation.
Other approaches include the isolation and integration into vectors of sequences with high transcriptional activity that can also be tedious and time-consuming. Selexis solution to this quandary is a marriage of basic molecular biology with a practical outcome.
According to Igor Fisch, Ph.D., CEO of Selexis, various chromosomal elements could increase gene expression when appropriately inserted into specific regions of the cells chromosomes. The Selexis investigators screened various possible sequences, including boundary elements, matrix attachment regions, and locus control regions for their ability to increase the expression of reporter genes in CHO cells.
Of those assayed, only the matrix attachment elements (MAR) were effective. MARs are sections of the genome that anchor the chromatin to the nuclear matrix proteins during interphase. S/MAR is the designation given to the DNA sequences that integrate into the chromosome of the host cell in such a fashion that they define boundaries of the genetic region in which they reside.
If a transgene is inserted next to the S/MAR element, the complex may form an independent region of the chromosome and undergo overexpression. The exact biochemical mechanism causing this overexpression is not yet fully understood, but is believed to be brought about by chemical-physical interactions, as well as regulatory proteins that are pulled into the complex.
According to Selexis, the S/MAR element will maintain high-level expression in protein production systems and is proving to be a versatile vehicle for high-level protein production.
For monoclonal antibodies, we routinely obtained over 50 pg per cell per day of protein production, stated Dr. Fisch. But more importantly, our experience has shown high-expression yields without need for gene amplification in a wide range of cell lines. Moreover the process is rapid, as we can generate high-productivity clones in less than eight weeks. The system also displays stability of the productivity in suspension cultures grown in fully synthetic media.
The Bottom Line
By picking away at the hide of the elephant, antibody production scientists have moved forward on all fronts. Cell culture densities have been increased by a factor of 30 in the past 15 years, and cellular productivity has increased 10 fold. Yields of protein of more than 1 g/L are common, and there are claims of even higher levels. While downstream processing may still lag behind, previously insurmountable challenges, posed by the high density of material that must be partitioned and separated are being addressed, and progress is evident.
All these developments are encouraging for the pharmaceutical industry in general. It is likely that the majority of new drugs that will be introduced in the foreseeable future will be biologics, many of which are targeted at chronic diseases that require repeated doses of the therapeutic agent. Technological gains that can make these treatments more widespread and affordable will be welcomed by the health care system.