May 1, 2012 (Vol. 32, No. 9)
Angelo DePalma Ph.D. Writer GEN
Five presentations on protein purification at last month’s “Bioprocess International Europe Conference” in Prague provided clear and detailed insights on the latest thinking on how to best handle therapeutic biomolecules.
Günther Jagschies, Ph.D., strategic customer relations leader at GE Healthcare, reviewed the pluses and minuses of affinity chromatography for both monoclonal antibodies and “novel applications.” These include purification of kappa and lambda Fab fragments, coagulation factors VII and VIII, the nonpathogenic adeno-associated virus, and alfa-1 antitrypsin. Competition has heated up for protein A resins, with several small players introducing new mAb capture resins with binding capacities approaching 60 g/L with low residence times and high caustic stability.
Most experts agree, and actual practice validates, that affinity chromatography is the capture step of choice for mAbs due to its high selectivity, volume reduction, suitability for platform separations, and reduction of subsequent purification steps.
“But cost becomes an issue when the capacity and lifetime features of affinity resins are not fully utilized, such as during clinical manufacturing or for products requiring a limited number of batches per year,” said Dr. Jagschies.
Correctly sizing a capture column, particularly for legacy mAbs, is the most straightforward way to minimize cost issues, he added.
Ligand leakage and resin lifetime were considered issues in the past, but are rarely problems with today’s well-designed and -supported resins. Subsequent purification steps almost always clear ligands, which in any case do not lead to significant capacity losses. Lifetime has reached the low hundreds of cycles for hydroxide-stable ligands and the fifties for resins that are not as stable to CIP.
Although costs dominate discussions of affinity resin negatives, according to Dr. Jagschies, interest in protein A and other affinity-based separations has never been higher. This is somewhat of a contradiction since costs dominate, especially at lower scale. “I do not believe that cost will prevent manufacturers from using affinity capture as it is too powerful a method to miss out on,” he explained.
Efficient Workflow for the Production of Unstable IgM Antibodies
Xuemei He, Ph.D., a process chromatography expert at Bio-Rad Laboratories, discussed a new three-column workflow for purifying a monoclonal IgM antibody. Because the protein was unstable below pH 5 and prone to aggregation, cation exchange and hydrophobic interaction chromatography was problematic.
By taking advantage of the novel mixed mode medium’s salt tolerance, Dr. He was able to first remove DNA and host cell proteins by flow through chromatography over AG1-X8 resin followed directly by binding the antibody onto the novel mixed mode medium at high salt and close to neutral pH. The third step, polishing with a ceramic hydroxyapatite (CHT), gave active protein free of aggregates.
Purifying IgMs is challenging due to the molecules’ large size and instability. Several capture modes have been tried, including ion exchange, HIC, size exclusion, and affinity chromatography. IgMs often denature or precipitate under these conditions.
Although size-exclusion chromatography is gentle it does not exhibit production-scale volumetric productivity. Affinity chromatography media for IgM purification are currently available, but the molecules frequently do not survive the harsh elution conditions, resulting in low recovery.
IgMs tend to be more stable in high-salt buffers, which makes them unsuitable for conventional ion-exchange purification. Dilution to reduce feedstream conductivity for IgM binding on an ion-exchange column may lead to target molecule precipitation, while the higher process volumes are problems in their own right.
Dr. He’s approach works around IgM’s large molecular mass and heavy glycosylation, which lead to slow diffusion through conventional media. Her development-stage resin is based on rigid macroporous polymeric base matrices that allow efficient mass transfer. The precise chemistry involved is still under wraps. Dr. He did disclose that it contains “structural elements that interact with target molecules via one or more modes of interaction that include one or more electrostatic or hydrophobic interactions and hydrogen bonding.
“Under optimized conditions, this mixed-mode media offers unique selectivity unmatched by single-mode chromatography.”
Bio-Rad is currently working with collaborators to develop an intuitive approach to method development using this medium.
CHT is the ceramic form of the ages-old, naturally occurring mixed-mode chromatography material, hydroxyapatite or (Ca5(PO4)3OH)2. Electrostatic interactions between calcium and phosphate groups and the surface charges on biomolecules are what make CHT the “polishing workhorse” for removing process- and product-related impurities such as host cell proteins/DNAs, viruses, endotoxins, leached ligand from affinity chromatography, and antibody aggregates/fragments.
“Our method has general applicability to a potentially large variety of proteins, particularly where affinity capture is not a viable option,” explained Dr. He. But depending on the specific process and proteins the sequence of chromatographic steps and purification conditions may need to be adjusted to achieve the best separation.
“The final process based on this medium is simple, with minimal conditioning steps, yet offers products with high purity and biological activity,” said Dr. He.
According to Jörg Peters, Ph.D., of Bayer Pharma global drug discovery, Bayer is the biopharmaceutical company with the most experience in protein crystallization, and its work seems about to bear fruit. The difficulties in crystallizing therapeutic proteins aside, the technique has lagged behind chromatography because the unit operations—borrowed from chemicals and small molecule drugs—are foreign to bioprocessors.
Bayer’s process is described as an “integrated, disposable, and closely operated” piece of equipment that integrates harvest, washing, drying, sampling, and storage of product. Because it is contained in a disposable system, cleanroom use and cross-contamination are minimized.
Dr. Peters explained that, in principle, protein crystallization is applicable to any kind of protein, peptide, or crystallizable virus. “However in certain cases finding the correct conditions in multi-dimensional space is difficult.”
But by no means is this a platform technology. The additives and physico-chemical conditions employed in protein crystallization are in the public domain, but the precise conditions for development-stage projects remain trade secrets. The “filtration-drying-storage” device was developed at Bayer Healthcare Pharmaceuticals in cooperation with sister unit Bayer Technology Services.
Interestingly the technique is not suitable for the equivalent of a capture step since those involve large volumes and relatively low titers. No chance, then, that this can replace expensive affinity media. “It’s a post-capture step and has been shown by Bayer and other companies to fit into the purification sequence,” noted Dr. Peters, who added that Bayer plans to make the technology available for licensing and perhaps for “further co-development.”
Applying a Toolbox to a Platform
Platform approaches to production and purification have been popularized by the emergence of mAb therapeutics. Matteo Costioli, Ph.D., associate manager for downstream processing at Merck Serono, described what he calls a “toolbox approach” for mAb purification platforms.
The toolbox idea can be thought of as adding versatility to manufacturing platforms, and thereby reducing costs. An example is the three-column platform for mAbs. Rather than stick with expensive unit operations like protein A capture, Merck Serono process engineers evaluate older, trusted technologies or new ones with the idea of plugging-and-playing them when the need arises.
“The main goals of our approach are to reduce costs, shorten timelines, and increase process knowledge,” said Dr. Costioli.
Regardless, this is a major undertaking involving significant behind-the-scenes development and probably some in-process tweaking as well. “Toolboxing” requires first identifying different purification technologies, matching them to processes on molecules, and testing with different mAbs under different conditions.
When a similar situation arises down the road—a highly hydrophobic protein or one that tends to aggregate—Merck Serono reaches into its toolbox, pulls out the right operation, and applies it. The company is accumulating a database of techniques and conditions, which it hopes to apply to its future bioseparations projects.
“Everything will be tested, known to work, and suitable for including in the separation platform,” pointed out Dr. Costioli.
His Prague presentation specifically focused on a process that involves precipitating impurities in-line with the bioreactor, then capturing on a cation exchange column, or capturing and clarifying in one step using a novel expanded bed adsorption technology.
Originally developed by Upfront Chromatography and now exclusively owned by DSM Biologics, the so-called second generation expanded bed adsorber has applications in food and feed industries as well as pharmaceuticals. Its main claim to fame is reducing two unit operations (harvest, capture) to one at high yield and throughput.
So why haven’t biotech companies been knocking down Dr. Costioli’s door to get a piece of these cost-saving technologies? “There’s always a balance between the effort you have to put in at the beginning, to change a technology, and the benefit at the end,” he said.
Sylvio Bengio, Ph.D., scientific communications manager for bioproduction at Pall Life Sciences, rounded out the talks on bioprocess separations with a case study on mixed-mode (or multimode) chromatography for mAb purification.
We tend to think of mixed-mode resins as novel but they are as old as column chromatography itself. In 1903 Mikhail Tsvet, then a newly minted professor at Warsaw University, discovered that hydroxyapatite could separate plant pigments. Tsvet did not realize it at the time, but the mineral’s ability to separate chlorophyll from carotenoids was based on a mixed-mode effect.
Tsvet’s last name, coincidentally, means “color” in Russian, the same word from which “chromatography” was derived. Tsvet’s discovery lay dormant for several decades, and he died without official recognition of his invention.
Pall introduced the first mixed-mode chromatography resin in 2000, and today numerous vendors sell mixed-mode resins combining nearly every combination of adsorption mode. Hydrophobic interaction plus some type of ion exchange predominate.
In a bioprocess setting, mixed-mode replaces mostly hydrophobic interaction chromatography. They provide the benefit of not requiring binding-promoting salts, at concentrations of up to 3 M, required to run hydrophobic interaction columns. “Salts are a problem at such high concentrations, and recycling them is a problem,” explained Dr. Bengio.
In addition to intermediate-step chromatography, mixed-mode ligands are useful in nonantibody purification as a low- or no-salt alternative to hydrophobic interaction chromatography. Once optimized, mixed mode’s resolving power is quite high. “We’ve published on a separation of two proteins differeing only by a methionine group,” noted Dr. Bengio.
The method has gained steam with the advent of high-throughput screening methods and design of experiment-based screening for both chromatography media.